WO2008005529A2 - Cell-mediated directed evolution - Google Patents

Abstract

The present invention relates to systems which harness the molecular biology of a living cell to direct evolution of biological entities of interest. According to the invention, a host cell is engineered to facilitate mutation of a nucleic acid target corresponding to that entity and select for desirable mutants. As applied to populations of host cells, the invention provides a means to generate and contemporaneously select mutants of interest, allowing for the production of extremely diverse libraries enriched in the most 'fit' mutants.

Description

CELL-MEDIATED DIRECTED EVOLUTION

PRIORITY CLAIM This application claims priority to United States Provisional

Application No. 60/891,833, filed February 27, 2007; United States Provisional Application No. 60/819,531 filed July 7, 2006, and 60/819,532 filed July 7, 2006, the contents of each of which are hereby incorporated by reference in their entireties.

GRANT INFORMATION

The subject matter herein was developed at least in part under National

Institute of Health Grant No. 2R01GM062867-06, National Institute of Health Grant No. GM62867-01A1 and National Science Foundation Grant No. CHE 0350183, and a National Science Foundation Graduate Research Fellowship to Jonathan Bronson, so that the United States Government has certain rights herein.

1. INTRODUCTION

The present invention relates to systems which harness the molecular biology of a living cell to direct evolution of biological entities of interest. According to the invention, a host cell is engineered to facilitate mutation of a nucleic acid target corresponding to that entity and select for desirable mutants. As applied to populations of host cells, the invention provides a means to generate and contemporaneously select mutants of interest, allowing for the production of extremely diverse libraries enriched in the most "fit" species of mutants. The methods of the invention, which include the generation of secondary libraries by genetic exchange {e.g. sexual reproduction) between "evolved" libraries, are calculated to be capable of producing libraries having a diversity of 10²⁴ or greater. In specific, non-limiting embodiments, the present invention provides for directed evolution of hydrolases useful to the energy sector. 2. BACKGROUND OF THE INVENTION

2.1 DIRECTED EVOLUTION

As compared to evolution in nature, a slow process in which mutations occur randomly and apt mutants persist through natural selection, directed evolution is a strategic approach for rapidly creating and selecting useful mutants in the laboratory. Directed evolution holds promise for creating designer enzymes for use as reagents in biomedical research and medical diagnostics, and perhaps even as therapeutics.

Briefly, the history of the development of directed evolution is as follows. In the 1990's, the ability to reverse transcribe and amplify RNA directly was successfully exploited to select for RNA molecules with de novo binding and catalytic functions from very large (>10¹² unique molecules) random libraries of RNA (Wilson and Szostak, 1999, Annu. Rev. Biochem. 68_i 61 1 -647;. Jenison et al., 1994, Science 263: 1425-1429; Tsang and Joyce, 1996, J. MoI. Biol. 262: 31-42; Sassanfar and Szostak, 1993, Nature 364: 550-553). Beginning with the introduction of phage display technology in 1985 ( Smith, 1985, Science 228: 1315-1317), the development to robust methods (Merryman et al., 2002, Chem. Biol. 9; 741-746; Hanes and Pluckthun, 1997, Proc. Natl. Acad. Sci. USA. 94: 4937-4942; Roberts and Szostak, 1997, Proc. Natl. Acad. Sci. USA. 94: 12297-12302; Colby et al., 2004, Methods Enzymol. 388: 348-358; Arnold, 2001 , Nature 409: 253-257) for directly or indirectly "tagging" a protein with its DNA sequence for very large library sizes brought momentum to the protein directed evolution field (Arnold, 2001, Nature 409: 253- 257; Lin and Cornish, 2002, Angew. Chem. Int. Ed. Engl. 4Jh 4402-4425; Aharoni et al., 2005, Curr. Opin. Chem. Biol. 9: 210-216; Magliery and Regan, 2004, Eur. J. Biochem. 271 : 1595-1608). These methods lend themselves naturally to selections for binding — for example, by repeated passage of the protein-DNA complex over a small-molecule ligand affinity matrix. The last several years have now seen multiple reports from different laboratories of the de novo directed evolution of binding proteins from large libraries (>10⁸; Binz et al., 2005, Nat. Biotechnol. 23÷ 1257- 1268). In 1999, Skerra and coworkers reported the de novo selection of a lipocalin with a 65 nM dissociation constant for fluorescein (Beste et al., 1999, Proc. Natl. Acad. Sci. USA. 96κ 1898-1903.). De novo binding proteins have also been reported using libraries of fϊbronectin, thioredoxin, and ankyrin and TPR repeat proteins (Chivers et al., 1996, EMBO J. JjJ: 2659-2667; Cortajarena et al., 2004, Protein Eng. Des. SeI. Hi 399-409; Rutledge et al., 2003, J. Am. Chem. Soc. 125: 14336-14347; Binz et al., 2004, Nat. Biotechnol. 22i 575-582).

The foregoing reports utilize a similar strategy. An existing, well- studied protein fold that naturally binds to diverse targets is used as a privileged binding "scaffold". Flexible loops in the protein that comprise most of the ligand- binding pocket are randomized to all 20 amino acids to create very large libraries

(>10²⁰). Active protein variants are selected from a subset of ca. 10⁸ of these variants based on binding to an affinity matrix using methods like phage display to link each protein variant to its mutatable and amplifiable DNA code.

While directed evolution of binding proteins has met with early success, directed evolution of catalytic proteins - enzymes- has been difficult. The problem for enzyme catalysis is that these methods like phage display that effectively create a protein-DNA complex are difficult to adapt to select for catalytic turnover (Arnold, 2001, Nature 409: 253-257; Lin and Cornish, 2002, Angew. Chem. Int. Ed. Engl. 4ii 4402-4425). Bringing the power of directed evolution to bear on more demanding applications in enzyme catalysis, such as changing the substrate specificity of an existing enzyme, appears to require larger library sizes that cannot be achieved even with automation of these traditional assays. Proof of the feasibility of changing enzyme substrate specificity when large library sizes (ca. 10⁸) can be achieved comes from the application of traditional genetic complementation assays to directed evolution. For example, Yano and coworkers were able to increase the activity of aspartate aminotransferase with β-branched substrates by five orders of magnitude using an Escherichia coli (E. colϊ) β-branched amino acid auxotroph, with multiple rounds of selection from libraries of 10⁶-10⁷ variants (Yano et al., 1998, Proc. Natl. Acad. Sci. USA. 95: 5511-5515). One of the early examples of test tube directed evolution was carried out with the classic TIM barrel triose-phosphate isomerase (Hermes et al., 1990, Proc. Natl. Acad Sci. 87:696-700). More recently the wealth of sequence data and sophisticated understanding of structure-function relationships in TlM barrel' enzymes has been exploited to interconvert existing TIM barrel enzymes with dramatic changes in substrate specificity and catalytic activity (Schmidt et al., 2003,

Biochemistry 42i 8387-8393; Jurgens et al., 2000, Proc. Natl. Acad. Sci. USA. 97: 9925-9930; Silverman et al., 2001, Proc Natl Acad Sci U S A. 98: 3092-309775-78; Hsu et al., 2005, Proc. Natl. Acad. Sci. USA. 102: 9122-9126).

Most recently, an αβ/ βα metal lohydrolase that hydrolyzes the thioester bond in S-D-lactoylglutathione was converted to a related αβ/ βα metal lohydrolase with β-lactamase activity by mutagenesis and randomization of four key active site loops from libraries of 10⁷-10⁸ variants using an ampicillin-resistance growth selection in E. coli (Park et al., 2006, Science 311 : 535-538). These complementation assays, however, limit directed evolution for the most part to the specific substrates and reactions already carried out by Nature (for an exception, see Wang et al., 2001, Science 292: 498-500). De novo test tube directed evolution of catalytic function may be predicted to require even larger library sizes.

2.2 CHEMICAL COMPLEMENTATION For the handful of reactions for which there are high-throughput selections for function (for example, the synthesis of a detectable reaction product), directed evolution has been used to change enzyme substrate specificity using large libraries (about 10⁸). For reactions which do not themselves provide a means for selection, the technique of "chemical complementation" has been developed to create suitable selection mechanisms (see, for example, United States Patent Application Publication No. 20040106154 by Cornish; Abida et al., 2002, Chembiochem 3:887- 895; Baker et al., 2002, Proc. Natl. Acad. Sci. 99: 16537-16542; Baker et al., 2003, Anal. Biochem 315:134-137; Lin et al., J. Am. Chem. Soc. 126:15051-15059; de Fillipe et al., 2004, Biochemistry 43:10353-10363;Senugupta et al., 2004, Biochemistry 43:3570-3581 ; Carter et al., 2005, Chembiochem. 6:2055-2067). In chemical complementation, enzyme catalysis is linked to cell survival or phenotype via covalent coupling of two small molecule three-hybrid ligands that reconstitute a transcriptional activator such that the chemistry can be readily varied. A heterodimeric small molecule bridges a DNA-binding domain (DBD)-receptor fusion protein and an activation domain (AD)-receptor fusion protein. When the DNA binding Domain and Activation Domain are joined by the small molecule bridge, the DNA-binding domain can bind to a corresponding binding site in a host cell which is operably linked to a reporter gene, and transcription of the reporter gene can occur. Catalytic activity of an enzyme E can be detected where the small molecule bridge is a substrate of the enzyme (see FIGURE 1). Enzyme catalysis of formation of a bond between the two small molecule ligands to create such a bridge is detected as activation of an essential reporter gene; whereas enzyme catalysis resulting in bond cleavage and destruction of the bridge may be detected by repression of a toxic reporter gene. The assay can be applied to new chemical reactions simply by synthesizing new small molecule substrates to bridge the DBD and AD domains.

2.3 SYSTEMS FOR SCREENING FOR CATALYTIC ACTIVITY

Several technologies have been developed to address the need for high- throughput screening of mutants for catalytic activity. As one example, methods like phage display and cell surface display have been further modified to detect catalysis by attaching not only the mutant protein, but also the substrate and product for the chemical reaction to the phage or cell surface (Pedersen et al., 1998, Proc. Natl. Acad. Sci. USA. 95: 10523-10528; Demartis et al., 1999, J. MoI. Biol. 286: 617-633; Olsen et al., 2000, Nat. Biotechnol. 18: 1071-1074). In another approach, as used by Tawfik and Griffiths, emulsion technology is used to create water droplets of crude cell extracts that encapsulate not only the protein and its encoding DNA but also multiple substrate and product molecules (Tawfik and Griffiths, 1998, Nat. Biotechnol. JJx 652-656; Aharoni et al., 2005, Chem. Biol. \2\ 1281-1289; Mastrobattista et al., 2005, Chem. Biol. Yh 1291-1300). For a review, see Cornish, 2006, Nature 440:156- 157.

2.4 THE NEED FOR MORE ACTIVE CELLULASE According to the United States Department of Energy, over 85 percent of the energy used in the United States comes from fossil fuels (April 30, 2006, www.energy.gov). A viable alternative energy source would reduce dependence on non-renewable fossil fuel resources. Ethanol derived from corn or sugarcane fermentation is the most common renewable fuel today {Id.) However, increasing energy needs can be expected to make these raw materials limited in the future. Therefore, generation of ethanol via biomass conversion is an attractive alternative. Currently, it is not cost effective to generate ethanol from biomass due to the high enzyme cost required for cellulose degradation. A ten-fold decrease in the enzyme cost, equivalent to a ten-fold increase in cellulase catalytic activity, is needed to make generation of ethanol from biomass commercially viable. Since cellulose is estimated to constitute virtually half of the biomass, increasing the specific activity of cellulases ten-fold arguably would be the most straightforward solution to the energy crisis. However, industrial efforts to generate cellulase mutants with increased catalytic activity either by rational design or by screening random mutants individually for increased activity (limited to 10⁴-10⁶ mutants even with automation techniques) have met with limited success.

3. SUMMARY OF THE INVENTION

In particular non-limiting embodiments, the present invention provides for a system for directing evolution of a biological entity of interest comprising a plurality of host cells, such that each host cell: (i) contains a nucleic acid target corresponding to the entity of interest;

(ii) contains at least one diversifying nucleic acid ("DINA") which can be used to direct mutation of the target; and

(iii) exhibits at least one mutator function which increases the likelihood that the target will be mutated by the DINA; wherein a plurality of host cells contain a plurality of different DINAs, and wherein said system further comprises a means for selecting for a mutant entity of interest target exhibiting a desired activity.

The present invention further provides for a method for directing evolution of a nucleic acid target, especially a protein, especially an enzyme, of interest which utilizes the foregoing system, whereby the plurality of host cells are subjected to conditions such that the mutator function operates to direct mutation of the target by the DINA and then host cells expressing a mutated target with altered activity are selected, with the optional step of permitting genetic exchange between genetically non-identical host cells (for example, but not limited to, sexual reproduction) followed by selection.

The present invention also provides for a primary population of host cells (a "primary library") which contains, as a result of operation of the mutator function(s) and DINAs, a diversity of mutant targets. Still further, the present invention provides for a secondary population ("secondary library") of host cells resulting from genetic exchange between two or more primary populations of cells.

Accordingly, the present invention provides for the creation of extremely large primary and secondary libraries of mutant targets of interest, by calculation having a diversity potentially of 10²⁴ and even higher, with "less fit" species of mutants being eliminated by selection, so that the selected library would contain a practically sustainable number of "more fit" species of mutants.

The entity of interest may be, for example and not by way of limitation, a protein, a functional RNA, a promoter element, an intron, a ribosome binding site, an endonuclease recognition site, or a repressor element. In other non- limiting embodiments, the entity of interest may be a metabolite, and the nucleic acid target corresponding to that metabolite may be an enzyme that produces it. Moreover, the invention may be used to essentially concurrently evolve more than one target, and therefore, in a specific non-limiting embodiment, may evolve a plurality of members of a molecular pathway that produces a metabolite of interest.

In particular, non-limiting embodiments, the system of the invention may be used to change the level or nature of the activity of a target enzyme, where the change, for example, is an increase or decrease in the activity of the enzyme, an alteration in the conditions required for enzyme activity, or a change in the substrate specificity of the enzyme.

In certain non-limiting embodiments of the invention, the basis for selection, rather than being the target itself, is the result of chemical complementation. According to such embodiments, mutation of the target results in a change in the amount of a compound which bridges DNA-binding and activation domains of a two- site receptor fusion protein and, as a consequence, a change in the amount of expression of reporter gene product, which forms the basis for selection.

In one specific non-limiting embodiment of the invention, the system comprises a yeast cell as the host cell, and the mutator function is a yeast HO endonuclease which acts to "shuffle" variant DINAs, for example in 2μ plasmids, where DINAs can replace one or more portion of a target by homologous recombination. The HO endonuclease operating on a plurality of DINAs efficiently increases the diversity of substitutions possible. The host yeast cell is then cultured under conditions which select for a desirable activity of a mutant target; for example, where the activity confers a survival advantage.

In particular, non-limiting embodiments, the present invention provides for systems for directed evolution of biomass converting enzymes, such as enzymes of interest to the energy sector.

In specific, non-limiting embodiments, the present invention provides for a system for directed evolution of cellulase which employs chemical complementation to link cellulase enzyme catalysis to cell survival. A selection step, which allows more than 10⁸ mutants to be tested (because mutants with enhanced activity have a survival advantage) increases the probability of producing a mutant with significantly enhanced cellulase activity. Indeed, the methods of the invention have generated a number of cellulase variants with approximately five-fold increases in activity compared to the parent enzymes. Accordingly, in various embodiments, the present invention provides for systems for directed evolution of cellulases (and other hyrdolases), as well as novel cellulases with enhanced activity. Directed evolution may be facilitated by cell-mediated selection as well as, in non-limiting embodiments, cell-mediated mutation and/or genetic exchange. In non-limiting embodiments, the novel cellulases of the invention may be employed in industrial processes involving cellulose breakdown, such as the generation of ethanol as an energy source.

4. BRIEF DESCRIPTION OF THE FIGURES FIGURE 1. Chemical complementation. A heterodimeric small molecule bridges a DNA-binding Domain (DBD)-receptor fusion protein and an Activation Domain (AD)-receptor fusion protein, activating transcription of a reporter gene downstream from the cognate DNA-binding site in vivo.

FIGURE 2. Cartoon of the 3-D structure of a canonical TIM barrel enzyme, taken from PDB Ia3h, prepared using Pymol (DeLano Scientific).

FIGURE 3 A-C. Extracellular N-glycans in vertebrates display three N-glycan subtypes: A. complex, B. high mannose, and C. hybrid. The figure is adapted from Grogan et al., 2002, Annu. Rev. Biochem. 71 :593-634.

FIGURE 4. Application of chemical complementation to detect glycosynthase and glycosidase activity.

FIGURE 5A-B. Chemical complementation links cellulase activity to URA3 reporter gene transcription in vivo. A. Yeast cells containing the DBD-DHFR and AD-GR fusion proteins and the URA3 reporter gene expressing either no enzyme, glycosidase (pacman) or an inactive glycosidase (circle) were grown in selective media with and without 1 μM Dex-Cel-Lac-Mtx. B. Starting with a 1 : 100 mixture of active:inactive glycosidase variants (top gel, restriction digest of 9 enzymes from independent colonies; bottom gel, restriction enzymes from the 9 fastest growing colonies after selection) the URA3 selection achieved a 200-fold enrichment in active glycosidase from a single round of selection. FIGURE 6A-B. A. Structure of the complex of the Bacillus agaradhaerens Cel5A (SwissProt: O85465; pdb: Ih5v) with methyl 4,4II,4III,4Iv- tetrathio-α-cellopentoside. The cassettes that have direct contact with the substrate are highlighted in red. Figure was prepared using Pymol (DeLano Scientific). B. Designed loop libraries from the sequences of family 5 glycosidases. Cellulase: B. agaradhaerens Cel5A (SwissProt: O85465; pdb: Ih5v); Chitosanase: Streptomyces griseus HUT 6037 chitosanase II (SwissProt: Q83VL5, 32% identity with Cel5A); Mannanse: Thermobifϊda fusca KW3 mannosidase (SwissProt: Q9ZF13; pdb: Ibqc, 14% identify with Cel5A). FIGURE 7. DERA naturally catalyzes the aldol addition reaction between acetaldehyde and D-glyceraldehyde-3 -phosphate. Evolved DERA variants with expanded substrate specificity are being sought, for example, for the synthesis of key chiral intermediate (6) in the synthesis of statins.

FIGURE 8. Detection of aldolase activity via chemical complementation. DERA should catalyze the addition of Dex-"acetaldehyde" and

Mtx-"glyceraldehyde". Spontaneous cyclization of the aldol product shouldresult in a stable Dex-Mtx aldol product that should dimerize the DBD-DHFR and AD-GR fusion proteins, activating the transcription of a downstream reporter gene.

FIGURE 9. Retro synthetic scheme of Dex-AAld and Mtx-GAld. FIGURE 10. Representation of the substrate binding pocket of the E. coli D-2-deoxyribose-5-phosphate aldolase (PDB: ljcj) showing the points of attachment for the Dex and Mtx ligands. There is a large open cavity that should readily accommodate derivatization of the 3-phosphate with Mtx. Four point mutations (L20A, Y49A, V73A and F76A) are proposed to create an opening for the linker from acetaldehyde to Dex. These positions are not implicated in catalysis and are not expected to disrupt protein fold. This figure was prepared using Pymol (DeLano Scientific).

FIGURE 1 IA-B. Simplified schematic diagram showing the effects of mutation and selection. Each rounded rectangular cell contains a plasmid which is either light grey or dark grey and an enzyme (middle grey circle missing a quadrant) having one or more mutation (identified as a light grey or dark grey quadrant). A. Mutation. Beginning with 10⁸ cells each encoding a different protein variant, if there were mutations at each round of cell division, there would be about 10¹⁶ cells containing 10¹⁶ different protein variants after a week of growth assuming doubling time of 6 hours. The problem, however, is that 10¹⁶ cells would require about

100,00OL of cell culture. B. Mutation and selection. The cell elegantly solves this problem through simultaneous mutation and selection. As illustrated here, if there were simultaneous mutations and selection at each round of cell division (cartooned as first selecting for light grey mutations, then light and dark grey mutations), simplistically after the same week of growth, there would be only 10⁸ cells containing the most fit 10⁸ protein variants, such that only the more fit of the two cells survived.

FIGURE 12A-B. In vivo mutagenesis. A. Loop libraries by homologous recombination. Active-site loops in the TIM barrel structure of the thermophilic TRPl may be mutagenized by homologous recombination of synthetic oligonucleotides encoding random loop libraries introduced via electroporation into yeast cells. B. In vivo mutagenesis with HO endonuclease. Removing the transformation step, the short oligonucleotide loop (DINA) libraries can be generated in vivo by galactose induced cleavage of a 2μ plasmid encoding the loop libraries embedded in HO endonuclease sites at each round of cell division. By tuning the levels of HO endonuclease induction, only a portion of the ca. 60 copies of 2μ plasmid present in each cell may be cleaved, allowing the loop library template plasmid to be replicated and restored at each cell division.

FIGURE 13A-B. The TRPl complementation selection distinguishes between the inactive HIS6 and the active HIS6:D127V variants in yeast. A. HIS6:D127V, but not HIS6, is able to complement a TRPl deficiency, growing on SC media lacking leucine and supplemented with only 10 μM tryptophan. B. TRPl mock selection. A 100:1 mixture of plasmid DNA encoding either inactive (His6) or active (His6:D127V) enzyme was grown under either nonselective or selective conditions to complement a TRPl deficiency. After seven days of selection, total pooled plasmid DNA was analyzed on a DNA gel following restriction digest (HIS6:D127V contains a unique restriction site). Lane 1 : lkb ladder; lane 2: selection on SC(LT-); lane 3: selection on SC(L-); lane 4: selection on SC(L-+10μM tryptophan).

FIGURE 14A-E. A. Schematic of HO-induced DNA cassette recombination to mutate the HI S 6 gene and confer ability to grow in low TRP medium, where HO enducuclease is expressed under the control of a galactose- inducible promoter. B. HIS6 recombinants in host cells induced with galactose. C. HIS6 recombinants in host cells which were not induced with galactose. D. Schematic of protocol for the steps taken to carry the HO induced DNA cassette recombination. E. DNA gel with restriction digest of the pool of HIS6 variants after HO induced recombination and TRPl selection. Lane 2: BssSI PCR digest of DNA pool induced with galactose for 30 min., Lane 3: non-disgested PCR of HIS6 DNA pool Lane 4: : BssSI PCR digest of DNA pool induced with galactose for 60min

FIGURE 15A-B. Results of an experiment to determine efficiency of mating of M ATa strain VM44 and MATa strain VM55, showing percent zygotes over time. (A) shows numerical results and (B) depicts those results in graph format. FIGURE 16. After galactose induction an equal dilution of cells is plated on selection media. Left is VM44 transformed with library 1. Right is VM45 transformed with library 2.

FIGURE 17. After mating overnight a dilution of cells is plated on selection media.

FIGURE 18. After four day selection, a dilution of cells is plated on selection media.

FIGURE 19. After sporulation and separation, a dilution of cells is plated on selection media. FIGURE 20. Growth after two days in selection medium in triplicate, the first three bars on the left representing, respectively, His D 127V, His WT, and ΔtrpF, the remaining bars showing activity of variants generated by the mutation plus mating experiments.

FIGURE 2 IA-B. A. Nucleic acid sequence of cellulase coding region of colony 1 (Cellulase 4) (SEQ ID NO:1). B. Amino acid sequence predicted from A (SEQ ID NO:3).

FIGURE 22 A-B. A. Nucleic acid sequence of cellulase coding region of Colony 2 (Cellulase 1 ) (SEQ ID NO:2). B. Amino acid sequence predicted from B (SEQ ID NO:4). FIGURE 23 A-C. A. Application of chemical complementation to detect cellulase activity. B. Starting from a 1: 100 mixture active:inactive cellulase mutants (top gel, restriction digest of 9 enzymes from independent colonies, bottom gel, restriction enzymes from the 9 fastest growing colonies after selection) the URA3 selection achieved a 200-fold enrichment in active glycosidase from a single round of selection. C. Steady state kinetic parameters of the parent cellulase and the improved mutant using p-nitrophenyl cellobioside.

FIGURE 24. Chemical complementation offers a selection for cellulase catalysis, an important target for the conversion of biomass to ethanol. FIGURE 25. Chemical complementation provides a growth selection for cellulase catalysis. In the presence of Mtx-Cel-Dex the transcriptional activator is reconstituted, activating transcription of a toxic reporter gene, thus leading to cell death. Introduction of an active cellulase which is able to cleave the β-l,4-glucosidic bond present in the cellulose linker in Mtx-Cel-Dex halts transcription of the toxic reporter gene, thus leading to cell survival. Bottom: The structure of the Mtx-Cel-Dex molecule.

FIGURE 26A-B. Reverse three-hybrid strain. A. Schematic construction of the reverse yeast-three-hybrid strain. B. The reverse three-hybrid strain was grown under URA3 counter selection conditions in the presence of 0, 1, 5, 1 OμM Mtx-Dex, and 1 μM Mtx-Cel-Dex. The presence of small molecule leads to cell death, measured as a decrease in cell density at OD600.

FIGURE 27A-C. Detection of cellulase activity and cellulase enrichment assay using the URA3 counter selection. A. The reverse three-hybrid strain is grown under counter selection conditions in the presence of empty vector, the Humicola insolens cellulase Cel7B and the inactive cellulase variant Cel7B:E197A. Under counter selection conditions, the presence of empty vector and Oμm Mtx-Cel- Dex leads to cell survival, the presence of cellulase and lμM Mtx-Cel-Dex leads to cell survival, while the presence of inactive cellulase Cel7B:E197A in the presence of lμM Mtx-Cel-Dex leads to cell death. Finally, presence of empty vector in the presence of iμM Mtx-Dex leads to cell death. B. Starting from a 1:100 mixture of active inactive cellulase mutants, the URA3 counter selection achieved a 200-fold enrichment in active cellulase from a single round of selection. Lane 1, active cellulase control. Top gel, lanes 2-10, restriction digest of 9 enzymes from independent colonies isolated before selection. Bottom gel, lanes 2-10, restriction digest of 9 independent colonies isolated after selection. C. Twenty-two cellulase variants isolated before selection and twenty-two variants isolated after five days of selection were tested for aldehyde formation using carboxymethylcellulose (CMC) in crude cell extracts. Aldehyde formation was measured as absorption at 660nm. The mean activity of cellulase variants isolated before and after 5 days of selection is plotted. The error bars represent the variance from the mean. The cellulase activity on CMC from cellulase variants isolated after 5 days of selection is higher than that of cellulases isolated before selection. This difference is statistically significant (p<0.005).

FIGURE 28A-B. Homology model of Cellulase 1 (CELLV AR2, see Table IA and Cel_5. (Cellulase 2 in Table IA) using B. agaradharens Cel5A (PDB: 1H5V). Left. Cellulase 1 (CELLVAR2) differs by nine mutations (pink), when compared to the most active starting cellulase, CeIN. Right. Cel5.7 (Cellualse 2) differs by eight mutations (pink), when compared to the CeIN.

FIGURE 29. Substrates of enzymes useful in biomass conversion.

5. DETAILED DESCRIPTION OF THE INVENTION For clarity of disclosure, and not by way of limitation, the detailed description of the invention is divided into the following subsections: (i) host cells;

(ii) targets; (iii) mutator functions; (iv) diversifying nucleic acids; (v) genetic exchange; (vi) selection;

(vii) libraries;

(viii) exemplary in vivo directed evolution systems; (ix) exemplary enzymes for evolution;

(x) directed evolution systems for improving enzymes for biomass conversion; and

(xi) cellulase variants.

5.1 HOST CELLS

Systems of the invention comprise a host cell containing a nucleic acid target corresponding to a biological entity of interest. In particular, non-limiting embodiments (see below) the target may be a gene ("target gene") encoding a protein of interest ("target protein"), operably linked to a promoter element.

The host cell may be a prokaryotic or eukaryotic cell, including, but not limited to, a yeast cell, a bacterial cell, an insect cell, a plant cell, a unicellular slime mold cell, an invertebrate cell (e.g., a worm cell), a vertebrate cell and/or a mammalian cell. In specific, non-limiting embodiments of the invention, the yeast cell is a yeast that contains a HO-endonuclease, for example Saccharomyces cerevisiae, Candidal glabrata, Kluveromyces delphiensis, or Saccharomyces castellii. In non-limiting embodiments of the invention, and in conjunction with the ability to select for desirable mutant targets, the host cell has a phenotype that provides a selection mechanism. The phenotype may be manifested by native host cells ("endogenous phenotype") or may be produced by genetic engineering or similar artificial means ("exogenous phenotype").

