WO2005116271A2

WO2005116271A2 - Selection of microorganisms with growth dependent upon extracytoplasmic enzymes

Info

Publication number: WO2005116271A2
Application number: PCT/US2005/018430
Authority: WO
Inventors: Zhilian Fan; John Mcbride; Lee R. Lynd
Original assignee: The Trustees Of Dartmouth College
Priority date: 2004-05-25
Filing date: 2005-05-25
Publication date: 2005-12-08
Also published as: WO2005116271A3

Abstract

A method for selection of microorganisms with improved growth on non-native substrates is disclosed. Strains expressing tethered, enzymes on their cell surface are grown on non-native substrates and mutants with improved growth characteristics relative to the parent strain can be advantageously selected. Both soluble and insoluble substrates may be employed.

Description

SELECTION OF MICROORGANISMS WITH GROWTH DEPENDENT UPON

EXTRACYTOPLASMIC ENZYMES

RELATED APPLICATIONS

This application claims the benefit of priority to U.S. provisional patent application serial no. 60/574,147, filed May 25, 2004, which is incorporated by reference herein.

GOVERNMENT RIGHTS

The U.S. government has certain rights in this invention as provided for by the terms of Grant No. 60NANB1 D0064, awarded by the National Institute of Standards and Technology.

BACKGROUND

Natural selection is primarily responsible for the properties of organisms found in natural environments. Improved laboratory selection processes are potentially a powerful approach for development of new microbial strains with desired properties. The general concept behind laboratory selection processes is to present a microorganism with an environmental challenge, such as the presence of a non-native growth substrate, and to select for organisms that demonstrate an ability to overcome the challenge. A variety of techniques are available to select for microorganisms in culture. Continuous culture techniques include those where a culture is maintained over a period of time that is sufficient to allow many cell divisions to occur. These processes are primarily used for developing both a fundamental understanding of selection in microbial systems and improved industrial strains. Classical chemostats are described, for example, in Herbert D., Elsworth R., Telling R.C. (1956) The continuous culture of bacteria; a theoretical and experimental study, J Gen Microbiol 14: 601 -622; Dykhuizen D.E., Hartl D.L. (1983) Selection in chemostats, Microbiol Rev 47:150-168; Zeng, A. P., Continuous culture, In Demain A.L., Davies J.E. (eds) (1999) Manual of industrial microbiology and biotechnology, ASM Press, Washington, DC. 151-178; and Gottschal J.C. (2000) Continuous culture Encycl Microbio\ 1 : 873-886. Other continuous culture configurations are described in Brown S.W., Oliver S.G. (1982) Isolation of ethanol-tolerant mutants of yeast by continuous selection. Eur J Appl Microbiol 16: 119-122; Fleming G., Dawson M.T., Patching J.W. (1988) The isolation of strains of Bacillus-subtilis showing improved plasmid stability characteristics by means of selective chemostat culture, J Gen Microbiol 134: 2095-2101 ; and Gostomski P., Muhlemann M., Lin Y.H., Mormino R., Bungay H. (1994) Auxostats for continuous-culture research, J Biotechnol 37: 167-177. Improved microbial growth of mutant strains has sometimes been observed on non-native substrates that support little or no growth in wild type strains. This improvement may be attributed to selection for increased production of particular intracellular enzymes in continuous culture. For example, Horiuchi T., Tomizawa J.I., Novick A. (1962) Isolation and properties of bacteria capable of high β-galactosidase synthesis, Biochem Biophys Acta 55: 152-7 reports that after growing E. coli for 80 generations in a lactose- limited chemostat, Jhe culture grew much faster on lactose, and that β- galactosidase expression represented more than 20% of cellular protein. As a result of the selection and the laboratory techniques employed, expression became constitutive rather than inducible. Similarly, Rigby P.W.J., Burleigh B.D., Hartley B.S. (1974) Gene duplication in experimental enzyme evolution, Nature 251 : 200-204, reports the use of chemostat selection to obtain a strain of Klebsiella aerogenes that was capable of growing on xylitol from a wild type which was not able to grow on this substrate. The ability of the selected strain to utilize xylitol was attributed to a roughly 20-fold increase in ribotol dehydrogenase. Schneider, K.H., Jakel G., Hoffmann R., Giffhorn F. (1995) Enzyme evolution in Rhodobacter spharoide-selection of a mutant expressing a new galactitol dehydrogenase and biochemical characterization of the enzyme, Microbiology 141 : 1865-1873 reports selecting a mutant strain of Phodobacter spharoides in a chemostat with the ability to grow on galacitol from a parent strain not able to grow on this substrate. Growth was attributed to a galacitol dehydrogenase that was not present in the wild type strain. In the above examples, it is apparent that mutant or adapted organisms with altered expression of catabolic enzymes may realize a selective advantage, as compared to a wild type population which does not demonstrate such altered expression. The mutant or adapted organisms thrive because the benefit(s) of altered production are available to the mutant to a greater extent than they are available to the wild type. This requirement is automatically met in the case of catabolic enzymes that are produced intracellularly, but is not necessarily fulfilled in the case of extracytoplasmic enzymes that are expressed either on the cell surface or secreted into the extracellular milieu.

This leaves open the question of whether selection will be effective for growth due to extracytoplasmic enzyme expression. In one example, Lelieveld H.L.M. (1982) The use of continuous cultures for selection and isolation of microorganisms producing extracellular enzymes adapted to extreme environments, Biotechnol Bioeng 24: 1419-1425 predicts that hyper- production of a particular secreted extracellular enzyme could be selected for in continuous culture by growing the organism on a target substrate requiring a particular enzyme. The article hypothesizes that the local concentration of extracellular enzyme would be greater in the immediate vicinity of the hyper- producing cell than in the vicinity of the wild type cells. This would result in higher concentration of substrate at the cell surface, and hence, higher growth rates. Despite this prediction by Lelieveld, the prior art does not report success in the use of continuous culture under the conditions proposed by Lelieveld to obtain strains with improved growth on soluble substrate by virtue of the change in enzymes that are expressed extracellularly and secreted to the medium.

