- Email: [email protected]
The recruitment of activation induced cytidine deaminase to the immunoglobulin locus by a regulatory element
Molecular Immunology 47 (2010) 1860–1865
Contents lists available at ScienceDirect
Molecular Immunology journal homepage: www.elsevier.com/locate/molimm
Short communication
The recruitment of activation induced cytidine deaminase to the immunoglobulin locus by a regulatory element Yonghwan Kim, Ming Tian ∗ Section of Molecular Genetics and Microbiology, University of Texas at Austin, 1 University Station, A5000, Austin, TX 78712, USA
a r t i c l e
i n f o
Article history: Received 13 January 2010 Accepted 25 February 2010 Available online 23 March 2010 Keywords: Immunoglobulin Gene conversion Transcription AID Enhancers
a b s t r a c t Activation induced cytidine deaminase (AID) is critical for the diversification of immunoglobulin (Ig). AID is generally thought to function by deaminating cytidines into uridines in the target DNA within the Ig loci, and the subsequent processing of the uridines, through DNA repair or replication, could lead to three different forms of Ig diversification events: class switch recombination, somatic hypermutation and gene conversion. Although AID is important for effective immunity, its mutagenic activity needs to be restricted to the Ig loci in order to avoid rampant mutations in the genome. In our previous studies, we have identified an Ig regulatory element (Region A) that is important for AID mediated gene conversion in chicken B cells, but its mechanism of function was unclear. In this report, we provide evidence that the regulatory element plays a role in recruiting AID to the Ig locus. © 2010 Elsevier Ltd. All rights reserved.
1. Introduction Activation induced cytidine deaminase (AID) is essential for three different pathways of immunoglobulin (Ig) diversification: class switch recombination, somatic hypermutation and Ig gene conversion (Muramatsu et al., 2000; Revy et al., 2000; Arakawa et al., 2002; Harris et al., 2002). There are different opinions regarding the mechanism of AID function (Shivarov et al., 2009; Di Noia and Neuberger, 2007), but the majority of evidence is consistent with the model that AID acts by deaminating cytidines into uridines in the target DNA within the Ig loci. A major unresolved issue is how the cytidine deamination activity of AID is targeted to the Ig loci to avoid widespread mutations throughout the genome (Odegard and Schatz, 2006). The question becomes less clear-cut with the increasing realization that AID in fact does mutate non-Ig genes to various extents in B cells, suggesting that the targeting of AID to the Ig loci is not as absolute as previously thought (Liu et al., 2008). In spite of the ambiguity, the Ig loci sustain substantially higher levels of mutation than the other loci, and the question remains as to why AID preferentially acts on the Ig loci. The answer does not come from the intrinsic specificity of AID itself. Although AID preferentially deaminates the WRC hotspot (W = A, T; R = A, G) (Pham et al., 2003), such motif is widespread in the genome. Therefore, the specificity of AID may rely on cofactors. In this regard, AID has been shown to interact with replication
∗ Corresponding author. Tel.: +1 512 471 5752; fax: +1 512 471 7088. E-mail address: [email protected] (M. Tian). 0161-5890/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.molimm.2010.02.025
protein A (RPA) (Chaudhuri et al., 2004) and CTNNBL1 (Conticello et al., 2008). The association of AID with RPA is dependent on the phosphorylation of S38 of AID by protein kinase A (PKA) (Basu et al., 2005), and AID, RPA and PKA colocalize to the switch regions, the target of AID during class switch recombination (Vuong et al., 2009). CTNNBL1 is a factor associated with spliceosomes, and given the link of AID function with transcription and splicing, the protein may play a role in this connection (Conticello et al., 2008). In spite of these progresses, it remains unclear how these AID interacting proteins target AID to the Ig loci as none of these proteins have been reported to exhibit binding specificity toward either DNA or RNA. The issue could also be addressed from the perspective of the DNA target; the Ig loci must exhibit unique features that can be recognized by AID and its cofactors. The switch regions represent such unique elements in the Ig loci. They can adopt unusual DNA structures (Tian and Alt, 2000; Yu et al., 2003; Duquette et al., 2004), which contain single stranded DNA, the obligate substrate for AID; additionally, the switch regions contain high densities of the AGCT motif, a hotspot for AID deamination. The combination of these two features renders the switch regions as prime substrates for AID (Shinkura et al., 2003; Zarrin et al., 2004), and may explain at least in part the preferential deamination of the switch regions by AID during class switch recombination. On the other hand, the Ig variable regions do not exhibit unique sequence features, and non-Ig DNA became the target of AID in the context of the Ig loci (Yélamos et al., 1995). Therefore, AID needs to be recruited to the Ig variable regions by other elements in the Ig loci, and there have been some recent progresses in identifying such elements in the chicken B cell line DT40 (Yang et al., 2006; Kothapalli et al., 2008; Blagodatski et
Y. Kim, M. Tian / Molecular Immunology 47 (2010) 1860–1865
al., 2009; Kim and Tian, 2009). In our previous study, we found that a 2.4-kb region (Region A) downstream of the constant region of Ig is important for AID mediated gene conversion, and more importantly this region is sufficient to target AID function to ectopic loci (Kim and Tian, 2009). In this report, we present further analysis of this region, and show that Region A may function by recruiting AID to the Ig locus. 2. Materials and methods 2.1. Cell culture and transfection experiments The DT40 cells were cultured at 37 ◦ C under 5% CO2 in RPMI media supplemented with 10% fetal bovine serum, 1% chicken serum, 100 units/ml penicillin, 100 g/ml streptomycin and 28 M -mercaptoethanol. For transfections, 10 g of linearized plasmid was electroporated into 10 × 106 cells; the electroporation condition was: 600 V and 25 F. After electroporation, the cells were distributed into two 96-well tissue culture plates. Selection drugs were added a day later. Stable clones were picked after approximately 10 days of culture. For targeted integration events, the desired clones were identified by Southern analysis of the genomic DNAs isolated from the stable clones. For inducible expression clones of I-SceI, Dam, or Dam fusion proteins, the expression of the relevant cDNAs were detected with Northern analysis of the RNAs from the stable clones. 2.2. Gene conversion experiments The gene conversion substrates have been described in detail previously (Kim and Tian, 2009). Briefly, the gene conversion reporter was derived from the puromycin resistance gene, which was inactivated by the insertion of an oligonucleotide containing
1861
an I-SceI site (Fig. 1A). The gene conversion reporter was transcribed with three different promoters derived from the chicken Ig, chicken EF1␣ and chicken -actin genes. A fragment of the puromycin resistance gene was positioned 2-kb upstream of the promoter to serve as the repair template in gene conversion. To measure gene conversion activity at the Ig locus, the substrate was integrated into a position 4-kb downstream of the transcription initiation site of the rearranged allele of Ig (L1 through L7, Fig. 1B). To measure gene conversion outside of the Ig locus, the substrate was integrated into a position 3.3-kb upstream of the hsc70 gene (Hsc, Fig. 1B). I-SceI was expressed with the tetracycline inducible system. For this purpose, the reverse Tet repressor was expressed with the actin promoter in a stably integrated transgene. The I-SceI cDNA was under the control of tetracycline inducible promoter, and the construct was stably integrated into the genomes of appropriate cell lines. For gene conversion experiments, the cells were subcloned in media containing 0.5 g/ml doxcycline. Single colonies were picked into 24 well tissue culture plates, and the cells were cultured for a total of 12 days from the start of subcloning. Gene conversion assay was performed as described previously (Kim and Tian, 2009). Briefly, an aliquot of cells from the 24 well plates was diluted into media containing 0.5 g/ml puromycin, and the cells were distributed into a 96 well tissue culture plate. As a control for plating efficiency, another aliquot of the cells was cultured in media without puromycin. The cell numbers used in each experiment were indicated in the Supplemental Table 1. After 6–7 days of culture, the colony numbers were counted under microscope. Gene conversion frequency was calculated as the percentage of puromycin resistant colonies relative to the total colonies. AID mediated gene conversion experiments (Supplemental Table 2) were performed in a similar fashion except that the cells were subcloned without doxcycline.
