The phenotype and survival of antigen-stimulated transgenic CD4 T cells in vivo: the influence of persisting antigen.

Naive and primed/memory CD4 T cells are distinguished by changes in the expression of activation/adhesion molecules that correspond with an altered function. Adoptively transferred TCR transgenic (tg) CD4 T cells specific for ovalbumin peptide (OVA-pep) were analysed for changing phenotype and the speed of change in vivo following antigen challenge with alum-precipitated (ap) OVA-pep, a conjugate that stimulated a Th2-type cytokine response. The change of CD45RB in relation to number of divisions showed that the transition from CD45RB(hi) (naive) to CD45RB(low) (primed/memory) was incremental; with each cell cycle the number of CD45RB(hi) molecules on the cell surface was diluted by approximately half and replaced by the low-weight isoform. Similarly, the change to CD44(hi) expression increased gradually during four rounds of proliferation. The loss of CD62L expression occurred early and was independent of cell division. CD69 was up-regulated quickly within 1-2 cycles, but down-regulated after about seven divisions. The expression of CD49d was not altered during the early rounds of division, although it was up-regulated on 30-60% of tg T cells dividing repeatedly (>or=8 cycles). When analysed on day 3 following stimulation, CD25 was no longer up-regulated. The intra-peritoneal injection of ap-OVA-pep stimulated tg T cells in the spleen and mesenteric lymph node one day in advance of those in more distant peripheral lymph nodes. Evidence indicated that residual antigen persisted for at least 4 weeks and was able to stimulate naive tg T cells. However, residual antigen had no net effect on extending or reducing survival of the transferred population.

Naive and memory T cells migrate in comparable numbers through the normal rat lung: only effector T cells accumulate and proliferate in the lamina propria of the bronchi.

T cells reach the lung via the pulmonary and bronchial arteries that supply the alveolar and bronchial regions. Although these regions are differentially affected by T cell-mediated diseases, the migration of T-cell subsets in these two regions has not been studied. Naive, memory, and effector T cells were injected into congenic rats and traced in sections of normal lung. All three T-cell subsets were found in large numbers in the alveolar region and exited again quickly. Only effector T cells accumulated in the lamina propria of the bronchi. Further, 72 h after injection 6% of the effector T cells still proliferated in the lung, whereas apoptotic effector T cells were only observed 1 h after injection (0.2%). Thus, not only effector and memory but also naive T cells continuously migrated through the lung. The preferential accumulation of effector T cells in the bronchial lamina propria may explain why some diseases preferentially affect the bronchial region.

Both CD45R(low) and CD45R(high) "revertant" CD4 memory T cells provide help for memory B cells.

During a primary response to thymus dependent antigens, B cells undergo a number of qualitative changes to become memory B cells - processes that require co-stimulatory signals and cytokine help from CD4 T cells. The question of whether distinct, antigen-experienced memory CD4 T cells are subsequently needed to program memory B cells into antibody synthesis has not been clearly resolved. Using an adoptive transfer model in which memory but not naive B cells were stimulated, we evaluated CD4 T cell help using lymphocytes obtained from primed or unprimed thymectomized donors and expressing a naive (CD45R(high)) or a memory (CD45R(low)) phenotype. Memory B cells, most of which were committed to the IgG1 (Th2) subclass, could be stimulated to produce antibody using help transferred by the CD45R(high) naive subset of unprimed donors (slow onset of response), the CD45R(low) subset of 7 day primed donors (large, rapid antibody response) or by both the CD45R(low) and the CD45R(high) "revertant" subsets of 6 month primed donors. We found that antigen primed CD45R(low) CD4 T cells reverted (defaulted) with time to a CD45R(high) resting state, a change that was prevented by persisting antigen. The evidence suggests that CD4 memory T cells are partitioned into two different functional states (CD45R(high) and CD45R(low)) and that these determine the characteristics of the memory B cell response in terms of speed, size and longevity.

Naive and memory T lymphocytes migrate in comparable numbers through normal rat liver: activated T cells accumulate in the periportal field.

