The second respiratory chain of Candida parapsilosis: a comprehensive study

Biochimie 71 (1989) 8 8 7 - 9 0 2 ~ ) Soci6t6 de C h i m i e biologique / Elsevier, Paris

887

The second respiratory chain of Candida parapsilosis" a comprehensive study Martine GUERIN, Nadine C A M O U G R A N D , Roland CAUBET, Souad ZNIBER, VELOURS, Stephen MANON, Emmanuel GUELIN and Alain C H E Y R O U

Gis61e

lnstitut de Biochimie Cellulaire et Neurochimie du CNRS, Universit~ de Bordeaux H, 1 rue Camille SaintSaChs, 33077 Bordeaux Cedex, France (Received 5-4-1989, accepted 16-5-1989)

S u m m a r y m The yeast C. parapsilosis CBS7157 is strictly dependent on oxidative metabolism for growth since it lacks a fermentative pathway. It is nevertheless able to grow on high glucose concentrations and also on a glycerol medium supplemented with antimycin A or drugs acting at tl-_z level of mitochondrial protein synthesis. Besides its normal respiratory chain C. parapsilosis develops a second electron transfer chain antimycin A-insensitive which allows the oxidation of cytoplasmic NAD(P)H resulting from glycolytic and hexose monophosphate pathways functioning through a route different from the NADH-coenzyme Q oxidoreductase described in S. cerevisiae or from the alternative pathways described in numerous plants and microorganisms. The second respiratory chain of C. parapsilosis involves 2 dehydrogenases specific for N A D H and N A D P H respectively, which are amytal and mersalyl sensitive and located on the outer face of the inner membrane. Since this antimycin A-insensitive pathway is fully inhibited by myxothiazoi, it was hypothesized that ~lpptrnne

a r ~ t r a n e f p ; - r , ~ r l t n n t n ~ l n t ~ n p n~e,d t h a t ;e ttiff~re~n~ f.~ar, a I h P ~ , i ~ ; t " M c.'a~,,n'x~amo ( ' l - - r v t t ' u - h r a r n ~

b cycle. Two inhibitory sites were evidenced with myxothiazoi, one related to the classical pathway, the other to the second pathway and thus, the second quinone pool could bind to a Q-binding protein at a specific site. Elimination of this second pool leads to a fully antimycin A-sensitive N A D H oxidation, whereas its reincorporation in mitochondria allows recovery of an antimycin A-insensitive, myxothiazol sensitive N A D H oxidation. The third step in thi~ secon/respiratory chain involves a specific pool of cytochrome c which can deliver electrons either to: a th,.'rd pb.osphorylation ~-ite ,or to ,?.~ alternative ";d o~'~ cyto"~". . . . ~on This cytochrome is inhibited by high cyanide concentrations and salicylhydroxamates. alternative oxidase / exogenous NAD(P)H / yeast mitochondria evolution Introduction

The genus Candida is composed of various species, including Candida parapsilosis, itself divided into several groups [1], some of which support pathogenicity [2]. For several years the laboratory has concentrated on the study of the respiratory metabolism of a C. parapsilosis strain, CBS 7154, allocated to the form I of the subgroup I, as defined by Montrocher and Claisse [1], by using a mitochondrial DNA restriction analysis [3].

This strictly aerobe yeast has interesting features, both at the genetic level since its mitochondrial DNA is linear, short, and presents an original organization [3, 4], and at the metabolic level, particularly at the oxidative level. Indeed, although C. parapsilosis is described as a fermentative yeast [5], in strain CBS 7154 grown on a glucose medium, as in other strains related to the same group (Guerin et al., unpublished results), no alcohol dehydrogenase I activity or immunoprecipitable protein have been detected [6] and consequently, C. parapsilosis is unable to

S88

M. Guerin et al.

reoxidize NADH at the fermentative level. However, it should be noted that whereas a constitutive alcohol dehydrogenase fails to be synthesized, an adaptative enzyme, analogous to alcohol dehydrogenase II of S. cerevisiae can be detected when cells are grown on a medium supplemented with glycerol [6]. Despite this constraint, this yeast is able to grow on high glucose concentrations as well on a non-fermentescible carbon source in the presence of inhibitors acting either on mitochondrial DNA encoded protein synthesis or at an oxidative phosphorylation level (see below).

Cellular metabolism Metabolism of glucose The apparent contradiction between lack of fermentative pathway and fast growth on high glucose concentrations led us to investigate the glucose metabolism of C. parapsilosis. Relative activities of glycolytic and hexose monophosphate pathways of C. parapsilosis and their regulation were studied and compared to those of S. cerevisiae, when these yeasts were grown under fermentative or oxidative conditions [7]. Whatever the growth conditions, enzymatic activities of glycolytic pathway, except that of glyceraldehyde-3-phosphate dehydrogenase, tvoro In'ut, or in g'~ n . . . . . ; i ~ ;~ ~-L---".-S . cerevi.................. Ut,,.//a~t:S~a ttiatl ill siae, whereas some enzymatic activities of gluconeogenic and hexose monophosphate pathways were greatly enhanced and showed low sensitivity to catabolic repression. Since the glucose metabolism of yeast grown on glycerol is essentially directed towards gluconeogenesis, 2 control steps of tile glycolytic pathway, phosphofructokinase and pyruvate kinase, were studied. These enzymes, located at the upper and the lower part of such pathways respectively, are allosteric with glycolytic intermediates as homotropic and heterotropic effectors. Phosphofructokinase is regulated by various metabolites, namely ATP and fructose-2,6-bisphophate [8, 9]; the enzyme of C. parapsilosis was less active and less ATP-sensitive than that of S. cerevisiae, but exhibited the same sensitivity towards fructose-2,6-bisphosphate. On the contrary, large differences appeared between both yeasts in pyruvate kinase activity. The maximal enzymatic activity was lower in C. parapsilosis than in S. cerevisiae, and more sensitive to gluconeogenic culture conditions [7].

