Unique properties of auranofin as a potential anti-rheumatic drug

Agents and Actions, vol. 19, 1/2 (1986)

0065-4299/86/020109-07 $ 2.90 9 1986 Birkh~tuser Vedag, Basel

Unique properties of auranofin as a potential anti-rheumatic drug M.L. Barrett and G. P. Lewis Department of Pharmacology, Royal College of Surgeons, Lincoln's Inn Fields, London WC2A 3PN

Abstract

Gold salts, auranofin (AF), aurothiomalate (ATM) and aurothioglucose (ATG) displayed immunosuppressive action in a series of in vitro assays which mimic the cell-cell interactions thought to occur in rheumatoid arthritis. The gold salts inhibited phytohaemagglutinin (PHA)-induced thymidine incorporation and y-IF production by peripheral blood mononuclear cells, as well as IL-2-induced proliferation of PHA-blasts. The separate addition of IL-2 and y-IF partly reversed the anti-proliferative effects of ATM and ATG; however, the addition of IL-1 had no effect. ATM and ATG inhibited PHAstimulated IL-1 production by mononuclear cells but not spontaneous or LPS-induced IL-1 production by adherent monocytes. It was concluded that ATM and ATG inhibited lymphocyte function and lymphocyte-amplification ofmacrophage function. The anti-proliferative effects of AF were partly reversed by IL-2 but not by y-IF or IL-1. AF inhibited PHA-stimulated IL-1 production by mononuclear cells as well as spontaneous and LPS-induced production by adherent cells. It appeared that AF inhibited lymphocyte and macrophage function directly. AF also displayed potential anti-inflammatory activity in that it inhibited PGE2 and collagenase production by proteolytically dispersed rheumatoid synovial cells. Introduction

Rheumatoid arthritis is a disease of chronic immunologically-mediated inflammation. It may result from a specific or polyclonal response to an unknown antigen. The disease is characterised by excessive antibody production which may be the result of B-lymphocyte hyperactivity driven by unrestrained T-lymphocyte helper/induced cells. Present in the joint synovial tissue are macrophages and macrophage-like synovial cells which express I a determinants of the Major Histocompatability Complex. Antigen presented in association with surface I a markers can serve as a stimulant in the mixed lymphocyte reaction. The importance of lymphocytes in the pathogenesis of

rheumatoid arthritis is supported by profound clinical improvement seen in rheumatoid patients following depletion of recirculating lymphocytes either by lymphophoresis [1] or thoracic duct drainage [2]. The development of adjuvant arthritis in rats has also been delayed by depletion of T-cells using anti-pan T-cell monoclonal antibodies (W3/13) [3]. Implication of the macrophage or antigen-presenting cell has come from a study in which antibodies against Ia (HLA-DR) were found to alleviate the symptoms of rheumatoid arthritis [4]. The expression of I a determinants on macrophages can be stimulated by lymphocytes through the production of y-interferon (y-IF) [5]. I a expression is thought to be implicated in macrophage production of in-

110 terleukin (IL)-I [6]. IL-1 is an amplifying factor in IL-2 receptor expression and IL-2 production in antigen and mitogen driven lymphocyte proliferation [7]. IL-2 functions as the growth factor supporting T-cell proliferation. If the pathogenesis of rheumatoid arthritis stems from macrophagelymphocyte interactions, then the above cytokines may play a modulating role in the disease. It follows that pharmacological agents which regulate the release or action of these cytokines may modulate the course of rheumatoid arthritis. Gold compounds are one of the few classes of drugs which retard the progress of the disease [8]. Aurothiomalate (ATM) and aurothioglucose (ATG), both injectable gold salts, are considered therapeutically equivalent. Auranofin (AF), an orally active gold preparation, appears to be slightly less potent, although clinical experience with this new agent has so far been limited [9]. It is, however, reported to be less toxic. We have studied the effects of these gold compounds in a series of in vitro assays which mimic the cell-cell interactions thought to occur in the rheumatoid synovium. All three gold preparations modified lymphocyte amplification of macrophage function while AF, in addition, affected macrophage function directly. AF also inhibited the generation of inflammatory mediators from rheumatoid synovial cells in culture. Materials and Methods

