Multi-specific proteinaceous prodrug constructs

Abstract

The present invention relates to proteinaceous prodrug constructs, e.g., proteinaceous fusion constructs that comprise a complement 3- and pregnancy zone protein-like, alpha-2 -macroglobulin domain-containing (CPAMD) protein (e.g., A2M) and one or more drugs and function as protease- activatable prodrugs. The invention further relates to a release mechanism from the CPAMD protein and the development of multispecific, such as bi-specific, drugs.

Description

MULTI-SPECIFIC PROTEINACEOUS PRODRUG CONSTRUCTSCross-reference to related application

[0001] This application claims priority to, and the benefit of, European patent application EP23180912.0 filed on 22 June 2023, the contents of which is hereby incorporated by reference in its entirety.Sequence listing

[0002] This application contains a Sequence Listing which has been submitted electronically in XML format and is hereby incorporated by reference in its entirety. The XML file, created on 20 June 2024, is named P52261WO-Sequence_listing.xml and is 487 bytes in size.Technical field of the invention

[0003] The present invention relates to proteinaceous prodrug constructs, e.g., proteinaceous fusion constructs that comprise a complement 3- and pregnancy zone protein-like, alpha-2 -macroglobulin domain-containing (CPAMD) protein (e.g., alpha-2 -macroglobulin (A2M)) and one or more drugs, and function as protease-activatable prodrugs. The invention further relates to a release mechanism from the CPAMD protein and the development of multi-specific, such as bi-specific, drugs.Background of the invention

[0004] When a biopharmaceutical drug is administered to patients in a form that is initially active, it can exert its biological effect both in diseased and healthy tissues. Drug effects in healthy tissue may be detrimental to patient health, quality -of-life, and / or treatment efficacy. One strategy to minimize these side effects is the derivation of drugs into a prodrug form that is initially inactive and becomes active in the diseased environment.

[0005] Proteases are enzymes that catalyze the hydrolysis of peptide bonds in other proteins. There are over 600 known human proteases and the majority are closely regulated under normal circumstances. In many diseases specific proteases become dysregulated and have an increased activity relative to healthy conditions. Many protease-activated prodrug technologies for biopharmaceutical drugs have been developed, where most are based on antibodies. Several technologies use masking moieties which block the antibody’s antigen-binding region (paratope), preventing it from binding to its cognate epitope on the target antigen and thus rendering it inactive. The masking moiety is attached to the antibody via a linker incorporating a protease-cleavable site. Cleavage of the linker by proteases separates the antibody from the masking moiety, liberating the paratope and restoring the antibody’s activity. Masking moieties may be designed to specifically bind to the antibody paratope (e.g., using phage display -derived peptides as in CytomX’s Probody technology) or may sterically encompass the paratope sufficiently to sequester it without any specific interactions (e.g. using long, bulky peptides as inAmunix’s XPAT technology). Another approach prevents functional VH / VL domain pairing in the prodrug by incorporating inactive VH and VL domains, attached to the antibody by a protease -sensitive linker (e.g. Maverick’s COBRA technology). The linker’s cleavage removes the inactive domains and enables the correct VH / VL pairing in the activated prodrug.

[0006] These diverse prodrug technologies all have their advantages and disadvantages. For example, CytomX’s Probody minimizes the use of non-human and potentially immunogenic sequences, but is not modular and requires the identification of affine masking moieties for every antibody incorporated into their platform. Amunix’s XPAT platform does not require specific mask / antibody interactions and furthermore has a difference in circulatory half-life before and after the prodrug’s activation, but this is accomplished by using long non-human peptides. Hence, new prodrug technologies combining the key advantages of multiple technologies are needed and would constitute meaningful advancement of the field.

[0007] In the co-pending patent application PCT / EP2023 / 052280, a novel technology for making prodrugs using CPAMDs and antibodies or other biopharmaceuticals is described. Such CPAMD-based prodrugs end up being large (>700 kDa), mono-specific, and tetra- valent.

[0008] In the treatment of cancer, the penetration of therapeutic molecules into the tumor can be limited by their size. Furthermore, some therapeutic approaches such as T cell redirection (e.g. using bi-specific T cell engagers a.k.a. BiTEs) require the use of bi-specific antibodies, which must additionally bind mono-valently to T cells.

[0009] There is thus a need for the development of prodrugs capable of being released in smaller sizes, and / or the possibility of the released drugs being bi-specific.Summary of the invention

[0010] In particular, the proteinaceous prodrug constructs (e.g., the proteinaceous fusion constructs) described herein enable specific drug delivery and controllable activity. This can be achieved by the proteinaceous prodrug construct undergoing a conformational change through which it is able to control the activity of one or more drugs (e.g., a first drug such as an antibody) comprised in it. In the native (uncleaved) state of the proteinaceous prodrug construct, the one or more drugs (e.g., the first drug / antibody) is / are not exposed and thus, inactive. In the active (cleaved) state of the proteinaceous prodrug construct, the drug (e.g., the first drug / antibody) is exposed and able to interact with its target. For example, a first drug may be positioned inside the Receptor Binding Domain (RBD) of a CPAMD protein (e.g., A2M) comprised in a proteinaceous prodrug construct of the invention. The CPAMD protein (e.g., A2M) includes a bait region comprising a first protease cleavage site. The conformational change from the native state to the active state is triggered by the cleavage of the first protease cleavage site in the bait region.

[0011] Typically, the proteinaceous prodrug constructs described herein comprise a second drug (e.g., a second antibody). The second drug can be fused to the C-terminal end of the RBD. At the C-terminal end, the second drug (e.g., the second antibody or other antigen-targeting moiety) is accessible in thenative as well as the active state and can be used to direct the proteinaceous prodrug construct to a specific tissue, a specific cell type, and / or a specific receptor.

[0012] A specific drug delivery and controllable activity is provided by a release mechanism of the drug(s) (e.g., the first and second antibodies) from the larger prodrug construct. For example, the RBD may comprise a second protease cleavage site at the N-terminal end of the RBD. The release mechanism is enabled by the cleavage of a second protease cleavage site. Typically, the RBD remains bound to the CPAMD protein (e.g., A2M), via non-covalent interactions, upon cleavage of the second protease cleavage site. Upon altering conformation following proteolytic cleavage of the first protease cleavage site, the RBD comprising the drugs (e.g., the first and second antibodies) is released from the prodrug construct.

[0013] A cleavage site(s) comprised within a proteinaceous prodrug construct can be modified to control where the drug becomes exposed. Depending on the cleavage site, the drug can be exposed only at the location where proteases recognizing that cleavage site(s) are present. When a specific protease is present and cleaves the first cleavage site, the conformation of the proteinaceous prodrug construct is changed from “native” to “active”. Accordingly, the present invention provides proteinaceous prodrug constructs (e.g., proteinaceous fusion constructs), where the activity and specificity can be controlled and directed towards a specific area (e.g., a particular tissue) of a subject in need of treatment with such constructs.

[0014] In the present disclosure, the inventors describe exemplary CP AMD-based prodrugs that incorporate a second protease cleavage site in the sequence of a CPAMD protein (A2M), allowing the moiety carrying a first drug to be released from the remainder of CPAMD protein after activation, resulting in both a smaller activated first drug and monovalent specificity (Example 12 and 13). The inventors further describe the simultaneous fusion of a CPAMD protein (A2M) with a first drug (a first antibody) and a second drug (a second antibody) to create a bi-specific prodrug, wherein the first drug is shielded by the CPAMD protein before release (Example 14). The inventors also demonstrate that the release of a smaller activated and multi-specific (i.e., bi-specific) RBD comprising the first and second drugs, i.e., the first and second antibodies (Example 14).

[0015] Specifically, in one aspect, the present invention relates to a proteinaceous prodrug construct comprising a complement 3- and pregnancy zone protein-like, alpha-2 -macroglobulin domaincontaining (CPAMD) protein, such as A2M, comprising a bait region comprising a first protease cleavage site and a Receptor Binding Domain (RBD) comprising a second protease cleavage site at the N-terminal end of the RBD, a first drug positioned inside the RBD, and a second drug fused to the C- terminal end of the RBD, wherein the CPAMD protein shields the first drug and the second drug is accessible, and the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD and thereby making the first drug accessible.

[0016] The inventors have found that the RBD remains bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the second protease cleavage site. Accordingly, in another aspect, theinvention relates to a proteinaceous prodrug construct comprising a complement 3- and pregnancy zone protein-like, alpha-2 -macroglobulin domain-containing (CPAMD) protein, such as A2M, comprising a bait region comprising a first protease cleavage site and a Receptor Binding Domain (RBD) comprising a first drug positioned inside the RBD and a second drug fused to the C-terminal end of the RBD, wherein the CPAMD protein shields the first drug and the second drug is accessible, the RBD is bound to the CPAMD protein, via non-covalent interactions, and the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, thereby releasing the RBD and making the first drug accessible.

[0017] Also provided is a nucleic acid encoding the proteinaceous prodrug as described herein and a vector, such as a plasmid, comprising the nucleic acid. These nucleic acids and vectors may accordingly be comprised in host cells, and thus another aspect relates to a host cell comprising the nucleic acid as described above or vector as described above. Several methods of production, methods of treatment and therapeutic uses of these aspects are also provided.Brief description of the figures

[0018] The following figures illustrate the invention with proteinaceous prodrug constructs that comprise alpha-2 -macroglobulin (A2M) as a CPAMD protein. A person of skill in the art of proteinaceous prodrug design will appreciate that other CPAMD proteins can take the place of A2M.

[0019] Figure 1A shows a schematic overview of a proteinaceous prodrug construct (1), e.g., a fusion protein, comprising a CPAMD protein (2), e.g., A2M, fused to one or more drugs (e.g., one or more nanobodies / biopharmaceutical moieties) (3) positioned inside or in the vicinity of the RBD domain of the CPAMD protein. The one or more drugs (3) are inaccessible when the bait region of the CPAMD protein has not been proteolytically cleaved (inactive or “native” conformation I). The one or more drugs (3) are accessible when the bait region is cleaved by a protease (4) (active conformation II). When the protease (4) cleaves the “bait region”, the protease (4) is trapped inside the proteinaceous prodrug construct (1). Figure IB shows a schematic overview of the different fusion strategies of the CPAMD protein (e.g., A2M) and the drug.

