2 June 19992 March 1998

MucosalImmunologyUpdateSMI Officers

PresidentCharles O. Elson

President-ElectWarren Strober

Secretary/TreasurerPeter B. Ernst

CouncillorsNadine Cerf-BensussanPatrick HoltHiroshi KiyonoLloyd Mayer

Society Business Office4340 East West Highway, Suite 401Bethesda, MD 20814–4411, USATel: (301) 718-6516; Fax: (301) 656-0989E-mail: smi@paimgmt.com

Mucosal Immunology Bulletin BoardIGA-L@nlists.ufl.edu

World Wide Webhttp://www.socmucimm.org

Publication CommitteeAllan Cripps

e-mail: cripps@science.canberra.edu.auHiroshi Kiyono, Editor

e-mail: kiyono@biken.osaka-u.ac.jpThomas T. MacDonald

e-mail: t.t.macdonald@mds.qmw.ac.ukVictor E. Reyes, Assistant Editor

e-mail: vreyes@utmb.eduDenis Snider

e-mail: snider@fhs.csu.mcmaster.ca

Design and LayoutMary Jo Durso SingletonGraphic Services, LogisticsUniversity of Texas Medical Branch at Galveston

Mucosal Immunology Update (ISSN 1068-7629) is published quarterly in March, June, September, and December by the Society forMucosal Immunology, 4340 East West Highway, Suite 401, Bethesda, MD 20814-4411, and is designed and printed by The University ofTexas Medical Branch, Logistics Graphic and Printing Services, 301 University Blvd., Galveston, TX 77555-1044. Periodicals postage paid atBethesda, MD and at additional mailing offices.

Postmaster: Send change of address to Mucosal Immunology Update, c/o: Peter Ernst, Secretary/Treasurer; 4340 East West Highway, Suite401, Bethesda, MD 20814-4411, USA. Canada Post International Publications Mail Product (Canadian Distribution) Sales Agreement No. 620173.

Copyright © 1998 by Society for Mucosal Immunology. All rights reserved. Printed in the U.S.A.

All members of the Society for Mucosal Immunology receive the Mucosal Immunology Update with their membership. Subsriptions tonon-members: personal subscriptions $60 in U.S., $70 elsewhere; institutional subscriptions $80 in U.S., $100 elsewhere; subsription freeto members; single copies, $25 for nonmembers. All prices include a nominal handling charge where applicable. New prices are effectiveJanuary 1, 1999. For subscribers outside the North America, add $3 for air freight. All issues to destinations outside North America areshipped via air freight—delivery within 7–12 days worldwide. Bulk rates available upon request. Payments should accompany all orders.

The Society of Mucosal Immunology cannot be held responsible for errors or for any consequences arising from the use of the informationcontained in this newsletter. The appearance of advertising in this newsletter does not constitute an endorsement or approval by The Society ofMucosal Immunology of the quality or value of the product advertised or of claims made for it by its manufacturer.

MIU is supported in part by a grant-in-aid for publication of Scientific Research Results by the Ministry of Education, Science, Sportsand Culture of Japan.

Table of Contents

Editorial ....................................................................... 3

Health scares—a very English pastime? ........... 3

Dendritic Cells are Key Regulators ofIntestinal Immune Responses .......................... 5

Commentary—How InflammationOvercomes Suppression .................................. 10

Dendritic Cells, Mucosal Immunityand Peripheral Tolerence ................................ 11

Mucosal Immunology AffinityGroup Workshop-Harrogate 1998 ................ 15

The First European MucosalImmunology Meeting ...................................... 17

Charles Elson, M.D., Presidentof the Society for Mucosal Immunology .... 18

Announcements ..................................................... 19

2 June 1999

The cover depicts a collage of photos of mucosal inductive sites.Counterclockwise from the top, a view of a lymphoid aggregatein the airway (courtesy of John Bienenstock, in the “Handbookof Mucosal Immunology”’ with permission); a line drawing ofthe Peyer’s patch from Peyer’s original manuscript, 1677; ascanning eletronmicrograph of M-cells overlying a patch(courtesy of Jacques Pappo). The background is anotherscanning electronmicrograph of M-cells.

Official Publication of theSociety for Mucosal Immunology

MucosalImmunologyUpdate

Vol. 7, No. 2, 1999

• Editorial 3

• Health scares—a very English pastime? 3

• Dendritic Cells are KeyRegulators of IntestinalImmune Responses 5

• Commentary—HowInflammation OvercomesSuppression 10

Inside

• Dendritic Cells, MucosalImmunity and PeripheralTolerence 11

• Mucosal Immunology AffinityGroup Workshop–Harrogate 1998 15

• The First European MucosalImmunology Meeting 17

• Charles Elson, M.D.,President of the Society forMucosal Immunology 18

• Announcements/Classifieds 23

Mucosal Immunology Update 3

EditorialIn the first issue of Mucosal Immunology Update

in 1998 ( Vol. 6 #1), there was a cry from the heart ofthe editorial committee to the effect that we wishedMIU to become much more participatory, and toinclude submissions of any kind from the member-ship. One year on, the membership have made theirviews clear–there have been no unsolicited articles.This outcome, although not unexpected, is extremelydisappointing. I was always under the impressionthat immunologists would do pretty much anythingto get their name in print, so either this premise isfalse, or perhaps the membership feels that an ar-ticle in MIU is on par with one in the the MongolianJournal of Proctology.

Unfortunately, therefore this issue follows theusual format, with some modifications. I have com-missioned two mini-reviews from Jo Viney and Gor-

don MacPherson on dendritic cells. I have also in-cluded two meeting reports. The first is the success-ful first European Mucosal Immunology Meeting atSt. Bartholomews Hospital and the second is the mu-cosal immunology symposium held as part of theBritish Society for Immunology’s annual meeting inHarrogate just before Christmas. There is also a com-mentary from Yoshi Ohtsuka on a seminal paperwhich appeared in Nature recently showing howinterferon-g interrupted TGFb signalling, of greatrelevance to mucosal immunology. Finally, there isa personal, more philosophical piece, on the dam-age done to UK mucosal immunology, IBD researchand public health by repeated scare stories in themedia (fuelled by clinical investigators) aboutCrohn’s disease.

Tom MacDonaldSt. Bartholomews Hospital, London

Health scares—a veryEnglish pastime?Tom MacDonaldSt. Bartholomews Hospital

I had the pleasure of attending the recent meet-ing on coeliac disease in Naples and was very im-pressed by the detailed analysis of the way thattransglutaminase can modify gliadin peptides to makethem stronger T cell immunogens. High quality sci-ence, good food, convivial hosts and a beautiful loca-tion made it a very pleasant few days. Arriving backin England, checking my e-mail at home, I found amessage from Ian Sanderson to let me know about arecent story in the British media. If you don’t careabout the UK then stop reading now because other-wise this piece will seem too parochial.

While I was in Naples, there had been a nationalTV news item, with an interview, and questions togovernment health ministers, about the possibilty thatCrohn’s disease is caused by Mycobacterium paratu-berculosis in the drinking water in household taps(a few years ago it was in milk). Thinking perhapsthat there was a blockbuster paper about to comeout on this topic, I went directly to the scientificsource of these allegations. There was no paper, i.e.,

no data. It was merely a rehashing of an old story. Thesource of the allegation felt duty bound, after muchpressure and soul searching, to go to the media withhis fears! I was overwhelmed by their sense of duty.

A terrible sense of déjà vu came over me–here wego again! We have just been through this in the UKwhere the publications and public pronouncementsof some individuals on measles and Crohn’s diseaseled to a major public health scare and a drop in thenumber of infants receiving MMR vaccine. The samegroup also implied in the media, but not in the scien-tific press, that MMR might cause autism. This wasnot buried in some late night regional bulletin, buton the main evening news from the BBC. At signifi-cant cost, the MRC and Department of Health, under-standably worried about the drop in vaccination, aretrying to allay the fears of the public over this issue.

For some reason, health scares happen frequentlyin the UK, although obviously when it deals withCrohn’s it has a particular resonance for GI immu-nologists. In recent years there have been fearsabout salmonella in eggs, listeria in soft cheeses, andmost famously, BSE in cattle. The very latest is fearsover the safety of genetically modified foods–socalled Frankenstein foods. In all cases the media have

4 June 1999

given these issues a very high profile and in a highlysensationalist way. It is very easy for scientists toblame the media for this and say they distort thefacts. However this is fallacious, journalists may usewords that scientists do not like, but they are onlyflushing out a story with an alleged scientific basis.Don’t shoot the messenger.

There is however a broader issue, is it somethingpeculiar about the United Kingdom that allows thissort of thing to happen? From my own knowledge,these ideas do not seem to be taken seriously else-where. Is it the UK tradition of tolerance of eccen-tricity that allows this kind of thing? I am not sosure of this because the UK-based researchers whohave come up with most of these ideas frequentlyare invited to international meetings. Is it in the in-terest of completeness, paranoia or as light relief thatthis occurs?

