CL- 2~ 7-

JOURNAL OF SURGICAL RESEARCH 56, 500-504 (1994)

Immediate-Early Gene Expression in EGF-Stimulated IntestinalEpithelial Cells

RICHARD A. HODIN, M.D., SHUFEN MENG, M.D., AND DEANNA NGUYEN

Department o{ Surgery, Beth Israel Hospital, Haruard Medical School, Boston, Massachusetts 02215

Presented at the Annual Meeting of the Association for Academic Surgery, Hershey, Pennsylvania, November 10-13, 1993

The mammalian small intestinal epithelium is a con-tinuously renewable population of cells which arisefrom a proliferative zone of undifferentiated stem cellswithin the crypts. Epidermal growth factor (EGF) isthought to playa role in the maintenance and prolifera-tive response of the intestinal epithelium. In order toinvestigate the early changes in intestinal gene expres-sion which occur in response to a mitogenic stimulus,we performed studies in two different celllines (IEC-6and HT-29), both of which have characteristics of in-testinal crypt cells. Cells were grown in DMEM + 10%FCS at 37°Cf5% CO2 and treated with cithcr EGF (10'ngfml) andfor cycloheximide (CHx) (10 p.gfml) forvarious times. Northcrn blot analyscs were pcrformedon total RNA using 32P-Iabelcd cDNA probcs corrc-sponding to various protooncogencs. Our resultsshowcd that EGF treatment causcd rapid incrcascs inc-fos, c-jun, and junB cxpression (P < 0.001) in bothccIllincs. c-fos and c-jun foIlowed similar time courses,pcaking at 30 min, whcrcasjunB levcls plateaucd at 1hr. Thc magnitude and time course of protooncogencinduction by EGF wcre similar in thc two ceIllincs. Adose-responsc cxperiment indicated peak EGF cffcctsat 10 ngfml. CHx treatment resulted in grcater andmore prolonged incrcases in protooncogcne expressionwhen compared to EGF alonc, indicating that proteinsynthesis is not required for protooncogene induction.These results suggest that c-fos, c-jun, andjunB are aIlpart of thc "immcdiate-early" genomic response toEGF within the intestinc, and that both IEC-6 and HT-29 ceIls may be suitable modcls to study the molecularmcchanisms underlying intestinal epithelial growth.@ 1994 Academic Press, Inc.

INTRODUCTION

The mammalian small intestine is lined by a simplecolumnar epithelium which forms finger-like projec-tions, termed villi, that extend into the lumen and pro-vide a large surface afea for absorptive functions. Theepithelium is a continuously renewable population ofcells which arise from a proliferative zone ofundifferen-

tiated stem cells within the crypts. The cells acquire adifferentiated phenotype, i.e., enterocytes, enteroendo-crine, or goblet, as they migrate towards the villus tip,eventually to be extruded into the lumen. Paneth cells,whose function is not well understood"tend to migratein the other direction to reside at the base of the crypts[1,2].

Renewal of the intestinal epithelium is required tomaintain the functional integrity of the mucosa and per-tubations in this normal process of cellular growth arecommon to many pathologic conditions including non-tropical sprue, Crohn's disease, radiation enteritis, smallbowellymphoma, and prolonged fasting. The abnormali-ties in intestinal epithelial growth lead to villus atrophyand its associated morbidity, including diarrhea, malab-sorption, and perhaps impairment in gut barrier func-tion [3]. Although various luminal and hormonal factorshave been shown to affect the growth state of the smallintestine, the molecular mechanismsunderlying this pro-cess are not well understood [4].

The growth of mammalian cells is in large part me-diated by polypeptide growth factors which bind to cellsurface receptors and trigger a complex cascade of cyto-plasmic and nuclear events. Second messenger systemstransmit a signal from the cell membrane into the nu-cleus, resulting in alterations in gene expression. Studiesutilizing mitogen-stimulated cells or tissues have eluci-dated the existence of two classes of genes whose expres-sion is increased at various times following exposure tothe growth signal. The "immediate-early" genes are in-duced within minutes following growth factor stimula-tion, and this increased expression occurs independentof protein synthesis. The so called "delayed-early"genes are induced hours following a growth stimulusand, in contrast to the immediate early genes, proteinsynthesis inhibition blocks the increase in delayed-earlygene expression [5, 6].

