Identification of 3',4',5'-trimethoxychalcone analogues as potent inhibitors of Helicobacter pylori-induced inflammation in human gastric epithelial cells

Article abstract of DOI:10.1016/j.bmcl.2010.07.094

Efforts to identify potent small molecule inhibitors of Helicobacter pylori led to the evaluation of 23 3',4',5'-trimethoxychalcone analogues. Some of the compounds displayed potent antibacterial activity against H. pylori. Three most active and selective

Full text of DOI:10.1016/j.bmcl.2010.07.094

Bioorganic & Medicinal Chemistry Letters 20 (2010) 5462–5465
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of 30,40,50-trimethoxychalcone analogues as potent inhibitors of
Helicobacter pylori-induced inflammation in human gastric epithelial cells
Chih-Ho Lai a,b, Yerra Koteswara Rao c, Shih-Hua Fang d, Yu-Ting Sing a,b, Yew-Min Tzeng c,
*
a Department of Microbiology, School of Medicine, China Medical University, Taichung, Taiwan, ROC
b Graduate Institute of Basic Medical Science, China Medical University, Taichung, Taiwan, ROC
c Institute of Biochemical Sciences and Technology, Chaoyang University of Technology, Wufeng, Taiwan, ROC
d Institute of Athletics, National Taiwan College of Physical Education, Taichung, Taiwan, ROC
a r t i c l e i n f o
a b s t r a c t
Article history:
Efforts to identify potent small molecule inhibitors of Helicobacter pylori led to the evaluation of 23
30,40,50-trimethoxychalcone analogues. Some of the compounds displayed potent antibacterial activity
against H. pylori. Three most active and selective compounds 1, 7, and 13 also showed the bactericide
activity against the reference as well as multidrug-resistant strains of H. pylori. Additionally, the afore-
mentioned three compounds potentially inhibited the H. pylori adhesion and invasion to human gastric
epithelial (AGS) cells. Furthermore, these selective compounds inhibited the H. pylori-induced gastric
inflammation by reduced inflammatory mediator’s nuclear factor kappa B activation, and the secretion
of interleukin-8.
Received 21 April 2010
Revised 16 July 2010
Accepted 21 July 2010
Available online 25 July 2010
Keywords:
30,40,50-Trimethoxychalcone analogues
Helicobacter pylori infection
Human gastric epithelial cells
Nuclear factor kappa B
Interleukin-8
Ó 2010 Elsevier Ltd. All rights reserved.
Helicobacter pylori is a microaerophilic Gram-negative bacte-
rium which inhabits approximately one-half of the global human
population.1 H. pylori binds to human gastric epithelial (AGS) cells
in the stomach via bacterial adhesions causes various gastric dis-
eases such as chronic gastritis, peptic ulcer, and gastric cancer.2
H. pylori induces inflammation-associated gene expression in AGS
cells, including activation of nuclear factor kappa B (NF-rB), and
production of inflammatory cytokines including interleukin (IL)-
8.3 IL-8 secret by AGS cells is an important host mediator inducing
neutrophil migration to the site of infection, and consider to be
crucial cytokine in the regulation of inflammation and immune
processes.4 NF-rB plays an important role in expression of IL-8 in
H. pylori-infected gastric inflammation.5 Thus in H. pylori-infected
AGS cells, inhibition of inflammatory mediators NF-rB activation
and IL-8 secretion might be a useful therapeutic strategy in chronic
gastritis.
Treatment for H. pylori stomach infection usually includes two
antibiotics clarithromycin (CLR), and amoxicillin (AMX) or metro-
nidazole (MTZ) that help to kill the bacteria, and one proton pump
inhibitor which causes the stomach to make less acid and help the
ulcer to heal—known as triple therapy. However, the successful
rate of H. pylori eradication with triple therapy has declined from
over 90% to about 60% during the past decade.6 In addition,
increased drug-resistance problems caused by extensive use of
antibiotics, adaptive survival mechanisms of pathogenic bacteria
to counteract currently used antimicrobials, lack of protection
against re-infection, and the necessity of taking a long-term pre-
scription for complete eradication has became a global issue.7 In
addition, the cost of this therapy is significant, and therefore the
development of new effective drugs with high selectivity, mini-
mum side effects and low manufacture costs is urgently required.
