- Email: [email protected]
Microbiological transformations of fusidane-type antibiotics a correlation between fusidic acid and helvolic acid
Tetrahedron Letters N0.47, pp. 4843-4846, 1968. Pergamon press. Printed in Great Britain.
MICROBIOLOGICALTRANSFORMATIONSOF FUSIDANE-TYPE ANTIBIOTICS A CORRELATION BETWEEN FUSIDIC ACID AND HELVOLIC ACID W. von Daehne, H. Lorcn and W.O. Godtfredsen Leo PharmaceuticalProducts, Ballerup, Denmark (Received in UK 10 August 1968; accepted for publication 16 August 1968) The structural work on the antibiotic helvolic acid culminated recently when Okuda et al.' proposed formula I on the basis of convincing chemical and spectroscopicevidence. According to this proposal the compound contains the same framework as fusidic acid (II)2a'b and thus belongs to the fusidane class of antibiotics. The structure of fusidic acid has been verified by X-ray analysis3, and a correlation between this compound and helvolic acid would therefore be desirable. Attempts to achieve this by chemical means have been unsuccessful so far, but we now wish to report an interrelation accomplistedby means of a combination of chemical and microbiological fl transformations . 'Ihisapproach became possible when we found that the fungus Acrocylindrium oryzae, Saw. Kominami produces helvolic acid when grown aerobically in deep culture. 'Ihanksto this observation, we now had at our disposal organisms capable of introducing oxygen functions in the fusidane framework at C-11 (Fusidiwn coccinewn) as well as at C6
and G7
f'hcrocylindrium
oryzae). When 7-deacetoxy-1,2-dihydrohelvolic acid (III)5 (loo ug/ml) was added to an actively growing culture of a mutant of Fusidium coccineum, K. Tubaki, it was rapidly transformed into *2 a more polar compound as indicated by thin layer chromatography (tic) . No trace of this compound could be detected in parallel fermentations without added 111. Isolation of the metabolite *2 was accomplishedby preparative layer chromatography (plc) of an ethyl acetate extract of the acidified culture fluid,
Treatment
of the amorphous acid (IVa) thus obtained with ethereal diazo-
methane gave a methyl ester C32H 4S07*3 (M+ = 544). m.p. 155-156'C, assigned structure IVb on the “1 An elegant interrelationbetween the twu antibiotics has independentlybean achieved by Okuda
Y
et
al.
4
For tic and plc, silica gel HF was used as adsorbent. The solvent system was 10% methanol 254 in methylene chloride.
*3 All compounds for which empirical formulae are given, gave satisfactorymicroanalyses.
4844
No.47
TABLE 1
(n.m.r. data)*
CH-6
CH
0
CH3-LO-
CH-24
(Y-11
Methyl fusidate
5.86/d J = 8.5
5.12/b J=7
4.36/m
3.75/m
3.63/s
1.98/s
IVb
5.84/d J = 6.5
5.06/b
4.42/m
3.82/m
3.61/s
1.95/s
IVd
5.78/d J = 8.5
5.05/b
4.35/m
3.70/m
3.60/s
1.94/s
IVe
5.81/d J = 8.5
5.08/b
4.38/m
3.65/s
1.97/s 2.01/s 2.07/s
IVf
5.83/d J = 8.5
5.06/t
4.47/m
3.62/s
1.97/s 1
V
5.88/d J = 8.5
5.05/b
3.63/s
1.97/s
Compound
CB-7
3.89/m J-8
CH-3
-0-E
CH-16
ca.4.92/m ca.4.9/m
3.90
3.90/m
3
* 'lhespectra were obtained with a Varian HA-100 (100 MC) instrument, CBC13 being used as solvent. lhe line positions are given in h-values and with TMS as internal reference. For characterization of the signals, the following abbreviationsare used: s (singlet), d (doublet), t (triplet), m (multiplet),b (broad, ill-defined signal).
0
I
n
Acrocvl.