For example, the host cell may have a nutritional requirement which the desirable mutant target directly or indirectly provides, or a particular compound may be toxic to the host cell, and the desirable mutant target may directly or indirectly alleviate said toxicity. In the foregoing, "directly" means that the mutant target itself is acting, and "indirectly" means that the mutant target is causing another agent to act. As a non-limiting example of indirect action, the mutant target may produce a small molecule bridging compound that links DNA-binding and activation domains in a chemical complementation assay, thereby producing a reporter gene which satisfies a nutritional requirement or degrades a toxic protein (or the converse if the mutant target cleaves a bridging molecule). In one specific, non-limiting embodiment, where the host cell is a yeast cell, the reporter gene may be URA3, which encodes a toxic product.

The host cell further comprises at least one diversifying nucleic acid ("DINA"). Said one or more DINA may be exosomal or may be integrated into genomic DNA. The DINA may be introduced into the host cell by any means known in the art, including, but not limited to, transformation, transfection, electroporation, microinjection, etc.. In particular non-limiting embodiments of the invention, one or more DINA may be introduced into a parental host cell, which is then propagated to produce a population of progeny host cells containing the DINA(s); in some instances, the DINA(s) may have undergone mutation between cellular generations.

In certain non-limiting embodiments, the host cell may further comprise elements supporting chemical complementation, for example, (i) a DNA binding site linked to a reporter gene; (ii) a nucleic acid encoding a receptor for a first small molecule linked to a DNA binding domain (DBD) which specifically binds to the DNA binding site, operably linked to a promoter element; and (iii) a nucleic acid encoding a receptor for a second small molecule linked to an activation domain, operably linked to a promoter element. Said promoters may be distinct or shared. The first small molecule, second small molecule, or a molecule comprising the first small molecule and second small molecule connected by a linker (e.g., a substrate bridge) may be introduced into the cell or may be generated by the target and/or mutants thereof. The first small molecule joined to the second small molecule forms the basis for association between DNA binding and activation domains of a fusion protein and is also referred to herein as a "bridging compound."

5.2 TARGETS

The present invention provides for a means for directing evolution of an entity of interest via a nucleic acid (DNA or RNA, preferably DNA) target corresponding to that entity. "Corresponding" refers to the relationship between nucleic acid and the target, as can be seen below. Suitable entities of interest(which can be molecules found in nature (wild-type) or naturally occurring or engineered mutants thereof) include, but are not limited to: a protein (especially an enzyme), where the corresponding target is a target gene encoding said protein or a portion thereof; a promoter, where the target is the promoter itself or a portion thereof; a functional RNA, where the target is the RNA itself or a DNA which may be transcribed to produce the functional RNA, or a portion thereof; an intron, where the target is the DNA or RNA version of the intron, or a portion thereof, a ribosome binding site, where the target is the DNA or RNA version of the ribosome binding site, or a portion thereof, an endonuclease recognition site, where the target is the DNA site or a portion thereof; a repressor element, where the target is the repressor element or a portion thereof; a metabolite; or a plurality or combination of the above, where the ultimate entity of interest is a product of their association (e.g. the entity of interest is a metabolite and the target is one or more component of a pathway that generates the metabolite).

The size of the target may be understood to be dictated by the nature of the target. "A portion thereof is defined herein to mean a region which represents less than 100 percent, less than 90 percent, less than 80 percent, less than 70 percent, less than 60 percent, less than 50 percent, less than 40 percent, less than 30 percent, less than 20 percent, less than 10 percent, or less than 5 percent of the size of the target. Examples of functional RNAs are provided in Alper et al., 2006, Science 314:1565-1568; Buskirk et al., 2003, Chemistry & Biology 10:533-540; and Buskirk et al., 2004, Chemistry & Biology l±: 1157-1163).

The nucleic acid target is alternatively referred to herein as a target gene. Where the target gene is not native to the host cell, it may be referred to as a heterologous target gene. The target gene may comprise a region, or "replaceable segment", which is desirably mutated by the DINA. The replaceable segment may optionally comprise a region which is engineered to facilate mutation by the DINA, for example, but not by way of limiation, via the incorporation of a homologous nucleic acid sequence and/or an endonuclease site.

Examples of metabolites which may be entities of interest include but are not limited to the antimalarial drug precursor artemisinic acid (Ro et al., 2006. Nature 440:940-943; hydrocortisone (Szczebara et al., 2003, Nature Biotechnol. 2J_: 143-149; resveratrol (Beekwilder et al., 2006, Appl. Environmental Microbiol. 72:5670-5672; and beta-carotene and lycopene (Yamano et al., 1994, Biosci. Biotechnol. Biochem. 58:1 1 12-1114).

In particular non-limiting embodiments, the target is a protein, especially an enzyme. It should be understood, however, that although the examples present herein usually refer to protein targets, analogous techniques may be used to evolve other classes of entites of interest (targets) as discussed above.

Virtually any protein may serve as the target in the present invention, provided that such protein may directly or indirectly (as described herein) provide a potential basis for selection. A protein of interest is alternatively referred to herein as the "target protein." The target protein may be a specific enzyme, the activity of which is desirably increased, decreased, or otherwise modified. The target gene of the target protein may encode a protein scaffold which is common to a number of enzymes, having one or more segment that may be replaced ("replaceable segment") without compromising the structure and/or stability of the enzyme, optionally engineered to comprise a marker sequence which facilitates its replacement (e.g., a sequence homologous to a portion of a DINA and/or an endonuclease specific recognition element). In non-limiting embodiments of the invention, the size of the replaceable segment may be between about 4 and 75 amino acids (between 12 and about 300 nucleotides in the corresponding target gene), or between about 5 and 10 amino acids (between 15 and 30 nucleotides in the target gene), or between about 5 and 25 amino acids (between 15 and 75 nucleotides in the target gene), or between about 10 and 50 amino acids (between 30 and 150 nucleotides in the target gene).

Homology, as that term is used herein, refers to a sequence which is at least about 75, at least about 80, at least about 85, at least about 90, at least about 95 or at least about 98 percent homologous relative to a comparison sequence, as may be determined by standard software such as BLAST or FASTA. A homologous region of a nucleic acid may be at least about 8, at least about 10 nucleotides, at least about 15 nucleotides, at least about 20 nucleotides, at least about 25 nucleotides, at least about 30 nucleotides, at least about 40 nucleotides, or at least about 50 nucleotides, in length. In one preferred non-limiting embodiment, the present invention utilizes as a target protein a TIM barrel scaffold. Over 10% of enzymes whose structures are deposited in the protein data bank (PDB) have the TIM barrel scaffold (Hocker, 2005, Biomol. Eng. 22:31-38). TIM barrel enzymes are promiscuous in their activity, catalyzing five of the six Enzyme Commission (EC) reaction classes — oxidations and reductions, transferase reactions, lyase reactions, hydrolase reactions, and isomerase reactions (Sterner and Hocker, 2005, Chem. Rev. 105:4038-4055). As illustrated in FIGURE 2, the classic TIM barrel structure is comprised of eight internal, parallel β-strands connected by loops to eight surrounding α-helices (Gerlt and Babbitt, 2001, Annu. Rev. Biochem. 70:209-246). Although any one of the eight loops may comprise or constitute a replaceable segment, because structural integrity appears to be maintained by the N-terminal half of the barrel and residues important for substrate binding and catalytic activity are found in the C-terminal half of the barrel in the loops and at the ends of the β-strands, loops in the C-terminal half of the barrel are preferred locations for replaceable segments. Randomization of the β- strand termini and loops may be used to support new or altered enzyme functions without, in theory, impairing the structural integrity of the protein. The sequence, structure, and function of TIM barrel enzymes are well characterized (Gerlt and Raushel, 2003, Curr. Opin. Chem. Biol. 7:252-264). In particular non-limiting embodiments, a subclass of TIM barrel may be used as the target gene, for example, but not by way of limitation, a TIM barrel glycosidase or a TIM barrel aldolase.

The target gene may be operably linked to a suitable promoter, which may be an endogenous host cell promoter or a promoter introduced by genetic engineering, such as a heterologous promoter (a promoter not associated with the target gene in nature). The promoter may be an inducible promoter, or may be constitutively active.

A host cell may comprise more than one target, where each target may be the same or may be different (for example, members of a molecular pathway). 5.3 MUTATOR FUNCTIONS

The host cells of the invention comprise one or more mutator function. At least one mutator function increases the likelihood that the target will be mutated by the DINA. An additional optional, "ancillary" mutator function may increase the diversity of potential mutations.

The mutator function may be native to the host cell ("endogenous mutator function") or may be engineered ("exogenous mutator function"). One non-limiting example of a mutator function is homologous recombination (reciprocal recombination).

Another non-limiting example of a mutator function is a function that facilitates homologous recombination (e.g., the formation of double strand breaks by the HO endonuclease of yeast; other homing endonucleases may be used to provide said mutator function; see, for example, Belfort & Roberts, 1997, Nucl. Acids Res. 25(17^379-3388I

Yet another non-limiting example of a mutator function is gene conversion (non-reciprocal recombination).

A non-limiting example of an ancillary mutator function is a tendency to commit replication errors such as is produced by an error-prone polymerase.

Another non-limiting example of an ancillary mutator function is the ability to decrease expression of a DNA damage-binding (DDB) protein (e.g., a repressible promoter controlling the expression of a DDB-encoding gene).

Another non-limiting example of an ancillary mutator function is a transposable element. A mutator function or ancillary mutator function may occur as a result of interaction of nucleic acids or as a result of interaction between one or more nucleic acid and one or more protein.

In non-limiting embodiments of the invention, the mutator function or ancillary mutator function is executed by one or more protein, wherein the expression of at least one such protein is controlled by an operably linked promoter element. Said promoter may be constitutively active or may be inducible or otherwise conditionally active (e.g., can be inhibited by binding of a repressor element). Most preferably, the promoter permits "titration" of the amount of mutator function operating. In a specific, non-limiting embodiment, the promoter is the Gall promoter.

In various specific non-limiting embodiments of the invention, the mutator function(s) comprise(s) (i) a gene encoding an HO endonuclease operably linked to a promoter element, wherein the endonuclease product of this gene acts at HO-specific recognition sequences flanking one or more diversifying nucleic acid present in the host cell genome or (preferably) in an episome, and/or at one or more HO-specific recognition sequence comprised in a replaceable segment of the target gene; (ii) a gene encoding a l-Scel gene operably linked to a promoter element, wherein the product of this gene acts at one or more l-Scel site -specific recognition sequence in the target gene (for example, using the delitto perfetto system described in Storici et al., 2003, Proc. Natl. Acad. Sci. U.S.A. JOO: 14994-14999); or (iii) genes encoding Ragl, Rag2, and optionally terminal deoxynucleotidyltransferase (TdT), which act at specific recognition sequences flanking one or more replaceable segment within the target gene and/or one or more diversifying nucleic acid sequence in the host cell. 5.4 DIVERSIFYING NUCLEIC ACIDS

As set forth above, a diversifying nucleic acid ("DINA") may be comprised in the host cell genome or in an episomal sequence {e.g., a 2μ plasmid). For example, one or more DINA may be comprised in the host cell genome outside of the target gene, or maintained in episomal form. In various non-limiting embodiments of the invention, a DINA comprises or is flanked by (i) a region homologous to one or more target gene sequence (so as to facilitate recombination) (e.g. , is homologous to a replaceable segment) and/or (ii) a recognition sequence recognized by an enzyme that promotes recombination (e.g., an endonuclease, for example HO endonuclease, Ragl or Rag2). A diversity of DINAs, referred to herein as a "DINA library" may be generated and provided to a population of host cells. For example, a diverse DINA library may be generated by standard chemical or genetic engineering techniques (e.g., error prone PCR), and may be introduced into a host cell either directly (e.g., by transformation) or via insertion into plasm ids, phage, virus, or a transposable element. Where such a library comprises diverse loops for introduction into a protein scaffold, it is referred to as a "loop library."

A DINA may be approximately commensurate in length relative to the portion of the target intended to be modified (for example, the "replaceable segment", e.g., a "loop" in a protein scaffold), may be smaller or may be larger (for example, comprising sequence homologous to a region outside of the replaceable segment).

More than one DINA may be used to replace or be inserted into a replaceable segment of the target gene. Accordingly, the length of a DINA may be between about 12 and about 500 nucleotides, or between about 9 and 15 nucleotides, or between about 15 and 30 nucleotides, or between about 15 and 75 nucleotides, or between about 30 and 150 nucleotides, or between about 50 and 200 nucleotides. The DINA may comprise a portion homologous to a target sequence and/or replaceable segment of the target gene or a portion homologous to a region of the target gene which is not part of a replaceable segment, where said portions may be at least about 8, at least about 10 nucleotides, at least about 15 nucleotides, at least about 20 nucleotides, at least about 25 nucleotides, at least about 30 nucleotides, at least about 40 nucleotides, or at least about 50 nucleotides in length; the homologous portion may promote homologous recombination with the target gene. Accordingly, the DINA itself may be sufficiently homologous as to promote a homologous recombination or gene conversion event.

5.5 GENETIC EXCHANGE

According to certain non-limiting embodiments of the invention, genetically non-identical host cells are permitted to engage in genetic exchange between their genomes, under conditions in which homologous recombination and/or gene conversion can occur. Genetic exchange may be achieved, for example and not by limitation, by sexual reproduction or cell fusion followed by division (e.g., protoplast fusion or plasmodia formation). Genetic exchange in an environment that provides a selection mechanism (see below) maintains an extremely diverse "virtual" library (by calculation, for example, having 10²⁴members) to manageable size for culture. Where the present invention is used to produce a diversified plurality

("library") of host cells, genetic exchange may occur between members of the library. Alternatively, a plurality of libraries may be prepared separately and then combined to allow genetic exchange to occur between members of different libraries. 5.6 SELECTION

Selection, according to the invention, may be achieved by any method known in the art. As stated above, selection may operate directly upon the activity of the target or mutant target or indirectly, for example via chemical complementation or via the product of a gene downstream from the protein of interest (or mutants thereof) in a functionally-linked pathway. Chemical complementation is described, for example, in United States Patent Application Publication No. US 2004/0106154 by Cornish. Without limitation, selection may be made based on, for example: (i) the ability of a host cell to thrive in a particular nutritional context {e.g., medium lacking a particular anabolic component); (ii) the ability of a host cell to survive in the presence or absence of a particular agent (e.g., the presence of a conditional toxin such as neomycin or hygromycin B); (iii) host cell volume; (iv) the production of a detectable product (e.g., a fluorescent protein such as green fluorescent protein as an antigen "tag"); (iv) host cell size; (v) host cell mobility; (vi) host cell invasiveness; (vii) host cell adhesiveness, etc..

5.7 LIBRARIES The term "library" as used herein refers to a collection of host cells containing a plurality of mutated versions of the target protein. The "diversity" may be quantitated as the number of individual host cells estimated to be required to represent all the different mutant target proteins expressed by the library.

Libraries may be categorized as "primary libraries" which are the result of operation of the invention and where genetic exchange may or may not occur between members of the library, or "secondary libraries" which are the result of allowing or promoting genetic exchange between two or more primary libraries prepared according to the invention.

For example, a primary library size equal to the product of the transformation efficiency and the number library variants capable of being produced by the DINA(s) and mutator function(s) in one cell may be generated, and a secondary library having a size equal to the product of the sizes of the parent primary libraries undergoing genetic exchange may be generated.

As a non-limiting example of how diversity in a primary library may be expanded according to the present invention, if one cell has one plasmid containing 10 discrete DINAs, each flanked with HO sites, each mutating one or more amino acids at different replaceable segments of a target protein, the target gene in that cell could be mutated in 10²³ different ways because each targeted region of DNA could either remain wild type or mutate; because there are two states that a targeted region of DNA can be in and 10 DINAs, there would be 2 ' ° = 10²⁴ different combinations of wild type and mutated regions (one of which is everything stays wild type, hence a possible 10²³ mutants). In order to realize all 10²³ combinations, a parent cell would theoretically need to divide to generate 10²³ different progeny cells, each carrying the DINA-containing plasmid. Accordingly, if 10⁸ cells are transformed with different plasmids having 10 unique DINAs, then those cells can be grown and induced to create a library of up to 10³ x 10⁸ (~10* '). As a non-limiting example of how diversity in a secondary library may be generated, if three groups of cells each containing a (primary) library of 10 generate a secondary library through genetic exchange, a diversity of up to 10⁸ x 10⁸ x 10⁸ (=10²⁴) in the resulting secondary library may be generated. Moreover, this method naturally lends itself to searching sequence space via genetic algorithms (i.e. by performing simultaneous mutagenesis and selection), so the potential library can be efficiently searched by using far fewer actual (selected) library members

5.8 EXEMPLARY IN VIVO DIRECTED EVOLUTION SYSTEMS

The present invention provides for a system for directing evolution of a entity of interest, for example a protein, especially an enzyme, of interest ("the target protein"), comprising a plurality of host cells, such that each host cell: (i) contains, in its genome, a nucleic acid corresponding to the entity of interest {e.g., encoding the target protein) ("the target gene"), optionally operably linked to a promoter element;

(ii) contains at least one diversifying nucleic acid ("DINA") which can be used to direct mutation of the target gene; and (iii) exhibits at least one mutator function which increases the likelihood that the target gene will be mutated by the DINA; wherein a plurality of host cells contain a plurality of different DINAs, and wherein said system further comprises a means for selecting for a mutant entity of interest exhibiting a desired activity. In one set of non-limiting embodiments, the present invention provides for a system for directing evolution of an entity of interest {e.g. a target protein) comprising a plurality of host cells, such that each host cell:

(i) contains, in its genome, a nucleic acid corresponding to the entity of interest {e.g., encoding the target protein) ("the target gene"), optionally operably linked to a promoter element; (ii) contains at least one diversifying nucleic acid ("DINA") which can be used to direct mutation of the target gene;

(iii) exhibits at least one mutator function which increases the likelihood that the target gene will be mutated by the DINA; (iv) a DNA binding site linked to a reporter gene;

(v) a nucleic acid encoding a receptor for a first small molecule linked to a DNA binding domain that specifically binds to the DNA binding site, operably linked to a promoter element; and

(vi) a nucleic acid encoding a receptor for a second small molecule linked to an activation domain, operably linked to a promoter element, wherein, whether or not the DNA binding domain and the activation domain are joined (to cause expression of the reporter gene) is related to mutation of the target gene; wherein a plurality of host cells contain a plurality of different DINAs; and wherein said system further comprises a means for selecting for a mutant entity exhibiting a desired activity based on the product of the reporter gene. For example, where the entity of interest is a protein, if the target gene is mutated in vivo to form a mutant protein with bond-forming activity, and a first small molecule and a second small molecule are provided to said host cell, the mutant protein links the first small molecule and the second small molecule, causing expression of the reporter gene. Alternatively, where the entity of interest is a protein, if the target gene is mutated in vivo to form a mutant protein with bond-cleaving activity, and a first small molecule linked to a second small molecule is provided to said host cell, the mutant protein cleaves the first small molecule from the second small molecule, thereby inhibiting or preventing expression of the reporter gene. In a further non-limiting embodiment, the present invention provides for a system of directing evolution of an entity of interest (e.g., a target protein) comprising a plurality of host cells, such that each host cell:

(i) contains, in its genome, a nucleic acid corresponding to the entity of interest (e.g., encoding the target protein) ("the target gene"), optionally operably linked to a promoter element;

(ii) contains at least one diversifying nucleic acid ("DINA") comprising a flanking pair of endonuclease specific recognition sequences, which can be used to direct mutation of the target gene; and (iii) exhibits at least one mutator function which increases the likelihood that the target gene will be mutated by the DINA; wherein the mutator function is an endonuclease which cleaves at the specific recognition sequences, and wherein a plurality of host cells contain a plurality of different DINAs, and wherein said system further comprises a means for selecting for a mutant entity of interest exhibiting a desired activity.

In a still further non-limiting embodiment, the present invention provides for a system of directing evolution of an entity of interest (e.g., a target protein) comprising a plurality of host cells, such that each host cell:

(i) contains, in its genome, a nucleic acid corresponding to the entity of interest (e.g., encoding the target protein) ("the target gene"), optionally operably linked to a promoter element;

(ii) contains at least one diversifying nucleic acid ("DINA") comprising a flanking pair of endonuclease specific recognition sequences, which can be used to direct mutation of the target gene; (iii) exhibits at least one mutator function which increases the likelihood that the target gene will be mutated by the DINA; wherein the mutator function is an endonuclease which cleaves at the specific recognition sequences, and wherein a plurality of host cells contain a plurality of different DFNAs; (iv) a DNA binding site linked to a reporter gene;

(v) a nucleic acid encoding a receptor for a first small molecule linked to a DNA binding domain that specifically binds to the DNA binding site, operably linked to a promoter element; and

(vi) a nucleic acid encoding a receptor for a second small molecule linked to an activation domain, operably linked to a promoter element, wherein, whether or not the DNA binding domain and the activation domain are joined (to cause expression of the reporter gene) is related to mutation of the target gene; wherein a plurality of host cells contain a plurality of different DINAs; and wherein said system further comprises a means for selecting for a mutant entity exhibiting a desired activity based on the product of the reporter gene. For example, where the entity of interest is a protein, if the target gene is mutated in vivo to form a mutant protein with bond-forming activity, and a first small molecule and a second small molecule are provided to said host cell, the mutant protein links the first small molecule and the second small molecule, causing expression of the reporter gene. Alternatively, where the entity of interest is a protein, if the target gene is mutated in vivo to form a mutant protein with bond-cleaving activity, and a first small molecule linked to a second small molecule is provided to said host cell, the mutant protein cleaves the first small molecule from the second small molecule, thereby inhibiting or preventing expression of the reporter gene. The present invention further provides for methods of using said systems to produce libraries, and the primary and secondary libraries so produced.

5.9 EXEMPLARY ENZYMES FOR EVOLUTION For example and not by way of limitation, the systems and methods of the present invention may be applied toward directed evolution of the following families of enzymes (and see section 5.10 below, relating to enzymes useful in biomass conversion).

5.9.1. DIRECTED EVOLUTION OF GLYCOSIDASE/GLYCOSYNTHASES Despite their fundamental role in biological processes and potential use as therapeutics, it still remains difficult to synthesize carbohydrates, glycoproteins, and glycolipids. In the past two decades, there has been tremendous progress in the chemical synthesis of complex carbohydrates (Paulsen, 1982, Angew Chem. Int. Ed. Engl. 21_, 155-173; Plante, et al., 1999, Org. Lett., I, 211-214; Hashimoto, et al., 1989, J. Chem. Soc. Chem. Commun., 685-687; Schmidt, R.R. and Michel, J., 1980, Angew. Chem. Int.Ed. Engl. \9_, 731-732; Schmidt, 1986, Angew. Chem., 98, 213- 236; Fraser-Reid, et al., 1992, Synlett, 927-942; Fraser-Reid, et al., 1988, J. Chem. Soc. Chem. Commun., 823-825; Kahne, et al., 1989, J. Am. Chem. Soc, JJl, 6881- 6882; Toshima and Tatsuta, 1993, Chem. Rev. 93, 1503-1531) (Seeberger and Danishefsky, 1998, Ace. Chem. Res. 3_i, 685-695; Danishefsky, et al., 1993, Science, 260, 1307-1309) However, chemical synthesis is still quite demanding other than for a handful of laboratories most skilled in the art. Natural glycosyltransferase and glycosidase enzymes are also used for carbohydrate synthesis (Flitsch, 2000, Curr. Opin. Chem. Biol. 4, 619-625; Koeller, and Wong, 2001, Nature, 409, 232-240). However, the human enzymes are often required, and many of these are difficult to express recombinantly, with several not even commercially available (Crout and Vic, 1998, Curr. Opin. Chem. Biol., 2, 98-11 1; Sears and Wong, 2001, Science, 29J_., 2344-2350; Thiem, J., 1995, FEMS Microbiol. Rev., 16, 193-21 1; Nilsson, 1988, Trends Biotechnol. 6, 256-264; Hamilton, 2004, Nat. Prod. Rep. 21, 365-385). Therefore, it is desirable to use directed evolution to develop a family of well behaved TIM barrel enzymes that could be used for the routine synthesis of carbohydrates and their conjugates for biomedical research by laboratories not specializing in carbohydrate synthesis and perhaps even for large-scale production of therapeutics. In a first set of non-limiting embodiments, the present invention provides for the development of a family of "glycosynthase'Vglycosidase pairs that can be used for the synthesis of ΛMinked glycans (FIGURE 3 A-C). ΛMinked glycosylation of proteins occurs in the Endoplasmic Reticulum and Golgi Apparatus and is integral to protein trafficking via the secretory pathway and serves as an important extracellular ligand for cell-cell communication (Helenius and Aebi, 2001, Science, 29JL, 2364-2369; Lechner and Wieland, 1989, Annu. Rev. Biochem., 58,

173-194; Kornfeld and Kornfeld, 1985, Annu. Rev. Biochem. 54, 631-664; Grogan, et al., 2002, Annu. Rev. Biochem. 71, 593-634; Essentials of glycobiology, eds. Varki, et al., Cold Spring Harbor Laboratory Press Cold Spring Harbor, NY, 1999). A core oligosaccharide is transferred to the side chain of Asn from a lipid precursor. This core ΛMinked glycan is then built-up through differential cleavage, modification, and further glycosylation to produce ΛMinked glycans of diverse structure (Kornfeld and Kornfeld, 1985, Annu. Rev. Biochem., 54, 631-664). Because of the complexity of ΛMinked glycan structure and the difficulty of reconstituting the lipid precursor and human enzymes for in vitro synthesis, ΛMinked glycoproteins have proven particularly challenging to synthesize. Currently ΛMinked glycoproteins are synthesized recombinantly (and in impure form) in mammalian cell lines (Grabenhorst, et al., 1999, Glycoconj. J., \6, 81-97). Partial reconstitution of the human glycosylation pathway in yeast shows promise (Hamilton, et al., 2003, Science, 301, 1244-1246) (the yeast pathway produces different glycosylation patterns), but also suffers from heterogeneity in the final product. Glycopeptides can be prepared by synthetic chemistry — often in concert with recombinant enzymes — or by in vitro reconstitution of the transglycosylation reaction using Endo-M with subsequent enzymatic elaboration (Sears and Wong, 2001, Science, 291, 2344-2350; Grogan, et al., 2002, Annu. Rev. Biochem. 71, 593-634; Live, et al., 2001, Org. Lett. 3, 851-854; Unverzagt, C, 1996, Angew Chem. Int. Ed. Engl. 35, 2350-2353;

O'Connor, et al., 2001, J. Am. Chem. Soc, 123, 6187-6188; Haneda, et al., 1998, Bioorg. Med. Chem. Lett. 8, 1303-1306). These methods, however, remain sufficiently challenging, so that the state of the art cannot routinely synthesize N- linked glycoproteins with defined structural variations to study the role of //-linked glycosylation on myriad biological pathways. Erythropoietin, for example, which has three N-linked and one O-linked glycans, is used therapeutically with heterogeneous glycosylation, even though this heterogeneity is known to affect its therapeutic efficacy (Fukuda, et al., 1989, Blood, 73, 84-89). With robust chemical methods for coupling the first carbohydrate as the glycosylamine to Asp, in various embodiments, the present invention provides for engineering a family of designer enzymes that allow ΛMinked glycans of diverse structure to then be built-up one sugar at a time. Alternatively, these designer enzymes could be used to selectively cleave and then build back up from a core oligosaccharide transferred using Endo-M. The genesis of such as family may begin with directed evolution of a glycosynthase/glycosidase pair for the β-l,4-Man-GlcNAc linkage found in the pentasaccharide core of all JV-linked glycans (Grogan, et al., 2002, Annu. Rev. Biochem., Jl, 593-634; Essentials of glycobiology, eds. Varki, et al.,Cold Spring Harbor Laboratory Press Cold Spring Harbor, NY, 1999). Without favorable neighboring group participation from the 2'- hydroxyl group of mannose, this linkage cannot be synthesized directly in high yield with existing chemical methods. There are no recombinant glycosyltransferases of glycosidases commercially available for the synthesis or selective hydrolysis of this linkage.

Chemical complementation was adapted as a selection for glycosynthase activity (FIGURE 4), starting with the published Cel7B:E197A glycosynthase variant from Humicola insolens. The effectiveness of chemical complementation in this context was experimentally determined (see Section 8, below).

