Molec. gen. Genet. 149, 2 3 - 3 1 (1976) C© by Springer-Verlag 1976
Cloning of Calf Thymus Satellite I DNA in Escherichia coli Fritz Gautier, Hubert Mayer, and Werner Goebel Gesellschaft for Biotechnologische Forschung mbH Abteilung Genetik, Mascheroder Weg 1, D-3300 Braunschweig-St6ckheim, Federal Republic of Germany
Summary. The 1400 base pair repeat produced by digestion of calf satellite I DNA ((p= 1.714 g/cm 3) with EcoRI,was cloned in E. coli. The hybrid plasmid (pGM 214) which contains the Co/E1-Ap vector (pSF 2124) and the 1400 base pair fragment replicates stably in E. coIi and can be amplified by chloramphenicol treatment. No clone was found in which more than one "repeat unit" of the satellite I DNA was present in the chimaera plasmid. Digestion of the original satellite I and the plasmid pGM 214 with R. Sinai shows that the satellite DNA replicated in E. coli is cleaved by the restriction endonuclease Sinai whereas the original satellite I DNA from calf thymus is not, suggesting that the satellite I contains a large amount of modified cytosine or guanosine, probably 5-methyl-cytosine. R.EcoRI* produces a number of fragments with the satellite I in the range of 300 base pairs to 1400 base pairs. A physical map of pGM 214 (and pSF 2124) with R. EcoRI, R. HincII, R.HindIII, R. SmaI, R. BamI and R. EclI was constructed. The 1400 base pair "repeat unit" in the pGM 214 is efficiently transcribed in vitro by purified RNA polymerase, starting from a pSF 2124 promoter. The restriction enzyme EcII produces a 350 base pair repeat with calf satellite II (~0=1,722 g/cm3), whereas the satellite I is not cut by this enzyme.
Introduction
Calf thymus DNA contains four satellites representing about 14% of the total DNA (Filipski, Thiery and Bernardi, 1973). The function and evolution of the satellites are still unclear. Several attempts have been made to analyze the satellite structure with res-
triction enzymes (e.g. Botchan, 1974; Philippsen et al., 1975). H6rz, Hess and Zachau (1974) investigated the cleavage of mouse satellite DNA with Hind restriction endonucleases and found that the satellite DNA was remarkably resistent to these restriction enzymes. They discussed this fact on the basis 'of mutation in the recognition sites or by methylation of the cytosine residues. Filipski et al. (1973) showed a number of "anomalous" physical properties for the calf thymus satellites compared to bacterial DNAs of identical G + C content. For further investigation of the structure of calf thymus satellites it seemed quite helpful to us to clone the satellite I in E. coli. After cloning in E. coli with a suitable cloning vector it is easy to isolate large amounts of pure satellite DNA (even highly labelled) and no problems with methylation of the DNA or other "anomalous" DNA modifications will effect subsequent analysis with restriction nucleases. So, Gill and Falkow (1975) constructed a cloning vector (ColE1-Ap=pSF 2124) with one R.EcoRI site in the structural gene for colicin El. Because the Co/E1-Ap plasmid has two transformation markers, Ap and colicine immunity, and one integration marker, colicin production, we used this plasmid for cloning calf thymus satellite I in E. coli.
Experimental Procedures Bacterial Strains. E. coli 5K, a restriction and modification negativ strain, was used for all transformations. E. coli C 600 carrying the ColE1-Ap r plasmid (pSF 2124) was obtained from S. Falkow. E. coli K 12/W8 t68c857 was obtained from Saedler. JC 309 (CoIE1) and JC 411 R64drdl 1 (ColE2) was obtained from J. Collins. Media, buffer, centrifugation conditions, sources of reagents for bacterial cultures and D N A preparation were the same described by Goebel and Bonewald (1975). DNAs: Purified calf satellite I DNA was a gift from H. Brinemann. L a m b d a - D N A was prepared from t68c857 in E. coli K12/ W8 using standard procedures. Plasmid D N A was isolated accord-
24
F. Gautier et al. : Cloning of Calf Thymus Satellite l DNA in Escherichia colt
ing to Goebel (1970). Col plasmids were amplified in broth in the presence of chloramphenicol (150 lag/ml) at a cell density of 110 Klett units for 12 h at 37° C.
Enzymes. EcoRI and SmaI were isolated according to H. Mayer (manuscript in prep.); HincII was prepared according Landy et al. (1974) by G. Luibrand, BamI according Wilson and Young (1975) by H. Hartmann, HindIII by D. Blohm, EclI by H. Hartmann and kindly given to us. EclI is a new restriction enzyme isolated from Enterobacterium cloacae.
After incubation for 5 rain at 37o C without ATP and GTP, the reaction was started by addition of ATP and GTP. After 12 rain DNAse was added to a final concentration of 20 gg/ml. The DNAse treatment was stopped by heating the reaction mixture to 85° C for 15 min. After cooling SDS was added to a.final concentration of 0.05%. The total assay was used for hybridisation. In a second assay the rate of RNA synthesis was followed by removing 5 gl samples at different times. The 5 ~tl samples were precipitated on Whatman filters in 10% cold TCA, 0.1 M Na4P20~, washed 4x in 5% TCA, 1 x in ethanol/ether (60/40) and 1 x in ether. The dried filters were counted in toluene based scintillation fluid.
Assays for Enzymes. Digestion with EcoRI was carried out in 50 mM Tris-HC1, pH 7.5, 50 mM NaCI and 10 mM MgC12, with Sinai in 15 mM Tris-HC1, pH 9, 15 mM KC1 and 6 mM MgClz, with HindIII in 50 mM Tris-HC1, pH 7.5, 20 mM NaC1 and 20 mM MgCI2, with HincII in 5 mM Tris-HC1, pH 7.9, 100 mM NaC1 and 10 mM MgClz, with BamI in 10 mM Tris-HC1, pH 7.4, 150 mM NaC1 and 10 mM MgClz, with EclI in 40 mM Tris-HC1, pH 7.5, 100 mM NaC1 and 20 mM MgC12, with EcoRI* in 25 mM Tris-HC1, pH 8.5, 2 mM MgC12 and 50% ethylene glycol. Except for EcoRI* (6 h) the incubation time was 30 min at 37° C. Ligation with T4 DNA ligase (purchased from Miles Lab.) was done in 66 mM Tris-HCl, pH 7.6, 6.6 mM MgCI2, 1 mM ATP, 0.66 mM dithiothreitol, 1 mg/ml BSA for 15 h at 0° C; 1 gl ligase (10 units/ml) was added to the reaction mixture (total volume 100 gl).
Hybridization Procedure. For hybridization of in vitro synthesized RNA to the EcoRI fragments of pGM 214 the EcoRI-cut plasmid was electrophoresed on 1% agarose gel. The gel was stained, photographed and cut into strips. The DNA was transferred to nitrocellulose filter (0.45 ~t, Schleicher & Schfill) according to the procedure described by Southern (1975). Prior to hybridization the filter was incubated in 0.01% BSA, 0.01% FicolI and 0.01% polyvinylpyrollidon in 4 x SSC for 5 h at 65° C. Hybridization and washing of the filters were carried out according to standard procedures (Gillespie and Spiegelman, 1965). The dried filters were fluorographed according to Laskey (1975). After fluorography, the filter was cut at the sides where bands could be seen and counted in a liquid scintillation counter.
Analytical Ultracentr~tgation. Analytical ultracentrifugation was Electrophoresis in Agarose Gels. 1% or 2% slab gels in a vertical electrophoresis apparatus (construction of our laboratory) were run in 36 mM Tris-HC1, 30 mM NaHzPO 4 and 10 mM EDTA, pH 7.5. The gels were stained with ethidium bromide (0.5 ~tg/ml) and photographed under UV light.
carried out in a Beckman Model E ultracentrifuge with UV scanner. The samples for centrifugation were made up to 1.705 g/cm 3 CsC1 in TES with 1 ~tg to 3 ~tg plasmid DNA and Micrococeus luteus DNA (~0= 1.731 g/cm 3) as internal marker.
