MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
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 Kondo, E.
Right arrow Articles by Fukushige, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kondo, E.
Right arrow Articles by Fukushige, S.

 Previous Article  |  Next Article 

Molecular and Cellular Biology, June 2005, p. 4388-4396, Vol. 25, No. 11
0270-7306/05/$08.00+0     doi:10.1128/MCB.25.11.4388-4396.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.

The Thymine DNA Glycosylase MBD4 Represses Transcription and Is Associated with Methylated p16INK4a and hMLH1 Genes

Emiko Kondo, Zhaodi Gu, Akira Horii, and Shinichi Fukushige*

Department of Molecular Pathology, Tohoku University School of Medicine, 2-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8575, Japan

Received 20 August 2004/ Returned for modification 14 October 2004/ Accepted 7 March 2005


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Epigenetic silencing through methyl-CpG (mCpG) is implicated in many biological patterns such as genome imprinting, X chromosome inactivation, and cancer development. In this process, the mCpG binding domain (MBD) proteins play an essential role in transmitting epigenetic information to downstream regulatory proteins. Among the five MBD proteins identified so far, MBD4 has been the only exception; it has long been thought to be a DNA repair protein. Herein we demonstrate that MBD4 has the ability to repress transcription through mCpG. Transcriptional repression by the MBD4 is histone deacetylase (HDAC) dependent, and MBD4 directly binds to Sin3A and HDAC1 at three central regions that overlap transcriptional repression domains. Furthermore, a chromatin immunoprecipitation assay clearly shows that MBD4 binds to hypermethylated promoters of the p16INK4a and hMLH1 genes. These results suggest that MBD4 is one of the essential components involved in epigenetic silencing in cancer and its repair activity is necessary for the maintenance of hypermethylated promoters.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Methyl-CpG (mCpG) binding domain protein 4 (MBD4) has been suggested to function as a tumor suppressor because MBD4 is frequently mutated in human colorectal tumors with microsatellite instability (3, 26). However, there are controversial reports about its function in tumorigenesis; MBD4 mutations are generally monoallelic, predominantly at the poly(A) tract, and have not been detected in DNA mismatch repair-proficient tumors. Moreover, a recent study demonstrated that loss of MBD4 does not alter the spontaneous mutation rate, the tumor onset, or the tumor spectrum in mismatch repair-deficient mice (29). Therefore, there is the possibility that MBD4 mutations by themselves may simply reflect microsatellite instability in mismatch repair-deficient tumors and may have no relationship with cancer development in these tumors. Thus, it is necessary to elucidate the function of MBD4 to understand its impact on human carcinogenesis.

Methylation at CpG dinucleotides in genomic DNA is a fundamental modification in epigenetic gene silencing (13). Recently a correlation between DNA hypermethylation, hypoacetylation of histones, tightly packed chromatin, and transcriptional repression has been demonstrated. Signals of DNA methylation are mediated through a protein family that binds to symmetrically methylated CpGs. Such proteins contain a specific domain, called the methyl-CpG (mCpG) binding domain (MBD), which consists of {approx}70 amino acid residues in an {alpha}/ß-sandwich fold built of three to four ß-twisted sheets and a helix with a characteristic hairpin loop in the opposite layer (5). So far, five MBD proteins have been identified: MBD1, MBD2, MBD3, MBD4, and MeCP2 (methyl CpG binding protein 2). Four of them are known to be associated with transcriptional repression. The remaining MBD protein, MBD4 contains a C-terminal DNA glycosylase catalytic domain in addition to an N-terminal MBD and thus it has been thought to be involved in DNA repair rather than transcriptional repression (6, 11).

In fact, in vitro biochemical analysis has shown that MBD4 is a thymine and uracil glycosylase specific for G-T and G-U mismatches resulting from the deamination of 5-methylcytosine and cytosine, respectively, at CpG sites (12). In addition, an increase in 5-methylcytosine to T mutations in Mbd4–/– Big Blue mice and an increased occurrence of colorectal tumors in Mbd4–/– ApcMin/+ mice has been demonstrated, however, the frequency of C to T transition mutations at CpG sites was increased only by a factor of two to three (20, 34). The relatively moderate mutator phenotype of Mbd4–/– mice probably indicates that another thymine DNA glycosylase, TDG, or equivalent glycosylases play a redundant role to repair G-T mismatches. Because spontaneous deamination of 5-methylcytosine occurs at a very high frequency in the genome, organisms appear to have evolved a redundant repair system to maintain original sequence information. In addition to G-T mismatch repair activity, MBD4 has shown to have 5-methylcytosine DNA glycosylase activity (37). The activity of 5-methylcytosine DNA glycosylase present in MBD4 was about 30 times lower than the thymine DNA glycosylase, and the biological significance of 5-methylcytosine DNA glycosylase activity has not yet been elucidated.

MBD4 is expressed in numerous human tissues, including ovary and testis, tissues that produce germ cells (11). MBD4-green fluorescent protein (GFP) localized within the foci of heavily methylated satellite DNA and this localization is disrupted in DNA methyltransferase-deficient embryonic stem (ES) cells that have a reduced level of CpG methylation (11). Although MBD4 was first identified as an MBD-containing protein with a region of similarity to bacterial DNA repair enzymes, it was independently isolated by a two-hybrid screening as one of the molecules that interacts with human mut L homolog 1 (hMLH1), the major mismatch repair protein (6). This interaction therefore suggested that MBD4 might function in mismatch repair. Recently, Cortellino et al. reported that MBD4 affects the cell cycle in the similar manner as mismatch repair components (8); Mbd4-deficient cells failed to undergo G2/M cell cycle arrest and apoptosis upon treatment with DNA-damaging agents. In addition, the amounts of several mismatch repair proteins are reduced in Mbd4-deficient cells due to a posttranscriptional mechanism. These observations may suggest that the interaction between MBD4 and hMLH1 plays an essential role to perform the mismatch repair-dependent DNA damage checkpoint (28).

