Cell Reports · PDF fileCell Reports Article Mnk2 Alternative Splicing Modulates the p38-MAPK Pathway and Impacts Ras-Induced Transformation Avraham Maimon,1 Maxim Mogilevsky,1

Please cite this article in press as: Maimon et al., Mnk2 Alternative Splicing Modulates the p38-MAPK Pathway and Impacts Ras-Induced Transfor-mation, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.03.041

Cell Reports

Article

Mnk2 Alternative Splicing Modulatesthe p38-MAPK Pathway and ImpactsRas-Induced TransformationAvraham Maimon,1 Maxim Mogilevsky,1 Asaf Shilo,1 Regina Golan-Gerstl,1 Akram Obiedat,1 Vered Ben-Hur,1

Ilana Lebenthal-Loinger,1 Ilan Stein,2 Reuven Reich,3 Jonah Beenstock,4 Eldar Zehorai,5 Claus L. Andersen,6

Kasper Thorsen,6 Torben F. Ørntoft,6 Roger J. Davis,7 Ben Davidson,8,9 David Mu,10 and Rotem Karni1,*1Department of Biochemistry and Molecular Biology, Institute for Medical Research Israel-Canada2Department of Immunology and Cancer Research, Institute for Medical Research Israel-Canada3Department of Pharmacology, Institute for Drug Research4Department of Biological Chemistry

The Hebrew University-Hadassah Medical School, Jerusalem 91120, Israel5Department of Biological Regulation, Weizmann Institute of Science, Rehovot 76100, Israel6Department of Molecular Medicine, Aarhus University Hospital, Skejby, 8200 Aarhus N, Denmark7Howard Hughes Medical Institute and Program in Molecular Medicine, University of Massachusetts Medical School, Worcester,

MA 01655, USA8Department of Pathology, Oslo University Hospital, Norwegian Radium Hospital, Ullernchausseen 70, Oslo 0310, Norway9University of Oslo, Faculty of Medicine, Institute of Clinical Medicine, Oslo 0424, Norway10Department of Microbiology and Molecular Cell Biology, Leroy T. Canoles Jr. Cancer Research Center, Eastern Virginia Medical School,

651 Colley Avenue, Norfolk, VA 23501, USA

*Correspondence: [email protected]://dx.doi.org/10.1016/j.celrep.2014.03.041

This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

SUMMARY

The kinaseMnk2 is a substrate of the MAPK pathwayand phosphorylates the translation initiation factoreIF4E. In humans, MKNK2, the gene encoding forMnk2, is alternatively spliced yielding two splicingisoformswith differing last exons:Mnk2a, which con-tains a MAPK-binding domain, and Mnk2b, whichlacks it. We found that the Mnk2a isoform is down-regulated in breast, lung, and colon tumors and istumor suppressive. Mnk2a directly interactswith, phosphorylates, activates, and translocatesp38a-MAPK into the nucleus, leading to activationof its target genes, increasing cell death and sup-pression of Ras-induced transformation. Alterna-tively, Mnk2b is pro-oncogenic and does not activatep38-MAPK, while still enhancing eIF4E phosphoryla-tion. We further show that Mnk2a colocalization withp38a-MAPK in the nucleus is both required andsufficient for its tumor-suppressive activity. Thus,Mnk2a downregulation by alternative splicing is atumor suppressor mechanism that is lost in somebreast, lung, and colon tumors.

INTRODUCTION

The serine/threonine kinases Mnk1 and Mnk2 were discovered

by their direct interaction with and activation by the mitogen-

activated protein kinases (MAPKs) extracellular signal-regulated

kinase (ERK) and p38-MAPK (Fukunaga and Hunter, 1997; Was-

kiewicz et al., 1997). Mnk1 and Mnk2 phosphorylate the transla-

tion initiation factor eIF4E on serine 209 (Fukunaga and Hunter,

1997; Waskiewicz et al., 1997). The eIF4E protein binds to the

50 cap structure of mRNAs and is essential for cap-dependent

translational initiation (Mamane et al., 2004). In mice lacking

both kinases (MNK-double-knockout [DKO] mice), eIF4E is

completely unphosphorylated on serine 209 (Shenberger et al.,

2007; Ueda et al., 2004). Intriguingly, these mice develop and

live normally, displaying no adverse phenotype (Shenberger

et al., 2007; Ueda et al., 2004). Mnk1 andMnk2 are 72% identical

in their amino acid sequence (Fukunaga and Hunter, 1997; Was-

kiewicz et al., 1997). Biochemically, it has been shown that

whereas Mnk1 activation is dependent on upstream MAPK

signaling, Mnk2 possesses intrinsic basal activity when intro-

duced into cells (Scheper et al., 2003). There is limited evidence

connecting Mnk1/Mnk2 to human cancer, none directly. For

example, Mnk1 was shown to be downregulated during differen-

tiation of hematopoietic cells and upregulated in lymphomas

(Worch et al., 2004). In a functional study, overexpression of a

constitutively active Mnk1 mutant, or a phosphomimetic eIF4E

mutant, promoted c-Myc-mediated lymphomagenesis in vivo

(Wendel et al., 2007). Recently, it has been demonstrated that

mice lacking both Mnk1 and Mnk2 are more resistant to tumor

development when crossed with phosphatase and tensin homo-

log (PTEN�/�) mice, and mouse embryonic fibroblast (MEF) cells

from these mice are resistant to transformation by several onco-

genes, suggesting that eIF4E phosphorylation or the presence of

at least one Mnk is required for transformation (Furic et al., 2010;

Ueda et al., 2010). The notion that Mnk1 and Mnk2 are positive

drivers in human cancer stems from the important role their

Cell Reports 7, 1–13, April 24, 2014 ª2014 The Authors 1

mailto:[email protected]

http://dx.doi.org/10.1016/j.celrep.2014.03.041

http://creativecommons.org/licenses/by-nc-nd/3.0/

A B

C

D

E

F

G

Figure 1. A Switch in MKNK2 Alternative

Splicing in Primary Tumors and Cancer

Cell Lines

(A) The human-splicing isoforms of MKNK2

contain a basic region important for eIF4G binding

in their N terminus as well as a putative NLS. The

catalytic domain contains two conserved threo-

nine residues (T197 and T202) in the activation

loop that need to be phosphorylated (P) by MAP

kinases for kinase activation (Scheper et al., 2003).

Mnk2a contains a binding site for MAP kinases

located in the C terminus (Scheper et al., 2003).

Mnk2b is generated by an alternative 30 splice site

in intron 14 that generates a shorter last exon

(14b), lacking the MAPK-binding site (Scheper

et al., 2003).

(B) RNA levels of Mnk2a and Mnk2b in the indi-

cated primary or immortal breast and breast

cancer cell lines.

(C) The ratios of Mnk2a/Mnk2b RNA in non-

transformed breast cell lines (n = 4) and breast

tumor samples (n = 13) are represented by a box

plot. Top and bottom box edges represent the

third and first quartile. Whiskers indicate 90th and

10th percentiles; yellow and blue dots represent

minimum and maximum points, respectively. p

value is based on t test two-tailed analysis.

(D) RNA levels of SRSF1 in breast tumor samples

(n = 13) and nontransformed breast cell lines (n = 4)

are represented by a box plot as depicted in (C).

Median SRSF1 levels from nontransformed breast

cell lines (4) and breast tumor samples (13) are

shown. p value is based on t test two-tailed anal-

ysis.

(E–G) Pie chart representation of percentage (%) of

normal and tumor breast (E), lung (F), and colon (G)

samples with Mnk2a/Mnk2b ratios %1 (yellow

color) or >1 (blue color). Analysis is based on RNA-

seq data from TCGA project (https://tcga-data.nci.

nih.gov/tcga/).


known substrate, eIF4E, plays in cancer (Mamane et al., 2004),

and not from direct functional evidence.

In humans, each of the MKNK1 and MKNK2 genes gives rise

to at least two distinct proteins, with different C termini, as a

consequence of 30 prime alternative splicing (Parra-Palau et al.,

2003; Parra et al., 2005; Scheper et al., 2003). The longer forms

of human Mnk1 and Mnk2, referred to as Mnk1a and Mnk2a,

respectively, possess a MAPK-binding motif that is absent

from the shorter isoforms Mnk1b andMnk2b. In the correspond-

ing murine proteins, no short forms have yet been identified

(Parra et al., 2005). A recent study showed that resistance of

pancreatic cancer cells to gemcitabine is mediated by SRSF1

upregulation and a switch in Mnk2 alternative splicing, which

enhances eIF4E phosphorylation, implicating this alternative

splicing event with chemotherapy resistance (Adesso et al.,

2013).

Here, we examined the expression of Mnk2-splicing isoforms

in several types of cancer and manipulated the expression of

each isoform in normal and transformed cells. We found that

the Mnk2a isoform is downregulated in human cancers and is

2 Cell Reports 7, 1–13, April 24, 2014 ª2014 The Authors

a tumor suppressor, which colocalizes, interacts with, phosphor-

ylates, and activates p38-MAPK, leading to activation of its

target genes and to p38a-mediated cell death. These results

suggest that Mnk2a is a p38-MAPK kinase and acts as an up-

stream activator of p38-MAPK. Moreover, Mnk2a antagonizes

Ras-induced transformation in vitro and in vivo. However, the

Mnk2b isoform, or an inactive kinase-dead version of Mnk2a

(Mnk2aKD), did not activate the p38-MAPK pathway and was

pro-oncogenic. Taken together, our results suggest that Mnk2

alternative splicing serves as a switch in several cancers to

downregulate a tumor suppressor isoform (Mnk2a) that activates

the p38-MAPK stress pathway and to induce an isoform (Mnk2b)

that does not activate this pathway and is pro-oncogenic.

RESULTS

A Switch inMKNK2 Alternative Splicing in Breast, Lung,and Colon CancersMKNK2 is alternatively spliced to yield two isoforms: Mnk2a and

Mnk2b (Figure 1A) (Parra-Palau et al., 2003; Parra et al., 2005;

https://tcga-data.nci.nih.gov/tcga/



Scheper et al., 2003). Recently, we have found that the splicing

factor oncoprotein SRSF1 (SF2/ASF) modulates the splicing of

MKNK2, to reduce Mnk2a and increase Mnk2b (Karni et al.,

2007). To examine if changes in MKNK2 splicing are a general

phenomenon in cancer, we compared immortal and primary

breast cells to breast cancer cell lines, as well as to breast tumor

samples. We detected higher or equal expression of Mnk2a

compared to Mnk2b in immortal (MCF-10A, HMLE, HMT-3522-

S1) (Itoh et al., 2007) and primary breast cells (human mammary

epithelial cells). In contrast, Mnk2a expression was significantly

decreased, and in some cases, Mnk2b increased in tumor cell

lines and tumor samples (Figures 1B, 1C, and S1A). In most of

these tumors, SRSF1 was elevated (Figure 1D). To expand our

analysis, we analyzed RNA-sequencing (RNA-seq) data from

normal and cancer samples available from The Cancer Genome

Atlas (TCGA) project (https://tcga-data.nci.nih.gov/tcga/). Com-

parison of reads covering Mnk2a or Mnk2b splice junctions in

these samples shows that whereas only two out of 106 (1.9%)

normal breast samples have a Mnk2a/Mnk2b ratio equal to or

less than 1:1, 104 out of 994 (10.5%) breast tumor samples

have such a ratio. This trend is also observed in lung and colon

samples (Figures 1E–1G). In normal lung samples, Mnk2a level

was higher and Mnk2b level was lower compared to lung tumors

(Figures S2A–S2C). In the samemanner, the Mnk2a/Mnk2b ratio

in normal colon samples was higher compared to colon tumors

(Figure S2E). These results suggest that a ratio of Mnk2a/

Mnk2b less than 1:1, due to either reduced levels of Mnk2a

and/or increased levels of Mnk2b, is much more prevalent in

tumors than in normal tissue samples. We also examined the

correlation of expression of several serine/arginine-rich (SR)

and heterogeneous nuclear ribonucleoprotein (hnRNP) proteins

with Mnk2a/Mnk2b ratios. We found that in breast tumors,

expression of several SR and hnRNP A/B proteins (SRSF1,

SRSF2, SRSF3, SRSF4, SRSF7, SRSF9, hnRNP A1, hnRNP

A2/B1, and Tra2b) had a negative correlation (Pearson correla-

tion smaller than �0.1) with Mnk2a/Mnk2b ratios. Expression

of SRSF5 had a positive correlation with Mnk2a/Mnk2b ratios,

whereas expression of SRSF6 had no correlation (Table S1).

We found a similar, but not identical, pattern in lung and colon

tumors (Tables S2 and S3). Expression of the factors that

showed negative correlation with Mnk2a/Mnk2b ratios was

also elevated in breast tumors that showed aMnk2a/Mnk2b ratio

equal to or less than 1:1 (Figure S1I) and in lung and colon tumors

when compared to corresponding normal tissues (Figures S2D

and S2F).

To examine if oncogenic transformation affects MKNK2

splicing, we compared MCF-10A cells to MCF-10A cells trans-

formed by a mutant Ras oncogene (H-RasV12). We found that

oncogenic Ras reduced Mnk2a and elevated Mnk2b as well as

SRSF1 expression (Figures S1B and S1C). Similarly, pancreatic

cancer cell lines harboring mutant K-Ras showed reduced

Mnk2a and elevated Mnk2b levels compared to pancreatic

cancer cell lines with wild-type (WT) K-Ras (Figure S1D). The

pancreatic cancer cell line Panc-1, which harbors a K-Rasmuta-

tion, showed elevated levels of SRSF1 and Mnk2b and reduced

Mnk2a at both the RNA and protein levels compared to BXPC-3

cells harboring WT K-Ras (Figures S1E–S1G). This altered

splicing is also shown in another oncogenic-transformation

model: immortal lung epithelial cells (BEAS-2B) transformed

by SRSF1 (Shimoni-Sebag et al., 2013) (Figure S1H). These

data suggest that oncogenic transformation, caused either by

activated Ras or another oncogene, affects MKNK2 alternative

splicing.