Although the liver is known to contain a significant number of lymphocytes, migration of these through the compartments of the liver, parenchyma and periportal field, has not been studied. The periportal field, in particular, is affected in several immunological disorders of the liver. Populations of labeled naive, activated, and memory T cells were injected into congenic rats. The recipient livers and draining lymph nodes were removed at various time points, and cryostat sections were analyzed for the presence of donor cells using quantitative immunohistology. Donor cell proliferation and apoptosis were examined in vivo by BrdU (5 microM 5-bromo-2-deoxyuridine) incorporation and the TUNEL technique, respectively. Early after injection (0.5-1 h), naive, activated, and memory T cells were localized to the parenchyma and periportal field in comparable numbers. With time, all T cell subsets left the parenchyma but remained or, in the case of activated T cells, significantly accumulated in the periportal field. Furthermore, 12% of activated donor T cells proliferated in vivo within the periportal field, and 0.5% showed evidence of apoptosis. Taken together, not only activated and memory, but also naive T cells continuously migrate through the liver, showing a preference for the periportal field, and activated T cells mainly proliferate there. This may explain why many immunological liver diseases predominantly affect the periportal field.

Lymphocyte trafficking: CD4 T cells with a 'memory' phenotype (CD45RC-) freely cross lymph node high endothelial venules in vivo.

Antigen encounter not only induces a change in surface expression of CD45RC isoforms in the rat from a high (CD45RC+) to a low molecular weight molecule (CD45RC-), but also stimulates changes in expression of adhesion molecules that regulate CD4 T-cell migration. T cells with an activated or 'memory' phenotype (CD45RC-) are thought to enter lymph nodes almost exclusively via afferent lymphatics whereas T cells in a resting state (CD45RC+) migrate across high endothelial venules (HEV). The present study monitored the rapid recirculation from blood to lymph of allotype-marked CD45RC T-cell subsets. Surprisingly, we found that CD45RC- CD4 T cells entered the thoracic duct slightly faster and reached peak numbers 3 hr earlier (18 hr) than did the CD45RC+ subset. To determine whether the entrance of CD45RC+ and RC- subsets was restricted to HEV and afferent lymphatics, respectively, recirculation of CD4 T cells was monitored in mesenteric lymphadenectomized (MLNx) rats (on healing the intestinal afferent lymphatics are joined directly to the thoracic duct), or in recipients that had had the mesenteric lymph node (MLN) acutely (2-3 hr) deafferentized (entry would be restricted to HEV). In these studies CD45RC- CD4 T cells entered the MLN across HEV on an equal basis with T cells expressing a CD45RC+ phenotype. Contrary to currently held dogma the results showed that, in vivo, CD4 T cells with a memory phenotype freely enter lymph nodes (LN) across HEV as well as via afferent lymphatics.

CD4+ T cells of both the naive and the memory phenotype enter rat lymph nodes and Peyer's patches via high endothelial venules: within the tissue their migratory behavior differs.

It is thought that naive T cells predominantly enter lymphoid organs such as lymph nodes (LN) and Peyer's patches (PP) via high endothelial venules (HEV), whereas memory T cells migrate mainly into non-lymphoid organs. However, direct evidence for the existence of these distinct migration pathways in vivo is incomplete, and nothing is known about their migration through the different compartments of lymphoid organs. Such knowledge would be of considerable interest for understanding T cell memory in vivo. In the present study we separated naive and memory CD4+ T cells from the rat thoracic duct according to the expression of the high and low molecular weight isoforms of CD45R, respectively. At various time points after injection into congenic animals, these cells were identified by quantitative immunohistology in HEV, and T and B cell areas of different LN and PP. Three major findings emerged. First, both naive and memory CD4+ T cells enter lymphoid organs via the HEV in comparable numbers. Second, naive and memory CD4+ T cells migrate into the B cell area, although in small numbers and continuously enter established germinal centers (GC) with a bias for memory CD4+ T cells. Third, memory CD4+ T cells migrate faster through the T cell area of lymphoid organs than naive CD4+ T cells. Thus, our study shows that memory CD4+ T cells are not excluded from the HEV route. In addition, "memory" might depend in part on the ability of T cells to specifically enter the B cell area and GC and to screen large quantities of lymphoid tissues in a short time.

