Immunomodulatory Effect of Methimazole on Inbred Mice

Immunobiol., vol. 180, pp. 23-32 (1989) Department of Anatomy and 2Department of Physiology, Faculty of Medicine, The Chinese University of Hong Kong, Shatin, NT, and 3 Department of Biology, Hong Kong Baptist College, Kowloon, Hong Kong 1

Immunomodulatory Effect of Methimazole on Inbred Mice WING-KEUNG Lrul, NAI-KI MAK3, and CHUEN-CHUEN WONd Received April 10, 1989 . Accepted in Revised Form August 15, 1989

Abstract The effects of antithyroid drug, methimazole (MMI, I-methyl-2-mercaptoimidazole), on the immune system of inbred mice, C57BL!6, were studied. The proliferative response of splenic lymphocytes to concanavalin A (Con A) and allogeneic stimulator cells was reduced when the mice were provided with 0.1 % MMI in tap water ad libitum for two to four weeks. The reduction of proliferative response was correlated with a lower frequency of proliferative spleen cells in the MMI-treated mice. The ability to produce macrophage activating factor of these spleen cells and the levels of hemolytic plaque-forming cells were also reduced. However, the mitogenic response of the splenic lymphocytes to lipopolysaccharide (LPS) was enhanced. In order to investigate whether the impaired cell-mediated and humoral immune systems would increase the susceptibility of the MMI -treated animal to the growth of tumor cells, mice were challenged with a lethal dose of Moloney virus-induced T cell lymphoma of C57BL!6 mice (MBL-2). The time required for 50 % of the animals to die was reduced from 15 days for the normal mice to 10 days for the MMI-treated mice.

Introduction Antithyroid drugs, such as 6-propyl-2-thiouracil (PTU) and methimazole, have long been used to inhibit the abnormal biosynthesis of thyroid hormones in patients with Graves' Disease (1-3). However, these drugs have been reported to exert regulatory effects on both the cellmediated and humoral immune responses of mice (4) and human (5). PTU suppresses in vitro mitogenic response of lymphocytes to phytohemagglutinin (PHA) and Con A (1), but MMI enhances the proliferative response of human peripheral blood lymphocytes (PBL) to Con A, PHA or pokeweed mitogen (PWM) when normal PBL were cultured at a high concentration of MMI in vitro (1, 6). Both drugs diminish total immunoglobulin synthesis in vitro (3, 5). Because most studies on immunoregulatory effect of MMI have been conducted using in vitro assay system, the present study was therefore designed to investigate the in vivo action of this antithyroid drug on the immune system of inbred mice provided with MMI for 2 to 4 weeks. Mitogenic response, frequency of the precursors of proliferative T cells (PPTC), mixed lymphocyte culture and secretion of macrophage activating factors (MAF) were used to measure the cell-

24 . WING-KEUNG Lru, NAI-KI MAK, and CHUEN-CHUEN WONG

mediated immune functions, while proliferative response of B cells to LPS and the formation of antibody secretion cells were used to estimate the humoral immune response. In addition, the development of a transplantable murine tumor in MMI-treated syngeneic mice was also studied.

Materials and Methods Mice Female inbred mice, C57BLl6 and BALB/c, were bred and housed in normal laboratory conditions (21 ° ± 2°C, 12/12-light/dark cycle). They were allowed free access to water and standard rodent chow for experimental use, at which time they were 8-10 wk old.

MMI treatment Two hundred mice (C57BLl6) were provided with 0.1 % MMI (Sigma M8502, St. Louis, MO, U.S.A.) in tap water ad libitum for 2 and 4 wk (MMI-2wk and MMI-4wk, respectively) and served as test animals. Control mice only received tap water.

Preparation of Lymphocyte Mice (C57BLl6) were sacrificed by cervical dislocation, and the spleens were asceptically excised. Splenocytes were obtained by gently pressing the tissue through a 1DO-mesh stainless steel sieve. The cells were washed, counted and resuspended at 1 X 107 cells/ml RPMI 1640 (GIBCO, NY, U.S.A.) supplemented with 10 % fetal calf serum (FCS), streptomycin sulfate (100 ""g/ml) and penicillin G (100 IU/ml).

