Biotherapy 5: 71-81, 1992. (~) 1992 Kluwer Academic Publishers. Printed in the Netherlands.

Cytolytic T lymphocytes: an overview of their characteristics David W. Lancki & Frank W. Fitch Department of Pathology, the Ben May Institute, and the Committee on Immunology, The University of Chicago, Chicago, Illinois 60637, U.S.A. Received 13 September 1991; accepted4 October 1991

Key words: Anti-TCR mAb, Cytotoxic T lymphocytes, mAb coated target cells, mechanisms of lysis, T cell clones Abstract

Cloned T cells have been useful for assessing the lytic potential of distinct T cell subsets and for determining the relative contribution of different effector mechanism involved in the lytic process. Alloreactive CD8 + murine T cell clones and cloned murine CD4 ÷ TH1 and TH2 T cells reactive with nominal antigen (ovalbumin) lysed nucleated target cells bearing antigen or coated with anti-CD3 monoclonal antibody in a short term 51Cr-release assay. These clones were also evaluated for their ability to lyse efficiently sheep erythrocyte (SRBC) target cells coated with anti-CD3 mAb by a mechanism (presumably involving membrane damage) that does not involve nuclear degradation. Three patterns of lysis were observed: CD8 + and some CD4 + TH2 effector cells lysed efficiently nucleated target cells and anucleated SRBC coated with anti-CD3 mAb. However, CD4 + TIal (and a few TH2 ) T cells which lysed nucleated target cells bearing antigen or coated with anti-CD3 mAb did not lyse efficiently the SRBC coated with anti-CD3 mAb. One CD4 beating Tn2 cell failed to lyse efficiently either nucleated target cells or anucleated SRBC coated with anti-CD3 mAb. These results indicate that both TH1 and TH2 clones have lytic capabilities. Furthermore, they suggest that some but not all TH2 murine T cell clones have lytic characteristics similar to those of conventional CD8 + CTL. However, it is not certain how these patterns of lysis of target cells in vitro relates to the capacity of CTL to lyse such target cells in vivo.

Abbreviations: ADCC: antibody-dependent cell-mediated cytotoxicity; APC: antigen presenting cell; CTL: cytotoxic T lymphocyte; FcR: Fc receptor for IgG; HTL: helper T lymphocyte; MHC: major histocompatibility complex; MLR: mixed lymphocyte reaction; LFA-I: Lymphocyte Function antigen1; LT: lymphotoxin PFP: pore-forming protein; SRBC: sheep red blood cell; TCR: T cell receptor for antigen; TNP: trinitrophenol.

Introduction

Thymus-derived T lymphocytes sensitized in vivo or in vitro have the potential to injure target cells bearing the sensitizing antigen (reviewed in [1]). These cytotoxic T lymphocytes (CTL) form a major part of the cell-mediated immune response that enables the host to survive encounters with infectious organisms, oncogenic agents, and other foreign challenges. Lysis mediated by

CTL can be quantified readily by measuring the release of 5~Cr-radiolabelled macromolecules from target cells [1, 2]. In general, this system provides a sensitive and accurate means for determining the extent of target cell injury and allows the relative lytic capability of different cell populations to be compared easily. However, this measure of lysis may provide a restricted view of cytotoxicity, which can almost certainly oversimplify the processes involved. This can

72 lead to the interpretation that "lysis is lysis". It is not certain how the capacity of a CTL effector to lyse a target cell in vitro as measured by the release of radiolabelled macromolecules relates to the capacity of CTL to kill such target cells in vivo. Several mechanisms have been proposed to explain cytotoxicity, and potential mediators of lysis have been identified (reviewed in [3, 4]). However, none of the potential mediators alone appears to explain adequately the collective observations on the lytic properties of CTL effector populations. As a result, several investigators have proposed that lysis may involve multiple mechanisms [5-8]. If indeed multiple effector mechanisms are involved in lysis, it seems likely that not all of these mechanisms will be expressed or utilized by each effector cell. Lytic activity assessed using heterogeneous populations of CTL may not permit identification of the relative contributions of a particular lytic mechanism. Perceptions of the properties of CTL are based to a large extent on the studies done with heterogeneous populations of alloreactive (predominantly CD8+) effector cells generated during stimulation in unidirectional mixed lymphocyte reaction (MLR) in vitro, or by priming in vivo with tumor cells bearing alloantigens. The use of these techniques to generate reproducible effector cell populations for the study of CTL properties facilitated the analysis of the effector and target cell interactions. However, as pointed out above, these populations have limitations in assessing the relative contributions of different cytotoxic effector mechanisms. An alternative approach to the analysis of cytoto:~ic effector mechanisms has relied on the derivation and use of clonal populations of CTL. Homogeneous clonal populations of T lymphocytes can be assessed for functional activities and these activities correlated with other characteristics. The participation of distinct subsets of the T lymphocytes in the execution of lytic function has been documented by investigators using human [9, 10] and murine [11, 12] systems. The development of clonal populations of effector cells has allowed a more detailed analysis of the capacity of members of the different T cell subsets to perform cytotoxic functions.

