A method for directed evolution and functional cloning of enzymes

A general scheme is described for the in vitro evolution of protein catalysts in a biologically amplifiable system. Substrate is covalently and site specifically attached by a flexible tether to the pIII coat protein of a filamentous phage that also displays the catalyst. Intramolecular conversion of substrate to product provides a basis for selecting active catalysts from a library of mutants, either by release from or attachment to a solid support. This methodology has been developed with the enzyme staphylococcal nuclease as a model. An analysis of factors influencing the selection efficiency is presented, and it is shown that phage displaying staphylococcal nuclease can be enriched 100-fold in a single step from a library-like ensemble of phage displaying noncatalytic proteins. Additionally, this approach should allow one to functionally clone natural enzymes, based on their ability to catalyze specific reactions (e.g., glycosyl transfer, sequence-specific proteolysis or phosphorylation, polymerization, etc.) rather than their sequence- or structural homology to known enzymes.

Mutagenesis of the regulatory domain of phenylalanine hydroxylase

The regulatory domain of phenylalanine hydroxylase (PAH, EC 1.14.16.1) consists of more than 100 amino acids at the N terminus, the removal of which significantly activates the enzyme. To study the regulatory properties controlled by the N terminus, a series of truncations and site-specific mutations were made in this region of rat PAH. These enzymes were expressed highly in Escherichia coli and purified through a pterin-conjugated Sepharose affinity column. The removal of the first 26 amino acids of the N terminus increased the activity by about 20-fold, but removal of the first 15 amino acids increased the activity by only 2-fold. Replacing serine-29 of rat PAH with cysteine from the same site of human PAH increased the activity by more than 4-fold. Mutation of serine to other amino acids with varying side chains: alanine, methionine, leucine, aspartic acid, asparagine, and arginine also resulted in significant activation, indicating a serine-specific inhibitory effect. But these site-specific mutants showed 30–40% lower activity when assayed with 6-methyl-5,6,7,8-tetrahydropterin. Stimulation of hydroxylase activity by preincubation of the enzyme with phenylalanine was inversely proportional to the activation state of all these mutants. Combined with recent crystal structures of PAH [Kobe, B. et al. (1999) Nat. Struct. Biol. 6, 442–448; and Erlandsen, H., Bjorgo, E., Flatmark, T. & Stevens, R. C. (2000) Biochemistry 39, 2208–2217], these data suggest that residues 16–26 have a controlling regulatory effect on the activity by interaction with the dihydroxypropyl side chain of (6R)-5,6,7,8-tetrahydrobiopterin. The serine/cysteine switch explains the difference in regulatory properties between human and rat PAH. The N terminus as a whole is important for maintaining rat PAH in an optimum catalytic conformation.

Orthogonal combinatorial mutagenesis: a codon-level combinatorial mutagenesis method useful for low multiplicity and amino acid-scanning protocols

We describe here a method to generate combinatorial libraries of oligonucleotides mutated at the codon-level, with control of the mutagenesis rate so as to create predictable binomial distributions of mutants. The method allows enrichment of the libraries with single, double or larger multiplicity of amino acid replacements by appropriate choice of the mutagenesis rate, depending on the concentration of synthetic precursors. The method makes use of two sets of deoxynucleoside-phosphoramidites bearing orthogonal protecting groups [4,4′-dimethoxytrityl (DMT) and 9-fluorenylmethoxycarbonyl (Fmoc)] in the 5′ hydroxyl. These phosphoramidites are divergently combined during automated synthesis in such a way that wild-type codons are assembled with commercial DMT-deoxynucleoside-methyl-phosphoramidites while mutant codons are assembled with Fmoc-deoxynucleoside-methyl-phosphoramidites in an NNG/C fashion in a single synthesis column. This method is easily automated and suitable for low mutagenesis rates and large windows, such as those required for directed evolution and alanine scanning. Through the assembly of three oligonucleotide libraries at different mutagenesis rates, followed by cloning at the polylinker region of plasmid pUC18 and sequencing of 129 clones, we concluded that the method performs essentially as intended.

