Combinatorial determination of sequence specificity for nanomolar DNA-binding hairpin polyamides

Development of sequence-specific DNA-binding drugs is an important pharmacological goal, given the fact that numerous existing DNA-directed chemotherapeutic drugs rely on the strength and selectivity of their DNA interactions for therapeutic activity. Among the DNA-binding antibiotics, hairpin polyamides represent the only class of small molecules that can practically bind any predetermined DNA sequence. DNA recognition by these ligands depends on their side-by-side amino acid pairings in the DNA minor groove. Extensive studies have revealed that these molecules show extremely high affinity for sequence-directed, minor groove interaction. However, the specificity of such interactions in the presence of a large selection of sequences such as the human genome is not known. We used the combinatorial selection method restriction endonuclease protection, selection, and amplification (REPSA) to determine the DNA binding specificity of two hairpin polyamides, ImPyPyPy-gamma-PyPyPyPy-beta-Dp and ImPyPyPy-gamma-ImPyPyPy-beta-Dp, in the presence of more than 134 million different sequences. These were verified by restriction endonuclease protection assays and DNase I footprinting analysis. Our data showed that both hairpin polyamides preferentially selected DNA sequences having consensus recognition sites as defined by the Dervan pairing rules. These consensus sequences were rather degenerate, as expected, given that the stacked pyrrole-pyrrole amino acid pairs present in both polyamides are unable to discriminate between A.T and T.A base pairs. However, no individual sequence within these degenerate consensus sequences was preferentially selected by REPSA, indicating that these hairpin polyamides are truly consensus-specific DNA-binding ligands. We also discovered a preference for overlapping consensus binding sites among the sequences selected by the hairpin polyamide ImPyPyPy-gamma-PyPyPyPy-beta-Dp, and confirmed by DNase I footprinting that these complex sites provide higher binding affinity. These data suggest that multiple hairpin polyamides can cooperatively bind to their highest-affinity sites.     

Combinatorial identification of a novel consensus sequence for the covalent DNA-binding polyamide tallimustine

Many agents successfully used in cancer chemotherapy either directly or indirectly covalently modify DNA. Examples include cisplatin, which forms a covalent adduct with guanines, and doxorubicin, which traps a cleavage intermediate between topoisomerase II and torsionally strained DNA. In most cases, the efficacy of these drugs depends on the efficiency and specificity of their DNA binding, as well as the discrimination between normal and neoplastic cells in their handling of the drug-DNA adducts. While much is known about the chemistry of drug-DNA adducts, little is known regarding the overall specificity of their formation, especially in the context of a whole human genome, where potentially billions of binding sites are possible. We used the combinatorial selection method restriction endonuclease protection, selection, and amplification (REPSA) to determine the DNA-binding specificity of the semisynthetic covalent DNA-binding polyamide tallimustine, which contains a benzoic acid nitrogen mustard appended to the minor groove DNA-binding natural product distamycin A. After investigating over 134 million possible sequences, we found that the highest affinity tallimustine binding sites contained one of two consensus sequences, either the expected distamycin hexamer binding sites followed by a CG base pair (e.g., 5'-TTTTTTC-3' and 5'-AAATTTC-3') or the unexpected sequence 5'-TAGAAC-3'. Curiously, we found that tallimustine preferentially alkylated the N7 position of guanines located on the periphery of these consensus sequences. These findings suggested a cooperative binding model for tallimustine in which one molecule noncovalently resides in the DNA minor groove and locally perturbs the DNA structure, thereby facilitating alkylation by a second tallimustine of an exposed guanine on another side of the DNA.     

Cloning of endogenous murine leukemia virus-related sequences from chromosomal DNA of BALB/c and AKR/J mice: identification of an env progenitor of AKR-247 mink cell focus-forming proviral DNA

