(12) INTERNATIONAL APPLICATION PUBLISHED UNDER THE PATENT COOPERATION TREATY (PCT)
`
`(19) World Intellectual Property Organization a
`International Bureau
`
`(43) International Publication Date
`2 March 2006 (02.03.2006)
`
` es) ATLAICO
`
`(10) International Publication Number
`WO 2006/023144 A2
`
`(51) International Patent Classification:
`C12P 21/06 (2006.01)
`CI2QN 15/74 (2006.01)
`
`(21) International Application Number:
`PCT/US2005/024140
`
`(22) International Filing Date:
`
`6 July 2005 (06.07.2005)
`
`(25) Filing Language:
`
`(26) Publication Language:
`
`(30) Priority Data:
`60/585,918
`
`English
`
`English
`
`6 July 2004 (06.07.2004)
`
`US
`
`(71) Applicant(for all designated States except US): BIOREN
`INC. [US/US], 100 Glenn Way, Suite #1, San Carlos, CA
`94070-6264 (US).
`
`(72) Inventor; and
`(75) Inventor/Applicant (for US only): CREA, Roberto
`[IT/US]; 700 Occidental Avenue, San Mateo, CA 94402
`(US).
`
`(74) Agents: REMILLARD,Jane, E.et al.; Lahive & Cock-
`field, LLP, 28 State Street, Boston, MA 02109 (US).
`
`(81) Designated States (unless otherwise indicated, for every
`kind of national protection available): AE, AG, AL, AM,
`AT, AU, AZ, BA, BB, BG, BR, BW, BY, BZ, CA, CH, CN,
`CO, CR, CU, CZ, DE, DK, DM, DZ, EC, EE, EG, ES, FI,
`GB, GD, GE, GH, GM, HR, HU, ID,IL, IN, IS, JP, KE,
`KG,KM,KP, KR, KZ, 1.C, IK, LR, LS, LT, LU, LV, MA,
`MD, MG, MK, MN, MW, MX, MZ, NA, NG,NI, NO, NZ,
`OM,PG, PH, PL, PT, RO, RU, SC, SD, SE, SG, SK, SL,
`SM,SY, TJ, TM, TN, TR, TT, TZ, UA, UG, US, UZ, VC,
`VN, YU, ZA, ZM, ZW.
`
`(84) Designated States (unless otherwise indicated, for every
`kind of regional protection available): ARIPO (BW, GH,
`GM, KE, LS, MW, MZ, NA, SD, SL, SZ, TZ, UG, ZM,
`ZW), Eurasian (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM),
`European (AT, BE, BG, CH, CY, CZ, DE, DK,EE, ES, FI,
`FR, GB, GR, HU,IE,IS, IT, LT, LU, LV, MC, NL, PL, PT,
`RO, SE, SI, SK, TR), OAPI (BF, BJ, CF, CG, CI, CM, GA,
`GN, GQ, GW, ML, MR, NE, SN, TD, TG).
`
`Published:
`
`without international search report and to be republished
`upon receipt of that report
`
`For two-letter codes and other abbreviations, refer to the "Guid-
`ance Notes on Codes and Abbreviations" appearing at the begin-
`ning ofeach regularissue of the PCT Gazette.
`
`(54) Title: LOOK-THROUGH MUTAGENESIS FOR DEVELOPING ALTERED POLYPEPTIDES WITH ENHANCED PROP-
`ERTIES
`
`
`
`
`
`23.144A2INDIVIDUNNARIOATATAOTA
`
`©} (57) Abstract: A method of mutagenesis by which a predetermined amino acid is introduced into each and every position of a
`selected set of positions in a preselected region (or several different regions) of a polypeptide to producea library of polypeptide
`&> analogsis disclosed. The method is based on the premise that certain amino acidsplay a crucial role in the structure and function
`©} ofproteins and thus is capable of identifying and distinguishing functional aminoacid residues("hot spots") from non-functional
`amino acids residues ("cold spots") within a polypeptide or portion thereof. Libraries can be generated which contain only desired
`polypeptide analogs and are of reasonable size for screening. The libraries can be used to study the role of specific amino acids
`in polypeptide structure and function and to develop new or improved polypeptides such as antibodies, antibody fragments,single
`
`2 chain antibodies, enzymes, and ligands.
