Pim1 Protein Expression Is Regulated by Its 5'-Untranslated Region and Translation Initiation Factor eIF-4E1

Vol. 8, 1371-1380, December 1997 Cell Growth & Differentiation 1371

Pim-1 Protein Expression Is Regulated by Its 5’-UntranslatedRegion and Translation Initiation Factor eIF-4E1

D. S. Hoover,2 D. G. Wingett,3 J. Zhang, R. Reeves,and N. S. Magnuson4Departments of Genetics and Cell Biology [0. S. H., R. R.], Microbiology[0. G. W., J. Z., N. S. M.], and Biochemistry and Biophysics [R. R.],Washington State University, Pullman, Washington 99164-4233

AbstractExpression of Pim-1 , an oncogenic serine/threoninekinase, is highly regulated at the transcriptional,posttranscriptional, and posttranslationah levels. Here,

we report that expression of Pim-1 kinase isadditionally regulated at the translational level. Pim-1protein expression did not increase in Hut-78lymphocytes in response to PMAI/ionomycinstimulation despite -20-fold increases in mRNA levels,suggesting that translation was repressed. Sequenceanalysis of the 5’-untranshated region (UTR) indicateda long (400 nucleotide), 76% G+C-rich region,characteristics known to inhibit translation. Deletion ofthe 5’-UTR of pim-1 increased translation of the Pim-1protein -10-fold in vitro in reticulocyte lysates and-1.6-fold in vivo in NIH-3T3 cells. When full-length5’-UTR-containing pim-1 cDNA constructs weretransfected into NIH-3T3 cells overexpressingeukaryotic translation initiation factor 4E (eIF-4E),--6-fold higher levels of Pim-1 protein were produced,as compared to that produced in control NIH-3T3 cells.Moreover, elF-4E overexpression had little effect in theabsence of the 5’-UTR, suggesting that it relieved 5’-UTR-mediated inhibition of Pim-1 expression.

IntroductionPim-i is an oncogenic serine-threonine protein kinase (1-3)

that is expressed predominantly in hematopoietic cells, inboth the fetus and the adult (4, 5), and in germ cells of thetestis (6). Overexpression of pim- 1 alone is not sufficient fortumorigenesis but, rather, requires the concomitant overex-pression of a cooperating gene such as c-myc, N-myc, or

bcl-2 (7-9). The mechanisms by which constitutive overex-

Received 5/19/97; revised 9/26/97; accepted 10!3i/97.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to mdi-cate this fact.1 This work was supported in part by United States Department of Agri-culture Grant 91-37206-6867, the Murdock Trust, the Washington StateUniversity Graduate School, and NIH Grants T32-A107025 and ROl-GM46352.2 Present address: Department of Veterinary Microbiology and Pathology,Washington State University, Pullman, WA 991 64-4233.3 Present address: Boise Veterans Affairs Medical Center, Boise, ID83705.4 To whom requests for reprints should be addressed. Phone: (509) 335-0966; Fax: (509) 335-1907; E-mail: [email protected].

pression of pim-1 contributes to the cellular processes that

promote malignant cell growth are not understood. pim-1

expression can be induced by mitogens and by variouscytokines (1 0-i 7). pim-1 expression appears to be requiredfor interheukin-3-mediated proliferative signals (i3) and is

important in influencing cell fate decisions between prohifer-

ation and apoptosis (1 8, 19). Collectively, these observationsstrongly suggest that Pim-i kinase plays an important role inregulating signaling pathways that are involved in cell prohif-

eration and apoptosis.In general, expression of genes that are involved with cell

growth is highly controlled because aberrant expression of-ten results in deregulated growth and malignant transforma-

tion of cells. In keeping with its possible role in regulating cellgrowth, the expression ofpim-1 has, likewise, been shown tobe highly regulated at three different levels: transcriptional,posttranscriptionah, and posttranslational. pim- 1 is induced

as an immediate early gene in a restricted, cell type-specificpattern of expression (1 0-i 7, 20-22). Despite features thatcharacterize its promoter as one belonging to a constitutivehyexpressed housekeeping gene (20), nuclear run-on analyseshave demonstrated that increases in the steady-state levels

of pim-1 mRNA observed after mitogenic stimulation are, inpart, the result of increases in transcriptional activity (23, 24).However, levels ofpim-1 mRNA are also controlled posttran-scriptionally by modulation of mRNA stability (i6, i7, 21).

pim-1 expression has also been shown to be regulated post-translationally because both the murine and human M,34,000 Pim-i proteins are rapidly degraded in vivo (T112 =

-10 and 6 mm, respectively; Refs. 1 and 22).� Here, wereport an additional point of regulation of Pim-i protein 1ev-

els, in the form of translational control.Translational control can be mediated by a variety of ele-

ments within the 5’-UTR6 of mRNAs, including leader length,minicistrons, pohypyrimidine tracts, secondary structure, andthe consensus sequence surrounding the initiating AUGcodon (25-28). Particularly, it has been shown that extensivesecondary structure within the 5’-UTR effectively inhibits

translation (29). Interestingly, expression of a number ofgenes encoding growth factor receptors and proto-onco-proteins has been shown to be regulated by elements withinthe 5’-UTR (30-39). It has been postulated that translationalregulation is an important molecular mechanism for control-hing levels of proteins that are critical to cell growth and

differentiation (25).In addition to cis-acting elements within the 5’-UTR, other

important components of translational control are protein

5 0. S. Hoover and N. S. Magnuson, unpublished observations.6 The abbreviations used are: UTR, untranslated region; PMA, phorboli 2-myristate B-acetate; TCA, trichloroacotic acid; f3-gal, f3-galactosidase;0Oc, omithino decarboxylase.

A.timehrs: 0 4 6

� �- p/rn-i

a� �-GAPOH

123

C timehrs: 0 4 8

� �- pfrn-i

I N �- GAPDH

123

D tlmehrs: 4 0 4 8PRE

Mr r IM -j---IMMUNE---1Kd

36- �

S. -.-Pim-1

29-

1 2 3 4

36-S

29-

#{176}Pim-1

1372 Translational Regulation of Pim-1

Fig. 1 . Pim-1 translation is inde-pendent of transcriptional in-creases in Hut-78 lymphocytestreated with PMA/ionomycin.Hut-78 cells were stimulated with10 ng/ml PMA plus 1 �u�i ionomycinfor 0, 4, and 6 h. U937 myelomono-cytic cells were treated with 100units/mI IFN-y. A and C, Northernanalyses of Hut-78 cells and U937cells, respectively. Total RNA wasisolated at each time point, and 20;Lg of each sample was analyzedby agarose gel, transferred to nitro-cellulose, and probed with a pim- 1cDNA sequence. Equivalent load-ing was verified by probing withGAPDH. B and D, immunoprecipi-tations of Hut-78 and U937 cells,respectively. [�5SjMethionine-la-beled cells were treated as do-scribed above, and equivalentamounts of TCA-precipitablecounts were immunoprecipitatedwith preimmune serum (Lane 1) orantibody to Pim-1 (Lanes 2-4), fol-lowed by SDS-PAGE analysis, asdescribed in “Materials and Meth-ods.”

B. timohrs: 0 0 4 6PRE

Mr I IM T IMMUNE -�

Kd

1 2 3 4

factors involved in protein synthesis (40, 41). One such factor

is the 5’-m7 GpppN cap-binding protein eukaryotic transla-

tion initiation factor 4E (eIF-4E; Ref. 42). elF-4E regulates

translation in at least two independent processes. First,

eIF-4E functions directly in formation of translation initiation

factor complex eIF-4F (43) and is the rate-limiting component

in formation of this complex. The eIF-4F complex facilitates

mRNA-ribosome binding, scanning and initiation of transla-

tion (43-46). Second, eIF-4E can increase translation by

facilitating mRNA export from the nucleus as has been

shown for cyclin Dl mRNA (47).

Here, we report that expression of pim-1 mRNA is regu-

lated by both the 5’-UTR and levels of eIF-4E. Our findings

demonstrate an additional level of cellular regulation of pim-1

that was not previously identified that may be an important

control point for Pim-1 protein kinase-related functions.

