Identification and characterization of pectenotoxin (PTX) 4 and PTX7 as spiroketal stereoisomers of two previously reported pectenotoxins

Article abstract of DOI:10.1021/jo971310b

Pectenotoxins (PTXs) isolated from the scallon Patinopecten yessoensis were shown to be involved in an episode of diarrhetic shellfish poisoning (DSP). A total of eight analogues (PTX1-PTX7 and PTX10) have been isolated to date, and the structures of four of these analogues (PTX1, PTX2, PTX3, and PTX6) have already been elucidated. Here, we report the characterization of PTX4 and PTX7 as 7-epi-PTX1 and 7-epi-PTX6, respectively, on the basis of NMR data and an acid-catalyzed chemical interconversion. The structures of two new artifacts, PTX8 and PTX9, produced following this treatment are also reported.

Full text of DOI:10.1021/jo971310b

J . Org. Chem. 1998, 63, 2475-2480
2475
Id en tifica tion a n d Ch a r a cter iza tion of P ecten otoxin (P TX) 4 a n d
P TX7 a s Sp ir ok eta l Ster eoisom er s of Tw o P r eviou sly Rep or ted
P ecten otoxin s
Katsunori Sasaki, J effrey L. C. Wright, and Takeshi Yasumoto*
Faculty of Agriculture, Tohoku University, Tsutsumidori-Amamiya, Aoba-ku, Sendai 981, J apan;
Institute for Marie Biosciences, National Research Council of Canada, 1411 Oxford Street, Halifax,
Nova Scotia, Canada B3H 3Z1
Received J uly 17, 1997 (Revised Manuscript Received December 3, 1997 )
Pectenotoxins (PTXs) isolated from the scallop Patinopecten yessoensis were shown to be involved
in an episode of diarrhetic shellfish poisoning (DSP). A total of eight analogues (PTX1-PTX7 and
PTX10) have been isolated to date, and the structures of four of these analogues (PTX1, PTX2,
PTX3, and PTX6) have already been elucidated. Here, we report the characterization of PTX4
and PTX7 as 7-epi-PTX1 and 7-epi-PTX6, respectively, on the basis of NMR data and an acid-
catalyzed chemical interconversion. The structures of two new artifacts, PTX8 and PTX9, produced
following this treatment are also reported.
In tr od u ction
The pectenotoxins (PTXs) (Figure 1) are a family of
cyclic polyether macrolide toxins that have been found
in the digestive glands of toxic scallops1-4 together with
okadaic acid (OA) and dinophysistoxin-1 (DTX1), the
principal toxins of the diarrhetic shellfish poisoning
(DSP).4 PTX2 (2) induces diarrhea and liver damage.5
Whereas OA and its analogues are powerful phosphatase
inhibitors,4,6 the PTXs do not possess this activity, and
their mode of action still remains to be determined.
Nevertheless, the poisoning associated with these toxins,
and possibly others, is spread worldwide, posing a serious
threat to public health and to the aquaculture industry.1
To date, a total of eight PTX analogues (PTX1-PTX7
and PTX10) have been isolated from toxic extracts of the
F igu r e 1. The structures of pectenotoxins (PTXs). To avoid
the interconversion arising from the presence of a carboxyl
group at C-43, all NMR measurements of the PTX6 (6), PTX7
(7), and PTX9 (9) congeners were carried out using the
corresponding phenacyl ester derivative.
scallop Patinopecten yessoensis.1-4 Five of these (PTX1-
PTX5) were obtained from neutral fractions of an Al2O3
column separation, and three others (PTX6, PTX7, and
PTX10) were found in acidic fractions.3 The compound
PTX1 (1) was the first of the pectenotoxin family to be
characterized, and its structure was secured by X-ray
crystallography (Figure 1).1 Later, the structures of three
other analogues 2, 3, and 6 (PTX2, PTX3, and PTX6)
were determined by spectroscopic methods and compari-
son with 1.1-3 The combined data indicated that the
structural variation between these analogues resides in
the nature of the substituent at C-18, where all stages
of oxidation from methyl to carboxyl are found: PTX2:
CH3, PTX1: CH2OH, PTX3: CHO, and PTX6: COOH
(Figure 1).4 Recently, the absolute configuration of PTX6
has been determined by NMR spectroscopy using phe-
nylglycine methyl ester (PGME) as a chiral anisotropic
reagent.7 As is often the case with families of shellfish
toxins, the compounds are not produced by the shellfish
themselves, but rather are products of toxigenic microal-
gae or phytoplankton ingested by the shellfish. Various
dinoflagellate species of the genera Dinophysis and
Prorocentrum are known to produce a variety of polyether
toxins.3,4,8 Strains of the dinoflagellate Dinophysis fortii
which produce OA and/or DTX1 were also found to
produce PTX2 (2) in J apan9 and Europe.10 Interestingly,
2 was the only PTX congener found in both studies,
leading to the proposal that 2 is the parent toxin which
(1) Yasumoto, T.; Murata , M.; Oshima, Y.; Sano, M.; Matsumoto,
G. K.; Clardy, J . Tetrahedron 1985, 41, 1019.
(2) Murata, M.; Sano, M.; Iwashita, T.; Naoki, H.; Yasumoto, T.