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`LOOK-THROUGH MUTAGENESIS FOR DEVELOPING ALTERED
`POLYPEPTIDES WITH ENHANCED PROPERTIES
`
`Related Information
`The entire contents of all other patents, patent applications, and references cited
`throughout the following specification also are incorporated by referenceherein in their
`entireties.
`
`Backgroundof the Invention
`Mutagenesis is a powerful toolin the study ofprotein structure and function.
`Mutations can be madein the nucleotide sequence of a cloned gene encoding a protein
`of interest and the modified gene can be expressed to produce mutantsofthe protein.
`By comparing the properties of a wild-type protein and the mutants generated,it is often
`possible to identify individual aminoacids or domains of aminoacids that are essential
`for the structural integrity and/or biochemical function of the protein, such asits binding
`and/or catalytic activity. The number of mutants that can be generated from a single
`protein, however, rendersit difficult to select mutants that will be informative or have a
`desired property, evenif the selected mutants that encompass the mutationsare solely in
`putatively importantregionsof a protein (e.g., regions that make up an activesite of a
`protein). For example, the substitution, deletion, or insertion of a particular amino acid
`may havea local or global effect on the protein.
`Previous methods for mutagenizing polypeptides have beeneither too restrictive,
`too inclusive, or limited to knocking out protein function rather than to gaining or
`improving function. For example, a highly restrictive approachis selective or site-
`directed mutagenesis whichis used to identify the presenceof a particular functional site
`or understand the consequences of making a very specified alteration within the
`functional site. A commonapplication ofsite directed mutagenesisis in the study of
`phosphoproteins where an aminoacid residue, that would ordinarily be phosphorylated
`and allow the polypeptide to carry out its function, is altered to confirm the link between
`phosphorylation and functional activity. This approachis very specific for the
`polypeptide and residue being studied.
`Conversely, a highly inclusive approach is saturation or random mutagenesisthat
`is designed to produce a large number of mutations encompassingall possible alterations
`within a defined region of a gene or protein. This is based on the principle that, by
`generating essentially all possible variants of a relevant protein domain,the proper
`arrangementof amino acidsis likely to be produced asone of the randomly generated
`mutants. However, in practice, the vast number of random combinations of mutations
`generated can preventthe capacity to meaningfully select a desired candidate because of
`the presenceof the so-called “noise” of so many undesired candidates.
`-l-
`
`20
`
`25
`
`30
`
`35
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`Another approach,referred to as “Walk Through” mutagenesis(see,e.g., U.S.
`Patent Nos: 5,830,650; 5,798,208) has been used to mutagenize a defined region of a
`polypeptide by synthesizing a mixture of degenerate oligonucleotides that, statistically,
`contain a desired set of mutations. However, because degenerate polynucleotide
`synthesis is employed, Walk-Through mutagenesis yields a number of undesired
`alterations in additionto the desired set of mutations. For example, to sequentially
`introduce a mutation across a defined region of only five aminoacid positions, a set of
`over 100 polynucleotide must be made (and screened)(see,e.g., Fig. 6). Accordingly,
`to make and screen, for example, two or three regions becomesincreasingly complex,
`i.e., requiring the making and screening of 200 to over 300 polynucleotides,
`respectively, for the presence of only 10 to 15 mutations.
`In yet another approach which has been used to mutagenize proteinsis alanine
`scanning mutagenesis, where an alanine residue is “scanned” througha portion of a
`protein to identify positions where the protein’s function is interrupted. However, this
`approach only looksat loss of protein function by way of substituting a neutral alanine
`residue at a given position, rather than gain or improvement of function. Thus,it is not a
`useful approach for generating proteins having improvedstructure and function.
`Accordingly, a need remainsfor a systematic way to mutagenizea protein for
`new or improved function.
`
`Summary of the Invention
`The invention pertains to a method of mutagenesis for the generation of novel or
`improved proteins (or polypeptides) andto libraries of polypeptide analogs and specific
`polypeptides generated by the methods. The polypeptide targeted for mutagenesis can
`be a natural, synthetic or engineered polypeptide, including fragments, analogs and
`mutant formsthereof.
`In one embodiment, the method comprises introducing a predetermined amino
`acid into essentially every position within a defined region (or several different regions)
`of the amino acid sequence of a polypeptide. A polypeptide library is generated
`containing polypeptide analogs which individually have no more than one
`predetermined aminoacid, but which collectively have the predetermined aminoacid in
`every position within the defined region(s). Alone, this method can be referred to as
`“look-through” mutagenesis because,in effect, a single, predetermined amino acid (and
`only the predetermined aminoacid) is substituted position-by-position throughout one or
`more defined region(s) of a polypeptide.
