MATERIALS AND METHODS RELATING TO LINKERS FOR USE IN PROTEIN DRUG CONJUGATES

Abstract

The present invention relates to protein drug conjugates, methods of manufacturing the same and their use in therapy. In particular, the present invention relates to protein drug conjugates comprising a globular protein, an improved linker and a drug for use in targeted drug delivery applications.

Claims

1. A protein drug conjugate comprising: a globular protein; a linker; a drug; and wherein the linker comprises a nitrogen containing heterocyclic aromatic ring comprising a vinyl substituent.

2. The protein drug conjugate according to claim 1, wherein the globular protein is an albumin.

3. The protein drug conjugate according to claim 2, wherein the albumin is human albumin.

4. The protein drug conjugate according to claim 1, wherein the globular protein is an antibody or fragment thereof.

5. The protein drug conjugate according to claim 4, wherein the antibody is a monoclonal antibody.

6. The protein drug conjugate according to claim 1, wherein the protein drug is selected from the group consisting of a cytotoxin, a therapeutic peptide, and a polypeptide.

7. The protein drug conjugate according to claim 6, wherein the cytotoxin is a biologically active cytotoxic material.

8. The protein drug conjugate according to claim 6, wherein the cytotoxin is an anticancer drug.

9. The protein drug conjugate according to claim 1, wherein the linker comprises a molecule having the general formula: ##STR00013## wherein: X and Y are independently selected from CH or N; R.sub.1 is selected from: (CH.sub.2).sub.n—C(O)—R, or, (CH.sub.2).sub.m—Z—R, or, (CH.sub.2).sub.m—Z—C(O)—R, or, (CH.sub.2).sub.n—C(O)—Z—R, or, (CH.sub.2).sub.m—Z—(CH.sub.2)—C(O)—R, or, (CH.sub.2).sub.m—Z—(CH.sub.2)—C(O)—Z—R, or, (CH.sub.2).sub.m—Z—C(O)—(CH.sub.2).sub.n—Z—(CH.sub.2).sub.n—C(O)—Z—R, or, (CH.sub.2).sub.n—CH(CO.sub.2R.sub.2).sub.2, or, (CH.sub.2).sub.m—Z—(CH.sub.2).sub.2CH(CO.sub.2R).sub.2, or, (CH.sub.2).sub.n—Z.sub.1, wherein: Z is independently selected from NH, O or S, Z.sub.1 is independently selected from N.sub.3 or OH, n is any integer from 0 to 10, m is any integer from 0 to 10, and R is H, OH, an amine or a poly(alkylene glycol) group; and R.sub.2 and/or R.sub.3 are selected from the same group of molecules as R.sub.1 or R.sub.2 and/or R.sub.3 are selected from hydrogen or an electron withdrawing group, such as halogen (F, Cl, or Br), —NO.sub.2, —CO.sub.2H, —CO.sub.2R.sub.4, COR.sub.4, —CHO, —CN, —CF.sub.3, —SO.sub.2NR.sub.4R.sub.5 where R.sub.4 and R.sub.5 are independently selected from hydrogen or C.sub.1-10 alkyl; or; R.sub.2 and/or R.sub.3 are selected from hydrogen, alkyl or phenyl; or R.sub.2 and R.sub.3 together form a fused (hetero) aromatic ring substituent which include, but is not limited to, an indole, indazole, benzimidazole, quinoline, isoquinoline, aziridine or a purine.

10. The protein drug conjugate according to claim 9 wherein the linker comprises a molecule having the general formula (I) with the proviso that when Z is O and R1, and optionally R2 and R3 when selected from the same group of molecules as R1, is in the 2 or 6 position on the ring, m is at least 3.

11. The protein drug conjugate according to claim 1, wherein the linker is a 4-vinylpyridine.

12. The protein drug conjugate according to claim 1, wherein the linker may be represented by the following formulae, ##STR00014## wherein: R.sub.1 is selected from; (CH.sub.2).sub.n—C(O)—R, or (CH.sub.2).sub.m—Z—R, or (CH.sub.2).sub.m—Z—C(O)—R, or (CH.sub.2).sub.n—C(O)—Z—R, or (CH.sub.2).sub.m —Z—(CH.sub.2).sub.n—C(O)—R, or (CH.sub.2).sub.m —Z—(CH.sub.2).sub.n—C(O)—Z—R, or (CH.sub.2).sub.m —Z—C(O)—(CH.sub.2).sub.n—Z—(CH.sub.2).sub.n—C(O)—Z—R, or, (CH.sub.2).sub.n—CH(CO.sub.2R.sub.2).sub.2, or, (CH.sub.2).sub.m—Z—(CH.sub.2).sub.2CH(CO.sub.2R).sub.2, or, (CH.sub.2).sub.n—Z.sub.1; wherein: R.sub.2 is selected from the same group of molecules as R.sub.1, hydrogen, an electron withdrawing group, alkyl or phenyl group, Z is independently selected from NH, O, or S, Z.sub.1 is independently selected from N.sub.3 or OH, n is any integer from 0 to 10, m is any integer between 0 and 10, R is a hydrogen (H), hydroxide (OH), amine or a poly-(alkylene glycol) group.

13. The protein drug conjugate according to claim 12, wherein the linker comprises a molecule having the general formula (III) with the proviso that when Z is O, m is at least 3.

14. The protein drug conjugate according to claim 1, wherein the linker comprises a poly-(alkylene glycol) group.

15. The protein drug conjugate according to claim 14, wherein the poly-(alkylene glycol) group is a polyethylene glycol (PEG).

16. The protein drug conjugate ADC according to claim 14, wherein the poly-(alkylene glycol) structure is provided with at least one reactive functional group including hydroxy, amine, carboxylic acid, alkyl halide, azide, succinimidyl, or thiol groups.

17. The protein drug conjugate according to claim 1, which further comprises an extender linker.

18. The protein drug conjugate according to claim 17, wherein said extender linker is enzyme cleavable.

19. A method of producing the protein drug conjugate of claim 1, the method comprising contacting the protein with the linker which is bound to the drug.

20. The method of claim 19, the method comprising contacting the protein having at least one reactive thiol group with the linker which comprises a functionalising reagent comprising a nitrogen containing heterocyclic aromatic ring having a vinyl substituent capable of reacting with at least one free thiol group of the protein, wherein the linker functionalising reagent is covalently linked to a poly-(alkylene glycol) molecule.

21. The method according to claim 19, the method comprising an initial step of reacting a precursor functionalising reagent comprising a nitrogen containing heterocyclic aromatic ring having a vinyl substituent with a poly-(alkylene glycol) molecule to produce the functionalising reagent.

22. The method according to claim 19, the method further comprising an initial step of modifying the protein to produce a variant polypeptide having a thiol group at at least one desired position of the polypeptide.

23. (canceled)

Description

BRIEF DESCRIPTION OF THE FIGURES

[0121] FIG. 1. Relative rates of reaction for example linkers in accordance with the present invention and glutathione.

[0122] FIG. 2. RP-HPLC analysis of PL13 free acid reactivity with N-acetyl cysteine (NAC) after 24 hours at three different pHs (7.0, 7.5 and 8.0).

[0123] FIG. 3. MS data to demonstrate formation of PL-13-NAC Adduct at pH 7.0.

[0124] FIG. 4. MS data to demonstrate formation of PL-13-NAC Adduct at pH 8.0.

[0125] FIG. 5. RP-HPLC analysis of PL13 free acid reactivity with NAC at pH 8.0.

[0126] FIG. 6. RP-HPLC analysis of PL13 free acid reactivity with a mixture of amino acids (Tyr/Lys/His/NAC) at pH 7.0.

