- Email: [email protected]
Estradiol regulation of thymic lymphocyte function in the rat: Mediation by serum thymic factors
J. .\rrroicl Bir&w. Vol. 16. pp. 683 lo 690. I982 Prinled in Great Briuin. All rights resrrved
0022-4731/82/050683-08#)3.00/0 Copyright Q 1982 Pergamon Press Ltd
ESTRADIOL REGULATION .OF THYMIC LYMPHOCYTE FUNCTION IN THE RAT: MEDIATION BY SERUM THYMIC FACTORS CHARLES J. GttosswN*t$, LEON J. SHOLITON+$, and GARY A. ROSELLEII~~ *Medical Research Service, VA Medical Center; tPhysiology Dept., UC College of Medicine, $Dept. of Biological Sciences, Xavier University, Cincinnati, OH, SDivision of Endocrinology and Metabolism, Dept. of Internal Medicine, UC College of Medicine, Cincinnati, OH and I~lnfectious Diseases Section, VA Medical Center and TDivision of Infectious Diseases, Dept. of Internal Medicine, UC College of Medicine, Cincinnati, OH, U.S.A.
(Received 29 June 1981)
SUMMARY Estradiol (E2) can depress the function of the thymic lymphocytes. To determine if this response to a gonadal steroid is regulated directly or indirectly, thymic lymphocytes were incubated in vitro for 3 days in the presence of the mitogens concanavalin A (Con A) or phytohaemagglutinin (PHA) + tissue culture media + 20:/, specific rat serum fractions and pulse labelled with tritiated thymidinc. Rat serum fractions were prepared from control, castrate, thymectomiad (TX) and castrate-TX animals as well as from similar groups of animals treated in uiuofor three days with physiological doses of El. It was found that there was a significcant enhancement of thymocyte blastogenesis in cultures incubated with castrate rat serum + Con A or PHA vs. control serum cultures (P < 0.001). Direct replacement of Es to castrate sera in vitro at physiological concentrations failed to depress thymocyte blastogenesis to noncastrate levels, Sera prepared from castrate animals treated with E2 at physiological concentrations was successful in depressing the blastogenic response to noncastrate levels. Sera from TX animals did not enhance Con A induced blastogenesis, but the PHA response was significantly increased (P < 0.01). This effect was lost utilizing sera from TX animals treated with Es. It is concluded that thymocyte function is regulated by serum factors which have their origin in the thymus, and that these factors are modulated by the gonadal steroid estradiol.
mune system, specifically through an effect on T-cells
INTRODUCTION
It has been well documented that the function of the cellular immune system is depressed by estradiol (E2) [l-6] and stimulated by castration [l-2,7-9] while thymic weight is increased upon castration and decreased by exogenous E2 treatment. Recently we demonstrated that the rat thymus contains specific, high affinity E2 receptors which are present only in the thymic reticuloepithelial matrix and not in the thymocyte [lO-133. As the T-cell is one of the major functional units of the cellular immune system, the action of Et may well be at this level. There are two possible modes of action by which this hormone could produce an affect on T-cells. There may be a direct effect via a receptor mediated process or an indirect effect via modulation of some intermediate substance produced at another site. The objective of the present study is to determine the mode of action by which estradiol modulates the cell mediated im* Presented. in part. at the 12th Annual Meeting of the Society jar the Study qf Reprodtrction. Quebec, Canada 1979 Address Correspondence to: Charles J. Grossman, Ph.D. VA Medical Center Medical Research Service (151) 3200 Vine Street Cincinnati. OH 45220. U.S.A. 683
of thymic origin.
METHODS
Anima’ mcde’ Male Holtzman rats were obtained from Harlen Industries of Indiana. The animals were l-3 months old and weighed approximately 150g at the time of use. The rats were maintained on a standard laboratory diet and a 13: 11 photoperiod flights on 0500 h to 1800). Surgical procedures
All surgical procedures were carried out under ether anesthesia. Groups consisting of 20 animals were either castrated, adrenalectomized or thymectomized 3 weeks prior to their use. If more than one surgical treatment was used, 2 weeks were allowed between procedures. Animals to be thymectomized were shaved and bathed in 70% ethanol. All surgical instruments were submerged in 70% ethanol between procedures. Thymectomies were performed as described by Sherman et al.[14] with two baby Allis foreceps used to grasp the thymus and remove it from
CHARLES J. GROSSMANet al.
684
the incision. All groups that were adrenalectomized were fed physiologic saline in place of water to compensate for their increased salt loss. For all procedures adequate sham operated controls were employed. Steroid treatment For some studies ex~~~ntal animals (20 per group) were surgically prepared and these animals along with intact controls were treated daily with 15 pg of Ez prepared by dissolving 50 ccgof the steroid in 1 ml of corn oil. Animals were injected subcutaneously with 0.3 ml of oil or steroid in oil/day for three days. Injection of 15 pg/day resulted in serum levels of 63 pg/ml (results kindly supplied by BioScience La~ratori~ of Van Nuys, ~~ifomia and Dr. U. Goebelsman of the University of Southern California) which is within the physiological range for the rat (6-80 pg/ml) [lS, 163. In all experiments, animals were sacrificed by d~pitation. Preparation
of thymacytes
Thymi were removed from intact male rats under sterile ~nditions and im~iately placed in Eagle’s minimal essential media with Earle’s salts (MEM) in Petri dishes. Each thymus was then placed on a sterile 80 mesh stainless steel screen and teased apart using two sterile dissecting needles. The released thymocytes were washed through the screen using MEM. collected in a sterile Petri dish, centrifuged at 4OOg for 10 min, and the supernatant decanted. The thymocyte pellet was then resuspended in fresh MEM and washed three times as described above. The washed cells were diluted in MEM containing I mM L-glutamine, 100 units/ml penicillin, 100 ccc/ml streptomycin, and 20% rat serum (as specified below) to a final concentration of 12.5 x 10s cells/ml. Serum selection and cell oiability
Prior to the start of these studies, standard in vitro assays were attempted using both human and fetal calf sera. Both of these sera proved to be unsatisfactory due to low cell viability after 3 days incubation. We therefore used rat serum which allowed greater than 95% of the cells originally placed in the culture to remain viable after 3 days incubation as measured by Trypan blue dye exclusion. Preparation
of rat serum
Sera were prepared from the various groups of rats by decapitating the animals, collecting the blood, and allowing it to clot at room temperature. It WBS thm centrifuged at 8OOg for 15 min. decanted and stored at -30°C. Before use, the sera were thawed, sterilized by filtration (0.45 micron filter, Nillipore Corp., Bedford, MA) and heat inactivated at 56°C for 3Omin in a constant temperature water bath. Bhstoyenic
ussays
Blastogenic assays were carried out in triplicate in
96 well microtiter plates (A,S Nunc. Roskilde, Denmark) using 0.2 ml of cell suspension per well (2.5 x 10’ cells). Ten microliters of varying concentrations of the mitogens phytohaema~lutinin-M (PHA, 1:s and A:2 dilutions: Difco Laboratories, Detroit, MI) and concanavalin A (Con A, 2 and 5 pg, Pharmacia Fine Chemical, Piscataway, NJ) were added to the appropriate cultures. These dilutions were chosen because they consistently yielded a strong response in this assay system. After three days of incubation at 37’C in a 5O, CO2 atmosphere, the cuttures were pulsed labelted for two hours with 1 &i[methyC3H]-thymidine (SA 6 Ciimmol, Schwarz-Mann, Orangeburg, NY). The cells were then collected on glass fiber filters using a multiple automated sample harvester (MASH II, Mi~robiological Products, Walkersville, MD). The amount of incorporated radioactivity was determined in a Beckman liquid scintillation counter and the results expressed as disintegrations per minute (d.p.m.1. In certain experiments, estradiol was added directly to the cell cultures so that the final concentration was in the physiological range (15 pg/ml) [is, 161 of culture medium. Controls consisted of cell cultures with all tested sera in the absence of mitogen. Statistical
analysis
Results are a compilation of several ex~irnen~ each of which included concomitantly run controls. Data were analyzed using the Student’s c-test for unpaired samples with 0.05 as the upper level of significance. In all cases, maximum d.p.m./serum sample/ mitogen dosage was used for statistical analysis. This method is utilized in immunological assay throughout the literature to eliminate biological variability between preparations. RESULTS
Control cells, regardless of the type of rat sera used, remained unstimulated and the results were not significantly different from each other (Table 1). Upon the addition of mitogen. however, significant differences between serum groups were observed. Castrate sera signifiantly enhanced DNA synthesis in response to either PHA or Con A. In vitro replacement of Er to the castrate sera at physiologic concentrations did not abolish the effects of the castrate sera with respect to normal sera (Table 2). However, when castrate animals were treated for 3 days with subcutaneous injections of 15 pg Ez per day, their sera were unable to enhance blastogenesis regardless of whether Con A or PHA was used (Tables 3 and 4). When animals were thymectomized, their sera did not enhance Con A stimulated blastogenesis (Table 3) but did significantly enhance the PHA response (Table 4). Treatment irt siuo with EZ (15 &day) did not affect the results using Con A as the mitogcn, but did depress the lymphocyte response to normal levels using PHA. Sera prepared from animals that were
Regulation of thymic lymphocyte function
685
Table I. The effects of normal vs. castrate serum on normal thymocytes with or without the addition of the mitogens concanavalin A (Con A) or phytohaemagglutinin (PHA) Control
n x (d.p.m.) SEM
PHA
Con A
Normal serum
Castrate serum
Normal serum
Castrate serum
Normal serum
Castrate serum
77 383 +40
56 439 +49
77 823 +73
56 1814 k159
77 1269 fl84
56 5646 f769
*P < 0.001
‘P < 0.001
NS
* Significant as measured between normal and castrate serum using the Student’s r-test. n = number of separate in oirro blastogenic assays. The assays were run utilizing approximately 15 separate serum pools prepared from 10 animals each.
