Editorial overview: CD27 and (TNFR) relatives in the immune system: their role in health and disease

seminars in I M M U N OL OG Y, Vol 10, 1998: pp 417]422

Article No. si980159

Editorial overview: CD27 and (TNFR) relatives in the immune system: their role in health and disease Wil A.M. Loenen

Relatives of the tumor necrosis factor receptor family are proving of crucial importance for an effective immune response as well as maintenance of homeostasis. The reviews in this issue summarise accumulating evidence for an emerging intricate interplay between these receptors themselves and with the CD28 pathway. It is hypothesised, that the lesser known CD27 receptor may overlap and synergise with CD40 and other relatives and may regulate the T-cell mediated activation cascade and control lymphocyte expansion by facilitating maturation and cell cycle progression. The diagnostic and prognostic value of the occurrence of these receptors and their soluble forms is becoming increasingly apparent, and recent data may herald new strategies for T cell manipulation via combinations of these receptors to ameliorate or prevent immune disease, coronary heart disease, transplant rejection, graft-versus-host disease, viral infections, and to promote tumor eradication.

an exclusive role in immunity and may have evolved by gene duplication as additional control proteins having their impact on onset, continuation andror termination of the response. Viruses known to manipulate antigen-recognition and processing and produce TNFR-like proteins for their own purposes, also affect other receptors of this family, which will help us to elucidate the role of these receptors in immunity. The first part of this issue presents an overview of all the receptors and ligands of the family known to date, and reviews on the well-known receptors CD40 and CD95 ŽApo-1rFas .. No review is included on NGFR, as the literature has become remarkably silent after the flurry of excitement in the early 1990s. The second part includes receptors, where many of the data have been collected by just a few laboratories: CD30, OX-40, 4-1BB and CD27. Recurrent themes list cross-talk with the CD28 route and overlap within the family with respect to tissue distribution and usage of signal transduction molecules, but these receptors may have additional and potentially crucial balancing roles affecting the outcome of immune disease, transplant rejection, graft-versus-host disease and tumor-specific immunity. Other factors influencing these diseases, including the role of TNF, have been reviewed elsewhere and it is becoming clear, that T cell suppression in the absence of knowledge on etiology or disease-triggering antigen may become a feasible approach for treatment via the co-stimulatory routes. Gravestein and Borst Žpage 423 of this issue. present an overview of the receptors of the family and their ligands known to date, summarising tissue distribution and function, and the phenotype of knockout mice. The receptors are characterised by homology in the ligand binding domain, but share few intracellular features. Information on signal transduction remained elusive for many years, but with the advent of new technologies these pathways are now rapidly being tracked down. On the outside, surface-associated and soluble forms of receptors and ligands apparently lead to reciprocal signalling result-

Key words: TNFR gene family r CD27r lymphocyte survival r cell cycle r immune homeostasis Q1998 Academic Press

IN ORDER TO MOUNT an effective immune response, naive T cells require presentation by antigen as well as additional co-stimulatory signals. Once the activation cascade is underway, unbridled proliferation hampering efficiency and specificity of the attack must be avoided, and finally, once the foreign antigen has been deleted, a proper termination of the response must take place, eliminating effector cells and generating memory. One co-stimulatory route to the AgR involves the CD28rB7 pathway, another the relatives of the nerve growth factor ŽNGF.rtumor necrosis factor ŽTNF. receptor family. The latter is proving of crucial importance for a proper response to antigen as well as immune homeostasis via their control of apoptosis. Some receptors appear to play

From the Department of Pathology, Maastricht University, P.O. Box 5800, 6202 AZ Maastricht, The Netherlands Q1998 Academic Press 1044-5323r 98r 060417q 06$30.00r 0

