This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hernandez-Ramon, E. E.
Right arrow Articles by Maitland, N. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hernandez-Ramon, E. E.
Right arrow Articles by Maitland, N. J.

 Previous Article  |  Next Article 

Journal of Virology, May 2008, p. 4853-4861, Vol. 82, No. 10
0022-538X/08/$08.00+0     doi:10.1128/JVI.02388-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Dimerization of the Human Papillomavirus Type 16 E2 N Terminus Results in DNA Looping within the Upstream Regulatory Region{triangledown}

Elena E. Hernandez-Ramon,1,{dagger} Julie E. Burns,1* Wenke Zhang,2,{ddagger} Hannah F. Walker,1 Stephanie Allen,2 Alfred A. Antson,3 and Norman J. Maitland1

YCR Cancer Research Unit, Department of Biology (Area 13), University of York, Heslington, York, YO10 5DD United Kingdom,1 Laboratory of Biophysics and Surface Analysis, School of Pharmacy, University of Nottingham, Nottingham, NG7 2RD United Kingdom,2 York Structural Biology Laboratory, Department of Chemistry, University of York, Heslington, York, YO10 5DD United Kingdom3

Received 5 November 2007/ Accepted 15 February 2008


arrow
ABSTRACT
 
Papillomavirus E2 proteins play a central role in regulating viral gene expression and replication. DNA-binding activity is associated with the C-terminal domain of E2, which forms a stable dimer, while the N-terminal domain is responsible for E2's replication and transactivation functions. The crystal structure of the latter domain revealed a second dimerization interface on E2 which may be responsible for DNA loop formation in the regulatory region of the human papillomavirus (HPV) genome. We investigated the biological significance of the N-terminal dimerization by introducing single amino acid substitutions into the dimerization interface. As expected, these substitutions did not influence the C-terminal dimerization and DNA-binding functions of E2. However, the mutations led to reduced transactivation of a synthetic E2-responsive reporter gene, while HPV DNA replication was unaffected. The effect of the mutations on DNA looping was visualized by atomic force microscopy. While wild-type E2 was able to generate DNA loops, all three mutant E2 proteins were defective in this ability. Our results suggest that N-terminal dimerization plays a role in E2-mediated transactivation, probably via DNA looping, a common mechanism for remote regulation of gene transcription.


arrow
INTRODUCTION
 
Several members of the human papillomavirus (HPV) family have been strongly implicated in the development of human cancers, particularly cervical cancer; of these, HPV type 16 (HPV-16) is the most common type found in tumors (25). Indeed, the presence of HPV DNA has been identified as the major risk factor for cervical carcinoma and is increasingly implicated in a number of other human tumors (4, 14). The E2 proteins of HPVs are central regulatory proteins in the viral life cycle. While considered to be a typical transcription factor, the E2 protein is equally important in the initiation of viral DNA replication. The loss or deletion of the E2 open reading frame (ORF) occurs frequently in cervical carcinoma cells when the genome of high-risk HPV types, such as HPV-16, becomes integrated into the cell genome (10). This has resulted in a hypothesis which states that the removal of E2 control results in deregulated expression of the HPV oncogenes E6 and E7. The presumed mode of action of HPV E2 is mediated by its binding to the consensus sequence AGGCN4GCCT, which is normally present in four copies in the upstream regulatory region (URR) of the genomes of genital HPVs at highly conserved locations relative to the transcriptional start site and origin of viral DNA replication (26). It has been proposed that binding to these sites has a hierarchical priority which at low concentrations of E2 results in transcriptional transactivation, whereas increasing amounts result in repression of the HPV early promoter (18, 29, 38). However, it has recently been shown that, in the case of HPV-16, E2 binding sites (E2BS) 1 and 2 (and 3 to a certain degree) are important for transcriptional repression, independent of binding affinities (37). Hence, the precise mechanism of E2 transcriptional control is not clearly understood, as the protein is capable of acting as a transcriptional transactivator as well as a repressor in cotransfection experiments with promoter-indicator-gene plasmids containing multiple E2BS (reviewed in reference 8).

E2 proteins are ~45-kDa nuclear phosphoproteins with a tripartite secondary structure. The structure of the intact E2 protein closely resembles that of other mammalian transcription factors, consisting of a DNA-binding/dimerization domain (DBD) connected by a flexible linker to a multiple-protein-binding transactivation domain (TAD). The three-dimensional structures of the C-terminal DBD (11, 16, 17) and N-terminal TAD (1, 5, 15, 30, 45) of several E2 proteins have been reported, both alone and in complex, revealing a tight dimer of the DBD bound to DNA and a characteristic L-shaped TAD structure. In the case of HPV-16, two TAD domains form a dimer; this additional dimerization interface has been proposed to link two E2 dimers bound via the DBD to two distant E2BS within the URR (5). Such interactions would induce a loop within the viral promoter (Fig. 1), supporting the model proposed by Knight et al. (21), who showed that full-length wild-type bovine papillomavirus type 1 (BPV-1) E2 protein, and not the truncated C-terminal E2 protein, formed stable DNA loops, visible with electron microscopy.


Figure 1
View larger version (37K):
[in this window]
[in a new window]

  		FIG. 1. Structure of HPV-16 E2 protein. Schematic representation of E2-mediated DNA loop formation showing possible N-terminal (E2NT) interactions between E2 molecules (shown in yellow and blue) bound through the C termini (E2CT) at different E2BS in DNA. (Adapted from reference 5 with permission).

Detailed examination of this second dimerization domain revealed the critical amino acids (5). The TAD dimer is formed by hydrogen-bonding interactions between the side chains of seven amino acids, Arg37, Ala69, Ile73, Gln76, Leu77, Glu80, and Thr81 (Fig. 2A), which are highly conserved in the many types of papillomavirus. In random mutagenesis studies of E2, changes in several of these amino acids resulted in a loss of biological activity (reviewed in reference 8). To test the biological significance of the TAD self-interaction, we have generated single-residue-mutant E2 proteins and assessed their biological activity. Proteins mutated in the TAD dimer interface failed to act as transcriptional transactivators, but their ability to initiate HPV genome replication was unaffected. Additionally, we visualized DNA looping by atomic-force microscopy (AFM) and demonstrated the ability of wild-type HPV-16 E2 to induce loops in the URR of the viral genome. Proteins mutated in the TAD dimer interface, although able to bind to DNA, failed to produce DNA loops.