ResultsPirn-1 Translation Is Independent of Transcriptional In-creases in mRNA in Hut-78 Lymphocytes Stimulated withPMA plus lonomycin. The first indication that cellular levels

of Pim-1 kinase might be controlled in vivo by the efficiency

of messenger RNA translation came from experiments in

which the levels of both pim-1 mRNA and protein in two

different human hematopoietic cell lines were compared. We

and others have shown that pim-1 mRNA expression is

inducible by both protein kinase C (23) and Jak/STAT-medi-

ated signal transduction pathways (1 6, 1 7, 48). We were

interested in extending these studies to include protein ex-

pression. We investigated the effects of PMA plus ionomycin

on the human T-cell line Hut-78 and IFN-y on the human

myelomonocytic cell line U937. At various time points fol-

lowing treatment, the stimulated cells were analyzed for both

pim-1 mRNA and protein levels by Northern blot and immu-

noprecipitation procedures, as described in “Materials and

Methods.”

Hut-78 cells were treated with 1 0 ng/ml PMA and 1 j.�M

ionomycin and U937 cells were treated with 1 00 units/mI

IFN-y for the times indicated (Fig. 1). Northern blot analyses

(Fig. 1 , A and C) revealed that both cell types exhibited an

approximately 20-fold increase in pim-1 mRNA levels by 4 h

poststimulation, and message levels remained high 6 or 8 h

after treatment. These results indicate that PMA plus iono-

mycin or IFN-y significantly increase the steady-state levels

of pim-1 mRNA in Hut-78 and U937 cells, respectively, by

relatively similar amounts.

In concurrent experiments, Pim-1 protein levels were

measured in treated Hut-78 and U937 cells. Immunoprecipi-

tations were performed on equal amounts of TCA-precipita-

ble counts of [35S]methionine-labeled cells (Fig. 1 ,B and D).SDS-PAGE analysis of the immunoprecipitates revealed a

surprising result. In experiments in which U937 cells were

treated with IEN-y, the cellular levels of Pim-1 protein, as

expected, increased, in concert with increasing concentra-

tions of cellular pim-1 mRNA, to levels -5-fold higher than

that present in unstimulated U937 cells by 8 h poststimula-

tion (Fig. 1D). In marked contrast, in treated Hut-78 cells,

Pim-1 protein levels did not increase, despite -�20-fold in-

creases in mRNA during the course of this induction (Fig. 1,

A and B). Additionally, we did not observe any change in

cellular Pim-1 protein levels when the Hut-78 cell stimulation

time course was extended for up to 1 2 h (data not shown).

This apparent failure of Pim-1 protein to concomitantly in-

crease with increasing mRNA levels could not be attributed

to changes in protein half-life because the T112 of Pim-1

protein in Hut-78 cells was determined to be -�6 mm in either

the presence or absence of PMA plus ionomycin (data not

shown). Additionally, we verified that the lack of increase in

GGACC4AGCA0CA0CA0CA0CA0CA0CA4C�cCACTAGcCTcCTGCCCCGCG0c0CT

�

OcGccCTcccGccGCCAGTCCCCGCAGCCCCCTCAOTTOTCCTccOACTCCcCCTcGOCC

CCCAOGCATAGcCTTcGGCACAGcCCcGCCTcCGOCTcCTGcOGCAOCTCCTCT000CAC

. . I . +i .CGTCCCTGCGCCGACATcCTGGAGGTTGCGATGCTCTTGTCC&&MTCAACT�0CTTGCC

Fig. 2. Structural analysis and secondary structurepredictions of the 5’-UTR of thepim-1 transcript. A,

the cONA sequence includes -400 nucleotides ofupstream sequence, including the initiating ATGcodon. #{176},putative cap site; 0, an octamer CAGrepeat; underilning, the initiating ATG (+1) codon;arrow, the site at which the 5’-UTR was truncated.B, secondary structure of the 5’-UTR of pim-1mRNA was predicted using a modification of theFOLD program (51). The overall theoretical Gibbsstandard free energy is -153 kcaVmol. C, diagramof the full-length (F5) and deleted (05) 5’-UTRpim-lconstructs used for in vitro (pKSpim.F5/.O5) and invivo assays (pRBKpim.F5/O5).

B

C

-390

+1

5’ UTR

Cell Growth & Differentiation 1373

A-390

-330

-270

-210

-150

-90

-30

-331

-271

-211

-151

-91

�31

29

+1 Codhig A#{216}onPIm-1 pKSpim. F5/

-I pRBKpim. F5ATG

-13 � CodIng Region PIm-1 pKSpIm. 05/

�5’LrrR

ATG

Pim-1 protein levels was not a result of our normalizationprocedures. Because a marked increase in protein synthesis

would increase the TCA-precipitable counts:protein ratio and

skew sampling, we compared total TCA-precipitable counts

at each time point for three different experiments. We found

that both treated and untreated cells had similar rates ofmethionine incorporation. Collectively, this data indicated

that pim-1 translation was repressed in Hut-78 cells upon

treatment with PMA and ionomycin.Sequence Analyses and Secondary Structure Predic-

tions of the 5’-UTR of the pim-1 Transcript. To determinethe translational control mechanisms regulating Pim-1 pro-tein levels, our attention focused on analysis of the 5’-UTR of

the pim-1 message. Analysis of the 5’-UTR of the pim-1cDNA sequence (Ref. 49; Fig. 2A) revealed several important

features that are known to influence translation. The 5’-UTRis unusually long, with transcription start sites mapping to the

region between -400 and -390 (20). This 400-nucleotide-long UTR is significant because 80-90% of known 5-UTRs

are 1 00 nucleotides or less in length (50). Additionally, it has

been shown that long UTRs play an important role in trans-

lational regulation (50). Importantly, the 5’-UTR is 76% G+C-

rich, suggesting a high degree of secondary structure. The

theoretical standard free energy of the 5’-UTR of the pim-1message was determined using a modification of the FOLD

program (Ref. 51 ; Fig. 2B). This program predicts the sac-ondary structure and minimum Gibbs free energy, i�G, for a

series of suboptimal conformations of RNA sequences,

based on assumed values of stacking and loop destabilizing

energies. Fig. 2B shows the predicted secondary structure

and the theoretical Gibbs free energy (�G = -1 53 kcaVmol)for the entire pim-1 5’-UTR. The average i�G for the stem

loop structures between nucleotides -350 and -265 was

calculated to be -48 kcaVmol. Because formation of sac-

ondary structure of -37 to -50 kcaVmol has been shown to

effectively inhibit translation, these predicted free energy

values are significant (26, 29, 52, 53). The sequence structure

predictions of Pim-i 5’-UTR are consistent with models in

which highly G+C-rich sequences form secondary struc-

tures that effectively reduce translational efficiency (26, 28,

52, 53). Further sequence analysis indicates there are no

upstream ATG codons in the pim-1 5’-UTR, indicating the

absence of traditional alternate upstream open reading

frames, which could decrease the overall efficiency of trans-

. � t- Pim-1

B.

MrKd

45-

36-

29-

25-

C.

Mr

Kd

45- �‘.

Fig. 3. Deletion of the 5’-UTR of pim-1 mRNA increases translationalefficiency in reticulocyte lysatos and in vivo in transfected cells. A, in vitrotranscribed full-length pim-1 5’-UTR, pKSpim.F5, or the 5-deleted ver-sion, pKSpim.D5 transcripts (2 pg), were analyzed on a 1 % agarose gel,stained in Sybrgreen, and visualized on a Fluorimager as described in“Materials and Methods.” B, equimolar concentrations of the above tran-scripts were translated in a reticulocyte lysato with a no-RNA reaction ascontrol. Translation reactions were analyzed by SOS-PAGE analysis andautoradiography. C, NIH-3T3 cells were cotransfected with 20 �g of pim-1plasmid containing oitherthe full-length 5’-UTR pRBKpim.F5 (Lanes 1 and3) or a plasmid in which the 5’-UTR had been deleted pRBKpim.D5 (Lanes2 and 4) and 5 j�g of pSV-�-gai. Transfected cells wore metabolicallylabeled and immunoprecipitated with Pim-1-specific antibody.


lation from the initiating AUG codon. However, there are

three in-frame CUG codons at -285 to -283, -228 to -226,and -i2 to -10. The context of sequence surrounding thesingle initiation codon, GGGAUGC, is not optimal for efficienttranslation, according to the consensus sequence estab-

hished by Kozak (50, 54), which would indicate a general,

decreased efficiency of pim-1 translation from this codon.There are also eight tandemly repeated copies of the se-quence CAG located at nucleotides -323 to -300 of the5’-UTR, which is located in a region predicted to be highlyfolded (Fig. 2B). This repeat is also conserved in the mouse,

and a database search (55) of this sequence shows that a

variety of mRNAs, including receptors, cytokines, and devel-opmentally regulated genes in Drosophila, share this re-peated sequence. Finally, sequence analyses of the humanand murine pim-1 5’-UTRs indicates 79% identity betweenthe two UTRs with calculated �G = -153 for human and- 132 kcal/mol for mouse. The evolutionary conservation ofthese 5’-UTRs suggests important biological functions forthese sequences. Together, these findings indicate that ele-ments within the 5’-UTR could modulate the efficiency ofpim-1 mRNA translation.

