Immunological Rev. (1978), Vol. 40 Published by Munksgaard, Copenhagen, Denniark No part may be reproduced by any process without written permission from the autbor(s)

The Regulation of Lymphocyte Functions by the Macrophage EMIL R. UNANUE I. INTRODUCTION

We summarize here part of the work done in our laboratories on the role of macrophages in immune induction- We will not analyze the extensive literature on the subject but will refer only to key pertinent papers. Other contributors to this volume will present their results and interpretations of the macrophage helper role. We are dividing our studies into three sections: antigen presentation for antibody production and T cell stimulation, secretion of lymphostimulatory molecules, and lastly, regulation of T cell difierwitiation. It is our view, one also shared by others, that macrophages play a key crucial helper role in immune induction. The evidence gathered, in particular during the past few years, argues strongly for a fundamental role of phagocytes in immune stimulation, antigen recognition, and in the control of lymphocyte proliferation and differentiation. This evidence has, to a great extent, been derived from tissue culture studies which admittedly must be interpreted with caution but which nevertheless are invaluable for examining aspects of cellular reactions that are impossible to study in vivo. We expect that the information derived from tissue culture experiments will help us to probe the in vivo process in a more meaningful way. The helper role of phagocytes comes as a result of various interrelated functions: 1) their ability to remove and destroy extracellular antigen and, therefore, to deplete the extracellular milieu of molecules that have the potential for inducing negative responses in the lymphocytes; 2) their capacity somehow to present antigen to lymphocytes and, therefore, start the recognition and immune inductive process; and 3) their capacity to release molecules that can regulate lymphocyte functions. These three functions qualify the macrophage as a central cell determining how much antigen the lympho-

Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115, U.S.A.

228

EMIL R. UNANUE

cyte recognizes, and in what form, and insuring that the lymphocyte response is a positive one. These functions of the phagcKytes are deeply interrelated and are affected by several variables: structure and amount of antigen and whether they are under Ir gene control, state of maturation or difierentiation of the cells, etc. The role of macrophages in immune induction, in our view, stems from the way in which immunity developed during phylogeny (Calderon & Unanue 1975). The phagocytes represent the descendents of the primitive phagocytic cells of invertebrate species which carried out the essential elimination of foreign material. Specific immunity developed as a function necessary to complement and make more efficient the essential phagocyte function of antigen uptake and elimination, at the same time insuring the lack of immunity to self. Thus, during evolution, lymphocytes 'learned' that those antigens taken by macrophages were of importance, were pathogenic, should be recognized as foreign, and eliminated; while, in contrast, those molecules of antigen found in the extracellular milieu did not require positive induction. Thus, the function of lymphocytes was to recognize macrophage-associated antigens, to insure regulatory function which, in a way, amplified macrophage function, and, at the same time, to suppress recognition of molecules which were not macrophage associated.

n. ANTIGEN PRESENTATION FUNCTION

Antigen presentation studies have been carried out in two basic systems: one in which antibody formation is assayed and requires T and B cell function; a second in which T cell recognition and stimulation is being monitored, in many cases in the absence of B lymphocytes. The results of both systems show some differences that are worth emphasizing since these differences may well pcrint to two different mechanisms of antigen presentation. Early Studies - Antigen Presentation for Antibody Production in vivo and in vitro: In our early experiments with Dr. B. A. Askonas in London we developed a system to assay macrophage function (Unanue & Askonas 1968a, 1968b)- This system was identical to tliat employed by N. A. Mitchison (1969) and Gallily & Feldman (1967). It consisted of obtaining populations of macrophages (peritoneal cells obtained after peptone or thiogly coll ate injections), pulsing these, usually with radioactive antigen, and then assaying the immime response to this antigen associated in live macrophages. Up till this time, most research on macrophages had utilized extracts of these cells to examine immunogenicity. The provocative experiments of Fishman, Adler,

LYMPHOCYTE REGULATION BY MACROPHAGE

229

and coworkers (Fishman & Adler 1963, Adler et al. 1968) had been among the most exciting, calling attention to a crucial role of macroph age-extracted antigen in lymphocyte recognition via an inununogenic fragment bound to RNA. The role of RNA and RNA-antigen complexes (Gottlieb 1969) from macrophages remains imexplained. Whether these molecules which are operative in particular assay systems are really functional in in vivo or in vitro assay with live macrophages has never been established and remains questionable (Roelant et al. 1971). In our early studies, immunogenicity was assayed in vivo by measuring antibody responses in mice recipients of syngeneic macrophages bearing a known amount of antigen - or in irradiated mice transplanted both with macrophages and lymph node cells. The results were unequivocal: antigen associated with live macrophages was immunogenic, by defiinition capable of producing a strong immune response. The strong immunogenicity of a macrophage-bound antigen was particularly evident when studying weak antigens like serum proteins. For example, albimiins bound to macrophages were 1,000- to lC,000-fold more immunogenic than when given in soluble form (Mitchison 1969, Spitznagel & Allison 1970, Schmidtke & Unanue 1971). In analyzing the system, several points became apparent: live macrophages were better than dead macrophages; cell contact was required; and, furthermore, the strong immunogenicity of macrophage-associated antigens was found with all kinds of antigen whether particulate or soluble, large or small proteins. Our revi^v of 1972 examined more than a dozen difEerent antigens assayed in this system (Unanue 1972).

DNP-KLH

100

1000

ttg DNP-

Figure 1. This experiment compares the in vitro antibody response to DNP-KLH presented to DNP-KLH-primed spleen cells in three forms; macrophage bound (DNP-KLH macrophages), soluble, or adherent to the dish. From: Katz & Unanue (1973).

230

EMIL R. UNANXJE

The same results were later carried out in an in vitro Systran (Katz & Unanue 1973). The macrophages were pulsed for a few minutes, this time with a haptenated protein, dinitrophenylated hemocyanin (DNP-KLH), then incubated with DNP-KLH-primed spleen cells for 4 days, after which the number of antibody-forming cells was assayed by a plaque method. Figure 1 shows a comparison of immunogenidty of the DNP-KLH molecules, free, macrophage-botmd, or adherent to a culture dish. Clearly, DNP-KLH bound to macrophages was highly immunogenic. Also, it was apparent that macrophages would not substitute for T lymphocytes, as would be expected. Similar restilts had been previously obtained in vivo, transfering macrophages with antigen into thymectomized recipients (Unanue 1970). Thus, the system required the recognition of both hapten and carrier determinants by B and T cells, respectively. In subsequent experiments (this time assaying in vivo), we showed that, in fact, both hapten and carrier determinants associated with the macrophages had to be present in the same molectile (Unanue & Katz 1973). Thus, the classical observation that for immunogenidty two foreign determinants had to be included in the same molecule (Rajewsky et al. 1969, Mitchison 1971) applied also to the macrophage-bound antigens. Other studies indicated that DNP-KLH was immunogenic when bound either directly to the macrophage stu^ace or as part of an immune complex (Katz & Unanue 1973). A further point of great interest concerned the relationship between soluble antigen and macrophage-boimd antigen. The immune response, in vitro to hapten-carrier conjugates, was highly sensitive to the dose of soluble antigen which depended, in part, upon the state of priming of the mouse. A high dose of antigen (varying from 10 /./g to 0.1 //g), in fact, suppressed the immune response. We found that administration of antigen bound to macrophages produced a strong response that was quenched by addition of free, soluble antigen (Katz & Unanue 1973). The experiment was interpreted as indicating two forms of antigen: an immunogenic-macrophage-bound - and a non-immunogenic one - represented by the soluble antigen. An excess of the latter overrode the immunogenic form of antigen. Results similar to this had been previously obtained by Spitznagel & Allison (1970) in in vivo systems: the priming of mice with macrophage-bound albumin was decreased by addition of soluble albumin. Two other points of interest were brought out in our experiments (Katz & Unanue 1973), as well as in further impublished experiments (Unanue & Kiely 1976). These examined whether allogeneic macrophages could present antigens to syngeneic primed T and B cells and whether other cells could substitute for macrophages. Addition of allogenic macrophages resulted in a strong response, wiiich was identical to that produced by syngeneic macro-