Preparation of lymphokines Twenty ml of normal spleen cells (1 x 107 cells/ml) from C57BLl6 mice were cultured in a 75 cm2 culture flask (Sterilin, England) in the presence of 10 ""g/ml Con A (C2631, Sigma, St. Louis, MO, U.S.A.) for 3 h. The culture medium was removed, and the lymphocyte monolayer was rinsed carefully with warm phosphate-buffered saline (PBS) to remove excess ConA. The cells were further incubated for an additional 24 h at 37°C with 20 ml of fresh supplemented RPM!. After incubation, the cell free culture medium was sterilised by passing through a 0.22 !lID millipore filter and the lymphokine-containing medium was collected for MAF assay.

Mitogen induced lymphocyte proliferation Proliferative response was determined by incubating aliquots (1 X 106 cells/D. 1 mllwell) of splenocytes in triplicate with an equal volume of appropriated dilution of ConA in 96-well flat-bottom microtest plates (Sterilin, England) at 37°C in a humidified atmosphere of 5 % CO2 • The cultures were incubated for 72 h; during the last 6 h, the cells were pulsed with 0.5 ""Ci/10 ""l/well of 3H-methyl-thymidine CH- TdR, specific activity 5 Ci/mmole, Amersham, England). Cells were harvested onto a glass fiber filter with a cell harvester (Titertek, Flow Laboratories, England), and the radioactivity was measured using a Beckman scintillation counter. The proliferative response was expressed as mean counts per minute (cpm). Procedures for measuring the proliferative response of splenocytes to LPS (L4391, Sigma, U.S.A.) were similar to those for Con A as described above, but the cells were pulsed with 3H-T dR for 12 h.

Mixed lymphocyte culture Splenocytes (1 x 108 cells/ml) of normal BALB/c mice were treated with 25 ""g/ml mitomycin C (M0503, Sigma, U.S.A.) for 30 min and were used as a source of stimulator cells.

Methimazole, An Immunomodulator . 25 Responder splenocytes (5 X 105 cells/O.1 mllwell) from normal or MMI-treated C57BLl6 mice were co-cultured with stimulator cells at the ratio of 1 :1, 1 :0.5 and 1 :0.25 at 3rC for 84 h before they were pulsed with 3H-TdR (0.5 !JoCi/well) for 12 h. The amount of radioactivity up taken by the lymphocytes of MMI-treated mice was counted and compared with the controls.

Limiting dilution analysis of the precursors of proliferative T cell frequency The frequency of PPTC was assayed according to the method of RYSER and MACDoNALD (7). Briefly, a series of concentration of splenocytes (2.5 X 104 to 1.25 X 10' cells/culture) from normal and MMI-treated mice were co-cultured in 24 replicates with 2 X 10 6 feeder cells in the presence of 10 !Jog/ml Con A in a final volume of 0.2 ml at 37°C for 114 h .. Feeder cells were syngeneic spleen cells priorly treated with 25 !Jog/ml mitomycin C for 30 min. During the last 6 h of incubation, the cultures were pulsed with 0.5 !JoCi 3H-TdR, and those having 3H-TdR uptake greater than three standard deviations above the control values (i.e. cultures with Con A and feeder cells) were scored as positive cell proliferation. The frequency of PPTC was estimated according to the method described by TASWELL (8).

Macrophage activating factor (MAP) assay The tumoricidal activity of MAF activated macrophages was assayed with a modified method of KOGA et al. (9). Elicited peritoneal macrophages (PEC) were produced by i.p. injection of 3 ml of 3 % (w/v) thioglycolate (T0632, Sigma, U.S.A.) in RPMI and after 5 days, the peritoneal cells were collected by peritoneal lavage twice with 5 ml of cold Hanks' balanced salt solution (HBSS). The cells were rinsed, counted and resuspended at 5 X 106 cells/ml supplemented RPMI. Aliquots (0.2 ml) of cell suspensions were allowed to settle onto 96-well flat-bottom culture plates at 37°C for 3 h. The cultures were washed vigorously with warm HBSS to remove non-adherent cells, and the adherent macrophages were activated by incubation with a series of dilution of lymphokines (1 :1,1:2 and 1 :4) for 24 h. The tumoricidal activity of the activated macrophages was assayed by a further incubation of the macrophages with P815 mastocytoma cells for 42 h. The cultures were pulsed for 6 h with 0.5 !JoCi 3H-TdR, and the uptake of 3H-TdR by P815 cells was determined.