Crohn’s health scares are quite recent but thesearch for the etiologic agent has a chequered his-tory. I therefore decided to take a trawl through theliterature on etiologic agents in Crohn’s disease overthe last 30 years, excluding the more fanciful onessuch as cornflakes and toothpaste. The results ofmy efforts are shown in the table. I have only shownthe main papers, in many cases the same groups havepublished a series of papers around the same puta-tive etiologic agent. Of the 27 papers shown (thereis not enough space to include the extensive num-ber of papers debunking most of this work), 11 werein the Lancet, seven were in Gastroenterology andthe work is dominated by UK investigors. Why theLancet, a London-based publication, and not the NewEngland Journal of Medicine? Is this again a reflec-tion of something peculiar about the UK? The Lan-cet is a very high profile journal and it must surelyrealize that its publications can have major implica-tions for public health. Who referees these papers? Ihave never met anyone who admits to this, althoughmany of my colleagues are very happy to announcethat they have refereed other papers which appearin prestige journals,

I have always also detected a sense of zealotry inthose who propound these new ideas. They are thevisionaries, helping humanity to prevent a devastat-ing disease such as Crohn’s. The rest of us areluddites, gnawing away at minor issues while they

deal with the big picture. The usual historical ex-amples of people who had major ideas which wererejected by their peers at the time is trotted out,with the unspoken implication that they areamongst these luminaries. What is forgotten is thatfor every correct idea before its time there are tensof thousands which are wrong.

What particularly saddens me is that whenever anew idea comes out, such as measles and Crohn’s, anumber of other investigators try to repeat the work.The fact that no measles genetic material can be de-tected in Crohn’s disease by PCR although it is al-leged to be detectable by immunostaining is enoughto set the alarm bells ringing. After the initial publi-cation, there is usually an unseemly debate in theletters section of the Lancet, and as the negative pa-pers start to appear in other journals, more corre-spondence about why the original observation couldnot be repeated. The idea then fades like the cheshirecat. Overall however, my major thought is of thesheer waste of time and effort in trying to confirmthe poor science invariably associated with these “breakthroughs”. Personally I refuse to become in-volved in any of this because I have enough prob-lems doing my own research without trying to con-firm someone elses half-baked ideas.

Whenever I get into a debate about the cause ofCrohn’s disease and recite the usual notions aboutthe normal flora and disordered T cell responses inthe gut wall, I try to base my argument largely onthe fact that in this area at least, different people indifferent places in the world get the same results.The mouse models are also very strong evidence.There is reproducibility of observations which thenform a firm basis for further studies. Looking at thetable it is also clear the idea that the normal floramight be important is quite old. But there is alwaysa whisper—remember Helicobacter pylori. Is it thefear of being the person who rejected the paper todiscover the cause of Crohn’s disease which explainsthe suspension of critical faculties in this area? Thisshould not matter, if it is true, the paper will getpublished somewhere, and anyway referees are usu-ally anonymous. I have always considered theH.pylori argument to be overstated. The failure torecognize H. pylori in the stomach says more aboutthe development of upper GI endoscopy and gas-troenterologists than anything else.

Mucosal Immunology Update 5

There is also a very real down-side to all of this.There is very little work funded by the major chari-ties or government on inflammatory bowel diseaseimmunology in the UK. The British Digestive Foun-dation supports some, but it is very minor league.Do the individuals who announce these health scaresand “breakthroughs” do it out of frustration at lackof funding and are they trying somehow to bypasspeer-review to some extent? Personally I think thatit exacerbates rather than remedies the problem. Itcannot be good for IBD research in the UK for thechief medical officer of health of England to writeto every doctor in the country saying that the fearsover MMR and Crohn’s are unfounded. I remembera few years ago, when I had a grant funded by theWellcome Trust, that I had to take out all of the workusing patient material and concentrate on mecha-nisms in model systems. There is a perception in theUK that gastroenterologists are not very academic(with apologies to those who are) and that the drugcompanies keep them well supported. As I writethis article, large numbers of UK-based gastroenter-ologists are being flown to Orlando for the AGA bypharmaceutical companies. The perception that GIresearch is somewhat inferior extends to others work-ing on mucosal inflammation, we are all tarred withthe same brush. Some aspects of mucosal immunol-ogy are well supported, such as mucosal vaccinationwhere there are a number of very eminent groups. Itis therefore not any particular bias against GI research,but specifically IBD immunology research.

This is a very bad situation because IBD is quitecommon in the UK and leads to substantial morbid-ity and economic cost to the country. Any new treat-ments which might help the patients have to bebased on understanding the disease mechanisms.Anti-TNF antibody is a very good example and willundoubtedly help a lot of patients. My only problemis that I cannot help but feeling that when I revisitthis area ten years hence, I will have a few morepapers to add and there will be a lot more healthscares. Then again, I could be wrong, perhaps, a causeof Crohn’s will be found, but I wouldn’t bet on it.Complex problems rarely have simple solutions.

Dendritic Cells areKey Regulators ofIntestinal ImmuneResponsesEilidh Williamson and Joanne L. VineyDepartment of Molecular ImmunologyImmunex Corporationemail: williame@immunex.com

jviney@immunex.com

The intestinal immune system has developedextremely elegant methods of immunoregulation.Healthy individuals are capable of mounting effec-tive immune responses to the many pathogenswhich may be encountered via the oral route, whileavoiding harmful inf lammatory responses toinnocous luminal antigens (Ag). Since we are toler-ant to the soluble dietary Ag we ingest with everymeal, the classical view of immunoregulation in the

Etiological agents in Crohn’s disease Journal

Granuloma-inducingtransmissable agents Lancet 1970 2:168

Transfer of disease to rabbitswith Crohn’s homogenates Lancet 1973 2:1120

Enterobacterial common antigen Gastro 1973 64:43Bacteroides Gastro 1974 66:1153Transfer of disease to rabbits

with Crohn’s homogenates Gastro 1975 69:618Granuloma-inducing

transmissable agents Lancet 1976 2:761Viral agent Lancet 1976 2:215Transfer of disease to rabbits

with Crohn’s homogenates Gut 1977 18:360Reovirus-like agent Lancet 1977 1:1169Granuloma-inducing

transmissable agents Gastro 1978 75:637Cell-wall defective Pseudomonas Gastro 1978 75:368Mycobacteria Lancet 1978 2:693Chlamydia Lancet 1979 1:19Mycobacteria Lancet 1979 1:444Eubacterium and Peptostreptococcus Anton van Leewenhock 1980 46:587Bacterial L-forms Gastro 1983 85:364Mycobacterium paratuberculosis Dig Dis Sci 1984 29:1073Coprococcus Eur J Immunol 1985 15:860Wolinella Infect and Immun 1986 53:671Mycobacterium paratuberculosis Gut 1992 33:1209Measles infection J Med Virol 1993 39:345Bakers yeast or related antigen Digestion 1994 55:40Perinatal measles infection Lancet 1994 344:508Mycobacterium paratuberculosis Gut 1994 35:506Listeria, E. coli and streptococci Gastro 1995 108:1396Measles vaccination Lancet 1995 345:1071In utero measles Lancet 1996 348:515

6 June 1999

gut is that soluble protein Ag elicit tolerance, whereasparticulate Ag generate active protective immunity.In reality, this is rather an over-simplification sincemost of the population are tolerant to their owncommensal bacterial flora, which are clearly non-soluble, non-dietary Ag. The fact that we can mountvigorous responses to foreign microorganisms andeven the commensal bacterial flora of others, whileremaining tolerant to our own commensal flora1,serves to highlight the extent to which intestinalimmune responses are regulated. It is clear that theseregulatory processes are paramount to healthy liv-ing. Common consequences of toler-ance breakdown are food allergies andinflammatory bowel disease (IBD),both of which can cause morbidityand, in severe cases, mortality.

Overall, the field of intestinalimmunoregulation is a fascinating areagiving rise to many pertinent questions.How does the intestinal immune systemcarry out this important homeostaticfunction and at what level is it regulated?What are the relative roles of the differ-ent antigen presenting cells (APC) in thegut and is it possible that the same APCcan function to promote both conse-quences of antigen feeding—toleranceand immunity. These are just a few ofthe many types of questions that stillneed to be addressed in mucosal immu-nology. In this review we will summa-rize some of the current views and hy-potheses that abound today.

Antigen presenting cells inthe intestine

The organized and diffuse compart-ments of the gut-associated lymphoidtissues (GALT) contain a variety of dif-ferent types of professional APC, in-cluding dendritic cells (DC), B cells and macroph-ages, as well as less conventional APC, such as intes-tinal epithelial cells (IEC). The Peyer’s patches (PP)are classically considered to be the major site forimmune induction in the intestine and it is gener-ally assumed that many luminal Ag gain access tothe GALT via M cells in the PP epithelium. PP con-tain many DC-like cells, which are particularly preva-lent beneath the dome area of the PP, as well as in

the interfollicular T cell zones of the PP2, 3, in addi-tion to B cells and macrophages. However, since thesurface area of the PP is small in comparison to thatof the villous epithelium, it is probable that otheralternative routes are important for Ag entry and sub-sequent presentation. Soluble Ag in particular mayenter the gut through conventional enterocytes bytranscytosis or endocytosis. Ag entering via theseroutes is likely to come into contact primarily withlamina propria (LP) APC. In the LP, there is an exten-sive network of MHC Class II+ cells with the mor-phology of DC4-6. MHC Class II+ cells with DC-like

morphology have also been detected residing be-tween IEC7, although whether these cells are DCremains unclear and their function has not beenstudied in detail. It should also be noted that MHCClass II+ IEC can themselves present Ag to T cells8-10,although whether and how epithelial cells can actas APC for MHC class II-restricted CD4+ T cells re-mains controversial. For the purposes of this over-view, we will focus on the phenotype and function

Mucosal Immunology Update 7

of DC in the gut and their role in tolerogenic andimmunogenic Ag presentation (Fig. 1).

