Volume 6 Number 9 1979

Nucleic Acids Research

Preparation of triple-block DNA polymers using recombinant DNA techniques

Erik Selsing+ and Robert D.Wells University of Wisconsin, Department of Biochemistry, College of Agricultural and Life Sciences, Madison, WI 53706, USA Received 12 April 1979 ABSTRACT The construction of several recombinant plasmid derivatives containing novel triple-block DNA sequence insertions is described. The protocol for these constructions involves synthesis of a heterogenous mixture of block oligomer duplexes, dA-dC-* '40 using pancreatic deoxyribonuclease and ter40 40 dT 0 & d C1-, minal trans erase. The synthetic duplexes were mixed with linearized and dG-tailed vectors and the DNA mixture used to transform E. coli. Triple-block sequences of the type dGidA.dC dGkdT.dC., characterized by DNA sequencing, were insertad tnto the BAm AI site of pBR322 and next to the lac wild-type and UVS promoter regions in pRW26 and pRW28. Similarly, sequences were inserted into the Sma I site of pACYC189 and could be excised by cleavage with Sma I since the procedure regenerates the recognition site. The approach provides a technique for the synthesis of a large family of defined sequence triple-block polymers in essentially unlimited amounts. Although these inserts contain sequences which have the potential for forming stable hairpin structures, the recombinant plasmids are stable and appear to replicate normally.

INTRODUCTION

Biochemical and biophysical studies on enzymatically synthesized DNA duplexes have shown that DNA properties at a given site depend on both the base sequence of the site and the base sequences of nearby regions (reviewed in 1 and 2-5). Studies with block DNA polymers of the type dC.dA.dT.dG. (6,7) have extended knowledge of DNA conformational behavior (8) and have led to a reconsideration of the theoretical framework for analysis of DNA melting behavior (9). Further studies with molecules with longer blocks and with other types of blocks would be interesting, however, synthesis of such DNA duplexes is timeconsuming and tedious, A major problem is the heterogenous nature of the products from most DNA synthetic reactions, C) Information Retrieval Limited 1 Falconberg Court London Wl V 5FG England

3025

Nucleic Acids Research Repeated isolation and characterization of each reaction product is required. Using recombinant DNA technology, we have developed a protocol for the construction of a new class of DNA block polymers having the general form, dG.dA.dC -dG dT dCi, as inserted sei j k k ii quences in plasmid DNAs. Plasmids containing these inserted block sequences are stable and can be used to obtain an essentially unlimited supply of the block sequences. Insertion of the triple-block polymer into a Sma I site of a plasmid allows excision using Sma I so that the block fragments can be studied in isolation or as part of a larger DNA. The protocol allows minute amounts of a heterogenous mixture of related DNA fragments to be used for the preparation of large amounts of individual but'related triple-block duplexes. The construction of these recombinant plasmids is significant since, despite the homopolymeric nature of the inserted fragments and their inherent potential for adopting stable hairpin structures, the plasmids are stable and amplifiable.

MATERIALS Escherichia coli Strains The following strains were used: C600 SF8 (r-m-, rec BC, kk + lop 11, lig ) (10) and JC1569 (rec A-) (11,12). Plasmids pBR322 (ampicillin and tetracycline resistance) (13) in strain C600 SF8 was obtained from H. Boyer while pACYC189 (chloramphenicol and kanomycin resistance, colicin El immunity) (14) in strain JC1569 was obtained from S. Cohen. Polynucleotides The homopolymer duplex dA ndTn was prepared in a templated reaction with r4icrococcus luteus DNA polymerase as previously described (15) and was characterized by analytical buoyant density centrifugation (16). Single-stranded dAn and dT were nn then prepared by preparative cesium chloride density gradient centrifugation of dAn-dTn duplexes in alkaline solution (17). Enzymes Calf thymus terminal deoxynucleotidyl transferase (terminal