Some factors influencing the assessment of cytotoxic activity of T cell subsets

Subpopulations of T lymphocytes have been defined by the expression of subset-specific cell surface molecules and by the demonstration of functional differences. CD8 and CD4 cell surface molecules are expressed on mutually exclusive subsets of mature peripheral T lymphocytes. The expression of these cell surface markers correlates well with the class of the major histocompatibility (MHC) antigen recognized by the T cell. Thus, CD4 + T cells generally recognize antigen plus class II MHC molecules, while CD8 + T cells generally recognize antigen plus class I MHC molecules. However, exceptions to this correlation have been observed [13-15]. A functional correlation has also been observed among these subsets. CD4 + cell functions have been predominantly attributed to the particular array of lymphokines (including B cell and/or T cell growth factors) that are produced by those cells. In contrast, cytolytic function has been primarily attributed to the CD8 + subset of T cells. There also are exceptions to these functional correlations. CD4 + T cells can also cytotoxic [16-18], and some CD8 + T cells may produce arrays of lymphokines similar to those produced by helper T lymphocytes (HTL) type I (TH1) [19, 20 I. Recognition of the lytic potential of many CD4 + T lymphocytes may depend on several factors relating to the nature of antigen recognized by such T cells. The methods currently used to measure cytotoxic ability (particularly in a short term assay) require efficient interaction between the effector cells and a high proportion of the target cells. CD8 + effector cells apparently recognize nominal antigens that are processed and presented in the context of class I MHC, but class I MHC molecules are more widely expressed in different tissues than are class II MHC molecules. The restricted recognition by most CD4 + cells of nominal antigen in the context of the class II MHC molecule places special constraints on the capacity to demonstrate cytotoxicity. In order for target cells to be recognized efficiently, they must express a sufficient quantity of class II MHC antigen. The array of cells (macrophage, dendritic cells, B cells, etc.) that

73 express sufficient levels of class II MHC antigen is relatively limited. The minimum amount of class II M H C molecule that is needed to functionally activate the cells has not been carefully quantified and may vary depending on the characteristics of the T cell receptor (TCR) of effector cells. Also, the target cell must be able to present a sufficient amount of the relevant nominal antigen in the context of the class II M H C molecules. For this to occur, the target cell must also be capable of properly "processing" the particular antigen for presentation on the cell surface. Target cell characteristics that might affect the efficient processing of antigen are not well understood. Homogeneous populations of generally efficient class II MHC-bearing APC that can be readily radiolabelled for use in standard cytotoxicity assays have been derived [21, 22]; these cells provide sensitive target cells for assessing lytic potential. In addition, target cells must express appropriate ligands for any accessary molecules that may be needed to stabilize interactions between effectot and target cells and/or to activate the cytotoxic function of effector cells. Conjugate formation appears to precede recognition events in CTL-target cell interactions [23, 24]. LFA-1 ( C D l l b / C D 4 9 ) , LFA-2 (CD2) which are expressed on lymphocytes, and ICAM-1 (CD54), and LFA-3 (CD58), which are expressed on target cells, represent adhesion molecules involved in interactions between immune cells and their targets or antigen-presenting cells (APC). Other structures, including CD8 and CD4 may contribute to cell adhesion between target cells bearing class I or class II MHC molecules respectively. However, these structures may also be involved in the regulation of signalling through the T C R [25, 26]. A similar role has been postulated for CD2 in human T cells [27, 28]. The restrictions described above have until recently limited the availability of sensitive and accurate assays for measuring the lytic activity of CD4 ÷ against target cells bearing class II MHC antigens. Many of the standard target cells used in CTL assays, such as P-815 mastocytoma cells and EL-4 lymphoma cells, bear only class I M H C molecules. In addition, the assay conditions for measuring some forms of CD4 ÷ cellmediated killing vary significantly [24, 29-31].

Cytotoxic assays involving long term incubation of T cells with the target cells (12-24h) may involve effector mechanisms (such as the secretion of cytokines synthesized de novo) that are distinct from those used by CD8 ÷ CTL in short term assays (3.5 - 6 h). Although the recognitive events associated with CTL-target cell interaction are mediated through the clonally distributed TCR, the precise mechanism by which the lethal hit is triggered is not known. In both human and mouse, the predominant TCR structure contains an a//3 heterodimer, which contains variable regions having a similar organization but which are distinct from those of Ig molecules. Cells expressing an alternative, less variable TCR structure, the 3,/3 heterodimer, have also been shown to function as lytic effector cells. The specificity of antigen recognition is determined by these highly polymorphic oc//3 or 3,/3 heterodimeric components of the T cell receptor. These components of the TCR are physically associated on the cell surface with the multimolecular CD3 complex which appears to be the primary structure involved in transmembrane signalling. The involvement of the T C R / C D 3 complex in triggering the lethal hit was first documented by studies showing that soluble anti-TCR/CD3 monoclonal antibodies inhibited lysis [32, 33]. The observation that immobilized anti-TCR/CD3 mAb activated other T cell functions including secretion of lymphokines led to the development of the "redirected lysis" assay in which target cells are coated with a monoclonal antibody which reacts with the TCR of the effector cell. Interactions of such antibody-coated target cells with CTL induces efficient lysis [34-38]. Since there appears to be no polymorphism in the CD3-e component of the TCR, anti-CD3-e mAb should react with the same affinity with the TCR of all CD3 ÷ T cells, thus avoiding the effects of possible differences in affinity of the TCR. This assay appears to provide an objective assessment of the cytotoxic activity of all T cells. Thus, monoclonal antibodies against the TCR have been used as a sensitive and convenient surrogate for antigen in the analysis of the lytic capacity of T cells. How well the anti-TCR/CD3 mAb mimic antigen in the stimulation of T cells is still controversial. There is growing evidence