Agric. Biol. Chem. 1986, 50, 2693.
(3) Yasumoto, T.; Murata, M.; Lee, J .-S.; Torigoe, K. in Bioactive
Molecules: Mycotoxins and Phycotoxins ’88; Natori, S., Hashimoto, K.,
Ueno, Y., Eds.; Elsevier: New York, 1989; Vol 10, p 375.
(4) Yasumoto, T.; Murata, M. Chem. Rev. 1993, 93, 1897.
(5) Ishige, M.; Satoh, N.; Yasumoto, T. Reps. Hokkaido Inst. Health,
1988, 38, 15.
(6) (a) Takai, A.; Bialojan, C.; Troschka M.; Ru¨egg, J . C. FEBS Lett.
1987, 217, 81. (b) Cohen, P.; Holmes, C. F. B.; Tsukitani, Y. Trends
Biochem. Sci. 1990, 15, 98.
(7) Sasaki, K.; Satake, M.; Yasumoto, T. Biosci. Biotech. Biochem.
1997, 61, 1783.
(8) (a) Shimizu, Y. Chem. Rev. 1993, 93, 1685. (b) Garson, M. J .
ibid. 1699.
(9) Lee, J .-S.; Igarashi, T.; Fraga, S.; Dahl, E.; Hovgaard, P.;
Yasumoto, T. J . Appl. Phycol. 1989, 1, 147.
(10) Draisci, R.; Lucentini, L.; Gianetti, L.; Boria, P.; Poletti, R.
Toxicon 1996, 34, 923.
S0022-3263(97)01310-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 03/28/1998

2476 J . Org. Chem., Vol. 63, No. 8, 1998
Ta ble 1. 1H a n d 13C Ch em ica l Sh ifts of th e P h en a cyl Ester s of P TX6, P TX7, a n d P TX9a
Sasaki et al.
PTX6
PTX7
PTX9
position
1H
2.61
1.27
3.58
1.24(ax)
1.35(ax)
1.78(ax)
1H′
13C
1H
2.68
1.30
3.97
1.43(ax)
1.92(ax)
1.67(ax)
1H′
13C
1H
2.84
1.11
3.65
1.33(ax)
1.69(ax)
1.41(ax)
1H′
13C
1
2
172.6
48.6
15.1
76.2
29.9
22.0
34.4
107.5
32.6
22.3
81.1
173.1
47.8
16.1
72.5
29.5
20.6
33.7
106.2
39.8
25.4
79.7
170.7
45.7
14.2
73.4
26.0
19.6
34.9
96.6
34.7
27.6
67.4
2-Me(41)
3
4
5
6
7
8
9
1.60(eq)
1.60(eq)
1.62(eq)
1.92(eq)
1.55(eq)
1.79(eq)
1.66(eq)
1.49(eq)
1.53(eq)
1.32
1.49
4.35
2.20
2.09
2.10
2.31
4.92
2.16
2.72
1.51(ax)
2.30(ax)
3.99
1.84(eq)
1.99(eq)
10
10-OH
11
11-OH
12
-
4.39
4.22
-
4.28
6.72
4.61
3.82
-
74.9
77.6
79.2
81.8
23.1
44.5
213.7
79.2
70.6
33.1
82.5
31.5
27.5
110.3
79.2
29.8
37.3
84.5
26.6
50.6
30.4
23.6
140.2
130.5
12.5
135.0
121.7
83.1
75.2
34.1
82.7
98.6
82.3
22.4
49.3
213.5
80.3
70.2
33.3
82.8
31.7
29.6
109.8
79.9
28.5
37.8
84.1
26.8
49.5
30.0
23.5
140.3
130.8
12.7
134.6
122.1
82.9
75.8
34.1
82.5
98.6
83.6
19.9
51.6
212.0
79.8
70.4
32.6
82.7
31.7
27.3
110.5
78.8
30.2
35.5
84.8
26.7
50.4
30.2
23.6
139.6
130.8
13.0
134.5
123.0
82.8
75.4
34.1
82.7
98.6
12-Me(42)
13
14
15
16
17
18
19
20
21
22
23
24
25
25-Me(44)
26
27
27-Me(45)
28
29
29-Me(46)
30
31
32
33
34
35
36
1.23
2.17
1.50
2.48
1.40
2.65
3.33
3.12
2.78
4.06
4.60
3.06(ax)
4.13
4.54
3.11(ax)
4.14
4.48
3.04(ax)
2.01(eq)
2.01(eq)
2.01(eq)
2.25
1.95
2.56
2.48
2.29
2.04
2.59
2.41
2.27
1.88
2.59
2.49
4.37
1.85
1.41
4.25
1.95
1.47
4.24
1.73
1.28
2.33
1.58
2.01
1.54
1.94
1.64
1.21
1.57
2.49
0.92
5.09
1.32
1.42
2.64
0.92
5.06
1.14
1.55
2.43
0.92
5.09
1.75
1.57
1.69
1.59
6.69
5.50
4.98
5.78
2.57
5.30
1.70
6.63
5.56
4.97
5.75
2.60
5.20
1.67
6.63
5.51
4.94
5.66
2.60
5.28
2.72
2.70
2.68
36-OH
7.09
7.04
7.06
37
3.86
71.2
3.82
71.0
3.84
71.1
37-OH
5.96
5.91
5.95
38
2.54
1.15
30.2
18.2
2.52
1.15
30.3
18.2
2.54
1.18
30.2
18.3
38-Me(47)
39
40
43
1.99(ax)
4.30(aq)
1.26(eq)
3.84(eq)
28.0
60.9
171.4
1.97(ax)
4.28(aq)
1.25(eq)
3.82(eq)
28.1
60.9
171.4
2.02(ax)
4.29(ax)
1.29(eq)
3.86(eq)
28.1
60.9
171.3
a
1H NMR chemical shifts measured in pyridine-d5 (ref δ 7.211). 13C NMR chemical sifts measured in pyridine-d5 (ref δ 123.5).
