Inhibition of [3H]nitrendipine binding by phospholipase A2

Life Sclences, Vol 37, pp. 1301-1308 Printed in the U S A.

INHIBITION OF [ ' H ] N I ~ I P I N E

Pergamon Press

BINDING BY PHOSPHCLIPASE A 2

Mark E. Goldman I' * and John J. P1sano 2 '+ 1Sectlon on Molecula~ Pharmacology, Cllnlcal Neurosc~ence Branch, NIMH, NIH, Bethesda, MD 20205, laboratory of Chemistry, I~KBI, NIH, Bethesda, MD 20205 (Received in flnal form July 25, 1985) 9mmmry Phosphollpase A 2 from several sources inhlblted ['H]nltrendiplne bindLng to membranes from braLn, heart and 11eal longitudinal muscle. The enzymes from bee venom and Russell's viper venom were most potent, having IC,o values of approxLmately 5 and 14 ng/ml, respectlvely, in all three membrane preparations. Inhlbition of binchng by bee venom phospholipase A 2 was time- and dose-dependent. Mastoparan, a known facllitator of phosphollpase A 2 enzymatic activity, shlfted the bee venom phospholipase A 2 dose-response curve to the left. Pretreatment of braLn membranes wlth bee venom phosphollpase A 2 (10 ng/ml) for 15 mln caused a 2-fold increase in the Kd wlthout changlng the B~= v compared with untreated membranes. Extension of the preLncuSM~ion period to 30 nun caused no further increase in the KA but signiflcantly decreased t h e n B ~ to 71% the value for untreated m~mbranes. ['H]N1trendlplne, prel ted wlth bee venom phosphollpase A2, was recovered and found to be fully active, indlcatLng that the phosphollpase A 2 did not modlfy the ligand. It is concluded that phospholxpase A 2 acts on the membrane at or near the ['H]nltrendipLne bindlng site and that phospholipids play a key role Ln the interactlons of 1,4 dlhydropyrldlne calc1~ channel antagonists with the dlhydropyridine binding site. The transport of calclum through calci~n channels is essentlal for many blological processes includlng neurotransmltter and hormone re/ease and muscle contractlon (1,2). Calc1~ channel blockers have become an nnportant drug class for the treatment cardiovascular dlsorders mncludlng hypertenslon, angina, atherosclerosls and mlgraLne (3). TWo classes of calc1~ channels have been postulated based upon pharm~cologxcal and physiological studmes. Voltage operated calclum channels are actlvated by membrane depolarlzation whereas receptor-operated calcium channels are activated through stimulation of menbrane receptors by certaxn neurotransmltters or thexr agonmsts (4). Voltageoperated channels are sensltlve to several classes of organlc calcl~ channel antagonists includlng 1,4-dihydropyrldlnes (nitrendlplne, nlfedlplne), phenylalkylamLnes (verapamml, D600) and benzothlazepLnes (diltlazem) (5). ['H]N1trendlplne xs a hxgh affmlty and specxflc llgand employed to label dlhydropyrldlne blndlng sltes in a saturable and stereoselective manner (6,7,8). In some tlssues, such as heart, zleal longmtudlnal nE~scle and bramn, *To whom correspondence should be addressed at Big. 10, Rm. 4N214, NIH, Bethesc]a, MD, 20205

