SAMase of Bacteriophage T3 Inactivates Escherichia coli's Methionine S-Adenosyltransferase by Forming Heteropolymers.

S-Adenosylmethionine lyase (SAMase) of bacteriophage T3 degrades the intracellular SAM pools of the host Escherichia coli cells, thereby inactivating a crucial metabolite involved in a plethora of cellular functions, including DNA methylation. SAMase is the first viral protein expressed upon infection, and its activity prevents methylation of the T3 genome. Maintenance of the phage genome in a fully unmethylated state has a profound effect on the infection strategy. It allows T3 to shift from a lytic infection under normal growth conditions to a transient lysogenic infection under glucose starvation. Using single-particle cryoelectron microscopy (cryo-EM) and biochemical assays, we demonstrate that SAMase performs its function by not only degrading SAM but also by interacting with and efficiently inhibiting the host's methionine S-adenosyltransferase (MAT), the enzyme that produces SAM. Specifically, SAMase triggers open-ended head-to-tail assembly of E. coli MAT into an unusual linear filamentous structure in which adjacent MAT tetramers are joined by two SAMase dimers. Molecular dynamics simulations together with normal mode analyses suggest that the entrapment of MAT tetramers within filaments leads to an allosteric inhibition of MAT activity due to a shift to low-frequency, high-amplitude active-site-deforming modes. The amplification of uncorrelated motions between active-site residues weakens MAT's substrate binding affinity, providing a possible explanation for the observed loss of function. We propose that the dual function of SAMase as an enzyme that degrades SAM and as an inhibitor of MAT activity has emerged to achieve an efficient depletion of the intracellular SAM pools. IMPORTANCE Self-assembly of enzymes into filamentous structures in response to specific metabolic cues has recently emerged as a widespread strategy of metabolic regulation. In many instances, filamentation of metabolic enzymes occurs in response to starvation and leads to functional inactivation. Here, we report that bacteriophage T3 modulates the metabolism of the host E. coli cells by recruiting a similar strategy: silencing a central metabolic enzyme by subjecting it to phage-mediated polymerization. This observation points to an intriguing possibility that virus-induced polymerization of the host metabolic enzymes is a common mechanism implemented by viruses to metabolically reprogram and subdue infected cells.

Viperin is an iron-sulfur protein that inhibits genome synthesis of tick-borne encephalitis virus via radical SAM domain activity.

Viperin is an interferon-induced protein with a broad antiviral activity. This evolutionary conserved protein contains a radical S-adenosyl-l-methionine (SAM) domain which has been shown in vitro to hold a [4Fe-4S] cluster. We identified tick-borne encephalitis virus (TBEV) as a novel target for which human viperin inhibits productionof the viral genome RNA. Wt viperin was found to require ER localization for full antiviral activity and to interact with the cytosolic Fe/S protein assembly factor CIAO1. Radiolabelling in vivo revealed incorporation of (55) Fe, indicative for the presence of an Fe-S cluster. Mutation of the cysteine residues ligating the Fe-S cluster in the central radical SAM domain entirely abolished both antiviral activity and incorporation of (55) Fe. Mutants lacking the extreme C-terminal W361 did not interact with CIAO1, were not matured, and were antivirally inactive. Moreover, intracellular removal of SAM by ectopic expression of the bacteriophage T3 SAMase abolished antiviral activity. Collectively, our data suggest that viperin requires CIAO1 for [4Fe-4S] cluster assembly, and acts through an enzymatic, Fe-S cluster- and SAM-dependent mechanism to inhibit viral RNA synthesis.

Compensatory evolution for a gene deletion is not limited to its immediate functional network.

BACKGROUND: Genetic disruption of an important phenotype should favor compensatory mutations that restore the phenotype. If the genetic basis of the phenotype is modular, with a network of interacting genes whose functions are specific to that phenotype, compensatory mutations are expected among the genes of the affected network. This perspective was tested in the bacteriophage T3 using a genome deleted of its DNA ligase gene, disrupting DNA metabolism. RESULTS: In two replicate, long-term adaptations, phage compensatory evolution accommodated the low ligase level provided by the host without reinventing its own ligase. In both lines, fitness increased substantially but remained well below that of the intact genome. Each line accumulated over a dozen compensating mutations during long-term adaptation, and as expected, many of the compensatory changes were within the DNA metabolism network. However, several compensatory changes were outside the network and defy any role in DNA metabolism or biochemical connection to the disruption. In one line, these extra-network changes were essential to the recovery. The genes experiencing compensatory changes were moderately conserved between T3 and its relative T7 (25% diverged), but the involvement of extra-network changes was greater in T3. CONCLUSION: Compensatory evolution was only partly limited to the known functionally interacting partners of the deleted gene. Thus gene interactions contributing to fitness were more extensive than suggested by the functional properties currently ascribed to the genes. Compensatory evolution offers an easy method of discovering genome interactions among specific elements that does not rest on an a priori knowledge of those elements or their interactions.