is metabolized in scallops to yield the congeners 1, 3, and
6.3
Resu lts
The PTX6/7 pair of congeners shared a common mo-
lecular ion in FABMS [m/z 887 (M - H)-], as did the
PTX1/4 pair [m/z 875 (M + H)+]. In addition, the LSI-
MS/MS spectra of the (M + Na)+ ion for PTX6 (6) and
PTX7 (7) were essentially identical, and a similar result
was observed for the MS/MS fragmentation spectra of
the (M + Na)+ ion of PTX1 (1) and PTX4 (4). Thus the
MS data indicated that PTX6 and PTX7 are isomeric, as
are PTX1 and PTX4. However, inspection of the NMR
data further extended the relationship between the two
pairs. The 1H and 13C NMR data for the PTX6/7
congeners and PTX1/4 congeners are shown in Table 1
and Table 2, respectively. Upon comparison of the data
for PTX7 (7) with that for PTX4 (4), and similarly from
previous comparisons between PTX6 (6) with PTX1 (1),3
it was clear in each case that these pairs of molecules
Although the structure and origin of PTX toxins 1-3
and 6 can be accounted for in this way, the situation with
the remaining four congeners, PTX4, PTX5, PTX7, and
PTX10, is more complex, and their structures have not
been determined as yet, primarily due to the small
amounts of the compounds currently available. However,
as the structural information of all the toxic constituents
found in the scallops is essential for designing proper
analytical methods and for understanding toxin metabo-
lism in shellfish, we renewed our effort to isolate and
determine the structures of these uncharacterized ana-
logues. In this paper we report the structural elucidation
of PTX4 (4) and PTX7 (7), as well as the isolation and
characterization of two new artifacts, PTX8 (8) and PTX9
(9).

Spiroketal Stereoisomers of Pectenotoxins
J . Org. Chem., Vol. 63, No. 8, 1998 2477
Ta ble 2. 1H a n d 13C Ch em ica l Sh ifts of P TX1, P TX4, a n d P TX8a
PTX1
PTX4
PTX8
position
1H
2.63
1H′
13C
1H
2.68
1H′
13C
1H
2.83
1H′
13C
1
2
172.6
48.6
15.2
76.2
29.9
22.0
34.4
107.5
32.6
22.3
81.0
75.1
81.7
23.2
44.5
213.8
79.7
71.1
32.1
84.0
29.3
27.9
109.6
79.6
29.8
37.4
84.2
26.7
50.9
30.4
23.6
140.2
130.5
12.5
135.1
121.7
83.0
75.1
34.1
82.6
98.6
71.1
30.2
18.3
28.0
60.9
67.3
173.1
47.8
16.2
72.5
29.4
20.6
33.7
106.3
39.8
25.5
79.8
77.6
82.1
22.6
49.2
213.7
80.9
70.8
32.5
84.2
29.4
30.1
109.1
80.1
28.6
37.5
83.9
26.9
49.7
30.1
23.5
140.3
130.8
12.6
134.9
122.1
82.9
75.7
34.1
82.5
98.6
71.0
30.2
18.3
28.0
60.9
67.3
170.7
45.8
14.2
73.4
26.1
19.6
34.9
96.5
34.6
27.5
67.4
79.2
83.5
20.0
51.7
212.1
80.4
70.9
31.6
84.2
29.4
27.8
109.7
79.2
30.3
35.5
84.5
26.8
50.6
30.2
23.6
139.6
130.8
13.0
134.9
123.1
82.8
75.4
34.1
82.7
98.6
71.1
30.2
18.3
28.1
60.9
67.1
2-Me(41)
1.30
3.58
1.22(ax)
1.34(ax)
1.78(ax)
1.30
3.98
1.42(ax)
1.92(ax)
1.67(ax)
1.08
3.63
1.30(ax)
1.69(ax)
1.38(ax)
3
4
5
6
7
8
9
10
11
12
1.61(eq)
1.65(eq)
1.60(eq)
1.90(eq)
1.54(eq)
1.80(eq)
1.63(eq)
1.48(eq)
1.53(eq)
1.29
1.47
4.33
4.39
2.19
2.13
2.08
2.33
4.94
4.25
2.18
2.77
1.48(ax)
2.30(ax)
3.97
1.81(eq)
1.98(eq)
3.79
12-Me(42)
13
14
15
16
17
18
19
20
21
22
23
24
25
25-Me(44)
26
27
27-Me(45)
28
29
29-Me(46)
30
31
32
33
34
35
36
1.22
