Hypermutation Rate Normalized by Chronological Time
http://www.100md.com
免疫学杂志 2005年第9期
Abstract
It is generally believed that in cells undergoing Ig somatic hypermutation, more cell divisions result in more mutations. This is because DNA synthesis and replication is thought to play roles in the known mechanisms–cytidine deamination and subsequent conversion to thymidine, uracil-DNA glycosylase-mediated repair, mismatch repair, and DNA synthesis by error-prone polymerases. In this study, we manipulated the number of cell generations by varying the rate at which cultures of a mouse cell line were replenished with fresh medium. We found that the frequency of mutants does not necessarily increase with the number of cell generations. On the contrary, a greater number of divisions can lead to a lower frequency of mutants, indicating that cell division is not a rate-limiting step in the hypermutation process. Thus, when comparing mutation rates, we suggest that rates are more appropriately expressed as mutations per day than per cell generation.
Introduction
The incidence of mutation often coincides with cell division. During the DNA synthesis that precedes cell division, imperfect copying and repair can lead to mutations. For this reason, mutation rates are almost always expressed with units of mutations per cell generation, be it bacteria (1, 2), yeast (3), plants (4), animals (5), or Ig somatic hypermutation (6, 7, 8, 9, 10, 11, 12, 13).
Activated B cells can mutate their DNA at a rate much higher than the normal, spontaneous rate–a phenomenon called Ig somatic hypermutation. Mutations that occur in the Ig V region can lead to better Ab affinity, allowing the immune system to further evolve and improve its Ab repertoire (as reviewed in Ref. 14). Somatic hypermutation requires activation-induced cytidine deaminase (AID)3 (15). AID is thought to deaminate cytosine bases on DNA (16, 17, 18) to produce uracil. The uracil base can be paired with adenine upon the next round of DNA synthesis, and then with one further round, that adenine will pair with thymine. Thus, DNA replication can parlay the activity of AID into a C to T mutation (14). Uracil-DNA glycosylase can excise the uracil base, and ensuing DNA synthesis opposite the apyridimic site can lead to mutations to G, A, or T (19). Mismatch repair mechanisms (20, 21, 22, 23) are believed to detect mismatched uracil bases (e.g., U-G) and excise stretches of ssDNA (24) near the site of cytidine deamination. The strand containing the uracil or even the strand opposite (21) the uracil might be excised, and error-prone polymerases (25, 26) repairing the section may introduce incorrectly paired bases, not only at the site of the original deaminated C but also other nearby bases. Mismatch repair coupled with error-prone polymerases thus can lead to more mismatches; when the DNA is next copied, these mismatches can lead to permanent mutations.
In all cases where uracil in the DNA leads to mutation, DNA synthesis is needed to finalize the mutation process. For this reason it is commonly believed that the rate of hypermutation increases with the rate of DNA replication, and thus also with that of cell division. We have created several reporter constructs and used them to quantify mutation rates. In this study, we investigate whether the number of cell generations increases the frequency of mutants.
Materials and Methods
Vector construction
Constructs were based on the Moloney murine leukemia virus retroviral vector contained in p102.21 (kindly provided by J. B. Lorens, Rigel Pharma, South San Francisco, CA). They contain the internal ribosome entry site (IRES) and puromycin resistance gene of pIRESpuro3 (BD Clontech). Mutations were introduced into enhanced GFP (EGFP) (see Fig. 2A) at the fluorescent chromaphore to create the reporter genes GFP-CT (EGFP C198G, T199C) and GFP-CG (EGFP C201T, G203C, C204T). GFP* contains the premature stop codon (TAG) at bases 319–321 and was used previously (27).
FIGURE 2. Measuring mutation rates with GFP-based reporters. A, Sequences of reporters GFP-CT, GFP-CG, and GFP* compared with GFP (EGFP of BD Clontech). Fluorescence was restored when the boxed nucleotides were mutated. Base locations are given for the starting and ending bases of the sequences. B, Reporters were expressed using retroviral vectors. Vector components: LTR, long terminal repeat; , retroviral packaging signal; GFP rep., GFP reporter gene; PuroR, puromycin resistance gene. C, Flow cytometry of GFP expression. The mean fluorescence of GFP-negative cells was set to 3.0. Cells with fluorescence >100 (bar labeled GFP+) were counted as GFP-positive.
Cell culture
Retroviral vectors were packaged using PhoenixEco cells (ATCC SD 3444; American Type Culture Collection). 18-81 cells were infected with the vectors and selected in the presence of 2.5 μg/ml puromycin for 7 days. To start experiments with zero fluorescent mutants, fluorescent cells were removed by FACS. After FACS sorting, cells were grown with 1.5 μg/ml puromycin and fluorescent cells were detected by flow cytometry. All 18-81 cultures were grown in RPMI 1640 medium with 10% FCS. Cells were split at the same time daily by different ratios to control the number of cell generations. The splitting ratios and volumes (milliliters) in terms of culture plus fresh media are: 1/2, 10 + 10; 2/5, 8 + 12; 1/3, 7 + 14; 2/7 6 + 21; 1/4, 5 + 15; 1/5, 4 + 16.
DNA sequencing
DNA from GFP-positive cells was isolated, and the GFP reporter genes were PCR-amplified using Pfu polymerase (Stratagene). The PCR products were incubated with Taq polymerase to add deoxyadenosine overhangs and then cloned into pCR2.1-TOPO (Invitrogen Life Technologies). The plasmids were then amplified in Escherichia coli and sequenced.