Although the only function of MBD4 has long been thought to be a thymine DNA glycosylase that repairs G-T mismatches originated by deamination of 5-methylcytosine at CpG sites, the C-terminal catalytic domain of MBD4 has the same substrate specificity as the intact protein. Therefore, the significance of tethering the mCpG binding and glycosylase domains within the same molecule remains to be elucidated (25). In addition, physical and functional linkages between DNA glycosylases and proteins that regulate gene transcription have recently been discovered. For example, another thymine DNA glycosylase, TDG, has been demonstrated to both activate and repress transcription by using a variety of mechanisms (21, 31). Methylpurine DNA glycosylase has a synergistic effect on gene silencing together with MBD1 and also interacts with estrogen receptor {alpha} to inhibit transcription (17, 33). In Arabidopsis thaliana, the DEMETER (DME) gene, which encodes DNA glycosylase, activates maternal allele expression of the imprinted MEDEA (MEA) Polycomb gene in the central cell (7). Thus, DNA glycosylases are linked to the process of gene regulation by physical interactions that modulate the activities of transcription factors, receptors, and chromatin-remodeling factors. Therefore, we tried to test a possible involvement of MBD4 in transcriptional regulation through mCpGs.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Plasmids. Escherichia coli strain DH5{alpha}F' was used to propagate all plasmids. Human colon and liver cDNA libraries were purchased from Stratagene. A fetal brain cDNA library was a generous gift from K. Yamakawa (RIKEN, Wako, Japan) (35). These libraries were amplified, and the DNAs were purified (pooled cDNA) by the methods described previously (36). The MBD4, MBD2a, MBD2b, MeCP2, Sin3A, and HDAC1 cDNA clones were PCR amplified by using the pooled cDNA as the template, and the primer sequences were designed on the basis of sequence database (GenBank accession numbers: NM_003925 for MBD4, NM_003927 for MBD2a, XM_054399 for MBD2b, NM_004992 for MeCP2, NM_015477 for Sin3A, and NM_004964 for HDAC1).

The sets of primers used are as follows: MBD4-F1 (5'-GATGGATCCATGGGCACGACTGGGCTG-3') and MBD4-R1 (5'-GATCTCGAGGATGAGCTTGAAAGCTGCAG-3'), MBD2a-F1 (5'-GATGGATCCATGCGCGCGCACCCGGG-3') and MBD2-R1 (5'-GATCTCGAGTTAGGCTTCATCTCCACTGTC-3'), MBD2b-F1 (5'-GATGGATCCATGGATTGCCCGGCCCTC-3') and MBD2-R1, MeCP2-F1 (5'-GATGGATCCATGGTAGCTGGGATGTTAGG-3') and MeCP2-R1 (5'-GATCTCGAGTCAGCTAACTCTCTCGGTCAC-3'), Sin3A-F1 (5'-GATCTCGAGATGAAGCGGCGTTTGGATGAC-3') and Sin3A-R1 (5'-GATGCGGCCGCTTAAGGGGCTTTGAATACTGTG-3'), HDAC1-F1 (5'-GATCTCGAGATGGCGCAGACGCAGGGC-3'), and HDAC1-R1 (5'-GATGCGGCCGCTCAGGCCAACTTGACCTCC-3'). These primers contain BamHI or XhoI or NotI site in the primers to facilitate cloning (italic).

A reporter plasmid pSF100 (L0), which contains the ß-galactosidase gene under the control of the human ß-actin promoter, was constructed from pSF18 described previously (9). Plasmid pSF18 was a generous gift from B. Sauer (Stowers Institute for Medical Research, Kansas City, MO). Another reporter plasmid, pSF100-LexA (L8), was also constructed by inserting 8 LexA operators in front of the ß-actin promoter. An effector plasmid, pcDNA-LexA-MBD4, was constructed by inserting the LexA sequence into HindIII and BamHI sites and by inserting MBD4 cDNA into BamHI and XhoI sites of pcDNA3.1/V5-His (Invitrogen). DNA fragments containing the promoter regions of human p16INK4a and MLH1 genes (GenBank accession numbers: X94154 for p16INK4a and U83845 for hMLH1) were amplified from human genomic DNA and subcloned into pSF100 in the place of the ß-actin promoter.

The sets of primers used are as follows: p16 pro-F1 (5'-GATGGCGCGCCACGGCGTCCCCTTGCCTGG-3') and p16 pro-R1 (5'-GATAAGCTTTCCATGCTGCTCCCCGCCGC-3'), hMLH1 pro-F1 (5'-GATGGCGCGCCGCTCGTAGTATTCGTGC-3'), and hMLH1 pro-R1 (5'-GATAAGCTTAGAAGAGCCAAGGAAACGTC-3'). These primers contain AscI and HindIII sites in the forward and reverse primers, respectively, to facilitate cloning (italic). We performed sequencing analysis in all the constructs to confirm identity. Plasmid pGV-C2 (Toyo Ink) was also used as an internal control reporter.

Transfection and reporter assays. HEK293T cells were seeded on six-well tissue culture dishes and were transfected with reporter (0.15 µg; with or without LexA operators), effector (1 µg; LexA-MBD proteins or deletion constructs), and internal control reporter (0.15 µg; simian virus 40 promoter/enhancer driving luciferase). Transfection was performed using the Lipofectamine reagents (Invitrogen) according to the supplier's recommendations. The cells were harvested 48 h after transfection. Then the ß-galactosidase and luciferase assays were performed as described previously (9, 23). The transfected cells were treated with trichostatin A (0.1 µg/ml) for 24 h before harvest. Transfection of CHO-K1 cells (obtained from Cell Resource Center for Biomedical Research, Institute of Development, Aging, and Cancer, Tohoku University) with reporter (0.5 µg; ß-galactosidase gene under the control of p16INK4a or hMLH1 promoter), effector (1 µg; MBD proteins), and internal control reporter (0.15 µg) was also performed. pSF100 reporter plasmid with p16INK4a or hMLH1 promoter was either mock-methylated (no enzyme added) or M.SssI (New England Biolabs) methylated. Complete methylation was checked by restriction digestion with BstUI (New England Biolabs).

