Antigen presentation, antigen-presenting cells and antigen processing

Antigen presentation, antigen-presenting cells and antigen processing

Antigen presentation, antigen-presenting cells and antigen processing D.R. Katz Histopathology Department, Bland-Sutton Institute, University College ...

812KB Sizes 2 Downloads 272 Views

Antigen presentation, antigen-presenting cells and antigen processing D.R. Katz Histopathology Department, Bland-Sutton Institute, University College and Middlesex School of Medicine, London, UK. Current Opinion

in Immunology

Introduction The literature of the period June 1987- May 1988 is particularly rewarding to review from the point of view of those interested in antigen presentation.The T cell itself has long recognized the importance of antigen-presenting cells and of processing,but it is only recently that the signihcanceof these areas has become apparent to research immunologists as well, making the field a central topic of interest rather than a fringe discipline examined only by eccentrics. Because of this increased interest, the literature in the field has grown considerably, and thus in this review it has been necessaryto focus on one type of presentation only, i.e. the one which results in T cell activation.B cell activation, the events in the germinal centre, and the complex question of presentation within the bone-marrow and myeloid series,have all had to be consideredin less detail.

Antigen-MHC

association

Recognitionof how antigen-presentationmight be investigated has been given a considerableimpetus as part of the excitement aroused by a paper which appeared in Nufure in the latter half of 1987.Described in an accompanying review as making ‘every immunologist’s pulse race’[l] , this paper ourlined the structural characterization of the major histocompatabilitycomplex (MHC) human leukocyte antigen (HIA)-A2 [2], and demonstrated that there is a groove in this structure which constitutes a binding site for an antigen. The study gives a tantalizing glimpse into the configuration of one immunogenic antigen-MHCcomplex, which is the key unit of presentation, and as a result many of the studies which have followed during the past year have tried to conform with this prototype: for example, by asking how any given immunologic peptide antigenmight (or might not) be made to fit into such a groove for subsequentpresentation.Fur-

1988, 1:213-219

thermore, this study has also opened new possibilities for future detailed investigationof the mechanismof association of antigenwith MHC (both class I and II); and in the more distant future, it should be possible to define highdinity ligands for the groove that could block associationwith the MHC molecule of an immunogenic antigen which might otherwise bind and be presented. Although the questions raised by the Bjorkman paper have considerableimplications for future researchinthe antigen-presentingfield, in a sensethe paper falls outside the speciiic remit of the present review, since the paper itself explored neither the role of the HIAA2 histocompatability locus itself, nor the bound peptide, in the context of antigen presentation.For an examination of this type of question, one has to look at other studies.

Presentation of self-HIA One of the most interesting of these other studies [31 also used I-HA-A2in the assaysystem.The basis for this study was a panel of allospecificT cell clones which were chosen becausethey recognizerespectivelyIRA-A2 only (included in a peptide composedof residues98-113from the native molecule>, a shared epitope (represented in residues56-69) between FILL42 and HIAB17, and HIAB17 only. One set of experiments with these rlones used the 98-l 13 I&h42 peptide pulsed on to an I-I&A2 cytotoxic T cell and as a result blocked the T cell clone from killing the I-&i& targets.An alternative98-113 peptide homologous to the same residues but from HIA-Aw68 rather than ISA-AZ:has no effect in the system.A common epitope peptide from A2/B17 (A2.56-69) pulsed on to the appropriate T cells was also inhibitory of effector function for that particular clone, suggestingthat such T cell blocking might represent a general rule. A second set of experiments with the same clones examined their interaction with a panel of B ce!ls of dif-

Abbreviations HlA-human Il-interleukin;

leucocyte antigen; M-K-major CM-CSF-granulocyte-macrophage

histocompatibility complex; colony stimulating factor.

@ Current Science Ltd ISSN 0952-7915

213

214

Immune response

ferent I-IIA types in the presence of the same peptides (A2.56-69) and A2.9S3-113.Most of the B cells (even those from closely-matched I-IL4 types) could not be sensitized to lysis by the peptides. However, it was possible to identify one such sensitization system: the clone which recognizes the shared epitope A2/B17 lysed HIAAw69 targets (but not others) in the presence of peptide A2.56-69. In contrast with the previous part of the study, pretreatment experiments led to lysis only if the targets (and not the T cells) were exposed to the peptide before the assay.

