Previous Article | Next Article ![]()
Journal of Virology, December 2008, p. 12030-12038, Vol. 82, No. 24
0022-538X/08/$08.00+0 doi:10.1128/JVI.01575-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Department of Microbiology and Molecular Genetics, Harvard Medical School, Boston, Massachusetts 02115
Received 24 July 2008/ Accepted 30 September 2008
|
|
|---|
|
|
|---|
Like cellular DNA, the genomes of DNA viruses that replicate within the nucleus also associate with chromatin, albeit to varied degrees (48). The genome of herpes simplex virus 1 (HSV-1) has been shown to associate with histones during both lytic infection of epithelial cells and latent infection of neurons (13, 32, 39). During latent infection, the viral genome is packaged into nucleosomes and forms a classical laddering pattern following digestion with micrococcal nuclease (13). Consistent with the silencing of lytic genes during a latent infection, lytic gene promoters show markers of heterochromatin, such as H3K9me2, and low levels of histone acetylation (43, 78). In contrast, during lytic infection, regular repeating nucleosome arrays of HSV-1 DNA have not been detected (39, 45, 47). However, at least a portion of the viral DNA associates with histones (32, 39), although one study showed that histone association with immediate-early (IE) gene promoters was low in the presence of functional VP16 (32). A recent study by Oh and Fraser reported that the HSV-1 genome associated with histones at early times postinfection and the proportion of viral DNA associating with histones decreased at late times in the infectious cycle. The authors attributed this decrease to an increase in newly synthesized genomes that were free of histones (55).
The histones present on viral DNA during lytic infection show modifications associated with active gene expression, namely, acetylation of histone H3 and di- and trimethylation of histone H3 lysine 4 (32, 36, 39). Given that regulation of chromatin structure has important consequences for DNA processes, it is likely that both viral and cellular gene products manipulate the HSV-1 chromatin structure to allow gene expression and DNA replication to take place. Consistent with this, proteins involved in both chromatin remodeling and histone modifications are recruited into viral replication compartments (75), which are the sites of late gene transcription and DNA replication (3, 9, 10, 41, 57, 62).
The activation domain of the HSV-1 VP16 virion transactivator has been shown to interact with ATPase-dependent chromatin remodeling proteins and histone acetyltransferases (54, 76). Deletion of the VP16 activation domain results in a decrease in recruitment of ATPase-dependent chromatin remodeling proteins and histone acetyltransferases to IE promoters and an accompanying increase in histone occupancy and decrease in histone acetylation (32). An additional candidate for modifying chromatin on the HSV-1 genome is the immediate-early protein ICP0. ICP0 is able to stimulate the expression of all three classes of viral genes in infected cells (4, 7) and cotransfected genes (14, 21, 53, 56, 61), although it does not bind DNA directly (17). The ability of ICP0 null viruses to replicate in tissue culture cells is dependent upon both the cell type and multiplicity of infection (MOI) (64, 72). ICP0 null viruses show the highest degree of attenuation within primary embryonic or fetal fibroblasts, in which a low multiplicity of infection results in a quiescent infection (15). Infection of a number of other cell types, such as HeLa cells, at a low multiplicity of infection results in attenuation of viral infection, in which gene expression is reduced compared to that in wild-type (WT) virus-infected cells (4, 7). The effects of ICP0 gene deletions on viral gene expression can be overcome in part by treatment with inhibitors of histone deacetylases (HDACs) (34, 59, 60), suggesting that ICP0 may act to increase levels of histone acetylation.
ICP0 has been found to interact with the class II HDACs and reduce their activity in vitro (49). ICP0 also interacts with the RE1 silencing corepressor (REST/CoREST)-HDAC1 complex and causes dissociation of HDAC1 from the complex. Disassociation of HDAC1 from the REST/CoREST complex is able to increase, at least in part, viral yields in the absence of ICP0 (24). Despite the observation that ICP0 interacts with HDACs, there is no direct evidence that ICP0 affects the chromatin structure on the HSV-1 genome during a lytic infection. We therefore employed chromatin immunoprecipitation (ChIP) to determine the effects of ICP0 on the chromatin structure of HSV-1 during lytic infection.
|
|
|---|
The WT strain of HSV-1 (KOS) used in this study was grown and titrated as described previously (42, 44). The ICP4– virus (n12) was provided by P. Schaffer (University of Arizona) (12) and grown and titrated on E11 cells (66). The ICP0– (7134) and 7134R rescued viruses, provided by P. Schaffer, were grown and titrated on U2OS cells (5). The DNA polymerase null, HP66, and HP66R rescued viruses, provided by D. Coen (Harvard Medical School), were grown and titrated as described previously (50).