Mnk2a Has Tumor-Suppressive Activity, whereasMnk2b Is Pro-oncogenic In VitroTo examine the role of Mnk2 alternative splicing in cellular trans-

formation, we seeded nontransformed breast MCF-10A cells

transduced with Mnk2-splicing isoforms into soft agar. Cells

expressing Mnk2b or a Mnk2aKD were transformed and gener-

ated colonies in soft agar, whereas cells expressing Mnk2a did

not (Figures 2A and 2C). Mnk2aKD probably acts in a domi-

nant-negative manner by competing with Mnk2a for substrate

binding, while incapable of phosphorylation. Similar results

were obtained in another transformation model of NIH 3T3 cells

(Figures S3A and S3B). Furthermore, when MCF-10A cells

expressing Mnk2-splicing isoforms were transformed by onco-

genic Ras, cells coexpressing Mnk2a showed reduced ability to

form colonies in soft agar, indicating that Mnk2a can block Ras-

induced transformation (Figures 2B and 2D). Similarly, Mnk2a in-

hibited colony formation in soft agar of the osteosarcoma cell line

U2OS (Figures S3C and S3D). Knockdown of Mnk2a enhanced

colony formation of MCF-10A and NCI-H460 cells in soft agar,

suggesting that Mnk2a is tumor suppressive (Figures S3E–

S3H). Neither Mnk2a nor Mnk2b expression changed signifi-

cantly cell proliferation or cell-cycle distribution (Figures S3I–

S3N).However, cellswithMnk2aknockdownhadaslightlyhigher

proliferation rate, indicating that Mnk2a reduction may enhance

proliferation (Figure S3J). Taken together, these results suggest

that the tumor-suppressive activity of Mnk2a is probably only

partly mediated through its effects on cellular proliferation.

Mnk2a Has a Tumor Suppressor Activity In VivoIn order to examine if Mnk2a possesses tumor suppressor

activity in vivo, we injected Ras-transformed MCF-10A cells

transduced with Mnk2-splicing isoforms into nonobese dia-

betic-severe combined immunodeficiency (NOD-SCID) mice.

We found that mice injected with Ras-MCF-10A cells expressing

either empty vector or Mnk2b formed tumors (six out of six),

whereas mice injected with Ras-MCF-10A cells expressing

Mnk2a did not form any tumors (zero out of eight) (Figure 2E).

Tumors from cells expressing Mnk2b showed an increased

mitotic index (Figures 2G and 2H) compared with tumors from

cells expressing Ras alone but did not show significant

enhanced tumor growth (Figure 2E). Inversely, mice injected

with Ras-MCF-10A cells expressing small hairpin RNA (shRNA)

against Mnk2a showed enhanced tumor growth rate (Figure 2F)

and increased mitotic index in the tumors (Figures 2I and 2J),

indicating that Mnk2a depletion cooperates with and enhances

Ras tumorigenicity. Collectively, our results suggest that

Mnk2a has tumor suppressor activity, and it can antagonize

Ras-mediated transformation in vitro and in vivo.

Mnk2a Sensitizes Cells to Stress-Induced Cell DeathBecause overexpression or downregulation of Mnk2a did not

affect cellular proliferation significantly (Figures S3I–S3N), we



A B

C D

E F

G H I J

Figure 2. Mnk2a Inhibits Ras-Induced

Transformation and Tumorigenesis

(A) Western blot of MCF-10A cells transduced

with the indicated retroviruses encoding for Mnk2

isoforms and Mnk2aKD.

(B) Western blot of cells described in (A)

transduced with a retrovirus encoding for

H-RASV12.

(C) Cells described in (A) were seeded into soft

agar (see Experimental Procedures) in duplicates,

and colonies were allowed to grow for 14 days.

Colonies in ten fields of each well were counted,

and the mean and SD of colonies per well of three

wells are shown (n = 3).

(D) Cells described in (B) were seeded into soft

agar and counted as in (C) (n = 3).

(E and F) Pools of MCF-10A cells transformed with

the indicated retroviruses encoding Mnk2 iso-

forms followed by H-RASV12 transduction (E) or

MCF-10A cells transduced with shRNA against

Mnk2a (Sh2a-2) followed by transformation by

H-RASV12 (F) were injected (23 106 cells/injection,

in Matrigel) subcutaneously into NOD-SCID

mice. Tumor growth curves were calculated as

described in Experimental Procedures. The num-

ber of tumors formed per injection is shown near

the legend bars. *p % 0.01.

(G–J) Formalin-fixed, paraffin-embedded tissue

sections from tumors derived from the indicated

cell pools described in (E) and (F) were stained with

anti-phospho-H3 to detect mitotic cells (G and I).

Graphs show the average and SD of p-H3-positive

cells from ten fields of three different tumors (H

and J). *p % 0.01.


hypothesized that Mnk2a might enhance the sensitivity of cells

to apoptosis. To examine the possible role of Mnk2-splicing

isoforms in the response to cellular stress, we challenged

immortalized breast cells (MCF-10A and Ras-transformed

MCF-10A cells) transduced with retroviruses encoding Mnk2a,

Mnk2b, or Mnk2aKD with different stress conditions. Although

Mnk2a enhanced apoptotic cell death in response to anisomycin

treatment, as measured by trypan blue exclusion and caspase-3

cleavage, Mnk2b and the Mnk2aKD protected against

apoptosis (Figures 3A and 3D). Moreover, knockdown of

Mnk2a protected MCF-10A cells from anisomycin-induced

apoptosis (Figures 3B and 3E). The correlation of apoptosis

and caspase-3 cleavage with p38-MAPK phosphorylation

(Figures 3D and 3E) suggests that Mnk2a proapoptotic activity

might involve activation of the p38-MAPK pathway. Mnk2a

also reduced survival of Ras-transformed MCF-10A cells


and MCF-10A cells forced to grow in

suspension or stimulated with osmotic

shock (Figures S4A and S4C). In contrast,

knockdown of Mnk2a protected cells

from osmotic shock, suggesting that

Mnk2a mediated this stress response

(Figure S4D). Mnk2a also inhibited

the survival of MCF-10A cells trans-

formed by oncogenic Ras when sparsely

seeded for colony survival assay (Figures 3C and 3F), sug-

gesting that Mnk2a sensitizes Ras-transformed cells to

low-density stress conditions. One of the stress pathways

induced by anisomycin and other cellular insults is the

p38-MAPK pathway (Benhar et al., 2001; Cuadrado and

Nebreda, 2010; Hazzalin et al., 1996). In order to examine

if p38-MAPK activation is involved in Mnk2a-enhanced cell

death, we blocked p38-MAPK kinase activity with the

specific kinase inhibitor ATP competitor SB203580 (Badger

et al., 1996; Benhar et al., 2001; Lee et al., 2000). p38-MAPK

inhibition partially rescued cells expressing Mnk2a from ani-

somycin-induced cell death (Figure S4B). Neither transient

expression nor stable expression of Mnk2a alone reduced

survival of MCF-10A cells, suggesting that Mnk2a exerts its

proapoptotic properties only under stress conditions (Figures

S4F and S4G).

A B C

D E F

G

H

Figure 3. Mnk2a Sensitizes MCF-10A Cells

to Stress-Induced Apoptosis

(A and B) Trypan blue exclusion assay of MCF-10A

cells transduced by the indicated retroviruses.

Following 24 hr in growth factor-free media, cells

were treated with 0.5 mM anisomycin for 24 hr (A)

or 48 hr (B). *p % 0.05; **p % 0.01.

(C) Colony formation assay of pools of MCF-10A

cells transduced with the indicated retroviruses

encoding Mnk2 isoforms and transformed by

H-RASV12. Average and SD of the number of col-

onies per well are shown (n = 3). *p % 0.05.

(D and E) Cells described in (A) and (B) were

analyzed by western blot. Cleaved caspase-3

served as a marker for apoptosis. Levels of p-p38,

p38, p-eIF4E, and eIF4E were detected with the

indicated antibodies. b-catenin served as loading

control. Numbers represent ratios of p-p38/total

p38 and p-eIF4E/total eIF4E normalized to that of

pBABE or MLP (arbitrarily set at 1) ± SD (n = 3).

(F) Representative wells with colonies described

in (C).

(G) MCF-10A cells transduced with the indicated

retroviruses followed by transduction with

H-RASV12 were seeded into soft agar in the pres-

ence or absence of the indicated concentrations of

SB203580, and colonies were counted 14 days

later. *p % 0.01; **p % 0.05 (n = 3).

(H) Photographs of representative fields of col-

onies in soft agar obtained as described in (G).


Mnk2a, but Not Mnk2b, Enhances p38a-Mediated CellDeath and Suppression of Ras-Induced TransformationBecause Mnk2a, but not Mnk2b, contains a MAPK-binding

domain (Figure 1A) and can be activated by ERK and p38-

MAPK (Parra et al., 2005; Scheper et al., 2003), we examined if

Mnk2a can mediate stress responses emanating from activated

p38-MAPK. MCF-10A cells expressing either Mnk2 isoforms or

knocked down for Mnk2a were transduced with a constitutively

active p38amutant (Askari et al., 2007; Avitzour et al., 2007; Dis-

kin et al., 2004) and grown in the absence or presence of the p38

Cell Reports 7,

kinase inhibitor SB203580. Cells express-

ing Mnk2a showed increased cell death

upon active p38-MAPK transduction,

which was inhibited by SB203580 (Fig-

ures S5A and S5B). Cells in which

Mnk2a was knocked down showed

increased protection from p38-induced

cell death (Figures S5C and S5D).

SB203580 efficiently inhibited p38

activity, as was measured by phosphory-

lation of its substrate MK2. However, it

did not affect phosphorylation of p38-

MAPK or eIF4E, indicating that

SB203580 treatment did not inhibit up-

stream Mnk2 or MKK3/MKK6 activity

(Figure S5E). These results suggest that

Mnk2a augments p38-MAPK stress

activity. Inhibition of p38-MAPK by

SB203580 rescued the ability of cells

cotransduced with Mnk2a and Ras to form colonies in soft

agar, suggesting that p38 activation by Mnk2a plays an impor-

tant role in its ability to suppress Ras-induced transformation

(Figures 3G and 3H).

Mnk2a Interacts with, Activates, and Induces NuclearTranslocation of p38-MAPKTo determine if Mnk2a activates p38-MAPK, we examined

the phosphorylation status of p38-MAPK in cells expressing

Mnk2-splicing isoforms (Enslen et al., 1998). As expected,

1–13, April 24, 2014 ª2014 The Authors 5

A

B

C

D

E

F

Figure 4. Mnk2a Interacts with p38-MAPK

and Leads to Its Activation and Transloca-

tion into the Nucleus

(A) Western blot analysis of MCF-10A cells trans-

duced with the indicated retroviruses encoding for

Mnk2 isoforms or Mnk2aKD. Numbers represent

the ratios of p-p38/total p38 and p-eIF4E/total

eIF4E normalized to pBABE (arbitrarily set at 1) ±

SD (n = 3).

(B) Coimmunoprecipitation of HA-p38a or T7-

Mnk2 isoforms from HEK293 cells cotransfected

with the indicated Mnk2 isoforms together with

HA-p38a-MAPK. Precipitated and input proteins

were subjected to western blot analysis. The

asterisk (*) represents a nonspecific band. As a

control, lysate from cells transfected with T7-

Mnk2a was incubated with beads in the absence

of the T7 antibody. IP, immunoprecipitation; IB,

immunoblotting.

(C) Western blot analysis of HeLa cells co-

transfected with either Mnk2a or MKK3 together

with HA-p38a-MAPK. b-catenin was used as a

loading control.

(D) Left: western blot analysis of MEF cells from

WT mice (MKK3/MKK6 WT) and MKK3/MKK6�/�

DKO mice. Right shows a western blot analysis of

MEF cells from MKK3/MKK6�/� DKO mice (MEF

MKK3/MKK6�/�) transduced with the indicated

retroviruses. b-catenin was used as a loading

control.

(E) In vitro kinase assay using recombinant Mnk2a,

recombinant kinase dead (KD) p38a, recombinant

eIF4E alone, or in the indicated combinations.

Reaction mixes were stopped, separated on

SDS-PAGE, and transferred onto nitrocellulose

membranes. Upper panel shows a radioactive

exposure of the membrane detected by a Fuji

phosphorimager cassette and visualized by

MacBass phosphorimager. Bottom panel shows

the same membrane as above that was probed

with the indicated antibodies and detected by

enhanced chemiluminescence.

(F) Cellular distribution of p38a in MCF-10A cells,

transduced with retroviruses encoding the indi-

cated Mnk2 isoforms. N, nuclear fraction; C,

cytoplasmic fraction. c-myc (nuclear) and

caspase-2 (cytoplasmic) served as fractionation

controls.


phosphorylation of a known substrate of Mnk2, serine 209 of

eIF4E, was induced by Mnk2a expression (Figures 4A, S3A,

and S3C). Although previous reports have suggested that

Mnk2b has a lower kinase activity than Mnk2a (Scheper et al.,

2003), we observed that it phosphorylates eIF4E to a similar

extent asMnk2a (Figures 4A, S3A, SC3, and S5E). TheMnk2aKD

did not enhance eIF4E phosphorylation (Figures 4A, S3A, and

S3C). In contrast, only cells expressing Mnk2a showed

increased p38-MAPK phosphorylation, suggesting that p38-

MAPK is activated by Mnk2a (Figures 4A, 4C, 4D, S5E, and

S6A) (Avitzour et al., 2007; Enslen et al., 1998). Moreover, knock-

down of Mnk2a in MCF-10A cells reduced p38-MAPK basal

phosphorylation level (Figures 3E and S3E). In addition, we

determined that the phosphorylation state of the p38-MAPK

substrate MK2 was enhanced only in cells expressing Mnk2a,


but not Mnk2b or Mnk2aKD (Figure 4A). Knockdown of Mnk2a

inhibited MK2 phosphorylation, and inhibition of p38-MAPK by

the kinase inhibitor SB203580 abolished MK2 phosphorylation

in cells expressing Mnk2a (Figure S5E). To further examine if

the kinase activity of Mnk2a is important for p38 phosphoryla-

tion, we treated MCF-10A cells with the Mnk1/Mnk2 kinase in-

hibitor CGP 57380 (Knauf et al., 2001) (Chrestensen et al.,

2007). Mnk1/Mnk2 kinase inhibition reduced p38-MAPK phos-

phorylation on the known MEK3/MEK6 phosphorylation sites

T180 and Y182, similar to its effect on eIF4E S209 phosphoryla-

tion (Figure S6B). Finally, we examined p38 phosphorylation in

immortal breast and breast cancer cell lines and observed that

in the latter (which tend to express lower Mnk2a levels) (Fig-

ure 1B), both p38-MAPK and eIF4E phosphorylation is lower

than in the nontransformed cells (Figures S6C–S6E).


We next examined whether Mnk2 isoforms can differentially

interact with p38a-MAPK in cells. Coimmunoprecipitation of

transfected or endogenous p38a from HEK293 cells demon-

strated that Mnk2a and Mnk2aKD, unlike Mnk2b, efficiently

bound p38a (Figure 4B). Importantly, even though Mnk2aKD

was bound to p38a, it did not cause activation of p38a, as

measured by p38a or MK2 phosphorylation (Figures 3D, S5E,

and 4A). Finally, to rule out the possibility that Mnk2 isoforms

compete with Mnk1 for p38a binding, we examined Mnk1 bind-

ing to p38a. Mnk1 was bound to HA-p38a; however, its binding

was not affected by any of the Mnk2 isoforms (Figure 4B). Taken

together, these results suggest that Mnk2a interacts with p38a

and leads to its activation.

