oo22-202X/ 79/ 730 1-0029$02.00/ 0 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY, 73:29-38, 1979 Vol. 73, No.1 Copyright © 1979 by The Williams & Wilkins Co. Printed in U.S.A. The : Clock for Immunologic Aging?

MARGUERITE M.B. KAY, M.D. Molecular and Clinical Laboratory, Geriatric R esearch, Education an.d Clinical Center (GRECC 691 / JI G), Veterans Administration Wadsworl.h Medical Cen.ter, Los Angeles, California and Deparl.ment of Medicine, University of California, Los Angeles, California U.S.A.

A brief overview of the effects of age on function C. As immunologic vigor decrea es, the incidence of infec­ is presented. Normal immune functions can begin to tions, autoimmune and immune-complex diseases, a nd cancel' decline shortly after an individual reaches sexual matu­ increases [8-12], e.g., in immunodeficient newborns and im­ rity. Foremost among the cellular changes are those in munosuppressed adults [13-15]. A strong piece of evidence the stem cells as reflected in their growth properties and linking decreased immunologic vigor to disease is that one can the availability of precursor T cells, and in the T cells, in prevent, or even reverse, the occurrence of immunodeficiency, which a shift in subpopulations may be occurring. Pres­ wasting disease, amyloidosis, and autoimmunity in neonatally ent evidence indicates that thymic precedes, thymectomized and genetically susceptible mice by reconstitut­ and therefore may be responsible for, the age-dependent ing them with young but not old, syngeneic thymus or spleen decline in the ability of the immune system to generate grafts [16,17]. Decreased T cell responsiveness and immune­ functional T cells. It now appears that the primary effect complex deposition have also been implicated as causative of thymic involution is on a T cell differentiation path­ factors in arteriosclerosis [18]. way; the more mature T cells are affected first, the less D. The immune system is amenable to both cellular and mature T cells only later. Thus, the thymus may be the molecular analysis. Therefore, it offers many opportunities for aging clock for the immune system. Current studies are experimental manipulation. centered on processes regulating growth and atrophy of E. Delay, reversal, or prevention of the decline in normal the thymus, and methods for restoring the impaired immune functions may delay the onset or lessen the severity of immune function of elderly individuals. diseases associated with aging. Present evidence suggests that the decline in immune func­ tions that accompanies aging is due primarily to changes in the Aging is chamcterized by a declining a bility of the individual T cell component of th e immune system, and that B cell to adapt to envi.ronmental stress. This decline is exemplified changes are minimal or secondary to those of the T cells [19]. physiologically by an inability to maintain homeostasis. There­ In this presentation, therefore, I will present an overview of the fore, many physiological systems are being experimentally scru­ rapidly developing area of the aging of the immune system by tinized as a part of the search for the mechanism or mechanisms first discussing the age-related cha nges in T cell activities, and responsible for the deregulation of homeostasis and for methods then by focusing on the possible mechanisms that may be to control diseases. Of these, the immune system is perhaps the responsible for these changes. Before doing 0, I will briefly most productive system, for the following reasons: review the major components of the immune system . A. Certain immune functions tend to normally decline with age in humans [1], guinea pigs [2], hamsters [3], rats [4], mice THE IMMUNE SYSTEM [5], dogs [6], and rabbits [7], although the onset, magnitude, The immune system protects the body in a highly specific and rate of decline vary with the type of immune function and manner against foreign invasion by viruses, bacteria, fungi, and the species. possibly one's own somatic cells that undergo neoplastic B. The immune system is "organismal" in that it is in con­ changes by seeking out and destroying them. Any factor or tinual contract with most, if not all, cell, tissue, and organ event that can decrease the normal policing activity of the systems within the body. Thus, a ny alteration in the immune immune system can promote the growth of invasive a ntigens system would be expected to affect all other systems. If one (e.g., bacteria and cancer cells), which can, in turn, disrupt views aging as a pertUl'bation of homeostasis, a mobile, dynamic various physiological functions. Hence, it should be apparent system such as the immune system is perhaps the ideal one in that the immune system plays a major role in the preservation which to study such perturbations and their consequences. of health and, therefore, in life sustenance. The immune system is organismal, i.e., its cells are distributed This is publication number 015 from the Laboratory of Molecular throughout the body and circulate within the blood and lymph. and Clinkal Immunology, and publication number 018 from Veterans The major lymphoid tissues are the thymus, lymph nodes, Administration Wadsworth GRECC. spleen, and (Fig 1). The stem cells, found primar­ Work conducted in my laboratory is supported by DOE Co ntract ily in the bone marrow, can differentiate into T cells, B cells, or No. EY-S-03-0034 and NIH Grant No. HL-22671. Reprint requests to: Marguerite M. B. Kay, M.D., Molecular and macrophages. Macrophages and B cells are generated in the Clinical Immunology Laboratory, Geriatric Hesearch, Education and bone marrow, but T cells originate from pre-T cells under the Clinical Center (GRECC 691/ 11G), Veterans Adm inistration Wads­ influence of the thymus. worth Medical Center, Los Angeles, CA 90024. Obviously, anything that affects the stem cells will affect the Abbreviations: other components of the immune system. Unlike stem cells CML: cell-mediated Iymphocytotoxicity undergoing passage in vivo [21-23], whose self-replicating abil­ Con A: concanavalin A ity can be exhausted, stem cells in situ can self-replicate DNCB: dinitrochlorobenzene throughout the natw-allife span of an individual. However, the GVH: graft-versus- host LPS: lipopolysaccharide ability of stem cells to expand clonally and their rate of division MLC: mixed lymphocyte cu lture decrease with age [24-26], as do their ability to repair x-ray­ MLS: mean life span induced DNA damage [27] and their ability to home into the NZB: New Zealand black thymus [28,29]. When investigators tried to reverse these age­ PHA: phytohemagglutinin related kinetic parameters by enabling "old" stem cells to self­ TdT: terminal deoxynucleotidyl transferase replicate in syngeneic young recipients for a n extended period,

29 30 KAY Vol. 73, No.1

the thymic lymphatic mass decreased with age, primarily as a result of atrophy of the cortex. The onset of this decrease coincided with the attainment of sexual matmity, and was found in both laboratory animals and huma ns [36,37). Subse­ BONE MARROW quently, atrophy of the epithelial cells and decreased levels of thymus hormone or hormones have been observed. Histologi­ cally, the cortex of an involuted thymus is sparsely populated • mL~ STEM (ELLS~ em, with lymphocytes that are replaced by numerous macrophages. filled with lipoid granules [38). In addition, infiltration of plasma THYMUS I and mast cells can be observed in the medulla as well as in the A cortex [38). Although the size of lymph nodes and spleen remains about the same after adulthood in individuals without . )CELL\ : lymphatic neoplasia [36,37], the cellular composition of these tissues shifts so that there are diminished numbers of germinal centers and increased numbers of plasma cells and macro­ SPLEEN LYMPH NODES phages, as well as increased a mounts of connective tissue [39, 40). In humans, the number of circulating T cells has been reported to decrease progressively after adulthood to a level that is 70% that of a young adult by the 6th decade of life, or to remain the same [41-43). Autologous erythrocyte binding, pre­ sumably to immatme T cells, increases with age [44). F Ie 1. Cellular traffic of the immune system. Reprinted from Mak­ B. Cellular Changes inodan, T., Immunity and Agin g, Handbook of the Biology of Aging, ediLed by J. E. Birren, Van Nostrand Rein hold Co., New Yo rk [20]. In vivo changes: In general, T-cell-dependent cell-mediated functions decline with age. However, there are conflicting re­ ports in the Iiteratme concerning certain functions in aging they fo und that the cell s still behaved as old stem cells kineti­ humans and animals [2,9,45-48]. For example, some investiga­ cally. However, when "young" stem cells were allowed to self­ tors have reported a decrease in delayed skin hypersensitivity replicate in syngeneic old recipients, they behaved as old stem to common test to which individuals had been previ­ cell s kinetically. These results indicate that the milieu of stem ously sensitized (such as purified protein derivative of tuber­ cells induces subtle but stable irreversible changes that affect culosis, streptokinase-streptodornase, Candida, Trichophyton, their responsiveness to differentiation homeostatic factors. etc.). Others have reported no decrease except in elderly persons T cells are responsible for protecting the body against viruses, with acute illness. One may argue that any decrease in delayed fungi, and certain types of bacteria; for preventing the growth hypersensitivity seen in the elderly reflects aging of the skin of certain neoplasms; and for regulating B cell antibody pro­ rather tha n the immune system. However, results of tumor cell duction to a large number of antigens. Human T cells are rejection tests in aging mice were comparable, and because identified morphologically by their ability to form rosettes with sheep erythrocytes, and mouse T cells by the presence of the