Zn/kOAc
oryzae Rl=R3=R4=H; R2=0 Rl=R3=H; R2=O; R4=CH3 Rl=R3=R4=H; R2=H, a-OH ~=R~=H; R2=H,a-OH; R4=CH3 ~=Ac; R2=H,a-OAc; R3=B; R4=CH3 Rl=H; R2=0; R3=OB; K@f3
4845
No.47 basis of the folloting
evidence.
l’he UV spectrum (AiioH
(dioxane), pect
vhich show
to the keto group)
riesl,
similar
but incompatible
acid
Series
fact
that
2b
vitb
. Ihe presence
the chemical
Table 1) correspond f4
methyl fusidate
That the latter
negative
to that reported
tbecurvesreported of a-orientated
shift
(vide --
infra)
afforded
6-keto
the pr+
and As3o2 - -3.54
due to the protons by oxidation
evidence
of the location
of
at G6
and C-7 in fusidic
also
&th
the
spectmrn (cf.
in the spectrum of
with Jones reagent
that IVb snd V could
acid se-
consistent
in the n,m.r,
at C-3 smd Cl1
res-
in the fusidic
at C-3 and C-11 is signals
vith
in the helvolic
3- or 11-monoketoderivatives groups
suggests
from the GDcurve
derivatives
the two BO-~~
characterized
clear-cut
vas inferred
(Ae310 = -3.42
for for
of
to the signals
. IVb was further
at C6
hydroxyl
and splitting
closely
is
Cotton effect
V, C32H4407, m.p. 187-188OC. ‘Ihe fact
ester acid
a strong
indicates
-1 in the infrared smd a band at 1705 cm
sence of the smse chromophoric system as in III, the presence of a six ring ketone.
220 mp (e = 10,150))
to the triketo
be obtained
from fusidic
the oxygen functions
in these
compounds. Introduction
of oxygen functions
Acrocylindriwn
oryzae.
transformation
products
more polar
of
!lhis procedure
me main metabolite
gave three
nals of
indicating
two of
hat
the third
afforded rial
room temp.)
*4 ‘Ihe G3 sidic
proton
is
also
hydroxyl
consistent
acetate
a methyl ester,
three
extract
without of
are
added II.
the acidified
C32 E&O7
follovs
coupling
The chemical protons
the vidth
the equatorial
of
with a neighbouring
aad therefore
b = 3.58,
shift
sig-
i*ld splitting
with
V. Ihe equato-
the n.m.r. proton.
(acetic
signal
due
lhe fomatim
anhydridc/pyri-
unhindered
and the Cl1
HO-{g
with Jones reagent
in every respect frun
m.p.
in methyl fusidate.
that oxidation
of IVd under mild. conditions
resonates-at
= 546),
This compollnd vas assigned
groups.
from the fact
group in IVd follow a diaxial
(k
spectrum (cf. Table 1) contains
187-188OC, identical
acid
(150 &ml)
polarity.
lhe n.m.r.
with
II
by means of
as mnorphous but homogeneous powders,
at 220 mu (c = 9,050).
on acctylation
in 3-epi-fusidic
acid at b I 3.80.
acids
tith
fermentations
to those due to the C-3 and Gil
which indicates (IVe)
ERIethyl
vas accomplished
these compounds, uhich all
in parallel
secondary hydroxyl
C32H4407, n.p.
of the G6
a 3,6,16_triacetate
dine;
three
None of
by plc of
diasomethae
group -is at G6
ester
at G6,
of
by tic.
be detected
decreasing
reasons.
correspond
hydroxyl
conformation
to the proton of
signals
a triketo
organism vas incubated
carboxylic
absorption
the following
the presence
these
of
A gave vith
165-166OC, showing ultraviolet IVd for
this
vas acccmplished
A, B, zmd C in order
structure
of
compound, could
the metabolites
fluid.
designated
a culture
were formed as indicated
than the parent
.Isolation culture
Len
acid
proton
nature
of
the
in 11-epi-fa+
4846
No.47
*5
c-6 hydroxyl group
.
