Endogenous dopamine synthesis and DOPA-decarboxylase activity in rat renal cortex

Molecular and Cellular Endocrinology,

Elsevier/North-Holland

21 (1982) 45-54

45

Scientific Publishers, Ltd.

ENDOGENOUS DOPAMINE SYNTHESIS AND DOPA-DECARBOXYLASE ACTIVITY

IN RAT RENAL CORTEX

F. WAHBE, J. HAGEGE, N. LOREAU and R. ARDAILLOU INSERM,

Unit6 64, H6pital Tenon, Paris (France)

Received 29 December 1981; revision received 15 February 1982; accepted 17 February 1982

The dopamine content of either cortical slices or isolated glomeruli prepared from rat kidneys and of their incubation medium was measured at different times both under basal conditions and in the presence of L-DOPA. Production of dopamine from L-DOPA by purified cytosolic proteins of the whole cortex was also measured. Dopamine synthesized by these 3 renal preparations accumulated linearly with time over 60 min. Dopamine produced by the glomeruli was more rapidly released into the incubation medium than that produced by the cortical slices. The dopamine synthetic rate was very low in the absence of L-DOPA but increased rapidly when L-DOPA was added to the incubation medium. No plateau was reached in the range of concentrations studied (O-100 PM) when cortical slices or cytosolic proteins were studied whereas dopamine production by isolated glomeruli reached an equilibrium above 10 PM L-DOPA. Dopamine synthesis in the presence of 100 PM L-DOPA was linearly related to the amount of renal protein. The synthetic rates were 43, 2.2 and 0.2 nmoles . h-’ . mg-’ for the cytosolic proteins, the cortical slices and the isolated glomeruli respectively. Dopamine synthesis by the cortical slices in the presence of 100 WM L-DOPA was progressively inhibited by increasing concentrations of or-methyl-DOPA. Cortical slices prepared from rats treated by benserazide, an inhibitor of L-DOPA decarboxylase, synthesized much less dopamine than those from control rats. These results show that rat renal cortex deprived of neuronal supply can synthesize dopamine in vitro from extracellular L-DOPA. Keywords:

dopamine; renal cortical slices; isolated glomeruli; L-DOPA decarboxylase.

It has been well established that dopamine is produced in vivo by the kidney. Fluorescence techniques provided evidence for dopamine-containing neuronal elements at the glomerular vascular poles of the canine kidney (Dinerstein et al., 1979). Further support for intra-renal production is that the amount of dopamine excreted in the urine is greater than can be accounted for by the total clearance of dopamine from the plasma perfusing the kidney (Ball et al., 1978; Oates et al., 1979). In addition, the kidney from several species contains a large amount of aromatic decarboxylase which converts 3(3,4dihydroxyphenyl)-L-alanine (DOPA) to dopamine (Christenson et al., 1970; Nagatsu et al., 1969; Srinivasan and Awapara, 1978). Using a specific and sensitive radioenzymatic assay (Da Prada and 0303-7207/82/0000-0000/$02.75

0 Elsevier/North-Holland

Scientific Publishers, Ltd.

46

F. Wahbe,J. Hagege,N. Loreau,R. Ardailluu

Ziircher, 1976), we have measured dopamine synthesis by rat renal cortical slices and isolated glomeruli under basal conditions and in the presence of the dopamine precursor, L-DOPA. These studies demonstrate that rat renal cortex can synthesize dopamine in the absence of intact nerves. They also show that glomerular cells share this synthetic function with the other components of the cortex.