theta a ntigen on their surfaces. Assays for T cell activity include I I I I responsiveness to plant mitogenic lectins such as phytohemag­ glutinin (PHA), concanavalin A (Con A), participation in de­ layed hypersensitivity (i.e., skin- test) reactions, ability to mount graft-versus-host (GVH) reactions, and ability to help or sup­ press the responses of other cells. B cells can be identified by the presence of immunoglobulin on their surface. The immunoglobulin can be detected by electron microscopy, or more commonly by immunofluorescent staining (Fig 2) . Generall y, B cells cannot produce antibody without the direction of T cells, unless the a ntigens consist of repetitive subunits such as those found in most carbohydrates or polysaccharides. Assays for B cell activity include respon­ siveness to certain mitogens such as lipopolysaccharide (LPS), antibody production in response to T-cell-independent and -dependent antigens, and colony-forming ability. Macrophages also participate in the immune response. They are derived from peripheral blood monocytes, which are derived from a bone marrow precursor cell [31). Macrophages cooperate with T and B cells in many immune responses; they phagocytize bacteria, some viruses, and senescent a nd damaged cells. Be­ cause macrophages usua lly confront antigens before the T and B cells participate in most immunologic activities, many of the earlier studies on the mechanism of loss in immunologic vigor with age were focused on macrophages. These studies showed I , that macrophages are not adversely affected by aging in their handling of antigens during both induction of immune responses and phagocytosis [32-35). FIG 2. Human B ce ll with flu orosce in -Iabeled antihuman immuno­ glob ulin bound to the immunoglobulins on its surface. Wh en exposed AGE-RELATED CHAN GES IN T CELL FUNCTION to ultraviolet light, the flu orescein-labeled molecul es fluoresce a bright A. Morphology green. The B ce ll that posed for this picture has immunoglobulin (represenLed by "Y") on its surface to which fluoroscein-labeled antiim­ The first hint that normal T cell functions might decline with munoglobulin antibodies (represented by "Y") have become bound age came from the findings of morphologists who showed that [taken from reference 30]. July 1979 THE THYMUS: CLOCK FOR IMMUNOLOGIC AGING 31