Virus infections. HeLa cells were plated into 100-mm dishes 24 h prior to infection to obtain 95% confluence at the time of infection. Virus was diluted in cold phosphate-buffered saline (PBS) containing 0.1% glucose and 1% heat-inactivated BCS in the presence or absence of 200 µg/ml sodium phosphonoacetate (PAA). After 1 h of adsorption at 37°C, cells were washed twice with PBS containing 0.1% glucose and 1% heat-inactivated BCS, twice with low-pH wash buffer (40 mM citric acid, 10 mM KCl, 135 mM NaCl, pH 3) (33), and twice with DMEM containing 1% heat-inactivated BCS. Cells were overlaid with DMEM containing 1% heat-inactivated BCS and incubated at 37°C.
Chromatin immunoprecipitation. ChIP assays were carried out as described previously (78) with a few modifications. At various times postinfection, cells were incubated with 1% formaldehyde for 10 min to cross-link the chromatin. Cells were washed twice with PBS and scraped into PBS containing Complete protease inhibitor tablets (Roche Diagnostics). Cells were resuspended in sodium dodecyl sulfate (SDS) lysis buffer (1% SDS-10 mM EDTA-50 mM Tris, pH 8.1) containing protease inhibitors and incubated on ice for 20 min. The samples were sonicated in 20-s pulses for a total of 4 min, using a Heat Systems XL sonicator, to yield DNA fragments of between 200 and 500 bp in length. The samples were clarified by centrifugation at 13,000 x g at 4°C for 10 min. The supernatant was diluted 10-fold in phosphate-buffered radioimmunoprecipitation assay buffer (0.1% SDS, 1% sodium deoxycholate, 150 mM NaCl, 10 mM Na2PO4, 2 mM EDTA, 1% Nonidet P-40) and precleared for 1 h at 4°C. At this point, 1% of the total volume was removed and reserved as the input. Immunoprecipitation was carried out at 4°C overnight with 5 µg rabbit immunoglobulin G (IgG; Millipore) as the negative control or 1.5 µg anti-histone H3 IgG (Abcam) or 5 µl anti-acetyl histone H3 lysine 9/18 IgG (Millipore).
Immunocomplexes were collected by incubation with a salmon sperm DNA-protein A-agarose slurry for 1 h at 4°C with rotation. Beads were washed for 4 min at 4°C with rotation, two times with low-salt wash buffer (150 mM NaCl, 20 mM Tris·HCl, pH 8.1, 2 mM EDTA, 1% Triton X-100, and 0.1% SDS), one time with high-salt wash buffer (500 mM NaCl, 20 mM Tris·HCl, pH 8.1, 2 mM EDTA, 1% Triton X-100, and 0.1% SDS), one time with lithium chloride wash buffer (0.25 M LiCl, 1% NP-40, 1% sodium deoxycholate, 1 mM EDTA, and 10 mM Tris-HCl, pH 8.1) and two times with Tris-EDTA buffer (10 mM Tris-HCl, pH 8, 1 mM EDTA). The DNA-protein complexes were eluted by adding 200 µl of elution buffer (1% SDS, 0.1 M NaHCO3) preheated to 65°C, rotating for 10 min at room temperature, incubating for 10 min at 65°C, and rotating for 10 min at room temperature. NaCl was added to a final concentration of 0.2 M to both the eluates and the inputs, and the samples were incubated at 65°C for 5 h in the presence of 1 µg RNase A (Ambion). The samples were then digested with proteinase K, and the DNA was purified by using a QIAquick PCR purification kit (Qiagen).
