Somatic Hypermutation in the Heavy Chain Locus Correlates with Transcription

Immunity, Vol. 9, 105–114, July, 1998, Copyright 1998 by Cell Press

Somatic Hypermutation in the Heavy Chain Locus Correlates with Transcription Yosho Fukita, Heinz Jacobs,*† and Klaus Rajewsky Institute for Genetics University of Cologne Weyertal 121 D-50931, Cologne Germany

Summary Three mutant immunoglobulin heavy chain (IgH) insertion mice were generated in which a targeted nonfunctional IgH passenger transgene was either devoid of promoter (pD) or was placed under the transcriptional control of either its own RNA polymerase II–dependent IgH promoter (pII) or a RNA polymerase I–dependent promoter (pI). While the transgene mutation-frequency (0.85%) in memory B cells of pI mice was reduced compared to that in pII mice (1.4%), the distribution and the base exchange pattern of point mutations were comparable. In pD mice, the mutation frequency was drastically reduced (0.09%). The mutation frequencies correlated with the levels of transgene-specific pre-mRNA expressed in germinal center B cells isolated from the mutant mice.

Introduction Rearranged immunoglobulin (Ig) genes can be further diversified by somatic hypermutation during T cell– dependent immune responses (see Rajewsky, 1996, for review). Somatic hypermutation occurs in germinal centers (GC) (Berek et al., 1991; Jacob et al., 1991), is characterized by an elevated frequency of mutations, and is restricted to the variable regions of Ig heavy and light chain genes (Kim et al., 1981; Gearhart and Bogenhagen, 1983; Lebecque and Gearhart, 1990; Weber et al., 1991). Analyses of Ig-transgenic mice have provided insights into the cis-acting DNA elements required to target somatic mutation to the V(D)J region (O’Brien et al., 1987). In the case of the k light chain gene, the k intronic and 39 enhancers but not the rearranged VkJk segment itself were found to be essential for hypermutation to occur (Betz et al., 1994; Yelamos et al., 1995). The V gene promoter can be exchanged by other RNA polymerase II (pol II)–dependent promoters without affecting hypermutation (Betz et al., 1994; Tumas-Brundage and Manser, 1997). A linkage between transcription initiation and the occurrence of somatic hypermutation has been suggested by Peters and Storb, who found that placing an additional Ig promoter immediately 59 to the constant portion of the Igk chain gene leads to the occurrence * To whom correspondence should be addressed (e-mail: Jacobs @BII.CH). † Present address: Basel Institute for Immunology, Grenzacherstrasse 487, CH-4005, Basel, Switzerland.

of somatic hypermutation within 2 kb downstream of each promoter but not between these two areas (Peters and Storb, 1996; Rada et al., 1997). In previous analyses of somatic hypermutation, conventional transgenic mice have often been used. An inherent problem of conventional transgenesis is the random integration of several copies of the transgene into the genome (Lacy et al., 1983). In contrast, gene targeting allows the targeted insertion of a single gene copy into any given locus. We have used the latter approach to study the dependence of somatic hypermutation in the IgH locus on transcription of the rearranged V HD HJ H element. For this purpose, we generated three IgH insertion mice that carry a rearranged V HD HJH gene segment that either lacked a promoter (pD mice) or was placed under the control of either the Ig heavy chain gene promoter (pII mice) or an RNA polymerase I (pol I)–dependent promoter (pI mice). The pol I–dependent promoter was chosen since it might allow one to distinguish between transcription per se and the coupling of a putative mutator to a pol II–specific component. The analysis of somatic hypermutation in these IgH insertion mice is the subject of this report. Results Generation of Mutant Mice Previously, a mutant mouse strain (designated B1-8i) was generated in which the JH locus and the DQ52 element were replaced by the VHDHJH segment of (4-hydroxy3-nitrophenyl) acetyl (NP)–binding antibody, B1-8 (Sonoda et al., 1997). The targeting vector used for the generation of this B1-8i mouse was modified to construct three different targeting vectors for the experiments presented here (Figures 1A–1H). The vectors share a termination codon in the leader exon of the rearranged B1-8 VHDHJH (VHB1-8) gene, rendering the inserted transgene inert to selection pressures imposed on productively rearranged heavy chain genes. In this way, information on the intrinsic specificity of the hypermutation process can be obtained (Sharpe et al., 1991; Betz et al., 1993). The first vector contained the physiological promoter of the VHB1-8 gene. The resulting mutant mice served as positive controls for the analysis of somatic hypermutation of the inserted V HB1-8 gene. The second vector was constructed such that the VH promoter upstream of the VHB1-8 gene was deleted. Analysis of this deletion mutant should reveal the relevance of the V H promoter for somatic hypermutation. In the third vector, the VH gene promoter was replaced by a bipartite 2.2 kb pol I promoter fragment (Kuhn et al., 1990). These targeting vectors were transfected individually into E14.1 ES cells (Ku¨hn et al., 1991). Homologous recombinants were identified by PCR screening and confirmed by Southern blot analysis. Recombinant ES cell clones were then injected into C57B/6 blastocysts to generate chimeras, which transmitted the mutant genes through the germline. Removal of the loxP-flanked neor gene in

Immunity 106

Figure 1. Targeted Insertion of the Modified V HB1-8 Gene into the Germline of the IgH Locus (A) Partial restriction endonuclease map of the wild-type IgH locus. The closed circle represents the IgH intronic enhancer, and closed boxes represent DQ52 and J H1-4 elements. B, BamHI; R, EcoRI; H, HindIII; P, PstI; probe A, HindIII–XhoI fragment; probe B, NaeI–EcoRI fragment; probe C, EcoRI–HindIII fragment. (B) Targeting cassette vector (Sonoda et al., 1997). This vector was used for the introduction of the modified B1-8 V HDHJ H (VHB1-8) gene mentioned below. The XhoI–EcoRI fragment containing the DQ52 and J H1-4 elements was replaced by VHB1-8 and the floxed neomycin gene. Closed triangles represent the loxP sites. (C, E, and G) Predicted structure of the targeted IgH locus of the VH promoter (pII) mice (C), the promoter deletion mice (E), and the pol I promoter (pI) mice (G) before deletion of the neomycin gene. Arrowheads indicate the position of PCR primers. The fragment sizes (in kilobases) detected by probes A, B, and C are indicated. (D, F, and H) Predicted structure of the targeted IgH locus of the pII mice (D), the pD mice (F), and the pI mice (H) after deletion of the neomycin gene. (I–K) Southern blot analysis of offspring of germline chimera. (I and K) HindIII- and (J) EcoRI-digested tail DNA from offspring of germline chimera before and after deletion of the neomycin gene was hybridized with probe A (I), probe B (J), and probe C (K). Lanes 1, wild-type 129 mouse; 2 and 3, the pII mice before and after deletion of the neomycin gene, respectively; 4 and 5, the pD mice before and after deletion of the neomycin gene, respectively; 6 and 7, the pI mice before and after deletion of the neomycin gene, respectively. Fragment sizes (in kilobases) are indicated.