74 that the activation signals provided through stimulation of the TCR alone may not accurately mimic the activation events associated with antigen stimulation [39]. Other factors may contribute to the differences observed in the capacity of CD4 + effector cells to lyse antigen bearing APC. Culture conditions may influence the properties of effector cells which are derived [40-42]. This has been demonstrated in the recent experiments performed by investigators in this laboratory; those findings suggest that the presence of IFN-3, during culture can profoundly influence the characteristics of murine HTL clones that are selected [41]. Differences in the activation requirements have been reported for normal T cells, and those primed in vitro [43]; and differences in characteristics relating to lysis have been reported for T lymphocytes derived following priming in vivo, as compared to those derived by priming in vitro [40, 42, 44]. For example, mice primed with antigen in vivo yielded a higher frequency of clones that were resistant to inhibition by antiCD8 mAb [40]. Also, peritoneal exudate lymphocytes (PEL) from mice primed in vivo lack appreciable expression of cytotoxic granules [44] or serine esterase activity or pefforin associated with these granules [42]. In addition, culture of T cells in vitro has been shown in some cases to modify the properties and functional capabilities of the cultured cells [45-47]. The extent to which these conditions influence the functional properties of CTL remains to be determined.

Properties of lysis by cytotoxic T lymphocytes of different subsets Although the cytotoxic activity of CD8 + T cells has been recognized for many years, the precise mechanisms by which lethal injury occurs has eluded definitive characterization (reviewed in [4, 48]). Some of the lytic activity of CD8 + (and CD4 ÷) T cells has been reported to involve the synthesis de novo of soluble factors (such as conventional lymphotoxin (LT) activity) [49]. However, lysis by CD8 + T cells appears to be largely independent of synthesis of lytic factors de novo. Lysis of target cells by many CD8 ~ T ceils appears to involve molecules (some of

which are capable of inducing membrane damage) that are present in intracytoplasmic granules. As described above, however, some highly lytic CD8 ÷ T cells appear to lack expression of these cytotoxic granules and their associated molecules. The controversy regarding the lytic potential of subsets of CD4 + T cells may be due in part to our lack of understanding of the lytic mechanisms and molecules that may be involved in lysis. CD4 + T cells having properties of the TH1 cells have been shown to lyse target cells bearing antigen or targeted for lysis with anti-CD3 mAb [37, 50], but CD4 + cells having properties of the TH2 have been reported not to lyse antigen-bearing APC [31, 50]. In an effort to better understand the basis for the differences that have been described in the lytic activity of T cell from distinct subsets we tested the capacity of several CD4 + T cell clones derived in our laboratory for their capacity to lyse in assays designed to assess 1) the general lytic potential of these clones against nucleated target cells that were coated with anti-CD3~ mAb, and 2) the capacity of these clones to lyse anucleated SRBC coated with the anti-CD3 mAb to assess their capacity to cause membrane damage in target cells. Lysis of nucleated target cells by T cell clones having properties of distinct T cell subsets

We have assessed the lytic potential of murine T lymphocytes derived in this laboratory that have properties of either the TH1 or of the TH2 subset. These clones are reactive with the nominal antigen ovalbumin in the context of I-A k. Lytic potential of these clones was tested against target cells generated in one of two ways: in some experiments, target cells were coated with anti-CD3E mAb (via the cell surface FcR on P-815 mastocytoma cells) or by pulsing LK35.2 APC with ovalbumin. The results presented in Figure 1 show that each of the TH1 clones derived from two separate cloning experiments (A04 or A05) could lyse P-815 target cells with varying degrees of efficiency in a short term (5 h) 5~Cr-release cytotoxicity assay. In addition, the results presented in Figure 1 show that each of the cells having properties of TH2 (producing IL-4, but not IL-2 or IFN-y) or Tnl.5 (cells

75

LK Production IL-2

IL-4

IFN-~'

CLONE

-t-

-

-I-

q-

,.

if-

==

q-

,=.

A 0 4 / M 14

--

"t-

-.,

/M23

_

+

_

/C15

.=

-~-

...

/C21

==

-!-

...

A05/M14

==

-I"

...

/M21

==

-~-

,,,,

/C13

_

+

_

/c22

_

-}-

-}-

==

-I-

"}"

/M5

--

-l-

"}"

/M8

--

-t-

d-

--

-I-

-I-

II ANTI-CD3

!_2

0CM

ACM1

A04/M4 IIIII _1

A05/M12 /M23 I 80

0

% LYSIS OF P-815 CELLS COATED WITH ANTI-CD3 mAb Fig. I. CD4 + clones having properties of either THI (producing IL-2 and IFN-y), TH2 cells (producing IL-4 and not IL-2 or IFN-y), or TH1.5 (producing IL-4 and IFN-y) can lyse target cells coated with anti-CD3 mAb. Clonal populations of T cells (following sedimentation on a ficoll-hypaque) were combined in V-bottom microtiter plates (Linbro) with the 51Cr labeled FcR bearing P-815 mastocytoma target cells (2.5 × 103) at an effector-to-target cell ratio of 20 : 1 in the absence or presence of culture medium (CM) or anti-CD3 mAb containing hybridoma supernatants (1/8 dilution) in a total volume of 200/xl. Cultures were centrifuged at 500 X g for 5 minutes and allowed to incubate for 5 hours. At that time 100 p.1 of the reaction supernatant was collected and the percent specific lysis determined in comparison with maximum release (freeze/thaw) using the formula: (Test Release - Spontaneous Release) % Specific Lysis = (Maximum Release - Spontaneous Release) × 100