The peripheral T-cell pool: regulation by non-antigen induced proliferation?

Despite continuous perturbations, the recirculating lymphocyte pool remains relatively constant. Expansion of the pool is compensated for by cell loss. When the T-cell pool is in deficit, either from a congenital defect or an acquired immunodeficiency, T-cell numbers are restored by extrathymic division-a response that occurs without deliberate provocation. We have considered how the recirculating pool may be stably maintained and how a T-cell deficit might be restored to equilibrium. Recent evidence suggests that depleted T-cell compartments are replenished by CD4 T-cell proliferation in the absence of specific antigen, a response that occurs without engaging the T-cell receptor.

Both activated and nonactivated leukocytes from the periphery continuously enter the thymic medulla of adult rats: phenotypes, sources and magnitude of traffic.

Although the thymus is primarily noted for the export of T cells to the periphery, a small influx of cells has also been observed. It is still a matter of debate whether entry into the thymus depends on prior activation. The phenotypes, sources and degree of immigration are largely unknown. We monitored by quantitative immunohistochemistry the entry of cells from the periphery into the rat thymus in three experimental models. We injected i.v. recirculating, small, nonactivated CD4+ T cell subsets, often referred to as naive (CD45RC+) and memory or antigen-experienced (CD45RC-) cells, purified from thoracic duct lymph of allotype-marked donors, allotype-marked leukocytes released from spleen or lung transplants, or leukocytes labeled in the periphery for 12 weeks during the S-phase of the cell cycle by oral application of 5-bromo-2-deoxyuridine (BrdUrd). Early after i.v. injection (0.5 h), significantly more antigen-experienced (CD45RC-) CD4+ T cells entered the thymus, and by 24 h four times as many cells from the CD45RC- subset as from the CD45RC+ subset had entered the thymus and localized to the medulla. None of the thymic entrants expressed the interleukin (IL)-2 receptor. Following spleen transplantation approximately 40% of donor cells entering the thymic medulla were T cells and approximately 55% were B cells. In contrast, from a lung transplant, approximately 85% of peripheral immigrants were T cells and approximately 10% were B cells. After both procedures, a small number of NK cells and monocytes/macrophages were found among the immigrants (< 5%). Rats were fed BrdUrd continuously for 12 weeks, a procedure which labeled approximately 30% of peripheral lymphocytes but not cortical thymocytes. BrdUrd-labeled cells were localized almost exclusively to the thymic medulla and represented approximately 10% of medullary cells. Of the thymic immigrants approximately 50% were T cells, approximately 30% were B cells (including approximately 15% IgD+ cells), approximately 15% were NK cells and the remainder (approximately 5%) were monocytes/macrophages. Only a quarter of BrdUrd-labeled cells expressed the IL-2 receptor. The thymus is continuously infiltrated by both activated and nonactivated leukocytes from the periphery, including T cells, B cells, NK cells and monocytes. These immigrants are supplied by lymphoid and nonlymphoid organs in a characteristic subset composition. Their entry is facilitated by prior antigen experience or activation. Thus, the participation of the thymic medulla in general leukocyte traffic suggests a mechanism by which the T cell repertoire could potentially be modulated by the peripheral tissues.

Migration pathways of CD4 T cell subsets in vivo: the CD45RC- subset enters the thymus via alpha 4 integrin-VCAM-1 interaction.

