Identification of cloned Y chromosomal DNA sequences from a lampbrush loop of Drosophila hydei

Proc. NatL Acad. Sci. USA Vol. 79, pp. 5132-5136, September 1982 Biochemistry

Identification of cloned Y chromosomal DNA sequences from a lampbrush loop of Drosophila hydei (spermatogenesis/recombinant DNA/gene structure/in situ hybridization)

PETER VOGT, WOLFGANG HENNIG, AND INGRID SIEGMUND Department of Genetics, Catholic University Nijmegen, Toernooiveld, 6525 ED Nijmegen, The Netherlands

Communicated by W. Beermann, May 11, 1982

ABSTRACT By recombinant DNA techniques, a Y chromosomal sequence of Drosophila .hydei was isolated. This DNA sequence of8.93-kilobase length is a member ofa family ofrepetitive sequences located in the short arm of the Y chromosome. Tissuespecific transcripts complementary to the cloned sequence were found in testes RNA. In situ hybridization demonstrated that such transcripts are present in the lampbrush loop pair "nooses" in primary spermatocyte nuclei-a loop pair that is associated with the only fertility gene in the short arm of the Y chromosome.

Details of the cloning procedures will be described elsewhere. All recombinant DNA experiments were carried out under CII (PII) containment conditions of the Dutch recombinant DNA rules, which essentially agree with the guidelines of-the National Institutes of Health. For dot hybridization, the assay of Kafatos et al. (14) was modified by spotting alkali-lysed plasmids (15) on nitrocellulose filtergrids under low vacuum. Before loading, the plasmids were nicked by a minimum amount of DNase, denatured in 0.2 M NaOH, and neutralized with 1 vol of 2 M ammonium acetate. The amount of DNA per spot was 1 pkg. The dot filters were hybridized at 680C in 0.2% polyvinylpyrrolidone/0.2% Ficoll/ 2% bovine serum albumin/300 mM Tris-HCl/30 mM EDTA/ 0.90 M NaCl, pH 8/0.2% NaDodSO4 with Escherichia coli tRNA (40 ,ug/ml), poly(A) (100 Ag/ml), and 5' end-labeled [32P]RNA from total testes or spermatocyte nuclei (106 cpm/ ml). After hybridization for 40 hr, we washed the filters for several hours at 60'C in large volumes of 0.3 M NaCl/0.03 M sodium citrate, pH 7/0.2% NaDodSO4. For restriction enzyme digestion, we followed the protocols of the suppliers. DNA blots were prepared as described by Southern (16). The blots were prehybridized and hybridized in 0.1% polyvinylpyrrolidone/0. 1% Ficoll/1% bovine serum albumin/200 mM Tris-HCI/20 mM EDTA/0.90 M NaCl, pH8/ 0.1% sodium pyrophosphate/25 mM phosphate buffer, pH 7.0, with 250 ug of sonicated and denatured calf thymus DNA per ml as carrier. To the hybridization mix we added the [32P]DNA probes (=106 cpm/ml) labeled by nick translation as described by Rigby et al. (17). After hybridization for 40 hr at 680C, we washed the filters for some hours at 60'C with 0.3 M NaCl/ 0.03 M sodium citrate, pH 7/0.2% NaDodSO4 or with 0.015 .M NaCI/0.0015 M sodium citrate, pH 7/0.2% NaDodSO4 at 65°C if high stringency was desired. RNA blotting was carried out as described by Thomas (18). For autoradiograms we used Kodak X-Omat film and Dupont Lightning Plus screens. Restriction maps of the clones were constructed from appropriate double and partial digests. The EcoRI sites were mapped as described by Smith and Birnstiel (19). For the transcript in situ hybridization experiments with primary spermatocytes, slides were prepared as described earlier (11). After removing the cover slip by freezing the slides in liq*uid nitrogen, the tissue was treated with ethanol/acetic acid, 3: 1 (vol/vol), for 3 min. After dehydration in a series ofethanol steps, hybridization was done without denaturation in 10 mM Tris/0.3 M NaCl with 2 x 105 cpm of [3H]cRNA per slide prepared from the plasmid clones as described by Wensink et al. (20). After hybridization for 4 hr at 68°C, the slides were washed for 4 hr at 50°C in 0.3 M NaCl/0.03 M sodium citrate, pH 7, dehydrated in a series of ethanol steps, and coated with Ilford L4 emulsion (diluted 1:1 with distilled water). The autoradiographs were developed in Kodak Dl9b for 2 min at 16°C. The

