Previous Article | Next Article ![]()
Molecular and Cellular Biology, February 2004, p. 997-1006, Vol. 24, No. 3
0270-7306/04/$08.00+0 DOI: 10.1128/MCB.24.3.997-1006.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Departments of Molecular Oncology,1 Immunology,2 Physiology,3 Pathology, Genentech Inc., South San Francisco, California4
Received 27 August 2003/ Returned for modification 9 October 2003/ Accepted 30 October 2003
|
|
|---|
|
|
|---|
, which is expressed only as a soluble molecule, TNF family members are expressed as cell surface proteins acting in a juxtacrine and autocrine manner. Proteolytic processing of some of the ligands generates their corresponding soluble forms. The majority of proteins of the TNF receptor family are composed of type I transmembrane molecules. Many of these receptors also exist in soluble forms generated by proteolytic cleavage of the cell surface protein or transcribed by alternative splicing mechanisms from the genes encoding the full-length receptors. The ligand-binding motif of the TNF receptor family consists of tandem cysteine-rich domains of about 40 amino acids in length. Each cysteine-rich domain contains several cysteines (typically six) and certain other residues in conserved positions. APRIL (a proliferation-inducing ligand, also known as TRDL-1, TALL-2 [12, 35], and TNFSF13A) is a member of the TNF family that has been shown to be capable of inducing the proliferation of certain tumor cell lines in vitro and in vivo (9). Together with a related member of the TNF family, BLyS (B-lymphocyte stimulator, also known as BAFF, TALL-1, zTNF4, THANK, and TNSF13B) (22, 23, 32, 35), APRIL shares two common receptors, TACI and BCMA (21, 29, 40, 45). However, unlike APRIL, BLyS also binds to BR3 (BLyS receptor 3 or BAFF-R), the least-conserved member of the TNF receptor family (39, 48). Both APRIL and BLyS are expressed by macrophages, monocytes, dendritic cells, and T cells (25, 32, 35, 37). Both ligands exist in cell surface as well as soluble forms.
Like most other TNF family members, soluble BLyS is created by cleavage of a transmembrane cell surface protein (18, 22, 32). In contrast, soluble APRIL is produced in the Golgi apparatus within the cell by a furin convertase (16). Furthermore, the transmembrane form of APRIL (named TWE-PRIL) is an unusual fusion product of two alternatively spliced RNAs, composed of exons encoding intracellular and transmembrane domains from the neighboring family member TWEAK [also called Apo3L or TNFSF12 (4, 20)] and exons from APRIL encoding the extracellular part of the molecule (28). BCMA, TACI, and BR3 are type III transmembrane proteins, lacking N-terminal signal sequences. BCMA and TACI contain intracellular TRAF binding motifs (reviewed in reference 17). The signaling mechanisms of these receptors are not fully characterized; however, they activate the NF-
B and mitogen-activated protein kinase pathways (reviewed in reference 17). All three receptors are expressed on B cells, while TACI and BR3 are also detected on the surface of some T cells (14, 39, 41, 46, 48).
While several reports document direct involvement of BLyS, TACI, and BR3 in regulating the development and function of B cells in vivo (reviewed in reference 17), the role of APRIL in immune regulation is not well defined. Alteration in the expression of BLyS or BR3 in the mouse (by gene knockout or naturally occurring mutation, respectively) leads to diminished numbers of mature B cells due to a block at the T1 stage of development (7, 31, 40, 49). In contrast, knockout of TACI results in accumulation of B cells, particularly pronounced in older mice with homogeneous genetic background (34). Elevated levels of BLyS in transgenic mice upregulate B-cell activity, leading to the development of a lupus-like autoimmune disorder (8, 13, 18). Humans with severe B-cell disorders or immunodeficiency virus infection have elevated serum levels of BLyS (3, 6, 38, 52).