METHODS Renal preparatiorxs Male Sprague-Dawley rats (200-250 g) purchased from Janvier (Le Genet, France) were used. Under pentobarbital anesthesia (Nembutal Abbott, 5 mg/lOO g BW i.p.) a cannula was inserted in the lower aorta and a buffered isotonic solution (pH 7.4, NaCl 130 mM, KC1 5 mM, MgS04 1 mM, CaClz 1 mM, NaCHaCOO 10 mM, glucose 5 mM, Na2HP04 4 mM, NaH2P04 1 mM) was perfused after clamping the aorta over the renal arteries and opening a renal vein or vena cava for drainage. Blanched kidneys were removed and used to prepare cortical slices, purified glomeruli or cytosolic proteins. Slices of cortex, approx. 0.2-0.3 mm thick, were prepared with a Stadie-Riggs microtome and cut into 1-2 mm’ fragments. GIomeruIi were isolated as previously described (Sraer et al,, 1979) with minor modi~cations. In short, cortices from 4 kidneys were dissected and minced to a paste-like consistency. The homogenate was pushed successively through a 106pm sieve which excluded the tubules and a 50+m sieve which retained the glomeruli. The suspension was then passed through a 25-gauge needle and centrifuged at 120 Xg for 90 sec. The supernate was discarded, the peIlet resuspended in the same buffer solution, passed again through the needle and centrifuged. This operation was repeated 3 times. Each individual preparation was checked for purity under light microscopy. The final pellet consisted in nearly pure isolated, decapsulated glomeruli. Tubular fragments were always below 2% of the total number of glomeruli. No afferent or efferent arterioles were observed. Recovery was about 5000-15 000 glomeruli per rat (roughly 0.5.--1.5 mg glomerular protein). Cytosolic proteins were prepared from the whole renal cortex according to Weber et al. (1976). The supernate obtained after centrifugation at 100 000 X g for 60 min was taken as the source of DOPA decarboxylase. Chemicals and radioenzymatic assay reagents The following materials were used: catechol-~-methyltransferase extracted from porcine liver (1000-2000 Ufmg protein), 3-hydroxytyram~e hydrochloride (dopamine) and 3-methoxytyramine hydrochloride from Sigma (St Louis, MO, U.S.A.); [‘HI S-adenosyl-L-methionine (S-15 Ci/mmole) from the “Centre d’fitudes NuclCaires” (Saclay, France); L-3,4dihydroxyphenylalanine (L-DOPA) from Fluka AC (Buch, Switzerland); Benserazide from Roche SA (Neuiliy, France); 2-methyl-3(3,4-d~ydroxyphenyl)-L-aIan~e (L-tnethyl-DOPA) from Merck, Sharp and Dohm

Renaldopamine (Paris, France). All the other reagents came from Merck (Darmstadt, many) or Prolabo (Paris, France) and were A grade.

41 West Ger-

Incubation and dopamine assay in the renal preparations and their incubation mediums Cortical slices (2-3 mg protein) and isolated glomeruli (l-2 mg protein) were incubated in 20 and 1 ml respectively of the same buffer as that used in the preparation steps in the presence or absence of LDOPA. A greater volume was necessary with cortical slices in order to obtain a homogeneous suspension. Incubation was carried out at 20-22°C with 02 bubbling from below and under continuous agitation. Incubation was stopped after 60 min by centrifugation at 3000 Xg for 10 min. The pellet and the supernate were separated, frozen and lyophilized. Cytosolic proteins were incubated in a volume of 1 ml under the same conditions except that their concentration in the medium (200 gg protein per tube) was markedly lower than those used for the whole tissue. Protein concentration was determined according to Lowry et al. (1951). The lyophilized renal preparations (renal slices, glomeruli and cytosolic proteins) were sonicated for 30 set in 1 ml of 0.3 M perchloric acid using a Sonifer B-12 apparatus (Branson, Danbury, CO, U.S.A.). The supernate obtained after centrifugation at 27 000 X g for 5 min was kept for dopamine determination. The lyophilized incubation mediums of the renal slices and isolated glomeruli were treated similarly except for the initial sonication. Dopamine was measured by a radioenzymatic procedure using S-[3H]adenosylmethionine and rat liver catechomethyltransferase. The methylated amines were separated by thin-layer chromatography on silica gel and the radioactive 3-methoxytyramine was measured by liquid-scintillation counting (Da Prada and Ziircher, 1976). Results were expressed as moles of dopamine synthesized per mg of tissue protein and per hour. Experiments were repeated 3-8 times. Each point of each experiment corresponds to the results of 2 incubation tubes run in parallel. In vivo studies Rats were injected intraperitoneally with 150 mg/kg benserazide, an inhibitor of aromatic acid decarboxylase, and were sacrificed 60 min later. The kidneys were removed and renal slices were prepared.

RESULTS Dopamine synthesis under basal conditions and in the presence of L-DOPA Dopamine synthesized in the presence of 100 PM L-DOPA by each of the 3 renal preparations accumulated with time (Fig. 1). When renal slices were studied, the concentration of dopamine in the tissue itself rapidly reached a plateau whereas dopamine accumulation in the incubation medium was linear for at least 120 min.

