MOLE ULAR H RA
BIND! G PROT

RIZ TIO OF TH
ODI
R
I T RA TIO WITH KRA 0

ION D T RMIN NT0
NE MRNA

by

eba tian J. Ma k d en ki
. c.

TH

111

r it

f

rthem

r1 ti h

lumbia, 20 12

I
BMITT D IN PARTIAL F LFILLM
THE REQ IREM TS FOR TH D RE
I
E
M TEROF
IN
BI HEMI TRY

UNIVERSITY OF N RTHERN BRIT! H
July 20 15

© eba tian J. Mack den ki, 2015

T

OLUMBIA

F

Abstract
Th

- P) pla

ding r gi n d t nninant binding pr t in (

r gulating gen

pr

RD - P binding t

R c nt
mutat dan or

m cane r

r, and tran lati nal repr

traffi

n thr ugh In

a r le in

m

- apr

nc gen frequently

d in an er, leading t up -regul ati n.

mRN

interacti n ther [! r

ent an v 1 therapeuti

i the aim of thi th

lectr ph retic m bilit and immun precipitati n experiment

pre ent d h r u mg

-mutan t

pp rtunity t

RD- P ariant

-dri en cancer and

h w D ur K-hom 1 gy dom ain are

required to bind a 57-nucleotide minimum binding eq u nee in the coding region ofKRA
RNA· the e mutati n d not c mpromi e pr teinD lding a e idenced by circul ar di chr i m
spectro copy. DNA-ba ed anti- en e oligonu cleotid inhibitor were devel ped ba ed on the
'minimun1 binding

vitro. Finally a

quence that ucce sfull y abroga te the

RD-BP-KRA mRNA intera ction in

RD-BP KHl toKH4 construct purified usin g a nove l "o n-co lumn "-re[i !ding

scheme suggests a pos ible role for RRM-domains a negati ve regulators of RNA-binding.

II

Table of Content
b tra t. ..................................................................................................................................· · · · ·· · 1
nt nt .................................................................................................................. ........ .11

i t f abl
Li t f

igur

v

··································································································································

······························································································································VII

kn w l dg rn nt ....................................................................................................................... X
·····································································································································XI

hapter 1 - Introduction
pre

n rn cancer. .................................... ........................................ I

1.1

lmp rian e f g n

1.2

P

1. 3

RN -bindin 0 pr tein and mR

t tran

1

ripti nal g n r gu lati n and m RN
tu rn

tability .................................................... 2

er ..................................................................... .. .

1.4

bri cf hi t ry f

1.5

RD -BP in earl y de elopment and carc in gene i ...... ................................ ....................... 6

1.6

VICKZ protein famil y tructure and function ....... .......... .... .............. .. ............................... .

1. 7

neogeni c KRA

RD -BP ................................... ......... ............... ....................................... 5

gene ex pre ion and

RD-BP .. ......................................... ........... ....... 12

1. 8

Targeting KRA in KRA -driven cancer ........... ....... .. ... ..... ........... .. ............................... 14

1.9

R e earch goal ......... ..... ............ ... ... .... .. ......... .. ....... ........ .......... ......................................... 19

Chapter 2 -In vitro characterization of KRA mRNA bindin g to CRD-BP and
development of antisense oligonucleotide inhibitor .
2.1

M eth dology - Mapping f
2.1.1
2. 1.2
2.1. 3
2 .1.4
2.1.5
2 . 1.6

RD-BP binding ite

n KRA

mRN ............................ 2 1

Tran formati n of BL2 1(D ) cell with p T 28 b- RD -BP p ia mid .................. 2 1
Indu cti nand rowth o f p T 28 b- RD-BP .c li ell ....................................... 22
Purification & quantifi ca tion f r c mbinant RD -BP pro tein ..... ........... ............ 22
ub-region cDN templ ate .................................. 24
en ration f KRA mRN
enerati n of [32P] -labe ll d KRA

RN

M A mapping of R -BP binding ite

ub trate by 1

............................... -6

n KR

.......................... ·····- 7

mR

Ill

2.2

M th d 1 gy - Anti n
lig nucl tid inhibit r ........................................................ 28
li gonu 1 tid ............. ...... 28
2.2.1
M
mp titi n a a
ith p cifi anti- n

2.

11 ............................................................................................. ......... 0
R ult and
m
ub-r gi n by I T. .................................................. 0
ynth
2. .1
RD-B affinity mapping f
2. .2
M
lig nucl tid inhibit r ............................. .45
pm nt f p ific anti n
2.
D
2 . .4 KRA RNAn1inimum binding qu n anal i ................................................. 53

C hapter 3 - A e in g th e rol e of
in vitro a nd in c IJ

RD-BP KH doma in in olved in KRA

mRN

binding

3.1

M th d 1 gy- RD-BP KH d main requirement D r binding KRA mR A in vitro ... 57
3.1.1 Purification f RD-BP KH d main ariant ....................................................... 57
ub trate by IVT .......................................... 58
3 .1.2
nerati n f [ 2P]-labelled RN
e perim nt with RD-BP KH d main variant ....................................... 88
3.1.3 EM

3.2

Methodology - CRD-BP KH domain requirements for KRA mRNA binding in cells .. 59
3.2.1 Preparation f :Unmune precipitated RNA from HeLa cell .................................. 59
3.2.2 cD A ynthesi fr m IP-RNA ............................................................................. 60
3.2.3 KRA mRNA RT-qPCR ..................................................... ............ ..................... 60

3.3

Results and Discu ion ....................................................................................................................... 62
3.3.1 CRD-BP ingle KH domain va1iant EMSA binding a ay ................................. 62
3.3.2 CRD-BP double KH domain mutant EMSA binding assays .. .............................. 66
3.3.3 KH domain requirements for KRAS mRNA binding in HeLa cells ..................... 69

C hapter 4 -Assessin g th e structural impa ct on GXXG muta tion s in KH d omains on CRDBP, conformation al cha n ge up on RNA binding, a nd KH1to4 crysta llization
4.1

Methodology- Multiple structural analyses of the CRD-BP protein-RNA interaction .... 75
4.1.1 Production of CRD-BP KH variants .. .. .. .. .............................................................. 75
4.1.2 Circular dichToism spectra .... .. ... .... ....... .. .... ........................................................... 76
4.1.3 CRD-BP secondary structure e timation ............................................................... 77
4.1.4 CD monitoring f CRD-BP confonnation upon RNA titration ............................ 77
4.1.5 Generation of 18S RNA ........................................................................................ 79
4.1.6 Thennal stability of RD-BP -RNA complexe .................................................... 79
4.1.7 Denaturing purification ofCRD-BP KH1to4 ........................................................ 0
4.1.8 Native purification ofCRD-BP KHlto4 ............................................................... 1
4.1.9 Denaturing on-column refolding purification of RD-BP KH l to4 ...................... 2

IV

4.1.1 0 Fluor c nee larizati n anal .i f KH 1t 4 binding .............................. .. ........ 4
ntru tin fp T-41 c( )- RD- PK lt 4plamid .............. ... .. ..... ..... .. .. .... .. 5
4.1.11
4.2

R ult and Di cu i n ..... ... ............................................ ....... .... .... ............. ..... .... .. ........... 91
4 .2.1
ir ular dichr i m anal i [
- P K mut nt ariant ..... ...... ........ .... ... .. .. .91
4 .2.2
ircular dichr i m m nit ring of
binding in itr ....... ... ...... ... .... 94
Th tmal tabilit f R - P-RN
mpl
........... ............................... 104
4 .2.
4 .2.4
RD-BP K.Hl t 4 pr t in cry tal. ........................................................................ 10

Chapter 5- General Di cu ion
5. 1

Pr j ct gen ral

5.2

R f KRA
Mapping f the RD-BP bindin g it on th e c din g reg i n and 3'
mRN u ing th le troph retic m bili ty hift a ay ............... .......................... ... ....... .. 116

5.3

T ting the pecifi c anti- en e lig nucl eotid e inhibit r again t the RD-BP-KRA
rnRNA interaction in itro u ing EM
..... ....... .... ..... ... ..... .... ........ ........ ......... .......... .. ... ! 19

5.4

Examining

5.5

IP-qRT PCR analy i ofCRD-BP KH domain requirements for KRAS 1nRNA
binding in HeLa cell ... .... ...... ...... .. .............. .. ........ .......... .... .... .. ............... ............. ... ... .. .. 123

5.6

Structural analysis of CRD -BP KH variant containing point mutation in the
GXXG motif at the KH domain ........... ... .. .. .. ... ... ... ........ ..... .. ......... ....... .... ....................... 124

5.7

Monitoring ligand-induced conformational change in full-l ength CRD-BP and the
KH 1to4 truncation variant. ............ ........... .... ....... .... .. .. ... .... ....... ...... ................................ 126

5.8

RRM1 -RRM2 didomain as a negative regulator of RNA-binding ... ....... ..... ................... l 29

5.9

Assessing CRD-BP-RNA complex stability u ing then11al shift analysi ....................... 129

5.10

CRD-BP KH 1t 4 protein cry tal.. .... ..... ..... ......... ..... ..... .. .. .... ... .. ..................................... 1 1

5.11

Summary & concluding remarks .. .... .. .. ....... ........ .. .... .. .. ..... ...... ....................................... l3 3

rvi w ....................... ...................................... ............ ..... ............... .. ... 11 5

RD-BP KH domain requirement for binding KRA mRNA in vitro .. .... . 12 1

v

List of Tables
r t inmutati n tudi

Tabl 1.1:
Tabl 2.1:

mJ

1t1 n f buff! r u

d in d naturing purificati n f r c mbinant

RD -BP pr tein fr m tran £1 1m d
Tabl 2.2:

omp

Tabl 2.3:

R -

)

-F £1 r in itr tran
r u

.c li c ll ............................... 22
rag

f purifi d

r t in ................................................................................ 2

P R prim r u ed fl r th gen rati n f K
t mplate

Tabl 2.4:

-21 (

iti n f buf.G r u ed fl r r £1 !ding nd

r c mbinant

..................................... 12

mRN

ub-r gi n c N

ription ................................................................. 24

d £1 r th e g n rati n f K

m

ub-r gi n c

A

- ( 1-4) £1 r in itr tran cripti n ... .............................................. 25
Table 2.5:

P R primer u ed .G r the gen rati n f KRA m
template KRA -

ub-r gi n c NA

4(a-f) fl r in v itro tran cription ..... ......................................... 25

Table 2 .6 :

IVT reagents used for KRA mRNA ub-regi n RN

Table 2 .7 :

EM A reagent u ed for KRA -CRD-BP affinity mapping and

synthe i ........ ................. 26

screening of anti en e oligonucleotide inhibitor ..... ............. ................... ............. 27
Table 2.8:

DNA sequence (5' to 3 ')of candidate inhibition anti sen se oligo
nucleotides used in EMSA inhibition assays ........... .......... ................. ................... 28

Table 2.9:

Concentration of antisense oligo nucleotides (AON ) used in EM A
inhibition as ays .......... .. .............. ..... .. .. .. ... .... ...... ........ ...... ... .................................. 29

Table 3.1:

Smrunary of generated CRD-BP KH domain variant ................ .......................... 5

Table 3 .2:

KRAS RT-qPCR amplification primers and TaqMan probe ................................. 6 1

Table 3.3:

Reagents used for KRAS qPCR reaction .. ... ..... ................................................... 62

Table 4 .1:

Mea urement parameter for CD sp ctra of RD-BP KH variants ...................... 76

Table 4.2 :

Buffers u ed in native purification of RD-BP KHlto4 ....................................... 2

VI

Tabl 4. :

Buf~ r

Tabl 4.4 :

P R r ag nt u

Tabl 4.5:

Re tricti n di ge t r ag nt

~ r K.H 1to4

Tabl 4.6:

Reagent u d fl r

KH 1t 4 pi a mid li gati n ........................... ........ ....... 90

Tabl 4.7:

Reag nt

~ r c

Table 4 .

Pr t in

c nd ary tructure

Table 5.1:

u d in d natming n-

I KZ mRN

lumn r fl !ding pr t in purifi

d t g n rate KH 1t 4-p T -41 (-

ti n .. ................ 4

) pla mid

ub -cl ning ............................................. 9

lon P R ere ning ............................... .. ....................... .............. 91

targ t iz

timati n fl r

RD - P vari ant ... ........... ............... 94

........................................................ .......... ........... ..... 11 8

VII

List of Figures
Figur 2.1 :

Pla mid map f p T 2 b + -

Figur 2.2 :

raphical r
( th.r ugh

- P ................................................................... 2 1

ntir c ding
.G r K

ub -r gi n lab !led
qu nee nd partial '

ub-r gi n

TR .............. 0

thr ugh F .................................. 0

Figur 2. :

I T

Figur 2.4:

M
a
ing
din g qu cnce RN
u -regi n
t
for
RD-BP binding ......................................................................... ......................... 3 1

Figur 2.5:

M

a
ing KR
RR
ub -regi n
t
for RD- P
binding .................................................................................................................. 32

Figur 2.6:

Kd plot for KRA mRN

Figure 2.7:

EM

Figure 2.8:

aturation binding curve for

Figure 2.9:

Graphical repre entation of KRAS -A(l through 4) RNAs ................................... 36

Figure 2.10:

KRAS ub-region eDNA template for in vitro tran cription ...... ........... ......... ..... 36

Figure 2.11:

EMSA assessing KRAS -A 3' truncated sub-regions A 13 A 12 and A 1 for
CRD-BP binding ......................... .. ......................................................................... 37

Figure 2.12:

EMSA assessing KRAS -A 5" truncated sub-regions A24, A34 and A4 for
CRD-BP binding ................................................. .. .............. .... ......... ...................... 3 8

Figure 2.13 :

EMSA assessing RNAs KRAS-Al2 and KRAS-A34 for CRD-BP binding ......... 39

Figure 2.14:

Saturation binding curve for CRD-BP vs KRA - 34 RNA ............................... .40

Figure 2.15 :

EMSA asse sing KRA -A3 RNA for CRD-BP binding ...................................... .40

Figure 2.16:

Graphical representation ofKRA -A34(a-f) RNAs ............................................. .42

Figure 2.17 :

EM A assessing the KRAS -A3 4(d, e and f) RNA , for

t mpl at

aturation binding

ub-region

, D and F ............................................. 34

p riment between

RD-BP and KRA -A .............. 34

RD -BP vs KRA -A InRNA ................................. 35

RD-BP binding .......... .43

VI II

- P binding ........................... .44

Figur 2.1

M

igure 2.1

M

Figure 2.20:

Bar graph f anti n e lig nu 1 tid m 1 ule M( 1- inhibit ry effect n
c mpl
n ............................................................ 4 7
RD- P-

Figur 2.21 :

M

a

mg
f anti

hara t rizing

4(a band

II r

n e li g nu 1 tid

- M(l - ) inhibit r ................. .46

compl

inhibiti n by

...............................................................................................................48

Figur 2.22 :
A

c mpl x inhibiti n by
charact rizing
- P-K
- M7 .............................................................................................................. 49

Figur 2.23:

Inhibiti n pl t of
M6 ( ) and M7 (B) again t RD-BP-KRA RN
mple fl nnation ........................................................................................ 50

Figure 2.24:

omputer imulated RN folding of KRA -A RN with KRA -A3 4(b) and
N inhibitor M6 and M7 target region indicated ........................................ 52

Figure 2.25:

Local DN alignment ofKRAS-A34 and putative binding equence for
IMP-1 and ZBPl CRD-BP orthologs ..................... ....... ............. ........................... 56

Figure 3.1:

EMSA a se ing CRD-BP KH variant with single point mutation for binding
to sub-region KRAS-A RNA; (A) KH l and KH2 variants, (B) KH3 and KH4
variants ....... ............................. ..... ................. .. ... ...... ...... .... ............................. ....... 64
EMSA assessing CRD-BP KH variants with single point mutation for binding
to KRAS-A and c-myc RNAs ; (A) KHl variant, (B) KH2 variant,
(C) KH3 variant, (D) KH4 variant. ........ .. ... ........................................................... 67

Figure 3.2:

Figure 3.3:

EMSA assessing the CRD-BP KH variant with two point mutation for binding
to sub-region K.RAS-A RNA; (A) KH1-2 , K.Hl-3 (B) KHl -4, KH2-3
(C) KH2-4 a11d KH3-4 ............ ... .... ..... .... .... .... ....................................................... 68

igure 3.4:

Box plot pf the RT -qPCR cycle threshold values from immune-precipitated
KRAS mRNA as ociated with CRD-BP KH variants ........................................... 70

Figure 3.5:

Relative KRAS mRNA associated with KH domain variant from immuneprecipitation of RD-BP in He La cells a determined by RIP R T -qPCR ............. 7 1

IX

Figur 4.1:

Diagram f n-

1gur 4.2:

p

lmnn r £ lding KH 1t 4 pr t in purificati n che1n ............... .
T pla mid multiple cl ning it

41 c(+) - n

Figur 4. :

p ctra f

ild -typ

Figure 4.4 :

p ctra f

R - P titr t d

Figur 4.5:

MRE at 222 run f

Figur 4.6:
Figure 4 .7 :
Figure 4 . :
Figure 4.9:

- P and KH mutant ariant ................................. .
ith ari u

titrated

D pe tra f KH It 4 titrated w ith RN

ith R

ub trat ....................... 99

ub trate ........................................... 101

MRE at 222 run f KHlt 4 titrated with RN
pl t

ub tra te ........................................ 96
ub trate .................... ............... 8

RD- P titrat d with
- P ampl

D-HT

qu nee ............................... 6

ub trate ............................. .. .... 102

f KHl to4 ampl e titrat d with RN

ub trat .. ..................... ] 03

Figure 4.10:

TM of KH1to4

RD -BPproteinin compl exwithRNA ub trate ................... 105

Figure 4.11 :

Plasmid pET41 c(+)-CRD-BP KH1to4 ... ........................................................... .. 108

Figure 4.12:

Comparison of native and denatu1ing KHl to4 protein purification ................... ! 09

Figure 4.13 :

CRD-BP KH1 to4 construct binding activity and protein cry tal.. ............. .......... 111

Figure 4.14:

Thermal aggregation buffer optimization for CRD-BP KH1 to4 ......................... 11 4

X

Acknowledgements
I

uld lik t thank fir t nd :fl r m

pportunit t w rk in cane r r

tim

earch .

r

tm

u

ed in a

i ntifi c nv1r nm ent.

a i tanc and

ften m ightful c n er ati n I a kn

tark a w ell a

eni t an R n b rg and Mark

I would al

like to

Young :fl r pr

tend m

thank t

r.

.G r gi ing me th

h lp d m gr w a an

p ent in hi lab h

a ad mic and all wed m t pa1ii ipat in m ea ningfu l r
kill r quir d t

h

r.

'

arch pr j ct w hil e dev J ping th
r th much apprecia t d techni cal
ndr a

agg1

n ell , Ma1iha

arne :fl r preparing the IP -

:fl r thi pr j e t.

r.

r . Jan

tephen R ader,

rent Munay, and

iding my recormnend at i n for admi i n int th e M . c . Bi ch emi try pr gram

at UNB , without whom thi e c ll ent e pen ence w uld not have been p
family and clo e fri nd - Bany, Lauri e, Kri

Li a, Brahm,

gratitude for their upport throu gh the rough tim

lex and

ibl . inall y my

orbin, de erve m y utmo t

, and company during the good .

1

Chapter 1
Introduction
llular r p n
regularl y e p

th

t

d t d p nd h a il

pr CI

c ntr 1

of

erting influenc

borneo ta i t bl

g1 al c nditi n that rga m m are

ide an

nth appr priat r gu lati n f g n
m

nth

ari u bi

ithin th cell i

n . uch

y tem f igna lling m I cule cap abl

h m i al p athway in

d cl tting t indu ti n f

vari u protein 1 el

c mpl

pr

d in cell acti vitie . rom

und h aling the ability t

el ctiv ly m dulate

ritical [! r appr priate re p n e t

timuli and optima]

functioning. Fir t d fm d in th lat 1950 , the central d gm a of m olec ul ar bi logy state that
D A i transcrib ed into an RN

m

nger m olecul e (mRNA) and ub equ entl y translated into

a tring of amino acids w ith defined tructur known a protein. Much work ha focu ed on the
fme details of tllis proce

and tho e efforts have delivered an ever-growing body of know ledge

regarding the preci e workings of how genetic sequences tored in D A tum into the full y
functioning mol ecul ar m achine and building blocks that cell ar e largely m ade of. As
astonishing as this proce s of protein production m ay be, it i not enough to imply have protein
being made in the cell ; to avoid diseases like can cer, they must also be regulated .

1.1 Importance of gene expression in cancer
Cancer is a particularly devastating disease among a host of other , and i there ult of
gen etic instability fo stering the deregul ation of bioch emical pathway effecting cell gro wth, cell
death (apoptosis), proliferation, repli cati ve inm1ortality, angio gen e i , and m etastasi (re iew ed
in Hanahan and W einberg, 2011) .

enetic mutation affecting tran c1iption or altering th

functioning of important nzyme in terms of reaction rat can lead to change in pathw ay
activity. When thes altered pathw ays affect ell cycl event , notab le abnon11aliti e can ari e

2

that impa t th h alth and
of 1nelan ma

2010 .
a

ignalling m 1 cui

ntain mutati n in th

ind u ti n f the
many gen

11-b ing fan afflict d indi idual. F r

P-Kina

in luding

bn nnal

iated with hyp

-

pr
ic di

receptor ignalling in ol

path

. hi lead t

hi h in tum au

nhan

a cular end th lial gr

rd r

nd n

din th un-g

ern d bl

n tituti

d tran ripti n and

pr

n f

a 1

and

amu 1 ,

)gn

1

th fact r (

ignalling, and p tentiat
d

imat ly 40%

-Raf r ulting in

chang

f th

gem

1nple appr

ftn

height n d c 11

el fo1mati n r quir d by aggre 1ve

olid tum ur (Ferrara 2010 . Th can nical epithelial -me nchymal tran iti n ( MT) i a key
tep in the initiation f carcin ma meta ta i and can re ult from the ct pic inducti n and
chronic o ere pr

n f e era! tran cription fac t r in luding Zeb 1/2 and

wi t a reviewed

by Yang and Weinberg (2008 . Gen expre ion i influenced by many cellular proc ses and
between tran ctiption and the final protein product po t-transcriptional regulation mechani m
provides a major source of gene control.
1.2 Post transcriptional ge ne regulation and mRNA stability
Control of cell activities, stability and even the complex tring of events that guide
embryo development i dependent on mRNA encoded information the trength of th ese signal
affected by steady-state mRNA levels- the balance between tnRNA production and destru ction
(Brewer 2003) . Initial transcription rates, RNA plicing, RNA transp011, and mRNA degradation
all play major roles in determining steady-state mRNA levels and in pm1icular the modulation of
the mo lecu le's longev ity, or half-life , within the cell is proving to be increasingly important
(reviewed in Filipowicz eta!., 2008). Evidence for the importance of tabilizing/destabilizing
effects on steady state mRNA level com from global eDNA array experiment comparing
stress-induced transcription rate and steady- tate mRNA level

f many gene , a well a kin tic

3

tudi

a £1 und t

de a . tabiliz ti n f m

fmR

man g n gr up und r an u

tr

r du

min d mamm ali an

g n e pr
mR

half-li£1

ng

act r af£1 ting mR

it in th

ha b n a rapid!

mg pr

), it ha b c m

tight! regulating mRN

2) . riti ca ll y, it ha b n

are n t a

in the

ll are di

e . With th di

I ar that p

ef a /. 1

, and ur und r tanding f th
ry

f

interil r nee by ire

t-tran cripti nall e el regul atJ n 1 a c r mea n

A tability. i -element are intrin ic

i t thr ugh ut th I ngth fa tran cript, £1 und in th e c ding regi n , 5'

and the 3' TR, that ar a

5).

11. In gen ral a dynamic c mbinati n f ci -element and

tran acting fa ct r pr du e the ari ati n in m
qu nee that

ll ( an ' fa/ . 2

th g ne e pre i n f

tabilit can dramati all alt r g n

n mall chang

and Mel! ( 19

hang

nhan

cia ted

ith mR

(in tability, with th e latt r being the m

c mmon region of de tabilizing i -element in the form f

R,
t

-ri ch equ ence element , r

ARE ( haw et a/. 19 6) . i -el ment can ex i t ingly or together with ther regulatory
element on a ingle tran cript and collecti vely provid e rapid , adaptive re p n

to a number of

different cell signal ( hen et a!. 199 ). Ci -element often act in concert with tran -acting
factors to achi eve their effect, a group con i ting of prot in and non-coding nucleic ac id , uch
as tibonuclea e , RNA-binding protein (RPB ) and mi croRNA capable of recognizing the
mature RNA tran cript and modulating it phy iological urvivability (Neugarbauer et al. 2002).
1.3 RNA bindin g protein s and mRNA turnover

Mechani m of gene regulati n are multipl e, and the pecific interplay b tween r gu lating
factors increa e in complexity in eukaryotic cell (Hayne 1999).

ntrol of gene expre ion by

mRNA regulati n begin irmnedi at ly foil wing tran cription in the nucl u , and ' n e p 11

f

4

int the yt pla m a het r g n u nu l ar rib nucl

f prot in call d RN -binding pr t in ( r

unp rtant la
50 kDa ann t fr

1 diffu

regulat

p rt are r qu ired

pore compl

(D nig

t a/. 2

and m dulat d

nRNP) patii 1 r qmr
r 200 ).

00) .

pr

an

prot in larger than

th nu lear m mbran , pecific nuclear

and are of particular int r t b cau
equen

pr t in

p rt ignal that

P :D r rec gniti nand tran it tl11· ugh the nuclear
pia an in trum ntaJ r 1 in R
a an adapt r betw

n mRN

metab li m ,
regulat ry

ion f th g ne . RBP ar a large f mily of pr t in and

hundred e i t (mo tly un hara t riz d) within the human gen m . Together with all the REPbinding ite within the tran criptome, the e RPB form an imp rtant tier of gene regulation
ba don equ nee el

ti it and p cific affinity

r gu latory cod "(Ke n 2007) .

v ral type ofRN

llecti ely t rmed the' po ttran criptional
binding protein have been identified and

each function in different way by identifying uniqu ci -elem nt , binding different mRNA
regions and altering tran lation rate or tabilizing/de tabilizing the mRNA transcript. P27Kipl
for example is a regulator of cyclin during G l phase, and its activity is down-regulated by the
association of its 5'UTR mRNA with AU-rich binding protein , HuR and HuD . Such physical
interaction restricts ribosome access to the tran cript and therein reduce tran lation (Yeh eta!.,
2008). BRF-1 is another RNA binding protein which functions not by regulating tran lation,
instead achieving its effect by destabilizing AU-rich mRNA target tran cripts (interleukin-3 ,
interleukin-6 and TNF-alpha) through deadenylation (Raineri eta/., 2004) . RBP canaL o erve
to stabilize mRNAs and increase relative gene expression, a phenomenon that when kept incheck serves to appropriately regulate various cell events. Enoneous over-expres ion of RBPs
however can bolster multipl e downstremn mRNA targets simultaneously, leading to global cell

5

m 1 ular pr fil

hift impli at d in th d

dy fun ti n, and

n anc r ( in t a/., 200

1.4 A brief hi tory of

t 1min nt

prot in that wa fir t hara t riz d aft r
imately 70

equ nc

9 %identity
VglRBP

a that

f c-myc mRN

protein fmnil (al

4·

b d

a t aI., 2 0 1 1) .

RD-BP

ding R gi n tability

appr

an t a I., 2

-cr

inding Pr t in ( R - P) i an RN -binding
linking e p eriment re

1 cti el bind and tabiliz
B ern t in eta!. 1 92) .

al d an

P of

th c ding r gi n d t nninant

R - P i a m mber f th "VI KZ"

kn wn a zipc de-binding pr tein ), a m all g r up f hi ghly con erved (77 -

it and

tandmi, 2002) orth I g u RBP with imilar functi n that include

RA (X nopu la e1•i ), IMP-1, 2,

(Homo apien ) and ZBP 1 ( Gallu

H om o api n )

RDBP (Mus musculus ), K

allu dom e ti u ) (revi ewed in Yi raeli 2005) . There is

however no e tablished tandard for the naming, and th de ignati n are ft n used
interchangeably depending on the laboratory. For example, Patel

t a!., (20 11) refer

to ZBP 1 a

IMP1 , with IGF2BP1 (another name for CRD-BP) li ted in bracket . Member ofthi family
have been identified as important po t-transcriptional mRNA regulators with roles in mRNA
stability, as well as translational control and mRNA localization involved in lamellipodia
formation and bleb extensions in motile cell types (Bern tein et al. 1992 ; Condeeli and Singer,
2005 ; Poincloix eta!., 2011) . VICKZ proteins are therefore each multifunctional and ha ve been
further implicated in early developmental stages where they are involved in critical events uch
as oogenesis, synaptogenesis, cell motility, and embryo polarization (Boylan et a!., 2008).
Of particular intere tis CRD-BP ' ability regulate gene e pres ion by affecting the half- li fe
of mRNA targets which influences protein level . Demon trating the potency of CRD-BP ' s
effect on steady-state n1RNA levels, an experiment utilizing tran genic mi ce over-e pre m g

6

RD-BP r p rt d a dramatic l 0-:fl ld incr a

2, h

prot in. i1nilarly B n1 t in

pr t in '

in I

11-fr

ffe t n target mR

le

I and path a

ntributi n t di a

i n t yet fu ll

pictur and it

and in ol em nt in

rtain path a

tabli h d targ t f th

1 v 1 wer

d that -m yc mRN

RD-BP in a

mRN

, an

II n

d ca a a . h re are num r u
ign I ling,

th

tabiliz d by at
amp!

f thi

it fi t int an verall

d.

are b ginning t r

al th imp rtance f

P.

1.5 CRD-BP in ea rl development a nd carcin o en e i
RD-BP i n rm all y und r tri t pati t mp ral c ntr 1 with
tage

pre i n p aking between th

n mbry rue day 12.5 in mice. In th day leading up to birth,

zygote and arl

mbr

CRD-BP becom

und tectabl or 1

pre

d at

ry 1 w le el in all r d nt ti ue (Han en et

al. 2004). in1ilarly, adult human ti ue h w li ttl IM P- 1 xpre ion comp ared to the robu t
level found in variou embryo tage (Leed et al. 1997 ·

iel en et al. 1999; L et a!. 20 12) .

Ilnportantly, CRD-BP expre ion in rats does not re-em erge during liver ti ue regeneration
(partial hepatectomy) agreeing with the standing theory that CRD -BP i norm all y limited to
developmental stage only and not ju t ti ue grow th/regeneration (Leed eta!. 1997) . Also
consi stent with the notion that CRD-BP i an important player in the earliest life stage in
mammals, CRD-BP-defi cient mice generated by gene trap insertion show severe abnormalitie
such as dwarfism, improper gut fonnation and higher perinatal mortality rates (Hansen eta!. ,
2004). On the other hand, over-expression ofCRD-BP in a mouse mod el all owed for nonnal
development, but produced populations with a high incidence of mamm ary adenocarcinomas
(95 % occurrence compared to no tumours in the non-transgenic mice) demonstrating a potent
proto-oncogene feature of CRD-BP (Tes i r et a!. 2004).

7

with

f anc r. Ind

v ra l typ

1 at d

RD- P IMP- I in 1 t r liD

f

R -e pr

pre

1

n f

1 a!. 200

f n n- mall

). In imilar tudi

en more in riminatin g

iti

trongl a

iated

% f human pri1nar br a t car in rn a , with - 1

%a

h r th

half f n ur

R -BP g n i 1 cat d

pith li al (brain tum ur , nearl y a third

lllung ar in rna , 7 % f m ali gn nt m

cancer w r fl und t te t p

ur and i

al d

a r ult of g ne duplicati ninth 17q2 1 band r gi n
(I annidi

an

a1nining r a t anc r pati ent tum ur r

d, a tud

-BP in

tag

fl r RD - P I annidi

en h yma l tum ur and 69% f varian
t a/. 200 1, 200

id n e r gardin g thi pr tin · mv

that found 81 % (1 7 out f 2 1) f t t d c 1 rectal tumour w r p

; K bel I a/. 2007).

m nt in n opla ia i a tud y
itive fl r ignificant

RD-BP

expre ion, whil e normal colon ti ue and infl ammatory b w 1 di ea ed ti ue howed no
detectable level (Ro

et al. 200 1). Rea on fo r thi re- xpre i n are mo tl y unkn wn howev r

recent tudy of CRD-BP in m elanom a cell cultured und er hypox ic conditi n that mimi c the
tumour micro environment ha

hown CRD -BP modulati on by th e HIF J a (hypox ia indu ced

factor 1 alpha) tran cription factor through acetylation of the CRD -BP gene. MT cell
proliferation assays in the same paper also uncovered a strong link between CRD -BP and the
melanoma cell' s pro liferative capacity, as sh-RNA ( h011 hairpin RNA) directed against

RD-

BP was ab le to abrogate the cancer 's growth (Crai g et al. 20 12) . Interestin g ly, HIF I a indu ction
is al o seen in other cancers where CRD-BP appears to pla y a role, such as breast and colorectal
tumours (Kieth et al. 201 2). Hypoxia-independent induction ha also been observed by Noubissi
et al. (2006) , where the researchers demonstrated that CRD-BP can be indu ced by en hanced ~­

catenin/Tcf signalling, a common feature in many colorectal cancer (Morin eta/. 1997).
The clinical value of RD-BP is evid ent in prognostic stati stics, and in general, human CRDBP expression level correlate with the dia gnosed grad of m alignancy for evera l cancer

8

Yani and Yi ra li 2

2 .

r du

d r CUIT n -fr

and

ingl

ariat and multi

riat

anan can

tr ngl

r pati nt

ith

dditi nall

n !at

it ha b n [! und that

rall u i al rat

m

it i un lear

anan cancer.

far b ed n
[! r

heth r a tu I pr gn ti

(K b 1 t a/. 2 07) . In

ith r e r h requirin g larg r

mple iz

h re

t r va l nt, p iti e te ting c n elated

R -

n1 m

nd ffcr a m lecul ar fact r 111 pr gn

erall aggr

(Dimitriadi

1 a/. 2

7 . Whil

th r , it i

lear that

R -BP i g nerall an imp rtant ingr di ent D r carcin g ne 1

me can er

it preci e ontributi n and m d
how preci el it intera

ith

lth ugh

facti n app ear t b c mple , kn wledge fit target and

ith th

progno tic and e en th rapeutic

m rc ignifi cant than in

ill all w cienti t t take aim at RD- P in a
n e.

nD rtunately th re i much gr und yet t tread in thi

realm, but h w thi RBP d e it j b i pr gr

ming t li ght.

1.6 VICKZ protein famil y tructure and function

I KZ protein have been h wn t bind multipl e target including the mRNA fi r c-myc,
IGF-II, beta-actin, Hl9, tau and other (reviewed in Yi raeli 2005). How th ey accompli h thi
through the multiple RNA-binding domain present in the e protein . RD-BP and other VI KZ
family member are composed of six RNA-binding domain ; 4 K-homology domain (KHl -4)
with a ~a.a~~a. topology, and 2 RNA recognition motif (RRM 1-2) with a 0 a.~ ~ a.~ topology that
make up the majority of th e protein'

tructure . Hi gh eq uence con ervation f the VI KZ

family RNA-binding domain , as well a the linker region between KH l -KH 2 and KH -KH4,
ugge t that the domain may function together a did main (KH 12 and KH34) (Git and
Standart, 2002). lnd ed, tudie

f the ZBP 1 hom log how that linker region amino acid ar

critical [i r proper protein function . In thi e periment, a trun a ted KH 4 ZBP 1 con tru twa

9

ynthe iz d and the link r armn a id j ining Z P 1 KH and KH4 p udo-d mam w r
replac d with th

e fr m th adjac nt KHl - KH2 link r r gi n.

a id wa chang d, maintaining th

ralll ngth f th link r.

id ntitie al n produc d a mark d nega ti
d tennin d thr ugh

M

af[i ct n th

nl the id ntity [ th amm
hang in the amin a id

b rved Kd D r zip

d binding a

, indi ating th imp rtant r le the e link r r gi n play in RN

binding ( hao t a!. 2010 . h

i 1 gical ignificance

a d termin d thr ugh ell-ba ed

experimentati n wh r a

FP-fu d ZBP 1 linker mutant full length pr t in c ntaining th

linker mutation)

ed in mou e embryonic fibr bla t t a

a

pr

am

any functional impact.

When analyzed by confocal micr copy the linker mutant n 1 nger r tained th e wild-type ubcellular localization and wa found t with th

in-b tween amin acid

erving a more than

ju t a hinge ( hao eta /. 201 0) .
Part of RD-BP' function i to bind and localize mR

within the cell. How it

accompli she thi task and the question of which, if any, individual domains are primarily
re ponsible for binding has been an wered with varying results by tudying CRD-BP and its
clo ely related orthologs. Granular RNP formation as ists in localizing vmiou mRNA
tran cripts within the cell, and systematic deletion of the individual domains of IMP 1 revealed
nuclear export signals (NES) in KH2 and KH4 domains (Nielsen et al. 2004) . To addre s the
question of domain redundancy in target-binding, Nielsen et al. (2002) de igned an experiment
using GFP-tagged deletion constructs of IMP 1 that had the RRM and KH domains
systematically removed, and subsequently analyzed for presence of IMP ! -containing RNP
granule fonnation, sub-cellular localization and mRNA binding capacity. It was found that in
terms of cytoplasmic trafficking, loss of both RRM domains had little impact on subcytopla mic
localization, with all four KH domain proving nece ary and uffi cient for nonnal IMP-1

10

btain d with regard t RN -binding

di tributi nand granular app aranc . imilar r ult
to kn wn targ t Hl

1 tr ph r tic m bilit

£ r f£ cti

h r th

mRN

binding

hift a a

an u d l ti n
d

M

n t a/. 2 02 .

t bind it targ t b ta-a tin mR
unabl t bind

Z

er a ayed u ing
r all £ und t b r quired

1 - th £ unding memb r fthe

p riment

nly r quir KH -4 d mam

ith the th r ~ ur d main (RRMl-2 and Kl 1-2 con truct)

ith an appreciable Kd ( ha

Jf

a/. 20 10) .

identical b ha i ur in am - ub trat binding a a
demon trat that

d main 1-4

n er el

I KZ pr tein fami1 , wa £ und through imilar

n truct

I KZ prot in (including

target ba i (VI KZ protein truncati n tudi

e pit high imilarity and therwi e

, the different d main r quirem nt

RD-BP) may interact differently nan individual
ummarized in Table 1.1).

Further in ight into the tructural detail of VI KZ family pr tein and h w they bind
RNA wa addre ed by mean of X-ray cry tallography.

tudying ZBPl in complex with beta-

actin mRNA, it was found that binding of RNA wa mediated by interaction with a minimum 29
nt RNA fragment in 2 spots with both the KH3 and KH4 domain arranged in an anti-parallel
fashion such that the RNA binding urface ran in oppo ite directions (revealed by a IMP-1
KH3-KH4 crystal structure). This didomain configuration was found to induce a unique 180°
looping of the RNA strand with two non-sequential, spaced RNA sequences bound. It i thought
that this is likely the in vivo confonnation and mechanism of binding for VICKZ proteins
because the interaction of the KH domains is stabilized by sequestering a large surface area and
several hydrophobic residues present in all VICKZ homologs (Chao eta/. 201 0). Adding another
level of complexity, surface plasmon resonance and NMR experiments have demonstrated that
the KH3-4 didomain contains a dimerization motif. Weak protein-protein interaction between
the separate KH-3-4 didomains was observed and this interaction is stabilized by the pre ence of

1

RN

1 ading t alt n1ati

c mp

iti n ( it

ibiliti

tan dart 2002 ·

analy i and ad an ed at mic D rc mi r

m

I KZ RNP

t a!. 2000) . In fur1h r upp rt f thi n ti n, kin ti
p

fl

1 ugg t p

int ra ti n in thi m d 1

characterized by a highl tran i nt int rm diat

iti

c

p rati ity with

n t a/. 20 4; J n

multipl IMP- 1 m 1 ul e binding t a ingl mR
2007). Pr t in-RN

and granul

ur

ith th fir t binding

f 1 w tabili ty, [! 11 w d by a

n t a!.

ent
c nd IMP 1

binding. Prot in-pr t in int ra ti n ia a KH -4 dimeriza ti n m ti f th n furth er tabilize the
RNP comple , l nding
target mRN

aero

pl anati n t the hyp r- tabl e

th c 11

P granul e re

n ibl e D r tran p rting

i 1 n eta /. 2004 ). Littl info rmation r garding the actual

tran port of the e protecti e RNP c mpl e

e i t . H wever, immun - taining and con[! cal

microscopy r v aled that IMP-I doe co-localize with mi cr tubul e and f-ac tin , u gge ting
cytoskel eton a ociation with

TP -dependent movem ent ( iel en eta!. 2002).