the targeted locus was achieved by crossing the animals to the deleter strain (Figures 1I–1K) (Schwenk et al., 1995). Somatic Hypermutation of VHB1-8 Depends on the Presence of the Promoter Spleen cells of heterozygous mutant mice were stained as described in the Experimental Procedures. In brief,

the cells were depleted of IgM1 B cells, T cells, and macrophages using magnetic cell sorting (Miltenyi et al., 1990), and the remaining cells were subsequently stained with IgM-, IgD-, Thy1-, CD11b-, CD45R-, and Vl1-specific antibodies. Single memory B cells were sorted from the IgM2, IgD2, Thy12, CD11b2 (FITC-negative), CD45R1 (allophycocyanin, APC-positive), Vl11

Hypermutation and Transcription of Mutant Ig Genes 107

Table 1. Somatic Mutation Analysis of pD, pI, and pII Mice

Mutant Strain

PCR Products with Point Mutation (#)/ Total PCR Products Sequenced

Mutation Frequencya

Point mutations per sequence (#)

Actual Mutation Frequencyb

Normalized Mutation Frequency c

pII mice (V HB1-8) pD mice (VHB1-8)

12/20 (60%) 8/35 (22%)

0.84% (103/12200) 0.056% (12/21350)

0–32 0–3

100%

pD mice (Vl1) pI mice (V HB1–8)

20/35 (57%) 14/25 (56%)

0.60% (51/8400) 0.47% (73/15250)

0–7 0–17

1.40% (103/7320) 0.24% (12/4880) [0.092% (12/12200)]d 1.06% (51/4800) 0.85% (73/8540)

Naive B cells from Pol II mice

2/20 (10%)

0.030% (3/12200)

0–2

6.5% e 60%

Total numbr of point mutations Total number of base pairs sequenced Total number of point mutations b Actual mutation frequency 5 Total number of base pairs of PCR products containing point mutations c Actual mutation frequency of the pII mice was normalized as 100%, and actual mutation frequency of the pD mice and pI mice were compared. d This number was calculated based on the finding that 57% of memory B cells of the pD mice were under the influence of a putative somatic hypermutation machinery, but due to a lack of promoter activity the actual point mutation process failed to occur. e This number was calculated based on d. a

Mutation frequency 5

(phycoerythrin, PE-positive) fraction. The V HB1-8 gene was amplified from the genomic DNA of these single cells, using primers that specifically hybridize to the corresponding sequences of the inserted VHB1-8 gene of the targeted allele and not to the endogenous V HB1-8 gene. PCR products were purified from agarose gels and directly sequenced. Coamplification and sequencing of the functionally rearranged Vl1 gene (Ford et al., 1994; Gonzalez-Fernandez et al., 1994) served as an internal control for the mutator activity. At least two independent experiments were performed for each mutant strain. The PCR error frequency of our experimental setup was controlled by sequencing the VHB1-8 amplificates from 20 naive (IgM1, IgD1, CD45R1, and Vl11) B cells derived from pII mice. Two PCR products were found to have one and two point mutations, respectively. The Taq polymerase error frequency was 0.03% (3 mutations/12200 nt). As expected, somatic hypermutation of the V HB1-8 gene occurred normally in pII mice. About 60% of the sequenced samples contained multiple point mutations (Table 1), consistent with the finding that 64% of Vl1

genes amplified from single cells of the same B cell fraction in wild-type mice are mutated (Jacobs et al., 1998). The mutation frequency is 0.84% (103 mutations/ 12200 nt [610 nt 3 20 samples]) in the V HB1-8 passenger transgene of pII mice. Taking only the mutated VHB1-8 genes into account, the mutation frequency is 1.4% (103 mutations/7320 nt [610 nt 3 12 samples]). The number of point mutations found in the amplificates of memory B cells varied from 0 to 32 (Figure 2). When analyzing somatic hypermutation in memory B cells derived from pD mice, the percentage of mutated V HB1-8 genes was reduced to 22%, compared to 60% found in pII mice (Table 1). In addition, the maximum number of point mutations detected in a mutated gene was 3, as compared to 32 in pII mice. To verify that we had indeed isolated memory B cells, we coamplified and sequenced the Vl1 genes from the same single cells. 57% of the Vl1 genes were found to be mutated, a value comparable to the percentage of mutated Vl1 genes in memory B cells of wild-type mice and mutated V HB1-8 genes in memory B cells of pII mice (Table 1). Therefore, although only 22% of the analyzed VHB1-8

Figure 2. The Percentage of V HB1-8 Genes with a Given Number of Point Mutations from pD, pI, and pII Mice Total number of mutations found were 12 in the pD mice, 73 in the pI mice, and 103 in the pII mice.

Immunity 108

Table 2. Base Exchange Pattern of Hypermutation pI

A G C T

A ——

——

G

to C

T

11 14 ——7 —— 10 15 4 —— 3 9 —— 14 — —— 5 6 5

%

n 5 72

31 28 25 16

A G C T

from

from

n 5 103

pD A ——

——

G

to C

T

0 6 ——13 —— 6 14 4 —— 4 4 —— 8 —— 3 2 8 —

genes were mutated, the analysis of the Vl1 genes from the same single cells indicated that hypermutation actually occurred in 57% of the samples. Taking this finding into account, the mutation frequency of the pD VHB1-8 gene further decreases from 0.24% (based on 22% mutated VH genes) to 0.09% (based on 57% mutated Vl1 genes). Thus, the frequency of V HB1-8 mutations in memory B cells of pD mice is 15-fold lower than in memory B cells of pII mice. In contrast, the mutation frequency of 1.1% at the l locus is consistent with a previous analysis (Jacobs et al., 1998), indicating that the hypermutator as such is unaffected in memory B cells of pD mice. As so far only 12 point mutations were found in the pD mice, the relevance of the frequent C nucleotide changes (7/12) remains unclear (Table 2). These results demonstrate the importance of the VH promoter in the process of somatic hypermutation. Can the Pol I Promoter Substitute the Pol II Promoter in the Control of Somatic Hypermutation? To answer this question, somatic hypermutation was analyzed in single memory B cells isolated from pI mice. Similar to the percentage of mutated VHB1-8 genes in pII mice and mutated Vl1 chain genes in pD mice, 56% of the amplificates were mutated (Table 1). However, the range of mutations per gene (0–17) was below that of pII mice (0–32). Similarly, the mutation frequency in memory B cells of pI mice was reduced to 0.84%, i.e., 60% of the level found in pII mice (Table I). As indicated in Figure 3, the pol I promoter did not have any significant impact on the distribution of point mutations compared to pol II promoter-controlled hypermutation. Identical hot spots were found in VHB1-8 genes of pII