Site-specifically located 8-amino-2′-deoxyguanosine: thermodynamic stability and mutagenic properties in Escherichia coli

2-Nitropropane (2-NP), an important industrial solvent and a component of cigarette smoke, is mutagenic in bacteria and carcinogenic in rats. 8-Amino-2′-deoxyguanosine (8-amino-dG) is one of the types of DNA damage found in liver, the target organ in 2-NP-treated rats. To investigate the thermodynamic properties of 8-amino-dG opposite each of the four DNA bases, we have synthesized an 11mer, d(CCATCG*CTACC), in which G* represents the modified base. By annealing a complementary DNA strand to this modified 11mer, four sets of duplexes were generated each containing one of the four DNA bases opposite the lesion. Circular dichroism studies indicated that 8-amino-dG did not alter the global helical properties of natural right-handed B-DNA. The thermal stability of each duplex was examined by UV melting measurements and compared with its unmodified counterpart. For the unmodified 11mer, the relative stability of the complementary DNA bases opposite G was in the order C > T > G > A, as determined from their –ΔG° values. The free energy change of each modified duplex was lower than its unmodified counterpart, except for the G*:G pair that exhibited a higher melting transition and a larger –ΔG° than the G:G duplex. Nevertheless, the stability of the modified 11mer duplex also followed the order C > T > G > A when placed opposite 8-amino-dG. To explore if 8-amino-dG opposite another 8-amino-dG has any advantage in base pairing, a G*:G* duplex was evaluated, which showed that the stability of this duplex was similar to the G*:G duplex. Mutagenesis of 8-amino-dG in this sequence context was studied in Escherichia coli, which showed that the lesion is weakly mutagenic (mutation frequency ∼10–3) but still can induce a variety of targeted and semi-targeted mutations.

Fission Yeast Aip3p (spAip3p) Is Required for an Alternative Actin-directed Polarity Program

Aip3p is an actin-interacting protein that regulates cell polarity in budding yeast. The Schizosaccharomyces pombe-sequencing project recently led to the identification of a homologue of Aip3p that we have named spAip3p. Our results confirm that spAip3p is a true functional homologue of Aip3p. When expressed in budding yeast, spAip3p localizes similarly to Aip3p during the cell cycle and complements the cell polarity defects of an aip3Δ strain. Two-hybrid analysis shows that spAip3p interacts with actin similarly to Aip3p. In fission yeast, spAip3p localizes to both cell ends during interphase and later organizes into two rings at the site of cytokinesis. spAip3p localization to cell ends is dependent on microtubule cytoskeleton, its localization to the cell middle is dependent on actin cytoskeleton, and both patterns of localization require an operative secretory pathway. Overexpression of spAip3p disrupts the actin cytoskeleton and cell polarity, leading to morphologically aberrant cells. Fission yeast, which normally rely on the microtubule cytoskeleton to establish their polarity axis, can use the actin cytoskeleton in the absence of microtubule function to establish a new polarity axis, leading to the formation of branched cells. spAip3p localizes to, and is required for, branch formation, confirming its role in actin-directed polarized cell growth in both Schizosaccharomyces pombe and Saccharomyces cerevisiae.

Two recombination-dependent DNA replication pathways of bacteriophage T4, and their roles in mutagenesis and horizontal gene transfer

Two major pathways of recombination-dependent DNA replication, “join-copy” and “join-cut-copy,” can be distinguished in phage T4: join-copy requires only early and middle genes, but two late proteins, endonuclease VII and terminase, are uniquely important in the join-cut-copy pathway. In wild-type T4, timing of these pathways is integrated with the developmental program and related to transcription and packaging of DNA. In primase mutants, which are defective in origin-dependent lagging-strand DNA synthesis, the late pathway can bypass the lack of primers for lagging-strand DNA synthesis. The exquisitely regulated synthesis of endo VII, and of two proteins from its gene, explains the delay of recombination-dependent DNA replication in primase (as well as topoisomerase) mutants, and the temperature-dependence of the delay. Other proteins (e.g., the single-stranded DNA binding protein and the products of genes 46 and 47) are important in all recombination pathways, but they interact differently with other proteins in different pathways. These homologous recombination pathways contribute to evolution because they facilitate acquisition of any foreign DNA with limited sequence homology during horizontal gene transfer, without requiring transposition or site-specific recombination functions. Partial heteroduplex repair can generate what appears to be multiple mutations from a single recombinational intermediate. The resulting sequence divergence generates barriers to formation of viable recombinants. The multiple sequence changes can also lead to erroneous estimates in phylogenetic analyses.