The 5’-UTR Inhibits Translation ofpim-1 mRNAin Vitroin Reticulocyte Lysates. To experimentally determine theinfluence of the 5’-UTR on pim-1 translation, we generated

pim-1 cDNA containing plasmids with either the full-length5’-UTR (pKSpim.F5) or a version that is truncated 13 nude-otides upstream of the ATG codon (pKSpim.D5; Fig. 2C).This deletion eliminated 377 nucleotides of upstream G+C-rich sequence, while maintaining the context of the sequence

flanking the ATG translation start codon. In vitro transcrip-

tions were performed on both the full-length pim-1 5’-UTR-containing plasmid and the truncated 5’-UTR-containingplasmid. Approximately equal amounts of the transcriptionreactions were run on a 1 % agarose gel to verify concentra-tion and the integrity of the transcripts. Fig. 3A shows both

transcripts, and the size difference between transcripts in ai % gel is consistent with the size of the nucleotide deletion

of the 5’-UTR. Equimolar concentrations of the two tran-scripts were translated in reticulocyte lysates. Fig. 3B showsa representative SDS-PAGE analysis of the translation reac-tion. We found an average -1 0-fold increase of the Mr

34,000 Pim-i protein translated from the 5’-deleted pKS-pim.D5 transcript, as compared to the protein translatedfrom the full-length pKSpim.E5 transcript. The predominantform of the protein is the predicted M� 34,000 protein, with alesser amount of a Mr 35,000 protein. A longer exposure ofthis autoradiograph indicates that a M, 35,000 protein is alsopresent in Lane 3. The Mr 35,000 protein is presumablyinitiating from CUG codon -iO to -12, as has been shown

for murine Pim-i (1). However, unlike the murine model, thereis no detectable CUG-initiated Mr �‘44,000 Pim-i proteinproduct. Several lower molecular weight proteins were ob-served in the in vitro translation products (Fig. 3B, asterisks).These smaller products may be produced from initiation at

downstream CUG codons because we have noted there arefive in-frame CUG codons immediately downstream of theAUG codon, three of which are in strong Kozak context (56)at positions +57 to +59, +99 to +ioi , and +252 to +254.

k ‘#{176} Q�

A.

Q� Qt,

36-

..�, �<t”#{176}�

(�1 �

- �1 2 3

�,, �.� � .� �

Q� qt�’ �

0 #{149} 4- Pim-1

29-

25- � �

�

1 2 3 4

Proteins initiating from these codons would have apparentmolecular masses of approximately 30, 31 , and 29 kDa,respectively. There is also a downstream AUG codon at

position +262 to +264, which is in a strong Kozak contextand could potentially yield a product of Mr 25,000. We have

F’ PRE____ _____Li. I lM ( IMMUNE

Mr + + - + - 3T3-4E(P2)

Kd � � �3&

2� � 11* �- Pim-1

. 1 2 3 4 5

Fig. 4. Overexpression of elF-4E in NIH-3T3 cells increases translationof Pim-1 protein. 3T3-4E(P2) (+) or 3T3 (-) cells were cotransfected withpRBKpim.F5 and pSV-f3-gal. Replicate plates from two independent setsof transfections were harvested for Northern analysis (Fig. 4, A and B,Lanes 3-6) and immunoprecipitation (Fig. 4C). Nontransfected controlcells are shown in Lanes 1 and 2. A, 20 �g of total RNA were loaded ineach lane, and blots were sequentially probed with pim-1 and p-gal. B,Lanes 1, 3, and 5 (+), cells overoxprossing elF-4E; Lanes 2, 4, and 6 (-),

3T3 control cells; Lanes 3 and 5 and 4 and 6 are replicates. C, immuno-precipitations were performed with anti-rPim-1 antibody on [�5S]methi-onine-labeled cells normalized to (3-gal activity, followed by SOS-PAGEanalysis as described in “Materials and Methods.” Lanes 1, 2, and 4 (+),

cells overexpressing elF-4E; Lanes 3 and 5 (-), 3T3 control cells; Lanes 2and 4 and 3 and 5 are replicates.


observed these lower molecular weight proteins in our trans-

fection assays (see Fig. 5, A and B, asterisks), indicating thattheir expression is not restricted to the conditions of the

reticulocyte hysate. These proteins are unlikely to be degra-

dation products of the Mr 34,000 Piml protein because they

were also present when the ATG codon was mutated to a

TTG and the Mr 34,000 protein was not expressed (data notshown).

We next determined whether a similar inhibition by the

5’-UTR also occurred in vivo. We constructed eukaryoticexpression vectors containing either the full-length pim- 1

cDNA (pRBKpim.F5) or the 5’-UTR-deleted cDNA (pRBK-

pim.D5) and transfected these plasmids into NIH-3T3 cells

(Fig. 3C). We had previously determined that pim-1 mRNAwas not detectable in these cells under our conditions (datanot shown and Fig. 4A, Lanes 1 and 2). A -1 .6-fold increasein translation of Pim-1 protein was observed when the 5’-

UTR was deleted (Fig. 3C, Lanes 2 and 4). We have observeda similar reproducible increase in transfected Wehi-168cells.5 We do not know the reason for differences in levels of

inhibition mediated by the 5’-UTR in reticulocyte lysates, as

compared to transfected cells; they likely reflect differencesbetween experimental systems. Collectively, these resultssuggested that Pim-1 translation was inhibited by the pros-once of the 5’-UTR.

Overexpression of elF-4E in NIH-3T3 Cells IncreasesPim-1 Translation. We have shown that the 5’-UTR ofpim-1 is highly G+C-rich, is predicted to contain extensive

secondary structure (Fig. 2B), and inhibits translation ofpim-1 in reticulocyte lysates and in transfected cells. Trans-

lation initiation factor ehF-4E has previously been shown to

increase expression of proteins containing extensive sec-

ondary structure in their 5’-UTRs (53). To determine whether

the 5’-UTR of the pim-1 mRNA is a downstream target ofeIF-4E action, we examined the effects of overexpression ofelF-4E on pim-1 mRNA translation. NIH-3T3 cells, stablytransfected with eIF-4E (57), were used for these transfectionexperiments. Endogenous pim- 1 message was not detectedin our Northern blot conditions (Fig. 4A, Lanes 1 and 2).Therefore, we determined the effects of elF-4E overexpres-

sion on pim- 1 translation using a pim- 1 transient transfectionassay. Stably transfected NIH-3T3 cells overexpressingelF-4E are referred to as 3T3-4E(P2) (58). These cells werecotransfected with pRBKpim.F5 and pSV-13-gal as a controlfor transfection efficiency. Replicate sets of transfections

were performed using the same pool of DNA/calcium phos-

phate crystals per replicate set. Both pim-1 mRNA and pro-tom levels were analyzed from each set of transfections. Two

representative samples of the RNA and protein are shown inFig. 4, A and C. pim- 1 translation was increased an average

-�6-fold in the 3T3-4E(P2) cells, as compared to the control3T3 cells (Fig. 4C, Lanes 2 and 4 versus 3 and 5). Althoughthere is a small increase in pim- 1 mRNA levels present in thetransfected 3T3-4E(P2) cells as compared to 3T3 cells alone

(Fig. 4A, Lanes 3 and 5 versus 4 and 6), this increase is minorwhen it is compared to the levels of increased translation

observed. Additionally, we have not seen a one-to-one cor-relation in increased pim-1 mRNA to protein under any con-ditions we have investigated. We also monitored frgal