`However, in a preferred embodiment, the LTM method is improved by usingit to
`identify functional amino acids(or so-called “hot spots”) from non-functional amino
`acids (or so-called “cold spots”) within a polypeptide, or portion thereof, to further
`
`-2-
`
`20
`
`25
`
`30
`
`35
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`reduce the numberof residuesto be altered in order to screen and obtain a desired
`
`property in a polypeptide. Accordingly, the improved method of look-through
`mutagenesis (LTM)(hereafter the improved LTM beingreferred to as LTM2)allowsfor
`the identification and building of a subset of candidate molecules representing only the
`mostrelevant functional alterations in the polypeptide which can then beefficiently
`screened free of any “noise”. Importantly, LTM2 also allowsfor the construction of an
`LTM2library having superior advantages overtraditionallibraries because it has been
`designed to include only alterations in the amino acid residues of the polypeptide most
`likely to have an effect on the function of the polypeptide and therefore, upon screening,
`more likely to yield an altered polypeptide having an enhanced property. Thus, LTM2
`allows one to “look-through”the structural and functional consequencesof separately
`substituting a predetermined aminoacid at each functional amino acid position within a
`defined region of the polypeptide, thereby segregating a specific protein chemistry to the
`defined region without any interference or “noise” from the generation of unwanted
`polypeptide analogs(i.e., analogs containing aminoacid substitutions other than those
`that follow the LTM2 scheme) (see, for example, Fig. 1).
`Accordingly, the present invention allowsfor highly efficient and accurate
`systematic evaluation of the role of a specific amino acid changein one or more defined
`regions of a polypeptide. This becomesparticularly important when evaluating (by
`mutating) two or more defined regions, such that the number of polypeptide analogs
`required greatly increases and, thus, the presence of undesired analogs also increases.
`The present invention obviates this problem by completely eliminating undesired
`analogs and,thus, the potential that any changesin protein structure or function
`observedare the result of anything but substitution of the predetermined aminoacid.
`Thus, the effect of segregating a specific protein chemistry to even multiple regions with
`a protein can be studied with high accuracy andefficiency. Importantly, this includes
`. Studying how mutagenesis can effect the interaction of such regions, thereby improving
`the overall structure and function of the protein.
`In a particular embedmentof the invention, the methodsof the invention are
`suitable for identifying a particular chemical motif that maps to one or more functional
`aminoacid resides or positions. The aminoacid residue(s) that contribute to such a
`chemical motif can occurat one or more positions that are contiguous, non-contiguous,
`within one or more CDRregions, and/or within one or more polypeptides, for example,
`antibody heavy and light chains. The methodsofthe invention allow for the further
`exploration of a chemical motif in that they allow for the systematic testing (or chemical
`profiling) of related amino acid chemistries at selected amino acid position(s) or defined
`region(s), Accordingly, in one embodiment, the invention provides a method for
`identifying a desired chemistry and then exploring the consequencesof incorporating
`
`-3-
`
`20
`
`25
`
`30
`
`35
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`related or unrelated chemistries to achieve either an enhanced property or to remove a
`deleterious property. Typical amino acid side chain chemistries suitable for profiling by
`the methodsof the invention are polar, positively charged, negatively charged, and
`hydrophobic aminoacid side chain chemistries. In one embodiment, a charged
`chemistryis identified as resident at a selected aminoacid reside(s), position, or defined
`region(s) and other charged aminoacids are substituted for the parental amino acid such
`that an alteration in a measurable property is achieved.
`In a preferred embodiment, the
`alteration in a measurable property is an enhancedproperty in an antibody, for example,
`improved antigen-binding or effector function.
`Accordingly, the invention also provides antibody libraries comprising related
`aminoacid side group chemistries introduced at selected amino acid positions(s)/
`defined regions having, for example, related chemistry, for the efficient screening of
`antibodies with improved properties.
`In another embodimentof the invention, the library of polypeptide analogsis
`generated and screenedbyfirst synthesizing individual polynucleotides encoding a
`defined region or regions of a polypeptide where, collectively, the polynucleotides
`representall possible variant polynucleotides according to the look-throughcriteria
`described herein. The methodis used to identify and distinguish functional amino acid
`residue(s) (positions) from non-functional amino acid residue(s) (positions). A subset of
`variant polynucleotides are expressed, for example, using in vitro transcription and
`translation and/or using a display technology, such as ribosomedisplay, phage display,
`bacterial display, yeast display, arrayed display, or any other suitable display system
`knownintheart.