Site-specific integration by adeno-associated virus.

Adeno-associated virus (AAV) has attracted considerable interest as a potential vector for gene delivery. Wild-type virus is notable for the lack of association with any human disease and the ability to stably integrate its genome in a site-specific manner in a locus on human chromosome 19 (AAVS1). Use of a functional model system for AAV DNA integration into AAVS1 has allowed us to conclude that the recombination event is directed by cellular DNA sequences. Recombinant junctions isolated from our integration assay were analyzed and showed characteristics similar to those found in latently infected cell lines. The minimal DNA signals within AAVS1 required for targeted integration were identified and shown to contain functional motifs of the viral origin of replication. A replication mediated model of AAV DNA integration is proposed.

Identification of three physically and functionally distinct binding sites for C3b in human complement factor H by deletion mutagenesis.

Human complement factor H controls spontaneous activation of complement in plasma and appears to play a role in distinguishing host cells from activators of the alternative pathway of complement. In both mice and humans, the protein is composed of 20 homologous short consensus repeat (SCR) domains. The size of the protein suggests that portions of the structure outside the known C3b binding site (SCR 1-4) possess a significant biological role. We have expressed the full-length cDNA of factor H in the baculovirus system and have shown the recombinant protein to be fully active. Mutants of this full-length protein have now been prepared, purified, and examined for cofactor activity and binding to C3b and heparin. The results demonstrate (i) that factor H has at least three sites that bind C3b, (ii) that one of these sites is located in SCR domains 1-4, as has been shown by others, (iii) that a second site exists in the domain 6-10 region, (iv) that a third site resides in the SCR 16-20 region, and (v) that two heparin binding sites exist in factor H, one near SCR 13 and another in the SCR 6-10 region. Functional assays demonstrated that only the first C3b site located in SCR 1-4 expresses factor I cofactor activity. Mutant proteins lacking any one of the three C3b binding sites exhibited 6- to 8-fold reductions in affinity for C3b on sheep erythrocytes, indicating that all three sites contribute to the control of complement activation on erythrocytes. The identification of multiple functionally distinct sites on factor H clarifies many of the heretofore unexplainable behaviors of this protein, including the heterogeneous binding of factor H to surface-bound C3b, the effects of trypsin cleavage, and the differential control of complement activation on activators and nonactivators of the alternative pathway of complement.

Random mutagenesis of Thermus aquaticus DNA polymerase I: concordance of immutable sites in vivo with the crystal structure.

Expression of Thermus aquaticus (Taq) DNA polymerase I (pol I) in Escherichia, coli complements the growth defect caused by a temperature-sensitive mutation in the host pol I. We replaced the nucleotide sequence encoding amino acids 659-671 of the O-helix of Taq DNA pol I, corresponding to the substrate binding site, with an oligonucleotide containing random nucleotides. Functional Taq pol I mutants were selected based on colony formation at the nonpermissive temperature. By using a library with 9% random substitutions at each of 39 positions, we identified 61 active Taq pol I mutants, each of which contained from one to four amino acid substitutions. Some amino acids, such as alanine-661 and threonine-664, were tolerant of several or even many diverse replacements. In contrast, no replacements or only conservative replacements were identified at arginine-659, lysine-663, and tyrosine-671. By using a library with totally random nucleotides at five different codons (arginine-659, arginine-660, lysine-663, phenylalanine-667, and glycine-668), we confirmed that arginine-659 and lysine-663 were immutable, and observed that only tyrosine substituted for phenylalanine-667. The two immutable residues and the two residues that tolerate only highly conservative replacements lie on the side of O-helix facing the incoming deoxynucleoside triphosphate, as determined by x-ray analysis. Thus, we offer a new approach to assess concordance of the active conformation of an enzyme, as interpreted from the crystal structure, with the active conformation inferred from in vivo function.