%

n 5 12

27 33 22 18

A G C T

from

pII

A ——

——

1 1 1

to C

G

T

%

1 —— — 0 0

1 1 25 0 8 ——0 —— 6 59 —— —— 8 0

and pI mice, notably at position 31II, 65II, and 105II of the VHB1-8 gene (numbering according to Kabat and Wu, 1991). These codons are typical hot spots with a RGYW consensus motif (where R is a purine base [A or G], Y a pyrimidine base [C or T], and W is A or T). This motif is a salient feature of major hot spots (Rogozin and Kolchanov, 1992). In line with previous results (Betz et al., 1993), these hot spots reflect intrinsic properties of the molecular mechanism of somatic hypermutation. As for the base exchange pattern, A and G nucleotides change to another nucleotide more frequently than C and T in both the pII mice and the pI mice (Table 2). These base substitution preferences are comparable to those seen in other passenger genes (Betz et al., 1993). Transcription of Mutant VHB1-8 Genes in GC B Cells of pI, pII, and pD Mice To explore a possible relation between somatic hypermutation and transcriptional activity at the IgH locus, we determined the pre-mRNA and mRNA levels of the mutant B1-8 transgenes in GC-derived B cells by comparative RT-PCR (Klein et al., 1997). Using random hexamers, cDNA libraries were synthesized from total RNA isolated from 1 3 105 CD45R1 PNAhigh B cells from each of the mutant mice. Two sets of oligonucleotides were designed such that they specifically primed the mutant VHB1-8 and the b-actin cDNAs, respectively. In both cases, the conditions for the amplification were chosen such that the accumulation of products was in the exponential phase. PCR products were gel-fractionated, blotted onto a nylon-membrane, and hybridized with a32 P-labeled V HB1-8 or b-actin probes. The 322 bp amplificates of the b-actin cDNA served as an internal standard to normalize the expression level of the VHB1-8 Figure 3. Location of Somatic Hypermutation in the Modified VHB1-8 Gene in pD, pI, and pII Mice The point mutations found in the VHB1-8 gene of memory B cells are given as percentage of mutations per codons sequenced and are plotted against the position of the VHB1-8 gene (numbering according to Kabat and Wu, 1991; point mutations in the leader intron and J H2 intron are also given). Percentages were calculated from eight mutated sequences of pD mice, 14 of pI mice, and 12 of pII mice.

Hypermutation and Transcription of Mutant Ig Genes 109

Table 3. Relative V HB1-8 mRNA Levels in PNAhigh B Cells

Pre-mRNA mRNA

pII Mice

pD Mice

pI Mice

100% 100%

8% 23%

72% 87%

All the data were first normalized to b-actin expression. For both pre-mRNA and mRNA, the amount of PCR products of the pD mice and the pI mice was normalized against the PCR products of the pII mice.

1993; Lozano et al., 1994; Peltz et al., 1994; Maquat, 1995; Aoufouchi et al., 1996).

Figure 4. Comparative PCR of cDNAs Generated from Splenic B2201PNAhigh B Cells from pD, pI, and pII Mice (A) Schematic figure illustrating the PCR approach to amplify the cDNA species from the mRNA and pre-mRNA of the V HB1-8 genes. Boxes indicate the leader exon and V HDHJH exon of the inserted VHB1-8 gene. Arrows indicate the location of primers used for the amplification of V HB1-8 cDNA. The thick lines labeled with 265 bp and 183 bp represent the coding sequence of the pre-mRNA and mRNA, respectively. The thin line indicates the leader intron. (B) Southern blot analysis of amplified cDNA corresponding to VHB1-8 genes from the mutant mice. Experiments were performed in duplicate. The upper band corresponds to pre-mRNA of VHB1-8 genes and the lower band to mRNA of VHB1-8 genes. The sizes of the bands are indicated. (C) Southern blot analysis of cDNA corresponding to the b-actin gene transcripts from the mutant mice. In (B) and (C), 5,000 cell equivalents of cDNA mixture were amplified in separate reaction with V HB1-8 and b-actin primers.

product. For the VHB1-8 cDNA, two PCR products of different size were observed in all three mutant strains (Figure 4A). As confirmed by sequencing, the 265 bp fragment corresponded to the pre-mRNA of VHB1-8 and the 183 bp fragment to the mRNA, lacking the 82 bp leader intron of VHB1-8 (Figure 4B). The ratio between the 265 and 183 bp fragments was similar in all samples. This argues against possible contamination of the samples by genomic DNA, as this would have favored the amplification of the 265 bp fragment only. Normalization of the intensity of each band of the pD mice and the pI mice against the corresponding bands of pII mice revealed that the level of pre-mRNA was 8% and 72% and the level of mRNA 23% and 87%, respectively (Table 3). As splicing of pre-mRNA is usually an efficient process and occurs rapidly after transcription (Zhang et al., 1994), the relatively high ratio of premRNA versus mRNA levels in this study is surprising but may be due to the premature termination codon inserted into the V HB1-8 transgene (Cheng and Maquat,

The Level of Mutant VHB1-8-Specific Pre-mRNA Correlates with the Frequency of Somatic Hypermutation To compare the pre-mRNA levels with the frequency of somatic hypermutation, we normalized the values obtained from pII mice to 100%. The pre-mRNA rather than the mRNA level was chosen, since the pre-mRNA level should relate more directly to the transcription rate (Zhang et al., 1994). The resulting relative mutation frequencies of the pD mice and the pI mice were 7% and 60%, respectively (Table 1). These figures correlate roughly with the relative amounts of pre-mRNA: 8% and 72% in the pD mice and pI mice, respectively (Table 3). VHB1-8 Transcripts in pI Mice are Polyadenylated In contrast to RNA pol II transcripts that are synthesized in the nucleoplasm, pol I transcripts of the ribosomal genes are synthesized in the nucleolus and are neither capped nor polyadenylated (Smale and Tjian, 1985). In the comparative RT-PCR experiment mentioned above, a random hexamer-primed cDNA was used as a template. To analyze whether VHB1-8 transcripts from the pI mice are polyadenylated, the same comparative RTPCR was performed using oligo(dT)-primed cDNA as template. The level of V HB1-8 transcripts in this preparation was comparable to that in the random hexamerprimed cDNA, indicating that the B1-8 transcripts from the pI mice are polyadenylated (data not shown). Discussion Gene Targeting Approach for the Analysis of Somatic Hypermutation To analyze the role of Ig transcription on somatic hypermutation of rearranged Ig genes, three different mutant IgH insertion strains were generated. The gene targeting approach ensures a site-specific insertion of a single copy into the targeted allele and excludes random integration of several copies into the genome. The latter are inherent problems of conventional transgenesis into oocytes (Lacy et al., 1983). The cis-acting elements that control somatic hypermutation at the IgH locus are not completely defined. However, somatic hypermutation occurs normally in a rearranged VHDHJH gene when targeted into its physiological position in the heavy chain locus (Taki et al., 1993). Consistent with these results, somatic hypermutation of the B1-8 passenger transgene in the IgH locus