Despite the numer us studi e di playing tight a ociation with whittl ed-down target
mRNA regions, and even X -ray stru ctural model

no con en us binding equence fo r VI KZ

proteins as a family or individually, has been identified with only case- p ecific gen eralizations
having been made that do not fully account for observed binding. For example, CRD-BP
interaction with c-myc mRNA has b een shown to predominantly occur in the region
corresponding to nucleotid es 1763 to 177 7 (5 '-AGCCACAGCAUACAU-3 '), but it i un known
which nucleotides comprise the actual binding motif (Coulis et al. 1999). Even between
experiments using identical proteins in similar conditions, different conclusions have been
drawn. Jonsen et al. (2007) , conclud ed that ZBPl has a high affinit y for guano ine-1ich and
cytosine-poor RNA regions, and putati ve ly identifi ed a 5' -CCYHH
U, H = A or

-3' bin d in g motif (Y=C or

or U) . U sing a different experimental approach, other re earchers later c n luded

12

that th ZBP l c n n u bindin g m tif
tran ript (Hafn r t a/., 2010 .

n nn hm nt f '-

a

I KZ pr t in R

tran cript are till1arg 1 t ntati

,h

afl r m nti n d " R
rna

ay

-binding r quir m nt ~ r mRN

r Pat 1 r al. (2011 , d m n trat d a tri t binding

r quir m nt fl r ju t th Z P 1 KH 4 did mam,
qu nc , and th KH d m 111 re

- ' r peal in th targ t

ith the KH4 d main r c gnizin g a 5 '-

. .
gmz mg

equ n e, in line with th

-1 ping" m d I (Pat I eta /. 2 11 ·

ha ct a/. 2010 . Tb

e di cr pancie

m thing ab ut th m th d u ed, r it ma b that the di f~ r nt id entifi d equ ence

tern fr m diffl r nt KH d main pr fl r n

. It ma al

be that the

multi-RN -binding-

domain pr tein ar targ ting a c rnm n thre -dimen iona1 tru ture a opp

ed t ju t a linear

equence.

Ta bl e 1.1.

ummary of mention ed VI CKZ protein mutation stu di e
Type of S tudy

Conclusion

R ef.

Cell-based,
GFP-tagged
In vitro EMSA

NES pre ent in KH2 &
KH4
Increa ed zipcode RNA
binding Kd .
Reduced granulation
and cytopla rnic sublocalization
KH domains 1-4 all
required for effective
binding ofH- egment
(173 nt fragment) H 19
RNA
KH 3-4 didomain
impm1ant in binding
beta-actin mRNA (54
nt binding region)

Niel en et al.
2004
Chao et al. 2010

CRD-BP
Ortholog
IMPl

Type of
Mutation
Deletion construct

ZBPl

KH3-4 linker

ZBPl

KH3 -4 linker

Cell-based,
GFP-tagged

IMPl

Deletion construct

In vitro, EMSA
Cell-based,
GFP-tagged

ZBPl

Deletion construct

In Vitro, EMSA

Chao et al. 2010

Nielsen et a/.
2002

Chao et al. 2010

13

ne

1.7 Onco enic KRA

pre ion and

r gulati n f riti al path
ar

f an

t th

r.Mn

f an

eta/. 2

nn e

J

5 .

nd ar

f p c ifi c
kah a hi ' I a/.

R path way

mm n t c ch

th e

1

a

n titu ent

ignalling pr t in all d th Kir t n rat arc rn a iral nc gene h m 1 g (K

Abn rmaliti

1 al. 200

).

ignall ing m le ul ar implica t d in 0% f pancrea ti ca ncer , 60%

in thi

color ctal anc r ( R ), and 50%

. Ma cau

h w up -rcgul ati n

including th Wn b ta- at nin . PI

1995 ; R dri gu z I a/. 19

ntr

cl

taL pan rea ti , nd m tri al,

r

that

ll-

11 har t

nc g 111 p th a

fth

mpl

TPa

d in c 11 ignalling and

m nt m

trun n

path a

RD -BP

).

flun g ca ncer

ergham

c n1n1 n mutati ninth glyc in e- 12 p

redu c

TPa e ac ti it , effec ti

preventing hydrol y i

f th e ac ti ating

"'

7

/

a! . 19 7; B

f

' I a/. 1987;

iti n t a parti e ac id tr ngly

I ekin g th e m lec ul e in th e" n" tate by

TP m ol cule nee it ha bound , and i co n id ered an

impotiant biological fa ctor in man y can cer (G e rge et a! . 1993; Pylayeva- upta el a ! . 20 11 ).
Thi in effect bypas e externally regulated ign alling and prov id e th e nece ary biochemi cal
env ironment to induce th e cancer phenotyp e and i in fa ct the ing le mo t con1m on acti ating
mutation in human cancer ( ha w el a /. 2011 ). R A interferen e experim ent ha e fu rth r
hown that chan ge in KRA expre ion can independ ently regul ate the down tr am effector of
all the aforementioned pathway
V

F, and apoptoti c fa ctor F X

therapi e (Miura l a/. 2005) .

u ch a the mitoti c fa ctor eye linD 1, tran lation enhancing 6K,
m akin g Kra a lucrativ target for future antineop la ti

14

r gulat d KRA a ti ity in man

an

r h

e

tud

f

th wild-typ D rm. Inde d a c mpr h
pati nt found that high
rat

and unfa

pre

ild-typ K

urabl tr atm nt utc m

1

n

r an al

an

fr m

r

1 i n 111 nd m trial carcin rna
ith

rrelat

r

than did an mutant ariati n . 1 ated

nand KRA g n amplifi ati n p ci I C lly c ne lat d

ith an aggre

1

e

11 a in r a d fr m prim ary t m ta tatic le i n ( irkeland t a/. 20 12).
for wild-typ K

ch

l11

KRA gen amplificati n c IT lat
and panitumumab in

me

R

ith anti-

db

r le

alt 1ia t a/. 20 13) wh dem n trated

FR antib dy th rapy re i tance with cetuximab

pati nt . KRA inducti n i ind pendentl y abl e t indu ce a

cancer phenotyp in rat pancrea , and kn ckd wn ha demon trated th at m any pancreatic cancer
cell line ar dep nd nt n KRA o er-e pre ion Tuve

n et a l. 2004; Zhang e/ a l. 2006)

IMP1 has been found to be a maj or pl ayer in th e regulati n of KRA level by binding directl y
to the mRNA transcript (Mongroo et al. 2011 ).
directly interacts with KRA mRN

V -cro slinking ha demonstrated that IMP -1

in th e codin g re gion ( D ) and the 3'

TR and th at down-

regulation of IMP 1 directl y results in the reduced mRN A level and protein level of Kras in

RC

cell lines (Mongroo et al. 2011 ). Positive regulation ofKRAS expre ion by IMP l!CRD-BP
therefore presents a novel opportunity to target KRAS in cancer that co-express CRD-BP .

1.8 Targetin g KRAS in KRAS -driven cancers
KRAS driven malignancies are very cmnmon, effecting up to a third of all cancers but most
prominent in leukemias, colorectal, pancreatic and lung cancers and as such, a diverse range of
anti-KRAS therapeutic strategies have been pursed. There i still no prescti babl e go-to drug for
targeting KRAS in cancer, with many experim ental treatments falling short du e to systemic
toxicity or poor inhibition, which ha limited the degree to which KRA cancer can be treated

15

(Bain

ral pr m1mga

tal.20ll).H

thu far. The pr t in·

hall

thi ar a t m lecul

in apabl

pr t in
gr

ith mall m 1 ul

ha pr

nan ill-fat d trat gy

and lack f binding p ck t ha e limit d th
ub tantiall m dul ating K

binding intera ti n (Maur r t a!. 2 12;

u ce

lll

acti ity du e t 1 w-affinity

un r a!. 20 1 ; hima et a!. 20 13 ). H wever,

tal. (201 ) mad a r c nt KRA targeting br akthr ugh
inhibit r

arch that ha

fr

rec ntl .

pr du ed h p ful candidat
Directly targ ting th

nu

h n

trem

p riment utilizing thi 1-based

h w d tr ng pr t in binding character. The e inhibit r f01m a di ulphide b nd with
nth KRA pr t in, and are abl t c ax KRA

tabilizing the

int it inactive D 1m by

DP-bound c n£ rmati n. Furthermore, th t p candidate m lecule di play an

inherent preference for cell ex pre ing the

12D mutant ver ion f KRA

commonly found in

many cancer and achieve ome pecific anti-proliferati n activity. How ever despite the
promising binding data, it is empha ized that the e thiol-based inhibitors are un uitable as cancer
drugs in the current form , a Ras ignalling wa only marginally reduced . Development of thi s
class of inhibitors remains an active area of re earch pursing chemical "warhead ''more capable
of inhibiting KRAS signalling.
Famesyl-transferase inhibitors (FTl's) have been thoroughly explored in the laboratory and
in the clininc (reviewed in Appels et al. 2005). This class of KRAS inhibitor di rupts a po ttranslational farnesylation modification of the KRAS protein, whereby non11ally a large 15carbon hydrophobic tail is added to allow effective membrane as ociation. A KRAS achieves it
signalling function through interactions with other proteins near the cell membrane, failure to
attach a lipid anchor results in a significant inhibition of KRAS signalling. In practice, FTis such
as Lonafarnib and Tipifamib only proved effective against the le

common H-RA and N-RA

16

me1nb r

f the RA

alt tnati

1y p

tran :D ra

nlik th

mp n at r r dund an

nzym

u I pr

a it turn

pr t in

th r

t-tran lati nail m difi d b geran 1 tran f ra

again t KRA dri
imu1tan

up rfamil .

ut c n b

n inhibiti n f fa1n

larg 1 limiting th

TI

ucce

r . Targ ting

th th f n

1 and g ran 1 tran fera e nzym

d 1 th 1 t n n-can

r u and c

c r u cell alike, a many pr tein

n an

ary £ r ba ic cell fun ti n ar

ub trat

f th

yJ

e am enz me .

Pre enting KRA fr m fun ti ning at the cell m mbrane r main d an enduring therapeutic
trat gy

ith th intr du ti n f P

pho phodi e t ra

inhibit r .

i th d Ita ubunit of c MP

and i a KRA binding partner that preferentiall y interact with fame ylated

KRAS and aid in th tran lo ation of KRA to the membrane ( handra eta/. 20 12). m all
molecul

dev loped to inhibit thi critica l tep, b

pec ifi ca ll

interruptin g P

ciati on

with the famesylated tail of KRA but till allow ing the actual fane ylation tep to occur
culminated in the introduction and pre-clinical trial of Deltarasin. Do e-dep end ent delocalization
of famesylated KRAS using deltrarasin was recently demonstrated in human pancreatic cancer
cell line and in a mouse model of pancreatic du ctal adenocarcinoma where tumour growth wa
noticeably reduced, providing promi ing evidence that thi strategy is a valid method to target
KRAS . However, the dru g's lack of specificity produces hi gh ystemic toxicity and future work
still needs to be done before pushing tllis proto-drug to human trials (Zimmerman eta!. 2013 ).
Salirasib (S-trans,trans-fatnesylthiosalicylic acid) is another small mol ecule that has been
shown to interfere with proper membrane as ociation of KRA . Salirasib however do e not
directly target any protein or enzym e important in the po t-translational modification or
intracellular transport of KRAS , instead binding th hydrophobic famesyl tail directly, reducing
the overall hydrophobicity of the tail and impairing it effectivene

to interact with the cell

17

m mbran (W i z tal. 1 99) . R cent pre-clinical and lini al d

11-t 1 rat d in pati nt

furth r h wn that them 1 cul i

ad n car in rna (PD ) and nhan e th

ffecti

n

1 pm nt n alira ib ha

ith m eta t ti pan reati du tal

gemcitabin

in c mbination ( ah ru t al. 2012 . imilar pr bl m limit th ef.[i cti
uni er ally aH ct all RA i D rm
Inhibit r d ign d t bl

kK

chang

KRA and ha been th

p

f

nly

enting th ac ti ati n f KRA by inhibiting the

F

, required t induce

TP -binding and

DP D r

ific targ t of Patgiri et a/. (20 11 ) wh have produ ced a mall

molecule, a ynthetic a-helix that mimic the binding m tif of the
with the KRA -Ra -GEF

f alira ib as it

fr m reaching. r int racting with th membrane ar

chan g fact r: R a -

turn KRA "o n '. Th

ne

ith n cane r cell p cifi ci ty.

trategy, but an th r pr mi ing a enu e i pr

RA guanin nucle tid

h moth rapy wh n u d

protein, that competes

protein-protein interaction blocking the exchange of GDP for

GTP and hindering KRAS activation through receptor tyrosine kinase stimulation. Whil e thi
work has been limited thus far to cell line only, it is undoubtedly a promising avenue of
research.
Attacking KRAS at a genetic level by altering the rate of transcription is a more recent
approach pioneered by Cogoi et al. (20 13 ). Within the promoter region of the KRAS gene, it was
detennined that stretches of the DNA fonn G4-quandruplex structures which are recognized by a
myc-associated zinc finger transcription factor (MAZ). MAZ protein binding promotes KRA
transcription, and by using a synthetic DNA oligonucleotide G4-quadrupl ex rnin1ic that acts as a
MAZ decoy, KRAS expression could be effectively reduced in human pancreatic cell lines
(Pane-l) and reduce KRAS-driven tumour growth in mice by 64%. The pancreati c cancer cell
were shown to display increa ed apopto i upon admini tration of the MAZ-decoy and validate

18

th con

£ cu

Z- p cifi

pt of

t bili

n in rea ing ph

achie e the

f£

c mp tit r m 1 ul
fth rn 1 cul

t pr umabl du t degr dati n fth

It i cl ar that
h

4 quadrupl

1

urr nt t pi , and

pit lit rall d

e

rd

mpl

ity in targeting thi

fr
TPa

functi nally di tinct memb r

ear h,

-b

a r quired to

larg d

d rn 1 cul .

trategic th rap utic t. rg t in th war n cane r
r rnam an lu i

target in th

linic.

ignalling m lecule ari e fr rn the true turall y imilar, but

f th Ra famil

target th oth r relat d pr t in unint nti nally.
have only rec ntly ho n m d rat

a

. urt.h r w rk in t.hi ar a will

u c

hereb targeting K

u uall y inadv rtently

ppr ache wh r by KRA

i dire tly targeted

, u ing a thi 1-ba ed binding cheme, but the e till

are limit d in ability t actuall kn ckd wn KRA acti ity. FTI

h wed gr at early promi e by

preventing a critical po t-tran lati nallipid modification nece ary for KRA m mbrane
a ociation, but the co-acti ity ofb th fame yl- and geranyl-tran fera e inhibitors pr ved toxic
to cell . Similar toxicity plagues the alira ib approach, where ob truction of the hydrophobic
tail of KRAS by a masking molecule delocalize the protein. Interfering with the proper transport
of fame ylated KRA protein to the inner cell membrane u in g a PDE8 inhibitor has hown
some promise in mou e model

but again ystemic toxicity and poor dru g tolerance ha limited

the effectiveness of this molecule and requires further optimization till . Inhibiting the activation
of KRAS showed moderate success. Ras-GEF SOS is required to induce KRAS to bind GTP, a
necessary step in KRA signa lling, and with the u e of ynthetic a.-helix mimic de igned to
mimic the KRAS binding site, KRAS docking with the guanine nucleotide exchange factor could
be inhibited and overall activity reduced. Perhaps the single mo t promi ing work, including cell
and mouse-model data wa the G4-quadiuplex MAZ-decoy approach, a it wa able to
specifically target KRAS at the genetic level, and not the other RAS i oforms, but ha not yet

19

d t human clinical trial .

pr gr
mpl

targ t KRA

n

1 ad ne

itating th n

d £ r fut1h r r

a w 11 a new no

il

n£ rtunat 1 d pite the di
r bull t ha b en pr

r

atTa

f trat gi

n alth ugh th r ar

ome I r m1 m g

n f

f

arch int th

1 appr a h

1. 9 R e ea rch Goal
RD-BP ha a prrm r le in th initiati nand pr gr
pat1icular int re t i th recent di c

ry that

ral ancer .

R - P bind th KRA mRN

tran cript,

11-e tabli hed nc g n pre alent inc lor ctal, pancreatic
and lung can r . Targ ting KRA -dri

n can er ha proven e tremely difficult and de pite

thorough in e tigati n of a wid variety f anti-KRA
target d therapy. Importantly,

trategie

there remain

till n effective

RD-BP i an one fetal pr tein , and i nonnally only expr

ed

during early zygotic and embryonic tage of development. Adult ti ue nonnall y ha virtuall y
no detectable levels of CRD-BP ; however robust re-expression occur in man y cancer cell s and
therein lies a novel therapeutic oppm1unity that targets the core of some cancers and also i
inherently specific. Abrogating the KRAS mRNA interaction with CRD-BP therefore may be an
effective means of attacking some KRAS-driven cancers, and is the focus of thi re earch.
The first major research aim was to confirm that CRD-BP bind s KRAS mRNA in the
coding region and 3'UTR using the EMSA assay and to further thi s work by mapping CRD-BP
binding sites on the KRAS mRNA transcript using pmified recombinant CRD-BP protein and
IVT generated KRAS sub-regions. Determination of a minimum binding sequence wa the end
goal for these experiment c ntributing elements toward a fmal CRD-BP consen u
and serv ing as a focus for further study of the mol ecular nature ofth

equence

RD-BP-KRA mRNA

interaction. The second major goal was the development of binding inhibitor , pearheaded u ing

20

D

-ba d lig nu 1 tid

c ntigu u

tr t h

that c m 1 m ntaril bind thr ugh Wat

dru g m
t u

11 a a

-dri

and m u

a minimum binding equ n

d main r quir ment
goal of

~ rK

inding.

or

ibl e u e a pr t -

n cancer . h third maj r g al wa

abr ga ting

R - P interacti n ar an ultimat

ledge f th unique KH d main r quirement II r each
ary pr perti e

mRNA targ t hed light nt

properti

p

nal anal i t ga in in ight int th KH

and

I KZ pr t in re earch kn

which ub- et fKH d main n

and ha

r c gniti n

pairing t

m nt

uld all

fa minimum binding

f th ir ability t bl ck K

n- ri k ba

mall m I cul e inhibit r mu t have -

db bl eked. The final bj ecti e wa t elu cidate tructural

RD-BP . Pr t in cry tallizati n fl r u e in X-ray di ffracti n wa to be explored t

re eal the preci

me hani m f

RD -BP

I KZ protein RN -binding.

dditi nall y, circul ar

dich.roi m wa to be employed in order to a es the po ibility fa dynamic

RD-BP pr tein

that undergoes confonnational hift in econdary structure upon target- pecific mRNA-bindin g.
Determining tructural change that occur upon binding target RNA would offer an additional
means of creening potential RNA target molecule u ing CD spectroscopy.

21

Chapter 2
In vitro characterization of KRA mRN binding to CRDBP and development of anti en e oligonucleotide inhibitors.
h [! ll

mg h

qu nc forK

mR

t r

ver

p rim nt

arried ut t d t m1ine a minimum binding

r quircd to bind 'R -BP.

dditi nally. the devel pment of anti-

int raction will b pre ent d and di cu ed.
2.1 M ethodolo gy - Mappin g of

RD -BP bindin g ite on KR

' mR A

2.1.1 Tran formati on of BL2l (D 3) cell with p T 28b- RD -BP pi a mid.

Pia mid containing wild-typ
F':Jll-l ngth mou

R -BP v.a pre iou I acquired fr m the

RD-BP coding equ n

cut with Nc I and Xhol wa

r. J. Ro ss lab .

ubcl ned int

p T28b( +) (No agen). -flank d by an
T p ro m o tt>r 68

amino tem1inu FLAG epitope and a

-

terminu 6x Hi -tag. The pla mid contain
a kanamycin resistance gene and
expression is driven by a T7 promoter as
depicted in Figure 2.1. Approxi1natel y 100

1 7 tt>rminator

2D·B

111 of competent BL21 (D 3) cells were
thawed from -80°C storage on ice for
one half hour before adding 10 ng of

Figure 2.1. Plas m id map of pET28b(+)-CRD-BP

p -< T28b-CRD-BP plasmid DNA. ells were ubsequently placed on ice for an additional 25
minute followed by 90 seconds at 42° . el l were immediately upplemented with 00 ~LI of

22

L.B. br th (In itr g n) and in ubat d at 7°
ntaining 25 ug/ml kanam cin ulfat ( i h r

(In itr g n) plat
w r th n all

d t in ubat

rnight D r appr

ci ntifi ) antibi ti . Plat

imat 1 1 h ur , i lding col nie

f

tran D nnant harb uring th

2.1.2 Induction and

rowth of pET28b - RD-BP

Tran D tmant c 1 nie

ntaining th

inoculat 100 ml f

R - P

e t r wer u edt

. . br th (In itr g n) c ntaining 25 ug/ml kanamycin ulfate ( i her

cientific in a 250 ml
ew Brun wick

.c li c II

rl enm

r f1a k.

ientific, m d 1 Inn

11 u p n i n wa plac d n an incubator- haker
a 40) et at 200 rpm, 1 inch tr ke, at 3 7°

for 3 h ur

before being tran £ n d to a 1L e111bach fla k containing 900 ml of L. . broth with 25 IJ. g/ml
kanamycin. Culture wa gr wn until cell

.D .60 o equal ed 0.5 at which point th e cell were

_induced with 1mM I opropyl-~-D-1-thiogala c t pyrano id e (IPT

Bio Ba ic lnc.) and in cubated

under the same conditions for another 6 hour . Centrifugation of 1L culture at 4 °C, 2, 100 xg, for
15 minute (sub-divided into 4 volume of 250 ml) produced 4 cell pell ets which were tored at 80 °C .

2.1.3 Purifica tion & qu antification of recombinant CRD-BP protein
R ecombinant CRD-BP purification was achieved using the following buffers and schem e
as outlined in table 2.1 below .

Table 2.1. Composition of buffers used in denaturin g purification of recombi nant CRD-BP
. f rom t ran sf orm e d BL -2 1(DE3) E .co r1 ce ll s.
pro t em
Tris-Cl
pH
Urea
Buffer
NaH2P04
lOmM
8
lOOmM
8M
B
6.5
lOmM
c
100mM
8M
6.3
10mM
lOOmM
8M
D
5.9
lOmM
100mM
E
8M
lOmM
4 .5
100mM
8M
F

23

)

11 pellet

u p nd dint 12 ml f 1

buf~ r

whil b ing agitat d u ing a

a

.c li

clear db

21 (

d n1

g~ r

ntrifu gati nat 2 1

buf~ r

ell in 1

r n m dium

up 111atant wa th n yring filter d 0.

fi r 15 min u t

i ntifi Pr

tting (

minut

and incubating th

£ r 1 h ur. R in-1 at m1

ti a id

1-

ml

f bu f~ r B, £ 11

buf~ rD .

a

lut d b pa ing 12 ml of bu ffer

l2mlofbufferF,colle ting 0.5ml frac ti n

T

d by 5 ml f bu ffe r

f pr tein

ad

agar

iag n)

a applied t a m1m-c lumn

(Qiag n) and wa hed u ing
RD-BP pr t in

y at w

J.ll11 ) nd pr par d [! r batch-binding by mi ing in l

nick 1-nitril tri

n1

u t ).

at 4 ° , pr du cing a p 11 t f c 11 debri .

1nl f pre- quilibrat d (bufD r
lurr

re in u bated .G r 1 h ur n 1

!uti n.

and finally 5 ml f

r th e c lumn £ 11 wed by
naturedprotein ampl e were

immediat ly tored at - 0 ° .
Purifi ed recombinant mou e

RD-BP ampl e were re- natured using 3,5 00 MW

mm1

dialysis units (Thermo cientific) in a multi- tep bu ffer exchange. D en atured pr tein samples
containing 8M urea were dialyzed again t a 500 -fold volume of refolding buffer for 24 hours.
Following the protein refoldin g stage, ampl e wa then di alyzed again t a 500 -fo ld volum e of
fmal buffer (buffer recipe below in table 2 .2) for 2 hours follow ed by another dialysis against a
2,500-fold volume of final buffer to compl etely remove all traces of urea. Finall y concentration
of protein was determined using a BCA Protein A ssay Kit (Thermo Scientific).
'
f ld'mg an d st orage of pun'fi1e d recom b mant
CRD -BP protem.
T a bl e 2..
2 B u ffers use dfor reo
Refolding buffer
Storage (final) buffer
200 mM NaCl
200 n1MNaCl
20 mM Tris-Cl
20 mM Tris-Cl
2 M urea
10% glycerol
10% glycerol
0.01 % triton-X
0.01 % triton-X
pH 7.4
1 mM Glutathione - reduced
0.1 mM Glutathione - oxidized
pH 7.4

24

2.1.4

ub-r

neration of KRA mR

K

In rd r t a

u -r g1 n ffinit [! r

mR

b

ntigu u

tr t h

ign d t

m lify
th full

hi ch inc lu
parti I · untran lat d r gi n

c ding equ n

ign d t am lif

Prim r pair

r d

pr m t r

qu n e Ill
t mplat

, mall r fra gm nt

R

hain re c ti n P R u ing prim r

fir t g nerat

f cD

t mplat

ion cD

[! r

p

ifi

5' nd
ub

qu nt u

R ( n g ne,

mR

ub -rcgi n
fall~ r

in binding a a

).

ard prim r t dri

c
111

with

7

itr tr n cripti n

(tabl e , 4 and 5) and P R reacti n

carri d uta c rding t th e t mplat
Templat
10 P R buf:D r
d TP (2.5 mM)
forward prim r
revere pnm r
Taq. Polymera e (5 ,000
Water

/ml)

10 ng
3. 5 J..d
.5 ~~
100 ng
100 ng
0.5 ~~
t 35 ~I

h nn cycler pr gram
94°
30 ec
30 cyc le
50°
30 ec
45 c
72°

Table 2.3 . PCR primer used for the generation of KRAS mRNA ub-region cD
templates A-F for in vitro transcription. KRA mRNA ub-regi n nucl tide po ition

111

bracket are relative to the KRAS A G tart codon.

Region
A(l-185)
B(175-401)
C(388-610)
D(568-793)
E(772-988)
F(971-1155)

Primer

DNA se9.uence

forward
rever e
forward
reverse
forward
rever e
forward
rever e
forward
rever e
:£1 rward

5'GGAT CTAATACGACT

TT 3'

5' -TCATGAC TG T TGT G-3'
-3'
5'-GGATCCTAATAC A TCA TATAGG G AGGTCATG GGA T
TGT-3'
5' -G TAAGT TGAG
GA-3'
5'-G
A GATT

5' 5'-G
5'5 '-

TT-3'

T A TATAGGAT
TT - '
T A T

__..,, .

T G

5'TA

TT T

TT

-3'

25

Table 2.4. P R prim r u ed f r th
n
template KRA - (1-4) f r in vitr tran
p iti n in br k t ar r lati t th

m
d n.

ub - r gion cD
ub-r gi n nu 1 tid

A3-4 93-185)
A4(139-185)
A3(93-138)

Table 2.5. P R primers u ed for the oeneration of KRA mRN
template KRA -A34(a-.f) for in vitro tran cription . KRA mRN
po ition in bracket are relati e to th KRA A
tart cod n.

ub-region cD
ub-regi n nucl e tid e

Re ion
A3-4a (111-185)

3'

A3-4b(l29-185)

3'

A3-4c(l4 7 -185)
A3-4d(93-167)
A3-4e(93-149)
A3-4f{93-131)

rever e
forward
reverse
forward
reverse
forward
reverse
forward
reverse

3'

3'

26

2.1.5 Generation of

eP)-labelled KRA R
2

In itr tran ripti n (IV ) f

e
2

P] -lab 11 d

mitur

TP

a u dt

w r prepar d £ r a h

ub trate b IVT
ub-r gi n c

t

pl at

in th pr

ub trat . I

int 111all lab 11

r acti n

ub -r

ion R

nthe i .

IVT reagent
5x tran cripti n buf£ r

mount in 20 ul reaction volume
4 )ll

100mMDTT
20 U/ul RNa in

2 )ll

10 mMATP

1 )ll
1 )ll

10 mMCTP

1 )ll

10mMGTP

1 Ill
2. 5 )ll

DEPC H 2 0

(3 2P]-UTP

f

t mpl at ace rding t tabl e 2 ..

Table 2.6. IVT reagent u d for KRA mRN

100 J.1M UTP
15 U/ul T7 RNA polyrnera e

nc

1 Jll
To 20 Jll
2. 5 )ll

IVT reactions were mixed and incubated at 3 7°C fo r 1 hour prior to a 10 min incubation with 10
)ll ofRNase-free DNase (l U/ )ll) at 37°C. IVT reactions were topped with a urea loading dye
(9M urea, 0.01 % bromophenol Blue, 0.01 % xylene cyanol , 0.01 % phenol) and RNA transcripts
gel purified using an 8% polyacrylamid e (29: 1 bi :acrylamide) denaturing gel (7M urea) ran at
25 rnA for 1 hour with 0.5x TBE buffer. Excised gel-slices w ere cru hed and RNA extracted in
400 )ll of H 2 0 at 70 °C for 5 minutes prior to further purifi cation by Perfo rma DTR gel-filtration
cartrid ges (Edge Bio) . RNA wa then subj ected to phenol -chloroform extraction and ethanol
precipitation before being resuspended in 25 )ll ofDEPC H 2 0 . Purified [32 P]-labelled IVT
produ cts were quantified u sing a liquid scintill ation counter (Hid ex) and stored at -20 ° .

27

2.1.6

M

mappm of RD-BP bindin

Binding r a tion
r g1 n RN

n KRA mRN
h g n rat d

e

2

]-lab ll d

4 nM. M

a c rding t tabl 2.7 and

n u d £ r mapping high affini t
RN

it

binding it

ntain d 20

0 cpm

pr b .

T able 2.7 . EM A r eagent u ed for KRA - RD-BP affini mappin g and er e nin g of
anti en e oli gonucl eotid e inhibitor .
Final C oncentration
Volume
E MSA binding reaction reagent (1 ul)
Binding buffer ( e b 1 r)
4 ~1
0.53 mg/ml
1 ~~
10 mg/ml bak r' ea t tRN
(Sigma Aldrich)
0.53 mg/ml
1 ~~
10 mg/ml B A
(New England Biolab ;)
1 ul
2 /~1
40 U/~1 rRNa in
(Promega)
X
Ot 540nM
CRD-BP
32
20,000 cpm/reaction
[ P]-labelled RNA probe
2 ~1
to 19 ~~
H20
Bindin g buffer r eagents (per 1 ml)
1M tris-Cl pH 7.4
0.5 M EDTA (pH 8.0)
100mM DTT
50% glycerol
10% triton X-100
H20

50~~

25 ~1
100 ~1
500 ~1
1 ~1
324 ~1

50InM
12. 5 mM
101nM
25%
0.01 %
-

RNA probes prior to reagent mixing were first heated to 55 ° for 5 minutes and allowed to cool
to room temperature for 7 minutes in order to facilitate proper fo lding. Reagents were added
separately in a 1.5 ml eppendorf tube and centrifu ged at 3000 g for 10 seconds to begin binding
reaction; mixtures wer then incubated at 3 7 ° for 10 minute follow d by a 5 mi nute

28

in ubati n n 1

, an

th r 1 minut

at 7 °

nd a final

11 i ting f 2
1 11

.2%

an 1 an 40%

t tailing

ri -H 1, .2°-o r m ph n 1 blue

mM

edb 1 ading l 5 ~1

h amp!

add d t

-r d n n-d naturing g 1 in 0.

nt an % p
[! r

e p

minute t p n 1

minut

gel w r

at 2 m . M

d and i uali z d u ing a
mplet d n

ftwar

er 1 11 4.0 Packard In trum nt

2.2 - Methodology 2.2.1 EM A

Inc .

nti en e oligonucleotide inhibitor
ompetition a ay with pecific anti- en e oligonucleotide

I3-adiolab lled KRA of

mpan

ptiqu a11t

(nt 1-1 5) wa

ynthe iz d u ing in vitro tran cripti n in th pre ence

eP] - TP (m th d de cribed in ecti n 2.1.5 and u ed in M
2

with the parallel addition f

competit r molecul
Table 2.8. DNA sequences (5' to 3') of candidate
inhibition antisense oligo nucleotide u ed in
EMSA inhibition assays.
AON

DNA sequence

SMl
SM2
SM3
SM4
SMS

5'ACCACAAGTTTATATTCAGT AT3'

.

M A

experim ental co nditi on wer identical t
pre iou I y de cribed
experiment and can be reviewed in
ection 2. 1.6. Anti- en e oligonucleotid
(A N ) of 23-24-m r were d igned to
cov r the entir ty of the KRA -

r g10n

SM6
SM7
SMS

Technologie Inc. (table 2.8). creening of th anti- n

in a compl em ntary fa hion nd-to-end,
and w re ynth ized by Integrat d D
lig nucleotid , labelled

- M(l

29

d d initial! u ing tw c nc ntrati n : 10-~ ld and 5 -~ ld r lativ t th

thr ugh
KRA -

RN

pr b c nc ntrati n.

mpl

r

n

abrogat the int raction b tw
upper high r m 1 cular

-

n

RN

g I band.

ight

aut radi graph u ing the

ntrati n f 1, th

- P a indicat d

and

ibit r

w r gauged by th ir ability t

f th indi idual

w uld b

th EM

D r apr b

th l

fan

inding data wa analyz d by d n it m try f

cl ne t rag Ph

ph r- y t m and

t quantify th d gr

f compl e inhibiti nand quantitati

candidate A

ul

ly a

th

ptiquant

ffectiven

ftware

f

. urther

range from lX to 4000X molar-fold

T abl e 2.9. oncentration of anti sen e oli go
nucleotid e (A O ) u ed in M A
inhibition a ays . M lar-fold c ncentration s
are ba d n a calculated 1 l 8 pM KRA probe.

concentration (equivalent to 11 pM - 4 72 nM,

A ON molar-

a

characterizati n
pur ued u ing a great r c ncentrati n

see table 2.9). Additionally AO

M-2 wa

fold
lX

sx

further characterized as a negative control for

lOX
25X

inhibition with concentration tested from 1X to

sox

1OOOX molar-fold (118 pM to 11.8 nM) .

lOOX
250X

Densitometry analysis using Optiquant software
was utilized to produce inhibition plots using
Ka leidaGraph™ software.

so ox

750X
lOOOX
2000X
4000X

Concentration (nM)
0.118
0.590
1.18
2.95
5.9
11.8
29 .5
59
88 .5
118
236
472

30

2.3 Re ult and Di cu ion
2.3 .1 ynthe i of KRA mRN
n ration f KRA
b m an
r pr

fP R

a

ucc

ub -r egion by IVT

ub-r gi n

fu l wi th all

1

t mplat

in in ' 'ifro tran cripti n

[I r u

fragm nt , 1 b 11 d

thr ugh

nt d graphi all in figur 2.2.

• (97 1- 1155)

0 (56 - 93)

A( l -185)

5-

and are

,...--------.,....-------.,....-------...,..-------"""T

..
..
.. ... ..... ..... ....

·,·

·. -3'

1200

900

F igure 2.2. G raphi cal r epresenta tion of KRA m RN
ub-region la b ell ed KRA -(A
thro ugh F) cove rin g th e entire codin g sequ nee a nd pa rti a l 3' UT R . ucleo tide p siti n m
bracket ar r lati e to the KRA mR
tart codon (
G) and th remainin g un tudi d 3'
UTR region i repre ented by the double-peak dotted line .

DNA bands were of the expected size
506

as determine d by ethidium bromide

: 6
~44

_9R

UV visualization on a 1% agarose gel

220
154

anal ysis (figure 2.3} . Subsequent T7-

Figure 2.3. IVT eD NA t empl ates for KRAS subregions A through F. DNA samples were run on a
1% agarose gel with invitrogen 1 Kb ladd er and
visualized with ethidium bromide staining.

driven in vitro transcription reactions
produ ced

eP]-i nterna lly labelled
2

ra d ioactive RNA probe s of the
expected relative size with the gel extraction purification method typically yielding 30 111 of 0.5 2 million cpm/~-tl. RNA sub-regions were de igned to contain a small amount of sequence overlap
with the neighboring fragments (approxin1ately 20 nucleotid e ) to avoid plitting a potential
binding-sequ ence or di srupting a pecifi c RNA tructure such as a hairpin loop. While the full

31

rati nal il r thi

in th fa t th t R -

id

ding r gi na

~ r~

th
fth

hundr d nu I tid

al. 2 09 ; Pr kipcak ' I a/. 1

4·
'

11' -m

r PL

ha

\) ith ~-

tin

I a/. 2

1 '

t

i all

ind

nd

-m

m

F

u 1

ubi i 7/

I a/. 2

pr be (20,
R w r titrat d with

din g r gi n and parti al

rin g th entir

r v ithin th fir t ~ w

d u ing el ctr ph r ti c
. Radi a ti

hift a a ( M

'

I a/. 2

ffil

RD-BP affinit il r
m bilit

ith r in th

f

RD-BP affini ty mappin

2.3.2 EM

red. he

nu I

in I ngth , nl th fir t

Ri a pr

KRA

0 cpr r acti n)
- P and th ir ability

ed b th £ rmati n fa band wi th r duced mi grati n di tan e,

to £ rm a c mpl e a

D rm d by binding f

indicati e of a hi gh r m 1 cul ar w ight c mpl

RD -BP with the RNA

fragment. Fir tl KRA c din g r gion fragm nt · '(nt 1- 1 5). · 8 ' ( I 75-40 I) and ' · (3

-6 l 0)

were te ted (figure 2.4.)
KRA revion B

KR

r gi n
( nt l- 185)

KR.\ . region
(nt 388-61 0)

(n t 175-40 1)

CRD-B P:
(nl\ J)

Bound [
R~

lJnhou nd
RN

2

3

4

5

6

7

8

9

10

II

12

13

1-t

15

Figure 2.4. EM SA asse in g KRAS codin g equ ence RNA ub-region
to for C RD-BP
2
bindin g. P] -labelled unbo und RNA (20,000 cpm/reacti n) with no prot in add d (lan L6
and 11) form a larg r m le ular weight compl up n titrati n with RD -B P, a indicat d b
the bound RNA fra cti n (Ian 2-5, 7- 10, 12- 15). Binding r a ti on wer in uba t d il r a total of
20 minu te at 37° and 10 minute on ic b £ re ing re lved n an % polyacr ]amide gel.

e

32

R lati

a D und £i r ub -r gi n K

-BP affinit

tr ng

b th

ing abl t £i rm a high r- rd r band u gge ting th fl 1mati

compl

. Wh n

-

mparing Ian

1-

(Ian

t 1n

-pr tein

-

K

and K

(1 an e 1 1-1 ) th tw
f the c ding

f the fir t l 5 nucle tid
qu nc fl rm a m r di tin t c mpl

at 540 nM f

larger b und fra ti n f R

and h

6-1 0) c n i tentl y h wed a weak affinity with the gradual

RD -BP . In c ntra t, KRA -B (lan

ut

up to 540 nM concentrati n, and thu

a c n ider d t be a relativ ly low affinity regi nand

inve ti gated no furth r. RN

ith an gligibl change in th unbound R

fracti n

app aranc of tw faint high r band

fragm nt repr

enting ub -r g io n of th partial KRA

R

3'

mRNA were n e t te ted for affinity in the arne manner.

r eo ion E
(nt 772-988)

KR

KR

region D
(nt 568-793)

CRD-BP:

~

'\,:

::0

,~

'\.

,..,~

~
,.,~

~

::0

'V

'o

,.:,~

reojon F

KRA

~

(nt · 971-1155)

'""~

'

~

(n~'l)

~

'\,.\:)

'V

,..,':¥'>.