Auranofin (Smith Kline and French Labs Ltd) was dissolved in absolute alcohol at 20 mg/ml and further diluted in Dulbecco's Modified Eagles Medium (DMEM) to the required concentrations. Aurothiomalate (10% Myocrisin, May and Baker Ltd) was diluted in DMEM as required. Aurothioglucose (Sigma) was dissolved at 8 mg/ml in DMEM and sterilized by membranefiltration (0.22 lira Millex) prior to dilution in DMEM. The cytokine preparations used were purchased from the sources indicated: interleukin-1 (Genzyme, UK, 100u/ml), )'-interferon (Ventrex, USA, 100 u/ml), T-cell growth factor (Associated Biomedics Systems Inc, USA). The isolation and culture of peripheral blood mononuclear cells were essentially as described in our previous paper in these Proceedings. Drugs and cytokine preparations were added at the beginning of the culture period. ),-IF in the culture

Agents and Actions, vol. 19, 1/2 (1986)

supernatant was measured directly after 72 h using a Centocor RIA kit. Adherent monocytes were prepared by incubating for 2 hours at 37 ~ and non-adherent cells removed by repeated washing. Drugs were then added and incubation continued for a further 48 hours. Adherent synovial cell cultures (RASC) were established essentially as described by other workers [10], using synovectomy material obtained from rheumatoid patients undergoing corrective surgery. Adherent synovial cells (3 x 10s/ ml), maintained in DMEM supplemented with 10% heat-inactivated foetal calf serum (HIFCS) release large amounts of prostaglandin E2 and collagenase during primary culture. Upon subculturing, the production of mediators declines to low or undetectable levels. These quiescent cells were used for assay of IL-1 action through stimulation of PGE2 release. After 48 h incubation periods at 37~ in 5% CO2, the cell-free supernatants were removed and the PGE2 content measured as previously described [11]. Mononuclear cell supernatants were assayed on synovial cells at 1 in 10 and 1 in 25 dilutions. At these dilutions, the effects of drugs and PGE2 production by mononuclear cells were insignificant. In collaboration with Dr W. Harvey (Eastman Dental Hospital, London), synovial cell supernatants were assayed for collagenase by digestion of 3H-acetylated rat skin collagen fibrils, using trypsin to activate latent enzyme. Results

All three gold salts, AF, ATM and ATG, inhibited both sub-optimal (0.1 ~tg/ml) and optimal (l.0~tg/ml) PHA-stimulated SH-thymidine incorporation in peripheral blood mononuclear cells as seen in Figure 1. The gold salts were more effective against sub-optimal than optimal PHAinduced stimulation. AF (ICso 0.84 ~tg/ml) was the most potent inhibitor, being approximately 160 and 480 times more active on a weight basis than ATM (ICso 138 ~tg/ ml) and ATG (ICs0 405 ~tg/ml) respectively. Exogenous IL-2 (T-cell growth factor, 1 u/ml) partly reversed the inhibitory effects of the gold salts upon sub-optimal PHA-stimulated cells as illustrated in Figure 2a. Recombinant IL-2 (100 u/ ml) had a similar effect (results not shown). The effect was notable against concentrations of gold

111

Agents and Actions, vol. 19, 1/2 (1986) PHA 0. I UG/ML (---) PNA I.e UG/ML ( - - )

1~2

iI

/

~

//

zo

/

z

ii /111

I

~ s0i

O EO'

'

'

30,

/

~ 30

3

s o

~z 9o

//

z

#

tO UIML

/ /

g

0:05

41:2

0:8

3:2

IZ:5

GOLD COMPOUNDCONCENTRATION(UG/ML)

Figure 1 Effects of auranofin ( l ) , sodium aurothiomalate ( e ) and aurothioglucose (&) on sub-optimal (0.1 pg/ml) (broken lines) and optimal (1.01zg/ml) (solid lines) PHA-induced ~H-TdR incorporation by mononuclear cells. Each value is the mean of 12 replicate cultures from 3 separate experiments; vertical lines indicate SEM.

