You are on page 1of 5

Proc. Natl. Acad. Sci. USA Vol. 77, No. 2, pp.

1106-1110, February 1980

Immunology

Organization of K light chain genes in germ-line and somatic tissue
(antibody diversity/germ-line genes/restriction map)

ROLF JOHO*, IRVING L. WEISSMAN*, PHILIP EARLYt, JIM COLEt, AND LEROY HOODt
*Laboratory of Experimental Oncology, Department of Pathology, Stanford University, Stanford, California 94305; and tDivision of Biology, California Institute of Technology, Pasadena, California 91125

Communicated by H. N. Eisen, October 5, 1979

ABSTRACI We studied the organization of the K light chain genes in germ-line (sperm) and somatic (embryo) tissues. We constructed a plasmid containing a DNA insert coding for the K chain MOPC 167 and used the Southern blotting technique to determine the organization of K variable and constant region genes. In the haploid genome of the mouse there is only one constant region gene detectable and it has the same organization in sperm and embryo DNAs. There are several variable region genes in sperm and embryo that are related to the Vk167 gene. The organization of the V genes in sperm and embryo DNAs is identical. These results show that there is no rearrangement of variable region genes (or "minigenes") during early embryogenesis. The polypeptides that form antibody molecules can be divided into two-distinct parts, the variable (V) and the constant (C) region. From RNA-DNA hybridization studies, only a few copies of a particular C region seem to be present in the genome (1-11). The number of V region genes present in the germ line is still not resolved. Because the V regions determine the antibody specificity, the question of the origin of antibody diversity is an integral part of the number and organization of the V region genes. Several hypotheses have been proposed to explain the origin of antibody diversity. The strict germ-line hypothesis proposes that there are multiple V region genes and each antibody-producing cell chooses to express only one of these genes from the inherited V region repertoire (12). The various somatic diversification hypotheses explain the origin of antibody diversity quite differently. It is assumed that the germ line contains only a few V region genes and these are somatically diversified during ontogeny in order to yield many different V regions. A strict mutation hypothesis proposes that mutations (simple base changes) occur during differentiation of precursor cells into antibody-producing cells (13, 14). These mutations may occur in the hypervariable part of the V region or they may occur randomly throughout the V region gene. The somatic recombination hypothesis postulates a limited number of V genes in the germ line which undergo unequal crossing-over during ontogeny to form recombined V genes, thereby expanding the V region information that is expressed (15, 16). Wu and Kabat (17) pointed out that amino acid sequence analysis of several hundred light chain V regions suggested a division of V regions into four "framework" regions with very little variability (FR1, FR2, FR3, and FR4) and three hypervariable regions (HV1, HV2, and HV3) that exhibit extensive sequence diversity. Kabat and others later proposed that each of these V region segments was a germ-line minigene that was inherited independently (18-20). By these hypotheses, subgenic
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. 1106

elements representing all possible V region framework and hypervariable sequences would be inherited in the germ line and would be rearranged during ontogeny to create diverse V region coding sequences. One version of this model postulates that the DNAs coding for the hypervariable segments are inserted into a preexisting framework region of the V gene (17). Other somatic rearrangements of inherited "minigene" V region fragments have been proposed (18-20), and it is possible to develop several models that would predict expansion, contraction, or insertion of hypervariable regions. However, two groups recently cloned and characterized mouse immunoglobulin light chain V region genes, and their results place serious constraints on the minigene hypothesis. Seidman et al. (21, 22) cloned and determined the sequence of two K light chain V region genes from neoplastic antibodyproducing cells (myelomas). Tonegawa et al. (23, 24) cloned and