Filter Binding. EcoRI-cut pGM 214 and EcoRI- and Sinai-cut Construction of the Hybrid Plasmids. Plasmid pSF 2124 (Cole 1-Ap) was cut to the linear form with EcoRI. Calf satellite I DNA was completely digested with EcoRI to the 1400 base pair repeat for one assay, and only partially digested with EcoRI for two other assays. The DNAs were precipitated with 60% (v/v) ethanol at - 20 ° C, centrifuged and dissolved in ligase buffer. A tenfold excess of satellite I fragments was added to the EcoRI-digested pSF 2124 DNA and ligated with T4 ligase. E. colt 5K was transformed with the whole iigation mixtures according to standard procedures (Cohen and Chang, 1973). The transformed cultures were incubated for 90 min at 37°C in ENB medium and aliquots were plated on agar plates with 50 gg/ml ampicillin. For testing colicin production single colonies were grown in ENB plates over night at 37° C. The colonies were killed with chloroform and overlayed with E. colt C in soft agar (0.2 ml from a overnight culture of E. colt C to 50 ml soft agar). After incubation for 6 h at 37° C, colonies could be detected which did not make clear lysis zones in the E. colt C overlay. Those non colicin producing colonies were taken from parallel plates and tested for colicin E1 resistance and colicin E2 sensitivity by cross-streaking on agar plates with colicin E1 on one side of the plate and colicin E2 on the other side. From all colonies which were Ap, E1 immune and not colicin producing, 1 ml cultures were grown in small tubes. After amplification with chloramphenicol (150 gg/ml) over night, cleared lysates were prepared, which were directly submitted to agarose gel electrophoresis. From all colonies which contained a plasmid larger than the original pSF 2124, DNA preparations were made according to Goebel (1970). The experiments were carried out under P2 conditions.
h~ vitro Transcription Assay. In vitro transcription with purified RNA polymerase was made in 0.15 M KC1, 0.01 M MgC12, 0.5 rag/ ml BSA, 0.01 M dithiothreitol, 0.04 M Tris-HC1 (pH 7.9), 10 ~ M EDTA, 4x 10 -4 M ATP, UTP, CTP and 50 gCi 3H-GTP, 1 I-tg DNA and 10 ~tg RNA polymerase in a total volume of 0.1 ml.
pGM 214 DNA was incubated for 10 min in the transcription buffer without the nucleotide triphosphates with a 10-fold excess of RNA polymerase. After 10 min at 37° C heat denatured E. coli DNA was added (30 ~tg/ml) and incubation was continued for 10 min at 37° C. The solution was filtered through a presoaked nitrocellulose filter. The filters were washed with 0.15 M KC1 in 4 mM Tris-HC1, pH 7.9, cut into small pieces with a scalpel and incubated with proteinase K (200 gg/ml) for 15 rain and with 0.01% SDS for another 15 min. The solution was electrophoresed on agarose gels and the gels were stained and photographed.
Melting Curve. pGM 214 was cut with EcoRI and precipitated with 60% (v/v) ethanol at - 2 0 ° C, centrifuged and dissolved in 50 mM Na2HPOJNaH2PO~, pH 7.4. A sample with 1.20D at 265 nm was melted in a Gilford spectrophotometer with a heating rate of 0.5° C/rain.
Results
Cloning o f the C a l f Satellite I D N A with p S F 2124 ( C o l E l - A p ) Figure 1a shows a densitometer tracing of an analytical d e n s i t y g r a d i e n t u l t r a c e n t r i f u g a t i o n o f t h e p u r i f i e d c a l f s a t e l l i t e I D N A . I t c a n b e s e e n t h a t t h e r e is still a b o u t 1 0 % c a l f s a t e l l i t e II D N A p r e s e n t in t h i s p r e p a r a t i o n . N o o t h e r D N A is p r e s e n t as u n i q u e s e q u e n c e s , a n d satellites III a n d IV are absent. Figure 1 b s h o w s t h a t t h e s a t e l l i t e I D N A is c u t b y E c o R I t o a u n i q u e size ( m o l e c u l a r w e i g h t o f 0.925 x 10 6 ( B o t c h a n , 1974)),
F. Gautier et al. : Cloning of Calf Thymus Satellite I DNA in Escherichia coli
25
5= 1,731g/cm 3 jo= 1,722 g/cm3 jo= 1,71/, g/cm 3
a
Increasing density
>
Fig, 1. a UV-scan of Analytical Ultracentrifugation of Calf Satel-
lite I DNA. Calf satellite I DNA was centrifuged in CsC1 with 44,000 rpm at 25°C for 40 h with Micrococcus luteus DNA ( = 1.731 g/cm 3) as internal marker. Ultraviolet-absorbance tracings of the analytical density gradient at 265 nm were taken after 40 h. b Electrophoresis of EcoRI-cut Satellite DNA. Electrophoresis in a 1% agarose slab gel with 100 mA/150 V for 3 h in electrophoresis buffer (experimental procedures) of calf satellite I DNA, partially digested (a) and totally digested (b) with EcoRI
or to oligomers of this fragment. The satellite II D N A is not cut by EcoRI (Philippsen et al., 1975) and stays near the top of the gel. The cloning vector pSF 2124 is cut once by EcoRI (So et al., 1975) to give the linear molecule (Fig. 2). Plasmid pSF 2124 and the EcoRI-cut satellite I DNA were joined by T4 ligase, and E. coli 5K was transformed with the ligation mixture. Three independent transformations were made. In one a complete
digest of satellite I DNA with EcoRI was used, and in two others partially digested DNA. Amongst 2000 ampicillin-resistent (Ap) transformants 300 were further screened for colicin production. Among those, 40 non-colicin-producing colonies were checked for resistance to colicin E1 and sensitivity to colicin E2 (i.e. E1 immune). D N A preparations were made from colonies which were Ap, E1 immune and not colicin producing. Figure 2 shows an agarose gel with plas-
Fig. 2. Electrophoresis of Chimaera Plasmids. Electrophoresis in a 1% agarose slab gel of different plasmids isolated from Ap, El, E2 s and not colicin producing colonies. (a)calf satellite I/EcoRI, (b) pSF 2124, (c) pSF 2124/EcoRI, (d) pGM 213, (e) pGM 213/EcoR1, (7")pGM 214, (g) pGM 214/EcoRI, (h) pGM 215, (i) pGM 215/EcoRI, (k) pGM 216, (l) pGM216/EcoRI, (m) pGM217, (n) pGM 217/EcoRI, (o) pGM 218, (p) pGM 218/EcoRI, (q) pGM 219, (r) pGM 219/EcoRI
26
F. Gautier et al. : Cloning of Calf Thymus Satellite I DNA in Escherichia coli
A OD AT
0s
! 78
80
I
910
~
I
~ T (°El
:"
Fig. 3. Electrophoresis of BamI-cut pGM 214. Electrophoresis in a 1% agarose slab gel of (a) Lambda DNA/EcoRI as molecular weight marker, (b) pSF 2124, (c) pSF 2124/BamI, (d) pGM 214 and (e) pGM 214/BamI
Fig. 4. Melting Curve of EcoRI-cut pGM 214. The differentia1 melting curve of EcoRI-cut pGM 214 DNA was made in 50 mM Na2HPOJNaHzPO4, pH 7.4, with a heating rate of 0.5 ° C/min. The change in UV absorbance was automatically recorded
mid DNAs from several independently isolated clones. All supercoiled DNAs ran slower than the original pSF 2124 DNA after electrophoresis in an agarose gel. An EcoRI digest of these new plasmid produces the linear pSF 2124 and the EcoRI-fragment of the satellite I DNA. R.BamI does not cut in the satellite DNA, but cuts in the ampicillin transposon, converting the chimaera plasmids into linear molecules (Fig. 3). For the linear molecules a molecular weight of 8.3 x 106 was estimated by mobility on agarose gel electrophoresis. This infers that the chimaeras contain only one "repeat unit" of the satellite I in addition to the 7.35 x 106 dalton pSF 2124 cloning vector. This was confirmed by agarose gel electrophoresis, sedimentation analysis and CsC1 density centrifugation in the analytical ultracentrifuge. The plasmid chimaeras were found to an S-values of 31.5. Density centrifugation shows a density shift to slightly increased G + C content and establishes that only one "repeat unit" is integrated in the chimaeras. For further investigations plasmid pGM 214 was selected since the experiments described indicate that the other plasmids are all the same size.