Immunoprecipitation. Immunoprecipitation was performed as described previously (15). The whole-cell extracts of HEK293T cells were immunoprecipitated using anti-MBD4 (H300) antibody (Santa Cruz) and analyzed by Western blotting using anti-mSin3A (K-20) polyclonal antibody (Santa Cruz) and anti-HDAC1 monoclonal antibody, clone 2E10 (Upstate), respectively. The HEK293T cells grown in 10-cm dishes were also transfected either with 4 µg of pFLAG-CMV-2 vector or pFLAG-MBD4 DNA. At 48 h posttransfection, cells were lysed, immunoprecipitated using anti-mSin3A (K-20) and anti-HDAC1 antibodies, and analyzed by Western blotting using anti-FLAG monoclonal antibody M2 (Sigma).

In vitro binding assay. The glutathione S-transferase (GST)-in vitro transcription and translation assay was performed as described previously (15). Glutathione S-transferase (GST) fusion proteins of MBD4 containing various regions amplified by PCR with Thermococcus kodakaraensis polymerase (Toyobo) were made using a pGEX-2TK vector (Amersham Pharmacia Biotech). Deletion constructs of Sin3A and HDAC1 containing various regions were also cloned into pcDNA3.1/V5-His.

Reverse transcription-PCR assay. Total RNAs were isolated from HeLa, DLD1, AN3CA, and HEK293T cells by RNeasy mini kit (QIAGEN). Multiplex PCR amplifications were performed in samples using primer sets for B2M (ß2 microglobulin) and p16INK4a or hMLH1. Nucleotide sequences for the primers are as follows: B2M-C1, 5'-GTGGAGCATTCAGACTTGTC-3'; B2M-C2, 5'-CAAACATGGAGACAGCACTC-3'; p16-C1, 5'-TGGCTGAGGAGCTGGGCAT-3'; p16-E3R, 5'-TTCCCGAGGTTTCTCAGAG-3'; hMLH1-LHF10, 5'-TCTGATTGACAACTATGTGCC-3'; and hMLH1-LHR6, 5'-GTTTGGAATGGAGCCAGGC-3'. The PCR products were analyzed on 3% agarose gels, and the bands were visualized by ethidium bromide staining.

Chromatin immunoprecipitation assay. Chromatin immunoprecipitations were performed as described previously (2) with HeLa, DLD1, AN3CA, and HEK293T cells. Chromatin was sonicated to an average length of 1 kb using a Tomy Ultrasonic Disruptor UD-200 with microtip. Sonicated chromatin was subjected to immunoprecipitation using a protocol provided by Upstate. Antibodies used for chromatin immunoprecipitation were anti-normal rabbit immunoglobulin G (Calbiochem), anti-MBD4 (H300), anti-MBD2 (Upstate), anti-dimethyl-histone H3 (Lys9) clone RR202 (Upstate), anti-acetyl-histone H3 (Upstate), and anti-acetyl-histone H4 (Upstate). PCR amplification was performed in 15 µl with the p16INK4a CpG island primers (18) and with the hMLH1 primers (14).

RNA interference. The sense sequence (5'-GGACUGAAGUUCAGAUCCA-3') corresponding to nucleotides 343 to 361 relative to the start codon was chosenas short interfering RNA (siRNA) targeting MBD4, and a double-stranded siRNA was synthesized by Japan Bioservice. The HeLa, DLD1, HEK293T, and AN3CA cells were seeded on six-well tissue culture dishes and were transfected with mock (without siRNA) or MBD4 siRNA (final concentration, 200 nM). Transfection was performed using the Oligofectamine reagents (Invitrogen) according to the supplier's recommendations. The cells were harvested 48 h after transfection and were analyzed by Western blotting using anti-MBD4 antibody (Abcam) and anti-ß-actin monoclonal antibody clone AC-15 (Sigma).


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
MBD4 represses transcription. To determine whether MBD4 has some ability to regulate transcription, we transiently transfected human HEK293T cells with a reporter plasmid, which either did (L8) or did not (L0) have LexA operator sequences in front of the human ß-actin promoter, in the presence or absence of three LexA-MBD fusion constructs (Fig. 1A). In the absence of MBD proteins, the ratio of L8 to control L0 expression was unaffected and was close to 1.0 (Fig. 1B). In the presence of the LexA-MBD4 protein, the L8/L0 expression ratio was between 0.42 and 0.5. These results indicate that MBD4, as well as two control repressors, MBD2b and MeCP2, has the ability to repress transcription (see Fig. 1B). Transcriptional repression activity of MBD4 was also observed when we used HeLa cells (data not shown). The repression activity of MBD4 is dose dependent and increased in parallel with the amounts of the effector plasmid (Fig. 1C). Next, we asked where the transcriptional repression domains exist in MBD4 protein. A tethering assay using deletion constructs containing various regions indicated that the central regions of MBD4 (residues 76 to 151 corresponding to MBD and residues 152 to 454) excluding the C-terminal glycosylase catalytic domain are transcriptional repression domains (Fig. 1D and 3A).



View larger version (27K):
[in this window]
[in a new window]
  		FIG. 1. Transcriptional repression activity and the location of transcriptional repression domains in MBD4. (A) Schematic diagram of two reporters that did or did not contain LexA operators (L8 and L0, respectively) and an effector that contains a fusion between the LexA and MBD proteins. (B to D) Human HEK293T cells were cotransfected with a reporter, an effector, and an internal control plasmid. (C) Dose dependency of transcriptional repression by the MBD4 is shown. The total amount of effector plasmid was normalized to 1.0 µg with insertless vector pcDNA-LexA. (D) Effectors containing various regions of MBD4 were used to identify transcriptional repression domains. Columns represent the mean ratio between ß-galactosidase activities of L8 and L0 with standard deviation in three independent experiments. Mock represents relative ß-galactosidase activity (L8/L0) in the effector vector without MBD proteins. 8 LexAop, 8 LexA operators; ß-actin pro, human ß-actin promoter; CMV pro, cytomegalovirus promoter.