determined the sequence of a X V region gene from embryonic DNA, assuming this to be representative of germ-line DNA. In contrast to the minigene hypothesis, they found an intact V region containing, in order, FR1-HV1-FR2-HV2FR3-HV3. Although FR4 was present in 12-day embryo DNA as a noncontiguous genetic element, consistent with part of the minigene hypothesis, the fact that the HV regions were in place ruled out HV insertions at a later stage in development. However, it was still possible that gene insertion or gene reassortment events could occur prior to day 12 of embryogenesis and that some pther organization of V region minigenes exists and is inherited via germ-line DNA. We think that, in order to make any conclusions as to the origin of antibody diversity, it is necessary to study the gene organization in the germ line. For this reason we have characterized the organization of mouse K light chain genes in sperm and embryo DNA. To do this we constructed a plasmid containing a DNA insert coding for the K light chain MOPC 167 and used the Southern blotting technique to determine the organization of K V and C region genes. This analysis demonstrates unequivocally that the Vk167 gene family is not altered in size from germ line to 15-day embryos and, therefore, places serious constraints on minigene hypotheses that would require insertions, expansions, or contractions of Vk genes during embryogenesis. MATERIALS AND METHODS Preparation of Light Chain mRNA. Membrane-bound polysomal poly(A)+ RNA was prepared from MOPC 167 tumors (25) and fractionated on an isokinetic 10-40% sucrose gradient. The peak of light chain mRNA was identified by in vitro translation in rabbit reticulocyte lysate (26).
Abbreviations: V, variable; C, constant; bp, base pair(s); kb, kilobase(s).

Scheller) and the cDNA were phosphorylated as described (29) and ligated to 0. Sperm DNA and BALB/c embryo DNA (15 days' gestation) were prepared according to Blin and Stafford Agarose Gel Electrophoresis and Transfer to Nitrocellulose Filter Paper. (40). and used for nick translation. BALB/c mouse sperm was purified from vas deferens and epididymis.C-T-G-C-AOH 3' amino acid sequence of MOPC 167 K chain (38). BALB/c sperm and embryo DNA were digested with BamHI. Preparation of Vk and Ck Probes. pH 7. These nucleotides presumably code for part of the "signal peptide" removed in vvo from the primary translation product (39). . Synthetic EcoRI dodecamer linkers (gift of R. From the DNA sequence data of Hamlyn et al. Acad. A partial sequence of 80 nucleotides of the 470-bp HincII fragment corresponded to the 3' untranslated region sequence of mouse mRNA determined by Hamlyn et al. respectively.5 Mug of EcoRI-digested pMB9. This generated three fragments of 2800.2% agarose in 40 mM Tris acetate. The eluted DNA was loaded onto a 0. 1). respectively.5/8mM MgCI2/4mM dithiothreitol/0. 3A). Sci. whereas the 3' untranslated part of the K gene insert is oriented toward the HincII site farther away from the EcoRI site. Natl. (40) it is evident that the HincII site at amino acid position 125 coincides with an Hpa I site and that there is a Hae III site at amino acid position 196.ug/ml).C. coli 1776 as described (30). In order to orient the K chain insert in pMB9 we digested the recombinant plasmid with HincHI. Transformants containing light chain sequences were identified by using the Grunstein-Hogness procedure (31) with kinase 32P-labeled MOPC 167 light chain mRNA for hybridization (32). From these data we can conclude that the 5' terminal part of the insert is oriented toward the HincII site closest to the EcoRI site.5% sarcosinate according to Witkin et al. (34). separate 32P-labeled probes specific for Vkl67 and Ck sequences were prepared and used in all subsequent experiments. Mouse DNA (10 ug) digested with restriction endonucleases (enzymes were obtained from Bethesda Research Laboratories) was electrophoresed through horizontal agarose gels in 70 mM Tris-HCl. This sequence contains nucleotides corresponding to the amino terminus of the K light chain sequence of MOPC 167 (38). EcoRI. The cDNA (0. Preparation of Genomic DNA. The DNA fragments were separated by electrophoresis.14 M 2-mercaptoethanol and digested with proteinase K (0. beginning five nucleotides from the 3' end [excluding the poly (A)] (results not shown). In pMB9.5-1 Mug) was incubated in 25 Ml of 60 mM Tris-HCl. The DNA sequence of the 100-base-pair EcoRI/Pst.5 . Fig. Organization of K Light Chain Genes in Germ Line and Embryo.Immunology: Joho et al. a very strong band was visible in addition to a few weak ones. pH 7. The arrow indicates the amino terminus of the mature light chain.