Quantitative RT-PCR
Expression of mouse AID and the GFP reporter constructs were measured using quantitative RT-PCR. RNA was extracted from cells and cDNA produced by standard application of reverse transcriptase and random hexamer primers. PCR was performed using the ABI PRISM 7700 machine. AID and reporter expression was normalized to the amount of expressed mouse -actin and hypoxanthine phosphoribosyltransferase (HPRT). PCR primers and probes used were AID: 5'-GAAAATTCTGTCCGGCTAACCA-3', 5'-TCGCAAGTCATCGACTTCGT-3', 5'-[6-FAM]-TCGGCGCATCCTTTTGCCCTT-[TAMRA]-3'; GFP reporters: 5'-CCACATGAAGCAGCACGACT-3', 5'-TGCGCTCCTGGACGTAGC-3', 5'-[6-FAM]-TTCAAGTCCGCCATGCCCGAA-[TAMRA]-3'; -actin: 5'-AGGTCATCACTATTGGCAACGA3', 5'-CACTTCATGATGGAATTGAATGTAGTT-3', 5'-[6-FAM]-TGCCACAGGATTCCATACCCAAGAAGG-[TAMRA]-3'; HPRT: 5'-TGGAAAGAATGTCTTGATTGTTGAA-3', 5'-AGCTTGCAACCTTAACCATTTTG-3', 5'-[6-FAM]-CAAACTTTGCTTTCCCTGGTTAAGCAGTACAGC-[TAMRA]-3'.
Results
Mutation rates were measured in 18-81 mouse cells. 18-81 is a cell line of B lineage that expresses endogenous AID and hypermutates its Ig V region (11). Like activated B cells, 18-81 divides quickly (as low as a 5 h doubling time) and also undergoes Ig class switching, generally a change from μ H chain to 2b. 18-81 was also ideal for this study because it is almost completely nonadherent and allowed cultures to be split precisely with reproducible growth conditions each day.
Because the cell line proliferates so rapidly, the cell population was able to reach the same stationary cell concentration each day, even when the cultures were split by different amounts daily. For this study, we found that the number of cell generations per day could be varied by the amount that the culture was split and replenished with fresh medium. For example, a culture split by 1/2 (1 part culture plus 1 part fresh medium), on average, undergoes one cell division by the next day. A culture split by 1/4 averages two cell generations per day. Staining with CFSE, a vital dye that enables tracking of cell division, and monitoring with flow cytometry confirmed that varying the ratio of cell splitting changed the average number of cell generations (Fig. 1). However, because the measurement of hypermutation rates in our experiments requires several weeks and because CFSE could only be detected in our cells during the first 4 days after staining, CFSE tracking could not be used concomitantly with our mutation rate measurement experiments.
FIGURE 1. Proliferation of cells determined by CFSE staining. 18-81 cells were spun down, stained with CFSE, washed, and resuspended into the original culture medium (supernatant) at the original cell concentration. After 1 day, the mean CFSE fluorescence per cell was measured by flow cytometry. Cell division led to loss of fluorescence; thus, cells having undergone more generations would have less fluorescence. Because cells split with a higher fraction of fresh media had less fluorescence, we conclude that the different cell splitting fractions led to different amounts of cell generations. We note though, that loss of fluorescence, in part, could also have occurred independent of cell generation (e.g., CFSE leakage).
The mutation reporter genes (Fig. 2A) were variants of GFP that become fluorescent when mutated. GFP-CT becomes fluorescent when base 199 mutates from C to T; GFP-CG becomes fluorescent when base 203 mutates from C to G; GFP* becomes fluorescent when the premature stop codon reverts to a codon assigned to an amino acid. The GFP reporter genes were expressed in 18-81 cells using a retroviral vector (Fig. 2B) and mutations that restored fluorescence were detected by flow cytometry (Fig. 2C). Sequencing the reporter genes from fluorescent cells verified that fluorescence was not restored by unexpected, proximal mutations. In all cases GFP-CT sequences mutated C to T (10/10) and GFP-CG sequences mutated C to G (9/9). Sequences of GFP* from fluorescent cells all contained mutations at base 321; these mutations were primarily G to C (11/12 sequences) but also G to T (1/12).
Accumulation of mutations was monitored continually over 32 days (Fig. 3A). Each measured time point consisted of data from 106 cells. Because many cells could be easily analyzed via flow cytometry, we were able to calculate mutation rates (Fig. 3B) more accurately than by mere DNA sequencing.
FIGURE 3. Rate of hypermutation at different rates of cell generation. A, Mutations causing reporters to become GFP-positive were monitored continuously. Reporters tracked accumulation of specific mutations; GFP-CT (C T), GFP-CG (CG), and stop codon reversion, GFP* (GFP rev.). The rate of cell generations (gen./day) was determined by the daily cell splitting ratio. B, Mutation rates as a function of cell generation rate. CT (), CG (?), stop codon reversion (). If mutation rate increased proportionally with that of cell generation, the mutation rate would have the slope of the dashed line.
We measured mutation frequencies from cultures ranging from 1.0 to 2.3 generations per day. From these mutation frequencies we calculated mutation rates. Under these culture conditions, cultures could be stably and continuously maintained. Cells appeared to be healthy and viable staining showed <1% dead cells. To our surprise, we found that the number of mutations did not correlate positively with the number of cell generations. This was true of C to T, C to G, and the TAG reversion mutations. In all cases, maximal mutation rates were observed when cells produced 1.3 generations per day and decreased as the number of cell generations per day was increased. We found that at one generation per day, the mutation frequency was low. Though the mutation rate of GFP* was approximately twice that of GFP-CT and GFP-CG, any similarities or differences in rates could be due to differences in sequence or position of the mutation site within the GFP reporter gene. Nonetheless, it was surprising that the C to T and C to G mutation rates were approximately the same over the range of cell growth conditions. Because C to T mutations presumably require only AID activity and DNA synthesis, and C to G mutations likely require AID plus additional downstream factors and DNA synthesis, this could indicate that AID is a rate-limiting factor while downstream factors (uracil-DNA-glycosylase, mismatch repair, error-prone polymerases) are not.
In our experiments we varied the media replenishing (i.e., splitting) ratio of the cultures, thereby controlling the number of cell generations per day. Clearly, more cell generations did not necessitate more mutations. When we measured the expression levels of AID (Fig. 4A) and the GFP reporter (Fig. 4B), we found that expression varied over our range of culture splitting conditions. AID expression was optimal when cultures were split by 2/5 and 1/3 (1.6 and 1.8 generations per day, respectively); reporter expression was highest when cultures were split by 1/2 and 1/3 (1.0 and 1.3 generations per day, respectively) and decreased if cells proliferated more.
FIGURE 4. AID and GFP expression at different rates of cell generation. Expression was determined by real-time quantitative RT-PCR and normalized to -actin () and HPRT (?) expression. Expression of (A) AID and (B) GFP reporter. C, Product of AID and GFP reporter expression levels.