'""');

Bound
Rl\A

..

nbound

RNA
2

3

4

5

6

7

8

9

10

11

12

13

14

15

F igure 2.5. EMSA assessin g KRAS 3' UTR RNA sub-regions D to F for CRD-BP binding.
32

P] -labelled unbound RNA (20,000 cpn11reaction) with no protein added (lane 1,6 and 11)
forms a larger mol ecular weight com plex upon titration with CRD-BP , a indicated by the bound
RNA fraction (lanes 2-5, 7 - l 0, 12- 15). Binding reactions were incubat d for a total of 20 minute
at 37° and 10 minute on ice b efore being resolved on an 8% p o lyacrylamide gel.
[

33

RN

KRA -

th di pla

and

chang in th unbound fracti n fRN

c un d

ith in rea ing

Inter tingly th KRA -F r gi nand t a le er

n
-

t nt th K

ith a faint band D nnin g ju t ab

high r rd r c mpl

- P and a ignificant

d th ability t bind

ntrati n

r gi n , D rm multiple

th unb und R

fra c ti n in

tandem with a m re pr min nt band ab
that fl r on] the
how

regi n (nt

er the phen m n n d

7 1- 11 5 and 772-

n rep rt d pr

IMP-1, which ha a

%

pecw to H 19 mRN

and app ar

II mR

ith

n t appear t c

binding m de ha e b

i u 1

quenc id ntity t

EMSA much like KRA -F and KRA -E (Niel

ccur eparately;

rail affinity fl r

ith ari u targ t of

- P.

th

KZ pr tein .

RD -BP , and i capabl e f binding a a ingle

ry imilar to th e

'zip ode" re g ion bind IMP-I in a

, tw binding e ent

M

hift pattern f KRA - . The I

qu ential dim riza ti n m d e a detennined by

n et al. 2002·

hao et al. 201 0) . Binding data

for all six fragments was analyzed by den itometry of the autoradiog:raph u ing the

yclone

Storage Phosphor-System and Optiquant oftware and Kd plot generated by fitting to the Hill
equation for quantitative comparison (figure 2.6). KRAS-C region was not included in the Kd
calculation due to repeated inability to achieve greater than 70% bound fraction in the EM A .
KRAS mRNA regions D (nts 568-793) and F (nts 971-1155) , representing the first 588
nucleotides of the 3' UTR, also revealed multiple binding sites for CRD-BP (figure 2.5, lanes 15, and 11-15 respectively) .

34

KRAS r eg 1on A

00
'0

c

::J

KRAS reg1on D

100
'0

c

80

::J

'0

c

80

::J

.c

.c

~ 60
a::

~ 60
a::

~
a::

0

0

0

c

:p
u

40

c

c

t; 40

tl

e

20

20
Kd=171

Kd =277 63

0± 771n

6

Cl =220 18 ± 21 45 nM

3 nM

0 ~~--~--L-~--~~

o ~~--~--~~--~~

0

100

200

300

400

500

40

u.

u.

20

0

0

~

"'.....
u.

80

0

0

.c

KRAS reg1on F

100

200

600

'00

400

500

0

600

CRD-BP concentration (nM)

100

200

300

400

500

00

CRD-BP concentratiOn (nM)

CRD-BP concentration (nM)

ub-r 1 n
D and F. ata p int ar a eraged fr m
Figure 2.6 Kd plot for KRA mR
p nm nt and fit t the Hill quati n D r Kd d terminati n
two bioi gi al t f triplicat
(Hill c effi i nt = 2) · IT r bar r pr ent tand ard d iati n .
a th K

mparing th calculat d Kd alue d ubl c nfirm d K
ub-region p

mRN

ing the highe t affinity binding ite f th e amin d fragment

the calculated

(171.10 ± 17.71 nM) being lower than KRA - and alm

t 1.6-fo ld 1 wer

Kd value£ r KRA -

than KRAS-D (220.1 ± 21.45 nM and 277 .63 ± 64 .13 nM, re pecti ve1y) . urther binding
characterization of the KRA -A fragment wa done with additional

RD-BP concentration to

more accurately detennine the binding character and similarly a aturation binding curve plotted
to determine the Kd and Hill Coefficient (figure 2.7 and 2.8, re pectively) .

RD-BP:
(oM)

Bound
R~ A

Unbound

~t

R.

2

3

4

5

6

•

7

8

9

JU

11

12

13

14

15

16

17

Figure 2.7. EMSA saturation bindin g experiment between CRD-BP and KRAS-A (nts 1185). RNA (20,000 cpm) with no protein add ed (lane 1) fon11s a larger molecular weight
compl ex upon titration with RD-BP, as indicat d by the bound RNA fraction (lane 2-17).

35

aturation bindin g curTe
.K:&
• l -185) YS
BP

0

80

0

"0

c:

:::1

0
.0

0

60

<
2

0

0:::

c:
0

~ 40

....
(o::l

lL

10

CRD-BP concentration (nM}

Figure 2.8. EM SA Saturation binding curve for CRD-BP vs KRAS-A (nts 1-185). Data wa
derived in t1iplicate from aturation EM
e periments where RD-BP wa inc ubated with
2
P]-labelled KRAS-A RNA(20 ,000 cpm) and fit to the Hill equation.

e

Further refmement of the KRAS binding equence was pursued in a imilar manner by
producing 5' and 3' truncations of the high affinity KRAS-A region. Small eDNA templates
covering arbitrarily assigned sub-regions ofKRAS-A (1 through 4) were succes fully generated
using PCR (figure 2.1 0) . If upon retention or removal of one of the four approximately 50
nucleotide regions resulted in drastic changes in CRD -BP affinity, it could be deduced by
combining data where the high affinity site(s) is located. IVT reactions canied out in the
presence of radio-labelled UTP were used to synthesize RNA trands corresponding to the
repre entative mRNA regions a depicted in figure 2.9.

36

KRAS-A4 ( 139- 185)

~1--~~~--~~0~~--~~~--~~0~~

5'-------~------~--~~----~~------------~----------~
-3
I
50
100
150
200
( .\ ' G)

~---------------------------------------------------------------------------------------------------------------

4

Figure 2.9. Graphical repre entati n of KRA - (1 through 4) R
qu nee and nucle tid p
).

iti n in brack t are relati ve t

B

A
,.,. ,.,.,,., .....,,.., .....'
'

bp

. ragm nt c ver th

- ,

,

.::...: \

~

~

~

C...,' \
......

~~" . ~ "j! ~\"- . ~~~ . ~

~

~

~'

....._.,~> .....~ .....~
, \ C...,' \
....... .....•

C...,'

~~ . ~ ~ . ~'\""·

,~

~

~

~

400 - 00 20 -

bp
400300 -

200 -

100-

100-

--

........,

~ . . ..

...

""'~~'/11f

Figure 2.10. KRAS sub-region eDNA templates for in vitro tran cription. ( ) 5' and "'
truncation of KRAS region A and (B) 5' and 3' truncati on of KRA
1% agaro e gel vi ualized by V -ethidium bromide taining .

regio n A34 r

Synthe ized RNA fra gment were agam a e ed for relativ affinity u ing EM
[

32

P] -labelled RNA probe (20 ,000 cpm/ rea ction) titrat d w ith

the bindin g a ay for the ' trun at d fra gment
lane 1-5, -10 and 11 - 15 r pecti e ly.

1 , A 12 and

I ed on a

and

RD-BP up to 540 nM . Re ult
I an b

e n in figure 2. 11 in

f

37

K RA

RO ~B P :

'

~

(n~l)

13oun d [
R .A

Un bound
R l

5

6

7

8

9

10

II

12

13

14

15

F igure 2.11. EMSA asses ing KRAS-A 3 trunca ted ub-r egion s A l 3, 12 and A I fo r RDBP bindin g. 2P]-labelled unbound RN 20,000 cpm/reacti n) with no pr tein add ed (lane
1 _6 and 11) form a larger molecular weight compl ex with RD -BP (lane 2-5, 7- 10, 12-15).
Binding reaction were incubated for a total of 20 minute at 3 7° and 10 minute on ice before
being re olved on an 8% polyacrylamide gel.

e

It is immediately apparent upon visual in pection of lane l to 5 (figure 2. 11 ) that only the

KRAS -A l3 region corresponding, to nucleotide po itions 1-141 ofKRAS mRNA retained the
ability to bind CRD-BP with measurabl e affi nity. From the systematic 3' truncation of KRA -A,
it was detennined that KRAS-A1 2, one full half ofKRAS-A consistin g of 92 nucleotides, did
not likely possess any binding information. Further EMSA binding experiments were carried out
usin g 5' truncated fra gments of KRA -A to examine the po sibility of a binding equence
present in the 3' end f KRAS -A (figure 2. 12).

38

RD-BP :
( n J)

Bound [
R, A

Unhound

R!'i.

2

3

5

6

8

10

II

12

13

14

15

Figure 2.12. EM A as es in g KRA -A 5' trun cated ub -r egion
24, 34 a nd A4 for R D32
BP bindin g. [ P] -labell ed unb und RNA (20 000 cpn reaction) with n pr tein add ed (lan e
1_ 6 and 11 ) form a larger m lecular weight complex with RD -BP (lanes 2-5, 7- 10, 12- 15) .
Binding reaction were incub ated fo r a total of 20 minute at 3 7° and 10 minute on ice before
being resolved on an 8% polyacrylamid e gel.

From th e EM A binding a ay with 5 · trun cati on of KRA -

in fi gu re 2. 12 It can be

seen that both KRAS -A24 and KRAS -A34 are both capable offmming a protein-RNA complex
while KRAS -A4, th e ta ilin g 47 nucleotid e 3' end , shows no interacti on. A faint band in KRA A4 can be seen directl y above the unbound RNA, however it is pre ent in the ab enc of an y
protein and its magnitude does not change with increasing CRD-BP concentration and i thus
considered a non-binding region. To further confinn the binding di parity between KRA -Al2
and KRAS -A34, another experiment was completed pitting the tw o fragments side-by-side on
the sam e gel and effectively doubling the protein concentration to achieve binding aturation for
Kd detennination (figures 2. 13 and 2. 14). Direct compari son ofthe 2 regions how that KRA A34 bind to

RD-BP more aggress iv ly than KRA -Al 2, de pite a ize diffl renee of only one

39

nu 1 tid

h

d that
affinit

imi lar t th

fra ti n

540 nM

nu I tid

th 1 arl

hil

ragm nt

tin zi

1 a/. 19

.K

ugg ting that p rha
fragm nt

it

it elf pr

te ted fo r RD-BP affinity b th

qu enc repr

und

dt b ap

icl en '/ a/.

r inding r gi n,
qu nee a

di pla ed binding. Hen , a mall
ith

an

1 m nt ar gr at r th an 100

th t c ntain d the bindin g

ntaining regi n-

nt h

nt in th

i raeli et a/. 1

a r gr n

prob (20 ,000 cpr r a ti n)

- 24 ( 1

f

ti

pr tein binding R

~-a

, ' ith th

4, it an b

- 2 and K

11 f th

ith n arl
nu

in that m t

200 ; Pr kip ak

R

fl nnati n, th 1 rg r

mp l

h

4 r gi n nd that th

gniz d l m nt in th

ifi binding. R turning t

affinity i n t u t n n-

with lit ratur

ifi r

ing that th r i a

nting r gion K

nJ y

eP]- Ia ell d RN
2

wa

ynth e ized and

II

12

am meth d (fi gur 2.15).
KRA

RD-BP:
(n\1)

Bound [
R~A

nbound [

·~ '
1

2

3

4

5

6

7

8

9

10

Figure 2.13. EM A assessing RNA KRA -Al2 and KRAS-A34 for CRD -BP bindi ng. (~P ]­
labelled unbound KRA -Al 2 and KRA -A3 4 (20,000 cpm/ reacti on) with no protein add d (la ne
1 and 7, re pectively) fmm a larger molecular w ight mpl upon titrati on with RD-BP fo r
the KRA -A3 4 RNA fragm nt but onl y to a limited e tent with KRA - 12 up t 1200 nM, a
- 12) .
indicated by th b und RNA fraction (KRA - 12: lane 2-6; KRA - 4: Ian

40

KRAS region A34

00
"0

c:

:J
0
.0

80

~ 60

0::
c:
0

+=i
u

rn

40

L.

u.

20
fd

= 37 78 ± 25 408 n
00

CRD-BP concentration (nM)

Figure 2. 14. aturation bindin g curve fo r RD-BP vs KRA - 34 R A (nt 93-185). Data
wa produced from aturation M
xperiment ho n in figure 13 where RD-BP was
32
incubated with [ P]-labelled KRA - 34 and KRA - 12 RNA(20 000 cpm) and fit to the Hill
equation.
KRA (nts 1-185)

RD-BP:

~

(n:\rl)

Bound[
R~

Unbound
R!\A

1

2

3

4

5

6

7

8

9

F igure 2.15. EMSA assessin g KRAS-A3 RNA for CRD-BP bindi ng. [32 P]-labelled unbound
RNA (20,000 cpm/reaction) with no protein added (lane 1 and 4) fom1 a larger molecular
weight complex upon titration with RD-BP only for the control KRA - RNA fragment but
not with A3 , as indicated by the bound RNA fraction (lane 2) .

41

KRA -

did n t h

972 nM. Thi un

p

n with

any ignifi ant inding

n requir m nt f nearby

ted r ult rna ha

qu nee r gi n ( p cifi

RD- P concentrati n u t

r n n- p cific) ut id e regi n

RNA :D lding :D r r c gniti n b

that ar in trum ntal in pr p er

uld b that inding f K

thr ugh a bipartit m d , wher b multipl binding ite e i t al ng th
full binding. In thi
may

c nari , a ingl

i t in th flanking regi n K

ite ma fall

trand t fa ilitat
whil an ther ite( )

- 2 and KRA - 4 uch that nl y a c mbinati n f tw

binding e ent ar r quir d :D r tabl binding.
and RlP- hip

ithin r gi n K

achi v

id n

:G r thi

p rim nt id nti fyi ng on erved 5 · and 3 ·

quen

cenari c me fr m

L X

in man amplified R A

target of I KZ prot in , ugg ting multipl e binding ite are i1np rtant :D r any given RNA
target. Howe er among the coh rt of amplified RNA target , ome equ ence were pre ent that
did not share any obviou

equenc

imilarity with the other identified (Farina et al. 2003 ; Patel

et al. 20 12).
Further mapping of the CRD-BP binding sequence identified in KRA -A34 wa canied
out in one more set of 5 · and 3' truncation s of KRA -A34 acco rding to the graphic below (figure
2.16). While the previous systematic truncation of KRAS-A into four rou ghly equally ized
segments (regions 1 through 4 ), this series of truncations aimed instead to simply have off
terminal nucleotides from both the 5' and 3' end of KRAS-A34, to fmd the smallest fragment
that can still produce a complex as detennined by EMSA.

42

Figure 2.1 6. Graphi ca l representation of KRAS-A34(a-f) RN . Fragment repre ent ubdivi ions of the KRA -A34 r gion covering nu cleotid e 93-185 f the codin g equ nee.
ucleotide po ition in bracket are relati e t the KRA mRNA tart c d n (A G) .

To further nanow down a minimum CRD-BP binding equ ence, an additional et f RNA
fragments representing ub-division of the parent KRAS-A34 positive binding region were
synthesized. The ability of the 3' truncated fragments to bind CRD-BP was a sessed by EMSA
and the results for RNAs labelled KRAS-A34( d, e or f) can be seen in figure 2.17. Analyzing the
results of the 3' truncated fragments of bindin g region KRA -A34 showed virtually no binding.
As with the previously discussed fraginent KRAS-A4 which presented with a faint upper band
but did not conelated with CRD-BP concentration it can be seen that KRA -A34(d)- covering
nucleotides 93-167, also presents with a similar pattern, and is concluded to be a non-binding
fragment.

43

RO- BP :
(n 1
Bound [

R:'\A

lnbound

R..'\ .

2

3

4

5

6

9

J ()

ll

12

13

14

IS

16

17

e

F igure 2.1 7. EM A a e in g th e KRA -A34(d e a nd f) RNA , for RD-BP binding. 2 P]labelled unbound RNA 20 000 cpr reac6on) with n pr t in add ed (lane 1, 3, 8 and 13) fonn
a larg r molecular weight c mpl upon titrati n with RD -BP onl y in theca e f the KRA -A
probe which wa u ed a po itive contr 1 and includ ed t d m n trate binding c ntra t for the e
nQn-binding fragm ent . Binding reactions were incubated fo r a total of 20 minute at 3 7° and
10 minute on ice before being r ol ed on an 8% polyacrylamid e gel.

To en ure that the lack of co mpl ex fo rm ati on in any of th e 3' truncated fragm ent was not
due to experimental eiTor, a positive control previously determined to bind (KRA S-A) wa
includ ed. From thi s, it was concluded th at the initi al 3' truncati on, th e large t of the three w hi ch
removed 19 nucleotides to produce region KRAS -A34(d) likely contained binding elements.
The compl ementary 5' truncated fragment , labell ed as KRA -A34(a, b and c) which retained
the aforementioned 19 nucleotide sequ ence, w ere also te ted for binding using EMSA (figure
2.18 .) Complex formation can be readily observed as a di tinct band above the unbound fraction
of RNA in fra gm ents KRAS -A34(a) and KRA -A34(b) with binding almo t entirely lo tin
KRAS-A34(c).

44

KR
RO-BP :
oM

'

fb

Bound [

RN

nbound

RNA

J

4

5

6

9

10

II

12

14

15

e

F igure 2.18. EM A a e sin g KRA - 34(a b a nd c) for RD-BP bindin g. 2 P]-labell ed
unbound RN (20,000 cpn reacti n) with n pr tein add d (lane 1, 6 and 11) .fi nn a larg r
mol cular weight c mple upon titration ith RD-BP a indicated by D rmati n of a band in
".Binding reacti n wer in ubat d :D r a total of20 minute at
the area labelled' b und R
3 7° and 10 minute on ice beD reb ing re lv d nan % polyacrylamid e g 1.

Oddly, the addition of CRD-BP appear to shift the unbound KRA -A34(c) RNA further
down the gel, increasing 1nigration di tance a indicated by the increa ingly den e
autoradiographic expo ure appearing below the free RNA in lanes 11 to 15 . While this may be
considered binding, a lack of a di tinct band sugge ts at mo t non-specific interaction and wa
left as an anomaly. Interestingly it is possible to di tingui h an intem1ediate band between the
upper shifted complexes, and the unbound RNA which was not readily apparent in the nontruncated KRAS-A34 fragment. Hence, the smallest region ofKRAS mRNA presented here
capable of binding CRD-BP as dete1mined by EMSA i the 57 nucleotide region labelled KRA A34(b) ; nucleotides 129-185.

45

2.3.3 Development of pecifi c anti n
D

n 1 p ram unt t

1 pm nt f m lecular inhibit r

th 1 ng t rm g al

f tud ing

1 w r und t

tabl 1

pr li:D rati

targ t gi

pr

arli

t

k

1 in later li:D , it pr

i n f thi

oligonu le tide

nc and

i and tap nng ff t e tr m 1y

ti ity in tum r ti ue 1 a clear anti-

c rtain cane r and it ha b n h wn that

ignalling
-BPi capable f

tal. 201 1) ingling ut th KRA - RD- P

interaction a an ideal target :D r m 1 cular inte1
ba ed anti en

g n

R - P und r trict

th path ay .

n it in

ha be n d mon trated t dri

targ t . With

RD- P nd it m

ntr 1 during th

pati t mporal

enhancing

oli onucl eotid e inhibi tor

nti n . T approach tlu ta k 23-24 mer

ere de igned t contigu u ly c ver the entire regi n f

KRA -A in a compl m ntary fa hion (nt 1-1 S) with the rational being that important RNA
sequence or tructural element within KRA -A, one bound thr ugh Wat on- rick base
pairing, will impede

RD-BP binding to KRA mRNA. Any di covered DNA-based

oligonucleotide inhibitors could ub equently be synthesized a an 0-methyl RNA
oligonucleotide derivative and used in the context of cell-based anti-proliferative as ays for
further development. Under binding conditions identical to the previou ly outlined EMSA
32

binding experiments, [ P]-labelled KRAS-A (nts 1-18S) RNA was incubated with 216 nM
CRD-BP in the presence of eight potential inhibitor oligos at 1OX and SOX molar-fold
concentrations (figure 2.19). Protein concentration wa cho en to only fractionally bind the RNA
probe (as opposed to full shift at S40 nM) to avoid over saturation of the RNA and increase the
protein-RNA complex sensitivity to the inhibitory effe t f the oligonucleotide for scr erung. In

vitro transcribed unlabelled KRAS-A RNA was also incubated with CRD-BP at 1OX and SOX
concentrations to erve a a positive control for inhibition (figure 2. 19, Ian s 3-4 ). Concentration

46

of radio-lab 11 d KRA w r mad t 1.1

mp tit rm 1 cui

wa d termin d t b 11 pM and thu

pr

10

nM and 5.9 nM r

ba ed n abilit t r du

and 0

th b und fra ti n f

r w r th n

. uc
in th

M

O ligon ucleotide Inhibitor

Bou nd [

RN

nbound [

RNA
1

4

5

6

7

10

11

12

13

14

15

16

17

I

19

20

e

Figure 2.19. EMSA cr een of anti sense oli gonucJ eotid e 0 -SM(l -8) inhibitor . 2P] labelled KRA - RN probe (-/lane 1) wa partiall y hifted by 216 nM RD-BP (+/lane 2) in
t~e pre ence of 23-24-mer anti n e oligonucleotide in 2 concentration : 1OX molar exces
relative to 118 pM radioactive probe (1.1 nM), and SOX (5 .9 nM) in lane 3-20. Effectivene
of potential inhibitor i evidenced by a reduction in the bound RNA fraction indicated on th
figure left. Enor bars are standard deviation.

Screening of SM1 to SM8 inhibitor oligonucleotide was completed in triplicate and when the
bound RNA in the protein-only reaction (lane 2) i compared to those with inhibitors added , it
can be seen in the representative figure that a reduction in complex fonnation occuned for
inhibitors SM6 (lanes 15-16), SM7 (lanes17-18) and to a lesser extent SM4 (lanes 11-1 2). Co ld,
unlabelled KRAS-A which was used as a positive control for inhibition did display a reduction
of the complex as expected, however the extent of inllibition wa

omewhat less than predicted

possibly due to error in RNA quantification of either the radioactive pr be, or the unlab lled
competitor RNA . Densitometry of the autoradiographs using the Cyclon e Storage Pho phorSystem and Optiquant software was used to construct a graphical repre ntation of the EM A

48

(figur 2 .21 , ian

1 - 17) .

in additi n t a ting a an n

g n

m

me cir um tan

th

am

f[i

nd

ran g

f th inhibit r li g nu

tid

M

c nditi n rem ain d th

am

a hie

rna imal KRA -

M 7 wa

fr m 1

;\Jolar-fold:

-

r

ting a
r m ll m 1 u 1e w ith

hift, and all

tid

r

3

4

panding th e c n entra ti n

n1 d ut b
, r 11

t

ep t [! r

- Pc n

p

t 1I n

ntrati n

a

chara t riza ti n f M and

KRA - RD -BP binding rang . H r , M 2
li gonucl

d ba g n

enefi ial u

Furth r qu ntifi ati n f

anti- en

c pr

inding u ing Ji g nu le tid

r, nhan mg
t rna ha

it an a t a a tum r UJ pr

rr lat d K

(Zhang t a/. 2001 ). In u h ca
a tum r uppr

a bit f a mi

tat d arli r in th intr du ti n

a al

nin g that ugg

fi gure 2 .2 1) .
ct at 54 nM t

M 7 ver th entire

includ ed [! r c mpan

t d it may fun cti n a a g

nb

ed nth initi al

d negati e c ntr 1.

+

'nbound [
R~

(KRA ~ - A)

2

5

6

7

8

9

JO ll

12

13

14

IS

16 17

F igure 2.21. EMSA cha ra cteri zin g C RD-BP-KRA -A RNA compl ex inhi bition b y A O N32
SM6. [ P] -labell ed RNA - regi n KRA -A prob e (unbound in lane 1) wa incub ated with 540
nM RD-BP (lane 2). Inhibitor M6 wa pre ent (lane 3- 12) in oth erwi iden ti cal ondition
to lane 2, ranging in concentration from 1X mo lar-fo ld relati ve to probe cone ntration ( 11 pM)
to 1000 mo lar-fold (11 nM) which COlT lated with a lo of RD -BP- bound RN . In contra t
A N- M 2 di play d dra ti cally r du ced ability to interrupt the RD -BP-K RA - RN
compl ex.

46

ofradi -labell d KRA -

pr b

a d t rmin d t b ll

wremad t 1.1 nMand .9 nMr pctiel
lect d ba d on ability t r du

10

M and thu c mp tit r m 1 cul

and50 ).

ful inhibit r w r then

uc

in th

th b und fracti n f

M

molar-fo ld : (-

Bou nd [

R.'i

nbou nd [

RNA
4

5

6

7

I0

II

J2

13

14

J5

16

J7

I

19

20

F igure 2.19. EMSA scr een of a nti en se olioonucl eotid e AO N- M (l -8) inhibitors. [32 P] labelled KRA -A RN probe (-/lane 1) wa partially hif1ed by 216 nM RD -BP (+/lane 2) in
the pre ence of23-24-mer anti en e oligonucleotide in 2 concentrations: lOX molar exce s
~elative to 11 pM radioactive probe (1.1 nM), and 50X (5.9 nM) in lane 3-20 . Effectivene s
of potential inhibitor i evidenced by a reduction in the bound RNA fraction indicated on the
figure left. Error bar are tandard deviation.

Screening of SM1 to SM8 inhibitor oligonucleotides wa completed in triplicate and when the
bound RNA in the protein-only reaction (lane 2) is compared to those with inhibitors added, it
can be seen in the representative figure that a reduction in complex formation occurred for
inhibitors SM6 (lanes 15-16), SM7 (lanesl7-18) and to a lesser extent SM4 (lanes 11-12). Cold,
unlabelled KRAS-A which was used as a positive control for inhibition did di splay a reduction
of the complex as expected, however the extent of inhibition was somewhat le

than predicted

possibly due to eiTor in RNA quantification of either the radioactive probe, or the unlabelled
competitor RNA. Den itometry of the autoradiograph u ing th
System and

yclone Storage Pho phor-

ptiquant oftware was used to construct a graphical repre entation of th EM

47

a r du ti n in th
hang and 10 % b ing

b ing n

mpl t

f th b und

mpl

inten it , with 0%

band (fi gur 2.2

.

*
60
50
c:
0

·.p
0

0

:J

"0
OJ
'X
OJ

30

a.
E

20

0

0

0

lZJ1 Ox

al ar excess

.

ol ar e c ss

50

~

0

- 0
Competitor oli go

Figure 2.20. Bar graph of anti en e oli gonucl eotid e molec ul e M(l-8) inhibitory effect on
CRD-BP-KRA R A compl e form ation . Data i pr ented a the av rage percent redu ction
of bound RNA compar d to a control w ith no inhibit r pre ent in an electr phoretic m bility
VA · F=3.47, p<0.02,
hift a ay. D ata wa ubj ected to tati ti cal an aly i u ing a ne-way
n=3 and stud ent' t-te t (p<O.OS ). Anti en e oligonucleotid e (23-24 m er) wer pre ent in 2
concentration : 1OX m olar exce relative to 11 8 pM radi oactive probe ( 1. 1 nM), and SOX ( S.9
nM).

Plotting the anti sen e inhibitor data a a bar graph in fi gure 2.20 more clearl y depict the
relatively trong ability for M 6 and

M7 to redu ce the

RD -BP-KRA RNA compl ex

formation . At SOX molar excess (S .9 nM), M6 and M 7 both achi eved appr
reduction f the bound RNA band , tw ic the ffectivene
Intere tingly,

imately half

of the ne t mo t ffecti v inhibitor.

M 2 and M 3 appeared to be capable f trength ening the KRA -A-CRD-BP

complex fonn ati n in the

experimental co nditi n . Thi phen m n n f increa ing the bound

RNA fracti n wa repeatable in the initial ere n, but wa not e id ent at high r con entration

48

(figur 2.2 1 Ian

1 - 17 .

it f

tat d arli r in th intr du ti n

in additi n t a ting a an n

g n

m

m e cir urn tan

it an act

m1

d bag n

a tum r uppr

r

(Zhang t a/. 2 01 ). In u h a
r, nhan mg

a tum r uppr
th

am

f[i

t

R -

binding u ing li g nu c l

nd

M7

a

rang

f the inhibit r li g nu 1 tid

fr m 1

t 1

M

c nditi n r main d th

a hi

;\lola ·-fold:

-

, r 11

charac t rizati n f

KRA - RD-BP binding rang . H r , M2
lig nu 1 tid

n1 d ut b e panding th e c ncentrati n
t 11

p

nM (fi gure 2.21).

R -BP c n cntrati n wa

pt [! r

am

hift, and all

n

with

n fi ial u

ma ha

Furth r quantifi ati n f

anti-

( r mall m 1 ul

tid

a al

M 6 and

et at 540 nM t

M7

includ ed .D r c mpari

cr enin g that ugge ted it m ay fun cti n a a g

er th e entire

n ba ed n th e initi al
d negati e c ntr J.

+

Bound [
Rl'iA

' nbound [
R~

(KRA -A)

2

3

4

5

6

7

8

9

10

11

)2

13

14

15

16 17

Figure 2.21. EMSA characterizing CRD-BP-KRA -A RNA complex inhibition by AON32
SM6. [ P] -labell ed RNA - region KRAS -A probe (unbound in lane 1) wa incub ated w ith 540
nM RD-BP (lane 2) . Inhibitor M6 wa pre ent (Ian 3- 12) in otherwi e identi cal condi tion
to lane 2, ranging in concentration from 1X molar-.D ld relative t pr be cone ntrati on ( 11 pM)
to 1000 molar-fold (118 nM) which coiTelat d with a lo of CRD-BP- bound R A . In ntra t
A N - M2 di played dra ti call y redu ed ability to inteiTupt the RD -BP -KRA - RN
compl ex .

49

M a a apabl K

In p cti n ffigure 2.21 re al

f£ t i mirr r d t

ntration rang . Whil thi

in th fre RN

and ugg t a d gre

f p

rv d

ifi it

- M7 hara t rizati n

Inhibit r

However in an art mJ t t rem

lig

M2 with a i ibl mer a

111

am c nc ntrati n ran g , it i cl ar M i farm r active

r th

fracti n

r th

f th b und fracti n f

e id nc d by the gradual di app aran
con

- P inhibit r, a

e the

f~ ct bey

nd non- p cific int racti n.

d in parallel under id ntical c ndition .

a al

mple entirely, th c ncentrati n rang

R - P-

of M 7 wa e t nd d to 4000-fold (4 72 nM) a can be e n in figure 2.22 .

Mo lar-fold:

-

+

Bou.nd [
R:'\

nbound [

RNA

(KRAS-A)

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

Figure 2.22. EMSA cha r acterizing CRD -BP-KRAS-A RNA compl ex in hibition by AON32
SM7. [ P] -labelled RNA- region KRAS-A probe (unbound in lane 1) was incubated with 540

nM CRD-BP (lane 2). Inhibitor SM7 was present (lanes 3-14) in otherwise identical conditions
to lane 2 ranging in concentration from lX molar-fold relative to probe concentration (118 pM)
to 4000 molar-fold (472 nM) which con-elated with a loss ofCRD-BP complexed bound RNA.
AON SM2 , in contrast displayed reduced ability to inten-upt the CRD-BP-KRA -A RNA
complex .

As can be seen in the SM7 inhibition assay (figure 2.22), a simi lar trend occurred with
SM7 as with SM6. Both inhibitors show a reduction of the bound RNA fraction and a
proportional increase in unbound RNA, indicative of a lo
RNA complex.

of stability of the CRD-BP-KRA

nee again the SM2 oligonucleotide displayed a imilar pattern a in figure 2.2 1,

50

ith n a ti ity again t

P .D 1mati n . Pu hing th u p r limit f th inhibit r c n

rang to 4000-:fi ld pr due d n1y a marginal furth r r du ti 11 in th b und RN
can be e n t b

wa complet d u ing th ph
up t
oftware a

thi

ffi ti e far

I0 0

ph imag r
n

(figur 2.2 ).

90

AO. -_ ~17

AO. ~- ~16
80

80

70

70

c:
::J
-o

60

JJ

so

Q.e

1J

c:

::J
0
JJ
Q.e

60
50

40

40

30

30
20

20
0

A

u1n rical data D r b th

ft ar .

ntrati n v a pi tt d gr phi ally u ing Kal ida raph ™

90

1J

fra ti nand

n it m try analy i of

nc ntr ti n lane 11 -1

tern and ptiquant

ntrati 11

800
200
400
600
S!\16 .\lolar-fold excess

1000

0

B

800
600
S?\17 .\lolar- fold excess

2(}0

400

1000

Figure 2.23. Inhibition plot of AON SM6 (A) and SM7 (B) against CRD-BP-KRAS-A RNA
complex formation. Graphs represent quantified data from densitometry analy i of
electrophoretic mobility hift assays with inhibitor oligonucleotide concentrations ranging from
118 pM at lX molar-fold to 118 nM at 1000X molar-fold excess. [32 P]-labelled KRA -A probe
was 118 pM with CRD-BP protein maintained at 540 nM.

Graphical presentation more clearly depicts a tr nd whereby both AON M6 and SM7
inhibitors show effective do e-dependent abrogation of the RNA-protein interaction; howe er
SM7 is more active at low r concentrations (I

50 of approximately 1OOX or 1.18 nM for

M 7,

51
rall d

c mpared t 400X r 4.72 nM £ r M ) and i
olig nu 1 tid

RN

f th

t.

111

n ther 1 kat th inhibit r in parall 1 ith

£ !ding f KRA - with n~j( ld lace M6 (lab 11

ibl impli

rmp rtant £ r

in th

RD-BP r c gniti n (fi gur 2.2 ).

r gion i indi cated in r d - th

EM A mapping ( ecti n 2 .. 2), and
Thi

mall

t

triking re ult complem nt th pre i u

conclu i n that the KRA direct contact with the

ting th

n

1 p tru tur

a

4(b)

frag1nent identifi ed tlu· ugh

ith b th th

M6 and M 7 inhibit r .

affinity m apping data, and reinforc

4(b) r gion i important for KRA mRN

the

binding and likely i

RD-BP pr t in, at lea tin the e in vitro a ay . Mongr

howed that th KRA mRN

N

mput r imulat d

arne fi gur , th K

R - P-binding

erl ap can be

ingl

in blu ) and M7 (lab 11 d in bla ck) a

compl m ntary t adjacent t m-1 p tru ture p

RN

t f[i c ti

d th m

li1

e/ al. (20 1 l )

coding region dir ctl y interact w ith IMP - I (Hum an rtholog f

CRD-BP) in ivo thr ugh UV cro s-linking and ba ed on the de cribed experim ents here, that
interaction likely occur in the KRA -A34(b) region betw een nu cleotid es 129- 185 relative to th e
start codon. To explain the reduction in overall affinity of this smaller 57 nucleotide region
compared to the larger KRAS -A (1 85 nt), it must be con idered that CRD-BP po sess multipl e
domains that each contribute to an overall Kd for a given RNA ub trate, as is theca e for
hnRNP K and FMRl where multiple KH-domains are required for a stabl e interaction with RNA
(Siomi et al, 1994). Further emphasis on the requirement for multiple KH-domain interactions
comes from experunents involving IMP -1, the Human ortholog ofCRD-BP , where all four KHdomains were shown to be necessary for proper cytoplasmic traffi cking and RNA binding
(Nielsen et al. 2002) . As a result of this property, it is pos ible that the smalle t equence
identified capable of forming a complex with the protein - KRA -A34(b) may only b

52

nteracting

ith a ub- t of th R

1a e r m

d ther lo

r-affinit

-binding K
it

domain . and th at th truncati n pr ce

~'/
.u·

U•

I

?

00

I

I
'

,.

llecti ely c ntribute t binding.

that w uld n rmall

I

ma

I

\

•

•I

~

g - u-

I
I

Region KRAS - A34(b)

I

\

I

.o

\

s

6

SM7
\ ... t

I
Figure 2 .24. Computer simulated RNA foldin g of KRA S-A RNA w ith KRA -A3 -' (b ) and
AON inhibitors SM6 and SM 7 target region s indicated. RNA fo ldi ng of KRA ~ - (nt 1-185)

wa co mpl ted u in g the mfold we b serve r: the prese nted tructu r hav ing a pred icted free
~n e rgy of d = -3 8.63 [initi all y -46.91 Kcal/mol.

52

inter ting with a ub- t of th R
ha e rem

d ther low r-affinit

-binding KI l d m in , nd that the truncati n proces rna
ite that

ould n rmall c llectivel

1

ontribute to binding.

_,r
I

I

t

.u·

U•

~

-

'

-/'

u-u- \

-u'-- \ .- \- _,.\ _ ~,~ ... ,-·
,
l _ u ...
"'

oo

I

I

,

D

-------u

r /
I

t

I

I

a
I

,

Region KRAS -A34(b)

s
SM7

6

I

,

Figure 2.24. C omputer simula ted RNA foldin g of KRA -A RNA wit h KRAS-A3 4(b) and
AO N inhibitors SM6 and SM7 ta rget r egions indicated. RNA folding of KRA - (nts 1-185)
was completed u in g the mfold web se rver~ the pre ented structure having a predicted free
energy of dG = -3 8.63 [initia ll y -46 .9] Kcal/mol.

53

2.3.4 KRA RN minimum bindin
Ph

i al

qu nee analy i

parati n f the KH d main and their pr ci e g

play a critical rol in th

ele ti n of RN

targ t

a well

m try relati

nucle tid

archit cture d e n tall w [! r rec gniti n f 1 ng equ nc

to n anoth r

equence. KH-d main

but in t ad direct int raction

betwe n an RNA ub trate and an indi idual d main i limit d to only to 4 or 5 nucleotide and
i of w ak affinity ( ha

tal. 201 0). T try and identify om

f the equence that may be

ilnportant in facilitating thi interacti n, a earch for putative RNA-binding equences for the
VICKZ prot in family wa undertaken and local sequence aligrnnent were completed u ing
previou ly identified con en us equences and the putative KRAS 1nRN A binding region. While
there is little agreement regarding a precise con ensu binding equence for thi family of
protein , with different experiments returning largely different conclu ions a few intere ting
results are mentioned here that positions KRA -A34(b) in relation to the available literature.
VglRBP is a closely related ortholog ofCRD-BP who e known mRNA targets : VLE and
V gJ, contain a sequencing clustering of 5 '-UUUCUA-3' or 5 '-UUCAC-3' (Git and

tandart,

2002). In KRAS-A34(b) RNA, a sequence very similar to those aforementioned, 5 '-UU

C-3 ·,

appears in close proximity to a repeat which runs in the opposite direction, separated by only 5
nucleotides. Strikingly Du et al. (2008) suggest that a minimum of exactly 5 nucleotide between
binding-sequences could fill the gap between KH-domains and i required for efficient binding
for hnRNP and P B l/2, which are related RNA-binding protein that contain only three KH
domains . Furthen11ore, the RNA target reported for hnRNP and P B 112 r quire a minimum of 2

54

r p at

qu nc : 5'

fa 19 nucl

tid

- ' equ n

found in KRA -

A
rtantly in d

the ambiguity f binding-

qu ence rd r

a!. (20 10) with nly th

rall p la1it r lati

Thi group al

d t nnin d fr m

deg 11 rat RN

library

length. h

e

X) fa

rand m mutati n intr due d int th targ t "zipcode"
d ub - qu n
e

: 5'-

-"''and 5'-A A-3.

the p1i r identified equence wi thin its 57 nu cleotide

r th latt r eq uenc appear d twice in partial v rlap a 5 '-A

not urpri in g gi

11 the

hao 1

t th KH -d main being an imp rtant factor.

f el ction and amplificati n (

equence, that ZBP 1 recogniz d tw
KRA -A34(b) regi n d e n t p

el ping am del for binding,

ithin a tran cript ha been d mon trat d by

r und

11 1 ting

'which c ntain th 5 '-

impli ity ofthe 5'-

A-3' which i

A-3' sequence and it likelih o d of appearing by

chance. How ever, later w rk with ZBP 1 completed by Patel eta!. (20 12), focu ing on the ZBPl
KH3 -4 didomain only, eparately determined imilar binding elements except for a lightl y
lengthier 3' bipartite sequ ence (5 · -ACAC-3 ') and thi s still matche th at found in the KRASA34(b) region. Additionally, u sing 1H- 15N H QC NMR it was also reported b y Patel that the
5 '-ACAC-3' motif interacts with th e IMP - 1 KH 3 domain as evid enced by resonance hifts
within the residues of the KH3 domain when free protein wa compl exed with this small
oligonucleotide. Pairing RNA sequ ence with a specific KH d01nain is useful informati n for
constructing a model of KRAS RNA binding.
PAR-CLIP (photoactivateable-ribonucleoside-enhanced crosslinking and
imrnunoprecipitation) experiments, where UV-reacti ve nucleosides are add ed to cell culture to
be incorporated into nlRNA transcripts, as well as other RNA molecules, revealed that IGF2BP 1
(aka IMP-1) co-precipitates with cross- linked RNA sequ ence: 5'- AUH-3' (where H=A,C or
U). This was in fact the only ingle consensus sequ ence found , from thi

tudy, however the

55

actual cro -link d r idu
mTing within a

nt range i1nplying

ntain thi 5 ' pr

i u ly t ha

m tif in th

H-

- · th at

ibility (Hafner t a/, 201 0) . KRA -

4(b) d e

r, th

I ightl larg r KRA - 24, m ntioned

ur

( cti n 2 .. 2), doe c ntain thi

nc at nucl

tid p

iti n 79 thu

ffering a

qu nc t r an up tr am binding equenc hyp the iz din ecti on 2 .3.2 to work
ther pre i u ly m nti ned

KRA - 12 which in Jud e th
m to bind

th r 10%

affinity imilar t th full! ngth K

coop rativ ly with th

not

me fl

m tif. H

~ rm of 5 ' -

p t ntial RRE

with th

ccurr d within thi m tif nly 90% fth tim

ithin K

n th at ha th

RD-BP, it i thu not uffic i nt n it

requir informati on in the KRA -

a~

4(b). Keeping in mind that

rementi ned 5 '-

H-3' motif, do

wn a a binding m tif and seems t

4 regi n, which doe bind

RD-BP, alb eit with reduced

affinity.
ielsen et al. ( 1999) propo ed a ingle putative binding sequ ence fo r IMP - 1: 5 'UUCACGUUCAC- 3 · ba ed on EM A and

V -crosslinking methods with mRNA target IGF- II.