LYMPHOCYTE REGULATION BY MACROPHAGE

231

TABLE I Immune response of F-KLH spleen cells to antigen bound to syngeneic or allogenic macrophages Macroph age-bound antigen IgM PFC Strain A A A A

(H-2") (H-2») (H-2*) (H-2»)

C57BL/6 (H-2*') C57BL/6 CH-2f) C57BL/6 (H-2*>) C57BL/6 (H-2*')

C57Biy6 CH-2*')

Antigen

IgG PFC

Amount (ng)

F-KLH F-KLH F-KLH F-RGG

1 0.1

593 (500- 780) 274 (220- 420)

0.01 3

120 ( 60- 240) 60 ( 20- 120)

F-KLH F-KLH F-KLH F-RGG

1

906 (780-1,020)

87 ( 20- 160) 643 (800-1,220)

0.1 0.01

280 (120- 420) 140 ( 20- 220) 125 ( 20- 240)

1,020 (800-1,220) 526 (480- 600) 200 (120- 380)

73 ( 40- 100)

3

1,813 933 620 80

(980-2,520) (880-1,020) (480- 860) ( 40- 140)

F-KLH spleen cells were obtained from A/St mice immunized with 50 /^g of F-KLH twice. The spleen cells were depleted of macrophages and IO' of them added to dishes containing 2 X 10* peritoneal macrophages. The macrophages had been pulsed 1 h on ice with the antigen and then washed thoroughly to eliminate the molecules. The amounts of macrophage-bound antigen were calculated on the basis of uptake studies made with radioactive antigens. Results represent the average number of plaque-forming cells (PFC) of two to four cultures (in parentheses are the ranges of values). F-RGG = fiuorescein bound to rabbit IgG. Culture conditions were those of Mishell & Dutton (unpublished experiments of Unanue & Kiely 1976).

phages. The response was dependent on specific T cell recognition of carrier molecules and was not explained by an allogeneic type of effect. The unpublished experiments showing this point are presented in Table I. This result has become critical in view of other experiments indicating that T cell proliferation was very much restricted to recognition of antigen on macrophages from the same strain used initially for immunization (Rosenthal & Shevach 1973). This point is analyzed further. Lastly, we found that in culture antigen presentation to T-B cell populations could, to some extent, take place by other cells like fibroblasts if these were able to bind the antigen. T Cell-Macrophage Interaction Our studies examining interaction of purified T cells with macrophages - in absence of B cell function - have been restricted to a system in which we assay for production of mediators following a brief interaction between T cells and macrophages (Unanue et al. 1976, Farr et al. 1977). The mediator that we have assayed is a 15,000-dalton mitogenic protein (MP), the characteristics of which are described in the

232

EMIL R. UNANUE

next section- Suffice to say, at the present moment, that the production of MP is dependent on the interaction between immune T cells, macrophages, and antigen in our case, the intracellular facultative bacteria, Listeria monocytogenes. The MP is produced by macrophages but probably also by T cells. The system of interaction is important inasmuch as it involves the immune response to an intracellular facultative bacteria that reqtiires the development of T cell as well as macrophage activation for protective immtmity. The present experimental system, originally devised with J. Calderon and J.-M. Kiely (1976), now has been throroughly explored by A. G. Farr. It involves producing an infection in mice with live Listeria monocytogenes, obtaining the peritoneal exudate T cells a few days later, and mixing these with nonnal macrophages together with dead Listeria organisms. Upon their interaction, a number of mediators are secreted in a matter of just a few hours. The effective production of MP requires specific an^-Listeria T cells, macrophages, and Listeria organisms. The fascinating point of this system is that the interaction requires homology at the MHC of the species - the H-2 complex in the case of the mouse - between the immune T cells and the normal macrophages. This was shown by experiments represented in Table II and Figure 2. Using congenic strains of mice, it became apparent that homology at the I-A subregion between the T cells and the macrophages was required. Thus, in the experiment of PEL-^BIO A

p^giQ^g Qp HOMOLOCT

MACROPHAGES K J i A ] : ! K

i

D_

MITOGEHIC STIMULATION

BIO BIOG BIO A(4R)

+ -f

BI0A(5R)

- - - - , .

BI0A(6R) -

-

_

-

BIOA(I5R) 4- + ^. + BIOA(IBR) BIOAQR

., _ + + + 0 2 4 6 8 10 12 (4 16 / / J THrUIDINE INCORPmATION (cpm X W^l

Figure 2. This bar graph shows that the interaction between T cells and macrophages leading to the production of MP requires homology at the I-A region of H-2. The T cells came from B10.A mice immunized with Listeria monocytogenes. The macrophages were obtained from the different congenic BIO strains depicted in the graph. Both T cells and macrophages were cultured for a day, and then the amount of MP in the culture fluid was assayed on thymocytes. This experiment was reported by Farr et al. (1977).

LYMPHOCYTE REGULATION BY MACROPHAGE

233

TABLE n The T cell-macrophage interaction resulting in the generation of mitogenic protein is regulated by the H-2 complex in the mouse

Experiment

Listeria Tcell

Normal macrophage

(H-21) BALB/c BALB/c (H-2»') AKR AKR

BALB/c AKR AKR BALB/c

53,700 13,300 88,700 17,600 900

± ± ± ± ±

BIO B10.D2 B10.D2 BIO

25,400 4,700 22,200 3,200 400

± 800 ± 100 ± 2,500 ± 100 ± 10

BALB/c DBA/2 DBA/2 BALB/c

5,900 9,100 41,400 42,700 500

± 200 ± 600 ± 900 ± 1,900 ± 50

None (H-2'') BIO BIO (H-21) B10.D2 B10.D2 None (H-20) BALB/c BALB/c (H-2'J) DBA/2 DBA/2 None

Generation of mitogenic activity 1,200 2,000 3,200 1,800 90

We have assayed for MP in the mixtures of activated T cells and normal macrophages. T cells alone did not secrete MP. Mitogenic activity was assayed at 25 % v/v. Results represent counts per minute of incorporation of H'-thymidine (± SEM) (from: Farr et al. 1977).

Figure 2, MP was produced when T lymphocytes from BIO.A strain of mice were added to syngeneic macrophages (BIO.A) or to macrophages from strains BIO.A (4R), BIO.A (15R), or BIO.AQR, aU sharing I-A subregion; but very little was produced when added to strains BIO.A (5R) and BIO A (6R), which share the right-hand side of the MHC. When we first developed the experimental system, we were surprised to find that addition ol Lwfen'a-immune T cells to the normal macrophages, without addition of Listeria organisms, resulted in high production of MP. Subsequently, Doctor Farr has found that the explanation for these results was that a response was made to fetal proteins carried with the Listeria organisms. In assays where there is no exposure to fetal proteins (the fetal calf serum of the medium), there was a strong response to Listeria, clearly restricted by the I-A subregion of Ia. These results are in agreement with the extensive studies of Rosenthal and associates in the guinea pig (Rosenthal & Shevach 1973, Bardnski & Rosen-

