man colon cancer cells were studied. DOX and INDO were used atclinically achievable concentrations. Beta-catenin levels were deter-mined by western blotting. Beta-catenin-dependent transcriptionwas measured by TOPFlash reporter assay. Cell viability was deter-mined by tiazolyl blue (MTT) reduction and growth was determinedby direct cell counting. In vivo, tumor growth was determined bymeasurement of subcutaneous xenografts in BALB/c nude mice.DOX was administered in drinking water, and INDO by IP injection.Results: In vitro, both DOX and INDO exposure result in a dosedependent down-regulation of beta-catenin protein and beta-catenin-dependent gene transcription, resulting in inhibition of cell viabilityand proliferation. A synergistic relationship of combining these twoagents is evident. In vivo studies show an enhanced anti-tumor effectof combined DOX and INDO treatment on SW480 xenograft growth(Figure 1). Conclusions: DOX and INDO suppress beta-cateninexpression and transcription activity in colon cancer cells, inhibitingcancer cell growth in vitro and in vivo. DOX and INDO may be ofclinical use in therapeutic or chemopreventative strategies for coloncancer.

41. Serum Response Factor is Alternatively Spliced in ColonCancer. L. C. Patten, M.D., N. S. Belaguli, Ph.D., M. Baek,M.D., D. H. Berger, M.D. Baylor College of Medicine, Houston,TX.

Introduction: Serum response factor (SRF) is a transcriptionfactor important in cellular differentiation and cell cycle regulation.SRF function is regulated in part by alternative splicing. Little isknown about the expression or role of these alternatively splicedforms during tumorigenesis. We hypothesized that there is a changein expression of splice variants during intestinal tumorigenesis, andthat this change in splice variant expression promotes the tumorphenotype. Methods: The expression of SRF was determined bywestern blotting of normal intestinal cells and human colon cancercell lines. To determine the effect of alternatively spliced forms ofSRF on intestinal growth and proliferation, the major alternativelyspliced isoform of SRF (SRF‚5) seen in cancer cells was transfectedinto IEC-6 cells. IEC-6 and IEC-6SRF‚5 cells were plated on matri-gel on day 0. Cell numbers were determined at four timepoints.Results: Western blotting demonstrates that full length SRF is thepredominant form of SRF in IEC-6 cells and the well-differentiatedhuman colon cancer cell line HT-29. In poorly differentiated humancolon cancer cells (WiDr, HCT 116, LoVo, and SW480) SRF‚5 is thedominant form expressed (Figure 1). There was a marked increase incell growth kinetics in IEC-6 cells transfected with SRF‚5 comparedto the parental cells (Figure 2). Conclusion: This data demonstratesthat an alternatively spliced form of SRF, SRF‚5 is expressed inhuman colon cancer cell lines. Additionally, this data demonstratesthat expression of SRF‚5 may contribute to the tumor phenotype byaffecting cell growth. This is the first study to document a change inalternative splicing of SRF in human malignancy.

RESIDENT AWARD SESSION

42. Treatment of Bile Acid Malabsorption Using Ileal StemCell Transplantation. J. R. Avansino, M.D., V. Hoagland, B.S.,J. Woolman, B.A., W. G. Haigh, Ph.D., M. Stelzner, M.D. Uni-versity Of Washington/VAPSHCS.

Purpose: To determine if ileal stem cells transplanted into asegment of jejunum can be used to treat bile acid malabsorption.Methods: In adult Lewis rats, a 15-cm segment of jejunum wasisolated with its blood circulation left intact and partially stripped ofenterocytes using luminal high-velocity perfusions with 3 mM EDTAsolutions. Continuity was restored by anastomosing the proximaland distal gut. Ileal stem cell clusters were harvested from neo-natalLewis rats and transplanted into these segments. After 4 weeks, ratsunderwent an ileectomy and the isolated segment was anastomosedin its place. After an additional 4 weeks, a 48 hour stool collectionwas performed. The engrafted segment was harvested for tauro-cholate uptake studies, ileal bile acid transporter (IBAT) protein (byimmunohistomorphometry) and IBAT mRNA quantitation (by RT-PCR). Data was analyzed by ANOVA/t-test. Rats undergoing ileec-tomy, jejunectomy or sham operation served as controls. Results:Total bile acid loss in the stool was significantly lower in rats with aneo-ileum compared to rats with an ileectomy (p � 0.004; Fig. 1).Na�-dependent taurocholate uptake was significantly increased inthe neo-ileum compared to the jejunum (p � 0.001). IBAT proteinsignal intensity was significantly higher in the neo-ileum comparedto jejunum (p � 0.001). IBAT mRNA amounts were significantlyhigher (p � 0.004) than in the jejunum however comparable to theileum. Conclusion: Ileal stem cells were used to establish a newzone of active bile acid uptake and IBAT expression in a jejunalsegment. This neo-ileum eliminated loss of bile acids in the stoolafter ileectomy. This is the first time intestinal stem cell transplan-tation has been shown to correct a clinical malabsorption syndrome.