The present investigation examines the localization and migration of purified T cell subsets in comparison with B cells, CD8 T cells and CD4+ CD8- single-positive thymocytes. CD4 T cell subsets in the rat are defined by mAb MRC OX22 (anti-CD45RC), which distinguishes resting CD4 T cells (CD45RC+) from those (CD45RC-) which have encountered antigen in the recent past--subpopulations often referred to as 'naive' and 'memory'. Purified, 51Cr-labelled CD45RC+ CD4 T cells broadly reflected the migration pattern of CD8 T cells and B cells. Early localization to the spleen was followed by a redistribution to mesenteric lymph nodes (MLN) and cervical lymph nodes (CLN), B cells migrating at a slightly slower tempo. There was almost no localization of these subpopulations to the small or large intestine [Peyer's patches (PP) excluded]. In contrast, CD45RC- CD4 T cells (indistinguishable in size from the CD45RC+ subset) localized in large numbers to the intestine; they were present here at the earliest time point (0.5 h), persisted for at least 48 h but did not accumulate, indicating a rapid exit. Numerically, localization of CD45RC- CD4 T cells in the MLN could be accounted for entirely by afferent drainage from the intestine. Unexpectedly, CD45RC- CD4 T cells (but not other subsets) localized and accumulated in the thymus. In vivo treatment with mAb HP2/1 against the integrin alpha 4 subunit inhibited almost entirely CD45RC- CD4 T cell migration into the PP (98.1%), intestine (87.1%), MLN (89.1%) and thymus (93.5%); migration into the CLN was only reduced by half. To distinguish between recognition of MAdCAM-1 and VCAM-1 by alpha 4-containing integrins, recipients were treated with mAb 5F10 against rat VCAM-1. Except for the thymus and a small reduction in CLN, localization of CD45RC- CD4 T cells was unaffected; entry to the thymus was almost completely blocked (92.3%) by anti-VCAM-1. The results indicated (i) that CD45RC- CD4 T cells alone showed enhanced localization to the gut and PP, probably via alpha 4 beta 7-MAdCAM-1 interaction; (ii) that many CD45RC- cells entered non-mucosal LN independently of alpha 4 integrin or VCAM-1; and (iii) that entry of mature recirculating CD45RC- CD4 T cells into the thymus across thymic endothelium was apparently regulated by alpha 4 integrin-VCAM-1 interaction.

Membrane CD45R isoform exchange on CD4 T cells is rapid, frequent and dynamic in vivo.

CD4 T cells bearing high (240-190 kDa) and low (180 kDa) molecular mass isoforms of the leukocyte common antigen CD45 define functionally distinct subsets which have been equated with naive and memory T cells. In the rat, CD4 T cells expressing a high molecular mass isoform [identified by monoclonal antibody MRC-OX22 (anti-CD45RC)] exchange this for the 180 kDa molecule (CD45RC-) when stimulated by antigen. Here we show, by transferring mature allotype-marked CD45RC- CD4 T cells (depleted of immature Thy-1+ CD45RC- recent thymic emigrants) into normal euthymic recipients, that many T cells re-express the high molecular mass isoform in less than 6 h. By 24 h, 30-60% of CD45RC- CD4 T cells became CD45RC+; within a week the entire cohort appeared to exchange the low for the high molecular mass isoform. Isoform exchange was dynamic and many CD4 T cells returned once again to the CD45RC- state. CD45RC- CD4 T cells declined in number more rapidly than the CD45RC+ subset after transfer. The results suggest that CD45R isoforms distinguish between resting T cells (CD45RC+) and those which have encountered antigen in the recent past. CD45R isoforms would appear to be unsuitable markers of naive and memory T cells.

Rapid re-expression of CD45RC on rat CD4 T cells in vitro correlates with a change in function.

Rat CD4+ T cells are divided phenotypically by the anti-CD45RC monoclonal antibody OX22 into subsets with contrasting functions. Stimulation of T cells in vitro is known to induce a change in isoform from CD45RC+ to CD45RC-. We have investigated the in vitro conditions which promote a switch in isoform in the opposite direction. We observed that a majority of CD45RC- CD4 T cells (> 90%) spontaneously re-expressed CD45RC during the first 1-3 days of culture in both the presence and absence of alloantigen. The T cells remained CD45RC+ when cultured for 7 days in serum-free growth medium. However, alloantigen-activated lymphocytes, expressing the interleukin-2 receptor (IL-2R), downregulated CD45RC by day 4 and remained CD45RC- during the course of the experiment. Using mixtures of allotype-marked CD45RC+ and CD45RC- T cells, it was demonstrated that each subset showed comparable survival, IL-2R expression and time courses of activation in response to alloantigen. The repertoire of neither subset was, therefore, deficient in terms of allorecognition. The rapid re-expression of CD45RC in culture was accompanied by a change in function: CD45RC+ "converts", obtained by overnight culture of CD45RC- T cells, induced significantly higher graft-versus-host responses. Thus, the transition in culture from CD45RC- to CD45RC+ reflects a major functional reprogramming of the cell and not a trivial modulation of a surface antigen.