3026

Nucleic Acids Research transferase) was a gift from Dr. Robert Ratliff and had a specific activity of 32,000. Pancreatic deoxyribonuclease was purchased from Worthington Biochemicals. Bacterial alkaline phophatase was purchased from Sigma (Type III-S) and was dialyzed to 10 mM Tris-HCl (pH 8) before use. Polynucleotide kinase was purchased from P-L Biochemicals. The restriction enzymes Eco RI and Hae III were gifts from S.C. Hardies and R.W. Blakesley, respectively; the isolation and properties of the enzyme preparations were described previously (18,19). Sma I was a gift from Drs. J. Slightom and 0. Smithies, and Bam HI was purchased from Bethesda Research Laboratories. Ribonuclease A was obtained from Worthington Biochemicals. The lysozyme preparation was described previously (20). Other Materials The sources of drugs were: ampicillin, Wyeth Laboratories; chloramphenicol, Sigma; kanomycin, Bristol Laboratories, tetracycline, Calbiochem; spectinomycin, Upjohn Company.

METHODS

Preparation of Plasmid DNA Highly purified plasmid DNA was prepared from 300-1000 ml cultures as previously described (19) except that strains containing drug resistant plasmids were grown in the presence of the appropriate antibiotics. Levels of antibiotics used were: ampicillin (100 pg/ml), tetracycline (10 pg/ml), chloramphenicol (25 pg/ml) and kanomycin (25 pg/ml). DNase Digestions of Polynucleotides Digestions of dAn and dTn to yield oligomers of the desired size range were performed at 37°C in small scale reactions (0.1 ml) containing 50 mM Tris-HCl (pH 8.0), 1.0 mM MnCl2, 6.0 mM DNA (concentrations of DNA expressed in terms of nucleotide residues) and 0.13 mg/ml pancreatic deoxyribonuclease. Aliquots of the digests were examined on 20% analytical polyacrylamide gels and the appropriate time of digestion was used in subsequent larger scale reactions (1.1 ml). Restriction Enzyme Digestions The restriction endonuclease reaction conditions used for

3027

Nucleic Acids Research Hae III were described (21). Bam HI and Eco RI were used under the conditions described for Eco RI (21). Terminal Transferase Reactions Polymerization of dC residues onto dA and dT oligomers using terminal transferase were performed at 370C. The reactions contained 200 mM potassium cacodylate (pH 7.2), 1.0 mM CoCl2, 2.5 mM 2-mercaptoethanol, 1.0 mM [ 3H]-dCTP (2.0 PCi), 0.025 mM dA or dT oligomers (oligomer concentrations expressed in terms of molecules) and 600 units/ml terminal transferase. After completion, the reactions were phenol extracted, ether extracted and dialyzed into 10 mM Tris-HICl (pH 8.0), 0.1 mM EDTA. Polymerization of dG residues onto linearized plasmids using terminal transferase were also performed at 370C. The reactions (0.1 ml) contained 50 mM potassium phosphate (pH 7.0), 1.0 mM CoCl 2', 2.5 mM 2-mercaptoethanol, 2.5 piM [ 3H]-dGTP (2.0 pCi), 18.8 pg/ml plasmid DNA and 300 units/ml terminal transferase. Reactions were stopped by addition of an equal volume of 0.5 M EDTA (pH 6.8). Transformations Approximately equimolar quantities of dA-idC4'40 '0 and dT '40-dC'40 oligomers were annealed by mixing in 100 mM Tris-HCl (pH 8.0), 0.1 mM EDTA. These duplexes were then mixed with dG-tailed plasmids in approximately 10:1 ratios of synthetic oligomer duplex molecules to plasmid molecules. Trahsformation of E. coli C600 SF8 with the plasmid-synthetic duplex mixture was performed as described by Hardies et al. (19) with the exception that different drug selections, ampicillin for pBR322 derivatives or chloramphenicol for pACYC189 derivatives, were used instead of the colicin selection employed for the pVIH51 deriva-

tives. Minipreps Screening of recombinant DNIA plasmids for the desired block polymer insertions required restriction enzyme analysis. In some cases, this was accomplished by isolation of plasmid DN4A from small (10 ml), chloramphenicol-amplified bacterial cultures. The procedure used to obtain plasmid DNA from these cultures which could be subsequently digested with restriction enzymes 3028

Nucleic Acids Research was described elsewhere (22).