Hctabolite C was assigned structure IVa since treatment with diazomethane gave a methyl ester, C32Hq8Q7t m-p. 155-156OC, identical in all respects with IW
obtained via the sequence I+III +
IVa+IVb. tietaboliteB gave with diazomethane an amorphous methyl ester, C32H4808 (M' = 560), tentatively assigned structure IVf on the basis of biogenetic considerationsand the following observations. The n.m.r. spectrus (cf. Table I) reveals the presence of three secondary hydroxyl groups. One of the HO-kg
signals is similar to that due to the C-11 proton in methyl fusidate. The renaining two
are overlapping, but have the same chemical shift as the CH-3 signal in methyl fusidate. Finally, the C&curve (&336 = - 1.90 and Ac326 = -2.56) corresponds closely to the curves reported for iketc+7a-hydroxyl compounds in the helvolic acid series. 'Iheformation of IVb and V from fusidic acid as vell as from helvolic acid indicates that ttheseantibioticshave a common framework and - in view of the good arguments previously set forth197 for the location and stereochemistryof the ring B substituents in helvolic acid - supports the correctness of formtiLaI. That I also represents the absolute configurationof helvolic acid is a consequence of the fact that II represents the absolute stereochemistryof fusidic acid2b,3.
l5 It is known that the C-11 hydroxyl group in the.fusidic acid series is not acetylated under 2b these conditions. Achrovledganent.We wish to thank Professor G. Ourisson, University of Strassbourg, for the c&curves, Dr. B.C. Das, Gif-sur-Yvette,for the mass-spectra, and Dr. U. Scheidegger, Varian ffi, Ziirich,for the 100 MHz n.m.r. spectra. REFERENCES 1.
S. Okuda, S. Ivasaki, U.I. Sair, Y. Macbida, A. Inoue zmd K. Tsuda, Tetrahedron Letters, 1967. 2295.
2.
a) D. Arigoni, U. van Daehnc, W.O. Godtfredsen, A. Melera, 'andS. Vangedal, Experientia, 20, 344 (1964). b) W.O. Godtfredsen, W. van Daehne, S. VagedaI, D. Arigoni, A. Marquet,and
A. Melera,
Tetrahedron,2l, 3505 (1965). 3. 4.
909 (1968). A. Cooper and D.z Hodgkin, Tetrahedron,3 s. Oktida,Y. Sato, T. Hattori, and H. Wakabayashi,Tetrahedron Letters, =,4847
5. s. Okuda, Y. Nakayama, and K. Tsuda, Chem.Pharm.Bull.,l4, 436 (1966). 6. W.O. Godtfredsen, Fusidic acid zmd Some Related Antibiotics, Copenhagen 1967. (Iamy, chem.connn., 19k7, 729.
MICROBIOLOGICALTRANSFORMATIONSOF FUSIDANE-TYPE ANTIBIOTICS A CORRELATION BETWEEN FUSIDIC ACID AND HELVOLIC ACID W. von Daehne, H. Lorcn and W.O. Godtfredsen Leo PharmaceuticalProducts, Ballerup, Denmark (Received in UK 10 August 1968; accepted for publication 16 August 1968) The structural work on the antibiotic helvolic acid culminated recently when Okuda et al.' proposed formula I on the basis of convincing chemical and spectroscopicevidence. According to this proposal the compound contains the same framework as fusidic acid (II)2a'b and thus belongs to the fusidane class of antibiotics. The structure of fusidic acid has been verified by X-ray analysis3, and a correlation between this compound and helvolic acid would therefore be desirable. Attempts to achieve this by chemical means have been unsuccessful so far, but we now wish to report an interrelation accomplistedby means of a combination of chemical and microbiological fl transformations . 'Ihisapproach became possible when we found that the fungus Acrocylindrium oryzae, Saw. Kominami produces helvolic acid when grown aerobically in deep culture. 'Ihanksto this observation, we now had at our disposal organisms capable of introducing oxygen functions in the fusidane framework at C-11 (Fusidiwn coccinewn) as well as at C6
and G7
f'hcrocylindrium
oryzae). When 7-deacetoxy-1,2-dihydrohelvolic acid (III)5 (loo ug/ml) was added to an actively growing culture of a mutant of Fusidium coccineum, K. Tubaki, it was rapidly transformed into *2 a more polar compound as indicated by thin layer chromatography (tic) . No trace of this compound could be detected in parallel fermentations without added 111. Isolation of the metabolite *2 was accomplishedby preparative layer chromatography (plc) of an ethyl acetate extract of the acidified culture fluid,
Treatment
of the amorphous acid (IVa) thus obtained with ethereal diazo-
methane gave a methyl ester C32H 4S07*3 (M+ = 544). m.p. 155-156'C, assigned structure IVb on the “1 An elegant interrelationbetween the twu antibiotics has independentlybean achieved by Okuda
Y
et
al.