Figure 2
View larger version (23K):
[in this window]
[in a new window]

  		FIG. 2. Computer modeling of the N-terminal dimerization interfaces of wild-type and mutant HPV-16 E2 proteins. Potential amino acid interactions in the TAD dimerization domain of the wild-type (A), R37A (B), E80A (C), and A69Q E2 proteins (D). Dotted lines represent potential hydrogen bonding between residues of two E2 TAD molecules (colored blue and red). Fourteen bonds can potentially be formed in the wild-type E2 dimer interface, six in the R37A mutant, and eight in the E80A mutant. The A69Q mutant is not predicted to disrupt hydrogen bonding in the interface but to cause steric hindrance to the opposing loop structure. (E) Apparent molecular mass (MW) of wild-type (WT) and R37A E2 proteins at various concentrations at 18,000 rpm. The R37A mutant appears to be monomeric at all concentrations. The wild-type protein is a mixture of monomeric and dimeric forms which may contain oligomers higher than the dimer. The wild type's curves cover the same molecular-mass range at different concentrations and do not overlap, suggesting that the mixture is not in equilibrium under these conditions.


arrow
MATERIALS AND METHODS
 
Cloning and expression of the HPV-16 E2 proteins. E2 ORFs were PCR amplified, using the reference clone pBR322:HPV-16 as template. The full-length HPV-16 E2 ORF had previously been cloned in pLNCX (Clontech) by PCR amplification using primers E2-Kozak-Flag-ATG (5'-TTTAAGCTTCCGCGGCAGCCACCATGGACTACAAGGACGACGATGACAAGATGGAGACTCTTTGCCAACG-3' [HPV-complementary sequence in bold]) and E2 TAG (5'-AAAAGAATTCTTTGGATCCTTTATCGATAAATCATATAGACATAAATCCAGTAGAC-3'), cloning into pGEM-T Easy (Promega), and HindIII/ClaI subcloning into pLNCX to generate pLNCX-E2. The E39A mutant was constructed by PCR with primers E2-Kozak-Flag-ATG and E39A rev (5'-TCCCATTCTCTGGCCTTGTAATAAATAGCACATGCTAG-3'), cloning of the amplicon in pGEM-T Easy, and replacement of the HindIII/BstXI fragment of pLNCX-E2. The other single-residue mutations were introduced into the HPV-16 E2 ORF by PCR mutagenesis with the following primers containing the desired mutations (mutated codons underlined): for the R37A mutant, 5'-GGAAACACATGGCCCTAGAATGT-3' and 5'-ACATTCTAGGGCCATGTGTTTCC-3' PCR; for the A69Q mutant, 5'-CAAAGAATAAACAGTTACAAGCAA-3' and 5'-TTGCTTGTAACTGTTTATTCTTTG-3'; and for the E80A mutant, 5'-GCAACTAACGTTAGCAACAATATATA-3' and 5'-TATATATTGTTGCTAACGTTAGTTGC-3'. The PCR products were cloned in pGEM-T Easy (Promega) or pCR 2.1-TOPO (Invitrogen) and subcloned into pLNCX (Clontech) and pTriEx 1.1 (Novagen) for mammalian expression or pET30 (Novagen) for protein expression in bacteria. The inserts included N-terminal FLAG tags and the E2 stop codon to prevent the expression of the C-terminal His tag in pET30. The C-terminal fragment encoding amino acids 277 to 365 of HPV-16 E2 (SCT) was previously cloned in pET15b by NdeI/BamHI subcloning from pT7Blue:E2CT (described in reference 29). All E2 inserts were fully sequenced to ensure that no other mutations were present. The construction of the pET15b plasmid encoding the N-terminal fragment encoding amino acids 1 to 201 of HPV-16 E2 (E2NT) was described in reference 9. pET15b:R37A NT was constructed by the same method, using pGEM-T Easy:E2 R37A as the template for PCR.

Soluble, recombinant, full-length E2 or CT proteins were expressed in Escherichia coli BL21(DE3) pLysS by induction with 1 mM isopropyl-β-D-thiogalactopyranoside for 3 h at 25°C, purified by nickel affinity followed by heparin-binding chromatography (modified from reference 29), and used immediately or snap-frozen and stored at –80°C. The final purified proteins were checked by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and Western blotting with rabbit primary antibody anti-E2SCT (described in reference 40). The E2NT protein was expressed and purified as described in reference 9, dialyzed into 20 mM Tris-HCl, pH 8.0, 0.2 mM EDTA, 300 mM NaCl, 2.5 mM Tris(2-carboxyethyl)phosphine hydrochloride (TCEP), snap-frozen, and stored at –80°C.

Cell culture. HeLa cells were grown in monolayer culture in Dulbecco's modified Eagle's medium plus 10% fetal calf serum (D10 medium) at 37°C and 5% CO2. C-33A cells were cultured in D10 medium supplemented with 10 mM HEPES. LNCaP cells (derived from a human prostate carcinoma) were maintained in monolayer culture in RPMI medium plus 10% fetal calf serum.

Transcriptional transactivation assays. The response plasmid p2x2E2BS:SV40min:EGFP was previously constructed by the ligation of a 513-bp fragment containing the 4E2BS and minimal simian virus 40 promoter from pGL2:2x2xE2BS:SV40min:luc (22) into pEGFP-1 (Clontech) (32).

For transactivation assays, cells were passaged 1:2 2 days before transfection, and six-well plates were seeded with 1.5 x 105 to 1.8 x 105 cells per well on the following day. After 24 h, cells were transfected with an E2 expression plasmid (pLNCX-E2 or pTriEx-E2) and the E2 response plasmid p2x2E2BS:SV40min:EGFP, using Fugene6 transfection reagent (Roche). Cells were observed by fluorescence microscopy and harvested after 48 h for analysis of enhanced green fluorescent protein (EGFP) expression by flow cytometry.