DNA Sequencing The procedures used for labelling 51-termini of DNA duplexes using polynucleotide kinase and y- 2P-ATP, for isolation of DNA fragments from polyacrylamide gels and for sequencing labelled DNA fragments were those described by Maxam and Gilbert

(23). Other Methods 5% polyacrylamide slab gel electrophoresis was performed as described previously (24). Non-denaturing 20% analytical polyacrylamide tube gel electrophoresis was as described by Burd and Wells (25). 20% polyacrylamide slab gels containing urea, used for DNA sequencing, were as described by Maxam and Gilbert (23).

RESULTS

Scheme for Plasmid Constructions Fig. 1 shows the protocol adopted for insertion of tripleoligoand dTblock DNA sequences into various plasmids. dAx x mers, generated by partial pancreatic deoxyribonuclease digestion of the respective dAn and dTn polymers, are separately "tailed" with dC residues using terminal transferase to give the and dT-dC-. single-stranded block oligonucleotides, dA-dCx x x x These block oligomers are mixed to yield short dA*dT duplexes of various lengths having dC tails of various lengths. The tailed dA-dT duplexes are mixed with linearized plasmids that have been tailed, using terminal transferase, with dG residues. The basepairing of dG and dC tails results in the circularization of some plasmid molecules. Transformation of competent E. coli cells with this DNA mixture results in the isolation of some bacterial clones containing plasmids with the triple-block sequence insertions. Since each colony results from one transformation event, the protocol allows isolation of many different plasmid molecules with different insertions. Each of these insertions will be an individual member of the family of sequences Once a clone is .dG k dT j dC.. having the general form dG.dA.dC j I. k I isolated and characterized, essentially unlimited amounts of the recombinant plasmids, and hence of the triple-block sequences, 3029

Nucleic Acids Research

dTn

dAn

O

PLASMID

DONASE LINEARIZE PLASMID

dTi

dAi

WITH RESTRICTION ENZYME

I TERMINAL TRANSFERASE dCTP

dAiI

ICy

ANNEAL

dTt dC.5TTTT

lllllTT T TERMINAL TRANSFERASE ~~~~~~~~~dGTP

+

dA

dC

dG

dT dC J~~LmJm~~~~mi

dG

TRANSFORM E COLI AND SCREEN FOR PLASMIDS WITH INSERTIONS

k

SdC4

Fig. 1. Scheme for the construction of plasmids containing triple-block sequences.

become available by large-scale growth of the bacterial strain. Preparation of dA- and dT- Oligomer Mixtures ___x

x

Digestions of either dAn or dTn with pancreatic deoxyribonuclease yield distributions of oligomer sizes with an average chain length x (25). By following the kinetics of the reactions using analytical 20% polyacrylamide gels, mixtures of oligomers with a particular value of x can be generated. A gel analysis of a mixture of dT oligomers, used in the cloning studies, is shown in Fig. 2. The absorbance peaks in such a gel scan correspond to oligomers differing in length by one nucleotide residue (25). Due to the logarithmic dependence of migration on oligomer size, it is difficult to precisely determine the average chain length of the dT- oligomer mixture shown in x

3030

Nucleic Acids Research

0

w

a:

E

0 a.

Ct c' H

cr

0 u z

()

-4

0L

z m

C) v

CID

I

l0

0

mn

z

w

C-) w

a-

DISTANCE MIGRATED (cm)

TIME (min)

Fig. 2. Steps in the preparation of dT -4-dCi-d. The left panel shows an analytical polyacrylamide gel ana ysis of a DNase digestion of dTn. Digestion was for 10 min as described in Methods. A portion of the digest (150 nmole) was mixed with 0.025 ml 40% sucrose, 0.1 mg/ml bromphenol blue and the sample (0.050 ml) was layered onto a 12-cm 20% polyacrylamide gel. The gel was subjected to electrophoresis for 3 hr, then sliced 9.5 cm from the top and scanned (25) at 270 nm. Assignment of oligomer sizes was made according to previously determined relative migration rates compared to bromphenol blue (25). The right panel shows the reaction of terminal transferase with dT4 0 and dCTP. The reaction was described in Methods and, at intervals, acid-insoluble radioactivity was determined on 0.005 ml samples from the reaction.