Fig. 1. Deletion of Region A increases I-SceI mediated gene conversion. (A) The gene conversion substrate. Puro-mu represents the puro reporter containing an I-SceI cleavage site, which is marked with a filled square. Puro-fr represents the wild-type puro fragment that serves as the repair template. P indicates promoter. (B) The gene conversion substrates integrated at the Ig or the hsc70 locus. The rearranged allele of the Ig locus is shown at the top. The integration site for the gene conversion substrate is marked with an arrowhead. The substrates integrated at the Ig locus (L1–L7) are shown below the diagram of Ig. Deletion of Region A is indicated with a bracket. The promoters for the substrates are: L1, L2: Ig; L4, L5: EF1␣; L6, L7: -actin. The gene conversion substrate integrated at the hsc70 locus is shown at the bottom; the EF1␣ promoter is used in this substrate. (C) This dot plot presents the results of the I-SceI mediated gene conversion (GC) analysis. Each dot represents the result of one experiment. The median of each data set is indicated with a bar in the plot, and the value is written below the plot. The GC substrates and addition of doxcycline are indicated below the plot.
1862
Y. Kim, M. Tian / Molecular Immunology 47 (2010) 1860–1865
2.3. Dam methylation experiments Dam and Dam fusion proteins were expressed with tamoxifen inducible constructs; compared with the tetracycline inducible system, the tamoxifen inducible expression constructs exhibit lower basal expression of Dam and Dam fusion proteins. Dam expression constructs were stably integrated into the AID locus of the appropriate cell lines; this step inactivates one allele of the AID gene, and the other allele of the AID gene was inactivated by a separate construct. Details of the targeting constructs are omitted, but the information is available upon request. Dam expression was induced by the addition of tamoxifen to 25 nM. Genomic DNAs were harvested from the cells that have been treated with tamoxifen for 8 or 12 h, and methylation status was determined by restriction digest with MboI together with other appropriate enzymes. The digested genomic DNAs were analyzed with Southern blotting. 3. Results 3.1. Region A is dispensable for chromatin accessibility at the Ig locus In our previous study, we carried out a deletion analysis of the Ig locus to identify regulatory elements involved in targeting AID (Kim and Tian, 2009). We found that deletion of a 2.4-kb region (Region A) downstream of the Ig constant region severely impaired AID mediated gene conversion, but the underlying cause of the defect was unclear. One possibility is that the chromatin accessibility at the Ig locus is reduced by the deletion of Region A. This model seems inconsistent with the fact that deletion of Region A only has minor effects on the transcriptional levels at the Ig locus (Kim and Tian, 2009). On the other hand, transcriptional activity is influenced by many factors, and chromatin accessibility is just one of them. Thus, it remains possible that deletion of Region A may reduce chromatin accessibility more severely than transcriptional levels. DNase I sensitivity assay is commonly used to analyze chromatin accessibility. A drawback of this assay is that the nuclei need to be permealized with detergent, which could potentially change chromatin structures. For this reason, it is preferable to carry out the nuclease digestion in live cells, and we used the homing endonuclease I-SceI for this purpose. I-SceI cleaves an 18-bp recognition sequence that is rarely found in the genome (Rouet et al., 1994). Since our gene conversion substrate contains an I-SceI site (Fig. 1A), the cleavage efficiency at this site ought to reflect its accessibility toward I-SceI. In this experiment, we expressed I-SceI with a tetracycline inducible construct in DT40 cells that contain an integrated copy of the gene conversion substrate at either the intact Ig locus (L1, Fig. 1B) or the Ig locus without Region A (L2, Fig. 1B). However, we were unable to detect any cleavage product with Southern analysis (data not shown). One potential explanation is that the DNA break was rapidly repaired, and if this was the case, the repair product may be easier to detect and could serve as an indirect readout for I-SceI cleavage activity. Since the DNA break can be repaired through gene conversion (Fig. 1A), we may be able to use gene conversion frequency as a measure for I-SceI cleavage activity. In the I-SceI inducible cell lines, the endogenous AID gene has been inactivated (see Section 2); therefore gene conversion should be dependent on I-SceI. Consistent with the predication, there is minimal gene conversion activity in L1 without doxcycline, and the low gene conversion activity is presumably due to the leaky expression of I-SceI under uninduced conditions (L1, -dox, Fig. 1C). Induction of I-SceI expression led to a clear increase in gene conversion activity, and in the puromycin resistant colonies the I-SceI site has been converted to wild-type puro sequence (data not shown). Thus, gene conversion in this system appears to reflect the cleavage
activity of I-SceI, and we used this assay to analyze the accessibility of the gene conversion substrate to I-SceI. We compared the I-SceI mediated gene conversion activity in L1 and L2 cells, and found unexpectedly that deletion of Region A increased gene conversion activity by approximately 3-fold (compare L2 with L1, Fig. 1C and Supplemental Table 1). This result is in contrast to what we observed in AID mediated gene conversion, which is severely decreased by the deletion of Region A (Kim and Tian, 2009). This unusual observation prompted us to test the generality of this phenomenon. In our previous study, we generated gene conversion substrates that are under the control of different promoters: Ig (L1, L2), EF1␣ (L4, L5) and -actin (L6, L7) (Fig. 1B). These substrates behaved similarly in AID mediated gene conversion, and it would be of interest to determine whether deletion of Region A also increases the gene conversion activity in the substrates with EF1␣ and actin promoters. Such indeed turned out to be the case (compare L5 with L4; L7 with L6; Fig. 1C and Supplemental Table 1). To further substantiate this point, we analyzed a gene conversion substrate that was integrated in a position that is 3.3-kb upstream of the hsc70 gene (Hsc, Fig. 1B), and this substrate would represent another situation where there is no Region A. This gene conversion substrate is transcribed with the EF1␣ promoter, and as expected from our earlier study (Kim and Tian, 2009), the substrate exhibits undetectable AID mediated gene conversion activity (data not shown). By contrast, I-SceI induced efficient gene conversion in this substrate, and interestingly the gene conversion frequency is close to those in L2, L5, and L7 (without Region A), but is higher than L1, L4, L6 (with Region A) (Fig. 1C and Supplemental Table 1). Thus, in all the situations that we have examined, Region A appears to exert opposite effects on AID versus I-SceI mediated gene conversion. We currently do not have an explanation for this phenomenon, but the result is inconsistent with the predication of the model that deletion of Region A reduces the accessibility of the gene conversion substrate. A caveat of the I-SceI experiment is that it relies on an indirect readout. Therefore, we would like to address this issue with a more direct assay; moreover, if an independent assay gives the same result, we would be more confident about the conclusions. For this purpose, we chose the Dam methyltransferase from Escherichia coli as a second probe for chromatin accessibility. The enzyme methylates the adenine within the GATC motif, and there is no equivalent enzyme in the eukaryotic cells. It has been shown that expression of Dam in eukaryotic cells leads to the methylation of GATC motifs throughout the genome, and the methylation efficiencies generally correlate with chromatin accessibility (Singh and Klar, 1992). We used a tamoxifen inducible expression construct to express Dam in DT40 cells (Fig. 2A). In this construct, the gpt drug selection marker interposes between the promoter and the Dam cDNA, and precludes the expression of Dam under this situation. The gpt cDNA is flanked with two loxP sites, and the DT40 cells used in our study contain a stably integrated transgene for tamoxifen inducible cre recombinase (Arakawa et al., 2001). Treatment of the cells with tamoxifen activates the cre recombinase, which excises out the gpt cDNA and leads to the expression of Dam (Fig. 2A). To compare the methylation pattern of the gene conversion substrate in L1 and L2 cells, we induced Dam expression with tamoxifen in these cells, isolated genomic DNA 8 or 12 h later, and digested the genomic DNA with the MboI restriction enzyme, which cleaves only unmethylated GATC motif. We focused our analysis on two MboI sites within the puro reporter, and found no obvious difference in the kinetics of methylation between L1 and L2 (Fig. 2B). As a control, we measured the methylation levels at the EF1␣ locus (Fig. 2C), which should not be affected by Region A, and would therefore reflect the Dam activities in the two cell types. We calculated the methylation ratio of Ig/EF1␣, and found only slight reductions in the methylation kinetics in L2 relative to L1 (Fig. 2D). To make sure that
Y. Kim, M. Tian / Molecular Immunology 47 (2010) 1860–1865
1863
Fig. 2. Deletion of Region A does not affect Dam mediated methylation of the gene conversion substrate. (A) The tamoxifen inducible expression construct for Dam. The addition of tamoxifen activates the cre recombinase expressed from a transgene in the DT40 cells. (B) The methylation analysis of the puro reporter integrated at the Ig locus. The autoradiograph is a Southern analysis of the DNA isolated from L1 and L2 cells that have been treated with tamoxifen for 0, 8, 12 h, as indicated at the top of the autoradiograph. The DNA was digested with MboI, EcoRI and XbaI, except for lane C, where MboI was omitted. The sizes of the DNA marker (Mr) are indicated to the left of autoradiograph. The DNA fragments are marked with a, b, and c on the right of the autoradiograph, and detailed descriptions of these fragments are shown in a diagram below. In the diagram, the puro represents the puro reporter integrated at the Ig locus. The location of the probe is indicated. The restriction sites are: E, EcoRI; M, MboI; X, XbaI. (C) The methylation analysis of the EF1␣ locus. This figure is illustrated the same way as (B). (D) This plot is a quantification of the relative methylation levels of L1 and L2, as determined by the analysis shown in (B) and (C). The value for each time point was the average of the results from three experiments, and the error bars represent standard deviation. The y-axis represents the relative methylation levels. This value was calculated by band-a (Ig)/band-a (EF1␣); the signals for band-a was derived by quantifying the blots as shown in (B) and (C) with PhosphoImager. The value for L1 cells at 12 h tamoxifen treatment was set as 1. Only band-a was used because it represents complete methylation of the DNA fragment. (E) The methylation analysis of the -globin locus. This figure is illustrated the same way as (B).
