Regulation of the class of immune response induced by antigen

CELLULAR

IMMUNOLOGY

Regulation

81,

345-356 (1983)

of the Class of Immune Response

Induced by Antigen

I. Specific T Cells Switch the in Vivo Response from a Cell-Mediated to Humoral Mode

P. A. BRETSCHER Department of Immunology University of Alberta, Received

March

and MRC Group Edmonton, Alberta

on Immunoregulation, T6G 2H7. Canada

29, 1983; accepted

June 2. I983

Unprimed murine spleen cells, when administered intravenously to irradiated recipients together with antigen for 7 days, are induced to display either DTH reactivity or to mount a humoral (IgM and IgG) response. The class induced depends on the number of spleen cells given to the irradiated host. A low number of cells does not support the induction of any response, a medium number only gives rise to substantial DTH reactivity, whereas a high number only mounts a humoral (IgM and IgG) response. Observations show that the higher number of T cells in a large inoculum of spleen cells, compared to the number present in a medium one, is responsible for the absence of DTH reactivity and the mounting of a humoral response. This finding suggests that the induction of DTH precursor cells may occur when fewer antigen-specific helper-T-celldependent signals are generated than the number of signals required to induce B-cell precursors of the IgM and IgG classes.This possibility is favored by further observations. The administration of in situ irradiated, primed helper T cells to mice reconstituted with a medium number of normal spleen cells, results both in the specific suppression of the DTH response that occurs in the absence of these primed cells and in the mounting of a humoral response.

INTRODUCTION Observations in the literature show that the conditions under which an antigen induces delayed-type hypersensitivity are different from those required to induce IgM and IgG humoral responses. For example, the generalization has been made that antigens with few foreign sites, either by virtue of their small size or of being slightly modified self-antigens, induce only cell-mediated responses and are nonimmunogenic for a humoral response (1). Some of the variables of administering an antigen with many foreign sites that are known to affect the class of response induced are the maturity of the recipient (2), the route by which the antigen is administered (3), the nature of the adjuvant employed (4), the time examined after immunization (5) and the dose of antigen given (2, 5). In addition, the past history of the animal with regard to the antigen in question is of importance; thus immunization leading to a humoral response can render animals unresponsive for the induction delayed-type hypersensitivity (DTH)’ (6-9) a phenomenon referred to as immune deviation, and animals ’ Abbreviations used: B cell, bone marrow- or bursederived cell; CRBC, chicken erythrocytes; DRBC, donkey erythrocytes; DTH, delayed-type hypersensitivity; PFC, plaque-forming cells; T cells, thymusderived cells; Thy- I, T-cell surface antigen. 345 000%8749/83 $3.00 Copyright All

rights

0 of

1983 reproduction

by Academic in any

Press.

Inc.

form

reserved

346

P. A. BRETSCHER

and cells immunized to display DTH reactivity are sometimes specifically unresponsive at the humoral level (2, 10). The purpose of the work reported here was the development of an in vivo system which allows an analysis of those cellular interactions responsible for determining the class of response an antigen induces when an unsensitized animal encounters an antigen. It is essential, in view of the phenomenon of immune deviation, to analyze responses mounted by unprimed cells when examining these interactions. It was also considered desirable to avoid the use of adjuvants in the absence of a detailed knowledge of how they function and to obtain a system in which the administration of antigen by the same route gives rise to the induction of DTH and humoral responses. The latter feature may allow one to analyze these cellular interactions without taking the complexities of lymphocyte traffic and lymphoid tissue organization into account. This paper describes an in vivo system with which the cellular interactions responsible for determining the class of response (i.e., DTH and IgM and IgG antibody) an antigen induces can be analyzed. The observations described show that the level of helperT-cell-dependent signals, generated under the particular conditions of immunization, determines the class of response induced and provides a means for specifically switching the class of response mounted, in an antigen-specific manner, from a cell-mediated to humoral mode. MATERIALS

