~ z"l.

Joumal of Surgical Research 82, 201-208 (1999) @Article ID jsre.1998.5520, available online at http://www.idealibrary.com on ID E ~L

Femur Fracture Induces Site-Speciflc Changes in T -Cell Immunity

Molly M. Buzdon, M.D., Lena M. Napolitano, M.D.,l Hui-Jun Shi, M.S., Douglas M. Ceresoli, Ph.D.,Ravi Rauniya, and Barbara L. Bass, M.D.

Department of Surgery, Baltimore VA Medical Center, University of Maryland School of Medicine, Baltimore, Maryland 21201

Submitted for publication July 17. 1998

h) alter injury. Changes in the GALT immune responsemay contribute to intestinal mucosal dysfunction andincreased susceptibility to gut-derived sepsis altertraumatic injury. C> 1999 Academic Press

INTRODUCTION

Background. Trauma is associated with altered hostdefense and susceptibility to infection, in part due tocytokine dysregulation and altered T-cell immunity.The gut-associated lymphoid tissue (GALT) provides adefense against infection and contributes to the pro-cess of mucosal healing by T-cell activation and cyto-kine production.

Objective. To determine whether femur fracture in-duces alterations in Peyer's patch and splenic T-cellphenotype, proliferative response, and cytokine ex-pression following traumatic injury.

Methods. Mice underwent femur fracture or shamprocedure and, 48 h later, lymphocytes were isolatedfrom spleen and Peyer's patches. Lymphocytes werecultured, and lipopolysaccharide (10 p.g/ml) was addedin some cultures. Cells and supernatant were har-vested at 48 h. Proliferation was analyzed by [3HJthy-midine, and interleukin-10 aL-lO) protein was mea-sured by ELISA in the culture supernatant. T-cellphenotype was determined by flow cytometry.

Results. Femur fracture induced a significant in-crease in proliferative response in Peyer's patch im-munocytes. In contrast, no significant differenceswere identified in splenocyte proliferative response48 h after femur fracture injury. Femur fracture in-duced a significant decrease in IL-10 protein expres-sion of both splenocytes and Peyer's patches. Femurfracture algo induced a significant increase in the frac-tion of CD3+, CD4+, and T-cell receptor-cx¡J Peyer'spatch immunocytes, whereas splenocytes demon-strated no significant phenotypic change.

Conclusion. Femur fracture is associated with sig-nificant alterations in Peyer's patch but not splenicT-cell phenotype and proliferative response early (48

1 To whom correspondence should be addressed at Department of

Surgery, Baltimore VAMC and University of Maryland School ofMedicine, Room 5C-116, 10 North Greene Street, Baltimore, MD21201. Fax: (410) 605-7919.

201 0022-4804/99 $30.00Copyright @ 1999 by Academic Press

Al! rights of reproduction in any form reserved.

The impact of trauma has far-reaching effects be-yond local tissue injury. Trauma has been shown tocause changes in intestinal permeability and immuno-globulin production, which may lead to increased sus-ceptibility to infection, systemic inflammatory re-sponse syndrome, and multiple organ dysfunction[1-4]. The gut-associated lymphoid tissue (GALT, in-cluding Peyer's patch, mesenteric lymph node, intra-epithelial, and lamina propria lymphocytes) plays acritical role in providing local defense againstintestine-derived infection through interactions be-tween B and T immunocytes, IgA production, and cy-tokine production. Little is known, however, of thelocal GAL T immune response induced by traumaticinjury .

Recent experimental studies have documented sig-nificant alterations in GALT immunity in murine mod-els of hemorrhagic shock and trauma. Xu et al. [5]documented a significant decrease in total viable cellyield and an increased apoptosis in Peyer's patch cells24 and 72 h fOllOwing hemorrhage or trauma (laparot-omy) plus hemorrhage in a murine modelo In addition,Abraham and Freitas [6] found decreased numbers ofBcells secreting IgM, IgA, and IgG2b as well as totalserum Ig levels after hemorrhage, and suggested thatthese indicators ofhumoral immunosuppression playasignificant role in the intestinal mucosal immune re-sponse following trauma and hemorrhage.

