key: cord-0005030-kwbc7cv2 authors: Strohal, Robert; Helmberg, Arno; Kroemer, Guido; Kofler, Reinhard title: MouseV(k) gene classification by nucleic acid sequence similarity date: 1989 journal: Immunogenetics DOI: 10.1007/bf02421180 sha: 7a8e460c3c89871c3c04b19c31565d68dc4a1301 doc_id: 5030 cord_uid: kwbc7cv2 Analyses of immunoglobulin (Ig) variable (V) region gene usage in the immune response, estimates ofV gene germline complexity, and other nucleic acid hybridization-based studies depend on the extent to which such genes are related (i. e., sequence similarity) and their organization in gene families. While mouseIgh heavy chainV region (V (H)) gene families are relatively well-established, a corresponding systematic classification ofIgk light chainV region (V (k)) genes has not been reported. The present analysis, in the course of which we reviewed the known extent of theV (k) germline gene repertoire andV (k) gene usage in a variety of responses to foreign and self antigens, provides a classification of mouseV (k) genes in gene families composed of members with >80% overall nucleic acid sequence similarity. This classification differed in several aspects from that ofV (H) genes: only someV (k) gene families were as clearly separated (by >25% sequence dissimilarity) as typicalV (H) gene families; mostV (k) gene families were closely related and, in several instances, members from different families were very similar (>80%) over large sequence portions; frequently, classification by nucleic acid sequence similarity diverged from existing classifications based on amino-terminal protein sequence similarity. Our data have implications forV (k) gene analyses by nucleic acid hybridization and describe potentially important differences in sequence organization betweenV (H) andV (k) genes. The ability of the immune system to recognize virtually any antigen is mediated by the enormous sequence variability in the amino-terminal region of immunoglobulin (Ig) heavy and light chains. Among other mechanisms, this diversity is generated by somatic juxtaposition of gene segments that are separated in the germline, termed variable (V), diversity (heavy chain only), and joining (J) gene segments (reviewed by Tonegawa 1983 , Alt et al. 1986 . V genes contribute all residues of the first and second complementarity determining region (CDR) of both heavy and light chains, as well as part of the light chain CDR-3, and hence contribute the majority of antigen contact residues . In mice, several hundred V n and V k (over 90% of all serum Ig is of the Igk isotype) gene segments exist in the germ line (Brodeur and Riblet 1984 , Livant et al. 1986 , Cory et al. 1981 , Kofler et al. 1989 . These genes can be very similar or may differ by over 40% nucleotides, and V region classifications based on nucleic and/or amino acid sequence similarity have been proposed (Brodeur and Riblet 1984 , Dildrop 1984 , Potter et al. 1982 . Thus, mouse V H genes have been grouped in 11 V H gene families in which members generally share > 80% of their nucleic acid sequence within, and < 70-75 % between, families (Brodeur and Riblet 1984 , Winter et al. 1985 , Kofler 1988 , Reininger et al. 1988 . Individual members of a given family cross-hybridize in nucleic acid hybridization assays only with members of their own family. These VH gene families correspond well with a VH region classification based on similarities at the protein level (Dildrop 1984) . Understanding V H gene relatedness on the nucleic acid sequence level has greatly facilitated studies regarding the expression of different V H gene families during ontogeny (Yancopoulos et al. 1984 , Perlmutter et al. 1985 and in response to foreign and self antigens (Manser et al. 1987b , Kofler et al. 1987a ). These studies have thus provided an important insight into B-cell repertoire generation. V k classifications reported to date are confined to the protein level. One attempt to systematically classify V l proteins was based on the partial amino acid sequence up to the invariant cysteine in position 23 (Cys23), leading to 26 V k subgroups, designated VkCyS (Potter 1977) . A modified classification, based on the length and similarity of the amino termini up to the invariant tryptophan 35 (Trp35) of 79 V k proteins, was introduced in 1982 (VkTrp subgroups; Potter et al. 1982) . Four of the VkCyS subgroups were condensed and two new groups were added, resulting in a total of 24 V k subgroups, six of which are still defined only by sequences up to Cys23. This classification has now been generally accepted and, although an extended comparison at the nucleic acid level has never been reported, the corresponding V k protein subgroups have been widely used synonymously with V k gene families. More recently, we have performed a detailed restriction fragment length polymorphism (RFLP) analysis with DNA probes corresponding to 16 V k protein subgroups, and obtained evidence that such protein groups may not necessarily correspond to gene families analogous to those described for V n genes (Kofler et al. 1989) . Since a large number of full-length V k nucleic acid sequences has been reported, it is now possible to address, by direct sequence comparison, the matter of whether V k genes can be organized into gene families, as has been accomplished with V n genes, and how such V k gene families relate to the existing V k protein groups. This issue is of considerable interest for V k gene usage determinations, repertoire estimates, genomic mapping, and similar studies using nucleic acid hybridization, since such procedures depend on relatedness between V k groups, gene families, and corresponding DNA probes. We compiled 248 full-length V k nucleic acid sequences from the literature and several databases, and assigned them to existing V k protein classifications with subsequent grouping into gene families comprised of members with >80% overall nucleic acid sequence similarity. Our analysis revealed that the current