Recombinant phages containing murine leukemia virus (MuLV)-reactive DNA sequences were isolated after screening of a BALB/c mouse embryo DNA library and from shotgun cloning of EcoRI-restricted AKR/J mouse liver DNA. Twelve different clones were isolated which contained incomplete MuLV proviral DNA sequences extending various distances from either the 5' or 3' long terminal repeat (LTR) into the viral genome. Restriction maps indicated that the endogenous MuLV DNAs were related to xenotropic MuLVs, but they shared several unique restriction sites among themselves which were not present in known MuLV proviral DNAs. Analyses of internal restriction fragments of the endogenous LTRs suggested the existence of at least two size classes, both of which were larger than the LTRs of known ecotropic, xenotropic, or mink cell focus-forming (MCF) MuLV proviruses. Five of the six cloned endogenous MuLV proviral DNAs which contained envelope (env) DNA sequences annealed to a xenotropic MuLV env-specific DNA probe; in addition, four of these five also hybridized to an ecotropic MuLV-specific env DNA probe. Cloned MCF 247 proviral DNA also contained such dual-reactive env sequences. One of the dual-reactive cloned endogenous MuLV DNAs contained an env region that was indistinguishable by AluI and HpaII digestion from the analogous segment in MCF 247 proviral DNA and may therefore represent a progenitor for the env gene of this recombinant MuLV. In addition, the endogenous MuLV DNAs were highly related by AluI cleavage to the Moloney MuLV provirus in the gag and pol regions.     

Characterization of proviruses cloned from mink cell focus-forming virus-infected cellular DNA

Two proviruses were cloned from EcoRI-digested DNA extracted from mink cells chronically infected with AKR mink cell focus-forming (MCF) 247 murine leukemia virus (MuLV), using a lambda phage host vector system. One cloned MuLV DNA fragment (designated MCF 1) contained sequences extending 6.8 kilobases from an EcoRI restriction site in the 5' long terminal repeat (LTR) to an EcoRI site located in the envelope (env) region and was indistinguishable by restriction endonuclease mapping for 5.1 kilobases (except for the EcoRI site in the LTR) from the 5' end of AKR ecotropic proviral DNA. The DNA segment extending from 5.1 to 6.8 kilobases contained several restriction sites that were not present in the AKR ecotropic provirus. A 0.5-kilobase DNA segment located at the 3' end of MCF 1 DNA contained sequences which hybridized to a xenotropic env-specific DNA probe but not to labeled ecotropic env-specific DNA. This dual character of MCF 1 proviral DNA was also confirmed by analyzing heteroduplex molecules by electron microscopy. The second cloned proviral DNA (designated MCF 2) was a 6.9-kilobase EcoRI DNA fragment which contained LTR sequences at each end and a 2.0-kilobase deletion encompassing most of the env region. The MCF 2 proviral DNA proved to be a useful reagent for detecting LTRs electron microscopically due to the presence of nonoverlapping, terminally located LTR sequences which effected its circularization with DNAs containing homologous LTR sequences. Nucleotide sequence analysis demonstrated the presence of a 104-base-pair direct repeat in the LTR of MCF 2 DNA. In contrast, only a single copy of the reiterated component of the direct repeat was present in MCF 1 DNA.     

Molecular cloning of unintegrated and a portion of integrated moloney murine leukemia viral DNA in bacteriophage lambda.

A covalently closed circular form of unintegrated viral DNA obtained from NIH 3T3 cells freshly infected with Moloney murine leukemia virus (M-MLV) and a port of the endogenous M-MLV from the BALB/Mo mouse strain have been cloned in bacteriophage lambda. The unintegrated viral DNA was cleaved with restriction endonuclease HindIII and inserted into the single HindIII site of lambda phage Charon 21A. Similarly high-molecular-weight DNA from BALB/Mo mice ws cleaved sequentially with restriction endonucleases EcoRI and HindIII and separated on the basis of size, and one of the two fractions which reacted with an M-MLV-specific complementary DNA was inserted into the HindIII site of Charon 21A. Recombinant clones containing M-MLV-reacting DNA were analyzed by restriction endonuclease mapping, heteroduplexing, and infectivity assays. The restriction endonuclease map of the insert derived from unintegrated viral DNA, lambda x MLV-1, was comparable to published maps. Electron microscope analysis of the hybrid formed between lambda x MLV-1 DNA and 35S genomic M-MLV RNA showed a duplex structure. The molecularly cloned lambda x MLV-1 DNA contained only one copy of the long terminal repeat and was not infectious even after end-to-end ligation of the insert DNA. The insert DNA derived from endogenous M-MLV, lambda x MLVint-1, contained a DNA stretch measuring 5.4 kilobase pairs in length, corresponding to the 5' part of the genomic viral RNA, and cellular mouse DNA sequences measuring 3.5 kilobase pairs in length. The viral part of the insert showed the typical restriction pattern of M-MLV DNA except that a single restriction site, PvuII, in the 5' long terminal repeat was missing. Reconstructed genomes containing the 5' half derived from the integrated viral DNA and the 3' half derived from the unintegrated viral DNA were able to induce XC plaques after transfection in uninfected mouse fibroblasts.     

