tion endonuclease fragment derived from the 3' end of theEcoRI fragment carrying the murine JH-region genes (22) (see Fig. 2). The VH-region probe (V probe) ...
Proc. Nati Acad. Sci. USA Vol. 80, pp. 4997-5001, August 1983 Genetics
Recombination between antibody heavy chain variable-region genes: Evidence for gene conversion (somatic diversification)
ULRICH KRAWINKEL, GABRIELE ZOEBELEIN, MARIANNE BRUGGEMANN , ANDREAS RADBRUCH, AND KLAus RAJEWSKY Institute for Genetics, University of Cologne, Weyertal 121, D-5000 Cologne 41, Federal Republic of Germany
Communicated by Elvin A. Kabat, April 13, 1983
ABSTRACT The murine hybridoma line B1-8.81 secretes monoclonal IgD Al antibodies specific for the hapten (4-hydroxy3-nitrophenyl)acetyl (NP). The variable (V) region of these antibodies is defined by a characteristic pattern of idiotopes. A spontaneous V-region variant (BI-8.V1) with altered idiotope pattern was selected. The structural variation is confined to the V region of the heavy chain. It was shown previously that the variant V region is encoded by a gene that was generated by a crossover between the rearranged VDJ gene of the wild type (B1-8. 61) and a neighboring germ-line VH gene. In the present study the nucleotide sequence of coding and flanking regions of the VH gene expressed in variant B1-8.V1 was determined. Wild-type and variant VH genes differ at 15 positions in a region between leader sequence and codon 66. The sequence of the region carrying the substitutions is identical to the sequence of the corresponding region in a neighboring germ-line VH gene. This implies that the variant VH gene was generated by a mechanism of recombination more complicated than single crossover. Gene conversion as the mechanism of the recombination is discussed.
(8), as the mechanism that generated the mutant phenotype of B1-8.VL. The recombination involved a crossover located between codons 66 and 70 of the genes V102.1 and V186.2 that is expressed by BL-8.61. However, because the genes V102.1 and V186.2 are translated into V regions exhibiting identical amino acid sequences up to position 20, amino acid sequence analysis of the BL-8.V1 heavy chain could not resolve whether a second crossover between V102.1 and V186.2 had occurred in the gene segments upstream of codon 21. We therefore determined the nucleotide sequence of the rearranged variant VH gene and its flanking regions. Comparison of this sequence to the wild-type VDJ gene sequence and to the sequence of the germ-line VH gene V102.1 demonstrates that a double crossover occurred in vitro between V102.1 and the VDJ gene of BI8. 81, thus creating a mutant rearranged VH gene. Alternatively, gene conversion between H chain V-region genes might account for the recombinant genotype of B1-8.V1. MATERIALS AND METHODS Cell Lines. The cell line B1-8.V1 is a somatic variant of the mouse hybridoma line B1-8. 61 (13) as described (15). High molecular weight DNA of B1-8.V1 was extracted from tissue-culture cells (16). Southern Filter Hybridization. Purified DNA probes were labeled to a specific activity of 2 x 108 cpm/fug by nick-translation (17). Transfer of DNA to nitrocellulose paper and hybridization were carried out as described (18). The filters were hybridized for 16-36 hr to 1 x 106 cpm of radioactive probe per ml at 650C in a shaking water bath. The salt conditions for hybridization were 450 mM NaCI/45 mM sodium citrate/0. 1% sodium dodecyl sulfate/10% dextran sulfate, sonicated denatured salmon sperm DNA at 100 jug/ml, and 0.2% each of Ficoll 400, bovine serum albumin, and polyvinylpyrrolidone 360. The unhybridized probe was removed from the filters by washing in 300 mM NaCI/30 mM sodium citrate/0.5% sodium dodecyl sulfate and 15 mM NaCl/1.5 mM sodium citrate/0.5% sodium dodecyl sulfate at 650C (19, 20). Hybridization was visualized by autoradiography using prefogged x-ray films (21). DNA Probes and Recombinant Phage DNAs. The JH-region probe (J probe) is a 0.6-kilobase-pair (kb) Xba I/EcoRI restriction endonuclease fragment derived from the 3' end of the EcoRI fragment carrying the murine JH-region genes (22) (see Fig. 2). The VH-region probe (V probe) used in this study is derived from plasmid pAB'y2a-1 (8), which contains the cDNA encoding