Two presenting allocytotoxicity

mechanisms

for

Taking these results together, peptide A2.56-69 has been implicated in lysis in two separate ways. Firstly, since inhibition occurs at the T cell level, and is common to both A2 and B17 targets, this inhibition may well be caused by competition between the exogenous peptide and the endogenous peptide for the antigen specihc T cell receptor. Put another way, processing pathways may generate plentiful peptides in L&JO,but assuming that these peptides do play a role in a secondary response, then irrespective of how much peptide is produced, the relative specificity and relative afllnity of the peptide for the T cell receptor is an important variable which helps to determine whether or not there is a speciIic response. Secondly, sensitization of the target via binding of the 56-69 peptide to HIAAw69, can probably be considered analogous to binding of antigen to MHC, as described for typical antigen presentation of viral antigens such as inIluenza Hypothetically the 56-69 peptide could easily adopt an a helical structure in the groove in the I-IIAAw69 binding site, and thus mimic the configuration of native HIAX, which would in turn make I-IIA-Aw69 cells a recognizable target The structure of the 98-113 peptide is such that it would be less likely to do this. If ‘molecular mimicry’ of 56-69- rather than 98-113-l&e peptides could be generated naturally from exogenous viral or bacterial sequences as a result of antigen processing, and then used to conIirm these findings, this model will have served as a very useful example of antigen spe cilic as well as allo-responses. But even in the allo-response itself the model is of interest. It illustrates that an HIA-derived peptide can, indeed, bind to intact HLA and be presented. Furthermore, at least in this example, allospecific T cells appear to recognize a portion of the HIA molecule itself The number of epitopes which could be generated from a single HIA molecule is quite considerable, and there is adequate opportunity for binding to I-IL4 on an individual cell, so this mechanism of presentation could account for alloreactiviry in general.

Presentation

of exogenous

peptides

The Clayberger study [3] used class I MHC as both peptide antigen and restriction element on the presenting cell. Class 1 MHC is an example of an endogenous peptide in contrast with the exogenous peptides which have been used in most studies of antigen presentation, chiefly to dissect the nature of the class II MHCantigen association. Ovalbumin responses presentation

the sheet model for

In an interesting example of this type of study, the structural characteristics of an antigen necessary for interaction with class II and for subsequent recognition were examined in the well-established ovalbumin model system [4]. For this project a series of truncated synthetic ovalbumin peptides were synthesized and used to see which shorter sequences might inhibit binding of radiolabelled peptide 323339 (of known immunogenic properties) in a competitive fashion. The results were expressed as relative binding capacity to the restriction element, I-Ad. Once the binding pattern had been established, amino acid substitutions within the peptides were used to map more precisely which particular residues are involved in the association, and peptide sequences from other antigens known to interact with I-Ad in an immunogenic fashion were examined by ‘best-fit’ to try to identify a common theme. In summary, the findings in this part of the study were consistent with the classic determinant selection hypothesis: a few class II molecules are able to bind to large numbers of variable foreign antigens which result from processing, and this association can lead to an immune response. The second part of this same study looked at the other side of the antigen, the pattern recognized by the T cell. The same set of peptides was used, but the results were monitored in terms of hybridoma responses rather than binding to class II. This T cell recognition was far more susceptible to single amino acid substitutions. While some of these changes in responsiveness could be explained by failure to bind to the class II on the presenting cell, others could only be accounted for by effects at the T cell level, and these were not due to differences in conformation. The results of this study favour a planar contiguration for antigen between the class II molecule and the T cell receptor, such that some of the residues can be involved in association with both sides of the sandwich. Influenza responses presentation

the helical model for

However, in a second study which also focused upon class II-associated responses, two influenza spe&c T cell