salts which caused less than 70% inhibition of the proliferative response. IL-2 did not reverse inhibition of mononuclear cells stimulated with an optimal P H A concentration. As illustrated in Figure 2b, the addition o f )'-IF (10 u/ml) partly reversed the anti-proliferative effects of ATM and ATG but had no effect upon the action of AF. Addition of IL-1 (5 u/ml) had no effect upon the inhibitory action of any of the gold compounds as depicted in Figure 2 c. Figure 3 illustrates the effects o f the gold compounds upon IL-2-induced proliferation of PHAblasts. Effective concentration ranges were similar to those that inhibited PHA-induced m o n o n u clear cell proliferation. ),-IF production was also inhibited in this concentration range as shown in Figure 4. Previously we reported studies on the effects of gold salts upon IL-1 production [11]. We found, as noted in Table 1, that although all gold compounds inhibited PHA-induced IL-1 production by mononuclear cells, only AF effectively inhibited 'spontaneous' generation by adherent mono-

/

/

/

T

"

0.05 a)

0.2

0.8

3.2

.+ ,/ 12.5

EO

200

8oo

GOLD COMPOUNDCONCENTRATION(UG/ML) 0"IF 1OU/ML

~ ~176 //!

~o TM ~ 00 i ~ +o i i 1o " o.o6

o12

ole

a:2

I:~.5

8~o

260

s6o

GOLD COMPOUNDCONCENTRATION(UG/ML)

b)

Oo I I! 7o

_~

/

50

Figure 2

Inhibition by auranofin (l), sodium aurothiomalate (e) and aurothioglucose (A) on sub-optimal PHA (0.1 Ixg/ml) stimulated 3H-TdR incorporation by mononuclear cells in the absence of (solid lines) and presence of (broken lines) (a) IL-2 (1 u/ml), (b)),-IF (10 u/ml) and (c) IL-1 (5 u/ml). Each value is the mean of 16 replicate cultures from 4 separate experiments; vertical lines indicate SEM.

/

1o

1t/

30

| lO

0.05 C)

0.2

0.8

3.2

12.5

50

200

GOLD COMPOUND CONCENTRATION(UG/ML)

EO0

112

Agents and Actions, vol. 19, 1/2 (1986)

/

9@

50

/

Table 1

//

9

0.8

3.2

I2.S

50

200

Assay

| IL-1 (MCF) production Spontaneous LPS 1 p_g/ml PHA 1 p_g/ml

I0 0.2

Effects of aurothiomalate (ATM), aurothioglucose (ATG) and auranofm (AF) upon IL-1 production and action. Each ICso value is the mean of 6 replicates from 2 separate experiments.

800

GOLD C O M P O U N D CONCENTRATION ( U G I M L )

Concentration of gold salt ICso (~tg/ml) ATM

ATG

AF

> 800 > 800 40

> 800 280

2.1 0.20 0.55

410

> 800

2.2

> 800 > 800

> 800 > 800

7.2 3.6

| IL-1 (MCF)

Figure 3 Effects of auranofin (ll), sodium aurothiomalate (~) and aurothioglucose (A) on IL-2-induced ~H-TdR incorporation by PHA-blasts. Each value is the mean of 9 replicate cultures from 3 separate experiments; vertical lines indicate SEM.

action on RASC (PGE2 production) | RASC 1o culture spontaneous generation PGFa Collagenase

O0

z 0

70

0

~ 3o

! 0.05

0.2

0[8

3.2

12.5

50

200

800

cytes. We now report that LPS-generated IL-1 production is also inhibited by AF but not by ATM. Primary culture synovial cells spontaneously release PGE2 and collagenase and both were inhibited by AF, as illustrated in Figure 5. ATM and ATG had no effect upon the release of either mediator. This result agrees with a previous finding depicted in Table 1 on the effects of the gold compounds upon IL-1 stimulation of quiescent synovial cells to produce PGE2.

