Immobilized anti-CD3-induced T cell growth: Comparison of the frequency of responding cells within various T cell subsets

CELLULAR

133,206-2 18 ( 1991)

IMMUNOLOGY

Immobilized Anti-CD34nduced T Cell Growth: Comparison of the Frequency of Responding Cells within Various T Cell Subsets’ THOMAS D. GEPPERT~ AND PETER E. LIPSKY Harold C. Simmons Arthritis Research Center, Department of Internal Medicine, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas 75235 Received August 16. 1990; accepted October 15, 1990 Human T cellscan be divided into subsetsbasedon the expressionof CD29, CD45RA, CD45R0, LFA-3, or CD1 la. It has been suggestedthat the subset of CD4+ T cells that expresseshigh densities of CD29, CD1 la, CD45R0, and LFA-3 contains “memory” T cells, whereasthe subset of ceils that expressesCD45RA contains “naive” T cells. In order to obtain a more complete picture of the functional capacities of human naive and memory CD4+ and CD8+ T cell subsets, highly purified T cells were activated with a uniform stimulus and responseswere examined in bulk cultures and under limiting dilution conditions. T cell activation was achieved with an immobilized mAb to the CD3 molecular complex, 64.1. In bulk cultures, immobilized 64.1 stimulated a vigorous response.Moreover, the number of cells entering the cell cycle, the magnitude of the [‘Hlthymidine incorporation, and the growth of the cells over 6 days in culture by naive and memory CD4+ T cells was comparable. To delineate the frequency of responsive cells in each subset more precisely, cells were cultured with immobilized 64.1 at limiting dilution and the precursor frequency of responding cells was assessedby examining wells microscopically for visible growth. Immobilized 64. I was able to induce some T cells from each subset to grow in the complete absenceof AC, when exogenous IL2 was present. The number of responding CD4+ and CDS+ cells was comparable. The percentage of naive cells responding in each population was approximately three times greater than the frequency of memory cells. IL4 could also support the growth of immobilized 64. l-activated CD4+ T cells, but the frequency of responding cells was much lower than that supported by IL2. The vast majority ofthe IL-4 responsive CD4+ cells resided within the naive cell subset. The data indicate that the response of CD4+ and CD8+ naive and memory T cell subsetsto immobilized anti-CD3 depends on the density of responding cells. Naive T cells have an enhanced capacity to grow when cultured in the absenceof other T cells or accessorycells.This ability may facilitate their expansion during primary immune responses. 0 1991 Academic

Press. Inc.

INTRODUCTION

Human T cells can be divided into subsetsbased on the expression of various cell surface molecules. Thus, for example, T cells can be partitioned into two distinct populations based on the expression of CD4 or CD8 (l-5). Moreover, both CD4+ and CD8+ T cells can be further subdivided into subpopulations based on their expression of CD29, CD45RA, and CD45RO or the density of LFA-3 and CD1 la (6-10). ’ This work was supported by U.S. Public Health ServicesGrants AR09989 and AR36 169. * Recipient of an Arthritis Investigator Award of the Arthritis Foundation. 206 OOOS-8749/91$3.00 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form rewved.

ANTKD34NDUCED

RESPONSES OF T CELL

SUBSETS

207

It has been suggested that the subset of CD4+ T cells that expresses high densities of CD29, CD 1la, CD45R0, and LFA-3 contains “memory” T cells, whereas the subset that expresses CD45RA contains the “naive” T cells ( 11, 12). Support for this notion comes from the observation that the population of CD4+ T cells expressing CD29, CD45R0, and CD1 la contains the bulk of the T cells that respond to recall antigens (6, 7, 9-l l), whereas both subsets respond comparably to mitogenic lectins (6, 7, 9, 11, 12). Moreover, preimmune cord blood T cells express predominantly CD45RA (9, 13). Finally, CD45RA-positive cells lose CD45RA expression following mitogen-induced activation and acquire increased expression of CD29, CD45R0, LFA-3, and CD1 la (9, 14-16). Recently, several studies have examined the activation requirements for memory and naive CD4+ T cell subsets (6, 7, 14, 17- 19). Although the results of these studies have frequently been conflicting, most have come to the conclusion that the activation requirements for naive CD4+ T cells are more stringent than those for memory CD4+ T cells. Thus, for example, most investigators have reported that soluble mAb to CD2 or CD3 stimulate proliferation of memory, but not naive CD4+ T cells (12, 14, 1719), but when more potent stimuli such as phytohemagglutinin (PHA) are employed, the subsets respond comparably (6, 7, 9, 14, 19). These studies assessed responses within the first one or two divisions of the responding cells and focused on stimuli that triggered activation of memory but not naive T cells. Thus, these previous studies do not assessthe requirements for memory and naive T cells to grow once activated. Moreover, since the experiments were carried out under bulk culture conditions the studies could not determine the potential role of contaminating accessory cells or interactions between CD4+ T cells in the induction of cell cycle entry and progression. Recent studies have demonstrated that immobilized mAb to CD3 can induce the activation of the majority of human T cells in bulk cultures (20). To determine the actual percentage of responding CD4+ T cells, the precursor frequency of immobilized anti-CD3-stimulated T cells that proliferated and produced lymphokines was examined using limiting dilution techniques (2 1). These studies demonstrated that immobilized anti-CD3 can stimulate approximately 26% of CD4f T cells to grow and proliferate in the complete absence of accessory cells when supplemental IL2 is present in the medium. IL4 could also support anti-CD3-induced T cell growth, although the frequency of responding cells was much lower. Immobilized anti-CD3 induced IL2 production from approximately 8% of CD4+ T cells. This response differed from proliferation in that it was nearly completely dependent on the presence of accessory cells. The finding that only a fraction of the T cells could grow following immobilized antiCD3-induced stimulation in limiting dilution cultures, whereas the majority became activated in bulk cultures, suggests that cooperation between T cells was necessary for some of them to be activated or that cells differed in their capacity to grow and produce cytokines after initial activation. Whether variability in the functional outcome of anti-CD3-induced activation reflected the naive versus the memory phenotype of the CD4+ T cells was not analyzed. The current studies, therefore, were carried out to determine the frequency of cells within these various T cell subsets that grow under limiting dilution conditions following stimulation with immobilized 64.1. The data demonstrate that following anti-CD3-induced activation, the majority of the CD4+ and CD8+ T cells that are capable of growing in response to IL2 or IL4 reside within the naive subset. The data indicate, therefore, that whereas naive T cells