Plaque-forming cell (PPC) assay Normal and MMI-4wk mice (C57BL) were primed i.v. with an aliquot (0.2 ml) of sheep red blood cells (1 X 107 SRBC, Serotec, U.K.) for 4 days, and the splenocytes were collected for PFC assay according to the method described by CUNNINGHAM and SZENBERG (10). Guinea pig serum preadsorbed with the appropriated erythrocytes was used as a source of complement. Splenocytes (1 X 105) were incubated in 4 replicates with 2 % SRBC and 10 % complement in a final volume of 90 !Jol at 37°C in a Cunningham Chamber for 1 h, and the number of hemolytic plaques formed by antibody secreting cells was counted· under a dissection mICroscope.

Effect of MMI treatment on the survival of tumor-bearing mice Syngeneic MBL-2 tumor cells were maintained in an ascites form by weekly passage in C57BL mice. Test mice, 10 mice per group, were injected i.p. with viable 1.5 X 107 MBL-2 cells and then maintained in normal laboratory conditions for the observation of their survival rate.

Results

Proliferative response to Can A To determine whether short-term MMI treatment in vivo interferes with T lymphocytes proliferation, splenocytes from MMI-2wk and MMI-4wk

26 . WING-KEUNG Lru, NAI-KI MAK, and CHUEN-CHUEN WONG

.-.2

wk wk 4wk

0-00

40000

i

~-~

'"

I

~ ~

"~"

30000

,~!~i

::> ~

~

1....

20000

/:~j

.<1

t'"l

I

II:

10000

a

/T/

0,.......'_______ '

.ti(;

a

10

20

30

40

Con A (jJ.g/ml)

Figure 1. Mitogenic response of lymphocytes from MMI-treated C57BLl6 mice to ConA.

mice were co-cultured with Con A, and Figure 1 shows that the mitogenic responses of these lymphocytes were significantly reduced at low Con A concentrations (5 to 10 ftg/ml), but they were reconstituted to control level at higher Con A concentrations (> 10 ftg/ml).

Mixed lymphocyte culture Mixed lymphocyte culture is a measure of functional T lymphocytes, and the primary allogeneic response of spleen cells from MMI-treated mice was investigated. The ability of test lymphocytes to proliferate in response to 15000

a '"

0 - 0 0 wk . - . 2 wk 6 . - 6 . 4wk

~

~

p.,

10000

::>

I [zl

...

l"l

I

5000

:xl

STIMULATOR/RESPONDER

Figure 2. Mixed lymphocyte culture of MMI-treated C57BLl6 mice.

Methimazole, An Immunomodulator . 27 Table 1. Effect of MMI treatment on the production of MAF P815

+

PEC

Source of lymphokines

+ +

+ + +

+

+

MMlb (2 wk)

+

+

MMlb (4 wk)

NSC"

NSC: normal spleen cells b Spleen cells from mice treated with MMI for 2 n=3

Dilution of lymphokines

3H-TdR uptake (cpm) (Mean ± S.D.)

1 :1 1:2 1:4 1 :1 1:2 1:4 1 :1 1:2 1:4

141,364 ± 6,407 58 1,053 ± 120,120 ± 5,601 39,967± 3,785 49,848 ± 3,521 70,316 ± 10,036 48,597 ± 3,907 39,269 ± 1,574 47,107 ± 3,438 115,102 ± 5,850 126,886 ± 5,725 121,398 ± 33,846

a

to

4 weeks

alloantigen stimulation was lower than that of control groups at all different ratios of stimulator to responder (Fig. 2).

MAP assay In this study, an in vitro assay for the inhibition of tumor cell growth by activated macrophages was used to estimate the ability of spleen cells from MMI-treated mice to produce MAF. Results in Table 1 show that normal PEC did not significantly inhibit the growth of P8iS mastocytoma cells, unless they were pre-incubated with lymphokines prepared from normal and MMI-2wk spleen cells. It is believed that normal spleen cells stimulated by Con A produced regulatory molecules which can activate macrophages (11-13). However, the tumoricidal activity of PEC treated with lymphokines prepared from MMI-4wk spleen cells was greatly reduced, indicating that the macrophage-activating ability of these lymphokines is lower than the controls.

Table 2. Frequency of proliferative T cells in MMI-treated mice Exp. No.

