Structure-Function Relations in the Inhibition of Murine Contact Hypersensitivity by Amiloride

Structure-Function Relations in the Inhibition of Murine Contact Hypersensitivity hy Amiloride Ann M. Lindgren, Richard D. Granstein, Jimichi Hosoi, and Richard L. Gallo* MGH/Harvard Cutaneous Biology Research Center, Department of Dermatology, and "The Boston Children's Hospital, Harvard Medical School, Boston, Massachusetts, U,S,A,

Topical application of amiloride, a potent inhibitor of the Na"'"/H"'' antiport, inhihits cutaneous inflammation induced hy ultraviolet radiation or contact hypersensitivity in mice. Amiloride analogues with greater and lesser inhihition of the Na"''/H"*' exchange were tested to determine whether anti-inflammatory effects correlate with this activity. Structural analogues of amiloride without significant activity at the Na"''/H"'" antiport (pyrazine, pyrazinamide, and chloropyrazine) failed to inhihit contact hypersensitivity. N-amidino-3-amino-5-dimethyl amino-6-chloropyrazinecarhoxamide (DMA) has a 23-fold greater affinity for the Na"'"/H''' antiport compared to amiloride, hut failed to inhihit contact hypersensitivity in this assay. 3,5-diamino-6-chloropyrazine-amido-guani-

A

miloride hydrochloride is used clinically as a potassium-sparing diuretic. Although its specific mechanism of action has not been elucidated, many studies have shown that amiloride inhibits ion transport, particularly at the Na'^/H *^ antiport [1-3]. Our laboratory demonstrated previously that topical application of amiloride inhibited tissue inflammation caused by contact hypersensitivity (CHS) or ultraviolet radiation [4]. Amiloride also has been shown to inhibit the release of several cytokines (interleukinslp, -6, and -8; tumor necrosis factor) hy pulmonary macrophages [5], The mechanism of action hy which amiloride inhihits inflammation remains unknown. Its effects may be due to the blockade of ion transport and thus intracellular signaling [6,7], inhibition of cell proliferation [8-10], or inhibition of protein synthesis [11-14]. This study investigated further the mechanism(,s) of amiloride action hy using amiloride analogues. Structural analogues, as well as analogues that are either more or less potent for inliibiting the N a ^ / H ^ antiport [3,15,16], were examined for the ability to inhihit CHS, penetrate murine skin, and affect protein synthesis. MATERIALS AND METHODS Reagents Amiloride hydrochloride was purchased from Sigma Chemical Co. (St, Louis, MO), Pyrazine, pyrazinamide, and chloropyrazine were Manuscript received July 18, 1994; accepted for publication September 6, 1994, Reprint requests to: Dr, IVichard D, Granstein, Massachusetts General Hospital-East, Cutaneous Biology Research Center, Bldg, 149, 3rd Floor, 13 th Street, Charlestown, MA 02129. Abbreviations: CHS, contact hypersensitivity; DCG. .3,5-diamino-6chloropyrazine-amido-guanidine; DMA, N-amidino-3-amino-5-dimethyl amino-6-chloropyrazinecarboxamide; TNCB, 2,4,6-tHnitrochlorobenzene,

dine (DCG), which has only 7% of the affinity of amiloride for the antiport, suppressed contact hypersensitivity as well as amiloride. Experiments examining the ahility of these agents to diffuse through mouse skin revealed amiloride to he superior to hoth DCG and DMA, which were approximately equal. DMA, w^ith greater inhihition of the Na^/H"*" antiport hut lesser ahility to inhihit contact hypersensitivity, inhihited protein synthesis and induced cell death more than amiloride or DCG. Amiloride and DCG hold promise as topical anti-inflammatory agents. Their anti-inflammatory properties do not correlate with affinity for the Na"''/H"*' antiport, ahility to penetrate murine skin, or inhihition of protein synthesis. J Invest Dermatol 104:38-41, 1995