Antigen presentation,

clones were used and the results led to a different strutturd model for recognition of antigen by T cells [5]. One of these influenza-specihc clones recognized a matrix protein determinant (sequence 17-31) and the other, a haemagglutinin determinant (sequence 306-329). The first part of this study was to map the essential stimulatory residues, using peptides with either common amino- or common carboxy-termini as a starting point. As a result the essential stimulatory residues for the matrix antigen were found to be at positions 18 and 19 at the amino-terminus, 29 (glutamic acid) at the carboxy-terminus, and 308 &sine> and 318 (alanine) for the haemagglutinin antigen, respectively. The hypothesis in the subsequent experiments rests upon alignment of these T cell epitopes, using two adjacent hydrophobic residues within the sequence as a starting point, in order to reveal other similarities in peptide structure. These similar hydrophobic regions should represent the DRl binding region within the epitope. If the antigenic epitope is then considered as a helix, then the opposite face from the hydrophobic regions should be the clot-tally specific T cell receptor binding region; and the two DRl binding hydrophobic regions themselves should be relatively interchangeable. Conversely, the successful achievement of this experiment, taken in conjunction with the linding that a histocompatibility antigen can indeed accommodate a peptide in a helical conformation without significant changes in the structure of the protein, as outlined above, would support the notion that DRl does bind an antigen in helix form as a critical event in presentation. In addition to these studies of the hydrophobic region and of the T cell epitope, Rothbard et aL [ 51 also examined the role of the other residues in the immunogenic peptide. They propose that the amino acids on the sides of the helix, while not forming part of the ‘essential epitope’, nonetheless have the potential to interact with residues on the histocompatibility antigens, and hence to influence T cell proliferation. This type of ‘neighbourhood’ effect is also illustrated in a slightly different way in other experiments in which an alanine (which does form hydrogen bonds readily) is substituted for proline (which does not form hydrogen bonds as readily, and hence mitigates against helical structure) within the synthetic peptide at a nonMIX binding site. As predicted, the alanine substituted peptide does induce T cell proliferation under conditions where the proline peptide does not.

antigen-presenting

cells and processing Katz

sponse itself (whether helper, cytotoxic or suppressor; and primary, secondary or hybridoma) may also play a role which has not yet been defined. Furthermore, in viva, where there are opportunities for multiple types of association of antigen with MI-X, and likewise many ways for an antigen to dissociate during processing, it is inherently likely that more than one mechanism may operate. Possibly an idealized antigen model should be a peptide that can be seen in both sheet and helical form? In a review which addressed some of these possibilities, Allen [6] suggested that one way to understand processing and presentation is to start with the premise that there are three different types of T cell determinants. Type I determinants (such as fibrinogen and listeria products) do not require processing and can nonetheless bind to both class II (via the ‘agretope’) and T cell receptor (via the T cell epitope). l&e II determinants are those which require unfolding to reveal the agretope and epitope regions. It follows that there will probably be other potential epitopes which are concealed for steric reasons within a Type II determinant. Lysozyme and myoglobin are two examples of this. However, these two antigens are also included, along with cytochrome C, ovalbumin and insulin, among the classic type III determinants, i.e. those which require proteolytic cleavage for recognition.

What is antigen processing? This model has considerable overall merit, but some recent studies have illustrated that it too may provide an incomplete picture in several respects. For example, one aspect of the question ‘what constitutes processing’ has been explored recently by examining the effects of membrane proteases on an ovalbumin antigen [7]. In this study the authors emphasize that processing in immunological terms does not start with a peptide, but with a whole antigen which is degraded to peptides and that furthermore one must recognize the functional differences between ‘physiological’ events such as membrane turnover, and the active selective processing pathway which leads to association with MI-K as part of the immune response. Along similar lines, Loenscher et al [8] have defined another minor variation in structure which nonetheless constitutes an important limiting circumstance for presentation. For this study they used their photo-afIinity labelled antigen system (which was itself the first clearcut model for antigen-MIX association in the prehistoric days of 1985).

Antigen processing The difference between these two detailed class II studies highlight the complexity of the system and raise several interesting possibilities. For example, the generation of relatively minor differences at single amino acid level during processing may be amplihed into considerable divergence in response patterns by the time that the T cell reaction is measured; and the nature of the T cell re-

Processing and the response to peptides However, perhaps the most direct study &cult to reconcile with a straightfotward models of sequential degradation and processing leading to presentation is some recent work using the well-known cytochrome c-driven T cell response 191.

215

In this study the initial experimentswere directed at identifylng the role of the residue at position 95 in the immunogenlc peptide 93-103. The peptide at position 99 is known to be part of the T cell epitope, and residue 103 as part of the agretope,but the effect of 95 was unknown. Whateverthe role of 95, the assumptionwas made that a minimal antigenic 93-103 peptide does not require processlng.Thus, it was surprlsii to find that if a tyrosine is substituted for isoleucine at position 95, then when the peptide is presentedby I-Eb presenting cells the 95-tyroshe variant proves ten times more potent than the parent peptlde; however, if the conventional I-Ek presenting cells are used then the 95-tyrosineis much less effective. If the presenting cells are pre-lixed then this difference is abolished, indicating that it is unlikely that the results are due to 95 being merely a lower aflinity component of the normal MIX binding agretope.Furthermore, chloroquine also abolishes the dil%erentialeffect; experiments usingF1 antigen-presentingcells indicate that both strains are capablein ptinciple of processingthe antigen equally efficienw, and the resultswere also coniirmed in oiuousing the same two mouse strains. One way to interpret the results with the cytochrome c peptides is to regird these experiments as representing an example of a residue substitution unmasking a functional polymorphic MI-K determinant This unexpected Ending needs to be conlirmed further, but it does raise important questions, particularly for the role of antigen and MHC in autoimmunity.