Real-time PCR. Real-time PCR was performed using the power Sybr green PCR master mix and a Prism 7300 sequence detection system (Applied Biosystems) according to the manufacturer's instructions. The final reaction volume was 25 µl, containing 2.5 µl DNA and 100 nM of each primer. The primers used in this study are shown in Table 1. The specificity of each primer pair was determined by running a dissociation curve of the PCR products. All DNA samples were run in duplicate, and relative copy numbers were determined by comparison with a standard curve generated by 10-fold dilutions of sonicated DNA from HSV-1-infected HeLa cells. The fraction of viral DNA immunoprecipitated compared to the input sample was normalized to the fraction of cellular DNA (glyceraldehyde 3-phosphate dehydrogenase [GAPDH] precipitated in the same reaction.
|
View this table: [in a new window] |
TABLE 1. Primers used for real-time PCR analysis
|
|
|
|---|
![]() View larger version (14K): [in a new window] |
FIG. 1. ChIP analysis of the fraction of HSV-1 DNA associated with histone H3. Total cell extracts were prepared at 3 hpi from HeLa cells infected with HSV-1 at an MOI of 0.1 PFU/cell. ChIP was carried out with an antibody specific to the C terminus of histone H3. The fraction of DNA immunoprecipitated compared to the input value was determined by real-time PCR, with the fraction immunoprecipitated by a nonspecific antibody subtracted, and is expressed as the fold enrichment over the fraction of GAPDH DNA immunoprecipitated. Primers were specific for regions of ICP4 (A), ICP8 (B), and UL44 (C) (see Table 1 for primer locations and sequences). Abbreviations: US, upstream of the transcriptional start site; SS, start site; DS, downstream of the transcriptional start site. The means and standard errors of the means of at least five independent experiments are shown.
|
![]() View larger version (16K): [in a new window] |
FIG. 2. Effects of MOI and inactivation of nonadsorbed virus on the proportion of viral DNA associated with histone H3. ChIP was carried out at 3 hpi with an antibody for histone H3, and real-time time PCR was undertaken using primers for the region downstream of the ICP4 transcriptional start site and analyzed as for Fig. 1. (A) HeLa cells were infected at MOIs between 0.1 and 5 PFU/cell. (B) HeLa cells were infected at an MOI of 0.1 PFU/cell. After a 1-h adsorption period, the medium was replaced without washing the cells, or the cells were washed with PBS alone or washed with PBS followed by a low-pH wash buffer. The means and standard errors of the means of at least four independent experiments are shown.
|
The temporal decrease in the association of viral DNA with histone H3 occurs independently of viral DNA replication. The proportion of viral DNA associated with histone H3 was shown to decrease at late times following infection at a high MOI (55). To determine if the association of viral DNA with histones also decreased in a low-MOI infection, we carried out ChIP assays at various times following infection and analyzed the proportion of the ICP4 gene that was associated with histone H3. ICP4 gene sequences were precipitated with histone H3-specific antibody at approximately the same efficiency as GAPDH at 3 hpi, but the association of viral DNA with histone H3 decreased between 3 and 8 hpi and was further decreased at 12 hpi (Fig. 3A).
![]() View larger version (31K): [in a new window] |
FIG. 3. Kinetics of histone H3 association with the HSV genome and effects of viral DNA replication. HeLa cells were infected at an MOI of 0.1 PFU/cell in the presence or absence of the DNA polymerase inhibitor PAA (A and B) or with either a DNA Pol– virus or the rescued DNA Pol+ virus (C and D). Cell lysates were prepared at 3, 8, and 12 h postinfection. (A and C) ChIP was carried out using anti-histone H3 antibody. The fraction of ICP4 downstream (DS) (A) or ICP8 DS (C) DNA immunoprecipitated as determined by real-time PCR was normalized to the fraction of GAPDH DNA immunoprecipitated. The HSV genome copy number determined by real-time PCR was normalized to the GAPDH copy number and then normalized to the results for the untreated 3-h time point (B and D). The means and standard errors of the means of at least three independent experiments are shown. A significant decrease (*) between the mean fraction of viral DNA immunoprecipitated between 3 and 8 hpi was determined using Student's t test (P < 0.05).
|
To rule out any nonspecific effects of PAA on the association of histones with viral DNA, we examined the association of histone H3 with the ICP8 gene in cells infected with a replication-defective, DNA polymerase null virus, HP66 (50), in parallel with the rescued virus, HP66R. Analysis of the genome copy numbers confirmed that viral DNA replication did not occur following infection with HP66 (Fig. 3D). Following infection with both the HP66 DNA polymerase null and the HP66R rescued virus, ICP8 DNA associated with histone H3 to high levels at 3 hpi, but the levels decreased by 8 hpi (Fig. 3C). These results confirmed that viral DNA replication was not a prerequisite for the reduction of histone association with HSV-1 DNA between 3 and 8 hpi.