range (transition breath 8° C) whereas a very sharp transition (0.5 ° C) was observed for the satellite DNA part which melts at a higher temperature due to its higher G + C content. Cleavage of the original satellite I DNA, ofplasmid pSF 2124, and of the chimaera plasmid pGM 214 with HinclI (Fig. 5) yields fragments which are in agreement with previous observations of Botehan (1974) and Philippsen et al. (1975). Satellite I DNA gives fragments of 0.125x 106 and 0.74 x 106. Plasmid pSF 2124 gives two fragments of 5.45 x 106 and 1.85 x 106. After integration of the EcoRI-fragment of satellite I in pSF 2124, cleavage with HinclI should yield four fragments. The small fragment (0.125x 106) derives from the satellite DNA. One of the three other fragments is identical with the 1.85 × 106 fragment of pSF 2124. The two other fragments are combinations between parts of pSF 2124 and satellite I (see map). The fact that we could not detect the 0.74 x 106 fragment cut from the original satellite I in the hybrid plasmid, further indicates that only one "repeat unit" of the satellite DNA was integrated in the hybrid plasmid. Figure 5 shows that the purified satellite I DNA from calf thymus is not cut by R. Sinai. After integration of the satellite DNA into plasmid pSF 2124, the resulting chimaera plasmid pGM 214 is cleaved into four fragments by Sinai. As Sinai cuts plasmid pSF 2124 only once (Fig. 6), three cleavage sites must therefore be located in the satellite part. The recognition sequence for Sinai is CCCGGG (Greene and Mulder, unpubl, observations). The overall methyla-
Characterization of pGM 214 by Melting Curves and Analyses with Restriction Enzymes EcoRL HinclI, SmaL EclI and EcoRI* Absorbance melting of plasmid pGM 214 after cleavage with EcoRI yields profiles with biphasic transitions (Fig. 4). The pSF 2124 part melts over a broad
F. Gautier et al. : Cloning of Calf Thymus Satellite I DNA in Escherichia coli
27
Fig. 5. Electrophoresis of Satellite I DNA and pGM 214 DNA cut with different Restriction Enzymes. Electrophoresis in a 2% slab gel of (a) lambda DNA/HindIII+EcoRI, (b) calf satellite I/EcoRI, (c) pGM 214/EcoRI, (d) pSF 2124/EcoRI, (e) satellite I/SmaL (/) pGM 214~Sinai, (g) satellite I/HincII, (h) pGM 214/HinclI, (/) pSF 2124/Ecli, (lc) pGM 214/EclI, (I) satellite I/EclI, (m) pSF 2124/HincII
Fig. 6. Electrophoresis ofpSF 2124 DNA. Electrophoresis in a 1% agarose slab gel of (a) pSF 2124/BamI, (b) pSF 2124/BamI + EcoRI, (c) pSF 2124/SmaI, (d) pSF 2124/Smai + EcoRI, (e) pSF 2124/Sinai + BamI + EcoRI and Q~ Lambda
DNA/EcoRI
28
F. Gautier et al. : Cloning of Caif Thymus Satellite I DNA in Escherichia coil
tion in calf DNA (5-methylcytosine) is known to be about 1% (Filipski et al., 1973). The ratio of 5-methylcytosine to cytosine is 0.05 (Deutsch, Razin and Sedat, 1976). We therefore propose that the original satellite I is not cut by SrnaI due to the methylation of most of the C-rich recognition sites for Sinai. To account for the complete lack of cutting of the original satellite by SrnaI the extent of methylation would have to be about 25% of the cytosine redidues, implying that the original satellite is differentially overmethylated compared to the bulk DNA. After replication in a modification free system (E. coli 5K) the Sinai sites are available for cutting. Figure 5 shows the cleavage patterns of pSF 2124, pGM 214 and the original satellite DNA with the restriction endonuclease EclI. Plasmids pSF 2124 and pGM 214 are both cut three times. The largest fragment of p G M 214 is 0.93 x 106 daltons larger than the largest fragment of pSF 2124, showing that the satellite part is not cut by EclI. The original satellite which still contains about 10%
satellite II DNA yields four faint bands with EclI of 0.23, 0.46, 0.92 and 1.8 x l06 molecular weight. The largest part of this DNA however is not cut. After a double digestion with EclI and EcoRI this mixture of satellites I and II is cut down to two fragments (Fig. 7) of 0.23 and 0.92 x 106. A double digest of pG M 214 with EclI and EcoRI (Fig. 7) gives three new fragments compared to EclI alone. One fragment has the same size as the EcoRI "repeat unit" of the satellite I. These data can be explained on the assumption that the satellite II (10%) gives a repeat unit of 350 base pairs with EclI and that the satellite I is not cut by EclI. EcoRI* cuts satellite I DNA and pGM 214 into may many fragments (Fig. 8). Those fragments derived from the satellite are in the range of 0.2 to 0.9 x 106. It is not sure that the EcoRI* digest was complete, although we incubated for 6 h under very efficient conditions (H. Mayer, manuscript in prep.). Under these conditions a unit fragment of low molecular weight was not detected.
Fig. 7. Electrophoresis of p G M 214 and Satellite D N A cut with EclI. Electrophoresis in a 1% agarose slab gel of (a) p G M 214/EclI, (b) p G M 214/EclI+EcoRI, (c) calf satellite/EclI, (d) calf satellite/ EclI+EcoRI and (e) ColE1/EcoRI* as molecular weight marker. The 0.23 x 106 band of calf sateIlite/EclI is hardly to see on the print
Fig. 8. Electrophoresis of p G M 214 and Calf Satellite I DNA cut with EcoRI*. Electrophoresis in a 1% agarose slab gel of (a) p G M 214/EclI, (b) p G M 214/EcoRI*, and (c) calf satellite I/EcoRI*
F. Gautier et al. : Cloning of Calf Thymus Satellite I DNA in Escherichia coli
0~
'
~
29
0
/ O
-f, O
X
3
E
2
/
o/;/ o/
/°
I
10
I
t(min)
20
Fig. 10. Electrophoresis of Rsc11/HincII and pGM 214/HincII. Electrophoresis in 1% agarose slab gel of (a) Rscll/HincII and (b) pGM 214/HincII
1t7 vitro Transcription of pGM 214
Fig. 9. a In vitro Transcription of pGM 214. The kinetics of transcription of pGM 214 with E. coli RNA polymerase was followed by precipitating 5 gl samples of the transcription assay with 10% TCA at indicated time points, b Filter Binding of DNA/RNA Polymerase Complexes. Electrophoresis in a 1% agarose slab gel of (a) filter bound pGM 214/SmaI+EcoRI (partial), (b) filter bound pGM 214/EcoRI. (a) and (b) were complexes of pGM 214 fragments and RNA polymerase (see experimental procedures); (c) not bound pGM 214/SmaI+EcoRI and (d) not bound pGM 214/EcoRI. The background is due to denatured E. coli chromosomal DNA
With purified RNA polymerase holo-enzyme pGM 214 was efficiently transcribed. The kinetics of RNA synthesis (Fig. 9a) shows that the transcription on the supercoiled template is nearly complete by 12 min. The same effect has been observed for other plasmids too (Gautier and Goebel, manuscript in prep.) and can be explained by the fact that part of the synthesized RNA remains hybridized to the template and unwinds the DNA. Reinitiation of RNA synthesis seems to be inhibited by the new structure of the DNA-RNA hybrid. After 12 min most of the synthesized RNA sediments at 16S and 21S in a sucrose gradient. The in vitro 3H-labelled RNA was hybridized to the EcoRI fragments of pGM 214 according to Southern (1975). The amount of radioactivity which was specifically bound to the large fragment (pSF 2124 part of pGM 214) was 140,000 cpm, whereas the small fragment (satellite I part of pGM 214) bound 13,500 cpm. This ratio of hybridization (10.7:1) corresponds reasonably well with the ratio of the molecular weights of pSF 2124 and satellite I repeat unit in
30
F. Gautier et al. : Cloning of Calf Thymus Satellite I D N A in Escherichia coli
pGM 214, which is 8 : 1, and suggests that the satellite part of pGM 214 is transcribed in vitro, most probably by read-through from a ColE1 promoter. A read-through mechanism is further supported by filter binding studies in which RNA polymerase is used to bind EcoRI fragments or Sinai plus EcoRI fragments of pGM 214 to cellulose membrane filters. The data indicate that RNA polymerase binds only to the ColE1-Ap part suggesting that the satellite I part has no promoter for E. coli RNA polymerase (Fig. 9b).
Physical Map of p G M 214 (and p S F 2124) with EcoRI, SmaL HinclI, HindlI1, BamI and EclI Different double digest and triple digests led to a physical map of pGM 214 and pSF 2124 (Fig. 11). As noted already the replication of the satellite I in E. coli 5K makes three SrnaI sites available for cutting. The transposition of the ampicillin transposon to ColE1 (So et al., 1975) introduces two new HincII sites compared to ColE1 and one new Barnl site. The mapping of the HincII, Sinai and EcoRI sites and the absence of a BamI site in ColE1 has previously been established (Collins et al., 1976). It is interesting to notice that the HincII site mapped by Collins et al. (1976) in ColE1 is not found in the ColE1-Ap plasmid. The two HincII sites of the Col E1-Ap plasmid create a fragment which is identical in size with one of the HincII fragments of the mini-R1 plasmid Rsc 11 (Goebel and Bonewald, 1975), which is also known to derive from the ampicillin transposon of R1 (Fig. 10).