purchased from Aldrich (Milwaukee, Wl), N-amidino-3-aniino-5-dimetliyI amino-6-chIoropyrazinecarboxamide (DMA) and 3,5-diamino-6-chloropyrazine-amido-guanidine (DCG) were the kind gifts of Shiseido Co,, Ltd, (Tokyo, Japan), The structures of DMA and DCG were verified by nuclear magnetic resonance. The chemical structures of these agents are shown in Figl, Animals Female mice of the strain C57BL/6, 6 - 8 weeks of age, were purchased from The Jackson Laboratory (Bar Harbor, ME). Outbred Swiss athymic nude mice (nu/nu) were obtained from the Radiation Biology Laboratoi-y at Massachusetts General Hospital, Mice were housed in a facility with a 12-h on/12-li off light setting and bad ad lihitum access to mouse chow and chlorinated water, CHS Reaction Mice were sensitized on their shaved abdomens with 20 /xl of 2% wt/vol 2,4,6-trinitrochlorobenzcne (TNCB) (Eastman Kodak, Rochester, NY) in a 4:1 solution of acetone and corn oil. Negative control animals received no abdominal coating, Elicitation was performed 5—7 d later by coating both dorsal and ventral ear surfaces with 5 /xl of 1% TNCB in a 4:1 solution of acetone and corn oil. The 24-li ear swelling was measured in a coded fashion with a spring-loaded micrometer (Fowler, Biggsfield, England), Amiloride and analogues were dissolved in distilled water hefore they were added to the proper weight of hydrated petrolatum (Denison Pharmaceuticals, Pawtucket, RI), Chloropyrazine, a liquid, was added directly to the hydrated petrolatum. At 1 h and 12 h after elicitation of the CHS reaction, both surfaces of the ears were covered with a thin layer of the indicated ointment. The negative and positive controls received hydrated petrolatum alone. Five mice were assigned to each group. The two-tailed Student t test for unpaired samples was performed to determine the statistical significance between each group and the positive control. Protein Synthesis Assay The murine transformed keratinocyte cell line PAM 212 was maintained in RPMl-1640 with addition of 1 mM sodium pyruvate, 10 mM Hepes buffer, 0,1 mM essential and 0,1 mM nonessential

0022-202X/95/S09,50 • SSDI0022-202X(94)00249-7 . Copyright © 1995 by The Society for hivestigative Dermatology, Inc, 38

VOL. 104, NO. 1 JANUARY 1995

INHIBITION OF CHS BY ANALOGUES OF AMILORIDE

NH,

Table I. Comparison of Amiloride and Analogues in Inhibition of Aniiloride-Sensitive ^^Na"^ Flux by Cultures of Chick Skeletal Muscle Cells Amiloride/ Analogue

Concentration in JLLM Necessarj- to Inbibit Na ' Flux by SO'V!,"

DMA Amiloride DCG

0.3 7 100

N PYRAZINE

N

CONH2

-NH2 ^NH2

PYRAZINAMIDE

C

DIMETHYLAMILORIDE (DMA)

Cl ^NH2

CHLOROPVRAZINE

3,5-DIAMINO-6-CHLOROPYRAZINEAMIDO-GUANIDINH (DCG)

Figure 1. Amiloride and analogue structures. Comparison of the structures of tbe pyrazine ring, the amide of the pyrazine ring, and a chlorinated pyraziue ring with amiloride, DMA, and DCG.

39

" From Vigne P, Frelin C, Cragoe EJ, Lazdunski M: Structure-activity relationships of amiloride and certain of its analogues in relation to the blockade of the Na / H exchange system. Mo! Pharmaco! 25:131-136, 19S4 |3].

failed to inhibit tbe CHS response compared to tbe positive control at any concentration used. However, DCG produced potent inbibition of tbe response in tbe same concentration range as amiloride. Indeed, in many experiments, a greater inbibition of tbe response was observed at low concentrations of DCG compared to equivalent concentrations of amiloride (data not sbown). Because tbe activity of tbese agents at the sodium-proton antiport does not appear to correlate with the ability to inbibit tbe elicitation of CHS, other structural analogues of amiloride, wbicb do not have activity at the sodium-proton antiport, also were examined for their ability to inbibit CHS. Tbree compounds were tested, representing tbe base pyrazine ring (pyrazine), tbe amide of tbe base pyrazine (pyrazinamide), and a cblorinated pyrazine ring (cbloropyrazine). As sbown in Fig 3, none of tbese compounds bad inbibitory activity at tbe concentrations tested.