Discussion
In addition to AID, hypermutation is thought to require transcription of the template DNA (6, 17, 28, 29). If levels of AID and reporter transcription both correlate positively with the mutation rate, and are independent of each other, then the product of both factors might also reflect the mutation rate. That is,
rate [AID]
rate [GFP]
rate [AID][GFP].
When the AID and reporter expression levels from our experiments are multiplied (Fig. 4C), the resulting curve does reflect the changes in mutation rate that occurred with varying the number of generations per day (Fig. 3B). By the terms of this model, in the cultures where cells divide on average only 1.3 times per day, high reporter transcription plus moderate levels of AID can result in a high mutation rate. Furthermore, high mutation rates can occur when there is high AID expression but less reporter transcription. The correlation of the mutation rate with AID and reporter expression suggests that AID and transcription are primary limiting factors in hypermutation. Although cell division is likely an integral downstream step in the mutation process, the fact that the correlation occurs independently of cell division (i.e., does not require cell generation number as a dependent variable) suggests that even if AID and GFP were held constant, varying the number of generations (under the conditions tested here) would not greatly affect the number of mutations. Thus the incidence of cell division is high enough that it is not a limiting factor for hypermutation under our laboratory growth conditions. In other words, under these conditions, cell division is not the "bottleneck" point in the hypermutation pathway. However, one might still predict that at some very low rate of cell division, the mutation rate would diminish. Our findings are in line with the observations of Bachl et al. (30) who noted that "variation in cell numbers seems to be much greater than the variation in the frequency of mutant cells. If these variations are accurate, they suggest that mutation is not highly coupled to cell division in these cells."
Our results do not imply that there cannot be other factors that affect the rate of hypermutation. For example, nuclear import and export (31, 32) and phosphorylation of AID (33) have recently been suggested to affect the rate of hypermutation. It has been suggested that other limiting factors exist specifically for mutation of Ig genes in B cells (31) and such factors may or may not play a role in our experimental system here.
The only variable that we sought to manipulate was the number of cell generations per day. However, the protocol also causes variation in the freshness of the growth medium (a highly diluted culture has more fresh medium), and time spent at cell growth saturation, which might influence phenotypic expression. The more diluted cultures spend less time at saturation before the GFP is measured. Could a variation in "phenotypic lag", i.e., a delay in gene expression, have influenced our results? A delay between the time when the mutation occurs and the time when a sufficient amount of GFP can be detected by flow cytometry could skew rate measurements over a period of time that is comparable in length to the phenotypic lag. However, our measurements of mutation frequencies were taken over a month, a period of time that is likely much longer than any lag that might occur between the incidence of mutation and expression of GFP fluorescence. Therefore, irrespective of any short lags in expression, the cultures will have reached a steady state, and the measured rates will not be affected by a phenotypic lag.
We found that fewer generations, and thus more time spent at saturation, gave higher frequencies of mutations. Accumulation of metabolites in the growth medium and time spent at saturation might be also mutagenic. Furthermore, variation of cell concentration or nutrients could affect hypermutation in ways unrelated to replication. But unless there are factors of a higher magnitude of importance than AID or DNA transcription, this would not invalidate our proposition that cell division is not a rate-limiting step for hypermutation.
In the study of mutations, whether Ig hypermutation or not, mutation rates are nearly always expressed on a per cell generation basis. This is because in most cases, mutations are introduced during DNA replication. However, in our study, we find that hypermutation is likely an important exception–cell division is not rate limiting. It follows then that rates should rather be reported on a per time (e.g., mutations/base pair/day) basis.
We have shown that increased cell generations not only do not necessitate increased mutation, but also that high cell proliferation can lead to decreased AID levels and mutation rates. Although our results were obtained using a cell line and may not accurately reflect an in vivo system, in the animal, it might make sense to decrease the mutation rate when cells are dividing quickly; once B cells with good Ab affinity are selected to undergo increased clonal proliferation, reducing AID levels could prevent further mutagenesis at the Ig loci or other locales throughout the genome. Although it has already been recognized that AID and transcription are important factors for hypermutation, we demonstrated here that AID activity and transcription alone are largely sufficient to explain the incidence of mutation. These activities, unlike cell division, are likely the dominant rate-limiting steps.
Acknowledgments
We are grateful to Frank Stahl for critically reading the manuscript; and for J. Lorens (Rigel Pharma, South San Francisco) for providing the retroviral vector; and to Desirée Yang, Freia Spillmann, Melissa Lodoen, Jessica Hamerman, and J. A. Bluestone for their efforts—cdub out.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This study was supported by National Institutes of Health (NIH) Grant AG20684. C.L.W. was supported by the NIH Immunology Training Grant (NIH/NIAID T32 AI07334).
2 Address correspondence and reprint requests to Department of Microbiology and Immunology, University of California, Box 0414, San Francisco, CA 94103-0414. E-mail address: cliffw@itsa.ucsf.edu
3 Abbreviations used in this paper: AID, activation-induced cytidine deaminase; IRES, internal ribosome entry site; EGFP, enhanced GFP; HPRT, hypoxanthine phosphoribosyltransferase.
Received for publication November 16, 2004. Accepted for publication February 8, 2005.
References
Luria, S. E., M. Delbrück. 1943. Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28:491.
Hashem, V. I., W. A. Rosche, R. R. Sinden. 2002. Genetic assays for measuring rates of (CAG): (CTG) repeat instability in Escherichia coli. Mutat. Res. 502:25
Wloch, D. M., K. Szafraniec, R. H. Borts, R. Korona. 2001. Direct estimate of the mutation rate and the distribution of fitness effects in the yeast Saccharomyces cerevisiae. Genetics 159:441.
Caetano-Anolles, G.. 1999. High genome-wide mutation rates in vegetatively propagated bermudagrass. Mol. Ecol. 8:1211.
Eckert, K. A., G. Yan, S. E. Hile. 2002. Mutation rate and specificity analysis of tetranucleotide microsatellite DNA alleles in somatic human cells. Mol. Carcinog. 34:140.