`
`The expressed polypeptides are then screened and selected using functional
`assays, such as binding assays or enzymatic/catalytic assays. In one embodiment, the
`polypeptides are expressed in association with the polynucleotide that encodes the
`polypeptide, thereby allowing for identification of the polynucleotide sequence that
`encodes the polypeptide.
`In yet another embodiment, the polypeptidesare directly
`synthesized using protein chemistry.
`In yet another embodimentof the invention, a combinatorial beneficial library of
`the V, and V,; CDR amino-acid sequencevariations is constructed. This second library
`is constructed by generating coding sequences having, at each aminoacid variation
`position, codonsfor the wildtype amino acid and for each of the previously identified
`beneficial variant amino acids atthat position.
`Thus, the present invention provides a methodofintelligent mutagenesis that can
`be used to generate libraries of polypeptide analogsthat are of a practical size for
`screening, in part, because the libraries are devoid of any undesired analog polypeptides
`or so-called noise. The method can be usedto studythe role of specific amino acidsin
`
`-4-
`
`30
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`polypeptide structure and function and to develop new or improved polypeptides such as
`antibodies, binding fragments or analogs thereof, single chain antibodies, catalytic
`antibodies, enzymes, and ligands. In addition, the method can be performed with the
`benefit of a priori information, e.g., via computer modeling, that can be used to select an
`initial subset of polypeptide analogs to be produced and studied using LTM2.
`Other advantages and aspects of the present invention will be readily apparent
`from the following description and Examples.
`
`BriefDescription of the Figures
`Figure J illustrates the advantages of improved LTM (LTM2) over LTM inthat
`functional amino acids are distinguished from non-functional amino acids such that a
`more beneficial subset of candidate molecules is obtained and screened.
`
`Figure 2 illustrates a general approach for the use of polymerase chain reaction
`(PCR)to build defined regions of an antibody heavy andlight chain for identifying
`functional aminoacid residues, into a larger gene context.
`Figure 3 illustrates the arrangementofvariable light-chain (V_.) and variable
`heavy chain (Vj) CDRsin a synthetic single chain antibody (scFv) anti-ovalbumin gene
`context.
`In the application of LTM,a leucine aminoacid is introduced into each of the
`fourteen residues 56-69 in Vy CDR2 ofthe antibody. Forthe application of LTM2,only
`those residues identified as functional are further explored by mutagenesis.
`Figure 4 illustrates the by single overlap extension polymerase chain reaction
`(SOE-PCR)for the production of an LTM Vy CDR2 library; the production of multiple
`LTM Vu CDRlibraries; and an array of LTM library combinations containing both Vy and
`V,_ CDRs
`Figure 5 illustrates the diversity of the libraries of the invention with the x and y
`axes of the matrix representing the CDRsofeachof the light and heavy chains wherein an
`“0” indicates a wild-type CDR and a “1!”indicates a mutated CDR andtheintersected
`numberrepresenting the complexity of the resultant subset library (e.g., 4 means four
`CDRsare simultaneously mutated).
`Figure 6 showsa schematic of a yeast expression vector for displaying proteins of
`interest, e.g., polypeptide analogs of the invention, on the surface of yeast for efficient
`identification of function (phenotype) and corresponding encoding sequence (genotype).
`Figure 7 represents a Fluorescence-Activated Cell Sorter (FACS™) plot of the
`binding of biotinylated ovalbumin and streptavidin FITC to wild type anti-ovalbumin
`scFv (gray line); pYD1 vector alone (solid gray area); and contro] scFv (blackline).
`Figure 8 represents Fluorescence-Activated Cell Sorter (FACS™) plots showing
`a selection gate (the R1 trapezoid) for identifying only those LTM clonesthat expressed
`the scFv fusion with a higher binding affinity to ovalbumin than the anti-ovalbumin wild
`
`-5-
`
`20
`
`25
`
`30
`
`35
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`type antibody (left panel), the distribution of binding affinities of the total LTM library
`(center panel), and a post sort FACSanalysis (right panel) to confirm that >80% of the
`pre-screen anti-ovalbumin scFv clones were within the predeterminedcriteria.
