US 2012O164068A1 (19) United States (12) Patent Application Publication (10) Pub. No.: US 2012/0164068 A1 Hudson et al. (43) Pub. Date: Jun. 28, 2012

(54) MMUNO-CONUGATES AND METHODS Publication Classification FOR PRODUCING THEM (51) Int. Cl. A 6LX 39/395 (2006.01) Inventors: Peter John Hudson, Blackburn CI2N 5/3 (2006.01) (76) CI2N 15/63 (2006.01) (AU); Paul Robert Sanders, CI2P 2L/00 (2006.01) Moonee Ponds (AU); Debra CI2N 5/071 (2010.01) Tamvakis, Camberwell (AU); A6II 5L/It (2006.01) Fabio Turatti, Docklands (AU); C07K I6/46 (2006.01) Michael Paul Wheatcroft, CI2N 5/62 (2006.01) Southbank (AU); David Leong, CI2N 5/10 (2006.01) Preston (AU) A6IP35/00 (2006.01) CI2N I/2 (2006.01) CI2N I/19 (2006.01) (21) Appl. No.: 13/383,578 C07K 6/00 (2006.01) A61R 49/00 (2006.01) PCT Fled: Jul. 2, 2010 (52) U.S. Cl...... 424/1.49; 530/387.1; 424/130.1; (22) 536/23.53; 435/320.1; 435/69.6; 435/375; 424/9.1:530/387.3; 424/133.1536/23.4: (86) PCT NO.: PCT/AU2O1 O/OOO847 435/328; 435/252.3:435/252.33:435/252.31; 435/252.34; 435/252.35:435/326; 435/254.23; S371 (c)(1), 435/254.21 (2), (4) Date: Mar. 16, 2012 (57) ABSTRACT The present invention provides an isolated protein compris Related U.S. Application Data ing an immunoglobulin variable region comprising at least two cysteine residues positioned within framework region 1 (60) Provisional application No. 61/223,353, filed on Jul. 6, such that if at least two of the cysteine residues are not 2009, provisional application No. 61/256,703, filed on conjugated to another compound a disulphide bond forms Oct. 30, 2009. between the cysteine residues. Preferably the protein com prises an immunoglobulin heavy chain variable region and an immunoglobulin light chain variable region, wherein at least one of the variable regions comprises the two cysteine resi (30) Foreign Application Priority Data dues. The present invention also provides a protein that binds to TAG72. The present invention also provides conjugates of Jul. 3, 2009 (AU) ...... 20099.03127 the protein and another compound. Patent Application Publication Jun. 28, 2012 Sheet 1 of 52 US 2012/O164068 A1

i. Patent Application Publication Jun. 28, 2012 Sheet 2 of 52 US 2012/O164068 A1

E. Patent Application Publication Jun. 28, 2012 Sheet 3 of 52 US 2012/O164068 A1

s

i i Patent Application Publication Jun. 28, 2012 Sheet 4 of 52 US 2012/O164068 A1

i Patent Application Publication Jun. 28, 2012 Sheet 5 of 52 US 2012/O164068 A1

Patent Application Publication Jun. 28, 2012 Sheet 6 of 52 US 2012/O164068 A1

a resisiae ass.

& 8

3:...'

ig:rk & Patent Application Publication Jun. 28, 2012 Sheet 7 of 52 US 2012/O164068 A1

i Patent Application Publication Jun. 28, 2012 Sheet 8 of 52 US 2012/O164068 A1

Patent Application Publication Jun. 28, 2012 Sheet 9 of 52 US 2012/O164068 A1

Patent Application Publication Jun. 28, 2012 Sheet 10 of 52 US 2012/O164068 A1

ir-stated FR w1 H13-H16

jr-rritated FR: 7- 4-7

Figure 7B Patent Application Publication Jun. 28, 2012 Sheet 11 of 52 US 2012/O164068 A1

888: 8& Patent Application Publication Jun. 28, 2012 Sheet 12 of 52 US 2012/O164068 A1

*************************

88:88:8: , ix:

igare:

Patent Application Publication Jun. 28, 2012 Sheet 13 of 52 US 2012/O164068 A1

Exckstratic

8 & &: 8: 8: 8: : 8: 88: 8: 38

88 (Xixx Patent Application Publication Jun. 28, 2012 Sheet 14 of 52 US 2012/O164068 A1

oord os? by? oz?

gotaanã, Patent Application Publication Jun. 28, 2012 Sheet 15 of 52 US 2012/O164068 A1

Patent Application Publication Jun. 28, 2012 Sheet 16 of 52 US 2012/O164068 A1

Patent Application Publication Jun. 28, 2012 Sheet 17 of 52 US 2012/O164068 A1

Patent Application Publication Jun. 28, 2012 Sheet 22 of 52 US 2012/O164068 A1

88.3:38 3. :

&&&&&&&&&

: 88: 88 38

8

×3×3×3×3&&3&??&*&

* ** * ? «; & * * & & 8 & ------******************* 8::::::::: 88: 88::::::::

Patent Application Publication Jun. 28, 2012 Sheet 23 of 52 US 2012/O164068 A1

8,833.83 8:::::::::::

3:8:88:8::::::::::x:y

& - 8 8 38 & 88:88: 3888:

*igi & 32 Patent Application Publication Jun. 28, 2012 Sheet 24 of 52 US 2012/O164068 A1 Ápoq?AVp?XeN

ÁÐEd—>

auen 8/9?7ZI Patent Application Publication Jun. 28, 2012 Sheet 25 of 52 US 2012/O164068 A1

es "es O b k t B Ch. O --

Patent Application Publication Jun. 28, 2012 Sheet 26 of 52 US 2012/O164068 A1

es es O 2 s k s o e CD

Patent Application Publication Jun. 28, 2012 Sheet 27 of 52 US 2012/O164068 A1

Patent Application Publication Jun. 28, 2012 Sheet 28 of 52 US 2012/O164068 A1

... 3.....8.-s :*:::::::::::::::::::::::::::::::::::::::::::::::::...... :...--

Counts vs DeConvoluted Mass (annu) Figure i is Patent Application Publication Jun. 28, 2012 Sheet 29 of 52 US 2012/O164068 A1

8,

Figure iC Patent Application Publication Jun. 28, 2012 Sheet 30 of 52 US 2012/O164068 A1

: *WXW-wx-w------&x &

XY

Patent Application Publication Jun. 28, 2012 Sheet 31 of 52 US 2012/O164068 A1

8x888-83-888,

$8-3.33 3:3:Xiy *:::::::::::::

E.

s : $8 gogo &%sorror 88: 3888 888 8x888-88-883.

3,333.38 ::::::::y

::::::::::::::: *g: 83.38

Patent Application Publication Jun. 28, 2012 Sheet 33 of 52 US 2012/O164068 A1

AVP02-101-PEG L8-L11 Diabody

to Time (min)

Figure 15C Patent Application Publication Jun. 28, 2012 Sheet 34 of 52 US 2012/O164068 A1

s

&

: : s: & Patent Application Publication Jun. 28, 2012 Sheet 35 of 52 US 2012/O164068 A1

?*****g**

********????

?| §§§§ |bodovi ?podozi •gogov No.goboz

83xx + x83888: 83:83 &883.383 Patent Application Publication Jun. 28, 2012 Sheet 36 of 52 US 2012/O164068 A1

Aw-- (scar six

&xs xx is:

. Kok Kitikey x. w: 8xxx & 3

is axed witxxy * is ixitx: 8x8xxy is is 8x88:graisix8 is $3.8x88g.8xxixe.

Figure 88.

Patent Application Publication Jun. 28, 2012 Sheet 38 of 52 US 2012/O164068 A1

a-assssssssss-XYYYXY^ssssss-YYYYYYYXYYYYYX'ssyr-Yss'ss'sswww's'ssssssssssssssssssssssssssssa-assssssaal-seas------a-a-a-a-a-assssssssssss

sessssssssssex-axssssss-XaX-axxxxxaaxe-wasssssssssssss-XXX-Sesssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssss---a-

s

8. 8.:::: s.s

sigrise 8 Patent Application Publication Jun. 28, 2012 Sheet 39 of 52 US 2012/O164068 A1

8

& i. &

: x 8 | Doravs-PEG,3888 AVP0407 Patent Application Publication Jun. 28, 2012 Sheet 40 of 52 US 2012/O164068 A1

(?oj??uou)zo-woday Patent Application Publication Jun. 28, 2012 Sheet 41 of 52 US 2012/O164068 A1

E - ... is is 9. . Patent Application Publication Jun. 28, 2012 Sheet 42 of 52 US 2012/O164068 A1

AVPO4-07 (non hio D

; : ; ; ; ; ; ; ; ; , , , 8 x: at exi witxy in six-xerxtraft rice Patent Application Publication Jun. 28, 2012 Sheet 43 of 52 US 2012/O164068 A1

AVPO4-07 (non Thio)

--Spleen -- Kidney

lung *** - Carcass

; ; ; ; ; ; ; ; 48 72 ge irrers 25 labeled Avibody in S-74 xenograft mice

Figure 22B Patent Application Publication Jun. 28, 2012 Sheet 44 of 52 US 2012/O164068 A1

:8

883 askie X-S- 34-Y-- Awixxy iris-execrat rice *igare: 2C Patent Application Publication Jun. 28, 2012 Sheet 45 of 52 US 2012/O164068 A1

Figure 23B Patent Application Publication Jun. 28, 2012 Sheet 46 of 52 US 2012/O164068 A1

woxiaixxy exo-axxy Ex. Costigate 3 xiiaxxy °C. cxx88 x-axxy 8834 (Exigex

s x 38 irre irs igrate 24 Patent Application Publication Jun. 28, 2012 Sheet 47 of 52 US 2012/O164068 A1

& x^-*.^e----- * 8 : 88x3-8-3-83-8:

3. 8:8:8: 8, ::::::::: 3. 8: 8. & XXXs 8

88-888-8-8-8 Patent Application Publication Jun. 28, 2012 Sheet 48 of 52 US 2012/O164068 A1

Figure 26A Patent Application Publication Jun. 28, 2012 Sheet 49 of 52 US 2012/O164068 A1

Figure 26B Patent Application Publication Jun. 28, 2012 Sheet 50 of 52 US 2012/O164068 A1

so so

« Catax.

Figure 26C Patent Application Publication Jun. 28, 2012 Sheet 51 of 52 US 2012/O164068 A1

8 ed Patent Application Publication Jun. 28, 2012 Sheet 52 of 52 US 2012/O164068 A1

Figure 27A

Figure 278

Figure 27C US 2012/0164068 A1 Jun. 28, 2012

IMMUNO-CONUGATES AND METHODS various conjugate species within the heterogeneous mixture FOR PRODUCING THEM resulting from a conjugation reaction. 0005. Furthermore, non-specific conjugation of a com CROSS-REFERENCE TO RELATED pound to an antibody or protein comprising an antigen bind APPLICATIONS ingdomain thereofmay reduce or completely prevent binding of the antibody/protein to an antigen, for example, if the 0001. This application claims priority from Australian compound is conjugated to a region required for antigen Patent Application No. 2009903127 entitled “Immuno-con binding. This risk is increased in proteins that comprise anti jugates and methods for producing them filed on 3 Jul. 2009: gen binding domains that are far Smaller than an intact anti U.S. Patent Application No. 61/223,353 entitled “Immuno body in which there may be few residues suitable for conju conjugates and methods for producing them filed on 6 Jul. gation that are not important for antigen binding. For 2009; and U.S. Patent Application No. 61/256,703 entitled example, proteins comprising little more than antigenbinding 'Variable domain molecules and methods of use filed on 30 domains of an antibody have few sites to which a compound Oct. 2009, the entire contents of which are incorporated by can be conjugated without reducing or preventing antigen reference. binding. 0006 Carbohydrate(s) on the Fc region of an antibody is a FIELD OF INVENTION natural site for attaching compounds. Generally, the carbo 0002 The present invention relates to proteins comprising hydrate is modified by periodate oxidation to generate reac immunoglobulin variable regions modified to facilitate con tive aldehydes, which can then be used to attach reactive jugation of a compound thereto or having a compound con amine containing compounds by Schiff base formation. As jugated thereto. the aldehydes can react with amine groups, reactions are carried out at low pH so that lysine residues in the antibody or BACKGROUND OF THE INVENTION antigen binding domain are protonated and unreactive. Hydrazide groups are most Suitable for attachment to the 0003. The highly specific binding nature of immunoglo aldehydes generated since they are reactive at low pH to form bulins, e.g., antibodies and antibody-like molecules (e.g., a hydrazone linkage. The linkage can then be further stabi camelid immunoglobulin or immunoglobulin new antigen lized by reduction with sodium cyanoborohydride to form a receptors (IgNARS) from cartilaginous fish) or proteins com hydrazine linkage (Rodwell et al., 1986). Disadvantages of prising antigenbinding domains thereof makes them particu this approach include the harsh conditions required for link larly Suitable for delivering molecules to specific targets in a age which can damage and aggregate Some antibody mol Subject. For example, immunoglobulins or proteins compris ecules. For example, methionine residues present in some ing antigen binding domains thereof can be conjugated to antibody variable regions may be particularly Susceptible to cytotoxic or cytostatic compounds e.g., drugs, to kill or oxidation by periodate which can lead to loss of antigen inhibit growth of cells, such as tumour cells (Lambert, 2005). binding avidity. Histidine and/or tryptophan residues are also Such a conjugate facilitates targeted delivery of the cytotoxic Susceptible to oxidation. Furthermore, many proteins com or cytostatic compounds to cells expressing the antigen to prising antigen binding domains of an antibody do not nec which the immunoglobulin or fragment binds, rather than essarily comprise a Fc region, meaning that they cannot be non-specifically throughout a subject. Such conjugates can conjugated to a compound using the foregoing process. permit use of compounds that are generally toxic to a subject 0007 Cysteine thiols are reactive at neutral pH, unlike by ensuring the delivery of toxic levels of the compound to the most amines which are protonated and less nucleophilic near site at which it is required rather than systemically within a pH 7. Since free thiol groups are relatively reactive, proteins Subject. Furthermore, conjugation of antibodies or proteins with cysteine residues often exist in their oxidized form as comprising antigen binding domains thereof to detectable disulfide-linked oligomers or have internally bridged disul compounds. Such as fluorophores or radioisotopes facilitates fide groups. Extracellular proteins generally do not have free detection of target molecules within a Subject, for example to thiols (Garman, 1997). Cysteine residues have been intro facilitate detection of diseased cells Such as cancer cells, e.g., duced into proteins by genetic engineering techniques to form using in vivo, imaging-based methods. covalent attachments to ligands or to form new intramolecu 0004 Conventional means of linking a compound to an lar disulfide bonds. However, inserting or Substituting cys antibody or a protein comprising antigen binding domain teine thiol groups into a protein is potentially problematic, generally leads to a heterogeneous mixture of molecules particularly in the case of those which are relatively acces where the compounds are attached at a number of sites on the sible for reaction or oxidation, i.e., positioned at sites useful antibody. For example, compounds have typically been con for conjugation of a compound. This is because, in concen jugated to an antibody or protein comprising antigen binding trated solutions of the protein, whether in the periplasm of domains thereof through the often-numerous lysine residues Escherichia coli, culture Supernatants, or partially or com in the antibody or antigen binding domain, generating a het pletely purified protein, cysteine residues on the surface of the erogeneous antibody-compound conjugate mixture. Depend protein can pair and oxidize to form intermolecular disulfides, ing on reaction conditions used, the heterogeneous mixture and hence protein aggregates. Such protein aggregation often typically contains a distribution of conjugates with from 0 to leads to poor yields of isolated protein that is in a useful form, about 8, or more, attached compounds. In addition, within e.g., having a desired biological activity. Furthermore, the each subgroup of conjugates with a particular integer ratio of protein oxidatively can forman intramolecular disulfide bond compounds to antibody or protein there is a potentially het between the newly engineered cysteine and an existing cys erogeneous mixture where the compound is attached at vari teine residue, which can render the protein inactive or non ous sites on the antibody or protein. Analytical and prepara specific by misfolding or loss of tertiary structure. Each of the tive methods are inadequate to separate and characterize the foregoing problems are exacerbated in antibodies and pro US 2012/0164068 A1 Jun. 28, 2012 teins comprising antigenbinding domains thereof which gen if at least one of the cysteine residues is not conjugated to erally comprise several cysteine residues that bond with one another compound a disulphide bond is capable of forming another to ensure correct folding and Stability and, as a con between the cysteine residues. sequence antigen binding activity. 0013. In an alternative or additional example, the present 0008. It will be clear to the skilled artisan from the fore invention provides an isolated protein comprising an immu going that there is a need in the art for proteins comprising noglobulin heavy chain variable region (V) and an immu antigen binding domains of immunoglobulins that are modi noglobulin light chain variable region (V), wherein at least fied so as to permit simple conjugation of a compound one of the variable regions comprises at least two cysteine thereto. Preferred proteins will facilitate recombinant pro residues positioned within framework region (FR)1, wherein duction in a variety of systems, preferably without resulting if at least two of the cysteine residues are not conjugated to in considerably levels of multimeric aggregates linked by another compound a disulphide bond is capable of forming intermolecular bonds. between the cysteine residues. I0014 Preferably, the protein comprises at least one of V, SUMMARY OF INVENTION and at least one of V in a single polypeptide chain. 0009. In work leading up to the present invention, the 00.15 Preferably, the cysteine residues are positioned such inventors sought to identify sites within a variable region of that the disulphide bond is present under non-reducing con an immunoglobulin, e.g., an antibody that permit conjugation ditions. of a compound thereto without preventing binding of the 0016 Preferably, the cysteine residues are positioned variable region to an antigen. As exemplified herein, the between residue 2 numbered according to the Kabat number inventors have determined that numerous sites within frame ing system and complementary determining region (CDR)1. work region 1 (FR1) of a variable region that are accessible 0017. In one example, the cysteine residues are positioned for conjugation, and are Sufficiently removed from the anti within one or more loop regions of FR1. gen binding site of the variable region that a compound con 0018. In an alternative or additional example, the cysteine jugated thereto is unlikely to interfere with or prevent antigen residues are within the V and are positioned between resi binding. These sites are conserved in both heavy chain vari dues 2 to 30 numbered according to the Kabat numbering able regions (V) and light chain variable regions (V). Based system. Preferably, the cysteine residues are positioned on this determination, the inventors produced various pro between residues 7-20 and/or residues 24-30 numbered teins comprising mutated variable regions in which two cys according to the Kabat numbering system, and more prefer teine residues are inserted into FR1. These cysteine residues ably positioned between residues 7-20. In a further example, are positioned such that a disulfide bond can form between the residues are positioned between residues 6-16 numbered them. During recombinant production and/or purification, the according to the Kabat numbering system. In a further cysteine residues are linked by a disulphide bond thereby example, the residues are positioned between residues 7-16 reducing or preventing those residues bonding with other numbered according to the Kabat numbering system. cysteine residues either within the same protein or in another 0019. In an alternative or additional example, the cysteine protein. This reduces the likelihood of production of linked residues are within the V, and are positioned between resi multimers and/or an aberrantly folded variable region, and dues 2 to 22 numbered according to the Kabat numbering permits production and/or isolation of functional protein. Fol system. Preferably, the cysteine residues are positioned lowing isolation, the cysteine residues are reduced or other between residues 7-20 numbered according to the Kabat wise broken permitting conjugation of a compound to the numbering system. In a further example, the residues are protein. The inventors have also demonstrated that conjuga positioned between residues 7-19 numbered according to the tion of numerous compounds to these proteins, including Kabat numbering system. In a further example, the residues bulky compounds such as polyethylene glycol (PEG) and are positioned between residues 7-17 numbered according to imaging compounds, such as radioisotopes does not prevent the Kabat numbering system. binding of the variable region to an antigen. 0020. In an exemplified form of the invention the cysteine 0010. In one example, the present invention provides an residues are additional to a conserved cysteine residue in the isolated protein comprising an immunoglobulin variable V and/or V. The skilled artisan will be aware that the region comprising at least two cysteine residues positioned conserved cysteine residue is at residue 23 in the V, and/or within framework region (FR)1, wherein if at least one of the residue 22 in the V, numbered according to the Kabat num cysteine residues is not conjugated to a compound a disul bering system in at least a majority of naturally occurring phide bond is capable of forming between the cysteine resi antibodies. dues. 0021. In one preferred form of the invention the cysteine 0011. In another example, the present invention provides residues are positioned N-terminal to the conserved cysteine an isolated protein comprising an immunoglobulin variable residue. Preferably, the cysteine residues are positioned at one region comprising at least two cysteine residues positioned or more of the following: within framework region (FR)1, wherein if at least two of the 0022 (i) residue 8 and residue 11 of a K V, numbered cysteine residues are not conjugated to a compound a disul according to the Kabat numbering system; phide bond is capable of forming between the cysteine resi 0023 (ii) residue 14 and residue 17 of a K V, numbered dues. according to the Kabat numbering system; 0012. In an alternative or additional example, the present 0024 (iii) residue 7 and residue 11 of a V, numbered invention provides an isolated protein comprising an immu according to the Kabat numbering system; noglobulin heavy chain variable region (V) and an immu 0025 (iv) residue 14 and residue 17 of a wV, numbered noglobulin light chain variable region (V), wherein at least according to the Kabat numbering system; one of the variable regions comprises at least two cysteine 0026 (v) residue 8 and residue 12 of a v V, numbered residues positioned within framework region (FR)1, wherein according to the Kabat numbering system; US 2012/0164068 A1 Jun. 28, 2012

0027 (vi) residue 7 and residue 10 of a V, numbered 0047. In one example, the PEG has a molecular weight no according to the Kabat numbering system; and/or greater than about 4000 Da, for example, a molecular weight 0028 (vii) residue 13 and residue 16 of a V, numbered no greater than about 2000 Da, such as a molecular weight no according to the Kabat numbering system. greater than about 1,500 Da. In one example, the PEG has a 0029. In another preferred example of the invention, the molecular weight no greater than 1,000 Da. Such as, no cysteine residues are positioned at one or more of the follow greater than 900 Da, for example, no greater than 800 Da. ing: such as, no greater than 600 Da. In one example, the PEG has 0030 (i) residue 13 and residue 19 of a K V, numbered a molecular weight from about 550 Da to about 1,000 Da. according to the Kabat numbering system; 0048. In another example, the PEG has no more than about 0031 (ii) residue 13 and residue 19 of a V, numbered 70 or 75 or 77 ethylene glycol units. For example, the PEG has according to the Kabat numbering system; no more than about 50 ethylene glycol units. Preferably, the 0032 (iii) residue 6 and residue 9 of a V, numbered PEG has no more than 48 ethylene glycole units. For according to the Kabat numbering system; and/or example, the PEG has no more than about 40 ethylene glycol 0033 (iv) residue 12 and residue 18 of a V, numbered units. For example, the PEG has no more than about 30 according to the Kabat numbering system. ethylene glycol units. For example, the PEG has no more than 0034. The present invention clearly contemplates modify about 27 ethylene glycol units. For example, the PEG has no ing additional residues within the variable region or protein more than about 24 ethylene glycol units. For example, the comprising same. For example, the present inventors have PEG has no more than about 15 ethylene glycol units. For clearly demonstrated that Substituting residues positioned example, the PEG has no more than about 12 ethylene glycol between cysteine residues or even replacing cysteine residues units. Preferably, the PEG comprises about 12 to 27 ethylene naturally occurring within CDRs does not prevent a protein of glycol units. the invention binding to an antigen. 0049. In one example, a protein of the invention comprises 0035. The present invention also provides an isolated pro at least one polypeptide comprising a sequence at least about tein comprising a Fv comprising at least one protein of the 80% or 90% or 95% or 96% or 97% or 98% or 99% or 100% invention in which at least one V, binds to at least one V to identical to the sequence set forth in SEQ ID NO: 57 and form an antigen binding site. having conjugated to at least one of the cysteine residues in 0036. One form of the protein comprises the V, and the V FR1 a short monodispersed PEG. Preferably, the PEG com which form the antigenbinding site being in a single polypep prises 15 to 30 ethylene glycol units and preferably 24 ethyl tide chain. For example, the protein is: ene glycol units. The protein can comprise one or more and 0037 (i) a single chain Fv fragment (sclv); preferably less than 10 or 5 or 4 or 3 or 2 substitutions, 0038 (ii) a dimeric sclv (di-scEv); or preferably conservative amino acid substitutions or deletions 0039 (iii) at least one of (i) and/or (ii) linked to a Fc or a or insertions. Exemplary changes to the recited sequence heavy chain constant domain (C)2 and/or C3. include deleting the N-terminal serine or substituting the 0040 Alternatively, the protein comprises the V, and the serine for another amino acid residue (preferably a conserva V, which form the antigen binding site being in different tive amino acid Substitution) and/or deleting or Substituting polypeptide chains. In one example, each polypeptide chain the C terminal lysine and/or arginine. in the protein comprises a V and a V. Preferably, such a 0050. The inventors have also modified proteins compris protein is: ing variable regions to include a serine or threonine residue at 0041 (i) a diabody; the N-terminus. This residue permits site-specific conjuga 0042 (ii) a triabody; or tion of a compound thereto. By combining the N-terminal 0043 (iii) a tetrabody. serine/threonine mutation with the cysteine mutations dis 0044. In another example, the protein of the present inven cussed above, the inventors have produced proteins to which tion is an immunoglobulin, preferably an antibody. Exem they can site-specifically conjugate at least two different com plary forms of immunoglobulins are described herein and are pounds. to be taken to apply mutatis mutandis to the present example 0051. Accordingly, an example of the invention provides a of the invention. protein of the invention additionally comprises at least one 0045. In some examples of the invention, the protein of the N-terminal threonine or serine residue. The serine or threo invention comprises the cysteine residues being linked by a nine residue may be added to the N-terminus of the protein disulphide bond. Alternatively, the protein of the invention (i.e., is additional to the sequence of the protein). Preferably, comprises a compound conjugated to at least one of the cys the serine or threonine residue replaces a naturally occurring teine residues, wherein conjugation of the compound does not amino acid residue at the N-terminus of the protein, i.e., is the prevent binding of the protein to an antigen. Exemplary com result of a substitutional mutation. Optionally, the threonine pounds include a compound selected from the group consist or serine residue is linked to a compound Such as a compound ing of a radioisotope, a detectable label, a therapeutic com described above. pound, a colloid, a toxin, a nucleic acid, a peptide, a protein, 0052. In one example, a protein of the invention comprises a compound that increases the half life of the protein in a a first compound conjugated to at least one of the cysteine subject and mixtures thereof. The skilled artisan will appre residues and a second compound conjugated to the threonine ciate that the term protein encompasses proteins comprising or serine residue, wherein the second compound is different one or more immunoglobulin variable regions, for example, to the first compound. an antibody or fragment thereof including an Fv containing 0053. The present invention contemplates a protein that protein such as is described herein. can bind specifically to any antigen. Preferred proteins of the 0046. In one example, a protein of the invention is conju invention bind specifically to an antigen selected from the gated to polyethylene glycol (PEG). For example, the PEG is group consisting of human epidermal growth factor (Her)2. monodisperse PEG. tumour associated glycoprotein Tag72, MUC1 or prostate US 2012/0164068 A1 Jun. 28, 2012 specific membrane antigen (PSMA). Other proteins bind to a tein comprises a sequence at least about 80% or 90% or 95% plurality of antigens, e.g. the previously listed antigens, by or 96% or 97% or 98% or 99% or 100% identical to the virtue of cross-reactivity or the protein being multi-specific. sequence set forth in any one of SEQID NOs: 57,63, 65, 75, 0054 Examples of proteins of the invention comprise a 77, 79, 81, 83, 85, 87, 89,91, 93, 95, 97,99, 101, 103,105, sequence 80% or 90% or 95% or 96% or 97% or 98% or 99% 119, 121, 123, 125, 127, 129, 131 or 133, provided that the or 100% identical to the sequence set forth in any one of SEQ protein comprises a N-terminal threonine or serine residue. ID NOS:55, 59, 61, 109, 115 or 117, modified to include the 0069. The present invention also provides the present two or more positioned within FR1. Suitable sites for modi invention additionally provides a protein comprising a modi fication are described herein and are to be taken to apply fied immunoglobulin variable region comprising at least two mutatis mutandis to this example of the invention. For cysteine residues positioned within framework region (FR) 1 example, the protein comprises a sequence at least about 80% and/or a N-terminal threonine or serine residue, and wherein or 90% or 95% or 96% or 97% or 98% or 99% or 100% the unmodified form of the variable region does not comprise identical to the sequence set forth in set forth in any one of at least one of the cysteine residues (preferably at least two of SEQID NOs:57,63,65, 75,77, 79,81, 83,85, 87, 89,91, 93, or all of the cysteine residues) and/or the threonine or serine 95, 97,99, 101, 103 or 105,119, 121, 123,125, 127, 129, 131 residue. Suitable sites for positioning the cysteine residues or 133, provided that the protein comprises the cysteine resi and/or threonine or serine residue are described herein and are dues in FR1. to be taken to apply mutatis mutandis to the present example 0055 As discussed above, the inventors have also pro of invention. duced proteins comprising a N-terminal threonine or serine 0070. In one example, a protein of the invention is human, residue. This site is also useful for conjugation of a com humanized, deimmunized or chimeric. pound, even in the absence of cysteine residues in FR1. 0071. The present invention also provides a composition 0056. Accordingly, the present invention also provides an comprising a protein of the invention and a pharmaceutically isolated protein comprising an immunoglobulin heavy chain acceptable carrier. variable region (V) and an immunoglobulin light chain vari 0072 The present invention also encompasses an isolated able region (V), wherein at least one of the variable regions nucleic acid encoding a protein of the invention. Exemplary comprises a N-terminal threonine residue or serine residue. nucleic acids include those having a sequence at least about The serine or threonine residue may be added to the N-termi 80% or 90% or 95% or 96% or 97% or 98% or 99% or 100% nus of the protein. Preferably, the serine or threonine residue identical to the sequence set forth in any one or more of SEQ replaces a naturally occurring amino acid residue at the N-ter ID NOS: 54,58, 60, 108, 114 or 116 altered to include codons minus of the protein, i.e., is the result of a Substitutional encoding at least two cysteine residues in FR1 of the encoded mutation. protein and/or to include a N-terminal serine or threonine 0057 Preferred proteins comprise a Fv comprising at least residue. In one example, a nucleic acid of the invention com one protein of the invention comprising a N-terminal threo prises a sequence at least about 80% or 90% or 95% or 96% nine or serine residue, wherein at least one V, binds to at least or 97% or 98% or 99% or 100% identical to the sequence set one V, to form an antigen binding site. forth in any one or more of SEQID NOs: 56,62, 64, 74,76, 0058. In one example, the V, and the V, which form the 78, 80, 82, 84, 86, 88,90, 92,94, 96, 98, 100, 102, 104,118, antigen binding site are in a single polypeptide chain. For 120, 122, 124, 126, 128, 130, 132, provided that the sequence example, the protein is: encodes a protein comprising at least two cysteine residues in 0059 (i) a single chain Fv fragment (sclv); FR1 and/or a N-terminal serine or threonine residue. The 0060 (ii) a dimeric sclv (di-scEv); or skilled artisan will be aware that due to the degeneracy of 0061 (iii) at least one of (i) and/or (ii) linked to a Fc or a codon usage, numerous nucleotide sequences can encode a heavy chain constant domain (C)2 and/or C3. protein of the invention. All Such nucleotide sequences are 0062 Alternatively, the V, and the V, which form the encompassed by the present invention. For example, a codon antigenbinding site are in different polypeptide chains. In one optimized nucleic acid can be produced to facilitate expres example, each polypeptide chain in the protein comprises a sion in a specific cell type or organism. V, and a V. Preferably, Such a protein is: 0073. A nucleic acid of the invention can be operably 0063 (i) a diabody; linked to a promoter to thereby produce an expression con 0064 (ii) a triabody; or struct. Such an expression construct or the nucleic acid is 0065 (iii) a tetrabody. preferably included in a vector, preferably a vector replicable 0066. In another example, the protein of the present inven in a cell, e.g., a plasmid or phagemid or cosmid or artificial tion is an immunoglobulin, preferably an antibody. Exem chromosome. plary forms of immunoglobulins are described herein and are 0074 The present invention also provides an isolated cell to be taken to apply mutatis mutandis to the present example comprising an exogenous nucleic acid or expression con of the invention. struct of the invention, preferably wherein the cell expresses 0067. In one example, the protein additionally comprises a a protein of the invention. Exemplary cells include, but are not compound conjugated to the threonine or serine residue. limited to, bacterial cells, yeast cells, mammalian cells or Exemplary compounds are described herein and are to be insect cells. taken to apply mutatis mutandis to the present example of the 0075. The nucleic acids and/or expression constructs and/ invention. or cells provided by the invention also provide a basis for 0068. In one example, a protein comprises a sequence at methods for producing proteins of the invention. Accord least about 80% or 90% or 95% or 96% or 97% or 98% or 99% ingly, the present invention also provides a method for pro or 100% identical to the sequence set forth in any one of SEQ ducing a protein of the invention, the method comprising ID NOS:55, 59, 61, 109, 115 or 117, modified to include the maintaining an expression construct of the invention under N-terminal threonine or serine residue. For example, the pro conditions sufficient for the encoded protein to be produced. US 2012/0164068 A1 Jun. 28, 2012

For example, the method comprises culturing a cell of the the invention are described herein and shall be taken to apply invention under conditions sufficient the encoded for the pro mutatis mutandis to the present example of the invention. tein to be produced. In one example, the method additionally I0088. The present invention additionally provides a comprises isolating the protein. The method can additionally method for delivering a compound to a cell, the method comprise testing the protein, e.g., for binding activity or affin comprising contacting the cell with a protein of the invention ity. The method can additionally comprise formulating the that is conjugated to the compound or a composition com protein into a pharmaceutical composition. prising same. In one example, the cell is contacted by admin 0076. The present invention also provides a method for istering the protein or composition to a subject. producing a conjugate comprising a protein of the invention, I0089. The present invention also provides imaging meth the method comprising: ods, such as a method for localizing or detecting an antigen in 0077 (i) obtaining a protein of the invention comprising at a Subject, said method comprising: least two cysteine residues positioned within framework 0090 (i) administering to a subject a protein of the inven region (FR)1; and tion for a time and under conditions sufficient for the pro 0078 (ii) conjugating a compound to at least one of the tein to bind to the antigen, wherein the protein is conju cysteine residues to thereby produce the conjugate. gated to a detectable label; and 0079. In one example, the cysteine residues in the protein 0091 (ii) localizing or detecting the detectable label in obtained at (i) are linked by a disulphide bond and the method vivo. additionally comprises reducing or otherwise breaking the 0092. The skilled artisan will recognize that the foregoing disulphide bond prior to linking the compound to the cysteine method is useful for localizing or detecting cells, groups of residue(s). Preferably, reducing or otherwise breaking the cells Such as tumours, tissues and organs or parts thereof disulphide bond generates a free thiol group in the protein and expressing the antigen. Exemplary antigens are described the compound has a thiol reactive group. By reacting the throughout this specification and are to be taken to apply compound with the thiol reactive group, the conjugate is mutatis mutandis to the present example of the invention. produced. 0093. The present invention also provides a method for 0080. In one example, the compound is conjugated to the diagnosing or prognosing a condition in a subject, the method protein using a maleimide. For example, the protein is con comprising contacting a sample from the Subject with a pro tacted with a compound comprising a maleimide functional tein or composition of the invention for a time and under group Such that conjugation occurs. conditions sufficient for the protein to bind to an antigen and I0081. In a further example of the invention, the protein form a complex and detecting the complex, wherein detection additionally comprises at least one N-terminal serine or of the complex is diagnostic or prognostic of the condition. threonine residue and the method additionally comprises con Preferably, the protein is conjugated to a detectable label and jugating a compound to the serine or threonine residue. Pref detection of the label is indicative of the complex. erably, the compound conjugated to the serine or threonine 0094. In one example, the method comprises determining residue is different to the compound conjugated to the cys the level of the complex, wherein an enhanced or reduced teine residue(s). level of said complex compared to a control sample is diag 0082. The present invention provides an alternative nostic or prognostic of the condition. method for producing a conjugate comprising a protein of the 0.095 The present invention additionally provides a invention, the method comprising: library comprising a plurality of proteins of the invention. In 0083 (i) obtaining a protein of the invention comprising a one example, the proteins are displayed on the Surface of a N-terminal threonine or serine residue; and particle (e.g., a phage or a ribosome) or a cell. Clearly, the 0084 (ii) conjugating a compound to at least one serine or present invention also provides a library of nucleic acids threonine residue at the N-terminus of the protein to encoding said library comprising a plurality of proteins of the thereby produce the conjugate. invention. 0085 Optionally, a method of the invention for producing 0096. The present invention additionally provides a a conjugate additionally comprises isolating the conjugate method for isolating a protein of the invention, the method and/or formulating the conjugate into a pharmaceutical com comprising contacting a library of the invention with an anti position. gen for a time and under conditions Sufficient for (or Such I0086. It will be apparent to the skilled artisan based on the that) a protein binds to the antigen and isolating the protein. foregoing that the inventors have produced reagents that are useful in a variety of applications, including, delivery of a 0097. The present invention additionally provides a toxic compound or a radioisotope to a diseased cell, tissue or method for producing a library comprising a plurality of organ (e.g., a cancer) and/or in vivo imaging and/or for proteins of the invention, the method comprising: increasing the stability of a protein. (i) obtaining or producing nucleic acids encoding a plurality 0087. Accordingly, the present invention also provides for of proteins comprising an immunoglobulin variable region, use of a protein or a composition of the invention in medicine. wherein the variable regions comprising at least two cysteine For example, the present invention provides for use of a residues positioned within framework region (FR) 1 and/or a protein of the invention in the manufacture of a medicament N-terminal threonine or serine residue; for treating or preventing a condition. The present invention (ii) producing a library of expression constructs comprising also provides a method of treating or preventing a condition in the following operably linked nucleic acids: a Subject, the method comprising administering a protein or (0.098 a) a promoter: composition of the invention to a Subject in need thereof. 0099 b) a nucleic acid obtained or produced at (i); and Exemplary conditions are described herein and are to be taken 0.100 c) a nucleic acid encoding a polypeptide that to apply mutatis mutandis to the present example of the inven facilitates display of the variable region containing pro tion. Furthermore exemplary conjugated forms of a protein of tein in?on the cells or particles; and US 2012/0164068 A1 Jun. 28, 2012