We have previously studied the growth response in therat small intestine using a fastingjrefeeding model anddemonstrated induction in the protooncogenes c-fos andc-jun shortly after the refeeding stimulus. Because thisincrease in protooncogene expression occurred indepen-

500

f"" ~ A:;..,..0022-4804/94 $5.00Copyright @ 1994 by Academic Press, Inc.Al! rights of reproduction in any for11l reserved.

¿-R-i.~ce..J~~ ~ -~~.:- M &"" ~F

~--= ~ :.- ~ ~ ~ L.:, ...~ ~ ~.4 e.~-"o lO? /--1

HODIN, MENG, AND NGUYEN: EARLY GENE EXPRESS ION IN INTESTINE 501

dent of protein synthesis, we concluded that c-fos andc-jun were immediate-early genes in the context of thesmall intestinal crypt cell hyperplasia that occurs withrefeeding [7]. Since it is difficult to study crypt cells inthe intact animal, the present studies were undertakento establish an in vitro system in which to study themolecular mechanisms governing intestinal epithelialproliferation.

c..Q".;~a.~uE.QIC:¡;E>.

.c1-

~METHODS Control EGF

FIG. 1. [3H]Thymidine incorporation in EGF-treated IEC-6cells. IEC-6 cells were treated with saline or 10 ng/ml EGF for 24 hrand then (3H]thymidine incorporation was measured. Depicted arethe mea n :t SEM (n = 4), in counts per minute, for control and EGF-treated cells. Statistical significance (P < 0.001) was determined usingStudent's unpaired t test.

[3 H] Thymidine 1 ncorporation

[3H]Thymidine (1 JlCi/ml medium) was added to thecells following a 24-hr treatment with either saline (con-trol) or 10 ng/ml epidermal growth factor (EGF), andincubation was carried out for 6 hr. The cells were thenwashed in ice-cold PBS and 5% TCA added for 30 min at4 °C. The cells were then washed with cold TCA followedby 95% ethanol, air-dried, and incubated with 0.5 NNaOH for 1 hr at 50°C. The cell lysates were then trans-ferred to scintillation vials for measurement of incorpo-rated radioactivity.

cpm/Jlg DNA. Conditions for hybridization were as fol-lows: 5x standard saline citrate (SSC)/50% (v/v) form-amide/l% (w/v) sodium dodecyl sulfa~e (SDS) at 42°C.Washing conditions were 2X SSC/O.l% SDS at 50°C.Relative changes in mRNA levels were determined bylaser densitometry of autoradiograms and normalizedfor actin mRNA. Statistical analyses were carried outusing Student's unpaired t test; P < 0.05 was consideredstatistically significant.

Cell Culture

IEC-6 and HT-29 cells were purchased from Ameri-can Type Culture Collection (ATCC) and grown in 160-cm2 plastic ftasks at 37°C/5% CO2 in Dulbecco's modi-fied Eagle's medium (Gibco) supplemented with 10% fe-tal bovine serum (Sigma), 2 mM L-glutamine, and 100U/mI penicillin/streptomycin. 1'he medium waschanged every 3 days and the cells split via trypsiniza-tion when they reached 80-90% conftuence. Experi-ments were performed on cells at 80% conftuence andthe medium was changed 24 hr prior to the start of eachexperimento The cells were treated with either 10 ng/mlEGF:t 10 J1.g/ml cycloheximide (CHx) or 5 J1.g/ml aniso-mycin (Anis) for periods of O, 0.5, 1, 4, or 24 hr. At theprescribed times, the medium was removed and the cellswere scraped in guanidinium thiocyanate (GTC) forRNA extraction.