tumor cells were injected intraperitoneally in these tests, aging This practice would reduce the number of conflicting reports of the skin was not the limiting factor [46). It is more likely regarding the effects of age on immune functions. that the conflicting results are attributable (a) to utilization of There are conflicting reports in the literature on the effects only 1 skin test to assess T cell function, (b) to selection of age on the MLC reaction. Some investigators have reported of skin test antigens (e.g., because only a fraction of the U. S. a marked decrease with age [54,58,71,72). Others have reported population has been immunized aginst tuberculosis, a negative that cells from old mice are equally, if not more, efficient than result cannot be interpreted as refl ective of defective immuno­ cells from young mice, as both responding and stimulating cells . logic memory), and (c) to selection of the population samples in the MLC reaction. However, the same cells showed a de­ (i.e., some studies have used hospitalized patients and medical l: reased GVH reaction index [52). Some of these discrepancies clinic populations). Bel:ause of these considerations, it is rec­ would be avoided if investigators would not use mitomycin­ ommended that a battery of common test a ntigens be utilized treated cells as the stimulating cells in the MLC reaction. for the assessment of secondary delayed skin hypersensitivity Mitomycin may leak from the stimulating cells, and responding reactions. cells from old a nimals may be more susceptible to this drug In contrast to its controversial role in secondary delayed skin than cells from young anin1als. A more sound approach to the reactions, T cell function in primary delayed skin reactions in utilization of the MLC index, when studying the effects of age response to antigens to which the individuals have not been on cell-mediated immunity, is the use of either x-rayed or sensitized previously, such as dinitrochlorobenzene (DNCB), hybrid cells from donors of 1 age group as the stimulating cells. declines with age. Studies performed in vivo with mice indicate that various T­ MECHANISMS OF THE AGE-RELATED DECLINE IN cell-dependent functions decline with age [46,49-51)' Thus, cells NORMAL IMMUNE FUNCTIONS from old mice have a decreased a bility to mount a G VH reaction [49,52], even when enriched T cell populations are utilized to One of the hallmarks of is the increase in compensate for the possibility that old animals may have fewer variability of immune indices [60,73). If more than 1 factor were T cells [52). Resistance to challenge with syngeneic and allo­ responsible for the increase in variability, it would not be geneic tumor cells in vivo decreases dramatically with age surprising to find that the decline in immune capacity of aging [49,53). It is of significance that, in the studies mentioned above, mice resulted from changes in the immune cells, changes in their milieu, or both. To differentiate between the influences of the decline in response to PHA measured in vitro approximated the decline in GVH reaction and tumor cell challenge measured cells a nd theil" milieu, investigators used the cell transfer in vivo [49). method, which assesses immunocompetent cells from young a nd old mice in immunologically inert old and young syngeneic In vitro changes: The findings from experiments performed in vitro show that the proliferative capacity of T cells of human recipients, respectively [46,74-76). The results revealed that beings and rodents, in response to PHA, Con A, and allogeneic both types of change affect the immune response, that about target cells, declines wi th age [1,3,49,54-63]. The PHA-stimu­ 10% of the normal age-related decline can be attributed to changes in the cellular milieu, and that 90% of the decline can lated lymphocytes from aged individuals bind less tritiated actinomycin than those from younger individuals [64,65). be attributed to changes intrinsic to the old cells [75,76). The decline in the cell-mediated lymphocytotoxicity (CML) A. Cellular Milieu index of T cells of long-lived mice has been reported to be moderate against allogeneic tumor cells and not readily appar­ The responsible factor (or factors) in the cellular envil"onment ent against certain syngeneic tumor cells [46). The discrepancy was shown to be systemic and noncellular [75,76). Spleen cells between results of tumor cytotoxicity tests performed in vivo from young mice were cultmed with the test antigen either in and in vitro may be due in part to inadequate culture conditions the young (or old) recipient's spleen by the cell transfer method, in the latter, as indicated by the demonstration that a significant or in the recipient's peritoneal cavity by the diffusion chamber decline with age in the mixed lymphocyte culture (MLC) re­ method [77). A 2-fold difference in response was observed action can be detected with improved culture methods [66). between young and old recipients at both sites, an indication There has been a single report that the response to PHA in that the factor is systemic. The fact that the effect was observed mice shows only a minimal decline with age [67). However, in in cells g1'own in cell-impermeable diffusion chambers further this report, "middle-aged" rather than old C57B1/6J mice were indicates that a noncellular factor is involved. A comparable 2- used; 15- to 20-mo-old mice were used when the mean life span fold difference was also observed when bone marrow stem cells (MLS) of the mice in that laboratory was 24 mo, a procedure were assessed in the spleens of young and old syngeneic recip­ equivalent to assaying 44- to 58-YJ'-old humans whose MLS is ients [27J, an indication that the systemic, noncellular factor 70 yr. This fact directs attention to an important point: the age influences both lympho- and hematopoietic processes. at which a mouse is "old" varies with the strain. For example, The factor could be a deleterious substance of molecular or in 1 laboratory the MLS of DBA/2 males is 86 weeks whereas viral nature, or ' it could be an essential substance that is the MLS ofC67B1/6J males is 130 weeks [62,68). FUlthermore, deficient in old mice. Factors of both types probably change the MLS of a strain can vary between laboratories because the with age, and fmthermore, several factors of each typ e may animal housing conditions can vary; e.g. , C57Bl/6J that have exist. an MLS of 120 weeks in some laboratories may have an MLS Unfortunately, this area of research has not progressed as of 97 weeks in others [69,70). Thus, a 90-week-old C57B1/6J rapidly as anticipated because a simple, sensitive in vitro assay mouse could be considered "old" in 1 laboratory but "middle­ to analyze mouse serum has not yet been perfected. For exam­ aged" in another. ple, it is unclear why normal adult mouse serum, but not fetal It should be emphasized that young adult mice (3 to 10 mo bovine serum, is toxic for mouse immuno co mpet~ nt cells g1"own old) rather than "young" mice (:S 2 months old) should be used in vitro. as a reference point for aging studies. The latter may be "immature" immunologically, according to the immunologic B. Cellular Changes index. For example, it was shown in 1 laboratory that the Ttu"ee types of cellular changes could cause a decline in antisheep erythrocyte response of BC3F, mice (MLS, 30 mol normal immune functions: (a) an absolute decrease in cell does not mature until 6 mo of age [5], but their response to number ttu"ough death, possibly caused by autoimmune cells, PHA peaks at 8 mo of age [58). A further complication encoun­ (b) a relative decrease in cell number as a result of an increase tered with immature young mice is that variability between in the number of "suppressor" cells, and (c) a decrease in individuals is greater than in young adults. The description of functional efficiency caused, perhaps, by somatic mutation. One mice as "old" or "young," then, should be based upon the approach to resolving this problem is to first estimate the survival and developmental pattern of each individual strain. frequency of old mice that exhibit these cellular changes. This 32 KAY Vol. 73, No.1 estimation can be accomplished through the assessm ent of the Recently, HU'okawa and Makinodan [80] tra nsplanted thy­ activity of reference immune cells from adult individuals in the mus lobes from mice 1 day to 33 mo of age into young adult, T­ presence of immune cells from old individuals. A response of cell-deprived recipients. They then kinetically assessed the young-old cell mixtUJ"es that was less than the sum of the emergence of T cells (Fig 3A, B). The thymectomized, x-irra­ responses given by pure young and pure old cells would indicate diated (and therefore T-cell-deficient) mice were given stem that the decreased response of old individuals was due to an cells from bone marrow so that they could produce all blood increase in suppressor cells. If the responses of the mixtures cells. (Because they did not have thymuses, they could not were comparable, it would indicate that the decreased response produce T cells from the pre-T cells made by the bone m arrow.) of old individuals was caused either by a decrease in theu' Each thymus graft was placed in the capsule of a kidney to functional efficiency or a general loss of unmune cells. If the provide it with a good blood supply. In this situation, pre-T response of the young-old mixture was higher than the sum of cells from the bone marrow of recipient mice migrated into the responses given by pure young and pUTe old cells, it would thymus grafts and were transformed into T cells if the thymus indicate that the decreased response of old individuals was had the ability to induce the transformation. The recipient mice caused by a selective loss of 1 type of immune cell that existed were examined (a) for repopulation of the paracortical (T-cell­ in excess in young individuals. The results showed that all 3 dependent) areas of the lymph nodes, (b) for the number of types of interactions can occur. This finding supports the con­ cells with the theta (T cell) antigen on their sUJ"faces, (c) for tention that although there may be only 1 underlying mecha­ theu' responses to the T cell mitogens PHA and succinyl-Con nism responsible for the loss of immunologic vigor with age, it A, (d) for T-cell-dependent antibody responses to sheep eryth­ is expressed differently by aging individuals, a nd this difference rocytes v ia a plaque assay, and (e) for mitogenic reactivity of contributes to the increased variability in immunologic per­ splenic cells to allogeneic lymphocytes (MLC reaction). The formance with age. thymic tissues lost the following influences: the influence on Because involution of the thymus precedes the age-related lymphocyte repopulation of T-cell-dependent areas of lymph decline in T cell function, a cause and effect relationship has nodes (compare step A in Fig 3A, B); the influence on mitogen been suspected (i .e., thymic involution results in a decreased responsiveness of splenic T cells to PHA and succinyl- Con A capacity of the system to generate functional T cells). Experi­ (see step B in Fig 3A, B); the influence on the number of spleen ments have shown that adult thymectomy accelerates the de­ cells with the theta (T cell) a ntigen (refer to step B in Fig 3A, cline in immune responsiveness by demonstrating a decreased B); a nd the influence on the mitogenic reactivity of spleen cells hemagglutinin response to sheep eryth.rocytes, particularly in to allogeneic lymphocytes. HU'okawa and Makinodan found (i) the early response (19S) phase [78]; a marked decrease in the that the ability of the thymus tissue to influence the differen­ GVH reaction; th e poor general condition of the mice; and a tiation or matUJ"ation of precursor cells into functional T cells reduced a ntibody response to bovine serum albumin [79). The decreased with increasing age a nd (ii) that the various T cell effect on the latter response was less pronounced than that on activities exhibited different susceptibilities to thymus involu­ the GVH reaction. tion [80).

No Action

A. Mi c roscopic section of the paracortlca J A. MicroscopIC section of the paraco rtical C. Spl c nl c T-ccll s responding to C. Sple ni c T -cell fa iling to r eSI)(} nd a rea or a lymph node to T-cell s pec Hic mit o~ en s PHA area o f :l l ymph nodc T -cell s pecific mlt o ~ c n s P IIA or s -ConA or s-COIlA

0' .~ D. Mice""co,,'C seelion 01 'ho s ", oo n~

TXO mouse with thym ic g ra n TXB mouse wit h thym ic graft from a I-day old donor fro m a 33- rnonth old dono r