AND

METHODS

Antigens and mice. CBA/CaJ mice, aged 6-18 weeks and bred at the University of Alberta, were used in this study. Donkey red cells (DRBC) were obtained from Colorado Serum Company and chicken red blood cells (CRBC) from one chicken (typed as B*I* for the B genetic region) housed at the University of Alberta. Preparation of cell suspensions and conditions for the adoptive transfer. Single-cell suspensions from the spleens of normal mice were prepared by the gentle passage of spleen fragments through a stainless-steel wire mesh under sterile conditions in the presence of Leibovitz medium (LM). After setting for 2 min, the supematant was harvested. The cells so obtained were washed twice in LM and the number of viable cells was determined using trypan blue exclusion. Single-cell suspensions from the spleens of irradiated, reconstituted mice were prepared in the same way, except that the LM contained approximately 10 wg/ml of beef pancreatic DNAase (Sigma) to prevent cell clumping (11). Recipient mice were irradiated at 850 R using a ‘?Zo source. They were reconstituted with varying numbers of spleen cells derived from normal donors and given either 0.1 ml of 10% CRBC or 0.1 ml of 50% DRBC within 2 hr of irradiation. In some cases, as described under Results, recipient mice received primed cells intravenously 24 hr before irradiation. The spleens of reconstituted mice were harvested 1 week later and their DTH and humoral responses determined as described below. Assays of humoral (IgA4 and IgG) and DTH responses. The humoral response to CRBC and DRBC was determined by enumerating the number of plaque-forming cells (PFC) using the standard plaque assay (12). A slightly lower density of CRBC was employed in the slide than customary as this gave rise to clearer PFC. Indirect PFC were obtained by adding an optimal amount of a rabbit anti-mouse Ig serum to enhance IgG PFC. The yield of cells from irradiated mice was between 200 and 400% 7 days after reconstitution with a low number of cells. When larger numbers of cells were used

T CELLS

SWITCH

RESPONSE

FROM

DTH

TO

HUMORAL

MODE

347

for reconstitution, the yield was sometimes less. The DTH reactivity of spleen cells from reconstituted mice was determined by injecting 10’ viable cells either with or without antigen into the hind footpads of normal mice, and the 24-hr antigen-dependent swelling measured as described elsewhere (13). The unit of swelling is 0.0 1 mm. The standard error of the differential swelling was calculated as the square root of the squares of the standard errors of the swelling observed with and without antigen. Previous work has shown that this assay of antigen-dependent swelling can be used as a quantitative measure of DTH reactivity. The 24-hr antigen-dependent swelling increases when increasing numbers of cells, specifically sensitized to display DTH reactivity, are injected into the footpad. A doubling in the number of sensitized cells administered led to an increase in antigen-dependent swelling of approximately 50% when the number of sensitized cells was varied over an eightfold range (13). The yield of cells from the irradiated mice was approximately linearly proportional to the input number. The swelling produced by lo7 cells from the different groups is thus a measure of the DTH reactivity obtained from the same number of input spleen and hence DTH precursor cells. It is therefore a measure of the normalized DTH response (see under Results). When the yield was relatively lower with larger inocula, the DTH reactivity, as measured by the antigen-dependent swelling produced by 10’ cells, is an overestimate of the relative normalized DTH reactivity. Even so, the DTH reactivity decreases as the size of the inoculum is increased above 10’ cells (see Results). Cell separation procedures. T cells were killed with the aid of the commercially available anti-Thy- 1 hybridoma antibody (New England Nuclear, anti-Thy-2, Lot FPA.019). Viable cells, at a concentration of 107/ml, were incubated on ice for 1 hr in the presence of the hybridoma antibody at a dilution of 104, and an equal volume of agarose-absorbed guinea pig complement ( 14) was added at a dilution of I/ 16. The cells were then incubated for 45 min at 37°C. Control cells were incubated under identical conditions except that the hybridoma antibody was not added. The cells were washed in LM and the number of viable cells was determined by trypan blue exclusion. Anti-Ly- 1.1 and anti-Ly-2.1 sera, kindly provided by Dr. I. F. C. McKenzie of Melbourne, Australia, were used at a fivefold higher concentration than that required to produce maximal lysis of T cells derived from normal spleen, by the same procedures used in the anti-Thy-l treatment described above. B cells, PFC, and macrophages were removed by passage through a nylon wool column (15) and the yield of viable cells was determined by trypan blue exclusion. RESULTS The DTH and Humorai Hosts