[0020] Figure 2 shows native PAGE (A) and SDS-PAGE (B) analysis of wildtype A2M and fusions constructs with A2M and antibody scFvs from Atezolizumab, Ipilimumab, and Nivolumab, as noted. Before analysis, samples were treated with methylamine (MA) or thermolysin, as indicated. (C) A schematic of the domain organization of A2M-antibody constructs, showing the size of products generated by thiol ester autolysis and bait region cleavage.

[0021] Figure 3 shows conformational dependence of antigen binding by A2M-antibodies, measured by biolayer interferometry. (A) The interaction between A2M-Atezolizumab (purified by one round of depletion using a PD-L1 resin, see Example 4) and immobilized PD-Ll-hFc. The control and methylamine-treated A2M-Atezolizumab show a ~ 149-fold difference in their effective concentration, calculated from the fitted kObs values for their association. (B) The interaction between A2M-EgAl (purified by two rounds of depletion using an LRP1 resin, see Example 4) and immobilized EGFR-hFc.The control and methylamine-treated or thermolysin -treated samples show a ~63-fold difference in their effective concentration. (C) The interaction between A2M-Ipilimumab (purified by three rounds of depletion using an LRP1 resin) and immobilized CTLA-4-hFc. (D) The interaction between A2M- Nivolumab (purified by three rounds of depletion using an LRP1 resin) and immobilized PD-l-hFc. (E) The interaction between A2M-KN035 (not enriched for native A2M) and immobilized PD-Ll-hFc. (F) The interaction between A2M-Urelumab (purified by three rounds of depletion using an LRP1 resin) and immobilized 4-lBB-hFc. (G) The interaction between A2M-Foralumab (not enriched for native A2M) and immobilized CD3ya-hFc. (H) The interaction between A2M-Muromonab (not enriched) and immobilized CD3y8-hFc. In panels G and H, the +thermolysin sensorgram has the signal from a biosensor associating with thermolysin only subtracted, due to the low-intensity responses. (I) The interaction between A2M-Adalimumab (not enriched) and immobilized TNFa.

[0022] Figure 4 shows enrichment of native A2M-antibodies using affinity depletion. (A) A2M- Atezolizumab was depleted using a resin coated with its cognate antigen, PD-L1. One round of depletion was performed. Biolayer interferometry was then used to compare antigen binding of the untreated sample before (left) and after (right) depletion. (B-D) A2M-Nivolumab, A2M-Ipilimumab, and A2M- Urelumab were depleted using LRP1 -coated resin. Three rounds of depletion were performed for each A2M-antibody, after which biolayer interferometry was used to compare their antigen binding before and after depletion. (E) A2M-Ipilimumab was depleted by three rounds with Protein L resin and biolayer interferometry was used to compare its antigen binding before and after.

[0023] Figure 5 shows immune checkpoint blockade by A2M-Atezolizumab in a cell bioassay of PD- 1 / PD-L1 blockade. PD-1+Jurkat T cells with a NFAT-driven luciferase gene to report NFKB signaling were co-cultured with PD-L1+CHO-K1 cells expressing a TCR agonist, in the presence of a dilution series of A2M-Atezolizumab in its native and methylamine-treated conformations, or Atezolizumab scFv fused to a human Fc region. The luminescence response, with the background from control cells subtracted and the subsequent response normalized to the maximum response, is shown. EC50 curves were fitted using linear regression and the maximum response and EC50 values from fitting are shown for each antibody.

[0024] Figure 6 shows the conformation and functionality of tabula rasa A2M which comprises a bait region that cannot be cleaved by proteases. (A) Sequences of the wildtype, tabula rasa (TR) and TR K704 bait regions. Basic residues (i.e. cleavage sites for trypsin or LysC) are highlighted. (B) Porelimited native PAGE of A2M incorporating the three given bait region sequences. All constructs originally demonstrated the slow electrophoretic mobility that is characteristic of A2M’s native conformation; upon methylamine aminolysis or bait region cleavage, A2M collapses and demonstrates a faster electrophoretic mobility. Wildtype A2M and A2M TR K704 were both collapsed by trypsin, while only A2M TR K704 was collapsed by LysC; A2M TR was not collapsed by either protease. (C) Reducing SDS-PAGE of the same A2M samples as in panel B. The thiol ester-dependent heatfragmentation bands (TE 120 and TE 60) disappeared upon methylamine treatment. Bait region cleavage of A2M gives its ~85 and ~95 N- and C-terminal fragment bands; the C-terminal fragment additionallyforms high-MW multimer products through thiol ester-mediated conjugation. When the bait region is not cleavable by trypsin or LysC, A2M can be cleaved outside of the bait region without any activation of its thiol ester. A2M TR K704 forms an intense -250 kDa band upon proteolytic activation due to thiol ester-mediated conjugation of the bait region lysine residue.

[0025] Figure 7 shows incorporation of MMP2 substrate sites into tabula rasa A2M. (A) Bait region sequences for wildtype A2M, TR A2M, and four TR bait regions each incorporating a different MMP2 substrate sequence (A21A, B74, C9, and SI). The MMP2 recognition sequence is highlighted in each sequence; cleavage occurs at the N-terminus of the bolded hydrophobic residue. (B) A2Ms with these 6 bait regions were digested by MMP2 and nine other human proteases and cleavage was assessed by SDS-PAGE. Proteases that cleave a bait region are indicated with a + in the case of full cleavage and (+) in the case of partial cleavage (relative to wildtype A2M). The TR bait region was not cleaved by any tested protease, whereas each MMP2 substrate was cleaved by every tested MMP. The TR SI bait region was not cleaved by proteases other than MMPs, indicating an increased selectivity of inhibition relative to the wildtype bait region. (C-D) Pore-limited native PAGE and reducing SDS-PAGE, respectively, of the six A2Ms with and without MMP2 cleavage. All constructs are similarly bait region- cleaved by MMP2, resulting in a conformational collapse and the appearance of high-MW multimer products in SDS-PAGE, with the exception of A2M TR.

[0026] Figure 8 shows optimization of the production and inhibitory capacity of A2M TR SI. (A) Several modifications of the MMP2 substrate bait region, tabula rasa SI, were tested for their ability to improve the formation of native A2M and its inhibitory capacity towards MMP2. TR SI QRT4 reintroduces the fourth quarter of the wildtype bait region. Two different SI positions (with cleavage at position 710 or 703) were tested in TRA7, which shortens the TR bait region by seven residues. (B) Pore-limited native PAGE of A2Ms with the indicated bait regions. A2M TR SI is expressed with a substantial amount of non-native A2M. This non-native A2M could be removed by depletion using LRP1 -conjugated resin. Alternatively, the native content was improved in TRA7 and TR QRT4. (C) The ability of the indicated A2Ms to inhibit MMP2’s digestion of DQ-gelatin was determined. Fitted curves calculated from the experimental data points by linear regression are shown as dotted lines. Error bars show the standard; n=3.

[0027] Figure 9 shows A2M-antibodies incorporating engineered bait regions. (A) Bait region sequences for the wildtype A2M bait region, the shortened MMP2 substrate bait region “TRA7 SI 1703” that was described in Example 6, and an additional engineered bait region “TRA7 SI 1703 P704.” (B) Pore limited native PAGE and (C) reducing SDS-PAGE of wildtype A2M, A2M-Atezolizumab with a wildtype bait region, and A2M-Atezolizumab with the TRA7 SI 1703 bait region. The A2Ms were analyzed untreated, methylamine-treated, or treated by a 0.5: 1 or 4: 1 molar ratio of MMP2:A2M, as indicated. (D) Biolayer interferometry was used to assess PD-L1 binding by A2M-Atezolizumab with the 3 bait regions shown in panel A (wildtype top left comer, TRA7 SI 1703 top right comer, TRA7 SI 1703 P704 bottom), before and after MMP2 cleavage. A biosensor associating with MMP2 only, withoutA2M-Atezolizumab, is included to account for this background binding. A2M-Atezolizumab with wildtype bait region was additionally cleaved with thermolysin for comparison.

[0028] Figure 10 shows: (A) Reducing SDS-PAGE analysis of purified A2M-PD1. A2M-PD1 is expressed and purified by the same protocol as wildtype A2M or A2M-antibodies, to a high purity. The formation of an internal thiol ester in A2M-PD1 causes heat-induced fragmentation at the thiol ester site under denaturing conditions, generating an N-terminal and C-terminal product band. (B) A2M-PD1 binding to immobilized PD-L1, as assessed by biolayer interferometry. PD-L1 binding by A2M-PD1 without any treatment to change its conformation or after methylamine- or thermolysin-treatment to collapse its conformation is shown. A reference biosensor where thermolysin was added without any A2M-PD 1 was included to account for non-specific binding of thermolysin to the biosensor surface, and has been subtracted from the A2M-PDl+thermolysin sensorgram. A2M-PD1 after LRP1 depletion was also included, without treatment and after methylamine treatment.

[0029] Figure 11 shows: (A) Reducing SDS-PAGE analysis of purified A2M-IL2. A2M-IL2 is expressed and purified by the same protocol as wildtype A2M or A2M-antibodies, to a high purity. The formation of an internal thiol ester in A2M-IL2 causes heat-induced fragmentation at the thiol ester site under denaturing conditions, generating an N-terminal and C-terminal product band. (B) A2M-IL2 binding to immobilized IL-2Ra, as assessed by biolayer interferometry. IL-2Ra binding by A2M-IL2 without any treatment to change its conformation or after methylamine- or thermolysin-treatment to collapse its conformation is shown. A2M-IL2 after three rounds of LRP1 depletion was assessed in the same manner.