`Figure 9 illustrates steps in the screening of scFv antibodies (e.g., anti-ovalbumin)
`formed in accordance with the present invention for improved bindingaffinity based on
`equilibrium binding kinetics (e.g., to ovalbumin).
`Figure 10 showsequilibrium binding curves for anti-ovalbumin scFv expressing
`cells prior to selection (circles), after one round of selection(light triangles), after two
`roundsofselection (dark triangles), and for the anti-ovalbumin wild-type reference
`antibody (black squares).
`Figure 1] illustrates typical steps for screening of antibodies formed in accordance
`with the present invention for high binding affinity based on particular binding kinetics,
`e.g., antibody K,constants, using the test antigen ovalbumin.
`Figure 12 showsthe identification of enhanced properties in two clones (i.e.,
`higher relative Kor as compared to an reference antibody (square)) using the methods of
`the invention.
`
`Figure 13 represents the enhanced properties (see fold better than wild type) of a
`subset of improved clones having lower ECs9 values with respect to an anti-ovalbumin
`wild-type reference antibody control (square).
`Figure 14 shows a matrix representing the functional (hot spots) and non-functional
`(cold spots) aminoacid positions of an exemplary antibody. Mutations associated with
`enhancedaffinity (relative to the reference wild type antibody) based on equilibrtum
`binding (ECs) and/or kinetic binding experiments are shown below each Vy and VL CDR
`position.
`
`Detailed Description of the Invention
`In order to provide a clear understanding of the specification and claims, the
`following definitions are provided below.
`
`Definitions
`Asused herein the term “analog”refers to a variant or mutant polypeptide (or a
`nucleic acid encoding such a polypeptide) having one or more aminoacid substitutions.
`The term “binding molecule” refers to any binding molecule, including proteins,
`polypeptides, and peptides that bind to a substrate or target.
`In one embodiment, the
`binding moleculeis an antibody or binding fragment thereof(e.g., a Fab fragment),
`single domain antibody, single chain antibody (e.g., scFv), or peptide capable of binding
`a ligand.
`In another embodiment, the binding molecule, in particular, binding molecules
`comprising CDRregion(s), can comprise nontraditional scaffolds or framework regions
`
`-6-
`
`20
`
`25
`
`30
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`derived from other antibodies, immunoglobulins, or immunoglobulin-like molecules
`(e.g., fibronectin), or be in part or in whole, of synthetic origin.
`The term “defined region”refers to a selected region of a polypeptide.
`Typically, the defined region includesall or a portion of a functionalsite, e.g., the
`bindingsite of a ligand, the bindingsite of a binding molecule or receptor, or a catalytic
`site. The defined region mayalso include multiple portions of a functional site. For
`example, the defined region can includeall, a portion, or multiple portions of a
`complementarity determining region (CDR), ¢.g., a single domain binding region, or a
`complete heavy and/or light chain variable region (Fv) of an antibody. Thus, a
`functional site may include a single or multiple defined regions that contribute to the
`functional activity of the molecule.
`The terms “functional amino acid(s)” and “non-functional amino acid(s)”refer
`to, respectively, the amino acid residues (or corresponding aminoacid residue position)
`within a polypeptide (or portion thereof) that are determined (using, for example, the
`methods of the invention) to contribute to a measurable property or activity of the
`polypeptide. Accordingly, a functional amino acid residue(s) (or corresponding
`position(s)) is referred to as a “hot spot(s)” as it is a residue or residue position that
`influences the activity of the polypeptide as compared to a non-functional residue(s) or
`position(s) which does notinfluence the activity of the polypeptide and therefore
`referred to as a “cold spot(s)”. A functional amino acid residue (or position) is
`distinguished from a non-functional amino acid residue (or position) as being suitable
`for mutagenesis. Typically, when applying the methodsof the invention to the
`investigation of an antibody molecule, amino acid residuesthat alter, for example,
`antigen binding, are considered functional residues/positions(i.e., hot spots) whereas
`residues that do not alter such binding are referred to as non-functional
`residues/positions(i.e., cold spots).
`The term “measurable property” refers to a functional property or activity of a
`polypeptide (or portion thereof) that can be measured, determined, or assayed for, using
`standard techniques and include, bindingactivity, kinase activity, catalytic activity,
`thermalstability, or enzymatic activity. Measurable properties of polypeptides that are
`antigen-binding polypeptides,e.g., antibodies, typically include binding specificity,
`binding avidity, binding affinity, Fc receptor binding, glycosylation, complement
`binding, half-life stability, solubility, thermal stability, catalytic activity, and enzymatic
`activity.