Responses Mounted

by Normal

Spleen Cefls in irradiated

Various numbers of normal spleen cells were injected intravenously into irradiated mice. In all cases, except where specified, the cells were injected with a constant amount of antigen per recipient. The number of mice per group was so chosen that at sacrifice, 7 days after transfer, the number of viable spleen cells harvested in each group was about 10’ cells. This number was needed to assay DTH reactivity. The number of direct and indirect PFC in the spleen of six individual mice, belonging to each group, was determined and the DTH reactivity for each group was measured with a pool of spleen cells as described under Materials and Methods.

348

P.

NUMBER

OF

NORMAL

A. BRETSCHER

SPLEEN

CELLS / RECIPIENT

x IO-’

I. The normalized DTH (unbroken line), IgM (dotted line), and IgG (broken line) responses mounted by various numbers of normal spleen cells to the antigen DRBC in the environment of an irradiated host 7 days after reconstitution. The scale for the IgG response is 20-fold higher than that for the IgM response. FIG.

Figures 1 and 2 show how the DTH and humoral responses depend on the number of spleen cells received by each recipient for the two antigens CRBC and DRBC. Both classes of responses are dependent on the recipient mice receiving antigen (unpublished results). The cells causing swelling are sensitive to anti-Thy-l treatment; swelling also requires the presence of antigen in the footpad of recipient mice (unpublished results). These findings confirm that the antigen-dependent swelling observed is a measure of DTH reactivity.

0 NUMBER

5

IO OF NORMAL

I5 SPLEEN

20

25

CELLS 1 RECIPIENT

30 x 1om6

2. The normalized DTH (unbroken line), IgM (dotted line), and IgG (broken tine) responses mounted numbers of normal spleen cells to the antigen CRBC in the environment of an irradiated host 7 days after reconstitution. FIG.

by various

T CELLS

SWITCH

RESPONSE

FROM

DTH

TO

HUMORAL

MODE

349

The size of the response in this adoptive transfer system depends in a complicated way on the number of normal spleen cells given to a recipient. In the absence of cellular interactions, the response per lo6 input cells, i.e., the normalized response, would be independent of the input number. In fact, the normalized response depends on the input number. The normalized response measures the effector activity (either cell-mediated in the form of DTH or humoral in the form of IgM or IgG PFC) obtained from the same number of precursors in inocula of different sizes and is thus a measure of the efficiency of precursor cell induction in these different inocula. The normalized responses for the two antigens are shown in Figs. 1 and 2. This kind of dependency of the response on the input cell number has been observed in more than 10 separate experiments. Three characteristics of the response are evident. (i) A smaller inoculum (less than 10’ cells) mounts a DTH response than the inoculum required (about 3 X 10’ cells) to obtain a large humoral response. (ii) The induction of a strong IgG response is associated with an absence of DTH reactivity, and vice versa. This system thus exhibits the usual exclusiveness between the induction of these different classes. (iii) The value of the normalized response initially increases with the size of the inoculum for both the DTH and humoral classes of immunity. This increase shows that the induction of both classes requires some form of cellular collaboration. Two other features of these observations are worth noting. The normalized IgG response regularly appears to be more sharply dependent on the input cell number than the IgM response, and this is probably due to the fact that it is more helper T cell dependent (16). Second, the dependency of the IgM response on the size of the donor inoculum is clearer for the antigen DRBC than for CRBC. This is likely to be due to the fact that the background response to CRBC is considerably higher than to DBRC (unpublished observations) and one would therefore expect there to be a larger population of partially induced IgM B cells specific for CRBC than for DRBC. This greater heterogeneity of the anti-CRBC B-cell population, with respect to the degree of background stimulation, may be responsible for the observation that the anti-CRBC IgM response is less sharply dependent on input cell number than the anti-DRBC IgM response. The Number of T Cells Determines the Class of Immune