Dispersing biofilms with engineered enzymatic bacteriophage.

Synthetic biology involves the engineering of biological organisms by using modular and generalizable designs with the ultimate goal of developing useful solutions to real-world problems. One such problem involves bacterial biofilms, which are crucial in the pathogenesis of many clinically important infections and are difficult to eradicate because they exhibit resistance to antimicrobial treatments and removal by host immune systems. To address this issue, we engineered bacteriophage to express a biofilm-degrading enzyme during infection to simultaneously attack the bacterial cells in the biofilm and the biofilm matrix, which is composed of extracellular polymeric substances. We show that the efficacy of biofilm removal by this two-pronged enzymatic bacteriophage strategy is significantly greater than that of nonenzymatic bacteriophage treatment. Our engineered enzymatic phage substantially reduced bacterial biofilm cell counts by approximately 4.5 orders of magnitude ( approximately 99.997% removal), which was about two orders of magnitude better than that of nonenzymatic phage. This work demonstrates the feasibility and benefits of using engineered enzymatic bacteriophage to reduce bacterial biofilms and the applicability of synthetic biology to an important medical and industrial problem.

Compensatory evolution in response to a novel RNA polymerase: orthologous replacement of a central network gene.

A bacteriophage genome was forced to evolve a new system of regulation by replacing its RNA polymerase (RNAP) gene, a central component of the phage developmental pathway, with that of a relative. The experiment used the obligate lytic phage T7 and the RNAP gene of phage T3. T7 RNAP uses 17 phage promoters, which are responsible for all middle and late gene expression, DNA replication, and progeny maturation, but the enzyme has known physical contacts with only 2 other phage proteins. T3 RNAP was supplied in trans by the bacterial host to a T7 genome lacking its own RNAP gene and the phage population was continually propagated on naive bacteria throughout the adaptation. Evolution of the T3 RNAP gene was thereby prevented, and selection was for the evolution of regulatory signals throughout the phage genome. T3 RNAP transcribes from T7 promoters only at low levels, but a single mutation in the promoter confers high expression, providing a ready mechanism for reevolution of gene expression in this system. When selected for rapid growth, fitness of the engineered phage evolved from a low of 5 doublings/h to 33 doublings/h, close to the expected maximum of 37 doublings/h. However, the experiment was terminated before it could be determined accurately that fitness had reached an obvious plateau, and it is not known whether further adaptation could have resulted in complete recovery of fitness. More than 30 mutations were observed in the evolved genome, but changes were found in only 9 of the 16 promoters, and several coding changes occurred in genes with no known contacts with the RNAP. Surprisingly, the T7 genome adapted to T3 RNAP also maintained high fitness when using T7 RNAP, suggesting that the extreme incompatibility of T7 elements with T3 RNAP is not an invariant property of divergence in these expression systems.

Transcriptional inhibition by an oxidized abasic site in DNA.

2-Deoxyribonolactone (dL) is an oxidized abasic site in DNA that can be induced by gamma-radiolysis, ultraviolet irradiation, and numerous antitumor drugs. Although this lesion is incised by AP endonucleases, suggesting a base-excision repair mechanism for dL removal, subsequent excision and repair synthesis by DNA polymerase beta is inhibited due to accumulation of a protein-DNA cross-link. This raises the possibility that additional repair pathways might be required to eliminate dL from the genome. Transcription-coupled repair (TCR) is a pathway of excision repair specific to DNA lesions present in transcribed strands of expressed genes. A current model proposes that transcription arrest at the site of DNA damage is required to initiate TCR. In support of this model, a strong correlation between transcription arrest by a lesion in vitro and TCR of the lesion in vivo has been found in most cases analyzed. To assess whether dL might be subject to TCR, we have studied the behavior of bacteriophage T3 and T7 RNA polymerases (T3RNAP, T7RNAP) and of mammalian RNA polymerase II (RNAPII) when they encounter a dL lesion or its "caged" precursor located either in the transcribed or in the nontranscribed strand of template DNA. DNA plasmids containing a specifically located dL downstream of the T3, T7 promoter or the Adenovirus major late promoter were constructed and used for in vitro transcription with purified proteins. We found that both dL and its caged precursor located in the transcribed strand represented a complete block to transcription by T3- and T7RNAP. Similarly, they caused more than 90% arrest when transcription was carried out with mammalian RNAPII. Furthermore, RNAPII complexes arrested at dL were subject to the transcript cleavage reaction mediated by elongation factor TFIIS, indicating that these complexes were stable. A dL in the nontranscribed strand did not block either polymerase.