Several p38-MAPK isoforms other than p38a exist and are

involved in stress signaling and apoptosis (Cuadrado and

Nebreda, 2010; Turjanski et al., 2007). Thus, we examined if

Mnk2a enhances the phosphorylation of other p38-MAPK family

members when cotransfected into HeLa cells. We found that

Mnk2a enhanced the phosphorylation of p38a and p38b, only

weakly elevated phosphorylation of p38g, and had no significant

effect on phosphorylation of p38d (Figure S6F). We next exam-

ined the interaction of Mnk2a with the other p38-MAPK mem-

bers by coimmunoprecipitation. In correlation with the ability of

Mnk2a to phosphorylate the different p38-MAPK isoforms, we

found that Mnk2a interacts with p38a and p38b, but we could

not detect interaction with p38g or p38d (Figures S6G and

S6H). These results suggest that Mnk2a kinase activity is spe-

cific to p38a and p38b.

Cotransfection of HeLa cells with p38a and WT Mnk2a

showed higher p38a phosphorylation levels than cells co-

transfected with WT MKK3, indicating that (without upstream

activation) Mnk2a’s ability to phosphorylate p38a is as strong

or stronger than MKK3 (Figure 4C). To examine if Mnk2a can

directly phosphorylate p38-MAPK in the absence of the known

p38-MAPK kinasesMKK3 andMKK6 (Enslen et al., 1998; Rincon

and Davis, 2009), we transduced MEF cells from MKK3/MKK6

DKO mice (Brancho et al., 2003) with retroviruses encoding for

Mnk2a or an empty vector. Mnk2a induced p38-MAPK phos-

phorylation in the absence of MKK3 and MKK6, suggesting

that it can directly phosphorylate p38-MAPK (Figure 4D). To

further examine if Mnk2a directly phosphorylates p38-MAPK,

we performed an in vitro kinase assay with recombinant

WT Mnk2a and kinase-dead p38a proteins and found that

Mnk2a can directly phosphorylate p38a-MAPK (Figure 4E).

Taken together, these results suggest that Mnk2a is a direct

p38a-MAPK activator kinase.

Mnk2a Colocalizes with p38-MAPK and Affects ItsCellular LocalizationUpon activation, p38-MAPK is translocated to the nucleus and

phosphorylates transcription factors that mediate some of its

stress response (Aplin et al., 2002; Gong et al., 2010; Pfundt

et al., 2001; Plotnikov et al., 2011). Using cytoplasmic and

nuclear fractionation, as well as immunofluorescent staining,

we observed that both Mnk2a and Mnk2b can be detected in

the nucleus (Figures 4F, S6I, and S7). However, cells that ex-

press Mnk2a showed an increased nuclear fraction of total and

phosphorylated p38-MAPK (Figures 4F, 5B, S6I, S7A, and

S7B), indicating that Mnk2a leads to both p38-MAPK activation

and its translocation into the nucleus.

In order to examine if Mnk2a affects p38-MAPK cellular local-

ization, we generated two Mnk2a mutants. In the first mutant

(KKR), we mutated the putative nuclear localization signal of

Mnk2a (69-KKRGKKKKR-77) to KKRGKKAAA, in which the last

KKR was replaced with three alanines (AAA). This mutant is ex-

pected to be mostly cytoplasmic, as was shown for the homolo-

gous mutation in Mnk1(Parra-Palau et al., 2003). In the second

mutant (L/S), we mutated the putative nuclear export signal

(NES) of Mnk2. Although in Mnk1, the NES motif is localized to

a different region (Parra-Palau et al., 2003), we identified a similar

motif (LxxxLxxL) in the C-terminal region of Mnk2 (starting at

amino acid 281 ofMnk2a) andmutated the last two lysines to ser-

ines. Transfection of these mutants into HeLa cells or transduc-

tion into MCF-10A cells showed the expected localization: the

nuclear localization of Mnk2a L/S was enhanced, whereas that

of Mnk2a KKR was decreased, when compared to that of

Mnk2a (Figures 5A and S7).When cotransfected with HA-tagged

or GFP-tagged p38-MAPK, Mnk2a colocalized with p38-MAPK.

Mnk2a L/S rendered p38-MAPK mostly nuclear and colocalized

with it in the nucleus. Mnk2a KKR colocalized with p38-MAPK in

both the cytoplasmandnucleus butwas less nuclear thanMnk2a

(Figures 5A, S7A–S7D, and S7F). Both Mnk2a KKR and the L/S

mutants can interact with HA-tagged and endogenous p38a,

as was demonstrated by coimmunoprecipitation, and can pull

down phospho-p38a (Figure S7E). In addition, the effects of

the L/S and KKR mutants on the localization of endogenous

p38a were similar to their effects on HA-p38 in MCF-10A cells

transduced with retroviruses expressing these mutants (Fig-

ure S7A). Overall, these results suggest that Mnk2a and p38-

MAPK are colocalized in both the cytoplasm and nucleus and

that Mnk2a can affect the cellular localization of p38a-MAPK.

Mnk2a Localization Affects Induction of p38-MAPKTarget Genes and ApoptosisWe next examined the expression of FOS and cyclo-oxygenase

2 (COX-2), both targets of p38-MAPK stress response (Ferreiro

et al., 2010). Expression of both genes was induced in MCF-

10A cells expressing Mnk2a and reduced in cells expressing

Mnk2b or Mnk2aKD (Figure 6A). Moreover, knockdown of

Mnk2a inhibited FOS and COX-2 expression (Figure 6B). Inter-

estingly, whereas Mnk2a L/S could activate p38-MAPK target

genes similarly to Mnk2a, the KKR mutant did not increase the

expression of FOS and COX-2 (Figure 6C). To examine if

Mnk1/Mnk2 kinase activity modulates the expression of p38a

target genes, we treated MCF-10A cells with the Mnk1/Mnk2

kinase inhibitor CGP 57380. Inhibition of Mnk1/Mnk2 kinase

activity, which reduced p38-MAPK phosphorylation (Fig-

ure S6B), also reduced the expression p38-MAPK target genes

(Figure 6D). MEF cells from mice with a DKO of both activators

of p38-MAPK, MKK3 and MKK6 (MKK3/MKK6 DKO), show

very low basal phosphorylation of p38-MAPK compared to WT

cells (Brancho et al., 2003) (Figure 4D). Introduction of Mnk2a

into MKK3/MKK6 DKOMEFs elevated p38-MAPK phosphoryla-

tion (Figure 4D) and also induced the expression of FOS and

COX-2 (Figure 6E). Mnk2b inhibited the expression of FOS and

COX-2 below the basal level (Figure 6A). Mnk2aKD inhibited


A

B

Figure 5. Mnk2a Colocalizes with p38-

MAPK and Affects Its Cellular Localization

(A) Immunofluorescence of HeLa cells co-

transfected with either empty pCDNA3 vector or

pCDNA3-based expression vectors for T7-tagged

Mnk2a, Mnk2aL/S, or Mnk2aKKR, together with

pCDNA3-GFP-p38a (WT) as described in Experi-

mental Procedures. T7 tag is red, and GFP-p38a

is green. Right view shows quantification of cyto-

plasmic/nuclear distribution of GFP-p38a in cells

similar to those described (n = 40 for each mutant).

*p % 0.01.

(B) Immunofluorescence of MKK3/MKK6�/� MEF

cells transduced with the indicated retroviruses.

T7 tag was stained red, and p38a was stained

green. Right view shows quantification of cyto-

plasmic/nuclear distribution of p38a in cells similar

to those described.( n = 57 for pBABE; n = 47 for

Mnk2a). *p = 9.6 3 10�17.


the expression of COX-2 and had no effect on the expression of

FOS (Figure 6A). This is possibly due to other regulatory ele-

ments in the FOS promoter making it less responsive to inhibition

of MNK2a, as seen from the kinase inhibitor experiment (Fig-

ure 6D). To examine if Mnk2a localization can affect its proapo-

ptotic activity, we treated cells expressing Mnk2a isoforms and

the KKR and L/S mutants with anisomycin and measured cell

death. We found that whereas Mnk2a and the L/S mutant sensi-

tized cells to apoptosis, the KKR mutant inhibited apoptosis and

did not significantly decrease colony survival, in correlation with

its inability to induce p38a target genes (Figures 6F and S4E).

Taken together, these results indicate that Mnk2a not only acti-

vates p38-MAPK, but it also regulates the expression of p38-

MAPK target genes, altering the transcriptional program of the

cells and leading to induction of apoptosis.

Modulation of MKNK2 Splicing by Splice Site-Competitive Antisense RNA Oligos Affects Sensitivity toApoptosis and Cellular TransformationTo examine if modulation of endogenous MKNK2 alternative

splicing affects the oncogenic properties of cells, we used two


experimental systems in which Mnk2a/

Mnk2b ratios were altered due to cellular

transformation: MCF-10A cells trans-

formed by oncogenic Ras (Figures S1B

and S1C), and immortal lung bronchial

epithelial cells, BEAS-2B, transformed

by SRSF1 overexpression (Shimoni-

Sebag et al., 2013) (Figure S1H). We

designed Cy5-labeled 20-O-met-RNA

antisense oligos to mask the MKNK2

splice sites, shifting the splicing balance

between Mnk2a and Mnk2b (Figures 7A

and 7B). Elevation of Mnk2a and reduc-

tion of Mnk2b by the oligo that

blocks production of Mnk2b (2b Block)

sensitized cells to anisomycin-induced

apoptosis, reduced colony survival, and

inhibited soft agar colony formation in both Ras-transformed

MCF-10A cells (Figure 7) and SRSF1-transformed BEAS-2B

cells (Figure S8). The 2b block oligo also enhanced anisomy-

cin-induced apoptosis of nontransformed MCF-10A cells (Fig-

ure 7G). Elevation of Mnk2b and concomitant reduction of

Mnk2a, by the oligo that competes with Mnk2a intron-exon junc-

tion (2a block), protected Ras-transformed MCF-10A cells and

SRSF1-transformed BEAS-2B cells from anisomycin-induced

apoptosis, and increased colony survival (Figures 7E, 7F, S8E,

and S8F). Introduction of the 2a block oligo also protected non-

transformedMCF-10A cells from anisomycin-induced apoptosis

(Figure 7G). The 2a block oligo did not increase anchorage-inde-

pendent growth in soft agar (Figures 7C, 7D, S8C, and S8D),

probably because cells expressing oncogenic Ras or SRSF1

are already highly invasive and tumorigenic (see Figure 2D for

Ras-transformed and Shimoni-Sebag et al., 2013 for SRSF1-

transformed cells). As in the case of Mnk2a or Mnk2b over-

expression (Figures S4F and S4G), in normal conditions without

stress, modulation ofMKNK2 splicing by the antisense oligos did

not induce apoptosis on its own (Figures S8G and S8H). These

data suggest that manipulation of the endogenous ratios of

A B

C D

E F

Figure 6. Mnk2a Localization and Kinase

Activity Are Required for Induction of p38a

Target Genes and Apoptosis

(A and B) Quantitative RT-PCR (qRT-PCR) of RNA

from MCF-10A cells transduced either with the

indicated Mnk2 isoforms or Mnk2aKD or with MLP

vectors encoding for shRNAs against Mnk2a iso-

form.

(C) qRT-PCR of RNA from MCF-10A cells trans-

duced with either the indicated Mnk2 isoforms or

the indicated Mnk2a mutants.

(D) qRT-PCR of RNA from serum-starved MCF-

10A cells treated with Mnk1/Mnk2 inhibitor CGP

57380 at the indicated concentrations for 8–10 hr.

(E). qRT-PCR of RNA from MKK3/MKK6�/� MEF

cells transduced with the indicated retroviruses.

*p % 0.01.

(F) Trypan blue exclusion assay of MCF-10A cells

described in (C) serum starved for 24 hr and then

treated with 0.5 mM anisomycin for 24 hr (graph).

Bottom panel shows a western blot analysis of the

cells described. Cleaved caspase-3 served as a

marker for apoptosis and b-catenin as a loading

control.


Mnk2a and Mnk2b affects the oncogenic potential of cells trans-

formed by Ras and SRSF1.

DISCUSSION

The process of alternative splicing is widely misregulated in

cancer, and many tumors express new splicing isoforms,

which are absent in the corresponding normal tissue (Venables

et al., 2009; Xi et al., 2008). Many oncogenes and tumor suppres-

sors are differentially spliced in cancer cells, and it has been

shown that many of these cancer-specific isoforms contribute

to the transformed phenotype of cancer cells (Srebrow and

Kornblihtt, 2006; Venables, 2004). Here, we have shown that

MKNK2 alternative splicing ismodulated in cancer cells to down-

regulate the expression of the tumor-suppressive isoformMnk2a

and enhance the expression of the pro-oncogenic isoform

Mnk2b. Both splicing isoforms phosphorylate the translation

initiation factor eIF4E. However, only Mnk2a binds to and acti-

vates p38-MAPK, leading to enhanced activation of the p38

Cell Reports 7,

stress pathway, induction of its target

genes, and enhanced cell death.

Previously, we identified the splicing

factor SRSF1 (SF2/ASF) as a potent

proto-oncogene and reported MKNK2

as one of its splicing targets (Karni et al.,

2007) (Anczukow et al., 2012). Enhanced

expression of SRSF1 reduced the levels

of Mnk2a and increased the levels of

Mnk2b, whereas knockdown of SRSF1

caused a reciprocal change in MKNK2

splicing (Karni et al., 2007) (Anczukow

et al., 2012). Here, we found that Mnk2a

is downregulated, whereas in some

cases, Mnk2b is upregulated, and the

level of SRSF1 is upregulated in breast cancer cell lines and

tumors (Figures 1A–1D). Moreover, oncogenic transformation

by Ras elevated the levels of SRSF1 and induced a shift in

MKNK2 splicing (Figures S1B and S1C). Pancreatic cancer cell

lines harboring mutant K-Ras (such as Panc-1) had elevated

SRSF1 levels and altered MKNK2 splicing compared to a

pancreatic cancer cell line with WT K-Ras (Figures S1D–S1G).

In another model system of oncogenic transformation, by

SRSF1 (Shimoni-Sebag et al., 2013),MKNK2 alternative splicing

was also shifted (Figure S1H), suggesting that changes in

MKNK2 splicing occur upon transformation with oncogenes

such as Ras and SRSF1.

