J. Mol.
Biol.
Mapping
(1978)
121, 299-310
of Mitochondrial4 S RNA Genes in Xenopus laevis by Electron Microscopy SEIGO Omt,
JOSE LUIS RAMIREZ$, WILLIAI\I AND IGOR B. DAWID~
B. UPHOLT$
Carnegie Institution of Washington Department of Embryology 115 West University Parkway Baltimore, Md 21210, V.8.A. (Received
25 October 1977)
The distribution of sites hybridizing with mitochondrial 4 S RNA molecules on mitochondrial DNA of Xenopus Zaevis has been mapped in relation to the ribosomal RNA genes and EcoRI restriction endonuclease sites. RNA molecules linked to ferritin were employed for t’his purpose. We have obt,ained evidence for 15 4 S RNA sites on t,he H-strand and six sites on the L-strand of X. Zaevis mtDNAl/. An indication of the possible existence of one additional site on the H-strand and four additional sites on the L-strand has been obtained. One 4 S RNA site is located in the gap between the two rRNA genes, and one site flanks each outside end of the rRNA genes. The other 4 S RNA sites are distributed almost evenly throughout both strands of the mtDNA. A comparison with the map of 4 S RNA sites on the mtDNA of HeLa cells (Angerer et al., 1976) suggests considerable evolutionary conservation of site organization.
1. Introduction Mitochondrial DNA carries the ribosomal RNA and transfer RNA genes for the protein synthesizing machinery in mitochondria. The total number of mt-tRNAs// in animal cells has not yet been determined wit’h certainty. One might expect a number that could provide a full set of tRNAs sufficient to read the genetic code. Early work suggested about 12 4 S RNA genes on HeLa cell mtDNA (Wu et al., 1972) or 15 genes on Xenopus Zaevis mtDNA (Dawid. 1972a). Mitochondrial 4 S RNAs include organelle-specifiq tRNAs (see Dawid, 19723). On more detailed investigation the number of known 4 S RNA sites in HeLa mtDNA has risen to 19, and the location of these genes has been mapped (Angerer et al., 1976). We have studied the distribution of 4 S RNA genes on X. luevis mtDNA with t,he aim of providing a map of these genes in relation to the positions of other functional regions in the DNA (Dawid et al., t Present address: Department of Anatomy and Cell Biology, School of Medicine, University of Pittsburgh, Pittsburgh, Penn. 16261, U.S.A. 1 Present address: Escuela de Biologia, de la U.C.V., Apdo 10098, Caracas, Venezuela. f Present address : University of Chicago, Departments of Pediatrics and Biochemistry and the Joseph P. Kennedy Mental Retardation Research Center, Chicago, Ill. 60637, U.S.S. T To whom correspondence should be addressed. ij Abbreviations used: mtDNA, mitochondrial DNA: mt-tRNA, mieochondrial tRNA; mtrRNA, mitochondrial rRNA. 299 00224
-2836/78/1213-9910
$02.00/O.
,~c;; 1078 Academic
Press (London)
Inc.
Ltd.
3llll
S. OHI,
.J. I,.
K:ZMIKEZ.
\V.
13. 1:I’HOLT
ANI)
1976; Ramirez & Dawid. 1978). Furt,her, we have RKA genes in Xenopus mtDl\u’A in t’his study. Wr feel tlhis number and report here evidence for at least 21 4 S RIVA sites in HeLa cells and Xerbopus suggests tionary conservation of gene arrangement between results have been presented briefly by Dawid et ~1.
2. Materials
I.
H.
l);\\V11)
re-examined t’ht, number of 4 S t,hat earliw work uudcrestimated sites. Comparison of the maps of a considerable extent of evolufrogs and man. Some of these (1976).