Characterization of bacteriophage T3 DNA ligase.

DNA ligases of bacteriophage T4 and T7 have been widely used in molecular biology for decades, but little is known about bacteriophage T3 DNA ligase. Here is the first report on the cloning, expression and biochemical characterization of bacteriophage T3 DNA ligase. The polyhistidine-tagged recombinant T3 DNA ligase was shown to be an ATP-dependent enzyme. The enzymatic activity was not affected by high concentration of monovalent cations up to 1 M, whereas 2 mM ATP could inhibit its activity by 50%. Under optimal conditions (pH 8.0, 0.5 mM ATP, 5 mM DTT, 1 mM Mg(2+) and 300 mM Na(+)), 1 fmol of T3 DNA ligase could achieve 90% ligation of 450 fmol of cohesive dsDNA fragments in 30 min. T3 DNA ligase was shown to be over 5-fold more efficient than T4 DNA ligase for ligation of cohesive DNA fragments, but less active for blunt-ended DNA fragments. Phylogenetic analysis showed that T3 DNA ligase is more closely related to T7 DNA ligase than to T4 DNA ligase.

Insertion of the T3 DNA polymerase thioredoxin binding domain enhances the processivity and fidelity of Taq DNA polymerase.

Insertion of the T3 DNA polymerase thioredoxin binding domain (TBD) into the distantly related thermostable Taq DNA polymerase at an analogous position in the thumb domain, converts the Taq DNA polymerase from a low processive to a highly processive enzyme. Processivity is dependent on the presence of thioredoxin. The enhancement in processivity is 20-50-fold when compared with the wild-type Taq DNA polymerase or to the recombinant polymerase in the absence of thioredoxin. The recombinant Taq DNA pol/TBD is thermostable, PCR competent and able to copy repetitive deoxynucleotide sequences six to seven times more faithfully than Taq DNA polymerase and makes 2-3-fold fewer AT-->GC transition mutations.

Identification and characterization of T3/T7 bacteriophage-like RNA polymerase sequences in wheat.

Using PCR-based methods, we assembled two wheat cDNA sequences, wheat-G and wheat-C, that encode T3/T7 bacteriophage-like RNA polymerases (RNAPs) sharing 45% amino acid identity. In phylogenetic analyses using maximum likelihood, parsimony and distance methods, the predicted protein sequence of wheat-G (1005 amino acids, 113 kDa) clusters with sequences of previously assigned mitochondrial RNAPs from dicotyledonous plants (Arabidopsis thaliana, Chenopodium album); likewise, in such analyses, the wheat-C sequence (949 amino acids, 107 kDa) affiliates specifically with the Arabidopsis sequence that encodes a phage-like RNAP thought to function in chloroplasts. To confirm biochemically the assignment of the gene encoding the putative wheat mitochondrial RNAP, we isolated a ca. 100 kDa wheat mitochondrial protein that is enriched in fractions displaying specific in vitro transcription activity and that reacts with an antibody raised against a recombinant maize phage-type RNAP. Internal peptide sequence information obtained from the 100-kDa polypeptide revealed that it corresponds to the predicted wheat-G cDNA sequence, providing direct evidence that the wheat-G gene (which we propose to call RpoTm) encodes the wheat mitochondrial RNAP.

Stereospecific differences in repair by human cell extracts of synthesized oligonucleotides containing trans-opened 7,8,9, 10-tetrahydrobenzo[a]pyrene 7,8-diol 9,10-epoxide N2-dG adduct stereoisomers located within the human K-ras codon 12 sequence.