Bachl, J., C. Carlson, V. Gray-Schopfer, M. Dessing, C. Olsson. 2001. Increased transcription levels induce higher mutation rates in a hypermutating cell line. J. Immunol. 166:5051.
Yoshikawa, K., I. M. Okazaki, T. Eto, K. Kinoshita, M. Muramatsu, H. Nagaoka, T. Honjo. 2002. AID enzyme-induced hypermutation in an actively transcribed gene in fibroblasts. Science 296:2033.
Poltoratsky, V., C. J. Woo, B. Tippin, A. Martin, M. F. Goodman, M. D. Scharff. 2001. Expression of error-prone polymerases in BL2 cells activated for Ig somatic hypermutation. Proc. Natl. Acad. Sci. USA 98:7976.
Lin, M. M., M. Zhu, M. D. Scharff. 1997. Sequence dependent hypermutation of the immunoglobulin heavy chain in cultured B cells. Proc. Natl. Acad. Sci. USA 94:5284.
Wabl, M., H. M. Jack, J. Meyer, G. Beck-Engeser, R. C. von Borstel, C. M. Steinberg. 1987. Measurements of mutation rates in B lymphocytes. Immunol. Rev. 96:91
Wabl, M., P. D. Burrows, A. von Gabain, C. Steinberg. 1985. Hypermutation at the immunoglobulin heavy chain locus in a pre-B-cell line. Proc. Natl. Acad. Sci. USA 82:479.
Zhu, M., J. L. Rabinowitz, N. S. Green, B. J. Kobrin, M. D. Scharff. 1995. A well-differentiated B-cell line is permissive for somatic mutation of a transfected immunoglobulin heavy-chain gene. Proc. Natl. Acad. Sci. USA 92:2810.
Baumal, R., B. K. Birshtein, P. Coffino, M. D. Scharff. 1973. Mutations in immunoglobulin-producing mouse myeloma cells. Science 182:164.
Neuberger, M. S., R. S. Harris, J. Di Noia, S. K. Petersen-Mahrt. 2003. Immunity through DNA deamination. Trends Biochem. Sci. 28:305.
Muramatsu, M., K. Kinoshita, S. Fagarasan, S. Yamada, Y. Shinkai, T. Honjo. 2000. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102:553.
Petersen-Mahrt, S. K., R. S. Harris, M. S. Neuberger. 2002. AID mutates E. coli suggesting a DNA deamination mechanism for antibody diversification. Nature 418:99.
Ramiro, A. R., P. Stavropoulos, M. Jankovic, M. C. Nussenzweig. 2003. Transcription enhances AID-mediated cytidine deamination by exposing single-stranded DNA on the nontemplate strand. Nat. Immunol. 4:452.
Pham, P., R. Bransteitter, J. Petruska, M. F. Goodman. 2003. Processive AID-catalysed cytosine deamination on single-stranded DNA simulates somatic hypermutation. Nature 424:103.
Di Noia, J., M. S. Neuberger. 2002. Altering the pathway of immunoglobulin hypermutation by inhibiting uracil-DNA glycosylase. Nature 419:43.
Rada, C., M. R. Ehrenstein, M. S. Neuberger, C. Milstein. 1998. Hot spot focusing of somatic hypermutation in MSH2-deficient mice suggests two stages of mutational targeting. Immunity 9:135.
Cascalho, M., J. Wong, C. Steinberg, M. Wabl. 1998. Mismatch repair co-opted by hypermutation. Science 279:1207.
Kim, N., G. Bozek, J. C. Lo, U. Storb. 1999. Different mismatch repair deficiencies all have the same effects on somatic hypermutation: intact primary mechanism accompanied by secondary modifications. J. Exp. Med. 190:21.
Martomo, S. A., W. W. Yang, P. J. Gearhart. 2004. A role for msh6 but not msh3 in somatic hypermutation and class switch recombination. J. Exp. Med. 200:61.
Bardwell, P. D., C. J. Woo, K. Wei, Z. Li, A. Martin, S. Z. Sack, T. Parris, W. Edelmann, M. D. Scharff. 2004. Altered somatic hypermutation and reduced class-switch recombination in exonuclease 1-mutant mice. Nat. Immunol. 5:224
Zeng, X., D. B. Winter, C. Kasmer, K. H. Kraemer, A. R. Lehmann, P. J. Gearhart. 2001. DNA polymerase is an A-T mutator in somatic hypermutation of immunoglobulin variable genes. Nat. Immunol. 2:537.
Faili, A., S. Aoufouchi, E. Flatter, Q. Gueranger, C. A. Reynaud, J. C. Weill. 2002. Induction of somatic hypermutation in immunoglobulin genes is dependent on DNA polymerase . Nature 419:944.
Wang, C. L., R. A. Harper, M. Wabl. 2004. Genome-wide somatic hypermutation. Proc. Natl. Acad. Sci. USA 101:7352.
Chaudhuri, J., M. Tian, C. Khuong, K. Chua, E. Pinaud, F. W. Alt. 2003. Transcription-targeted DNA deamination by the AID antibody diversification enzyme. Nature 422:726.
Peters, A., U. Storb. 1996. Somatic hypermutation of immunoglobulin genes is linked to transcription initiation. Immunity 4:57.
Bachl, J., M. Dessing, C. Olsson, R. C. von Borstel, C. Steinberg. 1999. An experimental solution for the Luria-Delbruck fluctuation problem in measuring hypermutation rates. Proc. Natl. Acad. Sci. USA 96:6847.
McBride, K. M., V. Barreto, A. R. Ramiro, P. Stavropoulos, M. C. Nussenzweig. 2004. Somatic hypermutation is limited by CRM1-dependent nuclear export of activation-induced deaminase. J. Exp. Med. 199:1235.
Ito, S., H. Nagaoka, R. Shinkura, N. Begum, M. Muramatsu, M. Nakata, T. Honjo. 2004. Activation-induced cytidine deaminase shuttles between nucleus and cytoplasm like apolipoprotein B mRNA editing catalytic polypeptide 1. Proc. Natl. Acad. Sci. USA 101:1975.
