Chromosoma (Berl.) 83, 145-158 (1981) CHROMOSOMA �9 Springer-Verlag 1981

Distribution of Spacer Length Classes and the Intervening Sequence Among Different Nucleolus Organizers in Drosophila hydei *

Werner Kunz, Gabriele Petersen, Renate Renkawitz-Pohl 1, Karl Heinz G1/itzer and Mireille Schfifer Institut ffir Genetik, Universitfit Dijsseldorf, D-4000 Dfisseldorf, Federal Republic of Germany; 1 present address : Institut ffir Molekulare Genetik, Universit/it Heidelberg, Im Neuenheimer Feld 230, D-6900 Heidelberg 1, Rederal Republic of Germany

Abstract. Drosophila hydei rRNA genes from different chromosomes and from different stocks have been studied by restriction enzyme analysis. In DNA from wild-type females, about half of the X chromosomal rRNA genes are interrupted by an intervening sequence within the 28S coding region. In contrast to D. melanogaster, the intervening sequences belong to a single size class of 6.0 kb. Although there are two nucleolus organizers on the Y chromosome, genes containing the intervening sequence seem to be restricted to the X chromosome. - As shown in four cloned rDNA fragments, the nontranscribed spacers differ in length by having varying numbers of a 242 base pair sequence located in tandem in the right section of the spacer. In genomic rDNA, the spacers also differ in length by a regular 0.25 kb interval. Spacers with between 5 and 15 subrepeats occur frequently within the X and Y chromosomal nucleolus organizers in different D. hydei stocks; shorter and longer spacers are also present but are relatively rare. - Although each genotype is characterized by different frequencies of some spacer classes, the prominent spacer length heterogeneity pattern is similar among the different nucleolus organizers and, therefore, seems to be conserved during evolution.

Introduction

The family of repeated rRNA genes has been shown to be heterogeneous in length in several but not all eukaryotes hitherto investigated (for review see Long and Dawid, 1980). This heterogeneity is attributed to intervening sequences inserted into the 28S rRNA coding region and to varying lengths of the nontran- scribed spacers. In Drosophila melanogaster (Long and Dawid, 1979) and Xenopus laevis (Boseley et al., 1979), it has been shown that different spacer lengths are mainly due to variable numbers of a short repeated sequence in the right portion of the spacer close to the 18S coding region.

* This paper is dedicated to Professor Dr. W. Beermann on the occasion of his 60th birthday

0009-5915/81/0083/0145/$02.80

146 W. Kunz et al.

T h e a n a l y s i s o f c l o n e d f r a g m e n t s o f Drosophila hydei r D N A 1 r e v e a l e d a

s i m i l a r l e n g t h h e t e r o g e n e i t y , d u e b o t h t o a n ivs I a n d to s p a c e r v a r i a t i o n s ( R e n -

k a w i t z - P o h l et al. , 1980). I n t h i s p a p e r we d e s c r i b e t h e d i s t r i b u t i o n o f t h e ivs

a n d o f t h e d i f f e r e n t s p a c e r sizes in g e n o m i c D N A . I n t h e g e n o m e o f D. hydei, t h e r R N A g e n e s a r e c l u s t e r e d a t t h r e e d i f f e r e n t loc i o n e N O 1 is o n t h e

X c h r o m o s o m e , t w o o t h e r s a r e l o c a t e d o n t h e Y c h r o m o s o m e ( H e n n i g et al. , 1975 ;

Sch / i f e r a n d K u n z , 1975). W e c o m p a r e d t h e d i s t r i b u t i o n o f t h e ivs a n d t he

s p a c e r sizes a m o n g n u c l e o l u s o r g a n i z e r s a t d i f f e r e n t c h r o m o s o m a l loci a n d in

d i f f e r e n t s tocks . E s p e c i a l l y we w a n t e d to e l u c i d a t e w h e t h e r s p a c e r s izes d i v e r g e

r a p i d l y in e v o l u t i o n o r w h e t h e r p a r t i c u l a r l e n g t h s a r e c o n s e r v e d .

Material and Methods

Drosophila hydei Stocks. Drosophila was cultivated at 23 ~ C in plastic bottles on a medium containing cornmeal, malt, molasses, soymeal, agar, and brewer's yeast. Laboratory inbred wild-type females as well as several translocation stocks were used in our experiments. Attached-X (XX) females of these translocation stocks contain either a complete Y chromosome (XXY) or only one of the two halves of a Y (~-~y~Ks, ~ y v P ) . The translocation stocks were constructed by O. Hess (1970) and kindly provided for our experiments.