 


View larger version (26K):
[in this window]
[in a new window]
  		FIG. 3. Domains of MBD4 interacting with Sin3A or HDAC1. (A) Maps of the MBD4 deletion constructs, indicating truncated proteins that did (+) or did not (–) bind to Sin3A and HDAC1. Regions representing transcriptional repression activity (TRDs) and regions of strong binding to Sin3A and HDAC1 (Sin3A and HDAC1 IDs) are indicated by the thick black lines (top). ND, not done. (B) In vitro-translated 35S-labeled Sin3A and HDAC1 were incubated with an immobilized fusion protein containing GST and different portions of MBD4. Numbers correspond to amino acid positions in the protein. Input (10% of total) protein and protein bound by GST alone and protein bound by each GST-MBD4 fusion protein were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The expression levels of each GST fusion protein were determined by immunoblot analysis using anti-GST antibody (Zymed). Asterisks indicate fusion proteins corresponding to expected protein sizes. Each lane also contains some degradation products.

 
Transcriptional repression of MBD4 is HDAC dependent. Because it is widely known that the HDAC-dependent pathway is involved in transcriptional repression by most MBD proteins (22, 23, 27), we next asked whether this is the case for MBD4 by using the HDAC inhibitor trichostatin A. The tethering assay described in Fig. 1 was performed in the presence or absence of trichostatin A, and transcriptional repression by each of the LexA-fused full-length MBD4 (residues 1 to 580), the N-terminal MBD (residues 76 to 151), and the C-terminal MBD4 lacking MBD (residues 152 to 580) was completely relieved by the trichostatin A treatment (Fig. 2A). This result shows that the transcriptional repression activity of MBD4 is mainly HDAC dependent.



View larger version (30K):
[in this window]
[in a new window]
  		FIG. 2. HDAC-dependent transcriptional repression by MBD4. (A) HEK293T cells were transfected as in Fig. 1. After 24 h, transfected cells were treated with trichostatin A (shaded bars) or left untreated (white bars) and incubated for a further 24 h. Relative ß-galactosidase activities were also obtained as in Fig. 1. (B) Western blots of immunoprecipitates prepared using antibodies against MBD4 and control normal rabbit immunoglobulin G. An aliquot of input protein (1%) was used as the control. Probes were anti-mSin3A (upper panel) and anti-HDAC1 (lower panel) antibodies. (C) Western blots of immunoprecipitates prepared using antibodies against Sin3A (upper panel) or HDAC1 (lower panel). HEK293T cells were transfected with either pFLAG-CMV-2 vector or pFLAG-MBD4. The probe was an anti-FLAG antibody. An aliquot of input protein (1%) was used as the control.

 
Therefore, we performed the immunoprecipitation experiments to determine whether MBD4 forms a protein complex with corepressor Sin3A and HDAC1, which are two major components in the HDAC-dependent pathway in vivo (1, 10). The immunoprecipitation was performed with anti-MBD4 antibody and control normal rabbit immunoglobulin G, followed by immunoblotting with anti-mSin3A or anti-HDAC1 antibody. The anti-MBD4 antibody precipitated Sin3A as well as HDAC1 (Fig. 2B). To confirm this result, we expressed a FLAG-tagged MBD4 construct in HEK293T cells, immunoprecipitated with either anti-mSin3A or anti-HDAC1 antibodies, and detected with anti-FLAG antibodies (Fig. 2C). Again, both anti-mSin3A as well as anti-HDAC1 antibody precipitated FLAG-tagged MBD4. These results clearly indicate that MBD4 forms a protein complex with Sin3A and HDAC1. The immunostaining with each fluorescent-labeled antibody for MBD4, Sin3A, and HDAC1 showed a similar staining pattern distributed throughout the whole nucleus (data not shown), however, the precise colocalization of MBD4, Sin3A, and HDAC1 has not yet been elucidated.

Regions of MBD4 for interacting with Sin3A and HDAC1. We next determined the possibility of direct binding between MBD4 and Sin3A or HDAC1 by using an in vitro GST pulldown assay. As indicated in Fig. 3, both Sin3A and HDAC1 directly interacted with MBD4. A detailed in vitro GST pulldown assay with various deletion constructs indicates that the same three regions located at the central region of MBD4 (residues 76 to 151, 212 to 352, and 413 to 454), which also overlapped with transcriptional repression domains, bind to Sin3A and HDAC1 (Fig. 3).

Regions of Sin3A and HDAC1 for interacting with MBD4. Furthermore, we determined domains on Sin3A and HDAC1 interacting with MBD4. An in vitro GST pulldown assay shows that two regions (residues 524 to 679 and 1014 to 1273) of Sin3A interact with MBD4 (Fig. 4A). Interestingly, these two regions are exactly the same ones on Sin3A interacting with MeCP2 (22). Therefore, it is possible that the structure that is an interacting interface with Sin3A is the same or very similar between MBD4 and MeCP2. The in vitro GST pulldown assay also showed that the interacting domain of HDAC1 with MBD4 is an HDAC family motif (residues 26 to 303) (Fig. 4B). This motif is a conserved domain in all known members of the class I HDACs, suggesting the possibility that MBD4 could bind to other members of the HDAC family proteins.



View larger version (31K):
[in this window]
[in a new window]
  		FIG. 4. Domains of Sin3A (A) and HDAC1 (B) interacting with MBD4. Maps of the Sin3A (A) or the HDAC1 (B) deletion constructs, indicating truncated proteins that did (+) or did not (–) bind to MBD4. Regions of strong binding to MBD4 are indicated by the thick black lines (top). Deletion constructs of Sin3A and HDAC1 were transcribed and translated in vitro to give 35S-labeled proteins and processed as described in Fig. 3. G, protein bound by GST alone; B, protein bound by the MBD4; I, input protein (10% of total).