The Y chromosome ofDrosophila has long been known to carry genes responsible for male fertility (1) but is otherwise relatively devoid of genes compared with its size (2). This is in accord with its heterochromatic appearance in tissues other than the male germ line, where it becomes active in primary spermatocytes (3). Although many arguments can be made for an essential function ofthe Y chromosomal genes in spermatogenesis, their actual role in securing male fertility is still unknown. Recent data from our laboratory indicate that at least some of the genes of the Y chromosome of Drosophila hydei are involved in regulating the expression ofautosomal genes producing sperm constituents (unpublished data). In the Y chromosome of D. hydei, a total of 16 complementation groups has been identified (4), all of which are essential for spermatogenesis. Five of the fertility genes develop lampbrush loops during their transcription in the premeiotic stage of spermatogenesis (5), comparable to the "giant granular loops" described by Callan and co-workers for the. chromosomes in the oocytes of amphibians (6). Although detailed cytological (7) and genetic (4) observations on the Y chromosomal lampbrush loops are available, knowledge of the molecular structure of these loci is restricted (see ref. 7). From earlier work (8-10), it is known that repetitive DNA sequences located on the Y chromosome are represented in testes RNA. This RNA is synthesized during the lampbrush stage of the Y chromosome but is still found in postmeiotic stages in which RNA synthesis has ceased (8, 11, 12). In this study, we describe the isolation and identification of a DNA sequence from one ofthe lampbrush loops of the Y chromosome by recombinant DNA techniques. This loop, formed in the short arm of the Y chromosome, is composed of repetitive DNA sequences of a complex structure. The cloned sequence is transcribed as shown by transcript in situ hybridization and other biochemical experiments. MATERIALS AND METHODS The preparation of Drosophila DNA from adult flies will be described elsewhere. Testes RNA was prepared as described by Glisin et aL (13). The publication costs ofthis 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.

Abbreviation: kb, kilobase(s).

5132

Biochemistry: Vogt et aL

Proc. Natl. Acad. Sci. USA 79 (1982)

time of exposure is indicated in the legends. Photographs were taken with a Zeiss photomicroscope III on Agfaortho 25 and Agfapan 100 film.

RESULTS Identification of Transcribed Y-Chromosomal Clones. The identification of Y chromosomal DNA sequences in a recombinant DNA clone bank of D. hydei will be described in detail in a separate paper. In summary, these clones were created by ligating partial Pst I digests of genomic DNA of males to Pst Idigested pBR322 DNA. Recombinant clones were screened with radioactive DNA from males or females of D. hydei. All clones preferentially or exclusively hybridizing with DNA of males were further studied with the aid of DNA blots after digestion of D. hydei DNA with Pst I. From a series of recombinant clones recovered in this way, the clone PY9 described in the present communication reacted exclusively with DNA of males. It had an insert length of 8.93 kilobases (kb). Hybridization of nick-translated PY9 DNA with Pst I-digested genomic DNA from males resulted in a hybrid pattern as shown in Fig. 1. This pattern remained unchanged when, after hybridization, the filters were washed in low salt (0.015 M NaCI/0.0015 M sodium citrate, pH 7). Therefore, we conclude that the DNA sequence, cloned in PY9, belongs to a