By aligning DNA equivalents of KRA -A34 with the propo ed I F-II binding equence, a
region of highest imilarity can be een and this fa ll within the KRAS -A34(b) region. The RNA
equivalent is 5 '-UUCUCGACAC -3' in the KRA -A 34(b) region. A propo ed binding sequ ence
for ZBP 1 is 5 '-ACACCC-3' (Ross et a!. 1997) and a similar alignment pl aced the region of
possible irnpm1ance within the sam e range. Po1y-C and poly-U pyrimidine homopolymers
compete significantly better than poly-A and poly-G purine homopolymer aga inst VLE mRNA
with V g lRBP (Git& Standat1 2002). Transcripts with more C/U content hould therefore bind
better, however the total composition of A3 4(b) doe not agree with high
/U and 56% AG . However binding a says with poly-

and high U; 44%

h mopolym er do how that the binding

mode with these oligonucleotides is different, being additi ve and competing away the a tual

56
LE targ t
whi h bind

quenc in a lin ar n n-

fa hi n a

pp

ed t th target VL

operati

Perhap it i n t

qu n

p rati

n t fall und r

mu h th ra

rall

c nt nt, but in t ad a c rtain

KZ pr t 111

that aid in binding

equ n e r peat in KRA -

qu enc

tab!

I -1i h ub-3'

m nti n d 5'-

4(b) a we ll a th 5'-

- ' ma fall in -line with thi

idea.

A

lignmen t f hvo D

CATGA

9

-- -- -- - - - --------------------------------------------------------------TTCACGTTCAC------------

94

A T ~TGATCC

·CAA ~GAGGA -T CC - AC GGAAGC

sequence

GTAGT AT-GATGGAGAAACCTGICTCTTGG~TATTC CG

II

-CACAGC~GG

I

·equence l
A A GA CC.A.ACAA AGAGGA- CC'ACAGGAAGCAAG.,-AG AA"GATGGAGAAACC GTC - C,....-GGA-A

c - CGA-CJ

Sequence 2
-

-------------ICACG

B

CAC--

Alignment of h\·o DNA sequences

AT TGA CCAA.CAA A.Gt<GGA T CCTACAGGAAGCAAGTAG AAr

GA GGAGAAACCTG CTC TT GGATt..TTCTCGACACAGU..GG CATGA

II

--------------------------------------------------------------------------- - -A.CACCC----------

?3

93

Sequence 1
ATA GATCCAACAATAGAGGA I CCTACAGGAAGCAAG AGIAAT GA- GGAGAAACc - G- CTCIGGA A--c CGACA

Sequence 2
-------------------------------ACACCC----

Figure 2.25. Local DNA alignment of KRAS-A34 and putative bindin g sequences for IMPI and ZBPI CRD-BP orthologs. (A) lMP- 1proposed binding equence and aligned with
KRAS -A3 4 and (B) ZBPl proposed binding equence aligned with KRAS-A 4. DN
equenc
were converted from their RNA sequ ence origins forth purpo e of the aligm11ent, and both
hi ghlight an overlapping pos ibl e binding region.

57

Chapter 3
Assessing the role of CRD-BP KH domains involved in
KRAS mRNA binding in vitro and in cells.
Thi
indi idual

hapt r utlin

th

perim nt und rtaken t d t rmin

RD-BP KH d mam in binding K

el ctroph r tic m bility hift a ay ( M
ariant . Th

f each f th .G ur

cond

cti n ~ cu e

xp riment u ing p int mutation

RD -BP pr t in

n d t rmining th imp rtance f KH d main in cell ,

u ing RNA irnrnuno-pr cipitati n c upl d with quantitati e rever e tran cripta

P R (RT-

qP R).
3.1 Methodology- CRD-BP KH domain requirements for bindin g KRA

mRNA in vitro.

3.1.1 Purification of CRD -BP KH domain va riants
CRD-BP variant w re previously generated by another member in Dr. Lee' lab u in g
site-directed mutagenesis again t the first glycine (to a partate) in each of the four G-X-X-G
motifs within each KH domain (G212D, G293D, G422D , G504D). The e motifs are critical for
the interaction of the KH domain with target mRNA, and by mutating each individually a well
as in tandem as double-KH mutants, individual KH domain importance in binding KRAS mRNA
cou ld be assessed. Proteins harbouring a specific single mutation were labelled as ''KH 1"
through "KH4" depending on the KH domain affected, with double mutant harbouring two
affected domains labe ll ed as "KH 1-2" through "KH3-4 "· information ummarized in Table 3.1.

58

Table 3.1. ummary of enerated RD-BP KH domain variant . KH d mam are
indi idu all af:D t d a a r ult f it -dir ct d mutati n in th can ni cal - - - m tif.
CRD-BP mutant

KH domain mutated
KHl KH2 KH3 KH4
•

K.Hl
KH2
KH3
KH4
KHl-2
KHl-3
KHl-4
KH2-3
KH2-4
KH3-4

•
•

•
•
•
•

•
•
•
•

•

•

•

•

•

E . oli BL2 1 (D 3) cell w re heat- h ck tran fonn ed with p T2 b- RD-BP variant
pla mid and purifi ed u ing a
200 mM

i- T

agar

e c Iumn D llowed by dialy i into th e fin al buffer:

aCl 20 mM Tri - 1, 10% glycero l, 0.0 1% Trit n-X, pH 7.4 as described in section

2.1.1. Quantification of each protein ample wa compl t d using a B A Protein As ay Kit
(Thenno Scientific) with

RD -BP protein samples typically ranging between 3 and 7 J.lM.

3.1.2 Generation of

eP]-lab ell ed RN A sub strates by IVT
2

Experiments with CRD-BP KH domain variants w ere completed using the KRAS -A
RNA fragment (nts 1-185) detennined to possess strong binding affinity (section 2.3 .2), a well
as the coding region dete1minant region of c-myc mRN A (nts 1705 -1886), an established mRN A
target of CRD-BP (Prokipcak et al. 1994) . T7 -driven in vitro transcripti on of both [32 P ]-labell ed
RNA regions from eDNA templates was completed as described previou ly in ection 2.1.5.

3.1.3 EMSA experiments with CRD-BP KH domain vari ants
MSA binding reactions were compl et d for each generated [32 P]-labe1led KRA
myc sub-region and contained

and c-

RD-BP KH-mutant concentration varied from 0 to 540 nM.

MSA reactions used for KH domain binding a

ment contained KRAS or -m c prob at

11

59

20 ,000 pm. Prot in-RN
re o l

mi ture w re incubated at 7 °

d on an % p lya r lrunid (29 : 1

TB

running buffer and ubj ct d t

t

cti n 2.1. .

pti uant

Bi -rad) n n-denaturin g g 1 in 0.5 x

crylamid

:D r 0 minut

at 25 1nA in a in1ilar fa hi n

iati n c n tant wer calcul ated ba d n data acquir d from

d n it m try analy i
and

:D r n ha lf h ur and c mpl

f th

M

aut radi graph u ing the

y 1n

torage Ph

ph r- y t m

ftw ar .

3.2 M ethodology -

RD -BP KH domain requirement for KRA mRNA bindin g in cell .

3.2.1 Preparation of IP-RNA from HeLa cell
Pr parati n of purifi ed imn1un precipitated (IP) RNA wa part f previ us pr j ect and
wa already on hand for the d

crib ed experim ent · a cond en ed description of m eth d u ed to

acquire the IP-RNA follow . CRD-BP KH dom ain mutant vari ant as de crib ed in th e previo u
ection (3 .1.1) w ere ub -cloned into a pcD A m ammalian expre ion vector (+FLA ) and
transfected into HeLa cells u ing Lipofectamine 2 000 reagent (Invitrogen) follow ed by a 48 hour
incubation. Cells w ere lysed with total cellly i (T L) buffer in th e pre ence of RNa in and
protease inhibitor tablet (Roche diagnostic ) and lysate incub ated with anti -FLAG antib dy
overnight at 4 °C . Immunoprecipitation (IP) of the different CRD-BP variants containing point
mutations in the GXXG motif (single and double mutants) using Protein G-agarose bead was
p erform ed and equal amounts of irmnune-precipitated CRD-BP w a recovered as detennined by
SDS -PAGE, follo wed by treatment with Proteinase Kanda final pheno l-chl oroform-isoamyl
alcohol extraction. IP-RNA was qu antified u ing a Nanodrop Spectrophotometer (Thermo
Scientific). RT-qPCR to m easure the relative am ounts ofKRA
the variants as compared to a w ild-type control.

mRN

associated with each of

60

3.2.2 cD

nthe i from IP-RN

11 ur

n

ntamin ti ninth RT-qP R r

ith th

pr
mp l

d. ir tl t

pr t in granul

h 1 ~g f I

-pr

a a a i1 able), 1 ~~

mad u

10 1-d r a ti n v lum

f

a

.

1(

a

a

incubati n p ri d ~ 11 w d . D acti ated rea ti n wer

drawn ff fi r u
ynth

and 10 ~I

in cD

yn th

i of

r m

f th

ha
a l kit

m

1 n) w a

KH

ari ant

mp l

hi ch n1 y 1 bi 1 gi a1
b u f~ r and l 2

) ( m i n), l 1-tl 1

re incub ated at 7 ° [! r 0 min ~ 11 wed by

. R a ti n

1 m 1 ul

"

ipitated w ith a h

th additi n f 2 ~1 f rD a e deacti ating reagent a ft r

dim nt rD a

-fr

hi h m

pt KH 2 and KH , f

(all ingl an d ubl mut nt in tri li
r pli at

th

'

ti 11

hi h a 2 min r

m temp rature

pun at I 0,000 g [! r 1.5 mint

up ern atant c ntaining th e purifi ed IP -R A wa

i pri r t qP R .

mpl m entary D A fr m purifi ed IP -RN

ampl e wa ace mpli hed

u ing a Biorad i cript cD A ynthe i kit; 20 ~~ reacti on prepared w ith 4 ~~ 5x i cript buffer
mix containing rand om hexam er prim er , 5 ~1 nu clea e-fr e H 2

10 ~I of D

-fre IP -RNA,

and 1 ~1 rever e tran cripta e (RTa e) a embled pecificall y in th at rd er. A

embl ed reactio n

,

w ere placed in a thermocycler and ubj ected to incub ati on at 25 ° [! r 5 min (rand m hexa m er
primer annealing), 42 °C fo r one half hour (eDNA ynthe i ), 85 °
inactiv ation) follo wed b y 4

°

torage until u e.

fo r 5 min (RTa e

a control, each KH vari ant IP amp le wa

also ubj ected to the arne procedure w ithout the addition f r ver e tran cripta e (- RT/no
template) to account for the po ibility o f genomi c DNA tability.

3.2.3 KRA mRNA RT -qP R
uantitative P R w a u ed t m ea ure relati ve amount
the

f KRA

mRN

a o iat d ~ ith

RD -BP vari ant t determin in ol em ent/ imp rtan c f indi idua l KH domain . Fir tl ,

61

KRA

ifi

mplifi ati n nm r d

ding

qu n

nth

t am lify a 12 nu

ign

ized b lnt gr ated

t

tid r gi n f

hn 1 g1c In . (l

a

), a

mpl m ntar TaqMan pr
q
ba d n d

rib d

nt

ith an r ati

g n tr n

Table 3.2. KRA

ripti n (

R prim r

nd

qM an pr b ( abl e .2) wer

ancer c lllin

ting

li g nucl e tid

g i I a/. 2 1 ).

RT -qP R amplificati n prim r an d

aqMa n prob

Primer/probe
Forward primer:
KRA TaqMan probe:

uantitati

P R amp!

were etup a 25 ~1 reac ti on and c mpleted in tripli ca te [! r

each bi l gical repli at t tall y 9 reacti n p r

RD -BP ari ant.

a ingle biolog i a! replicate wa a ail able for KH 2 and KH 3
3 valu e w re obtained fo rth

e.

pr v i u ly m nti ned , nl y

RD -BP va ri ant and a

uch, nJ y

dditi nall y, D r each RIP ampl e a no RT co ntr I wa

m ea ured in an otherwi e id enti cal fa hi on to en ure no gen omi c D

contaminati on . In a 96

w ell P R microplate (Axygen cientifi c Inc .) rea ti on were a em bled according to tabl e .3
from an iQ™

upermi

kit (Bi o-R ad ). In an attempt to redu ce pipetting error and inter- amp le

variability, all qP R exp eriment were completed by fir t creating a m a ter-mix th at contain d
all reagent except the cDN

. T o each ample well of th e P R micro pl ate, 24 111 of ma tcr mix

w a add ed, with the final 1 111 eDNA vo lum e add ed la t.

11 wing additi on of all reagent , th

u ed P R mi c ropl ate w 11 w re ea l d w ith ealing film uitabl e for rea l time P R ( x g n
c ientifi c Inc). A s a final te p, th e P R mi cropl ate w ere c ntrifuged at 200 g for 1 minute at
ro m temp eratur t en ure pr per mi ing.

o il wing rea ti n a emb l , qP R wa perf nncd

62
d u ing i 5 ptical
y t m

ftwar (Bi -Rad) .

Table 3.3. Reagent u ed for KRA qP R r ea ction . I up nni i fr m th
i -Rad i TM
up 1mi kit and contain all n c
buf:D r ingr di ent d P and
ym . 11 r acti n
w re a mbled in triplicat D r a h
- P mutant ariant and mi d a a rna t r mi prior to
adding final D
qPCR reagent
Forward primer (lOJ.tM)

Volume u ed per qPCR reaction

Rever e primer (1 0 J.tM)

0.5 J-tl
0.5 J-tl

KRA TaqMan probe (5 J.tl\1)
IQ™ uperm1 (2X)
eDNA

H20
Total volume

0.5 J-tl

12.5 J-tl

1 J-tl
10 J-tl
25 J-tl

3.3 Results and Di cussion - KH domain requirements for bindin g KRA mRNA
3.3.1 CRD-BP in gle KH domain va riant EM A bindin g assays
To inve tigate the functional rol e of each individual KH domain in binding KRA
mRNA, CRD-BP variant harboring mutation in each GXXG motif of the four KH domain
were used in EMSA binding a ays with the high affinity KRAS sub-region A ( ection 2.3.2)
[

32

P]-labelled probe. EMSA containing the KH mutant and KRAS-A probe was can·ied out with

all single and double mutants beginning with KHl through KH4 single mutant variant and were
performed in a manner identical to chapter 2 describing KRAS mRNA affinity mapping. Re ult
from the ingle KH mutant displayed a stark contrast in binding capacity when compared to
wild-type CRD-BP (figure 3.1). Beginning with figure 3.1A, it can be seen that the wild-type
RD-BP fully shifts the free KRAS -A RNA probe within the experimental protein cone ntration
ran ge of 0-540 nM (lanes 1-6). How ever, with a single GXXG motif point mutation in either
KH 1 or KH2 (lanes 7- 11 and 12- 16 respectively) , the binding pattern change con iderably, both

63
mutant di playing limit d mobility hift f th fr
ha

uniqu binding patt rn · KHl r main capab le ffl1ming a p cific mol cular weight

compl

of th

am

1z a wild-typ , albeit at a reduc d quantity figure .1

while KH2 app ar t ha
b

a c mplet ab ence f thi band.

n n ar th t p fth gel fl r KH2 (figur

KHl Ian

a

. lnt re tingly KH 1 and KH2 b th

but till pre ent in th

ed for KRA -

RN

found. A with KHl a

.1 , lan

faint , much larg r complex can

12-1 ) a a 1n ar that i ab nt in the

iJd -type hift patt m . RD-B variant KH3 and KH4 were

binding capacity in th
X

lane 7-11 ),

am

perim nt and imilar r ult were

mutati n in the KH d main yield d dra ti cally reduced binding

ability with only an e tremely faint trace of the pecific compl x that is so prominent in the
wild-type er ion of RD-BP , and to a 1

r ex tent the KHl mutant variant (figure 3. 1B, lanes

7-11). Higher molecular weight mearing wa al o not a prominent when compared to wildtype. KH4 domain mutation in CRD -BP produced a very clear re ult with no hint of any
significant binding whatsoever, neither a a pecific complex band , nor a a higher molecular
weight smearing. This high molecular weight mearing observed may sugge t partial binding of
the protein to KRAS RNA.
Due to the striking reduction in KRAS-A probe binding affinity when any single CRDBP KH domain GXXG motif is mutated, we wondered if perhap these mutant protein amples
had somehow mis-folded or were otherwise producing artifact data . To test this possibility, an
additional set of EMSA experiments were completed utilizing a positive control where the
binding pattern with CRD-BP KH mutants was already known. c-myc mRNA contains a region
known as the coding-region instability detenninant (CRD) , and is a well establi hed RNA targ t
sequence of RD-BP (Ben1stein eta!. 1992; Prokipcak ct a!. 1994; Leed eta!. 1997).

64

\\ T

h..lll

I RO- RP jn. 1:

Fr

R:"A

[

lane

B
I RD -BPj n\1 :

Round
R;\ \

Free

R~\

[

lane:

1

3

4

F igure 3.1. E M SA assessin g CRD-BP KH va ri a nt with single poin t mutation for bind ing to
sub-region KRAS-A RNA; (A) KH I a nd KH2 variants, (B) KH3 and KH4 va ri a n ts. C2 P]labelled unbound KRAS -A (nts 1-1 85). RNA probe (2 0,000 cpm/reaction) with no protein added
(lanes 1, A and B) fonns a larger molecul ar weight complex upon titrati on with WT CRD-BP
and single KH vmiants po sessing GXXG motif mutation (first glycine to aspartate), a indicated
by the bound RNA fraction (lanes 2- 16, A and B) .
This RNA probe consisting of 182 nucleotide , was synthesized by IVT and wa u ed in
additional EMS As (figure 3 .2). Finally, it is worth noting that higher radi oactivity (80,000
cpm/reaction) was used in this latter et of EMSA experiment in ord er to increa e en itivity to
complex formation, and the re ults can be

en a the pr ence of a weak KRA -CRD-BP

complex fonn ation in all four KH mutant sa mple . This is in co ntra t to the previou set of

65

M

nl K.Hl £ 1m d a i ibl band f n tabl int n ity and wa abl t

(figur

r due th amount f fre RN . H w
hift d RN
nee
M
data

int n it

erall patt 111 fKHl pr du ing th

till remain un hang d, al

u gge ting that th K 1 d m ain ha a

ary (£ r full binding) but reduc d r 1 in binding the KRA p riment inc rp rating th
hich

a

ry imilar to th pr

-my RN
1

u

a ap

ub-r gi n fragm ent.

c ntr 1 (fi gur

iti

et f in gle KH mutant

pro id d om compari

n r garding the binding m de f KRA RN

RNA. Fir tl y, it can b

en that

ingle mutation in an

n KH-dom ain fa iling t

Furtherm r ,
as a dimer with two

trong t

.2,

- ) yield d

RD - P variant , but al
a c mpar d to c-m yc

RD -BP di pl ay r dund ancy in binding c-m yc mRN

with a

tr ngly affect verall binding.

RD -BP app ar t bind c-myc (and the KH mutant vari ant ) cooperatively
parate Kd valu e a evid ence by two distinc t band clea rl y v i ibl e at 108

riM (figure 3.1 A -D lane 8). First, RD -BP binding likely occur a a single protein c mpl x,
with the dimer formin g ub equently. On the other hand , KRA -A RNA does not appear to form
a strong complex with a single CRD -BP m olecul e, with onl y an extrem ely faint interm edi ate
band visible in the wild -typ e lanes (fi gure 3.1 A -D , lane 2- 6). The c-m yc and KRAS RNA
formed a dominant upp er compl ex w ith simil ar migra ti on distance, and th e RNA m olecules
b eing of similar size (1 82 and 185 nucleotides respectively), hence it is also likely that KRAS-A
RNA bind s almost exclusively in the dimerized fmm with 2 CRD -BP molecul es attached in one
complex. It could be that an initial binding event with a single CRD -BP m olecule indu ces an
RNA conformation change that produces or reveals a second

RD -BP binding site of very high

affinity, which is immedi ately bound . In tllis induced fit model, the initial binding ev nt forming
complex 1 (one molecul e bound) is rate-limiting and the second

RD -BP m olecul e (comple 2)

bind s almost imultaneou ly. Thi s id ea is in parti al contras t to a pr viou ly propo ed binding

66

dim rizati n n th
I F-II mRN

targ t,

quential fl rmati n f c mpl
mpl

target f

II. In thi

u c

cifi c and a r lati

ful ~ rm ati on f

. If thi m d l i ind

mpl

1) and a much 1 w r, cl arl y

am m d 1 put ~ rw ard by

1 i c n idered n n-

intermediat , and it i th
tabl e RNP compl

nt (

ith a fa t initial binding

d

mpl

i 1 en

t a/. (2004 ), th

ly 1 w tability pr cur

2 th at " I

r

k " th targ t mRN

int a

ub-r gi n i a b na fid

ITect then the K

RD-BP with uniqu binding haract r in that th dimeriza ti n ev nt occur either

imultan ou ly with th fl rmation f
po sibility i that up n an iru tial

mpl x l or at th

ery lea t, extrem ely rapidl y. An ther

-BP binding, not the RN , but rather the pr tein

experien e a dynamic confl nnati nal hift, a mecharu m ugge t d by W ang et a /. ( 1999) for
the binding of a rel ated KH domain c ntaining protein: Poly-

binding protein (PABP ) with

target a-globin mRNA . Thi change in fo lding character may then lead to the creati on f a
protein-protein interaction ite recognized by the second

RD-BP that can lead to the

dimerization event apparent in the EMSA fi gure presented here. In support of either model (i)
RNA or (ii) protein conformation change, exp eriment conducted to determine if IMP -1 can
dimeri ze independent of an RNA ubstrate indicate that RNA-binding i a prerequisite for
dimerization and that complex I precedes the formation of complex II (Niel en et al. 2004).

3.3.2 CRD-BP double KH domain mutant EMSA binding assays
Further work in assessing the pecific requirements of KH domain in binding KRA
mRNA was also completed u sing the EMSA. CRD -BP KH variants harbouring point mutations
at two GXXG motif sites located in separate KH domain

ere u ed in a simi lar fa hi on to

assess whether or not all traces of binding could be aboli hed. Indeed with the mutation of any
tw o KH domains, absolutely no binding wa obs rved (figure .3 A -D).

67

\\I

[( IW- IW [n\l :

Bound [
H '\ \

I rt•e
R'\ \

[

lanl

B

\\ I

hi I.!

[< RO- BJ> )u\1 :

Bound [
I{

\

,.. fl"C

R

\

[

Jan

RI>- IW [n\1 :

Bouud [
H '\ \

1 r l'

H

\

[
~

!.me

4

X

(l

')

Ill

II

12

I'

1-l

15

16

,..,

D
hll4
j( RI>- BI'[ n\1 :

Round [
R'\ \

1- n •l'
It'\ ,\

I an~

[
2

·'

4

5

(!

7

X

I)

10

II

12

I1

14

15

I C.

l.

Figure 3.2. EM A as essing CRD-BP KH variants with ingle point mutation for· binding to
KRA -A and c-myc RNA ; (A) KHI variant, (B) KH2 variant, ( ) KH3 va.-iant, (D) KH4
32
variant. [ P] -Iabelled unb und KRA - (nt 1- 1 5) and -m yc (nt 1705- 1 6) RN
( 0,000
c pm/ reacti n) with no protein add d (lane I and 7,

-D)

68

I R O-BI j o \1 :

Bound [

R \

l nbound [
R'\ \

Jan' .

B

\\ I

I"

16

15

IC>

1\.lll -4

I RO-UP j o'\ 1:

Bound [

R' \

KR. \ , [
(19. -J )

lane-

c
-..."v

').·

Bound [
R~A

":~

it-

'

,~
"y

KH 2-4
,..'>-..."...., ~

'). -

~~

'

,b

.....

ww
w

1

L nbound [
R"\ A

lane:

KH J-4
'l-

6

7

9

10

II

12

13

14

F igure 3.3. EMSA assessin g th e CRD-BP KH variant with two point mu tations for bindin g
to sub-region KRAS-A RNA; (A) KHI -2, KHI -3 (B) KHI-4, KH 2-3 (C) KH2-4 and KH3-4.
32
[ P] -labelled unbound KRAS -A (nts 1-185) RNA probe (8 0,000 cpm/reaction) with no protein
add ed (lanes 1 in all gel images) fom1s a larger molecular weight compl x upon titration with
WT RD-BP and doubl e KH variants posses ing GXXG motif mutations (fir t glycin to
aspartate), a indicated by the bound RNA fraction (lane 2- 16, A and B). Binding reaction were
incubated for a total of 20 minutes at 3 7 ° and 10 minutes on ic before being resol ed on an
8% polyacrylamide gel.

69

m pl te li1nina ti n f binding in the d ubl K l mutant
!ling

that KRA

mRN

int ra t primaril and lik 1

p regi n , and n t b an

th r n n-can ni al m

3.3.3 . KH domain r equirement for KRA

mR

RD-BP KH d main binding requir ment [! r K
thr ugh th u

[

ild-typ

nti - L

rib nucle pr t inc mple

precipitated with

hani m .

a al

a e ed in

upl d

ith R -qP R . H

II
a cell

KH mutant ariant a w II

antib dy \ a then u ed t immun e-pr cipitate th
mRN

-precipitated K

a

f th

ciated with ach

a c ntr 1 [! r ea h KH ariant ample a n

BP KH mutant wa th n analyz db RT-qP R .
rever e tran cripta

ntir ly thr ugh the

mRN

immun -pr c ipitati n (RIP m th d

n.

ide

bindin in H e a cell

pia mid c ntainin g
a a

p nm nt pr

M

(- RT/n t mplat ) amp! c n i ting nly f the purifi d mR

co-

RD -BP wa run und er identical qP R co nditi n . Thi co ntrol pr du ced n

detectable fluore c nee within the 40 cycle limit of the y tem and co nfirmed that no genomi c
D A c ntamination wa pre ent in any of th

amp!

te ted.

The KRA qP R primer were te ted for effi c iency u ing pla mid dilution c ntaining
2

KRAS eDNA template and d termined to be 97. %with an R of 0.99 . Data group
(figure 3.4) from RT-qP R ample were then analyzed u ing a

t alue

ne Way Analy i of Variance

(one-way AN OVA) te t and a <0.0001 p-value (a suming null hypothe i of all group being of
the same population) wa calculated with an F- tat value of 19 . 1, indicating a ignificant
difference between one or more of the analyzed group . From th rej
that all sample were equal r from the same population, a p
c mpari on te twa calculated fl reach

tion of null hypoth

t-hoc Tuke ' I I D multiple

RD-BP KH mutant pair, and re ult re

level ofKRA mRNA as ociated with wild-type

al d that

RD - P were ignificantl dif[i r nt than the

70

KI-f -mutant . Fw1h tm r

KHl KRA mRN

than th

RD -BP mutant (p=O.Ol ). Finally an unpair d t-t twa c ndu t d t

th r r maining

compar th

t- alu

betw

n

ere al

ild -type and ach ind i idu al

analy i al o bowed that all f th qP R data gr up (K
ignificantl diffi rent fr m

r lati

di ffi r nee in m

le el b tw

cal

b

ignifi cantly lli gh r

RD- P mutant ampl

. Thi

ingle nd d uble mutant

were

ild -typ at p=O.O1. ycle thr h ld alu

RD-BP and pl tt d n a 1 garithmi

typ

[t und t

wer n rm aliz d t wild -

hich wa r quired c nsidering the large

n wi ld-type and the d ubl e KH mutant (figure 3.5) .

--

u
~

0

..c
Vl

-~

-

-=
~

L

t.J

;,.....

u

JO

I

WT

Kil l

Kll ::!

KI D

K11 4 Kill ] KH IJ K l ll 4 KH1 3 KIL~-l K IU-l

C R O- BP vari ant

Figure 3.4. Box plot pf the RT -qPCR cycl e threshold valu es from immun e-precipitated
KRAS mRNA associated with CRD -BP KH variants. H01izontal middle line in the
encompassing box represents the overall m ean, the top box repre ents the third quartile and
bottom box the first quartile. Bars are the absolute minimum and m aximum Ct val ue obtain d.

71

*
1

*
<;(

z

01
p <. 0. l

~

E
(.f)

~

~

Cl>

0 01

>
·::::
«l
Cl>

~

0 00 1

0 0001
WT

H

H2

KH3

KH4

-2

1-3

1-4

2-3

2

3

CRO -BP KH -variant

Figure 3.5. R elative KRAS mRN A ass oci a ted with KH domain variants from immun e
precipitation of CRD-BP in H eL a cells as determin ed by RlP RT -qPC R . ach cycle
threshold value i from averaging 3 biological ampl e in triplicate and en or bar h re represent
the standard deviation, with the exception of KH2 and KH3 of whi ch only a single bi ological
replicate was available for RT-qPCR . One-way AN VA determined P<O.OOOl and po t-hoc
Tukey HSD and Student' s t-test revea led signifi cance between pairs.

To understand the molecular nature of how

RD-BP bind KRAS mRNA, experiment

in the form of EMSA gel shift assays and also RT -qPCR analysis were undertaken. Data from in
vitro EMSA experin1ents conelated well with the re ult from IP experiments u ing HeLa cell .
Together, the data presented in thi s chapter trongly upport a mod el where all four KH domains
of CRD-BP are required for KRAS mRN

to maximally bind

RD-BP . Mutation of th fir t

XXG motif glycine to aspmiate in of any on of the four KH domain

resulted in a near 1000-

fold reduction in pull-down KRAS mRNA, save for KHI. The conclu ion of a redu ed role for
KH 1, that wa initiall y ugge ted by the M A data (fi gure 3. 1A and 3.2A), i fu1iher echoed by

72

tb RT-qP R data .

mparing KH 1 ( t -

0.00 ) it can b

e n that KRA tnRN

1 vel ar

about 1 -fl ld 1 w r than wild-type ( t-27 . 10) 1

th

th r

appro imatel 1000- fl ld 1

M

and IP-qP R data upp rt the

n ti n that ind

r (fi gur

.5).

g th r, th

d all fl ur KH d m ain ar r quir d [! r

RD- P KH mutant are

- P t bind KRA mRN , with

p rhap a miti gat d but till nece ary r

fl r th KHl d main. T hi c nclu ion infl r that th

int racti n i highly p cific· [! ur RN

qu ence 1 ment

vid ntl y r quir d to contact each

KH domain fl r binding, and i in-lin wi th th lit rature n ti n that many RN -binding pr tein
achieve th ir p cifi cit through int racti n w ith multipl d m ain .
R ult by Bam e
tnRNA . B th mRN

t a/. (20 1 ) fo und

like KRA

imi lar R -qP R re ult with c-m yc and

D44

were fo und t di play redu ced as ociation with KH single

and double mutant v ruiant of RD-BP ( e cept KH4, which retained affinity [! r c-m yc ).
However in stark contra t to KRA data, EM SA CRD -BP binding result fo r these sam e
mRNAs (c-myc and CD44) di play an oppo ite result, with single KH mutations having no
effect on binding (except KH4 did not bind CD44 mRNA) . RNP granule f01mation in zebrafish
was also analyzed to search fo r a con elation with in vitro binding data, with inj ection of the
identical CRD-BP KH mutant' s mRNAs directl y into th e embryo. T hese in vivo res ult better
support the EMSA binding data, w ith the single KH mutant CRD-BP granules appearing in cell s
at the same rate as wild -typ e, and KH double mutants having much lower numbers of grruml e
(except for KHl- 2 and KH3-4; the CRD-BP didomains). KRAS mRNA , which did not bind
strongly with any single KH-mutant, ma y therefore not contribute signi ficantly to the CRD-BP
granule core structure, as the granules continue to form under mutant conditions that do not
allow KRAS mRNA to bind. While KRA S mRNA, at least in vitro and in HeLa cell doe not
have a pattern that conelate well with granul e form ation in zebrafi sh, it cellular rele an e i

73

not 1 t. KRA mRN
granul

and many oth r mRN

but may i1nply b pre nt in 1

may n t [! 1m th c r

r c py numb r

granul a t rminal a ociati n within th m h

tru tur

r inter-di p r d thr ugh ut th

rk n t [! nning the c r

Intere tingly a a id n t , the e z brafi h granul e data all w for the p
KH did mam p
[i rmati

ing n mutati n (eith r KH 1-2 r KH -4) i

n, but that mutati n in both (all th er K

f th

caffl ldin g tructur
ibility that an intact

uffici nt [! r granul e

d ubl mutant ) did mam may mo t

trongly impair functi n.
Thi conclu i n i

irnilarl

upp rted by

i 1 en

I a!. (2002) wh

h wed w ith IMP- 1

truncation tudi e (KH 1to KH4 , KH2 t KH4, and KH 1 t KH 3) that both intact KH did main
(KHl -2 and KH 3-4) are required fo r IMP - 1 to bind H l9 RN

in imilar M A studi

. Whil e

CRD-BP truncation were not a e ed fo r bind ing to KRA mRN A, the re ult do compl em ent
the IMP- l -Hl9 binding re ults in that all KH domain are required. The group also [! und that
only the wild-type and KHl to4 recombinant protein (no RRM domains) was able to fonn RNP
granules and exhibit lamellipodia anchoring as detennined by confo cal micro copy in NIH 3T3
cells (mouse embryo), unlike the Barnes et a!. (2 015 ) paper where the KH3to4 truncation could
still form granules. Given the evolutionary distance betw een zebrafi h and mice, the compared
KH3to4 data could be clues to a specie -sp ecific intracellular regulator of CRD -BP . The e two
groups also chose different amino-tetn1inus truncation position for the KH 3to4 construct;
Barnes et al. (2015): 409-577 and N ielsen eta!. (2002): 344-577, re pectively and perhap the
unique granule fonning characteristics li e in the differential re idue .

It has b en demon trated that in th Xenopus ortholog Vg lRBP , all fo ur KH domains are
required for high affinity binding to a 340 nucleotid e region of V g l mRNA, and in fa t the two
didomains (KHl t 2 and KH3to4) can act cooperatively, much like what appear to b the ca e

74

for KRA mRN

( it and tandart, 2002).

nth

th r hand, nly the KH t 4 did mam f

IMP-1 i r quired to bind th beta-a tin 54 nucle tid "z ip
th full-length ver i n
utilize

nly the tw

mRNA (Farina
in

ha

tal. 201 0) . imilarl in hicken embry

ut nn t -te1minal KH d main : K

t a!. 200

. While tudi

ith affinity imi lar to
, th Z P 1 1ih log

and KH4 t bind beta -a tin

pertaining p cifi call y t th KH -d main inv lvem nt

I KZ pr tein binding th ir ari u m

conclud d that difD r nt mRN

d ' r gion

targ t are limi ted, it can alr ad y be c nfid ently

reqmr uniqu e KH-d main combination , with KRA

appearing to requir all D ur d main t achi e e a high affini ty interacti n.

75

Chapter 4
Assessing the structural impact on GXXG mutations in KH
domains on CRD-BP, conformational change upon RNA
binding, and KH1to4 crystallization
p rim ntal chap t r, the

m lecule i

ubject d t multiple

. T h e ffe t mutati n f th e ·

" m ti f ha

n verall gl bal

In thi final
tru
tru tur

f

RD -BP i in e ti ga ted by m an

aforem enti n d ingl and d ubi KH mutant
wild-type pecu·a.

ing

f ir ul ar dichr i m ( D) p ctr

condary

c p y· the

R -BP ari ant an alyzed and c mpared to a

D p ctro c py in a imilar fa hion, the po ibility of RD -BP

undergoing ignifi ant conformati nal change up n RN
what effect said binding m ay have on overall

binding i al o addre ed, a well as

RD -BP therm al tability. Finall y, attempt to

produce a cry tal tructure of RD -BP areal o detail ed and cover vari ou a pect of the process
including pla mid ub-cloning recombinant protein expre ion, and the multiple purifi ca ti on
technique emplo yed. While no diffraction data wa ultimately achieved fr om a CRD -BP pro tein
crystal, much was learned that m ay h elp pave the way to eventual success with this particular
ongoing proj ect; the finding appearing at the end of thi s chapter.

4.1 Methodology - Multipl e stru ctu ral analyses of the CRD -BP protein-RNA interaction
4.1.1 Produ cti on of CRD-BP KH variants
CRD-BP mutants harbouring the id entical

XXG mutations previou ly descri bed in

section 3.3 .1 and 3.3 .2 (same clones), were u ed for structural analysis. D e-natured amples
stored in 8 M urea at -8 0 ° were dialyzed again t refoldin g buffe r: 200 1nM Na L 20 mM Tris-

l, 10% glycerol, 0.01 % Triton-X I 00, pH 7.4 as de cribed in ection 2.1. 1 however an
additional di aly is buffer exchange tep was includ d to optimize CD pectra data acqui iti n, as

76

uruqu buf[i r
uf:D r :D r

n traint
pr t in

p
m

uantifi ati n f a h pr t in
ith

4.1.2

whi hi kn

nt

R -BP pr t in ampl

i all rangin g bet

a

length pe tra

ann d in th far
protein ample

I, 10% glyc r 1, pH 7.4

ignal.

n

Pr t m
n

nm) . inal

a Kit ( henn

and 7 J.lM .

ircular dichroi m p ctra

larim tr

canmn g

a p r[i nn d n wild -typ

w 11 a th KH ingl and d ubi mutant u ing a Ja c J- 15

11

- 2

ntribut it

mpl \ a c mpl t d u ing a

ircul ar dichr i m p ctr

Inc.).

1

anning in th [; r-

f 2 0 mM ri - 1, 20 mM

and la ked th additi n f rit n- 1

ntifi )

tral

pectr p larim ter (J a c

r mea ured in a .1-cm path-length quartz cu

rang (260 - I 5 nm ) to a

er dilut d int

tte; ampl e

a- heli and 0- h et c ntent.