MMI Treatment (wk)

Frequency of proliferative T cells (X 10-4)

o

1.11 0.83

4

DAD

28 71

4

0.50 0.23

54

2 2

% of Reduction

o

28 . WING-KEUNG Lru, NAI-KI MAK, andCHUEN-CHUEN WONG

Cells (xl 000)/ culture

o

50

25

75

100

125

100+T--------~--------~--------~------~~----~~

rn

Q)

'"

::I

37

+'

::I t)

Q)

....+'I>

l\I till

10

Q)

I:l ~

till 0

...:l

o Owk l> 2 wk o 4wk

Figure 3. Comparison of frequencies of proliferative T cells in the spleens of normal (0), and mice treated with MMI for 2 weeks (,0,) or 4 weeks (0). Dotted line indicates 37 % negative cultures.

Limiting dilution analysis of proliferative T-cell precursor frequency The frequency analysis was used to investigate whether the reduced response of test splenocytes to Con A and the specific alloantigen are due to a decrease of PPTC in MMI-treated mice. Results in Table 2 and Figure 3 show that the frequency of PPTC in MMI-2wk and MMI-4wk mice was only 28 % and 54-74 % of the control.

Mitogenic response to LPS The proliferative response of B cells to LPS, a polyclonal B cell activating substance, is shown in Figure 4. B cells of MMI-4wk mice increased significantly in the presence of high levels of LPS (2 to 4 fA,g/ml) in comparison with controls. No significant difference was observed between the MMI-2wk group and the controls. Table 3. Production of PFC (mean MMI treatment (wks)

4

n=4

± S.

D.) in MMI-treated mice

Priming with SRBC

+

+

<1

67.0 ± 8.5 43.3 ± 1.6

Methimazole, An Immunomodulator . 29 20000

s

o-oOwk .-.2wk t . - t . 4wk

p.

~ 15000 Q)

~ .... p. p

Q)

J:I

10000

;a

.[ E5

t'J

I

5000

::r: O+-----.-----r-----r----~

o

2

3

4

LPS (J.'g/ml)

Figure 4. Mitogenic response of lymphocytes from MMI-treated C57BLl6 mice to LPS.

Production of direct IgM plague-forming cells (PFC) assay Normal and MMI-4wk C57BLl6 mice were primed with SRBC to test the drug's effect on the capacity of these mice to produce antibody against SRBC. Table 3 shows that a high level of PFC was induced in normal mice after priming with SRBC for 4 days, but 35 % reduction of hemolytic plagues was observed in MMI-treated mice.

Effect of MMI treatment on the survival of tumor bearing mice Normal and MMI-treated C57BLl6 mice were challenged with a lethal dose of syngeneic MBL-2 tumor cells (1.5 X 107 cells/mouse) to observe the 100T-------~~~ I

90

b-o

80 70

I I I I

:

b-El ~ •••• I

060

I

>

I

·~50

.,

"'-0I

::l

_40

o

~30

20 10

I

I

•

~--, I

r'~

I: b-;"

··~:-TI : I

O~--~----~--~~~~---

5 10 15 20 25 Time after tumor cell injection (days)

Figure 5. Survival rate of mice treated with MMI for 2 weeks (_) or 4 weeks (0) and normal mice (e) after injection of tumor cells (MBL2).

30 . WING-KEUNG LID, NAI-KI MAK,

and

CHUEN-CHUEN WONG

survival rate of these mice. Figure 5 shows the time required for MBL-2 cells to cause lethality in 50 % of normal and MMI-2wk mice were 15 and 14 days in comparison to 10 days of MMI-4wk mice.