and Methods
(a) DXA and RNA mtDNA and mtRNA from X. Zaevis ovaries were prepared as described by Dawid (1972a). The cloning of the EcoRI fragment is described by Ramirez & Dawid (1978). All work with recombinant DNA followed t,he National Institutes of Health guidelines. Experiments with mtDNA clones were carried out under PZ-EKl conditions. Mitochondrial 4 S RNA was purified by sucrose gradient centrifugation and gel filtration over Sephadex GlOO, and mt-rRNA was separated by sucrose gradient centrifugation (Dawid, 1972a). A marker DNA fragment was derived from a digest, of cloned EcoRI A fragment with the restriction endonuclease HpaII. One fragment of 1300 base-pairs occurs in the digest which hybridizes to mt-rRNA and coincides almost precisely with the sequence coding for the large rRNA. The explicit evidencr for this map location is contained in the histogram of 4 S RNA sites on the L-strand (R,esldts; Fig. 6). This HpaII fragment was purified by agarose gel electrophoresis. (b) Strand separation Since mtDNA is degraded in alkali, we used the following procedure (Szybalski et al., 1971). Singly nicked mtDNA (Greenfield et al., 1975) was made up to 13 pg/ml in CsCl of density 1.73 g/cm3 with a minimal amount of buffer present. The solution was stirred at 0°C and a KOH solution was added to bring the pH to 12.2, as measured with a Radiometer pH meter 22 with a GK2302B electrode. After 10 to 15 s, poly(I,G) (1% mg/ml, previously incubated in 0.1 M-KOH at 37°C for 4 rnin and neutralized) was added to bring the poly(I,G)/DNA ratio to 1.3. The solution was diluted immediately with a solution of CsCl (density 1.73 g/cm3) containing 30 mm-Tris*HCl (pH 8.5) and sufficient HCl to neutralize t’he KOH, to give a final DNA concn of 1.8 pg/ml. The final density was adjusted to between 1.73 and 1.74 g/cm3. The solution was centrifuged to equilibrium, using 7 ml of solution per thick-walled polyallomer tube, at 33 x lo3 revs/min and 20°C in a Beckman 65 or 40 rotor for 48 11. Figure 1 shows the separation that was obtained. The H-strand does not bind poly (1,G) and was obtained essentially pure and largely intact. The L-strand, however, is complexed
IO
Fro. 1. Separation of the gradient
is described
strands
in the text.
20 Fraction
30 no.
of X. Zaevis mtDNA
with
poly(I,G).
The
preparation
of the
MITOCHONDRIBL
with poly(I,G) without some
so firmly
4 S RNA
that
it proved
degradation of the DNA. EcoR,I A fragment, which was separated centrifugation (Ramirez & Dawid, 1978).
difficult Therefore
IN
XEYOPUS
to remove we
used
of 4 S RLVA
301
the synthetic
L-strand
into H and L-strands
coupling
(c) Covalent
MAP
from
by alkaline
polymer the
cloned
CsCl gradient
to few&in
WC used the BI procedure of Wu & Davidson (1973), which links oxidized tRNA to amino groups in ferritin by a Schiff base reaction, followed by stabilization of the complex by reduction with NaBH,. The procedure was modified slightly by a threefold increase in the concentration of reactant RNA and ferritin. In most experiments 10% of the input RNA could be linked to ferritin. The complex was stahlr: at) - 70°C for at least, 1 year. (d) Hybridizatio,~ T11e hybridization mixture had the following composition. H-strand of mtDNA, 3 pp/ml; ferritin--4 S RNA, 40 pg of RNA/ml; small mt-rRNA, 2 pg/ml; large mt-rRNA, 4 pg/ml ; 4076 formamide (Fisher certified) ; 0.15 BI-Na phosphate buffer (pH 7.0) ; NaCl to adjust cation concn to 0.5 M. In the case of the L-strand the rRNA was omitted and 3 pg of denatured HpaII marker/ml were included. Incubation was for 2 II at 37°C. In every case the ferritin-4 S RNA complex was purified immediately before hybridization by prctcipitation from 60% ammonium sulfate. This step removed free 4 R RNA t,hat could h:tvc~ been present as a result of degradation of the complex. (e) Removal
of excess few&n-l
S RLNA
Two procedures were used for this purpose. The first was the electrophoretic method of Wu & Davidson (1973). We found that extraction with phenol, also used by these authors, resulted in total loss of the complex in our hands. A second and more convenient method involved chromatography on hydroxylapatite. For this procedure the hybridization mixture was diluted 20-fold with 0.15 M-sodium phosphate (pH 7.0) and a small amount of labeled HeLa cell DNA was added to facilitate detection of DNA-containing fractions (any labeled DNA would be suitable). The mixture was applied to a 0.5 cm x 0.6 cm hydroxylapatite column (Bio Rad) equilibrated with 0.15 M-phosphate buffer. The column was washed with the same buffer until no ferritin eluted, as judged by visible color. The DNA complex was then eluted with 0.5 M-phosphate, 0.5 tir-NaC1, and small samples from the peak fractions were direct,ly spread for electron microscopy. (f) Electron
microscopy
The spreading
solutions contained 50% formamidc and the other constituents described by Wellauer 8z Dawid (1977) and Wellauer et al. (1976): the hypophase was water. The methods for electron microscopy and evaluation of micrographs are described in the satnc references. In all spreadings DNA from phape 4X 171 was inclltdrd as a standard.
