A novel aspect of the reaction of ivermectin aglycon producing 13-substituted-14Z derivative

Article abstract of DOI:10.1246/bcsj.75.2509

During research to synthesize 13-deoxy-13β-substituted ivermectin aglycons, unexpected unique products, 13β-substituted-14Z ivermectin aglycons, were obtained. We will report the results of that series of studies and a possible reaction mechanism.

Full text of DOI:10.1246/bcsj.75.2509

c. Jpn.
© 2002 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 75, 2509–2515 (2002) 2509
e Chemical
ciety of Ja-
n
e Chemical
ciety of Ja-
n
A Novel Aspect of the Reaction of Ivermectin Aglycon Producing
13-Substituted-14Z Derivative
13-Substituted-14Z Ivermectin Derivatives
Y. Sugiyama et al.
Yoko Sugiyama, Hiroaki Takahashi, and Akio Saito*
02
09
15
SJA8
09-2673
162
Medicinal Chemistry Research Lab., Sankyo Co., Ltd., 2-58, Hiromachi 1-chome, Shinagawa, Tokyo 140-8710
(Received June 12, 2002)
During research to synthesize 13-deoxy-13β-substituted ivermectin aglycons, unexpected unique products, 13β-
substituted-14Z ivermectin aglycons, were obtained. We will report the results of that series of studies and a possible re-
action mechanism.
02
02
.2
Milbemycins, which are isolated from Streptomyces hygro-
scopicus,1 have attracted a great deal of interest, as they are
known to be potent anthelmintic products.2 Ivermectin (22,23-
dihydroavermectin)3 is also known to possess activity against
endoparasites similar to that of milbemycins. These two com-
pounds have a similar structure, both having a 16-membered
ring. The significant structural differences are the substituents
at the 13-position and the 25-position. That is, ivermectin has
a 4-O-α-L-oleandrosyl-α-L-oleandrosyl group at the 13-posi-
tion with an α-configuration, while milbemycin has no func-
tional group in this position (Fig. 1), and ivermectin has an s-
butyl group at the 25-position while milbemycins have a meth-
yl or ethyl group there.
It was reported by Mrozik et al.4 that the substituent at the
13-position greatly contributes to the anthelmintic activity of
ivermectins. Thus a great number of 13-substituted milbemy-
Fig. 1. Structures of milbemycins and ivermectin. The 3D structure of 13-hydroxy milbemycin A4 was calculated by MM2 and
MOPAC based on the X-ray structure of milbemycin A.

2510 Bull. Chem. Soc. Jpn., 75, No. 11 (2002)
13-Substituted-14Z Ivermectin Derivatives
have established a practical method to synthesize 13β-substi-
tuted milbemycin derivatives from 15-hydroxymilbemycin via
allyl cation 1 (Fig. 2) in our previous paper;6 2) The reaction of
the 13α-hydroxyivermectin derivative was also likely to pro-
ceed in a similar manner via allyl cation to produce the desired
13β-substituted derivative; and 3) The steric hindrance of the
α-side of the reaction site (Fig. 2) would also facilitate obtain-
ing the desired 13β-substituted derivatives from the 13α-hy-
droxy derivative.
Regardless of the predictions above, the reaction did not
proceed as we expected. It gave a mixture of three derivatives
(Fig. 3): a small quantity of desired 13β-14E derivative 2,
some undesired 13α-14E derivative 3, and, to our surprise,
13β-14Z derivative 4. Those structures were determined by
NMR spectra and mass spectra:7 2 has a distinct doublet as α-
C-13H at 4.91 ppm, with a large coupling constant (10.2 Hz);8
3 has a distinct multiplet as β-C-13H at 5.05 ppm;8 and 4 has a
doublet as α-C-13H at 5.23 ppm, approximately 0.3 ppm low-
er than 14E-derivative,9 and also has a multiplet as C-17H at
3.82 ppm, approximately 0.3 ppm lower than 14E-derivative.9
Thus we examined the series of reactions to obtain the desired
product as described below.
Fig. 2. Space filling model of the usual allyl cation.
cin derivatives also have been synthesized and their activities
evaluated, and they were found to have anthelmintic activity as
potent as ivermectin, although there is a difference between
their configurations at the 13-position, α (ivermectin) and β
(13-substituted milbemycins).
