Relative reactivities of aminoglycosides and their synthetic equivalents in the C-glycosylation of aromatics. Synthesis of the pseudoaglycone (D-C-glycoside) of the benzanthrins

Article abstract of DOI:10.1016/S0040-4020(00)00868-1

The pseudoaglycone of the benzanthrin antibiotics was prepared by a short sequence in which C-glycosylation of dehydrorabelomycin dimethyl ether served as the key step. The use of a 3-azido 2,3,6-trideoxy pyranose as an activated equivalent of a 3-dimethyl-amino 2,3,6-trideoxy sugar was demonstrated in this reaction. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00868-1

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 10255±10261
Relative Reactivities of Aminoglycosides and their Synthetic
Equivalents in the C-Glycosylation of Aromatics. Synthesis of the
Pseudoaglycone (d-C-Glycoside) of the Benzanthrins
*
Kathlyn A. Parker and Qing-jie Ding
Department of Chemistry, Brown University, Providence, RI 02912, USA
Dedicated to the memory of Arthur G. Schultz, a creative chemist and a nice person
Received 10 March 2000; accepted 6 May 2000
AbstractÐThe pseudoaglycone of the benzanthrin antibiotics was prepared by a short sequence in which C-glycosylation of dehydro-
rabelomycin dimethyl ether served as the key step. The use of a 3-azido 2,3,6-trideoxy pyranose as an activated equivalent of a 3-dimethyl-
amino 2,3,6-trideoxy sugar was demonstrated in this reaction. q 2000 Elsevier Science Ltd. All rights reserved.
Introduction
Retrosynthetic analysis of the structures of the benzanthrins
suggested O-glycosylation of the pseudoaglycone
3
(l-glycoside or d-glycoside), the non-sugar hydrolysis
product of benzanthrins A and B. This aryl C-glycoside
might be available from the coupling of dehydrorabelomy-
cin (4a) or one of its derivatives (e.g. 4b) with the rare
amino sugar angolosamine (5, drawn here as the d-sugar)4
or its synthetic equivalent. This approach required good
sources of both key intermediates (Scheme 1).
Benzanthrins A and B are structurally novel angucyclines
from Nocardia lurida. They exhibit strong antimicrobial
activities, especially against Gram-positive bacteria1 and
are active in tumor cell lines.2 Analysis of the spectroscopic
data for benzanthrins A and B allowed assignment of
structures 1 and 2, respectively, in which relative (but not
absolute) stereochemistry is designated for each of the
gylcoside moieties.
With rapid access to angucycline nuclei (including dehydro-
rabelomycin itself) from Michael addition chemistry,5 we
were poised to examine the C-glycosylation of dehydro-
rabelomycin derivatives. Studies focused on optimization
of the coupling of 4b and equivalents of amino pyranose 5
have now revealed the relative reactivities of 2,3-dideoxy
3-amino sugar synthons in Friedel±Crafts chemistry.6 These
results have been incorporated in the ®rst successful synthe-
sis of the benzanthrin pseudoaglycone (3, d-C-glycoside).
Among the angucycline natural products, the structures of
the benzanthrins are unusual in several respects. First and
most obvious, they have four fully aromatized rings.3
Second, in the benzanthrins, the tetracyclic nucleus bears
both O- and C-glycoside appendages. Third, the C-glyco-
sidic linkage is attached at C-2 in ring A rather than at C-9 in
ring D as for other angucyclines. Finally, unlike other
anguclycines, the benzanthrins contain amino sugars.
Keywords: amino sugars; glycosidation; quinones.
* Corresponding author. E-mail: kathlyn_parker@brown.edu
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00868-1

10256
K. A. Parker, Q. Ding / Tetrahedron 56 (2000) 10255±10261
catalysis by TMSOTf/AgClO4, to effect C-glycosylation
of b-naphthol in excellent yield.8 The protected and acti-
vated angolosamine 16 was prepared from the differentially
functionalized 10 which had been obtained by the literature
sequence.9 Debromination of 10 was best effected by hydro-
genolysis with Raney nickel (rather than with palladium),
giving the known 119 in quantitative yield (Scheme 3).
Scheme 1. Retrosynthesis of the benzanthrin pseudoaglycone 3 (d-C-glyco-
side).
In our hands, activated pyranose 16 glycosylated b-naphthol
in the presence of SnCl4 in 86% yield, presumably via an
O-glycosylation followed by an O- to C-glycoside migra-
tion.10 However, this reagent did not glycosylate substrate
4b even when extended times and higher temperatures were
employed (see Scheme 4). Likewise neither catalysis by
TMSOTf/AgClO4 nor by Cp2HfCl2/AgClO4, which has
effected glycosylation of b-naphthol with a similar angolo-
samine derivative,10 led to the formation of detectable
coupling product from substrates 16 and 4b.
Results
Dehydrorabelomycin dimethyl ether 4b was chosen as an
appropriate tetracyclic substrate for glycosylation studies.
This compound was readily available from hatamarubigin
A (6) by methylation and application of a previously
described two-step aromatization procedure (Scheme 2).5
In their descriptions of studies with b-naphthol, both Suzuki
and Toshima reported that amino sugars are less reactive as
glycosylating agents than are `neutral' sugars. We specu-
lated that our benzanthrin precursor 4b was a particularly
unreactive substrate because of steric hindrance and internal
The most convergent approach to the benzanthrin pseudo-
aglycone would involve the direct incorporation of angolo-
samine itself, necessarily protected on the C-4 oxygen.7 In
our ®rst approach, therefore, we examined the utility of
the known derivative 16 which has been reported, under
Scheme 2. Preparation of dimethyl dehydrorabelomycin 4b.