GOLD COMPOUNO C O N C E N T R A T I O N 0JGIML)

Figure 4 Inhibition of sub-optimal PHA (0.1 ~tg/ml) stimulated y-IF production by auranofin (m), sodium aurothiomalate (O) and aurothioglucose (A). Each value is the mean of 6 replicate cultures from 2 separate experiments; vertical lines indicate SEM. PGE z (---) COLLAGENASE ( - - )

o 0.1115

0105 012 0.8 122 12~ST"---~SO AURANOFIN CONCENTRATION (UGIML)

Figure S Effects of auranofin upon PGE2 release (broken lines) and collagenase release (solid lines) from primary culture rheumatoid synovial cells. Each value is the mean of 3 replicates from 1 experiment; vertical lines indicate S.E.M.

Discussion

This study is an extension of previous work examining the effects of gold salts on lymphocytemacrophage interactions [11]. It provides additional evidence that, whereas AF interferes with both lymphocyte and macrophage functions in mitogen-driven mononuclear cell cultures, ATM and ATG appear to affect only the lymphocytemediated amplification of macrophage function. The ability &gold salts to inhibit PHA-stimulated mononuclear cell proliferation has been previously demonstrated by us and others. AF was much more active that ATM which was more active than ATG. However, activity was not accounted for in terms of gold, as AF consists of 30% gold by weight and both ATM and ATG consist of 50% gold by weight.

Agents and Actions, vol. 19, 1/2 (1986)

Cytokines IL-2, y-IF and IL-1 are established mediators of intercellular reactions between subpopulations of mononuclear cells. We were therefore interested in the effect of their addition upon the anti-proliferative effects of the gold compounds. IL-2 was able to partly restore the anti-proliferative effects of all three gold compounds in their lower effective concentration range. However, this effect was limited, due to a direct action by the drugs upon IL-2 responder cells. Inhibition of IL-2 proliferation of PHA-blasts was demonstrated in about the same concentration range as that which inhibited PHA-induced mononuclear cell proliferation. The ability of y-IF to restore low-dose anti-proliferative effects of ATM and ATG suggested a role for this lymphokine in the inhibitory effects of these drugs. Furthermore, ATM and ATG inhibited the release of y-IF in these concentration ranges. AF also inhibited the release of y-IF from mononuclear cell populations but the anti-proliferative effect of this drug was not affected by exogenous lymphokine, y-IF has been reported to be produced by T-cells of the helper/inducer subset and to possess macrophage-activating properties [12, 13]. Therefore, it appeared that, whereas all of the gold compounds blocked y-IF production, AF additionally affected the ability of the macrophage to respond to the lymphokine. As mentioned previously, macrophage activation by y-IF involves expression of I a determinants which are thought to proceed or be required for IL-1 release. IL-1 is capable of stimulating lymphocytes, modifying receptor expression in some cases and promoting the generation of T-cells through IL-2 release [14]. However, the addition of IL-1 failed to reverse the anti-proliferative effects of the gold salts, as would have been expected had inhibition of its release played a prominent role in the action of gold compounds. The results of studies on the effect of gold salts upon IL-1 production separated the conventional injectable compounds from the orally active drug. AF was active against 'spontaneous' and LPS-induced IL-1 production by adherent monocytes, as well as PHA-induced IL- 1 production by mononuclear cells. In contrast, ATM and ATG did not significantly affect spontaneous IL-1 generation or that produced by the macrophage stimulant LPS. They did, however, inhibit IL-1