3. Results (a) 4 X RNA
sites
on th H-strand
The mapping of 4 S RNA molecules by electron microscopy requires the attachment of a bulky marker since the 4 S RNA-DNA duplex itself is too short to be visualized directly. We used the method of Wu & Davidson (1973) in which 4 S RNA is linked covalently to ferritin. In these experiments we used a preparation of separated H-strand purified from mtDNA with the aid of CsCl gradient centrifugation in the presence of poly(I,G) (see Materials and Methods). These DNA samples consisted mainly of circular and fulllength linear mtDNA strands but also contained some broken molecules. The DNA sequences which hybridize with the two rRNA molecules were used as markers in
302
S. OHI,
J.
I,.
RAMIREZ,
\V.
H.
l’l’HOl,‘P
.-\NI)
I.
H.
l)ALVIl)
(a)
FIG. 2. H-strands of mtDNA with attached ferritin-4 S RNA molecules. The 4 S RNA sites are numbered as in Fig. 8. The rRNA duplex regions used for reference are identified as Lg (large) of Sm (small).
these measurements. H-strand samples were annealed with the rRNAs and ferritin4 S RNA, as described in Materials and Methods, and spread for electron microscopy. Any single-stranded DNA molecule that contained duplex regions formed by both rRNAs and at least two ferritin-4 S RNA molecules were photographed and measured. Figure 2 shows two circular H-strand molecules. The molecule in (a) carries an unusually large number of ferritin molecules while the molecule in (b) is more typical of the population (see below). The ferritin label that occupies the position between the two mt-rRNA regions is immediately apparent. Measurements based on 152 H-strand molecules are summarized in the histograms
MITOCHONDRlAL
Circular
4 S RNA
and full-length
lmear
M.tP
length
molecules
60
40 % Of mean
I
303
60
% Of full
, 0
XE’KOP’,:S
molecules
40
Broken
IN
2
full
length
3 Lenqlh
4
5
6
t 6.6
(pm)
FIG. 3. Histogram of 4 S RNA &es on the H-strand. Molecules in (a) were full-length, while (b) presents data obtained with broken molecules. The stippled regions show the positions of the rRNA duplex regions in each mtDNA molecule. Numbers above peaks refer to the numbers of ferritin-4 S RNAs in the peak.
of Figure 3. Figure 3(a) shows measurements on intact molecules, including circles and linear strands measuring between 6-O and 7.5 pm. This range was used for linear molecules since it was the range of lengths of circles. The strands were assumed to be full-length and measurements are expressed as a percentage of the genome. In Figure 3(b) we show measurements with broken molecules. To allow comparative measurements as percentages of the genome we corrected the lengths of some molecules in the following way. Distances along the mtDNA were measured relative to nearby $X174 circles, and were then recalculated as percentages of the total mtDNA by using the ratio of the average length of +X174 DNA and a single-stranded mtDNA. Other broken strands were measured without close-by standards and the average lengths were used in the calculations without corrections. As expected, the broken strands yielded reliable values close to the rRNA markers and less useful data at greater distances from them.
the
FIG. 4. Part left of the
of H-strand with occupied doublet Figure. Symbols are as in Fig. 2.
site.
The
full-length
linear,
strand
continues
to
Assignments of 4 S RNA sites on the H-strand are indicated above the histogram of Figure 3. Each clearly defined peak is considered a 4 S RNA site. In some cases two close sites could not be well separated in the histogram. In these cases we base the assignment on the presence of molecules in which two ferritin-4 S RNA molecules are seen attached to the H-strand close to each other at the appropriate map positions. Table 1 summarizes these cases. Sites H3 and H4 are so close that only a single peak was obtained in the histogram. Nevertheless, the assignment of t’wo sites at this position is warranted since doublets were often observed; one example is shown in Figure 4. Sites H8 and H9 are separated in the histogram and doublets were seen at a high frequency. Sites Hll and HI2 are separated in the histogram but only two doublets were seen, most likely because the frequency of occupation of these two sites is low. In the case of HlO an ambigiity remains: while two doublets were seen, we obtained a single peak in the histogram with moderately high occupancy and
1
TABLE
Frequency Site
Location Frequency Distance
(map
H3
units)
sites
and
0.15,
of observation between
of doublets of fewdin-
9/158 (bases)
H4
0.16 =
0.057
2001-528
The numbering of sites refers to Fig. 8. Map t Listed as a single site since no separation doublets were observed. $ Listed as 2 sites because of the separation 5 Average and standard deviation are given.