Response Induced

Recent work from this and other laboratories has shown that the induction of DTH requires antigen-specific helper T cells (17, 18). Preliminary observations (J. Tucker and P. A. Bretscher) suggest that the low level of DTH reactivity, obtained on reconstituting a small number (2-3 X 106) of normal spleen cells from irradiated mice, is due to the limiting number of helper T cells present. The lack of help is likely to be the reason for the increased size of the normalized DTH response as the size of the inoculum of normal spleen cells is increased from a small (2-3 X 106) to a medium (5-7.5 X 106) number. A possible explanation for the dependency of class induced on the size of the donor inoculum is that a medium inoculum (5-7.5 X 106) contains a sufficient number of helper T cells to mount a DTH response, whereas a larger number of helper T cells is required to mount a humoral one. The observations in Table 1 test the idea that the class of response mounted by a large inoculum, as opposed to a medium one, is due to the higher number of T cells present in the large

350

P. A. BRETSCHER TABLE 1

The Higher Number of T Cells Present in a Large as Opposed to a Medium Inoculum of Unprimed Spleen Cells Is Responsible for the Humoral Response and Lack of DTH Reactivity Response* (PFC/106 input + SE) No. of unprimed cells given” (x10-6) 7.5 7.5 + 27.5 Anti-Thy-l treated 7.5 + 27.5 C’ treated 7.5 + 27.5 untreated

DTH f SE 78 84 22 24

+ 7 f 5 zk 3 f 4

IgM l7f 20& 117 f 107 +

kG 7 6 16 12

lo* 4 19f 4 750 + 146 973 + 44

’ Yield of cells after C’ treatment 82%, after anti-Thy-l treatment 52%. * Normalized response to CRBC (i.e., activity from the same number of input precursor cells), assuming C’ treatment kills no cells, and anti-Thy-1 treatment kills all T but no B cells.

inoculum. Irradiated mice were given either a medium inoculum of cells (7.5 X 106) that induces only a DTH response, a large inoculum (7.5 + 27.5 X lo6 = 35 X 106) that induces only a humoral response, or a medium inoculum together with a number of T-depleted normal cells to give a total inoculum equivalent to a large inoculum (7.5 X lo6 cells and 27.5 anti-Thy-l-treated cells). This latter group only mounts a DTH response, showing that it is the larger number of T cells present in a large, as opposed to a medium, inoculum of normal spleen cells that is responsible for the suppression of DTH and the induction of a humoral response. Further experiments were designed to characterize the mechanism of action of these T cells. The increase in the humoral and the decrease in the DTH response that occurs as the inoculum of cells is increased above 10 X lo6 cells could be due to the induction of antigen-specific T cells able to help the humoral and suppress the DTH response. Specific T Cells Derived from Mice That Are Mounting Only a Humoral Response Can Help a Humoral Responseand Concomitantly SuppressDTH

Several instances are known in which the activity of induced regulatory T cells is relatively radiation resistant (19, 20). This is most likely to be due to the fact that such cells do not have to divide and/or differentiate to express their biological activity. It therefore seemed reasonable to suppose that the spleens of mice, reconstituted with antigen and a large inoculum of normal cells that mounts only a humoral response, might contain radiation resistant cells able to help a humoral response and suppress the induction of DTH. The results in Fig. 3 show that cells, obtained from spleens of irradiated mice reconstituted 7 days earlier with 3.5 X IO’ spleen cells and antigen, can, when given to recipient mice 24 hr before irradiation, specifically cause a small inoculum to mount a humoral response with the concomitant suppression of DTH. Cells primed to CRBC allow a small inoculum to mount an anti-CRBC humoral response with the concomitant suppression of the induction of DTH to CRBC, and did not significantly affect the DTH and humoral responses to DRBC, and vice versa. The cells(s) responsible for switching the class are sensitive to anti-Thy-l treatment as shown in Table 2. The activity of these T cells is also partially abrogated by antiLy-1 and unaffected by anti-Ly-2 treatment as shown in Table 3.