A

B

kb

-23.67

PY9-

-

9.46

- 4.26

5133

family of repetitive DNA sequences with a relatively small degree of sequence divergence. Screening of Y chromosomal clones in a dot-hybridization procedure (14) (Fig. 2) with RNA extracted from either testes

or spermatocyte nuclei showed that the clone PY9 was

transcribed. Localization of the Clone PY9 on the Y Chromosome. The determination of the position of Y chromosomal DNA sequences within the chromosome became possible with the aid of suitable X/Y translocation chromosomes carrying only parts of the Y chromosome (for details, see ref. 4; unpublished data). With this method, the clone PY9 can be shown to be derived from the short arm of the Y chromosome. From earlier studies it is evident that the lampbrush loops in primary spermatocytes are transcriptionally particularly active loci (11). To determine whether or not the transcribed sequence of clone PY9 is located and expressed in one ofthe lampbrush loops, we carried out transcript in situ hybridization experiments as described by Pukkila (21). Slides with primary spermatocytes ofwild-type males were hybridized with [3H]cRNA of the clone PY9 as probe. RNA-RNA hybrids were detected by autoradiography. We found that the radioactive probe binds to the lampbrush loop pair "nooses" (Fig. 3). This loop pair has been associated with the only genetic site detected in the short arm of the Y chromosome (4, 5). Because artifacts have been described in transcript in situ hybridization experiments with amphibian lampbrush loop chromosomes (23), we performed a series of experiments, making use of males with partially deficient Y chromosomes, to exclude artificial binding of the RNA to lampbrush loops. From strains obtained in our genetic studies (4), the following male genotypes could be constructed: (i) males carrying only the long arm of the Y chromosome [In spermatocytes, all lampbrush loop pairs except the nooses (and all genetic loci except Q) are present. This chromosome region showed no crossreaction with the clone PY9 in DNA blots (data not shown). ]; (ii) males containing exclusively the short arm of the Y chromosome [This results in the expression of only the loop pair nooses (locus Q) in primary spermatocytes. In this chromosome region, all DNA sequences homologous to our clone PY9 were detected in DNA blots.]; (iii) males without Y chromosome (X/0 males). 4

S i

0

S.

- 1.96

FIG. 1. Hybridization pattern of clone PY9 to DNA blots of Pst Idigested DNA of wild-type males (lane A) and females (lane B) of D. hydei (3 pg of DNA per lane). For size comparisons, selected HindIu restriction fragments of wild-type DNA are indicated at the right. The position of the cloned fragment is indicated at the left. As repeated experiments indicate, the complex pattern is not the result of a partial digest. The label in the range of high molecular size is due to the presence of undigestable DNA, as it is often obtained in single-fly DNA preparations.

FIG. 2. Dot-hybridization experiment of spotted Y chromosomal plasmid clones (PY series) hybridized with 5' end-labeled total testes [32P]RNA. Each spot contains 1 ug of DNA. Hybridization was in 300 mMTris HCl/30 mM EDTA/0.90 M NaCl, pH 8, at 70C for 40 hr, with washing in 0.3 M NaCl/0.03 M sodium citrate, pH 7, at 65°C. DNA spots that hybridized more strongly than did the pBR322 spot (arrowhead) are defined as containing DNA sequences expressed in testes transcripts. The clone PY9 is indicated by an arrow.

Proc. Natl Acad. Sci. USA 79 (1982)

Biochemistry: Vogt et aL

5134

1) did not show any predominance in the hybridization pattern. The fragment pattern in the low molecular weight range showed some regular spacing reminiscent of digest patterns of simple tandemly repetitive DNA sequences (satellite DNA). However, the pattern as a whole may be more complicated (for example, as indicated by the clustering of fragments in the 5- to 7-kb

range).

xTh

1.

Ns

N

D .i

.

*

C4

,Au

/.

..

.\

.4:

..

....

F.