classification in V k protein groups or subgroups frequently did not reflect relatedness on the nucleic acid sequence level. Furthermore, V k gene family organization differed in important aspects from that of V n gene families; only some of the V k gene families were clearly separated by sequence dissimilarity of > 25 %, as is usually observed in V~/gene families. The remaining families were more similar to each other and, in several instances, large portions of genes from different families shared > 80 % of their sequences, leading to cross-hybridization between those families in hybridization analyses. In addition, although ancillary to the primary aim of this study, we reviewed the specificities encoded by the various V k gene families and estimated their germline gene complexity. V k nucleic acid sequence bank. A database was constructed consisting of V k nucleic acid sequences from the Genetic Sequence Data Bank (GenBank, Los Alamos, New Mexico), E. A. Kabat's collection , and other publications. Only sequences encoding the entire mature V k protein were included in the database. If applicable, sequence portions encoding untranslated region, leader sequence, introns, or J segments were removed prior to comparisons. This primary database of 248 full-length V k sequences was then condensed to a final database of 109 ( Fig. 1 ) by deleting duplicate sequences and those differing by only 1 to 4 base pairs (bp). V k protein groups and subgroups. All nucleic acid sequences were translated into amino acids and organized into V k protein groups and subgroups. Assignment to V k protein groups (labeled I to VII) was based on the length of the amino-terminal sequence up to the invariant Trp35 (41,40, 39, 36, 35, 34, and 33 residues, respectively; Kabat et al. 1987) . Organization into V k protein subgroups was based on < 13 substitutions up to Trp35 (VkTr p subgroups; Potter et al. 1982) . Sequences meeting assignment criteria for more than one subgroup were assigned to the subgroup with the best match. V k gene families. Analogous to V n gene families, we defined a "V k gene family" as a group of nucleic acid sequences that exhibit > 80% overall sequence similarity with every member of this family, and < 80 % with V~ genes from other families. In nucleic acid hybridization analyses under defined stringency conditions (Brodeur and Riblet 1984) , all members of a gene family can be expected to cross-hybridize with each other. The V k gene family nomenclature proposed in this study was adjusted as far as possible to that used for V k protein subgroups, in order to minimize confusion in the literature; when V k protein subgroups and V k gene familes corresponded to each other (e.g., Vk21), the V k subgroup designation was used for the V k gene family as well. V k gene families comprising two or more V k protein subgroups were given the designation of the respective subgroups (e. g., the Vk4/5 gene family comprised Vk4 and Vk5 protein subgroups). Addition of capital letters to the designation indicates that a V k protein subgroup included members from two distinct V k gene families (e. g., the Vk9 protein subgroup comprised members from two distinct V k gene families, termed Vg9A and Vt9B, respectively). VeRF and (tentatively) Vk38C were two new gene families that could not be related unambiguously to any V k protein subgroup and, hence, were named after a prototypic sequence. The major goal of this study was to investigate the organization of mouse V k genes in terms of nucleic acid sequence similarity, and to determine the relationship of such organization to existing V k protein classifications. To this end, we first compiled 109 distinct (i. e., >4 bp different), full-length V k nucleic acid sequences that were used as a database for subsequent analyses (Fig. 1) . The sequences were translated into amino acids (Fig. 2 ) and assigned to protein groups and subgroups (Table 1) . Classification into protein groups was based on the number of residues up to the invariant Trp35 and, hence, was unambiguous in all instances. However, this classification was of limited practical value, since it frequently did not reflect structural relatedness (i. e., sequence similarity) between V k sequences. For example, group V included members of several, sometimes quite dissimilar, V k gene families (V23, Vk12/13, VkRF, Vk11, Vk9A, Vk9B, VklO, Ve38C, Vk19/28 i0 20 30 40 50 60 70 80 90 AACATTGTGCTGACCCAATCTCCAGCTTCTTTC, GCTGTGTCTCTAGGGCAGAGGGCCACCATATCCTGC AGAGCCAGTGAAAGTGTTGATAGT. GACATTGTGCTGACCCAATCTCCAGCTTCTTTGGCTGTGTCTCTAGGGCAGAGGGCCACCATATCCTGC GACATTGTGCTGACACAGTCTCCTGCTTCCTTAGCTGTATCTCT~AGAGGGCCACCATCTCATGC GACATTGTGCTAACACAGTCTCCTGCTTCCTTAGCTGTATCTCTGGGGCAGA~CACCATCTCATGC GACATTGTGCTGACCCAATCTCCAGCTTCTTTGGCTGTGTCTCTAGGACAGAGAGCCACTATCTTCTGC GACATTGTGCTGACCCAATCTCCAGGATCTTTGGCTGTGTCTCTAGGGCAGA~CACCATATCCTGC AAAATTGTGCTGACCCAATTTCCAGCTTCTTTGGCTGTGTCTCTAAGGCAGAGGGCCACCATATCCTC, C AGAGCCAGTGAAAGTGTTGATAGT, AGGGCCAGCAAAAGTGTCAGTACA. AGGGCCAGCCAAAGTGTCAGTACA. AGAGCCAGCCAGAGTGTCGATTAT. AGAGCCAGTGAAAGTGTTGAAAGT. AGAGCCAGTGAAAGTGTTGATAGT. TTCAGTGGCAGTGGGTCTAGGACAGACTTCACCCTCACCATTGATCCTGTGGAGGC TGATGATGCTGCAACCTAT TACTGT 002 TTCAGTGGCAGT~TCTAGGACAGACTTCACCCTCACCATTAATCCTGTGGAGGC TGATGATGTTGCAACCTAT TACTGT 003 TTCAGTGGCAGTGGGTCTGGGACAGACTTCACCCTCAACATCCATCCTGTGGAGGAGGAGGATGCTGCAACCTATTAC TGT 004 TTCAGTC, GCAGTGGGTCTGC, GACAGACT TCACCCTCAACATCCATCCTGTGGAGGAGGAGGATACTGCAACATATTACTGT 005 TTCAGTGGCAGTGC, GTCTGC.4~ACAGACTTCACCCTCAACATCCATCCTGTGGAGGAGGAAGATGCTGCAACC TATTACTGT 006 TTCAGTGGCAGTGC4~TCTATGACAGAC T TCACCCTCACCATTAATCC TGTGGAGGCTGATGATGT TGCAACATATTACTGT 007 TTCAGTGGCAGTGGGTCTAGGACAGACTTCACCCTCACCATTGATCCTGTGGAGGCTGATGATGGTGCAACCTATTACTGT 008 TTTAGTGC:K:AGTGGGTCTGGGACAGACT TCAGCCTCAACATCCATCC TGTGGAGGAGGATGATATTGCAGTGTATTTCTGT 009 TTCAGT~AGTGGGTCTGGGACAGACT TCACCC TCAACATCCATCC TGTGGAGGAGGAGGATGCTGCAACCTAT TACTGT 010 TTTAGTGGCAGTGGATCAGGGACAGATTT TACTC TTAGCATCAACAGTGTGGAGTCTGAAGATATTGCAGATTAT TACTGT 011 TTTAGTGGCAGTGGATCAGGGACAGATTT