A. � - 4 - + - 3T3-4E(P2)

,up,u, �- pirn-1

123456

B. � - + - + - 3T3-4E(P2)

!!� �- �3-gaI

123456

mRNA expression by reanalyzing the Northern with a a-gal

probe. It was evident that /3-gal mRNA levels were similar(Fig. 4B), and the ratio of 13-gal mRNA to pim- 1 mRNA wasproportionately equivalent in both cell types. Measurementof 13-gal activity suggests that the increased expression me-diated by elE-4E on pim-1 is specific. If a-gal activity (as ameasure of protein concentration) was also increased by

elF-4E, then normalizing for this increased activity, by reduc-ing sample, would reduce Pim-1 protein, relative to controls,

and this is not observed. Additionally, others have previouslyshown, with these same cell lines, that eIF-4E overexpres-sion appears to increase translation of only a specific subsetof proteins because proteins such as S-adenosylmethionine,

Plm-1

4-q�r29- �-�- *

25-

C

� ‘�. 0

Mr � �

kO

45-

36-

4E

29- �

25-

B �;BK;m.D:i

Mr� �

kO � �

45-

36-Plm.1

� 1�29- *

25-

,-, “I #{163}�1 0E � ��:�7. � 0) (�)

.� 6�0>510

0’

,�0

0C)

C,,


Fig. 5. elF-4E overcomes translational re-pression mediated by the 5’-UTR. A and B,immunoprecipitations and SOS-PAGEanalysis of r55]methionine-labeled 3T3-4E(Ala), 3T3-4E(P2), and 3T3 control cellstransfected with thefull-length 5’-UTR con-taming plasmid pRBKpim.F5 (A) or the5-deleted UTR containing plasmid pRBK-pim.D5 (B). Columns, the data after normal-ization to Pim-1 levels present in the 3T3control cells within each set of transfec-tions. C, Western blot analysis of elF-4Eoveroxpression in 3T3-4E(Ala), 3T3-4E(P2),and 3T3 control cells. Total protein (75 �g)was loaded in each lane and run on SOS-PAGE, as described in “Materials andMethods.”

A +pRBKpim.F5

I��1‘T Q.

� � 0

Mr

kD � �

45-

36-

,�.�

#{176}- 0) �‘)

00

S

0

0>

0

0cc

Rb, and actin do not increase in response to overexpressionof elE-4E in the 3T3-4E(P2) cells (36, 58). Such a finding isconsistent with the idea that mRNAs are translated with

different efficiencies in a competitive situation (59, 60). This

data suggests the increased Pim-1 protein levels are, in fact,dependent on elE-4E levels, suggesting a translational reg-uhatory mechanism that is independent of increases in pim-1

mRNA levels.

Increased elF-4E Expression Can Overcome the Inhib-itory Effect of the 5’-UTR on Pim-1 Translation. Becausethe presence of the 5’-UTR of pim-1 repressed translation(Fig. 3) and overexpressed eIF-4E increased expression ofPim-1 in transfections using plasmids containing the full-length 5’-UTR (Fig. 4), we investigated whether deletion of

the 5’-UTR altered elE-4E-mediated increased expression.We compared Pim-1 protein expressed in transient transfec-

tion assays using either the full-length 5’-UTR-containing

plasmid (pRBKpim.F5) or the 5’-UTR-deleted version (pRBK-

pim.D5). These plasmids were transfected into 3T3-4E(P2)

cells, which overexpress elF-4E, NIH-3T3 control cells, and,

as an additional control, 3T3-4E(Ala) cells. The 3T3-4E(Ala)cells contain a Ser-53 to alanine mutant of elF-4E (57), whichis thought to reduce elE-4E function. Immunoprecipitationsare shown in Fig. 5, A and B. Samples were normalized fortransfection efficiency independently for Fig. 5, A and B.Pim-1 protein levels were found to be -�7-fold higher in the

3T3-4E(P2) cells when the 5’-UTR was present (Fig. 5A), ascompared to Pim-1 expressed in the control cells, which didnot overexpress elF-4E. In contrast, the effect of elF-4E

overexpression was much less in the absence of the 5’-UTR.When the 5’-UTR deletant was transfected into cells over-expressing elF-4E, only a --1 .7 fold additional increase was

seen (Fig. SB). Taken together, these data suggest that the

5’-UTR inhibits basal Pim-1 expression but that elE-4E canact on the 5’-UTR to overcome this inhibitory effect. It isinteresting to note that in each case, pim- 1 translation was

less in the mutant 3T3-4E(Ala) cells than it was in the 3T3control cells. This decrease in protein translation has also

been observed by others (61) and may suggest dominant-negative effect of the elF-4E alanine mutant. A Western blot

(Fig. 5C) demonstrates that eIF-4E levels are -8 fold higherin the elF-4E stably transfected NIH-3T3 cells than those

observed in the control cells.

Discussion

The data reported here demonstrate an additional level of

control over the expression of the pim- 1 proto-oncogene.

Overexpression of the pim- 1 gene has been shown to actsynergistically with other overexpressed protein factors toinduce neoplastic transformation of lymphoid cells and in-crease susceptibility to chemically induced lymphomagen-

esis in vivo (7, 62). Therefore, elucidation of the molecular

mechanisms regulating the cellular levels and activity of the

Pim-1 kinase are important for an understanding of how this

proto-oncogene induces cancer. Here, we have providedevidence that one aspect of control of cellular Pim-1 kinaselevels involves pim-1 protein expression.

The 5’-UTR ofpim-1 mRNA contains sequences that pro-dict translational regulation, including a highly structured5’-UTR sequence, upstream CUG codons, and a poor Kozakconsensus sequence. To determine the effects of the up-stream sequence on translation, we created constructs thatwere deleted of the 5’-UTR. We found, in in vitro transcrip-

Cell Growth & DIfferentiation 1377

tion/transhation and transfection experiments, that the 5’-

UTR inhibited translation. On the basis of these observations,

we were interested in determining if elF-4E, which has beenshown to increase efficient translation of mRNAs with en-cumbered leader sequences (53), might influence translation

of pim-1 mRNA. This would suggest an additional mecha-nism for controlling Pim-1 kinase levels and Pim-1 -associ-ated malignant transformation. We found that elF-4E over-

expression in 3T3-4E(P2) cells significantly increased

expression of pim-1 mRNA, producing at least -6-fold

higher levels of protein, as compared to the 3T3 control cells.This increase was independent ofpim-1 mRNA levels, which

were essentially the same in both 3T3-4E(P2) and 3T3 controlcells. This pronounced effect of elF-4E was only seen in the

intact, 5’-UTR containing mRNA, suggesting that ehF-4E actsby relieving the inhibitory effect of the 5’-UTR.

The importance of elF-4E and proteins that are subject to

its control was demonstrated when its overexpression wasshown to induce malignant transformation of NIH-3T3 cells

(57) and aberrant growth and morphology of HeLa cells (63).Additionally, microinjection of elF-4E mRNA into Xenopusoocytes leads to the induction of mesoderm from ectodermalexplants (64). At least one target of elF-4E-mediated trans-

lational increases, ODC, has been shown to cause neoplastic

transformation as a result of this overexpression (36). In thecase of ODC, elF-4E is thought to function primarily at the

level of translation initiation and loading of the mRNA to the

ribosome (47). However, an additional function for elF-4E is

the regulation of the cell cycle-associated protein, cydlin Dl.Here, elF-4E functions in nucleocytoplasmic transport andfacilitates export of cydlln Dl mRNA to the cytoplasm,

thereby increasing accessibility to the translation machinery(47, 56). We are presently investigating the specific mecha-nism of eIF-4E in pim-1-associated translational increases.