+Deceased 0024-3205/85 $3 00 +

00

1302

Phosphol~pase A 2 and [3H]N~trend~p~ne

Vol

37, No

14, 1985

there is a close correlatlon between the abil~ty of dlhydropyrldlne analogs to •nhlblt ['H]nltrend~plne b~ndlng and calc~um-medlated events or ~ C a fluxes (9,10,11). Calci~n is required for ['H]nltrendlp~ne b~nd~ng; calcium chelators or lanthanum inhlblt binding (12). Verapamll interacts wlth nltrend~plne binding sites in a negative heterotrop~c manner and causes a decrease ~n the ['H]nltrendiplne blndlng afflnlty whlle d~it~azea, acting at the sane allosterlc complex, enhances [3H]n~trendlplne binding (13,14). PhospholIpase A 2 , among several enzymes tested, has been shown to Inh~b~t nltrendlplne blnd~ng to ileal longitudinal muscle membranes (8). However, the nature and potency of thls ~nhlb~t~on were not determlned. To understand better the ~nteract~on between the 1,4-dlhydropyrld~ne class of calcium channel blockers and thelr b~nd~ng sites, we have studied the effect of phosphol~pase A2 from various sources o~ the b~nd~ng of ['H]nltrendlp~ne to membranes from brain, heart and ileal longitudinal muscle. We now report that phosphollpase A~ caused a dose- and tLme-dependent inhlb~t~on of b~nd~ng. As the enzymatic reaction progressed, there was flrst a decrease ~n b~ndlng affinity and then a decrease In both affinity and n~nber of blnd~ng sites.

Materlals and Methods Materlals - Drugs and chemlcals were obtalned from the followlng sources phospholipases (honey bee (Apls mellifera) venom, 1510 U/mg, Russell's vlper (V1pera russelll) venom, 6 U/rag, cobra (Na3a na3a) venom, 465 U/mg, rattlesnake (Crotalus adamanteus) venom, 600 U/mg, and porclne pancreas, 790 U/mg, (Sigma Chemical Co., St. Louis, MO), [3H]nltrendlp~ne ([5-methyl-3H], speclflc actlvlty = 70-78 Ci/mmol), Aquasol (New ~ g l a n d Nuclear, Boston, MA), nifedlplne (generously provided by Pflzer Lahoratorles, New York, NY), mastoparan (Peninsula Laboratorles, Belmont, CA), male rats (200-250g, Sprague Dawley, Taconlc Farms, Germantown, NY), female guinea plgs (200-250g, Hartley, NIH Stock). All other chemicals were reagent or ultrapure grade. Preparatlon of membranes - Rats were decapltated and whole bralns were homogenlzed in 20 volumes of ice-cold Trls-HCI buffer (50 raM, pH 7.7 at 23°C) wlth a Polytron (PT i0 probe, settlng 5.6) for 15 sec. The homogenate was centrifuged at 1,100xg for 10 nun and the supernatant fluld was recentrlfuged at 34,000xg for 20 mln. The resulting pellet was resuspended In 10 volumes of Trls buffer and I00 ~i aliquots were taken for blndlng assays. Gulnea plg ileal longltudlnal muscle was isolated as descrlbed prevlously (8) with the exceptlon that the lletnn was placed in Trls buffer prior to removal of the longitudinal muscle. Membranes were prepared from the longltudlnal muscle by the procedure used for braln. Heart membranes were prepared by a prevlously described procedure (15). Brlefly, rat hearts were hc~nogenlzed with a Polytron In 20 vol of Trls buffer and centrlfuged at 20,000xg for 20 nun. The pellet was rehomogenlzed a second tLme in the same voltmle of buffer and centrlfuged. The pellet was resuspended in 10 vol of Trls buffer and 100 ~i allquots were used in blndlng assays. Standard ['H]nltrendlplne blndln~ assay. Dupllcate or trlpllcate incubatlons were carrled out wlth i00 ~i of [~H]nltrendlplne, I00 ~i of membrane preparatlon (0.2 - 0.3 mg proteln) and 1.8 ml of 50 mM Trls buffer in the presence or absence of drugs or enzymes (7,8,13). The flnal [3H]nltrendlplne concentratlon was I00 pM. Non-speclflc blndlng was determlned by the addltlon of i pM nlfedlplne and speclflc blndlng was taken to be the dlfference between total and non-speclflc blndlng. Samples were incubated for 45 mln at 23°C in