Mucosal DCIn the small intestine, various DC populations have

been described in both the PP and LP2-6. In the PP,two distinct populations can be distinguished usingthe DC-reactive monoclonal antibodies N418, NLDC-145, M342 and 2A12, 3. The first DC population ap-pears to be relatively undifferentiated, is negative forNLDC-145 and M342 and is found in the subepithe-lial dome region of the PP and throughout the fol-licle, sparing the germinal centre. The second, an in-terdigitating population, appears to be more matureand is found in the interfollicular T cell regions of thePP and reacts with antibodies to all four DC-specificmarkers. Functional studies reveal that PP DC canprime T cells after pulsing with Ag in vitro, or afteroral administration of Ag in vivo3, 6. However, it shouldbe noted that the DC described in those studies maywell have become activated during culture in vitro,thereby altering their intrinsic functional properties.Although DC have also been described in the LP, theirpaucity has hindered detailed analysis of their phe-notype. In order to learn more about the nature ofDC in the gut and in the LP in particular, we haverecently utilized Flt3 ligand (Flt3L), a growth factorthat dramatically expands DC in vivo. Utilization ofFlt3L to expand DC has permitted us to phenotypi-cally and functionally analyze intestinal DC in theirnatural, unmanipulated environment 11.

Effects of Flt3L on intestinal DCMice treated with Flt3L for 10 consecutive days

show a dramatic increase in the levels of CD11c+veMHC class II+ DC in both the organized and diffuselymphoid tissues. In the LP, increased DC numbersare visible in both the villous and crypt regions11.In the PP, the distinct populations of DC localizedto the dome and interfollicular regions are equallyexpanded, with no apparent preferential increasein either subpopulation11. From our studies, it ap-pears that Flt3L simply expands mucosal DC, result-ing in more abundant DC populations which main-tain exactly the same phenotype as are present inthe normal intestine. It is also clear that Flt3L doesnot induce DC activation. Multi-color f lowcytometrical studies have demonstrated that the ma-jority of freshly isolated cells from GALT of Flt3L-treated mice express high levels of MHC Class II and

low levels of the costimulatory molecules CD80 andCD8611; 12—a phenotype consistent with mature, rest-ing DC. Despite their resting phenotype, these cellsare fully functional, as evidenced by their ability toprocess and present Ag in in vitro Ag loading assays(Viney, unpublished observations). In further sup-port of the functionality of Flt3L-expanded DC,we have observed that these cells are fully respon-sive to inflammatory signals in vitro and in vivo.Such inflammatory stimuli promote DC activationand g ive r ise to the associated increase incostimulatory molecules11, 12

Mucosal DC are intrinsically tolerogenicAPC in vivo

As we have discussed, there are many indicationsthat DC are constitutively present throughout theGALT, yet the default response to orally administeredAg is tolerance, not activation. This may indicate oneof several possible scenarios i) that intestinal DC arenon-functional in normal, healthy individuals ii) thatalternative intestinal APC types are preferentiallyinvolved in tolerance induction or iii) that intestinalDC can function in a tolerogenic manner. In orderto test these hypotheses and examine whether theincreased numbers of DC would influence toleranceinduction, we treated mice with Flt3L to expandintestinal DC prior to feeding soluble ovalbumin(OVA)11. We found that mice treated with Flt3L priorto OVA feeding exhibit more profound tolerancethan that seen in equivalent OVA fed control mice.The increased tolerance in Flt3L-treated animals isevident in terms of both DTH and Ag-specific prolif-erative responses. In addition, Flt3L-treated miceshow greater inhibition of OVA-specific serum IgG1and IgG2a Ab titres after OVA feeding than equiva-lent PBS controls. In all cases, the enhanced toler-ance exhibited by Flt3L-treated mice is most dra-matic in mice fed low doses of Ag which are gener-ally ineffective in control animals.

These results indicate that DC may play a pivotalrole in tolerogenic antigen presentation in the in-testine. The fact that there does not appear to be apreferential increase in any individual DC subset af-ter Flt3L-treatment, or any deviation from the nor-mal DC tissue localization, suggests that the en-hanced tolerance observed in Flt3L-treated mice ismost likely attributable to the increased probabilityof a naive Ag-specific T cell coming into contact with

8 June 1999

a tolerogenic DC in the intestine (Fig. 2). Our find-ings are, however, in direct contrast to previous re-ports that intestinal DC isolated from Ag-fed miceare immunostimulatory in vitro13. We believe thatthese apparently disparate results may simply reflectthe difference between analyzing DC function insitu using Flt3L and analyzing isolated DC in vitro,since the procedure of removing DC from their lo-cal tissue microenvironment is known to promoteDC activation14, 15.

of immune reactivity occurring in a normal intes-tine at any given time. This raises the question ofwhether intestinal DC are in some way functionallydistinct from their peripheral counterparts and areperhaps unable to induce active immunity. An alter-native question, which we have chosen to addressexperimentally, is “can the regulation of bothtolerogenic and immunogenic responses be medi-ated by this single intestinal APC type?” (Fig. 1). Wedecided to test whether intestinal DC expanded byFlt3L are intrinsically tolerogenic, or whether theycan support active immune responses by examin-ing responsiveness in Flt3L-treated mice followingimmunization with the potent mucosal immunogenand adjuvant cholera toxin (CT). We found that Flt3L-expanded DC can not only support, but actuallyenhance the protective response to CT 12. Comparedwith PBS-treated controls, Flt3L-treated mice thathave been orally immunized with a sub-optimal doseof CT show dramatically increased CT-specific pro-tection against subsequent intestinal challenge. Moststrikingly, Flt3L-treated animals exhibit significantanti-CT protective responses when immunized withvery low doses of CT which are essentially ineffec-tive in PBS-treated controls. Furthermore, the in-creased resistance to CT-challenge afforded by Flt3Ltreatment is accompanied by enhanced local andsystemic CT-specific IgA Ab levels.

How can intestinal DC be both tolerogenicand immunogenic APC?

The results described above clearly indicate thatboth immunogenic and tolerogenic intestinal re-sponses can be heightened in the presence of in-creased numbers of intestinal DC, but how is thisachieved? Recently, it has become apparent that thelevels of the costimulatory molecules CD80 andCD86 on the APC surface can critically determinewhether the outcome of an antigenic encounterleads to tolerance or active immunity16, 17. APC whichexpress low levels of CD80/86 appear to promote Tcell tolerance by preferentially signalling throughthe high affinity CTLA-4 receptor on T cells, whileAPC expressing high levels of CD80/86 deliver posi-tive, stimulatory signals to the T cell via the CD28receptor on T cells16. We have evidence to suggestthat CT can enhance CD80/86 expression on Flt3L-expanded intestinal DC, thereby inducing their acti-vation12. Thus, we believe that the augmented anti-

Intestinal DC can be converted fromtolerance promoting APC intoimmunogenic APC in vivo

Although our findings indicate that the probabledefault activity for intestinal DC is tolerogenic anti-gen presentation following oral administration ofsoluble protein Ag, there is clearly an inordinate level

Mucosal Immunology Update 9

CT protection observed in Flt3L-treated mice immu-nized perorally with CT can most likely be attrib-uted to the increased number of DC now capableof providing an efficient costimulatory signal to in-testinal T cells through the CD28 receptor. In fur-ther support of this hypothesis, we have found that acti-vating Flt3L-expanded DC with the proinflammatorycytokine, IL-1a at the time of feeding soluble OVAmimics the adjuvant effects of CT in vivo. As such,treating mice with Flt3L in combination with IL-1aovercomes the default tolerogenic response normallyassociated with Ag feeding and instead promotes apowerful immunogenic response.

SummaryThese studies suggest that DC may be well-posi-

tioned to regulate the qualitative nature of intesti-nal immune responsiveness and provide us with aninteresting working model (Fig. 1). As such, intesti-nal DC which encounter Ag in the absence of aninflammatory signal elicit the default response toan orally administered Ag i.e. profound systemic tol-erance. Naturally, in the presence of greatly expandedDC numbers, such as after Flt3L-treatment, the levelof tolerance induced by Ag feeding is dramaticallyenhanced (Fig. 2). In contrast, if intestinal DC en-counter Ag together with an appropriate inflamma-tory signal, such as that delivered by CT, or IL-1a,the DC become activated to express high levels ofthe costimulatory molcules CD80/86 and the out-come of the response is active immunity (Fig. 1).Again, this response is exaggerated in the presenceof increased DC numbers, leading to increased re-sponsiveness (Fig. 2).