amino acids, 2 mM L-glutamine, 1f)0 U/ml penicillin, 100 /xg/ml streptomycin, and 10% heat-inactivated fetal bovine serum (Gibco Laboratories, Grand Island, NY). Cells were grown in tissue culture at 37°C in a 5% CO-, humidified incubator. Cells were washed and plated in 96-well roundbottom plates at a concentration of 5 X lO' cells/well. Twenty-four bours later, wliile the cells were still stibconfluent, fhe medium was removed and new medium added witb indicated additions. Emetine, a potent inhibitor of protein syntbesis | I 7 | , was added to parallel cell cultures to inhibit protein synthesis as a control. Cells were ctiltured for an ndditioi-.al 18 b and tben labeled with ^^H-leucine for 6 b. Wells were harvested with a semiautomated cell harvester, and ladioactivify was assayed by liquid scintillation counting. Counts per minute were normalized to cell number. Cells from parallel cultures were used for cell counting. All groups were done in triplicate. Skin Penetration Tbe subctitaneous fat was removed mecbanically from the dorsal skin of nude mice and placed between dual glass diffusion chambers. Solutions of amiloride or analogues in 1 niM Dulbecco's phosphate-buffered saline (Gibco) were placed in the epidermal-side chamber, and phosphate-buffered saline alone was placed on the dermal side. The glass chambers tben were placed in a 31°C water bath. At tbe times indicated, difFusion of the test substance tlirough tbe dermis was measured by a spectroHuorimeter. The fluorescence measured was compared to standard concentration curves for each solution prepared simultaneously. Emission scans determined ma.xininni excitation at the wavelength 370 nm, wliicli was used for eacb experiment. RESULTS Inhibition of CHS To ascertain wbether the activity of amiloride in inhibiting sodium-proton excbange correlates witb antiinflammatory activity, we examined amiloiide and two of its analogues, witb different potency in inbibition of tbe sodiumproton antiport, for tbeir anti-inHammatory activity in tbe CHS assay. Table I compares tbe inbibitory activity of tbe Na'VH^ antiport. One of tbese, DMA, has a 23-fold greater ability to inbibit tbe sodium-proton antiport compared to tbe parent compound [3,15,1 6J. Tbe otber, DCG, bas only 7% of tbe inbibitory activity of amiloride for the antiport (3]. As shown by the data in Fig 2, DMA

0.0

1.0 concentration %

Figure 2. Inhibition of the Na*/H^ antiport does not correlate with inhibition of CHS. Mice were sensitized to TNCB and 5-7 d later challenged with .5 fil of 1% TNCB to both dorsal and ventral ear surfaces. At 1 h and 1 2 b later, both ear surfaces were coated witb a thin layer of tbe indicated ointment at different wt/wt concentrations, and CHS was assessed. DMA (closed cirdes) was unable to inbibit tbe CHS response at any concentration. DCG (closed .'itjiiares) significantly suppressed CHS in a dose-dependent fnsbion from O..S% to 2% wt/wt (2%, p < O.OOI; 1%, p = 0.017; O..S%, p < 0.001; 0.1%, p = 7.33) Amiloride (ii;)ra squares) also inhibited tbe CHS response at 0.5% to 2% (2%, p = 0.001; 1"/., p = 0.002; 0.5%, p = 0.023; 0.1%., p = 0.945). These data represent one of four experiments performed, n = 5 for all groups, except for 2% amiloride and O.rXi DMA, for which n = 4. Data points represent the mean ± standard error of the mean.

40

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

LINDGREN ET AL

20000

1.0 2,0 concentration % Iwt/wt)

0,0

1,0

2,0

4,0

concentration (mM)

Figure 3. Structural analogues of amiloride do not inhibit CHS. At

Figure 5. The effect of amiloride and analogues on protein syn-

1 h and 12 h after elicitation of CHS to 1% TNCB, amiloride and amiloride analogue ointment,* were applied as described in Materials and Methods, and CHS was assessed. The base ring pyrazine (closed triatigles) failed to inliibit CHS at any concentration, Pyrazinamide (open circles) and chloropyrazine (closed circles) were unable to inhibit CHS, Amiloride (opett squares) significantly inhibited CHS at all concentrations (2%, p < 0,001; 1%, p < 0,001; 0,1%, p = 0,004), Data points represent the mean ± standard error of the mean.

thesis. PAM 212 cells were plated in a 96-well flat-bottom plate. After 24 h incubation, the culture medium was changed to medium containing amiloride (open squares) or the analogue, DMA (closed circles) or DCG (close square.^). Cells were cultured further for 18 h and labeled with ''H-leucine for 6 h. Cells from parallel cultures were used for cell counting. This experiment represents one of two experiments performed, •'H-leucine incorporation = 18,861 cpm in untreated cells and 273 cpm in cells treated with 3.6 X 10~^ niM emetine as a control for maximal inhibition of protein synthesis.