TACTCCTAGCATCAACAGTGTGGAGTCTGAAGATATTGCAGAATAT TACTGT 012 TTTAGT~AGTGGATCAGGGACAGAC TT TACTCTTAGCATCAACAGTGTGGAGTCTGAAGATGTTGCAGATTAT TACTGT 013 TTCAGTGGCAGTGGATCAGGGACAGATTTCACTCTCAGTATCAACAGTGTGGAGACTGAAGATTTTGGAATGTAT TTCTGT 014 TTCAGTGGCAGTGGATCAGGGACAGATT TCACTCTCAGTATCAACAGTGTGGAGACTGAAGATTTTGGAATGTAT T TCTGT 015 TTCAGTGGCAGTGGATCAGGGACAGAT T TCNCTC TCAT TATCAACAATG TGGAGACTGAAGATTTTGGAATGTATT TCTGT 016 T TCAGTGGCAGTGGATCAGGGACAGAT T TCACTCTCAGTATCAACAGTGTGGAGAC TGAAGATTTTGGAATGTATT TCTGT 017 TTCAGTGGCAGTGGGTCTGGGACCTCT TATTCTC 210 220 230 240 250 260 270 280 055 TTCAGTGGNAGTGGGNCTGGNAACTCTTACTCTCTCACGATCAGCAGCATGGAGGC NGAAGATGT TGCCACT TATTACTGT 056 TTCAGTGGNAGTGGGTCT~ACC TCTTACTCTCTCACAATCAGCAGAGTGGAGGCTGAAGATGCTGCCACTTATTACTGC 057 TTCAGTGGAAGTGGGTCTGGGACCTCTTAC TCTCTCACAATCAACAGAGTGGAGGCTGAAGATC, C TGCCACTTAT TAC TGC 058 TTCAGTGC4CAGTGGATCAGGAACACAATATTCTCTCAAGATCAACAGCCTGCAGCCTGAAGATTT TGGGAGT TATTACTGT 059 TTCAGTGC-,CAGTGGATCAGGAACACAATATTC TC TCAAGATCAACAGCCAC, CAGCCGGAGGATT T TGGGAGTTATTACTGT 060 TTCAGTGC4:AGTGGATCAC, C, CACACAGTTTTCTCTGAAGATCAACAGCCTGCAGCCTGAAGATTT TGGGAGTTAT TAC TGT 061 TTCAGTGGCAGTGGATCAGC4:ACACAGT T TTCTC TGAAGATCAACAGCCTC, CAGCCTGAAGATTT TGGGAGTTATTACTGT 062 TTCAGTGC4•AGTGGATCTGGTACAGATTTcACTCTCACCATCAGTAGCCTGGAGCCTGAAGATTTTGCAATGTATTACTGT 063 TTCAGTGGAACTGGATAT~ACAGATTTCACT TTCACCATCAGCAGCCAGGAGGAAGAAGATGTGTCAACT T ATTTCTGT 064 TTCAGTGGCAGTAGGTCTGGGTCAGATTATTCTCTCACCATCAGCAGCCTTGAGTCTGAAGATTTTGTAGACTATTACTGT 065 TTCAGT~AGTAGGTCTGGGTCAGATTATTCTCTCATTATCGGCAGCCTTGAGTCTGAAGATTTTGCAGACTATTACTGT 066 TTCAGTGGCAGTAGGTCTGGGTCAGATTATTCTC TCACCATCAGCAGCCT TGAGTCTGAAGATTT TGCAGAC TAT TACTGT 067 TTCAGTC, GCAGTGGGTCTGGGTCAGATTAT TCTCTCACCATCAC, CAGCCTAGAGTCTGAAGATTT TGCAGAC TATTACTGT 068 TTCAGTGC, CAGTGGATCTGGGCAAGATTATTCTCTCACCATCAGCAGCCTGGAGTATGAAGATATGGGAAT TTATTATTGT 069 TTCAGTGGCAGTGGATCTGC, CM:AAGAATAT TC TCTCACCATCAGCAGCC TGGAGTATGAAGATATGGGAAT T TATTTT TGT 070 TTCAGTC47,CAGTGGATCTGC4~NAAGATTATTCTCTAACCATCAGCAGCCTGGAGTCTGACGATACAGCAACTTATTAcTGT 071 TTCAGTGGCAGTGGGTCTGGAACAGAT T AT TC TC TCACCAT TAGCAACCTGGAGCAAGAAGATAT TGCCACTTACTTTTGC 072 TTCAGT~AGTGGGTCTGGAACAGATTATTCTTTCACCATTAACAACCTGGAGTAAGAAGATGTCGCCACTTATTCTTGA 073 TTCAGTGGAAGTGGGTCTGGGAGAGAT TATTCCT TCAGCATCAC, CAACCTGGAGCCTGAAGATAT TGCAACT TAT TATTGT 074 TTCAGTGGAAGTGGGTCT~AGAGATTAT TCC TTCAGCATCAGCAACCTGGACGCGGAAGAGATTGCAAC T TATTATTGT 075 TTTAGTGGCAGTGGGTCAGGAACAGATT TCACCC TGGAAATCAGTAGAGTGAAGGCTGAGGATGTGGGTGTGTATTACTGT 076 TTCAGTGGCAGTGGGTCAGGAACTGATTTCACACTGAGAATCAGTAGAGTGGAGC4ZTGAGGATGTGGGTGT TTATT ACTGT 077 TTCAGTAGCAGTGGGTCAGGAACCGAcTTCACACTGAGAATCAGCAGAGTGGAGGCTGAGGATGTGGGTGTTTATTACTGT 078 TTCAGTGGCAGTGGGTCAGGAACTGCTTTCACACTGAGAATCAGTAGAGTGGAGGCTGAGGATGTGGGTATTTAT T ATTGT 079 TTCAGTC,GCAGTGGGTCAGGAACTGCTTTCACACTGAGAATCAG~AGAGTGGAGGCTGAGGATGTGGGTATTTATTATTGT 080 TTCAGTGGCAGTGGGTCAGGAACTGCTTTCACACTGAGAATCAGTAGAGTGGAGGCTGAGGATGTGGGTGTTTATTACTGT 081 T TCAGTGC, CAGTGAGTCAGGAACTGATT TCACACTGAGAATCAGCAGAGTGGAGGCTGAGGATGTGGGTGTT TATTAC TGT 082 T TCAGTGGCAGTGGATCAGGGACAGAT TTCACACTCAAGATCAGCAGAGTGGAGGCTGAGGATCTGGGAGT TTAT T TCTGC 083 TTCAGTGGCAGTGGATCAGGGACAGATTTCACACTCAAGATCAGCAGAGTGGAGGC TGAGGATCTGGGAGTT TATTAC TGC 084 TTCAGTGGTAGTGGATCAGGGACAGATTTCACACTGAAAATCAGCAGAGTGGAGGCTGAGGATT TGGGAGTT TATTTCTGC 085 TTCAGTGGCAGTGGT TCA~ACAGAT TTCACACTCAAGATCAACACAATAAAGCCTGAGGAC TT~AATGTAT TACTGC 086 TTCAGTC, Cg=AGTGGATCAGGGACAGATT TCACACTCAATATCAGCAGAGTGGAGGC TGAGGATATGGGAGT T TATTACTGC 087 TTCAGTGGTAGTGGATCAGGGACAGAT TTCACACTGAAAATCAGCAGAGTGGAGGCTGAGGATT TGGGAGTT TATTTC TGC 088 TTCAGT~AGTGGATCAGGGACAGATTTCACACTCAAGATCAGCAGAGTGGAGGCTGAGGATGT AGGAAT T T ATTACTGT 089 TTCACTGGCAGTGGATCAGGGACAGATTTCACACTGAAAATCAGCAGAGTGGAGGCTGAGGATTTGGGAGTTTATTATTGC 090 TTCATA~AGTGGATCTGGGACAGATTTCAC TC TTACCATCAGCAGTGTGCAGGCTGAAGACCTGGCAGATTACTTCTGT 091 T TCACAGGCAGTGGATCTGGGACAGATTTCACTCTCACCATCAC, CAGTGTGAAGGCTGAAGACC TGGCAGTT TATTAC TGT 092 TTCACAGGCAGTGGATCTGGAACAGATTTCACTCTCACCATCAGCAGTGTGCAGGC TGAAGACCT GGCAGT T TATTAC TGT 093 TTCACAC, C, CAGTGGATCTGGAACCGATTTCACTCTTACCATCAGCAGTGTGCAGGCTGAAGACCTGGCAC T TTAT TACTGT 094 TTCACAC,GCAGTGGATCTGGGACAGATTTTACTCTTACCATCAGCAGTGTACAAGCTGAAGACCTGGCAGTTTATTACTGT 095 TTCACGGGCAGTGGATCTGC43ACAGATT TCACTCTCACCATCAGCAGTGTGC~CTGAAGACCTGGCAGT T TATTAC TGC 096 TTCACAGGAAGTGGTTCT~AGNGATTATACTCTCACAGTCAGCAGTGTGAAGGCTGAAGACCTGGCACTTTACTACTGT 097 TTCACAGGCAGTGGATCTGGGACAGATTTCACTCTGACCATCAGCAGTGTACAGGTTGAAGACCTCACACATTATTACTGT 098 TTCACTGGCAGTGGATATGGGACGGATTTCACTTTCACCATCAGCACTGTC, CAGGCTGAAGACCTGGCAG T T TAT T TCTGT 099 TTCACTGGCAGTGGATCTGGGACAGATTTCACTTTCACCATCAC•CAGTGTGCAGGTTGAAGACCTGGCAGTTTATTTCTGT I00 TTCACAGGCAGTGGATCT~ACAGATTTCACTCTCACCATTAGCAATGTGCAGTCTGATGACTTGGCAGATTATTTCTGT i01 TTCACAGGCAGTGGATCTGGGACAGAT TTCACTCTCACCATCAGCAATATGCAGTCTGAAGACCTGGCAGAT TATTTCTGC 102 TTCTCAGGCAGTGGATCTGC4~ACAGATTTCA~TCTCACCATCAGCAATGTGCAGTCTGGAGACTTGGCAGAGTATTTCTGT 103 TTCACAGGCAGTGGATCTGGGACAGATTTCACTCTCACCATTACCAATGTGCAATCTGAAGACCTGGCAGATTATTTCTGT 104 TTCACAGGCAGTGGATCTGGGACAGATTTCACTCTCACCATCAGCAATGTGCAGTCTGAAGACTTGGCAGAGTATTTCTGT 105 TTCAcAGGCAGTGGATCTGGGACGGATTTCACTCTCACCATcAGCAATGTGcAGTCTGAAGACTTGGCAGTGTATTTCTGT 106 TTCACTGGCAGTGGATCTGGGACGGATTTCACTTTCACCATCAGCAGTGTGCA~TGAAGACCTGGCAGTTTATTACTGT 107 TTCACTGC4=AGCGGATCT~ACGGAT TTCACTTTCACCATCAGCAGTGTGCAC, C4~TGAAGACC TGC4CAGTT TATT AC TGT 108 TTCACAGGCAGTGGATCTGCAACAGATTTCACTCTGACCATCAGCAGTGTGCA~TGAAGACCTTGCAGAT TATCAC TGT 109 TTCACAGC4 Organization of V k proteins into subgroups using < 13 mismatches up to Trp35 as a criterium (Potter et al. 1982) better reflected primary structure similarities, although such organization frequently led to multiple assignments, in which cases only a single assignment for the sequence representing the best match was included (Table 1) . Moreover, as will be shown below, this classification repeatedly failed to adequately reflect overall similarity at the nucleic acid sequence level. Finally, some sequences (discussed below) could not be assigned unambiguously to any existing VkTrp subgroup. We then determined whether V k nucleic