But the main difference concerned the regulation of the enzyme. In fermentative yeast, such as S. cerevisiae or C. tropicalis, pyruvate kinase activity is controlled by changes in the enzyme concentration and by allosteric mechanisms: the enzyme is largely inactive in the absence of fructose-l,6-bisphosphate, unless very large amounts of phosphoenolpyruvate are present [10] since fructose-l,6-bisphosphate converts the enzyme into a form with a higher affinity for its substrate. In C. parapsilosis, this effector was inefficient on pyruvate kinase, the activity depending only on phosphoenolpyruvate concentration (Figure 1). Consequently, the glycolytic pathway of C. parapsilosis is regulated like that of other oxidative yeasts, such as C. lypolitica [11]. On the other hand, the relative insensitivity to catabolite repression of fructose-l,6-bisphosphatase, and high activities of glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase [7] argued in favour of a very active hexose monophosphate pathway, even when cells were grown under repressed culture conditions. Consequently it appeared that in C. parapsilosis, catabolism of glucose led to large amounts of N A D H and NADPH.

Growth resistance to drugs acting on oxidative metabolism Since growth of C. parapsiiosis was only a result of respiratory metabolism, we tested on a medium supplemented with glycerol, the effect of a large spectrum of drugs known to act on different mitochondrial functions [12]: chloramphenicol, erythromycin and paromomycin bind to mitochondrial ribosomes and inhibit mitochondrial protein synthesis [13]; antimycin A and diuron inhibit electron transfer between cytochromes b and cl [14]; oligomycin inhibits oxidative phosphorylation [15] and CCCP is an uncoupler [16]. Experiments were carried out with several strains of C. parapsilosis (form I) exhibiting some differences due to the natural heterozygosis of the nuclear genome, such as, for example, their growth requirement for synthetic media or their resistance to vanadate or to allyl alcohol, a substrate of alcohol dehydrogcnase, the aldehyde of which was demonstrated to be toxic to S. cerevisiae [17]. All the strains tested showed a high level of tolerance to all the mentioned drugs except for the uncoupler, some of the strains being CCCP-sensitive. Growth rate measurements in liquid to which these drugs had been added showed that except for oligomy-

The second respiratory chain of C. parapsilosis

889

--O

~m

o',.: U) (V c.

0

jl~

LI

j¢

O. ¢D

I~ IX

1

2

,

,

3

4

,

//

Phosphoenolpyruvate 5ram

I

10

Fig. 1. Effect of fructose-1,6-biphosphate on saturation kinetics of pyruvate kinase by phosphoenolpyruvate. Cells were grown on glucose. S. cerevisiae ~ without or • • with i m M fructose-l,6-biphosphate; C. parapsilosis ~ without or o----o with 1 m M fructose-1,6-biphosphatc. Reproduced with the permission of Caubet et al. [7].

cin, they increased the doubling-time. Further experiments demonstrated that oligomycin resistance was not due to a non-permeability of the cell membrane to the drug, but in fact to an oligomycin-insensitivity of ATPsynthase (see below). On the other hand, modifications observed in the different cytochrome spectra (Fig. 2A) and in the pattern of electrophoretic analysis of mitochondrially encoded proteins confirmed that drugs acted on their target. Cellular respiration and its sensitivity to cyanide was then measured on cells grown in the presence of each of the drugs (Fig. 2B). Values of cellular respiration showed that the drug effect varied in the following order, erythromycin being the less inhibitory: erythromycin < paromomycin < chloramphenicol < antimycin A < diuron. The main result concerned the cyanide-insensitivity of cell respiration indicating the relative contribution of an alternative pathway, also drug-dependent: (i) cells grown in the presence of erythromycin or paromomycin exhibited the same sensitivity to cyanide as the

control; (ii) most of the weak respiration observed with diuron was partly inhibited by low cyanide concentrations; (iii) respiration of cells grown in the presence of chloramphenicol, and above all ;.rl the presence of antimycin A, was very resistant to cyanide [12]. All these results and other observations resuiting from experiments carried out nn isolated mitochondria (see below) led us to hypothesize the existence of a mitochondrial alternative pathway specific to oxidation of cytop!asmic nicotinamide adenine dinucleotides produced by oxidation of sugars.

Determination at the cellular level of the branching level of the alternative pathway Since Candida parapsilosis like numerous microorganisms and plants, also possesses an alternative pathway (for review see [18]), the problem was to determine where this pathway merged into the classical respiratory chain. Most of the known alternative pathways are in fact branched at the ubiquinone level, and then the following oxidoreduction step involves reactions

890

M. Guerin et al.

independent of the functioning of the cytochrome pathway; for example, growth of C. lypolitica or C. albicans in the presence of antimycin A is probably due to the activity of the first phosphorylation site [19, 20]. The problem was to determine whether the growth of C. parapsilosis in the presence of antimycin A was due to the functioning of the first or the tbi~d site. Indeed, unlike S. cerevisiae which possesses 2 phosphorylation sites [21], most of the Candida species possess 3 sites [22, 23]. Physiological studies on microorganisms have related the dry weight of living material synthesized by growing cells to the amount of substrate, or to the number of mol ATP produced by catabolism of the energy source [24]. Although the Elsden constant [24], as was initially believed, is not suitable for all the microorganisms or independent of culture conditions, nevertheless, under defined cxperi-

mental conditions, 2 yeasts can be compared. Consequently, experiments were carried out at the cellular level, as compared to those of S. cerevisiae, to determine the relative energetic growth yield obtained from energy sources which involved 2 or 3 phosphorylation sites, and in the presence or in the absence of antimycin A

[75!

Cells were grown in a natural exhausted medium to which various amounts of glucose or glycerol were added to ensure conditions such that the energy source was the limiting factor. Under these conditions, glucose catabolism first involved glycolysis and then oxidation through the respiratory chain of the by-products was coupled to 2 or 3 phosphorylation sites. Glycerol utilization involved specific pathways, but in all cases its oxidation involved only respiration coupled to the second and the third phosphorylation sites [25]. Some energetic parameters were

arbitrary

units

~Control

aa a

b c

100

100 100

notom 100 66

142

10g

I O.C

~~,,,~ a ~ -

°-'-~ • ,

f

o

o

a

~

(~ONTROL PAR0~IONYCI N ERYrHROMYCIN

" --

L~

~DIURON

Q0Z Imgd.w 125_'L-15 1124._15 120'-15

0

/.0_--/-2

c o

:/

oin.

174

27

215

"5

,f - - - ~

CHLORANPHENICOL 1004-10

.c c 5C ~4 32

/~ ~Chloramphenicol

17

72

84

138

/

1B3

31

0-5

~

2

• - &/~~

ANTIHYCIN A .

.

.