[0127] FIG. 7. MS data to demonstrate formation of PL13-NAC in reactivity with amino acid mixture at pH 7.0.

[0128] FIG. 8. RP-HPLC analysis of PL13 free acid reactivity with 10 equivalents of Lysine at pH 7.0.

[0129] FIG. 9. RP-HPLC analysis of PL13 free acid reactivity with 10 equivalents of aspartic acid.

[0130] FIG. 10. Structure of PL13-val-cit-4-aminobenzoyl-MMAE.

[0131] FIG. 11. MS analysis of PL13-val-cit-4-aminobenzoyl-MMAE.

[0132] FIG. 12. HIC and UV-Vis profiles of Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE after 1 hour of reaction.

[0133] FIG. 13. HIC and UV-Vis profiles of the Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE conjugate at 1, 8 and 16 hours of incubation with 1.25 molar excess of drug-linker over thiol.

[0134] FIG. 14. HIC profile of the Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE conjugate at 16 hours of incubation with 5 molar excess of drug-linker over thiol.

[0135] FIG. 15. HIC profile of Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE obtained via conjugation of 10 molar excess of PL13-val-cit-4-aminobenzoyl-MMAE after 16 hours of incubation.

[0136] FIG. 16. PLRP profile of Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE conjugated at 10 molar excess of PL13-val-cit-4-aminobenzoyl-MMAE after 16 hours of incubation.

[0137] FIG. 17. PLRP profile of double desalted Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE.

[0138] FIG. 18. HIC analysis of the Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE conjugate in the presence of NAC at 0, 24 and 48 hours.

[0139] FIG. 19. PLRP profile of the Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE conjugate in the presence of NAC at 0, 24 and 48 hours.

[0140] FIG. 20. HIC analysis of the Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE conjugate in the presence of NAC at 0, 24 and 48 hours.

[0141] FIG. 21. Analysis of Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE (A) and Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE (B) conjugates by SEC chromatography.

[0142] FIG. 22. Non-reducing SDS-PAGE analysis of Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE and Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE.

[0143] FIG. 23. Structure of PL13-NH-PEG4-OSu linker.

[0144] FIG. 24. ESI-MS analysis of PL13-NH-PEG4-COOH linker intermediate.

[0145] FIG. 25. Analysis of Trastuzumab cross-linking via SMCC or PL13-NH-PEG4-OSu linker by reducing SDS-PAGE.

[0146] FIG. 26. Structure of PL13-NH-PEG4- val-cit-4-aminobenzoyl-MMAE linker.

[0147] FIG. 27. HIC profile of Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE (A) and Trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE (B).

[0148] FIG. 28. PLRP profile of Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE (A) and Trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE (B).

[0149] FIG. 29. DAR calculation for Trastuzumab-PL13-NH-PEG4-val-cit-MMAE and Trastuzumab-mal-val-cit-MMAE based on the PLRP elution profile of FIG. 28.

[0150] FIG. 30. SEC analysis of Trastuzumab-mal-val-cit-MMAE (A) and Trastuzumab-PL13-NH-PEG4-val-cit-MMAE (B).

[0151] FIG. 31. Representative HIC (A) and PLRP (B) profiles of Trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE achieved after process optimisation at 1 mg scale.

[0152] FIG. 32. HIC profiles of Trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE (A) and Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE (B) achieved at 150 mg scale.

[0153] FIG. 33. SEC profiles of Trastuzumab-PL13-NH-PEG4-val-cit-aminobenzoyl-MMAE (A) and Trastuzumab-maleimide-val-cit-aminobenzoyl-MMAE (B) achieved at 150 mg scale.

[0154] FIG. 34. HIC profile of Thiomab-PL13-NH-PEG4-val-cit-MMAE at pH 7.0 (solid line) and pH 7.5 (dashed line).

[0155] FIG. 35. PLRP profile to demonstrate de-drugging for trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE.

[0156] FIG. 36. PLRP profile to demonstrate de-drugging for trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE with a DAR of 3.8.

[0157] FIG. 37. In vitro stability of ADCs: trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE, Mal-ADC (A) vs trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE, PL13-ADC (B)

[0158] FIG. 38. In vivo stability of ADCs: PL13-ADC vs Mal-ADC

[0159] FIG. 39. Cell kill data for SKBR3 cell line (A), BT474 cell line (B) and JIMT-1 cell line (C).

[0160] FIG. 40. Multi-dose xenograft data to demonstrate ADC toxicity.

[0161] FIG. 41. Multi-dose xenograft data to show ADC effect on tumour growth.

[0162] FIG. 42. Tumour maturity histopathology data

[0163] FIG. 43. Comparative data for conjugation of 20 KDa PEG-PL12 and PEG-PL13 versus 20 KDa PEG-maleimide to reduced albumin at pH 7.4

[0164] FIG. 44. Optimisation of PL11 (A), PL12 (B) and PL13 (C) conjugation to albumin.

[0165] FIG. 45. Analysis of PL11 (A), PL12 (B) and PL13 (C) conjugation to albumin by ESI-MS.

[0166] FIG. 46. Stability of Albumin-PL-12 conjugate in the presence of 1 mM GSH demonstrated by ESI-MS.

[0167] FIG. 47. Structure of PL13-NH-PEG4-val-cit-4-aminobenzoyl-Doxorubicin linker.

[0168] FIG. 48. Comparative data for conjugation of PL13-NH-PEG4-val-cit-4-aminobenzoyl-doxorubicin linker to albumin, thioalbumin (single mutant) and thioalbumin (double mutant).

[0169] FIG. 49. Non-reducing SDS-PAGE analysis of albumin-PL13-NH-PEG4-val-cit-4-aminobenzoyl-doxorubicin conjugates. (1) Thioalbumin-double mutant-DOX; (2) Thioalbumin-single mutant-DOX; (3) Albumin-DOX; (M)-protein marker.

[0170] FIG. 50. SEC analysis of albumin-PL13-NH-PEG4-val-cit-4-aminobenzoyl-doxorubicin conjugates: Albumin (A) and albumin-DOX conjugate (B); Thioalbumin (single mutant) (C) and Thioalbumin-DOX (single mutant) (D); Thioalbumin (double mutant) (E) and Thioalbumin-DOX (double mutant) (F).

[0171] The methods of the present invention are capable of providing protein drug conjugates, such as ADCs or albumin drug conjugates, in accordance with the present invention which comprise an antibody or albumin bound to a linker, which in turn is bound to a drug, such as a cytotoxin or a therapeutic peptide or polypeptide.

[0172] In the present invention, references to antibodies include immunoglobulins whether natural or partly or wholly synthetically produced. The term also covers any polypeptide or protein comprising an antigen binding domain. Antibody fragments are also contemplated which comprise antigen binding domains including Fab, scFv, Fv, dAb, Fd fragments, diabodies, triabodies or nanobodies. It is possible to take monoclonal and other antibodies and use techniques of recombinant DNA technology to produce other antibodies or chimeric molecules which retain the specificity of the original antibody. Such techniques may involve introducing DNA encoding the immunoglobulin variable region, or the complementarity determining regions (CDRs), of an antibody to the constant regions, or constant regions plus framework regions, of a different immunoglobulin. See, for instance, EP 0 184 187 A, GB 2,188,638 A or EP 0 239 400 A. Antibodies can be modified in a number of ways and the term should be construed as covering any specific binding member or substance having an antibody antigen-binding domain with the required specificity. Thus, this term covers antibody fragments and derivatives, including any polypeptide comprising an immunoglobulin binding domain, whether natural or wholly or partially synthetic. Chimeric molecules comprising an immunoglobulin binding domain, or equivalent, fused to another polypeptide are therefore included. Cloning and expression of chimeric antibodies are described in EP 0 120 694 A and EP 0 125 023 A.