[0030] Figure 12 shows: (A) The interaction between 5 nM A2M-fusion-EgAl, before and after methylamine treatment, with immobilized human EGFR, measured during one hour association and one hour dissociation using biolayer interferometry. (B) The interaction between 5 nM A2M-iRBD-EgAl, before and after methylamine or thermolysin treatment, with immobilized human EGFR, measured during one hour association and one hour dissociation using biolayer interferometry. (C-E) The interactions between 5 nM of A2M-miRBD-EgAl, A2M-miRBD-KN035, and A2M-miRBD- Atezolizumab, before and after methylamine treatment, with immobilized EGFR or PD-L1, measured during one hour association and one hour dissociation (or two hours of association and 10 minutes dissociation, in the case of A2M-miRBD-Atezolizumab) using biolayer interferometry. (F) The interaction between 10 nM of A2M-tRBD-EgAl, before and after methylamine treatment, with immobilized EGFR, measured during one hour association and dissociation.

[0031] Figure 13 shows: The RBD domain (residues 1335-1474 of SEQ ID NO: 1) of A2M, with emphasis on four proposed sites that can be used for the insertion of drugs to achieve conformationdependent binding. These sites are residues 1392-1404 or 1391-1405 (loop 2), as demonstrated by the ciRBD, iRBD, miRBD, and tRBD fusion approaches, as well as residues 1368-1379 (loop 1), 1420- 1426 (loop 3), and 1450-1457 (loop 4), all of which are flexible linkers between beta strands that are spatially close to 1392-1404 and facing the same direction on the RBD domain. In contrast, residue 1468 defines the position of the inserted drug in the A2M-fusion-EgAl construct, where conformation-dependent binding was not obtained, indicating that this opposite side of the RBD domain is unsuited to achieve conformation-dependent binding. The RBD domain structure (from PDB accession code 7V0N) is represented as a cartoon, while the Caatoms of the indicated residues are shown as spheres. The RBD domain is shown from two different angles, as indicated.

[0032] Figure 14 shows: Illustrations of further modified A2M proteins. Schematic representations of the native structure, and the collapsed structures induced by proteolytic cleavage of the bait region, are shown. Additionally, an overview of the proteins’ domain organization is given (note that all proteins are identical in the N-terminal region spanning from MG1 to CUB2. The cleavage site indicated by arrows on the domain organization is distinct from the bait region and can be cleaved without resulting in a conformational change of A2M. The use of a furin cleavage site and a TEV protease cleavage site are shown in the Examples.

[0033] Figure 15 shows: SDS-PAGE (A) and native PAGE (B) analysis of wildtype A2M purified from plasma or recombinant A2M furinRBD, which has a furin cleavage site between its CUB and MG8 domains. Both proteins were included for analysis as either untreated samples, or after methylamine- induced aminolysis of the thiol ester or cleavage of the bait region with thermolysin. A2M furinRBD was completely intracellularly processed by furin, produced with an intact thiol ester and native conformation, and was induced to undergo its conformational change to a collapsed conformation by either thiol ester amino lysis or bait region cleavage.

[0034] Figure 16 shows: (A) Size exclusion chromatography (SEC) of A2M furinRBD, either untreated or following methylamine amino lysis of the thiol ester to induce A2M’s conformational change. To the right, a chromogram zooming in on the elution of the MG8 domain is shown, as the MG8 domain has a very low extinction coefficient and gives a low signal when measuring absorbance at 280 nm. (B) SDS- PAGE of A2M furinRBD samples, before and after the SEC separation shown in panel A. The MG8 domain elutes in fraction 2 after methylamine treatment, but mostly elutes with the remainder of the A2M protein before methylamine treatment. This shows that the MG8 domain is released by A2M’s conformational change.

[0035] Figure 17 shows: (A) SDS-PAGE of A2M tevRBD after overnight cleavage at room temperature with a titration series of TEV protease added to indicated the weight / weight ratio. TEV protease cleavage of A2M tevRBD was near complete at a 2.5 : 1 w / w ratio and results in a truncation of the intact A2M band, the C-terminal autolytic band, and the C-terminal bait region cleavage product. (B) SDS-PAGE and (C) native PAGE analysis of wildtype A2M purified from plasma or recombinant A2M tevRBD before or after TEV protease cleavage and removal of TEV protease by SEC. All proteins were included for analysis as either untreated samples, or after methylamine-induced aminolysis of the thiol ester or cleavage of the bait region with thermolysin. A2M tevRBD kept its thiol ester and native conformation after TEV protease cleavage, and was induced to undergo its conformational change to a collapsed conformation by either thiol ester aminolysis or bait region cleavage with thermolysin.

[0036] Figure 18 shows: (A) Size exclusion chromatography (SEC) of A2M tevRBD after its complete cleavage by TEV protease and a preceding SEC purification step to remove the majority of the TEVprotease. TEV-cleaved A2M tevRBD was either untreated or treated with methylamine for aminolysis of the thiol ester, to induce A2M’s conformational change. To the right, a chromogram zooming in on the elution of the MG8 domain’s elution is shown. (B) SDS-PAGE of A2M tevRBD samples, before and after the SEC separation shown in panel A. The MG8 domain elutes in fraction 2 and 3 (depending on its glycosylations) after methylamine treatment but elutes with the remainder of the A2M protein before methylamine treatment. This shows that the MG8 domain is released by A2M’s conformational change.

[0037] Figure 19 shows: (A) SDS-PAGE and native PAGE (B) of wildtype A2M purified from plasma and recombinant A2M with a TEV protease-cleavable site (tevRBD), a C-terminal nanobody, and a ciRBD-positioned scFv (tevRBD+2xAb). Both A2M proteins are analyzed either untreated, after methylamine aminolysis of the thiol ester, or after bait region cleavage with thermolysin. A2M tevRBD+2xAb was produced with an intact thiol ester, as determined by the presence of autolytic fragments in SDS-PAGE. Thermolysin cleavage occurred predominantly in the tevRBD+2xAb bait region, but also occurred in the TEV protease site (ENLYFQS (SEQ ID NO: 226), which contains several hydrophobic residues that are recognized and cleaved by thermolysin. This results in a ladder of 5 distinctly migrating bands seen for thermolysin-cleaved tevRBD+2xAb in native PAGE, corresponding to the removal of 0-4 MG8 domains and the attached antibody fragments by thermolysin (as indicated on the image). - (C) SDS-PAGE and native PAGE (D) of wildtype A2M purified from plasma and recombinant A2M with a modified TEV protease-cleavable site (tevRBD), a C-terminal nanobody, and a ciRBD-positioned scFv (tevRBD+2xAb_2). The recombinant A2M with two antibodies and a TEV site was initially native, containing thiol ester and adopting a slow mobility in native PAGE. Upon TEV protease cleavage, the MG8 domain remained non-covalently associated with A2M, which remained native (with intact thiol ester and slow electrophoretic mobility). Induction of TEV -processed A2M with two antibodies using methylamine resulted in a conformational change which released the MG8 domain if TEV site cleavage had been performed.

[0038] Figure 20 shows: (A) Illustration of the domain structure of a further modified, bispecific A2M protein, i.e., an exemplary generalized domain structure of an A2M-BiTE. Anti-CD3 = antibody targeting CD3. Anti-TAA = antibody targeting a tumor associated antigen. A first cleavage site located in the bait region and a second cleavage located at the N-terminus of the MG8 domain are indicated by arrows. (B) Corresponding schematic representation of the “inactive” closed confirmation of the A2M BiTE, and the “active” open conformation induced by proteolytic cleavage of the first cleavage site, are shown. Anti-TAA is accessible in the closed and open conformations. Anti-CD3 is shielded until the first cleavage site is cleaved. Cleavage of the second cleavage site by a protease (shown in light grey) “primes” the A2M-BiTE for release of the anti-CD3-MG8-anti-TAA BiTE. Release of the BiTE is dependent on cleavage of the first cleavage site, as indicated, leaving behind an activated A2M with the protease which cleaved the first cleavage site (shown in dark grey) trapped inside the molecule.

[0039] Figure 21 shows: (A) a pore-limited gel and (B) SDS-PAGE showing TEV -digested A2M and A2M-BiTE construct BiTE2-TEV4 (SEQ ID NO: 258) samples. The pore-limited gel in (A) shows that,after digest with TEV, the anti-CD3-MG8-anti-EGFR domain of the A2M-BiTE can be released by the conformation change (induced with Methylamine or trypsin), giving an “empty” A2M that migrates the same as unmodified A2M. Release of the MG8 domain comprising the two antibodies required subjecting the A2M-BiTE to TEV-protease cleavage and methylamine- or trypsin induced conformational change.

[0040] Figure 22 shows: Level of ERK1 / 2 phosphorylation in Jurkat cells stimulated by A2M-BiTEs in different conformations (A) BiTE2-TEV4 (SEQ ID NO: 258), (B) BiTE3-TEV5 (SEQ ID NO: 259) in the presence of EGFR-expressing human colon carcinoma (HCT116) cells as target cells, or (C) BiTE2-TEV4 (SEQ ID NO: 258) used with wild-type or EGFR-expressing CHO cells as target cells. Conformational change, here induced by trypsin, to unshield the anti-CD3 of the BiTEs and the presence of target cells expressing the tumor associated antigen (EGFR) were required to induce ERK1 / 2 phosphorylation in the Jurkat cells.

[0041] Figure 23 shows: Antigen-specific target cell killing by primary T cells obtained from PBMCs once the conformation of the A2M-BiTE construct is changed so that the anti-CD3 is accessible; conformation change was induced either by (A) Thermolysin, with CHO as target cells or (B) methylamine or trypsin, with HCT116 as target cells. Conformational change was required to induce an effective cytotoxic T cell response.

[0042] The present invention will now be described in more detail in the following.General

[0043] In order for the present invention to be more readily understood, certain terms are first defined below. Additional definitions for the following terms and other terms are set forth throughout the specification.

[0044] As used in this specification and the appended claims, the singular forms “a,” “an” and “the” include plural referents unless the context clearly dictates otherwise. For example, “a biopharmaceutical moiety” is understood to represent one or more biopharmaceutical moieties. As such, the terms “a” (or “an”), “one or more”, and “at least one” can be used interchangeably herein.