The structure-activity relationships of milbemycin and iver-
mectin derivatives have also been studied by examining the
substituent at the 25-position.5 That is, the derivatives of mil-
bemycin A3 (C-25 methyl), milbemycin A4 (C-25 ethyl), and
ivermectin (C-25 s-butyl) were synthesized and their activities
were compared.5 In the course of this study, it was necessary
for us to establish a method to synthesize 13β-substituted-13-
deoxy-ivermectin aglycon from ivermectin aglycon, because it
is known that the 13α-substituted derivative has less activity
than the 13β-substituted derivative.4 There seemed to be no
difficulty in this synthesis according to these reasons: 1) We
Results
Reaction at Room Temperature. First of all, a reaction
of ivermectin aglycon (5) and carboxylic acid 6 at room tem-
perature was examined. When the reaction took place at room
temperature, the products were a mixture of three derivatives:
2, 3, and 4 (Fig. 3). The ratio of these three derivatives was
2:3:4 = 8:14:17 according to the analysis of the peak areas
by HPLC, showing that the desired product was obtained at
very low yield.
Time Course Study of the Reaction. Then, to maximize
the ratio of the desired product, time course analysis of the ra-
tio of each compound was carried out. At the time points of 0,
5, 10, 30, 60, 120, and 180 min, a part of the reaction mixture
was extracted and the ratio of the three compounds (2, 3, and
Fig. 3. Reaction of ivermectin aglycon. a) CF3SO3H/CH2Cl2.

Y. Sugiyama et al.
Bull. Chem. Soc. Jpn., 75, No. 11 (2002) 2511
rivative 3 was suggested, 13β-14Z derivative 4 itself was puri-
fied and examined to see whether the transformation would
take place (Fig. 6). As we expected, both 2 and 3 were ob-
tained almost instantly when 4 was subject to the reaction con-
ditions (Fig. 7). Thus, this result strongly supports that there
must be conversion of 4 into 3.
α
Conversion of the 13 -14E Derivative to Other Deriva-
tives. Regarding the results in the reactions above, also the
transformation from 13α-14E derivative 3 to other products (2
and 4) was suggested. So 13α-14E derivative 3 itself was iso-
lated and examined to see whether it produces those com-
pounds under the same reaction conditions as mentioned above
(Fig. 6). The result showed that transformation occurred from
the 13α-14E derivative 3 to products 2 and 4, although the ra-
tio of the transformation was lower than that from 13β-14Z de-
rivative 4 (Fig. 8).
Fig. 4. Relative ratios of 2, 3, and 4 under room temperature.
Conversion of the 14Z Derivative in the Presence of Car-
boxylic Acid 7. As those conversions were presumed to be
via cation(s), the reaction of 13β-14Z derivative 4 with carbox-
ylic acid 7 was examined to verify that it was due to intermo-
lecular reaction, not intramolecular reaction (Fig. 6). As the α-
position of the carbonyl group in 7 is less hindered than that in
6, it was predicted that exchange of the acyloxy group at the
13-position would take place readily. To our satisfaction, ex-
change certainly took place to produce 8 and 9 as the only
products; 2, 3, 4, and 10 were not detected (Fig. 6).
4) was analyzed by HPLC (Fig. 4). Unsatisfactorily, the de-
sired product 2 was again obtained in low yield. Here, the ra-
tio of 13β-14Z derivative 4 was inversely dependent on the re-
action time, thus transformation of 13β-14Z derivative 4 to
13α-14E derivative 3 was presumed.
Effect of the Reaction Temperature. If a transformation
from 4 to 3 occurs, the reaction temperature could have an ef-
fect on the ratio of 2, 3, and 4 because the transformation
should take place via cation(s). Also the ratio of the desired
product 2 was expected to rise when the reaction took place at
higher temperature because its structure is presumed to be
thermodynamically stable. Thus, the ivermectin aglycon (5)
and carboxylic acid 6 were mixed under the same reaction con-
ditions as above, except the temperature was 40 °C, and time
course analysis of the ratio of the three compounds was exam-
ined again. This time, at the time points of 5, 10, 20, 60, 120,
and 180 min, a part of the reaction mixture was extracted and
the ratio of the three compounds was analyzed by HPLC (Fig.
5). Compared to the result at room temperature, the ratio of
13β-14Z derivative 4 was apparently decreased, and the ratio
of 13α-14E derivative 3 was increased. Although the yield of
the desired compound was not satisfactory, the transformation
of 4 to 3 was enhanced here.