[0173] In the prior art it has been shown that fragments of a whole antibody can perform the function of binding antigens. Examples of binding fragments are (i) the Fab fragment consisting of VL, VH, CL and CH1 domains; (ii) the Fd fragment consisting of the VH and CH1 domains; (iii) the Fv fragment consisting of the VL and VH domains of a single antibody; (iv) the dAb fragment (Ward, E. S. et al., Nature 341, 544-546 (1989)) which consists of a VH domain; (v) isolated CDR regions; (vi) F(ab′)2 fragments, a bivalent fragment comprising two linked Fab fragments (vii) single chain Fv molecules (scFv), wherein a VH domain and a VL domain are linked by a peptide linker which allows the two domains to associate to form an antigen binding site (Bird et al, Science, 242; 423-426, 1988; Huston et al, PNAS USA, 85: 5879-5883, 1988); (viii) bispecific single chain Fv dimers (PCT/U.S.92/09965) and (ix) “diabodies”, multivalent or multispecific fragments constructed by gene fusion (WO 94/13804; Holliger et al, P.N.A.S. USA, 90: 6444-6448, 1993). Fv, scFv or diabody molecules may be stabilised by the incorporation of disulphide bridges linking the VH and VL domains (Reiter et al, Nature Biotech, 14: 1239-1245, 1996). Minibodies comprising a scFv joined to a CH3 domain may also be made (Hu et al, Cancer Res., 56: 3055-3061, 1996). Accordingly, such binding fragments are contemplated by the present invention.

[0174] In preferred embodiments, the methods disclosed herein employ reagents and conditions that are well adapted for binding a protein, such as an antibody, via a linker to a drug, such as a cytotoxin. In particular, the reaction conditions that are used in the present method helps to avoid the problems that tend to occur when using prior art reagents such as maleimide, which have a tendency to produce a mixture of different products with a range of different properties. More especially, as can be seen from the experimental data below, methods of producing protein drug conjugates, such as ADCs, which utilise linkers according to the present invention, avoid the problems associated with maleimide crosslinking.

[0175] As mentioned above, the protein of interest (e.g. the antibody) for the protein drug conjugate (e.g. ADC) composition may be bound to the linker using existing thiol groups or by introducing thiol groups in an initial step of the method, for example by reacting one or more functional groups of the protein (e.g. antibody) to produce a thiol group, or by introducing a thiol group or a precursor thereof into the protein (e.g. antibody). By way of example, this may involve the step of introducing a cysteine residue into the protein (in the example an antibody) at a site where it is desired to bind the linker to the protein (e.g. antibody). This may be useful in situations where a convenient cysteine residue for reaction according to the present invention is not present in a starting or wild-type polypeptide/protein. Conveniently, this may be achieved using site directed mutagenesis of the protein, such as an antibody polypeptide, the use of which is well established in the art.

[0176] Alternatively or additionally, an initial reduction step may be required where the protein is commercial grade albumin as the majority of cysteine thiols present are capped, as discussed in further details below

[0177] Experimental Data and Discussion

[0178] The following experimental data was mainly produced utilising the linker molecule identified as, and referred to herewith as, PL13, in its free acid or NH -pegylated form, as shown below.

##STR00010##

1. Demonstration of PL11, PL12 and PL13 Linker Reactivity with Glutathione.

[0179] Glutathione contains a thiol group which is readily available for conjugation, and can provide a good model for proteins to assess the suitability of linkers for use in the present invention for utility in protein drug conjugates, such as ADCs. FIG. 1 shows three linker group examples according to the present invention and demonstrates their ability to conjugate to the thiol groups present in glutathione. The examples in FIG. 1 are PL11, PL12 and PL13, and their structures are shown below.

##STR00011##

2. Determination of PL13 Free Acid Selectivity

[0180] The data given below demonstrates that linkers in accordance with the present invention show specificity for cysteine groups. PL13 free acid is identified as;

##STR00012##

Reactivity of PL13 Free Acid with N-Acetyl Cysteine at pH 7.0, 7.5 and 8.0

[0181] PL13 free acid was reacted with 2 molar equivalents of N-acetyl cysteine (NAC) at three separate pHs: 7.0, 7.5 and 8.0. Reactions were carried out in a methanol/phosphate buffered saline (PBS) solution (ratio of 9:1), buffered to the appropriate pH at room temperature (RT). RP-HPLC analysis with a gradient of 1% to 50% B (Acetonitrile with 0.1% TFA) over 15 minutes and detection at 254 nm was undertaken to monitor addition of free thiol to the vinyl group of PL13 over time.

[0182] Method:

[0183] 210 μL of 2.61 mM PL13 free acid in methanol was mixed with 22 μL of 50 mM NAC in buffer to give a final concentration of 2.37 mM PL13 free acid and 4.74 mM NAC.

[0184] Results:

[0185] The PL13-NAC adduct was eluted at retention time of 5.1 minutes at all pHs after 24 hours (see FIG. 2). The amount of product (PL13-NAC adduct) was similar at pH 7.0 and 8.0. The reactivity of PL13 free acid at pH 7.5 was slower. In addition, an unidentified peak was eluted at 4 minutes.

[0186] ESI-MS analysis confirmed the presence of the PL13-NAC adduct at all analysed pHs. MS trace data is included herewith to show this analysis as performed at pH 7.0 (FIG. 3.), and pH 8.0 (FIG. 4.).

Reactivity of PL13 Free Acid with N-Acetyl Cysteine at pH 8.0

[0187] PL13 free acid was reacted with 2 molar equivalents of NAC in methanol/PBS solution (9:1), pH 8.0. The progress of the reaction was monitored by RP-HPLC using a gradient of 1% to 50% B over 15 minutes with detection at 254 nm. Samples were analysed at 0, 1, 4 and 72 hours of incubation at RT (see FIG. 5.).

[0188] Method:

[0189] 100 μL of 2.61 mM PL13 free acid in methanol was mixed with 11 μL of 50 mM NAC in buffer at pH 8.0 to give a final concentration of 2.35 mM PL13 free acid and 5 mM NAC.

[0190] Results:

[0191] Addition of PL13 free acid to NAC is slow over the first 4 hours. After 72 hours of incubation ˜70% of PL13 free acid was converted to product (PL13-NAC adduct) at pH 8.0.

Reactivity of PL13 Free Acid with 10 Equivalents of Tyrosine, Histidine, Lysine and N-Acetyl cysteine at pH 7.0

[0192] PL13 free acid was challenged with 10 equivalents of each amino acid (tyrosine (Tyr), histidine (His), Lysine (Lys) and N-acetyl cysteine (NAC)) at pH 7.0. The reaction was analysed by RP-HPLC using a gradient of 1 to 50% B over 15 minutes with detection set at 254 nm.

[0193] Method:

[0194] 50 μL of 2.61 mM PL13 in methanol was mixed with 260 μL of 20 mM Tyr/His/Lys and NAC in buffer to give a final concentration of 0.42 mM PL13 and 4.2 mM of each of Tyr, His, Lys and NAC.

[0195] Results:

[0196] PL13 free acid reacted selectively with NAC in the presence of Tyr, His and Lys at pH 7.0 (see FIG. 6). FIG. 7 shows the MS data to demonstrate that only the PL13 NAC adduct was formed.

[0197] The addition reaction between PL13 and NAC increased significantly at 10 molar excess of amino acid. Conversion of PL13 free acid to the desired PL13-NAC adduct is 90% complete after 4 hours at room temperature (RT) and is fully complete in <18 hours.

[0198] The results obtained demonstrate the selectivity of the free acid form of the linker molecule in accordance with the present invention for cysteine reactivity via the vinyl group of the linker molecule.