It has previously been determined tha,t"tissue injuryinduces significant alterations in systemic immunity

JOURNAL OF SURGICAL RESEARCH: VOL. 82, NO. 2, APRIL 1999202

and cytokine regulation that differ from studies inhemorrhagic shock models. We have previously docu-mented that traumatic injury induces a suppression insystemic interleukin-1O aL-lO) expression, and is as-sociated with increased gut permeability [7, 8]. To ourknowledge, no previous studies have been performed todate investigating the alteration in mucosal immune(GALT) function after soft tissue trauma. In thesestudies we investigated the effect of tissue trauma onintestinal mucosal immunity in a murine femur frac-ture model by evaluation of Peyer's patch immuno-cytes.

dex calculated as mean cpm in LPS-stimulated cultures divided byme~n cpm for unstimulated conditions, thereby refiecting alterationsin both unstimulated and stimulated proliferative responses.

IL-1O quantitation. IL-10 was measured by ELISA (PerSeptiveBiosystems, Framingham, MA) in Peyer's patch and splenic lympho-cytes cultured in sterile 96-well fiat-bottom microtiter plates with5 x lO. cells per well in RPMI medium and 10 ¡Lg/ml LPS in vitro in

the culture supernatant.Flow cytometry. To determine the phenotype oflymphocytes iso-

lated from Peyer's patch and spleen, lO. cells were placed in 50 ¡Ll ofF ACS buffer (PBS/2% fetal calf serum/ 0.08% NaN 3) in V -bottom96-well plates (Linbro, Aurora, OH). Fc Block (Pharmingen, SanDiego, CA) was added at a concentration of 1:100 and kept on ice for5 mino The proper antibody concentrations were added singly or incombination: fiuorescein isothiocyanate (FITC) anti-CD3 (clone 145-2C11) (Pharmingen), allophycocyanin (APC) anti-CD8 (clone 53-6.7)(Pharmingen), FITC anti-TCR a{3 (clone H57-597) (Pharmingen),FITC anti-TCR-y8 (clone GL3) (Pharmingen), R-phycoerythrin (R-FE) anti-CD4 (clone GK1.5) (Pharmingen). AlI antibodies were inF ACS buffer; incubations were for 30 min on ice. Following staining,the cells were washed twice in HBSS containing 1% bovine serumalbumin. Three-color fiow cytometry analysis was performed on aBecton Dickinson FACSTAR cell sorter for analysis of lO' cells persample. These experiments were repeated on three separate occa-sions (n = 4 animals per group) to confirm the results obtained.

Statistical analysis. Data are expressed as means :t SEM. Anal-ysis of variante (ANOV A) was performed to assess the differenceamong all experimental groups. A Fisher protected least significancedifference test was used to make post hoc comparisons. A P value ofless than 0.05 was considered to be significant.

MATERIALS AND METHODS

RESULTS

Lymphocyte Proliferation

Femur fracture induced a significant increase in pro-liferative res pon se in Peyer's patch immunocytes, withapproximately a twofold increase in proliferative indexcompared with normal and sham controls. In contrast,no significant differences were identified in splenocyteproliferative response 48 h after femur fracture injury(Fig. 1).

Animals. Male Balb/C mice (Taconic Farms, New York, NY), 20to 30 weeks old and weighing 20-25 g, were used in this study.Animals were maintained on a 12-h lighUdark cycle and were redstandard chow and water ad libitum. AlI procedures, anesthesia, andeuthanasia were approved by the Institutional Animal Care and UseCornmittee of fue University of Maryland at Baltimore. AlI experi-mental groups (control, sham, femur fracture) included 12 animals

per group.Experimental designo The animals were anesthetized with inha-

lational methoxyfiurane. The right hind leg was shaved and preppedwith 70% alcohol. Femur fracture (FFx) animals underwent a forcefracture at the midshaft of the left femur with a hemostat, inducingboth soft tissue and boDe crush injuries, and were resuscitated with1.0 cc normal saline administered subcutaneously at the base oftheneck and maintained on a heating pad until full recovery fromanesthesia .Sham control s underwent anesthesia with methoxyfiu-ralle and resuscitation alone. Animals were allowed standard chowand water ad libitum in the postoperative period and were sacrificed48 h after femur fracture.