113 generated indirectly through the lymphocyte mitogen PHA. It appeared from our studies that inhibition of y-IF release could play an important role in the action of ATM and ATG upon rheumatoid arthritis. Immune interferon has been found in the serum of patients with rheumatoid arthritis and there appears to be correlation between interferon titres and disease activity [15]. 7-IF could be implicated in persistent antigen presentation through stimulation of I a expression. As mentioned previously, therapy using preparations containing antibodies against I a (HLA-DR) alleviated the symptoms of this disease. Lipsky and Ziff provided good evidence that ATM intefered with macrophage accessory function but the precise mechanism was not elucidated [16]. However, it would appear from our study that this effect did not involve IL-1, but perhaps the cell-cell contact inherent in antigen-Ia presentation [ 17]. It has been reported that y-IF plays a role in stimulating IL-1 production possibly through Ia expression. Both IL-1 and y-IF have been reported to stimulate synovial fibroblast proliferation, perhaps causing synovial hyperplasia [14, 18]. The rheumatoid synovium is characterised by a high proportion of T-helper cells (T4 § in close association with those expressing Ia (DR) and an apparent lack of or reduction in T-suppressor (T8 § cells [19]. IL-1 has been reported to promote the generation of T-helper over T-suppressor cells [20]. y-IF itself has also been reported to directly enhance T-helper cell proliferation while T-suppressor cells remain unaffected [21]. In summary, ATM and ATG may alter the course of rheumatoid arthritis through suppression of y-IF release and T-cell proliferation. In support of this theory is clinical evidence which reports reduced lymphocyte counts in patients receiving standard gold therapy [22]. In contrast AF, which was consistently more active than ATM or ATG in every test system, also displayed additional properties. It directly inhibited the 'spontaneous' IL-1 production by adherant mononuclear cells which occurred during handling, as well as that stimulated by endotoxin. IL-1 is thought to be an important inflammatory mediator in synovial tissue as it stimulates PGE2 and collagenase release, both of which are involved in inflammatory processes [23]. Pros-

114

taglandins are believed to be important mediators of vasodilatation, oedema and hyperalgesia associated with inflammation, as well as being implicated in bone resorption through stimulation of osteoclasts [24]. Collagenase and other neutral proteases may cause degradation of extracellular matrix molecules. Freshly isolated proteolytically dispersed synovial cells are a heterogenous population which is thought to contain IL-1 secreting cells which trigger the release of PGE2 and collagenase [25]. AF was found to inhibit the spontaneous release of both PGE2 and collagenase from primary culture synovial cells, while ATM and ATG had no effect upon either. AF also inhibited IL-l-induced PGE2 production by quiescent synovial cells. The ability of AF to inhibit both PGE2 and collagenase release is shared by the glucocorticoids which are another group of drugs reported to modify the course of rheumatoid arthritis. Nonsteroidal anti-inflammatory drugs which provide only symptomic relief in this disease, inhibited PGE2 release but either increased or had no effect upon collagenase release. In conclusion, the gold salts tested displayed immunosuppressive activity in vitro which may represent their mode of action against rheumatoid arthritis. ATM and ATG inhibited lymphocyte function and lymphocyte-mediated amplification of macrophage function. AF inhibited both lymphocyte and macrophage function directly, and in addition inhibited production of mediators thought to be involved in rheumatoid joint destruction. We gratefully acknowledge the support of the Arthritis and Rheumatism Council.

References [1] J. Karsh, J. H. Klippel, P. H. Plotz, J. L. Decker, D. G. Wright and M. W. Flye, Lymphophoresis in rheumatoid arthritis, Ann. Rheum. Dis. 33, 361 (1981). [2] H. E. Paulus, H. I. Machleder, S. Levine, D. T. Y. Yu and N. S. MacDonald, Lymphocyte involvement in rheumatoid arthritis, Arth. Rheum. 20, 1249 (1977). [3] P. Larsson, R. Holmdahl, L. Dencker and L. Klareskog, In vivo treatment with W3/13 (anti-pan T) but not with 0)(8 (anti-suppressor/ cytotoxic T) monoclonol antibodies impedes the development of adjuvant arthritis in rats, Immunology 56, 383 (1985). [4] J. Sany, J. Clot, M. Ponneau and M. Andory, Antibodies against HLA-DR alleviate symptoms of rheumatoid arthritis, Arth. Rheum. 25, 17 (1982).