The role of APRIL in the immune system was previously investigated by analysis of APRIL transgenic mice (37). These mice display augmented T-cell-independent B-cell responses and increased survival of T cells. To assess whether APRIL is important for embryonic development and postnatal immune function, we deleted the mouse APRIL gene by homologous recombination.
|
|
|---|
![]() View larger version (39K): [in a new window] |
FIG. 1. Targeting of the mouse APRIL gene. (A) Structure of the mouse APRIL genomic locus. Boxes correspond to the genomic regions containing the TWEAK (white bars), APRIL (black bars), and SMT3IP1 (grey bars) genes. The orientations of the three genes are marked by arrows. (B) Schematic representation of the targeting construct designed to replace part of the first exon and all downstream exons of the APRIL gene with a neo cassette. (C) Structure of the mutated region in the APRIL gene. The positions of the 5' and 3' external probes used for Southern blot analysis of ES cell are indicated by bars. The positions of the primer sets used for genotype analysis of mouse tail DNA are indicated by black (external) and grey (internal) arrowheads. (D) Southern blot analysis of recombination of the APRIL gene. Analysis of BsmI (DI) and SpeI (DII) digested DNA derived from several ES cell clones. DNA was digested and fractionated on a 0.7% agarose gel, blotted onto a nylon membrane, and hybridized with 5' (DI) and 3' (DII) probes. (E) Genotyping of APRIL-/- mice by PCR. Tail-derived genomic DNA was subjected to PCR amplification with nested external and internal sets of primers to visualize wild-type and deletion mutant APRIL genes as 3.2-kb and 3.0-kb fragments, respectively. (F) Expression of APRIL in total splenocytes derived from APRIL+/+ and APRIL-/- mice was determined by fluorescence-activated cell sorting with anti-mouse APRIL monoclonal antibody (black) or an isotype-matched irrelevant antibody control (grey line and filled area). (G) Quantitative real-time PCR analysis of APRIL (black bars), TWEAK (white bars), and SMT3IP (grey bars) mRNA expression in spleens of APRIL+/+ (WT), APRIL+/- (HET), and APRIL-/- (KO) mice. All values were normalized to an RPL19 RNA internal control. Standard deviations were calculated from triplicate reactions.
|
R1 embryonic stem cells (24) were transfected with the linearized vector by electroporation, and G418-resistant clones were screened for the presence of the expected recombination event by Southern blot analysis with 5'- and 3'-specific DNA probes (as shown in Fig. 1). Two independent APRIL-/- cell lines were microinjected into C57BL/6 blastocytes. Germ line transmission in mice generated by crossing chimeric males with C57BL/6 females was detected by coat color and confirmed by two-step genomic PCR (Fig. 1) with the following external (E) and internal (I) primer sets: E forward, TGCCCTAAGCCAGTCTACACCCAGTATTCCTTC; E reverse, TGGCCTGAAAGAAATGCTCACACTATCACCAAC; I forward, CTTAGAACCAGCCGTAGGAAGGATT; and I reverse, GTGCCAGGGCGTCCAGTACATACAA. APRIL knockout animals were backcrossed five times into the C57BL/6 background.
Examination of APRIL, TWEAK, and SMT3IP1 mRNA expression. Total RNA was extracted from splenocytes of APRIL-/-, APRIL+/-, and APRIL+/+ mice with the RNeasy mini kit (Qiagen Inc.). Real-time PCR (Taqman) was carried out in a 25-µl reaction containing 25 ng of total RNA, 0.6 µM each of the gene-specific forward and reverse primers, and 0.2 µM gene-specific fluorescence probe. APRIL-specific primers (forward: CCATGGCAGAGCCTCTGG; reverse: GAAGGATGGGGCGAAATCTC; probe: AGCAGAGTCCTGATGTCCTGGAAGCCT), TWEAK-specific primers (forward: ACTTGCTGGTGAACGGTGTG; reverse: AGAATTCTCAGCCACAGCAGCAAGC; probe: TGGCCCTGCGCTGCCTGG), and SMT3IP1-specific primers (forward: TCTGGTACTACAGCTGATCCAGTCTT; reverse: TATGGTAAGGGGCTTCCGG; probe: TCAGCGGATGCCAGGCAACGC) amplify nucleotides 468 to 532 of TWEAK (AF030100), 780 to 846 of APRIL (AF294825), and 1056 to 1122 of SMT3IP1 (AF194031). Gene-specific PCR products were measured with ABI Prism 7700 sequence detection system following the manufacturer's instructions (PE Corp.). The relative levels of APRIL, TWEAK, and SMT3IP1 mRNAs were normalized to RPL19 levels measured in the same samples.