208

GEPPERT

AND

LIPSKY

may have more rigorous activation requirements, they exhibit a greater intrinsic pacity to grow following activation than memory T cells. MATERIALS

AND

ca-

METHODS

MAb. A variety of mAb to determinants expressed by various T cell subsets were employed including 153.13 (gift from Dr. Nigel Staite, Upjohn, Kalamazoo, MI), an IgGl mAb to CD4; OKT8 [American Type Culture Collection (ATCC)], an IgG2a mAb directed at CDS; TS2/9.1.1 (ATCC), an IgG 1 mAb directed at LFA-3 (22); 2H4 (Coulter, Hialeah, FL), an IgGl mAb directed at CD45RA (7); UCHLl (DAKO, Santa Barbara, CA), an IgG 1 mAb directed at CD45RO ( 10); and 4B4-RD 1 (Coulter), an IgGl mAb to CD29 (6). In addition, 64.1, an IgG2a mAb directed at the CD3 molecular complex on mature T cells (23); L243 (ATCC), an IgG2a mAb directed at a monomorphic determinant present on HLA-DR molecules; and B73.1 (a gift from Dr. Giorgio Trinchieri), an IgGl mAb directed at CD16 (24) were used. B73.1 and 64.1 were purified by passage over a column of Sepharose 4b coupled with staphylococcal protein A. L243, TS2/9.1.1, 153.13, and OKT8 were used as culture supernatants. Reagents and mitogens. Recombinant IL2 was provided by The Cetus Corporation (Emeryville, CA). Recombinant IL4, was purchased from Genzyme Corporation (Boston, MA). PHA was obtained from Wellcome Reagents Division, Burroughs Wellcome Co. (Research Triangle Park, NC). Cell preparation. PBMC were obtained from healthy adult volunteers by centrifugation of heparinized venous blood over sodium diatrizoate/ficoll gradients (Isolymph, Gallard Schlesinger Chemical Mfg. Corp., Carle Place, NY). Cells were washed once in Hanks’ balanced salt solution (HBSS) and twice in saline before further processing. Preparation of T cell subsets. T cell subsets were prepared by first enriching the population for lymphocytes. The lymphocyte-enriched population was isolated by either removing those cells adherent to glass dishes, passing the cells over a nylon wool column, or rosetting with neuraminidase-treated SRBC, isolating the rosette forming cells, and passing them over a nylon wool column (25). It has been demonstrated that these methods do not alter the percentage of cells responding to immobilized antiCD3 in this system (2 1). T cell subsets were then isolated by negative or positive selection using either the fluorescence activated cells sorter (FACS) or panning. T cell subsets were isolated using a panning technique as described (26). Briefly, the cells were reacted with saturating concentrations of the CD 16-specific mAb, B73.1, an anti-HLA-D mAb, IVA I2 and L243, and either an anti-CD4 mAb ( 153.13) or an anti-CD8 mAb (OKT8) alone or in combination with mAb to CD45RA or LFA-3. MAb to HLA-DR were utilized to remove IaS accessory cells and Ia+ T cells. Ia+ T cells were removed because they likely represent activated T cells whose activation requirements may be different from freshly isolated resting T cells. After washing, the cells were added to goat anti-mouse immunoglobulin (GaMIg)-coated panning dishes and were incubated for 90 min at 4°C. Afterward, the nonadherent cells were gently aspirated and were panned a second time on another GaMIg-coated petri dish. In each case, only the negatively selected cells were employed. T cell subsets were also isolated by negative selection using the FACS as described (27). PBMC or T lymphocytes were reacted with saturating concentrations of B73.1, IVA12, L243, and either 153.13, or OKT8 alone or in combination with TS2/9.1.1,