X RNA
H8
and
0.48, lo/158
H9
HlOt
0.51 =
on the H-strand
0.59,
0.063
608 -t. 293
2/158
§
Hll
0.61 =
of peaks
in
the
0.65,
0.013
histogram
HlZ$
0.69
2/158
270
positions refer to Fig. 3. of peaks was seen in the
and
== 0.013 580
hist,ogram (Fig.
3).
and
only
2
MITOCHONDRIAL
4 S RNrl
M;\P
IN
Xh’NOt’~‘,S
305
yet the doublet frequency is low. Therefore, a single site was assigned to this position in the final map (Fig. 8: see below). The distribution of 4 S RNA sites on the H-strand is quite even along the strand and is almost symmetrical with respect to an axis through map positions 0 and FjO?;, (Figs 3 and 8). We were concerned about the chance that actually all 4 S RNA sibes are contained in one half of the molecule but misassignment of bhe lengths of the two rRNA duplex regions would create an artifact& mirror image of the sites. This situation does not obtain since many strands were found which contained ferritin-4 S RNAs between both 0 and 50 and 50 and lOOq/, on the map (see Fig. 2(a)). The frequency of attachment of ferritin-4 S RNA molecules per intact strand was measured. We observed an average of 3.4 ferritin labels ppr strand which corresponds t)o an occupation frequency of 0.23 per site. This value is lvithin the range observed 1)) Wu P/ nl. (1972) and Angerer d al. (1976). WP measured the background of fortuitous coincidence of ferritin-4 8 RNA or released ferritin with DNA strands by counting ferritin molecules associated with 4X174 DNA molecules. After removal of excess ferritin by the electrophoretic method (Wu it nl.. 1972) the frequency a-as 0.041 ferritins per three +X174 strands% which equal ia length one mtDNA strand. This background frequency is too low to affect the rcbsults. (b) 4 S RNA sites ONthP L-strand Because of bhe difficulties of obtaining intact L-strand from mtDNA isolated from t,he frog we used the cloned large EcoRI A fragment, which represents 877; of the genome (see Materials and Methods, and Ramirez & Dawid, 1978). The L-stra.nd of the cloned fragment was purified by centrifugation in alkaline CsCl gradients. A rest,rict,ion fragment derived with HpalI was used as a marker. This fragment, which hybridizes with labeled mt-rRNA, forms a duplex w&h the L-strand of the EcoRI A fragment close to one end (Figs 5 and 7). The position of this marker duplex (see histogram in Fig. 6) corresponds quite closely to the gene region for the large mtrRNA (Ramirez & Dawid, 1978). Thus, the HpdI fragment is a useful marker. Figure 5 shows an example of an L-strand with three ferritin-4 S RNAs attached t-o it. From the measurement of L-&and molecules we derived the histogram shown in Figure 6. As in the case of the H-strand there is a region with very close sites. This region around map position 30 contains at least two sites since a number of doublets were seen (Fig. 7). We believe t,hat this region does in fact contain three sites, because of the high frequency of its occupation and because of the morphology of doublets. Such doublets occurred in two classes: in one class the two ferritin labels were separated by a distance of about 200 bases (Fig. 7(a)), while t’he other class (7(b)) the ferritfin molecules almost touched each other. While two types of doublets suggest t,hrer sites no triplet was seen and alternative interpretations are possible. We indicate the tent,atively assigned third site by a small arrow in the histogram of Figure 6. The other small arrows in that histogram denote tentative assignments. The number of ferrit’in labels at these sites was low, both absolute and relative to the frequency at other sites. It is difficult to determine whether these are in fact 4 S RNA sites corresponding to low-abundance RNAs or whether there is a background, perhaps from fragments of larger RNA molecules. We use arrows at these positions to point out the possibility that 4 S RNA sites can occur at these locations.
Bl)!i
S. OHI,
.I.
I>. I{;\hllliIC%,
FIG. 5. L-strand with 3 ferritin+4 was used as a referrnce is indicated of the tentatively identified sites.
\\‘.
13. l~I’llol,‘l’
.\SI)
I.