Isolation and Purification of DNA from Flies. Nuclei from frozen adult flies of different genotypes were isolated, and the DNA was extracted and purified as described elsewhere (Renkawitz-Pohl et al., 1980). DNA from single flies was prepared following the method of Endow and Glover (1979) for single brains.

Cloning of D. hydei DNA, Restriction Enzyme Analysis and Hybridization. These methods have been described in detail elsewhere (Renkawitz-Pohl et al., I980). D. hydei rDNA fragments have been cloned in the plasmid vector pBR322 in E. coli strain K 12 HB101. The recombinant plasmid DNA was labelled with ~-32P-dATP (400 600 Ci/mmol) by nick-translation to specific activities of 2-5 x I06 cpm/~g DNA (Maniatis et al., 1975).

Restriction endonuclease digestion of genomic or cloned DNA was carried out in 6.6 mM Tris-HC1 (pH 7.9), 6.6 mM /~-mercaptoethanol, 6.6 mM MgClz, and 60 mM NaC1 for 6-15 h at 37 ~ C. The fragments were separated either on 1% agarose or on 5% acrylamide gels. Molecular weights were determined by including Hind III-digested 2-phage DNA as a standard. DNA was transferred from agarose gels to nitrocellulose filters by the method of Southern (1975). Hybridization of 32P-DNA probes to DNA on filters was done overnight at 65~ in a solution containing 2 x SSC, Denhardt's buffer (1966), and 0.2% SDS.

Electron Microscopy. Heteroduplex formation between EcoRI fragments of cloned rDNA was carried out following the method of Davis et al. (1971). The DNA was denatured in 0.1 N NaOH at room temperature for 15 rain, and then neutralized with HC1. Heteroduplexes were formed in a solution containing 12.5 ~tg DNA/ml, 70% formamide, 100 mM Tris-HC1, and 10 mM EDTA (pH 8.5) at 24~ for 2.5 h. The reaction was terminated by dialysis against 10 mM Tris-HC1, 1 mM EDTA (pH 8.5).

For spreading, a solution containing 0.05 0.1 ~tg heteroduplex DNA/mI, 50% formamide, 0.2 M Tris-HC1, 0.02 M EDTA, and 100 gg/ml of cytochrome c (pH 8.5) was spread onto a distilled water hypophase (Wellauer and Dawid, 1977). For internal calibration, double stranded cpX 174 DNA was added to the spreading solution. Grids were stained with uranyl acetate (Davis et al., 1971), and shadowed with Pt/C. Electron micrographs were taken with a Zeiss EM 9S-2 at 60 kV. The

1 Abbreviations used: NO nucleolus organizer; rDNA ribosomal RNA genes; ivs intervening se- quence; spacer the entire DNA segment between the 28S and 18S coding regions which includes both the nontranscribed spacer and a possible external transcribed spacer region

Spacers and Intervening Sequence in D. hydei 147

negatives were projected with a 15-fold enlargement, and the lengths of the molecules were measured with a Numonics Graphic Calculator.

Results

D. hydei has one NO on the X chromosome and two separate NOs on the Y chromosome, one at the tip of the short arm (TKS-NO), and the other at the tip of the long arm (FP-NO) (Hennig et al., 1975; Schfifer and Kunz, 1975). The X-NO has been studied in females of our wild-type stock (XX~) in XO males whose X chromosome has been propagated independently from the wild-type stock for three years (XO1), and in XO males f rom a stock which has been kept separately for at least twenty years (XO2). The Y chromo- somal NOs were investigated in females from five different translocation stocks which each contain the attached-X chromosome (XX). Since the RX chromo- somes of all five stocks have completely lost their r ibosomal DNA, the XXY females contain only Y chromosomal NOs (Kunz and Schiller, 1976). Three of these stocks (XXYI, XXY2, XXY3) possess a complete Y chromosome including both NOs, the other two ( ~ y r K s , ~-~vP)each contain only one of the two halves of the Y chromosome and therefore only one NO. All " Y - N O stocks" have been kept separately from each other for at least fifteen to twenty years.