Immunity 110

occurred normally with respect to frequency and distribution of point mutations, and base exchange pattern. About 60% of the IgH passenger transgenes amplified from single class switched Vl11 memory B cells were mutated, a value that is reproducibly found in Vl1 genes amplified from class switched Vl11 B cells of wild-type and other mutant mouse strains (Jacobs et al., 1998). The possibility that the unmutated amplificates are due to a contamination with naive Igm1l11 B cells is unlikely, since the sorted cells expressed high levels of the Vl1 light chain at their surface and stained negative for IgMand IgD-specific MAb. Therefore, z40% of class switched Vl11 B cells likely carry unmutated VlJl or VHDHJH gene segment. Compared to the pII mice, the mutation frequency was decreased 15-fold in the promoter-deficient passenger transgene amplified from class switched Vl11 B cells isolated from pD mice (Table 1). However, this lower mutation frequency was clearly above the Taq polymerase error, as determined for our experimental setup. In addition, the point mutations found are not compatible with the most frequent errors introduced by the Taq polymerase, i.e., T to C or A to G exchanges (Tindall and Kunkel, 1988; Keohavong and Thilly, 1989). The 15fold decrease in the mutation frequency of the VHB1-8 gene in pD mice compared to pII mice is strikingly similar to the 12.5-fold decreased frequency of somatic hypermutation in DHJH rearrangements (0.2%) in comparison to functional VHDHJH joints in hybridoma cells (2.5%) (Roes et al., 1989). Although the VH promoter of the V HB1-8 gene was deleted in the pD mice, a noticeable level of passenger transgene-specific transcripts was found in the PNAhigh B cells sorted from these mice. In all targeting experiments, the downstream portion of the DQ52 promoter (Alessandrini and Desiderio, 1991; Kottmann et al., 1992, 1994), the DQ52 element, and the JH locus were replaced by the modified VHB1-8 gene. Early in B cell development, the DQ52 promoter appears to be necessary to generate full accessibility for the V(D)J recombination machinery (Thompson et al., 1995). According to in vitro studies on the DQ52 promoter (Kottmann et al., 1994), the truncated form of the DQ52 promoter as present in the targeted alleles is transcriptionally inactive. The next nearest identified D segment is Dsp2, which locates 17 kb upstream of the DQ52 element (Wood and Tonegawa, 1983). It is not known whether the Dsp2 element possesses a promoter, or other promoter elements are present between Dsp2 and DQ52. Since somatic hypermutation occurs preferentially in a 2 kb window downstream of a defined promoter (Lebecque and Gearhart, 1990; Weber et al., 1991; Rothenfluh et al., 1993; Rogerson, 1994; Peters and Storb, 1996), we speculate that the remaining truncated DQ52 promoter fragment may preserve a low transcriptional activity in vivo, which would imply that the rules for transcriptional regulation at the DQ52 promoter as determined in vitro do not apply to GC B cells analyzed ex vivo. The RNA Polymerase I Promoter Allows Somatic Hypermutation When analyzing somatic hypermutation of the pol I promoter-controlled V HB1-8 passenger gene, the mutation

frequency was 60% of that found in memory B cells of pII mice. To date, promoter exchange experiments have analyzed somatic hypermutation of Ig genes under the control of the human b globin promoter (Betz et al., 1994) and the mouse B29 promoter (Tumas-Brundage and Manser, 1997). Both promoters are pol II–dependent, the enzyme known to control the synthesis of all precursor messenger RNA (pre-mRNA) in mammalian cells (Zawel and Reinberg, 1992; Zawel et al., 1995; Orphanides et al., 1996). In contrast, the pol I promoter controls the synthesis of the ribosomal RNA precursor in the nucleolus, which subsequently is processed into the 28S, 18S, and 5S ribosomal RNA (rRNA) species (Reeder, 1990). In analogy to pol II promoters, the pol I promoter has a bipartite organization, an upstream control element (UCE), and a start site proximal core (Grummt and Skinner, 1985; Smale and Tjian, 1985). The UCE contains 13 repeats of a 140-bp element, which enhances the activity of the pol I promoter (Kuhn et al., 1990). Exchanging UCE by pol II transactivator sequences such as GAL4 DNA binding sites, E2 protein recognition sequences, or an octamer motif, did not allow transcription from the core pol I promoter (Schreck et al., 1989), suggesting that the transactivation mechanism of pol I promotercontrolled genes is different from that of pol II promotercontrolled genes. However, when placing both elements of the bipartite pol I promoter upstream of the V HB1-8 gene, a high level of polyadenylated unspliced premRNA and spliced mRNA was found in GC B cells. These observations suggest that either the pol II transcription machinery can transcribe a pol I promoter-controlled heavy chain gene in the nucleoplasm, or the primary transcripts are indeed synthesized by the pol I transcription machinery, and the altered subnuclear localization of the polymerase I transcripts (nucleoplasm instead of nucleolus) allows the described posttranscriptional modifications. Since we cannot exclude either possibility, the question whether the putative mutator is physically associated with the RNA polymerase II complex, or the level of transcription just reflects the accessibility of the putative mutator to the IgH locus, remains unanswered.

Somatic Hypermutation Correlates with Transcription In the previous promoter exchange (Betz et al., 1994; Tumas-Brundage and Manser, 1997) and promoter duplication experiments (Peters and Storb, 1996), a slight reduction of the mutation frequency was observed with heterologous promoters. In these reports, the quantitation of specific mRNA was determined in hybridomas, but it is unknown whether mRNA levels of Ig genes in these cells correlate with mRNA levels in GC B cells. Furthermore, it is difficult to correlate transgene-specific mRNA and somatic hypermutation, because transgenes tend to integrate at random in a concatemeric fashion into the genome, and often the RNA levels do not correlate with the copy number of the transgene. The latter problem not applying to the gene targeting approach taken in this study, we compared the relative levels of RNA expressed from the mutant VHB1-8 genes in GC B cells, taken ex vivo, using semiquantitative RTPCR. In contrast to non-GC-derived B cells, B cells of