234

EMIL R. UNANUE

thai 1977) and Schwartz & Paul (1976) and Erb & Feldman (1975) in the mouse showing the MHC control of T cell-macrophage interaction. Recently, Pierce and associates have reported MHC-restrictions in secondary antibody responses in vitro to antigens under Ir gene control (Pierce et al. 1976). The Immunogenic Moiety: Though extensive work carried out with macrophages pulsed with antigen has definitely established that macrophage-associated antigens are immimogenic and may play an essential role, particularly insofar as T cells are concerned, the question remains how a macrophage, which is a cell programmed for the uptake and elimination of antigens, is still able to present antigen molecules to lymphocytes. It is well known that the bulk of antigen molecules is taken up, internalized, and degraded by the lysosomal system. How, then, is it possible that macrophages, aside from antigen elimination, also present immunogenic moieties? This question has been analyzed extensively in our laboratories as weU as by Alan Rosenthal and his associates using different assays of lymphocyte functions. The results, although not comparable, are not mutually exclusive and indicate, to us, two pathways of antigen presentation by the macrophages. Assaying for antibody production, we examined whether some antigen molecules bound to macrophages could escape lysosomal digestion, at least for a period of time, and become available to lymphocytes. Most antigens first were bound to the macrophage membrane and then were interiorized in vesicles, and finally, underwent extensive lysosomal digestion- In exploring the metabolism of antigens and correlating with the immunogenicity of antigens, we found two possible ways in which proteins could escape extensive degradation and potentially become available to lymphocytes. One involved few molecules of antigen that remained surface-botind; the other, a few molecules of partially degraded molecules that were released from the cell. With an antigen like hemocyanin, for example, we foxmd a small number of membrane-bound molecules remaining on the membrane despite the interiorization of the bulk of molecules (Unanue et al. 1969, Unanue & Cerottini 1970). Removal of these surface-bound molecules by treatment with trypsin or with antibody decreased significantly the immunogenicity of macrophageassociated hemocyanin. Recent experiments reevaluating these points are shown in Table III. Thus, we speculated that interiorization of soluble antigen bound to macrophage surface was not 100 % complete and that the few molecules that were slowly interiorized served an antigen-presentation function. Furthermore, with antigens that were totally interiorized, one could also identify a small number of molecules escaping complete digestion and somehow reverting back to the extracellular milieu by a process of exocytosis

LYMPHOCYTE REGULATION BY MACROPHAGE

235

TABLE m Effects of trypsin on immunogenicity of macrophage-bound antigen Macrophage-bound antigen Fresh macrophages

0 1 ng 0.1 ng 0.01 ng

After 24 h with antigen

24 h antigen ^-trypsin

— — — —

— — —

0 1 ng 0.1 ng 0.01 ng

0 1 ng 0.1 ng 0.01 ng

IgM PFC

IgG PFC

60 320 187 87

( 40- 60) ( 80-680) (160-240) ( 80-100)

40 527 273 120

( 20- 60) (24(^940) (140-380) ( 60-140)

13 420 40 80

( 10- 40) (200-640) ( 10-120) ( 40-100)

13 680 100 53

( 5-40 (660-700) ( 60-180) ( 20- 80)

10 47 67 20

( 0- 20) ( 20- 60) ( 60- 80) ( C^- 40)

40 153 67 33

(40) (100-220) ( 40-100) ( 20- 40)

The spleen cells (10^) were obtained from A/St mice primed with F-KLH ia alum and cultured for 4 days with 2 X 10' macrophages. Three sets of peritoneal macrophages were added to the spleen cells. Set 1 (Column 1) consisted of macrophages freshly harvested, pulsed for 30 min with F-KLH, and washed; Set 2 (Column 2) consisted of macrophages pulsed the same way with F-KLH hut then cultured for 24 h, washed, and then used in the assay with spleen cells; Set 3 (Column 3) is the same as Set 2 except the macrophages were trypsinized just before addition of the spleen cells. This experiment shows that the immunogenicity of F-KLH in macrophages is relatively long lived and explained, in great part, hy trypsin-removable molecules (unpublished experiments of Unanue & Kiely 1976).

(Cruchaud & Unanue 1971, Calderon & Unanue 1974). Calderon examined this release of antigen at depth, being concerned with the possibility that the released antigen could come from dying macrophages. He found that the antigen was released continuously for long periods of time, although usually in small amounts. The antigen was in the form of partially fragmented molecules still able to bind antibodies. Furthermore, there was no relationship between the release of antigen and the release of a cytoplasmic enzyme, like lactate dehydrogenase, an indication that the antigen was not coming from lysed cells. The released antigen derived from intracellular sources. This was shown by exposing macrophages to antigen and allowing interiorization of most molecules; the few remaining on the cell surface-bound antigen mole-

EMH- R. UNANUE

236

cules were then removed by trypsin, therefore leaving antigen only inside the cell. This treatment did not reduce the release of antigen molecules (Figure 3). Finally, the released antigen, when recovered in the medium, was clearly as immunogenic as the native molecule. These results contrasted wtih those of Rosenthal and associates (Ellner & Rosenthal 1975, Ellner et al. 1977) where the fimctional antigen molecule from macrophages required for T cell prolferation was trypsin insensitive and not recognized by antibody. It would appear that the antigen presentation to T cells, a step somehow regulated by the MHC of the species, apparently involves some kind of ill-defined processing which eventually allows the immunogenic molecule to 'surface' for T cell recognition in a form that is no longer recognized by antibody, perhaps because the molecule has been extensively changed. This point has been elaborated upon by Rosenthal and associates and forms the essence of his review. The experiments of Erb & Feldman (1975) identifying an antigen-MHC immunogenic complex are, of course, also highly relevant in this regard. One critical issue is whether the

IN SUPERNATANT

20'

pg'"j-KLII IN SUPERNATANT

0-4

4-21

21-27

PERIOD OF INCUBATION (hrs)

Figure 3. This graph demonstrates the release of I^^-hemocyanin after uptake by macrophages. Dishes with macrophages were incubated with I^^-KLH for 1 h at 37" C, at which time a set was treated with trypsin in order to remove molecules bound to the membrane. Treated and control dishes after washing were then placed in culture for the times indicated; the culture fluids were examined for macromolecular I^-hemocyanin. Hemocyanin is released from the macrophages presumably from the interiorized molecules that escaped trypsin treatment (Calderon & Unanue 1974).

LYMPHOCYTE REGULATION BY MACROPHAGE

237

TABLE rV Comparison of two forms of macrophage-bound antigen MHC-restricted form

Non-MHCrestricted form

Ass&y

Usually, macrophages pulsed with antigen are added to T cells which are assayed for proliferation or secretion of mediators.

Macroph age-pulsed antigen is added to mixtures of T-B cells which are then assayed for a plaqueforming response

Duration of immunogenicity

Relatively short

Relative long-lived

Sensitivity to trypsin

Not sensitive

Sensitive

Abrogation by antibody

Not effective

Eflfective

Substitution by fibroblast

Not possible

Yes, to a great extent

Macrophage requirement in assay system

Strict

Less strict

This table compares data obtained from our laboratories with those obtained mostly from Rosenthal and associates. References are in the discussion in the text.

MHC gene products regulate a step in the handling of the antigen or whether the antigen becomes involved in physical association with the MHC product, or both- These points have not been resolved and need to be integrated with known involvement of the MHC of the species in T-B cell collaboration (Katz et al. 1973) and in interactions involving cytotoxic T cells (Zinkemagel & Doherty 1974). We are proposing (Table IV) two pathways of antigen presentation of macrophages to lymphocytes - one involves a molecule that is not extensively changed and which becomes accesible for relatively long periods of time; the second involves a molecule where its handling and/or presentation is somehow under the regulation or control of the MHC of the species. Not only should we reconcile the discrepant data on macrophage-antigen presentation on the two systems analyzed above but also the observations indicating that B cells recognize antigen molecules in their native state (Sela et al. 1967), while T cells may not discriminate between native and denatured forms of antigen (Gell & Benacerraf 1959, Ishizaka et al. 1974, Shirrmacher & Wigzell 1974). Yet both hapten and carrier determinants must be present in the same molecule for B-T cell interaction!