I
Biol.
Mapping
(1978)
121, 299-310
of Mitochondrial4 S RNA Genes in Xenopus laevis by Electron Microscopy SEIGO Omt,
JOSE LUIS RAMIREZ$, WILLIAI\I AND IGOR B. DAWID~
B. UPHOLT$
Carnegie Institution of Washington Department of Embryology 115 West University Parkway Baltimore, Md 21210, V.8.A. (Received
25 October 1977)
The distribution of sites hybridizing with mitochondrial 4 S RNA molecules on mitochondrial DNA of Xenopus Zaevis has been mapped in relation to the ribosomal RNA genes and EcoRI restriction endonuclease sites. RNA molecules linked to ferritin were employed for t’his purpose. We have obt,ained evidence for 15 4 S RNA sites on t,he H-strand and six sites on the L-strand of X. Zaevis mtDNAl/. An indication of the possible existence of one additional site on the H-strand and four additional sites on the L-strand has been obtained. One 4 S RNA site is located in the gap between the two rRNA genes, and one site flanks each outside end of the rRNA genes. The other 4 S RNA sites are distributed almost evenly throughout both strands of the mtDNA. A comparison with the map of 4 S RNA sites on the mtDNA of HeLa cells (Angerer et al., 1976) suggests considerable evolutionary conservation of site organization.
1. Introduction Mitochondrial DNA carries the ribosomal RNA and transfer RNA genes for the protein synthesizing machinery in mitochondria. The total number of mt-tRNAs// in animal cells has not yet been determined wit’h certainty. One might expect a number that could provide a full set of tRNAs sufficient to read the genetic code. Early work suggested about 12 4 S RNA genes on HeLa cell mtDNA (Wu et al., 1972) or 15 genes on Xenopus Zaevis mtDNA (Dawid. 1972a). Mitochondrial 4 S RNAs include organelle-specifiq tRNAs (see Dawid, 19723). On more detailed investigation the number of known 4 S RNA sites in HeLa mtDNA has risen to 19, and the location of these genes has been mapped (Angerer et al., 1976). We have studied the distribution of 4 S RNA genes on X. luevis mtDNA with t,he aim of providing a map of these genes in relation to the positions of other functional regions in the DNA (Dawid et al., t Present address: Department of Anatomy and Cell Biology, School of Medicine, University of Pittsburgh, Pittsburgh, Penn. 16261, U.S.A. 1 Present address: Escuela de Biologia, de la U.C.V., Apdo 10098, Caracas, Venezuela. f Present address : University of Chicago, Departments of Pediatrics and Biochemistry and the Joseph P. Kennedy Mental Retardation Research Center, Chicago, Ill. 60637, U.S.S. T To whom correspondence should be addressed. ij Abbreviations used: mtDNA, mitochondrial DNA: mt-tRNA, mieochondrial tRNA; mtrRNA, mitochondrial rRNA. 299 00224
-2836/78/1213-9910
$02.00/O.
,~c;; 1078 Academic
Press (London)
Inc.
Ltd.
3llll
S. OHI,
.J. I,.
K:ZMIKEZ.
\V.
13. 1:I’HOLT
ANI)
1976; Ramirez & Dawid. 1978). Furt,her, we have RKA genes in Xenopus mtDl\u’A in t’his study. Wr feel tlhis number and report here evidence for at least 21 4 S RIVA sites in HeLa cells and Xerbopus suggests tionary conservation of gene arrangement between results have been presented briefly by Dawid et ~1.
2. Materials
I.
H.
l);\\V11)
re-examined t’ht, number of 4 S t,hat earliw work uudcrestimated sites. Comparison of the maps of a considerable extent of evolufrogs and man. Some of these (1976).