ANTI-CD3-INDUCED

RESPONSES OF T CELL

SUBSETS

209

UCHLl, 4B4, or 2H4. The cells were then counterstained with FITC-GaMIg and sorted directly into microtiter wells based on their relative fluorescence intensity using an automatic cell deposition unit. When positive selection was used CD4+ T cells were first purified from T cells by panning and then reacted with TS2/9.1.1, UCHL 1, 4B4, or 2H4 and sorted directly into microtiter wells based on their fluorescence intensity. Adherence of anti-CD3 to culture wells. 64.1 was diluted in 0.1 M Tris buffer (pH 9.5). One-half microgram of mAb diluted in 50 ~1 of buffer was placed in each of the wells of a 96-well flat-bottom polystyrene microtiter plate (No. 3596; Costar, Cambridge, MA) and incubated for at least 3 hr at room temperature. The wells were washed twice with saline to remove nonadherent mAb before the addition of cells. Techniquesof cell culture. Cells were cultured in medium RPM1 1640 supplemented with 10% heat-inactivated normal human serum (NHS),3 penicillin G (200 units/ml), gentamicin (10 pg/ml), and L-glutamine (0.3 mg/ml). All cultures were carried out in a total volume of 100-200 ~1. 13H]Thymidine incorporation was assessedby culturing 1 X lo5 T cells with immobilized anti-CD3 (0.5 pg/well). Cell growth and lymphokine production were assessedby culturing varying numbers of CD4+ T cells alone or with immobilized 64.1 (0.5 pg/well). Preliminary experiments, using various concentrations of immobilized 64.1, indicated that 0.5 pg/well of 64.1 yielded optimal CD4+ T cell responsiveness (2 1). Determination of the percentage of CD4-k T cells proliferating in response to immobilized 64.1. When the T cell subset was prepared by panning, the percentage of responding cells was determined as described (2 1). Briefly, T cells were suspended in culture medium such that 100 ~1 of medium containing 6, 4, 3, 2, or 1 cell. 48 wells were then seeded with 100 ~1 from each dilution of cells. The cells were cultured in the presence of immobilized 64.1 and with various cytokines including IL2 (40 units/ ml) or IL4 ( l-2 rig/ml). Preliminary experiments indicated that these concentrations of cytokines yielded optimal responses. Following 6- 10 days in culture, the wells were examined by a blinded observer using an inverted light microscope. Positive wells were identified by the presence of a tightly approximated colony of cells containing more than a three- to fourfold increase in the number of cells originally seeded. The number of negative wells was used to determine the frequency of responding T cells by the procedure of maximum likelihood as described (28). Data are expressed as the precursor frequency of responding T cells & the standard error of the frequency determined by maximum likelihood (SEr). When the T cell subsets were prepared using the FACS, the cells were sorted directly into the wells using the automated cell deposition unit. One or three cells were sorted into each well of a 96-well microtiter plate. Positive wells were identified as above and the frequency of responding cells determined using the formula: X = (1 - F)“, where F is the frequency of responding cells, n is the number of cells added to each well, and X is the fraction of cells that are negative. Standard deviation was determined from the variance (F), which is derived by expanding 1 - X(‘ln) in a Taylor-series approximation. Thus, the variance (F) = (d(F)/d(X))2 X variance (X), where d(f )/ d(x) is evaluated at the observed frequency of negative wells.

3 Abbreviations used: FBS, fetal bovine serum: GaMIg, goat anti-mouse immunoglobulin; inactivated normal human serum: PHA. phytohemagglutinin.

NHS, heat-

210

GEPPERT

AND

LIPSKY

RNA and DNA content. Cellular content of RNA and DNA was determined after acridine orange (AO) staining using the technique described by Darzynkiewicz et al. (29). Briefly, 3 X lo5 cells per sample were collected from cultures in microtiter wells after 30 hr of incubation, washed once, and suspended in 100 ~1 fresh medium containing 10% NHS. The cells were made permeable in a chilled solution containing 0.1% (vol/vol) Triton X-100 (Sigma Chemical Co.), 0.2 M sucrose, 10e4A4 EDTA, and 2 X lop2 A4 citrate phosphate buffer, at pH 3.0. After 6 days, the cells were stained by adding a second solution of 0.002% purified A0 (Polysciences Inc., Warrington, PA), 0.1 M NaCl, and 10m2A4 citrate phosphate buffer, at pH 3.8. The cells were analyzed with a cytofluorograph (System 50-HH; Ortho Diagnostic Instruments, Westwood, MA) using an argon laser setting of 488 nm at 50 mW. The red (RNA) and green (DNA) fluorescenceemissions from each cell were separatedand quantitated by individual photomultipliers. The data are based on the analysis of 9 X 1O3cells per sample. Cell growth. T cells (5 X 104)were cultured with IL2 in the presence or absenceof immobilized 64.1 and IL2 and the number of T cells in the well was determined as described (29). Briefly, after 6 days in culture, an aliquot of the lysing agent (ZapIsoton) was added to the cultures to disrupt the plasma membrane. Free nuclei were then counted using a Model D2 Coulter Automatic Blood Cell Counter (Coulter) and the total number of cells in each well was calculated. The data are expressed as the means of triplicate determinations f SEM. RESULTS Anti-CD3-induced f3H]thymidine incorporation by memory and naive CD4+ T cells in bulk cultures. Figure 1 depicts 64. l-stimulated proliferation of both memory and naive CD4+ T cells. 64.1-stimulated naive and memory CD4+ T cell proliferation was comparable regardlessof the day the experiment was harvested (Days 3-6). When the results from seven experiments harvested on Day 3 were compared, no statistically significant differences were noted (77,000 ? 33,000 cpm for memory CD4+ T cells and 58,000 f 25,000 for naive CD4+ T cells, mean f SEM, P > 0.638). Frequency of cells within CD4+ subsetscapable of growth in responseto immobilized anti-CD3 and ZL2. The frequency of 64. l-stimulated CD4+ T cell subsetsgrowing in the presence of IL2 was determined using a modified limiting dilution technique. Initial experiments utilized CD4+ T cell subsetsprepared by panning. As can be seen in Fig. 2, the frequency of responding naive CD4+ T cells responding was markedly greater than the frequency of memory CD4+ T cells. Similar results were obtained with subsetspurified with the FACS. The fluorescence histograms of CD4+ T cells stained for CD45RO (UCHL l), CD29 (4B4), and CD45RA (2H4) are depicted in Fig. 3. As can be seen, CD4+ T cells can be subdivided based on the fluorescence intensity of staining with these mAb, as has been described previously. The size of these subpopulations varies somewhat from donor to donor (data not shown). This is most clear when the cells are stained with UCHLl and 4B4, but can also be seen when the cells are reacted with 2H4. The histograms were then divided into regions based on fluorescence intensity as depicted in Fig. 3. With the exception of the UCHLl-stained cells within region 2, each region contained approximately 25% of the CD4+ T cells analyzed. The UCHL lstained CD4+ T cells within region 2 represent 50% of the cells analyzed. The frequency