acid sequences could be organized into gene families (analogous to V H genes), and how such families related to V k protein groups and subgroups. For this purpose, all V k genes in the data bank were arranged in groups of > 80 % sequence similarity, which were termed V k gene families. The characteristics of these families and their relationship to Vk protein groups and subgroups are detailed below. A quick summary outlining how the different classifications correspond to each other is presented in Table 2 . Vk21 genefamily. All Vk21 genes fulfilled the criteria for a typical V gene family, i.e., all members were >80% similar (mostly >90%) and differed from all other V k sequences by at least 25 %. This gene family corresponded completely to protein subgroup Vk21 which, in turn, coincided with V k protein group III. Five germline genes have been cloned (Heinrich et al. 1984 ) and approximately ten expressed sequences have been published. Vk21 genes were used in response to influenza hemagglutinin (Clarke et al. 1985 , Meek et al. 1989 ) and major histocompatibility complex class II antigens (Devaux et al. 1985) , and encoded some lupus-associated autoantibodies (Shlomchik et al. 1987c ). The P3-X63-Ag8.653 myeloma line, a derivative of the MOPC21 myeloma that has lost the ability to express Ig heavy and light chain proteins and is frequently used in hybridoma technology (Kearney et al. 1979) , also expressed a non-functional Vk21 mRNA (Strohal et al. 1987) . With the exception of an MRLlpr/lpr rheumatoid factor (RF, anti-Ig) V k sequence (AM12; Shlomchik et al. 1987c ), which differed from all known Vk21 germline genes by > 30 bp and may have derived from an unknown germline gene, all other expressed sequences were very similar to, and hence probably derived from, known V~21 germline genes. RFLP (Kofler et al. 1989 ) and gene cloning analyses (Heinrich et al. 1984) suggested an estimated 6 to 13 Vk21 germline genes in the genome of most inbred strains of mice. Finally, an incomplete Vk21 sequence (VM201, Meek et al. 1989) , which was therefore not included in our data bank, should be mentioned as it lacked two codons in CDR-1 in comparison to other Vk21 sequences. Unless caused by somatic events, this would make the corresponding germline gene the only Vk gene with 37 codons up to Trp35. V23 genefamily. Similar to Vk21, Vk23 sequences were well separated from all other V k sequences, and formed a gene family that corresponded entirely to its protein counterpart, the Vk23 subgroup (protein group V). One germline gene has been reported (Pech et al. 1981 ) that was subsequently observed in RFs from BALB/c mice (Shlomchik et al. 1987a) , and that probably encoded an (NZB • NZW)F 1 RNA-speciflc autoantibody (Eilat et al. 1988) . Additional Vk23 genes, more distant from the above germline gene but closely related to each other, possibly derived from a second V23 germline gene and encoded nitrophenyl-specific anti-idiotypes (Sablitzky and Rajewsky 1984) and a creatine-kinase-specific antibody (Buckel et al. 1987) . A nonfunctional Vk23 member was cloned from an MRL/n RF-producing hybridoma and might correspond to another Vk23 (pseudo) gene (Kofler et al. 1989 ). Our previous RFLP analyses suggested the presence of four to eight V~23 germline genes in the genome of most inbred strains of mice. However, this may represent an over-estimate due to cross-hybridization of the more conserved 3' portion of the Vk23 probe with Vkl sequences (Kofler et al. 1989 , and below). Kabat et al. 1987 . + Codes of V k genes are given in legend to Figure 1 . * V~, V, gene family (this report); Group, V k protein groups ; .. Subgroup, VkTrp35 subgroups (Potter et al. 1982 (Heinrich et al. 1984) ; 2, (Clarke et al. 1985) ; 3, (Shlomchik et al. 1987c ); 4, (Strohal et al. 1987) ; 5, (Pech et al. 1981) ; 6, (Altenburger et al. 1980) ; 7, (Eilat et al. 1988 ); 8, (Kofler et al. 1989 ); 9, (Sablitzky and Rajewsky 1984) ; 10, (Buckel et al. 1987) ; 11, (Even et al. 1985) ; 12, (H6chtl et al. 1982) ; 13, (Kofler et al. 1987b); 14, (Kofler et al. 1988) ; 15, ; 16, ; 17, (Shlomchik et al. 1987a) ; 18, (Nahmias et al. 1988 ); 19, ; 20, (Parslow et al. 1984) ; 21, (Liu 1987) ; 22, (Kaartinen et al. 1983 ); 23, (Griffiths et al. 1984) ; 24, (Seidman et al. 1978) ; 25, (Kelley et al. 1985) ; 26, (Seidman et al. 1979) ; 27, (Max et al. 1980) ; 28, (Darsley and Rees 1985) ; 29, (Near and Haber 1989) ; 30, (Reininger et al. 1987) ; 31, ; 32, (Campbell 1987) ; 33, (Meek et al. 1989) ; 34, (Selsing and Storb 1981) ; 35, (Joho et al. 1984) ; 36, (Lutz and Davie 1988) ; 37, (Corbet et al. 1987) ; 38, (Borden and Kabat 1987); 39, (Schiffet al. 1983 ); 40, (Oilier et al. 1985) ; 41, (Riley et al. 1986 ); 42, (Kwan et al. 1981) ; 43, (Boyd et al. 1986 ); 44, (Cabilly et al. 1984) ; 45, (Hawley et al. 1982) ; 46, (Sahagan 1986 ); 47, (Beidler et al. 1988 ). (Glu 6~ Ala 3~ Tyr 10) system (Oilier et al. 1985) . In RFLP analyses, two strongly and several weakly hybridizing restriction fragments were observed (Kofler et al. 1989) . Whether the latter corresponded to additional, more distant, Vk12/13 germline genes or are due to high similarity (>80%) in portions of the probe with other V~ genes (particularly those of Vk gene families 9A, 9B, 10, and 11) remains to be determined. I(~RF gene family. The MRL-RF24 Vk protein (Kofler et al. 1987b ), a member of the large protein group V, had 12 mismatches up to Trp35 from two Vk12-13 proteins (K2 and MOPC129), but differed from the remaining Vk12-13 proteins (and all other V k proteins) by > 12 residues. Thus, this protein could not be unambiguously assigned to known Vk subgroups. Its nucleic acid sequence differed from all V k sequences by > 25 %, thus forming a distinct V k gene family, termed Vf~F. Used as a probe, this gene identified a single restriction fragment that was absent in haplotype Igk f (Kofler et al. 1989 ). The corresponding (as yet uncloned) germline gene probably also encoded a BALB/c (Bruck et al. 