Differential cytotoxicity of activated lymphocytes on allogeneic and xenogeneic target cells. I. Activation by tuberculin and by Staphylococcus filtrate

Normal lymphocytes activated by mitogens such as phytohemagglutinin (PHA), by Staphylococcus filtrate (SF), or lymphocytes from sensitized individuals stimulated by antigen (PPD, etc.) are cytotoxic to tissue culture cells of different origins. In this and the following paper, the results of a detailed quantitative analysis of the specificity of this cytotoxic reaction are presented. Effector cells were human or mouse lymphocytes, activated by PHA, SF, PPD, or serum factors in the culture medium. Cells from established cell lines of human, mouse, hamster, or rabbit origin, or primary human or rat embryonic fibroblasts were used as target cells. Lysis was quantitated by release of ⁵¹Cr from labeled target cells.Purified human blood lymphocytes, activated by PPD, SF, or otherwise, preferentially damaged allogeneic target cells. Lysis of xenogeneic target cells was weak or did not occur. A close correlation was noted between target cell destruction and blastoid transformation of the lymphocytes, but the slope of the regression lines of xenogeneic cytotoxicity was much smaller than that of allogeneic cytotoxicity when plotted as a function of blastoid transformation.Lymph node or spleen cells from CBA mice were stimulated by PPD to transformation and DNA synthesis. CBA lymphocytes also showed an increased degree of blast transformation in medium containing fetal calf serum or certain batches of fresh human serum. Mouse lymphocytes activated in these ways damaged allogeneic L cells but had no effects on xenogeneic Chang cells.These results indicate that lymphocytes activated by various means preferentially damage target cells from their own species. The recognition mechanisms which determine the specificity of the reactions are not known.     

Human cytotoxic T lymphocytes (CTL) against Epstein-Barr virus (EBV) infected cells: EBV specificity and involvement of major histocompatibility complex determinants in the lysis exerted by anti-EBV CTL toward HLA-compatible and allogeneic target cells

Human peripheral blood lymphocytes (PBL), from anti-Epstein-Barr virus (EBV)-seropositive donors, were stimulated by EBV and were shown to be cytotoxic toward autologous, HLA-compatible, and fully allogeneic EBV-transformed target cells. The lysis was not due to natural killer (NK) cells since the target cells used were resistant to lysis by fresh PBL and by virus-stimulated PBL-depleted of AET-SRBC-rosetting T cells (the latter being still fully cytotoxic on K562 NK-susceptible target cells). Conversely only E-rosette-purified (T) lymphocytes killed EBV-transformed HLA-compatible and allogeneic target cells. Moreover, anti-MHC antibodies inhibited the cytotoxicity exerted by EBV-induced cytotoxic T lymphocytes (CTL) on both autologous and allogeneic target cells. Finally the lysis was EBV specific since PHA blasts were not killed and since only EBV-transformed cells could compete for lysis with the EBV-positive target cells. Efficient competition was achieved by EBV-transformed cells autologous or allogeneic to the targets, even when effector and target cells were fully allogeneic. All together, the data suggest that human anti-EBV CTL may recognize nonpolymorphic HLA determinants on the target cells in association with the virus-induced antigens.     

Graft-versus-host reactivity of mouse thymocytes: effect of in vitro treatment with anti-TL serum.

The lymph ducts efferent from prefemoral nodes of sheep were cannulated and the lymph flow monitored during immune responses to injected allogeneic lymphocytes or xenogeneic murine P815 mastocytoma cells. Changes in the lymph began 5–6 days after injection of allogeneic cells but at 3–4 days after injection of xenogeneic cells, in both systems the number of large cells in the lymph increased to reach peak values of up to 40% of the total. The in vitro cytotoxic activity of lymph cells, cell supernatants, or cell free lymph was determined by measuring the release of ⁵¹Cr from prelabeled target lymphocytes or P815 cells. The cytotoxic mechanisms that were detected in the allogeneic and xenogeneic systems were similar; in both cases the lymph cells were cytotoxic only during the large cell response, and when the immunoblast numbers had returned to normal levels in the lymph no further cell-mediated cytotoxic effects were detected. During the blast response lymphocytes alone caused some target cell damage but their cytotoxic effector function was greatly increased in cultures containing complement or normal blood white cells. It was concluded that the lymph immunoblasts caused some target cell damage by direct action, but the majority of their cytotoxic activity was associated with synthesis and secretion of complement-dependent antibody (C.D.A.) and leukocyte-dependent antibody (L.D.A.).     