p r o d u c i n g IL-4 and I F N - 7 ) derived f r o m the s a m e cloning e x p e r i m e n t s as the TH1 cells were also able to lyse efficiently the P-815 cells in a redirected lytic assay. In the absence of anti-CD3 m A b , no appreciable lysis of these target cells was observed. T h e capacity o f cells f r o m b o t h H T L subsets to lyse target cells does not a p p e a r to be an artifact o f the assay using P-815 cells c o a t e d with anti-CD3 m A b , since in a separate

e x p e r i m e n t we d e m o n s t r a t e d that TH1 and T n 2 cells were b o t h able to lyse target cells bearing t h e relevant antigen (Table 1) and cells targeted with the a n t i - C D 3 m A b . H o w e v e r , not all TH2 cells were able to lyse target cells. A s described a b o v e , s o m e TH2 clones have b e e n r e p o r t e d to be u n a b l e to lyse target cells c o a t e d with antiC D 3 m A b [51]. We have b e e n unable to induce lysis of nucleated target cells using the TH2 clone

76 Table I. CD4 ÷ T cell clones can lyse target cells bearing antigen or coated with anti-CD3 mAb SUBSET (SPECIFICITY)

CLONE

T~ i (I-Ak/OVA)

GL2 GL3 GL10 L9 L11 L12

TH2 (I-Ak/OVA)

LK35.2 2 0 1 0 0 0

% SPECIFIC LYSIS OF TARGET CELL:a LK35.2OVA P-815 50 44 52 33 49 58

2 1 1 4 2 4

P-815/aCD3 59 51 53 54 45 55

aCD4+ clones having properties of either TH1 (producing IL-2 and IFN-7), or TH2 cells (producing IL-4 and not IL-2 or IFN-y) were combined in V-bottom microtiter plates (Linbro) with the SlCr labeled FcR bearing P-815 mastocytoma target cells (2.5 × 103) at an effector-to-target cell ratio of 20:1 in the absence or presence of anti-CD3 mAb containing hybridoma supernatants (1/8 dilution) in a total volume of 200 p,l. Cultures were centrifuged at 600Xg for 5 minutes and allowed to incubate for 5 hours. D10 (provided by Dr. C. Janeway) (Figure 2). It is not clear if such clones represent an additional distinct subset of H T L . Ability o f T cells to cause membrane damage in target cells. Evidence that discrete membrane lesions occur in the target cell membrane is based on the inverse relationship between the effective molecular size of intracellular components being released and the time of release following injury [52]. This led to the proposal that cell death resulted from "colloid osmotic" forces resulting from the influx of water through the lesions. Studies using fura-2 imaging, suggest that antigen-specific lysis by cloned CTL may involve transmembrane fluxes leading to increases of calcium to millimolar levels in target cells [53]. Electron microscopy provided evidence for the formation of "pores" in erythrocyte ghosts following lysis by lymphocytes in an A D C C reaction [54]. These findings were extended to cloned CTL cells [55]. Two types of lesions were found, similar to those caused by complement but with larger internal diameters. Evidence that such lesions can be induced directly in cell membranes has been provided in experiments using non-nucleated cells. Erythrocytes which lack nuclei can be lysed by conventional CTL when coated with a n t i - T C R / C D 3 mAb [8, 56]. Collectively, these and other data suggest that direct m e m b r a n e damage can be induced as a result of the interaction between at least some cytotoxic lymphocytes and target cells during cellmediated cytotoxicity. It is not clear, however, if

this is an obligate element in the lytic process of these effector cells. Our initial studies using anucleated SRBC coated with anti-CD3 mAb indicated that although CD8 + C T L clones could indeed lyse SRBC targeted in this way [8], a number of CD4 + clones having characteristics of the TH1 subset [57] were unable to lyse the SRBC targets efficiently [8]. Since that time, the lytic potential of those CD4 + clones against nucleated target cells such as P-815 mastocytoma cells coated with anti-CD3 mAb or antigen pulsed APC has been established. T o determine if the failure of some CD4 + cells to lyse SRBC coated with anti-CD3 m A b was a general property of CD4 + cells that have cytotoxic potential or if this represented a property of a subset of the CD4 + cells, we examined the capacity of several other CD4 + cells having properties of either T u l or TH2 cells [41] to lyse anucleated target cells using a heteroconjugated anti-DNP(and TNP)/anti-CD3 antibody preparation, (kindly provided by Dr. David Segal) to coat TNP-treated SRBC. This appears to be an efficient method of immobilizing anti-CD3 mAb on the SRBC targets. This approach resulted in identification of two patterns of lytic behavior against SRBC coated with the anti-CD3 m A b [58]: 1) SRBC coated with anti-CD3 mAb were efficiently lysed by our CD8 + C T L clones (B18 in Figure 2), and 2) a relatively inefficient pattern of lysis was observed with T u l cells (GL18 and PGL2, also shown in Figure 2). Interestingly, some TH2 cells were able to lyse efficiently the SRBC targets (L16 and PL3), while other TH2 cells (D10) could not. A m o n g those cells that did not efficiently lyse

77 110 P-815

I// ///

/// /// ///

,--I --I

7A ///

l.U

/// /..// /// /// /// /// /// /// /// /// i// V-- I//

0 n-" B18

U..