ANTI-CD3-INDUCED

EXF’T A

RESPONSES OF T CELL

211

SUBSETS

TIME CELLS HARVESTED (DAYS)

1

3

2

7

i

5

2

0

100

3H-thymidine

200

incorporation

300

400

@pm x lo-‘)

FIG. 1. Immobilized anti-CD3-induced proliferation of memory and naive CD4+ T cell subsets. T lymphocytes were stained with mAb to OKT8 and either CD45RA or LFA-3 and panned to deplete the population of cells bearing these determinants. The remaining CD4+ T cell subsets (IO’ cells/wells) were cultured with immobilized 64. I for 3 days and [3H]thymidine incorporation was determined. The data are expressed as the mean cpm of triplicate determinations -t SEM.

of cells from within each region that grew in response to immobilized anti-CD3 and IL2 is depicted in Table 1. As can be seen, the highest frequency of responding cells was found in the subsets with the greatest densities of CD45RA and the least expression of CD29 and CD45RO. The results of eight experiments using cells from eight individuals are depicted in Table 2. The cell populations used in these experiments were isolated in a variety of ways to ensure that the results were not artifacts of the method employed to purify the cells. Although the absolute frequency of responding cells varies among the various donors, the relative frequency of responding cells within the naive CD4+ T cell subset was always greater than that within the memory subset, regardless of the donor or the method used to isolate the specific subset.

Frequency of 64.1~stimulated CD4+ T cell subsetsgrowing in the presence of IL4. Memory cells are known to produce IL4 (30). It was therefore possible that this population would require IL4 rather than IL2 for cell growth. The frequency of anti-CD3stimulated memory and naive T cells growing in response to IL2 and IL4 was, therefore, examined and compared. As can be seen in Table 3, the vast majority of the IL4 responsive cells as well as the IL2 responsive cells were found in the naive cell subset. In both subsets, the frequency of IL4 responsive cells was less than the frequency of IL2 responsive cells.

Comparison of the responseof immobilized 64.l-stimulated naive and memory CD4+ T cells in bulk cultures with their response under limiting dilution conditions. The response of naive and memory CD4+ T cells to immobilized 64.1 in bulk cultures was compared with their response under limiting dilution conditions (Table 4). The percentage of immobilized 64.1 -stimulated naive and memory CD4+ T cells entering

212

GEPPERT

AND

LIPSKY

LEXPT I

I 1

0

I

20

I

40

I

I

..I....

60

80

PERCENTAGE OF RESPONDING CELLS

FIG.2. Frequency of cells within CD4+ subsets capable of growth in response to immobilized anti-CD3 and IL2. Memory and naive CD4+ T cell subsets were prepared from T cells by panning to remove CDSpositive and either CD45RA- or LFA-3-positive cells. Varying numbers (1-5 cells/well) were cultured with immobilized 64.1 (500 rig/well) and IL2 (40 units/ml) for 7 days. The percentage of wells with or without growth was determined by light microscopy and the percentage of responding cells was determined by the method of maximum likelihood analysis.

the cell cycle was comparable in bulk culture. Similarly, the growth of immobilized 64.1-stimulated naive and memory CD4+ T cells was comparable in bulk cultures. In contrast, the percentage of responding naive CD4f T cells was nearly threefold greater than the percentage of responding memory T cells under limiting dilution conditions. Frequency of various 64. l-stimulated T8 cell subsetsgrowing in the presenceof IL2. The next experiments examined the capacity of CD8+ T cell subsetsto grow in response to immobilized CD3 and IL2. Results of initial experiments examining the frequency

484

2H4

t

L.---A 23

Relative

Fluorescence

Intensity

4

(log scale)

FIG. 3. Histograms portraying the staining patterns of CD4+ T cells with UCHL 1, 4B4, and 2H4. CD4+ T cells were reacted with the indicated mAb, counterstained with a fluorescein-conjugated GaMIg, and analyzed with the fluorescence-activated cell sorter. The cells are divided into groups based on the intensity of staining and the responsiveness of these groups examined. Data are shown in Table I.