Cytotoxic and proliferative lymphocyte responses to allogeneic and xenogeneic antigens in vitro.

In vitro lymphoproliferative responses to foreign histocompatibility antigens are phylogenetically restricted. Responses occur most readily to allogeneic or closely related xenogeneic leucocytes, but not to unrelated xenogeneic cells. Specific cytotoxic T cell responses to foreign histocompatibility antigens show the same phylogenetic restriction. This lack of xenoreactivity is not due to a lack of precursor cells for the xenoantigens; guinea-pig lymphocytes, although normally unresponsive to mouse antigens, have a similar precursor frequency for these antigens as do lymphocytes of allogeneic mouse strains. Specific cytotoxic responses of guinea-pig lymphocytes to mouse antigens can be generated if a factor released from con A stimulated guinea-pig spleen cells is added to the culture medium. The factor produced by con A-activated spleen cells (CS) is also phylogenetically restricted in its action; CS must be obtained from animals homologous with the donor of the responding lymphocytes.     

Cytotoxic lymphocytes induced by soluble factors derived from the medium of leukocyte cultures.

Studies were undertaken to evaluate the cytotoxic capacity of human peripheral blood lymphocytes activated by either supernatants (CFM) derived from lymphocyte cultures or lymphocytes treated for 60 min at 45 degrees C. The effect of the addition of heat-treated cells on the cytotoxic activity of CFM-induced effector cells was also studied. CFM from either unmixed or mixed cultures of lymphocytes was capable of activating cytotoxic effector cells. These effector cells could kill any allogeneic target cells but failed to effect cytotoxicity on the target cells autologous to the responding cells. Both the heat-treated cells and CFM from cultures of these cells also activated lymphocytes to cytotoxic effector cells having specific receptors for nonself antigens. The question of whether heat-treated cells activate cytotoxic cells by themselves or through secreted soluble factor cannot yet be clearly answered. The findings of the present investigation suggest that expression of cytotoxicity induced in MLC is not necessarily restricted to the target cells syngeneic to the stimulator cells, but can be extended to any allogeneic target cells by the indirect effect of soluble factor secreted from stimulated cells that causes a polyclonal activation of cytotoxic precursors in the responding cell populations. The present findings also emphasize the need for caution in the use of heat-treated lymphocytes as innocent-bystander cells in MLC to provide additional cytotoxic specificities in the responder cells, since heat-treated cells alone can activate lymphocytes to cytotoxic effector cells that kill any allogeneic target cells.     

LY-2+ effectors cytotoxic for syngeneic tumor cells: generation by allogeneic stimulation and by supernatants from mixed leukocyte cultures

The present study was aimed at gaining insight into means by which stimulation of mouse spleen cells with allogeneic normal cells in mixed leukocyte cultures (MLC) can result in the generation of effector cells cytotoxic for syngeneic tumor or transformed cells. Stimulation of lymphocytes from BALB/c or C3H mice for 5 days with cells from mice of every allogeneic strain tested, in medium containing mouse serum and lacking xenogeneic serum, resulted in the activation of effectors cytotoxic for syngeneic cells transformed spontaneously or by SV40, polyoma or adenovirus. In each experiment, all of the syngeneic transformed cell lines, as well as clones derived from these lines, were lysed to the highest degree by effectors obtained from the same culture, and therefore stimulated with cells from the same allogeneic strain. Although the particular allogeneic sensitizing strain that induced the highest cytolytic activity varied between experiments, effectors obtained from the culture with the highest cell recovery always exhibited the greatest cytotoxicity against all the syngeneic transformed cells and clones. Lysis was mediated predominantly by Ly-2+ effectors; total lytic units of cytotoxicity recovered after treatment with monoclonal anti-Ly-2 antibody and complement (C) were reduced by 85 to 90% compared to cells treated with C alone. Lysis of syngeneic tumor cells by the allosensitized effectors in cytotoxicity assays was not inhibited by the addition of unlabeled "blocking" lymphocytes from the allogeneic strain used for sensitization. In addition, it was found that lymphocytes cultured without stimulating cells for 5 days in medium supplemented with supernatants from secondary MLC that are known to contain high levels of lymphokines, mediated high levels of cytotoxicity on all the transformed cells tested, but lacked detectable cytotoxic activity for syngeneic or allogeneic Con A blasts. The MLC supernatant-activated effectors that lyse the transformed cells are phenotypically CTL, because treatment with anti-Ly-2 and C reduced lytic activity by approximately 75%. Taken together, these findings suggest that the generation in MLC of Ly-2+ effector cells cytotoxic for syngeneic transformed cell lines might not be due, in some cases, to lymphocyte responses to particular alloantigens on the stimulating cells that are cross-reactive with "alien" histocompatibility antigens on transformed cells, but rather is due to effector cell activation by lymphokines produced during allogeneic stimulation.     