Because of the internal complexity of the cloned sequence (see Fig. 5) and the heterogeneous Pst I fragment pattern (Fig. 1), a determination of the number of repeats was complicated. We estimate from comparisons of the signal in the dot-hybridization experiment with the intensity of signals of other clones that there may be =80 Pst I copies (of variable length). If the cloned DNA sequence is part of a satellite sequence cluster, restriction enzymes not cutting within the cloned sequence should produce a few bands or even only a single restriction band characteristic for the total length of a cluster of repeated sequences. However, for the two enzymes tested (Sal I and Xba I), the opposite was observed (Fig. 4). In total digests of genomic DNA with Sal I or Xba I, a larger number of restriction fragments hybridizing with the cloned sequence were obtained. The hybridization patterns had remarkable similarities: a strong hybrid formation with a small fragment (0.6 kb for Sal I or 0.8 kb for Xba I) was accompanied by strong hybridization at the top region of the gel. No regular spacing could be established from these blots. At present, two possibilities of sequence organization may be postulated. In one type of arrangement, alternating short and long Sal I fragments (or Xba I fragments or both) are tandemly arranged. At least two Pst I restriction sites must be present in the long fragments. In an alternative arrangement, small Sal I (or Xba I) fragments (or both) are tandemly arranged forming clusters of different

A FIG. 3. In situ hybridization of [3H]cRNA of the clone PY9 to the transcripts of the lampbrush loop pair nooses, located in the short arm of the Y chromosome. NO, nucleolus; Ns, nooses; C1, clubs; Tr, tubular ribbons; Ps, pseudonucleolus; Th, threads; Au, autosomes. (A and E) Phase-contrast autoradiography without staining. (C, D, andF) Phase contrast after Giemsa staining. The strongly stained regions represent autosomal material (22). (A) Primary spermatocyte nucleus of wildtype D. hydei. (B) Schematic presentation of the nucleus in C and D. (C) Same nucleus as in D after removing the silver grains. (E) Cytological picture of a spermatocyte nucleus displaying only nooses. A phase-contrast picture is shown because the nucleus shown in F, after removing the silver grains, does not permit recognition of the nooses. Even in fresh cytological preparations, nooses are often hardly visible. It is assumed that the more refractive appearance, as seen inE, reflects an artificial contraction of the ribonucleoprotein along the loop axis. (F) In situ hybridization with a spermatocyte nucleus containing only nooses, with hybridization occurring strongly and exclusively to nooses. It is remarkable that labeling in nuclei of this constitution is always lower than in the wild-type nuclei. The reasons are unknown. Exposure of the autoradiographs was 10 days. (Bar = 5 gm.)

Autoradiographic label in transcript in situ hybridization experiments with the clone PY9 was found only in spermatocytes of those males that carry the short arm of the Y chromosome with unfolded nooses in the spermatocyte nuclei (Fig. 3). Therefore, we conclude that the clone PY9 carries an DNA insert homologous to the transcribed region of the nooses. Genomic Organization of DNA Sequences in the Lampbrush Loop Nooses. When the clone PY9 was hybridized to a Pst I digest of genomic DNA from males, a ladder of fragments ranging from 0.7 to 21 kb and no hybridization at all on DNA from females were seen. The Pst I fragment cloned in PY9 (Fig.

kb 23.67-

B

t*

-.

946-

4.26-

1.96-

FIG. 4. Hybridization pattern of clone PY9 on a DNA blot of DNA from wild-type males digested with Sal I (lane A) or Xba I (lane B). The

0.5 9-*4-

molecular size in kb is indicated on the left by the position of selected HindHI fragments of A phage wildtype DNA.

Biochemistry: Vogt et al. PDV r, . LI 0I

.