This model has important implications in termsof mucosal adjuvant design. Clearly, a major obstaclein the design of mucosal vaccines is that most solubleAg encountered via the oral route promote toler-ance, unless administered with an appropriate adju-vant. Although CT is widely recognized as a power-ful immunogen and adjuvant that can preventtolerization to coadministered soluble proteins inrodents18-20, it is not well tolerated in humans. Mu-tant CT molecules have been generated that re-tain adjuvant activity without the associated tox-icity 21, 22. Nevertheless, these molecules are immu-nogenic when administered to research animals.This may prevent their repeated use as mucosaladjuvants, since pre-existing immunity to CT re-

duces its effectiveness as an adjuvant. There is there-fore a need to develop alternative, safe and effectivemucosal adjuvants. Our studies suggest that Flt3L,used in conjunction with an inflammatory media-tor, such as IL-1a, may be a reagent useful in thedesign of mucosal immunization strategies.

AcknowledgmentsThe authors would like to thank Gary Carlton

for excellent assistance with graphics for Figures1 and 2.

References1. Duchmann R, Kaiser I, Hermann E, et al. Clin. Exp.

Immunol. 1995; 102: 448-455.

2. Ruedl C, Rieser C, Bock G, Wick G and Wolf H, Eur. J.Immunol.1996; 26:1801-1806.

3. Kelsall B, and Strober W, J. Exp. Med. 1996; 183, 237-247.

4. Pavli E, Woodhams CE, Doe WF, and Hume DA, Immunol-ogy 1990; 70:40-47.

5. Mayrhofer G, Pugh CW, and Barclay AN, Eur. J. Immunol.1983; 13:112-122.

6. Liu L, and MacPherson GG, Immunology 1995; 85:88-93.

7. Maric I, Holt PG, Perdue MH, and Bienenstock J,J. Immunol. 1996; 156: 1408-1414.

8. Kaiserlian D, Vidal K, and Revillard JP, Eur. J. Immunol.1989; 19:1513-1519.

9. Mayer L, and Shlien R, J. Exp. Med. 1987; 166:1471-1483.

10. Bland PW, and Warren LG, Immunology 1986; 58:9-14

11. Viney JL, Mowat AM, O’Malley JM, Williamson E, and FangerNA, J. Immunol. 1998; 160:5815-5825.

12. Williamson E, and Viney JL, submitted 1999.

13. Liu L, and MacPherson GG, J. Exp. Med. 1993; 177:1299-1307.

14. Zhou LJ, and Tedder TF, J. Immunol. 1995; 154P: 3821-3825.

15. O’Doherty U, Peng M, Gezelter S, et al. Immunology 1994;82:487-493.

16. Perez VL, Van Parijs L, Biuckians A, et al. Immunology1997; 6:411-417.

17. Samoilova EB, Horton JL, Zhang H, et al. Int. Immunol .1998; 10: 491-498.

18. Elson CO, and Ealding W, J. Immunol. 1984;133: 2892-2897.

19. Elson CO, and Ealding, W, J. Immunol. 1984; 132:2736-2741.

20. Lycke N, and Holmgren J, J. Immunol. 1986; 59:301-308.

21. Fontana MR, Manetti R, Ciannelli V, et al. Infect. Immunity.1995; 63:2356-2360.

22. Yamamoto S, Takeda Y, Yamamoto M, et al. J. Exp Med. 1997;185:1203-1210.

10 June 1999

Commentary

How InflammationOvercomesSuppression

Inhibition of transforming growth fac-tor-b/SMAD signalling by the interferon-g/STAT pathway.

Luis Ulloa, Jacqueline Doody andJoan Massague

Cell Biology Program and HowardHughes Medical Institute, MemorialSloan-Kettering Cancer CenterNew York, New York 1002, USA

Transforming growth factor-b (TGF-b) andinterferon-g (IFN-g) have opposite effects ondiverse cellular functions, but the basis for thisantagonism is not known. TGF-b signals througha receptor serine kinase that phosphorylatesand activates the transcription factors Smads 2and 3, whereas the IFN-g receptor and its asso-ciated protein tyrosine kinase Jak1 mediatephosphorylation and activation of the transcrip-tion factor Stat1. Here we present a basis forthe integration of TGF-b and IFN-g signals. IFN-g inhibits the TGF-b-induced phosphorylationof Smad3 and its attendant events, namely, theassociation of Smad3 with Smad4, the accumu-lation of Smad3 in the nucleus, and the activa-tion of TGF-b-responsive genes. Acting throughJak1 and Stat1, IFN-g induces the expression ofSmad7, an antagonistic SMAD, which preventsthe interaction of Smad3 with the TGF-b recep-tor. The results indicate a mechanism of trans-modulation between the STAT and SMAD sig-nal-transduction pathways.

Nature 1999, 397: 710-713

CommentTransforming growth factor (TGF)-b is an impor-

tant cytokine at mucosal surfaces and throughoutthe body. TGF-b knock out mice develop chronicinflammation in many tissues, including the gas-trointestinal tract1. TGF-b is produced by several cell-types in the intestinal mucosa including epithelialcells and macrophages. Its expression is particularly

increased in inflamed mucosa of patients with ul-cerative colitis (UC) and Crohn’s disease (CD)2, lo-calized mostly to inflammatory cells of the laminapropria. It is considered that TGF-b is also very im-portant in modulating epithelial cell restitution. TGF-b is also known as an antigen-non-specific suppres-sor cytokine, released from the regulatory T cellsthat mediate active suppression against orally admin-istered antigen. In studies of multiple sclerosis (MS)patients, short term cultures of blood lymphocytesresulted in an increase in the frequency of MBP spe-cific T cells that secreted TGF-b1 (Th 3 cells) in theMBP-fed patients compared with that of the nonMBP-fed patients3. There are a large number of mousemodels of IBD. Most of these models show a largeincrease in cells with a Th1 type cytokine pattern.Transfusing CD4+CD45RBhi T cells into SCID micecauses colitis, which can be prevented by the simul-taneous infusion of CD4+CD45RBlo cells. The pro-tective effect of the RBlo cells is inhibited by anti-TGF-b, but not anti-IL-44. Mucosal administration of

2,4,6-trinitrobenzene sulfonic acid (TNBS) also in-duces IBD like colitis, with infiltration of Th1 type Tcells. However, feeding haptenated protein abrogatesthis Th1 responsiveness and increases productionof TGF-b by PP and lamina propria T cells5. Diseaseis exacerbated in tolerant mice by treatment withanti-TGF-b antibody5, 6. These studies indicate thatinflammatory responses in the gut can be regulatedthrough the balance between IFN-g/Th1 cell re-sponses and TGF-b/Th3 cell responses.

In this article, Ulloa et al have worked out howIFN-g can overcome the immunosuppressive effectof TGF-b. It was, first, shown that IFN-g inhibits theTGF-b signal-transduction pathway in U4A/Jak1 cellswhich expressed Jak1 (IFN-g receptor associatedprotein tyrosine kinase) and Stat1 (transcription fac-tor of IFN-g). Second the antagonistic SMAD, Smad7,was switched on after administration of IFN-g. IFN-g also increased the level of Smad7 bound to theTGF-b receptor complex. Since Smad 3 phosphory-lation and its interaction with Smad4 is a key eventin TGF-b signalling, inhibition of Smad 3 binding tothe TGF-b receptor complex by Smad 7 effectivelydown-regulates the transmission of information fromthe TGF-b receptor to the nucleus. IFN-g inhibition

Mucosal Immunology Update 11

requires de novo protein synthesis, and was inde-pendent of MAP kinase Erk. Treatment of cells witha Smad7 antisense oligonucleotide, prevented theIFN-g mediated increase in Smad7 and the effectof IFN-g on Smad3 phosphorylation. These resultstherefore demonstrated that IFN-g signallingthrough the Jak1/Stat1 pathway rapidly increasesthe expression of Smad7, causing the inhibition ofTGF-b-mediated Smad3 phosphorylation and anattendant loss of TGF-b signalling to the nucleus.This paper nicely illustrates the inhibitory effectof IFN-g/STAT1 signalling on TGF-b/SMAD pathway,and may help to explain the signalling pathwaysthat abrogate tolerance.

In the context of GI disease in man, IFN-g can beseen to have two related effects. It is highly pro-in-flammatory, inducing the production of TNF-a andIL-1b by macrophages, and at the same time, itswitches off homeostatic immunosuppressive path-ways mediated by TGF-b. Understanding the molecu-lar basis for this now will allow the development ofdrugs which can prevent the activity of SMAD7thereby indirectly inhibiting the inflammatory ef-fects of interferon-g.

References1. Shull MM, Ormsby I, Kier AB, Pawlowski S, Diebold RJ, Yin

M, Allen R, Sidman C, Proetzel G, Calvin D. et al. Nature1992;359:693-699.