• )2 2 , 0'60 " MW =8,3.106
.
~ ~ ~ ''~ 0,44 0.33 0,40
/__
EcoR, ~rHinc II BQmI --..,,..~Ecl [
Fig. 11. Physical Map of p G M 214 with HinclI, EcoRI, Sinai, BamI and EclI. p G M 214 has no HindIII site
Discussion
In this paper, we have demonstrated that a repeat unit ofa eucaryotic satellite (satellite I of calf thymus) DNA comprising 1400 base pairs can be stably propagated in the procaryotic E. coli. As with other eucaryotic DNA segments cloned in E. coli (e.g. Morrow et al., 1974) this was achieved by incorporation of the EcoRI fragment of purified satellite I DNA into a suitable plasmid vector (Co/E1-Ap=pSF 2124). It should be emphasized at this point that the satellite I DNA of calf thymus used in these experiments was highly purified by procedures previously developed (H. Bfinemann, pers. communication). The preparation used in these studies still contained some satellite II DNA but no other DNA with unique sequences. All clones examined showed a homogenous population of recombinant plasmid DNA with one repeat unit of satellite 1 DNA incorporated into the EcoRI site of the cloning vector ColE1-Ap. There were no chimaera plasmids detected which carried multimers of the repeat unit even in experiments where the satellite DNA was only partially digested and multimeric repeat units prevailed in the digest before ligation. The reason for the obvious "selective advantage" of the monomeric units is unknown, but one can speculate that tandem repeats of such relatively short nucleotide sequences may not be tolerated by the E. coli cell. Our experiments, however, do not rule out the possibility that clones carrying hybrid plasmids with more than one repeat unit may occur with a low frequency. As shown by cleavage with several restriction enzymes the molecular properties of the satellite repeat unit are not altered after prolonged propagation in the procaryotic background with one important exception. Restriction endonuclease SrnaI which recognizes the nucleotide sequence CCCGGG does not cleave the original satellite I DNA but introduces three breaks into the satellite part of the chimaera plasmid pGM 214. It has been observed before that some restriction enzymes do not completely degrade satellite DNA (H6rz et al., 1974) and it has been argued that this may be due to mutations in the satellite or to methylation of this DNA. The data obtained with Sinai enzyme support the latter idea since calf thymus DNA is known to be methylated with a ratio of 5-methylcytosine to cytosine being 0.05 (Deutsch et al., 1976). The recognition site for Sinai is very rich in cytosine and one can therefore expect that many SmaI sites may be blocked due to the methylation of this base. Beside the three SmaI recognition sites the repeat unit of satellite I carries a number of EcoRI* sites which creates several small fragments of this unit
F. Gautier et al. : Cloning of Calf Thymus Satellite I DNA in Escherichia coli
when isolated either from the eucaryotic or the procaryotic background. This fact supports the idea of Botchan (1974) that the repeat unit has a short 8-20 base pair substructure with enough divergence in this basic sequence to cause an apparent complexity of 200-250 base pairs. The newly isolated restriction endonuclease EclI (H. Hartmann, unpublished results) does not cleave satellite I DNA but introduces breaks into satellite II of calf thymus DNA creating a repeat unit of 350 base pairs which has not been found before with other restriction enzymes. The function of the satellite DNA in the eucaryotic chromosome is not understood and it is unknown whether satellite DNA may ever become transcribed within the eucaryotic cell. The in vitro transcription studies performed with the chimaera plasmid indicate that E. coli RNA polymerase can efficiently transcribe the satellite DNA part starting from a strong promoter on the Co/E1-Ap vector. As expected no binding for RNA polymerase was found on the satellite DNA. One great advantage of cloning parts of a eucaryotic chromosome in E. coli, especially with multicopy plasmids as cloning vectors, is the possibility of easily isolating large amounts of homogenous eucaryotic DNA fragments. In E. coli this DNA can be highly labelled with radioactive nucleotides. This possibility facilates some biophysical investigations e.g. sequencing, hybridization studies or DNA-Protein interaction studies. These studies would not be effected by "anomalous" physical properties of the eucaryotic DNA when the DNA is replicated in the bacterial cell, although the methylation may affect proteinDNA interactions. Acknowledgements. We are indebted to W. Ziarlik for expert technical assistance, to H. Schilling for operating the ultracentrifuge, to S. Falkow, J. Collins and Saedler for bacterial strains and to H. Bfinemann for purified calf thymus satellite I DNA. We thank our collaborators in our laboratory for giving us several restriction enzymes. We are grateful to J. Collins for helpful discussions and critical reading of the manuscript.
31
Cohen, S.N., Chang, A.C.Y. : Recirculation and autonomous replication of a sheared R-factor DNA segment in Escherichia coli transformants. Proc. nat. Acad. Sci. (Wash.) 70, 1293-1297 (1973) Collins, J., Jorgensen, P., Karlstr6m, H.O., Lindenmaier, W., Johnson, M., Sj6berg, B. : Restriction endonuclease mapping of the ColE1 plasmid and of Eseherichia coli genes carried on the transducing phage )~dnrd, ).dilv5 and 2drtfD 18 or on hybrid plasmids derived from them. DNA-Insertions, Editor Targum, C. Cold Spring Harbor: Cold Spring Harbour publication (in press) Deutsch, J., Razin, A., Sedat, J.: Analysis of 5-methylcytosine in DNA. Analyt. Biochem. 72, 586 592 (1976) Filipski, J., Thiery, J.-P., Bernardi, G.: An analysis of the bovine genome by Cs2SO4-Ag + density gradient centrifugation. J. molec. Biol. 80, 177 197 (1973) Gillespie, D., Spiegelman, S. : A quantitative assay for DNA-RNA hybrids with DNA immobilized on a membrane. J. molec. Biol. 12, 829-842 (1965) Goebel, W. : Studies on extrachromosomal DNA elements - Replication of the colicinogenic factor ColE1 in two temperature sensitive mutants of Escherichia coli defective in DNA replication. Europ. J. Biochem. 15, 311 320 (1970) Goebel, W., Bonewald, R.: Class of small multicopy plasmids originating from the mutant antibiotic resistance factor Rldrd19B2. J. Bact. 123, 658 665 (1975) H6rz, W., Hess, J., Zachau, H.G.: Highly regular arrangement of a restriction-nuclease-sensitive site in rodent satellite DNAs. Europ. J. Biochem. 45, 501-512 (1974) Landy, A., Ruedisueli, E., Robinson, L., Foeller, C., Ross, W.: Digestion of deoxyribonucleic acids from bacteriophage T7, 2 and q~80 with site-specific nucleases from Hemophilus influenzae strain Rs and strain Rd. Biochem. 13, 2134-2142 (1974) Laskey, R.A., Mills, A.D.: Quantitative film detection of 3H and 1~C in polyacrylamid gels by fluorography. Europ. J. Biochem. 56, 335-341 (1975) Morrow, J.F., Cohen, S.N., Chang, A.C.Y., Boyer, W.H., Goodman, H.M., Helling, R.B.: Replication and transcription of eukaryotic DNA in Escherichia coli. Proc. nat. Acad. Sci. (Wash.) 71, 1743-1747 (1974) Philippsen, P., Streek, R.E., Zachau, H.G., Miiller, W. : Investigation of the repetitive sequences in calf DNA by cleavage with restriction nucleases. Europ. J. Biochem. 57, 55-68 (1975) So, M., Gill, R., Falkow, S. : The generation of a ColEI-Ap r cloning vehicle which allows detection of inserted DNA. Molec. gen. Genet. 142, 239 249 (1975) Southern, E.M. : Detection of specific sequences among DNA fragments seperated by gel electrophoresis. J. molec. Biol. 98, 503 517 (1975) Wilson, G.A., Young, F.E. : Isolation of a sequence-specific endonuclease (BamI) from Bacillus amyloliquefaciens H. J. molec. Biol. 97, 123-125 (1975)
References
Communicated by F. Kaudewitz
Botchan, M.R. : Bovine satellite I DNA consists of repetitive units 1400 base pairs in length. Nature (Lond.) 251, 288-292 (1974)
Received and accepted August 4, 1976
Cloning of Calf Thymus Satellite I DNA in Escherichia coli Fritz Gautier, Hubert Mayer, and Werner Goebel Gesellschaft for Biotechnologische Forschung mbH Abteilung Genetik, Mascheroder Weg 1, D-3300 Braunschweig-St6ckheim, Federal Republic of Germany
Summary. The 1400 base pair repeat produced by digestion of calf satellite I DNA ((p= 1.714 g/cm 3) with EcoRI,was cloned in E. coli. The hybrid plasmid (pGM 214) which contains the Co/E1-Ap vector (pSF 2124) and the 1400 base pair fragment replicates stably in E. coIi and can be amplified by chloramphenicol treatment. No clone was found in which more than one "repeat unit" of the satellite I DNA was present in the chimaera plasmid. Digestion of the original satellite I and the plasmid pGM 214 with R. Sinai shows that the satellite DNA replicated in E. coli is cleaved by the restriction endonuclease Sinai whereas the original satellite I DNA from calf thymus is not, suggesting that the satellite I contains a large amount of modified cytosine or guanosine, probably 5-methyl-cytosine. R.EcoRI* produces a number of fragments with the satellite I in the range of 300 base pairs to 1400 base pairs. A physical map of pGM 214 (and pSF 2124) with R. EcoRI, R. HincII, R.HindIII, R. SmaI, R. BamI and R. EclI was constructed. The 1400 base pair "repeat unit" in the pGM 214 is efficiently transcribed in vitro by purified RNA polymerase, starting from a pSF 2124 promoter. The restriction enzyme EcII produces a 350 base pair repeat with calf satellite II (~0=1,722 g/cm3), whereas the satellite I is not cut by this enzyme.