Overall, limited successes have been obtained when selecting for improved expression or performance of extracytoplasmic enzymes. Francis J.C., Hansche P.E. (1971) Directed evolution of metabolic pathways in microbial populations. I. Modification of the acid phosphatase pH optimum in S. cerevisiae, Genetics 70: 59-73 reports a study involving the use of well- mixed cultures to obtain strains with improved growth on a soluble substrate by virtue of changes in extracytoplasmic enzyme expression. The enzyme in this study, acid phosphatase, was a cell wall-associated enzyme that metabolizes orthophosphate. A 1.7-fold increase in the acid phosphatase mutant was achieved after maintenance in chemostat with the enzyme substrate (orthophosphate) supplied. Naki D., Paech C, Ganshaw G.,

Schellenberger V. (1998) Selection of a subtilisin-hyperproducing Bacillus in a highly structured environment, Appl Microbiol Biotechnol 49: 290-294, reports selection-based improvement of secreted enzyme expression through a different means ~ using a hollow fiber apparatus to control the cell density at one clone per fiber and limit transport of enzymes and reaction products away from the cell.

SUMMARY

The instrumentalities reported herein advance the art by providing a method for selection of microorganisms where growth depends upon expression of a tethered extracytoplasmic enzyme. Selection results are particularly improved when an insoluble target substrate is provided for contact with microorganisms that potentially express the tethered extracytoplasmic enzyme.

In one aspect, a method according to these instrumentalities provides for selection of microorganisms where a relative growth enhancement depends upon expression of a tethered extracytoplasmic enzyme. The extracytoplasmic enzymes are generally tethered to a surface of a microbial cell, for example, by an "anchor protein" attached to the cell wall. The method includes growing a cell in a culture medium that contains a non-native target substrate. Mutants may be identified by a relative growth enhancement, apparent for example by increased representation of the mutant population over time.

As used herein, an "extracytoplasmic enzyme" is an enzyme that is bonded to or protrudes from the external cell membrane wall. This enzyme may be a heterologous enzyme in the sense of an enzyme that is expressed by the microorganism from heterologous or xenogenic DNA, such as an organism that has been transformed with recombinant DNA. Heterologous enzymes may be genetically engineered and expressed on the surface of a non-native microorganism utilizing techniques known in the art. Alternatively, this enzyme may occur naturally in the organism, but may be expressed at low levels. In this case for example, the selection process may be used to select organisms that have increased production of a gene of interest.

As used herein, the term "parent" strain refers to a microbial strain that is naturally occurring or has been genetically engineered to express a heterologous enzyme but has not yet been given the opportunity to evolve under conditions of stress that are amenable to selection processes. A

"mutant" as used herein is a phenotypically distinct strain that arises from the original parent.

The term "non-native target substrate" shall refer to a substrate that does not normally promote growth and/or reproduction of an organism. Generally, an organism produces enzymes that catalyze the conversion of a target substrate into metabolizable products, which are used by the organism for growth and/or reproduction. If an organism does not normally produce sufficient amounts of an enzyme to convert a target substrate into metabolizable products, the substrate is a non-native substrate. Target substrates useful in the practice of the disclosed methods include cellulose, hemicellulose, chitin, starch and protein.

It is generally considered that the wild type or parent strain has been shown not to express the tethered extracytoplasmic enzyme, or to express such enzyme at relatively low levels as compared to the strain(s) after selection. Thus, a stress condition may be created where growth of the microorganism is nutritionally limited by the non-native target substrate. Selectable microorganisms that thrive in this type of environment have adapted to express enzymes permitting the microorganisms to grow upon the non-native target substrate. Particularly preferred tethered extracytoplasmic enzymes contemplated for selection by the present method act on insoluble target substrates and include cellulases, xylanases, hemicellulases, chitinases, amylases and proteinases. There exists broad applicability of the method, such that the method is not limited to any particular type of microorganism or enzyme. In a narrower and more preferred sense, specific candidate organisms include Escheria coli, Klebsiella oxytoca, Bacillus subtilis, Thermanaerobacter thermosaccharolyticum, Thermoanaerobacterium saccharolyticum,

Zymomonas mobilis, Clostridium thermocellum, Clostridium cellulolyticum, Ruminococcus albus, Ruminococcus flavefaciens, Fibrobacter succinogenes, Trichoderma reesei, Thermobifida fusca, Cellulomonas fimi, Candida glycerinogenes, Debaryomyces hansenii, Candida tropicalis, Schizosaccharomyces pombe, Candida albicans, Aspergillus fumigatus, Aspergillus nidulans, Cryptococcus neoformans, Magnaporthe grisea, Neurospora crassa, Pneumocystis carinii, Pichia stipitis, Pichia pastoris, Candida shehatae and Pachysolen tannophilus.

The preferred methodology may be used, for example, to select for tethered extracytoplasmic enzymes in the categories of cellulases, xylanases, hemicellulases, chitinases, amylases and proteinases. This is done by placing a target substrate in the growth medium and selecting for enhanced growth where the normal candidate microorganism in unmutated form either does not digest the target substrate or inefficiently processes the target substrate. Enhanced growth is an indicator of enhanced expression of an enzyme that is complementary to the target substrate.

The effectiveness of this selection technique may be confirmed by comparative mathematical modeling. Selection results are more certain, the magnitude of growth improvement is relatively more enhanced, and the duration of the selection process is diminished, for example when using chemostats to select for microorganisms containing tethered enzyme targeting an insoluble target substrate.

The examples below show four comparative embodiments that have been evaluated involving microorganisms growing on either soluble or insoluble target substrates. The examples evaluate the effectiveness of selection for organisms expressing enhanced amounts of extracytoplasmic enzymes, either with or without tethering to the cell surface. In a first embodiment, enzyme was released to the culture medium without significant binding to the cell surface ("free enzyme") where the target substrate was substantially soluble. A second embodiment examined free enzyme with a substantially insoluble target substrate, while third and fourth embodiments respectively featured enzyme that was bound to a cell surface, i.e., "tethered enzyme", with a substantially soluble target substrate, and tethered enzyme with a substantially insoluble target substrate. In the third and fourth embodiments, it was assumed that cells adhered to the substrate by virtue of enzymes on the cell surface that bind to the substrate in the course of carrying out their catalytic function. Results for the two embodiments involving tethered enzymes demonstrated a clear probability of selective advantage for mutants with enhanced enzyme production, with the best case for selection being that of using a tethered enzyme in combination with an insoluble target substrate. Analyses indicate that it is possible to select for faster growing cells in less than three months. In this regard, faster growing cells are hereby defined as faster growing mutant cells representing 1 % of the population.