The significance of antigen-presenting

cells

Irrespective of the precise interpretation of these experiments, the study would have been clearly impossible using inert membraneswith known MHC molecules attatched, which has been suggestedas an optimum in u&u presentation model. The cellular components (e.g. intemalimtion processes,kinetics, and in situ localization of antigen) must all be considered as integral parts of presentationand processing studies [lo]. A recent study which begins to addressthis question has used gold labelled antigens to tty to demonstrateinternalization and recycling events dire& and this approach may provide much useful information about control of processing in the future, provided that the sensitivityof the method can be cof.llimled [ll]. Reevaluatingthe studies cited above in this context, only the paper discussing the influenza peptides placed particular emphasis on chamcterization of the presenting cell pathway in any detail. Three types of presenting cell were used: (1) Epstein-Barrvirus-transformedautologous cells; (2) histocompatible DRl positive peripheral blood mononuclear cells, and (3) L cells transfectedwith both a and fl DRl chains. Anti-DRl antibodies inhibited all three types of presentation. Thus In the influenza peptide experiments the nature of the presentingcell has been consideredas a variable,and ca&ully excluded. However, a different conclusion was

reached in another system recently [12], indicating that a presenting cell should not always be assumed to be playing a constant role in every type of response. In this other system the assayconsisted of induction of IL-2 synthesis in murine T cell hybridomas which are known to react to allogenic I-E determinants using the Vl3 17avariable element of the T cell receptor. As stimulators, B cell lines, macrophagecell lines, bone marrowderived macrophages,spleen cells, and I-Ek transfected fibroblast cell lines were compared in the same experiment. The results were quite unequivocal. The Vl3 17a and lines respond to some fresh macrophagesbut not agedmacrophages,and B cell I-E, but not I-E expressed on the other presenting cell types. T cells that do not express this particular Vg chain can respond to the other cells which express I-E determinants and the responses to I-E restricted antigenic peptides are not impaired. Antibodies againstknown accessoryadhesion molecules did not inhibit this B cell Induced response to any greater extent. h&u-rackand Kappler [ 121propose that these B cells express I-E complexed with an endogenouspeptide, which may itself be cell type-specilic.For example, the peptide might be a breakdown product of some protein which is synthesizedby B cells in particular. The resultant selfpeptide might then associatewith I-E within a groove [ 21, in an analogousway to the other self-peptideswhich have been shown to provide an immunogenic signal [3]. This combination would then constitute an antigen suitable for recognition by the selectedT cell population. The authors [ 121 prefer this model to others, such as that Isuggesting that the B cell I-E has either altered conformation or altered glycosylation or the suggestionthat there is some as yet uncharacterizedB cell-specific presentation associationsurface molecule.

Heterogeneity of antigen-presenting

cells

Whichever of these hypotheses proves correct, the observation that one cell can present better than another cell does focus further attention on the issue of presenting cell phenotype. During the period under review, the notion (now generallyaccepted) that most cells have the potential capacity to act as presenting cells has not been challenged. In addition the role of the macrophage in processing has remained pre-eminent, if not unique. The role of the B cell continues to be emphasized in some studies [13], although in this particular latest analysis ‘cell-free’class II molecules were also effective. Neither quantitative Ia nor discrepanciesin Ia glycosylation could account for the differences between activated B cells which present in alloresponses,and resting B cells, which do not. In the germinal centre, B cell presentation has also been described [ 141,and a mechanismhas been proposed for the transfer of antigen from follicular dendritic cells to other presenting cells which are more eflicient at antigen processing [15].

