Immunological Tolerance: A Reassessment of Mechanisms of the Immune Response

Immunological Tolerance: A Reassessment of Mechanisms of the Immune Response presents the basic biological phenomena of immunological tolerance. This book discusses immunology as a critical field for the analysis of molecular features of mammalian cell genetics, biosynthesis, differentiation, and interactions among cell types. Organized into six chapters, this book begins with an overview of the relationship between antigen structure and its ability to induce tolerance. This text then examines the essentiality of antigen for the proliferation of activated antibody-forming clones and discusses the role of antibody in homeostasis and suppression. Other chapters consider the restoration and transfer of immunological responsiveness by thoracic duct lymphocytes. This book discusses as well the distribution of antigen in tissues and cells. The final chapter deals with the origin of carrier antibody and the significance of maternal transfer of antibodies and antigens. This book is a valuable resource for immunologists, microbiologists, scientists, and clinicians.

• Language: English
• Publisher: Academic Press
• Release date: Jun 28, 2014
• ISBN: 9781483270654

Book preview

1969

I

PROPERTIES OF ANTIGENS IN RELATION TO RESPONSIVENESS AND NON-RESPONSIVENESS

Publisher Summary

This chapter describes a few generally accepted facts concerning the relationship between antigen structure and its ability to induce tolerance. In an adult, an antigen whose physicochemical properties allow its free diffusion in body fluids is more tolerogenic than immunogenic. In contrast, antigens whose physicochemical properties cause their concentration in the cells of the reticuloendothelial system or on the surface of macrophages are recognized as being more immunogenic. Two types of explanations have been proposed to account for these observations relating localizing properties and fate of antigens, with tolerance or immunity. One of the explanations is that in adult animals, tolerance is achieved by freely diffusible antigens because of the direct contact of such antigens with specific cells of thymic origin—before they encounter antigen in or on macrophages. Thus, contact with antigen before it goes to the macrophages leads to tolerance. When it is tolerant, such a cell is inhibited to differentiate or proliferate. Other explanation accounts for the immunogenic difference between particulate and soluble antigens; the explanation is that particulate antigens act on macrophages as adjuvants, and as a consequence of their uptake by these cells, macrophages produce some nonspecific material that induces the sensitive cell, which has reacted with antigen in its environment to differentiate and to proliferate.

Particulate vs. soluble antigens – Activation, antibody synthesis and proliferation of antibody-forming cells as separate events – Antibodies of differing affinity as a factor in non-responsiveness – Thermodynamic considerations – Role of the carrier – Genetic aspects of non-responsiveness – Dosage effects – Termination of tolerance – Silence vs. elimination of immuno-competent cells in non-responsiveness – Relationship between the tolerant and the productive state of lymphocytes – Exhaustive differentiation as a factor in non-responsiveness – Fate of antigen in vivo – Recruitment vs. proliferation of antibody-forming cells – Purging of cell-receptors.

DR. BENACERRAF: I will describe first some generally accepted facts concerning the relationship between antigen structure and its ability to induce tolerance, then propose a hypothesis to account for these facts, and lastly, point out some questions that I feel need to be asked and discussed.

I feel that among the most important of these facts is the observation that in an adult, an antigen whose physicochemical properties allow its free diffusion in body fluids is more tolerogenic than immunogenic. In contrast, antigens whose physicochemical properties (or the presence of antibodies) cause their concentration in the cells of the reticuloendothelial system or on the surface of macrophages, are recognized as being more immunogenic. I shall not take the time to describe the numerous experiments carried out by many of us to substantiate this point, as they are known to most of you.

Two types of explanations have been proposed to account for these observations relating localizing properties and fate of antigens, with tolerance or immunity. One of the explanations is that in adult animals tolerance is achieved by freely diffusible antigens because of the direct contact of such antigens with specific cells of thymic origin–before they encounter antigen in or on macrophages. Thus, contact with antigen before it goes to the macrophages would lead to tolerance. When it is tolerant, such a cell is inhibited to differentiate or proliferate. On the other hand, if a specific cell reacts with antigen after it is processed by macrophages, or after it has been localized on the surface of macrophages, then the environment in the lymphoid tissue where this occurs, or the contact with macrophage-bound antigen, causes it to differentiate and proliferate.

But there is another explanation that has been proposed by Dr. Claman to account for the immunogenic difference between particulate and soluble antigens; that is that particulate antigens act on macrophages as adjuvants, and as a consequence of their uptake by these cells, macrophages produce some non-specific material which induces the sensitive cell, which has reacted with antigen in its environment to differentiate and to proliferate. Evidence from experiments on the production of tolerance in vitro by direct contact of lymphocytes with antigen should settle this problem.

