33
Gene, 30 (1984) 33-39 Elsevier GENE
1054
Tandem repeated DNA in an intergenic region of Herpes simplex virus type 1 (Patton) (Recombinant DNA; fragment polymorphism;
Southern transfer hybridization)
K. Umene *, R.J. Watson * and L.W. Enquist * Laboratory of Molecular Virology, National Cancer Institute, NIH. Bethesda, MD 20205 (U.S.A.) (Received
January
(Accepted
April 30th, 1984)
12th, 1984)
SUMMARY
When the entire Us region of HSV-1 (Patton) was cloned as an EcoRI fragment in bacteriophage AgtWES, the BamHI B6B5 fragment was observed to vary in size among independent isolates [Umene and Enquist, Gene 13 (198 1) 25 l-2681. This fragment polymorphism also occurred in DNA of HSV-1 single plaque isolates. We report here that this heterogeneity is due to variation in copy number of a 15-bp tandem repeat of sequence 5’-CCACTCCCCACCCAC-3’, which apparently lies in an intergenic region of the HSV-1 DNA.
INTRODUCTION
The HSV-1 DNA is a linear duplex molecule of approx. 150 kb consisting of two covalently linked components, L and S, comprising 82% and 18% of the genome, respectively (Fig. 1A). The ends of the DNA molecule have short, direct repeats (the “a” sequence) that are present in the inverse orientation
* Present
addresses:
Medicine,
Kyushu
0118192641-l Molecular
(K.U.) Department University
101 (main
Genetics,
of Virology,
60, Fukuoka
Univ. number);
Faculty
812 (Japan)
(R.J.W.
of Tel.
and L.W.E.)
Inc., 10320 Bren Road East, Minnetonka,
MN 55343 (U.S.A.) Tel. (612) 935-7335. Reprint
requests
Abbreviations: pairs;
should be sent to R.J. Watson. HSV-1,
kb, kilobases
internal TRs/IR,,
repeat
Herpes simplex virus type 1; bp, base
or kilobase
sequences
terminal
pairs;
of
and internal
region of HSV-1; Us, unique
the
TRJIR,, long
repeat
sequences
terminal
region
sequences
of
of the short
of the short region of
HSV-1.
0378-l 119/84/$03.00
0
1984 Elsevier
and
HSV-1;
Science
Publishers
at the L-S junction, and are implicated in isomerization of the L and S segments (Hayward et al., 1975; Clements et al., 1976; Roizman, 1979). The “a” sequence displays size heterogeneity between HSV-1 strains and between intrastrain plaque isolates (Wagner and Summers, 1978; Mocarski and Roizman, 1981; Davison and Wilkie, 1981). Size variation was attributed to differences in copy number of small G + C-rich tandem reiterations. Variation in copy number of two further multiple reiterations located in adjacent TRJR, sequence also was reported (Mocarski and Roizman, 1981; Davison and Wilkie, 1981). We and others have reported that a BamHI DNA fragment (the B6B5 or BamHI-Z fragment) that maps entirely within HSV-1 Us DNA also exhibited size heterogeneity between different human isolates and within cloned DNA fragments and single-plaque isolates derived from single strains (Lonsdale et al., 1980; Umene and Enquist, 1981). Here, we demonstrate that size variation of the B6B5 DNA fragment is due to variation in copy number of a G + C-rich
34
15bp tandem reiteration. This tandem reiteration maps in an intergenic region of the HSV-1 genome.
MATERIALS
AND METHODS
supplemented with 5 y0 fetal bovine serum. Stocks of HSV-1 (Patton strain) were grown from singleplaque purified virus using low multiplicities of infection (0.01 plaque forming units/cell) as described in Graham et al. (1978).
(a) llgtWES HSV-1 clones
(g) Analysis of mRNA
The description of >WES recombinant bacteriophage strains carrying the EcoRI-H fragment of HSV-1 (strain Patton) is found in Umene and Enquist (198 1). DNA was isolated from 2 phage that were purified through glycerol gradients as described by Umene and Enquist (1981).
The methods for analyzing mRNA species by Northern blot analysis and by nuclease Sl and exonuclease VII analysis were described in Watson et al. (1983).