4
For tic and plc, silica gel HF was used as adsorbent. The solvent system was 10% methanol 254 in methylene chloride.
*3 All compounds for which empirical formulae are given, gave satisfactorymicroanalyses.
4844
No.47
TABLE 1
(n.m.r. data)*
CH-6
CH
0
CH3-LO-
CH-24
(Y-11
Methyl fusidate
5.86/d J = 8.5
5.12/b J=7
4.36/m
3.75/m
3.63/s
1.98/s
IVb
5.84/d J = 6.5
5.06/b
4.42/m
3.82/m
3.61/s
1.95/s
IVd
5.78/d J = 8.5
5.05/b
4.35/m
3.70/m
3.60/s
1.94/s
IVe
5.81/d J = 8.5
5.08/b
4.38/m
3.65/s
1.97/s 2.01/s 2.07/s
IVf
5.83/d J = 8.5
5.06/t
4.47/m
3.62/s
1.97/s 1
V
5.88/d J = 8.5
5.05/b
3.63/s
1.97/s
Compound
CB-7
3.89/m J-8
CH-3
-0-E
CH-16
ca.4.92/m ca.4.9/m
3.90
3.90/m
3
* 'lhespectra were obtained with a Varian HA-100 (100 MC) instrument, CBC13 being used as solvent. lhe line positions are given in h-values and with TMS as internal reference. For characterization of the signals, the following abbreviationsare used: s (singlet), d (doublet), t (triplet), m (multiplet),b (broad, ill-defined signal).
0
I
n
Acrocvl.
Zn/kOAc
oryzae Rl=R3=R4=H; R2=0 Rl=R3=H; R2=O; R4=CH3 Rl=R3=R4=H; R2=H, a-OH ~=R~=H; R2=H,a-OH; R4=CH3 ~=Ac; R2=H,a-OAc; R3=B; R4=CH3 Rl=H; R2=0; R3=OB; K@f3
4845
No.47 basis of the folloting
evidence.
l’he UV spectrum (AiioH
(dioxane), pect
vhich show
to the keto group)
riesl,
similar
but incompatible
acid
Series
fact
that
2b
vitb
. Ihe presence
the chemical
Table 1) correspond f4
methyl fusidate
That the latter
negative
to that reported
tbecurvesreported of a-orientated
shift
(vide --
infra)
afforded
6-keto
the pr+
and As3o2 - -3.54
due to the protons by oxidation
evidence
of the location
of
at G6
and C-7 in fusidic
also
&th
the
spectmrn (cf.
in the spectrum of
with Jones reagent
that IVb snd V could
acid se-
consistent
in the n,m.r,
at C-3 smd Cl1
res-
in the fusidic
at C-3 and C-11 is signals
vith
in the helvolic
3- or 11-monoketoderivatives groups
suggests
from the GDcurve
derivatives
the two BO-~~
characterized
clear-cut
vas inferred
(Ae310 = -3.42
for for
of
to the signals
. IVb was further
at C6
hydroxyl
and splitting
closely
is
Cotton effect
V, C32H4407, m.p. 187-188OC. ‘Ihe fact
ester acid
a strong
indicates
-1 in the infrared smd a band at 1705 cm
sence of the smse chromophoric system as in III, the presence of a six ring ketone.
220 mp (e = 10,150))
to the triketo
be obtained
from fusidic
the oxygen functions
in these
compounds. Introduction
of oxygen functions
Acrocylindriwn
oryzae.
transformation
products
more polar
of
!lhis procedure
me main metabolite
gave three
nals of
indicating
two of
hat
the third
afforded rial
room temp.)
*4 ‘Ihe G3 sidic
proton
is
also
hydroxyl
consistent
acetate
a methyl ester,
three
extract
without of
are
added II.
the acidified
C32 E&O7
follovs
coupling
The chemical protons
the vidth
the equatorial
of
with a neighbouring
aad therefore
b = 3.58,
shift
sig-
i*ld splitting
with
V. Ihe equato-
the n.m.r. proton.