238

EMIL R. UNANXJE

We envision that the T cell recognizes antigen presented by macrophages in a special form - in the broad sense of the word, i.e. processed, associated with an MHC structure or whatever, but in a way in which an MHC genecontrolled stq5 has played a part The T cell then responds, proliferates, and/or makes mediators which influence other T cells, macrophages, or B cells. In the context of the T-B cell collaboration, which may take place so effectively with macrophage-associated antigen under no MHC control, we envision that the B cell first recognizes the native antigen remaining on the macrophage cell surface - as our experiments have indentified it - and then somehow processes it. The B cell becomes the cell responsible for handling antigen and presenting it to the T cell. This handling step by the B cell involves an MHC-controlled step. This scheme would explain the B cell recognizing native antigen via surface Ig, yet allowing, in a second step, the T cell to enter into interaction. This hypothesis requires that T cells may recognize antigens processed by either macrophages and/or B cells and that the MHCcontrolled step insofar as antigen handling and/or presentation is concerned could be exerted at the level of both macrophages and B cells. The experiments of Katz et al. (1973), showing that, with antigens under Ir gene control. Fx-primed T cells can only interact with DNP- primed B cells of the responder strain is compatible with the assumption that the antigen-handling step by the B cell is critical. This experiment resembles the same situation with antigen presentation of macrophages to T lymphocytes. Finally, we should note that both B cells and macrophages can process antigen in very similar ways - insofar as gross parameters of antigen handling are concerned (reviewed by Schreiner & Unanue 1976). One main difference is that, while the macrophage recognizes many antigens, the B cell recognizes only that with specificity for its surface Ig.

m . SECRETION OF LYMPHOREGULATORY MOLECULES

Macrophages secrete a number of molecules that can have a profound influence on lymphocyte functions. Only a few of these molecules have been characterized, although more and more information is becoming available concerning their biological mode of action and their mechanisms of secretion. It is now recognized that macrophages are highly secretory cells (reviewed in Unanue 1976). Through secretion, phagocytes can exert considerable influence on their extracellular environment. The secretory function of phagocytes may be as important as their function of intracellular handling of foreign materials. In fact, both functions are deeply interrelated since the major stimulus for secretion is the process of uptake and phagocytosis of a

LYMPHOCYTE REGULATION BY MACROPHAGE

239

TABLE V Lymphoregulatory activities found in macrophage culture fluids Molecule

Biochemistry

Effects

Mitogenic protein (LAF)

Approximately 15,000 daltons trypsin-resistant molecule

Increases DNA synthesis in thymocytes, to lesser extent in T and B cells, increased T cell helper/suppressor acUvity (?)

B cell-differentiating molecule

Two activities: 140,000 and 15,000

Differentiates primed B cell to antibody-secreting cells in absence of antigen T cells

T cell-activating molecule

Same as MP plus a heterogenous group of 40,000 to 60,000 molecules

Increases in helper/suppressor cell activity, depending upon state of priming

Thymic-differentiating factor (TDF)

Approximately 40,000

Differentiates immature thymocytes to mature T lymphocytes

This table shows only the lymphostimulatory molecules studied in our laboratory.

particle. Molecules secreted by macrophages include complement proteins, lysosomal enzymes, neutral proteases, interferon, lysozyme, etc. A few of these molecules are secreted continuously, but most others are secreted following the uptake of particles. Some, like the plasminogen activators, for example, are only secreted following activation of the macrophages and stimulation by phagocytosis (Gordon et al. 1974). Thus, secretion is influenced by the state of maturation of the phagocytes as well as by their function at a particular time. The lymphoregulatory molecules secreted by macrophages which we have been studying are outlined in Table V. Macrophage culture fluid contains the MP - mitogenic protein - referred to before, an activity that promotes B cell difierentiation (Calderon et al. 1975) and a thymic-differentiating factor discovered by D. I. Beller (Beller & Unanue 1977). Less well characterized is a factor that causes angiogenesis (Polverini et al. 1977) and another that produces lymphocyte chemotaxis (Ward et al. 1977). Antigen fragments, as analyzed before, are also released. We have not analyzed the antigen-containing molecule bearing MHC determinants discovered by Erb & Feldman. The molecules shown in Table V all restilt in lymphocyte stimulation. However, macrophage culture fluids also contain some less well-characterized

240

EMIL R. XJNANUE

inhibitory molecules. Some of these are of macromolecular nature and have been found only in high-density cultures of macrophages (for example see Chen et al. 1977). Of great interest among the inhibitory molecules secreted by the macrophage are nucleosides, which we will analyze before the lymphostimulatory molecules. Secretion of Nucleosides: Macrophages secrete a nujnber of nucleosides into their extracellular milieu. These include the purine nucleosides as well as thymidine (Stadecker et al. 1977). Although their release can result from catabolism of pliagocytized dead cells, as originally reported by Opitz et al. (1975), it is of considerable interest that they are also released as a result of synthesis and secretion. When analyzing the effects of macrophage culture fluids on proliferation of various cells, Calderon and I confirmed previous studies (Nelson 1973, Waldmann & Gottlieb 1973, Ulrich 1974, Kasahara & Shiori-Nakano 1976) that showed macrophage fluids containing materials that inhibited tritiated thymidine incorporation by the dividing cells (Calderon et al. 1974). The macrophage fluids also contained an easily dialyzable molecule that inhibited cell growth. The inhibition of cell growth was conspicuous on EL-4 thymic leukemia but was not apparent on other dividing cells. The EL-4 cells exposed to the inhibitor recovered as soon as the inhibitor was removed. In our most recent studies with Stadecker, Calderon & Kamovsky (1977), we proceeded to assay in the macrophage cultures for molecules that inhibited both radiolabeled thymidine uptake and actual cell growth. The results using a series of biological and biochemical approaches clearly pointed to thymidine as the molecule responsible for both efEects when testing on EL-4 cells. The thymidine, in great part, was synthesized by the macrophages; macraphages pulsed with C^-'-formate released radiolabeled thymidine. The amounts of thymidine made by the macrophages were cotistant for at least 48 h of culture and were surprisingly high. In fact, no other natural cell in biology that we know of secretes nucleosides. The amount of thymidine - about 0-3 to 1 /fg per 5 X 10^ macrophages per ml of culture per 24 h - was sufficient to compete for the radiolabeled thymidine used in standard assays for DNA synthesis and, on some cells, could result in true thymidine blockade. Thymidine blockade refers to a phenomenon whereby excess thymidine blocks the conversion of cytidylate to deoxycytidilate. The amounts necessary to produce thymidine blockade usually vary in various reports in the literature, from 10"^ to 10"^ M.When examining various cell types - tumors, lymphocytes - we found that the amounts of thymidine required to produce thymidine blockade varied greatly - the EL-4 line was, in fact, blocked by solutions of about 1(H M thymidine, about three to four logs l^s than other cells!

tYMPHOCYTE REGULATION BY MACROPHAGE

241

This high susceptibility of this tumor cell explains why the macrophage culture fluids had actually reduced its growth. As would be expected, the effect of the thymidine secreted by macrophages on EL-4 was bypassed by addition of 2-deoxycytidine to the medium. Why do macrophage secrete thymidine? The explanation may lie in the experiments published by Green and associates (Chan et al. 1974). They found that tissue culture lines lacking thymidine kinase released thymidine. Thymidine kinase behaves as a salvage enzyme that phosphorylates the thymidine formed by the cell, thus preventing the molecule from escaping. Stadecker foimd, in collaboration with Kauffman & Davidson in the Department of Microbiology, that macrophages lacked thymidine kinase activity. The observations that macrophages are making thymidine but not DNA suggest that these cells may be ready to go into DNA synthesis at any moment This, in fact, can happen, although this area of macrophage physiology has not been studied thoroughly. Has the macrophage-secreted thymidine any biological function? From a practical point of view, it Is a nuisance for immunologists. The presence of thymidine may well explain many of the in vitro phenomena seen when culturing cells with macrophages. Whether the secreted thymidine has any physiological ro!e, perhaps by producing cytostasis in local inflammatory reactions where macrophages concentrate in high numbers, is a possibility, although remote. A flnal point to take into consideration is that macrophages also secrete adenosine and guanosine, although the secretion of these has not been evaluated (Stadecker et al. 1977). The Lymphostimulatory Molecules — The Mitogenic Protein: The MP secreted by macrophages was discovered by Gery et al. (Gery et al. 1972, Gery & Waksman 1972), and termed lymphocyte-activating factor- It is a small protein, of about 15,000 molecular weight, resistant to trypsin but sensitive to chymotrypsin and papain (Calderon et al. 1975, Blyden & Handschumacher 1977). Whether the MP has enzymatic activity is not clear. MP appears not to be a serine protease, since it is resistant to treatment with diisopropylfluorphosphate. A very similar mitogenic material has been isolated from cell preparations rich in neutrophils and claimed to have enzymatic activity (Nakamura et al. 1976). The neutrophil MP was of low activity, and to the extent that macrophages contributed to its formation, was not determined. A mitogenic protein, also of low molecular weight, about 20,000, has been isolated from cultures of immune guinea pig lymphoid cells cultured with antigen (Gately et al. 1975). The MP that we have been studying is particularly active in promoting cell division of thymocytes and, to a lesser extent, in T and B lymphocytes. We have found no activity against Immunological