RESULTS

[3 H] Thymidine 1 ncorporation

As shown in Fig. 1, a 24-hr treatment with EGFcaused an approximate 41 % increase in IEC-6 cellularproliferation (P < 0.001). Similar experiments were per-formed in HT-29 cells, algo revealing an increase in [3H]-thymidine incorporation in response to EG F. These mea-surements verified the mitogenic effects of EGF shownby others [16].

Northern Blot Analyses

The cells were homogenized in GTC and total RNAwas purified by ultracentrifugation through a CsCI gra-dient [8]. Northern blot analyses were then performedusing standard techniques [9]. The c-fos probe was a 0.5-kb NcoI/SalI fragment derived from the mouse cDNA[10]. The c-jun probe was a 0.7-kb PstI insert from themouse cDNA [11]. The junB probe was a 1.5-kb insertfrom the mouse cDNA [12]. The c-fos, c-jun, andjunBcDNA-containing plasmids were purchased fromATCC. The actin probe was a 1.0-kb PstI fragment de-rived from a {J-actin cDNA [13]. Complementary DNAprobes were 32p labeled by the method of Feinberg andVogelstein [14], typically to a specific activity of 5 X 108

Protooncogene Expression in IEC-6 Cells

EG F treatment of IEC-6 cells caused rapid and signifi-cant increases in both c-fos and c-jun expression at 30mino The le veis decreased markedly by 1 hr and returnedto baseline by 4 hr (Fig. 2). There was no difference inc-fos or c-jun mRNA levels between 4 and 24 hr of EGFtreatment (datanot shown). The magnitude ofthe EGF-induced c-jun increase was 1.6-fold (P < 0.001), and thec-fos increase could not be accurately quantified since itsmRNA is undetectable in the control state. The sameNorthern blots were also probed withjunB, demonstrat-ing that this protóoncogene isalso induced (fivefold, P <0.001) byEGF (data not shown). ThejunB levelspeakedat 1 hr, as opposed to the 30-min peak seen with c-fosand c-jun.

Figure 3 depicts the results of experiments using theprotein synthesis inhibitor CHx. These studies demon-

502 JOURNAL OF SURGICAL RESEARCH: VOL. 56, NO. 6, JUNE 1994

EGF (hrs) EGFCHxAnis

o 0.5 1 4 -+--+ +--+-+----+-+-- -

Jun B .. '...c-jun ---

Actin -

FIG.4. Effects of CHx and Anis on EGF-stimulated IEC-6 cells.Northern analyses were performed on IEC-6 cells treated with 10 ng/mI EGF :t 10 IJg/ml CHx or 5 IJg/ml Anis for 30 mino Probes specificfor thejunB and actin mRNAs are described under Methods. Twentymicrograms of total RNA wasloaded in each lane.

FIG.2. Protooncogene induction in IEC-6 cells. Northern analy-sea were performed on IEC-6 cells treated with 10 ng/ml EGF for theindicated times. Probes specific for the c-fos, c-jun, and actin mRNAsare described under Methods. Twenty micrograms of total RNA wasloaded in each lane.

(Anis). As illustrated in Fig. 4, neither CHx or Anisblocked the EGF-induced increase in junB expression.In fact, as was seen with CHx, Anis l~d to a superinduc-tion in junB mRNA levels. An EGF dose-response ex-periment was performed and revealed maximum effectsat the 10 ngfml clase (Fig. 5). The dose-response curvefor c-Ios and c-jun were identical to that shown for junB.

Protooncogene Expression in HT -29 Cells

All of the above studies were carried out in HT-29cells, in order to determine if this intestinal cellline re-sponded to EG F in the same way as the IEC-6 cells. Boththe EGF and CHx effects on c-fos, c-jun, andjunB wereseen in HT-29 cells, just as was the case in the IEC-6cells. Figure 6 depicts the effects on junB expression,illustrating the EGF induction, as well as the CHx su-perinduction.

strated that inhibition ofprotein synthesis did not blockprotooncogene induction, and, in fact, resulted ingreater increases in c-fos and c-jun expression thanthose seen with EGF alone. By itself, CHx causedmarked increases in c-fos and c-jun mRNA levels, butthese changes occurred later and were more prolongedthan those caused by EG F. In combination, CHx andEGF led to a superinduction in c-fos and c-jun expres-sion at 0.5, 1, and 4 hr. The effects of CHx on junBexpression were similar to those seen with c-fos and c-jun (data not shown).