A restriction site map of D. hydei r D N A has recently been established by analysis of cloned r D N A (Renkawitz-Pohl et al., 1980) (Fig. 1). The clone pDh2- A4 contains a D. hydei r D N A fragment of 6.0 kb (A-fragment) which includes a minor part of the 18S, the complete 28S coding sequence, and the left fifth of the spacer. The clone pDh2-D7 contains the other large EcoRI fragment of the r D N A repeating unit (B-fragment, 5.1 kb) which includes most of the spacer and the 18S coding region.

In order to compare genomic r D N A from different genotypes, we digested

A - f r a g m e n t B - f r a g m e n t [ I1 i

'=I . , ; H I I ,

s p a c e r 1 8 S

ivs Fig. 1. Summary of the mapping data from cloned rDNA and direct genome analysis (Renkawitz- Pohl et al., 1980). Restriction sites for HpaII and HaeIII were not determined for the entire rDNA unit. The two large EcoRI fragments of the rDNA repeating unit (A-fragment and B-fragment), which have been cloned, are indicated. BamHI (v), EcoRI (~), HaeIII ('[), HindIII (~), HpaII(I), PstI (~), SalI (T), XbaI (~)

148 w. Kunz et al.

nuclear DNA with different restriction enzymes. The DNA fragments were separated by agarose gel electrophoresis, transferred to nitrocellulose filters according to the procedure of Southern (1975), and hybridized separately with 32p-rDNA from the clones pDh2-A4 and pDh2-DT, respectively. Hybridization with the A-fragment has been used to demonstrate heterogeneity within the 28S sequence, whereas hybridization with the B-fragment has been used to show spacer length variation.

Intervening Sequence in Genomic rDNA

With EcoRI-digested genomic DNA from flies of genotypes containing only Y chromosomal NOs (X-XY1, ~-~yTKS, ~"~yFP), the 6.0 kb cloned A-fragment hybridizes only with a band of equal size. In EcoRI-digested DNA from females containing the X chromosomal NO, this 6.0 kb band is also present, but the cloned A-fragment hybridizes additionally with bands of 8.5 and 2.4 kb (Fig. 2a). EcoRI digestion of the D. hydei rDNA clone pDh2-H8 produces the same two bands: Airs1 and Aivs2, respectively (Renkawitz-Pohl et al., 1980). Analysis of this clone has shown that both bands result from the insertion of an ivs of 6.0 kb containing additional EcoRI sites within the 28S gene. A third EcoRI fragment known from the clone pDh2-H8 (1.1 kb), which is located within the right section of the ivs, is not visible in Figure 2a, because it does not share sequences with the A-fragment used as a hybridization probe. To demon- strate this fragment in genomic DNA, we could not use the ivs from the pDh2-H8 clone as a hybridization probe, because this sequence hybridizes to several other DNA sequences located outside the nucleolus organizer (unpublished results).

All visible EcoRI-fragments appear as sharp single bands. This indicates that within the limits of resolution of our autoradiograms the ivs is constant in length within a population of rDNA repeating units in the X chromosomal NO. However, with only the EcoRI digestion experiments we cannot exclude a possible length heterogeneity within the short 1.1 kb fragment at the right end of the ivs. The restriction enzyme PvuII allows us to decide whether this ivs-region also is constant or whether it varies in length. PvuII digests chromo- somal rDNA into four fragments: 9.5, 4.5, 1.6, and 1.1 kb (Fig. 2b). A mapping through the use of our pDh2-H8 clone has shown that the 4.5 kb fragment contains more than half of the ivs (including the 1.1 kb EcoRI fragment), the 28S/32 coding sequence, and some two to three hundred bases at the 5' end of the spacer (data not shown). As shown after PvuII digestion of XX DNA and hybridization to the cloned A-fragment, only the spacer containing 9.5 kb band varies in size, whereas the 4.5 kb fragment appears as a sharp band (Fig. 2b). This demonstrates that, in contrast to D. melanogaster (Wellauer and Dawid, 1978), only a single length class of intervening sequences occurs inserted in the rDNA of D. hydei and it is restricted to the X chromosomal NO. Prelimi- nary data indicate that like in D. virilis (Barnett and Rae, 1979) the ivs may be present as a tandem duplication in some of the X chromosomal rRNA genes, since the restriction enzyme HpaI that cuts once in the rDNA unit but not in the ivs, produces fragments of both 17 and 23 kb that hybridize with the A-fragment.