The economical, facile, and rapid synthesis of chalcones make
them attractive as potential drug candidates to fight against
H. pylori.8 Although many studies have been reported the wide
variety of pharmacological activities of chalcones include antican-
cer, angiotensin converting enzyme inhibitors, anti-inflammatory,
anti-leishmanial,9 however, there are few reports on chalcones as
inhibitors of H. pylori-induced inflammation.8 Chalcones obtained
from the Claisen–Schmidt condensation reaction between the cor-
responding aromatic aldehydes and acetophenones have a similar
structure to sofalcone, an oral gastrointestinal medication. Their
structural differences exist only in the substituent pattern of two
aromatic rings. We have previously reported the 30,40,50-trimeth-
oxychalcone analogues (TMCs) as potent dual inhibitors of nitric
oxide production and tumor cell proliferation.10 Encouraged by
the aforementioned information and in an attempt to find more
potent compounds against H. pylori-induced gastric inflammation,
herein we report our results which lead to the identification of po-
tent growth inhibitors of reference as well as multidrug-resistant
* Corresponding author. Tel.: +886 4 23323000x4471; fax: +886 4 23395870.
E-mail address: ymtzeng@cyut.edu.tw (Y.-M. Tzeng).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.07.094

C.-H. Lai et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5462–5465
5463
strains of H. pylori; the inhibition of H. pylori adhesion and invasion
to AGS cells; and suppression of inflammatory mediators NF-rB
activation and IL-8 release in H. pylori-infected AGS cells.
intracellular activity reflects the ability of H. pylori to evade antibi-
otic treatment. Based on the previous kinetics of antibiotics AMX
and MTZ were not administered the bactericide effect on intracel-
lular H. pylori.13,14 Although CLR is known to concentrate in the
cells and has antibacterial activity in both extra- or intracellularly,
however, H. pylori can easily develop resistance during exposure to
CLR.15 Additionally, CLR is more expensive (around sevenfold cost)
as compared with AMX and MTZ in the clinical therapy of H. pylori-
related gastrointestinal diseases. Thus, we used lower concentra-
tion of CLR and AMX at 0.05 mg/ml, and higher concentration of
MTZ at 0.8 mg/ml as the positive control, and the tested TMCs
1–23 were at 10-fold concentration of authentic antibiotics. In this
letter, the TMCs 1–23 were examined at 10-fold lower concentra-
tion as compared with positive control (0.5 mg/ml vs 0.05 mg/ml),
that is, generally used in the evaluation of chemical compounds to
examine the H. pylori growth inhibition. At the experimental con-
centrations, chalcones 3, 6, 7, 11, 13, and 23 exhibited potent
inhibitory effect than the three positive controls, while 4, 5, 10,
12, 17, 19, and 21 showed nearly similar inhibition zone range
11–16 mm to that of AMX (Table 1). In contrast, the TMCs 8, 9,
20, and 22 had negligible effect on H. pylori growth inhibition.
Thus, these data indicated that TMCs may useful to develop poten-
tial growth inhibitors against H. pylori.