and Methods
(a) DXA and RNA mtDNA and mtRNA from X. Zaevis ovaries were prepared as described by Dawid (1972a). The cloning of the EcoRI fragment is described by Ramirez & Dawid (1978). All work with recombinant DNA followed t,he National Institutes of Health guidelines. Experiments with mtDNA clones were carried out under PZ-EKl conditions. Mitochondrial 4 S RNA was purified by sucrose gradient centrifugation and gel filtration over Sephadex GlOO, and mt-rRNA was separated by sucrose gradient centrifugation (Dawid, 1972a). A marker DNA fragment was derived from a digest, of cloned EcoRI A fragment with the restriction endonuclease HpaII. One fragment of 1300 base-pairs occurs in the digest which hybridizes to mt-rRNA and coincides almost precisely with the sequence coding for the large rRNA. The explicit evidencr for this map location is contained in the histogram of 4 S RNA sites on the L-strand (R,esldts; Fig. 6). This HpaII fragment was purified by agarose gel electrophoresis. (b) Strand separation Since mtDNA is degraded in alkali, we used the following procedure (Szybalski et al., 1971). Singly nicked mtDNA (Greenfield et al., 1975) was made up to 13 pg/ml in CsCl of density 1.73 g/cm3 with a minimal amount of buffer present. The solution was stirred at 0°C and a KOH solution was added to bring the pH to 12.2, as measured with a Radiometer pH meter 22 with a GK2302B electrode. After 10 to 15 s, poly(I,G) (1% mg/ml, previously incubated in 0.1 M-KOH at 37°C for 4 rnin and neutralized) was added to bring the poly(I,G)/DNA ratio to 1.3. The solution was diluted immediately with a solution of CsCl (density 1.73 g/cm3) containing 30 mm-Tris*HCl (pH 8.5) and sufficient HCl to neutralize t’he KOH, to give a final DNA concn of 1.8 pg/ml. The final density was adjusted to between 1.73 and 1.74 g/cm3. The solution was centrifuged to equilibrium, using 7 ml of solution per thick-walled polyallomer tube, at 33 x lo3 revs/min and 20°C in a Beckman 65 or 40 rotor for 48 11. Figure 1 shows the separation that was obtained. The H-strand does not bind poly (1,G) and was obtained essentially pure and largely intact. The L-strand, however, is complexed
IO
Fro. 1. Separation of the gradient
is described
strands
in the text.
20 Fraction
30 no.
of X. Zaevis mtDNA
with
poly(I,G).
The
preparation
of the
MITOCHONDRIBL
with poly(I,G) without some
so firmly
4 S RNA
that
it proved
degradation of the DNA. EcoR,I A fragment, which was separated centrifugation (Ramirez & Dawid, 1978).
difficult Therefore
IN
XEYOPUS
to remove we
used
of 4 S RLVA
301
the synthetic
L-strand
into H and L-strands
coupling
(c) Covalent
MAP
from
by alkaline
polymer the
cloned
CsCl gradient
to few&in
WC used the BI procedure of Wu & Davidson (1973), which links oxidized tRNA to amino groups in ferritin by a Schiff base reaction, followed by stabilization of the complex by reduction with NaBH,. The procedure was modified slightly by a threefold increase in the concentration of reactant RNA and ferritin. In most experiments 10% of the input RNA could be linked to ferritin. The complex was stahlr: at) - 70°C for at least, 1 year. (d) Hybridizatio,~ T11e hybridization mixture had the following composition. H-strand of mtDNA, 3 pp/ml; ferritin--4 S RNA, 40 pg of RNA/ml; small mt-rRNA, 2 pg/ml; large mt-rRNA, 4 pg/ml ; 4076 formamide (Fisher certified) ; 0.15 BI-Na phosphate buffer (pH 7.0) ; NaCl to adjust cation concn to 0.5 M. In the case of the L-strand the rRNA was omitted and 3 pg of denatured HpaII marker/ml were included. Incubation was for 2 II at 37°C. In every case the ferritin-4 S RNA complex was purified immediately before hybridization by prctcipitation from 60% ammonium sulfate. This step removed free 4 R RNA t,hat could h:tvc~ been present as a result of degradation of the complex. (e) Removal
of excess few&n-l
S RLNA
Two procedures were used for this purpose. The first was the electrophoretic method of Wu & Davidson (1973). We found that extraction with phenol, also used by these authors, resulted in total loss of the complex in our hands. A second and more convenient method involved chromatography on hydroxylapatite. For this procedure the hybridization mixture was diluted 20-fold with 0.15 M-sodium phosphate (pH 7.0) and a small amount of labeled HeLa cell DNA was added to facilitate detection of DNA-containing fractions (any labeled DNA would be suitable). The mixture was applied to a 0.5 cm x 0.6 cm hydroxylapatite column (Bio Rad) equilibrated with 0.15 M-phosphate buffer. The column was washed with the same buffer until no ferritin eluted, as judged by visible color. The DNA complex was then eluted with 0.5 M-phosphate, 0.5 tir-NaC1, and small samples from the peak fractions were direct,ly spread for electron microscopy. (f) Electron
microscopy
The spreading
solutions contained 50% formamidc and the other constituents described by Wellauer 8z Dawid (1977) and Wellauer et al. (1976): the hypophase was water. The methods for electron microscopy and evaluation of micrographs are described in the satnc references. In all spreadings DNA from phape 4X 171 was inclltdrd as a standard.