Variable (V) regions of immunoglobulin heavy (H) and light (L) chains are encoded by polymorphic gene segments, which join as a result of DNA rearrangements in the precursors of antibody-producing cells (1-5). The combination of V, diversity (D), and joining (J) segments in active heavy chain V-region genes generates diversity particularly in the third hypervariable region (3, 4). Additional diversity is introduced into rearranged V-region genes by somatic mutations. One mechanism to generate diversity has been shown to be point mutations (2, 6-9); a possible other one may be recombination between V-region genes (10, 11). A case of somatic recombination between antibody H chain V-region genes has been described by Dildrop et at (12). The present study extends this analysis. We use as a model system the hybridoma line B1-8.61 (13), which synthesizes an IgD Al antibody specific for the hapten (4-hydroxy-3-nitrophenyl)acetyl (NP). B1-8. 61 was derived as a spontaneous class-switch variant from the hybridoma Bl-8.Lu (IgM, Al) (14). The H chain variable region of B1-8.,a and BL8.61 is encoded by the germ-line V-region gene V186.2, which belongs to the NPe-VH gene cluster (8). The spontaneous variant Bl-8.V1 that lost an antigenic V-region determinant (idiotope) has previously been isolated from B1-8. 61 by cell sorting using anti-idiotope antibodies. Serological and biochemical analysis demonstrated that the structural difference between variant and wild type is confined to the V region of the antibody H chain (15). Amino acid sequence analysis (12) showed recombination between the VDJ region of Bl8.61 and a neighboring germ-line H chain V-region gene, VI 02.1
Abbreviations: H chain, immunoglobulin heavy chain; VH region, variable region of the immunoglobulin H chain; VH gene, nucleotides coding for the VH region through FR3; FR, framework segment of the V region; D, diversity; J, joining; CDR, complementarity-determining region; IVS, intervening sequence; NP, (4-hydroxy-3-nitrophenyl)acetyl; kb, kilobase pair(s); bp, base pair(s).
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 4997
4998
Genetics: Krawinkel et al.
the entire H chain of anti-NP antibody S43 (14). The V probe contains a 0.35-kb Pst I fragment carrying the VH segment cloned in pABy2a. Both J and V probe were kindly given to us by A. L. M. Bothwell, Cambridge, MA. A probe containing the leader region (codons -19 to +3) of the VH gene expressed by BL-8.,u (14) was a gift of P. H. Schreier, Cologne, Federal Republic of Germany. The leader region probe represents the 85-base-pair (bp) Pst I fragment of plasmid pABA1l1 (8). Plasmids and phage M13 replicative forms were purified on CsCl gradients. Plasmids pV186.2 and phage AV102 were also gifts from A. L. M. Bothwell. They contain germ-line VH genes V186.2 and V102.1/ 2, respectively, of the NPb-VH gene family (8). Construction and Screening of a A Phage Library. DNA extracted from B1-8.V1 was digested to completion with EcoRI. Restriction fragments with an average length of 5 kb were enriched on a sucrose gradient, ligated to the arms of A phage Charon 30 (23), and packaged in vitro (24). The resulting library of -3 x 104 recombinant phages was directly screened by in situ hybridization to the nick-translated J probe (25). A second library screening was performed after phage amplification. Replicas of the phage plates on nitrocellulose filters were hybridized, washed, and autoradiographed, using conditions described above. Positively hybridizing phages were grown as described (26). DNA Sequence Analysis. Nucleotide sequence analysis was carried out by using the single-stranded phage M13 vector mp701 (gift of D. Bentley, Oxford). The dideoxy chain termination procedure was employed in all experiments (27, 28).