(b) Bacteriological
RESULTS
methods
Media, phage growth conditions and bacterial strains are described in Umene and Enquist (198 1). All media contained 10 mM MgSO, unless otherwise specified. (c) Cloning DNA fragments in pBR322
BumHI fragments were cloned in pBR322 using standard procedures (Silhavy et al., 1984). The E. coli host for these experiments was LE392. (d) Gel electrophoresis
and Southern hybridization
Restriction endonucleases were purchased from New England Biolabs, Inc. (Beverly, MA) or Bethesda Research Laboratories, Inc. (Rockville, MD) and used in buffers recommended by the vendors. DNA markers, agarose and acrylamide gel procedures, Southern transfer and hybridization conditions have been presented by Denniston et al. (1981). (e) DNA sequencing
The Maxam-Gilbert (1980) chemical degradation system was used as described by Watson and Vande Woude (1982). (f) HSV methodology
African green monkey kidney (Vero) cells were grown to confluency in modified Eagles medium
AND DISCUSSION
(a) Size variability of the HpaII-Mb011
fragment
The HSV-1 Us DNA region is contained entirely within the EcoRI-H fragment that has been cloned in a bacteriophage 1 vector in both orientations (Enquist et al., 1979). We observed previously that secondary plaque isolates of these AgtWES . EcoRIH hybrid bacteriophage (i.e., Dee 24-SA, Dee 36-SA, Dee 36-202, Dee 33-206, MME24 and SRS-12) displayed size heterogeneity within the B6B5 subfragment of the HSV-1 DNA insert (Umene and Enquist, 1981). To investigate whether the phenomenon of B6B5 fragment size heterogeneity was also present in HSV-1 genomic DNA, single plaque isolates of HSV-1 strain Patton were first selected. The DNAs from primary (SP21, SP22, SP23, SP24 and SP27) and secondary (SP22-4 and SP22-6) HSV-1 plaque isolates were subjected to BamHI digestion and analysis by agarose gel electrophoresis. We found that the B6B5 DNA fragment contained by each of these plaque isolates had a different, characteristic mobility on agarose gels (not shown). Variation in size of the B6B5 fragment within single plaque isolates of HSV-1 strains has also been reported by Lonsdale et al. (1980). To identify the specific region subject to size variation, B6B5 DNA fragments isolated from a number of >WES *EcoRI-H hybrid bacteriophage and from HSV-1 plaque isolates were cloned in pBR322. The designation of these plasmid isolates and the origin of the cloned B6B5 fragments are given in Table I. The B6B5 fragments carried by each
35 TABLE I
of these plasmids were then analyzed by digestion with various restriction endonucleases, followed by polyacrylamide gel electrophoresis. It was found that a single HpaII-MboII subfragment displayed size heterogeneity (Fig. 2). This heterogeneous HpaIIMb011 subfragment mapped near the middle of B6B5 (Fig. 1C) and was estimated to vary in size from 145 bp (pUK166) to 360 bp (pUK164).
Plasmid constructions Plasmid
Insert derivation*
Reference
pUK160 pUK161 pUK162 pUK164 pUK165 pUK166 pUK173 pUK175 pUK176
Phage 1, Dee 24-SA Phage L, Dee 36-SA Phage 1, Dee 36-202 Phage I, Dee 33-206 Phage I, MME24 Phage a, SRS-12 HSV-1, SP22-4 HSV-1, SP22-6 HSV-1, SP23
Umene and Umene and Umene and Umene and Umene and Umene and This report This report This report
Enquist Enquist Enquist Enquist Enquist Enquist
(1981) (1981) (1981) (1981) (1981) (1981)
(b) Sequencing
of the heterogeneous
fragments
To determine the cause of variability, the HpaIIMb011 fragments from plasmids pUK173, 175 and 176 (all containing a B6B5 fragment cloned directly in pBR322 from HSV-1 plaque isolates) were sequenced using the Maxam and Gilbert (1980) technique. It was found that these DNA fragments had identical DNA sequences, but varied in the copy number of a 15-bp tandem repeat of sequence 5’-
a The plasmids carry a BumHI fragment (B6B5) from the indicated HSV-1 or I isolates.