Chaudhuri, J., C. Khuong, F. W. Alt. 2004. Replication protein A interacts with AID to promote deamination of somatic hypermutation targets. Nature 430:992.(Clifford L. Wang and Matt)
It is generally believed that in cells undergoing Ig somatic hypermutation, more cell divisions result in more mutations. This is because DNA synthesis and replication is thought to play roles in the known mechanisms–cytidine deamination and subsequent conversion to thymidine, uracil-DNA glycosylase-mediated repair, mismatch repair, and DNA synthesis by error-prone polymerases. In this study, we manipulated the number of cell generations by varying the rate at which cultures of a mouse cell line were replenished with fresh medium. We found that the frequency of mutants does not necessarily increase with the number of cell generations. On the contrary, a greater number of divisions can lead to a lower frequency of mutants, indicating that cell division is not a rate-limiting step in the hypermutation process. Thus, when comparing mutation rates, we suggest that rates are more appropriately expressed as mutations per day than per cell generation.
Introduction
The incidence of mutation often coincides with cell division. During the DNA synthesis that precedes cell division, imperfect copying and repair can lead to mutations. For this reason, mutation rates are almost always expressed with units of mutations per cell generation, be it bacteria (1, 2), yeast (3), plants (4), animals (5), or Ig somatic hypermutation (6, 7, 8, 9, 10, 11, 12, 13).
Activated B cells can mutate their DNA at a rate much higher than the normal, spontaneous rate–a phenomenon called Ig somatic hypermutation. Mutations that occur in the Ig V region can lead to better Ab affinity, allowing the immune system to further evolve and improve its Ab repertoire (as reviewed in Ref. 14). Somatic hypermutation requires activation-induced cytidine deaminase (AID)3 (15). AID is thought to deaminate cytosine bases on DNA (16, 17, 18) to produce uracil. The uracil base can be paired with adenine upon the next round of DNA synthesis, and then with one further round, that adenine will pair with thymine. Thus, DNA replication can parlay the activity of AID into a C to T mutation (14). Uracil-DNA glycosylase can excise the uracil base, and ensuing DNA synthesis opposite the apyridimic site can lead to mutations to G, A, or T (19). Mismatch repair mechanisms (20, 21, 22, 23) are believed to detect mismatched uracil bases (e.g., U-G) and excise stretches of ssDNA (24) near the site of cytidine deamination. The strand containing the uracil or even the strand opposite (21) the uracil might be excised, and error-prone polymerases (25, 26) repairing the section may introduce incorrectly paired bases, not only at the site of the original deaminated C but also other nearby bases. Mismatch repair coupled with error-prone polymerases thus can lead to more mismatches; when the DNA is next copied, these mismatches can lead to permanent mutations.
In all cases where uracil in the DNA leads to mutation, DNA synthesis is needed to finalize the mutation process. For this reason it is commonly believed that the rate of hypermutation increases with the rate of DNA replication, and thus also with that of cell division. We have created several reporter constructs and used them to quantify mutation rates. In this study, we investigate whether the number of cell generations increases the frequency of mutants.
Materials and Methods
Vector construction
Constructs were based on the Moloney murine leukemia virus retroviral vector contained in p102.21 (kindly provided by J. B. Lorens, Rigel Pharma, South San Francisco, CA). They contain the internal ribosome entry site (IRES) and puromycin resistance gene of pIRESpuro3 (BD Clontech). Mutations were introduced into enhanced GFP (EGFP) (see Fig. 2A) at the fluorescent chromaphore to create the reporter genes GFP-CT (EGFP C198G, T199C) and GFP-CG (EGFP C201T, G203C, C204T). GFP* contains the premature stop codon (TAG) at bases 319–321 and was used previously (27).
FIGURE 2. Measuring mutation rates with GFP-based reporters. A, Sequences of reporters GFP-CT, GFP-CG, and GFP* compared with GFP (EGFP of BD Clontech). Fluorescence was restored when the boxed nucleotides were mutated. Base locations are given for the starting and ending bases of the sequences. B, Reporters were expressed using retroviral vectors. Vector components: LTR, long terminal repeat; , retroviral packaging signal; GFP rep., GFP reporter gene; PuroR, puromycin resistance gene. C, Flow cytometry of GFP expression. The mean fluorescence of GFP-negative cells was set to 3.0. Cells with fluorescence >100 (bar labeled GFP+) were counted as GFP-positive.
Cell culture
Retroviral vectors were packaged using PhoenixEco cells (ATCC SD 3444; American Type Culture Collection). 18-81 cells were infected with the vectors and selected in the presence of 2.5 μg/ml puromycin for 7 days. To start experiments with zero fluorescent mutants, fluorescent cells were removed by FACS. After FACS sorting, cells were grown with 1.5 μg/ml puromycin and fluorescent cells were detected by flow cytometry. All 18-81 cultures were grown in RPMI 1640 medium with 10% FCS. Cells were split at the same time daily by different ratios to control the number of cell generations. The splitting ratios and volumes (milliliters) in terms of culture plus fresh media are: 1/2, 10 + 10; 2/5, 8 + 12; 1/3, 7 + 14; 2/7 6 + 21; 1/4, 5 + 15; 1/5, 4 + 16.
DNA sequencing
DNA from GFP-positive cells was isolated, and the GFP reporter genes were PCR-amplified using Pfu polymerase (Stratagene). The PCR products were incubated with Taq polymerase to add deoxyadenosine overhangs and then cloned into pCR2.1-TOPO (Invitrogen Life Technologies). The plasmids were then amplified in Escherichia coli and sequenced.
Quantitative RT-PCR
Expression of mouse AID and the GFP reporter constructs were measured using quantitative RT-PCR. RNA was extracted from cells and cDNA produced by standard application of reverse transcriptase and random hexamer primers. PCR was performed using the ABI PRISM 7700 machine. AID and reporter expression was normalized to the amount of expressed mouse -actin and hypoxanthine phosphoribosyltransferase (HPRT). PCR primers and probes used were AID: 5'-GAAAATTCTGTCCGGCTAACCA-3', 5'-TCGCAAGTCATCGACTTCGT-3', 5'-[6-FAM]-TCGGCGCATCCTTTTGCCCTT-[TAMRA]-3'; GFP reporters: 5'-CCACATGAAGCAGCACGACT-3', 5'-TGCGCTCCTGGACGTAGC-3', 5'-[6-FAM]-TTCAAGTCCGCCATGCCCGAA-[TAMRA]-3'; -actin: 5'-AGGTCATCACTATTGGCAACGA3', 5'-CACTTCATGATGGAATTGAATGTAGTT-3', 5'-[6-FAM]-TGCCACAGGATTCCATACCCAAGAAGG-[TAMRA]-3'; HPRT: 5'-TGGAAAGAATGTCTTGATTGTTGAA-3', 5'-AGCTTGCAACCTTAACCATTTTG-3', 5'-[6-FAM]-CAAACTTTGCTTTCCCTGGTTAAGCAGTACAGC-[TAMRA]-3'.