Agents and Actions, vol. 19, 1/2 (1986) [5] P. S. Steep, R. N. Moore, H. M. Johnson and J. J. Oppenheim, Regulation of murine macrophage Ia antigen expression by a lymphokine with immune interferon activity, J. Exp. Med. 156, 1780 (1982). [6] R. C. Newton, Effect of interferon on the induction of human monocyte secretion of interleukin-1 activity, Immunology 56, 441 (1985). [7] K. A. Smith, L. B. Lachman, J. J. Oppenheim and M. F. Favata, The functional relationship of the interleukins, J. Exp. Med. 151, 1551 (1980). [8] Empire Rheumatism Council Sub-Committee, Gold therapy in rheumatoid arthritis: report of a multi-centre controlled trial, Ann. Rheum. Dis. 19, 95 (1960). [9] H. A. Capell, D. S. Cole, K. K. Manghani and R. W. Morris (Eds.), A uranofin, proceedings of a Smith Kline and French International Symposium, Excerpta Medica, Amsterdam 1983. [10] J. M. Dayer, S. M. Krane, G. G. Russel and D. R. Robinson, Production of collagenase and prostaglandins by isolated adherent rheumatoid synovial cells, Proc. Natl. Acad. Sci. USA 73, 945 (1976). [ll] D. Gordon and G. P. Lewis, Effects of piroxicam on mononuclear cells, comparison with other anti-arthritic drugs, Inflammation, Supp. to 8, 87 (1984). [12] O. Martinez-Maza, U. Andersson, J. Andersson, S. Britton and M. DeLey, Spontaneous production of interferon-y in adult and newborn humans, J. Immunol. 132 (l), 251 (1984). [13] L. P. Severdeshy, C. V. Benton, W. H. Berger, E. Rinderknecht, 1L N. Harkins and M. A. Palladino, Biological and antigenic similarities of murine interferon-y and macrophage-activating factor, J. Exp. Me d., 159, 812 (1984). [14] C. A. Dinarello, Interleukin-l, Rev. Infect. Dis. 6, 51 (1984). [15] J. J. Hooks, H. M. Moutsopoulos, S. A. Geis, N. I. Stahl, J. L. Decker and A. L. Nokins, Immune interferon m the circulation of patients with autoimmune disease, N. Engl. J. Med. 301, 5 (1979). [16]P. E. Lipsky and M. Ziff, Inhibition of antigen- and mitogeninduced human lymphocyte proliferution by gold compounds, J. Clin. Invest. 57, 455 (1977). [17] A. V. Haq, D. G. Mayernite, C. Orosz and J. J. Rinehart, Interleukin-I secretion is not required for human macrophage support of T-cell proliferation, Cell. Immunol. 87, 517 (1984). [18] C. E. Brinkerhoff and P. M. Guyre, Increased proliferation of human synovial fibroblasts treated with recombinant immune interferon, J. Immunol. 134 (5), 3142 (1985). [19] L. Klareskog, U. Forsum and D. Kubelitz, Immune function of human synovial cells: phenotypic and T-cell regulatory properties of macrophage-like cells that express HIM-DR, Arthr. Rheum. 25, 488 (1982). [20] D. R. Green, B. Clive, T. A. Ferguson, K. D. Beaman and P. M. Flood, Production of an antigen-specific component of suppressor inducer factor by a helper T-cell clone: possible role of IL-1 at the interface of immunity and tolerance, Br. J. Rheumat. 24 (Supp.), 105 (1985). [21] P. Scheurich, U. Ucer, M. Killian and K. Pfizenmaier, Differential effects of gamma-interferon on human T-cells during primary activation in vitro. In: Cellular and Molecular Biology of Lymphokines, p. 63. Academic Press, New York 1985. [22] J. G. Hanley and B. Bresnihan, Reduction of peripheral blood lymphocytes in patients receiving gold therapy for rheumatoidarthritis, Ann. Rheum. Dis. 44, 299 (1985).

115

Agents and Actions, vol. 19, 1/2 (1986) [23] S. B. Mizel, J. M. Dayer, S. M. Krane and S. E. Mergenhagen, Stimulation of rheumatoid synovial cell collage-

nase and prostaglandin production by partially purified lymphocyte_activating faction (interleukin 1), Proc. Natl. Acad. Sci. USA, 78 (4), 2474 (1981). [24] D. R. Robinson, A. H. Tashjian and L. Levine, Prostaglandin-stimulated bone resorption by rheumatoid synovia, J. Clin. Invest. 56, 1181 (1975).

[25] S. M. Krane, S. R. Goldring and J. M. Dayer, Interactions

among lflmphocytes, monocytes and other synovial cells in the rheumatoid synovium. In Lymphokines, 7, p. 75 (Ed. E. Pick). Academic Press, New York 1982.