ANTI-CDII-INDUCED

RESPONSES OF T CELL TABLE

Immobilized

213

SUBSETS

I

Anti-CD3-Induced CD4+ T Cell Growth: Response of Various CD4+ T Cell Subsets Isolated Using The Fluorescence-Activated Cell Sorter Group

mAb

I

2

3

4

(‘% responding cells k SD) UCHLl 4B4 2H4

60.3 f 5.2 50.0 + 4.5 19.0 f 2.6

38.8 f 3.8 50.0 f 4.5 23.4 + 2.8

ND 19.1 f 2.6 ND

19.5 f 2.6 22.2 IL 2.8 37.8 f 3.8

Note. CD4-positive T cells were reacted with saturating concentrations of mAb to CD45RO (UCHLI), CD45RA (2H4), or CD29 (484) and counterstained with a fluorescein isothiocyanate-conjugated GaMlg. CD4+ T cells were separated into various groups based on the intensity of the staining with the mAb as shown in Fig. 3, and sorted directly into microtiter wells (3 cells/well) coated with immobilized anti-CD3 (64.1). Following 6 days in culture with IL2 (40 units/ml), the wells were examined microscopically for growth and the frequency of responding cells determined.

of 64.1 -stimulated CD4+ and CD8+ T cells growing in the presence of IL2 are depicted in Table 5. The frequency of responding cells varied substantially from individual to individual as described previously for CD4+ T cells. However, in each experiment, the frequency of CD4+ and CDS+ T cells responding was the same regardless of the method employed to isolate the T cell subsets. Although it appears from these four experiments that the frequency of responding cells was higher when the cells were isolated using the FACS, this was not a reproducible finding. As shown in Table 6, the frequency of responding CD8+ cells that were CD45RO-, LFA-3bIM, or CD29was significantly greater than that of CD45RA- cells. DISCUSSION The current studies compare the response of CD4+ and CD8+ naive and memory T cell subsets to immobilized mAb to CD3. The data demonstrate that the response of these subsets to immobilized anti-CD3 depends on the density of cells. When T cells are stimulated in bulk cultures with an immobilized mAb to CD3 (64.1) the percentage of cells entering the cell cycle and initial DNA synthesis were comparable. Moreover, the absolute growth of naive and memory T cell subsets was comparable. In contrast, when limiting dilution culture conditions were employed and the capacity of the cells to grow into a colony of cells was assessed,the frequency of immobilized anti-CD3-stimulated T cells responding was dramatically greater in the naive T cell subset regardless of the expression of CD4 or CDS. Since a uniform and potent stimulus was employed to activate highly purified subpopulations of T cells, the data suggest that once-activated naive T cell subsets exhibit a greater intrinsic capacity to grow and produce IL2. The current studies are the first to report an enhanced ability of naive T cells to grow under limiting dilution conditions. The number of cells entering the cell cycle, the initial DNA synthesis, and the absolute cell growth of naive and memory T cells stimulated by immobilized 64.1 is comparable. This result conflicts with previously reported findings that mAb to CD3

214

GEPPERT

AND

LIPSKY

TABLE 2 Immobilized

Anti-CD3-Induced

CD4+ T Cell Growth: Frequency of Naive and Memory Cells Responding in The Presence of IL2 CD4+ T cell subset

Naive Expt. I 2 3 4 5 6 -I 8

Memory

Phenotype

%++SD

LFA-3CD29CD29CD29CD45ROCD29LFA-3nIM LFA-3nIM

18.1 26.5 6.0 26.9 58.0 58.0 25.1 43.2

f 2.5 f 3.1 + 1.4 + 3.1 + 5.1 -c 5.1 + 3.0 f 4.0

Phenotype CD45RACD45RACD45RACD45RACD45R.k CD45RACD45RALFA-LUGHT

%+fSD 3.4 * 1.1 0.5 ?I 0.4 1.0 + 0.6 12.0 * 2.0 27.5 -+ 3.0 14.0 * 2.0 4.9 -+ 1.3 2.6 + 0.9

Note. Nonadherent lymphocytes (experiment I), nylon wool passed lymphocytes (experiments 2-5) lymphocytes that formed rosettes with neuraminidase-treated SRBC and did not adhere to a nylon wool column (experiments 6 and 7) and CD4-positive T cells (experiment 8) isolated from rosette-positive nylon wool nonadherent T cells by panning were reacted with mAb to CD29, CD45RA, CD45R0, and LFA-3 and counterstained with fluorescein isothiocyanate-conjugated GaMlg. T cells were sorted directly into wells (3 cells/well) containing IL2 (40 units/ml) and immobilized 64.1 (500 &well) based either on the absence of staining (experiments l-7) or on the density of staining (experiment 8). Following 6-8 days in culture, the wells were examined microscopically for growth and the frequency of responding cells was determined.