[0045] Unless specifically stated or obvious from context, as used herein, the term “or” is understood to be inclusive and covers both “or” and “and”. Furthermore, “and / or” where used herein is to be taken as specific disclosure of each of the two specified features or components with or without the other. Thus, the term “and / or” as used in a phrase such as “A and / or B” herein is intended to include “A and B”, “A or B”, “A” (alone), and “B” (alone). Likewise, the term "and / or" as used in a phrase such as “A, B, and / or C” is intended to encompass each of the following aspects: A, B, and C; A, B, or C; A or C; A or B; B or C; A and C; A and B; B and C; A (alone); B (alone); and C (alone).

[0046] Throughout this specification and embodiments, the words “have” and “comprise”, or variations such as “has”, “having”, “comprises”, or “comprising” will be understood to imply the inclusion of a stated element, feature, or integer, or group of elements, features, or integers, but not the exclusion of any other elements, features, or integers or group of elements, features, or integers. It is furtherunderstood that wherever embodiments are described herein with the language “comprising” or “having” of grammatical equivalents thereof, otherwise analogous embodiments described in terms of “consisting of’ and / or “consisting essentially of’ are also provided.

[0047] As used herein, the term “about” refers to an interval of accuracy that a person skilled in the art will understand to still ensure the technical effect of the feature in question. The term indicates a deviation from the indicated numerical value of ±10%. In some embodiments, the deviation is ±5% of the indicated numerical value. In certain embodiments, the deviation is ±1% of the indicated numerical value.

[0048] The terms “variant” and “homolog” are used interchangeable to refer to proteins in which at least one function of the reference protein is preserved (e.g., to undergo a conformational change upon cleavage by a protease). In some embodiments, a variant or homolog is at least 70%, at least 80%, at least 85%, at least 90%, at least 91%, at least 92%, at least 93%, at least 94%, at least 95%, at least 96%, at least 97%, at least 98%, or 99% identical to a wildtype version of the reference protein (e.g., a CP AMD protein such as A2M, e.g., human A2M comprising or consisting of the amino acid sequence set forth in SEQ ID NO: 1).

[0049] As used herein, the term “fragment” refers to a protein that is truncated (e.g., N-terminally and / or C-terminally) by one or more amino acids or comprises one or more deletions of amino acids while preserving at least one function of the reference protein (e.g., to specifically bind an antigen or receptor, for instance in case of an antibody or cytokine, or to undergo a conformational change upon cleavage by a protease, for instance in case of a CPAMD protein such as A2M).

[0050] As used herein, the terms “therapeutic” and “therapeutically active” refer to any pharmaceutical, drug or composition that can be used to treat or prevent a disease, illness, condition or disorder or bodily function.

[0051] As used herein, the term “substantially” refers to the qualitative condition of exhibiting total or near-total extent or degree of a characteristic or property of interest. One of ordinary skill in the biological arts will understand that biological and chemical phenomena rarely, if ever, go to completion and / or proceed to completeness or achieve or avoid an absolute result. The term “substantially” is therefore used herein to capture the potential lack of completeness inherent in many biological and chemical phenomena.

[0052] As used herein, the term “in vitro” refers to events that occur in an artificial environment, e.g., in a test tube or reaction vessel, in cell culture, etc., rather than within a multi-cellular organism.

[0053] As used herein, the term “in vivo” refers to events that occur within a multi-cellular organism, such as a human and a non-human animal. In the context of cell-based systems, the term may be used to refer to events that occur within a living cell (as opposed to, for example, in vitro systems).Definitions

[0054] Prior to discussing the present invention in further detail, the following terms and conventions will first be defined:Alpha-2-macroglobulin (A2M)

[0055] The term “A2M” is to be understood as referring to an alpha-2 -macroglobulin protein, or variants or fragments thereof, that comprise (1) a bait region with at least one protease cleavage site, and (2) a Receptor Binding Domain (RBD), and are capable of altering conformation upon proteolytic cleavage of the at least one protease cleavage site. A2M is also known as C3 and PZP-like alpha-2 - macroglobulin domain-containing protein 5 (CPAMD5). Preferably A2M may be the human A2M protein (NCBI #9606, Uniprot P01023). The amino acid sequence of human A2M is given in SEQ ID NO: 1, with the naturally occurring polymorphisms I1000V and N639D. Unless indicated otherwise, residue numbers that are provided herein to identify specific amino acids or regions of A2M refer to the residues as set forth in SEQ ID NO: 1. It will be apparent to the skilled person that the numbering may differ in A2M variants that comprise one or more of the modifications described herein.CPAMD

[0056] The term “CPAMD” is to be understood as referring to a member of the C3 and PZP-like alpha- 2 -macroglobulin domain-containing protein (CPAMD) protein family, to which A2M belongs. An illustrative list of CPAMD proteins is provided in Table 1. In some embodiments, a proteinaceous prodrug construct of the invention may comprise a variant or fragment of a naturally occurring CPAMD protein. Such variants or fragments retain the capability of shielding the one or more drugs and altering their conformation upon proteolytic cleavage of the at least one protease cleavage site comprised in them to make the one or more drugs comprised in the proteinaceous prodrug construct accessible. Typically, such proteins, and variant or fragments thereof, form multimers, specifically homodimers or homotetramers.RBD domain

[0057] The terms “RBD” or “RBD domain” is to be understood as the receptor-binding domain of a CPAMD protein (e.g., A2M). In the native human A2M protein, the RBD is located at its C-terminus and spans amino acids 1335-1474 of A2M. Y1452 and Y1453 are involved in the formation of thiol ester groups. The thiol ester groups stabilize the molecule in its “native” conformation.

[0058] The amino acid sequence of the RBD domain of native human A2M is given in SEQ ID NO: 3. The RBD domain is also known as the macroglobulin 8 (MG8) domain (which is the term used in the examples). In the proteinaceous prodrug constructs described herein, one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) is positioned inside or in the vicinity of the RBD such that the CPAMD protein (e.g., A2M) remains capable of altering conformation upon proteolytic cleavage of at least one protease cleavage site (e.g., the first protease cleavage site) that is comprised in the bait region of the CPAMD protein.Inaccessible

[0059] The term “inaccessible” is to be understood as a drug (e.g., the first drug or antibody) of the proteinaceous prodrug construct possessing a decreased ability to interact with its binding partner (e.g., a target antigen) when the construct is in a “closed” conformation (i.e., the protease cleavage site in the bait region is not proteolytically cleaved). In some embodiments, the ability of the drug to interact with its binding partner is decreased by 90% or more (e.g., 95% or 99% or more) in the closed conformation relative to the unshielded (or “open”) conformation. Thus, the drug is “inaccessible” to its binding partner (e.g., in the case the drug is an antibody such as scFv or nanobody).

[0060] Thus, the term “inaccessible” may also be understood as the drug being “inactive”, in an “inactivated state”, or “shielded”.

[0061] Thus, in an embodiment, a. the one or more drugs is inaccessible when the bait region in the CPAMD protein (e.g., A2M) has not been proteolytically cleaved; and b. the one or more drugs is accessible when the bait region in the CPAMD protein (e.g., A2M) has been proteolytically cleaved.Accessible

[0062] The term “accessible” is to be understood as a drug (e.g., an antibody such as scFv or nanobody) of the proteinaceous prodrug construct being capable of interacting with its binding partner (e.g., a target antigen). Thus, the term “accessible” may also be understood as the drug being “unshielded”.

[0063] For example, a drug fused to the C-terminal end of the RBD is capable of interacting with its binding partner whether the construct is in a “closed” conformation (i.e., the protease cleavage site in the bait region is not proteolytically cleaved) or in an “open” conformation, i.e., its accessibility is independent of protease cleavage. A drug positioned inside the RBD becomes accessible in a proteasedependent manner, i.e., requires a protease to cleave a protease cleavage site located in the bait region of the proteinaceous prodrug construct described herein. Cleavage alters the conformation of the proteinaceous prodrug, such that CPAMD protein changes from being in a “closed” conformation to being in an “open” conformation.Bait region

[0064] The term “bait region” is to be understood as the region of a CPAMD protein (e.g., A2M) that comprises at least one protease cleavage site. In the native human A2M protein, the bait region spans amino acids 690-728 of A2M. The sequence of the bait region of native human A2M is given in SEQ ID NO: 4. The bait region of native human A2M is preferentially cleaved by most proteases, and bait region cleavage triggers A2M’s conformational change. The bait region sequence may be modified in order to change the selection of proteases that are able to cleave the bait region and trigger conformational change of the CPAMD protein.Biopharmaceutical moiety

[0065] The term “Biopharmaceutical moiety” is to be understood as a protein or fragment of a protein (e.g., a peptide or polypeptide) with therapeutic properties that can be incorporated into a proteinaceous prodrug construct with a CP AMD protein (e.g., A2M) in order to produce a proteolytically activatable prodrug. The term is used interchangeably herein with the term “drug”. Examples of biopharmaceutical moieties include antibody fragments such as single-domain antibodies (e.g. nanobodies) or single-chain variable fragments (scFvs), cytokines, or fragments of cell surface receptors or ligands. In a typical embodiment, the drug is an antigen-targeting moiety such as an antibody, e.g., a single-domain antibody such as a nanobody or scFv. In some embodiments, the first and the second antibodies are antigentargeting moieties such as antibodies. In some embodiments, the first antibody is a scFv and the second antibody is a nanobody (or vice versa).

[0066] Example sequences are given for the EGFR-binding nanobody EgAl (SEQ ID NO: 27), the scFv from PDL1 -binding Atezolizumab (SEQ ID NO: 28), the PDL1 -binding nanobody KN035 (SEQ ID NO: 29), the scFv from PDl-binding Nivolumab (SEQ ID NO: 30), the scFv from CTLA-4-binding Ipilimumab (SEQ ID NO: 31), the scFv from CD3-binding Foralumab (SEQ ID NO: 32), the scFv from CD3-binding Muromonab (SEQ ID NO: 33), the scFv from 4-lBB-binding Urelumab (SEQ ID NO: 34), the scFv from TNFa-binding Nivolumab (SEQ ID NO: 35), the IL2 cytokine (SEQ ID NO: 36), the extracellular region of the PD1 receptor (SEQ ID NO: 39), or a CD3-binding scFv derived from Tebentafusp (SEQ ID NO: 252).