RESULTS Isolation and Characterization of a A Phage Clone Containing the Expressed VH Gene of B1-8.V1. High molecular weight DNA from B1-8. 81 and B1-8.V1 was analyzed by Southern blot hybridization using the J probe. This probe detects a 6.6-kb and a 4.4-kb EcoRI restriction fragment in the genome of both cell lines (data not shown). As demonstrated in a previous study, the 6.6-kb EcoRI fragment of B1-8.81 originates from the genome of the X63.Ag8 fusion partner (29) (note that the B1-8. 81 hybridoma arose from the fusion between a BALB/ c-derived X63 cell and a C57BL/6 lymphocyte). We thus conclude that the VDJ gene of wild-type B1-8. 81 and of variant Bi8.V1 is located on the 4.4-kb EcoRI fragment. This conclusion is confirmed by Southern blot hybridization analysis of wild-type and variant genomic DNAs using the V probe. A 4.4-kb EcoRI fragment is detected that is absent from the genome of C57BL/ 6 liver cells (data not shown). To isolate the expressed VH gene of B1-8.V1 we enriched the 4.4-kb EcoRI restriction fragment from Bl-8.V1 DNA on a sucrose gradient and cloned it in the A phage vector Charon 30. The resulting phage library was screened without previous amplification by using the J probe. Positive phages were rescreened with the V probe, and one clone that hybridized to both J and V probes was isolated. This clone, designated AVAR13, was subjected to restriction mapping, hybridization, and sequence analysis. A second screening of the amplified library yielded seven clones that also were positive with the J and V probes (AVAR1-AVAR7). Four of these clones were studied by restriction mapping. The amino acid sequence of the Bl-8.Vl H chain showed that the variant VH gene arose by recombination between the wild-type VDJ gene of B1-8. 81 and another member of the NPb_ VH gene family, V102.1 (12). However, amino acid sequence analysis could not resolve whether a single crossover or a more complicated mechanism was involved in the generation of the variant. Because a single crossover between V186.2 and V102.1 should result in a variant V gene with V102.1-like 5' flanking regions, we mapped AVAR13 with restriction enzymes able to discriminate between the 5' flanks of V186.2 and V102.1. The
Proc. NatL Acad. Sci. USA 80 (1983) A
B
C
D
kb 1.9-
1.51.3-
FIG. 1. Southern blot hybridization of the leader region probe to AVAR13, AV102, and pV186.2 DNA. pV186.2 DNA was digested with Pst I and Sac I (lane A). AVAR13 DNA was digested withPst I and Sac I (lane B) andPstI andHindlll (lane C). AV102 DNA was digested with Pst I andHindIII (lane D). Lane A was exposed for a shorter period and lane D was exposed for a longer period than lanes B and C in order to make band intensities more similar.
Southern blot hybridization analysis of AVAR13 DNA digested with the restriction enzymes Sac I, HindIII, and Pst I (which cuts both V186.2 and V102.1 at the same position) is shown in Fig. 1. A restriction map of AVARL3 is presented in Fig. 2. The VDJ gene was located within the clone by the V probe, whereas the probe hybridizing to the V-region leader sequence (see Materials and Methods) was used to determine gene orientation. Fig. 1 shows double digests of AVAR13 DNA with Pst I plus Sac I (lane B) and Pst I plus HindIII (lane C), a Pst I/Sac I double digest of plasmid V186.2 (lane A) and a Pst I/HindIII double digest of phage AV102 DNA (lane D) analyzed by Southern blot hybridization using the leader region probe. Lanes A and B both show a hybridizing 1.3-kb Pst I/Sac I fragment indicating that the 5' flanking region of the VDJ gene cloned in AVAR13 carries the diagnostic Sac I site of V186.2. The diagnostic 1.5-kb Pst I/HindIII fragment of AV102 DNA (lane D) does not show up in the corresponding digest of AVAR13 DNA (lane C). The 1.9-kb Pst I/HindIII fragment found here is flanked by a HindIII site that maps outside of the VDJ-containing EcoRI fragment in AVAR13. This analysis indicates that the variant VDJ region cloned in AVAR13 is flanked by V186.2-like and not V102.1-like sequences. Further restriction mapping detected a deletion of 1.4 kb between JH3 and the Xba I site of the VDJ-containing EcoRI fragment in AVAR13 (broken line in Fig. 2). This deletion probably represents a cloning artefact because it is not observed in other phages of the AVAR series. Nucleotide Sequence of the Rearranged VH Gene of B1-8.V1. AVAR13 DNA was digested with Pst I and Sau3A. The resulting
AVAR 13
Sa
1
P VH
DJH2
FIG. 2. The expressed VDJ region of B1-8V1. Restriction map of the AVAR13 insert. R, EcoRI; B, BamHI; P, Pst I; S, Sac I; X, Xba I; Sa, Sau3A. Protein-encoding regions are shown by raised boxes. The L (leader), VH, D, JH2, and JH3 regions were identified by nucleotide sequence analysis. The region cloned in the J probe is shown as an open box. Arrows indicate our sequencing strategy.
