J. Mol. Bid. (1971) 50, 555-563
Studies of Fractionated II.
HeLa Cell Metaphase Chromosomes
Chromosomal Distribution of Sites for Transfer RNA and 5 s RNA?
YOSEF ALONI, LOREN E. J~ATLEN AND GIUSEPPE ATJXRDI
Division of Biology, Cdijorniu In&i&de of Technology Pasadena, Calij. 91109, U.S.A. (Received17 August 1970, and in revisedjomn 11 November1970) DNA extracted from HeLa cell metaphase chromosomes fractionated on the basis of sedimentation velocity in glycerol-sucrose gradients has been tested for the capacity to hybridize with highly purified tRNA$ and 5 s RNA. The results obtained indicate that the sites for these two RNA classes, in contrast to those for the high molecular weight rRNA, are distributed among chromosomes of all
1. Introduction The development of methods for the isolation of intact metaphase chromosomesfrom mammalian cells and their fractionation on the basis of sedimentation velocity (Huberman & Attardi, 1967; Maio t Schildkraut, 1966, 1969; Mendelsohn, Moore t Salzman, 1968) has opened the way for investigating the chromosomal distribution of certain types of genes either by molecular hybridization with the corresponding primary gene products or by analysis of their physical properties. By using the former mentioned approach, the chromosomal distribution of the genesfor the high molecular weight rRNA in HeLa cells was previously analyzed (Huberman & Attardi, 1967). It was found that the DNA complementary to the high molecular weight rRNA is confined to the smaller HeLa cell chromosomes, which include those carrying a nucleolar organizer. In the preceding paper (Hatlen & Attardi, 1971), evidence has been presented for redundancy of the DNA complementary to tRNA and 5 s rRNA in HeLa cells. In the present paper, RNA-DNA hybridization experiments have been carried out with DNA from fractionated HeLa cell metaphasechromosomesin order to investigate the chromosomal distribution of the redundant DNA sites for tRNA and 5 s RNA. The results obtained indicate that, in contrast to the genesfor the high molecular weight rRNA, the DNA complementary to 4 s RNA0 and 5 s RNA is distributed in chromosomesof all size ranges. t Part I was Huberman & Attardi, 1967. $ Abbreviations used: tRNA, transfer RNA;
dodecyl sulfate. 8 In the present work the expressions 4 s RNA although the assumption that all 4 s RNA isolated the tRNA class has not been proved. 36
and tRNA have been used from the cytoplasm of HeLa
interchangeably cells belongs
L. E. HATLEN
2. Materials and Methods (a) Cells and methoda of gmwth The method of growth of HeLa cells (S3 clone1 strain) in suspension previously (Amaldi & Attardi, 1968).
has been described
(b) Chromoeme isolation and fractionation Chromosomes were isolated from cells blocked in meta.phase with vinblaatine sulfate and then fractionated in glycerol-sucrose gradient according to the procedure previously described (Huberman & Attardi, 1967). (c) Isolation and denatura;tion of DNA DNA was prepared from fractionated chromosomes by a modification of the Marmur procedure (Marmur, 1961). Suspensions of chromosomes in 0.026 M-EDTA (pH 8.0) 0.037 M-N&~ were homogenized with a tight-fitting Dounce homogenizer (Knotas Glass Co., Vineland, N.J.) to break up clumps, and then incubated with 500 pg pronase/ml. (Calbiochem, B grade, 45,000 units/g) and 0.1% dodecyl SO4 at 37’C for 10 hr. At the end of the incubation, the dodecyl SO4 concentration was raised to 0.5% and the isolation of the DNA continued according to the Marmur procedure. Denaturstion of the DNA and removal of RNA was carried out by dialysis against 0.3 N-KOH for 18 hr at room temperature followed by dialysis for 48 hr against 2 x SSC (SSC is O-15 M-N&I, 0.015 M-sodium citrate.) (d) Labeling conditions and isolation of RNA Reference is made to the preceding paper (Hatlen & Attardi, 1971) for the methods used to prepare highly purified 4 s and 5 s RNA from cells labeled with [32P]orthophosphate. The specific activity of the find preparations varied between 7 x lo5 and lo6 cts/min/pg. For labeling high molecular weight rRNA with [3H]uridine, HeLs cells (6 x104/ml.) were incubated for 48 hr in the presence of 2.6 ,&nl. [6-3H]Uridine (Radiochemical Centre, Amershrtm, 23 c/m-mole), then subjected to a chase for 18 hr with 2 x 10e3 M-unlabeled uridine. The 28 s