Directed Evolution of a β-l,4-Man-GlcNAc Glycosynthase/Glycosidase Pair Directed evolution may be carried out via chemical complementation to modify the substrate specificity of a glycosynthase and a glycosidase for the synthesis and hydrolysis, respectively, of the β-l,4-Man-GlcNAc linkage found in the core of N- 1 inked glycans. As the protein scaffold, family 5 glycosidases — monomeric TIM barrel enzymes, with known specificity for β-l,4-linked oligomers of GIu, Man, and GIcNAc, may be used. Chemical complementation may be adapted for β-1 ,4-Man- GIcNAc substrate specificity simply through the synthesis of the necessary Mtx-Man- F donor, DeX-(GIcNAc)₂ acceptor, and Mtx-Man-(GlcNAc)₂-Dex trisaccharide substrates. Directed evolution, for example, but not by way of limitation, from TIM barrel loop libraries may then be performed using yeast LEU2 and URA3 5-FOA growth selections. The activity and substrate specificity of evolved protein variants may then be characterized in vitro. Further, the best β-l,4-Man-GlcNAc glycosynthase and glycosidase variants may then be incorporated into a synthesis of an ΛMinked glycopeptide.

Substrate Synthesis. Chemical complementation may be adapated to detect synthesis of the β-l,4-Man-GlcNAc linkage with the^' LEU2 selection using DeX-(GIcNAc)₂ and Mtx-Man-F as substrates. With subsequent enzymatic synthesis of Dex-(GlcNAc)₂-Man-Mtx, chemical complementation may then be used to detect hydrolysis of the β-l ,4-Man-GlcNAc linkage using the URA3 counter selection. Man monosaccharide may be used as the α-fluoro donor because ultimately it may be desirable to add the Man monosaccharide to the GlcNAc-GlcNAc disaccharide in the ΛMinked glycan core. Use of the monosaccharide may also simplify the synthesis of the substrate. If need be, however, one could start with a disaccharide Glu-Man-F donor and then subsequently modify it to accept a monosaccharide Man-F donor.

TJM Barrel Libraries. Family 5 endo-glycosidases may be used as the protein scaffold for these directed evolution experiments (Henrissat and Bairoch, 1993, Biochem. J., 293 (Pt.3\ 781-788; Henrissat, 1991, Biochem. J.. 280 (Pt. 2). 309-316) (FIGURE 6A). Family 5 glycosidases are monomeric TIM barrel enzymes, and the endo-type enzymes in this family should readily incorporate the Dex and Mtx CIDs required for chemical complementation. The sequences of over 500 family 5 glycosidases are known. There are 157 known cellulases, with 10 different high- resolution structures in the PDB; 24 mannosidases, with 5 high-resolution structures; and 1 chitosonase (Coutinho and Henrissat, 1999). The in vitro activities and substrate specificities of several of these family 5 glycosidases have been extensively characterized (Coutinho and Henrissat, 1999). Chemical complementation may be used to "clone" a new glycosynthase with β-l,4-Man-GlcNAc synthesis activity simply by screening active- site GIu to GIy variants of these known family 5 glycosidases using chemical complementation. Activity of this cloned β-1 ,4-Man-GlcNAc glycosynthase TIM barrel may be improved by randomizing the β-strand termini and loops involved in substrate recognition and catalysis. For illustration, FIGURE 6B presents the loop libraries the could be used in the B. agaradhaerens Cel5A (O85465, Ih5v) based on alignment with the S. griseus chitosanase II (Q83VL5, wild-type enzyme shows weak transglycosylation activity to accept (GLcN Ac)3 as the acceptor substrate) and T. fusca mannosidase (Q9ZF13, lbqc). The design of these libraries is based largely on recent successes in the de novo directed evolution of binding proteins (Beste, et al., 1999, Proc. Natl. Acad. Sci., USA, 96, 1898-1903; Binz, et al., 2004, Nat. Biotechnol., 22, 575-582; Silverman, et al., 2005, Nat. Biotechnol., 23, 1558-1561) and changes in substrate specificity with traditional genetic assays and extensive characterization of the TIM barrel scaffold (Dwyer, et al., 2004, Science, 304, 1967- 1971). A "privileged scaffold" is used for enzyme catalysis. The libraries are focused on the loops that directly contact the substrate and do not disrupt the structural integrity of the protein scaffold. Very diverse libraries of random sequence may be used, even though these libraries vastly exceed the amount of sequence space that can be searched even with a selection.

Chemical Complementation Selections. LEU2 selection may be used for glycosynthase activity and URA3 5-FOA may be used for counter selection for glycosidase activity. The stringency of both selections can be readily tuned, simply by varying the concentration of leucine or 5-FOA in the yeast media, to select for improvements in activity. False positives may be ruled out simply by carrying out a secondary three-hybrid screen with and without small-molecule substrate.

Characterization of Evolved Protein Variants. The activity of successful clones may be confirmed in vitro with purified enzyme based on confirmation of the product by HPLC purification and LR-MS and ¹H NMR and specific activity measurements (Lin, et al., 2004, J.Am. Chem. Soc, 126, 15051- 15059). It may also be desirable to carry out further NMR experiments to confirm the regio- and stereochemistry of the products. Michaelis-Menten parameters may be determined by measuring fluoride anion release in the case of the glycosynthases and by measuring the release ofp-nitrophenol from a tetrasaccharide linked to nitrophenol for the glycosidases (Lin., et al., 2004, J. Am. Chem. Soc, 126, 15051-15059; Ly and Withers, 1999, Annu. Rev. Biochem. 68, 487-522).

5.9.2 DIRECTED EVOLUTION OF ALDOLASES The aldol reaction is one of the key carbon-carbon bond forming reactions in organic synthesis (Evans, et al., 1982, Topics in Stereochemistry, 1-1 15; Mahrwald, 1999, Chem. Rev., 99, 1095-1 120; Heathcock, 1981, Comprehensive Organic Synthesis, 133; Heathcock, 1991, Comprehensive Organic Synthesis, 181). Aldolases, natural enzymes that catalyze this reaction, have emerged as important tools for aldol synthesis, particularly on the industrial scale (Henderson and Toone, 1999, Comprehensive Natural Product Chemistry, 367-441) (Wymer and Toone, 2001, Curr. Opin. Chem. Biol., 4, 110-1 19) (Gijsen, et. al., 1996, Chem. Rev. 96, 443- 474). Aldolases typically catalyze aldol addition reactions with defined regio- and stereo-chemistry. Significantly, this control of reactivity is independent of neighboring group effects and instead depends on the orientation of the donor and acceptor substrates in the enzyme active site (Franke, et al., 2004, Methods Enzymol, 388, 224-238). On the industrial scale, enzymes offer advantages over synthetic catalysts or auxiliaries that can be expensive to prepare and can generate toxic waste products. According to the present invention, chemical complementation may be used to carry out directed evolution of D-2-deoxyribose-5 -phosphate aldolase (DERA) variants with expanded substrate specificity for chemical synthesis. DERA is the only natural aldolase known to be able to carry out the aldol condensation of two aldehydes (Barbas, et al., 1990, J. Am. Chem. Soc, U2, 2013-2014; Chen, et al., 1992, J.Am. Chem. Soc, 114, 741-748; Gijsen and Wong, 1995, J.Am. Chem. Soc. 1 17, 7585-7591), a task that has been difficult to solve through synthetic methods (Heathcock, et al., 1980, J.Org. Chem., 45, 1066-1081 ; Mahrwald, et al., 1997, Tetrahedron Lett., 38, 4543-4544; Yachi, et al., 1999, J.Am. Chem. Soc. 121; Denmark and Ghosh, 2001, Angew. Chem. Int. Ed. Engl. 40, 4759-4762). The E. coli DERA enzyme has a TIM barrel fold, it is a monomer, its substrate specificity is well defined, and several high-resolution structures are available (Heine, et al., 2001, Science, 294, 369-374; Heine, et al., 2004, J. MoI. Biol. 343, 1019-1034). Current approaches to improve the catalytic activity or change the substrate specificity of DERA rely on in vitro screens, where only ca. 10³— 10⁴ variants can be tested at a time (Greenberg, et al., 2004, Proc. Natl. Acad. Sci. USA, HM, 5788-5793), although efforts are being made to engineer genetic selections for DERA activity (DeSantis, et al., 2003, Bioorg. Med. Chem. H, 43-52). Finally, there is significant interest in modifying the substrate specificity of DERA for a tandem synthesis with 3- azidopropinaldehyde and two acetaldehyde units for the formation of the key chiral lactone 6 present in the cholesterol-lowering drugs Lipitor® (atorvastatin) and Crestor® (rosuvastatin) (Greenberg, el ah, 2004, Proc. Natl. Acad. Sci. USA, 101. 5788-5793; DeSantis, et al., 2003, Bioorg. Med. Chem. JJL 43-52; Liu and Wong, 2004, Tetrahedron Lett. 45, 2439-2441) (FIGURE 7).

By analogy to the adaptation of chemical complementation to detect glycosynthase and glycosidase activity, chemical complementation may be modified to detect aldolase activity simply by adding a vector encoding DERA and the appropriate Dex and Mtx aldehyde substrates (FIGURE 8). Aldol addition may be used to generate the Dex-aldol-Mtx product leading to dimerizatioπ of the DBD- DHFR and AD-GR fusion proteins and activation of the LEU2 reporter gene. Because the reaction is reversible, aldol substrates are designed such that upon aldol addition a cyclic product will be formed, which is not a substrate for DERA (Barbas, et al., 1990, J.Am. Chem. Soc, H2, 2013-2014; Chen, et al., 1992, J.Am. Chem. Soc, JJI4, 741-748; Gijsen and Wong, 1995, J. Am. Chem. Soc. Ji7, 7585-7591). Alternatively, retro-aldolase activity may be selected for using the URA3 counter selection.

Substrate Synthesis. Based on the natural DERA substrates acetaldehyde and D-glyceraldehyde-3-phosphate (Heine, et al., 2001, Science, 294, 369-374), the Dex-"acetaldehyde" (Dex-AAld) and Mtx-"glyceraldehyde" (Mtx- GAId) substrates shown in FIGURE 9 may be used to detect DERA activity using chemical complementation. A phosphodiester "glyceraldehyde" acceptor substrate may be used in order to achieve better cell permeability properties than a phosphate derivative. The recent directed evolution of DERA to accept D-glyceraldehyde with a 2.5-fold improvement in catalytic efficiency using an in vitro screen establishes the feasibility of evolving a DERA variant with altered substrate specificity and the need for further improvement (DeSantis, et al., 2003, Bioorg. Med. Chem., Jl, 43-52). Alternatively, a DERA recently cloned from a soil bacterium that appears to have broader substrate specificity (Greenberg, et al., 2003, Bioorg. Med.Chem., I Y_x 43-52) may be used. Based on inspection of a high-resolution structure of the E. coli DERA with product bound (Heine, et al., 2001, Science, 294, 369-374), Dex-AAld and Mtx- GAId may be readily accommodated to DERA with hydrocarbon linkers if four point mutations outside the catalytic site are made (FIGURE 10). The activity of the substrates with purified enzyme may be verified in vitro based on product synthesis as judged by HPLC. The Mtx substrate may also be tested for cell permeability based on competition of Dex-Mtx transcription activation (this does not work for Dex substrates, see Abida et al.). DERA may also be used to synthesize the Dex-aldol-

Mtx product to ensure that it is cell permeable and activates transcription in our yeast three-hybrid strain.

Directed Evolution of DERA Variants With Specificity for Unphosphorylated Substrates. Directed evolution may be used to produce DERA variants that work efficiently with a Mtx-GAld substrate in which the phosphodiester is replaced with a simple methylene linker. Briefly, loop libraries of the DERA TIM barrel may be selected for increased activity with the unphosphorylated acceptor substrate using the LEU2 growth selection. Evolved variants may be purified and characterized in vitro based on their activity with known fluorogenic substrates (Greenberg, et al., 2004, Proc. Natl. Acad. Sci., HM, 5788-5793). The best-evolved variants may be tested for their ability to synthesize lactone 6 using 3- azidopropinaldehyde or 3-chloropropinaldehyde as an acceptor substrate.

Directed Evolution of a "C-Glycosynthase ". Features of the glycosynthase and aldolase TIM barrels may be combined to create a "C- glycosynthase". Directed evolution may be used to produce a DERA variant that can catalyze addition of a Dex-AAld substrate to a Mtx-Lac-F electrophile, building in the randomized TIM barrel loops that interact with the Lac-F electrophile.

5.10 DIRECTED EVOLUTION SYSTEMS FOR IMPROVING ENZYMES FOR BIOMASS CONVERSION

The systems and methods described in the foregoing sections may be used to "evolve" enzymes for use in biomass conversion. A list of enzymes, and their substrates, relevant to biomass conversion, ("BCEs") which may be "evolved" according to the present invention, are set forth in Tables 5 A and 5B at the end of the Examples section of this document, and in FIGURE 29.

In non-limiting embodiments of the invention, a BCE may be evolved in a system whereby a library of BCE variants is generated (where diversity is generated in vitro and/or in vivo) and then selected by survival in cell culture, using chemical complementation whereby the DNA binding domain ("DBD") and the activation domain ("AD") are joined via the substrate for the BCE. In non-limiting embodiments, the DBD may be comprised in a first fusion molecule which further comprises a first receptor for a first ligand, the AD may be comprised in a second fusion molecule which further comprises a second receptor for a second ligand, and the substrate for the BCE may be comprised in a fusion molecule which comprises the first ligand and the second ligand, such that cleavage of the substrate prevents DBD and AD from being joined and triggering expression of a directly or conditionally toxic reporter gene.

DBD molecules which may be used according to the invention include, but are not limited to LexA, the GAL4 DBD (e.g., residues 1-147) (Yang et al., 1995, Nucl. Acids Res. 23: 1152-1 156); ZFHDl and zinc finger DBDs (Pabo et al., 2001, Annu. Rev. Biochem. 70:313-340).

AD molecules which may be used according to the invention include, but are not limited to, B42, the GAL4 AD (e.g., residues 768-881) (Yang et al., 1995, Nucl. Acids Res. 23:1 152-1156), NF-κB, and HSV VP 16 AD (Burcin et al., 1998, Front. Biosci. 3:cl-7) and the AD used in Chakraborty et al., 1992, J. Biol. Chem. 267:17498-17501).

Receptor/ligand pairs which may be utilized according to the invention include, but are not limited to, DHFR/antifolate (e.g., DHFR/trimethoprim; see Gallagher et al., 2007, Anal. Biochem. 363JJQ: 160-162); glucocorticoid receptor/steroid (or glucocorticoid receptor/agonist or glucocorticoid receptor/antagonist); TET-repressor/tetracycline; penicillin binding proteins/penicillin or cephalosporin (fluorescently labeled penicillins are commercially available, such as BOCILLIN FL and BOCILLIN 650/665 (Molecular Probes, Inc., Oregon)); acetylcholinesterase/acetylcholine (fluorescently labeled acetylcholine is commercially available, such as Amplex Red acetylcholine (Molecular Probes Inc., Oregon)); carboxypeptidase A/MTX; cyclophilin prolyl isomerase/cyclosporin; FK506-binding protein (FKBP)/FK506 and rapamycin; beta-lactamase/clavulinate; DNA binding site/DNA binding protein; DNA alkyltransferase/O⁶-benzylguanine- O⁶ alkylated guanine (Gendreizig et al., 2003, J. Am. Chem. Soc. 125(49): 14970- 14971 ): estrogen receptor/estrone (Muddana and Peterson, 2004, Org. Lett. 6(9): 1409- 1412): streptavidin/biotin (Hussey et al., 2003, J. Am. Chem. Soc. I2503J-.3692-3693); and hemoglobin/heme. For cyclophilin prolyl isomerase/cyclosporin and FK506-binding protein (FKBP)/FK506 and rapamycin, strategies for redesigning ligand/protein interfaces and modified structures are set forth in Clackson, 1998, Curr. Opin. Struct. Biol. 8:451 -458.

BCEs which may be "evolved" according to said system include any enzyme which degrades a naturally occurring substrate, including but not limited to carbon- oxygen lyase enzymes, including but not limited to enzymes listed in Tables 5A and 5B, below, and/or that operate on substrates set forth in Table 5A, Table 5B, and/or FIGURE 29.

Where diversity is generated in vitro, for example by the creation of a diversified phage or plasmid library, standard techniques may be used. Where diversity is generated in vivo, a system as set forth herein may be employed which utilizes a plurality of host cells, wherein each host cell contains a nucleic acid encoding the BCE; contains at least one DNA which can be used to direct mutation of the BCE, and exhibits at least one mutator function.

The foregoing systems may be used to generate a plurality of primary libraries comprising mutants of a BCE, which then may, through genetic exchange

{e.g., sexual reproduction) may be used to produce a secondary library of substantially amplified diversity.

5.1 1 CELLULASE VARIANTS The present invention provides for cellulase variants having increased activity relative to their parent cellulase. The present invention provides for isolated nucleic acids comprising sequences encoding said variants and portions thereof, as well as purified proteins. See for example, Table IA, which includes SEQ ID NOS: 1- 60. In one set of non-limiting embodiments, the present invention provides for a nucleic acid encoding the cellulase variant of Colony 1 described in Section 13, below (termed "CELLVARl", Cellulase 4 in Table IA, below), wherein the nucleic acid sequence comprises SEQ ID NO:1 (FIGURE 21 A), or comprises the nucleic acid sequence of residues 1 through 924 of SEQ ID NO:1, or the nucleic acid sequence of residues 61 through 924 of SEQ ID NO: 1 , or the nucleic acid sequence of residues 109 through 924 of SEQ ID NO:1, or the nucleic acid sequence of residues 1 15 through 924 of SEQ ID NO:1 ; or a nucleic acid which hybridizes to any of the foregoing under stringent conditions, defined as e.g., hybridization in 0.5 M NaHPO₄, 7 percent sodium dodecyl sulfate ("SDS"), 1 mM ethylenediamine tetraacetic acid

("EDTA") at 65°C, and washing in O.lx SSC/0.1 percent SDS at 68°C (Ausubel et al., 1989, Current Protocols in Molecular Biology, Vol. I, Green Publishing Associates, Inc., and John Wiley & Sons, Inc. New York, at p. 2.10.3). The present invention further provides for a nucleic acid which is at least 90 percent, at least 95 percent, at least 96 percent, at least 97 percent, at least 98 percent, or at least 99 percent homologous to one or more of SEQ ID NO:1 (FIGURE 21 A) or residues 1 through 924 of SEQ ID NO:1 or residues 61 through 924 of SEQ ID NO:1 or residues 109 through 924 of SEQ ID NO:1, or residues 115 through 924 of SEQ ID NO:1, where homology may be determined relative to a second sequence using standard software such as BLAST or FASTA, and where said homologous nucleic acid encodes a protein having cellulase activity. Of note, CELLVARl had previously been believed to show a 33-fold improvement over its parental strains, but this was later determined to have been an error, as a parental strain that was not considered in that determination had a high level of activity. In further non-limiting embodiments, the present invention provides for a nucleic acid encoding an amino acid sequence as set forth for CELLVARl in SEQ ID NO:3 (FIGURE 21B), or for a nucleic acid sequence comprising a nucleic acid sequence encoding the amino acid sequence of residues 1 through 308 of SEQ ID NO:3, or residues 25 through 300 of SEQ ID NO:3, or residues 21 through 308 of SEQ ID NO:3, or residues 37-308 of SEQ ID NO:3, or residues 39-308 of SEQ ID NO:3, optionally further encoding additional residues of SEQ ID NO:3 and/or amino acid sequence not found in SEQ ID NO:3, and where the protein encoded by said nucleic acid has cellulase activity. In another set of non-limiting embodiments, the present invention provides for a nucleic acid encoding the cellulase variant of Colony 2 described in Section 8, below (termed "CELLV AR2"), wherein the nucleic acid sequence comprises SEQ ID NO:2 (FIGURE 22A), or comprises the nucleic acid sequence of residues 1 through 924 of SEQ ID NO:2, or the nucleic acid sequence of residues 61 through 924 of SEQ ID NO:2, or the nucleic acid sequence of residues 109 through 924 of SEQ ID NO:2, or the nucleic acid sequence of residues 1 15 through 924 of SEQ ID NO:2; or a nucleic acid which hybridizes to any of the foregoing under stringent conditions, defined as e.g., hybridization in 0.5 M NaHPθ₄, 7 percent sodium dodecyl sulfate ("SDS"), 1 mM ethylenediamine tetraacetic acid ("EDTA") at 65°C, and washing in 0.1 x SSC/0.1 percent SDS at 68⁰C (Ausubel et al., 1989,

Current Protocols in Molecular Biology, Vol. I, Green Publishing Associates, Inc., and John Wiley & Sons, Inc. New York, at p. 2.10.3). The present invention further provides for a nucleic acid which is at least about 90 percent, at least 95 percent, at least 96 percent, at least 97 percent, at least 98 percent, or at least 99 percent homologous to one or more of SEQ ID NO:2 (FIGURE 22A) or residues 1 through 924 of SEQ ID NO:2 or residues 61 through 924 of SEQ ID NO:2 or residues 109 through 924 of SEQ ID NO:2, or residues 1 15 through 924 of SEQ ID NO:2, where homology may be determined relative to a second sequence using standard software such as BLAST or FASTA, and where said homologous nucleic acid encodes a protein having cellulase activity.

In further non-limiting embodiments, the present invention provides for a nucleic acid encoding an amino acid sequence as set forth for CELLVAR2 in SEQ ID NO:4 (FIGURE 22B), or for a nucleic acid sequence comprising a nucleic acid sequence encoding the amino acid sequence of residues 1 through 308 of SEQ ID NO:4, or residues 25 through 300 of SEQ ID NO:4, or residues 21 through 308 of SEQ ID NO:4, or residues 37-308 of SEQ ID NO:4, or residues 39-308 of SEQ ID NO:4, optionally further encoding additional residues of SEQ ID NO:4 and/or amino acid sequence not found in SEQ ID NO:4, and where the protein encoded by said nucleic acid has cellulase activity. Additional non-limiting examples of cellulase variants according to the invention are set forth in Table IA, below, which includes nucleic acid and amino acid sequences as well as, in certain instances, kinetic data showing enzyme activity (summarized in Table IB). Further, the present invention provides for cellulase variants having a nucleic acid sequence which is least about 90 percent, at least 95 percent, at least 96 percent, at least 97 percent, at least 98 percent, or at least 99 percent homologous to a nucleic acid sequence of a variant as set forth in Table IA, where homology may be determined relative to a second sequence using standard software such as BLAST or FASTA, and where said homologous nucleic acid encodes a protein having cellulase activity. In still further non-limiting embodiments, the present invention provides for a nucleic acid encoding an amino acid sequence as set forth for a cellulase variant in Table IA, and where the protein encoded by said nucleic acid has cellulase activity.

Any of the foregoing cellulase variant-encoding nucleic acids may be operably linked to a promoter, which may be a native cellulase promoter or a promoter which is not a native cellulase promoter (i.e., a heterologous promoter). Any suitable promoter, including constitutively active promoters, inducible promoters, eukaryotic promoters, prokaryotic promoters, viral promoters, etc., may be used. Further, said CELLVARl or CELLV AR2-encoding nucleic acids may further be modified or operably linked to other elements which aid expression. For example, but not by way of limitation, amino acid residues 1-20, or 1-36, or 1-38 of SEQ ID NO:3 or SEQ ID NO:4 may be replaced by a heterologous (e.g., aritficial or designed based on the signal sequence of another protein) signal sequence. Any of the foregoing nucleic acids may be comprised in a vector molecule. The invention further provides for a host cell containing a vector molecule which comprises a nucleic acid encoding CELLVARl or CELLV AR2, where the host cell may be a eukaryotic or prokaryotic cell, and in particular may be a bacterial cell, a yeast cell, an insect cell, or a mammalian cell.

The present invention further provides for a CELLVARl protein with cellulase activity comprising the amino acid sequence (SEQ ID NO:3) or portions thereof such as amino acid residues 1 through 308 of SEQ ID NO:3, or amino acid residues 25 through 300 of SEQ ID NO:3, or amino acid residues 21 through 308 of SEQ ID NO:3, or amino acid residues 37 through 308 of SEQ ID NO:3, or amino acid residues 39-308 of SEQ ID NO:3, optionally further encoding additional residues of SEQ ID NO:3 and/or amino acid sequence not found in SEQ ID NO:3 (e.g., a heterologous signal sequence). The present invention further provides for a CELLVAR2 protein with cellulase activity comprising the amino acid sequence (SEQ ID NO:4) or portions thereof such as amino acid residues 1 through 308 of SEQ ID NO:4, or amino acid residues 25 through 300 of SEQ ID NO:4, amino acid residues 21 through 308 of SEQ ID NO:4, or amino acid residues 37 through 308 of SEQ ID NO:4, or amino acid residues 39-308 of SEQ ID NO:4, optionally further encoding additional residues of SEQ ID NO:4 and/or amino acid sequence not found in SEQ ID NO:4 (e.g., a heterologous signal sequence).

The present invention still further provides for a cellulase variant with cellulase activity comprising an amino acid sequence as set forth in Table IA, and for cellulase variants having cellulase activity which are at least about 90 percent, at least 95 percent, at least 98 percent, or at least 99 percent homologous thereto.

Such proteins may be produced using the nucleic acid molecules disclosed above and protein expression methods known in the art. The foregoing cellulases may be used to break down cellulose, for example as part of a method to produce ethanol.

In a specific, non-limiting embodiment of the invention e.g., see Section 14, below), the following chemical complementation system may be used: a DNA binding domain such as LEX a may be fused to the Mtx receptor protein dihydrofolate reductase (DBD-DHFR) and an activation domain such as B42 may be fused to the Dex receptor protein glucocorticoid receptor (AD-GR). In the presence of Mtx-Dex, AD-GR is recruited to DBD-DHFR, which is bound upstream of the reporter gene, effectively reconstituting the transcriptional activator, and leading to transcription of the downstream reporter gene. To adapt the yeast three-hybrid system to read out cellulase catalysis, the linkage between Mtx and Dex may be replaced with the substrate for the reaction, cellulose, and the cellulase enzyme may be added as the fourth component to the system. Both the DBD-DHFR and AD-GR fusion proteins may be expressed from a Gall promoter in a 2μ plasmid, and the reporter gene may be URA3, which in the presence of 5-fluoroorotic acid (5-FOA) produces 5- fluorouracil (5-FU) and so is toxic. Cell survival is achieved by cleavage of the heterodimeric small molecule, disrupting expression of the URA3 gene. Said chemical complementation may be used in a diversifying system to generate cellulase variants of increased activity, which further comprises at least one DINA and at least one mutator function, as described above. Furthermore, a first primary library generated by said chemical complementation/diversifying system may be used to generate a secondary library by genetic exchange with a second primary library generated by said chemical complementation/diversifying system, for example by sexual reproduction. Analogous systems may be used to improve activity of other hydrolases or other enzymes which cleave their substrates.

TABLE IA

CELLULASE 1 = CELLVAR2

(3.7 fold improvement over starting genes)] Kcat = 4.3xl0³ sec^'' Km = 3.2xl0^"3M

Kcat/Km = 1.4x10⁶ sec^{" 1}M^"1

Nucleic acid sequence cellulase 1 fCELL V AR2): SEO ID NO: 2 ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAATGTATTCCGCGTCGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGCGGCGCAAAGCCTCGGCGTTTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGCGGCGTGACCTGGAACGGGCAGATTCGG CCTTATGCGCTGGAAGTGACTGAAACTATCCGTAGTAAAGATCCTGATAA TCTGATTATCGTTGGCACGGGTACCTGGAGTCAGGATATCCATGATGCGG CGGACAATCTGTTGCCCGATCCGAATACGATGTACGCGCTGCATTTCTATG CGGGCACGCACGGGCAATTCCTGTGCGATCGTATTGACTATGCGCAAAGC CGTGGCGCGGCGATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCGG CAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGA ATAACCGTGGCATAAGCTGGGTGAACTGGTCGCTTACCGATAAGTCAGAG GCGTCCGCCGCGCTGGCTCCAGGGGCGAGTAAATCAGGCGGTTGGACGGA GCAGAATTTGTCGGCGTCAGGAAAATTTGTCAGAGCACAGATTCGCGCGG CTGCGAATCTAAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 1 (CELLVAR2V SEO ID NO:4 MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAAAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTETIRSKDPDNLIIVGTGTWSQDIHDAADNLLPDPNTMYALHFYAGTHGQF LCDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGISWVN WSLTDKSEASAALAPGASKSGGWTEQNLSASGKFVRAQIRAAANLSGGDHH HHHH*

CELLULASE 2 (5.7-fold improvement over starting genes)] Kcat = 5.2xlθ³ sec^"'

Km = 2.5x10^'3M

Kcat/Km = 2.IxIO¹¹ SeC ^•-l¹χM/f-1

Nucleic acid sequence cellulase 2 (SEO ID NO:5>

ATGGCCACACCGGTGGAAACGCATGGCCAACTGTCCATTGAAAATGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAATGTATTCCGCGTCGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTCAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTTTACATCATCATCGACTGGCA TATCTTGTCGGATAACGATCCCAATATTTATAAAGCACAGGCAAAAACCT TCTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGCGGCGTGACATGGAACGGGCAGATTCG GCCTTATGCGCTGGATGTTACTGACACTATCCGTAGCAAAGATCCTGATAA CCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGACATCCATGACGCGG CGGATAATCAGCTGCCCGATCCGAATACGCTGTACGCGCTGCATTTCTATG CGGGTACGCACGGGCAGTTCCTGCGCGATCGTATTGACTATGCGCAAAGC CGTGGCGCGGCGATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCGG CAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGA ATAACCGTGGTGTAAGCTGGGTAAACTGGTCACTAAGCGATAAGTCTGAG ACGTCTGCGGCGCTGACTCCAGGGGCGAGTAAATCAGGCGGTTGGACGGA GCAGAATCTGTCGACGTCAGGAAAATTTGTCAGAGAGCAGATTCGTGCAG GTGCGAATCTGGGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 2 (SEQ ID NO:6~)

ATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVN KDSMKWL RDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHILS DNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYALD VTDTIRSKDPDNLIIVGSGTWSQDTHDAADNQLPDPNTLYALHFYAGTHGQFL RDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVNW SLSDKSETSAALTPGASKSGGWTEQNLSTSGKFVREQIRAGANLGGGDHHHH

HH*

CELLULASE 3

Kcat = 5xlO³ sec^'1 Km = 3.6x10^'3M Kcat/Km = 1.4x10⁶ sec^'1 M^"1

Nucleic acid sequence cellulase 3 (SEO ID NO:?)

ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG

ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAATGTATTCCGCGTCGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGCGGCGCAAAGCCTCGGCGTTTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCCAATATTTATAAAGCACAGGCAAAAACCT TCTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGCGGCGTGACCTGGAACGGGCAGATTCGG CCTTATGCGCTGGAAGTGACTGAAACTATCCGTAGTAAAGATCCTGATAA TCTGATTATCGTTGGCACGGGTACCTGGAGTCAGGATATCCATGATGCGG CGGACAATCTGTTGCCCGATCCGAATACGATGTACGCGCTGCATTTCTATG CGGGCACGCACGGGCAATTCCTGTGCGATCGTATTGACTATGCGCAAAGC CGTGGCGCGGCGATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCGG CAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGA ATAACCGTGGCATAAGCTGGGTGAACTGGTCGCTTACCGATAAGTCAGAG GCGTCCGCCGCGCTGGCTCCAGGGGGCGAGTAA Amino acid sequence cellulase 3 (SEO ID NO:8)

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW

LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAAAQSLGVYIIIDWHIL

SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL

EVTETIRSKDPDNLIIVGTGTWSQDIHDAADNLLPDPNTMYALHFYAGTHGQF

LCDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGISWVN

WSLTDKSEASAALAPGGE-

CELLULASE 4 = CELLVARl

Kcat= 1.6x10⁴ sec-l Km=I .3x10³ M Kcat/Km= 1.1 x 10⁷ M ¹ sec- '

Nucleic acid sequence cellulase 4 (CELLVARU fSEO ID NO: 11

ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG

ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATAAAATG GCTGCGTGACGACTGGGGGATCAATGTGTTTCGCGTTGCCATGTACACGG CGGAGAATGGTTATATTGCTAACCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGCGGCGCAAAGCCTCGGCGTCTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCCAATACTTATAAAGCACAGGCAAAAATCT TCTTTGCCGAAATGGCTGGGCTGTATGGCAGCTCACCGAACGTGATTTATG AAATCGCCAATGAGCCAAACGGTGGCGTGACGTGGAACGGACAGATTCG GCCTTATGCGCTGGAAGTGACTGACACCATCCGTAGCAAAGATCCTGATA ACCTTATTATCGTCGGTACCGGCACCTGGAGTCAGGATATCCACGATGCA GCGGATAACCAACTGCCCGATCTGAATACCCTATACGCGCTGCATTTCTAT GCGGGCACGCACGGGCAATTCCTGTGCGATCGTATTGACTATGCGCAAAG CCGTGGCGCGGCGATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCG GCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AATAACCGTGGCGTAAGCTGGGTGAACTGGTCACTAAGCGATAAGTCTGA GACGTCTGCGGCGCTGACTCCAGGGGCGAGCAAATCAGGCGGCTGGACG GAGCAGAATTTGTCGACGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGC GGGGGCGGGTCTGAGCGGTGGTGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 4 (CELL V ARl) (SEO UD NO:3) MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSIKWL RDDWGINVFRVAMYTAENGYIANPSLANKVKEAVAAAQSLGVYIIIDWHILS DNDPNTYKAQAKIFFAEMAGLYGSSPNVIYEIANEPNGGVTWNGQIRPYALE VTDTIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDLNTLYALHFYAGTHGQFL CDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVNW SLSDKSETSAALTPGASKSGGWTEQNLSTSGKFVREQIRAGAGLSGGDHHHH HH*

CELLULASE 5

Kcat= 6.7x10³ se Km=1.5xl0^"3 M Kcat/Km=4.5xl O⁶ M^{^}'sec-¹

Nucleic acid sequence cellulase 5 (SEO ID NO:9)

ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATAAAATG GCTGCGTGACGACTGGGGGATCAATGTGTTTCGCGTTGCCATGTACACGG CGGAGAATGGTTATATTGCTAACCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGCGGCGCAAAGCCTCGGCGTCTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCCAATACTTATAAAGCACAGGCAAAAATCT TCTTTGCCGAAATGGCTGGGCTGTATGGCAGCTCACCGAACGTGATTTATG AAATCGCCAATGAGCCAAACGGTGGCGTGACGTGGAACGGACAGATTCG GCCTTATGCGCTGGAAGTGACTGACACCATCCGTAGCAAAGATCCTGATA ACCTTATTATCGTCGGTACCGGCACCTGGAGTCAGGATATCCACGATGCA GCGGATAACCAACTGCCCGATCTGAATACCCTATACGCGCTGCATTTCTAT GCGGGCACGCACGGGCAATTCCTGTGCGATCGTATTGACTATGCGCAAAG CCGTGGCGCGGCGATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCG GCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AATAACCGTGGCGTAAGCTGGGTGAACTGGTCACTAAGCGATAAGTCTGA GACGTCTGCGGCGCTGACTCCAGGGGCGAGCAAATCAGGCGGCTGGACG GAGCAGAATTTGTCGACGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGC GGGGGCGGGTCTGAGCGGTGGTGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 5 (SEQ ID NO: 10) MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSIKWL RDDWGINVFRVAMYTAENGYIANPSLANKVKEAVAAAQSLGVYIIIDWHILS DNDPNTYKAQAKIFFAEMAGLYGSSPNVIYEIANEPNGGVTWNGQIRPYALE VTDTIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDLNTLYALHFYAGTHGQFL CDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVNW SLSDKSETSAALTPGASKSGGWTEQNLSTSGKFVREQIRAGAGLSGGDHHHH HH*

CELLULASE 6

Kcat= 2.1xl0² sec-l Km= 0.96xl0-³ M

Kcat/Km=2.2xlO⁵ M^-1 sec-¹

Nucleic acid sequence cellulase 6 (SEQ ID NO: 1 O ATGGCCACGCCAGTAGAAACACATGGCCAACTGTCCATCGAAAATGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGATTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAATGTATTCCGCGTCGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGTGGCACAAAGCCTCGGCGTTTACATCATCATCGACTGGCA TATCTTGTCGGATAACGATCCCAATACTTATAAAGCACAGGCAAAAACCT TCTTTGCTGAAATGGCTGGGCTGTATGGCAGCTCACCGAACGTGATTTATG AAATCGCCAATGAGCCAAACGGTGGCGTGACCTGGAACGGGCAGATTCG GCCTTATGCGCTGGAAGTGACTGACACCATCCGTAGCAAAGATCCCGATA ACCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGATATCCATGATGCA GCGGATAACCAACTGCCCGATCCGAATACGCTGTACGCGCTGCATTTCTAT GCGGGCACGCACGGGCAGTTCCTGCGCGATCGCATTGATTATGCACAAAG CCGCGGTGCCGCGATTTTTGTCAGCGAGTGGGGAACCAGCGATGCGTCCG GCAATGGCGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AATAACCGTGGCGTAAGCTGGGTGAACTGGTCGCTTACCGATAAATCAGA GGCGTCCGCCGCGCTGGCTCCGGGAGCGAGCAAATCTGGTGGCTGGACGG AGCAGAATCTGTCGACGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGCA GGTGCGAATCTGGGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 6 (SEQ ID NO: 12) MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRV AMYTAADGYISNPSLANKVKEAVAV AQSLGVYIIID WHIL SDNDPNTYKAQAKTFFAEMAGLYGSSPNVIYEIANEPNGGVTWNGQIRPYAL EVTDTIRSKDPDNLIIVGSGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQF LRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEASAALAPGASKSGGWTEQNLSTSGKFVREQIRAGANLGGGDHH HHHH*

CELLULASE 7 Kcat= l.lxlθ² sec-l Km= 1.4x10^"2 M Kcat/Km=8.12xlO³ M^"1 sec-¹

Nucleic acid sequence cellulase 7 (SEO ID NO: 13) ATGGCCACGCCAGTAGAAACGCATGGCCAACTGTCCATCGAAAATGGGCG ACTGGTGGATGAGCAGGGAAAAAGAGTGCAACTGAGAGGAATCAGCTCG AACGGGTTGCAGTGGTTTGGTGACTACGTCAACAAAGATTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAACGTATTCCGTGTTGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTCAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTCTACATCATCATCGACTGGCA TATCTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGCGGCGTGACATGGAACGGGCAAATTCG GCCTTATGCGCTGGATGTTACTGACACCATCCGTAGCAAAGATCCCGATA ACCTCATTATCGTCGGTACCGGCACCTGGAGTCAGGATATCCACGATGCA GCGGATAACCAACTGCCCGATCCGAATACGCTGTATGCGCTGCATTTCTAT GCCGGTACGCACGGGCAGTTCCTGCGCGATCGCATTGACTATGCACAAAG CCGCGGTGCTGCAATTTTCGTCAGCGAGTGGGGAACAAGCGATGCGTCCG GCAACGGTGGGCCGTTCCTGCCAGAATCGCAAACCTGGATCGATTTCCTG AATAACCGTGGCATAAGCTGGGTTAACTGGTCGCTTACCGATAAATCAGA GGCGTCTGCGGCGCTGGCACCGGGAGCGAGCAAATCTGGCGGTTGGACAG AGCAGAATTTGTCGGCGTCAGGAAAATTTGTCAGAGCACAGATTCGCGCG GCTGCGAATCTGGGCGGTGGCGATCACCATCACCATCACCATTAA Amino acid sequence cellulase 7 (SEQ ID NO: 14)

MATPVETHGQLSIENGRLVDEQGKRVQLRGISSNGLQWFGDYVNKDSMKWL RDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHILS DNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYALD VTDTIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQFL RDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGISWVNW SLTDKSEASAALAPGASKSGGWTEQNLSASGKFVRAQIRAAANLGGGDHHH HHH*

CELLULASE 8

Kcat= 8.4x10² sec-l Km= 4.3xlO ³ M Kcat/Km=2xlO⁵ M°sec-' Nucleic acid sequence cellulase 8 (SEO ID NO: 15}

ATGGCCACGCCAGTAGAAACGCATGGCCAACTGTCCATTGAAAATGGGCG ACTGGTGGACGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGTGACTACGTCAACAAAGATTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAATGTATTCCGCGTCGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGCGGCGCAAAGCCTCGGCGTTTACATCATCATCGACTGGCA TATCTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGCGGCGTGACATGGAACGGGCAAATTCG GCCTTATGCGCTGGAAGTGACTGAAACTATCCGTAGTAAAGATCCTGATA ACCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGACATCCATGATGCG GCAGACAATCAGTTGCCCGATCCGAATACGATGTACGCGCTGCATTTCTAT GCCGGCACGCACGGGCAGTTCCTGCGCGATCGTATCGATTATGCGCAAAG CCGCGGCGCCGCGATTTTTGTCAGCGAGTGGGGCACAAGCGATGCGTCCG GCAACGGCGGACCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AACAACCGTGGTGTGAGCTGGGTTAACTGGTCGCTTACCGATAAGTCAGA GGCGTCTGCGGCGCTGGCACCGGGAGCGAGCAAATCAGGTGGCTGGACG GAGCAGAATCTGTCGACGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGC AGGTGCGAATCTGGGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 8 (SEQ ID NO: 16)

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAAAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTETIRSKDPDNLIIVGSGTWSQDIHDAADNQLPDPNTMYALHFYAGTHGQF LRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEASAALAPGASKSGGWTEQNLSTSGKFVREQIRAGANLGGGDHH HHHH*

CELLULASE 9

Kcat= 7.1xl0² sec-l

Km= 4.6x10³ M

Kcat/Kπv=1.5xlO⁵ M^{^}'sec-¹

Nucleic acid sequence cellulase 9 (SEQ ID NO: 17")

ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG

ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG

CACGGTTTGCAGTGGTTTGGTGACTACGTCAACAAAGATTCGATGAAATG GCTGCGCGATGACTGGGGGATTAACGTATTCCGTGTTGCCATGTACACGG CAGCGGATGGCTATATTTCCAATCCTTCCCTTGCGAATAAGGTCAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTTTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAGCCAAACGGTGGCGTGACCTGGAACGGGCAGATTCG GCCTTATGCGCTGGAAGCCACTGACACCATCCGTAGCAAAGATCCCGATA ACCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGATATCCATGATGCG GCAGACAATCAGTTGCCCGATCCGAATACTCTGTATGCGCTGCATTTCTAT GCGGGTACGCACGGGCAGTTCCTGCGCGATCGTATTGACTATGCACAAAG CCGCGGTGCTGCAATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCG GCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AATAACCGTGGCGTAAGCTGGGTGAACTGGTCGCTTACCGATAAGTCAGA GGTGTCCGCCGCGCTGGCTCCAGGGGCGAGTAAATCAGGCGGTTGGACGG AGCAGAATCTGTCGACGTCAGGAAAATTTGTCAGAGCACAGATTCGCGCG GCTGCGAATCTAAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 9 (SEQ ID NO: 18)

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW

LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EATDTIRSKDPDNLIIVGSGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQF LRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEVSAALAPGASKSGGWTEQNLSTSGKFVRAQIRAAANLSGGDHH HHHH*

CELLULASE 10

Kcat= 87 sec-1 Km= 0.76x10^'3 M Kcat/Km=l .lxlO⁵ M^"'sec-'

Nucleic acid sequence cellulase 10 (SEQ ID NO: 19)

ATGGCCACGCCAGTAGAAACACATGGCCAACTGTCCATCGAAAACGGGCG

ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGTGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGCGATGACTGGGGGATTAACGTATTCCGTGTTGCCATGTACACGG CAGCGGATGGCTATATTTCCAATCCTTCCCTTGCGAATAAGGTCAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTCTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTCGCCGAAATGGCTGGGCTGTACGGTAACTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGTGGCGTGACCTGGAACGGGCAAATTCGG CCTTATGCGCTGGAAGTGACTGAAACTATCCGTAGTAAAGATCCTGATAA CCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGACATCCATGATGCGG CGGACAATCTGTTGCCCGATCCGAATACGATGTACGCGCTGCATTTCTATG CGGGTACGCACGGGCAGTTCCTGCGCGATCGCATTGATTATGCACAAAGC CGCGGTGCCGCGACTTTTGTCAGCGAGTGGGGCACAAGCGATGCGTCCGG CAATGGCGGACCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGA ATAACCGTGGCGTAAGCTGGGTTAACTGGTCGCTTACCGATAAATCAGAG GCGTCTGCGGCGCTGGCTCCAGGGGCGAGCAAATCAGGTGGCTGGACGGA GCAGAATTTGTCAACGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGCGG GTGCAGACCTGAGTGGTGGCGTTCACCATCACCATCACCATTAA

Amino acid sequence cellulase 10 (SEQ ID NO:20^") MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTETIRSKDPDNLIIVGSGTWSQDIHDAADNLLPDPNTMYALHFYAGTHGQF LRDRIDY AQSRGAATFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEASAALAPGASKSGGWTEQNLSTSGKFVREQIRAGADLSGGVHH HHHH*

CELLULASE 11

Nucleic acid sequence cellulase 11 (SEQ ID NO:2P

ATGGCCACACCGGTGGAAACACATGGCCAACTGTCCATCGAAAATGGGCG ACTGGTGGATGAGCAGGGAAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGGTTGCAGTGGGTTGGTGACTACGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAACGTATTCCGTGTTGCCATGTACACGG CAGCGGATGGCTATATTTCCAATCCTTCCCTTGCGAATAAGGTCAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTTTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCTAATATTTATAATGCACAGGCAAAAACCTT CTTTGCTGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAGCCAAACGGTGGCGTGACCTGGAACGGACAGATTCG GCCTTATGCGCTGGAAGTGACTGACACCATCCGTAGCAAAGATCCCGATA ACCTCATTATCGTCGGTACCGGCACCTGGAGTCAGGATATCCACGATGCA GCGGATAACCAACTGCCCGATCCGAATACCCTATACGCGCTGCATTTCTAT GCGGGCACGCACGGGCAGTTCCTGCGCGATCGTATCGATTATGCGCAAAG CCGTGGCGCGGCGATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCG GCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AATAACCGTGGCGTAAGCTGGGTGAACTGGTCGCTTACCGATAAGTCAGA GGCGTCCGCCGCGCTGGCTCCAGGGGCGAGCAAATCTGGCGGTTGGACAG AGCAGAATTTGTCGGCGTCAGGAAAATTTGTCAGAGCACAGATTCGCGCG GCTGCGAATCTAAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 1 1 (SEQ ID NO:22^')

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWVGDYVNKDSMKW

LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHIL SDNDPNIYNAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTDTIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQF LRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEASAALAPGASKSGGWTEQNLSASGKFVRAQIRAAANLSGGDHH HHHH*

CELLULASE 12

Nucleic acid sequence cellulase 12 (SEQ ID NO:23) ATGGCCACGCCAGTAGAAACACATGGCCAACTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGGGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGTGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGCGATGACTGGGGGATTAACGTATTCCGTGTTGCCATGTACACGG CAGCGGATGGCTATATTTCCAATCCTTCCCTTGCGAATAAGGTCAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTCTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTCGCCGAAATGGCTGGGCTGTATGGTAACTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGTGGCGTGACCTGGAACGGGCAAATTCGG CCTTATGCGCTGGAAGTGACTGAAACTATCCGTAGTAAAGATCCTGATAA CCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGACATCCATGATGCGG CGGACAATCTGTTGCCCGATCCGAATACGATGTACGCGCTGCATTTCTATG CGGGTACGCACGGGCAGTTCCTGCGCGATCGCATTGATTATGCACAAAGC CGCGGTGCCGCGATTTTTGTCAGCGAGTGGGGCACAAGCGATGCGTCCGG CAATGGCGGACCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGA ATAACCGTGGCGTAAGCTGGGTTAACTGGTCGCTTACCGATAAATCAGAG GCGTCTGCGGCGCTGGCTCCAGGGGCGAGCAAATCAGGTGGCTGGACGGA GCAGAATTTGTCAACGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGCGG GTGCAGACCTGAGTGGTGGCGTTCACCATCACCATCACCATTAA

Amino acid sequence cellulase 12 (^"SEQ ID NO:24)

MATPVETHGQLSIENGRLVDEQGKRVQLGGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTETIRSKDPDNLIIVGSGTWSQDIHDAADNLLPDPNTMYALHFYAGTHGQF LRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEASAALAPGASKSGGWTEQNLSTSGKFVREQIRAGADLSGGVHH HHHH*

CELLULASE 13

Nucleic acid sequence cellulase 13 (SEQ ID NO:25) ATGGCCACGCCAGTAGAAACGCATGGCCAGCTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGTGACTACGTCAACAAAGATTCGATGAAATG GCTGCGCGATGACTGGGGGATTAACGTATTCCGTGTTGCCATGTACACGG CAGCGGATGGCTATATTTCCAATCCTTCCCTTGCGAATAAGGTCAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTTTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAGCCAAACGGTGGCGTGACCTGGAACGGGCAGATTCG GCCTTATGCGCTGGAAGCCACTGACACCATCCGTAGCAAAGATCCCGATA ACCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGATATCCATGATGCG GCAGACAATCAGTTGCCCGATCCGAATACTCTGTATGCGCTGCATTTCTAT GCGGGTACGCACGGGCAGTTCCTGCGCGATCGTATTGACTATGCACAAAG CCGCGGTGCTGCAATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCG GCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AATAACCGTGGCGTAAGCTGGGTGAACTGGTCGCTTACCGATAAGTCAGA GGTGTCCGCCGCGCTGGCTCCAGGGGCGAGTAAATCAGGCGGTTGGACGG AGCAGAATCTGTCGACGTCAGGAAAATTTGTCAGAGCACAGATTCGCGCG GCTGCGAATCTAAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 13 (SEO ID NO:26) MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EATDTIRSKDPDNLIIVGSGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQF LRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEVSAALAPGASKSGGWTEQNLSTSGKFVRAQIRAAANLSGGDHH HHHH*

CELLULASE 14

Nucleic acid sequence cellulase 14 (SEQ ID NO:27)

ATGGCCACGCCAGTAGAAACGCATGGCCAACTGTCCATTGAAAATGGGCG ACTGGTGGATGAGCAGGGAAAAAGAGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGATTCGATGAGATG GCCGCGCGATGACTGGGGGATTGACGTATTCCGTGTTGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTCAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTCTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCCAATACTTATAAAGCACAGGCAAAAACCT TCTTTGCCGAAATGGCGGGGCTGTATGGCAGCTCACCGAACGTGATTTAT GAAATCGCCAATGAGCCAAACGGTGGCGTGACACGGAACGGGCAAATTC GGCCTTATGCGCTGGATGTGACTGAAACTATCCGTAGTAAAGATCCTGAT AATCCGATTATCGTCGGCAGCGGGACCTGGAGCCAGGACATCCACGATGC GGCGGACAATCTGTTGCCCGATCCGAATACGATGTACGCGCTGCATTTCTA TGCCGGTACGCACGGGCAGTTCCTGTGTGATCGTATCGATTATGCGCAAA GCCGCGGTGCTGCAATTTTCGTCAGCGAGTGGGGCACAAGCGATGCGTCC GGCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCT GAACAACCGTGGTGTGAGCTGGGTTAACTGGTCGCTTACCGATAAATCAG AGGCGTCTGCGGCGCTGGCACCGGGAGCGAGCAAATCTGGCGGTTGGACA GAGCAGAATTTGTCGGCGTCAGGAAAATTTGTCAGAGCACAGATTCGCGC GGCTGCGAATCTAAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 14 (SEQ ID NO:28)

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMRW

PRDDWGIDVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHIL SDNDPNTYKAQAKTFFAEMAGLYGSSPNVIYEIANEPNGGVTRNGQIRPYAL DVTETIRSKDPDNPIIVGSGTWSQDIHDAADNLLPDPNTMYALHFYAGTHGQF LCDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEASAALAPGASKSGGWTEQNLSASGKFVRAQIRAAANLSGGDHH HHHH*

CELLULASE 15

Nucleic acid sequence cellulase 15 (SEO ID NO:291 ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGGTTGGTGACTATGTCAACAAAGATTCGATGAAATG GCTGCGTGACGACTGGGGGATCAATGTGTTTCGCGTTGCCATGTACACGG CGGAGAATGGTTATATTGCTAACCCTTCTCTCGCCAATAAAGTAAAAGAG GCTGTTGCGGCGGCGCAAAGCCTTGGCGTCTACATCATCATCGACTGGCA TATCTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCT CCTTTGCCGAAATGGCGGGGCTGTATGGTAACTCGCCGAACGTGATTTAT GAAATCGCCAATGAGCCAAACGGCAGTGTGACATGGAACGGGCAAATTC GGCCTTATGCGCTGGAAGCCACTGACACCATCCGTAGCAAAGATCCCGAT AACCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGACATCCATGATGC GGCGGACAATCTGTTGCCCGATCCGAATACGCTGTACGCGCTGCATTTCTA TGCGGGCACGCACGGGCAGTTCCTGCGCGATCGCATTGATTATGCACAAA GCCGCGGTGCCGCGATTTTTGTCAGCGAGTGGGGCACAAGCGATGCGTCC GGCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCT GAACAACCGTGGTGTGAGCTGGGTTAACTGGTCGCTTACCGATAAATCAG AGGCGTCTGCGGCGCTGACTCCAGGGGCGAGTAAATCAGGCGGTTGGACG GAGCAGAATCTGTCGACGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGC AGGTGCGAATCTAAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 15 (SEQ ID NO:30)

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWVGDYVNKDSMKW LRDDWGINVFRVAMYTAENGYIANPSLANKVKEAVAAAQSLGVYIIIDWHIL SDNDPNIYKAQAKTSFAEMAGLYGNSPNVIYEIANEPNGSVTWNGQIRPYAL EATDTIRSKDPDNLIIVGSGTWSQDIHDAADNLLPDPNTLYALHFYAGTHGQF LRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEASAALTPGASKSGGWTEQNLSTSGKFVREQIRAGANLSGGDHHH HHH

CELLULASE 16

Nucleic acid sequence cellulase 16 (SEQ ID NO:31) ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAACGTATTCCGTGTTGCCATGTACACGG CAGCGGATGGCTATATTTCCAATCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGTGGCACAAAGCCTCGGCGTTTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTTGCCGAAATGGCGGGGCTGTATGGTAACTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGTGGCGTGACCTGGAACGGGCAAATTCGG CCTTATGCGCTGGATGTTACTGAAACTATCCGTAGTAAAGATCCTGATAAT CTGATTATCGTTGGCACGGGTACCTGGAGTCAGGATATTCATGACGCGGC GGATAATCAGCTGCCCGATCCGAATACGCTGTACGCGCTGCATCTCTATGC CGGTACGCACGGGCAATTCCTGTGCGATCGTATTGACTATGCGCAAAGCC GCGGCGCCGCGATTTTTGTCAGCGAGTGGGGGACAAGCGATGCGTCCGGC AATGGCGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGAAT AACCGTGGCGTAAGCTGGGTGAACTGGTCGCTTACCGATAAGTCAGAGGC GTCCGCCGCGCTGGCTCCAGGGGCGAGCAAATCAGGTGGCTGGACGGAGC AGAATATGTCGACGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGCAGGT GCGAATCTAAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 16 fSEQ ID NO:32) MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL DVTETIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDPNTLYALHLYAGTHGQF LCDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEASAALAPGASKSGGWTEQNMSTSGKFVREQIRAGANLSGGDHH HHHH*

CELLULASE 17

Nucleic acid sequence cellulase 17 (SEO ID NO:33^")

ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG ACTGGTGGACGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATCAATGTGTTTCGCGTTGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGCGGCGCAAAGCCTCGGCGTTTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCCAATACTTATAAAGCGCAGGCAAAAACCT TCTTTGCTGAAATGGCGGGGCTGTATGGCAGCTCACCGAACGTGATTTATG AAATCGCCAATGAGCCAAACGGTGGCGTGACCTGGAACGGGCAGATTCG GCCTTATGCGCTGGAAGTGACTGACACCATCCGTAGCAAAGATCCTGATA ACCTTATTATCGTCGGTACCGGCACCTGGAGTCAGGATATCCACGATGCA GCGGATAACCAACTGCCCGATCCGAATACCCTATACGCGCTGCATTTCTAT GCGGGCACGCACGGGCAATTCCTGTGCGATCGTATTGACTATGCGCAAAG CCGTGGCGCGGCGATTTTCGTCTGCGAGTGGGGCACCAGCGATGCGTCCG GCAATGGCGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AATAACNGTGGCGTAAGCTGGGTGAACTGGTCGCTTAGCGATAAATCAGA GGCGTCCGCCGCGCTGGCTCCAGGCGCGAGTAAATCAGGTGGCTGGACGG AGCAGAATCTGTCGACGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGCA GGTGCGAATCTGGGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 17 (SEQ ID NO:34^')