Introduction
Calf thymus DNA contains four satellites representing about 14% of the total DNA (Filipski, Thiery and Bernardi, 1973). The function and evolution of the satellites are still unclear. Several attempts have been made to analyze the satellite structure with res-
triction enzymes (e.g. Botchan, 1974; Philippsen et al., 1975). H6rz, Hess and Zachau (1974) investigated the cleavage of mouse satellite DNA with Hind restriction endonucleases and found that the satellite DNA was remarkably resistent to these restriction enzymes. They discussed this fact on the basis 'of mutation in the recognition sites or by methylation of the cytosine residues. Filipski et al. (1973) showed a number of "anomalous" physical properties for the calf thymus satellites compared to bacterial DNAs of identical G + C content. For further investigation of the structure of calf thymus satellites it seemed quite helpful to us to clone the satellite I in E. coli. After cloning in E. coli with a suitable cloning vector it is easy to isolate large amounts of pure satellite DNA (even highly labelled) and no problems with methylation of the DNA or other "anomalous" DNA modifications will effect subsequent analysis with restriction nucleases. So, Gill and Falkow (1975) constructed a cloning vector (ColE1-Ap=pSF 2124) with one R.EcoRI site in the structural gene for colicin El. Because the Co/E1-Ap plasmid has two transformation markers, Ap and colicine immunity, and one integration marker, colicin production, we used this plasmid for cloning calf thymus satellite I in E. coli.
Experimental Procedures Bacterial Strains. E. coli 5K, a restriction and modification negativ strain, was used for all transformations. E. coli C 600 carrying the ColE1-Ap r plasmid (pSF 2124) was obtained from S. Falkow. E. coli K 12/W8 t68c857 was obtained from Saedler. JC 309 (CoIE1) and JC 411 R64drdl 1 (ColE2) was obtained from J. Collins. Media, buffer, centrifugation conditions, sources of reagents for bacterial cultures and D N A preparation were the same described by Goebel and Bonewald (1975). DNAs: Purified calf satellite I DNA was a gift from H. Brinemann. L a m b d a - D N A was prepared from t68c857 in E. coli K12/ W8 using standard procedures. Plasmid D N A was isolated accord-
24
F. Gautier et al. : Cloning of Calf Thymus Satellite l DNA in Escherichia colt
ing to Goebel (1970). Col plasmids were amplified in broth in the presence of chloramphenicol (150 lag/ml) at a cell density of 110 Klett units for 12 h at 37° C.
Enzymes. EcoRI and SmaI were isolated according to H. Mayer (manuscript in prep.); HincII was prepared according Landy et al. (1974) by G. Luibrand, BamI according Wilson and Young (1975) by H. Hartmann, HindIII by D. Blohm, EclI by H. Hartmann and kindly given to us. EclI is a new restriction enzyme isolated from Enterobacterium cloacae.
After incubation for 5 rain at 37o C without ATP and GTP, the reaction was started by addition of ATP and GTP. After 12 rain DNAse was added to a final concentration of 20 gg/ml. The DNAse treatment was stopped by heating the reaction mixture to 85° C for 15 min. After cooling SDS was added to a.final concentration of 0.05%. The total assay was used for hybridisation. In a second assay the rate of RNA synthesis was followed by removing 5 gl samples at different times. The 5 ~tl samples were precipitated on Whatman filters in 10% cold TCA, 0.1 M Na4P20~, washed 4x in 5% TCA, 1 x in ethanol/ether (60/40) and 1 x in ether. The dried filters were counted in toluene based scintillation fluid.
Assays for Enzymes. Digestion with EcoRI was carried out in 50 mM Tris-HC1, pH 7.5, 50 mM NaCI and 10 mM MgC12, with Sinai in 15 mM Tris-HC1, pH 9, 15 mM KC1 and 6 mM MgClz, with HindIII in 50 mM Tris-HC1, pH 7.5, 20 mM NaC1 and 20 mM MgCI2, with HincII in 5 mM Tris-HC1, pH 7.9, 100 mM NaC1 and 10 mM MgClz, with BamI in 10 mM Tris-HC1, pH 7.4, 150 mM NaC1 and 10 mM MgClz, with EclI in 40 mM Tris-HC1, pH 7.5, 100 mM NaC1 and 20 mM MgC12, with EcoRI* in 25 mM Tris-HC1, pH 8.5, 2 mM MgC12 and 50% ethylene glycol. Except for EcoRI* (6 h) the incubation time was 30 min at 37° C. Ligation with T4 DNA ligase (purchased from Miles Lab.) was done in 66 mM Tris-HCl, pH 7.6, 6.6 mM MgCI2, 1 mM ATP, 0.66 mM dithiothreitol, 1 mg/ml BSA for 15 h at 0° C; 1 gl ligase (10 units/ml) was added to the reaction mixture (total volume 100 gl).
Hybridization Procedure. For hybridization of in vitro synthesized RNA to the EcoRI fragments of pGM 214 the EcoRI-cut plasmid was electrophoresed on 1% agarose gel. The gel was stained, photographed and cut into strips. The DNA was transferred to nitrocellulose filter (0.45 ~t, Schleicher & Schfill) according to the procedure described by Southern (1975). Prior to hybridization the filter was incubated in 0.01% BSA, 0.01% FicolI and 0.01% polyvinylpyrollidon in 4 x SSC for 5 h at 65° C. Hybridization and washing of the filters were carried out according to standard procedures (Gillespie and Spiegelman, 1965). The dried filters were fluorographed according to Laskey (1975). After fluorography, the filter was cut at the sides where bands could be seen and counted in a liquid scintillation counter.
Analytical Ultracentr~tgation. Analytical ultracentrifugation was Electrophoresis in Agarose Gels. 1% or 2% slab gels in a vertical electrophoresis apparatus (construction of our laboratory) were run in 36 mM Tris-HC1, 30 mM NaHzPO 4 and 10 mM EDTA, pH 7.5. The gels were stained with ethidium bromide (0.5 ~tg/ml) and photographed under UV light.
carried out in a Beckman Model E ultracentrifuge with UV scanner. The samples for centrifugation were made up to 1.705 g/cm 3 CsC1 in TES with 1 ~tg to 3 ~tg plasmid DNA and Micrococeus luteus DNA (~0= 1.731 g/cm 3) as internal marker.