~ , @<;;\ compleme n' No Act ion ..J;<. . --+ @)- ~compl c mcllt g @ ~ --+ W"''''':-..( ----+ Plaque Fo rmation D. Spl e nic T -cells not b c in~ very helpful to an ull certain B-cell, who then leaves s heep r ed bl ood cell s "lone D. Sple ni c T-cclls h clpln~ an uncertain D-cell mount :t n ani ibody r esponse to s hee p red blood cells A B 1. ·,. 1...... , ", I PIG 3, The extent to whi ch age-related in vo lu tion of the thymus affects its capac ity to transform pre-T cells in to T ce lls (see reference 80 for details) , Thymectomized mice were exposed to x- irradi ation for destruction of any remaining T cells. They then were given stem cells fro m bone marrow so th at they could produce all blood ce lls. These thymectomized, x-irradiated, bone-marrow-reconstituted (TXB) mice received transpl ants of thymuses from donor mice between 1 day and 33 mo of age. The TXB (recipient) mice were then examin ed for various responses, such as mitogeni c reactivity of splenic cells to allogenic lym phocytes (mixed lymph ocyte culture reaction) and responses to the T cell mitogens phytohemagglu tin in (PHA) and sll ccinyl-concanavalin A (S,COIl,A). A, transpl antation of thymuses from I-day-old donors. B, transplantation of thymuses from 33-mo-old donors. July 1979 THE THYMUS: CLOCK FOR IMMUNOLOGIC AGING 33

antigen decreases with age, as does the amount of theta a ntigen A. on cell surfaces; yet there is no compensatory increase in B lymphocytes, nor does the lymphocyte number change signifi­ cantly (Fig 5) . These facts suggest an increase in T cells that do not carry detectable theta receptors on their surfaces [83). 2. Although PHA-induced blastogenesis of cells from old mice is significantly reduced (Fig 4), these cells bind 1 ~r, I-l a b e l e d PHA just as well as cells from young mice [84). Because there seems to be no significant decrease in binding affinities or o receptor sites for PHA on cells of old mice, the defect cannot be in their membrane receptors. 3. The cyclic nucleotide 3',5'-guanosine monophosphate, which has been shown to increase when T cells are stimulated B. by mitogens, is found in rela tively small concentrations in mitogen-stimulated T cells of old mice (Fig 6). The levels of cyclic nucleotides are hormone-dependent [85,86], an indication that the defect in T cells is intracellular and may be hormone­ dependent. 4. The life span of hypopituitary dwarf mice, which are T­ cell-deficient, can be extended from 4 mo to 12 mo by a single intraperitoneal injection of 150 X 10(; lymph node cells rich in

oi.n" ... ~.ofe I naiv G L - c..c lL A. B. FIG 4. Postulated effects of age on T cell differentiation. Present evidence suggests that thymi c involu tion leads to a decrease in normal peripheral effector functions by altering the proportion ofT ce ll subsets. It seems that so me T cell subsets, which are necessary for the initiation of certain immunolog ic activi ties, decrease wi th age. It has bee n hy­ pothesized that certain T cells, under the influence of a thymic hormone or hormones (shown here as the teacher, Thymic H), differentiate in to mature T cells (A) . With age, the level of the thymic hormone dec reases. Thus, some T cell subsets do not become "educated" and are not effective in the process of life maintenance (B) [taken from reference 30).

In recent experiments performed in my laboratory with duo­ mosome markers and para biotic mice between 8 to 12 weeks of age, the movement of stem cells into the thymus seemed to cease [28). Thus, at least during this growth phase, it appears that the cells within the thymus are self-perpetuating. This may also be true of peripheral T cells. The cessation of stem cell traffic into the thymus prior to adolescence is probably due to changes in the thymus since stem cells are capable of migrating into thymic transplants, as demonstrated by Hiro­ kawa a nd Makinodan's study [80]. However, participation of an extrathymic regulatory m echanism cannot be excluded. It is significant that the cessation of stem cell to T cell traffic occms dming a growth phase of the individual, and before detectable thymic involu tion or a decrease in the number of T cells in the thymus, lymph nodes, or spleen [28). In this regard, terminal deoxynucleotidyl transferase (TdT), peak II, decreases in the thymuses of long-lived mice from 4 mo of age onward [81). The TdT catalyzes the polymerization of deoxynucleotides in the presence of a primer, in vitro. However, the function of this enzyme in vivo is not known. It has been postulated that it may be a somatic mutagen for immunologic diversification [82). These observations indicate that thymic involution precedes, and is responsible for, the age-dependent decline in the ability of the immune system to generate functional T cells. But what is the mechanism by which thymic involution leads to a de­ crease in normal peripheral effector cell functions? It appears that involution of the thymus alters the proportion of T cells, possibly through decreased synthesis of differentiation hor­ tsbskor-; m:J. mones, through synthesis of factors that can alter cells (Fig 4), FIG 5. T cell aging. This figure shows a T cell aging before your or through both mechanisms. The observations listed below eyes. A, a T ce ll that has decreased amoun ts of theta or T cell antigen support this view. on its surface. B, a T cell that loses theta as it ages [taken from 1. The proportion of mouse lymphocytes bearing the theta reference 30l 34 KAY Vol. 73, NO.1

reduced capacity for repeated cell division, the decline is prob­ A. B. ably due to a reduced number of responding cells. 7. The absolute number of colony-forming, circulating hu­ man T cells decreases with age (55 to 82 yr) to a level that is < 17% that of adults 21 to 35 yr of age (Fig 8 [91]) , In addition, colony-forming T cells of older individuals exhibit a delay in initiating division and decreased proliferative capacity [91]. These T cells exhibit morphological and histochemical charac­ teristics consistent with those of the subpopulation of T cells that show helper, rather than suppressor, activity, as described by others [92,93]. This evidence suggests that with age there is a shift in lymphocyte subpopulations that leads to a decrease in a responsive subset of T cells. + """""ptoetl,.nol 8. DenSity distribution anaJysis, with bovine serum aJbumin or ...... po~mJC. \ co·+'d~ and ficoll discontinuous gradients, shows that the frequency of less dense cells (p 1.06 to 1.08) increases at the expense of the more dense cells (p 1.10 to 1.12) [94]. This density shift within the lymphocyte population can also be seen in young mice shortly after they have been immunized with foreign erythro­ cytes or allogeneic lymphocytes, and in tumor-bearing mice. In

J\ A

v(~ b 'O-- @@ -rih(jrnus cells +- bone m2/'rOW cdls

FIG 6. Effect of age and chemicals on cy clic nucleotide levels of phytohemagglutinin-stimulated (PHA-stimulated) T cells. A, when T cells from young individuals are stimulated with PHA, the level of t he cyclic nucleotide 3',5'-guanosine monophosphate (shown here as blach dots within the membrane) increases. B, when T cells from old individ­ uals are stimulated with PHA, the cyclic nucleotide level and the amount of cel.l pmliferation remain relatively low. However, addition of 2-mercaploethanol or po)ynuc!eotides to the cul ture significantly in creases the blastogenic response of old cells to PHA [taken ii'om reference 30]. mature T cells, but a single injection of an equal number of ce Ll s 6'om a thymus deficient in mature T cells, or 50 X 10" bone marrow cell s, is ineffective (Fig 7 [17,87]) . Comparable life prolongation has been demonstrated by injection of growth factors and thyroxine in untreated, but not in thymectomized, dwarf mice [16,17]. This prolongation suggests that mature T cells are req uired for the approxjmation of homeostasis, but that their activities cannot be sustained throughout the normal life span of the species without replenishment from a precursor pool and, perhaps, a favorable hormonal environment. 12- month 5 5. Transfer of 5 X 107 to 5 X 10~ thymus cells from immu­ nologically mature congenic donors into athymic mice recon­ stitutes th e recipients' capacity to reject allogeneic skin grafts, whereas transfer of an equal number of thymus cells from \~a; ,\ newborn mice does not [88]. Presumably, the ability of thymus cell suspensions fTom adult mice to reconstitute graft rejections FI G 7. Effect of T cells on the life span of hypopi tuitary dwarf mice. depends on the number of "competent" lymphocytes residing Once upon a time in the laboratory of Drs. Fabris, Pierpaoli , and in the thymic medulla or those from the recU'cuJating pool that Sorkin, there were hypopituitary dwarf mice. Besides being dwarves are trapped in the thymus cell suspensions. (which was bad enough) , these poor beasts were also T-ceU-deficient. 6. Cell-cycle experiments on the proliferative response of T As a result of the latter condition, they died at 4 mo of age. The good cells to mitogens and alIoantigens demonstrate (a) that T cells doctors wracked their brains trying to find a way to prolong the lives of their mice. They injected thymus or bone marrow cell suspensions, but of young and old mice have the same generation time (i .e., "" 9 these treatments were futile because the cell suspensions lacked mature hI') and (b) that the length of each stage of the cell cycle T cells. In desperation, th e.Y injected I'y mph node cells containing probably is the same for both young and old cells [89,90]. mature T cells. Although the dwarf mice did not live happily ever after, Because the age-related decline in the proliferative responses they did live to be 12 mo old (i .e., 3 times lheil' normalliIe span) [taken cannot be attributed to an altered generation time or to a from reference 30l July 1979 THE THYMUS: CLOCK FOR IMMUNOLOGIC AGING 35