Total amounts of HSV-1 DNA associated with histone H3 increase as a consequence of viral DNA replication. Although levels of histone H3 on viral DNA decreased prior to or in the absence of viral DNA replication, this did not rule out the possibility that histones were also removed from viral DNA during DNA replication, as reported previously (55). At 12 h postinfection, an increase in the proportion of viral DNA associated with histone H3 could be seen in PAA-treated cells compared to the untreated cells (0.033% in the PAA-treated cells compared to 0.0069% in the untreated cells) (Fig. 3A). However, due to the large differences in viral genome copy numbers (approximately 1,000-fold more DNA in the untreated cells) (Fig. 4B), it was difficult to determine whether the slightly decreased proportion of DNA associating with histone H3 in the untreated cells occurred because of an increase in viral DNA load. Therefore, we examined the total amount of viral DNA sequences immunoprecipitated with the H3 antibody normalized to the total amount of GAPDH sequence immunoprecipitated at the time of DNA replication. If histones were removed during replication, the total amount of DNA associated with histone H3 would decrease. Surprisingly, instead of seeing a decrease in the amount of DNA associated with histone H3 following DNA replication, there was a 90-fold increase in the amount of viral DNA associated with histone H3 in the untreated cells in comparison to the PAA-treated cells (Fig. 4A). Taken together, these data suggested that following DNA replication there were increases in the absolute amounts of both histone H3-associated and histone-free viral DNA.
![]() View larger version (20K): [in a new window] |
FIG. 4. Effects of DNA replication on the total amount of viral DNA associated with histone H3. HeLa cells were infected at an MOI of 0.1 PFU/cell in the presence or absence of the DNA polymerase inhibitor PAA. Cell lysates were prepared at 12 hpi. (A) ChIP assays were carried out using anti-histone H3 antibody. The copy number of ICP4 downstream DNA associated with histone H3 was determined by real-time PCR and normalized to the copy number of GAPDH associated with histone H3 and then normalized to the results for PAA-treated cells. (B) Total HSV genome copy number determined by real-time PCR using primers for the ICP4 gene was normalized to the GAPDH copy number and then normalized to results with the PAA-treated cells. The means and standard errors of the means are shown. Samples with mean values that vary significantly (P < 0.05, Student's t test) are indicated (*).
|
![]() View larger version (17K): [in a new window] |
FIG. 5. Kinetics of histone H3 association with the ICP8 gene following infection with an ICP4 nonsense mutant virus. HeLa cells were infected at an MOI of 0.1 PFU/cell with either WT (A) or ICP4– (B) virus. Cell lysates were prepared at 3 and 8 hpi. The fraction of ICP8 downstream DNA immunoprecipitated, as determined by real-time PCR, was normalized to the fraction of GAPDH DNA immunoprecipitated. Lines connect values for individual experiments. Samples with mean values that varied significantly (P < 0.05, paired Student's t test) are indicated (*).
|
![]() View larger version (30K): [in a new window] |
FIG. 6. Kinetics of histone H3 association with the ICP4 and ICP8 genes following infection with a ICP0– virus. HeLa cells were infected at an MOI of 0.1 PFU/cell with either an ICP0– virus (7134) (B and D) or rescued virus (7134R) (A and C). Cell lysates were prepared at 3 and 8 hpi. The fraction of ICP4 downstream (A and B) or ICP8 downstream (C and D) DNA immunoprecipitated determined by real-time PCR was normalized to the fraction of GAPDH DNA immunoprecipitated. Lines connect values for individual experiments. Samples with mean values that varied significantly (P < 0.05, paired Student's t test) are indicated (*).
|
![]() View larger version (17K): [in a new window] |
FIG. 7. ChIP for total histone H3 and acetylated histone H3 on the ICP8 gene in the presence and absence of ICP0. HeLa cells were infected an MOI of 0.1 PFU/cell with either 7134 ICP0– virus or 7134R rescued virus (ICP0+). Cell lysates were prepared at 6 hpi, and ChIP was carried out using anti-histone H3 (A and B) and anti-acH3 (K9/18) (B) antibodies. The proportion of acetylated histone H3 associating with viral DNA was determined as a fraction of DNA associating with acH3 K9/14, normalized to the fraction associating with total histone H3. Lines connect values for individual experiments. Samples with mean values that varied significantly (P < 0.05, paired Student's t test) are indicated (*).