Vol. 37, No. 14, 1985

Phosphollpase A 2 and [3H]Nltrendipine

1303

the dark and filtered over Whatman GF/B filters. The filters were washed two tlmes wlth 5 ml of Ice-cold Trls buffer, placed in I0 ml of ~guasol liquid scintillation cocktail and counted in a liquid scintillation counter. Protein was determined by the method of Lowry et al., (16) with bovine sert~ albumln as the standard. Prelncubatlon of brain membranes and ['H]nltrendlplne. M~abranes were suspended in 11 ml of 50 mM Tris buffer containing I mM c a l c i ~ chloride. Five ml aliquots were incubated in the presence or absence of 50 ng bee venom phosph~ lipase A 2 for 15 or 30 min at 30°C in a shaking water bath. At the completion of the prelncubatlon, 30 ml of ice-cold, calcl~n-free Trls buffer, was added and the suspension was centrifuged at 34,000xg for 20 mln. The pellet was resuspended in 30 ml of fresh buffer and the centrlfugatlon was repeated. The pellet was suspended in 6 ml of Trls buffer. To show that the bee venom phospholipase A, dld not chemically alter the ['H]nltrendlplne, the ligand was prelncubated in the standard manner (45 min, 23°C) wlth I) membranes, 2) membranes and bee venom phospholipase A2, 3) buffer or 4) buffer wlth bee venom phosphollpase A 2 . After the prelncubatlon, the iigand was extracted and compared wlth fresh iigand in the standard assay or in the standard assay containing the phospholipase A 2 solution used in the prelncubation step (to show that the enzyme was active). ['H]Nitrendlplne was extracted from the prelncubated sample after removal of the menbranes by centrlfugation at 34,000xg for 20 min. The supernatant fluid was transferred to a 50 ml glass tube and the ['H]nltrendiplne was extracted into 5 ml of chloroform by shaking for 5 mln. The mixture was centrifuged and the lower chloroform layer was evaported to dryness wlth a nitrogen stream. Buffer was added to the residue and trltlom equlvalent to I00 pM ['H]nltrendlplne was taken for assay In the standard manner. Analysls of Data - The percent inhlbitlon by phosphollpase A 2 was determlned by divldlng the amount of speclflcally-bound ['H]nltrendlplne in the presence of enzyme by the amount bound in t h e a b s e n c e o f enzyme thenmultlplylng by 100. IC,o values were determlnedby Problt analysis. The results were expressed as mean ± standard error of the mean (SE). The Scatchard transformation of the specific ['H]nitredlplne blnchng curve was employed and a stralght llne was fitted by linear regression. The mean K a and Bmau values from each experlment were analyzed statlstlcallyby StudenE's tes~-~or differences between groups of data. Results Phosphollpase A2 from varlous sources inhlblted ['H]nltrendlplne blndlng to braln membranes (Fig. i). Bee venom phosphollpase A2 (IC,o = 9.8 + 1.6 ng/2 ml, n = 7) and Russell's vlper venom, (IC,o = 28.7 ± 5.5 ng/2 ml, n = 3) were most potent. The enzyme from oobra venom (IC,o = 424.0 • 6.5 ng/2 ml, n = 3) also completely inhiblted blndlng at hlgher concentratlons whereas pancreatic phosphollpase A2 inhlbltlon plateaued at approxlmately 62~. The rattlesnake enzyme was least potent, gxvlng only 47-62~ Lnhlbltlon at 100 ~g/2 ml, the hlghest concentration tested.