Results
Mutation rates were measured in 18-81 mouse cells. 18-81 is a cell line of B lineage that expresses endogenous AID and hypermutates its Ig V region (11). Like activated B cells, 18-81 divides quickly (as low as a 5 h doubling time) and also undergoes Ig class switching, generally a change from μ H chain to 2b. 18-81 was also ideal for this study because it is almost completely nonadherent and allowed cultures to be split precisely with reproducible growth conditions each day.
Because the cell line proliferates so rapidly, the cell population was able to reach the same stationary cell concentration each day, even when the cultures were split by different amounts daily. For this study, we found that the number of cell generations per day could be varied by the amount that the culture was split and replenished with fresh medium. For example, a culture split by 1/2 (1 part culture plus 1 part fresh medium), on average, undergoes one cell division by the next day. A culture split by 1/4 averages two cell generations per day. Staining with CFSE, a vital dye that enables tracking of cell division, and monitoring with flow cytometry confirmed that varying the ratio of cell splitting changed the average number of cell generations (Fig. 1). However, because the measurement of hypermutation rates in our experiments requires several weeks and because CFSE could only be detected in our cells during the first 4 days after staining, CFSE tracking could not be used concomitantly with our mutation rate measurement experiments.
FIGURE 1. Proliferation of cells determined by CFSE staining. 18-81 cells were spun down, stained with CFSE, washed, and resuspended into the original culture medium (supernatant) at the original cell concentration. After 1 day, the mean CFSE fluorescence per cell was measured by flow cytometry. Cell division led to loss of fluorescence; thus, cells having undergone more generations would have less fluorescence. Because cells split with a higher fraction of fresh media had less fluorescence, we conclude that the different cell splitting fractions led to different amounts of cell generations. We note though, that loss of fluorescence, in part, could also have occurred independent of cell generation (e.g., CFSE leakage).
The mutation reporter genes (Fig. 2A) were variants of GFP that become fluorescent when mutated. GFP-CT becomes fluorescent when base 199 mutates from C to T; GFP-CG becomes fluorescent when base 203 mutates from C to G; GFP* becomes fluorescent when the premature stop codon reverts to a codon assigned to an amino acid. The GFP reporter genes were expressed in 18-81 cells using a retroviral vector (Fig. 2B) and mutations that restored fluorescence were detected by flow cytometry (Fig. 2C). Sequencing the reporter genes from fluorescent cells verified that fluorescence was not restored by unexpected, proximal mutations. In all cases GFP-CT sequences mutated C to T (10/10) and GFP-CG sequences mutated C to G (9/9). Sequences of GFP* from fluorescent cells all contained mutations at base 321; these mutations were primarily G to C (11/12 sequences) but also G to T (1/12).
Accumulation of mutations was monitored continually over 32 days (Fig. 3A). Each measured time point consisted of data from 106 cells. Because many cells could be easily analyzed via flow cytometry, we were able to calculate mutation rates (Fig. 3B) more accurately than by mere DNA sequencing.
FIGURE 3. Rate of hypermutation at different rates of cell generation. A, Mutations causing reporters to become GFP-positive were monitored continuously. Reporters tracked accumulation of specific mutations; GFP-CT (C T), GFP-CG (CG), and stop codon reversion, GFP* (GFP rev.). The rate of cell generations (gen./day) was determined by the daily cell splitting ratio. B, Mutation rates as a function of cell generation rate. CT (), CG (?), stop codon reversion (). If mutation rate increased proportionally with that of cell generation, the mutation rate would have the slope of the dashed line.
We measured mutation frequencies from cultures ranging from 1.0 to 2.3 generations per day. From these mutation frequencies we calculated mutation rates. Under these culture conditions, cultures could be stably and continuously maintained. Cells appeared to be healthy and viable staining showed <1% dead cells. To our surprise, we found that the number of mutations did not correlate positively with the number of cell generations. This was true of C to T, C to G, and the TAG reversion mutations. In all cases, maximal mutation rates were observed when cells produced 1.3 generations per day and decreased as the number of cell generations per day was increased. We found that at one generation per day, the mutation frequency was low. Though the mutation rate of GFP* was approximately twice that of GFP-CT and GFP-CG, any similarities or differences in rates could be due to differences in sequence or position of the mutation site within the GFP reporter gene. Nonetheless, it was surprising that the C to T and C to G mutation rates were approximately the same over the range of cell growth conditions. Because C to T mutations presumably require only AID activity and DNA synthesis, and C to G mutations likely require AID plus additional downstream factors and DNA synthesis, this could indicate that AID is a rate-limiting factor while downstream factors (uracil-DNA-glycosylase, mismatch repair, error-prone polymerases) are not.
In our experiments we varied the media replenishing (i.e., splitting) ratio of the cultures, thereby controlling the number of cell generations per day. Clearly, more cell generations did not necessitate more mutations. When we measured the expression levels of AID (Fig. 4A) and the GFP reporter (Fig. 4B), we found that expression varied over our range of culture splitting conditions. AID expression was optimal when cultures were split by 2/5 and 1/3 (1.6 and 1.8 generations per day, respectively); reporter expression was highest when cultures were split by 1/2 and 1/3 (1.0 and 1.3 generations per day, respectively) and decreased if cells proliferated more.
FIGURE 4. AID and GFP expression at different rates of cell generation. Expression was determined by real-time quantitative RT-PCR and normalized to -actin () and HPRT (?) expression. Expression of (A) AID and (B) GFP reporter. C, Product of AID and GFP reporter expression levels.