2.16
1.48
2.45
1.36
2.64
3.34
3.12
2.73
4.07
4.53
2.78(ax)
4.12
4.46
2.80(ax)
4.10
4.34
2.71(ax)
1.46(eq)
1.46(eq)
1.44(eq)
1.78
1.93
2.12
2.43
1.81
2.01
2.12
2.35
1.78
1.86
2.11
2.40
4.35
1.86
1.42
4.22
1.93
1.49
4.20
1.72
1.28
2.37
1.61
2.05
1.52
1.93
1.63
1.22
1.64
2.51
0.93
5.09
1.31
1.50
2.63
0.92
5.10
1.10
1.59
2.44
0.91
5.11
1.81
1.64
1.73
1.60
6.69
5.50
4.97
5.79
2.57
5.31
1.70
6.65
5.55
4.97
5.77
2.58
5.21
1.67
6.64
5.50
4.92
5.66
2.58
5.28
2.73
2.72
2.68
37
38
3.86
2.56
1.15
2.00(ax)
4.31(aq)
3.85
3.83
2.53
1.14
1.97(ax)
4.28(aq)
3.85
3.85
2.54
1.15
2.02(ax)
4.28(ax)
3.83
38-Me(47)
39
40
43
1.27(eq)
3.86(eq)
3.93
1.25(eq)
3.82(eq)
3.93
1.25(eq)
3.83(eq)
3.91
a
1H NMR chemical shifts measured in pyridine-d5 (ref δ 7.211). 13C NMR chemical shifts measured in pyridine-d5 (ref δ 123.5). OH
signals were not observed because H2O signal was irradiated.
shared a common carbon skeleton, differing only in the
nature of the substituent at C-18: In 4 (and 1) it is
hydroxymethyl, while in 7 (and 6) it is carboxyl,3 leading
to the conclusion that 4 and 7 are stereoisomers of 1 and
6, respectively.
Acid-Catalyzed In ter con ver sion of th e P TXs. Upon
standing in aqueous acetonitrile solution, PTX7 (7)
underwent gradual transformation to PTX6 (6) whereas
PTX4 (4) remained unchanged under the same condi-
tions. Since 7 contains a carboxyl group and 4 does not,
it was suspected that the rearrangement was acid
catalyzed. Indeed, addition of TFA (0.1% v/v) to an
F igu r e 2. Acid-catalyzed Interconversion of the PTXs.
aqueous acetonitrile solution of 7 not only accelerated the
conversion of 7 to 6, but also resulted in formation of an
additional isomeric product coded PTX9 (9) (Figures 1
and 2). In addition, 6, which is stable during the
purification and isolation process, also underwent equili-
bration to 7 and 9 following similar treatment with TFA
(Figures 1 and 2). The proportions of isomers at equili-
bration (after 48 h) were PTX6:PTX7:PTX9 (40:16:44).
Under the same acid conditions, a similar equilibration
was induced between the previously stable derivatives
PTX1 (1) and PTX4 (4), together with formation of a new
compound PTX8 (8) (Figure 2). Thus treatment of 1 with
TFA yielded the equilibrium mixture of PTX1:PTX4:
PTX8 (29:14:57). As further evidence for the role of acid
in the rearrangement process, the phenacyl ester of PTX7
(7), which is normally stable upon standing under neutral

2478 J . Org. Chem., Vol. 63, No. 8, 1998
Sasaki et al.
F igu r e 3. Equivalent portions of the PTX1/6 and PTX4/7 molecules, showing the NOEs around the spiroketal rings A and B.
conditions and does not undergo spontaneous rearrange-
ment, rapidly gave the equilibration mixture PTX6 ester:
PTX7 ester:PTX9 ester (30:9:61) following exposure to
TFA.
differences observed for a few other resonances elsewhere
in the molecule (e.g. H-22, H-23, H-26, and H-27).
In PTX6 (6) and PTX7 (7), coupling patterns of H-3
(ddd, 11, 11, and 2 Hz), were similar and consistent with
their axial configuration, indicating that the conformation
of the tetrahydropyran ring A in PTX7 (7) is not flipped.