|
|
|
|---|
HSV-1 DNA associates to higher levels with histone H3 early in infection at a low MOI. In this study, we extended previous observations that at least a portion of viral DNA associates with histone H3 during lytic infection (32, 39) to conclude that physiologically relevant levels of histone H3 associate with viral DNA early in infection. The association of viral DNA with histones following infection appears to be similar to that seen for transiently transfected DNA, in which the DNA associates with histones but fails to form a regularly spaced nucleosomal structure (31). The lack of a regularly spaced nucleosomal structure on transfected DNA is possibly due to the lower representation of the linker histone, histone H1. It is not known whether histone H1 is found on viral DNA at early time points, but during later time points in infection it has been shown to be excluded from replication compartments (71).
We showed that the level of histones on viral DNA is dependent upon the multiplicity of infection, with an increased proportion of viral DNA associated with histone H3 at a low MOI. Possible reasons for the dependence on MOI are that the cell is unable to assemble histones efficiently onto higher quantities of incoming viral DNA, a limiting pool of available histone H3 for assembly onto increasing amounts of viral DNA, and/or the presence of increased quantities of viral proteins able to reduce the histone association with the viral genome. A study by Yager and Bachenheimer (81) showed that both histone H3 mRNA levels and H3 synthesis rapidly decline following viral infection, which may limit the pool of histones available for loading onto viral DNA at higher MOIs.
Two viral tegument proteins which may limit the histone association with the viral genome are VP16 and VP22. VP16 has been found to interact with members of the SWI/SNF family of ATP-dependent chromatin remodeling proteins (28, 51, 54). In addition, deletion of the VP16 activation domain was found to result in increased levels of histone H3 on IE promoters (32). VP22 has been found to interact with the chromatin remodeling protein TAF-1 (77) and prevent TAF-1-mediated nucleosome deposition onto naked DNA in vitro.
A role for ICP0 in reducing histone H3 occupancy. Following initial loading of histones onto the viral DNA, we observed a decrease in association of histone H3 with viral DNA. The reduction in histone load is analogous to the loss of histones from promoters following activation of the HS82 genes and the PHO5 genes in Saccharomyces cerevisiae (63, 82). Little is known about eviction of histones from the promoters of higher eukaryotic promoters. However, there is evidence that nucleosome disruption precedes transcriptional activation in the HIV long terminal repeat (reviewed in reference 30) and the human T-cell leukemia virus 1 promoter (46).
Interestingly, histone loss following activation of the HS82, PHO5, and human T-cell leukemia virus 1 promoters occur independently of transcription (46, 68, 69, 82). The loss of histones from an early promoter in the absence of ICP4, and thus the absence of transcription, demonstrated that transcription is also not required for the reduction in histone association on the ICP8 promoter. However, histone loss is dependent upon ICP0. A role for ICP0 in reducing histone association to allow access to the DNA for transcription factor binding is consistent with its role in promoting the expression of genes in the context of both viral infection and transfection (4, 7). Furthermore, the reactivation of viral genomes from at least a subset of latently infected neurons in vivo and quiescently infected cells in vitro can be promoted by ICP0 (25, 26, 83). Given that high levels of histones are present on the HSV-1 genome in latently infected neurons (reviewed in reference 40), it is possible that a function of ICP0 in the reactivation from latency involves the reduction in histone association with the viral genome. A recent study by Coleman et al. found no decrease in histone association with the viral genome following ICP0-mediated derepression in vitro (8). However, this does not rule out the possibility that ICP0 mediates removal of histones from latently infected neuronal cells.