Recently, it has been shown that Mnk2, but not Mnk1, inhibits

protein translation through its negative effect on eIF4G Ser1108

phosphorylation and by inhibiting mammalian target of rapamy-

cin activity (Hu et al., 2012). An earlier study detected similar

effects of Mnk2 on translation (Knauf et al., 2001). These results

suggest that Mnk2 might possess a tumor-suppressive activity

by inhibiting translation, an additional mechanism to the one

1–13, April 24, 2014 ª2014 The Authors 9

A B

C D E

F G

Figure 7. Modulation of MKNK2 Splicing by

Splice Site-Competitive Antisense RNA

Oligos Affects Sensitivity to Apoptosis and

Cellular Transformation

(A) Schematic diagram of splice site-interfering

RNA oligonucleotides. 20-O-methyl RNA oligos are

labeled with Cy5 at the 50 and interfere with U1/U2

binding to the 30 splice site of either exon 14a (2a

Block) or exon 14b (2b Block).

(B) RT-PCR of RNA from Ras-transformed MCF-

10A cells transfected with the indicated Cy5-

labeled antisense 20-O-methyl RNA oligos.

Numbers ± SD represent ratio of Mnk2a/Mnk2b

normalized to that of cells transfected with the

scrambled (Scr) oligo (arbitrarily set at 1) (n = 3).

(C) Soft agar growth assay of cells transfected as

in (B). Colonies in ten fields of each well were

counted, and themean and SD of colonies per well

of three wells are shown (n = 3).

(D) Photographs of representative fields of col-

onies in soft agar obtained as described in (C).

(E) Colony formation assay of cells transfected as

in (B). Average and SD of the number of colonies

per well are shown (n = 3).

(F) Cell survival assay and western blot analysis of

MCF-10A-Ras cells transfected as in (B) incubated

in the presence of anisomycin for 36 hr. Cleaved

caspase-3 (Clv. Casp-3) is a marker for apoptosis,

and b-catenin was used as a loading control.

(G) Cell survival assay and western blot analysis of

MCF-10A cells transfected and treated as in (F).


proposed in our study. Analyzing RNA-seq data from hundreds

of tumors from patients with breast, colon, and lung cancer

deposited in the TCGA database, we found that the MKNK2

alternative splicing switch occurs in a significant portion of

breast, lung, and colon tumors (Figures 1E–1G and S1) and is

either weakly correlated or inversely correlated with expression

of several splicing factors from the SR and hnRNP A/B families

(Tables S1–S3). These results suggest that a combinatorial

network of splicing factors (rather than a single splicing factor)

controls MKNK2 alternative splicing in tumors. Upregulation of

oncogenic-splicing factors such as SRSF1 can be caused by

several mechanisms such as gene amplification (Karni et al.,


2007) or transcriptional activation (Das

et al., 2012) and may lead to a change in

MKNK2 alternative splicing as has been

shown previously (Karni et al., 2007).

To elucidate the role of Mnk2-splicing

isoforms in cancer development, we

examined the oncogenic activity of

Mnk2-splicing isoforms in vitro and

in vivo. Results of these experiments

suggest that Mnk2a possesses a tumor-

suppressive activity in vitro and in vivo,

and this activity of Mnk2a is mediated

by activation of the p38-MAPK pathway.

Interestingly, both Mnk2a and Mnk2b,

but not the Mnk2aKD, phosphorylated

eIF4E to a similar extent, suggesting

that eIF4E phosphorylation cannot account for their different

biological activity (Figures 4A, S3A, and S3C). In cancer cells,

when alternative splicing results in reduced expression of

Mnk2a and increased expression of Mnk2b or when it is manip-

ulated artificially, as we have done here, there is still phos-

phorylation of eIF4E (Figures 3D and 4A), but no activation of

the p38-MAPK pathway, which is mediated by Mnk2a. Thus,

our results suggest that Mnk2b uncouples eIF4E phosphoryla-

tion from activation of the p38-MAPK stress pathway and thus

sustains only the pro-oncogenic arm of the pathway (Figures 3,

4, S5, and S6). However, we cannot rule out the possibility that

Mnk2b can phosphorylate other substrates (yet to be found)


that contribute to its oncogenic activity and, in this manner, acts

as a pro-oncogenic factor. We further found that Mnk2 isoforms

can differentially interact with p38a-MAPK in cells. In coimmuno-

precipitation assays, Mnk2a binds p38a, whereas Mnk2b does

not bind p38a efficiently, suggesting that this interaction might

be important for p38a activation by Mnk2a (Figure 4B). The

ability of Mnk2a to phosphorylate p38-MAPK was similar in

strength or even stronger than MKK3 (Figure 4C), and Mnk2a

can induce p38-MAPK phosphorylation in the absence of

MKK3 and MKK6 (Figure 4D). These results combined with the

results of an in vitro kinase assay (Figure 4E) suggest that

Mnk2a is a direct p38a-MAPK activator kinase. To the best of

our knowledge, Mnk2a is the first direct kinase, other than

MKK3, MKK6, and MKK4, that can phosphorylate and activate

p38-MAPK. Some scaffold proteins can enhance p38-MAPK

autophosphorylation and induce its activation (De Nicola et al.,

2013). However, this cannot explain our finding because we

used a kinase-dead recombinant p38a and recombinant

Mnk2a in the in vitro kinase assay (Figure 4E). We also found

that Mnk2a can bind and phosphorylate both p38a and p38b,

but not p38g or p38d, suggesting that Mnk2a has specificity to-

ward these two substrates (Figures 4A, 4B, and S6F–S6H).

We further confirmed that the kinase activity of Mnk2a is

required for activation of p38a and its target genes because

both the Mnk2aKD or application of the Mnk1/Mnk2 kinase

inhibitor CGP 57380 (Knauf et al., 2001; Chrestensen et al.,

2007) inhibited these activities (Figures S6B and 6). Taken

together, these results indicate that Mnk2 kinase activity is

required for the activation of p38-MAPK and its downstream

targets.

In all of our experimental systems, we demonstrated that

whereas both p38 phosphorylation and eIF4E phosphorylation

are enhanced by Mnk2a overexpression and reduced by its

knockdown, only p38 phosphorylation correlates with the de-

gree of apoptosis (Figure 3). Results from these gain- and loss-

of-function experimental systems suggest that p38-MAPK, but

not eIF4E, phosphorylation/activation determines the fate of

these cells. Finally, in the breast cancer cell lines examined in

this paper, there are lower levels of p38 (and of eIF4E) compared

with two immortal breast cell lines (Figure S6), suggesting clinical

relevance of these findings. The role of the tumor suppressor

activity of p38a was recently demonstrated in hepatocellular

carcinoma and colon cancer development. In both cases,

tissue-specific knockout of p38a led to cancer development

in vivo (Sakurai et al., 2013) (Wakeman et al., 2012). Moreover,

a recent study showed that the gene encoding MKK3, a known

p38-MAPK upstream kinase, is lost in some breast tumors and

that MKK3 acts as a tumor suppressor in breast cancer (MacNeil

et al., 2014). Because Mnk2a acts as another p38-MAPK kinase,

it is reasonable to assume that it acts as a tumor suppressor in a

similar manner.

The generation of Mnk2a mutants mutated in either the NLS

or NES allowed us to examine if Mnk2a affects p38-MAPK

cellular localization and if this effect mediates p38a activation.

When cotransfected with HA-tagged or GFP-tagged p38-

MAPK, Mnk2a colocalized with p38-MAPK both in the cyto-

plasm and nucleus (Figures 5A and S7C). However, transduction

of MKK3�/�/MKK6�/� MEFs where p38-MAPK is unphosphory-

lated (Figure 4D) with Mnk2a induced nuclear accumulation

of p38-MAPK (Figure 5B). These results suggest that physiolog-

ical expression levels of Mnk2a can induce translocation of

the endogenous p38-MAPK into the nucleus. Both Mnk2a

mutants can interact with p38a, including with its phosphory-

lated form, as was measured by coimmunoprecipitation (Fig-

ure S7E). However, onlyMnk2a and the nuclear L/Smutant could

activate both p38-MAPK target genes (Figure 6C) and induce

apoptosis (Figure 6F), whereas the cytoplasmic mutant KKR

could not activate either and even reduced apoptosis below

the empty vector levels (Figures 6C and 6F). These results sug-

gest that Mnk2a’s ability to translocate p38a into the nucleus

is both required and sufficient to mediate its tumor suppressor

activity as an inducer of p38a-MAPK target genes and

apoptosis. Mnk2b inhibited the expression of FOS and COX-2

below the basal level. Because Mnk2b does not bind p38a, we

hypothesize that it does not act in a dominant-negative manner

to Mnk2a. Rather, other downstream effects of this isoform

may affect the promoters of FOS and COX-2 to inhibit their tran-

scription. Further investigation is required to examine this

question.

WemanipulatedMKNK2 alternative splicing by antisense RNA

oligonucleotides that mask either the Mnk2a or theMnk2b splice

sites (Figure 7A). Results from these experiments suggest that

manipulation of the endogenous ratios of Mnk2a and Mnk2b

affects the oncogenic potential of cells transformed by Ras

and SRSF1. Moreover, these results suggest that MKNK2 alter-

native splicing is a critical event in both Ras- and SRSF1-induced

transformation, similar to other oncogenic alternative splicing

events regulated by SRSF1 (Ben-Hur et al., 2013).

An alternative splicing switch to eliminate a tumor suppressor

is fast and cost effective and serves as an additional level of

regulation. Several examples of such regulation of important

tumor suppressors already exist: BIN1, MDM2, Caspase-9,

BCL-2 family members, and others (Bae et al., 2000; Ge et al.,

1999; Hossini et al., 2006; Kim et al., 2008; Srinivasula et al.,

1999; Steinman et al., 2004). Many tumor suppressor genes

are deleted from the genomes of cancer cells. Such an event

would be unfavorable in the case of Mnk2 because absence of

both isoforms would result in reduced eIF4E phosphorylation,

and cells would be less oncogenic as in the case of Mnk1/

Mnk2-DKO cells (Furic et al., 2010; Ueda et al., 2010). However,

eliminating only Mnk2a and expressing Mnk2b instead would

sustain eIF4E phosphorylation without activation of the p38

stress pathway. Although we did not analyze MKNK2 DNA

copy number variation in tumors, we expect that the main mech-

anism to eliminateMnk2a in tumors will be throughmodulation of

alternative splicing rather than deletion of the MKNK2 gene.

In conclusion, we have identified amechanism inwhichMnk2a

interacts with, phosphorylates, and induces translocation of

p38-MAPK into the nucleus, and thus induces transcription of

its target genes, which results in increased apoptosis. Both

Mnk2a and Mnk2b phosphorylate eIF4E on serine 209, which

contributes to cellular transformation, but Mnk2b, which cannot

bind p38-MAPK, uncouples this phosphorylation from induction

of the p38-MAPK stress response. Our results identify Mnk2

alternative splicing as a mechanism for elimination of a tumor

suppressor (Mnk2a), which is a modulator of the p38-MAPK



stress pathway, and for generating the pro-oncogenic isoform

(Mnk2b).

EXPERIMENTAL PROCEDURES

Animal Care

All animal experiments were performed in accordance with the guidelines of

the Hebrew University committee for the use of animals for research. Veteri-

nary care was provided to all animals by the Hebrew University animal care

facility staff in accord with AAALAC standard procedures and as approved

by the Hebrew University Ethics committee.

In Vitro Kinase Assay

In vitro kinase assay was performed using recombinant Mnk2, eIF4E (Abcam),

and kinase-dead p38a (Diskin et al., 2007) as described by Askari et al. (2007)

and Avitzour et al. (2007). In brief, 300 ng of recombinantMnk2awas incubated

alone or with either 5 mg of recombinant kinase-dead p38a or 1 mg of recom-

binant eIF4E, in 20 ml reaction buffer containing 20 mM cold ATP, 0.5 mCi 32P

ATP, 30 mM MgCl2, 10 mM HEPES (pH 7.5), 50 mM EGTA, 10 mM b-glycer-

ophosphate, 5 mM NaVO4, 50 mM b-mercaptoethanol, and 0.5 mM dithio-

threitol (DTT). Reactions were shaken for 1 hr at 30�C. Reactions were stopped

by the addition of 50 ml of cold dialysis buffer (12.5 mM HEPES [pH 7.5],

100 mM KCL, 0.5 mM DTT, and 6.25% glycerol) followed by the addition of

30 ml 43 Laemmli buffer. A total of 20 ml of the final volume was separated

by SDS-PAGE, transferred to nitrocellulose by western blotting, and exposed

to Fuji phosphorimager. After the radioactive exposure, the membrane was

probed with the indicated antibodies to visualize the recombinant proteins in

the reaction.

Modulation of MKNK2 Splicing by Antisense-RNA Oligos

Cy5-labeled 20-O-methyl-modified RNA oligonucleotides were synthesized by

Sigma-Aldrich. A total of 2.5 mM of each oligo was transfected using Lipofect-

amine 2000 according to the manufacturer’s instructions. For determination of

RNA levels, BEAS-2B, MCF-10A-Ras, or MCF-10A cells were harvested 48 hr

after transfection, and levels of Mnk2a and Mnk2b were analyzed by RT-PCR.

For biological assays (colony survival, soft agar, and apoptosis), cells were

transfected as described and 24 hr after transfection were seeded for the indi-

cated assay. Oligo sequences, cell lines, plasmids, shRNA, primer sequences,

and additional experimental procedures used are described in Supplemental

Experimental Procedures.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures,

eight figures, and three tables and can be found with this article online at

http://dx.doi.org/10.1016/j.celrep.2014.03.041.

AUTHOR CONTRIBUTIONS

A.M. and R.K. designed the experiments. A.M., A.S., A.O., M.M., V.B.-H., I.S.,

E.Z., R.G.-G., and D.M. performed experiments. J.B., C.L.A., K.T., T.F.Ø.,

R.R., R.J.D., I.L.-L., and B.D. contributed reagents and technical help. A.M.

and R.K. analyzed the data and wrote the manuscript.

ACKNOWLEDGMENTS

The authors wish to thank Prof. Rony Seger for the GFP-p38a construct and

comments on the manuscript; Prof. David Engelberg for the constitutively

active p38a mutant and other p38-MAPK proteins and fruitful discussions;

Prof. Eli Pikarsky for his help in tumor pathological analysis; Prof. Oded Meyu-

has, Dr. Zahava Kluger, and Dr. Ittai Ben Porath for comments on the manu-

script; members of the Engelberg and Ben Porath labs for sharing reagents

and discussions; and members of the R.K. lab for helpful discussions. This

work was supported by the US-Israel Bi-national Science Foundation (grant

no. 2009026) and Israeli Science Foundation (grant nos. 780/08 and 1290/

12) (to R.K.).


Received: August 14, 2013

Revised: February 13, 2014

Accepted: March 13, 2014

Published: April 10, 2014

REFERENCES

Adesso, L., Calabretta, S., Barbagallo, F., Capurso, G., Pilozzi, E., Geremia, R.,

Delle Fave, G., and Sette, C. (2013). Gemcitabine triggers a pro-survival

response in pancreatic cancer cells through activation of the MNK2/eIF4E

pathway. Oncogene 32, 2848–2857.