GL18

PGL2

L16

PL3

D10

011o ~ ~11

[ ] CM + CLONEDCELLS • ANTI-CD3+ CD8+ CTL [ ] ANTI-CD3+ CD4+ TH1 [ ] ANTI-CD3+ CD4+ TH2

anucleated SRBC targets among the subsets of H T L cells. The T cell growth factor IL-2 has been shown to induce expression of hemolytic granules, perforin activity and serine esterase activity during culture in vitro [45]. In the present studies, the variation in hemolytic potential (and B L T serine esterase levels, data not shown) cannot be accounted for simply on the basis of exposure to IL-2, since clones having either high or low lytic activity against SRBC were maintained under identical conditions. The molecular basis for the observed differences in the capacity to lyse efficiently SRBC coated with anti-CD3 m A b is currently under investigation; these differences may involve variability in the expression of gene products reported to have hemolytic activity as described below.

Molecules implicated in CTL-mediated membrane damage in target cells

0

B18

GL18

PGL2

L16

PL3

D10

CLONE

Fig. 2. T cell mediated lysis of SRBC coated with anti-CD3 mAb. T cell clones having properties of CD8 ÷ CTL (B18), or CD4 ÷ clones having properties of either TH1 (GL18 and PGL2), or TH2 (LI6, PL3, and D10) were tested for their capacity to lyse TNP-derivitized, 51Crlabeled SRBC (2 x 10~ cells) coated with anti-CD3/anti-TNP heteroconjugated antibody at a saturating concentration of 0.3/xg/ml. The 51Cr labeled FcR bearing P-815 mastocytomatarget cells (5 × 103) were incubated with effector cells either in the absence or in the presence of anti-CD3 mAb containing hybridoma supernatants (1/8 dilution). The effector and target cells were combined in a V-bottom microtiter plate as described above at an effector-to-target cell ratio of 25 : 1 and incubated for 6 hours. The % specific lysis was determined as described above. S R B C were T cell clones (GL18 and PGL2) that could lyse antigen bearing target cells or certain nucleated cells that were coated with anti-CD3 mAb. Also among the cells that did not efficiently lyse the SRBC was one TH2 clone (D10) which did not lyse nucleated target cells coated with anti-CD3 mAb (shown in Figure 2). These results indicate that there is variability in the expression of an efficient capacity to lyse

The biochemical analysis of the molecules that may be involved in the formation of membrane lesions has become possible through the development of clonal populations of lymphocytes. A characteristic feature of many CTL and NK cells is the presence of large electron-dense granules within the cytoplasm. Isolation of these granules from effector cells results in an enrichment for hemolytic and serine esterase activities. These and other observations led to the proposal that these granules may be responsible for the lesions on the target cell membrane [59]. Molecules having hemolytic activity, referred to as cytolysin, perforin, or lymphocyte PFP when isolated from C T L or N K cells appear to be very similar if not identical. Proteolytic enzymes have also been implicated in the lytic activity of CTL based on inhibition of lytic activity by various protease inhibitors [60]. The role of these enzymes in lysis is not clear; however, they may serve to process granule-related molecules to generate active lytic structures. Stimuli that trigger lysis in CTL also trigger the exocytosis of granule associated serine esterase activity [8, 61]. Based on these observations a model for CTL granule exocytosis triggered by T C R cross-linking involving the activation of protein kinase C and increased intracellular levels of calcium [61]. The involvement of calcium in the lytic process was further

78 indicated through studies showing increased levels of free intracellular calcium due to release of intracellular stores and the influx of extracellular calcium.

Mechanism(s) of lysis that appear to be independent of the perforin/exocytosis mechanism Although the release of pefforin and other lysisrelated molecules by granule exocytosis may provide one mechanism by which CTL can damage target cells, several lines of evidence suggest that CTL may have other mechanisms of lysis. First, as described above, certain populations of alloreactive, highly lytic peritoneal exudate CTL stimulated in vivo appear to lack appreciable expression of cytotoxic granule, hemolytic activity (perforin), or serine esterase activity [42, 44]. It is not clear if these cells primed in vivo represent a subset of CTL distinct from those obtained by sensitizing in vitro; the basis for the differences that do exist in cells derived using different conditions has yet to be determined. Secondly, recent studies indicate that there may not be an obligatory role for extracellular calcium in the lethal hit involved in certain lytic events [6, 52, 62]. Calcium appears to be required for exocytosis and is absolutely required for the perforin pathway, yet several investigators have demonstrated lytic activity of certain target cells in the apparent absence of extracellular calcium or exocytosis of granular material [5, 6, 8]. These findings suggest that different mechanisms may be involved in lysis of target cells by CTL. CI'L induce rapid DNA damage in target cells that is not observed in complement-mediated lysis of cells [63]. Based on this observation, Russell proposed an "internal disintegration" model in which CTL activate target cell endonucleases which digest DNA yielding 150 to 180 pair units. The effector molecule produced by the CTL that triggers DNA damage has not been identified; however, several investigators have shown that lymphocytes produce factors that are cytotoxic. LT [64], tumor necrosis factor (TNF), and TNF/LT-Iike molecules such as cytotoxin [7] have been described as potential effector molecules produced by CTL. LT-containing supernat-

ants produced by CTL can induce DNA fragmentation in susceptible target cells [49]. LT-like molecules have been found associated with granules [7], and LT-like molecules are produced by T cells [64]. The presence of several of these molecules has been demonstrated in distinct subsets of T cells [20], and this has now been documented at the level of gene expression [65]. However, the role of these molecules in CTL lysis is controversial.