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW

LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAAAQSLGVYIIIDWHIL SDNDPNTYKAQAKTFFAEMAGLYGSSPNVIYEIANEPNGGVTWNGQIRPYAL EVTDTIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQF LCDRIDYAQSRGAAIFVCEWGTSDASGNGGPFLPESQTWIDFLNNXGVSWVN WSLSDKSEASAALAPGASKSGGWTEQNLSTSGKFVREQIRAGANLGGGDHH HHHH*

CELLULASE 18

Nucleic acid sequence cellulase 18 (SEO ID NO:35> ATGGCCACGCCAGTAGAAACGCATGGCCAACTGTCCATTGAAAATGGGCG ACTGGTGGATGAGCAGGGAAAAAGAGTGCAACTGAGAGGAATCAGCTCG AACGGGTTGCAGTGGGTTGGTGATTACGTAAACAAAGATTCGATGAAGTG GCCGCGTGATGACTGGGGGATTAATGTGTTTCGCGTTGCCATGTACACGG CGGCGGATGGCTATATTTCGAATCCTTCCCTCGCCAATAAGGTCAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTCTACATCATCATCGACTGGCA TATTTTGTCGGACAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTTGCCGAAATGGCGGGGCTGTATGGCAACTCGCCGAATGTGATTTATG GAATCGCCAATGAGCCAAACGGCGGCGTGACATGGAACGGGCAAATTCG GCCTTATGCGCTGGAAGTGACTGAAACTATCCGTAGTAAAGATCCTGATA ATCTGATTATCGTCGGCAGCGGGACCTGGAGCCAGGATATTCATGACGCG GCGGATAATCAGCTGCCCGATCCGAATACGCTGTACGCGCTGCATTTCTAT GCTGGCATGCACGGGCAATTCCTGTGCGATCGTATTGACTATGCGCAAAG CCGTGGCGCGGCGATTTTCGTCATCGAGTGGGGCACCAGCGATGCATCCG GCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AACAACCGTGGCATAAGCTGGGTAAACTGGTCACTAAGCGATAAGTCTGA GACGTCTGCGGCGCTGGCTCTAGGGGCGAGTAAATCAGGCGGCTGGACAG AGCAGAATTTGTCGGCGTCAGGAAAATTTGTCAGAGAGCAGATTCGTGCA GGTGCGAATCTGGGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 18 (SEQ ID NO:36)

MATPVETHGQLSIENGRLVDEQGKRVQLRGISSNGLQWVGDYVNKDSMKW PRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYGIANEPNGGVTWNGQIRPYAL EVTETIRSKDPDNLIIVGSGTWSQDIHDAADNQLPDPNTLYALHFYAGMHGQF LCDRJDYAQSRGAAIFVIEWGTSDASGNGGPFLPESQTWIDFLNNRGISWVNW SLSDKSETSAALALGASKSGGWTEQNLSASGKFVREQIRAGANLGGGDHHH HHH*

CELLULASE 19

Nucleic acid sequence cellulase 19 (SEO ID NO^ ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG ACTGGTGGATGAACGGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTACAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAATGTATTCCGCGTCGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTCAAAGAA GCCGTTGCGGCGGCGCAAAGCCTCGGCGTTTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAATCTT CTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AATTCGCCAATGAACCTAACGGCGGCGTGACCTGGAACGGGCAGATTCGG CCTTATGCGCTGGAAGTGACTGAAACTATCCGTAGTAAAGATCCTGATAA TCTGATTATCGTCGGTACCGGCACCTGGAGTCAGGATATCCACGATGCAG CGGATAACCAACTGCCCGATCTGAATACCCTATACGCGCTGCATTTCTATG CGGGCACGCACGGGCAATTCCTGTGCGATCGTATTGACTATGCGCAAAGC CGTGGCGCGGCGATTTTCGTCCGCGAGTGGGGCACCAGCGATGCATCCGG CAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGA ATAACCGTGGCGTAAGCTGGGTGAACTGGTCGCTTACCGATAAGTCTGAG ACGTCTGCGGCGCTGACTCCAGGGGCGAGCAAATCAGGCGGCTGGACGG AGCAGAATTTGTCGGCATCAGGAAAATTTGTCAGAGCACAGATTCGCGCG GCTGCGAATCTAAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 19 (SEQ ID NO:38^*) MATPVETHGQLS1ENGRL VDERGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAAAQSLGVYIIIDWHIL SDNDPNIYKAQAKIFF AEMAGLYGNSPNVIYEF ANEPNGGVTWNGQIRPYAL EVTETIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDLNTLYALHFYAGTHGQF LCDRIDYAQSRGAAIFVREWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSETSAALTPGASKSGGWTEQNLSASGKFVRAQIRAAANLSGGDHH HHHH*

CELLULASE 20

Nucleic acid sequence cellulase 20 (SEQ ID NO:39)

ATGTTTCGCGTTGCCATGTACACGGCGGCGGATGGTTATATTTCGAATCCT TCCCTCGCCAATAAGGTCAAAGAAGCCGTTGCGGTGGCGCAAAGCCTCGG CGTCTACATCATCATCGACTGGCACATTTTGTCGGATAACGATCCTAATAT TTATAAAGCACAGGCAAAAACCTTCTTTGCCGAAATGGCGGGGCTGTACG GTAATTCGCCGAACGTGATTTATGAAATCGCCAATGAACCTAACGGCGGC GTGACCTGGAACGGGCAGATTCGGCCTTATGCGCTGGAAGTGACTGAAAC TATCCGTAGCAAAGATCCTGATAACCTTATTATCGTCGGCAGCGGGACCT GGAGCCAGGACATCCATGACGCGGCGGATAATCAGCTGCCCGATCCGAAT ACGCTGTACGCGCTGCATTTCTATGCGGGCACGCACGGGCAGTTCCTGCG CGATCGTACTGACTATGCGCAAAGCCGTGGCGCGGCGATTTTCGTCAGCG AGTGGAGCACCAGCGATGCATCCGGCAACGGTGGGCCGTTCCTGCCTGAA TCGCAGACCTGGATCGATTTCCTGAATAACCGTGGCATAAGCTGGGTGAA CTGGTCGCTTACCGATAAGTCAGAGGCGTCCGCCGCGCTGGCTCCAGGGG CGAGTAAATCAGGCGGTTGGACGGAGCAGAATTTGTCGACGTCAGGCAAA TTTGTCAGAGAGCAGATTCGTGCGGGGGCGGGTCTGAGCGGTGGCGATCA CCATCACCATCACCATTAA

Amino acid sequence cellulase 20 (SEQ ID NO:40) MFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHILSDNDPNIY KAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYALEVTETIRSK DPDNLIIVGSGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQFLRDRTDYA QSRGAAIFVSEWSTSDASGNGGPFLPESQTWIDFLNNRGISWVNWSLTDKSEA SAALAPGASKSGGWTEQNLSTSGKFVREQIRAGAGLSGGDHHHHHH*

CELLULASE 21

Nucleic acid sequence cellulase 21 (SEQ ID NO:41) ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCA CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAATGTATTCCGCGTCGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGCGGCGCAAAGCCTCGGCGTTTACATCATCATCGACTGGCA TATCTTGTCGGATAACGATCCCAATATTTATAAAGCACAGGCAAAAACCT TCTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGCGGCGTGACCTGGAACGGGCAGATTCGG CCTTATGCGCTGGAAGTGACTGAAACTATCCGTAGTAAAGATCCTGATAA TCTGATTATCGTTGACACGGGTACCTGGAGTCAGGATATCCATGATGCGG CGGACAATCTGTTGCCCGATCCGAATACGACGTACGCGCTGCATTTCTATG CGGGCACGCACGGGCAGTTCCTGCGCGATCGTATTGACTATGCGCAAAGC CGTGGCGCGGCGATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCGG CAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGA ATAACCGTGGCATAAGCTGGGTGAACTGGTCGCTTACCGATAAGTCTGAG ACGTCTGCGGCGCTGACTCCAGGGGCGAGCAAATCAGGCGGCTGGACGG AGCAGAATTTGTCGACGTCAGGCAAATTTGTCAGAGAACAGATTCGTGCG GGGGCGGGTCTGAGCGGTGGCGATCACCATCACCATCACCATTAA Amino acid sequence cellulase 21 (SEQ ID NO:42)

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAAAQSLGVYHIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTETIRSKDPDNLIIVDTGTWSQDIHDAADNLLPDPNTTYALHFYAGTHGQF LRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGISWVN WSLTDKSETSAALTPGASKSGGWTEQNLSTSGKFVREQIRAGAGLSGGDHHH HHH*

CELLULASE 22

Nucleic acid sequence cellulase 22 (SEO ID NO:43)

ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG

ACTGGTGGATGAACAGGGGAAAAGGGTGCAATTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAATGTATTCCGCGTCGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTCAAAGAG GCCGTTGCGGCGGCGCAAAGCCTCGGCGTTTACATCATCATCGACTGGCA TATCTTGTCGGATAACGATCCCAATATTTATAAAGCACAGGCAAAAACCT TCTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAGCCAAACGGTGGCGTGACGTGGAACGGACAGATTCG GCCTTATGCGCTGGAGGTGACTGACACTATCCGTAGCAAAGATCCTGATA ACCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGACATCCATGACGCG GCGGATAATCAGCTGCCCGATCCGGATACGCTGTACGCGCTGCATTTCTAT GCGGGCACGCACGGGCAATTCCTGTGCGATCGTATTGACTATGCGCAAAG CCGTGGCGCGGCGATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCG GCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AATAACCGTGGCATAAGCTGGGTGAACTGGTCGCTTACCGATAAGTCAGA GGCGTCCGCCGCGCTGGCTCCAGGGGCGAGTAAATCAGGCGGCTGGACGG AGCAGAATTTGTCGACGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGCG GGGGCGGGTCTGAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 22 (SEO ID NO:44^')

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAAAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTDTIRSKDPDNLIIVGSGTWSQDIHDAADNQLPDPDTLYALHFYAGTHGQF LCDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGISWVN WSLTDKSEASAALAPGASKSGGWTEQNLSTSGKFVREQIRAGAGLSGGDHH HHHH*

CELLULASE 23 Nucleic acid sequence cellulase 23 (SEO ID NO:45~)

ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAATGTATTCCGCGTCGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGCGGCGCAAAGCCTCGGCGTTTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCCAGTATTTATAAAGCACAGGCAAAAACCT TCTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGCGGCGTGACATGGAACGGGCAGATTCG GCCTTATGCGCTGGAAGTGACTGAAACTATCCGTAGTAAAGATCCTGATA ATCTGATTATCGTTGGCACGGGTACCTGGAGTCAGGATATCCATGATGCG GCGGACAATCTGTTGCCCGATCCGAATACGATGTACGCGCTGCATTTCTAT GCGGGCACGCACGGACAATTCCTGTGCGATCGTATTGACTATGCGCAAAG CCGTGGCGCGGCGATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCG GCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AATAACCGTGGCATAA

Amino acid sequence cellulase 23 (SEQ ID NO:46) MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAAAQSLGVYIIIDWHIL SDNDPSIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTETIRSKDPDNLIIVGTGTWSQDIHDAADNLLPDPNTMYALHFYAGTHGQF LCDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGISWVN WSLTDKSEASAALAPGASKSGGWTEQNLSASGKFVRAQIRAAANLSGGDHH HHHH*

CELLULASE 24 Nucleic acid sequence cellulase 24 (SEO ID NO:47~)

ATGGCCACGCCAGTAGAAACGCATGGCCAACTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGTGACTACGTCAACAAAGATTCGATGAAATG GCTGCGCGATGACTGGGGGATTAACGTATTCCGTGTTGCCATGTACACGG CGGAGAATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTAAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTCTACATCATCATCGACTGGCA TATCTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAATCTT CTTTGCCGAAATGGCGGGGCTGTATGGCAACTCGCCGAACGTGGTTTATG AAATCGCCAATGAACCTGACGGCGGCGTGACCTGGAACGGGCAAATTCGG CCTTATGCGCTGGAAGCCACTGACACTATCCGTAGCAAAGATCCTGATAA CCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGACATCCATGATGCGG CGGACAATCTGTTGCCCGATCCGAATACGATGTACGCGCTGCATTTCTATG CGGGTACACACGGGCAGTTCCTGCGCGATCGTATCGATTATGCGCAAAGC CGCGGTGCCGCGATTTTTGTCAGCGAGTGGGGCACCAGCGATGCATCCGG CAACGGCGGACCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGA ACAACCGTGGCGTAAGCTGGGTGAACTGGTCGCTTACCGATAAATCAGAG GCGTCTGCGGCGCTGGCACCGGGGGCGAGCAAATCTGGCGGTTGGACGGA GCAGAATTTGTCGGCGTCAGGAAAATTTGTCAGAGCACAGATTCGCGCGG CTGCGAATCTAAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 24 (SEO ID NO:48)

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAENGYISNPSLANKVKEAVAVAQSLGVYIΠDWΉIL SDNDPNIYKAQAKIFFAEMAGLYGNSPNVVYEIANEPDGGVTWNGQIRPYAL EATDTIRSKDPDNLIIVGSGTWSQDIHDAADNLLPDPNTMYALHFYAGTHGQF LRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEASAALAPGASKSGGWTEQNLSASGKFVRAQIRAAANLSGGDHH

HHHH*

CELLULASE 25

Nucleic acid sequence cellulase 25 (SEQ ID NO:49) ATGGCCACGCCAGTAGAAACGCATGGCCAACTGTCCATTGAAAATGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGTGACTACGTCAACAAAGATTCGATGAAATG GCTGCGTGACGACTGGGGGATCAATGTGTTTCGCGTTGCCATGTACACGG CGGAGAATGGTTATATTTCCAATCCTTCCCTTGCGAATAAGGTCAAAGAA GCCGTTGCGGCGGCGCAAAGCCTCGGCGTTTACATCATCATCGACTGGCA CATTTTGTCGGATAACGATCCCAATACTTATAAAGCACAGGCAAAAACCT TCTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAGCCAAACGGTGGCGTGACCTGGAACGGGCAGATTCG GCCTTATGCGCTGGAAGTGACTGACACCATCCGTAGCAAAGATCCCGATA ACCTCATTATCGTCGGTACCGGCACCTGGAGTCAGGATATCCACGATGCA GCGGATAACCAACTGCCCGATCCGAATACCCTATACGCGCTGCATTTCTAT GCTGGCACGCACGGGCAGTTCCTGCGCGATCGCATTGACTATGCGCAAAG CCGCGGTGCCGCGATTTTTGTCAGCGAGTGGGGCACCAGCGATGCATCCG GCAATGGCGGGCCGTTCCTGCCAGAATCGCAAACCTGGATCGATTTCCTG AATAACCGTGGCGTAAGCTGGGTGAACTGGTCGCTTACCGATAAGTCAGA GGCGTCTGCGGCGCTGACTCCAGGGGCGAGTAAATCAGGCGGTTGGACGG AGCAGAATCTGTCGACGTCAGGAAAATTTGTCAGAGCACAGATTCGCGCG GCTGCGAATCTGGGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 25 (SEO ID NOtSOl MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAENGYISNPSLANKVKEAVAAAQSLGVYIIIDWHIL SDNDPNTYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTDTIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQF LRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVN WSLTDKSEASAALTPGASKSGGWTEQNLSTSGKFVRAQIRAAANLGGGDHH HHHH*

CELLULASE 26

Nucleic acid sequence cellulase 26 (SEO ID NO:5 P

ATGGCCACGCCAGTAGAAACACATGGCCAACTGTCCATCGAAAATGGGCG

ACTGGTGGACGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGTATAAATGTGTTCCGCGTCGCTATGTACACGG CGGAGAATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTAAAAGAG GCCGTTGCGGCGGCGCAAAGCCTTGGCATCTACATCATTATCGACTGGCA CATTTTGTCGGATAACGATCCTAATATTTATAAAGCGCAGGCAAAAACCTT CTTTGCTGAAATGGCGGGGCTGTATGGTAATTCGCCGAACGTGATTTATGA AATCGCCAATGAACCTAACGGCGGCGTGACATGGAACGGGCAGATTCGGC CTTATGCGCTGGAAGCCACTGACACCATCCGTAGCAAAGATCCCGATAAC CTCATTATCGTCGGTACCGGCACCTGGAGTCAGGATATTCATGACGCGGC AGACAATCAGCTGCCCGATCCGAATACGCTGTACGCGCTGCATTTCTATGC GGGTACGCACGGGCAGTTCCTGCGCGATCGTATCGATTATGCGCAAAGCC GCGGTGCTGCAATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCGGC AACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGAA CAACCGTGGTGTGAGCTGGGTTAACTGGTCGCTTACCGATAAATCAGAGG CGTCCGCCGCGCTGGCTCCAGGGGCGAGTAAATCAGACGGTTGGACGGAG CAGAATCTGTCGACGTCAGGAAAATTTGTCAGAGCACAGATTCGCGCGGC TGCGAATCTAAGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 26 (SEQ ID NO:52^") MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAENGYISNPSLANKVKEAVAAAQSLGIYIIIDWHILS DNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQTRPYALE ATDTIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQFL RDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVNW SLTDKSEASAALAPGASKSDGWTEQNLSTSGKFVRAQIRAAANLSGGDHHHH HH*

CELLULASE 27 Nucleic acid sequence cellulase 27 (SEO ID NO:53^>)

ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGCGACTATGTCAACAAAGACTCGATGAAATG GCTGCGTGATGACTGGGGGATTAATGTATTCCGTGTTGCCATGTACACGGC AGCGGATGGCTATATTTCCAATCCTTCCCTTGCGAATAAGGTCAAAGAAG CCGTTGCGGTGGCGCAAAGCCTTGGCATCTACATCATCATCGACTGGCAC ATTTTGTCGGATAACGATCCTAATATTTATAAAGCGCAGGCAAAAACCTTC TTTGCTGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATGA AATCGCCAATGAACCTAACGGCGGCGTGACATGGAACGGGCAGATTCGGC CTTATGCGCTGGATGTTACTGACACTATCCGTAGCAAAGATCCTGATAACC TTATTATCGTCGGTACCGGCACCTGGAGTCAGGATATTCATGACGCGGCG GATAATCAGCTGCCCGATCCGAATACGCTGTACGCGCTGCATTTCTATGCG GGCACGCACGGGCAGTTCCTGCGCGATCGCATTGACTATGCACAAAGCCG CGGTGCTGCAATTTTTGTCAGCGAGTGGGGCACAAGCGATGCATCCGGCA ACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGAATA ACCGTGGTATAAGCTGGGTAAACTGGTCGCTTACCGATAAGTCTGAGACA TCCGCGGCGCTGGCACCGGGAGCGAGCAAATCTGGTGGTTGGACAGAGCA GAATTCGTCGGCGTCAGGAAAATTTGTCAGAGAGCAGGTTCGTGCGGGTG CAGACCTGAGTGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 27 (SEQ ID NO:54)

MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADG YISNPSLANKVKEA V A VAQSLGIYIIID WHILS DNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYALD VTDTIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQFL RDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGISWVNW SLTDKSETSAALAPGASKSGGWTEQNSSASGKFVREQVRAGADLSGGDHHH

HHH*

CELLULASE 28

Nucleic acid sequence cellulase 28 (SEQ ID NO:55) ATGGCCACGCCAGTAGAAACGCATGGCCAACTGTCCATCGAAAATGGGCG ACTGGTGGATGAGCAGGGAAAAAGAGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGTGACTATGTCAACAAAGACTCGATGAAGTG GCTGCGTGATGACTGGGGGATTAACGTATTCCGTGTTGCCATGTACACGG CAGCGGATGGCTATATTTCCAATCCTCCCCTTGCGAATAAGGTCAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTCTACATCATCATCGACTGGCA TATCTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGCGGCGTGACATGGAACGGGCAAATTCG GCCTTATGCGCTGGAAGTGACTGACACTATCCGTAGCAAAGATCCTGATA ACCTTATTATCGTCGGCAGCGGGACCTGGAGCCAGGACATCCATGATGCG GCAGACAATCAGTTGCCCGATCCGAATACGCTGTACGCGCTGCATTTCTAT GCGGGCACGCACGGGCAATTCCTGTGCGATCGTATTGACTATGCGCAAAG CCGTGGCGCGGCGATTTTCGTCAGCGAGTGGGGCACCAGCGATGCATCCG GCAACGGCGGACCGTTCCTGCCTGAATCGCAAACCTGGATCGATTTCCTG AATAACCGTGGCATAAGCTGGGTTAACTGGTCGCTTAGCGATAAGTCTGA GACGTCTGCGGCGCTGACTCCAGGGGCGAGTAAATCAGGCGGTTGGACGG AGCAGAATCTGTCGGCGTCAGGAAAACTTGTCAGAGCACAGATTCGTGCA GGTGCGAATCTGGGCGGTGGCGATCACCATCACCATCACCATTAA

Amino acid sequence cellulase 28 (SEQ ID NO:56^") MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSMKW LRDDWGINVFRVAMYTAADGYISNPPLANKVKEAVAVAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTDTIRSKDPDNLIIVGSGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQF LCDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGISWVN WSLSDKSETSAALTPGASKSGGWTEQNLSASGKLVRAQIRAGANLGGGDHH HHHH*

CELLULASE 29

Nucleic acid sequence cellulase 29 (SEO ID NO: 57)

ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG

ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGTGGTTTGGTGACTACGTCAACAAAGGTTCGATGAAATG GCTGCGCGATGACTGGGGGATTAACGTATTCCGTGTTGCCATGTACACGG CGGCGGATGGTTATATTTCGAATCCTTCCCTCGCCAATAAGGTAAAAGAA GCCGTTGCGGTGGCACAAAGCCTCGGCGTCTACATCATCATCGACTGGCA TATCTTGTCGGATAACGATCCTAATATTTATAAAGCACAGGCAAAAACCTT CTTTGCCGAAATGGCGGGGCTGTACGGTAATTCGCCGAACGTGATTTATG AAATCGCCAATGAACCTAACGGCGGCGTGACATGGAACGGGCAAATTCG GCCTTATGCGCTGGAAGTGACTGAAACTATCCGTAGTAAAGATCCTGATA ATCTGATTATCGTTGGCACGGGTACCTGGAGTCAGGATATTCATGACGCG GCGGATAATCAGCTGCCCGATCCGAATACGATGTACGCGCTGCATTTCTAT GCGGGTACGCACGGGCAGTTCCTGCGCGATCGCATTGATTATGCACAAAG CCGCGGTGCCGCGATTTTTGTCAGCGAGTGGGGCACCAGCGATGCATCCG GCAACGGTGGGCCGTTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTG AATAACCGTGGCGTAAGCTGGGTGAACTGGTCGCTTACCGATAAATCANA GGCGTCCGCCGCGCTGGCTCCAGGG GCGAGTAAATCTGGCGGTTGGACGGANCANAATCTGTCGACGTCAGGCAA ATTTGTCANANAGCAGATTCGTGCAGGTGCGAATCTGGGCGGTGGCGATC ACCATCACCATCACCATTAA

Amino acid sequence cellulase 29 (SEO ID NO:581 MATPVETHGQLS1ENGRL VDEQGKRVQLRGVSSHGLQWFGDYVNKGSMKW LRDDWGINVFRVAMYTAADGYISNPSLANKVKEAVAVAQSLGVYIIIDWHIL SDNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYAL EVTETIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDPNTMYALHFYAGTHGQ FLRDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWV NWSLTDKSXASAALAPGASKSGGWTXXNLSTSGKFVXXQIRAGANLGGGDH HHHHH*

CELLULASE 30

Nucleic acid sequence cellulase 30 (SEO ID NO:591

ATGGCCACGCCAGTAGAAACGCATGGTCAGCTGTCCATCGAAAACGGGCG ACTGGTGGATGAACAGGGGAAAAGGGTGCAACTGAGAGGGGTCAGTTCG CACGGTTTGCAGT GGTTTGGTGACTACGTCAACAAAGATTCGATAAAATGGCTGCGTGACGAC TGGGGGATCAATGTGTTTCGCGTTGCCATGTACACGGCGGAGAATGGTTA TATTGCTAACCCTTCTCTCGCCAATAAGGTAAAAGAGGCCGTTGCGGCGG CGCAAAGCCTCGGCGTTTACATCATCATCGACTGGCATATCTTGTCGGATA ACGATCCTAATATTTATAAAGCACAGGCAAAAACCTTCTTTGCCGAAATG GCTGGGCTGTACGGTAATTCGCCGAACGTGATTTATGAAATCGCCAATGA ACCTAACGGCGGCGTGACATGGAACGGGCAAATTCGGCCTTATGCGCTGG AAGTGACTGACACCATCCGTAGCAAAGATCCCGATAACCTCATTATCGTC GGTACCGGCACCTGGAGTCAGGATATCCACGATGCAGCGGATAACCAACT GCCCGATCCGAATACCCTATACGCGCTGCATTTCTATGCCGGTACGCACGG GCAGTTCCTGCGCGATCGTATTGACTATGCACAAAGCCGCGGTGCCGCGA TTTTCGTCAGCGAGTGGGGCACAAGCGATGCGTCCGGCAACGGCGGACCG TTCCTGCCTGAATCGCAGACCTGGATCGATTTCCTGAATAACCGTGGCGTA AGCTGGGTGAACTGGTCGCTTACCGATAAGTCAGAGACGTCTGCGGCGCT GGCACCGGGAGCGAGCAAATCTGGCGGTTGGACAGAGCAGAATTTGTCG GCGTCAGGCAAATTTGTCAGAGAGCAGATTCGTGCAGGTGCGAATCTGGG CGGTGGCGATCACCATCACCATCACCATTAA

Aminoacid sequence cellulase 30 (SEQ ID NO:60) MATPVETHGQLSIENGRLVDEQGKRVQLRGVSSHGLQWFGDYVNKDSIKWL RDDWGINVFRVAMYTAENGYIANPSLANKVKEAVAAAQSLGVYIIIDWHILS DNDPNIYKAQAKTFFAEMAGLYGNSPNVIYEIANEPNGGVTWNGQIRPYALE VTDTIRSKDPDNLIIVGTGTWSQDIHDAADNQLPDPNTLYALHFYAGTHGQFL RDRIDYAQSRGAAIFVSEWGTSDASGNGGPFLPESQTWIDFLNNRGVSWVNW SLTDKSETSAALAPGASKSGGWTEQNLSASGKFVREQIRAGANLGGGDHHH HHH*

TABLE IB

The present invention is illustrated by the following non-limiting working examples. The contents of the working examples is hereby incorporated by reference into this detailed description of the invention section.

6. EXAMPLE: IN VIVO MUTAGENESIS WITH HO ENDONUCLEASE

Directed evolution requires two steps — mutagenesis of the DNA encoding the protein of interest, followed by selection of the fittest variants. Both these steps may be carried out by a living cell. Harnessing the cell's powerful mutagenesis and selection machineries for "unnatural" chemistry has real advantages for searching very large library sizes and ultimately being able to achieve the de novo directed evolution of enzymes (FIGURE 1 1).

The key to the directed evolution of ever more dramatic changes in function is the ability to search very large libraries of protein variants. Changes in enzyme substrate specificity typically have required libraries of about 10^s protein variants (Yano, et al., 1998, Proc. Natl. Acad. Sci. USA, 95, 5511-5515; Park, et al., 2006, Science, 31 1 , 535-538). It is reasonable to assume that the de novo directed evolution of enzyme activity requires even larger library sizes. So what limits library size? Hypothetical Iy, beginning with 10⁸ cells, each containing DNA encoding a unique protein variant (a library size currently achievable by DNA transformation in yeast). If the yeast has a doubling time of 6 hours, and the DNA could be mutated at each round of cell division, after 1 week there would be ca. 10^l6cells, with 10¹⁶ different protein variants (FIGURE 1 IA). The problem is that given there are about 10⁸ cells/mL in a rapidly growing and dividing yeast cell culture ("late log phase"), 10¹⁶ cells would require 100,000 L of material! This thought experiment nicely illustrates that the problem is not really technical issues like DNA transformation efficiency, but simply the volume of material required (for in vitro systems there is a similar limit imposed by the amount of DNA that can realistically be produced). The cell's elegant solution to this problem is to carry out mutation and selection simultaneously. For example, simplistically, if only the fitter variant survived at each round of cell division, after a week of mutation and selection, there would be only 10⁸ cells containing the 10⁸ fittest protein variants (FIGURE 1 IB). Thus, as detailed below, a yeast strain may be engineered that can carry out both mutation and selection for "unnatural" chemistry to allow a virtual search of very large library sizes.