Lymphocyte isolation. Lymphocytes were harvested by excisingPeyer's patches from the serosal sirle of the small bowel, mincing,then rinsing through 40-¡Lm filter with RPMI 1640 medium (Gibco)supplemented with 10% heat-inactivated fetal bovine serum (Ry-clone), 1000 IU/ml penicillin, and 1000 ¡J.g/mlstreptomycin (Gibco).The cell suspension was centrifuged at 1000 rpm for 5.min, washedtwice with RPMI medium, and then re suspended in 5 mI of RPMImedium. Splenic tissue was teased apart with blunt forceps, filteredthrough a 40-¡J.m filter, and centrifuged at 1000 rpm for 5 mino Redblood cells were lysed with R2O supplemented with 0.15 M NR.CI, 1mM KHCO3, and 0.1 mM Na2EDTA. Lymphocytes were washedtwice and resuspended in 5 mI of RPMI medium. Peyer's patch andsplenic lymphocytes were cultured in sterile 96-well round-bottommicrotiter plates (Corning Glass Works, Coming, NY) with 5 x 105cells per well in RPMI medium with 5% fetal bovine serum (FBS)(the unstimulated group) or stimulated with 10 ¡J.g/mllipopolysac-chancle (LPS) in vitro. LPS was chosen as the stimulant for theselymphocyte cutlures based on our previous studies [7, 81 which doc-umented that three stimulants (LPS, concanavalin A, and phytohe-magluttinin) had the same effect on splenocyte proliferative responseafter FFx. Cultures were incubated at 37°C in a humidified atmo-sphere of 5% CO2. Both cells and supematant were harvested at 48 hof culture. Tritiated thymidine (0.25 ¡J.Ci/ml of culture medium) wasadded for the final 12 h of incubation. Cells were harvested anta aglass-fiber filter (Xtalscint, Ready Filter, Beckman, Fullerton, CA)with a Skatron semiautomatic cell harvester (Skatron Inc., Sterling,VA) and tritiated thymidine uptake was measured with a scintilla-tion counter. Each set of cultures had three replicates performed,and mean counts per minute (cpm) for each culture condition werecalculated. [3R]Thymidine results are presented as proliferative in-

IL-IO Protein Expression

Constitutive expression ofIL-lO in Peyer's patch andsplenic immunocytes was no different in all three ex-perimental groups. Peyer's patch immunocytes fromnormal controls secrete small amounts of IL-IO in re-sponse to LPS, approximately 70 pg/ml. IL-IO proteinexpression in Peyer's patch immunocytes was signifi-cantly downregulated in femur fracture animals, witha decrease of approximately 65% relative to normaland sham contraIgo Splenocyte IL-IO production wasapproximately 20-fold greater than Peyer's patch IL-IOproduction in all experimental groups. Femur fracturealgo induced a significant decrease in splenocyte IL-IOprotein expression 48 h postinjury, approximately a60% reduction compared with normal and sham con-trols (Fig. 2).

1:

203BUZDON ET AL.: FRACTURE-INDUCED CHANGES IN IMMUNITY

A pression increased from 15% in normal and shamcontrols to 23% in the experimental femur fracturegroup (Fig. 4). Peyer's patch immunocytes bearingTCRa/3 were algo significantly increased in the femurfracture group (Fig. 5). There was no difference inTCR'Y5-positive cells in Peyer's patch lymphocytes,which were les s than 0.3% of total in all experimen-

50

40xQ)'CC

Q)>,-...=~Q)o,-,--Q~c..