CD45R CD4 T cell subset-reconstituted nude rats: subset-dependent survival of recipients and bi-directional isoform switching.

High and low molecular weight variants (CD45R) of the leukocyte common antigen (CD45) divide CD4 T helper cells into subpopulations which display distinct characteristics. In vitro and in vivo evidence suggested that the presence of the high molecular weight splice variants CD45RA or CD45RB distinguished naive CD4 T cells from memory T cells which underwent an irreversible switch to the low molecular weight isoform as a consequence of antigen encounter. In the rat monoclonal antibody MRC OX22 identifies an epitope on the CD45RB molecule. We investigated this proposed differentiation pathway by reconstituting athymic nude rats with highly purified OX22+ or OX22- CD4 T lymphocyte subsets obtained from the thoracic duct (TDL) of euthymic, congenic, allotype-marked donors. Injection of CD4+CD45RB- (45R-) cells ensured long-term survival of nude recipients; recipients of CD4+CD45R+ (45R+) cells died within 2-3 months of injection. Early after transfer (3-4 weeks) the progeny of both 45R+ and 45R- TDL were uniformly 45R-. With time (by 2 months) progeny of both parental types expressed the high molecular weight CD45RB isoform. Nude recipients of 45R- TDL always generated progeny a proportion of which bore the 45R+ phenotype; 3 months to 2 years post-injection, 30%-50% of the donor-derived CD4 T cells were 45R+, 45R- progeny isolated from primary recipients of either 45R- or 45R+ cells transferred into secondary nude recipients induced skin allograft rejection with equal effectiveness and also generated 45R+ offspring. The results indicated that CD4 T cell subsets switched between CD45R isoforms and that the change between high and low molecular weight expression was bi-directional. The splice variants apparently are not lineage or maturation markers, but rather identify CD4 T cells that exist transiently in different functional states.

Interconversion of CD45R subsets of CD4 T cells in vivo.

T lymphocytes express multiple forms of the leukocyte common antigen CD45, transcribed by alternative usage of leukocyte-common antigen exons 4-6. Species-specific monoclonal antibodies against restricted epitopes (CD45R) of the antigen subdivide CD4 T cells into reciprocal subsets expressing either the high molecular weight isoforms CD45RA or RB or a molecule in which exons 4-6 have been spliced out (CD45R0). CD45R+ or RB+ CD4 T cells are potent in graft-versus-host reactions, and interleukin-2 related activities, whereas the CD45R0+ subset responds in vitro to recall antigens and provides help for antibody synthesis. It is unclear whether CD45R subsets derive from separate lineages, or are products of unidirectional or reversible differentiation. We show by transferring CD45R+ or CD45R- allotype-marked CD4 T cells into athymic nude rats that both subsets routinely generate cells of the opposite phenotype with a function that follows phenotype, not parentage. The recent equation of CD45R subsets as maturation stages representing 'naive' and 'memory' T cells is difficult to reconcile with this finding.

Allograft rejection in athymic nude rats by transferred T cell subsets. II. The response of naive CD4+ and CD8+ thoracic duct lymphocytes to an isolated MHC class I disparity.