~~~~~~~~~~~~~~~~~~~~~EcoRIIPst w the corresponding I fragment with I Ser Val Ser lie Ser Cys T-C-A-G-T-T-T-C-C-A-T-C-T. and the 90base-pair insert was purified and end-labeled with [32P]phosphate by use of polynucleotide kinase. The DNA fragments containing Vk or Ck sequences were electro-eluted according to Tabak and Flavell (37) except that hydroxyapatite in the trough was replaced by DEAEcellulose (Whatman DE-52) suspended in electrophoresis buffer. Digestion of the end-labeled K chain insert with HincIl yielded two terminally labeled fragments of approximately 310 and 470 bp. The supernatant was removed and the pelleted tissue fragments were resuspended in the same medium. both supernatants were combined and freed from contaminating tissue by filtration through a nylon screen.6 M NaCl/50 mM Tris-HCl. pH 7. The mixture was used directly to transform E. (33). 1% NaDodSO4 and exposed for 4-14 days on Kodak X-Omat XR5 film with a Dupont intensifying screen at -70°C. precipitated with ethanol.5 mg/ml) at 500C for 3 hr. Bgl II. If not otherwise stated.8/60 mM Na phosphate/2 mM EDTA at room temperature. it was cut with EcoRI.2-ml DEAE-cellulose column and the DNA fragment was eluted with 0.2 mM deoxynucleotide triphosphates with 5 units of Escherichia coli polymerase I (Boehringer Mannheim) at 12'C for 10 min to produce flush ends (28). The same band pattern was obtained when sperm or embryo DNA was digested with a particular restriction enzyme (Fig. 2200. respectively. cDNA Cloning. The spermatocytes were pelleted by centrifugation for 10 min at 2500 rpm and contaminating somatic cells were removed by lysis in 1. and 1300 bp. The DNA was transferred to nitrocellulose filters as described by Southern (35). In order to clarify the origin of strong and weak bands. Mouse sperm cells were lysed in the presence of 2% NaDodSQ4 and 0. transferred to nitrocellulose filter paper. USA 77 (1980) Gly Val Ser 1107 Gly Asp lie Val 5' A-A-T-T-C-G-G-G-T-T-C-T-G-G-A-T-C-T-C-T-G-G-A-G-T-C-A-G-T-G-G-G-G-A-T-A-T-T-G-T-Glie Thr Gln Asp Glu Leu Ser Asn Pro Val Thr Ser Gly Glu E If -T-A-A-C-C-C-A-G-G-A-T-G-A-A-C-T-C-T-C-C-A-A-T-C-C-T-G-T-C-A-C-T-T-C-T-G-G-A-G-A-A. Purification of Mouse Sperm. pH 7. With each type of restriction digest. After centrifugation (1 min at 1000 rpm).I fragment was determined according to Maxam and Gilbert (32) (Fig. Adjacent to the terminal linker oligonucleotide there are an additional 25 nucleotides preceding the sequence coding for the amino terminus of the K light chain.5 mM Na citrate/0. and hybridized to 32P-labeled plasmid pl67kRI containing Vk167 and Ck sequences. The filters were hybridized to nick-translated 32P-labeled DNA (36). Phe Trp lie Ser Proc. The labeled insert was digested with either Pst I or HincII to produce terminally labeled fragments. The purified K chain insert was found to contain one Hpa I and one Hae III'site at the corresponding amino acid positions 125 and 196. In order to compare the organization of the K light chain genes in germ-line and somatic tissue. The plasmid pl67kRI was found to contain an EcoRIexcisable insert of approximately 900 base pairs (bp). 1% Na pyrophosphate/0. the two HincII sites are 910 and 2350 bp from the single EcoRI site into which the 900 bp K light chain sequence was inserted (41).7/1 mM EDTA/ethidium bromide (0. RESULTS The Plasmid pl67kRI Contains a Complete K Light Chain Insert. the filters were given a final wash in 45 mM NaCl/4.5/1 mM EDTA. Microscopic inspection of the sperm suspension revealed no detectable contamination with somatic cells (less than 1 nucleated cell per 200 spermatocytes). The tissue was minced with a razor blade and suspended in medium 199 containing 5% fetal calf serum for 15-30 min at room temperature. Double-stranded cDNA was synthesized from MOPC 167 light chain mRNA (27). These enzymes do not cleave the K chain insert in pl67kRI. To determine what portion of the mRNA sequence was contained in this plasmid. and HindIII. HincII-digested plasmid pl67kRI was electrophoresed through 1. 2 shows a restriction map of pl67kRI.