D bu ffer (20 mM Tri - l, 2 0 mM

7.4) to betw een 1 and 2 J.lM ace rding to th B

R - Pa

R - P

a 1 10% glycer 1, pH

protein a ay (Th ermo i cher cientifi c Inc.)

and canned at a con tant temp rature of 25 ° maintained by a Pelti er t mperature c ntroll r. A
blank reading coni ting of D bu ffer onl y wa

ubtracted fr m each pectrum. Ja co J-8 15 D

pectra acqui ition parameters were et accord ing to the :D ll owing (Tabl e 4.1):
Table 4.1. Measurement parameters for CD spectra of CRD-BP KH variant .
Photometric mode:
Measure range:
Data pitch:
Sensitivity:
D.I.T.:
I
Bandwidth:
Start mode:

D, HT
240-195 run
0.5 nm
tandard
2 econd
1.00 run
in1n1edi ately

Scanning speed:
Baseline correction:
Shutter control:
CD detector:
PMT voltage:
accumulations:

Final D pectra data point were co nverted fr m raw

20 nrnlminut
ba eline
manu al
PMT
auto

D ignal (mdeg) to mean

elipticity (deg• cm2 dm r 1) u ing the built-in pe traManager oft ware uite (Jasco In .) by

77
m luding th

ptical param ter

dif£ r n

RD-BP quantifi d c nc ntrati n c uld be ace unted [i r pri r t a final

m

f

RD-BP

pe tra pl t. Finally, m lar elipticity alu

nc ntrati nand uvet1 path-1 ngth; h r

wer c nvert d t m ean r idu e lipticity, th

c mmon unit t r pre nt pr tein ec nd ary tru cture analy i
were g nerat d u in g Kal ida raphT

4.1.3

mall

u ing Mi cr

ft xcel.

ata pl t

oft are.

RD-BP econdary tructure e timation

In order to

tract additi nal quanti ta ti

RD-BP and the ariou

data fr m th

ingl and d ubi KH mutant variant ,

toe timate the fra ti on f a-h Ii . ~- h
ample fr m the mean re idu

ftwa re alg rithm were u d

t and un-coil ed m tif pre ent within each

lipti ity (deg • cm 2 dm r') value pl ot.

RD-BP

umerical data points at

1 nm interval were determined from a curve trace u ing built-in Ja co Spectra analysis oftw are
from 195-240 nm and convert d to

£, or mo lar circu Jar di chro i m (deciliter moJ!' -1 cm/\- 1), and

used a the raw input values for econd ary tructure analy i w ith th e K2D3 circular di chroi m
web server (Louis-Jeune et al. 2011) using generated protein CD reference data sets from
DichroCalc (Bulheller and Hirst 2009).

4.1.4 CD monitoring of CRD-BP conformation upon RNA titration
CD spectropolarimetry was used to monitor econdary structure content of CRD-BP
following titration with multiple target mRNA regions previously found to bind through EMSA .
KRAS , C-myc and Gli all had binding and non-binding fragm ents that were used for CD
comparison, the sequ ences of which follow below .
c-myc(+)

1705- 1792
c-myc( -)

1815- 1886
li (+)
Gli(-)

AC AGAUCCCGGAGUUGGAAAACAAUGAAAAG C
AA UA UUAU CUUA AA
AGC ACAG AUACAUCCUGU CGU AAG
GAAGAGGA UUGUUG GAAA GA GAGAA AGUUGAAA A AAACUUGA CAG UA
GGAA U UUGUGCG
A U
U GUAG
AU AA U G GAU C
AG U
UAG UUUUUU AU A U G GAUG A

78

t:run at d er 1 n f
KH 1 to KH4 wa al

ild-typ

RD -BP lacking th RRM d mam

u din id nti al

p rim nt with th rational b ing that if th RRM
binding, r m

d mam ar in rt and d not pa1ti ipat m
chang in

D ignal occuning

ithin th KH d mam

e p rm1 nt 0.1 em quartz cuv tt
pectr p larimet r

re u d for all

a u d nee again

indicat d (4.1.2) xc pt

an p

c ntaining only

al f them will amplify any

b rv d up n

titrati n.

r the e

ampleanaly i and th Ja coJ-815

ith imilar param t r

et; all fi ld a pre iou ly

d wa incr a ed to 100 m minut and nly 3 accumulations

perfonned. Thi chang in data acqui ition param ter wa mad e becau e the higher protein
concentration u ed in the e e perim nt c mpared to the mutant library analy i yielded lower
noi e and pr duced cleaner pectra. oncentrated t ck
into CD buffer (20 mM

aH 2P0 4 , 200 mM

RD-BP KHlt 4 ample were diluted

aF, pH 7.4) to a final concentration of0 .1 mg/ml

(2 .33 11M) in 110 111 amples that al o included target RNA at three different concentration
protein: RNA ratios: 16 :1 (0.1 5 11M) 4:1 (0.58 J.!M) , and 1: 1 (2 .33 J.LM) . Protein-RNA mixture
were incubated at 35 °C for 10 minutes, followed by 5 minute on ice, another 10 minute at 35
°C and a fmal 5 minutes on ice immediately prior to CD scanning at 25 °C. For each RNA tested,
three RNA-blank samples containing the target RNA molecule at identical final concentrations
(three protein-RNA ratios each) were canned, and the resulting RNA-only pectrum ubtracted
from the protein-RNA sample mixtures to judge the change specific to CRD-BP. For a more
straight forward visual comparison of the various RNA's effect on CRD-BP conformation, the
222 nm CD value for each sample, proportionate to the a-helical content of the mea ured protein,
was plotted as a bar graph to help di tingui h any RNA that had a uniqu e effect.

79

4.1.5 Generation of 18 RN A
1

RN

lin ariz d pT7- l

wa gen rat d fr m a M
pla mid t mplat . inear 1

di g t d in cir ular D 1m
D

wa in ubat d

foll w d by pr t ina

ith Hind Ill.

ith 20 unit

4

imately 50 Ill) 5 ~1 f

nbi n) u ing

t mpl ate uitable .D r tran ripti n wa fir t

~1 reacti

f I indiii (

K di g ti n in 0.5%

chloroform wa add ed and v rt
(appr

cri t T7 tr n ripti n kit (

w

n c ntaining (2 ll g) f p T7- 1
ngland

pla mid

i lab Inc.) D r 4 h ur at 3 7 °

diurn d cec l ulfa te D r 1 h ur at 50 ° . DN

d -D 1m2 minute at 1 ,00 0 rpm . To the upp r pha e
M

dium acetate and 100 Ill of 100% ethanol wa add ed,

mixed and incubated at -20 ° fo r ne half h ur to all ow linearized D

t precipitate. DNA

wa sub equentl y wa hed with 200 Ill of 70% ethanol, and fo llow ing centrifu gati on the DNA
p~ll et wa

re-su pend ed in 20 ~l of nuclease-free wa ter. Purified linearized DNA wa th en u ed

as a templ ate in the MEGAscript kit to generate approximately 100 11g of pure RNA .

4.1.6 T hermal stability of CRD-BP-RN A compl exes
Overall stability ofCRD-BP wa a essed u ing a thermal ramp (Pelti er temperature
controller) coupled with CD monitoring (Jasco J-8 15) at 222 nn1 . The wavelength 222 nm i
proporti ona l to the overall a.-helical content of the protein and i a good m easure of protein
denaturation. Protein melting temperature was calculated by fitting the molar elipticity to
equation 1, u sing the built-in Jasco Spectral Analysis software for determining Tm , as defined by
the temperature where 50% of the total elipticity i lo t.
Equation 1.

y =

[min + ( max- min ) ]
1 + e[- k (x- Tm ) ]

80

RD-BP pr tein wa add d to a 110 ul r a ti n t a final c n
in a 20 111M Tri - 1 and 200 mM
rela ed farab

ab

a l buf[i r. Th

wit h back t

a 1wa rationalized by th

rbanc r quir m nt £ r mea uring nl at 222 nm ( chl rid i n

rb p larized light and int r£ r

ith p

trongly

tra acqui iti n bel w 2 0 run) . emp ratur wa

incr a ed fr m 20 ° to 5 ° at a 2 ° /minute ramp-rate
pitch and

ntr ti n f0.4 m g/ml (9. J..LM)

0

c nd quilibration delay 0.5 °C

th rwi e id ntical data acqui iti n param ter a previ u ly indicated (4 .1.2) .

buffer- nly blank ubtracti n wa

arried ut t

tabli h a ba eline, as nl y the chang in

ignal i m a ured to det rmine m elting temperatur and it inclu i n w uld have not chang d
re ult .
RNA ample w re added to

RD -BP in 4 different rati

pr tein toRN

4:1 16: l and 64 : 1). Th rmal denaturation xpenm nt included KRA

rati

( 1:1,

ub-regi n A3 4 as the

PC? itive binding RNA, and KRA -Al2, 18 and tRNA as negative binding te t RNAs . ach
experimental condition was repeated in triplicate.

4.1.7 Denaturing purification of CRD-BP KH1to4
BL21 (DE3) E.coli cells were transfonned with pET-4l c-KHl to4 plasmid and expre sed
in an identical fashion to section 2.1.3 , as with the full -length CRD-BP. Frozen pellet (4) were
re-suspended in a denaturing buffer containing 8 M urea, and purified using a Qiagen minicolmnn in a manner identical to the full -length CRD-BP protein producti on except "buffer E''
(pH 4.9) was skipped , proceedin g directly to " buffe r F" (pH 4. 5). This provided higher
concentration in the elution step wruch was followed by dialysis into a 2 M urea refolding buffer
and fmally into a 0 M urea storage buffer .

81

4.1.8 Native purification of
B 21 (D

)

.

RD -BP K.Hlto4

li ell w r h at- h k tran D rm d (2 .1.1) with pla mid p

(4.l.ll)containingth truncat d

dingregi n f

agar containing 25 ~ g/ml kanam

in anti-bi tic and 11

in ubat r.

ppr

imat l 20 c 1 m

u d to in culate 1

f

br th

er

ide (IPT

- P KHl t KH4) and plat d n B

icked fr m th

d t gr w
ucc

etnight in a 7 °

ful tran D tmant and were

ntainin g 25 ).l g/ml kanam ycin and incubated at 7 ° until

D 6oo = 0.5 at which p int the cultur
thiogala topyran

-4lc

a indu ed with 1 mM I pr pyl-0- - 1-

Bi Ba ic Inc. foll w d by a 6 h ur expre i n period at 37 °

ell were pelleted by entrifugati n at 4 ° , 2,100 xg, for 15 minute ( ub -divided into 4
volume

f 250 m1) which pr du

d 4 cell pell et which were t red at - 0 ° .

ell pellet wer thawed on ice for 15 minute andre u pended in 12 ml non-denaturing
lysis buffer (table 4.2). Ly ozyme (1 m g/ml) wa added to there u pend ed cell pellet and
incubated for 30 minute on ice followed by

nication ly i con i ting of 6 cycles of 20 second

bursts (power level 5) - 40 second cool-down time . Lysate was ubsequently centrifuged for 30
minutes at l 0,000 xg 4 °C . Lysate supernatant was mixed with 2 n1l of 50% Ni-NT A agarose
resin slurry (Qiagen) and incubated on ice for 1 hour on a waver. The lurry wa then loaded
onto a Qiagen mini-column and the beads washed twice with 4 ml of wa h buffer (table 4.2)
followed by elution with 4 ml of elution buffer; collecting 0.5 ml fractions off the column. SD PAGE analysis was completed to assess the presence and purity of the fraction containing CRDBP KHl to4 with a final 3-stage dialysis step against a storage buffer consisting of 50 mM
NaH 2 P0 4 , 200 mM Na 1, pH 7.4 using l 00 ~1 n1ini-dialy i units (Thermo cientific) .

82

Table 4.2. Buffer u ed in native purification of

RD -BP KHI to4
Elution buffer

4.1.9 Denaturing on-column refolding purification of
High c ncentrati n
refolding m thod

RD -BP KRlto4

RD - P KH It 4 pr tein required the u

fan ' on-

I umn " pr tein

hereby pr tein are d natured and th en re-£ ld ed while still b und to aNi-

TA c lumn t a oid inter-mol cular aggr gate £ rrnin g. The pr ce
al o a vi ual guide can be r fened to (figure 4.1). BL21 (D 3)

i de cribed here and

.coli cell were heat-shock

transfonned (2 .1.1) with pla mid pET -41 c (4.1.11) containing the truncated c ding region of
CRD-BP (K.Hl to KH4) and plated n LB agar containing 25 ~ g/ml kanamycin anti-bi tic and
allowed to grow overnight in a 37 °C incubator. Approximately 20 colonie were picked from
the successful transfonnant and were u ed to inoculate 4x 1 L (4 L total culture) ofLB broth
containing 25 ~ g/ml kanamycin and incubated at 3 7 °C until OD 600 = 0.5-0.7 (each 1 L culture
varied slightly at time of reading) at which point the culture were induced with 1 mM IPTG
(Bio Basic Inc.) followed by a 6 hour expression period at 37 °C . Cells were pelleted by
centrifugation at 4 °C, 2100 xg, for 15 minutes and cell pel lets stored at -8 0 ° . Cell pellet were
resuspended and lysed in 48 ml of 8 M urea denaturing lysi buffer (table 4.3) and placed on a
waver for 1 hour on ice. Following celllysi , the ly ate wa centrifuged for one half hour at
10,000 xg (4 ° ) followed by filtration with a 0.45 ~m syring filter ( MD Millipore) yielding
approximately 45 ml of cleared lysate.

83

Sample inje t

Eq.

Wash 1
(JJ

0

N

ro

"U

·-E

·Figure 4.1. Diagram of on -column refoldin g KHlto4 protein purification cheme. rea
con ntrati n rang fr m a rna f M t minimum f 0 M (% - denaturing ly i buffer) and
i1nidazol rang fr In a rna
f 250 mM t a minimum f 10 mM with an intennediate wa h
c nc ntrati n of 52 . m
!eared 1 at wa load d nt an
machine (FPL

E Healthcar Life

Healthcare Life cience ) ampl holder.
Hi -trap™ co lumn (

KT prime plu fa t-protein liquid chr matography
u ing a 50 ml

KTAprime

up rl oop™ (

!eared ly ate wa then inj ected over a 5 ml

i-NT A

Healthcar Life cience ) at a flow rat of 0.5 mllmin , with 50 ml f

denaturing ly i buffer pumped through the

up rl oop™, behind the amp le plun ger to en ure no

inadvertent nlixing with non-denaturing buffer. CRD-BP bound to the His-trap™ column was
wa hed with 5 column volumes (25 ml) of denaturing ly i buffer at 5 ml/min. On-column
protein refolding was then initiated, where urea concentration wa slowly reduced from 8 M urea
to 0 M urea in the presence of oxidized/reduced glutathiones at a rate of 0.5 mllmin ( 100 ml
program span) by progressively increasing the percentage ofrefolding/wa h buffer (table 4.3)
passed over the column. This was accomplished by utilizing the AKTAprime plus mixing valve
(from 0 to 100% refolding/wash buffer). A second wash step passing 10 colunm volumes (50 ml)
over the Ni-NTA column at 5 ml/min was programmed followed by the final elution tep passing 25 m l of elution buffer at 1 ml/min over the Ni-NTA Hi -trap™ co lumn , collecting I .0
ml fractions. Collected fractions were analyzed by SD -PAGE and the si pure t fraction
pooled prior to a 3-stage dialy is against a final working buffer consi ting of 100 mM Tri - 1,

84

ca

tt

(Th nn

cientific).

p ctr ph tom t r (Th rm

EXP

y

bt

1.
'

h 4 kDa pr tein wa d ubly quantifi d
c· ntifi ) u ing a al u1at d

a t ig r t 1 2005) f

prot in quantificati n kit (Pi r
fw1h r

nc ntrat d t

h rm

ith a Nan dr p

tin ti n c - ffici nt (Pr t aram -

94 M- 1 m-1 and additi nally u ing a B

ci ntifi ). or cr

1 mg/ml u ing 10,00

tallizati n trial , the pr t in wa
MW ( 10

a) Amic n

ltra 2

ml cent:Iifugal filt r ( MD Millip r c rp .).

Table 4.3. Buffer u ed in denaturin g on -column refoldin g protein purification.
Hidiz d D rm glutathione.
reduced fonn glutathi n
Refolding/wash buffer
Denaturing lysis buffer
Elution buffer
50 mMNaH 2P 4
50JnM aH2P04
50 mMNaH2P04
200mM a 1
200 lnM a 1
200 lnM NaCl
52.8 lnM imidazole
10 lnM imidazole
250 1nM imidazole
'
0 M urea
8 M urea
0 M urea
1 mM GSH/0.1 mM GSSG
pH 8.0
pH 8.0
pH 7.4

4.1.10 Fluorescence polarization analysis of CRD-BP binding KRAS RNA.
Fluorescence polarization (FP) spectra copy was used to confirm the RNA binding
activity of CRD-BP KH1 to4 protein prepared by on-colunm refolding . A fluore cein-labelled
KRAS probe was designed and synthesized by Integrated DNA Technologies Inc (IDT) ba ed on
the KRAS-A34(b) fragment found to bind using EM A in ection 2.3 .2. The mfold RNA web
server was used to fold the KRAS-A34(b) RNA fragment (57 nucleotides) in sili ca, and ba ed on
the secondary structure prediction, a 44-nucleotide fragment containing the nucleotide
compri ing the computed structured r gion was chosen. Binding reaction were a sembl d a 19
~1 volumes containing

RD-BP KHlto4 protein (0-1000 nM), KRAS FP probe (10 nM) in a

85

buffer imilar t

M

binding a a

2.1.6)· 20 1nM Tri - 1, 200 mM

a 1 1 H 7.4.

Imn1 diat ly fi 11 wing r a ti n a embly, c mp nent w r mi ed by mild centrifugati nat

000 rpm in a tabl t p c ntrifuge fi r 10
po 1 th r action. Binding r acti n
fi 11 w d b an ther 5

minut

nd , mi

d by fli king, nd entrifuged again to

r th n incubated at 5 ° fi r 10 minut

° fi r 1 minut

and 5 tninute

on ice for 5

n ice. lnt a

4-well

microplat , 17 ~l f ach ampl reacti n m1

a add d and ub qu ently th t tal flu re cence

and parall lip rpendicular Ou re cenc

alu

w re mea ur d u ing a y:nergy 2 Multi -mode

Mi roplat Read rand

ntr 1 and Data

Inc.).

n5 Read r

oftware (Bi tek In trument

ptical excitation filter and mi i n filt r were et t 4 5/20 nm and 52 /20 nm

re pecti ely and y t m ga in et to auto-adju t with a read-height of7 .00 tmn . Fluore cence
anisotropy valu e w replotted for each CRD -BP KHl to4 concentration ample in
K~l ei daGraph™ graphing

oftware to determine Kd.

sapo itive control, wild-type full -length

CRD-BP was u ed for compari on with Bovine Serum Albumin (BSA, Sigma Aldrich) being
used as a negative control in otherwise identical experimental conditions.

4.1.11 Construction of pET -41c(+)-CRD-BP KH1to4 plasmid
Crystallization ofCRD-BP was al o attempted with a truncated f01m containing only the
K.Hl to KH4 binding module (RRMl and RRM2 removed) which was sub-cloned into a
pET4l c(+) that had been previously modified to contain no N-tetminu GST-tag sequ ence. The
pl asm id vector was generou ly donated from Dr. Natalie trynadka 's X-ray cry tallography
laboratory at the University of British Columbia and i ideal for protein expression oriented
towards stru ctural studi es due to a thrombin cleavable 8x his-tag and no G T -tag. Ndei and
Hindiii restriction enzymes (New England Bi lab Inc) were u ed to accomplish ub-cloning of
K.Hlto4

RD-BP into p T4l c(+) plasmid .

86

IZ promotor

.,

3 tta t a c ga c t c c t t gggg

BamHI

BsrG!

I ac operator

Xba!

tt g t g gc gg t a tt cccc t c t g

Asci

Pstl

cbs

at

Sal!

at cggg t ccga tt c t g t c ggcc tt ggcgcgcc t gc ggcg gc t ccg t cg~ c

tttt g ttt a c ttt

g

.t:i!1eL

gg g t t c t t g

Htn dll

g~ ttg t gcc cgcgg t g t gcnycy~~yc
thrombin

X hoi

t c g g c cca cca cca cca cca cca cc c t
t g tt
_ _ _ _ _ _ r.,·l
~__:_.:.....:..:..:......;

8HIS Tsg

'igure 4.2. p T41c(+) - no G T pia mid multipl e clonin g ite equ ence.

ote the thrombin
lea able 8 Hi -tag and lack of
equen e . d 1 and Hindlll restricti n ite wer utilized
r ub-cloning full-length and KH l t 4 R -BP coding equence .

:RD-BP K.Hlto4 insert DNA generation by P R
P R wa utilized to generat the D

in ert (table 4.4): with wild-type

RD-BP pia mid

JET28b backbone) erving a th rea tion template . Primer wer de igned to amplify a 1J 1
ucleotide region of the coding equence encompassing KH 1 to KH4 a out! ined below.

~ RD-BP K.Hl to4 forward primer :

>esign cheme: 5' [handle]-[Ndel]-(complementary region] 3'
>NA sequence: 5' [ACCA ]-(CATA TG]-[ A TCCCTCT CGGCTCCTG] 3'
ina! sequence: 5'-ACCACATATGATCCCTCTCCGGCTCCTG-3'

:RD-BP K.Hl to4 rever se p r im er :
>esign cheme: 5'[ complementary region ]-[1 Iind ll l]-[handle] 3'
on-R . . DNA sequence: 5'[AAGCAA AGCAC AGAAG]-[ AG TT]-[AC A] 3'
AAAG T CTTCT GT
TGTTGCTT-3 '
inal DNA sequence: 5'-A
*R.C. - reverse complement
*handle (A
A) not rever e-complement

86

T7 promotor

~

aa tta t c ga c t c c t a t gggg 1

BamHI

EcoBI

t a t cgggat ccg
I

BsrGI

cbs

Xbal

tt cccc t c t g

Asci

Pstl

Sac!

a t aa ttttg tttaa c ttt

Sal!

tt c t gt c ggcctt ggcgcgcct gc ggcg gc t ccg t cgac

g ., rgg g t

~

t ac t

tg

Hmdll

g6 ttgt gcc cgcggt g tgcn ycy~~gc

l
thrombin

X hoi

c t c g g c cca cca cca cc cca cca cca c t
·-~..:...:..:..:...:......----.:..-- F

t g tt
t •

8HIS Tsg

Figure 4.2. pET41c(+) - no G T pia mid multiple clonin g . ite equ ence. N te the thr mbin
equenc . de 1 and 1 indllf re triction itc w reutilized
clea abl 8 Hi -tag and lack of
for ub-cloning fuJI-length and KH 1t 4 R -BP coding equencc .

CRD-BP K.Hlto4 in ert DNA ge ner ation by P R
PCR wa utilized to gen rate the D A in ert (table 4.4)~ with wild-type

RD-BP pia mid

(pET28b backbone) erving a the reaction template. Primer were de igned to amplify a 1J 13
nucleotide region of the coding equence encompassing KH 1 to KH4 as outlined below.

CRD-BP KH1to4 forward primer :
Design scheme: 5' [handle]-[Ndel]-[complementary region] 3'
DNA sequence: 5' [ACCA]-[CATATG]-[ATCCCTCT CGGCTCCTG] 3'
TG-3'
Final sequence: 5'-ACCACATATGATCCCT TCCGGCT

CRD-BP KHl to4 reverse prim er :
Design cheme: 5'[ complementary region ]-[Hindiii]-[handle] 3'
non-R . . DNA equence: 5'[AAGCAA AGCACCAGA G]-[AAG TT]-[AC A] 3'
AA
TG TGCT TTGCTT- '1!
Final DNA seq uence : 5'-A
*R. . - reverse complement
*handle (A
A) not reverse-complement

87

RD-BP KH1to4 amplicon DN
etc cc gg ct
cc a tcc g aaa
c g ggcgctgc
g c a agatgat
t cccctgaa
ggaacctgaa
acctcacgct
gcagggccga
ccatgagctt
cagcttcatc
gctccttcat
tgggcgccat
cctccatcaa
ctggaccccc
agaatttctt
cagcagccgg
cc g cagctga
t taagatcat
tggctcaagt

cctggtgcct
catcacaaaa
ggagaaggcc
cttggagatt
gatcctggct
gaagg ggag
ctataaccc
gcaggagatc
gcagtcccac
cagcgc gtc
gcaggc ccg
cat ggcaag
gattgcacca
agaggctcag
tggtcccaag
ccgtgtca c
gg gg agtg
cggacatttc
taagcaacag

equence:
acgcagtatg
cagacgcagt
atcagcgtgc
atgcacaagg
ca aacaac
caggacacag
gagaggacca
atgaagaaag
ctca ccctg
cctcctcctc
gagcaggaga
aagggccagc
ccagaaacac
tcaaggccc
gaggaagtaa
ggcaaaggcg
ccaagagacc
tatgccagcc
caccagaag

aggcgctat
ccaaaataga
attcaacccc
aggcaaagga
tcgtcgggcg
agacgaagat
cactgtgaa
ttcgagaggc
ggc taacct
ccagcagtgt
tggtacaagt
aca caaaca
c gac ccaa
agggaagaa
age agagac
gcaaaacgg
agaccccgga
agatggctca

cattggcaag
cgtgcatagg
tgaaggc gc
caccaaaacg
actca tggc
caeca ctca
gggcgcca t
ttacgagaac
ggc gc g a
caccggggc
gttcatcccc
ac ctcccgc
ag cgaatg
ctatggcaaa
ccacatacgg
gaatgagc g
tgagaacgac
gcggaaga c

a tee
gagg g t g cca
aaggagaatg
tcctccgcgt
gcagatgaag
aaggaagggc
cgctccag
gagaactgtt
gacg ggccg
ggtctcttcc
gctcccta a
gcccaggctg
ttcgccagcg
gtcgtca ca
ctaaaagaag
gttccggct
cagaacttga
caag cattg
cgagacatcc

PCR mixture were a embled in triplicate and later combined to provide high working
concentration. Each 25 ~-tl PCR reaction wa a embled according to table x and a then11ocycler
scheme programmed for optimal amplification. High-fidelity Phu ion™ DNA po lymera e

ew

England Biolabs Inc.) wa used in order to reduce likelihood of mutation during in ert
amplification. PCR reactions upon completion were then run on a 1% agaro e gel at 120 V
stained with 0.5 1-1-g/ml ethidium bromide and ubsequently gel-purified u ing a QIAEX II Gel
Extraction Kit (Qiagen) following the recmnmended protocol.

PCR thermocycle program:
Step

Condition

1

98 °C 30 sec

2

98°C 15 sec

3

55 °C 30 sec

4

72 °

2 min

6

72 °

10 min

7

40

forever

step 5 - repeat 25x

88

Table 4.4. PCR reagent u ed to generate KHlto4-pET-41c(-G T) pia mid vector.
volum
2.5 Ill

ector pla mid (p T41 c) £ r ub-cl nmg,
concentrati on, mall
£ llowing heat- h

hi h wa d nated fr m

lum e ampl e, wa amplifi din a 10 ml

B

a a1 w

H5a ~ .c li ce ll culture v !urn

k tran fonnati n (2. 1.1) and plating onto an LB agaro e plate with 25 ll g/ml

kana1nycin antibiotic.

vernight-gr wth c loni e w re cultured in 10 ml LB broth volume (25

llg/ml kanamycin) for 16 h ur , followed by pl a mid purifi cati on using a QIAprep

pin

mini prep kit (Qiagen). Due to pET41 c being a low-copy plasmid tw o 5 ml culture volum e were
prepared eparate ly, each acco rdin g to th e manufacturer' reco mm endati ons and th en co mbined
by passing both plasn1id preparations over a ingle QIAprep spin column and eluted together to
yield approximately 5 ll g of plasmid D A (1 00 ng//ll, 50 )ll).
Double-digest restriction endonuclea e reactions were setup for th e in ert and pET4 1c
empty vector DNA sampl e in preparation for sub -cloning. Digest reaction were setup in 20 )ll
volumes according to table 4.5 and allowed to continue for 2 hours at 37 °C. Digested DNA
products were resolved on a 1% agaro e gel and gel-purifi ed from excised DNA bands using a
QIAEX II

el xtraction Kit ( iagen) fo llowing the recomn1end ed protocol (note that KH 1to4

insert DNA elution i at room temperature whil e the larg r pET4 l c v ctor DNA is eluted at 50

90

Table 4.6. Reagent u ed for RD-BP KH1to4 pia mid ligation. Pla mid p T4l c(- T) MW
appr imat ly 5 kb and KHlt 4 in rt
1.2 kb ; alu u d t calculat d molar rati s.
Reagent
lOX T4 buf:D r (NEB)
KH1to4 DN_
p T41 D
T4D

rt.

l Jll
to 20 Jll

1 Jll
t 20 Jll

fu i li ga ti n c ntaining th e

RD -BP KH1 to4

10 ng

5 ng

50 ng
1 Jll

3:1

t 20 Jll

cr

n ~ r ucc

w ere am pl ed dir ctl

ff th L

P R
m

1 Jll
to 20 Jll

2 J.tl

liga e (NEB)

H2

Vector onl~

50 ng

7:1
2 )..tl
85 ng
50 ng

1:1
2j..d

a u ed t

inoculate mall P R rea ti n tube .

agar

2 Jll
0 ng
50 ng

pl ate u ing a t rile loop and u ed to
mbl ed D r each colony ampled

n reaction wa a

according to tabl e 4 _6 and run on a 1% agaro e gel for UV ethidium bromid e det ction of
appropriate band. u cce ful tran form ant were then u ed to in ocul ate l 0 ml LB broth culture
and pla mid pmified u ing the QIAprep pin miniprep kit (Qi agen) as de crib ed previou ly for
the empty pET4l c vector.
DNA equencing w as completed fo r three po iti ve clone plasmids to further confi rm the
presence of correct DNA insert, and also to ensure no mutations were present in the equ ence.
Plasmid samples were diluted to 100 ng/Jll and 10 J.!l volumes were shipped to M acrogen for
next generation sequ encing u sing three different equencing primers for each pl asmid : T7
promoter forward prim er, T7 termin ator rever e primer and a custom designed " KH I P" fo rward
prim er that ann eal s to th e 3' end of KH 1 DNA sequ ence.

Primers used for DNA sequ encing of pET4lc-KHlto4:
A

T7 promoter forward primer sequ ence: 5'- TAA TA

T A

T7 te1minator reverse primer equ ence: 5 ' -

T AGTTATTG

"KI-f 1P'' forward prim r equ ence:

G T

5 '-

T

TT

TATAGGG -3 '

G

T

91

Table 4.7. Rea gent for colon P R creenin g.
PCR reaction mix

Thermocycle program

reagent
lOXP R bufD r
p (2.5 mM)
d

volume

ilil!

temQera ture

.5 ~1

1

.5 ~1

2

K.Hlt 4_ F prim r

1 ~~

K.Hl t 4_R primer

1 ~1

4

°
95 °
54 °
72 °

Taq.P 1.

0.5 ~1

5

H2

25 .5 ~~

6

time
5 1nin

95

t

linin
90 ec
2 min

t p 2 - 30X
72

°

5 min

4.2 Re ults and Di cu ion
4.2.1 Circular Dicbroi m analy i of CRD -BP KH mutant va ri ant
RD-BP KH single and double mutant variants harbouring the
. previou ly for EM

XX

mutation u ed

binding experim nt (3.3.1 and 3.3.2), were canned u ing a circular

dichroism (CD) sp ectropolarimeter and far-UV pectra acquired from 240-190 m11 for each
mutant protein. All protein amp les were freshly dialyzed and refolded into CD buffer and were
all between 1 and 3 11M concentration prior to loading into a 0.1 em quartz cuvette for

D

scanning. Acquired CD spectra for each KH mutant were remarkably imilar to wild-type CRDBP following conver ion from raw CD milli-degrees to mean residue elipticity (figure 4 .3). The
mean residue elipticity unit accounts for differences in scanned protein concentration and return
the average CD signal per amino acid residue, making inter-protein sample compari on more
feasib le. Far-UV wavelength cans of wild-type CRD-BP reveal a pectrum characteri tic fa
protein with mixed alpha-helix/beta-sheet structure wi th a strong broad p ak at 222 nm,
indicative of a large alpha-helical component. This finding matche known tructur

of KH

92

d mam-c ntaining pr t in a w 11 a
domain pr t in ( ha

t a. I 201 0;

ther

pectra d ri ed from ingle and tand n1 KH

al erd eta!. 200 7;

It i inun diat ly lear up n in p ction f the
global tructur
p ctra p ak
up n cl
The

lu11iel t a!. 2006) .

R -BP KH mutant

D p ctra that th

ar rem arkabl y imilar and do not h w any large di fD r nee in the locati n of
r the amplitud e f

m p

di f~ r nc

ti on, m

ignal.

ubtl e dif:G r nee d e i t betw een the pectra

t n ta bly at wa elength I wer th an 205 run, and higher than 23 0 nm .

are m all and tr ngly ugge t tha t the verall fold of the vari u KH

XX

mutant app ear to be m aintain d. H w

r as far- V pectra analy i i not a mea ure of the

f pro tein , it remain p

ible that relati ve change in alph a-helix/beta-sheet

terti ary tructur

orientation ha e occurTed a a re ult of intr du cing

XXG mutation . A well , shift in th e

preci s a-h e l i and ~- heet res idue bound aries are pos ibl e, so long as the overall fraction of
each motif remain con tant. W ild -type CRD-BP wa scanned at 90 °C as a contro l for denatured
protein (figure 4 .3D ), and far-UV pectra showed weaker CD ign al a expected, notably in the
208 run and 222 nm regions where protein econdary structure contributes strongly.
Further tructural analysis involving econdary structure calcul ati ons fro m CRD -BP CD
spectra w ere also completed with the aid of the K2 D3 econdary structure estim ation web erver,
and the Dichrocalc online CD reference-spectra data base . Inputtin g th e L1£ va lu e , or mo lar
circular dichroism (deciliter moJ!'- 1 cm"'- 1) at 1 run interva ls for each mutant produced hi ghly
similar final secondary structure estimati ons, varying b y le s than a percent b tw een each ample
(table 4. 8). Given the ti ght overlap in CD spectra between wild-type

RD-BP and th e va1i ou

mutant fonns, especially at the criti al wavelength s used t r omputing econdary tructure
e timations, this result was expected and brings numetical cope a to th likenes of proteins
within the

RD-BP mutant library.

93

A

B

o'
0
E

'0

5 5000

RD B WT
H1
CRD B
CRD BP KH2
CRD 8 KH3
CRD B
H-1

0
CD

~

~
~

0

0.

..

0
E
'0

0
CD

o
.,

'0

~

~~------~------~-

270

Lii

RD BP VVT
RO t:5 KH1 2
0 -B · KH1 -3
0 OP KH1 4
RO BP KH2 3

5 ':iOOO

730

Wave length (nm)

220

CIJ

230

__.i-4 0

/J""'

Wavelength (nm)

:J

~ 5000

/~

Q.l

cr:

1.1

c

p'

Cll

CD

~)l

o'

~

~~

c

D
o'

-0

-0

E

E

'0

CRO BP WT
(;>CRO BP KH2 4
...... - CRO BP KH 3 4
- ·0
CRO BP Y5A
<l
CRO BP 0 . :>? F

~

E 5000

u

C>
CD

u

..

~

0

u

220

Lii

Wavelength (nm )

230

CD

:J

-;;

5000

Cl>

cr:

~

c

"'CD

::E -1

-

'0

E
u

o·
;;:r

-~~

rf

/

rr
_g

i'!O

CR

5000

OP W1

Cl
Q.l

'C

..

~
u

a.

u

o·

Wavelength (nm)

0

a.

200

210

240

Lii
CD

:J

'C

';OOQ

"'

G.~-&e~

zy-0

CD

cr:
c
.,"'

:'!:

10.

Figure 4.3. CD spectra of wild-type CRD-BP a nd KH mutant variants. ircular dichroi m
spectra of(A) wild-type, KHl , KH2, KH3 and KH4; (B) wild-type, K.Hl-2 , KHl- , KHl-4,
KH2-3 ; (C) wild-type, KH2-4 , KH3-4, Y5A, and D527E. (D) Wild-type CRD-BP pectral can
at 90 °C as po itive control for denatured protein. All spectra generated from 8 accumulation
using a Ja co J-8 15 ; protein scanned in 20 mM Tris- 1, 10% glycerol, 200 mM Na I, pH 7.4
buffer.

94

Table 4.8. Protein econdary tructur e timation for RD-BP variants. K2D n ural n t
ir ular di h ri m tructural d tenninati n w b erver wa u d inc njunction with ichr calc
p tra d tab a t g nerat
timati n fr m £ ( d cilit r m }/\- 1 m/\- 1)
nlin

/ o a-helL~

o/o B- beet

Wild-type

67.51

9.00

23.49

KHI

67. 2

. 9

23.59

KH2

67.7

9.10

23.17

KH3

7.51

9.02

23.47

KH4

67.51

9.00

23.49

KHI-2

67.55

9.12

23.33

KHI-3

67.53

9.00

23.47

KHI-4

67.4

9.06

23.46

KH2-3

67.51

9.00

23.49

KH2-4

67.51

9.00

23.49

KH3-4

67.51

9.02

23.47

CRD-BP variant

0

0

/ o Random coil

4.2.2 Circular di chroism monitoring of CRD-BP-RNA binding in vitro
Assessing whether or not CRD-BP undergoes a conformation change upon RNA
interaction is impor1ant, as it may serve as a measure or indicator for RNA binding. Also, it
sheds light on how the protein functions and can highlight the different relevancie of
crystallographic structures publi hed with and without RNA bound . To determine if CRD-BP
has such dynamic character, CD spectroscopy was u ed a before in a eries of RNA titration
experunents with both high and low affinity RNA substrates. In addition to the full-length wildtype form, a truncated ver ion of CRD-BP was created containing only the KH I to KH4 binding
modu le (referred to from now on simply a "KH I to4"), and used

·t n i el

in the e

experiments. The rationale behind this move to include KH 1to4 wa that the fun tionall in 11

95
RRM don1ain ha
lik 1 t

b n

t nnined not t c ntribut t RN

hift fr m n c nformati n t an th r. H nee, inclu ion f th

wa anticipat d a p t ntiall dulling an

b erved change in

domain and th re£ r r m ved. ull -length
mplet farmyc

binding affinity, and thu ar not

p

R -BP

D ignal temming fr m the KH

a e amin ed fir t, mea uring th

trum fr m 1 0-260 nm and titrated with KRA

RD -r gi n, a n n-binding -m c regi n, a well a a pair f

lig nucl otid e with relati el
GLI(-) re p cti

4, KRA Al 2, the cLI competit r

trong competiti n and w ak c mp etiti n, labell ed

1 . Furth rmor , a a n gati e c ntr I fo r

B A wa titrated with KRA

tatic RRM domain

4 RN

Ll ( ) and

-ob ervable tructural changes,

and p ctra obtained in an otherwise identi cal fashi n.

Each ample wa pr par d in term of pro tein :RN

molar rati o equivalents; 4: 1 and 1: 1

preparation were completed and cann ed fo r ac h RN

ubstrate. It can be seen in figure 4.4,

that the addition of any RNA ub trate- high affini ty or low affmi ty appears to induce a change

in the CD spectra in a imilar fa hion. At all wavelengths, the mea ured di chroism reduced upon
titration with RNA, how ever while there is a lack of specificity fo r the general effect, it can be
seen that in every experiment (figure 4.4 A-C) , the stronger binding fragment wa able to induce
a larger change at the same concentration. As RNA alone possesses CD character, it is important
to subtract the RNA spectra measured at the same concentration from the titration sampl es; an
RNA-CD blank. In fi gure 4.4D, BSA was measured with KRAS A34 and A l 2, and the RNA
spectra subtracted, produ cing es entially identi cal

D spectra refl ecting both the unchanging

nature of BSA in the presence of RNA, and also validating the spectral sub traction techniqu e in
principle. A reduction in absolute molar elipti city, as observed in all CRD -BP-RNA titration ,
refl ects an un-coiling effect of the va rious alpha-helice compri ing th protein a well as betasheets. The effect i not absolute, as the

D signal does not ba e-line lo ing only about 25%

96

A

8
1 10

1 10'

no RNA
KRAS A34(•) 4 1
KRAS A34(•) 1
KRAS A12( ) 4 1
KRAS A12() 1 1

\;

E

E

no RNA
c myc(>)4 1
c myc(+) 1 1
c myc( }4 1
c myc( ) 1 1

~

0

E

' E0.