Proc. Natl. Acad. Sci. USA 80 (1983)
Genetics: Krawinkel et al.
fragments were ligated at random into the M13 vector mp701. Recombinant phages were screened with the V probe and positive phages were subjected to nucleotide sequence analysis. The sequencing strategy is indicated by arrows in Fig. 2. Fig. 3 shows the complete nucleotide sequence of the coding and flanking regions of the rearranged VH gene of B1-8.V1. The sequence of the variant VH gene-designated V1 in Fig. 3-is compared to the sequence of the wild-type B1-8 and to the sequences of germ-line genes V186.2 and V102.1 (8). The result of the sequence comparison can be summarized as follows: (i) The amino acid sequence deduced from the nucleotide sequence of BL-8.V1 matches the amino acid sequence published previously (12). SEA
VI . . GA.TgACTGTT b 14
.......
..........
V18b-2.... .
VI........
.............. ..........
.............. ...................... ............. --
---
...............
V102-1
-9 CTCTTTACAG TTACTGAGCA CACAGGACCT CACCATGGOA
TUGAOACTIfA
(ii) B1-8.V1 and B1-8 VH genes are both rearranged to DFL16 (30) and JH2 (iii) The VH region of B1-8.V1 differs in 15 positions from Bi8, which expresses the germ-line sequence V186.2. (iv) The substitutions in the B1-8.V1 sequence are confined to a segment that starts at codon 11 and ends at codon 66. The
sequence of this segment matches exactly the corresponding sequence of V102.1. Positions diagnostic for the beginning and the end of the V102.1-like sequence are marked by W. (v) The V102.1-like segment of B1-8.V1 is flanked at the 5' end by a region of V186.2-like sequence up to position 34 (*) of the IVS between leader region and VH region. The following sequence up to codon 11 is shared between V186.2 and VI 02.1 40
VI .
.
.......... ..........
.....
B1-8.-A-
V102-1
---------
V1U2-1 ....
. ----------
---0--- -AA-AG-0-
A-
G--- -- ----------
60
AGTTCAAGGO
VI .. TGATACTAAC TACAATCAAA --o--G ------.G-G-
----.C.-------T---- --
CACAGTAGCA
GGCTTGAGGT
.....................
............
V186-2
....
VI12-1
....
CTGOACATAT .. .... ....
--------A-
----------
--
ACATGUGTGA CAATGACATC ...... ....
..........
-C
G- --------A-
--------
---------- ------
.-
-
-----
80
VI .
ACAAACCCTC CAGCACAUCC TACATUCAOC TCAGCAGCCT GACATCTGAG
B1-8 .----
TCCCAGGTCC
V12-1 ....-----T----
Psti
AACTGCAGCA
BVI86..... ..........
..........
....
----------
----------
--____
---------- --
---------- ---------- ---------- --------
90
98
B1-8.--
------
_________
---
--------
---
----------
-----
D
a
VI........ GACTCTGCGG TCTATTATTG TGCAAGATAC GATTACTACO GTAGTAGCTA -
V1W2-1 ....
CTGACTUTAG
-
- ----------
4
----
*70
V186-2 .... ------------------ ---------
VI........ CACTTTOCCT TTCTCTCCAC AGGTGTCCAC
Via6-2
CAAGGCCACA
----
V12- 1....
B148
--AA-AG-0-
----------
------
V186-2 .... --I-I.U -----.
VI........
- 0-
----------.-
--- ----
....
B318 .
V186-2 --
-
-5
IL TCATGCTCTT CTTUUCAGCA ACAGCTACAG GTAAGGOUCT
CDR 2
o50
AGGCCT0GCC AAGGCCTTGA GTGATTOAAUGATTCATC CTTCTGATAG
V186-2....
..........
4999
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
---
B186-2....V102-1-.-. -------C------T.... .......... ----------
10 *
VI .1 CC.TGGCT GAGCTGGTGA AGCCT
C TTCAGT
31-8......----
------
V186-2 ....
V102-i....
---. ---T
.
------
---
20
GTGTCCTGCA ---
.------
.
----
-
. -
..........
JH2 VI........ CTTGACTAC TGGGGCCAAG GCACCACTCT CACAGTCTCC TCAGGTGAGT VI
CCTTACAACC TCTCCTTCT ATTCAGCTTA AATAGATTTT ACTGCATTTG
VI........ TTGGG VI .
A AATOTGTGTA TCTGAATTTC AOGTCATAAG GACTAGGAC
ACCTtGGGAG TCAOAAAGGG TCATTGGGAG CCCTOTOACG CAGACAGACA
Bam H1
VI........ TCCCAGCTC CCATACTTCA TGOCCAGAGA TTTATAGGGA TCCTGGCCAG
CDR I VI .