In arder to verify that the CHx results were due toprotein synthesis inhibition, and not so me other nonspe-cific CHx effect, we performed another set of experi-ments using the protein synthesis inhibitor anisomycin

DISCUSSION

A number of different hormones and growth factorshave been shown to be trophic for the small intestinalmucosa, including gastrin, bombesin, and neurotensin

EGF

CHx

c-fos -

c-jun

--

Jun B -actin

--

FIG. 3. Effects of EGF and CHx on protooncogene expression inIEC-6 cells. Northern analyses were performed on IEC-6 cells treatedwith 10 ng/ml EGF and/or 10 ¡tg/ml CHx for the indicated times.Probes specific for the c-fos, c-jun, and actin mRNAs are describedunder Methods. The faint bands under the c-fos mRNA are due tononspecific hybridization with 18S ribosomal RNA. Twenty micro-grams of total RNA was loaded in each lane.

FIG. 5. EGF dose-response in IEC-6 cells. Northern analyseswere performed on IEC-6 cells treated with various doses of EGF, asindicated, for 30 mino Probes specific for the junB and actin mRNAsare described under Methods. Twenty micrograms of total RNA wasloaded in each lane.

~

HODIN, MENG, AND NGUYEN: EARLY GENE EXPRESSION IN INTESTINE 503

0.5 hr 1 hr 4hr

EGFCHx

-+-++-++-+--++-++-++

Jun B ---

Actin

FIG.6. Protooncogene expression in HT-29 cells. Northern anal-yses were performed on HT-29 cells treated with 10 ng/ml EGF and/or 10 "g/mi CHx for the indicated times. Probes specific for the junBand actin mRNAs are described under Methods. Twenty microgramsof total RNA was loaded ineach lane.

sponse to mitogens within the rat small in'testine. As isthe case in other tissues, e.g., serum-stimulated fibro-blasts [26] and regenerating liver [25], it is likely that anumber of other genes will be found to playa part in theintestinal mitogenic response.

Cycloheximide has previously been shown to superin-duce the expression ofthe immediate early genes, proba-bly through effects on both transcription as well asmRNA stability [27]. So, toa, we have found that eithercycloheximide or anisomycin alone significantly in-creases protooncogene expression, as well as augment-ing the effects of EGF.

Previous studies on early-response genes have usedserum-starved cells treated with Borne growth stimulus,either serum itself or a specific growth factor, whereasthe present studies were performed on cells maintainedin standard medium. It is likely t.hat the protooncogeneinduction in these intestinal cells may be even more dra-matic if the EGF treatment were to be given underserum-starved conditions. .

Since both the EG F -induced cellular proliferation andthe increased protooncogene expression that we haveseen in IEC-6 and HT-29 cells algo occur in other celltypes [28], the physiologic relevance of these changesremains to be determined. As is the case in regeneratingliver [29], it is likely that one or more intestine-specificgenes will be found to play an important role in the cryptcell hyperplastic response.

ACKNOWLEDGMENTS

This work was supporled by lhc Deparlmenl of Surgery. Belh IsraelHospilal, and a granl from lhc Harvard Digcslivc Diseases Cenler,Harvard Mcdical School (NIH Granl P30 DK34854). R. Hodin is anAGA Foundalion/Induslry Research Scholar.

REFERENCES

[15]. EG F is a well-characterized peptide that stimulatescellular proliferation in a number of different tissues andcell culture systems, including small intestine [16, 17].The presence of this growth factor in amniotic fluid, co-lostrum, mother's milk, and saliva has suggested a physi-ologically important rol e in gastrointestinal growth anddevelopment. The fact that adult jejunal crypt cells areable to bind, internalize, and degrade EGF has led to theproposition that EGF may playa role in the maintain-ance and proliferative response of the adult intestinal,epithelium [18]. Recent studies have suggested thatEGF may function in promoting epithelial restitutionfollowing injury [19], and a role in reducing the atrophycaused by fasting has been suggested [20].