70_4-5

. 10

Fig. 2. A: Cytochrome spectra of C. parapsilosis. Spectra were recorded at 77oK on a cell suspension dithionite-reduced versus H20~ oxidized cells. Cells were grown in a 2% liquid medium, supplemented / or not with 34.5 m g / I diuron or 2 mg / mi antimycin A, or 5 g / I erythromycin or 4 g / l chloramphenicol. Cells were harvested in the late exponential phase of growth, washed and suspended in H20 at the same concentration: 156 mg (d.w.)/ml. B: Cyanide sensitivity of the cellular respiration of C. parapsilosis CBS 7154 grown in a 2% glycerol medium supplemented or not with the different drugs. The concentrations of the drugs are the same as in Figure 2A. Cells (4 mg dry weight) were suspended in 3 ml of 50 mM sodium phosphate, 50 raM sodium phtalate, 2% glucose buffer, pH 4.7. Reproduced with the permission of Camougrand et al. [12].

The second respiratory chain of C. parapsilosis

cose or glycerol as substrate, only the third phosphorylation site was functional, consequently the cyanide insensitive pathway merged into the cytochrome pathway at the third phosphorylation site level.

determined, among them Ys = g (d.w.) of cells synthesized/mol carbon source catabolized. From Table 1, it appears that: (i) glucose is a very poor energy source for C. parapsilosis; (ii) in the presence of antimycin A, ¥ s represents one third of that calculated for the control when glucose is the energy source and one half when glycerol is the substrate. Since 3 sites were available for glucose and 2 for glycerol, it was deduced that, in presence of antimycin A, with glu-

The second respiratory chain of C. parapsi-

Iosis Comparative effect o f respiratory chain inhibitors on respiration and phosphorylation From previous experiments, it appeared that growth on high glucose concentrations could be due to the relative insensitivity of the respiratory chain, and mainly that of cytochrome c oxidase, to the catabolic repression [6]. A number of inhibitors acting at various points of the respiratory chain (see Scheme 1) were tested on the oxidation and phosphorylation rates of different substrates involving dehydrogenases located either on internal or external face of the inner mitochondrial membrane. Experiments were carried out on mitochondria isolated following the enzymatic method [26], from C. parapsilosis grown in the absence (condition 1, control mitochondria) or in the presence (condition 2, anti A-mitochondria) of antimycin A added to a complete medium supplemented with glycerol. Results are reported in Table II, from which several lines of evidence can be drawn: (i) Substrates involving internal dehydrogen-

Table 1. Comparative growth-yield of C. parapsilosis CBS 7154 and S. cerivisiae as a function of the energy source. Energy source

Strain

Glucose

C. parapsilosis 172 - 20 S. cerevisiae 230 + 30

Ys

Glucose + anti- C. parapsilosis mycin A S. cerevisiae

61 - 5 22 --- 2

C. parapsilosis S. cerevisiae

107 +- 10 120 - 10

anti- C. parapsilosis S. cerevisiae

53 --- 0 0

Glycerol Glycerol mycin A

+

Ys: molar yieldcoefficient= g (d.w.) cells formedper moi. substrate catabolized. Reproduced with the permission of Camougrandet aL [25].

NADH

ANTIMYCIN

~

/

1

891

A

OH2 / O _ . . . J \

8utanedione Amyta.I

ilS- ~

,,-,-:J .......

I-, ,o i

"

/~ N A D H if

,

.Ct

Scheme1. The secondrespiratorypathwayof C. parapsilosis.

MYXOTHIAZOL

a a3

.

I

I~il -- G ~ I . limlI" ~

NAD~H - - ~v

C _ '

2

KCN

- _ 10raM

M. Guerin et al.

892

by salycil h y d r o x a m a t e ( S H A M ) a k n o w n inhibitor o f alternative p a t h w a y s [28] o r by high cyanide c o n c e n t r a t i o n s ( > 8 m M ) . (iii) A T P synthesis is fully inhibited by antim y c i n A or low c y a n i d e c o n c e n t r a t i o n but n o t b y amytal or SHAM. (iv) A d d i t i o n o f b o t h inhibitors, a m y t a l a n d a n t i m y c i n A , fully inhibits respiration w h e r e a s t h a t o f a m y t a l and S H A M d o e s n o t , indicating that b o t h the f o r m e r d o n o t act on the s a m e p a t h w a y , w h e r e a s b o t h the latter do. (v) In the p r e s e n c e o f antimycin A plus S H A M residual respiration occurs, leading to a

ases are p o o r l y respired by c o n t r o l - m i t o c h o n d r i a and their o x i d a t i o n is fully a n t i m y c i n A sensitive. A n t i m y c i n A - m i t o c h o n d r i a d o not oxidize these substrates. (ii) O x i d a t i o n rate o f substrates involving external d e h y d r o g e n a s e s such as N A D H , N A D P H o r g l y c e r o l - 3 - p h o s p h a t e , is faster t h a n that of internal substrates (5 times f o r N A D H ) and in particular, it is partly a n t i m y c i n A-insensitive. T h e residual respiration (the m a i n o n e for antimycin A - m i t o c h o n d r i a ) , is inhibited either by amytal, which inhibits the N A D ( P ) H d e h y d r o g e n a s e s related to this p a t h w a y [27], or

Table !!. Mitochondrial respiration and phosphorylation rates of C. parapsilosis CBS 7154. lnhibitors

None

Culture conditions

t~-keto glutarate

Ethanol

Succinate

NADH pH6.8

NADPH pH 5.3

VO2

V02

V02

V02

VPho V O 2

1 2

25 0

20 0

31 0

150 70

135 14

89 .

Glycerol 3-phosphate

VPho V O 2

VPho

72

36

.

89 .

.

Antimycin A 0.2p.g./mg protem

1

0

0

0

24

0

25

2

-

-

-

56

.

.

.

.

.

Amytal 2 mM

1 2

5 -

-

-

117 20

130 .

24 .

.

70 .

.

KCN 0.2 mM

1 2

0 -

-

0 -

23 20

0 0

26 .

0 0

.

0 . . .

KCN10mM

1 2

S H A M 1 mM 1 2 Amytal + SHAM

1 2

Antimycin A + SHAM

1 2

25 -

.

31 -

120 22

125 5

-

-

-

-

-

-

110 17

. .

-

KCN 0.2 mM 1 + SHAM 2 Amytal + antimycin A

-

-

.

.

.

. .

12

-

-

8

-

-

14

-

-

0

0

--

--

0

.

0

-

0

nm

-

-

-

-

-

-

.

.

. .

.

. .

.