Because elF-4E overexpression increases expression of

only a subset of mRNAs (59, 60), it is likely that these down-stream effectors are important for cell growth and develop-

ment. In particular, it has been suggested that translation ofproteins that are required for G1 cell cycle progression ishighly cap-dependent and, thus, more sensitive to changesin eIF-4E levels and/or activity (65). ODC and cychin Dl , two

proteins known to be regulated by increased expression of

elF-4E (36, 61), are expressed during G1 and are necessary

for G1-S transition (61 , 66). We have previously shown that,

likewise, Pim-1 protein is induced during G1, by prolactin, ina T-ceII lymphoma line (17). Thus, Pim-1 is a new example of

a G1 protein, the expression of which can be modulated byelF-4E levels. Collectively, these observations indicate that

translation of proteins involved in cell growth is highly sen-sitive to changes in elF-4E levels and suggest that Pim-1 is

a downstream target for eIF-4E-associated, deregulated

growth. Additionally, as might be predicted for proteins thatare translationally regulated, the half-lives of Pim-1 , ODC,

and cydlin Dl are quite short: -‘1 0 mm, i o mm, and <30 mm,respectively (1 , 67, 68). These data support the contention

that proteins involved in proliferative responses are highly

regulated at the translational and posttranshational levels.Translational regulation was suggested by data that

showed a discrepancy in the levels of induced pim-1 mRNA

and Pim-1 protein produced by treatment of Hut-78 cellswith PMA and ionomycin. In these cells, pim-1 transcriptionwas increased 20-fold, whereas Pim-1 protein expression

did not change. The decreased protein:mRNA ratio was notdue to changes in protein stability, suggesting that transla-

tion of pim-1 was down-regulated. We have also observed

discrepancies in mRNA:protein ratios in other cell lines (17).�

Because elF-4E modulated pim-1 expression in transfected

NIH-3T3 cells, we determined eIF-4E levels in PMMonomy-

cm-treated Hut-78 cells. We found no change in steady-stateelF-4E levels (data not shown). This suggests the transha-

tional repression ofpim-1 mRNA is not entirely dependent on

steady-state eIF-4E levels but may also be dependent on

other factors, such as the phosphorylation state of elF-4E,

additional regulatory proteins, such as 4E-BP1(PHAS-l) and

4E-BP2 (69, 70), which bind to elF-4E and inhibit cap-de-

pendent translation (71 , 72), or reduced transport of pim-1mRNA from the nucleus to the cytoplasm.

Because of the deleterious effects of dysregulated cell

growth, oncoproteins, such as Pim-1 , are controlled at mul-

tiple levels. These forms of regulation are controlled byvarious signaling pathways. In Hut-78 T cells, PMAiionomy-cm activates protein kinase C and pim-1 transcription (23)

but does not appear to activate and may inhibit pathwaysnecessary for efficient translation of Pim-1 (73). In contrast, in

U937 myeloid cells treated with IFN--y, the Jak kinase path-

way may mediate transcriptional regulation of pim-1,

whereas P1-3 kinase pathways, which increase elF-4E activ-

ity through 4E-BP1 phosphorylation (60), may be involvedwith the eIF-4E-mediated translation of Pim-1 . Further ax-

amination of the various signaling pathways involved withpim-1 translation should provide important information as to

how Pim-1 protein expression is regulated under both nor-

mal and neoplastic conditions in hematopoietic cells.

Materials and MethodsCell Culture. Hut-78 T lymphocytes and U937 myolomonocytic cellswere obtained from American Type Culture Collection (Rockville, MD) andmaintained in RPMI 1640 (Life Technologies, Inc.) with 10% FCS (Hy-clone), 100 units/mI penicillin, 100 �g/ml streptomycin, 3 m� glutammne,and 10 mM Hepes (pH 7.3). NIH-3T3, NIH-3T3-4E(P2), and NIH-3T3-4E(Ala) cells were a gift of Or. N. Sonenberg (McGill University). TheNIH-3T3-4E(P2) and the NIH-3T3-4E(AIa) cells are referred to as 3T3-4E(P2) and 3T3-4E(Ala), and they overexpross the wild-type elF-4E andthe Ser-53 to Ala mutant, respectively (57). These cells were maintained

on 100-mm plates in OMEM (Life Technologies, Inc.), supplemented asdescribed above with the addition of 100 m�i sodium pyruvate. 3T3-4E(P2)and 3T3-4E-ALA cells were additionally supplemented with Geneticin(G41 8; Life Technologies, Inc.) at concentrations of 200 and 100 �g/ml,respectively.

IFN-’y (Boehringer Mannhelm) was diluted in RPMI to a stock concen-tration of 1 x 10� units/mI. PMA (Sigma Chemical Co.) was diluted in

ethanol to a final stock concentration of 2 mg/mI. lonomycin (Calbiochem)was a 5 mM stock solution in DMSO.

Antibodies. Polyclonal antibodies to Pim-1 were made against re-combinant glutathione-S-transferase-Pim-1 (gst-Pim-1) fusion protein,

which was purified as described (3). New Zealand white rabbits wereimmunized, and the sera were collected according to Harlow and Lane(74). Anti-eIF-4E was a gift of Or. N. Sonenberg.

Plasmid Constructions. Standard protocols were used for all recom-binant DNA techniques (75, 76). A full-length 2.67-kb pim-1 cONA (49)with flanking EcoRl linkers was subcloned into an EcoRI-digosted pBlue-

script-KS plasmid (pKS; Stratagene). This construct contains the full-length 5’-UTR referred to as pKSpim.F5 (Fig. 2C). To assess the effects of


the 5’-UTR on translational efficiency, a construct containing a deletion of

the 5’-UTR was constructed using PCR-mediated mutagenesis. Thesense PCR primer GTGCTCGATATCCCTGGAGGUGGGATG includedan EcoRV site (underlined) 13 nucleotides upstream from the ATG trans-lation start codon, which was used for subcloning (Fig. 2A). The antlsensePCR primer included nucleotides +1 19 to +89. The PCR-generated frag-

ment(143 nucleotides)wasthen digested and subcloned into EcoRV- andNail-digested pKSpim.F5. The resultant plasmid contains a deletion of377 nucleotides of the 5’-UTR and is referred to as pKSpim.D5 (Fig. 2C).

The expression vector pRBK (Invitrogen), which uses the Rous sarcomavirus promoter, was used for cloning and expression of pim-1 in mam-malian cells. The full-length pim-1 cONA was digested from a pKSpimplasmid, in which pim-1 was in a 5’-to-3’ orientation relative to the T7promoter using Nofl and ligated into a NotI-digested site in the pRBKpolylinker. This plasmid that contains the full-length pim-1 5’-UTR wasdesignated pRBKpim.F5. A pRBK plasmid containing the deleted 5’-UTR

pim-1 cDNA was created from pKSpim.D5. An EcoRV- and NotI-digestedpKSpim.O5 2.3-kbpim-1 fragment was ligated into a Nofi-digested pRBKplasmid, followed by a fill-in reaction, to which the EooRV site was bluntend-ligated. The resultant plasmid was designated pRBKpim.O5. Se-quences of plasmid constructs were verIfied by dideoxy chain terminationsequencing according to manufacturer’s instructions (United States Blo-chemical).

RNAlsolatlon and Northern BlotAnalysis. Cells weregrown, treated(where indicated), and harvested as described below for immunopreclpi-tation and metabolic labeling. Approximately 5 x iO� cells were collected

atthe indicated times. Foradherent cells, two 100-mm plateswere pooled

for each analysis. The RNA was isolated using the guanidinium isotbo-cyanate-cesium chloride method (75). The filters were hybridized wIth a

pim-1 (49), a-gal (Promega)or GAPOH (American Type Culture Collection)probe, as indicated In the figure legends.

In Vitro Transcription and Translation. Transcription reactions of theplasmids pKSpim.F5 and pKSpim.O5 were performed using T3 RNApotymerase according to recommendations by the supplier (Stratagene).

RNAtranscripts were quantitated, and 2 �g of each franSCIIpt were run ona 1% agarose-formaldehyde gel. The gel was stained in Syb green (Mo-lecular Probes), and the RNA was visualized on a Huorimager (Molecular

Dynamics). Equimolar concentrations of each transcript were then trans-lated in a 25-pJ total reaction volume of a rabbit reticulocyte lysate(Promega) for 30 mm. Translation reactIons were diluted with a 5X SOSsample buffer, boiled, and analyzed by SOS-PAGE. The gel was placed influor, dried, and quantitated using a Phosphorimager (Molecular Dynam-

ice).Transfections and �3-gaI Assays. Transient transfections were per-

formed as described (77). Calcium phosphate/DNA crystals containing 15hg of pim-1 plasmid DNA constructs plus 5 �sg of �-gaI were added tocells. Crystals remained on cells for 18 h. Cells were washed and refed

with fresh media and incubated for another 30 h before harvesting. �-gaIactMty was determined on cells cotransfected with the approprlatepim-1

cONA plasmid and pSV-p-gal (Promega) according to the protocol out-lined by Promega, with some modifications, as described below. Cellswere metabolically labeled for immunopreolpitations prior to a-gal anal-ysis. After labeling, the cells were washed three times with PBS, scrapedinto Eppendorf tubes, placed on ice for 10 mm, and then centrifuged. The

supematants were carefully removed, and the cells were resuspended byvortexing in exactly 1 ml of PBS. i5Jiquots of 50 �zJ in dupl�ate wereremoved from each sample, centrifuged, and resuspended in 250 m� Tris(pH 8.0). The a-gal assays were then performed according to Promega.The remaining cell pellets were frozen and the lysates were normalized to