In vivo HPV DNA replication. Transient replication assays were carried out as described by Sakai et al. (28). C-33A cells were split 1:2 2 days before transfection and seeded on the following day into 10-cm-diameter plates at a density of 6 x 105 cells per plate. After 24 h, cells were cotransfected by using a calcium phosphate transfection kit (Invitrogen) with 1 µg of plasmid p16ori containing the HPV-16 replication origin, 5 µg of the HPV-16 E1 expression plasmid pCMV-E116, and 100 ng of the HPV-16 E2 expression plasmid pLNCX:E2. Seventy-two hours after transfection, low-molecular-weight DNA was extracted and digested with 10 U of DpnI and/or 10 U of XmnI overnight. Digested DNAs were separated on agarose gels, transferred to positively charged nylon membranes (Roche), and probed by using a 32P-radiolabeled, 700-bp, ori-containing HPV-16 fragment released from PvuII-digested p16ori.

Analytical ultracentrifugation. Sedimentation analyses were carried out in an Optima XL-I ultracentrifuge (Beckman-Coulter, CA), using scanning UV optics. Three concentrations of recombinant E2NT were loaded into cells containing 12-mm path length, six-channel Epon centerpieces with quartz windows. The solvent was 20 mM Tris-HCl, pH 8.0, 0.2 mM EDTA, 300 mM NaCl, 2.5 mM TCEP. Data were obtained at rotor speeds of 18,000 and 25,000 rpm, and the time to equilibrium was typically 10 to 12 h. All runs were carried out at 20°C, and all radial scans were at a wavelength of 280 nm. Molecular weights were estimated by using the Beckman ultracentrifuge software.

EMSA. Complementary 20-mer oligonucleotides, including E2BS 4 (most distal to the P97 promoter of HPV-16; E2BS4 sense, 5'-TTCAACCGAATTCGGTTGCA-3', and E2BS4 antisense, 5'-TGCAACCGAATTCGGTTGAA-3'), were end-labeled with biotin using a biotin 3'-end DNA-labeling kit (Perbio) and annealed. Binding reactions were carried out by using a LightShift chemiluminescent electrophoretic mobility shift assay (EMSA) kit (Perbio) according to the manufacturer's instructions. Twenty-microliter reaction mixtures containing 1x binding buffer (10 mM Tris, 50 mM KCl, 1 mM dithiothreitol, pH 7.5), 2.5% glycerol, 0.05% NP-40, 25 to 60 ng (0.5 to 1.33 pmol) of purified protein, and 50 fmol labeled, double-stranded oligonucleotide were incubated for 20 min at room temperature. Reaction mixtures lacking protein or containing cold competitor (10 pmol unlabeled oligonucleotide) were included as controls. Reaction products were separated on 8% native polyacrylamide gels and transferred to positively charged nylon membrane for chemiluminescent detection.

AFM of DNA-protein complexes. (i) DNA template. In order to generate the template for DNA looping, a fragment from the HPV-16 URR containing E2BS 3 and 4 was amplified by PCR using the primers 5'-GTGTGTTTGTATGTATGGTA-3' (nucleotides 7239 to 7258 in the HPV-16R sequence) and 5'-TACGCCCTTAGTTTTATACA-3' (nucleotides 15 to 34 in the HPV-16R sequence). The PCR product was gel purified by using GeneClean (Q-BIOgene) according to the manufacturer's instructions.

(ii) Protein preparation. Proteins used for AFM studies were cleaved overnight with enterokinase (pET30) or thrombin (pET15b) to remove the N-terminal His tag, using a recombinant enterokinase or recombinant thrombin kit (Novagen) according to manufacturer's protocols. Cleavage was monitored by sodium dodecyl sulfate-polyacrylamide gel electrophoresis.

(iii) DNA binding. Binding reaction mixtures containing 50 ng (0.11 pmol) of double-stranded DNA template, 0.55 pmol of cleaved protein, 1x binding buffer from a LightShift chemiluminescent EMSA kit, 2.5% glycerol, and 0.05% NP-40 were incubated at 37°C for 30 min. Samples were kept on ice for 30 min to a few hours until they were immobilized on a NiCl2-treated, freshly cleaved mica surface (Agar Scientific) as described below.

(iv) AFM imaging. Imaging experiments were carried out using a multimode AFM with a Nanoscope IIIa controller (Veeco, CA), using a tip holder for tapping-mode imaging in air. To enable imaging, 5 to 10 µl of the protein-DNA complex solution was deposited on NiCl2-treated mica and incubated at room temperature for 30 seconds, followed by three gentle rinses with deionized water (100 µl/rinse). The sample was then dried under a gentle flow of dry N2, attached to a magnetic sample puck, and positioned in the AFM sample stage for imaging. Rectangular silicon tapping-mode cantilevers (Olympus) with resonant frequencies in the range of 280 to 360 kHz were employed during the experiments. The scanning frequencies were typically 3 Hz per line. The AFM images were analyzed by using the instrument analysis software or SPIP (Image Metrology A/S, Denmark).


arrow
RESULTS
 
R37A E2 is monomeric in solution. Structural modeling of E2 using the software package QUANTA (Accelrys, Inc.) was used to predict amino acid substitutions that would disrupt the dimerization interface without causing major alterations in the overall protein structure (Fig. 2B to D). Wild-type E2 genes and mutants containing arginine 37 to alanine (R37A), alanine 69 to glutamine (A69Q), and glutamic acid 80 to alanine (E80A) substitutions were generated by PCR mutagenesis and cloned into vectors for expression in bacterial and mammalian cells.

In order to assess whether mutant E2 could self-interact, wild-type and R37A E2 N-terminal truncated proteins were expressed in E. coli, purified by immobilized metal-ion affinity chromatography and ion exchange chromatography, and analyzed by analytical ultracentrifugation. The results presented in Fig. 2E show that the R37A mutant exists in solution as monomers, unlike the wild-type protein, which shows both monomeric and dimeric forms, although the mixture was not in equilibrium under the conditions used. Similar results were obtained at 25,000 rpm (data not shown). The solvent density and partial specific volume of the protein were calculated as 1.0113 and 0.7306, respectively, using the program SEDNTERP according to the method of Laue et al. (23).