Antigen presentation,antigen-presentingcells and proce!ssingKatz Dendritic

cells

What has come to the fore, instead of the previous approach which looked chiefly at exclusive antigen-presenting cells, is the studies which have begun to address the question of selectivemechanismapplied to the dendritic cell, which has been described as the main ‘professional (and most potent) presentingcell type identifiable in most lymphomedullary tissues. A new approach in this regard has been to examine the role of soluble mediators (including lymphokines) on dendritic cell function. For example, our own study of the human dendritic cell [ 161 examined the relationship between these cells and a known monocyte product, the active metabolite of vitamin 4. We suggested that interaction between monocytes and dendritic cells may be better envisagedas a linked relationship with respect to mediators rather than by regardingthe dendritic cell as an independent all-embracingpresentingcell type. This type of mediator pathway is probably a more realistic form of synergy than the previous suggestion that macrophagesprocessedand dendritic cells presentedthe resultant peptides. It is interesting that this previous suggestion would also have been inconsistent with both the cytochrome C [9] and the B cell - Vj317aT cell [12] studies, noted previously.

Dendritic

cells and 11-l

The obvious macrophage-derivedmediator which might act in conjunction with a dendritic cell is IL-l. From earlier work it appearsthat dendntic cells do not synthesize this lymphokine, thus delineatinganother potentially sigr&ant difference between the two cell types. However, the effects of adding exogenousIL-1 into dendritic cell-T cell cultures had not been investigatedthoroughly. The role of added exogenousIL-1was examinedin detail in thymocyte responsesby Inaba et al [17]. These studies demonstrated that dendritic cell function is indeed amplified by IL-l, and that dendritic cells pre-pulsedwith IL-1 are even more effective.Interestingly,the pulsed dendritic cells did not appear to express IL-l activity themselves,suggestingthat the II-1 acts directly on the dendritic cells rather than on the T cells. How this effect is mediated is not as yet clear. In addition, the same group were also able to demonstrate the induction of typical dendritic cells from classII negativethymic precursors by adding IL-1 into the culhues for a 48 h period. However, this is not the only lymphokine which has been implicated directly in the generation of dendritic cells. In parallel studies [18l Iangerhans cells from the skin could serve as dendritic cell precursors if they are cultured in the presence of granulocyte-~c~p~ colony stimulating factor (GMCSF).Furthermore IL-1will synergizewith GM-CSFin this regardwhich raisesinteresting and unresolvedquestions as to the relative lymphokine/cytokine receptor expression on these different presenting cells.

lineage relationships of antigen-presenting cells These studies raisecomplex issuesof lineageamong presenting cells which have not yet been fully resolved. In this field, perhapsthe most signi&ant prospect for future studieswas a report describing the featuresof transgenic mice expressing GM-CSF[ 191.Antigen presentation&Y sehas not been investigatedin this study but the different types of antigen-presentingcells which might be isolated from congenic animals expressing selected lymphokine genesprovides an imaginativeway to investigatepresenting cell interrelationships and lineagesin the future. What this study did show was that the effect of ‘biasing’ the myeloid lineage in this fashion is unpredictable. For example,the mice develop muscle damageand blindness and have many multi-nucleatecells, but do not manifest the changesin T cells which might be anticipated as secondary to the myeloid lineage changes. An alternativemethod to examinelineageistouseretroviral transfer to insert a relevant gene into a lethally irradiated animal and then monitor the development of the antigen-presentingcells. Again a suitable model has been identiiied [201 and again there are some unexpected results. In these mice v-fms has been transferred. The c-fins would normally encode for the CSF-1receptor present on proliferating myeloid cells, but instead the study demonstratesmultilineagedistribution and no proliferative advantagefor myeloid cells.

Summary To summadze,during the period under review there have been considerableadvancesin our understandingof how antigen is associatedwith MHC on the surface of a presenting cell. Basic rules which govern this association have been confirmed as including both the nature and the conliguration of the antigen. Although the different processing pathways which generate the antigen-MHC complex have remainedill-defined, it is now much mote clear what the optimum end point of such processQ must be. As a result of these studies, several new questions can now be addressedin the field of antigen presentation DIM@ the next few years it should become apparent why there should be such an unusual differential pre senting cell requirement for T cells which are expressing a selectedvariable region of the T cell receptor,and whether other T cells expressingdifferent receptors have different optimum presentation requirements.It should become clear also whether di&ent antigenshavesimilar residuesat the sides of the a helix which are important in MHC binding, and whether the sheet or he@ models are more important in viva A rational explanation for dominanceof allomactivity has been proposed, and kinetic studies of self-MHC processing are essentialto dehe this further. Furthermore

217

2l8

Immune reswnse the possibility of an antigen unmasking an MHC polymorphism (which in turn could influence presentation> will have to be considered seriously in the pathogenesis of auto-immunity.How lymphokines affect patterns of antigen-MHCassociation and hence presentation needs tobeinvestigatedbothininvitrosystemandinthenew animal models. Finally, studieswhich examine the mechanism im&ed in dendritic cell function should be able to establish why these cells are particularly &cient as ‘rmhml adjuvants’for presentation to T cells.