F. Wahbe, J. Hagege, N. Loreau, R. Ardaillou

48 WHOLE

ISOLATED

CORTEX

GLOMERULI

CYTOSOLIC

PROTEINS

IMpamine

(1wnol.

mg protein-l)

r . 4 J

2_

6

3

\ i

9

l-

”

I

I

k

IO

30

60

120 (mid

10

30

60

120 (min)

10

30

60

120 (mNY

Fig. 1. Dopamine synthesis by whole cortex, isolated glomeruli and cortical cytosolic proteins in the presence of 100 PM LDOPA as a function of time. Dopamine was assayed in the renal tissues (closed circles) and in the incubation mediums (open circles) for the two first preparations. The sum of both is indicated as total dopamine synthesis (open triangles). Each point is the mean and each vertical bar twice the s.e.m. of 3 individual values from different experiments.

After 40-min incubation, there was more dopamine in the incubation medium than in the tissue. Total dopamine production per min was constant over 120 min since total dopamine concentration increased linearly with time. Glomerular synthesis of dopamine was clearly smaller than the total cortical synthesis. A small amount of dopamine, never exceeding 0.05 nmole/mg protein, even after 120-min incubation, was found in the glomeruli. The majority of the newly synthesized dopamine was released into the incubation medium. Contrary to what was observed with the cortex, the glomerular synthesis reached a plateau after 60-min incubation. It reached avout one fourth of the cortical production. Relative dopamine synthesis by the cortical cytosolic proteins was the greatest, 4-5 times more than that by the whole tissue. Dopamine accumulated progressively with time in the incubation medium during the first 60 min. There was then a clear decrease in the total amount of dopamine formed. The differences in dopamine synthesis by the three renal preparations are shown on Fig. 2. Dopamine synthesis in the presence of 100 PM L-DOPA was linearly related to the amount of total protein whatever the preparation studied. These experiments showed that dopamine did not accumulate within the glomerular cells. All the regression lines started from the origin. The slopes expressed as nmoles . h-’ . mg-’ were 2.2,0.22 and 43 for dopamine productions by the whole cortex (r = 0.98,~ < O.Ol), the isolated glomeruli (r = 0.99, p < 0.01)

49

Renal dopamine WHOLE

ISOLATED

CORTEX

GLOYERULI

CYtOSOLlC

PROTEINS

totopsmin*

I:nmoLhr-1)

1

L 1

5

8

protshr

1

2

3

4

@w

5 pfotrk, tmrB

0.1

Q2

a.3

a.4

0.5 prot* Fo)

Fig. 2. Dopamine synthesis by whole cortex, isolated glomeruli and cortical cytosolic proteins in the presence of 100 PM L-DOPA after 60-min incubation as a function of the amount of renal protein added to the incubation medium. Dopamine was assayed in the renal tissues (closed circles) and in the incubation mediums (open circles) for the two first preparations. The sum of both is indicated as total dopamine synthesis (open triangles). Each point is the mean and each verticat bar twice the s.e.m, of 6 individual values from different experiments.

WHOLE

CORTEX

Dopamino [nmol.hr-l.mS

ISOLATED

GLOYERULI

I

CYTOSOLIC IO

PROTEINS

4i

Prot4ln-1)

w L DOPA fM1

-i

10-S

-8 * 10-s

10-4 L DO?A (MI

l

/

. 10-e

10-S

10-4 L DOPA (Y1

Fig. 3. Dopamine synthesis by whole cortex, isolated glomeruli and cortical cytosolic proteins after 60-min incubation as a function of L-DOPA concentration in the incubation medium. Dopamine was assayed in the renal tissues (closed circles) and in the incubation mediums (open circles) for the two first preparations. The sum of both is indicated as total dopamine synthesis (open triangles). Each point is the mean and each vertical bar twice the s.e.m. of 8 individual values from different experiments.