Hypermutation and Transcription of Mutant Ig Genes 111

the GC have more binding sites for the lectin peanut agglutinin (PNA). Thus, GC B cells can be distinguished and sorted from other B cells using fluorescein-conjugated PNA. In these cells, somatic hypermutation is known to be ongoing (Berek et al., 1991; Jacob et al., 1991). The relative frequency of somatic hypermutation correlated well with the relative pre-mRNA level of the passenger transgenes found in CD45R1 PNAhigh B cells isolated from the mutant strains. Since nonsense codons can reduce the abundance of nuclear mRNA without affecting the abundance of pre-mRNA or the halflife of cytoplasmic mRNA (Cheng and Maquat, 1993; Peltz et al., 1994; Maquat, 1995), the relatively high VHB1-8 pre-mRNA levels may be due to the presence of a termination codon in the leader exons in the various VHB1-8 transgenes, resulting in a rapid degradation of the corresponding mRNA species. Alternatively, a nonsense codon was found to inhibit RNA splicing in a B cell hybridoma (Lozano et al., 1994; Aoufouchi et al., 1996). The mechanism for this translation independent mechanism remains unclear. According to previous findings, the 59 boundary for somatic hypermutation is located at or near the transcription start site, suggesting a linkage between transcription and somatic hypermutation (Lebecque and Gearhart, 1990; Weber et al., 1991; Rothenfluh et al., 1993; Rogerson, 1994; Peters and Storb, 1996). Furthermore, in case of the k light chain gene, the k intronic and 39 enhancers are critical cis-elements for hypermutation (Betz et al., 1994; Yelamos et al., 1995). The results presented in this report indicate that somatic hypermutation correlates with the transcriptional activity in the IgH locus. This result is expected, if the transcription machinery is physically associated with (part of) the hypermutator (Peters and Storb, 1996; Goyenechea et al., 1997). We had hoped to support this model by the demonstration that hypermutation is only seen in the case of pol II– and not pol I–dependent transcription. Its occurrence in both cases, and the ambiguity of that result makes us unable to exclude the possibility that the apparent correlation of somatic hypermutation with the pre-mRNA levels might simply reflect the accessibility of the putative mutator to the IgH locus. Experimental Procedures Construction of the Targeting Vectors The targeting vector pIVhL2neor was constructed as described (Sonoda et al., 1997). This vector was used as a targeting cassette vector to introduce the modified B1-8 V HDHJ H (VHB1-8) gene. A VHB1-8 gene with its V H promoter was cloned into the HincII site of pBluescriptII vector (Stratagene), and the V H internal heptamer was modified as described (Sonoda et al., 1997). To make this VHB1-8 gene a passenger transgene, TGT/ATC at codon 215 and 214 (numbering according to Kabat and Wu, 1991) was modified to TGA/CTC through site-directed mutagenesis. The plasmid DNA pUC-SP/SalI, which contains mouse rDNA promoter and enhancer of pMrE-SP (Kuhn et al., 1990), was kindly provided by Dr. Ingrid Grummt (German Cancer Research Center, Heidelberg, Germany). A 2.2 kb HindIII–EcoRI fragment of this vector was subcloned into HindIII–EcoRI sites of pBluescriptII vector (Stratagene), and then a 2.2 kb XhoI–BamHI fragment was subcloned into a XhoI–BamHI site of pSL1190 vector (Pharmacia). A 1.3 kb XhoI–NcoI VH promoter fragment upstream of the V HB1-8 gene was replaced by a 2.2 kb XhoI–NcoI fragment of the rDNA promoter from the above pSL1190 vector. Subsequently, a 2.9 kb XhoI–ClaI

fragment containing the rDNA promoter-controlled VHB1-8 gene was subcloned into the SalI–ClaI sites of pIVhL2neor to generate the pol I promoter (pI) construct. To generate the promoter deletion (pD) construct, a 0.7 kb NcoI– EcoRI fragment of the VHB1-8 gene was subcloned into a NcoI– EcoRI site of pSL1190 vector (Pharmacia), released as a 0.7 kb SalI–ClaI fragment, and subcloned further into the SalI–ClaI sites of pIVhL2neor, which generates the pD construct. One loxP site was inserted into the BalI–ClaI site of the J H2 intron of the V HB1-8 gene, and then a 2.0 kb XhoI–ClaI fragment was subcloned into the SalI–ClaI sites of pIVhL2neor to generate the VH promoter (pII) construct. Generation of the Targeted ES Cells NotI-linearized targeting vectors (30 mg) of each construct were transfected by electroporation into 1 3 107 E14.1 ES cells, respectively (Ku¨hn et al., 1991). The transfected cells were selected with G418 (350 mg/ml) and gancyclovir (2 mM). Double resistant colonies were screened by PCR using a 59 primer (59-ACGATTACTACGGTAG TAGCTAC-39) located at the DFL16.2 element in V HB1-8 and a 39 primer (59-GGAAACTAGAACTACTCAAGCTA-39) located 61 bp 39 of the EcoRI site and 59 of the Em enhancer (Taki et al., 1993). PCR amplification was performed for 40 cycles; each cycle consisted of 1 min at 958C, 1 min at 618C, and 2 min at 728C. Homologous recombinant transfectants, positive for a 1.3 kb amplified fragment, were further verified by Southern blot analysis. Genomic DNA from PCR-positive colonies was EcoRI digested and hybridized with probe A (Figure 1A). Homologous recombinants were identified by a 6.5 kb band corresponding to the wild-type allele and an additional 1.9 kb band in the case of the pol I promoter (pI) targeted allele, or an additional 1.8 kb band in the case of the promoter deletion targeted allele, or an additional 1.2 kb band in the case of the VH promoter (pII) targeted allele. Out of 96 G418 and gancyclovir double resistant clones of each targeted ES cell, two carried the targeted allele, in the case of the pI construct and the pD construct, and one carried the targeted allele, in the case of the pol II construct. Generation of the Mutant Mice For each construct, recombinant ES cell clones were injected into C57BL/6-derived blastocysts and transplanted into uteri of F1 (BALB/c X C57BL/6) foster mothers. Chimeric mice were mated to C57BL/6 or the deleter strain (Schwenk et al., 1995) of C57BL/6 background. Offspring generated through germline transmission of the ES cell genome, as indicated by coat color, were analyzed by PCR on tail DNA, prepared as described by Laird et al. (1991) in combination with the same primer set used for the screening of homologous recombinant ES cells. Mice carrying the targeted allele were confirmed by Southern blot analysis. Nonimmunized neor gene-deleted heterozygous mice were analyzed at the age of 8–12 weeks for the single-cell sorting and subsequent analysis of somatic hypermutation. For the isolation of RNA, each mutant mouse was immunized with 100 mg NP-CG alumn in 200 ml PBS 10 days before analysis. Mice are kept under ordinary laboratory conditions. Cell Separation and Flow Cytometry Single-cell suspensions from spleen were prepared by gently teasing the cells from the splenic capsule in medium (Dulbecco’s modified Eagle’s medium [DMEM] supplemented with 5% fetal calf serum). They were treated with Tris-buffered 0.155 M NH 4Cl for 5 min to remove erythrocytes and washed by centrifugation through phosphate-buffered saline containing 1% bovine serum albumin and 0.1% NaN3 (PBS, BSA, and NaN3). The splenic cells were stained with a combination of rat anti-mouse IgM anti-CD5 (53-7.3; Ledbetter and Herzenberg, 1979), and anti-Mac1 (M1/70.15.11; Springer et al., 1979) antibodies coupled to magnetic beads (Miltenyi Biotech) to deplete the IgM 1 cells, T cells, and macrophage fractions by using the MACS system (Miltenyi et al., 1990). The cells passing through the column in the magnetic field were collected and further stained by a combination of fluorescein isothiocyanate (FITC)-conjugated R33-24-12 (anti-m; Gru¨ tzmann, 1981), FITC-conjugated 1.3-5 (anti-d; Roes et al., 1995), FITC-conjugated G7 (anti-Thy1; Gunter et al., 1984), FITC-conjugated M1/70.15.11 (anti-Mac1; Springer et al., 1979), phycoerythrin (PE)-conjugated LS136 (anti-l; Reth et