T CELLS SWITCH Immune (CRBC or DRBC) cells gtven before wadmtlon

Group

RESPONSE FROM DTH TO HUMORAL

800 1

Immunlrlng Antigen

t 0

I

20

DTH

40 Umts

PFC / IO6 Input 400 I

600 I

of

I

60 Swellhng

3.51

MODE

80

200 /

0 I

1

100

FIG. 3. Cells derived from mice mounting a humoral response specifically switch the response, mounted by adoptive recipients, from a cell-mediated to humoral mode. Cells from the spleen of irradiated mice, reconstituted with 3.5 X IO’ normal spleen cells and either CRBC or DRBC, were harvested on Day 7. Cells (2 X 10’) were given to recipient mice 24 hr before irradiation as shown in the figure. Mice challenged with CRBC were reconstituted with 5 X lo6 normal spleen cells after irradiation, whereas those challenged with DRBC were given 8 X lo6 cells. Seven days later the DTH and humoral responses of mice belonging to the different groups were determined as described under Materials and Methods. K!, IgN PFC: LB, IgG PFC.

The Humoral Response Observed, When the Class Is Switched by SpeciJc T Cells, Is Due to the Induction of B Cells in the Normal and Not in the Primed and Irradiated Spleen Cell Population It is possible that the humoral response is due to B cells, present in the primed population, even though they are irradiated. An experiment was performed to test whether this possibility is likely. B cells, PFC, and macrophages were removed from half the primed population by passage through a nylon wool column (15), and the ability of the depleted cells to cause a switch in the class induced was compared to TABLE 2 T Cells Are Responsible for the Switch in Class Induced No. of normal cells given” (X 10-6)

PFC/106 input f SEb 2 X 10’ nylon-passed immune cells

8

-

8 8

Anti-Thy-l + C’ treated C’ treated

DTH f SE 43 + 3 41 +4 3rt3

I&f 196 k 28 354 + 32 1003 k 340

16 446 z!I 90 550 f 145 3164 +- 805

“Cells, from the spleen of irradiated mice 7 days after reconstitution with 3.5 X 10’ spleen cells and CRBC, were passed through a nylon wool column. Recovery of cells from the columns was 12%. Recipient mice were given the equivalent of 2 X 10’ immune cells that were treated with either guinea pig complement (yield 65%) or anti-Thy-l and complement (yield 20%). b Assayed 7 days after reconstitution.

352

P. A. BRETSCHER TABLE 3 The T Cells Responsible for the Switch in Class Induced are Ly-I+ and Ly-2.

No. of normal cells given” (x10-6)

2

7

X

PFC/106 + Se

IO7 Nylonpassed” cells

DTH + SE 68 26 66 29

-

1 7 7

Untreated Anti-Ly-I treated Anti-Ly-2 treated

w

t 3 -+ 4 -t 3 k 7

JgG

3Ok 9 1161t 3 45 f 13 83 -e 22

13tr 2 208 -+ 40 44 2 18 282 -+ 81

LICells, from the spleen of irradiated mice 7 days after reconstitution with 35 X IO* normal spleen cells and DRBC, were passed through nylon wool columns. Recovery of cells from columns was 20%. Cells were either untreated or anti-Ly- I. I or anti-Ly-2. I and complement treated before being given to mice 24 hr before irradiation and reconstitution with 7 X IO6 normal spleen cells and DRBC. The number of cells given per mouse before irradiation was derived from 2 X IO’ immune cells before passage through the nylon wool column.

that of the undepleted population. The results of such an experiment are shown in Fig. 4. In this case two different doses of primed, and primed and depleted, cells were given to separate recipients 24 hr before irradiation. Two features of these results are noteworthy. First, removal of B cells and PFC from the primed population did not PFC/ CRBC -Immune cells administered -*, before irradiation f ~10 a

1000 I

800 I

IO6 Input

400

600 I

I

jw

200 I

.‘.‘.‘.