Scheme 3. Synthesis of activated angolosamine synthons 13, 14, and 16.
Scheme 4. Glycosylation experiments with angolosamine derivative 16.

K. A. Parker, Q. Ding / Tetrahedron 56 (2000) 10255±10261
10257
Scheme 5. Glycosylation of b-naphthol and of substrate 4b with a `neutral' sugar.
hydrogen bonding and we concluded that our attempts to
couple 16 and 4b were suffering from the combination of an
unreactive reagent and an unreactive substrate.
The prospect of testing a synthetic equivalent of 16 in the
glycosylation of 4b then appeared particularly attractive as
we had both intermediates 13 and 14 in hand. Mesylate 13
proved to be particularly reactive in coupling with substrate
4b, providing a 92% yield of the b-C-glycoside in 2 h at
2208 (Scheme 6). However, conversion of this product to
the C-30 azide 24 proceeded in poor yield and, although this
transformation was not optimized, we regarded it as
unpromising.
Assuming that glycosylation of 4b with a `neutral' sugar
would be more feasible, we prepared the known olivose
derivative 19.11 Although condensation was slow compared
to that with b-naphthol (Scheme 5), this `neutral' sugar and
4b, in the presence of SnCl4 at 08C, afforded a relatively
impressive 71% yield (total) of C- and O-glycosylation
products, 21 and 22.
The desired 24 could be obtained more ef®ciently by the
Scheme 6. Ef®cient glycosylation of substrate 4b with the mesyloxy sugar 13 and introduction of the C-30 N by azide displacement.
Scheme 7. Glycosylation of substrate 4b with the azido sugar 14 and elaboration to benzanthrin pseudoaglycone 3 (d-glycoside).

10258
K. A. Parker, Q. Ding / Tetrahedron 56 (2000) 10255±10261
direct coupling of substrate 4b with the activated azido
sugar 14 (Scheme 7). Although this glycosylation was
much slower than the corresponding condensation of
azido sugar 14 with b-naphthol, it afforded good yields of
the advanced intermediate 24 after extended times at
2258.12
the mixture was extracted with methylene chloride. The
organic phase was dried over Na2SO4 and concentrated to
yield a yellow solid. This material was immediately
dissolved in dry acetonitrile and the solution was treated
with palladium (II) acetate (54 mg, 0.24 mmol). After stir-
ring at room temperature overnight, solvent was removed
and the residue was chromatographed (2% ether/methylene
chloride) to yield 57 mg (82%) of a dark red solid.
Completion of the synthesis of the benzanthrin pseudo-
aglycone 3 as the d-C-glycoside now required functional
group manipulation. Debenzoylation was accomplished
with sodium methoxide in methanol affording 26. Successful
reduction of the azide group required the testing of a number
of sets of conditions13 but was neatly accomplished with the
SnCl2/thiophenol/triethylamine reagent.14 N,N-Dimethyl-
ation of the primary amine 27 with NaH2PO3 and aqueous
formaldehyde gave the dimethylamino compound 28.
Finally, unmasking of the phenolic hydroxyl groups with
aluminum bromide gave the pseudoaglycone of the benzan-
thrins (3) as the d-C-glycoside.
1
IR(CHCl3) 1675, 1589 cm21; H NMR (CDCl3) d 2.44 (s,
3H), 4.03 (s, 3H), 4.04 (s, 3H), 6.95 (s, 1H), 7.01 (s, 1H),
7.26 (d, 1H, Jˆ8.4 Hz), 7.49 (s, 1H), 7.63 (t, 1H, Jˆ8.4 Hz),
7.75 (d, 1H, Jˆ8.4 Hz), 10.96 (s, 1H); 13C NMR (CDCl3) d
21.3, 56.7, 56.8, 115.3, 117.3, 117.4, 118.4, 119.5, 120.1,
123.2, 131.5, 132.7, 134.5, 136.8, 139.4, 141.6, 154.5,
155.0, 158.1, 183.4, 192.0. HRMS calcd for C21H16O5
348.0997, found 348.0987.
Methyl 4-O-benzoyl-3-O-methanesulfonyl-2,6-dideoxy-
a-d-ribo-hexopyranoside (11).9 A solution of methyl hexo-
pyranoside 10 (4.62 g, 10.94 mmol) in methanol (120 mL)
containing Raney nickel (15 mL, 50% slurry in water) and
triethylamine (1 mL) was hydrogenated at room tempera-
ture for 12 h. The solid was ®ltered off and washed with
methanol and methylene chloride. The ®ltrate was concen-
trated and the residue was partitioned between methylene
chloride and water. The organic layer was dried with
sodium sulfate and concentrated to give the pure product
(3.62 g, 96%). 1H NMR (CDCl3) d 1.26 (d, 3H, Jˆ6.3 Hz),
2.15 (m, 1H), 2.30 (dd, 1H, Jˆ15.2, 3.0 Hz), 2.96 (s, 3H),
3.39 (s, 3H), 4.40 (m, 1H), 4.75 (d, 1H, Jˆ4.2 Hz), 4.81 (dd,
1H, Jˆ9.8, 3.1 Hz), 5.18 (q, 1H, Jˆ3.0 Hz), 7.46 (t,
Jˆ7.5 Hz, 2H), 7.60 (t, 1H, Jˆ7.5 Hz), 8.04 (d, 2H,
Jˆ7.5 Hz).