 
Presence of MBD4 in hypermethylated p16INK4a and hMLH1 promoters. The data presented in the early literature showed that the subcellular localization of MBD2 and MBD4 among MBD proteins was very similar and concentrated on heavily methylated regions (11). Because MBD2 binds to hypermethylated promoters in cancer-related genes, we sought to test the possible involvement of MBD4 in this context. To analyze the involvement of MBD4 in hypermethylated promoters seen in cancer-related genes, we performed a chromatin immunoprecipitation assay using cell lines carrying hypermethylated promoters of p16INK4a and hMLH1 genes, which are both highly methylated in many cancers. Because a previous report had shown that MBD2 protein could bind to hypermethylated p16INK4a promoter (18), we used MBD2 as the positive control. In the DLD1 cell, which contains hypermethylated p16INK4a promoter (18, 24), MBD4 as well as MBD2 bound to its promoter sequence, whereas, in the HeLa cell that does not contain hypermethylated p16INK4a promoter (18), and therefore expresses p16INK4a mRNA, neither MBD4 nor MBD2 bound to its promoter (Fig. 5A).



View larger version (50K):
[in this window]
[in a new window]
  		FIG. 5. Chromatin immunoprecipitation analysis of the occupancy by MBD4 and histone modification status of hypermethylated p16INK4a and hMLH1 promoters. (A) p16INK4a transcripts were coamplified by RT-PCR, with the B2M transcripts as an internal control. PCR products (181 bp for p16INK4a and 341 bp for B2M) were separated by electrophoresis on a 3% agarose gel (left panel). PCR analysis of DNA in chromatin immunoprecipitated with anti-MBD4, anti-acetylated histone H3 (AcH3) and H4 (AcH4), and anti-dimethyl histone H3 Lys9 (MeH3) for the p16INK4a promoter region (395 bp) was performed on DLD1 and HeLa cells (right panel). The input and two negative controls, bound fractions of the no-antibody (NAB) and the normal rabbit immunoglobulin G, are also shown. (B) Expression of hMLH1 was analyzed by RT-PCR after coamplification of hMLH1 transcript (208 bp) with B2M (341 bp) using HEK293T, AN3CA, and HeLa cells (left panel). As in A, binding to MBD4, the acetylation status of histone H3 and H4, and methylation status of histone H3 Lys9 for the hMLH1 promoter region (606 bp) were also performed using the chromatin immunoprecipitation assay (right panel). Lane M contains MspI-digested pBluescript SK(+) (Stratagene) DNA and was used as the size marker for agarose gel electrophoresis.

 
To see the relationship between the binding to MBD4 protein, acetylation, and methylation in histone proteins, we have performed the chromatin immunoprecipitation assay using anti-acetylated histone H3 and H4 antibodies as well as anti-dimethylated histone H3 Lys9 antibody. The result clearly shows that histone H3 and H4 are deacetylated and histone H3 Lys9 is methylated in hypermethylated p16INK4a promoters of DLD1 cells as described previously (16, 18). Therefore, these results show a clear correlation between MBD4 binding, deacetylation of histone H3 and H4, and methylation of histone H3 Lys9. In addition, we analyzed the involvement of MBD4 protein in the hypermethylated hMLH1 promoter. Again, MBD4 bound to hypermetylated hMLH1 promoters in HEK293T (30) as well as AN3CA cells (32), and these promoters have deacetylated histone H3 and H4 and dimethylated histone H3 Lys9 (Fig. 5B).

To further confirm the binding of MBD4 protein in the hypermethylated p16INK4a and hMLH1 promoters, we expressed the FLAG-tagged MBD4 construct as well as the FLAG vector in HeLa, DLD1, HEK293T, and AN3CA cells and performed the chromatin immunoprecipitation assay using anti-FLAG antibody. Again, FLAG-MBD4 specifically bound to hypermethylated p16INK4a promoter in DLD1 cell and to hypermethylated hMLH1 promoters in HEK293T as well as AN3CA cells, but not to unmethylated promoters in HeLa cell (data not shown).

Transcriptional repression of hypermethylated p16INK4a and hMLH1 promoters by MBD4 in mammalian cells. To further confirm the link between the mCpG binding ability and the transcriptional repression activities of MBD4, we cloned p16INK4a and hMLH1 promoters upstream of ß-galactosidase reporter gene, methylated them by SssI methyltransferase, and analyzed the effects of MBD2 and MBD4 proteins (Fig. 6). Without methylation, neither MBD4 nor MBD2 have any effects on reporter gene expression located downstream of p16INK4a (Fig. 6A) and hMLH1 promoters (Fig. 6B). However, with methylation, MBD4 and MBD2 could repress transcription from both the p16INK4a and hMLH1 promoters (Fig. 6C). This result was further confirmed by using the constructs in which only the p16INK4a and hMLH1 promoters are SssI methylated and then ligated to the backbone plasmid (data not shown). In this experiment, both MBD4 and MBD2 proteins could repress the ß-galactosidase activities by more than 90% compared with the mock transfection.



View larger version (26K):
[in this window]
[in a new window]
  		FIG. 6. Effect of MBD4 and MBD2 on unmethylated and methylated p16INK4a and hMLH1 promoters. Unmethylated (M-) or SssI-methylated (M+) pSF100 reporter with the p16INK4a or hMLH1 promoter was cotransfected with pGV-C2 and MBD2a- or MBD4-expressing plasmid (pcDNA-MBD2a or pcDNA-MBD4) or insertless plasmid pcDNA3.1/V5-His. The relative ß-galactosidase activity of unmethylated or methylated pSF100 in combination with pGV-C2 and pcDNA3.1/V5-His (mock) was normalized to 100, and the relative ß-galactosidase activities (mean plus standard deviation) were determined after correcting the transfection efficiency by pGV-C2.

 
In addition, the knockdown of MBD4 was performed by siRNA technology to see whether MBD4 by itself could play an essential role in transcriptional repression from the hypermethylated p16INK4a and hMLH1 promoters in mammalian cells (Fig. 7). The MBD4 siRNA could reduce the protein level up to 79% in AN3CA, 50% in HEK293T, and 70% in DLD1 cells. However, the recovery of p16INK4a and hMLH1 expression in these knockdown cells was not observed.