Next, the effect of TMCs 1–23 on gastric epithelial cell (AGS
cells) viability was analyzed using MTT method.16 At the maximum
concentration tested for H. pylori growth inhibition (0.5 mg/ml),
although some of the TMCs exhibited cytotoxic effect, however,
the potent chalcones 2,3,30,40,50-pentamethoxychalcone (1), 7,
and 30,4,40,50-tetramethoxychalcone (13) had negligible effect on
AGS cell viability. Therefore, the three selective and potent com-
pounds 1, 7, and 13 were chosen for the further experiments. We
subsequently determined the minimum bactericidal concentration
(MBC) for the chalcones 1, 7, 13 and three standard antibiotics
against H. pylori reference strain Hp 26695, and multidrug-resis-
tant strains v1254 and v1354.17 As shown in Table 2, the chalcone
TMCs 1–23 are synthesized in accordance with our established
synthetic route (Table 1), and the details for synthesis and charac-
terization have been described in our previous report.10 In this let-
ter, TMCs 1–23 at a concentration of 0.5 mg/ml were evaluated for
their growth inhibitory effect against H. pylori reference strain Hp
26695 by the disk agar diffusion method.11 As shown in Table 1,
tested compounds showed a wide range of potency with inhibition
zone in a range from 0 to 26 mm. The chalcone 7, which substi-
tuted with a methoxy group in position R2 and a hydroxyl group
in R3 showed the most potent inhibitory effect against H. pylori
growth with inhibition zone of 26 mm. This inhibitory effect was
conserved when these substitutions reversed as in 14 (20 mm). A
similar inhibitory trend as in 7 was observed for chalcones bearing
3,4,5-tri-OMe (3), 3,4-di-OH (11), 4-OH (6), 4-OMe (13), and 4-CHO
(23) in ring-B with inhibition zone of 23, 23, 22, 22, and 22 mm,
respectively. Furthermore, the TMCs substituted in ring-B with
2,3-di-OMe (1), 3-OH, 4-OMe (14), 3-OMe (15), and 4-NO2 (18) also
displayed potent anti-H. pylori activity with a inhibition zone be-
tween 18 and 21 mm (Table 1). At a concentration of 0.05 mg/ml,
the commonly used antibiotic drugs CLR and AMX for treatment
of H. pylori infection inhibited the bacterial growth with inhibition
zone of 21 mm and 14 mm, respectively. Previous studies in the
infection model have revealed that H. pylori could survive and rep-
licate in the intracellular niche.12 The need for an antibiotic with
Table 1
Chemical structures of 30,40,50-trimethoxychalcone analogues 1–23, and their inhib-
itory effect on H. pylori growth using disk agar diffusion test
R1
O
2
R2
H3CO
1
7 showed the highest activity with MBC of 12.5 lM against the ref-
B
A
erence and multidrug-resistant strains of H. pylori. Noticeably, this
chalcone has comparable inhibitory effect against the reference as
well as multidrug-resistant strains. The MBC of chalcone 1 against
R3
H3CO
5
R4
OCH3
Hp 26695, v1254, and v1354 strains was 25.0, 50.0, and 25.0
lM,
Compound
R1
R2
R3
R4
Inhibition
zonea (mm)
respectively, while the same for 13 was 100 M against each strain
l
(Table 2). Noticeably, the MBCs of chalcones 1 and 7 were compa-
rable with standard antibiotics CLR and MTZ (Table 2). These re-
sults suggest that chalcones 1, 7, and 13 were not only having
the bactericide activity against the antibiotic-susceptible H. pylori
strain but also inhibited the multidrug-resistant strains. The re-
sults of this study was in parallel to our previous report,10 that
chalcones 1, 7, and 13 are potent inhibitors of nitric oxide produc-
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
AMX
CLR
MTZ
OMe
H
H
H
OMe
H
H
OMe
OH
OH
H
H
H
H
H
H
H
H
H
OMe
OMe
OMe
H
H
H
OMe
H
H
OMe
OH
OH
H
OH
OMe
H
H
H
H
H
H
CHO
H
H
H
H
OMe
H
OMe
H
H
H
OH
H
H
H
H
H
H
H
H
H
H
21
16
23
12
16
22
26
0
OMe
OMe
N(CH3)2
OMe
OH
OH
OMe
H
H
OH
H
OMe
OMe
H
F
Br
NO2
Me
H
H
H
CHO
0
13
23
11
22
20
19
17
11
18
15
0
14
0
22
14
21
7
Table 2
Minimum bactericidal concentration (MBC) of chalcones (1, 7 and 13) and standard
antimicrobial agents (AMX, CLR and MTZ) against H. pylori reference strain Hp 26695,
and multidrug-resistant strains v1254 and v1354
Chalcone or antibiotica
MBCb
Hp 26695c
v1254c
v1354c
1
7
13
AMX
CLR
MTZ
25.0
12.5
100.0
12.5
1.6
50.0
12.5
100.0
12.5
50.0
400.0
25.0
12.5
100.0
12.5
25.0
800.0
OH
CHO
H
NO2
H
H
H
H
25.0
a
AMX: amoxicillin, CLR: clarithromycin, and MTZ: metronidazole.
b
Concentrations of chalcones 1, 7, and 13, and standard antimicrobial agents
M.
were in
l
AMX: amoxicillin, CLR: clarithromycin, and MTZ: metronidazole.
c
a
Hp 26695 was a reference strain. Strains v1254 and v1354 were clinical isolates,
Concentration of compounds 1–23 was at 0.5 mg/ml and AMX at 0.05 mg/ml,
which show resistant to both metronidazole and clarithromycin
CLR at 0.05 mg/ml and MTZ at 0.8 mg/ml. DMSO was used as a negative control.