1304

Phospholipase A 2 and [3H]Nitrendlplne

100

~

/

--t¢.:

7~

Vol. 37, No

I

I

I

I

lO 6

lO 5

lO 4

14, 1985

I

o "ru.I

o

10 lO

10 9

10 8

10 7

PHOSPHOLIPASE A 2 (g)

Figure i. Inhibition of ['H]nitrendlplne binding to brain membranes by phospholipase A S from various sources. ~ z y m e s and iigand were incubated in the standard manner. Data represent the mean of duplicate or triplicate determinations obtained In a single experlment. SE bars were omltted slnce they were always less than 10~ of the mean. Key bee venom (b--A) Russell's vipers (O---O), cobra venom (A ~&), porclne pancreas (e---e), rattlesnake venom (~ ~). The inhlbitlon of ['H]nltrendipine binding by bee venom phosphollpase A S (I ng/2 ml or 3 ng/2 ml, n = 3) was tlme-dependent. As the incubation period increased (30, 60, 120 mln) the percent inhibition of ['H]nltrendlplne increased linearly (Table I). This effect contrasts the tLme-lndependent inb/bltlon of ['H]nltrendlplne binding by nlfedlplne (I or 3 nM). TABLE I Time Course of Inhlbltlon of ['H]Nitrendlplne B1ndlng to Brain Membranes by Bee Venom Phosphollpase As, Mastoparan and Nifedlplne % Inhibition of ['H]Nitrendlplne B1ndlng Time (m~n) Inhibitor (concentration) 30 60 120 Bee Venom Phospholipase (I ng/2 ml) 9 21 33 Bee Venom Phosphollpase (3 ng/2 ml) 30 51 73 Mastoparan (3.4 ~4) 18 24 34 Mastoparan (8.5 ~M) 46 58 69 Nifedlplne (I.0 nM) 54 49 49 Nifedlplne (3.0 nM) 83 74 72 ['H]Nltrendiplne blndlng was determined in the standard manner. Results were expressed as percent inhlbltlon of control ['H]nltrenchplne binding (no inhibitor) at each incubation tlme. The SE were less than I~ of the mean values and were omltted. Similar results were found in other experlments. Scatchard analysls of ['H]nltrendlplne binding to rat braln membranes indlcated the presence of a single population of blnchng sites (Fig 2). In slx separate experiments, the apparent blndlng affinity (Kd) was 158.9 + 14.7 p~4 and the blnchng capacity (Bma ~) was 2 3 . 5 . 8 ± 1 3 . 2 fmol/mg prot (Fig. 2 ) . PreLncubatlon of membranes wlt/%-75ee venom phosphollpase A S (50 ng/Sml, 15 mln

Vol. 37, No. 14, 1985

Phosphollpase A 2 and [3H]Nltrendlplne

1305

30°C) caused a szsnifzcant 2-fold increase of the K~ wlth no change in B,m Y (Fzg. 2A). Extenszon of the prelncubatzon to 30 mi~, caused a statzstzc~l-fyszgnzfzcant decrease zn the Bma x to 71% of the control value as well as a 2-fold zncrease in the Ka (Fig. 2B). Following preincubatzon of brazn membranes wlth 250 ng/Sml o~ bee venom phosphollpase A 2 (30 min at 30°C) there was cemplete loss of speczflc ['H]nltrendzpzne bznd~ng sites (results not shown).