Anczukow, O., Rosenberg, A.Z., Akerman, M., Das, S., Zhan, L., Karni, R.,

Muthuswamy, S.K., and Krainer, A.R. (2012). The splicing factor SRSF1 regu-

lates apoptosis and proliferation to promote mammary epithelial cell transfor-

mation. Nat. Struct. Mol. Biol. 19, 220–228.

Aplin, A.E., Hogan, B.P., Tomeu, J., and Juliano, R.L. (2002). Cell adhesion

differentially regulates the nucleocytoplasmic distribution of active MAP

kinases. J. Cell Sci. 115, 2781–2790.

Askari, N., Diskin, R., Avitzour, M., Capone, R., Livnah, O., and Engelberg, D.

(2007). Hyperactive variants of p38alpha induce, whereas hyperactive variants

of p38gamma suppress, activating protein 1-mediated transcription. J. Biol.

Chem. 282, 91–99.

Avitzour, M., Diskin, R., Raboy, B., Askari, N., Engelberg, D., and Livnah, O.

(2007). Intrinsically active variants of all human p38 isoforms. FEBS J. 274,

963–975.

Badger, A.M., Bradbeer, J.N., Votta, B., Lee, J.C., Adams, J.L., and Griswold,

D.E. (1996). Pharmacological profile of SB 203580, a selective inhibitor of cyto-

kine suppressive binding protein/p38 kinase, in animal models of arthritis,

bone resorption, endotoxin shock and immune function. J. Pharmacol. Exp.

Ther. 279, 1453–1461.

Bae, J., Leo, C.P., Hsu, S.Y., and Hsueh, A.J. (2000). MCL-1S, a splicing

variant of the antiapoptotic BCL-2 family member MCL-1, encodes a proapo-

ptotic protein possessing only the BH3 domain. J. Biol. Chem. 275, 25255–

25261.

Benhar, M., Dalyot, I., Engelberg, D., and Levitzki, A. (2001). Enhanced ROS

production in oncogenically transformed cells potentiates c-Jun N-terminal

kinase and p38 mitogen-activated protein kinase activation and sensitization

to genotoxic stress. Mol. Cell. Biol. 21, 6913–6926.

Ben-Hur, V., Denichenko, P., Siegfried, Z., Maimon, A., Krainer, A., Davidson,

B., and Karni, R. (2013). S6K1 alternative splicing modulates its oncogenic

activity and regulates mTORC1. Cell Rep. 3, 103–115.

Brancho, D., Tanaka, N., Jaeschke, A., Ventura, J.J., Kelkar, N., Tanaka, Y.,

Kyuuma, M., Takeshita, T., Flavell, R.A., and Davis, R.J. (2003). Mechanism

of p38 MAP kinase activation in vivo. Genes Dev. 17, 1969–1978.

Chrestensen, C.A., Shuman, J.K., Eschenroeder, A., Worthington, M., Gram,

H., and Sturgill, T.W. (2007). MNK1 and MNK2 regulation in HER2-overex-

pressing breast cancer lines. J. Biol. Chem. 282, 4243–4252.

Cuadrado, A., and Nebreda, A.R. (2010). Mechanisms and functions of p38

MAPK signalling. Biochem. J. 429, 403–417.

Das, S., Anczukow, O., Akerman, M., and Krainer, A.R. (2012). Oncogenic

splicing factor SRSF1 is a critical transcriptional target of MYC. Cell Rep. 1,

110–117.

De Nicola, G.F., Martin, E.D., Chaikuad, A., Bassi, R., Clark, J., Martino, L.,

Verma, S., Sicard, P., Tata, R., Atkinson, R.A., et al. (2013). Mechanism and

consequence of the autoactivation of p38a mitogen-activated protein kinase

promoted by TAB1. Nat. Struct. Mol. Biol. 20, 1182–1190.

Diskin, R., Askari, N., Capone, R., Engelberg, D., and Livnah, O. (2004). Active

mutants of the human p38alpha mitogen-activated protein kinase. J. Biol.

Chem. 279, 47040–47049.

Diskin, R., Lebendiker, M., Engelberg, D., and Livnah, O. (2007). Structures of

p38alpha active mutants reveal conformational changes in L16 loop that

induce autophosphorylation and activation. J. Mol. Biol. 365, 66–76.

http://dx.doi.org/10.1016/j.celrep.2014.03.041


Enslen, H., Raingeaud, J., and Davis, R.J. (1998). Selective activation of p38

mitogen-activated protein (MAP) kinase isoforms by the MAP kinase kinases

MKK3 and MKK6. J. Biol. Chem. 273, 1741–1748.

Ferreiro, I., Joaquin, M., Islam, A., Gomez-Lopez, G., Barragan, M., Lombar-

dıa, L., Domınguez, O., Pisano, D.G., Lopez-Bigas, N., Nebreda, A.R., and

Posas, F. (2010). Whole genome analysis of p38 SAPK-mediated gene expres-

sion upon stress. BMC Genomics 11, 144.

Fukunaga, R., and Hunter, T. (1997). MNK1, a new MAP kinase-activated

protein kinase, isolated by a novel expression screening method for identifying

protein kinase substrates. EMBO J. 16, 1921–1933.

Furic, L., Rong, L., Larsson, O., Koumakpayi, I.H., Yoshida, K., Brueschke, A.,

Petroulakis, E., Robichaud, N., Pollak, M., Gaboury, L.A., et al. (2010). eIF4E

phosphorylation promotes tumorigenesis and is associated with prostate

cancer progression. Proc. Natl. Acad. Sci. USA 107, 14134–14139.

Ge, K., DuHadaway, J., Du, W., Herlyn, M., Rodeck, U., and Prendergast, G.C.

(1999). Mechanism for elimination of a tumor suppressor: aberrant splicing of a

brain-specific exon causes loss of function of Bin1 in melanoma. Proc. Natl.

Acad. Sci. USA 96, 9689–9694.

Gong, X., Ming, X., Deng, P., and Jiang, Y. (2010). Mechanisms regulating the

nuclear translocation of p38 MAP kinase. J. Cell. Biochem. 110, 1420–1429.

Hazzalin, C.A., Cano, E., Cuenda, A., Barratt, M.J., Cohen, P., and Mahade-

van, L.C. (1996). p38/RK is essential for stress-induced nuclear responses:

JNK/SAPKs and c-Jun/ATF-2 phosphorylation are insufficient. Curr. Biol. 6,

1028–1031.

Hossini, A.M., Geilen, C.C., Fecker, L.F., Daniel, P.T., and Eberle, J. (2006). A

novel Bcl-x splice product, Bcl-xAK, triggers apoptosis in human melanoma

cells without BH3 domain. Oncogene 25, 2160–2169.

Hu, S.I., Katz, M., Chin, S., Qi, X., Cruz, J., Ibebunjo, C., Zhao, S., Chen, A., and

Glass, D.J. (2012). MNK2 inhibits eIF4G activation through a pathway involving

serine-arginine-rich protein kinase in skeletal muscle. Sci. Signal. 5, ra14.

Itoh, M., Nelson, C.M., Myers, C.A., and Bissell, M.J. (2007). Rap1 integrates

tissue polarity, lumen formation, and tumorigenic potential in human breast

epithelial cells. Cancer Res. 67, 4759–4766.

Karni, R., de Stanchina, E., Lowe, S.W., Sinha, R., Mu, D., and Krainer, A.R.

(2007). The gene encoding the splicing factor SF2/ASF is a proto-oncogene.

Nat. Struct. Mol. Biol. 14, 185–193.

Kim, E., Goren, A., and Ast, G. (2008). Insights into the connection between

cancer and alternative splicing. Trends Genet. 24, 7–10.

Knauf, U., Tschopp, C., and Gram, H. (2001). Negative regulation of protein

translation by mitogen-activated protein kinase-interacting kinases 1 and 2.

Mol. Cell. Biol. 21, 5500–5511.

Lee, J.C., Kumar, S., Griswold, D.E., Underwood, D.C., Votta, B.J., and

Adams, J.L. (2000). Inhibition of p38 MAP kinase as a therapeutic strategy.

Immunopharmacology 47, 185–201.

MacNeil, A.J., Jiao, S.C., McEachern, L.A., Yang, Y.J., Dennis, A., Yu, H., Xu,

Z., Marshall, J.S., and Lin, T.J. (2014). MAPK kinase 3 is a tumor suppressor

with reduced copy number in breast cancer. Cancer Res. 74, 162–172.

Mamane, Y., Petroulakis, E., Rong, L., Yoshida, K., Ler, L.W., and Sonenberg,

N. (2004). eIF4E—from translation to transformation. Oncogene 23, 3172–

3179.

Parra, J.L., Buxade, M., and Proud, C.G. (2005). Features of the catalytic

domains and C termini of the MAPK signal-integrating kinases Mnk1 and

Mnk2 determine their differing activities and regulatory properties. J. Biol.

Chem. 280, 37623–37633.

Parra-Palau, J.L., Scheper, G.C., Wilson, M.L., and Proud, C.G. (2003).

Features in the N and C termini of the MAPK-interacting kinase Mnk1 mediate

its nucleocytoplasmic shuttling. J. Biol. Chem. 278, 44197–44204.

Pfundt, R., van Vlijmen-Willems, I., Bergers, M., Wingens, M., Cloin, W., and

Schalkwijk, J. (2001). In situ demonstration of phosphorylated c-jun and p38

MAP kinase in epidermal keratinocytes following ultraviolet B irradiation of

human skin. J. Pathol. 193, 248–255.

Plotnikov, A., Zehorai, E., Procaccia, S., and Seger, R. (2011). The MAPK

cascades: signaling components, nuclear roles and mechanisms of nuclear

translocation. Biochim. Biophys. Acta 1813, 1619–1633.

Rincon, M., and Davis, R.J. (2009). Regulation of the immune response by

stress-activated protein kinases. Immunol. Rev. 228, 212–224.

Sakurai, T., Kudo, M., Umemura, A., He, G., Elsharkawy, A.M., Seki, E., and

Karin, M. (2013). p38a inhibits liver fibrogenesis and consequent hepatocarci-

nogenesis by curtailing accumulation of reactive oxygen species. Cancer Res.

73, 215–224.

Scheper, G.C., Parra, J.L., Wilson, M., Van Kollenburg, B., Vertegaal, A.C.,

Han, Z.G., and Proud, C.G. (2003). The N and C termini of the splice variants

of the humanmitogen-activated protein kinase-interacting kinaseMnk2 deter-

mine activity and localization. Mol. Cell. Biol. 23, 5692–5705.

Shenberger, J.S., Zhang, L., Hughlock, M.K., Ueda, T., Watanabe-Fukunaga,

R., and Fukunaga, R. (2007). Roles of mitogen-activated protein kinase signal-

integrating kinases 1 and 2 in oxidant-mediated eIF4E phosphorylation. Int. J.

Biochem. Cell Biol. 39, 1828–1842.

Shimoni-Sebag,A., Lebenthal-Loinger, I.,Zender, L., andKarni,R. (2013).RRM1

domain of the splicing oncoprotein SRSF1 is required for MEK1-MAPK-ERK

activation and cellular transformation. Carcinogenesis 34, 2498–2504.

Srebrow, A., and Kornblihtt, A.R. (2006). The connection between splicing and

cancer. J. Cell Sci. 119, 2635–2641.

Srinivasula, S.M., Ahmad, M., Guo, Y., Zhan, Y., Lazebnik, Y., Fernandes-

Alnemri, T., and Alnemri, E.S. (1999). Identification of an endogenous domi-

nant-negative short isoform of caspase-9 that can regulate apoptosis. Cancer

Res. 59, 999–1002.

Steinman,H.A., Burstein, E., Lengner, C., Gosselin, J., Pihan,G., Duckett, C.S.,

and Jones, S.N. (2004). An alternative splice form of Mdm2 induces p53-inde-

pendent cell growth and tumorigenesis. J. Biol. Chem. 279, 4877–4886.

Turjanski, A.G., Vaque, J.P., and Gutkind, J.S. (2007). MAP kinases and the

control of nuclear events. Oncogene 26, 3240–3253.

Ueda, T., Watanabe-Fukunaga, R., Fukuyama, H., Nagata, S., and Fukunaga,

R. (2004). Mnk2 and Mnk1 are essential for constitutive and inducible phos-

phorylation of eukaryotic initiation factor 4E but not for cell growth or develop-

ment. Mol. Cell. Biol. 24, 6539–6549.

Ueda, T., Sasaki, M., Elia, A.J., Chio, I.I., Hamada, K., Fukunaga, R., and Mak,

T.W. (2010). Combined deficiency for MAP kinase-interacting kinase 1 and 2

(Mnk1 and Mnk2) delays tumor development. Proc. Natl. Acad. Sciences

USA 107, 13984–13990.

Venables, J.P. (2004). Aberrant and alternative splicing in cancer. Cancer Res.

64, 7647–7654.

Venables, J.P., Klinck, R., Koh, C., Gervais-Bird, J., Bramard, A., Inkel, L.,

Durand, M., Couture, S., Froehlich, U., Lapointe, E., et al. (2009). Cancer-asso-

ciated regulation of alternative splicing. Nat. Struct. Mol. Biol. 16, 670–676.

Wakeman, D., Schneider, J.E., Liu, J., Wandu, W.S., Erwin, C.R., Guo, J.,

Stappenbeck, T.S., and Warner, B.W. (2012). Deletion of p38-alpha

mitogen-activated protein kinase within the intestinal epithelium promotes

colon tumorigenesis. Surgery 152, 286–293.

Waskiewicz, A.J., Flynn, A., Proud, C.G., and Cooper, J.A. (1997). Mitogen-

activated protein kinases activate the serine/threonine kinases Mnk1 and

Mnk2. EMBO J. 16, 1909–1920.

Wendel, H.G., Silva, R.L., Malina, A., Mills, J.R., Zhu, H., Ueda, T., Watanabe-

Fukunaga, R., Fukunaga, R., Teruya-Feldstein, J., Pelletier, J., and Lowe, S.W.

(2007). Dissecting eIF4E action in tumorigenesis. Genes Dev. 21, 3232–3237.

Worch, J., Tickenbrock, L., Schwable, J., Steffen, B., Cauvet, T., Mlody, B.,

Buerger, H., Koeffler, H.P., Berdel, W.E., Serve, H., and Muller-Tidow, C.

(2004). The serine-threonine kinase MNK1 is post-translationally stabilized

by PML-RARalpha and regulates differentiation of hematopoietic cells. Onco-

gene 23, 9162–9172.