1986 ) and a C57BL/6 (Sablitzky and Rajewsky 1984) idiotypespecific antibody, as well as an (NZB • NZW)F~ DNAspecific autoantibody (Eilat et al. 1988 ). Vkll, 9A, 9B, 10, and 38C gene families. The V k gene families discussed thus far were clearly separated from all other V k genes by >25% overall sequence dissimilarity and in this respect resembled V n gene families. The following five gene families, distantly related to Vk12/13 and Vf, F, were less well separated from one another. V~ll gene family. For this gene family with four to six germline genes by RFLP analysis (Kofler et al. 1989 ), a single nucleic acid sequence corresponding to a nonfunctional rearrangement from an NZB myeloma (Kelley et al. 1985) was present in the data bank. This sequence fulfilled protein assignment criteria for V k protein subgroups 9, 10, and 11; however, it best matched Vkll proteins. Comparisons with the entire data bank (including some Vk9 and Vkl0 sequences) revealed matches of only 76 % or less at the nucleic acid level, making this sequence the prototype for the Vk11 gene family. Vkll proteins were observed in the beta 2, 1 fructosan response . Vk9A gene family. The Vk9 protein subgroup, another member of the large protein group V (Potter et al. 1982) , comprised sequences that, at the nucleic acid level, fell into two distinct gene families, termed Vk9A and V~9B. The Vk9A gene family included two germline genes (Seidman et al. 1979 , Max et al. 1980 , one of which may be expressed in hen egg lysozyme antibodies (Darsley and Rees 1985) . Another expressed Vk9A gene from an NZB x NZW F~ anti-DNA IgM ) was only 88% similar to the other germline gene and probably derived from an unknown Vk9A germline gene. In addition, Vk9A genes have been observed in GAT-idiotypespecific antibodies (Oilier et al. 1985) . Vk9B gene family. The T1 sequence and its germline counterpart, V-L6 (Pech et al. 1981) , both assigned to the Vk9 protein subgroup (Potter et al. 1982) , differed from Vk9A (and all other Vk) nucleic acid sequences by > 20 % and, hence, formed a separate family, termed Vk9B. Genes from this family encoded antibodies specific for digoxin Margolies 1987, Near and Haber 1989) and Escherichia coli (Pennell et al. 1988) , and bromelain-treated red blood cell autoantibodies from lupus and normal mice (Reininger et al. 1987 ). VklO gene family. This family corresponded to the Vkl0 subgroup (protein group V). RFLP data suggested two to three VklO germline genes (Kofler et al. 1989) , one of which has been cloned Capra 1987, Wysocki et al. 1987 ) and probably encoded arsonate (Manser et al. 1987a , Meek et al. 1987 , oxazolone , oligosaccharide (Matsuda and Kabat 1989) , bromelain-treated red blood cell (Pennell et al. 1988 ) and RF-like (Shlomchik et al. 1987c ) antibody responses. A more distant Vkl0 sequence, with multiple in-frame stop codons, has been observed as a nonfunctional allele of an NZB myeloma (Kelley et al. 1985) , and might correspond to one of the uncloned VklO germline genes. sequences encoding the 38C13 lymphoma (Campbell 1987 ) and the VMll3 anti-hemagglutinin hybridoma (Meek et al. 1989 ) light chains, respectively, were > 20 % different from all other V k nucleic acid sequences in the database and, hence, could not be assigned to any Vk gene family; the closest matches (77-78 %) were observed with a VklO germline gene Capra 1987, Wysocki et al. 1987 ). At the amino acid level, members of four VkTrp subgroups (Vk9, Vkl0, Vkl 1, and Vk12/13) exhibited equally distant relatedness (nine and more residues difference in the N-terminal 35 amino acids), making unambiguous assignment at the protein level impossible. Whether these sequences were the representatives of a new V k gene family or corresponded to highly mutated (VklO) genes remains to be determined. Vk24/25, Vgl, and Vk2 gene families. The next three families were grouped together based on sequence similarity of up to 78% between V24/25 members and Vkl and Vk2 genes, respectively, and because the overall similarity between Vk2 and some Vkl genes exceeded 80%. The latter observation, i.e., similarity of > 80% between some, but not all, members of two gene families, obviously constitutes a problem in this type of V k gene classification (see below). Vk24/25 gene family. Originally, only a single Vk24 germline gene (involved in the phosphocholine response; Malipiero et al. 1987, Gearhart and Bogenhagen 1983) had been reported (Selsing and Storb 1981) . Other investigators have cloned this, a related pseudogene, and two additional Vk24 germline genes (Joho et al. 1984) . The latter were only about 82-83% similar to the Vk24 prototype and may have encoded Streptococcus group A carbohydrate antibody light chains previously assigned to the Vk25 subgroup (Lutz and Davie 1988) . Hence, these two VkTrp subgroups (protein group II) were probably encoded by distant members of a single V k gene family. In addition to the four cloned Vk24/25 germline genes, evidence was obtained for the presence of at least two more germline genes in this family: firstly, RFs from autoimmune and normal mice (Shlomchik et al. 1987a (Shlomchik et al. , 1987c ) has yet to be isolated. (already previously condensed with Cys23 subgroups Vk3 and Vk26; Potter et al. 1982) and Vk2 were encoded by sequences that, using a stringent family definition, precluded classification into either a single, or two distinct, gene families; all Vkl nucleic acid sequences were > 80% similar, yet the three almost identical Vk2 nucleic acid sequences reported (Akolkar et al. 1987 , Panka et al. 1988 ) shared up to 81.7% similarity with some, but only about 75 % with other, Vkl members. Moreover, sequence similarity in the 3' portion of several Vkl and V~2 genes was around 90%. These two "gene families" were, therefore, partially overlapping. However, for reasons of clarity, we have retained them as separate Vk gene families. Three V~I germline genes (Corbet et al. 1987 ) and approximately 40 expressed Vkl sequences