43. Cigarette Smoke Increases Intimal Hyperplasia (IH) andHomocysteine (Hcy) in a Rat CEA Model. J. A. Davis, M.D.,A. T. Brown, Ph.D., H. Chen, M.D.. Y. Wang, B.S., J. F. Eidt,

252 ASSOCIATION FOR ACADEMIC SURGERY—ABSTRACTS

M.D., M. M. Moursi, M.D. University of AR for Med Sci/CentralAR Veterans Healthcare Sys.

Purpose: Evaluate cigarette smoke exposure on post CEA IH anddefine Hcy’s role in this effect. Introduction: HyperHcys and smok-ing are independent risks for CVD, however the relationship of thesetwo factors with post CEA IH are unclear. We performed a CEA inrats exposed to cigarette smoke with the hypothesis that smokingwould increase IH that may relate to an elevated plasma Hcy. Folicacid (FA) was used to test for the significance of Hcy elevation.Methods: All rats underwent an open CEA with direct removal ofthe intima and suture closure of the carotid. N � 13 rats received“smoke” exposure and N � 12 received “no smoke”. Smoke exposurewas 8 cigarettes/day 2 weeks prior, and 2 weeks post CEA. The ‘nosmoke’ and ‘smoked’ rats were given either control (N � 7, N � 6) or25 ppm added FA (N � 6, N � 6), resulting in 4 groups of rats. Ratswere sacrificed at 2 weeks post CEA; liver, urine, blood and carotidarteries samples were obtained. Endpoints were: IH (% luminalstenosis), plasma Hcy and 2 enzymes responsible for Hcy metabolismMTHFR and CBS, as well as urinary cotinine (a nicotine metabolite).Results: Smoke exposure increased IH vs. no-smoke controls bynearly 120% (57.8 � 6.2 vs. 26.8 � 5.4 %, p � 0.005). Smoke exposedrats had an increased plasma Hcy vs no smoke controls (8.3 � 0.8 vs5.7 � 0.8 umol/L, p � 0.015). Regression analysis showed a positivecorrelation between IH and plasma Hcy (R � 0.41, p � 0.039). FA inthe no smoke group produced no changes in IH or Hcy. Smoked ratsgiven FA had decreased plasma Hcy from smoke group and similar tono smoke group. Along with reductions in Hcy, FA eliminated theincrease in IH seen with smoke exposure (33.5 � 6.2 vs. 57.8 � 6.2%, p � 0.003). CBS activity decreased in smoked rats by nearly 20%vs no smoke rats (474.2 � 12.2 vs. 592.0 � 30.1 u/mg, p � 0.004). FAsupplementation in smoked rats increased CBS activity to control nosmoke group levels. There were no changes in MTHFR activityamong the groups. Smoked rats had increased urinary peak cotininelevels vs no smoke rats (1111.1 � 302.7 vs. 39.7 � 12.5, p � 0.04)which decreased with FA (p � 0.01). Conclusion: Smoking in-creases plasma Hcy and post-CEA IH. This suggests Hcy has anetiological role in the IH increase observed with smoking, since bothwere negated with FA.

44. Endotoxin Inhibits Enterocyte Migration by IncreasingIntegrin Function Through PI3 Kinase. F. Qureshi, M.D.,H. R. Ford, M.D., S. Cetin, M.D., J. Li, B.S.C., P. Boyle, B.S.C.,L. Sysko, B.S.C., J. Upperman, M.D., D. J. Hackam, M.D., Ph.D.Children’s Hospital of Pittsburgh, Pittsburgh, PA.