T CEll C£l1YiI ES however, playa major role in the genesis of the disease of NZB. The causes and mechanisms of the decline in normal immune YOUNG OLD functions, and of immunodeficient diseases in these animal models, may be different from those in long-lived mice and DAY I elderly humans. We may be observing phenotypic cru:icatw-es of old-age immune deficiency that are analogous to the pheno­ typic featw-es of accelerated aging seen in progeric humans. Thus, although a decrease in suppressor T cell activity with age can account for the emergence of autoantibodies in older short­ lived mice, it cannot account for the emergence of autoantibod­ ies in older long-lived mice. In the latter mice the relative number of suppressor T cells increases slightly with age DAY 3 [75,76,102,103]. Increases in the number of regulator T cells with suppressor activity can interfere with the B cell response to antigenic stimulation by reducing either the number of antigen-respon­ sive immunocompetent units or the magnitude of immunologic

DAY 3 burst size. The notion that the number of regulator cells with inhibitory activities increases with age was initially tested through assessment of the antisheep erythrocyte response of spleen cells from young mice in the presence and in the absence DAY G of spleen cells from old mice [27]. T he ratio of the observed response to that which would have been expected if the re­ sponse of young and old cells had been additive was <1. These

1. 5 . b,~md. results indicate that spleens of old mice contain regulator cells that can interfere wi th the immunologic activities of spleen cells FIG 8. Effect of age on coLony-forming, circulating human T cells. from young mice. The proliferative response of T cells from Kinetic studies have revealed that colonies fo rmed by T ceUs fi·om young mice to PHA and allogeneic target cells was also assessed young individuals achieve maximum number by 3 days postculture and in a similar manner. These initial studies have been confirmed that those from older individuals do not reach max imum number until and extended with cells from both aging mice and humans 6 days. Fewer coloni es are formed by T cells fro m older individuals. In addition, the size of the colonies produced by cells from older individuals [63, 103,104]. is smaller than that produced by cells from young individuals. The helper function of T cells declines with age. This fact has been demonstrated in intact animals as well as through assays performed both in vivo and in vitro [35,75,76,101]. The these latter mice, however, the spleen cell number increases, ability of canier-primed T cell populations to collaborate with whereas in unimmunized old mice it does not. This phenomenon young mouse B cells primed with hapten (2,4-dinitrophenol) s uggests that there is a relative increase wi th age in immatw·e decreases dramatically with age [105]. T cells at the expense of mature T cells. Finally, T cells may also regulate hematopoiesis [106], for it 9. Shortly after the thymus begins to involu te and atrophy, has been shown that hematopoiesis of parental stem cells in th e level of serum thymic hormone or hormones decreases with heavily irradiated F 1 recipients can be augmented in the pres­ age [95]. It would seem reasonable to assume that this hormone ence of parental T cells. is necessary for terminal differentiation of T cells, which could lead to a decrease in certain normal effector cell functions, to an increase in suppressor function, and thus to a deficit in c. Thymus overall T cell responsiveness (Fig 4) . All the relevant studies to date indicate that the process of 10. The PHA response of T cells from old mice can be involution and atrophy of the thymus is the key to the aging of significantly increased in vitro by the addition of certain chem­ the immune system. It fo llows, then, that the search for the icals (e.g., mercaptoethanol and polynucleotides) to the cul­ causes and mechanisms of aging of the immune system should t w-es, as can the antibody response to sheep erythrocytes (Fig be oriented ru·ound the thymus. The causes may be either 6A, B [24,96,97]). This effect supports the view that old mice extrinsic or intrinsic to the thymus. The most likely extrinsic are defi cient in a hormone that is essential for the differentiation cause is a breakdown in the regulation of the thymus by the of T cells to become responsive to a mi togen and also suggests hypothalamic-pituitru-y new-oendocrine axis [16,17,107]. Intrin­ t hat simple chemicals can be substituted for this hormone. sic causes might be found at either the DNA level or th e non­ How these regulatory T cells function as suppressors and DNA level [for moleculru· theories of aging, see reference 108]. h elpers is not known; therefore, studies on age-related changes It has been found that stress and other psychological factors in T cells involved in the regulation of B cell immune responses can affect the immune response as measured by skin graft have not been extensive. This dearth is due in part to the rejection, and by primary and secondru·y antibody responses. complexities of the system. The recent development of antisera Lesions in the anterior basal hypothalamus, but not in the directed against allelic T cell determinants (Ly differentiation median or posterior hypothalamus, can depress delayed hyper­ antigens), which allows identification of 3 T cell subsets, should sensitivity and reduce the severity of anaphylactic reactions greatly facilitate research in the area of age effects on T cell [for review, see reference 109]. and insulin suppressor function [98,99]. It seems that T cells expressing the have been shown to act preferentially on T-dependent in1mune Ly 1 antigen are responsible for helper and delayed hypersen­ functions. Thyroxine and sex hormones influence both T and B sitivity effects and that cells expressing Ly 2,3 are responsible cell responses [for review, see references 16 and 17]. On the for suppressor effects and for cell- mediated cytolysis [98,99]. other hand, there is evidence that the thymus can modulate No specific function has yet been assigned to Ly 1,2,3 cells [for hormone levels [16,17,110,111]. The area of thymic-neuroen­ a complete review of suppressor cells in aging, see reference docrine interactions should be an exciting and fruitful subject 100]. for future aging reseru-ch. The evidence that suppressor T cell activi ty declines with With respect to intrinsic causes, 3 possible mechanisms of age was derived from studies on short-li ved New Zealand black involution and atrophy can be proposed. One is clonal exhaus­ (NZB) and related mice [101]. ViTal of the C type, tion [112]; thymus ceUs might have a genetically programmed 36 KAY Vol. 73, No.1 clock mechanism to self-destruct and die after undergoing a 8. Good RA, Yunis EJ: Association of autoimmunity, immunodefi­ fix ed number of divisions. The mechanism would be similar to ciency and aging in man, rabbits and mice. Fed Proc 33:2040- 2050, 1974 that of the Hayfli ck phenomenon, which is seen, in vitro, in 9. Gross L: Immunologic defect in aged population and its relation human fibroblasts. It would require eit her that the thymus to cancer. Cancer 18:201-204, 1965 count the cells leaving it or that the cells count the divisions 10. Mackay IR: Ageing and immunological function in man. Geron­ t hey have undergone, or both. Another possible mechanism is tologia 18:285- 304, 1972 11 . Mackay IR, Whittingham SF, Mathews JD: The immunoepide­ an alteration in thymus cell DNA, either randomly or through miology of aging, Immunity and Aging. Edited by T. Makinodrul, viral [113]. Various stable alterations in DNA, includ­ E Yunis. New York, Plenum Press, 