Xi, L., Feber, A., Gupta, V., Wu, M., Bergemann, A.D., Landreneau, R.J., Litle,

V.R., Pennathur, A., Luketich, J.D., and Godfrey, T.E. (2008). Whole genome

exon arrays identify differential expression of alternatively spliced, cancer-

related genes in lung cancer. Nucleic Acids Res. 36, 6535–6547.


Cell Reports, Volume 7

Supplemental Information

Mnk2 Alternative Splicing Modulates

the p38-MAPK Pathway and Impacts

Ras-Induced Transformation

Avraham Maimon, Maxim Mogilevsky, Asaf Shilo, Regina Golan-Gerstl, Akram Obiedat,

Vered Ben-Hur, Ilana Lebenthal-Loinger, Ilan Stein, Reuven Reich, Jonah Beenstock,

Eldar Zehorai, Claus L. Andersen, Kasper Thorsen, Torben F. Ørntoft, Roger J. Davis,

Ben Davidson, David Mu, and Rotem Karni

Maimon et. al.

Fig. S1

A

Mnk2a Mnk2b

No

rma

lize

d e

xp

res

sio

n

0

0.5

1

1.5

2

2.5

0

0.2

0.4

0.6

0.8

1

1.2

Mnk2a Mnk2b

0

2

4

6

8

10

No

rma

lize

d

ex

pre

ss

ion

SRSF1 C D

W.T.

Ras

Mut-K-Ras

GAPDH

Mnk2a Mnk2b

MKNK2

B

No

rma

lize

d

ex

pre

ss

ion

E

SRSF1 F

0

20

40

60

80

No

rma

lize

d

ex

pre

ssio

n

RT

-PC

R Mnk2a

Mnk2b

GAPDH

G SRSF1

β-Actin

Mnk2a

Mnk2b

Weste

rn B

lot

GAPDH

Mnk2a

Mnk2b

H

BEAS-2B cells

MKNK2

0

5000

10000

15000

20000

0

5000

10000

15000

0

5000

10000

15000

20000

0

2000

4000

6000

8000

0

5000

10000

15000

0

5000

10000

15000

20000

0

2000

4000

6000

0

1000

2000

3000

0

2000

4000

6000

8000

0

2000

4000

6000

8000

0

5000

10000

15000

20000

25000

0

10000

20000

30000

40000

SRSF1 SRSF2 SRSF3 SRSF4

SRSF5 (not significant) SRSF6 (not significant) SRSF7 SRSF8

SRSF9 TRA2β hnRNPA1

Maimon et. al.

Fig. S1

p = 0.04 p = 0.0001 p = 1.42E-08 p = 1.42E-08

p = 0.2 p = 0.84 p = 2.92E-08 p = 0.0001

p = 5.42E-09 p = 0.04 p = 3.3E-12 p = 2.2E-05

BRCA

No

rma

lize

d e

xp

res

sio

n

No

rma

lize

d e

xp

res

sio

n

No

rma

lize

d e

xp

res

sio

n

I

hnRNPA2B1

0

2000

4000

6000

8000

10000

0

1000

2000

3000

4000

5000

6000

7000

0

2000

4000

6000

8000

10000

12000

0

2000

4000

6000

8000

10000

0

500

1000

1500

2000

2500

0

1000

2000

3000

4000

5000

6000

7000

0

5000

10000

15000

20000

0

10000

20000

30000

40000

SRSF1 SRSF2 SRSF5

SRSF6 SRSF8 SRSF9

hnRNPA1 hnRNPA2B1

Normal Tumor Normal Tumor Normal Tumor

Normal Tumor Normal Tumor Normal Tumor

Normal Tumor Normal Tumor

LUAD

0

200

400

600

800

1000

0

100

200

300

Mn

k2

a e

xp

res

sio

n

Mn

k2

b e

xp

res

sio

n p = 0.005 p = 8.8E-23

A B

0

2

4

6

8

10 p = 0.03

Mn

k2

a/M

nk

2b

C

D p = 3.8E-36 p = 1.8E-14 p = 2.7E-07

p = 0.008 p = 9.48E-06 p = 3.6E-39

p = 5E-13 p = 3.9E-25

No

rma

lize

d e

xp

res

sio

n

No

rma

lize

d e

xp

res

sio

n

No

rma

lize

d e

xp

res

sio

n

Maimon et. al.

Fig. S2

Related to Fig. 1

COAD

0

2000

4000

6000

8000

0

2000

4000

6000

8000

10000

0

1000

2000

3000

4000

0

2000

4000

6000

8000

10000

12000

14000

0

2000

4000

6000

8000

10000

12000

0

2000

4000

6000

8000

0

2000

4000

6000

8000

10000

0

1000

2000

3000

4000

5000

6000

0

10000

20000

30000

40000

0

5000

10000

15000

20000

25000

30000

35000

SRSF1 SRSF2

SRSF4


Normal Tumor

SRSF5 SRSF6

SRSF7


Normal Tumor

SRSF9 TRA2β

hnRNPA1 hnRNPA2B1



0

5

10

15

Mn

k2

a/M

nk

2b

p = 2.5E-05

E F

Normal Tumor

p = 0.02 p = 7.6E-11

p = 7.6E-06 p = 1.5E-06 p = 0.00041

p = 3.5E-10 p = 1.99E-12 p = 0.00049

p = 9.3E-66 p = 0.00046

No

rma

lize

d e

xp

res

sio

n

No

rma

lize

d e

xp

res

sio

n

No

rma

lize

d e

xp

res

sio

n

Maimon et. al.

Fig. S2

Related to Fig. 1

No

rma

lize

d e

xp

res

sio

n

0 100 200 300 400 500 600 700 800

Co

lon

ies

\we

ll

0

1000

2000

3000

4000

5000

A B

C D

E

SRp55

p-eIF4E

eIF4E

T7 Mnk2a T7 Mnk2b

β-Catenin

p-eIF4E

T7 Mnk2a T7 Mnk2b

0

1000

2000

3000

4000

5000

6000

7000

Co

lon

ies\w

ell

C

olo

nie

s\w

ell

Maimon et al.

Figure S3

0

200

400

600

800

β-catenin

p-p38α

p38α

Mnk2a Mnk2b

p-eIF4E

eIF4E

0

0.2

0.4

0.6

0.8

1

1.2

1.4

No

rma

lize

d e

xp

res

sio

n

Mnk2a Mnk2b

Co

lon

ies

\we

ll

F

H G

pBABE

Mnk2a

Mnk2b

Mnk2aKD

I

K L

G1 = 61.58%

S = 5.07%

G2/M = 23.62%

G1 = 59.03%

S = 6.4%

G2/M = 25.4%

G1 = 60.23%

S = 6.69%

G2/M = 27.68%

G1 = 60.64%

S = 7.36%

G2/M = 26.27%

0 5

10 15 20 25 30 35

1 2 3 4

Re

lati

ve

A6

50

Time (days)

M N

x

pBABE

Mnk2a

Mnk2b Mnk2aKD

J

0 5

10 15 20 25 30 35 40

1 2 3 4

MLP (-)

shMnk2a-1

shMnk2a-2

Time (days)

Rela

tive

A6

50

MLP Sh2a-1 Sh2a-2

Maimon et al.

Figure S3

0

5

10

15

20

25

30

35

% C

ell

Dea

th

MCF-10A + Ras

0

10

20

30

40

50

60

% S

urv

iva

l

+ SB203580

+ Vehicle

A B

C D

0

20

40

60

80

100

0

10

20

30

40

50

60

70

% C

ell

dea

th

% C

ell

dea

th

E

Nu

mb

er

of

co

lon

ies

/we

ll

0

10

20

30

40

50

60

*

*

*

* **

Maimon et al.

Figure S4

Related to Fig. 3

0

2

4

6

8

10

12

GFP +

% o

f ga

ted

(G

FP c

ells

)

0

2

4

6

8

10

12

24 48 Time (hours)

72

pBABE

Mnk2a Mnk2b

% C

ell

de

ath

MCF-10A

pCDNA3

Mnk2a

Mnk2b

9.13% 6.21%

6.7% 5.29%

8.36% 6.65%

MCF-10A

F

G

Maimon et al.

Figure S4

Related to Fig. 3

pWZL

p38α

p38α +

SB203580

pBABE Mnk2a Mnk2b Mnk2aKD

B

A

C

D

No

rma

lize

d

ce

ll d

ea

th

*

*

MLP sh2a-1

pWZL(-)

p38α

p38α + SB

sh2a-2

Vehicle +SB203580

Vehicle +SB203580

0

0.5

1

1.5

2

2.5

3

0

0.2

0.4

0.6

0.8

1

1.2

No

rma

lize

d

ce

ll d

ea

th

Maimon et al.

Figure S5

Related to Fig. 3

β-catenin

p-MK-2

MK-2

β-catenin

p-MK-2

MK-2

An

iso

mycin

A

nis

om

ycin

+

SB

203580

p-p38

p38

p-eIF4E

eIF4E

E

A B

p-p38α

p38α

1±

0.1

2.8±

0.3

0.4±

0.2

1.2±

0.4

MCF-10A-Ras

MCF-10A

p-p38α

p38α

p-eIF4E

eIF4E

5 10 CGP (µM):

C

p-eIF4E

eIF4E

p-p38α

p38α

GAPDH

E

I

0

0.5

1

1.5

2

0

0.5

1

1.5

p-eIF4E/eIF4E p-p38α/p38α

No

rma

lize

d

ph

osp

ho

ryla

tio

n

No

rma

lize

d

ph

osp

ho

ryla

tio

n

T7-Mnk2a C N

p38α

Myc

Caspase-2

C N C N C N

p-p38α

T7-Mnk2b

T7-Mnk2a: + - + - + - + -

Mnk2a

p-p38

HA

β-catenin

α β,δ

γ

α β,δ

γ

D

Mnk2a

β-catenin

HA-p38

T7

HA IB

Input

Mnk2a

HA-p38

- + - + - +

IP:T7Mnk2a

T7-Mnk2a:

IB

- + - + - +

β,δ γ

F

H

T7-Mnk2a:

G

T7

HA

Maimon et al.

Figure S6

Related to Fig. 5

A B M

nk

2a

M

nk

2a

L/S

M

nk

2a

KK

R

DAPI p38α T7 Mnk2 Merged p

BA

BE

MCF10A

0

20

40

60

80

100

120

Cytoplasmic staining Nuclear staining

% o

f p

38

α

po

sit

ive

ce

lls

*

MCF-10A

Mn

k2

a

Mn

k2

a L

/S

Mn

k2

a K

KR

DAPI HA-p38α T7 Mnk2 Merged

pC

DN

A3

p

CD

NA

3

HA-p38α

0

50

100

150

* * *

Cytoplasmic staining Nuclear staining

% o

f H

A-p

38

α

po

sit

ive

ce

lls

C D

HeLa

HeLa

Maimon et al.

Figure S7

0

10

20

30

40

50

0 10 20 30 40 50

% staining of Cy3 (T7-Mnk2a)

in cytoplasm

% s

tain

ing

of

FIT

C

(GF

P-p

38α

) in

cyto

pla

sm

r=0.358

p = 0.0376

F

Maimon et al.

Figure S7

T7

p-p38α

p38α

Input

HA-p-p38α p-p38α

T7-Mnk2

HA-p38α p38α

IB

HA-p-p38α p-p38α

T7-Mnk2

HA-p38α p38α

IP:T7-Mnk2

T7

p-p38α

p38α

IB

E

Mnk2a Mnk2b

β-actin

Scr

2a

block

2b

block

40 400 1000

Scr 2a Block 2b Block

40

400

0

20

40

60

80

100

120

140

Co

lon

ies

/ W

ell

C

olo

nie

s / W

ell

0

1000

2000

3000

4000 Scr 2a Block 2b Block

A B

C D

E F

1±

0.2

0.6±

0.1

2±

0.6

Maimon et al.

Fig. S8

Fold Mnk2a:

Scr 2a Block 2b Block

0

20

40

60

80

100

0

20

40

60

80

100

0

20

40

60

80

100

% S

urv

iva

l

% S

urv

iva

l

% S

urv

iva

l

24h 48h 96h

G

H

Ras MCF10A

Normal conditions

Maimon et al.

Fig. S8

Tables S1-S3

Related to Fig. 1

Table S1: correlation between splicing factors expression and Mnk2a/Mnk2b ratios in breast tumors

Splicing factor Correlation (R value) P value Number of tumor samples

SRSF1 -0.152 4.87E-06 888

SRSF2 -0.171 2.9E-07 888

SRSF3 -0.298 < 1E-08 888

SRSF4 -0.255 < 1E-08 888

SRSF5 0.113 0.00066 888

SRSF6 <0.1 >0.05 non significant 888

SRSF7 -0.217 < 1E-08 888

SRSF9 -0.237 < 1E-08 888

TRA2β -0.167 5.1E-07 888

hnRNP A1 -0.28 < 1E-08 888

hnRNP A2/B1 -0.14 2.6E-05 888

Table S2: correlation between splicing factors expression and Mnk2a/Mnk2b ratios in lung tumors


SRSF1 -0.137 0.002218 490

SRSF2 -0.159 0.000407 490

SRSF3 -0.236 1.1E-07 490

SRSF4 0.05 >0.05 non significant 490

SRSF5 0.164 0.000261 490

SRSF6 -0.02 >0.05 non significant 490

SRSF7 -0.22 4.8E-07 490

SRSF9 -0.35 < 1E-08 490

TRA2β -0.162 09000000 490

hnRNP A1 -0.148 0.001001 490

hnRNP A2/B1 -0.19 1.38E-05 490

Table S3: correlation between splicing factors expression and Mnk2a/Mnk2b ratios in colon tumors





SRSF4 -0.132 0.033 192



SRSF7 -0.17 0.01461 192

SRSF9 -0.247 0.00055 192

TRA2β -0.0458 >0.05 non significant 192

hnRNP A1 0.052426 >0.05 non significant 192

hnRNP A2/B1 -0.05888 >0.05 non significant 192

Supplementary Figure Legends and Experimental procedures

Figure S1. A switch in MKNK2 alternative splicing in breast, colon and lung

tumors.