BRIEF DESCRIPTION OF THE DRAWINGS

Systems and methods will be described in further detail with reference to the following detailed description and the accompanying drawings, in which:

FIG. 1 illustrates free enzyme with soluble and insoluble substrates and tethered enzyme with soluble and insoluble substrates;

FIG. 2 illustrates boundary conditions for an insoluble substrate/enzyme tethered case;

FIG. 3 graphically depicts sensitivity to ks in soluble and insoluble substrate cases;

FIG. 4 graphically illustrates selection time for a mutant at different Rs values; FIG. 5 graphically illustrates selection time for a mutant having 2 times greater sufficiency for different initial mutation frequencies; FIG. 6 graphically illustrates selection time for mutants with only increased base area compared to the parent strain, and for mutants with only increased enzyme production (3-fold) (i.e., mutant and parent have the same percentage of base area); FIG. 7 graphically illustrates selection time for a mutant having 2 times greater sufficiency for different gap distances;

FIG. 8 graphically illustrates selection time for a mutant with 2 times greater sufficiency for different boundary layer thicknesses; and

FIG. 9 graphically illustrates selection time for a mutant with 2 times greater sufficiency for different values of the diffusivity of glucose in water.

DETAILED DESCRIPTION

There will now be shown and described a comparative evaluation of four model systems that may be used to select for enhanced growth of microorganisms expressing extracytoplasmic enzymes in continuous culture. The embodiments are based upon Saccharomyces cerevisiae, but may be used with any compatible microorganism and growth culture media. The selection process is confirmed by mathematical modeling to show that use of an insoluble substrate in conjunction with a tethered enzyme is associated with increased expression of extracytoplasmic enzyme. In a preferred sense, specific candidate organisms include Escheria coli, Klebsiella oxytoca, Bacillus subtilis, Thermanaerobacter thermosaccharolyticum, Thermoanaerobacterium saccharolyticum, Zymomonas mobilis, Clostridium thermocellum, Clostridium cellulolyticum, Ruminococcus albus, Ruminococcus flavefaciens, Fibrobacter succinogenes, Trichoderma reesei, Thermobifida fusca, Cellulomonas fimi, Candida glycerinogenes, Debaryomyces hansenii, Candida tropicalis, Schizosaccharomyces pombe, Candida albicans, Aspergillus fumigatus, Aspergillus nidulans, Cryptococcus neoformans, Magnaporthe grisea, Neurospora crassa, Pneumocystis carinii, Pichia stipitis, Pichia pastoris, Candida shehatae and Pachysolen tannophilus. The discussion below illustrates by way of example, and not by limitation. In one aspect, a mathematical model was created to evaluate a variety of comparative selection systems. Specifically, the comparative selection systems included: (1) the microorganism expressing secreted (non-tethered) enzyme in a growth medium that contains a soluble non-native target substrate; (2) the microorganism expressing secreted (non-tethered) enzyme in a growth medium that contains an insoluble non-native target substrate; (3) the microorganism expressing tethered enzyme in a growth medium that contains a soluble non-native target substrate; and (4) the microorganism expressing tethered enzyme in a growth medium that contains an insoluble non-native target substrate.

In one embodiment, substantially soluble substrate with free and tethered enzyme, as shown in FIG. 1 , was evaluated for β-glucosidase- dependent utilization of cellobiose. In other embodiments, substantially insoluble substrates with free and tethered enzymes were evaluated for cellulase-dependent utilization of cellulose. Specifically, this model system addressed Saccharomyces cerevisiae on a growth-limiting soluble substrate that was generated by the action of one or more saccharolytic enzymes acting on either a soluble substrate or an insoluble substrate.

Localization of expression affects the concentration of substrate in the local region of the cell, whether the localization is intra-cytoplasmic, extracytoplasmic and associated with the cell, or extra-cytoplasm ic but secreted into the medium. This occurs because the products of an intracellular reaction are contained within the cell; the products of a reaction occurring in the bulk fluid equilibrate throughout the bulk fluid; and the products of reactions occurring within the boundary layer of the cell diffuse into the bulk fluid, unless they are restricted in some way, for example by attachment to an insoluble substrate. The level of expression affects the concentration of metabolizable substrate and thus the growth rate of an individual; however, the effect relative to other organisms is dependent on a combination of factors.

The extent to which selection for improved growth on non-native substrates is achievable for microorganisms expressing extracytoplasmic enzymes depends on a variety of factors. A number of these factors have been identified and explored. They are categorized according to whether they impact production of metabolizable substrate or the transport/uptake of that substrate. Those factors affecting usable substrate conversion, e.g., soluble saccharides in the case of cellulose, include level of expression, specific activity and the rate of mutation relevant to the phenotype of interest. Factors affecting substrate transport/uptake include: the value of the saturation constant, ks, cell-substrate gap width, location of expression, cell shape and physical properties in the gap. Cell shape determines whether the cell-substrate distance at the location of adherence is realized only at a point - in one extreme - or over a broad surface. Physical properties in the gap, such as the diffusivity of the substrate in the fluid, may be impacted by extracellular polysaccharide, chemical features of the cell wall surface and protuberances from cells that direct usable substrates to the cells, for example, as reported by Shoham Y., Lamed R., Bayer E.A. (1999) The cellulosome concept as an efficient microbial strategy for degradation of insoluble polysaccharides. Trends Microbiol 7: 275-281.

Binding of the organism to the substrate gives rise to a gap between the substrate surface and the cell which can act to limit the diffusion of hydrolysis products away from an organism. The shape of the organism may influence the cell's ability to generate and capture the products of substrate degradation. By way of example, flatter cells have more of their attached enzyme exposed to the substrate and the proximity of the cell and substrate are such that products of substrate degradation must diffuse a longer distance to escape to the bulk solution.