A)
0
L
K
5
10
1
I
15
1 Kilobase pairs K-
H
;=
01
00
us
Cl
0
06 I
1
MbMbSlM~
1
TRS
a
2
1
I
pestn-2
4 W
//A
t@
Kilobilse pairs
85 I
repeats
Fig. 1. Maps ofHSV-1 DNA. (A) Structure ofthe HSV-1 genome. HSV-1 DNA consists oftwo covalently linked components designated L and S. Each component consists of unique sequences (U, and Us) bracketed by inverted repeat sequences (TR,, IR,, IR, and TR,). The short “a” sequences are indicated. The EcoRI-H and -K fragments are designated as H and K respectively. (B) Map of the S region of HSV-1. BarnHI sites (Bl-B8), EcoRI sites (RI) and a Sal1 site on the EcoRI-H and -K fragments are indicated. The variable BarnHI fragment (B6B5) is shown as a hatched segment. The SalI-B5 region cloned into pBR322 (pBS127-2) is indicated by a horizontal bar. (C) Map of the variable BamHI fragment (B6B5). Mb011 sites (Mb), HpaII sites (Hp) and a Sal1 site (Sl) are shown. The region of the 15-bp tandem array is shown as a hatched segment.
36
CCACTCCCCACCCAC-3’. Plasmids pUK173, pUK175 and pUK176 contained 16, 13 and 20 copies of this tandem repeat, respectively. The number of tandem repeats within plasmid subclones of B6B5 fragments derived from &t WES . EcoRI-H clones, were estimated from the observed sizes of HpaII-Mb011 DNA fragments on the polyacrylamide gel electropherogram (Fig. 2). The copy number was estimated to range from 8 (pUK166) to 22 (pUK164). A continuous 2560-bp sequence extending leftwards from the EcoRI cleavage site in TR, (Fig. 1) and including the entire immediate-early gene encoding IEmRNA-5, has been published previously (Watson and Vande Woude, 1982). The polyadenylation signal of the IEmRNA-5 gene, which is tran-
Ml23456789M
353 078 072
-
-
%I? = 271 234
-
194
-
116
-
72
-
Variable
I
HPaII/MboiI DNA Fragment
of a HpaII-MbolI
Fig. 2. Heterogeneity fragment
derived
IgtWES~ EcoRI-H chimeric
plasmids
and MboII, acrylamide bromide pUK161;
single-plaque
hybrids.
staining
fragments
fragments
were electrophoresed were visualized
and UV illumination.
lane 3, pUK162;
of the B6B5
of HSV-I
and from
in Table I were digested with HpaII
described
gel. DNA
fragment isolates
The B6B5 DNA
and the products
lane 6, pUK166; pUK176.
from
Lane 1, pUK160;
lane 4, pUKl64;
lane 7, pUK173;
on a 5% by ethidium lane 2,
lane 5, pUKl65;
lane 8, pUK175;
lane9,
The Hue111 digest of $X1 74 DNA was run in parallel
as a size marker
(M).
maps approx. 550 bp from the BamHI-5 (B5) site within the SalI-B5 DNA fragment. Using the procedure of Maxam and Gilbert (1980), we have now completed the sequence of the SalI-B5 fragment. That region of this sequence extending from the IEmRNA-5 polyadenylation signal to the Sal1 recognition site and including the variable HpaII-Mb011 fragment is represented in Fig. 3. The nucleotides are numbered consistently with an extension of the reported IEmRNA-5 gene sequence (Watson and Vande Woude, 1982). (c) Transcriptional studies The 15-bp tandem repeat is located 271 bp 3 ’ to the IEmRNA-5 polyadenylation signal (Fig. 4), and is not, therefore, represented in mRNA transcribed from this promoter. To elucidate whether the 15-bp tandem repeat is transcribed at any stage of the virus replication cycle, we analyzed the mRNAs which map within the S&-B5 fragment by hybridization of [32P]DNA probes to “Northern blots” containing immediate-early, early, and late HSV-1 mRNAs. The S&-B5 DNA fragment hybridized to three early mRNAs of estimated sizes 2.8, 1.6 and 1.3 kb, in addition to IEmRNA-5 (Fig. 4A). In contrast, a 53 1-bp TuqI fragment (nucleotides 2389-2800; 411 + 120 bp; Fig. 3) spanning the 9 x 15-bp tandem repeats and approx. 