From previous studies and based on the X-ray crystal-
lographic data for PTX1 (1),1 it was concluded that ring
A in PTX6 (6) contained an equatorial oxygen at C-7. This
is reflected in the ROESY data which displayed NOE
correlations from H-8 to H-3 as well as to H-5 in 6 (Figure
3). In contrast, these key NOEs were not observed in
PTX7 (7), and instead NOEs from H-10 to H-3 as well as
to H-41 (Me) were noted (Figure 3). Since the NOE
correlations from H-3 to H-4 and H-5 are still observed
in 7, it was determined that the oxygen substituent in 7
is axial. Other NOE correlations in ring B (e.g. H-8 to
H-9 and H-9 to H-10) supported this assignment and
hence PTX7 (7) was identified as the C-7 epimer of 6
(Figure 1). In support of this, the resonances for H-3,
H-4eq, and H-5ax in PTX7 (7) were shifted by ∆ +0.39,
+0.32, and +0.57 ppm, respectively, compared with those
of PTX6 (6), consistent with the effect of an axial oxygen
substituent on ring A in 7.12
Th e Str u ctu r e of P TX7 (7). To avoid any intercon-
version catalyzed by the C-18 carboxyl group, all NMR
measurements of the PTX6 and PTX7 congeners were
carried out using the corresponding phenacyl ester
derivatives. As a prelude to the structural assignment
of the phenacyl ester of PTX7, it was necessary to
completely assign the 1H and 13C signals for the phenacyl
ester of the characterized PTX6 isomer using conven-
tional 2D NMR experiments. These data were then used
in the assignment of the PTX7 resonances and the data
for both compounds are shown in Table 1. On the basis
of the NMR connectivity data it was concluded that PTX7
(7) possessed the same planar structure as PTX6 (6). For
example, the 1H-1H COSY spectra for 7 and 6 both
displayed similar patterns of cross-peaks, although some
differences in chemical shifts were observed. The con-
nectivities around the quaternary carbons established by
the HMBC data were also identical between compounds
though the single correlation H-9/C-7 observed in 6 was
absent in the HMBC spectrum of 7. In the latter case,
C-7 was assigned by default as the last remaining
unassigned spiroketal carbon. Further careful compari-
Th e Str u ctu r e of P TX9 (9). The negative ion FABMS
data for PTX9 [m/z: 887 (M - H)-] revealed that the
compound was isomeric with PTX6 (6) and PTX7 (7).
Once again, in the structural determination work, all
NMR measurements were performed with the phenacyl
esters in order to compare and contrast with the corre-
sponding NMR data of 6 and 7, as well as prevent the
possibility of any spontaneous interconversion. The
assignments for PTX9 (9) shown in Table 1, based on the
usual series of 1D and 2D spectra, show chemical shift
differences for the resonances H-2 through H-13 com-
pared with the corresponding resonances in 6 and 7. In
particular, the greatest differences are observed for the
protons H-6 through H-11, though paradoxically the two
spin systems H-41 (Me) through H-6 and H-8 through
1
son of the NMR data of 6 and 7 revealed a series of H
chemical shift differences (in the range ∆ 0.2-0.7 ppm)
between the corresponding protons H-3 to H-13 in the
region around rings A-C. The most significant differ-
ences were observed for H-3, and H-8 through H-10,
though smaller differences were also observed for H-22,
H-23, H-26, and H-27. In each isomer, the corresponding
carbons in the region C-3 through C-15 also displayed
chemical shift differences, some quite significant (e.g. C-3,
C-8, C-9, C-11, and C-13), whereas there was little
difference in the 13C resonances for carbons in the region
C-22 through C-27. Together, these data suggested that
the structural difference between the two isomers was
in the region around rings A and B which form a spiro-
[4.5] system. Since it is known that acid treatment of
spiroketal systems can lead to epimerization at the
spiroketal carbon,11 it was suspected that acid-catalyzed
ring opening and reclosure of this system might lead to
epimerization at C-7. Such a stereochemical change in
this region of the PTX molecule would be expected to
result in conformational changes to the entire macrocyclic
lactone ring and would explain the minor chemical shift
1
H-11 were still present. Whereas the H NMR spectra
of 6 and 7 both displayed resonances for all the exchange-
able hydroxyl protons (HO-11, HO-36, and HO-37), only
the resonances for two hydroxyl signals (HO-36, HO-37)
could be accounted for in the spectra of 9. Instead, a new
hydroxyl signal (4.61 ppm, broad singlet) was observed,
and this was assigned to position 10, based on the
correlations appearing in the 1H-1H COSY spectrum.
Thus the oxygen at C-11 in 9 must be part of an ether
link, arising through expansion of ring B to a tetrahy-
(11) For
a
detailed review of spiroketal chemistry: Perron, F.;
(12) Satake, M.; Ishibashi, Y.; Legrand, A. M.; Yasumoto, T. Biosci.
Biotech. Biochem. 1997, 60, 2103.
Albizati, K. F. Chem. Rev. 1989, 89, 1617.

Spiroketal Stereoisomers of Pectenotoxins
J . Org. Chem., Vol. 63, No. 8, 1998 2479
pairs led to the idea that PTX8 (8) possessed the same
skeleton as PTX9 (9). Once again this was supported by
the observation that the 1D and 2D NMR for 8 closely
resembled those of 9 (Table 2), and the NOE data as
shown in PTX9 (9) (Figure 4) confirmed the biaxial
orientation of C-O bonds in the spiro[5.5] system in
PTX8 (8) (Figure 1).