The geometric shape of the cell-substrate gap may affect the diffusion of reaction products and thus the ability of a cell to retain those products. Both the geometry of the gap between the organism and the substrate and the physical properties affect the ability of the organism to retain the product(s) of substrate hydrolysis. The distance of this gap determines the extent to which the products of hydrolysis escape to the bulk solution. The closer the organism is to the substrate surface, the less likely it is that diffusion will carry the products of hydrolysis laterally away from the cell. The physical properties of the area between the cell and the substrate surface also affect the diffusion of hydrolysis products. Thus, in the course of selection on the basis of comparatively enhanced growth, changes in shape and perhaps other variables may accompany increased expression levels and/or activity of tethered enzymes.

It is possible for selection to occur in a comparatively rapid time span under conditions where the enzyme is tethered to the cell surface, the substrate is insoluble and the cell sits flatly upon and very close to the insoluble substrate surface. The maximum difference in growth rates between a mutant and a parent strain occurs when cross-feeding between the two is totally eliminated. Thus, an improvement in the ability of an individual cell to retain the products of hydrolysis increases the effectiveness of the selection scheme.

A recombinant yeast expressing a tethered cellulase is examined herein. The yeast can be engineered with a short linker region between the recombinant cellulase and the cell, for example, as reported by Murai T., Ueda M., Yamamura M., Atomi H., Shibasaki Y., Kamasawa N., Osumi M., Amachi T., Tanaka A. (1997) Construction of a starch-utilizing yeast by cell surface engineering, Appl Env Microbiol 63: 1362-1366, which would keep the cell in close proximity to the insoluble substrate surface.

The present method may be used by those pursuing the development of organisms via selection in continuous culture. This method of selection may be used, for example, to develop organisms for dairy and alcohol fermentation processes, for developing new systems to probe physiological pathways, for genetically modifying crops and for fuel production via bioconversion.

EXAMPLE 1 MATHEMATICAL MODELING AND QUANTIFICATION OF SELECTION

PERFORMANCE STANDARDS Mathematical modeling according to one embodiment may be performed on a personal computer utilizing Matlab PDE toolbox (Mathworks, Natrick, MA) software.

The Monod equation may be used to describe the dependence of the cell growth rate, μ, upon the concentration of soluble substrate (Monod J. 1942. Recherches sur la crossance des cultures bacteriennes. Paris: Hermann.):

where, G_c = rate limiting substrate concentration at the cell surface (g/L), ks = saturation constant (g/L) and

maximal growth rate (hr ^~1).

For situations in which the substrate concentration at the surface of the cell is not uniform, the growth rate μ may be calculated by considering the effect of local substrate concentration over the entire surface of the cell, using the following equation:

where A is the total surface area of the cell, and G_c,iocai is the local substrate concentration at the cell surface.

Selection due to differences in substrate availability arising from different activity and or location of saccharolytic enzymes was examined. Thus, the same values of μ,^ and k were used for the mutant and parent strains. Enzyme Expression for the Parent and Evolved Strains

The extent of enzyme expression may be defined in terms of a dimensionless sufficiency parameter, S. As defined by McBride et al. (McBride J., Zeitsman J., Van Zyl W., Lynd L. 2005. Utilization of cellobiose by recombinant β-glucosidase-expressing strains of Saccharomyces cerevisiae: characterization and evaluation of the sufficiency of expression. Enzyme and Microb Tech. In Press.), this parameter is equal to the ratio of the cell-specific enzyme activity (a = grams usable substrate produced/ (g cells ^* hr)) and the maximum cell-specific substrate consumption rate (q_max = M ax/Yx/s, where Yχ_/S is the cell yield):

S = -2- (3)

"max The sufficiency ratio, Rs, is defined as:

Rs ⁼ Smutan Spareπt (4)

The value of S for the parent strain was assumed to be 0.05. This value corresponded to growth being strongly limited by extracytoplasmic enzyme expression, since the rate of usable substrate supply was assumed to only be about 5% of the rate of substrate consumption for a culture growing at μ_maχ- The value of Rs is, therefore, representative of the relative expression of the enzyme being expressed. A range of Rs values from 1.1 to 6, consistent with empirically observed values for increased enzyme expression due to selection, was considered. Empirical values include a 4-fold, 5-fold and 2.8- fold increase in enzyme expression in the studies of Francis and Hansche (1971), Naki et al. (1998) and Brown et al. (Brown S.W., Oliver S.G., 1982. Isolation of ethanol-tolerant mutants of yeast by continuous selection. Eur J Appl Microbiol 16: 119-122.), respectively.

Mutation Frequency, f

The mutation frequency, /, is defined as the number of mutants per cell per division. Data from studies of Saccharomyces cerevisiae focusing on "gain of function" mutations, including auxotrophic reversions and antibiotic resistance, are presented in Table 1. Similar mutation frequencies are observed for other Saccharomyces species and other genera. For phenotypes that require change of one specific amino acid, the average frequency is about 1.07 ^* 10^"7 - 2.8 ^* 10^"7. For the case where a number of mutations could have an effect on phenotype, the average frequency is about 1.3 ^* 10^"3 - 2.3 ^* 10^"3. A mutation frequency of 5 ^* 10^~7 was used in the present calculations, with sensitivity of results evaluated based on a range of frequency values from l ^* 10^"4 to 1 ^* 10^"9. Table 1. Phenotypic mutation frequencies for Saccharomyces cerevisiae, reported as mutants/cell/division.

BP changes for

Selection For Frequency effect Source

Cycloheaarride resistance 0-1.5E-6 average -0.9E-6 1 Adams 1985

Met+ fromMet- 1.60BO8 1 agni 1962

Met+ fromMet- 4.10BO8 1 Magni 1962

His+ fromHis- 1.00BO9 1 Magni 1962

His+ fromHis- 2.00EO9 1 Magni 1962

Flucanozole resistance 4.70EO3 >1 Anderson 2003

Flucanozole resistance 3.43&04 >1 Anderson 2003

His+ froml-Ds- (not selection induced- single base pair nutation) 9.00E-11 1 Hall 1992

Lys+ftomLys- 3.00EO8 1 Steele 1992

Lys+fiomLys- 3.00EO8 1 Storchova 1999

His +fϊomHis - 5.00&08 1 Storchova 1999

CAN resistance 3.20B07 >1 Zeyl 2000

Cyclohexairide resistance 8.00E-10 1 Zeyl2000

5-flouroorotic acid resistance 3.30E08 >1 Zeyl2000

Adams J., Paquin C, Oeller P.W., Lee L.W. 1985. Physiological characterization of adaptive clones in evolving populations of the yeast Saccharomyces cerevisiae. Genetics 110: .173-185.