The following relates to developing an in vivo mutagenesis system with a traditional TRPl yeast genetic selection, although such a system may be extended to incorporate chemical complementation technology. By analogy to ampicillin-resistance selections in E. coli, TRPl is one of the classic selectable markers used in yeast genetics (Jones and Fink, 1982, The molecular biology of the yeast sacchromyces (eds. Strahern, J., Jones, E. &Broach, J.) 181-299, Cold Spring Harbor Laboratory, Cold Spring Harbor; Braus, G.H., 1991 , Microbiol. Rev., 55, 349-370). TRPl encodes phosphoribosyl anthranilate isomerase (PRAI), which carries out the isomerisation of phosphoribosyl anthranilate to carboxyphenylamino deoxyribulose phosphate-the third step in tryptophan biosynthesis from chorismic acid (Braus, G.H., 1991, Microbiol. Rev., 5_5, 349-370). Thus, a yeast trpl auxotroph can be used to select for PRAI activity simply by growing cells on media lacking tryptophan (Perfect, et al., 1992, Gene, 122, 213-217; Nakai, et al., 2000, FEMS Microbiol. Lett. 190, 51-56; Cosano, et al., 1998, Yeast, 14, 861-867). PRAI is a TIM barrel enzyme, and its sequence, structure, and function are very well characterized (Sterner, et al., 1995, Embo. J., 14, 4395-4402; Thoma, et al., 2000, Structure, 8, 265-276; Henn-Sax, et al., 2002, Biochemistry, 4J_, 12032-12042; Sterner, et al., 2001, Methods Enzymol, 33_i, 270-280). The enzyme is well behaved, and the activity of the purified protein can be readily monitored in vitro based on the change in fluorescence at 400 nm upon conversion of substrate to product (Hommel and Kirschner, 1995, Biochemistry, 34_» 5429-5439).

A system may be developed to direct evolution so as to increase the activity of the thermophilic PRAI from Thermotoga maritima at ambient temperature (the T. maritima PRAI is impaired about 10-fold in its catalytic efficiency at ambient temperature compared with the S. cerevisiae enzyme (Thoma, et al., 2000, Structure, 8, 265-276; Hommel and Kirschner, 1995, Biochemistry, 34, 5429-5439) and so is expected to begin with a weak selective advantage in the yeast auxotroph). Alternatively the directed evolution of phosphoribosyl imidazolecarboxamide isomerase (HIS6) to PRAI (TRPl) (Jurgens, et al., 2000, Proc. Natl. Acad. Sci. USA, 97, 9925-9930) reported previously could be used.

Because of the efficiency of homologous recombination in S. cerevisiae (Krogh and Symington, 2004, Annu. Rev. Genet., 38, 233-271) (homologous recombination is significantly less efficient, for example in E. colϊ), TIM barrel loop libraries may be produced in vivo simply by transforming linear oligonucleotides encoding the loop libraries into yeast containing DNA encoding the PRAI TIM barrel. Homologous recombination is used routinely as a tool in yeast genetics for creating "knock-outs" by replacing a segment of the chromosomal DNA with a homologous piece of linear synthetic DNA by recombination (Baudin, et al., 1993, Nucleic Acids Res., 3_i, 3329-3330). Homologous recombination has long been used to create plasmid in yeast (Ma, et al., 1987, Gene, 58, 201-216) (homologous recombination has been used to engineer yeast three-hybrid strains (Baker, et al., 2003, Anal. Biochem. 315, 134-137)). More quantitatively, Sexson and co-workers showed that using 30-base pair homologous ends he could achieve efficient plasmid reconstitution via homologous recombination in yeast (Raymond, et al., 1999, Biotechniques, 26, 134-138; 140-131). Building on this work, homologous recombination has been exploited for in vivo "DNA shuffling" (Swers, et al., 2004, Nucleic Acids Res. 32, e36; Cherry, et al., 1999, Nat. Biotechnol., 17, 379-384; Abecassis, et al., 2000, Nucleic Acids Res. 28, E88). The TIM barrel scaffold lends itself readily to loop mutagenesis by homologous recombination since the β-sheet and α-helix on either side of each loop provide sufficient homology regions for efficient recombination. Currently, the synthesis of libraries in which multiple, nonadjacent loops are randomized by in vitro PCR methods is cumbersome, expensive, and low yielding (see Beste et al.). Synthesis of the TIM barrel loop libraries may be used to create a family of libraries in which different combinations of the loops are randomized at different stages in the selection process (FIGURE 12A).

To produce a suitable host cell, the gene encoding the PRAI from T. maritima may be placed under control of the tunable MET promoter into a yeast trpl auxotroph, for example by homologous recombination. Random loop libraries for loops 3 and 5, which form much of the substrate-binding pocket for PRAI (Thoma, et al., 2000, Structure, 8, 265-276; Hennig, et al., 1997, Biochemistry, 36, 6009-6016), may be prepared. Linear oligonucleotides ca. 100 base pairs in length encoding the randomized loops surrounded on either side by 30-base pair PRAI homology regions may be prepared from synthetic oligonucleotides by standard molecular biology methods. These oligonucleotide loop libraries may be transformed into yeast by electroporation. The efficiency of homologous recombination may be confirmed by sequencing PRAI genes obtained by colony PCR from colonies grown under nonselective conditions. Then the loop libraries may be used to increase the activity of the T. maritima PRAI at ambient temperatures with a trpl growth selection. In Vivo Mutagenesis with HO Endonuclease. As discussed above, much larger library sizes could be accessed, however, if the mutagenesis could be carried out entirely inside the cell, with a round of mutagenesis and selection at each cell division. Since the mutagenized DNA would not have to be reintroduced to the cell at each round of mutagenesis, in vivo mutagenesis would remove the limit on library sizes currently imposed by DNA transformation efficiency. Classic yeast genetic experiments in which the HO endonuclease is placed under control of the tunable Gal promoter provide a route to achieving this goal (Nasmyth, K., 1987, Embo. J. 6, 243-248; Connolly, et al., 1988, MoI. Cell. Biol. 8, 2342-2349; Elias- Arnanz, et al., 1996, Mol.Gen. Genet, 252, 530-538). In studies of yeast recombination, it has been shown that the highly-specific HO endonuclease, which naturally is involved in mating-type switching in yeast, can be knocked-out and then reintroduced into the yeast chromosome under control of the Gal promoter without impairing yeast viability (Connolly, et al., 1988, MoI. Cell. Biol. 8, 2342-2349; Nasmyth, K., 1993, Curr. Opin. Genet. Dev., 3, 286-294; Jensen and Herskowitz, 1984, Symp. Quant. Biol., 49, 97-104). HO endonuclease can then be induced after each cell division to cleave a 2μ plasmid engineered to have a 28-base pair-HO endonuclease recognition site (Connolly, et al., 1988, MoI. Cell. Biol. 8, 2342-2349). By tuning the timing and concentration of galactose induction, the extent of plasmid cleavage can be controlled such that a percentage of the plasmid should remain uncleaved (Galactose Induction Protocol., yvww.bio.brandeis.edu/haberlab/iehsite/protocol.htmt) and so be copied and restored to the ca. 60-copies per cell typical for 2μ plasmids at each cell division (Volkert, et al., 1989, Microbiol. Rev. 53, 299-317). Accordingly, linear loop libraries may be generated in vivo simply by constructing a 2μ plasmid encoding each loop library embedded in HO endonuclease cleavage sites. Loop libraries may be generated afresh after each cell division simply by cycling HO endonuclease induction.

7. EXAMPLE: ESTABLISHMENT OF A YEAST HOST STRAIN

Based on a published report that a HIS6:D127V mutant could complement a TRPl deficiency in E. coli, the directed evolution of HIS6 may be used to increase its activity with the analogous indole substrate based on TRPl complementation in yeast. The data presented herein show that the HIS6:D127V enzyme can also complement a TRPl deficiency in yeast, validating the proof of principle target.

HIS6:D127V has been previously identified to complement a TRPl deficiency. Since this evolution was carried out in E.coli, it was first determined whether the TRPl selection in yeast could differentiate between the inactive H1S6 and the active HIS6:D127V variants. After 4 days of selection on solid media, HIS6 did not complement a TRPl deficiency, while the HIS6:D127V variant did, surviving on plates containing 10 μM tryptophan (FIGURE 13A). The TRPl selection was then used to carry out a mock selection. Starting from a 1 : 100 ratio of active (HIS6:D127V) to inactive (HIS6) gene, the TRPl selection achieved >100-fold increase in TRPl activity after seven days of selection (FIGURE 13B). The results validate the conversion of HIS6 to HIS6:D127V to complement a TRPl deficiency in yeast. 8. EXAMPLE: USE OF CHEMICAL COMPLEMENTATION TQ DETECT GLYCOSYNTHASE AND GLYCOSIDASE ACTIVITIES

8.1 DETECTION OF GLYCOSYNTHASE ACTIVITY Experimental results. First, it was shown that chemical complementation could link Cel7B:E197A glycosynthase activity to transcription of a LEU2 reporter gene in the presence of the Dex-cellobiose (Dex-Cel) and Mtx-lactose-F substrates (Mtx-Lac-F). Then the LEU2 selection was used to enrich for the Cel7B:E197A glycosynthase from a 1 :100 mixture of active:inactive glycosynthase variants, giving a 400-fold enrichment in a single round of selection. Finally, the LEU2 selection was used to select the most active variants from a saturation library at position El 97 in Cel7B. Significantly, false positives (ca. 50% of the clones) could be readily ruled out by a simple secondary three-hybrid screen with and without the Dex-Cel and Mtx-Lac-F substrates. Two of the most active variants, Cel7B:E197S and an unencoded Cel7B:N196D,E197A double mutant, were purified and their activity was determined in vitro based on product synthesis as verified by HPLC and the specific activity of fluoride ion release as reported (Lin, et al., 2004, J. Am. Chem. Soc, 126. 15051- 15059). The Cel7B:E197S variant was found to have a five-fold increase in glycosynthase activity (Table 2). It should be emphasized that this variant had high activity with unmodified substrates, suggesting that use of the Dex and Mtx linkers in the chemical complementation selection should not impair the activity of evolved variants with the desired in vitro substrates. TABLE 2 Glycosynthase activities and protein purification yields for CeI17B variants

For the results in Table 2, glycosynthase activity for tetrasaccharide synthesis from α-lactosyl fluoride and p-nitrophenyl β-cellobioside (PNPC) was measured for the Humicola isolens Veil 7B variants in sodium phosphate buffer, pH 7.0, at room temperature. Specific activities were determined by measuring the fluoride ion release rate by a fluoride ion selective electrode. The protein purification yields were determined by western analysis from total cell culture.

8.2 DETECTION OF GL YCOSIDASE ACTIVITY Using chemical complementation for the directed evolution of glycosidases required synthesis of the Mtx-Lac-Cel-Dex tetrasaccharide substrate, which was achieved using the Cel7B:E197A glycosynthase enzyme. Mtx-Lac-Cel- Dex was purified by HPLC and confirmed by UV spectroscopy and LRMS. LacZ transcription assays confirmed that this tetrasaccharide substrate was cell permeable and able to activate transcription in a yeast three-hybrid system. A counter selection method was required. A counter selection is more demanding because enzyme catalysis must relieve transcription of a toxic reporter gene. The chemical complementation system was adapted to provide a URA3 counter selection for bond cleavage reactions (Huang, et al., 1997, Proc. Natl. Acad. Sci. USA, 94, 13396-13401; Vidal, et al., 1996, Proc. Natl. Acad. Sci. USA, 93, 10315-10320). Briefly, the URA3 gene product converts 5-fluoroorotic acid (5-FOA) to 5-fluorouracil (5-FU), which is toxic, allowing for counter selection in the presence of 5-FOA (a classic yeast genetics experiment). Synthesis of the URA3 gene from a weak SPO 13 promoter attenuates toxicity. Toxicity is then controlled by transcription activation of the URA3 gene by the artificial yeast three-hybrid transcriptional activator. First, the URA3 yeast three-hybrid counter-selection strain was developed using a Dex-Mtx CID with a simple methylene linker. Then, it was shown that a 5-FOA selection could distinguish the active Cel7B glycosidase from the inactive Cel7B:E197A variant in this strain background (FIGURE 5A). Finally, the URA3 counter selection was used to enrich 200- fold for the Cel7B glycosidase from a 1 : 100 mixture of activeύnactive glycosidase variants (FIGURE 5B). These results establish that chemical complementation can be used to detect glycosidase activity, serving as the basis for directed evolution experiments.

9. VALIDATION OF HP-INDUCED DNA CASSETTE

RECOMBINATION IN VIVO

The following experiment was performed to test the system schematically depicted in FIGURE 14A. According to this system, short oligonucleotide (DINA) libraries may be generated in vivo by galactose-induced cleavage of a 2 μ plasmid encoding the loop libraries embedded in HO endonuclease sites at each round of cell division.

To determine both the level of HO-induced recombination and the background level of recombination under non-inducing conditions, recombination of a DNA cassette encoding the Dl 27V mutation into the wt HIS6 gene was scored based on growth in a yeast trpl knockout strain (strain VM46Y). The wtHISό gene is encoded on a standard yeast expression plasmid under control of the MET promoter. The D 127V cassette is encoded on a second plasmid, with 30-bp of homology to the HIS6 gene on either side of the D 127V mutant and HO cleavage sites on either side of the cassette. The yeast strain is a trpl knockout with HO endonuclease under control of the GAL promoter integrated into the chromosome. VM46Y was grown under both inducing and non-inducing conditions, and then the percent recombinants was measured as the ratio of cells growing on T plates (10 μM Trp, only HIS6:D127V grows under these conditions) versus non-selective plates (both HIS6 and HIS6:D127V grow under these conditions). Briefly, an overnight culture of VM46Y taken from a a fresh patch was used to seed a culture that was then grown to an ODδoo = 0.6. These cells were then either induced by 1 hour growth in 2% lactate followed by 1 hour growth in 2% galactose ("induced cells") or grown under non-inducing conditions by 1 hour growth in 2% lactate plus 2% glucose followed by 1 hour growth in 2% glucose ("non- induced cells").

The results of these experiments is shown in FIGURE 14B-C. Based on serial dilutions, the percent recombinants were measured as induced cells = 2% and non-induced cells = 0.004%. It was estimated that the true number of recombinants is probably higher, because the HIS6:D127V complementation of the yeast trpl auxotroph is incomplete. Specifically, when a plasmid with a LEU2 selectable marker encoding the HIS6:D127V variant is transformed into the yeast trpl auxotroph, the ratio of cells growing on a L^' versus a LT plate is not 1 :1 as would be expected, but rather is 50:1. Given that HIS6:D127V complementation of the trpl auxotroph is only 2%, the true recombination in induced cells may be estimated to be 100% and in non-induced cells may be estimated to be 0.2%. To further evaluate the extent of recombination, the following experiment, performed in liquid culture rather than on selective medium, was performed.

Strain VM39 is a trp- auxotroph, containing the Ho-endonuclease under control of the galactose promoter, and carrying the plasmid coding for HIS6 wildtype. VM39 was transformed with a pool of plasmids ("library plasmid") coding the D127NNS cassette with 30bp of homology to the HIS6 gene on either side of the D127NNS mutant and the HO cleavage sites in either side of the cassette. A silent restriction site (BssSI) was incorporated 5 base pairs away from the 127 position in the cassettes to more easily analyze the pool of HIS6D127 variants after selection.

After transformation of the VM39 cells with the pool of plasmids, the transformed cells were incubated for lhr at 30⁰C in YPD at 200rpm for recovery. Next, the cells were pelleted, washed with distilled water, and resuspended in 2ml induction media (synthetic medium containing galactose and lacking leucine and uracil) and shaken for lhr at 30⁰C (see FIGURE 14D for schematized protocol).

Then, the cells were pelleted, washed with sterile distilled water, and resuspended in 2ml selection media (synthetic media containing glucose and lOuM tryptophan) and shaken at 30⁰C for 8 days. Finally, lOOul of the selection media were used to inoculate a 5ml amplification culture (synthetic media containing glucose and lacking leucine) to amplify the winners of the TRPl selection. After 5 days the DNA from the cells in the amplification culture was extracted and used as a template to PCR the pool of HIS6 variants. The PCR of the pool of HIS6 variants after selection was digested overnight with BssBI. FIGURE 14E shows a DNA gel of the pool of HIS6 variants after digestion with BssBI. As FIGURE 14E shows, after 8 days of selection, the pool of HIS6 variants that underwent 60min of galactose induction has been enriched for the HIS6:D127NNS variants when compared to the pool of HIS6 variants induced with galactose for only 30 min. Since the His6:D127NNS variants carry a BssSi restriction site, it is possible to identify them in a pool of HIS6 wiltype strains by digestion with BssSi which generates a smaller DNA fragment. The enrichment occurs because during the 60min of galactose induction, the HO endonuclease is expressed form the galactose promoter and effectively cleaves the cassettes. The recombination machinery of yeast then effectively recombine the cassettes at the 127 position in the His6 wild type incorporating not only the NNs mutation but also the BssSi restriction site.

10. EXAMPLE: GENERATION OF A SECONDARY (DIVERSIFIED) LIBRARY BY MATING (DIVERSIFIED) PRIMARY LIBRARIES Testing of Mating and Sporulation. Strain VM44 is a MATα type, trp- auxotroph, containing the HO-endonuclease under control of the galactose promoter, and the HisA wild-type plasmid with a leucine marker. Strain VM45 is the MATa type version of VM44. MATa and MATα cells mate with each other and create zygotes. These zygotes spawn diploids that contain both MATa and MATα mating types and cease to mate. To test the efficiency of mating liquid cultures, an overnight culture of each VM44 and VM45 was grown to saturation. ODβoo readings determined cell counts and equal cell amounts were then mixed into 2ml of media, in a 14ml culture tube. The tubes shook at 30°, 200RPMs and then 1 :10 dilutions with sterile water were taken at several time intervals; 30min, 60min, 90min, 120min, 210min, 270min, 300min. Using a microscope at 40Ox total magnification, zygotes vs. non zygotes were counted in at least 10 large hemocytometer squares, per dilution. Each dilution is calculated as the average of the 10 squares. Figure 15 shows the percentage of zygotes which peaks around 210min.

Based on the above experiment, it was concluded that these strains need no more than 6 hours to fully mate. To test the efficiency of sporulation in triplicate, an overnight culture of each VM44 and VM45 was grown to saturation. ODβoo readings determined cell counts and equal cell amounts were mixed into 2ml of media, in a 14ml culture tube. The tubes were then shaken at 30°, 200RPMs for 6 hours to ensure full mating. ODβoo readings were established and the newly mated culture was used to inoculate to ODe_Oo=- 100, three 14 ml culture tubes with 2ml of 10% YPD media (which is yeast extract-peptone, media with 10% glucose). These tubes were shaken at 22°, 350RPM until the OD₆₀o=-600-.800, which is about 16 to 18 hours. Each tube was pelleted and washed twice with sterile water, and then re- suspended with 2ml SPO2 sporulation media (0.5% KAc). Each tube went back into 22° shaker at 350RPM for another 48 hours. The cells were than washed with water, diluted 1 :10, and viewed under 400X total magnification with a hemocytometer to determine the number of spores versus unsporulated cells. Each tube was calculated as the average of 10 hemocytometer squares counted. The three tubes were then averaged to determine the percent sporulation in triplicate after 48 hours, which is 47%.

Mating Of Primary (Diversified) Libraries. The following experiment was performed, whereby two primary libraries (diversified by in vivo recombination) were combined such that mating and sporulation could occur, thereby producing a secondary library of amplified diversity, with a goal of producing a HIS mutant with improved Trp activity. VM44 was transformed with library 1 , a pool of plasmids that randomized sites for loop 5 (S125NNS, D127NNS) and loop 8 (I221NNS, G223NNS, F226NNS). VM45 was transformed with library 2, a pool of plasmids that randomized sites for loop 2 (H48NNS, V50NNS, L52NNS, A55NNS, I56NNS), the loops being flanked by HO endonuclease cleavage sites. After transformation the cells were incubated for 1 hr at 30⁰C in YPD at 200rpm for recovery. Next, the cells were pelleted, washed with distilled water, and resuspended in 5ml pre-induction media (synthetic medium containing 2% lactate and lacking leucine and uracil) and shaken for lhr at 30⁰C. Then, the cells were pelleted, washed with sterile distilled water, and resuspended in 5ml induction media (synthetic media containing galactose and lacking leucine and uracil, to induce production of the HO- endonuclease) and shaken for lhr at 30⁰C. Finally, the cells were pelleted, washed with sterile distilled water, and resuspended in 5ml selection media (synthetic media containing glucose and lOuM of tryptophan) and shaken at 30⁰C for 2 more hours. lOOul of a dilution of OD₆oo=-300 from each tube was plated on selection plates (synthetic media plates containing glucose and lOuM of tryptophan). Figure 16 is a picture of these plates taken after 4days of 30⁰C incubation.

Each 5ml cell culture was than mixed together in 50ml flask and shook overnight at 30°, 200RPM to ensure full mating. lOOul of a dilution of OD₆oo=-300 from each tube was plated on selection plates. Figure 17 is a picture of these plates taken after 4 days of 30° incubation. All the cells were used to inoculate a large IL culture of selection media to ODeOo= IOO and shaken at 30°, 200RPM. ODβoo readings were taken at several time intervals to determine whether selection was working. The selection increased to

3.35 after four days which implied that the selection worked well. lOOul of a dilution of OD6oo⁼-300 from each tube was plated on selection plates. Figure 18 is a picture of these plates taken after 4 days of 30° incubation.

OD_6Oo= 1 -00 number was used to inoculate to OD₆₀O=- 100, 10% YPD media. The flask was shaken at 22°, 35ORPM until the OD₆oo=-600-.800, which is about 16 to 18 hours. The cells were then pelleted and washed twice with sterile water, and then re-suspended with SPO2 sporulation media (0.5% KAc). The flask went back into 22° shaker at 350RPM for another 96 hours. A sample of cells was then washed with water, diluted 1 :10, and viewed under 400X total magnification with a hemocytometer to determine the number of spores versus unsporulated cells. The sporulation efficiency after 96 hours was calculated as the average of 10 hemocytometer squares counting spores vs. non-spores, which is 94%. The spores were separated by pelleting the cells down, washing them with sterile water and resuspending the cells in ImI of sterile water. 5OuI of Zymolase (24Units/ul) was added and incubated in 37° waterbath for lhr to break the ascus. To complete separation, the cells were washed with 1% SDS solution and sonicated for five minutes. lOOul of a dilution of OD₆oo= 300 from each tube was plated on selection plates. Figure 19 is a picture of these plates taken after 4 days of 30° incubation.

In order to further analyze the results, the cells were used to prepare a maxiprep so the pool of plasmids could be electroporated into Lucigen E cloni. Supercompetent cells. Another maxiprep was prepared from the E.coli and transformed back into a generic trp- yeast strain (Δtrpl). This was done to minimize the incidences of false positives. 118 colonies were picked and grown overnight in SC(L-) (synthetic media lacking leucine) in 2x96 well plates. The overnight culture was used to dilute 1 :10 into selection media (synthetic media with 5uM tryptophan) and OD_6Oo were read after 2 days. The growth of these colonies were compared to HisWT and HisD127 mutants and colonies which grew faster than HisD127V were isolated for to compare growth in triplicate. As seen in FIGURE 20, several colonies grow better then the HisD127V mutant, which implies that they have higher TrpF catalytic activity.

1 1. EXAMPLE: DIRECTED EVOLUTION OF GL YCOSIDASE ACTIVITY

Using chemical complementation for the directed evolution of glycosidases required synthesis of the Mtx-Lac-Cel-Dex tetrasaccharide substrate, which was achieved using the Cel7B:E197A glycosynthase enzyme. Mtx-Lac-Cel- Dex was purified by HPLC and confirmed by UV spectroscopy and LRMS. LacZ transcription assays confirmed that this tetrasaccharide substrate was cell permeable and able to activate transcription in a yeast three-hybrid system. A counter selection method was required. A counter selection is more demanding because enzyme catalysis must relieve transcription of a toxic reporter gene. The chemical complementation system was adapted to provide a URA3 counter selection for bond cleavage reactions (see FIGURE 23 A for schematic drawing; Huang, et al., 1997, Proc. Natl. Acad. Sci. USA, 94, 13396-13401; Vidal, et al., 1996, Proc. Natl. Acad. Sci. USA, 93, 10315-10320). Briefly, the URA3 gene product converts 5-fluoroorotic acid (5-FOA) to 5-fluorouracil (5-FU), which is toxic, allowing for counter selection in the presence of 5-FOA (a classic yeast genetics experiment). Synthesis of the URA3 gene from a weak SPOJ 3 promoter attenuates toxicity. Toxicity is then controlled by transcription activation of the URA3 gene by the artificial yeast three- hybrid transcriptional activator. First, the URA 3 yeast three-hybrid counter-selection strain was developed using a Dex-Mtx CID with a simple methylene linker. Then, it was shown that a 5-FOA selection could distinguish the active Cel7B glycosidase from the inactive Cel7B:E197A variant in this strain background (FIGURE 5A). Finally, the URA3 counter selection was used to enrich 200-fold for the Cel7B glycosidase from a 1 :100 mixture of active:inactive glycosidase variants (FIGURE 5B). These results establish that chemical complementation can be used to detect glycosidase activity, serving as the basis for directed evolution experiments.

12. EXAMPLE: DIRECTED EVOLUTION OF CELLULASE I

As above, chemical complementation was adapted to detect cellulase activity, utilizing a URA3 counter selection for bond cleavage reactions, allowing detection of cellulase activity (FIGURE 23 A; Vidal et al., 1996, Proc Natl Acad Sci U S A. 93: 10315-10320; Huang and Schreiber, 1997, Proc Natl Acad Sci U S A 94: 13396-13401). Also as above, using the URA3 counter selection, 200-fold enrichment for Humicola insolens Cellulase7B from a 1 :100 mixture of active:inactive cellulase mutants was (FIGURE 23B). Further, URA3 counter selection was applied to the directed evolution of cellulases and a cellulase variant was identified having 2.6 fold increased activity when compared to the parent gene (FIGURE 23C).

13. EXAMPLE: DIRECTED EVOLUTION OF CELLULASE II Materials and Methods: Construction of Reporter Yeast Strain for Cellulase Selection. The Ura3

ORF from pMWl 12 was amplified using primers VC 1077 and VC 1640 and was, by homologous recombination, introduced into the ura3-52 locus of the yeast strain FY251 (MATa trplΔ63 his3A200 ura3-52 leu2Δ I GaI+) and selected on 5- floroorotic acid plates to give strain VC2169. Next, the plasmids pKB521 and pBC398* carrying the DNA-binding domain-dihydrofolate reductase fusion protein (LexA-eDHFR) and the glucocorticoid receptor-activation domain protein fusion (GR--B42), were transformed into VC2169 to give strain PPY9. The pUra3- LexA(4op)-Pspol3-Spol3(15aa)-Ura3 fragment was created by fusion PCR of the LexAop(8x) fragment from pMW 112, amplified using primers VC 1671 and VC 1672, and the Pspol3-Spol3(15aa)-Ura3 fragment from MaV95² (Invitrogen), amplified using primers VC 1669 and VC 1670. The pUra3-LexA(4op)-Pspol3-Spol3(15aa)- Ura3construct was introduced into p425Met25 to create pPPY2176. Finally the pUra3-LexA(4op)-Pspol3-Spol3(15aa)-Ura3 fragment from pPPY2176 was amplified using primers VCl 677 and VC 1678 and integrated into PPY9 to create VPPYl 2, the reporter yeast strain for cellulase selection. The successful integration was confirmed by PCR and phenotypically.

Cellulase mutagenesis. The DNA shuffling vector, pPPY2148, was constructed by digesting p425Met25 (ATCC) with HindIII and Pstl and introducing an 800bp stuffer flanked by Sfil sites. The catalytic domain of four family 5 cellulases with 77-90% sequence identity, E. carotovora CeIN (Olsen et al., 1996,

Biotechnology (N Y) 14:71-76), CeIVI (Mae and Palva, 1995, MoI Gen Genet. 247:17-26). CeIA (Park et al., 1997, Biochem Biophys Res Commun. 241 :636- 641)and CeIT (cloned from the E. carotovora subs, atroceptica (Bell et al., 2004, Proc Natl Acad Sci U S A. K)Il 1 1 105-11110) genomic DNA) were cloned into pPY2148 to create plasmids pPPY2230, pPPY2232, pPPY2234 and pPPY2236 respectively. A ⁶His-tag was added at the C-terminus of the four parent genes to facilitate protein purification after selection. The catalytic domain of these cellulases was identified by sequence comparison with the catalytic domain of Bacillus agaradhaerens Cel5A, which has 68% sequence identity at the amino acid level with the parent genes and which three dimensional X-ray structure is available (Davies et al., 1998, Biochemistry 37i 1926-1932; Davies et al., 1998, Biochemistry 37: 11707-11713: Varrot et al., 2000, J MoI Biol. 297i 819-828; Varrot et al., 2001, Acta Crystallogr D Biol Crystallogr. 57i 1739-1742; Varrot and Davies, 2003, Acta Crystallogr D Biol Crystallogr. 59^ 447-452). The cellulase genes were shuffled using standard procedures with slight variantions (Zhao and Arnold, 1997, Nucleic Acids Res. 25l 1307-1308). To introduce the shuffled product into the shuffling vector via homologous recombination, 30bp homology to the shuffling vector was introduced at the 5' and 3' end of the shuffled product by carrying an extra amplification step of the shuffling product primers VC1658-VC1659. 7.5ng of shuffled product (1.4El 1 shuffled product molecules assuming each base pair is 33OM and the shuffled product was lOOObp) was used as a template in a lOOul amplification reaction. The product was purified in a 2% agarose gel.

