An HIV RNase H inhibitory 1,3,4,5-tetragalloylapiitol from the African plant Hylodendron gabunensis

Article abstract of DOI:10.1021/np0702279

A new compound, 1,3,4,5-tetragalloylapiitol (1), was isolated from the aqueous extract of the plant Hylodendron gabunensis and was found to be a potent inhibitor of RNase H enzymatic activity. The structure of 1 was elucidated by NMR analyses to be an apiitol (2) sugar moiety substituted with four gallic acid residues. Optical rotation measurements of the free sugar following basic hydrolysis indicated that the 35 absolute configuration was the same as that of D-apiitol. Compound 1 inhibited HIV-1, HIV-2, and human RNase H with IC ₅₀ values of 0.24, 0.13, and 1.5 μM, respectively, but it did not show inhibition of E. coli RNase H at 10 μM.

Full text of DOI:10.1021/np0702279

J. Nat. Prod. 2007, 70, 1647–1649
1647
An HIV RNase H Inhibitory 1,3,4,5-Tetragalloylapiitol from the African Plant
Hylodendron gabunensis
Kentaro Takada,† Alun Bermingham,† Barry R. O’Keefe,† Antony Wamiru,†,‡ John A. Beutler,† Stuart F. J. Le Grice,§
John Lloyd, Kirk R. Gustafson,*,† and James B. McMahon†
Molecular Targets DeVelopment Program, Center for Cancer Research, National Cancer Institute, Building 1052, Room 121,
Frederick, Maryland 21702-1201
ReceiVed May 14, 2007
A new compound, 1,3,4,5-tetragalloylapiitol (1), was isolated from the aqueous extract of the plant Hylodendron
gabunensis and was found to be a potent inhibitor of RNase H enzymatic activity. The structure of 1 was elucidated by
NMR analyses to be an apiitol (2) sugar moiety substituted with four gallic acid residues. Optical rotation measurements
of the free sugar following basic hydrolysis indicated that the 3S absolute configuration was the same as that of D-apiitol.
Compound 1 inhibited HIV-1, HIV-2, and human RNase H with IC50 values of 0.24, 0.13, and 1.5 µM, respectively,
but it did not show inhibition of E. coli RNase H at 10 µM.
Human immunodeficiency virus-type 1 (HIV-1) reverse tran-
scriptase (RT) has two distinct enzymatic domains, and these
domains separately carry out a RNA-dependent/DNA-dependent
DNA polymerization reaction and ribonuclease H (RNase H)
hydrolytic activity. RNase H specifically hydrolyzes the RNA strand
of a RNA/DNA heteroduplex.1–3 The RNase H function of HIV
RT is required to effectively incorporate viral genetic information
into the host cell genome.4,5 While both of the RT activities are
critical for viral infectivity, only the polymerase activity has been
successfully exploited as a target for commercial drugs.6 Therefore,
HIV-1 RNase H remains an attractive molecular target for develop-
ing new anti-HIV agents for potential chemotherapeutic applica-
tions.7 As part of a screening campaign to find compounds that
can inhibit RNase H activity, a library of natural product extracts
(n ) 82 067) was screened by our group. Active extracts from the
initial RNase H screen were confirmed and then prioritized using
a series of secondary assays that included a rapid fluorescence-
based capillary electrophoresis analysis, dose–response testing using
a panel of different RNase H enzymes, and a cell-based anti-HIV
assay.8 In this paper we describe the isolation, structure elucidation,
and biological characterization of a new compound from an extract
of the plant Hylodendron gabunensis Taub. (Fabaceae) that has
potent activity against HIV RNase H.
groups in 1 was revealed by the presence of four singlet aromatic
resonances (δH 7.85, 7.82, 7.80, 7.78) that each integrated for 2H,
four ester carbonyl carbons (δC 167.6, 166.9, 166.9, 166.3), and
four groups of 13C NMR resonances with characteristic chemical
shifts that clustered around δC 147.4, 141.1, 120.6, and 110.3. This
assignment was supported by the observation that each of the
aromatic protons showed HMBC correlations to two phenolic
carbon resonances, a nonprotonated aromatic carbon around δ
120.6, and an ester carbonyl. Further HMBC correlations from H-1
to C-1′, H-3 to C1′′, H-4 to C-1′′′, and H-5 to C-1′′′′ indicated that
the four galloyl groups were connected to C-1, C-3, C-4, and C-5
of the apiitol core.
