|
|
||||||||
B Is Decisively Involved in the Expression of IL-91



*
Institute of Immunology, Johannes Gutenberg University, Mainz, Germany;
Gesellschaft für Strahlen und Umweltforschung-National Research Center for Environment and Health, Institute of Clinical Molecular Biology and Tumor Genetics, Munich, Germany; and
Institute of Pathology, University of Wurzburg, Wurzburg, Germany
| Abstract |
|---|
|
|
|---|
B, which is shown by an NF-
B-driven
reporter gene construct. In the presence of an inhibitor of NF-
B
activation, the production of IL-9 is strongly decreased, whereas the
expression of IL-13 is hardly reduced, and that of IL-4 is not affected
at all. NF-
B drives the expression of IL-9 via three NF-
B binding
sites within the IL-9 promoter, which we characterize using gel shift
analyses and reporter gene assays. In the light of recent reports that
strongly support critical roles for IL-9 and IL-13 in allergic lung
inflammation, our results emphasize the potential clinical importance
of LPS as an enhancer of mast cell-derived IL-9 and IL-13 production in
the course of inflammatory reactions and allergic
diseases. | Introduction |
|---|
|
|
|---|
mRNA
(6). Besides their ability to secrete mediators such as histamine, leukotrienes, and PGs, which directly mediate inflammatory reactions, mast cells can also produce a variety of cytokines that partly overlap with the cytokine pattern produced by Th cells of the Th2 subset (1).
IL-9, originally termed P40, T cell growth factor III, or mast cell growth-enhancing activity (7, 8, 9), is a multifunctional cytokine produced by activated Th cells (10) and activated mast cells (11). Although initially described as a T cell growth factor, naive T cells do not respond to IL-9, but recent data suggest that it is a critical factor for the early stages of intrathymic T cell maturation from precursor cells (12). It was also reported that IL-9 exerts growth-enhancing activity on bone marrow-derived murine mast cell (BMMC) lines (13, 14). Subsequently, it has been demonstrated that elevated levels of IL-9 in vivo lead to pronounced mastocytosis, which enhances the resistance to infections with nematodes (15, 16, 17).
Among other biological activities (10), it was recently suggested on the basis of genetic linkage analyses that IL-9 might play an important role in the pathogenesis of asthma (18, 19). This assumption was supported by the lung-specific expression of IL-9 in transgenic mice, which was accompanied by airway inflammation, bronchial hyper-responsiveness, and mast cell hyperplasia (20) as well as the increased expression of IL-9 and its receptor in human patients with atopic asthma (21).
IL-13, originally designated P600, was initially described as
a cytokine, which is produced by activated Th2 cells, but its
expression by activated mast cells has also been reported
(22, 23, 24, 25). By using the IL-4R
chain and STAT6 for
signaling, IL-13 shares some biological activities with IL-4, such as
the promotion of human B cell growth and the switching of B
cells to the IgE isotype (26). Furthermore, it has been
shown that IL-13, like IL-9, is a critical mediator in experimental
models of allergic asthma (27, 28). The selective
neutralization of IL-13 ameliorated asthma symptoms, including airway
hyper-responsiveness, eosinophil recruitment, and mucus
overproduction. These findings were further corroborated by the
pulmonary expression of IL-13 (29).
In this paper we report that LPS acts as a strong costimulator for the
production of IL-9 and IL-13 by primary BMMC activated with either
ionomycin or cross-linked IgE, whereas the production of IL-4 remains
unaffected. LPS treatment of BMMC leads to an increased activation of
NF-
B, as demonstrated using an NF-
B-dependent reporter gene
construct. Gel shift analyses and reporter gene assays reveal the
presence of three binding sites for NF-
B within the IL-9
promoter.
| Materials and Methods |
|---|
|
|
|---|
BALB/cAnn mice were originally obtained from the Zentralinstitut für Versuchstierforschung (Hannover, Germany), bred in our animal facility, and used at 510 wk of age. BALB/c congenic mice, which express the LPSd allele from the LPS hyporesponsive strain C3H/HeJ (30), were provided by Chris Galanos (Max Planck Institut für Immunbiologie, Freiburg, Germany).