Yeast transformation and cellulase selection. The reverse three hybrid strain was transformed using standard high transformation efficient procedures (Colby et al., 2004, Methods Enzymol. 388: 348-358). After recovery for an hour in YPD, the cells were centrifuge for 10 sec at lO.OOOrpm, washed with sterile distilled water. The pellet was then resuspended in selection media (SC(HTL^"), 2% galactose, 0.2% 5- FOA, IuM Dex-Cel-Lac-Mtx) for a final volume of 2ml. A sample of the selection was plated under non-selective conditions on day 9. In vitro kinetics. Ten colonies selected on day 8 and 9 were characterized in vitro using para-nitrophenyl cellobioside (pNPC) as a susbtrate. The cellulase chimeras were amplified from the shuffling vector using primers PPYlOO and PPY 103 and subcloned into the T7-expression vectors pAED4 between Ndel and HindIII sites. The cellulase chimeras were over expressed in E.coli and purified using a Nickel affinity column. Protein concentration was determined using the extinction coefficient of the parent genes ε=76890 and absorption at A₂So- To determine the kinetic constants for the hydrolysis of pNPC, the cellulase chimeras were incubated at room temperature in phosphate buffer (25mM K₂HPO₄, 10OmM NaCl, pH=7) containing various concentrations of the substrate, and the release of p-nitrophenol (ε_PH=f= 4000M^-1Cm^'1 , A420) was recorded continuously in a SpectraMax Plus 384 spectrophotometer.

Results

TABLE 3

The coding regions of the cellulase genes for Colony 1 and Colony 2 were sequenced and determined to be SEQ ID NO: 1 and SEQ ID NO:2, respectively. Of note, because a parental strain of colony 1 (cellulase 4; CELLVARl) was not considered, the increase in activity was determined to be 33-fold, which was an error.

14. EXAMPLE: DIRECTED EVOLUTION OF CELLULASES VIA CHEMICAL COMPLEMENTATION A reverse yeast three-hybrid strain was developed that performs efficiently in liquid culture, thereby enabling cellulase selection in liquid culture which allows for small increases in catalytic activity to be amplified and reduces the amount of small molecule required per experiment. Once the yeast three-hybrid strain was developed, cellulase activity from Humicola insolens Cel7B, which catalyzes the hydrolysis of β-l,4-linked glucosidic bonds in cellulose, was detected using the URA3 counter selection. In the experiments described in this section, it is demonstrated that chemical complementation can read out cellulase activity and can produce a 200-fold enrichment in active cellulases from a 1 :100 active:inactive cellulase pool after one round of selection. The URA3 counter selection was then applied to the directed evolution of cellulases using three cellulases from Erwinia carotovora: CeIN and CeIA and CeV (Olsen, O. et al., 1996, Biotechnology (N Y) 14; 71-6; Park, et al., 1997, Biochem. Biophys. Res. Commun. 241 : 636-41; Bell, et al., 2004, Proc. Natl. Acad. Sci. U S A 101 :11105-10). Using DNA family shuffling as the mutagenesis and chemical complementation as the selection platform, cellulase variants were isolated with 3.6-fold and 5.3-fold increases (Cellulase 1 (CELLV AR2) and Cellulase 2 of Table IA, respectively)in cellulase activity on/7-nitrophenyl cellobioside (pNPC). This selection methodology enables a high-throughput assay that may be used for directed evolution or cloning of cellulase variants with increased catalytic activity for the conversion of biomass to fuel ethanol (FIGURE 24).

14.1 METHODS

Reporter yeast strain construction. The URA3 gene from pMWl 12 was recombined at the ura3-52 locus of FY251 (MATa trplΔ63 his3Δ200 ura3-52 leu2Δ 1 GaR) and selected on plates lacking uracil to give strain VC2169Y. Vectors pBC398 and pKB521 carrying the GR-B42 and Lex A-DHFR constructs were transformed into VC2169Y to give strain VC2291Y. The LexA(4op)-URA3 reporter construct was obtained by fusion PCR of the LexA(4op) from pMWl 12 and the Spol3-URA3 fusion gene from MaV95 (Vidal et al., 1996, Proc. Natl. Acad. Sci. U S A 93 : 10315-20). The reporter construct was inserted into pPP Y2148 to create pPPY2176. The reporter construct was amplified with oligos carrying 30bp homology to the URA3 promoter and the URA3 gene, recombined at the URA3 locus of VC2291 Y, and selected on 5 -FOA plates to give the reverse yeast three-hybrid strain VC2204Y. Recombination was confirmed phenotypically and by colony PCR. Cellulase enrichment assay. A library containing 1 : 100 ratio of active cellulase Cel7B to inactive cellulase (Cel7B:E197A) variants was transformed into the reverse yeast three-hybrid strain to give a library size of 104. Nine random colonies were analyzed to establish the library integrity. The library was plated under non-selective conditions lacking the corresponding auxotrophs and incubated at 30⁰C. After three days, the cells were resuspended in sterile distilled water to an OD600=1 and 1 Oμl were used to carry out the enrichment assay in synthetic media containing 2% galactose, 2% raffinose, 0.2% 5-FOA, and lμM Mtx-Cel-Dex, and lacking histidine, tryptophan, and leucine, in a final volume of 200μl. After 5 days of selection, a sample was plated under non-selective conditions lacking the corresponding auxotrophs and nine random colonies were analyzed by colony PCR and restriction digest. A Ncol site was incorporated into the Cel7B:E197A variant when the glutamic acid nucleophile was mutated to an alanine in order to facilitate the restriction analysis.

Cellulase mutagenesis. The DNA shuffling and expression vector pPPY2148 was constructed by digesting p425Met25 (ATCC 87323) with HindIII and Pstl and introducing an 800bp stuffer flanked by Sfil sites. A C-terminus 6-Histag was added to the catalytic domains, signal peptides removed, of CeIN, CeIA, and CeIV, and subcloned into pPPY2148 to create vectors pPPY2230, pPPY2234, and pPPY2236 respectively. The cellulase genes were shuffled using standard procedures with slight modification (Zhao& Arnold, 1997, Nucleic. Acids. Res. 25: 1307-8). The shuffled product was re-amplified using oligos containing 30bp homology to the promoter and terminator regions of the shuffling vector and the product was purified in a 2% agarose gel.

Yeast transformation and cellulase selection. The reverse yeast three-hybrid strain, VC2204Y, was transformed with the 12.5μg of shuffled cellulases and 3.6μg of cut vector via electroporation using high transformation efficiency procedures with slight variations to give a library of 10⁸ (Colby, et al., 2004, Methods Enzymol. 388: 348-58). The library was resuspended in ImI YPD and recovered in a 30⁰C shaker for lhour. After recovery, lOμl were used to determine the library size and the rest was resuspended in selection media (synthetic media at pH 5 containing 2% galactose, 2% raffinose, 0.2% 5-FOA, and lμM Mtx-Cel-Dex, lacking histidine, tryptophan, and leucine) to a final volume of 2ml. The selection was run for nine days, and samples were taken on days 5, 6 and 8. To isolate selected variants, samples of the selection were diluted and plated under non-selective conditions. Secondary screen for aldehyde formation using CMC. The secondary screen for aldehyde formation using CMC was adapted to be carried out in a 96-well plate format (Jue et al., 1985, J. Biochem. Biophys. Methods. 11 : 109-15). The twenty-two colonies isolated before selection and the twenty-two colonies isolated after 5 days of selection were grown in 10ml synthetic media lacking leucine for two days. The same number of cells (2x107) from each cell culture were arrayed on a 96-well plate and lysed using lOOμl YPER (Novagen). The cell extract (80μl) was transferred onto a second 96-well plate containing 120μl 2% CMC (low viscosity, Sigma) in 0.1 M sodium acetate buffer pH 5, and the reaction was incubated at 37°C for 24hrs. The next day, lOμl of the reaction mixture were incubated with 90μl of 0.1% tetrazolium blue (0.5M potassium sodium tartrate, 0.05M sodium hydroxide) and heated at 98⁰C for lOmin. The reaction was cooled to room temperature and absorption at 660nm was taken immediately.

In vitro kinetic characterization with pNPC. The three starting cellulases and the improved cellulase variants were subcloned into the E.coli T7 vector, pAED4, between Ndel and HindIII sites to give vectors pPPY2256 (CeIN), pPPY2258 (CeIA), pPPY2257 (CeIV), pPPY2291 (Cel_3.7), and pPPY2292 (Cel_5.7). The cellulases were purified using Quick spin Ni-NTA affinity column (Qiagen). Protein concentration was determined using the extinction coefficient of the starting genes (ε=76670) and absorption at 280nm. To determine the kinetic constants for the hydrolysis of pNPC, the cellulase chimeras were incubated at room temperature in phosphate buffer (25mM K.2HPO4, 10OmM NaCl, pH 7) containing seven different concentrations of pNPC ranging from 12mM to 0.185mM. The release of p-nitrophenol (εpH 7= 4000M-lcm-l, A420) was recorded continuously in a SpectraMax Plus 384 spectrophotometer at 420nm. 14.2 RESULTS

In the yeast three-hybrid system the DBD and the AD are split in half and each half is fused to a protein receptor for one of the small molecules. In the Mtx- Dex yeast three-hybrid system, the DBD is fused to the Mtx receptor protein dihydrofolate reductase (DBD-DHFR) and the AD is fused to the Dex receptor protein glucocorticoid receptor (AD-GR). In the presence of Mtx-Dex, AD-GR is recruited to DBD-DHFR, which is bound upstream of the reporter gene, effectively reconstituting the transcriptional activator, and leading to transcription of the downstream reporter gene.

To adapt the yeast three-hybrid system to read out cellulase catalysis, the linkage between Mtx and Dex is replaced with the substrate for the reaction, cellulose, and the cellulase enzyme is added as the fourth component to the system. For the cellulase growth selection, LexA is used as the DBD and B42 as the AD. Both the DBD-DHFR and AD-GR fusion proteins are expressed from a Gall promoter in a 2μ plasmid. The reporter gene is URA3, which in the presence of 5-fluoroorotic acid (5-FOA) produces 5-fluorouracil (5-FU) and so is toxic. Cell survival is achieved by cleavage of the heterodimeric small molecule, disrupting expression of the URA3 gene (Figure 25). For the directed evolution of cellulases, the catalytic domains, signal peptides removed, of the three starting cellulases, E. carotovora, CeIN, CeIA, and CeIV, were shuffled and introduced into a 2μ yeast expression vector under the repressible Met25 promoter

Mtx-Cel-Dex substrate. The cellulose substrate for the chemical complementation selection must incorporate the Mtx and Dex ligands, yet still be efficient substrates for cellulases. Since E. carotovora CeA, CeIN, and CeIV, belong to the family 5 subfamily 2 glycosylhydrolase, we took advantage of the extensive crystallographic data on the cellulase Bacillus agaradharens CeI 5 A, which shares 68% amino acid sequence identity with the starting cellulases, to guide the design of the Mtx-Cel-Dex substrate (Davies,et al.,1998, Biochemistry Υh 1926-32; Davies, et al., 1998, Biochemistry 37: 1 1707-13; Varrot, et al., 2000, J. MoI. Biol. 297: 819-28; Varrot, et al.₅ 2001, Acta Crystallogr. D Biol. Crystallogr. 57j_. 1739-42; Varrot & Davies, 2003, Acta Crystallogr. D Biol. Crystallogr. 59: 447-52).The high-resolution structure of B. agaradharens Cel5A predicts five subsites in the active site that accommodate five glucose units (-3, -2, -1, 1, 2). Given that in other cellulases four subsites (-2, -1 , 1, 2) contribute most to the binding energy, with the fifth contributing only slightly, the cellulase substrate was constructed with four saccharides (Schou et al., 1993, Eur. J. Biochem. 217: 947-53). The synthesis of Mtx-Cel-Dex was carried out chemoenzymatically. Mtx-Lac-F and Dex-Cel were synthesized using previously published strategies and the linkage between the two halves was performed by Cel7B:E197A glycosynthase (Lin et al., 2004, J. Am. Chem. Soc. 126:15051-9).

Reverse yeast three-hybrid strain construction. Previously, chemical complementation has been adapted to detect bond cleavage reactions as a screen using lacZ as the reporter gene (Baker, K. et al., 2002, Proc. Natl. Acad. Sci. U SA 99_i 16537-42). Modifying chemical complementation to detect bond cleavage reactions as a growth selection is not trivial because, in a counter-selection, basal transcription of the toxic reporter gene may stress the cells, not only affecting their growth but also decreasing the dynamic range of the system. To tailor chemical complementation to detect cellulase activity as a growth selection, a counter-selection yeast strain had to be developed in which enzyme catalysis relieves transcription of a toxic reporter gene. In yeast, the URA3 gene product, OMP decarboxylase, converts 5-FOA to 5-FU, which is toxic to the cell, allowing for counter selection in the presence of 5-FOA. With the idea of carrying the counter selection in liquid culture, we developed a URA 3 reverse yeast three-hybrid strain that performs efficiently in this medium. Precedent in the literature of a URA3 reverse yeast two-hybrid strain supported the feasibility of constructing a URA3 reverse yeast three-hybrid strain (Vidal, et al., 1996, Proc. Natl. Acad. Sci. USA 93: 10315-20; Huang, et al., 1997, Proc. Natl. Acad. Sci. USA 94: 13396-401).

To reduce the basal expression of the toxic URA3 reporter gene, the gene was placed under control of the tightly regulated Spol3 promoter(Vidal, et al., 1996, Proc. Natl. Acad. Sci. USA 93_i 10315-20), in front of which four LexA operons (LexA(4op)) would have to be placed in order for the DBD to be recruited upstream of the reporter gene. One-step introduction of a LexA(4op)-Spol3-URA3 construct into our chemical complementation yeast strain was not convenient because our yeast strain has a ura3-52 background, conferring a Ura^* phenotype. Simple introduction of the LexA(4op)-Spol3-URA3 construct into this strain would also lead to a Ura^" phenotype. As there is no phenotypic change, this introduction cannot be selected or screened. Therefore, we resorted to a three-step approach to construct the reverse yeast three-hybrid strain (FIGURE 26A).

First, we recombined the URA3 gene at the ura3-52 locus of our chemical complementation yeast strain, FY251, and selected for Ura⁺ phenotype on medium lacking uracil to create VC2169. Next, the vectors carrying the DBD-DHFR and GR-B42 constructs were introduced into VC2169 and selected on plates lacking uracil, histidine, and tryptophan to create VC2291. Finally, we introduced the LexA(4op) and the Spol 3 promoter upstream of the URA3 gene and selected for Ura^" phenotype on medium containing 5-FOA. This selection provided numerous reverse three-hybrid colonies. Three colonies were used to optimize the reverse yeast three- hybrid system for the highest fold increase in activation in the presence of 1 OμM Mtx- Dex by measuring the difference between cell survival and cell death. Mtx-Dex was used, instead of Mtx-Cel-Dex, for the optimization of the system as the cellulose linker is not necessary for the system's functioning, and the synthesis of Mtx-Dex is simpler than that of Mtx-Cel-Dex. The system was optimized by varying the level of galactose, which affects the concentration of DBD-DHFR and AD-GR in the system, and 5-FOA, which controls the toxicity level of the reporter gene. Galactose levels ranging from 1 to 2% and 5-FOA concentrations ranging from 0.1 to 0.5% were tested simultaneously. The largest fold activation was obtained with synthetic media containing 2% galactose, 2% raffinose, 0.2% 5-FOA, and lOμM Mtx-Dex, and lacking histidine and tryptophan. We observed that 5-FOA concentrations beyond 0.3% are detrimental for cell growth, while reducing the concentration to 0.1% leads to cell growth independent of Mtx-Dex. At 0.2% 5-FOA, cell growth was Mtx-Dex dependent and the inherent toxicity of 5-FOA was not overly detrimental to the cell, thus allowing us to indirectly increase the stringency of the URA3 counter selection by increasing the 5-FOA concentration up to 0.3% if necessary. With the optimized conditions for the reverse three-hybrid system in hand, we collected two hundred and ten colonies from the reverse yeast three-hybrid selection to isolate the reverse yeast three-hybrid strain with the largest fold activation in liquid medium. We found that the increase in activation was strain dependent and that five days of selection generated strains with increases in activation spanning 0.7 to 5-fold. The ten strains with the highest increase in activation were tested in triplicate in the presence of 0 and lOμM Mtx-Dex. To determine the minimum small molecule concentration necessary to elicit transcription activation we tested the strain with the highest fold increase in activation, VC2204, with 0, 1, 5, and lOμM Mtx-Dex (FIGURE 26B). The reverse three-hybrid strain showed an on-off behavior between 1-1 OμM Mtx-Dex, meaning that with 1 μM Mtx-Dex all binding sites for the small-molecule are saturated, thus allowing us to carry the URA3 counter selection experiments at lμM small molecule concentration. We then tested VC2204 with 1 μM Mtx-Cel-Dex, the small-molecule to be used in the cellulase selection, to confirm activation of the system. At lμM Mtx- Cel-Dex, the reverse yeast three hybrid system shows a only 2-fold increase in activation, which we attribute to poor cell permeability of the Mtx-Cel-Dex substrate (FIGURE 26B). However, knowing that small changes in enzyme activity are amplified in liquid selections, we believed that a 2-fold increase in activation would be sufficient to detect cellulase activity.

Chemical complementation links cellulase activity to URA3 transcription in vivo. The reverse yeast three-hybrid system was modified to detect cellulase activity by introducing the cellulase from H. insolens Cel7B as a fourth component to the system. Using Mtx-Cel-Dex as the substrate for the cellulase, we optimized the URA3 counter selection to detect H. insolens Cel7B activity by changing the pH of the media, which controls enzyme behavior inside the cell. The activity of the cellulase in the URA3 counter selection was tested in media ranging from pH 4 through pH 7 in order to optimize the behavior of the cellulase inside the cell. Maximal cellulase activity was detected at pH 5. The final conditions for the URA3 counter selection were synthetic media at pH 5 containing 2% galactose, 2% raffinose, 0.2% 5-FOA, and lμM Mtx-Cel-Dex, and lacking histidine, tryptophan, and leucine. Using the optimized conditions for the URA3 counter selection, we showed that expression of H. insolens Cel7B in the presence of Mtx-Cel-Dex conferred a growth advantage to the reverse yeast three-hybrid selection strain, presumably because the Cel7B cellulase catalyzed the cleavage of Mtx-Cel-Dex, thus halting the expression of the toxic URA3 reporter gene. We performed this experiment in liquid culture as we were planning to carry out the directed evolution of cellulases in liquid culture. FIGURE 27A shows cell density measured as absorption at 600nm in the absence of Mtx-Cel-Dex and the presence of empty vector, in the presence of lμM Mtx-Cel-Dex and presence of active Cel7B, in the presence of lμM Mtx-Cel-Dex and the presence of an inactive mutant of Cel7B (Cel7B:E197A), obtained by mutation of the nucleophilic glutamic acid to alanine, and in the presence of lμM Mtx-Dex and the presence of empty vector. Under selection conditions, presence of empty vector in the absence of Mtx-Cel-Dex leads to cell growth, hence an enhanced absorption at 600nm. Next, we showed that transcription activation is dependent on the catalytic activity of Cel7B. The reverse yeast three-hybrid strain expressing either Cel7B or the inactive cellulase variant Cel7B:E197A was grown in the presence of IuM Mtx-Cel- Dex. Cellulase activity cleaves Mtx-Cel-Dex, disrupting dimerization and halting transcription of the URA3 reporter gene, leading to increased cell growth and hence an increase in OD_6OO- The inactive cellulase mutant Cel7B:E197A does not cleave the Mtx-Cel-Dex substrate, leading to transcription of the toxic URA3 reporter gene and hence a lack of increase in ODβoo- Finally, presence of empty vector in the presence of 1 μM Mtx-Dex leads to cell death, hence a lack of increase in ODβoo- This second control was performed as a redundancy, with the more readily synthesized Mtx-Dex instead of Mtx-Cellulose-Dex as the more stringent control, inactive cellulase, was performed in the presence of Mtx-Cellulose-Dex. Together these data indicate that Cel7B activity can be detected using chemical complementation. To further confirm that the URA3 counter selection could effectively detect cellulase activity in liquid culture, we carried out an enrichment assay for cellulase activity (FIGURE 27B). A library containing 1 :100 ratio of Cel7B to inactive Cel7B:E197A variants was transformed into the reverse yeast three-hybrid strain. To facilitate the colony analysis, a unique Ncol restriction site was introduced in the Cel7B:E197A variant at the A 197 position. Before selection, nine colonies were analyzed by colony PCR and restriction mapping to determine the integrity of the library. As expected from a 1 :100 ratio, all nine colonies were Cel7B:E197A. After five days of selection in liquid culture, a sample of the selection was plated, and nine random colonies were analyzed. Of these nine colonies, six contained Cel7B and three contained the Cel7B:E197A inactive cellulase variant, which corresponds to an enrichment of 200-fold enrichment after a single round of selection.

Directed evolution of cellulases. The URA3 counter selection was used to evolve cellulases with increased catalytic efficiency. We used DNA family shuffling as the mutagenesis technique because it is technically simple and has been used to obtain the largest increases in catalytic activity in recent years(Castle, et al., 2004, Science 304: 1 151-4; Park, et al., 2006, Science 3JLL: 535-8). Therefore, we acquired three cellulases, E. carotovora CeIN, CeIA, and CeIV, which share 77-90% sequence identity at the nucleotide level, to generate a library of cellulase chimeras. To reach a large library size, we took advantage of the efficient recombination machinery of 5. cerevisiae to introduce the library of shuffled cellulases into the expression vector via in vivo homologous recombination (Colby, et al., 2004, Methods Enzymol. 388: 348-58; Abecassis, et al., 2000, Nucleic. Acids. Res. 28: E88). The shuffled cellulase library was amplified using primers carrying 30 base pairs identical to the promoter and terminator regions of the expression vector to ensure successful recombination. The shuffled cellulase library was co-transformed with linearized expression vector via electroporation to produce a library of 10⁸.

The selection was run for a total of nine days. Twenty-two variants were isolated before selection and twenty-two variants were isolated after five days of selection. To determine whether longer exposure to selection conditions leads to the isolation of cellulase variants with even higher catalytic activities, we also isolated eight variants after six days of selection and eight variants after eight days of selection. It is worth noting that we could have isolated any number of cellulase variants on any day of the selection as the selection was run in liquid culture and required only the taking of a small sample and its plating under non-selective conditions to isolate more cellulase variants. Sequencing of the 40 variants isolated after selection showed an average of 5 to 12 crossovers. To determine whether the URA3 counter selection was indeed selecting for variants with increased cellulase activity, we retransformed the 62 variants into the URA3 counter selection strain and tested the crude cell extract for cellulase activity using a colorimetric screen based on the quantification of aldehyde formation using carboxymethylcellulose (CMC) as the substrate (Jue et al., 1985, J. Biochem. Biophys. Methods. U \ 109-15). As FIGURE 27C shows, the mean cellulase activity of the twenty-two variants isolated after five days of selection is higher than the mean cellulase activity of variants isolated before selection. This difference is statistically significant (p<0.005), indicating that that the URA 3 counter selection is indeed selecting for cellulase variants with increased catalytic activity. There is no statistical difference, however, between the cellulase activities of colonies isolated after five days of selection and those isolated after six and eight days of selection. In Vitro Characterization of Evolved Cellulase Variants. We carried out the in vitro kinetic characterization of four cellulases isolated on day eight using pNPC, the standard substrate for measuring cellulase kinetics. The four variants were overexpressed in E. coli and purified using a Ni-NTA affinity column. Two variants had very low expression levels with the conditions used and no in vitro kinetics were measured. The other two variants, Cel_3.7 (cellulase 1) and Cel_5.7 (cellulase 2) showed very good expression levels and 3.7-fold (cellulase 1) and 5.7- fold (cellulase 2) increases in catalytic efficiency, respectively, over the best starting cellulase, CeIN . Table 4 shows the kinetics for the starting cellulase genes and the improved cellulase variants.

Table 4.

Kinetics of the parent genes CeIN, CeIA, CeIV and the evolved variant on pNPC Cellulase k_cat (sec^"1) Km (M) K_c8IZKIn (M ¹SeC ¹)

Cellulase N 1.3xl0^J 3.4x10 ^J 3.7x10⁵

Cellulase A 1.4x10³ 7.2 xl 0^'3 1.92 xlO⁵

Cellulase V l.lxlO³ 4.6 xlO-³ 2.4 xlO⁵

Cel_3.7 (cellulasel) 4.3xlO³ 3.2 xlO^'3 1.4x10⁶

Cel_5.7 (cellulase 5.2x10³ 2.47 xlO^"3 2.IxIO⁶

2)

CELV=CELT from before. Between last year and now they actually name that variant CELV.

Sequencing of the improved cellulase variants revealed nucleotide segments from CeIV, CeIA and CeIN. Cel_3.7 (cellulase 1) contains nine crossovers. When compared to the most active starting cellulase, CeIN, Cel_3.7 (cellulase 1) differs by nine point mutations: S126N, D157E, Q186L L193M, V254I, T288A, E295A, G30A, and G304S. A homology model of Cel_3.7 (cellulase 1) was made, utilizing the B. agaradherans Cel5A structure (SwissProt: O85465; PDB: Ih5v) with the inhibitor methyl 4,4II,4III,4IV-tetrathio-α-cellopentoside bound to the active site, using Swiss-Model (FIGURE 28A). The eight mutations accessible with the homology model map onto the surface of the enzyme, away from the extended carbohydrate binding site. Although the B. agaradharens CeI 5 A X-ray structure does not cover the region where the ninth mutation is positioned, we manually extended the chain by three more residues and concluded that the last mutation is on the N- terminus of the TIM barrel and away from the C-terminus catalytic region.

When Cel_5.7 (cellulase 2) is compared to the most active starting cellulase, CeIN, Cel_5.7 differs by eight point mutations: A86V, S126N, E154D, T172S, C208R, T262S, A267T, and A272T. A homology model of Cel_5.7 (cellulase 2) was made, again utilizing the same structure of B. agaradherans Cel5A, using Swiss-Model (Fig. 28b)³⁸. Four mutations are in the C-teminus loops that compose the extended carbohydrate binding site, while the other 4 map onto the surface of the enzyme. Although mutation T172S is located in C-terminus loop 5, it is 15A away from closest hexose ring in the substrate inhibitor. Mutations T262S, A267T, A272T are located in C-terminus loop 8. The mutation T262S is the closest to the substrate inhibitor, 6A away from the 6'-OH of the hexose ring in the -3 subsite. Interestingly, only one mutation is present in both of the evolved variants, S126N. This mutation is located at the N-terminus of the barrel and points towards the solvent.

Various publications are cited herein, the contents of which are hereby incorporated by reference in their entireties.