Spacers and Intervening Sequence in D. hydei 149

Fig. 2a, b. Fragments of ribosomal DNA from flies of D. hydei genotypes containing only the X (a: left, and b) or the Y chromosomal NOs (a: right) hybridized with the A-fragment. The DNA has been digested with EcoRI (a) or PvuII (b), separated on I% agarose gels, transferred to nitrocellulose filters, and hybridized with nick-translated 3zp DNA from clone pDh2-A4. Only the X chromosomal NO shows ivs-containing bands: 8.5 and 2.4 kb in the case of EcoRI; 4.5 and 1.1 kb in the case of PvulI. Note that these bands do not vary in size

Fur thermore , we were interested to know the percentage of ivs-conta ining genes within the X-NO. Therefore, we double digested the r D N A from wild-type females with EcoRI and PstI. PstI shortens the EcoRI -A- f ragmen t f rom 6.0 to 5.5 kb, whereas the Aivsl-fragment is shortened from 8.5 to 4.8 kb (see Fig. 1). The 5.5 kb f ragment results f rom r R N A genes without an ivs; it includes a part of the internal t ranscribed spacer, the 28S coding region, and a part of

150

O.D.

5 ! i

5.5 4,8 kb

W. Kunz et al.

Fig. 3. Microdensitometer tracing of an autoradiogram of EcoRI-PstI double digested DNA fron'l wild-type females. The DNA fragments were electrophoresed on 1% agarose, transferred, and hybridized with 32P-DNA from the clone pDh2-A4 which contains the A-fragment. The left band (5.5 kb) results from rRNA genes without ivs, the right band (4.8 kb) represents the respective fragment from ivs- containing genes. 100% of the left band and 69% of the right band hybridize with the A-fragment, After correction for these different hybridization percentages, it has been calculated that 45% of the rRNA genes in the X-NO contain the ivs

the nontranscribed spacer. This fragment is complementary to the A-fragment throughout its entire length. The 4.8 kb fragment represents the respective frag- ment from ivs-containing genes, including, in addition to parts of the internal transcribed spacer and the 28S region, a left portion of the ivs; 69% of this fragment is complementary to the A-fragment. Both fragments differ in size by only 0.7 kb. The use of this double digestion reduces the danger of an un- equal transfer to nitrocellulose of fragments having large differences in molecular weight.

After electrophoresis, transfer, and hybridization with the A-fragment, au- toradiograms were selected which gave a linear response to increasing exposures when traced on a Joyce-Loeble Chromoscan 200 (Fig. 3). Taking the different hybridization percentages into consideration, it has been calculated that 45% of the rRNA genes in the X-NO contain the ivs. However, this percentage refers to the DNA from whole flies containing diploid as well as polyploid tissues. The number of ivs-containing genes in diploid tissue is actually much larger, since in polyploid tissue ivs-genes are more underreplicated than the rest of the rRNA genes (for D. melanogaster: Endow and Glover, 1979; for D. hydei: our own unpublished results).

Spacer Length Heterogeneity in D. hydei rDNA

In contrast to the constant length of the A- and the AivJragments within genomic rDNA, the B-fragment varies considerably in size (Fig. 4). The major B-bands of EcoRI-digested rDNA of different genotypes have lengths from 4.5 to 5.1 kb.

This length heterogeneity reflects either a heterogeneity among the rRNA genes in a single fly, or the heterogeneous pattern within the population would result from the combination of different flies having distinct spacer length pat- terns. To decide between these alternatives, we analysed the rDNA of single flies (Fig. 5). In addition, we compared fly pools from different lines which were started from single pair matings from the same population. No differences in rDNA pattern were detected between individual flies and between different lines. This demonstrates that the spacer length heterogeneity pattern of single

Fig. 4. Patterns of spacer length heterogeneity from flies of eight different genotypes of D. hydei. Each DNA sample was prepared from a large number of flies. DNA was digested with EcoRI (left) or HaeIII (right), separated on 1% agarose gels, transferred to filters, and hybridized with 3zP-labelled DNA from clone pDh2-D7 which contains the B-fragment. Only those bands which include the variable part of the spacers are visible in this figure; some small HaeIII fragments of the 18S region which also hybridize with the B-fragment are not shown. The HaeIII fragments include almost the entire spacer (see Fig. 1); the EcoRI bands include additionally most of the 18S region, but this does not vary in length (see Fig. 6). The EcoRI bands are 0.3 kb larger than the HaeIII fragments; in the figure, they are lined up with the respective HaeIII bands. The spacer classes, which represent stepwise integral multiples of an 0.25 kb interval, are indicated