At this juncture, an observation and an ®nal comment are
appropriate. First, our studies of b-naphthol and our
benzanthrin intermediate 4b indicate the relative reactivities
of the 2,3,6-trideoxy 3-amino pyranose equivalents to be 13
(30-OMs).19 (30-OBz).14 (30-N3).16 (30-NMe2). Thus
3-azido 2,3-dideoxy sugars can be considered to be reactive
surrogates for 3-amino 2,3-dideoxy sugars in electrophilic
susbstitution reactions. Also, in experiments with substrate
4b, neither tri¯uoroacetamide 29 nor urethane 30 afforded
more than trace amounts of C-aryl glycoside.
4-O-Benzoyl-3-O-methanesulfonyl-2,6-dideoxy-b-d-ribo-
hexopyranose (12). A solution of methyl hexopyranoside
11 (1.01 g, 2.94 mmol) in aqueous acetic acid (80%, 23 mL)
containing 2N HCl (6 mL) was stirred at re¯ux for 1.5 h.
The mixture was cooled and the reaction was quenched with
sodium bicarbonate. The mixture was extracted with meth-
ylene chloride and the organic layer was dried with sodium
sulfate and concentrated below room temperature. The resi-
due was ®ltered through a short column of silica gel to yield
a colorless oil (810 mg, 84%). IR (CDCl3) 3356, 1725,
In conclusion, benzanthrin pseudoaglycone 3, d-C-glyco-
side, was prepared from the readily accessible hatamarubi-
gin A dimethyl ether 4b in 59% overall yield by a 5-step
sequence in which the crucial C-aryl glycosylation utilized a
3-azido-2,3,6-trideoxy sugar.
1
1602, 1452, 1365, 1265, 1179 cm2l. H NMR (CDCl3) d
1.43 (d, 3H, Jˆ6.0 Hz), 1.6 (s, OH?), 1.91 (m, 1H), 2.40
(dm, 1H, Jˆ<12 Hz), 2.90 (s, 3H), 4.15 (m, 1H), 4.80 (dd,
1H, Jˆ10.0, 3.0 Hz), 5.25 (m, 2H), 7.46 (t, 2H, Jˆ7.5 Hz),
7.60 (t, 1H, Jˆ7.5 Hz), 8.0 (d, 2H, Jˆ7.5 Hz).
Experimental
Melting points were taken on a Thomas±Hoover capillary
apparatus and are uncorrected. IR spectra were recorded on
a Perkin±Elmer 1600 series FT-IR. 1H and 13C NMR spectra
were recorded on either a Bruker WM 250 or a Bruker AM
400 WB instrument. Ether refers to diethyl ether. All
reactions were conducted under an atmosphere of N2 unless
otherwise indicated.
Acetyl 4-O-benzoyl-3-O-methanesulfonyl-2,6-dideoxy-b-
d-ribo-hexopyranoside (13). Methyl hexopyranoside 11
(1.03 g) was dissolved in aqueous acetic acid (80%, 5 mL)
and the mixture was stirred at re¯ux for 40 min. The solu-
tion was then cooled and neutralized with saturated sodium
bicarbonate solution. The mixture was extracted with
methylene chloride and the organic phase was dried (sodium
sulfate) and concentrated to yield an off-white solid. The
solid was immediately dissolved in pyridine (2.5 mL) and
acetic anhydride (1 mL) was added at 08C. The resulting
mixture was stirred at 08C for 14 h. The excess pyridine
was removed at reduced pressure and the residue was
subjected to column chromatography (50% ethyl acetate
1-Hydroxy-3-methyl-6,8-dimethoxy-benz[a]anthracene-
7,12-dione (4b). To a stirred solution of 6-O-methyl
hatomarubigin A5 (70 mg, 0.20 mmol) in 3 mL of methyl-
ene chloride were added at 2408C triethylamine (40 mg, 55,
0.40 mmol) and trimethylsilyl tri¯uoromethanesulfonate
(53 mg, 46 mL, 0.26 mmol). The reaction was quenched
15 min later with the addition of saturated NaHCO3 and

K. A. Parker, Q. Ding / Tetrahedron 56 (2000) 10255±10261
10259
in hexanes) to give 970 mg (87%) of an off-white solid. IR
(CDCl3) 1726 cm21 1H NMR (CDCl3) d 1.32 (d, 3H,
Jˆ6.3 Hz), 2.13 (s, 3H), 2.10±2.20 (dm, 1H, Jˆ<12 Hz),
2.48 (m, 1H), 2.95 (s, 3H), 4.25 (m, 1H), 4.84 (dd, 1H,
Jˆ9.5, 9.5 Hz), 5.37 (m, 1H), 6.10 (dd, 1H, Jˆ9.4,
2.2 Hz), 7.41 (t, 2H, Jˆ7.5 Hz), 7.52 (t, 1H, Jˆ7.5 Hz),
8.10 (d, 2H, Jˆ7.5 Hz). 13C NMR (CDCl3) d 17.8, 21.0,
35.4, 38.5, 69.1, 72.0, 75.3, 89.9, 128.6, 129.3, 129.7,
137.7, 165.5, 168.9. HRMS calcd for C16H20SO8
372.0879, found 372.0881.