1977, pp 35-49 ing cross-linking and breaks, can occur [for review, see reference 12. Gardner ID, Remington JS: Age-related decline in the resistance 114]. The 3rd possible mechanism is a stable molecular altera­ of mice to infection with intracellular pathogens. Infect Immu­ nol 16:593- 598, 1977 tion at the non-DNA level through subtle error-accumulating 13. Fudenberg HH: Genetically determined immune defi ciency as the mecha nisms. predisposing cause of "autoimmunity" and lymphoid neoplasia. At th e level of t he genes, evidence is mounting that genes of Am J Med 51:295-298, 1971 the H -2 system in mice a nd the HLA region in humans influence 14 . Fudenberg HH, Good RA, Goodman H C, Hitzig W, Kundel H G, Roitt 1M, Rosen FS, Rowe OS, Seligmann M, Sooth ill J R: immune responsiveness and disease susceptibility [115-118]. Primary immunodeficiencies. Bull WHO 45:125-142, 1971 15. Penn I, Starzl TE: Malignant tumors ru'ising de novo in immuno­ CONCLUDING REMARKS suppressed organ transplant recipients. Transplantation 14:407- 417,1972 An attempt has been made to summarize present knowledge 16. Fabris N: Hormones and aging, Immunology and Aging. Edited and to show potential future progress in the areas of age-related by T Makinodan, E Yunis. New York, Plenum Press, 1977, pp changes in T cell immune function and mecha nisms of the age­ 73- 89 related decline. 17. PiantaneLli L, Fabris N: Hypopituitary dwru·f and athymic nude mice and the study of t he relationships among thymus, hor­ N ormal immune functions can begin to decline shortly after mones, and aging, Genetic Effects on Aging. Edited by D H ru'­ an individual reaches sexual maturity. Although changes in the rison. New York, Alan R. Liss, Inc., 1977, pp 315-333 environment of the cells are partially responsible, the decline is 18. Lattime EC, Strausser HR: Arteriosclerosis: Is stress induced due primarily to changes in t he cells. Foremost among the immune suppression a risk factor? Science 198:302- 303, 1977 19. Kay MMB, M akinodan T : Immunobiology of aging: Evaluation of cellular changes are those seen in the stem cells as reflected in current status. Clin Immunol Immunopathol 6:394-413, 1976 their growth properties, and in T cells, where a shift in subsets 20. Makinodan T: Immunity and aging, Hahdbook of the Biology of may be occurring with age. It appears that the process or Aging. Edited by CE Finch, L Hayllick. New York, Van Nos­ processes regulating involution a nd atrophy of the thymus trand-Reinhold Company, 1977, pp 379- 408 could be the key to immunosenescence. Future studies on the 21. Siminovitch L, Till JE, McCulloch EA: Decline in colony-forming ability of mruTOw ce LIs subjected to serial transplantation into mechanism or mechanisms responsible for decreased function irradiated mice. J Cell Physiol 64 :23- 31, 1964 are expected, therefore, to focus as much on the area of the 22. Lajtha LJ, Schofield R: Regul ation of stem cell renewal and neuroendocrine-thymic axis as they do on possible intrinsic differentiation: Possible significance in aging. Advances in Ger­ mechanisms. The increasing size of the aged population is ontological Reseru'ch 3:131-146, 1971 23. Harrison DE, Astle CM, Doubleday JW: Stem cells from old rapidly becoming one of the most critical socioeconomic issues immunodefi cient mice give normal responses in young recipi­ on this planet. Innumerable diseases are associated with age ents. J ImmunolI8:1223-1227, 1977 and with the loss of immunologic vigor. The immune system is 24. Makinodan T , Deitchman JW, Stoltzner GH, Kay MMB, Hiro­ one of t he most attractive systems for attempting intervention kawa K: Restoration of the declining normal immune fu nctions of aging mice. Proceedings of the 10th International Congress in the debilitative processes associated with aging, for many Gerontology 2:23, 1975 reasons. It is a distinct, semiautonomous system and an organ­ 25. Davenport FM, H ennessy AV , Francis T Jr: Epidemiologic and ismal one that interacts with, protects, a nd affects all other immunologic significance of age distribution of antibody to systems. The age-related decline in immune functions precedes, antigenic vru'iants of influenza virus. J Exp Med 98:641-656, 1953 and presumably causes a predisposition to, many of the diseases 26. Albright J , Makinodan T: Decline in the growth potential of seen in elderly individuals. The immune system is easily ap­ spleen-colonizing bone marrow stem cells of long- lived aging proached mechanistically and is amenable to restorative ma­ mice. J Exp Med 144:1204-1213, 1976 nipula tion. As understanding of the aging immune system in­ 27. Chen MG: Age-related changes in hematopoietic stem cell popu­ lations of a long- lived hybrid mouse. J Cell Physiol 78:225- 232, creases, I anticipate that methods will be developed to predict, 1971 minimize, delay, and prevent the debilitative processes associ­ 28. Kay MMB: Effect of age on T cell differentiation. Fed Proc 37: ated with aging. 124 1-1244, 1978 29. T yan ML: Age-related decrease in mouse T-cell progeni tors. J I am grateful to J erry Sproul for his excell ent assistance in preparing Immunol, in press this manuscript and to Dr. Lynn Baker for drawing Figures 2-7. 30. Kay MMB, Baker L: Cell changes associated with declining im­ mune function, Physiology and Cell Biology of Aging. Edi ted by A Cherkin, N Karasch, T Makinodan, B StrehleI', C Finch, S REFERENCES Scott. New York, Raven Press, in press, pp 27-49 1. Roberts-Thomson I, Whittingham S, Youngchaiyud U, Mackay 31. Volkman A, Gowans JL: The origin of macrophages from bone IR: Agein g, immune response, and mortali ty. Lancet 2:368- 370, marrow in the rat. Br J Exp P athol 46:62-70,1965 1974 32. Aoki T , T ell er MN, Robitaille ML: Aging and cancerigenesis. II. 2. Baer H, Bowser RT: Antibody production and development of Effect of age on phagocytic activity of the reticuloendothelial contact skin sensitivity in guinea pigs of various ages. Science system and on tumor growth. J Nat! Cancer Inst 34:255- 264, 140:1211-1212, 1963 1965 3. Mathies M, Lipps L, Smith GS, Walford RL: Age-related decline 33. Aoki T , T eller MN: Aging and cancerigenesis. III. Effect of age on in response to phytohemagglutinin and pokeweed mitogen by isoantibody formation. Cancer Res 26: 1648-1652, 1966 spleen cells from hamsters and a long- lived mouse strain. J 34. T ell er MN: Age changes and immune resistance to cancer. Ad­ Gerontol 28:425-430, 1973 vances in Gerontological Research 4:25-43, 1972 4. Bilder G, Denckla WD: Restoration of ability to reject xenografts 35. Heidrick ML, Makinodan T : Presence of impairment of humoral and c1 eru' carbon after hypophysectomy of adult rats. Mech immunity in nonadherent spleen cells of old mice. J Immunol Ageing D ev 6:153-163, 1977 111:1502-1506, 1973 5. Makinodan T, Peterson JW: Relative antibody-forming capacity 36. Andrew W: Cellular Changes with Age. Springfield, IUinois, of spleen cells as a fu nction of age. Proc Nat! Acad Sci USA 48: Chru'les C Thomas, 1952 234-238, 1962 37. Santisteban GA: The growth and involution of lymphatic tissu 6. J aroslow BN, Suhrbier KM, Fritz TE: Decline and restoration of and its interrelationships to aging and to the growth of the antibody fo rming capacity in aging beagle dogs. J Immunol 112: adrenal glands and sex organs in CBA mice. Anat Rec 136:117- 1467-1476, 1974 126, 1960 7. Nomaguchi TA, Okuma-Sakurai Y, Kimura I: Changes in immu­ 38. Hirokawa K: The thymus and aging, Immunology and Aging. nological potential between juvenile and presenil e rabbits. Mech Edited by T Makinodan, E Yunis. New York, P lenum Medical Ageing Dev 5:409-417, 1976 Book..Co., 1977, pp 51-72 July 1979 THE THYMUS: CLOCK FOR IMMUNOLOGIC AGING 37