0

c:

ro

110' ~~----~~----~----~----~
730
210
240
200
220

110' ~~------~----~----~----~
210
2'30
240
200
2?0

wavelength (nm)

wavetengt (nm)

c

D
1 10'

L....

0

E

'0

"E

no RNA

'3 2 1

Gh •) 4
Ghl•) I 1

24 1

no RtJA
KRAS A14(• J 4 1
KRAS A34(•) 1 1
KRAS A12( 4 1
KRAS A12() 1

Gil() 4 t
Gh ) 1 1

sooo

1 ti 1

u

0

1l

8C
0
• UJ

5
;;

800(.

r

0::

c::

ro

::::
1 10' ~~------L-----~----~------

200

210

220

1

10

24 1

240

wavelenqtt nm

200

21

2 IC

240

wavelength 'nm

Figure 4.4. CD spectra of CRD-BP titrated w ith vanou s substrates. All ample are
presented as a molar ratio of protein to RNA. with RNA concentration increasing from 4 :1 to
1:1. (A) KRAS A34 i a high affinity substrate while KRA A 12 bind poorl y. RNA ub trate
also included: (B) c-myc (+ )and c-myc (-)are strong and weak binding sub trate re pectively .
(C) Gli ( + ) and Gli (-) are strong and weak oligonucleotide competitor re pectively 'vVith (D)
containing scans of BSA acting as a negative conformational hift control.
structure, however a loss in structure doe reDect and le s stable and higher free energ)
conformation. This is by itself an unlikely circumstance as RN

binding is spontaneou, and doe

not require energy input, however this energy barrier may be overcome by the free-energ)
co ntribution of RNA binding, ati sfying charge and po sibly orienting h) drophobic rcsidm: s
into more energy favourab le po itions .

97

urth r

p nm ntal data

a

tracted fr m the e p ctra, c mparing mo t dir ctly the

alpha-h li al cont nt f each titration ampl a 222 nm bar chart (figure 4.5).
p

tra wa m

ured in tripli ate and graphed a~ the a erag

each

f 11 a cumulati n , the 222 nm

value in figur 4 .5 i an averag \\-ith err r bar repre enting tandard deviati n. Thi
222 nm data allo~ [! r r latively ea
protein tructur . It can b

c mpan on of th R

D

ub trate effect n

tracted
RD-BP

a ily c n in the bar chart that indeed the tronger binding

ub trate . mo t n tab!~ c-myc. labelled with "( + )" arc rei iabl

more able to reduce the

D

ignal trength. alb it n marginally more than their vv aker binding"(-)" counterparts. As can be
quick] d due d fr m the ra\\-

0- pectra in figure 4.40. B A here sh w no distingui hable

pattern in prot in econdary tructur chang

. with neither ub trate affinity (KRA -A ...,4 vs

KRA - 12) n r R A cone ntration correlating Vvith the relatively mall change in 222 nm
elipticity.
Another important accompanying piece of data in CD experiments i the record d high
tension voltage (HTV), essentially a mea ure of the required signal amplification or gain
required to di tinguish noise from CD signal. Certain buffer or protein may absorb trongly in
the far UV, resulting in higher voltages being required to acquire spectra. However, protein
precipitating from solution also typically how strong increases in HTV. Patel e/ a/ . (20 12)
proclai1n that ZBPl bound to RNA possesse reduced solubility compared to the prot in alone,
which has been a major stumb ling block in the quest for a ful l-length cry tal tructure of ZBP 1
or any of it homo logs including

RD-BP . Reduced olubility of the CRD-BP-RN

in the experiments presented here would indeed how lo

complexes

of CD signaL as i ob erved .

98

A
~

-~....

B
1 w•

E

E
0

.....
u

• 4 1 Protem RNA

0

•

9000

E

1 1 Prote1n RNA

1 10

• 4 1 Protem RNA
• I 1 Protem RNA

9000

"e-o

8000
7000

7000

000

6000

sooo

'>000

4000

4000

3000

'3000

2000

2000
no RNA

krasa34(•)

krasa12( )

no RNA

CRD BP substrat

c Myc(+

c Myc( )

CRO BP substrat

c

D
I 10•

•
•

0
E
"E

4 1 Protem RtJA
I 1 Pro1em RNA

• 4 I Protem RNA
• 1 1 Protem RNA

000

0

8000

7000

'

' v;

<D

a:

::l

'J

6000

c:

c:

'll

~'

ro

ttl

::E
E

5000

f

c:

c:

('

~.

('

('.,

g:

4000
no RriA

Gil(•)

rRo BP substrate

Ghr)

19 1

18 1

l 1
1 t 1(l

no RNA

krasa34(•)

krasa12( l

BSA substratP

Figure 4.5. MRE at 222 nm of CRD -BP titrated w ith RNA ub trates . All sample are
presented as a molar ratio of protein to RNA with 4: l having lower RN concentration than 1:1
and mean residue elipticity (MRE) presented a absolute value . (A) KRA A34 was dete1111ined
to be a positive binding fragment while KRAS A12 binds poorly . (B) c-myc (+ )and c-myc (-)
are strong and weak binding substrates respectively. ( ) GLI ( +) and GLI (-) ar . trong and weak
oligonucleotide con1petitor respectively with (D) containing 222 nm trace of B. A acting as a
negative conformational shift control .

99

A

B
1000

no RNA

KRAS A34(•) 4 1
KRAs A ( +) 1
KRAS A12( 4
KRAS A12( ) 1 1

()()

800
'l)

700

0

0

'=

no RNA
c ycl • ) 4 1
c Myc(•)l1

800

c Myc( ) 4 1
c Myc( )1 1

00

~

(!)

>

1000

600

>

soo

J:

400

400
;?()('

300
200

0

200

240

220

210
wavelengt

1000

no Rt A

900

Gl1 • 4 1
Gh(•J 1 1

800

Gh( J 4
Gh 1 1

~00

230

240

no RNA
KRAS AJ4(•) 4 1

KRAs A14(• ) 1 1
KRAS A12 l 4 1
KRAS A12( l 1 1

4'>0

700

:r
0

400

ro

(!)

>

600

J:

soo

.....

220

D

0

-a

210

wavelength (nm)

(nrn)

c

Qi

200

>

..

150

400
:\)I!

300
2'>0

200

200

21

no

24

l.JC

wavelength (nm)

11

220

2'30

240

wavelength (nm)

Figure 4.6. CD-HTV plots of CRD-BP sa mple titrated with RN A substrates. All samples
are presented as a molar ratio of protein to RNA with 4:1 having lower RNA concentration than
1:1 and measurements plotted as high tensions voltage (HT -voltage) . ( ) KRAS A34 wa
detem1ined to be a high affinity substrate while KRA Al2 binds poorly. (B) c-myc (+ )and cmyc (-)are strong and weak binding substrates respectively . (C) Gli (+ )and Gli (-)are trong
and weak oligonucleotide competitors re pectively with (D) B A acting a a negative
confonnational shif1 control.
How v r, th HTV or gain would increase proportional] and appear different than that
of wi ld type alo ne. As can be e n in figure 6, no uch difference ar

vi dent ugg sting that

low co mp lex o lubi lity is not likely to be responsible for the CD reduction observed . I lowe\ er,
co nsi dering

RD-BP form . large multimer complexe a visible granules in cell ' , it cannot be

100
di

unted that the formation

f th

larger

mple

ould be re pon ibl for lo

of

ignal, th r [! r p t ntiating th data a am a ure f granu l :D rmation and not actual tructural
chang

ithin th pr t -in.

it int racti n

tructural change in

RD-BP and it homo! gs how veri lik ly, a

ith it e lf a a dim r. and with ther pr tcin

micro-tubul a

ciat granul

i R

dependent (B

uch as d nein and kin

in w ithin

ian eta/. 200 ; ll avi n eta!. 1998 ;

loannidi eta!. 2005; Ni 1 en era!. ~004) and pcrhap a protein-protein interaction m otif is
r

ealed up n R

binding a

ugge ted by Wang ef a/. (1999).

Furth r tructural tudie \\ere imilarly carried out with a Kll 1 to Kll4
truncation c mpri ing only the R

-binding module . The e

RD-BP

p riment were done in an

otherwi e id ntical fa hion except for an additional 16 : I, 1 \\- RN

concentration amp le wa

included to te t for a lovver thre hold concentration boundary for the

D effect. KRA -A34( )

and KRA -Al2(-) w reemployed again a well a several alternative ubstrate that were not
expected to bind: '18S', 'Orflb'. · pike' and 'tRNA' which were all u ed to broaden the earch
for potential RNA ub trates that do not cause the ob erved change in 'D spectra, which wou ld
offer much tronger support for the notion that these conformation changes are induced by RNA
binding.

D monitoring of KH 1to4 secondary structure titrated with any of the RNA ubstrates

however did not uncover any particular RNA that wa ineffective at producing the previously
observed reduction in

D signal (figure 4.7 and 4.8) .

101

A

B
1 10'

0

6000

E

e

6000

0
·D

4000

1 10'

noR A
RA5 A34(•) 4 1
KRA5 A34(•) 1 1
KRA5 A12() 4 1
KRA5A12() 1 1

-._
0

no RNA
185 1 1
185 4 1
185 1 1

6000

E
6000
4000

2000

0
2000

2000

4000

4000

6000

fiOOO
190

200

210

?'}I

wavelengt

2 0

' o

4

2oo

210

220

?30

240

?50

2 0

wavel ngth (nm)

c

D
1 10.

'o

700

(nm)

1 10

no Rt A
orf1b 1 1
orf1b 4 1

6000

E

0

no RNA
SPike 1 1
SPike 4 1
sp1ke 1 1

800C

f:

orflb 1 1

6000

6000

4000
2

0

2000
4000

4()()(.

000

f\000

2'>0

7

wavelenqttJ (nm

200

10

220

2'3C

240

2'

2

wavelength (nm)

E
1 10

0

E
'C

••E

Figure 4.7. CD spectra ofKHlto4
titrated with RNA ubstrates. All
samples are presented as a molar ratio of
protein to RNA, with RNA concentration
increasing from 16: 1 to 1:] for ubstrate
(A) KRA S A 3 4 and 12: (B) 18 : ( )
orb 1b: (D) pike: and ( ) tRNA .

no RtJA
IRIJA Hi 1
tRNA4 1
!RNA 1 1

6000
6000

u
0

(J}

4000

'C

2000

0
2000
4000
6000
19(.

700

710

72<

0

wavelength (nrn)

240

2 0

2h0

102

A
-

B
6000

E
E
<..)

-0

• 4 1 Prot m RNA
• 1 1 Prot m RNA

0

E

..

5000

E:

<..)

0

0

Q)

~

Q)

4000

.!(

?

<.;;

:=

Jf

~

()()()

(jj

J)

'])

::;

il

•11

cr

::;

v.
Ill

2000

0:

c;

ro

Q)

·r
:::i:
E
c:

1000

c-.
c-.

C'\1

C'\o
N

c-.

0
CRD BP

H 1104 substrate

CRD BP KH1to4 substrate

c
-._

D
-600
~._

0

E
500

'E

1.)

'])

'])

-~

J( 1

iii

1C 1

J(

~
:t
w

r

;;.
l1

::;

;;

2(

l'

0:

c;

ro

'll

E
c;

200C

c

ro

::;

3(

'!)

::;

· a:

5(J

"'0

0

:::
~

6000

0

E

"e

2C

c;

ro

::;
E
c;

?000

::0

1001

(ll

E.

c;

('

,..

('.

N

c-.

CRD BP KH 1104 substrate

CRD BP KH11o4 substrate

E

--

6001

E

'0

'e

50or

1.)

0

'])

.l(!)C

~

"

:=

~

iii

'\C,l

Q)

::;

;;.

a>

0:

2000

c;

ro

(])

::::
E

00

c:

('"

N
N

no RtJA

Hi 1 !RNA

.t 1 IR~JA

CRO BP KH 1to4 substrate

1 1 !RNA

F igure 4.8. M RE at 222 nm of KHl to4
titrated with RNA substrates. All
samples are presented a a molar ratio of
protein to RNA with RNA concentration
increasing from 16: 1 to 1: 1 and mean
r sidue elipticity (MR ) pre ented a
absolut value . (A) KRA A 4 i a high
affinity sub trate while KRA
12 bind
p rJy . RNA sub trate: (B) 18 , ( )
Orfl b, (D) Spike, and ( ) tRNA are not
expected to po ses high affinit) binding
character.

103

B
soo

no RNA
KRAS A

500

no RNA
18S 1 1
18S 4 1
18S 1 1

(+) 4 1

KRAs A34(•) 1 1
KRAS A12( ) 4 1
KRAS A12() 1 1

4

0

.o

400

(!)

~
•:X:

10

200

210

220

230

240

2SO

2 0

wavelenglh (nm)

wave ng h (nm

c

D
00

no RNA

550

orf1b 1 1
orfl 4 1
orf1b 1 1

5()('

no RNA

4r0

!>prke 16 1
!>pike 4 1
SPike 1 1

500
4

)

0

>
1-

400

3'l0

J:

3">0

300

190

200

21J

221

2'lC

23

26

E

ID
0

4()()

(!)

>
3">0

300

190

200

210

220

230

wavelenqth (nm)

22r

~

--'-4

----L-

____,j

26(,

All sample are pre ented a a molar ratio
of protein to RNA with RN
concentration increasing from 16 : 1 to 1: 1
and measurement plotted a high ten ion
voltage (l--IT-voltage). (A) KRA A34 i
a high affinity ubstrate 'While KR
A12 bind poorly . RNA sub trates (8)
18 . (C) Orfl b. (D) Spike. and (E) tR A
are not expected to po ess high affinit)
binding character.

-';;

~

210

Figure 4.9. C D-HTV pl ots of KH 1 to4
samples titrated w ith RNA substrates.

no Rt A
tRNA 1'
IRt A 4 1
!RNA 1 1

450

--'-----1---'----

wavelength (nm)

wavelenqth tnm)

500

2'>0 L----'--1qr
200

4

't

104

4.2.3 Thermal tability of

RD -BP-RNA RBP comple e

To furth r anal z the p t ntial for circular dichr i m to be u ed diagno tically in
hara t rizing

R -BP binding 'Aith vari u R

target . m lting temp ratur

f

d fined a 50o/o unii ld d prot in. 'A a det rmin d in the pr ence of various R
target
tabili t

[ th

f

mpl , b) reducing free-en rgy as are ult f pontaneou binding and increa e
pectro copy was again utilized to monitor the protein

tructure at 222 nm in th pr ence f diffi rent RNA pcci
increa ed ufficiently to melt the pr t in. KR
with KHl to4 truncated
.

target . In

R -BP may ha e the potential to increase the overal l

them lting t mp rature of R - P.

m I cui

RD-BP. a

, and the temp rature wa

- 12 and KRA , -A34 w reanalyzed in compl x

RD-BP. a w II a n gative control ubstr te 18 and tR

with the pr vi u

monitoring

RNA

periment . multipl R A concentration were

t ted including 1:1, 4:1. 16 :1. and 64:1 molar ratio . Importantly. protein concentration wa
increa ed to 0.4 mg/ml (- 10 flM) for scanning. Thi d ci ion wa mad due to the ob erv d
dramatic reduction in CD ignal at 222 nm (figure 4.7) with increasing RNA concentration.
Because the ba eline

0 ignal prior to melting would be quite low and prone to noise a a result

of RNA presence already. the overall quantity ofprotein-RNA complex wa increased
ufficiently to achieve approximately 30 mdeg of raw CD at 222 nm. Re ult can be

n 111

figure 10, as determined melting temperatures (Tm) which were computed by Ja co Spectral
Manager Thermal-denaturation software.

nly the 64:1 and 16:1 amples (figure 4.10 A and B.

respectively) were scanned due to visible precipitation of the

RD-BP-RNA com pi x . in the

4:1 and 1:1 sample ets. This precipitation wa not pre ent in the CD spectral analy i · sample
and is attributab le mainly to the increa ·ed concentration of both protein and RN
apparent reduced so lubility ofthe complc cs compared to protein alone.

u ed and th

105

KH1to4-RNA (64 :1) complex Tm

--

()

B

45

45

44

44

43

43
0

0
.........

0

E

1-

KH1to4-RNA (16 :1) complex Tm

E

1-

42

41

42

41

40

40
no RNA 18S tRNA A12

A34

no RNA 18$ tRNA A12

RNA substrate

RNA substrate

c
~----+---:::-=~--""'---,..._

CO [ deg)

-20

.tO

A34

GO
empera ure [CJ

HP
· noR A
"1111 A e of denatur hon
20 66 · 75 59 ICI
(2) Fllllf19 ne of stable range
A Y • 10 137062 • 1·32 3582)
8 y ·10 0119679 + (·8 63574)
(31 Re.sadual of htt~ng curve
SIGMA 01826
141Tm "' 40 67 • · 0 051091 (CJ
15JdH
• 259251 • 4651 8 [J/moll
(61 dS
= 826 12 • · 14 82331J!moV I

__

80 .1]

Figure 4.10. TMs of KHl to4 CRD-BP protein in complex with RNA substra tes. Plotted
values were determined from 222 nm
0

D amp le trac

over temperature-ramping from 20 to 80

and Tm calculated based on los of half CD signal. (A) RN A sub trate pre ent in a 64 :1

protein :RNA mol ar ratio and (B) 16 :1 in otherwise identical conditions. (C) xample Tm
calcu lation from thermal melting curve of RD-BP KH 1to4 protein ample with a sociated
computed valu es.

106

ther
and pr
alon

r littl differ nee in m lting t mperature between the protein al ne

ampl
nd ample

(le

ith R

r, th differenc

than 4 ° ).

a pr due d at th higher RN
111 m

re mo t pron unced at the low r R
electi ity can be e n in b th
n gli gibl dif[i rene

p cted , th gr at t difD r nc between protein
c n entration ( 16 :1 prot in to

ltin g temperatur bet een KRA -A 12 and KRA -A34
c nc ntrati n (64 : 1 protein t RN ). While

ncentrati on range .

in mea ured T m. both KR

L, _

ith both 18 ~ and tRNA addition producin g
12 and KRA -A 4 ampl es ge ne rall y

app ared to b within err r f each the r ~ ith po ibl a li ght bi a t
ub trate K

4 in the I wer R

me

ard the stronger binding

concentrati on ampl e . Toe pl a in th e more pr valent

differ nee in KRA - 12 and KRA - 34 at lo\\- r co ncentratio ns co mpared to hi gh r, it co uld
be that und r the

D e perim ental cond ition . there is a degr e of non- pecifi c bindin g, alth ugh

it cannot be complete! non- pecific a 18

and tRNA compl exe hav

ignifi cantl y lowe r T m .

In the context of u ing m elting temperature to assess C RD-BP compl ex tability and RNA
sub trate binding, the reduced so lubility of the compl exe at the hi gh co ncentrati on required fo r
Tm determination render the method limi ted . Improvements m ay fo ll ow by impro in g compl ex
solubility by experimenting w ith alternati ve buffe r content , or p erhap engineerin g a more
solubl e CRD-BP variant by geneti call y fusin g a tag such a Fh8 or G T.

4.2.4 CRD-BP KH l to4 protein crysta l
C rystalli zati on of RD -BP was fir t attempted u ing our lab 's ex i ting \\-i ld-type. fu lllength

RD-BP plas1nid con truct and stand ard prote in pro du ti o n protoco L v. hercb) a

denaturing prote in purifi cati o n scheme is empl oyed and refold ed by removing urea \\ ith a 3stage buffer exchan ge process. ll owever, re fo ldin g in the di a lys is units appeared to have an
effecti ve concentrati o n limit o f 0.7 mg/ml w ith additi o na l prote in prec ipitating out of so lution.

107
ombin d

ith increa d difficult con

ntrating (clogg d filter-concentrator ) wh n thi limit

wa approached, thi

tandard pr parati n method wa det rmin d to b un uitable for cry tal

pr t in pr parati n.

dditionall

nt to

r.

~ rg

tr nadka' laborat ry at th

d mea ur , a 15 ml denatured ample at 0.7 mg/ml was
m er it

facilit wh reattempt to refold the protein ia dial

f

riti h

olumbia X-ray crystallography

al o fail d for imilar rea ons.

I d cid d t con truct an alternative con struct whereb th RRM d main wer removed
(figur 4.11

), a th re are large flexible region connecting the RRM domain to the Kl-11 to4

RNA-binding module that l rea oned may affect the ability tore~ ld by our standard method . A
PCR amplicon co ering th KH I t KH4 d main r gion wa generated for use in ub-cloning
and pro ed t b th e pect d ize ( 1,122 nt) and of high purity as determined by agaro e gel
analy i (figure 4.11 A). Both the D A in ert and p
dig ted with

41 c( + ) pia mid vector were ucce

d I and Hindlll producing ticky end and ligat d with DHSa

fully

.coli

tran formant able to grow on LB-agar kanam ycin plates. Pre ence of the KH 1to4 insert wa
confirmed u ing a digest analysi again on mini-prepped plasmid amples from uccessful
transformants (figure 4.11 B).
Expression in BL21 E .coli cell s successfully produced protein with imilar purity and
expression levels a with the full-length con truct. Direct ize com pari on of the full-length
CRD-BP (67 kDa) and KHlto4 truncation confirms the ab ence ofRRM1 and RRM2 domain ,
appearing at the expected 43 kDa (figure 4.1 2A). However, dialy is-ba ed refolding u ing our 3stage buffer exchange method again proved limited to appro imately 0.7 mg/ml. implying that
the limiting aggregation factor is likely present within the KH 1 to KII4 region.

108

5( 00-

~ 000-

•

I ( 5)-

10005065000400-

r- tennmator 609 ~

c

La

perator l - - _ r - prom tor -l-5l

pET41c - KH1to4
6166 bp

Figure 4.11. Plasmid pET41c(+ )-CRD- BP KH1to4. (A) DNA agaro c gel analysi s ofPCR

produ t pannin g d main KI I 1 to KI 14 of ' RI -B P ge ne. (B) Endo nuc lease di gc t anal) 'is of
pE 4 1c( +)-C RD-BP KI 11 to4 using Nde l and J li nd Ill rc 'Oived on 1.5° o agarosc gel. Uncut and
cut v ctor band can he seen at 5 kh and in ert DNA hands appear.' at c~pcctcd I I ~- hp (C )
Pi a mid map of pET4 1c( -+ )- ' RD-BP KJ II to4 .

109

ati e prot in purifi ati n
purifi d b thi meth d
111

a

ub equentl pur ued a de cribed in 4.1 .8. Fraction

ntain d ignifi antly le

protein. with mo t of the protein in th

lubl fra ti n found in th p llet (figur 4.12 ). imilar r trictive re ult for native

purificati n f

R - P in r c mbinant

tern ha e been dem n trat d previ u ly with

Pr kipcak e/ a/. ( 199 ) and Du era/. (... 00 ) that pecificall indicat that most of the protein
nd up in th in

lubl fraction .

B

A
kDa
43--

~ative- KH lto4 0.2 mg/ml

Refolded - KH 1to4 0.2 mg/ml
Ill"

522
5.00-

260 nm

1.40 -

I
!

::~
o.eo

j oo}z

50·

060-

o.o0.20-

o.oo-

1 ooosoo.oo-

mWndonvll>
zio z9o ~ ,io l2o m
nm

31 OBDI4 .00111121'9

Figure 4.12. Comparison of native and denaturing KHl to4 protein purifications. (A)

Denaturing protein purification - refolding facilitated by gradual removal of ur a via dialy 1 m
3-stage buffer exchange cheme. Below: UV -spectrum of protein ample showing dominant 280
nm peak; 260/2 80 : 0.61 (B) Native protein purification with imidazole eluti on. Below: UVspectrum of purified protein howing dominant 260 nm peak: 260/280 : 2.08.

While the reduced yield wa not entirel y limiting. endoge nou s bacteria RN

\\ere found

to co-purify with the protei n in large amounts a ' determined by UV -spcc troscop; (figur 4.12B)

110

whi h

uld lik 1 interfer

ith r tallization and reduc

lubility .

pproache including

n n- pe ifi RNa e

uch a b nz na e tr atment t rid th pr tein f R A were c n id r d,

ho e

ith the 1 wer ield f thi method ,

r c mbin d

approach

ith uperior re ult .

periment c ntinued t find anoth r

third approach call ed n-

lumn r [! !ding was tri ed. llere th

denaturing purificati n tep en ure olubili ati on of th e large oth rwi e in oluble fraction of the
protein . Hi -tag binding to the
to 0 M urea vera m

i- T

co lumn occur in

M urea, which i then lowl y reduced

th gradi ent prior t eluti on u in g co ne ntrated imidaz le a d scrib din

cti n 4.1. 9. R uJ t fr m thi m thod v...e re e. cell ent.
approach d 2 mg/ml with n

ff th e co lumn. protein concentration

1gn of pr cipitatio n. As we lL optimi zati on of th e proce s

e entuall y1 lded >95°/o purity ample vvith mor

trin gent wa h-step . and furth er

concentration up to 10 mg/mJ proved impl e with no clogged filters (fi gure 4.13 ). To en ure
this alternati e method v... a providing active protein till capabl e of binding RN A, fluore cence
polarization (FP ) technique was emplo ed. Thi technique expl oits the tim e-scal e whereby a
fluorophore re-emit plane-polari sed li ght and can detect the increa in gly random di tributi on of
un-bound, free tumbling labelled-RNA co mpared to protein-bound RNA measured a ani sotropy.
A fluorescein-labelled KRAS RNA probe based on previou EM A work (2.3.2) wa
synthesized by IDT and used to as ess its interaction with the CRD-BP KH1to4 protein ampl e.
It is evident according to the FP analysis in fig ure 13A that unlike the BSA negative bind ing

controL the degree of ani sotropy with full-l ength CRD-BP and the KHl to4 con truct increase
considerably with concentrati ons similar to those seen in

M A e ·per iment . Ind eed upon clo e

inspection. it can be seen that the KH 1to4 vari ant appears to bind th e RN

substrat \\ ith hi gher

affinity than the natural form . Calcul ation of the di . oc iati on con tant fro m an RN

bou nd

fraction plot (fi gure 4.13B) reveal a 3- fold in crease in substrate arfinity for the Kill to4 variant

111

( 11. ± 7.1 nM

mpar d lo 9 .6 ± 4. nM), po ibl implicating the RRM domain as a

re tri ting 1 m nt in binding.

A

B

200

.s>- 150

'0

a.

:;J

I

Q)

__.,_ KH1 o4

c:
Q)

--BSA

u

•

u

I

ctl
04
._
u.

CRD BP Kd 311 8

50

_.__ KH 1to4

1 1

Kd=93 6t34 1

02
o ~--~----~----~----~--~

0 ~--~----~----~----~--~

0

:?00

400

600

800

0

1000

200

400

600

800

1000

Protein concentration (nM)

Protem concentration (nM)

c

I
£l

c

VJ

::I

/

0
;

Q)

u::

/

~

u

...0

/

/

z
a: 06

-f '-- CRD-BP

I

100

08

0
.0

~

VJ

Cll

c

~

0
~
0

c:

---

--

~

12

D

RD-BP KHJt 4
kDa

- 250

- 130
- 100
-70
- 55

- 35
-25

- IS
Figure 4.13. CRD-BP KH1to4 construct binding activity and protein crystal. ll fi gure
pertain to the CRD-BP construct containing only the KHl to KH4 domain binding module,
excluding theN-terminal RRM 1 and RRM2 dom ai n . (A) Fluorescence pol ari zation/ ani otropy
RNA-binding activity as ay. CRD-BP KH 1to4 confirmed to bind flu ore cein-1abelled KRAS
RNA probe u ing fu ll -length CRD-BP as a po itive contro l and B A a a negative binding
control. (B) KH1to4 and fu ll -length CRD-BP Fluorescence ani otropy data u ing KR S RN
probe co nverted to fraction RNA bound plot for di sociation con tant determination. (C) 0 , PA
analys is of co lumn-refo lded RD-BP KJ I 1to4 protein. (D) Po ' itive hit protein Ct). tal of
Kl I1to4 in 0.4 f.! I moth er liquor.

112

Perhap a fun ti n f the RRM d mam i t confl r enhanc d electivity .G r a giv n
R

ub trat , a ting a n gati e r gulat r

f binding in om circum tanc . 1t ha be n

d m n trat d that the RRM d main d n t c ntribute t a l wer Kd, h we
acr

th

ari u orth l g impli

r gulation po tulat d h re.

a u eful functi n, n

r their conservation

f which c uld be the negative

umer u Kl I It 4 pr tein preparation were delivered to Dr.

tr nadka' X-ray cry tall graphy lab rator for cr tallizati n trial . lli gh concentration
ample

f I 0-11 mg/ml pr t in \\er

hipped from our lab [i r direct entry into crystallization, a

well a a 1 mg/ml prcparati n v.here a ubscquent thrombin digest allow d for removal of the 8
X hi -tag pri r t c ncentration and cry tallization . Two trial cone ntration were etup (4 mg/ml
and 8 mg/ml) B r cry. tal crcening v.ith three eparate crystal cr ening kit : Qiagen
uite, Qiagen P

Ia ic

T uite. and Molecular Dimen ion JCSG-Piu ™ MD1-37 uite. N protein

cry tals were produced from amples in our original buffer (20 mM NaH 2P

4· 200 mM Na

L

pH 7.4), ho e\'er a ingle cry tal wa achieved that wa dialyz d into 0.2M lithium ulphate,
0.1 M Tri - I pH 8.5, 40°/o v/v P G400 buffer and confirmed to fluore cc under

V li ght,

indicating that it was comprised of protein (figure 4.130). Unfortunately the phereolyte type
crystal bore un uitable characteri tics for further study. including a twinned structure and thu no
diffraction pattern was obtained despite the lengthy attempt. Further attempt to reproduce th
crystal did not prove successfuL however data obtained from subsequent pre-cry tallization
optimization tests did reveal some u efu l information pertinent to future attempt at crystalli~:ing
thi s protein (figure 4.14 ). Buffer factor including sa lt, pH. and the buffer chemical it elf v.erc
varied in a thermal aggregation experiment, whereby stabe li zing elements \\ere ~ ystematicall)
tested for ability to prevent non-specific protein interaction. From these experiments conducted
at the trynadka laboratory, we discovered that KI II to4 CRD-BP is ' tabili7 ~d significant!) b)

113

c nferred a profound tabilizing ffi ct. High alt up t 500 mM Na 1 wa also

lower pH; pH
d termin d t al

111 r a

the temperature f aggr galion, albeit to a I

t fthe alt ~a al

r e tent than the pH

ffect, and th

f~

buffi r

h i e had a tr ng impact n protein tabilit a w lL with M

dep ndent on the buffer y tern u d. The pecific

gr at t d gree of tabilizati n a cording to the temperatur of aggregation
endea

providing the
p riments. Future

r int cr) tallizing ' RD-BP will ertainl benefit from this kno ledge, a it is likely an

id al torag and tran port buffer. a \Veil a being optimal for cr tallization trial .

114

1(

A

0

.~
Ill

-

100 n MIll PI') pi I I 0. '•()nM N. 11 I

-

100 mM <JI

-

IOOrnM M£'1 pll fi 0 J'iOr M

Cl

"'..ao ,,

-

100 mM I f PIC, n~ 7 0, J'JI')nM

.1U

-

100 mM Ill P( '> plf

'lOrnM

0

ct

-

100m

Glycin pH !J 0 }'J()nM

JCI

c

..,<U 1':;:]
c
c

.2
..,
<U

gs

Q.

=
........
c

n plf 'J 0 'JOn

0.

100 mM Ml S pll b 0 ~
)~

-

IOOmM HIP(<; Jllf 7 0 'J

mM
nM

(I

.tO

0::

u

100 mM {,I

ne pi 9 0 ~M

0

40

30

so

41

Temperature (°C)

B

r

-u

Ill

-

4'

0

41

I

Figure 4.14. Thermal aggregation buffer optimization for CRD-BP KHlto-1 . ( ) Prot in

aggregation a a function of temperature in various buffer oluti on .. Data \~a produced from
pectro copy experiments conducted at the Strynadka X-ray cry ' tallograph) laborator) . (B) f agg
(temperature of aggregation , defined as temperature at v. . hich 50° o or protein is aggregated)
calculated from aggregation intensity trace for CRD-BP in different buffer sy tems.

115

Chapter 5
General Discussion
5.1 Proj ect general overview
he

I KZ famil y f RN -binding pr t in include

pr t in (IMP

K

, ZBP 1, and

ra) have multifaceted rol e including intracellular

1 alizati n, nucl ar e p rt f mRN , tran ript tabiliza tion and tran lati nal

mRN

regulati n. Their fun cti n i achi
4

g l RBP

RD -BP a well a it ortholog u

d thr ugh the interacti on of 2 N- tenninal RRM dom ains and

-t rminal hnRNP (heterogen u nu clear rib nucle protein)-K -homology domain (KH

domain for hort) whi ch fun cti on in combinati n to bind a vari ety of mRN A targets including cm yc,

LI MDRl , I F-II am ong t other (Pr kip cak et al. 1994;

and Lee 2007 ·

oubi i et al. 2009; paranese

iel en et al. 1999). CRD-BP is spatiotemporall y regulated, nonnall y expre ed

during the first 12 week of embryo gene is and disappearing into complete absence or scarcity in
adulthood (Leeds et al. 1997). Human CRD -BP (IMP- 1) i often re-expres ed in m any cancer
and has been dem on strated to be important for tum our cell urvival as ummarized by B ell et al.
(20 13).
KRAS , a well established genetic driver of m any cancers including colon, breast and lung
tumours has recently been suggested to be an important oncogenic 1nRNA target of human CRDBP, increasing expression of KRAS protein in colon cancer cell line (M o ngroo et al. 2011 ).
Furthennore, while m any studi e foc us on mu tated ga in-of-functi on KRA variants uch a
KRAS

12

v which con titutively activa tes downstream targets such a the PI3K,

MAPK pathway , gene amplifica tion and overexpr
aggre iveness and poor re ponse to anti-

FR, and

i n of wild-type KRA al o a o iate with

R therapy in col r ctra l, ndometrial and non-

116

mall c lllung cane r (Valtorta tal. 2013; Birkeland tal. 2012; a aki et al. 2011). De pite
num r u and di
ribozyme , and p

attempt t targ t th KRA pr t in u ing mall molecule , antioncogene
t-tran lational modification inhibit r , th re remain no effective mean

treatm nt D r KRA dri
th r i ne d :tl r n

n cane r (r

1 appr ach

iewed in Jancik et al. 2011 ;

pearh ad an att mpt t inten-upt the

pecific anti en e ligonu l

5.2

llin et al. 2014) . Hence,

t an ti -KRA therapies. The goa l of thi inve tigation wa to

confirm and furth r charact 1ize them lecul ar interacti n betw een
a in cell , a well a

f

RD-BP and KRA in vitro

RD-BP-KRA mRNA interaction

tide molecule .

M appin g of the CRD-BP bindin g sites on th e cod ing region a nd 3' UTR of KRA

mRNA usin g th e electroph or etic mobili ty shift assay

Determining the binding potential ofKRA mRNA and determining a small er segment
of the mRNA molecule that retained affinity for CRD-BP was an important endeavor. It offered
the opportunity to confinn previously reported results regarding the interaction between KRA
mRNA and CRD-BP protein using an alternative experimental method, and also to nan·ow down
a more precise sub-region containing the nece ary sequence infonnation required for binding.
Knowledge of a minimum binding sequence not only contributes to the goal of detennining an
mRNA target consensus sequence, but opens up options for other methodologi es to tudy CRDBP-KRAS mRNA interaction. For in tance, to use the flum·e cence polarization spectroscopy to
study nucleic acid-protein interaction, nucl eic acid mu t be larger than 90-mer (Hafner eta!.
2008).
12

Tlu·ough the use of PCR and in vitro transcription, [ P] -labelled fra gment
mRNA were synthe iz d and tested for

RD-BP affinity u ing M

f KRA

f the ~ i cour

117
fragm nt

ach appr

imat ly 200 nu leotide in length fl ur were di c v r d to bind with

m d rat affinity in th nan m lar rang ( ecti n 2.3 .2). Thi finding confirmed previou ly
publi h d r ult that indi at IMP-1 (9 % id ntity t
coding r gion and al
attachment ite
atta hm nt it
ik

the 3'

TR (Mongr

ithin both fthe e mRN
1

in the

eta/ . 2011) but al o lucid ated multiple
reg1 n .

n t urpri ing and thi notion i

a tal . (2006) wh r

RD-BP) bind KRA mRN

umer u

RD-BP-KRA mRNA

upp rted by Niel en et al. (1999) and

e eral IMP-1 binding ite were found n the 5' UTR of IGF-Il

mRNA and '

TR f

FMR1 mRN

in which multiple IMP1 protein were ugge ted to bind in the coding region and

3

44 mR

re p ctiv ly.

imilar conclu ion w r drawn regarding

TR (Rackham and Br wn 2004 ).
KRAS n1RNA ub-regio n ' ' po

ed the hi ghest affinity with a measured kD of

177 .9± 18.4 nM, with the other binding fragments :

, D and F having at least 2-fold lower

affinity (figure 2.6). This fragment size compare well to literature defined target (table 5.1 ).
Further truncation analysi of the KRAS-A sub-region identified ' KRA -A34 , a 93 nucleotide
fragment which binds CRD-BP with similar affinity. KRAS-A34(b), a smaller 57 nucleotide
region located near the 3' end of the parent KRAS-A fragment, also retained ability to fmm a
larger molecular weight complex with CRD-BP, albeit with reduced affinity as detetmined by
EMSA (figure 2.18). It also became apparent that with the progressive whittling down of KRASA to KRAS-A34(b) sequence length has a general impact on the overall affinity of this RNA
target and specificity only becomes apparent when comparing RNA fragment

f imilar ize a

inKRAS-A12/KRAS -A34 (92 and 93 nt, respectively - figure 2.13) and KRA -A34(e)/KRA A34(b) (57 and 57 nt - figure 2.17 and 2.18). Indi ations that other targ t may follow thi
pattern stem from a imi1ar truncation study u ing P I, a KH domain

ntaining prot in in

118

hila r lat

al

t

R -BP ,

h r additi nal

qu nee flankin g a c r

lement

ntributed to

r Kd ( hmi l t al. 2

Table 5.1. VI KZ mR

tar

RNA-binding protein

t lZ

mRN
I

IMPl

target
-II

Target size (nt)

Reference

162

Niel en(l99)

CRD-BP

-myc

182

Bern tein ( 1992)

IMPI

Tau

600

A tl a (2004)

ZBPl

~-actin

54

Farina (2003)

IMPl

FMRl

7

Rackham (2004)

IMP- I

P BP

22

Patel (2005)

YglRBP

gl (

340

Mowry ( 1992)

~)

IMPI

Hl9

17

Runge (2000)

PSI

II

69

Siebel (1990)

KRA -

4(b) RN

quence wa analyzed u ing mfold RNA-:fl lding oftware to a e

tructure recogniti n in the :fl nn of tern I op m tif: .

po ibl

mo l cui c mbin d with inclu ion of putati

- ' 5 ._
'

C

GACA

-3 · and 5'-

tructured core of the RNA

bindin g I ment id ntifi d fr m literature: 5'AC

- ' (2.3 .4) ub equently a ll owed the

identification of a 44-m er fragment u ing flu re cence polariza tion pectro
2012 Git and

tandart 2002, Ro

opy (Patel eta!.

eta/. 1997). Development of the KRA -A34(b)-ba ed

fluore cence polarization probe wa in trumental in the binding capacity a e sment forth
truncated

RD-BP KH 1to4 protein, a well a o ther molecular end ea or in ur lab ratory

including mall molecule inhibit r cr ening. Notably none f th e m ention d
appear to be able t bind the protein by it elf, how ver it ha be n propo

quence Iem nt

d th at they do

contribute in tandem with other elements, to a hi gh affinity interacti on lik ly ea h mediated b a
eparate KH d main opening th e pos ibility of man combination binding ·cqu nee ( hao eta/.
2010) .