3343, 1716, 1622, 1602 cm2l. 1H NMR (CDCl3) d 1.41 (d,
3H, Jˆ6.2 Hz), 2.05 (m, 1H), 2.23 (m, 1H), 2.31 (s, 6H),
3.40 (m, 1H), 3.90 (m, 1H), 5.16 (dd, 1H, Jˆ10, 10 Hz),
5.53 (dd, 1H, Jˆ2, 11.4 Hz), 7.13 (d, 1H, Jˆ8.8 Hz), 7.30 (t,
1H, Jˆ8.7 Hz), 7.47 (t, 3H, Jˆ<6 Hz), 7.60 (t, 1H, Jˆ
<6 Hz), 7.72 (t, 2H, Jˆ9.6 Hz), 7.78 (d, 1H, Jˆ6.9 Hz),
8.08 (d, 2H, Jˆ<7 Hz), 8.88 (s, 1H).
C-Aryl glycoside (20). To a stirred solution of 2-naphthol
Ê
(9 mg, 0.033 mmol), 4 A molecular sieves (50 mg) and tin
(IV) chloride (1 M solution in methylene chloride, 0.08 mL)
at 2788C, acetyl 3,4-di-O-benzoyl-2,6-dideoxy-a-d-
arabino-hexopyranoside 19 (11 mg, 0.028 mmol) in
1.2 mL methylene chloride was added via a syringe. The
mixture was stirred for 5 min at 2788C and then the
temperature was gradually increased to 2208C during
40 min. The reaction was quenched with saturated sodium
bicarbonate and the mixture was extracted with methylene
chloride. The organic phase was dried with sodium sulfate
and concentrated to give an oil (12 mg, 91%) after column
Azido hexopyranoside 14. Acetyl hexopyranoside 13
(300 mg, 0.81 mmol), sodium azide (80 mg, 1.23 mmol)
and DMF (3 mL) were combined and the mixture was stir-
red at 758C for 34 h. The mixture was cooled and poured
into water and extracted with benzene. The benzene extracts
were dried with sodium sulfate and concentrated to yield a
viscous oil which was passed through a short pad of silica
gel (33% ethyl acetate and hexanes as eluent). Concentra-
tion of the ®ltrate gave a pale yellow oil (23 mg, 91%).
[a]25Dˆ2268 (cˆ0.1, CHCl3); IR (CDCl3) 2103, 1762,
1
chromatography (1:2 ethyl acetate and hexanes). H NMR
1
1730 cm21; H NMR (CDCl3) d 1.27 (d, 3H, Jˆ6.2 Hz),
(CDCl3) d 1.47 (d, 3H, Jˆ6.2 Hz), 2.28 (apparent q, 1H,
Jˆ<11 Hz), 2.70 (dm, 1H, Jˆ<11 Hz), 4.03 (m, 1H), 5.43
(dd, 1H, Jˆ9.6, 9.6 Hz), 5.63 (m, 1H), 5.70 (dd, 1H,
Jˆ11.7, 2.3 Hz), 7.11 (d, 1H, Jˆ8.6 Hz), 7.24±7.60 (m,
8H), 7.75 (m, 3H), 7.85 (d, 2H, Jˆ8.5 Hz), 8.02 (d, 2H,
Jˆ7.5 Hz), 8.72 (s, 1H).
1.81 (q, 1H, Jˆ10 Hz), 2.15 (s, 3H), 2.31 (dm, 1H,
Jˆ<10 Hz), 3.71 (m, 1H), 4.91 (dd, 1H, Jˆ9.6, 9.6 Hz),
5.80 (dd, 1H, Jˆ10, 2.2 Hz), 7.45 (m, 2H), 7.58 (t, 1H,
Jˆ7.5 Hz), 8.04 (d, 2H, Jˆ7.5 Hz); 13C NMR (CDCl3) d
17.5, 21.0, 35.0, 59.7, 72.0, 74.8, 91.4, 128.5 (2 C), 129.1,
129.8 (2 C), 133.5, 165.5, 169.0.
C-Glycoside 21 and O-glycoside 22. To a stirred solution
of the dimethyl ether of dehydrorableomycin (4b, 13 mg,
0.037 mmol) in methylene chloride (1 mL) containing
Dimethylamino hexopyranoside 16. To a stirred solution
of the azido sugar 14 (3 mg, 9 mmol) in dry ethyl acetate
(2 mL), palladium on activated carbon (10%) was added and
the mixture was hydrogenated at room temperature for
30 min. The catalyst was ®ltered off and the ®ltrate was
concentrated. The residue was immediately dissolved in
acetonitrile (2 mL) and treated with aqueous formaldehyde
(37%, 17.2 mL) and sodium cyanoborohydride (4 mg).