I

I

t

100

200

I

I

i

100

200

BOUND Figure 2. Scatchard transformation of specific ['H]nztrendzpzne bindzng to brazn membranes prelncubated wzth bee venom phospholzpase A=. Brain membranes were incubated wzth (O) or wzthout (O) bee venom phosphollpase A, (10 ng/ml) as desczbed zn the Materzals and Methods for 15 nun (Panel A) or 30 nun (Panel B). Data shown are from one of three experzments each carrzed out zn trzplzcate. Pan,el A_, Control, Kd. = 142.0 -+ 1.9 pM, Bma x = 238.2 ± 17.1 fmol/mg protein (n = ~J. ~nospnoizpase-treated, K d = 292.4 t 2 1 . 5 pM, Bma x = 194.8 ± 6.8 fmol/mg protezn (n = 3). The Kd value s were dlfferent from e a ~ other at the p < 0.01 level of slgniflcance Dut the B,~, values were not slgnlflcantly different. ~anel s control, K a = 175.9 ±"~'8.2 pM, B,~ v = 193.4±4 ±8.0 fmol/mg protein mean -+ SE, (n = 3). l~hospholzpase A,-treat'~q~; Kd = 397.1 ± 7.4 pM, B_,,. = 137.0 -+ 18.6 fmol/mg prot (n=3). The .Kd and Bma x values were dlfferen~at the p < 0.01 and p < 0.05 levels of szgnzficance, respectively. There was not a slgnzfzcant chfference in the Lowry proteln values in the two groups. Mastoparan was a weak inhibitor of ['H]nztrendzplne blndlng. At concentrations of 3.4, 8.5 or 17 pM, mastoparan caused 18, 50 or 60 percent inhibition of blnding, respectzvely (results not shown). At 3.4 pM the peptlde caused a szgnzfzcant (p < 0.05) increase zn the potency of bee venom phospholipase A= reduclng the IC,o from 10.4 + 2.2 ng/2 ml to 4.1 • 0.7 ng/2 ml (n --5) (Fzg. 3). ~'ne lnhlbltzon of binding by mastoparan (3.4 pM, 8.5 pM) was tlmedependent. As the incubation time increased (30, 60, 120 mln), there was a iznear increase in the percent inhzbltlon of ['H]nztrendzplne bzndlng (Table I).

1306

Phospholipase A 2 and [3H]N1trendlplne

100-

-I,'; I

Vol. 37, No

I

I

I

I

10 10

10 9

10 8

10 7

14, 1985

(3

"~'~ 75

-g~- so

z~"

25

r,r

0

10-11

I

I

BEE VENOM PHOSPHOLIPASE A2 (g)

Figure 3. Mastoparan enhancement of the inhibition of ['H]nitrendlplne blndlng by bee venom phospholipase A 2 . Incubatlons were carried out wlth 0.01 - i00 ng bee venom phosphollpase A 2 in the standard manner in the presence (O) or absence (0) of 3.4 ~M mastoparan. Data are from one of flve experlments. Each point is the mean of trlpllcate samples. The SE bars were Omltted slnce they were less than 10~ of the mean. The bee venom phospholipase A 2 dose-response curves carried out in the presence of mastoparan were calculated as a percent of mastoparan alone. Mastoparan alone caused a 17.7 + 2.3 ~ (n = 5) Lnhlbltlon of ['H]nltrendlplne binding. Bee venom and Russell's viper venom phospholipases inhlblted blnding of ['H]nltrend~pine to rat braln membranes as well as to rat heart and guinea plg ileal longitudinal muscle membranes. Using bee venom phospholipase A~ in the standard assay, the IC,o values for at least three separate experlments were 9.8 + 1.6 ng/2 ml, 9.0 ± 1.1 ng/2 ml and 9.7 ± 0.9 ng/2 ml for rat braln, rat heart and guinea plg 11eal longitudinal muscle membranes, respectively. Using Russell's viper venom, the IC,o values were 28.7 ± 5.5 ng/2 ml, 35.2 -+ 8.5 ng/2 ml and 34.6 + 3.2 ng/2 ml, respectively. Furthermore, the shape of the doseresponse curves were superlmposable for the three tissues (data not shown). Prelncubatlon of ['H]nmtrendlplne with bee venom phosphollpase A z dld not alter its binding to brain membranes. ['H]Nltrendlplne was prelncubated wlth membranes, membranes and phosphollpase (i0 ng/2ml), buffer or buffer and phosphc~ llpase (I0 ng/2ml). At the eompletlon of the prelncubatlon, the free ['H]nltren ~ dlplne was extracted and placed in a second incubatlon wlth fresh membranes. Under these eondltlons, the amount of nlfedlplne-sensltlve speclflc blnchng was 25.2 + 3.4, 23.5 + 3.8, 23.4 + 3.9 or 25.1 ± 6.5 fmollmg prot., respectively (n = 3). This is comparable to 29.1 • 1.6 fmol/mg prot. bound when a fresh ['H]nmtrendlplne solutlon was incubated with fresh membranes. The phosph~ iipase Az solutlon used during the prelncubatlon was fully active since when included in the second incubatlon caused an 88% inhlbltlon of ['H]nltrenchplne blndlng.