-I--T

Proc. Natl. Acad. Sci. USA 79 (1982) fjd jbj I v

v

e

v

jbj e

f

j

v

1I kb

9T:I

-

d

5135

Jbi d

----RNA

FIG. 5. Restriction map of PY9. The pBR322 regions are indicated by thick bars. The thin lines under the map show the EcoRI fragments that hybridize with testes RNA. The interrupted lines indicate the EcoRI fragments that hybridize only slightly with testes RNA (see also Fig. 6). The small letters b-g characterize internal EcoRI elements of various lengths (letters a and c are not present because they define other fragments in different clones, not discussed in this paper). Identical letters in the different fragments do not imply identity of sequence but of length. Symbols for restriction sites: 4, EcoRI; I, Hincd; 9, HinduII; K, Hinfl; t, Hpa I; 9, Pst I; 7, Pvu II; 4, Sph I. (HinfI cuts pBR322 in 10 fragments. The map displays only the two sites closest to the Pst I sites of the Drosophila insert.)

lengths. These clusters must be bordered by long Sal I (or Xba

I) fragments (or both) with or without Pst I restriction sites. In any case, the smallest Sal I and Xba I fragments must have sequence homology to parts of the cloned Pst I fragment PY9. Characterization of the Cloned Sequences and Their Transcripts. In establishing a restriction map of the clone PY9, only 5 out of 25 restriction enzymes cut the cloned insert (Fig. 5). The insert was characterized by a series of EcoRI restriction fragments that differed in length (from 0.54 to 1.33 kb). From cross-hybridization experiments (data not shown), we found that most if not all EcoRI fragments showed extensive crosshomology. To identify the sequence-parts being transcribed, we hybridized a DNA blot of the EcoRI-digested clone with 5' endlabeled total testes [32P]RNA (Fig. 6). All EcoRI elements hybridized except the EcoRI fragment of 3.7 kb at one of the pBR322/insert borders. The length of the insert region in this fragment was presumably too small (10-100 base pairs) for hybridization to show up in this kind of experiment. The variable hybridization intensity of the other bands is of interest because it indicates some degree of sequence heterogeneity between the various EcoRI elements. Such a difference in the sequences was particularly evident for the 1.33-kb EcoRI fragment from the clustering of restriction sites (Fig. 5) not found in the other EcoRI fragments. These data indicate that regions f and g are either only partially homologous to the testes RNA or that they are present at considerably lower redundancy than RNA sequences homologous to the other EcoRI elements of the cloned fragment. Because in this experiment total RNA from testes was used, which mainly consists of cytoplasmic RNA, the hybrid pattern reflected preferentially the composition of cytoplasmic RNA. To obtain information on the length of the nooses transcripts that were homologous to our clone, we prepared glyoxal-denatured total RNA (25) from testes and carcass (males after dissection of testes) and separated it by electrophoresis on a 1.2% agarose gel. After transfer to nitrocellulose filters (18), we hybridized the RNA blot with labeled DNA of PY9 (Fig. 7A). Whereas carcass RNA did not hybridize at all (Fig. 7B), a wide molecular weight range of RNA molecules appeared to be homologous to PY9, including molecules from 1 to 10 kb. A maximum density of hybridization was found around 3 kb, but a distinct size class was not detected. It is difficult to exclude the possibility that the heterogeneous distribution of hybrids was due to RNA degradation, but 18S and 28S ribosomal RNA molecules were not much degraded. They are seen as white bands in Fig. 7A. Hybridization of the same filters with a ribosomal DNA clone indicated with a higher level of sensitivity that only minor degradation of ribosomal RNA had occurred (data not shown).

DISCUSSION In this paper, we describe a transcribed DNA sequence obtained from the Y chromosome of Drosophila. It is related to the lampbrush loop pair nooses in the short arm of the Y chro-

mosome and, hence, is associated with the only gene detected in this chromosome arm (4). This chromosome region is composed of a family of repetitive sequences. The existence of repetitive elements associated with this locus and others close to the kinetochore has been postulated earlier (4) from genetic arguments. These repetitive sequences are represented by a series of Pst I fragments (0.7 to >20 kb), which may be added up to a total of at least 150 kb. The basic structure of the different Pst I fragments appears to be rather similar as indicated by the restriction patterns of two other members of this family (unpublished data). In principle, this sequence family is constructed from relatively short EcoRI elements (390-1,350 bases in the clones mapped so far). In a large scale arrangement, these EcoRI elements are included in the various Pst I fragments. Although some similarities in these features are reminiscent of the structure of DNA sequences separating the histone gene clusters in Notophthalmus (26), the structure of the sequences ofthe nooses must be more complex. The restriction analysis and the transcription pattern of the EcoRI elements in PY9 indicate sequence heterogeneity. The cloned sequence is complementary to transcripts formed