XX X'-'~y I ~'~yTKS ~'~yFP

Fig. 5. Length pattern of B-fragments of the DNA from single flies of four different genotypes. The DNA was digested with EcoRI, fractionated on I% agarose, transferred to a filter, and hybridized with 32p-DNA from pDh2-D7 which contains the B-fragment. The main B-band (Bg) is indicated. All clearly visible bands are the same in single flies as in fly pools from whole populations (compare with Fig. 4)

152 W. Kunz et al.

Fig. 6. a EcoRI B-fragments of different lengths are present in the clones. D. hydei chromosomal DNA (right) and DNA of five clones were digested separately with EcoRI, fractionated on 1% agarose, transferred to nitrocellulose, and hybridized with 32P-DNA from pDh2-D7 which contains the B-fragment. The lower band, which is constant in length, is the vector pBR322 (4.4 kb) (v); the other bands increasing stepwise in size from the left (4.1 kb) to the right (5.1 kb) are the different rDNA B-fragments. The clone pDh2-B6 contains two B-fragments. b DNA pattern after double digestion with EcoRI and HaeIII of the same clones as in A (stained with ethidium bromide). c DNA pattern after digestion of the clone pDh2-D7 and the vector pBR322 with HpaII. The DNA fragments were separated on 5% acrylamide gels and stained with ethidium bromide

flies is representa t ive of the whole popu la t ion . Therefore , in some of the geno- types inves t iga ted we d id no t select ind iv idua ls bu t s tudied the D N A f rom a large n u m b e r o f flies.

Leng th va r i a t ion mus t be res t r ic ted to the midd le or the r ight sect ion of the spacer and does no t occur within the shor t 1.1 kb segment in the left pa r t o f the spacer ad jacen t to the 28S sequence, because the A - f r a g m e n t in E c o R I digests is cons tan t in length (Fig. 2a).

The length he te rogene i ty o f the B- f ragment m a y reflect ei ther length differ- ences be tween different r D N A repea t units or add i t i ona l res t r ic t ion sites in some o f the units. To decide between these a l ternat ives , five different clones con ta in ing E c o R I B-f ragments o f different lengths were invest igated. F o u r c lones con ta in one B-f ragment each (pDh5-E4, pDh4-C10, pDh4-F11 , and pDh2-D7) ,

Spacers and Intervening Sequence in D. hydei 153

and one clone includes two B-fragments (pDh2-B6). Two of these clones (pDh2-D7 and pDh2-B6) have been characterized in detail previously (Renkawitz-Pohl et al., 1980). The B-fragments in the five clones vary in length in regular intervals of about 0.25 kb (Fig. 6a). The shortest fragment is 4.1 kb (in pDh2-B6), the longest 5.1 kb (in pDh2-D7 and pDh2-B6). A comparison with total D. hydei DNA (right in Fig. 6a) shows that fragments of the same sizes represent t h e major B-fragment classes in genomic DNA.

To localize the region of length heterogeneity within the B-fragment, all five clones were double digested with EcoRI and HaeIII. HaeIII cuts off the 18S region from the B-fragment; therefore, the products of double digestion, as shown in Figure 6b, include only the spacer. This experiment shows that the length variation does not occur in the 18S coding region but is located in the spacer segment of the B-fragment.