[0067] Terms such as “biopharmaceutical moiety”, “drug”, “therapeutic peptide”, “therapeutic polypeptide” or “therapeutic protein”, “active agent” are used herein to refer to proteinaceous compounds that can be used to treat or prevent a disease, illness, condition, or disorder of bodily function.Antigen-targeting moiety

[0068] The term “antigen-targeting moiety” of the invention includes single-chain variable fragment, monoclonal, recombinant, chimeric, humanized, fully human, single-chain, single-domain and / or bi- specific antibodies including antibody fragments. Examples of such fragments include Fab F(ab'), F(ab)', Fv, and sFv fragments. The term is used herein interchangeably with the term “antigen-binding moiety”.

[0069] The antibodies may be generated by enzymatic cleavage of full-length antibodies or by recombinant DNA techniques, such as expression of recombinant plasmids containing nucleic acid sequences encoding antibody variable regions. In a typical embodiment, the antigen-targeting moieties in the proteinaceous prodrug constructs of the invention are single-chain antibodies such as scFvs or single-domain antibodies (e.g., nanobodies)

[0070] A “Single-chain Fv”, “sFv” or “scFv” antibody comprises a VH domain and a VL domain in a single polypeptide chain. The VH and VL are typically linked by a peptide linker. Any suitable linkermay be used. In some embodiments, the linker is a (GGGGS)n (SEQ ID NO: 223) or a (GGS)n. In some embodiments, n = 1, 2, 3, 4, 5, or 6.

[0071] The term “single-domain antibody” refers to an antigen-targeting moiety in which one variable domain of an antibody specifically binds to an antigen without the presence of another variable domain. Single domain antibodies include nanobodies.

[0072] An antigen is a molecule or a portion of a molecule capable of being bound by an antibody, which is additionally capable of inducing an animal to produce antibody capable of binding to an epitope of that antigen. An antigen can have one or more epitopes. The specific reaction referred to above is meant to indicate that the antigen will react, in a highly selective manner, with its corresponding antibody and not with the multitude of other antibodies, which can be evoked by other antigens. Accordingly, the antigen-targeting moieties in the proteinaceous prodrug constructs of the invention specifically bind the respective target antigens.

[0073] The antigen-targeting moieties for use in the proteinaceous prodrug constructs of the invention can be obtained in various ways known to a person of skill in the art of antibody production. For example, monoclonal antibodies (mAbs) contain a substantially homogeneous population of antibodies specific to antigens, which population contains substantially similar epitope binding sites. Such antibodies may be of any immunoglobulin class including IgG, IgM, IgE, IgA, IgD and any subclass thereof. A hybridoma producing a monoclonal antibody for use in the present invention may be cultivated in vitro, in situ, or in vivo. Production of high titers in vivo or in situ is a preferred method of production.

[0074] Chimeric antibodies are molecules in which different portions are derived from different animal species, such as those having a variable region derived from a murine monoclonal antibody and a human immunoglobulin constant region.

[0075] The term "chimeric antibody", as used herein, includes monovalent, divalent or polyvalent immunoglobulins. A monovalent chimeric antibody is a dimer (HL) formed by a chimeric H chain associated through disulfide bridges with a chimeric L chain. A divalent chimeric antibody is tetramer (H2L2) formed by two HL dimers associated through at least one disulfide bridge. A polyvalent chimeric antibody can also be produced, for example, by employing a CH region that aggregates (e.g., from an IgM H chain, or [micro] chain).

[0076] Murine and chimeric antibodies, fragments and regions of the present invention may comprise individual heavy (H) and / or light (L) immunoglobulin chains.

[0077] Selective binding agents, such as antibodies, fragments, or derivatives, having chimeric H chains and L chains of the same or different variable region binding specificity, can also be prepared by the appropriate association of the individual polypeptide chains.

[0078] In some embodiments, the term “antibody” as used herein refers to a single-chain or singledomain antibody.Receptor binding Domain (RBD)

[0079] The present disclosure several ways by which a biopharmaceutical moiety (e.g., a first drug or antibody) can be positioned inside the RBD to shield it or render it inaccessible until a protease cleavage site (e.g., a first protease cleavage site) located in the bait region of the CPAMD protein is cleaved. ciRBD

[0080] The term “ciRBD” is to be understood as proteinaceous fusion constructs between a CPAMD protein (e.g., A2M) and a biopharmaceutical moiety (e.g., a therapeutic peptide, polypeptide or protein), where the biopharmaceutical moiety is placed into the RBD domain at a position between the residues that correspond to residues 1402 and 1403 of native human A2M, without removing any of residues of the CPAMD protein. Linker sequences may be used to connect the N-terminus of the biopharmaceutical moiety with the carboxyl end of residue 1402 (SEQ ID NO: 78) and to connect the C-terminus of the biopharmaceutical moiety with the amino end of residue 1403 (SEQ ID NO: 79). An example of a ciRBD fusion construct incorporating the EgAl nanobody (SEQ ID NO: 27) into A2M is given in SEQ ID NO: 5-6. iRBD

[0081] The term “iRBD” is to be understood as proteinaceous fusion constructs between a CPAMD protein (e.g., A2M) and a biopharmaceutical moiety (e.g., a therapeutic peptide, polypeptide or protein), where the biopharmaceutical moiety replaces the residues of the RBD domain corresponding to the residues spanning from and including position 1392, to and including 1403 in native human A2M. The biopharmaceutical moiety is connected to residue 1391 by an N-terminal linker (SEQ ID NO: 80) and to residue 1404 by a C-terminal linker (SEQ ID NO: 81). An example of an iRBD fusion construct incorporating the EgAl nanobody (SEQ ID NO: 27) into A2M is given in SEQ ID NO: 84-85. miRBD

[0082] The term “miRBD” is to be understood as proteinaceous fusion constructs between a CPAMD protein (e.g., A2M) and a biopharmaceutical moiety (e.g., a therapeutic peptide, polypeptide or protein), where the biopharmaceutical moiety replaces the residues of the RBD domain corresponding to the residues spanning from and including position 1393, to and including 1395 of native human A2M. The biopharmaceutical moiety is connected to residue 1392 by an N-terminal linker (SEQ ID NO: 82) and to residue 1396 by a C-terminal linker (SEQ ID NO: 83). An example of a miRBD fusion construct incorporating the EgAl nanobody (SEQ ID NO: 27) into A2M is given in SEQ ID NO: 86-87. tRBD

[0083] The term “tRBD” is to be understood as proteinaceous fusion constructs between a CPAMD protein (e.g., A2M) and a biopharmaceutical moiety (e.g., a therapeutic peptide, polypeptide or protein),where the biopharmaceutical moiety is incorporated into a position C-terminal to the RBD domain. Furthermore, residues 1393 to 1402 of the RBD domain, or the corresponding residues of the RBD domain of another CPAMD protein, are modified to enable the formation of an a-helix with a sequence that is complementary to that of another a-helix that is positioned at the N-terminus of the biopharmaceutical moiety. The RBD domain a-helix and the a-helix at the N-terminus of the biopharmaceutical moiety are designed to interact with each with coiled-coil interactions. These coiled- coil interactions bring the biopharmaceutical moiety into a position relative to the RBD domain which facilitates shielding of the biopharmaceutical moiety by the CPAMD protein (e.g., A2M). The biopharmaceutical moiety is connected at its N-terminus to the C -terminus of its adjacent a-helix by a 2-residue linker, and the a-helix itself is connected at its N-terminus to the C-terminus of the RBD domain by a 15 -residue linker. An example of a tRBD fusion construct incorporating the EgAl nanobody (SEQ ID NO: 27) into A2M is given in SEQ ID NO: 92-93.Epitope

[0084] In the present context, the term “epitope” refers to the part of an antigen that is recognized by the immune system.Eukaryotic expression vector

[0085] In the present context, the term “eukaryotic expression vector” refers to a tool (e.g., a nucleic acid) used to introduce a specific coding polynucleotide sequence into a target cell, comprising expression control sequences (e.g., a suitable promoter sequence) operatively linked to a nucleotide sequence to be expressed.Sequence identity

[0086] In the present context, the term "sequence identity" is here defined as the sequence identity between genes or proteins at the nucleotide, base or amino acid level, respectively. Specifically, a DNA and an RNA sequence are considered identical if the transcript of the DNA sequence can be transcribed to the corresponding RNA sequence.

[0087] Thus, in the present context, "sequence identity" is a measure of identity between proteins at the amino acid level and a measure of identity between nucleic acids at nucleotide level. The protein sequence identity may be determined by comparing the amino acid sequence in a given position in each sequence when the sequences are aligned. Similarly, the nucleic acid sequence identity may be determined by comparing the nucleotide sequence in a given position in each sequence when the sequences are aligned.

[0088] To determine the percent identity of two amino acid sequences or of two nucleic acids, the sequences are aligned for optimal comparison purposes (e.g., gaps may be introduced in the sequence of a first amino acid or nucleic acid sequence for optimal alignment with a second amino or nucleic acid sequence). The amino acid residues or nucleotides at corresponding amino acid positions or nucleotidepositions are then compared. When a position in the first sequence is occupied by the same amino acid residue or nucleotide at the corresponding position in the second sequence, then the molecules are identical in that position. The percent identity between the two sequences is a function of the number of identical positions shared by the sequences (i.e., % identity = # of identical positions / total # of positions (e.g., overlapping positions) x 100). In one embodiment, the two sequences are the same length.

[0089] In another embodiment, the two sequences are of different length and gaps are seen as different positions. One may manually align the sequences and count the number of identical amino acids. Alternatively, alignment of two sequences for the determination of percent identity may be accomplished using a mathematical algorithm. Such an algorithm is incorporated into the BLASTN and BLASTX programs of (Altschul et al. 1990). BLAST nucleotide searches may be performed with the NBLAST program, to obtain nucleotide sequences homologous to a nucleic acid molecule of the invention. BLAST protein searches may be performed with the BLASTX program, to obtain amino acid sequences homologous to a protein molecule of the invention.