417

Wil A.M. Loenen

ing in expansion of the activation cascade. The role of the soluble forms is often far from clear and is important to resolve, as discussed below. The receptors basically fall into two groups, one involved in cell survival Žincluding CD40., the other, characterised by the presence of a so-called death domain Že.g. CD95., in apoptosis via the caspase route. Some receptors, such as TNFR-I p55, share features with both groups, thus in part explaining the pleiotropic biological effects by TNFs. The receptors without death domain associate with a family of TNFR-associated factors ŽTRAF., that link via the NIK and JNK routes to key transcription factors such as NF-kB and c-jun, thus controlling many aspects of the response to antigen. The division in cell death receptors and those promoting survival is not strict, probably a reflection of the emerging and intricate array of stimulators, inhibitors and bifurcations, that control both apoptosis and TRAF routes at different levels. TRAF knock-out mice are viable but develop serious organ defects leading to premature death, while the phenotype of receptor-deficient mice indicates that these receptors are not uniquely required for early lymphoid development. Vogel and Noelle Žpage 435 of this issue. review the role of CD40 in immunity. Originally defined some 10 years ago as a B cell-specific receptor, CD40 as well as its ligand CD40L are now known to be expressed on many other cells and act at virtually all stages in both cell-mediated and humoral responses. Important information on the way cell survival and apoptosis-promoting receptors compete is slowly being assembled. Co-ligation of CD40 and CD95 on B cells leads to either expansion or apoptosis and a decisive role for the AgR has been proposed, tipping the balance towards continued survival and concomitant differentiation if the B cells are truly antigenspecific, but eliminating inappropriately stimulated bystander or tolerant B cells. Interplay between the CD40 and CD28 routes involves mutual effects on the induction of ligands, affecting interactions between T cells and APCs, extravasation of lymphocytes, and directing other T-cell mediated events and decisions about anergy and tolerance. CD40 triggering leads to activation of tyrosine kinases as well as the TRAF route. Details on varying ratios of TRAF molecules in the signalling complex upon CD40 triggering in conjunction with other stimuli are emerging, and link to different biological responses. Such data resemble the formation of different transcription factor complexes of, e.g. Fos and Jun, discovered a decade ago,

and would allow the cell to react or alter its response quickly. In line with these pleiotropic activities, immunotherapeutic interventions in diseases with known defects in this route, but also in other illnesses, including possibly coronary heart disease, appear to be able to block inflammatory reactions, autoantibody formation, restore imbalances in T helper ratios and induce tolerance via its interplay with the CD28 route. Treatment seems to benefit both early and late events, as one would have hoped to see, but short periods of treatment have sometimes surprisingly longlasting effects, especially when combined with CD28. The authors end with the dependence of tumor-specific immunity on CD40 and this topic is enlarged upon by Toes, Schoenberger, Van der Voort, Offringa and Melief Žpage 443 of this issue. in their review on cytotoxic T cell priming and anti-tumor immunity. The detailed mechanism behind this route is slowly being unravelled, from the processing and presentation of tumor-specific antigens by specialised APC to the necessity for T helper cells and epitope linkage with the reactive CTL. The involvement of CD40 in T cell help has triggered research in its role in these interactions and the authors review recent evidence that CD40 signals can replace the helper cells. It supports a model in which CTL priming is a CD40-dependent two-step event: helper T cells activate the APC, the activated APC in turn the CTL, and in this way temporal separation between helper T and CTL cell allows two-step control as well as expansion of the response. These results indicate, that restoration of the response of potentially tumor-reactive lymphocytes in patients, due to lack of T cell help to the APC, may be achieved by directly stimulating the APC via the CD40 route, leading to CTL-mediated tumor eradication. Moulian and Berrih-Aknin Žpage 449 in this issue. review the stages at which CD95 and its ligand CD95L may participate in immune homeostasis, and how disruption can lead to autoimmunity and organ-destruction. The authors also mention, that tumor tolerance, in addition to lack of antigen-recognition and co-stimulation, may be due to a counterattack by CD95L-expressing tumor cells on infiltrating ŽCD95q. T cells. Some observed CD95 effects may relate to the emerging fight for control over life and death with its survival-promoting relatives, as discussed above with respect to CD40. In this way, CD95 would indirectly affect T-cell mediated toxicity, activation-induced cell death and immune privilege, though a more direct way cannot be excluded. In addition to this struggle 418

Editorial overview: CD27 and (TNFR) relatives in the immune system: their role in health and disease