30

A

20

10

oFemur FxNor mal Sham

~ 80EO)c.C 60

c,-...=1-...C 40~UCCuO 20-I

..JB

oNormal Sham Femur Fx

B.2000

-EO) 1500Q.

CO

:::;=~ 1000

CQj1.)CO1.)O 500-I

..J

Nor mal Sham Femur Fx

FIG.1. Proliferative index ofPeyer's patch (A) and splenic (B)immunocytes in normal, sham, and femur fracture mice, deter-mined by [3HlthymidiI1e. Proliferative index was calculated asmean counts per minute (cpm) in LPS-stimulated cultures dividedby mean cpm for unstimulated conditions. Femur fracture induceda significant increase in proliferative response 48 h postinjury inPeyer's patch (A) immunocytes, with approximately a twofold;~~?OO~~ :- -?~I:l"o-~':'.- ;n~o~ n~-"",?o.1 ~,;." ...n~ol ~_.. a"~- oti.~."..o~ .ti l"v",~"""V~ """'A "V"'l'~'~~ v «. U o "

controls (* P < 0,001 vs 'normal and sham controls), There was nosignificant differences in splenocyte (B) proliferative response48 h after'femur fracture injury,

T -Gell Phenotype

Femur fracture induced a significant increase inthe fraction of CD3 + Peyer's patch immunocytes

(23% vs 15% in both normal and sham controls, Fig.3). Similarly, CD4 + Peyer's patch T-Iymphocyte ex-

Normal Sham Femur Fx

FIG. 2. Expression of IL-IO protein in Peyer's patch (A) andsplenic (B) immunocytes in normal, sham, and femur fracture mice48 h after injury. IL-IO protein expression in Peyer's patch (A)immunocytes was significantly downregulated in femur fracture an-imals, with a decrease of approximately 65% relative to normal andsham controls (* P < 0.05 vs normal and sham controls). Femurfracture induced a significant decrease in splenocyte (B) IL-IO pro-tein expression 48 h postinjury, approximately a. 60% reductioncompared with normal and sham controls (* P < 0.03 vs normal andsham controls).

204 JOURNAL OF SURGICAL RESEARCH: VOL. 82, NO. 2, APRIL 1999

BA

1coa

rCD8

c

CD8

CD3 ..

FIG.3. Flow cytometric analysis ofT-cell phenotype in Peyer's patch lymphocytes with two-color immunostaining. Expression ofCD3+(lower right quadrant) and CD8+ (upper left quadrant) receptora in normal (A). sham (B), and femur fracture (C) mice. Femur fractureinduced a significant increase in Peyer's patch lymphocyte CD3 expression (23% vs 15% in both normal and sham controla). Data shown arerepresentative of three separate experimenta.

tal groups. Femur fracture did not induce any signif-icant change in splenocyte CD3, CD4, TCRaf3, orTCR'Yc5 expression, in contrast to the significant al-terations demonstrated in Peyer's patch T-cell phe-notype (data not shown).

er's patches, which are macroscopic lymphoid aggre-gates that extend from the mucosa into the submucosaof the intestine. The Peyer's patch is a specialized Bitefor initiation of mucosal immune responseso The ger-minal centers ofPeyer's patches are composed primar-ily ofB cellso By 19 weeks of gestational age, the B cellsare identifiable in distinct follicular aggregateso Adja-cent to the germina! centers are interfollicular zonesthat contain predominantly CD4 + T cells. Intestinal

antigen uptake stimulates naive B and T cells in thePeyer's patch to migrate to mesenteric lymph nades,undergo further division, and differentiate into laminapropria and intraepithelial lymphocytes. B lympho-

DISCUSSION

GALT is the largest immune cell mass in the bodyand provides immunological protection against resi-dent microbial flora and potential infectious pathogens[9-11]. The initial site of entry of immature lympho-cytes into the gastrointestinal tract is through the Pey-

BUZDON ET AL.: FRACTURE-INDUCED CHANGES IN IMMUNITY 205

A B

CD8

rCD8

CD4 ..

c

CD8

CD4 .FIG.4. Flow cytom~tric analysis ofT-cell phenotype in Peyer's patch lymphocytes with two-color immunostaining. Expression of CD4 +

(lower right quadrant) nd CD8+ (upper left quadrant) receptors in normal (A), sham (B), and femur fracture (C) mice. Femur fractureinduced a significant in rease in Peyer's patch lymphocyte CD4 expression (23% vs 15% in both normal and sham controls). Data shown arerepresentative of three separate experiínents.