[0090] To obtain gapped alignments for comparison purposes, Gapped BLAST may be utilized. Alternatively, PSI-Blast may be used to perform an iterated search that detects distant relationships between molecules. When utilizing the BLASTN, BLASTX, and Gapped BLAST programs, the default parameters of the respective programs may be used. See http: / / www.ncbi.nlm.nih.gov. Alternatively, sequence identity may be calculated after the sequences have been aligned e.g. by the BLAST program in the EMBL database (www.ncbi.nlm.gov / cgi-bin / BLAST). Generally, the default settings with respect to e.g. “scoring matrix” and “gap penalty” may be used for alignment. In the context of the present invention, the BLASTN and PSI BLAST default settings may be advantageous.

[0091] The percent identity between two sequences may be determined using techniques similar to those described above, with or without allowing gaps. In calculating percent identity, only exact matches are counted. An embodiment of the present invention thus relates to sequences of the present invention that has some degree of sequence variation.Subject

[0092] The term "subject" comprises humans of all ages, other primates (e.g., cynomolgus monkeys, rhesus monkeys); mammals in general, including commercially relevant mammals, such as cattle, pigs, horses, sheep, goats, mink, ferrets, hamsters, cats and dogs, as well as birds. Preferred subjects are humans.

[0093] The term “subject” also includes healthy subjects of the population and, in particular, healthy subjects, who are exposed to pathogens and in need of protection against infection, such as health personnel.

[0094] It should be noted that embodiments and features described in the context of one of the aspects of the present invention also apply to the other aspects of the invention.

[0095] All patent and non-patent references cited in the present application, are hereby incorporated by reference in their entirety.Detailed description of the invention

[0096] In certain aspects, the present disclosure relates to a fusion protein (e.g., proteinaceous prodrug construct) comprising a complement 3- and pregnancy zone protein-like, alpha-2 -macroglobulin domain-containing (CPAMD) protein, such as A2M, comprising:(a) a bait region comprising at least one first protease cleavage site;(b) a Receptor Binding Domain (RBD);(c) a second protease cleavage site introduced at the N-terminal end of the CPAMD protein’s RBD domain, wherein the RBD domain remain bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the second protease cleavage site; and(d) at least one biopharmaceutical moiety positioned C-terminal to the second protease cleavage site, wherein cleavage of the first protease cleavage site results in release of the RBD domain, and thereby the at least one biopharmaceutical moiety, from the CPAMD protein.

[0097] In some embodiments, a fusion protein comprising a complement 3- and pregnancy zone protein-like, alpha-2 -macroglobulin domain-containing (CPAMD) protein is provided, comprising:(a) a bait region comprising a first protease cleavage site;(b) a Receptor Binding Domain (RBD) comprising: a second protease cleavage site at the N-terminal end of the RBD; and a drug fused to the C-terminal end of the RBD; wherein:(i) the RBD remains bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the second protease cleavage site; and(ii) the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site releasing the RBD.In these embodiments, the drug is accessible.

[0098] Without wishing to be bound by any particular theory, the inventors believe that the release of the RBD comprising the drug in a protease-activatable manner can improve tissue penetration. For example, the first cleavage site (and optionally the second protease cleavage) may be such that it is cleaved by one or more proteases present in a tissue affected by a disease (e.g., a tissue comprising a tumor or a site of inflammation).

[0099] In some embodiments, the first and second protease cleavage sites are different. For example, it may be advantageous to “pre-cleave” the second protease cleavage site during the manufacturing of a fusion protein or proteinaceous prodrug construct disclosed herein.

[0100] One suitable protease is furin. Furin is expressed in the Golgi apparatus of mammalian cells. When a furin cleavage site is used as the second protease cleavage site in a fusion protein or proteinaceous prodrug disclosed herein, cleavage of the second protease cleavage site occurs during recombinant expression in a mammalian host cell.

[0101] Another suitable protease is a Tobacco etch virus (TEV) protease. TEV proteases are highly site-specific. TEV cleavage sites are commonly included recombinantly produced proteins to allow, e.g., removal of affinity tags used for the purification of the recombinantly produced protein from host cell component.

[0102] As shown by example 12 and 13, specific prodrugs can be generated wherein the second protease cleavage site at the N-terminus of the RBD domain is either a furin or TEV cleavage site. Thus, in a specific aspect, the invention relates to a proteinaceous prodrug construct comprising a complement 3- and pregnancy zone protein-like, alpha-2 -macroglobulin domain-containing (CPAMD) protein, such as A2M, comprising:(a) a bait region comprising at least one first protease cleavage site;(b) a Receptor Binding Domain (RBD);(c) a fiirin protease cleavage site according to SEQ ID NO: 225 or a TEV protease cleavage site according to SEQ ID NO: 226 introduced at the N-terminal end of the CPAMD protein’s RBD domain, such as from residue position 1334 to 1340 of A2M, wherein the RBD domain remain bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the fiirin or TEV protease cleavage site; and(d) at least one biopharmaceutical moiety positioned C-terminal to the second protease cleavage site, wherein cleavage of the first protease cleavage site results in release of the RBD domain, and thereby the at least one biopharmaceutical moiety, from the CPAMD protein.

[0103] Such fusion proteins or proteinaceous prodrug constructs can be turned into multi-specific, such as bi-specific prodrugs, by the introduction of a additional biopharmaceutical moiety (e.g., an antibody). In some embodiments, a first biopharmaceutical moiety (e.g., a first drug or antibody) is positioned inside the RBD region, and a second biopharmaceutical moiety (e.g., a second drug or antibody) is positioned C-terminally to the CPAMD protein (e.g., fused to the C-terminal end of the RBD). In such embodiments, the CPAMD protein shields the first biopharmaceutical moiety and the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD, and thereby the first- and second biopharmaceutical moiety, from the CPAMD protein, and making the first biopharmaceutical moiety accessible.

[0104] The skilled person will appreciate that the functioning of a bi-specific prodrug may not be dependent on the introduction of a second protease cleavage site. Accordingly, in certain aspects, the invention provides a proteinaceous prodrug construct (e.g., a multi- or bispecific proteinaceous prodrug construct) comprising a complement 3- and pregnancy zone protein-like, alpha-2 -macroglobulin domain-containing (CPAMD) protein, such as A2M, comprising:(a) a bait region comprising at least one first protease cleavage site;(b) a Receptor Binding Domain (RBD); and(c) at least two biopharmaceutical moieties (e.g., first and second antibodies) positioned C-terminal to the RBD, wherein at least one first biopharmaceutical moiety (e.g., a first antibody) is positioned inside the RBD domain and at least one second biopharmaceutical moiety (e.g., a second antibody) is positioned in a non-shielded position C-terminal to the RBD domain, wherein the CPAMD protein shields the first biopharmaceutical moiety (e.g., the first antibody) and the CPAMD protein or fragment thereof is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, making the first biopharmaceutical moiety unshielded.

[0105] In some embodiments, a second protease cleavage site is introduced at the N-terminal end of the CPAMD protein’s RBD domain, wherein the RBD domain remains bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the second protease cleavage site and releases the RBD upon cleavage of the first protease cleavage site, thereby releasing the first- and second biopharmaceutical moiety from the CPAMD protein.

[0106] As such, these aspects may be combined and thus in another aspect, the invention relates to a proteinaceous prodrug construct comprising a complement 3- and pregnancy zone protein-like, alpha- 2-macroglobulin domain-containing (CPAMD) protein, such as A2M, comprising:(a) a bait region comprising at least one first protease cleavage site;(b) a Receptor Binding Domain (RBD);(c) a second protease cleavage site introduced at the N-terminal end of the CPAMD protein’s RBD domain, wherein the RBD domain remain bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the second protease cleavage site; and(d) at least two biopharmaceutical moieties positioned C-terminal to the second protease cleavage site, wherein at least one first biopharmaceutical moiety is positioned inside the RBD domain and at least one second biopharmaceutical moiety is positioned in a non-shielded position C-terminal to the RBD domain, wherein the CPAMD protein or fragment thereof shields the first biopharmaceutical moiety and the CPAMD protein or fragment thereof is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD, and thereby the first- and second biopharmaceutical moiety, from the CPAMD protein, and making the first biopharmaceutical moiety unshielded.

[0107] In particular embodiments, a proteinaceous prodrug construct is provided that comprises a complement 3- and pregnancy zone protein-like, alpha-2-macroglobulin domain-containing (CPAMD) protein, comprising:(a) a bait region comprising a first protease cleavage site; and(b) a Receptor Binding Domain (RBD) comprising: a second protease cleavage site at the N-terminal end of the RBD,a first biopharmaceutical moiety (e.g., a first drug or antibody) positioned inside the RBD, and a second biopharmaceutical moiety (e.g., a second drug or antibody) fused to the C- terminal end of the RBD; wherein:(i) the CPAMD protein shields the first biopharmaceutical moiety and the second biopharmaceutical moiety is accessible (i.e., unshielded);(ii) the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD and thereby making the first biopharmaceutical moiety accessible.In these embodiments, the RBD remains bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the second protease cleavage site

[0108] In one specific embodiment, the first biopharmaceutical moiety is an antigen-binding moiety that specifically binds CD3, and the second biopharmaceutical moiety is an antigen binding moiety that specifically bind a tumor cell surface antigen.

[0109] In specific embodiments, the first biopharmaceutical moiety is atoxic protein (e.g., a cytotoxic protein), and the second biopharmaceutical moiety is an antigen binding moiety that specifically bind a tumor cell surface antigen.

[0110] Typically, the CPAMD protein is human. In some embodiments, the CPAMD protein is A2M. In some embodiments, the CPAMD protein is human A2M.Targeted prodrugs[OHl] The proteinaceous prodrug construct according to the present invention may be engineered such as to direct the location at which the prodrug performs its function. In some embodiments, the proteinaceous prodrug construct comprises a biopharmaceutical moiety that is capable of directing the proteinaceous prodrug construct to a specific tissue (e.g., cancerous or inflamed tissue), a specific cell type (e.g., an immune cell or a tumor cell), and / or a specific receptor (e.g., a T cell receptor or a tumor cell surface antigen). In particular embodiments, such a biopharmaceutical moiety is fused to the C- terminal end of the RBD.