Filter Binding. EcoRI-cut pGM 214 and EcoRI- and Sinai-cut Construction of the Hybrid Plasmids. Plasmid pSF 2124 (Cole 1-Ap) was cut to the linear form with EcoRI. Calf satellite I DNA was completely digested with EcoRI to the 1400 base pair repeat for one assay, and only partially digested with EcoRI for two other assays. The DNAs were precipitated with 60% (v/v) ethanol at - 20 ° C, centrifuged and dissolved in ligase buffer. A tenfold excess of satellite I fragments was added to the EcoRI-digested pSF 2124 DNA and ligated with T4 ligase. E. colt 5K was transformed with the whole iigation mixtures according to standard procedures (Cohen and Chang, 1973). The transformed cultures were incubated for 90 min at 37°C in ENB medium and aliquots were plated on agar plates with 50 gg/ml ampicillin. For testing colicin production single colonies were grown in ENB plates over night at 37° C. The colonies were killed with chloroform and overlayed with E. colt C in soft agar (0.2 ml from a overnight culture of E. colt C to 50 ml soft agar). After incubation for 6 h at 37° C, colonies could be detected which did not make clear lysis zones in the E. colt C overlay. Those non colicin producing colonies were taken from parallel plates and tested for colicin E1 resistance and colicin E2 sensitivity by cross-streaking on agar plates with colicin E1 on one side of the plate and colicin E2 on the other side. From all colonies which were Ap, E1 immune and not colicin producing, 1 ml cultures were grown in small tubes. After amplification with chloramphenicol (150 gg/ml) over night, cleared lysates were prepared, which were directly submitted to agarose gel electrophoresis. From all colonies which contained a plasmid larger than the original pSF 2124, DNA preparations were made according to Goebel (1970). The experiments were carried out under P2 conditions.
h~ vitro Transcription Assay. In vitro transcription with purified RNA polymerase was made in 0.15 M KC1, 0.01 M MgC12, 0.5 rag/ ml BSA, 0.01 M dithiothreitol, 0.04 M Tris-HC1 (pH 7.9), 10 ~ M EDTA, 4x 10 -4 M ATP, UTP, CTP and 50 gCi 3H-GTP, 1 I-tg DNA and 10 ~tg RNA polymerase in a total volume of 0.1 ml.
pGM 214 DNA was incubated for 10 min in the transcription buffer without the nucleotide triphosphates with a 10-fold excess of RNA polymerase. After 10 min at 37° C heat denatured E. coli DNA was added (30 ~tg/ml) and incubation was continued for 10 min at 37° C. The solution was filtered through a presoaked nitrocellulose filter. The filters were washed with 0.15 M KC1 in 4 mM Tris-HC1, pH 7.9, cut into small pieces with a scalpel and incubated with proteinase K (200 gg/ml) for 15 rain and with 0.01% SDS for another 15 min. The solution was electrophoresed on agarose gels and the gels were stained and photographed.
Melting Curve. pGM 214 was cut with EcoRI and precipitated with 60% (v/v) ethanol at - 2 0 ° C, centrifuged and dissolved in 50 mM Na2HPOJNaH2PO~, pH 7.4. A sample with 1.20D at 265 nm was melted in a Gilford spectrophotometer with a heating rate of 0.5° C/rain.
Results
Cloning o f the C a l f Satellite I D N A with p S F 2124 ( C o l E l - A p ) Figure 1a shows a densitometer tracing of an analytical d e n s i t y g r a d i e n t u l t r a c e n t r i f u g a t i o n o f t h e p u r i f i e d c a l f s a t e l l i t e I D N A . I t c a n b e s e e n t h a t t h e r e is still a b o u t 1 0 % c a l f s a t e l l i t e II D N A p r e s e n t in t h i s p r e p a r a t i o n . N o o t h e r D N A is p r e s e n t as u n i q u e s e q u e n c e s , a n d satellites III a n d IV are absent. Figure 1 b s h o w s t h a t t h e s a t e l l i t e I D N A is c u t b y E c o R I t o a u n i q u e size ( m o l e c u l a r w e i g h t o f 0.925 x 10 6 ( B o t c h a n , 1974)),
F. Gautier et al. : Cloning of Calf Thymus Satellite I DNA in Escherichia coli
25
5= 1,731g/cm 3 jo= 1,722 g/cm3 jo= 1,71/, g/cm 3
a
Increasing density
>
Fig, 1. a UV-scan of Analytical Ultracentrifugation of Calf Satel-
lite I DNA. Calf satellite I DNA was centrifuged in CsC1 with 44,000 rpm at 25°C for 40 h with Micrococcus luteus DNA ( = 1.731 g/cm 3) as internal marker. Ultraviolet-absorbance tracings of the analytical density gradient at 265 nm were taken after 40 h. b Electrophoresis of EcoRI-cut Satellite DNA. Electrophoresis in a 1% agarose slab gel with 100 mA/150 V for 3 h in electrophoresis buffer (experimental procedures) of calf satellite I DNA, partially digested (a) and totally digested (b) with EcoRI
or to oligomers of this fragment. The satellite II D N A is not cut by EcoRI (Philippsen et al., 1975) and stays near the top of the gel. The cloning vector pSF 2124 is cut once by EcoRI (So et al., 1975) to give the linear molecule (Fig. 2). Plasmid pSF 2124 and the EcoRI-cut satellite I DNA were joined by T4 ligase, and E. coli 5K was transformed with the ligation mixture. Three independent transformations were made. In one a complete
digest of satellite I DNA with EcoRI was used, and in two others partially digested DNA. Amongst 2000 ampicillin-resistent (Ap) transformants 300 were further screened for colicin production. Among those, 40 non-colicin-producing colonies were checked for resistance to colicin E1 and sensitivity to colicin E2 (i.e. E1 immune). D N A preparations were made from colonies which were Ap, E1 immune and not colicin producing. Figure 2 shows an agarose gel with plas-
Fig. 2. Electrophoresis of Chimaera Plasmids. Electrophoresis in a 1% agarose slab gel of different plasmids isolated from Ap, El, E2 s and not colicin producing colonies. (a)calf satellite I/EcoRI, (b) pSF 2124, (c) pSF 2124/EcoRI, (d) pGM 213, (e) pGM 213/EcoR1, (7")pGM 214, (g) pGM 214/EcoRI, (h) pGM 215, (i) pGM 215/EcoRI, (k) pGM 216, (l) pGM216/EcoRI, (m) pGM217, (n) pGM 217/EcoRI, (o) pGM 218, (p) pGM 218/EcoRI, (q) pGM 219, (r) pGM 219/EcoRI
26
F. Gautier et al. : Cloning of Calf Thymus Satellite I DNA in Escherichia coli
A OD AT
0s
! 78
80
I
910
~
I
~ T (°El
:"
Fig. 3. Electrophoresis of BamI-cut pGM 214. Electrophoresis in a 1% agarose slab gel of (a) Lambda DNA/EcoRI as molecular weight marker, (b) pSF 2124, (c) pSF 2124/BamI, (d) pGM 214 and (e) pGM 214/BamI
Fig. 4. Melting Curve of EcoRI-cut pGM 214. The differentia1 melting curve of EcoRI-cut pGM 214 DNA was made in 50 mM Na2HPOJNaHzPO4, pH 7.4, with a heating rate of 0.5 ° C/min. The change in UV absorbance was automatically recorded
mid DNAs from several independently isolated clones. All supercoiled DNAs ran slower than the original pSF 2124 DNA after electrophoresis in an agarose gel. An EcoRI digest of these new plasmid produces the linear pSF 2124 and the EcoRI-fragment of the satellite I DNA. R.BamI does not cut in the satellite DNA, but cuts in the ampicillin transposon, converting the chimaera plasmids into linear molecules (Fig. 3). For the linear molecules a molecular weight of 8.3 x 106 was estimated by mobility on agarose gel electrophoresis. This infers that the chimaeras contain only one "repeat unit" of the satellite I in addition to the 7.35 x 106 dalton pSF 2124 cloning vector. This was confirmed by agarose gel electrophoresis, sedimentation analysis and CsC1 density centrifugation in the analytical ultracentrifuge. The plasmid chimaeras were found to an S-values of 31.5. Density centrifugation shows a density shift to slightly increased G + C content and establishes that only one "repeat unit" is integrated in the chimaeras. For further investigations plasmid pGM 214 was selected since the experiments described indicate that the other plasmids are all the same size.