Purpose: Experimental necrotizing enterocolitis (NEC) is charac-terized by circulating endotoxin and impaired enterocyte migration.Because migration requires the detachment of cells from the under-lying matrix and the inhibition of surface integrins, we hypothesizedthat impaired enterocyte migration is due to increased integrin func-tion. Methods: NEC was induced in newborn rats after formulagavage and hypoxia, from which terminal-ileal mucosal scrapingswere prepared. IEC-6 cells were treated with LPS (50 ug/ml,12 h) �/-PI3K inhibitor LY(25 uM). Integrin expression was determined byconfocal microscopy or SDS-PAGE of cells sub-fractionated by cen-trifugation at 100,000 g. Integrin activity was determined by adher-ence of 15um fibronectin coated latex beads to IEC-6 monolayers.Migration was measured by time-lapse microscopy of IEC-6 cellsmoving into a scraped wound. Results: Expression of �3 and �1integrins was increased 3-fold in ileal mucosal scrapings from NEC-rats vs. ctrls (n � 3, p � 0.05), and 4-fold in LPS-treated IEC-6 cellsvs. ctrls (n � 5, p � 0.05). LPS caused a significant shift of �3 and �1integrins from the cytosolic to plasma membrane fraction by SDS-PAGE and confocal microscopy (see Figure, arrows point to �1 ex-pression), which was reversed by LY. Increased integrin expressionwas associated with a PI3K-dependent increase in integrin functiondetermined by bead adhesion (beads/cell: ctrl: 56 � 4, LPS: 77 � 4,

LPS � LY: 35 � 3, LY 45 � 5, n � 3 experiments, p � 0.001). Finally,maintenance of internalized integrins reversed the LPS-induced in-hibition of migration (ctrl: 7 � 2 um/h, LPS: 0.5 � 1 um/h, LY 4 � 2um/hour, p � 0.05). Conclusion: Enterocyte migration is inhibitedby LPS through increased surface expression of integrins in a PI3Kdependent manner. Modulation of enterocyte migration via integrinexpression may provide novel insights into the pathogenesis of NEC.

PARALLEL SESSION II

CARDIOTHORACIC PARALLEL SESSION

45. Inhaled Carbon Monoxide Abrogates Pulmonary Hyper-tension. E. P. Nadler, M.D., B. S. Zuckerbraun, M.D., L. E.Otterbein, M.D., T. R. Billiar, M.D., S. Kanno, M.D. Departmentof Surgery, University of Pittsburgh Medical Center, Pittsburgh,PA.

Background: Pulmonary hypertension poses a significant clinicalchallenge. Our current therapies are limited and not efficacious.Carbon monoxide (CO), which is produced endogenously by hemeoxygenases, has been shown to possess vasoregulatory properties.Therefore, we hypothesized that inhaled low dose CO would preventand reverse pulmonary vascular hyperplasia and right ventricularhypertrophy (RVH) in an animal model of pulmonary hypertension.Methods: Monocrotaline (MCT)-treated rats were divided into 4groups (n � 3–6 per group). Group A received MCT (50 mg/kg, s.c.)alone. Group B was treated with 1 hour daily of inhaled CO (250ppm) days 1–14 after MCT administration. Group C received COfrom days 15–28 and Group D received CO from days 29–42. Allanimals were sacrificed on day number 43 and their hearts and lungsharvested for morphometric and histologic evaluation. Body weight,right and left ventricular masses, and mean pulmonary arterialpressure (mPAP) were evaluated. Results: MCT caused progressivepulmonary hypertension. By day 42, MCT-treated rats had a mPAPof 35 � 3.3 compared to untreated controls whose mPAP was 16 �2.1 and CO-treated rats which had a mPAP of 20 � 7.1 (p � 0.05,ANOVA). CO prevented RVH in all treatment groups, even whennormalized for body weight (Table) (p � 0.05, Fisher’s Least Signif-icant Difference). Rats treated with MCT alone displayed significant

TABLE—ABSTRACT 45

CO(days given)

Body wt.(gm)

RV wt.(mg)

RV/BW(mg/g)

Group A None 219 � 3.6 414 � 33.5 1.89 � 0.19Group B 1–14 275 � 8.3* 157 � 14.7* 0.57 � 0.06*Group C 15–28 266 � 9.7* 180 � 30.1* 0.68 � 0.14*Group D 29–42 236 � 11.7* 193 � 67.2* 0.82 � 0.3*

* p � 0.05 v. Group A, Fisher’s Least Significant Difference,ANOVA.

253ASSOCIATION FOR ACADEMIC SURGERY—ABSTRACTS