Vol. 37, No

14, 1985

Phosphollpase A 2 and [3H]Nltrendipine

1307

Arachldonlc acld, a possible active product of phospholipase A2, was a weak inhlbltor of ['H]nltrendiplne binding to brain membranes. In three separate experiments, the ICso was 38.0 + 4 ~M (data not shown). The inhibition of binding by arachldonlc acid was not tlme-dependent. Discussion we have shown that phosphollpases A 2 from various sources were potent inhibltors of ['H]nltrendipine binding to brain membranes. Most potent were the enzymes from bee venom and Russell's viper venom: cobra venom phospholipase A, had intermediate activity, and the enzymes from rattlesnake venom and porcine pancreas were very weak Inhlbitors. Thls wide variation in potency probably is a reflection of the we/l-known difference In substrate speciflcltles of phospholipases (17,15). The bee and Russell's viper venom phospholipases inhibited blnchng to heart and ileal membranes wlth equivalent potencies as for inhibiting binding to brain membranes. Thus, it appears that all three preparations have slmilar phospholipase-sensltive sites. Bee venom phospholipase A z inhibited binding to brain membranes in a tlmeas well as dose-dependent manner. Initially, bee venom phospholipase A 2 caused only a decrease in the affinity. As the incubation period was increased, there was a decrease In both the affinity and B ~ v . In keeping with the concept that inhibition of binding was due to enzyt~ic activity, we observed that mastoparan increased the potency of bee venom phospholipase A,. Thls tetradecapeptlde, discovered in wasp venom (19), binds to phospholiplds and makes them more susceptible to hydrolysis (20). Mastoparan has been shown previously to st~ulate phospholipases in tissues as well as the purified enzyme from various sources (21). At the concentrations employed in the present study, mastoparan had a weak inhibitory effect on binding which could have been due either to stimulation of endogenous phospholipase or to a direct effect on the membrane. The enhanced inhibition by bee venom phospholipase A, observed in the presence of mastoparan supports the premise that destruction of membrane phosphollplds mediates the decrease in ['H]nltrendiplne binding caused by phospholipase A 2 treatment. Since phosphollpases hydrolyze certain synthetic esters as well as phosphoIiplds (17), it was necessary to demonstrate that the enzyme dld not hydrolyze the iIgand whlch contains an ester group. ['H]Nitrendiplne prelncubated with bee venom phospholipase A, was recovered by chloroform extraction and it was fully active in a standard binding assay. Furthermore, all radioactivity was extracted from the aqueous layer by chloroform. No counts were lost when the extract was evaporated to dryness, Lndicatlng no generation of ['H]methanol, the potential radioactive hydrolysis product. The mechanl~ of phosphollpase A2 inhibition of binding is not established. It is possible that the decreased binding of ['H]nitrendiplne occurred because the enzyme hydrolyzed an essential membrane phospholipld in close proxlmlty to the ['H]nmtrendipine binding site. AlterD~tlvely, phospholipase A, activity could have liberated an inhibitory fatty acid or lysophosphollpld (22). Our results do not support the latter possibility since the putative inhlbltors would have been lost durlng the membrane washing step prlor to the incubation wlth ['H]nltrendiplne. Furthermore, arachldonlc acld was only a weak inhibitor of ['H]nltrendiplne binding to brain membranes (IC,o = 35 raM). It is also conceivable that the actions of phospholipase A2 are not related to phospholipid hydrolysis (23).