(iii) expressing proteins encoded by the expression constructs 0117. In one example of the invention: Such that they are displayed in?on the cells or particles. 0118 (i) at least one of the polypeptides comprises a V 0101 Suitable sites for positioning the cysteine residues comprising complementarity determining regions (CDRS) and/or threonine or serine residue are described herein and are of a V comprising a sequence set forth in SEQID NO: to be taken to apply mutatis mutandis to the present example 111; and of invention. 0119 (ii) at least another of the polypeptides comprises a 0102. In one example, the amino acids in the CDRs of the V, comprising CDRS of a V comprising a sequence set protein are random or semi-random or are derived from a forth in SEQID NO: 113. human antibody. I0120 Exemplary CDRs are as follows: 0103) In one example, the method additionally comprises I0121 (i) CDRH1 comprises a sequence set forth in amino isolating nucleic acid encoding the protein. Such a nucleic acids 31 to 35 of SEQID NO: 111; acid can be introduced into an expression construct. Option 0.122 (ii) CDRH2 comprises a sequence set forth in amino ally, the protein can be expressed. acids 50 to 66 of SEQID NO: 111; 0104. The present inventors also produced a protein com I0123 (iii) CDRH3 comprises a sequence set forth in prising an immunoglobulin variable region capable of spe amino acids 99 to 104 of SEQID NO: 111; cifically binding to tumour antigen TAG72. The inventors 0.124 (iv) CDRL1 comprises a sequence set forth in amino found that this protein is stable in vivo. acids 24 to 40 of SEQID NO: 113: 0105. Accordingly, the present invention additionally pro 0.125 (V) CDRL2 comprises a sequence set forth in amino vides an isolated protein comprising a plurality of polypep acids 56 to 62 of SEQID NO: 113; and tides each comprising an immunoglobulin V, and an immu 0.126 (vi) CDRL3 comprises a sequence set forth in amino noglobulin V, linked by a region comprising an insufficient acids 95 to 103 of SEQID NO: 113. number of amino acids to permit the V, and the V, to asso I0127. In one example, an anti-TAG72 protein of the inven ciate with one another, wherein: tion comprises at least two polypeptides both of which com 0106 (i) at least one of the polypeptides comprises a V prise: comprising a sequence set forth in SEQ ID NO: 111 or a I0128 (i) a V comprising a sequence set forth in SEQID sequence at least about 60% identical thereto; and NO: 111 or a humanized or de-immunized form thereof; 0107 (ii) at least another of the polypeptides comprises a and V, comprising a sequence set forth in SEQID NO: 113 or I0129 (ii) a V, comprising a sequence set forth in SEQID a sequence at least about 60% identical thereto, NO: 113 or a humanized or de-immunized form thereof. 0108 wherein the V of the polypeptide at (i) and the V, of 0.130. An example of the invention provides an anti the polypeptide at (ii) associate to form a Fv capable of TAG72 protein comprising at least two polypeptides each specifically binding to tumourantigenTAG72. In the descrip comprising a sequence set forth in SEQID NO:55, 115 or tion herein, this protein is referred to as an anti-TAG72 pro 117. tein. However, any description herein in relation to a “protein I0131. In one example at least one of the polypeptides in the of the invention' equally applies to this protein unless the anti-TAG72 protein comprises at least two cysteine residues context indicates otherwise. positioned within framework region (FR) 1 as described herein-above. Exemplary sequences are set out above. For 0109. In one example, the anti-TAG72 protein comprises a example, the sequence is at least about 80% or 90% or 95% or region linking the V and the V, having 6 or fewer amino 96% or 97% or 98% or 99% or 100% identical to the sequence acids, for example 5 or fewer amino acids, such as 4 or fewer set forthin any one or more of SEQID NOS:57, 63,75, 77,99, amino acids, e.g., 3 or fewer amino acids. Such as 2 or fewer 103, 119, 121, 123, 125, 127, 129, 131 or 133 and comprises amino acids, for example, 1 or 0 amino acids. the cysteine residues. Preferably, the sequence is at least 0110. In an example, the anti-TAG72 protein of the inven about 80% or 90% or 95% or 96% or 97% or 98% or 99% or tion comprises a V, comprising a threonine at position 5 100% identical to the sequence set forth in SEQID NO: 57 numbered according to the Kabat numbering system and/or a and comprises the cysteine residues. threonine at position 53 according to the Kabat numbering 0.132. In one example, the protein comprises a single system and/or a glutamic acid at position 79 numbered polypeptide and comprises a sequence at least about 80% or according to the Kabat numbering system. 90% or 95% or 96% or 97% or 98% or 99% or 100% identical 0111. In an alternative or additional embodiment, the anti to the sequence set forth in SEQID NO: 101. TAG72 protein of the invention comprises a V comprising a I0133. In one example, an anti-TAG72 protein of the inven leucine at position 80 numbered according to the Kabat num tion comprises a compound conjugated thereto. For example, bering system. the compound is conjugated to a cysteine residue or a serine 0112. In one example, an anti-TAG72 protein comprises at residue or a lysine residue in at least one of the polypeptides least two polypeptides, each comprising: in the protein. 0113 (i) a V comprising a sequence set forth in SEQID I0134. In one example, at least one of the polypeptides in NO: 111 or a sequence at least about 60% identical thereto; the anti-TAG72 protein comprises at least two cysteine resi and dues positioned within FR1, wherein if at least two of the 0114 (ii) a V, comprising a sequence set forth in SEQID cysteine residues is not conjugated to a compound then a NO: 113 or a sequence at least about 60% identical thereto, disulphide bond is capable of forming between the cysteine 0115 wherein the V of one polypeptide and the V, of residues and/or a N-terminal serine and/or threonine residue, another polypeptide associate to form a Fv capable of spe and wherein the compound is conjugated to at least one of the cifically binding to TAG72. cysteine residues and/or to the serine residue. For example, 0116. An exemplary anti-TAG72 protein is a diabody, a the polypeptide comprises the cysteine residues and the triabody or a tetrabody. N-terminal serine and/or threonine residue, and wherein the US 2012/0164068 A1 Jun. 28, 2012 compound is conjugated to at least one of the cysteine resi 0145 The present invention additionally provides a dues and a different compound is conjugated to the serine method for producing an anti-TAG72 protein of the invention, residue. the method comprising maintaining an expression construct 0135 An exemplary compound is selected from the group encoding same such that the encoded polypeptide and for the consisting of a radioisotope, a detectable label, a therapeutic protein are produced. compound, a colloid, a toxin, a nucleic acid, a peptide, a 0146 In one example, the method comprises culturing the protein, a compound that increases the half life of the protein cell under conditions sufficient the encoded for the protein to in a Subject and mixtures thereof. be produced. 0136. The inventors have found that conjugation of poly 0.147. In one example, the method additionally comprises ethylene glycol (PEG) to a protein of the invention substan isolating the protein. tially increased its stability and tumour uptake. The inventors 0.148. The present invention additionally provides a found that monodisperse PEGs and/or PEG having small method for producing an anti-TAG72 protein of the invention molecular weights increased stability and/or tumour uptake comprising a compound conjugated to at least one of the to Substantially the same degree as higher molecular weight cysteine residue, the method comprising: PEGs. This appears counter-intuitive since larger molecules 0149 (i) obtaining an anti-TAG72 protein of the invention are generally cleared from a subject slower than Smaller mol comprising cysteine residues in FR1; and ecules. Since larger PEGs generally comprise mixtures of 0150 (ii) conjugating a compound to at least one of the molecules of various molecular weights, the use of Smaller cysteine residues in the FR1 of the polypeptide(s) to and/or monodisperse PEGs provides an advantage in so far as thereby produce the protein. it facilitates production of protein conjugates similar at the 0151. In one example, the cysteine residues in the molecular level. Such conjugates are desirable for in vivo polypeptide(s) in the protein obtained at (i) are linked by a applications. disulphide bond and the method additionally comprises 0.137 Accordingly, in one example, an anti-TAG72 pro reducing or otherwise breaking the disulphide bond prior to tein of the invention is conjugated to polyethylene glycol conjugating the compound to the cysteine residue(s). For (PEG). For example, the PEG is monodisperse PEG. example, reducing or otherwise breaking the disulphide bond 0.138. In one example, the PEG has a molecular weight no generates a free thiol group in the protein and the compound greater than about 4000 Da, for example, a molecular weight has a thiol reactive group permitting conjugation of the com no greater than about 2000 Da, such as a molecular weight no pound to the protein. greater than about 1,500 Da. In one example, the PEG has a 0152. In one example, at least one polypeptide in the pro molecular weight no greater than 1,000 Da. Such as, no tein comprises at least one N-terminal serine or threonine greater than 900 Da, for example, no greater than 800 Da. residue and the method additionally comprises conjugating a such as, no greater than 600 Da. In one example, the PEG has compound to the serine or threonine residue. a molecular weight from about 550 Da to about 1,000 Da. 0153. In another example, the present invention provides a 0.139. In another example, the PEG has no more than about method for producing an anti-TAG72 protein of the invention 70 or 75 or 77 ethylene glycol units. For example, the PEG has comprising a compound conjugated to a N-terminal serine or no more than about 50 ethylene glycol units. Preferably, the threonine residue, the method comprising: PEG has no more than 48 ethylene glycole units. For 0154 (i) obtaining an anti-TAG72 protein of the invention example, the PEG has no more than about 40 ethylene glycol comprising a N-terminal serine and/or threonine residue; units. For example, the PEG has no more than about 30 and ethylene glycol units. For example, the PEG has no more than O155 (ii) conjugating a compound to at least one serine or about 27 ethylene glycol units. For example, the PEG has no threonine residue at the N-terminus of the polypeptide to more than about 24 ethylene glycol units. For example, the thereby produce the protein. PEG has no more than about 15 ethylene glycol units. For 0156. In one example, a method for producing a protein example, the PEG has no more than about 12 ethylene glycol conjugated to a compound comprises conjugating the protein units. Preferably, the PEG comprises about 12 to 27 ethylene to PEG. glycol units. 0157. The present invention additionally provides a 0140. In one example, the PEG is conjugated to an addi method for localizing and/or detecting and/or diagnosing tional compound. Such as a chelating agent, e.g., a macrocy and/or prognosing a cancer in a Subject, said method com clic chelating agent, such as 1,4,7,10-tetraazacyclododecane prising: 1,4,7,10-tetraacetic acid (DOTA). (i) administering to a subject an anti-TAG72 protein of the 0141. The present invention also provides an anti-TAG72 invention or composition comprising same such that it binds protein of the invention and a pharmaceutically acceptable to tumour antigen TAG72, if present, and carrier. (ii) detecting the proteinbound to the TAG72 in vivo, wherein 0142. The present invention additionally provides an iso detection of the bound protein localizes and/or detects and/or lated nucleic acid encoding one or more of the polypeptides in diagnoses and/or prognoses the cancer. an anti-TAG72 protein of the invention. 0158. The present invention additionally provides a 0143. The present invention further provides an expres method of diagnosing or prognosing a cancer in a Subject, the sion construct comprising a nucleic acid encoding an anti method comprising: TAG72 protein of the invention or a polypeptide therefrom. (i) contacting a sample from the subject with an anti-TAG72 0144. The present invention additionally provides an iso protein of the invention or a composition comprising same lated cell expressing an anti-TAG72 protein of the invention. such that it binds to tumour antigen TAG72, if present; and For example, the cell comprises a nucleic acid encoding an (ii) detecting the protein bound to the TAG72, wherein detec anti-TAG72 protein of the invention or polypeptide thereform tion of the bound protein is diagnostic or prognostic of the and/or expression construct comprising same. CaCC. US 2012/0164068 A1 Jun. 28, 2012

0159. In one example of either of the previous two error bars show standard deviation. H=heavy chain. L-light examples, the protein is conjugated to a detectable label and chain. Numbers indicate positions of cysteine residues (if the method comprises detecting the label to detect the protein present), H7-H10=AVP04-84, L8-L11=AVP04-50, bound to TAG72. HB-H16=AVP04-85, L14-L17=AVP04-78. 0160 The present invention additionally provides a 0169 FIG. 6A is a diagrammatic representation showing method of treating a cancer, the method comprising admin the in silico homology modeled intra-Framework 1 disul istering an anti-TAG.72 protein of the invention or a compo phide insertion mutations in the V, and V of the AVP02-60 sitions comprising same Such that it binds to tumour antigen diabody (comprising a polypeptide comprising a sequence set TAG72 on cancer cells and treats the cancer. forth in SEQID NO: 61). Heavy chain shown in black, light 0161 In one example, the protein or a compound conju chain shown in grey, in silico mutated disulphide insertion gated thereto induces death of the cancer cells. mutant side chains shown as ball and stick. Left hand side: 0162. In one example, the protein is conjugated to a com aligned FVs from un-mutated and two disulphide insertion pound induces death of the cancer cells. mutant diabody models. Middle: as for left hand side but only showing FR1. Right hand side as for middle but rotated on the BRIEF DESCRIPTION OF THE DRAWINGS horizontal axis by 100°. 0163 FIG. 1 is a diagrammatic representation showing the 0170 FIG. 6B is a diagrammatic representation showing in silico homology modeled, un-mutated AVP04-07 diabody the in silico homology modeled intra-Framework 1 disul (comprising a polypeptide comprising a sequence set forth in phide insertion mutations in the V, and V of the AVP02-60 SEQ ID NO: 57). Framework residues are shown in grey, diabody (comprising a polypeptide comprising a sequence set CDR residues shown in black, potential disulphide insertion forth in SEQID NO: 61). Depicted are models of FR1 com residues identified for mutation are shown in white. Arrows prising various mutations as indicated. H heavy chain. identify residues in one of the Fvs with Kabat numbering. L=light chain. Numbers indicate positions of cysteine resi 0164 FIG. 2 is a diagrammatic representation showing the dues (if present). Un-mutated FR1 H/Un-mutated FR1 in silico homology modeled, un-mutated AVP07-17 diabody L=AVP02-60, H7-H10=AVP02-104, L8-L11=AVPO2-101, (comprising a polypeptide comprising a sequence set forth in HB-H16=AVP02-105, L14-L17=AVP02-102. SEQ ID NO: 59). Framework residues are shown in grey, 0171 FIG. 7A is a diagrammatic representation showing CDR residues shown in black, potential disulphide insertion the in silico homology modeled intra-Framework 1 disul residues identified for mutation are shown in white. Arrows phide insertion mutations in the V, and V of the AVP07-17 identify residues in one of the Fvs with Kabat numbering. diabody (comprising a polypeptide comprising a sequence set 0.165 FIG.3 is a diagrammatic representation showing the forth in SEQID NO. 59). Heavy chain shown in black, light in silico homology modeled, un-mutated AVP02-60 diabody chain shown in grey, in silico mutated disulphide insertion (comprising a polypeptide comprising a sequence set forth in mutant side chains shown as ball and stick. Left hand side: SEQ ID NO: 61). Framework residues are shown in grey, aligned FVs from un-mutated and two disulphide insertion CDR residues shown in black, potential disulphide insertion mutant diabody models. Middle: as for left hand side but only residues identified for mutation are shown in white. Arrows showing FR1. Right hand side as for middle but rotated on the identify residues in one of the Fvs with Kabat numbering. horizontal axis by 100°. 0166 FIG. 4A is a diagrammatic representation showing 0172 FIG. 7B is a diagrammatic representation showing the in silico homology modeled intra-Framework 1 disul the in silico homology modeled intra-Framework 1 disul phide insertion mutations in the V, and V of the AVP04-07 phide insertion mutations in the V, and V of the AVP07-17 diabody (comprising a polypeptide comprising a sequence set diabody (comprising a polypeptide comprising a sequence set forth in SEQID NO:57). Heavy chain is shown in black, light forth in SEQID NO: 59). Depicted are models of FR1 com chain shown in grey, in silico mutated disulphide insertion prising various mutations as indicated. H heavy chain. mutant side chains shown as ball and stick. Left hand side: L=light chain. Numbers indicate positions of cysteine resi aligned FVs from un-mutated and two disulphide insertion dues (if present). Un-mutated FR1 H/Un-mutated FR1 mutant diabody models. Middle: as for left hand side but only L=AVP07-17, H7-H10=AVP07-90, L7-L11=AVP07-88, showing FR1. Right hand side as for middle but rotated on the HB-H16=AVP07-91, L14-L17=AVP07-89. horizontal axis by 100°. 0173 FIG. 8 is a graph of the average per residue solvent 0167 FIG. 4B is a diagrammatic representation showing ASAs for the CDR groups and disulphides of the AVP02-XX the in silico homology modeled intra-Framework 1 disul Avibodies, V derived columns are shown in grey, V, derived phide insertion mutations in the V, and V of the AVP04-07 columns are shown in white. Shown as average per residue diabody (comprising a polypeptide comprising a sequence set ASA for residues in CDR groups, conserved inter-sheet dis forth in SEQID NO:57). Depicted are models of FR1 com ulphide residues and disulphide insertion mutation residues, prising various mutations as indicated. H heavy chain. error bars show standard deviation. H=heavy chain. L-light L=light chain. Numbers indicate positions of cysteine resi chain. Numbers indicate positions of cysteine residues (if dues (if present). Un-mutated FR1 H/Un-mutated FR1 present), H7-H10=AVP02-104, L8-L11=AVP02-101, L=AVP04-07, H7-H10=AVP04-84, L8-L11=AVP04-50, HB-H16=AVP02-105, L14-L17=AVP02-102. HB-H16=AVP04-85, L14-L17=AVP04-78. 0.174 FIG. 9 is a graph of the average per residue solvent 0168 FIG. 5 is a graph of the average per residue solvent ASAs for the CDR groups and disulphides of the AVP07-XX accessible surface areas (ASA) for the CDR groups and dis Avibodies, V derived columns are shown in grey, V, derived ulphides of the AVP04-XX Avibodies, V, derived columns are columns are shown in white. Shown as average per residue shown in grey, V, derived columns are shown in white. ASA for residues in CDR groups, conserved inter-sheet dis Shown as average per residue ASA for residues in CDR ulphide residues and disulphide insertion mutation residues, groups, conserved inter sheet disulphide residues (H22-H92 error bars show standard deviation. H=heavy chain. L-light and L23-L88) and disulphide insertion mutation residues, chain. Numbers indicate positions of cysteine residues (if US 2012/0164068 A1 Jun. 28, 2012

present), H7-H10=AVP07-90, L7-L11=AVP07-88, 0185 FIG. 14A includes a graphical representation of the HB-H16=AVP07-91, L14-L17=AVP07-89. atomic mass unit (AMU) of anti-HER2 scEv AVP07-71 (SEQ 0175 FIG. 10A is a graphical representation showing the ID NO: 105) showing a mass proportional to the protein plus 280 nm chromatograph of AVP04-50 (comprising polypep the addition of two molecules of PEG following conjuga tide comprising a sequence set forth in SEQ ID NO: 57) tion (30078.36 amu). His-Tag affinity chromatography purification. Arrow indi 0186 FIG. 14B includes a graphical representation of the cates elution peak of interest. Dotted line indicates proportion AMU of anti-TAG72 AVP04-50 (SEQID NO:57) showing a of elution buffer. mass proportional to the protein plus the addition of two 0176 FIG. 10B is a graphical representation showing molecules of PEG following conjugation (28166.84 amu) results of cation purification of AVP04-50 (comprising per monomer-chain. polypeptide comprising a sequence set forth in SEQID NO: 0187 FIG. 14C includes a graphical representation of the 57). Arrow indicates elution peak of interest. Dotted line AMU of anti-HER2 diabody AVP07-88 (SEQ ID NO: 87) a indicates proportion of elution buffer. mass proportional to the protein plus the addition of two 0177 FIG. 10C is a graphical representation showing molecules of PEG following conjugation (29196.72 amu) results of size exclusion chromatography of AVP04-50 (com per monomer-chain. prising polypeptide comprising a sequence set forth in SEQ 0188 FIG. 14D includes a graphical representation of the ID NO: 57). Arrow indicates elution peak of interest. Dotted AMU of anti-MUC1 diabody AVP02-101 (SEQID NO: 79) lines outline fractions of interest. showing one molecule of PEG conjugated (24749.05 amu) 0.178 FIG. 10D is a graphical representation showing per monomer-chain. results of post-purification size exclusion chromatography of 0189 FIGS. 15A-C include graphical representations of a AVP04-50 (comprising polypeptide comprising a sequence column shift assay used to determine immunoreactivity of set forth in SEQID NO:57). Arrow indicates elution peak of PEGylated Avibodies mentioned herein (as indicated, interest. nomenclature corresponds to that used throughout the text 0179 FIG.10E is a copy of a photographic representation and in the sequence listing). Each graph comprises two over showing results of a reducing SDS-PAGE gel showing the laid size exclusion chromatography profiles; of the Avibody purity of AVP04-50 (comprising polypeptide comprising a PEG conjugate either in the presence (solid line) or absence sequence set forth in SEQID NO: 57) post purification. Lane (dotted line) of antigen. 1: Invitrogen Benchmark Pre-stained molecular weight 0.190 FIG. 16 includes a graphical representation of a marker, Lane 2: AVP04-50 protein band. column shift assay used to determine immunoreactivity of 0180 FIGS. 11A-C include graphical representations of europium-conjugated AVP04-50 (SEQID NO:57) Avibody. the purified Avibodies mentioned herein (as indicated, The representation comprises two overlaid size exclusion nomenclature corresponds to that used throughout the text chromatography profiles; of the Avibody-Europium conju and in the sequence listing) following size exclusion chroma gate either in the presence (dotted line) or absence (solid line) tography. of antigen. 0181 FIGS. 12A-C include graphical representations of a 0191 FIG. 17 includes a graphical representation of column shift assay used to determine immunoreactivity of immunoreactivity of europium-conjugated AVP04-50 (SEQ Avibodies mentioned herein (as indicated, nomenclature cor ID NO. 57) as determined by cell binding assay on antigen responds to that used throughout the text and in the sequence positive (LS174T, solid line) and negative (SK-OV-3, dotted listing). Each graph comprises two overlaid size exclusion line) cell lines. chromatography profiles; of the Avibody incubated either in 0.192 FIG. 18A is a graph showing the biodistribution of the presence (Solid line) or absence (dotted line) of antigen. 'I and '''In-labelled AVP04-07 (comprising polypeptide 0182 FIG. 13A includes a copy of a photographic repre comprising a sequence set forth in SEQID NO: 55) in estab sentation showing results of Avibody conjugation with mal lished LS-174TXenograft mice. Open arrow–Tumour uptake PEG-MeOH indicating an increase in molecular weight curve, closed arrow-kidney uptake curve, circle arrow-blood with respect to unpegylated control (“naked') Avibody by clearance curve. SDS-PAGE. Lane 1 and 10: Benchmark pre-stained molecu 0193 FIG. 18B is a graph showing the biodistribution of lar weight standard, Lane 2 and 3: Naked AVP07-71 and 'I and '''In labelled AVP04-50 (comprising polypeptide AVP07-71-PEG respectively. Lane 4 and 5: Naked AVP07 comprising a sequence set forth in SEQID NO: 57) in estab 88 and AVP07-88-PEG respectively. Lane 6 and 7: Naked lished LS-174TXenograft mice. Open arrow–Tumour uptake AVP02-101 and AVP02-101-PEG respectively. Lane 8 and curve, closed arrow-kidney uptake curve, circle arrow-blood 9: Naked AVP04-50 and AVP04-50-PEG respectively. clearance curve. 0183 FIG. 13B includes a copy of a photographic repre 0194 FIG. 19 is a graph showing the biodistribution of sentation showing results of Avibody conjugation with mal 'I labelled AVP07-17 (comprising polypeptide comprising PEG-MeOH indicating an increase in molecular weight a sequence set forth in SEQ ID NO: 59) and AVP07-63 with respect to unpegylated control (“naked') Avibody by (comprising polypeptide comprising a sequence set forth in SDS-PAGE. Lane 1: Benchmark pre-stained molecular SEQID NO: 65) in established SKOV3 xenograft mice. Dark weight standard, Lane 2 and 3: Naked AVP04-70 and AVP04 grey bars=AVP07-63, open bars=AVP07-17. 70-PEG respectively. Lane 4 and 5: Naked AVP04-84 and 0.195 FIG. 20 is a graph showing the biodistribution of AVP04-84-PEG respectively. Panel A: '''In-labelled AVP04-07 (comprising polypeptide 0184 FIG. 13C includes a copy of a photographic repre comprising a sequence set forth in SEQID NO: 55) conju sentation showing results of Avibody conjugation with mal gated with PEG3400 to random surface lysines and Panel B: PEG-MeOH by SDS-PAGE. Lane 1: Benchmark pre '''In-labelled AVP04-50 (comprising polypeptide compris stained molecular weight standard, Lane 2: AVP04-74 ing a sequence set forth in SEQID NO: 57) conjugated with PEG conjugate. Lane 3: AVP04-78-PEG conjugate. size-monodispersed PEGs conjugated to engineered intra US 2012/0164068 A1 Jun. 28, 2012

Framework 1 disulphide mutations in established LS-174T 0206 FIG. 26C is a graphical representation showing bio Xenograft mice. Open arrow–Tumour uptake curve, closed distribution of '''In labelled PEG conjugated AVP04-07 arrow-kidney uptake curve, circle arrow-blood clearance diabody in nude mice bearing LS174T xenografts. CUV. 0207 FIG. 26D is a graphical representation showing bio 0.196 FIGS. 21A and B are graphical representations distribution of I-labelled PEG conjugate conjugated showing the PET biodistribution of Cu-labelled, unpegy AVP04-07 diabody in nude mice bearing LS174TXenografts. lated AVP04-07 (comprising polypeptide comprising a 0208 FIG.27A is a graphical representation showing PET sequence set forth in SEQID NO: 55) AVP04-07 conjugated images of a time course of imaging for a single mouse with size-monodispersed PEG, to random surface lysines injected with Cu-labelled intact AVP04-07 diabody at 1 h,4 and AVP04-50 (comprising polypeptide comprising a h, 2 hand 46 h post-injection. sequence set forth in SEQID NO: 57) conjugated with size (0209 FIG. 27B is a graphical representation showing PET monodispersed PEGs conjugated to engineered intra images of a time course of imaging for a single mouse Framework 1 disulphide mutations in established LS-174T injected with Cu-labelled PEG, AVP04-07 diabody. Xenograft mice. Upward facing arrows location of tumour, 0210 FIG.27C is a graphical representation showing PET Downward facing arrows-location of kidney. images of a time course of imaging for a single mouse 0.197 FIG.22A is a graphical representation showing bio injected with Cu-labelled PEG, AVP04-07 diabody. distribution of '''In-labelled intact AVP04-07 diabody in nude mice bearing LS174T xenografts. Open arrow–Tumour KEY TO SEQUENCE LISTING uptake curve, closed arrow-kidney uptake curve, circle 0211 SEQID NO: 1-amino acid sequence of FR1 of a arrow-blood clearance curve. human antibody heavy chain; 0198 FIG.22B is a graphical representation showing bio SEQ ID NO: 2-amino acid sequence of FR1 of a human distribution of 'I-labelled intact AVP04-07 diabody in nude antibody heavy chain; mice bearing LS174T xenografts. Open arrow–Tumour SEQ ID NO:3—amino acid sequence of FR1 of a human uptake curve, closed arrow-kidney uptake curve, circle antibody heavy chain; arrow-blood clearance curve. SEQ ID NO. 4 amino acid sequence of FR1 of a human 0199 FIG.22C is a graphical representation showing bio antibody heavy chain; distribution of '''In-labelled AVP04-07 diabody conjugated SEQ ID NO: 5-amino acid sequence of FR1 of a human to PEG3400 in nude mice bearing LS174T xenografts. Open antibody heavy chain; arrow–Tumour uptake curve, closed arrow-kidney uptake SEQ ID NO: 6-amino acid sequence of FR1 of a human curve, circle arrow-blood clearance curve. antibody heavy chain; 0200 FIG. 23A is a copy of a photographic representation SEQ ID NO: 7 amino acid sequence of FR1 of a human showing results of isoelectric focusing of 1, intact AVP04-07 antibody heavy chain; diabody: 2, AVP04-07-PEG, conjugate; 3, AVP04-07 SEQ ID NO: 8 amino acid sequence of FR1 of a human PEG conjugate (20:1 ratio): 4, AVP04-07-PEG conjugate antibody heavy chain; (50:1 ratio); 5, AVP04-07-PEG3400 conjugate. SEQ ID NO: 9 amino acid sequence of FR1 of a human 0201 FIG. 23B is a copy of a photographic representation antibody heavy chain; showing results of SDS gel electrophoresis of 1, intact SEQ ID NO: 10 amino acid sequence of FR1 of a human AVP04-07 diabody: 2, AVP04-07-PEG, conjugate; 3, antibody heavy chain; AVP04-07-PEG conjugate (20:1 ratio): 4, AVP04-07 SEQ ID NO: 11—amino acid sequence of FR1 of a human PEG conjugate (50:1 ratio); 5, AVP04-07-PEG3400 conju antibody heavy chain; gate. SEQ ID NO: 12-amino acid sequence of FR1 of a human 0202 FIG. 24 is a graphical representation showing antibody heavy chain; results of size exclusion chromatography of ''In-radiola SEQ ID NO: 13—amino acid sequence of FR1 of a human belled intact AVP04-07 diabody and PEG conjugates (radio antibody heavy chain; activity shown as arbitrary units). Peaks from Right to Left: SEQ ID NO: 14—amino acid sequence of FR1 of a human labelled no. 1: DOTA-intact diabody, no. 2: DOTA-PEG antibody K light chain; Cys-VS-diabody conjugate, no. 3: DOTA-PEG7-Cys-VS SEQ ID NO: 15-amino acid sequence of FR1 of a human diabody conjugate, no. 4: DOTA-Cys-VS-PEG3400-diabody antibody K light chain; conjugate. SEQ ID NO: 16—amino acid sequence of FR1 of a human 0203 FIG. 25 is a diagrammatic representation of the antibody K light chain; scheme for synthesizing DOTA-PEG-Cys-VS. FMOC SEQ ID NO: 17—amino acid sequence of FR1 of a human amido-PEG-acid was conjugated to S-t-butyl cysteine on antibody K light chain; Wang resin using standard activation chemistry (DCC/ SEQ ID NO: 18-amino acid sequence of FR1 of a human HOBt). The FMOC was removed with piperidine and conju antibody K light chain; gated to DO3AtEuAc using standard activation chemistry SEQ ID NO: 19—amino acid sequence of FR1 of a human (DCC/HOBt). The product was removed from the resin with antibody K light chain; TFA, purified by reverse phase HPLC, reacted with excess SEQ ID NO: 20 amino acid sequence of FR1 of a human vinyl sulfone in DMF, and repurified by reverse phase HPLC. antibody K light chain; 0204 FIG. 26A is a graphical representation showing bio SEQ ID NO: 21—amino acid sequence of FR1 of a human distribution of '''In-labelled PEG, conjugated AVP04-07 antibody K light chain; diabody in nude mice bearing LS174T xenografts. SEQ ID NO: 22—amino acid sequence of FR1 of a human 0205 FIG. 26B is a graphical representation showing bio antibody K light chain; distribution of 'I-labelled PEG, conjugated AVP04-07 SEQ ID NO: 23-amino acid sequence of FR1 of a human diabody in nude mice bearing LS174T xenografts. antibody K light chain; US 2012/0164068 A1 Jun. 28, 2012

SEQ ID NO: 24 amino acid sequence of FR1 of a human SEQ ID NO: 55-amino acid sequence of AVP04-07 anti antibody K light chain; TAG72 diabody; SEQ ID NO. 25—amino acid sequence of FR1 of a human SEQ ID NO. 56—nucleotide sequence encoding modified antibody K light chain; AVP04-07 anti-TAG72 diabody designated AVP04-50 com SEQ ID NO: 26—amino acid sequence of FR1 of a human prising cysteine residues in FR1 and a N-terminal serine; antibody K light chain; SEQ ID NO: 57 amino acid sequence of modified AVP04 SEQ ID NO: 27 amino acid sequence of FR1 of a human 07 anti-TAG72 diabody designated AVP04-50 comprising antibody K light chain; cysteine residues in FR1 and a N-terminal serine; SEQ ID NO: 28—amino acid sequence of FR1 of a human SEQ ID NO: 58 nucleotide sequence encoding AVP07-17 antibody K light chain; anti-Her2 diabody; SEQ ID NO: 29 amino acid sequence of FR1 of a human SEQ ID NO: 59-amino acid sequence of AVP07-17 anti antibody K light chain; Her2 diabody; SEQ ID NO:30 amino acid sequence of FR1 of a human SEQ ID NO: 60 nucleotide sequence encoding AVP02-60 antibody K light chain; anti-MUC1 diabody; SEQ ID NO: 31—amino acid sequence of FR1 of a human SEQ ID NO: 61-amino acid sequence of AVP02-60 anti antibody K light chain; MUC1 diabody SEQ ID NO: 62 nucleotide sequence SEQ ID NO: 32—amino acid sequence of FR1 of a human encoding modified AVP04-07 anti-TAG72 diabody desig antibody K light chain; nated AVP04-84 comprising cysteine residues in FR1 and a N-terminal serine; SEQ ID NO: 33—amino acid sequence of FR1 of a human SEQID NO: 63—amino acid sequence of modified AVP04 antibody K light chain; 07 anti-TAG72 diabody designated AVP04-84 comprising SEQ ID NO. 34 amino acid sequence of FR1 of a human cysteine residues in FR1 and a N-terminal serine; antibody K light chain; SEQ ID NO: 64 nucleotide sequence encoding modified SEQ ID NO: 35—amino acid sequence of FR1 of a human AVP07-17 anti-Her2 diabody designated AVP07-63 compris antibody K light chain; ing cysteine residues in FR1, removal of cysteine residues in SEQ ID NO: 36—amino acid sequence of FR1 of a human CDR3H and a N-terminal serine; antibody K light chain; SEQ ID NO: 65-amino acid sequence of modified AVP07 SEQ ID NO: 36—amino acid sequence of FR1 of a human 17 anti-Her2 diabody designated AVP07-63 comprising cys antibody K light chain; teine residues in FR1, removal of cysteine residues in CDR3H SEQ ID NO:37 -amino acid sequence of FR1 of a human and a N-terminal serine; antibody w light chain; SEQID NO: 66 nucleotide sequence of mutagenic primer SEQ ID NO: 38—amino acid sequence of FR1 of a human for substituting the N-terminal Glin residue with a Ser residue antibody w light chain; in AVP04-07: SEQ ID NO: 39-amino acid sequence of FR1 of a human SEQID NO: 67 nucleotide sequence of mutagenic primer antibody w light chain; for substituting the N-terminal Glin residue with a Ser residue SEQ ID NO: 40 amino acid sequence of FR1 of a human in AVP04-07: antibody w light chain; SEQID NO: 68 nucleotide sequence of mutagenic primer SEQ ID NO: 41-amino acid sequence of FR1 of a human for replacing cysteine for alanines residues into AVP07-17: antibody w light chain; SEQID NO: 69 nucleotide sequence of mutagenic primer SEQ ID NO: 42-amino acid sequence of FR1 of a human for replacing cysteine for alanines residues into AVP07-17: antibody w light chain; SEQID NO: 70 amino acid sequence of human Her2: SEQ ID NO: 43—amino acid sequence of FR1 of a human SEQID NO: 71—amino acid sequence of human PSMA: antibody w light chain; SEQ ID NO: 72-amino acid sequence of an isoform of SEQID NO: 44 amino acid sequence of FR1 of a camelid human MUC1; and immunoglobulin; SEQ ID NO: 73—amino acid sequence of an isoform of SEQID NO: 45—amino acid sequence of FR1 of a camelid human MUC1 expressed in several forms of cancer. immunoglobulin; SEQ ID NO: 74 nucleotide sequence encoding modified SEQID NO: 46—amino acid sequence of FR1 of a camelid AVP04-07 anti-TAG72 diabody designated AVP04-85 com immunoglobulin; prising cysteine residues in FR1 and a N-terminal serine; SEQID NO: 47 amino acid sequence of FR1 of a camelid SEQ ID NO: 75-amino acid sequence of modified AVP04 immunoglobulin; 07 anti-TAG72 diabody designated AVP04-85 comprising SEQID NO: 48 amino acid sequence of FR1 of a camelid cysteine residues in FR1 and a N-terminal serine; immunoglobulin; SEQ ID NO: 76 nucleotide sequence encoding modified SEQID NO: 49 amino acid sequence of FR1 of a camelid AVP04-07 anti-TAG72 diabody designated AVP04-78 com immunoglobulin; prising cysteine residues in FR1 and a N-terminal serine; SEQID NO: 50 amino acid sequence of FR1 of a camelid SEQ ID NO: 77 amino acid sequence of modified AVP04 immunoglobulin; 07 anti-TAG72 diabody designated AVP04-78 comprising SEQ ID NO: 51—amino acid sequence of FR1 of a spiny cysteine residues in FR1 and a N-terminal serine; dogfish shark IgNAR; SEQ ID NO: 78 nucleotide sequence encoding modified SEQ ID NO: 52—amino acid sequence of FR1 of a nurse AVP02-60 anti-MUC1 diabody designated AVP02-101 com shark IgNAR; prising cysteine residues in FR1 and a N-terminal serine; SEQID NO: 53—amino acid sequence of a linker; SEQ ID NO: 79-amino acid sequence of modified AVP02 SEQ ID NO: 54 nucleotide sequence encoding AVP04-07 60 anti-MUC1 diabody designated AVP02-101 comprising anti-TAG72 diabody; cysteine residues in FR1 and a N-terminal serine; US 2012/0164068 A1 Jun. 28, 2012