Magni G.E., Von Borstel R.C. 1962. Different rates of spontaneous mutation during mitosis and meiosis in yeast. Genetics: 1097-1108.

Anderson J.B., Sirjusingh C, Parsons A.B., Boone C, Wickens C, Cowen L.E., Kohn L.M. 2003. Mode of selection and experimental evolution of antifungal drug resistance in Saccharomyces cerevisiae. Genetics 163: 1287-1298.

Hall B.G. 1992. Selection-induced mutations occur in yeast. P Natl Acad Sci USA 89: 4300-4304.

Steele D.F., Jinks-Robertson S. 1992. An examination of adaptive reversion in Saccharomyces cerevisiae. Genetics 132: 9-21.

Storchova Z., Vondrejs V. 1999. Starvation-associated mutagenesis in yeast Saccharomyces cerevisiae is affected by Ras2/cAMP signaling pathway. Mutat Res- Fund Mol M 431 : 59-67.

Zeyl C, DeVisser J.A.G.M. 2001. Estimates of the rate and distribution of fitness effects of spontaneous mutation in Saccharomyces cerevisiae. Genetics 157: 53-61.

Population Dynamics

The difference between populations of cells is characterized by their relative enzyme expression as represented by the parameter, Rs. Mutations that give a particular value of Rs are predicted to occur with a constant frequency at every cell division. To simplify the model, the case where successive mutations occur is not considered. A material balance on a constant volume, well-mixed continuous culture containing parent and mutant cell populations, assuming that mutations occur at a constant frequency, /, and that there are no backward mutations yields: parent i

dX mmuuttaanntt _ .... γ y ,, d j ₇t,, " mmuuttaanntt mmuuttaanntt

where D = dilution rate (hr ¹), X_parent = parent cell concentration (mg/L), X_mutant = mutant cell concentration (mg/L), μparent = growth rate of the parent and μmutant = growth rate of the mutant.

Because the changes from parent strain to mutant strain occur over long time periods during which almost all of the cell population is comprised of the parent strain X_parent » Xmutant- It is therefore reasonable to apply the parent pseudo steady state assumption to Xparent, in which case D = μpare - ^• -* Substituting this result into equation (6) yields:

" rωtant _ ■ ■ y _ „ y J -parenty J -parentγ = (A/ΔX Jr ∞rmty i, r^mut&nt mutant r^pαrenV' mutant , >y. parent , vy_* mutant -T^J '^mutant , _Λ\ pαren,

fμ where, Δμ = (μ_mutant ~ M_Parent+ _t ^p,^αZ )^■ ln(2)

The concentration of a mutant population at time t is obtained by integrating equation (7 ): y (t - ( Y _I J ■"" ppaarreenntt y _y , , ( ⁽AAμμxx((.tt--tt₀)))) J J r r~¹ ppaarreenntt _γ ,„. Λ mu tan t ^¹ ) ~ ^Λ mu tan tO ^ l ,n( ,2~)Δ . ./. ^{Λ p}p^aa^rr^eeⁿn^,t ⁾ )^{e ~} \ ,n( ,2„),.A . μ . ^{" Λ p}p^aa^rr^eeⁿn't °)

Assuming X_paren_t is constant, the time required for the mutant to make up 1% of the total cell population can be calculated. A final percentage of 1% was chosen because while mutants could easily be screened for at that percentage, the effect of the mutant population on the parent population and the bulk substrate concentration was negligibly small. This gives:

where F₀ is the initial fraction of mutants. F₀ was chosen to be /, and the ratio of X_mu_faπ_f S Xparen_t at time to is also /, which gives the minimum fraction of mutant cells present initially. To prevent the situation where fractions of cells were included in the calculation, it was assumed that 1// was significantly less than the total number of cells in the reactor. This condition can be easily met in practice by choosing an appropriate cell concentration and reactor volume.

The value of Δμ depends on the local substrate concentration at the surface of the cells for the mutant and the parent strain. The diffusion boundary layer, cell geometry and mass transfer are important for calculating these concentrations.

Calculation of Substrate Concentrations

When the dilution rate for a pseudo steady state continuous culture is chosen, the growth rate for the parent strain is also determined and the implied surface substrate concentration around the parent cell is calculated from (1). Values for the corresponding bulk substrate concentration were calculated by numerically solving a differential material balance for diffusion of substrate near the parent cell. This bulk substrate concentration was then used to calculate the cell surface concentration of a mutant cell also by a material balance. Details of the material balance solutions depend on the scenario considered and are examined below. With the surface substrate concentration of both the parent and mutant cells known, the growth rate of parent cells and mutant cells was determined using equation (2). The selection time was calculated using equation (9). Boundary Layer Thickness

The boundary layer is an area of fluid surrounding a particle. The boundary layer is a thin layer of fluid near the surface of a particle where transport due to diffusion in the absence of convection is assumed to take place.

For the soluble substrate case, the thickness of the boundary layer around the cell may be calculated assuming that cells are spherical according to the Frossling correlation:

Sh =

where k_\ = mass transfer coefficient (m/s), d_p = the diameter of the particle (m), Ds = the diffusivity of the substrate of interest in water (m²/S), Sh = Sherwood number, Re = Reynolds number and Sc = Schmidt number.

Since the mass transfer coefficient is defined as:

k, = ^- (1 1 ) o

where δ is the boundary layer thickness. Equations 10 and 1 1 can be combined to give:

5= 4

(2 + 0.6Re^1/2 Sc^{, 3}) (12) Since the relative velocity of cells and the surrounding fluid is very small, the Reynolds number approaches zero and δ = dp/2 approaches the cell radius. This correlation also gives a minimum value for the Sherwood number of 2 (Kirwan. 1987. "Mass Transfer Principles" in Rousseau, Handbook of Separation Process Technology, John Wiley and Sons.). The minimum value for the Sherwood number gives the largest value for the boundary layer and subsequently the slowest mass transfer. If selection is not efficient in this case, then it will also not be efficient for thinner boundary layers.