The general statement can be made that tolerance results from the contact of a sensitive cell with antigen when, either in the absence of proper localization or processing of the antigen or because of the introduction of immunosuppressive agents (agents that tend to stop proliferation), the clonal proliferation of the sensitized cell is abolished. This, it seems to me, is the basis for all the mechanisms of tolerance induction.

After these considerations concerning the relationship between the diffusible or particulate nature of antigens and tolerance and immunity, we might try to see whether on a similar basis we can explain the fascinating phenomenon of low-dose tolerance to protein antigens described by Dr. Mitchison; this phenomenon is observed most particularly with soluble antigens. Indeed, one can easily understand that low-dose tolerance may be the result of the fact that when too little of a soluble and diffusible antigen is injected, it contains an insufficient amount of the aggregated fraction capable of being localized by the macrophages. Then the soluble unaggregated fraction acts as a tolerizing antigen. Thus the prediction can be made that with a preparation of a soluble antigen that has been freed of particulates, the phenomenon of low-dose tolerance should not be observed since all doses should then induce tolerance; this, I believe, is the case. The phenomenon of low-dose tolerance should depend upon the competition between the most immunogenic and most tolerogenic fractions of soluble antigens. Within a certain dose range enough of the immunogenic fraction is administered to stimulate immunity.

The next question we can ask is: how is tolerance induced, and how do the general properties of antigen affect tolerance or affect immunity? By this I don’t mean to refer to the fate of the antigen but rather to the chemical properties of the antigen itself. Three hypotheses can be proposed in this respect. First, tolerance can result from the destruction of a specific cell by antigen as originally suggested by Dr. Burnet. Second, it can result from the inhibition of the differentiation of the cell by antigen, the cell remaining dormant for its normal lifetime, with the possibility that it may eventually revert to a non-tolerant state. According to the third mechanism, tolerance may result when an antigen causes the differentiation of the cell to an antibody-producing cell without proliferation or with minimal proliferation. This last possibility was originally suggested by Dr. Sterzl who provided some data in favor of such an interpretation. Of the three hypotheses, I think that, at the present time, evidence exists only for the last two.

There are indeed antigens to which tolerance is produced by virtue of the fact that antigen can stimulate antibody synthesis but not clonal proliferation. This particular mechanism has been observed with the pneumococcus polysaccharide by Dr. Howard in mice and by Drs. Paul and Siskind of our laboratory in rabbits. I would like now to illustrate this point. We have confirmed the original observation of Dr. McLeod that the pneumococcus polysaccharide is capable of stimulating a secondary response in rabbits that have been immunized previously with the whole organism. Table 1 shows the nature of the secondary response in rabbits immunized with pneumococcus type III when boosted with a half milligram of SSS III. A booster response is obtained with the polysaccharide only after 6 months have elapsed. A considerable secondary response is observed with pneumococcus polysaccharide late in immunization because such a response presumably requires that enough memory cells accumulate from the primary immunization to respond to the polysaccharide. Figure 1 shows the nature of the response curve; it is very sharp, it peaks at 8 days and the level of antibody is down at 16 days. The response is not sustained, suggesting that specific cells have not been stimulated to proliferate by the polysaccharide. This interpretation is supported further by the observation that the spleen cells of such animals, maintained in vitro, are not stimulated to synthesize DNA by a wide range of polysaccharide concentrations whereas under the same circumstances protein antigens would produce such stimulation. These results suggest that the specific cells that have been stimulated to produce antibody do not proliferate and therefore tolerance to the polysaccharide results. This is also shown in Fig. 2 which presents results of repeated challenges with the same 0.5 mg dose of polysaccharide. A typical response is seen only after the first challenge; repeated challenges demonstrate that tolerance has been established by the first challenge. This again indicates that the memory cells that have been produced by primary immunization are capable of being boosted by the polysaccharides alone, but are not stimulated to proliferation. This state of affairs results in exhaustion of the specific cell population as far as this system is concerned; this then is in keeping with the data to be presented later by Dr. Howard.

TABLE 1

The secondary response of rabbits immunized with pneumococcus Type III and boosted with SSS III

aAnimals boosted at 1 month received 0.4 mg S III intraperitoneally and 0.1 mg intravenously. All others received 0.5 mg S III intravenously. Anti-S III antibody concentration was measured by quantitative precipitin analysis.