Fox BS, CARBONE FFt,GERMAINEN, PATERSON Y, Scttwunz EH: (Letter to the Editor) Processing of a minimal antigenic pcptide alters its interaction with MHC molecules. Nahrre 1988, 331:538-540. Substitution of a single residue in qtochrome C peptide alters the pattern of responsiveness.Thus even minimal peptide antigen can be modulated in uifro and in uiua

9.

e

l

DELOV~~CH TI, SEMP~EJW, Puttups ML Iniiuence of antigen processing on immune responsiveness. Immunol Today 1988, 9216-217. Antigen processing is essential for presentation in most systemswhich havebeen examined properly. This article summarizes the need for further cell biologicai and biochemical sNdies of processing in order to explain presentation and outlines some ~UNIE! objectives for these 10.

l

undies.

Annotated references and recommended reading l l e

Of interest Of outstanding interest

1.

TOWNSEND A, M.uzMtcsi~R~A Those images that get Besh images met. Naturn 1987,329&l-483. Discussion of the implications which result from the chamcterixation of the HIAX molecukz The authors review succincdy the different pos siblepathwayswhichmayleadtodassIandclassBassodationwith =aF= 2. BJORRMAN PJ, SAPRR m Qt.at~ot~ B, B~~mrr WS, S~OMINGER l a JI+ Wn.m DC: Stmcture of the human ciaas I histocompatibility antigen, HlAA2. Nature 1987, 329506-511. Stnumral chamcterisanon of HlA&! reseais a groove which provides antigen-binding site. The lint critical demonstration of the threedknensional cxmliguration of an example of the key molecule in presentation l

J, LuowrC DS, SCHOO~MK GK, KRENSKYAM: (Letter to the Editor) HLA&! peptides can regulate qtolysis by human allogeneic T lymphocytes. Nature 1987, 330~763765. A peptide from an H&A2 molecule in associationwith another Class I molecule is seen on a target cell. AUo-T celi clones are cytotoxic and this can be blocked by speciiic peptides at the TceU level. The paper is a demonstration of a classI peptide seen on a class E MHC molecule =maneigen 4. SRI-IX 4 BUUS S, CtXON S, S~UIH JA, Mnes C, GRRY HM: l Structural chvzctalstlcs of an antigen required for its interaction with Ia and recognition by T cezils.Nature 1987, 3.

CIAYBRRGER C, PARHAMP, Rm

l

328:395399.

Ovalubumin peptides used to dissect nature of ciass II T-ceil binding. ItUMMD ]B, &HlER RI, HOWIANDK, BAL V, Ecwts DD, SEKALYR, LONGEO, TAYLORWE, L+bt~JR Structural model of HLA-DRl restricted T celi antigen recognition. fXll988, 52:515-523. Inawnzapepeidescanbeusedto~eanahelixwithanimmunogenie epitope and an interchangeablebinding region which associates with DEl. This is critical paper in anaiy& of what constitutes immunogenicity at the amino acid level. 5.

l e

&EN PM: Ant&m processiag at the molecular level. Immud Today 1987, 8:278-273. Descdbes three possible coniigumdons for antigen which might be recognhedbyTceUs.Summa&es the molecular events which may generate an antigen-MHCcomplex

6. l

Chum BM, SHAWMA: Membranea proteases invoked in antigen processing. J cell Bfkbem 1988,128:283. EelI ftee system aUows study of molecules invoived in processing.

7.

8.

~ENSCHERIF, AUEN PM, UNANLIRER: Binding of photoreactive lysoayme peptides to murine histocompatibility class II molecules. Pnx Natl Acad Sci LISA1988, 85871-874. Altering the photomactive moiety changes the ai%nityof a peptide for a class II MHC molecule. Strucmrai variation in the non-MHC binding tegion may be important in generating an immunogenic complex. l e

HN J, BERZOFXYJA, DELOMTCHTL Ultrastructural study of intemalixation and recycling of antigen by antigen presenting cells. J Mel Cell Immurwl1988, 3321-343. Ulustrateskinetics of biding, pathway of internalization and recycling of gold labelled insulin and myoglobin. Provides a new view of how processing and association with MHC might operate at the single cell level. 11.

l

12.