E2 N-terminal dimerization disruption affects transactivation. For measurement of their transactivation activities, mutant E2 genes were cloned into the expression vectors pTriEx and pLNCX and cotransfected into the human epithelial cell lines HeLa, C-33A, and LNCaP together with an E2-responsive EGFP reporter vector. After optimization of the relative dosage of E2 and target plasmids in each cell type (data not shown), a standard transactivation assay was carried out in LNCaP epithelial cells, in which E2 cytotoxicity was lowest in our cell bank. Forty-eight hours after transfection, the EGFP fluorescence of the cells was visualized by inverted phase fluorescence microscopy, as shown in Fig. 3A. Overall transfection efficiencies of approximately 50% were achieved (measured by transfection of a cytomegalovirus-EGFP construct, not shown). The cells were harvested, and the relative amounts of EGFP fluorescence were quantified by fluorescence-activated cell sorter analysis (Fig. 3B). The results confirmed the ability of wild-type E2 to transactivate the minimal promoter, with EGFP expression increased almost 70-fold compared to basal levels. A complete abrogation of transactivation was observed with the R37A and A69Q mutants. Although the E80A mutant showed increased transactivation (about fivefold) in comparison to that of the other two mutants, it was less than 7% of that achieved using wild-type E2. When the expression vector pTriEx was used, similar transactivation levels were obtained, as shown in Fig. 3B.


Figure 3
View larger version (15K):
[in this window]
[in a new window]

  		FIG. 3. In vitro transcriptional transactivation of a synthetic E2-responsive promoter by E2 mutants. (A) EGFP expression in LNCaP cells 48 h after cotransfection with pLNCX-E2 expression and reporter plasmids. Cells shown in the panel labeled "no E2" were transfected with reporter vector and empty expression plasmid only. (B) Fluorescence-activated cell sorter quantification of E2-mediated transactivation 48 h posttransfection. Means and standard deviations of the results of three experiments using pLNCX (black) and pTriEx (gray) E2 expression plasmids are shown. Results were normalized to "basal expression" (i.e., that of cells transfected with empty expression plasmid), set at a value of 1. WT, wild type.

The results of the assays performed in cell lines C-33A and HeLa showed lower overall transactivation levels (up to fivefold) than in LNCaP cells, but a significant difference was still observed between wild-type and mutant E2-mediated transactivation (data not shown).

N-terminal dimerization disruption does not affect replication. In general, papillomavirus DNA replication requires both the full-length E2 protein and the E1 protein. While E1 is recognized as the major papillomavirus viral replication protein, it is widely accepted that E2 plays an important auxiliary role (43). In order to assess the capacities of the mutant E2 proteins to replicate viral DNA, transient replication assays were carried out in C-33A cells, as described above. As shown in Fig. 4, wild-type, R37A, A69Q, and E80A E2 proteins were replication competent, while no replication was observed with E39A mutant E2. This mutant was previously established as replication defective (28) and was used in our study as a negative control.


Figure 4
View larger version (75K):
[in this window]
[in a new window]

  		FIG. 4. In vitro HPV DNA replication in the presence of HPV-16 E2 and E1. Replication assays were performed as described in Materials and Methods. Low-molecular-weight DNA was digested with XmnI (X) or XmnI/DpnI (XD), and the products probed for HPV-16 ori. The lane labeled "ori" contained only plasmid p16ori. The band at 4 kb represents linearized plasmid p16ori. The presence of this band in XD lanes represents newly replicated, DpnI-resistant plasmid. WT, wild type.

R37A, A69Q, and E80A mutant E2 proteins bind to DNA. To further confirm that all E2 mutant proteins could bind to DNA, EMSAs were performed. The DNA template used for these assays contained E2BS 4 (the most-distal E2BS to the P97 promoter) from the HPV-16 URR. Bacterially produced E2 proteins were prepared, purified to virtual homogeneity using immobilized metal-ion affinity chromatography and heparin affinity, and incubated with double-stranded, biotin-labeled template. As shown in Fig. 5, C-terminal, truncated, wild-type, R37A, A69Q, and E80A E2 proteins clearly shifted the labeled template, confirming that the DNA-binding function was intact, while controls with unlabeled competitor and supershifting with anti-E2 antiserum confirmed that the shifted bands were E2 specific (data not shown).


Figure 5
View larger version (14K):
[in this window]
[in a new window]

  		FIG. 5. DNA-binding capacities of recombinant E2 proteins. Binding reactions were performed, visualized by EMSA as described in Materials and Methods, and separated on an 8% native acrylamide gel. +, lane contained E2 protein; –, lane contained no E2 protein; CT, C terminus; WT, wild type.

HPV-16 E2 protein induces DNA looping. Preformed, full-length BPV-1 E2 dimers were shown by electron microscopy to form stable DNA loops through the N-terminal region when bound to isolated DNA fragments containing widely spaced E2BS (21). We investigated the capacity of the HPV-16 E2 protein to induce DNA loops by using AFM. The advantage of AFM over conventional electron microscopy is that heavy metal staining and excessive drying are not required.

The DNA template was generated by PCR and comprised a 700-bp HPV-16 URR fragment (nucleotides 7239 to 34 of the HPV-16 genome) containing the two E2BS (3 and 4) most distal to the P97 promoter (Fig. 6A). Purified proteins were incubated with the DNA template, immobilized, and imaged. Two hundred DNA molecules were analyzed for each protein examined.


Figure 6
View larger version (103K):
[in this window]
[in a new window]

  		FIG. 6. DNA-looping abilities of wild-type and mutant HPV-16 E2 proteins. (A) Diagram of the HPV-16 URR. The template for AFM was generated by PCR amplification of the region between the arrows, containing E2BS 3 and 4. (B) Predicted E2-DNA structures and measurements. E2 is represented schematically as TAD (gray ovals) linked to DBD (dark circles) (upper schematic) or as gray ovals (lower schematic). (C) Large-area AFM image views of DNA structures complexed with E2 proteins. The scale bars represent 160 nm. The Z ranges for the images shown are as follows: wild type (WT), Z range of 2 nm; R37A mutant, Z range of 2.5 nm; A69Q mutant, Z range of 3 nm; E80A mutant, Z range of 2 nm; and C terminus (CT), Z range of 3 nm. (D) Zoomed images of the representative structures observed for the DNA-E2 protein complexes (images of various dimensions). Arrows indicate the positions of E2 binding to the DNA template (image of R37A is a three-dimensional representation rather than a planar view as for the other images). (E) Zoomed (100 x 100 nm) images and measurements of three DNA loops formed by wild-type E2 protein. (F) Results of analysis of 200 E2-DNA molecules. *, P value of <0.001 using Fisher's exact test. WT, wild type; C-term, C terminus.