SEQ ID NO: 80 nucleotide sequence encoding modified SEQ ID NO: 98 nucleotide sequence encoding modified AVP02-60 anti-MUC1 diabody designated AVP02-104 com AVP04-07 anti-TAG72 diabody designated AVP04-51 com prising cysteine residues in FR1 and a N-terminal serine; prising cysteine residues in FR1 and a N-terminal serine; SEQ ID NO: 81-amino acid sequence of modified AVP02 SEQ ID NO:99-amino acid sequence of modified AVP04 60 anti-MUC1 diabody designated AVP02-104 comprising 07 anti-TAG72 diabody designated AVP04-51 comprising cysteine residues in FR1 and a N-terminal serine; cysteine residues in FR1 and a N-terminal serine; SEQ ID NO: 82 nucleotide sequence encoding modified SEQ ID NO: 100 nucleotide sequence encoding modified AVP02-60 anti-MUC1 diabody designated AVP02-102 com AVP04-07 anti-TAG72 scFv designated AVP04-70 compris prising cysteine residues in FR1 and a N-terminal serine; ing cysteine residues in FR1 and a N-terminal serine; SEQ ID NO: 83-amino acid sequence of modified AVP02 SEQID NO: 101-amino acid sequence of modified AVP04 60 anti-MUC1 diabody designated AVP02-102 comprising 07 anti-TAG72 scFv designated AVP04-70 comprising cys cysteine residues in FR1 and a N-terminal serine; teine residues in FR1 and a N-terminal serine; SEQ ID NO: 84 nucleotide sequence encoding modified SEQ ID NO: 102 nucleotide sequence encoding modified AVP02-60 anti-MUC1 diabody designated AVP02-105 com AVP04-07 anti-TAG72 triabody designated AVP04-74 com prising cysteine residues in FR1 and a N-terminal serine; prising cysteine residues in FR1 and a N-terminal serine; SEQ ID NO: 85-amino acid sequence of modified AVP02 60 anti-MUC1 diabody designated AVP02-105 comprising SEQID NO: 103-amino acid sequence of modified AVP04 cysteine residues in FR1 and a N-terminal serine; 07 anti-TAG72 triabody designated AVP04-74 comprising SEQ ID NO: 86 nucleotide sequence encoding modified cysteine residues in FR1 and a N-terminal serine; AVP07-17 anti-HER2 diabody designated AVP07-88 com SEQ ID NO: 104 nucleotide sequence encoding modified prising cysteine residues in FR1, removal of cysteine residues AVP07-17 anti-HER2 schv designated AVP07-71 compris in CDR3H and a N-terminal serine; ing cysteine residues in FR1, removal of cysteine residues in SEQ ID NO: 87 amino acid sequence of modified AVP07 CDR3H and a N-terminal serine; 17 anti-HER2 diabody designated AVP07-88 comprising SEQID NO: 105-amino acid sequence of modified AVP07 cysteine residues in FR1, removal of cysteine residues in 17 anti-HER2 schv designated AVP07-71 comprising cys CDR3H and a N-terminal serine; teine residues in FR1, removal of cysteine residues in CDR3H SEQ ID NO: 88 nucleotide sequence encoding modified and a N-terminal serine; AVP07-17 anti-HER2 diabody designated AVP07-90 com SEQID NO: 106 nucleotide sequence of mutagenic primer prising cysteine residues in FR1, removal of cysteine residues for introducing cysteine residues at Kabat positions L8 and in CDR3H and a N-terminal serine; L11 of the FR1 region of the VL chain in AVP04-07: SEQ ID NO: 89-amino acid sequence of modified AVP07 SEQID NO: 107 nucleotide sequence of mutagenic primer 17 anti-HER2 diabody designated AVP07-90 comprising for introducing cysteine residues at Kabat positions L8 and cysteine residues in FR1, removal of cysteine residues in L11 of the FR1 region of the VL chain in AVP04-07: CDR3H and a N-terminal serine; SEQ ID NO: 108 nucleotide sequence encoding modified SEQ ID NO: 90 nucleotide sequence encoding modified AVP07-17 anti-HER2 diabody designated AVP07-86 replac AVP07-17 anti-HER2 diabody designated AVP07-89 com ing CDR3H Cysteine residues Cys 104 (Kabat numbering prising cysteine residues in FR1, removal of cysteine residues H100) and Cys 109 (H10OE) with Alanines and comprising a in CDR3H and a N-terminal serine; N-terminal serine; SEQ ID NO: 91-amino acid sequence of modified AVP07 SEQID NO: 109-amino acid sequence of modified AVP07 17 anti-HER2 diabody designated AVP07-89 comprising 17 anti-HER2 diabody designated AVP07-86 replacing cysteine residues in FR1, removal of cysteine residues in CDR3H Cysteine residues Cys104 (Kabat numbering H100) CDR3H and a N-terminal serine; and Cys 109 (H10OE) with Alanines and comprising a N-ter SEQ ID NO: 92 nucleotide sequence encoding modified minal serine; AVP07-17 anti-HER2 diabody designated AVP07-91 com SEQ ID NO: 110 nucleotide sequence encoding V of prising cysteine residues in FR1, removal of cysteine residues AVP04-07 anti-TAG72 diabody; in CDR3H and a N-terminal serine; SEQID NO: 111-amino acid sequence of V of AVP04-07 SEQ ID NO:93-amino acid sequence of modified AVP07 anti-TAG72 diabody; 17 anti-HER2 diabody designated AVP07-91 comprising cysteine residues in FR1, removal of cysteine residues in SEQ ID NO: 112 nucleotide sequence encoding V, of CDR3H and a N-terminal serine; AVP04-07 anti-TAG72 diabody; SEQ ID NO: 94 nucleotide sequence encoding modified SEQID NO:113-amino acid sequence of V, of AVP04-07 AVP02-60 anti-MUC1 diabody designated AVP02-103 com anti-TAG72 diabody; prising cysteine residues in FR1 and a N-terminal serine; SEQID NO: 114 nucleotide sequence encoding AVP04-07 SEQ ID NO: 95-amino acid sequence of modified AVP02 anti-TAG72 diabody lacking a linker sequence (designated 60 anti-MUC1 diabody designated AVP02-103 comprising AVP04-69); cysteine residues in FR1 and a N-terminal serine; SEQ ID NO: 115-amino acid sequence of AVP04-07 anti SEQ ID NO: 96 nucleotide sequence encoding modified TAG72 diabody lacking a linker sequence (designated AVP07-17 anti-HER2 diabody designated AVP07-68 com AVP04-69); prising cysteine residues in FR1, removal of cysteine residues SEQID NO: 116 nucleotide sequence encoding AVP04-07 in CDR3H and a N-terminal serine; anti-TAG72 diabody lacking a linker sequence and amino SEQ ID NO: 97-amino acid sequence of modified AVP07 acid N-terminal to linker in V (designated AVP04-09); 17 anti-HER2 diabody designated AVP07-68 comprising SEQ ID NO: 117 -amino acid sequence of AVP04-07 anti cysteine residues in FR1, removal of cysteine residues in TAG72 diabody lacking a linker sequence and amino acid CDR3H and a N-terminal serine; N-terminal to linker in V (designated AVP04-09); US 2012/0164068 A1 Jun. 28, 2012

SEQ ID NO: 118 nucleotide sequence encoding modified SEQID NO: 134 is an amino acid sequence of a linker; and AVP04-07 anti-TAG72 diabody designated AVP04-50 com SEQID NO: 135 is an amino acid sequence of a linker. prising cysteine residues in FR1 of V: SEQID NO: 119-amino acid sequence of modified AVP04 DETAILED DESCRIPTION OF PREFERRED 07 anti-TAG72 diabody designated AVP04-50 comprising EMBODIMENTS cysteine residues in FR1 of V: SEQ ID NO: 120 nucleotide sequence encoding modified General AVP04-69 anti-TAG72 diabody designated AVP04-50 com prising cysteine residues in FR1 of V, and lacking a linker 0212. Throughout this specification, unless specifically stated otherwise or the context requires otherwise, reference Sequence; to a single step, composition of matter, group of steps or group SEQID NO: 121—amino acid sequence of modified AVP04 of compositions of matter shall be taken to encompass one 69 anti-TAG72 diabody designated AVP04-50 comprising and a plurality (i.e. one or more) of those steps, compositions cysteine residues in FR1 of V, and lacking a linker sequence: of matter, groups of steps or group of compositions of matter. SEQ ID NO: 122 nucleotide sequence encoding modified 0213 Those skilled in the art will appreciate that the AVP04-09 anti-TAG72 diabody designated AVP04-50 com invention described herein is susceptible to variations and prising cysteine residues in FR1 of V, and lacking a linker modifications other than those specifically described. It is to sequence and amino acid N-terminal to linker in V, be understood that the invention includes all such variations SEQID NO: 123—amino acid sequence of modified AVP04 and modifications. The invention also includes all of the steps, 09 anti-TAG72 diabody designated AVP04-50 comprising features, compositions and compounds referred to or indi cysteine residues in FR1 and lacking a linker sequence and cated in this specification, individually or collectively, and amino acid N-terminal to linker in V, any and all combinations or any two or more of said steps or SEQ ID NO: 124 nucleotide sequence encoding modified features. AVP04-07 anti-TAG72 diabody with a N-terminal serine resi 0214. The present invention is not to be limited in scope by the specific embodiments described herein, which are due in V, intended for the purpose of exemplification only. Function SEQID NO: 125-amino acid sequence of modified AVP04 ally-equivalent products, compositions and methods are 07 anti-TAG72 diabody with a N-terminal serine residue in clearly within the scope of the invention, as described herein. V. 0215. Any embodiment herein shall be taken to apply SEQ ID NO: 126—nucleotide sequence encoding modified mutatis mutandis to any other embodiment unless specifically AVP04-69 anti-TAG72 diabody lacking a linker sequence stated otherwise. and comprising a N-terminal serine residue in V, 0216. Unless specifically defined otherwise, all technical SEQID NO: 127 amino acid sequence of modified AVP04 and scientific terms used herein shall be taken to have the 69 anti-TAG72 diabody lacking a linker sequence and com same meaning as commonly understood by one of ordinary prising a N-terminal serine residue in V, skill in the art (for example, in cell culture, molecular genet SEQ ID NO: 128 nucleotide sequence encoding modified ics, immunology, immunohistochemistry, protein chemistry, AVP04-09 anti-TAG72 diabody lacking a linker sequence biochemistry and homology modeling). and amino acid N-terminal to linker in V and comprising a N 0217. Unless otherwise indicated, the recombinant pro terminal serine residue in V, tein, cell culture, and immunological techniques utilized in the present invention are standard procedures, well known to SEQID NO: 129-amino acid sequence of modified AVP04 those skilled in the art. Such techniques are described and 09 anti-TAG72 diabody lacking a linker sequence and amino explained throughout the literature in Sources Such as, J. acid N-terminal to linker in V and comprising a N terminal Perbal, A Practical Guide to Molecular Cloning, John Wiley serine residue in V, and Sons (1984), J. Sambrook et al. Molecular Cloning: A SEQ ID NO: 130 nucleotide sequence encoding modified Laboratory Manual, Cold Spring Harbour Laboratory Press AVP04-50 anti-TAG72 diabody comprising cysteine residues (1989), T. A. Brown (editor), Essential Molecular Biology: A in FR1 of V, and a N-terminal serine residue in V and Practical Approach, Volumes 1 and 2, IRL Press (1991), D. lacking a linker sequence; M. Glover and B. D. Hames (editors), DNA Cloning: A Prac SEQID NO: 131-amino acid sequence of modified AVP04 tical Approach, Volumes 1-4, IRL Press (1995 and 1996), and 50 anti-TAG72 diabody comprising cysteine residues in FR1 F. M. Ausubeletal. (editors), Current Protocols in Molecular of V, and a N-terminal serine residue in V and lacking a Biology, Greene Pub. Associates and Wiley-Interscience linker sequence; (1988, including all updates until present), Ed Harlow and David Lane (editors) Antibodies: A Laboratory Manual, Cold SEQ ID NO: 132 nucleotide sequence encoding modified Spring Harbour Laboratory, (1988), and J. E. Coligan et al. AVP04-50 anti-TAG72 diabody comprising cysteine residues (editors) Current Protocols in Immunology, John Wiley & in FR1 in V, and a N-terminal serine residue in V and Sons (including all updates until present). lacking a linker sequence and amino acid N-terminal to linker 0218. The description and definitions of variable regions in V, and parts thereof, immunoglobulins, antibodies and frag SEQID NO: 133-amino acid sequence of modified AVP04 ments thereof herein may be further clarified by the discus 50 anti-TAG72 diabody comprising cysteine residues in FR1 sion in, for example, Kabat (1987 and/or 1991), Bork et at in V, and a N-terminal serine residue in V and lacking a (1994) and/or Chothia and Lesk (1987 and 1989) or Al linker sequence and amino acid N-terminal to linker in V, Lazikani et at (1997). US 2012/0164068 A1 Jun. 28, 2012

0219. The term “and/or”, e.g., “X and/or Y shall be antibody or a primate (preferably human) antibody. The term understood to mean either “X and Y” or “X or Y and shall be 'antibody also encompasses humanized antibodies, prima taken to provide explicit Support for both meanings or for tized antibodies, human antibodies and chimeric antibodies. either meaning Proteins related to antibodies, and thus encompassed by the 0220. As used herein, the term “between in the context of term “immunoglobulin' include domain antibodies, camelid defining the positioning of an amino acid residue or nucle antibodies and antibodies from cartilaginous fish (i.e., immu otide residue shall be taken to mean any residues located noglobulin new antigen receptors (IgNARS)). Generally, between the two recited residues and the two recited residues. camelid antibodies and IgNARS comprise a V, however lack For example, the term “between residues 8-11” shall be a V and are often referred to as heavy chain immunoglobu understood to include residues 8, 9, 10 and 11 in the context lins. As used herein, the term “immunoglobulin' does not of a KV, or a V and/or the term “between residues 8-12 in encompass T cell receptors and other immunoglobulin-like the context of a wV, shall be understood to mean residues 8, 9, 11 and 12 since a J. V., does not contain residue 10 in the domain containing proteins that are not capable of binding to Kabat numbering system. an antigen, e.g., by virtue of an antigen binding site compris 0221) Throughout this specification the word “comprise’, ing a variable region. Furthermore, the term “immunoglobu or variations such as “comprises” or “comprising, will be lin' does not encompass a protein comprising an immuno understood to imply the inclusion of a stated element, integer globulin domain that does not comprise a FR1, since the or step, or group of elements, integers or steps, but not the invention cannot be performed with Such a protein. exclusion of any other element, integer or step, or group of 0224. As used herein, “variable region” refers to the por elements, integers or steps. tions of the light and heavy chains of an antibody or immu 0222. As used herein the term "derived from shall be noglobulin as defined herein that includes amino acid taken to indicate that a specified integer may be obtained from sequences of CDRs; i.e., CDR1, CDR2, and CDR3, and FRs. a particular source albeit not necessarily directly from that In the case of IgNARs the term “variable region' does not SOUC. require the presence of a CDR2. V. refers to the variable region of the heavy chain. V, refers to the variable region of Selected Definitions the light chain. According to the methods used in this inven 0223) As used herein, the term “immunoglobulin' shall be tion, the amino acid positions assigned to CDRS and FRS are taken to mean an antibody or any antibody-related protein. defined according to Kabat (1987 and 1991). The skilled The skilled artisan will be aware that an antibody is generally artisan will be readily able to use other numbering systems in considered to be a protein that comprises a variable region the performance of this invention, e.g., the hyperVariable loop made up of a plurality of polypeptide chains, e.g., a light chain numbering system of Chothia and Lesk (1987 and/or 1989) variable region (V) and a heavy chain variable region (V). and/or Al-Lazikani et al (1997). An antibody also generally comprises constant domains, 0225. As used herein, the term “heavy chain variable which can be arranged into a constant region or constant region” or “V, shall be taken to mean a protein capable of fragment or fragment crystallisable (Fc). Antibodies can bind binding to one or more antigens, preferably specifically bind specifically to one or a few closely related antigens. Gener ing to one or more antigens and at least comprising a FR1 ally, antibodies comprise a four-chain structure as their basic comprising at least about 30 amino acids. Sequences of exem unit. Full-length antibodies comprise two heavy chains (-50 plary FR1 from a heavy chain are provided herein (see, for 70 kD) covalently linked and two light chains (-23 kD each). example, SEQID NOs: 1 to 13). Preferably, the heavy chain A light chain generally comprises a variable region and a comprises three or four FRS (e.g., FR1, FR2, FR3 and option constant domain and in mammals is either a K light chain or a ally FR4) together with three CDRs. Preferably, a heavy chain w light chain. A heavy chain generally comprises a variable comprises FRs and CDRs positioned as follows residues 1-25 region and one or two constant domain(s) linked by a hinge or 1-30 (FR1), 31-25 (CDR1), 36-49 (FR2), 50-65 (CDR2), region to additional constant domain(s). Heavy chains of 66-94 (FR3), 95-102 (CDR3) and 103-113 (FR4), numbered mammals are of one of the following types C, 6, e, Y, or L. according to the Kabat numbering system. In one example, Each light chain is also covalently linked to one of the heavy the heavy chain is derived from an immunoglobulin compris chains. For example, the two heavy chains and the heavy and ing said heavy chain and a plurality of (preferably 3 or 4) light chains are held together by inter-chain disulfide bonds constant domains or linked to a constant fragment (Fc). and by non-covalent interactions. The number of inter-chain 0226. As used herein, the term “light chain variable disulfide bonds can vary among different types of antibodies. region' or “V, shall be taken to mean a protein capable of Each chain has an N-terminal variable region (V or V, binding to one or more antigens, preferably specifically bind wherein each are ~110 amino acids in length) and one or more ing to one or more antigens and at least comprising a FR1 constant domains at the C-terminus. The constant domain of comprising about 23 amino acids. Sequences of exemplary the light chain (C, which is ~110 amino acids in length) is FR1 from a light chain are provided herein (see, for example, aligned with and disulfide bonded to the first constant domain SEQID NOs: 14 to 43). Preferably, the light chain comprises of the heavy chain (C, which is ~330-440 amino acids in three or four FRS (e.g., FR1, FR2, FR3 and optionally FR4) length). The light chain variable region is aligned with the together with three CDRs. Preferably, a light chain comprises variable region of the heavy chain. The antibody heavy chain FRs and CDRs positioned as follows residues 1-23 (FR1), can comprise 2 or more additional C. domains (such as, C2, 24-34 (CDR1), 35-49 (FR2), 50-56 (CDR2), 57-88 (FR3), C3 and the like) and can comprise a hinge region can be 89-97 (CDR3) and 98-107 (FR4), numbered according to the identified between the C1 and Cm constant domains. Anti Kabat numbering system. In one example, the light chain is bodies can be of any type (e.g., IgG, IgE. IgM, Ig|D, IgA, and derived from an immunoglobulin comprising said light chain IgY), class (e.g., IgG, IgG, IgG, IgG, IgA and IgA) or linked to one constant domain and/or not linked to a constant subclass. Preferably, the antibody is a murine (mouse or rat) fragment (Fc). US 2012/0164068 A1 Jun. 28, 2012

0227. In some examples of the invention the term “frame residues 89-97. These CDRs can also comprise numerous work regions' will be understood to mean those variable insertions, e.g., as described in Kabat (1987 and/or 1991). region residues other than the CDR residues. Each variable 0232. The term “constant region' (syn. CR or fragment region of a naturally-occurring immunoglobulin (e.g., anti crystallizable or Fc) as used herein, refers to a portion of an body) typically has four FRS identified as FR1, FR2, FR3 and immunoglobulin comprising at least one constant domain and FR4. If the CDRs are defined according to Kabat, exemplary which is generally (though not necessarily) glycosylated and light chain FR (LCFR) residues are positioned at about resi which binds to one or more F receptors and/or components of dues 1-23 (LCFR1), 35-49 (LCFR2), 57-88 (LCFR3), and the complement cascade (e.g., confers effector functions). 98-107 (LCFR4). Note that LCFR1 does not comprise resi The heavy chain constant region can be selected from any of due 10, which is included in KLCFR1. Exemplary heavy the five isotypes: C, Ö, e, Y, or L. Furthermore, heavy chains of chain FR(HCFR) residues are positioned at about residues various Subclasses (such as the IgG subclasses of heavy 1-30 (HCFR1), 36-49 (HCFR2), 66-94 (HCFR3), and 103 chains) are responsible for different effector functions and 113 (HCFR4). thus, by choosing the desired heavy chain constant region, 0228. For all immunoglobulin variable regions of the proteins with desired effector function can be produced. Pre invention, “framework region 1’ (FR1) is defined as the resi ferred heavy chain constant regions are gamma 1 (IgG1), dues between the natural N-terminal residue and the start of gamma 2 (IgG2) and gamma 3 (IgG3). the complementarity determining region No. 1 (CDR1). 0233. A “constant domain is a domain in an immunoglo These residues have been numbered by at least two nomen bulin the sequence of which is highly similar in immunoglo clatures being 1) Kabat (1987 and/or 2001) and 2) Chothia bulins/antibodies of the same type, e.g., IgG or IgM or IgE. A and Lesk (1987, 1989 and Al-Lazikani et at 1997). The constant region of an immunoglobulin generally comprises a Chothia and Lesk numbering system was based on the well plurality of constant domains, e.g., the constant region of Y, C. established Kabat system and attempted to correct the num and ö heavy chains comprise three constant domains and the bering of light chain CDR1 and heavy chain CDR1 sequence Fc of Y. C. and Öheavy chains comprise two constant domains. length variability in the immunoglobulin variable regions to A constant region of L and e heavy chains comprises four better fit their actual position in the three-dimensional struc constant domains and the Fc region comprises two constant ture. The CDR-specific numbering adopted by Chothia and domains. Lesk was later modified in 1989 but then reverted in 1997. 0234. As used herein, the term “Fv' shall be taken to mean There are subtle differences between these numbering sys any protein, whether comprised of multiple polypeptides or a tems when dealing with residues found within CDR loops. single polypeptide, in which a V, and a V associate and form 0229. As the skilled person will appreciate, within frame a complex having an antigen binding site, i.e., capable of work region 1, and thus prior to CDR1, a single highly specifically binding to an antigen. The V and the V, which conserved cysteine residue (Cys) is generally present. Within form the antigen binding site can be in a single polypeptide both kappa and lambda variable light chains, this conserved chain or in different polypeptide chains. Furthermorean Fv of cysteine is invariantly in Kabat position 23 and forms a dis the invention (as well as any protein of the invention) may ulphide bond with another highly conserved cysteine residue, have multiple antigen binding sites which may or may not invariantly in Kabat position 88, within the region defined as bind the same antigen. This term shall be understood to framework region 3, between CDR2 and CDR3. However, encompass fragments directly derived from an immunoglo the present invention contemplates indels, generally man bulin as well as proteins corresponding to Such a fragment made indels, e.g., of one, two or three amino acids, which may produced using recombinant means. In some examples, the alter the position of the conserved cysteine relative to other V is not linked to a heavy chain constant domain (C)1 amino acids of FR1. and/or the V, is not linked to a light chain constant domain 0230. The pairing of highly conserved cysteines is subtly (C). Exemplary Fv containing polypeptides or proteins different in variable heavy chains, occurring between con include a Fab fragment, a Fab' fragment, a F(ab') fragment, a served cysteines in invariant Kabat positions 22 (within FR1) ScFv, a diabody, a triabody, a tetrabody or higher order com and 92 (within FR3). However, this pairing is almost perfectly plex, or any of the foregoing linked to a constant region or conserved in all immunoglobulins, Suggesting this disulfide domain thereof, e.g., C2 or C3 domain. A "Fab fragment” bond was probably already present at the beginning of Ig-loop consists of a monovalent antigen-binding fragment of an diversification and was maintained under selective pressure. immunoglobulin, and can be produced by digestion of a The almost perfect conservation of the disulfide bond further whole immunoglobulin with the enzyme papain, to yield a Suggests that it contributes significantly to the stability of the fragment consisting of an intact light chain and a portion of a Ig-loop. heavy chain or can be produced using recombinant means. A 0231. As used herein, the term “complementarity deter “Fab' fragment of an immunoglobulin can be obtained by mining regions' (syn. CDRs; i.e., CDR1, CDR2, and CDR3 treating a whole immunoglobulin with pepsin, followed by or hypervariable region) refers to the amino acid residues of reduction, to yield a molecule consisting of an intact light an immunoglobulin variable region the presence of which are chain and a portion of a heavy chain. Two Fab' fragments are necessary for antigen binding. Each variable region typically obtained per immunoglobulin treated in this manner. A Fab' has three CDR regions identified as CDR1, CDR2 and CDR3. fragment can also be produced by recombinant means. A Each CDR may comprise amino acid residues from a “F(ab')2 fragment of an immunoglobulin consists of a dimer “complementarity determining region' as defined by Kabat of two Fab' fragments held together by two disulfide bonds, (1987 and/or 1991). For example, in a heavy chain variable and is obtained by treating a whole immunoglobulin mol region CDRH1 is between residues 31-35, CDRH2 is ecule with the enzyme pepsin, without Subsequent reduction. between residues 50-65 and CDRH3 is between residues A 'Fab fragment is a recombinant fragment comprising two 95-102. In a light chain CDRL1 is between residues 24-34, Fab fragments linked using, for example a leucine Zipper or a CDRL2 is between residues 50-56 and CDRL3 is between C3 domain. A “single chain Fv' or “scFv' is a recombinant US 2012/0164068 A1 Jun. 28, 2012 molecule containing the variable region fragment (FV) of an meant by the term “selective binding. Generally, but not immunoglobulin in which the variable region of the light necessarily, reference to binding means specific binding, and chain and the variable region of the heavy chain are covalently each term shall be understood to provide explicit support for linked by a suitable, flexible polypeptide linker. A detailed the other term. discussion of exemplary Fv containing proteins falling within 0243 The term the terms “preventing”, “prevent' or “pre the scope of this term is provided herein below. vention' in the context of binding of a protein of the invention 0235. As used herein, the term “antigenbinding site' shall to an antigen shall be taken to mean complete abrogation or be taken to mean a structure formed by a protein that is complete inhibition of binding to the antigen. capable of specifically binding to an antigen. The antigen binding site need not be a series of contiguous amino acids, or Variable Region Containing Proteins even amino acids in a single polypeptide chain. For example, in a Fv produced from two different polypeptide chains the 0244. The present invention contemplates any protein that antigen binding site is made up of a series of regions of a V comprises an immunoglobulin variable region that specifi and a V that interact with the antigen and that are generally, cally or selectively binds to one or more antigens and that is however not always in the one or more of the CDRs in each modified as described herein according to any embodiment. variable region. Preferred proteins comprise at least one V and at least one 0236. By “Kabat numbering system’ is meant the num V. Exemplary immunoglobulin variable regions are variable bering system to determining the position of FRs and CDRs in regions from antibodies and modified forms thereof (e.g., a variable region of an immunoglobulin as set out in Kabat humanized antibodies) and heavy chain antibodies, such as, (1987 and/or 1991). camelid immunoglobulin and IgNAR. 0237. The term “protein' shall be taken to include a single polypeptide chain, i.e., a series of contiguous amino acids Immunoglobulin Variable Regions linked by peptide bonds or a series of polypeptide chains Antibody Variable Regions covalently or non-covalently linked to one another (i.e., a polypeptide complex). For example, the series of polypeptide 0245. As will be apparent to the skilled artisan based on chains can be covalently linked using a Suitable chemical or a the description herein, the proteins of the invention can com disulphide bond. Examples of non-covalent bonds include prise one or more variable regions from an antibody modified hydrogen bonds, ionic bonds, Van der Waals forces, and to comprise at least two cysteine residues in FR1. The present hydrophobic interactions. A non-covalent bond contemplated invention also provides antibody molecules. Such antibodies by the present invention is the interaction between a V and a may be produced by first producing an antibody against an V, e.g., in some forms of diabody or a triabody or a tetra antigen of interest and modifying that antibody (e.g., using body. recombinant means) or by modifying a previously produced 0238. The term “polypeptide chain' will be understood to antibody. mean from the foregoing paragraph to mean a series of con 0246 Methods for producing antibodies are known in the tiguous amino acids linked by peptide bonds. art. For example, methods for producing monoclonal anti 0239. The skilled artisan will be aware that a “disulphide bodies, such as the hybridoma technique, are by Kohler and bond is a covalent bond formed by coupling of thiol groups. Milstein, (1975). In a hybridoma method, a mouse, hamster, The bond is also called an SS-bond or disulfide bridge. In or other appropriate host animal, is typically immunized with proteins, a disulphide bond generally occurs between the thiol an immunogen or antigen or cell expressing same to elicit groups of two cysteine residues to produce cystine. lymphocytes that produce or are capable of producing anti 0240. The skilled artisan will also be aware that the term bodies that will specifically bind to the immunogen or anti “non-reducing conditions' includes conditions sufficient for gen. Lymphocytes or spleen cells from the immunized ani oxidation of Sulfhydryl ( SH) groups in a protein, e.g., per mals are then fused with an immortalized cell line using a missive for disulphide bond formation. Suitable fusing agent, such as polyethylene glycol, to form a 0241. As used herein, the term “antigen' shall be under hybridoma cell (Goding, 1986). The resulting hybridoma stood to mean any composition of matter against which an cells may be cultured in a suitable culture medium that pref immunoglobulin response (e.g., an antibody response) can be erably contains one or more Substances that inhibit the growth raised. Exemplary antigens include proteins, peptides, or survival of the unfused, immortalized cells. For example, if polypeptides, carbohydrates, phosphate groups, phosphor the parental cells lack the enzyme hypoxanthine guanine peptides or polypeptides, glyscosylated peptides or peptides, phosphoribosyl transferase (HGPRT or HPRT), the culture etc. medium for the hybridomas typically will include hypoxan 0242. As used herein, the term “specifically binds' shall thine, aminopterin, and thymidine (“HAT medium'), which be taken to mean a protein of the invention reacts or associates substances prevent the growth of HGPRT-deficient cells. more frequently, more rapidly, with greater duration and/or Other methods for producing antibodies are also contem with greater affinity with a particular antigen or antigens or plated by the present invention, e.g., using ABL-MYC tech cell expressing same than it does with alternative antigens or nology described generically in detail in Largaespada (1990) cells. For example, a protein that specifically binds to an or Weissinger et al. (1991). antigen binds that antigen with greater affinity, avidity, more 0247 Alternatively, the antibody, or sequence encoding readily, and/or with greater duration than it binds to other same is generated from a previously produced cell expressing antigens. It is also understood by reading this definition that, an antibody of interest, e.g., a hybridoma or transfectoma. for example, a protein that specifically binds to a first antigen Various sources of Such hybridomas and/or transfectomas may or may not specifically bind to a second antigen. As such, will be apparent to the skilled artisan and include, for “specific binding does not necessarily require exclusive example, American Type Culture Collection (ATCC) and/or binding or non-detectable binding of another antigen, this is European Collection of Cell Cultures (ECACC). Methods for US 2012/0164068 A1 Jun. 28, 2012

isolating and/or modifying sequences encoding variable 0274 DIVMTQTPLSLSVTPGQPASISC (SEQ ID N regions from antibodies will be apparent to the skilled artisan 26); and/or described herein. 0275 EIVLTQSPDFQSVTPKEKVTITC (SEQ ID N 0248 Following antibody production and/or isolation of a 27); sequence encoding same, the antibody is modified to include 0276 ETTLTQSPAFMSATPGDKVNISC (SEQ ID N cysteine residues in FR1 at sites as described herein according 28); to any embodiment. Generally, this involves isolating the (0277 AIRMTQSPFSLSASVGDRVTITC (SEQ ID N nucleic acid encoding the antibody, modifying the sequence 29); thereof to include codons encoding cysteine residues (i.e., 0278 AIQLTQSPSSLSASVGDRVTITC (SEQ ID N TGT or TGC) at the requisite sites in a FR1 encoding region 30): and expressing the modified antibody. 0279 NIQMTQSPSAMSASVGDRVTITC (SEQIDN 0249 Exemplary human antibody heavy chain FR1 31); sequences comprise a sequence selected from the group con 0280 DVVMTQSPLSLPVTLGQPASISC (SEQ IDN sisting of 32); (0250 QVQLVQSGAEVKKPGASVKVSCKASGYTFT (0281 DIVMTQTPLSSPVTLGQPASISC (SEQ ID N (SEQID NO: 1); 33): 0251 QVQLVQSGAEVKKPGASVKVSCKVSGYTLT 0282 DVVMTQSPAFLSVTPGEKVTITC (SEQIDN (SEQID NO: 2): 34); 0252 QMQLVQSGAEVKKTGSSVKVSCKASGYTFT (0283 VIWMTQSPSLLSASTGDRVTISC (SEQ ID N (SEQID NO:3); 35); and 0253). QMQLVQSGPEVKKPGTSVKVSCKASGFTFT (0284 AIRMTQSPSSFSASTGDRVTITC (SEQ ID N (SEQID NO: 4): 36). 0254) QVQLVQSGAEVKKPGSSVKVSCKASGGTFS 0285 Exemplary human antibody w light chain FR1 (SEQ ID NO. 5): sequences comprise a sequence selected from the group con 0255 QVTLKESGPVLVKPTETLTLTCTVSGFSLS sisting of QSVLTQPPSVSAAPGQKVTISC (SEQ ID NO: (SEQID NO: 6): 37); QSVLTQPPSASGTPGQRVTISC (SEQ ID NO: 38); 0256 QITLKESGPTLVKPTQTLTLTCTFSGFSLS QSALTQPASVSGSPGQSITISC (SEQ ID NO. 39); (SEQ ID NO: 7); QSALTQPRSVSGSPGQSVTISC (SEQ ID NO: 40); SYV 0257 QVTLRESGPALVKPTQTLTLTCTFSGFSLS LTQPPSVSVAPGKTARITC (SEQ ID NO: 41); (SEQID NO: 8): SYELTQPPSVSVSPGQTASITC (SEQ ID NO: 42); and 0258 QVQLVESGGGLVKPGGSLRLSCAASGFTFS QLVLTQSPSASASLGASVKLTC (SEQID NO: 43). (SEQ ID NO: 9); 0286 The foregoing sequences are merely exemplary of 0259 EVQLVESGGGLVQPGGSLRLSCAASGFTFS sequences that may be used to perform the invention and are (SEQID NO: 10); not an exhaustive list of Such sequences. These examples are 0260 EVQLVESGGGLVKPGGSLRLSCAASGFTFS provided for the purposes of describing the invention and not (SEQID NO: 11): limiting the invention. It is within the capability of the skilled 0261) EVQLVESGGGVVRPGGSLRLSCAASGFTFD artisan to determine the sequence of an additional FR1 using (SEQID NO: 12); and known methods and/or based on the disclosure in, for 0262 EVOLLESGGGLVQPGGSLRLSCAASGFTFS example, Kabat (1987 and/or 2001). (SEQID NO: 13). 0287. The foregoing examples of FR1 regions are readily 0263. Exemplary human antibody K light chain FR1 modified to include two or more cysteine residues at positions sequences comprise a sequence selected from the group con as described herein in any example or embodiment. sisting of DIQMTQSPSSLSASVGDRVTITC (SEQID NO: 0288 The skilled artisan will be readily able to determine 14); DIQMTQSPSTLSASVGDRVTITC (SEQID NO: 15); the sequence of nucleic acid encoding a FR1 based on knowl 0264. EIVMTQSPATLSVSPGERATLSC (SEQ ID NO: edge in the art and/or sequences set forth herein. 16); 0265 EIVLTQSPATLSLSPGERATLSC (SEQ ID N Chimeric, Deimmunized, Humanized and Human Antibodies 17); 0289. The proteins of the present invention may be derived 0266 EIVLTQSPGTLSLSPGERATLSC (SEQ ID N from or may be humanized antibodies or human antibodies or 18); variable regions derived therefrom. The term “humanized 0267 DIVMTQSPDSLAVSLGERATINC (SEQ ID N antibody' shall be understood to refer to a chimeric molecule, 19); generally prepared using recombinant techniques, having an 0268 DIVMTQSPLSLPVTPGEPASISC (SEQ ID N antigen binding site derived from an antibody from a non 20): human species and the remaining antibody structure of the 0269 DIVMTQSPSSLSASVGDRVTITC (SEQ ID molecule based upon the structure and/or sequence of a 21): human antibody. The antigen-binding site preferably com 0270 EIVMTQSPATLSLSPGERATLSC (SEQ ID prises CDRs from the non-human antibody grafted onto 22); appropriate FRS in the variable regions of a human antibody 0271 DIQMTQSPDFLAVSLGERATINC (SEQ ID N and the remaining regions from a human antibody. Antigen 23); binding sites may be wild type or modified by one or more 0272 EIVLTQSPSSLSASVGDRVTITC (SEQ ID N amino acid substitutions. In some instances, framework resi 24); dues of the human immunoglobulin are replaced by corre 0273 DIVMTQTPLSLPVTPGEPASISC (SEQ ID N sponding non-human residues. Humanized antibodies may 25); also comprise residues which are found neither in the recipi US 2012/0164068 A1 Jun. 28, 2012 ent antibody nor in the imported CDR or framework epitopes, e.g., B cell epitopes or T cell epitopes removed (i.e., sequences. In general, the humanized antibody will comprise mutated) to thereby reduce the likelihood that a subject will substantially all of at least one, and typically two, variable raise an immune response against the protein. Methods for regions, in which all or substantially all of the CDR regions producing deimmunized proteins are known in the art and correspond to those of a non-human immunoglobulin and all described, for example, in WO00/34317, WO2004/108158 or substantially all of the FR regions are those of a human and WO2004/064724. For example, the method comprises immunoglobulin consensus sequence. Methods for humaniz performing an in silico analysis to predict an epitope in a ing non-human antibodies are known in the art. Humaniza protein and mutating one or more residues in the predicted tion can be essentially performed following the method of epitope to thereby reduce its immunogenicity. The protein is U.S. Pat. No. 5,225,539, U.S. Pat. No. 6,054,297 or U.S. Pat. then analyzed, e.g., in silico or in vitro or in vivo to ensure that No. 5,585,089. Other methods for humanizing an antibody it retains its ability to bind to an antigen. Preferable an epitope are not excluded. The skilled artisan will understand that a that occurs within a CDR is not mutated unless the mutation protein of the invention that is not a complete antibody can is unlikely to reduce antigen binding. Methods for predicting also be humanized, e.g., a variable domain can be humanized. antigens are known in the art and described, for example, in 0290 The term “human antibody” as used herein in con Saha (2004). Exemplary potential epitopes in AVP04-07 nection with antibody molecules and binding proteins refers occur at the following positions of SEQ ID NO: 55: 35-41: to antibodies having variable and, optionally, constant anti 68-77: 84-90; 109-119; 122-128; 160-169; and 185-194. body regions derived from or corresponding to sequences Residues that may be mutated to potentially reduce immuno found in humans, e.g. in the human germline or somatic cells. genicity include K38, T71, A72, K74, T87, T112, V113, The “human' antibodies can includeamino acid residues not S114, S115, G116, T125, Q163, Q164, P166, F188, T189, encoded by human sequences, e.g. mutations introduced by G190 or S191. random or site directed mutations in vitro (in particular muta tions which involve conservative Substitutions or mutations in Heavy Chain Immunoglobulins a small number of residues of the antibody, e.g. in 1, 2, 3, 4 or 0293 Heavy chain immunoglobulins differ structurally 5 of the residues of the antibody, preferably e.g. in 1,2,3,4 or from many otherforms of immunoglobulin (e.g., antibodies), 5 of the residues making up one or more of the CDRs of the in so far as they comprise a heavy chain, but do not comprise antibody). These “human antibodies’ do not actually need to a light chain. Accordingly, these immunoglobulins are also be produced by a human, rather, they can be produced using referred to as “heavy chain only antibodies’. Heavy chain recombinant means and/or isolated from a transgenic animal immunoglobulins are found in, for example, camelids and (e.g., a mouse) comprising nucleic acid encoding human cartilaginous fish (also called IgNAR). antibody constant and/or variable regions. Human antibodies 0294 The variable regions present in naturally occurring or fragments thereof can be produced using various tech heavy chain immunoglobulins are generally referred to as niques known in the art, including phage display libraries “V. domains' in camelid Ig and V-NAR in IgNAR, in order (e.g., as described in U.S. Pat. No. 6,300,064; U.S. Pat. No. to distinguish them from the heavy chain variable regions that 5,885,793; U.S. Pat. No. 6,204,023; U.S. Pat. No. 6,291, 158: are present in conventional 4-chain antibodies (which are or U.S. Pat. No. 6,248,516), or using transgenic animals referred to as “Vidomains”) and from the light chain variable expressing human immunoglobulin genes (e.g., as described regions that are present in conventional 4-chain antibodies in WO2002/066630; Lonberget al. (1994) or Jakobovits et al. (which are referred to as “V, domains'). (2007)). 0295 Heavy chain immunoglobulins do not require the 0291. In one example an protein of the invention is a presence of light chains to bind with high affinity and with chimeric antibody or part thereof, e.g., a Fab fragment. The high specificity to a relevant antigen. This feature distin term "chimeric antibody” refers to antibodies in which a guishes heavy chain immunoglobulins from Some conven portion of the heavy and/or light chain is identical with or tional 4-chain antibodies, which comprise both V and V, homologous to corresponding sequences in antibodies domains. This means that single domain binding fragments derived from a particular species (e.g., murine, such as can be derived from heavy chain immunoglobulins, which are mouse) or belonging to a particular antibody class or Sub easy to express and are generally stable and soluble. Heavy class, while the remainder of the chain(s) is identical with or chain immunoglobulins and variable regions domains thereof homologous to corresponding sequences in antibodies domains derived therefrom can also comprise long Surface derived from another species (e.g., primate, such as human) loops (particularly CDR3), which facilitate penetration of and or belonging to another antibody class or Subclass, as well as binding to cavities often found in antigens such as enzymes fragments of Such antibodies, so long as they exhibit the and on the Surface of proteins of viruses and agents causative desired biological activity (U.S. Pat. No. 4,816,567). Typi of infectious diseases. cally chimeric antibodies utilize rodent or rabbit variable 0296. A general description of heavy chain immunoglo regions and human constant regions, in order to produce an bulins from camelids and the variable regions thereof and antibody with predominantly human domains. For example, a methods for their production and/or isolation and/or use is chimeric antibody comprises a variable region from a mouse found inter alia in the following references WO94/04678, antibody modified according to the present invention any WO97/498.05 and WO 97/49805. embodiment fused to a human constant domain and/or a 0297 Exemplary sequences of framework 1 domains human constant region. The production of Such chimeric from heavy chain immunoglobulins from camelids include antibodies is known in the art, and may be achieved by stan the following, dard means (as described, e.g., in U.S. Pat. No. 5,807,715; 0298 GGSVQTGGSLRLSCEISGLTFD (SEQ ID NO: U.S. Pat. No. 4,816,567 and U.S. Pat. No. 4,816,397). 44); 0292. The present invention also contemplates a deimmu 0299 GGSVQTGGSLRLSCAVSGFSFS (SEQ ID NO: nized protein. De-immunized proteins have one or more 45); US 2012/0164068 A1 Jun. 28, 2012 19