119
Whil a mall ub

t f

qu nee lement pr p

d do help explain KRA binding,

many r c gniti n elem nt pr claim d :fl r th VI KZ pr t in family don t xplain KRA 4(b) binding
high

f

nu 1

ith

RD-BP .

id c ntent and

tably, a fa irl y con i tent n ti n i that RNA target contain
it and tandart (2002) howed that ind eed poly- /p ly-U

h m p lyn1 r ar m re capabl of c mp ting away Vg lRBP -VL _. mRNA interaction in vitr
KRA -

4(b) attribute

nl y 44% f it c mp

iti on to uridin or cyto ine, an alternative

m chani m than imply high

c ntent mu t be re ponsibl e fo rth binding of thi region. It i

lik ly that multipl m echani m

xi t fo rb th binding and granul e fo rmation in vivo . KRA a

well a c-m yc both app ar to form a hom dimer with an intermedi ate band between free RNA
and the larger compl

(figu re 3.2) while ZBP1 how no uch intennediate band and

corre ponding ize exclu ion chromatography data al o ugge t a monomeric specie ( hao et
a!. 201 0). The ame group al o found no cooperativity with a Hill coefficient of approximately 1,

in contra t to KRAS RNA binding (Hill coeffic ient approximately 1.9 - see fi gure 2.8) as well
binding to c-m yc and CD44 RNAs (Barne et al. 20 14 ). Krissinel and Henrick (2007) computed
a free-energy of dis ociation calculation and al o suggest that homodimerization of VI KZ
proteins is unfavourable and thus the phenomenon may be an artifact of the experimenta l
conditions u sed in EMSA as suggested previou ly by N ielsen eta!. (2004 ).

5.3

Testing the specific anti-sense oligonucleotide inhibitors again st the CRD-BP-KRAS

mRNA in teraction in vitro using EMSA
The benefit of elim inating

RD-BP interacti n wi th its vari ous target mRNA are largely

apparent (Iatmidis el a!. 2005 ; Elcheva eta!. 2008 ; Mongroo eta!. 2011 ), however IMP 1
knockdown in T47D and MDA23 1 human breast cancer cell cau d increa ed in a iven

due

120
t r du ed I

alizati n f

-cadh rin mRN

( u t a!. 2011) . -cadherin, along with another

RD-BP target ~- a ctin ha a c ll-cell adhe i n r le, fannin g the core structure of focal
adh ion that redu ce migrati n and mitiga te inva ivene

(B erx t a!. 1995 ; Birchmeier and

Behr n 1 94) and becau e IMPl i able t bind and tran Jationally repr ss E-cadherin mRNA,
I KZ pr t in m ay ha

anti-pr l i~ rati v function a well. Furthenn re, direct knockdown of

RD -BP in hum an K 562 Jeuk mia cell r ulted in increa ed cell proliferation (Liao et a!. 2004 ).
Whil complica ting a th rap utic trategy con id erably, the mixed nature of CRD-BP ' progr wth and anti -gr

th fac tor ju tifie the target- pecific approach our lab pur ue , as opposed

to imply trying to down-regulate or otherw i e inhibit CRD -BP function entirely, although many
cancer would till r p nd po itively from a general inhibition of

RD-BP due to increased

chemo- en iti ity ( raig eta!. 20 12). A target specific approach would allow a per onalized
treatment of CRD-BP-po itive can cers, preferenti ally driving down pecific oncogene
expre s1on.
Work on identifying effective, pecific molecular inhibitors of the CRD-BP KRA

mRN A interaction was an ultimate goal of thi s re earch and the attempt was spearheaded by
designing multiple 23 to 24-mer DNA-based anti - ense oligonucleotide molecule that were
complementary to sub-regions spanning the entire length of the KRA S-A region in a contiguou
fashion (Table 2.8). EM SA assays were used to test the effecti veness of each oligonucleotide at
32

breaking the interaction in vitro, as evidenced by the loss of super hi fted [ P]-labelled b und
fra ction RNA in compl ex with CRD-BP; an approach adopted from King eta!. (2014) . By
titrating increasing amounts of anti-sense oligonucleotide mol cule in to a CRD-BP-KRA
mRNA binding reaction, from 1 to I 000 fo ld mo lar excess, two candidate were re ealed a
good molecular inhibitors. Labelled a SM-6 and M-7, each pr

ed c n id rabl m r efTe tive

121

than any f th

th r candidate m lecule , panning KRA nucleotide po ition 116- 138 and

1 9-161 f the

ding r gi n re pectiv ly (fi gure 2. 19). M-2 di played virtually no ability to

c mp t away th binding int racti n a det rmined by EM
hi gh

qu enc

p cificity £ r int rrupting th KRA - RD-BP interaction under the e

e pcrim ntal c nditi n
the olig nucle tid
the KRA Altemati

and i considered evid ence of

t

figure 2.2 1 and 2.22 lane 13- 17).
th

M -6 and M-7 p

tran cript and thu pre
ly, a ignificant RN

binding c uld al

iti on

nt d proper

omplem ntary annealing of

ffectively rna ked the RNA nucleotid es in
RD-BP acce s to the binding sequ ence.

tructure change induced a a re ult of the oligonu cleotid es

explain the effect, however thi model require that

RD-BP recogni ze

pecifi c RNA tru ctural element a a mechani m of binding whi ch i as of yet unconfirmed.
Regard!

of the mechani m, the re-app earance of this region a a potenti al CRD-BP interaction

ite furth er support earli er KRA mapping finding which sugge ted that the KRA -A34(b)
region covering nucleotide positions 129- 185 fa lls within the M-6 and SM-7 coverage, and
likely contains important binding information (refer to figure 2.24 for graphic representati on).
The SM-6 and SM-7 DNA molecule were sub equently used as po itive controls for inhibitor
of the KRAS-CRD-BP interaction in small molecule screens, as well as erving a potential
proto-drugs in cell-based KRAS inhibition as ays with SM-2 continuing to serve as a negative
control (Wang et a!. 201 5).

5.4

Examining CRD-BP KH domain requirements for binding KRAS mRNA in vitro

VI KZ fan1 ily of proteins contain 2 N -terminal RRM domain (labelled RRM 1 and 2)
and 4

-terminal KH domains (labell ed KH 1 thro ugh 4). The imporianc of d tennining th

sub-set ofKH domains required for high affi nity bi nd ing to KRA mRNA t m from a ba ic

122

1en
1m

p r p

ti e in und r tanding th

t1ant in a plied re ar h a it pr

inhi it r again t th

n genic fun ction f RD-BP. uch kn wledge i al

id

riti cal in[! nnati n for the de ign of p cific

n g ni c fun ti n f RD -BP .

differ nt RN

ha e dif£ r nt KH d main req uir m nt il r bind ing, m lecul
KH d main rn a pr di ctab l

d main r qu irem nt il r

nl affe t a ub et f RN
hR

tran cript appear t

efil cti e at bl ekin g a pecific

targ t . Kn wledge f th

uniqu e KH

m I ul w uld all w Ia ifi ca ti n f the vari u RN A

and p tenti ate pr di ti n f th multipl e targ t a given inhibit r r c mbination thereof w uld
ha
T appr ac h th
multipl e pla mid

pe ific KH d main requirement il r RD-BP t bind KRA mRNA,

n truct con taining RD-B P variant each harbouring either a ingle

mutation in n of th il ur KH domain (KH 1, KH2, KH3 and KH 4 ), r combin ati on
GXX

XX
f2

mutation -each imu ltaneou ly pre ent in 2 eparate KH d main (KH 1-2 , KH 1-3 ,

KH 1-4, KH 2-3, KH 2-4 and K H3-4) a de cribed in greater detai l ( ecti on 3.1. 1). Each

XXG

motif exi t a a fl e ible loop between the 1st and 2n d a-helix f type I KH domain and directly
interact with the RNA target of KH domain containing protein ( ri hin eta!. 2001 ). Mutation
of the fir t glycine to an a partate (table 3.1) ha previou ly b en demon trated to ab li h the
ability to bind RNA, pre umably through teric hindrance from the larger ide chain, or from
charge repulsion - both a partate and the RN

backbone po e

negative charge (Lew i ct a/.

2000; Lin el a /. 1997; Paziewska eta/. 2004 ; Fenn eta/. 2007 ; Jone. and chedl 1995).

Furthe1m ore, limiting the mutation to 2

XX

motif: i rationa lize by the fact that no R A

target ob erved thu far ha been demonstrated t bind
ach mutant pro tein wa then subj ected to M

ith d1ything le , than 2 Kll domain,.

binding a , ay u ·ing the KR , -

probe and

123
th

f£ ct ach

mutati n c mpared t wild -type then a

r g1 n d t rminant - wh r

RD -BP bind ) RN

u ing any f th e ingl
2

RD (coding

wa utilized a a po itive contr l.

RD -BP KH mutant al ne abrogated th ability f

RD -BP t bind th [ P]- lab lled KRA RN
KH 1 wa r p at dl

d. Th c-myc

alm

t entirely (fi gure .1 and 3 .2). Mutation in

een t impact the bind ing t a lightl y le er ext nt, but wa

till required

£ r th high affinity binding di p layed by th wi ld -type control. Hollingworth et al. (201 2)
imil arl y c n lud d that th K.H l d main played a miti ga ted role forK RP RNA binding (an
RNA binding pr t in im ilar to ZBPl in th e tud y, c ntain.ing 4 type I KH domain ). The KH4
mutati on nth

th r hand appeared to ha e the large t effect, ind epend ently reducing the bound

fra ti n of RNA aln1o t entirely. Intr du ction of a econd mutati on in any combination of KH
dom ain proved ufficient to entirely inhibit any trace of binding under conditions u ed in the
electrophoretic a ay . This re ult was expected given the impact of a single KH mutati on on
KRAS binding.

5.5

IP-qRT PCR analysis of CRD-BP KH domain requirements for KRAS mRNA
binding in HeLa cells
Complem entary exp e1iments using

RD -BP-FLAG immu ne-precipitation and RT-qPCR

also strongly agreed with the KRAS EMSA data. Mammali an expression vectors (pcDNA)
harbouring each of the KH mutant variant genes were transfected into HeLa cell and ubj cted
to anti-fl ag CRD -BP pull-down combined with RT -qPCR u ing KRA

p cific primer ( ction

3.2.3 ). R esults revealed a similar pattern as with EM A experiment , with ingle and double KH
mutants displ aying relatively less association with KRA mRNA in cell than the wild-typ fOJm
(P=0.01) . Notabl y w hile the KHI mutant precipitated appro imately 10-fold le

KRA mRN

compared to the wild -type, the other single KH mutant were nearly I 000-fold I

, miiToring th

124

pre 1 u ly di u ed
th

P rim nt

M

r ult (figure 3.5). Intere tingly the conclu ion drawn from both

upp rt an ti n that all KH d main are requir d for RNA binding, at lea t

to th high affinity r gi n f KRA mRN
Thi
mRN

nclu i n i ech

d wi th

amined in thi

tudy.

g lRBP rep011edly binding the VL

equence in Vg l

u ing all £1 ur KH d main ( it and tanda11 2002). imilarly, IMP -1 truncation

e p riment by

iel

n el a!. (2002) dem n trate clearly that neither th RRMl -2, KH 1-2 or

KH -4 dido main were capable of hifting H 19 RNA u ing M A , with the minimum
requir m ent being all fi ur KH domain . In th
inhibitory effects on granu le formation in

arne study, removal of any KH domain also had

IH 3T3 cell . Bridging the evolutionary gap to the

Dro ophila ortholog P I, once again only the pre ence of all four KH domains wa conducive to
RNA binding, and an extended interaction urface along the tandem KH domain wa propo ed
that highlighted the inter-KH domain geometry as in important element in binding (Chmi el et al.
2006). In contra t other group have focussed on the KH3-4 didomain and show that KH3 and
KH4 are nece ary and sufficient for binding of some mRNA targets (Farina eta!. 2003; Nielsen

eta!. 2004; Chao eta!. 201 0) . Intere tingly, Barnes eta!. (20 14) wa able to how that
simultaneous mutation in KH3 and KH4 did not abrogate binding ofCRD-BP with CD44
mRNA, yet taken together, the culmination ofCRD-BP/VICKZ experiments focussing on KH
domain involvement in RNA binding strongly suggest a sub trate specific model of binding with
perhaps two classes: i) KH3-4 didomain dependency and ii) KH 1, 2, 3 and 4 dependency.

5.6

Structural analysis of CRD-BP KH va riants containing point mutation in the
GXXG motif at the KH domain
As the introduction of mutations in a protein can ha

a/~ topo logy, an important feature of the

drastic folding on~ equ en ·e for

RD -BP KH mutation e pcriment wa

tn.1 ·tural

125
anal

i u ing circular dichr i n1 ( D) pectr polarim try of each mutant variant. In ection

4 .2. 1 th

D p

b

ary minimall y b tw

en t
a

n the pr tein wi th only mall variation predominantl y at

tral cm

th

RD-BP mutant wa

time ( accumul ati n ) and xperiment compl et d in triplicate; the re ulting far-UV
s ar repre ntati e of 24 can and while the averag d spectra did how very small

de iati n b twe n ampl
H n

RD-BP, ingl KH mutant and double KH mutants can

1 ngth b low 205 nm and high r than 23 0 nm (fi gure 4. ). ach

canned
p

tra forth wild-typ e

, the v ra ll [i ld i lik ly identical and within enor of pi petting.

D e idenc for pr per KH d mai n fo ld ing strongly sugge t that the ob erved

change in KRA affinity with the ari u GXX

mutations ( ecti on 3.3 .1) were not due to

perturbed global tru cture or de tabilization of neighborin g CRD-BP domains and further
validate the relati ve ro le for each KH dom ain in binding KRA mRNA .
GXXG t GDDG mutation in the KH 3 and KH4 domains within ZBPl (GKKG and
GKGG to GDDG re pectively, to b e more specific) has been demonstrated to not have a
significant impact on dom ain structure as detetmined by superimposition of 15N- l H HSQC
spectra and CD thermal denaturation experiments. In the same paper, all four KH domain in
KSRP , a related RNA binding protein with hi gh sequence similarity to CRD-BP, were similarly
shown to m aintain stru ctural integrity with the introduction of the GDDG doubl e mutation
(Hollingworth eta!. 20 12) . In PSI protein KH domains , a highl y conserved isoleucine po itioned
irmnediately adj acent to the canonical GXXG motif was mutated to asparagine, and CD analy i
of both secondary structure content and thermal sta bility also revealed that no major t1uctural
changes resulted. Structural tability as ev id nc d from the

RD-BP mutation

perim ent

presented here and from other groups, strongly sugge ts that whi le the ubiquitou G .~ XG loop i

126
ab

lut ly riti cal :D r tabl RN

int ra ti n, it doe n t app ar to contribute to the ttuctural

integrity fKH d mam .

5.7

Monitoring ligand-induced conformational change in full-length CRD-BP and the
KH1to4 truncation var iant

Pr t in c nfi rmati n change are an inherent property of many nucl eic acid (NA)
bindin g pr tein

u h

tran cripti on fa tor , whi ch can und ergo large ca le shift in a -helical

content upon binding
D

, uch a

equ nee (M 11 t a/. 2002).

REB - a b/Zip type tran ctiption factor that bind a 32 bp
D pectro co py i a choice technique for anal yzing

conformation change in bi omolecules, e peciall y proteins w here concentration and stability
re traint can be limiting factor . CD pectro copy also allow the examination of secondary
tructure hift in many pro teins with regard to binding of li ga nds, including sm all rn olecul

,

bio-polymer and nucleic acid and is u ed extensively in thi area (Gray 201 2). As CRD -BP is
an RNA binding protein, we investigated the possibility of it und ergoin g a conformational hi ft
in secondary stru cture upon ligand interaction w ith KRAS RNA. This avenue, if fruitful would
open up additional mean of as essing sequence specificity for CRD -BP by CD monitoring
secondary structural changes associated with high affinity RNA target , as detenn ined by EMSA
or other methods. Fmihennore, characterizing CRD-BP structural changes upon binding would
complem ent crystallographic and NMR data with and without RNA pr sent for the entire
VICKZ famil y of proteins.
To approach the assessm ent of a possible
far-UV (190-25 0 nrn)

RD-BP stru tu ra l change upon RNA binding,

D spectra were gathered fo r both th wild -type protein and a KH I to4

truncated mutant containing only the fou r KH domain RN A binding module . Truncated

RD -BP

wa generated fo r thi s puJ1JO e, as it ha been how n th at the RRM domain d n t b th m 'lve

127
bind RN

t any ap r ciabl affinity (mM rang ) and thu it wa rationalized that pre nee of

th RRM domain w uld lik ly nly mitigate any change in
1 en t a!. 2002).
int

r both

lution at pr t in:RN

u ing a J a co J- 15

RD-BP ariant

D ignal ( it and tandart 2002 ·

a library of potential RNA targets were titrated

m lar ratio varying from 16: 1 t l : J and ub equently catmed

D p ctr polarimeter t

arch for evid ence f econdary tructure

change . Du e t th inherent

D ac ti ity f nucleic acids, oluti ns of the RNA sub trate alone

were al

pectra ubtract d fro m the protein-RNA c mbination pectra u ing

cann d and the

an in-hou

ftware (Ja co pectraManager) o a to i olate changes a sociated with

Examinati n of th various full-length and KH lto4

RD-BP .

RD -BP pectra revealed intere ting

finding . Fir tly, both ver ion of CRD-BP di played a rapid reduction of CD ignal at the 208
nm and 222 nm peak up on additi on of RNA (figure 4.4 and 4.7). The chan ge in the pectrum
equat s to a reducti on in ov rail a/~-topology, achieving a more unstructured fonn.

uch change,

while le s comn1on among NA-binding protein (as opposed to und ergoing an increa e in
secondary structure) ha been reported previou ly for the human tt·anslin protein. Translin binds
both RNA and DNA targets with a particularly high affinity for mi crosatellite repeat and
telomeric repeats and reduces in total a-helicity from 67% to 50% upon binding as detennined
using CD spectroscopy (Jacob et al. 2004; Kaluzhny eta!. 2005). However unlike translin, the
reduction in total a-helicity in CRD-BP was mo tly non-specific, as non-binding negati e control
RN As were also able to generate the effect.
Upon close exmnination of the protein-RNA CD spectra howe er, there wa a trend for
the hi gher affinity RNA to more effectively induce the lo s of a-helicity at equi alent
concentrations (figures 4.5 and 4.8 ). Thi wa the case for all te t d RNA , and th ere uJt
suggest that an equilibrium exists between bound and un-bound

RD-BP conD nnational form ,

128
with hi gh r affinit targ t
rall f£
pr

t

hifting

RD - P t a b und tate m re readily. Howev r becau e the

a g n rall y n n- p cific, it i

rganizing it elf in the pr

nc

f any R

u nc , p rhap ~ r 1-dim n i nal
binding pr t in ha

th e

RD-BP i capabl e f tru cturally

ttling d wn n a hi gh affinity

annmg r ther purp

. lnd e d th r nucleic acid-

o d n n- p cifi c pre rganiz ing capacity,

a t tran cri pti n fa t r

4 in whi ch n r lati n w a

vid ent betw en

N4

ub tr t affin it and ind u ti n [ in r a ed a- heli city. The reported increa e in

a-heli it did h
ca

prior t

be n h wn t ha e th pr
ti c

u h a th eukar

lik ly that

e er ar fr m 22-5 %depend ing n ub trate binding affinity, mu ch like th e
be ~

app ar t

r

RD -BP ( han cl a/. 2007).

D e pit th larg change in
actu all y change

D igna l, it till cannot be aid w ith ce1iainty th at

tru tural ly upon RN

RD -BP

binding becau e imiJ ar e ffec t are ob erved when

c mpl e e begin to precipitate or aggregate. A Patel eta!. (20 12) reported a redu cti n in Z BP 1
protein olubility when bound to RNA, it i po

ibl e that the ob erved redu cti on in tru cture

pre ented here could re ult from micro-p recipitation, aggrega ti on, or granul e format i n, although
no precipitate wa detected after centrifugation of the sample ( orrea and Ramo 200 ).
Furtherm ore,

hmiel el a/. (2006) perfonned imilar experiments with the Dro ophila P I

protein (73% id enti ty, 80% imilarity in the KH 1 to KH4 region) and found no ub tantial
change in secondary tructure upon binding target RNA, although the
the published work fo r further cru tiny. Inter tingly, regard le

of h w the ob

i occurring, the KH 1to4 truncation did not appear to have the am
inducti

11 of th e a-helical loss, a

D data wa n t hown in
rv d

D change

ub. trate -pecificity for

both KRA -A34 and KRA - 12 at I : I molar ratio w r abl

to achi eve the effect with equal competiti cnc. (figure 4.7 ), pcrhap. implicatmg the RRM
domain a binding regu lator .

129

5.8

RRM1 -RRM2 didomain a a nega ti ve r egulato r of RNA-binding
I pr P

a a n ga ti
binding

that th RRM1 -RRM2 did main may increa e targ t electivity by functioning

r gulat r f RN
perim nt

binding. Thi wa first evid nt after fluore cence polarization

nduct d on the KH 1t 4 truncation protein. The e exp riment

r p atedly c nfirmed (thi binding a ay wa c mpleted fo r every KHI to4 protein ampl
deli er d to th

B

X-ray

KRA RNA with nhan

ry tall ography facility) that the KH l to4 va riant wa able to bind

d affinity, wi th a Kd nearly l/3rd that of full -length

RD-BP ; KHlto4

Kd = 93.6 ± 4. nM and fu ll -1 ngth Kd = 3 11.8 ± 67 .1 nM re pectively (fi gure 4.13B) . While
RRMl and RRM2 dom ain in
tandart 2002;

RD-BP are not capable of binding RNA independ entl y (Git and

iel n eta/. 2002), they may recognize certai n equ ences and affect the KH I to4

binding module through inter-domain cooperativity. Alternati vely they m ay pl ay a more static
role providing additional steric restrictions to a lock and key fi t fo r CRD -BP RNA target .
However finiher expe1i m ents with additional RNA targets will be needed to test thi hypothe i .
If substantiated a regulatory role for the RRM domains may explain the aggres ive clonogeni c
growth of several breast cancer cell lines that have only recently been di covered to expre s a
KHl to KH4 CRD-BP isofonn (Fakhraldeen eta/. 2015) .

5.9

Assessing CRD-BP-RNA complex stability using thermal shift analysi
Stability of a protein is another fa ctor that can change with ligand binding and i oft n

used as a tool to aid protein crys tal fonnation (Vedadi eta/. 2006) .

n incr ase in the melting

temperature of a protein, or a th rmal shift, i a c mmon mea ure of prot in tability and change
in Tm are typicall y m easured using either th indire t thenno.Ouor method, whi h u

n n-

130
P

ific flu re

nt dy

m nit ring of
at 222 nm

that bind hydr ph bic prot in urface , or the more direct

D

ndary tructur (re iewed in Lavind r tal. 2009). Measuring the

a el ngth 111

t r pre

nlati

D signal

of a-helical cont nt) over a temperature range allows

the prot in un:fl lding t be m nitored accurately and a Tm (midpoint of thermal denaturation
CUI

)

h m

an b

ub

g , and al

th

qu ntly calculated fr m the melting curv to compare mutant protein ,
pr t in tability change a are ult of ligand binding (Layton eta/. 2011).

u l e pl red

wh th r r n t
b tw en R

tructural analy i revealed no clear way of distinguishing

RD -BP wa binding a p tential RNA target, and although a minor correlation
affin it and a-helicity reduction was apparent, and no ligand-induced thermal hift

xperim nt ha e been reported for VI KZ protein , the

D thennal hift a say was pur ued in

another attempt to de elop a novel method of a se sing RNA affinity for CRD-BP. In theory a
high affinity interaction may stabilize the protein significantly and be detectable with tlli
method as a hift in m elting temperature.
Initial work in developing thjs method proved difficult a the solubility of the proteinRNA complex appeared to reduce with increasing amounts of RNA approaching a 1:1
protein:RNA ratio and large amounts of precipitate could be recovered from cent1ifugation. An
increased concentration of the complex compared to the CD structural scanning was nece ary
however for determining melting temperature, due to the fact that the necessary CD ignal at 222
nm was so strongly reduced upon addition of the RNA (refer to ection 4.2 .2) . Thus the ratio of
protein to RNA was kept quite high at 16:1 and 4: 1 with protein con entration at appro imately

10 JlM. D ete1mination of melting temperature wa accompli shed as per section 4.1.6 and th
resulting Tm 's co mpared in figure 4.10 where it can be seen that while there was a small incr a
in melting temperature of 3.5 °C upon addition of KRA -A34 RN , the cf~ twa not 'P ciii,

131

a KRA - 12 a]
mpar d t th
RN
p

h

1

° and 42. 0 ° re p ctively) negativ contr 1

hile

M

wa able t

- 12 (figur 2.1 ), it wa

- , indicating th at K
a a

1 w affinity RN

h w that KRA -A 4 had hi gher
till! wer than that fth e initial

and c uld b re p n ible .G r th e imilar Tm change

ith different
ca n a!

m degr e f

- 12 may have c ntained a weak binding ite

cl

. Whil in rea ing th R

the effect f RN

ugg

tih n ting that

id nt in th

betw

(Tm = 41 .5

a a larger in r a e in Tm indicating that th r i

RD -BP th n K

par nt fra gm nt K
n t

nd tRN

r, ther

ifi it . It i

affinit .G r

d a imilar change at b th c ncentrati n (Tm = 4 .9 ° for both) .

indu

b erved

r t a I : 1 rati may b tter di tin gui h

RD-BP affi niti e , th e difference in Tm between th e hi gh and

be een to decrea e with increa ing concentration of th RNA,

tin g that th m th d b c me even le

en itive a more R A i add ed . Thi ran

mewhat c unter-intuiti e to our initial e pectation and impli e a no n- pecific effect that may
be explainabl e by non- p ecific RNA binding or ubiquitous lo e m olubility, although
precipitation in the e ample wa not ob erved . In genera l, thermal hift analy i of RD-BP a
a method for a e ing R A target affini ty proved limited, a the difference in Tm were very
mall and did not con·elate well with RNA binding.

5.10

CRD-BP KHlto4 protein cry tal
No complete cry tal stru cture for CRD-BP or any of the VI KZ protein current! exi t ,

with the closest be ing a tru cture olved for the KH and KH4 didomain of IMP I b und to a
ma ll egm ent of RNA ( hao eta!. 20 10) . Whi le thi

tructure has clucidat d an R

binding m echani m, with anti-parallel binding surfa

on oppo itc sides of the di -domain, it ha

n t ex plained the di ver ity of mRNA target.

-lo ping

r been suffi c ient to gcn 'rate a c nscnsus bm hn g

132
'-"-'l-., ..,itating th

BP truncati n ati ant wa

n ed D ram r c mpl te mod 1. For thi purp se, a KH1 to4

ynthe iz d containing a1nin acid 196 to 566 (N BI R ference

qu nc
h

RD-

e1i ng the D ur KH d main . The preci e po ition of truncation
en ba ed n pr

1

u

tudi

; the

-te1m inu p

ition ba ed on the ucce ful

hao et

al. (2 010) IMP! cry tal c n tiuct, and the N- t rminal baed on KHl to KH4 functional studi es
iel n t al. 2002) . Phy i 1 gi al rele anc of a KHl to KH4 cry tal stru cture tems from a
recent wo rk re ealing the pre nee of a

RD-BP plice vari ant in human brea t cancer cell that

nly co ntain th K.Hl t KH4 regio n and i nece ary [! r clon genic grow th (Fakhraldeen et al.
2015 ). hi kingly thi di covery ugge ts that
tudie

a antibodi e

RD-BP i und errepre ented in m any cancer

pecific to the N-tetminal deleti on will not detect this CRD-BP i oform .

Al o the K.Hl to KH4 vmiant retain fu ll RNA binding capacity as well as the full -length 's
propen ity for protein-protein dimerization (Niel en et al. 2004 ).
onventional native preparation of the fu ll-length CRD-BP and K.Hl to4 truncation
yielded low qu antiti es of solubl e protein, a complication previou ly reported by Du et al. (2008)
Git and Stand art (2002) and Prokipcak et al. (1993) as well as hi gh levels of RNA contamination
(fi gure 4.12B). presum abl y from endogenously bo und E.co li RNA. Furtherm ore, our lab' s
existing denaturing protein purification protocol also proved limited to refolding concentrations
of less than 0.7 mg/m l as well as issues concentrating to high level . Eventually a denaturing, oncolumn refolding method was developed that allowed for both hi gh concentration of

RD -BP

with no RNA contamination (4.1.9) and thi s method was used to begin cry tallization tri al in
collab rati n with the Dr. Strynadka X-ray crystallography laboratory at UBC . Though "'e era!
attempts produ ced onl y fa l e positi ve (mo tly in pho phatc buf~ r) , a single hit with cond ition ·
0.2M lithium sulphate, O.lM Tris pH 8.5, 40% v/v P

400 produced aU -flu orescent cry tal

133
but wa not f high en ugh qu ality£ r further tudy (figure 4.1 3D). Optimizati n experiment
calculated a t mp ratur

f aggregati n £ r

RD-BP KH 1to4 of near 4 1 ° in mo t buffer

te t d (fi gur 4 .14) whi ch matche perfectly with the melting temperature a determined by
th rmal hift

p riment (4.1 ° ) c nfirming the high temperatur

D

en itivity of this protein

which rna e plain th diffi ultie encount r d in th e e cry talliza tion attempts. In addition, the
KH1 -KH2 didomain ha been tu died u ing oluti n NMR , and wa determined to po e
inherent fl e ibility not pr

nt in th KH -KH-4 di dom ain, poss ibly impacting the ability to

fonn a table cry tal nucleati n ite (Du et al. 2008) . Whil th cry tal hit wa ultimately
un uitable for X-ray diffraction data coll ection, thermal aggregation experiments found that pH
6, high alt and ME buffer (4 .2.4) were all tabilizing fa ctor and are hi ghly relevant fo r future
CRD-BP cry tal pur uits .
5.11

S ummary & concludin g remarks

CRD-BP and the orthologou VICKZ protein famil y member , including human IMP- I ,
have been implicated in several fom1 of cancer, as well as playing criti cal ro le in the early
stages of embryogenesis (Elcheva et al. 2008 ; Fakhraldeen et al. 20 15; Joannidi s eta/.
2001 ;Ni elsen et a!. 1999) . CRD-BP achieves its effects ( confening enhanced transcript stability,
translational regulation, and mRNA localization) through binding nlRNA in the cytoplasm and
nucleus (Doyle eta/ . 2000) . Despite demonstrated importance in gene regulation and expan i e
international research efforts, there remains no mod el that fully characteriz s VICKZ protein . In
this work the molecular intera ction between CRD-BP and target mRNA was tud ied, focusing
'
on KRAS - a proto-oncogene that when mutated (G 12V) or overe pres ed lead. to
mi sregulation of cell growth throu gh induction of the downstream effector pathway : mTOR,
PI3K, MA PK, and

FR (Radzioch cl a/. 2010; Birkeland ct a /. 201 2; Va ltm1a ct a!. _0 13;

134

a aki t a!. 2011 ). It ha b n d mon trat d by M ngroo
KRA

regulat

pr

i n in c lorectal cancer cell lin

t a!. (2011) that

RD-BP positively

. Knockd wn with

RD -BP- iRNA

di play d a 5 % reducti n in KRA in W -480 and indu ction of RD-BP with a ma1nmalian
e pr

ect r n arl tripl ed KRA level in LIM2405 cell . Uncoupling

with KRA

ignalling u ing inhibit r m l cu] e m ay therefore pr ve to be a valuable strategy in

RD -BP expr sion

battling KRA -dri en ca ncer .
Mol cul ar d tail p rta ining t the
bindin g
a!. 2011 ).

ith th c din g reg i n and

RD-BP-KRA mRNA interaction how direct

' TR, a dete nnin ed th ro ugh

V-cr

slinking (Mongroo et

o further pecifi city of a binding equ ence or indication of actual binding affinity is

known which prompted the KRA mapping ex periments for CRD -BP binding (section 2.3.2).
KRA mRNA wa fo und to bind CRD-BP in both th e codin g reg ion and 3'UTR usin g M A,
confun1ing previous fmdings of M ongroo eta!. (20 11 ). Ind eed multiple binding ite were
di scovered within both regions with at lea t two present in the coding region (1 -185 and 38861 0) and two present in the 3 'UTR (568-793 and 97 1-11 55). The hi ghest affinity RNA fragment
"KRAS -A" covered nu cleotide pos itions 1-185 in the coding region and the di sociation con tant
calculated to be 177.9 ± 18.4 nM. Further EMSA ana lysis of KRAS-A revealed a 57 nucleotide
minimum CRD-BP binding sequence (129- 185) KRAS-A34(b), as well as fragments posses ing
virtually no affinity for CRD-BP, confinning orne specificity of binding. The predicted n?fold
secondary structure of the KRAS-A34(b) ub-rcgion, covering nucleotides 129-185, a well as
the raw prnnary RNA sequ ence, will add to a growing li brary of earchable RN
sequences for determination of a consensus bi nding motif for

tructure and

RD -BP. Furthem1ore, th

size of the KRA binding region allow for fu ture fluore c n e p larization ba ed hi ghthroughput creens for potential mall molecule inhibitors.

mall

135
RD-BP lik many RBP utiliz

multiple RNA-attachment ite to achieve high affinity,

high p cificity interacti n . F ur K -h m logy (KH) domain con titute the primary RNA
binding m dule, with tw

-terminal RRM d main al o pre ent; the precise purpose of the e

tw domain r main unclear. pecific KH d main requirem ent for binding appear unique to
each mRN

tran cript, implying that ari u

hart binding element on the RNA act in

c mbination with th pr tein . The high affinity KRA -A sub-region (nts 1-185) wa examined in
conjunction with KH mutant vari ant

f

RD -BP to detennine which if any domain wa

pr d minantly re p n ibl e for binding. B th

M

and i1nmune-precipitation coupled with RT-

qP R analy i in H La cell revealed a imilar patten1 of KRA a ociation, trongly supporting
the notion that KRAS require all four KH domains to tabili ze the interaction, with a omewhat
mitigated role for KH 1 domain . Thi re ult optimistically su ggests that dru gs found capable of
blocking any of the fo ur KH domains wi ll strongly inhibit the CRD-BP-KRAS mRNA
interaction, in cells and in vi tro, making KRAS a lucrative opp01iunity among CRD-BP target .
Traditional drug discovery often proceed with random te ting of generated sm all
molecule libraries, searching for a desired effect among thousand or millions of candidate
(reviewed in Hugh es eta!. 20 10) . The use of specific anti-sense oli gonucleotide inhibitors that
bind mRNA and reduce translation, or interaction with nascent proteins is relatively new, and
requires specific knowledge of the target mRNA but can indirectly inhibit a protein target
(Henning and B est, 2002). To date there i no published work on inhibitors designed to abrogate
the interaction between CRD-BP and KRAS mRNA, in vitro or in vivo . Thi the is pre ent , two
DNA-based anti-sen e o ligonucleotide inhibitors (section 2.3 .3, M6 and M7) again t th
RD-BP-KRAS mRNA interaction that were developed with th knowledge gained from th
KRAS minimum binding equence (section 2 .. 1), the mo t ef[i ' ti

w1th an I ~o of 11 . nM .

136
Th m

t pr babl e m chani m ofth ir inhibition is complementary Wat on- rick base

pairing, ma king th KRA RN
inhibiting int racti n.
with th anti-

n

binding

RD -BP and th reby

rl ap f th n1inimum binding equ nee di c v red in ection 2.3.1

lig nucle tid inhibit r target regi ns doubly impli ate the KRA -A34(b)

r g1 n (nt : 129- 1 5) a imp rtant b r
DN -ba

qu en e/ tru cture from

RD -BP interaction with KRA RNA. There earch into

d li g nucl e tid inhibit r pre ented here ha laid the groundwork for future cell-

ba ed a ay u in g M6 and
therap eutic trat gi

M7 to target KRA , and may join the other hopeful anti-KRA

with fur1h er d

el pment.

dditi onall y,

M 7 and M 2 have found fu11her

u e a a po iti e control in flu ore cence polarization based creenin g of proto-drug for antiRD-BP activity in our laboratory.
econdary tructure analy is of the

RD -BP protein revealed a hi ghly stru ctured protein

with primarily a-helical (69%) compo ition, in-line with other KH domain-containing protein
(Chao eta!. 2010; Valverde et al. 2007; Chmi el eta!. 2006). Further an alysis of the CRD-BP KH
mutant library conclud ed that introduction of the GXXG mutations used to study KH domain
requirements in RNA binding did not compromise the global structure of the protein.
Remarkably little variation was apparent between the CRD -BP KH mutant proteins CD pectra
and provides compe llin g evidence that the GXXG loop region between th e fir t and econd ahelix do es not contribute any structural integrity to the protein. FUiiher CD analysis hawed
quantifiable changes in secondary structure upon binding RNA implying that CRD-BP may hav
multiple forms, one conformation stabilized by RNA binding. T he effect wa general! nonp ecific with all RNA abl e to induce the hift, but the degree of change did app ar to cotTelat
with RNA target affinity suggesting ome specificity (4.2.2). Intere tin gly with the rem val of
the RRM dom ains, the pecifi city wa mitigated betw en KRA - l ~a nd KR

-A3 4 RNA

137

targ t (high and 1 w affinity re p ectively) ugge ting ome role in pecificity for the RRM
d main . upp rting thi n ti n i the flu re cence p lariza tion binding as ay completed,
c mparing th full -length
compari

n re

RD-BP t a g nerat d KHlto4 truncati n variant (4.2.4) . Kd

al a triking 3-fl ld increa e in affinity with the removal of the RRM domains.

no functi n ha

et be n attributed to the RRM domain in VICKZ proteins, confirmation of

thi ef~ ct with th r RN

target may a cribe a novel negative binding regul ator fun ction to the

RRM dom ain .
lnterd main g
predicting

metry w ithin th e ''KH I to4" binding modul e is key to und erstandin g and

RD-BP mRN

target , a

ugge ted by Chao et al. (2 01 0) and evid enced by the

appearan ce of bipartite equ ence requirement fo r ~-actin "zipcode" RNA. Currentl y there i no
cry tal tructure of full -length CRD -BP or any of the VICKZ protei ns, with onl y a tructure of
the KH3-KH4 did om ain availabl e to scru tinize. However, as KRAS requires all four KH
domain , a more complete tru cture is necessary to understand how binding occurs. Despite a
thorough attempt to obtain a diffraction pattern and crystal structure of CRD -BP, no cry tal
suitabl e for structure dete1mination were obtained. A single large pro tein crystal posse sing
twinned character did grow in 0.2M lithium sulphate, O.l M Tris-C l pH 8.5, 40% v/v PEG400
buffer, however it was deemed no t sui ta ble for X-ray diffracti on. Fonnation of a crystal doe
however suggest that the specific KH 1to4 construct described in this thesi can potentially
crystallize and future efforts in compl ex with an RNA target may be more successful.

xi

References
tla , R., B har, ., Hiott, . inzburg, I. (2004) . The in ulin-lik growth fa ctor mRNA
binding-prot in IMP - 1 and the Ra -r gulat ry pr tein 3BP a ociate with tau 1nRNA and HuD
pr t in in difD r ntiated P 1 neuronal c 11 . J eurochem. 89 : 613 -626.
pp 1
. M ., B ijnen I. H. and chell en , J. H. (2005) .
inhibit r : are 1 w . neola i t. 10: 565- 57 .
Bain , ., Xu, D ., D r
Futur U d.

velopm ent of fmne yl tran D rase

. (20 11 ). Inhibition f Ra for cancer treatment: the earch continue .
7- 1 0 .