In the insoluble substrate case with tethered enzyme expression, the cell is attached to the surface of an insoluble substrate present within a boundary layer whose thickness may be determined by the size of the cellulose particle. Cellulose is used as a model insoluble substrate, assuming maximum particle dimensions between 50 and 200 μm and noting that particles in this range can be obtained from commercial suppliers. Idealizing cellulose particles as spheres, the estimated boundary layer thickness is in the range of about 10-20 μm for a 100 μm diameter particle, depending on the relative velocity of the particle to the fluid; the higher value corresponds to the case where the terminal settling velocity of the particle is used, and the smaller value corresponds to the case where the centripetal acceleration due to stirring is used. In the simulations presented herein, a conservative estimate of 10 μm for the diffusion boundary layer thickness was used for the insoluble substrate case with tethered enzyme, with sensitivity of results evaluated based on a range of frequency values from 10 to 40 μm. Soluble Substrate Case When the substrate is substantially soluble, the enzymatic hydrolysis of cellobiose to glucose is used as a model system. In the case considered, cellobiose was supplied to the system with a concentration that was far greater than the K_m, and hence β-glucosidase production was operating at V_max and glucose production was substantially the function of the volumetric enzyme concentration, β-glucosidase production was assumed to be proportional to cell production.

Mass transfer in the boundary layer around a spherical cell can be described by material balances including diffusion and reaction terms,

_^ 1 d , ₂ dY, „ _Λ D_γ — — (r² — ) + R_Y = 0 (13) r dr dr where Y may be either the substrate concentration or enzyme concentration. Boundary conditions are divided into three types: concentration is equal to bulk concentration at the outer surface of the boundary layer; rate of transport is equal to the rate of consumption and/or generation at the cell surface; glucose concentration at the cell surface has to be sufficient to support a growth rate equal to the dilution rate plus the maintenance rate. The boundary conditions for equation (13) in enzyme non-tethered and tethered cases are presented in Table 2.

Table 2. Summary of boundary conditions for soluble substrate case used in

α : dry mass per cell having a radius of R (g) γ: substrate generation rate by the enzyme around the cell surface(g/hr)

Enzyme Not-Tethered Case

The change of glucose and enzyme concentrations for a quasi steady state continuous culture of parent cells can be described by the following equations: d^EB _ = ά _Λj( ^{m c} + m)X parent - D. E_B = 0 ⁴) dt + G, ^T = -^-( ^ + ^m^_Paren_t - D * G_B + k . E_B = 0 _{( 1 5}) d^{t Y}X /S ^Ks ⁺ c where G_c = glucose concentration on cell surface (mg/L), m = maintenance glucose consumption (hr ¹), EB = bulk enzyme concentration (mg/L), GB= bulk glucose concentration (mg/L), k = enzyme turnover number (mg glucose/hr/mg enzyme) and ω = coupling factor of enzyme production and growth rate. Solving equations (14) and (15) at steady state, leads to:

_E = 2 _G 6⁾

^B (k - D/(ω* Y_{x /s})) ^B

GB and E_B were solved by the following iterative procedure. By substituting a guessed value of GB in equation (6), a value for E_B was obtained. With E_B known, the two differential equations for glucose and enzyme presented in Table 3 were solvable. Hence it was possible to find G_B from the concentration profile. GB was then used to calculate E_B using equation (6). Iteration was performed until the values for G_B converged. With the bulk GB and EB known, Gc .on the surface of the mutant cells could be calculated by solving the reaction-diffusion equation (16).

Enzyme Tethered Case

GB was directly solvable from equation (13) using the boundary conditions provided for the parent cells in Table 2. With GB known, G_c on the mutant cells was obtained by solving for the glucose profile in the boundary layer of the mutant cell.

Insoluble Substrate

Enzyme Tethered Case

For the insoluble substrate case, it was assumed that, because the cell is relatively small with respect to the particle and because there is likely to be more than one cell attached to a substrate particle, it is possible to approximate the geometry of the system by having a cell with the shape shown in FIG. 2 adhered to a substrate surface which is a flat plate with a diameter of 40 μm. Sensitivity analysis on the size of the flat plate was conducted to determine the effect of this variable.

Cell Shape and Gap Distance

Binding of cell-attached enzymes to insoluble substrates likely leads to changes in cell shape, which are manifested either instantaneously and/or over time as a result of selection. A radially-symmetric cell shape was assumed with overall volume of 65μm³ corresponding to that of a 5 μm diameter sphere, typical of a yeast cell (Johnston G.C., Ehrhardt C.W., Lorincz A., Carter B.L.A. 1979. Regulation of cell size in the yeast Saccharomyces cerevisiae. J Bacteriol 137: 1-5.). A radius of curvature (R) of 1.47 μm and a base diameter/length (L) of 2.94 μm (IJR=2, percentage contact area 8.2%) were assumed for the base case (see FIG. 2), which was intended to reflect a moderate degree of deformation due to interactions between cell-bound enzymes and the substrate. The cell shape was also allowed to vary parametrically in sensitivity studies. Results were evaluated with respect to a range in the R/L ratio from 0.2 to 4 (percentage contact area: 37.5% to 3.3%), while keeping the overall volume constant at 65 μm³.

The gap between the cell base and the insoluble substrate surface (H) is potentially amenable to manipulation, e.g., by altering the length of the enzyme "tether" (Murai et al., 1997), and may be varied parametrically from 10 to 100 nm. A cell-substrate gap of 10 nm has been observed for cellulose- adherence in naturally-occurring cellulolytic microorganisms (Kudo H., Cheng K.J., Costerton J.W. 1987. Electron microscopic study of the methylcellulose- mediated detachment of cellulolytic rumen bacteria from cellulose fibers. Can J Microbiol 33: 267-271.), and was used for the present evaluation.