[0112] Specific cells may, for example, be immune cells, such as when engineering bi-specific proteinaceous prodrug constructs (as illustrated in, e.g., Examples 14-17). Accordingly, in some embodiments, the proteinaceous prodrug construct comprises a biopharmaceutical moiety that is capable of directing the proteinaceous prodrug construct to immune cells. In particular embodiments, such a biopharmaceutical moiety is fused to the C-terminal end of the RBD. In some embodiments, the immune cells are NK cells, macrophages, T cells, or dendritic cells. In particular embodiments, the immune cells are T cells. In a specific embodiment, the biopharmaceutical moiety is an antibody that specifically binds to a T cell receptor (e.g., CD3).

[0113] Since the prodrugs can provide both shielded and unshielded moieties, it is possible to select whether the above described directing is occurring before or after the conformational change in the CPAMD protein. In some embodiments, the at least one biopharmaceutical moiety has specificity against a target and is able to bind to its target before cleavage of the first protease cleavage site. In some embodiments, the at least one biopharmaceutical moiety has specificity against a target and is not able to bind to its target before cleavage of the second protease cleavage site. In some embodiments, the at least one biopharmaceutical moiety is a targeting moiety.

[0114] As detailed below, a particularly preferred embodiment is the generation of BiTEs, and an essential part of BiTEs is the T cell specificity, however without being bound by theory, prodrugs may also be manufactured with T cell specificity without being a BiTE. In some embodiments, at least one biopharmaceutical moiety, e.g., the first biopharmaceutical moiety (i.e., the first antibody), is T cell specific, such as a T cell specific moiety. In some embodiments, the T cell specific moiety is specific against a receptor expressed at increased levels on T cells, such as CD3, CD4 and / or CD8, preferably CD3. In some embodiments, the T cell specific moiety is an anti-CD3 moiety.

[0115] It is noted in particular, that when a prodrug is developed with specificity against certain T cell receptors, such as with anti-CD3 BiTEs, using a multimeric CPAMD protein, it is in particular important to develop a proteinaceous prodrug comprising a second protease cleavage site that is processed before introduction into the circulation of a subject, to avoid premature activation of T cells through crosslinking due to receptor binding by multivalent antibodies.

[0116] In some embodiments, the biopharmaceutical moiety is NK cell-specific, such as an NK cellspecific moiety. In some embodiments, the NK cell-specific moiety is specific against a receptor expressed at increased levels on NK cells, such as CD 16. In some embodiments, the NK cell-specific moiety is an antibody against CD 16.

[0117] In some embodiments, the biopharmaceutical moiety is macrophage specific, such as a macrophage specific moiety. In some embodiments, the macrophage specific moiety is specific against a receptor or molecule expressed at increased levels on macrophages, such as SIRPa. In some embodiments, the macrophage-specific moiety is an SIRPa inhibitory antibody, such as a SIRPa antibody that blocks CD47.

[0118] In some embodiments, the biopharmaceutical moiety is specific towards dendritic cells, such as a dendritic cell specific moiety. In some embodiments, the biopharmaceutical moiety is specific towards a receptor or molecule expressed at increased levels on dendritic cells, such as the DNGR1 receptor. In some embodiments, the dendritic cell-specific moiety is an antibody against the DNGR1 receptor.Multi-specific drugs

[0119] As presented herein, the proteinaceous prodrugs may also be multi-specific prodrugs, such as bi-specific prodrugs. In some embodiments, the RBD is a multi-specific drug upon release from the CPAMD protein. In some embodiments, the RBD is a bi-specific drug upon release from the CPAMD protein. In particular embodiments, the RBD comprises a first antibody and a second antibody that areboth accessible upon release from the CMAPD family protein. In some embodiments, the CPAMD protein is a multi-specific drug upon conformational change. In some embodiments, the CPAMD protein is a bi-specific drug upon conformational change. In particular embodiments, the CPAMD protein comprises a first antibody and a second antibody that are both accessible upon conformational change.

[0120] The at least two biopharmaceutical moieties chosen to create a multi-specific prodrug, such as a bi-bispecific prodrug, may be any combination of the biopharmaceutical moieties, drugs, etc. as described further herein.

[0121] When a proteinaceous prodrug is designed to comprise a first biopharmaceutical moiety in any of loops 1-4 of the RBD, and a second biopharmaceutical moiety at the C-terminal end of the RBD, the first biopharmaceutical moiety will be shielded until cleavage of the bait region, whilst the second biopharmaceutical moiety will not be shielded. This is dependent on the biopharmaceutical moiety at the C-terminal end not being further adapted such as to being tethered into the RBD domain, as described further below.

[0122] In some embodiments, the bi-specific drug comprises one biopharmaceutical moiety (e.g., a first antibody) capable of directing the proteinaceous prodrug construct to immune cells as described in the previous section and a drug (e.g., a second antibody that specifically binds a tumor cell surface) as described further below.

[0123] Such proteins can be engineered to be directed towards specific immune cells. In some embodiments, the bi-specific drug (e.g., a bispecific antibody prodrug construct) is a Bi-specific T cell engager (BiTE), Bi-specific NK cell engager (BiKE), Bi-specific Macrophage engager (BiME), or a Bi- specific dendritic cell engager (BiDE). In preferred embodiments, the bi-specific drug is a Bi-specific T cell engager (BiTE). These acronyms are somewhat established terms within the field. The acronyms cover a bi-specific molecule comprising a moiety targeting a specific cell type, for instance an anti-CD3 antibody fragment and a moiety targeting a therapeutic target (e.g., a tumor cell surface antigen), such as EGFR.

[0124] A bi-specific drug (e.g., a bispecific antibody prodrug construct) comprises at least two drugs (e.g., two antibodies) as described further below.

[0125] In some embodiments, a bispecific antibody prodrug construct comprises a CPAMD protein comprising:(a) a bait region comprising a first protease cleavage site;(b) a Receptor Binding Domain (RBD) comprising: a second protease cleavage site at the N-terminal end of the RBD; a first antibody positioned inside the RBD; and a second antibody fused to the C-terminal end of the RBD; wherein:(i) the CPAMD protein shields the first antibody and the second antibody is accessible;(ii) the RBD remains bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the second protease cleavage site; and(iii) the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD and thereof making the first antibody accessible.

[0126] In some embodiments, the second protease cleavage site is cleaved (such as during manufacturing of the proteinaceous prodrug construct, e.g., during or after recombinant expression of the construct in a host cell), yielding a bispecific, antibody prodrug construct comprising a CPAMD protein, comprising:(a) a bait region comprising a first protease cleavage site;(b) a Receptor Binding Domain (RBD) comprising a first antibody positioned inside the RBD; and a second antibody fused to the C-terminal end of the RBD; wherein:(i) the CPAMD protein shields the first antibody and the second antibody is accessible;(ii) the RBD is bound to the CPAMD protein, via non-covalent interactions; and(iii) the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD and thereof making the first antibody accessible.CPAMD proteins

[0127] The proteinaceous prodrug construct described herein comprises a complement 3- and pregnancy zone protein-like, alpha-2 -macroglobulin domain-containing (CPAMD) protein. The CPAMD protein comprises a bait region with at least one protease cleavage site and a Receptor Binding Domain (RBD).

[0128] In some embodiments, the one or more drugs are positioned inside or in the vicinity of any one of loops 1-4 of the RBD (e.g., loop 1, loop 2, loop 3, or loop 4). In one specific embodiment, the one or more drugs are positioned inside loop 2 (which is the most convenient site in the RBD for generating a prodrug construct disclosed herein that shields one or more drugs until cleavage of a protease site within the bait region). In another specific embodiment, the one or more drugs are position inside loop 3 (which, as shown herein, is a suitable alternative to loop 2). In yet another specific embodiment, the one or more drugs are positioned inside loop 4. In a further specific embodiment, the one or more drugs is positioned in the vicinity of loop 2 of the RBD.

[0129] In a proteinaceous prodrug construct of the invention, the CPAMD protein shields the drug positioned inside the RBD. The CPAMD protein, is capable of altering conformation upon proteolytic cleavage of at least one protease cleavage site comprised in the bait region, making the drug accessible.

[0130] While the invention is described in more detail in reference to proteinaceous prodrug constructs in which the CPAMD protein is an alpha-2 -macroglobulin (A2M), or a variant or functional homolog thereof, a person of skill in the art of proteinaceous prodrug design will appreciate that other CPAMD proteins can take the place of A2M.

[0131] In some embodiments, the proteinaceous prodrug construct is a fusion protein. In one embodiment, the proteinaceous fusion construct, comprises a member of the CPAMD family fused to one or more drugs; or a modified member of the CPAMD family (2) fused to one or more drugs; wherein the one or more drugs are positioned inside or in the vicinity of the RBD domain of A2M. Proteinaceous fusion construct and proteinaceous prodrug are used interchangeably herein.

[0132] In some embodiments, the one or more drugs are inserted in any one of loops 1-4 of the RBD. In some embodiments, the loop is modified by addition, substitution, or deletion of one or more amino acids to accommodate the one or more drugs. In some embodiments, the one or more drugs replace one or more amino acids of the loop. In particular embodiments, the loop is loop 2 of the RDB. In some embodiments, the loop is loop 3 of the RBD. In further embodiments, the loop is loop 4 of the RBD.

[0133] As discussed herein, placing one or more drugs in the vicinity of loop 2 of the RBD can be accomplished by insertion of the one or more drugs inside or within 5 amino acid residues of loop 2 (e.g., by replacing one or more residues, or by direct insertion). Similarly, this can be accomplished by insertion of the one or more drugs inside or within 5 amino acid residues of loop 1, loop 3, or loop 4 (e.g., by replacing one or more residues, or by direct insertion). Loops 1, 3 and 4 have respective distances of 27 A, 21 A, and 25 A to loop 2, as calculated from their centers of mass. In the ciRBD fusion approach described herein, the shortest restraint between the drug and loop 2 is the 15 -residue C-terminal linker. From an average length of 3.5 A per amino acid residue, it can be calculated that the one or more drugs can be positioned about 52 A (e.g., about 50 A, about 40 A, about 30 A, or about 20 A) away from loop 2 and occupy a position where its accessibility is dependent on the conformation of the CPAMD protein (e.g., A2M).