. . . .

.

. . . .

. .

. .

0

-

. .

-

60

0 .

. .

.

30

-

. .

.

.

.

.

Cells were grown in the absen!e (conditior. 1) or in the presence (condition 2) of antimycin A (2 mg/i) added to a complete medium supplemented with glycerol. Mitochondria were prepared following the enzymatic method [26]. VO: respiration rate: n-atom 0-min I-mg-t. VPho: pnosphorylation rate: 32p-ATP synthesis: nmol.min-l.mg-l, nm: not measurable.

The second respiratory chain of C. parapsilosis weak ATP synthesis, whereas addition of low cyanide concentration plus S H A M fully inhibits respiration and ATP synthesis. Therefore, the obligate transfer of electrons from alternative dehydrogenase to cytochrome c oxidase, when alternative oxidase is blocked led to ATPsynthesis.

Oligomycin-insensitivity of A TPase When tested on cell resistance to drugs, oligomycin does not inhibit growth of C. parapsilosis [12]. Several inhibitors of ATPsynthase, DCCD [291, triethyltin [30] and oligomycin were assayed on ATPase activity and mitochondrial respiration of C. parapsilosis, as compared to S. cerevisiae. C. parapsilosis mitochondria exhib-

893

ited the same sensitivity as S. cerevisiae rnitochondria towards DCCD and triethyltin, but were insensitive to oligomycin (not shown). Effect of oligomycin on a range of concentration was tested at 2 temperatures, 18° and 28oC. Results reported in Figure 3 show that, as expected, 5 - 8 /xg oligomycin/mg protein fully inhibits state 3 of respiration in S. cerewsiae mitochondria whatever the temperature, whereas C. parapsilosis mitochondria are partly inhibited by oligomycin, even at high inhibitor concentrations. However, the inhibitory effect of oligomycin increased with the temperature, suggesting a possible relation with the membrane conformation. Indeed, the structure of C. parapsilosis ATPsynthase seemed to differ

A

!

B 30c

"i

\

v

E

II

O 200-

Ii

[]

[]

0

lb

'

'

.50

''

bopg

tcln'

[]

'

'

~o

Fig. 3. Oligomycin sensitivity of mitochondria from S. cerevisiae and C. parapsilosis. Mitochondria (1 m g / m l ) were suspended in a 0.65 M mannitol, 0.36 mM EGTA, 10 mM Tris-maleate buffer (pH 6.7), 0.3% bovine serum albumin (respiration buffer). State 3 (~---1 C. parapsilosis; e----e S. cerevisiae) and state 4 (t~---~ C. parapsilosis; oo S. cerevisiae) were measured with N A D H as substrate. Experiments were carried out at 18oC (A) or at 28°(2 (B).

894

M. Guerin et al.

slightly from that of S. cerevisiae in 2 ways: polypeptide composition and attachment F1-F0: immunoprecipitation of ATPase from C. parapsilo~is with an antiserum anti-F1 from S. cerevisiae [31] led to precipitation of Fl-subunits but not of F0-sector (M. Guerin et al., unpublished results). These results prompted us to enquire about the structure of subunit 6 of F0-sector of ATPase, since if DCCD reacts only with subunit 9, oligomycin-sensitivity requires an adequate arrangement between subunits 6 and 9 [32]. The ~_ene of Su6 was cloned and seauenced, leading t~o the following conclusions: "(i) the primary sequence was ~ 50% homologous to that of S. cerevisiae; (ii) the known sites conferring the oligomycin sensitivity in S. cerevisiae (Oli 2 and Oli 4 [33]) were not modified in C. parapsilosis; consequently the oligomycin resistance wz~snot due to a modification at these loci (Guelin et al. submitted for publication). All these data led us to the following conclusions: besides its normal respiratory pathway, C. parapsilosis develops an alternative electron transfer chain which allows the oxidation of NAD(P)H through a second route different to that already described for S. cerevisiae [34], or C. utilis [35], or plants [36]: - It is specific for external substrates and not for internal substrates. ~""is antimycin A-iL-~ensitive, and .merelore - ~ ..... does not use the NADH-coenzyme Q oxidoreductase described in S. cerevisiae [34]. - I t is not phosphorylating per se, but branched at the third phosphorylation site level, since ATP synthesis at this very site can be evidenced as well at the cellular level as at the mitochondrial level, - A m y t a l inhibits entry of electrons, and SHAM, like high cyanide concentrations, acts after the branching point of the alternative pathway, on a putative oxidase. To differentiate between the known alternative pathways and that of C. parapsilosis, the latter was termed "second respiratory chain". T h e different segments o f the s e c o n d respiratory chain o f C. parapsilosis

The specific NA DH and NA DPH dehydrogenases Unlike mammalian mitochondria, mitochondria from plants, fungi and yeasts are capable to oxidize externally added NADH. Since NADH cannot cross over the mitochondrial inner mem-

brane [34], the NADH-binding site of the N A D H dehydrogenase responsible for oxidation of cytosolic N A D H must face the intermembrane space in these organisms. C. parapsilosis possesses 2 routes of electron transfer from exogenous NAD(P)H to oxygen. One involves the NAD(P)H-Co-enzyme Q oxidoreductase already described in S. cerevisiae [34] and recently isolated [37], and the other is related to the second respiratory chain. Since involvement of N A D H dehydrogenase located on the outer membrane, which catalyses an antimycin A-insensitive and piericidin-insensitive oxidation of N A D H , had already been described in some organisms [38, 39], we first localized NAD(P)H dehydrogenases of C. parapsilosis [40]. NAD(P)H-ferricyanide oxidoreductase activities and NAD(P)H oxidase activities were measured on mitochondria and on intact mitoplasts isolated from cells grown in the absence or in the presence of antimycin A. Whatever the culture conditions, mitoplasts retained N A D H and NADPH oxidation, which was partly amytal-sensitive, partly cyanide-sensitive. Moreover, NAD(P)H dehydrogenase activities were inhibited by mersalyl, a thiol reagent which does not permeate the inner membrane [41], thus indicating that dehydrogenases involved in the second respiratory chain were located on the outer face of the inner mitochondrial membrane. On the other hand, several lines of evidence indicate that the second chain involves dehydrogenases specific for N A D H and NADPH respectively [40], as distinct from the NAD(P)H: uLiquinone oxidoreductase [34]. (i) pH optima were 6.8 and 5.3 for NADH and NADPH, respectively. (ii) Kin=3 + / - 2 / z M and 817 + / - 2 0 0 / z M for N A D H and NADPH respectively. (iii) In contrast to plant mitochondria [42], both N A D H and N A D P H oxidations were amytal-sensitive, but the former less than the latter. (iii) N A D H dehydrogenase was butanedionesensitive, whereas N A D P H dehydrogenase was not. (iv) N A D H dehydrogenase was more mersalyl-sensitive than N A D P H dehydrogenase, especially in the presence of butanedione. On the other hand, studies on the effect of N A D P + on N A D H oxidation showed that NADH: ubiquinone oxidoreductase had Michaelis-Menten kinetics and was inhibited by NADP ÷, whereas N A D H dehydrogenase re-