(Data reproduced with permission of Journal of Immunology)

Fig. 1 Anti-S III secondary response to 0.5 mg of S III in rabbits immunized with type III pneumococci 6 to 9 months earlier. Open circles indicate average values for the serum anti-S III concentrations of 6 rabbits boosted with type III pneumococci; closed circles represent average values for the serum anti-S III concentrations of four normal rabbits who received an intravenous injection of 0.5 mg S III. (Data reproduced with permission of Journal of Immunology)

Fig. 2 The secondary tertiary and quaternary responses to 0.5 mg of SSS III in rabbits immunized with type III pneumococci 9 months earlier.

The question must be asked: what are the properties of antigens which, aside from stimulating specific antibody synthesis by specific cells, can also evoke their proliferative response? The carrier molecule in the pneumococcus polysaccharide system has these properties; in other systems they are probably also associated with the carrier.

The other question one may ask is: are there other systems aside from those involving non-metabolizable antigens of the pneumococcus polysaccharide type, in which the ratio of cells that go to proliferation to cells that go to differentiation (and antibody synthesis) may be influenced by the properties of the antigen? The answer to this question is yes. This type of effect has been observed with protein antigens as a function of the dose. If large amounts of antigen are injected, more cells tend to go to antibody synthesis and differentiation than when small amounts of antigen are employed. In contrast, when low doses of antigens are used, less cells mature to antibody synthesis and more cells are stimulated to proliferate in preparation for the secondary response. I think there are ample data from many laboratories, including our own, demonstrating this phenomenon.

The last point I want to address myself to, deals with the fact that antibody-producing cells are a heterogeneous population of cells synthesizing antibodies of different affinity. Since the audience is composed almost exclusively of immunobiologists with very few immunochemists, and since those immunochemists that are here are distinguished by their understanding of biology, I think it is not necessary to deal with the problem of whether these cells are precommitted to the synthesis of a specific antibody before antigens are injected. However, it is a generally accepted fact that immunocompetent cells are committed to the synthesis of a specific immunoglobulin before the antigen is introduced. Indeed, if this were not the case, it would be very difficult to explain tolerance as was recognized by Dr. Burnet very early in the game. This being the case, the same conclusions that apply to antibody-producing cells must also apply to cells that can be rendered tolerant by antigen. Since for a given antigen the antibody-producing cells are a heterogeneous population of cells, each producing an antibody with different affinity for the antigen, the same thing can be said about the cells that are rendered tolerant. Therefore, one would expect that the same thermodynamic considerations that apply to the binding of antigens by specific antibody and to the stimulation of cells for the immune response, should also apply for tolerance induction. By this it is meant that if tolerance is achieved, with low doses of antigen, we should expect that the tolerance has a narrow specificity since tolerance would have been achieved by affecting only those cells that are most specific for the antigen. Under such a regimen cross-reacting cells should not be rendered tolerant.

On the contrary, if a large amount of antigen is used to achieve high-dose tolerance, or if the low tolerance regimen is increased progressively so that the animals are eventually exposed to a large concentration of antigen, a much larger width of tolerance specificity should be observed since many more low-affinity cells are rendered tolerant. Tables 2 and 3 illustrate this point.

TABLE 2

Antibody produced by BSA-tolerant rabbits immunized with DNP10 BSA

TABLE 3

Antibody produced by BSA-tolerant rabbits immunized with DNP10 BSA

I hasten to add that the data are similar to data from Dr. Cinader’s laboratory, both groups having reached the same conclusions. Table 2 shows an experiment that was carried out by Drs. Paul, Siskind and Thorbecke. Rabbits were rendered tolerant by a low-dose regimen in which the highest dose injected was one mg of BSA. Rabbits rendered tolerant were identified by the fact that they did not respond to alum-precipitated BSA. They were injected with DNP10 BSA to break tolerance by a method originally introduced by Dr. Weigle. The antibodies specific for BSA, DNP-BSA and DNP-fibrinogen were measured. It is evident that when one mg constitutes the largest dose (a total of nine mg were given to a 2–3 kg rabbit) most animals make a considerable amount of anti-BSA when challenged with DNP10 BSA; the resulting antibodies, however, have a lower affinity for BSA than the antibody produced by a normal animal. These animals also make a fair amount of anti-DNP. Rabbits also can be brought from a state of low-dose tolerance to a state of high-dose tolerance by weekly injections of 100 mg of BSA. The results obtained after challenge with DNP-BSA are shown in Table 3. Almost no anti-BSA was produced; at a time when there is no BSA in the circulation, there is also almost no anti-DNP produced because the degree of tolerance is now so profound that cells only partially specific for BSA and also specific for DNP-BSA have been affected.