so

F. Wahbe, J. Hagege, N. Loreau, R. Ardaillou

and the cytosolic proteins (v = 0.99, p < 0.01) respectively. Expressed as percentage of L-DOPA transformed per hour and per mg of protein, dopamine synthesis was 0.11 and 0.22% for the cortical slices and the isolated glomeruli respectively whereas a much greater value, 43%, was calculated for the cytosolic proteins. In the absence of L-DOPA in the incubation medium, dopamine synthesis was very low. Total production reached 33.2 2 12.1,9.5 * 1.5 and 0.13 + 0.03 pmoles f h-’ . mg-’ for the whole cortex, the isolated glomeruli and the cytosohc proteins respectively. Dopamine synthesis increased rapidly when L-DOPA was added to the incubation medium (Fig. 3). For the slices of whole cortex, it reached about 2 nmoles . h-r ’ mg-’ in the presence of 100 MM L-DOPA. No plateau was reached in the range of L-DOPA concentrations studied (O-100 PM). At any concentration of L-DOPA, there was more dopamine in the incubation medium than in the tissue. For the isolated glomeruli, total synthesis was smaller than for the whole cortex (0.4 nmoles - h-” . mg-’ at 10-100 I.IM L-DOPA). A plateau was reached above 10 PM L-DOPA. Almost all the dopamine synthesized was found in the incubation medium. Dopamine synthesis was the greatest with the cytosolic proteins since it reached 10 nmoles . h-’ . mg-’ m . the presence of 100 PM L-DOPA even if this value was lower than that derived from the slope of the regression line shown on Fig. 2. No plateau was observed in the range of L-DOPA concentrations studied. For the slices of whole cortex and the cytosolic proteins there was considerable variation in the results obtained from individual experiments at the greatest concentrations of L-DOPA. L-DOPA decarboxylase was in part released from the cortical slices into the incubation medium as shown by the results shown in Table I. In this experiment, cortical slices were initially incubated for 1 h without L-DOPA, the slices were then removed and 100 PM L-DOPA was added to the medium. There was some transformation of the L-DOPA into dopamine, approx. 12% of the amount transformed in the presence of the slices.

Table 1 Compared studies of dopamine synthesis (pmoles *h-r . mg-‘1 within the cortical slices and in the incubation mediums after exposure to cortical slices Results are given as means it s.e.m. from 3 individual values. Basal conditions

With 100 yM I,-DOPA

Two-phase experiments a

Cortex

5.8 f 3.0

262 ?: 21

3.5 + 2.3

Incubation medium

3.5 + 1.9

524 * 12

98.2 f 3.5

a For this experiment, “cortex” corresponds to dopamine accumulation within the cortical slices under basal conditions, and “incubation medium” corresponds to dopamine accumulation in the incubation medium after 1 h in the presence of the cortical slices without L-DOPA followed by 1 h without cortical slices with 100 MML-DOPA.

51

Renal dopamine

Effects of inhibitory agents on dopamine synthesis Dopamine synthesis by cortical slices was studied at increasing concentrations of ~-me~yl-DOPA, a specific competitive inhibitor of L-DOPAdecarboxyiase, Basal synthesis of dopamine was not modified when 100 MMa-methyl-DOPA was added. On the contrary, synthesis of dopamine in the presence of 100 MM L-DOPA progressively diminished at increasing concentrations of a-methyl-DOPA. 50% inhibition was observed around 5 PM and approximately 14% of the rate of synthesis persisted with 100 I.IM cy-methyl-DOPA (Fig. 4, left). The effect of benserazide, an inhibitor of the aromatic aminoacid decarboxylases, was also investigated. Dopamine synthesis by renal cortical slices prepared from control rats and rats treated with benserazide were compared. There was no difference when the cortical slices were incubated without L-DOPA or with low concentrations of this aminoacid (1 FM) whereas a marked fall of dopamine synthesis by the slices of the treated rats was observed when these slices were incubated with 1O-1 00 PM L-DOPA (Fig. 4, right).

5M

Dopsmino pmol.hr-l.m@ protein”

DI

I

I

t 401

B

with in viva Eenssraride

cf

without

20

ithout I. Dow

1O-4

U

methyl DqaaiMI

0

104 10-5 to-6

1 Dopa M

Fig. 4. Dopamine synthesis by cortical slices incubated in the presence of increasing concentrations of o-methyl-DOPA with or without 100 PM L-DOPA (left). Dopamine synthesis by cortical slices prepared from rats injected or not with benserazide incubated in the presence of increasing concentrations of LDOPA (right). The heights of the rectangles indicate the means and the vertical bar twice the s.e.m. of 4 individu~ values from different experiments.