The role of ICP0 in promoting histone acetylation is consistent with previous observations that ICP0 is able to cause the dissociation of HDACs from the REST/CoREST complex (23) and that ICP0 is able to promote histone acetylation during derepression of the quiescent viral genome in vitro (8). Furthermore, previous observations that HDAC inhibition can increase viral gene expression in the absence of ICP0 suggest that ICP0 functions to inhibit HDACs which would otherwise try to silence viral gene expression. However, in some studies inhibition of HDAC administered at the time of infection had no effect on viral replication or gene expression (18, 60). HDAC inhibition has a broad spectrum of effects, including a decrease in the levels and a redistribution of the heterochromatin protein HP1, as well as decreased histone H3 acetylation and lysine 4 methylation at the nuclear periphery (2), which is an early site of HSV-1 genome localization (16, 70). Hence, it is difficult to determine whether the effects of HDAC inhibition are due to direct effects on histone acetylation or indirect effects due to remodeling of the cell nucleus. Further studies are therefore required to determine the effects of HDAC inhibition on the HSV-1 chromatin structure.
A role for ICP0 in reducing total histone load has not been proposed previously. It is possible that through promoting histone acetylation or preventing deacetylation, ICP0 indirectly reduces histone association. Indeed, histone acetylation results in a change in the net charge on the nucleosome, which could loosen histone-DNA interactions (20). Furthermore, acetylated histones have been shown to be easier to displace from reconstituted nucleosomes in vitro (6, 27, 37) in addition to serving as a binding site for chromatin remodeling proteins (1, 29, 38). Conversely, ICP0 perhaps functions directly to reduce histone levels on viral DNA, and the finding that a higher proportion of histones are acetylated is instead a consequence of the reduced histone occupancy or selective removal of nonacetylated histones.
Components of PML bodies, such as PML, Sp100, and hDaxx, play a role in reducing herpesvirus gene expression (18, 19, 79). One protein found within PML bodies, hDaxx, interacts with the core histones and chromatin-remodeling proteins (35, 80). Although hDaxx does not appear to be degraded by ICP0, hDaxx localization to nucleoprotein complexes is inhibited in the presence of ICP0 (19). Hence, a function of PML bodies may be to prevent chromatin remodeling on the viral genome. However, to date there is no direct evidence for a link between PML component repression of HSV-1 gene expression and chromatin structure. It will be of interest to define the roles of cellular proteins and the dependence of histone acetylation on the reduction of histone levels on the HSV-1 genome to determine not only how ICP0 functions to regulate viral gene expression but also the mechanisms underlying nucleosome eviction in higher eukaryotes.
The association of progeny DNA with histone H3. In a recent study Oh and Fraser reported that viral progeny DNA does not associate with histones (55). In contrast, we observed that the total amount of viral DNA associated with histones increased with viral DNA replication. The different observations for the role of DNA replication could possibly be due to the differing cell types used: Oh and Fraser carried out experiments in a neuroblastoma cell line, whereas in this study HeLa cells were used. A further possibility is that the different observations resulted from the different multiplicities of infection used: Oh and Fraser infected cells at an MOI of 5 PFU/cell, whereas in this study cells were infected at an MOI of 0.1 PFU/cell. Given that the proportion of viral DNA associating with histone H3 at early times was found to decrease at higher MOIs, it is also possible that reduced amounts of histones are loaded onto replicating DNA following high-MOI infections. In addition, the study by Oh and Fraser examined the proportion of progeny DNA associating with histone H3 by performing consecutive ChIPs for histone H3 and bromodeoxyuridine. However, it is possible that the incorporation of bromodeoxyuridine into replicating DNA altered the chromatin structure, as observed previously (52, 73).
It is important to examine the total amount of DNA associating with histones and not just the proportion, which may give misleading results due to increased viral genome copy number. Our data suggest that as DNA replication takes place there is an increase in both histone-free and histone-associated viral DNA. The nature of the two populations of viral DNA following DNA replication possibly reflects the fate of replicated DNA. Given that HSV-1 DNA packaged within capsids is free from histones (22, 55, 58), the histone-free portion of DNA may represent either packaged DNA or DNA destined for packaging. It is tempting to speculate that the histone-associated DNA represents DNA serving as a template for late viral gene expression and that a change in the chromatin structure of viral DNA following DNA replication is involved in the regulation of late gene expression. Therefore, it is possible that a temporal change in histone association throughout the course of infection is a mechanism by which the virus regulates its gene expression throughout the lytic infectious cycle.
We thank Priscilla Schaffer for providing the ICP0 and ICP4 mutant viruses, Don Coen for the DNA polymerase mutant viruses, and Kevin Bryant and Megan Horn for comments on the manuscript.
Published ahead of print on 8 October 2008. ![]()
|
|
|---|
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»