A 3 g portion of the organic solvent extract of H. gabunensis
was partitioned between CH2Cl2 and H2O, which concentrated the
RNase H inhibitory activity into the H2O-soluble fraction. The
aqueous layer was then extracted with n-BuOH, and the active
n-BuOH-soluble material was sequentially fractionated on ODS and
on a polyamide resin. Final purification was carried out by ODS
HPLC to yield compound 1 as the principal active component. The
molecular formula of 1 was determined to be C33H28O21 by
HRESIMS measurement (obsd [M – H]- m/z 759.1050, calcd for
Compound 1 was hydrolyzed with 2 N NaOH/MeOH (1:1) for
12 h to liberate the free sugar moiety, and the hydrolysate was
then purified by a combination of anion and cation exchange column
chromatography to afford apiitol (2). The 1H and 13C NMR data of
this material showed good consistency with the literature data for
2, which was previously isolated from the plant Torillis japonica.9
The specific rotation of 2 ([R]20D –5.5) that we recovered from the
hydrolysate of 1 indicated that the absolute configuration of C-3
was S (lit. value [R]26D –4.0)9 rather than R (lit. value [R]21D +4.3).9
Thus the central core of 1 consists of D-apiitol, which has previously
been shown to be biosynthesized via reduction of D-apiose (3-C-
hydroxymethyl-aldehydo-D-glycero-tetrose).10 While D-apiose is
widely distributed in plant cell wall polysaccharides,11 compound
1 is the first secondary metabolite reported that contains the reduced
sugar apiitol.
1
C33H27O21 759.1045). Analysis of both the H NMR and HSQC
spectroscopic data obtained in C5D5N showed an oxymethine (δH
6.44/δC 72.7) and three oxymethylene groups (δH 5.31, 5.07/δC 63.7,
5.11, 4.88/65.5, and 4.97, 4.92/65.5). A COSY spectrum revealed
correlations between H-3 and H-4 and between the geminal
methylene protons. HMBC correlations from H-1a, H-1b, H-3,
H-5a, and H-5b to a nonprotonated carbon at δ 74.3 (C-2) implied
that the sugar moiety was apiitol (2). The presence of four galloyl
* To whom correspondence should be addressed. Tel: 301-846-5391.
Fax: 301-846-6851. E-mail: gustafson@ncifcrf.gov.
† Molecular Targets Development Program, NCI.
Compound 1 was tested for in Vitro activity against RNase H
from HIV-1, HIV-2, human, and E. coli using a fluorescence
resonance energy transfer (FRET) based assay.8 It potently inhibited
both HIV-1 and HIV-2 RNase H with IC50 values of 0.24 and 0.13
‡ Basic Research Program, SAIC-Frederick Inc.
§ Drug Resistance Program, CCR, NCI.
Proteomics and Mass Spectrometry Facility, NIDDK, Bethesda, MD.
10.1021/np0702279 CCC: $37.00
2007 American Chemical Society and American Society of Pharmacognosy
Published on Web 10/13/2007

1648 Journal of Natural Products, 2007, Vol. 70, No. 10
Notes
Table 1. NMR Spectroscopic Data (500 MHz C5D5N) for
1,3,4,5-Tetragalloylapiitol (1)
δH
(J in Hz)
position
δC
gCOSY
gHMBC
1a
1b
2
4.88 d, (11.6)
5.11 d, (11.6)
65.5 H1b
H1a
74.3
C1′
C1′
3
6.44 brd, (8.0) 72.7 H4a, H4b C1′′
4a
5.07 dd, (8.0,
12.0)
63.7 H4b, H3
C1′′′
4b
5a
5b
5.31 brd, (12.0)
4.92 d, (10.6)
4.97 d, (10.6)
H4a, H3
65.5 H5b
H5a
C1′′′
C1′′′′
C1′′′′
galloyl 1
1′
2′
3′
4′
Figure 1. COSY and HMBC correlations of 1. The bold line and
the arrows represent a COSY cross-peak and HMBC correlations,
respectively.