3. Results (a) 4 X RNA
sites
on th H-strand
The mapping of 4 S RNA molecules by electron microscopy requires the attachment of a bulky marker since the 4 S RNA-DNA duplex itself is too short to be visualized directly. We used the method of Wu & Davidson (1973) in which 4 S RNA is linked covalently to ferritin. In these experiments we used a preparation of separated H-strand purified from mtDNA with the aid of CsCl gradient centrifugation in the presence of poly(I,G) (see Materials and Methods). These DNA samples consisted mainly of circular and fulllength linear mtDNA strands but also contained some broken molecules. The DNA sequences which hybridize with the two rRNA molecules were used as markers in
302
S. OHI,
J.
I,.
RAMIREZ,
\V.
H.
l’l’HOl,‘P
.-\NI)
I.
H.
l)ALVIl)
(a)
FIG. 2. H-strands of mtDNA with attached ferritin-4 S RNA molecules. The 4 S RNA sites are numbered as in Fig. 8. The rRNA duplex regions used for reference are identified as Lg (large) of Sm (small).
these measurements. H-strand samples were annealed with the rRNAs and ferritin4 S RNA, as described in Materials and Methods, and spread for electron microscopy. Any single-stranded DNA molecule that contained duplex regions formed by both rRNAs and at least two ferritin-4 S RNA molecules were photographed and measured. Figure 2 shows two circular H-strand molecules. The molecule in (a) carries an unusually large number of ferritin molecules while the molecule in (b) is more typical of the population (see below). The ferritin label that occupies the position between the two mt-rRNA regions is immediately apparent. Measurements based on 152 H-strand molecules are summarized in the histograms
MITOCHONDRlAL
Circular
4 S RNA
and full-length
lmear
M.tP
length
molecules
60
40 % Of mean
I
303
60
% Of full
, 0
XE’KOP’,:S
molecules
40
Broken
IN
2
full
length
3 Lenqlh
4
5
6
t 6.6
(pm)
FIG. 3. Histogram of 4 S RNA &es on the H-strand. Molecules in (a) were full-length, while (b) presents data obtained with broken molecules. The stippled regions show the positions of the rRNA duplex regions in each mtDNA molecule. Numbers above peaks refer to the numbers of ferritin-4 S RNAs in the peak.
of Figure 3. Figure 3(a) shows measurements on intact molecules, including circles and linear strands measuring between 6-O and 7.5 pm. This range was used for linear molecules since it was the range of lengths of circles. The strands were assumed to be full-length and measurements are expressed as a percentage of the genome. In Figure 3(b) we show measurements with broken molecules. To allow comparative measurements as percentages of the genome we corrected the lengths of some molecules in the following way. Distances along the mtDNA were measured relative to nearby $X174 circles, and were then recalculated as percentages of the total mtDNA by using the ratio of the average length of +X174 DNA and a single-stranded mtDNA. Other broken strands were measured without close-by standards and the average lengths were used in the calculations without corrections. As expected, the broken strands yielded reliable values close to the rRNA markers and less useful data at greater distances from them.
the
FIG. 4. Part left of the
of H-strand with occupied doublet Figure. Symbols are as in Fig. 2.
site.
The
full-length
linear,
strand
continues
to
Assignments of 4 S RNA sites on the H-strand are indicated above the histogram of Figure 3. Each clearly defined peak is considered a 4 S RNA site. In some cases two close sites could not be well separated in the histogram. In these cases we base the assignment on the presence of molecules in which two ferritin-4 S RNA molecules are seen attached to the H-strand close to each other at the appropriate map positions. Table 1 summarizes these cases. Sites H3 and H4 are so close that only a single peak was obtained in the histogram. Nevertheless, the assignment of t’wo sites at this position is warranted since doublets were often observed; one example is shown in Figure 4. Sites H8 and H9 are separated in the histogram and doublets were seen at a high frequency. Sites Hll and HI2 are separated in the histogram but only two doublets were seen, most likely because the frequency of occupation of these two sites is low. In the case of HlO an ambigiity remains: while two doublets were seen, we obtained a single peak in the histogram with moderately high occupancy and
1
TABLE
Frequency Site
Location Frequency Distance
(map
H3
units)
sites
and
0.15,
of observation between
of doublets of fewdin-
9/158 (bases)
H4
0.16 =
0.057
2001-528
The numbering of sites refers to Fig. 8. Map t Listed as a single site since no separation doublets were observed. $ Listed as 2 sites because of the separation 5 Average and standard deviation are given.