stimulate only memory T cell DNA synthesis and with similar studies that have demonstrated that naive CD4+ T cells are unable to respond to anti-CD2 mAb (12, 14, 17- I9,3 1). These previous studies either employed soluble mAb to CD3 or suboptimal immobilized anti-CD3. Thus, some studies utilized mAb to CD3 conjugated to Sepharose beads that deliver a less intense activation signal than immobilized 64.1 (32) whereas others utilized mAb to CD3 that when immobilized deliver a suboptimal activation signal (3 1). Since immobilized 64.1 has been shown to stimulate the majority of CD4+ T cells to enter the cell cycle (32) it does not seem surprising that it would stimulate both subsets to respond. The capacity of immobilized 64.1 to stimulate comparable naive and memory CD4+ T cell proliferation is similar to the responses observed when the cells are stimulated by lectins such as concanavalin A and PHA (6, 7, 9, 14, 19). The observation that naive T cells respond to some, but not all mitogens that activate memory T cells suggests that the activation requirements for naive T cells may be more rigorous than those for memory T cells. The current studies demonstrate, however, that both subsets can respond to immobilized anti-CD3 in bulk cultures. The finding that CD4+ naive and memory cells respond comparably in bulk cultures, but a higher frequency of naive cells grow under limiting dilution conditions has several potential explanations. First, it is possible that memory cells have an intrinsic deficiency in the capacity to grow sufficiently to produce an identifiable colony in vitro. This does not appear to be the case, however, since the growth of these cells in bulk cultures was comparable. Moreover, previous studies have found that nearly every T cell can be induced to undergo clonal growth in vitro, regardless of the phe-

ANTI-CD3-INDUCED

RESPONSES

OF T CELL

215

SUBSETS

TABLE 3 Immobilized

Anti-CD3-Induced

CD4+ T Cell Growth: Frequency of Naive and Memory Cells Responding in the Presence of IL4 CD4+ T cell subset Memory

Naive Expt.

Cytokine

1

IL2 IL4 IL2 IL4

2

Phenotype

%++-SD

CD29-

58.0 13.9 43.2 2.2

LFA-3oIM

k f + +

5.1 1.0 4.0 0.3

Phenotype CD45RALFA-~BRIGHT

%++SD 14.0 0.9 2.6 0.0

f f f f

2.0 0.8 0.9 0.2

Note. Rosette-positive T cells that did not adhere to nylon wool (experiment 1) or CD4+ T cells were reacted with mAb to CD29, CD45RA, and LFA-3 and counterstained with fluorescein isothiocyanateconjugated GaMIg. T cells were sorted directly into wells containing the indicated cytokine and immobilized 64.1 (500 rig/well) based either on the absence of staining (experiment I) or on the density of staining (experiment 2). Wells containing IL2 received three cells, whereas seven cells were placed into wells containing IL4. Following 6 days in culture, the wells were examined microscopically for growth and the frequency of responding cells was determined.

notype, when stimulated by PHA and supplemental IL2 and supported by large numbers of feeder cells (33). This suggests that there is no inherent deficiency in clonal growth of memory T cells in vitro. Alternatively, it is possible that the frequency of immobilized 64.1 -stimulated memory cells with the capacity to proliferate is less than the corresponding frequency of immobilized 64. l-stimulated naive T cells, but that the memory T cells divide more efficiently thereafter. In this way a smaller number

TABLE 4 Immobilized 64. l-Stimulated Naive and Memory CD4+ T Cells Become Activated and Grow Comparably in Bulk Cultures, but Naive Cells Grow More Effectively under Limiting Dilution Conditions

Expt.

I

T cell subset

Growth

Cell cycle progression

% Responding cells

Memory Naive

(Cells/well X IO-‘) 256.7 f 3.2 240.6 f 6.3

(% cells in Cl,, Gz + M or S) 55 69

25.3 f 3.6 69.9 + 9.0

Note. T cells were reacted with mAb to HLA-DR and CD8 panned with GaMIg-coated dishes to remove contaminating HLA-DR and CD8-positive cells. The CD4+ T cells were reacted with mAb to CD16, HLADR, OKT8, and UCHL-I or 2H4 and negatively selected using the FACS. For cell cycle progression and growth 5 X lo4 of each CD4+ T cell subset was then cultured with immobilized 64. I and IL2. Cell cycle progression was determined by acridine orange staining and analysis using the FACS 30 h following stimulation, whereas growth was measured by determining the number of cells in the wells 6 days following stimulation. Unstimulated cultures containing memory and naive CD4+ T cells had 2 X IO4 and 2.5 X lo4 cells/well, respectively, 6 days following stimulation. The percentage of responding cells was determined by aliquoting three cells into wells containing IL2 (40 units/ml) and immobilized 64.1 (500 ng/weII). Following 6 days in culture, the wells were examined microscopically for growth and the frequency of responding cells was determined.

216

GEPPERT

AND

LIPSKY

TABLE 5 Immobilized

Anti-CD3-Induced T Cell Growth: Frequency of CD4+ and CD8+ T Cells Responding in the Presence of IL2 CD4+

Expt.