Discussion
In addition to AID, hypermutation is thought to require transcription of the template DNA (6, 17, 28, 29). If levels of AID and reporter transcription both correlate positively with the mutation rate, and are independent of each other, then the product of both factors might also reflect the mutation rate. That is,
rate [AID]
rate [GFP]
rate [AID][GFP].
When the AID and reporter expression levels from our experiments are multiplied (Fig. 4C), the resulting curve does reflect the changes in mutation rate that occurred with varying the number of generations per day (Fig. 3B). By the terms of this model, in the cultures where cells divide on average only 1.3 times per day, high reporter transcription plus moderate levels of AID can result in a high mutation rate. Furthermore, high mutation rates can occur when there is high AID expression but less reporter transcription. The correlation of the mutation rate with AID and reporter expression suggests that AID and transcription are primary limiting factors in hypermutation. Although cell division is likely an integral downstream step in the mutation process, the fact that the correlation occurs independently of cell division (i.e., does not require cell generation number as a dependent variable) suggests that even if AID and GFP were held constant, varying the number of generations (under the conditions tested here) would not greatly affect the number of mutations. Thus the incidence of cell division is high enough that it is not a limiting factor for hypermutation under our laboratory growth conditions. In other words, under these conditions, cell division is not the "bottleneck" point in the hypermutation pathway. However, one might still predict that at some very low rate of cell division, the mutation rate would diminish. Our findings are in line with the observations of Bachl et al. (30) who noted that "variation in cell numbers seems to be much greater than the variation in the frequency of mutant cells. If these variations are accurate, they suggest that mutation is not highly coupled to cell division in these cells."
Our results do not imply that there cannot be other factors that affect the rate of hypermutation. For example, nuclear import and export (31, 32) and phosphorylation of AID (33) have recently been suggested to affect the rate of hypermutation. It has been suggested that other limiting factors exist specifically for mutation of Ig genes in B cells (31) and such factors may or may not play a role in our experimental system here.
The only variable that we sought to manipulate was the number of cell generations per day. However, the protocol also causes variation in the freshness of the growth medium (a highly diluted culture has more fresh medium), and time spent at cell growth saturation, which might influence phenotypic expression. The more diluted cultures spend less time at saturation before the GFP is measured. Could a variation in "phenotypic lag", i.e., a delay in gene expression, have influenced our results? A delay between the time when the mutation occurs and the time when a sufficient amount of GFP can be detected by flow cytometry could skew rate measurements over a period of time that is comparable in length to the phenotypic lag. However, our measurements of mutation frequencies were taken over a month, a period of time that is likely much longer than any lag that might occur between the incidence of mutation and expression of GFP fluorescence. Therefore, irrespective of any short lags in expression, the cultures will have reached a steady state, and the measured rates will not be affected by a phenotypic lag.
We found that fewer generations, and thus more time spent at saturation, gave higher frequencies of mutations. Accumulation of metabolites in the growth medium and time spent at saturation might be also mutagenic. Furthermore, variation of cell concentration or nutrients could affect hypermutation in ways unrelated to replication. But unless there are factors of a higher magnitude of importance than AID or DNA transcription, this would not invalidate our proposition that cell division is not a rate-limiting step for hypermutation.
In the study of mutations, whether Ig hypermutation or not, mutation rates are nearly always expressed on a per cell generation basis. This is because in most cases, mutations are introduced during DNA replication. However, in our study, we find that hypermutation is likely an important exception–cell division is not rate limiting. It follows then that rates should rather be reported on a per time (e.g., mutations/base pair/day) basis.
We have shown that increased cell generations not only do not necessitate increased mutation, but also that high cell proliferation can lead to decreased AID levels and mutation rates. Although our results were obtained using a cell line and may not accurately reflect an in vivo system, in the animal, it might make sense to decrease the mutation rate when cells are dividing quickly; once B cells with good Ab affinity are selected to undergo increased clonal proliferation, reducing AID levels could prevent further mutagenesis at the Ig loci or other locales throughout the genome. Although it has already been recognized that AID and transcription are important factors for hypermutation, we demonstrated here that AID activity and transcription alone are largely sufficient to explain the incidence of mutation. These activities, unlike cell division, are likely the dominant rate-limiting steps.
Acknowledgments
We are grateful to Frank Stahl for critically reading the manuscript; and for J. Lorens (Rigel Pharma, South San Francisco) for providing the retroviral vector; and to Desirée Yang, Freia Spillmann, Melissa Lodoen, Jessica Hamerman, and J. A. Bluestone for their efforts—cdub out.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This study was supported by National Institutes of Health (NIH) Grant AG20684. C.L.W. was supported by the NIH Immunology Training Grant (NIH/NIAID T32 AI07334).
2 Address correspondence and reprint requests to Department of Microbiology and Immunology, University of California, Box 0414, San Francisco, CA 94103-0414. E-mail address: cliffw@itsa.ucsf.edu
3 Abbreviations used in this paper: AID, activation-induced cytidine deaminase; IRES, internal ribosome entry site; EGFP, enhanced GFP; HPRT, hypoxanthine phosphoribosyltransferase.
Received for publication November 16, 2004. Accepted for publication February 8, 2005.
References
Luria, S. E., M. Delbrück. 1943. Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28:491.
Hashem, V. I., W. A. Rosche, R. R. Sinden. 2002. Genetic assays for measuring rates of (CAG): (CTG) repeat instability in Escherichia coli. Mutat. Res. 502:25
Wloch, D. M., K. Szafraniec, R. H. Borts, R. Korona. 2001. Direct estimate of the mutation rate and the distribution of fitness effects in the yeast Saccharomyces cerevisiae. Genetics 159:441.
Caetano-Anolles, G.. 1999. High genome-wide mutation rates in vegetatively propagated bermudagrass. Mol. Ecol. 8:1211.
Eckert, K. A., G. Yan, S. E. Hile. 2002. Mutation rate and specificity analysis of tetranucleotide microsatellite DNA alleles in somatic human cells. Mol. Carcinog. 34:140.
Bachl, J., C. Carlson, V. Gray-Schopfer, M. Dessing, C. Olsson. 2001. Increased transcription levels induce higher mutation rates in a hypermutating cell line. J. Immunol. 166:5051.