`- The term “look-through mutagenesis” or “LTM”refers to a method for
`introducing a predetermined aminoacid into essentially every position within a defined
`region (or several different regions) of the amino acid sequence of a polypeptide. A
`polypeptide library is generated containing polypeptide analogs whichindividually have
`
`-7-
`
`20
`
`25
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`20
`
`25
`
`no more than one predetermined aminoacid, but which collectively have the
`predetermined aminoacid in every position within the defined region(s).
`The term “improved look-through mutagenesis” or “LTM2”refers to LTM
`conductedsoas to identify or distinguish functional amino acid residues(hot spots) from
`non-functional amino acid residues (cold spots). Accordingly, the LTM2 methodallows
`for selectively introducing a predetermined aminoacid into the functional amino acid
`residue positions within a polypeptide (or portion thereof). Corresponding LTM2
`libraries are therefore enriched for polypeptides analogs having aminoacid alterations
`mostlikely to confer an altered or enhanced property. LTM2 can becarried out
`subsequent to LTM or based on a priori informationasto the functionality of a given
`aminoacid residue or residue position.
`The term “library” refers to two or more molecules mutagenized according to the
`method of the invention. The molecules of the library can be in the form of
`polynucleotides, polypeptides, polynucleotides and polypeptides, polynucleotides and
`polypeptidesin a cell free extract, or as polynucleotides and/or polypeptides in the
`context of a phage, prokaryotic cells, or in eukaryotic cells. Libraries of the invention
`can contain 2 or more molecules or polypeptide analogs, for example about 2 to 10,
`about 10 to 50, about 50 to 10”, about 10°, about 10%, about 10°, about 10°, about 107,
`about 10°, about 10°, about 10'°, about 10!', about 10'* , about 10'%, or more, or any
`interval or range of the foregoing.
`The term “mutagenizing”refers to the alteration of an amino acid sequence.
`This can be achieved by altering or producing a nucleic acid (polynucleotide) capable of
`encoding the altered amino acid sequence,or by the direct synthesis of an altered
`polypeptide using protein chemistry.
`The term “mutagenesis” refers to, unless otherwise specified, any art recognized
`technique for altering a polynucleotide or polypeptide sequence. Preferred types of
`mutagenesis include walk-through mutagenesis (WTM), beneficial walk-through
`mutagenesis, look-through mutagenesis (LTM), improved look-through mutagenesis
`(LTM2), or combinations thereof.
`The term “combinatorial beneficial mutagenesis” refers to a combination library
`of coding sequencesthat encode degenerate mixtures of Vi and/or Vj; CDR amino-acid
`sequencevariations initially identified from the predetermined LTM aminoacid
`mutagenesis screen as having an alteration on a measurable property.
`In the
`combinatorial beneficial mutation approach, oligonucleotide coding sequences are
`generated which represent combinationsof these beneficial mutationsidentified by LTM.
`These combinations may be combinationsofdifferent beneficial mutations within a single
`CDR,mutations within two or more CDRswithinasingle antibody chain, or mutations
`within the CDRsofdifferent antibody chains.
`
`30
`
`35
`
`-8-
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`The term “polynucleotide(s)” refers to nucleic acids such as DNA molecules and
`RNA molecules and analogs thereof (e.g., DNA or RNA generated using nucleotide
`analogs or using nucleic acid chemistry). As desired, the polynucleotides may be made
`synthetically, e.g., using art-recognized nucleic acid chemistry or enzymatically using,
`e.g.,a polymerase. Typical modifications include methylation, biotinylation, and other
`art-known modifications.
`In addition, the nucleic acid molecule can be single-stranded
`or double-stranded and, where desired, linked or associated (e.g., covalently or non-
`covalently) to a detectable moiety.
`The term “variant polynucleotide” refers to a polynucleotide encoding a
`corresponding polypeptide analog (or portion thereof) of the invention. Thus, variant
`polynucleotides contain one or more codonsthat have been changedto result in
`expression of a different amino acid.
`The term “polypeptide(s)” refers to two or more amino acids joined by a peptide
`bond,e.g., peptides (e.g., from 2 to ~50 aminoacid residues), as well as longer peptide
`sequencesé.g., protein sequences which typically comprises amino acid sequences from
`as few as 50 aminoacid residues to more than 1,000 aminoacid residues.