Athymic PVG-rnu/rnu (RT1c) rats were grafted with skin bearing isolated MHC disparities 7-14 days in advance of cell transfer. The ability of naive CD4+ or CD8+ thoracic duct lymphocytes to induce rejection was assessed by adoptive transfer of one or both T cell subsets into nude recipients bearing congenic PVG.r1 (MHC class I-only disparity, Aa) or PVG.r19 (class I and II-only disparity, Aa B/Da) skin grafts. Recipients of purified CD4+ TDL always rejected r19 allografts, whereas CD8+ TDL were ineffective against this class I + II difference. Neither the injection of CD4+ TDL nor CD8+ TDL alone resulted in destruction of r1 skin grafts. However, rejection of r1 tissue was observed in 63% of cases (19/30) when both CD4+ and CD8+ TDL were present in the nude recipients. Rejection of r1 skin was also induced in some recipients when CD8+ TDL were transferred 8 weeks in advance of CD4+ TDL. In contrast, sequential transfer in the reverse order apparently induced tolerance in the CD4+ population--i.e., surviving r1 skin grafts on 8 week CD4+ T cell-reconstituted nude recipients were not rejected following the subsequent transfer of CD8+ TDL. We conclude that CD4+ T cells were required for rejection of both class I and class II differences. In the presence of a class II difference, CD4+ T cells function autonomously to initiate both the inducer and effector stages of rejection. When the disparity is confined to class I, CD8+ T cells are required (probably at the effector stage) but are dependent on CD4+ T cells for help. There was no evidence of CD4+ effector T cells that could recognize class I directly within the graft.

Allograft rejection in athymic nude rats by transferred T-cell subsets. I. The response of naive CD4+ and CD8+ thoracic duct lymphocytes to complete allogeneic incompatibilities.

PVG.rnu/rnu nude rats were pre-grafted with two allogeneic skin grafts, AO(RTlu) and BN(RTln), 6-14 days in advance of cell transfer. Cellular requirements for rejection were established by transferring graded numbers of B cell-depleted (Ig-) thoracic duct lymphocytes (TDL) or purified W3/25+ (CD4+) or OX8+ (CD8+) TDL subsets. Allografts were rejected by 10(5) to 5 x 10(6) Ig- TDL in a dose-dependent fashion. A similar dose-response relationship was found by transferring 5 x 10(5) to 5 x 10(6) Ig- OX8- TDL (purified by depletion of B cells and OX8+ cells). Larger numbers of Ig- OX8- TDL (10-30 x 10(6)) did not significantly accelerate rejection. W3/25+ TDL alone (10(5)), highly purified by fluorescence-activated cell sorting (FACS), were sufficient to induce allograft rejection in this athymic nude rat model. In contrast, 10 times more FACS purified OX8+ TDL (10(6)) were unable to initiate skin graft rejection despite the complete class I and class II MHC incompatibilities. Furthermore, the addition of 10(6) OX8+ cells did not accelerate or retard the rejection induced by 10(5) W3/25+ cells alone. Pre-grafted nude recipients, irradiated (500 R) 2 hr before W3/25+ TDL injection, in order to eliminate putative nude T cells, rejected allografts on the same day as unirradiated controls. We conclude that when confronted with complete MHC disparities, CD4+ T cells are necessary and sufficient to induce skin allograft rejection whereas CD8+ T cells do not appear to contribute.

The origin of T cells in permanently reconstituted old athymic nude rats. Analysis using chromosome or allotype markers.

Athymic PVG-rnu/rnu rats receiving a single intravenous injection of syngeneic euthymic thoracic duct lymphocytes (TDL) develop normal levels of CD4+ T lymphocytes and survive for more than 2 years in a conventional animal house. We investigated the origin of the T cells (and B cells) in reconstituted nude recipients by transferring TDL carrying either the 3T chromosome marker or the RT6b + Igk-1b allotype or the RT7b (leucocyte-common) allotype markers. Karyotype analysis of spleen and lymph node (LN) cells from 1- to 2-year-old PVG-3T/3T-reconstituted nude recipients, stimulated in vitro with phytohaemagglutinin (PHA), unexpectedly revealed that a majority (79-97%) of dividing cells were of nude origin. However, extensive nude cell division was also recorded in PHA-stimulated cultures using mixtures of euthymic (PVG-3T/3T) and unreconstituted nude spleen cells; the assumption that only T cells divide in PHA-stimulated cultures thus appears to be erroneous. In contrast to the karyotype analysis, sIg- RT6b+ LN cells obtained from nude recipients reconstituted 2 years earlier with PVG-RT6b allotype-marked TDL, were all of donor origin with no indication of a nude-derived sIg- RT6a+ population. Igk-1b+ donor B cells were not found in these same recipients. Dual fluorescence analysis of TDL from 18- to 20-month RT7b-reconstituted nudes showed that 91-100% of CD4+ cells were donor-derived. When tested functionally, sIg- RT7b+ (donor) cells, but not sIg- RT7b- (nude-derived) cells, were able to reject skin allografts and induce local graft-versus-host (GVH) responses. Donor T cells, in contrast to CD4+ cells of nude origin, divided extensively in nude recipients; FACS-purified RT7b+ (donor) TDL retransferred from 17-month primary reconstituted nude rats, expanded further (60-100-fold) in secondary nude recipients. In conclusion, only the donor-derived CD4+ cells in reconstituted nude rats displayed T-cell function; evidence to the contrary from karyotype analysis was flawed. At no stage in their life did uninjected or T-cell reconstituted nude rats develop endogenous cells that in any way resembled CD4+ products of the thymus.