The potent environmental carcinogen benzo[a]pyrene (BaP), following enzymatic activation to enantiomeric pairs of bay-region 7,8-diol 9, 10-epoxides (the benzylic 7-hydroxyl group and epoxide oxygen are cis for DE-1 diastereomers and trans for DE-2 diastereomers), reacts with DNA to form covalent adducts predominately at the exocyclic amino groups of purines. Specific adducts, corresponding to the trans opening of each of the four optically active BaP DE isomers at C-10 by the N 2-amino group of dG, were synthesized as appropriately blocked phosphoramidites and were incorporated at either the first or second G of codon 12 within the G-rich sequence of human K-ras codons 11-13: GCT G1G2T GGC. The adducted oligonucleotides were incorporated into plasmids by primer extension, followed by purification of the covalently closed circular constructs. Adducts derived from either (+)- or (-)-DE-2, placed at either G1 or G2, presented strong blocks to in vitro transcription elongation by bacteriophage T3 RNA polymerase, but only moderately blocked transcription elongation by human RNA polymerase II in nuclear extracts. Adducts derived from all four DEs, placed on either G1 or G2, were used as substrates in a DNA repair synthesis assay using human whole cell extracts. Adducts derived from three of the DE stereoisomers exhibited significant amounts of repair synthesis, but the (-)-DE-2 adduct experienced no repair synthesis above that of the control. Constructs containing a pre-existing nick at the sixth phosphodiester bond 3' to either (+)-DE-2 or (-)-DE-2 adducts exhibited increased repair synthesis.

Rapid mutagenesis and purification of phage RNA polymerases.

We have developed plasmid-based expression systems that encode modified forms of T7 RNA polymerase (RNAP) having 6-12 histidine residues fused to the amino terminus. The histidine-tagged RNAPs (His-T7 RNAPS) are indistinguishable from the wild-type (WT) enzyme in nearly all biochemical assays. Similar plasmids that encode His-tagged T3 and SP6 RNAPs have also been constructed. To facilitate site-directed mutagenesis of the RNAP gene, the size of the target plasmid was minimized by using T7 RNAP itself as a selectable marker. BL21 (DCAT4) cells (which carry a chromosomal copy of the chloramphenicol acetyltransferase cat gene under control of a T7 promoter) are resistant to chloramphenicol when functional T7 RNAP is expressed, thus allowing the selection and maintenance of the target plasmid in these cells. Mutagenesis is accomplished by denaturing the plasmid, annealing mutagenic DNA primers, and repairing the plasmid with T4 DNA polymerase. Two DNA primers are used: one corrects a defect in the bla gene, the other introduces the desired mutation into the RNAP gene; 30-85% of the ampicillin-resistant transformants carry the desired mutation in the RNAP gene. By using BL21 (DCAT4) cells as a recipient for transformation the functional integrity of the RNAP gene may conveniently be monitored by assessing the level of chloramphenicol resistance in vivo. Methods for rapid, simultaneous purification of multiple samples of modified (His-tagged) and conventional RNAPs are described. Together, these developments greatly enhance our ability to characterize this important class of enzymes.

A bacteriophage T3 promoter can be linked to a lethal gene without detectable toxicity for eukaryotic cells. Interest for inducible transgenes.

The bacteriophage T3 promoter can be selectively transcribed by the corresponding RNA polymerase in eukaryotic cells. A toxic gene can in principle be linked to this promoter in a "dormant" and innocuous transgene in a transgenic animal. In this scheme, the activating strain expresses the RNA polymerase. When expression of the gene is needed in the progeny, the 2 lines are crossed. However, when a single molecule is sufficient to kill the cell--as with the diphtheria toxin--transcriptional "leakage" from the promoter may not be tolerated by the cell, even when extremely weak. Therefore, prior to more elaborate studies, diphtheria toxin, as a prototype of a gene toxic to the organism, has been linked to the bacteriophage T3 promoter in a T3-E-DTA construct. The T3-E-DTA plasmid has been transiently transfected into human embryonic kidney derived cells together with a lacZ plasmid. By co-transfection, the T3-E-DTA cells can be readily identified as lacZ positive, and their fate followed by the production of beta-galactosidase at the single cell or overall population level. In spite of the extreme toxicity of the toxin, the cells tolerate the presence of the T3-E-DTA construct, and are only killed--with a high efficiency--when the T3 RNA polymerase is present. Transactivation is usually restricted to the auxiliary factors of transcription. With this study, the promoter and the polymerase are revealed as potential and efficient inducible and activating elements of a very simple binary system.