100 bp of the IEmRNA-5 3’ transcribed region, was found to hybridize to IEmRNA-5, but not to hybridize to any of the above early mRNAs (Fig. 4A). These and previous data (Watson et al., 1983) indicate that these early mRNAs map to the left of the 15-bp tandem repeats. To determine the direction of transcription of these early mRNAs, the SalI-B5 DNA fragment, uniquely 32P-labeled at either the 5’ or 3’ SalI termini, was hybridized to early RNA using the Berk and Sharp (1978) procedure, and the resultant RNA/DNA hybrids were treated with nuclease S 1. The DNA fragments protected against nuclease digestion were analyzed by electrophoresis on a denaturing urea-polyacrylamide gel (Fig. 4B). Protection of a single 205-base DNA fragment was observed with the 3’-32P-labeled probe only. These data indicate that the 2.8, 1.6 and 1.3 kb early mRNAs comprise a family of 3’ coterminal transcripts that are polyadenylated using a signal 205 bp from the Sal1 site, and which are transcribed from
37
2400 GACACACGCA
GGTGCTGTCT
TCGACGTGTT
CGTACGCGGG
GCTGTACTCG
ACCCACTGCC
TaqI TGCATCTGTT
TGGTGCGTTT
GGGTGTGGGG
ACCCGGCCCT
AACCCCACCC
2460 CTGTGCTAGG
GCAATTTGTA
CCCTTAATAA
ATTTCACAAA
(AIn C!%GATTTTAT CGCATCGTGT
2520 CTTATTGCGG
GGGAGAAAAC CGATGTCGGC
ATAGAAAACC
GCCATGATTC
TAAGACGTCC
2580 GAACGCGAGT
GGGTGGGGAA CAACCCATAC
CGGACAGATG
CCGATGAGCC
ACCCGCACCC
2640 TTGGGTGCGG
GAGGTACGGG GTGGTTTGTT
CATCCTATGG
TTCCGACCCC
ACAAACAGCC
2700 CCCAGAG’KG
GTTTGGGTAT
GGTTACATTT
TCTGTCTGGT
GGTCGGGCTT
GTTTCTTCCT MboII
CCACCCAC’J n CCAAAAATCA
ACCGGGAGAC HpaII
GTTAAAGGCT
GGGTGCAAAT
TGCGGGGTGA
TGGGGGGGAA GAGAGACGAC Mb011
CGCGTGTCGA
TGCGGTCTTT
TAGCGGAGCA
GCCACATCAG
GAGCGCCCCA
CCATGCCGAC
GCAGCGGGTG
0% AACAT&CCA
ATCGAACCCA TaqI
2759 TG[CCACTCC 2817 ATTTAATGTA -
, ACAGAACGGC CACGAGGAGA CAGGCGATCA
2077 AAGAAGGACG 2937 AATCCGCCCG 2997 CGTCTGCGTC Sal1
-GAC Fig 3. Nucleotide
sequence
from the EcoRI cleavage by horizontal by square
arrows)
of the 15-bp tandem
site in TR, (Watson
are indicated.
repeats
and of flanking
and Vande Woude,
The recognition
sites for
regions.
The nucleotides
1982) (Fig. 1). The polyadenylation
are numbered
extending
leftwards
sites [(A)n] and signals (underlined
TuqI, Mb011 and HpaII are underlined. The IS-bp tandem array is bounded
brackets
the strand opposite from IEmRNA-5. A potential polyadenylation signal (a CA dinucleotide preceded by the sequence 5’-ATTAAA-3’) is present at the predicted position (Fig. 3, nucleotides 2793-2814: note that the sequence of the opposite DNA strand is presented). The position of the IEmRNA-5 and early mRNA polyadenylation sites (Fig. 4), and the absence of hybridization of the 531-bp pBS127-2 TuqI fragment to any mRNA other than IEmRNA-5 (Fig. 5), indicate that the 15-bp repeats lie in an intergenic region of HSV-1 DNA.