Immunity 112

al., 1979), and allophycocyanin (APC)-conjugated RA3-6B2 (antiCD45R/B220; Coffman, 1982) in staining medium. By using a dual laser flow cytometer (FACStar Plus, Becton-Dickinson), FITC-negative, PE-positive, and APC-positive single cells were directly sorted into 0.5 ml microtubes containing 20 ml PCR buffer (GIBCO-BRL, with 2.5 mM MgCl2) supplemented with 1 mg/ml 5S-ribosomal RNA from Escherichia coli. Cells present in the lymphocyte gate (Fo¨ rster et al., 1989) were analyzed, while the dead cells were excluded from analysis by staining with propidium iodide. For the isolation of RNA from the splenic cells, they were stained with RA3-6B2 (anti-CD45R/B220; Coffman, 1982) coupled to magnetic beads (Miltenyi Biotech), followed by magnetic cell sorting as described above. The cells attached to the column in the magnetic field were collected and further stained by a combination of FITCconjugated PNA (Rose et al., 1980) and PE-conjugated RA3-6B2 (anti-CD45R/B220; Coffman, 1982). By using a dual laser flow cytometer (FACStar Plus, Becton-Dickinson), 1 3 105 cells of B2201 PNAhigh fraction from each mutant strain were collected in 1.5-ml Eppendorf tubes. At least two tubes were prepared from each mutant mouse. Single-Cell PCR To prepare DNA for amplification, microtubes with single-cells were overlaid with mineral oil, 1 ml proteinase K (10 mg/ml; Boehringer) was added, and the samples were digested for 1 hr at 558C. The proteinase K was subsequently inactivated for 10 min at 958C. PCR amplification was carried out in two rounds. The first reaction contained two sets of primers: for the amplification of VHB1-8, 59 leader exon 1 primer [59-CACCATGGGATGGAGCTGAC-39] and 39 JC intron 1 primer [59-CTCTGAAATCGATACCGTCG-39] were used for the pD and the pI mice, and the same 59 leader exon primer and 39 JC intron 2 primer [59-TATGATCGGAATTCCTGCAG-39] were used for the pII mice. These primers hybridize specifically to the corresponding sequences in the inserted VHB1-8 of the targeted allele. The 59 primers recognize the termination codon in the leader exon of the V HB1-8 gene, and the 39 primers recognize the sequence of plasmid DNA that is integrated into the downstream of the JH2 intron fragment of the targeting vector during the subcloning procedures. For the amplification of the Vl1 gene, Vl external primer [59-GGGTATGC AACAATGCGCATCTTGTC-39] and Jl1 primer [59-CACGGACAGGAT CCTAGGACAGTCAGTTTGGT-39] were used. Amplifications were done at the initial cycle of 958C for 2 min, 758C for 5 min, and 728C for 1.5 min, followed by 28 cycles (1 min at 948C, 1 min at 608C, and 1.5 min at 728C). The Taq polymerase was added after the first denaturation step. The second-round PCR amplification was performed in separate reactions for the V HB1-8 and Vl1 genes, using 1.5 ml of the first-round reaction mixture in a 50 ml volume. For the amplification of the V HB1-8 gene, the nested 59 leader exon 2 primer [59-GATGGAGCTGACTCATCCTCTTCTTGG39] and 39 JC intron 1 primer were used for the pD and the pI mice, and the nested 59 leader exon 2 primer and 39 JC intron 3 primer [59-CGGAATTCCTGCAGCCATGA-39] were used for the pII mice. For the amplification of the Vl1 gene, the nested Vl internal primer [59GCGAAGAGAAGCTTGTGACTCAGGAATCTGCA-39] and Jl1 primer were used. PCR consists of 30 cycles (1 min at 948C, 1 min at 638C, and 1.5 min at 728C). The amount of primers used were 2 pmol each for the first round PCR and 7 pmol for the second-round PCR. All PCRs contained dATP, dCTP, dTTP, and dGTP (Pharmacia) at 200 mM each, and 5 U of Taq DNA polymerase and PCR buffer containing 2.5 mM MgCl2 (GIBCO-BRL). The final volume was 50 ml. Each PCR was followed by a 10 min incubation at 728C. 10 ml of the second-round PCR product were analyzed on 1% agarose gels stained with ethidium bromide. For sequencing, bands were cut from preparative 1% TAE agarose gels (Seakem GTG agarose, Biozym), and the DNA was isolated using Spin-X columns (Costar). Cycle sequencing was performed using the Ready Reaction DyeDeoxy Terminator Cycle sequencing kit (Applied Biosystems) following the manufacturer’s instructions. Sequencing primers for the VHB1-8 and Vl1 genes were the same as the primers used for the second-round PCR amplification. Sequencing was done using an ABI 373A or 377A automatic sequencer (Applied Biosystems). RNA Isolation and cDNA Synthesis Cellular RNA was isolated from 1 3 10 5 cells of B2201 PNAhigh fraction from each mutant strain, as described (Chomczynski and Sacchi,

1987). Total cellular RNA was hybridized to random hexamer or oligo(dT) primer (Boehringer Mannheim). First-strand cDNA synthesis was performed using SuperScript II Reverse Transcriptase (GIBCO-BRL). Two cDNA mixtures were prepared from each mutant strain. Comparative RT-PCR Comparative RT-PCR of cDNA mixtures was done as described (Klein et al., 1997) with brief modifications. One-twentieth of each first-strand cDNA mixture, corresponding to about 5000 cell equivalent, was amplified using the 59 leader exon 1 primer and the VHB1-8 exon primer [59-GCTTCAACCCAGTGCATCCA-39]. In separate PCR reactions, one-twentieth of a cDNA mixture was amplified using 59 b-actin primer [59-GCACCACAACCTTCTACAATGAGCTG-39] and 39 b-actin primer [59-CACGCTCGGTCAGGATCTTCATGAG-39]. PCR consists of 20 cycles (1 min at 958C, 1 min at 638C, and 1 min at 728C) for the amplification of V HB1-8 cDNA, and 18 cycles for the amplification of b-actin cDNA with the same PCR condition. One-fifth of each PCR product was loaded onto a 2% agarose gel. After gel-electrophoresis, PCR products were transferred to a Nylon membrane according to standard procedures (Sambrook et al., 1989). Membranes were hybridized with a 32P-labeled PCR products. For the generation of DNA probe, one-twentieth of first-strand cDNA from the pII mice was amplified with the 59 leader exon 1 primer/VHB1-8 exon 2 primer or the 59 b-actin primer/39 b-actin-2 primer for 30 cycles under the conditions described above. PCR products were cut from preparative 2% agarose gels, and the DNA was isolated using Spin-X columns (Costar). DNA was labeled with a32P dATP using random primer according to standard procedure (Sambrook et al., 1989). After hybridization with the labeled PCR products, final washing of the corresponding membrane was performed with 0.1 3 SSC/0.1% sodium dodecyl sulfate at 608C. Densitometric analysis of membranes for quantitation of band intensities was performed using a phosphoimager (Fuji, Tokyo, Japan). Acknowledgments The authors are grateful to A. Egert, R. Zoebelein, and C. Go¨ttlinger for their excellent technical help, G. Esposito, L. Pao, and C. Schmedt for comments and critical reading, I. Grummt for providing the mouse rDNA promoter/enhancer fragment, and W. Mu¨ ller for providing the DNAPLOT program. DNAPLOT is part of IMGT (BioB104-CT96-0037). This work was supported by the Deutsche Forschungsgemeinschaft through SFB 243, the Land NordrheinWestfalen, and the European Community (B104-CT96-0077). Y. F. and H. J. were supported by postdoctoral fellowships of the Alexander von Humboldt Foundation and the European Molecular Biology Organization, respectively. Received April 10, 1998; revised June 15, 1998. References Alessandrini, A., and Desiderio, S.V. (1991). Coordination of immunoglobulin DJH transcription and D-to-JH rearrangement by promoterenhancer approximation. Mol. Cell. Biol. 11, 2096–2107. Aoufouchi, S., Yelamos, J., and Milstein, C. (1996). Nonsense mutations inhibit RNA splicing in a cell-free system: recognition of mutant codon is independent of protein synthesis. Cell 85, 415–422. Berek, C., Berger, A., and Apel, M. (1991). Maturation of the immune response in germinal centers. Cell 67, 1121–1129. Betz, A.G., Milstein, C., Gonzalez-Fernandez, A., Pannell, R., Larson, T., and Neuberger, M.S. (1994). Elements regulating somatic hypermutation of an immunoglobulin kappa gene: critical role for the intron enhancer/matrix attachment region. Cell 77, 239–248. Betz, A.G., Neuberger, M.S., and Milstein, C. (1993). Discriminating intrinsic and antigen-selected mutational hotspots in immunoglobulin V genes. Immunol. Today 14, 405–411. Cheng, J., and Maquat, L.E. (1993). Nonsense codons can reduce the abundance of nuclear mRNA without affecting the abundance of pre-mRNA or the half-life of cytoplasmic mRNA. Mol. Cell. Biol. 13, 1892–1902.