The second respiratory chain of C. parapsilosis lated to the second chain had allosteric properties: N A D H is a negative effector and is displaced from its regulatory site by NADP ÷ or NAD÷ (Fig. 4)[40].

recorded. In Figure 5 it can be seen that addition of SHAM promotes only a very little extra reduction of cytochromes when internal substrates were used, whereas with N A D H , SHAM induced a reduction of cytochromes c and aa3. In a second set of experiments, the cytochrome reduction level obtained with N A D H as substrate was measured as a function of cyanide concentration (Fig. 6). The maximal level of cytochrome b reduction occurred at low cyanide concentration, whereas those of cytochromes c and aa3 were dependent on inhibitor concentration. Reduction of cytochrome aa3 proceeded in 2 steps: one part was reduced by 0.1 mM cyanide, the second by 8 mM cyanide. Cytochrome c reduction increased along the range of cyar_dde

Evidence for two pools of cytochrome c Spectrophotometric studies were carried out on antimycin-A mitochondria, either on steady state or on kinetics of the cytochrome reduction resulting from addition of internal or external substrates [43]. First, mitochondria were divided into 2 aliquots, one of these being preincubated with SHAM; then both lots were substrate-reduced and spectra of SHAM + substrate-reduced versus substrate-reduced mitochondria were

NADH

a

895

A

B 100

4-

" - ~28

e,0 m 4.1

NADH

,m

~. 50

b

0 NADH

4

a

le NADP+ rnM)

C

" ~ " I00T 22"-.

AMYTAL

n ,atom O~

!

! 2 min

Fig. 4, Inhibition of NADH oxidation by NADP +. A: Mitochondria from normal cells were purified on a sucrose gradient and suspended in respiration buffer. Respiration was initiated with 1 mM NADH: additions of 9 mM NADP+; (a) respiration monitored in the absence of inhibitors; (b) mitochondria pre-incubated in the presence of 1 mM SHAM and 2 mM amytal; (¢) mitochondria pre-incubated in the presence of 0.2/zg antimycin A.mg protein i. Values in bold type are respiration rates (n-atom oxygen min -1.rag protein-I), B: Experimental conditions are the same as in A, (a). Reproduced with the permission of Camougrand et al. [40].

896

M. Guerin et al.

A

B

A=O.O02

k/

\ I

I

I

:!

550

!

I

~

!

600

!

i

nm

550

~

L

i

i

1

600

l

I

nm

Fig. 5. Cytochrome spectra of antimycin A-mitochondria. A: (a) pyruvate + malate or 2-oxoglutarate-reduced v e r s u s oxidized mitochondria; (b) 1 mM SHAM + substrate-(pyruvate + malate or 2-oxoglutarate) reduced v e r s u s substrate-reduced mitochondria; (e) 10 mM cyanide + substrate-reduced v e r s u s substrate-reduced mitochondria. B: (a) NADH or succinate-reduced v e r s u s oxidized mitochondria; (b) SHAM + NADH-reduced v e r s u s NADH-reduced mitochondria; (¢) SHAM + succinate-reduced v e r s u s succmate-refuced mitochondria. Reproduced with the permission of Guerin and Camougrand [43I.

concentrations to reach a maximum at 10 mM cyanide. This result prompted us to investigate reduction kinetics of cytochrome c with ascorbate (+ TMPD) or N A D H as substrate (Fig. 7). As a control experiment, S. cerevisiae mitochondria were reduced with ascorbate/TMPD and then with NADH: no extra reduction occurred after NADH addition (not shown). However, in antimycin-A mitochondria from C. parapsilosis, only part of cytochrome c was reduced after ascorbate / TMPD addition. Subsequent addition of N A D H promoted a further reduction of cytochrome c, which was inhibited by amytal. Successive reductions of cytochrome c were

independent of the order of substrate addition. The sum of these 2 pools represented only 68% of total endogenous cytochrome c (as determined after dithionite addition) instead of the 83% reduced in S. cerevisiae. These results demonstrated the presence of 2 pools of cytochrome c, one accessible to ascorb a t e / T M P D , and the other reducible by N A D H , which is sensitive to amytal, SHAM ~r~d high cyanide concentrations, and therefore implicated in the second respiratory pathway. The problem which arose concerned the relationships existing between these 2 pools of cytochrome c. Whatever the order of N A D H or ascorbate / TMPD addition, the total amount of

The second respiratory chain o f C. parapsilosis

no different maximal absorption peak could be detected and up to now assays for separating both cytochromes c have not been reproducible. However, previous determination of halfreduction potentials of the different cytochromes was carried out with mitochondria isolated from cells grown in the absence or in the presence of antimycin A. Measurements were made with the following wave-length pairs: 540-550 nm, 558-575 nm and 600-630 nm for cytochrome e, cytochromes b and cytochrome aa3 respectively. Under these conditions, 2 cytochromes b (Era 6.8 = -95 -+ 5 mV and +40 _+ 10 mV respectively) and 2 cytochromes c (Era 6.8 = +70 -+ 10 and +240 +- 10 mV respectively could be detected. Also both hemes a and a3 (Em 6.8 = 240 -+ 10 and 360 +_ 20 mV respectively) were evidenced (Guerin and Ohnishi, unpublished results).

o--

0

~. 0.2 E

/

0

E e,o

~ 0.1 0 0

'

.1

.2

'

.3

"//

.

.

1

897

.

.

3

.

.

.

5

.

.

.

7

.

.