Ba111 M . an Ren burg, . Li W ., Mehmo d, K ., Mackedenski, ., han, ., King, T .,
Mill r, ., ee, H . (20 14 ). M lecul ar in ight into the codin g region determinant-binding
prot in-RN interacti n thr ugh ite-directed mutagene i in the heterogen ou nuclear
ribonu cl pr t in-K-hom logy domains. J Bioi. h m. 290: 625-639.
Bell J. Wachter, K. Mi.ihleck, B., Paza iti , N ., Ka hn, M ., Lederer, M ., Huttelmai er, . (201 3) .
In ulin-like gr wth fac tor 2 mRNA- binding protein (IGF2BPs): post-transc1i ptional dtivers of
cancer progre i n?. Cell. Mol. L{f( Sci. 70: 2657-267 5.
Bernstein, P ., H nick, D . Prokipcak, R., Ro s, J. (1 992) . ontrol of c-m yc mRNA half-life in
vitro by a protein capable of binding to a coding region stability determinant. Genes D ev. 6(4):
642-654 .
Berx, G ., Cleton-Jan en, A . M . Noll et, F. , de Leeuw, W . J. , van d Vij ver, M ., Com eli e, C.
and van Ro y, F. ( 1995). E-cadherin i a tumour/invasion suppressor gene mutated in hum an
lobular breast cancers. EMB O J. 14 : 6 107-6 11 5.
Birchmeier, W ., Behrens, J . (1994). Cadherin expression in carcinoma : role in the fo rm ation of
cell junctions and the prevention of in vasivene . Biochim. Biophy . A eta . 1198 : 11 -26 .
Birkeland, E ., W ik, E ., Mjos, S., Hoivik, E ., Trovik, J., We111er, H,. Ku onmano, K ., Peter on,
K ., Raeder M ., Holst, F., Oyan, A ., Halland, K ., Akslen, L. , Simon, R., Krakstad, ., alve en,
H . (20 12). KRAS gene amplification and overexpression but not mutation associates with
aggressive and n1etastatic endometrial cancer. Br. J. Cancer. 107: 1997-2004 .
Bos, J. L. Fearon, E. R ., Han1ilton, S. R., Verlaan-de Vries, M ., an B om, J. H., an der Eb,
J., & Vogelstein, B . (1987) . Preva lence of ra gene mutations in huma n colore tal cancer .
Na ture. 327(61 20) : 293 -297.
Boyerinas, B ., Park, S., homron, N ., Hed gaard, M ., Vinth r, J., nder on, J. , Feig, ., u, J .,
Burge, ., Peter, M . (2008) . ldentification f L t-7-r gul at don ofeta1 gene . an ccr Re.. 68( ):
2587-25 91 .

xii

B ylan, K . Mi h
i, M . Marque
. M rin X., hia, W ., Hay T. (2008) . Motility
r n id ntifie
r
phila I F- II mRN -binding pr t in-zipc d -binding pr tein acting in
and ynaptogen i . PLoS
n t . 4(2) : 36.
Bre er,
625.

. (200

.M

eng r

Bulh ller, B ., Hir t J. (200 ).
25 : 9-540 .

d cay during ageing and d v 1 pm ent. A

i hr

tn

R . 1: 607-

ale-circul ar and linear dich.roi m online. Bioinfonnatic .

Bunn r ., L b L. (1 9 ). Mutation in the KRA 2 neogene during progre 1ve ta ges of
human c 1 n carcin ma. Pr c. at/. A ad. S i . 86 (7) : 2403- 7.
han, I. , hahra an . F dor a,
hin J. (200 ). The bZIP target overlappin g D A
ub ite within a half- it , re ulting in increa d bind ing affiniti e . Biochemi fly . 46 : 1663- 167 1.
handra , ., recco, H . . Pi upati, V. , Perera, D ., as id y, L., koulidi , F., et al. (201 2). The
GDI-lik olubilizing fac t r PDEdelta u tain the patial organi -za tion and ignalling ofRas
famil y protein . at. Cell Bioi. 14 : 14 - 158.
Chao, J., Pat ko ky, Y. , Pate l, V . (20 10) . ZBP 1 recogni tion f ~ -acti n zipcode induces RNA
looping. Gene D e1•. 24 : 148- 158 .
Chen C . Y ., Del G atto-Konczak, F . Wu, Z., & Karin, M . (1998) . tabilization of interleukin-2
mRNA by the c-Jun N H2-terminal kina e pathway. Science. 280 : 1945- 1949 .
Chmiel, N ., Rio, D ., Doudna, J. (2006) . Di tinct contributions of KH domains to ub trate
binding affini ty of drosophila P-element om atic inhibitor pro tein. RNA . 12 : 283-29 1.
Cogoi S., Zorzet, S., Rapozzi, V ., Geci, I. , Peder en, E., Xodo, L. (2 01 3 ). MAZ-binding G4decoy with locked nucleic acid and twisted intercalating nucleic acid modifica tion suppre e
KRAS in pancreatic cancer cell and delays tumor growth in mice. Nucleic Acids Res. 41 (7) :
4049-4064.
Collins, M ., M agliano, M (20 14 ). Kras as a key oncogene and therapeutic targ t in pancreatic
cancer. F ront. Physiol. 4: 407 .
Condeelis, J. , Singer, R. H ., & Segall, J. E . (2005) . The great e cape: when cancer cell s hijack
the genes for chemotaxis and motility. Annu. Rev. Cell Dev. Bioi. 21 : 695 -7 18.
Costa, S. (20 13 ). D evelopment of a Novel Fusion ystem D r Recombinant Protein Produ ction
and Purification in Escherichia coli. Ph.D . Thesi , University of Minho, Braga, Portu ga l.
ouli , ., Lee, ., Nardone, V ., Prokipcak, R . (2000) . Inhibition of -m yc pres ion in ce ll
by targeting an RNA-protein int raction u ing antisense oligonu 1 otid - . Mo l. Phanna I.
57(3 ): 485-494.

xiii

ratg

pi gelman Y . . (20 12) . Inhibiti n f coding region determinant binding
m elan rna c 11 t h motherapeutic ag nt . Pigm nt ell Melanoma Re .

11

n J . iel n, . vruch, J. (20 13) . mT R complex 2 pho phorylate IMPl
prom t I 2 pr ducti nand the proliferati n of mou e embryonic
v. 2 7: 3 0 1-3 12.

Da i , M . ., and amu 1 , Y . (20 10).
m Ian ma. n o n . 29 : 5545- 5555 .

naly i of the genome to personalize therapy for

D an J. ., ully, ., lark, . R ., & aklatvala, J . (2004) . The inv lvem ent of U-rich
el m nt-binding pr tein in p mit gen-activa ted protein kinase pathway-m ediated mRNA
tabili ation.
//. i nal. 16( 10) : 111 3- 11 2 1.
D rgham . T. , Du gan, M . ., Ku way, R ., Du , W ., Kamarau kiene, D . S., Vaitkevicius, Y . K .,
.. . & arkar, F . H . (1997) . Pre alence and clinical ignifi cance f combin ed K -ras mutati nand
p53 aberrati n in pancreatic adenocarcinoma. Int. J . Pancreatol. 21( 2): 127- 143 .
D errigo, M ., Ce telli , A ., av ttieri, ., Diliegro, I. (2000) . RNA-protein interactions in th e
control of tability and localization ofmRNA. Int. J Mol. M ed. 5(2): 111 -123 .
De hler, J ., High ett, M ., A bramson, T., Schnapp, B . (1998) . A highl y con erved RNA-binding
protein for cytopl a mic mRNA locali zation in vertebrate . Curr. Bioi. 8 : 489-496 .
Dimitiiadis, E ., Trangas, T ., Milatos, ., Fouka , P . G., Gioulbasanis, I. , Courtis, N ., ... &
Ioannidi , P . (2007). Ex pre ion of oncofetal RNA-binding protein CRD-BP/IMP 1 predict
clinical outcome in colon cancer. J. Int. Cancer. 121 (3) : 486-494.
Doyle, G ., B etz, N., Leeds, P ., Fleisig, A ., Prokipcak, R ., Ross, I . (1998). The c-m yc coding
region determinant-binding protein: a member of a fami ly of KH domain RNA-binding protein .
Nucleic Acids R es. 26: 5036-5044.
Doyle, G ., Bourdeau-Heller, J ., Coulthard, S., Meisner, L. , Ro , I . (2000) . Amplification in
human breast cancer of a gene encoding a c-myc mRNA binding protein. Can cer Re . 60 : 27562759.
Du , z., Fenn, S. , Tijhen, R ., James, T. (2008) . tru cture of a construct of a human poly( C)binding protein containing the first and econd KH domain reveal insight into it regul atory
mechanism s. J. Bioi. Chern. 283 : 28757 -28766.
Elcheva, I. , Tara pore, R ., Bhatia, N ., piegelma n, V . (2008) . 0 ere pres ion of mRN A-binding
protein CRD-BP in malignant melanoma . On cogene. 27 : 5069-50 74.
Icheva , I. , oswarmi, S., Noubi ssi, F., piegclman, V. (2009) . RD -BP prot ct the coding
reg ion f ~Tr p 1 mRN A from miR- 18 -mediated degradation . Mol. ell. 35(2): 240-246.

xiv

Faklnald n, . lark R ., R pra, ., hin, ., Huang, W.
J. Wi in ki, K ., Kim, T .,
pi .g lman, V . le and r, . (20 15). Two i form f the RNA binding protein, oding
R g10n D t 1minant-binding Pr tein ( RD -BP/I 2BP1), ar expr ed in brea t pithelium
and upp 11 clon g nic gr wth fbrea t tum r c 11 . J Bioi. hem . 290(21) : 13386- 13400.
Farina, K . Hutt lmai r . Mu unuru, K ., Dam 11 R ., inger, R. (2003) . Two ZBP 1 KH
d main facilitate ~ - a c tin mRN localizati n granule formati n and cytoskel etal attachment. J.
II Bi I. 160(1 ) : 77- 7.
F nn, . Du Z. L
J., Tj hen R ., troud , M., Jame , T. (200 7). ry tal tructure of the third
KH domain f human p ly( )-binding protein-2 in compl x with a -rich strand of human
telomeric D
at 1.
re luti n. u lei A cids Res . 35(8 ): 2651 -2 660.
F rrara , . (20 10) . Path ay m di ating V
r wth Fa tor R v . 21: 2 1- 26.

F-indep endent tumor angio gene i . y tokin e

Filipowicz W ., Bhatta h aryya, ., onenberg, N. (200 8). M echani sm s of po t-tran criptional
regulation by mi croRN : ar e the an wer in ight?. Na t. R ev. Genet. 9(2): 102- 14.
Gasteig r E. Hoogland, . Gattiker, A ., Duvaud, S. Wilkin , M ., ppel, R ., Bairoch, A. ;
Protein Identification and A naly sis Too l on the ExPASy Server; (In) John M . Walker (ed): The
Proteomics Protocol H andbook, Humana Pre (2005 ). pp. 57 1-607.
Georges R .N., Mukhop adhyay, T. , Zhan g, Y. , Yen, N. , & Roth, J. A . (1993). Prevention of
m1hotopic human lung cancer growth by intratracheal instillation of a retroviral antisense K-ras
construct. Cancer Res. 53 (8) : 1743 -1746 .
Gherzi, R ., Lee, K .Y ., Briata, P ., W egmull er, D ., Moroni, C., K arin, M ., and Chen, C. Y . (2004) .
A KH domain RNA binding protein, KSRP , promotes ARE-directed mRNA turnover by
recruiting the degradation m achinery. Mol. Cell. 14 : 57 1- 583 .
Git, A ., Standart, N . (201 2) T he KH domains ofXenopus Vg lRBP m edi ate RNA bindi ng and
self-association. RNA . 8: 13 19- 1333 .
Goolsby, K ., Shapiro, D . (2003) . RNA i-mediated depletion of the 15 KH dom ain protein, vigilin,
induces death of di viding and non-dividing hum an cells but does not initially inhibit protein
synthesis. N ucleic A cid Res. 31(1 9) : 5644-5653 .
Goswanni, S., T arapore, R., TeS laa, J., Grinblat, Y., etalUii, V., piegelman, S. (20 l 0).
MicroRN A-340-m ediated degradation of m icrophthalmi a-a sociated tran cription fa tor mRN
i inhibited by the coding region detenninantebinding protein. J. Bioi. Ch em . 285(27): .... 05 3220540.

XV

ray, D . ircular dichr i m f pr tein-nucl ic acid interaction . In omprehen ive chiroptical
py V lume 2: Applicati n in ter ch mical analy is of ynthetic compound ,
p tr
natural pr du ct and bi molecul B ro a ., P lavarapu P., Nakani hi K ., Woody, R ., d.;
Wiley:
w Y rk, 20 12; 1. 2· p 615-6 3.
ri hin . (200 1). KH domain: n m tif, two fold . uc/eic Acids Res. 29 : 638- 643 .
u W . Katz, Z. Wu, Bin., Park, H ., Li D., Lin, tanley., Well , A ., inger, R. (2011) .
R gulati n of local pr i n of ell adhe i n and motility-related mRNA in breast cancer
cell by IMP1 /ZBP1. J.
ll S i. 125 : 81-91.
Hafn r M. Landthaler, M . Burg r, L. , Kh r hid, M ., Hau er, J., Berrunger P., RothbaUer, A.,
cano, M ., Jungkamp , ., Mun chauer, M ., !rich, ., Wardle ., Dewell, ., Zavolan, M .,
Tu chl, T . (20 10) . Tran ciptome-wide identification of RN -b inding protein and microRNA
target ite byP R- IP. ell. 141 : 129-141.
Hall M ., Huang ., and Black, D . (2004). Differentiation-induced colocalization of the KH-type
plicing regulatory pr t in with polypyrimidine tract binding protein and the c- rc pre-mRNA.
Mol. Bioi. ell. 15 : 774-786.
Hmnilton K . oubi i, F., Katti P., Hahn, ., Davey, S., Lund mith, ., Klein- zanto, A.,
Rhim, A., piegelman, V ., Rustgi A . (2013). IMPl promotes tumor growth, dissemination and a
tumor-initiating cell phenotype in colorectal cancer cells xenograft . Carcinogene is. 34(11 ):
2647-2654 .
Hananan, D ., Weinberg, R. (2011). Hallmarks of cancer: the next generation. ell. 144(5 ): 646674.
Hansen, T. , Hammer, N ., Nielsen , J. , Madsen, M ., Dalbaeck, C. , Wewer, U., Christiansen, J. ,
Nielsen, F. (2004). Dwarfism and impaired gut development in insulin-like growth factor II
mRNA-binding protein !-deficient mice. Mol Cell Biol 24( 10) : 4448-4464
Havin, L., Git, A., Elisha, Z., Obennan, F., Yaniv, K. , Schwartz, S., Standart, N ., Yi eraeli , J.
(1998). RNA-binding protein conserved in both microtubule- and microfilament-ba ed RNA
localization. Genes D ev. 12 : 1593-1598 .
Haynes, S. (1999). RNA-Protein interactions. Totowa, New Jer ey, Humana Pre
Hennessy, B . T ., Smith, D . L. , Rmn, P. T., Lu , Y ., & Mill , G . B. (2005) . Expl iting the
PBK/AKT pathway for cancer dru g discovery. Nat. R ev. Drug Discover;. 4(1 2): 9 8-l 004.
Henning SW, Be te G (2002) . Loss-of-function trategie in dru g target va lidati on. Cun. Drug
Discov. May: 17-2 1.
Hollingworth, D ., andel, A., Nicastro, G., Martin, ., Briata, P., Gherzi, R., Ramo , . (~01 ~) .
KH domains with impaired nucleic acid binding a a tool for fun cti ona l anal si ~ . uclcic A cicls
R e . 40(14) : 6873 -6886 .

xvi

Hutt lmaier ., Zenklu en D . L d rer M ., Diet nberg, J. Lorenz, M ., M ng, X ., Ba ell, G.,
nd li , 1. , ing r, R. (2005) . patial re gulati n f ~-actin translation by re-d pend nt
ph phorylati n ofZBPl. atur. 438 : 512-515.
I annidi P . Tran g a T. Dimitriadi
ami taki M . Kriazoglou , I. , Tsiapa1i , ., Kittas, .,
gnanti
iel en F. iel en, J., hri tian n, J. , Pandi N. (2001). -myc and I F-II
mRN -binding pr tein ( RD-BP/IMP- 1) in b nign and malignant me enchymal tumor . Int . J
an r. 94 : 4 0-4 4.
I annidi P ., Mahaira, ., Papadop ul u, . Teixeira, M . R ., Heim , ., And er en, J. A., ... &
Tranga T. (200 ). q24
py number gain and expr ion of the c-myc mRNA tabilizing
pr tein RD-BP in primary br a t carcinoma . Int. J ancer. 104(1) : 54-59.
Ioannidi , P . Mahaira, L., Perez, ., ritzapi , ., otiropoulou, P ., Kavalaki s G ., Ant akli , A.,
Ba vani t, ., Papamichail , M . (2005). RD -BP/IMPl expre ion characterizes cord blood
D34+ tem cell and effe t c-myc and I F-II expre sion in M F-7 cancer cells. J Bioi. Chem .
280 (20): 200 6-20093 .
Jacob E., Puc han ki, L., Zeruya, E., Baran N., Manor, H . (2004) . The human protein tran lin
pecifically binds ingle-stranded micro atellite repeats, d(GT)n, and G-strand telomeric r peat ,
d(TTA GG)n: a tud y of the binding parameters. J. Mol. Bioi. 344 : 939-950.
Jancfk, S., Drabek, J., Rad zioch, D . and Haj du ch, M . (20 10). Clinical Relevance ofKRAS in
Human Cancers. J Biomed Biotech no!. vol. 20 10, Article ID 150960, 13 page , 2010.
doi : 10.1155/20 10/ 150960
Johnson, S., Gros hans, H ., Shingara, J., Byrom, M. , Jarvis, R ., Cheng, A ., Labourier, E.,
Reinert, K ., Brown, D ., Slack, F. (2005) . RAS is regulated by the let-7 microRNA fami ly. Cell.
120 : 635-647 .
Jones, A ., Schedl T. ( 1995). Mutations in gld-1, a female genn cell-specific tumor suppre or
gene in Ca enorhabditis elegans , affect a conserved domain al o found in rc-as ociated protein
Sam68 . Genes D ev. 9: 1491-1504.
Jonson, L. , Vikesaa, J., Krogh, A., Nielsen, L., Hansen, T ., Borup, R ., Johnsen, A ., Christiansen,
J. Nielsen, F. (2007). Molecular compo ition of IMP 1 ribonucleoprotein granules. Mol. Cell.
Proteomics. 6: 798-811 .
Kaluzhny, D ., Laufman, 0 ., Timofeev, E., Bori ova, 0 ., Manor, H ., hchyolkina, . (2005).
Conformational changes induced in the human protein tran lin and in the single- 'tranded
oligod oxynucle tid es d( T)(l2) and d(TT G G)(5) upon binding of these
oligodeoxynucleotides by translin. J Biomo/. Struct. D_vn. 23 : 257-265.
Kato, T. , Hayama, S., Yamabuki , T., Is hikawa, N., Miyam to, M ., Ito, T., T uchi a, ., Kondo,
S., Nakamura, Y. , Dai go, Y . (2007). Increa de pre ion fin ulin-like grovvth fa tor- II

xvii

m
nger RN -binding pr tein 1 i a ciated w ith tum r pr gre
ancer. lin . an r R . 13(2): 4 4-442.

n in patient with lung

J. . (2007) . RN regu l n : c rdination f p t-tran cripti nal v nt . Na ture R vi w
17 tic. ' 8(7) : 5
-5 4 .
K ith, ., J hn n, R . ., & im
. (20 I 1). HI F l a and HIF2a: iblin g rivalry in hypo ic
tum ur gr w th and pr gr
at. R e . ancer. 12( 1): 9-22.
Ke n

Kim M ., lack, . (20 14 Mi cr R
n ol. 7: 4.

-m di ated regul ati n of KRA

King, D., Baine , M ., Th m en, ., Lee,
mall m le ul antibioti Ii r the ability t
£ . 9( ): 15

in cancer. J J!emato/.

e ing pecific ligo nucleo tid e and
RD -BP- D44 RN A interacti n . PLoS

Kj r m J., Ikd ah1 , T ., uren, T. , k lund , .,
rbye, H ., Hamfj ord J. , Pfeiffer, P ., G limeliu ,
I ang, H ., T eit, K ., Kure, . (20 12) . Let-7 miRN A-bincling ite
B ., K r t n, .
' TR; co l re ta l ca ncer reenin g popul ati on preva lence and
po l 111 rphi 111 in th e KR
influ enc on c linical outc me in patient wi th m ta tatic co l r ct I can cer treated with 5fl ourouracil and o alipl atin +/- cetuxim ab . BMC ancer . 12: 53 4 .
Kobel, M ., Weiden dorfer, D ., Reinke, ., Lederer, M ., chmi tt, W . D ., Zeng, K ., ... &
Hi.ittelmaier S. (2007). Ex pre ion of the RNA- binding pro tein IMP 1 correlate w ith p or
progno i in ovarian carcinoma. neogene. 26(54) 7584-7589.
Laheru, D ., hah , P . Rajeshkumar, N . V. , Mca ll i ter, F., Taylor, ., Gold weig, H ., et al. (2012) .
Integrated preclinical and clin ica l developm ent of S-trans, trans- Fam e ylthio alicylic Acid
(FTS , alira ib ) in p ancreati c cancer. Im ,est. New Drug 30: 239 1- 2399.
Lavind er, J ., Hari, S. , Sullivan, B ., Magliery, T. (2008) . High-throughput th rmal creenin g: a
gen eral, rapid dye-bind ing thennal shift screen for protein engineering. JAm. Ch em. Soc. 131:
3794-37 95 .
Layton, C ., H ellinga, H . (20 1 1). Quantitation of protein-protei n interaction by thetmal tability
shift analysi . Protein Sci. 201 : 1439-1450.
Le, H ang., Sorrell, A ., Sidd le, K . (2012) . Two i f01m s of the m RNA binding protein I FBP2
are generated by alternative tran lational initia tion. PLoS 0 E . 7(3) : 33140.
Lebedeva, S., Jen , M ., Theil, K ., chw anhau ser, B ., elbach, M ., Landthaler, M .,
Raje\v ky,
N . (20 11 ). T ranscriptom -wid e anal ys is of regulatory int rac tion of th RNA- binding prot in
HuR. M ol. cell. 43 (3) : 340-35 2.
Le c1 , P ., Kren , B ., B ylan, J. , Betz, N ., tcer, ., Gruppu o, P., Ro , J. ( 1997). De\ lopmental
in vtt ro. n ·ogene.
regul ation of R -BP, an RN -binding pr tcin tha t tabi lize c-m y mR
14: 1279- 1286.

xviii

.,

h n H ., Darnell R ., Burley . (2000) .
main: implicati n for paraneopla tic di ea e

B ., Pat 1 M . Hu , Y ., harl
., H erri ck, D ., Br w r, . (2004). Targeted knockd wn of
-binding pr t in RD -BP pr m t cell pr liferation ia an in ulin-like growth factor
Il-dep nd nt path ay in human K5 2 cell . J. Bioi. hem. 279(47) : 4 716-48724 .
. ( 1997).
h m . 272(4 ): 27274-272 0.

pecificity and det rminant

f am68 RN

binding.

Lin, F ., hen, Y ., Lin, Y ., T ai, J ., hen, J., Wang H ., Lin, . (2006) . The Rol e of Human
nti g n R , an RN -binding Pr tein, in M diating th
tabilizati n fToii-Like R ceptor 4
mRN Indu ed by nd t in
1 Mechani m Invo lved in Ya cular Inflammation .
Art ri
1 r. Thromb . a c., 26( 12): 2622-2629.
Linker, K . Pautz, ., Fechir, M ., Hubri ch, T., reeve, J ., and Kl einert, H . (2005) . Involvement
of K RP in th p t-tran crjptional regul a- ti n of hum an i
xpres ion -c mplex interpla y of
K RP with TTP and HuR. ucleic Acid. Res. 33 : 4813--4827 .
Loui -Jeune, ., Andrade- a arro, M ., Perez-Iratxeta, . (2011) . Predi ction of protein
econdary tru cture from circular di chroism u ing theo retica ll y deri ved pectra . Pr teins:
Structure, Fu n lion, and Bioi1~(ormatic . 80 (2): 374-38 1.
Ma caux, C ., Iannino, ., Martin, B ., Pae m ans, M. , Berghman , T. , Dusart, M ., ... & culier, J.
P . (2004). The role ofRAS oncogene in urvi val of pati ent w ith lung cancer: a y tem atic
review of th e literature wi th m eta-analy i . Br. J. Cancer. 92( 1): 13 1-139.
Maurer, T ., Garrenton, L. , Oh, A ., Pitt , K ., Ander on, D ., kelton, N. , Fauber, B ., Malek, .,
Stokeoe, D ., Ludl am , M ., Bowman, K ., Gianetti, A ., tarova nik, M ., Mellman, I. , Jack on, P.,
Rudolph, J ., Wang, W ., Fang, G . (20 12). Sma ll-molecule ligand bind to a di tinct pocket in Ra
and inhibit SOS-mediated nucleotide exchange activi ty. Pro c at/. A cad. Sci. USA. 109(14) :
5299-5304.
Miura, Y ., hnami, S., Yoshida, K., Oha hi , M ., Nakano, M ., hnami , ., oki, K . (2005).
Intraperitoneal inj ection of adenov iru expre ing antisen e K-ra RN
uppre e p riton al
dis emination f hamster syngeneic pancreati c cancer wi thout y t mi to i ity. Cancer Lett.
218(1 ): 53-62 .
Moll J. , Acharya, A ., al, J. , Mir, A ., Yin on, . (2002) . Magn ium i required for pecific
DNA binding of th e R B B -Z IP domain. N ucleic Acids Res. 30: 1240- 1246.
Mongroo, P ., Noubi i, F., C uatr ca a , M . (20 11 ) lMP - 1 di pla ys cro, -talk with K-Ras and
modulate c Ion cancer cell urvival throu gh th e novel proap ptotic prot in YFIP2. anccr
R "S. 71 : 2 172-2 182 .

xix

M rin P . park , . K rin k, V . Barker,
r , H ., Vogel tein B., K inzler K . (19 7) .
ti at i n f ~- at nin- cf ignaling in c 1 n an er by mutati n in B-eat nin r P . ien e.
275 : 17 7-17 0.
(1992).
1 m ent in X en pu

g tal me
cyte .

nger RN 1 cali zati n dir ct d by a 340-nt RN
. 255 : 91 - 994.

Mu 11 r-Pilla ch F., Lach r
W allrapp
Micha, ., Zimm rhack1, F . Ham i ter, H.,
arga ., Fri
H ., Bu ehl r, M .,
g r H ., il a, M ., dler, ., r , T. (1997) . 1 ning fa
g n hi ghl y
r pre d in an r c ding Li r a novel KH -d m ain containing pr tein .
17 . 14: 272 -27
N ug rbau er K . (2002).
7 1.

nth importance of being c -tran cripti nal. J

II. Sci. 11 5: 3865 -

i 1 en F ., i 1 n, J. , Kr i ten en, M ., Koch
lu·i ti ansen, J. (2 002). ytopla rnic
-binding pr tein by c n erved KH domains. J ell Sci. 115: 2087traffi king of I
2097 .
iel n, J. , Kri t n n , ., Will emoes, M . N iel en F., hri ti an en, J. (2004). equenti al
dirneti zati on fhuman zipc de- bindi 1g pr tein IMPl on RNA: a cooperative m echani m
providing RNP tability. ucleic Acid Re . 32(14): 4368-4376.
Nie1 en , J ., lu·i ti ansen , J . Lykke-Andersen, J ., John n, A., W ewer, U. N ie! en, F . (1999) . A
tran 1ation in late
famil y of in ulin-like grow th factor II mRN -binding proteins repre
development. M ol. Cell. B ioi. 19(2): 1262- 1270.
iel en, J ., Adolph, S., R ajpert de M eyt , E., Lykke-Ander en , J ., Koch, ., hri tian en, J .,
Niel en , F. (200 3 ). N uclear tran it of human zipcode-binding p rotein IMP 1. Biochem J 376:
383-391 .
Noubissi, F ., Elcheva, I. , Bhati a, N., hakoori, A. Ougolkov, A., Liu, J., Minamo to, T., Ro , J.,
Fuchs, ., Spiegelma n, V . (2006). RD -BP medi ate tabiliza ti on of fJ-Tr P 1 and c-my mRN
in respo nse to B-catenin sign alling. Nature. 44 : 898-90 1.
N oubissi, F ., Goswami, ., Sanek, N., K awakam i, K ., Minam oto, T., Mo r, A ., rinblat, Y .,
Spiegelman, S. (2009). Wnt igna1ing timu1 ate tran cripti ona1 outcom of the hedgehog
pathw ay by stabilizing GLil m RNA. ancer Re.. 69(22): 572- 57
leynikov, Y ., Singer, H . (2003). Rea l-ti m vi ua li za ti n f ZBP I a ociation vvith P-actin
mRNA during tran cripti nand loca lization. Curr Bioi. 13 : 199-207 .
trem, J ., Pet r , U., o, M ., Wells, J., hokat, K . (20 1 ). K-Ra ( d2 ) inhibitor
allosteri call y co ntro l T P affinity and effec tor interaction . a lure. 503( 74 77): 54 -55 1.

XX

Pan, ., Hi.itt ]mai er, ., ing r, R .
mRN during tran ri ti n. Mol.

u, W. (2007) . ZBP2 facilitate binding f ZBP I t P-actin
/l. Bioi. 27(2 ): 340- 51 .

ith p ly( )-binding pr tein (P ABP) and the
. Bag J. (200 ). IMP 1 int ract
aut r gu lat ry tran lati na1 c ntr 1 el m nt fp BP-mRNA thr ugh th KI III-IV d main.
FEB J. 273 : 5 7 -5 90.
Pat 1, ., Mitra , . Hani , R ., Bu baum,
i nn t, T., Bren witz M . irvin, M . L vy, M. ,
lm , ., inger, R ., ha , J. (20 12). patia1 arrangement fan RNA zi1 c d id ntifie mRNA
und r p t-tran ripti na1 c ntr I.
D v. 26: 43-53 .
Patgiri . Yada , K ., r ra, P . arint ra ti n. at. hem.. Bioi. 7 : 5

. (2011) . An rth

- teric inhibit r fth Ra- o

Pazie ka, ., Wyrwi z, ., Bujnicki, J.,
m ztyk, K.,
tr w ki, J. (2004). o perative
binding of th hnRNP K three KH d main to mRN target . FEES Lett . 577 : 134-140.
Poinclou , R .,
llin, ., Lizarraga, F. R mao, M ., Debray, M . Piel, M ., & havrier, P . (20 11 ).
ontractility f th c 11 rear dti ve inva i n of brea t tumor cell in 3D Matri gel. Proc. at/.
Acad. S i. U.S.A. 108(5) : 1943 - 194 .
Prokip ak, R ., H nick , D. , Ro , J. ( 1994 ). Purification and properti es of a protein that bind to
the C-terrninal coding region of human c-myc mRNA. J Bioi. Chem . 269(12) : 9261-9269 .
Py1ay va-Gupta Y , Grabocka E, Bar- agi D (2011). RA oncogene : weaving a tum ti genic
web. Nat R ev Cancer 11(11 ): 761 - 774 .

,.
Rackham, 0 ., Brown, C . (2004) . Visualizaton of RNA-protein interactions in li ving cell : FMRP
and IMP1 interact on mRNA . EMBO J. 23 : 3346-3355 .
Raineri I. , Wegmu ell er, D ., Gros , B ., Certa, U., & Moroni , . (2004). Rol of UFl i oform ,
HuR and BRF1 in ARE-dependent mRNA turnover tudi ed by RN interference. ucleic acids
re . 32(4): 1279-1 288.
Rodriguez-Viciana, P ., Watne, P . H ., Vanhaesebroeck, B ., Waterfield , M . D ., & Downward, J.
(1996). Activation of pho sphoinositide 3-kinase by interaction with Ra and by point mutation.
EMB J. 15(1 0) : 2442-2451.
Ro s, A ., leynikov, Y., Ki Iau ki . ., Ta neja, K.. inge r, R . ( 1997). haract rization of a 0actin mRNA zipcode-binding protein. Mol. ( II. Bioi. 17(4) 2 15 -2 165.
Ros , J., Lemm, I. , & Berberet, B. (200 1). Ov re pre sion of an mRN -binding prot in in
human colorectal cancer. On cogene. 20( 45) : 6544.

xxi

a aki H ., Hik aka Y., Kawano . M riyama ., Yano, M . Fuji, Y. (2011). valuation of
Kra g n mutati nand copy numb r gain in non- mall cell lung cancer. J Thorac. Oneal. 6(1 ):
15-20.
haw . Kam n, R . ( J 986) .
F mRN m diate
lecti

con er d
equ nee from the 3' untran Jated reg ion f
mRN degradati n. ell. 46(5) : 659-667 .

M-

haw . T ., Win l w, M . M ., Magendantz, M . uyang, ., Dowdle J. ubramanian A ., .. . &
1a k , T. (20 11 ). 1 cti e killing of K-ra mutant cancer cell by mall mol cule inducers of
xi dati e tre . Pr . at/. A cad. i. U.S.A. 108(21 ): 8773- 778.
hima F. Yo hikawa Y ., Y M . raki, M ., Mat umoto , ., Liao , J. , Hu, L. , ugim to, T. , Ijiri,
Y . Tak da
i hi yama, Y ., at , ., Murako , ., Tamura, A .,
oda , T. , T uda, K.,
Miyakawa, T. , Fukuni hi H ., himada, J., Kuma aka, T., Yamamoto M ., Katoaka, T. (2013) . In
ili co di c ery f mall-m lecule Ra inhi it r that di play antitumor activity by blocking the
Ra -effect r interaction. Proc. at/ Acad. Sci. USA. 110 : 8182- 187.
hin K ., Bae, ., Hong, H ., Kim, R ., Kang, M . Park, . (20 11 ). miR-181 a how tumor
uppre ive effect again t ral quamou cell carcinoma cell s by downregulating K-ras.
Biochem. Biophy . Re . ommun . 404 : 896-902 .
iebel C., Rio, D . (1990). Regulated splicing of the Dro ophila P transposable element third
intron in vitro : Somatic repression. Science . 248 : 1200-1208 .
Sparanese, D ., Lee, C. (2007) . CRD-BP hields c-myc and MDR-1 RNA from endonucleolytic
attack by a marrunalian endoribonuclease. Nucleic Acid Res. 35(4) : 1209-122 1.
Sun, Q . Burke, J. , Phan, J., Bums, M ., Olejniczak, E., Water on, A ., Lee, T. , Rossane e, 0 .,
Fesik, S. (20 12). Di covery of small molecules that bind to K-Ra s and inhibit sos-mediated
activation. Angew. Ch em. Int. Edn. 51 : 6140-6143.
Suzuki, M ., Imani hi, S., Dohmae, N ., Asanuma, M. , Mat umoto, S. (201 0) . Identifi cation of a
male- pecific RNA binding protein that regulates sex-specific plicing of Bmd x by increasing
RNA binding activity of BmPSI. Mol. Cell. Bioi. 30(24 ): 5776-5786.
Takahashi, Y ., Kitadai , Y ., Bucana , C . D ., leary, K . R ., and Ellis, L. M. (1995) . E pre ion of
vascular endothelial grow th factor and it receptor KDF, con elates with asculmity, meta tasis,
and proliferation of human co lon cancer. Can cer Re . 55 : 3964-3968.
orio, A.,
Tanic, M ., Yanowsky, K ., Rodriguez-An tona , C., Andres, R ., Marqu ez-Roda , I.,
Benit z, J. , Maiiin s-Delgado, B. (20 12). Deregul ated miRNA in heredi tary breast ca nc er
revealed a role for miR-3 0c in regulating KRA oncog ne. PLo. ONE. 7(6): e 84 7.
Te ier, ., Doyle, G ., lark, B., Pitot, H., Ro , J. (2004) . Mammary tumor indu tion in
transgenic mi ce expres ing an RN -binding pr t in. anccr Res. 64( 1): 209-2 14.

xxii

Tu
n , D ., haw , A ., Willi ,
il r, D ., Ja k on, ., hang . M ercer, K ., r ch w, R .,
H k H ., r wl
. tal. (2004) . nd g n u nc g nic K-ra ( 12D) timulate
pr li£ rati n and wid e pr ad ne la ti and de el pm ental d feet . an r ell . 5: 75- 7
agt gaa l. I. Paraf F., auric lla, ., Dimartino,
. V r n e . Laurent-Puig, P .,
1 na ,
.
, M ., Punt ., ambac rta, M ., Bardelli, A . Nic lant nio, F.
(201 ). K
gen amplifi ati n in c 1 r ctal cancer and impact n re p n e t
FR-targeted
th rap y. In t. J an r. 133 : 1259- 1266 .
.,

al rd , R ., dward ,
275 : 27 12-272 .

., R ga n L. (200 ). tructur and fun cti n f KH dom ain . FEES J.

V al rd R . P zdnyak a, I. , Kajander, T. , Venka tram an, J., Regan, L. (2 007). ragil X
m ental r tard ati n ndr m : tructure f th KH 1-KH 2 domain of fragil e X menta l retard ati n
pr tein . Stru tur . 15 : 1090- 109 .
V dadi , M . ie en , F ., A llali-Ha ani, A ., Fed r v ., inerty, P ., W a ney, ., Yeung, R .,
AlTow mith, ., Ball , L., B rglund , H ., Hu i, R . Mar den, B ., ordlund P ., und tr m , M .,
W eigelt, J. , dward , M . (2006) . hemi cal creening m eth d to identify ligand th at promote
protein tability, pr tein cry talli za ti on, and tructure determinati n. PNAS. 103 : 15 83 5- 15840.
Vike a, J. , Han n, T. , J n on, L. (2006) . RN -binding IMP promote cell adhe i n and
invadopodia fo rmati on. EMBO J 25 (7) : 1456- 1468 .
W ang, Z ., D ay N ., Tri fi lli P ., Kil edji an , M . (1999) . A n mRNA tability compl ex functions
with Poly(A)- Binding Protein to tabilize mRNA in vitro. Mo l. Cell. Bioi. 19(7) : 4552-4560.
W ang, Q., Zhang, Z. , Blackwell, K ., Carmichael, G . (2005) . Vigilin bind t promi cuou ly to-I-edited RNAs and are involved in the f01m ati on of heterochromatin. urr. Bioi. 15 : 3 4- 391 .
W ang, C ., M ackedenski. S., Lee, C . Inhibitor ofCRD -BP -KRA mRNA int raction. RiboWe t
conference, Edmonton A B (2015)
W eisz, B ., Giehl, K ., Gana-Weisz, M ., Egozi, Y ., Ben-Baru h, ., M arciano, D ., et al. (1 999). A
new functional R as antagonist inhibit human pancreatic turn r growth in nude mice. On coe;cne.
18 : 257 9- 2588 .
Wilkie, ., Davis, I. (200 1). D rosophila wingless and pair-rul e tran cript loca lize api all y by
dynein-m di ated tran port of RNA J article . Cell. 105 : 209-2 19.
Yanagihara, M ., I hikawa , ., Naito, M ., Nakajima , J ., bura tani , I-I , Hatakc ama, M. (~00 ).
Paired-like homeoprotein
X R 1 act a a , equence- p ifi tran cripti onal r "pressor o f the
human K-ra gene. On cogene. 24 : 5878-5 87 .

xxiii

Yang J. and Weinb rg R . . (200 ). pithelial -m enchymal tran ition: At the cro road of
d el p1n nt and tumor m ta ta i .
ll. 14 : 818- 29.
Yaru , K ., & Yi raeli, J. K . (2002) . The involvement of a con rved family of RNA binding
pr t in in embry nic d
1 pm nt and carcin gene i . G n . 287(1) : 49-54.
Y h, . H ., Hung, L. Y ., H u ., Le . Y ., ee P . T., Lia W . L. , .. . & T eng, J. T. (2008).
RN -binding Pr t in HuR Int ract
ith Thromb modulin 5' ntran lated R egion and
R epre e Internal Rib orne ntry ite- m diat d Translation und er IL- l~ Tr atment. Mol. Biol.
1!. 19(9): 3 12-3 22.
Yi ra li, J. , (2005) . VI KZ protein : a multi-talent d family of regulatory RNA -binding
prot in . Bioi
!1. 97 : 7-96 .
Yu ., Lu , Z., Liu, ., M ng, Y . Ma, Y ., Zha , W ., Liu , J . Yi J ., hen, J. (201 0). rniRNA -96
uppre e KRA and function as a tum r uppre r gen in pancreatic cancer . Cancer Res.
70(14): 6015-6025 .
Zhang,Y.
munaiti ,J., amuel, .K . hen ,P . and h en,Y. (2006) . Antitumor activity of an
oncolytic adenoviru -deli vered oncogene m all interference RNA . Cancer Re . 66:973 6-9743 .
Zilnm rmann, ., Papke, B ., I mail, ., Vartak, ., Chandra, A ., Hoffmann M ., et al. (2013) .
Small mol cul e inhibition of the KRA -PDEdelta inter- action impair oncogenic KRAS
signalling. Nature. 497 : 638- 642 .