castrate vs adrenalectomized PHA or Con A as mitogen.
both castrate and thymectomized, or castrate-thymectomized and E2 treated (I 5 &day) produced no enhancement when either Con A or PHA was used as
mitogen. To ascertain the effects of adrenal steroids on this in t:itro blastogenic system, animals were either adrenalectomized or castrate-andrenalectomized and after sufficient recovery time were sacrificed and the sera prepared as described above. As can be seen in Table 5, the general effect of sera from either adrenalectomized alone or castrate-adrenalectomized animals was enhancement of DNA synthesis regardless of the mitogen used. It can also be seen that there were no significant differences between the effects of sera from
Although a number of interpretations can be drawn from these studies, three conclusions may be made. First thymocyte function as measured by a standard in oitro blastogenic assay is modulated by serum substances. !Second, the production, release, activation or inactivation of these substances is affected by El. Third, the thymus is involved in some way, possibly as the source of these substances.
(E,) to cultures of normal
Concanavaien
A
Normal Serum
R (d.p.m.) SEM n
Castrate Serum
15 p&ml E2
Control
15 pg/ml EZ
1269
626
5646
5108
k184 77
k 105 8
f 169 56
f 1047.3 23
P-NS*
P-NS* P > 0.001t Phytohaemagglutinin
Normal Serum
X (d.p.m.) SEM n
using either
DISCUSSION
Table 2. The effect of in vitro addition of estradiol thymocytes
Control
animals
Castrate Serum
Control
15 pgiml E,
823
761
1814
1424
+74 77
f166 8
*159 56
k213.9 23
P-NS*
Control
15 pg/ml E,
P-NS’ P > 0.001t
* There was no significant difference between the control and 15 pg/ml E2 within any group. t There was a significant difference between control vs. control, control vs 15 pg/ml E2 and 15 pg.ml E2 vs 15 pg/ml E2 between normal and castrate groups. n = number of separate in citro blastogenic assays. The assays were run utilizing approximately 15 separate serum pools prepared from 10 animals each. Student’s r-test was utilized for statistics.
686
CHARLESJ.
GROSSMANer al.
Table 3. Concanavalin A SIONICICANCI
OOWADS
rnvaus
r csll
DPM ff f SIYJ
NOINALS
CASTRATI
WYMlCTOMY
rNraLCro*Y t Ij Is
CASIRAll rNvMKroM*
cAsr1)*r# rnwecroav
t r, 88
cl r
0
cl 0
COMlARlD NOIYALS
10’
CAsrRArf!
I169 c e 0.001
f w ” = 77 5444
cl
0
El
0
$I
0
E%
0
H
0
* ¶97
‘$I
0
t 135
* 749
l <
0.001
n 0 86 790 2 II5
NS
c < 0.05
NS
r * 0.001
NS
P * 0.05
NI
c * 0.01
NI
c * 0.01
II=6 1091 f 143 n = 26 (86 t 191 n = 10 1014 0 * 18 547 I) = to
Pooled sera were prepared from groups of normal or surgically manipulated rats. After clotting, centrifqation and millipore filtering, the sera were heat inactivated at 56’C for 30 min. Thymocytes were extracted from normal rat thymi and incubated in blastogenic assay at a final concentration of 2.5 x 10’ cells per assay. Each assay well also contained 2OS: rat serum plus the mitogen Con A, minimal exsential media with Earles salts. 1 mM L-glutamine, penicillin and streptomycin. Incubation was for 3 days at 37°C after which time the cultures were pulse labelled for 2 h with tritiated thymidine, collected on glass fiber filters and counted for incorporated radioactivity.
Inspection of Tables 3 and 4 reveal variations in the depending on whether PHA or Con A is employed as mitogen. Since these differences may be due to a differential effect of these mitocens on thymocyte subpopulations, the results will be evaluated separately. results
Con A Removal of sex steroids by castration results in a significant increase in thymic size [7-93 and in our studies, the production of a serum substance which enhances the response to the mitogen, Con A (Table 3). Replacement by subcutaneous injection with physiological levels of El reverses these efkcts. Smce these e&c& take place in the ma& rat, it is conceivable that the gonadal T is aromatized into E2 in the periphery. While such an tiect has never been satisfactorily proven in the rat, it has been clearly demonstrated in other tissue[17-19) indudiag those of the human [20,21]. Measurement of steroid levels in our normal male rats resulted in a level of
8.8 pg/ml for Ez and 410 ng/DI for T. These levels decreased in the castrate male rat to 3.9 pg/ml for E2 and c 10 ng/DI for T (results kindly supplied by BioScience Laboratories, Van Nuys, California and Dr U. Goebelsman of the University of Southern California). While these results do not absolutely prove that T is aromatized to E2, they do suggest this to be the case especially in the light of the recent findings from our laboratory showing that rat thymic homogenates can metabolize testosterone 1223. Thymcctomy abolished the effects of castration and therefore, the thymus is probably the source of the active serum factor. Additional evidence for a thymic source is supplied by a variety of surgical manipulations. Serum fractions prepared from animals after thymectomy + castration, thymectomy + Ez injection, or thymectomy + castration + Er injection produced no greater effect than thymectomy alone. These results support the hypothesis that Ez functioning through thymic reticuloepithelial E2 receptors can depress the production and/or release of a thymic
687
Regulation of thymic lymphocyte function Table 4. Phytohaemagglutinin SIONICICANCI r CELL
COBCARCD
NORMALS 82J
0
NOlMAlS
t 73
c l 0.001
I) = 77
00
CASIIAll
c<
0
CAsrlArt
l
L* Rx
0.01
c<
0.05
624 f 140
NS
7 * 0.01
NS
c S 0.001
NS
c * 0.01
I) = IO
743 t7 I) = I8
614
0
m*rrcrorr
c<
I) = 24
0
rnwtcromr
c < 0.05
12J9 f 157
0
cAsr8Arf
0.001
NS
00 rnvmcrohw + t, Ia
ro* CAsrRAres
f I49 II = IO
Pooled sera and thymocytes were prepared as described in Table 3. Each assay well also contained 20:< rat serum plus the mitogen PHA minimal essential media with Earles salts, 1 mM L-glutamine, penicillin and streptomycin. Incubation was for 3 days at 37°C after which time the cultures were pulse labelled for 2 h with tritiated thymidine, collected on glass fiber filters and counted for incorporated radioactivity. Table 5. Effects of adrenalectomy
(Adx) and castration on serum enhancement blastogenesis
of
Concanaualin A
Normal serum x (d.p.m.) 1269 SEM +184 n 77 Significant with respect to: Normals Castrate
z (d.p.m.) SEM
Normal serum 823 +73
II 77 Significant with respect to: Normals Castrate
Castrate serum
Adx serum
Castrate-Adx serum
5646 + 769 56
7647 f 2702 8
10644 f 2269 8
P < 0.001 -
P < 0.001 NS
P < 0.001 P < 0.05
Phyrohaemagglu~inin Castrate serum Adx serum 1814 2695 *159 f631
Castrate-Adx serum 3429 f 558
56
8
8
P < 0.001
P < 0.001
P < 0.001 P < 0.001
-
NS
II = number of separate in cirro blastogenic assays. The assays were run utilizing approximately 15 separate serum pools prepared from 10 animals each. Student’s r-test was utilized for statistics.