1308

Phosphollpase A 2 and [3H]N1trendlplne

Vol. 37, No

14, 1985

In conclusion, phosphollpases A2, especially froa bee venoa and Russell's venom, were potent znhlbitors of ['H]nztrenchpzne blndlng to brain, heart and ileal longztud2nal muscle membranes. Inhibition is due to an ezymatlc reactlon(s). Our data indicate that phospholzpzds at or near the ['H]nltrendlpzne binding slte are crucial for maintaining binding. Furthermore, changes In endogenous phosphollpase activity (and subsequently the menbrane phospholipid environment) may regulate physiologically the dlhydropyrldine binding slte.

References

i.

R.P. RUBIN, Calc1~n and Cellular Secretzon, pp. 45-78, Phenua Press, N.Y.

2. 3. 4. 5. 6.

B. KATZ and R. MIL~DI, J . P h y s i o l . 203, 459-487 (1976). R.A. JANIS and D . J . TRIGGLE, Drug D e v e l . Res. 4, 257-274 (1984). T.B. BCLTON, P h y s z o l . Revs. 59, 607-718 (1979~. R.A. JANIS and D . J . TRIGGLE, J . Med. Chem. 2-9, 775-785 (1983). F . J . EHLERT, W.R. ROESKE, E. ITOGA and H . I . YAMAW3RA, L i f e S c l . 30, 21922202 (1982). K.M.M. ~13RPHY and S.H. SNYDER, Eur. J . Pharmacol. 77, 201-202 (1982). G.T. BC~G~R, P. Gnk~30, R. KLOCEDWSKI, E. LUCRGWSKI, H. SIBGEL, R.A. JANIS, A.M. TRIGGLE and D . J . TRIGGLE, J . Pharmacol. Exp. Th er . 225, 291-309 (1983). R.W. TSIEN, Ann. Rev. P ~ s l o l . 4__5, 341-358 (1983). L.B. I~3SHqBtR~3ER, M.K. TICI~ and D . J . TRIGGLE, Canad. 5. P h y s i o l . l~larmacol. 57, 333-347 (1979). D.N. ~DE~EI~ISS and H. SPEDDING, N a t u r e 314, 94-96 (1985). R.J. GOULD, K.M.M. MURPHY and S.H. SNYDER, Proc. Natl. Acad. Scz. USA 7__9, 3656-3660 (1982). H.I. ~ R A , H. S(~OEMAKER, R.G. BCLES and W.R. ROESKE, Bzochem. Bzophys. Res. (br~a. 108, 640-646 (1982). K.M.M. MUPPHY, R.J. GOULD, B.L. LARGEST and S.H. SNYDER, Proc. Natl. Acad. Sci. USA 80, 860-864 (1983). R.J. GOULD, K.M.M. MURPHY and S.H. SNYDER, Mol. Pharmacol. 25, 235-241 (1984). O.H. LORRY, N.J. ROSEBROUGH, A.L. FARR and R.J. RANDALL, J. Biol. Chem. 193, 265-275 (1951). M.K. JAIN, M. ~ , J. ROGERS and G.H. D~4AAS, Bzochem. Pharmacol. 3_/3, 2541-2551 (1984). H. VAN DEN BOSCH, Ann. Rev. Biochem. 43, 243-277 (1974). Y. HIRAI, Y. ~ , T. YASUHARA, H. YOSHIDA and T. NAKAJIMA, Blomed. Res. I, 185-197 (1980). T. HIGASHIJIMA, K. WAKAMAT~3, M. TAK~MITSU, M. FJJINO, T. NAKA3IMA and T. MIYAZAWA, FEBS Lett. 152, 227-230 (1983). A. ARGIOLAS and J.J. PISANO, J. Biol. Chem. 258, 13697-13702 (1983). L.J. RCB~I~TS, II, R.A. LHNIS, J.A. OATES and K.F. AUSTEN, Bxochzm. Blophys. Acta 575, 185-192 (1979). K.R. SOONS, C.C. YANG and P. R O S ~ G , Bzochem. Pharmaeol. 3_/3, 3914-3917 (1984).

( 1982 ) .

7. 8.

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.