Based upon our demonstration in intact rats thatsmall intestinal c-fos and c-jun mRNAs were induced byrefeeding [7], we wanted to establish an in vitro modelsystem in which to study the mechanisms governingcrypt cell hyperplasia. The present studies employed therat jejunal crypt-derived IEC-6 cells [21] and the humancolon cancer-derived HT-29 cells [22], both ofwhich arethought to closely resemble small intestinal crypt cells.Since EGF has trophic effects upon intestinal epitheliaboth in vivo and in vitro, we used this growth factor tostudy the crypt cell genomic response to a mitogenicstimulús.

The protooncogenes, c-fos and c-jun, which comprisethe well-characterized AP -1 transcription factor, arethought to play an important role in the growth anddifferentiation of a number of tissues [23]. The expres-sion of cotos and c-jun is increased in many different celltypes upon exposure to various growth factors [24, 25].The increases in c-fos and c-jun expression that we havedemonstrated in EGF-stimulated intestinal cells areconsistent with the presumed role that these protoonco-genes play in cell replication.

The fact that c-fos, c-jun, and junB expression is in-creased very early following EGF, and that the increasesare not inhibited by cycloheximide, confirm that theseprotooncogenes are part of the immediate-early re-

1. Leblond, C. P., and Stevens, C. E. The constant renewal of theintestinal epithelium in the albino rato Anat. Rec. 100: 357, 1948.

2. Gordon, J. l. Intestinal epithelial cell differentiation: New in-sights from chimeric and transgenic mice. J. Cell Biol. 108: 1187,1989.

3. Earnest, D. L. Gastrointestinal mucosal changes from starvationand stress. The Eighth Ross Conference On Medical Research1988. P. 11.

4. Williamson, R. C. N. Intestinal adaptation (Parts I & 11). N.Engl. J. Med. 298(25): 1393/298(26): 1444, 1978.

5. Rozengurt, E. Early signals in the Mitogenic Response. Science234: 161, 1986.

6. Lanahan, A., Williams, J. B., Sanders, L. K., and Nathans, D.Growth factor-induced delayed early response genes. Mol. Cel/Biol. 12: 3919, 1992.

7. Hodin, R. A., Graham, J. R., Meng, S., and Upton, M. P. Tem-poral pattern of rat small intestinal gene expression with refeed-ing. Am. J. Physiol. 266: G83, 1994.

8. Chirgwin, J. M., Przybla, A. E., MacDonald, R. J., and Rutter,W. J. Isolation of biologically active ribonucleic acid fromsources enriched in ribonuclease. Biochemistry 18: 5294, 1980.

504 JOURNAL OF SURGICAL RESEARCH: VOL. 56, NO. 6, JUNE 1994

20. Bragg, L. E., Hollingsed, T. C., and Thompson,J. S. Urogastronereduces gut atrophy during parenteral alimentation. J. Parenter.Enteral Nutr. 14: 283, 1990.

21. Quaroni, A., Wands, J., Trelstad, R. L., and Isselbacher, K. J.Epitheloid cell cultures from rat small intestine. J. CeU Biol. 80:248, 1979.

22. Le Bivic, A., Hirn, M., and Reggio, H. HT-29 cel\s are an in vitromodel for the generation of cel\ polarity in epithelia during em-bryonic differentiation. Proc. Natl. Acad. Sci. USA 85: 136,1988.

23. Curran, T., Franza, R., Jr. ros and Jun: The AP-1 connection.CeU 55: 395, 1988.

24. Almendral, J. M., Sommers, D., Macdonald-Bravo, H., Burck-hardt, J., Perea, J., and Bravo, R. Complexity of the early geneticresponse to growth factors in mouse fibroblasts. Mol. CeU Biol. 8:2140, 1988.