Flow cytometry analyses. Single-cell suspensions were obtained from thymus, spleen, and lymph nodes by dissociation of the isolated tissues with plastic mesh and rubber stoppers from syringes. Suspensions were treated with red blood lysis buffer (Biosource International) following the manufacturer's instructions. Single-cell suspensions were stained with conjugated monoclonal antibodies (BD, Palo Alto, Calif.) for 30 min at 4°C. Staining of splenocytes with mouse monoclonal antibodies generated against mouse APRIL was done as described (27). Cell-associated fluorescence was analyzed with a FACScan instrument and associated Cell Quest software (Becton Dickinson, San Jose, Calif.).
T- and B-cell proliferation. Single-cell suspensions of total splenocytes were prepared as described above. For T-cell proliferation, total splenocytes (5 x 105/well) were stimulated with increasing amounts of concanavalin A or anti-mouse CD3e. For B-cell proliferation, splenic B cells were positively or negatively selected with anti-CD19-coated or anti-CD43-coated magnetic beads, following the manufacturer's instructions (Miltenyi Biotec, Auburn, Calif.). Purified B cells (5 x 105/well) were stimulated with increasing amounts of lipopolysaccharide or anti-mouse CD40 in the presence of recombinant mouse interleukin-4 (100 ng/ml). Proliferation was measured by pulsing cells with 1 µCi of [3H]thymidine for the final 18 h (T cells and positively selected B cells) or 8 h (negatively selected B cells) of the 3 (T cells and negatively selected B cells) or 5 (positively selected B cells) days of culture and measuring radioactivity uptake.
In vitro production of immunoglobulins by B cells. B cells from total splenocytes were purified and stimulated as described above. Culture supernatants were removed after 5 days and analyzed for IgM and IgG1 isotypes by enzyme-linked immunosorbent assay (ELISA) (BD).
Determination of serum immunoglobulin concentrations. Naive mice were bled at 12 weeks of age, and their serum was analyzed for various immunoglobulin isotypes by ELISA (BD). Serum samples from four mice per group were serially diluted (1:100 to 1:2,000) in triplicate and tested by ELISA, and results in the linear range of detection were used to determine relative immunoglobulin levels in APRIL-/- and APRIL+/+ mice.
In vivo antigen-specific antibody production. For analysis of type 2 T-cell-independent antigen-specific antibodies, 6-week-old APRIL-/- and APRIL+/+ mice were immunized intraperitoneally with 100 mg of trinitrophenyl-keyhole limpet hemocyanin (TNP-KLH) in complete Freund's adjuvant, bled 10 days later, and serially diluted serum samples were analyzed for trinitrophenyl-specific antibodies by ELISA (BD). For analysis of thymus-dependent antigen-specific antibodies, 6-week-old APRIL-/- and APRIL+/+ mice were immunized intraperitoneally with 100 mg of NP23-conjugated chicken gamma globulin (4-hydroxy-5-nitrophenyl hapten conjugated to chicken gamma globulin at a ratio of protein to hapten of 23:1; Biosource Technologies, Vacaville, Calif.) precipitated in alum. Total and high-affinity nitrophenyl-specific serum antibodies of various isotypes were measured by ELISA as described above with plates coated with NP23- and NP2-conjugated bovine serum albumin (BSA), respectively (BSA-to-hapten ratios of 23:1 and 2:1, respectively).
Ex vivo KLH-specific T-cell responses. Six-week-old APRIL-/- and APRIL+/+ mice were immunized via the hind footpads with 100 µg of KLH in saline, in a 1:1 emulsion with complete Freund's adjuvant that contained Mycobacterium tuberculosis strain H37Ra (1 mg/ml, Difco Laboratories, Detroit, Mich.). After 5 days, the popliteal lymph nodes were removed and cell suspensions were prepared as described above. T cells (5 x 105/well) were stimulated with increasing amounts of KLH. For interleukin-2 measurement, culture supernatants were collected after 24 h and analyzed by ELISA (BD). Interleukin-4 and gamma interferon levels were measured after 4 days of culture by ELISA (BD). Proliferation was measured by pulsing cells with 1 µCi of [3H]thymidine for the final 18 h of the 5 days of culture and measuring radioactivity uptake.