and 18 s RNA components were isolated from pursed ribosomal subunits as previously described (Am&i t Attardi, 1968). To eliminate any possible DNA contaminant, the RNA samples were treated with pancreatic DNaee (20 pg/ml.) in 0.05 M-Tris buffer (pH 6.7 at 25”C), 0.025 M-KCl, O-0025 M-Mgcl, for 60 min at 22’C, re-extracted with dodecyl SO,-phenol, precipitated with ethanol, dissolved in 2 x SSC and run on 55 cm columns of Sephadex GlOO equilibrated with 2 x SSC st 22°C. The specific activity of the 3H-labeled high molecular weight rRNA was 5 x lo4 cts/min/pg. (e) RNA-DNA hybridization procedure Hybridization of 3aP-labeled 4 s or 5 s RNA to DNA from fractionated chromosomes and isolation of RNA-DNA hybrids were carried out as described in the preceding paper (Hatlen & Attardi, 1971); in all cases, incubation was for 4hr at 72”C, and RNaee digestion was carried out with 10 pg of enzyme/ml. for 1 hr at 22°C. Each incubation mixture contained, as an internal control for efficiency of chromosome fractionation and availability of DNA sites for hybridization, 3H-labeled 28 s RNA or both 3H-labeled 28 8 and 3H-labeled 18 s RNA, as specsed below. As a control for non-spec& background, denatured DNA was incubated separately from labeled RNA and mixed with it before the RNese digestion step. All the hybridization values reported here have been corrected for the background, determined as described above.
3. Results (a) Chromosome jracttiim Hubermen & Attardi (1967) have shown that isolated HeLa cell m&phase chromosomescan be fractionated on the basisof sedimentation velocity in a glycerolsucrosegradient. The results of one such fractionation, in terms of chromosomedis-
5 s RNA
tribution, are shown in Figure l(a). Phase contrast microscopy revealed that an effective separation of the chromosomes on the basis of size was achieved by this procedure: most of the chromosomes appeared to be single, although some clusters of small chromosomes were seen among the larger chromosomes. In most experiments, the chromosome distribution in the glycerol-sucrose gradient was arbitrarily cut into four sections, as shown in the inserts of Figures 2 and 4, and the chromosomes of each class were collected by centrifugation and used for DNA extraction. In the experiment shown in Figure l(a), the faster and slower sedimenting halves of the chromosome distribution were separately pooled, as indicated by arrows, and subjected to a second cycle of glycerol-sucrose gradient centrifugation (Fig. 1(b) and (c)). Notice that the heavier chromosomes tend to sediment more slowly in the second fractionation: this may result from some change occurring in these chromosomes during the additional operations required for the rerun. The two chromosome distributions obtained in the second fractionation were each divided into two approximately equal cuts for DNA extraction, as shown in Figure l(b) and (c). To provide sufficient DNA for the hybridization experiments described in this paper, corresponding chromosome classes from two separate fractionations were pooled for direct DNA extraction or for a second cycle of glycerol-sucrose density gradient centrifugation.
Fm. 1. Distribution of chromosomes after centrifugation through a glycerol-sucrose density gradient (a), and recentrifugation of chromosome classes through the same type of gradient (b) and (a). Chromosomes were isolated from about 1.6 x 10s HeLa cells layered onto 2 separate 140~ml. linear gradients from 30% (w/w) glycerol in 0.02 M-Tris buffer, pH ‘7.0, 0*002 ~-c&l,, 0.06% Saponin (at the top) to 30% ( w / w ) sucrose in the same medium (at the bottom), and centrifuged for 40 mm at 460 g (4°C) (Huberman BEAttardi, 1967). 4-ml. fractions were collected through thin glass tubing inserted at the bottom of the gradient, and the chromosome concentration in even numbered fractions was determined by counting in a bacterial counting chamber. In (a) the distribution of chromosomes in one of the gradients is shown. The cut-off points for pooling the fractions into 2 chromosome classes are indicated by arrows. (b) The faster sedimenting chromosomes (I) from 2 gradients a8 in (a) were pooled and recentrifuged as described above. (c) The slower sedimenting chromosomes (II) from two gradients were pooled and recentrifuged as described above.