Penetration o f Murine Skin To examine the possihility that the discrepancy in the activities of DMA and DCG for the sodium-proton antiport and the relative potencies of these agents in inhibiting CHS could be due to differences in the ability of these

c o 0)

o c

0

4

8 time (hours)

Figure 4. Transit of amiloride and analogues through murine skin. One-millimolar solutions of amiloride (open squares), DMA (closed circle.^), and DCG (open triangles) in Dulbecco's phosphate-buffered saline were placed on the epidennal side of dual glass diffu,sion chambers. Diffusion through the dermis was measured by a spectrofluorimeter at the times indicated. This experiment represents one of four experiments performed. Emission scans determined maximum excitation at the wavelength 370 nm, which was used for all experiments. Concentradon curves for each compound were done simultaneously.

agents to penetrate murine skin, we designed experiments to assess directly transcutaneous movement of these compounds. As shown by the data in Fig 4, amiloride was found to diffuse through murine skin more rapidly then DMA or DCG. DMA and DCG penetrated with approximately equal facility. Of interest, the lesser penetration of DCG contrasted with its equal or better CHS-inhihitory activity compared with amiloride. Because DMA and DCG penetrated the skin with approximately equal ability, the relative concentrations at putative sites of action would not be expected to he different. Therefore, neither differences in activity at the sodium-proton antiport alone nor penetration ability can explain the relative differences in anti-inflammatory activity ohserved for these compounds. Inhibition of Protein or DNA Synthesis Amiloride is known to have activities other than inhibition of ion exchange. For example, amiloride inhihits both protein and DNA synthesis at some concentrations [10-12], We examined whether the relative abilities of amiloride, DMA, and DCG to inhibit protein or DNA synthesis correlated with anti-inflammatory activity. With regard to effects on protein synthesis, DMA was the most potent agent in inhibiting protein synthesis (Fig 5), Similarly, DMA was the most potent agent in inhibiting thymidine uptake by PAM 212 cells, whereas amiloride and DCG were roughly equivalent (data not shown). For all agents tested, concentrations that inhibited protein synthesis also affected cell viability as assessed by trypan blue exclusion, DMA was more potent than amiloride or DCG in inhibiting piotein synthesis and in induchig cell death. However, despite the efficiency of DCG in inhibiting CHS, DCG was much less potent than amiloride for inhibition of protein synthesis and for induction of cell death. As a whole, these data suggest that neither inhibition of protein synthesis or DNA synthesis, nor effects on cell viahility, account for the relative potencies of these agents in inliihiting inflammation.

VOL. 104, NO. 1 JANUARY 1995

DISCUSSION Tbese experiments address tbe mechanism(s) of action of topical amiloride in inhibiting CHS. Affinity for the Na^/H"*" antiport failed to correlate with the ability to inhibit CHS. DMA, witb a 23-fold greater inhibitory activity tban amiloride for the Na'*^/H^ antiport, was unable to inhibit tbe CHS response. However, DCG, witb only 1V« of the activity of amiloride for the antiport, similarly suppressed CHS. This suggests that inliibition of tbe Na'' / H ^ antiport activity does not play a crucial role in the inliibition of CHS observed with these agents. The abilities of amiloride, DCG, and DMA to inhibit protein synthesis and to induce cell death did not correlate witb anti-inflamfnatory activity. Similarly, differences in the ability of tbese compounds to pefietrate murine skin do not explain tbe relative anti-inflammatory activities observed. Tbese in vitro results, altbough important, must be interpreted appropriately. Although tbe effects of amiloride on Na^/H"'" exchange and on protein synthesis are not thought to be cell-typespecific, tbe actual relevant targets of amiloride and its analogues in vivo remain unclear, as do the concentrations over time of these agents at putative relevant sites in vivo. Clearly, botb topical amiloride and its analogue DCG have promise as effective topical anti-inflammatory agents. Their precise mechanism(s) of action warrant further investigation.