Trypanosome U-deletional RNA editing involves guide RNA-directed endonuclease cleavage, terminal U exonuclease, and RNA ligase activities.

We have studied the mechanism of accurate in vitro RNA editing of Trypanosoma brucei ATPase 6 mRNA, using four mRNA-guide RNA (gRNA) pairs that specify deletion of 2, 3, or 4 U residues at editing site 1 and mitochondrial extract. This extract not only catalyzes deletion of the specified number of U residues but also exhibits a novel endonuclease activity that cleaves the input pre-mRNA in a gRNA-directed manner, precisely at the phosphodiester bond predicted in a simple enzymatic model of RNA editing. This cleavage site is inconsistent with a chimera-based editing mechanism. The U residues to be deleted, present at the 3' end of the upstream cleavage product, are then removed evidently by a 3' U-specific exonuclease and not by a reverse reaction of terminal U transferase. RNA ligase can then join the mRNA halves through their newly formed 5' P and 3' OH termini, generating mRNA faithfully edited at the first editing site. This resultant, partially edited mRNA can then undergo accurate, gRNA-directed cleavage at editing site 2, again precisely as predicted by the enzymatic editing model. All of these enzymatic activities cofractionate with the U-deletion activity and may reside in a single complex. The data imply that each round of editing is a four-step process, involving (i) gRNA-directed cleavage of the pre-mRNA at the bond immediately 5' of the region base paired to the gRNA, (ii) U deletion from or U addition to the 3' OH of the upstream mRNA half, (iii) ligation of the mRNA halves, and (iv) formation of additional base pairing between the correctly edited site and the gRNA that directs subsequent nuclease cleavage at the next editing site.

Single-copy transgenic mice with chosen-site integration.

We describe a general way of introducing transgenes into the mouse germ line for comparing different sequences without the complications of variation in copy number and insertion site. The method uses homologous recombination in embryonic stem (ES) cells to generate mice having a single copy of a transgene integrated into a chosen location in the genome. To test the method, a single copy murine bcl-2 cDNA driven by either a chicken beta-actin promoter or a human beta-actin promoter has been inserted immediately 5' to the X-linked hypoxanthine phosphoribosyltransferase locus by a directly selectable homologous recombination event. The level of expression of the targeted bcl-2 transgene in ES cells is identical in independently isolated homologous recombinants having the same promoter yet varies between the different promoters. In contrast, the expression of bcl-2 transgenes having the same (chicken beta-actin) promoter varies drastically when they are independently integrated at random insertion sites. Both promoters direct broad expression of the single-copy transgene in mice derived from the respective targeted ES cells. In vitro and in vivo, the human beta-actin promoter consistently directed a higher level of transgene expression than the chicken beta-actin promoter.

Mini-chromosomes derived from the human Y chromosome by telomere directed chromosome breakage.

We have used telomeric DNA to break two acrocentric derivatives of the human Y chromosome into mini-chromosomes that are small enough to be size- fractionated by pulsed-field gel electrophoresis. One of the mini-chromosomes is about 7 Mb in size and sequence-tagged site analysis of this molecule suggests that it corresponds to a simple truncation of the short arm of the Y chromosome. Five of the mini-chromosomes are derived from the long arm, are all rearranged by more than a simple truncation, and range in size from 4.0 Mb to 9 Mb. We have studied the mitotic stabilities of these mini-chromosomes and shown that they are stably maintained by cells proliferating in culture for about 100 cell divisions.

Fos and Jun do not bend the AP-1 recognition site.