with 4s RNA
As mentioned in Materials and Methods (section (e)), all hybridization mixtures contained 3H-labeled 28 s RNA at an RNA/DNA ratio of 1 : 20 either alone or in combination with 3H-labeled 18 s RNA at an RNA/DNA ratio of 1 : 50. These ratios are sufficient to saturate the rRNA sitesin HeLa cell DNA (Jeanteur & Attardi, 1969). Since it was shown previously (Huberman & Attardi, 1967) that the DNA complementary to rRNA is confined to the smaller HeLa cell chromosomes,which include those carrying a nucleolar organizer, the 3H-labeled rRNA speciesprovided a convenient internal control to determine the etliciency of the chromosomefractionation as well as the availability for hybridization of the DNA purified from the various chromosomefractions. In the experiment presented in Figure 2, 3H-labeled 28 s was
~~~~-~~ I” 0
Input RNA/DNA lb. 2. Saturation curves of DNA from fractionated chromosomes by HeLa cell 32P-labeled 4 s RNA. The incubation mixtures contained 20 pg chromosome1 DNA, various amounts of HeLa cell 3’P-labeled 4 s RNA and 1 pg HeLa cell sH-labeled 28 s RNA in I.0 ml. 2 x SSC. In the insert, the distribution of chromosomes after centrifugation through a glycerol-sucrose density gradient is shown (----). The cut-off points for pooling the fractions into separate chromosome classes are indicated by arrows. - xx-, Saturation level by 3aPelabeled 4 s RNA of the DNA saturation level by 3H-labeled extracted from the various chromosome classes; [email protected]
, 28 s RNA of the DNA extracted from the various chromosome classes.
used as an internal control. As shown in the insert, the smallest chromosome(class4) contains per unit DNA four times more DNA sites complementary to 28 s RNA than the DNA extracted from the largest chromosomes(class 1). l?rom the hybridization values obtained with DNA extracted from the four chromosome classes,after normalization for the relative proportion of each class, an average DNA saturation value for 28 s RNA of about 2.0 x lob4 was calculated: this figure is in good agreement with the saturation level of total HeLa cell DNA by 28 s RNA previously found (1.8 x 10e4, Jeanteur & Attardi, 1969). In contrast to the hybridization values obtained with 28 s RNA, the saturation levels for 4 s RNA appear to befairly constant in the DNA from the various chromosome classes,ranging from 1.0 x loss for the
5 s RNA
chromosomes of class 3 to about 1.2 x 10m5 for the chromosomes of class 1, with an average of 1.1 x 10e5, as found for total HeLa cell DNA (see preceding paper). The curves of saturation by 32P-labeled 4 s RNA of DNA purified from various chromosome classes are shown in Figure 2. The saturation curves rise slowly and finally plateau around an input RNA to DNA ratio of O-040. The results of the hybridization experiments between 32P-labeled 4 s RNA and the DNA extracted from the fractionated chromosome classes obtained after two cycles of centrifugation in glycerol-sucrose density gradients, as described in Figure 1, are hown in Figure 3. In this experiment, both 3H-labeled 28 s RNA and 3H-labeled 18
input RNA/DNA FIQ. 3. Saturation curves of DNA from recentrifuged chromosome classes by HeLa 32P-labeled 4 s RNA. The incubation mixtures contained 20 pg chromosomal DNA (extracted from the 4 chromosome classes obtained after 2 cycles of centrifugation in glycerol-sucrose density gradients as described in Fig. l), various amounts of HeLa 32P-labeled 4 s RNA, 1 pg 3H-labeled 28 s RNA and 0.4 pg sH-labeled 18 s RNA in 1.0 ml. 2 x SSC. In the insert, the saturation levels by esP-labeled 4 s RNA (- XX-) and by combined sH-labeled 28 s RNA and 3H-labeled 18 s RNA (-.---.-) are shown.
s RNA were used as an internal control. The DNA purified from the smallest chromosomes (class 4) gave a level of hybridization with these RNA species which is about five times higher than that obtained with the DNA extracted from the largest chromosomes (class 1) ; this suggests that a small improvement in chromosome fractionation was achieved by subjecting the chromosomes to a second cycle of centrifugation in a glycerol-sucrose density gradient, The weighted average for the saturation levels of DNA purified from the various chromosome classes by the combined 28 s and 18 s RNA was found to be 2.3 x 10W4, in good agreement with the summation of the saturation levels of HeLa cell DNA by these RNA species previously reported (2.5 x 10e4, Jeanteur & Attardi, 1969). It can be seen from Figure 3 that recentrifugation of the chromosomes in a glycerol-sucrose density gradient did not appreciably change the distribution of the chromosomes carrying the DNA sites complementary to 4 s RNA. The saturation levels ranged between about 1.2 x 10e5 for DNA purified from the chromosomes of class 3, and 1.4 x lob5 for DNA purified from the chromosomes of classes 1 and 4.