After the mixture was stirred for 15 min, acetic acid was
added to make the solution close to neutral (checking with a
piece of wet pH paper). The reaction was continued for an
additional 40 min and quenched with saturated sodium
bicarbonate. The mixture was extracted with methylene
chloride and the organic phase was dried and concentrated
to yield a white crystal, 3 mg (quantitative). IR (CDCl3)
Ê
molecular sieves (4 A) and tin (IV) chloride (1 M solution
in methylene chloride, 0.1 mL) at 2788C, hexopyranoside
19 (13 mg, 0.034 mol) in 0.5 mL methylene chloride was
added via a syringe and the temperature was gradually
increased to 08C. The reaction was continued at that
temperature for 9 h before it was quenched with saturated
sodium bicarbonate. The mixture was extracted with
methylene chloride and the organic phase was dried with
sodium sulfate and concentrated to yield a dark solid, which
was chromatographed (80:1 methylene chloride and ether)
to afford both the C-glycoside (11 mg, 45%) and the
O-glycoside (6 mg, 26%). C-Glycoside 21 IR (CHCl3)
1
3450, 1766, 1723, 1675, 1591 cm2l. H NMR (CDCl3) d
1
1720, 1602, 1269 cm21; H NMR (CDCl3) d 1.22 (d, 3H,
1.38 (d, 3H, Jˆ6.1 Hz), 2.33±2.62 (m, 2H), 2.63 (s, 3H),
3.90 (m, 1H), 4.00 (s, 3H), 4.03 (s, 3H), 5.35 (dd, 1H, Jˆ9.5,
9.5 Hz), 5.50 (m, 2H), 7.13 (s, 1H), 7.24±7.55 (m, 8H), 7.95
(d, 2H, Jˆ7.0 Hz), 8.05 (d, 2H, Jˆ7.1 Hz), 10.71(s, 1H).
13C NMR (CDCl3) d 18.2, 21.3, 35.0, 56.45, 56.5, 73.2,
73.7, 75.0, 75.1, 115.4, 115.5, 117.6, 119.4, 121.6, 122.9,
123.0, 128.25±128.3 (2 C), 129.6±129.7 (2 C), 130.1,
133.0, 133.2, 134.0, 134.1, 136.8, 137.7, 139.5, 152.7,
154.7, 158.0, 165.9, 170.0, 183.0, 190.8. HRMS calcd for
C41H34O10 686.2152, found 686.2148. O-Glycoside 22, a-
and b-anomers, ratio <3: 1, IR(CHCl3) 1762, 1724, 1674,
1592 cm21. 1H NMR (CDCl3) d 0.84 major and 1.15 minor
(d, Jˆ7.0 Hz and d, Jˆ7.0 Hz, total 3H), 2.10 (m, 1H), 2.20
minor and 2.60 major (s and s, total 3H), 2.68 (m, 1H), 3.50
(m, 1H), 4.00 (m, 1H), 4.01(s, 3H), 4.02 (s, 3H), 5.10 major
and 5.22 minor (bs and m, 1H total), 5.75 minor and 6.10
major (s and s, total 1H), 7.10±8.20 (m, 15H).
Jˆ6.2 Hz), 1.52 (m, 1H), 2.05 (m, 1H), 2.14 (s, 3H), 2.27 (s,
6H), 2.90 (m, 1H), 3.60 (dd, 1H, Jˆ9.2, 6.2 Hz), 4.90 (dd,
1H, Jˆ9.8, 9.8 Hz), 5.78 (dd, 1H, Jˆ9.7, 2.2 Hz), 7.42±7.48
(t, 2H, Jˆ7.5 Hz), 7.55 (t, 1H, Jˆ7.5 Hz), 8.02 (d, 2H,
Jˆ<7 Hz); HRMS calcd for C17H23NO5 321.1576, found
321.1579.
Dimethylamino C-aryl glycoside 17.8 To a stirred solution
of 2-naphthol (29 mg, 0.20 mmol) in 1 mL of methylene
chloride containing tin (IV) chloride (0.31 mL, 1 M solution
in methylene chloride, 0.30 mmol) at 2788C, sugar 16
(32 mg, 0.1 mmol) in 0.5 mL of methylene chloride was
added and the mixture was gradually warmed to rt and
stirred overnight. The reaction was quenched with sat.
sodium bicarbonate and the resulting mixture was extracted
with methylene chloride. The solvent was removed and the
residue was chromatographed (10% ether in methylene
chloride) to give a viscous oil (35 mg, 86%). IR (CDCl3)
Methanesulfonyl C-aryl glycoside 23. To a stirred solution

10260
K. A. Parker, Q. Ding / Tetrahedron 56 (2000) 10255±10261
of the aromatic compound 4b (12 mg, 0.034 mmol), ®rst tin
(IV) chloride (1 M solution in methylene chloride, 0.12 mL,
3.5 equiv.) and then hexopyranoside 13 (18 mg,
0.034 mmol) were slowly added. The mixture was stirred
at 2788C for 15 min and then it was gradually warmed to
2208C. The reaction was continued at this temperature until
TLC (2.5% ether in methylene chloride) showed no
aromatic starting material (17 h). The reaction was then
quenched with sat. sodium bicarbonate and the mixture
was extracted with methylene chloride. The organic phase
was dried with sodium sulfate and concentrated to yield a
dark red solid. Column chromatography (25% ether in
methylene chloride) provided the C-glycoside (23 mg,
92%) as the b-anomer. IR(CDCl3) 3324, 1722, 1675,
C-Glycoside 25. To a stirred solution of 2-naphthol (4.5 mg,
30 mmol) in methylene chloride (1 mL) containing tin (IV)
chloride (1 M solution in CH2Cl2, 45 mmol) at 788C was
hexopyranoside 14 (5 mg, 15 mmol) with 1 mL methylene
chloride. The mixture was stirred at 2788C for 10 min and
then the temperature was gradually increased to 2358C and
kept overnight. The reaction was quenched with saturated
sodium sulfate and the mixture was extracted with methyl-
ene chloride. The organic phase was dried and concentrated
to yield a solid, which gave a colorless solid after column
chromatography (1:3 ethyl acetate and hexanes as eluent).