α
Reaction of 13 -Hydroxy Derivative with Carboxylic
Acid 7. To compare the result of the reaction of 4 with car-
boxylic acid 7, 13α-hydroxy derivative 5 was used as a sub-
strate in the reaction conditions with carboxylic acid 7. The
products were in the ratio of 55% 13α-14E derivative 9, 30%
13β-14Z derivative 8, and 10% 13β-14E derivative 10 accord-
ing to the analysis of the peak areas by HPLC.
Discussion
According to the results above, we gave up obtaining the
13β-substituted-13-deoxy ivermectin aglycon from ivermectin
aglycon at high yield. But in that series of the studies, we en-
countered a very interesting compound, the 13β-substituted-
14Z derivative. Thus we suggest some hypotheses about the
reaction mechanism to obtain the 14Z derivative 4 as follows.
The Possibility of the Conversion between E-Derivative
and Z-Derivative. It is very likely that there is a path, which
enables the conversion from E to Z, and the path is completely
different from the one that is in the reaction of 13β-hydroxy
milbemycin. If the reaction occurs in the same path via the al-
lyl cation as reported in our previous paper,6 the product
should be only 13β-substituted product regarding its steric hin-
drance.
Conversion of the 14Z Derivative to the 14E Derivative.
As transformation from 13β-14Z derivative 4 to 13α-14E de-
Possibility of a Homoallylic Cation. Similar rearrange-
ment at the 13-position was also examined by Merck research-
ers using the 13α-substituted derivative as a substrate.4 In their
report, the configuration at the 13-position was retained when
solvolysis took place on the 13α-substituted derivative, and a
homoallylic cation was suggested as an explanation of the re-
tention of the stereochemistry. But they did not mention any
unusual by-products such as the 13β-substituted-14Z deriva-
tive. Besides, the mechanism for obtaining these by-products
cannot be explained by a homoallylic cation.
Fig. 5. Relative ratios of 2, 3, and 4 at 40 °C.

2512 Bull. Chem. Soc. Jpn., 75, No. 11 (2002)
13-Substituted-14Z Ivermectin Derivatives
Fig. 6. Examined reactions between the products. a) CF3SO3H/CH2Cl2.
Fig. 7. Actual concentrations of 2, 3, and 4 and their total concentration in the reaction mixture. The 13β-substituted-15Z (4) deriv-
ative was used as a substrate. The reaction took place under room temperature.
Mechanism to Produce the Unique Configurations. Al-
E to Z as presumed above, we regard that it is the inflexibility
though it seems very difficult to change the configuration from
of the rigid 16-membered ring (Fig. 1) that causes a different

Y. Sugiyama et al.
Bull. Chem. Soc. Jpn., 75, No. 11 (2002) 2513
Fig. 8. Actual concentrations of 2, 3, and 4 and their total concentration in the reaction mixture. The 13α-substituted-15E (3) deriv-
ative was used as a substrate. The reaction took place under room temperature.
double bond, thus it would produce the allyl cation readily,
which configuration is almost the same configuration as the
13β-hydroxy derivative to produce the corresponding product
2. On the other hand, if the hydroxy group is in the α-position,
the situation is completely different from the one in the β-posi-
tion (Fig. 9). That is, the conversion of the configuration from
E to Z occurs supposedly in the following manner: First, the
hydroxy group at the 13α-position is protonated and eliminat-
ed (Fig. 9-ꢀ). The vacant orbital is almost vertical to the p or-
bital of the adjacent double bond, thus it is unable to produce
the allyl cation readily, so the C-13 changes its orbital from sp3
to sp2 (Fig. 9-ꢁ). Then, as the p orbital at C-13 is vertical to
the adjacent p orbital, a bond pair between C-13 and -14, and
C-15 and -16, rotates to bring about the conjugation between
the cation and the double bond (Fig. 9-ꢁ) to produce the allyl
cation (Fig. 9-ꢂ). Those two bonds are required for simulta-
neous rotation because the 16-membered ring is too rigid to al-
low rotation of the bonds. At this point, the α-side of C-13 is
less hindered than the β-side; thus the 13α-substituted product
should be produced here (path a), and maybe a small amount
of 13β-product is produced despite the steric hindrance (path
b). Second, the double bond migrates to the bond between C-
13 and C-14 (Fig. 9-ꢃ), then a bond pair between C-14 and
-15 and C-16 and -17 rotates to make a Z-like configuration
(Fig. 9-ꢃ to ꢄ). However, this configuration (ꢄ) is also very
unstable. Thus the bond pair, between C-13 and -14 and C-16
and -17 rotates again to counter the strain in the 16-membered
ring, resulting in production of the allyl cation (Fig. 9-ꢅ). At
this point, the β-side of C-13 is less hindered than the α-side,
allowing only the 13β-14Z-product (path c). The 15-substitut-
ed product could be obtained (path d) here, but it would readily
Fig. 9. Presumed reaction mechanism producing 15-Z deriv-
convert into the 13-subsituted product as we reported in our
previous paper.6
ative.