Hypermutation and Transcription of Mutant Ig Genes 113

Chomczynski, P., and Sacchi, N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156–159. Coffman, R.L. (1982). Surface antigen expression and immunoglobulin gene rearrangement during mouse pre-B cell development. Immunol. Rev. 69, 5–23. Ford, J.E., McHeyzer-Williams, M.G., and Lieber, M.R. (1994). Analysis of individual immunoglobulin lambda light chain genes amplified from single cells is inconsistent with variable region gene conversion in germinal-center B cell somatic mutation. Eur. J. Immunol. 24, 1816–1822. Fo¨rster, I., Vieira, P., and Rajewsky, K. (1989). Flow cytometric analysis of cell proliferation dynamics in the B cell compartment of the mouse. Int. Immunol. 1, 321–331. Gearhart, P.J., and Bogenhagen, D.F. (1983). Clusters of point mutations are found exclusively around rearranged antibody variable genes. Proc. Natl. Acad. Sci. USA 80, 3439–3443. Gonzalez-Fernandez, A., Gupta, S.K., Pannell, R., Neuberger, M.S., and Milstein, C. (1994). Somatic mutation of immunoglobulin lambda chains: a segment of the major intron hypermutates as much as the complementarity-determining regions. Proc. Natl. Acad. Sci. USA 91, 12614–12618. Goyenechea, B., Klix, N., Yelamos, J., Williams, G.T., Riddell, A., Neuberger, M.S., and Milstein, C. (1997). Cells strongly expressing Ig(kappa) transgenes show clonal recruitment of hypermutation: a role for both MAR and the enhancers. EMBO J. 16, 3987–3994. Grummt, I., and Skinner, J.A. (1985). Efficient transcription of a protein-coding gene from the RNA polymerase I promoter in transfected cells. Proc. Natl. Acad. Sci. USA 82, 722–726. Gunter, K.C., Malek, T.R., and Shevach, E.M. (1984). T cell-activating properties of an anti-Thy-1 monoclonal antibody. Possible analogy to OKT3/Leu-4. J. Exp. Med. 159, 716–730. Jacob, J., Kelsoe, G., Rajewsky, K., and Weiss, U. (1991). Intraclonal generation of antibody mutants in germinal centres. Nature 354, 389–392.

and Berns, A. (1991). Simplified mammalian DNA isolation procedure. Nucleic Acids Res. 19, 4293. Lebecque, S.G., and Gearhart, P.J. (1990). Boundaries of somatic mutation in rearranged immunoglobulin genes: 59 boundary is near the promoter, and 39 boundary is approximately 1 kb from V(D)J gene. J. Exp. Med. 172, 1717–1727. Ledbetter, J.A., and Herzenberg, L.A. (1979). Xenogeneic monoclonal antibodies to mouse lymphoid differentiation antigens. Immunol. Rev. 47, 63–90. Lozano, F., Maertzdorf, B., Pannell, R., and Milstein, C. (1994). Low cytoplasmic mRNA levels of immunoglobulin kappa light chain genes containing nonsense codons correlate with inefficient splicing. EMBO J. 13, 4617–4622. Maquat, L.E. (1995). When cells stop making sense: effects of nonsense codons on RNA metabolism in vertebrate cells. Rna 1, 453–465. Miltenyi, S., Mu¨ ller, W., Weichel, W., and Radbruch, A. (1990). High gradient magnetic cell separation with MACS. Cytometry 11, 231–238. O’Brien, R.L., Brinster, R.L., and Storb, U. (1987). Somatic hypermutation of an immunoglobulin transgene in kappa transgenic mice. Nature 326, 405–409. Orphanides, G., Lagrange, T., and Reinberg, D. (1996). The general transcription factors of RNA polymerase II. Genes Dev. 10, 2657– 2683. Peltz, S.W., He, F., Welch, E., and Jacobson, A. (1994). Nonsensemediated mRNA decay in yeast. Prog. Nucleic Acid Res. Mol. Biol. 47, 271–298. Peters, A., and Storb, U. (1996). Somatic hypermutation of immunoglobulin genes is linked to transcription initiation. Immunity 4, 57–65. Rada, C., Ye´lamos, J., Dean, W., and Milstein, C. (1997). The 59 hypermutation boundary of k chains is independent of local and neighbouring sequences and related to the distance from the initiation of transcription. Eur. J. Immunol. 27, 3115–3120.

Jacobs, H., Fukita, Y., van der Horst, G.T.J., de Boer, J., Weeda, G., Essers, J., de Wind, N., Engelward, B.P., Samson, L., Verbeek, S., et al. (1998). Hypermutation of immunoglobulin genes in memory B cells of DNA repair-deficient mice. J. Exp. Med. 187, 1735–1743.

Rajewsky, K. (1996). Clonal selection and learning in the antibody system. Nature 381, 751–758.

Kabat, E.A., and Wu, T.T. (1991). Identical V region amino acid sequences and segments of sequences in antibodies of different specificities. Relative contributions of VH and VL genes, minigenes, and complementarity-determining regions to binding of antibody-combining sites. J. Immunol. 147, 1709–1719.