2. Babyatsky NW, Rossiter G, Podolsky DK, Gastroenterology1996;110:975-984.

3. Fukaura H, Kent SC, Pietrusewicz MJ, Khoury SJ, Weiner HL,Hafler DA, J. Clin. Invest. 1996;98:70-77.

4. Powrie F, Carlino J, Leach NW, Mauze S, Coffman RL.,J. Exp. Med. 1996;183:2669-2674.

5. Neurath MF, Fuss I, Kelsall BL, Stuber E, Strober W,J. Exp. Med. 1995;182:1281-1290.

6. Ludviksson BR, Ehrhardt RO, Strober W,J. Immunol. 1997;159:3622-3628.

Yoshikazu Ohtsuka, M.D.Dept. of Paediatric Gastroenterology,St.Bartholomew’s and the Royal LondonSchool of Medicine and Dentistry, London

Dendritic Cells,Mucosal Immunityand PeripheralToleranceFang-Ping Huang and G. Gordon MacPhersonSir William Dunn School of Pathology,University of Oxford, South Parks Road,Oxford OX1 3RE, United Kingdom

A classical and fundamental question which im-munologists are still finding hard to answer is howmucosal tolerance versus immunity is precisely regu-lated. It is far from clear, in the gut for example, whatthe exact regulatory mechanism or mechanisms arewhich determine the generation of protective im-mune responses to mucosally-related pathogens orharmful substances while maintaining a profoundtolerance to numerous food antigens (Ag) and tonormal gut flora. This has become particularly in-triguing following the demonstration that dendriticcells in the gut are capable of acquiring orally- orintestinally-administered soluble antigens (Ag) andcan subsequently activate sensitized T-cells in vitroand naive T-cells in vivo 1,2.

Mucosal immunity versus tolerance:still a ‘gut feeling’

Oral tolerance, one form of peripheral tolerance,is well-recognized but poorly understood. Oral ex-posure to soluble Ag may result in systemic immu-nological hyporesponsiveness, although under cer-tain conditions it leads to systemic priming. Thisappears to reflect, at least in part, the special way inwhich the immune system operates in the gut. Oraltolerance has therefore been suggested as one ofthe mechanisms to explain life-long immunologicaltolerance to dietary Ag 3 and similar mechanisms mayregulate immune responses in other mucosal tis-sues 4. However, many suggested mechanisms for oraltolerance do not seem to provide satisfactory ex-planations. The association of oral tolerance withlocal mucosal immune responses, e.g. IgA produc-tion5,6 to soluble Ag remains controversial7. The roleof gd T-cells in maintaining mucosal tolerance hasbeen emphasized and TCR gd knock-out mice haveclear defects in oral tolerance 8, yet gd T cells from

12 June 1999

the gut may also abrogate oral tolerance9. Moreover,dose-dependent tolerance (i.e. high dose tolerance)or priming certainly does not alone explain thisphenomenon. Assuming that the ability of the im-mune system to differentiate the majority of foodAg from harmful ones is not genetically programmed,the control of mucosal responses would have to relycritically on the nature and the doses of Ag upontheir initial encounter with the system. Unless moth-ers ensure that the food they are giving to their ba-bies for the first time contains multiple antigens, allat high enough doses (“tolerizing” dose), systemicimmune responses to these Ag upon their subse-quent encounter would inevitably follow. This isclearly not the case since food allergies, which arebelieved to be the result of a breakdown of oral tol-erance, occur only in a minority of children (2% in-tolerance to cow’s milk in Europe)10 despite expo-sure to a full range of dietary Ag at various concen-trations in baby food. Although the introduction ofcow’s milk-based infant formula is thought to bepossibly associated with a higher risk of type I dia-betes, only those genetically predisposed to the dis-ease would be at risk11. Hence, mechanisms mustexist at birth to regulate neonatal responses to di-etary Ag that are maternally derived or contained inartificially formulated baby food in the cases ofbottle-fed babies. Interestingly, evidence from ani-mal models indicates that in neonatal rats, systemicsensitization rather than tolerance may be inducedafter the administration of certain oral Ag12. Sincethis inability to induce tolerance can be partiallyrestored with adult splenocytes, it has been attrib-uted to a regulatory imbalance in the neonatal gut.Taking together the fact that food-allergic diseasesare indeed more a feature of childhood, it is obviousthat such innate regulatory mechanisms also requirea post-birth host maturation process.

DC heterogeneity and functionsDendritic cells are not only the most potent Ag

presenting cells but also the only cell type that iscapable of activating naive T-cells in vivo13. It is nowwell established that DC acquire Ag in peripheraltissues and transport them to draining lymph nodeswhere processed Ag are presented to T-cells to ini-tiate immune responses. Until recently, DC have beenportrayed solely as initiators of immune responses.

However there is now increasing evidence that thefunctions of DC are much more complex than pre-viously thought18.

DC differing in their stages of maturation, expres-sion of cell surface markers, functional propertiesand tissue/organ origins have been described bothin vitro and in vivo. Langerhans cells (LC) may bedistinct from tissue DC19 and DC found in marginalzones of the spleen differ from those in T cell ar-eas 20. Recently, three populations of DC have beenisolated from murine lymph nodes21. The most im-portant aspect of DC classification is its relevanceto their functional differences. As opposed to ma-ture DC, immature DC are actively endocytic. Thisability, which is subsequently lost upon maturation,enables them to acquire Ag in peripheral tissues ef-fectively. DC of a mature phenotype however ex-press highly upregulated cell surface MHC class II,as well as MHC class I, and co-stimulatory moleculessuch as CD80 and CD86. The latter are known to becrucial for effective Ag presentation. The difficultywith understanding the relationships between thesedifferent populations of DC arises because DC un-dergo dramatic phenotypic and functional changesduring maturation and activation22,23 and even aftershort periods in culture24,25. It is usually not clearwhether identified DC subpopulations representdistinct lineages or stages of maturation within asingle lineage.

In the mouse, however, it is now believed that atleast two distinct lineages with different ontogenicorigins exist in vivo. These comprise the “classical”DC described by Steinman26, and the CD8aa DCfound in thymus and other lymphoid organs that cankill CD4+ T cells in a Fas/FasL-dependent manner27. Inhumans, CD34+ bone marrow-derived DC and periph-eral blood monocyte-derived DC have been well docu-mented although their lineage origins are yet to bedefined. Nevertheless, CD4+CD11c-CD3-CD45RA+

cells purified from human tonsil have been shown todevelop into DC after culture with IL-3 and CD40L(28), conditions known to give rise to lymphoid DCor DC2 in the mouse29.

Phenotypically and functionally distinctDC subsets in the gut

Functional DC have been identified in many mu-cosal tissues. These include the Peyer’s patches 30, 31,

Mucosal Immunology Update 13

the mesenteric nodes and the lamina propria 2, 32; theairway mucosa23 and the urinary tracts 34.

In rat, we have observed two phenotypically andfunctionally distinct subsets of DC in lymph (L-DC)draining the gut35. These cells constitutively and con-tinuously migrate from the intestine to the me-senteric nodes. Collected under near-physiologicalconditions without enzymatic treatment or incuba-tion above 4oC, L-DC can acquire oral Ags understeady-state conditions and present them to naiveT-cells 1, 2, 35. The two L-DC subsets differ in co-expres-sion of CD4 and OX41. CD4+OX41+ L-DC (CD4+ L-DC for short) are stronger APC for naive as well assensitized T cells, and survive longer in culture. CD4-

OX41- L-DC (CD4- L-DC for short) are weak APCwhen freshly isolated and survive poorly in culture.They lose the ability to process/present native Agcompletely after brief culture, but their potency asstimulators of an allogeneic MLR is increased. Oneof the characteristic differences of the CD4- L-DC isthat they contain large cytoplasmic inclusions andstrikingly high levels of non-specific esterase (NSE).We have recently obtained strong evidence that CD4-

L-DC contain apoptotic cell debris derived from in-testinal epithelial cells (IEC) and that the NSE in thesecells is derived from IEC (MacPherson et al., in prepa-ration). Under steady conditions, CD4- L-DC migrateinto T cell areas of mesenteric lymph nodes, areasfrom which CD4+ L-DC may be excluded(MacPherson et al., unpublished data). These cellsare, therefore, constitutively transporting oral as wellas self Ag to lymph node T-cell areas - the very loca-tion where peripheral tolerance is believed to takeplace. It is of interest that freshly collected cells ofboth populations appear to be DC of mature phe-notype with high surface MHC class II and B7 ex-pression. The lineage origins of the CD4+and CD4-

L-DC are however currently unclear.

Potential roles of DC in mediating self-and oral-tolerance

Mechanisms that have been described or sug-gested for peripheral tolerance include ignorance,suppression, anergy caused by direct interactionwith peripheral Ag in the absence of co-stimulation,and activation-induced cell death (AICD) involvingprofessional APC 36,37 which is also known as crosstolerization (reviewed by Heath et al.)18.