RP-HPLC Analysis of PL13 Free Acid Reactivity with Lysine.

[0199] PL13 free acid was challenged with 10 equivalents of Lys in PBS buffer, pH 7.4 at RT. The reaction was analysed by RP-HPLC using a gradient of 12 to 50% acetonitrile in water over 30 minutes with detection set at 270 nm.

[0200] Method:

[0201] 10 μL of 4 mM PL13 in methanol was mixed with 50 μL of 8 mM Lys solution and 40 μL of PBS buffer pH 7.4 to give a final concentration of 0.4 mM PL13 and 4.0 mM Lys. The reaction was analysed by RP-HPLC using a gradient of 12 to 50% acetonitrile in water over 30 minutes with detection set at 270 nm.

[0202] Results:

[0203] PL13 free acid did not react with Lys at pH 7.4 as no peak is observed in the RP-HPLC analysis corresponding to a PL13-Lys adduct (see FIG. 8).

[0204] The results obtained demonstrate that no adduct is formed by the incubation of PL13 free acid with Lys. This data supports the cysteine group selectivity of the free acid form of linkers of the present invention. It should be appreciated that no such cysteine specificity is exhibited by maleimide and as such the linkers of the present invention present a benefit in this regard.

[0205] While the above data shows the cysteine specificity exhibited by the linkers of the present invention, it is readily appreciated that lysine specificity may alternatively be achieved if the linker is modified by methods well known in the art.

Reactivity of PL13 Free Acid with Aspartic Acid

[0206] PL13 free acid was treated with 10 equivalents of aspartic acid (Asp) at pH 7.0 and at RT. Analysis was performed by RP-HPLC at gradient of 1% to 50% B over 15 minutes with detection at 254 nm.

[0207] Method:

[0208] 50 μL of 2.61 mM PL13 free acid in methanol was mixed with 280 μL of 50 mM Asp in buffer to give a final concentration of 0.42 mM PL13 free acid and 4.2 mM Asp.

[0209] Results:

[0210] PL13 free acid is stable in the presence of Asp at pH 7.0 over 18 hours at RT (see FIG. 9).

[0211] MS spectra substantiates that no adduct is formed by the incubation of PL13 free acid with Asp (data not shown). Again, this data supports the cysteine group selectivity of the linkers of the present invention. It should be appreciated that no such cysteine specificity is exhibited by maleimide and as such the linkers of the present invention present a benefit over this.

3. Synthesis of PL13-Val-Cit-4-aminobenzoyl-MMAE Cytotoxic Drug Linker

[0212] PL13-val-cit-4-aminobenzoyl-MMAE was synthesised by a fragment approach, which will be familiar to the person skilled in the art.

[0213] Method:

[0214] PL13 free acid was coupled to the free amino terminal of H-val-cit-4-aminobenzoyl-MMAE, an exemplary cytotoxin, via a HOBt active ester method.

[0215] Results:

[0216] 3.1 mg of PL13-val-cit-4-aminobenzoyl-MMAE (see FIG. 10) was synthesised. The material was solubilised in dimethyl-acetamide (DMA) to afford a drug linker solution at a concentration of 50 mM.

[0217] RP-HPLC confirmed purity of the PL13-val-cit-4-aminobenzoyl-MMAE linker (data not shown).

[0218] ESI-MS analysis confirmed the identity of the cytotoxic drug linker (expected MW is 1296.67 Da, experimental result gave MW of 1296.5 Da) (see FIG. 11).

4. Generation of Trastuzumab-PL13-Val-Cit-4-aminobenzoyl-MMAE and Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE

[0219] 20 mg of Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE conjugate, an exemplary ADC of the present invention, and 20 mg of Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE conjugate, an ADC comprising a known maleimide linker, were produced.

Conjugation of PL13-val-cit-4-aminobenzoyl-MMAE and Maleimide-val-cit-4-aminobenzoyl-MMAE to Trastuzumab

[0220] Method:

[0221] Trastuzumab was reduced to allow conjugation of 3-4 drugs per trastuzumab molecule. A detailed method for the reduction of trastuzumab is not provided as this should be well known by the person skilled in the art.

[0222] Maleimide-val-cit-4-aminobenzoyl-MMAE was conjugated to Trastuzumab at 1.25 molar excess over free thiol at pH 7.0. The reaction was performed for 1 hour at RT. The conjugation reaction was quenched by an excess of NAC. Analysis of the Trastuzumab conjugate was accomplished by Hydrophobic Interaction Chromatography (HIC) using a Tosoh butyl-NPR (4.6×3.5, 2.5 μm) column and by UV-VIS spectroscopy. MMAE has a distinctive UV absorbance at 248 nm (ε.sub.248=1500 M.sup.−1 cm, ε.sub.280=15900 M.sup.−1 cm).

[0223] PL13-val-cit-4-aminobenzoyl-MMAE was coupled to Trastuzumab at 1.25, 2.5, 5 and 10 molar excess over free thiol. Reactions were carried out for 16 hours at pH 7.0. Conjugation reactions were analysed by both HIC (separation on a Tosoh butyl-NPR column) and PLRP chromatography (separation on a PLRP column—2.1 mm×5 cm, 5 μm). Trastuzumab conjugates were examined by UV-VIS spectroscopy. Both the MMAE cytotoxic drug and PL13 linker contribute to UV absorbance at 248 nm.

[0224] Results:

[0225] Reaction of maleimide-val-cit-4-aminobenzoyl-MMAE with Trastuzumab was completed within 1 hour using a ratio of 1.25 drug over free thiol (see FIG. 12). The numbers in FIG. 12 designate the amount of drug conjugated to a full length antibody. The inlet is the UV-Vis profile which represents an increase of absorbance at 248 nm due to conjugation of maleimide-val-cit-4-aminobenzoyl-MMAE to Trastuzumab. 10 mg of the trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE conjugate (Drug to Antibody Ratio (DAR) 2.2) was obtained and set aside for in vitro studies, as discussed further below.

[0226] The rate and efficiency of the PL13-val-cit-4-aminobenzoyl-MMAE reaction is much slower than the maleimide-val-cit-4-aminobenzoyl-MMAE due to low solubility of the linker (see FIG. 13). Conjugation was performed at 1.25 molar excess of drug-linker over thiol. The inlet is the UV-Vis profile which depicts an increase of absorbance at 248 nm due to coupling of PL13-val-cit-4-aminobenzoyl-MMAE to Trastuzumab within 16 hours. A slow increase in the rate of PL13-val-cit-4-aminobenzoyl-MMAE conjugation to Trastuzumab was observed by applying 5 molar excess of drug-linker over thiol group (see FIG. 14). In FIG. 14, the numbers designate the amount of drug conjugated to a full length antibody. DL indicates unconjugated drug.

[0227] Conjugation of PL13-val-cit-4-aminobenzoyl-MMAE to Trastuzumab at 10 molar excess and in the presence of 50% propylene glycol resulted in 90% yield within 4 hours (see FIG. 15 and FIG. 16). In FIG. 15, the numbers designate the amount of drug conjugated to a full length antibody. DL indicates unconjugated drug. In FIG. 16, the numbers designate the amount of drug conjugated to light (L) or heavy (H) chain of the antibody. 7.2 mg of Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE conjugate (DAR 3.03), was obtained and set aside for in vitro studies, as discussed further below.

Determination of Drug Antibody Ratio (DAR) for Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE and Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE.

[0228] HIC and PLRP chromatography are often applied to characterize average drug-load and drug-load distribution for cysteine-linked ADCs. Determination of average drug-load and drug-load distribution is a crucial attribute as it effects the potency and pharmacokinetics of the ADC.