CD8+ (‘% responding cells f SD or SE,)

1 2 3 4

25.1 33.0 4.2 6.3

+ f rt i

6.3 7.5 1.2 1.5

25.1 25.1 1.4 6.1

+ + + +

6.3 6.5 1.9 1.5

Note. T cells were reacted with mAb to HLA-DR and CD4 or CD8 and either counterstained with FITCGaMIg (experiments 1 and 2) or panned with GaMIg-coated dishes (experiments 3 and 4) to remove contaminating HLA-DR and CD4- or CD8-positive cells. Various numbers of panned T cells were aliquoted into wells (4, 2, or I cell/well) containing IL2 (40 units/ml) and immobilized 64. I (500 rig/well), whereas cells counterstained with FITC-GaMIg were sorted (3 cells/well) directly into wells based on the absence of staining. Following 6 days in culture, the wells were examined microscopically for growth and the frequency of responding cells determined. (SE,) in experiments 1 and 2 and SD in experiments 3 and 4.

of precursor cells might give rise to a comparable number of progeny. This seems unlikely, however, as the immobilized 64.1-stimulated naive T cells formed larger colonies than the memory T cells (data not shown). Moreover, the initial DNA synthesis and the number of cells entering the cell cycle was comparable in the bulk cultures. The most likely explanation for these findings is that the growth of memory T cells is more dependent on the presence of adjacent cells than the growth of naive T cells. Thus, it is possible that the response of memory T cells is dependent on lymphokines releasedby adjacent T cells or on accessorycell-derived cytokines. Alternatively, their response may be dependent on contact with activated T cells or contaminating accessory cells. The possibility that the response of memory cells is dependent on the presence of a contaminating accessorycell seemssomewhat unlikely, however, as the

TABLE 6 Immobilized

Expt. I

2

Anti-CD3-Induced T8 Cell Growth: Frequency of CD8+ T Cell Subsets Responding in the Presence of IL2 CD8+ T cell subset CD45RACD45ROLFA-3nn, CD45RACD29CD45RO-

Frequency of responding cells f SD 6.8 44.6 43.5 5.2 32.3 17.5

+- 1.5 2 4.2 +- 4.1 + 1.3 t 3.4 +- 2.5

Note. Nylon wool passed T lymphocytes were reacted with mAb to CD45RA, CD45R0, CD29, and LFA3 and counterstained with fluorescein isothiocyanate-conjugated GaMIg. MAb-negative T cells were sorted directly into microtiter wells (3 cells/well) containing IL2 (40 units/ml) and immobilized 64. I (500 rig/well). Following 6 days in culture, the wells were examined microscopically for growth and the frequency of responding cells was determined.

ANTI-CD3-INDUCED

RESPONSES OF T CELL

SUBSETS

217

responding cells in the bulk cultures were highly purified T cells. The data suggest, therefore, that the growth of memory T cells may depend on soluble factors released by, or contact with, adjacent T cells whereas the growth of naive T cells is independent of these additional signals. Since memory cells grew poorly in response to both IL2 and IL4, signals delivered by cell-to-cell contact rather than just cytokines appear to be the more likely requirement. Regardless of the exact requirements for the growth of memory cells, the data demonstrate that once activated by immobilized anti-CD3, naive cells have an enhanced capacity to grow under limiting dilution conditions in the presence of IL2. The finding that naive T cells have an enhanced capacity to grow under limiting dilution conditions may have physiologic importance. Thus, during primary antigenic challenge, when the precursor frequency of responding cells is low, it is important to expand the population of antigen responsive cells. Moreover, it may be necessary that these cells grow in the absence of adjacent activated T cells. In contrast, during a secondary immune response when the frequency of antigen responsive cells is high, expansion of memory cells may be less critical and also likely to occur in an environment containing other activated T cells. Previous studies have demonstrated that CD8+ T cells like CD4+ T cells can be subdivided based on the expression of CD45RA, CD45R0, CD29, and LFA-3 (34). Moreover, CD8+, CD45RAS cells, like CD4+, CD45RA+ cells, lose CD45RA expression with activation. Finally, the subset of CD8+ cells with CTL activity following allogeneic stimulation do not arise from the CD45RA+ population. The data suggest that these markers may also be useful in separating CD8f T cells into naive and memory subsets. The current studies demonstrate that, like the corresponding CD4+ subsets, the vast majority of the anti-CD3-stimulated IL2 responsive cells reside within the subset of cells that does not express CD29, CD45R0, and LFA-3 or the CD8+ naive T cell subset. The finding that the requirements for the growth of naive and memory CD8+ T cell subsets are similar to those of the corresponding CD4+ subsets supports the conclusion that these may represent comparable functional subpopulations of T cells. The current studies demonstrate that naive and memory T cells can both respond to immobilized anti-CD3. Moreover, they demonstrate that under conditions which favor cellular cooperativity and an optimal stimulus, the number of cells entering the cell cycle, the initial DNA synthesis, and the absolute growth of the cells stimulated by immobilized anti-CD3 is comparable, whereas under conditions which limit cellular cooperativity, naive cells exhibit an enhanced ability to grow. The differential capacity of memory and naive T cells to grow in the absence of cellular cooperativity may be important in facilitating their respective roles in the immune response.