range (transition breath 8° C) whereas a very sharp transition (0.5 ° C) was observed for the satellite DNA part which melts at a higher temperature due to its higher G + C content. Cleavage of the original satellite I DNA, ofplasmid pSF 2124, and of the chimaera plasmid pGM 214 with HinclI (Fig. 5) yields fragments which are in agreement with previous observations of Botehan (1974) and Philippsen et al. (1975). Satellite I DNA gives fragments of 0.125x 106 and 0.74 x 106. Plasmid pSF 2124 gives two fragments of 5.45 x 106 and 1.85 x 106. After integration of the EcoRI-fragment of satellite I in pSF 2124, cleavage with HinclI should yield four fragments. The small fragment (0.125x 106) derives from the satellite DNA. One of the three other fragments is identical with the 1.85 × 106 fragment of pSF 2124. The two other fragments are combinations between parts of pSF 2124 and satellite I (see map). The fact that we could not detect the 0.74 x 106 fragment cut from the original satellite I in the hybrid plasmid, further indicates that only one "repeat unit" of the satellite DNA was integrated in the hybrid plasmid. Figure 5 shows that the purified satellite I DNA from calf thymus is not cut by R. Sinai. After integration of the satellite DNA into plasmid pSF 2124, the resulting chimaera plasmid pGM 214 is cleaved into four fragments by Sinai. As Sinai cuts plasmid pSF 2124 only once (Fig. 6), three cleavage sites must therefore be located in the satellite part. The recognition sequence for Sinai is CCCGGG (Greene and Mulder, unpubl, observations). The overall methyla-
Characterization of pGM 214 by Melting Curves and Analyses with Restriction Enzymes EcoRL HinclI, SmaL EclI and EcoRI* Absorbance melting of plasmid pGM 214 after cleavage with EcoRI yields profiles with biphasic transitions (Fig. 4). The pSF 2124 part melts over a broad
F. Gautier et al. : Cloning of Calf Thymus Satellite I DNA in Escherichia coli
27
Fig. 5. Electrophoresis of Satellite I DNA and pGM 214 DNA cut with different Restriction Enzymes. Electrophoresis in a 2% slab gel of (a) lambda DNA/HindIII+EcoRI, (b) calf satellite I/EcoRI, (c) pGM 214/EcoRI, (d) pSF 2124/EcoRI, (e) satellite I/SmaL (/) pGM 214~Sinai, (g) satellite I/HincII, (h) pGM 214/HinclI, (/) pSF 2124/Ecli, (lc) pGM 214/EclI, (I) satellite I/EclI, (m) pSF 2124/HincII
Fig. 6. Electrophoresis ofpSF 2124 DNA. Electrophoresis in a 1% agarose slab gel of (a) pSF 2124/BamI, (b) pSF 2124/BamI + EcoRI, (c) pSF 2124/SmaI, (d) pSF 2124/Smai + EcoRI, (e) pSF 2124/Sinai + BamI + EcoRI and Q~ Lambda
DNA/EcoRI
28
F. Gautier et al. : Cloning of Caif Thymus Satellite I DNA in Escherichia coil
tion in calf DNA (5-methylcytosine) is known to be about 1% (Filipski et al., 1973). The ratio of 5-methylcytosine to cytosine is 0.05 (Deutsch, Razin and Sedat, 1976). We therefore propose that the original satellite I is not cut by SrnaI due to the methylation of most of the C-rich recognition sites for Sinai. To account for the complete lack of cutting of the original satellite by SrnaI the extent of methylation would have to be about 25% of the cytosine redidues, implying that the original satellite is differentially overmethylated compared to the bulk DNA. After replication in a modification free system (E. coli 5K) the Sinai sites are available for cutting. Figure 5 shows the cleavage patterns of pSF 2124, pGM 214 and the original satellite DNA with the restriction endonuclease EclI. Plasmids pSF 2124 and pGM 214 are both cut three times. The largest fragment of p G M 214 is 0.93 x 106 daltons larger than the largest fragment of pSF 2124, showing that the satellite part is not cut by EclI. The original satellite which still contains about 10%
satellite II DNA yields four faint bands with EclI of 0.23, 0.46, 0.92 and 1.8 x l06 molecular weight. The largest part of this DNA however is not cut. After a double digestion with EclI and EcoRI this mixture of satellites I and II is cut down to two fragments (Fig. 7) of 0.23 and 0.92 x 106. A double digest of pG M 214 with EclI and EcoRI (Fig. 7) gives three new fragments compared to EclI alone. One fragment has the same size as the EcoRI "repeat unit" of the satellite I. These data can be explained on the assumption that the satellite II (10%) gives a repeat unit of 350 base pairs with EclI and that the satellite I is not cut by EclI. EcoRI* cuts satellite I DNA and pGM 214 into may many fragments (Fig. 8). Those fragments derived from the satellite are in the range of 0.2 to 0.9 x 106. It is not sure that the EcoRI* digest was complete, although we incubated for 6 h under very efficient conditions (H. Mayer, manuscript in prep.). Under these conditions a unit fragment of low molecular weight was not detected.
Fig. 7. Electrophoresis of p G M 214 and Satellite D N A cut with EclI. Electrophoresis in a 1% agarose slab gel of (a) p G M 214/EclI, (b) p G M 214/EclI+EcoRI, (c) calf satellite/EclI, (d) calf satellite/ EclI+EcoRI and (e) ColE1/EcoRI* as molecular weight marker. The 0.23 x 106 band of calf sateIlite/EclI is hardly to see on the print
Fig. 8. Electrophoresis of p G M 214 and Calf Satellite I DNA cut with EcoRI*. Electrophoresis in a 1% agarose slab gel of (a) p G M 214/EclI, (b) p G M 214/EcoRI*, and (c) calf satellite I/EcoRI*
F. Gautier et al. : Cloning of Calf Thymus Satellite I DNA in Escherichia coli
0~
'
~
29
0
/ O
-f, O
X
3
E
2
/
o/;/ o/
/°
I
10
I
t(min)
20
Fig. 10. Electrophoresis of Rsc11/HincII and pGM 214/HincII. Electrophoresis in 1% agarose slab gel of (a) Rscll/HincII and (b) pGM 214/HincII
1t7 vitro Transcription of pGM 214
Fig. 9. a In vitro Transcription of pGM 214. The kinetics of transcription of pGM 214 with E. coli RNA polymerase was followed by precipitating 5 gl samples of the transcription assay with 10% TCA at indicated time points, b Filter Binding of DNA/RNA Polymerase Complexes. Electrophoresis in a 1% agarose slab gel of (a) filter bound pGM 214/SmaI+EcoRI (partial), (b) filter bound pGM 214/EcoRI. (a) and (b) were complexes of pGM 214 fragments and RNA polymerase (see experimental procedures); (c) not bound pGM 214/SmaI+EcoRI and (d) not bound pGM 214/EcoRI. The background is due to denatured E. coli chromosomal DNA
With purified RNA polymerase holo-enzyme pGM 214 was efficiently transcribed. The kinetics of RNA synthesis (Fig. 9a) shows that the transcription on the supercoiled template is nearly complete by 12 min. The same effect has been observed for other plasmids too (Gautier and Goebel, manuscript in prep.) and can be explained by the fact that part of the synthesized RNA remains hybridized to the template and unwinds the DNA. Reinitiation of RNA synthesis seems to be inhibited by the new structure of the DNA-RNA hybrid. After 12 min most of the synthesized RNA sediments at 16S and 21S in a sucrose gradient. The in vitro 3H-labelled RNA was hybridized to the EcoRI fragments of pGM 214 according to Southern (1975). The amount of radioactivity which was specifically bound to the large fragment (pSF 2124 part of pGM 214) was 140,000 cpm, whereas the small fragment (satellite I part of pGM 214) bound 13,500 cpm. This ratio of hybridization (10.7:1) corresponds reasonably well with the ratio of the molecular weights of pSF 2124 and satellite I repeat unit in
30
F. Gautier et al. : Cloning of Calf Thymus Satellite I D N A in Escherichia coli
pGM 214, which is 8 : 1, and suggests that the satellite part of pGM 214 is transcribed in vitro, most probably by read-through from a ColE1 promoter. A read-through mechanism is further supported by filter binding studies in which RNA polymerase is used to bind EcoRI fragments or Sinai plus EcoRI fragments of pGM 214 to cellulose membrane filters. The data indicate that RNA polymerase binds only to the ColE1-Ap part suggesting that the satellite I part has no promoter for E. coli RNA polymerase (Fig. 9b).
Physical Map of p G M 214 (and p S F 2124) with EcoRI, SmaL HinclI, HindlI1, BamI and EclI Different double digest and triple digests led to a physical map of pGM 214 and pSF 2124 (Fig. 11). As noted already the replication of the satellite I in E. coli 5K makes three SrnaI sites available for cutting. The transposition of the ampicillin transposon to ColE1 (So et al., 1975) introduces two new HincII sites compared to ColE1 and one new Barnl site. The mapping of the HincII, Sinai and EcoRI sites and the absence of a BamI site in ColE1 has previously been established (Collins et al., 1976). It is interesting to notice that the HincII site mapped by Collins et al. (1976) in ColE1 is not found in the ColE1-Ap plasmid. The two HincII sites of the Col E1-Ap plasmid create a fragment which is identical in size with one of the HincII fragments of the mini-R1 plasmid Rsc 11 (Goebel and Bonewald, 1975), which is also known to derive from the ampicillin transposon of R1 (Fig. 10).