2

I-

t

1I

I

0

20

40

DTH

Untts

of

,

60

1

80

Swelling

FIG. 4. Effect of passage through nylon wool columns on the activity of cells able to switch the response from a cell-mediated to humoral mode. Cells from the spleen of irradiated mice reconstituted with 3.5 X IO’ normal spleen cells and CRBC were harvested on Day 7. Half the cells were passed through nylon wool columns, and depleted and undepleted cells were given to recipients as indicated in the figure. Recipient mice were irradiated 24 hr later and given 5 X IO6 normal spleen cells and CRBC. The DTH and humoral responses of the recipient mice were assayed 7 days after irradiation. q , IgG PFC: k!?~IgM PFC.

T CELLS

SWITCH

RESPONSE

FROM

DTH

TO

HUMORAL

353

MODE

diminish their ability to help a humoral and suppress a DTH response but slightly enhanced (about threefold) these activities. This suggests that the humoral response observed is due to the induction of B cells in the small inoculum of unit-radiated normal cells. Second, the degree of suppression of DTH, and the degree of help for a humoral response, appear to go hand in hand, i.e., the more complete the suppression of DTH the larger the humoral response. It can be seen from the observations recorded in Fig. 4 that 7 X lo6 primed cells, yielding as few as lo6 cells after passage through nylon wool, have significant activity in switching the class induced when given to mice 24 hr before irradiation. It can be seen from the observations in Fig. 4, as well as those in Figs. 1 and 2, that the production of IgG antibody is more helper T cell dependent than is that of IgM. Kinetics of the DTH and Humoral

Response

Table 4 shows the kinetic appearance of the DTH and humoral responses when mice are reconstituted with antigen and either a medium or large inoculum of normal spleen cells. It can be seen that DTH reactivity is already present in the spleen of mice reconstituted with a medium but not a large inoculum 4 days after reconstitution and immunization, whereas a small humoral response is first detectable on Day 5 in mice reconstituted with a large number of cells. These observations suggest that the absence of DTH reactivity, in mice reconstituted with a large inoculum, does not require the production of antibody; the large inoculum displays insignificant DTH reactivity or PFC on Day 4 at the time when the small inoculum express DTH reactivity. This inference is consistent with other observations (22) as described under Discussion. DISCUSSION The in vivo experimental system described in this report was developed to allow an analysis of those interactions between unprimed lymphocytes that are responsible for determining whether an antigen induces a strong DTH or humoral response. Substantial antigen-dependent responses are induced in this system when antigen is administered intravenously to mice without adjuvant, and different conditions give TABLE Kinetics

of DTH

PFC/l@

No. of normal cells/recipient y (x10-6) 5

35

and Humoral of Normal

4

12*

3 (0)

862 18 (3 + 3) 79 + 14 (134 + 36)

a Percentage yields for medium inoculum: 4. 19; Day 5, 38; Day 6, 116; Day

Responses Mounted by Medium Spleen Cells in Irradiated Hosts

and High Numbers

input IgM (IgG) on dayb

5

33 k 14 (5 2 3)

4

6 40+ (47 t

DTH 7

15 9)

136k 28 (1042 f 340) Day 4 was 34; Day

63+ (20 +

10 13)

116 + 27 (1044 + 214)

on day’

4

5

6

45 + 5

75 f 4

77 + 6

78-r-

15 +- 3

18 f4

II +4

17 2 8

5, 54; Day 6, 153; Day

7, 236; for large inodum:

7, 113.

b Six mice per groupper day immunized to CRBC. ’ Mean antigen-dependent swelling + SE in units of 10d2 mm produced

by the same number

of input

cells.