[0229] Results:

[0230] HIC characterisation of Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE resulted in a DAR calculation of 2.2 with 1.8% unconjugated Trastuzumab (see FIG. 12). The DAR calculation from the HIC profile is determined by well known methods in the art, as discussed below in relation to FIG. 27.

[0231] HIC analysis of Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE resulted in peaks which did not resolve well enough to support DAR determination (see FIG. 15). However, this data is included here for completeness. HIC was used to calculate the amount of Trastuzumab with DAR 0 calculated to be ˜8.4%.

[0232] Separation of dithiothreitol (DTT) reduced Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE via a PLRP column afforded well resolved peaks (FIG. 16) corresponding to unconjugated or drug conjugated antibody light and heavy chains. A DAR of 3.0 was calculated for Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE (see FIG. 17). The table in FIG. 17 depicts the DAR calculation for Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE based on the percentage area of unconjugated and drug loaded light and heavy chains.

5. Stability Studies of Trastuzumab-PL13-Val-Cit-4-aminobenzoyl-MMAE and Trastuzumab-maleimide-Val-Cit-4-aminobenzoyl-MMAE

[0233] Stability of drug-linker for the trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE and trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE conjugates was evaluated in the presence of NAC in PBS buffer.

[0234] Method:

[0235] Each ADC (at concentration of 1-2 mg/mL), was incubated with 1 mM NAC in PBS buffer for 24 hours at 37° C.

[0236] Results:

[0237] The trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE conjugate was not influenced by the presence of NAC in buffer (see FIGS. 18 and 19). Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE showed decreased stability in the presence of NAC (see FIG. 20). In FIG. 18, the numbers designate the amount of drug conjugated to a full length antibody. DL indicates unconjugated drug.

[0238] FIG. 19 in particular demonstrates there is no profile change for the PL13 containing ADC resulting from challenge from free thiol, indicating stability of the construct.

6. SDS-PAGE and SEC Analysis of Trastuzumab-PL13-Val-Cit-4-aminobenzoyl-MMAE and Trastuzumab-maleimide-Val-Cit-4-aminobenzoyl-MMAE Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE and Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE were evaluated by SEC chromatography and non-reducing SDS-PAGE.

[0239] Results:

[0240] Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE is present as 99.5% monomer (see FIG. 21 (A)). Trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE is present as 93.5% monomer (see FIG. 21 (B)).

[0241] SDS-PAGE analysis under non-reducing conditions showed the presence of multiple bands in both the trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE and Trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE samples due to the disruption of inter-chain disulphide bonds in the antibody during alkylation of cysteine residues with linker-drug moieties (see FIG. 22). In FIG. 22, lane 1 represents the analysis of a 5 μL of sample of trastuzumab-PL13-val-cit-4-aminobenzoyl-MMAE loaded on the gel at a concentration of 1.8 mg/mL and lane 2 represents the analysis of a 5 μL of sample of trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE loaded on the gel at a concentration of 5 mg/mL. The integrity of full length antibody is still sustained due to strong non-covalent interactions between heavy and light chains which was confirmed by SEC analysis.

7. Synthesis of PL13-NH-PEG4-OSu Heterobifunctional Linker

[0242] The hydrophilic PEG (PEG4) was introduced to the linker molecule linker arm portion, to provide PL13-NH-PEG4-OSu, to improve linker solubility over the example given under section 1 above (see FIG. 23).

[0243] Method:

[0244] PL13 free acid (as described above) was coupled to 1-Amino-3, 6, 9, 12-tetraoxapentadecan-15-oic acid. Activation of the carboxyl group to the desired succinimidyl ester was undertaken using N,N-dicyclohexylcarbodiimide (DCC) 2.5 eq./HOSu coupling in anhydrous dichloromethane (DCM). Unreacted DCC and a diisopropylurea by-product were removed via filtration from cold DCM.

[0245] Results:

[0246] ESI-MS of PL13-NH-PEG4-COOH (see FIG. 24), confirmed the linker identity.

Comparison of Cross-reactivity of PL13-NH-PEG4-OSu and SMCC Linkers

[0247] The extent of Trastuzumab cross-linking via succinimidyl trans-4-(maleimidylmethyl)cyclohexane-1-carboxylate (SMCC), a known linker, and PL13-NH-PEG4-OSu was evaluated by reducing SDS-PAGE.

[0248] Method:

[0249] Trastuzumab at a final concentration of 2 mg/mL was incubated with 10 fold excess of SMCC or PL13-NH-PEG4-OSu. Samples were incubated at RT.

[0250] Results:

[0251] It is evident from the SDS-PAGE that PL13-NH-PEG4-OSu shows less high molecular weight bands than the SMCC linker (see FIG. 25, lanes 5-9). There are some higher molecular weight bands on the PL13-NH-PEG4-OSu SDS-PAGE but these are present to some extent in the starting Trastuzumab lane as well (see FIG. 25, lane 5).

[0252] In contrast to the linkers of the present invention, the SMCC linker induces trastuzumab cross-linking due to non specific reactivity of the maleimide group particularly towards amine side chains of lysine residues.

8. Synthesis of PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE

[0253] To improve further the solubility of the previously synthesised PL13-val-cit-4-aminobenzoyl-MMAE cytotoxic drug linker, a short hydrophilic PEG unit was introduced to the molecule (see FIG. 26). The PEG unit also introduces a ‘spacer’ to the ADC linker arm, increasing the separation of the MMAE cytotoxic payload and the antibody attachment point through PL13 by some 39 atoms (versus 22 atoms previously in PL13-vcMMAE). It is believed this “spacer” should increase the rotational flexibility of the PL13 unit, which may influence the kinetics of the Michael addition to free thiol.

[0254] Method:

[0255] 40 mg of crude Fmoc-val-cit-4-aminobenzoyl-MMAE was purified. The N-Fmoc group was then removed by aminolysis and the amine functionalised cytotoxic compound was purified to afford 24 mg of material. This was used to couple PL13-NH-PEG4-COOH via standard HOBt active ester chemistry.

[0256] Results:

[0257] PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE was synthesised. This molecule represents an example of a cleavable ADC system in accordance with the present invention.

9. Conjugation of PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE and Maleimide-val-cit-4-aminobenzoyl-MMAE to Trastuzumab.

[0258] Method:

[0259] Trastuzumab was reduced with 2.4 times excess of tris(2-carboxythyl)phosphine (TCEP) for 1 hour at RT in the presence of ethylenediaminetetraacetic acid (EDTA). The antibody was re-buffered into PBS. Trastuzumab, at a concentration of 20 mg/mL, was conjugated to 20 fold excess of PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE in the presence of 5% v/v DMA at RT for 16 hours. Additionally, maleimide-val-cit-4-aminobenzoyl-MMAE linker was coupled to Trastuzumab (20 mg/mL) at 6 molar excess at RT for 1 hour. The ADCs obtained were analysed by HIC, PLRP and SEC chromatography.

[0260] Results:

[0261] HIC characterisation of trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE resulted in a DAR calculation of 4.29 (see FIG. 27 A).

[0262] HIC characterisation of trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE resulted in a DAR calculation of 3.15 (see FIG. 27 B).

[0263] In FIGS. 27A and 27B, the numbers designate the amount of drug conjugated to a full length antibody. The table represents the relative percentage of each peak area from the HIC elution profile. The average DAR is calculated by multiplying the percentage peak area by the corresponding drug load to obtain a weighted peak area. The weighted peak areas are summed and divided by 100 to give a final DAR value.

[0264] DAR analysis of DTT reduced trastuzumab-maleimide-NH-PEG4-val-cit-4-aminobenzoyl-MMAE by PLRP confirmed the average drug load to be 4.3 (see FIG. 28 A and FIG. 29).