Aberrant presentation of Ags by non-professionalAPC, such as enterocytes, lacking co-stimulatorymolecules may be one explanation for the induc-tion of oral tolerance3 but it is difficult to under-stand how naive T-cells which recirculate throughlymph nodes would come into contact with thesecells. Alternatively, T-cell tolerance to parenchymalself-antigens has also been thought to be explicableby “ignorance” 38,39. It was thus of great interest whenspecific T-cell tolerance to a viral haemagglutininexpressed as a self Ag on parenchymal cells wasshown to require a bone marrow-derived APC, ratherthan parenchymal cells expressing the peptide-MHCcomplexes17. It was concluded that this type oftolerance induction involves professional APC andactivation of T cells prior to their tolerization. In-deed, T-cells tolerance to hen egg lysozyme ex-pressed as self-Ag is also Fas-dependent40 andCTLA-4 signalling by CD80/CD86 is evidently in-volved in the induction of peripheral tolerance41.It is noteworthy that both the CD4+ and CD4- L-DCsubsets in the rat are of mature phenotype by thetime they reach the draining mesenteric nodes. Wesuggest that these mature CD4- L-DC, under steadyconditions carry oral and self Ag and have directaccess to naive T-cells, and are thus able to induceand to maintain tolerance via apoptosis42. It wouldbe interesting to study the expression of functionalmolecules such FasL on these cells, particularly thoseof the CD4-OX41- phenotype.

A new model of DC-mediated self-tolerance basedon the two DC lineages identified in mouse26 hasrecently been put forward formally by Fazekas de StGroth43. According to the theory, immunogenic ortolerogenic properties of DC can be linked directlyto cells of the two lineages: the myeloid DC(CD8a -, 33D1+, DEC-205lo) being immunogenicand the lymphoid DC (CD8a+, 33D1-, DEC+205hi)being tolerogenic. The important question thenbecomes how these cells, or L-DC of CD4- andCD4+ phenotypes found in the rat, are selectivelyinduced or regulated? An inflammatory “danger”signal44 mediated via DC is a possible answer.Fazekas de St Groth’s model takes into account themicroanatomical importance of the lymphoid or-gans, thus offering a plausible explanation as to howDC might mediate the induction of tolerance ver-sus immune responses: CD8a+ DC found in T-cell

14 June 1999

areas under steady conditions are important in tol-erance induction; myeloid CD8a - DC enter T areasto initiate an immune response only after they re-ceive the “danger” signals. Thus, a germline-encodedproperty of these two subsets of DC with distinctontogenic origins is now postulated to explain theDC-mediated self/non-self or normal/danger dis-crimination.

It has been well documented that DC migrationfrom skin45, small intestine46 and solid organs47 canbe stimulated by LPS, TNFa and IL-1. DC can beinduced to migrate from hepatic sinusoids to co-eliac nodes by particulate stimuli48. Within thespleen, LPS or parasitic Ag induces migration of DCfrom the marginal zone into T cell areas49,50. Thus,another question follows: do DC remain in periph-eral tissues until an inflammatory signal stimulatesthem to migrate? In the gut, as mentioned previ-ously, DC are continually migrating from periph-eral tissues to lymph nodes in the absence of in-flammation. Two subpopulations of L-DC migrateconstitutively in bovine skin lymph51. This is alsoevidenced by the presence of DC in all mamma-lian peripheral lymph sampled, including renal andhepatic lymph in normal sheep52-54. This steady statemigration is presumably essential for the inductionof tolerance to dietary antigens. There has to be amechanism by which DC can migrate and intro-duce food Ag to the immune system without rely-ing on inflammation. It can be argued however that,at least in the gut, this migratory process might beregulated by normal flora. To answer this defini-tively, further study on CD4- and CD4+ L-DC migra-tion using germ-free animals, which have beenshown to have defects in the induction of toler-ance to oral Ag55, would be informative. Alterna-tively, inflammatory signals might modify DC prop-erties in terms of switching from one phenotypeto the other. Under steady state, CD4- L-DC arefound to be the dominating cell type (about 75%)of the DC migrating constitutively in the pseudo-afferent lymph. It will be interesting to determinehow the ratio of CD4+ to CD4- L-DC and their otherfunctional characteristics change following stimu-lation by inflammatory stimuli.

No matter which one, or combination, of theabove mechanisms is responsible for the selectiveinduction of DC mediated tolerance versus immu-

nity, a process of dynamic balance within the im-mune system is clear —another example explainedperfectly by the ‘Ying/Yang’ theory. Upon stimula-tion by bacterial products, the so-called myeloidDC have been shown to enter T-cell areas, but anearly increase in lymphoid DC in the same areas isalso evident49. Similarly, in response to LPS or para-site extracts (e.g. Toxoplasma gondii) most of theactivated DC (CD11c+), as indicated by IL-12 ex-pression, detected in T-cell areas were ofCD8a+DEC+205hi phenotype50, although it was notclear how the cell ratio between CD8a+ and CD8a-

might be changed. Cytokines are likely to be themediators for the reciprocal control of DC and T-helper cell differentiation29. New evidence hasemerged indicating that differential regulation ofimmunity and tolerance is associated with a bal-ance of signalling between CD80 and CD86 respec-tively, and such a balance is also reciprocally con-trolled by Th1 and Th2 cytokines56. With the recentfinding of HLA-G, a tolerance-related MHC class Imolecule which is expressed on DC and can beregulated by IL-10 and IFNg57, there is little doubtthat DC, apart from their role in initiating protec-tive immunity and in central tolerance, are also cru-cial mediators of many forms of peripheral toler-ance by providing a close but highly regulated linkbetween innate and adaptive immune mechanisms.

AcknowledgementWe wish to thank Emma Turnbull for her help-

ful discussion and critical proof reading.

References1. Liu LM., and MacPherson GG, Immunology 1991;

73:281-286.

2. Liu LM., and MacPherson GG, J. Exp. Med. 1993;177:1299-1307.

3. Strobel S. and Mowat AMcl., Immunol. Today 1998;4:173-181.

4. McMenamin C, Schonhegrad M, Oliver J, Girn B and HoltPG, Int. Arch. Aller. A. Imm. 1991; 94:78-82.

5. Challacombe SJ, and Tomasi TBJ, J. Exp. Med. 1980;152:1459-1472.

6. Richman LK, Graeff AS, Yarchoan R and Strober W,J. Immunol. 1981; 126:2079-2083.

7. MacDonald TT, Eur. J. Immunol. 1983; 13:138-142.

8. Ke Y, Pearce K, Lake JP, Ziegler HK, and Kapp JA,J. Immunol. 1997; 158:3610-3618.

9. Fujihashi K, Taguchi T, McGhee JR, Elderdge JH, Bruce MG,Green DR, Singh B, and Kiyono H, J. Immunol. 1990;145:2010-2019.

10. Madsen C, Enviromental Toxicology and Pharmacology1997; 4:163-167.

Mucosal Immunology Update 15

11. Kolb H, and Pozzilli P, Immunol. Today 1999; 20:108-110.

12. Miller A, Lider O, Abramsky O, and Weiner HL,Eur. J. Immunol.1994; 24:1026-32.

13. Banchereau J, and Steinman RM, Nature 1998; 392:245-252.

14. Crowley M, Inaba K, and Steinman R, J. Exp. Med. 1990;172:383-386.

15. Finkelman FD, Lees A, Birnbaum R, Gause WC, and MorrisSC, J. Immunol. 1996; 157:1406-1414.

16. Viney JL, Mowat AM, Omalley JM, Williamson E, and FangerNA, J. Immunol. 1998; 160:5815-5825.

17. Adler A, Marsh DW, Yochum GS, Guzzo J, Nigam A, NelsonWG, and Pardoll DM, J. Exp. Med. 1998; 187:1555-1564.

18. Heath WR, Kurts C, Miller JFAP, and Carbone FR,J. Exp. Med. 1998; 187:1549-1553.

19. Schuler G, Romani N, and Steinman RM,J. Invest. Dermatol. 1985; 85:99.

20. Metlay JP, Witmer-Pack MD, Agger R, Crowley MT, LawlessD, and Steinman RM, J. Exp. Med. 1990; 171:1753-1771.

21. Salomon B, Cohen JL, Masurier C, and Klatzmann D,J. Immunol. 1998; 160:708-717.

22. Schuler G, and Steinman RM, J. Exp. Med. 1985; 161:526-546.

23. Ruedl C, and Hubele S. Eur. J. Immunol. 1997; 27:1325-1330.

24. MacPherson GG, Immunology 1989; 68:102-107.

25. MacPherson GG, Fossum S, and Harrison B, Immunology1989; 68:108-113.

26. Steinman RM, Annu. Rev. Immunol. 1991; 9:271-296.

27. Suss G, and Shortman K, J. Exp. Med. 1996; 183:1789-1796.

28. Grouard G, Rissoan MC, Filgueira L, Durand I, Banchereau J,and Liu YJ, J. Exp. Med. 1997; 185:1101-1111.

29. Rissoan MC, Soumelis V, Kadowaki N, Grouard G, Briere F,de Waal Malefyt R, and Liu YJ, Science 1999; 283:1183-1186.

30. Spalding DM, Williamson SI, Koopman WJ and McGhee JR,J. Exp. Med. 1983; 157:1646-1659.

31. Kelsall BL, and Strober W, Ann. NY Acad. Sci. 1996; 778:47-54.

32. McWilliam AS, Nelson D, Thomas JA, and Holt PG,J. Exp. Med. 1994; 179:1331-1336.

33. Pavli P, Woodhams C, Doe W, and Hume D,Immunology 1990; 70:40-47.

34. Hart DNJ and Fabre JW, Transplantation 1981; 31:318-325.

35. Liu LM, Zhang M, Jenkins C, and MacPherson GG,J. Immunol. 1998; 161:1146-1155.

36. Heath WR, Allison J, and Miller, Autoimmunity 1993;15:293-298.

37. Kurts C, Heath WR, Kosaka H, Miller JF, and Carbone FR,J. Exp. Med. 1998; 188:415-420.

38. Ohashi PS, Oehen S, Buerki K, Pircher H, Ohashi CT,Odermatt B, Malissen B, Zinkernagel RM, and Hengarten H,Cell 1991; 65:305-317.