View larger version (45K):
[in this window]
[in a new window]
  		FIG. 7. Influence of siRNA targeting MBD4 on the suppression of MBD4 expression at protein level in cells containing the hypermethylated p16INK4a and hMLH1 promoters. (A) Treatment with mock or siRNA targeting MBD4 was performed using HeLa, DLD1, AN3CA, and HEK293T cells and the immunoblot assay for MBD4 and ß-actin was performed 48 h after transfection. The reduction rate of MBD4 protein by siRNA shown below the gel was calculated by using an LAS-1000 Plus (Fujifilm). (B) p16INK4a (left panel) and hMLH1 (right panel) transcripts were coamplified by RT-PCR with the B2M transcript as described for Fig. 5 using HeLa, DLD1, AN3CA, and HEK293T cells treated with siRNA targeting MBD4.

 

   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In this study, we found that MBD4 has an additional role in transcriptional repression through methyl-CpG. Our results presented here seem to teleologically oppose a mutation avoidance function of MBD4 because transcriptional repression at hypermethylated promoters of cancer-related genes leads to cancer development. However, given the fact that all MBD proteins, including MBD4, have transcriptional repression activity, their inherent role in cancer could be very similar.

In order to understand the role of MBD4 in cancer, it is important to consider the linkage between the two functional domains, MBD and the glycosylase catalytic domain, because these two domains are tethered within the same molecule. An early report showed that MBD4 as well as MBD2 among other MBD proteins colocalizes to highly methylated satellite DNA in murine cells (11). The MeCP1 complex, which contains MBD2 as a mediator of mCpG information, requires at least 15 consecutive mCpGs for binding (19), and MBD2 is involved in hypermethylated promoters in cancer-related genes such as p16INK4a, BRCA1, and MGMT (4, 18). In this context, it is conceivable that MBD4 also stays at the hypermethylated promoters of cancer-related genes and that the role of the glycosylase catalytic domain may be the maintenance of mCpG sites to allow preferential binding. Our discovery that MBD4 in fact binds to the hypermethylated promoters in p16INK4a and hMLH1 genes and has the ability to repress transcription strongly supports this idea.

As shown in Fig. 7, the knockdown of MBD4 in DLD1, AN3CA, and HEK293T cells by siRNA did not change the repression status in p16INK4a and hMLH1 genes. In addition, we constructed an MBD4 knockout cell line originated from DLD1 cell by using the conventional homologous recombination method and examined the recovery of p16INK4a expression. However, the MBD4 knockout cell did not change the repression status in the p16INK4a gene again (data not shown). There are several possibilities to explain these results. First, each MBD protein is suggested to play a redundant role and the hypermethylated promoters of p16INK4a and hMLH1 may be occupied by several different MBD proteins. In fact, as shown in Fig. 5, both MBD4 and MBD2 bound to the hypermethylated promoter of the p16INK4a gene. Second, as suggested by Bachman et al., histone modification (especially histone H3 methylation) is important to repress transcription from the hypermethylated promoters and the role of MBD proteins accompanying DNA methylation may be the maintenance of repression status established by histone modifications (2).

As shown in Fig. 3, the central region of MBD4 (residues 152 to 454) interacted with both Sin3A and HDAC1 and overlapped with one of the transcriptional repression domains. However, as described by Hendrich et al., the central region of MBD4 is less conserved between human and mouse compared with the MBD and glycosylase domains (12). To clarify whether the central region of MBD4 in the mouse (residues 139 to 428 corresponding to residues 152 to 454 in the human) is involved in transcriptional repression, we tested its interacting ability with Sin3A and HDAC1 by the GST pulldown assay. The central region of mouse MBD4 also interacted with Sin3A and HDAC1, suggesting the conservation of binding ability of this region between human and mouse to both proteins (data not shown).

It is interesting that MBD4 represses hypermethylated hMLH1 promoters because MBD4 interacts with hMLH1 (6). The interacting domain of MBD4 with hMLH1 is located at residues 413 to 454 (Kondo et al., unpublished observation) and this region corresponds to one of the domains interacting with Sin3A and HDAC1. Therefore it is possible to think that transcriptional repression of hMLH1 gene by the MBD4 further accelerates the binding of MBD4 for Sin3A and HDAC1. Further analysis would clarify whether hMLH1 regulates transcriptional repression by the MBD4.

Recently, Ballestar et al. provided the evidence that gene-specific profiles of MBDs exist for each hypermethylated promoter, whereas a common pattern of histone modifications is shared (4). In addition, they used the combination of chromatin immunoprecipitation assay of MBDs and a CpG island microarray to identify novel targets of epigenetic silencing in human cancer. This approach is applicable for MBD4 to discover cancer-causing hypermethylation. The role of MBD4 in cancer and other biological aspects such as genome imprinting and heterochromatin formation should be examined in detail, taking account of the transcriptional repression activity identified in this study.


   ACKNOWLEDGMENTS
 
We thank B. Sauer for the plasmid, K. Yamakawa for fetal brain cDNA library, and B. L. S. Pierce (a professor with the University of Maryland University College) for editorial work in the preparation of the manuscript.

This work was supported by Grant-in-Aid 12213010 from the Ministry of Education, Culture, Sports, Science and Technology of Japan to A.H.


   FOOTNOTES
 
* Corresponding author. Mailing address: Department of Molecular Pathology, Tohoku University School of Medicine, 2-1 Seiryo-machi, Aoba-ku, Sendai, Miyagi 980-8575, Japan. Phone: 81 22 717 8043. Fax: 81 22 717 8047. E-mail: shinichi{at}mail.tains.tohoku.ac.jp. Back


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 
1. Alland, L., R. Muhle, H. Hou, Jr., J. Potes, L. Chin, N. Schreiber-Agus, and R. A. DePinho. 1997. Role for N-CoR and histone deacetylase in Sin3-mediated transcriptional repression. Nature 387:49-55.[CrossRef][Medline]

2. Bachman, K. E., B. H. Park, I. Rhee, H. Rajagopalan, J. G. Herman, S. B. Baylin, K. W. Kinzler, and B. Vogelstein. 2003. Histone modifications and silencing prior to DNA methylation of a tumor suppressor gene. Cancer Cell. 3:89-95.[CrossRef][Medline]