Cytokines, cytokine assays, mAbs, and reagents
Mouse IL-9 (mIL-9) was assayed by a specific sandwich ELISA with
reference standard curves using known amounts of mIL-9. To detect
mIL-9, we used mAb 229.4 and biotinylated mAb D9302.C12 (provided by J.
Van Snick, Ludwig Institute for Cancer Research, Brussels, Belgium).
This ELISA detects biologically active mIL-9, as confirmed using an
IL-9-specific bioassay (8). The detection limit of this
ELISA is 50100 pg/ml. Human rIL-1
(lot 693-98) was provided by
Seiler (Behringwerke, Marburg, Germany).
Murine IL-3 was isolated from supernatants of myelomonocytic WEHI-3B
cells using DEAE chromatography. The cDNA of His-tagged murine KL
(provided by G. W. Bornkamm (31)) was expressed in
Escherichia coli, and KL was affinity purified using the
QIAexpress system, according to the manufacturers directions (Qiagen,
Dusseldorf, Germany). The biological activity of KL was verified using
a proliferation assay measuring [3H]thymidine
uptake by mast cells. Murine rIL-4 was a gift from W.
Müller (Department of Experimental Immunology,
Braunschweig,Germany). For the detection of IL-4 via ELISA, we used
mAbs BVD4-1D11 and BVD6-24G2, which were gifts from A. OGarra (DNAX
Research Institute, Palo Alto, CA), and 1 U/ml of IL-4 corresponds to
100 pg/ml (standard from R&D Systems, Minneapolis, MN). IL-13 was
detected by ELISA with MAP413 and BAF413 using known amounts of rmIL-13
(R&D Systems). Pyrrolidinedithiocarbamate (PDTC), an antioxidant
inhibitor of NF-
B activation, LPS (E. coli serotype
055:B5), ionomycin, and polymyxin B were purchased from Sigma
(Steinheim, Germany). LPS from Salmonella abortusequi was
donated by Chris Galanos (Max Planck Institut für
Immunbiologie).
Generation of BMMC
The mice were sacrificed by cervical dislocation, intact femurs and tibias were removed, and bone marrow cells were harvested by repeated flushing with MEM.
The cell culture was established at a density of 3 x 106 cells/ml in IMDM, supplemented with 10% FCS (inactivated at 56°C), 2 mM L-glutamine, 1 mM pyruvate, 100 U/ml penicillin, 100 µg/ml streptomycin, 20 U/ml mIL-3, and 50 U/ml mIL-4. Alternatively, cells were grown in the presence of IL-3 alone or in combination with additional KL (100 ng/ml). Nonadherent cells were transferred to fresh culture plates every 23 days for a total of at least 21 days to remove adherent macrophages and fibroblasts. FACS analyses using an anti-CD13 Ab (R3-242, PharMingen, San Diego, CA) (32) and IgE plus anti-IgE mAb (33, 34, 35) as well as May-Grünwald-Giemsa and toluidine blue staining revealed that the resulting cell population consisted of >99% BMMC (data not shown).
In vitro cell stimulation
Culture medium was IMDM supplemented with 5% FCS (previously inactivated at 56°C), 1 mM pyruvate, 2 mM L-glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin. Stimulations were conducted in triplicate using 96-well plates with 105 cells/well in a final volume of 200 µl, including 0.5 µM ionomycin alone or in combination with LPS, IL-1, PDTC, or polymyxin B as indicated in the figure legends.
For stimulation of BMMC via their Fc
receptors, mast cells (5
x 105/well) were incubated with the
IgE-anti-DNP Ab A2 (33, 34) for 4872 h and
subsequently cultured in 48-well plates previously coated with DNP-BSA
(2 µg/ml in PBS). After 48 h the supernatants were tested for
the presence of IL-9 by ELISA.