Schematic representations of the DNA loops and predicted sizes are shown in Fig. 6B. As shown in Fig. 6C and D, wild-type HPV-16 E2 protein formed loops with the template DNA. Figure 6E shows three enlarged DNA loops for which tail and loop sizes were calculated, resulting in measurements very close to the predicted dimensions. The results presented in Fig. 6F show that this E2-mediated DNA looping was observed in 10% of the molecules (20 loops in 200 analyzed molecules). Other intermolecular structures were seen with wild-type E2 only, suggesting interactions between proteins bound to different DNA molecules (data not shown).

DNA looping is mediated by the N-terminal dimerization interface of HPV-16 E2. The identification of a second dimerization interface located within the E2 N-terminal domain suggested that looping may result from interactions between DNA-bound E2 dimers (5) (Fig. 1). To test this hypothesis, we evaluated the abilities of R37A, A69Q, E80A, and C-terminal E2 proteins to form DNA loops by using AFM as described above. R37A E2 protein was completely defective in forming DNA loops (Fig. 6C). This result was not due to a binding deficiency, since all the proteins bound to E2BS DNA (Fig. 5) and protein-DNA complexes were observed with AFM (Fig. 6D). The A69Q and E80A E2 proteins both showed a reduced ability to form DNA loops compared with that of wild-type E2 (Fig. 6C): loops were observed in only 2% and 1% of the analyzed molecules, respectively (Fig. 6F), showing a statistically significant difference (P < 0.001, using Fisher's exact test) in comparison to the number of loops formed by wild-type E2. These proteins were also observed bound to the DNA template (Fig. 6D). As expected, an N-terminally deleted C-terminal fragment of E2 was unable to form DNA loops while retaining the ability to form protein-DNA complexes, visualized by using AFM (Fig. 6C and D).


arrow
DISCUSSION
 
The crystal structure of the N-terminal TAD of HPV-16 E2 protein revealed a novel and extensive second dimerization surface (5). Although the role of this dimerization has not been determined, it seemed to be of major importance for protein function, as most mutations in HPV-16 E2 that compromise transactivation without affecting replication were found to map in the dimerization interface (28). Although the HPV-11 E2 protein crystal structure has since been determined to be monomeric, probably due to different crystallization conditions (45), dimer formation in the HPV-16 E2 TAD was verified in solution, using biophysical techniques (5).

Recently, Sanders et al. (30) demonstrated that the BPV-1 E2 protein also dimerizes through the TAD. Although the residues important for HPV-16 TAD dimerization are highly conserved, the BPV-1 interaction is mediated through redox interactions involving a different interface and residues that are not conserved between BPV and HPV E2 proteins. Nevertheless, these results support the hypothesis of N-terminal dimerization as a more-general mechanism to regulate papillomavirus E2 activity.

There is an accumulated literature reporting the effects of single-residue mutations on the transactivation and replication functions of the E2 TAD (reviewed in reference 8). However, many of these amino acid changes are predicted to disrupt the overall protein structure, leading to inconclusive results. Information derived from the three-dimensional structure of the N-terminal TAD of the HPV-16 E2 protein and computational modeling allowed us to identify important residues in the dimerization interface. In the HPV-16 E2 TAD dimer, arginine 37 (R37) makes several hydrogen bonds with glutamic acid 80 (E80) and threonine 81 (T81) of a separate E2 molecule (Fig. 2A) (5). We replaced R37 and E80 with alanine (R37A and E80A, respectively) in order to abolish those interactions. These changes were predicted to cause destabilization of the dimer interface (Fig. 2B and C) and minimal perturbation of the protein conformation, since the R37 side chain is exposed and alanine is compatible with all secondary structures (13). Alanine 69 is a highly conserved residue, with all papillomavirus E2 proteins having glycine or alanine at this position (http://hpv-web.lanl.gov). A69 was mutated to glutamine (A69Q) to evaluate the replacement of a nonpolar, small amino acid with a polar side-chain amino acid, and as shown in Fig. 2D, the glutamine side chain is predicted to interfere with dimer formation due to steric hindrance.

It has been suggested that mutations which reduce or eliminate one function but not another should not disrupt protein structure extensively. All three mutants were able to support the replication of a plasmid containing an HPV-16 origin of replication (Fig. 4), consistent with previous reports for R37A and the analogous HPV-18 Q80A mutant (15, 28) and confirming that the mutant E2 proteins have no major structural alterations.

For the transactivation function of E2, our results showed that the R37A mutation has a major effect on transactivation by HPV-16 E2, as previously demonstrated (28). Alanine substitutions for this invariant amino acid in BPV-1 and HPV-11 and -31 E2 proteins have shown similar results (3, 12, 41). The replacement of glutamine 80 with alanine in HPV-18 (15) resulted in a 50% reduction in transactivation. In our work, the replacement of glutamic acid 80 with alanine resulted in an even more striking reduction in transactivation. Replacing A69 with glutamine (A69Q) severely affected E2-mediated transactivation.

It had been previously reported that the full-length wild-type BPV-1 E2 protein formed stable DNA loops that were visible by electron microscopy and that the TAD was necessary for this looping (21). Antson et al. suggested that the loop formation may be mediated by TAD dimerization forming tetramers between E2 molecules bound as dimers to widely separated E2BS in the URR of the virus genome (5) (Fig. 1). Such a structure could bring tissue-specific enhancers into close proximity to the core transcription complex, as reported for other transcriptional regulators (27). The redox-dependent TAD dimerization since reported for BPV-1 E2 (30) has been proposed as an intradimeric, and not as a tetrameric, interaction. However, tetrameric forms in BPV E2 are not ruled out and are likely to be the structural mechanism by which this virus induces DNA loops (21).