Dam methylation levels do correlate with chromatin accessibility in our system, we analyzed the methylation status of the -globin locus, which is not expressed in DT40 cells, and should represent a less accessible locus relative to the Ig and EF1␣ loci, which are actively transcribed in DT40 cells. Consistent with the expectation, the methylation kinetics of the -globin locus is obviously slower compared with those in both the Ig and the EF1␣ loci (compare Fig. 2E with B and C). This result shows that the Dam methylation assay is capable of revealing differences in chromatin accessibility in our system; if there were major reductions in chromatin accessibility in L2, we should have detected it with this assay. In this regard, the results from both the Dam methylation and I-SceI experiments are in agreement. Thus, we conclude that Region A is dispensable for chromatin accessibility at the Ig locus, with the assumption that AID, I-SceI, and Dam access chromatin similarly. 3.2. Region A facilitates the recruitment of AID to the Ig locus The second model for Region A function is that the element may recruit AID to the Ig locus. We used the DamID assay to test this model. In this method, the protein of interest is fused to Dam; if the protein binds specifically to a particular site in the genome, that locus would be preferentially methylated by the fusion protein rel-
ative to other loci (van Steensel and Henikoff, 2000). To apply this method to our study, we fused AID with Dam and expressed the fusion protein in DT40 cells with the tamoxifen inducible system. The AID-Dam fusion is functional in gene conversion experiments, and its activity is dependent on Region A (Fig. 3A and Supplemental Table 2). We compared AID-Dam mediated methylation of the gene conversion substrate in L1 and L2 cells, and found that the methylation level in L2 cells was lower than that in L1 cells (Fig. 3B). As a control, we analyzed the methylation pattern at the EF1␣ locus, and found no difference between L1 and L2 cells (Fig. 3B). We calculated the methylation ratio of Ig/EF1␣, and found an approximately 50% reduction in the methylation levels of gene conversion substrate in L2 cells (Fig. 3C); although the reduction is modest, it is reproducible in three experiments. This result suggests that the deletion of Region A impairs the interaction of AID-Dam fusion with the gene conversion substrate at the Ig locus. To test whether the recruitment effect of Region A is specific to AID, we generated a Dam fusion with Apobec-1, which is the closest homologue of AID within the cytidine deaminase family (Wedekind et al., 2003). In spite of the homology, Apobec-1 is unable to substitute for AID in Ig diversification, potentially due to the lack of recruitment to the Ig loci (Fugmann et al., 2004). We expressed the Apobec-1-Dam fusion in L1 and L2 cells, and as shown in Fig. 3D, Apobec1-Dam
1864
Y. Kim, M. Tian / Molecular Immunology 47 (2010) 1860–1865
Fig. 3. Deletion of Region A reduces the interaction of AID-Dam with the gene conversion substrate at the Ig locus. (A) This dot plot shows the gene conversion analysis of AID−/− cells complemented with AID-Dam. This figure is illustrated the same way as Fig. 1C. (B) The methylation analysis with AID-Dam. This figure is illustrated the same way as Fig. 2B. (C) This plot is a quantification of the methylation analysis as shown in (B). The value for each time point was the average of the results from three experiments, and the error bars represent standard deviation. The y-axis represents the relative methylation levels. This value was calculated by band-a (Ig)/band-b (EF1␣); the signals for band-a or b was derived by quantifying the blots as shown in (B) with PhosphoImager. The reason to use band-b for EF1␣ is because band-a is not detectable. Since the EF1␣ signal only serves as normalization, the use of band-b should not affect the outcome of the analysis. The value for L1 cells at 12 h tamoxifen treatment was set as 1. (D) This figure shows the methylation analysis with Apobec1-Dam. This figure is illustrated the same way as Fig. 2B. (E) This histogram compares the methylation of the gene conversion substrate (L1) mediated by AID-Dam and Apobec1-Dam after 12 h of tamoxifen treatment as shown in (B) and (C). Relative methylation levels are calculated the same way as in (C). The data were derived from three experiments; the error bar represents standard deviation. The average methylation levels mediated by AID-Dam or Apobec1-Dam are indicated below the plot, and the value for AID-Dam is set as 1.
methylates L1 and L2 to essentially the same extents. Moreover, relative to AID-Dam, Apobec1-Dam methylates the gene conversion substrate poorly; after normalization against the methylation levels at the EF1␣ locus, AID-Dam methylates the gene conversion substrate 11-fold more efficiently than Apobec1-Dam (Fig. 3E). In summary, these results suggest that AID-Dam is preferentially recruited to the Ig locus, and this recruitment involves Region A. 4. Discussion In our previous study, we identified a regulatory region (Region A) that is important for AID mediated gene conversion. Our
present work is aimed at understanding how Region A facilitates AID function. We tested two models: first, Region A may render local DNA more accessible to AID; second, Region A may recruit AID to adjacent DNA. Our results are consistent with the second hypothesis. It remains to be determined how Region A recruits AID to the Ig locus. The region contains binding sites for several transcription factors, including E2A, NFB, IRF4, PU.1, Mef2, and Octamer binding proteins; some of these factors have been implicated in AID function (Schoetz et al., 2006; Kim and Tian, 2009). It is possible that these factors may form an enhanceosome that recruits AID to the Ig locus.