A. RNA from the indicated immortal breast cell lines and primary breast tumors was

extracted and the levels of Mnk2a and Mnk2b were detected by Q-RT-PCR. B. RNA

from MCF-10A cells transduced with retroviruses encoding either pWZL empty or

pWZL-H-RasV12

was extracted and the levels of Mnk2a and Mnk2b were detected by

Q-RT-PCR. GAPDH was used for normalization. C. RNA from cells described in (B)

was extracted and the levels of SRSF1 mRNA were analyzed by Q-RT-PCR. D. RNA

from pancreatic cancer cell lines harboring W.T. K-Ras (ASPC-1, BXPC-3) or mutant

K-Ras (Patu8902, Patu8988T and Panc-1) was extracted and RT-PCR was performed

to detect the levels of Mnk2a and Mnk2b transcripts. GAPDH mRNA was used as a

control. E. RNA from pancreatic cancer cell lines BXPC3 or Panc-1 was extracted

and the levels of Mnk2a and Mnk2b were detected by RT-PCR. GAPDH was used as

a control. F. RNA from cells described in (E) was extracted and the levels of SRSF1

were analyzed by Q-RT-PCR. G. Cells described in (E) were lysed and after Western

blotting membranes were probed with the indicated antibodies. β-actin was used as a

loading control. H. RNA from the indicated immortal lung bronchial epithelial cells

(BEAS-2B) transduced either with pBABE or SRSF1 was extracted and RT-PCR was

performed as described in (E). GAPDH mRNA was used as a control. I. Box plot

representation of SRSF1-9, TRA2β, hnRNPA1 and hnRNPA2B1 expression levels in

breast tumors with Mnk2a/Mnk2b ratios ≤ 1 (n=104) or with Mnk2a/Mnk2b ratios 1

(n=890). Top and bottom box edges represent the third and first quartile. Whiskers

indicate 90 and 10 percentiles, asterisks represent minimum and maximum points. p

value is based on T-test two tail analysis.

Figure S2. A switch in MKNK2 alternative splicing in colon and lung tumors.

A-C. Box plot representation RNA levels of Mnk2a (A), Mnk2b (B) and

Mnk2a/Mnk2b ratio (C) in lung normal (n=58) and tumor (n=490) tissues. Top and

bottom box edges represent the third and first quartile. Whiskers indicate 90 and 10

percentiles, asterisks represent minimum and maximum points. p value is based on T-

test two tail analysis. D. Box plot representation of SRSF1-9, TRA2β, hnRNPA1 and

hnRNPA2B1 RNA levels in lung tumors (n=490) compared to normal tissue (n=58).

Top and bottom box edges represent the third and first quartile. Whiskers indicate 90

and 10 percentiles, asterisks represent minimum and maximum points. p values are

based on T-test two tail analysis. E. Box plot representation of Mnk2a/Mnk2b RNA

ratio in colon normal (n=23) and tumor (n=192) tissues. Top and bottom box edges

represent the third and first quartile. Whiskers indicate 90 and 10 percentiles, asterisks

represent minimum and maximum points. p values are based on T-test two tail

analysis. F. Box plot representation of SRSF1-9 TRA2β, hnRNPA1 and hnRNPA2B1

RNA levels in colon tumors (n=192) compared to normal tissue (n=23). Top and

bottom box edges represent the third and first quartile. Whiskers indicate 90 and 10

percentiles, asterisks represent minimum and maximum points. p values are based on

T-test two tail analysis. Analysis based on RNA-seq data from The Cancer Genome

Atlas (TCGA) project (https://tcga-data.nci.nih.gov/tcga/).

Figure S3. Both Mnk2 isoforms phosphorylate eIF4E, yet Mnk2a suppresses

whereas Mnk2b promotes tumorigenesis in vitro.

A. NIH 3T3 cells were transduced with the indicated retroviruses encoding either for

Mnk2 isoforms or a kinase-dead version of Mnk2a (Mnk2aKD) and after selection for

puromycin resistance were lysed and subjected to Western blot analysis with the

indicated antibodies. B. Cells described in (A) were seeded into soft agar in duplicate

and colonies were allowed to grow for 14 days. Colonies in ten fields of each well

were counted and the mean ± standard deviation of colonies per well in a

representative experiment is shown. C. U2OS osteosarcoma cells were transduced

with the indicated retroviruses encoding either for Mnk2 isoforms or a kinase-dead

version of Mnk2a (Mnk2aKD) and after selection were lysed and subjected to

Western blot analysis with the indicated antibodies. D. Cells described in (C) were

analyzed for growth in soft agar as described in (B). E. Pools of MCF-10A cells

transduced with retroviruses encoding either empty vector (MLP) or the indicated

shRNAs against Mnk2a were lysed and lysates were subjected to Western blot

analysis with the indicated antibodies. F. Cells described in (E) were seeded into soft

agar and 20 days later colonies were counted n=3. G. RNA was extracted from cells


described in (E) and the levels of Mnk2a and Mnk2b were measured by Q-RT-PCR.

H. NCI-H460 lung carcinoma cells transduced with the indicated retroviruses

encoding for isoform-specific shRNAs against Mnk2a were seeded and analyzed for

growth in soft agar as described in (B). I-J. MCF-10A cells transduced with

retroviruses encoding either for Mnk2 isoforms (I) or isoform-specific shRNAs

against Mnk2a (J) were seeded in sixplicates in 96 well plates and growth curves were

measured as described in Experimental procedures (n=6). K-N. Cells described in (I)

were stained with PI and subjected to flow cytometry analysis for DNA content

assessment. Percent of cells gated in each phase is indicated.

Figure S4. Mnk2a sensitizes cells to stress-induced apoptosis.

A. MCF-10A cells transduced with the indicated retroviruses encoding either for

Mnk2 isoforms or a kinase-dead version of Mnk2a (Mnk2aKD) were transduced with

H-RASV12

. After selection with hygromycin transductants were maintained in growth

factor-free medium in suspension for 48 h. Cell death was measured using the trypan

blue exclusion assay. B. Human pancreatic cancer cells Panc-1 were seeded on 96-

well plates (6000 cells/well). Following starvation cells were treated with 0.5 μM

anisomycin in the presence or absence of 20 μM SB203580 and % survival was

determined after 24 hours (* p ≤ 0.01). C. MCF-10A cells were transduced as in (A).

After selection cells cells were serum starved in a growth factor-free medium for 24h,

treated with 0.5 M Sorbitol (to induce osmotic shock) and 24 h later were stained with

trypan blue (n=2). D. MCF-10A cells transduced with the indicated retroviruses

encoding for shRNAs against Mnk2a were seeded as in (C). Following starvation

cells were treated with 0.25 M Sorbitol and 24 h later were fixed and stained with

trypan blue (n=2) (*p ≤ 0.01, ** p ≤ 0.05). E. Pools of MCF-10A cells co-transduced

with the indicated retroviruses and oncogenic Ras were plated at a low density (400

cells/well). 10 days post selection with hygromycin cells were assessed for colony

formation with methylene blue staining. n=3. (* p ≤ 0.01). F. MCF-10A cells were

co-transfected with either empty vector (pCDNA3) or Mnk2 isoforms together with

pCDNA3-GFP. 48h after transfection cells were subjected to flow cytometry analysis

to detect GFP-positive cells. Percent of cells gated is indicated in each dot plot. (n=2).

Right; representative graph showing the % of GFP gated cells from 2 experiments

described in (F) ± standard deviation. G. MCF-10A cells transduced with the

indicated retroviruses encoding for Mnk2 isoforms. After selection with puromycin

transductants cells were maintained in normal growth factor medium for the indicated

times. Cell death was measured using the trypan blue exclusion assay.

Figure S5. Mnk2a but not Mnk2b enhances p38α-mediated cell death.

A. MCF-10A cells transduced with the indicated Mnk2 isoforms and a kinase-dead

version of Mnk2a (Mnk2aKD) were subsequently transduced with either a

constitutively active form of p38α or empty vector (pWZL), and selected with

hygromycin in the presence or absence of 20 µM SB203580 for 48 hours.

Morphology of the cells was analyzed using a light microscope. B. Cells treated as in

(A) were stained with methylene blue as described in Experimental procedures. % cell

death in each column was normalized to that of the empty vector (pBABE). n=4 (* p

≤ 0.05). C-D. MCF-10A cells transduced with the indicated viruses expressing

isoform-specific shRNAs against Mnk2a were transduced with a constitutively active

form of p38α or an empty vector and analyzed as described in (A) and (B),

respectively. Cell death in (B, D) was calculated by normalizing the absorbance

values for cells co-transduced pWZL-active p38 mutant and Mnk2 isoforms to that of

cells co-transduced pWZL and Mnk2 isoforms. n=4, (* p ≤ 0.05). E. Western blot

analysis of MCF-10A cells transduced with retroviruses containing either the

indicated Mnk2 isoforms or a kinase dead version of Mnk2a (Mnk2aKD) or isoform

specific shRNAs against Mnk2a treated with 1 µM anisomycin, in the presence or

absence of 10 µM SB203580.

Figure S6. Phosphorylation of p38-MAPK and eIF4E is regulated by Mnk2a

kinase activity and modulates p38α localization.

A. MCF-10A cells were transduced by H-RASV12

and after selection cells were lysed

and subjected to Western blot analysis with the indicated antibodies. Numbers

representing ratio of p-p38/total p38 were calculated normalizing to the level in MCF-

10A cells set as one (n=2). B. MCF-10A cells were seeded in 6-well plates and treated

either with DMSO (veh) or with the Mnk1/2 kinase inhibitor CGP 57380 (5 or 10

M) for 4h. Total cell lysates were subjected to Western blot analysis with the

indicated antibodies. C. Cell lines from breast immortal (MCF-10A, HMLE) or cancer

(MCF-7, MDA-MB-231, SUM-149, SUM-159, MDA-MB-468) cell lines were seeded in

6-well plates and 24 h later were lysed and subjected to Western blot analysis with the

indicated antibodies. D,E. Quantitation of p38 (D) and eIF4E (E) phosphorylation levels.

The level of phosphorylated p38 and eIF4E was normalized to that of their total level.

Results are presented as an average ± SD (n=2). F. HeLa cells were co-transfected with

empty vector (pCDNA3) or Mnk2a with either HA-p38β, or HA-p38γ, or HA-p38δ.

Levels of phospho- and total p38-MAPK as well as Mnk2a and p38-MAPK isoforms

(alpha, beta, gamma and delta) were measured by Western blotting. G-H. HeLa cells

were co-transfected as in (F) with either HA-p38β or HA-p38γ or HA-p38δ together

with either pCDNA3 or T7-Mnk2a. T7-Mnk2a was co-immunoprecipitated from

lysates with anti-T7 antibody. Input (G) and immunoprecipitated (H) proteins were

subjected to Western blot analysis with the indicated antibodies. Lysates from cells

transfected with T7-Mnk2a immunoprecipitated without T7 antibody was used as

control. I. Cellular distribution, cytoplasmic (C) and nuclear (N) fractions, of p38α, p-

p38α and T7-Mnk2 isoforms in MCF-10A cells transduced with Mnk2 isoforms or a

kinase-dead version of Mnk2a (Mnk2aKD). C-myc (nuclear) and caspase-2 (cytoplasmic)

served as fractionation controls.

Figure S7. W.T and mutants of Mnk2a colocalize, interact with and can induce

translocation of p38α into the nucleus.

A. MCF-10A cells transduced with the indicated retroviruses were seeded, fixed and

subjected to immunofluorescence to detect endogenous p38α. (T7-tag was stained red

and p38α was stained green). B. Quantification of cytoplasmic/nuclear distribution of

endogenous p38α in cells similar to those described in (A). n=100 for empty vector

and Mnk2a. (*p ≤ 0.01). C. HeLa cells co-transfected with HA-tagged p38α and

either empty vector pCDNA3 or pCDNA3 encoding for either T7-Mnk2a, Mnk2aL/S

or Mnk2aKKR. 24 h posttransfection cells were fixed with 4% paraformaldehyde and

subjected to immunofluorescence, as described in Experimental procedures (Mnk2

was stained green and HA-tag was stained red). D. Quantification of

cytoplasmic/nuclear distribution of HA-p38α in cells similar to those described in (C).

n=35 cells for each mutant. (*p ≤ 0.01, ** p ≤ 0.05). E. HeLa cells were co-

transfected with HA-tagged p38α and pCDNA3 vectors encoding for either T7-

Mnk2a, Mnk2aL/S or Mnk2aKKR. Cells were lysed 48 h posttransfection and the

Mnk2a mutants and p38α were co-immunoprecipitated as described in the

Experimental procedures. Input and immunoprecipitated proteins were subjected to

Western blot analysis with the indicated antibodies. F. Correlation between T7-

Mnk2a localization (Cy3) and GFP-p38α localization (FITC) in 40 HeLa cells co-

transfected with T7-Mnk2a and GFP-p38α as described in Fig. 6A (Pearson

correlation = 0.358, p = 0.0376).

Figure S8. Modulation of MKNK2 splicing by splice-site competitive antisense

RNA oligos, affects survival and cellular transformation.

A. BEAS-2B cells transformed with SRSF1 were transfected with the indicated Cy5-

labeled antisense 2’-O-methyl RNA oligos. 48 hours later RNA was extracted and the

levels of Mnk2a and Mnk2b were measured by RT-PCR. Numbers represent ratio of

Mnk2a/Mnk2b normalized to that of the scrambled oligo (arbitrarily set at 1) ± SD

(n=3). B. Representative photographs of cells described in (A), cells were visualized

under fluorescent microscope 24h after oligo transfection. C. Cells transfected as in

(A) were seeded into soft agar (see Experimental procedures) in triplicates and

colonies were allowed to grow for 14 days. Colonies in ten fields of each well were

counted and the mean and standard deviation of colonies per well of 3 wells is shown

(n=3). D. Photographs of representative fields of colonies in soft agar obtained as

described in (C). E. Cells transfected as in (A) were plated at low density (40 and 400

cells/well) and assessed for colony formation with methylene blue staining 20 days

later. Average and standard deviation of the number of colonies per well is shown.

n=3. F. Representative wells with colonies described in (E). G. Ras transformed

MCF-10A cells were transfected with the indicated Cy5-labeled antisense 2’-O-

methyl RNA oligos. Following 24 hours cells were visualized under fluorescent

microscope and representative photographs were taken. H. Cells transfectd with the

indicated Cy5-labeled antisense 2’-O-methyl RNA oligos described in (G) were

grown for the indicated times. Cell viability was measured by trypan blue exclusion

using a BioRad cell counter.

Table S1: Correlation analysis between Mnk2a/Mnk2b ratio and splicing factor

levels in breast cancer

Correlation analysis between Mnk2a/Mnk2b ratio and SRSF1-9, TRA2β, hnRNPA1

and hnRNPA2B1 RNA levels in breast cancer samples (n=890). 94 samples with the

highest Mnk2a/Mnk2b ratios and 10 samples with the lowest values were excluded

from the analysis. Pearson correlation value (r) and p-value are shown for each

correlation. Analysis based on RNA-seq data from The Cancer Genome Atlas

(TCGA) project (https://tcga-data.nci.nih.gov/tcga/).


levels in lung cancer


and hnRNPA2B1 RNA levels in lung cancer samples (n=490). Pearson correlation

value (r) and p-value are shown for each correlation. Analysis based on RNA-seq data

from The Cancer Genome Atlas (TCGA) project (https://tcga-data.nci.nih.gov/tcga/).


levels in colon cancer


and hnRNPA2B1 RNA levels in colon cancer samples (n=192). Pearson correlation

value (r) and p-value are shown for each correlation. Analysis based on RNA-seq data

from The Cancer Genome Atlas (TCGA) project (https://tcga-data.nci.nih.gov/tcga/).