M&RRACK P, KAPPLRR J: (better to the Editor) T cells can distinguish between allogeneic major histocompatibiity complex products on different cell types. Nature 1987, 332840843. AUogeneicT celi Linesexpressing Vg17a, respond to B ceils but not other presenting cells. The authors demonstrate differential role of pm senting ceils and propose that a ceUproduct binding to self may be the key feature. l e

KIUXERJ, JEM~ DM, CHESNUTEW, GRRYHM: Studies on the capacity of intact cells and puriiied Ia from ditfercnt B cell sources to function in antigen presentation to T cells. J Immund 1988, 140:388394. Different B cell population and purified Ia molecules were used as stimulators for ailogenic T ceils. The resuks confirm that activated B cells present and resting B cells do not, but the reasonsfor this remain spec ularive. 13.

l

Kosco MH, SZAKU AK, T!zw JG: In vivo obtained antigen presented by germinal center B cells to T cells in vitro. J Immunol 1988, 140:354360. B cells were exposed to antigen in viva in the form of complexes on follicular dendritic cells. The activated primed B cells were potent presenting cells in u&u 14.

l

E%?AKAL AK, Kosco MH, Tr?iv JG: A novel in vivo follicular dendritic ceil-dependent iscosome-mediated mechanism for delivery of antigen to antigen-processing cells. J Immund 1988, 140341353. Immune complex bodies were detected in follicuiar dendritic cells in u&a Morphological evidence suggeststhat this is transferred to B cells, and the mechanism is important for B ceU memory. 15.

l

BRENNANA, KATZ DR, NUNN JD, B~uue~ S, HRuI~~N M, F~WIERLJ, O’EIORDANJLH: Dendritic cells from human tissues express receptors for the immunorcguiatory vitaminD3 metabolite, dihydroxycholecalcifeferol. Immunology 1987, 61:457-&l. Isolation of human tissue dendritic cells and a receptor for a momxyte product in these cells. Demonstratesa possible monocyte-dendritic ceU coUaborativepathway which can regulate immune function. 16.

l

17.

INABAK, W~TMJZR-PACK MD, INABAM, MURAMATSU S, S~-E~MAN RM: The function of Ia+ dendritic celis and Ia- dendritic cell precursors in thymocyte mitogenesis to lectin and lectin plus IL-l. J Exp Med 1988, 167~149-162. Describes IL-1 synergy with dendritic ceils, cbrifyiq the role of IL-1 in relationship to dendritic cells; and demonstratesa novel pathway of differentiation into antigen-presentingcelis. l

18. l

HEUPLERC, KOCH F, SCXIJUZR G: Granulocyte/macrophage colony-stimulating factor and interleukin 1 mediate the maturation of murine epidermai Langerhans celis into potent irnmunostimulatory dendritic cells. J Eap Med 1988, 167:700-705.

Antigen presentation, ~angerhanscells are not effective as antigen-presentingcells, but the two cytokines can induce presenting activity. This study demonstrates a link between cytokines and presentingceU function. E, KElso A, KAh’NOUllAKts 19. cUlHLSERTSON RA, LYONS1, STANLEY G, W-N DJ, KUNIWORTHGK, GONDATJ, DUNN AR: 0 Transgenic mice expressing a haemopoeitic growth fatfor gene (GM-C%) develop accumulation of macrophages, blindness and a fatal syndrome of tissue damage. cell 1987, 5k675-6136. Transgenicmice expressingGM-CSFdevelop lesions probably associatedwith excessektor macrophagefunctions. This ttansgenicmouse Is an exampleas to what kind of model could be useful in future SNdies

antigen-presenting

cells and processing Katz

of presentation. 20. HEARDJM, ROUSSEL MF, ~NME~ER CW, SHERRCJ: Multilineage hematopoietic disorders induced by transplantation @ of hone marrow cells expressing the v-fax3 oncogene. cell 1987, 5 1663-673. Rewwiral uansfer of v-fms into lethally inadiated recipients resuhs in mukUineageproliferation, but v-fms expressingcells do not have a sekcdve proliferative advantage.A possible alternativeto the ttansgenic model in the sNdy of antigen-presentingcd lineage.

219