[0265] Analysis of DTT reduced trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE by PLRP confirmed an average drug load to be 3.8 (see FIG. 28 B and FIG. 29).

[0266] In FIGS. 28A and 28B, the numbers designate the amount of drug conjugated to light (L) or heavy (H) chain of the antibody.

[0267] Analysis of both trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE (HER-MAL-MMEA) and trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE (HER-PL-13-MMEA) on SEC chromatography showed that both samples are represented as monomeric species (see FIG. 30 A and B). The table below

[0268] FIGS. 30A and 30B shows the calculation of high molecular weight species (HMW), monomeric and low molecular weight species (LMVV).

10. Optimised Conditions for the Conjugation of PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE

[0269] Method:

[0270] Trastuzumab (23.5 mg/mL) was reduced with 2.4 equivalents of TCEP for 2 hours at RT to yield an average of 4.5 free thiols. The reduced Trastuzumab was conjugated to 20 fold excess of PL13-NH-PEG4-val-cit-MMAE in the presence of 10% v/v DMA at 30° C. for 18 hours. The conjugate was analysed by HIC and PLRP chromatography (see FIG. 31 A and B).

[0271] Results:

[0272] The HIC and PLRP data show a substantial improvement in conjugation efficiency. The higher level of conjugation efficiency has also resulted in a significant reduction in the level of ‘odd’ DAR species and a reduction in the level of unconjugated antibody.

11. Conjugation of Trastuzumab-PL13-NH-PEG4-val-cit-aminobenzoyl-MMAE (A) and Trastuzumab-maleimide-val-cit-aminobenzoyl-MMAE (B) at 150 mg Scale, Demonstrating Scalability of the Process

[0273] Method:

[0274] Trastuzumab (24.53 mg/mL) was reduced with 2.1 equivalents of TCEP for 2 hours at RT. The reduced Trastuzumab was conjugated to 20 fold excess of PL13-NH-PEG4-val-cit-MMAE in the presence of 10% v/v DMA at 30° C. for 18 hours.

[0275] Trastuzumab at concentration of 25.68 mg/ml was reduced with 1.95 equivalents of TCEP for 2 hours at RT. The reduced Trastuzumab was conjugated to 6 fold excess of maleimide-val-cit-MMAE in the presence of 10% v/v DMA at RT for 1 hour.

[0276] The conjugates were analysed by HIC and SEC chromatography (see FIGS. 32 and 33).

[0277] Results:

[0278] The HIC profiles showed similar conjugation efficiency for Trastuzumab-PL13-NH-PEG4-val-cit-aminobenzoyl-MMAE (FIG. 32A) and Trastuzumab-maleimide-val-cit-aminobenzoyl-MMAE (FIG. 32B). SEC (FIGS. 33A and 33B) confirmed that both conjugates are monomers with only a small amount of high molecular weight species present (1.4-2.4%). This demonstrates the scalability of the conjugation process developed using the linkers of the present invention.

12. Conjugation of PL13-NH-PEG4-val-cit-aminobenzoyl-MMAE to Thiomab, Demonstrating Successful Conjugation to Engineered Trastuzumab (Thiomab)

[0279] Method:

[0280] Trastuzumab with a cysteine mutation (V205C) introduced in the antibody light chain was conjugated at concentration of 10 mg/ml to 40 fold excess of PL13-NH-PEG4-val-cit-MMAE in the presence of 10% DMA in buffer pH 7.4. Conjugation reactions were incubated at 35° C. for 48 hours.

[0281] Results:

[0282] The HIC profile of Trastuzumab (V2015C)-PL13-NH-PEG4-val-cit-MMAE demonstrates that PL13 can be successfully conjugated to antibodies with engineered Cys residues at both pHs (FIG. 34). The rate and extent of the conjugation reaction appears to be influenced by the thiol microenvironment.

13. Utility of ADCs According to the Present Invention as ADCs.

[0283] ADC conjugation products, as described above in section 8, were further subjected to in vivo and in vitro testing to demonstrate their utility as commercial ADCs.

In Vitro Stability of Trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE and Trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE in the Presence of NAC

[0284] Method:

[0285] Stability of drug-linker for the trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE and trastuzumab-PL13-N H-PEG4-val-cit-4-aminobenzoyl-MMAE conjugates was evaluated in the presence of NAC in PBS buffer.

[0286] Results:

[0287] The trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE conjugate was not influenced by the presence of NAC in the buffer (FIG. 36). Trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE showed decreased stability in the presence of NAC (see FIG. 35). FIG. 36 in particular demonstrates there is no profile change for the PL13 containing ADC resulting from challenge from free thiol, indicating stability of the construct.

In Vitro Stability of Trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE and Trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE in Mouse Plasma

[0288] The in vitro serum stability of the ADCs Trastuzumab-PL13-NH-PEG4-val-cit-aminobenzoyl-MMAE and Trastuzumab-maleimide-val-cit-aminobenzoyl-MMAE was compared in mouse plasma.

[0289] Method:

[0290] Trastuzumab-PL13-NH-PEG4-val-cit-aminobenzoyl-MMAE and Trastuzumab-maleimide-val-cit-aminobenzoyl-MMAE were separately spiked into filtered athymic nude mouse plasma to a concentration of 0.2 mg/mL. The solutions were mixed and triplicate aliquots of 50 μL were taken and snap frozen in liquid nitrogen (time point—0 h). The plasma solutions of the ADCs were incubated at 37° C. for 7 days. Aliquots of 50 μL were pulled in triplicate for each time point: 1, 2, 3, 4, 5, 6 and 7 days, and stored at −80° C. until analysis.

[0291] Results:

[0292] ADC stability was monitored by LC-ESI/MS analysis. Plasma samples were pre-purified and trypsin/CNBr digested before MS analysis. Four non-conjugated peptides (two from heavy chain and two from light chain) and two conjugated peptides (one from heavy chain and one from light chain) were selected to monitor the stability of the ADCs. Averaged data from non-conjugated peptides correspond to the stability of the antibody component in mouse plasma (Total-Ab) and data from conjugated peptides show the stability of the ADC conjugate (LC conjugated, HC conjugated peptide).

[0293] This in vitro data revealed that Trastuzumab-maleimide-val-cit-aminobenzoyl-MMAE undergoes de-drugging at both the light chain and heavy chain, whereas the antibody component is stable (FIG. 37A). Loss of drug from the heavy chain peptide was more rapid when compared to the light chain peptide, indicating that the rate of drug loss is affected by the conjugation site.

[0294] The Trastuzumab-PL13-NH-PEG4-val-cit-aminobenzoyl-MMAE conjugate retained MMAE drug throughout the 7 day incubation period, showing that PL13 confers stability to both conjugation sites (FIG. 37B). The observed variability for both conjugates is due to efficiency of sample recovery for the ESI-MS method applied for sample analysis.

In Vivo Stability of Trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE and Trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE in Mice

[0295] Method:

[0296] In vivo stability was evaluated in mice injected with 5 mg/kg of the trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE and trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE conjugates. Plasma samples were pre-purified and trypsin/CNBr digested before LC-MS analysis. Two conjugated peptides (one from heavy chain and one from light chain) were selected to monitor the stability of the ADCs.

[0297] Results:

[0298] The stability of conjugated peptides is a result of two events; 1) catabolic degradation of ADCs and 2) loss of drug due to linker instability. Conjugated peptides derived from trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE show better stability at 1 day in mouse plasma than peptides derived from trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE. The observed difference is likely due to more rapid drug loss from conjugated peptides derived from trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE than from trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE which is consistent with the in vitro mouse plasma data, showing rapid de-drugging for trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE within the first four days of incubation (FIG. 38); approximately 70% of the drug attached to the light chain and 90% of the drug on the heavy chain was lost within the first two days. In contrast, trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE retains more drug due to the linker stability over a longer period of time.