Fig. 2. However, the value of x clearly lies between 30 and 50 and therefore, for convenience, we denote the mixture as dT4-. 40 A similar protocol of digestion and gel analysis provided a dAT40 sample. Terminal Transferase Reactions Fig. 2 (right panel) shows the kinetics of polymerization of dC residues by calf thymus deoxyribonucleotidyl transferase onto the 3'-OH ends of the dT-4 mixture of oligomers. In the reaction shown, complete polymerization of the dCTP substrate 3031

Nucleic Acids Research would result in synthesis of dTv-dC4-0 block oligomers. As indicated, complete polymerization does occur after 40 min. of incubation. This result is similar to other studies involving addition of deoxyribonucleotide residues to oligomer primers (25). When dA4-b oligomers were used as primers in similar reactions, the kinetics and extent of the reactions were idenblock oligomers. tical, resulting in synthesis of dA--dC440 40 Terminal transferase was also used to polymerize dC residues onto the 3'-OH ends of linearized vector DNAs. Polymerization of homopolymeric "tails" onto plasmid DNAs has been used extensively in construction of recombinant DNAs (26). Under the reaction conditions used (see Methods), after 5 min. of incubation the reaction was complete and 30% of the dGTP substrate in the reaction had been polymerized into acid-insoluble polynucleotides. This percentage of incorporation indicated an average of 50 dG residues had been polymerized onto each end of the plasmid molecules present. When dCTP (the same preparation used for "tailing" the oligomer mixtures above) replaced dGTP in these reactions, similar levels of incorporation (X30%) were found. Thus, the low percentage of incorporation upon completion of the reaction was not due to an impure triphosphate preparation or to the self-limiting nature of dGTP incorporation onto oligomer primers seen previously (25). We have not studied this reaction further since the level of incorporation observed was sufficient for the purposes of this investigation. Construction and Clharacterization of Recombinant Plasmids Insertion of Triple-Block Sequences into pBR322 - For our preliminary studies, we chose to use the Bam HI site in pBR322 as the site of insertion since it was uncertain whether insertion of triple-block sequences could give rise to stable recombinant clones. This allowed screening of candidate clones by their ability to grow on tetracycline containing media. After transforrnation with the plasmid-block polymer DNA mixture (see Methods), 50 ampicillin resistant candidate clones were selected and tested for tetracycline resistance. 45 were found to be sensitive to tetracycline; of these, six were chosen for restriction enzyme analysis of the plasmid DNA. In all six cases, digestion of the plasmid DNA with Hae III indicated that DNA 3032

Nucleic Acids Research fragments ranging in size from 40 to 80 bp in length had been inserted into the Bam HI site of pBR322. The analysis of two of these plasmids, pRW701 and pRW702, is shown in Fig. 3e As can be seen, the 104 bp Hae III fragment of pBR322, which contains the Bam HI cleavage site, has been replaced in the recombinant plasmids by a larger fragment. In pRW701, the fragment is 157 bp in length, while in pRW702, it is 184 bp in length, indicating insertions of 53 and 80 bp, respectively. Neither pRW701 nor pRW702 are cleaved by Bam HI (data not shown). The DNA fragments inserted into pRW701 and pRW702 were further characterized by DNA sequencing analysis. Fig. 4 shows that both of these recombinant DNA plasmids contain triple-block inserted sequences; details on the inserted sequences are summarized in Table 1. The sequences flanking the triple-block insertions correspond to the sequences surrounding the Bam HI C'

-

M~)

f-

Qi

CL

rb Q

84

Fig. 3. Hae III restriction analysis of pRW7Ol and pRW702. As described in Methods, approximately 0.002 mg of the plasmid DNAs (as indicated in the figure) were digested with 2 units of Hae III for one hr, subj-ected to electrophoresis on a 5% polyacrylamide gel and visualized by ethidium bromide staining. Sizes of pBR322 Hae III fragments indicated are those reported by Sutcliffe (30). 3033