An analysis of the clone pDh2-D7 by partial digestion has shown that nine Hpa II sites follow each other in tandem (Renkawitz-Pohl et al., 1980). This regular order of HpaII sites starts a distance of 2.1 kb from the 3' end of the 28S coding sequence and ends 1.0kb before the 5' end of the 18S coding region (Fig. ~). Our clones of different spacer lengths offered the possibili- ty to investigate whether varying numbers of HpaII fragments may be the basis for the observed length variation. Digestion of the B-fragments from four clones (pDh5-E4, pDh4-C10, pDh4-F11, and pDh2-D7) with HpaII demon- strates the existence of a 242 bp fragment whose staining intensity indicates a repeated occurrence (Fig. 6c). To determine the number of these segments within the clones, the ethidium bromide stained gels were photographed, and the negatives were scanned with a Chromoscan 200 (Joyce-Loeble) (Fig. 7). The HpaII bands of the vector (pBR322) were used as a reference so that DNA concentrations and exposure time were within the linear range of film response. This quantification has indeed shown that clones with longer spacers include proportionally more 242 bp segments than shorter ones. Since clone pDh2-D7 has nine HpaII sites in tandem, it can be calculated that pDh4-F11 contains eight, pDh4-C10 seven, and pDh5-E4 six HpaII sites in regular 242 bp intervals. Variation in number of 242 bp segments is sufficient to account for the length differences of the four spacers.

We were further interested in investigating whether the 242 bp segments within the spacer are homologous subrepeats. Therefore, we compared the spacers of the two clones pDh5-E4 and pDh2-DT, which differ in length by 0.7 kb. Electron microscopic analysis of heteroduplex molecules formed from the B-fragments of these two clones shows that the spacers are indeed homolo- gous. A deletion loop of 0.7 kb in length was observed; its position varied throughout the length of a region from about 500 to 2,000 bp from the short end of the heteroduplex molecule (Fig. 8). This region corresponds to the range where the HpaII restriction sites occur in regular order. Since no unpaired regions have been observed, we conclude that the 242 bp segments represent a cluster of tandem internal sequence repetitions in that part of the spacer.

Among different nucleolar organizers in genomic DNA, the most frequent length class is a spacer of 5.1 kb. This spacer is called the B 9 spacer, because it has the same length as that of the clone pDh2-D7 which contains nine

154 W. Kunz et al.

pDh5-E4

/

JU Jt_J <

pDh4-F11

p D h 2 ~

pDh4-ClO

Fig. 7. Microdensitometer tracings of the DNA patterns of the plasmid pBr322 and of four recombi- nant clones after digestion with HpaII and electrophoresis on 5% acrylamide gels. Note that the 242 base pair band (arrow) varies in relative amount among the different clones. Differences in the other bands are explained by additional sequences besides the B-fragment in pDh4-C10 and by an opposite orientation of pDh2-D7 within the vector

Spacers and Intervening Sequence inD. hydei 155

40

ul 13. o o 30 r

o

20 7D

E 10 Z

I-I J I

0.5

I I I I I

1.0 2.0 3.0 4.0 Posit ion of delet ion Loops (kb)

Fig. 8. About 230 heterodupiex molecules formed with the B-fragments from pDh5-E4 and pDh2-D7 (the insert shows one example) have been measured with respect to the location of the deletion loop. The total length of the abscissa (4.4 kb+0.45) represents the length of the heteroduplex molecule. Deletion loops have been observed in a range between about 0.5 and 2.0 kb from the short end of the molecule (arrows). The bar in the insert represents 1.0 kb

subrepeats. Other spacer classes in genomic D N A differ in length from this main class by integral multiples of an 0.25 kb interval. Therefore, in genomic r D N A like in the clones the different spacer size classes seem to be based on varying numbers of internal subrepeats. Four frequent spacer classes are shorter than the B 9 spacer: Bs, BT, B6, and B5 (in each case the index designates the number of postulated internal subrepeats) (Fig. 4). In some particular geno- types, additional smaller spacer classes have been detected at the positions B,, B3, and B2, but these spacers occur at a very low frequency.

Among the spacers larger than B9, six classes have frequently been found: Blo, Bll, B12, B13, B14, and B~s. Some additional spacer classes that are even larger (up to 23 kb in length) have been detected in most of the genotypes. Assuming that these extremely large spacers are also related to the main spacer by a summation of 0.25 kb subrepeats, a 23 kb spacer would contain as much as 80 subrepeats. Spacers of comparably large sizes have also recently been detected in the r D N A of D. melanogaster (Indik and Tartof, 1980).

Discussion

The spacers of D. hydei and D. melanogaster are very different in sequence as shown by cross hybridization experiments and by a comparison of restriction enzyme patterns (Renkawitz-Pohl et al., 1980). But the spacers of both species resemble each other in having a regular 0.25 kb interval between sites recognized

156 w. Kunz et al.

by a particular restriction enzyme; this is AluI in D. melanogaster (Long and Dawid, 1979) and HpaII in D. hydei. Both, the size of the interval (0.25 kb) and the most frequent number of sequence repetitions (between 6 and 9), are the same in the two species. Furthermore, in both cases the length heterogeneity in spacers is due, at least in part, to varying numbers of these repeated sequence elements.