Plasmids, transfection, and reporter gene assays
The 5'-region of the murine IL-9 gene (36)
encompassing nucleotides -610 to +32 was amplified from genomic DNA by
PCR with 5'-CCg gat ccT CAA GGC CAA TGC TAG C-3' and 5'-GTG Taa gct tGA
CGG GAG TCT GGA ACT C-3' as primers, and this was verified by DNA
sequencing. Lowercase letters indicate the authentic BamHI
site and the artificial HindIII site, which was introduced
to allow forced cloning into the promoterless pGL3 Basic luciferase
reporter gene vector (Promega, Mannheim, Germany). Mutageneses of the
three potential NF-
B sites was performed using the QuickChange
site-directed mutagenesis kit (Stratagene, Amsterdam Zuidoost, The
Netherlands) and was verified by DNA sequencing (Table I
).
|
B), provided by Ralf
Marienfeld (Institute of Molecular Pathology, Wurzburg, Germany),
comprises four copies of an NF-
B binding site from the murine c-Myb
intronic enhancer (37). BMMC (2 x
106 cells in 0.2 ml of serum-free IMDM) were
transfected with 10 µg of pTATALUC+(4xNF-
B)
or the same amount of pTATALUC+ as
a control (38). For transfection of the
mIL-9 promoter construct or the derived NF-
B mutants 20 µg of the
respective plasmid was used. Transfections were performed by
electroporation in 0.2-cm cuvettes at room temperature using a
Bio-Rad Gene Pulser (Hercules, CA) set at 350 V, R =
, and 960
µF. Cells were allowed to recover for 2 h in IMDM supplemented
with FCS, glutamine, and pyruvate (described above), then were
harvested, washed with IMDM, and stimulated under different
conditions as outlined in the figure legends. To exclude differences in
transfection efficiency, cells were cotransfected with 200 ng of pRL-TK
(Promega), which contains the thymidine kinase promoter region upstream
of the Renilla reniformis luciferase. Cells were lysed after
24 h, and luciferase activity was measured by a luminometer
(Berthold, Germany) using the dual luciferase reporter assay system
from Promega. Data were standardized according to the
Renilla luciferase activity. EMSA
EMSAs were performed as described previously (39)
using 5 µg of nuclear protein and 0.7 µg of poly(dI-dC). For
competition, 30 ng of an oligonucleotide containing the NF-
B site
5'-GGGAATTTCC-3' from the mouse invariant chain MHC class II
promoter was added (40). In supershiftEMSAs, 1 µg of
Abs specific for NF-
B p65, c-Rel, RelB, p50, and p52 were
added to the incubation mixture (Santa Cruz Biotechnology, Santa
Cruz, CA).
| Results |
|---|
|
|
|---|
BMMC were generated using IL-3 in combination with IL-4 as
described in Materials and Methods according to a recent
paper (41). Activation of these BMMC by the
Ca2+ ionophore ionomycin induced a low, but
significant, production of IL-9 as described recently (11)
as well as moderate levels of IL-13. Fig. 1
demonstrates that addition of LPS
strongly augmented the production of both IL-9 and IL-13 in a
dose-dependent fashion, with 10 ng/ml LPS already exhibiting a
significant effect.5 The
addition of polymyxin B, a cationic peptide capable of neutralizing LPS
(42), significantly reduced the costimulatory function of
LPS. In contrast to IL-9 and IL-13, the production of IL-4, which was
induced by ionomycin, remained unaffected regardless of whether LPS or
LPS plus polymyxin B were present. Also, at lower doses of ionomycin,
leading to a decreased production of IL-4, no effect of LPS on the
expression of IL-4 could be observed (data not shown). With LPS alone
no cytokine production was detected, indicating the requirement for
Ca2+ signaling as a primary stimulus.
|
|
mRNA was
stimulated by LPS (6) suggests that LPS acts on the
production of IL-9 and/or IL-13 via endogenously induced IL-1
.
To exclude this possibility, we stimulated BMMC with ionomycin
plus LPS in the presence of either human rIL-1R antagonist as described
recently (11) or a combination of neutralizing polyclonal
anti-IL-1
and anti-IL-1
Abs. Under these conditions,
compared with the respective controls, we could not detect any
significant decrease in cytokine production (data not shown).