Nucleic Acids Research

4b

4 - -~ ~~ _6S ~ ~ -

*dI 4 ~ ~~

~

~

~

~

~

~~|.S

Fig. L4. DNIA sequencing analysis of pRW70l and pRW702. For each plasmid, the Taq I fragment containing the insertion was isolated from preparative 5% polyacrylamide gels and labelled using polynucleotide kinase as in Methods. The labelled fragments were then digested with Hae III, and the fragments containing the insertions again-isolated using 8% polyacrylamide gel electrophoresis. DNA sequencing of the isolated fragments was as described by Maxam and Gilbert (23). pRW70l and pRW702 were analyzed on 30 x 40 x 0.4 cm 20% polyacrylamide slab sequencing gels (panels A and B). For these gels, the xylene cyanol FF tracking dye (known to migrate with DNA fragments approximately 28 bases in length) was 30 cm from the top of the gel. Further analysis of pRW702 on a 16 x '40 x 0.1 cm sequencing gel was performed (panel C). In this panel, the DNA fragments were allowed to migrate further into the gel to provide increased resolution of larger fragments. site in pBR322. In pRW701, the sequencing data only allows comparison with pBR322 sequences that lie to the left (that is,

3034

Nucleic Acids Research TABLE 1 Triple-block sequences inserted into

plasmid

Plasmid

Vector

Site of Insertion

Sequence of Insertiona

pRW701

pBR322

Bam HI

(G)

pRW702 pRW712

pBR322

Bam HI

pRW26

Bam HI - Hind III

pRW716

pRW26

Bam HI - Hind III

pRW718

pRW26

Bam HI - Hind III

DNAs.

Insertion Size by Restriction Enzyme Analysis

( (T)21b

80

()27

I)16 (G)20b (T) 30b(S)c

100

(G)30

() 20b(G)15b

63

~cG) 20

T ~A0t 14 ()12

46

C 18

53

G12b6

aSizes

of homopolymer tracts are given when sequence analysis was unambiguous and the error is not more than +1 base residue.

bThe bands were not well-resolved on the sequencing gels or anomalous migration behavior of the fragments was observed (see text); thus, a more conservative error of +5 base residues is estimated in these cases.

cDue to limited sequencing data and the large size of this insertion, only the presence of a CG-tract is indicated here. Length determination was not attempted. have lower residue numbers according to the convention described by Bolivar et al. (13)) of the Bam HI site. From the data, it is apparent that the four residues comprising the protruding 5'-terminus of the Bam HI site have been removed during the transformation process. In pRW702, the sequences to the left of the block insertion are identical to those in pRW701. In addition, due to an anomalous migration behavior of the labelled DNA

fragments on the 20% polyacrylamide gels, it is apparently possible, for this plasmid, to compare the sequences to the right of the insertion with the corresponding sequences in pBR322. Again, in pRW702, the four protruding residues of the Bam HI site were lost on both sides of the inserted tripleblock sequence. We do not know why the longer labelled DNA fragments in the pRW702 sequencing gel exhibit the anomalous migration behavior which allows the sequence to be determined to the right of the insertion, but it may be due to the ability of these fragments to adopt stable hairpin structures. The close agreement between the insertion sizes determined restriction mapping and DNA sequencing indicates that no by additional DNA was inserted. From the sequencing data, it is also apparent that the 3035

Nucleic Acids Research triple-block sequences have inserted in both orientations iLnto the pBR322 vector. In pRW701, the labelled DNA strand contains dT residues while in pRW702, this strand contains dA residues. Insertion of Triple-Block Sequences into pRW26 and pRW28 - The two recombinant plasmids (pRW26 and pRW28) which contain the 95 bp Alu I fragments bearing the E. coli lac wild-type and UVS promoters, respectively, were described previously (27). In these plasmids, the portion of pBR322 between the Bam HI and Sal I sites was replaced with the lac promoter fragments, maintaining both the Bam HI and Sal I sites. These two plasmids were linearized by a double-digestion with Bam HI + Hind III and then tailed with dG residues using terminal transferase. After mixing with the synthetic oligomer duplex and transformation as described above, sixteen candidate clones for each plasmid, obtained using ampicillin selection, were chosen for further analysis. For pRW26, fifteen of the clones gave blue colony positive test results on XG indicator plates, indicating multiple cbpies of the lac operator sequence (19). For pRW28, thirteen clones gave positive results. Restriction endonuclease analysis of these clones by a small-scale DN4A isolation procedure described previously (22) indicated that all the colonies giving positive tests on XG plates contained small (40-110 bp) insertions next to the lac promoter sequence while the colonies giving negative tests had deletions of the original plasmid vehicle. Fig. 5 shows Hae III restriction analysis of the plasmid DNAs isolated (using standard procedures) from eight of the pRW26 candidate clones. Since digestion of pRW26 and pRW28 with Bam HI + Hind III removes a fragment of DNA from the vectors, there is no control digest in Fig. 5 for comparison to the recombinant DNA digestion patterns. However, knowledge of the Hae III restriction maps of pRW26 and pRW28 allows calculation of the Hae III fragment size that would result if the plasmids recircularize after digestion with Bam HI + Hind III (arrow in Fig. 5). As can be seen, the digestion patterns of the isolated recombinant plasmids show no Hae III digestion bands with this size. Instead, each has a more slowly migrating band, indicating small insertions into the plasmid vehicle. 3036