SI Experimental procedures

Plasmids

Mnk2a and Mnk2b cDNAs were amplified by RT-PCR from HeLa RNA extracts

using primers coding for an N-terminal T7 tag and subcloned into the EcoRI site of

the pCDNA3.1 and pBABE plasmids. Mnk2a-kinase dead (Mnk2aKD), Mnk2aL/S

and Mnk2aKKR were generated by site directed mutagenesis. Mnk2aKD: lysine 113

(K113) was replaced by alanine. Mnk2aL/S: leucines 281/285 were replaced by

serines. Mnk2aKKR: lysines 60/61 and arginine 62 were replaced by alanines.

Mutagenesis primers are described in Table S1. pWZL-HA-p38α(D176A+F327S)

was generated by sub-cloning of HA-p38α(D176A+F327S) from pCDNA3.1 (Askari

et al., 2007; Avitzour et al., 2007; Diskin et al., 2004) into the EcoRI site of pWZL-

hygro. shMnk2a-1 and shMnk2a-2 were constructed in the MLP vector (Dickins et

al., 2005). shRNA sequences are shown in Table S1.

Cells




HEK293, MCF-7, MDA-MB-231, Panc-1, MEF MKK3/6 wild type and MKK3/6-/-

cells were grown in DMEM supplemented with 10% (v/v) FBS, penicillin and

streptomycin. Human breast cells: MCF-10A, were grown in DMEM/F12

supplemented with 5% (v/v) horse serum (HS, Biological Industries, Israel), 20 ng/ml

epidermal growth factor (EGF) (Sigma), 10 μg/ml insulin (Biological Industries,

Israel), 0.5 μg/ml hydrocortisone (Sigma), 100 ng/ml cholera toxin (Sigma), penicillin

and streptomycin. HMLE cells were grown in MEBM/DMEM/F12 supplemented as

described above. HMT-3522-S1 cells were grown in DMEM/F12 supplemented with

250 ng/ml insulin,10 µg/ml transferrin, 5 µg/ml prolactin, 10 ng/ml EGF, 10-10

M

71β-estradiol, 10-8

M sodium selenite, 0.5 µg/ml hydrocortisone. HMEC cells derived

from normal breast tissue and were purchased from Cell Lonza and were grown in

human mammary epithelial cells serum-free medium (ECACC). MDA-MB-468 cells

were grown in Leibovitz-F12 supplemented with 10% (v/v) FBS, penicillin and

streptomycin. SUM159 cells were grown in Ham's F12 with 5% calf serum, 5 μg/ml

insulin and 1 μg/ml hydrocortisone. Human pancreatic cancer cells BXPC-3 were

grown in RPMI1640 supplemented with 10% (v/v) FBS, penicillin and streptomycin.

All cell lines were grown at 37 °C with 5% carbon dioxide. To generate stable cell

pools, NIH 3T3, U2OS and MCF-10A cells were infected with pBABE-puro

retroviral vector (McCurrach and Lowe, 2001) expressing T7-tagged human Mnk2

isoform cDNA. Medium was replaced 24 h after infection, and 24 h later, infected

cells were selected for by the addition of puromycin (2 µg/ml) or hygromycin (200

µg/ml) for 72-96h. In the case of double infection with pWZL-hygro-Ras (McCurrach

and Lowe, 2001) or pWZL-hygro-p38α, cells were treated with hygromycin for 72 h

after selection with puromycin . In the case of infection with MLP-puro-shRNAs

vectors, MCF-10A cell transductants were selected for with puromycin (2 μg/ml) for

96 h.

Survival and apoptosis assays

MCF-10A or Panc-1 cells were transduced with the indicated retroviruses. Following

selection, 1×104 cells per well were seeded in 96-well plates. 24h later, the cells were

serum starved for another 24 hours. At 24 hours (before treatment) one 96-plate was

fixed and served as normalizing control ("Time 0"). After starvation the medium was

replaced with starvation medium containing the indicated concentration of

anisomycin and the cells were incubated for an additional 24h. Cells were fixed and

stained with methylene blue as described previously (Karni et al., 2007) and the

absorbance at 650 nm of the acid-extracted stain was measured on a plate reader

(BioRad) and was normalized to cell absorbance at "Time 0". For apoptosis, MCF-

10A cells were seeded on 6-well plates (500×103 cells/well). 24 hours later cells were

incubated with 0.5 or 1 µM anisomycin for 24 or 48 hours. Medium and PBS washes

were collected together with cells trypsinized from each well into 15ml tubes and

centrifuged at 1000g for 5 min. Cells were washed with PBS and after another

centrifugation were resuspended in 100 µl of PBS. 10 µl of the cell suspension was

mixed with 10 µl of 4% trypan blue solution and live/dead cells were counted using a

Bio-Rad TC-10 Automated Cell Counter. After counting the remaining 90 µl of cell

suspension was centrifuged, PBS was discarded and cells were resuspended in 90 µl

of Laemmli buffer. Lysates were separated on SDS-PAGE and after Western blotting

membranes were probed with antibodies against cleaved caspase 3 (Cell Signaling) to

evaluate induction of apoptosis.

RT-PCR

Total RNA was extracted with Tri reagent (Sigma) and 2 µg of total RNA was reverse

transcribed using the AffinityScript (Stratagene) reverse transcriptase. PCR was

performed on 1/10 (2 µl) of the cDNA, in 50 µl reactions containing 0.2 mM dNTP

mix, 10× PCR buffer with 15 mM MgCl2 (ABI), 2.5 units of TaqGold (ABI) and 0.2

mM of each primer; 5% (v/v) DMSO was included in some reactions. PCR conditions

were 95 °C for 5 min, then 33 cycles of 94°C for 30 s, 57°C for 30 s and 72°C for 45

s, followed by 10 min at 72°C. PCR products were separated on 1.5% or 2% agarose

gels. Primers are listed in Table S1.

Q-RT-PCR

Total RNA was extracted with Tri reagent (Sigma) and 2 µg of total RNA was reverse

transcribed using the AffinityScript (Stratagene) reverse transcriptase. Quantitative

PCR was performed using SYBER Green (SYBR® Premix Ex Taq™ # RR041A) and

the CFX96 (Bio-Rad) machine. Unknown samples were compared to a standard

curve, which is established by serial dilutions of a known concentration of cDNA.

The standard that was used is β-Actin. Ct- Threshold cycle from β-Actin and target

gene were inserted into the standard curve formula and the final value was the ratio

between the target gene divided by the β-Actin standard gene. Primers are listed in

Table S1. The PCR reaction is composed of the following steps: 1 cycle at 95ºC for

10 seconds; 40 cycles of 95ºC for 5 seconds and 48ºC for 20 seconds.

Immunoprecipitation

Cells were lysed with CHAPS buffer (Sarbassov et al., 2004), the lysates were cleared

by centrifugation at 12,000g for 15 min, and the supernatant was passed through a

0.45-mm filter. For each cell line, 30 µl of 50% (v/v) protein G–Sepharose beads was

incubated with 2 µg of antibody against HA (Santa Cruz) or 2 µg of antibody against

T7 (Novagen) for 1.5 h at 4°C. The beads were then incubated while rotating with 0.7

mg of total protein from each lysate for 4 h at 4°C. The beads were washed three

times with CHAPS buffer and boiled for 5 min in 50 µl of 2×SDS Laemmli buffer.

After SDS-PAGE and transfer the membranes were probed with antibodies against

Mnk2 and p38α.

Immunofluorescence

Cells plated on a cover glass were rinsed twice with 1 X PBS+ containing Mg

++/Ca

++

and fixed with 4% PFA at room temperature for 30 minutes. Cells were permeabilised

with 0.5% NP40 (Fluka) and then washed with 1X PBS+ containing 0.1% Tween 20

(PBS+T). Following permeabilization, cells were blocked with PBS

+T containing 20%

fetal bovine serum for 45 minutes. Fixed cells were incubated with the following

primary antibodies in 1:1000-1:2000 dilutions: Mouse anti T7-tag (Novagen), Rabbit

anti p38α (Santa Cruz), Mouse anti HA-tag (Santa Cruz) over night at 4°C. Following

incubation, cells were washed once with 1% ammonium chloride and then with

PBS+T, and incubated with the following secondary antibodies in 1:1000 dilution;

Alexa 594 goat anti Mouse (Invitrogen), Alexa donkey anti rabbit 488 (Jackson) and

incubated for 50 minutes at room temperature in the dark. Cover-glasses were washed

with 1% ammonium chloride and then with PBS+T. Cover-glasses were placed onto

slides with mounting solution (Thermo Scientific Cat# TA-030-FM) containing DAPI

(1:1000, Sigma Cat#D9564)). Microscopy was performed on NIKON ECLIPSE Ti /

IntenseLight, C-HGFI using the NIS Elements digital system. Quantification of

cytoplasmic and nuclear cell distribution: Nuclear area was marked by the DAPI

stain. The intensity the FITC and Cy3 channels of each cellular fraction was

calculated using NIS-Elements/Annotations and Measurements software (Nikon).

Dividing the intensity by the area in each cell, we obtained values corresponding to a

specific cellular staining in each cell.

Primers and sequences

Knockdown shRNA oligos

shMnk2a-1:

CAGTGATTCCATGTTTCGTAA

shMnk2a-2:

CAGGTTTGAAGACGTCTACCA

RT-PCR

Mnk2:

Mnk2a/2b Forward: GCTGCGACCTGTGGAGCCTGGG

Mnk2a Reverse: GATGGGAGGGTCAGGCGTGGTC

Mnk2b Reverse: GAGGAGGAAGTGACTGTCCCAC

GAPDH Forward: ATCAAGAAGGTGGTGAAGCAG

GAPDH Reverse: CTTACTCCTTGGAGGCCATGT

p38-MAPK target genes (Mouse):

mCOX2 Forward: TACAAGCAGTGGCAAAGGC

mCOX2 Reverse: CAGTATTGAGGAGAACAGATGGG

mc-FOS Forward: GGCTTTCCCAAACTTCGACC

mc-FOS Reverse: GGCGGCTACACAAAGCCAAAC

mGAPDH Forward: AATCAACGGCACAGTTCAAGGC

mGAPDH Reverse: GGATGCAGGGATGATGTTCTGG

Q-PCR

p38-MAPK target genes (Human):

COX2 Forward: CCGAGGTGTATGTATGAGTGT

COX2 Reverse: CTGTGTTTGGAGTGGGTTTC

c-FOS Forward: GAACAGTTATCTCCAGAA

c-FOS Reverse: TTCTCATCTTCTAGTTGG

SRSF1:

SF2 Forward: GAGTTCGAGGACCCGCGAGACG

SF2 Reverse: GAGCTCCGCCACCTCCAC

Mnk2:

Mnk2a/2b Forward: TCCGTGACGCCAAGCAG

Mnk2a Reverse: GGTCTTTGGCACAGCTG

Mnk2b Reverse: GAGGAAGTGACTGTCCCAC

Actin:

Actin Forward: CCCAGCACAATGAAGATCAA

Actin Reverse: TAGAAGCATTTGCGGTGGAC

GAPDH:

GAPDH For: TCACCACCATGGAGAAGGC

GAPDH Rev: GCTAAGCAGTTGGTGGTGCA

Mnk2 Mutagenesis:

Mnk2aKD:

L113A Forward: CAGGAGTACGCCGTCGCGATCATTGAGAAGCAG

L113A Reverse: CTGCTTCTCAATGATCGCGACGGCGTACTCCTG

Mnk2a L/S:

L/S Forward:

TGCGACCTGTGGAGCAGTGGCGTCATCAGTTATATCCTACTCAGCG

L/S Reverse:

CGCTGAGTAGGATATAACTGATGACGCCACTGCTCCACAGGTCGCA

Mnk2a KKR:

KKR/AAA Forward:

GACGCCAAGAAGAGGGGCAAGAAGGCGGCGGCCGGCCGGGCCACCGACA

GCTTCTC

KKR/AAA Reverse:

GAGAAGCTGTCGGTGGCCCGGCCGGCCGCCGCCTTCTTGCCCCTCTTCTTG

GCGTC

Antisense RNA oligos (all bases 2'-O-methyl):

MKNK2 2a block oligo: CAGCTGTTCCTGGGAAACGGGG

MKNK2 2b block oligo: GGAAGTGACTGTCCCACCTTCAGA

SCRAMBLED (Scr): GCAATCCGCAATCCGCAATCC

Experimental procedures References

Askari, N., Diskin, R., Avitzour, M., Capone, R., Livnah, O., and Engelberg, D.

(2007). Hyperactive variants of p38alpha induce, whereas hyperactive variants of

p38gamma suppress, activating protein 1-mediated transcription. The Journal of

biological chemistry 282, 91-99.

Avitzour, M., Diskin, R., Raboy, B., Askari, N., Engelberg, D., and Livnah, O.

(2007). Intrinsically active variants of all human p38 isoforms. The FEBS journal

274, 963-975.

Dickins, R.A., Hemann, M.T., Zilfou, J.T., Simpson, D.R., Ibarra, I., Hannon, G.J.,

and Lowe, S.W. (2005). Probing tumor phenotypes using stable and regulated

synthetic microRNA precursors. Nat Genet 37, 1289-1295.

Diskin, R., Askari, N., Capone, R., Engelberg, D., and Livnah, O. (2004). Active

mutants of the human p38alpha mitogen-activated protein kinase. The Journal of

biological chemistry 279, 47040-47049.

Karni, R., de Stanchina, E., Lowe, S.W., Sinha, R., Mu, D., and Krainer, A.R. (2007).

The gene encoding the splicing factor SF2/ASF is a proto-oncogene. Nature structural

& molecular biology 14, 185-193.

McCurrach, M.E., and Lowe, S.W. (2001). Methods for studying pro- and

antiapoptotic genes in nonimmortal cells. Methods in cell biology 66, 197-227.

Sarbassov, D.D., Ali, S.M., Kim, D.H., Guertin, D.A., Latek, R.R., Erdjument-

Bromage, H., Tempst, P., and Sabatini, D.M. (2004). Rictor, a novel binding partner

of mTOR, defines a rapamycin-insensitive and raptor-independent pathway that

regulates the cytoskeleton. Curr Biol 14, 1296-1302.

Documents

Cell Reports · PDF fileCell Reports Article Mnk2 Alternative Splicing Modulates the p38-MAPK Pathway and Impacts Ras-Induced Transformation Avraham Maimon,1 Maxim Mogilevsky,1