Differential interleukin secretion by in vitro activated human CD45RA and CD45RO CD4+ T cell subsets

CELLULAR

IMMUNOLOGY

141, 10-20 (1992)

Differential lnterleukin Secretion by in Vitro Activated Human CD45RA and CD45RO CD4+ T Cell Subsets JOANA MARIA FERRER, ARESIO PLAZA, MIGUEL KREISLER, AND FERNANDO D~Az-ESPADA .%wicio de Inmunologia,

Clinica Puerta de Hierro, Madrid,

Spain

Received May 24, 1991; accepted November 20. 1991

The CD45RA and CD45RO isoforms have been reported to define complementary subsets among CD4+ T cells: CD45RA CD4+ T cells are considered “virgin T cells” and CD45RO “primed T cells.” We investigated the secretion of lymphokines by human CD4+ CD45RO and CD4+ CD45RA T helper cells after mitogen stimulation. CD45RA and CD45RO CD4+ T cells were isolated by negative immunoselection using magnetic beads.CD45RO cells, but not CD45RA cells, proliferate well in responseto pokeweed mitogen (PWM) or insoluble antXD3. Both subpopulations produced interleukin (IL)-2, K-6, and interferon (IPN)-7 when stimulated with PWM for l-4 days. Only Day 1 supematants from CD45RO cells contained moderate amounts of IL4. After 14 days of continuous culture and stimulation with PWM, the CD45RA subset had lost the expression of CD45RA and gained that of CD45RO. When long-term cultured CD45RA or CD45RO cells were treated with insoluble anti-CD3, they incorporated [‘Hlthymidine at similar levels, but only CD45RO cells secretedIL-4 and significantly increased their secretion of IFN-7. Thesedata indicate that despitephenotype conversion,the two subpopulations maintain functional differencesin the secretionof lymphokines, thus suggestingthat circulating CD45RA and CD45RO cells may represent different lines of differentiation. 0 1992 Academic PB. Inc.

INTRODUCTION Mouse helper T cell clones have been divided into TH 1, which secretesIL-2 and interferon (IFN)-7, and TH2, which secretesIL-4 and IL-5. This has led to the assumption that the activation of one or another subset in vivo may determine whether a humoral or cellular responseis elicited (I). Heterogeneity among normal mouse T helper lymphocytes has also been demonstrated. A monoclonal antibody against a subset of CD45 (T200) molecules that requires the expression of the second variable exon can be usedto separateCD4 T cells into two subpopulations. The brightly stained subpopulation produces IL-2 but not IL-4, while the weakly stained population produces IL-4 but not IL-2 (2). Human CD4+ T cell clones show a more heterogeneous pattern of lymphokine secretion, making it very difficult to subdivide human T cell clones on the basis of their lymphokine secretion (3-6). Functional differencesamong human CD4+ T cells have been described and related to the expression of two isoforms of the CD45 molecule. These isoforms are expressed by distinct types of cells, and their differential expression is regulated by alternative splicing of three exons from a single RNA precursor (7). CD4+ T cells expressing 10 0008-8749192$3.00 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

ACTIVATION OF CD4+ T CELL SUBSETS

11

CD45RA and CD45RO isoforms have been proposed to represent “virgin” cells and “memory” cells, respectively (8- 11). The in vitro long-term activation of CD4+ T cell subpopulations results in the expression of different sets of interleukins encoding mRNAs and in the induction of a phenotypic conversion from CD45RA into CD45RO (9, 12, 13). Despite this phenotypic conversion, the acquisition of functional properties resembling those of the CD45RO subpopulation remains controversial. In this study we investigated the patterns of IL-2, IL-4, IL-6, and IFN-7 secretion by in vitro-activated CD45RA and CD45RO subpopulations. These lymphokines are considered to be representative products of distinct T cell subsets. We found that despite phenotypic conversion, differences in IL-4 and IFN--y secretion exist between the two subpopulations, which could be considered as separate lineages of differentiation. MATERIAL