Nucleic Acids Research

-CIL

a.

0~ w Cr w F)( 0ri a Q a a Q

w

a

CO x

a

58 -

4 34 = 267i 84 -_

89

Fig. 5, Hae III restriction analysis of pRW26 derivatives containing insertions next to the lac promoter sequence. Conditions for the analyses are the same as described in Fig. 3. The meaning of the arrow is described in Results. The sizes of insertions are: pRW711, 96 bp; pRW712, 100 bp; pRW713, 110 bp; pRW714, 71 bp; pRW715, 66 bp; pRW716, 63 bp; pRW717, 56 bp and pRW718, 46 bp. DNA sequence analyses of pRW712, pRW716 and pRW718 indicated triple-block sequence insertions similar to those in pRW701 and pRW702 and are summarized in Table 1. In the cases where the data was available, the protruding ends of the restriction sites had been lost during the transformation process.

DISCUSSION The results presented indicate that DNA recombinant techniques may be used to obtain DNA fragments having novel base sequences from synthetically prepared heterogenous samples. Despite the highly repeated base sequences of the block polymers described, and the potential for stable hairpin structures inherent in these sequences, stable, amplifiable plasmids containing the triple-block sequences could be constructed. It is not apparent why difficulties encountered by workers in this 3037

Nucleic Acids Research (19) and other laboratories (28,29) when cloning either repeated-sequence or palindromic DNA fragments were not observed in this case. Perhaps the presence of as few as 10-20 bp intervening between the inverted repeats (which may be unpaired in the cruciform structure) is crucial. The plasmids containing triple block polymer insertions and having pBR322 as the parental vehicle were stable, as judged by restriction enzyme analysis, after at least four subclonings of the host bacteria and after secondary transformation into E. coli strain MO (data not shown). Such stability was not seen in similar experiments with a different plasmid vehicle, pACYC189, containing block inserts (see below). Apparently, the choice of a host-vector system for insertion of unusual sequences into plasmid DNAs may be critical. We have inserted triple-block sequences into pBR322 and two pBR322 derivatives, pRW26 and pRW28. We have also, using the same protocol, inserted fragments into pACYC189, using the Sma I side present in the kanomycin-resistance element of this plasmid. While difficulties with the stability of these plasmids (later found also to be present in the parental vehicle itself) prevented DNA sequencing analysis of the fragments inserted into the Sma I site, the insertions could be excised from the vehicle using Sma I (data not shown). This indicates that the triple-block sequences can be studied when imbedded in plasmid DNA or as isolated fragments. Such studies will further our knowledge of long-range interactions in DNA and assist in understanding the roles of these effects in gene expression. In addition, the ability to perform such studies will no longer depend on the time-consuming and tedious enzymatic and chemical approaches to prepare such molecules.

ACKNOWLEDGEMENTS

This work was supported by grants from the National Institutes of Health (CA 20279) and the National Science Foundation (PCM 77-15033). E.S. was supported during a portion of these studies by a postdoctoral fellowship from the National Institutes of Health (GM 05468). 3038

Nucleic Acids Research ABBREVIATIONS The nucleic acid subscript n designates a long chain polymer (longer than 100 nucleotide residues) where the molecules are heterogenous in length and the subscripts i,j,k and m designate oligomers (shorter than 100 residues) where the length is well-defined. The subscript x indicates a mixture of different sized oligomers. Present address: Department of Microbiology and Immunology, SC-42, University of Washington, Seattle, Washington 98195.

REFERENCES 1.

2. 3.

4.

5.

6. 7. 8.

9.

10. 11. 12. 13. 14.