`The term “pooling” refers to the combining of polynucleotide variants or
`polypeptide analogs to form libraries representing the look-through mutagenesis (LTM)
`or improved look-though mutagenesis (LTM2)ofan entire polypeptide region. The
`molecules maybe in the form of a polynucleotide and/or polypeptide and may coexistin
`the form of a sublibrary, as molecules on a solid support, as moleculesin solution,
`and/or as molecules in one or more organisms(e.g., phage, prokaryotic cells, or
`eukaryotic cells).
`The term “predetermined amino acid”refers to an amino acid residue selected
`for substitution at each position within a defined region of a polypeptide to be
`mutagenized. This doesnotinclude position(s) within the region that already (e.g.,
`naturally) contain the predetermined aminoacid and, thus, which need not be substituted
`with the predetermined aminoacid. Accordingly, each polypeptide analog generated in
`accordance with the present invention contains no more that one “predetermined amino
`acid” residue in a given defined region. However, collectively, the library of
`polypeptide analogs generated contains the predetermined aminoacid at each position
`within the region being mutagenized, and in a preferred embodiment, at amino acid
`positions determined to be functional (hot spots). Typically, a predetermined amino
`acid is selected for a particular size or chemistry usually associated with the side group
`of the amino acid. Suitable predetermined amino acids include, for example, glycine
`and alanine (sterically small); serine, threonine, and cysteine (nucleophilic); valine,
`
`-9.
`
`20
`
`25
`
`30
`
`35
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`leucine, isoleucine, methionine, and proline (hydrophobic); phenylalanine, tyrosine, and
`tryptophan (aromatic); aspartate and glutamate (acidic); asparagine, glutamine, and
`histidine (amide); and lysine and arginine (basic). Use of non-traditional amino acid
`residues (e.g., homocysteine) are also within the scope of the invention and can be
`introduced using any art recognized techniques.
`
`Detailed Description
`The study of proteins has revealed that certain amino acids play a crucial role in
`their structure and function. For example, it appears that only a discrete numberof
`amino acids participate in the binding of an antibody to an antigen or are involved in the
`catalytic event of an enzyme.
`Thoughit is clear that certain amino acidsarecritical to the activity or function
`of proteins, it is difficult to identify which aminoacids are involved, how they are
`involved, and what substitutions can improve the protein’s structure or function.
`In part,
`this is due to the complexity of the spatial configuration of amino acid side chains in
`polypeptides and theinterrelationship of different portions of the polypeptide that
`contribute to form a functional site. For example, the interrelationship between the six
`CDRsofthe variable heavy and light chain regions of an antibody contribute to the
`antigen or ligand-binding pocket.
`Previous mutagenesis methods, such as selective (site-directed) mutagenesis and
`saturation mutagenesis, are of limited utility for the study of protein structure and
`function in view of the enormous numberofpossible variations in complex
`polypeptides. This is especially true given that desirable combinationsare often
`accompanied by the presence of vast amounts of undesirable combinationsor so-called
`noise.
`
`The methodof this invention provides a systematic, practical, and highly
`accurate approach for evaluating the role of particular amino acids and their position,
`within a defined region of a polypeptide, in the structure or function of the polypeptide
`and, thus, for producing improved polypeptides.
`
`I. Selecting a Defined Region
`In accordance with the present invention, a defined region or regions within a
`protein are selected for mutagenesis. Typically, the regions are believed to be important
`to the protein’s structure or function. This can be deduced, for example, from what
`structural and/or functional aspects are known or can be deduced from comparing the
`defined region(s) to what is known from the study of other proteins, and may be aided
`by modeling information. For example, the defined region can be onethat has a role ina
`functionalsite, e.g., in binding, catalysis, or another function.
`In one embodiment, the
`
`-10-
`
`20
`
`25
`
`30
`
`35
`
`

`

`WO 2006/023144
`
`PCT/US2005/024140
`
`defined region is a hypervariable region or complementarity determining region (CDR)
`of an antigen binding molecule (see, ¢.g., Fig. 1). In another embodiment, the defined
`regionis a portion of a complementarity determining region (CDR).
`In other
`embodiments, two or more defined regions, e.g., CDRsor portions thereof, are selected
`for mutagenesi