AND METHODS

Reagents. The monoclonal anti-CD45RO antibody UCHLl (14) was kindly provided by Dr. Beverley (ICRF, London) and anti-CD45RA, 1 1 l- 1C5 ( 15) was kindly provided by Dr. Vives (Hospital Clinico, Barcelona). Other monoclonal antibodies used were specific for CD23 (B6; Coulter Clone, UK), CD8 [clone B9.4.2. (16)], CD3 [clone SPVT3b. (17)], and CD14 (M02; Coulter Clone, UK). Pokeweed mitogen (PWM) was purchased from GIBCO Laboratories (Grand Island, NY) and phytohemagglutinin (PHA) was obtained from Difco Laboratories (Detroit, MI). Recombinant IL-2 was obtained from Behringwerke (Germany) and recombinant IL-4 was obtained from Amersham (UK). Alternatively, IL-2 was obtained from the MLA- 144 ( 18) supernatant after ammonium sulfate precipitation. T and B cell preparations. Mononuclear cells were separated on Ficoll-Hypaque (Lymphoprep, Nycomed AS, Norway) from the peripheral blood of healthy volunteers. The cell-containing fraction was collected and rosetted with neuraminidase (Behringwerke, Germany)-treated sheep red blood cells (SRBC), and rosetting and nonrosetting fractions were separated after Ficoll-Hypaque centrifugation. T cells were obtained from the rosetting fraction after lysis of the SRBC with distilled water. CD4+ T cells were obtained by treating the T cells with anti-CD8 and anti-MO2 monoclonal antibodies for 30 min at 4°C followed by lysis with a rabbit complement (Sera Lab, UK) for 45 min at 37°C. CD8- T cells were loaded in a nylon column (Type 200L; Du Pont, Wilmington, DE) and eluted with RPMI-10% fetal calf serum (FCS). Cell preparations contained less than 1% monocytes and 2% CD8+ cells. B cells were obtained from the nonrosetting fraction after depletion of adherent cells by a 1-hr incubation in plastic petri dishes at 37°C in RPM1 containing 20% FCS. Nonadherent cells were treated with anti-CD3 and anti-MO2 monoclonal antibodies for 30 min at 4°C followed by lysis with a rabbit complement for 45 min at 37°C. CD45RA and CD45RO cell preparations. CD4+ T cells (25 X 106) were incubated with anti-CD45RO (UCHLl) or anti-CD45RA (11 l-lC5) for 20 min at 4°C. Cells were washed twice in buffered salt solution (BSS)-1% FCS, resuspended in 150 ~1 of medium, and incubated with 150 ~1 of a suspension of sheep anti-mouse IgG immunobeads (Dynabeads; Dynal, Norway) for 20 min at 4°C. The samples were diluted with BSS- 1% FCS and positive cells were removed with a magnet (Dynal, Norway). The remaining antigen negative cells were concentrated by centrifugation, and a second cycle of negative selection was performed. Cell culture. The cells were cultured in RPM1 1640 (Microbiological Associates Bioproducts, Bethesda, MD) supplemented with 10% FCS (Sera Lab. UK), 25 mM

12

FERRER

ET AL.