(B) Ck sequences. The faint bands visible in the first experiment were no longer detected. Restriction map of pl67kRI. 2. -2.7. respectively. In order to locate the positions of the various restriction sites near the Ck sequences in the mouse genome. Analysis was as described in Fig.4A B Hpa Hpa s e s s e s e e Hpa s e /. Organization of Region Surrounding Ck Gene.46 -6. 5). and only one intense band corresponding in position to the intense one in the above experiment could be seen. 4B). The Ck sequences are located on a 1. Organization of Vk Genes. and 4. Because the intense bands in Fig. which is compatible with DNA-RNA hybridization data (1-10). FIG. the plasmid pl67kRI was replaced with a Ck specific probe. The experiment described above was repeated but instead of plasmid pl67kRI. There is one additional Pst site at amino acid 43 (not shown on map). (A) Vk and Ck sequences. Bgl II. The orientation of transcription was determined by the known position of the Hpa I site in the amino acid coding sequence. or HindIII and analyzed with the Ck specific probe (Fig.64 -4. Sperm (lanes s) and embryo (lanes e) DNAs were subjected to two consecutive digestions as indicated. (C) Vk167 related sequences. The HindIII-digested sperm DNA shows some partial digestion products. HindIII-digested bacteriophage X DNA was used as a size marker and nick-translated X DNA was included during the hybridization. EcoRI. Sperm and embryo DNAs yielded the same size fragment when digested with a particular restriction enzyme. EcoRr. USA 77 (1980) Bam / 890 BP L VKI67 CK 11 0 3' UT 1001 Eco RI Pst I 2?0 1 210 260 Hinc 11 Hinc 11 Has III Eco RI Hpa I FIG. we took advantage of the fact that there is an Hpa I site at the codon for amino acid 125.34 Bam Bgl + + Eco + Bam Bgl + Hin + + + A Bgl Eco Eco se s e s e I Hin Hin Hin s e AI . The DNA sequence coding for the constant part of the K chain was in a 0. Hpa I-digested sperm and embryo DNA were further digested with BamHI.1108 Immunology: joho et al. We conclude that the organization of the Ck region is the same in the germ line as in the embryo (as far as one can tell by mapping several restriction sites). Thus. Proc. 4A). 5). Bgl II. transferred to nitrocellulose filter paper.8-kb HindIII/Bgl II fragment could contain as many as three Ck genes. we performed double digestions on BALB/c sperm and embryo DNAs and analyzed the DNA digests with a Ck specific probe (Fig. Sequences in restriction endonuclease-digested BALB/c DNA. 4. 3B. 3B). 2.26 -1. it was reasonable to assume that all the faint bands visible contained Bam Bam Bgl + + -23. BALB/c sperm (lanes s) and embryo (lanes e) DNA were digested with BamHI. By mapping this Hpa I site it should be possible to locate any other Ck genes within the 1. In order to determine how many Ck region genes are present in the germ line. a Vkl67 specific probe was used during hybridization. and hybridized to the nick-translated plasmid pl67kRI. and HindIII. and HindIII. Sci. Natl Acad. This Hpa I site is therefore close to the beginning of the Ck gene (close to the V-C junction). 3A corresponded to the fragment carrying the Ck region. The lengths of the fragments were 13. The resulting fragments were electrophoresed through 0. If we calculate 321 bp for the coding part of the Ck gene and 211 bp for the 3' untranslated region (40). the 1.6 _0 _ -9. 