Discu ssion
To date, only PTX2 (2) has been found in extracts of
Dinophysis cells, and it has been proposed that the PTX
congeners 1, 3, and 6 are produced through oxidation of
the C-43 methyl group of 2 in the digestive gland of the
shellfish.3 Following the identification of PTX4 (4) and
PTX7 (7) as the 7-epi -derivatives of PTX1 and PTX6,
respectively, the question arises as to whether the former
pair of congeners are artifacts of the isolation procedure.
According to the fluorometric HPLC analytical data using
9-anthryldiazomethane (ADAM) as a fluorescence label-
ing reagent,14 the digestive glands of scallops collected
from Mutsu Bay, J apan, contained almost equal amounts
of PTX6 (5.7 µg/g digestive gland) and PTX7 (5.0 µg/g
digestive gland). Although acidic conditions do not
prevail during the normal recovery process, a recovery
test was performed using digestive glands of nontoxic
scallops spiked with PTX6 (57 µg/g) and no PTX7 was
detected. This indicates that PTX6 (6) is stable under
these conditions and is not transformed to PTX7 (7)
during the extraction and analysis procedure. Thus it
can be concluded that PTX7 (7) and hence PTX4 (4) are
not artifacts of the extraction process but occur naturally
in extracts of scallop digestive glands. Of further sig-
nificance is the fact that if these compounds were
artifacts, it might be expected that PTX8 (8) and PTX9
(9) would also be observed in the mixture, but careful
examination of the fluorometric HPLC data failed to show
any evidence of PTX9 in the shellfish extracts.
The occurrence of both epimers of a spiroketal ring
system in polyethers such as the pectenotoxin group, is
rare. Although a number of dinoflagellate polyether
metabolites containing spiroketal rings have been re-
ported,4,11 to our knowledge the only other precedent of
naturally occurring spiroketal epimers is the recently
identified epimers ciguatoxin (CTX) 4A and CTX4B from
Gambierdiscus toxicus.12 However, from a consideration
of the factors affecting the orientation of the spiro C-O
bonds, the occurrence of both spiroketal epimers among
the pectenotoxins may not be surprising. In tetrahydro-
pyran rings the equatorial preference for substituents is
important, but other factors including steric effects,
anomeric effects, intramolecular hydrogen bonding, and
other chelation effects influence this selection.11 From
the studies of many spiroketal systems by different
groups, it is generally accepted that in spiro[5.5]- and
spiro[4.5] systems, the biaxial arrangement of spiro C-O
bonds, maximizing the anomeric effect, is favored in
compounds containing saturated or unsaturated ether
ring systems. Indeed, a biaxial orientation of C-O bonds
is found in several marine polyether metabolites includ-
ing the DSP toxins,4 halichondrin,15 and aplysiatoxin.16
In contrast to this trend, the X-ray crystallographic
evidence for PTX1 (1) reveals an equatorial spiro C-O
F igu r e 4. Portions of the PTX8/9 molecules, showing the
NOEs around the spiroketal rings A and B.
dropyran ring and formation of a new spiro[5.5] system.
In accord with this proposal, the resonance for C-10 in
the 13C NMR spectra of 6 and 7, which falls in the
expected region of 78-85 ppm for carbons comprising a
saturated spiro[4.5] system,13 experienced a considerable
upfield shift (ca 13 ppm) in the spectrum of 9 (Table 1).
Further support for structure 9 comes from the magni-
tude and patterns of coupling for H-10 (ddd, 10, 10 and
4 Hz) and H-11 (d, 10 Hz) which are consistent with their
axial configuration. The stereochemistry around C-7
with respect to ring A in 9 was defined by a key NOE
between H-3 and H-11 (Figure 4) which indicated that
the oxygen at C-7 is axially orientated, and that C-8 is
equatorially oriented, as in the case of PTX7 (7). Com-
parison of the NMR data and inspection of the remainder
of the NOE data for 9 indicated that there were no
differences at the other stereocenters compared with 6
and 7, and hence the stereochemistry for 9 is as shown
(Figure 1).
Th e Str u ctu r es of P TX4 (4) a n d P TX8 (8). The LSI-
MS/MS spectra of the [m/z 897 (M + Na)+] ion for both
PTX1 (1) and PTX4 (4) were identical, providing further
evidence that the molecules were stereoisomers. Since
1
the H NMR spectrum of PTX4 (4) showed such a close
similarity with that of PTX7 (7), it was possible to match
many of the crucial assignments for the backbone reso-
nances which greatly facilitated the structural elucida-
tion process (Table 2). In particular, the resonances for
C-3, C-4, C-5, C-6, C-7, C-8, C-9, and C-10 were very
similar to those of the corresponding resonances in PTX7
(7), indicating that the configuration around C-7 was the
same in both molecules. On the other hand, the reso-
nances for the same positions in PTX1 (1) were similar
to the corresponding positions in PTX6 (6), supporting a
structural relationship between these two congeners.