6 mg, 95%. IR (CDCl3) 3368, 2102, 1728, 1624, 1602,
1524, 1264 cm2l; 1H NMR (CDCl3) d 1.40 (d, 3H,
Jˆ6.2 Hz), 2.08 (m, 1H), 2.44 (dm, 1H, Jˆ<12 Hz), 3.88
(m, 2H), 5.10 (dd, 1H, Jˆ9.7, 9.7 Hz), 5.56 (dd, 1H,
Jˆ11.5, 1.7 Hz), 7.11 (d, 1H, Jˆ8.8 Hz), 7.32 (t, 1H,
Jˆ8.0 Hz), 7.40±7.78 (m, 7H), 8.07 (d, 2H, Jˆ8.0 Hz),
8.60 (s, 1H); 13C NMR (CDCl3) d 18.1, 36.0, 61.1, 75.3,
76.4, 76.5, 113.7, 113.8, 120.0, 120.3, 123.1, 127.0, 128.6,
128.8, 129.1, 129.9, 130.3, 130.6, 133.6, 153.7, 165.6.
1
1618, 1591, 1365, 1267 cm21; H NMR (CDCl3) d 1.33
(d, 3H, Jˆ6.7 Hz), 2.30 (dm, 1H, Jˆ<12 Hz), 2.60 (s,
3H), 2.60±2.70 (m 2H), 3.05 (s, 3H), 4.10 (s, 3H), 4.11 (s,
3H), 4.20 (m, 1H), 5.03 (dd, 1H, Jˆ10, 3 Hz), 5.41 (m, 1H),
5.78 (dd, 1H, Jˆ11.8, 2 Hz), 7.12 (s, 1H), 7.22 (m, 1H), 7.38
(s, 1H), 7.50 (t, 2H, Jˆ7.5 Hz), 7.60 (m, 2H), 7.65 (t,
1H, Jˆ7.5 Hz) 8.10 (d, 2H, Jˆ7.5 Hz), 10.68 (s, 1H);
13C NMR (CDCl3) d 18.2, 21.2, 29.3, 35.0, 38.8, 56.5,
70.4, 71.0, 73.0, 77.1, 115.3, 115.5, 117.6, 119.4, 121.5,
122.6, 122.9, 128.6, 129.3, 129.8, 130.1, 133.6, 134.0,
134.1, 136.8, 137.7, 139.5, 152.4, 154.8, 158.1, 165.7,
183.0, 190.7. HRMS calcd for C35H32SO11 660.1665,
found 660.1671.
Azido C-glycoside 26. To the stirred solution of product 24
(10 mg, 16.5 mmol) in a mixture of methanol and THF (4:1,
6 mL), sodium methoxide (20 mg) was added and the
mixture was stirred at room temperature for 30 h. The
solvent was removed at reduced pressure and the residue
was partitioned between methylene chloride and diluted
hydrochloric acid (0.1N). The organic phase was dried
(sodium sulfate) and concentrated to give a dark solid,
which was passed through a short column of silica gel to
yield the pure product. IR(CD2Cl2) 2105, 1680, 1590,
Azido C-aryl glycoside 24. To a stirred solution of the
Ê
aromatic 4b (15 mg, 42.5 mmol), 4 A molecular sieves
(80 mg) and tin (IV)chloride (0.15 mL, 1 M solution in
methylene chloride) in 1 mL methylene chloride, acetyl
azido sugar 14 in 1 mL of methylene chloride was added
via a syringe. The mixture was kept at 2788C for 15 min
and then gradually warmed to 2208C and then kept in the
refrigerator (at approximately 2208C) for 5 days. The reac-
tion was quenched by the addition of saturated sodium
bicarbonate and the mixture was extracted with methylene
chloride. The organic layer was dried with sodium sulfate
and concentrated. The residue was subject to thin layer
chromatography (2% ether in methylene chloride) to
provide three compounds. The ®rst component proved to
be unreacted aromatic 4b (3 mg). The second component
was C-glycoside 24 (16 mg, 78%, dark red) and the third
component was para-C-glycoside 240 (2.4 mg, 12%, dark
red). Spectra of ortho-C-glycoside 24: IR (CD2Cl2) 3333,
2104, 1727, 1678, 1590 cm21; 1H NMR (CDCl3) d 1.30 (d,
3H, Jˆ6.1 Hz), 2.20±2.31 (m, 2H), 2.64 (s, 3H), 3.78 (m,
1H), 3.95 (m, 1H), 4.00 (s, 3H), 4.02 (s, 3H), 5.04 (dd, 1H,
Jˆ9.6, 9.6 Hz), 5.52 (dd, 1H, Jˆ11.2, 2.2 Hz), 7.16 (s, 1H),
7.30 (d, 1H, Jˆ8.3 Hz), 7.51 (s, 1H), 7.53 (m, 2H), 7.62 (m,
3H), 8.10 (d, 2H, Jˆ7.5 Hz), 10.96 (s, 1H); 13C NMR
(CDCl3) d 18.3, 21.5, 35.0, 56.7, 56.8, 62.4, 74.2, 75.8,
76.3, 115.7, 116.2, 118.2, 119.8, 122.1, 123.2, 123.8,
128.9. 130.0 (2 C), 130.1, 131.3 (2 C), 133.7, 133.8,
134.5, 137.1, 138.0, 140.2, 152.2, 154.9, 158.2, 166.1,
183.1, 191.5. Spectra for the para-C-glycoside 240
1
1279 cm2l; H NMR (CDCl3) d 1.38 (d, 3H, Jˆ6.2 Hz),
2.00 (m, 2H), 2.55 (s, 3H), 3.15 (m, 1H), 3.40 (m, 1H),
3.50 (m, 1H), 3.99 (s, 3H), 4.01(s, 3H), 5.32 (dd, 1H,
Jˆ2.0, 10.5 Hz), 7.13 (s, 1H), 7.27 (m, 1H), 7.41 (s, 1H),
7.64 (m, 2H), 10.58 (s, 1H); 13C (CDCl3) d 18.4, 21.3, 34.8,
56.7, 56.8, 64.9, 74.4, 76.3, 77.3, 115.4, 115.6, 118.0, 119.4,
121.7, 123.4 (2 C), 130.5, 134.5 (2 C), 137.3, 138.0, 139.8,
152.3, 155.0, 158.3, 182.9, 190.9; HRMS calcd for
C27H25N3O7 503.1625, found 503.1627.