9

KCN mM

Fig. 6. Measurementof the cytochromereduction level as a function of cyanide concentration. AntimyeinA-mitochondria (3.8 mg protein/ml) were suspendedin a 0.65 M mannitol 0.36 mM EGTA, 10 mM Tris-maleate buffer, pH 6.8 and oxidized at 0°C by shaking. Cytochromeswere recorded at room temperature; NADH-reducedversus oxidized mitochondria. Aliquots of KCN were added in the reduced sample. Cytochrome content was determined between the followingwavelengthpairs: cytochromeaa3, ZIG 600-630 = 24 mM-l.cm-I (e--o); cytochrome b, zl~ 560-575 = 18 mM-~ .cm-z (l---i); cytochrome c, /t~ 540-550 = 18 mM-]-cm-] (V--V). Reproduced with the permission of Guerin and Camougrand[43].

cytochrome c remained unchanged. However, we observed a more significant reduction of cytochrome c by NADH when this former substrate was added. Moreover, when mitochondria were preincubated with SHAM, cytochrome c was almost completely reduced; only a residual extra reduction was obtained after NADH addition. Subsequently, it seems that if each pool of cytochrome c can be attributed to a particular pathway, either the main or the secondary one, substrates specific to these pathways could reduce part of the cytochrome c in the other pool. Some experiments were carried out to characterize this second cytochrome c. Unfortunately,

Intermediate step between N A D ( P ) H dehydrogenases and eytoehrome c Several lines of evidence account for this from experiments described above: cytoplasmic NADH and NADPH, resulting from glycolytic and hexose monophosphate pathways functioning, or from glycerol oxidation [7], are partly oxidized through specific NAD(P)H dehydrogenases [40]. Then, these electrons are delivered to the third phosphorylation site at the level of a specific cytochrome c [43], through an unknown step which does not seem to involve the classical coenzyme Q-eytochromes b cycle since it is antimycin A-insensitive. These results prompted us to inquire about this intermediate step between dehydrogenases and eytochrome c, by studying the potential role of quinones in this pathway. First, mitochondria were progressively depleted from endogenous quinone according the method of Nordling et al. [44], and at each step, antimycin A-insensitivity of residual respiration was measured. Then, the organic extract containing quinones was reincorporated according to Nordling et al. [44] and the same measurements were carried out. From results reported in Table lII, it can bc seen that: (i) lyophilisation of mitochondria decreased the rate of N A D H oxidation but did not affect the functioning of the antimyein-insensitive pathway; (ii) first extraction with pentane led to a 42% residual respiration, which was fully inhibited by antimycin A; (iii) reincorpotation of the pentane extract in mitochondria promoted 63% recovery of respiration (of the initial rate), but the main result was that this

898

M. Guerin et al.

respiration was again partly antimycin A-insensitive: (iv) after 2 pentane extractions, 80% of the respiration was lost, but reincorporation of organic extract led to recovery of respiration that was partly antimycin A-insensitive. The same experiment was carried out with antimycin A-mitochondria, but after pentane extraction, leading to 95% loss of respiration, the organic extract failed to be reincorporated. In a second set of experiments, myxothiazol, another inhibitor of the bcl complex, was assayed. Indeed, antibiotic inhibitors of the b c l complex that have been characterized up to now can be classified into one of the 2 groups, depending on which of the 2 pathways of cytochrome b reduction they block. In mammalian mitochondria, it has been shown that antimycin blocks the pathways of b reduction through center "i", in the terminology of the Q cycle [45], by

A*T V

----I

a I N A DH

A-T T

binding to a site close to the heme of b-562 [46, 47, 48] and destabilizing ubisemiquinone [49]. Myxothiazol blocks the pathway of cytochrome b reduction through center "o" by binding proximal to the heme of b-566 [50] and apparently blocking oxidation of ubiquinol at a site on the iron-sulfur protein of the b c l complex [51]. The effect of myxothiazol was tested on N A D H oxidati,~n as a function of the inhibitor concentration in the presence or in the absence of antimycin A (Fig. 8). First, it can be seen that myxothiazol alone fully inhibits N A D H oxidation, but this inhibition proceeds in 2 steps with half-inhibition constants equal to 0 . 0 2 5 / z g / m g protein and 0 . 7 / ~ g / m g protein respectively. In the presence of antimycin A, only high inhibitor concentrations were effective. On the contrary, when mitochondria were treated with pentane (leading to a 100% antimycin A sensitive respi-

Aft f

tb

I~,~.N

NADH

~,.T

-T-SHAM e

A DH

allly,~au

~ ADH

s2°; A.T

A=O005 lmin Fig. 7. Kineticmeasurements of cytochromec reduction. Antimycin A mitochondria (2.5 mg protein/mi) were suspended in the buffer described in Figure 6 and oxidized by shaking at (PC. Measurements were recorded at 5oC between 540-550 nm. Additions: 5 mM NADH; 5 mM ascorbate/TMPD (A + T); 2 mM amytal; 2 mM SHAM. (a) Mitoehondria from S. cerevisiae; (b-e) mitoehondria from C. parapsilosis. Reproduced with the permission of Guerin and Camougrand [43].

The second respiratory chain of C. parapsilosis

25o1

Table Ul. Mitochondrial respiration after quinone extraction.

";=..2O0 C

Respiration % rate n at 0-min-~-mg-]

•

= e

t~ i. Q.150

Control

E

0.1O~ I

i

L,-

5

0 0

899

0.2

,

0.4

~

0.6

1

Myxot h iazo I

1.2

pg

Fig. 8, Effect of myxothiazol on respiration rate of C. parapsilosis mitocbondria. Mitochondria were suspended in a 0.65 M mannitol, 10 mM Tris-maleate pH 6.7, 0.36 mM EGTA. Respiration was started by addition of substrate and then increments of myxothiazol were added. Average of 2 experiments. ~ NADH 2 mM; B - - I mitochondria were preincubated with 0.2 p,g antimycin A before N A D H addition; e - - e mitochondria were submitted to one extraction with pentane (see Table III); & - - & substrate = 2 - oxoglutarate.

ration) only the high affinity site was operative. When inhibition of oxoglutarate oxidation was assayed, only one inhibitory site was evidenced, with a half-inhibition constant close to the high affinity site, and corresponding to the very site evidenced in mammalian mitochondria (not shown). From these results it appeared that the second respiratory chain could branch the main one at the level of the Rieske protein (which is the target of the myxothiazol [51]. Since it has been postulated that myxothiazol could displace the quinone from the iron-sulfur protein [51], one can suppose that 2 quinone pools could exist in C. parapsilosis, one related to the classical coenzyme Q - c y t o c h r o m e b cycle and the other to the second respiratory chain, able to bind to the Rieske protein or to an unknown Q-binding protein.