PY. 9 kb 4.26-

1.62-g

f -e

0.850.54-

-b

A

B

C

FIG. 6. Hybridization pattern of 5' end-labeled total testes [32P]RNA on a DNA blot of the clone PY9 digested with EcoRI (lane B). For comparison, the corresponding fluorescence pattern of the gel is shown after staining with EtBr (lane A). The fragment lengths indicated at the left are selected from a Hindu digest of wild-type A phage DNA and a Hae HI/Tha I digest of M13 DNA (24). The schematic pattern (lane C) refers to Fig. 5.

5136

quence in RNA lead to three alternative interpretations. Possibly some parts ofthe cloned sequence occur with much higher frequency in testes RNA than do other sequence regions. This could be due to a more frequent representation of these repeated elements of the cloned inserts in other parts ofthe loop. Alternatively, partial transcription of the clone might be envisaged. A third possibility is that the regions hybridizing less intensely might be removed by processing events after transcription of long stretches of the loop. In this context, it should be recalled that earlier electron microscopic studies (27-29) revealed extremely large transcripts in spermatocyte nuclei. Such large transcripts hardly can be isolated without degradation. The function of the gene associated with the lampbrush loop nooses is so far unclear. We also have no proof that the DNA sequence responsible for the function of gene Q is represented in our clone. The fact that RNA complementary to PY9 is found in the poly(A)-containing fraction of the cytoplasm (unpublished data), might be taken as an argument for translation of this RNA. This also has been postulated from the existence ofa conditional mutation in this gene (cf. ref. 4). The availability ofa wide range of mutants, part of them clearly affected in a regulatory region oflocus Q (4), will allow an extensive study of this region of the Y chromosome.

B

A

kb

2.01.6 -

FIG. 7. Hybridization patterns of clone PY9 to an RNA blot of glyoxal-denatured total testes RNA (lane A) and carcass RNA (lane B). Internal ribosomal RNA markers: 2.0 kb for denatured 28S RNA and 1.6 kb for 18SRNA. Possible contamination with DNA has been excluded by applying the method of Glisin (13) and checking the RNA preparation with gel electrophoresis (data not shown).

in the lampbrush loop pair nooses in the short arm ofthe Y chromosome. Because carcass RNA does not hybridize detectably, the RNA appears to be testes specific. The existence of such a RNA species has been demonstrated by competition experiments (8, 12). In the cytoplasm, its transcripts are found as

molecules of heterogeneous lengths. We cannot yet exclude completely the possibility that some degradation occurred during RNA extraction that preferentially would affect large transcripts. However, the different frequencies of the EcoRI elements of the cloned sequence in RNA from total testes supports a model of heterogeneous size of the cytoplasmic RNA derived from this genetic site. Sequence elements f and g in PY9 are under-represented in the testes RNA. Of course, we cannot disregard the possibility that none of the cloned sequences is transcribed but that the hybridization is due to crossreaction with transcripts from other sequences showing homology. However, in in situ hybridization, no region of the nooses hybridizes exclusively or preferentially with cRNA copied from clone PY9. This suggests that a major part of the loop pair nooses is transcribed from DNA homologous to the PY9 sequence. Hence, we may assume that the cloned sequence itself is represented in transcripts. It is of some interest that the molecular weight distribution of the RNA molecules in total testes RNA is compatible with the size range of Pst I elements in a genomic DNA digest. This might be in support of a model based on a discontinuous representation of the loop sequences in the cytoplasm. The recent observation of heterogeneity in size might explain why our earlier approaches to the identification of Y chromosomal RNA species of discrete molecular weight failed (12).