Reth, M., Imanishi-Kari, T., and Rajewsky, K. (1979). Analysis of the repertoire of anti-(4-hydroxy-3-nitrophenyl)acetyl (NP) antibodies in C 57 BL/6 mice by cell fusion. II. Characterization of idiotopes by monoclonal anti-idiotope antibodies. Eur. J. Immunol. 9, 1004–1013.

Reeder, R.H. (1990). rRNA synthesis in the nucleolus. Trends Genet. 6, 390–395.

Keohavong, P., and Thilly, W.G. (1989). Fidelity of DNA polymerases in DNA amplification. Proc. Natl. Acad. Sci. USA 86, 9253–9257.

Roes, J., Huppi, K., Rajewsky, K., and Sablitzky, F. (1989). V gene rearrangement is required to fully activate the hypermutation mechanism in B cells. J. Immunol. 142, 1022–1026.

Kim, S., Davis, M., Sinn, E., Patten, P., and Hood, L. (1981). Antibody diversity: somatic hypermutation of rearranged VH genes. Cell 27, 573–581.

Roes, J., Mu¨ ller, W., and Rajewsky, K. (1995). Mouse anti-mouse IgD monoclonal antibodies generated in IgD-deficient mice. J. Immunol. Methods 183, 231–237.

Klein, U., Ku¨ppers, R., and Rajewsky, K. (1997). Evidence for a large compartment of IgM-expressing memory B cells in humans. Blood 89, 1288–1298. Kottmann, A.H., Brack, C., Eibel, H., and Kohler, G. (1992). A survey of protein-DNA interaction sites within the murine immunoglobulin heavy chain locus reveals a particularly complex pattern around the DQ52 element. Eur. J. Immunol. 22, 2113–2120. Kottmann, A.H., Zevnik, B., Welte, M., Nielsen, P.J., and Kohler, G. (1994). A second promoter and enhancer element within the immunoglobulin heavy chain locus. Eur. J. Immunol. 24, 817–821. Kuhn, A., Deppert, U., and Grummt, I. (1990). A 140-base-pair repetitive sequence element in the mouse rRNA gene spacer enhances transcription by RNA polymerase I in a cell-free system. Proc. Natl. Acad. Sci. USA 87, 7527–7531.

Rogerson, B.J. (1994). Mapping the upstream boundary of somatic mutations in rearranged immunoglobulin transgenes and endogenous genes. Mol. Immunol. 31, 83–98. Rogozin, I.B., and Kolchanov, N.A. (1992). Somatic hypermutagenesis in immunoglobulin genes. II. Influence of neighbouring base sequences on mutagenesis. Biochim. Biophys. Acta 1171, 11–18. Rose, M.L., Birbeck, M.S., Wallis, V.J., Forrester, J.A., and Davies, A.J. (1980). Peanut lectin binding properties of germinal centres of mouse lymphoid tissue. Nature 284, 364–366. Rothenfluh, H.S., Taylor, L., Bothwell, A.L., Both, G.W., and Steele, E.J. (1993). Somatic hypermutation in 59 flanking regions of heavy chain antibody variable regions. Eur. J. Immunol. 23, 2152–2159.

Ku¨hn, R., Rajewsky, K., and Muller, W. (1991). Generation and analysis of interleukin-4 deficient mice. Science 254, 707–710.

Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press).

Lacy, E., Roberts, S., Evans, E.P., Burtenshaw, M.D., and Costantini, F.D. (1983). A foreign beta-globin gene in transgenic mice: integration at abnormal chromosomal positions and expression in inappropriate tissues. Cell 34, 343–358.

Schreck, R., Carey, M.F., and Grummt, I. (1989). Transcriptional enhancement by upstream activators is brought about by different molecular mechanisms for class I and II RNA polymerase genes. EMBO J. 8, 3011–3017.

Laird, P.W., Zijderveld, A., Linders, K., Rudnicki, M.A., Jaenisch, R.,

Schwenk, F., Baron, U., and Rajewsky, K. (1995). A cre-transgenic

Immunity 114

mouse strain for the ubiquitous deletion of loxP-flanked gene segments including deletion in germ cells. Nucleic Acids Res. 23, 5080– 5081. Sharpe, M.J., Milstein, C., Jarvis, J.M., and Neuberger, M.S. (1991). Somatic hypermutation of immunoglobulin kappa may depend on sequences 39 of C kappa and occurs on passenger transgenes. EMBO J. 10, 2139–2145. Smale, S.T., and Tjian, R. (1985). Transcription of herpes simplex virus tk sequences under the control of wild-type and mutant human RNA polymerase I promoters. Mol. Cell. Biol. 5, 352–362. Sonoda, E., Pewzner-Jung, Y., Schwers, S., Taki, S., Jung, S., Eilat, D., and Rajewsky, K. (1997). B cell development under the condition of allelic inclusion. Immunity 6, 225–233. Springer, T., Galfre, G., Secher, D.S., and Milstein, C. (1979). Mac1: a macrophage differentiation antigen identified by monoclonal antibody. Eur. J. Immunol. 9, 301–306. Taki, S., Meiering, M., and Rajewsky, K. (1993). Targeted insertion of a variable region gene into the immunoglobulin heavy chain locus. Science 262, 1268–1271. Thompson, A., Timmers, E., Schuurman, R.K., and Hendriks, R.W. (1995). Immunoglobulin heavy chain germ-line J H-C mu transcription in human precursor B lymphocytes initiates in a unique region upstream of DQ52. Eur. J. Immunol. 25, 257–261. Tindall, K.R., and Kunkel, T.A. (1988). Fidelity of DNA synthesis by the Thermus aquaticus DNA polymerase. Biochemistry 27, 6008–6013. Tumas-Brundage, K., and Manser, T. (1997). The transcriptional promoter regulates hypermutation of the antibody heavy chain locus. J. Exp. Med. 185, 239–250. Weber, J.S., Berry, J., Manser, T., and Claflin, J.L. (1991). Position of the rearranged V kappa and its 59 flanking sequences determines the location of somatic mutations in the J kappa locus. J. Immunol. 146, 3652–3655. Wood, C., and Tonegawa, S. (1983). Diversity and joining segments of mouse immunoglobulin heavy chain genes are closely linked and in the same orientation: implications for the joining mechanism. Proc. Natl. Acad. Sci. USA 80, 3030–3034. Yelamos, J., Klix, N., Goyenechea, B., Lozano, F., Chui, Y.L., Gonzalez Fernandez, A., Pannell, R., Neuberger, M.S., and Milstein, C. (1995). Targeting of non-Ig sequences in place of the V segment by somatic hypermutation. Nature 376, 225–229. Zawel, L., Kumar, K.P., and Reinberg, D. (1995). Recycling of the general transcription factors during RNA polymerase II transcription. Genes Dev. 9, 1479–1490. Zawel, L., and Reinberg, D. (1992). Advances in RNA polymerase II transcription. Curr. Opin. Cell Biol. 4, 488–495. Zhang, G., Taneja, K.L., Singer, R.H., and Green, M.R. (1994). Localization of pre-mRNA splicing in mammalian nuclei. Nature 372, 809–812.