have been reported. With the exception of an anti-dextran V k gene (W3129; Borden and Kabat 1987) with > 15 % differences from any known V~I gene, all expressed sequences were highly homologous to one of the above germline genes, suggesting that the total Vkl germline gene number may not exceed four. A more direct complexity estimate in our previous RFLP analysis was hampered by crosshybridization of the Vkl probe to non-Vkl genes due to > 80% sequence similarity in the 3' region of V~I and other V k genes (see below and Kofler et al. 1989) . Vkl genes were used in a variety of responses to foreign and self antigens (reviewed by Schiff et al. 1988 , Kofler et al. 1987a . V#2 germline genes have not yet been reported; the three expressed sequences encoded antibodies to dextran (Akolkar et al. 1987) , digoxin (Panka et al. 1988) , and DNA ). Vk8, Vk22, and Vfl9/28 gene families. The following three gene families were separated from each other by >20%, and from all other Vk genes by >25%, overall sequence similarity; however, large portions (codons 35 to 94) of their genes had between 80 % and 89 % common nucleotides, leading to extensive cross-hybridizations (Kofler et al. 1989 ). Vk8 (protein group I) were around 90% similar and shared up to 78% of their nucleotides with VJ9/28 and Vk22 genes. Similarity in codons 35-94 was even higher, reaching 87 % with V28 genes. The complexity of this gene family was difficult to assess by RFLP analyses due to possible cross-hybridization, however, at least half of the 13-20 fragments hybridizing to a Vk8 probe probably belonged to this large family (Kofler et al. 1989) . Vk8 genes encoded antibodies to phosphocholine (Malipiero et al. 1987) , dinitrophenyl (Riley et al. 1986) , and hen egg lysozyme (Darsley and Rees 1985) , as well as RF-like (Shlomchik et al. 1987a (Shlomchik et al. , 1987c and DNA-specific (Eilat et al. 1988) autoantibodies. E~22 gene family. The only two, almost identical, Vk22 (protein group I) sequences available for comparison, S107A (Kwan et al. 1981 ) and HPCA97 (Berek 1984) , revealed between 80% and 89% similarity with a large portion (codons 35 to 94) of all ~19/28 genes. The remaining nucleotides were, however, only < 70 % similar, resulting in an overall similarity of 72%-75%, thus refuting assignment of F~22 and Fk19/28 genes to a common gene family. Similarity with ~8 genes was in the range of 75%-77% and mismatches were distributed evenly over the entire gene. RFLP analyses suggested one to two I/~22 germline genes; additional weak restriction fragments hybridizing to a V~22 probe on Southern blots probably corresponded to genes from the ~19/28 and ~8 families (Kofler et al. 1989 ). F~22 genes encoded phosphocholine antibodies (Malipiero et al. 1987 by > 25 %. Thus, they were combined to a single ~ gene family, which was termed ~19/28. However, this F~ gene family (like some other F~ gene families, see below) behaved atypically in nucleic acid hybridization studies as compared to Fn gene families: different DNA probes from this family, i.e., a ~19 and a ~28 probe, did not hybridize to an identical, but to an overlapping, set of restriction fragments (Kofler et al. 1989 ). This could be explained by cross-hybridization of the 1,128, but not the ~19, probe with ~8 genes. RFLP data suggested four to six ~19/28 germline genes (Kofler et al. 1989 ), one of which, a F~28 germline gene, also known as VkSer, from haplotypes Igk-VSer ~, lgk-VSer b, lgk-VSe/, and Igk-VSer d, has been cloned (Boyd et al. 1986 , Ponath et al. 1989 . F~19/28 genes encoded antibodies to trinitrophenyl (Hawley et al. 1982) , carcinoembryonic antigen (Cabilly et al. 1984 , Beidler et al. 1988 , human breast/lung/colon cancer cells (Sahagan 1986) , influenza hemagglutinin (Meek et al. 1989) , and an RNA-specific (Eilat et al. 1988 ) and some RF-like autoantibodies (Kofler et al. 1989 , Shlomchik et al. 1987a , 1987c . Figure 3 shows the relatedness between different ~ gene families as reflected by overall nucleic acid sequence similarity. A significant difference from 1/tt gene families was apparent, since the latter are generally more distantly Fig. 3 . Sequence similarity between different ~ gene families; comparison of known germline genes and derivatives of putative germline genes (i. e., sequences differing from known germline genes by > 10% and primarily of the IgM isotype). Indicated are the highest and lowest percentages of nucleic acid sequence similarity between members from two families; single percentage resulted from comparisons yielding identical percentages. Shading intensities highlight increased overall similarity between the respective families. related by sequence similarity. Obviously, if members from different families are only a few percent less similar than those from within a family, cross-hybridizations might occur, particularly if these differences are not evenly distributed over the entire sequence. As described above, large sequence portions with high degrees of similarity were indeed observed in genes from families 8, 22, and 19/28, and thus explain the previously observed cross-hybridizations between those families. Closer scrutiny of the similarities between portions of V k sequences from different families revealed that the 3' region (particularly codons 57-88, corresponding to frame work region 3) were generally more closely related than the remaining sequence, and this portion might precipitate unexpected cross-hybridizations, even between otherwise distant V k gene families. For example, ~10 and Vk9A genes had a [35 bp 3' sequence with 83 % similarity, and VkRF and Vk9B genes shared 84% of 103 nucleotides at the 3' end. As a further complication, different genes from a given family may exhibit more or less cross-hybridizations with genes from other families. Because of the differences in the organization of V n and V k genes, nucleic acid sequence hybridization assays with V k DNA probes require particular care in the selection of