Hepes,200 mML-glutamine (Flow Laboratories) 5 X 10-5M2-mercaptoethanol (Sigma Chemical Co., St. Louis, MO), and antibiotics (penicillin/streptomycin; Difco). CD45RA or CD45RO CD4f T cells (5 X 104) were cultured in 0.2 ml of culture medium in the absenceor presenceof either PWM ( 1%) or PHA ( 1%)in 96-well flatbottom plates for 72 hr. For anti-CD3 treatment, culture wells were precoated by overnight incubation at 4°C with protein A-Sepharose-purified anti-CD3 MoAb (5 pg/ml) diluted in carbonate/bicarbonate buffer (pH 9.5). The cultures were pulsed with 1 &i/well of [3H]thymidine for the last 18 hr, and incorporation of radioactivity was determined by liquid scintillation counting after cell harvesting on fiberglasspaper. Monocytes (5%) were added to the culture when the stimulus was PWM or PHA. Immunoglobulin secretion was determined in B and CD4+ T cell subsetcocultures (5 X lo4 each) stimulated with PWM. The presenceof IgM and IgG antibodies in 9day supernatants was determined by an ELISA method. Briefly, 96-well Dynatech ELISA plates were preincubated overnight at 4°C with 1 @g/ml of affinity-purified class-specificgoat anti-human antibodies (Sigma) in carbonate/bicarbonate buffer (pH 9.5). The plates were washedwith PBS-0.05%Tween 20 and the samples(supematants diluted in PBS-O.1% BSA-0.05% Tween 20) were added to the wells and incubated for 2 hr at 4°C. The plates were rinsed and incubated for an additional 2 hr with peroxidase-conjugated affinity-purified goat anti-human IgM or IgG (Sigma). After washing the plates, substrate o-phenylenediamine dihydrochloride (Sigma) was added and the plates were read in a multichannel spectrophotometer (TiterTek; Flow) at 492 nm. T cell supernatants were obtained from 3 X lo6 cells/ml cultures stimulated with PWM in the presence of monocytes after several periods or after restimulation with insoluble anti-CD3 after 14-day cultures in the presenceof PWM. In 14-day cultures, rIL-2 (10 U/ml) was added at Days 3 and 10 of culture. Flow cytometry. Cells (2 X 105)were incubated with the corresponding monoclonal antibodies for 30 min at 4°C. Cells were washed three times in BSS-1% FCS and incubated with a 1:80 dilution of fluoresceinated goat anti-mouse Ig (Kallestad) for 30 min at 4°C. Cells were washed and analyzed in an EPICS PROFILE (Coulter). Cytokine analysis. The presenceof IL-2 wasdetermined by the induction of CTLL2 proliferation (19). CTLL2 cells (5 X 103/well) were incubated with dilutions of test supernatants, control IL-2, or human IL-4 in 96-well plates for a total of 24 hr. The cultures were pulsed with [3H]thymidine for the last 6 hr. The CTLLZ cell line responded to human IL-2, but not to human IL-4. The presence of IL-4, IL-6, and IFN-~ was determined by an ELISA detection system (Intertest-4/6/IFN-y; Genzyme). The calibration curves were linear from 0.2 to 3 rig/ml (IL-4) 160 to 2500 pg/ml (IL-6) and 200 to 3200 pg/ml (IFN-y); 0.1 ng/ ml is the minimum concentration of IL-4 detectable in a biological assay of B cell costimulation with insoluble anti-IgM reagents(data not shown). In some experiments, induction of the expression of the CD23 antigen on resting B cells was also used as a detection assay for IL-4. Resting B cells (2 X 105/well) were incubated for 48 hr at 37°C in U-shaped 96-well plates with 50% test supernatants, 100 U/ml rIL-4, or 10 U/ml rIL-2. Expression of CD23 was determined by flow cytometry, as indicated. Fractionation of T cell supernatants. PWM-stimulated supernatants of CD4+ cells were obtained at Day 1, concentrated 1O-fold by ultrafiltration (Amicon PM- 10 membranes), and fractionated by molecular exclusion chromatography in a FPLC Pharmacia 10 X 30 Superosa 12 column, equilibrated with RPMI-0.1% PEG 6.000 at a

ACTIVATION OF CD4+ T CELL SUBSETS

13

flow rate of 0.5 ml/min. Fractions (0.5 ml) were collected and FCS (10%) was added before performing the CTLL2 assay, as described. Molecular weight markers were human IgG (150 kDa), BSA (68 kDa), OA (44 kDa), and cytochrome c (12.5 kDa). RESULTS Purijcation of CD45RA and CD45RO CD4+ T cell subsets. CD4+ T cells from normal individuals display a heterogeneous pattern of expression of CD45 isoforms. An example corresponding to two donors is given in Fig. lA, where a high proportion of cells from individual 1 coexpresses CD45RA and CD45R0, whereas donor 2 shows a normal distribution of expression of CD45RA and CD45RO. Only donors expressing equivalent numbers of each isoform were selected for further studies. After negative selection with immunobeads (Fig. lB), CD45RA- cells express high levels (92%) of CD45RO antigen, while only a small percentage stains weakly with 11 l-lC5 (MoAb against CD45RA antigen). On the other hand, the CD45RO- population shows high levels of expression of CD45RA antigen (95%) and only a few cells stain weakly with UCHLl MoAb detecting CD45RO antigen. The percentage of the contaminating CD45RO’” cells ranged from 4 to 11% in a series of four different experiments. Proliferation and cooperation in Ig secretion of CD4-k T cell subsets. In order to ascertain the properties of the cell subsets used in this study, CD45RA and CD45RO subpopulations were purified and cultured for 72 hr in the presence of PWM, PHA, or insoluble anti-CD3 (Table 1). Monocytes (5%) were added when the stimulus was PWM or PHA. CD45RA cells do not proliferate in response to insoluble anti-CD3 and display a lower response to PWM than CD45RO cells, which show a high degree of proliferation in response to both insoluble anti-CD3 and PWM. When CD45RA and CD45RO subpopulations were cocultured for 9 days with B cells in the presence or absence of PWM (and 5% monocytes), only the CD45RO