CONCLUSIONS
We have shown that size heterogeneity in the BamHI-B6B5 fragment of HSV-1 DNA is due to differences in the copy number of a 15-bp tandem repeat (5’-CCACTCCCCACCCAC-3’). The num-
ber of iterations was invariably rather high (8 to 22). Size heterogeneity of the B6B5 DNA fragment was observed during propagation of llgt WES . EcoRI-H hybrids in E. coli (Umene and Enquist, 198 1) as well as during propagation of H SV- 1 in Vero cells. While the repeats were fairly stable in B6B5-containing plasmids cloned in E. coli, evidence of gain and loss was observed in Southern blots of most DNA preparations. An analogous size variation phenomenon has been reported within the HSV-1 “a” sequence (Wagner and Summers, 1978; Mocarski and Roizman, 1981; Davison and Wilkie, 1981). Here, also, variation was due to differences in copy number of tandemly reiterated sequences. The sequences of these repeats, -CCGCTCCTCCCCand -CCGCCCCTCGCCCCCTC(strain 17; Davison and Wilkie, 1981), and -CGCTCCTCCCCC(DR2) and -CGCTCCTCCCCGCTCCCGCGGCCCCGCCCCCAACGCC- (DR4) (strain F; Mocarski and
38
(A>
53tbp
Sal -65 IE
IE
Early late
TaqI
Early
(W
Late
kb
2.8-
310, 28r/,,= 234” Is+
I .6I .3-
118-
Fig. 4. Analysis of mRNAs encoded by B6BS. (A) Polyadenylated cytoplasmic immediate-early (IE), early, and late HSV-1 RNAs were resolved by electrophoresis on methylmercury agarose gels and blotted onto diazobenzyloxymethyl paper. The RNA blots were hybridized with the SafI-B5 and 531-bp Tog1 subfragments of B6B5 “‘P-labeled by nick-translation. Hybridization was visualized by autoradiography. Indicated to the left of the autoradiographs are the positions of the 2.0-kb IEmRNA-5 and of the 2.8, 1.6 and 1.3-kb early mRNAs. (B) S 1 endonuclease analysis of early mRNAs. The S&I-B5 DNA fragment was uniquely 32P-labeled at either the 5’ or 3’ SnlI terminus. These DNA probes were denatured and incubated under hybridization conditions with early HSV-1 RNA ( + ) or with control yeast tRNA ( - ). Following incubation, the mixtures were treated with endonudease S 1, and DNA fragments protected from digestion were analyzed by electrophoresis on a denaturing polyacrylamide gel and autoradiography. The 32P-labeied Hue111 fragments of $X174 RF DNA were co-electrophoresed as size markers (M).
DNA a
synthesist
5’--TTGCCACTCCCCACCCACCCA ----AACGGTGAGGGGTGGGlGETTTA----
-
3’
Skppoge I ONA synthesis CCCCA T c CC A C AC B -
5’-
I” C~ACTC~C~A~CCACC~AAA~3’ --TT - -AAC WGAGGGGTGGGTGmTTTA Repair Second
5’c
or round
---of DNA
synthesff
--TTGCCACTCCCCACCCACCCACTCCCCACCCACCCAAA 3’-- GGTGAGGGGTGGGTGGGTGAGGGGTGGGTGGSTT -
--- --
Fig. 5. A slippage-repair model for generating the 15-bp tandem repeat. (A) Indicated is a single copy of the IS-bp tandem repeat Banked by the underscored 3-bp incomplete direct repeat. It is proposed that during DNA replication the right-most 3-bp repeat present in the upper nascent DNA strand was copied and then base paired with the complementary sequence present in the lower template strand. (B) Continued DNA synthesis resulted in formation of a tandem repeat of the 15bp sequence in the nascent strand. (C)By some repair mechanism, or during a second round of DNA synthesis, the nascent strand served as a template for generation of a tandem repeat of the 1S-bp sequence in the complementary DNA strand.
39
Roizman, 1981) are similar to the 15bp repeat unit described in this report in that they are all extremely G + C rich. A variety of internal repeats and symmetries can be observed in the 15bp tandem array, the signilicance of which remains to be seen. One short repeat, however, may give insight into the formation of this tandem array. In Fig. 5, we have drawn a portion of the reiterated region with only one repeat. We note the 3-bp sequence, 5’-CCA-3’) both at the left junction and at the end of the 15-bp repeat. We speculate that a primordial HSV-1 genome had a single 15-bp sequence bounded by this short repeat. By slippage during replication and pairing of the CCA segments, as shown, a duplication of the 15-bp sequence occurred. By repeating the process and, perhaps by unequal recombination between repeats, the number of iterations grew to the present variable number. With only one exception, all S-region-defective genomes analyzed by Denniston et al. (1981) contained a novel recombination joint in the B6B5 fragment, and it was suggested that this recombination joint was generated by interaction of the S region “a” sequence and DNA near the B5-B4 restriction sites (i.e., within the B6B5 DNA fragment). The repeated G + C-rich sequences within the “a” sequence and the 15-bp repeat may be involved in the recombination event that leads to defective DNA formation. Alternatively, the large number of small reiterations may be a reflection, rather than a cause, of regions of active recombination. We have recently observed novel duplications in HSV-1 stocks that, by DNA sequence analysis, arose by a non-homologous interaction between specific sites in the B6B5 DNA (including the 15-bp tandem array) and sites within Us and TR, (KU. and L.W.E., in preparation). The B6B5 DNA region is thus implicatedin anumber ofrearrangements ofHSV-1 DNA involving DNA recombination.