Fidelity of the repertoire in T cell reconstituted athymic nude rats. Preservation of a deficit in alloresponsiveness over one year.

A single intravenous injection of a relatively small number of T cells contained in the population of rat thoracic duct lymphocytes (TDL) is sufficient to restore to normal the peripheral T cell pool of athymic PVG.rnu/rnu nude rats. The donor T cells expand greater than 10-15-fold, self-renew, and restore immunocompetency to nude recipients permanently (greater than 2 yr). We asked whether the T cell repertoire was affected by the expansion and self-renewal process. Nude recipients were injected with syngeneic PVG TDL that had been allospecifically depleted (negatively selected) by consecutive passage from blood to thoracic duct lymph through two irradiated (DAxPVG)F1 intermediate rats. Negatively selected TDL were tested before transfer by the P----F1 popliteal LN GVH assay and showed a greater than 90% depletion of specific reactivity to DA alloantigens. Surviving cells or their progeny were recovered from LN or TDL of nude recipients 8 and 12 mo after transfer. The deficit in GVH reactivity to the DA haplotype persisted, but normal GVH activity was demonstrated against a third party (AOxPVG)F1 alloantigen. The "hole" in the repertoire could not be attributed to tolerance induced by the co-transfer of contaminating irradiated F1 TDL. PVG TDL passaged consecutively through (AOxPVG)F1 and (DAxPVG)F1 intermediates and devoid of (AOxPVG)F1 cells remained specifically depleted to both AO and DA haplotypes when recovered from nude recipients 4 and 13 mo later, but displayed GVH activity to a third-party (BNxPVG)F1 alloantigen. Thus the exact specificity of the T cell repertoire of the original inoculum was faithfully maintained in nude recipients throughout the initial phase of rapid expansion and the continued self-renewal of the mature peripheral T cell pool.

Major histocompatibility complex control of NK-related allogeneic lymphocyte cytotoxicity in rats. The contributions of strong and medial transplantation antigens.

Allogeneic lymphocyte cytotoxicity (ALC) describes the elimination of allogeneic lymphocytes in vivo by an NK-related activity. There is evidence that ALC is demonstrable between donor and recipient when these are incompatible at MHC gene loci alone. Since ALC is a property of T cell-deficient nude rats, the role of the MHC in this rejection process needs further study. We have determined the contribution of the MHC to ALC using congenic and recombinant rats. In our analysis we have assumed that ALC involves the recognition of classic alloantigens by clonally distributed effector cells as for other examples of transplant rejection, although this is not yet proved. Strong ALC was measured between congenic rats that differed for MHC genes only. Non-MHC incompatibility alone did not elicit ALC. In the presence of MHC incompatibility the strength of ALC generated in a recipient was dependent on non-MHC genes. The PVG background generated high ALC responses whereas ALC was not measured in the DA rat. However ALC was measured in the congenic PVG-RT1avl (DA) rat. The contributions of classic class I (RT1.A), class II (RT1.B/D), and medial transplantation (RT1.C) regions of the rat MHC were determined by comparing different recombinant donors into the same recipient strain. Single region differences alone in any of these three MHC regions did not elicit full ALC. In two sets of transfers a combination of RT1.B/D and RT1.C region incompatibility was sufficient to generate a full allogeneic response. It can be concluded that the controlling element for allogeneic lymphocyte cytotoxicity is in the RT1.B/D-RT1.C region of the rat MHC.