688
CHARLESJ. GROSSMANet al.
castrate animals with Er reverses the increased thymic weight [7-9). Estrogens have also been reported to suppress the PHA ceil mediated immune system. For example, treatment with E2 depresses skin reactivity in guinea pigs [4,5]. PHA (Table 4) stimulation of thymocytes is significantly enhanced by a serum substance produced in delays corneai aliograft rejection in rabbits [6] and skin ailograft rejection in mice Cl], while reduction in castrate rats compared to serum from intact animals. circulating steroid levels following gonadectomy subE2 treatment in uiuo at physiological concentrations stantially enhances tissue rejection in male and female inhibits the production of this substances in the mice [ 1,2]. thymus, resulting in a thymocyte biastogenic response In pregnant humans depression of the cell mediated that is not significantly different from that seen using immune system results in lengthening of skin homoserum from noncastrate animals. Interestingly, serum graft rejection times [3] and a decrease in reactivity to from thymectomized animals produced a significant malignancies, certain micro-organisms [3 1,321 and a enhancement of PHA induced DNA synthesis above reduction of phytohaemogglutinin induced iymphonormals (1239 d.p.m. vs 823 d.p.m.). This enhancement was not seen utilizing sera from any of the foi- cyte transformation [32]. Further, in pregnant mice lowing groups: castrate rata treated in uiuo with Ez contact allergy sensitivity to picryl chloride is decreased [33,34] and mixed lymphocyte reactions are (15 pg/day) (624 d.p.m.), castration-thymectomy depressed [35]. (743 d.p.m.), or castration-thymectomy-Ez (15 pg/ These previous studies are essentially in agreement day) treatment (614d.p.m.). These observations are with our present work, and as reported earlier [36], in different from those seen with Con A where serum that we show an E2 reversible effect on in vitro thyfrom thymectomized animals had no effect on thymocyte DNA synthesis. We attempt to explain these mocyte function mediated by a serum substance possresults by reporting that while Con A has been shown ibly originating in thymic epitheiial ceils. These data to stimulate a subpopulation of suppressor would also support the work done by Kruisbeek[37] T-ceils [23,24], PHA is less specific. Presumably horwho showed a substance produced by rat thymic mones produced by the thymus could effect these subepitheiiai cells in tissue culture that effects T-cell populations differently and thus the absence of those maturation. hormones in thymectomized serum might also elicit It is well established that the thymus is the source different responses in these subpopulations. of hormones such as thymosin [38] and others [39]. Measurement of serum fractions for thymosin a 1 As expected, serum from adrenaiectomized animals also enhanced mitogen induced DNA synthesis even levels indicated that there was no significant difference between this hormone in normal serum in the absence of castration (Table 5). Castration and (138pg/lOO~i f 3.79, n = 10) and castrate serum adrenaiectomy were additive and enhanced thytno(146 pg/loO pi It 7.66, n = 9) as determined by radiocyte biastogenesis significantly more than sera preimmunoassay (results kindly supplied by Dr John pared from castrate or adrenalectomized animals McClure of Dr Alan Goldstein’s group at George alone. Washington University). Thus the thymic serum subIn the rat the adrenal produces some estrogen [2S] stance modulated by estradiol is something other and therefore, adrenaiectomy would assist in decreasthan thymosin a 1. ing plasma concentrations of this hormone. Since rat adrenals are also a source of glucocorticoids, adrenaIt has also been reported that substances produced by the thymus can affect the function of other organ lectomy would reduce circulating levels of these hormones as well. Giucocorticoids have been reported to systems. Thymectomy in the golden hamster in the prenatal period results in a wasting disease presumdirectly affect T-ceils through cytopiasmic receptor interactions [26] and the absence of giucocorticoids in ably due to a lack of some thymic derived substance [14]. In a recent paper by Michael, it is sera could enhance biastogenesis. Since sera prepared reported that a thymic-pituitary-ovarian axis exists from adrenaiectomized-castrate animals has minimal levels of sex steroi,ds and glucocorticoids, it is to be which could explain ovarian disturbances in thymectomized animals 140-J.Since our research has shown a expected that this group would have a maximal biasdirect interaction of the gonads on thymocyte functogenic response. tion, this axis may be bidirectional and might even be Gonadai influences on the thymus are well estabregulated by feedback of the sex hormones on the lished. E2 can directly influence the thymic epithelium thymic epithelium. as well as thymocytes [27,28] and treatment of rats The concept of sex hormones acting at the level of with El causes destruction of thymocytes, atrophy of thymic lobes, an increase in thymic fat as observed by the thymic epithelium is not new. DeVries and Hijmans[41] postulated such a mechanism from studies microscopy [9] and thymocyte pyknosis and karyorin NZB mice which spontaneously develop a haemorhexis[29]. Treatment of male and female rats with The fact that E2 results in a decrease in thymic weight and an acute iytic anemia or glomerulo-nephritis. lymphopenia [30] while castration produces a signifi- such mice also show thymic medullary lesions seemed cant increase in thymic weight [7-93. Treatment of to indicate a possible connection between the thymic
source of a serum factor that enhances biastogenesis induced by Con A.
689
Regulation oJ thymic lymphocyte function and the autoimmune disease. As has been pointed out by Raveche (‘f u/.[42] differentiation of particular lymphocyte subpopulations could be stimulated by factors from the thymic epitheiium as a result of interactions with sex hormones. Certainly the presence of specific thymic sex steroid hormone recep tors lends support to this hypothesis. In summary, we believe that in the rat, E2 can effect the function of the cell mediated immune system by modulating serum factors produced by the thymus and which, in turn, can enhance Con A and PI-IA sensitivity oJ normal thymocytes. The clinical significance of these results has not, as yet. been clarified but may have widespread implications in the etiology of a variety of disease states.
epithelium
Acknolllrdgc~n?rnt.s--The authors wish to express their appreciation to Mrs Patricia Uebel and Mrs Peggy Warth Jor their expert technical assistance. to Mr Richard L. Clark for his help with the preparation of the tables, and to Mrs Novie Fricke and MS Donna Denham for typing of this manuscript. This work was supported by funds supplied by the Veterans Administration. REFERESCES
1. GraKR. J.. Hildemann W. H. and Snell G. D.: Histocompatibility genes of mice. Tr~lnsplantnrion 4 (1966) 425-437. 2. Graff R. J.. Lappe M. A. and Snell G. D.: The influence of the gonads and adrenal glands on the immune response to skin grafts. Trcmspltrnrution 7 (1969) 105-l 11. 3. Anderson R. H. and Monroe C. W.: Experimental study of the behavior of adult human skin homografts during pregnancy. Am. J. Ubsrer. G_rnecof. 84 (1962) 1096-I 103. 4 Lurie M. B.. Harris 1. N.. Abramson S. and Allison J. M.: Constitutional factors in resistance to infection II. The effect of estrogen on tuberculin skin sensitivity and on the allergy of the internal tissues. Am. Rev. Tuberculosis 59
(19491186-197.