In vivo T-independent antigen-specific plasmablast differentiation. For analysis of T-cell-independent antigen-specific plasmablast differentiation, 10-week-old APRIL-/- and APRIL+/+ mice were injected with saline (control) or a high dose (108) or low dose (2 x 107) of heat-inactivated Streptococcus pneumoniae bacteria intravenously. After 3 days, splenocytes were collected and analyzed by fluorescence-activated cell sorting for presence of phosphorylcholine-reactive, CD138-positive plasmablasts.
Histology and immunohistochemistry. Tissues of 10-week-old APRIL-/- and APRIL+/+ mice were paraffin fixed and 3-µm sections were prepared. Hematoxylin and eosin-stained sections were analyzed for gross histological abnormalities. Peanut agglutinin (Vector Research, Burlingame, Calif.)-stained frozen sections were analyzed for structure of germinal centers.
|
|
|---|
Germ line-competent chimeric mice were generated by injection of two independent targeted ES cell lines into C57BL/6 blastocysts. These founders were used to generate two lines of APRIL-/- mice, which were born at the expected Mendelian ratio and with indistinguishable phenotypes, as described below. Further comparison of the phenotypes of two independent lines of APRIL-/- mice APRIL gene deficiency was verified by genomic PCR (Fig. 1E). In addition, fluorescence-activated cell sorting analysis with specific anti-mouse APRIL antibodies revealed lack of APRIL protein expression in total splenocytes isolated from APRIL-deficient mice but not from aged-matched wild-type littermates (Fig. 1F). The genomic locus that encodes APRIL also carries the TWEAK gene about 1 kb upstream of APRIL, while the first exon of another gene, SMT3IP1, is about 2 kb downstream of APRIL (Fig. 1A). To ensure that our APRIL targeting construct did not alter the TWEAK or SMT3IP1 genes, we analyzed the mRNA expression of all three genes in selected tissues by real-time PCR. This revealed altered expression of APRIL but not of TWEAK or SMT3IP1 (Fig. 1G).
Our observation that APRIL-deficient mice are viable appears to contradict a recent comment by Mackay et al., who noted in a review article that they have observed embryonic lethality of APRIL knockout mice (17). The striking discrepancy between these two knockout studies is difficult to reconcile without further details. Knockout of the same gene in different strain backgrounds on rare occasions may lead to distinctive phenotypes. For example, mice with knockout of the manganese superoxide dismutase gene generated on the C57BL/6J background die in utero due to dilated cardiomyopathy, while the same mutation introduced on the DBA/2J background leads to the development of a distinct severe phenotype and survival over several weeks after birth (10). Embryonic lethality may also be due to the very sensitive nature of the embryonic stem cells, which are able to acquire unrecognized mutations during in vitro manipulations. Alternatively, modification of the gene of interest may unexpectedly affect the expression of neighboring genes.
Of note, no other TNF family gene knockout reported to date caused embryonic lethality. Moreover, we have generated mice lacking both TWEAK and APRIL together by direct genetic manipulations in ES cells (rather than through crossing of single-gene knockout animals); these mice were also viable. Hence, it is unlikely that inadvertent deletion of TWEAK together with APRIL would cause lethality. On the other hand, SMT3IP1 encodes an isopeptidase involved in modification of SMT3b, a member of the sentrin family of ubiquitin-like proteins (26). It remains formally possible that unintended change in the structure or control of SMT3IP1 downstream of the APRIL locus may cause embryonic lethality. Regardless, our data suggest that APRIL is not critical for embryonic development.
Gross phenotypic and histologic analysis of APRIL-/- mice. To avoid potential phenotypic bias due to inbred genetics, we backcrossed the knockout animals 5 times into the C57/BL/6 strain background. APRIL-/- mice appeared phenotypically normal and fertile. APRIL signals by binding and activating two receptors: BCMA, expressed only on B lymphocytes (14, 19), and TACI, expressed both on T and B cells (41, 46). BCMA binds with high affinity to murine APRIL but not BLyS (33). Mice without BCMA lack any detectable immune abnormality (31, 47). In contrast, knockout of mouse TACI, a receptor that prefers binding to mouse BLyS over APRIL (33), leads to B-cell expansion and splenomegaly (42, 50). Surprisingly, administration of recombinant APRIL to mice results in a phenotype similar to that of TACI knockout (51). This probably results from exposure to high, nonphysiological APRIL doses that may trigger otherwise silent BCMA activity or perhaps prevent BLyS-TACI association, thereby directing BlyS more toward BR3.