In conclusion,

View publication stats

Proc. Natl. Acad. Sci. USA 79 (1982)

Biochemistry: Vogt et aL

our

data on the representation of the PY9 se-

We thank Mr. J. Hermans and P. Krimpenfort for help in restriction mapping; Dr. J. Hackstein for supplying Drosophila mutants; and Drs. R. Brand, P. Boender, and T. Hulsebos for extensive discussions; Miss T. M. M. ten Hacken and Mr. W. Janssen for technical help; and Miss E. C. Dolk for typing the manuscript. 1. Bridges, C. B. (1916) Genetics 1, 1-52 and 107-163. Brosseau, G. E. (1960) Genetics 44, 257-274. 3. Meyer, G. F., Hess, 0. & Beermann, W. (1961) Chromosoma 12, 676-716. 4. Hackstein, J. H. P., Leoncini, O., Beck, H., Peelen, G. & Hennig, W. (1982) Genetics, in press. 5. Hess, 0. (1965) Chromosoma 16, 222-248. 6. Macgregor, H. C. (1980) Heredity 44, 3-35. 7. Hennig, W. (1978) Ent. Germ. 4, 200-210. 8. Hennig, W. (1968) J. Mol. Biol. 38, 227-239. 9. Hennig, W. (1972)J. Mol. Biol. 71, 407-417. 10. Renkhwitz, R. (1978) Chromosoma 66, 225-236. 11. Hennig, W. (1967) Chromosoma 22, 294-357. 12. Hennig, W., Meyer, G. F., Hennig, I. & Leoncini, 0. (1974) Cold Spring Harbor Symp. Quant. Biol 38, 673-683. 13. Glisin, V., Crvenjakov, R. & Byus, C. (1974) Biochemistry 12, 2633-2637. 14. Kafatos, F. C., Weldon Jones, C. W. & Efstradiadis, A. (1979) Nucleic Acids Res. 7, 1541-1552. 15. Birnboim, H. C. & Doly, J. (1979) NucleicAcids Res. 7,1513-1523. 16. Southern, E. M. (1975) J. Mol Biol 98, 503-517. 17. Rigby, P. W., Dieckmann, M., Berg, P. & Rhodes, C. (1977) J. Mol. Biol. 113, 237-251. 18. Thomas, P. S. (1980) Proc. Nati Acad. Sci. USA 77, 5201-5205. 19. Smith, H. 0. & Birnstiel, M. L. (1976) Nucleic Acids Res. 3, 2387-2398. 20. Wensink, P. C., Finnegan, D. J., Donelson, J. E. & Hogness, D. S. (1974) Cell 3, 315-325. 21. Pukkila, P. J. (1975) Chromosoma 53, 71-89. 22. Yamasaki, N. (1977) Chromosoma 60, 27-37. 23. Callan, H. G. & Old, R. W. (1980) J. Cell Sci. 41, 115-123. 24. van Weezenbeek, P. M. G. F., Hulsebos, T. J. M. & Schoenmakers, J. G. G. (1980) Gene 11, 129-148. 25. Carmichael, G. G. & McMaster, G. K. (1980) Methods Enzymol 65, 380-391. 26. Diaz, M. O., Barsacchi-Pilone, G., Mahon, K. A. & Gall, J. G. (1981) Cell 24, 649-659. 27. Meyer, G. F. (1963) Chromosoma 14, 207-255. 28. Meyer, G. F. & Hennig, W. (1974) in The Functional Anatomy of the Spermatozoon, ed. Afzelius, B. A. (Pergamon, New York), pp. 69-75. 29. Glatzer, K. H. (1980) Exp. Cell Res. 125, 519-523. 2.