C

DowAwl

log

fluorescence

intensity

FIG. 1.Distinct patterns of CD45RA and CD45RO expression in CD4+ T cells from two different donors (A) and control of purification of negatively selected CD45 RA- and CD45RO- CD4+ T subpopulations (B). Cells were stained by indirect immunofluorescence with the CD45RA MoAb 11 l-lC5 (- - -) and the CD45RO MoAb UCHLl (-). Binding was detected with FITC-conjugated goat anti-mouse IgG and cytofluorographic analysis. Solid area corresponds to cells stained with secondary reagents alone. Numbers refer to percentage of positive cells, over the control.

FERRER ET AL.

14

TABLE 1 Proliferation of CD4+ T Cell Subsets Expt 1 Expt 2

CD45RA+ a CD45ROt CD45RA+’ CD45ROt

2.0 f 4.0 f 1.0 f 2.0 2

0.1 l’ 2.0 o.26 0.1

PWM

PHA

icuCD3C

252 1 15f 4 22+ 3 290 f 16

406 iz 42 185 f 18 540 +- 13 748 f 93

35* 7 202? 27 lo* 2 1260 + 103

dCell populations used were CD45RAt and CD45ROt CD4+ T cell subsets. b Results represent mean cpm X lo-’ f SD of triplicate cultures measured after 72 hr of incubation. c Cells were stimulated with PWM or PHA in the presenceof 5% monocytes or with plastic immobilized anti-CD3 MoAb (iolCD3).

subpopulation was capable of cooperating with B cells in the secretion of Igs (seeTable 2). We must note that the lack of cooperation in the CD45RA subpopulation cannot be explained by a lack of responseof these cells to PWM. These results indicate that despite the small contamination of CD45R” cells in the CD45RA preparation, the in vitro cell responseswere identical to those expected for CD45RA and CD45RO subsets(11). Lymphokines produced by short-term stimulated CD4i T cell subsets. CD45RA and CD45RO subpopulations were activated with PWM for several days to elicite cytokine release. IL-2 is detected by the CTLL2 assay in the supernatants obtained from stimulated CD45RA or CD45RO cells (Table 3). This activity is greatly reduced after a 4-day period in cultures of CD45RO cells, possibly due to interleukin uptake by the growing cells. CD45RA cells are also able to secretesubstantial amounts of IL2, despite their low proliferation in responseto PWM. When concentrated supematants of 1-day cultures are fractionated by Superosa12 molecular sieving (Fig. 2), the CTLL2inducing activity elutes as a single peak corresponding to 15- 19 kDa, in accordance with the expected size for interleukin 2. Only the CD45RO subset secretes(at Day 1) moderate amounts of IL-4, as assessedby an ELISA method (Table 4). IL-6 was detected only in Day 1 supernatants and was similar for the two subpopulations while the secretion of IFN-7 was observed after longer periods of incubation (Day 3). The CD45RO subsetsecretedhigher amounts of IFN-7 than the CD45RA subset(Table 5). Effect of long-term stimulation of CD4 T cell subsets: Phenotypic and functional changes. When CD4+ T cell subsets are stimulated for 14 days in the presence of TABLE 2 B Cell Differentiation Induced by CD4+ T Cell Subsetsin the Presenceof PWM

W/W M/ml) B t CD45RA+n B + CD45RO+

700160b 140170

PWM 870/180 80100/15790

’ Cultures consisted of purified B cells (5 X 104/well) mixed with CD45RAt or CD45RO+ CD4t T cell subsets(5 X 104/well) in the presenceof 5% monocytes. bSecretedIgM and IgG were measured in the supernatant fluid after 9 days of culture. Values represent the mean of triplicate cultures.