Calculation

Substrate concentration was omitted from the denominator in equation (1) because the surface substrate concentration was less than ks by at least 10-fold for all cases examined herein and this simplification greatly decreased the complexity of the calculation. The Matlab PDE toolbox (Mathworks, Natrick, MA) was used to solve a two-dimensional diffusion equation for a parent cell. The 2-D treatment was sufficient because of the radial symmetry of the cell. Referring to the boundaries shown in FIG. 2, the conditions at those boundaries were as follows: a-b: G = G_B; b-c, c-d, e-f , and f-a: VG = 0; Cell surface: VG = (α D_g / A) • [(μ_maχ)(G)(1/ks)+m]; d-e: VG = R_s»γ/ D_g . (Ab: base contact area of the cell, m²). To solve, a value of GB was guessed with iteration until the condition μp_ar_en_t= D was met. With GB known, the two dimensional PDE equation was solved again for the mutant (where dG/dx at the substrate surface was Rs times higher), resulting in the glucose concentration profile for the mutant. Soluble Substrate

Enzyme not Tethered

The calculated glucose concentration at the surface of the mutant cell was essentially the same as the concentration at the surface of the parent cell at different dilution rates (data not shown). Hence, it is practically impossible to select the mutant from the parent strain in continuous culture with a soluble substrate and non-tethered enzyme. Enzyme Tethered

In the soluble substrate and enzyme tethered case, there is a higher glucose concentration at the cell surface of mutants that produces more tethered enzyme. The selection time is highly dependent on the value of ks (see FIG. 3), however, cells with a smaller ks are able to derive a larger benefit from small differences in the concentration of glucose as compared to cells with a larger ks.

Insoluble Substrate Enzyme not Tethered

If the enzymes are not tethered to the cell surface, they may bind to the insoluble substrate surface due to their substrate binding ability, and release the hydrolyzed sugar to the bulk solution. The mutant that produces more or better enzyme will not receive any benefit because glucose reaches them by diffusion from the bulk solution, just as for the parent. Thus selection of strains with improved enzyme expression is not expected to be successful. Enzyme Tethered

Selection times of less than 2 months may be possible for selection carried out on solid substrates with tethered enzymes, due to the fact that mutants with more tethered enzyme have a substantially higher substrate concentration at the cell surface. Also, the selection time decreases with decreasing ks much as for the soluble substrate/tethered enzyme case.

Sensitivity Analysis for the Insoluble Substrate/Enzyme Tethered Case

To analyze the sensitivity of the simulation parameters each variable was independently modified (Table 3) with all other parameters held constant (see Table 4). It was determined that some simulation parameters are more important than others in determining the selection time for the insoluble substrate/enzyme tethered cases. Of the parameters affecting usable substrate production, R_s has the largest effect on selection time, resulting in an almost 30-fold decrease in selection time for a 6-fold increase in the parameter (Table 3). On the other hand, selection time is not very sensitive to the initial mutation frequency, which results because selection time is a function of the natural logarithm of the initial frequency in equation (9). Among the substrate transport/uptake parameters, the model is most sensitive to the percent of the total cell surface that is in close proximity to the substrate surface. Selection time decreases rapidly for mutants with flatter shapes (Table 4). ks is the next most important variable in the simulation, the value of which reflects the cells ability to use small concentrations of substrate rapidly. The potential to select for lowered values of ks for a particular substrate has been demonstrated in a number of studies (Dykhuizen, 1983). Since the diffusivity of the fluid in the gap between the cell and the insoluble substrate surface affects the speed of diffusion of hydrolysis products, decreasing this diffusivity decreases selection time. Similarly, selection time decreases as the gap distance decreases because the cells are able to retain a larger fraction of the newly created metabolizable substrate. Because selection is more effective as the boundary layer thickness increases, factors such as decreasing the stirring speed of the chemostat and increasing the viscosity of the fluid help to shorten the selection time. The diffusivity of glucose was included in the sensitivity analysis because the physical properties of water near the cell surface may not be the same as those in the bulk solution.

*(tl%, max/ t-|%, min)/ (Vhigh/ V|_ow)

Table 4. Base case parameters used in simulations.

Perry R.H., Green D.W. 1997. Perry's Chemical Engineers' Handbook. New York : McGraw- Hill 2581 p.

Rogers P.J., Stewart P.R. 1974. Energetic efficiency and maintenance energy characteristics of saccharomyces cerevisiae (wild type and petite) and Candiada parapsilosis grown aerobically and micro-aerobially in continuous culture. Arch Microbiol 99: 25-46.

Decker C.H., Visser J., Schreier P. 2001. β-glucosidase from Aspergillus tubingensis CBS 643.92: purification and characterization of four β-glucosidases and their differentiation with respect to substrate specificity, glucose inhibition and acid tolerance. Appl Microbial Biotechnol 55: 157-63.

Ghose T.K., Tyagi R.D. 1979. Rapid ethanol fermentation of cellulose hydrolysate 2. Product and substrate inhibition and optimization of fermentor design. Biotech Bioeng 21 : 1387-1400.

Calculation Results

Sensitivity to R_≥

FIG. 4 shows the effect of decreasing R_s. If R_s varies from 6 to 1.1 , the corresponding time needed to select a mutant from the parent strain varies from about 1.1 months to 34 months. Sensitivity to Initial Mutation Rate

Selection time is not very sensitive to the initial mutation rate (see FIG. 5), because selection time is a function of the natural logarithm of the initial frequency in equation (9). Sensitivity to Cell Shape

FIG. 6 presents results of a comparison of the effect of (a) mutations that occur only with regard to shape and (b) mutations that increase enzyme production for parents and mutants sharing a particular shape. Selection time decreases rapidly for mutants with flatter shapes. For parent/mutant pairs with substantially the same shape, an approximately 2-fold increase in enzyme expression is more quickly selected for in pairs with a flatter shape. Sensitivity to Gap Distance

FIG. 7 shows the effect of increasing gap distance on selection time. If the distance of the cell surface to the insoluble substrate surface varies from 5 nm to 100 nm, the corresponding time needed to select the mutant from the parent varies from about 1.9 months to 6 months. Thus, selection is expected to be more effective as the gap decreases. Sensitivity to Boundary Layer Thickness

As shown in FIG. 8, if the thickness of the boundary layer increases from 5 μm to 40 μm, the selection time decreases from 2.7 months to 2.2 months. Since selection is expected to be more effective as the boundary layer thickness increases, factors such as decreasing the stirring speed of the chemostat and increasing the viscosity of the fluid may help to shorten selection time. Sensitivity to Diffusivity

As demonstrated in FIG. 9, if the diffusivity of glucose in the gap between the cell and the substrate surface varies from 100% to 20% of the diffusivity of glucose in water, the corresponding time needed to select a mutant from the parent strain decreases from about 2.6 months to 0.8 months.