52

F. Wahbe,J. Hagege, N. Loreau, R. Ardaillou

DISCUSSION Rat kidney was found producing dopamine in vitro in the absence of neuronal supply. Dopamine synthesis might therefore be both dependent and independent of the sympathetic innervation of the kidney. The endogenous dopamine synthetic rate was greater for the whole cortex than for isolated glomeruli. The site of synthesis may be the epithelial cells which are common to both tubules and glomeruli. Cytosolic proteins were at least 5-10 times more efficient than cortical slices to produce dopamine. The lesser degree of puri~cation of L-DOPA-de~arboxylase and the limited diffusion of L-DOPA into the tissue when slices were used might account for this difference. The physiological role of locally synthesized dopamine is proably similar to the effects which have been demonstrated following vascular infusion of exogenous dopamine. Injected dopamine has multiple actions: it produces renal vasodilatation both in the dog (Goldberg, 1972) and in the rat (Imbs et al., 1979), renin release (Imbs et al., 1975), natriuresis (M~Giff and Burns, 1967) and phosphate diuresis (Cuche et al., 1976). The latter two effects can be explained to a large extent by the renal vasodilatation but a direct tubular action of dopamine cannot be excluded. The data in this report represents the first direct evidence for dopamine synthesis by isolated glomeruli. Glomerular dopamine may stimulate the Iocal production of renin (Imbs et al., 1975) and aIso directly act on the giomerular micro~irculation. It has been shown that in dogs with hemorrhagic shock (Nagakawa et al., 1976) and in rats with clamped renal arteries (Iaina et al., 1977), dopamine reduces the severity of acute renal failure. There are few reports of the renal dopamine concentration under basal conditions. The enclosed data (33 + 12 pmoleslmg protein) are similar to those of Bell and Gillespie (198 1) (approx. 20 pmoles~mg protein) and sIi~tly greater than those of Ball et al. (1978) (3-4 pmolesjmg protein). In the mouse kidney, Wagner et al. (1979) observed similar values around 20 pmoles/mg protein. No reports were found on dopamine synthesis by renal slices in the presence of L-DOPA added in vitro. The only data are those of Romero et al. (1973) who measured renal dopamine concentration in rats previously treated with L-DOPA (100 mg per kg BW and per day during 10 days). Possibly due to the low sensiti~ty of their technique, they could not detect dopamine in the kidneys of the control rats and found approx. 450 pmoles/mg protein in those of the treated rats. Our own results for dopamine generation in cortical pieces are markedly dispersed (0.4-2.2 nmoles * h-r ’ mg-r). Th’rs can be due to variation in accessibility of L-DOPA to the enzyme and also to the difficulty of the radioenzymatic technique itself which includes several steps: lyoph~ization, extraction, enzymatic methylation and ~hromatography, each of them contributing to the magnitude of the final error. L-DOPA decarboxylase has been previously purified from the hog kidney (Srinivasan and Awapara, 1978; Christenson et al., 1970). The K, of the purified enzyme was 190 PM, close to the concentration of L-DOPA used in the present experiments. At this concentration, the enzyme activity was around 400 nmoles * min-’ . mg-’ (Chris-

Renal

dopamine

53

tenson et al., 1970) thus much greater than that observed in our study with cytosolic proteins from the rat cortex (43 nmoles . h-’ * mg-‘). This smaller value may be accounted for by the lesser degree of puri~~ation of the enzyme and the absence of pyridoxal phosphate, cofactor required for a full activity. The experiments with cortical slices or isolated glomeruli investigate several stages of dopamine synthesis, namely transport of L-DOPA into the cells, L-DOPA decarboxylase activity and release of the synthesized dopamine into the incubation medium since only a small part of L-DOPA was transformed in the medium by the released enzyme. In the present study, L-DOPA was found to easily cross the membrane of the renal cells in vitro as already demonstrated from in vivo studies (Romero et al., 1973; Tyce and Owen, 1979). The activity of L-DOPA decarboxylase was blunted by competitive inhibitors such as cr-methyl-DOPA or general inhibitors of this enzyme family as benserazide. Total dopamine synthesized by the slices accumulated linearly with time over 120 min. On the contrary, there was a plateau or even a slight decrease between 60 and 120 min for the glomeruli. Accumulated dopamine synthesized by the cytosolic proteins reached a maximum at 60 min and then clearly decreased. The effect of the enzymes metabolizing dopamine was thus not apparent during the first 60 min of incubation. Dopamine was rapidly released out of the glomerular cells but more slowly from the cortical slices. This was probably due to the ratio surface of exchange over volume which was greater for the glomeruli than for the slices. Although this study clearly shows that renal tissue deprived of neuronal supply can synthesize dopamine in vitro from extracellular L-DOPA, it does not demon strate that this event occurs in vivo. The physiological concentration of L-DOPA in human plasma, approx. 8 nmoles/l (Johnson et al., 1978), is much lower than the concentrations of this aminoa~id which we added in vitro. Since tyrosine hydroxylase is probably intra-neuronal (Nagatsu et al., 1969), local high concentrations of L-DOPA may be produced at the neuronal ends for transformation into dopamine by the extra-neuronal L-DOPAdecarboxylase. Thus the fact that dopamine production may be only under neuronal control in the alive animal cannot be excluded.