Augmentation of generation of human allospecific cytotoxic T lymphocyte by PPD in in vitro sensitization culture

Summary PPD augmented human lymphocyte blastogeneic response to allogeneic lymphocytes in the mixed lymphocyte reaction (MLR) and generation of human cytotoxic lymphocytes against allogeneic human lymphocytes in in vitro sensitization (IVS) culture. The augmenting effect of PPD in the MLR was unequivocally synergistic at its lower concentrations (0.05 and 0.01 g/ml). The augmentation of MLR was observed following addition of a supernatant of culture medium of lymphocytes which had been precultured with PPD for 24 h then washed free of PPD and recultured without PPD for another 24 h. PHA and Con A, in contrast, suppressed both MLR and the generation of alloreative cytotoxic cells. The alloreactive cytotoxic lymphocytes whose generation was augmented by PPD belonged to the SRBC-rosette forming fraction and passed through a nylon-wool column. The NK cell-like activities of the alloreactive cytotoxic lymphocytes were not augmented by PPD. Analysis of the alloreactive cytotoxic lymphocytes whose generation was augmented by PPD by competitive inhibition assay with unlabeled cells indicated that the same allogeneic lymphocytes used as sensitizing cells in IVS culture inhibited the cytotoxicity, while MOLT-4 cells, which are frequently used as target cells for the human NK-cell assay, did not. When lymphocytes with known HLA-A and HLA-B were used in the IVS culture and the cytotoxicity assay, PPD was found to augment the cytotoxicity only against the target lymphocytes that possessed the same HLA as the sensitizing lymphocytes in IVS.     

Effects of in vivo priming on in vitro induction of cytotoxicity. I. Non-specific augmentation by in vivo presensitization with allogeneic or xenogeneic cells.

In vivo presensitization of donor mice of responding cells with third party cellular antigens augmented in vitro generation of cytotoxic T lymphocytes in allogeneic and xenogeneic combinations. In vitro induction of detectable cytotoxicity in presensitized responding cells required the incubation period needed for in vitro primary response. However, such cytotoxic T lymphocytes were generated after in vitro stimulation with monolayers of methylcholanthrene-induced tumor cells, UV-irradiated or heated spleen cells which had proved to be effective in secondary but not in primary response. Presensitized responding cells exposed to 600R-irradiation did not augment in vitro induction of cytotoxicity in normal responding cells. The augmenting effect of presensitized responding cells may be attributable to radiosensitive T cells which are in a transitional state in differentiation from typical unprimed cells to typical primed cells.     

Continuously proliferating T killer cells specific for H-2b targets: selection and characterization.

BALB/c (H-2d) thymus-derived lymphocytes sensitized to C57BL/6 (H-2b) alloantigens have been propagated in vitro for over 9 months. These T lymphocytes are specifically cytotoxic to H-2b target cells but are stimulated to proliferate by both H-2b and H-2k spleen cells. This indicates that for these selected cells the antigen requirements for cell proliferation are different from those for cell-mediated cytotoxicity. If not continuously stimulated with allogeneic spleen cells, the cytotoxic cultures fail to divide and rapidly lose their cytotoxic activity. Allogeneic erythrocytes do not stimulate cell proliferation in "quiescent" cell cultures and allogeneic tumor cells do so only in the presence of spleen cells. However, "quiescent" cell cultures display cytotoxicity in the presence of phytohemagglutinin A as do cell cultures which have lost their cytotoxic activity although they proliferate upon allogeneic stimulation. The significance of these findings is discussed.     

Effects of sodium periodate modification of lymphocytes on the sensitization and lytic phases of T cell-mediated lympholysis.