Discussion Although methimazole, an active metabolite of the carbimazole antithyroid agent, has been widely used in the treatment of hyperthyroidism (2), it has an immunomodulatory effect on the patients. It enhanced the mitogenic response of PBL to ConA, PHA and PWM (1, 6), and augmented the activity of natural killer cells to malignant B cell in vitro (6). On the contrary, MMI in vivo significantly inhibited the mitogenic response of splenic lymphocytes to Con A and specific allogeneic cells (Figs. 1 and 2). The results in this study further demonstrate that the frequency of proliferative precursor T cells from spleen of MMI-4wk mice also decreased by 2-fold in the limiting dilution assay. A similar observation was reported on PTU which altered the ratio of T cell subpopulations in hypothyroid rats (14). Whether the lower frequency of splenic PPTC is due to a suppression of cell production, an increase in cell death or a selective migration of T cells to other lymphoid organs in MMI-treated mice is unanswered in this study; the inhibitory effect of these antithyroid drugs on the cell-mediated immune functions is, however, specific to T cells, as indicated by the augmentation of B cells to a polyclonal B cell activator (Fig. 4). Surprisingly, MMI enhanced mitogenic response of lymphocytes to LPS but decreased the generation of direct (IgM) PFC to SRBC in vivo (Table3). A similar phenomenon was observed in human lymphocytes incubated with MMI or PTU in vitro at a non-cytotoxic concentration (10- 4 M) (5). SRBC is a T cell-dependent antigen, and the generation of primary humoral immune response to SRBC requires the presence of T helper cells (15). A lower frequency of PPTC (Table 2) might account for a reduced production of direct (IgM) PFC in MMI-treated mice. On the other hand, LPS is a well-documented polyclonal B cell activator, and its mitogenic stimulation is independent of T cells (12). Therefore, a decrease in the total T cell population in MMI -treated mice might not interfere with the mitogenic response of B cells· to LPg. The precise mechanism by which MMI stimulates the humoral immune response is currently under further investigation. The role of macrophages in the induction of immune response has been studied extensively (16). Monocyte-macrophages can be activated to become cytotoxic effector cells against a variety of microorganisms and neoplastic cells. The production and secretion of reactive oxygen metabolites are usually associated with the cytotoxic activity of macrophages (17). WEETMAN et al. and other authors (3, 4, 18-20) found that the presence of peroxidase in the monocyte-macrophages is an important factor

Methimazole, An Immunomodulator . 31

for the intracellular accumulation of MMI which inhibits production of oxygen free radicals and subsequently alters cell functions. In addition, the activity of lysosomal enzymes in macrophages of MMI -treated rats is reduced by 50 %, which can reduce their cytotoxicity (unpublished data). In this study, we further demonstrated that lymphokines from MMI-4wk mice contained a lower level of MAF, and thus could not activate normal macrophages to inhibit tumor cell growth. All these factors may explain the reduced tumoricidal activity of macrophages from MMI -treated mice. Our animal survival test with transplantable MBL-2 tumor cells also agrees with the fact that both specific and non-specific immune responses are important in the host defence against infectious and neoplastic cells. Interestingly, despite a decrease of mitogenic and allogeneic responses, our data show that treatment of mice with MMI for 2 weeks had no significant effect on mitogenic response of B cells to LPS, nor had it affected the killing ability of transplantable tumor cells. Only when the mice received an antithyroid drug treatment for 4 weeks were marked immunomodulatory effects seen. Whether the action of this drug is caused by a decrease of cellular thyroid hormone, and the subsequent influence of hormonenuclear receptor complexes to the corresponding genes for different immune responses (21, 22) or is a combination of both direct and indirect influences of this antithyroid drug on the test animals remains unclear. Mice and rats treated with MMI for more than four weeks have lower body weight, 50 % of lytic enzymes and reduced tumoricidal activity in macrophages (unpublished data) and an inhibitory effect on lymphocytes, indicating that this antithyroid drug has a wide range of action on the test animals. Further investigation is therefore required to obtain an insight of the action of the antithyroid drug and the regulatory effects of thyroid hormone on the macrophages and lymphocytes. Acknowledgements The authors thank Ms. MBL-2 tumor cells.

J. TAM for technical assistance and Dr. K. N.

LEUNG for providing

References 1. HALLENGREN, B., A. FORSGRAN, and A. MELANDER. 1980. Effects of antithyroid drugs on lymphocyte function in vitro. J. Clin. Endocrinol. Metab. 51: 298. 2. RATANACHAIYAVONG, S., and A. M. MCGREGORY. 1987. Pharmacology and clinical uses of antithyroidal agents: In «Pharmacology and Clinical Uses of Inhibitors of Hormone Secretion and Action». FURR, B. J. A., and A. E. WAKELING (Eds.). Bailliere Tindall. pp. 554-584. 3. WEETMAN, A. P., A. M. MCGREGORY, and R. HALL. 1983. Methimazole inhibits thyroid autoantibody production by the action on accessory cells. Clin. Immunol. Immunopathol. 28: 39. 4. WEETMAN, A. P., M. E. HOLT, A. A. CAMPBELL, R. HALL, and A. M. MCGREGORY. 1984. Methimazole and generation of oxygen radicals by monocytes: potential role in immunosuppression. Br. Med. J. 288: 518.