In Vitro Stability of Trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE and Trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE in Various Cell Lines

[0299] In FIGS. 39A to 39C SKBR3, BT747 and JIMT-1 cell kill data are shown for the ADCs described above. The SKBR3, BT747 and JIMT-1 cell lines were selected in this test, as they are known to express the Her2 receptor. Additionally, the JIMT-1 cell line is known to be resistant to trastuzumab. In FIGS. 39A to 39C “ADC mal” relates to trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE, and “ADC-PL13” relates to trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE. Relative in vitro activities of tested conjugates against three cell lines indicate the data for the trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE are comparable to the trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE. Accordingly, this data shows that the ADC of the present invention is active and so the presence of PL13 does not have a negative effect on cell killing.

In Vivo Stability of Trastuzumab-malemide-val-cit-4-aminobenzoyl-MMAE and Trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE in Mice Exhibiting Breast Cancer Cell Tumours

[0300] Xenograft data is shown in FIGS. 40, 41 and 42 in relation to the effect of the ADCs on mice exhibiting breast cancer cell tumours. The breast cancer cell line utilised was BT474 (Her2 expresser, with high susceptibility to trastuzumab).

[0301] Mouse weight is used (as shown in FIG. 40) to measure the toxicity of the ADC in vivo. It can be seen from the data obtained that the trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE (ADC-mal) (FIG. 40A) and the trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE (ADC-PL13) (FIG. 40B) have a negligible effect on mouse weight over the course of the therapy period. This is indicative of the ADCs of the present invention being of suitable toxicity levels for use commercially.

[0302] FIG. 41 shows the effectiveness of the tested ADCs at three doses (0.1, 5 and 10 mg/kg) on tumour growth over a 21 day therapy period. Tumour growth was determined by changes in tumour volumes. Mice treated with both compounds at 0.1 mg/kg experienced tumour growth delay relative to the untreated group. In contrast, all of the animals dosed with the trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE (ADC-mal) and the trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE (ADC-PL13) at 5 and 10 mg/kg showed complete responses over the therapy period.

[0303] FIG. 42 shows the effectiveness of the tested ADCs at three doses (0.1, 5 and 10 mg/kg) on tumour maturity histological phenotype over a 10 day therapy period. Tumour tissues were examined microscopically after haematoxylin and eosin staining and graded as follows:

0: no xenograft mass; occasionally, isolated tumour cells in sub-cutis or fat tissue
1: small fragmented mass; incomplete epithelial development sequence
2: xenograft showing marked cell loss
3: zonal cell loss with areas of mature tumour
4: intact xenograft mass showing full epithelial development sequence and associated pathology

[0304] No reduction in the tumour maturity was observed for either trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE (ADC-PL13) or trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE (ADC-mal) dosed at 0.1 mg/kg, with no significant differences observed between the two ADCs. However, trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE dosed at 5 and 10 mg/kg revealed significantly reduced tumour maturity starting at days 3 (grade 3 and grade 3/2, respectively). This effect was more pronounced at day 10 (grade 1 for 5 mg/kg and grade 1/0 for 10 mg/kg) with some samples showing no tumour mass. The trastuzumab-maleimide-val-cit-4-aminobenzoyl-MMAE dosed at 5 and 10 mg/kg reduced tumour maturity (grade 4/3 and grade 4/3/2, respectively) starting at 3 days, but the effects were less pronounced in comparison with animals dosed at corresponding doses of trastuzumab-PL13-NH-PEG4-val-cit-4-aminobenzoyl-MMAE at 3 days. Furthermore, there was a markedly greater tumour heterogeneity for ADC-mal than observed for ADC-PL13 for 5 and 10 mg/kg doses at 10 days.

14. Demonstration of PL11-PEG.sub.20 KDa, PL12-PEG.sub.20 KDa and PL13-PEG.sub.20 KDa Linker Reactivity with Albumin

[0305] As indicated above, the majority of the cysteine thiols present in commercial grade albumin are capped, and so a reduction step is necessary prior to conjugation with linker molecules.

[0306] Method:

[0307] Albumin (221 μM) was reduced with dithiothreitol (DTT) (1 mM) in phosphate-buffered saline (PBS) pH 7.0 containing ethylenediaminetetraacetic acid (EDTA) (1 mM) for 1 hour at room temperature (RT), followed by desalting on a Nap-5 column and quantification by UV measurement. Conjugation of 20 KDa PEG-PL-12 (10 molar excess) or 20 KDa PEG-PL-13 (10 molar excess) to reduced albumin (20 μM) was performed in PBS, pH 7.4, 1 mM EDTA for 1 day at RT. Similarly, conjugation of 20 KDa PEG-maleimide (1.1 molar excess) to reduced albumin (20 μM) was performed in PBS, pH 7.4 containing 1 mM EDTA for 1 day at RT. The reactions were analysed on SDS-PAGE gel and quantification performed by analysis of protein band density on SDS-PAGE gel using ImageLab software (Biorad).

[0308] Results:

[0309] Only moderate conjugation yields were observed with reduced commercial grade albumin after 1 day. The conjugation of PL13 proceeded in a comparable yield to the maleimide control, whereas PL12 gave a slightly lower yield over the same time period. This is shown in FIG. 43.

15. Optimisation of PL11, PL12 and PL13 Conjugation to Albumin

[0310] The conjugation of the linker to albumin was determined at pH 5.5 for PL-11 and at pH 5.5, 6.5, 7.5, 8.0 and 8.5 for PL-12 and PL -13.

[0311] Method:

[0312] Conjugation of PL-11 (10 molar excess) to albumin (50 μM) was performed in 50 mM acetate-Na pH 5.5 for 24 hours at 37° C. Conjugation of PL-12 and PL-13 (10 molar excess) to albumin (50 μM) was performed in the following buffers: 50 mM acetate-Na pH 5.5; PBS 1mM EDTA pH 6.5; PBS 1mM EDTA pH 7.5; 100 mM phosphate; 1mM EDTA pH 8.0; and 100 mM carbonate-Na, 1mM EDTA pH 8.5; for 24 hours at 37° C.

[0313] Upon completion, all albumin samples were desalted on a Nap-5 column into PBS buffer (pH 7.4). The conjugation efficiency was quantified by Ellman's assay. Ellman's assay detects and quantifies free cysteine residues by reacting of 5,5′-Dithio-bis-(2-nitrobenzoic acid) with free thiol groups. The conjugation of linker molecules to the free thiol group of 34Cys in albumin blocks the detection of these groups by Ellman's reagent in comparison to the unconjugated albumin control. A decrease in the concentration of detectable free thiols was used to quantify amount of linker molecule conjugated to the albumin.

[0314] Results:

[0315] Conjugation of PL-11 to albumin at pH 5.5 proceeded in an 82% yield. The optimum pH for conjugation of PL-12 (91%) and PL-13 (94%) was determined to be pH 7.5. These results are shown in FIG. 44.

16. Demonstration of linker-albumin conjugate stability by ESI-MS

[0316] In addition, Albumin-PL11, Albumin-PL12 and Albumin-PL13 conjugates were analysed by ESI-MS.

[0317] Method:

[0318] Conjugation of PL-11 (20 molar excess) to albumin (50 μM) was performed in PBS containing EDTA (1 mM) at pH 6.5 for 24 hours at 37° C. Conjugation of PL-12 and PL-13 (10 molar excess) to albumin (50 μM) was performed in PBS containing EDTA (1 mM) at pH 7.5 for 24 hours at 37° C. The albumin-linker conjugates were analysed by ESI-MS.