Sequences homologous to yeast mitochondrial and bacteriophage T3 and T7 RNA polymerases are widespread throughout the eukaryotic lineage.

Although mitochondria and chloroplasts are considered to be descendants of eubacteria-like endo- symbionts, the mitochondrial RNA polymerase of yeast is a nucleus-encoded, single-subunit enzyme homologous to bacteriophage T3 and T7 RNA polymerases, rather than a multi-component, eubacterial-type alpha 2 beta beta' enzyme, as encoded in chloroplast DNA. To broaden our knowledge of the mitochondrial transcriptional apparatus, we have used a polymerase chain reaction (PCR) approach designed to amplify an internal portion of phage T3/T7-like RNA polymerase genes. Using this strategy, we have recovered sequences homologous to yeast mitochondrial and phage T3/T7 RNA polymerases from a phylogenetically broad range of multicellular and unicellular eukaryotes. These organisms display diverse patterns of mitochondrial genome organization and expression, and include species that separated from the main eukaryotic line early in the evolution of this lineage. In certain cases, we can deduce that PCR-amplified sequences, some of which contain small introns, are localized in nuclear DNA. We infer that the T3/T7-like RNA polymerase sequences reported here are likely derived from genes encoding the mitochondrial RNA polymerase in the organisms in which they occur, suggesting a phage T3/T7-like RNA polymerase was recruited to act in transcription in the mitochondrion at an early stage in the evolution of this organelle.

Reverse 5' caps in RNAs made in vitro by phage RNA polymerases.

We show that about one-third of the RNAs produced in vitro by viral RNA polymerases in the presence of m7GpppG dinucleotides have unusual 5' caps. In these RNAs, the initiating dinucleotide is incorporated in an orientation opposite to that expected so that the 7-methyl guanine (m7G) nucleotide is adjacent to the body of the RNA, making a "reverse" cap. The doubly methylated dinucleotide, m7GpppGm, containing a 2' O-methylated guanine (Gm) is incorporated only in the reverse orientation. Precursors of U1 snRNAs containing reverse caps are recognized by antibodies specific for the m7G cap structure. When injected into Xenopus laevis oocyte nuclei, reverse-capped pre-U1 RNAs are exported considerably more slowly than normal. Furthermore, U1 RNAs with reverse caps exhibit a striking defect in nuclear import that can be attributed to the failure of reverse caps to be hypermethylated to m2,2,7G caps. Thus, the presence of reverse-capped RNAs in RNA preparations may affect conclusions about the efficiency and extent of certain m7G cap-dependent processes.

Structural and functional domains of the large subunit of the bacteriophage T3 DNA packaging enzyme: importance of the C-terminal region in prohead binding.

During head assembly of phage T3, DNA is packaged into a preformed protein shell, called the prohead, with the aid of non-capsid packaging proteins, the products of genes 18 and 19 (gp18 and gp19). We have developed a defined system, composed of purified gp18,gp19 and proheads for in vitro packaging of T3 DNA. Our previous results using the defined in vitro system indicate the sequential events in DNA packaging: the packaging proteins, gp18 and gp19, bind DNA and proheads, respectively. These complexes associate to form a direct precursor complexes for DNA translocation into the head. The formation of the precursor complexes requires ATP as an allosteric effector. Subsequent DNA translocation is driven by ATP hydrolysis. gp19 is an ATP binding protein that plays multiple roles in DNA packaging through interaction with ATP. gp19 changes its conformation by binding to ATP, as judged from the analysis of limited proteolysis. Sites cleaved by limited proteolysis were determined and mapped on the gp19 polypeptide (586 amino acid residues) to image the conformational change of gp19 induced by ATP. C-Terminal fragments generated by trypsin digestion bound the prohead and inhibited DNA packaging by intact gp19 in a competitive manner. On the other hand, N-terminal fragments did not bind the prohead nor did they inhibit DNA packaging. These results define a prohead binding domain at the C terminus of gp19. To identify the prohead binding domain more precisely, deletion mutants lacking the last 10 and 15 amino acids (gp19-delta C10 and gp19-delta C15, respectively) of the extreme C terminus of gp19 were constructed. Limited tryptic digestion patterns of these mutant proteins in the presence or absence of ATP were basically the same as those of gp19-wt, indicating that the conformation and its ATP response were not changed by these deletions. gp19-delta C15 lacked prohead binding activity and, therefore, DNA packaging activity. gp19-delta C10 had significant DNA packaging activity although it was reduced to one-tenth of that of gp19-wt. These results indicate that a C-terminal region of residues L571 to D576 of gp19 is crucial for prohead binding and that the last ten residues D577 to W586 of the C terminus seems to be important in stable binding of gp19 to the prohead.