17.7% agarose. The possibility that the entire region shown in the restriction map may be repeated in the genome cannot be eliminated. Only One Ck Gene Is Present in the Mapped Region of the Genome. These results also permit us to construct a map of the restriction sites flanking the Ck sequence (Fig.3 kilobases (kb) when digested with BamHI.98 FIG. Bgl II.5-kb segment flanked on one side by the Hpa I site and on the other side by the Bgl II site (Fig.8-kb HindIII/Bgl II restriction fragment. . The hatched area of the insert -represents DNA for which the sequence has been determined. a 32P-labeled DNA fragment containing Ck sequences corresponding to the sequence from amino acid 125 to the carboxy terminus (including the 3' untranslated region) was A B C Bam BgI Eco Hin Bam BgI Eco Hin Bam Bgl Eco Hin Hin s e s e se s e sese s e se sesese s e kb used as a probe for hybridization (Fig. 3. we can expect 532 bp for the entire Ck region gene (provided that there are no intervening sequences present). These experiments show that there is only one Ck gene in the region mapped in germ line and embryo DNAs.8-kb fragment terminated by HindIII and a Bgl II site.

ruling out major size rearrangements of the Vk167 related V region genes during embryogenesis.5 mM Na citrate). although the exact number was -not known (1-11). a simple picture emerged when Hae III digests of sperm and embryo DNAs were analyzed with the Vk167 probe (Fig. E. it would have been difficult to detect small size differences (which could be involved in rearrangements or insertion of hypervariable subgenic elements). except that the filters were washed under various salt conditions: A. size differences as small as 20 bp would easily have been detected. A map of several restriction sites flanking the Ck region gene was established and shown to be the same for sperm and embryo. Vk167 related sequences in Hae III-digested BALB/c DNA. The location of several restriction sites around the Ck gene is shown and drawn to scale. DISCUSSION Because immunoglobulin genes from embryo and myeloma DNAs have been cloned and analyzed. Map of several restriction sites flanking the germ-line Ck gene. we thought that it was important to qompare the gene organization in the germ line to that in these somatic tissues.5 mM/Na citrate. Natl. unless the entire 23-kb unit shown in Fig. Yet there is a remarkable feature about the Vk167 related V region genes. only two bands were found when the filters were washed under the same salt conditions (45 mM NaCI/4. A similar finding has been reported for V region genes related to the K light chain gene expressed in MOPC 149 and MOPC 41 (21. w al. 5. These results can be interpreted that the mouse haploid genome contains several K V region genes that are related to the Vk167 gene which is expressed in MOPC 167. Sperm (lanes s) and embryo (lanes e) DNAs were digested with Hae III and the resulting fragments were electrophoresed through 2% agarose. EcoRI. Ba. even a small contamination of the Vk167 probe with Ck probe shows in this type of analysis. one can conclude that during embryogenesis the Ck gene retains its germ-line organization as far as the flanking restriction sites are concerned. The right two lanes in C show some differences in band intensity but not in band pattern. Sperm and Embryo Ba Proc. BamHI. All the bands had the same position in sperm and embryo DNAs. The arrow indicates the orientation of the Ck gene (direction of transcription). the same experiment as described in Fig. C. Hp. Because the various restriction fragments carrying Vk167 related sequences were large (up to several kilobases). The Ck gene including the 3' untranslated region is indicated by black box. In order to solve this problem we used a restriction enzyme that produced smaller fragments but did not cut the V region Vk167 of the plasmid (Fig.5 mM Na citrate. 