Methylated DNA-binding protein from human placenta recognizes specific methylated sites on several prokaryotic DNAs.

Methylated DNA-binding protein (MDBP) from human placenta recognizes specific DNA sequences containing 5-methylcytosine (m5C) residues. Comparisons of binding of various prokaryotic DNAs to MDBP indicate that m5CpG is present in the recognition sites for this protein but is only part of the recognition sequence. Specific binding to MDBP was observed for bacteriophage XP12 DNA, which naturally contains approximately 1/3 of its residues as m5C, and for Micrococcus luteus DNA, M13mp8 replicative form (RF) DNA, and pBR322 when these three DNAs were methylated at CpG sites by human DNA methyltransferase. Five DNA regions binding to MDBP have been localized by DNase I footprinting or restriction mapping in methylated pBR322 and M13mp8 RF DNAs. A comparison of their sequences reveals a common 5'-m5CGRm5CG-3' element or closely related sequence in which one of the m5C residues may be replaced by a T. In addition to this motif, one upstream and one downstream m5CpG as well as other common residues over an approximately 20-bp long region may be recognized by MDBP.     

Duplicated region of the mouse genome containing a cytoplasmic gamma-actin processed pseudogene associated with long interspersed repetitive elements

The structures of two cloned recombinants of bacteriophage lambda and mouse genomic DNA (lambda mA14 and lambda mA36) were compared by electron microscopic analysis of various heteroduplex DNAs, restriction endonuclease mapping and nucleotide sequence determination. Each clone was shown to be derived from a distinct region of the mouse genome, but the two exhibited structural similarity over a region of at least 11,000 bases which included a cytoskeletal gamma-actin processed pseudogene of approximately 1800 bases. It is concluded that the two genomic regions were derived from a common ancestral region by duplication or amplification. The homologous regions of the two clones contained members of the long interspersed repetitive L1Md (long interspersed repeated sequence 1 of Mus domesticus) family lying in opposite orientation to one another, so that single-stranded DNA from the clones could form intra-molecular heteroduplexes. The complete nucleotide sequences of three L1Md members in lambda mA14 were determined. The longest of these (L1Md-14LH) had inserted into the gamma-actin processed pseudogene and, although it contained internal deletions, appeared to possess intact 5' and 3' ends. A second L1Md member (L1Md-14RH1) also appeared to have an intact 5' end but had lost most of its 3' portion, and a third member (L1Md-14RH2) was an internal fragment. The repeated sequence at the 5' ends of L1Md-14LH and L1Md-14RH1 showed these to be members of the L1Md-A family.     

Avian acute leukemia virus OK 10: analysis of its myc oncogene by molecular cloning.

Several DNAs representing the genome of the avian acute leukemia virus OK 10 were isolated by molecular cloning from a transformed quail cell line, 9C, which contained at least six OK 10 proviruses. Recombinant lambda phages harboring the OK 10 genome and additional flanking cellular DNA sequences were studied by restriction endonuclease mapping and hybridization to viral cDNA probes. Six of the clones represented complete proviruses with similar, if not identical, viral sequences integrated at different positions in the host DNA. The organization of the OK 10 genome was determined by electron-microscopic analysis of heteroduplexes formed between the cloned OK 10 DNA and DNAs representing the c-myc gene and the genomes of two other avian retroviruses, Rous-associated virus-1 and MC29. The results indicated that the OK 10 proviral DNA is about 7.5 kilobases in size with the following structure: 5'-LTR-gag-delta polmyc-delta env-LTR-3', where LTR indicates a long terminal repeat. The oncogene of OK 10, v-mycOK 10, forms a continuous DNA segment of around 1.7 kilobases between pol and env. It is similar in structure and length to the v-myc gene of MC29, as demonstrated by restriction endonuclease and heteroduplex analyses. Two of the OK 10 proviruses were tested in transfection experiments: both DNAs gave rise to virus with the transforming capacities of OK 10 when Rous-associated virus-1 was used to provide helper virus functions.     

Titration of integrated simian virus 40 DNA sequences, using highly radioactive, single-stranded DNA probes.