REFERENCES I. Kung, P., Goldstein, G., Reinherz, E. L., and Schlossman, S. F., Science 206, 347, 1979. 2. Reinherz, E. L., Kung, P. C., Goldstein, G., and Schlossman, S. F., J. Immunol. 123, 2894, 1979. 3. Reinherz, E. L., Kung, P. C., Goldstein, G., and Schlossman, S. F., Proc. Natl. Acad. Sci. USA 76,406 I, 1979. 4. Reinherz, E. L., Kung, P. C., Breard, J. M., Goldstein, G., and Schlossman, S. F., J. Immune/. 124, 1883, 1980. 5. Reinherz, E. L.. Kung, P. C., Goldstein, G., and Schlossman, S. F., J. Immunol. 124, 1301, 1980.

218

GEPPERT

AND

LIPSKY

6. Morimoto, C., Letvin, N. L., Boyd, A. W., Hagan, M., Brown, H. M., Komacki, M. M., and Schlossman, S. F., J. Irnmunol. 134, 3762, 1985. 7. Morimoto, C., Letvin, N. L., Distaso, J. A., Aldrich, W. R., and Schlossman, S. F., J. Irnmunol. 134, 1508, 1985. 8. Rudd, C. E., Morimoto, C., Wong, L. L., and Schlossman, S. F., J. Exp. Med. 166, 1758, 1987. 9. Sanders, M. E., Makgoba, M. W., Sharrow, S. O., Stephany, D., Springer, T. A., Young, H. A., and Shaw, S., J. Immunol. 140, 1401, 1988. IO. Smith, S. H., Brown, M. H., Rowe, D., Callard, R. E., and Beverley, P. C., Immunology 58, 63, 1986. 11. Tedder, T. F., Cooper, M. D., and Clement, L. T., J. Immunol. 134,2989, 1985. 12. Sanders, M. E., Makgoba, M. W., and Shaw, S., Immunol. Today 9, 195, 1988. 13. Gerli, R., Bertotto, A., Spinozzi, F., Cernetti, C., Grignani, F., and Rambotti, P., Clin. Immunol. Immunopathol. 40,429, 1986. 14. Byrne, J. A., Butler, J. L., and Cooper, M. D., J. Immunol. 141, 3249, 1988. 15. Clement, L. T., Yamashita, N., and Martin, A. M., J. Immunol. 141, 1464, 1988. 16. Akbar, A. N., Terry, L., Timms, A., Beverley, P. C., and Janossy, G., J. Immunol. 140, 2171, 1988. 17. Huet, S., Bournsell, L., Dausset, J., Degas, L., and Bernard, A., Eur. J. Immunol. 18, 1187, 1988. 18. Gerli, R., Bertotto, A., Crupi, S., Arcangeli, C., Marinehi, I., Spinozzi, F., Cernetti, C., Angelella, P., and Rambotti, P., J. Immunol. 142, 2583, 1989. 19. Sanders, M. E., Makgoba, M. W., June, C. H., Young, H. A., and Shaw, S., Eur. J. Immunol. 19, 803, 1989. 20. Geppert, T. D., and Lipsky, P. E., J. Clin. Invest. 81, 1497, 1988. 21. Vine, J. B., Geppert, T. D., and Lipsky, P. E., Cellular Immunol. 124, 2 12, 1989. 22. Sanchez-Madrid, F., Krensky, A. M., Ware, C. F., Robbins, E., Strominger, J. L., Burakoff, S. J., and Springer, T. A., Proc. Natl. Acad. Sci. USA 79, 7489, 1982. 23. Hanson, J. A., Martin, P. J., Beatty, P. L., Clark, E. A., and Ledbetter, J. A., In “Leukocyte Typing” (A. Bernard, L. Bournsell, J. Dausett, L. Milstein, and S. F. Schlossman, Eds.), p. 195. SpringerVerlag Publications. Berlin, 1984. 24. Perussia, B., Trinchieri, G., Jackson, A., Warner, N. L., Faust, J., Rumpold, H., Kraft, D., and Lanier, L. L., J. Immunol. 133, 180, 1984. 25. Moreno, J., and Lipsky, P. E., J. Clin. Immunol. 6, 9, 1986. 26. Wysocki, L. J., and Sato, V. L., Proc. Natl. Acad. Sci. USA 75, 2844, 1978. 27. Herzenberg, L. A., and Herzenberg, L. A., In “Handbook of Experimental Immunology” (D. M. Weir, Eds.), p. 22 1. Blackwell Scientific Publications, Oxford, 1978. 28. Taswell, C., J. Immunol. 126, 1614, 1981. 29. Darzynkiewicz, Z., Traganos, F., Sharpless, T., and Melamed, M. R., Proc. Nat/. Acud. Sci. USA 73, 2881, 1976. 30. Salmon, M., Kitas, G. D., and Bacon, P. A., J. Immunol. 143, 907, 1989. 31. Wasik, M. A., and Morimoto, C., J. Immunol. 144, 3334, 1990. 32. Geppert, T. D., and Lipsky, P. E., J. Immunol. 138, 1660, 1987. 33. Patel, S. S., Duby, A. D., Thiele, D. L., and Lipsky, P. E., J. Immunol. 141, 3726, 1988. 34. Yamashita, N., and Clement, L. T., J. Immunol. 143, 15 18, 1989.