3C). Instead of finding several smaller bands corresponding to as many larger bands that could be seen before. This experiment therefore shows clearly that there is no detectable difference in the gene organization for Vk167 related V region genes between germline and embryo tissues. 6. The results presented here imply that the Ck gene including the 3' untranslated region is present only once in the haploid mouse genome.Immunology: Joho et al. Sci. it was still visible to a varying degree. the poss e sibility that a Vk gene is close to the Ck gene or present on a restriction fragment of similar size cannot be ruled out.) it is a reasonable assumption to expect a restriction fragment carrying a Vk167 s e s e bp I ~ ! -4260 ~ . From RNA-DNA hybridization studies it was clear that there were only a few (one to four) Ck genes present in the genome. The Vk167 related sequences were analyzed as described in Fig. Therefore. only the nearest restriction site of a particular enzyme is shown. 45 mM NaCl/4. On either side of the Ck gene. 3.3 M NaCl/0. Surprisingly. . Although it is somewhat difficult to interpret band intensities on a Southern blot (it depends on transfer efficiency of the DNA. In this experiment. We therefore think that the bands visible at the Ck position when a Vk167 probe was used were due to contaminating Ck sequences in the probe. specific activity of the probe. Multiple restriction fragments that hybridized to the Vk167 probe were detected. The Vk167 probe used in these experiments was slightly contaminated by Ck probe. 15 mM NaCl/1. whereas under less stringent conditions (0. A Vk167 probe yielded only very faint bands on a Southern blot compared to a strong signal obtained with a Ck probe.03 M Na citrate) as many as seven bands could be detected. it is clear that large intervening sequences cannot be present. -2250 -1960 W 0 4W -590 '4. We used the Southern blotting technique to analyze Vk167 and Ck sequences. However. 2). USA 77 (1980) 1109 E B HHpBBa H Ck: Length. Under more stringent washing conditions. H. Bgl II. 5 is repeated a few times. etc. length of the probe available for hybridization. and sperm and embryo DNAs revealed the same band patterns when analyzed with any particular restriction enzyme. A B c FIG. Because the Ck gene does not comprise more than about 500 bp. None of these bands was due to Ck sequences because a similar analysis with the Ck probe yielded two bands of different length (results not shown). Vk sequences related to or identical with the V region sequence expressed in MOPC 167. The band patterns of sperm and embryo DNA restriction digests appeared to be identical. B. In order to test this assumption. kb -6-4 -2 0 1 2 3 4 5 -10 10 FIG. Hpa I. 6). Acad. only one band of 950-bp was seen. B. 22). Although the strong band containing the Ck gene had disappeared for the most part.03 M Na citrate. The same faint bands were visible. 3B was done except that the Ck probe was replaced by a Vk167 probe (including the V region sequence from the amino terminus to amino acid 93) (Fig.3 M NaCl/0. HindIII. 0. Whether the germ-line and embryo genes are different or not in their DNA sequences cannot be decided from the data presented here. Analysis of the V region gene(s) was less straightforward.