Special thanks to Sheseido Go. for synthesis of the DMA and DGG compounds and nuclear magnetic resonance of their structures.

REFERENCES 1. 2.

Sariban-Sohraby S, Benos DJ: Tbe amiloride-sensidve sodium channel. Am f Physiol 250:Ci75-Cl90, 1986 Benos DJ: Amiloride: a molecular probe of sodium transport in tissues and cells. Am f Physio! 242:C131-C145, 1982

INHIBITION OF CHS BY ANALOGUES OF AMILORIDE

3.

4.

5.

6.

7. 8.

9.

10. 11.

12.

13.

14. 15.

16.

17.

41

Vigne P, Frelin C, Cragoe EJ, Lazdunski M: Structure-acdvity relationsliips of amiloride and certain of its analogues in relation to tbe blockade of the N a * / H * exchange system. Mo! Phannaco! 25:131-136, 1984 Gallo RL, Granstein RD: Iiiliibidon of allergic contact dermadds and ultraviolet radiation induced tissue swelling in the mouse by topical amiloride. Arch Dermato! 125:502-506, 1989 Rolfe MW, Kunkel SL, Rowens B, St.™diford TJ, Cragoe EJ, Strieter KM: Suppression of human alveolar liL-icropbage-derived cytokines by amiloride. Amf Respir GeU Mo! Bio! 6:576-582, 1992 Swanson PD, Schellenberg GD, Anderson L: Effects of amiloride and analogs on Na /Ca e.xcliange in synaptic plasnialenima vesicles and on metabolism in cerebral slices. J Nmroc/icm l:SI60C, 1983 Tang C, Presser F, Morad M: Amiloride selecdvely blocks die low threshold (T) calcium channel. Science 240:213-215, 1988 Mills GB, Cheung RK, Cragoe EJ, Grinstein S, Gelfand EW: Acdvadon of the Na / H andport is not required for lecdn-induced proliferadon of human T lymphocytes. J/mmrimi/ 136:1150-1154, 1986 Hesketb TR, Moore JP, Morris JDH, Taylor MV, Rogers J, Smith GA, Metcalfe JC: A common sequence of calcium and pH signals in the mitogenic sdmulation ofeukaryodc cells. Nnlnre 313:481-484, 1985 Kocb KS, Leffert HL: Increased sodium ion influx is necessary' to inidate rat hep,itocyte proliferation. CeU 18:15.3-163, 1979 Rijezewijk J, Bauer F, Boezeman J, Happle R, Mier PD: Recruitment of quiescent (Go) cells following epidermal injury is initiated by acdvadon of the phosphoinositol cycle._/ Invest Dennato! 90:44-47, 1988 Yaniaguclii DT, Sakai R, Bahii L, Cragoe EJ, Jordan SC: Amiloride inbibidon of DNA synthesis and immunoglobulin production by activated buman peripheral blood mononuclear cells is independent of sodium/bydrogen andport. f Inwwno! 137:1300-1304, 1986 L'Allemain G, Francbi A, Crague E, Pouyssegnr J: Blockage of the N a * / H * antiport abolisbes growth factor-induced DNA synthesis in fibroblasts. J Bio! Chem 259:4313-4319, 1984 Lubiu M, Calm F, Coutermarch BA: Amiloride, protein synthesis, and acdvadon of quiescent cells. J GcU Physio! 113:247-251, 1982 Renner EL, Lake JR, Cragoe EJ, Scharschmidt BF: Amiloride and amiloride analogs inliibit Na+/KH--transpordug ATPase and Na*-coupled alanine transport ill rat hepatocytes. Biocliim Biop!iys Acta 938:386-394, 1988 Vairo G, Argyriou S, Bordun A, Goiida TJ, Cragoe EJ, H,nmilton JA: Na + /H"^ exchange involvement in colony-stimulating factor-l-stimulated macrophage proliferadon. JBiW Chem 265:16929-16939, 1990 GroUman AP: Structural basis for inhibidon of protein synthesis by emetine and cyclohexamide based on an analogy between ipecac alkaloids and glutarimide .indbiodcs. Proc Natl Acad Sci USA 56:1867-1874, 1966