within the family, there is some evidence emerging for control of the apoptotic activity of CD95 by downstream intracellular regulators, while on the outside ligand-mediated signals are influenced by levels of soluble CD95 andror membrane-bound and soluble CD95L in the vicinity. Attempts to use the information on CD95L-mediated killing of infiltrating lymphocytes to prevent rejection by expressing CD95L in the transplant, have met with both positive and negative results and such immunotherapeutic strategies will require more fundamental knowledge on the role of CD95 in these processes. Aberrant CD95 function can lead to immune disease in two ways. Decreased or absent activity may lead to lymphoproliferative disease and eventual autoimmunity, due to survival of potentially autoreactive cells. In other illnesses, high rather than absent CD95, e.g. as a result of lymphocyte activation, correlates with disease severity. As activation also induces its relatives, dissection of their individual receptor contributions to these diseases is essential. The authors continue with a summary on CD95related autoimmune diseases and altered susceptibility to others, and finish with their recent data on CD95 in the prototype B-cell mediated autoimmune disease myasthenia gravis ŽMG.. MG is associated with different thymic abnormalities, and autoreactive T lineage cells are known to play a role in the autoantibody-mediated attack at the neuromuscular junction. An enlarged population of CD95high cells has been identified in the thymus that comprise autoreactive cells, thus linking aberrant CD95 expression with this disease. In the periphery, the authors also found significant alterations in the CD4q subset, as previously also recorded for various other autoimmune diseases. The tentative conclusion could be that these CD95 high autoreactive cells have somehow failed to enter the apoptotic pathway, thus leading to recognition of self. In the second part of this issue, Horie and Watanabe Žpage 457 of this issue. survey the literature on CD30. This receptor was originally defined as a marker in Hodgkin’s lymphoma and present on the enigmatic giant multinuclear Reed]Sternberg ŽRS. cells characteristic for this disease. Expression in normal tissue appeared to be extremely limited, but using more sensitive fluorescence techniques, other cells are now known to express CD30. This extended tissue distribution may reflect the pleiotropic biological effects of CD30 signalling, some of which overlap with CD40, and the authors suggest that CD30 may play a wider role in the transition of naive to commit-

ted effector cells and interactions between T and B cells, than originally thought. In addition to this role in the periphery, CD30 knock-out mice establish a role for CD30 in negative selection, as such animals have an increased number of thymocytes, and altered susceptibility to anti-CD3 mAb-mediated apoptosis. A soluble form of CD30 ŽsCD30. in serum has proven to be an important factor for poor prognosis in various illnesses including viral infections, autoimmune diseases with a link to Th2 responses, and lymphomas. In the case of HIV, a second link exists with CD30, as surface expression is induced by the virus, and subsequent CD30 cross-linking allows HIV replication in latently infected cells Žprobably via its effect on NF-kB .. Whether sCD30 is simply the proteolytic result of spent surface receptor, or has an additional role to play once released from its original cell, remains to be elucidated. In addition to the prognostic value of sCD30, the surface receptor provides a tool for diagnosis and classification of lymphomas, and immunotoxins based on CD30 may allow therapeutic intervention. Finally, the authors perhaps allow us glimpses of the future, as they describe an aberrant CD30 protein in some cells, a potential cytosolic competitor. On the basis of the observed CD30-mediated depletion of TRAFs, they propose a model in which CD30 limits its own ability to transduce cell survival signals through proteolysis of these transducers. Finally, depletion of intracellular TRAF2 and associated proteins increased sensitivity to apoptosis during activation of death-promoting receptors such as TNFR-1. Such results open up new avenues of research that will take many more years to complete. Fundamental questions in T-cell mediated responses concern both the mechanisms via which the various subsets of effector cells are generated, but also how they reach and effect their function at the target site. Weinberg, Vella and Croft Žpage 471 of this issue. discuss the role of the OX-40 protein during this later phase in the immune response. First identified in Oxford in rats a decade ago as a receptor on activated CD4q T cells, research on OX-40 is now rapidly gaining momentum. Its ligand, OX-40L, proved to be identical to the HTLV-1-regulated gp34 protein, linking this family member, like CD30, to retroviruses that affect the immune system. As expected OX-40 provides a strong co-stimulatory signal with the AgR, but appears to affect only later stage effector and memory cells. Activation of this pathway enhances mutual responses of OX-40q T cells and OX-40Lq APC. The presence of OX-40L on vascular endothelial cells suggested a potential role in ex419