Sensitization of mouse splenic lymphocytes in vitro with sodium borohydride, suggesting that the biologic effects of sodium periodate are-treated autologous spleen cells stimulated a one-way mixed lymphocyte reaction and led to the generation of thymus-derived cytotoxic effector cells. These effectors were capable of lysing in 4 hr periodate-treated syngeneic and, to a lesser extent, periodate-treated allogeneic target cells. These results suggest that sensitization by periodate-treated autologous cells could result either from a specific reaction to modified self components or from a nonspecific mitogenic stimulation. Effector cells generated by allogeneic sensitization were detected on periodate-modified targets, irrespective of the H-2 antigens expressed by the targets. The effects of periodate modification on both stimulator and target cells were reversible by sodium periodate are dependent on the formation of a free aldehyde group on cell surface glycoproteins. Pretreatment of stimulator cells with neuroaminidase prevented the effect of periodate treatment, suggesting that the sensitization involves oxidized sialic acid residues. During the 4-hour 51Cr-release assay periodate-treated targets could be used to detect cytotoxic effector cells of any specificity. Fresh spleen cells and lymphocytes cultured for 5 days without antigen or in the presence of lipopolysaccharide did not lyse periodate-treated targets. An increasing level of cytotoxicity was detected on periodate-treated targets when the effector cells were generated, respectively, by stimulation with concanavalin A, by sensitization with periodate-modified autologous cells. Although the lysis of periodate-treated targets is itself nonspecific, effector cell specificity could be determined by selective blocking of the lytic phase with cells syngeneic to the stimulators. These results indicate that a nonspecific interaction can occur between lymphocytes and periodate-treated target cells, but that this interaction leads to lysis only when the lymphocytes were activated to become cytotoxic effectors.     

Production of alpha-lymphotoxin by human T-cell subsets

Human T cells were isolated from peripheral blood lymphocytes (PBL) and sensitized to allogeneic PBL in a one-way mixed-lymphocyte culture. These sensitized T cells were fractionated on the basis of their possession of Fc receptors for IgG (TG+) or IgM (TM+), or the absence of both IgG and IgM receptors (TG-M-). When restimulated with alloantigen of the same derivation, TG+, TM+, and TG-M- cells yielded almost equal amounts of cytotoxin. Anti-alpha-lymphotoxin serum neutralized most of this cytotoxic activity indicating that alpha-lymphotoxin (alpha-LT) constituted most of this activity. Although TG-M- cells function as effectors in allogeneic cytotoxicity, TG+ cells lyse IgG-coated targets in an antibody-dependent cell-mediated cytotoxic (ADCC) reaction, which has been shown to be mediated in part by alpha-LT. Whether TM+ cells can be cytotoxic is not clear. In addition, freshly isolated human T-cell subsets were stimulated with phytohemagglutinin-P (PHA-P). After PHA stimulation, TG+, TM+, and TG-M- cells produced similar amounts of soluble cytotoxin, which was largely neutralized by anti-alpha-LT. The TG+ cells incorporated less thymidine than the TM+ or TG-M- cells. Likewise, OKT4+ and OKT8+ subsets, isolated with the aid of monoclonal OKT8 or OKT4 antibody and complement, yielded lymphotoxin after stimulation with PHA. It is shown that all T-cell subsets, as defined here, can produce lymphotoxin. Furthermore, depending on the assay system, cytotoxicity can be clearly demonstrated in all of these subsets, except in TM+ cells, where positive and negative results have been reported.     

Alloantibody-induced cytotoxicity of macrophages.

Alloantibodies substantially alter the in vitro nonspecific cytotoxic effect of macrophages on bystanding allogeneic and xenogeneic target cells. Although the specific IgM alloantibody increased significantly the ability to parental macrophages to damage target cells, IgG alloantibody had an opposite effect and suppressed the macrophage cytotoxicity. Macrophages of F1 hybrid mice were less affected by this treatment unless alloantibodies against both parental strains were added together. Parental macrophages exposed in vitro to the concomitant action of both IgM and IgG alloantibodies exhibited a diminished cytotoxicity toward target cells. It is proposed that IgM and IgG alloantibodies induce different conformational changes of the macrophage surface.     