0300 GGSEQGGGSLRLSCAISGYTYG (SEQ ID NO: duce a bispecific protein or the V, and V can be different in 46); each polypeptide chain so as to form a trivalent protein. 0301 GGSVQPGGSLTLSCTVSGATYS (SEQ ID NO: (0309 As used herein, the term “tetrabody” shall be taken 47); to mean a protein comprising four associated polypeptide 0302 GGSVQAGGSLRLSCTGSGFPYS (SEQ ID NO: chains, each polypeptide chain comprising the structure 48); V, X-V or V X V, wherein V, is an immunoglo 0303 GGSVQAGGSLRLSCVAGFGTS (SEQ ID NO: bulin light chain variable region, V is an immunoglobulin 49); and heavy chain variable region, X is a linker comprising insuf 0304 GGSVQAGGSLRLSCVSFSPSS (SEQ ID NO: ficient residues to permit the V and V, in a single polypep 50). 0305. A general description of heavy chain immunoglo tide chain to associate (or form an Fv) or is absent, and bulins from cartilaginous fish and the variable regions thereof wherein the V of one polypeptide chain is associated with and methods for their production and/or isolation and/or use the V, of another polypeptide chain to thereby form a tet is found inter alia in WO2005/118629. An exemplary con rameric protein (a tetrabody). The V, and V associate so as sensus sequence for a Type 3 spiny dogfish shark IgNARFR1 to form an antigen binding site, i.e., a Fv capable of specifi comprises the sequence AWVEQTPRTAKETGESLT cally binding to one or more antigens. For example, the V of INCVLT (SEQ ID NO: 51). An exemplary consensus a first polypeptide chain is associated with the V of a second sequence for a Type 3 nurse shark IgNAR FR1 comprises the polypeptide chain, the V of the second polypeptide chain is sequence ARVDQTPKTITKETGESLTINCVLS (SEQ ID associated with the V, of a third polypeptide chain, the V of NO:52). the third polypeptide chain is associated with the V, of a fourth polypeptide chain and the V of the fourth polypeptide Variable Region Containing Proteins chain is associated with the V, of the first polypeptide chain. The V, and V can be the same in each polypeptide chain (i.e., Diabodies, Triabodies, Tetrabodies to produce a monospecific tetrabody) or the V, and V can be 0306 Exemplary preferred proteins comprising an immu of one type in two polypeptide chains and a different type in noglobulin variable region are diabodies, triabodies, tetra the other two polypeptide chains to produce a bispecific tet bodies and higher order protein complexes such as those rabody or the V, and V can be different in each polypeptide described in WO98/044001 and WO947007921. chain so as to form a tetraspecific tetrabody. 0307 As used herein, the term "diabody” shall be taken to 0310. The skilled artisan will be aware of diabodies, tria mean a protein comprising two associated polypeptide bodies and/or tetrabodies and methods for their production. chains, each polypeptide chain comprising the structure Generally, these proteins comprise a polypeptide chain in V, X-V or V X-V, wherein V, is an immunoglo which a V and a V are linked directly or using a linker that bulin light chain variable region, V is an immunoglobulin is of insufficient length to permit the V and V, to associate. heavy chain variable region, X is a linker comprising insuf The V and V can be positioned in any order, i.e., V, V, or ficient residues to permit the V and V, in a single polypep V. V.,. The V and V, are readily obtained, e.g., by isolat tide chain to associate (or form an Fv) or is absent, and ing nucleic acid encoding these polypeptide chains from a wherein the V of one polypeptide chain binds to a V, of the cell expressing an immunoglobulin comprising one or more other polypeptide chain to forman antigenbinding site, i.e., to variable region(s) of interest (including an antibody or a chi form a Fv molecule capable of specifically binding to one or meric antibody or a humanized antibody or a human anti more antigens. The V, and V can be the same in each body) or from a recombinant library expressing V, and V, polypeptide chain or the V, and V can be different in each polypeptide chains (e.g., a scFv library, e.g., as described in polypeptide chain so as to form a bispecific diabody (i.e., EP02394.00 or U.S. Pat. No. 4,946,778). The V, and/or V, comprising two FVS having different specificity). can then readily be modified to include the requisite cysteine 0308 As used herein, the term “triabody' shall be taken to residues as described herein according to any embodiment. mean a protein comprising three associated polypeptide 0311 Proteins comprising V, and V, associate to form chains, each polypeptide chain comprising the structure diabodies, triabodies and/or tetrabodies depending on the V, X-V or V X V, wherein V, is an immunoglo length of the linker (if present) and/or the order of the V and bulin light chain variable region, V is an immunoglobulin V, domains. Preferably, the linker comprises 12 or fewer heavy chain variable region, X is a linker comprising insuf amino acids. For example, in the case of polypeptide chains ficient residues to permit the V and V, in a single polypep having the following structure arranged in N to C order tide chain to associate (or form an Fv) or is absent, and V—X V, wherein X is a linker, a linker having 3-12 wherein the V of one polypeptide chain is associated with residues generally results in formation of diabodies, a linker the V, of another polypeptide chain to thereby formatrimeric having 1 or 2 residues or where a linker is absent generally protein (a triabody). For example, a V of a first polypeptide results in formation of triabodies. In the case of polypeptide chain is associated with the V of a second polypeptide chain, chains having the following structure arranged in N to Corder the V of the second polypeptide chain is associated with the V, X V, wherein X is a linker, a linker having 3-12 V, of a third polypeptide chain and the V of the third residues generally results in formation of diabodies, a linker polypeptide is associated with the V, of the first polypeptide having 1 or 2 residues generally results in formation of dia chain. The V, and V associate so as to form an antigen bodies, triabodies and tetrabodies and a polypeptide lacking a binding site, i.e., a Fv capable of specifically binding to one or linker generally forms triabodies or tetrabodies. more antigens. The V, and V can be the same in each 0312 Linkers for use in fusion proteins are known in the polypeptide chain (i.e., to produce a monospecific triabody) art. Linker sequence composition could affect the folding or two of the V, and two of the V can be the same and the stability of a fusion protein. By indirect fusion of proteins third of each different in the third polypeptide chain to pro through a linker not related to the fused proteins, the steric US 2012/0164068 A1 Jun. 28, 2012 20 hindrance between the two proteins is avoided and the free limited to those that bind to a specific antigen. Exemplary dom degree for the linking is achieved. antigens are described herein for the purposes of illustration 0313. It is often unfavorable to have a linker sequence with and not limitation. high propensity to adopt C.-helix or B-strand structures, which 0322 Exemplary diabodies, triabodies and/or tetrabodies could limit the flexibility of the protein and consequently its comprise a V sequence set forth in amino acids 1-115 of functional activity. Rather, a more desirable linker is a SEQID NO:55 or amino acids 1-129 of SEQID NO. 59 or sequence with a preference to adopt extended conformation. amino acids 1-120 of SEQID NO: 61 or amino acids 1-129 of In practice, most currently designed linker sequences have a SEQID NO: 109, which are modified to include two or more high content of glycine residues that force the linker to adopt cysteine residues in FR1 and/or a N-terminal threonine/serine loop conformation. Glycine is generally used in designed residue. For example, the V comprises a sequence set forth linkers because the absence of a B-carbon permits the 1. polypeptide backbone to access dihedral angles that are ener (i) amino acids 1-115 of SEQID NO:57: getically forbidden for other amino acids. (ii) amino acids 1-115 of SEQID NO: 63: 0314. In one embodiment, the linker is a glycine rich (iii) amino acids 1-115 of SEQID NO: 75: linker. Preferably, the linker is a glycine linker that addition (iv) amino acids 1-115 of SEQID NO: 77: ally comprises alanine and/or serine. Such linkers provide (v) amino acids 1-115 of SEQID NO: 99; flexibility, enhance hydrophilicity and are relatively protease (vi) amino acids 1-129 of SEQID NO: 65; resistant, see, e.g., Kortt et al., 2001. (vii) amino acids 1-129 of SEQID NO: 87: 0315. The conformational flexibility imparted by glycine (viii) amino acids 1-129 of SEQID NO: 89: may be important at the junction between C terminus of the (ix) amino acids 1-129 of SEQID NO: 91: protein and the N terminus of the linker. Accordingly, linkers (x) amino acids 1-129 of SEQID NO: 93: that comprise glycine in the region adjacent to the C terminus (xi) amino acids 1-129 of SEQID NO: 97: of the protein are preferred. In this regard, this does not impart (xii) amino acids 1-120 of SEQID NO: 79; a requirement that the first amino acid residue of the linker (xiii) amino acids 1-120 of SEQID NO: 81; need be a glycine. (xiv) amino acids 1-120 of SEQID NO: 83: 0316 Proline residues can be incorporated into the linker (XV) amino acids 1-120 of SEQID NO: 85; and/or to prevent the formation of significant secondary structural (xvi) amino acids 1-120 of SEQID NO: 95. elements by the linker. For example, a linker comprises the 0323. The diabodies, triabodies and/or tetrabodies com sequence Gly-Pro-Gly, where n is a number between about prise a V, sequence set forth in amino acids 121-234 of SEQ 1 and about 5. ID NO: 55 or amino acids 135-245 of SEQ ID NO. 59 or amino acids 126-232 of SEQ ID NO: 61 or amino acids 0317 Preferred linkers include a sequence selected from 135-245 of SEQID NO: 109, which are modified to include the group consisting of G: GG: GGG: GGGG (SEQID NO: two or more cysteine residues in FR1 and/or a N-terminal 134); GGGGS (SEQID NO: 135): S: SG: SGG; and SGGG. threonine/serine residue. For example, the V comprises a 0318 Diabodies and higher order multimers can also com sequence set forth in: prise proteins that are covalently linked, e.g., by virtue of a (i) amino acids 121-234 of SEQID NO:57: disulphide bond between the proteins, e.g., as described in (ii) amino acids 121-234 of SEQID NO: 63: WO2006/113665. (iii) amino acids 121-234 of SEQID NO: 75: 0319 Multispecific diabodies and higher order multimers (iv) amino acids 121-234 of SEQID NO: 77: can be produced through the noncovalent association of two single chainfusion products comprising V domain from one (v) amino acids 121-234 of SEQID NO: 99; immunoglobulin connected by a short linker to the V, domain (vi) amino acids 135-245 of SEQID NO: 65; of another immunoglobulin, thereby forming two FVs, each (vii) amino acids 135-245 of SEQID NO: 87: from a different immunoglobulin, see, for example, Hudson (viii) amino acids 135-245 of SEQID NO: 89: and Kortt (1999). Similarly, multispecific triabodies can be (ix) amino acids 135-245 of SEQID NO: 91: produced by noncovalent association of three single chain (x) amino acids 135-245 of SEQID NO: 93: fusion proteins as follows: (xi) amino acids 135-245 of SEQID NO: 97: (i) a first protein comprising a V. domain from a first immu (xii) amino acids 126-232 of SEQID NO: 79; noglobulin connected by a short linker to the V, domain of a (xiii) amino acids 126-232 of SEQID NO: 81; second immunoglobulin; (xiv) amino acids 126-232 of SEQID NO: 83: (XV) amino acids 126-232 of SEQID NO: 85; and/or (ii) a second protein comprising a V. domain from the second (xvi) amino acids 126-232 of SEQID NO: 95. immunoglobulin connected by a short linker to the V, domain 0324. The V, and V, described in the foregoing para of a third immunoglobulin; and graphs can be arranged in any order and linked by a Suitable (iii) a third protein comprising a V. domain from the third linker as described herein. For a diabody, the linker preferably immunoglobulin connected by a short linker to the V, domain comprises the sequence GGGS. For a triabody or tetrabody, of the first immunoglobulin. preferably there is no linker or a single glycine residue. 0320. The skilled artisan will readily be able to determine 0325 In one example, a diabody binds to TAG72 and Suitable modifications to the foregoing to produce bispecific comprises at least one polypeptide chain comprising (and triabodies, bispecific tetrabodies, trispecific tetrabodies and preferably two polypeptide chains each comprising) a tetraspecific tetrabodies. sequence set forth in SEQID NO: 55 which are modified to 0321. The present invention contemplates a diabody, tria include two or more cysteine residues in FR1 and/or a N-ter body, tetrabody or higher order multimer against any antigen minal threonine/serine residue. For example, a diabody com or combination thereof, and is not to be construed to be prises at least one polypeptide chain comprising (and prefer US 2012/0164068 A1 Jun. 28, 2012

ably two polypeptide chains each comprising) a sequence set (v) amino acids 1-129 of SEQID NO: 93; or forth in one or more of SEQID NO: 57, 63, 75, 77 or 79. (vi) amino acids 1-129 of SEQID NO: 97. 0326 In one example, a triabody binds to TAG72 and and the V comprises a sequence set forth in one of comprises at least one polypeptide chain comprising (and (i) amino acids 135-245 of SEQID NO: 65; preferably two or three polypeptide chains each comprising) (ii) amino acids 135-245 of SEQID NO: 87: a sequence set forth in SEQID NO: 102. (iii) amino acids 135-245 of SEQID NO: 89: 0327. In another example, a diabody binds to Her2 and (iv) amino acids 135-245 of SEQID NO: 91: comprises at least one polypeptide chain comprising (and (v) amino acids 135-245 of SEQID NO: 93; or preferably two polypeptide chains each comprising) a (vi) amino acids 135-245 of SEQID NO: 97. sequence set forth in SEQID NO: 109 which are modified to 0333. In another example, a schv binds to HER2 and com include two or more cysteine residues in FR1 and/or a N-ter prises a sequence set forth in SEQID NO: 105. minal threonine/serine residue. For example, a diabody com 0334. In a further example, the scFv binds to MUC1 and prises at least one polypeptide chain comprising (and prefer the V comprises a sequence set forth in one of: ably two polypeptide chains each comprising) a sequence set (i) amino acids 1-120 of SEQID NO: 79; forth in one or more of SEQID NO: 65, 87, 89,91, 93 or 97. (ii) amino acids 1-120 of SEQID NO: 81; 0328. In another example, a diabody binds to MUC1 and (iii) amino acids 1-120 of SEQID NO: 83: comprises at least one polypeptide chain comprising (and (iv) amino acids 1-120 of SEQID NO: 85; or preferably two polypeptide chains each comprising) a (v) amino acids 1-120 of SEQID NO: 95. sequence set forth in SEQID NO: 61 which are modified to and the V comprises a sequence set forth in one of include two or more cysteine residues in FR1 and/or a N-ter (i) amino acids 126-232 of SEQID NO: 79; minal threonine/serine residue. For example, a diabody com (ii) amino acids 126-232 of SEQID NO: 81; prises at least one polypeptide chain comprising (and prefer (iii) amino acids 126-232 of SEQID NO: 83: ably two polypeptide chains each comprising) a sequence set (iv) amino acids 126-232 of SEQID NO: 85; or forth in one or more of SEQID NO: 79, 81, 83, 85 or 95. (v) amino acids 126-232 of SEQID NO: 95. Single Chain Fv (scFv) Fragments 0335 The present invention also contemplates a disulfide 0329. The skilled artisan will be aware that schvs com stabilized Fv (or diFv or dsRV), in which a single cysteine prise V and V regions in a single polypeptide chain. Pref residue is introduced into a FR of V and a FR of V, and the erably, the polypeptide chain further comprises a polypeptide cysteine residues linked by a disulfide bond to yield a stable linker between the V and V, which enables the scFv to form Fv (see, for example, Brinkmann et al., 1993). the desired structure for antigen binding (i.e., for the V and 0336 Alternatively, or in addition, the present invention V of the single polypeptide chain to associate with one provides a dimeric Schv, i.e., a protein comprising two scEv another to forma Fv). This is distinct from a diabody or higher molecules linked by a non-covalent or covalent linkage. order multimer in which variable regions from different Examples of such dimeric scFv include, for example, two polypeptide chains associate or bind to one another. For ScFvs linked to a leucine Zipper domain (e.g., derived from example, the linker comprises in excess of 12 amino acid Fos or Jun) whereby the leucine Zipper domains associate to residues with (Gly-Ser) (i.e., GGGGSGGGGSGGGGS form the dimeric compound (see, for example, Kostelny 1992 (SEQ ID NO. 53)) being one of the more favored linkers for or Kruif and Logtenberg, 1996). Alternatively, two scFVs are a scFV. linked by a peptide linker of sufficient length to permit both 0330 Exemplary scFvs comprise a V sequence set forth ScFVs to form and to bind to an antigen, e.g., as described in in amino acids 1-115 of SEQID NO:55 or amino acids 1-129 US20060263367. In a further example, each schv is modified of SEQID NO:59 or amino acids 1-120 of SEQIDNO: 61 or to include a cysteine residue, e.g., in the linker region or at a amino acids 1-129 of SEQID NO: 109, which are modified to terminus, and the Sclvs are linked by a disulfide bond, e.g., as include two or more cysteine residues in FR1 and/or a N-ter described in Albrecht et al., (2004). minal threonine/serine residue. In one example the sclv binds 0337 Modified forms of scFv are also contemplated by to TAG72 and the V comprises a sequence set forth in one of the present invention, e.g., ScPV comprising a linker modified (i) amino acids 1-115 of SEQID NO:57: to permit glycosylation, e.g., as described in US623322. (ii) amino acids 1-115 of SEQID NO: 63: 0338. The skilled artisan will be readily able to produce a (iii) amino acids 1-115 of SEQID NO: 75: sch v or modified form thereof comprising a suitable modified (iv) amino acids 1-115 of SEQID NO: 77; or V, and/or V, according to the present invention based on the (v) amino acids 1-115 of SEQID NO: 99; disclosure herein. Exemplary sequences of V and/or V are and the V, comprises a sequence set forth in one of: described hereinandare to be taken to apply mutatis mutandis (i) amino acids 121-234 of SEQID NO:57: to this embodiment of the invention. (ii) amino acids 121-234 of SEQID NO: 63: 0339. Additional description of scFv is to be found in, for (iii) amino acids 121-234 of SEQID NO: 75: example, U.S. Pat. No. 5,260,203. (iv) amino acids 121-234 of SEQID NO: 77; or (v) amino acids 121-234 of SEQID NO: 99. Minibodies 0331. In one example, a scFv binds to TAG72 and com 0340. The skilled artisan will be aware that a minibody prises a sequence set forth in SEQID NO: 101. comprises the V and V, domains of an immunoglobulin 0332. In another example, the schv binds to Her2 and the fused to the C2 and/or C3 domain of an immunoglobulin. V comprises a sequence set forth in one of Optionally, the minibody comprises a hinge region between (i) amino acids 1-129 of SEQID NO: 65; the Vanda V. Sometimes this conformation is referred to as (ii) amino acids 1-129 of SEQID NO: 87: a Flex Minibody (Hu et al., 1996). A minibody does not (iii) amino acids 1-129 of SEQID NO: 89: comprise a C1 or a CL. Preferably, the V and V, domains (iv) amino acids 1-129 of SEQID NO: 91; are fused to the hinge region and the C3 domain of an US 2012/0164068 A1 Jun. 28, 2012 22 immunoglobulin. Each of the regions may be derived from cal coupling in vitro to form the bispecific protein comprising the same immunoglobulin. Alternatively, the V and V. variable regions. The bispecific protein thus formed was able domains can be derived from one immunoglobulin and the to bind to cells expressing the relevant antigen and normal hinge and C2/C3 from another, or the hinge and C2/C3 human T cells, as well as trigger the lytic activity of human can also be derived from different immunoglobulins. The cytotoxic lymphocytes against human breast tumour targets. present invention also contemplates a multispecific minibody 0348. Additional variable region containing proteins comprising a V and V from one immunoglobulin and a V include, for example, domain antibodies (dAbs) and fusions and a V, from another immunoglobulin. At least one of the thereof (e.g., as described in U.S. Pat. No. 6,248,516), single variable regions of said minibody comprises cysteine resi chain Fab (e.g., Hust et al., 2007) or a Fab (e.g., as described dues in FR1 as described herein. in EP 19930302894). 0341 The skilled artisan will be readily able to produce a minibody of the invention using methods known in the art Constant Domain Fusions together with the teaching provided herein. 0349 The present invention encompasses proteins com 0342 Based on the foregoing, the skilled artisan will prising a variable region and a constant region (e.g., Fc) or a appreciate that minibodies are small versions of whole immu domain thereof, e.g., C2 and/or C3 domain. For example, noglobulins encoded in a single protein chain which retain the the present invention provides a minibody (as discussed antigenbinding region, the C3 domain (or a C2 domain) to above) or a schv-Fc fusion or a diabody-Fc fusion or a tria permit assembly into a bivalent molecule and the immuno body-Fc fusion or a tetrabody-Fc fusion or a scFc-C2 fusion globulin hinge to accommodate dimerization by disulfide or a diabody-C2 fusion or a triabody-C2 fusion or a tetra linkages. body-C2 fusion or a scFV-C3 fusion or a diabody-C3 0343 Exemplary minibodies and methods for their pro fusion or a triabody-C3 fusion or a tetrabody-C3 fusion. duction are described, for example, in WO94/09817. Any of these proteins may comprise a linker, preferably an immunoglobulin hinge region, between the variable region Other Variable Region Containing Proteins and the constant region or constant domain. 0344 U.S. Pat. No. 5,731, 168 describes molecules in 0350. As used herein, the term “hinge region' includes the which the interface between a pair of Fv is engineered to portion of a heavy chain molecule that joins the C1 domain maximize the percentage of heterodimers which are recov to the C2 domain. This hinge region comprises approxi ered from recombinant cell culture to thereby produce bi mately 25 residues and is flexible, thus allowing the two specific proteins. The preferred interface comprises at least a N-terminal antigen binding regions to move independently. part of a C3 domain. In this method, one or more Small Hinge regions can be subdivided into three distinct domains: amino acid side chains from the interface of the first protein upper, middle, and lower hinge domains (Roux et al. 1998). are replaced with larger side chains {e.g., tyrosine or tryp 0351. As used herein, the term “C2 domain” includes the tophan). Compensatory "cavities of identical or similar size portion of a heavy chain immunoglobulin molecule that to the large side chain(s) are created on the interface of the extends, e.g., from between about positions 231-340 accord second protein by replacing large amino acid side chains with ing to the Kabat EU numbering system. Two N-linked Smaller ones (e.g., alanine or threonine). branched carbohydrate chains are generally interposed 0345 Bispecific proteins comprising variable regions between the two CH domains of an intact native IgG mol include cross-linked or "heteroconjugate' proteins. For ecule. In one embodiment, a protein of the invention com example, one of the proteins in the heteroconjugate can be prises a C2 domain derived from an IgG1 molecule (e.g. a coupled to avidin, the other to biotin. Such proteins have, for human IgG1 molecule). In another embodiment, a protein of example, been proposed to target immune system cells to the invention comprises a C2 domain derived from an IgG4 unwanted cells (U.S. Pat. No. 4,676.980). Heteroconjugate molecule (e.g., a human IgG4 molecule). proteins comprising variable regions may be made using any 0352. As used herein, the term “C3 domain includes the convenient cross-linking methods. Suitable cross-linking portion of a heavy chain immunoglobulin molecule that agents are known in the art, and are disclosed in U.S. Pat. No. extends approximately 110 residues from N-terminus of the 4,676.980, along with a number of cross-linking techniques. C2 domain, e.g., from about position 341-446b (Kabat EU 0346 Bispecific proteins comprising variable regions can numbering system). The C3 domain typically forms the also be prepared using chemical linkage. Brennan (1985) C-terminal portion of the immunoglobulin. In some immu describe a procedure wherein intact antibodies are proteolyti noglobulins, however, additional domains may extend from cally cleaved to generate F(ab')2 fragments. These fragments C3 domain to form the C-terminal portion of the molecule are reduced in the presence of the dithiol complexing agent, (e.g. the C4 domain in the L chain of IgM and the e chain of sodium arsenite, to stabilize vicinal dithiols and prevent inter IgE). In one embodiment, a protein of the invention com molecular disulfide formation. The Fab' fragments generated prises a C3 domain derived from an IgG1 molecule (e.g., a are then converted to thionitrobenzoate (TNB) derivatives. human IgG1 molecule). In another embodiment, a protein of One of the Fab'-TNB derivatives is then reconverted to the the invention comprises a C3 domain derived from an IgG4 Fab'-thiol by reduction with mercaptoethylamine and is molecule (e.g., a human IgG4 molecule). mixed with an equimolar amount of the other Fab'-TNB 0353 Constant domain sequences useful for producing derivative to form the bispecific protein. the proteins of the present invention may be obtained from a 0347 Progress has facilitated the direct recovery of Fab'- number of different sources. In preferred embodiments, the SH fragments from E. coli, which can be chemically coupled constant region domain or portion thereof of the protein is to form bispecific proteins comprising variable regions. derived from a human immunoglobulin. It is understood, Shalaby (1992) describe the production of a fully humanized however, that the constant region domain or portion thereof bispecific F(ab') molecule. Each Fab' fragment was sepa may be derived from an immunoglobulin of another mamma rately secreted from E. coli and subjected to directed chemi lian species, including for example, a rodent (e.g. a mouse, US 2012/0164068 A1 Jun. 28, 2012 rat, rabbit, guinea pig) or non-human primate (e.g. chimpan 0358. The protein of the invention may comprise any num Zee, macaque) species. Moreover, the constant region domain ber of constant region domains of different types. or portion thereof may be derived from any immunoglobulin 0359 The constant region domains or portions thereof class, including IgM, IgG, Ig), IgA and IgE, and any immu making up the constant region of a protein may be derived noglobulin isotype, including IgG1, IgG2, IgG3 and IgG4. In from different immunoglobulin molecules. For example, a a preferred example, the human isotype IgG1 is used. protein may comprise a C2 domain or portion thereof 0354) A variety of constant region gene sequences (e.g. derived from an IgG1 molecule and a C3 region or portion human constant region gene sequences) are available in the thereof derived from an IgG3 molecule. form of publicly accessible deposits or the sequence thereof is 0360. In another example of the invention, the protein of available from publicly available databases. Constant region the invention comprises at least a region of an Fc Sufficient to domains can be selected having a particular effector function confer FcRn binding. For example, the portion of the Fc (or lacking a particular effector function) or with a particular region that binds to FcRn comprises from about amino acids modification to reduce immunogenicity. 282-438 of IgG1, according to Kabat numbering. 0355. As used herein, the term “effector function refers to 0361. In one example, a protein of the invention comprises the functional ability of the Fc region orportion thereof (e.g., an altered synthetic constant region wherein or more constant C2 domain) to bind proteins and/or cells of the immune region domains therein are partially or entirely deleted (“do system and mediate various biological effects. Effector func main-deleted constant regions'). The present invention also tions may be antigen-dependent or antigen-independent. encompasses modified Fc regions or parts there having Antigen-dependent effector function” refers to an effector altered, e.g., improved or reduced effector function. Many function which is normally induced following the binding of Such modified Fc regions are known in the art and described, an immunoglobulin to a corresponding antigen. Typical anti for example, in U.S. Pat. No. 7,217,797; U.S. Pat. No. 7,217, gen-dependent effector functions include the ability to binda 798; or US20090041770 (having increased half-life) or complement protein (e.g. C1q). For example, binding of the US2005037000 (increased ADCC). C1 component of complement to the Fc region can activate the classical complement system leading to the opsonisation Mutations to Proteins and lysis of cell pathogens, a process referred to as comple ment-dependent cytotoxicity (CDCC). The activation of 0362. The present invention contemplates the use of complement also stimulates the inflammatory response and mutant forms of a protein of the invention. For example, Such may also be involved in autoimmune hypersensitivity. Other a mutant polypeptide comprises one or more conservative antigen-dependent effector functions are mediated by the amino acid Substitutions compared to a sequence set forth binding of immunoglobulins, via their Fc region, to certain Fc herein. In some examples, the polypeptide comprises 10 or receptors (“FcRs) on cells. There are a number of Fc recep fewer, e.g., 9 or 8 or 7 or 6 or 5 or 4 or 3 or 2 conservative tors which are specific for different classes of immunoglobu amino acid Substitutions. A "conservative amino acid Substi lin, including IgG (gamma receptors, or IgwRs), IgE (epsilon tution' is one in which the amino acid residue is replaced with receptors, or IgeRS), IgA (alpha receptors, or IgCRs) and IgM an amino acid residue having a similar side chain and/or (mu receptors, or IguRs). Binding of immunoglobulin to Fc hydropathicity and/or hydrophilicity. receptors on cell Surfaces triggers a number of important and 0363. In a preferred example, a mutant protein has only, or diverse biological responses including endocytosis of not more than, one or two or three or four conservative amino immune complexes, engulfment and destruction of immuno acid changes when compared to a naturally occurring protein. globulin-coated particles or microorganisms (also called anti Details of conservative amino acid changes are provided body-dependent phagocytosis, or ADCP), clearance of below. As the skilled person would be aware, such minor immune complexes, lysis of antibody-coated target cells by changes can reasonably be predicted not to alter the activity of killer cells (called antibody-dependent cell-mediated cyto the polypeptide when expressed in a recombinant cell toxicity, or ADCC), release of inflammatory mediators, regu 0364 Families of amino acid residues having similar side lation of immune system cell activation, placental transfer chains have been defined in the art, including basic side and control of immunoglobulin production. chains (e.g., lysine, arginine, histidine), acidic side chains 0356. As used herein, the term “antigen-independent (e.g., aspartic acid, glutamic acid), uncharged polar side effector function” refers to an effector function which may be chains (e.g., glycine, asparagine, glutamine, serine, threo induced by an immunoglobulin, regardless of whether it has nine, tyrosine, cysteine), nonpolar side chains (e.g., alanine, bound its corresponding antigen. Typical antigen-indepen Valine, leucine, isoleucine, proline, phenylalanine, methion dent effector functions include cellular transport, circulating ine, tryptophan), B-branched side chains (e.g., threonine, half-life and clearance rates of immunoglobulins, and facili Valine, isoleucine) and aromatic side chains (e.g., tyrosine, tation of purification. A structurally unique Fc receptor, the phenylalanine, tryptophan, histidine). “neonatal Fc receptor or “FoRn, also known as the salvage 0365. The present invention also contemplates one or receptor, plays a critical role in regulating half-life and cel more insertions or deletions compared to a sequence set forth lular transport. Other Fc receptors purified from microbial herein. In some examples, the polypeptide comprises 10 or cells (e.g. Staphylococcal Protein A or G) are capable of fewer, e.g., 9 or 8 or 7 or 6 or 5 or 4 or 3 or 2 insertions and/or binding to the Fc region with high affinity and can be used to deletions. facilitate the purification of the Fc-containing protein. 0357 Constant region domains can be cloned, e.g., using Positioning of Cysteine Residues the polymerase chain reaction and primers which are selected to amplify the domain of interest. The cloning of immuno 0366. The present invention contemplates positioning of globulin sequences is described in for example, in U.S. Pat. cysteine residues in FR1 at any site as described herein in any No. 5,658,570. embodiment or example. US 2012/0164068 A1 Jun. 28, 2012 24