Y. Kim, M. Tian / Molecular Immunology 47 (2010) 1860–1865
Deletion of Region A reduces, but does not abolish the targeting of AID to the Ig locus; as shown in Fig. 3, even in L2 cells (without Region A), AID-Dam still methylates the gene conversion substrate more efficiently than Apobec1-Dam. This observation suggests that, besides Region A, there may be other AID targeting elements at the Ig locus. In this regard, we showed previously that deletion of a 7.4-kb region (Region B) downstream of Region A also reduces AID mediated gene conversion, albeit to a lesser extent relative to the deletion of Region A. Both Region A and Region B overlaps with regulatory elements identified in other studies, including the 4.1-kb 3 regulatory region (Kothapalli et al., 2008) and the 9.8-kb diversification activator (DIVAC) (Blagodatski et al., 2009). The overall picture is that multiple elements contribute to the targeting of AID function, which may explain the modest reduction of AID recruitment to the Ig locus in the absence of Region A. Deletion of Region A appears to have a more prominent impact on AID mediated gene conversion (a reduction of 15-fold, Kim and Tian, 2009) than on AID recruitment (2-fold). One potential explanation is that gene conversion frequency does not correlate linearly with the levels of AID recruitment, but a more interesting possibility is that Region A contributes to AID function in additional ways. The speculation is prompted by the intriguing observation that Region A appears to interfere with I-SceI mediated gene conversion. Since the I-SceI experiments were done in AID−/− cells, the phenomenon has nothing to do AID recruitment, and might reflect the impact of Region A on the gene conversion substrate that would favor AID mediated deamination, but interferes with I-SceI mediated cleavage of the DNA target. The resolution of this issue in the future may reveal novel regulatory mechanisms of AID function by Region A as well as other cis-acting elements in the Ig locus. Acknowledgements We thank Dr. Arakawa and Dr. Buerstedde for providing with us the DT40cre1 cell line. This work was supported with funding from the University of Texas at Austin. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.molimm.2010.02.025. References Arakawa, H., Lodygin, D., Buerstedde, J.M., 2001. Mutant loxP vectors for selectable marker recycle and conditional knock-outs. BMC Biotechnol. 1, 7. Arakawa, H., Hauschild, J., Buerstedde, J.M., 2002. Requirement of the activationinduced deaminase (AID) gene for immunoglobulin gene conversion. Science 295, 1301–1306. Basu, U., Chaudhuri, J., Alpert, C., Dutt, S., Ranganath, S., Li, G., Schrum, J.P., Manis, J.P., Alt, F.W., 2005. The AID antibody diversification enzyme is regulated by protein kinase A phosphorylation. Nature 438, 508–511. Blagodatski, A., Batrak, V., Schmidl, S., Schoetz, U., Caldwell, R.B., Arakawa, H., Buerstedde, J.M., 2009. A cis-acting diversification activator both necessary and sufficient for AID-mediated hypermutation. PLoS Genet. 5, e1000332. Chaudhuri, J., Khuong, C., Alt, F.W., 2004. Replication protein A interacts with AID to promote deamination of somatic hypermutation targets. Nature 430, 992–998.
1865
Conticello, S.G., Ganesh, K., Xue, K., Lu, M., Rada, C., Neuberger, M.S., 2008. Interaction between antibody-diversification enzyme AID and spliceosome-associated factor CTNNBL1. Mol. Cell 31, 474–484. Di Noia, J.M., Neuberger, M.S., 2007. Molecular mechanisms of antibody somatic hypermutation. Annu. Rev. Biochem. 76, 1–22. Duquette, M.L., Handa, P., Vincent, J.A., Taylor, A.F., Maizels, N., 2004. Intracellular transcription of G-rich DNAs induces formation of G-loops, novel structures containing G4 DNA. Genes Dev. 18, 1618–1629. Fugmann, S.D., Rush, J.S., Schatz, D.G., 2004. Non-redundancy of cytidine deaminases in class switch recombination. Eur. J. Immunol. 34, 844–849. Harris, R.S., Sale, J.E., Petersen-Mahrt, S.K., Neuberger, M.S., 2002. AID is essential for immunoglobulin V gene conversion in a cultured B cell line. Curr. Biol. 12, 435–438. Kim, Y., Tian, M., 2009. NF-B family of transcription factor facilitates gene conversion in chicken B cells. Mol. Immunol. 46, 3283–3291. Kothapalli, N., Norton, D.D., Fugmann, S.D., 2008. Cutting edge: a cis-acting DNA element targets AID-mediated sequence diversification to the chicken Ig light chain gene locus. J. Immunol. 180, 2019–2023. Liu, M., Duke, J.L., Richter, D.J., Vinuesa, C.G., Goodnow, C.C., Kleinstein, S.H., Schatz, D.G., 2008. Two levels of protection for the B cell genome during somatic hypermutation. Nature 451, 841–845. Muramatsu, M., Kinoshita, K., Fagarasan, S., Yamada, S., Shinkai, Y., Honjo, T., 2000. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102, 553–563. Odegard, V.H., Schatz, D.G., 2006. Targeting of somatic hypermutation. Nat. Rev. Immunol. 6, 573–583. Pham, P., Bransteitter, R., Petruska, J., Goodman, M.F., 2003. Processive AID-catalysed cytosine deamination on single-stranded DNA simulates somatic hypermutation. Nature 424, 103–107. Revy, P., Muto, T., Levy, Y., Geissmann, F., Plebani, A., Sanal, O., Catalan, N., Forveille, M., Dufourcq-Labelouse, R., Gennery, A., Tezcan, I., Ersoy, F., Kayserili, H., Ugazio, A.G., Brousse, N., Muramatsu, M., Notarangelo, L.D., Kinoshita, K., Honjo, T., Fischer, A., Durandy, A., 2000. Activation-induced cytidine deaminase (AID) deficiency causes the autosomal recessive form of the Hyper-IgM syndrome (HIGM2). Cell 102, 565–575. Rouet, P., Smih, F., Jasin, M., 1994. Introduction of double-strand breaks into the genome of mouse cells by expression of a rare-cutting endonuclease. Mol. Cell Biol. 14, 8096–8106. Schoetz, U., Cervelli, M., Wang, Y.D., Fiedler, P., Buerstedde, J.M., 2006. E2A expression stimulates Ig hypermutation. J. Immunol. 177, 395–400. Shinkura, R., Tian, M., Smith, M., Chua, K., Fujiwara, Y., Alt, F.W., 2003. The influence of transcriptional orientation on endogenous switch region function. Nat. Immunol., 435–441. Shivarov, V., Shinkura, R., Doi, T., Begum, N.A., Nagaoka, H., Okazaki, I.M., Ito, S., Nonaka, T., Kinoshita, K., Honjo, T., 2009. Molecular mechanism for generation of antibody memory. Philos. Trans. R. Soc. Lond. B Biol. Sci. 364, 569–575. Singh, J., Klar, A.J., 1992. Active genes in budding yeast display enhanced in vivo accessibility to foreign DNA methylases: a novel in vivo probe for chromatin structure of yeast. Genes Dev. 6, 186–196. Tian, M., Alt, F.W., 2000. Transcription-induced cleavage of immunoglobulin switch regions by nucleotide excision repair nucleases in vitro. J. Biol. Chem. 275, 24163–24172. van Steensel, B., Henikoff, S., 2000. Identification of in vivo DNA targets of chromatin proteins using tethered dam methyltransferase. Nat. Biotechnol. 18, 424–428. Vuong, B.Q., Lee, M., Kabir, S., Irimia, C., Macchiarulo, S., McKnight, G.S., Chaudhuri, J., 2009. Specific recruitment of protein kinase A to the immunoglobulin locus regulates class-switch recombination. Nat. Immunol. 10, 420–426. Wedekind, J.E., Dance, G.S., Sowden, M.P., Smith, H.C., 2003. Messenger RNA editing in mammals: new members of the APOBEC family seeking roles in the family business. Trends Genet. 19, 207–216. Yang, S.Y., Fugmann, S.D., Schatz, D.G., 2006. Control of gene conversion and somatic hypermutation by immunoglobulin promoter and enhancer sequences. J. Exp. Med. 203, 2919–2928. Yélamos, J., Klix, N., Goyenechea, B., Lozano, F., Chui, Y.L., González Fernández, A., Pannell, R., Neuberger, M.S., Milstein, C., 1995. Targeting of non-Ig sequences in place of the V segment by somatic hypermutation. Nature 376, 225–229. Yu, K., Chedin, F., Hsieh, C.L., Wilson, T.E., Lieber, M.R., 2003. R-loops at immunoglobulin class switch regions in the chromosomes of stimulated B cells. Nat. Immunol. 4, 442–451. Zarrin, A.A., Alt, F.W., Chaudhuri, J., Stokes, N., Kaushal, D., Du Pasquier, L., Tian, M., 2004. An evolutionarily conserved target motif for immunoglobulin class-switch recombination. Nat. Immunol. 5, 1275–1281.