5. Kappas A.. Jones H. E. H. and Roitt I. M.: Effects of steroid sex hormones on immunological phenomena. iVnt1tre 198 (1963) 90’. 6. Waltman S. R.. Burde R. M. and Berrois J.: Prevention of cornea1 homograft rejection by estrogens. Transplantatiorl 11 (19711 194-196. 7. Westphak U.: Infbienck of various endogenous factors of CBG activity in the rat. In ~reroid-Prorein Inreraerinrrs. SpringerlVerlag. New York Monogr. Endocr. 4 (1971) 249-296.
8. Chiodi H.: Thymus hypertrophy induced by castration in old male rats and mice (abstract) Fed. Proc. 35 t 1976) 277. 9. Choidi H.: The relationship between the thymus and the sexual organs. Endorrinolog!~ 26 (1940) 107-l 16. 10. Grossman C. J. and Nathan P.: Specific estrogen receptor in rat and heel thymus (abstract) Fed. Proc. 36 (19771 5119. 1336. 11. Grossman C. J. and Nathan P.: Specific estrogen receptor in epithelial-reticular cells of the rat thymus. Phpioloyisr 20 (1977) 38. 12. Grossman C. J.. Sholiton L. J. and Nathan P.: Rat thymic estrogen receptor-I. Preparation. location and physiochemical properties. J. steroid Biochem. 11 (1979) 1’33-1240.
13. Grossman C. .I.. Sholiton L. J.. Blaha G. C. and Nathan P.: Rat thymic estrogen receptor-II. Physio-
logical
properties. J. srwoid
Binchcm
I I (1979)
1241.-2146.
14. Sherman J. D.. Adner M. D. and Dameshek W.: Elects of thymectomy on the colden hamster (~esf~~ri~~rus aururus). I. Wasting Disease. Blood 22 ( 1963) 252-271. 15. Kalra S. P. and Kalra P. S.: Temporal interrelationships among circulating levels of estradiol. progesterone and LH during the rat estrous cycle: effects of exogenous progesterone. ~ndo~rjn~~ogy 95 (1974) 1711-1718. 16. Butcher R. L., Collins W. E. and Fugo N. W.: Plasma concentration oJ LH, FSH, prolactin. progesterone and estradiol-17/I throughout the 4-day estrous cycle of the rat. Endocrinology 94 ( 1974) 1704 1708. 17. George F. W., Tobleman W. T., Milewich L. and Wilson J. D.: Aromatase activity in the developing rabbit brain. Endocrinobav 102 (1978) 86-91. 18 Sholiton L. J., Taylor 8. B. and Lewis H. P.: The uptake and metabolism of labelled testosterone by the brain and pituitary of male rhesus monkey (Mocaca mularra). Sreroids 24 (1974) 537-547. 19. Wilson J. D. and Gloyna R. E.: The intranuclear metabolism of testosterone in the accessory organs of reproduction. Recenr Prog. harm. Res. 26 (1970) 309-330. 20. Kelch B. P., Jenner M. R.. Weinstein R.. Kaplan S. L. and Grumbach M. M.: Estradiol and testosterone secretion by human, simian and canine Testes, in males with hy~gonadism and in male p~udohermaphrodites with the feminizing testes syndrome. J. clin. Invest. 51 (1972) 824-830. 21. Payne H. H.. Kelch R. P., Musich S. S. and Halpern M. E.: Intratesticular site of aromatization in the human. J. clin. Endocr. Metub. 42 (1976) 1081-1087. 22. Sholiton L. J., Grossman C. J. and Taylor B.: Rat thymic homogenates convert testosterone to androgenic metabolites. J. steroid Biochem. 13 (1980) 1365-1367. 23. Rich R. R. and Pierce C. W.: Biological expressions of lympho~te activation. III Suppression of plaqueforming cell responses in virro by supernatant Ruids Jrom concanavalin A-activated spleen cell cultures. J. fmmunol. 112 (1974) 1360-1368. 24. Hallgren H. M. and Yumis E. J.: Suppressor lymphocytes in young and aged humans. J. ~mmu~o~. 11% (1977) 2004-2008. 25. Shaikh A. A. and Shaikh S. A.: Adrenal and ovarian steroid secretion in the rat esters cycle temporally related to gonadotropins and steroid levels found in peripheral plasma. ~~do~rjnology 96 (1975) 37-44. 26. Crabtree G. R.. Gillis S.. Smith K. A. and Munck A.: Glucocorticoids and immune responses. Arthris. Rheum. 22 (1979) 12461256. 27. Gregoire C. and Duchateau G.: A study of lymphoethelial symbosis in thymus. Reactions of the lymphatic tissue to extracts and to implants oJ epithelial components of thymus. Archs Biol. 67 (1956) 269-296. 28. Ahlqvist J.: Endocrine influences on the lymphatic organs. immune responses, inflammation and autoimmunity. Acra. endocrin, ~o~enh 83 (1976) Suppl. 206. 29. Kuriharra N., Takokoro S. and Ogawa H.: Studies on hormonal actions of dihydroxyprogesterone acetophenide, estradiol and their mixtures. Jpn. J. Pharmacol. .I
.
22 (1972) 43-58.
C. R., Perkins F. T. and Bishop D.: The effect 30. Franks ^ . O! sex hormones on the growth of HELA tumor nodules in male and female mice. Br. J. Cancer 31 (1975) 100-l 10. 31. Finn R., St. Hill C. A., Govan A. J., RolJs 1. G., Gurney F. J. and Denye V.: Immunological responses in pregnancy and survival of fetal homograft: Br. Med.-J. 3 (1972) 150-152. 32. Purtilo D. T., Hal&en H. M. and Tunis D. J.: Depressed maternal lymphocyte response to phytohaemo-
690
33. 34.
35.
36.
37.
38.
CHARLES J. GRCSSMAN
glutinin in human pregnancy. Lancer I (1972) 762771. Febris N.: Immunological reactivity during pregnancy in the mouse. Experienta 29 (1973) 610-612. Carter J.: The effect of progesterone, oestradiol and HCG on cell-mediated immunitv in oreunant mice. J. Reprod. Fert. 46 (1976) 21 l-216: - Kasakura S.: Is cortisoi responsible for inhibition of MLC reactions on pregnancy plasma. Nature 246 (1973) 496-497. Grossman C. J., Sholiton L. J, Nathan P. and Roselle G. A.: Estradiol depression of thymic lymphocyte function: mediation by a serum factor. Biol. Reprod. 20 (abstract) (1979).13A. Kruisbeek A. M.: Thymic factors and T-ceil maturation in vitro: a comparison of the effects of thymic epithelial cultures with thymic extracts and thymus dependent serum _ factors. Thymes 1 (1979) 163-185. White A. and Goldstein A. L.: The endocrine role of
et al.
the thymus and its hormone, thymosine. in the regulation of the growth and maturation of host immunological competence. Adr. hferab. Disord. 8 (1975) 359-374. _ 39. Comsa J.: Hormonal interactions of the thymus. Thymic Hormones. (Edited bv T. D. Luckev) Universitv Park Press, BalGmore (1953) pp. 59-89. . 40. Michael S. D.: The role of the endocrine thymus in female reproduction. Arthritis Rheum. 22 (1979) 1241-1245. 41. DeVries M. J. and Hijmans W.: A deficient deveiopment of the thymic epithelium and autoimmune disease in NZB mice. J. Path. Bact. 91 (1966) 487494.