Summary Cytotoxic T lymphocytes (CTL) are thought to be the principle effector cells in the killing of tumor cells or virus infected cells, as well as in some forms of allograft rejection; as such, they represent an important element in the immune responses against these challenges. The properties of T lymphocytes having cytotoxic potential, and the mechanism(s) by which CTL injure target cells are important in developing strategies for modifying these immune responses. The ability to identify the lytic capabilities of T cell populations has depended on the availability of sensitive and accurate assay systems in vitro in which to measure their responses. Several recent reports indicate that cytotoxicity is an effector function that is distributed among the cells of distinct T cell subsets. We have tested the lytic potential of cloned T cells expressing either the CD8 or CD4 cell surface molecules. Our results indicate that CD8 and CD4 bearing T cells that have properties of either the T~I or the TH2 subset are capable of lysing nucleated target cells bearing antigen or coated with anti-CD3 mAb in a short term 51Cr release assay. We also tested the capacity of these clones to lyse efficiently SRBC coated with anti-CD3 mAb by a mechanism (presumably involving membrane damage) that does not involve nuclear degradation. It appears that such a mechanism is present in CD8 ÷ CTL and at least some CD4 ÷ T cell subsets. The role in vivo of such cytotoxic activity has not been determined. However, regulation of cellular immunity appears to be an inherent function of many (and perhaps all) of the T cell subsets. It is possible that the cytotoxic activity observed in some of

79 the T cell subsets may have a role in regulating the functional interactions between the members of subpopulations of T cells, and between T lymphocytes and other cells of the immune system.

12.

13.

Acknowledgements The authors wish to acknowledge Chyi-Song Hsieh, Peggy Schneider, and Yukio Hamada for their technical assistance. We also are grateful for the generous gift of the D10 cell line from Charles Janeway, and of mAb producing hybridoma cell lines and mAb preparations from Jeffrey Bluestone and David Segal. The secretarial assistance of Frances Mills also was greatly appreciated. This research was supported by Grant CA-44372 from the National Institutes of Health, U.S. Public Health Service.

14.

15.

16.

17.

References 1. Cerottini J-C, Brunner KT. Cell-mediated cytotoxicity, allograft rejection, and tumor immunity. Adv Immunol 1974; 18: 67-132. 2. Perlmann P, Holm G. Cytotoxic effects of lymphoid cells in vitro. Adv Immunol 1969; 11: 117-93. 3. Clark WR, Golstein P, eds. Mechanisms of cell-mediated cytotoxicity. New York: Plenum Press, 1982: 1-597. 4. Young JD. Killing of target ceils by lymphocytes: a mechanistic view. Physiol Rev 1989; 69: 250-314. 5. Ostergaard HL, Kane KP, Mescher MF, Clark WR. Cytotoxic T lymphocytes mediated lysis without release of serine esterase. Nature 1987; 330: 71-2. 6. Trenn G, Takayama H, Sitkovsky MV. Exocytosis of cytolytic granules may not be required for target cell lysis by cytotoxic T-lymphocytes. Nature 1987; 330: 72-4. 7. Liu C-C, Steffan M, King F, Young JD-E. Identification, isolation, and characterization of a novel cytotoxin in murine cytolytic lymphocytes. Cell 1987; 51: 393-403. 8. Lancki DW, Kaper BP, Fitch FW. The requirements for triggering of lysis by cytolytic T lymphocyte clones. II. Cyclosporin A inhibits TCR-mediated exocytosis but only selectively inhibits TCR-mediated lyric activity by cloned CTL. J Immunol 1989; 142: 416-24. 9. Wagner H, Starzinski-Powitz A, Jung H, Rrllinghoff M. Induction of I region-restricted hapten-specific cytotoxic T lymphocytes. J Immunol 1977; 119: 1365-8. 10. Moretta A, Mingari MC, Haynes BF, Sekaly RP, Moretta L, Fauci AS. Phenotypic characterization of human cytolytic T lymphocytes in mixed lymphocyte culture. J Exp Med 1981; 153: 213-8. 11. Swain SL, Dennert G, Wormsley S, Dutton RW. The Lyt

18.

19.

20.

21.

22.

23.

24.