Contents lists available at ScienceDirect
Molecular Immunology journal homepage: www.elsevier.com/locate/molimm
Short communication
The recruitment of activation induced cytidine deaminase to the immunoglobulin locus by a regulatory element Yonghwan Kim, Ming Tian ∗ Section of Molecular Genetics and Microbiology, University of Texas at Austin, 1 University Station, A5000, Austin, TX 78712, USA
a r t i c l e
i n f o
Article history: Received 13 January 2010 Accepted 25 February 2010 Available online 23 March 2010 Keywords: Immunoglobulin Gene conversion Transcription AID Enhancers
a b s t r a c t Activation induced cytidine deaminase (AID) is critical for the diversification of immunoglobulin (Ig). AID is generally thought to function by deaminating cytidines into uridines in the target DNA within the Ig loci, and the subsequent processing of the uridines, through DNA repair or replication, could lead to three different forms of Ig diversification events: class switch recombination, somatic hypermutation and gene conversion. Although AID is important for effective immunity, its mutagenic activity needs to be restricted to the Ig loci in order to avoid rampant mutations in the genome. In our previous studies, we have identified an Ig regulatory element (Region A) that is important for AID mediated gene conversion in chicken B cells, but its mechanism of function was unclear. In this report, we provide evidence that the regulatory element plays a role in recruiting AID to the Ig locus. © 2010 Elsevier Ltd. All rights reserved.
1. Introduction Activation induced cytidine deaminase (AID) is essential for three different pathways of immunoglobulin (Ig) diversification: class switch recombination, somatic hypermutation and Ig gene conversion (Muramatsu et al., 2000; Revy et al., 2000; Arakawa et al., 2002; Harris et al., 2002). There are different opinions regarding the mechanism of AID function (Shivarov et al., 2009; Di Noia and Neuberger, 2007), but the majority of evidence is consistent with the model that AID acts by deaminating cytidines into uridines in the target DNA within the Ig loci. A major unresolved issue is how the cytidine deamination activity of AID is targeted to the Ig loci to avoid widespread mutations throughout the genome (Odegard and Schatz, 2006). The question becomes less clear-cut with the increasing realization that AID in fact does mutate non-Ig genes to various extents in B cells, suggesting that the targeting of AID to the Ig loci is not as absolute as previously thought (Liu et al., 2008). In spite of the ambiguity, the Ig loci sustain substantially higher levels of mutation than the other loci, and the question remains as to why AID preferentially acts on the Ig loci. The answer does not come from the intrinsic specificity of AID itself. Although AID preferentially deaminates the WRC hotspot (W = A, T; R = A, G) (Pham et al., 2003), such motif is widespread in the genome. Therefore, the specificity of AID may rely on cofactors. In this regard, AID has been shown to interact with replication
∗ Corresponding author. Tel.: +1 512 471 5752; fax: +1 512 471 7088. E-mail address: [email protected] (M. Tian). 0161-5890/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.molimm.2010.02.025
protein A (RPA) (Chaudhuri et al., 2004) and CTNNBL1 (Conticello et al., 2008). The association of AID with RPA is dependent on the phosphorylation of S38 of AID by protein kinase A (PKA) (Basu et al., 2005), and AID, RPA and PKA colocalize to the switch regions, the target of AID during class switch recombination (Vuong et al., 2009). CTNNBL1 is a factor associated with spliceosomes, and given the link of AID function with transcription and splicing, the protein may play a role in this connection (Conticello et al., 2008). In spite of these progresses, it remains unclear how these AID interacting proteins target AID to the Ig loci as none of these proteins have been reported to exhibit binding specificity toward either DNA or RNA. The issue could also be addressed from the perspective of the DNA target; the Ig loci must exhibit unique features that can be recognized by AID and its cofactors. The switch regions represent such unique elements in the Ig loci. They can adopt unusual DNA structures (Tian and Alt, 2000; Yu et al., 2003; Duquette et al., 2004), which contain single stranded DNA, the obligate substrate for AID; additionally, the switch regions contain high densities of the AGCT motif, a hotspot for AID deamination. The combination of these two features renders the switch regions as prime substrates for AID (Shinkura et al., 2003; Zarrin et al., 2004), and may explain at least in part the preferential deamination of the switch regions by AID during class switch recombination. On the other hand, the Ig variable regions do not exhibit unique sequence features, and non-Ig DNA became the target of AID in the context of the Ig loci (Yélamos et al., 1995). Therefore, AID needs to be recruited to the Ig variable regions by other elements in the Ig loci, and there have been some recent progresses in identifying such elements in the chicken B cell line DT40 (Yang et al., 2006; Kothapalli et al., 2008; Blagodatski et
Y. Kim, M. Tian / Molecular Immunology 47 (2010) 1860–1865
al., 2009; Kim and Tian, 2009). In our previous study, we found that a 2.4-kb region (Region A) downstream of the constant region of Ig is important for AID mediated gene conversion, and more importantly this region is sufficient to target AID function to ectopic loci (Kim and Tian, 2009). In this report, we present further analysis of this region, and show that Region A may function by recruiting AID to the Ig locus. 2. Materials and methods 2.1. Cell culture and transfection experiments The DT40 cells were cultured at 37 ◦ C under 5% CO2 in RPMI media supplemented with 10% fetal bovine serum, 1% chicken serum, 100 units/ml penicillin, 100 g/ml streptomycin and 28 M -mercaptoethanol. For transfections, 10 g of linearized plasmid was electroporated into 10 × 106 cells; the electroporation condition was: 600 V and 25 F. After electroporation, the cells were distributed into two 96-well tissue culture plates. Selection drugs were added a day later. Stable clones were picked after approximately 10 days of culture. For targeted integration events, the desired clones were identified by Southern analysis of the genomic DNAs isolated from the stable clones. For inducible expression clones of I-SceI, Dam, or Dam fusion proteins, the expression of the relevant cDNAs were detected with Northern analysis of the RNAs from the stable clones. 2.2. Gene conversion experiments The gene conversion substrates have been described in detail previously (Kim and Tian, 2009). Briefly, the gene conversion reporter was derived from the puromycin resistance gene, which was inactivated by the insertion of an oligonucleotide containing
1861
an I-SceI site (Fig. 1A). The gene conversion reporter was transcribed with three different promoters derived from the chicken Ig, chicken EF1␣ and chicken -actin genes. A fragment of the puromycin resistance gene was positioned 2-kb upstream of the promoter to serve as the repair template in gene conversion. To measure gene conversion activity at the Ig locus, the substrate was integrated into a position 4-kb downstream of the transcription initiation site of the rearranged allele of Ig (L1 through L7, Fig. 1B). To measure gene conversion outside of the Ig locus, the substrate was integrated into a position 3.3-kb upstream of the hsc70 gene (Hsc, Fig. 1B). I-SceI was expressed with the tetracycline inducible system. For this purpose, the reverse Tet repressor was expressed with the actin promoter in a stably integrated transgene. The I-SceI cDNA was under the control of tetracycline inducible promoter, and the construct was stably integrated into the genomes of appropriate cell lines. For gene conversion experiments, the cells were subcloned in media containing 0.5 g/ml doxcycline. Single colonies were picked into 24 well tissue culture plates, and the cells were cultured for a total of 12 days from the start of subcloning. Gene conversion assay was performed as described previously (Kim and Tian, 2009). Briefly, an aliquot of cells from the 24 well plates was diluted into media containing 0.5 g/ml puromycin, and the cells were distributed into a 96 well tissue culture plate. As a control for plating efficiency, another aliquot of the cells was cultured in media without puromycin. The cell numbers used in each experiment were indicated in the Supplemental Table 1. After 6–7 days of culture, the colony numbers were counted under microscope. Gene conversion frequency was calculated as the percentage of puromycin resistant colonies relative to the total colonies. AID mediated gene conversion experiments (Supplemental Table 2) were performed in a similar fashion except that the cells were subcloned without doxcycline.