0367. In one example, the present invention provides an another. Methods for predicting the proximity of residues in a isolated protein comprising an immunoglobulin variable protein and/or predicting the likelihood of disulphide bond region comprising at least two cysteine residues positioned formation will be apparent to the skilled artisan and/or within framework region (FR)1, wherein the cysteine resi described herein. dues are positioned Such that at least one of the residues is 0375 Thus, in one example, a protein of the invention capable of being conjugated to a compound and wherein if at comprises at least two cysteine residues positioned within least one of the cysteine residues is not conjugated to a com framework region (FR)1, wherein the cysteine residues are pound a disulphide bond is capable of forming between the within about 2-9 A of one another, preferably, within about cysteine residues. 6-7A of one another. 0368. In another example, the present invention provides 0376. In another example, the cysteine residues are posi an isolated protein comprising an immunoglobulin variable tioned at residues in a protein at which their side chains will region comprising at least two cysteine residues positioned be exposed to solvent. Methods for determining solvent expo within framework region (FR)1, wherein the cysteine resi Sure or solvent accessible Surface area are known in the art dues are positioned Such that at least one of the residues is and include, for example, the Shrake-Rupley algorithm or the capable of being conjugated to a compound and wherein if at LCPO method. least two of the cysteine residues are not conjugated to a 0377 Thus in another example, a protein of the invention compound a disulphide bond is capable of forming between comprises at least two cysteine residues positioned within the cysteine residues. framework region (FR)1, wherein the cysteine residues are 0369. In an alternative or additional example, the present positioned such that their side chains (preferably their thiol invention provides an isolated protein comprising an immu groups) are exposed to Solvent. noglobulin heavy chain variable region (V) and an immu 0378. By “exposed to solvent” shall be understood to noglobulin light chain variable region (V), wherein at least mean that the side chains of the cysteine residues are on the one of the variable regions comprises at least two cysteine surface of a protein when folded such that they are capable of residues positioned within framework region (FR)1, wherein being in contact with a solvent in which the protein is present the cysteine residues are positioned such that at least one of or suspended. Preferably, at least one (or one or both) of the the residues is capable of being conjugated to a compound side chains are Sufficiently exposed to solvent such that a and wherein if at least one of the cysteine residues is not compound can be conjugated thereto. conjugated to another compound a disulphide bond is capable 0379 Preferably, the protein of the invention comprises at of forming between the cysteine residues. least two cysteine residues positioned at one or more of 0370. In an alternative or additional example, the present preferably two or more of preferably all of: invention provides an isolated protein comprising an immu (i) positioned such that their side chains are angled towards noglobulin heavy chain variable region (V) and an immu one another; noglobulin light chain variable region (V), wherein at least (ii) positioned Such that their side chain atoms are exposed to one of the variable regions comprises at least two cysteine Solvent; and/or residues positioned within framework region (FR)1, wherein (iii) positioned such that their Co. carbon atoms are about 6-7 the cysteine residues are positioned such that at least one of A of one another. the residues is capable of being conjugated to a compound 0380. The proteins of the present invention (as described and wherein if at least two of the cysteine residues are not hereinaccording to any one or more example of the invention) conjugated to another compound a disulphide bond is capable thus provide at least two cysteine residues positioned within of forming between the cysteine residues. framework region 1 (FR1) that can form a disulphide bond 0371. In each of the above examples of the invention, it is within FR1 and which can alternatively be reduced for sto preferable that at least two or the at least two cysteine residues ichiometric conjugation of compounds. These products of the are positioned such that they are capable of being conjugated invention have an advantage over other cysteine conjugation to a compound. strategies that do not provide at least two cysteine residues 0372. In one example of the invention, the cysteine resi positioned within framework region (FR) 1 that can form a dues are positioned within a loop region of FR1 As used disulphide bond within FR1. These prior and ineffective strat herein, the term “loop region of FR1” shall be taken to mean egies include single cysteine residues (Kim et al., 2008), a sequence of amino acids within FR1 that is provides flex C-terminal cysteine residues (Sirk et al., 2008) and single ibility for two regions and/or two amino acids of FR1 to cysteine residues in intactantibodies (Junutula et al., 2008) all associate with or bind to one another (e.g., by virtue of a of which result in poor expression yield, variable conjugation hydrogen bond), e.g., that provides sufficient flexibility for and complications for large scale processing. Furthermore, two amino acids in a beta sheet to associate with orbind to one antibodies that are conjugated on cysteine residues by partial another. A loop region of FR1 is not part of the CDR1. reduction of interchain-disulfide bonds have variable stoichi 0373) In another example, the cysteine residues in a FR1 ometry (Zero to eight drugs per antibody) and potentially are positioned so as to permit formation of a disulfide bond yield>100 species (Junutula et al., 2008). between the residues. 0381 Methods for predicting loops and/or the position of 0374. By “positioned so as to permit formation of a disul residues within a folded protein will be apparent to the skilled phide bond shall be understood to mean that two cysteine artisan and include in silico methods. For example, structural residues are positioned within a protein such that when the features of a protein are determined using appropriate soft protein folds they are sufficiently close for a disulphide bond ware available on the website of the National Center for to be formed between the residues. For example, the distance Biotechnology Information (NCBI) at the National Institutes between two carbon atoms in two cysteine residues may be of Health, 8600 Rockville Pike, Bethesda Md. 20894 such as, within about 6-7 A of one another or 2-9 A of one another, for example, through the NCBI Molecules Modelling Data such as about 3.5-6.8A of one another, e.g., about 4A of one base (MMDB) including three-dimensional biomolecular US 2012/0164068 A1 Jun. 28, 2012

structures determined using X-ray crystallography and/or residues are included. For example, the cysteine residues are NMR spectroscopy. The NCBI conserved domain database paired, i.e., combinations of two residues are arranged Such (CDD) includes domains from the known Smart and Pham that a disulphide bond can form between them. collections, with links to a 3D-structure viewer (Cn3D). The 0386 Preferably, a protein of the invention does not com NCBI Conserved Domain Architecture Retrieval Tool prise a free thiolin FR1 under non-reducing conditions and/or (CDART) uses precalculated domain assignments to neigh does not comprise a cysteine residue that is not linked to bor proteins by their domain architecture. another cysteine residue or to a compound under non-reduc 0382. Additional methods for predicting protein or pep ing conditions. tide secondary structure are known in the art and/or described, for example, in Moult, 1996; Chou et al., 1974: Protein Production Chou et al., 1974; Chou et al., 1978; Chou et al., 1978; or Mutagenesis Chou et al., 1979. 0383. Additionally, computer programs are currently 0387 DNA encoding a protein comprising a variable available to assist with predicting secondary structure of a region is isolated using standard methods in the art. For protein or peptide. One such method of predicting secondary example, primers are designed to anneal to conserved regions structure is based upon homology modeling. For example, within a variable region that flank the region of interest, and two proteins that have a sequence identity of greater than those primers are then used to amplify the intervening nucleic 30%, or similarity greater than 40%, often have similar struc acid, e.g., by PCR. Suitable methods and/or primers are tural topologies. The recent growth of the protein structural known in the art and/or described, for example, in Borrebaeck database (PDB) has provided enhanced predictability of sec (ed), 1995 and/or Froyen et al., 1995. Suitable sources of ondary structure, including the potential number of folds template DNA for such amplification methods is derived within the structure of a protein (Holm et al., 1999). For from, for example, hybridomas, transfectomas and/or cells example, methods for determining the structure of a protein expressing proteins comprising a variable region, e.g., as are described, for example, in US20020150906, or using a described herein. computer program or algorithm, such as, for example, MOD 0388. Following isolation, the DNA is modified to include ELLER, (Sali and Blundell, 1993). These techniques rely cysteine residues at the requisite locations by any of a variety upon aligning the sequence of a protein with the sequences of of methods known in the art. These methods include, but are proteins that have a characterized structure. Such alignment not limited to, preparation by site-directed (or oligonucle algorithms are known in the art and are accessed through otide-mediated) mutagenesis, PCR mutagenesis, and cassette software packages such as, for example BLAST at NCBI. mutagenesis of an earlier prepared DNA encoding the pro Structural information, i.e., three-dimensional structure, of a tein. Variants of recombinant proteins may be constructed query protein is then be predicted based upon structural infor also by restriction fragment manipulation or by overlap mation corresponding to the sequence or Subsequences extension PCR with synthetic oligonucleotides. Mutagenic aligned in the proteins or peptides that have previously been primers encode the cysteine codon replacement(s), for characterized. In this way it is possible to generate a library of example include residues that make up a codon encoding three-dimensional structures of proteins corresponding to a cysteine (i.e., TGT or TGC). Standard mutagenesis tech FR1 region of an immunoglobulin. niques can be employed to generate DNA encoding Such 0384 Additional methods of predicting secondary struc mutant DNA. General guidance can be found in Sambrook et ture include, for example, “threading’ (Jones, 1996), “profile all 1989; and/or Ausubel et al 1993. analysis” (Bowie et al., 1991: Gribskov et al., 1990; Gribskov 0389. Site-directed mutagenesis is one method for prepar et al., 1989), and “evolutionary linkage'. Conventional ing Substitution variants, i.e. mutant proteins. This technique threading of protein sequence is used to predict the 3D struc is known in the art (see for example, Carter et al 1985; Ho et ture scaffold of a protein. Typically, threading is a process of all 1989; and Kunkel 1987). Briefly, in carrying out site assigning the folding of the protein by threading (or compar directed mutagenesis of DNA, the starting DNA is altered by ing) its sequence to a library of potential structural templates first hybridizing an oligonucleotide encoding the desired (e.g., known structures of Fv or Fabs or FR1) by using a mutation (e.g., insertion of one or more cysteine encoding scoring function that incorporates the sequence as well as the codons) to a single strand of such starting DNA. After hybrid local parameters such as secondary structure and solvent ization, a DNA polymerase is used to synthesize an entire exposure (Rost et al. 1997; Xu and Xu 2000; and Panchenko second strand, using the hybridized oligonucleotide as a et al. 2000). For example, the threading process starts from primer, and using the single strand of the starting DNA as a prediction of the secondary structure of the amino acid template. Thus, the oligonucleotide encoding the desired sequence and solvent accessibility for each residue of the mutation is incorporated in the resulting double-stranded query sequence. The resulting one-dimensional (1D) profile DNA. Site-directed mutagenesis may be carried out within of the predicted structure is threaded into each member of a the gene expressing the protein to be mutagenized in an library of known 3D structures. The optimal threading for expression plasmid and the resulting plasmid may be each sequence-structure pair is obtained using dynamic pro sequenced to confirm the introduction of the desired cysteine gramming. The overall best sequence-structure pair consti replacement mutations. Site-directed protocols and formats tutes the predicted 3D structure for the query sequence. include commercially available kits, e.g. QuikChange R. Threading is made relatively simple in the present case Multi Site-Directed Mutagenesis Kit (Stratagene, La Jolla, because of the number of Fv and Fab fragments of immuno Calif.). globulins for which the secondary structure has been solved. 0390 PCR mutagenesis is also suitable for making amino 0385. In the case of proteins comprising more than two acid sequence variants of the starting protein. See Higuchi, cysteine residues, it is preferred that an even number of cys 1990; Ito et al 1991; Bernhard et al 1994; and Vallette et al teine resides are included, e.g., 4 or 6 or 8 or 10 cysteine 1989. Briefly, when small amounts oftemplate DNA are used US 2012/0164068 A1 Jun. 28, 2012 26 as starting material in a PCR, primers that differ slightly in (e.g., upstream activating sequences, transcription factor sequence from the corresponding region in a template DNA binding sites, enhancers and silencers) that alter expression of can be used to generate relatively large quantities of a specific a nucleic acid, e.g., in response to a developmental and/or DNA fragment that differs from the template sequence only at external stimulus, or in a tissue specific manner. In the present the positions where the primers differ from the template. context, the term “promoter' is also used to describe a recom 0391 Another method for preparing variants, cassette binant, synthetic or fusion nucleic acid, or derivative which mutagenesis, is based on the technique described by Wells et confers, activates or enhances the expression of a nucleic acid al, 1985. The starting material is the plasmid (or other vector) to which it is operably linked. Preferred promoters can con comprising the starting protein DNA to be mutated. The tain additional copies of one or more specific regulatory ele codon(s) in the starting DNA to be mutated are identified. ments to further enhance expression and/or alter the spatial There must be a unique restriction endonuclease site on each expression and/or temporal expression of said nucleic acid. side of the identified mutation site(s). If no such restriction 0396. As used herein, the term “operably linked to means sites exist, they may be generated using the above described positioning a promoter relative to a nucleic acid Such that oligonucleotide-mediated mutagenesis method to introduce expression of the nucleic acid is controlled by the promoter. them at appropriate locations in the starting DNA. The plas 0397 Cell free expression systems are also contemplated mid DNA is cut at these sites to linearize it. A double-stranded by the present invention. For example, a nucleic acid encod oligonucleotide encoding the sequence of the DNA between ing a protein of the invention is operably linked to a suitable the restriction sites but containing the desired mutation(s) is promoter, e.g., a T7 promoter, and the resulting expression synthesized using standard procedures, wherein the two construct exposed to conditions Sufficient for transcription Strands of the oligonucleotide are synthesized separately and and translation. Typical expression vectors for invitro expres then hybridized together using standard techniques. This sion or cell-free expression have been described and include, double-stranded oligonucleotide is referred to as the cassette. but are not limited to the TNT T7 and TNT T3 systems This cassette is designed to have 5' and 3' ends that are (Promega), the pEXP1-DEST and pEXP2-DEST vectors (In compatible with the ends of the linearized plasmid, such that vitrogen). it can be directly ligated to the plasmid. This plasmid now 0398. Many vectors for expression in cells are available. contains the mutated DNA sequence. Mutant DNA contain The vector components generally include, but are not limited ing the encoded cysteine replacements can be confirmed by to, one or more of the following: a signal sequence, a DNA sequencing. sequence encoding protein of the present invention (e.g., 0392. Single mutations are also generated by oligonucle derived from the information provided herein), an enhancer otide directed mutagenesis using double stranded plasmid element, a promoter, and a transcription termination DNA as template by PCR based mutagenesis (Sambrook and sequence. The skilled artisan will be aware of suitable Russel, 2001; Zoller et al 1983; Zoller and Smith, 1982). sequences for expression of a protein. For example, exem plary signal sequences include prokaryotic secretion signals Recombinant Expression (e.g., pelB, alkaline phosphatase, penicillinase, Ipp, or heat 0393. In the case of a recombinant protein, nucleic acid stable enterotoxin II), yeast secretion signals (e.g., invertase encoding same is preferably placed into expression vectors, leader, C. factor leader, or acid phosphatase leader) or mam which are then transfected into host cells, preferably cells that malian secretion signals (e.g., herpes simplex g|D signal). can produce a disulphide bridge or bond, such as E. coli cells, 0399. Exemplary promoters include those active in yeast cells, insect cells, or mammalian cells, such as simian prokaryotes (e.g., phoA promoter, B-lactamase and lactose COS cells, Chinese Hamster Ovary (CHO) cells, or myeloma promoter systems, alkaline phosphatase, a tryptophan (trp) cells that do not otherwise produce immunoglobulin protein, promoter system, and hybrid promoters such as the tac pro to obtain the synthesis of proteins in the recombinant host moter). These promoter are useful for expression in prokary cells. Review articles on recombinant expression in bacteria otes including eubacteria, such as Gram-negative or Gram of DNA encoding the immunoglobulin include Skerra et al. positive organisms, for example, Enterobacteriaceae such as (1993) and Plückthun, (1992). Molecular cloning techniques Escherichia, e.g., E. coli, Enterobacter, Erwinia, Klebsiella, to achieve these ends are known in the art and described, for Proteus, Salmonella, e.g., Salmonella typhimurium, Serratia, example in Ausubel or Sambrook. A wide variety of cloning e.g., Serratia marcescans, and Shigella, as well as Bacilli and in vitro amplification methods are suitable for the con Such as B. subtilis and B. licheniformis, Pseudomonas Such as struction of recombinant nucleic acids. Methods of producing P. aeruginosa, and Streptomyces. Preferably, the host is E. recombinant immunoglobulins are also known in the art. See coli. One preferred E. coli cloning host is E. coli 294 (ATCC U.S. Pat. No. 4,816,567; U.S. Pat. No. 5,225,539, U.S. Pat. 31,446), although other strains such as E. coli B. E. coli X No. 6,054,297, U.S. Pat. No. 7,566,771 or U.S. Pat. No. 1776 (ATCC 31,537), and E. coli W3110 (ATCC 27,325), 5,585,089. DH5a or DH1 OB are Suitable. 0394 Following isolation, the nucleic acid encoding a 04.00 Exemplary promoters active in mammalian cells protein of the invention is preferably inserted into an expres include cytomegalovirus immediate early promoter (CMV sion constructor replicable vector for further cloning (ampli IE), human elongation factor 1-cc promoter (EF1), Small fication of the DNA) or for expression in a cell-free system or nuclear RNA promoters (U1a and U1b), C-myosin heavy in cells. Preferably, the nucleic acid is operably linked to a chain promoter, Simian virus 40 promoter (SV40), Roussar promoter, coma virus promoter (RSV), Adenovirus major late promoter, 0395. As used herein, the term “promoter' is to be taken in B-actin promoter, hybrid regulatory element comprising a its broadest context and includes the transcriptional regula CMV enhancer/B-actin promoter oran immunoglobulin pro tory sequences of a genomic gene, including the TATA box or moter or active fragment thereof. Examples of useful mam initiator element, which is required for accurate transcription malian host cell lines are monkey kidney CV1 line trans initiation, with or without additional regulatory elements formed by SV40 (COS-7, ATCC CRL 1651); human US 2012/0164068 A1 Jun. 28, 2012 27 embryonic kidney line (293 or 293 cells subcloned for growth Where the protein is secreted into the medium, supernatants in suspension culture; baby hamster kidney cells (BHK, from Such expression systems are generally first concentrated ATCC CCL 10); or Chinese hamster ovary cells (CHO). using a commercially available protein concentration filter, 04.01 Typical promoters suitable for expression in yeast for example, an Amicon or Millipore Pellicon ultrafiltration cells such as for example a yeast cell selected from the group unit. A protease inhibitor such as PMSF may be included in comprising Pichia pastoris, Saccharomyces cerevisiae and S. any of the foregoing steps to inhibit proteolysis and antibiot pombe, include, but are not limited to, the ADH1 promoter, ics may be included to prevent the growth of adventitious the GAL1 promoter, the GAL4 promoter, the CUP1 pro contaminants. moter, the PHO5 promoter, the nmt promoter, the RPR1 0408. The protein prepared from the cells can be purified promoter, or the TEF1 promoter. using, for example, hydroxyl apatite chromatography, gel 0402 Typical promoters suitable for expression in insect electrophoresis, dialysis, and affinity chromatography, with cells include, but are not limited to, the OPEI2 promoter, the affinity chromatography being the preferred purification insect actin promoter isolated from Bombyx muri, the Droso technique. The Suitability of protein A as an affinity ligand phila sp. dsh promoter (Marsh et al 2000) and the inducible depends on the species and isotype of any immunoglobulin Fc metallothionein promoter. Preferred insect cells for expres domain that is present in the protein (if present at all). Protein sion of recombinant proteins include an insect cell selected A can be used to purify immunoglobulins that are based on from the group comprising, BT1-TN-5B1-4 cells, and human y1, y2, or y4 heavy chains (Lindmark et al. 1983). Spodoptera frugiperda cells (e.g., Sf19 cells, Sf21 cells). Suit Protein G is recommended for all mouse isotypes and for able insects for the expression of the nucleic acid fragments human Y3 (Guss et al. 1986). Otherwise affinity purification include but are not limited to Drosophila sp. The use of S. can be performed using the antigen or epitopic determinant to frugiperda is also contemplated. which a variable region in a protein of the invention binds or 0403. Means for introducing the isolated nucleic acid mol was raised. The matrix to which the affinity ligand is attached ecule or a gene construct comprising same into a cell for is most often agarose, but other matrices are available. expression are knownto those skilled in the art. The technique Mechanically stable matrices Such as controlled pore glass or used for a given cell depends on the known Successful tech poly(styrenedivinyl)benzene allow for faster flow rates and niques. Means for introducing recombinant DNA into cells shorter processing times than can be achieved with agarose. include microinjection, transfection mediated by DEAE-dex Other techniques for protein purification Such as fractionation tran, transfection mediated by liposomes such as by using on an ion-exchange column, ethanol precipitation, Reverse lipofectamine (Gibco, Md., USA) and/or cellfectin (Gibco, Phase HPLC, chromatography on silica, chromatography on Md., USA), PEG-mediated DNA uptake, electroporation and heparin SEPHAROSETM chromatography on an anion or cat microparticle bombardment such as by using DNA-coated ion exchange resin (such as a polyaspartic acid column), tungsten or gold particles (Agracetus Inc., WI, USA) amongst chromatofocusing, SDS-PAGE, and ammonium sulfate pre others. cipitation are also available depending on the protein to be 04.04 The host cells used to produce the protein of this recovered. invention may be cultured in a variety of media, depending on 04.09. The skilled artisan will also be aware that a protein the cell type used. Commercially available media such as of the invention can be modified to include a tag to facilitate Ham's F10 (Sigma), Minimal Essential Medium ((MEM), purification or detection, e.g., a poly-histidine tag, e.g., a (Sigma), RPM1-1640 (Sigma), and Dulbecco's Modified hexa-histidine tag, or a influenza virus hemagglutinin (HA) Eagle's Medium (DMEM), Sigma) are suitable for culturing tag, or a Simian Virus 5 (V5) tag, or a FLAG tag, or a glu mammalian cells. Media for culturing other cell types dis tathione S-transferase (GST) tag. Preferably, the tag is a hexa cussed herein are known in the art. his tag. The resulting protein is then purified using methods known in the art, such as, affinity purification. For example, a Isolation of Proteins protein comprising a hexa-his tag is purified by contacting a 04.05) A protein of the present invention is preferably iso sample comprising the protein with nickel-nitrilotriacetic lated. By "isolated' is meant that the protein is substantially acid (Ni-NTA) that specifically binds a hexa-his tag immobi purified or is removed from its naturally-occurring environ lized on a solid or semi-solid Support, washing the sample to ment, e.g., is in a heterologous environment. By “substan remove unbound protein, and Subsequently eluting the bound tially purified” is meant the protein is substantially free of protein. Alternatively, or in a addition a ligand or antibody contaminating agents, e.g., at least about 70% or 75% or 80% that binds to a tag is used in an affinity purification method. or 85% or 90% or 95% or 96% or 97% or 98% or 99% free of 0410. Following any preliminary purification step(s), the contaminating agents. mixture comprising the protein of the invention and contami 04.06 Methods for purifying a protein of the invention are nants may be subjected to low pH hydrophobic interaction known in the art and/or described herein. chromatography. 0407. When using recombinant techniques, the protein of the invention can be produced intracellularly, in the periplas Protein Synthesis mic space, or directly secreted into the medium. If the protein 0411 A protein of the present invention is readily synthe is produced intracellularly, as a first step, the particulate sized from its determined amino acid sequence using stan debris, either host cells or lysed fragments, is removed, for dard techniques, e.g., using BOC or FMOC chemistry. Syn example, by centrifugation or ultrafiltration. Carter et al. thetic peptides are prepared using known techniques of solid (1992) describe a procedure for isolating antibodies which phase, liquid phase, or peptide condensation, or any combi are secreted to the periplasmic space of E. coli. Briefly, cell nation thereof, and can include natural and/or unnatural paste is thawed in the presence of sodium acetate (pH 3.5), amino acids. Amino acids used for peptide synthesis may be EDTA, and phenylmethylsulfonylfluoride (PMSF) over standard Boc (NC.-amino protected No.-t-butyloxycarbonyl) about 30 min. Cell debris can be removed by centrifugation. amino acid resin with the deprotecting, neutralization, cou US 2012/0164068 A1 Jun. 28, 2012 28 pling and wash protocols of the original Solid phase procedure proliferation (such as TNP-470, squalamine, 2-methox of Merrifield, 1963, or the base-labile No.-amino protected yestradiol, combretastatins, endostatin, angiostatin, penicil 9-fluorenylmethoxycarbonyl (Fmoc) amino acids described lamine, SCH66336 (Schering-Plough Corp, Madison, NJ), by Carpino and Han, 1972. Both Fmoc and Boc NC.-amino R115777 (Janssen Pharmaceutica, Inc., Titusville, NJ) and protected amino acids can be obtained from various commer similar agents), antagonists of angiogenic growth factors cial sources, such as, for example, Fluka, Bachem, Advanced (such as such as ZD6474, SU6668, antibodies againstangio Chemtech, Sigma, Cambridge Research Biochemical, genic agents and/or their receptors (such as VEGF, bFGF, and Bachem, or Peninsula Labs. angiopoietin-1), thalidomide, thalidomide analogs (such as CC-5013), Sugen 5416, SU5402, antiangiogenic ribozyme Conjugates (such as angiozyme), interferon C. (Such as interferon C2a). 0412. The present invention also provides conjugates of Suramin and similar agents), VEGF-R kinase inhibitors and proteins described herein according to any embodiment. other anti-angiogenic tyrosine kinase inhibitors (such as Examples of compounds to which a protein can be conjugated SUO11248), inhibitors of endothelial-specific integrin/sur are the compound is selected from the group consisting of a vival signaling (such as vitaxin and similar agents), copper radioisotope, a detectable label, a therapeutic compound, a antagonists/chelators (such as tetrathiomolybdate, captopril colloid, a toxin, a nucleic acid, a peptide, a protein, a com and similar agents), carboxyamido-triazole (CAI), ABT-627, pound that increases the halflife of the protein in a subject and CM101, interleukin-12 (IL-12), IM862, PNU145156E as mixtures thereof. Exemplary therapeutic agents include, but well as nucleotide molecules inhibiting angiogenesis (such as are not limited to an anti-angiogenic agent, an anti-neovas antisense-VEGF-cDNA, cDNA coding for angiostatin, cularization and/or other vascularization agent, an anti-pro cDNA coding for p53 and cDNA coding for deficient VEGF liferative agent, a pro-apoptotic agent, a chemotherapeutic receptor-2) and similar agents. Other examples of inhibitors agent or a therapeutic nucleic acid. of angiogenesis, neovascularization, and/or other vascular 0413 A toxin includes any agent that is detrimental to ization are anti-angiogenic heparin derivatives and related (e.g., kills) cells. For a description of these classes of drugs molecules (e.g., heperinase III), temozolomide, NK4, mac which are known in the art, and their mechanisms of action, rophage migration inhibitory factor (MIF), cyclooxyge see Goodman et al., Goodman and Gilman's The Pharmaco nase-2 inhibitors, inhibitors of hypoxia-inducible factor 1, logical Basis of Therapeutics, 8th Ed., Macmillan Publishing anti-angiogenic Soy isoflavones, oltipraz, fumagillinandana Co., 1990. Additional techniques relevant to the preparation logs thereof. Somatostatin analogues, pentosan polysulfate, of immunoglobulin-immunotoxin conjugates are provided in tecogalan sodium, dalteparin, tumstatin, thrombospondin, for instance Vitetta (1993) and U.S. Pat. No. 5,194.594. NM-3, combrestatin, can statin, avastatin, antibodies against Exemplary toxins include diphtheria A chain, nonbinding other relevant targets (such as anti-alpha-V/beta-3 integrin active fragments of diphtheria toxin, eXotoxin A chain (from and anti-kininostatin mAbs) and similar agents. Pseudomonas aeruginosa), ricin A chain, abrin A chain, 0416) In one example, a protein as described herein modeccin. A chain, alpha-sarcin, Aleurites fordii proteins, according to any embodiment is conjugated or linked to dianthin proteins, Phytolaca americana proteins (PAPI, another protein, including another protein of the invention or PAPII, and PAP-S), momordica charantia inhibitor, curcin, a protein comprising an immunoglobulin variable region, crotin, Sapaonaria officinalis inhibitor, gelonin, mitogellin, Such as an immunoglobulin or a protein derived therefrom, restrictocin, phenomycin, enomycin and the tricothecenes. e.g., as described herein. Other proteins are not excluded. See, for example, WO 93/21232. Additional proteins will be apparent to the skilled artisan and 0414 Suitable chemotherapeutic agents for forming include, for example, an immunomodulator or a half-life immunoconjugates of the present invention include taxol. extending protein or a peptide or other protein that binds to cytochalasin B, gramicidin D, ethidium bromide, emetine, serum albumin amongst others. mitomycin, etoposide, tenoposide, Vincristine, vinblastine, 0417 Exemplary immunomodulators include cytokines colchicin, doxorubicin, daunorubicin, dihydroxy anthracin and chemokines. The term “cytokine' is a generic term for dione, mitoxantrone, mithramycin, actinomycin D, 1-de-hy proteins or peptides released by one cell population which act drotestosterone, glucocorticoids, procaine, tetracaine, on another cell as intercellular mediators. Examples of cytok lidocaine, propranolol, and puromycin, antimetabolites (such ines include lymphokines, monokines, growth factors and as methotrexate, 6-mercaptopurine, 6-thioguanine, cytara traditional polypeptide hormones. Included among the cytok bine, fludarabin, 5-fluorouracil, decarbazine, hydroxyurea, ines are growth hormones such as human growth hormone, asparaginase, gemcitabine, cladribine), alkylating agents N-methionyl human growth hormone, and bovine growth (such as mechlorethamine, thioepa, chlorambucil, mel hormone; parathyroid hormone, thyroXine, insulin, proinsu phalan, carmustine (BSNU), lomustine (CCNU), cyclophos lin, relaxin, prorelaxin, glycoprotein hormones Such as fol phamide, buSulfan, dibromomannitol, Streptozotocin, dacar licle stimulating hormone (FSH), thyroid stimulating hor bazine (DTIC), procarbazine, mitomycin C, cisplatin and mone (TSH) and luteinizing hormone (LH), hepatic growth other platinum derivatives, such as carboplatin), antibiotics factor; prostaglandin, fibroblast growth factor, prolactin, pla (such as dactinomycin (formerly actinomycin), bleomycin, cental lactogen, OB protein, tumour necrosis factor-C. and -B; daunorubicin (formerly daunomycin), doxorubicin, idarubi mullerian-inhibiting Substance, gonadotropin-associated cin, mithramycin, mitomycin, mitoxantrone, plicamycin, peptide, inhibin, activin, vascular endothelial growth factor, anthramycin (AMC)). integrin, thrombopoietin (TPO), nerve growth factors such as 0415 Examples of suitable angiogenesis inhibitors (anti NGF-B, platelet-growth factor, transforming growth factors angiogenic agents) include, but are not limited to, urokinase (TGFs) such as TGF-C. and TGF-B, insulin-like growth fac inhibitors, matrix metalloprotease inhibitors (such as marim tor-I or -II, erythropoietin (EPO), osteoinductive factors, astat, neovastat, BAY 12-9566, AG 3340, BMS-275291 and interferons such as interferon-C. -B, or -y, colony stimulating similar agents), inhibitors of endothelial cell migration and factors (CSFs) such as macrophage-CSF (M-CSF), granulo US 2012/0164068 A1 Jun. 28, 2012 29 cyte-macrophage-CSF (GM-CSF); and granulocyte-CSF dextrans, are commonly conjugated to proteins to increase (G-CSF), interleukins (ILS) such as IL-1, IL-1C, IL-2, IL-3, stability or size, etc., of the protein. IL-4, IL-5, IL-6, IL-7, IL-8, IL-9, IL-IO5 IL-I1, IL-12. IL-13, 0425 PEG, PEO or POE refers to an oligomer or polymer IL-14, IL-15, IL-16, IL-17, IL-18, IL-21 and LIF. of ethylene oxide. In the case of PEG, these oligomers or 0418 Chemokines generally act as chemoattractants to polymers are produced by, e.g., anionic ring opening poly recruit immune effector cells to the site of chemokine expres merization of ethylene oxide initiated by nucleophilic attack sion. Chemokines include, but are not limited to, RANTES, of a hydroxide ion on the epoxide ring. One of the more useful MCAF, M1P1-alpha or MIP1-Beta. The skilled artisan will forms of PEG for protein modification is monomethoxy PEG recognize that certain cytokines are also known to have (mPEG). chemoattractant effects and could also be classified under the 0426 Preferred PEGs are monodisperse or polydisperse, term chemokines. preferably monodisperse. The skilled artisan will be aware 0419 Exemplary serum albumin binding peptides or pro that PEG can be polydisperse or monodisperse. Polydisperse tein are described in US20060228.364 or US2O080260757. PEG comprises a mixture of PEGs having different molecular 0420 A variety of radionuclides are available for the pro weights. In the case of polydisperse PEGs, reference to a duction of radioconjugated proteins. Examples include, but specific molecular weight will be understood to refer to the are not limited to, low energy radioactive nuclei (e.g., Suitable number average molecular weight of PEGs in the mixture. for diagnostic purposes), such as 'C, N, H, °1, ''1, The size distribution is characterized statistically by its 'Tc, K, Fe, 'Ga, Ga, '''Inand the like. Preferably, the weight average molecular weight (MW) and its number aver radionuclide is a gamma, photon, or positron-emitting radio age molecular weight (Mn), the ratio of which is called the nuclide with a half-life suitable to permit activity or detection polydispersity index (Mw/Mn). MW and Mn are measured, after the elapsed time between administration and localiza in certain aspects, by mass spectroscopy. Most of the PEG tion to the imaging site. The present invention also encom protein conjugates, particularly those conjugated to PEG passes high energy radioactive nuclei (e.g., for therapeutic larger than 1 KD, exhibit a range of molecular weights due to purposes) such as 21, 1, 21, In, Rh, 'Sm,7Cu, a polydisperse nature of the parent PEG molecule. For Ga, Ho, ' Lu, Re and 'Re. These isotopes typi example, in case of mPEG2K (Sunbright ME-020HS, NOF), cally produce high energy C- or B-particles which have a short actual molecular masses are distributed over a range of path length. Such radionuclides kill cells to which they are in 1.5-3.0 KD with a polydispersity index of 1.036. close proximity, for example neoplastic cells to which the 0427 Based on the foregoing, the skilled artisan will be conjugate has attached or has entered. They have little or no aware that monodisperse PEG comprises a mixture of PEGs effect on non-localized cells and are essentially non-immu comprising Substantially the same molecular weight. Mono nogenic. Alternatively, high-energy isotopes may be gener disperse PEGs are commercially available, e.g., from Poly ated by thermal irradiation of an otherwise stable isotope, for example as in boron neutron-capture therapy (Guan et al., pure AS, Norway. 1998). 0428 The average or preferred molecular weight of the 0421. In another embodiment, the protein is conjugated to PEG will range from about 500 Da to about 200 kDa. For a “receptor' (such as streptavidin) for utilization in cell pre example, the molecular weight of the PEG is from about 1 to targeting wherein the conjugate is administered to the patient, about 100 kDa, from about 1.5 to about 50 kDa, from about followed by removal of unbound conjugate from the circula 1.5 to about 10 kDa, from about 1.5 kDa to about 5 kDa, from tion using a clearing agent and then administration of a about 1.5 kDa to about kDa, from about 1.5 to about 2 kDa. “ligand” (e.g., avidin) that is conjugated to atherapeutic agent 0429 Preferably, the PEG is monodisperse and has a (e.g., a radionucleotide). molecular weight of about 500 Da. Preferably, the PEG has a 0422 The proteins of the present invention can be modi molecular weight of about 1.5 kDa. Preferably, the PEG has a fied to contain additional nonproteinaceous moieties that are molecular weight of about 2 kDa. known in the art and readily available. Preferably, the moi 0430 Preferably, the PEG comprises a reactive group, eties suitable for derivatization of the protein are water such as a maleimide group. Preferably, the PEG is PEG soluble polymers. Non-limiting examples of water soluble maleimide. polymers include, but are not limited to, polyethylene glycol 0431. The physiologically acceptable polymer molecule (PEG), polyvinyl alcohol (PVA), copolymers of ethylene gly is not limited to a particular structure and is, in various col/propylene glycol, carboxymethylcellulose, dextran, poly aspects, linear (e.g. alkoxy PEG or bifunctional PEG), vinyl alcohol, polyvinyl pyrrolidone, poly-1,3-dioxolane, branched or multi-armed (e.g. forked PEG or PEG attached to poly-1,3,6-trioxane, ethylene/maleic anhydride copolymer, a polyol core), dentritic, or with degradable linkages. More polyaminoacids (either homopolymers or random copoly over, the internal structure of the polymer molecule is orga mers), and dextran or poly(n-vinyl pyrrolidone)polyethylene nized in any number of different patterns and is selected from glycol, propropylene glycol (PPG) homopolymers, prolypro the group consisting of homopolymer, alternating copolymer, pylene oxide/ethylene oxide co-polymers, polyoxyethylated random copolymer, block copolymer, alternating tripolymer, polyols (e.g., glycerol; POG), polyvinyl alcohol, and mix random tripolymer, and block tripolymer. tures thereof. Polyethylene glycol propionaldehyde may have 0432. The number of polymers attached to the protein may advantages in manufacturing due to its stability in water. vary, and if more than one polymer is attached, they can be the 0423. The polymer molecules are typically characterized same or different molecules. In general, the number and/or as having for example from about 2 to about 1000, or from type of polymers used for derivatization can be determined about 2 to about 300 repeating units. based on considerations including, but not limited to, the 0424 For example water-soluble polymers, including but particular properties or functions of the protein to be not limited to PEG, poly(ethylene oxide) (PEO), polyoxyeth improved, whether the protein derivative will be used in a ylene (POE), polyvinyl alcohols, hydroxyethyl celluloses, or therapy under defined conditions, etc. US 2012/0164068 A1 Jun. 28, 2012 30