A varying number of internal subrepeats as the basis of spacer length varia- tion has also been shown in the rDNA of other species including Xenopus (Boseley et al., 1979), mouse (Grummt and Gross, 1980), and Calliphora (Schfifer et al., in press). Also, the spacers of some oocyte-specific 5S RNA gene families in Xenopus differ in size due to varying numbers of copies of a tandemly repeated simple sequence DNA (Fedoroff and Brown, 1978). Since some somatic 5S RNA gene families have unique sequence spacers that are homogeneous in length (Petersen et al., 1980), a general correlation between length heterogene- ity and the occurrence of internal subrepeats can be made.

In contrast to Xenopus laevis and D. melanogaster, spacer length variation in D. hydei appears to be generated by varying numbers of internally repeated sequences in only one area of the spacer. Therefore, different spacer lengths are directly related to the respective numbers of these tandemly repeated sub- units. Three major spacer size classes containing 6, 8, and 9 subrepeats (B6, Bs, and B9) have been found in similar relative frequencies not only in all Y chromosomal NOs but also in the X-NO of D. hydei. The only exception is the absence of the B 9 spacer in one of the eight investigated genotypes (XO2). In the three genotypes, containing only X chromosomal NOs, two addi- tional spacers are prominent: B7 and B~o. Both are represented in the Y-NO only at a low frequency or not at all. It is interesting to note that these two spacer classes are mainly linked to genes containing the ivs (Renkawitz-Pohl et al., in press).

The other spacer classes occur less frequently but are not distributed random- ly around the main spacer bands. In most genotypes a ladder pattern of B~I to B15 spacers is present, but individual bands of this ladder differ in frequency or are missing among the different genotypes. Stepwise integral multiple lengths do not occur above the B15 spacer size, however, each genotype has a few bands of much larger size. It is striking that a special size class of spacers of ap- proximately 9.5 kb is rather frequent in four of the Y-NO genotypes (Fig. 4).

Whereas the spacer length heterogeneity pattern in the XXY genotypes repre- sents always the summation of the patterns of both Y chromosomal NOs, the TKS-NO and the FP-NO could be studied separately from each other in the genotypes XXY TKs and XXY rP, respectively. A comparison between the two nucleoli reveals that the two major spacers B9 and B6 do not differ significantly in frequency, but the third prominent spacer, Bs, is more abundant in the FP-NO than in the TKS-NO. Furthermore, there are differences between these two NOs in the pattern of the Ba~ to B~5 spacers and also in the larger spacers.

It seems possible that different spacer lengths have evolved by duplication or deletion of some of the internal subrepeats due to unequal sister chromatid exchange. In yeast, a high level of unequal recombination occurs between ribo-

Spacers and Intervening Sequence in D. hydei 157

somal D N A gene clusters on sister chromatids (Petes, 1980; Szostak and Wu, 1980). If such events were frequent within the rDNA clusters in D. hydei, various spacer length classes would be generated continuously so that particular spacer sizes would disappear, and new ones would emerge.

However, our results show that only a limited range of variation in spacer size is tolerated. Although a comparison of the pattern of spacer length hetero- geneity reveals some specific differences among the different nucleolus organiz- ers, the prominent spacer lengths are conserved during evolution.

Acknowledgments. This research was supported by the Deutsche Forschungsgemeinschaft by grant Ku 282/7. The experiments shown in Fig. 3 have been carried out by G. Franz. We thank Mrs. S. Drewniok for valuable technical assistance, Prof. O. Hess for providing us with the appropriate Drosophila stocks and for excellent working facilities, and Mrs. J. Morris for help in translating the manuscript.