Fig. 1. Deletion of Region A increases I-SceI mediated gene conversion. (A) The gene conversion substrate. Puro-mu represents the puro reporter containing an I-SceI cleavage site, which is marked with a filled square. Puro-fr represents the wild-type puro fragment that serves as the repair template. P indicates promoter. (B) The gene conversion substrates integrated at the Ig or the hsc70 locus. The rearranged allele of the Ig locus is shown at the top. The integration site for the gene conversion substrate is marked with an arrowhead. The substrates integrated at the Ig locus (L1–L7) are shown below the diagram of Ig. Deletion of Region A is indicated with a bracket. The promoters for the substrates are: L1, L2: Ig; L4, L5: EF1␣; L6, L7: -actin. The gene conversion substrate integrated at the hsc70 locus is shown at the bottom; the EF1␣ promoter is used in this substrate. (C) This dot plot presents the results of the I-SceI mediated gene conversion (GC) analysis. Each dot represents the result of one experiment. The median of each data set is indicated with a bar in the plot, and the value is written below the plot. The GC substrates and addition of doxcycline are indicated below the plot.
1862
Y. Kim, M. Tian / Molecular Immunology 47 (2010) 1860–1865
2.3. Dam methylation experiments Dam and Dam fusion proteins were expressed with tamoxifen inducible constructs; compared with the tetracycline inducible system, the tamoxifen inducible expression constructs exhibit lower basal expression of Dam and Dam fusion proteins. Dam expression constructs were stably integrated into the AID locus of the appropriate cell lines; this step inactivates one allele of the AID gene, and the other allele of the AID gene was inactivated by a separate construct. Details of the targeting constructs are omitted, but the information is available upon request. Dam expression was induced by the addition of tamoxifen to 25 nM. Genomic DNAs were harvested from the cells that have been treated with tamoxifen for 8 or 12 h, and methylation status was determined by restriction digest with MboI together with other appropriate enzymes. The digested genomic DNAs were analyzed with Southern blotting. 3. Results 3.1. Region A is dispensable for chromatin accessibility at the Ig locus In our previous study, we carried out a deletion analysis of the Ig locus to identify regulatory elements involved in targeting AID (Kim and Tian, 2009). We found that deletion of a 2.4-kb region (Region A) downstream of the Ig constant region severely impaired AID mediated gene conversion, but the underlying cause of the defect was unclear. One possibility is that the chromatin accessibility at the Ig locus is reduced by the deletion of Region A. This model seems inconsistent with the fact that deletion of Region A only has minor effects on the transcriptional levels at the Ig locus (Kim and Tian, 2009). On the other hand, transcriptional activity is influenced by many factors, and chromatin accessibility is just one of them. Thus, it remains possible that deletion of Region A may reduce chromatin accessibility more severely than transcriptional levels. DNase I sensitivity assay is commonly used to analyze chromatin accessibility. A drawback of this assay is that the nuclei need to be permealized with detergent, which could potentially change chromatin structures. For this reason, it is preferable to carry out the nuclease digestion in live cells, and we used the homing endonuclease I-SceI for this purpose. I-SceI cleaves an 18-bp recognition sequence that is rarely found in the genome (Rouet et al., 1994). Since our gene conversion substrate contains an I-SceI site (Fig. 1A), the cleavage efficiency at this site ought to reflect its accessibility toward I-SceI. In this experiment, we expressed I-SceI with a tetracycline inducible construct in DT40 cells that contain an integrated copy of the gene conversion substrate at either the intact Ig locus (L1, Fig. 1B) or the Ig locus without Region A (L2, Fig. 1B). However, we were unable to detect any cleavage product with Southern analysis (data not shown). One potential explanation is that the DNA break was rapidly repaired, and if this was the case, the repair product may be easier to detect and could serve as an indirect readout for I-SceI cleavage activity. Since the DNA break can be repaired through gene conversion (Fig. 1A), we may be able to use gene conversion frequency as a measure for I-SceI cleavage activity. In the I-SceI inducible cell lines, the endogenous AID gene has been inactivated (see Section 2); therefore gene conversion should be dependent on I-SceI. Consistent with the predication, there is minimal gene conversion activity in L1 without doxcycline, and the low gene conversion activity is presumably due to the leaky expression of I-SceI under uninduced conditions (L1, -dox, Fig. 1C). Induction of I-SceI expression led to a clear increase in gene conversion activity, and in the puromycin resistant colonies the I-SceI site has been converted to wild-type puro sequence (data not shown). Thus, gene conversion in this system appears to reflect the cleavage
activity of I-SceI, and we used this assay to analyze the accessibility of the gene conversion substrate to I-SceI. We compared the I-SceI mediated gene conversion activity in L1 and L2 cells, and found unexpectedly that deletion of Region A increased gene conversion activity by approximately 3-fold (compare L2 with L1, Fig. 1C and Supplemental Table 1). This result is in contrast to what we observed in AID mediated gene conversion, which is severely decreased by the deletion of Region A (Kim and Tian, 2009). This unusual observation prompted us to test the generality of this phenomenon. In our previous study, we generated gene conversion substrates that are under the control of different promoters: Ig (L1, L2), EF1␣ (L4, L5) and -actin (L6, L7) (Fig. 1B). These substrates behaved similarly in AID mediated gene conversion, and it would be of interest to determine whether deletion of Region A also increases the gene conversion activity in the substrates with EF1␣ and actin promoters. Such indeed turned out to be the case (compare L5 with L4; L7 with L6; Fig. 1C and Supplemental Table 1). To further substantiate this point, we analyzed a gene conversion substrate that was integrated in a position that is 3.3-kb upstream of the hsc70 gene (Hsc, Fig. 1B), and this substrate would represent another situation where there is no Region A. This gene conversion substrate is transcribed with the EF1␣ promoter, and as expected from our earlier study (Kim and Tian, 2009), the substrate exhibits undetectable AID mediated gene conversion activity (data not shown). By contrast, I-SceI induced efficient gene conversion in this substrate, and interestingly the gene conversion frequency is close to those in L2, L5, and L7 (without Region A), but is higher than L1, L4, L6 (with Region A) (Fig. 1C and Supplemental Table 1). Thus, in all the situations that we have examined, Region A appears to exert opposite effects on AID versus I-SceI mediated gene conversion. We currently do not have an explanation for this phenomenon, but the result is inconsistent with the predication of the model that deletion of Region A reduces the accessibility of the gene conversion substrate. A caveat of the I-SceI experiment is that it relies on an indirect readout. Therefore, we would like to address this issue with a more direct assay; moreover, if an independent assay gives the same result, we would be more confident about the conclusions. For this purpose, we chose the Dam methyltransferase from Escherichia coli as a second probe for chromatin accessibility. The enzyme methylates the adenine within the GATC motif, and there is no equivalent enzyme in the eukaryotic cells. It has been shown that expression of Dam in eukaryotic cells leads to the methylation of GATC motifs throughout the genome, and the methylation efficiencies generally correlate with chromatin accessibility (Singh and Klar, 1992). We used a tamoxifen inducible expression construct to express Dam in DT40 cells (Fig. 2A). In this construct, the gpt drug selection marker interposes between the promoter and the Dam cDNA, and precludes the expression of Dam under this situation. The gpt cDNA is flanked with two loxP sites, and the DT40 cells used in our study contain a stably integrated transgene for tamoxifen inducible cre recombinase (Arakawa et al., 2001). Treatment of the cells with tamoxifen activates the cre recombinase, which excises out the gpt cDNA and leads to the expression of Dam (Fig. 2A). To compare the methylation pattern of the gene conversion substrate in L1 and L2 cells, we induced Dam expression with tamoxifen in these cells, isolated genomic DNA 8 or 12 h later, and digested the genomic DNA with the MboI restriction enzyme, which cleaves only unmethylated GATC motif. We focused our analysis on two MboI sites within the puro reporter, and found no obvious difference in the kinetics of methylation between L1 and L2 (Fig. 2B). As a control, we measured the methylation levels at the EF1␣ locus (Fig. 2C), which should not be affected by Region A, and would therefore reflect the Dam activities in the two cell types. We calculated the methylation ratio of Ig/EF1␣, and found only slight reductions in the methylation kinetics in L2 relative to L1 (Fig. 2D). To make sure that
Y. Kim, M. Tian / Molecular Immunology 47 (2010) 1860–1865
1863
Fig. 2. Deletion of Region A does not affect Dam mediated methylation of the gene conversion substrate. (A) The tamoxifen inducible expression construct for Dam. The addition of tamoxifen activates the cre recombinase expressed from a transgene in the DT40 cells. (B) The methylation analysis of the puro reporter integrated at the Ig locus. The autoradiograph is a Southern analysis of the DNA isolated from L1 and L2 cells that have been treated with tamoxifen for 0, 8, 12 h, as indicated at the top of the autoradiograph. The DNA was digested with MboI, EcoRI and XbaI, except for lane C, where MboI was omitted. The sizes of the DNA marker (Mr) are indicated to the left of autoradiograph. The DNA fragments are marked with a, b, and c on the right of the autoradiograph, and detailed descriptions of these fragments are shown in a diagram below. In the diagram, the puro represents the puro reporter integrated at the Ig locus. The location of the probe is indicated. The restriction sites are: E, EcoRI; M, MboI; X, XbaI. (C) The methylation analysis of the EF1␣ locus. This figure is illustrated the same way as (B). (D) This plot is a quantification of the relative methylation levels of L1 and L2, as determined by the analysis shown in (B) and (C). The value for each time point was the average of the results from three experiments, and the error bars represent standard deviation. The y-axis represents the relative methylation levels. This value was calculated by band-a (Ig)/band-a (EF1␣); the signals for band-a was derived by quantifying the blots as shown in (B) and (C) with PhosphoImager. The value for L1 cells at 12 h tamoxifen treatment was set as 1. Only band-a was used because it represents complete methylation of the DNA fragment. (E) The methylation analysis of the -globin locus. This figure is illustrated the same way as (B).