ACTIVATION

15

OF CD4+ T CELL SUBSETS TABLE 3

IL-2 Secretion by CD45RA and CD45RO Cells0 CD45RA

SN 1/8b SN l/l6 SN l/32 U/mid

Dl

D4

261’

219 265 235 5.0

187 107 2.2

CD45RO D15

Dl

11

D4

266 220 119 2.3

17

7 0.02

23

5 9 0.08

D15 6 7 I
’ CD45RA or RO cells were cultured for 1 (Dl) or 4 (D4) days in the presence of PWM or for 14 days in PWM plus 1 day with olCD3 (D15). Supematants were assayedon CTLL-2 cells (5 X lo3 cells/well). b Supematant dilution. ’ [‘H]Thymidine uptake in cpm X IO-*. Basal proliferation of CTLL-2 was less than 700 cpm; 1 U/ml of rIL-2 gives 3 X 104.rIL4 (100-1000 U/ml) gives 700-1000 cpm. d Units of IL2/ml.

PWM, rIL-2, and 5% monocytes, they evidence changes in the expression of CD45 isoforms. Cytofluorometric studies show that the CD45RA subpopulation losesthe high level expression of CD45RA antigen and gains that of CD45RO after 14 days of culture (Fig. 3). These results are similar to those recently described (9, 12, 13). CD45RO cells show an increase in the intensity of expression of CD45RO antigen. Long-term stimulated cells incorporate [3H]thymidine when stimulated for 48 hr with insoluble anti-CD3 and pulsed with [3H]thymidine for the last 6 hr (Table 6). Proliferation increases if rIL-2 is added to the culture, although this effect can be explained by the persistence of the IL-2 receptor in long-term activated T cells. In

25

0 6

9

10

11

12

13

14

15

16

17

16

Ve ml FIG. 2. Gel filtration on a Superosa 12 FPLC column of the concentrated (X 10) supernatant from PWMtreated CD4+ T cells. The column was equilibrated with RPMI-0.1% PEG. FCS (10%) was added to each fraction before IL-2 (CTLLZ assay) was determined. The downward arrows indicate the elution positions of the MW standards.

FERRER ET AL.

16

TABLE 4 IL-4 Secretion by PWM-Activated CD4+ T Cell Subsets IL-4 (rig/ml)

CD45RA+” CD45RO+

Dl

D4

o.oo* 0.40

0.22 0.16

’ CD45RA+ and CD45RO+ T cell subsetswere cultured in the presence of PWM and 5% monocytes. The presenceof IL-4 was analyzed in Day 1 (Dl) or Day 4 (D4) supernatants. bResults are given as rig/ml determined by an ELISA method.

fact, IL-2 addition alone induces an increasein [3H]thymidine incorporation. Although CD45RA cells did not proliferate in responseto insoluble anti-CD3 (Table I), 14-day precultured cells incorporate [3H]thymidine after this treatment at levels similar to those found in CD45RO cells. Supernatants obtained after a 24-hr restimulation of long-term stimulated cells with insoluble anti-CD3 contained only small amounts of IL-2 and no IL-6 (Tables 3 and 5). Restimulated CD45RO cells secreteIL-4 and considerable amounts of IFN-7, as determined either by an ELISA method (Table 5) or, in the caseof IL-4, by the ability of the supernatants to induce the expression of CD23 antigen on resting B cells (Fig. 4). CD45RA subpopulations do not secreteIL-4 when cultured under similar conditions, and no increase in the secretion of IFNq was observed if compared with the Day 3 secretion in primary cultures. DISCUSSION Functional and phenotypic differences have been observed in human CD4+ subpopulations, according to severalcriteria. CD45RO CD4+ T cells (UCHLI +) respond to recall antigens and provide help for B cell differentiation in PWM-driven cultures, and are thus considered a “primed” population containing “memory” T cells. On the other hand, CD45RA CD4+ T cells do not respond to recall antigens but they show a high response to several mitogens, such as Con A or PHA, and are considered “virgin” T cells ( 11, 20-23). TABLE 5 IFN-7 and IL-6 Secretion by CD45RA and CD45RO Cells Dl IFN-7 CD45RA+” CD45RO+