166.9
120.6
110.3
147.4
141.1
7.82 sa
7.80 s
7.78 s
7.85 sa
C1′,2′,3′,4′,5′
5′
galloyl 2
1′′
2′′
3′′
4′′
5′′
galloyl 3
1′′′
2′′′
3′′
166.3
120.6
110.3
147.4
141.1
C1′′,2′′,3′′,4′′,5′′
C1′′′,2′′′,3′′′,4′′′,5′′′
C1′′′′,2′′′′,3′′′′,4′′′′,5′′′′
167.6
120.6
110.3
147.4
141.1
4′′′
5′′′
galloyl 4
1′′′′
2′′′′
3′′′′
4′′′′
5′′′′
Figure 2. Dose–response curve for compound 1 tested against
HIV-1 RNase H. Data were collected at a substrate concentration
of 125 nM, over an inhibitor concentration range of 0 to 10 µM in
10 2-fold dilution steps.
166.9
120.6
110.3
147.4
141.1
µM, respectively (Figure 2). Compound 1 showed significantly
weaker inhibition against human RNase H (IC50 ) 1.5 µM) and
no detectable inhibition of the E. coli enzyme at the highest test
concentration (10 µM). To assess whether the observed inhibition
resulted from nonspecific formation of compound–enzyme ag-
gregates, the dose–response experiment was repeated in the presence
of 0.1% Triton X-100 detergent, a known suppressor of small
molecules that work via this mechanism.12 Addition of Triton X-100
to the reaction buffer produced no significant effect on the observed
IC50 or Hill slope for compound 1 (data not shown). This suggests
that the activity of 1 is not due to compound aggregation or some
other nonspecific hydrophobic interaction with RNase H. In a cell-
based assay using virus strain HIV-1RF compound 1 did not afford
any cytoprotective effects at concentrations up to 40 µM.13
a These assignments may be reversed.
evaporated to give 56.5 g of organic extract. A 5 g portion of the extract
was partitioned between H2O and CH2Cl2. The aqueous layer was
extracted with n-BuOH, and the n-BuOH-soluble material was separated
on an ODS column to afford a potent active fraction (eluted with 50%
aqueous MeOH). The active fraction was passed through polyamide
resin with MeOH, and the eluant was purified by reversed-phase C18
HPLC (Cosmosil AR-II i.d. 10 × 250 mm) with a gradient elution
from 10% to 30% aqueous MeOH to give 1 (12.5 mg).
1,3,4,5-Tetragalloylapiitol (1): pale purple solid; [R]20 –18.6 (c
D
0.1, MeOH); UV (MeOH) λmax (log ε) 219 (4.96), 278 (4.58) nm;
HRESIMS [M – H]- m/z 759.1045 (calcd for C33H27O21, 759.1050, ∆
1
-0.5 mmu); H and 13C NMR data, see Table 1.
Hydrolysis of 1. A 5 mg portion of 1 was hydrolyzed with 2 N
NaOH–MeOH (1:1) at room temperature for 12 h. The reaction mixture
was neutralized with HCl and then passed through an anion exchange
column (DEAE Sephadex A-50) and then a cation exchange column
(BIO-RAD AG 50W-X2) to afford 2 in the eluate, which was isolated
as a colorless solid: [R]20D –5.5 (c 0.09, MeOH); 1H NMR (C5D5N) δ
4.61 (H-3), 4.51 (H-4a), 4,45 (H-4b), 4.44 (H-1), 4.40 (H-5); 13C NMR
(C5D5N) δ 76.4 (C-2), 74.9 (C-3), 65.1 (C-5), 64.1 (C-1), 63.3 (C-4).
RNase H FRET Assay. This assay was performed with DMSO
solutions of the chromatography fractions and pure compounds, as
previously described.8 Data were collected at a final substrate concen-
tration of 125 nM (approximately 5 × Km value), over a 12-point
inhibitor concentration range of 0 to 10 µM in 2-fold dilution steps.
DMSO was present at a final concentration of 1% in all of the samples.