The alternative oxidase Measurements of the cellular respiration (Fig. 2) indicated that in cells grown in the presence of antimycin A, the second respiratory pathway became predominant and was highly cyanide-

65%

Aleter

lyophilisation 219

100

53%

92

42

100%

Extract I + supernatant I 140

63

64%

Second extraction

19

100%

52

50%

First extraction

0.7

273

Antimycin sensitivity

Extract II + supernatant II

41

Cells were grown on a glycerol medium and mitochondria were prepared following the glass beads method [56]. They were lyophilized and then treated with pentane according to Nordling et al. [44], either once (extract I) or twice (extract II). Organic extract (supernatant I or If) was then reincorporated [44]. Mitochondria were then suspended in respiration buffer and respiration initiated with 3 mM N A D H , in the presence or in the absence of 0.2/~g antimycin A / m g protein.

resistant. Under these culture conditions, cellular respiration was 85% inhibited with 1 mM SHAM, whereas it was SHAM insensitive when cells were grown on glycerol only (not shown). When cytochrome spectra were recorded on whole cells that had been glucose-reduced in the presence of 10 mM cyanide versus glucosereduced in the presence of 0.1 mM cyanide, only 2 peaks appeared: cytochrome c at 548 nm and a peak at 590 nm. The height of the peak between 575 nm and 590 nm was dependent on the cyanide concentration and reached a maximum at 10 mM cyanide [43]. The same experiment was carried out on isolated mitochondria, which had been N A D H reduced in the presence of 10 mM cyanide versus N A D H reduced in the presence of 0.2 mM cyanide, the cytochrome spectrum being recorded at 77°K (Fig. 9). It can be seen that, besides classical cytochromes b, c, cl, aa3, a peak appeared

M. Guerin et al.

000

.

i

~

I

_

~

450

•

I

i

i

I

I

600

550

L '

,-

c,

o

l

n m

Fig. 9. Cytochrome spectra were recorded at 77"K on mitochondria from cells grown on a glycerol medium, and harvested in the early stationary phase, NAOH-reduced + 10 mM KCN v e r s u s NAOH-reduced + 0.2 mM KCN.

at 590 nm, with a corresponding Soret band at 449 nm, close to that of cytochrome a at 444 nm. A similar result was previously described [43] when cytochrome spectra of NADH-reduced +_ SHAM versus oxidized mitochondria was recorded. Addition of SHAM to antimycin Amitochondria induced the reduction of a cytochrome absorbing at 590 nm. This increase in the reduction at 590 nm suggests that SHAM, which is a metal chelator [28], could be a specific inhibitor to this cytochrome. From these experiments it appeared that growth on antimycin A induced the synthesis of a cytochrome absorbing at 590 nm. This cytochrome has been described in numerous prokaryotes, and named cytochrome al [52]. But, except for Nitrobacter sp., cytochromes al and aa3 do not generally occur in the same organism. In an attempt to determine whether cytochrome 590 should be a modified form of cyto-

chrome c oxidase, this latter enzyme was isolated and purified [53]. Although its polypeptide composition was different from that of S. cerevisiae, namely at the level of nuclear DNAencoded subunits, the spectral and enzymatic characteristics of the cytochrome c oxidase of C. parapsilosis were close to those of S. cerevisiae.

Conclusion Among yeasts, the genus Candida has been less studied than the genus Saccharomyces, and only few species have been investigated. The Candida parapsilosis strain (CBS 7154) described in this paper presents a number of interesting features both at the genetic level [3, 4] and at the oxidative metabolism level. Indeed, this strictly oxidative yeast, unable to ferment [6] but able to grow

The second respiratory chain of C. parapsilosis

on high glucose concentrations is resistant to a large spectrum of drugs inhibiting either mitochondrial D N A encoded protein synthesis or functioning of the cytochrome pathway, due to the presence of a unique second respiratory chain, different from that described in numerous organisms which is termed alternative or branched respiratory chain (for review see [18]). Some of the alternative routes described in the literature implicate an iron-sulfur containing protein [54] or involve a direct oxido-reduction of the quinone cycle [55]. These cyanide-insensitive pathways branch off from the main respiratory chain at the ubiquinone level, the alternative oxidase being non phosphorylating. It will be noted that C. parapsilosis can also develop such an alternative pathway when grown under specific conditions, which thus forms an addition to classical and second respiratory pathways [12], The capacity of C. parapsilosis to grow on non fermentescible carbon source in the presence of antimycin A offers a potent tool for studying the different segments of its second chain which is specific to the reoxidation of cytoplasmic NADH and N A D P H produced in large amounts by functioning of glycolytic and hexose monophosphate pathways [7]. It involves 2 dehydrogenases, specific for N A D H and NADPH respectively, which are amytal and mersalylsensitive [40], and which deliver electrons to a specific cytochrome c [43] through a quinone pool different from the coenzyme Q - cytochrome b cycle. From cytochrome c, which has been postulated to be an intermediate step between the cytochrome pathway and the second respiratory chain, electrons are delivered either to cytochrome aa3 or to cytochrome 590 (or cytochrorne al). The specific characteristics of C. parapsilosis raised the question of its place in the evolutionary process: indeed, although C. parapsilosis resembles other oxidative yeasts in numerous activities (for example those related to glucose catabolism), it also present some analogies to plant mitochondria, namely as far as NAD(P)H dehydrogenases are concerned, and with bacteria as regards the second respiratory chain: in such organisms, several types of branched respiratory systems exist, which involve participation of several cytochromes, according to culture conditions. Moreover, organisation of cytochrome c oxidase is more complex than that of lower eukaryotes and resembles the mammalian enzyme, indicating a complex regulation.

91H

C. parapsilosis presents similarities to prokaryotes as well as to lower and highcr cukaryotes, plant or mammals, and could constitute a step at the cross-roads of evolution.