Yoshikawa, K., I. M. Okazaki, T. Eto, K. Kinoshita, M. Muramatsu, H. Nagaoka, T. Honjo. 2002. AID enzyme-induced hypermutation in an actively transcribed gene in fibroblasts. Science 296:2033.
Poltoratsky, V., C. J. Woo, B. Tippin, A. Martin, M. F. Goodman, M. D. Scharff. 2001. Expression of error-prone polymerases in BL2 cells activated for Ig somatic hypermutation. Proc. Natl. Acad. Sci. USA 98:7976.
Lin, M. M., M. Zhu, M. D. Scharff. 1997. Sequence dependent hypermutation of the immunoglobulin heavy chain in cultured B cells. Proc. Natl. Acad. Sci. USA 94:5284.
Wabl, M., H. M. Jack, J. Meyer, G. Beck-Engeser, R. C. von Borstel, C. M. Steinberg. 1987. Measurements of mutation rates in B lymphocytes. Immunol. Rev. 96:91
Wabl, M., P. D. Burrows, A. von Gabain, C. Steinberg. 1985. Hypermutation at the immunoglobulin heavy chain locus in a pre-B-cell line. Proc. Natl. Acad. Sci. USA 82:479.
Zhu, M., J. L. Rabinowitz, N. S. Green, B. J. Kobrin, M. D. Scharff. 1995. A well-differentiated B-cell line is permissive for somatic mutation of a transfected immunoglobulin heavy-chain gene. Proc. Natl. Acad. Sci. USA 92:2810.
Baumal, R., B. K. Birshtein, P. Coffino, M. D. Scharff. 1973. Mutations in immunoglobulin-producing mouse myeloma cells. Science 182:164.
Neuberger, M. S., R. S. Harris, J. Di Noia, S. K. Petersen-Mahrt. 2003. Immunity through DNA deamination. Trends Biochem. Sci. 28:305.
Muramatsu, M., K. Kinoshita, S. Fagarasan, S. Yamada, Y. Shinkai, T. Honjo. 2000. Class switch recombination and hypermutation require activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102:553.
Petersen-Mahrt, S. K., R. S. Harris, M. S. Neuberger. 2002. AID mutates E. coli suggesting a DNA deamination mechanism for antibody diversification. Nature 418:99.
Ramiro, A. R., P. Stavropoulos, M. Jankovic, M. C. Nussenzweig. 2003. Transcription enhances AID-mediated cytidine deamination by exposing single-stranded DNA on the nontemplate strand. Nat. Immunol. 4:452.
Pham, P., R. Bransteitter, J. Petruska, M. F. Goodman. 2003. Processive AID-catalysed cytosine deamination on single-stranded DNA simulates somatic hypermutation. Nature 424:103.
Di Noia, J., M. S. Neuberger. 2002. Altering the pathway of immunoglobulin hypermutation by inhibiting uracil-DNA glycosylase. Nature 419:43.
Rada, C., M. R. Ehrenstein, M. S. Neuberger, C. Milstein. 1998. Hot spot focusing of somatic hypermutation in MSH2-deficient mice suggests two stages of mutational targeting. Immunity 9:135.
Cascalho, M., J. Wong, C. Steinberg, M. Wabl. 1998. Mismatch repair co-opted by hypermutation. Science 279:1207.
Kim, N., G. Bozek, J. C. Lo, U. Storb. 1999. Different mismatch repair deficiencies all have the same effects on somatic hypermutation: intact primary mechanism accompanied by secondary modifications. J. Exp. Med. 190:21.
Martomo, S. A., W. W. Yang, P. J. Gearhart. 2004. A role for msh6 but not msh3 in somatic hypermutation and class switch recombination. J. Exp. Med. 200:61.
Bardwell, P. D., C. J. Woo, K. Wei, Z. Li, A. Martin, S. Z. Sack, T. Parris, W. Edelmann, M. D. Scharff. 2004. Altered somatic hypermutation and reduced class-switch recombination in exonuclease 1-mutant mice. Nat. Immunol. 5:224
Zeng, X., D. B. Winter, C. Kasmer, K. H. Kraemer, A. R. Lehmann, P. J. Gearhart. 2001. DNA polymerase is an A-T mutator in somatic hypermutation of immunoglobulin variable genes. Nat. Immunol. 2:537.
Faili, A., S. Aoufouchi, E. Flatter, Q. Gueranger, C. A. Reynaud, J. C. Weill. 2002. Induction of somatic hypermutation in immunoglobulin genes is dependent on DNA polymerase . Nature 419:944.
Wang, C. L., R. A. Harper, M. Wabl. 2004. Genome-wide somatic hypermutation. Proc. Natl. Acad. Sci. USA 101:7352.
Chaudhuri, J., M. Tian, C. Khuong, K. Chua, E. Pinaud, F. W. Alt. 2003. Transcription-targeted DNA deamination by the AID antibody diversification enzyme. Nature 422:726.
Peters, A., U. Storb. 1996. Somatic hypermutation of immunoglobulin genes is linked to transcription initiation. Immunity 4:57.
Bachl, J., M. Dessing, C. Olsson, R. C. von Borstel, C. Steinberg. 1999. An experimental solution for the Luria-Delbruck fluctuation problem in measuring hypermutation rates. Proc. Natl. Acad. Sci. USA 96:6847.
McBride, K. M., V. Barreto, A. R. Ramiro, P. Stavropoulos, M. C. Nussenzweig. 2004. Somatic hypermutation is limited by CRM1-dependent nuclear export of activation-induced deaminase. J. Exp. Med. 199:1235.
Ito, S., H. Nagaoka, R. Shinkura, N. Begum, M. Muramatsu, M. Nakata, T. Honjo. 2004. Activation-induced cytidine deaminase shuttles between nucleus and cytoplasm like apolipoprotein B mRNA editing catalytic polypeptide 1. Proc. Natl. Acad. Sci. USA 101:1975.
Chaudhuri, J., C. Khuong, F. W. Alt. 2004. Replication protein A interacts with AID to promote deamination of somatic hypermutation targets. Nature 430:992.(Clifford L. Wang and Matt)