Nick-translated simian virus 40 (SV40) [32P]DNA fragments (greater than 2 X 10(8) cpm/micrograms) were resolved into early- and late-strand nucleic acid sequences by hybridization with asymmetric SV40 complementary RNA. Both single-stranded DNA fractions contained less than 0.5% self-complementary sequences; both included [32P]-DNA sequences that derived from all regions of the SV40 genome. In contrast to asymmetric SV40 complementary RNA, both single-stranded [32P]DNAs annealed to viral [3H]DNA at a rate characteristic of SV40 DNA reassociation. Kinetics of reassociation between the single-stranded [32P]DNAs indicated that the two fractions contain greater than 90% of the total nucleotide sequences comprising the SV40 genome. These preparations were used as hybridization probes to detect small amounts of viral DNA integrated into the chromosomes of Chinese hamster cells transformed by SV40. Under the conditions used for hybridization titrations in solution (i.e., 10- to 50-fold excess of radioactive probe), as little as 1 pg of integrated SV40 DNA sequence was assayed quantitatively. Among the transformed cells analyzed, three clones contained approximately one viral genome equivalent of SV40 DNA per diploid cell DNA complement; three other clones contained between 1.2 and 1.6 viral genome equivalents of SV40 DNA; and one clone contained somewhat more than two viral genome equivalents of SV40 DNA. Preliminary restriction endonuclease maps of the integrated SV40 DNAs indicated that four clones contained viral DNA sequences located at a single, clone-specific chromosomal site. In three clones, the SV40 DNA sequences were located at two distinct chromosomal sites.     

Some restriction endonucleases tolerate single mismatches of the pyrimidine.purine type.

DNAs from phage mutants M13mp18 and M13mp18/MP-1 were used to construct two closed circular heteroduplexes. One of them carried the sequence 5'-CCTGGG-3' 3'-GGGCCC-5' with a T.G mismatch at the position 6248. The other carried the sequence 5'-CCCGGG-3' 3'-GGACCC-5' with a C.A mismatch at the same position. Heteroduplexes were exposed to 7 restriction endonucleases having recognition sites within the sequence 5'-CCCGGG-3' 3'-GGGCCC-5' and to 1 restriction endonuclease having a recognition site within the sequence 5'-CCTGGG-3' 3'-GGACCC-5'. All tested enzymes cleaved at least one mismatch-containing sequence although with reduced efficiency. Smal and Xmal tolerated both mismatch-containing sequences. Aval, Hpall, Mspl, Ncil and Nsplll were able to tolerate only the T.G containing sequence, while BstNl was able to tolerate only the C.A containing sequence. It is inferred that the tolerance displayed by Smal and Xmal depends on the presence of either the original purines or the original pyrimidines in mismatches of both the T.G and C.A type and that all other tested enzymes require the presence of the original purines in mismaches of both types.     

Isolation and computer-aided characterization of MmeI, a type II restriction endonuclease from Methylophilus methylotrophus.

A Type II restriction endonuclease, MmeI, has been purified from the obligate methylotroph, Methylophilus methylotrophus. The enzyme was shown to have the non-palindromic recognition sequence 5'-T C C Pu A C (N)20-3', 3'-A G G Py T G (N)18-5' and to cleave (as indicated) on the 3' side, generating a two nucleotide 3' projection. Determination of the recognition sequence was achieved using two new computer programs; RECOG, which predicts recognition sequences from the pattern of restriction fragments obtained from DNAs of known sequence, and GELSIM, which generates graphical simulations of DNA band patterns obtained by gel electrophoresis of restriction digests of sequenced DNA molecules.     

Long terminal repeat of murine retroviral DNAs: sequence analysis, host-proviral junctions, and preintegration site.