Using AFM, we found that the HPV-16 E2 protein indeed formed DNA loops with an HPV-16 URR fragment containing E2BS 3 and 4 and that the loop formation required the TAD. When mutant E2 proteins were tested for their ability to form DNA loops, TAD dimerization interface mutant E2 proteins were completely or mostly defective in forming DNA loops in comparison to loop forming by wild-type E2 (P < 0.001). This was not due to an inability to bind specifically to DNA. Since truncated C-terminal E2 or mutant E2 proteins, presumably defective in TAD dimerization, were unable to form DNA loops, we conclude that the E2 N-terminal dimerization must be responsible for the DNA looping.

A recent study has shown that HPV-11 E2 also forms loops with URR fragments containing E2BS, with a frequency similar to that in our results (36). As with BPV-1 and HPV-16, this study found that the full-length protein was necessary for loop formation, and the results suggested that the N-terminal TAD was responsible. It was suggested that ori DNA remodeling by E2 could facilitate the binding of the E1 protein and melting of DNA during viral replication, while the formation of long-distance loops may inhibit this process. Although the URR fragment we used did not include the promoter-proximal E2BS 1 and 2, this model is not inconsistent with our results in which mutants which failed to form DNA loops were still competent for E1-dependent replication.

The results of this work suggest that N-terminal dimerization may play a role in regulating the transactivation function of the HPV-16 E2 protein, possibly via DNA looping. It would not be surprising if the E2 protein, as a transcriptional regulator, employs the DNA-looping mechanism to regulate gene transcription. It is believed that this mechanism is widely used in gene regulation (7, 44), particularly by transcriptional enhancers (31), bringing distally bound transcription factors close to the site of transcription initiation. Furthermore, it has been reported that multiple transcription factors function by oligomerizing and self-associating to form DNA loops (20, 39, 42, 47).

During the course of this investigation, E2 was reported to interact directly with the cellular bromodomain protein Brd4 (2). Brd4, besides tethering the viral genome to mitotic chromosomes in BPV-1 (6, 46), was proposed as the major cellular partner required for E2 transcriptional activation in a number of papillomaviruses, using either a heterologous or a homologous promoter in transactivation assays (19, 24, 34, 35). However, the results of a more-recent study suggested that HPV-18 and -11 E2-mediated repression of the HPV promoter does not require Brd4 (33). Although we have not assessed the ability of our mutants to bind Brd4, the R37A mutant and other mutants with mutations in the dimer interface fail to interact with cellular Brd4 (2, 6, 34, 35). Since the transactivation activity was assessed in vivo, it is likely that the lack of transactivation may be due to an inability to interact with cellular Brd4. DNA looping, however, was studied in vitro in the absence of Brd4 or any other proteins.

The participation of the N-terminal dimerization in DNA looping provides more insight into the mechanisms of HPV-16 E2 function. Ultimately, analysis of the crystal structures of these mutants is still necessary in order to fully confirm that the dimerization disruption is indeed occurring, as well as examination of the interaction of Brd4 with residues E80 and A69 and its consequences for loop and complex formation.


arrow
ACKNOWLEDGMENTS
 
We thank Yorkshire Cancer Research for funding this work; Andrew Leech, Technology Facility, Department of Biology, University of York, for analytical ultracentrifugation; Iain Morgan, University of Glasgow, for the gift of p16ori and pCMV-E116 plasmids; and John Benson, Harvard Medical School, for pGL2:2x2xE2BS:SV40min:luc.

Elena E. Hernandez-Ramon was funded by the Consejo Nacional de Ciencia y Tecnologia, Mexico, and by the Secretaria de Educacion Publica, Mexico. Wenke Zhang was funded by the Biotechnology and Biological Sciences Research Council.


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: YCR Cancer Research Unit, Department of Biology (Area 13), University of York, Heslington, York, YO10 5DD United Kingdom. Phone: 44 1904 328707. Fax: 44 1904 328710. E-mail: jeb10{at}york.ac.uk Back

{triangledown} Published ahead of print on 12 March 2008. Back

{dagger} Present address: Genomic Medicine Unit, Hospital General de México, Dr. Balmis 148, Col. Doctores, Del. Cuauhtémoc, México D.F., CP 06726 Mexico. Back

{ddagger} Present address: State Key Lab for Supramolecular Structure and Materials, Jilin University, 2699 Qianjin Street, Changchun, 130012 People's Republic of China. Back