To further examine whether the costimulatory effect of LPS can
also be observed under the conditions of a physiological stimulation
via the high affinity Fc
receptor, BMMC were activated by
cross-linked IgE. Without additional costimuli only low amounts of IL-9
and IL-13 were detectable, but the addition of LPS strongly enhanced
the production of both cytokines (Fig. 3
).
|
|
Expression of IL-9 depends on the activation of NF-
B
A well-documented signaling pathway for LPS leads to the
activation of NF-
B. To examine whether mast cells also respond in
this way to LPS, we measured the activity of an NF-
B-driven reporter
gene, pTATALUC+(4xNF-
B). It comprises the
luciferase gene under the control of a core promoter encompassing a
TATA box and four NF-
B binding sites. BMMC were transfected with
this construct via electroporation and stimulated as detailed in Fig. 5
. Activation solely with LPS had only a
marginal effect on the luciferase gene expression, whereas stimulation
with ionomycin led to a significant activity of the reporter gene.
Activation of BMMC with ionomycin and LPS further increased luciferase
activity
5-fold, indicating a strong coactivation of NF-
B by LPS
in mast cells. These results could be confirmed using cells stimulated
with IgE-Ag complexes after transfection with the NF-
B-dependent
reporter gene (Fig. 6
). To elucidate
whether the enhanced production of IL-9 and/or IL-13 depends on the
activation of NF-
B, BMMC were stimulated with ionomycin and
ionomycin plus LPS in the presence of PDTC, an inhibitor of I
B-
phosphorylation (43), and the levels of IL-9, IL-13, and
IL-4 were determined by ELISA. Fig. 7
depicts the strong dose-dependent inhibition of IL-9 production by
PDTC, suggesting an important role for NF-
B regarding activation of
the IL-9 gene. The fact that PDTC also inhibits the ionomycin-induced
production of IL-9 might be explained by a low level activation of
NF-
B by ionomycin, which is also obvious in Fig. 5
. In contrast to
IL-9, the production of IL-13 is hardly reduced, and the level of IL-4
remains unaffected in the presence of the inhibitor. These data are
in good agreement with the recent finding that the expression of mast
cell-derived IL-4 is NF-
B independent (44). It should
be noted that PDTC exhibited toxic effects, leading to impaired cell
viability at concentrations >100 nM.
|
|
|
B
We have recently shown that the 5' region of the IL-9
gene from -610 to +32 is sufficient for the inducible activation of
the IL-9 gene (41). In the human T cell line C5 MJ2,
transformed with leukemia virus type I, an NF-
B site at -59 to -50
was shown to be involved in basal and PHA-induced expression of the
IL-9 gene (45). Using the TRANSFAC database
(www.transfac.gbf-braunschweig.de), we identified three potential
binding sites for NF-
B within the murine IL-9 promoter, termed
NF-
B13. To examine whether all potential NF-
B sites contribute
to transcription of the IL-9 gene, we introduced point mutations into
each site and compared the activities of these mutants with
that of the wild-type sequence using reporter gene assays of
transfected BMMC stimulated with ionomycin and LPS. Point mutations
were carefully chosen to avoid the generation of potential binding
sites for other transcription factors, based on the TRANSFAC database.
Fig. 8
demonstrates that mutation of
NF-
B1 (-429 to -420, antisense strand) and NF-
B3 (-47 to -38,
sense strand) reduced reporter gene activity
50% in each case,
which fits with the observation that both binding sites comprise
identical sequences. Mutation of NF-
B2 (-167 to -158, sense
strand) also resulted in significant and reproducibly lower luciferase
activity, although the reduction was only in the range of 25% compared
with that in the wild type. NF-
B2 represents a predicted high score
binding site for NF-
B, which differs in sequence from the identical
sites NF-
B1 and NF-
B3 (Table I
). Mutation of all three NF-
B
binding sites does not ablate the activity of the IL-9 promoter,
indicating that other factors must be involved in the inducible
expression of the IL-9 gene.
|
B family members to the IL-9 promoter could be
confirmed by EMSAs using double-stranded oligonucleotides corresponding
to the three putative binding sites (Table I
B binding
site and using the mutated sequences of all three NF-
B sites.