D3

D15

IL-6

IFN-,y

IL-6

IFN-7

IL-6

160 700

340 950

<150
300 2200


r7CD45RA or RO cells were cultured for 1 (Dl) or 3 (D3) days in the presenceof PWM or for 14 days in PWM plus 1 day with cKD3 (D15). b Results are given as pg/ml determined by ELISA methods. The detection limits of the assaywere 150 pg/ml for IL-6 and 100 pg/ml for IFN-y.

ACTIVATION

OF CD4+ T CELL SUBSETS

17

CD451vI
--MY 0 -MY 14

e f

CD45M

” .,.‘. t.,;:..‘. .:

n

:

., ,p’ “I.. ..,+ (,,, ‘. ....

U

log

fluorescence

intensity

FIG. 3. PWM-induced changesin CD45RA and CD45RO expression. Negatively selectedCD45RA+ and CD45RO+ CD4+ T lymphocytes were analyzed at Day 0 (broken lines), stimulated with PWM, and subsequently analyzed at Day 14 (continuous line) as described under Materials and Methods.

In this study we have shown that both CD45RA and CD45RO subpopulations secrete substantial amounts of IL-2, but not IL-4, when stimulated with PWM for 1 or 4 days. In keeping with the results of others (24), CD45RO cells produce more EN-7 than CD45RA cells. The pattern of IL-6 secretion is similar in both subpopulations. The requirement for additional signals, such as those recently described in anti-CD3-treated CD45RA cells (25), can explain the low level of proliferation found in PWM-treated CD45RA cells rather than an intrinsic unresponsiveness of this type of cells. In fact, that small degree of proliferation can explain the limited uptake of lymphokines and the persistence of IL-2 in 4-day supernatants of PWM-activated CD45RA cells. The lack of cooperation of the CD45RA subpopulation in the Ig secretion induced by PWM cannot be explained solely by a different pattern of lymphokine secretion upon a first stimulation. The ability of CD45RO cells to help B cells may reside in the induction of certain adhesion molecules after activation (26), which should confer on CD45RO cells the capacity to participate in the cognate activation of B cells. The description of human CD4+ T cell subsets, defined by CD45 MoAbs, differs in some regards from that reported in their putative mouse counterparts, in which differences in effector functions (humoral vs cellular responses),rather than the state of priming, are adscribed to each subset (1). In humans, the distinction between CD45RA and CD45RO subpopulations has received considerable support from experiments showing that phenotypic conversion from CD45RA to CD45RO is elicited in vitro after prolonged mitogen activation (9). Moreover, this phenotypic conversion is accompanied by the induction of mRNA for IL-4, which is normally associated with CD45RO cells (12, 13). These data have been suggestedto imply the existence of a functional conversion between the two subpopulations related to priming. Some data indicating that only T cells primed by contact with relevant non-B antigen-pre-

FERRER ET AL.

18

BCEUCD23 0msIow. -

SND15

lx 0.1

b U.16

CD45JM SNDi5 cD45BDJ”2-\

IL-4 ngh’

2!a 2,4t

FIG. 4. IL-4 secretion by long-term restimulated CD4+ T cell subsets.CD45RA+ or CD45RO+ CD4+ T cell subsetscultured for 14 days in the presenceof PWM were stimulated for 24 hr with insoluble antiCD3. IL-4 in the supernatants (SN) was determined either by its ability to induce CD23 expression on resting B cells or by an IL-4 ELISA method. Negative controls include the effect of RPM1 alone (---) and 10 U/ml of rIL-2 (IL-2). Positive control includes the effect of 100 U/ml of rIL-4.