• )2 2 , 0'60 " MW =8,3.106
.
~ ~ ~ ''~ 0,44 0.33 0,40
/__
EcoR, ~rHinc II BQmI --..,,..~Ecl [
Fig. 11. Physical Map of p G M 214 with HinclI, EcoRI, Sinai, BamI and EclI. p G M 214 has no HindIII site
Discussion
In this paper, we have demonstrated that a repeat unit ofa eucaryotic satellite (satellite I of calf thymus) DNA comprising 1400 base pairs can be stably propagated in the procaryotic E. coli. As with other eucaryotic DNA segments cloned in E. coli (e.g. Morrow et al., 1974) this was achieved by incorporation of the EcoRI fragment of purified satellite I DNA into a suitable plasmid vector (Co/E1-Ap=pSF 2124). It should be emphasized at this point that the satellite I DNA of calf thymus used in these experiments was highly purified by procedures previously developed (H. Bfinemann, pers. communication). The preparation used in these studies still contained some satellite II DNA but no other DNA with unique sequences. All clones examined showed a homogenous population of recombinant plasmid DNA with one repeat unit of satellite 1 DNA incorporated into the EcoRI site of the cloning vector ColE1-Ap. There were no chimaera plasmids detected which carried multimers of the repeat unit even in experiments where the satellite DNA was only partially digested and multimeric repeat units prevailed in the digest before ligation. The reason for the obvious "selective advantage" of the monomeric units is unknown, but one can speculate that tandem repeats of such relatively short nucleotide sequences may not be tolerated by the E. coli cell. Our experiments, however, do not rule out the possibility that clones carrying hybrid plasmids with more than one repeat unit may occur with a low frequency. As shown by cleavage with several restriction enzymes the molecular properties of the satellite repeat unit are not altered after prolonged propagation in the procaryotic background with one important exception. Restriction endonuclease SrnaI which recognizes the nucleotide sequence CCCGGG does not cleave the original satellite I DNA but introduces three breaks into the satellite part of the chimaera plasmid pGM 214. It has been observed before that some restriction enzymes do not completely degrade satellite DNA (H6rz et al., 1974) and it has been argued that this may be due to mutations in the satellite or to methylation of this DNA. The data obtained with Sinai enzyme support the latter idea since calf thymus DNA is known to be methylated with a ratio of 5-methylcytosine to cytosine being 0.05 (Deutsch et al., 1976). The recognition site for Sinai is very rich in cytosine and one can therefore expect that many SmaI sites may be blocked due to the methylation of this base. Beside the three SmaI recognition sites the repeat unit of satellite I carries a number of EcoRI* sites which creates several small fragments of this unit
F. Gautier et al. : Cloning of Calf Thymus Satellite I DNA in Escherichia coli
when isolated either from the eucaryotic or the procaryotic background. This fact supports the idea of Botchan (1974) that the repeat unit has a short 8-20 base pair substructure with enough divergence in this basic sequence to cause an apparent complexity of 200-250 base pairs. The newly isolated restriction endonuclease EclI (H. Hartmann, unpublished results) does not cleave satellite I DNA but introduces breaks into satellite II of calf thymus DNA creating a repeat unit of 350 base pairs which has not been found before with other restriction enzymes. The function of the satellite DNA in the eucaryotic chromosome is not understood and it is unknown whether satellite DNA may ever become transcribed within the eucaryotic cell. The in vitro transcription studies performed with the chimaera plasmid indicate that E. coli RNA polymerase can efficiently transcribe the satellite DNA part starting from a strong promoter on the Co/E1-Ap vector. As expected no binding for RNA polymerase was found on the satellite DNA. One great advantage of cloning parts of a eucaryotic chromosome in E. coli, especially with multicopy plasmids as cloning vectors, is the possibility of easily isolating large amounts of homogenous eucaryotic DNA fragments. In E. coli this DNA can be highly labelled with radioactive nucleotides. This possibility facilates some biophysical investigations e.g. sequencing, hybridization studies or DNA-Protein interaction studies. These studies would not be effected by "anomalous" physical properties of the eucaryotic DNA when the DNA is replicated in the bacterial cell, although the methylation may affect proteinDNA interactions. Acknowledgements. We are indebted to W. Ziarlik for expert technical assistance, to H. Schilling for operating the ultracentrifuge, to S. Falkow, J. Collins and Saedler for bacterial strains and to H. Bfinemann for purified calf thymus satellite I DNA. We thank our collaborators in our laboratory for giving us several restriction enzymes. We are grateful to J. Collins for helpful discussions and critical reading of the manuscript.
31
Cohen, S.N., Chang, A.C.Y. : Recirculation and autonomous replication of a sheared R-factor DNA segment in Escherichia coli transformants. Proc. nat. Acad. Sci. (Wash.) 70, 1293-1297 (1973) Collins, J., Jorgensen, P., Karlstr6m, H.O., Lindenmaier, W., Johnson, M., Sj6berg, B. : Restriction endonuclease mapping of the ColE1 plasmid and of Eseherichia coli genes carried on the transducing phage )~dnrd, ).dilv5 and 2drtfD 18 or on hybrid plasmids derived from them. DNA-Insertions, Editor Targum, C. Cold Spring Harbor: Cold Spring Harbour publication (in press) Deutsch, J., Razin, A., Sedat, J.: Analysis of 5-methylcytosine in DNA. Analyt. Biochem. 72, 586 592 (1976) Filipski, J., Thiery, J.-P., Bernardi, G.: An analysis of the bovine genome by Cs2SO4-Ag + density gradient centrifugation. J. molec. Biol. 80, 177 197 (1973) Gillespie, D., Spiegelman, S. : A quantitative assay for DNA-RNA hybrids with DNA immobilized on a membrane. J. molec. Biol. 12, 829-842 (1965) Goebel, W. : Studies on extrachromosomal DNA elements - Replication of the colicinogenic factor ColE1 in two temperature sensitive mutants of Escherichia coli defective in DNA replication. Europ. J. Biochem. 15, 311 320 (1970) Goebel, W., Bonewald, R.: Class of small multicopy plasmids originating from the mutant antibiotic resistance factor Rldrd19B2. J. Bact. 123, 658 665 (1975) H6rz, W., Hess, J., Zachau, H.G.: Highly regular arrangement of a restriction-nuclease-sensitive site in rodent satellite DNAs. Europ. J. Biochem. 45, 501-512 (1974) Landy, A., Ruedisueli, E., Robinson, L., Foeller, C., Ross, W.: Digestion of deoxyribonucleic acids from bacteriophage T7, 2 and q~80 with site-specific nucleases from Hemophilus influenzae strain Rs and strain Rd. Biochem. 13, 2134-2142 (1974) Laskey, R.A., Mills, A.D.: Quantitative film detection of 3H and 1~C in polyacrylamid gels by fluorography. Europ. J. Biochem. 56, 335-341 (1975) Morrow, J.F., Cohen, S.N., Chang, A.C.Y., Boyer, W.H., Goodman, H.M., Helling, R.B.: Replication and transcription of eukaryotic DNA in Escherichia coli. Proc. nat. Acad. Sci. (Wash.) 71, 1743-1747 (1974) Philippsen, P., Streek, R.E., Zachau, H.G., Miiller, W. : Investigation of the repetitive sequences in calf DNA by cleavage with restriction nucleases. Europ. J. Biochem. 57, 55-68 (1975) So, M., Gill, R., Falkow, S. : The generation of a ColEI-Ap r cloning vehicle which allows detection of inserted DNA. Molec. gen. Genet. 142, 239 249 (1975) Southern, E.M. : Detection of specific sequences among DNA fragments seperated by gel electrophoresis. J. molec. Biol. 98, 503 517 (1975) Wilson, G.A., Young, F.E. : Isolation of a sequence-specific endonuclease (BamI) from Bacillus amyloliquefaciens H. J. molec. Biol. 97, 123-125 (1975)
References
Communicated by F. Kaudewitz
Botchan, M.R. : Bovine satellite I DNA consists of repetitive units 1400 base pairs in length. Nature (Lond.) 251, 288-292 (1974)
Received and accepted August 4, 1976