Since the key NOEs which were noted in the PTX6/7 pair
were also observed in PTX1/4 (Figure 3), PTX4 (4) was
identified as 7-epi-PTX1 (Figure 1). Extrapolation of the
structural correlations between the PTX1/4 and PTX6/7
(13) Relevant examples can be found in: (a) Uemura, D.; Takahashi,
K.; Yamamoto, T.; Katayama, C.; Tanaka, J .; Okumura, Y.; Hirata, Y.
J . Am. Chem. Soc. 1985, 107, 4796. (b) Murata, M.; Legrand, A. M.;
Scheuer, P. J .; Yasumoto, T. Tetrahedron Lett. 1992, 33, 525. (c) Pettit,
G. R.; Ichihara, Y.; Wurzel, G.; Williams, M. D.; Schmidt, J . M.;
Chapuis, J . J . Chem. Soc., Chem. Commun. 1995, 383.
(14) Lee, J .-S.; Yanagi, T.; Kenma, R.; Yasumoto, T. Agric. Biol.
Chem. 1987, 51, 877.

2480 J . Org. Chem., Vol. 63, No. 8, 1998
Sasaki et al.
bond in the spiroketal arrangement of rings A and B, and
presumably other factors including the conformational
restraints of the macrocyclic lactone ring exert an influ-
ence over the spiroketal configuration at C-7.1 Acid
treatment of PTX6 (6), which contains the same equato-
rial spiro C-O bond as found in 1, results in spiroisomer-
ization at C-7 leading to formation of PTX7 (7) which
possesses a biaxial arrangement of C-O bonds. Interest-
ingly, the isomer PTX9 (9) formed concomitantly during
this process, also contains the more common biaxial
conformation of spiro C-O bonds in the expanded 1,7-
dioxo[5.5]undecane system.
shown to share the same skeleton as PTX9 (9), including
the spiro[5.5] system.
Exp er im en ta l Section
Gen er a l Meth od s. All solvents were HPLC grade unless
otherwise stated. The 1H NMR spectra were measured at 400
or 500 MHz or 600 MHz, and the 13C NMR spectra were
recorded at 100 or 150 MHz. NMR samples were dissolved in
C5D5N. FAB mass spectra were obtained using 3-nitrobenzyl
alcohol as matrix. PTXs were extracted from the digestive
glands of DSP-contaminated scallops, Patinopecten yessoensis,
collected in 1988, from Mutsu Bay, J apan, and were purified
and isolated following the previously reported procedures.3 To
avoid any interconversion catalyzed by the C-18 carboxyl
group, measurements of chemical properties of the PTX6/7/9
congeners were carried out using the corresponding phenacyl
ester derivatives.
It is likely that the spiroisomerization occurs within
the digestive gland of the scallop, perhaps selectively
mediated by a scallop-derived enzyme as opposed to a
random acid-catalyzed process, since the additional 1,7-
dioxo[5.5]undecane derivatives (e.g. PTX8 and PTX9) are
not observed in the scallop extracts. Other explanations
based on biogenetic considerations are less plausible: A
putative linear polyketide chain could form the substrate
for two different cyclases, which differ only in the
stereochemistry of formation of the ring A/B spiroketal
system. However, if this were the case, it would be
expected that both epimers would be observed in the
toxin-producing dinoflagellate.
P TX4 (4): white amorphous solid; [R]20 +2.07 (c 0.193,
D
MeOH); UVmax 235 nm (ꢀ 12000, MeOH); IR (KBr) 3500, 1760,
1
1730 cm-1; H and 13C NMR data are shown in Table 2.
P TX6 (6) p h en a cyl ester : white amorphous solid; [R]20
D
+8.77 (c 0.114, MeOH); UVmax 237 nm (ꢀ 37000, MeOH); IR
(KBr) 3500, 1760, 1730, 1720, 1700 cm-1
data are shown in Table 1.
;
1H and 13C NMR
P TX7 (7) p h en a cyl ester : white amorphous solid; [R]20
D
+11.5 (c 0.131, MeOH); UVmax 237 nm (ꢀ 37000, MeOH); IR
(KBr) 3500, 1760, 1730, 1720, 1700 cm-1
data are shown in Table 1.
;
1H and 13C NMR
1
P TX8 (8): white amorphous solid; [R]20 +19.8 (c 0.126,
From the perturbation of the H shifts in other parts
D
of the macrocyclic lactone (see Table 1), and from simple
modeling experiments, it is clear that spiroisomerization
at C-7 in 4 and 7 affects the overall conformation of the
macrocyclic lactone ring, as does expansion to form the
spiro[5.5] system. Not surprisingly perhaps, these changes
also affect the toxicity of these PTX derivatives in the
mouse bioassay. The lethal toxicity of PTX4 was 770 µg/
kg (mouse, i.p.),3 which is approximately one-third that
of PTX1, while that of PTX7 was more than 5 mg/kg
(mouse, i.p.). The toxicities of PTX8 and PTX9 were also
diminished, at more than 5 mg/kg (mouse, i.p.). Interest-
ingly, acid treatment of homohalichondrin B, a sponge
polyether, caused epimerization at the central spiroketal
carbon resulting in significant conformational changes
and concomitant changes in cytotoxicity.17
MeOH); UVmax 237 nm (ꢀ 12000, MeOH); IR (KBr) 3500, 1760,
1
1730 cm-1; H and 13C NMR data are shown in Table 2.