senting cells can cooperate with B cells (27) could explain the suggestedproperties (primed or memory cells) attributed to the CD45RO subpopulation. Nevertheless, some discrepancies need to be explained. For instance, if CD45RA and CD45RO subpopulations display different effector functions, virgin and memory cells must be present in both subpopulations. Moreover, the genetically determined lack of CD45RA- CD4+ T cells in healthy individuals renders questionable the significance of the loss of CD45RA in the acquisition of memory functions (28). More importantly, bidirectional interconversion between both subpopulations of CD4+ T cells has been described in viva (29). Recent reports indicate that CD45RA+ and CD45RA- subsets maintain their original functions (suppression/help in immunoglobulin synthesis), despite prolonged activation and phenotypic conversion (30). We show in this paper that after a 14-day stimulation with PWM in the presence of exogenous IL-2, the CD45RA cells expressCD45RO antigen instead of CD45RA antigen, but that they are unable to secreteIL-4 after further stimulation. On the other hand, long-term cultured CD45RO cells expresshigher amounts of CD45RO isoform and secreteIL-4 after restimulation with insoluble anti-CD3. Our data indicate that despite phenotype conversion (CD45RA + CD45RO) after prolonged activation, secretion of IL-4 only occurs in activated (original) CD45RO cells. Moreover, if compared to unrestimulated cultures (Days 1 and 3) secretion of EN-7 increases in restimulated CD45RO cells, but maintain similar levels in the CD45RA subpopulation (Table 5). It is important to note that although uncultured CD45RA cells do not

ACTIVATION

OF CD4+ T CELL SUBSETS

19

TABLE 6 Proliferation of Restimulated CD4+ T Cell Subsets

CD45RO+’ CD45RA+

-

iaCD3

iaCD3 + IL-2

IL-2

1.5 2.8

33 f 2b 23 k 5

108 f 9 61 ?3

44 f 0.5 9 + 0.6

’ Cell populations used were CD45RA+ and CD45RO+ CD4+ T cell subsetsprecultured for 14 days in the presenceof PWM, 5% monocytes, and rIL-2. b Results represent mean cpm X lo-* + SD of triplicate cultures measured after 24 hr of incubation with plastic immobilized anti-CD3 MoAb (icuCD3) and/or 10 U/ml rIL-2 (IL-2). The wells were pulsed with [)H]thymidine for the last 6 hr.

respond to insoluble anti-CD3 (Table l), long-term activated cells proliferate after this treatment at levels comparable to CD45RO cells (Table 6). Assuming that IL-4 mRNA is present in both long-term activated CD45RA and CD45RO cells (12, 13), our data concerning IL-4 secretion suggestthat important differences persist between the activated products of the two subpopulations. It is likely that CD45RO cells derived in vitro from CD45RA precursors differ from activated “original” CD45RO cells, at least in their requirements for the secretion of IL-4, which can reflect differences in the signals required for translation/secretion, similar to those initiated through certain T cell surface molecules (CD28) (3 1). In any case,phenotypic conversion in vitro does not produce equivalent cellular products, suggesting that circulating CD45RA and CD45RO subsetsmay represent independent lineages of differentiation. It is possible that circulating CD45RO CD4+ T cells may not evolve from previously virgin CD45RA cells, but rather represent a distinct subpopulation. This subpopulation of “naive” CD45RO cells requires in vitro activation (priming?) to acquire new functions, represented by changes in the set of interleukins secreted(IL-2 + IL-4). It is evident from our work that a further heterogeneity is present within the CD45RO subset (CD45RA or CD45RO derived), which is in keeping with recently describedphenotypic and functional data (26, 30). ACKNOWLEDGMENTS The authors thank Luis Alvarez for the cytofluorometric analysis and Martha Messman for her editorial assistance.