Detailed necropsy and histologic analysis of various organs from APRIL-null animals compared to wild-type littermates, including primary and secondary lymphoid organs such as spleen and mesenteric lymph nodes (Fig. 2), did not reveal any detectable abnormalities. We also did not observe any notable differences between aged APRIL-deficient and wild-type animals. Thus, APRIL does not have a critical contribution to the development and structural organization of primary and secondary lymphoid organs. Other tissues examined and considered histologically normal included: brain, liver, kidneys, intestines, pancreas, bones, and gonads.
![]() View larger version (143K): [in a new window] |
FIG. 2. APRIL-/- mice have normal lymphoid organ structure. Histologic analysis of APRIL-/- mice. Sections from spleens (A) and mesenteric lymph nodes (B) from APRIL+/+ mice (left panel) and APRIL-/- mice (right panel) were stained with hematoxylin and eosin. Stained sections were photographed under a light microscope at 40x.
|
To assess if this contribution is essential, we first examined the effect of APRIL gene inactivation on T- and B-cell development. Fluorescence-activated cell sorting analysis of cells derived from the thymus, spleen, lymph nodes and peripheral blood with monoclonal antibodies to cell-specific markers did not reveal any gross differences in the number of T cells, neutrophils, natural killer (NK) cells or monocytes between APRIL-null and wild-type mice (Table 1, Fig. 3, and data not shown). Expression of the T-cell activation markers CD62L, CD44, CD25, and CD69 was similar on the surface of APRIL knockout and wild-type T cells (data not shown). Flow cytometric analysis of B-cell types, including pro- (B220+, CD43+, IgM-), pre- (B220+, CD43-, IgM-), immature (B220+, IgM+, IgD+), and recirculating mature (B220+, IgMlo, IgD+) B cells from bone marrow also did not reveal differences in cell number (Table 2). In contrast to the uniform increase of cellular components in the B-cell compartment in secondary lymphoid organs observed in TACI-null mice, the number of mature B cells in APRIL-/- mice was indistinguishable from that of the wild-type age-matched controls (Table 2). Together, these data suggest that elimination of the APRIL gene does not impair development, maturation and distribution of T and B lymphocytes.
|
View this table: [in a new window] |
TABLE 1. Distribution of T-cell subsetsa
|
![]() View larger version (31K): [in a new window] |
FIG. 3. APRIL-/- mice have normal T-cell development. Thymic T-cell populations in APRIL+/+ and APRIL-/- mice were determined by fluorescence-activated cell sorting. Numbers indicate the percentages of cells within the lymphocyte forward and side scatter gates. The data are representative of at least four analyses.
|
|
View this table: [in a new window] |
TABLE 2. B-cell subpopulations in APRIL+/+ and APRIL-/- micea
|
![]() View larger version (39K): [in a new window] |
FIG. 4. T and B cells from APRIL-/- mice have normal responses to different types of stimulation in vitro. (A) Total splenic T cells from APRIL+/+ ( ) or APRIL-/- mice ( ) were cultured in the presence of monoclonal antibody to CD3 (a) or concanavalin A (ConA) (b). Proliferation was measured by [3H]thymidine uptake on day 3. (B) Positively (a and b) and negatively (c) selected B cells from APRIL+/+ and from APRIL-/- mice were cultured in the presence of lipopolysaccharide (LPS) (a) or monoclonal antibody to CD40 plus interleukin-4 (IL4) (b and c). Proliferation was measured by [3H]thymidine uptake on day 5 (a and b) or day 3 (c). (C) B cells from APRIL-/- mice produce normal amounts of immunoglobulins upon in vitro stimulation. Purified B cells from APRIL+/+ and APRIL-/- mice were cultured in the presence of lipopolysaccharide (a) or monoclonal antibody to CD40 plus interleukin-4 (a and b). Five days later, culture supernatants were examined for the presence of IgG1 (a and b). Data are means ± standard deviation of four mice in each group.
|
![]() View larger version (35K): [in a new window] |
FIG. 5. APRIL-/- mice have normal T-cell responses to KLH. (A) APRIL+/+ mice ( ) and APRIL-/- mice ( ) were immunized with KLH in complete Freund's adjuvant. After 5 days, dendritic lymph nodes were collected and restimulated in vitro with KLH at the indicated concentrations. Proliferation was measured by [3H]thymidine uptake on day 5. (B to D) Cytokine secretion was measured in KLH-activated T cells by ELISA.