This mark shows the carbon, which is almost im-
possible to move because of the rigid 16-membered ring.
The allyl cation ꢂ is different from the allyl cation, which
is shown in Fig. 2.
Discussion on the Cations. According to the result of the
reaction between 4 and 7 (Fig. 6), the cation(s) are certainly
produced during the reaction, although the cation(s) them-
selves cannot be isolated or identified.
reaction from the one in the linear molecule. The hydroxy
group at the 13-position was protonated and eliminated to pro-
duce the cations. If the hydroxy group is in the β-position, the
vacant orbital would be parallel to the p orbital of the adjacent
Summary. In the course of our study, we encountered a
unique reaction producing a 14Z derivative that is usually not
observed. Although the exact configurations of the cations
were not verified, a plausible reaction mechanism to obtain the

2514 Bull. Chem. Soc. Jpn., 75, No. 11 (2002)
13-Substituted-14Z Ivermectin Derivatives
HPLC to identify the products 2 and 4.
14Z-ivermectin derivatives was suggested.
The Reaction of 4 with 7. Carboxylic acid 7 (451.8 mg, 2.49
mmol) was dissolved in dichloromethane (10 mL), trifluoro-
methanesulfonic acid (30 µL) was added, and the mixture was
stirred for 5 min at room temperature. Then a solution of 4 (500
mg, 0.62 mmol) in dichloromethane (3 mL) was slowly added and
the mixture was stirred at room temperature for 40 min. The reac-
tion mixture was diluted with a mixture of ethyl acetate and 4%
aqueous solution of NaHCO3 to quench the reaction, extracted
with ethyl acetate, washed with a 10% aqueous solution of
NaHCO3, and with water, dried over Na2SO4, and evaporated in
vacuo. The residue was chromatographed on ODS with eluent
(90% acetonitrile) to obtain a mixture of 8 (122.6 mg, 25.8%
yield) and 9 (128.4 mg, 27.4% yield). The compounds 8 and 9
Experimental
The Reaction of 5 with 6. Carboxylic acid 6 (160.9 mg, 0.68
mmol) was dissolved in dichloromethane (3.0 mL), trifluoro-
methanesulfonic acid (10 µL) was added, and the mixture was
stirred for 5 min at room temperature. Then a solution of 5 (100
mg, 0.17 mmol) in dichloromethane (2 mL) was slowly added and
the mixture was stirred at room temperature for 40 min. The reac-
tion mixture was diluted with a mixture of ethyl acetate and 4%
aqueous solution of NaHCO3 to quench the reaction, extracted
with dichloromethane, washed with a 10% aqueous solution of
NaHCO3 and with water, dried over Na2SO4, and evaporated in
vacuo. The residue was chromatographed on ODS with eluent
(100% acetonitrile) to obtain a mixture of 2, 3, and 4. The ratio of
2, 3, and 4 was estimated at 2:3:4 = 8:17:14 by analyzing the
area ratio of the HPLC spectrum. The mixture was chromato-
graphed on ODS with eluent (100% acetonitrile), again to isolate
the pure part of each of the compounds. The purified compounds