Nau.3881-3885. 31. Jeffreys... (1966) Nature (London) 211.. S. R. Tilghman. & Schuller. (1975) Annu. H.1110 Immunology: joho et al. 0. 28. P. Shine. (1978) Biochemistry 17.203-207. S. 33.. Med. S. 9. Dube. 88. USA Wickens. 11. A.L. Sci. Stavnezer. Leder. I... N. Natl. P. Hamlyn. (1976) Proc.4027-4031. M. Seidman et al (21. Vasalli.. & Bernardini. Maxam. & Milstein. 18. T. USA 71. 22) reported that the flanking sequences of two cloned related K V region genes are similar. Natl. Nat!. Stavnezer. Lacy. (1978) Nature (London) 276.353-357. & Gally. C. is Postdoctoral Fellow of the Swiss National Foundation. N. one would expect different restriction patterns of sperm and embryo DNAs. & Stafford.. & Milstein. Rev.. 30. R. Nature (London) 270. P. 26. Sci. D. J. P. P. Tonegawa.. Nau. (1978). C. Yonkovich. Gen. 3. W. Sim. & Boyer.. M. 1231-1236. Genet. (1974) Proc. C. Aellen. A. E. G. J. & Flavell. 305-353. 98. J. P. S. Pelham. Edelman. Honjo. (1978) Proc. Natl.. respectively. M. Acad. & Bishop. (1978) Science 202. Tizard. E. Packman. 7. Kindt. Faust. USA 73. Acad. Natl. Jarvis. (1978) J. 32. 3. Edelman and Gally (16) suggested that the diversity of V region genes is created by somatic recombination of a few germ-line genes.. Sci. Biol. & Mach.. R.. Tonegawa. M. Acad. 41.486-494. M. T.. Campbell.. Natl. Therefore. Acad. H. A.. 2483-2495. USA 75.). (1979) J. 1485-1489.. Natl. . S. Sci.285288. R... Sci. Seidman. E.. P. Seeburg. C. T. E. 164. R. M. USA gene to yield a signal of similar intensity as a restriction fragment bearing the Ck gene. 429-439. Because we do not find any differences between sperm and embryo DNAs. 16. O'Connell. 43-63. Southern. 38. J.. 72. S.. J. (1974) Proc. Nau. Huang. D. 3659-3663. Proc. Tiemeier.. Natl. We consider these possibilities to be unlikely. (1975) Cell 6. & Bendich. (1976) Biochemistry 15. I. Leder. R. J. 253. 21. A.. Cesari. Brenner. Lauer. K. Cheng.. H. H. J.780-784. G. H. T. & Potter. M. Natl. 125-133.5109-5114. Hood. L. 12. Natl. Honjo. This work was supported by American Cancer Society Grant IM56 (to I. A. 2321-2332. Natl. W.H. 1067-1075. & Norman. F.. Hardison. J. Acad. P. Acad. Sci. J.. Tait. Rabbitts. Kabat. & Schimke. Seidman. Chem. Acad. Biol. USA 71.. If related V region genes have similar or identical flanking sequences. N. 29. L. P.3727-3731. Leder. 3A) when a 32P-labeled probe of the same specific activity for Vk167 and Ck was used may indicate the absence of a V region gene identical to the one expressed in MOPC 167. G. (1978) Biochemistry 17. Marshall. (1974) Proc. Brack. Wu. Sci. 57. 309321. C. G. D. (21.3961-3965. E. Acad. 149. Sci.. 15. Tabak. Brownlee. 4. W. (1970) J. (24) and Seidman et al. Weigert. The results shown in Fig. & Hogness. 22) we know that the V region genes in embryo DNA contain sequences already coding for the hypervariable regions but lack sequences coding for the fourth framework. S. A. M. 503-517. E. R. & Capra. K. Briand. C.3518-3522. P. A. A. Recently. Proc. H. S. Nat!. Marcu. Cohn. 23. Maniatis. (1978) Nucleic Acid Res. Acad. 687701. H. Sci. 5-12. R. A. R. 1. Norman. S. D. T. Sci. 37. C. H. A. J.. (1963) Nature (London) 199.. S. (1977). The fact that V genes in embryo DNA consist only of the V region through HV3 is in accord with this hypothesis. E. & Schechter. If assembly of V genes happened during embryogenesis. 10. O.. (1977) Proc. 36. L.. (1978) Cell 15. M. Biochem.. 1723-1733. (1975) J. M.. C. Efstratiadis. Biochem.. M. W.727-731. 