The sensitivity analyses reveal a number of interesting points. In the case where an organism is growing on a soluble substrate by virtue of a tethered enzyme, the value of ks is the determining factor in the effectiveness of selection (see FIG. 3). The parameter with the greatest impact for the enzyme tethered, insoluble substrate case appears to be R_s (an almost 10- fold decrease in selection time relating to a 2-fold increase in the parameter — see FIG. 4). Another interesting result is that the mutation frequency does not have a large effect on selection time. This is somewhat dependent on the assumption that there can be enough cells in the beginning of the experiment that at least one mutant cell will be present. However, this requirement is not difficult to meet, even for frequencies as low as 10^'11 for yeast and 10^"7 for bacteria.

EXAMPLE 2 MUTANT SELECTION FROM CONTINUOUS CULTURE A mutant strain with enhanced cellulase activity may be selected starting with a strain that expresses cellulase enzymes at low levels. The parent cells are grown in batch culture at 37°C in anaerobic seram vials

(Bellco, Vineland, New Jersey), in suitable growth medium (e.g., Delft medium or YPD). Phosphoric acid swollen cellulose (PASC) is provided as the carbon source (1%). Continuous cultures are grown in a BIOFLO 3000 fermentor (New Brunswick Scientific, Edison, NJ) in a working volume of 1.5 liter at 37°C. Agitation is kept constant at 100 rpm. The establishment of steady- state conditions is assumed when the culture has been grown with constant feeding for a period of at least 3 generations in which the cell density monitored by measuring the total protein concentration of samples, the dry weight, and/or the rate of base addition remains unchanged for at least 1 generation.

Upon reaching steady state conditions, selection is performed by allowing the culture to grow over a period of time. During this time, mutant cells with enhanced enzyme production arise spontaneously and detection of such mutants is undertaken by screening colonies obtained from the continuous culture for enhanced growth on the target substrate. For example, enhanced enzyme production may be screened for by growing samples of the culture on solid media culture plates and determining which colonies are growing fastest by examining the size of colonies after a given period of time. Larger colonies are identified as those containing cells of a desirable phenotype with enhanced enzyme expression. Further characterization might include measuring enzyme activity in a cellulase assay. Changes may be made in the above methods and systems without departing from the scope hereof. It should thus be noted that the matter contained in the above description or shown in the accompanying drawings should be interpreted as illustrative and not in a limiting sense. The following claims are intended to cover all generic and specific features described herein, as well as all statements of the scope of the present method, which, as a matter of language, might be said to fall there between.

All references and publications cited are expressly incorporated by reference herein.

Claims

CLAIMS We claim:

1. A method of selecting a microorganism with growth dependent upon one or more tethered extracytoplasmic enzymes expressed on a surface of a microbial cell, comprising: growing the cell on a non-native substrate bindable by the one or more tethered extracytoplasmic enzymes expressed on the surface of the cell.

2. The method of claim 1 , wherein the microbial cell is selected from the group consisting of Saccharomyces cerevisiae, Escheria coli,

Klebsiella oxytoca, Bacillus subtilis, Thermanaerobacter thermosaccharolyticum, Thermoanaerobacterium saccharolyticum, Zymomonas mobilis, Clostridium thermocellum, Clostridium cellulolyticum, Ruminococcus albus, Ruminococcus flavefaciens, Fibrobacter succinogenes, Trichoderma reesei, Thermobifida fusca, Cellulomonas fimi, Candida glycerinogenes, Debaryomyces hansenii, Candida tropicalis, Schizosaccharomyces pombe, Candida albicans, Aspergillus fumigatus, Aspergillus nidulans, Cryptococcus neoformans, Magnaporthe grisea, Neurospora crassa, Pneumocystis carinii, Pichia stipitis, Pichia pastoris, Candida shehatae and Pachysolen tannophilus.

3. The method of claim 1 , wherein the tethered extracytoplasmic enzyme is selected from the group consisting of cellulases, xylases, hemicellulases, chitinases, amylases and proteinases.

4. The method of claim 1 , wherein the non-native substrate is soluble in a growth medium used for continuous culture.

5. The method of claim 4, wherein the soluble substrate is selected from the group consisting of cellobiose and protein.

6. The method of claim 1 , wherein the non-native substrate is substantially insoluble in a growth medium used for continuous culture.

7. The method of claim 6, wherein the substantially insoluble substrate is selected from the group consisting of cellulose, hemicellulose, chitin, starch and protein.

8. A mutant selected by the method of claim 1.

9. A method of selecting a mutant strain of microorganisms with growth that depends upon one or more tethered, extracytoplasmic enzymes expressed on a surface of a microbial cell, comprising: growing the cell in continuous culture on a non-native substrate bindable by the one or more tethered, extracytoplasmic enzymes expressed on the surface of the cell; removing the one or more mutant cells from the continuous culture; and identifying one or more mutant cells with an increased growth rate.

10. A mutant selected by the method of claim 9.

11. The method of claim 9, wherein the microbial cell is selected from the group consisting of Saccharomyces cerevisiae, Escheria coli, Klebsiella oxytoca, Bacillus subtilis, Thermanaerobacter thermosaccharolyticum, Thermoanaerobacterium saccharolyticum,

12. The method of claim 9, wherein the tethered, extracytoplasmic enzyme is selected from the group consisting of cellulases, xylases, hemicellulases, chitinases, amylases and proteinases.

13. The method of claim 9, wherein the non-native substrate is soluble in a growth medium used for continuous culture.

14. The method of claim 13, wherein the soluble substrate is selected from the group consisting of cellobiose and protein.

15. The method of claim 9, wherein the non-native substrate is substantially insoluble in a growth medium used for continuous culture.

16. The method of claim 15, wherein the substantially insoluble substrate is selected from the group consisting of cellulose, hemicellulose, chitin, starch and protein.