42. Raveche E. S.. Tjio J. H., Boegel W. and Steinberg A. D.: Studies of the effects of sex hormones on autosomal and X-linked genetic control of induced and spontaneous antibody production. Arrhriris Rheum. 22 (1979) 1177-1187.
0022-4731/82/050683-08#)3.00/0 Copyright Q 1982 Pergamon Press Ltd
ESTRADIOL REGULATION .OF THYMIC LYMPHOCYTE FUNCTION IN THE RAT: MEDIATION BY SERUM THYMIC FACTORS CHARLES J. GttosswN*t$, LEON J. SHOLITON+$, and GARY A. ROSELLEII~~ *Medical Research Service, VA Medical Center; tPhysiology Dept., UC College of Medicine, $Dept. of Biological Sciences, Xavier University, Cincinnati, OH, SDivision of Endocrinology and Metabolism, Dept. of Internal Medicine, UC College of Medicine, Cincinnati, OH and I~lnfectious Diseases Section, VA Medical Center and TDivision of Infectious Diseases, Dept. of Internal Medicine, UC College of Medicine, Cincinnati, OH, U.S.A.
(Received 29 June 1981)
SUMMARY Estradiol (E2) can depress the function of the thymic lymphocytes. To determine if this response to a gonadal steroid is regulated directly or indirectly, thymic lymphocytes were incubated in vitro for 3 days in the presence of the mitogens concanavalin A (Con A) or phytohaemagglutinin (PHA) + tissue culture media + 20:/, specific rat serum fractions and pulse labelled with tritiated thymidinc. Rat serum fractions were prepared from control, castrate, thymectomiad (TX) and castrate-TX animals as well as from similar groups of animals treated in uiuofor three days with physiological doses of El. It was found that there was a significcant enhancement of thymocyte blastogenesis in cultures incubated with castrate rat serum + Con A or PHA vs. control serum cultures (P < 0.001). Direct replacement of Es to castrate sera in vitro at physiological concentrations failed to depress thymocyte blastogenesis to noncastrate levels, Sera prepared from castrate animals treated with E2 at physiological concentrations was successful in depressing the blastogenic response to noncastrate levels. Sera from TX animals did not enhance Con A induced blastogenesis, but the PHA response was significantly increased (P < 0.01). This effect was lost utilizing sera from TX animals treated with Es. It is concluded that thymocyte function is regulated by serum factors which have their origin in the thymus, and that these factors are modulated by the gonadal steroid estradiol.
mune system, specifically through an effect on T-cells
INTRODUCTION
It has been well documented that the function of the cellular immune system is depressed by estradiol (E2) [l-6] and stimulated by castration [l-2,7-9] while thymic weight is increased upon castration and decreased by exogenous E2 treatment. Recently we demonstrated that the rat thymus contains specific, high affinity E2 receptors which are present only in the thymic reticuloepithelial matrix and not in the thymocyte [lO-133. As the T-cell is one of the major functional units of the cellular immune system, the action of Et may well be at this level. There are two possible modes of action by which this hormone could produce an affect on T-cells. There may be a direct effect via a receptor mediated process or an indirect effect via modulation of some intermediate substance produced at another site. The objective of the present study is to determine the mode of action by which estradiol modulates the cell mediated im* Presented. in part. at the 12th Annual Meeting of the Society jar the Study qf Reprodtrction. Quebec, Canada 1979 Address Correspondence to: Charles J. Grossman, Ph.D. VA Medical Center Medical Research Service (151) 3200 Vine Street Cincinnati. OH 45220. U.S.A. 683
of thymic origin.
METHODS
Anima’ mcde’ Male Holtzman rats were obtained from Harlen Industries of Indiana. The animals were l-3 months old and weighed approximately 150g at the time of use. The rats were maintained on a standard laboratory diet and a 13: 11 photoperiod flights on 0500 h to 1800). Surgical procedures
All surgical procedures were carried out under ether anesthesia. Groups consisting of 20 animals were either castrated, adrenalectomized or thymectomized 3 weeks prior to their use. If more than one surgical treatment was used, 2 weeks were allowed between procedures. Animals to be thymectomized were shaved and bathed in 70% ethanol. All surgical instruments were submerged in 70% ethanol between procedures. Thymectomies were performed as described by Sherman et al.[14] with two baby Allis foreceps used to grasp the thymus and remove it from
CHARLES J. GROSSMANet al.
684
the incision. All groups that were adrenalectomized were fed physiologic saline in place of water to compensate for their increased salt loss. For all procedures adequate sham operated controls were employed. Steroid treatment For some studies ex~~~ntal animals (20 per group) were surgically prepared and these animals along with intact controls were treated daily with 15 pg of Ez prepared by dissolving 50 ccgof the steroid in 1 ml of corn oil. Animals were injected subcutaneously with 0.3 ml of oil or steroid in oil/day for three days. Injection of 15 pg/day resulted in serum levels of 63 pg/ml (results kindly supplied by BioScience La~ratori~ of Van Nuys, ~~ifomia and Dr. U. Goebelsman of the University of Southern California) which is within the physiological range for the rat (6-80 pg/ml) [lS, 163. In all experiments, animals were sacrificed by d~pitation. Preparation
of thymacytes
Thymi were removed from intact male rats under sterile ~nditions and im~iately placed in Eagle’s minimal essential media with Earle’s salts (MEM) in Petri dishes. Each thymus was then placed on a sterile 80 mesh stainless steel screen and teased apart using two sterile dissecting needles. The released thymocytes were washed through the screen using MEM. collected in a sterile Petri dish, centrifuged at 4OOg for 10 min, and the supernatant decanted. The thymocyte pellet was then resuspended in fresh MEM and washed three times as described above. The washed cells were diluted in MEM containing I mM L-glutamine, 100 units/ml penicillin, 100 ccc/ml streptomycin, and 20% rat serum (as specified below) to a final concentration of 12.5 x 10s cells/ml. Serum selection and cell oiability
Prior to the start of these studies, standard in vitro assays were attempted using both human and fetal calf sera. Both of these sera proved to be unsatisfactory due to low cell viability after 3 days incubation. We therefore used rat serum which allowed greater than 95% of the cells originally placed in the culture to remain viable after 3 days incubation as measured by Trypan blue dye exclusion. Preparation
of rat serum
Sera were prepared from the various groups of rats by decapitating the animals, collecting the blood, and allowing it to clot at room temperature. It WBS thm centrifuged at 8OOg for 15 min. decanted and stored at -30°C. Before use, the sera were thawed, sterilized by filtration (0.45 micron filter, Nillipore Corp., Bedford, MA) and heat inactivated at 56°C for 3Omin in a constant temperature water bath. Bhstoyenic
ussays
Blastogenic assays were carried out in triplicate in
96 well microtiter plates (A,S Nunc. Roskilde, Denmark) using 0.2 ml of cell suspension per well (2.5 x 10’ cells). Ten microliters of varying concentrations of the mitogens phytohaema~lutinin-M (PHA, 1:s and A:2 dilutions: Difco Laboratories, Detroit, MI) and concanavalin A (Con A, 2 and 5 pg, Pharmacia Fine Chemical, Piscataway, NJ) were added to the appropriate cultures. These dilutions were chosen because they consistently yielded a strong response in this assay system. After three days of incubation at 37’C in a 5O, CO2 atmosphere, the cuttures were pulsed labelted for two hours with 1 &i[methyC3H]-thymidine (SA 6 Ciimmol, Schwarz-Mann, Orangeburg, NY). The cells were then collected on glass fiber filters using a multiple automated sample harvester (MASH II, Mi~robiological Products, Walkersville, MD). The amount of incorporated radioactivity was determined in a Beckman liquid scintillation counter and the results expressed as disintegrations per minute (d.p.m.1. In certain experiments, estradiol was added directly to the cell cultures so that the final concentration was in the physiological range (15 pg/ml) [is, 161 of culture medium. Controls consisted of cell cultures with all tested sera in the absence of mitogen. Statistical
analysis
Results are a compilation of several ex~irnen~ each of which included concomitantly run controls. Data were analyzed using the Student’s c-test for unpaired samples with 0.05 as the upper level of significance. In all cases, maximum d.p.m./serum sample/ mitogen dosage was used for statistical analysis. This method is utilized in immunological assay throughout the literature to eliminate biological variability between preparations. RESULTS
Control cells, regardless of the type of rat sera used, remained unstimulated and the results were not significantly different from each other (Table 1). Upon the addition of mitogen. however, significant differences between serum groups were observed. Castrate sera signifiantly enhanced DNA synthesis in response to either PHA or Con A. In vitro replacement of Er to the castrate sera at physiologic concentrations did not abolish the effects of the castrate sera with respect to normal sera (Table 2). However, when castrate animals were treated for 3 days with subcutaneous injections of 15 pg Ez per day, their sera were unable to enhance blastogenesis regardless of whether Con A or PHA was used (Tables 3 and 4). When animals were thymectomized, their sera did not enhance Con A stimulated blastogenesis (Table 3) but did significantly enhance the PHA response (Table 4). Treatment irt siuo with EZ (15 &day) did not affect the results using Con A as the mitogcn, but did depress the lymphocyte response to normal levels using PHA. Sera prepared from animals that were
Regulation of thymic lymphocyte function
685
Table I. The effects of normal vs. castrate serum on normal thymocytes with or without the addition of the mitogens concanavalin A (Con A) or phytohaemagglutinin (PHA) Control
n x (d.p.m.) SEM
PHA
Con A
Normal serum
Castrate serum
Normal serum
Castrate serum
Normal serum
Castrate serum
77 383 +40
56 439 +49
77 823 +73
56 1814 k159
77 1269 fl84
56 5646 f769
*P < 0.001
‘P < 0.001
NS
* Significant as measured between normal and castrate serum using the Student’s r-test. n = number of separate in oirro blastogenic assays. The assays were run utilizing approximately 15 separate serum pools prepared from 10 animals each.