Rev. (1978), Vol. 4 0

X5

242

EMIL R. UNANUE

fibroblasts. Preparations apparently pure in MP when added to mixttires of T and B cells enhance antibody formation (Calderon et al. 1975) (see next section). The conditior^ for secretion of MP are most interesting and indicate the very fine control of the macrophage secretory process. There is a small, usually variable secretion of MP under basal conditions from macrophages harvested from the peritoneal cavities of normal mice or mice following injection of peptone or thioglycollate broth. The secretion of MP is increased by: 1) a phagocytic stimulus; 2) exposure to endotoxin; 3) addition of activated lymphocytes or their products; and 4) by judiciotis use of inhibitors of protein synthesis. Addition to macrophages of opsonized red cells, of latex particles, of bacteria, stimulates a brief burst of secretion of MP for a period of 24 to 48 h (Unanue et al. 1976) (Figure 4). Non-stimtilated macrophages and peptonestimulated macrophages respond b ^ t of all, while thioglycollate-induced macrophages - which are highly activated - are poor secretors. In this case, there is an inverse relationship between macrophage activation and secretion of the product. Addition of E. coli lipopolysaccharide is a most potent stimulus for all macrophages (Gery et al- 1972). As was discussed before, the addition of immune T cells with antigen results in the generation of very high amounts of MP (Unanue et al. 1976).

10

LiMtria Organisms

Ko Challenge

Figure 4. The graph shows the increase in production of MP by macrophages after phagocytosis. Macrophages were incubated 24 h with Listeria organisms, opsonized red cells, latex beads, or beryllium and the culture fluids examined every 24 h (Unanue et al. 1976).

LYMPHOCYTE REGULATION BY MACROPHAGE

243

In these instances, intimate cell-to-cell contact is required (Farr et al. 1977). A small increase in MP can be generated by addition to macrophages of fluids from mixed lymphocyte cultures, but this increase - about two-- to four-fold over backgroxmd - is small compared with the previous situation where it rose from 50- to 100-fold over background (Unanue et al. 1976, Meltzer & Oppenheim 1977). The secretion of MP shows a superinduction phenwnenon (Unanue & Kiely 1977). Superinduction refers to the paradoxical increase in the formation of a cell product as a result of inhibition of RNA or protein synthesis (McAuslan 1963). It has been observed in several cell types, i.e. in liver cells secreting tyrosine aminotransferase (Steinberg et al. 1975), in fibroblast forming interferon (Vilcek & Ng 1971) etc. It is thought that the fonnation and/or secretion of the product under study is regulated by a control protein which is affected by the inhibitors. We found the following sequence of events with regard to the MP of macrophages. Macrophages fr^hly harvested from the peritoneal cavity did not contain any MP that could be isolated from their cytoplasm. A few minutes after planting, MP could be isolated and extracted from the microsomal fraction of the cell. The cytoplasmic MP

Cycloheximide [a]

35-

30 Supernoiont

I 2.0

^ i m i d e Treated Macrophages Unlrealed Maciophages

Supetnatanl

0.5

^

0

1

2

3

4

5

6

7

8

HOURS

Figure 5. The graph shows the superinduction phenomenon in the secretion of MP. Macrophages were harvested and planted in dishes for 1 h, after which one set received 10 j4g per ml of cycloheximide for 6 h. Cell extracts and culture supernatants were assayed for content of MP. Supernatant was collected at 8 h. Note that after 1 h MP is found in the cells, but the activity then decays and little is released out. In contrast, the MP in macrophages treated with cycloheximide remains stable; and furthermore, a portion is secreted (Unanue & Kiely 1977).

244

EMIL R. UNANUE

increased in amounts reaching peak levels by 1 to 2 h; it then declined steadily, disappearing altogether by 24 h. The decline in cytoplasmic MP did not correlate with its release from the cell. In other words, the MP had been internally degraded or somehow changed; and very little was secreted by the cell after the stimulus for its formation, which, in this case, was contact with a plastic surface. Inhibition of protein synthesis had two contrasting effects, depending on whether the dmgs were added immediately after harvesting the macrophages or 1 h later - at a time when cytoplasmic MP had been made. Cycloheximide added immediately after harvesting stopped cell synthesis of MP, but its addition 1 to 2 h after planting resulted in an increase in secretion. In these instances, the MP in the cell did not decline, remaining at the same level, and yet secretion of the MP was markedly increased (Figure 5). We have postulated that the secretion of MP involves a control protein, perhaps an enzyme which is somehow involved in the expression and/or degradation of the cellular protein. By shutting off the synthesis of the control protein, the MP persists in the cell and is secreted. The persistence of MP in the cell and its increased secretion (Figure 5) in the continuous presence of inhibitors of protein synthesis raises the possibility that the MP is in two forms, a precursor and an active molecule, the change from precursor to active moiety not involving protein synthesis. The following hypothetical scheme is consistent with our results: Control Protein

(I)

(3)

m

Stimuli -*•-*"-*- Precursor MP -^ -^ -*- MP -^ -»--*- Degradation ^

(4)

Release The stimuli lead to a brief formation of a pool of precursor MP (i, above), which then, in a way that does not involve protein synthesis, converts to active MP (2). MP is degraded (3), but some is released (4). The control protein acts at Stages 2 and/or 3. When no control protein is made, the pool of pre-MP keeps on converting to the active product. The fact that highly activated macrophages show less cellular and secreted MP may be a reflection of an increased content of the control protein. Phagocytosis, which stimulated MP production and secretion may act by increasing Steps 7 and 4. The Lymphostimulatory Molecules - Effects on Antibody Formation: Several

LYMPHOCYTE REGULATION BY MACROPHAGE

245

investigators have found activities in macrophage culture fluids which enhance B cell function (Schrader 1973, Calderon et al. 1975, Opitz et al. 1976, Wood & Cameron 1976). Our experience indicates that macrophage culture fluids contain at least two distinct activities, one that produces the differentiation of immune - memory - B cells into antibody-secreting cells, in the apparent absence of both exogenous antigen and T lymphocytes; and a second that, depending upon the immune status of the spleen cells, increases or decreases antibody production when antigen is added. We believe that the first activity is exerted most likely directiy on B lymphocytes, while the second one involves both B and T cells. Our experiments have employed spleen cells from mice immunized to a haptenated protein, fiuorescein (F) on hemocyanin (F-KLH) (Figure 6). The immune spleen cells cultured in critical amounts of the macrophage culture fluids differentiate to plaque-forming cells. The addition of antigen in an

10 25

50

0 10 25

50

25

50

% SUPERNATANT

Figure 6. Spleen cells from mice primed with F-KLH were incubated with different amounts of macrophage culture fluids and assayed for a plaque response after 4 days. The B cell-differentiating activity refers to the increase in PFC noted in the presence of F in an unrelated carrier - similar results were obtained without addition of any antigen (Calderon et al. 1975).