5464
C.-H. Lai et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5462–5465
tion from peritoneal macrophage cells.10 In addition, the potential
of chalcones 7 to eradicate H. pylori is also supported by previously
reported antibacterial activity against other bacteria strains,
including Staphylococcus aureus and Escherichia coli.18 To the best
of our knowledge, this was the first report to demonstrate that
the TMCs could significantly inhibited the H. pylori.
H. pylori adhesion to gastric mucosa of epithelial cells is an
important initial event in the pathogenesis of gastric malignancies,
and inhibition of bacterial adhesion may prevent certain pathogen-
related diseases, such as duodenal ulcer and gastric cancer.1 In this
letter, chalcones 1, 7, and 13 were further assayed with regard to
their ability to inhibit the adhesion of H. pylori to AGS cells.11 As
shown in Figure 1A, chalcone 7 displayed most potent anti-adhesion
activity with a reduction of 88.2% and 98.8%, at concentrations of 2.5
centration of 2.5
fect H. pylori invasion into AGS cells, however, treatment with
higher concentrations of 5.0 and 10.0 M resulted a moderate inhi-
bition of bacterial invasion into AGS cells (Fig. 1B). When treated
with AMX and CLR at a concentration of 2.5 M, the inhibitory ef-
fects of H. pylori invasion and adhesion to AGS cells were more than
90%, while only slight inhibitory effect (around 20%) of MTZ was ob-
served (Fig. 1). DMSO vehicle had no effect on adhesion and invasion
assays. Results from this study demonstrated that chalcones 7 fol-
lowed by 1 have the potential to inhibit the adhesion and invasion
of H. pylori into AGS cells.
lM, compounds 1 and 13 did not significantly af-
l
l
We further examine the effects of chalcones 1, 7, and 13 on H.
pylori-induced inflammation by the determination inflammatory
mediators production NF-
the H. pylori-infected AGS cells.17 We transfected AGS cells with
an NF- B reporter construct (NF-rB-luc), and used to determine
luciferase expression following treatment of tested compounds in
a concentration range between 2.5 and 10 M. Treatment of trans-
jB activation and IL-8 secretion from
and 5.0
fect with a reduction of 66.6%, 73.3%, and 99.9% at 2.5, 5.0, and
10.0 M, respectively. On the other hand, it is interesting to observe
that chalcone 13 had a minor inhibitory effect at relatively higher
concentrations 5.0 and 10.0 M, compared to the control (Fig. 1A).
lM, respectively. The chalcone 1 has such anti-adhesion ef-
j
l
l
l
fected AGS cells with chalcones 1, 7, and 13 and H. pylori infection
led to a dose-dependent reduction in the stimulation of luciferase
activity (Fig. 2A). The chalcone 7 displayed a dramatic luciferase
inhibitory activity by 86.1%, 98.6%, and 99.9% at concentrations
In addition, we further assayed chalcones 1, 7, and 13 abilities to in-
hibit the invasion of H. pylori to AGS cells.11 As expected, chalcone 7
showed the highest inhibition of H. pylori invasion into AGS cells by
57.6%, 91.0%, and 100% at the concentrations of 2.5
l
m, 5.0
l
m, and
of 2.5, 5.0, and 10.0
trol DMSO. However, three antimicrobial agents had no effect on
induction of NF- B activity assay at a concentration of 2.5
(Fig. 2A). At the concentrations of 2.5, 5.0, and 10.0 M, chalcones
lM, respectively, compared with that of con-
10.0 M, respectively (Fig. 1B). Comparing the effect of chalcone 7
l
on H. pylori invasion into AGS cells, we observed a similar pattern
of inhibitory effects on adhesion of bacteria to AGS cells. A low con-
j
lM
l
Figure 1. Effect of 30,40,50-trimethoxychalcone analogues 1, 7, and 13 on H. pylori
adhesion (A), and invasion (B), to human gastric epithelial cells. The bacteria to
gastric epithelial cells that was treated or untreated with chalcones 1, 7, and 13
prior infection with H. pylori at a MOI of 50 for 6 h. The standard antibiotics
amoxicillin (AMX), clarithromycin (CLR), and metronidazole (MTZ) were used as
Figure 2. Inhibitory effects of 30,40,50-trimethoxychalcone analogues on H. pylori-
infected AGS cells inflammation. Induction of NF-rB activation (A) and IL-8
secretion (B) were inhibited by treatment of chalcones 1, 7, and 13. DMSO or the
standard antibiotics amoxicillin (AMX), clarithromycin (CLR), and metronidazole
positive control at
a
concentration of 2.5
l
M. Each experiment was shown
(MTZ) at a concentration of 2.5 lM were used as positive control. The luciferase
represent mean values SD of at least six independent experiments. Statistical
significance was calculated using Student’s t-test when compared to DMSO treated
cells. P-values were determined by Student’s t-test. The significant difference was
activity and IL-8 levels in supernatants was determined (see Supplementary data).