castrate vs adrenalectomized PHA or Con A as mitogen.
both castrate and thymectomized, or castrate-thymectomized and E2 treated (I 5 &day) produced no enhancement when either Con A or PHA was used as
mitogen. To ascertain the effects of adrenal steroids on this in t:itro blastogenic system, animals were either adrenalectomized or castrate-andrenalectomized and after sufficient recovery time were sacrificed and the sera prepared as described above. As can be seen in Table 5, the general effect of sera from either adrenalectomized alone or castrate-adrenalectomized animals was enhancement of DNA synthesis regardless of the mitogen used. It can also be seen that there were no significant differences between the effects of sera from
Although a number of interpretations can be drawn from these studies, three conclusions may be made. First thymocyte function as measured by a standard in oitro blastogenic assay is modulated by serum substances. !Second, the production, release, activation or inactivation of these substances is affected by El. Third, the thymus is involved in some way, possibly as the source of these substances.
(E,) to cultures of normal
Concanavaien
A
Normal Serum
R (d.p.m.) SEM n
Castrate Serum
15 p&ml E2
Control
15 pg/ml EZ
1269
626
5646
5108
k184 77
k 105 8
f 169 56
f 1047.3 23
P-NS*
P-NS* P > 0.001t Phytohaemagglutinin
Normal Serum
X (d.p.m.) SEM n
using either
DISCUSSION
Table 2. The effect of in vitro addition of estradiol thymocytes
Control
animals
Castrate Serum
Control
15 pgiml E,
823
761
1814
1424
+74 77
f166 8
*159 56
k213.9 23
P-NS*
Control
15 pg/ml E,
P-NS’ P > 0.001t
* There was no significant difference between the control and 15 pg/ml E2 within any group. t There was a significant difference between control vs. control, control vs 15 pg/ml E2 and 15 pg.ml E2 vs 15 pg/ml E2 between normal and castrate groups. n = number of separate in citro blastogenic assays. The assays were run utilizing approximately 15 separate serum pools prepared from 10 animals each. Student’s r-test was utilized for statistics.
686
CHARLESJ.
GROSSMANer al.
Table 3. Concanavalin A SIONICICANCI
OOWADS
rnvaus
r csll
DPM ff f SIYJ
NOINALS
CASTRATI
WYMlCTOMY
rNraLCro*Y t Ij Is
CASIRAll rNvMKroM*
cAsr1)*r# rnwecroav
t r, 88
cl r
0
cl 0
COMlARlD NOIYALS
10’
CAsrRArf!
I169 c e 0.001
f w ” = 77 5444
cl
0
El
0
$I
0
E%
0
H
0
* ¶97
‘$I
0
t 135
* 749
l <
0.001
n 0 86 790 2 II5
NS
c < 0.05
NS
r * 0.001
NS
P * 0.05
NI
c * 0.01
NI
c * 0.01
II=6 1091 f 143 n = 26 (86 t 191 n = 10 1014 0 * 18 547 I) = to
Pooled sera were prepared from groups of normal or surgically manipulated rats. After clotting, centrifqation and millipore filtering, the sera were heat inactivated at 56’C for 30 min. Thymocytes were extracted from normal rat thymi and incubated in blastogenic assay at a final concentration of 2.5 x 10’ cells per assay. Each assay well also contained 2OS: rat serum plus the mitogen Con A, minimal exsential media with Earles salts. 1 mM L-glutamine, penicillin and streptomycin. Incubation was for 3 days at 37°C after which time the cultures were pulse labelled for 2 h with tritiated thymidine, collected on glass fiber filters and counted for incorporated radioactivity.
Inspection of Tables 3 and 4 reveal variations in the depending on whether PHA or Con A is employed as mitogen. Since these differences may be due to a differential effect of these mitocens on thymocyte subpopulations, the results will be evaluated separately. results
Con A Removal of sex steroids by castration results in a significant increase in thymic size [7-93 and in our studies, the production of a serum substance which enhances the response to the mitogen, Con A (Table 3). Replacement by subcutaneous injection with physiological levels of El reverses these efkcts. Smce these e&c& take place in the ma& rat, it is conceivable that the gonadal T is aromatized into E2 in the periphery. While such an tiect has never been satisfactorily proven in the rat, it has been clearly demonstrated in other tissue[17-19) indudiag those of the human [20,21]. Measurement of steroid levels in our normal male rats resulted in a level of
8.8 pg/ml for Ez and 410 ng/DI for T. These levels decreased in the castrate male rat to 3.9 pg/ml for E2 and c 10 ng/DI for T (results kindly supplied by BioScience Laboratories, Van Nuys, California and Dr U. Goebelsman of the University of Southern California). While these results do not absolutely prove that T is aromatized to E2, they do suggest this to be the case especially in the light of the recent findings from our laboratory showing that rat thymic homogenates can metabolize testosterone 1223. Thymcctomy abolished the effects of castration and therefore, the thymus is probably the source of the active serum factor. Additional evidence for a thymic source is supplied by a variety of surgical manipulations. Serum fractions prepared from animals after thymectomy + castration, thymectomy + Ez injection, or thymectomy + castration + Er injection produced no greater effect than thymectomy alone. These results support the hypothesis that Ez functioning through thymic reticuloepithelial E2 receptors can depress the production and/or release of a thymic
687
Regulation of thymic lymphocyte function Table 4. Phytohaemagglutinin SIONICICANCI r CELL
COBCARCD
NORMALS 82J
0
NOlMAlS
t 73
c l 0.001
I) = 77
00
CASIIAll
c<
0
CAsrlArt
l
L* Rx
0.01
c<
0.05
624 f 140
NS
7 * 0.01
NS
c S 0.001
NS
c * 0.01
I) = IO
743 t7 I) = I8
614
0
m*rrcrorr
c<
I) = 24
0
rnwtcromr
c < 0.05
12J9 f 157
0
cAsr8Arf
0.001
NS
00 rnvmcrohw + t, Ia
ro* CAsrRAres
f I49 II = IO
Pooled sera and thymocytes were prepared as described in Table 3. Each assay well also contained 20:< rat serum plus the mitogen PHA minimal essential media with Earles salts, 1 mM L-glutamine, penicillin and streptomycin. Incubation was for 3 days at 37°C after which time the cultures were pulse labelled for 2 h with tritiated thymidine, collected on glass fiber filters and counted for incorporated radioactivity. Table 5. Effects of adrenalectomy
(Adx) and castration on serum enhancement blastogenesis
of
Concanaualin A
Normal serum x (d.p.m.) 1269 SEM +184 n 77 Significant with respect to: Normals Castrate
z (d.p.m.) SEM
Normal serum 823 +73
II 77 Significant with respect to: Normals Castrate
Castrate serum
Adx serum
Castrate-Adx serum
5646 + 769 56
7647 f 2702 8
10644 f 2269 8
P < 0.001 -
P < 0.001 NS
P < 0.001 P < 0.05
Phyrohaemagglu~inin Castrate serum Adx serum 1814 2695 *159 f631
Castrate-Adx serum 3429 f 558
56
8
8
P < 0.001
P < 0.001
P < 0.001 P < 0.001
-
NS
II = number of separate in cirro blastogenic assays. The assays were run utilizing approximately 15 separate serum pools prepared from 10 animals each. Student’s r-test was utilized for statistics.