phenotype of a long-term allospecific T cell line. Both helper and killer activities to IA are mediated by Ly-1 cells. Eur J Immunol 1981; 11: 175-80. Tite JP, Janeway CA Jr. Cloned helper T cells can kill B lymphoma cells in the presence of specific antigen: Ia restriction and cognate vs. non cognate interactions in cytolysis. Eur J lmmunol 1984; 14: 878-86. Vidovic D, Juretic A, Nagy ZA, Klein J. Lyt Phenotypes of primary cytotoxic T cells generated across the A and E region of the H-2 complex. Eur J Immunol 1981; 11: 499-504. Ball E J, Stasny P. Cell-mediated cytotoxicity against HLA-D-region products expressed in monocytes and B lymphocytes. IV. Characterization of effector cells using monocional antibodies against human T-cell subsets. Immunogenetics 1982; 16: 157-69. Shinohara N, Kojima M. Mouse alloantibodies capable of blocking cytotoxic T cell function. V. The majority of I region-specific CTL are Lyt-2 ÷ but are relatively resistant to anti-Lyt-2 blocking. J Immunol 1984; 132: 578-83. Spits H, Ijssel H, Thompson A, De Vries JE. Human T4 ÷ and T8 ÷ cytotoxic T lymphocyte clones directed at products of different class II major histocompatibility complex loci. J Immtmol 1983; 131: 678-82. Meuer SC, Hodgdon JC, Cooper DA, Hussey RE, Fitzgerald KA, Schlossman SF, Reinherz EL. Human cytotoxic T cell clones directed at autologous virus-transformed targets: Further evidence for linkage of genetic restriction to T4 and T8 surface glycoproteins. J Immunol 1983; 131: 186-90. Pierres A, Naquet P, Van Agthoven A, Bekkhoucha F, Denizot F, Mishal Z, Schmitt-Verhulst A-M, Pierres M. A rat anti-mouse T4 monoclonal antibody (H129.19) inhibits the proliferation of Ia-reactive T cell clones and delineates two phenotypically distinct (T4 +, Lyt-2,3-, and T4-, Lyt-2,3 ÷) subsets among anti-Ia cytolytic T cell clones. J Immunol 1984; 132: 2775-82. Swain SL, Panfili PR. Helper cells activated by allogenic H-2K or H-2D differences have a Ly phenotype distinct from those responsive to I differences. J Immunol 1979; 122: 383-91. Fong TA, Mosmann TR. Alloreactive murine CD8 ÷ T cell clones secrete the Thl pattern of cytokines. J Immunol 1990; 144: 1744-52. Kim KJ, Kanellopoulos-Langevin C, Merwin RM, Sachs DH, Asofsky R. Establishment and characterization of BALB/c lymphoma lines with B cell properties. J Immunol 1979; 122: 549-54. Glimcher LH, Sharrow SO, Paul WE. Serologic and functional characterization of a panel of antigen-presenting cell lines expressing mutant I-A class II molecules. J Exp Med 1983; 158: 1573-88. Spits H, van Schooten W, Keizer H, van Seventer G, van de Rijn M, Terhorst C, De Vries JE. Alloantigen recognition is preceded by nonspecific adhesion of cytotoxic T cells and target cells. Science 1986; 232: 403-5. Blanchard D, Van Els C, Aubry JP, De Vries JE, Spits H. CD4 is involved in a post-binding event in the cytolytic reaction mediated by' human CD4 ÷ cytotoxic T lymphocyte clones. J Immunol 1988; 140: 1745-52.

80 25. Bank I, Chess L. Perturbation of the T4 molecule transmits a negative signal to T cells. J Exp Med 1985; 162: 1294-303. 26. Saizawa K, Rojo J, Janeway CA Jr: Evidence for a physical association of CD4 and the CD3t~/3 T-cell. Nature 1987; 328: 260-3. 27. Palacios R, Martinez-Maza O. Is the E receptor on human T lymphocytes a "negative signal receptor"?. J Immunol 1982; 129: 2479-85. 28. Meuer SC, Hussey RE, Fabbi M, Fox D, Acuto O, Fitzgerald KA, Hodgdon JC, Protentis JP~ Schlossman SF, Reinherz EL. An alternative pathway of T-cell activation: a functional role for the 50 kd T l l sheep erythrocyte receptor protein. Cell 1984; 36: 897-906. 29. Golding H, Munitz TI, Singer A. Characterization of antigen-specific, Ia-restricted, L3T4 ÷ cytolytic T lymphocytes and assessment of thymic influence on their self specificity. J Exp Med 1985; 162: 943-61. 30. Kaplan DR, Griffith R, Braciale VL, Braciale TJ. Influenza virus-specific human cytotoxic T cell clones: heterogeneity in antigenic specificity and restriction by class II MHC products. Cell Immunol 1984; 88: 193-206. 31. Erb P, Grogg D, Troxler M, Kennedy M, Fluri M. CD4 ÷ T cell-mediated killing of MHC class II-positive antigenpresenting cells. I. Characterization of target celt recognition by in vivo or in vitro activated CD4 + killer T cells. J Immunol 1990; 144: 790-5. 32. Lancki DW, Lorber MI, Loken MR, Fitch FW. A clone specific monoclonal antibody which inhibits T cellmediated cytolysis. Adv Exp Med Biol 1982; 146: 55762. 33. Meuer SC, Hodgdon JC, Hussey RE, Protentis JP, Schlossman SF, Reinherz EL. Antigen-like effects of monoclonal antibodies directed at receptors on human T cell clones. J Exp Med 1983; 158: 988-93. 34. Ertl HCJ, Greene MI, Noseworthy JH, Fields BN, Nepom JT, Spriggs DR, Finberg RW. Identification of idiotypic receptors on reovirus- specific cytolytic T cells. Proc Natl Acad Sci USA 1982; 79: 7479-83. 35. Lancki DW, Ma DI, Havran WL, Fitch FW. Cell surface structures involved in T cell activation. Immunol Rev 1984; 81: 65-94. 36. Staerz UD, Kanagawa O, Bevan MJ. Hybrid antibodies can target sites for attack by T cells. Nature 1985; 314: 628-31. 37. Spits H, Yssel H, Leeuwenberg J, De Vries JE. Antigenspecific cytotoxic T cell and antigen-specific proliferating T cell clones can be induced to cytolytic activity by monoclonal antibodies against T3. Eur J Immunol 1985; 15: 88-91. 38. Leo O, Sachs DH, Samelson LE, Foo M, Quinones R, Gress R, Bluestone JA. Identification of monoclonal antibodies specific for the T cell receptor complex by Fc receptor-mediated CTL lysis. J Immunol 1986; 137: 3874-80. 39. Hsi ED, Siegel JN, Minami Y, Luong ET, Klausner RD, Samelson LE. T cell activation induces rapid tyrosine phosphorylation of a limited number of cellular substrates. J Biol Chem 1989; 264: 10836-42.