References I Montrocher R. & Claisse M.L. (1984) Cell Mol. Biol. 30,291-301 2 Smith S.M., Lee E.Y,, Cobbs C.J. & Eng R.H.K. (1987) Arch. Path. Lab. Med. 111, 71-73 3 Camougrand N., Mila B., Velours G., Lazowska J. & Guerin M. (1988) Curr. Genet. 13, 445 - 449 '1 Kovac L,, Lazowska J. & Slonimski P.P. (1984) Mol. Gen. Genet. 135. 367-37l 5 Lodder J. (1970) The Yeast, a Taxonomic Study. Elsevier North-Holland, Amsterdam, 2nd edn. 6 Guerin M., Camougrand N., Velours G. & Guerm B. (1982) Eur. J. Biochem. 124,457-463 7 Caubet R,, Guerin B. & Guerin M. (1988) Arch. Microbiol. 149,324-329 8 Gancedo C., Sa[as M.L., Giner A. & Sols A. (1965) Biochem. Biophys. Res. Commun. 20, 15-20 9 Bartrons R., Van Schaftingen F., Vissers S. & Hors H.G. (1982) FEBS Lett. 143, 137-140 10 Haeckel R., Hess B., Lauterborn N. & Wuster K.H. (1968) Hoppe Seyler's Z. Physiol. Chem. 349, 699-714 11 Hirai M., "Tanaka A. & Fukui S. (1985) Biochim. Biophys. Acta 391,282-291 12 Cz~,nougrandN., Velours G. & Guerin M. (1986) Biol. Cell 58, 71-78 13 Vasquez D. (1974) FEBS Lett. 40, 563-584 14 Convent B. & Briquet M. (1978) Eur. J. Biochem. 82,473-481 15 Lardy H.A., Johnson D. & De Murray W.C. (1958) Arch. Biochem. Biophys. 78,587-597 16 Heytler P.G. (1936) Biochemistry 2, 357-361 17 Wills C. & Phelps J. (1975) Arch. Biochem. Biophys. 167, 627-637 18 Degn H., Lloyd D. & Hill G.C. (1977) (eds.) Functions of Alternative Terminal Oxidases. Pergamon Press, Oxford, New York 19 Henry M.F., De Troostenberg J.C. & Nyns E.J. (1977) in: Functions of Alternative Terminal Oxidases (H. Degn, D. Lloyd and G.C. Hill, eds.) Pergamon Press, Oxford, New York, pp. 55-65 20 Sheperd M.G., Chin M.C. & Sullivan P.A. (1978) Arch, Microbiol. 116, 61-67 21 Ohnishi T., Kawagushi K. & Hagihara B. (1966) J. Biol. Chem. 241, 1797-1806 22 Downie J.A. & Garland P.B. (1973) Biochem. J. 124, 123-134 23 Gallo M. & Azoulay E. (1974) Biochimie 56, 1129-1143

t)02

M. Guerin et al.

24 Bauchop I. & Elsden S.R. (1960) J. Gen. Mitrobiol. 23,457-469 25 Camougrand N., Velours G. & Guerin M. (1987) Biol. Cell61, 171-175 26 Guerin B., Labbe P. & Somlo M. (1979) Methods Enzymol. 55, 149-159 27 Camougrand N., Caubet R. & Guerin M. (1983) Eur. J. Biochem. 135,367-371 28 Schonbaum G.R., Bonner W.D., Storey B.T. & Bahr J.T. (1971) Plant Physiol. 47, 124-128 29 Sebald W., Graf T. & Lukins H.B. (1979) Eur. J. Biochem. 93,587-599 30 Dawson A.P. & Selwyn M.J. (1975) Biochem. J. 152,333-339 31 Esparza M., Velours J. & Guerin B. (1981) FEBS Lett. 134, 63-66 32 Nagley P. (1988) T.I.G. 4, 46-52 33 Senior A.E. & Wise J.G. (1983)J. Membr. Biol. 73, 105-124 34 Von Jagow G. & Klingenberg M. (1970) Eur. J. Biochem. 12,583-592 35 Bruinenberg P.M., Van Dijken J.P., Kuenen J.G. & Scheffers W.A. (1985) J. Gen. Microbiol. 131, 1043-1051 36 Laties G.G. (1982) Ann. Rev. Plant. Physiol. 33, 519-555 37 De Vries S. & Grivell L.A. (1988) Eur. J. Biochem. 176,377-384 38 Ohnishi T., Sottocasa G. & Ernster L. (1966) Bull. Soc. Chim. Biol. 48, 1189-1203 39 De Santis A. & Melandri B.A. (1984) Arch. Biochem. Biophys. 232,354-365 40 Camougrand N., Cheyrou A., Henry M.F. & Guo6n ~ t l O ~ JI. Gen. lA*:__-L.'-, ,,* VlltlUOlOl, I,~¢~ 3195-3204 . . . . . . . . . .

k = ~ u ]

41 Klingenberg M., Durand R. & Guerin B. (1974) Eur. J. Biochem. 42, 135-159 42 Koeppe D.E. & Miller J. (1972) Plant Physiol. 49,353-357 43 Guerin M. & Camougrand N. (1986) Eur. J. Biochem. 159, 519-524 44 Nordling B., Glazek E., Nelson B.D. & Ernster L. (1974) Eur. J. Biochem. 47,475-482 45 Mitchell P. (1976) J. Theor. Biol. 62,327-367 46 Berden J.A. & Opperdoes F.R. (1972) Biochim. Biophys. Actu 267, 7-14 47 Dutton P.L., Erecinska M., Sato N., Mukai Y., Pring M. & Wilson D.F. (1972) Biochim. Biophys. Acta 267, 15-24 48 Bowyer J.R. & Trumpower B.L. (1981) J. Biol. Chem. 256, 2245-2251 49 Ohnishi T. & Trumpower B.L. (1979) J. Biol. Chem. 255, 3278-3284 50 Becket W.F., Von Jagow G., Anke T. & Steglich W. (1981) FEBS Lett. 132, 329-333 51 Von Jagow G., Ljungdahl P.O., Graf P., Ohnishi T. & Trumpower B.L. (1984) J. Biol. Chem. 259, 6318-6326 52 Ingledew W.J. (1977) in: Function of Alternative Terminal Oxidases (H. Degn et al., eds.), Pergamon Press, Oxford, pp. 79-87 53 Camougrand N., Kadenbach B. & Guerin M. (1987) J. Bioenerg. Biomemb. 19,495-503 54 Henry M.F., Bonner W.D. Jr. & Nyns E.J. (1977) Biochim. Biophys. Acta 460, 94-100 55 Rich P.R. & Moore A.L. (1976) FEBS Lett. 65, 339-344 56 Lang B., Burger G., Doxiadis I., Thomas D.Y., . . . . . . . . . . .w. & K audewitz F . (i977) Ana.i BioDa,u=uw chem. 77, 110-121