0433. The skilled artisan will be aware that prior to con include those that are sensitive to tumour-associated pro jugation to a protein a polymer (e.g., PEG) may need to be teases such as Cathepsin B or plasmin. Cathepsin B cleavable activated by preparing a derivative having a functional group sites include the dipeptide sequences valine-citrulline and at one or both termini. phenylalanine-lysine. 0434 Particularly preferred compounds for conjugation to the protein of the present invention are set out in Table 1. Conjugation Methods Conjugation to Cysteine (Thiol) TABLE 1. 0438 Various methods are known in the art for conjugat Preferred compounds for coniugation ing a compound to a cysteine residue are known in the art and Group Detail will be apparent to the skilled artisan. Reagents for such conjugation typically bear reactive functionality which may Radioisotopes 123, 125, 1301, 1331, 135, 47Sc, 72 As, 72Sc, 90Y. 88y 97.Ru, (either directly 100Pd, 101mRh, 101mRh, 19Sb, 128Ba, 197Hg, 21 At react (i) directly with a cysteine thiol of a cysteine to form the or indirectly) 212 Bi, lism, 169Eu, 212Pb, 109 Pd, 11 In, 'Gu, 68Gu, labelled protein, (ii) with a linker reagent to form a linker 67Cu, 75Br, 76Br, 77Br, 99mTc, IIC, 13N, 15O, 18I. 188Rc, label intermediate, or (iii) with a linker protein to form the 203Pb, 64Cu, 105Rh, 198Au, 199Ag O 177Lu labelled protein. In the case of a linker several routes, employ Halflife Polyethylene glycol extenders Glycerol ing organic chemistry reactions, conditions, and reagents are Glucose known to those skilled in the art, including: (1) reaction of a Fluorescent Phycoerythrin (PE) cysteine group of the protein of the invention with a linker probes Allophycocyanin (APC) reagent, to form a protein-linker intermediate, via a covalent Alexa Fluor 488 Cy5.5 bond, followed by reaction with an activated compound; and Biologics Fluorescent proteins such as Renilla luciferase, GFP (2) reaction of a nucleophilic group of a compound with a mmune modulators linker reagent, to form compound-linker intermediate, via a Toxins covalent bond, followed by reaction with a cysteine group of An Immunoglobulin Halflife extenders such as albumin a protein of the invention. As will be apparent to the skilled Chemo- Taxol artisan from the foregoing, bifunctional linkers are useful in therapeutics 5-FU the present invention. For example, the bifunctional linker Doxorubicin comprises a thiol modification group for covalent linkage to darubicin the cysteine residue(s) and at least one attachment moiety (e.g., a second thiol modification moiety) for covalent or 0435. In one example of the invention, a spacer moiety is non-covalent linkage to the compound. included between the compound and the protein to which it is 0439 A variety of proteins and compounds, (and linkers) conjugated. The spacer moieties of the invention may be can be used to prepare a conjugate of the invention. Cysteine cleavable or non-cleavable. For example, the cleavable spacer thiol groups are nucleophilic and capable of reacting to form moiety is a redox-cleavable spacer moiety, such that the covalent bonds with electrophilic groups onlinker reagents or spacer moiety is cleavable in environments with a lower compound-linker intermediates or drugs including: (i) active redox potential. Such the cytoplasm and other regions with esters such as NHS esters, HOBt esters, haloformates, and higher concentrations of molecules with free sulfhydryl acid halides; (ii) alkyl and benzyl halides. Such as haloaceta groups. Examples of spacer moieties that may be cleaved due mides; (iii) aldehydes, ketones, carboxyl, and maleimide to a change in redox potential include those containing dis groups; and (iv) disulfides, including pyridyl disulfides, via ulfides. The cleaving stimulus can be provided upon intrac Sulfide exchange. Nucleophilic groups on a compound or ellular uptake of the conjugated protein where the lower linker include, but are not limited to amine, thiol, hydroxyl, redox potential of the cytoplasm facilitates cleavage of the hydrazide, oxime, hydrazine, thiosemicarbazone, hydrazine spacer moiety. carboxylate, and arylhydrazide groups capable of reacting to 0436. In another example, a decrease in pH causes cleav form covalent bonds with electrophilic groups on linker moi age of the spacer to thereby release of the compound into a eties and linker reagents. target cell. A decrease in pH is implicated in many physiologi 0440 Preferred labelling reagents include maleimide, cal and pathological processes, such as endoSome trafficking, haloacetyl, iodoacetamide Succinimidyl ester, isothiocyan tumour growth, inflammation, and myocardial ischemia. The ate, sulfonyl chloride, 2,6-dichlorotriazinyl, pentafluorophe pH drops from a physiological 7.4 to 5-6 in endosomes or 4-5 nyl ester, and phosphoramidite, although other functional in lysosomes. Examples of acid sensitive spacer moieties groups can also be used. which may be used to target lysosomes or endosomes of 0441 Maytansine may, for example, be converted to May cancer cells, include those with acid-cleavable bonds such as SSCH, which can be reduced to the free thiol, May-SH, and those found in acetals, ketals, orthoesters, hydrazones, trityls, reacted with a protein of the invention (Chari et al., 1992) to cis-aconityls, or thiocarbamoyls (see for example, U.S. Pat. generate a maytansinoid-immunoconjugate with a disulfide Nos. 4,569,789, 4,631,190, 5,306,809, and 5,665,358). Other linker. Maytansinoid conjugates with disulfide linkers have exemplary acid-sensitive spacer moieties comprise dipeptide been reported (WO 04/016801; U.S. Pat. No. 6,884,874; and sequences Phe-Lys and Val-Lys. WO 03/068144). The disulfide linker SPP is constructed with 0437 Cleavable spacer moieties may be sensitive to bio linker reagent N-Succinimidyl 4-(2-pyridylthio)pentanoate. logically supplied cleaving agents that are associated with a 0442 Another exemplary reactive functional group is particular target cell, for example, lysosomal or tumour-as N-hydroxysuccinimidyl ester (NHS) of a carboxyl group sub Sociated enzymes. Examples of linking moieties that can be stituent of a compound, e.g. biotin or a fluorescent dye or a cleaved enzymatically include, but are not limited to, peptides toxin or a protein. The NHS ester of the compound may be and esters. Exemplary enzyme cleavable linking moieties preformed, isolated, purified, and/or characterized, or it may US 2012/0164068 A1 Jun. 28, 2012 beformed in situ and reacted with a nucleophilic group of the 0445 Prior to linkage it is preferred that the protein of the protein. Typically, the carboxyl form of the compound is invention is made reactive for conjugation with linker activated by reacting with some combination of a carbodiim reagents by treatment with a reducing agent such as DTT ide reagent, e.g. dicyclohexylcarbodiimide, diisopropylcar (Cleland's reagent, dithiothreitol) or TCEP (tris(2-carboxy bodiimide, or auronium reagent, e.g. TSTU (O-(N-Succin ethyl)phosphine hydrochloride; Getz et al., 1999; Soltec Ven imidyl)-N.N.N',N'-tetramethyluronium tetrafluoroborate, tures, Beverly, Mass.). Disulfide bonds can be re-established HBTU (O-benzotriazol-1-yl)-N.N.N',N'-tetramethyluro between cysteine residues that are not required for linkage nium hexafluorophosphate), or HATU (O-(7-azabenzotria with dilute (200 nM) aqueous copper sulfate (CuSO) at room Zol-1-yl)-N.N.N',N'-tetramethyluronium hexafluorophos temperature. Other oxidants, i.e. oxidizing agents, and oxi phate), an activator, such as 1-hydroxybenzotriazole (HOBt), dizing conditions, which are known in the art may be used. Ambient air oxidation is also effective. This mild, partial and N-hydroxysuccinimide to give the NHS ester of the com reoxidation step forms intrachain disulfides efficiently with pound. In some cases, the compound and the protein, may be high fidelity. coupled by in situ activation of the compound and reaction with the protein to form the conjugate in one step. Other Conjugation to Threonine/Serine activating and coupling reagents include TBTU (2-(1H-ben Zotriazo-1-yl)-1-1,3,3-tetramethyluronium hexafluorophos 0446 Methods are also known in the art for conjugating a phate), TFFH(N,N',N',N'-tetramethyluronium 2-fluoro compound to a threonine or serine residue. For example, hexafluorophosphate), PyBOP (benzotriazole-1-yl-oxy-tris Zhang and Tam (1996) described a method in which carbonyl pyrrolidino-phosphonium hexafluorophosphate, EEDQ precursors are derived from the 1,2-aminoalcohols of serine (2-ethoxy-1-ethoxycarbonyl-1,2-dihydro-quinoline), DCC or threonine, which can be selectively and rapidly converted (dicyclohexylcarbodiimide); DIPCDI (diisopropylcarbodi to aldehyde form by periodate oxidation. Reaction of the imide), MSNT (1-(mesitylene-2-sulfonyl)-3-nitro-1H-1,2,4- aldehyde with a 1,2-aminothiol of cysteine in a compound to be attached to a protein of the invention forms a stable thia triazole, and aryl Sulfonyl halides, e.g. triisopropylbenzene Zolidine product. This method is particularly useful for label sulfonyl chloride. ling proteins at N-terminal serine or threonine residues. 0443) Additional conjugation methods include, for example, the use of maleimides, iodoacetimides or halo PEGylation Methods acetyl/alkyl halides, aziridine, acryloyl derivatives to react with the thiol of a cysteine to produce a thioeter that is reactive 0447 Various methods are known in the art for conjugat with a compound (e.g., Schelte et al., 2000 (use of maleim ing compounds, e.g., PEG, to a protein. In the case of PEG, ides); Reddy et al., 1988 (use of maleimide derivatives); Ram the molecule can be activated to facilitate its binding to seier and Chang, 1994 (use of iodacetamide); Eisen et al., amines or imidazoles, a carboxylic group, a hydroxyl group 1953 (use of 2,4-dinitrobenzeneulfonic acid); Grossman et or a sulfhydryl group. al., 1981 (use of aziridine); orYemet al., 1992 (use of acryloyl 0448. For example, Abuchowski et al (1977) activated derivatives). Disulphide exchange of a free thiol with an PEG using cyanuric chloride to produce a PEG dichlorotri activated piridyldisulphide is also useful for producing a con azine derivative. This derivative can react with multiple func jugate (King et al., 1978 and references cited therein, e.g., use tional nucleophilic functional groups, such as lysine, serine, of 5-thio-2-nitrobenzoic (TNB) acid). Preferably, a maleim tyrosine, cysteine and histidine. A modified form of this pro ide is used. tocol produced PEG-chlorotriazine, which has lower reactiv 0444 With respect to the use of radiolabelled conjugates, ity and conjugates more selectively with lysine or cysteine proteins of the invention may be directly labelled (such as residues (Mutsushima et al., 1980). through iodination) or may be labelled indirectly through the 0449. Two widely used forms of PEG used to conjugate to use of a chelating agent. As used herein, the phrases “indirect proteins are succinimidyl carbonate PEG (SC-PEG; Zalipsky labelling” and “indirect labelling approach both mean that a et al., 1992) and benzotriazole carbonate PEG (BTC-PEG: chelating agent is covalently attached to a protein and at least U.S. Pat. No. 5,560,234). Both of these compounds react one radionuclide is associated with the chelating agent. Such preferentially with lysine residues to form carbamate link chelating agents are typically referred to as bifunctional ages, however are also known to react with hystidine and chelating agents as they bind both the protein and the radio tyrosine. SC-PEG is slightly more resistant to hydrolysis than isotope. Exemplary chelating agents comprise 1-isothiocyc BTC-PEG. matobenzyl-3-methyldiothelene triaminepentaacetic acid 0450 Another PEG useful for conjugating to proteins is (“MX-DTPA) and cyclohexyl diethylenetriamine pentaace PEG-propionaldehyde (U.S. Pat. No. 5.252.714). An advan tic acid (“CHX-DTPA) derivatives, or DOTA. Linker tage of this chemistry is that under acidic conditions (about reagents such as DOTA-maleimide (4-maleimidobutyrami pH5) it is largely selective for N-terminal O-amine thus avoid dobenzyl-DOTA) can be prepared by the reaction of ami ing potential problems with non-specific conjugation. A nobenzyl-DOTA with A-maleimidobutyric acid (Fluka) acti acetal derivative of PEG-propionaldehyde, i.e., PEG-acetala vated with isopropylchloroformate (Aldrich), following the ldehyde provides an additional benefit in so far as it provides procedure of Axworthy et al. (2000). DOTA-maleimide for longer storage than PEG-propionaldehyde (U.S. Pat. No. reagents react with free cysteine amino acids of the proteins 5.990.237). of the invention and provide a metal complexing ligand 0451 Active esters of PEG carboxylic acids are probably thereon (Lewis et al., 1998). Chelating linker labelling one of the most used acylating agents for protein conjugation. reagents such as DOTA-NHS (1,4,7,10-tetraazacyclodode Active esters react with primary amines near physiological cane-1,4,7,10-tetraacetic acid mono (N-hydroxySuccinimide conditions to form stable amides. Activation of PEG-car ester) are commercially available (Macrocyclics, Dallas, boxylic acids to Succinimidyl active esters is accomplished by Tex.). reacting the PEG-carboxylic acid with N-hydroxysuccinim US 2012/0164068 A1 Jun. 28, 2012 32 ide (NHS or HOSu) and a carbodiimide. Exemplary carboxy disease or disorder (i.e., a condition) e.g., associated with or lic acid derivatives of PEG include carboxymethylated PEG expressed by a cancer or cancerous/transformed cell and/or (CM-PEG; Zalipsky et al., 1990), butanoic acid derivatives associated with an autoimmune disease and/or associated and propionic acid derivatives (U.S. Pat. No. 5,672,662). with an inflammatory disease or condition and/or associated Changing the distance between the active ester and the PEG with a neurodegenerative disease and/or associated with an backbone by the addition of methylene units can dramatically immune-deficiency disorder. influence reactivity towards water and amines (e.g., by reduc ing hydrolysis). Alternatively or in addition, hydrolysis can 0461 Exemplary antigens against which a protein of the be reduced by introducing an C.-branching moiety to the car invention can be produced include BMPR1B (bone morpho boxylic acid. genetic protein receptor-type IB, Dijke. et al 1994, 0452 PEGylation of free cysteine residues in a protein is WO2004063362); E16 (LAT1, SLC7A5, WO2004.048938); useful for site-specific conjugation (e.g., using a protein STEAP1 (six transmembrane epithelial antigen of prostate; modified to include cysteine residues as described herein). WO2004065577); CA125 (MUC16, WO2004.045553); MPF Exemplary PEG derivatives for cysteine conjugation include (MSLN, SMR, megakaryocyte potentiating factor, mesothe PEG-maleimide, PEG-vinylsulfone, PEG-iodoacetamide lin, WO2003101283): Napi3b (NAPI-3B, NPTIIb, and PEG-orthopyridyl disulfide. Exemplary methods for con SLC34A2, solute carrier family 34: WO2004022778); Sema jugating PEG to cysteine residues are described in Goodson 5b (F1110372, KIAA1445, SEMA5B, SEMAG, Semaphorin and Katre (1990) and/or above. Exemplary methods for con 5b, sema domain, seven thrombospondin repeats (type 1 and jugation using PEG-vinylsulfone are described, for example, type Hike), transmembrane domain (TM) and short cytoplas mic domain, (semaphorin) 5B, WO2004000997); PSCA in Li et al. (2006). (US2003129.192); ETBR (Endothelin type B receptor, 0453 U.S. Pat. No. 5,985,263 describes methods for con WO2004.045516); MSG783 (RNF124, WO2003104275); jugating PEG to the secondary amine group of histidine, STEAP2 (HGNC 8639, IPCA-I, PCANAP1, STAMP1, which has a lower pKa than the primary amine. An advantage STEAP2, STMP prostate cancer associated gene 1, prostate of this approach is that the acyl-histidine bond is not stable cancer associated protein 1, six transmembrane epithelial meaning that the protein is slowly released (i.e., the conjugate antigen of prostate 2, six transmembrane prostate protein, behaves as a slow release formulation or a pro-drug). WO2003087306); TrpM4 (BR224.50, F1120041, TRPM4, 0454. Another approach for PEGylation is to take advan TRPM4B, transient receptor potential cation channel, sub tage of a N-terminal serine or threonine, which can be con family M, member 4, US2003143557); CRIPTO (CR, CR1, verted to periodate as discussed above. Using this approach, CRGF, CRIPTO, TDGF1, teratocarcinoma-derived growth PEG has been conjugated to bioactive proteins (e.g., Gaertner factor, US2003224411); CD21 (CR2 (Complement receptor and Offord, 1996). T) or C3DR (C3d/Epstein Barr virus receptor) 0455 PEG can also be conjugated to carbohydrate groups. WO2004.045520); CD79b (CD79B, CD79B, IGb (immuno 0456. The present invention also encompasses the use of globulin-associated beta), B29, WO2004016225); FcRH2 reversible PEGylation strategies. (DFGP4, IRTA4, SPAP1A (SH2 domain containing phos phatase anchor protein Ia), SPAP1B, SPAP1C, Uses WO2004016225); HER2 (ErbB2, WO2004.048938): NCA 0457. The proteins of the present invention are useful in a (CEACAM6, WO2004063709); MDP (DPEP1, variety of applications, including research, diagnostic and WO2003016475): IL20RC (IL20Ra, ZCYTOR7, therapeutic applications. Depending on the antigen to which EP1394274); Brevican (BCAN, BEHAB, US2003 186372); the protein binds it may be useful for delivering a compound EphB2R (DRT, ERK, Hek5, EPHT3, Tyro5, to a cell, e.g., to kill the cell or prevent growth and/or for WO2003042661); ASLG659 (B7h, US20040101899); imaging and/or for in vitro assays. In one example, the protein PSCA (Prostate stem cell antigen precursor, is useful for both imaging and delivering a cytotoxic agent to WO2004022709); GEDA (lipoma HMGIC fusion-partner a cell, i.e., it is conjugated to a detectable label and a cytotoxic like protein WO2003054152); BAFF-R (B cell-activating agent or a composition comprises a mixture of proteins some factor receptor, BLyS receptor 3, BR3, WO20040583.09): of which are conjugated to a cytotoxic agent and some of CD22 (B-cell receptor CD22-B isoform, BL-CAM, Lyb-8, which are conjugated to a detectable label. Lyb8, SIGLEC-2, FLJ22814, WO2003072036), CD79a 0458. The proteins described herein can also act as inhibi (CD79A, CD79a, immunoglobulin-associated alpha, a B tors to inhibit (which can be reducing or preventing) (a) cell-specific protein that covalently interacts with Ig beta binding (e.g., of a ligand, an inhibitor) to a receptor, (b) a (CD79B) and forms a complex on the surface with IgM receptor signaling function, and/or (c) a stimulatory function. molecules, transduces a signal involved in B-cell differentia Proteins which act as inhibitors of receptor function can block tion: WO2003088808); CXCR5 (Burkitt's lymphoma recep ligand binding directly or indirectly (e.g., by causing a con tor 1, a G protein-coupled receptor that is activated by the formational change). CXCL13 chemokine, functions in lymphocyte migration and humoral defense, plays a role in HIV-2 infection and perhaps Antigens development of AIDS, lymphoma, myeloma, and leukemia WO2004040000); HLA-DOB (Beta subunit of MHC class II 0459. The present invention contemplates a protein com molecule (Iaantigen) that binds peptides and presents them to prising at least one variable region comprising at least two CD4+ T lymphocytes; WO9958658); P2X5 (Purinergic cysteine residues in FR1 capable of specifically binding to receptor P2X ligand-gated ion channel 5, an ion channel any antigenCs), i.e., an example of the invention is generic as gated by extracellular ATP may be involved in synaptic trans opposed to requiring a specific antigen. mission and neurogenesis, deficiency may contribute to the 0460) Examples of the present invention contemplate a pathophysiology of idiopathic detrusor instability; protein that specifically binds to an antigen associated with a WO2004.047749); CD72 (B-cell differentiation antigen US 2012/0164068 A1 Jun. 28, 2012

CD72, Lyb-2: WO2004.042346); LY64 (Lymphocyte antigen variety of unit dosage forms depending upon the method of 64 (RP 105), type I membrane protein of the leucine rich administration. For example, unit dosage forms suitable for repeat (LRR) family, regulates B-cell activation and apopto oral administration include powder, tablets, pills, capsules sis, loss of function is associated with increased disease activ and lozenges or by parenteral administration. It is recognized ity in patients with systemic lupus erythematosis: that the pharmaceutical compositions of this invention, when US2002 193567); FcRH1 (Fc receptor-like protein 1, a puta administered orally, should be protected from digestion. This tive receptor for the immunoglobulin Fc domain that contains is typically accomplished either by complexing the proteins C2 type Ig-like and ITAM domains, may have a role in B-lym with a composition to render it resistant to acidic and enzy phocyte differentiation WO2003077836); IRTA2 (Immuno matic hydrolysis or by packaging the compound in an appro globulin Superfamily receptor translocation associated 2, a putative immunoreceptor with possible roles in B cell devel priately resistant carrier Such as a liposome. Means of pro opment and lymphomagenesis; deregulation of the gene by tecting proteins from digestion are known in the art. translocation occurs in some B cell malignancies; 0466 Typically, a therapeutically effective amount of the WO2003077836); TENB2 (TMEFF2, tomoregulin, TPEF, protein will beformulated into a composition for administra HPP1, TR, putative transmembrane proteoglycan, related to tion to a subject. The phrase “a therapeutically effective the EGF/heregulin family of growth factors and follistatin: amount refers to an amount Sufficient to promote, induce, WO2004074320): CD20 (WO94/11026); VEGF-A (Presta et and/or enhance treatment or other therapeutic effect in a al., 1997); p53: EGFR: progesterone receptor; cathepsin D; Subject. As will be apparent, the concentration of proteins of Bcl-2: E cadherin; CEA, Lewis X: Ki67, PCNA; CD3: CD4; the present invention in these formulations can vary widely, CD5; CD7: CD11c: CD11d; c-Myc; tau: PrPSC; or AB. and will be selected primarily based on fluid volumes, vis 0462 Preferably, the protein of the invention specifically cosities, body weight and the like in accordance with the binds to Her2 (e.g., comprising a sequence set forthin SEQID particular mode of administration selected and the patient's NO: 70), MUC1 (e.g., comprising a sequence set forth in SEQ needs. Depending on the type and severity of the disease, a ID NO: 72 or 73), TAG72 (a high molecular weight mucin therapeutically effective amount may be about 1 lug/kg to 15 like protein e.g., as described in Johnson et al., 1986) or mg/kg (e.g. 0.1-20 mg/kg) of molecule, whether, for example, PSMA (e.g., comprising a sequence set forth in SEQID NO: 71). For example, the protein of the invention specifically by one or more separate administrations, or by continuous binds to Her2. For example, the protein of the invention infusion. A typical daily dosage might range from about 1 specifically binds to MUC1. For example, the protein of the ug/kg to 100 mg/kg or more. An exemplary dosage of the invention specifically binds to TAG72. For example, the pro protein to be administered to a patient is in the range of about tein of the invention specifically binds to PSMA. 0.1 to about 10 mg/kg of patient weight. For repeated admin 0463. Other exemplary antibodies from which a protein of istrations over several days or longer, depending on the con the invention can be derived will be apparent to the skilled dition, the treatment is Sustained until a desired Suppression artisan and include, for example, Rituximab (C2B8; WO94/ of disease symptoms occurs. An exemplary dosing regimen 11026); or bevacizumab (humanized A.4.6.1; Presta et al., comprises administering an initial loading dose of about 4 1997)). mg/kg, followed by a weekly maintenance dose of about 2 0464 Exemplary bispecific proteins may bind to two dif mg/kg of the protein. Other dosage regimens may be useful. ferent epitopes of the antigen of interest. Other such proteins The progress of this therapy is easily monitored by conven may combine one antigen binding site with a binding site for tional techniques and assays. another protein. Alternatively, an anti-antigen of interest 0467 Alternatively, the protein of the invention is formu region may be combined with a region which binds to a lated at a concentrated does that is diluted to a therapeutically triggering molecule on a leukocyte Such as a T-cell receptor effective dose prior to administration to a subject. molecule (e.g., CD3), or Fc receptors for IgG (FcyR), such as FcyRI (CD64), FcyRII (CD32) and/or FcyRIII (CD16), so as 0468. The pharmaceutical compositions of this invention to focus and localize cellular defense mechanisms to the cells are particularly useful for parenteral administration, e.g., for expressing the antigen of interest. Bispecific proteins may mulated for injection via the intravenous, intramuscular, Sub also be used to localize cytotoxic agents to cells which cutaneous, transdermal, or other Such routes, including peri express the antigen of interest. These proteins possess a staltic administration and direct instillation into a tumour or region that binds the antigen of interest and a region which disease site (intracavity administration). The compositions binds the cytotoxic agent (e.g., saporin, anti-interferon-C. for administration will commonly comprise a solution of the Vinca alkaloid, ricin A chain, methotrexate or radioactive proteins of the present invention dissolved in a pharmaceuti isotope hapten). WO 96/16673 describes a bispecific anti cally acceptable carrier, preferably an aqueous carrier. A vari ErbB2/anti-FcyRIII antibody and U.S. Pat. No. 5,837,234 ety of aqueous carriers can be used, e.g., buffered saline and discloses a bispecific anti-ErbB2/anti-FcyRI antibody. A the like. Other exemplary carriers include water, Saline, Ring er's Solution, dextrose solution, and 5% human serum albu bispecific anti-ErbB2/FcC. antibody is shown in WO98/ min. Nonacqueous vehicles Such as mixed oils and ethyl oleate 02463. U.S. Pat. No. 5,821,337 teaches a bispecific anti may also be used. Liposomes may also be used as carriers. ErbB2/anti-CD3 antibody. The vehicles may contain minor amounts of additives that enhance isotonicity and chemical stability, e.g., buffers and Pharmaceutical Compositions and Methods of Treatment preservatives. The compositions may contain pharmaceuti 0465. The proteins of the present invention (syn. active cally acceptable auxiliary Substances as required to approxi ingredients) are useful for parenteral, topical, oral, or local mate physiological conditions such as pH adjusting and buff administration, aerosol administration, or transdermal ering agents, toxicity adjusting agents and the like, for administration for prophylactic or for therapeutic treatment. example, Sodium acetate, sodium chloride, potassium chlo The pharmaceutical compositions can be administered in a ride, calcium chloride, Sodium lactate and the like. US 2012/0164068 A1 Jun. 28, 2012 34

0469 Techniques for preparing pharmaceutical composi Sufficient to reduce or eliminate at least one symptom of a tions are generally known in the art as exemplified by Rem specified disease or condition. ington's Pharmaceutical Sciences, 16th Ed. Mack Publishing 0477 As used herein, the term “subject' shall be taken to Company, 1980. mean any animal including humans, preferably a mammal. 0470 WO2002/080967 describes compositions and Exemplary Subjects include but are not limited to humans, methods for administering aerosolized compositions com primates, livestock (e.g. sheep, cows, horses, donkeys, pigs), prising proteins for the treatment of, e.g., asthma, which are companion animals (e.g. dogs, cats), laboratory test animals also suitable for administration of protein of the present (e.g. mice, rabbits, rats, guinea pigs, hamsters), captive wild invention. animals (e.g. fox, deer). Preferably the mammal is a human or 0471) Suitable dosages of compounds of the present primate. More preferably the mammal is a human. invention will vary depending on the specific protein, the 0478 As used herein, a “condition' is a disruption of or condition to be diagnosed/treated/prevented and/or the sub interference with normal function, and is not to be limited to ject being treated. It is within the ability of a skilled physician any specific condition, and will include diseases or disorders. to determine a suitable dosage, e.g., by commencing with a In an example, the condition is a cancer or an immunopatho Sub-optimal dosage and incrementally modifying the dosage logical disorder. to determine an optimal or useful dosage. Alternatively, to 0479. Exemplary cancers include, but are not limited to, determine an appropriate dosage for treatment/prophylaxis, carcinoma, lymphoma, blastoma, sarcoma, and leukemia or data from cell culture assays or animal studies are used, lymphoid malignancies. More particular examples of Such wherein a Suitable dose is within a range of circulating con cancers include squamous cell cancer (e.g. epithelial squa centrations that include the ED50 of the active compound mous cell cancer), lung cancer including Small-cell lung can with little or no toxicity. The dosage may vary within this cer, non-small cell lung cancer, adenocarcinoma of the lung range depending upon the dosage form employed and the and squamous carcinoma of the lung, cancer of the perito route of administration utilized. A therapeutically/prophylac neum, hepatocellular cancer, gastric or stomach cancer tically effective dose can be estimated initially from cell including gastrointestinal cancer, pancreatic cancer, glioblas culture assays. A dose may beformulated in animal models to toma, cervical cancer, ovarian cancer, liver cancer, bladder achieve a circulating plasma concentration range that cancer, hepatoma, breast cancer, colon cancer, rectal cancer, includes the IC50 (i.e., the concentration of the compound colorectal cancer, endometrial or uterine carcinoma, salivary which achieves a half-maximal inhibition of symptoms) as gland carcinoma, kidney or renal cancer, prostate cancer, determined in cell culture. Such information can be used to Vulval cancer, thyroid cancer, hepatic carcinoma, anal carci more accurately determine useful doses in humans. Levels in noma, penile carcinoma, as well as head and neck cancer. plasma maybe measured, for example, by high performance Preferably a cancer is breast cancer or ovarian cancer or liquid chromatography. prostate cancer. 0472. A protein of the invention may be combined in a 0480. In one example of the invention, the cancer pharmaceutical combination formulation, or dosing regimen expresses Her2. Exemplary cancers include breast cancer, as combination therapy, with a second compound. The second ovarian cancer, stomach cancer or uterine cancer, preferably compound of the pharmaceutical combination formulation or breast cancer. Such a cancer can be treated, for example, with dosing regimen preferably has complementary activities to a protein of the invention that binds to Her2. the protein of the combination such that they do not adversely 0481. In another example of the invention, the cancer affect each other. expresses PSMA. Exemplary cancers include prostate cancer. 0473. The second compound may be a chemotherapeutic Such a cancer can be treated, for example, with a protein of agent, cytotoxic agent, cytokine, growth inhibitory agent, the invention that binds to PSMA. anti-hormonal agent, and/or cardioprotectant. Such mol 0482 In a further example of the invention, the cancer ecules are suitably present in combination in amounts that are expresses Tag72. Exemplary cancers include carcinomas, effective for the purpose intended. A pharmaceutical compo Such as colorectal cancer, gastric cancer, pancreatic cancer, sition containing a protein of the invention may also have a ovarian cancer, endometrial cancer, breast cancer, non-Small therapeutically effective amount of a chemotherapeutic agent cell lung cancer, and prostate cancer. Such a cancer can be Such as a tubulin-forming inhibitor, a topoisomerase inhibi treated, for example, with a protein of the invention that binds tor, or a DNA binder. Pharmaceutical “slow release' capsules to Tag72. or compositions may also be used. Slow release formulations 0483. In a further example of the invention, the cancer are generally designed to give a constant drug level over an expresses MUC1, preferably a glycoform of MUC1 associ extended period and may be used to deliver compounds of the ated with cancer. Exemplary cancers include carcinomas, present invention. Such as colorectal cancer, gastric cancer, pancreatic cancer, 0474 The present invention also provides a method of breast cancer, lung cancer, and bladder cancer. Such a cancer treating or preventing a condition in a Subject, the method can be treated, for example, with a protein of the invention comprising administering a therapeutically effective amount that binds to MUC1. of a protein of the invention to a subject in need thereof. 0484 Immunopathology is the study of disease having an 0475. As used herein, the terms “preventing”, “prevent' or immunological cause and immunologic disease is any condi “prevention' in the context of preventing a condition include tion caused by the reactions of immunoglobulins to antigens. administering an amount of a protein described herein suffi Thus, an “immunopathological disorder can be defined as a cient to stop or hinder the development of at least one symp disorder arising from reaction of a Subject's immune system tom of a specified disease or condition. to antigens. Immunopathological disorders include autoim 0476. As used herein, the terms “treating”, “treat' or mune diseases and hyperSensitivity responses (e.g. Type I: “treatment include administering a therapeutically effective anaphylaxis, hives, food allergies, asthma; Type II: autoim amount of an inhibitor(s) and/or agent(s) described herein mune haemolytic anemia, blood transfusion reactions; Type US 2012/0164068 A1 Jun. 28, 2012