Amino C-glycoside 27. A solution of azido C-glycoside 26
(6 mg, 0.012 mmol) in 0.5 mL tin (II) chloride solution
(prepared by dissolving 285 mg SnCl2, 0.62 mL thiophenol
and 0.62 mL triethylamine in 10 mL THF) was stirred at
room temperature for 20 min before another 0.30 mL of
the tin (II) chloride solution was added. The reaction was
continued for another 50 min before it was quenched with
1N NaOH solution. The pH was adjusted to 9.5 with 0.5N
HCl solution and the mixture was extracted with methylene
chloride. The organic phase was dried with sodium sulfate
and concentrated to yield a dark solid, which was passed
through a short silica gel column, eluted ®rst with methylene
chloride to get rid of the thiophenol and then with 15%
methanol in methylene chloride to give a dark solid
(5.7 mg, 98%). IR (CDCl3) 3295, 1673, 1619, 1591, 1382,
1
1277, 1235, 1075 cm21; H NMR (CDCl3) d 1.40 (d, 3H,
IR(CD2Cl2) 2103, 1727, 1676, 1588 cm21
;
1H NMR
Jˆ6.0 Hz), 2.1 (m, 2H), 2.40 (s, 3H), 3.10 (m, 1H), 3.20 (dd,
1H, Jˆ9.7, 8.9 Hz), 3.52 (m, 1H), 4.10 (s, 3H), 4.11 (s, 3H),
5.24 (d, 1H, Jˆ9.6 Hz), 7.10 (s, 1H), 7.20 (d, 1H,
Jˆ8.2 Hz), 7.30 (s, 1H), 7.50 (m, 2H); 13C (CDCl3) d
19.1, 21.4, 40.4, 55.3, 56.6, 56.7, 75.7, 77.8, 112.9, 114.5,
116.5, 117.6, 119.7, 121.6, 122.6, 127.6, 133.2, 134.8,
(CDCl3) d 1.30 (d, 3H, Jˆ6.1 Hz), 2.15 (m, 2H), 2.49 (s,
3H), 3.80 (m, 2H), 4.00 (s, 3H), 4.11 (s, 3H), 5.05 (dd, 1H,
Jˆ9.6, 9.6 Hz), 5.23 (dd, 1H, Jˆ11.9, 2.7 Hz), 6.90 (s, 1H),
7.25 (d, 1H), 7.40 (m, 2H), 7.52 (m, 2H), 7.71 (m, 1H), 8.07
(d, 2H, Jˆ7.5 Hz), 8.77 (s, 1H), 10.55 (s, 1H).

K. A. Parker, Q. Ding / Tetrahedron 56 (2000) 10255±10261
10261
2. Benzanthrin A was inactive in animal models for sarcoma,
leukemia and lung tumors. See Ref. 1b above.
136.5, 136.6, 137.9, 151.1, 153.6, 156.8, 180.7, 187.6.
HRMS calcd for C27H27NO7 477.1788, found 477.1782.
3. Only antibiotic C104 shares this structural feature. See
Matsumoto, T.; Yamaguchi, H.; Suzuki, K. Tetrahedron 1997,
53, 16533±16544 and references therein.
Dimethylamino C-glycoside 28. The above primary amine
27 (6 mg, 0.0123 mmol) was dissolved in 1,4-dioxane
(0.5 mL) and treated with formaldehyde (20 mL) and an
aqueous solution of NaH2PO3 (1N, 0.5 mL). The mixture
was stirred at 608C for 5 h and the reaction was quenched
with 0.5N NaOH solution. The mixture was extracted with
methylene chloride and the organic phase was dried with
sodium sulfate and then concentrated. The residue was
passed through a short column of silica gel and washed
with 15% methanol in methylene chloride to give a dark
orange solid (5.3 mg, 88%) IR (CDCl3) 3294, 1674, 1621,
4. For convenience, we have employed d-sugars in our synthetic
studies.
5. See the previous paper in this series.
6. An excellent discussion of the electrophilic substitution
approach to C-aryl glycosides is contained in a review by
Jaramillo, C.; Knapp, S. Synthesis, 1994, 1±20.
7. C-Glycosylations with unprotected sugars (not amino sugars)
have been reported; see Matsuo, G.; Nakata, M.; Matsumura, S.;
Toshima, K. J. Org. Chem. 1999, 64 7101±7106. However, yields
are generally better with protected, C-1 activated sugars.