246

EMEL R. UNANUE

unrelated carrier protein does not enhance antibody formation any further. Thus, the activity acts as a polyclonal type of differentiating molecule most effective on immune spleen cells. It does not behave like a T-cell-repIacing factor. This differentiating activity does not require T lymphocytes inasmuch as it is also evident with cells depleted of T cells by treatment with anti-0 antibodies. The B cell-differentiating activity is generated best in long-term cultures of macrophages and, as with the MP, in cultures of macrophages that are not highly activated; it increases after phagocytosis but not always in parallel with the increase found for MP (Unanue et al. 1976). Thus, the data support the idea that B cells, most likely at a certain stage of their maturation, can readily start to differentiate to secreting cells upon interaction with the macrophage factor. A second activity noted in the macrophage culture fluid enhances, or suppresses, antibody production to the hapten-protein conjugate when the antigen is added to the culture. The enhancement requires specific challenge with the specific hapten-carrier conjugate used for priming the B and T cells. Usually, primed spleen cells from mice immunized weeks before harvesting for in vitro culture respond by an increase in antibody production. This increase is far beyond that seen without the addition of antigen and probably denotes an expansion of T helper function. The addition of the macrophage culture ffuids to spleen cells from freshly primed mice may result in enhancement or suppression of the in vitro antibody response, depending upon the amounts of macrophages culture fluids (Calderon et al. 1975). Figure 7 shows a representative experiment where, at 25 % v/v, the macrophage culture fluid enhanced the response but suppressed it at higher concentrations. These results likely reflect expansion of either helper or suppressor T cell populations. The molecular resolution of these various biological activities has not been adequately resolved. Our experience indicates that the B cell-differentiating activity usually resolves in G-lOO fractionation into two main fractions of about 140,000 and 15,000 molecular weights. The second is at the same position as the MP. However, in limited experiments, not all culture fluids contain both activities. In contrast, the activities that increase antibody production in the presence of antigen and T cells resolve with the MP as well as with a heterogeneous fraction of 40,000 to 60,000 molecular weigjit Further studies are warranted. Although our experience with other assays of antibody formation has been limited, we have done some experiments with the in vitro response to red cells. So far, we have been unable to substitute for macrophages by using macrophage culture fluids containing the activities described above. Thus, xising A/St spleen cells, the response to SRBC is ablated by removal of

LYMPHOCYTE REGULATION BY MACROPHAGE

247

adherent cells and requires the addition of tbese adherent cells for reconstitution, which cannot take place with the macrophage-conditioned media. In limited tests so far, we have not fotmd that other cell lines secrete the two activities that we have described above. Moller et al. (1976), did find a polyclonal activity in some fibroblast linesOther Activities: Conditioned media from macrophages following stimulation contain two activities that have been defined only in biological terms. One increases neovascularization (Polverini et al. 1977). It is our hypothesis that the endotheiial and vasctilar changes taking place in chronic infiammation and, in particular, during an immune response (Herman et al. 1972) could well be caused by the activated macrophages. The seccmd activity is one that produces lymphocyte chemotaxis, particularly of T cells (Ward et al.

4000

25 50 75

0

25 50 75

0

25 50 75

X SUPERNATANT

Figure 7. This is the same experimental setup as in Figure 6 but using freshly primed spleen cells. (The macrophage culture fluids were generated from peptone macropahges treated or not treated with anti-© and C to eliminate contribution of T cells to its generation). Note an increase or suppression in the F-KLH response, depending on dosage; note also the B cell-diflerentiating activity (no antigen coltimns) (Calderon et al. 1975).

248

EMIL R. UNANUE

1977). These two activities could place the macrophage as playing an important role in the anatomical changes in lymph nodes following the entrance of antigen. IV. CONTROL OF DIFFERENTIATION BY MACROPHAGES In the previous section we discussed the apparent effects of macrophage secretory products on one stage of B cell difierentiation. The question addressed in this section is whether the macrophage may regtilate the normal development of a paitictilar cell line. Hematologists have been aware of a colonystimtilating factor from macrophages that allows for the differentation of stem cells to myeloid elements in culture (for example, see Chervenick & LoBuglio 1972). We stmimarize now the results of experiments of Dr. D. I. Beller in our laboratory (Beller & Unanue 1977, Beller et al. 1977) which point to macrophages as regulating a step in normal thymic mattiration. The project started when we questioned whether macrophage culture fluids rich in MP and stimulating thymocyte proliferation were also difierraitating the thymocytes to mature T cells. Beller proceeded to isolate immature thymocytes - by separating the cells in bovine albumin gradients - and to culture them in macrophage culture fluids. The thymocytes, after 48 to 72 h of culture, were found to change their membrane properties, acquiring H-2 antigens, losing TL by cytotoxicity, and, more importantly, becoming responsive in mixed Ieukocj^e culttires. Thus, an impressive degree of mattiration had been achieved. Several points were established: 1) the phenomenon was not explained by selective growth of a few contaminating mature thymocytes within the immature cells - the change In H-2, for example, could be achieved in immature thymocytes treated with mitomycin C and not dividing; 2) the change in phenotype was stable; that is to say, the cells could be cultured with the thymic-differentiating factor, matured, the factor removed, and the cells would still respond to allogeneic stimulation; 3) the mattiration changes required long periods of ctilture (48 to 72 h) and clearly were not reproduced by addition of 2-mercaptoethanoI to the culture medium; 4) the maturing principle could be differentiated from interferon; and 5) the active factor was separable by size from the MP. The thymic-difierentiating factor (TDF) was about 40,000 daltons (Beller & Unantie 1977)- A representative result is shown in Figure 8. Is there any biological significance to a thymic-differentiating principle secreted by macrophages? We believe that this may be so. Macrophages are prominent in the thymus, particularly in the thymic-medullary boundary. It may well be that the sequence of thymic maturation from stem cell -^ prothymocyte -> immattire thymocyte -^ mature T cell is regulated at various stages

LYMPHOCYTE REGULATION BY MACROPHAGE

249

by different elements. Certainly, some of the thymic hormones can produce conversion of a stem cell to acquire thymocyte surface antigens. The step regulated by the macrophages may follow the action of the putative thymic hormones, driving the immature thymocytes to a state of maturity. In support of this multi-signal hypothesis are recent unpublished experiments of Beller showing that the macrophages isolated from the thymus, and not fibroblasts, mature thymocytes in a process also requiring 48 to 72 h. The macrophages isolated from the thymus have similar characteristics to those found in serous cavities but clearly are less activated, although highly active in differentiating the thymocyte. In unpublished observations, Beller has found that preparations of active thymopoietin obtained from Dr. G. Goldstein and active in conversion of stem cells to prothymocytes were not able to produce the changes in H-2 and TL induced by the macrophages. This can be taken to support the multi-signal hypothesis of thymic maturation.

0 0 0 •"

o • A *

Band Band Band Band

1 Ttiymocytes 1 Thytnocyies + MCF 1 Thymocyies + MCF + MitomycJn C 3 Thymocytei

aoo 600

£ 400

200

Vi

~-l

0 2 WMfff/f

4

6

8

10

OF ABSORBING CELLS

Figure 8. This graph depicts the change in the H-2 in thymocytes incubated with macrophage culture fluids (MCF). H-2 was measured by an absorption assay - a standard amount of antibody was absorbed with a variable number of cells (shown in tbe horizontal axis) and then titrated against Cr"-labelcd spleen cells. Band 1 thymocytes represent the immature cells which contain small amounts of H-2. Incubation with MCF makes the Band 1 cells develop as much H-2 as that found in the mature - Band 3 cells. This process is not affected by suppression of DNA synthesis (Beller & Unanue 1977).

250

EMIL R. UNANUE V. SUMMARY

Macrophages may exert a regulatory influence at various stages in the life of the lymphocyte - they may influence the non-antigen-driven differentiation of lymphocytes - as exemplifled by the effects on thymic differentiation; they may establish the mode and form of antigen to be presented or recognized by the lymphocyte; may regulate the lymphocyte's antigen-driven functions. Each of these critical regulatory steps needs explaining in molecular terms and integrated and placed in the context of the other regulatory functions of lymphocytes. The control of secretion of MP is an eloquent example of the molecular complexities and the intricate control mechanisms - internal and external - operating at each step of each regulatory process. A flnal comment concerns the question of macrophage heterogeneity. Is the same cell performing all the functions of degradation, presentation, and secretion - or cytotoxicity? Or do we have subpopulations, each with a different role? This issue is not settled. The Unitarians argue that the phagocytes pass through different stages of differentiation and that each function may become more or less prominent at each stage. Certainly, the manner in which each macrophage function is assayed can condition the outcome: for antigen presentation, one adds 1 % of macrophages to cultures of spleen cells; for cytotoxic Essays, the figure is 50 to 100 macrophages per tumor cell! It is our feeling that until such time as membrane molecules are identified and used as probes for differentiation or for identification of subsets we will not resolve this issue. Along these lines, macrophages have been found to have Ia antigens (Hammerling et al. 1975, Schwartz et al. 1976) and can be divided into two sets on the basis of the presence or absence of la- Dorf and I have found - by cytotoxicity - that only about 35 to 50 % of peritoneal macrophages bear Ia molecules (Dorf & Unanue 1977). Exceptionally, some exudates will bear up to 75 % positive cells. Neither la-positive nor lanegative macrophages change significantly after prolonged periods of culture. Whether these results indicate two defined subsets of macrophages is now being investigated.