P TX9 (9) p h en a cyl ester : white amorphous solid; [R]20
D
+15.2 (c 0.105, MeOH); UVmax 239 nm (ꢀ 37000, MeOH); IR
(KBr) 3500, 1750, 1730, 1720, 1700 cm-1
data are shown in Table 1.
;
1H and 13C NMR
P r ep a r a tion of P h en a cyl Ester s. In a typical procedure,
a solution of PTX7 (7) (7.1 µmol, 6.3 mg) in acetone (200 µL)
containing triethylamine (36 µmol, 5 µL) was treated with
phenacyl bromide (15 µmol, 3 mg) at 40 °C for 2 h. The
resulting ester was isolated using an ODS column (Cosmosil
5C18-AR, 10 mm × 250 mm; Nacalai Tesque, Kyoto, J apan)
and elution with 80% MeCN/H2O, flow rate 2 mL/min. PTX7
phenacyl ester eluted at rt 14.5 min (3.5 mg, yield 49%);
LSIMS 1029 (M + Na)+. The phenacyl ester of PTX6 prepared
in the same way, eluted at rt 10.5 min (7.2 mg, yield 64%);
LSIMS m/z 1029 (M + Na)+.
Acid -Ca ta lyzed Rea r r a n gem en ts. In a typical experi-
ment, the phenacyl ester of PTX7 (20 µg) was dissolved in 90%
MeCN/H2O (40 µL) containing 0.1% TFA (v/v) at room tem-
perature. Reaction progress was monitored by LC-UV (Cos-
mosil 5C18-AR, 4.6 mm × 250 mm; eluant 70% MeCN/H2O,
flow rate 1 mL/min; detection at 235 nm). Under these
conditions the order of elution was PTX6 ester (rt 7.5 min),
PTX7 ester (rt 11.0 min) and PTX9 ester (rt 15.5 min).
La r ge Sca le P r ep a r a tion of P TX8 (8) a n d P TX9 (9).
These compounds were obtained by acid treatment of 1 and
6, respectively, following the reaction conditions described
above. The products were purified by HPLC (Cosmosil 5C18-
AR, 10 mm × 250 mm; 70% MeCN/H2O for 8, 80% MeCN/
H2O supplemented with 0.1% (v/v) acetic acid for 9 as eluant).
PTX8 (8) was obtained in 29% yield from PTX1 (1) and PTX9
(9) 35% yield from PTX6 (6).
Con clu sion
Both PTX6 (6) and PTX7 (7) undergo autocatalyzed
equilibration, and the process is accelerated by the
addition of TFA, whereupon a new derivative PTX9 (9)
is also produced. Detailed analyses of the NMR data
indicated that 7 is the 7-epi-derivative of 6 and contains
the anomerically favorable biaxial arrangement of C-O
bonds at the C-7 spiroketal position. The third product
PTX9 (9), which is formed concomitantly, arises by
opening of the spiro[4.5] system and recyclization to form
an enlarged spiro[5.5] system. In an identical manner,
PTX1 (1) and PTX4 (4) undergo equilibration upon
treatment with TFA, and 4 was identified as the 7-epi-
derivative of 1. An additional third product PTX8 (8) was
Ack n ow led gm en t. We are grateful to Dr. L. Glen-
denning for his assistance during preparation of this
manuscript, to Dr. J . Curtis for recording some of the
LSIMS data, and to Mr. H. Onodera for NMR measure-
ments of the PTX1/4/8 congeners. Participation of the
second author (J .L.C.W.) in this study was supported
by a scholarship from the J apanese Society for Promo-
tion of Science. This work was supported by a grant-
in-aid from the Ministry of Education, Science, Sports,
and Culture, J apan (NO. 07102002).
(15) (a) Uemura, D.; Takahashi, K.; Yamamoto, T.; Katayama, C.;
Tanaka, J .; Okumura, Y.; Hirata, Y. J . Am. Chem. Soc. 1985, 107, 4796.
(b) Pettit, G. R.; Tan. R.; Gao, F.; Williams, M. D.; Doubek, D. L.; Boyd,
M. R.; Schmidt, J . M.; Chapuis, J .-C.; Hamel, E.; Bai, R.; Hooper, J .
N. A.; Tackett, L. P. J . Org. Chem. 1993, 58, 2538. (c) Litaudon, M.;
Hart, J . B.; Blunt, J . W.; Lake, R. J .; Munro, M. H. G. Tetrahedron
Lett. 1994, 35, 9435.
(16) Moore, R. E.; Blackman, A. J .; Cheuk, C. E.; Mynderse, J . S.;
Matsumoto, G. K.; Clardy, J .; Woodard, R. W.; Craig, J . C. J . Org.
Chem. 1984, 49, 2484.
(17) Hart, J . B.; Blunt, J . W.; Munro, M. H. G. J . Org. Chem. 1996,
61, 2888.
J O971310B