688
CHARLESJ. GROSSMANet al.
castrate animals with Er reverses the increased thymic weight [7-9). Estrogens have also been reported to suppress the PHA ceil mediated immune system. For example, treatment with E2 depresses skin reactivity in guinea pigs [4,5]. PHA (Table 4) stimulation of thymocytes is significantly enhanced by a serum substance produced in delays corneai aliograft rejection in rabbits [6] and skin ailograft rejection in mice Cl], while reduction in castrate rats compared to serum from intact animals. circulating steroid levels following gonadectomy subE2 treatment in uiuo at physiological concentrations stantially enhances tissue rejection in male and female inhibits the production of this substances in the mice [ 1,2]. thymus, resulting in a thymocyte biastogenic response In pregnant humans depression of the cell mediated that is not significantly different from that seen using immune system results in lengthening of skin homoserum from noncastrate animals. Interestingly, serum graft rejection times [3] and a decrease in reactivity to from thymectomized animals produced a significant malignancies, certain micro-organisms [3 1,321 and a enhancement of PHA induced DNA synthesis above reduction of phytohaemogglutinin induced iymphonormals (1239 d.p.m. vs 823 d.p.m.). This enhancement was not seen utilizing sera from any of the foi- cyte transformation [32]. Further, in pregnant mice lowing groups: castrate rata treated in uiuo with Ez contact allergy sensitivity to picryl chloride is decreased [33,34] and mixed lymphocyte reactions are (15 pg/day) (624 d.p.m.), castration-thymectomy depressed [35]. (743 d.p.m.), or castration-thymectomy-Ez (15 pg/ These previous studies are essentially in agreement day) treatment (614d.p.m.). These observations are with our present work, and as reported earlier [36], in different from those seen with Con A where serum that we show an E2 reversible effect on in vitro thyfrom thymectomized animals had no effect on thymocyte DNA synthesis. We attempt to explain these mocyte function mediated by a serum substance possresults by reporting that while Con A has been shown ibly originating in thymic epitheiial ceils. These data to stimulate a subpopulation of suppressor would also support the work done by Kruisbeek[37] T-ceils [23,24], PHA is less specific. Presumably horwho showed a substance produced by rat thymic mones produced by the thymus could effect these subepitheiiai cells in tissue culture that effects T-cell populations differently and thus the absence of those maturation. hormones in thymectomized serum might also elicit It is well established that the thymus is the source different responses in these subpopulations. of hormones such as thymosin [38] and others [39]. Measurement of serum fractions for thymosin a 1 As expected, serum from adrenaiectomized animals also enhanced mitogen induced DNA synthesis even levels indicated that there was no significant difference between this hormone in normal serum in the absence of castration (Table 5). Castration and (138pg/lOO~i f 3.79, n = 10) and castrate serum adrenaiectomy were additive and enhanced thytno(146 pg/loO pi It 7.66, n = 9) as determined by radiocyte biastogenesis significantly more than sera preimmunoassay (results kindly supplied by Dr John pared from castrate or adrenalectomized animals McClure of Dr Alan Goldstein’s group at George alone. Washington University). Thus the thymic serum subIn the rat the adrenal produces some estrogen [2S] stance modulated by estradiol is something other and therefore, adrenaiectomy would assist in decreasthan thymosin a 1. ing plasma concentrations of this hormone. Since rat adrenals are also a source of glucocorticoids, adrenaIt has also been reported that substances produced by the thymus can affect the function of other organ lectomy would reduce circulating levels of these hormones as well. Giucocorticoids have been reported to systems. Thymectomy in the golden hamster in the prenatal period results in a wasting disease presumdirectly affect T-ceils through cytopiasmic receptor interactions [26] and the absence of giucocorticoids in ably due to a lack of some thymic derived substance [14]. In a recent paper by Michael, it is sera could enhance biastogenesis. Since sera prepared reported that a thymic-pituitary-ovarian axis exists from adrenaiectomized-castrate animals has minimal levels of sex steroi,ds and glucocorticoids, it is to be which could explain ovarian disturbances in thymectomized animals 140-J.Since our research has shown a expected that this group would have a maximal biasdirect interaction of the gonads on thymocyte functogenic response. tion, this axis may be bidirectional and might even be Gonadai influences on the thymus are well estabregulated by feedback of the sex hormones on the lished. E2 can directly influence the thymic epithelium thymic epithelium. as well as thymocytes [27,28] and treatment of rats The concept of sex hormones acting at the level of with El causes destruction of thymocytes, atrophy of thymic lobes, an increase in thymic fat as observed by the thymic epithelium is not new. DeVries and Hijmans[41] postulated such a mechanism from studies microscopy [9] and thymocyte pyknosis and karyorin NZB mice which spontaneously develop a haemorhexis[29]. Treatment of male and female rats with The fact that E2 results in a decrease in thymic weight and an acute iytic anemia or glomerulo-nephritis. lymphopenia [30] while castration produces a signifi- such mice also show thymic medullary lesions seemed cant increase in thymic weight [7-93. Treatment of to indicate a possible connection between the thymic
source of a serum factor that enhances biastogenesis induced by Con A.
689
Regulation oJ thymic lymphocyte function and the autoimmune disease. As has been pointed out by Raveche (‘f u/.[42] differentiation of particular lymphocyte subpopulations could be stimulated by factors from the thymic epitheiium as a result of interactions with sex hormones. Certainly the presence of specific thymic sex steroid hormone recep tors lends support to this hypothesis. In summary, we believe that in the rat, E2 can effect the function of the cell mediated immune system by modulating serum factors produced by the thymus and which, in turn, can enhance Con A and PI-IA sensitivity oJ normal thymocytes. The clinical significance of these results has not, as yet. been clarified but may have widespread implications in the etiology of a variety of disease states.
epithelium
Acknolllrdgc~n?rnt.s--The authors wish to express their appreciation to Mrs Patricia Uebel and Mrs Peggy Warth Jor their expert technical assistance. to Mr Richard L. Clark for his help with the preparation of the tables, and to Mrs Novie Fricke and MS Donna Denham for typing of this manuscript. This work was supported by funds supplied by the Veterans Administration. REFERESCES
1. GraKR. J.. Hildemann W. H. and Snell G. D.: Histocompatibility genes of mice. Tr~lnsplantnrion 4 (1966) 425-437. 2. Graff R. J.. Lappe M. A. and Snell G. D.: The influence of the gonads and adrenal glands on the immune response to skin grafts. Trcmspltrnrution 7 (1969) 105-l 11. 3. Anderson R. H. and Monroe C. W.: Experimental study of the behavior of adult human skin homografts during pregnancy. Am. J. Ubsrer. G_rnecof. 84 (1962) 1096-I 103. 4 Lurie M. B.. Harris 1. N.. Abramson S. and Allison J. M.: Constitutional factors in resistance to infection II. The effect of estrogen on tuberculin skin sensitivity and on the allergy of the internal tissues. Am. Rev. Tuberculosis 59
(19491186-197.