ACKNOWLEDGMENTS

This paper has summarized part of our work with phagocytes, which has involved many individuals. It started at the National Institute for Medical Research in London with Dr. B. A. Askonas. I am deeply grateful to her for introducing me to the field and for her continuous support Our work continued at Scripp Clinic in La JoUa with Doctors Cerottini, Cruchaud, and

LYMPHOCYTE REGULATION BY MACROPHAGE

251

Schmidtke, and now at Harvard Medical School with Drs. J. Calderon, D. I. Beller, A. G. Farr, D. H. Katz, J.-M. Kiely, C. L. Sidman and M. J. Stadecker. I am most grateful to all for their dedication and their efforts. Most importantly, we all thank the phagocytes for being such a wonderful and tricky cell capable of keeping us busy for part of the day. Our work has been supported by National Institutes of Health Grants AI 10091, NCI 14723, and by a grant from the Council for Tobacco Research.

REFERENCES Adler, F. L., Fishman, M. & Dray, S. (1968) Antibody formation in vitro. IH. Antibody formation and allotypic specificity directed by ribonucleic acid from peritoneal exudate cells. J. Immunol. 97, 554-558. Barcinski, M. A. & Rosenthal, A. S. (1977) Immune response gene control of determinant selection. I. Intramolecular mapping of the immunogenic sites on instilin recognized by guinea pig T and B cells. /. exp. Med 145, 726-742. Beller, D. I., Farr, A. G. & Unanue, E. R. (1978) The regulation of lymphocyte proliferation and differentiation by macrophages. Fed. Proc. 37, 91-96. Beller, D. I. & Unanue, E. R. (1977) Thymocyte maturation in vitro by a secretory product from macrophages. 7. exp. Med. 118, 1780-1787. Blyden, G. & Handschumacher, R. E. (1977) Purification and properties of human lymphocyte activating factor (LAF). J. Immunol. 118, 1631-1638. Calderon, J., Kiely, J.-M., Lefko, J.L. & Unanue, E.R. (1975) The modulation of lymphocyte functions by molecules secreted by macrophages. I. Description and partial biochemical analysis. /. exp. Med. 142, 151-164. Calderon, J. & Unanue, E. R. (1974) The release of antigen molecules from macrophages - characterization of the phenomena. /. Immunol. 112, 1804—1814. Calderon, J. & Unanue, E. R. (1975) An evaluation of the role of macrophages in immune induction. Fed. Proc. 34, 1737-1742. Calderon, J., Williams, R. T. & Unanue, E. R. (1974) An inhibitor of cell proliferation released by cultures of macrophages. Proc. nat. Acad. Sci. (Wash.) 71, 4273-4277. Chan, T. S., Meuth, M & Green, H (1974) Pyrimidine excretion by cultured fibroblasts: Effect on mutational deficiency in pyrimidine salvage enzymes. /. Cell. Physiol. 83, 263-270. Chen, P. C , Gaetjens, E. & Broome, J. D. (1977) Macromolecular inhibitory factor for lympboid cells produced by mouse macrophages. Immunology 33, 391-398. Chervenick, P.A. & LoBuglio, A.F. (1972) Human blood monocytes: stimulators of granulocyte and mononuclear colony formation in vitro. Science 178, 164-166. Cnichaud, A. & Unanue, E. R. (1971) Fate and immunogenicity of antigens endocytosed by macrophages: a study tising foreign red cells and immunoglobulin G. /. Immunol. 107, 1329-1349. Dorf, M. E. & Unanue, E. R. (1977) Subpopulations of peritoneal macrophages identified with anti-la sera. In: Ir Genes and la Antigens. Proceedings of the Third Ir Gene Workshop, ed. McDevitt, H. O., Academic Press, New York, in press. EUner, J. J., Lipsky, P. E, & Rosenhtal, A. S. (1977) Antigen handling by guinea pig

252

EMIL R. UNANUE

macrophages: further evidence for the sequestration of antigen relevant for activation of primed T lymphocytes. /. Immunol. 118, 2053-2057. Ellner, J. J. & Rosenthal, A. S. (1975) Quantitative and immunologic aspects of the handling of 2,4 dinitrophenyl guinea pig albumin by macropbages. J. Immunol. 114, 1563-1569. Erb, P. & Feldmann, M. (1975) The role of macrophages in the generation of T helper cells. III. Influence of macrophage-derived factors in helper cell induction. Eur. J. Immunol. 5, 759-766. Farr, A. G., Dorf, M.E. & Unanue, E.R. (1977) Secretion of mediators following T lymphocyte-macrophage interaction is regulated by the major histocompatibility complex. Proc. nat. Acad. Sci. (Wash.) 74, 3542-3546. Fishman, M. & Adler, F. L. (1963) Antibody formation in vitro. IL Antibody synthesis in X-irradlated recipients of diffusion chambers containing nucleic acid derived from macrophages incubated witb antigen. /. exp. Med. 117, 595-603. GalUly, R. & Feldman, M. (1967) The role of macrophages in the induction of antibody in X-irradiated animaJs. Immunology 12, 197-206. Gately, C. L., Gately, M. K. & Mayer, M. M. (1975) The molecular dimensions of mitogenic factor from guinea pig lymph node cells. /. Immunol 114, 10-16 Gell, P. G. H. & Benacerraf, B. (1959) Delayed hypersensitivity to denatured proteins in guinea pigs. Immunology II 64, Gery, I. & Waksman, B. H. (1972) Potentiation of T-lymphocyte response to mitogens. II. The cellular source of potentiating mediator(s). /. exp. Med. 136, 143-155. Gery, I., Gershon, R. K. & Waksman, B. H. (1972) Potentiation of the T-lymphocyte response to mitogens. I. The responding cell. J. exp. Med. 136, 128-142. Gordon, S., Unkeless, T. & Cohn, Z. A. (1974) Induction of macrophage plasminogen activator by endotoxin stimulation and phagocytosis. Evidence for a two-stage process. /. exp. Med. 140, 995-1010. Gottlieb, A. A. (1969) Studies on the binding of soluble antigens to a unique ribonuclcoprotein fraction of macrophage cells. Biochem. 8, 2111-2116. Hammerling, G.., Mauve, G., Goldberg, E. & McDevitt, H. O. (1975) Tissue distribution of la antigens: Ia on spennatoza, macrophages, and epidermal cells. Immunogenetics 1, 428-442. Herman, P. G., Yamamoto, I. & Mellins, H. Z. (1972) Blood microcirculation in the lymph node during tbe primary immune response. /. exp. Med. 136, 697-714. Ishizaka, K., Kishimoto, T., Delesperse, G. & King, T. P. (1974) Immunogenic properties of modified antigen E. I. Presence of specific determinants for T cells in denatured antigen and polypeptide chains. /. Immunol. 113, 70-74. Kasahara, T. & Shioiri-Nakano, K. (1976) Splenic suppressing factor: purification and characterization of a factor suppressing thymidine incorporation into activated lymphocytes. /. Immunol. 116, 1251-1256. Katz, D. H. & Unanue, E. R. (1973) Critical role of determinant presentation in the induction of specific responses in immunocompetent lymphocytes. /. exp. Med. 137, 967-990. Katz, D.H., Hamaoka, T., Dorf, M.E., Maurer, P.H. & Benacerraf, B. (1973) Cell interactions between histoincompatible T and B lympbocytes. IV. Involvement of the immune response (Ir) gene in the control of lymphocyte interactions in responses controlled by the gene. /. exp. Med. 138, 734-739. McAuslan, B. R. (1963) The induction and repression of thymidine kinase in the poxvirus-infected HeLa cell. Virology 21, 383-389.