25. Mohn, K. L., Laz, T. M., Melby, A. E., and Taub, R. Immediate-early gene expression differs between regenerating liver, insulin-stimulated H-35 cel\s, and mitogen-stimulated BALB/c 3T3cel\s. J. Biol. Chem. 265(35): 21914, 1990.

26. Lau, L. F., and Nathans, D. Expression of a set ofgrowth-relatedimmediate-early genes in BALB/c 3T3 cel\s: Coordinate regula-tion with c-fos or c-myc. Proc. Nat' Acad. Sci. USA 84: 1182,1987.

27. Edwards, D. R., and Mahadevan, L. C. Protein Rynthesis inhibi-tara differential\y superinduce c-fo.o¡ and c-jun by three distinctmecbanisms: Lack of evidence for labile repressorR. EMBO J.11(7): 2415, 1992.

28. Quantin, B., and Breathnllch, n. Epidermal ~rowth fm:tor Rtimu-lates transcription 01' the c-jun prolo-onco~ene in rlll fibrollln"l".Nature 334 (11): 1988.

29. Mobn, K. L., Laz, T. M., \-I"u, ,J-C., Melliy, A. r~., llfllvo, Il..,nllclTaub, R. The immedinle-enrly ¡:rowth re"pon"e in re¡:eneralin~liver and inRulin-stimulnled \-1-3:, cell,,: (~ompnri"on wilh Rcrum.stimululed 3'1':1 ccl\" und iclentificutilm of 41 novcl immeclillle-curly¡:cnc8.M'II.Cc'UlJi'll.ll(I):381,1!)!)I.

9. Maniatis, J., Fritsch, E. F., and Sambrook, J. Molecular Cloning:A Laboratory Manual. Cold Spring Harbor Laboratory. ColdSpring Harbor, NY, 1982. P. 545.

10. Miller, A. D., Curran, T., and Yerma, l. M. coros protein caninduce cellular transformation: A novel mechanism of activationof a cellular oncogene. Cell 36(1): 51, 1984.

11. Bohmann, D., Bos, T. J., Admon, A., Nishimura, T., Vogt, P. K.,and Tjian, R. Human proto-oncogene c-jun encodes DNA bind-ing protein with structural and functional properties of tran-scription factor AP-1. Science 238: 1386, 1987.

12. Ryder, K., Lau, L. F., and Nathans, D. A gene activated bygrowth factors is related to the oncogene v-jun. Frac. Natl. Acad.Sci. USA 85(5): 1487, 1988.

13. Cleveland, D. W., Lopata, M. A., MacDonald, R. J., Cowan, N. J.,Rutter, W. J., and Kirschner, M. W. Number and evolutionaryconservation of alpha- and beta-tubulin and cytoplasmic beta-and gamma-actin genes using specific clones cDNA probes. Cel/20(1): 95,1980.

14. Feinberg, A. P., and Vogelstein, B. A Technique for radiolabelingDNA restriction endonuclease fragment to high specific activity.Anal. Biochem. 132: 6, 1983.

15. Evers, B. M., Izukura, M., Townsend, C. M., Uchida, T., andThompson, J. C. Differential effects of gut hormones on pancre-atic and intestinal growth during administration of an elementaldiet. Ann. Surco 211(5): 630,1990.

16. Caben, S. The epidermal growth factor (EGF). Cancer 51: 1787,1983.

17. Ulshen, M. H., Lyn-Cook, L. E., and Roasch, R. H. EffectR ofintraluminal epidermal growth factor on mucosal proliferationin the small intestine of adult rats. Ga.~troenterolugy 91: 1134,1986.

18. I~ebenthal, E. (Ed.) Human Ga.'1trointe.'1tinal Development. Ra.ven Press, New York, 1989.

19. Wright, N. A., Pike, C., and Elia, G. Induction of a novel epider-mal growth factor-secreting celllineage by mucosal ulceration inhuman gnRtruintestinnl stem cells. Nature 343: 82, 1990.