39. Chino F, Makinodan T, Lever WH, Peterson WJ: The immune mouse strains. J Gerontol 21:404-409, 1966 system of mice reared in clean and dirty conventional laboratory 70. Walford HL: When is a mouse "old"? J Immunol 117:352-353, farms. I. Life expectancy and pathology of mice with long life 1976 spans. J Gerontol 26:497-507, 1971 71. Gerbase-DeLima M, Meredith P, Walford R: Age-related changes, 40. Peter CP: - Possible immune origin of age-related pathological including synergy and suppression, in the mixed lymphocyte changes in long-lived mice. J Gerontol 28:265-275, 1973 reaction in long-lived mice. Fed Proc 34:159-161, 1975 41. Augener W, Cohnen G, Reuter A, Brittinger G: Decrease of T 72. Meredith P, Tittor W, Gerbase-DeLima M, Walford R: Age-re­ lymphocytes during aging. Lancet 1: 1164, 1974 lated changes in the cellular immune response of lymph node 42. Carosella ED, Monchanko K, Braun M: Ro ette-forming T cells and thymus cells in long- lived mice. Cell Immunol 18:324-330, in human peripheral blood at different ages. Cell Immunol 12: 1975 323- 325, 1974 73. Ma kinodan T , Chino F, Lever WE, Brewen BS: The immune 43. Diaz-Jouanen E, Williams RC Jr, Stri ckl and RG: Age-related systems of mice reru'ed in clean and dirty co nventional labora­ changes in T and B cells. Lancet 1:688- 689, 1974 tory farms. II. Primru'y antibody-forming activi ty of young a nd 44. Fournier C, Charreire J: Increase in autologous erythrocyte bind­ old mice with long life-spans. J Gerontol 26:508-514, 1971 ing by T cells with aging in man. Clin Exp Immunol 29:468-473, 74. Albright JF, Makinodan T: Growth and senescence of antibody­ 1977 forming cells. J Cell Physiol 67 (Suppl 1) :185-206, 1966 45. Giannini D, Sloan HS: A tuberculin survey of 1,285 adults with 75. P rice GB, Makinodan T: Immunologic deficiencies in senescence. special reference to the elderly. Lancet 1:525-527, 1957 I. Chru'acterization of intrinsic deficiencies. J Immunol 108:403- 46. Goodman SA, Makinodan T: Effect of age on cell -mediated im­ 412, 1972 munity in long-lived mice. Clin Exp Immunol 19:533-542, 1975 76. Price GB, Ma kinodan T: Immunologic defi ciencies in senescence. 47. Grossman J , Baum J , Fusner J, Condemi J: The effect of aging II. Characterization of extrinsic deficiencies. J Immunol 108: a nd acute illness on delayed hypersensitivity. J ALlerg Clin 413-417, 1972 Immunol 55:268- 275, 1975 77. Goodman SA, Chen MG, Makinodan T: An improved primary 48. Novick A, Novick I, Potoker S: Tuberculin skin testing in a response from mouse spleen cells cultured in vivo in diffusion clU"onically sick aged population. J Am Geriatr Soc 20:455-458, chambers. J Immunol 108:1387-1399, 1972 1972 78. Metcalf 0: Delayed effect of thymectomy in adul t life on immu­ 49. Perkins EH, Cacheiro LH: A multiple-para meter comparison of nological competence. Nature 208: 1336, 1965 immunocompetence and tumor resistance in aged Balb/c mice. 79. T aylor RB: Decay of immunological responsiveness after thymec­ M ech Ageing Dev, in press tomy in adult life. Nature 208:1334-1335, 1965 50. Stutman 0, Yunis EJ, Good RA: Deficient immunologic functions 80. Hirokawa K, Makinodan T: Thymic invo lution: Effect on T cell of NZB mice. Proc Soc Exp BioI Med 127:1204-1207, 1968 differentiation. J Immunol 114:1659-1664, 1975 51. Teague PO, Yunis EJ, Rodey G, Fish AJ , Stutman 0 , Good RA: 81. Pazmino NH, Ihle IN: Strain-, age-, and tumor-dependent distri­ Autoimmune phenomena a nd renal disease of mice. Role of bution of terminal deoxynucleotidyl transferase in thymocytes thymectomy, aging, and involu tion of immunologic capacity. of mice. J Immunol 117:620- 625, 1976 Lab Invest 22:121-130, 1970 82. Baltimore D: Is terminal deoxynucleotidyl transferase a somatic 52. Walters CS, Cla man HN: Age related changes in cell mediated mutagen in lymphocytes? Nature 248:409, 1974 immunity of Balb/ c mouse. J Immunol 115:1438-1443, 1975 83. Brennan PC, Jaroslow BN: Age-associated decline in theta antigen 53. Stjernsward J : Age dependent tumor-host barrier and effect of on spleen thymus-derived lymphocytes of B6CF, mice. Cell carcinogen initiated immune depression of rejection of iso­ Immunol 15:51-56, 1975 grafted methylcholanthrene induced sar coma cells. J Natal Can­ 84 . Hung C-Y, Perkins EH, Yang W-K: Age-related refractoriness of cer Inst 37:505-512, 1966 PHA-induced lymphocyte transformation. II. '"" I-P HA binding 54. Adler W, T akiguchi T , Smith RT: Effect of age upon primary to spleen cells from young and old mi ce. Mech Ageing Dev 4: aUoantigen recognition by mouse spleen cell s. J Immunol 107: 103-112, 1975 1357-1362, 1971 85. Lyngbye J , Kroll J: Quantitative immunoelectrophoresis of pro­ 55. Fernandez LA, MacSween JM, Langley GR: Lymphocyte re­ teins in serum from normal popula tion: Season-, age-, a nd sex­ sponses to phytohaemagglutinin: Age-related effects. Immunol­ related vru·iations. Clin Chern 17:495- 500, 1971 ogy 31:583- 587, 1976 86. Heidrick ML: Imbalanced cyclic-AMP and cyclic-GMP levels in 56. H a llgren HM, Buckley EC Ill, Gilbertsen VA, Yunis EJ: Lympho­ concanavalin-A stimulated spleen cells from aged mice. J Cell cyte phytohemagglutinin responsiveness, immunoglobulins and BioI 57:139a, 1973 a utoantibodies in aging humans. J Immunollll:1101-1107, 1973 87. Fabris N, Pierpaoli W, Sorkin E: Lymphocytes, hormones and 57. H eine KM: Die reaktionsfahigkeit der lymphozyten im alter. Folia aging. Nature 240:557-559, 1972 . Haematol (Leipz) 96:29-33, 1971 88. Pierpaoli W: Ina bility of thymus cells from newborn donors to 58. H ori Y, Perkins EH, Halsall MK: Decline in phytohemagglutinin restore transplantation immunity in athymi c mice. Immunology responsiveness of spleen cells from aging mice. Proc Soc Exp 29 :465-468, 1975 Bioi Med 144:48-53, 1973 89. Merhav S, Gershon H: The mixed lymphocyte response of senes­ 59. Kay MMB: Immunological aging patterns: Effect of pru'ainfluenza cent mice: Sensitivity to a Ll oantigen and cell replication time. type I virus infection on aging mice of 8 strains and hybrids, Cell Immunol, in press Genetic Effects on Aging. Edited by 0 Bergs ma, 0 Harrison. 90. Abraham C, Tal Y, Gershon H: Reduced in vitro response to N ew York, Alan R. Liss, Inc., 1978, pp 213-240 concanavalin A and lipopolysaccharide in senescent mice: A 60. Kay MMB, Mendoza J , Denton T, Union N, Diven J , Lajiness M: function of reduced number of responding cells. Eur J Immunol Age-related changes in the immune system of 8 medium and 7:301-304, 1977 long-lived strain and hybrids. !' Weight, cellular and activity 91. Kay MMB: Effect of age on human immunological parameters changes. Mech Ageing Dev, in press including T and B cell colony formation. P roceedings Xlth 61. Konen TG, Smith GS, Walford RL: Decline in mixed lymphocyte International Congress of Gerontology, T okyo. Amsterdam, Ex­ reactivity of spleen cells from aged mice of a long-lived strain. cerpta Medica, in press J ImmunolllO:1216-1221, 1973 92. Moretta L, Ferarini M, Mingari M, Moretta A, Webb S: Subpop­ 62. Pisciotta AV, Westring OW, Deprey C, Walsh B: Mitogenic effect ulations of human T cells identified by receptors for immuno­ of phytohaemagglutinin at diffe rent ages. Nature 215:193-194, globulins. J Immunol 117:2 171-2174, 1976 1967 93. Grossi CE, Webb SR, Zicca A, Lydyard P, Moretta L, Mingari M, 63. Inkeles B, Innes JB, Kuntz MM, Kadish AS, Weksler ME: Im­ Cooper MD: Morphological a nd histochemical analyses of two munological studies of aging. III. Cytokinetic basis for the human T-ceU subpopulations beru'ing receptors for IgM or IgG. impaired response of lymphocytes from aged humans to plant J Exp Med 147:1405-1417, 1978 lectins. J Exp Med 145:1158-1168, 1977 94 . Ma kinodan T , Adler W: The effects of aging on the differentiation 64. Preumont AM, Gansen P, Brachet J : Cytochemical study of a nd proliferation potentials of cells of the immune system. Fed human lymphocytes stimulated by PHA as function of donor Proc 34:153-158, 1975 age. Mech Ageing Dev 7:25-32,1978 95. Bach JF, Dardenne M, Salomon JC: Studies on thymus products. 65. Hildemann WH, Reddy AL: Phylogeny of immune responsiveness: IV. Absence of serum "thymic activity" in adul t NZB and (NZB Marine invertebrates. Fed P roc 32:2188- 2194, 1973 x NZW)F, mice. Clin Exp Immunol 14:247-256, 1973 66. Hirano T , Nordin A: Age-associated decline in the in vitro devel­ 96. Braun W, Yajima Y, Ishizuka M: Synthetic polynucleotides as opment of cytotoxic lymphocytes in NZB mice. J Immunol 11 7: restorers of normal antibody forming capacities in aged mice. J 1093-1098, 1976 R eticuloendothel Soc 7:418-424, 1970 67. Shigemoto S, Kishimoto S, Ya mamura Y: Cha nge of cell-mediated 97. Han IH, Johnson AG : Regulation of the immune system by cytotoxicity with aging. J Immunol 115:307-309, 1975 synthetic polynucleotides. VI. Amplification of the immune 68. Smith GS, Walford RL: Influence of the H-2 and H-1 histOCOm­ response in young and aging mice. J Immunol 11 7:423-427, 1966 patibility systems upon li fespan and spontall eous cancer inci­ 98. Cantor H, Boyse EA: Functional subclasses of T lymphocytes dence in congenic mice, Genetic Effects on Aging. Edited by 0 bearing different Ly antigens. I. The ge neration of functionally Bergsma, D Harrison. New York, Alan R. Liss, Inc., 1977, pp distinct T-cell subclasses is a differentiative process independent 281-312 of antigen. J Exp Med 141:1376-1389, 1975 69. Storer JB: Longevity and gross pathology at death in 22 inbred 99. Cantor H , Boyse EA: Functional subclasses of T lymphocytes 38 KAY Vol. 73, No.1