III: Serum sickness, necrotizing vasculitis, glomerulonephri order, wherein preferably there is a time period while both (or tis, rheumatoid arthritis, lupus; Type IV: contact dermatitis, all) active agents simultaneously exert their biological activi graft rejection). Autoimmune diseases include rheumatologic ties. disorders (such as, for example, rheumatoid arthritis, 0488 Prior to therapeutic use, a protein of the invention is Sjogren's syndrome, Scleroderma, lupus such as SLE and preferably tested in vitro and/or in vivo, e.g., as described lupus nephritis, polymyositisfdermatomyositis, cryoglobu below. linemia, anti-phospholipid antibody syndrome, and psoriatic arthritis), osteoarthritis, autoimmune gastrointestinal and In Vitro Testing liver disorders (such as, for example, inflammatory bowel diseases (e.g., ulcerative colitis and Crohn's disease), autoim 0489. In one example, a protein of the invention binds to mune gastritis and pernicious anemia, autoimmune hepatitis, an antigen, even if conjugated to a compound. The protein primary biliary cirrhosis, primary Sclerosing cholangitis, and may bind to the antigen at least as well as the protein from celiac disease), vasculitis (such as, for example, ANCA-as which it is derived. Alternatively, the protein or conjugate Sociated vasculitis, including Churg-Strauss vasculitis, comprising same binds to the antigen with at least about 10% Wegener's granulomatosis, and polyarteritis), autoimmune or 20% or 30% or 40% or 50% or 60% or 70% or 80% or 90% neurological disorders (such as, for example, multiple scle of the affinity or avidity of the protein from which it is derived rosis, opSoclonus myoclonus syndrome, myasthenia gravis, or a form of the protein lacking the cysteine residues and/or neuromyelitis optica, and autoimmune polyneuropathies), not conjugated to the compound. renal disorders (such as, for example, glomerulonephritis, 0490 Exemplary methods for determining binding affin Goodpasture's syndrome, and Berger's disease), autoimmune ity of a protein include a simple immunoassay showing the dermatologic disorders (such as, for example, psoriasis, urti ability of the protein to block the binding of the unmodified caria, hives, pemphigus Vulgaris, bullous pemphigoid, and protein or unconjugated protein to a target antigen, e.g., a cutaneous lupus erythematosus), hematologic disorders competitive binding assay. Competitive binding is deter (such as, for example, thrombocytopenic purpura, thrombotic mined in an assay in which the protein under test inhibits thrombocytopenic purpura, post-transfusion purpura, and specific binding of a reference protein to a common antigen. autoimmune hemolytic anemia), atherosclerosis, uveitis, Numerous types of competitive binding assays are known, for autoimmune hearing diseases (such as, for example, inner ear example, Solid phase direct or indirect radioimmunoassay disease and hearing loss), Behcet's disease, Raynaud's Syn (RIA), Solid phase direct or indirect enzyme immunoassay drome, organ transplant, and autoimmune endocrine disor (EIA), sandwich competition assay; solid phase direct biotin ders (such as, for example, diabetic-related autoimmune dis avidin EIA: solid phase direct labelled assay, solid phase eases such as insulin-dependent diabetes mellitus (IDDM), direct labelled sandwich assay; solid phase direct label RIA Addison's disease, and autoimmune thyroid disease (e.g., using 'I label; solid phase direct biotin-avidin EIA; or direct Graves disease and thyroiditis)). More preferred such dis labelled RIA (see, for example, Harlow and Lane, 1988). eases include, for example, rheumatoid arthritis, ulcerative Typically, such an assay involves the use of purified antigen colitis, ANCA-associated vasculitis, lupus, multiple Sclero bound to a solid surface or cells bearing either of these, an sis, Sjogren's syndrome, Graves disease, IDDM, pernicious unlabelled test protein and a labelled reference protein. Com anemia, thyroiditis, and glomerulonephritis. petitive inhibition is measured by determining the amount of 0485. In another embodiment, the disorder is an inflam label bound to the solid surface or cells in the presence of the matory disease. Inflammation is a protective response of body test protein tissues to irritation or injury- and can be acute or chronic. 0491. The present invention also encompasses methods Thus, inflammatory disorders include diseases involving neu for testing the activity of a protein of the invention. Various trophils, monocytes, mast cells, basophils, eosinophils, mac assays are available to assess the activity of a protein of the rophages where cytokine release, histamine release, oxidative present invention in vitro. For example, a protein of the burst, phagocytosis, release of other granule enzymes and present invention is administered to a cell or population chemotaxis occur. Hypersensitivity responses (defined above thereof to determine whether or not it can bind to said cell under immunopathological disorders) can also be regarded as and/or be internalized by said cell. Such an assay is facilitated inflammatory diseases (acute or chronic) since they often by labelling the protein of the present invention with a detect involve complementactivation and recruitment/infiltration of able label (i.e., producing a conjugate), however this is not various leukocytes Such as neutrophils, mast cells, basophils, essential since the protein of the present invention can also be etc. detected with a labelled protein. Such an assay is useful for 0486 The compositions of the present invention will be assessing the ability of a protein of the present invention to administered in a manner compatible with the dosage formu deliver a compound (i.e., a payload) to a cell and/or its utility lation and in Such amount as is therapeutically/prophylacti in imaging. Preferably the cell expresses an antigen to which cally effective. Formulations are easily administered in a the protein of the present invention binds and more preferably variety of manners, e.g., by ingestion or injection or inhala is a cell line or primary cell culture of a cell type that it desired tion. to be detected or treated. 0487. Other therapeutic regimens may be combined with 0492 Generally, the cytotoxic or cytostatic activity of a the administration of a protein of the invention. The combi protein of the present invention, e.g. conjugated to a cytotoxic nation therapy may be administered as a simultaneous or molecule is measured by: exposing cells expressing an anti sequential regimen. When administered sequentially, the gen to which the protein of the present invention binds to the combination may be administered in two or more adminis protein of the present invention; culturing the cells for a trations. The combined administration includes co-adminis Suitable period for the protein to exert a biological effect, e.g., tration, using separate formulations or a single pharmaceuti from about 6 hours to about 5 days; and measuring cell cal formulation, and consecutive administration in either viability, cytotoxicity and/or cell death. Cell-based in vitro US 2012/0164068 A1 Jun. 28, 2012 36 assays useful for measure viability (proliferation), cytotoxic to the protein. This permits determination of the in vivo ity, and cell death are known in the art. stability of the protein of the present invention. 0493 For example, the CellTiter-Glo R. Luminescent Cell 0497. A protein of the present invention can also be Viability Assay is a commercially available (Promega Corp., administered to an animal model of a human disease. The Madison, Wis.), homogeneous assay method based on the skilled artisan will be readily able to determine a suitable recombinant expression of Coleoptera luciferase (U.S. Pat. model based on the antigen to which the protein of the present Nos. 5,583,024; 5,674,713 and 5,700,670). This cell prolif invention binds. Exemplary models of, for example, human eration assay determines the number of viable cells in culture cancer are known in the art. For example, mouse models of based on quantitation of the ATP present in a cell, an indicator breast cancer include mice overexpressing fibroblast growth of metabolically active cells (U.S. Pat. No. 6,602,677). Alter factor 3 (Mulleret al., 1990); TGF-alpha (Matsuiet al., 1990); natively, cell viability is assayed using non-fluorescent resa erbB2 (Guy, et al., 1992); RET-1 (Iwamoto et al., 1990) or Zurin, which is added to cells cultured in the presence of a transplantation of human breast cancer cells into SCID mice. protein of the present invention. Viable cells reduce resaZurin Models of ovarian cancer include transplantation of ovarian to red-fluorescent resorufin, easily detectable, using, for cancer cells into mice (e.g., as described in Roby et al., 2000); example microscopy or a fluorescent plate reader. Kits for transgenic mice chronically secreting luteinising hormone analysis of cell viability are available, for example, from (Risma et al., 1995); or WX/Wv mice. Mouse models of Molecular Probes, Eugene, Oreg., USA. Other assays for cell prostate cancer are also known in the art and include, for viability include determining incorporation of H-thymidine example, models resulting from enforced expression of SV40 or 'C-thymidine into DNA as it is synthesized is an assay for early genes (e.g., the TRAMP model that utilizes the minimal DNA synthesis associated with cell division. In Suchanassay, rat probasin promoter to express the SV40 early genes or a cell is incubated in the presence of labelled thymidine for a transgenic mice using the long probasin promoter to express time sufficient for cell division to occur. Following washingto large T antigen, collectively termed the LADY model or remove any unincorporated thymidine, the label (e.g. the mice expressing c-myc or Bcl-2 or Fgf8b or expressing domi radioactive label) is detected, e.g., using a scintilation nant negative TGFB (see, Matusik et al., 2001, for a review of counter. Alternative assays for determining cellular prolifera transgenic models of prostate cancer). tion, include, for example, measurement of DNA synthesis by 0498. A protein of the present invention can also be BrdU incorporation (by ELISA or immunohistochemistry, administered to an animal model of a disease other than kits available from Amersham Pharmacia Biotech). Exem cancer, e.g., NOD mice to test their ability to suppress, pre plary assays for detecting cell death include APOPTEST vent, treat or delay diabetes (e.g., as described in Tang et al. (available from Immunotech) stains cells early in apoptosis, (2004)) and/or to a mouse model of GVHD (e.g., as described and does not require fixation of the cell sample. This method in Trenado (2002)) and/or to amouse model of psoriasis (e.g., utilizes an annexin V antibody to detect cell membrane re Wang et al. 2008) and/or to a model of rheumatoid arthritis configuration that is characteristic of cells undergoing apop e.g., a SKG Strain of mouse (Sakaguchi et al.), rat type II tosis. Apoptotic cells stained in this manner can then be sorted collagen arthritis model, mouse type II collagen arthritis either by fluorescence activated cell sorting (FACS), ELISA model or antigen induced arthritis models in several species or by adhesion and panning using immobilized annexin V (Bendele, 2001)) and/or a model of multiple sclerosis (for antibodies. Alternatively, a terminal deoxynucleotidyl trans example, experimental autoimmune encephalomyelitis ferase-mediated biotinylated UTP nick end-labelling (EAE; Bradl and Linington, 1996)) and/or inflammatory air (TUNEL) assay is used to determine the level of cell death. way disease (for example, OVA challenge or cockroach anti The TUNEL assay uses the enzyme terminal deoxynucleoti gen challenge (Chen et al. 2007; Lukacs et al. 2001) and/or dyl transferase to label 3'-OH DNA ends, generated during models of inflammatory bowel disease (e.g., dextran sodium apoptosis, with biotinylated nucleotides. The biotinylated sulphate (DSS)-induced colitis or Muc2 deficient mouse nucleotides are then detected by using streptavidin conju model of colitis (Van der Sluis et al. 2006). gated to a detectable marker. Kits for TUNEL staining are available from, for example, Intergen Company, Purchase, Diagnostic/Prognostic Methods NY. 0499. In one example, the present invention provides 0494 Stability of a protein of the present invention can methods for diagnosing or prognosing a condition. also be assessed by exposing a protein of the present invention 0500. As used herein, the term “diagnosis', and variants to serum and/or cells and Subsequently isolating the protein of thereof such as, but not limited to, "diagnose”, “diagnosed’ or the present invention using, for example, immunoaffinity "diagnosing includes any primary diagnosis of a clinical purification. A reduced amount of recovered protein of the state or diagnosis of recurrent disease. present invention indicates that the protein of the present 0501) “Prognosis”, “prognosing and variants thereof as invention is degraded in serum or when exposed to cells. used herein refer to the likely outcome or course of a disease, 0495. In another example, the ability of the protein of the including the chance of recovery or recurrence. present invention to block binding of a ligand to a receptor is 0502. In one example, the method comprises determining assessed using a standard radio-immunoassay or fluorescent the amount of an antigen in a sample. Thus, the proteins of the immunoassay. invention have utility in applications such as cell sorting (e.g., flow cytometry, fluorescence activated cell sorting), for diag In Vivo Testing nostic or research purposes. For example, a sample is con 0496 A protein of the present invention can also be tested tacted with a protein of the invention for a time and under for its stability and/or efficacy in vivo. For example, the conditions sufficient for it to bind to an antigen and form a protein of the present invention is administered to a subject complex and the complex is then detected or the level of and the serum levels of the protein is detected over time, e.g., complex is determined. For these purposes, the proteins can using an ELISA or by detecting a detectable label conjugated be labelled or unlabelled. The proteins can be directly US 2012/0164068 A1 Jun. 28, 2012 37 labelled, e.g., using a method described herein. When unla able by imaging. For example, the detectable label may be a belled, the proteins can be detected using Suitable means, as in protein, a radioisotope, a fluorophore, a visible light emitting agglutination assays, for example. Unlabelled antibodies or fluorophore, infrared light emitting fluorophore, a metal, a fragments can also be used in combination with another (i.e., ferromagnetic Substance, an electromagnetic emitting Sub one or more) Suitable reagent which can be used to detect a stance a substance with a specific MR spectroscopic signa protein, such as a labelled antibody (e.g., a second antibody) ture, an X-ray absorbing or reflecting Substance, or a Sound reactive with the protein or other Suitable reagent (e.g., altering Substance. labelled protein A). 0508. The protein of the present invention can be admin 0503 Preferably, a protein of the invention is used in an istered either systemically or locally to the tumour, organ, or immunoassay. Preferably, using an assay selected from the tissue to be imaged, prior to the imaging procedure. Gener group consisting of immunohistochemistry, immunofluores ally, the protein is administered in doses effective to achieve cence, enzyme linked immunosorbent assay (ELISA), fluo the desired optical image of a tumour, tissue, or organ. Such rescence linked immunosorbent assay (FLISA) Western blot doses may vary widely, depending upon the particular protein ting, RIA, a biosensor assay, a protein chip assay and an employed, the tumour, tissue, or organ Subjected to the imag immunostaining assay (e.g. immunofluorescence). ing procedure, the imaging equipment being used, and the 0504 Standard solid-phase ELISA or FLISA formats are like. particularly useful in determining the concentration of a pro 0509. In some embodiments of the invention, the protein tein from a variety of samples. of the invention is used as in vivo optical imaging agents of 0505. In one form such an assay involves immobilizing a tissues and organs in various biomedical applications includ biological sample onto a solid matrix, Such as, for example a ing, but not limited to, imaging of tumours, tomographic polystyrene or polycarbonate microwell or dipstick, a mem imaging of organs, monitoring of organ functions, coronary brane, or a glass Support (e.g. a glass slide). A protein of the angiography, fluorescence endoscopy, laser guided Surgery, invention that specifically binds to an antigen of interest is photoacoustic and Sonofluorescence methods, and the like. brought into direct contact with the immobilized sample, and Exemplary diseases, e.g., cancers, in which a protein of the forms a direct bond with any of its target antigen present in invention is useful for imaging are described herein and shall said sample. This protein of the invention is generally labelled be taken to apply mutatis mutandis to the present embodiment with a detectable reporter molecule, such as for example, a of the invention. In one example, the protein conjugates of the fluorescent label (e.g. FITC or Texas Red) or a fluorescent invention are useful for the detection of the presence of semiconductor nanocrystal (as described in U.S. Pat. No. tumours and other abnormalities by monitoring where a par 6.306,610) in the case of a FLISA or an enzyme (e.g. horse ticular protein of the invention is concentrated in a Subject. In radish peroxidase (HRP), alkaline phosphatase (AP) or f3-ga another embodiment, the protein of the invention is useful for lactosidase) in the case of an ELISA, or alternatively a laser-assisted guided Surgery for the detection of micro-me labelled antibody can be used that binds to the protein of the tastases of tumours upon laparoscopy. In yet another embodi invention. Following washing to remove any unbound protein ment, the protein of the invention is useful in the diagnosis of the label is detected either directly, in the case of a fluorescent atherosclerotic plaques and blood clots. label, or through the addition of a substrate, such as for 0510 Examples of imaging methods include magnetic example hydrogen peroxide, TMB, or toluidine, or 5-bromo resonance imaging (MRI), MR spectroscopy, radiography, 4-chloro-3-indol-beta-D-galaotopyranoside (X-gal) in the CT, ultrasound, planar gamma camera imaging, single-pho case of an enzymatic label. Such ELISA or FLISA based ton emission computed tomography (SPECT), positron emis systems are particularly Suitable for quantification of the sion tomography (PET), other nuclear medicine-based imag amount of a protein in a sample, by calibrating the detection ing, optical imaging using visible light, optical imaging using system against known amounts of a protein standard to which luciferase, optical imaging using a fluorophore, other optical the protein binds, such as for example, an isolated and/or imaging, imaging using near infrared light, or imaging using recombinant protein or immunogenic fragment thereof or infrared light. epitope thereof. 0511 Certain examples of the methods of the present 0506. In another form, an ELISA or FLISA comprises of invention further include imaging a tissue during a Surgical immobilizing a protein of the invention or an antibody that procedure on a Subject. binds to an antigen of interest on a Solid matrix, Such as, for 0512. A variety of techniques for imaging are known to example, a membrane, a polystyrene or polycarbonate those of ordinary skill in the art. Any of these techniques can microwell, a polystyrene or polycarbonate dipstick or a glass be applied in the context of the imaging methods of the Support. A sample is then brought into physical relation with present invention to measure a signal from the detectable said protein of the invention, and the protein to which said label. For example, optical imaging is one imaging modality compound binds is bound or captured. The bound protein is that has gained widespread acceptance in particular areas of then detected using a labelled protein of the invention that medicine. Examples include optical labelling of cellular com binds to a different protein or a different site in the same ponents, and angiography Such as fluorescein angiography protein. Alternatively, a third labelled antibody can be used and indocyanine green angiography. Examples of optical that binds the second (detecting) antibody. imaging agents include, for example, fluorescein, a fluores cein derivative, indocyanine green, Oregon green, a deriva Imaging Methods tive of Oregon green derivative, rhodamine green, a derivative 0507 As will be apparent to the skilled artisan from the of rhodamine green, an eosin, an erytlirosin, Texas red, a foregoing, the present invention also contemplates imaging derivative of Texas red, malachite green, nanogold Sulfo Suc methods using a protein of the invention. For imaging, protein cinimidyl ester, cascade blue, a coumarin derivative, a naph of the invention is conjugated to a detectable label, which can thalene, a pyridyloxazole derivative, cascade yellow dye, be any molecule or agent that can emit a signal that is detect dapoxyl dye. US 2012/0164068 A1 Jun. 28, 2012

0513 Gamma camera imaging is contemplated as a 0520 Closely related to PET is single-photon emission method of imaging that can be utilized for measuring a signal computed tomography, or SPECT. The major difference derived from the detectable label. One of ordinary skill in the between the two is that instead of a positron-emitting Sub art would be familiar with techniques for application of stance, SPECT uses a radioactive tracer that emits high-en gamma camera imaging. In one embodiment, measuring a ergy photons. SPECT is valuable for diagnosing multiple signal can involve use of gamma-camera imaging of an 'In illnesses including coronary artery disease, and already some or "Tc conjugate, in particular '''In-octreotide or "Tc 2.5 million SPECT heart studies are done in the United States Somatostatin analogue. each year. 0514 Computerized tomography (CT) is contemplated as an imaging modality in the context of the present invention. 0521 For PET, a protein of the invention is commonly By taking a series of X-rays from various angles and then labelled with positron-emitters such as 'C, 13N, O, F, combining them with a computer, CT made it possible to Rb, Cu, and Ga. Proteins of the invention are labelled build up a three-dimensional image of any part of the body. A with positron emitters such as 'mTc, 'Tl, and '7Ga, '''In computer is programmed to display two-dimensional slices for SPECT. from any angle and at any depth. The slices may be combined 0522 Non-invasive fluorescence imaging of animals and to build three-dimensional representations. humans can also provide in vivo diagnostic information and 0515. In CT, intravenous injection of a radiopaque con be used in a wide variety of clinical specialties. For instance, trast agent conjugated to a protein of the invention, which techniques have been developed over the years including binds to an antigen of interest can assist in the identification simple observations following UV excitation of fluorophores and delineation of soft tissue masses when initial CT scans are up to Sophisticated spectroscopic imaging using advanced not diagnostic. Similarly, contrast agents aid in assessing the equipment (see, e.g., Andersson-Engels et al., 1997). Specific vascularity of a soft tissue lesion. For example, the use of devices or methods known in the art for the in vivo detection contrast agents may aid the delineation of the relationship of of fluorescence, e.g., from fluorophores or fluorescent pro a tumour and adjacent vascular structures. teins, include, but are not limited to, in vivo near-infrared 0516 CT contrast agents include, for example, iodinated fluorescence (see, e.g., Frangioni, 2003), the MaestroTM in contrast media. Examples of these agents include iothala Vivo fluorescence imaging system (Cambridge Research & mate, iohexyl, diatrizoate, iopamidol, ethiodol, and Instrumentation, Inc.; Woburn, Mass.), in vivo fluorescence iopanoate. Gadolinium agents have also been reported to be imaging using a flying-spot Scanner (see, e.g., Ramanujam et of use as a CT contrast agent, for example, gadopentate. al, 2001), and the like. 0517 Magnetic resonance imaging (MRI) is an imaging 0523 Other methods or devices for detecting an optical modality that uses a high-strength magnet and radio-fre response include, without limitation, visual inspection, CCD quency signals to produce images. In MRI, the sample to be cameras, video cameras, photographic film, laser-scanning imaged is placed in a strong static magnetic field and excited devices, fluorometers, photodiodes, quantum counters, epif with a pulse of radio frequency (RF) radiation to produce a net luorescence microscopes, Scanning microscopes, flow magnetization in the sample. Various magnetic field gradients cytometers, fluorescence microplate readers, or signal ampli and other RF pulses then act to code spatial information into fication using photomultiplier tubes. the recorded signals. By collecting and analyzing these sig 0524. In some examples, an imaging agent is tested using nals, it is possible to compute a three-dimensional image an in vitro or in vivo assay prior to use in humans, e.g., using which, like a CT image, is normally displayed in two-dimen a model described herein. sional slices. The slices may be combined to build three dimensional representations. Articles of Manufacture 0518 Contrast agents used in MRI or MR spectroscopy imaging differ from those used in other imaging techniques. 0525. The present invention also provides an article of Examples of MRI contrast agents include gadolinium che manufacture, or "kit, containing a protein of the invention. lates, manganese chelates, chromium chelates, and iron par The article of manufacture optionally, comprises a container ticles. For example, a protein of the invention is conjugated to and a label or package insert on or associated with the con a compound comprising a chelate of a paramagnetic metal tainer, e.g., providing instructions to use the protein of the selected from the group consisting of Scandium, titanium, invention in a method described herein according to any Vanadium, chromium, manganese, iron, cobalt, nickel, cop embodiment. Suitable containers include, for example, per, molybdenum, ruthenium, cerium, indium, praseody bottles, vials, Syringes, blister pack, etc. The containers may mium, neodymium, promethium, Samarium, europium, be formed from a variety of materials such as glass or plastic. gadolinium, terbium, dysprosium, holmium, erbium, thu The container holds a protein of the invention composition lium, and ytterbium. A further example of imaging agents and may have a sterile access port (for example the container useful for the present invention is halocarbon-based nanopar may be an intravenous Solution bag or a vial having a stopper ticle such as PFOB or other fluorine-based MRI agents. Both pierceable by a hypodermic injection needle). Alternatively, CT and MRI provide anatomical information that aid in dis or additionally, the article of manufacture may further com tinguishing tissue boundaries and vascular structure. prise a second (or third) container comprising a pharmaceu 0519 Imaging modalities that provide information per tically-acceptable buffer, such as bacteriostatic water for taining to information at the cellular level. Such as cellular injection (BWFI), phosphate-buffered saline, Ringer's solu viability, include positron emission tomography (PET) and tion and dextrose solution. It may further include other mate single-photon emission computed tomography (SPECT). In rials desirable from a commercial and userstandpoint, includ PET, a patient ingests or is injected with a radioactive sub ing other buffers, diluents, filters, needles, and Syringes. stance that emits positrons, which can be monitored as the 0526. The present invention is described further in the Substance moves through the body. following non-limiting examples. US 2012/0164068 A1 Jun. 28, 2012 39

Example 1 included for the CDRs to facilitate comparison of the disul phides to a group of generally exposed residues. The intra Molecular Modelling sheet conserved disulphide average ASA was used to facili 1.1 Generation of Molecular Models for Avibodies tate comparison of the disulphides to a group of generally 0527 Avibodies are recombinant proteins comprising buried or inaccessible residues. variable domains of antibodies. Avibodies utilize the variable 0534 For the work presented here, no significant differ domains of monoclonal antibodies by fusing them into a ences were apparent in the accessibility of the disulphide single polypeptide chain interspersed by a short linker region mutants modeled with respect to the ASA as an indication of in either V-to-V, or V-to-V orientation. Depending on the conjugation probability whether they were in monomer, linker length, these Avibodies are designed to form stable, dimer or trimer configuration or whether the respective sch vs biologically active monobodies (sclv), diabodies, triabodies were in V to V, orientation or V, to V orientation. or tetrabodies containing one, two, three or four functional binding sites respectively. 1.2 Generation of a V to V, Linked Molecular Model for the 0528. The V and V, domain sequences of the Avibodies AVP04-07 Diabody modeled were used to search the RCSB PDB Data bank 0535 The AVP04-07 Avibody (SEQ ID NO. 55) is a (www.pdb.org) using both BLAST and/or FASTA searches. recombinant diabody with a theoretical pl/Mw: 8.0/51 KDa, The structure hits with the highest sequence identity, resolu a V.K light chain and a subgroup IV, chain. AVP04-07 tion and completeness were selected for use as templates for recognizes the tumour associated antigen TAG72. the Fv domains of the modeled Avibodies. If the asymmetric 0536. This Avibody utilizes the variable regions of the unit in a pdb file contained more than one template model all murine monoclonal antibody CC49, fusing them in sequence templates were used and treated identically. to form a stable, biologically active diabody containing two 0529. For Avibody diabodies and triabodies, quaternary functional binding sites. The variable domains of CC49 have templates were used to set the arrangement of the template been modified (Roberge, et al., 2006) in amino acid sequence Fvs in space and allow modeling of these Avibodies. For the in order to achieve a high-expressing and highly stable diabodies 1 LMK (Perisicet. al., 1994) or 1 MOE (Carmichael recombinant molecule with exceptional in vitro and in vivo et. al., 2003) were variously used and for the triabodies 1NQB properties. (Pei et al., 1997) was used to arrange the templates in qua 0537) Searching the PDB with the V and V, domain ternary space for modeling. sequences of the AVP04-07 highlighted one antibody in the 0530 For quaternary arrangement, copies of the core PDB, 17A6 (Larson et al., 2005), which had an 82% identity coordinate set generated by Israel Gelfand for the Fv domain match with AVP04-07 in both V and V, domains in an (Gelfand et. al., 1998a) were least squares aligned to the un-gapped alignment. quaternary template to form a “core' homo-dimer or homo 0538. The 17A6 template encodes the structure of an anti trimer. The selected Fv templates for each Avibody were then tumour CH2-domain-deleted humanized antibody. This least squares aligned to each Fv in this “core' homo-dimer or recombinant humanized antibody also recognizes the TAG72 homo-trimerto form template homo-dimers or homo-trimers. antigen. These files were subsequently edited to reflect the connectiv ity required for modeling the various Avibodies. 0539. The Fv structure in the 17A6 pdb file was used to 0531. In all cases, the “core' quaternary models were not model the Fv domains of the AVP04-07 diabody. The 1 LMK used for the Fv domain modeling in the final modeling runs described above was used for the quaternary spatial align and the linking residues were modeled “ab initio” as loops. ment of the templates to form an AVP04-07 diabody in the 0532 Molecular models of Avibodies were generated method described above. The selected highest scoring model using Discovery Studio (DS) Software (v2.5. Accelrys, CA, of the AVP04-07 diabody is shown in FIG. 1 and represents USA) using the MODELLER algorithm (Sali and Blundell, the “un-mutated configuration of this Avibody dimer. 1993) embedded in the software and evaluated using the 1.3 Generation of a V to V, Linked Molecular Model for the scoring functions contained in the Software. The best model AVP07-17 diabody was selected on the basis of the presence of a high ranking (0540. The AVP07-17 Avibody (SEQ ID NO. 59) is a score in each of the MODELLER generated Probability Den recombinant diabody with a theoretical pl/Mw: 6.4/55 KDa, sity Function (PDF) for total and physical energy and the an exceptionally long CDRH3 loop a V, light chain and a Discrete Optimized Protein Energy (DOPE) score, (Shen et. subgroup IV, chain. AVP07-17 recognizes the tumour asso al., 2006). The selected model was written out to a pdb file for ciated antigen HER2. further analysis. Images of the resulting models were also (0541 AVP07-17 has lower identity with the structures generated using DS. available in the RCSB pab when using standard Fasta and 0533. Further analysis of each selected model included Blast searches compared to the AVP04-07. No Fv pair of VL visual inspection on a graphics workstation and calculation of and VH showed as high an identity with AVP07-17 when the solvent accessible surface area (ASA) of relevant resi compared with the results obtained for AVP04-07. dues. The ASA is used here as an assessment of the modeled 0542. Alternative methods of searching the PDB were disulphide mutant's ability to be available for conjugation. tested to improve template selection for entire Fv domains. The standard deviation from the mean ASA (calculated in The MATRAS server (Kawabata 2003, Kawabata, et. al. Excel) was then used as an indication as to whether both or 2000) uses a standard sequence homology search against the one of the Cysteine residues in a disulphide or other group of current PDB using the BLAST program with a graphical residues (i.e., CDRs) were similarly exposed. For example, a representation of the aligned regions to assist in template large standard deviation indicates that one of the residues in selection. This method revealed two good templates, both the disulphide may be less exposed for reduction and/or con with greater than 64% sequence identity in both the V, and V. jugation. An average per residue average ASA is also domains. US 2012/0164068 A1 Jun. 28, 2012 40

0543. The selected Fv templates were contained in the pdb disulphide bond. This region should thus allow free access for files of a) 2B1H (Stanfield et al., 2006) which had 80.6% conjugation chemistry to disulphide containing constructs. identity to AVP07-17 excluding the linker residues and 0551 Avibodies which contain engineered intra-frame CDRH3 and b) 3G04 (Sanderset. al., 2007) which had 73.5% work 1 cysteine replacements are herein referred to as a identity to AVP07-17 excluding the linker residues and “Thiolated Avibody”. CDRH3. 0552. The V, and V. domains of antibodies are firstly (0544 The 1LMK diabody described above was used for members of the Immunoglobulin Superfamily classically the quaternary spatial alignment of the template FVS to form containing 7-10 fistrands in two sheets with a typical topol an AVP07-17 (“un-mutated') diabody in the method ogy and connectivity. These domains are secondly members described above. The long CDRH3 loop length of AVP07-17 of the V-type immunoglobulins showing symmetry of the was also problematic for modeling as no homologous struc B-sheets within the domain axis (Halaby, et. al., 1999). The tures could be found for use as templates. These were mod antibody V-type or V-set domains are divided V (type 1-4), eled as loops with no template constraints (essentially ab V.K. V. domains in online databases such as SCOP (http:// initio) and assessed for structural violations after modeling. Scop.mrc-limb.cam.ac.uk/scop/data/scop.b.c.b.b.b.html, In all cases presented here, the CDR3 loops are modeled with Murzin, et. al., 1995), InterPro (http://www.ebi.ac.uk/inter low confidence levels and are not included in some analyses profiEntry?ac=IPR013 106, Hunter et. al., 2009) and Pfam as they were not considered to affect the overall structure or (http://pfam.sbc.su.se/family/PFO7686, Bateman, et. al., framework regions of the Avibodies. 2004). (0545. The selected highest scoring model of the AVP07 0553 Given these well defined similarities, it is reason 17 diabody is shown in FIG. 2 and represents the “un-mu able to assume that all potential intra-framework 1 cysteine tated configuration for this Avibody dimer. pair replacement mutations identified in the AVP04-07 V.K domain should be structurally transferable to both V (type 1-4) domains and V, domains. 1.4 Generation of a Molecular Model for the AVPO2-60 Dia 0554. The V, domain of the model generated for AVP04 body 07 was inspected on a graphics workstation for pairs of resi 0546. The AVP02-60 Avibody (SEQ ID NO: 61) is a dues that could be mutated to Cysteine and be able to forman recombinant diabody with a theoretical pl/Mw: 8.47/50.1 intra-framework 1 disulphide bridges as well as be available kDa, a V, chain kappa and a subgroup III V, chain. It is based for reduction and Subsequent conjugation to payloads. Pairs on the primary mouse monoclonal C595 antibody that recog of residues in FR1 which have side chains generally angled nizes a breast cancer associated mucin encoded by the MUC1 towards each other, which have side chains atoms generally gene, CD227 (Gendleret. al., 1990). It recognizes the epitope exposed to solvent and which have Co. carbon atoms were RPAP within the protein core of the mucin, a motif repeated ~6-7A apart were considered as good candidates for intra Some 40 times in the sequence. framework 1 cysteine insertions capable of forming a disul (0547 Blast and Fastasearching with theV, or V revealed phide bond. several templates with high identity scores that contained 0555 Fortransfer to V. domains, a structural alignment of both the V, and V, domains. However, only one template had the V and V, domains of the Gelfand core co-ordinates a V with Sufficient identity in sequence and length to model (Gelfandet. al., 1998a; Gelfandet. al., 1998b) was generated. the CDRH3. Hence two templates were selected for V, and These aligned V, and V cores were Subsequently used to V, modeling while one extra template was selected for V. align the V and V, domains from each un-mutated FV only modeling. The templates selected were: a) 1 MHPV, domain model. This structural alignment was used for map and V (86.9% identity, 89.6% homology; Karpusas, et. al., ping the identified Cysteine mutant pairs from the V, to the 2003), b)2B2XV, and V (85.7% identity, 88.3% homology: V. sequences that were then used for modeling the V Clark, et. al., 2006) and c)2ADGV: (86.8% identity, 96.5% domain cysteine insertion mutants. In each case, a single homology; Zhou et. al., 2005) which was the only template modeling run was used to generate a double cysteine insertion with an un-gapped alignment for CDRH3, the V, domain of mutant model containing single analogous cysteine mutant this Fv was not used in the modeling. pairs in each of the V, and V. domains. (0548. Overall the templates the AVP02-60 has 88.4% and 1.6 Framework 1 Cysteine Insertion Positions Identified for identity and 91.1% homology. The 1 LMK diabody described Engineering Cysteine Replacement Mutations and Molecular above was used for the quaternary spatial alignment of the Modelling in AVP04 Avibody Diabodies. template Fvs to form an AVP02-60 (“un-mutated') diabody in the method described above. 0556. The un-mutated AVP04-07 model was the starting (0549. The selected highest scoring model of the AVP02 point for the mapping of intra-framework 1 cysteine insertion 60 diabody is shown in FIG. 3 and represents the “un-mu mutants as described above. This mapping showed that the tated configuration of this Avibody dimer. identified V, cysteine insertion positions could indeed be structurally mapped to the V, domain with ease and that these 1.5 Identification of Framework 1 Cysteine Insertion Posi residues were likely to form a disulphide bond. tions for Engineering Replacement Cysteine Mutations and 0557. In Framework 1 V. (Kabat residues 1 to 23 inclu sive) and Framework 1 V. (Kabat residues 1 to 30 inclusive) Molecular Modelling of the Same. of AVP04-07, the preferred positions for intra-framework 1 0550 Framework 1 (FR1) in the architecture of an immu cysteine insertions were identified as: noglobulin V domain is a good candidate for engineering 0558 Light chain framework 1 Kabat residues L8 and cysteine replacements because it is located on the edge of one L11 (AVP04-50, SEQID NO:57) of the two B-sheets and so is generally well exposed to solvent 0559 Heavy chain framework 1 Kabat residues except for residues adjacent to the conserved inter-domain H7-H10 (AVP04-84, SEQID NO: 63) US 2012/0164068 A1 Jun. 28, 2012

0560 Light chain framework 1 Kabat residues L14 and 101 Avibodies, translates structurally to L7-L11. In AVP07, L17 (AVP04-78, SEQID NO: 77) the preferred intra-framework 1 cysteine insertion positions 0561 Heavy chain framework 1 Kabat residues Hf were thus identified as: H16 (AVP04-85, SEQID NO: 75) 0573 Light chain framework 1 Kabat residues L7 and 0562 Modelling was repeated using the method outlined L11 (AVP07-88, SEQID NO: 87) for the AVP04-07 model (Example 1.2) using the same input 0574 Heavy chain framework 1 Kabat residues parameters except for the sequence input which reflected the H7-H10 (AVP07-90, SEQID NO: 89) desired mutations above. Model assessment was also carried 0575 Light chain framework 1 Kabat residues L14 and out as for the AVP04-07 models. L17 (AVP07-89, SEQID NO:91) 0563 Each cysteine insertion was subjected to modeling 0576 Heavy chain framework 1 Kabat residues Hf with one V, cysteine pair mutant and its analogous V, cys H16 (AVP07-91, SEQID NO: 93) teine pair mutant included in each modeling run. The results 0577. The V, , in the Kabat numbering system is missing of the cysteine insertion modeling onto the AVP04-07 FR1 the L10 sequence position (Johnson and Wu, 2000), which structure are shown in FIGS. 4A and B. FIG. 4B shows that necessitated the renumbering mentioned above. Structurally there was little structural change in the vicinity of the engi however, in the context of the current invention, V.K L8-L11 neered intra-framework 1 cysteine mutations even when a is equivalent to VW L7-L11 with no residues missing. This disulphide bond is formed. can clearly be seen when comparing the models for the 0564. As expected, with the aim of defining mutatable AVP04-07 V, FR1 and AVP07-17 V, FR1 in FIG. 4 and FIG. residue pairs that would be available for reduction and sub 7 respectively where in these specific cases the residue dele sequent conjugation to payloads, the solvent accessible Sur tion in the sequence is N-terminal the cysteine insertion. face area (ASA) values for each cysteine mutant pair was (0578. Following the trend observed with the cysteine shown to be significantly higher than the highly conserved, insertion positions outlined for AVP04-07, the accessible sur and structurally buried, cysteine pairs H22-H92 and L23-L88 face area (ASA) values for each preferred cysteine insertion, and similar to the structurally exposed CDR residues (FIG.5). in AVP02 (FIG. 8) and AVP07 (FIG. 9) Avibodies, were shown to be significantly higher than the highly conserved, 1.7 Framework 1 Cysteine Insertion Positions Identified for and structurally buried/inaccessible, cysteine pairs H22-H92 Engineering Cysteine Replacement Mutations and Molecular and L23-L88 and similar to the structurally exposed CDR Modelling in AVP07-XX and AVP02-xx Avibody Diabodies. residues. 0565. As outlined above, structural similarity between the 0579. The V domains of antibodies (VandV) have long VH (type 1-4), V.K. V.W. domains across the antibody fami been subdivided into sequential and structural subtypes, for lies is known and accepted. Due to this structural similarity, example the V, types I-IV and the V.K and Wk. These sub the cysteine insertion positions identified from the model of types are largely based on differences in the FR1 sequence AVP04-07, were structurally transferred to the AVP02-XX and and structure of these domains (Lefranc, 2001a; LeFranc, AVP07-XX cysteine insertion Avibody models. 2001b: Pallares, 1999). The current work shows that despite 0566. In the case of AVP02-60, the preferred Kabat posi this subdivision the preferred disulphide insertion positions tions for cysteine insertion were identical to that of AVP04 are readily transferrable in models of a variety of these sub 07, namely: types without significant distortion of the non-mutated model 0567 Light chain framework 1 Kabat residues L8 and frameworks. L11 (AVP02-101, SEQID NO: 79) 0580. However, the inventors noted that during visual 0568 Heavy chain framework 1 Kabat residues inspection of un-mutated structural models, some V/V, Sub H7-H10 (AVP02-104, SEQID NO: 81) types contained additional (alternate) positions where intra 0569 Light chain framework 1 Kabat residues L14 and framework 1 cysteine insertions could be placed, generally by L17 (AVP02-102, SEQID NO: 83) shifting the preferred disulphide insertion positions outlined (0570 Heavy chain framework 1 Kabat residues HB above by 1-2 residues towards the N-terminus (for example H16 (AVP02-105, SEQID NO: 85) AVP02-103, with FR1 H6-H9 mutation, SEQID NO: 95) or 0571. The AVP02-60 model has a V, type III domain that 1-2 residues towards the C-terminus of the polypeptide chain is structurally a little different, in the region of Kabat H7-H10, (for example corresponding clones AVP07-63 (SEQID NO: to the V, type I domain of AVP04-07 and AVP07-17. How 65) and AVP07-68 (SEQID NO: 97), with FR1 L8-L12). ever the structural transfer of the AVP04-07 cysteine insertion 0581. In all cases discussed above however, the preferred could still be achieved becausea) the Co. positions of H7 and and/or alternate positions identified as being compatible with H10 were very similar and b) although the intervening H8 and intra-framework 1 disulphide insertions all met the key mod H9 residue positions differed, both of these residues were eling constraints outlined in Example 1.5 Gly, a small and more flexible residue with respect to other amino acids. Based on the V-gene germline sequences pro 2 Synthesis of Avibody Constructs vided by the online Immunogenetics Database; IMGT (http:// imgt.cines.fr), 98% of human sequences have a Gly at H8 and 2.1 Synthesis of “Un-mutated Avibodies Without Engi 48% have Gly at H9. This flexibility allowed the structural neered Intra-Framework 1 Disulphide Insertions. transfer of the H7-H10 disulphide from AVP04-84 to AVP02 0582 DNA constructs encoding the V, and V, regions of 104. The results of the intra-framework 1 cysteine insertion/ a mouse mAbspecific for TAG72 (SEQID NO. 54), a human disulphide bond formation modeling onto the AVP02-60 FR1 mAbspecific for HER2 (SEQID NO:58) and a murine mAb structure is shown in FIGS. 6A and B. specific for MUC1 (SEQID NO: 60) were synthesized with 0572. In the case of AVP07-XX. Avibodies, as this construct the appropriate restriction sites and cloned into puC57 by contains a VW chain, the cysteine insertion at Kabat position GenScript (Piscataway, N.J., USA). Although Avibodies have L8-L11, found in both V.K containing AVP04-50 and AVPO2 been isolated in either orientation of V region i.e. V-Linker US 2012/0164068 A1 Jun. 28, 2012 42