(acetic
signal
due
lhe fomatim
anhydridc/pyri-
unhindered
and the Cl1
HO-{g
with Jones reagent
in every respect frun
m.p.
in methyl fusidate.
that oxidation
of IVd under mild. conditions
resonates-at
= 546),
This compollnd vas assigned
groups.
from the fact
group in IVd follow a diaxial
(k
spectrum (cf. Table 1) contains
187-188OC, identical
acid
(150 &ml)
polarity.
lhe n.m.r.
with
II
by means of
as mnorphous but homogeneous powders,
at 220 mu (c = 9,050).
on acctylation
in 3-epi-fusidic
acid at b I 3.80.
acids
tith
fermentations
to those due to the C-3 and Gil
which indicates (IVe)
ERIethyl
vas accomplished
these compounds, uhich all
in parallel
secondary hydroxyl
C32H4407, n.p.
of the G6
a 3,6,16_triacetate
dine;
three
None of
by plc of
diasomethae
group -is at G6
ester
at G6,
of
by tic.
be detected
decreasing
reasons.
correspond
hydroxyl
conformation
to the proton of
signals
a triketo
organism vas incubated
carboxylic
absorption
the following
the presence
these
of
A gave vith
165-166OC, showing ultraviolet IVd for
this
vas acccmplished
A, B, zmd C in order
structure
of
compound, could
the metabolites
fluid.
designated
a culture
were formed as indicated
than the parent
.Isolation culture
Len
acid
proton
nature
of
the
in 11-epi-fa+
4846
No.47
*5
c-6 hydroxyl group
.
Hctabolite C was assigned structure IVa since treatment with diazomethane gave a methyl ester, C32Hq8Q7t m-p. 155-156OC, identical in all respects with IW
obtained via the sequence I+III +
IVa+IVb. tietaboliteB gave with diazomethane an amorphous methyl ester, C32H4808 (M' = 560), tentatively assigned structure IVf on the basis of biogenetic considerationsand the following observations. The n.m.r. spectrus (cf. Table I) reveals the presence of three secondary hydroxyl groups. One of the HO-kg
signals is similar to that due to the C-11 proton in methyl fusidate. The renaining two
are overlapping, but have the same chemical shift as the CH-3 signal in methyl fusidate. Finally, the C&curve (&336 = - 1.90 and Ac326 = -2.56) corresponds closely to the curves reported for iketc+7a-hydroxyl compounds in the helvolic acid series. 'Iheformation of IVb and V from fusidic acid as vell as from helvolic acid indicates that ttheseantibioticshave a common framework and - in view of the good arguments previously set forth197 for the location and stereochemistryof the ring B substituents in helvolic acid - supports the correctness of formtiLaI. That I also represents the absolute configurationof helvolic acid is a consequence of the fact that II represents the absolute stereochemistryof fusidic acid2b,3.
l5 It is known that the C-11 hydroxyl group in the.fusidic acid series is not acetylated under 2b these conditions. Achrovledganent.We wish to thank Professor G. Ourisson, University of Strassbourg, for the c&curves, Dr. B.C. Das, Gif-sur-Yvette,for the mass-spectra, and Dr. U. Scheidegger, Varian ffi, Ziirich,for the 100 MHz n.m.r. spectra. REFERENCES 1.
S. Okuda, S. Ivasaki, U.I. Sair, Y. Macbida, A. Inoue zmd K. Tsuda, Tetrahedron Letters, 1967. 2295.
2.
a) D. Arigoni, U. van Daehnc, W.O. Godtfredsen, A. Melera, 'andS. Vangedal, Experientia, 20, 344 (1964). b) W.O. Godtfredsen, W. van Daehne, S. VagedaI, D. Arigoni, A. Marquet,and
A. Melera,
Tetrahedron,2l, 3505 (1965). 3. 4.
909 (1968). A. Cooper and D.z Hodgkin, Tetrahedron,3 s. Oktida,Y. Sato, T. Hattori, and H. Wakabayashi,Tetrahedron Letters, =,4847
5. s. Okuda, Y. Nakayama, and K. Tsuda, Chem.Pharm.Bull.,l4, 436 (1966). 6. W.O. Godtfredsen, Fusidic acid zmd Some Related Antibiotics, Copenhagen 1967. (Iamy, chem.connn., 19k7, 729.