0. Sci. 1045-1047. C. the cloning and sequencing of these germ-line genes will resolve this question. USA 74. Sci. Maxam. Therefore.. Y. P. Wu. Tonegawa..J. USA 75. A..E. Packman. 39. T. is a Faculty Research Awardee of the American Cancer Society. Proc. B. (1975). USA 71. F. 3. S. A.. Broome. Bernard. & Milstein. Weigert. 75. Mol. B. R. 2780-2785. 2491-2495. R. Sci. A. E. Burstein. Med. M. & Perry. 14. S..2429-2433. & Efstradiadis. 9. (1975) Proc. B. 3295-3299.L. Naber. Lomedico. T. From the work of Tonegawa et al. we consider somatic recombination during embryogenesis to be a very unlikely event. M.. D. Exp. W. & 72. Baxter. & Kabat. K. & Flavell. 23032308. (1978) Proc. & Elgin.560-564. Rudikoff. Tizard. C. Chick. 20) presented evidence that this might be the case. J.. USA 75. (1978) Proc. T. Acad. D. 242243. T. Amino acid sequence comparison of more than 500 light chains indicated that the different framework parts of the V region genes might be inherited independently and be put together as a functional V region gene during cell differentiation (early embryogenesis or B-cell differentiation). & Gilbert. W. USA. & Perry. & Bilofsky. J. H. C. R. USA 22. B. 132. A. (1967) Proc. B. Kabat et al.. T. Valbuena. Martial. C. 247-256. W. Villa-Komaroff. Polsky. Southern blotting. 6 indicate that this does not happen because sperm and embryo DNAs yield the same Vk carrying fragments after Hae III restriction. 13.W. USA 71. J. an episomal insertion mechanism as suggested by Wu and Kabat (17) or a gene interaction hypothesis as proposed by Kindt and Capra (19) would have to occur during early embryogenesis. is supported by National Institutes of Health Training Grant GM 07616. after restriction and 17. Gilbert. 35. cleavage with Hae III may yield multiple DNA fragments of the same length carrying related Vk genes. 20. Tonegawa. Mol. Acad.. & Leder. 25. several types of hypotheses have been put forward to explain antibody diversity by somatic recombination. Exp. Wu and Kabat (17) proposed that DNA segments coding for the hypervariable regions are joined to framework segments in assembling an intact V region. Packman. H. R. A. 23922400... M. (1977) Proc. & Mach. (1975) Eur.2703-2707. USA 74. USA 77 (1980) 5. M. T. O. F. Recently. Smithies. 2. & Bilofsky. T. (1978) Cell 15. T. 1299-1313. H.) and National Science Foundation Grant PCM 76-81542 (to L. P. & Leder. (1974) J..73. Acad. M. P. Steinberg. G. Faust. Sci. Buell. M. 34. Marcu. G. Swan. 67. D. A. S. A. 40. Acad. & Goodman. R. Therefore. (1978) Immunogenetics 6. Wu. Natl.211-250. J. R. B. & Gilbert. Blin. Gait. Swan D. Natl. 19. Farace. Swan. T.W. 8. It is still possible that hypervariable insertions could replace germ-line fragments of the same size.. (18. Genet.. Sci. Acad. Witkin. M. 24. P. (1976) Proc. & Milstein. C. Rabbitts. Natl. Quon. Grunstein. B.. (1978) Biochemistry 17.. C. (1970) Nature (London) 228. M. 6. (1978) Mol.. M. (1976) Eur. Valbuena. Korngold. 52. Natl Acad. Sci. T. & Leder. Honjo. G. Acad. R. & Jackson. (1976) Nucleic Acid Res. C. T. H. then one would expect restriction sites to be conserved within these sequences. Biol. S. I. 27. (1974) Proc. the results found (Fig. Kabat. H. 11-17. D. H. S.. Over the years. (1977) Cell 12. Whether somatic recombination happens during B-cell differentiation remains unclear. When Hae III digests of sperm and embryo DNAs were analyzed for the V region genes only two major bands could be detected with a length of 780 and 950 bp. P. Hozumi. J. S. Diggelmann.. & Leder...