REFERENCES I. Mosmann, T. R., and Coffman, R. L., Annu. Rev. Immunol. I, 145, 1989. 2. Bottomly, K., Luqman, M., Greenbaum, L., Carding, S., West, J., Pasqualini, T., and Murphy, D. B., Eur. J. Immunol. 19, 617, 1989. 3. Palliard, X., De Waal Malefijit, R., Yssel, H., Blanchard, D., Chretien, I., Abrams, J., De Vries, J., and Spits, H., J. Immunology 141, 849, 1988. 4. Maggi, E., Del Prete, G., Macchia, D., Parronchi, P., Tiri, A., Chretien, I., Ricci, M., and Romagnani, S., Eur. J. Immunol. 18, 1045, 1988. 5. Mingari, M., Moretta, A., Maggi, E., Pantaleo, G., Gerosa, F., Romagnani, S., and Moretta, L., Eur.J. Immunol. 14, 1006, 1984. 6. Umetsu, D. T., Jabara, H. H., De Kruyeff, R. H., Abbas, K. A., Abrams, J. S., and Geha, R. S., J. Immunol. 140,4211, 1988. 7. Thomas, M. L., Annu. Rev. Immunol. 7, 339, 1989.

20

FERRER ET AL.

8. 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. 9. Akbar, A., Terry, L., Timms, A., Beverley, P. C. L., and Janossy,G., J. Immunol. 140, 2 171, 1988. 10. Serra, H. M., Krowka, J. F., Ledbetter, J. A., and Pilarski, L. M., J. Immunol. 140, 1435, 1988. 11. Sanders,M. E., Makgoba, M. W., and Shaw, S., Immunol. Today 9, 195, 1988. 12. Salmon, M., Kitas, G. D., and Bacon, P. A., J. Immunol. 143, 907, 1989. 13. Bettens, F,, Walker, C., Gauchet, J. F., Wyss, T., and Pichler, W. J., Eur. J. Immunol. 19, 1569, 1989. 14. Terry, L. A., Brown, M. H., and Beverley, P. C. L., Immunology 64, 331, 1988. 15. Cobbold, S., Hale, G., and Waldman, H., Leucocyte typing. III. White cell differentiation antigens. Oxford University Press,p. 378, 1987. 16. Malisen, B., Rebai, N., Liabeuf, A., and Mawas, C., Eur. J. Immunol. 12, 739, 1982. 17. Spits, H., Yssel, H., Leewemberg, J., and DeVries, J. E., Eur. J. Immunol. 15, 88, 1985. 18. Rabin, H., Hopkins, R. F., Ruscetti, F. W., Nebauer, R. H., Brown, R. L., and Kawanaki, T. G., J. Immunol. 127, 1852, 1981. 19. Gillis, S., Ferm, M. M., Ou, W., and Smith, K. A., J. Immunol. 120, 2027, 1978. 20. Arthur, R. P., and Mason, D., J. Exp. Med. 163, 774, 1986. 21. Morimoto, C., Letvin, N. L., Boyd, A. W., Hagman, M., Brown, H. M., Komacli, H., and Schlossman, S. F., J. Immunol. 134, 3762, 1985. 22. Morimoto, C., Letvin, N. L., Distaso, J. A., Aldrich, W. R., and Schlossman, S. F., J. Immunol. 134, 1508, 1985. 23. Moore, K., and Nesbitt, A. M., Immunology 61, 159, 1987. 24. Dohlsten, M., Hedlund, G., Sjogren, H., and Carlsson, R., Eur. J. Immunol. 18, 1173, 1988. 25. Horgan, K. J., Van Seventer, G. A., Shimizu, Y., and Shaw, S., Eur. J. Immunol. 20, 1111, 1990. 26. Tedder, T. F., Matsuyama, T., Rothstein, D., Schlossman,S. F., and Morimoto, C., Eur. J. Immunoi. 20, 1351, 1990. 27. Lassila, O., Vainio, O., and Matzinger, P., Nature 334, 253, 1988. 28. Schwinzer, R., and Woniget, K., J. Exp. Med. 171, 1803, 1990. 29. Bell, E. B., and Sparshot, S. M., Nature 348, 163, 1990. 30. Rothstein, D. M., Sohen, S., Daley, J. F., Schlossman,S. F., and Morimoto, C., CeN. Immunol. 129, 449, 1990. 31. June, C. H., Ledbetter, J. A., Linsley, P. S., and Thompson, C. B., Immunol. Today 11, 211, 1990.