32 . WING-KEUNG LIU, NAI-KI MAK, and CHUEN-CHUEN WONG 5. WEISS, 1., and T. F. DAVIES. 1981. Inhibition of immunoglobulin-secreting cells by antithyroid drugs. J. Clin. Endocrinol. Metab. 53: 1223. 6. SHARMA, B. S., and A. N. ELIAS. 1987. Effects of methimazole on human lymphocyte proliferation and natural killer cell activity. Gen. Pharmacol. 18: 449. 7. RYSER, J. E., and H. R. MACDoNALD. 1979. Limiting dilution analysis of alloantigenreactive T lymphocytes. 1. Comparison of precursors frequencies for proliferative and cytolytic responses. J. Immunol. 122: 1691. 8. TASWELL, C. 1981. Limiting dilution assays for the determination of immunocompetent cell frequencies. J. Immunol. 126: 1614. 9. KOGA, T., M. MITSUYAMA, T. HANDA, Y. WATANABE, and K. NOMOTO. 1987. Gamma interferon mediated increase in the number of la-bearing macrophages during infection with Listeria monocytogenes. Infect. Immunity 55: 2300. 10. CUNNINGHAM, A. J., and A. SZENBERG. 1968. Further improvements in the plague technique for detecting single antibody-forming cells. Immunology 14: 599. 11. COFFMAN, R. 1., and J. CARTY. 1986. A T cell activity that enhances polyc1onal IgE production and its inhibition by interferon. J. Immunol. 136: 949. 12. KOHNO, K., T. KAKIUCHI, M. TAKEUCHI, S. TAIRA, and H. NARIUCHI. 1987. Accessory cell function in a Con A response: role of la and interleukin 1. Cell. Immunol. 106: 250. 13. MURRAY, H. W. 1988. Interferon-gamma, the activated macrophage, and host defense against microbial challenge. Ann. Intern. Med. 108: 595. 14. PACINI, F., H. NAKAMURA, and 1. J. DEGROOT. 1983. Effect of hypo- and hyperthyroidism on the balance between helper and suppressor T cells in rats. Acta Endocrinol. 103: 528. 15. MOTA, 1. 1981. Activity of immune cells. In: Fundamentals of Immunology. BIER, O. G., W. D. DA SILVA, D. GOTZE, and 1. MOTA (Eds.). Springer-Verlag New York Inc. pp. 33-57. 16. PERSSON, U., H. E. MOLLER, G. MOLLER, and C. 1. E. SMITH. 1978. The role of adherent cells in Band T lymphocyte activation. Immunol. Rev. 40: 78. 17. WILSON, C. B., V. TSAI, and J. S. REMINGTON. 1980. Failure to trigger the oxidative metabolic burst by normal macrophages: possible mechanism for survival of intracellular pathogens. J. Exp. Med. 151: 328. 18. BALAZA, C., E. KIss, A. Leovey, and N. R. FARID. 1986. The immunosuppressive effect of methimazole on cell-mediated immunity is mediated by its capacity to inhibit peroxidase and to scavenge free oxygen radicals. Clin. Endocrinol. 25: 7. 19. TAYLOR, J. J., R. 1. WILSON, and P. K. TAYLOR. 1984. Evidence for direct interactions between methimazole and free radicals. Fed. Eur. Biochem. Soc. 176: 337. 20. WEETMAN, A. P., C. GUNN, R. HALL, and A. M. MCGREGORY. 1984. The accumulation of 35S-methimazole by monocytes and macrophages. Acta Endocrinol. 107:: 366. 21. SAMUELS, H. H., B. M. FORMAN, Z. D. HOROWITZ, and Z. S. YEo 1988. Regulation of gene expression by thyroid hormone. J. Clin. Invest. 81: 957. 22. OPPENHEIMER, J. H., H. 1. SCHWARTZ, C. N. MARIASCH, W. B. KINLAW, N. C. W. WONG, and H. C. FREAKE. 1987. Advances in our understanding of thyroid hormone action at the cellular level. Endocrin. Rev. 8: 288. Dr. W. K. LIU, Dept. of Anatomy, Faculty of Medicine, The Chinese University of Hong Kong, Shatin, NT, Hong Kong