�-ga1 activity prior to use.MetabOliC LabelIng. Exponentially growing Hut-78 T lymphocytes

were treated by addition of 10 ng/mI PMA and 1 � lonomycin or leftuntreated. Exponentially growing U937 cells were treated with 100

unltWmloflFN-’y. Attheappropriatetlme, treated and untreated cells werecentrifuged and washed once with PBS kept at 37#{176}Cand then resus-

pended in methionine-/cystelne-free RPMI (Life Technokgies, Inc.), sup-plemented as described above, except 2% dIalyzed FCS was used forlabeling. PMMonomycin or IFN-ywas also added to the labeling mediumat the concentrations indicated above. The cells, at 5 x 106 cells/mI, werestarved for 15 mm. [�SJMethionmne (New England Nuclear) was then

added to a final concentration of 100 �CVmI and the cells werelabeled for50 mm. Cells were harvested by centrifugation and washed once in

ice-cold PBS and frozen at -70#{176}C.For each of the indicated time points,cells were also harvested for Northern analysis. The cells were treated inan Identical manner for both Northern analysis and metabolic labeling,

with the exception that “cold” methionine was added in place of theradiolabeled methionine. NIH-3T3, 3T3-4E(P2), or 3T3-4E(Ala) cells wereplated at cell densities of 5 x 10� per 100-mm plate. Transfected ornontransfected cells were washed once in 37#{176}CPBS and “starved” inmethionine-/cystelne-free OMEM (Life Technologies, lnc.)for 15 mm. This

medium was then removed, and the cells were refed with 3 ml of methi-onine-/cystelne-deflcient OMEM containing 100 �CVmI [�S]mothionine

and labeled for 1 h. Cells were then harvested, and p-gal activity wasdetermined as described above.

Immunopreclpitations. Hut-78 and U937 cell lyses were performedon -1 x io� cells/lysate in approximately 300 �I of modified RIPA buffercontaining 250 mM NaCI, 1 % Nonidot P-40, 0.25% dooxycholate, and 50mM Tris (pH 8.0), plus the following protease inhibitors: 2 m� EOTA, 1 mp�iphenylmethylsulfonyl fluoride, 1 �g/ml leupeptmn, and 0.2 units/mI apro-

tinin for 30 mm on ice. Following centrifugation at 4#{176}C,the supematantswere removed, and SOS was added to a final concentration of 0.1 %.

Lysates were precleared by the addition of 6 �d of antiserum from an

unrelated rabbit and a250-pJ cell pellet of Pansorbin (Cal-Biochem)for2 hat 4#{176}C,with rotation. The lysato wasthen centrifuged, and the supernatantwas removed to a fresh tube. TCA precipitates were performed on au-

quots, in duplicate, and each lysate was normalized to equal amounts ofincorporated label (-5 x 10� cpm). Anti-recombinant Pim-1 serum (6 �I)or prelmmune serum (6 �.4J), plus 30 �sI of protoin-G Sepharose (Pierce),were added to each lysate for2 h at 4#{176}C,with rotation. NIH-3T3 cells werelysed, precleared, and incubated with antibody and protein-G, as de-scribed above for suspension cells, except cell concentration was -1 x

106. Cells transiently transfected were normalized to p-gal activity (seeabove). Immune complexes were collected with low-speed centrifugationand washed four times in lysis buffer and once in 25 m� Hepos (pH 7.5).Samples were subjected to 10% SOS-PAGE. The gel was stained, placedIn fluor, and dried prior to autoradiography. Film was scanned using anImagequant densitometer (Molecular Dynamics).

Western Blots. 3T3, 3T3-4E(P2), and 3T3-4E(AIa) cells were lysed inTris buffer [20 m� Tris (pH 7.5), 150 m� KCI, 1 m� O1T, and 1 m�i EOTA)by three freeze-thaw cycles, and following centrifugation at 4#{176}Cproteinconcentrations in the supematants were quantitatod using the Bio-Radprotein assay. Total cellular protein (75 gig) from each cell line was as-sayed by SOS-PAGE analysis. The proteins were transferred to nitrocel-lulose. Blots were blocked for 2 h in 2% powdered milk-0.05% Tween 20in PBS and then incubated with anti-elF-4E (a gift of Or. N. Sonenberg) ata 1:2000 dIlution in PBS-Tweon 20 for 2 h. Blots were washed three timesfor 30 mm each in PBS-Tween 20. The secondary antibody goat antirab-bit, conjugated to horseradish peroxidase (BioRad), was used at a 1:6000dilution in PBS and applied for 1 h. Blots were washed, and immunecomplexes were detected using enhanced chemiluminescence (Amor-

sham). Films wore scanned using an Imagequant densitometer.

Acknowledgments

We thank Or. N. Sonenberg for the gift of the 3T3-4E(P2) and 3T3-4E(Ala)

cells and eIF-4E antibody. We also thank Or. Glenn Cantor for the criticalreading of the manuscript, Or. Sushma Ogram for the thoughtful discus-sions, and Steve Thompson and the Center for Visualization, Analysis, andDesign in the Molecular Sciences lab for sequence analysis.

References1 . Saris, C. J., Domen, J., and Bems, A. The pim-1 oncogene encodestwo related protein-sermno/threonine kinases by alternative initiation at

AUG and CUG. EMBO J., 10: 655-664, 1991.

2. Padma, R., and Nagarajan, L The human PIM-1 gene product is aproteln setine kinase. Cancer Res., 51: 2486-2489, 1991.

3. Hoover, 0., Fnedmann, M., Reeves, R., and Magnuson, N. S. Recom-binant human pim-1 proteln exhibits sermno/threonine kinase activity.J. Biol. Chem., 266: 14018-14023, 1991.

4. Bems, A., Cuypers, H. T., Selten, G., and Domen, J. In: N. 0. Kjeldaardand J. Forchhammer (eds.), Viral Carcinogenesis, pp. 21 1-230. Copen-hagen: Raven Press, 1987.


5. Amson, R., Sigaux, F., Przedborski, S., Flandrin, G., Givol, 0., and

Telerman, A. The human protooncogene product p33pim Is expressedduring fetal hematopolosis and in diverseleukemias. Proc. NatI. Acad. Sd.

USA, 86: 8857-8861 , 1989.

6. Sorrentino, V., McKinney, M. 0., Giorgi, M., Geremia, R., and Fleissner,

E. Expression of cellular protooncogenes in the mouse male germ line: a

distinctive 2.4-kilobase pim-1 transcript is expressed in haplold postmel-

otic cells. Proc. NatI. Aced. Scm. USA, 85: 2191-2195, 1988.

7. van Lohuizen, M., Verbeek, S., Krimpenfort, P., Oomen, J., Saris, C.,Radaszkiewicz, T., and Bems, A. Predisposition to lymphomagenesis in

pim-1 transgenic mice: cooperation with c-myc and N-myc in munneleukemia virus-Induced tumors. Cell, 56: 673-682, 1989.

8. Breuer, M., Wlentjens, E., Verbeek, S., Slebos, R., and Barns, A.Carcinogen-induced lymphomagenesis in pim-1 transgenlc mice: dosedependence and involvement of myc and ras. Cancer Res., 51: 958-963,

1991.

9. Acton, 0., Domen, J., Jacobs, H., Vlaar, M., Korsmeyer, S., and Bems,A. Collaboration of pim-1 and bcl-2 in lymphomagenesis. Curr. Top.Microbiol. Immunol., 182: 293-298, 1992.

10. Dautry, F., Weil, 0., Yu, J., and Oautry Varsat, A. Regulation of pimand myb mRNA accumulation by interleukin 2 and interleukin 3 in murinehematopoietic cell lines. J. Biol. Chem., 263: 17615-17620, 1988.

11 . Lilly, M., Le, T., Holland, P., and Hendrickson, S. L Sustained ex-

pression of the pim-1 kinase is specifically induced in myeloid cells bycytokines whose receptors are structurally related. Oncogene, 7: 727-732, 1992.

12. Sato, N., Sakamaki, K., Terada, N., Arai, K., and Miyajima, A. Signaltransduction by the high-affinity GM-CSF receptor two distinct cytoplas-mic regions of the common � subunit responsible for different signaling.EMBO J., 12: 4181-4189, 1993.