7 15

Day

354

P. A. BRETSCHER

rise to the induction of different classes of response. This system displays the inverse relationship between the induction of cell-mediated and humora] responses that is usually observed when animals are immunized with inert antigens administered without adjuvant. Three modes of response are apparent when lethally irradiated mice are reconstituted with different numbers of unprimed spleen cells but the same antigen challenge. A low number of cells does not support the induction of either a DTH or humoral response, a medium number supports only a substantial DTH response, whereas a large number supports only a humoral response. Observations are presented to show that the larger number of T cells, present in a large inoculum as compared to a medium one, is responsible for suppressing the cell-mediated response and giving rise to the induction of a humoral one. The mechanism of action of these cells was further analyzed. Antigen, given to irradiated animals reconstituted with a large number of normal spleen cells, induces not only a humoral response but also specific T cells that coordinately suppress the induction of a DTH response and help a humoral one. These activities are demonstrated when cells from such mice are given to recipients 24 hr before irradiation and reconstitution with a medium number of cells. The DTH response the recipient mice would mount in the absence of the in situ irradiated primed cells is suppressed and a humoral response is induced. Treatment of the primed cells before administration to recipients with anti-Thy-l serum shows that the cells responsible for switching the class of response are T cells. Observations are also presented to show that they are antigen specific. These observations suggest that more helper-T-cell-dependent signals are required to induce B cells than the number required to induce DTH precursor cells. Evidence for an inverse relationship between the induction of strong DTH and humoral responses has been documented in a number of experimental systems (2, 3, 5-10, 2 l-23). Bianchi and colleagues have described observations showing that cloned T-cell lines, able to help a humoral response, also mediate specific DTH reactions (24). These authors suggest that one cell type has the function to mediate both DTH reactions and help the induction of B cells; they point out, however, that their interpretation is difficult to reconcile with the known inverse relationship between the induction of strong DTH and humoral responses. This paradox rests on the two assumptions that only one cell type is present in their T-cell lines, and that such cells always express DTH and helper function coordinately. Neither of these assumptions is beyond reasonable doubt. Metcalf and Unman have shown that a single B cell, on appropriate immunogenic stimulation, can produce progeny secreting IgM and IgG antibody (25). This observation by itself is quite consistent with others showing that different conditions of immunization can favor the induction of these different classes of humoral response, and that most cells secrete antibody of only one class. These observations suggest that one B cell can give rise to different classes of progeny that synthesize different classes of immunoglobulin. Similarly, it is possible that a single T cell may give rise to progeny expressing different effector functions or, less likely, to one class of cell with two activities that can be separately regulated. In this case, the observations of Bianchi et al. would not be inconsistent with the well-known inverse relationship between the induction of cell-mediated and humoral responses or the observations described here. The mechanism(s) by which the DTH response is abrogated during the course of a humoral one has been the subject of a considerable amount of investigation. It is

T CELLS

SWITCH

RESPONSE

FROM

DTH

TO

HUMORAL

MODE

355

sometimes suggested that the absence of DTH reactivity during a humoral response may require the production of antibody or antibody-antigen complexes. Antibody can inhibit the induction of DTH, but the question here is whether it does so under circumstances that allow a humoral response to take place. Observations from several systems are difficult to reconcile with this possibility. Thus T cells, obtained from mice mounting a humoral response, were shown to specifically suppress the induction or expression of DTH, whereas the Ig+ cells were unable to do so, as was early immune or hyperimmune serum (21). Furthermore, such suppression of DTH can occur without the production of antibody in B-cell-depleted mice (2 1). Observations obtained in the system described here also suggest that antibody production is not required to inhibit DTH. Irradiated mice reconstituted with a large inoculum of spleen cells express insignificant DTH or humoral immunity 4 days after reconstitution, whereas those reconstituted with a medium inoculum already express substantial DTH reactivity at this time. This suggests that the lack of DTH reactivity does not require the production of antibody. A successful attempt has been made to obtain an in vitro analog of the in vivo observations described here (P. A. Bretscher, J. Zmmunol., in press). Three modes of response are observed when unprimed cells are cultured with the same antigen concentration but at different cell densities. A low density does not support the induction of any response, a medium density supports a transient IgM and substantial DTH response, whereas a high density only supports a humoral response. Of particular interest here is that a low density, on being supplemented with irradiated primed T cells, results in the induction of DTH. These cells act by the linked recognition of antigen in helping the induction of DTH (17). Preliminary experiments also suggest that mice given primed cells, before irradiation and reconstitution with a low number of normal spleen cells, mount a DTH response. The primed cells act in an antigenspecific manner (Tucker and Bretscher, unpublished observations). These results suggest that mice reconstituted with a low number of cells do not mount DTH response due to an insufficiency of helper T cells. The observations described in this report provide in vivo evidence that the number of helper-T-cell-dependent signals generated determines whether a cell-mediated or humoral response is mounted; fewer such signals are required for the induction of a cell-mediated than for a humoral response. This interpretation is supported by the development of a procedure for specifically switching the class of response induced from a cell-mediated to humoral mode by providing more specific helper T cells. The in vivo system described here should provide a means for ascertaining whether the same class of T cell can both specifically suppress the induction of DTH and help in the humoral response and to characterize the properties of these cells. The activities of the primed T cells, able to switch the response from a cell-mediated to humoral one, appear to be almost completely abograted by anti-Ly-1 and barely affected by anti-Ly-2 treatment (see Table 3) confirming earlier work characterizing the markers on T cells able to help B cells (26) and those able to suppress DTH responses (23). These observations are consistent with the possibility that both functions are mediated by the same class of T cell. ACKNOWLEDGMENTS I am grateful to Mr. M. Dhalla for his help in this work. Canada, to the MRC Group on Immunoregulation.