arrow
REFERENCES
 
    1
  1. Abbate, E. A., J. M. Berger, and M. R. Botchan. 2004. The X-ray structure of the papillomavirus helicase in complex with its molecular matchmaker E2. Genes Dev. 18:1981-1996.[Abstract/Free Full Text]
    2. 2
  2. Abbate, E. A., C. Voitenleitner, and M. R. Botchan. 2006. Structure of the papillomavirus DNA-tethering complex E2:Brd4 and a peptide that ablates HPV chromosomal association. Mol. Cell 24:877-889.[CrossRef][Medline]
    3. 3
  3. Abroi, A., R. Kurg, and M. Ustav. 1996. Transcriptional and replicational activation functions in the bovine papillomavirus type 1 E2 protein are encoded by different structural determinants. J. Virol. 70:6169-6179.[Abstract]
    4. 4
  4. Akgül, B., J. C. Cooke, and A. Storey. 2006. HPV-associated skin disease. J. Pathol. 208:165-175.[CrossRef][Medline]
    5. 5
  5. Antson, A. A., J. E. Burns, O. V. Moroz, D. J. Scott, C. M. Sanders, I. B. Bronstein, G. G. Dodson, K. S. Wilson, and N. J. Maitland. 2000. Structure of the intact transactivation domain of the human papillomavirus E2 protein. Nature 403:805-809.[CrossRef][Medline]
    6. 6
  6. Baxter, M. K., M. G. McPhillips, K. Ozato, and A. A. McBride. 2005. The mitotic chromosome binding activity of the papillomavirus E2 protein correlates with interaction with the cellular chromosomal protein, Brd4. J. Virol. 79:4806-4818.[Abstract/Free Full Text]
    7. 7
  7. Bulger, M., and M. Groudine. 1999. Looping versus linking: toward a model for long-distance gene activation. Genes Dev. 13:2465-2477.[Free Full Text]
    8. 8
  8. Burns, J. E., C. L. Burn, E. E. Hernandez-Ramon, C. Schmitz, A. Lloansi Vila, and N. J. Maitland. 2003. Forms and functions of papillomavirus E2 proteins. Recent Res. Dev. Biochem. 4:795-823.
    9. 9
  9. Burns, J. E., O. V. Moroz, A. A. Antson, C. M. Sanders, K. S. Wilson, and N. J. Maitland. 1998. Expression, crystallization and preliminary X-ray analysis of the E2 transactivation domain from papillomavirus type 16. Acta Crystallogr. D 54:1471-1474.[CrossRef][Medline]
    10. 10
  10. Choo, K. B., C. C. Pan, and S. H. Han. 1987. Integration of human papillomavirus type 16 into cellular DNA of cervical carcinoma: preferential deletion of the E2 gene and invariable retention of the long control region and E6/E7 open reading frames. Virology 161:261.
    11. 11
  11. Cicero, D. O., A. D. Nadra, T. Eliseo, M. Dellarole, M. Paci, and G. de Prat-Gray. 2006. Structural and thermodynamic basis for the enhanced transcriptional control by the human papillomavirus strain-16 E2 protein. Biochemistry 45:6551-6560.[CrossRef][Medline]
    12. 12
  12. Cooper, C. S., S. N. Upmeyer, and P. L. Winokur. 1998. Identification of single amino acids in the human papillomavirus 11 E2 protein critical for the transactivation or replication functions. Virology 241:312-322.[CrossRef][Medline]
    13. 13
  13. Cunningham, B. C., and J. A. Wells. 1989. High-resolution epitope mapping of hgh-receptor interactions by alanine-scanning mutagenesis. Science 244:1081-1085.[Abstract/Free Full Text]
    14. 14
  14. Gillison, M. L., W. M. Koch, R. B. Capone, M. Spafford, W. H. Westra, L. Wu, M. L. Zahurak, R. W. Daniel, M. Viglione, D. E. Symer, K. V. Shah, and D. Sidransky. 2000. Evidence for a causal association between human papillomavirus and a subset of head and neck cancers. J. Natl. Cancer Inst. 92:709-720.[Abstract/Free Full Text]
    15. 15
  15. Harris, S. F., and M. R. Botchan. 1999. Crystal structure of the human papillomavirus type 18 E2 activation domain. Science 284:1673-1677.[Abstract/Free Full Text]
    16. 16
  16. Hegde, R. S., and E. J. Androphy. 1998. Crystal structure of the E2 DNA-binding domain from human papillomavirus type 16: implications for its DNA binding-site selection mechanism. J. Mol. Biol. 284:1479-1489.[CrossRef][Medline]
    17. 17
  17. Hegde, R. S., S. R. Grossman, L. A. Laimins, and P. B. Sigler. 1992. Crystal structure at 1.7 Å of the bovine papillomavirus-1 E2 DNA-binding domain bound to its DNA target. Nature 359:505-512.[CrossRef][Medline]
    18. 18
  18. Hou, S. Y., S. Y. Wu, and C. M. Chiang. 2002. Transcriptional activity among high and low risk human papillomavirus E2 proteins correlates with e2 DNA binding. J. Biol. Chem. 277:45619-45629.[Abstract/Free Full Text]
    19. 19
  19. Ilves, I., K. Maemets, T. Silla, K. Janikson, and M. Ustav. 2006. Brd4 is involved in multiple processes of the bovine papillomavirus type 1 life cycle. J. Virol. 80:3660-3665.[Abstract/Free Full Text]
    20. 20
  20. Jackson, P., I. Mastrangelo, M. Reed, P. Tegtmeyer, G. Yardley, and J. Barrett. 1998. Synergistic transcriptional activation of the MCK promoter by p53: tetramers link separated dna response elements by DNA looping. Oncogene 16:283-292.[CrossRef][Medline]
    21. 21
  21. Knight, J. D., R. Li, and M. Botchan. 1991. The activation domain of the bovine papillomavirus E2 protein mediates association of DNA-bound dimers to form DNA loops. Proc. Natl. Acad. Sci. USA 88:3204-3208.[Abstract/Free Full Text]
    22. 22
  22. Kovelman, R., G. K. Bilter, E. Glezer, A. Y. Tsou, and M. S. Barbosa. 1996. Enhanced transcriptional activation by E2 proteins from the oncogenic human papillomaviruses. J. Virol. 70:7549-7560.[Abstract]
    23. 23
  23. Laue, T. M., B. D. Shah, T. M. Ridgeway, and S. L. Pelletier. 1992. Computer-aided interpretation of analytical sedimentation data for proteins, p. 90-125. In S. E. Harding, A. J. Rowe, and J. C. Horton (ed.), Analytical ultracentrifugation in biochemistry and polymer science. Royal Society of Chemistry, Cambridge, United Kingdom.
    24. 24
  24. McPhillips, M. G., A. G. Oliveira, J. E. Spindler, R. Mitra, and A. A. McBride. 2006. Brd4 is required for E2-mediated transcriptional activation but not genome partitioning of all papillomaviruses. J. Virol. 80:9530-9543.[Abstract/Free Full Text]
    25. 25