LYMPHOCYTE REGULATION BY MACROPHAGE

253

Meltzer, M. S. & Oppenheim, J. J. (1977) Bidirectional amplification of macrophagelymphocyte interactions: enhanced lymphocyte activation factor production by activated adherent mouse peritoneal cells. /. Immunol. 118, 77-82. Mitchison, N. A. (1969) The immunogenic capacity of antigen taken up by peritoneal exudate cells. Immunology 16, 1-14. Mitchison, N. A. (1971) The carrier effect in the secondary response to hapten-protein conjugates. II. Cellular cooperation. Eur. J. Immunol. 1, 18-27. Moller, G., Lemke, H. & Opitz, H. G. (1976) The role of adherent cells in the immune response: fibroblasts and products released by fibroblasts and peritoneal cells can substitute for adherent cells. Scand. J. Immunol. S, 269-280. Nakamura, S., Yoshinaga, M. & Hayashi, H. (1976) Interaction between lymphocytes and inflammatory exudate cells. 11. A proteolytic enzyme released by PMN as a possible mediator for enhancement of thymocyte response. / . Immunol. 117, 1-6. Nelson, D. S. (1973) Production by stimulated macrophages of factors depressing lymphocyte transformation. Nature 246, 306-307. Opitz, H. G., Niethammer, D., Lemke, H., Flad, H. D. & Huget, R. (1975) Inhibition of 'H-thymidine incorporation of lymphocytes by a soluble factor from macrophages. Cell. Immunol. 16, 379-388. Opitz, H. G., Opitz, U., Lemke, H., Huget, R. & Flad, H. D. (1976) Polyclonal stimulation of lymphocytes by macrophages. Eur. J. Immunol. 6, 457—467. Pierce, C. W., Kapp, J. A. & Benacerraf, B. (1976) Regulation by the H-2 gene complex of macrophage-lymphoid cell interactions in secondary antibody responses in vitro. J. exp. Med. 144, 371-381. Polverini, P. J., Cotran, R. S., Gimbrone, M. A. Jr. & Unanue. E. R. (1977) Activated macrophages induce vascular proliferation. Nature 269, 804—806. Rajewsky, K., Schirrmacher, V., Nase, S. & Jerne, N. K. (1969) The requirement for more than one antigenic determinant for immunogenicity. J. exp. Med. 129, 11311143. Roelants, G. E., Goodman, J. W. & McDevitt, H. O. (1971) Binding of a polypeptide antigen to ribonucleic acid from macrophage, HeLa and E. coli cells. /. Immunol. 106, 1222-1226. Rosenthal, A. S. & Schevach, E. M. (1973) Function of macropbages in antigen recognition by guinea pig T lymphocytes. I. Requirement for histocompatible macrophages and lymphocytes. /. exp. Med. 138, 1194-1212. Scbmidtke, J. R. & Unanue, E. R. (1971) Macrophage-antigen interaction: uptake, metabolism, and immunogenicity of foreign albumin. /. Immunol. 107, 331-338. Schrader, J. W. (1973) Mechanims of activation of bone marrow lymphocytes. HI. A distinction between a macrophage-produced triggering signal and the amplifying effect on triggered B lymphocytes of allogeneic interactions. /. exp. Med. 138, 1466-1480. Schreiner, G. F. & Unanue, E. R. (1976) Membrane and cytoplasmic changes in B lymphocytes induced by ligand-surface immunoglobulin interaction. Adv. Immunol. 24, 38-165. Schwartz, R. H., Dickler, H. B., Sachs, D. H. & Schwartz, B. D. (1976) Studies of la antigens on murine peritoneal macrophages. Scand. J. Immunol. 5, 731-743. Schwartz, R. H. & Paul, W. E. (1976) T lymphocyte-enriched murine peritoneal exudate cells, n . Genetic control of antigen-induced T lymphocyte proliferation. /. exp. Med. 143, 529-540. Sela, M., Schechter, B., Schechter, I. & Borek, F. (1967) Antibodies to sequential and

254

EMIL R. UNANUE

conformational determinants. Cold Spring Harbor Symp. Quant. Biol. 32, 537-552. Shirrmacher, V. & Wigzell, H. (1974) Immune responses against native and chemically modified albumins in mice. n . Effect of alteration of electric charge and conformation on the humoral antibody response and on helper T cell responses. /. Immunol. 113, 1635-1643. Spitznagel, J. K. & Allison, A. C. (1970) Mode of action of adjuvants: effects of antibody responses to macrophage-associated bovine serum albumin. J. Immunol. 104, 128-139. Stadecker, M. J., Calderon, M., Kamovsky, M. L. & Unanue, E. R. (1977) Synthesis and release of thymidine by macrophages. J. Immunol. 119, 1738-1743. Steinberg, R. A., Levinson, B. B. & Tomkins, G. M. (1975) 'Superinduction' of tyrosinc aminotransferase by actinomycin D: a reevaluation. CeU 5, 29-44. Ulrlch, F. (1974) A dialyzable protein synthesis inhibitor released by mammalian cells in vitro. Biochem. Biophys. Res. Comm. 60, 1453-1459. Unanue, E. R. (1970) Thymus dependency of the immune response to hemocyanin - an evaluation of the role of macrophages in thymectomized mice. J. Immunol. 105, 1339-1343. Unanue, E. R. (1972) The regulatory role of macrophages io antigetiic stimulation. Adv. Immunol 15, 95-165. Unanue, E. R. (1976) Secretory function of mononuclear phagocytes. Amer. J. Pathol. 83, 395^17. Unanue, E. R. & Askonas, B. A. (1968a) The immune response of mice to antigen in macrophages. Immunology 15, 287-296. Unanue, E. R. & Askonas, B. A. (1968b) Persistence of immunogenicity of antigen after uptake by macrophages. J. exp. Med. 127, 915-926. Unanue, E. R. & Cerottini, J.-C. (1970) The immunogenicity of antigen bound to the plasma membrane of macrophages. /. exp. Med. 131, 711-725. Unanue, E. R., Cerottini, J.-C. & Bedford, M. (1969) The persistence of antigen on the surface of macrophages. Nature (Lond.) 222, 1193-1145. Unanue, E. R., Dorf, M. E., David, C. S. & Benacerraf, B. (1974) The presence of I region associated antigen on B cells in molecules distinct from Ig and H-2K and H-2D. Proc. nat. Acad. Sci. (Wash.) 71, 5014-5016. Unanue, E. R. & Katz, D. H. (1973) Immunogenicity of macrophage-bound antigens: the requirement for hapten and carrier determinants to be on the same molecule for T and B lymphocyte collaboration, Eur. /, Immunol. 3, 559-563. Unanue, E. R. & Kiely, J.-M. (1976) Unpublished experiments. Unanue, E. R. & Kiely, J.-M. (1977) Synthesis and secretion of a mitogenic protein by macrophages: description of a superinduction phenomenon. J. Immunol. 119. 925-931. Unanue, E. R., Kiely, J.-M. & Calderon, J. (1976) The modulation of lymphocyte functions by molecules secreted by macrophages. IL Conditions leading to increased secretion. /. exp. Med. 144, 155-166. Vilcek, J. & Ng, M. H. (1971) Post-transcriptional control of interferon synthesis. /. Virol. 7, 588-594. Waldman, S. R. & Gottlieb, A. A. (1973) Macrophage regulation of DNA synthesis in cells: Effects of a soluble factor from macrophages. Cell. Immunol. 9, 142-156. Ward, P. A., Unanue, E. R., Goralnick, S. & Schreiner, G. F. (1977) Chemotaxis of rat lymphocytes. /. Immunol. 119, 419-421.

LYMPHOCYTE REGULATION BY MACROPHAGE

255

Wood, D. D. & Cameron, P. M. (1976) Stimulation of the release of a B cell-activating factor from human monocytes. Cell. Immunol. 21, 133-145. Zinkernagel, R. M. & Doherty, P. C. (1974) Immunological surveillance against altered self components by sensitised T lymphocytes in lymphocytic choriomeningitis. Nature 251, 547-548.