References

Barnett, T., Rae, P.M.M. : A 9.6 kb intervening sequence in D. virilis rDNA, and sequence homology in rDNA interruptions of diverse species of Drosophila and other diptera. Cell 16, 763 775 (1979)

Boseley, P., Moss, T., M/ichler, M., Portmann, R., Birnstiel, M.: Sequence organization of the spacer DNA in a ribosomal gene unit of Xenopus laevis. Cell 17, 19-31 (1979)

Davis, R.W., Simon, M., Davidson, N. : Electron microscope heteroduplex methods for mapping regions of base sequence homology in nucleic acids. In : Methods in Enzymology (L. Grossman, K. Moldave, eds.) vol. 21D, pp. 413-428. New York: Academic Press 1971

Denhardt, D.T. : A membrane-filter technique for the detection of complementary DNA. Biophys. biochem. Res. Comm. 23, 641~46 (1966)

Endow, S.A., Glover, D.M.: Differential replication of ribosomal gene repeats in polytene nuclei of Drosophila. Cell 17, 597~05 (1979)

Fedoroff, N.V., Brown, D.D.: The nucleotide sequence of oocyte 5S DNA in Xenopus laevis. I. The AT-rich spacer. Cell 13, 701-716 (1978)

Grummt, I., Gross, H. : Structural organization of mouse rDNA: Comparison of transcribed and nontranscribed regions. Molec. gen. Genet. 177, 223-229 (1980)

Hennig, W., Link, B., Leoncini, O. : The location of the nucleolus organizer regions in Drosophila hydei. Chromosoma (Berl.) 51, 57 63 (1975)

Hess, O.: Genetic function correlated with unfolding of lampbrush loops by the Y chromosome in spermatocytes of Drosophila hydei. Molec. gen. Genet. 106, 328 346 (1970)

Indik, Z.K., Tartof, K.D.: Long spacers among ribosomal genes of Drosophila melanogaster. Nature (Lond.) 284, 477-479 (1980)

Kunz, W., Schgfer, U. : Variations in the number of the Y chromosomal rRNA genes in Drosophila hydei. Genetics 82, 25-34 (1976)

Long, E.O., Dawid, I.B. : Restriction analysis of spacers in ribosomal DNA of Drosophila melano- gaster. Nucleic Acids Res. 7, 205~15 (1979)

Long, E.O., Dawid, I.B. : Repeated genes in eukaryotes. Ann. Rev. Biochem. 49, 727-764 (1980) Maniatis, T., Jeffrey, A., Kleid, D.G. : Nucleotide sequence of the rightward operator of phage 2.

Proc. nat. Acad. Sci. (Wash.) 72, 1184 1188 (1975) Peterson, R.C., Doering, J.L., Brown, D.D. : Characterization of two Xenopus somatic 5S DNAs

and one minor oocyte-specific 5S DNA. Cell 20, 131-142 (1980) Petes, T.D. : Unequal meiotic recombination within tandem arrays of yeast ribosomal DNA genes.

Cell 19, 765-774 (1980) Renkawitz-Pohl, R., Gl~itzer, K.H., Kunz, W.: Characterization of cloned ribosomal DNA from

Drosophila hydei. Nucleic Acids Res. 8, 4593-4611 (1980) Renkawitz-Pohl, R., Glfitzer, K.H., Kunz, W. : Ribosomal RNA genes with an intervening sequence

are clustered within the X chromosomal rDNA of Drosophila hydei. J. molec. Biol. (in press, 1981)

158 W. Kunz et al.

Sch/ifer, U., Kunz, W. : Two separated nucleolus organizers on the Drosophila hydei Y chromosome. Molec. gen. Genet. 137, 365-368 (1975)

Sch/ifer, M., Wyman, A.R., White, R. : Length variation in the nontranscribed spacer of Calliphora erythrocephala ribosomal DNA is due to a 350 base pair repeat. J. molec. Biol. (in press, 1981)

Southern, E.M. : Detection of specific sequences among DNA fragments separated by gel electro- phoresis. J. molec. Biol. 98, 503 517 (1975)

Szostak, J.W., Wu, R. : Unequal crossing over in the ribosomal DNA of Saccharomyces cerevisiae. Nature (Lond.) 284, 426-430 (1980)

Wellauer, P.K., Dawid, I.B. : The structural organization of ribosomal DNA in Drosophila melano- gaster. Cell 10, 193-212 (1977)

Wellauer, P.K., Dawid, I.B. : Ribosomal DNA in Drosophila melanogaster. II. Heteroduplex map- ping of cloned and uncloned rDNA. J. molec. Biol. 126, 769-782 (1978)

Received September 2, 1980-January 12, 1981 / Accepted by W. Beermann Ready for press February 12, 1981