39. Oldstone MB, Nerenberg M, Southern P, Price J, andLewicki H, Cell 1991; 65:319-331.

40. VanParijs L, Peterson DA, and Abbas AK, Immunity 1998;8:265-274.

41. Perez VL, VanParijs L, Biuckians A, Zheng XX, Strom TB, andAbbas AK, Immunity 1997; 6:411-417.

42. MacPherson GG, Huang FP, Platt N, Liu LM, Jenkins CD,Major J, Powell T, and Wykes M, J. Leukocyte Biology 1998;Supplement 2:51 (meeting abstract).

43. Fazekas de St Groth B, Immunology Today 1998; 19:448-454.

44. Matzinger P, Annu. Rev. Immunol. 1994; 12:991-1045.

45. Cumberbatch M, and Kimber I, Immunology 1992;75:257-263.

46. MacPherson GG, Jenkins CD, Stein MJ, and Edwards C,J. Immunol. 1995; 154:1317-1322.

47. Roake JA, Rao AS, Morris PJ, Larsen CP, Hankins DF, andAustyn JM, J. Exp. Med. 1995; 181:2237-2247.

48. Kudo S, Matsuno K, Ezaki T, and Ogawa M, J. Exp. Med.1997; 185:777-784.

49. De Smedt T, Pajak B, Muraille E, Lespagnard L, Heinen E,De Baetselier P, Urbain J, Leo O, and Moser M,J. Exp. Med. 1996; 184:1413-1424.

50. Reis e Sousa CR, Hieny S, Scharton-Kersten T, Jankovic D,Charest H, Germain RN, and Sher A, J. Exp. Med. 1997;186:1819-1829.

51. Howard CJ, Sopp P, Brownlie J, Kwong LS, Parsons KR, andTaylor G, J. Immunol. 1997; 159:5372-5382.

52. Morris PB, Nouvelle Rev. Francaise d’hemat. 1968; 8:145.

53. Smith JB, McIntosh GH, and Morris B, J. Anat. 1970; 107:87.

54. McIntosh GH, and Morris B, J. Physiol. 1971; 214:365.

55. Wannemuehler MJ, Kiyono H, Babb JL, Michalek SM, andMcGhee JR, J. Immunol. 1982; 129:959-965.

56. Stremmel C, Greenfield EA, Howard E, Freeman GJ, andKuchroo VK, J. Exp. Med. 1999; 189:919-930.

57. Carosella ED, Rouas-Freiss N, Paul P, and Dausset J,Immunology Today 1999; 20:60-62.

Mucosal ImmunologyAffinity GroupWorkshop–Harrogate 1998

Interactions between innate and acquiredimmunity at mucosal surfacesAfter an exhausting week at Harrogate it was a

great surprise to see such a good turnout for theMucosal Immunology Affinity Group workshop. TheBSI somewhat over-estimated the interest in thetopic however, putting us in the main hall.Neverthless there were about 250 participants atthe morning session. Going along with the maintheme of the meeting, it was decided to try to alsocover innate and acquired immunity in the gut. Theoriginal programme had to be changed somewhatbecause of several late withdrawals by US speakers(whose names are now engraved in the never-to-be-invited-again files), but the late replacements weremore than adequate.

The meeting began with an overview of trefoilpeptides by Ray Playford from Leicester. These are afamily of 3 very interesting small peptides made byepithelial cells whose main feature is their ability topromote epithelial restitution after gut injury. Theywill shortly be going into patients to determine if

16 June 1999

they promote ulcer healing. This was then followedby a talk by Ian Sanderson from Bart’s on chemokineproduction by gut epithelial cells and its modula-tion by nutrients and gut bacteria. Ian described thetransgenic mouse that Yoshi Otsuka has made in hislab where MIP2 is overexpressed in the gut epithe-lium using an epithelial specific promoter. The finaltalk before coffee was an excellent review by CharlesParkos from Emory who covered all of the work heand Jim Madara have done on identifying the epi-thelial surface ligands which neutrophils recogniseas they bind to and cross the gut epithelium. In ad-dition he covered the functional work where it hasbeen shown that adenosine secreted by neutrophilsin the gut lumen is the stimulus for secretion of ionsby gut epithelia. These studies have direct relevanceto the diarrhea seen in diseases such as ulcerativecolitis, amebiasis and shigellosis where neutrophilrecruitment into the gut is a prominent feature. Themorning session ended with a technicolor feast ofpictures by Per Brandtzaeg from Oslo who coveredthe expression of homing molecules by gut cells, aswell as some of his work on the function of gut en-dothelial cells. As the worlds’ best at multi-color im-munofluorescence on tissue sections, Brandtzaeg’swork now resembles a mixture of Jackson Pollockmeets Roy Liechtenstein, the latter by the fact thatthe colors are now so complex that he has to pro-vide multi-color cartoons to explain to the audiencewhat they are seeing. Seeing these pictures though,few could dispute that a4b7 is the gut homingmolecule in man.

At lunch the affinity group held their AGM and itwas decided that Kingston Mills would organise nextyears symposium, with more focus on the lung. Thisis particularly appropriate since next years meetingis with the allergists.

The afternoon session was more sparsely at-tended, as people departed Harrogate. However asthe audience got smaller, the talks got better andthe session was one of the best I have ever attended.First of all, Hugh Miller showed a lot of new data onthe control of protease production by mucosal mastcells. He showed clearly that TGFb is the key cytok-ine responsible for protease upregulation in thesecells, re-inforcing the message that it is a key cytok-ine at mucosal surfaces. Gordon Dougan then gavea memorable talk split into 2 sections. In the firsthalf he covered current work by Gad Frankel and

himself on intimin of enteropathogenic E. coli(EPEC). When these bacteria infect the gut they in-ject a receptor into gut epithelial cells to whichintimin on EPEC then binds. The structure of intiminhas just been solved by NMR and although a bacte-rial protein, it has a terminal C type lectin domainand 2 Ig-like loops. In the second part of his talk hecovered the use of non-toxic mutants of cholera tox-ins as oral adjuvants —the great unfulfilled promiseof mucosal vaccination. The final presentation be-fore coffee was a magnificent talk by K-E Magnussonfrom Sweden who covered his studies on the wayin which the protease from Vibrio cholera could de-grade epithelial tight junction proteins.

In the final session the high standard was main-tained and even exceeded. Delphine Guy-Grandfrom Paris gave an amazingly good talk on mouseintraepithelial lymphocytes, covering every aspectof their origin and function and making those in theaudience who also work on IEL wonder why theybother. It was really a tour de force and it was some-what regrettable that Delphine had booked a cheapflight home necessitating a Saturday night stay inHarrogate, something that none of us would wishto suffer. The short straw for the last talk fell to JoSpencer from St. Thomas’ Hospital who wiselyrealised that V region sequences and somatic hyper-mutation in gut B cells were beyond the capacity ofa tired audience. However in her sophisticated mo-lecular analysis she showed that it was very unlikelythat there was a B1 lineage of cells in humans, un-like mice, because all human IgVh genes in the gut,regardless of isotype, are highly mutated. Thus thereare no natural antibodies encoded by germ line genesin human gut. The session wound up, on time, with50 dedicated souls remaining to the end.

Overall the session was a success although weprobably failed in our aim of interdigitating innateand acquired immunity in the gut. Nonetheless mostof the talks were very good and some were quitesuperb (even the non-immunologic ones likeDougan and Magnusson). There clearly is a great dealof interest in mucosal immunology and the standardof talks deserved a bigger audience. Coming at theend of a tiring week is not the best time to havesuch meetings and so in the future we will try torun them earlier.

Tom MacDonaldSt. Bartholomews Hospital

Mucosal Immunology Update 17

The First European Mucosal Immunology Meeting“People will come!”. And to the surprise and de-

light of the Mucosal Immunology Affinity Group,who organized this meeting, they certainly did. TheGroup first identified the need to co-ordinate thefield within Europe some three years ago, and, on2nd and 3rd of October, 270 participants from 17European countries finally met at St. Bartholomew’sHospital in London. The level of enthusiasm for thetopic can be judged by the fact that there were over400 applications to attend, but this had to be lim-ited to 270 because of the size of the venue. Atten-dance was on a first come-first served basis. An im-pressive list of international speakers was assembled,who were happy to pay their own expenses to themeeting. In addition, there were 45 poster presenta-tions and 16 short oral presentations of submittedabstracts, all of a very high standard. Based on thetheme “The cells and molecules important inmucosal tolerance and inflammation”, theFirst European Mucosal Immunology Meeting ad-dressed the vital questions: What maintains mu-cosal immune homeostasis and what causes itsbreakdown in inf lammation?