3. Bader, S., M. Walker, B. Hendrich, A. Bird, C. Bird, M. Hooper, and A. Wyllie. 1999. Somatic frameshift mutations in the MBD4 gene of sporadic colon cancers with mismatch repair deficiency. Oncogene 18:8044-8047.[CrossRef][Medline]

4. Ballestar, E., M. F. Paz, L. Valle, S. Wei, M. F. Fraga, J. Espada, J. C. Cigudosa, T. H.-M. Huang, and M. Esteller. 2003. Methyl-CpG binding proteins identify novel sites of epigenetic inactivation in human cancer. EMBO J. 22:6335-6345.[CrossRef][Medline]

5. Ballestar, E., and A. P. Wolffe. 2001. Methyl-CpG-binding proteins. Targeting specific gene repression. Eur. J. Biochem. 268:1-6.[Medline]

6. Bellacosa, A. L. Cicchillitti, F. Schepis, A. Riccio, A. T. Yeung, Y. Matsumoto, E. A. Golemis, M. Genuardi, and, G., Neri. 1999. MED1, a novel human methyl-CpG-binding endonuclease, interacts with DNA mismatch repair protein MLH1. Proc. Natl. Acad. Sci. USA 96:3969-3974.[Abstract/Free Full Text]

7. Choi, Y., M. Gehring, L. Johnson, M. Hannon, J. J. Harada, R. B. Goldberg, S. E. Jacobsen, and R. L. Fischer. 2002. DEMETER, a DNA glycosylase domain protein, is required for endosperm gene imprinting and seed viability in Arabidopsis. Cell 110:33-42.[CrossRef][Medline]

8. Cortellino, S., D. Turner, V. Masciullo, F. Schepis, D. Albino, R. Daniel, A. M. Skalka, N. J. Meropol, C. Alberti, L. Larue, A. Bellacosa. 2003. The base excision repair enzyme MED1 mediates DNA damage response to antitumor drugs and is associated with mismatch repair system integrity. Proc. Natl. Acad. Sci. USA 100:15071-15076.[Abstract/Free Full Text]

9. Fukushige, S., and, B., Sauer. 1992. Genomic targeting with a positive-selection lox integration vector allows highly reproducible gene expression in mammalian cells. Proc. Natl. Acad. Sci. USA 89:7905-7909.[Abstract/Free Full Text]

10. Heinzel, T., R. M. Lavinsky, T.-M. Mullen, M. Söderström, C. D. Laherty, J. Torchia, W.-M. Yang, G. Brard, S. D. Ngo, J. R. Davie, E. Seto, R. N. Eisenman, D. W. Rose, C. K. Glass, and M. G. Rosenfeld. 1997. A complex containing N-CoR, mSin3 and histone deacetylase mediates transcriptional repression. Nature 387:43-48.[CrossRef][Medline]

11. Hendrich, B., and, A. Bird. 1998. Identification and characterization of a family of mammalian methyl-CpG binding proteins. Mol. Cell. Biol. 18:6538-6547.[Abstract/Free Full Text]

12. Hendrich, B., U. Hardeland, H.-H. Ng, J. Jiricny, and A. Bird. 1999. The thymine glycosylase MBD4 can bind to the product of deamination at methylated CpG sites. Nature 401:301-304.[CrossRef][Medline]

13. Jaenisch, R., and A. Bird. 2003. Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals. Nat. Genet. 33:245-254.

14. Kondo, E., T. Furukawa, K. Yoshinaga, H. Kijima, S. Semba, T. Yatsuoka, T. Yokoyama, S. Fukushige, and A. Horii. 2000. Not hMSH2 but hMLH1 is frequently silenced by hypermethylation in endometrial cancer but rarely silenced in pancreatic cancer with microsatellite instability. Int. J. Oncol. 17:535-541.[Medline]

15. Kondo, E., A. Horii, and S. Fukushige. 2001. The interacting domains of three MutL heterodimers in man: hMLH1 interacts with 36 homologous amino acid residues within hMLH3, hPMS1 and hPMS2. Nucleic Acids Res. 29:1695-1702.[Abstract/Free Full Text]

16. Kondo, Y., L. Shen, P. S. Yan, T. H.-M. Huang, and J.-P. J. Issa. 2004. Chromatin immunoprecipitation microassays for identification of genes silenced by histone H3 lysine 9 methylation. Proc. Natl. Acad. Sci. USA 101:7398-7403.[Abstract/Free Full Text]

17. Likhite, V. A., E. I. Cass, S. D. Anderson, J. R. Yates, and A. M. Nardulli. 2004. Interaction of estrogen receptor {alpha} with 3-methyladenine DNA glycosylase modulates transcription and DNA repair. J. Biol. Chem. 279:16875-16882.[Abstract/Free Full Text]

18. Magdinier, F., and A. P. Wolffe. 2001. Selective association of the methyl-CpG binding protein MBD2 with the silent p14/p16 locus in human neoplasia. Proc. Natl. Acad. Sci. USA 98:4990-4995.[Abstract/Free Full Text]

19. Meehan, R. R., J. D. Lewis, S. McKay, E. L. Kleiner, and A. P. Bird. 1989. Identification of a mammalian protein that binds specifically to DNA containing methylated CpGs. Cell 58:499-507.[CrossRef][Medline]

20. Millar, C. B., J. Guy, O. J. Sansom, J. Selfridge, E. MacDougall, B. Hendrich, P. D. Keightley, S. M. Bishop, A. R. Clarke, and, A. Bird. 2002. Enhanced CpG mutability and tumorigenesis in MBD4-deficient mice. Science 297:403-405.[Abstract/Free Full Text]

21. Missero, C., M. T. Pirro, S. Simeone, M. Pischetola, and R. Di Lauro. 2001. The DNA glycosylase T:G mismatch-specific thymine DNA glycosylase represses thyroid transcription factor-1-activated transcription. J. Biol. Chem. 276:33569-33575.[Abstract/Free Full Text]