2, 3, and 4 were analyzed by NMR to identify their structure. 2; 1H
NMR δ 0.77 (3H, d, J = 5.9 Hz, C-24 CH3), 0.84 (3H, d, J = 6.6
Hz, C-12 CH3), 0.94 (3H, t, J = 7.7 Hz, C-28 H), 1.88 (3H, s, C-4
CH3), 3.14 (1H, m, C-25 H), 3.53 (1H, m, C-2 H), 3.57 (1H, m, C-
17 H), 3.83 (1H, s, C-6 H), 3.96 (1H, broad s, C-7 OH), 4.69 and
4.73 (2H, ABq, J = 14.4 Hz, C-8 CH2), 4.91 (1H, d, J = 10.2 Hz,
C-13 H), 5.77 (1H, dd, J = 14.3 and 11.7 Hz, C-10 H), 5.85 (1H,
dt, J = 11.7 and 2.4 Hz, C-9 H), 6.55 (1H, s, C-3 H), 7.50 (2H, d,
J = 8.8 Hz, Ph H), 8.16 (2H, d, J = 8.8 Hz, Ph H). 3; 1H NMR δ
0.88 (3H, d, J = 6.6 Hz, C-12 CH3), 1.01 (3H, t, J = 7.3 Hz, C-28
H), 1.89 (3H, s, C-4 CH3), 3.16 (1H, m, C-25 H), 3.84 (1H, s, C-6
H), 3.94 (1H, broad s, C-7 OH), 4.69 and 4.74 (2H, ABq, J = 14.5
Hz, C-8 CH2), 4.78 (1H, m, C-15 H), 5.05 (1H, m, C-13 H), 5.39
(1H, m, C-19 H), 5.47 (1H, dd, J = 14.7 and 10.2 Hz, C-11 H),
5.71 (1H, dd, J = 14.7 and 11.0 Hz, C-10 H), 5.88 (1H, dt, J =
11.0 and 2.2 Hz, C-9 H), 6.58 (1H, s, C-3 H), 7.59 (2H, d, J = 8.8
1
were analyzed by NMR to identify their structure. 8; H NMR δ
0.82 (3H, d, J = 6.7 Hz, C-12 CH3), 0.90 (3H, t, J = 7.4 Hz, C-28
H), 1.89 (3H, s, C-4 CH3), 3.10 (1H, m, C-25 H), 3.55 (1H, m, C-
2 H), 3.74 (2H, s, benzyl H), 3.83 (1H, m, C-17 H), 3.86 (1H, s, C-
6 H), 4.65 (1H, broad s, C-7 OH), 4.71 and 4.77 (2H, ABq, J =
14.7 Hz, C-8 CH2), 5.36 (1H, d, J = 10.9 Hz, C-13 H), 5.47 (1H,
m, C-5 H), 6.51 (1H, s, C-3 H), 7.45 (2H, d, J = 8.7 Hz, Ph H),
8.20 (2H, d, J = 8.7 Hz, Ph H). 9; 1H NMR δ 0.88 (3H, d, J = 6.6
Hz, C-12 CH3), 0.98 (3H, t, J = 7.3 Hz, C-28 H), 1.90 (3H, s, C-4
CH3), 3.19 (1H, m, C-25 H), 3.57 (1H, m, C-2 H), 3.63 (1H, m, C-
17 H), 3.86 (1H, s, C-6 H), 3.87 and 3.83 (2H, ABq, J = 15.4 Hz,
benzyl H), 4.72 and 4.76 (2H, ABq, J = 14.5 Hz, C-8 CH2), 4.84
(1H, m, C-15 H), 5.17 (1H, m, C-13 H), 5.36–5.44 (1H, m, C-19
H), 5.64 (1H, dd, J = 14.8 and 10.0 Hz, C-11 H), 5.78 (1H, dd, J
= 14.8 and 11.1 Hz, C-10 H), 5.91 (1H, dt, J = 11.1 and 2.3 Hz,
C-9 H), 6.58 (1H, m, C-3 H), 7.51 (2H, d, J = 8.7 Hz, Ph H), 8.23
(2H, d, J = 8.7 Hz, Ph H).
The Reaction of 5 with 7. Carboxylic acid 7 (144.9 mg, 0.8
mmol) was dissolved in dichloromethane (3.0 mL), trifluoro-
methanesulfonic acid (10 µL) was added, and the mixture was
stirred for 5 min at room temperature. Then a solution of 5 (116.9
mg, 0.2 mmol) in dichloromethane (2 mL) was slowly added and
the mixture was stirred at room temperature for 60 min. A part of
the reaction mixture was extracted, diluted with a mixture of ethyl
acetate and 4% aqueous solution of NaHCO3, and analyzed by
HPLC to identify the products 8, 9, and 10. The structure of 10
was compared to the authentic sample to identify its structure. 10;
1H NMR δ 1.89 (3H, s, C-4 CH3), 3.14 (1H, m, C-25 H), 3.55 (1H,
m, C-2 H), 3.60 (1H, m, C-17 H), 4.72 and 4.76 (2H, ABq, J =
14.3 Hz, C-8 CH2), 4.98 (1H, d, J = 10.9 Hz, C-13 H), 6.57 (1H,
m, C-3 H).