Interestingly, the sequence element NF-
B2 displays the lowest
affinity for NF-
B, which is in good agreement with the finding that
the corresponding mutant reduced the activity of the IL-9 promoter to a
lesser extent compared with the mutations of NF-
B1 and NF-
B3 (see
Fig. 8
B1 from
the IL-9 promoter indicate the binding of p50, p65, and
c-Rel; the binding of p65 is only detectable upon strong
stimulation with ionomycin and LPS (Fig. 9
|
| Discussion |
|---|
|
|
|---|
B (49). Moreover, activation of mitogen-activated
protein kinases and AP-1 has been reported (50, 51, 52). The
latter might account for our observation that LPS enhances the
expression of IL-13 independently of NF-
B.
The strong up-regulation of mast cell-derived IL-9 and IL-13 by LPS
suggests a potential role for these cytokines at sites of inflammation.
Of special interest might be the earlier finding that IL-9 serves as a
growth factor for mast cells, thereby possibly supporting their local
proliferation via an autocrine loop. The growth-promoting activity of
IL-9 has also been reported for the B-1 subpopulation of B cells, which
are able to produce Abs against common bacterial Ags without
Ag-specific T cell help (53). In addition, mast cells
might be decisively involved in the recruitment of eosinophils at sites
of inflammation through the secretion of IL-9, IL-13, and IL-5. In this
context it has been shown that IL-9 induces the expression of the IL-5R
-chain on eosinophils (54) and that IL-13 induces
eotaxin expression by several cell types, including airway epithelial
cells (55, 56, 57, 58). Furthermore, it has been demonstrated that
IL-9 induces the expression of CC chemokines, also acting as eosinophil
chemotactic factors, by epithelial cells (59). We propose
that LPS could be a powerful coactivating stimulus for mast cells in
the course of inflammatory reactions and infectious diseases caused by
Gram-negative bacteria, often following viral infections.
Interestingly, inflammatory and bronchial obstructive responses to
inhalation of endotoxin in asthmatic patients have been described, and
it has been supposed that bacterial Ags can potentiate the action of
inhalant Ags (60, 61, 62). In asthmatics, extravasation of LPS
binding protein and soluble CD14 into the bronchoalveolar compartment
after allergen inhalation has also been reported, implying a role for
LPS in amplification of the inflammatory response
(63, 64, 65). During late asthmatic reactions, an accumulation
of mast cells occurs in lung in parallel with an increased IgE level in
serum that promotes IgE-mediated activation of mast cells.
In the light of these findings, our observation that LPS enhances the
production of IL-9 and IL-13 by mast cells activated with cross-linked
IgE might be of pathophysiological importance, because it has been
shown that both cytokines are central mediators of allergic asthma.
Furthermore, the strong dependence of IL-9 expression on the activation
of NF-
B might partly account for the inability of NF-
B-deficient
mice to develop airway hyper-responsiveness and allergic pulmonary
inflammation (66).
| Acknowledgments |
|---|
| Footnotes |
|---|
2 M.S. and C.M. contributed equally to this work. ![]()
3 Address correspondence and reprint requests to Dr. Edgar Schmitt, Institut für Immunologie, Hochhaus am Augustusplatz, D-55101 Mainz, Germany. ![]()
4 Abbreviations used in this paper: KL, kit ligand; BMMC, bone marrow-derived murine mast cells; mIL-9, mouse IL-9; LPSd, LPS nonresponder; PDTC, pyrrolidinedithiocarbamate; TLR, Toll-like receptor. ![]()
5 LPS derived from Escherichia coli was used throughout this study, although LPS from Salmonella abortusequi proved to be equally effective. ![]()
Received for publication July 7, 2000. Accepted for publication January 24, 2001.
| References |
|---|
|
|
|---|
receptor I crosslinking in cultured mast cells. Cell. Immunol. 185:146.[Medline]
B (p50/relB) binding to a c-myb intronic enhancer correlates with c-myb up-regulation and inhibition of erythroleukemia cell differentiation. Oncogene 15:1859.[Medline]
B-like factor to a promoter element. Mol. Cell Biol. 10:4146.
B and stabilizes a newly phosphorylated form of I
B-
that is still bound to NF-
B. EMBO J. 13:5433.[Medline]
B activity. J. Allergy Clin. Immunol. 105:500.[Medline]
B and AP-1 in human astrocytoma U373 cells. Biochem. Biophys. Acta 232:568.