5 B RNA
Figure 4 showsthe saturation curves by 32P-labeled5 s RNA of DNA purified from the four chromosomeclassesindicated by arrows in the insert. In comparison with the saturation curves by 4 s RNA, those obtained with 5 s RNA rise fairly steeply and level off already at an input RNA to DNA ratio of 0.008, as observed with unfractionated DNA (see preceding paper). The DNA saturation levels range between 2-O x 10e5 for the chromosomesfrom class 3 and 2.5 x 10m5for the chromosomesof class1. The weighted average of the saturation levels for 5 s RNA obtained with DNA purified from the four chromosomeclassesis about 2.1 x 10e5, which is clos eto the saturation level found with total HeLa cell DNA (25 x 10e5, Hatlen & Attardi,
] I 0 005
I 0 015.
InputRNAIDKA FIG. 4. Saturation curve8 of DNA from fractionated chromosomes by HeLa cell 3aP-labeled 6 s RNA. The incubation mixtures contained 20 pg of chromosomal DNA, variow amounts of HeLa cell 32P-labeled 6 s RNA and 1 pg HeLa cell 3H-labeled 28 s RNA in I.0 ml. 2 x SSC. In the insert, the dashed line represents the distribution of chromosomes after centrifugation through a glycerol-sucrose density gradient. The cut-off points for pooling the fractions into separate ohromosome classes are indicated by arrows. - x x -, Saturation level by 3aP-labeled 6 8 RNA of the DNA extracted from the various chromosome classes; -e---e-, saturation level by 3H-labeled 28 s RNA of the DNA extracted from the various chromosome classes.
In an attempt to improve the slight enrichment in sites complementary to 5 s RNA which was reproducibly found in the DNA extracted from the largest chromosomes (class l), a second cycle of chromosome fractionation through a glycerol-sucrose density gradient was performed, as described in the legend of Figure 1. As appears from Figure 5, the DNA extracted from the largest chromosomes (class 1) was saturated by 5 s RNA to a level about 25% higher than the DNA extracted from the chromosomesof class 3, therefore without any substantial change from the results obtained after the first fractionation. Again, the weighted average saturation level of the DNA purified from four chromosomeclassesby 6 s RNA was about 2.1 x 10m5.
5 s RNA
FIQ. 5. Saturation curves of DNA from recentrifuged chromosome classes by HeLa T’-labeled 5 s RNA. ‘Ike incubation mixtures contained 20pg of chromosomal DNA (extracted from the 4 chromosome classes obtained 8fbr 2 cycles of centrifugation in glycerol-sucrose density gradients as described in Fig. l), various amounts of He18 31P-labeled 5 s RNA, 1 pg. 3H-labeled 28 s RNA and O-4 pg. 3H-labeled 18 s RNA in 1.0 ml. 2 x SSC. In the insert, the saturation levels by 3sP-labeled 6 s RNA (-x-x-) and by combined 3H-labeled 28 s RNA and 3H-labeled 18 s RNA (-e--a-) are shown.
4. Discussion (a) Chromosmml
of sites for tRNA
The results presented above indicate that the DNA sites for tRNA in HeLa cells are distributed in chromosomes of all size ranges. The present observations are therefore in agreement with previously published evidence with regard to the lack of exclusive location of the 4 s sites in the chromosomes carrying a nucleolar organizer. Thus, on the basis of the lack of effect on 4 s RNA synthesis of low doses of actinomycin D which inhibit nucleolar RNA synthesis, Perry (1962) had concluded that the synthesis of 4 s RNA takes place in the extra-nucleolar region of the nucleus. Likewise, 4 s RNA had been found to be synthesized in homozygous anucleolate mutants of Xenopus luevis which cannot synthesize high molecular weight rRNA (Brown & Gurdon, 1964). Furthermore, no difference in the amount of DNA complementary to tRNA had been detected among Drosophila mdmogmter stocks possessing different doses of the nucleolar organizer region, indicating that none or very little of this DNA is in the region of the genome containing the cluster of genes for the high molecular weight rRNA (Ritossa, Atwood & Spiegelman, 1966). In the same organism, a possible location of the tRNA cistrons at the genetic loci of the dominant markers known as “minutes”, which are scattered among the X chromosome and the autosomes, has been postulated (Ritossa, et al., 1966). (b) Chrommomul distribution
of sites for 5 s RNA
Somewhat surprising has been the finding that t’he sites for 5 s rRNA, in contrast to those for the high molecular weight rRNA, are not concentrated in the chromo-