5. Kappas A.. Jones H. E. H. and Roitt I. M.: Effects of steroid sex hormones on immunological phenomena. iVnt1tre 198 (1963) 90’. 6. Waltman S. R.. Burde R. M. and Berrois J.: Prevention of cornea1 homograft rejection by estrogens. Transplantatiorl 11 (19711 194-196. 7. Westphak U.: Infbienck of various endogenous factors of CBG activity in the rat. In ~reroid-Prorein Inreraerinrrs. SpringerlVerlag. New York Monogr. Endocr. 4 (1971) 249-296.
8. Chiodi H.: Thymus hypertrophy induced by castration in old male rats and mice (abstract) Fed. Proc. 35 t 1976) 277. 9. Choidi H.: The relationship between the thymus and the sexual organs. Endorrinolog!~ 26 (1940) 107-l 16. 10. Grossman C. J. and Nathan P.: Specific estrogen receptor in rat and heel thymus (abstract) Fed. Proc. 36 (19771 5119. 1336. 11. Grossman C. J. and Nathan P.: Specific estrogen receptor in epithelial-reticular cells of the rat thymus. Phpioloyisr 20 (1977) 38. 12. Grossman C. J.. Sholiton L. J. and Nathan P.: Rat thymic estrogen receptor-I. Preparation. location and physiochemical properties. J. steroid Biochem. 11 (1979) 1’33-1240.
13. Grossman C. .I.. Sholiton L. J.. Blaha G. C. and Nathan P.: Rat thymic estrogen receptor-II. Physio-
logical
properties. J. srwoid
Binchcm
I I (1979)
1241.-2146.
14. Sherman J. D.. Adner M. D. and Dameshek W.: Elects of thymectomy on the colden hamster (~esf~~ri~~rus aururus). I. Wasting Disease. Blood 22 ( 1963) 252-271. 15. Kalra S. P. and Kalra P. S.: Temporal interrelationships among circulating levels of estradiol. progesterone and LH during the rat estrous cycle: effects of exogenous progesterone. ~ndo~rjn~~ogy 95 (1974) 1711-1718. 16. Butcher R. L., Collins W. E. and Fugo N. W.: Plasma concentration oJ LH, FSH, prolactin. progesterone and estradiol-17/I throughout the 4-day estrous cycle of the rat. Endocrinology 94 ( 1974) 1704 1708. 17. George F. W., Tobleman W. T., Milewich L. and Wilson J. D.: Aromatase activity in the developing rabbit brain. Endocrinobav 102 (1978) 86-91. 18 Sholiton L. J., Taylor 8. B. and Lewis H. P.: The uptake and metabolism of labelled testosterone by the brain and pituitary of male rhesus monkey (Mocaca mularra). Sreroids 24 (1974) 537-547. 19. Wilson J. D. and Gloyna R. E.: The intranuclear metabolism of testosterone in the accessory organs of reproduction. Recenr Prog. harm. Res. 26 (1970) 309-330. 20. Kelch B. P., Jenner M. R.. Weinstein R.. Kaplan S. L. and Grumbach M. M.: Estradiol and testosterone secretion by human, simian and canine Testes, in males with hy~gonadism and in male p~udohermaphrodites with the feminizing testes syndrome. J. clin. Invest. 51 (1972) 824-830. 21. Payne H. H.. Kelch R. P., Musich S. S. and Halpern M. E.: Intratesticular site of aromatization in the human. J. clin. Endocr. Metub. 42 (1976) 1081-1087. 22. Sholiton L. J., Grossman C. J. and Taylor B.: Rat thymic homogenates convert testosterone to androgenic metabolites. J. steroid Biochem. 13 (1980) 1365-1367. 23. Rich R. R. and Pierce C. W.: Biological expressions of lympho~te activation. III Suppression of plaqueforming cell responses in virro by supernatant Ruids Jrom concanavalin A-activated spleen cell cultures. J. fmmunol. 112 (1974) 1360-1368. 24. Hallgren H. M. and Yumis E. J.: Suppressor lymphocytes in young and aged humans. J. ~mmu~o~. 11% (1977) 2004-2008. 25. Shaikh A. A. and Shaikh S. A.: Adrenal and ovarian steroid secretion in the rat esters cycle temporally related to gonadotropins and steroid levels found in peripheral plasma. ~~do~rjnology 96 (1975) 37-44. 26. Crabtree G. R.. Gillis S.. Smith K. A. and Munck A.: Glucocorticoids and immune responses. Arthris. Rheum. 22 (1979) 12461256. 27. Gregoire C. and Duchateau G.: A study of lymphoethelial symbosis in thymus. Reactions of the lymphatic tissue to extracts and to implants oJ epithelial components of thymus. Archs Biol. 67 (1956) 269-296. 28. Ahlqvist J.: Endocrine influences on the lymphatic organs. immune responses, inflammation and autoimmunity. Acra. endocrin, ~o~enh 83 (1976) Suppl. 206. 29. Kuriharra N., Takokoro S. and Ogawa H.: Studies on hormonal actions of dihydroxyprogesterone acetophenide, estradiol and their mixtures. Jpn. J. Pharmacol. .I
.
22 (1972) 43-58.
C. R., Perkins F. T. and Bishop D.: The effect 30. Franks ^ . O! sex hormones on the growth of HELA tumor nodules in male and female mice. Br. J. Cancer 31 (1975) 100-l 10. 31. Finn R., St. Hill C. A., Govan A. J., RolJs 1. G., Gurney F. J. and Denye V.: Immunological responses in pregnancy and survival of fetal homograft: Br. Med.-J. 3 (1972) 150-152. 32. Purtilo D. T., Hal&en H. M. and Tunis D. J.: Depressed maternal lymphocyte response to phytohaemo-
690
33. 34.
35.
36.
37.
38.
CHARLES J. GRCSSMAN
glutinin in human pregnancy. Lancer I (1972) 762771. Febris N.: Immunological reactivity during pregnancy in the mouse. Experienta 29 (1973) 610-612. Carter J.: The effect of progesterone, oestradiol and HCG on cell-mediated immunitv in oreunant mice. J. Reprod. Fert. 46 (1976) 21 l-216: - Kasakura S.: Is cortisoi responsible for inhibition of MLC reactions on pregnancy plasma. Nature 246 (1973) 496-497. Grossman C. J., Sholiton L. J, Nathan P. and Roselle G. A.: Estradiol depression of thymic lymphocyte function: mediation by a serum factor. Biol. Reprod. 20 (abstract) (1979).13A. Kruisbeek A. M.: Thymic factors and T-ceil maturation in vitro: a comparison of the effects of thymic epithelial cultures with thymic extracts and thymus dependent serum _ factors. Thymes 1 (1979) 163-185. White A. and Goldstein A. L.: The endocrine role of
et al.
the thymus and its hormone, thymosine. in the regulation of the growth and maturation of host immunological competence. Adr. hferab. Disord. 8 (1975) 359-374. _ 39. Comsa J.: Hormonal interactions of the thymus. Thymic Hormones. (Edited bv T. D. Luckev) Universitv Park Press, BalGmore (1953) pp. 59-89. . 40. Michael S. D.: The role of the endocrine thymus in female reproduction. Arthritis Rheum. 22 (1979) 1241-1245. 41. DeVries M. J. and Hijmans W.: A deficient deveiopment of the thymic epithelium and autoimmune disease in NZB mice. J. Path. Bact. 91 (1966) 487494.
42. Raveche E. S.. Tjio J. H., Boegel W. and Steinberg A. D.: Studies of the effects of sex hormones on autosomal and X-linked genetic control of induced and spontaneous antibody production. Arrhriris Rheum. 22 (1979) 1177-1187.