probes and in data interpretation. While in general any member of a V n gene family used as a probe will recognize its entire family, but will not cross-hybridize with other families, our previous RFLP analyses and the current study strongly suggest that V k probes may often behave differently. As a rule, probes devoid of the more "promiscuous" 3' sequences will be more specific; however, such probes may not always hybridize to all members of their gene families, and therefore require the use of two or more genes to probe the entire family. Another question addressed in this study regards the total number of V k genes in the genome of inbred mice. We estimated the complexity of known V k gene families by using RFLP criteria (Kofler et al. 1989 ) and by taking into account expressed and germline genes identified for each family. Regarding expressed sequences, we assumed that [gM sequences with > 6, and IgG sequences with > 30 mismatches from known germline genes may have derived from as yet unknown germline genes. Allelic differences were also considered, however this was a minor concern as the majority of sequences in the database (91/109) derived from the same haplotype (IgU). This approach led to a total of about 70-140 genes (Table 3) . Obviously, such estimates need to be taken with caution due to the peculiarities of Vk gene probes discussed above, and to inherent limitations of the RFLP technique (discussed by Kofler et al. 1989 ). Furthermore, * References to cloned V~ germline genes are given throughout the texti , The one-member VtRF family is deleted in haplotype Igk f mice (Kofler et al. 1989 ). possible additional, as yet uncloned, V~ genes and gene families in the mouse genome have not been included. However, although evidence for some additional V k genes exists, their number might be limited. For two VkTr p subgroups, Vk27 (group I) and Vk20 (group VII), nucleic acid sequences have not been identified, but the corresponding Vk gene families may be small since only a single sequence for each subgroup has been reported to date. D'Hoostelaere published another novel V k gene family (pC9-26) with approximately six members as suggested by RFLP analyses (D'Hoostelaere et al. 1988 ), but whether or not this family related to either of the two subgroups above, or to Vk38C, is unknown. Nevertheless, the large number of responses to foreign and se~f antigens investigated at the nucleic acid sequence level and repeated isolation of identical sequences, suggest that the majority of the mouse V k germline repertoire might now be known. Different V L and V H germline genes are used to produce similar combining sites with specifiCity for alpha (1 --> 6) dextrans Regulation of genome rearrangement events during lymphocyte differentiation Functional and non-functional joining in immunoglobulin light chain genes of a mouse myeloma Cloning and high level expression of a chimeric antibody with specificity for human carcinoembryonic antigen The D segment defines the T15 idiotope: the immunoresponse of A/J mice to Pneumococcus pneumoniae Mutation drift and repertoire shift in the maturation of the immune response Molecular events during maturation of the immune response to oxazolone Activation of memory and virgin B cell clones in hyperimmune animals Nucleotide sequence of the cDNAs encoding the variable region heavy and light chains of a myeloma protein specific for the terminal nonreducing end of alpha (1-6) dextran Structural differences in a single gene encoding the VkSer group of light chains explain the existence of two mouse light-chain genetic markers The immunoglobulin heavy chain variable region (Igh-V) locus in the mouse. I. One hundred Igh-V genes comprise seven families of homologous genes Nucleic acid sequence of an internal image-bearing monoclonal anti-idiotype and its comparison to the sequence of the external antigen Cloning and nucleotide sequence of heavy-and light-chain cDNAs from creatine-kinase-specific monoclonal antibody Generation of antibody activity from immunoglobulin polypeptide chains produced in Escherichia coli Idiotype vaccination against murine B cell lymphoma. Humoral and cellular responses elicited by tumor-derived immunoglobulin M and its molecular subunits Inter-and intraclonal diversity in the antibody response to influenza hemagglutinin Two V k germline genes related to the GAT idiotypic network (Abl and Ab3/Abl') account for the major subfamilies of the mouse Vk-1 variability subgroup Sets of immunoglobulin V k genes homologous to ten cloned V k sequences: implications for the number of germline V k genes Nucleotide sequence of five antilysozyme monoclonal antibodies Preferential expression of Vk21E on IdX Ia.7 positive monoclonal anti-l-E antibodies The Igk L chain allelic groups among the Igk haplotypes and Igk crossover populations suggest a gene order A new classification of mouse V H sequences V region sequences of anti-DNA and anti-RNA autoantibodies from NZB/NZW F1 mice Light chain germ-line genes and the immune response to 2-phenyloxazolone Clusters of point mutations are found exclusively around rearranged antibody variable genes Somatic mutation and the maturation of immune response to 2-phenyloxazolone Mutant immunoglobulin genes have repetitive DNA elements inserted into their intervening sequences Somatic mutation creates diversity in the major group of mouse immunoglobulin kappa light chains Amino acids at the site of Vk-J k recombination not encoded by germline sequences Recombined flanks of the variable and joining segments of immunoglobulin genes Evolution of a multigene family of V k germ line genes Functional analogues of the VkOxl gene in different strains of mice: evolutionary conservation but diversity based on V-J joining mRNA sequences define an unusually restricted IgG response to 2-phenyloxazolone and its early diversification Sequences of Proteins of Immunological Interest, 4th edn, US Department of Health and Human Services A mouse myeloma cell