1
Hz, Ph H), 8.19 (2H, d, J = 8.8 Hz, Ph H). 4; H NMR δ 0.82
(3H, d, J = 6.5 Hz, C-12 CH3), 0.90 (3H, t, J = 7.3 Hz, C-28 H),
1.88 (3H, s, C-4 CH3), 3.11 (1H, m, C-25 H), 3.53 (1H, m, C-2 H),
3.82 (1H, m, C-17 H), 3.84 (1H, s, C-6 H), 4.63 (1H, broad s, C-7
OH), 4.68 and 4.74 (2H, ABq, J = 14.6 Hz, C-8 CH2), 5.23 (1H,
d, J = 11.0 Hz, C-13 H), 5.25 (1H, dd, J = 13.6 and 10.6 Hz, C-
11 H), 5.35 (1H, m, C-19 H), 5.41 (1H, m, C-15 H), 6.50 (1H, s,
C-3 H), 7.50 (2H, d, J = 8.8 Hz, Ph H), 8.17 (2H, d, J = 8.8 Hz,
Ph H).
The Reaction of 4 to 3. Carboxylic acid 6 (29.3 mg, 0.13
mmol) was dissolved in dichloromethane (3.0 mL), trifluoro-
methanesulfonic acid (5 µL) was added, and the mixture was
stirred for 5 min at room temperature. Then a solution of 4 (25.0
mg, 0.03 mmol) in dichloromethane (1 mL) was slowly added and
the mixture was stirred at 40 °C for 120 min. A part of the reac-
tion mixture was extracted, diluted with a mixture of ethyl acetate
and 4% aqueous solution of NaHCO3, and analyzed by HPLC to
identify the product 3.
We would like to thank Dr. Shuichi Miyamoto and Yoriko
Iwata for helpful discussions during the course of this work.
References
1
H. Mishima, J. Ide, S. Muramatsu, and M. Ono, J. Anti-
biot., 36, 980 (1983).
2
Y. Takiguchi, H. Mishima, M. Okuda, M. Terao, A. Aoki,
The Reaction of 3 to 2 and 4. Carboxylic acid 6 (23.5 mg,
0.10 mmol) was dissolved in dichloromethane (3.0 mL), trifluo-
romethanesulfonic acid (5 µL) was added, and the mixture was
stirred for 5 min at room temperature. Then a solution of 3 (20.0
mg, 0.025 mmol) in dichloromethane (1 mL) was slowly added
and the mixture was stirred at 40 °C for 120 min. A part of the re-
action mixture was extracted, diluted with the mixture of ethyl ac-
etate and 4% aqueous solution of NaHCO3, and analyzed by
and R. Fukuda, J. Antibiot., 33, 1120 (1980).
3
M. H. Fisher and H. Mrozik, “Macrolide Antibiot.,” ed by
S. Omura, Academic Press (1984), p. 553.
H. Mrozik, B. O. Linn, P. Escola, A. Lusi, A. Matzuk, F. A.
4
Preiser, D. A. Ostlind, J. M. Schaeffer, and M. H. Fisher, J. Med.
Chem., 32, 375 (1989).
5
In preparation.

Y. Sugiyama et al.
Bull. Chem. Soc. Jpn., 75, No. 11 (2002) 2515
6
(2001).
7
Y. Sugiyama and A. Saito, Bull. Chem. Soc. Jpn., 74, 1319
obtained in their reaction.
Y. Tsukamoto, K. Sato, T. Kinoto, and T. Yanai, Bull.
Chem. Soc. Jpn., 65, 3300 (1992).
B. Frei, P. Huxley, P. Maienfisch, H. B. Mereyala, G. Rist,
and A. C. O’Sullivan, Helv. Chim. Acta, 73, 1905 (1990).
8
The 14Z configuration has already been proved to be
produced during the reaction of 15-hydroxymilbemycin by
9
O’Sullivan,9 although the 13-substituted-14Z derivative was not