Results are shown by mean values SD from at least three independent experi-
*
ments. Statistical significance was evaluated using Student’s t-test with P <0.05;
*
**
**
set at P <0.05; P <0.01.
P <0.01.

C.-H. Lai et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5462–5465
5465
1 and 13 reduced the luciferase activity in the range 39.8–90.6%
and 14.7–66.0%, respectively (Fig. 2A).
and NSC 98-2811-M-324-001), China Medical University
(CMU97-CMC-019, CMU97-346), and by Tomorrow Medical Foun-
dation. We thank Chen-Che Lu and Li-Yu Chen for their expert
technical assistance.
To identify the mechanism whereby chalcones 1, 7, and 13
inhibited NF-rB activation, we further determined the effect of
these compounds on IL-8 production from H. pylori-infected AGS
cells.17 H. pylori-induced IL-8 expression is NF- B dependent.19
j
Supplementary data
The production of IL-8 in H. pylori-infected AGS cells was measured
in the presence or absence of chalcones 1, 7, and 13. In consistent
with NF-rB activity assay, IL-8 production in AGS cells infected with
H. pylori was dramatically reduced when cells were pre-treated
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2010.07.094.
with chalcone 7 (Fig. 2B). Even pretreatment of 2.5 lM inhibited
IL-8 production comparable with that of DMSO or three positive
References
1. Del Giudice, G.; Malfertheiner, P.; Rappuoli, R. Expert Rev. Vaccines 2009, 8,
1037.
2. Hatakeyama, M. J. Gastroenterol. 2009, 44, 239.
3. Nedrud, J. G.; Blanchard, S. S.; Czinn, S. J. Helicobacter 2002, 7, 24.
4. Crabtree, J. E.; Xiang, Z.; Lindley, I. J.; Tompkins, D. S.; Rappuoli, R.; Covacci, A. J.
Clin. Pathol. 1995, 48, 967.
5. Lo, Y. C.; Shih, Y. T.; Wu, D. C.; Lee, Y. C. Inflamm. Res. 2009, 58, 329.
6. Herrera, V.; Parsonnet, J. Clin. Microbiol. Infect. 2009, 15, 971.
7. Gisbert, J. P.; Calvet, X.; O’Connor, J. P.; Megraud, F.; O’Morain, C. A. Expert Opin.
Pharmacother. 2010, 11, 905.
controls treated cells, and the inhibition reached from 81.9% to
99.7% at concentration of 10.0 lM. On the other hand, chalcones 1
and 13 have relative lower IL-8 inhibitory effect with inhibition of
24.6%, 65.2%, and 87.0% and; 8.0, 23.4, and 59.4%, at concentrations
of 2.5, 5.0, and 10.0 lM, respectively (Fig. 2B). The results from our
study indicated that reduction of IL-8 by chalcones 1, 7, and 13 pre-
treatment might contribute to attenuate the NF-rB activity by AGS
cells in response to H. pylori infection.2 To the best of our knowl-
edge, we demonstrate here for the first time on chalcones 1, 7,
and 13 inhibited H. pylori-associated inflammation in AGS cells.