line that has lost immunoglobulin expression but permits the construction of antibody-searching hybrid cell lines Nonproductive kappa immunoglobulin genes: recombinational abnormalities and other lesions affecting transcription, RNA processing, turnover, and translation A new murine Ig V n gene family The genetic origin of autoantibodies Molecular analysis of the murine lupus-associated anti-self response: involvement of a large number of heavy and light chain variable region genes Immunoglobulin k light chain variable region gene complex organization and immunoglobulin genes encoding anti-DNA autoantibodies in lupus mice Complexity, polymorphism and connectivity of mouse V k gene families Nucleic acid and protein sequences of phosphocholinebinding light chains Production of a mouse-human chimeric monoclonal antibody to CD20 with potent Fc-dependent biologic activity Chimeric mouse-human IgG1 antibody that can mediate lysis of cancer cells One heavy chain variable region gene segment subfamily in the BALB/c mouse contains 500-1000 or more members Genetics and primary structure of V k gene segments encoding antibody to Streptococcal group A carbohydrate. Comparison of V k gene structure with idiotope expression Somatic mutation in anti-phosphorylcholine antibodies Somatically mutated forms of a major anti-p-azophenylarsonate antibody variable region with drastically reduced affinity for p-azophenylarsonate. By-products of an antigen-drive immune response? Evolution of antibody variable region structure during the immune response Variable region cDNA sequences and antigen binding specificity of mouse monoclonal antibodies to isomaltosyl oligosaccharides coupled to proteins. T-dependent analogous of tx(1-->6)dextran Variation in the crossover point of kappa immunoglobulin gene V-J recombination: evidence from a cryptic gene Identity of the Vkl0-Ars-A gene segments of the A/J and BALB/c strains Nucleotide changes in sequential variants of influenza virus hemagglutinin genes and molecular structures of corresponding monoclonal antibodies specific for each variant The immune response toward beta-adrenergic ligands and their receptors. VIII. Extensive diversity of V H and V L genes encoding anti-alprenolol antibodies Characterization of the heavy and light chain immunoglobulin variable region genes used in a set of anti-digoxin antibodies Organization and complete sequence of identical embryonic and plasmacytoma kappa V-region genes The idiotypic network and the internal image: possible regulation of a germ-line network by paucigene encoded Ab2 (anti-idiotypic) antibodies in the GAT system Complete variable region sequences of five homologous high affinity anti-digoxin antibodies Variable region framework differences result in decreased or increased affinity of variant anti-digoxin antibodies Struc; ture of the 5' ends of immunoglobulin genes: a novel conserved sequence Differences between germ-line and rearranged immunoglobulin V k coding sequences suggest a localized mutation mechanism Restricted Ig variable region gene expression among Ly-1 + B cell lymphomas Developmentally controlled expression of immunoglobulin V H genes Structural and evolu~ tionary comparisons of four alleles of the mouse immunoglobulin kappa chain gene, lgk-VSer Antigen-binding myeloma proteins of mice Classification of mouse V k groups based on the partial amino acid sequence to the first invariant tryptophan: impact of 14 new sequences from IgG myeloma proteins Novel V genes encode virtually identical variable regions of six murine monoclonal anti-bromelain-treated red blood cell autoan~ tibodies A member of a new V n gene family encodes anti-brnmelinized mouse red blood cell autoantibodies Preferential expression of variable region heavy chain gene segments by predominant 2,4-dinitrophenyl-specific BALB/c neonatal antibody clonotypes Molecular basis of an isogeneic antiidiotypic response A genetically engineered murine/human chimeric an~ tibody retains specificity for human tumor-associated antigen V k and Jk gene segments of A/J Ars-A an~ tibodies: somatic recombination generates the essential arginine at the junction of the variable and joining regions Interstrain conservation of the murine GAT-specific antibody V k repertoire as analyzed at the germline gene level The Ig germline gene repertoire: economy or wastage? Multiple related immunoglobulin variable-region genes identified by cloning and sequence analysis A kappa-immunoglobulin gene is formed by site-specific recombination without further somatic mutation Somatic mutation of immunoglobulin lightchain variable-region genes Variable region sequences of murine IgM anti-IgG monoclonal autoan, tibodies (rheumatoid factors). II. Comparison of hybridomas derived by lipopolysaccharide stimulation and secondary protein immunization Structure and function of anti-DNA autoantibodies derived from a single autoimmune mouse The role of clonal selection and somatic mutation in autoimmunity Sequences of variable regions of hybridoma antibodies to alpha (1 -> 6) dextran in BALB/c and C57BL/6 mice Complete variable region sequence of a non-functionally rearranged kappa light chain transcribed in the non-secretor P3-X63-Ag8.653 myeloma cell line Somatic generation of antibody diversity Members of novel V n gene families are found in VDJ regions of polyclonally activated B-lymphocytes Single germline V H and V k genes encode predominating antibody variable regions elicited in strain A mice by immunization with p-azophenylarsonate Preferential utilization of the most Jn-proximal V n gene segments in pre-B-cell lines Acknowledgments. We thank Dr. G. Wick for critically reviewing, and Ms_ M. Kat Occhipinti for editing, this manuscript. The work reported herein was supported by a grant from the Austrian Research Council (Project S-41/06). R. Strohal is supported by a fellowship from the Emil Boral Stiftung and G. Kroemer receives an Erwin Schr6dinger Stipend (J0307M) from the Austrian Research Council.