Guinea pigs immunized with xenogeneic cells: skin test, in vitro stimulation; cytotoxicity of lymph node cells

Guinea pigs were immunized with cells of a human lymphoblastoid line. While nonsensitized lymph node cells were not influenced, those derived from immunized animals were markedly stimulated in vitro by the human cells. Maximal activity was found with lymphocytes harvested on day 6. The response was paralleled with the strength of skin induration elicited by the human cells and with the efficiency of the cytotoxic effect of lymphocytes on human target cells. The stimulation and cytotoxicity was specific while the skin test demonstrated cross reactivity with mouse cells.     

Response of human fetal lymphocytes in xenogeneic mixed leukocyte culture: Phylogenetic and ontogenetic aspects

Cells from liver, thymus, and spleen of human fetuses at different stages of development were capable of a proliferation response against xenogeneic and allogeneic lymphocytes. The kinetics of fetal responses against rat lymphocytes were identical to those of fetal and adult responses against allogeneic cells. With all of the cell types studied, including adult lymphocytes, allogeneic responses were stronger than xenogeneic. Xenogeneic responses against lymphocytes from rat, mouse, or sheep were stronger than those against lymphocytes from rabbit, chicken, snake, or frog. These results are interpreted to indicate that recognition of foreign lymphocytes by human lymphocytes depends on the phylogenetic position of the species used as a source of stimulating cells. The degree of recognition decreases as the phylogenetic distance increases. Specific elimination of responding cells and restimulation with another cell population was used to study the specificity of proliferation responses against mouse and rat lymphocytes. Responses by prethymic liver cells from human fetuses were not due to the existence of specifically recognizing subpopulations. Thymus and spleen at 16 weeks' gestation contained specific subpopulations capable of differentiating between xenogeneic and allogeneic cells, as well as between xenogeneic cells with different intraspecies histocompatibility patterns. Generation of receptor diversity on T lymphocytes is discussed briefly in the light of these findings.     

Enhancement of the functional activities of human T cells after their interaction with SRBC

We have examined the functional consequences of the exposure of human lymphocytes to sheep red blood cells (SRBC). Peripheral blood lymphocytes (PBL) were incubated wih SRBC under optimal conditions for their interaction and, after lysis of the erythrocytes, the unfractionated PBL were examined for several T cell functions. Only after exposure to SRBC, but not after incubation with mouse, ox, chicken, or human erythrocytes, the unfractionated PBL showed an increased reactivity in the following functions: 1) Production of T cell growth factor, after PHA stimulation; 2) mitogenic response to suboptimal doses of PHA and Con A, and 3) response in mixed lymphocyte culture. Other functional activities, such as natural cytotoxicity (NK) and antibody-dependent cellular cytotoxicity (ADCC) were also enhanced by the interaction of PBL with SRBC, but the increases in cytotoxic activities were not consistently detected. Taken together, these results indicate that the interaction of PBL with SRBC has functional consequences in the reactivity of T cells producing an enhancement of several in vitro T cell functions.     

Lymphocyte cytotoxicity to influenza virus-infected cells. II. Requirement for antibody and non-T lymphocytes.

Peripheral blood lymphocytes (PBL) obtained from humans were cytotoxic for influenza virus-infected target cells. The PBL were shown to have associated influenza virus anti-hemagglutinin antibody (AHAb) detectable only by radioimmunoassay. This antibody could be removed by incubating PBL at 37 degrees C for 30 min. The lymphocyte population that was effective in this system was nonadherent and nonphagocytic cells. PBL gave comparable levels of cytotoxicity when tested by using either a xenogeneic or allogeneic virus-infected target cell. These results indicate that lymphocyte cytotoxicity to influenza virus infected cells may be mediated by small quantities of antibody and by lymphocytes that possess characteristics of K cells. No evidence for T cell-mediated cytolysis was found with this xenogeneic system.     

Nonspecific cytotoxic effects of antigen-transformed lymphocytes. III. Relationship to specific cytotoxicity.

Human peripheral blood lymphocytes, cultured either with PPD or in one-way mixed leucocyte reactions, develop a nonspecific cytotoxic potential which can be expressed on autologous as well as on unrelated xenogeneic target cells. This nonspecific cytotoxicity develops at an early stage in the mixed leucocyte reaction, before specific cytotoxic effects are demonstrable, and decreases as specific cytotoxicity increases.     