Wil A.M. Loenen

travasation, like its relative CD40, and in line with this, adhesion of activated T cells to the endothelial cell wall can be blocked by mAb to OX-40 or OX-40L. In line with its relatives, manipulation of T cell function via OX-40 could have profound effects on the clinical outcome of disease, and the authors report their results, mainly published within the past 2 years, that are starting to confirm such speculations. Tumor-specific immunity and memory was enhanced in mice after in vivo co-stimulation through OX-40. In the EAE rodent model for multiple sclerosis OX-40q and OX-40Lq cells in the CNS correlate with auto-antigen specificity and reactivity and studies with OX-40-based immunotoxins and fusion proteins support the notion, that OX-40 plays a critical role in this T-cell mediated inflammatory event in the CNS. Blocking this interaction may not have a lasting effect on T cell function, which perhaps relates to the efficacy of inhibiting the early response via manipulations of, e.g. the CD40 and CD28 routes and longlasting effects may be enhanced by combining these regimes. Vinay and Kwon Žpage 481 of this issue. review the current status of the 4-1BB receptor. Like its relative OX-40, 4-1BB was discovered as a receptor with an extremely limited distribution on activated T-lineage cells. Originally identified in mice, its human homologue as well as human and murine ligands have been cloned in more recent years. 4-1BB resembles OX-40 as a receptor involved in the effector phase of the immune response, with evidence for interplay with the CD28 route and bidirectional signalling benefitting both 4-1BBq T cells and 4-1BBLq APC. The authors suggests that this receptorrligand pair may be very exclusive for this interaction, without a wide ranging effect on other cell types involved in the immune response. However, in the light of CD30 expression on a range of other cells using more sophisticated detection techniques, and as the authors comment themselves on a potential role for 4-1BB in T cell memory, this warrants further investigation. While CD30 may promote Th2 responses, 4-1BB may favor Th1 cells. Like CD30 and OX-40, it shares features with CD40 with respect to lymphocyte maturation and adhesion, indicating potential overlap or synergistic effects of these receptors on those cells, on which both occur. Such synergy may be important with respect to the aforementioned struggle for survival between CD40 and CD95. The authors also describe a link between 4-1BB and retroviruses, influencing HIV pathogenesis, probably involving the NF-kB route, like CD30. The first results

on in vivo treatment with 4-1BB mAb are being published and fit in with tentative predictions that these would aggravate transplant rejection and graft versus host disease, but promote tumor-eradication. The authors finish with their most recent data on 4-1BB-deficient mice. These mice have elevated Ig responses and will be useful in further dissection of the respective roles of 4-1BB and its relatives in disease. In the final review of this issue, Lens, Tesselaar, Van Oers and Van Lier Žpage 491 of this issue. discuss data on the CD27 receptor and its ligand. In line with CD27 tissue distribution on most T-lineage cells, and subsets of B and NK cells, a substantial body of evidence, mainly generated by the Dutch and Boston groups, has been generated through studies on human peripheral blood cells, that support the notion that this receptor]ligand pair may control the size and function of lymphocyte populations upon antigenic triggering. This would fit in with the model proposed in my thesis on CD27 1 in which CD70 would release a CD27-imposed brake on cell division, as further discussed below. The pathway appears to be influenced by similar stimuli as its relatives, including interplay with the CD28 and CD40 routes, and signalling via TRAFs, as expected, but additionally appears to link to a novel death domain protein, Siva. Triggering of the CD27 route affects differentiation and proliferation of various subpopulations in different ways and this supports the notion, that CD27 facilitates maturation processes without interfering with programmed details of differentiation andror proliferation of each cell type. Like its relatives, a link between abnormalities in the CD27 pathway and various diseases is emerging. The soluble form of CD27, sCD27, released upon activation, can be used as disease marker in autoimmune disease and cancer, similar to, e.g. sCD30. The human CD27 ligand proved to be identical to the lymphoma antigen CD70, first described well over 10 years ago and known to be overexpressed in many B cell neoplasia. Such tumors often coexpress CD27 and other relatives, probably reflecting the activated phenotype of such malignancies. Additional reports on abnormal CD70 expression on macrophages in autoimmune inflamed organs await further conformation, but may be relevant in connection with earlier reports on the presence of activated T cells and sCD27 in synovial fluid of rheumatoid arthritis patients, and high sCD27 levels in the cerebrospinal fluid of people with multiple sclerosis. Finally, CD70 expression in atherosclerotic plaques has been re420

Editorial overview: CD27 and (TNFR) relatives in the immune system: their role in health and disease

ported, which if confirmed may prove to be an important observation in the light of the emerging usefulness of manipulations of the CD40 route in coronary heart disease Žcf. Vogel and Noelle Žpage 435 of this issue... As the authors comment, the issue of the role of CD27rCD70 in the unbridled expansion of malignant lymphocytes is unsettled. However, studies in B cell lines do provide some support for such a notion2 Žand unpublished observations., as do transfection studies in the T-ALL Jurkat 1 Žand unpublished results.. Jurkat does not tolerate overexpression of either receptor or ligand, but clonal expansion can occur when both cDNAs are expressed. Intriguing observations in this respect are the fact that overexpression of CD70 causes apoptosis, while that of CD27 appears to block cell division without killing the cell. The double transfectants are highly heterogeneous, with many giant multinuclear cells, reminiscent of the Reed]Sternberg cells in Hodgkin’s disease. In the light of reported links Žboth in vivo and in vitro . between multinucleation and defects in genes such as p53, mdm2 and cyclins, this suggests CD27 may not only facilitate survival but control cell cycle progression, as previously reported for CD40. Finally, a potential link with CD2 should be mentioned, as CD70-positive clones can be generated in the above experiments, which invariably lack CD2. As this adhesion receptor is involved in cell polarisation and promotes interactions between both helper and cytotoxic T cells and APC, this warrants exploration of cross-talk between these two routes in interactions between peripheral lymphocytes with their APC. These transfection data led to the hypothesis that CD27 controls peripheral lymphocyte expansion and support the model, that CD27 and CD70 interactions can contribute to autocrine growth of malignant lymphocytes. Initial data on CD27 knock-out mice do not show major disruption of lymphoid function. This could point to the existence of additional back-up proteins with overlapping andror synergistic functions, expected to exist, if CD27 indeed plays such a pivotal role. The current data on CD27 suggest synergism with CD40 in control of the whole T and B-cell mediated activation cascade, which will necessitate extensive cross-talk with both the CD40 and CD28 routes. The parallel with CD40 warrants investigation into potential overlap with its other relatives during later stages of the response. In this respect, 4-1BB and the recently identified GITR receptor which both