The nucleotide sequence of the long terminal repeat (LTR) of three murine retroviral DNAs has been determined. The data indicate that the U5 region (sequences originating from the 5' end of the genome) of various LTRs is more conserved than the U3 region (sequences from the 3' end of the genome). The location and sequence of the control elements such as the 5' cap, "TATA-like" sequences, "CCAAT-box," and presumptive polyadenylic acid addition signal AATAAA in the various LTRs are nearly identical. Some murine retroviral DNAs contain a duplication of sequences within the LTR ranging in size from 58 to 100 base pairs. A variant of molecularly cloned Moloney murine sarcoma virus DNA in which one of the two LTRs integrated into the viral DNA was also analyzed. A 4-base-pair duplication was generated at the site of integration of LTR in the viral DNA. The host-viral junction of two molecularly cloned AKR-murine leukemia virus DNAs (clones 623 and 614) was determined. In the case of AKR-623 DNA, a 3- or 4-base-pair direct repeat of cellular sequences flanking the viral DNA was observed. However, AKR-614 DNA contained a 5-base-pair repeat of cellular sequences. The nucleotide sequence of the preintegration site of AKR-623 DNA revealed that the cellular sequences duplicated during integration are present only once. Finally, a striking homology between the sequences flanking the preintegration site and viral LTRs was observed.     

Site-specific DNA binding of nuclear factor I: analyses of cellular binding sites.

Nuclear factor I is a cellular site-specific DNA-binding protein required for the efficient in vitro replication of adenovirus DNA. We have characterized human DNA sequences to which nuclear factor I binds. Three nuclear factor I binding sites (FIB sites), isolated from HeLa cell DNA, each contain the sequence TGG(N)6-7GCCAA. Comparison with other known and putative FIB sites suggests that this sequence is important for the binding of nuclear factor I. Nuclear factor I protects a 25- to 30-base-pair region surrounding this sequence from digestion by DNase I. Methylation protection studies suggest that nuclear factor I interacts with guanine residues within the TGG(N)6-7GCCAA consensus sequence. One binding site (FIB-2) contained a restriction endonuclease HaeIII cleavage site (GGCC) at the 5' end of the GCCAA motif. Digestion of FIB-2 with HaeIII abolished the binding of nuclear factor I. Southern blot analyses indicate that the cellular FIB sites described here are present within single-copy DNA in the HeLa cell genome.     

Identification of the avian myeloblastosis virus genome. II. Restriction endonuclease analysis of DNA from lambda proviral recombinants and leukemic myeoblast clones.

Two lambda proviral DNA recombinants were characterized with a number of restriction endonucleases. One recombinant contained a complete presumptive avian myeloblastosis virus (AMV) provirus flanked by cellular sequences on either side, and the second recombinant contained 85% of a myeloblastosis-associated virus type 1 (MAV-1)-like provirus with cellular sequences adjacent to the 5' end of the provirus. Comparing the restriction maps for the proviral DNAs contained in each lambda hybrid showed that the putative AMV and MAV-1-like genomes shared identical enzyme sites for 3.6 megadaltons beginning at the 5' termini of the proviruses with respect to viral RNA. Two enzyme sites near the 3'-end of the MAV-1-like provirus were not present in the putative AMV genome. We also examined a number of leukemic myeloblast clones for proviral content and cell-provirus integration sites. The presumptive AMV provirus was present in all the leukemic myeloblast clones regardless of the endogenous proviral content of the target cells or the AMV pseudotype used for conversion. Multiple cellular sites were suitable for integration of the putative AMV genome and the helper genomes. The proviral genomes were all integrated colinearly with respect to linear viral DNA.     

Hemoglobin switching in sheep. Synthesis, cloning, and characterization of DNA sequences coding for the beta B, beta C, and gamma-globin mRNAs.

Synthetic double-stranded DNAs (sDNAs) were prepared from sheep globin mRNA templates isolated from reticulocytes producing either hemoglobin B (HbB) (alpha 2 beta B2), HbC (alpha 2 beta C2), or HbF (alpha 2 gamma 2). These DNAs were inserted into the Eco RI site of plasmid pMB9 by the homopolymer tailing method and used to transform Escherichia coli X1776 to tetracycline resistance. Recombinant clones were identified by colony hybridization and further characterized by molecular hybridization and restriction endonuclease analysis. All plasmids analyzed thus far contained either beta- or gamma-globin DNA sequences. Moreover, sDNAs used for cloning yielded restriction endonuclease fragments consistent with the presence of predominantly beta- or gamma-sDNA, indicating that formation of double-stranded alpha-sDNA proceeds much less efficiently under our conditions than the formation of non-alpha-sDNAs. Three recombinant plasmids, pS beta B2, pS beta C69, and pS gamma 56, were selected for detailed study. These were shown to contain, respectively, beta B-, beta C-, and gamma-DNA sequences by molecular hybridization and by protection of the appropriate cDNAs from S1 nuclease digestion. Each contained all of the restriction endonuclease sites defined for the synthetic sDNAs and protected at least 90% of the sequence length of homologous cDNA. Restriction endonuclease maps of the beta B- and beta C-globin genes were identical at all 12 sites that were mapped, whereas four differences were identified in the gamma gene compared to the two others; three of these corresponded to differences in amino acid sequence of the globins. A method was developed to isolate the anti-mRNA strand of the insert for use as a specific molecular hybridization probe analogous to complementary DNA.     