8. Toshima, K.; Matsuo, G.; Tatsuta, K. Tetrahedron Lett. 1992,
34, 2175±2178.
1
1591, 1278, 1074 cm21; H NMR (CDCl3) d 1.44 (d, 3H,
Jˆ6.0 Hz), 1.92 (m, 2H), 2.30 (d, 1H, OH, Jˆ8.5 Hz), 2.42
(s, 6H), 2.53 (s, 3H), 2.80 (m, 1H), 3.30 (dd, 1H, J1ˆ9.5 Hz,
J2ˆ9.0 Hz), 3.51 (m, 1H), 4.01 (s, 3H), 4.07 (s, 3H), 5.27
(dd, 1H, J1ˆ10.0, 3.3 Hz), 7.10 (s, 1H), 7.24 (d, 1H,
Jˆ8 Hz), 7.34 (s, 1H), 7.60 (dd, 1H, Jˆ8.0, 7.9 Hz), 7.67
(d, 1H, Jˆ7.9 Hz), 10.37 (s, 1H); 13C NMR (CDCl3) d 18.7,
21.1, 25.9, 40.3, 56.4, 56.5, 67.7, 71.4, 75.5, 77.2, 78.2,
114.2, 114.4, 115.4, 117.3. 118.9, 121.1, 123.0, 129.1,
134.1, 134.9, 137.1, 137.6, 138.9, 152.1, 154.8, 158.1,
182.8, 190.1; HRMS calcd for C29H31NO7 505.2101,
found 505.2108.
9. Bartner, P.; Boxler, D. L.; Brambilla, R.; Mallams, A. K.;
Morton, J. B.; Reichert, P.; Sancilio, F. O.; Surprenant, H.;
Tomalesky, G. T.; Lukacs, G.; Olesker, A.; Thang, T. T.; Valente,
L.; Omura, S. J. Chem. Soc., Perkin Trans. 1 1979, 1600±1624.
10. Matsumoto, T.; Hosoya, T.; Suzuki, K. Tetrahedron Lett.
1990, 32, 4629±4632.
11. Matsumoto, T.; Katsuki, M.; Jona, H.; Suzuki, K. J. Am.
Chem. Soc. 1991, 113, 6982±6992.
12. In this reaction, a small amount of the para-substituted
C-glycoside 240 was recovered along with traces of the tetracyclic
starting material. Higher temperature (08C) resulted in the produc-
tion of larger amounts of 240 and also decomposition products.
Benzanthrin pseudoaglycone (3, d-C-glycoside). Di-
methylamino C-glycoside 28 (6 mg, 0.012 mmoL) was
treated with aluminum bromide (0.25 mL, 1 M solution in
CH2Br2) solution at 08C. The mixture was stirred at 08C for
1 h and the reaction was quenched with saturated sodium
bicarbonate. The mixture was extracted with methylene
chloride (9 times). The organic phase was dried with sodium
sulfate and concentrated to yield a dark green solid (5.3 mg,
93%). IR (CDCl3) 1630, 1606, 1457.8, 1277 cm21; 1H NMR
(CDCl3) d 1.40 (d, 3H, Jˆ6.2 Hz), 2.00 (m, 2H), 2.35 (s,
6H), 2.53 (s, 3H), 2.70 (m, 1H), 3.20 (dd, 1H, Jˆ10, 10 Hz),
3.42 (m, 1H), 5.20 (dd, 1H, Jˆ9.7, 2.5 Hz), 7.00 (s 1H), 7.20
(m, 1H), 7.35 (s, 1H), 7.60 (m, 2H), 9.85 (s, 1H), 11.71 (s,
1H), 11.86 (s, 1H). HRMS calcd for C27H27NO7 477.1788,
found 477.1782.
13. (a) Chromium (II) chloride: Kondo, J.; Nakai, H.; Goto, T.
Tetrahedron, 1973, 29, 1801. We had successfully employed this
reagent for the reduction of the b-naphthol C-glycoside. However,
with 26, chromium (II) chloride reduces the quinone to hydro-
quinone and further reduction is sluggish. (b) Triphenyl phosphine,
aqueous THF: Vaultier, M. Knouzi, N.; Carrie, R. Tetrahedron
Lett. 1983, 24, 763±764. This reaction was slow and afforded
numerous products. (c) Transfer hydrogenation over 5% Pd/C:
Gartiser, T.; Selve, C.; Delpuech, J. J. Tetrahedron Lett. 1983,
24, 1609±1610. These conditions gave an unidenti®ed product.
(d) Hydrogenation over Lindlar catalyst: Corey, E. J.; Nicolaou,
K. C.; Balanson, R. D.; Machida, Y. Synthesis, 1975, 590±591.
The reduction followed by N,N-dimethylation with formaldehyde
and NaH2PO3 afforded 20% of the desired 28.
Acknowledgements
This work was supported by the National Science Founda-
tion (NSF CHE-9521055) and by the National Institutes of
Health (CA 87503).
References
1. (a) Theriault, R.; Rasmussen, R.; Kohl, W. L.; Prokop, J. F.;
Hutch, T. B.; Barlow, G. J. J. Antibiot. 1986, 39, 1509.
(b) Rasmussen, R. R.; Nuss, M. E.; Scherr, M. H.; Mueller, S. L.;
McAlpin, J. P. J. Antibiot. 1986, 39, 1515.
14. Bartra, M.; Romea, P.; Urpi, F.; Vilarrasa, J. Tetrahedron
1990, 46, 587±594.