bearing diffe rent Ly antigens. II. Cooperation between sub­ restriction theory of aging a nd development. J Theor Bioi 33: classes of Ly+ cells in the generation of killer activity. J Exp 429-474, 1971 Med 141:1390-1399, 1975 109. Stein M, Schiavi HC, Camerino M: Inf'luence of brain and behavior 100. Gershon RK, Mebder CM: Suppressor cells in aging, Immunity on the immune system. Science 191:435 -440, 1976 and Aging. E dited by T Makinoda n, E Yunis. New York, Plenum 110. Besedovsky H, Sorkin E: Changes in blood hormone levels during Press, 1977, pp 103-110 the immune response. Proc Soc Exp Bioi Med I50:466-470, 1975 101. H ardin JA, Chuseo TM, Steinberg AD: Suppressor cells in the 111 . Pianta ne Lli L, Basso A, Muzzioli M, Fabris N: Thymus-dependent graft vs. host reaction. J Immunol 111:650-651, 1973 reversibility of physiological and isoproterenol evoked age-re­ 102. Goidl EA, Innes J B, Weksler ME: Immunological studies of aging. lated parameters in athymic (nude) and old normal mice. Mech II. Loss of IgG and high avidity pl aque-forming cells and in­ Ageing Dev 7:171-182, 1978 creased suppressor cell activity in aging mice. J Exp Med 144: 11 2. Hayflick L: The limited in vitro lifetime of human diploid cell 1037-1048, ]976 strains. Exp Cell Res 37:614 -636, 1965 103. Segre D, Segre M: Humoral immunity in aged mice. II. Increased Il3. Proffi tt MH, Hirsch MS, Black PH: Murine leukemia: A virus: suppressor T cell activi ty in immunologically defi cient old mice. induced autoimmune di ease? Science 182:821-823, 1973 J Immunol 116:735- 738, 1976 ] 14 . Price GB, Ma kinoda n T : Aging: Alterations of DNA-protein in­ J04 . Antel JP, Weinrich M, Arnson BGW: Circulating suppressor cells formation. Gerontologia 19:58-70, 1973 in man as a function of age,. Clin Immunol lmmunopathol 9:134- 115. Benacerraf B, McDevitt OH: Histocompatibility-linked immune 141, 1978 response genes. Science 175:273-279, 1972 105. Krogsrud RL, Perkins EH: Age-related changes in T cell function. 116. M eredith P, Walford HL: Effect of age on the response to T and J Immunol 118: 1607-1611, 1977 B cell mitogens in mice congenic at the H-2 locus. Immunoge­ l06. Goodman JW, Shinpock SG: Influence of thymus ceLIs on eryth­ netics, in press ropoiesis of parental marrow in irradiated hybrid mi ce. Proc Soc 117. Walford HL: Huma n B cell a lloantigenic systems: Their medical Exp Bioi Med 129:4 17, 1968 a nd biological significance, HLA System-New Aspects. Edited 107. Bearn JG: The thymus and the pituita ry ad.renal axis in anen­ by GB Ferrara. North-Holl a nd/Elsevier, 1977, pp 105-127 cephaly. Br J Exp Pathol 49:136-]44, 1968 118. Smith GS, Walford RL: Influence of the main histocompatibility 108. Strehler B, Hil'sch G, Gusseck D, J ohnson R, Bick M: Codon- complex on ageing in mice. Nature 240:727-729, 1977