Isolation and characterization of genomic clones for two chicken phenobarbital-inducible cytochrome P-450 genes

A cDNA clone specific for a chicken phenobarbital-inducible cytochrome P-450 was used to screen a chicken genomic library. Twenty-nine clones were isolated, restriction mapped, and divided into two non-overlapping groups. The cDNA clone hybridized to 12 kilobases of DNA from both groups. Both groups contained restriction fragments which hybridized to both 5' and 3' fragments of the cDNA clone, and it was concluded that the two groups were derived from two separate genes. Southern transfer analysis of individual chicken DNAs and quantitative hybridization analysis indicated that these two genes are independent and are present as single copies/haploid genome. Comparison of restriction digests of the cloned DNAs and total genomic DNA discounted the possibility that other closely related P-450 genes are present in the chicken genome.     

Intracisternal A-particle genes in Mus musculus: a conserved family of retrovirus-like elements.

The structural organization of intracisternal A-particle genes has been studied, using isolates from a mouse gene library in lambda phage Charon 4A. The predominant gene form among the isolates was 7.3 kilobases (kb) in length. R-loops between the 7-kb (35S) A-particle genomic ribonucleic acid and several of these genes were colinear, with no visible evidence of intervening deoxyribonucleic acid sequences. One recombinant was found with an A-particle gene that contained a 1.7-kb deletion. Using the deletion as a reference, the deoxyribonucleic acid and ribonucleic acid homology regions were localized with respect to one another and to the restriction map: the 5' terminus of the ribonucleic acid was several hundred base pairs within the 5' end of the deoxyribonucleic acid homology region. Restriction endonuclease fragments encompassing the 5' and 3' regions of one 7.3-kb gene were separately subcloned into pBR322. Heteroduplexes between the two subclones revealed an approximately 300-base pair segment of terminally redundant sequences. The cloned 3' fragment hybridized with restriction fragments from the 5' end of several other A-particle genes, demonstrating the presence of common (though not necessarily identical) terminally repeated sequences. A-particle genes varied in the occurrence of specific restriction sites at characteristic internal loci. However, heteroduplexes between several variant 7.3-kb genes showed continuous homology regions even when spread under stringent hybridization conditions. The relative abundance of restriction site variants was highly conserved in 12 laboratory strains of Mus musculus, in embryonic and adult tissues of a single inbred strain, and in the SC-1 cell line of feral mouse origin, but appeared to differ in a feral Japanese substrain, Mus musculus molossinus. Some evidence suggests that subsets of A-particle genes may have similar flanking sequences. The results are discussed in terms of the evolution of this multigene family.     

Cloning of random-sequence oligodeoxynucleotides

Methods are described for cloning random or highly degenerate nucleotide (nt) sequences. The procedures use synthetically derived mixtures of oligodeoxynucleotides (oligos) whose heterogeneous central portions are bounded at their 5' and 3' ends by sequences recognized by restriction endonucleases. Oligo collections of defined length and nt composition are synthesized by utilizing appropriate concentrations of all four nucleotide precursors during each addition step for the central region. Single-stranded oligos with appropriate 5' and 3' ends can be ligated directly, although inefficiently, into double-stranded (ds) DNA molecules with complementary 5' and 3' extensions produced by restriction endonuclease cleavage. A more general and efficient method is to convert the oligo into a ds form by incubating it with the Klenow (large) fragment of Escherichia coli DNA polymerase I. If the 3' ends are palindromic, two oligo molecules will serve as mutual primers for polymerization. The resulting products are ds molecules containing two oligo units separated by the original 3' restriction site and bounded at each end by the original 5' restriction site. After appropriate restriction endonuclease cleavage, oligo units can be cloned by standard procedures. Analysis of 26 recombinant M13 phages indicates that the nt sequences of the cloned oligos are in good accord with what was expected on a random basis.