13. Oomen, J., van der Lugt, N. M., Laird, P. W., Saris, C. J., Clarke, A. R.,Hooper, M. L, and Barns, A. Impaired intetleuldn-3 response in Pim-1-deficient bone marrow-derived mast cells. Blood, 82: 1445-1452, 1993.

14. Quelle, F. W., Sato, N., Wltthuhn, B. A., Inhom, R. C., Eder, M.,

Miyajima, A., Griffin, J. 0., and lhle, J. N. JAK2 associates with the � cchain of the receptor for granulocyte-macrophage colony-stimulating fac-tor, and its activation requires the membrane-proximal region. Mol. Cell.Biol., 14: 4335-4341, 1994.

15. Miura, 0., Miura, V., Nakamura, N., Quelle, F. W., Witthuhn, B. A., IhIe,J. N., and Aoki, N. Induction of tyrosine phosphorylation of Vav andexpression of Pim-1 correlates with Jak2-mediated growth signaling fromthe erythropoietin receptor. Blood, 84: 4135-4141 , 1994.

16. Yip Schneider, M. T., Horie, M., and Broxmeyer, H. E. Transcriptional

induction of pim-1 protein kinase gene expression by interferon y andposttranscriptional effects on costimulation with steel factor. Blood, 85:3494-3502, 1995.

17. Buckley, A. R., Buckley, 0. J., Leff, M. A., Hoover, 0. 5., and

Magnuson, N. S. Rapid induction of pim-1 expression by prolactmn andinterleukin-2 in rat Nb2 lymphoma cells. Endocrinology, 136: 5252-5259,

1995.

18. Moroy, T., Grzeschiczek, A., Petzold, S., and Hartmann, K. U. Ex-

pression of a Pim-1 transgene accelerates lymphoproliferation and inhib-its apoptosis in lpr/Ipr mice. Proc. NatI. Acad. Sd. USA, 90: 10734-10738,1993.

19. Takahashi, C., Harada, Y., Ariga, H., and Iguchi Ariga, S. M. Involve-ment of PIM-1 in DNA fragmentation in mouse NS-1-detived cells. Blo-

chem. Biophys. Res. Commun., 215: 538-546, 1995.

20. Meeker, T. C., Loeb, J., Ayres, M., and Sellers, W. The human Pim-1gene is selectively transcribed in different hemato-lymphoid cell lines inspite of a G+C-rich housekeeping promoter. Mol. Cell. Biol., 10: 1680-

1688, 1990.

21 . Wingett, 0., Reeves, R., and Magnuson, N. S. Stability changes inpim-1 proto-oncogene mRNA after mitogen stimulation of normal lym-

phocytes. J. Immunol., 147: 3653-3659, 1991.

22. Uang, H., Hittelman, W., and Nagarajan, L Ubiquitous expression andcellcydle regulation ofthe protein kinase PIM-1 . Arch. Biochem. Biophys.,330: 259-265, 1996.

23. Wingett, 0., Long, A., Kelleher, 0., and Magnuson, N. S. pim-1 proto-oncogene expression in anti-CO3-mediated T cell activation is associatedwith protein kinase C activation and is independent of Raf-1 . J. Immunol.,156: 549-557, 1996.

24. Rohwer, F., Todd, S., and McGuire, K. L The effect of IL-2 treatmenton transcriptional attenuation in protooncogenes pim-1 and c-myb inhuman thymic blast cells. J. Immunol., 157: 643-649, 1996.

25. Kozak, M. An analysis of vertebrate mRNA sequences: intimations oftranslational control. J. Cell. Blol., 115: 887-903, 1991.

26. Kozak, M. Structural features in eukaryotic mRNAs that modulate the

initiation of translation. J. Biol. Chem., 266: 19867-19870, 1991.

27. Geballe, A. P., and Morris, 0. R. Initiation codons within 5’-leaders ofmRNAs as regulators of translation. Trends Biochem. Sci., 19: 159-164,

1994.

28. Avni, 0., Shama, S., Loreni, F., and Meyuhas, 0. Vertebrate mRNAswith a 5’-terminal pyrimidine tract are candidates for translational repres-slon in quiescent cells: characterization of the translational cis-regulatoryelement. Mol. Cell. Biol., 14: 3822-3833, 1994.

29. Pelletier, J., and Sonenberg, N. Insertion mutagenesis to increasesecondary structure within the 5’ noncoding region of a eukaryotic mRNAreduces translational efficiency. Cell, 40: 515-526, 1985.

30. Rao, C. 0., Pech, M., Robbins, K. C., and Aaronson, S. A. The 5’untranslated sequence of the c-sis/platelet-derived growth factor 2 tran-script is a potent translational inhibitor. Mol. Cell. Biol., 8: 284-292, 1988.

31 . Marth, J. D., Overell, A. W., Meier, K. E., Krebs, E. G., and Perlmuttor,A. M. Translational activation of the Ick proto-oncogeno. Nature (Lond.),332: 171-173, 1988.

32. Manzella, J. M., and Blackshear, P. J. Regulation of rat omithinedecarboxylase mRNA translation by its 5’-untranslated region. J. Biol.Chem., 265: 11817-11822, 1990.

33. Bates, B., Hardin, J., Than, X., Drickamer, K., and Goldfarb, M.

Biosynthesis of human fibroblast growth factor-5. Mol. Cell. Biol., 11:1840-1845, 1991.

34. Arrick, B. A., Lee, A. L, Grendell, R. L, and Derynok, R. Inhibition oftranslation of transforming growth factor-a 3 mRNA by its 5’ untranslatedregion. Mol. Cell. Blol., 11: 4306-4313, 1991.

35. Fagan, R. J., Lazaris Karatzas, A., Sonenberg, N., and Rozen, R.Translational control of ornithine aminotransferase. Modulation by Initia-tion factor elF-4E. J. Biol. Chem., 266: 16518-16523, 1991.

36. Shantz, L M., and Pegg, A. E. Overproduction of omithine decarbox-ylase caused by relief of translational repression is associated with nec-plastic transformation. Cancer Res., 54: 2313-2316, 1994.

37. Ruan, H., Hill, J. R., Fatemie Nainie, S., and Morris, 0. R. Cell-specifictranslational regulation of S-adenosylmethionine decarboxylase mRNA.Influence of the structure of the 5’ transcript leader on regulation by the

upstream open reading frame. J. Biol. Chem., 269: 17905-17910, 1994.

38. Kevil, C., Carter, P., Hu, B., and DeBenedetti, A. Translational en-

hancement of FGF-2 by elF-4 factors, and alternate utilization of CUG and

AUG codons for translation initiation. Oncogene, 11: 2339-2348, 1995.

39. Mizokami, A., and Chang, C. Induction of translation by the 5’-un-translated region of human androgen receptor mRNA. J. Biol. Chem., 269:25655-25659, 1994.

40. Hershey, J. W. Translational control in mammalian cells. Annu. Rev.Biochem., 60: 717-755, 1991.

41 . Sonenberg, N. Cap-binding proteins of eukaryotlc messenger RNk

functions in initiation and control of translation. Prog. Nucleic Acid Res.Mel. Biol., 35: 173-207, 1988.

42. Sonenberg, N., Morgan, M. A., Mervick, W. C., and Shatldn, A. J. Apolypeptide in eukaryotic initiation factors that crosslinks specifically tothe 5-terminal cap in mRNA. Proc. NatI. Aced. Scm. USA, 75: 4843-4847,

1978.

43. Pause, A., M’ethot, N., Svitkin, V., Merrick, W. C., and Sonenberg, N.

Dominant negative mutants of mammalian translation initiation factor

eIF-4A define a critical role for eIF-4F in cap-dependent and cap-inde-pendent initiation of translation. EMBO J., 13: 1205-1215, 1994.


44. Pause, A., and Sonenberg, N. Mutational analysis of a DEAD box RNAhelicase: the mammalian translation Initiation factor eIF-4A. EMBO J., 11:2643-2654, 1992.

45. Grifo, J. A., Tahara, S. M., Leis, J. P., Morgan, M. A., Shatkin, A. J.,and Merrick, W. C. Characterization of eukaryotic initiation factor 4A, aprotein involved in ATP-dependent binding of globin mRNA. J. Blol.Chem., 257: 5246-5252, 1982.

46. Ray, B. K., Lawson, T. G., Kramer, J. C., Cladaras, M. H., Grifo, J. A.,Abramson, R. 0., Memck, W. C., and Thach, R. E AlP-dependent un-winding of messenger RNA structure by eukaryotic initiation factors.J. Biol. Chem., 260: 7651-7658, 1985.