Wells, R.D., Blakesley, R.W., Burd, J,F., Chan, H.W., Dodgson, J.B., Hardies, S.C., Horn, G.T., Jensen, K.F., Larson, J.E., Nes, I.F., Selsing, E. and Wartell, R.M., (1977) Critical Reviews in Biochemistry 4, 305-340. Selsing, E., Wells, R.D., Early, T.A. and Kearns, D.R. (1978) Nature 275, 249-250. Patient, R.K., Hardies, S.C., Larson, J.E,, Inman, R.B., Maquat, L.E. and Wells, R.D., (1979) J. Biol. Chem. in press. Selsing, E. and Wells, R.D., (1979) J. Biol. Chem. in press. Selsing, E., Wells, R.D., Alden, C.J. and Arnott, S. (1979) J. Biol. Chem. in press. Burd, J.F., Wartell, R.M., Dodgson, J.B. and Wells, R.D., (1975) J. Biol. Chem. 250, 5109-5113. Burd, J.F., Larson, J.E. and Wells, R.D. (1975) J. Biol. Chem. 250, 6002-6007. Early, T.A., Kearns, D.R., Burd, J.F., Larson, J.E. and Wells, R.D. (1977) Biochemistry 16, 541-551. Wartell, R.M. and Burd, J.F. (1976) Biopolymers 15, 14611480. Struhl, K., Cameron, J.R. and Davis, R.W. (1976) Proc. Natl. Acad. Sci. U.S.A. 73, 1471-1475. Chang, A.C.Y. and Cohen, S.N. (1978) J. Bacteriol. 134, 1141-1156. Clark, A.J. and Chamberlin, M. (1966) J. Mol. Biol. 19, 442-454.

Bolivar, F., Rodriguez, R.L., Greene, P.J., Betlach, M.C., Heyneker, H.L., Boyer, H.W., Crosa, J.H. and Falkow, S. (1977) Gene 2, 95-113. Chang, S. and Cohen, S.N. (1977) Proc. Natl. Acad, Sci. U.S.A. 74, 4811-4815. 3039

Nucleic Acids Research 15. 16.

17. 18.

19. 20.

21. 22. 23. 24.

Wells, R.D., Larson, J.E., Grant, R.C., Shortle, B.E. and Cantor, C.R. (1970) J. Mol. Biol. 54, 465-497. Wells, R.D. and Larson, J.E. (1972) J. Biol. Chem. 247, 3405-3409. Wells, R.D. and Blair, J.E. (1967) J. Mol. Biol. 27, 273-288. Blakesley, R.W., Dodgson, J.B., Nes, I.F. and Wells, R.D. (1977) J. Biol. Chem. 252, 7300-7306. Hardies, S.C., Patient, R.K., Klein, R.D., Ho, F., Reznikoff, W.S. and Wells, R.D, (1979) J. Biol. Chem. in press. Harwood, S.J., Schendel, P.F. and Wells, R.D. (1970) J. Biol. Chem. 245, 5614-5624. Hardies, S.C. and Wells, R.D. (1976) Proc. Natl. Acad, Sci. U.S.A. 73) 3117-3121. Klein, R.D., Selsing, E. and Wells, R.D. (1979) Plasmid submitted for publication. Maxam, A.M. and Gilbert, W. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 560-564. Blakesley, R.W. and Wells, R.D. (1975) Nature 257, 421-

1422.

25.

Burd, J.F. and Wells, R.D. (1974) J. Biol, Chem. 249,

26. 27.

70914-7101. Sinsheimer, R.L. (1977) Ann. Rev. Biochem. 46, 415-438. Klein, R.D., Selsing, E. and Wells, R.D. (1979) in press

28.

29. 30.

3040

for XIth International Congress of Biochemistry. Sadler, J.R., Betz, J.L., Techlenberg, M., Goeddel, D.V., Yansura, D.G. and Caruthers, N.H. (1979) Gene, in press. Bolivar, F., Betlach, M.C., Heyneker, H.L., Shine, J., Rodriguez, R.L. and Boyer, H.W. (1977) Proc. Niatl. Acad. Sci. U.S.A. 74, 5265-5269. Sutcliffe, J.G. (1978) Niucleic Acids. Res. 5, 2721-2728.