[0319] Results:

[0320] FIG. 45 shows the ESI-MS results obtained for the albumin-PL-11 conjugation (A), albumin-PL-12 conjugation (B) and the albumin-PL-13 conjugation (C).

[0321] The following albumin species were detected and quantified based on MS signal intensity (note that MW stands for molecular weight):

[0322] Albumin-PL11 [0323] Unconjugated albumin—6%

[0324] Expected MW of unconjugated Albumin—66440 (Detected MW=66445 Da) [0325] Albumin-PL11 conjugate—90%

[0326] Expected MW of albumin-PL11—66649 Da (Detected MW=66650 Da) [0327] Albumin conjugated to two PL11˜4%

[0328] Expected MW of albumin-2-PL11—66858 Da (Detected MW=66858 Da)

[0329] Albumin-PL12 [0330] Unconjugated Albumin—1%

[0331] Expected MW of unconjugated Albumin—66440 (Detected MW=66441 Da) [0332] Albumin-PL12—92%

[0333] Expected MW of albumin-PL12—66663 Da (Detected MW=66664 Da) [0334] Albumin conjugated to two PL12—7%

[0335] Expected MW of albumin-2-PL12—66886 Da (Detected MW=66886 Da)

[0336] Albumin-PL13 [0337] Unconjugated Albumin—5%

[0338] Expected MW of unconjugated Albumin—66440 (Detected MW=66441Da) [0339] Albumin-PL13—94%

[0340] Expected MW of albumin-PL13—66631 Da (Detected MW=66632Da) [0341] Albumin conjugated to two PL13—1%

[0342] Expected MW of albumin-2-PL13—66822 Da (Detected MW=D66822a)

[0343] In particular, with reference to FIG. 45, peaks (4501), (4504) and (4507) correspond to unreacted and oxidised albumin, peaks (4502), (4505) and (4508) correspond to single conjugate albumin and peaks (4503), (4506) and (4509) correspond to double conjugate albumin.

17. Demonstration of Stability of Albumin-linker Conjugates in Glutathione

[0344] The stability of albumin-linker molecule conjugates was determined in the presence of excess glutathione and analysed by ESI-MS over a period of 7 days.

[0345] Method:

[0346] Samples of the albumin-PL-12 conjugate (0.5 mg/mL) were incubated with reduced glutathione (1 mM) in PBS (pH 7.4) at 37° C. for 7 days. The samples were analysed on ESI-MS.

[0347] Results:

[0348] FIG. 46 shows the results obtained for the albumin-PL-12 conjugate as a representative example. A similar trend was observed for the albumin-PL-11 and albumin-PL-13 conjugates.

[0349] FIG. 46 shows the ESI-MS results at day 0, 1, 4 and 7. In FIG. 46, peaks (4601), (4603), (4605) and (4608) correspond to albumin, peaks (4602), (4604), (4606) and (4609) correspond to albumin-PL12 conjugate and peaks (4607) and (4610) correspond to GSH-albumin. Accordingly, it is shown that: [0350] The major species present is the albumin-PL12 conjugate (MW=˜66664 Da) [0351] Upon incubation with 1 mM glutathione, there is a very slow increase in Albumin-GSH signal (˜8%, MW=66752 Da) starting at 4 day of incubation with glutathione and reaching ˜13% on 7.sup.th day of incubation.

[0352] It can be concluded that albumin-linker conjugates of the present invention demonstrate good stability in a competitive environment.

18.Generation of Albumin-doxorubicin Conjugates Using Linkers According to the Present Invention

[0353] Method:

[0354] Albumin, Thioalbumin (single cysteine mutant) and Thioalbumin (double cysteine mutant) were conjugated to 14, 20 and 30 fold excess of PL13-NH-PEG4-val-cit-4-aminobenzoyl-doxorubicin linker (FIG. 47), respectively. Reactions were carried out in 147 mM sodium phosphate pH 7.5, in the presence of 10% v/v DMA and 0.6 mM EDTA, at 37° C. for 2 days in the dark. After this time, the reactions were desalted on PD-10 columns equilibrated in PBS buffer pH 7.4.

[0355] The conjugation efficacy was determined by UV-Vis spectroscopy using the doxorubicin absorbance and extinction coefficient at 495 nm (ε=8030 M.sup.−1, cm.sup.−1) and that of albumin at 280 (ε=34445 M.sup.−1, cm.sup.−1), with correction for doxorubicin absorbance at 280 nm according to the equation:

[00001]$\mathrm{Albumin}.Math..Math.\mathrm{concentration}=\frac{\left(\mathrm{Abs}.Math..Math.280.Math..Math.\mathrm{nm}-\left(0.724\times \mathrm{Abs}.Math..Math.495.Math..Math.\mathrm{nm}\right)\right)}{\mathrm{.Math.280}.Math..Math.\mathrm{nm}}$

[0356] The extent of albumin-doxorubicin conjugate aggregation was determined by analysis on non-reducing SDS-PAGE and SEC chromatography.

[0357] 10 μl of each albumin-doxorubicin conjugate in SDS-PAGE loading buffer was loaded on NuPAGE 4-12% Bis-Tris gel and run for 45 minutes at 200 V. The native fluorescence properties of doxorubicin were used to visualise albumin-doxorubicin conjugates before staining of a gel with Coomassie dye for protein species.

[0358] 5 μl of each conjugate was loaded on to a SEC HPLC column equilibrated with 150 mM sodium phosphate buffer at pH 7.0 and eluted at flow rate of 1 mL/min with detection at 280 nm.

[0359] Results:

[0360] Conjugation of doxorubicin to wild-type albumin and albumin mutants resulted in yields of 55% for albumin, 32% for thioalbumin-single mutant and 14% for thioalbumin-double mutant (see FIG. 48).

[0361] SDS-PAGE analysis revealed the presence of a small amount of high molecular weight species (HMWS) in all albumin-doxorubicin samples (see FIG. 49). This data is consistent with that observed by SEC analysis (see FIG. 50). In addition, SDS-PAGE analysis confirmed the conjugation of doxorubicin to albumin observed by UV-Vis spectroscopy (see FIG. 49, DOX fluorescence)

[0362] Albumin-doxorubicin and thioalbumin-doxorubicin conjugates were analysed along with non-conjugated albumin and non-conjugated thioalbumin controls by SEC chromatography. Analysis of the doxorubicin conjugates showed that the monomer content was approximately 97.6% for albumin-DOX, 86.8% for albumin-DOX (single mutant) and 88.2% for albumin-DOX (double mutant), (see FIG. 50 B, D and F). Non-conjugated albumin and thioalbumin controls (see FIGS. 50 A, C and E) showed similar monomer and high molecular weight species content to the doxorubicin conjugated species. In particular, FIG. 50 shows peaks at (5001), (5003), (5005), (5007), (5009) and (5011) corresponding to the presence of the HMWS and peaks at (5002), (5004), (5006), (5008), (5010) and (5012) corresponding to the monomer content.

[0363] Both thioalbumin mutants show a higher tendency to aggregate than wild-type albumin: 13.2-14.5% for thioalbumin (single mutant) and 11.8-16.6% for thioalbumin (double mutant), compared to 2.4-2.6% for native albumin. Analysis of these samples confirmed that conjugation of doxorubicin via PL13-PEG4-val-cit-4-aminobenzoyl is well tolerated and produced no additional aggregation aggregation (see FIG. 50 A, C and F).

[0364] Accordingly, it has been shown that the protein drug conjugates of the present invention provide suitable alternatives to known protein drug conjugates. Moreover, the linker incorporated into the protein drug conjugates of the present invention enables safe and effective drug delivery by successfully binding the drug to the protein and retaining the drug until the target tissue is reached.