At the first level of interaction of antigen withthe mucosal immune system, new light is being shedon the control by B cell products of the differentia-tion of antigen sampling epithelial M cells overlyingPeyer’s patches from absorptive epithelial cells (N.Debard, Switzerland). Exciting new work is show-ing the active participation of the epithelium in di-recting the mucosal immune response and respond-ing to antigen by altered gene expression (I.Sanderson, UK; R. Hershberg, USA). The inductionof an appropriate immune response to antigen fol-lowing its interaction with the epithelium is to alarge extent dependent on the nature of the mucosalantigen presenting cell (M. Bailey, UK) and modula-tion of responses can be achieved by targeting alter-native processing and presentation routes usingcholera toxin (N. Lycke, Sweden).

Downstream of the first interactive events withantigen and dependent on the route of entry, andthe concentration and the structure of the antigenwhich passes the epithelial barrier, the response isregulated by T cell anergy or suppression (A. Mowat,UK). Specific interactions between developmentalsignalling receptors, such as Notch, and their ligandson naive and regulatory T cells (G. Hoyne, UK) mayprovide the molecular key to understanding theregulation of mucosal responses. IgE mucosal re-

sponses are differently regulated from IgG responses(G. Kraal, Netherlands). It is perhaps not surprising,then, that the window of antigen concentrationwhich permits successful induction of tolerance,both in experimental models and in recent humantrials, is very restricted (G. Panayi, N. Staines, UK).The capacity of cells to lodge in the gut mucosa iscontrolled by unique gut-homing adhesion moleculeprofiles expressed on Peyer’s patch lymphocytes (I.N.Farstad, Norway; A. Hamann, Germany) and thesubsequent homing of activated cells to effector sitesin turn depends on the site of induction (M. Quiding-Jarbrink, Sweden).

The mucosal immune response to luminal anti-gens is tightly regulated and when things start togo wrong, as in chronic inflammatory bowel dis-ease, TNF-a is at the centre of things. From its rolein regulating apoptosis in the mucosa (M. Boirivant,Italy), to its multiple actions in full-blown Crohn’sdisease, its importance is revealed by the successof anti-TNF-a antibody treatment of Crohn’s disease(T. ten Hove, Netherlands). Recent work definingregulation of TNF-a-induced signalling and activa-tion mechanisms (M. Neurath, Germany; F. Pallone,Italy) and the emerging understanding of regula-tion of TNF-a transcription through repressor ele-ments (S. Schreiber, Germany) will point the wayto more specific therapies of mucosal inflamma-tion. Equally, analysis of the complexities of the in-flammatory process using model systems such asthe CD4+ T cell-transplanted SCID mouse (F. Powrie,UK; J. Reimann, Germany), or the CD3e transgenicmodel (S. Simpson, Ireland) shows great potentialfor defining anti-inflammatory therapy. The vital in-teraction between genetic and environmental fac-

First EMI, continued on page 18

18 June 1999

tors, both in maintaining immune homeostasis inthe mucosa and in the pathogenesis of mucosal in-f lammation was elegantly stressed by BalfourSartor (USA) in the closing keynote lecture.

Plans are already afoot to build on the success ofthis meeting by establishing regular European Mu-cosal Immunology meetings, with the next tenta-tively planned for 2000 at a venue outside the UK.The success of the meeting was due to a number offactors. It was very focused and there was no regis-tration fee. The willingness of senior figures to at-tend at their own expense to attract participantswas undoubtedly important. Flights into London aregood and the meeting was held at a central loca-

Charles Elson, M.D. matory bowel disease that hasbeen a continuing theme forthe past 20 years. In recentyears this has taken the form ofdevelopment of experimentalmouse models such as the C3H/HeJBir mouse, which developscolitis spontaneously under cer-tain conditions. Recent pub-lished studies demonstrate thatCD4 T cells reactive to com-mensal enteric bacterial anti-gens mediate this colitis.

Dr. Elson maintains an activeclinical practice providing con-sultative care to patients withgastrointestinal disorders, par-

ticularly inf lammatory bowel diseases. He hasserved on numerous foundation and NIH study sec-tions and advisory panels and currently is Chair-man of the Grants Council of the Crohn’s and Coli-tis Foundation of America. He is a member of a

number of national organizations including theAmerican Association of Immunologists, AmericanSociety of Microbiology, American College of Phy-sicians, and has recently been elected to the Asso-ciation of American Physicians. He is a co-founderof the Society for Mucosal Immunology and servedas its first Secretary-Treasurer. With this extensivebackground in mucosal immunology, we all lookforward to his leadership as President of the Soci-ety for Mucosal Immunology.

First EMI,continued from page 17

tion, easily reached from all London airports. Themeeting started on Friday afternoon and ended lateSaturday afternoon so that participants could getcheap flights, flying in Friday morning and leavingon Sunday. Reasonably priced hotels were identifiedfor the participants. The social event on Fridayevening was a great success. Finally, there was tre-mendous support from the BSI, Society for MucosalImmunology, Astra Hassle, Cantab Pharmaceuticals,Pfizer Inc, and Centocor. We hope that this is thefirst of many.

Paul W. Bland, Allan McI. Mowat,Thomas T. MacDonaldMucosal Immunology Affinity Group, BSI

is Professor of Medicine and Mi-crobiology and Director of theDivision of Gastroenterology andHepatology at the University ofAlabama at Birmingham and heholds the Basil I. HirschowitzChair in Gastroenterology. Dr.Elson received his undergraduatedegree from the University ofNotre Dame and his M.D. fromWashington University School ofMedicine in St. Louis. He did hisclinical training in internal medi-cine and gastroenterology atCornell University and the Uni-versity of Chicago. Followingthat he worked for four years doing basic immu-nology research at the National Institutes of Healthin Bethesda Maryland mainly in the ImmunologyBranch working with Dr. Warren Strober. While atNIH he provided some of the first direct experi-mental evidence demonstrating IgA-specific T cellregulation. Subsequent work in mucosal immunol-ogy has centered about regulation of mucosal im-mune responses, much of it using cholera toxin asprobe. These studies led to the first observation thatcholera toxin can act as a mucosal adjuvant; themechanism of such adjuvanticity has been the fo-cus of a number of subsequent studies.

During his Fellowship in Gastroenterology he de-veloped an interest and research focus into inflam-

New President for SMI

Mucosal Immunology Update 19

Society for Mucosal Immunology — Application for MembershipMembership in the Society for Mucosal Immunology is open to all immunologists, physicians, dentists, veterinarians,biochemists, or other scientists who do research in or who have an active interest in mucosal immunology, and who havepublished at least one first–authored paper in a peer–reviewed journal. Society membership includes a subscription toMucosal Immunology Update. (Please Type or Print)

Name: ________________________________________________ Degree: _____________________________ Title/Position:

Institution: ____________________________________________ Address: _______________ City/State/Country/Zip Code:

Primary Specialty: ______________________________________ First-Authored Publication: _________________________

_____________________________________________________________________________________________________

Please mail completed application form with $10 application fee and $60 annual dues (total $70) payable by check in U.S.dollars or by VISA or MasterCard account number (please include expiration date and MasterCard 4-digit interbanknumber) to Dr. Peter Ernst, Secretary-Treasurer, The Society for Mucosal Immunology, 4340 East West Highway, Suite 401,Bethesda, MD 20814-4411, U.S.A. Or e-mail to smi@paimgmt.com

Mucosal Immunology Update 19

AnnouncementsSociety Business

• Postdoctoral Research Assistant and Research Tech-nician—Department of Immunology, University ofGlasgow, Scotland. The project will use a combina-tion of in vivo and in vitro techniques to investi-gate the cellular and molecular basis of the immu-nogenicity of these potential vaccine vectors. Thelaboratory has a range of interests in mucosal im-munology and immunoregulation. Applicationsshould be sent to Dr. Allan Mowat, Department ofImmunology, Western Infirmary, Glasgow G11 6NTby 14th May, to whom preliminary enquiries shouldalso be addressed (Tel: 0141 211 2728; FAX: 0141337 3217; E-mail: a.m.mowat@clinmed.gla.ac.uk)

Job Postings

• Recently, the Governing Council approved the suggestion by Drs. Kagnoff and Kiyono to createan educational fund in support of mucosal immunology in developing countries. The goals ofthis fund are to facilitate the development of educational efforts in these countries. Currently,the priorities are to increase the cash available to support of these efforts, hence, a request forcontributions from members was included in your renewal notices. Interested individuals areasked to seek contributions from industry, government or private founda tions to augment thefund as well as provide suggestions for their use. Consider action of requests for projects willbegin in 1999. Please contact the Secretary- Treasurer, Peter Ernst, at the Society Business Officewith your suggestions and requests.

• Recently, members should have received their new SMI Membership Directory. With the in-crease in new members that the 10th International Congress for Mucosal Immunology hasattracted, we plan to publish an updated edition early in 2000. Please carefully read yourdescription in the directory. If any information is incorrect or missing, forward the correctionson to Mr. Peno by email at smi@paimgmt.com.