AGGTTCTW
CTACACCTTC ACCAGCTACT OGACA OON AAGCAG
VI........ CATTOCCGCT AGGTCCCTC? CTTCTATGCT TTCTTTGtCC CTCACTGGCC
VI........ TCCATCGAG ATCACCTGA
V1862
...._
VI.
OTCTAA!CAT
TOITTCACA
OCCCTAGCCA AGGATCATTT ATTGTCAOGG
JH3 ATGTOCCTOO TTGCTTACT
Pati
coooCAAGG
VI.......-G9AC!C??OTC ACTctCBCTO C-..
FIG. 3. Nucleotide sequence of the VDJ gene segment cloned in AVAR13. The sequence of the VDJ region of B1-8.V1 (Vi) is compared to the sequences of VH genes B1-8, V186.2, and V102.1 (see ref. 8). A broken line means sequence identity and a dotted line means the sequence was not determined. A * indicates a diagnostic position at the end or at the beginning of V186.2 segments in the B1-8.V1 sequence. A * marks a diagnostic position at the beginning or at the end of the V102.1-derived segment in the B1-8.V1 sequence. Restriction sites used to generate fragments for sequencing are underlined. The codons of the protein-encoding regions are numbered. IVS, intervening sequence; CDR, complementarity-determining region.
Genetics: Krawinkel et aL
5000
LI-
v
drI
lb
Proc. Natl. Acad. Sci. USA 80 (1983) V 102.1
+zzr3 ID A
-5
m
66 70 --
10
-41-
I I
B1-8. A
i
B1-8.V1
FIG. 4. Generation of the variant B1-8.V1 VDJ region by recombination between V102.1 and the VDJregion of wild-type B1-8.81. The segment indicated by arrows in the VDJ gene of B1-8.81 is replaced by the corresponding segment of V102.1. The left point of recombination is confined to the region between base pair 34 of the intervening quence separating leader (L, ends at codon -5) from VH region (V) and codon 10. The right point of recombination maps to the region between codons 66 and 70. Protein encoding regions are shown by wider bars. se-
and thus
cannot
to either V186.2 or V102.1-like segment of B1-8.V1,
be assigned specifically
Vi02.1. At the 3' end of the
the V186.2 sequence can be identified starting at codon 70 (*). No discrimination between Vi 02.1 and V186.2 is possible in the region between codons 66 and 70. In conclusion, the V-region variant B1-8.V1 expresses a variant V186.2 gene whose inner segment, at least between codons 11 and 66, apparently has been replaced by the corresponding region of the Vi02.1 gene (Fig. 4). The sequence of B1-8.V1 extends into theJH region, ending at the Pst I site of fH3. Both exons, JH2 andJH3, match the corresponding germ-line nucleotide sequence of the JH locus of the a haplotype (31). Because no reference for the germ-line sequence of the b haplotypeJH locus is known so far, we do not know whether the IVS between JH2 and JH3 of B1-8.V1 represents a germ-line sequence. DISCUSSION Variant Genotype. We determined the nucleotide
sequence
of the VDJ segment and its flanking regions of the variant cell line B1-8.V1. The 15-bp differences between mutant and wildtype VDJ regions are confined to a segment of 164 bases with long regions of extensive sequence homology between B1-8.V1 and V186.2 on either side. We show by sequence comparison with another member of the NP"-VH gene cluster (8) that the variant 164-bp region in B1-8.V1 is identical to the corresponding region in germ-line gene V102.1. It is unlikely that the mutant genotype arose by frequent point mutations rendering a part of the wild-type V186.2 sequence into one that matches precisely the sequence of the corresponding part of VI 02.1. Rather, we think that a recombination between the rearranged VDJ segment of B1-8. 81 and germ-line VH gene Vi02.1 generated the rearranged VH segment expressed by BL-8.V1. Our result confirms the conclusion drawn from the previous amino acid sequence analysis of the B1-8.V1 H chain V region (12). Beyond this conclusion we now show that the recombination between the VDJ region of B1-8. 81 and V102.1 possibly involved a more complicated mechanism than a single crossover because the segment that in the VDJ region of B1-8.V1 expresses the V102.1 sequence is flanked on both sides by wild-type-like regions. However, we cannot exclude that the mutant VH gene expressed by B1-8.V1 represents the result of a single crossover between a yet-unknown germ-line gene of the NP_-VH gene family and Vi86.2. Another possible mechanism to generate the recombinant VDJ segment of BL8.V1 is a transchromosomal single crossover between Vi86.2 and a VH gene on a BALB/c-derived chromosome in B1-8.81 cells. A single crossover between V186.2 and an unknown C57BL/