22. Nan, X., H.-H. Ng, C. A. Johnson, C. D. Laherty, B. M. Turner, R. N. Eisenman, and A. Bird. 1998. Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature 393:386-389.[CrossRef][Medline]

23. Ng, H.-H., Y. Zhang, B. Hendrich, C. A. Johnson, B. M. Turner, H. Erdjument-Bromage, P. Tempst, D. Reinberg, and, A. Bird. 1999. MBD2 is a transcriptional repressor belonging to the MeCP1 histone deacetylase complex. Nat. Genet. 23:58-61.[Medline]

24. Paz, M. F., M. F. Fraga, S. Avila, M. Guo, M. Pollan, J. G. Herman, and M. Esteller. 2003. A systematic profile of DNA methylation in human cancer cell lines. Cancer Res. 63:1114-1121.[Abstract/Free Full Text]

25. Petronzelli, F., A. Riccio, G. D. Markham, S. H. Seeholzer, J. Stoerker, M. Genuardi, A. T. Yeung, Y. Matsumoto, and A. Bellacosa. 2000. Biphasic kinetics of the human DNA repair protein MED1 (MBD4), a mismatch-specific DNA N-glycosylase. J. Biol. Chem. 275:32422-32429.[Abstract/Free Full Text]

26. Riccio, A., L. A. Aaltonen, A. K. Godwin, A. Loukola, A. Percesepe, R. Salovaara, V. Masciullo, M. Genuardi, M. Paravatou-Petsotas, D. E. Bassi, R. A. Ruggeri, A. J. P. Klein-Szanto, J. R. Testa, G. Neri, and A. Bellacosa. 1999. The DNA repair gene MBD4 (MED1) is mutated in human carcinomas with microsatellite instability. Nat. Genet. 23:266-268.[CrossRef][Medline]

27. Saito, M., and F. Ishikawa. 2002. The mCpG-binding domain of human MBD3 does not bind to mCpG but interacts with NuRD/Mi2 components HDAC1 and MTA2. J. Biol. Chem. 277:35434-35439.[Abstract/Free Full Text]

28. Sansom, O. J., J. Zabkiewicz, S. M. Bishop, j. Guy, A. Bird, and A. R. Clarke. 2003. MBD4 deficiency reduces the apoptotic response to DNA-damaging agents in the murine small intestine. Oncogene 22:7130-7136.[CrossRef][Medline]

29. Sansom, O. J., S. M. Bishop, A. Bird, and A. R. Clarke. 2004. MBD4 deficiency does not increase mutation or accelerate tumorigenesis in mice lacking MMR. Oncogene 23:5693-5696.[CrossRef][Medline]

30. Trojan, J., S. Zeuzem, A. Randolph, C. Hemmerle, A. Brieger, J. Raedle, G. Plotz, J. Jiricny, and G. Marra. 2002. Functional analysis of hMLH1 variants and HNPCC-related mutations using a human expression system. Gastroenterology. 122:211-219.[CrossRef][Medline]

31. Um, S., M. Harbers, A. Benecke, B. Pierrat, R. Losson, and P. Chambon. 1998. Retinoic acid receptors interact physically and functionally with the T:G mismatch-specific thymine-DNA glycosylase. J. Biol. Chem. 273:20728-20736.[Abstract/Free Full Text]

32. Veigl, M. L., L. Kasturi, J. Olechnowicz, A. Ma, J. D. Lutterbaugh, S. Periyasamy, G.-M. Li, J. Drummond, P. L. Modrich, W. D. Sedwick, and S. D. Markowitz. 1998. Biallelic inactivation of hMLH1 by epigenetic gene silencing, a novel mechanism causing human MSI cancers. Proc. Natl. Acad. Sci. USA 95:8698-8702.[Abstract/Free Full Text]

33. Watanabe, S., T. Ichimura, N. Fujita, S. Tsuruzoe, I. Ohki, M. Shirakawa, M. Kawasuji, and M. Nakao. 2003. Methylated DNA-binding domain 1 and methylpurine-DNA glycosylase link transcriptional repression and DNA repair in chromatin. Proc. Natl. Acad. Sci. USA 100:12859-12864.[Abstract/Free Full Text]

34. Wong, E., K. Yang, M. Kuraguchi, U. Werling, E. Avdievich, K. Fan, M. Fazzari, B. Jin, A. M. C. Brown, M. Lipkin, and W. Edelmann. 2002. Mbd4 inactivation increases C->T transition mutations and promotes gastrointestinal tumor formation. Proc. Natl. Acad. Sci. USA 99:14937-14942.[Abstract/Free Full Text]

35. Yamakawa, K., S. Mitchell, R. Hubert, X. N. Chen, S. Colbern, Y. K. Huo, C. Gadomski, U. J. Kim, and J. R. Korenberg. 1995. Isolation and characterization of a candidate gene for progressive myoclonus epilepsy on 21q22.3. Hum. Mol. Genet. 4:709-716.[Abstract/Free Full Text]

36. Yatsuoka, T., T. Furukawa, M. Sunamura, S. Matsuno, and A. Horii. 2004. TU12B1-TY, a novel gene in the region at 12q22-q23.1 frequently deleted in pancreatic cancer, shows reduced expression in pancreatic cancer cells. Oncol. Rep. 12:1263-1268.[Medline]

37. Zhu, B., Y. Zheng, H. Angliker, S. Schwarz, S. Thiry, M. Siegmann, and J.-P. Jost. 2000. 5-methylcytosine DNA glycosylase activity is also present in the human MBD4 (G/T mismatch glycosylase) and in a related avian sequence. Nucleic Acids Res. 28:4157-4165.[Abstract/Free Full Text]

Molecular and Cellular Biology, June 2005, p. 4388-4396, Vol. 25, No. 11
0270-7306/05/$08.00+0     doi:10.1128/MCB.25.11.4388-4396.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:


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 Kondo, E.
Right arrow Articles by Fukushige, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kondo, E.
Right arrow Articles by Fukushige, S.

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
J. Bacteriol. J. Virol. Eukaryot. Cell
Microbiol. Mol. Biol. Rev. Clin. Vaccine Immunol. All ASM Journals