The reaction was initiated through the addition of 2 nM enzyme and
allowed to run for up to 60 min at 25 °C. Dose–response rate data
were calculated using GraphPad Prism software version 4.00 for
Windows, GraphPad Software, San Diego CA, www.graphpad.com.
Experimental Section
General Experimental Procedures. Optical rotations were mea-
sured on a Perkin-Elmer 241 polarimeter, and UV spectra were obtained
1
on a Beckman DU640 spectrophotometer. H and 13C spectra and all
2D NMR data were recorded on a Varian INOVA NMR spectrometer
1
at 500 MHz for H and 125 Hz for 13C. Spectra were acquired at 27
°C, and chemical shifts were referenced to the residual solvent
resonances for C5D5N: δH 7.19 and δC 123.3. Electrospray ionization
(ESI) mass spectra were obtained on a Waters LCT Premier time-of-
flight (TOF) mass spectrometer. Fluorescence measurements for the
FRET bioassay were made using a Tecan Safire fluorescence plate
reader.
Plant Material. Leaf samples of Hylodendron gabunensis Taub.
(Fabaceae) were collected in January 1988 in the Southern Bakundu
Reserve Forest, Cameroon. The voucher number for this collection is
Q66R75, and a voucher sample is maintained at the Smithsonian
Institution in Washington, D.C.
Acknowledgment. We thank the Natural Products Branch, NCI, for
coordinating sample collections and T. McCloud (SAIC-Frederick) for
extractions. This research was supported by the Intramural Research
Program of the NIH, National Cancer Institute, Center for Cancer
Research, and by NCI contract N01-CO-12400. The content of this
Isolation of 1,3,4,5-Tetragalloylapiitol (1). The dried H. gabunensis
leaves (349 g dry weight) were ground to a fine powder and extracted
overnight with approximately 1000 mL of CH2Cl2–MeOH (1:1). After
draining off the solvent, the plant material was washed with 100%
MeOH, which was then combined with the earlier solvent mixture and

Notes
Journal of Natural Products, 2007, Vol. 70, No. 10 1649
(4) Hansen, J.; Schulze, T.; Moelling, H. J. Biol. Chem. 1987, 262, 12393–
12396.
(5) Tisdale, M.; Schulze, T.; Larder, B. A; Moelling, K. J. Gen. Virol.
1991, 72, 59–66.
(6) Imamichi, T. Curr. Pharm. Des. 2004, 10, 4039–4053.
(7) Klumpp, K.; Mirzadegan, T. Curr. Pharm. Des. 2006, 12, 1909–12.
publication does not necessarily reflect the views or policies of the
Department of Health and Human Services, nor does mention of trade
names, commercial products, or organizations imply endorsement by
the U.S. government.
Supporting Information Available: 1H, 13C, and 2D NMR spectra
for compound 1. This material is available free of charge via the Internet
at http://pubs.acs.org.
(8) Parniak, M. A.; Min, K.-L.; Budihas, S. R.; Le Grice, S. F. J.; Beutler,
J. A. Anal. Biochem. 2003, 322, 33–39.
(9) Kitajima, J.; Suzuki, N.; Ishikawa, T.; Tanaka, Y. Chem. Pharm. Bull.
1998, 46, 1583–1586.
(10) Neal, D. L.; Kindel, P. K. J. Bacteriol. 1970, 101, 910–915.
(11) Duff, R. B. Biochem. J. 1965, 94, 768–772.
(12) Feng, B. Y.; Shoichet, B. K. Nat. Protocols 2006, 2, 550–553.
References and Notes
(1) Kohlstaedt, L. A.; Wang, J.; Friedman, J. M.; Rice, P. A.; Steitz, T. A.
Science 1992, 256, 1783–1790.
(2) Arnold, E.; Jacobo-Molina, A.; Nanni, R. G.; Williams, R. L.; Lu,
X.; Ding, J. Nature 1992, 357, 85–89.
(3) Jacobo-Molina, A.; Ding, J.; Nanni, R. G.; Lark, A. D.; Lu, X.; Tantillo,
C. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 6320–6324.
(13) Gulakowski, R. J.; McMahon, J. B.; Staley, P. G.; Moran, R. A.; Boyd,
M. R. J. Virol. Methods 1991, 33, 87–100.
NP0702279