REFERENCES Berk, A.J. and Sharp, P.A.: Spliced early mRNAs of simian virus 40. Proc. Natl. Acad. Sci. USA 75 (1978) 1274-1278.
Clements, J.B., Co&i, R. and Wilkie, N.M.: Analysis of herpesvirus DNA substructure by means of restriction endonucleases. J. Gen. Viral. 30 (1976) 243-256. Davison, A.J. and Wilkie, N.M.: Nucleotide sequences of the joint between the L and S segments of Herpes simplex virus types 1 and 2. J. Gen. Virol. 55 (1981) 315-331. Denniston, K.J., Madden, M.J., Enquist, L.W. and Vande Woude, G.: Characterization of coliphage lambda hybrids carrying DNA fragments from Herpes simplex virus type 1 defective interfering particles. Gene 15 (1981) 365-378. Enquist, L.W., Madden, M.J., Schiop-Stansly, P. and Vande Woude, G.F.: Cloning of Herpes simplex type 1 DNA fragments in a bacteriophage lambda vector. Science 203 (1979) 541-544. Graham, B.J., Bengali, Z. and Vande Woude, G.F.: Physical map of the origin of defective DNA in Herpes simplex virus type 1 DNA. J. Virol. 25 (1978) 878-887. Hayward, G.S., Jacob, R.J., Wadsworth, S.C. and Roizman, B.: Anatomy of Herpes simplex virus DNA: evidence for four populations of molecules that differ in the relative orientations of their long and short components. Proc. Natl. Acad. Sci. USA 72 (1975) 4243-4247. Lonsdale, D.M., Brown, S.M., Lang, J., Subak-Sharpe, J.H., Koprowski, H. and Warren, K.: Variations in HSV isolated from human ganglia and the study of cloned variations in HSV-1. Ann. N.Y. Acad. Sci. 353 (1980) 291-308. Maxam, A.M. and Gilbert, W.: Sequencing end-labelled DNA with base-specific chemical cleavages, in Grossman, L. and Moldave, K. (Eds.), Methods in Enzymology, Vol. 65. Academic Press, New York, 1980, pp. 499-560. Mocarski, E.S. and Roizman, B.: Site-specific inversion sequence of the Herpes simplex virus genome: Domain and structural features. Proc. Natl. Acad. Sci. USA 78 (1981) 7047-7051. Roizman, B.: The structure and isomerization of Herpes simplex virus genomes. Cell 16 (1979) 481-494. Silhavy, T.J., Berman, M.L. and Enquist, L.W.: Experiments with Gene Fusions. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1984. Umene, K. and Enquist, L.W.: A deletion analysis of lambda hybrid phage carrying the Us region of Herpes simplex virus type 1 (Patton), I. Isolation of deletion derivatives and identification of &i-like sequences. Gene 13 (1981) 251-268. Wagner, M.J. and Summers, W.C.: Structure of the joint region and the termini of the DNA of Herpes simplex virus type 1. J. Virol. 27 (1978) 374-387. Watson, R.J. and Vande Woude, G.F.: DNA sequence of an immediate-early gene (IEmRNA-5) of Hepes simplex virus type 1. Nucl. Acids Res. 10 (1982) 979-991. Watson, R.J., Colberg-Poley, A.M., Marcus-Sekura, C.J., Carter, B.J. and Enquist, L.W.: Characterization of the Herpes simplex virus type 1 glycoprotein D mRNA and expression ofthis protein in Xenopus oocytes. Nucl. Acids Res. 11 (1983) 1507-1522. Communicated by W.C. Summers.