X RNA
H8
and
0.48, lo/158
H9
HlOt
0.51 =
on the H-strand
0.59,
0.063
608 -t. 293
2/158
§
Hll
0.61 =
of peaks
in
the
0.65,
0.013
histogram
HlZ$
0.69
2/158
270
positions refer to Fig. 3. of peaks was seen in the
and
== 0.013 580
hist,ogram (Fig.
3).
and
only
2
MITOCHONDRIAL
4 S RNrl
M;\P
IN
Xh’NOt’~‘,S
305
yet the doublet frequency is low. Therefore, a single site was assigned to this position in the final map (Fig. 8: see below). The distribution of 4 S RNA sites on the H-strand is quite even along the strand and is almost symmetrical with respect to an axis through map positions 0 and FjO?;, (Figs 3 and 8). We were concerned about the chance that actually all 4 S RNA sibes are contained in one half of the molecule but misassignment of bhe lengths of the two rRNA duplex regions would create an artifact& mirror image of the sites. This situation does not obtain since many strands were found which contained ferritin-4 S RNAs between both 0 and 50 and 50 and lOOq/, on the map (see Fig. 2(a)). The frequency of attachment of ferritin-4 S RNA molecules per intact strand was measured. We observed an average of 3.4 ferritin labels ppr strand which corresponds t)o an occupation frequency of 0.23 per site. This value is lvithin the range observed 1)) Wu P/ nl. (1972) and Angerer d al. (1976). WP measured the background of fortuitous coincidence of ferritin-4 8 RNA or released ferritin with DNA strands by counting ferritin molecules associated with 4X174 DNA molecules. After removal of excess ferritin by the electrophoretic method (Wu it nl.. 1972) the frequency a-as 0.041 ferritins per three +X174 strands% which equal ia length one mtDNA strand. This background frequency is too low to affect the rcbsults. (b) 4 S RNA sites ONthP L-strand Because of bhe difficulties of obtaining intact L-strand from mtDNA isolated from t,he frog we used the cloned large EcoRI A fragment, which represents 877; of the genome (see Materials and Methods, and Ramirez & Dawid, 1978). The L-stra.nd of the cloned fragment was purified by centrifugation in alkaline CsCl gradients. A rest,rict,ion fragment derived with HpalI was used as a marker. This fragment, which hybridizes with labeled mt-rRNA, forms a duplex w&h the L-strand of the EcoRI A fragment close to one end (Figs 5 and 7). The position of this marker duplex (see histogram in Fig. 6) corresponds quite closely to the gene region for the large mtrRNA (Ramirez & Dawid, 1978). Thus, the HpdI fragment is a useful marker. Figure 5 shows an example of an L-strand with three ferritin-4 S RNAs attached t-o it. From the measurement of L-&and molecules we derived the histogram shown in Figure 6. As in the case of the H-strand there is a region with very close sites. This region around map position 30 contains at least two sites since a number of doublets were seen (Fig. 7). We believe t,hat this region does in fact contain three sites, because of the high frequency of its occupation and because of the morphology of doublets. Such doublets occurred in two classes: in one class the two ferritin labels were separated by a distance of about 200 bases (Fig. 7(a)), while t’he other class (7(b)) the ferritfin molecules almost touched each other. While two types of doublets suggest t,hrer sites no triplet was seen and alternative interpretations are possible. We indicate the tent,atively assigned third site by a small arrow in the histogram of Figure 6. The other small arrows in that histogram denote tentative assignments. The number of ferrit’in labels at these sites was low, both absolute and relative to the frequency at other sites. It is difficult to determine whether these are in fact 4 S RNA sites corresponding to low-abundance RNAs or whether there is a background, perhaps from fragments of larger RNA molecules. We use arrows at these positions to point out the possibility that 4 S RNA sites can occur at these locations.
Bl)!i
S. OHI,
.I.
I>. I{;\hllliIC%,
FIG. 5. L-strand with 3 ferritin+4 was used as a referrnce is indicated of the tentatively identified sites.
\\‘.
13. l~I’llol,‘l’
.\SI)
I.
I