In conclusion, we have screened 23 30,40,50-trimethoxychalcone
analogues against H. pylori, and found some of them displayed po-
tent anti-H. pylori activity. The chalcones bearing 2,3-di-OMe (1),
3-OMe, 4-OH (7), and 4-OMe (13) in ring-B displayed potent
growth inhibition on reference as well as multidrug-resistant
strains of H. pylori. Additionally, the chalcones 1, 7, and 13 inhib-
ited the adhesion and invasion of H. pylori to AGS cells. In addition
8. Isomoto, H.; Furusu, H.; Ohnita, K.; Wen, C. Y.; Inoue, K.; Kohno, S. World J.
Gastroenterol. 2005, 11, 1629.
9. (a) Rao, Y. K.; Fang, S. H.; Tzeng, Y. M. Bioorg. Med. Chem. 2004, 12, 2679; (b)
Bonesi, M.; Loizzo, M. R.; Statti, G. A.; Michel, S.; Tillequin, F.; Menichini, F. Bioorg.
Med. Chem. Lett. 2010, 20, 1990; (c) Bandgar, B. P.; Patil, S. A.; Gacche, R. N.;
Korbad, B. L.; Hote, B. S.; Kinkar, S. N.; Jalde, S. S. Bioorg. Med. Chem. Lett. 2010, 20,
730; (d) Bandgar, B. P.; Gawande, S. S.; Bodade, R. G.; Totre, J. V.; Khobragade, C. N.
Bioorg. Med. Chem. 2010, 18, 1364; (e) Aponte, J. C.; Castillo, D.; Estevez, Y.;
Gonzalez, G.; Arevalo, J.; Hammond, G. B.; Sauvain, M. Bioorg. Med. Chem. Lett.
2010, 20, 100.
10. Rao, Y. K.; Fang, S. H.; Tzeng, Y. M. Bioorg. Med. Chem. 2009, 17, 7909.
11. Geethangili, M.; Fang, S. H.; Lai, C. H.; Rao, Y. K.; Lien, H. M.; Tzeng, Y. M. Food
Chem. 2010, 119, 149.
12. Bjorkholm, B.; Zhukhovitsky, V.; Lofman, C.; Hulten, K.; Enroth, H.; Block, M.;
Rigo, R.; Falk, P.; Engstrand, L. Helicobacter 2000, 5, 148.
13. Hulten, K.; Gibreel, A.; Skold, O.; Engstrand, L. Antimicrob. Agents Chemother.
1997, 41, 2550.
14. Matysiak-Budnik, T.; Heyman, M.; Candalh, C.; Lethuaire, D.; Megraud, F. J.
Antimicrob. Chemother. 2002, 50, 865.
15. Piccolomini, R.; Di Bonaventura, G.; Picciani, C.; Laterza, F.; Vecchiet, J.; Neri, M.
Antimicrob. Agents Chemother. 2001, 45, 1568.
these three chalcones diminish the activation of NF-jB, and secre-
tion of IL-8 in H. pylori-infected AGS cells. Therefore, the most po-
tential of chalcone 7 to inhibit H. pylori and its associated
inflammation, and the chemical nature of this low-molecular-mass
compound that translates to lower cost of synthesis compared to
cationic antibacterial peptides suggest that it has potential to de-
velop new therapeutic drug for the prevention of H. pylori-related
diseases.
16. Rao, Y. K.; Fang, S. H.; Tzeng, Y. M. Bioorg. Med. Chem. 2005, 13, 6850.
17. Lai, C. H.; Fang, S. H.; Rao, Y. K.; Geethangili, M.; Tang, C. H.; Lin, Y. J.; Hung, C. H.;
Wang, W. C.; Tzeng, Y. M. J. Ethnopharmacol. 2008, 118, 522.
18. Batovska, D.; Parushev, S.; Stamboliyska, B.; Tsvetkova, I.; Ninova, M.;
Najdenski, H. Eur. J. Med. Chem. 2009, 44, 2211.
Acknowledgements
19. Aihara, M.; Tsuchimoto, D.; Takizawa, H.; Azuma, A.; Wakebe, H.; Ohmoto, Y.;
Imagawa, K.; Kikuchi, M.; Mukaida, N.; Matsushima, K. Infect. Immun. 1997, 65,
3218.
This study was supported by National Science Council of Taiwan
(NSC 96-2113-M-324-002-MY3, NSC 97-2313-B-039-003-MY3,