share cytoplasmic homology with CD27 are perhaps the most likely candidates. Finally, other exciting novel regulatory mechanisms appear on the T cell scene, that may control CD27 function, such as the reported extracellular post-translation modification of CD27 by the NADdependent ADP-ribosyltransferase protein3. As NAD apparently inhibits proliferation, cytotoxicity and cytokine secretion, and concomitant with the observed ribosylation, T cell homing to secondary lymphoid organs is affected, this observation may lead to more insight in the exact nature of CD27 function. Where is research on this receptor family taking us? A view from the sky would see CD27 and CD40 as receptors controlling the antigenic response from the surface of their respective T and B cells, facilitating differentiation and progression through the cell cycle of all but final stage immune effector cells, in a continuous struggle with apoptosis-promoting receptors such as CD95 c.s. to keep cell death away, in which fight they are on the losing end whenever inappropriate or insufficient signals are received Žreceptors involved in the cell survival route are not necessarily limited to this family, but could include other routes such as CD28 and interleukins.. Once expansion is allowed, other receptors including 4-1BB and OX-40 join forces, while perhaps additional death receptors are recruited on behalf of their opponents of the apoptotic route. Other relatives such as CD30, NGFR, TNFRs and some of the other newly identified relatives also somewhere enter the scene, and in the whole process, both intracellular cross-talk and control mechanisms as well as extracellular soluble and surface receptors and ligands play their role. The issue of the function of individual soluble receptors has been a matter of much debate and centers on a crucial question: Do they play a functional role or are they simply spent receptors, no longer needed by the cell? Evidence is accumulating for their biological importance, but also for the spent receptor hypothesis. In this respect the elegant work by Barclay and co-workers should not go unmentioned4 . Using BIAcore equipment, these scientists showed high affinity binding of soluble OX-40L to its receptor, immobilised on the BIAcore sensor chip, but low affinity binding in the reverse situation, using soluble receptor. The conclusion from these and other experiments was, that the receptor once cleaved from the cell membrane, will release its ligand and is unlikely to compete for ligand with other surface OX-40 molecules due to this low affinity. Obviously, it 421

Wil A.M. Loenen

References

is imperative to carry out such experiments on the other receptors in order to solve this highly important issue, as differential affinity of soluble and surface receptors for their ligands and vice versa will greatly affect expansion or limitation of the response, and thus influence decisions on immunotherapeutic treatment regimes. In the light of recent successes with recombinant CTLA4-Ig and blocking of the CD40 pathway in amelioration of autoimmune disease as well as prolonged graft acceptance and tumor rejection, and the combined data reviewed here, it is likely that the next decade will see major breakthroughs in the development of new experimental strategies to prevent or ameliorate these illnesses via manipulations of the receptors of this family.

1. Loenen WAM Ž1997. Molecular analysis of CD27 to elucidate the role in lymphocyte development. PhD Thesis, Faculty of Medicine, Leiden University, The Netherlands 2. Prasad KVS, Ao Z, Yoon Y, Wu MX, Rizk M, Jacquot S, Schlossman SF Ž1997. 3. Okamoto S, Azhipa O, Yu Y, Russo E, Dennert G Ž1998. Expression of ADP-ribosyltransferase on normal T lymphocytes and effects of nicotinamide adenine dinucleotide on their function. J. Immunol 160;4190]4198 4. Al-Shamkhani A, Mallett S, Brown MH, James W, Barclay AN Ž1997. Affinity and kinetics of the interaction between soluble trimeric OX40 ligand, a member of the tumor necrosis factor superfamily and its reactor OX40 on activated T cells. J Biol Chem 272:5275]5282

422