/ / / /

/ J l

EC Enzyme Linkage Substrate

3.2.1.1 Alpha-amylase 1,4-alpha-D-glucosidic maltopentaose

3.2.1.2 Beta-amylase 1,4-beta-D-glucosidic maltotetraose

3.2.1.3 Glucan 1,4-alpha-glucosidase terminal 1,4-linked alpha-D-glucose cyclodextrin

3.2.1.4 Cellulase 1,4-beta-D-glucosidic cellotetraose

1,3- or 1,4-beta-D-glucans, requires C-3 substitution of glucose whose reducing NAG-NAG-BMA-MAN-MAN

3.2.1.6 Endo-l,3(4)-beta-glucanase group is involved in the hydro Iy zed linkage MAN-MAN

3.2.1.7 Inulinase 2,1-beta-D-fructosidic inulin

3.2.1.8 Endo- 1 ,4- beta-xy lanase 1,4-beta-D-xylosidic xylobiose

3.2.1.10 Oligo-l,6-glucosidase 1,6-alpha-D-glucosidic isomaltose

3.2.1.11 Dextranase 1,6-alpha-D-glucosidic dextran

3.2.1.14 Chitinase N-acetyl-beta-D-glucosaminide 1,4-beta- chitin

3.2.1.15 Polygalacturonase 1,4-alpha-D-galactosiduronic pectate

1,4-beta-linkages between N-acetylmuramic acid and N-acetyl-D-glucosamine residues in a peptidoglycan, and between N-acetyl- D

3.2.1.17 Lysozyme glucosamine residues in chitodextrins alpha-(2->3)-_; alpha-(2->6)-, alpha-(2->8)- glycosidic linkages of terminal sialic

3.2.1.18 Exo-alpha-sialidase residues terminal, non-reducing 1,4-linked alpha-D-

3.2.1.20 Alpha-glucosidase glucose

3.2.1.21 Beta-glucosidase terminal, non-reducing beta-D-glucose dimboa-beta-d-glucoside

3.2.1.22 Alpha-galactosidase terminal, non-reducing alpha-D-galactose melibiose terminal non-reducing beta- D-ga lactose

3.2.1.23 Beta-galactosidase residues in beta-D- galactosides lactose terminal, non-reducing alpha-D-mannose

3.2.1.24 Alpha-mannosidase residues in alpha-D- mannosides

3.2.1.25 Beta-mannosidase terminal, non-reducing beta-D-mannose terminal non-reducing beta-D-

3.2.1.26 Beta-fructofuranosidase fructofuranoside raffinose

3.2.1.28 Alpha,alpha-trehalase alpha, alha-trehalose alpha,alha-trehalose

3.2.1.31 Beta-glucuronidase beta-D-glucuronoside beta-D-glucuronoside

1,3-beta-D-glycosidic linkages in 1,3-beta-D

3.2.1.32 Xylan endo-l,3-beta-xylosidase xylans l->6)-alpha-D-glucosidic branch linkages in

3.2.1.33 Amylo-alpha-l,6-glucosidase glycogen phosphorylase limit dextrin l->4-linkages between N-acetyl-beta-D- glucosamine and D-glucuronate residues in

3.2.1.35 Hyalurononglucosaminidase hyaluronate hyaluronan

1,3-linkages between beta-D-glucuronate and N-acetyl-D-glucosamine residues in

3.2.1.36 Hyaluronoglucuronidase hyaluronate hyaluronan

1,3-beta-D-glycosidic linkages in 1,3-beta-D

3.2.1.37 Xylan 1,4-beta-xylosidase xylans terminal non-reducing beta^:D-fucose

3.2.1.38 Beta-D-fucosidase residues in beta-D- fucosides

1,3-beta-D-glucosidic linkages in 1,3-beta-

3.2.1.39 Glucan endo-l,3-beta-D-glucosidase D-glucans laminariπ, laminaribiose terminal non-reducing alpha-L-rhamnose

3.2.1.40 Alpha-L-rhamnosidase residues in alpha-L- rhamnosides

(l->6)-alpha-D-glucosidic linkages in pullulan and in amylopectin and glycogen, and the alpha- and beta-limit dextrins of

3.2.1.41 Pullulanase amylopectin and glycogen pullulan, maltotriose

3.2.1.42 GDP-glucosidase GDP-glucose GDP-glucose terminal, non-reducing beta-L-rhamnose

3.2.1.43 Beta-L-rhamnosidase residues in beta-L- rhamnosides

3.2.1.44 Fucoidanase 1,2-alpha-L-fucoside linkages in fucoidan fucoidan

3.2.1.45 Glucosylceramidase D-glucosyl-N-acylsphingosine D-glucosyl-N- D-galactosyl-N-

3.2.1.46 Galactosylceramidase D-galactosyl-N-acylsphingosiπe acylsphingosine D-galactosyl-D-galactosyl-D-glucosyl-N- D-galactosyl-D-galactosyl-

3.2.1.47 Galactosylgalactosylglucosylceramidase acylsphingosine D-glucosyl-N-

3.2.1.48 Sucrose alpha-glucosidase sucrose, maltose sucrose, maltose

terminal non-reducing N-acetyl-D- galactosamine residues in N-acetyl-alpha-D-

3.2.1.49 Alpha-N-acetylgalactosaminidase galactosaminides N-acetyl-d-glucosamine erminal non-reducing N-acetyl-D- glucosamine residues in N-acetyl-alpha-D-

3.2.1.50 Alpha-N-acetylglucosaminidase glucosaminides

3.2.1.51 Alpha-L-fucosidase alpha-L-fucoside alpha-L-fucoside terminal non-reducing N-acetyl-D- hexosamine residues in N-acetyl-beta-D-

3.2.1.52 Beta-N-acetylhexosaminidase hexosaminldes Di(n-acetyl-d-glucosamine) terminal non-reducing N-acetyl-D- galactosamine residues in N-acetyl-beta-D-

3.2.1.53 Beta-N-acetylgalactosaminidase galactosaminides

3.2.1.54 Cyclomaltodextrinase cyclomaltodextrin cyclomaltodextrin terminal non-reducing alpha-L- arabinofuranoside residues in alpha-L-

3.2.1.55 Alpha-N-arabinofuranosidase arabinosides arabinoxylan Glucuronosyl-disulfoglucosamine 3-D-glucuronosyl-N(2)-,6-disulfo-beta-D- 3-D-glucuronosyl-N(2)-,6-

3.2.1.56 glucuronidase glucosamine disulfo-beta-D-glucosamine

W 3.2.1.57 Isopullulanase panose panose

2,4-dinitrophenyl 2-deoxy- beta-D-glucose units from the non-reducing 2-fluoro-beta-d-

3.2.1.58 Glucan 1,3-beta-glucosidase ends of 1,3-beta-D-glucans glucopyranoside 1,3-alpha-D-glucosidic linkages in

3.2.1.59 Glucan endo-l,3-alpha-glucosidase isolichenin, pseudonigeran and nigeran 1,4-alpha-D-glucosidic linkages in

3.2.1.60 Glucan 1,4-alpha-maltotetraohydrolase amylaceous polysaccharides maltotetraose

1,4-alpha-D-glucosidic linkages in alpha-D-

3.2.1.61 Mycodextranase glucans containing both 1,3- and 1,4-bonds

3.2.1.62 Glycosylceramidase glycosyl-N-acylsphingosine glycosyl-N-acylsphingosine methyl-2-alpha-L- methyl-2-alpha-L-fucopyranosyl-beta-D- fucopyranosyl-beta-D-

3.2.1.63 1,2-alpha-L-fucosidase galactoside galactoside

3.2.1.64 2,6-beta-fructan 6-levanbiohydrolase 2,6-beta-D-fructofuranan 2,6-beta-D-fructofuranan

2,6-beta-D-fructofuranosidic linkages in 2,6- beta-D- fructans (levans) containing more

3.2.1.65 Levanase than 3 fructose units kestose?

3.2.1.66 Quercitrinase quercitrin quercitrin

3.2.1.67 Galacturan 1,4-alpha-galacturonidase (l,4-alpha-D-galacturonide)(n) (l->6)-alpha-D-glucosidic branch linkages in glycogen, amylopectin and their beta-

3.2.1.68 Isoamylase limit dextrins amylopectin (l->6)-alpha-D-glucosidic linkages in l->6-

3.2.1.70 Glucan 1,6-alpha-glucosidase alpha-D-glucans

1,2-glucosidic linkages in 1,2-beta-D-

3.2.1.71 Glucan endo-l,2-beta-glucosidase glucans

3.2.1.72 Xylan 1,3-beta-xylosidase non-reducing termini of 1,3-beta-D-xylans

1,4-beta-D-glucosidic linkages in beta-D-

3.2.1.73 Licheninase glucans containing 1,3- and 1,4-bonds lichenin?

3.2.1.74 Glucan 1,4-beta-glucosidase 1,4-beta-D-glucans

1,4-alpha-D-glucosidic linkages in

3.2.1.75 Glucan endo-l,6-beta-glucosidase amylaceous polysaccharides

3.2.1.76 L-iduronidase 1,6-beta-D-glucans lutean, pustulan mannan, galactomannan,

3.2.1.77 Mannan l,2-(l,3)-alpha-mannosidase 1,4-beta-D-mannosidic glucomannan mannan, galactomannan,

3.2.1.78 Mannan endo-l,4-beta-mannosidase 1,4-beta-D-mannosidic glucomannan terminal, non-reducing 2,1- and 2,6-linked

3.2.1.80 Fructan beta-fructosidase beta-D- fructofuranose inulin, levan, sucrose

3.2.1.81 Beta-agarase 1,4-beta-D-galactosidic linkages in agarose

3.2.1.82 Exo-poly-alpha-galacturonosidase pectic acid pectic acid

1,4-beta-D- between D-galactose 4-sulfate

3.2.1.83 Kappa-carrageenase and 3,6-anhydro-D-galactose kappa-carrageenans

3.2.1.84 Glucan 1,3-alpha-glucosidase terminal 1,3-alpha-D-glucosidic

6-phospho-beta-D-

3.2.1.85 6-phospho-beta-galactosidase 6-phospho-beta-D-galactoside galactoside

6-phospho-beta-D-glucosyl-(l,4)-D- 6-phospho-beta-D-

3.2.1.86 6-phospho-beta-glucosidase glucose glucosyl-(l,4)-D-glucose

1,3-alpha-D-galactosidic linkages in

Capsular-polysaccharide endo-l,3-alpha- Aerobacter aerogenes capsular

3.2.1.87 galactosidase polysaccharide

3.2.1.88 Beta-L-arabinosidase beta-L-arabinoside Arabinogalactan endo-l,4-beta- 1,4-beta-D-galactosidic linkages in

3.2.1.89 galactosidase arabinogalactans arabinogalactan

1,4-beta-D-glucosidic linkages in cellulose

3.2.1.91 Cellulose 1,4-beta-cellobiosidase and cellotetraose cellotetraose

3.2.1.92 Peptidoglycan beta-N-acetylmuramidase terminal, non-reducing N-acetylmuramic

3.2.1.93 Alpha, alpha-phosphotrehalase Alpha, alpha-trehalose 6-phosphate

3.2.1.94 Glucan 1,6-alpha-isomaltosidase 1,6-alpha-D-glucosidic

3.2.1.95 Dextran 1,6-alpha-isomaltotriosidase 1,6-alpha-D-glucosidic dextran high-mannose giycopeptides and

Mannosyl-glycoprotein endo-beta-N- glycoproteins containing

3.2.1.96 acetylglucosaminidase N,N'-diacetylchitobiosyl the -(Man(GlcNAc)(2))Asn- Glycopeptide alpha-N- terminal D-galactosyl-N-acetyl-alpha-D- glycopeptides,

3.2.1.97 acetylgalactosaminidase galactosaminidic glycoproteins

3.2.1.98 Glucan 1,4-alpha-maltohexaosidase 1,4-alpha-D-glucosidic amylaceous

3.2.1.99 Arabinaπ endo-l,5-alpha-L-arabinosidase 1,5-alpha-arabinofuranosidic 1,5-arabinans

3.2.1.100 Maπnan 1,4-mannobiosidase 1,4-beta-D-mannosidic 1,4-mannans

3.2.1.101 Maπnan endo-l,6-alpha-mannosidase 1,6-alpha-D-mannosidic 1,6-mannans Blood-group-substance endo-l,4-beta- blood group A and b

3.2.1.102 galactosidase 1,4-beta-D-galactosidic substances

3.2.1.103 Keratan-sulfate endo-l,4-beta- 1,4-beta-D-galactosidic keratan sulfate

3.2.1.104 Steryl-beta-glucosidase cholesteryl-beta-D-glucoside cholesteryl-beta-D-

3.2.1.105 Strictosidine beta-glucosidase strictosidine strictosidine

3.2.1.106 Mannosyl-oligosaccharide glucosidase non-reducing terminal glucose Glc(3)Man(9)GlcNAc(2) protein alpha-D-glucosyl-

Protein-glucosylgalactosylhydroxylysine protein alpha-D-glucosyl-l,2-beta-D- 1,2-beta-D-galactosyl-L-

3.2.1.107 glucosidase galactosyl-L-hydroxylysine hydroxylysine

3.2.1.108 Lactase lactose lactose poly(D-galactosamine)

3.2.1.109 Endogalactosaminidase 1,4-alpha-D-galactosaminidic (e.g. hyaluronan)

D-galactosyl-3-(N-acetyl-

D-galactosyl-3-(N-acetyl-beta-D- beta-D-galactosaminyl)-L-

3.2.1.110 Mucinaminylserine mucinaminidase galactosaminyl)-L-serine seriπe 1,3-linkages between alpha-L-fucose and N-

3.2.1.111 1,3-alpha-L-fucosidase acetylglucosamine glycoproteins

3.2.1.112 2-deoxyglucosidase 2-deoxy-alpha-D-glucoside 2-deoxy-alpha-D-glucoside

3.2.1.113 Mannosyl-oligosaccharide 1,2-alpha- terminal 1,2-linked alpha-D-mannose Man(9)(GlcNAc)(2) Mannosyl-oligosaccharide 1,3-1,6-alpha- terminal 1,3- and 1,6-linked alpha-D-

3.2.1.114 mannosidase mannose Man(5)(GlcNAc)(3)

3.2.1.115 Branched-dextran exo-l,2-alpha- 1,2-alpha-D-glucosidic, at branch point d extra n

3.2.1.116 Glucan 1,4-alpha-maltotriohydrolase 1,4-alpha-D-glucosidi amylaceous

3.2.1.117 Amygdalin beta-glucosidase (R)-amygdalin (R)-amygdalin

3.2.1.118 Prunasin beta-glucosidase (R)-prunasin (R)-prunasin

3.2.1.119 Vicianin beta-glucosidase (R)-vicianin (R)-vicianin

3.2.1.120 Oligoxyloglucan beta-glycosidase 1,4-beta-D-glucosidic oligoxyloglucans

3.2.1.121 Polymannuronate hydrolase D-mannuronide polymannuronate

3.2.1.122 Maltose-6'-phosphate glucosidase maltose 6'-phosphate maltose 6'-phosphate

3.2.1.123 Endoglycosylceramidase oligoglycosylglucosylceramide oligoglycosylglucosylcerami

3-deoxy-D-manno-2- octulosonate in capsular

3.2.1.124 3-deoxy-2-octulosonidase beta-ketopyranosidic polysaccharides

3.2.1.125 Raucaffricine beta-glucosidase raucaffricine raucaffricine

3.2.1.126 Coniferiri beta-glucosidase coniferin coniferin

1,6- between alpha-L-fucose and N-acetyl-D

3.2.1.127 1,6-alpha-L-fucosidase glucosamine glycopeptides

3.2.1.128 Glycyrrhizinate beta-glucuronidase glycyrrhizinate glycyrrhizinate

3.2.1.129 Endo-alpha-sialidase (2->8)-alpha-sialosyl oligo- or poly(sialic) acids

3.2.1.130 Glycoprotein endo-alpha-l,2-mannosidase alpha-D-glucosyl-(l,3)-D-mannosyl GlcMan(9)(GlcNAc)(2)

3.2.1.131 Xylan alpha-l,2-glucuronosidase alpha-D-l,2-(4-0-methyl)glucuronosyl hardwood xylan beta-l,4-linkages between

3.2.1.132 Chitosanase beta-1,4- between D-glucosamine residues D-glucosamine residues

3.2.1.133 Glucan 1,4-alpha-maltohydrolase (l->4)-alpha-D-glucosidic

Bis-D-fructose 2",1 :2,1'-

3.2.1.134 Difructose-anhydride synthase Bis-D-fructose 2',l:2,r-dianhydride dianhydride

3.2.1.135 Neopullulanase pullulan pullulan

3.2.1.136 Glucuronoarabinoxylan endo-l,4-beta- 1,4-beta-D-xylosyl some

3.2.1.137 Mannan exo-l,2-l,6-alpha-mannosidase 1,2-alpha-D- and 1,6-alpha-D- yeast mannan 3.2.1.139 Alpha-glucuronidase alpha-D-glucuronoside alpha-D-glucuronoside Beta-D-Gal-(l->3)-beta-D-

Beta-D-Gal-(l->3)-beta-D-GlcNAc-(l->3)- GlcNAc-(l->3)-beta-D-Gal-

3.2.1.140 Lacto-N-biosidase beta-D-Gal-(l->4)-D-Glc (l->4)-D-Glc 4-alpha-D-((l->4)-alpha-D- 4-alpha-D- ((l->4)-alpha-

3.2.1.141 glucano)trehalose trehalohydrolase alpha-(l->4)-D-glucosidic D-glucanosyl)(n) trehalose alpha- and beta-limit dextrins of amylopectin and glycogen, and in

3.2.1.142 Limit dextrinase (l->6)-alpha-D-glucosidic amylopectin and pullulan

3.2.1.143 Poly(ADP-ribose) glycohydrolase poly(ADP-ribose) at glycosidic (l"-2')

3-deoxyoctulosonyl-

3.2.1.144 3-deoxyoctulosonase 3-deoxyoctulosonyl-lipopolysaccharide lipopolysaccharide

(l->3)-beta-D-

3.2.1.145 Galactan 1,3-beta-galactosidase terminal, non-reducing beta-D-galactose galactopyranans alpha 6-O-(beta-D- alpha 6-O-(beta-D-xylopyranosyl)-beta-D- xylopyranosyl)-beta-D-

3.2.1.146 Beta-galactofuranosidase glucopyranoside glucopyranoside

3.2.1.147 Thioglucosidase thioglucoside thioglucoside alpha 6-O-(beta-D- alpha 6-O-(beta-D-xylopyranosyl)-beta-D- xylopyranosγl)-beta-D-

3.2.1.149 Beta-primeverosidase glucopyranoside glucopyranoside

3.2.1.150 Oligoxyloglucan reducing-end-specific

3.2.1.151 Xyloglucan-specific endo- beta- 1,4- 1,4- beta-D-glucosidic xyloglucan alpha-D-mannosyl-(l->6)-beta-D-mannosyl'

Mannosylglycoprotein endo-beta- (l->4)-beta-D- N-acetylglucosaminyl-(l-

3.2.1.152 mannosidase >4)-beta-D-N-acetylglucosaminyl terminal, non-reducing 2,1-linked beta-D-

3.2.1.153 Fructan beta-(2,l)-fructosidase fructofuranose fructan terminal, non-reducing 2,6-linked beta-D-

3.2.1.154 Fructan beta-(2,6)-fructosidase fructofuranose fructan

3.2.1.155 Xyloglucan-specific exo-beta-l,4-glucanase 1,4- beta-D-glucosidic xyloglucan

1,4-beta-D-xylose residues from reducing

3.2.1.156 Oligosaccharide reducing-end xylanase end of oligosaccharides

1,4-beta-D- between D-galactose 4-sulfate

3.2.1.157 Iota-carrageenase and 3,6-anhydro-D-galactose-2-sulfate iota-carrageenans

3.2.1.158 Alpha-agarase 1,3-alpha-L-galactosidic agarose

3.2.1.159 Alpha-neoagaro-oligosaccharide hydrolase 1,3-alpha-L-galactosidic neoagaro- oligosaccharides

3.2.1.160 Xyloglucan-specific exo-beta-l,4-glucanase xyloglucan xyloglucan

7-(beta-D-apiofuranosyl- (l->6)-beta-D-

7-(beta-D-apiofuranosyl-(l->6)-beta-D- glucopyranosyloxy)isoflavo

3.2.1.161 Beta-apiosyl-beta-glucosidase glucopyranosyloxy)isoflavonoid noid beta- 1,4- in the backbone of lambda-

3.2.1.162 Lambda-carrageenase carrageenan lambda-carrageenan

OO

EC Enzyme Linkage Substrate Notes

4.2.2.1 Hyaluronate lyase hyaluronate hyaluronate

(l->4)-alpha-D-

4.2.2.2 Pectate lyase (l->4)-alpha-D-galacturonan galacturonan beta-D-mannuronate containing

4.2.2.3 Poly(beta-D-mannuronate) lyase polysaccharides

1,4-beta-D- hexosaminyl and 1,3-

4.2.2.5 Chondroitin Ac lyase beta-D-glucuronosyl

4-(4-deoxy-beta-D-

4-(4-deoxy-beta-D-gluc-4- gluc-4-enuronosyl)-

4.2.2.6 Oligogalacturonide lyase enuronosyl)-D-galacturonate D-galacturonate

1,4-linked D-glucuronate or L- iduronate residues and 1,4-alpha- linked 2-sulfoamino- 2-deoxy-6-

4.2.2.7 Heparin lyase sulfo-D-glucose between N-acetyl-D- glucosamine

4.2.2.8 Heparin-sulfate lyase and uronate

4-(4-deoxy-alpha-D-galact-4-

S 4.2.2.9 Pectate disaccharide-lyase eπuronosyl)-D- galacturonate pectate

(l->4)-alpha-D-galacturonan

4.2.2.10 Pectin lyase methyl ester

4.2.2.11 Poly(alpha-L-guluronate) lyase terminal alpha-L- guluronate terminal beta-D-mannosyl-beta-

4.2.2.12 Xanthan lyase D-1,4- glucuronosyl xanthan

4.2.2.13 Exo-(l->4)-alpha-D-glucan lyase linear alpha-glucan linear alpha-glucan

4.2.2.14 Glucuronan lyase (l->4)-beta-D-glucuronans

N-acetylneuraminic acid

4.2.2.15 Anhydrosialidase glycosides terminal D-fructosyl-D-fructosyl

4.2.2.16 Levan fructotransferase (DFA-IV-forming) disaccharide terminal D-fructosyl-D- fructosyl

4.2.2.17 Inulin fructotransferase (DFA-I-forming) disaccharide terminal D-fructosyl-D-fructosyl

4.2.2.18 Inulin fructotransferase (DFA-III-forming) disaccharide

dermatan sulfate containing 1,4- beta-D- hexosaminyl and 1,3- beta-D-glucurosonyl or 1,3-alpha-

4.2.2.19 Chondroitin B L-iduronosyl beta-l,4-galactosaminic between N-acetylgalactosamine and either D-glucuronic acid or L-iduronic

4.2.2.20 Chondroitin-sulfate-ABC endolyase acid non-reducing ends of both polymeric chondroitin sulfates and their oligosaccharide

4.2.2.21 Chondroitin-sulfate-ABC exolyase fragments

Kt

O

Claims

WE CLAIM:

1. A system for directing evolution of a biological entity of interest comprising a plurality of host cells, such that each host cell:

(i) contains a nucleic acid target corresponding to the entity of interest;

(ii) contains at least one diversifying nucleic acid ("DINA") which can be used to direct mutation of the target; and

(iii) exhibits at least one mutator function which increases the likelihood that the target will be mutated by the DINA; wherein a plurality of host cells contain a plurality of different DFNAs, and wherein said system further comprises a means for selecting for a mutant target exhibiting a desired activity.

2. The system of claim 1 , wherein the host cell is selected from the group consisting of a prokaryotic cell and a eukaryotic cell.

3. The system of claim 2, wherein the host cell is selected from the group consisting of a yeast cell, a bacterial cell, an insect cell, a plant cell, a unicellular slime mold cell, an invertebrate cell, a vertebrate cell, and a mammalian cell.

4. The system of claim 1, wherein the entity of interest is selected from the group consisting of a protein, a functional RNA, a promoter element, an intron, a ribosome binding site, an endonuclease recognition site, a repressor element, and a molecular pathway to generating a product of interest.

5. The system of claim 1, which comprises more than one nucleic acid target.

6. The system of claim 1, wherein the mutator function is selected from the group of homologous recombination, a molecule which facilitates homologous recombination, and gene conversion.

7. The system of claim 1, further comprising an ancillary mutator function selected from the group consisting of an error-prone polymerase and a transposable element.

8. The system of claim 1 , wherein the means of selecting a mutant target exhibiting a desired activity is a detectable reporter gene.

9. The system of claim 1, wherein the means of selecting a mutant target gene exhibiting a desired activity is a toxic reporter gene.

10. The system of claim 1 , wherein the means for selecting a mutant target exhibiting a desired activity is a chemical complementation system.

1 1. A primary library of variants of a nucleic acid target corresponding to an entity of interest, produced using a system for directing evolution of a biological entity of interest comprising a plurality of host cells, such that each host cell:

(i) contains a nucleic acid target corresponding to the entity of interest;

(ii) contains at least one diversifying nucleic acid ("DINA") which can be used to direct mutation of the target; and

(iii) exhibits at least one mutator function which increases the likelihood that the target will be mutated by the DINA; wherein a plurality of host cells contain a plurality of different DINAs, and wherein said system further comprises a means for selecting for a mutant target exhibiting a desired activity.

12. The primary library of claim 1 1 , wherein the host cell is selected from the group consisting of a prokaryotic cell and a eukaryotic cell.

13. The primary library of claim 12, wherein the host cell is selected from the group consisting of a yeast cell, a bacterial cell, an insect cell, a plant cell, a unicellular slime mold cell, an invertebrate cell, a vertebrate cell, and a mammalian cell.

14. The primary library of claim 1 1, wherein the entity of interest is selected from the group consisting of a protein, a functional RNA, a promoter element, an intron, a ribosome binding site, an endonuclease recognition site, a repressor element, and a molecular pathway to generating a product of interest.

15. The primary library of claim 1 1 , wherein the mutator function is selected from the group of homologous recombination, a molecule which facilitates homologous recombination, and gene conversion.

16. The primary library of claim 1 1 , which comprises more than one nucleic acid target.

17. The primary library of claim 1 1 , further comprising an ancillary mutator function selected from the group consisting of an error-prone polymerase and a transposable element.

18. The primary library of claim 1 1 , wherein the means of selecting a mutant target exhibiting a desired activity is a detectable reporter gene.

19. The primary library of claim 1 1 , wherein the means of selecting a mutant target gene exhibiting a desired activity is a toxic reporter gene.

20. The primary library of claim 11, wherein the means for selecting a mutant target exhibiting a desired activity is a chemical complementation system.

21. A secondary library of variants of a nucleic acid target of an entity of interst, produced by genetic exchange between a plurality of primary libraries according to any one of claims 1 1 through 20.

22. A system for improving the activity of a biomass converting enzyme, comprising a plurality of host cells, wherein each host cell

(a) contains a nucleic acid encoding a variant of the biomass converting enzyme; and

(b) contains a chemical complementation system comprising: (i) a DNA binding domain comprised in a first fusion molecule which further comprises a first receptor for a first ligand;

(ii) an activation domain comprised in a second fusion molecule which further comprises a second receptor for a second ligand; and

(iii) the substrate for the biomass converting enzyme comprised in a fusion molecule which comprises the first ligand and the second ligand, such that cleavage of the substrate prevents the DNA binding domain and activation domain from being joined and triggering expression of a toxic reporter gene; wherein the variant of the biomass converting enzyme is generated in vivo by the operation of a mutator function on a diversifying nucleic acid.

23. The system of claim 22, produced by the genetic exchange of two primary libraries of biomass converting enzyme variants.

24. The system of claim 22 or claim 23, wherein the biomass converting enzyme is cellulase.

25. The system of claim 22 or claim 23, wherein the biomass converting enzyme is selected from the list of enzymes in Table 5A or Table 5B.

26. A primary library of variants of a nucleic acid target corresponding to an entity of interest, produced using a system for improving the activity of a biomass converting enzyme, comprising a plurality of host cells, wherein each host cell

(a) contains a nucleic acid encoding a variant of the biomass converting enzyme; and

(b) contains a chemical complementation system comprising: (i) a DNA binding domain comprised in a first fusion molecule which further comprises a first receptor for a first ligand;

(ii) an activation domain comprised in a second fusion molecule which further comprises a second receptor for a second ligand; and

(iii) the substrate for the biomass converting enzyme comprised in a fusion molecule which comprises the first ligand and the second ligand, such that cleavage of the substrate prevents the DNA binding domain and activation domain from being joined and triggering expression of a toxic reporter gene; wherein the variant of the biomass converting enzyme is generated in vivo by the operation of a mutator function on a diversifying nucleic acid.

27. The primary library of claim 26, produced by the genetic exchange of two primary libraries of biomass converting enzyme variants.

28. The primary library of claim 26 or claim 27, wherein the biomass converting enzyme is cellulase.

29. The primary library of claim 26 or claim 27, wherein the biomass converting enzyme is selected from the list of enzymes in Table 5 A or Table 5B.

30. A secondary library of variants of a nucleic acid target of an entity of interst, produced by genetic exchange between a plurality of primary libraries according to any one of claims 26 through 29.

31. A cellulase variant comprising the amino acid sequence, SEQ ID NO:4.

32. A cellulase variant comprising the amino acid sequence, SEQ ID NO:6.