B) which are involved in IL-6 regulation. Leukemia 10:1308.[Medline]
and IL-4 or IL-13. Invest. Ophthalmol. Vis. Sci. 41:1448.
synergistically induce eotaxin production in human nasal fibroblasts. Clin. Exp. Allergy 30:348.[Medline]
B/Rel transcription factors: c-Rel promotes airway hyperresponsiveness and allergic pulmonary inflammation. J. Immunol. 163:6827.This article has been cited by other articles:
![]() |
G. P. Katsoulotos, M. Qi, J. C. Qi, K. Tanaka, W. E. Hughes, T. J. Molloy, R. Adachi, R. L. Stevens, and S. A. Krilis The Diacylglycerol-dependent Translocation of Ras Guanine Nucleotide-releasing Protein 4 inside a Human Mast Cell Line Results in Substantial Phenotypic Changes, Including Expression of Interleukin 13 Receptor {alpha}2 J. Biol. Chem., January 18, 2008; 283(3): 1610 - 1621. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Wen, Y. Nong, J. G. Morgan, P. Gangurde, A. Bielecki, J. DaSilva, M. Keaveney, H. Cheng, C. Fraser, L. Schopf, et al. A Selective Small Molecule I{kappa}B Kinase beta Inhibitor Blocks Nuclear Factor {kappa}B-Mediated Inflammatory Responses in Human Fibroblast-Like Synoviocytes, Chondrocytes, and Mast Cells J. Pharmacol. Exp. Ther., June 1, 2006; 317(3): 989 - 1001. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Tesfaigzi Roles of Apoptosis in Airway Epithelia Am. J. Respir. Cell Mol. Biol., May 1, 2006; 34(5): 537 - 547. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. B. Corbeil, C. M. Campero, J. C. Rhyan, M. L. Anderson, L. J. Gershwin, D. W. Agnew, L. Munson, and R. H. BonDurant Uterine Mast Cells and Immunoglobulin-E Antibody Responses During Clearance of Tritrichomonas foetus Vet. Pathol., May 1, 2005; 42(3): 282 - 290. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Tanaka, M. Konno, S. Muto, N. Kambe, E. Morii, T. Nakahata, A. Itai, and H. Matsuda A novel NF-{kappa}B inhibitor, IMD-0354, suppresses neoplastic proliferation of human mast cells with constitutively activated c-kit receptors Blood, March 15, 2005; 105(6): 2324 - 2331. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Taube, X. Wei, C. H. Swasey, A. Joetham, S. Zarini, T. Lively, K. Takeda, J. Loader, N. Miyahara, T. Kodama, et al. Mast Cells, Fc{epsilon}RI, and IL-13 Are Required for Development of Airway Hyperresponsiveness after Aerosolized Allergen Exposure in the Absence of Adjuvant J. Immunol., May 15, 2004; 172(10): 6398 - 6406. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. E. Foster, K. Gott, M. R. Schuyler, W. Kozak, and Y. Tesfaigzi LPS-induced neutrophilic inflammation and Bcl-2 expression in metaplastic mucous cells Am J Physiol Lung Cell Mol Physiol, August 1, 2003; 285(2): L405 - L414. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Masuda, Y. Yoshikai, K. Aiba, and T. Matsuguchi Th2 Cytokine Production from Mast Cells Is Directly Induced by Lipopolysaccharide and Distinctly Regulated by c-Jun N-Terminal Kinase and p38 Pathways J. Immunol., October 1, 2002; 169(7): 3801 - 3810. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. E. Applequist, R. P. A. Wallin, and H.-G. Ljunggren Variable expression of Toll-like receptor in murine innate and adaptive immune cell lines Int. Immunol., September 1, 2002; 14(9): 1065 - 1074. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Caamano and C. A. Hunter NF-{kappa}B Family of Transcription Factors: Central Regulators of Innate and Adaptive Immune Functions Clin. Microbiol. Rev., July 1, 2002; 15(3): 414 - 429. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Speirs, J. Caamano, M. H. Goldschmidt, C. A. Hunter, and P. Scott NF-{kappa}B2 Is Required for Optimal CD40-Induced IL-12 Production but Dispensable for Th1 Cell Differentiation J. Immunol., May 1, 2002; 168(9): 4406 - 4413. [Abstract] [Full Text] [PDF] |
||||