  25. Muñoz, N., F. X. Bosch, S. de Sanjosé, R. Herrero, X. Castellsagué, K. V. Shah, P. J. F. Snijders, and C. J. L. M. Meijer. 2003. Epidemiologic classification of human papillomavirus types associated with cervical cancer. N. Engl. J. Med. 348:518-527.[Abstract/Free Full Text]
    26. 26
  26. O'Connor, M., S.-Y. Chan, and H.-U. Bernard. 1995. Transcription factor binding sites in the LCR of genital HPVs, p. III-21-III-46. In G. Myers, H.-U. Bernard, H. Delius, C. Baker, J. Icenogel, A. Halpern, and C. Wheeler (ed.), Human papillomaviruses 1995. Los Alamos National Laboratory, Los Alamos, NM.
    27. 27
  27. Ptashne, M. 2005. Regulation of transcription: from lambda to eukaryotes. Trends Biochem. Sci. 30:275-279.[CrossRef][Medline]
    28. 28
  28. Sakai, H., T. Yasugi, J. D. Benson, J. J. Dowhanick, and P. M. Howley. 1996. Targeted mutagenesis of the human papillomavirus type 16 E2 transactivation domain reveals separable transcriptional activation and DNA replication functions. J. Virol. 70:1602-1611.[Abstract]
    29. 29
  29. Sanders, C. M., and N. J. Maitland. 1994. Kinetic and equilibrium binding studies of the human papillomavirus type-16 transcription regulatory protein E2 interacting with core enhancer elements. Nucleic Acids Res. 22:4890-4897.[Abstract/Free Full Text]
    30. 30
  30. Sanders, C. M., D. Sizov, P. R. Seavers, M. Ortiz-Lombardía, and A. A. Antson. 2007. Transcription activator structure reveals redox control of a replication initiation reaction. Nucleic Acids Res. 35:3504-3515.[Abstract/Free Full Text]
    31. 31
  31. Schleif, R. 1992. DNA looping. Annu. Rev. Biochem. 61:199-223.[CrossRef][Medline]
    32. 32
  32. Schmitz, C. 2000. Transient and stable expression of the human papillomavirus type 16 (HPV-16) early protein 2 (E2) in human keratinocytes. D. Phil. thesis, University of York, United Kingdom.
    33. 33
  33. Schweiger, M.-R., M. Ottinger, J. You, and P. M. Howley. 2007. Brd4 independent transcriptional repression function of the papillomavirus E2 proteins. J. Virol. 81:9612-9622.[Abstract/Free Full Text]
    34. 34
  34. Schweiger, M.-R., J. X. You, and P. M. Howley. 2006. Bromodomain protein 4 mediates the papillomavirus E2 transcriptional activation function. J. Virol. 80:4276-4285.[Abstract/Free Full Text]
    35. 35
  35. Sénéchal, H., G. G. Poirier, B. Coulombe, L. A. Laimins, and J. Archambault. 2007. Amino acid substitutions that specifically impair the transcriptional activity of papillomavirus E2 affect binding to the long isoform of Brd4. Virology 358:10-17.[CrossRef][Medline]
    36. 36
  36. Sim, J., S. Ozgur, B. Y. Lin, J.-H. Yu, T. R. Broker, L. T. Chow, and J. Griffith. 2008. Remodeling of the human papillomavirus type 11 replication origin into discrete nucleoprotein particles and looped structures by the E2 protein. J. Mol. Biol. 375:1165-1177.[CrossRef][Medline]
    37. 37
  37. Soeda, E., M. C. Ferran, C. C. Baker, and A. A. McBride. 2006. Repression of HPV16 early region transcription by the E2 protein. Virology 351:29-41.[Medline]
    38. 38
  38. Steger, G., and S. Corbach. 1997. Dose-dependent regulation of the early promoter of human papillomavirus type 18 by the viral E2 protein. J. Virol. 71:50-58.[Abstract]
    39. 39
  39. Stenger, J. E., P. Tegtmeyer, G. A. Mayr, M. Reed, Y. Wang, P. Wang, P. V. C. Hough, and I. A. Mastrangelo. 1994. p53 oligomerization and DNA looping are linked with transcriptional activation. EMBO J. 13:6011-6020.[Medline]
    40. 40
  40. Stevenson, M., L. C. Hudson, J. E. Burns, R. L. Stewart, M. Wells, and N. J. Maitland. 2000. Inverse relationship between the expression of the human papillomavirus type 16 transcription factor E2 and virus DNA copy number during the progression of cervical intraepithelial neoplasia. J. Gen. Virol. 81:1825-1832.[Abstract/Free Full Text]
    41. 41
  41. Stubenrauch, F., A. M. E. Colbert, and L. A. Laimins. 1998. Transactivation by the E2 protein of oncogenic human papillomavirus type 31 is not essential for early and late viral functions. J. Virol. 72:8115-8123.[Abstract/Free Full Text]
    42. 42
  42. Su, W., S. Jackson, R. Tjian, and H. Echols. 1991. DNA looping between sites for transcriptional activation: self-association of DNA-bound Sp1. Genes Dev. 5:820-826.[Abstract/Free Full Text]
    43. 43
  43. Ustav, M., and A. Stenlund. 1991. Transient replication of BPV-1 requires two viral polypeptides encoded by the E1 and E2 open reading frames. EMBO J. 10:449-457.[Medline]
    44. 44
  44. Vilar, J. M. G., and S. Leibler. 2003. DNA looping and physical constraints on transcription regulation. J. Mol. Biol. 331:981-989.[CrossRef][Medline]
    45. 45
  45. Wang, Y., R. Coulombe, D. R. Cameron, L. Thauvette, M.-J. Massariol, L. M. Amon, D. Fink, S. Titolo, E. Welchner, C. Yoakim, J. Archambault, and P. W. White. 2004. Crystal structure of the E2 transactivation domain of human papillomavirus type 11 bound to a protein interaction inhibitor. J. Biol. Chem. 279:6976-6985.[Abstract/Free Full Text]
    46. 46
  46. You, J., J. L. Croyle, A. Nishimura, K. Ozato, and P. M. Howley. 2004. Interaction of the bovine papillomavirus E2 protein with Brd4 tethers the viral DNA to host mitotic chromosomes. Cell 117:349-360.[CrossRef][Medline]
    47. 47
  47. Zeller, R. W., J. D. Griffith, J. G. Moore, C. V. Kirchhamer, R. J. Britten, and E. H. Davidson. 1995. A multimerizing transcription factor of sea urchin embryos capable of looping DNA. Proc. Natl. Acad. Sci. USA 921:2989-2993.

Journal of Virology, May 2008, p. 4853-4861, Vol. 82, No. 10
0022-538X/08/$08.00+0     doi:10.1128/JVI.02388-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Cardenas-Mora, J., Spindler, J. E., Jang, M. K., McBride, A. A. (2008). Dimerization of the Papillomavirus E2 Protein Is Required for Efficient Mitotic Chromosome Association and Brd4 Binding. J. Virol. 82: 7298-7305 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hernandez-Ramon, E. E.
Right arrow Articles by Maitland, N. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hernandez-Ramon, E. E.
Right arrow Articles by Maitland, N. J.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»