6- or BALB/c-derived germ-line VH gene appears unlikely because one has to assume that in the C57BL/6 or BALB/c genome a VH gene exists whose sequence up to codon 10 precisely matches the V186.2 germ-line sequence, whereas the region between codons 11 and 66 exhibits a sequence identical to V102.1. No C57BL/6- or BALB/c-derived VH gene fulfilling this requirement has been found among the genes related to the NP"VH gene family (refs. 8, 32, and 33; T. Blankenstein and R. Dildrop, personal communication), although one has to concede that the collection of examined genes probably is incomplete. We favor a more complicated mechanism of recombination to account for the genotype of B1-8.V1. A reciprocal exchange of gene segments between V186.2 and V102.1 could have been generated by a double crossover. Alternatively, gene conversion, which has been already suggested as a mechanism to generate or limit diversity in multigene families (11, 34-37), might have transferred genetic information from V102.1 to V186.2. Recombination Between V-Region Genes. So far somatic mutations identified in antibody V genes have been interpreted as point mutations generated by a hypothetical mechanism that operates at some step of B-cell differentiation (8, 9, 38). However, somatic mutation has been studied only in families of antibodies that had similar antigen-binding specificities. Recombinations that generate an antibody with changed antigen-binding specificity, as encountered in variant B1-8.V1 (15), may therefore have gone unnoticed. Our present data provide evidence that recombination between V-region genes may contribute to the somatic generation of antibody diversity. The mutant VDJ gene of B1-8.V1 was generated by insertion of a 164- to 247-bp segment into V186.2, which is the VH gene expressed by the wild-type B1-8.81. This inserted segment originates from germ-line VH gene Vi02.1. Both genes belong to the b haplotype of the IgH locus. Only one copy of the IgHb locus is present in the genome of Bl-8.81 cells (29). Recombination between V186.2 and VI 02.1 thus might have occurred between sister chromatids or between the two genes on the same chromatid. Because of extensive sequence homology between V186.2 and Vi02.1 the points of recombination in the VDJ gene of B1-8.V1 cannot be mapped precisely. However, one breakpoint is confined to the region between base pair 34 of the IVS separating leader from VH region and codon 10. The other breakpoint maps to the region between codons 66 and 70 (see Figs. 3 and 4). The finding that B1-8.V1 whose VDJ gene segment contains two points of recombination was isolated as the only idiotope loss variant from =106 cells of the original Bl-8.81 cell population (15) suggests that the insertion of a V102.1-derived gene segment involving the region between codons 10 and 66 into the VDJ gene segment of B1-8. 81 was as frequent as or even more frequent than a single crossover between the genes in the region between codons 66 and 70. There is no bias against the latter type of recombination in our method to select variants because such a crossover would have generated the same phenotype as the one observed in B1-8.V1. As a mechanism of recombination to account for the recombinant genotype of B1-8.V1 we suggest gene conversion, which is reported to occur frequently in mitotic yeast cells among homologous DNA sequences of allelic and nonallelic genes located on the same or different chromosomes (39-41). In gene conversion information is transferred in a nonreciprocal way from a donor to a recipient gene. Both genes retain their integrity and their physical locations but an alteration in the structure of the recipient gene occurs in that the nucleotide sequence of part or allof the recipient gene becomes identical to that of the donor gene. Gene conversion between antibody V-region genes exhibiting regions of sequence homology may be generated by sin-
Genetics: Krawinkel et al.