V, and V-Linker-V (Carmichaeletal., 2003), all constructs 0589 AVP04-51 Diabody nucleic acid sequence (SEQ described herein were arranged as V-Linker-V. ID NO: 98), forming the Avibody mutated in Kabat 0583 All DNA manipulations were carried out according residues L13 and L19 (SEQID NO: 99). to standard protocols with reagents purchased from New 0590 AVP04-78 Diabody nucleic acid sequence (SEQ England Biolabs (Ipswich, Mass., USA). Diabody encoding ID NO: 76), forming the Avibody mutated in Kabat DNA constructs were excised from pUC57 with the appro residues L14 and L17 (SEQID NO: 77). priate restriction enzymes, resolved on a 1% (w/v) agarose 0591 AVP04-84 Diabody nucleic acid sequence (SEQ gel and purified from the gel using the Qiaquick gel extraction ID NO: 62), forming the Avibody mutated in Kabat residues H7 and H10 (SEQID NO: 63). kit (Qiagen). Constructs were ligated into similarly prepared 0592 AVP04-85 Diabody nucleic acid sequence (SEQ pET22b expression vectors and the ligation mixtures trans ID NO: 74), forming the Avibody mutated in Kabat formed by the electroporation method into E. coli XL1-Blue residues H13 and H16 (SEQID NO: 75). cells. Miniprep DNA was extracted from transformants using 0593 AVP04-70 sclv nucleic acid sequence (SEQ ID the Qiagen miniprep spin kit and recombinant clones identi NO: 100), forming the Avibody mutated in Kabat resi fied by sequencing with T7 promoter and terminator primers dues L8 and L11 (SEQID NO: 101). using Dye Terminator Cycle Sequencing kits with AmpliTaq. 0594 AVP04-74 Triabody nucleic acid sequence (SEQ The clone containing the V regions of the anti-TAG.72 mAb in ID NO: 102), forming the Avibody mutated in Kabat the V-Gly-Ser-V, orientation was designated AVP04-07 residues L8 and L11 (SEQ ID NO: 103). (SEQID NO. 54). The clone containing the V regions of the anti-HER2 mAb in the V-Gly-Ser-V, orientation was des AVP07-XX Family Template Sequences (HER2-Specific): ignated AVP07-17 (SEQ ID NO: 58). The clone containing 0595 AVP07-88 Diabody nucleic acid sequence (SEQ the V regions of the anti-MUC1 mAb in the V-Gly-Ser-V, ID NO:86), forming the Avibody mutated in Kabat resi orientation was designated AVP02-60 (SEQ ID NO: 60). dues L7 and L11 (SEQID NO:87). These three clones formed the base parental sequences from 0596 AVP07-71 sclv nucleic acid sequence (SEQ ID which all other Thiolated Avibodies were derived. NO: 104), forming the Avibody mutated in Kabat resi 0584) This method of cloning allowed for the insertion of dues L8 and L12 (SEQID NO: 105). an amino-terminal pelB leader sequence for periplasmic 0597 AVP07-63 Diabody nucleic acid sequence (SEQ expression of the target protein and either a carboxy-terminal ID NO: 64), forming the Avibody mutated in Kabat (His) tag or a carboxy-terminal Myc+(His) tag. The addi residues L8 and L12 (SEQ ID NO: 65). tion of an affinity tag, such as (His) was routinely used to 0598 AVP07-68 Diabody, identical to AVP07-63 but streamline downstream purification processes and is known containing a carboxy-terminal (His) tag and not a car to be neutral in biological activity. boxy-terminal Myc+(His) tag, nucleic acid sequence 0585. In some cases, an identical V/V, Avibody (SEQ ID NO: 96), forming the Avibody mutated in sequence was constructed in both carboxy-terminal (His) tag Kabat residues L8 and L12 (SEQID NO: 97). and carboxy-terminal Myc+(His) versions. One example of this is AVP07-63 (SEQ ID NO: 64) containing the Myc+ AVP02-XX Family Template Sequences (MUC1-Specific): (His) tag, and AVP07-68 (SEQID NO: 96) containing only the (His) tag. Although these two Avibodies had different 0599 AVP02-101 Diabody nucleic acid sequence (SEQ carboxy-terminal tags, the V and V sequences were com ID NO: 78), forming the Avibody mutated in Kabat resi pletely identical and thus these two constructs were used dues L8 and L11 (SEQ ID NO: 79). interchangeably. 0600 These thiolated Avibodies were exemplified herein to demonstrate that the preferred intra-framework 1 engi 2.2 Introduction of Intra-Framework 1 Engineered Cysteines neered cysteine insertion mutations were a) functionally transferable between V, and V. domains and different sub and N-terminal Serine Substitution by Site-Directed types thereof, b) compatible with proteins (e.g., Avibodies) Mutagenesis. containing a single (scFv) or multiple (diabody/triabody) Fv 0586 Based on modeling data generated, the intra-frame domains and c) robust enough to allow for movement (e.g., work 1 engineered cysteine insertion mutations were intro +/-1-2 residues) either side of the preferred intra-framework duced into the un-mutated Avibody sequences of AVP04-07. 1 disulphide position without abrogating functionality, as AVP07-17 and AVP02-60 to form the following thiolated Suggested as feasible by modeling (refer to Example 1.7). Avibodies: 0601. In all cases, cysteine residues were introduced by altering the nucleotide sequences encoding for the specific AVP04-XX Family Template Sequences (TAG72-Specific): amino acid of interest using a QuikChange R site-directed mutagenesis method (Stratagene) as per instructions. Using 0587. As discussed herein the use of the symbols “XX' in the AVP04-07 Avibody as an illustration, the Proline residue the context of an Avibody name indicates that numerous at Kabat position L8 (FR1 VL region) is encoded by the Avibodies of the same series exist and that the description sequence CCG and the Leucine residue at Kabat position L11 relates to all Avibodies in that series. Replacement of the “xx' (FR1 VL region) is encoded by the sequence CTG. The with a specific number indicates that the description refers to QuikChange R site-directed mutagenesis technique was used the Avibody having that number (based on the nomenclature to alter both of these nucleotide sequences to TGC, which used herein). encodes Cysteine. 0588 AVP04-50 Diabody nucleic acid sequence (SEQ 0602. The QuikChange(R) site-directed mutagenesis PCR ID NO: 56), forming the Avibody mutated in Kabat based method uses two complementary synthetic oligonucle residues L8 and L11, (SEQ ID NO:57). otides that contain the desired mutations as primers and plas US 2012/0164068 A1 Jun. 28, 2012 mid DNA as the template to synthesize the double-stranded clone was engineered from the parent AVP04-50 by deletion mutant PCR product. Using the example above, to introduce mutagenesis using mutagenic primers encoding the desired cysteine residues at Kabat positions L8 and L11 of the FR1 sequence above. region of the VL chain in AVP04-07, the following sequence 5 GATATC GTG ATG ACC CAGAGCTGC AGC AGC 3. Expression and Purification of “Un-Mutated” and Thi TGCCCG GTGAGC GTG GGC GAAAAAG 3' (SEQID olated Avibodies Using Bacterial Expression NO: 106) was used as the forward primer and 5' CTTTTTC 0607. The DNA of individual Avibody constructs was GCC CAC GCT CACCGG GCAGCT GCT GCA GCTCTG transformed into chemically competent E. coli BL21 cells GGT CAT CAC GAT ATC 3' (SEQID NO: 107) was used using the manufacturer's standard protocol (Stratagene). The as the reverse primer. Amplification was performed using the E. coli BL21 expression strain served as the major expression following conditions in sequence: 95°C. for 30 sec; 18 cycles strain for all Avibodies exemplified. Expression was by consisting of 95°C. for 30 sec, 55° C. for 30 sec and 68°C. for means of two interchangeable approaches depending on 13 min; a final extension of 68°C. for 7 min. The template was expected yield requirements; either bacterial shake-flask digested with DpnI at 37°C. for 1 hour. Transformants were expression or bacterial fed-batch fermentation. Quality obtained following the manufacturer's instructions and iden assessment on protein Avibody from either method clearly tified by DNA sequencing as described above. indicated that the two methods were interchangeable and 0603 Similar mutagenesis approaches were utilized to protein quality and properties were comparable. generate all thiolated Avibodies or were employed to replace 3.1 Bacterial Shake-Flask Expression the native N-terminal residue of the protein with a Serine residue. N-terminal Serine substitution was carried out either 0608. A single transformant colony was inoculated into before or after introduction of the intra-framework 1 disul 500 ml 2xYT containing 1% D-glucose and 100 ug/ml ampi phide mutations. An example nucleotide sequence used for cillin and incubated at 37°C. overnight, shaking at 220 rpm. substituting the N-terminal Glin of AVP04-07 with a Ser resi 18 L of the same media was seeded with the overnight culture to a final ODoo of 0.1 and incubated at 30°C. until the ODoo due in is presented in SEQID NO: 66 and SEQID NO: 67. was between about 0.6-0.8. The cultures were transferred to 12°C. and shaking continued until the induction temperature 2.3 Sequence Modification of Avibody Constructs was reached. Protein expression was induced with the addi 0604 Standard molecular biology techniques known to tion of 0.2 mMIPTG and the cultures incubated at 12° C. for those skilled in the art were employed for all other modifica 15 hours. Bacterial pellets were prepared by centrifugation at tions to DNA sequences described. Where an Avibody 10,000xg, harvested, weighed and stored at -20°C. sequence contained native Cysteine residues in hyperVari 0609 Bacterial pellets containing expressed protein from able CDR regions, positions that were likely to be surface this expression system averaged approximately 6 g/L of cul exposed as suggested by modeling data (refer to FIG. 5, FIG. ture media. 8 and FIG. 9), these residues were mutated to alternative, non-thiol-containing amino acids by site-directed mutagen 3.2 Bacterial Fed-Batch Fermentation esis essentially as described above. As an example, the paren 0610 Seed cultures were grown in 2 L baffled Erlenmeyer tal clone for the AVP07 family; AVP07-17, contained two flasks containing 500 mL of a complex medium and incu such Cysteine residues; Cys 104 (Kabat numbering H100) bated at 37° C. shaking at 200 rpm for 16 h; the complex and Cys 109 (H10OE) within the V, CDR3 region. These medium contained (per L): Tryptone, 16 g;Yeast Extract, 5 g; residues were Substituted to Alanine using standard NaCl, 5 g; ampicillin, 200 mg. Defined medium was used for Quikchange R site-directed mutagenesis using mutagenic protein expression and contained (per L): KHPO, 10.64 g; (NH4)HPO, 4.0 g; and citric acid monohydrate, 1.7 g; glu primers SEQ ID NO: 68 and SEQ ID NO: 69, forming cose 25 g; MgSO4.7H2O, 1.25 g; PTM4 trace salts, 5 mL.; AVP07-86 (SEQID NO: 109). All AVP07-xx Thiolated Avi ampicillin, 200 mg; thiamine-HCl, 4.4 mg. PTM4 trace salts bodies contain this extra modification of VCDR3, rendering contained (per L): CuSO4.5H2O, 2.0 g; NaI, 0.08 g; MnSO. the AVP07-XX family compatible with the intra-framework 1 H2O, 3.0 g; NaMoO.2H2O, 0.2 g; HBO, 0.02 g; CoCl disulphide mutation strategy. 6H2O, 0.5 g., ZnCl2, 7.0 g; FeSO.7H2O, 22.0 g; CaSO. 0605. Thiolated Avibodies were also generated with modi 2H2O, 0.5 g; HSO 1 mL. All media and additives were fied linker lengths in order to generate thiolated versions of sterilized by autoclaving at 121° C. for 30 minutes except sch v or Triabodies. It is well known from published literature PTM4 trace salts, thiamine hydrochloride and amplicillin in the antibody field that modification of linker composition which were filter sterilised. and length can affect formation of Avibody multimers (Kortt 0611 Protein expression was completed in 2 L glass Bio et al. 1997). Promotion of schv formation was engineered by stat B bioreactors (Sartorius Stedim Biotech, Germany) con modifying the linker length of the diabody parent from five taining 1.6 L of defined medium. The dissolved oxygen con residues, typically GGGGS (SEQ ID NO: 135) to fifteen, centration was maintained at 20% by automatically varying GGGGSGGGGSGGGGS (SEQ ID NO. 53) using a the agitation rate between 500 and 1,200 rpm and the aeration mutagenic primer encoding the extra residues. rate (air supplemented with 5% oxygen) between 0.3 and 1.5 0606 Similarly, triabody formation was encouraged by L min'. Oxygen supplementation of the air flow was manu removal of the linker residues and, in Some cases, even ally increased as required. The pH of the culture was con removal of up to two residues of the preceding variable trolled at 7.0 via automatic addition of 10% (v/v) HPO, or domain. For example, the nucleic acid encoding the AVP04 10% (v/v) NH solution and foam was controlled by the 74 Avibody (SEQID NO: 102), encodes a triabody with the automatic addition of antifoaming agent 10% (v/v) polypro residues VTVSS-DIVM instead of the linker region. This pylene 2025). Unless specified otherwise, the vessel tem US 2012/0164068 A1 Jun. 28, 2012 44 perature was maintained at 37° C. Bioreactors were inocu ranges. Avibodies were dialysed into more than 200x volume lated with seed culture to attain a starting optical density of buffer with three buffer exchanges no less than 2 hours (measured at 600 nm) of 0.25. apart. Dialysis was performed using Spectrapor 6-8000 Da 0612. After complete utilization of the glucose added to MW cut-off dialysis tubing at 4°C. the medium, nutrient solution (feed) containing (per L): glu 0.617 Following dialysis, the protein sample was centri cose, 600 g; and MgSO4.7H2O 22.4g, was pumped into the fuged at 3220xg for 10 minutes to pellet denatured insoluble bioreactor at a flow rate of 40 mL h". Two hours after initia material prior to ion exchange. Ion exchange was performed tion of the feed the vessel temperature was slowly reduced to using the AKTA purifier 10, employing up to two 5 mL 20° C. over a 2.5 hour period (6.8° C. h") after which protein HiTrapTMSPHP columns in series, passing the cleared dialy expression was induced by the addition of 0.2 mMIPTG and sed material through the column via a P960 external pump. the feed rate was decreased to 6 mL h". Cultures were har Following this step, the column was washed with 10 column Vested 12 hours after induction and typically optical densities Volumes of ion-exchange buffer prior to commencement of a (measured at 600 nm) reached 110 and approximately 330 g linear buffer gradient (salt gradient) for elution of the protein of wet cell paste was recovered from each 2L culture. from the column. In this process, the ion exchange buffer was 3.3 Purification of Avibodies Expressed in E. coli replaced over a linear gradient with the identical buffer with 0613 Irrespective of the expression approach that was the addition of NaCl to 1M final concentration. The elution implemented, all Avibody proteins were purified essentially gradient was performed over 300 mL with a final concentra as outlined below. tion of 600 mMNaCl. 0614 Bacterial pellets harvested from expression culture 0618. Fractions corresponding to the eluted diabody (as (approximately 50-400 g depending on expression method) determined by the 280 nm absorbance profile on Unicorn were lysed, protein extracted and Subsequently purified by Software) were pooled and quantified. A typical ion exchange standard chromatographic techniques. 5 mL of His-Tag affin elution profile for AVP04-50 is presented in FIG. 10B. The ity chromatography lysis buffer (20 mM phosphate, 500 mM AVP04-50 diabody routinely eluted at a salt concentration of NaCl, 20 mM Imidazole, 0.25 mg/ml Lysozyme, 1 mM approximately 37 mS/cm or 32% B in which the major PMSF, 50 ug/ml DNAsel, pH 7.4) for every gram of bacterial dimeric isoform (arrow) of AVP04-50 could be easily sepa pellet was used to resuspend the cell pellet prior to lysis by rated from other charge and size variants. The diabody clones, mechanical homogenisation then either Sonicated (6x30 sec even those from different families, routinely eluted at similar ond pulses on ice) or by three passages through an Emulsi point in the salt gradient. In some cases, analytical size exclu flex-C5 cell disruptor (AVESTINInc., Canada). The bacterial sion using a calibrated Superdex 200 10/300 column (GE lysate was Subsequently incubated at room temperature for 1 LifeSciences) in 1xRBS buffer (137 mMNaCl, 2.7 mMKC1, hour prior to centrifugation (16,000xg, 30 min) and filtration 8.1 mM NaHPO, 1.47 mM. KHPO, pH7.4) was carried (0.45 um filter membrane). out to confirm peak identity of the desired species or compo 0615. His-Tag affinity chromatography purification using sition of specific fractions before pooling. The elution frac the AKTA Purifier 10 (GE LifeSciences) was then used to tions containing the major isoform of interest were pooled for purify diabodies from filtered bacterial lysate. Between one downstream purification. and four 5 mL HisTrapTM (GE LifeSciences) crude FF col 0619. Following ion exchange, eluted protein material umns were employed in series for purification depending on was concentrated to approximately 3 mg/mL at 4°C. prior to the scale of purification. Lysate was passed through the His gel filtration. Gel filtration was performed using the Pharma TrapTM column via an external P960 pump. HisTrapTM col cia Amersham (GE LifeSciences) Superdex(R) 7526/60 prep umns were washed with 10 column volumes of His-Tag affin grade column in PBS on the AKTA Purifier 10. Using the ity chromatography extraction buffer (20 mM sodium AVP04-50 diabody as an example, the diabody eluted at phosphate, 500 mM. NaCl, 20 mM Imidazole, pH7.4). Puri approximately 140 ml (or 53.5 minutes) post injection (FIG. fied protein was eluted in 50% His-Tag affinity chromatogra 10C). Diabody variants, both within the AVP04 family and phy elution buffer (20 mM sodium phosphate, 500 mMNaCl, others, routinely eluted at similar elution volumes as expected 500 mM Imidazole, pH7.4) and 50% His-Tag affinity chro of any globular protein with a molecular weight of approxi matography extraction buffer (a final concentration of 260 mately 54 kDa. Fractions within the margins outlined in FIG. mM Imidazole). Fractions containing eluted proteins (as 10C, corresponding to the eluted AVP04-50 dimer, were determined by 280 mM absorbance on AKTA Unicorn soft pooled and concentrated to between 0.5-3 mg/ml using Ami ware) were collected, pooled, protein concentration deter con Ultrafree spin concentrators with a 10K MWCO (Milli mined and dialysed in the appropriate ion exchange buffer. A pore, USA) at 3200xg, 4° C. typical His-Tag affinity chromatography elution profile, 0620. The final purity of the purified product was routinely using the TAG72-specific AVP04-50 (SEQ ID NO: 57) dia assessed by gel filtration chromatography on a SuperdeX(R) body is shown in FIG. 10A. All Avibodies described herein 200 10/300 column and SDS-PAGE electrophoresis. As showed similar elution profiles. example, the purification method of AVP04-50 routinely 0616 Partially purified Avibodies were subsequently returned protein with purities resulting in a single clean elu dialysed in a buffer 1.0-1.5 pH units lower than the calculated tion peak on gel filtration (FIG. 10D) and a single defined pI of the protein (for cation exchange) or 1.0-1.5 pH units species on SDS-PAGE electrophoresis (FIG. 10E). higher than the pl of the protein (for anion exchange). Typi 0621. The purification strategy and resultant purity pro cally, Avibodies with a pl of 7.0-8.0 were dialysed in MES files did not differ significantly between any of the Avibodies buffer (50 mMMES, pH 6.0 for cation exchange), those with tested. FIGS. 11A-C highlight the final size exclusion chro a pi of 8.0-9.0 were dialysed in phosphate buffer (50 mM matography profiles of Avibodies described herein and as phosphate, pH 7.0 for cation exchange) and those with a pi of indicated in the Figures. As expected, aside from a small 5.0-6.5 were dialysed in Tris buffer (20 mM Tris-HCl, pH 8 degree of variance both within and between different Avibody for anion exchange). All Avibody p values fell within these families, the elution times of the Avibodies corresponded well US 2012/0164068 A1 Jun. 28, 2012

to the expected molecular size; triabodies eluted earlier than functionally active. The presence of engineered intra-Frame diabodies which eluted later than scFvs. All Avibodies work 1 disulphide bridges in thiolated Avibodies, and their described herein could be functionally expressed and purified availability to reduction, was assessed by a colorimetric to Substantial homogeneity. The presence of intra-Framework assay. 1 cysteine replacement mutations did not have any effect on 0628. Thiolated Avibodies were incubated with 3.8 mM of the ability to functionally express and purify the Avibody to TCEP (Tris(2-carboxyethyl)phosphine hydrochloride) Substantial uniformity, partially confirming modeling data (Pierce, Rockford, Ill., USA) in PBS for 25 min at RT. Fol Suggesting the placement of engineered cysteines within lowing reduction, TCEP was removed with a PD10 desalting Framework 1 of Thiolated Avibodies did not cause detrimen column pre-equilibrated with 100 mM phosphate buffer+1 tal structural conformational changes leading to Avibody mM EDTA pH 6.5, collecting 0.5 mL fractions. Peak protein destabilization. fractions were identified by UV spectroscopy and pooled. Example 4 0629. To test reactive thiols, 50-75 ug of reduced protein In Vitro Immunoreactive Assessment of Diabodies was diluted in 100 mM sodium phosphate buffer, 1 mM 0622 Binding activity to soluble antigen was established EDTA, pH 8.0 with 5 ul of 4 mg/mL Ellman's reagent (5.5'- by a column shift assay using size exclusion chromatography. Dithio-bis(2-nitrobenzoic acid); DTNB) (Pierce, Rockford, The antigen for the AVP04-XX. Avibodies is TAG72, available Ill.). The reaction was allowed to proceed at ambient tem in soluble form from bovine submaxillary mucin (BSM) perature for 15 min. Reactive sulphydryl concentration was (Sigma). For the AVP07-XX. Avibodies, the soluble antigen is quantified by spectroscopy, assuming the molar extinction recombinant HER2 ectodomain. For the AVP02-XX. Avibod coefficient of TNB in this buffer system at 412 nm, is 14,150 ies, the soluble antigen is recombinant full length MUC1. M' cm. Estimation of reactive sulphydryl groups per dia Irrespective of Avibody orantigen, the column shift assay was body was obtained by dividing the molar concentration of performed essentially as described below. sulphydryls by the molar concentration of diabody. Intact IgG 0623. At least two times molar excess of soluble antigen to and a non-Thiolated Avibody not containing intra-Frame diabody was incubated for 1 hr in PBS buffer at ambient work 1 cysteine replacement mutations (interchangeably temperature. Binding activity was determined by comparing the resulting Avibody-antigen complex peak to the free dia either AVP04-07 or AVP02-60) were used as standardizing body peak. A positive binding result was regarded as the controls. depletion of the peak corresponding to free Avibody and/or 0630 Under these reducing conditions, the conserved dis increased size of the peak corresponding to an Avibody ulphide bond between invariant Kabat positions L23 and L88 antigen complex following incubation. The elution profiles of and invariant Kabat positions H22 and H92 are not reactive the Avibody or Avibody-antigen complex was monitored and are not available for conjugation as expected. though absorbance at 280 nm. In all cases, Avibody alone 0631. Using the reducing conditions outlined above, an eluted between 28-33 minutes, and Avibody-Antigen com intact IgG control indicated on average, 8 reactive thiols plexes eluted at 10-25 minutes. following reduction, as expected from sequence analysis 0624 The immunoreactivity of all Avibodies described (data not shown). In contrast, control Avibody Such as herein was assessed using the protocol described above and AVP04-07 and AVP02-60 consistently showed no free or the results depicted in FIGS. 12A-C. In all cases, the forma reactive thiols. tion of an Avibody-antigen complex, evidenced by a signifi 0632 Table 2 shows a subset of AVP04-XX, AVP02-xx and cant shortening of elution times in gel filtration, and/or AVP07-XX. Thiolated Avibodies with intra-framework 1 cys reduced amount of unbound Avibody was observed; indicat teine replacement mutations in V, L8-L11 (or L7-L11 in the ing Avibodies are immunoreactive. Sclv Avibodies have just case of the V, containing AVP07-XX family). In all cases, one binding site on each molecule so are expected to display unreduced Avibodies containing intra-Framework 1 cysteine weaker binding properties than diabodies and triabodies that replacement mutations appeared to have less than 0.5 reactive have multiple binding sites and as such, display avid binding. thiols on the Surface, clearly indicating that the intra-Frame As expected, AVP04-70 schv forms a less stable Avibody: work 1 cysteine replacement mutations, in the unreduced Antigen complex than the diabody or triabody clones, as state, did indeed form a disulphide bridge. evidenced by a different SEC profile. However, depletion of 0633. This disulphide bridge could be readily reduced as the unbound AVP04-70 peak and formation of an Avibody: described, making them available for conjugation. In the Antigen complex peak is evident in the profile, indicating it is reduced state, all Avibodies containing intra-Framework 1 immunoreactive. cysteine replacement mutations displayed an average of 4 0625 Complex formation was not observed when Avibod reactive thiols. ies were incubated with an irrelevant antigen indicating a specific binding interaction occurred. TABLE 2 0626. The presence or absence of intra-Framework 1 cys teine replacement mutations in thiolated Avibodies did not Quantification of reactive thiols by molar absorptivity. abrogate binding, further indicating that the intra-Framework Average 1 cysteine replacement mutations sites were engineered in Protein Conc. Conc. S-H Number of positions which had little or no effect on the functional prop Sample (mol/L) (mol/L) reactive Thiols erties of the Avibody. AVP04-50 native 1838E-OS 7.915E-06 0.4 AVP04-50 reduced 1.268E-OS 5.853E-OS 4.6 Example 5 AVP2-101 native 1.179E-05 S.497E-06 O.S Quantification of Free Sulphydryls in Thiolated Avi AVP2-101 reduced 6.369E-OS 2.869E-04 4.5 bodies AVP07-88 native 8.758E-06 1.187E-OS 1.4 AVP07-88 reduced 7.532E-06 3.769E-05 S.O 0627. Thiolated Avibodies could be routinely expressed and purified to Substantial homogeneity and were shown to be US 2012/0164068 A1 Jun. 28, 2012 46

Example 6 Stoichiometrically defined conjugation to engineered cys teines was occurring with high efficiency. The following data Payload Conjugation to Reduced Engineered Disul show that immunoreactivity was not abrogated when the phides in Thiolated Avibodies intra-framework 1 cysteine replacement mutations were 0634. The availability of engineered intra-Framework 1 selectively reduced and small, thiol-reactive payloads were disulphide bridges in thiolated Avibodies to reduction indi conjugated. cated that any of a number of thiol-reactive payloads could be 0640 To this end, Thiolated Avibodies payloaded with conjugated to the exposed and reduced cysteines. maleimide-PEG-methoxy (Example 6) were assessed for 0635 To demonstrate this ability, a maleimide-PEG immunoreactivity essentially as described in Example 4. methoxy payload was conjugated to the reduced engineered 0641. In all cases, Avibody-antigen complex formation, intra-framework 1 cysteines essentially as described herein. evidenced by a significant shortening of elution times in gel 0636. Following the reduction of Thiolated Avibodies and filtration, was observed (FIGS. 15A-C). In all cases, Avibody removal of reducing agent, an excess of maleimide-PEG4 alone eluted between 28-33 minutes, and Avibody-Antigen methoxy (mal-PEG-OMe) (Quanta Biodesign, OH, USA) complexes eluted at 10-25 minutes. As expected, complex was added at 20 equivalents per Avibody and allowed to react formation was not observed when Avibodies were incubated overnight at 4°C. Following PEGylation, unreacted PEG was with an irrelevantantigen. This result indicated that conjuga removed by extensive dialysis and success of PEG loading tion of relatively small payloads to reduced intra-Framework was determined by SDS-PAGE and mass spectroscopy. 1 cysteine replacement mutations in thiolated Avibodies did 0637 For SDS-PAGE analysis, 2 ug of total protein was not abrogate binding. loaded per well and resolved using NuPAGE 4-12% bis-tris gel in MES-SDS buffer (Invitrogen). The resulting protein 0642 Payloading to engineered intra-framework 1 cys bands were visualized using Coomassie Blue stain. Success teine replacement mutations is however not limited to PEG or ful PEGylation reproducibly exhibited an approximate mass PEG-like molecules. increase of 5 kDa per monomeric-chain (FIG. 13 A-C). 0643) To show that thiolated Avibodies could be conju 0638 For mass spectroscopy analysis, an Agilent esiTOF gated with payloads very different from the above exempli mass spectrometer with a MassPREP on-line desalting car fied PEG-conjugates, again without abrogating binding, tridge (Waters Corporation, USA) was used to record mass AVP04-50 was payloaded with detectable Label Europium. spectra of PEGylated Avibodies. The system was equilibrated (0644. The Eu" chelate of 1-(p-iodoacetamidobenzyl)di for 1 min with 5% CHCN, followed by an elution gradient ethylenetriamine-N-N',N',N',N'-pentacetic acid (DTPA) from 5-95% acetonitrile over 9 min. PEGylated Avibodies (PerkinElmer, Turku, Finland) was used to conjugate to typically eluted at 7 min. MassHunter software was used to reduced intra-Framework 1 cysteine replacement mutations determine average mass of the sample by deconvolution of in AVP04-50 according to manufacturer's instructions. the relevant m/z charge peaks produced. Data is reported in Briefly, protein was concentrated to 3 mg/ml in 50-100 mM Table 3 Summarizes the average monomeric-chain Avibody sodium hydrogen carbonate buffer+4 mM EDTA, pH 8.5. mass obtained following deconvolution of mass spectra. The Eu-DTPA was added at 30 times (Eu-DTPA: protein) molar formula mass of PEG is reported as 1239.44 g/mol, there excess to reduced AVP04-50. The reaction was completed fore an increase of at least 2478.88 mass units indicates full following 3-16 hrs at 4°C. Unreacted Eu-DTPA was sepa conjugation to engineered cysteines. Examples of typical rated from the protein by gel filtration on a Superdex(R) 200 mass spectrum for AVP07-71, AVP04-50, AVP07-88 and 10/300 column, pre-equilibrated with Tris-buffered saline, AVP02-101 are shown in FIGS. 14A-D respectively. pH 7.4. Each resulting fraction was diluted in Enhancement Solution (PerkinElmer, Turku, Finland) and assayed for TABLE 3 Europium counts using a Victor time resolved fluorometer. Peak Europium counts corresponding with peak protein frac PEG loading on thiolated Avibodies as determined by mass spectroscopy. tions were pooled and stabilized with 0.1% of highly pure Observed Mass PEGylated Mass Mass Increase PEG BSA, and stored at 4°C., protected from light. Concentration Construct (Da.) (Da.) (Da.) loaded of incorporated Eu-DTPA was determined by calculating Eu AVPO4-SO 2S685.99 281 66.84 2480.85 2 counts of the sample relative to a 100 nM Eu standard Sup AVPO4-70 27453.01 29933.96 248O.95 2 plied with the kit. AVPO4-74 26SO6.93 2.8987.58 2480.65 2 (0645. Eu3+-DTPA-conjugated AVP04-50 was shown to AVPO4-78 2681632 29296.94 2480.62 2 AVPO4-84 2681632 29297.16 2480.84 2 be immunoreactive by methods essentially as described in AVPO7-68 S456488 S7046.4 2481.52 2 Example 4. Eu3+-DTPA-conjugated AVP04-50 showed AVPO7-71 27597.92 3OO78.36 2480.44 2 specificity to BSM as indicated by the formation of an Anti AVPO7-88 26780.09 291.96.72 2416.63 2 gen: Avibody complex. This complex formation was evi AVPO7-89 26343.95 28824.74 2480.79 2 denced by a shortening of protein elution times in gel filtra AVPO2-101 23509.78 24749.OS 1239.27 1 tion chromatography on a Superdex(R) 200 10/300 column (FIG.16). 0646. Immunoreactivity of Eu3+-DTPA-conjugated Example 7 AVP04-50 was determined by cell binding assay. The labelled In Vitro Immunoreactive Assessment of Payload Avibody was incubated with TAG72 positive (LS174T) and Conjugated Thiolated Avibodies negative (SK-OV-3) cell lines. Following the incubation period (1 hr., ambient temperature), cells were washed exten 0639. Thiolated Avibodies could be expressed, purified, sively and assayed for europium activity. Europium labelled and were shown to be immunoreactive in their native (un AVP04-50 showed intact immunoreactivity and antigen conjugated) state. Data reported above clearly indicated that specificity as indicated by a significant increase in fluorescent US 2012/0164068 A1 Jun. 28, 2012 47 intensity on LS174T (TAG72 positive) compared with SK 8.3 In vivo Biodistribution Using LS-174T Xenografts OV3 (TAG72 negative) cells line (FIG. 17). 0652 For the mouse xenograft model, female, athymic 0647. Taken together, these data suggest that it is possible nu/nu mice (Charles River Laboratories), 6-8 weeks old, were to site-specifically conjugate payloads to engineered intra injected with LS-174T cells (ATCC) (10) subcutaneously in framework 1 cysteine replacement mutations in Thiolated the flank, and tumours were allowed to grow for about 10 days Avibodies without abrogating binding to antigen. prior to study. Mice bearing LS-174T xenografts were injected intravenously with a mixture (200 ul) of 370 kBq of 'I and 150 kBq of '''In-labelled AVP04-07 or AVP04-50 Example 8 (2-6 ug of total protein) for biodistribution studies. Mice (4-6 mice/group) were euthanized at various time points and the In Vivo Performance of Radiolabelled AVP04-XX tumour, blood and major organs were collected, weighed and Avibodies counted. The counts were corrected for crossover of ''In counts in the 'I channel. Percentages of the injected dose (0648. The Thiolated Avibody AVP04-50, was used in vivo per gram of tissue (% ID/g) were calculated for each radio in a mouse Xenograft model to show that Intra-Framework 1 nuclide. Engineered Cysteines mutations in Thiolated Avibodies did 8.4. In Vivo b3iodistribution of Avibodies not affect in vivo stability or performance. Biodistribution 0653. The biodistribution of ''I and '''In-DOTA using ('I or '''In) of AVP04-50 was compared to that of the labelled Avibodies was measured in athymic mice bearing “un-mutated” parental AVP04-07. LS-174T xenografts. 8.1 Radiolabelling of Avibodies with 'I 0654) The performance of AVP04-07 and AVP04-50 Avi 0649. Radioiodination of the AVP04-07 and AVP-04-50 bodies in Vivo were absolutely comparable, Suggesting the Avibodies with I (PerkinElmer) was performed using the introduction of a modeling-defined intra-Framework 1 disul standard Iodogen method (Yazaki et al., 2001). The required phide bridge did not destabilize or negatively impact Avibody volume (5-10 uL) of NaI(26 mBq) was added to 200g of performance (FIGS. 18A and B). Avibody in a tube pre-coated with 20 Lug Iodogen (Pierce). 0655 The two radiotracers ('I—and '''In) were cleared After incubation at RT for 3 min, the labelled material was from the blood in a similar manner with only a marginal purified by FPLC using a Superdex-75 or 200 column as difference observed between clearance of AVP04-50 and described above. The column eluate was fractionated and AVP04-07. Approximately 50% cleared by 1 hpost-injection counted, after which peak fractions were pooled and used for and about 6-12% still in circulation at 4 h. As expected for in vitro and in vivo studies. Radiolabelling yields were typi proteins of this size, there was considerable kidney uptake cally 80-100%. (>100% ID/g at 24 h) for the '''In-labelled, but not for the 8.2 Conjugation of NHS-DOTA to Avibodies and Radiola 'I-labelled Avibody, demonstrating that the kidney was the belling with '''In major route of clearance. 0650 AVP04-07 was conjugated to NHS-DOTA (1,4,7, 0656. For '''In-AVP04-07, there was significant uptake in 10-Tetraazacyclododecane-1,4,7,10-tetraacetic acid mono the tumour, with over 25% ID/g observed as early as 4h post (N-hydroxysuccinimide ester). AVP04-07 and AVP04-50 injection and more than 20% ID/g still in the tumour at 48 h. were concentrated to 3-6 mg/mL using an Amicon Centricon The tumour uptake of '''In-AVP04-50 matched that of ''In YM-10 (10 kDa MWCO) centrifugal filter device (Millipore AVP04-07 in the 0-4 hour time period post injection, but then Corp, Bedford, Mass., USA) by centrifugation at 4,000 rpm at surpassed that of ''In-AVP04-07 to reach a maximum 4°C. (Allegra X15R, SX4750 rotor, Beckman Coulter). To tumour uptake of greater than 30% ID/g at 24 hours. Tumour remove metal contaminants, the protein (0.4 mL) was dia uptake was for both '''In-AVP04-07 and '''In-AVP04-50 lyzed against 14 Volumes conjugation buffer (0.1M sodium remained comparable at 48 hrs post injection. Tumour to bicarbonate, 5 mM 9 diethylenetriamene pentaacetate blood ratio for '''In-AVP04-07 was >50:1 at 24 h. In the case (DTPA), pH 8.5) for 2 hours, using a modified ultrafiltration of ''In-AVP04-50, tumour to blood ratio was increased to cell with a Biomax ultrafiltration membrane (Millipore, >60:1 at 24 h. Iodine-125 labelled Avibodies exhibited some PBQK02510). 19 uL of NHS-DOTA (B-280; Macrocyclics, what lower tumour to blood ratios and tumour uptake (about Dallas, Tex., USA) at 10 mg/mL (0.19 mg, 250 nmole) in 17% and 10% ID/g at 4 and 48 h respectively for AVP04-07 plasmagrade water (Fisher Scientific, Waltham, Mass., USA) and 19% and 10% ID/g at 4 and 48 h respectively for AVP04 was added to the Avibodies (2.5 mg. 48 nmole) at a 5-fold 50). As expected, some '''In was retained in the spleen, liver molar excess in the ultrafiltration cell and stirred for 1 hat RT. and carcass, while 'I-labelled Avibody was not retained in The protein was Subsequently dialyzed against 14 volumes of these tissues. 250 mM sodium acetate, pH 7.2 and stored at 4°C. 0657. In addition to the biodistribution of AVP04-07 and 0651 Radiometal labelling of DOTA-conjugated Avibody AVP04-50, a thiolated HER2-specific Avibody, AVP07-63 (AVP04-07-DOTA and AVP04-50-DOTA) was performed (SEQID NO: 65), was also shown to be as efficient in vivo as using 'InCl2 (Trace Life 11 Sciences, Denton,Tex., USA). its non-thiolated parental AVP07-17. In a typical experiment, 19 mBq of ''InCl2 was diluted with 0658 Inthese in vivo experiments, AVP07-17 and AVP07 additional 0.1M HCl and added to 125 ug DOTA-conjugated 63 were radiolabelled with “I using the chloramine T AVP04-07 in 0.25 Mammonium acetate pH 7.0 (final pH method ('I: protein ratio, 1:10), essentially as described adjusted to 5.5). After incubation at 43° C. for 45 min, the (Adams et al. 1993). The quality and immunoreactivity of the solution was adjusted to 0.1 mM DTPA to bind any residual radiopharmaceuticals were evaluated by SDS-PAGE and in a 'In and incubated at RT for an additional 10 min. Radiola live cell-binding assay as described (Adams et al. 1993). belling yields were typically 70-90%. The labelled material CB.17 Icr scid mice, 6-8 weeks of age, were implanted sub was then purified by HPLC using a Superdex-75 or 200 cutaneously on the abdomen with SKOV3 cells (2.5x10). column, and the column eluate was fractionated and counted. When the tumours had achieved a size of 50-200 mg (ap