Dam methylation levels do correlate with chromatin accessibility in our system, we analyzed the methylation status of the -globin locus, which is not expressed in DT40 cells, and should represent a less accessible locus relative to the Ig and EF1␣ loci, which are actively transcribed in DT40 cells. Consistent with the expectation, the methylation kinetics of the -globin locus is obviously slower compared with those in both the Ig and the EF1␣ loci (compare Fig. 2E with B and C). This result shows that the Dam methylation assay is capable of revealing differences in chromatin accessibility in our system; if there were major reductions in chromatin accessibility in L2, we should have detected it with this assay. In this regard, the results from both the Dam methylation and I-SceI experiments are in agreement. Thus, we conclude that Region A is dispensable for chromatin accessibility at the Ig locus, with the assumption that AID, I-SceI, and Dam access chromatin similarly. 3.2. Region A facilitates the recruitment of AID to the Ig locus The second model for Region A function is that the element may recruit AID to the Ig locus. We used the DamID assay to test this model. In this method, the protein of interest is fused to Dam; if the protein binds specifically to a particular site in the genome, that locus would be preferentially methylated by the fusion protein rel-
ative to other loci (van Steensel and Henikoff, 2000). To apply this method to our study, we fused AID with Dam and expressed the fusion protein in DT40 cells with the tamoxifen inducible system. The AID-Dam fusion is functional in gene conversion experiments, and its activity is dependent on Region A (Fig. 3A and Supplemental Table 2). We compared AID-Dam mediated methylation of the gene conversion substrate in L1 and L2 cells, and found that the methylation level in L2 cells was lower than that in L1 cells (Fig. 3B). As a control, we analyzed the methylation pattern at the EF1␣ locus, and found no difference between L1 and L2 cells (Fig. 3B). We calculated the methylation ratio of Ig/EF1␣, and found an approximately 50% reduction in the methylation levels of gene conversion substrate in L2 cells (Fig. 3C); although the reduction is modest, it is reproducible in three experiments. This result suggests that the deletion of Region A impairs the interaction of AID-Dam fusion with the gene conversion substrate at the Ig locus. To test whether the recruitment effect of Region A is specific to AID, we generated a Dam fusion with Apobec-1, which is the closest homologue of AID within the cytidine deaminase family (Wedekind et al., 2003). In spite of the homology, Apobec-1 is unable to substitute for AID in Ig diversification, potentially due to the lack of recruitment to the Ig loci (Fugmann et al., 2004). We expressed the Apobec-1-Dam fusion in L1 and L2 cells, and as shown in Fig. 3D, Apobec1-Dam
1864
Y. Kim, M. Tian / Molecular Immunology 47 (2010) 1860–1865
Fig. 3. Deletion of Region A reduces the interaction of AID-Dam with the gene conversion substrate at the Ig locus. (A) This dot plot shows the gene conversion analysis of AID−/− cells complemented with AID-Dam. This figure is illustrated the same way as Fig. 1C. (B) The methylation analysis with AID-Dam. This figure is illustrated the same way as Fig. 2B. (C) This plot is a quantification of the methylation analysis as shown in (B). The value for each time point was the average of the results from three experiments, and the error bars represent standard deviation. The y-axis represents the relative methylation levels. This value was calculated by band-a (Ig)/band-b (EF1␣); the signals for band-a or b was derived by quantifying the blots as shown in (B) with PhosphoImager. The reason to use band-b for EF1␣ is because band-a is not detectable. Since the EF1␣ signal only serves as normalization, the use of band-b should not affect the outcome of the analysis. The value for L1 cells at 12 h tamoxifen treatment was set as 1. (D) This figure shows the methylation analysis with Apobec1-Dam. This figure is illustrated the same way as Fig. 2B. (E) This histogram compares the methylation of the gene conversion substrate (L1) mediated by AID-Dam and Apobec1-Dam after 12 h of tamoxifen treatment as shown in (B) and (C). Relative methylation levels are calculated the same way as in (C). The data were derived from three experiments; the error bar represents standard deviation. The average methylation levels mediated by AID-Dam or Apobec1-Dam are indicated below the plot, and the value for AID-Dam is set as 1.
methylates L1 and L2 to essentially the same extents. Moreover, relative to AID-Dam, Apobec1-Dam methylates the gene conversion substrate poorly; after normalization against the methylation levels at the EF1␣ locus, AID-Dam methylates the gene conversion substrate 11-fold more efficiently than Apobec1-Dam (Fig. 3E). In summary, these results suggest that AID-Dam is preferentially recruited to the Ig locus, and this recruitment involves Region A. 4. Discussion In our previous study, we identified a regulatory region (Region A) that is important for AID mediated gene conversion. Our
present work is aimed at understanding how Region A facilitates AID function. We tested two models: first, Region A may render local DNA more accessible to AID; second, Region A may recruit AID to adjacent DNA. Our results are consistent with the second hypothesis. It remains to be determined how Region A recruits AID to the Ig locus. The region contains binding sites for several transcription factors, including E2A, NFB, IRF4, PU.1, Mef2, and Octamer binding proteins; some of these factors have been implicated in AID function (Schoetz et al., 2006; Kim and Tian, 2009). It is possible that these factors may form an enhanceosome that recruits AID to the Ig locus.