A functional chimeric DNA primase: the Cys4 zinc-binding domain of bacteriophage T3 primase fused to the helicase of bacteriophage T7.

Two colinear bacteriophage T7 gene 4 proteins provide helicase and primase functions in vivo. T7 primase differs from T7 helicase by an additional 63 residues at the amino terminus. This terminal domain contains a zinc-binding motif which mediates an interaction with the basic primase recognition sequence 3'-CTG-5'. We have generated a chimeric primase in which the 81 amino-terminal residues are derived from the primase of phage T3 and the 484 carboxyl-terminal residues are those of phage T7 helicase. The amino-terminal domain of T3 primase is 50% homologous with that of T7 primase. The resulting T3/T7 chimeric protein is a functional primase in vivo. While the primase activity of the purified protein is about one-third that of T7 primase, the recognition sites used and the oligoribonucleotides synthesized from these sites are identical. We conclude that the residues responsible for the interaction with the sequence 3'-CTG-5' are conserved between the chimeric and T7 proteins.

Reduced ethylene synthesis by transgenic tomatoes expressing S-adenosylmethionine hydrolase.

We have utilized a gene from bacteriophage T3 that encodes the enzyme S-adenosylmethionine hydrolase (SAMase) to generate transgenic tomato plants that produce fruit with a reduced capacity to synthesize ethylene. S-adenosylmethionine (SAM) is the metabolic precursor of 1-aminocyclopropane-1-carboxylic acid, the proximal precursor to ethylene. SAMase catalyzes the conversion of SAM to methylthioadenosine and homoserine. To restrict the presence of SAMase to ripening fruit, the promoter from the tomato E8 gene was used to regulate SAMase gene expression. Transgenic tomato plants containing the 1.1 kb E8 promoter bore fruit that expressed SAMase during the breaker and orange stage of fruit ripening and stopped expression after the fruit fully ripened. Plants containing the 2.3 kb E8 promoter expressed SAMase at higher levels during the post-breaker phases of fruit ripening and had a substantially reduced capacity to synthesize ethylene.

The thumb's knuckle. Flexibility in the thumb subdomain of T7 RNA polymerase is revealed by the structure of a chimeric T7/T3 RNA polymerase.

We have solved the structure of a chimeric T7/T3 RNA polymerase (RNAP) in an orthorhombic crystal by molecular replacement with the T7 RNAP structure determined from a monoclinic crystal. The structure of the protruding "thumb" subdomain of the polymerase appears very different in these two crystals apparently because of differences in packing contacts made by the thumb subdomain. These observations support the proposal that the thumb subdomain is flexible and can wrap around bound template to obstruct polymerase: template dissociation during processive synthesis.

Continuous in vitro evolution of bacteriophage RNA polymerase promoters.

Rapid in vitro evolution of bacteriophage T7, T3, and SP6 RNA polymerase promoters was achieved by a method that allows continuous enrichment of DNAs that contain functional promoter elements. This method exploits the ability of a special class of nucleic acid molecules to replicate continuously in the presence of both a reverse transcriptase and a DNA-dependent RNA polymerase. Replication involves the synthesis of both RNA and cDNA intermediates. The cDNA strand contains an embedded promoter sequence, which becomes converted to a functional double-stranded promoter element, leading to the production of RNA transcripts. Synthetic cDNAs, including those that contain randomized promoter sequences, can be used to initiate the amplification cycle. However, only those cDNAs that contain functional promoter sequences are able to produce RNA transcripts. Furthermore, each RNA transcript encodes the RNA polymerase promoter sequence that was responsible for initiation of its own transcription. Thus, the population of amplifying molecules quickly becomes enriched for those templates that encode functional promoters. Optimal promoter sequences for phage T7, T3, and SP6 RNA polymerase were identified after a 2-h amplification reaction, initiated in each case with a pool of synthetic cDNAs encoding greater than 10(10) promoter sequence variants.