Gryglewski et al. [18] recently reported that GALTphenotype was altered in a murine model of majorsurgery (partial gastrectomy). These studies docu-mented a significant postgastrectomy Day 3 increase inthe percentage of TCR-yS in Peyer's patches (546% ofsham control values) and a concomitant decrease inTCRa{3 (55% of sham control values). There was asimultaneous decline in both TCRa{3 and TCR-yS cellsin peripheral blood (45 and 58% of normal values,respectively). In contrast, leg amputation was associ-ated with no change in Peyer's patch TCR-yS or TCRa{3.These authors suggested that major surgical stress(partial gastrectomy) mar disturb the normal cell traf-tic selectively, with increased TCR-yS cell homing in

cytes in the lamin~ propria produce and secrete IgA.The IgA response Ibas algo be en found to be T celldependent.

Most T lymphocytes recognize antigen vía TCRa{3cell receptor. TCRy8 cells appear in the epithelium by18 weeks of gestatipn, when the human gut is sterile.Thus, migration O~ T cells in the fetal intestine is inpart antigen indep ndent. TCRy8 cell function is notwell characterized, but these cells have been found inlarger quantities in Peyer's patches than in peripheralblood. TCR y8 cells can decrease [12-14] or augment[15-17] immunological response in specific situationsand their immunomodulatory role is strongly sug-gested.

206 JOURNAL OF SURGICAL RESEARCH: VOL. 82, NO. 2, APRIL 1999

BA

CD8

rCD8

c

rCD8

FIG.5. Flow cytometric analysis ofT-cell phenotype in Peyer's patch lymphocytes with two-color immunostaining. Expression ofT-cellreceptor-ap+ (lower right quadrant) and CD8+ (upper left quadrant) receptors in normal (A), sham (B), and femur fracture (C) mice. Femurfracture induced a significant increase in Peyer's patch lymphocyte T-cell receptor-ap expression (22% vs 15.5% in both normal and shamcontrols). Data shown are representative of three separate experiments.

intestinal Peyer's patches and lymph nades (GALT)and with cell displacement from peripheral blood tolymphatic organs.

Furthermore, Xu and colleagues [5] determined thathemorrhagic shock alone or trauma coupled with hem-orrhage in mice produces decreased cell yield and in-creased apoptosis in Peyer's patch cells, which wasassociated with increased Fas expression. Thesechanges were noted to be predominantly in the B-celllineage. These studies suggest that the induction ofapoptosis or "programmed cell death" may algo playanimportant role in intestinal mucosal immune dysfunc-tion after trauma- hemorrhage.

In the studies presented herein we sought to inves-tigate alterations in the gut-associated immune re-

sponse after traumatic tissue injury instead of hemor-rhagic shock. Murine femur fracture induced asignificant increase in Peyer's patch lymphocyte CD3,CD4, and TCRa{3 expression early (48 h) postinjury,whereas splenocyte phenotype was unchanged. Itshould, however, be clarified that the studies of T -cellphenotype were performed on unstimulated immuno-cytes and not in those immunocyte criltures that wereexposéd to LPS as a stimulant. These data suggest thatGALT T-cell phenotype is affected very early after in-jury, when other constitutive responses such as prolif-eration and IL-IO production are unaltered.

Proliferative response of Peyer's patch immunocytesstimulated with LPS was significantly increased afterfemur fracture, algO suggesting locallymphocyte acti-

BUZDON ET AL.: FRACTURE-INDUCED CHANGES IN IMMUNITY 207

lymphocyte CD3, CD4, and TCRa/3 expression and anassociated significant decrease in Peyer's patch .IL-IOprotein expression, suggesting a greater tendency to-ward an increased local intestinal inflammatory re-sponse early postinjury. Changes in the GAL T immuneresponse may contribute to intestinal mucosal dysfunc-tion and increased susceptibility to gut-derived sepsisafter traumatic injury.