40. MacDonald HR, Thiernesse N, Cerottini J-C. Inhibition of T cell-mediated cytolysis by monoclonal antibodies directed against Lyt-2. Heterogeneity of inhibition at the clonal level. J Immunol 1981; 126: 1671-5. 41. Gajewski TF, Joyce J, Fitch FW. Anti-proliferative effect of IFN-3, in immune regulation. III. Differential selection of TH1 and TH2 murine helper T lymphocyte clones using recombinant IL-2 and recombinant IFN-7. J Immunol 1989; 143: 15-22. 42. Dennert G, Anderson CG, Prochazka G. High activity of Na-benzylocarbonyl-L-lysine thiobenzyl ester serine esterase and cytolytic perforin in cloned cell lines is not demonstrable in in vivo-induced cytotoxic effector cells. Proc Natl Acad Sci USA 1987; 84: 5004-8. 43. H~iyry P, Anderson LC. Generation of T memory cells in one-way mixed lymphocyte culture. IV. Primary and secondary responses to soluble and insoluble membrane preparations and to ultraviolet-light-inactivated stimulator cells. Scand J Immunol 1976; 5: 391-9. 44. Berke G, Rosen D. Are lytic granules and perforin 1 involved in lysis induced by in vivo-primed peritoneal exudate cytolytic T lymphocytes. Transplant Proc I987; 19: 412-16. 45. Berke G, Rosen D. Highly lyric in vivo primed cytolytic T lymphocytes devoid of lyric granules and BLT-esterase activity acquire these constituents in the presence of T cell growth factors upon blast transformation in vitro. J Immunol 1988; 141: 1429-36. 46. Brooks CG, Urdal DL, Henney CS. Lymphokine-driven "differentiation" of cytotoxic T-cell clones into cells with NK-like specificity: Correlations with displays of membrane macromolecules. Immunol Rev 1983; 72: 43-72. 47. Havele C, Bleackley RC, Paetkau V. Conversion of specific to nonspecific cytotoxic T lymphocytes. J Immunol 1986; 137: 1448-54. 48. Mtller G ed. Molecular mechanisms of T cell-mediated lysis. Immunol Rev 1988; 103: 1-211. 49. Schmid DS, Tite JP, Ruddle NH. DNA fragmentation: Manifestation of target cell destruction mediated by cytotoxic T-cell lines, lymphotoxin-secreting helper T-cell clones, and cell-flee lymphotoxin-containing supernatant. Proc Natl Acad Sci USA 1986; 83: 1881-5. 50. Tite JP, Powell MB, Ruddle NH. Protein-antigen specific h-restricted cytolytic T cells: Analysis of frequency, target cell susceptibility, and mechanism of cytolysis. J Immunol 1985; 135: 25-33. 51. Ju ST, Ruddle NH, Strack P, Doff ME, DeKruyff RH. Expression of two distinct cytolytic mechanisms among murine CD4 subsets. J Immunol 1990; 144: 23-31. 52. Henney CS. Studies on the mechanism of lymphocytemediated cytolysis. J Immunol 1973; 110: 73-84. 53. Poenie M, Tsien RY, Schmitt-Verhulst AM. Sequential activation and lethal hit measured by [Ca2+]i in individual cytolytic T cells and targets. EMBO J 1987; 6: 2223-32. 54. Dourmashkin RR, Deteix P, Simone CB, Henkart P. Electron microscopic demonstration of lesions in target cell membranes associated with antibody-dependent cellular cytotoxicity. Clin Exp Immunol 1980; 42: 554-60.

81 55. Dennert G, Podack ER. Cytolysis by H-2 specific T killer cells. Assembly of tubular complexes on target membranes. J Exp Med 1983; 157: 1483-95. 56. Lanzavecchia A, Staerz UD. Lysis of nonnucleated red blood cells by cytotoxic T lymphocytes. Eur J Immunol 1987; 17: 1073-4. 57. Glasebrook AL, Fitch FW. Alloreactive cloned T cell lines. I. Interactions between cloned amplifier and cytolyric T cell lines. J Exp Med 1980; 151: 876-95. 58. Lancki DW, Hsieh C-S, Fitch FW. Mechanisms of lysis by Cytotoxic T lymphocyte clones. Lytic activity and gene expression in cloned antigen-specific CD4 ÷ and CD8 ÷ T lymphocytes. J. Immunol 1991; 146: 3242-49. 59. Henkart MP, Henkart PA. Lymphocyte mediated cytolysis as a secretory phenomenon. Adv Exp Med Biol 1982; 146: 227-47. 60. Chang TW, Eisen HN. Effects of N alpha-tosyl-L-lysyslchloromethylketone on the activity of cytotoxic T lymphocytes. J Immunol 1980; 124: 1028-33.

61. Takayama H, Sitkovsky MV. Antigen receptor-regulated exocytosis in cytoxic T lymphocytes. J Exp Med 1987; 166: 725-43. 62. Tirosh R, Berke G. T-Lymphocyte-mediated cytolysis as an excitatory process of the target. I. Evidence that the target cell may be the site of Ca2 + action. Cell Immunol 1985; 95: 113-23. 63. Russell JH. Internal disintegration model of cytolytic lymphocyte-induced target damage. Immunol Rev 1983; 72: 97-118. 64. Conta BS, Powell MB, Ruddle NH. Activation of Lyt-1 ÷ and Lyt-2 ÷ T cell cloned lines: Stimulation of proliferation, lymphokine production, and self-destruction. J Immunol 1985; 134; 2185-90. 65. Jongeneel CV, Nedospasov SA, Plaetinck G, Naquet P, Cerottini JC. Expression of the tumor necrosis factor locus is not necessary for the cytolytic activity of T lymphocytes. J Immunol 1988; 140: 1916-22.