It would seem that in discussing tolerance one has to take into consideration thermodynamic considerations affecting the reaction of antigen with cells producing antibodies of different affinity. I hope that Dr. Siskind will present some data showing that when tolerance is induced in a system in which antibody affinity can be measured, and in which tolerance is either only partial or is disappearing, the antibody produced is of very low affinity since only high-affinity cells are rendered tolerant.

It is therefore very important, that one recognizes that a cell has basically a receptor (antibody) and that tolerance involves a reaction of antigen with this cell-associated receptor, the reaction being governed by considerations of affinity and concentration. There are, however, some tolerance experiments which cannot be explained simply by such thermodynamic considerations. More specifically, some studies carried out in our laboratory indicate that tolerance may sometimes involve two cells and not one cell, in a manner that is still not understood.

Thus, you are all familiar with the fact that there are guinea pigs which do not respond to DNP-polylysine (DNP-PLL). Nevertheless, these guinea pigs can be induced to form large amounts of anti-DNP-PLL if they are immunized with this antigen bound to a foreign carrier which is by itself immunogenic, such as ovalbumin or BSA. However, when these guinea pigs are made tolerant to the carrier albumin, these animals are no longer capable of making anti DNP-PLL when immunized with complexes of DNP-PLL and the tolerated carrier protein. This result was obtained despite the fact that we could not detect any degree of specificity for the carrier BSA in the anti-DNP-PLL produced by the control guinea pigs that had not been rendered tolerant. Thus, in this system we find that tolerance to an antigen suppressed its ability to act as a carrier for a hapten.

These are the issues that I want to raise with respect to the nature and the state of antigen and tolerance. I do not leave you with a single theory but rather with many questions to discuss and to resolve.

DR. SISKIND: I would like to say something about the effect of partial immunological tolerance on antibody-binding affinity. Dr. Thies and I made rabbits partially tolerant to DNP-horse serum albumin. This was done either in newborn animals by giving large doses of antigen or in adult animals by a low-dose method. In both cases immunization with DNP-HSA in complete Freund’s adjuvants elicited formation of a small amount of anti-DNP that had extremely low affinity. The binding energy for DNP-PLL was reduced by approximately 2 kilocalories as compared with the binding energy of the antibody produced in normal animals. It thus appears that cells that make antibody of the highest binding affinity for the antigenic determinant are the cells that are most readily rendered tolerant by an appropriate tolerance-inducing procedure.

This point can be carried one step further. Drs. Werblin, Benacerraf and I made animals tolerant to DNP-rabbit serum albumin and then challenged them with DNP-BGG in incomplete Freund’s adjuvants. Under these circumstances, the amount of anti-DNP made by the tolerant animals was approximately equal to that produced by the normal rabbit. However, the affinity of this antibody for DNP-PLL was slightly reduced (approximately one kilocalorie) as compared with that made by normal animals. Thus again cells of highest avidity for the antigenic determinant were rendered tolerant. This emphasizes the role of antigen interacting with cell-associated antibody in the process of tolerance induction.

DR. SCHWARTZ: Did you, in addition to measuring the affinity, identify the molecular class of the antibody being produced?

DR. SISKIND: All of this work was carried out in rabbits and the antibody involved was presumably IgG.

DR. LESKOWITZ: I would like to raise the point of whether we are really dealing with a problem of differences in affinity of antibodies to a single determinant, or perhaps with different categories of antibodies directed to different antigenic determinants. I imagine that all affinity measurements were made with DNP-epsilon-N-lysine or something of this sort. When you immunize with DNP-protein, it is possible to get DNP, let us say on tyrosine, to function as a determinant, and it is conceivable that some antibody might be made to this. Presumably this kind of antibody would have a lower affinity to DNP-epsilon-N-lysine. I wonder if there were any efforts made to find out whether the antibody of low affinity to DNP-lysine might have had a higher affinity for another potential determinant on the immunizing antigen, such as O-DNP-tyrosine?

DR. SISKIND: I think that there is no reason at all to doubt that some antibody molecules might be stimulated by the DNP-lysine determinant and bind to some other related compound with higher affinity than they bind to DNP-lysine. This, however, would not necessitate assuming that they were stimulated by DNP on tyrosine. They might very well have arisen in response to DNP on lysine which they also bind strongly. I don’t think that this can be clearly