This work

was supported

by a grant

by MRC,

356

P. A. BRETSCHER

REFERENCES 1. 2. 3. 4. 5.

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

Pearson, M. N., and Ralfel, S., J. Exp. Med. 133, 494, 1977. Parish, C. R., Transplant. Rev. 13, 35, 1972. Lagrange, P. H., Ma&mess, G. B., and Miller, T. E., J. Exp. Med. 139, 528, 1974. Freund, J., and McDermott, K., Proc. Sot. Exp. Biol. Med. 49, 548, 1942. Salvin, S. B., J. Exp. Med. 107, 109, 1958. Asherson, G. L., and Stone, S. H., Immun. 9, 205, 1962. Battisto, J. R., and Miller, J., Proc. Sot. Exp. Biol. Med. 111, 1I 1, 1962. Crowle, A. T., and Hu, C. C., Gin. Exp. Immunol. 1, 323, 1966. Loewi, G. E., Holborow, E. J., and Temple, A., Immunology 10, 339, 1966. Ramshaw, I. A., Woodsworth, M., Wright, K., and McKenzie, I. F. C., Cell. Immunol. 125, 197, 1980. Keller, G., Longenecker, M., and Diener, E., J. Cell. Physiol. 102, 35 1, 1980. Cunningham, A. J., and Szenberg, A., Immunology 14, 599, 1968. Bretscher, P. A., Eur. J. Immunol. 9, 3 11, 1978. Cohen, A., and Schlesinger, M., Transplantation (Baltimore) 10, 130, 1970. Julius, M. H., Simpson, E., and Herzenberg, L. A., Eur. J. Immunol. 3, 645, 1973. Cheers, C., and Miller, J. F. A. P., J. Exp. Med. 136, 1661, 1972. Tucker, M. J., and Bretscher, P., J. Exp. Med. 155, 1037, 1982. Leung, K. N., and Ada, G. L., J. Exp. Med. 153, 1029, 198 1. Pilarski, L. M., J. Exp. Med. 145, 709, 1977. Kettman, J., and Dutton, R. W., Proc. Natl. Acad. Sci. USA 68, 669, 1971. Ramshaw, I. A., Bretscher, P. A., and Parish, C. R., Eur. J. Immunol. 6, 674, 1976. Ramshaw, I. A., Bretscher, P. A., and Parish, C. R., Eur. J. Immunoi. 7, 180, 1977. Ramshaw, I. A., McKenzie, I. F. C., Bretscher, P. A., and Parish, C. R., Cell. Immunol. 31, 364, 1977. Bianchi, A. T. J., Hooijkaas, H., Benner, R., Tees, R., Nordin, A. A., and Schrier, H. M., Nature (London) 290,62, 198 1. Metcalf, E. S., and Klinman, N. R., J. Exp. Med. 143, 1327, 1976. Vadas, M. A., Miller, J. F. A. P., McKenzie, I. F. C., Chism, S. E., Shen, F. W., Boyse, E. A., Gamble, J. R., and Whitelaw, A. M., J. Exp. Med. 144, 10, 1976.