47. Rousseau, 0., Kaspar, R., Rosenwald, I., Gehrke, L, and Sonenberg,N. Translation initiation of omithine decarboxylaso and nudleocytoplasmictransport of cyclin 01 mRNA are increased in cells overexpressing eu-

karyotic initiation factor 4E. Proc. NatI. Aced. Sci. USA, 93: 1065-1070,

1996.

48. OamelI, J. E, Jr., Kerr, I. M., and Stark, G. R. Jak-STAT pathways andtranscriptional activation in response to IFNs and other extracellular sig-naling proteins. Science (Washington, DC), 264: 1415-1421 , 1994.

49. Reeves, R., Spies, G. A., Kiefer, M., Barr, P. J., and Power, M. Primarystructure of the putative human oncogene, pim-1. Gene, 90: 303-307,1990.

50. Kozak, M. An analysis of 5’-noncoding sequences from 699 verte-brate messenger RNAs. Nucleic Acids Res., 15: 8125-8148, 1987.

51 . Zuker, M., and Stiegler, P. Optimal computer folding of large RNAsequences using thermodynamics and auxiliary information. Nucleic Ac-ids Res., 9: 133-148, 1981.

52. Kozak, M. Influences of mRNA secondary structure on initiation byeukaryotic ribosomes. Proc. NatI. Aced. Sci. USA, 83: 2850-2854, 1986.

53. Koromilas, A. E, Lazaiis Karatzas, A., and Sonenberg, N. mRNAscontaining extensive secondary structure in their 5’ non-coding regiontranslate efficiently in cells overexpressing initiation factor elF-4E (Pub-

lished erratum in EMBO J., 11: 5138, 1992). EMBO J., 11: 4153-4158,

1992.

54. Kozak, M. Recognition of AUG and alternative initiator codons is

augmented by G in position +4 but is not generally affected by thenucleotides in positions +5 and +6. EMBO J., 16: 2482-2492, 1997.

55. Genetics ComputerGrouplnc. SequenceAnalysis Software Program,Version 7. MadIson, WI: Genetics Computer Group Inc., 1991.

56. Rosenwald, I. B., Kaspar, R., Rousseau, 0., Gehrke, L, Leboulch, P.,Chen, J. J., Schmidt, E. ‘I., Sonenberg, N., and London, I. M. Eukaryotictranslation initiation factor 4E regulates expression of cyclin 01 at tran-scnptional and post-transcriptional levels. J. Biol. Chem., 270: 21176-21180, 1995.

57. Lazaris Karatzas, A., Montine, K. S., and Sonenberg, N. Malignanttransformation by a eukaryotic initiation factor subunit that binds to mRNA

5’ cap. Nature (Lend.), 345: 544-547, 1990.

58. Rosenwald, I. B., Rhoads, 0. B., Callahan, L 0., Isselbacher, K. J.,and Schmidt, E V. Increased expression ofeukaryotictranslation Initiation

factors eIF-4E and eIF-2 a in response to growth induction by c-myc.Proc. NatI. Aced. Sci. USA, 90: 6175-6178, 1993.

59. Ray, B. K., Brendler, T. G., Adya, S., Daniels McQueen, S., Miller,J. K., Hershey, J. W., Grifo, J. A., Merrick, W. C., and Thach, R. E. Role ofmRNA competition in regulating translation: further characterization ofmRNA discriminatory initiation factors. Proc. NatI. Acad. Sci. USA, 80:

663-667, 1983.

60. Mendoz, R., Myers, M. G., Jr., White, M. F., and Rhoads, R. E.Stimulation of protein synthesis, eukaryotic translation initiation factor 4Ephosphorylation, and PHAS-I phosphorylation by insulin requires insulinreceptor substrate 1 and phosphatidylinositol 3-kinase. Mol. Cell. BioI.,

16: 2857-2864, 1996.

61 . Rosenwald, I. B., Lazaris Karatzas, A., Sonenberg, N., and Schmidt,E. V. Elevated levels of cyclin 01 protein in response to increased expres-

sin of eukaryotic initiation factor 4E. Mol. Cell. Biol., 13: 7358-7363,1993.

62. Breuer, M., Slebos, R., Verbeek, S., van Lohuizen, M., Wientjens, E.,and Bems, A. Very high frequency of lymphoma induction by a chemicalcarcinogen in pim-1 transgenic mice. Nature (Lond.), 340: 61-63, 1989.

63. Do Benedetti, A, and Rhoads, R. E. Overexpression of eukaryotic

protein synthesis initiation factor 4E in HeLa cells results in aberrantgrowth and morphology. Proc. NatI. Acad. Sd. USA, 87: 8212-8216,1990.

64. Klein, P. 5., and Melton, 0. A. Induction of mesoderm in XenopusIaevis embryos by translation initiation factor 4E. Science (WashingtonDC), 265: 803-806, 1994.

65. Brunn, G. J., Williams, J., Sabers, C., Wiedorrecht, G., Lawrence,J. C., Jr., and Abraham, R. T. Direct inhibition of the signaling functions ofthe mammalian target of rapamycin by the phosphoinositido 3-kinaseinhibitors, wortmannin and LY294002. EMBO J., 15: 5256-5267, 1996.

66. Bowlin, T. L, McKown, B. J., and Sunkara, P. S. Omithine decarbox-

ylase induction and polyamine biosynthesis are required for the growth ofinterleukin-2- and interleukin-3-dependent cell lines. Cell. Immunol., 98:341-350, 1986.

67. Bercovich, 1, Rosenberg-Hasson, Y., Ciechanover, A., and Kahana,C. Degradation of omithine decarboxylase in reticulocyte lysate is ATP-dependent but ubiquitin-independent. J. Biol. Chem., 264: 15949-15952,1989.

68. Sherr, C. J., Matsushime, H., and Roussel, M. F. In: J. Marsh (ed).Regulation of the Eukaryotic Cell Cycle, pp. 209-226. West Sussex: JohnWiley & Sons, 1992.

69. Haystead, T. A., Haystead, C. M., Hu, C., Un, T. A., and Lawrence,J. C., Jr. Phosphorylation of PHAS-I by mitogen-activated protein (MAP)

kinase. Identification of a site phosphorylated by MAP kinase in vitro andin response to insulin in rat adipocytes. J. Biol. Chem., 269: 231 8�-231 91,1994.

70. Un, T. A., Kong, X., Haystead, T. A., Pause, A., Belsham, G.,Sonenberg, N., and Lawrence, J. C., Jr. PHAS-I as a link between mito-gon-activated protein kinase and translation initiation. Science (Washing-ton DC), 266: 653-656, 1994.

71 . Rousseau, 0., Gingras, A. C., Pause, A., and Sonenberg, N. TheeIF4E-binding proteins 1 and 2 are negative regulators of cell growth.Oncogene, 13: 2415-2420, 1996.

72. Beretta, L, Gingras, A. C., Svitkin, V. V., Hall, M. N., and Sonenberg,N. Rapamycin blocks the phosphorylation of 4E-BP1 and inhibits cap-dependent initiation of translation. EMBO J., 15: 658-664, 1996.

73. Hutchcroft, J. E., Franklin, 0. P., Tsai, B., Harrison Findik, 0.,Varticovski, L, and Bierer, B. E. Phorbol ester treatment inhibits phos-phatidylinositol 3-kinase activation by, and association with, CO28, a

T-Iymphocyte surface receptor. Proc. NatI. Acad. Sci. USA, 92: 8808-8812, 1995.

74. Harlow, E, and Lane, 0. Antibodies: A Laboratory Manual, pp. 53-

135. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, Now York.1988.

75. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, 0. 0., Seidman,J. G., Smith, J. A., and Struhl, K. (eds.). Current Protocols in MolecularBiology, pp. 3.1.1-4.10.9. New York: John Wiley & Sons, 1988.

76. Sambrook, J., Frisch, E F., and Maniatis, T. Molecular Cloning: ALaboratory Manual, pp. 61-207. Cold Spring Harbor, NY: Cold SpringHarbor Laboratory, 1989.

77. Corsaro, C. M., and Pearson, M. L Enhancing the efficiency of DNA-mediated gene transfer in mammalian cells. Somatic Cell Genet., 7:603-616, 1981.

Pim1 Protein Expression Is Regulated by Its 5'-Untranslated Region and Translation Initiation Factor eIF-4E1

Documents