gle-strand exchanges followed by mismatch repair or gap repair
in the resulting heteroduplexes (42-44). The extensive homology of the regions flanking the gene segment between codons 10 and 66 could have promoted conversion of V186.2 by V102.1 in hybridoma B1-8. 81. Indications for gene conversion between nonallelic genes have been found in the human fetal globin gene family (35) and in the murine major histocompatibility gene complex (45). Recent data indicate that gene conversion may play a role in the evolution of antibody constant region (46, 47) and V genes (48, 49). Amino acid sequence analysis of antibody L chain V regions shows evidence that.a particular FRi or FR2 framework segment assorts with different CDR1s and CDR2s (summarized by Kabat and Wu in refs. 50 and 51). Kabat and Wu put forward the hypothesis that FR and CDR segments are separate "minigenes" in the genome and are able to join independently. The amino acid sequence of the present variant B1-8.V1 appears to confirm this hypothesis in that a new CRD2 is inserted between FR2 and FR3 of the B1-8. 81 H chain, although one has to assume additional point mutations to explain the alterations at position 20 (FRI) and 43 (FR2) of the variant V region. The result of the nucleic acid sequence analysis of the VDJ gene of B1-8.V1 is incompatible with the interpretation that insertion of a "CDR2 minigene" into the gene segment encoding the framework of the B18. 81 VDJ region generated the variant B18.V1 H chain. It is shown in this report that a gene segment composed of FR1, CDR1, .FR2, and CDR2 of a neighboring VH gene is inserted at the corresponding position of the VH gene expressed by B1-8.V1. Assortment of a particular FR sequence with different CDRs (50, 51) may thus be the result of conversion between V genes that encode similar FRs but differ in the segments encoding CDR1 and CDR2. Such genes are exemplified by some members of the NP"-VH gene family (8). In a recent report Wu and Kabat suggest that a gene conversion mechanism may move segments encoding CDRs from one gene to another (52). Gene conversion may explain recombination events between homologous genes in mammalian cells, but all observations of recombination in higher eukaryotes-recombination between V186.2 and V102.1 included-may be explained by unequal double crossovers as well. Such a mechanism could generate the variant VDJ segment of B1-8.V1 by a reciprocal exchange
of gene segments between the wild-type VDJ segment and V102.V1 However, it is impossible to examine both V186.2 and V102.1 after the postulated gene conversion once the two participants in the recombination are located on segregating chromatids. We thank E. Winter for help with DNA sequence analysis, T. Blankenstein for screening the amplified B1-8.V1 phage library, P. H. Schreier for the gift of the leader region probe, and D. Bentley for the gift of phage M13 mp701. We gratefully acknowledge the generosity of A. L. M. Bothwell, who provided probes and valuable unpublished information. This work was supported by the Deutsche Forschungsgemeinschaft through SFB 74 and a Heisenberg-Stipendium to U.K. 1. Brack, C., Hirama, M., Lenhard-Schuller, R. & Tonegawa, S. (1978) Cell 15, 1-14. 2. Bernard, O., Hozumi, N. & Tonegawa, S. (1978) Cell 15, 11331144. 3. Sakano, H., Huppi, K., Heinrich, G. & Tonegawa, S. (1979) Nature (London) 280, 288-294. 4. Max, E., Seidman, J. G. & Leder, P. (1979) Proc. Natt Acad. Sci. USA 76, 3450-3454. 5. Early, P. W., Huang, H., Davis, M., Calame, K. & Hood, L. (1980) Cell 19, 981-992. 6. Weigert, M. & Riblet, R. (1976) Cold Spring Harbor Symp. Quant. BioL 41, 837-846. 7. Pech, M., Hochtl, J., Schnell, H. & Zachau, H. G. (1981) Nature (London) 291, 668-670.