We have used a solution-based DNA cyclization assay and a gel-phasing method to show that contrary to previous reports [Kerppola, T. K. & Curran, T. (1991) Cell 66, 317-326], basic region leucine zipper proteins Fos and Jun do not significantly bend their AP-1 recognition site. We have constructed two sets of DNA constructs that contain the 7-bp 5'-TGACTCA-3' AP-1 binding site, from either the yeast or the human collagenase gene, which is well separated from and phased by 3-4 helical turns against an A tract-directed bend. The cyclization probabilities of DNAs with altered phasings are not significantly affected by Fos-Jun binding. Similarly, Fos-Jun and Jun-Jun bound to differently phased DNA constructs show insignificant variations in gel mobilities. Both these methods independently indicate that Fos and Jun bend their AP-1 target site by <5 degrees, an observation that has important implications in understanding their mechanism of transcriptional regulation.

Mutagenesis in the C-terminal region of human interleukin 5 reveals a central patch for receptor alpha chain recognition.

Cassette mutagenesis was used to identify side chains in human interleukin 5 (hIL-5) that mediate binding to hIL-5 receptor alpha chain (hIL-5R alpha). A series of single alanine substitutions was introduced into a stretch of residues in the C-terminal region, including helix D, which previously had been implicated in receptor alpha chain recognition and which is aligned on the IL-5 surface so as to allow the topography of receptor binding residues to be examined. hIL-5 and single site mutants were expressed in COS cells, their interactions with hIL-5R alpha were measured by a sandwich surface plasmon resonance biosensor method, and their biological activities were measured by an IL-5-dependent cell proliferation assay. A pattern of mutagenesis effects was observed, with greatest impact near the interface between the two four-helix bundles of IL-5, in particular at residues Glu-110 and Trp-111, and least at the distal ends of the D helices. This pattern suggests the possibility that residues near the interface of the two four-helix bundles in hIL-5 comprise a central patch or hot spot, which constitutes an energetically important alpha chain recognition site. This hypothesis suggests a structural explanation for the 1:1 stoichiometry observed for the complex of hIL-5 with hIL-5R alpha.

(Table 1) Summary of basalt unit thickness, vesicularity, number of flow units, and average porosity at ODP Site 135-839

An inflatable drill-string packer was used at Site 839 to measure the bulk in-situ permeability within basalts cored in Hole 839B. The packer was inflated at two depths, 398.2 and 326.9 mbsf; all on-board information indicated that the packer mechanically closed off the borehole, although apparently the packer hydraulically sealed the borehole only at 398.2 mbsf. Two pulse tests were run at each depth, two constant-rate injection tests were run at the first set, and four were run at the second. Of these, only the constant-rate injection tests at the first set yielded a permeability, calculated as ranging from 1 to 5 * 10**-12 m**2. Pulse tests and constant-rate injection tests for the second set did not yield valid data. The measured permeability is an upper limit; if the packer leaked during the experiments, the basalt would be less permeable. In comparison, permeabilities measured at other Deep Sea Drilling Project and Ocean Drilling Program sites in pillow basalts and flows similar to those measured in Hole 839B are mainly about 10**-13 to 10**-14 m**2. Thus, if our results are valid, the basalts at Site 839 are more permeable than ocean-floor basalts investigated elsewhere. Based on other supporting evidence, we consider these results to be a valid measure of the permeability of the basalts. Temperature data and the geochemical and geotechnical properties of the drilled sediments all indicate that the site is strongly affected by fluid flow. The heat flow is very much less than expected in young oceanic basalts, probably a result of rapid fluid circulation through the crust. The geochemistry of pore fluids is similar to that of seawater, indicating seawater flow through the sediments, and sediments are uniformly underconsolidated for their burial depth, again indicating probable fluid flow. The basalts are highly vesicular. However, the vesicularity can only account for part of the average porosity measured on the neutron porosity well log; the remainder of the measured porosity is likely present as voids and fractures within and between thin-bedded basalts. Core samples, together with porosity, density, and resistivity well-log data show locations where the basalt section is thin bedded and probably has from 15% to 35% void and fracture porosity. Thus, the measured permeability seems reasonable with respect to the high measured porosity. Much of the fluid flow at Site 839 could be directed through highly porous and permeable zones within and between the basalt flows and in the sediment layer just above the basalt. Thus, the permeability measurements give an indication of where and how fluid flow may occur within the oceanic crust of the Lau Basin.