This work was supported by grants from the “Institut National de la Sante et de la Recherche Medicale” (80 5 017) and the “Faculte de Medecine Saint-Antoine”. Mrs. Morin and Miss Knobloch provided excellent secretarial assistance.

REFERENCES Ball, S.G., Lee, M.R.,and Oates, N.S. (1978) Clin. Sci. Mol.Med. 55,343P. Ball, S.G., Oates, N.S., and Lee, M.R. (1978) Clin. Sci. Mol. Med. 55, 167.

54

F. Wahbe, J. Hagege, N. Loreau, R. Ardaillou

Bell, C., and Gillespie, J.S. (1981) J. Neurochem. 36, 703. Christenson, J.G., Dairman, W., and Undenfriend, S. (1970) Arch. Biochem. Biophys. 141,356. C&he, J.L., Marchand, G.R., Greger, R.F., Lang, F.C., and Knox, F.G. (1976) J. Clin. Invest. 58, 71. Da Prada, M., and Ziircher, G. (1976) Life Sci. 19, 1161. Dinerstein, R.J., Vannice, J., Henderson, R.C., Roth, L.J., Goldberg, L.I., and Hoffmann, P.C. (1979) Science 205,497. Goldberg, L.I. (1972) Pharmacol. Rev. 24,l. Iaina, A., Solomon, S., Gavendo, S., and Eliahou, H.E. (1977) Biomedicine 27, 137. Imbs, J.L., Schmidt, M., and Schwartz, J. (1975) Europ. J. Pharmacol. 33,151. Imbs, J.L., Schmidt, M., Ehrhardt, J.D., and Schwartz, J. (1979) in: Peripheral Dopaminergic Receptors, (J.L. Imbs and J. Schwartz, Eds.) Advances in the Biosciences, Vol. 20, Pergamon, Oxford, pp. 331-343. Johnson, G.A., Gren, J.M., and Kupiecki, R. (1978) Clin. Chem. 24, 1927. Lowry, O.H., Rosebrough, N.J., Farr, A.L., and Randall, R.J. (1951) .I. Biol. Chem. 193,265. McCiff, J.C., and Burns, CR. (1967) J. Lab. Clin. Med. 70,892. Nagakawa, B., Goldberg, L., Mac Cartney, J., and Matsumoto, T. (1976) Surg. Gynecot. Obstet. 142,871. Nagatsu, T., Rust, L.A., and De Quatro, V. (1969) Biochem. Pharmacol. 18, 1441. Oates, N.S., Ball, SC., Perkins, C.M., and Lee, M.R. (1979) Clin. Sci. 56,261. Romero, J.A., Lytle, L.D., Ordonez, L.A., and Wurtman, R.J. (1973) J. Pharmacol. Exp. Therap. 184,67. Sraer, J,, Sraer, SD., Chansel, D., Russo-Marie, F., Kouznetzova, B., and Ardaillou, R. (1979) Mol. Cell. Endocrinol. 16,29. Srinivasan, K., and Awapara, J. (1978) Biochim. Biophys. Acta 526,597. Tyce, G.M., and Owen Jr., CA. (1979) Biochem. Pharmacol. 28,327l. Wagner, J., Palfreyman, M., and Zraika, M. (1979) J. Chromatog. 164,41. Weber, P.C., Larsson, C., and Scherer, B. (1977) Nature (London) 266,65.