Characteristics of natural killer cells (NK cells) and features of the cytotoxic test in Syrian hamsters

The cytotoxic test in vitro with the use of xenogeneic target cells of human myeloma, strain K-562, labeled with 51Cr has demonstrated natural cytotoxicity of lymphoid cells from noninbred Syrian hamsters. This cytotoxicity occurs at the cost of non-adherent splenocytes. NK may be isolated over the gradient density of ficoll (1.078), selective for large granular lymphocytes. To detect the maximal lytic activity of NK from Syrian hamsters in the cytotoxic test in vitro, they should be brought into 10-12 hour contact with sensitive target cells K-562. In Syrian hamsters, the highest natural cytotoxicity is shown by the cells of the blood and spleen. In the bone marrow and thymus, it is little pronounced and is virtually absent from the peripheral lymph nodes.     

Stimulation with autologous lymphoblastoid cell lines: lysis of Epstein-Barr virus-positive and -negative cell lines by two phenotypically distinguishable effector cell populations

Human lymphocytes, stimulated in vitro for 6 days with x-irradiated or glutaraldehyde-treated autologous Epstein-Barr (EB) virus-transformed lymphoblastoid cell lines (LCL), are cytotoxic for autologous and allogeneic EB+ LCLs as well as for several EB- cell lines that are also susceptible to lysis by interferon-activated natural killer (NK) cells. To determine whether the apparent nonspecific lysis mediated by LCL-stimulated cells is due to a mixture of effector cells directed against different target cells, advantage was taken of our recent finding that monoclonal antibody OKT8 reacts with human cytotoxic T lymphocytes but not with NK cells or NK-like cells generated in mixed leukocyte cultures. The depletion of OKT8+ cells from LCL-stimulated cultures by treatment with OKT8 and complement abolished or markedly depleted cytotoxicity against all EB+ target cells tested, whereas cytotoxicity against EB-, NK-sensitive cell lines including K562, MOLT-4 and HSB-2 was not or only minimally reduced. These results indicate that stimulation with autologous LCL results in the generation of OKT8+ cytotoxic T lymphocytes that lyse EB virus-transformed LCL and OKT8- NK-like cells that lyse EB-, NK-sensitive cells.     

Functional heterogeneity among herpes simplex virus-specific human CD4+ T cells.

One hundred thirteen HSV-specific CD4+ T cell clones were established from the PBL of a healthy person and their functional heterogeneity was investigated. All clones proliferated in response to stimulation with HSV in the presence of autologous APC. Among those, 48 clones showed cytotoxic activity to HSV-infected autologous EBV-transformed lymphoblastoid cell line, but not to HSV-infected autologous fibroblasts, HSV-infected allogeneic cells, or K562 cells (group 1). Five clones showed cytotoxicity against HSV-infected autologous cells as well as HSV-infected allogeneic cells and K562 cells (group 2). The cytotoxicity of these clones was found to be mediated by the direct killing but not by the "innocent bystander" killing of target cells. Sixty clones showed no cytotoxic activity, however, among these, 23 revealed HLA-unrestricted and nonspecific cytotoxicity in the presence of PHA in culture (group 3), and the remaining 37 did not show any cytotoxic activity even in the presence of PHA (group 4). The cytotoxic patterns of these clones did not change in activated and resting phases, suggesting that the difference in cytotoxic ability does not depend on cell cycles. The cytotoxic activity of group 1 was inhibited by addition of anti-HLA-DR or anti-CD3 mAb to the culture, whereas these mAb had no effect on the cytotoxicity of group 2. All four groups of clones had helper activity for anti-HSV antibody production by autologous B cells. Moreover it was found that all groups of clones simultaneously produced IL-2, IL-4, and IFN-gamma after culture with APC followed by HSV Ag stimulation. The surface phenotype of all clones was uniformly CD2+, CD3+, CD4+, CD8-, CD29+, CD45RA-, but expression of Leu 8 was varied. These data therefore indicate that HSV-specific human CD4+ T cells are classified into at least four groups according to the presence and specificity of cytotoxicity, i.e., Th cells with HSV-specific and HLA-class II-restricted cytotoxicity, Th cells with HLA-unrestricted and nonspecific cytotoxicity, Th cells with lectin-dependent cytotoxicity, and Th cells without cytotoxic activity. The present finding of functional heterogeneity among virus-specific human CD4+ T cells might shed light on the pathogenesis of CD4+ T cell immunodeficiency, such as human retrovirus infections.