Proc. Natl. Acad. Sci. USA 80 (1983)
5001
8. Bothwell, A. L. M., Paskind, M., Reth, M., Imanishi-Kari, T., Rajewsky, K. & Baltimore, D. (1981) Cell 24, 625-637. 9. Crews, S., Griffin, J., Huang, H., Calame, K. & Hood, L. (1981) Cell 25, 59-66. 10. Smithies, 0. (1967) Cold Spring Harbor Symp. Quant. Biol. 32, 161-168. 11. Edelman, G. & Gally, J. (1970) The Neurosciences Second Study Program, ed. Schmitt, F. (Rockefeller Univ. Press, New York). 12. Dildrop, R., Bruggemann, M., Radbruch, A., Rajewsky, K. & Beyreuther, K. (1982) EMBO J. 1, 635-640. 13. Neuberger, M. S. & Rajewsky, K. (1981) Proc. Natt Acad. Sci. USA 78, 1138-1142. 14. Reth, M., Hammerling, G. J. & Rajewsky, K. (1978) Eur. J. Immunol. 8, 393-400. 15. Bruggemann, M., Radbruch, A. & Rajewsky, K. (1982) EMBO J. 1, 629-634. 16. Steffen, D. & Weinberg, R. A. (1978) Cell 5, 1003-1010. 17. Rigby, P. W. J., Dieckmann, M., Rhodes, C. & Berg, P. (1977)J. Mol. Biot 113, 237-257. 18. Southern, E. M. (1975) J. Mot Biol. 98, 503-517. 19. Denhardt, D. T. (1966) Biochem. Biophys. Res. Commun. 23, 641646. 20. Jeffreys, A. J. & Flavell, R. A. (1977) Cell 18, 559-568. 21. Laskey, R. A. & Mills, A. D. (1977) FEBS Lett. 82, 314-316. 22. Alt, F. W., Rosenberg, N., Lewis, S., Thomas, E. & Baltimore, D. (1981) Cell 27, 381-400. 23. Rimm, D. L., Horness, D., Kucera, J. & Blattner, F. R. (1980) Gene 12, 301-309. 24. Hohn, B. & Murray, K. (1977) Proc. Natt Acad. Sci. USA 74, 32593263. 25. Benton, W. D. & Davis, R. W. (1977) Science 196, 180-182. 26. Bentley, D. L. & Rabbitts, T. H. (1981) Cell 24, 613-633. 27. Schreier, P. H. & Cortese, R. (1979)1. Mol. Biol. 129, 169-172. 28. Sanger, F., Coulson, A. R., Barrell, B. G., Smith, A. J. H. & Roe, B. A. (1980)J. Mot Biol. 143, 161-178. 29. Sablitzky, F., Radbruch, A. & Rajewsky, K. (1982) Immunol. Rev. 67, 59-72. 30. Kurosawa, Y. & Tonegawa, S. (1982) J. Exp. Med. 155, 201-218. 31. Gough, N. M. & Bernard, 0. (1981) Proc. Nat. Acad. Sci. USA 78, 509-513. 32. Cohen, J. B., Effron, K., Rechavi, G., Ben-Neriah, Y., Zakut, R. & Givol, D. (1982) Nucleic Acids Res. 11, 3353-3365. 33. Loh, D. Y., Bothwell, A. L. M., White-Scharff, M. E., ImanishiKari, T. & Baltimore, D. (1983) Cell 33, 85-93. 34. Seidman, J. G., Leder, A., Nau, M., Norman, B. & Leder, P. (1978) Science 202, 11-17. 35. Slightom, J. L., Blechl, A. E. & Smithies, 0. (1980) Cell 21, 627638. 36. Egel, R. (1981) Nature (London) 290, 191-192. 37. Baltimore, D. (1981) Cell 24, 592-594. 38. Gearhart, P. J., Johnson, N. D., Douglas, R. & Hood, L. (1981) Nature (London) 291, 29-34. 39. Klar, A. J. S. & Fogel, S. (1979) Proc. Nat. Acad. Sci. USA 76, 4539-4543. 40. Jackson, J. A. & Fink, G. R. (1981) Nature (London) 292, 306310. 41. Ernst, J. R., Stewart, J. W. & Sherman, F. (1981) Proc. Natt Acad. Sci. USA 78, 6334-6338. 42. Holliday, R. (1964) Genet. Res. 5, 282-304. 43. Meselson, M. S. & Radding, C. (1975) Proc. Natl. Acad. Sci. USA 72, 358-361. 44. Szostak, J. W., Orr-Weaver, T. L., Rothstein, R. J. & Stahl, F. W. (1983) Cell 33, 25-33. 45. Evans, G. A., Margolies, D. H., Carmerini-Otero, R., Ozato, K. & Seidman, J. G. (1982) Proc. Natl. Acad. Sci. USA 79, 1994-1998. 46. Schreier, P. H., Bothwell, A. L. M., Muller-Hill, B. & Baltimore, D. (1981) Proc. Natt. Acad. Sci. USA 78, 4495-4499. 47. Ollo, R. & Rougeon, F. (1983) Cell 32, 515-523. 48. Clarke, S. H., Claflin, L. & Rudikoff, S. (1982) Proc. Nat. Acad. Sci. USA 79, 3280-3284. 49. Bentley, D. L. & Rabbitts, T. H. (1983) Cell 32, 181-189. 50. Kabat, E. A., Wu, T. T. & Bilofsky, H. (1978) Proc. Natt. Acad. Sci. USA 75, 2429-2433. 51. Kabat, E. A., Wu, T. T. & Bilofsky, H. (1979) J. Exp. Med. 149, 1299-1313. 52. Wu, T. T. & Kabat, E. A. (1982) Proc. Natl. Acad. Sci. USA 79, 5031-5032.