REFEREN CES

vation in the intestinal microenvironment. Previousstudies in our laboratory [7] investigated splenocyteproliferative responses to LPS and two other stimu-lants (concanavalin A and phytohemagglutinin) 24 and96 h after femur fracture. A similar response was notedwith all three stimulants, documenting a significantincrease in splenocyte proliferation at 24 h and a sig-nificant decrease in proliferation at 96 h. Additionallong-term studies determined that the depressedsplenocyte proliferative response persisted for up to 2weeks following femur fracture in this murine modeloThere is, therefore, a time-dependent alteration insplenocyte proliferative response after femur fractureinjury. We have investigated Peyer's patch prolifera-tive response at one other time point after injury (72 h)and a significant increase in proliferative response wasalgo identified (proliferative index = 45).

Interleukin-10 expression by Peyer's patch andsplenic immunocytes was algo examined after tissuetrauma. IL-10 is an important anti-inflammatory cyto-kine, and plays a critical role in intestinal immunity asshown in IL-10-deficient mice who develop fatal entero-colitis. The primary defect in these mice is a failure tocontrol normal intestinal immune responses to entericantigens, leading to excessive inflammatory cytokinerelease in the intestinal milieu. Acquired local IL-10deficiency in the gut after trauma may algo result incytokine dysregulation and an exaggerated intestinalinflammatory response. Lagoo and colleagues [19] an-alyzed cytokine gene expression in several murine lym-phoid tissues and demonstrated that Peyer's patchesconstitutively expressed higher levels of multiple in-flammatory cytokine transcripts compared with mostother tissues examined. This is consistent with thecontinuous overexposure of these GALT immunocyi;esto intestinalluminal antigen including bacterial endo-toxin.

We have previously documented that trauma ticinjury induces a suppression in systemic IL-10 ex-pression, and is associated with intestinal mucosaldysfunction and increased intestinal permeability [7,8]. We hypothesized that GALT IL-10 expressionwould algo be downregulated after tissue trauma,and confirmed this in the studies presented here.Constitutive IL-10 expression in Peyer's patch andsplenic immunocytes was no different in all threeexperimental groups. In contrast, Peyer's patch andsplenocyte IL-10 protein expression in response toLPS stimulation was significantly decreased in micethat sustained femur fracture compared with normaland sham contraIgo

CONCLUSION

Femur fracture (as a model of traumatic tissue in-jury) induced a significant increase in Peyer's patch

1. Napolitano, L. M., Koruda, M. J., Zimmennan, M. S., McCowan,K., and Meyer, A. A. Chronic ethanol intake and burn injury:Evidence for synergistic alteration in gut and immune integrity.J. Trauma 38: 198, 1995.

20 Deitch, E. A., Morrison, Jo, Berg, Ro and Specian, R. D. Effect ofhemorrhagic shock on bacterial translocation, intestinal mor-phology, and intestinal penneability in conventional andantibiotic-decontaminated rats. Grito Care Med. 18: 529, 1990.

30 Deitch, E. A. The role ofintestinal barrier failure and bacterialtranslocation in the development of systemic infection and mul-tiple organ failure. Arch. Surco 125: 403, 1990.

4. Willoughby, R. P. No, Harris, K. A., Carson, M. W., Martin,C. M., Troster, M., DeRose, G., Jamieson, S. W., and Potter,R. F. Intestinal mucosal permeability to .'Cr-ethylene-diaminetetraacetic acid is increased after bilaterallower ex-tremity ischemia-reperfusion in the rato Surgery 120: 547,1996.