|
APRIL-deficient mice have normal humoral responses. Transgenic APRIL expression results in increased serum levels of IgM but not IgG (37). Conversely, TACI gene knockout leads to decreased levels of IgM and not IgG (50). However, in another study, TACI deficiency marginally increased levels of IgM and some IgGs (42). Comparison of total IgG levels in the serum of naive APRIL-null and wild-type mice did not reveal any significant difference (Fig. 6A). Challenge of APRIL transgenic mice with specific antigens resulted in a significant increase in thymus-independent type 2 T-cell independent antibody responses, manifested by elevated production of IgM and IgGs, while only IgM levels show increases in the case of thymus-dependent B-cell triggering (37). Consistent with activation of TACI-mediated signaling by transgenic APRIL expression, TACI knockout mice show severe diminution of B-cell type 2 T-cell-independent responses (42, 50). In one of these reports, however, considerable elevation of immunoglobulin titers due to thymus-dependent B-cell responses was noted only for IgGs (50). We compared thymus-dependent responses in APRIL-/- and wild-type mice by challenging them with nitrophenyl-conjugated chicken gamma globulin. Titers of both high-affinity (Fig. 6B) and total (Fig. 6C) antibodies of the IgM, IgG1, IgG2a, IgG2b, IgG3, IgA, and IgE classes were similar in both groups of mice. The type 2 T-cell-independent B-cell responses in APRIL-/- and wild-type mice examined by injecting TNP-conjugated KLH were also similar (Fig. 6D).
![]() View larger version (37K): [in a new window] |
FIG. 6. B cells from APRIL-/- mice have normal humoral responses to different types of stimulation in vivo. (A) Basal serum titers for various immunoglobulin isotypes were measured from age-matched APRIL+/+ ( ) and APRIL-/- mice ( ) determined by ELISA. (B-D) Mice were immunized with NP23-chicken gamma globulin (a thymus-dependent antigen) or trinitrophenyl (TNP)-KLH (a type 2 T-cell-independent antigen). Antinitrophenyl serum titers in NP23-chicken gamma globulin-immunized and antitrinitrophenyl serum titers in TNP-KLH-immunized mice were measured 10 days after challenge. ELISA was used to measure the indicated immunoglobulin isotypes with high affinity to antigen (B) and total immunoglobulin (C) in anti-NP23-chicken gamma globulin-immunized mice and in anti-trinitrophenyl-KLH-immunized mice (D). White bars, APRIL+/+ mice; black bars, APRIL-/- mice. Data are means ± standard deviations of four mice in each group.
|
![]() View larger version (18K): [in a new window] |
FIG. 7. Thymus-independent antigen-specific plasmablast differentiation in APRIL-/- mice is normal. APRIL+/+ mice ( ) and APRIL-/- mice ( ) were injected intravenously with 108 (high dose) or 2 x 107 (low dose) heat-inactivated Streptococcus pneumoniae bacteria or phosphate-buffered saline. After 3 days, splenocytes were collected and analyzed by fluorescence-activated cell sorting for phosphorylcholine-positive (PC+) CD138+ plasmablasts.
|
While APRIL does not appear to play an obligatory role in development or immune function, its elevated expression in tumors and ability to promote cell proliferation remain intriguing (9). Constitutive expression of APRIL in mouse NIH 3T3 fibroblasts enhances their proliferation in vitro and tumorigenicity in vivo (9, 29). Since 3T3 cells lack expression of TACI and BCMA, this raises the possibility that an APRIL-specific yet unidentified receptor is expressed by these cells (29). In addition, it has been suggested that APRIL provides prosurvival signals in a glioblastoma cell line by inhibiting proapoptotic caspase activity and promoting expression of the inhibitor of apoptosis protein XIAP (30). In 3T3 fibroblasts and HT29 colon adenocarcinoma cells, APRIL induces expression of the prosurvival molecules Bcl-2 and Bcl-xl (17). Future studies will make use of APRIL knockout mice to investigate whether APRIL plays a specific role in causing cancer.
E.V. and F.K. contributed equally to this work. |
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»