somes carrying a nucleolar organizer, but are scattered among chromosomes of different size ranges. Indeed, evidence already existed which suggested that in eukaryotic cells, in contrast to bacteria, where there are observations pointing to a close linkage of the 5 s cistrons to the 16 s and 23 s cistrons (Colli & Oishi, 1969; Pato & Meyenburg, 1970), the 5 s genes are not located close to those for the high molecular weight rRNA. Thus, homozygous anucleolate mutants of X. 2aevis have been shown to contain a normal complement of DNA homologous to 5 s RNA (Brown & Weber, 1968). Furthermore, the DNA complementary to 5 s RNA in Xenopus has beenfound to have a different density from the DNA which contains the genesfor 18 s and 28 s RNA (Brown & Weber, 1968). In Drosophila, an analysis of the complement of 5 s sites in stocks with different dosesof the nucleolar organizer region has likewise led to the conclusion that the genesspecifying 5 s rRNA are not linked to the 28 s and 18 s genes (Perry, Greenberg & Tartof, 1970). The present observations confirm and extend the above mentioned findings, indicating a chromosomallocation different from that of the 28 s and 18 s RNA genesfor the majority, if not all, the 5 s genes in HeLa cells.From the previously published structural evidence indicating the presence of 5’-, di- and triphosphate groups in the majority of the 5 s molecules(Hatlen, Amaldi & Attardi, 1969), one would have to conclude that the multiple 5 s sites in each chromosomeare not clustered or, if clustered, are transcribed as individual units. The scattered distribution of the 5 s sites in the genome and the apparently large excess of the 5 s genesover the genes for 28 s and 18 s RNA (seepreceding paper) suggestthat the synthesis of 5 s RNA and that of the high molecular weight rRNA, though in general co-ordinated, may be subject to different control mechanisms,and furthermore hint at the possibility that 5 s RNA may have some as yet unknown function besidesthat of being a structural component of ribosomes. (c) Hybridization
In a previous study on the chromosomal distribution of the DNA sites for 28 s and 18 s RNA, the level of hybridization of chromosomal DNA with these RNA species was found to be appreciably lower than that obtained with total HeLa cell DNA. This apparent lossof high molecular weight rRNA siteshasnot beenobserved in the present work. A likely explanation for this difference is the use in the present work of pronase digestion in the DNA extraction procedure, which presumably resulted in a more complete recovery of nucleolar DNA from the isolated chromosomes.Also with 4 s and 5 s RNA the hybridization capacity of chromosome1DNA found here was very close to that observed with total HeLa cell DNA. This investigation was supported by a grant (GM-11726) and a fellowship (GM-00086) from the U.S. Public Health Service and was undertaken during the tenure of a Research Training Fellowship awarded by the International Agency for Research on Cancer to one of the authors (Y. A., on leave from the Weizmann Institute of Science, Rehovot, Israel). The excellent technical assistance of Mrs La Verne Wenzel and Mrs Bennota Keeley ie gratefully acknowledged.
REFERENCES Ameldi, Brown, Brown,
F. & Attardi, G. (1968). J. Mol. Biol. 32, 737. D. D. & Gurdon, J. B. (1964). Proc. Nat. Acad. Sci., D. D. & Weber, C. S. (1968). J. Mol. Biol. 34, 661.
Colli, W. & Oishi, M. (1969). Proc. Na;t. AC&. Sci., Wash. 64, 642. Hatlen, L., Amaldi, F. & Attardi, G. (1969). Btichem&vy, 8, 4989. Hatlen, L. & Attardi, G. (1971). J. Mol. Biol. 56, 536. Huberman, J. A. & Attardi, G. (1967). J. Mol. Biol. 29, 487. Jeanteur, Ph. & Attardi, G. (1969). J. Mol. BioZ. 45, 306. Maio, J. J. & Schildkraut, C. L. (1966). In Methods in Cell Physiology, ed. by Prescott, vol. 2, p. 113. New York: Academic Press. Maio, J. J. & Schildkraut, C. L. (1969). J. Mol. BioZ. 40, 203. Marmur, J. (1961). J. Mol. BioZ. 3, 208. Mendelsohn, J., Moore, D. E. & Salzman, N. P. (1968). J. Mol. BioZ. 32, 101. Pato, M. L. & Meyenburg, K. v. (1970). Cold Spr. Ho& Symp. &ant. BioZ. vol. the press. Perry, R. P. (1962). Proc. Nat. Acad. AS%., Wash. 48, 2179. Perry, R. P., Greenberg, J. R. & Tartof, K. D. (1970). Cold Spr. Hurb. Symp. BioZ. vol. 35, in the press. Ritossn, F. M., Atwood, K. C. & Spiegelman, S. (1966). Genetics, 54, 663.