Y. Kim, M. Tian / Molecular Immunology 47 (2010) 1860–1865
Deletion of Region A reduces, but does not abolish the targeting of AID to the Ig locus; as shown in Fig. 3, even in L2 cells (without Region A), AID-Dam still methylates the gene conversion substrate more efficiently than Apobec1-Dam. This observation suggests that, besides Region A, there may be other AID targeting elements at the Ig locus. In this regard, we showed previously that deletion of a 7.4-kb region (Region B) downstream of Region A also reduces AID mediated gene conversion, albeit to a lesser extent relative to the deletion of Region A. Both Region A and Region B overlaps with regulatory elements identified in other studies, including the 4.1-kb 3 regulatory region (Kothapalli et al., 2008) and the 9.8-kb diversification activator (DIVAC) (Blagodatski et al., 2009). The overall picture is that multiple elements contribute to the targeting of AID function, which may explain the modest reduction of AID recruitment to the Ig locus in the absence of Region A. Deletion of Region A appears to have a more prominent impact on AID mediated gene conversion (a reduction of 15-fold, Kim and Tian, 2009) than on AID recruitment (2-fold). One potential explanation is that gene conversion frequency does not correlate linearly with the levels of AID recruitment, but a more interesting possibility is that Region A contributes to AID function in additional ways. The speculation is prompted by the intriguing observation that Region A appears to interfere with I-SceI mediated gene conversion. Since the I-SceI experiments were done in AID−/− cells, the phenomenon has nothing to do AID recruitment, and might reflect the impact of Region A on the gene conversion substrate that would favor AID mediated deamination, but interferes with I-SceI mediated cleavage of the DNA target. The resolution of this issue in the future may reveal novel regulatory mechanisms of AID function by Region A as well as other cis-acting elements in the Ig locus. Acknowledgements We thank Dr. Arakawa and Dr. Buerstedde for providing with us the DT40cre1 cell line. This work was supported with funding from the University of Texas at Austin. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.molimm.2010.02.025. References Arakawa, H., Lodygin, D., Buerstedde, J.M., 2001. Mutant loxP vectors for selectable marker recycle and conditional knock-outs. BMC Biotechnol. 1, 7. Arakawa, H., Hauschild, J., Buerstedde, J.M., 2002. Requirement of the activationinduced deaminase (AID) gene for immunoglobulin gene conversion. Science 295, 1301–1306. Basu, U., Chaudhuri, J., Alpert, C., Dutt, S., Ranganath, S., Li, G., Schrum, J.P., Manis, J.P., Alt, F.W., 2005. The AID antibody diversification enzyme is regulated by protein kinase A phosphorylation. Nature 438, 508–511. Blagodatski, A., Batrak, V., Schmidl, S., Schoetz, U., Caldwell, R.B., Arakawa, H., Buerstedde, J.M., 2009. A cis-acting diversification activator both necessary and sufficient for AID-mediated hypermutation. PLoS Genet. 5, e1000332. Chaudhuri, J., Khuong, C., Alt, F.W., 2004. Replication protein A interacts with AID to promote deamination of somatic hypermutation targets. Nature 430, 992–998.
1865
Conticello, S.G., Ganesh, K., Xue, K., Lu, M., Rada, C., Neuberger, M.S., 2008. Interaction between antibody-diversification enzyme AID and spliceosome-associated factor CTNNBL1. Mol. Cell 31, 474–484. Di Noia, J.M., Neuberger, M.S., 2007. Molecular mechanisms of antibody somatic hypermutation. Annu. Rev. Biochem. 76, 1–22. Duquette, M.L., Handa, P., Vincent, J.A., Taylor, A.F., Maizels, N., 2004. Intracellular transcription of G-rich DNAs induces formation of G-loops, novel structures containing G4 DNA. Genes Dev. 18, 1618–1629. Fugmann, S.D., Rush, J.S., Schatz, D.G., 2004. Non-redundancy of cytidine deaminases in class switch recombination. Eur. J. Immunol. 34, 844–849. Harris, R.S., Sale, J.E., Petersen-Mahrt, S.K., Neuberger, M.S., 2002. AID is essential for immunoglobulin V gene conversion in a cultured B cell line. Curr. Biol. 12, 435–438. Kim, Y., Tian, M., 2009. NF-B family of transcription factor facilitates gene conversion in chicken B cells. Mol. Immunol. 46, 3283–3291. Kothapalli, N., Norton, D.D., Fugmann, S.D., 2008. Cutting edge: a cis-acting DNA element targets AID-mediated sequence diversification to the chicken Ig light chain gene locus. J. Immunol. 180, 2019–2023. Liu, M., Duke, J.L., Richter, D.J., Vinuesa, C.G., Goodnow, C.C., Kleinstein, S.H., Schatz, D.G., 2008. Two levels of protection for the B cell genome during somatic hypermutation. Nature 451, 841–845. Muramatsu, M., Kinoshita, K., Fagarasan, S., Yamada, S., Shinkai, Y., Honjo, T., 2000. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102, 553–563. Odegard, V.H., Schatz, D.G., 2006. Targeting of somatic hypermutation. Nat. Rev. Immunol. 6, 573–583. Pham, P., Bransteitter, R., Petruska, J., Goodman, M.F., 2003. Processive AID-catalysed cytosine deamination on single-stranded DNA simulates somatic hypermutation. Nature 424, 103–107. Revy, P., Muto, T., Levy, Y., Geissmann, F., Plebani, A., Sanal, O., Catalan, N., Forveille, M., Dufourcq-Labelouse, R., Gennery, A., Tezcan, I., Ersoy, F., Kayserili, H., Ugazio, A.G., Brousse, N., Muramatsu, M., Notarangelo, L.D., Kinoshita, K., Honjo, T., Fischer, A., Durandy, A., 2000. Activation-induced cytidine deaminase (AID) deficiency causes the autosomal recessive form of the Hyper-IgM syndrome (HIGM2). Cell 102, 565–575. Rouet, P., Smih, F., Jasin, M., 1994. Introduction of double-strand breaks into the genome of mouse cells by expression of a rare-cutting endonuclease. Mol. Cell Biol. 14, 8096–8106. Schoetz, U., Cervelli, M., Wang, Y.D., Fiedler, P., Buerstedde, J.M., 2006. E2A expression stimulates Ig hypermutation. J. Immunol. 177, 395–400. Shinkura, R., Tian, M., Smith, M., Chua, K., Fujiwara, Y., Alt, F.W., 2003. The influence of transcriptional orientation on endogenous switch region function. Nat. Immunol., 435–441. Shivarov, V., Shinkura, R., Doi, T., Begum, N.A., Nagaoka, H., Okazaki, I.M., Ito, S., Nonaka, T., Kinoshita, K., Honjo, T., 2009. Molecular mechanism for generation of antibody memory. Philos. Trans. R. Soc. Lond. B Biol. Sci. 364, 569–575. Singh, J., Klar, A.J., 1992. Active genes in budding yeast display enhanced in vivo accessibility to foreign DNA methylases: a novel in vivo probe for chromatin structure of yeast. Genes Dev. 6, 186–196. Tian, M., Alt, F.W., 2000. Transcription-induced cleavage of immunoglobulin switch regions by nucleotide excision repair nucleases in vitro. J. Biol. Chem. 275, 24163–24172. van Steensel, B., Henikoff, S., 2000. Identification of in vivo DNA targets of chromatin proteins using tethered dam methyltransferase. Nat. Biotechnol. 18, 424–428. Vuong, B.Q., Lee, M., Kabir, S., Irimia, C., Macchiarulo, S., McKnight, G.S., Chaudhuri, J., 2009. Specific recruitment of protein kinase A to the immunoglobulin locus regulates class-switch recombination. Nat. Immunol. 10, 420–426. Wedekind, J.E., Dance, G.S., Sowden, M.P., Smith, H.C., 2003. Messenger RNA editing in mammals: new members of the APOBEC family seeking roles in the family business. Trends Genet. 19, 207–216. Yang, S.Y., Fugmann, S.D., Schatz, D.G., 2006. Control of gene conversion and somatic hypermutation by immunoglobulin promoter and enhancer sequences. J. Exp. Med. 203, 2919–2928. Yélamos, J., Klix, N., Goyenechea, B., Lozano, F., Chui, Y.L., González Fernández, A., Pannell, R., Neuberger, M.S., Milstein, C., 1995. Targeting of non-Ig sequences in place of the V segment by somatic hypermutation. Nature 376, 225–229. Yu, K., Chedin, F., Hsieh, C.L., Wilson, T.E., Lieber, M.R., 2003. R-loops at immunoglobulin class switch regions in the chromosomes of stimulated B cells. Nat. Immunol. 4, 442–451. Zarrin, A.A., Alt, F.W., Chaudhuri, J., Stokes, N., Kaushal, D., Du Pasquier, L., Tian, M., 2004. An evolutionarily conserved target motif for immunoglobulin class-switch recombination. Nat. Immunol. 5, 1275–1281.