50 Xu, Y.X.,Ayala,A., Monfils,B., Cioffi,W. Go, andChaudry, l. H.Mechanism ofintestinal mucosal dysfunction following trauma--hemorrhage: Increased apoptosis associated with elevated Fasexpression in Peyer's patches. Jo Surco Res. 70: 55, 1997.

6. Abraham, E., and Freitas, A. Hemmorhage in mice inducesalterations in immunoglobulin-secreting B cellso Grito CareMed. 17: 1015, 1989.

7 o Napolitano, L. M., Koruda, M. J., Meyer, A. A., and Baker, C. C.The impact of femur fracture with associated soft tissue injuryon immune function and intestinal penneability. Shock 5(3):202, 1996.

8. Napolitano, L. M., Campbell, C., and Bass, B. L. Injury-inducedinterleukin-l0 deficiency is associated with increased nitric ox-ide release and gut mucosal disruption. In G. T. Keusch and M.Kawakami (Eds.), Cytokines, Cholera and the Gut, Proceedingsof the NIH U.S.-Japan Cooperative Medical Science Programon Gut and Cytokines, Malnutrition Panel, 1995. Amsterdam:lOS Press. Pp. 113-119.

9. MacDonald, T. T., and Spencer, J. Ontogeny of the gut-associated lymphoid system in mano Acta Paediatr Supplo 395:3,1994.

10. Williams, M. B., and Butcher, E. C. Homing ofnaive and mem-ory T lymphocyte subsets to Peyer's patches, lymph nodes andspleen. J. Immunolo 159: 1746, 1997.

11. Farstad, l. N., Norstein, J., and Brandtzaeg, P. Phenotypes ofBand T cells in human intestinal and mesenteric lympho Gastro-enterology 112: 163, 1997. ,.0..:

12. Schattenfroh, N. D., Hoffman, R. A., McCarthy, S. Ao, andSimmons, R. L. Phenotypic analysis of donor cells infiltratingthe small intestinal epithelium and spleen during graft-versus-host disease. Transplantation 27: 268, 1995.

13. McMenamin, C. M., McKersey, M., Kuhnlein,P., Hunig, T o, andHolt, P. G. -yfJ T cells downregulate primary IgE responses in

208 JOURNAL OF SURGICAL RESEARCH: VOL. 82, NO. 2, APRIL 1999

rats to inhaled soluble protein antigens. J. Immunol. 154: 4390,1995.

14. Vaessen, L., Ousehand, A., Baan, C., Jutte, N., Balk, A., Clas,F., and Weimer, W. Phenotypic and functional analysis ofTCR-y8-bearing cells isolated from human heart allografts.J. Immunol. 147: 846, 1991.

15. Broker, B., Lydyard, P. M., and Emmrich, F. The role ofTCR-y8cells in the normal and disordered immune system. Klin.Wochenschr. 68: 489, 1990.

16. Ptak, W., Szczepanik, M., Ramabhadran, R., and Askenaze,P. W. Immune or normal TCR-y8 that assist aJ3 T cells inelicitation of contact sensitivity preferentially use V 5 and V4 variable region gene segments. J. Immunol. 156: 976,1996.

17. Askenaze, P.W., Szczepanik,M.,Ptak, M., andPtak, W. TCR-y8cells in normal spleen assist immunized a{3 t cells in the adop-tive transfer of contact sensitivity: Effect of Bordetella pertussiscyclophosphamide, and anti-suppressor cell antibodies. J. Im-munol. 154: 3644, 1995.

18. Gryglewski, A., Szczepanik, M., Majcher, P., Popiela, T., andPtak, W. Different patterns of -y8 and a{3 T cell redistribution inthe mouse after partial gastrectomy. J. Surg. Res. 73: 137,1997.

19. Lagoo, A. S., Eldridge, J. H., Lagoo-Deenadaylan, S., Black,C. A., Ridwan, B. U., Hardy, K. J., McGhee, J. R., and Beagley,K. W. Peyer's patch CD8 + memory T cells secrete T helper type

1 and type 2 cytokines and provide help for immunoglobulinsecretion. Eur. J. Immunol. 24: 3087, 1994.