[0134] As an alternative to direct fusion, approaches can be designed to place the drug within an equivalent distance to loop 2 and with a similar orientation relative to the RBD domain as achieved by the direct fusion approach, through other means. For example, as described herein, coiled-coil interactions or high-affinity interactions can be used to anchor a drug to loop 2 (e.g., as in the tRBD approach described herein).

[0135] Table 1 provides an illustrative list of CPAMD proteins that may be used to implement the invention and also indicates the positions and sequence of each loop with the CPAMD protein.Table 1. Positions of loops 1-4 in various CPAMD proteins

[0136] In some embodiments, the CP AMD protein is selected from the group consisting of C3, C4A, C4B, C5, PZP, A2ML1, CD109, CPAMD8, Ovostatin homologue 1, Ovostatin homologue 2, and A2M. In some embodiments, the CPAMD protein is selected from A2M, PZP, Ovostatin 1, and Ovostatin 2, and functional homologs thereof. In some embodiments, the CPAMD protein is human A2M, or a functional homolog thereof, e.g., a mammalian A2M. In a particular embodiment, the CPAMD protein is A2M.

[0137] In one embodiment, the CPAMD protein is a human CPAMD protein, such as one of the proteins listed in Table 1, or a variant thereof. In some embodiments, the human CPAMD protein is a variant that has been modified as described herein, e.g., the variant may comprise a modified bait region.

[0138] In some embodiments, the CPAMD protein has at least about 70% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1. In some embodiments, the CPAMD protein has at least about 75% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least about 80% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least about 85% sequence identity to at least one of the full- length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least about 90% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1.

[0139] In one embodiment, the CPAMD protein has at least about 91 % sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least about 92% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least about 93% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least about 94% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least 95 % sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least about 96% sequence identity to at least one of the full- length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least about 97% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least about 98% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1. In one embodiment, the CPAMD protein has at least about 99% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1.

[0140] In further embodiments, the CPAMD protein is a human CPAMD protein, such as the proteins listed in Table 1, with the proviso that a. the bait region is modified as described herein; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0141] In another embodiment, the CPAMD protein has at least about 70% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1, with the proviso that a. the bait region is modified as described herein; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0142] In one embodiment, the CPAMD protein has at least about 80% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1, with the proviso that a. the bait region is modified as described herein; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0143] In one embodiment, the CPAMD protein has at least about 85% sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1, with the proviso thata. the bait region is modified as described herein; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0144] In one embodiment, the CPAMD protein has at least about 90% (e.g., at least 91%, at least 92%, at least 93% or at least 95%) sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1, with the proviso that a. the bait region is modified as described herein; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0145] In one embodiment, the CPAMD protein has at least about 95% (e.g., at least 96%, at least 97%, at least 98% or about 99%) sequence identity to at least one of the full-length CPAMD protein sequences listed in Table 1, with the proviso that a. the bait region is modified as described herein; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0146] The RBD domains and bait regions of the CPAMD proteins listed in Table 1 are described in Table 2. The RBD domains (also referred to as the “MG8 domain”) and bait regions (also referred to as “anaphylactic domain” in some CPAMD proteins) were identified on the basis of their functional equivalence to the corresponding domain / region of human A2M.Table 2. Positions of the RBD and bait regions in various CPAMD proteins* The available Ovostatinl sequence (Q6IE37.2) is of poor quality and likely incomplete.

[0147] When one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are introduced into the RBD, they are sterically hindered from interacting with other proteins such as their therapeutics targets. The RBD domain is itself a small domain (~16 kDa). Without wishing to be bound by any particular theory, the inventors believe that it is unlikely that the RBD domain is able to sterically hinder the one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) on its own, especially considering that linkers are typically present between the one or more drugs and the RBD domain. Without wishing to be bound by any particular theory, the inventors therefore believe that other portions or multiple copies of the CP AMD protein contribute to the surrounding and sequestering of the one or more drugs. For example, naturally occurring CPAMD proteins (e.g., A2M) form homotetramers.

[0148] In some embodiments, CPAMD protein (e.g., A2M) forms a multimer (e.g., a dimer or tetramer). In some embodiments, the multimer comprises identical subunits (e.g., homodimers or homotetramers). Without wishing to be bound by any particular theory, the inventors believe it to be possible that contributions from one or more adjacent subunits contribute to the sequestering of the one or more drugs.

[0149] Two human CPAMD proteins are known to form dimers (typically stabilized by one or more disulfide bridges), namely A2M and pregnancy zone protein (PZP, a.k.a. CPAMD6). In A2M, the disulfide-bridged dimer participates in additional non-covalent interactions with another disulfide- bridged dimer, primarily through their LNK regions, to form a tetramer. This tetramer formation is also seen in ovostatins, such as those that have been characterized in ducks, chickens, and frogs. The two human ovostatins, ovostatin 1 and ovostatin 2 are also predicted to be tetramers.

[0150] Accordingly, in some embodiments, a proteinaceous prodrug construct in accordance with the invention is capable of forming a multimer, e.g., a dimer or a tetramer. In some embodiments, the multimer is a heteromultimer (e.g., a heterodimer or heterotetramer). More typically, the multimer is a homodimer or homotetramer.

[0151] In some embodiments, the multimer (e.g., dimer or tetramer) formation occurs via a LNK region of the CPAMD protein. In some embodiments, a tetramer is formed by two disulfide-bridged dimers (e.g., two homodimers). In preferred embodiments, the CPAMD protein is a multimer.

[0152] The cysteines which form the inter-subunit disulfide bonds that are responsible for the disulfide- bridged dimer are found in two loops, one of which is located on the MG3 domain of the CPAMD protein and one of which is located on the MG4 domain of the CPAMD protein. These loops are defined in Table 3. The LNK region that has been shown to participate in interactions between the two disulfide- bridged dimers in tetramer-forming CPAMD proteins is also defined in Table 3.Table 3. Regions for multimer formation in various CPAMD proteins.

[0153] iRBD, miRBD, ciRBD, and tRBD as described herein create proteinaceous prodrug constructs by “locking” the location of a drug (e.g., a peptide, polypeptide or protein) in the vicinity of loop 2 (residues 1392-1405) on the RBD of CPAMD protein (e.g., A2M), either by direct fusion in the iRBD / miRBD / ciRBD approaches or by anchoring of the drug to this location with coiled-coil interactions in the tRBD approach. Other approaches that are able to anchor the drug in this general location relative to the RBD domain will be apparent to the skilled person.

[0154] In some embodiments, the proteinaceous prodrug construct comprises a first interaction domain and the one or more drugs comprise a second interaction domain, wherein the first and second interaction domains form a complex positioning the one or more drugs in the vicinity of any one of loops 1-4 (e.g., loop 1, loop 2, loop 3, or loop 4) of the RBD. In one specific embodiment, the first and second interaction domains form a complex positioning the one or more drugs in the vicinity of loop 2. In another specific embodiment, the first and second interaction domains form a complex positioning the one or more drugs in the vicinity of loop 4. In some embodiments, the first interaction domain and the second interaction domain form a coiled coil structure.

[0155] Without wishing to be bound by any particular theory, the inventors believe that the use of first and second interaction domains to position the one or more drugs in the vicinity of loop 2 of the RBD allows the CPAMD protein to take on its “native” conformation, thereby sequestering the one or more drugs inside it (thus, shielding it from interactions with one or more targets). Spatial proximity may be achieved, e.g., by inserting the first interaction domain in loop 2 of the RBD, or in one of loops 1-4 (e.g., loop 3) of the RBD.

[0156] One approach is the “docking” of a drug into the CPAMD protein (e.g., A2M). This can be done using a molecule that has an inherent affinity for loop 2 of the RBD, such as a functional fragment of the LRP1 receptor or an antibody (e.g., a nanobody) which recognizes a loop 2 epitope.

[0157] Alternatively, the RBD of the CPAMD protein (e.g., A2M) could be modified to facilitate such docking. For example, a tag sequence could be introduced into the RBD (e.g., in the “ciRBD” position), and the drug could be fused to an antibody (e.g., a nanobody or similar small binding domain) which recognizes the tag.

[0158] Accordingly, in some embodiments, the first interaction domain is a tag or epitope sequence within loop 2 of the RBD and the second interaction domain is a functional fragment of a receptor or antibody that is capable of binding specifically to the tag or epitope sequence.A2M

[0159] Alpha-2-macroglobulin (A2M) is a protein found at high concentrations (normally 1-5 g / L) in human plasma. A2M is a protease inhibitor with a well-characterized mechanism of action. First, proteases cleave an exposed and vulnerable stretch of sequence called the bait region, which is permissive to cleavage by most human proteases. Bait region cleavage triggers a conformational change in A2M that causes A2M to collapse around the protease, trapping the protease within A2M and preventing it from accessing additional large protein substrates (figure 1). Up to two proteases can be inhibited by a single A2M protein if cleavage is rapid and sequential. In addition to the trapping of the instigating protease(s), there are two additional consequences of the triggered conformational change: (i) a cryptic binding site on A2M for the LRP1 receptor is exposed, resulting in the binding of A2M- protease complex by cell surface LRP1 and the rapid clearance ofthese complexes, e.g. from circulation, by LRP1 -expressing hepatocytes, and (ii) a reactive thiol ester moiety is exposed on A2M, allowing the formation of covalent bonds to the trapped protease.

[0160] The present invention describes the incorporation of biopharmaceutical moieties into A2M in such a manner that the binding ability of the biopharmaceutical moiety is regulated by the conformation of A2M. Biopharmaceutical moieties suitable for use with the present invention include therapeutic peptides, polypeptides or proteins such as antibodies (e.g., single-chain or single domain antibodies such as scFvs and nanobodies). In the native conformation of A2M, the incorporated biopharmaceutical moiety occupies a shielded position where it has a decreased ability to interact with its therapeutic target. After the conformation of A2M is altered by proteolytic cleavage of the bait region (or, alternatively, by aminolysis of the thiol ester of A2M using methylamine, which triggers a similar conformational change), the biopharmaceutical moiety demonstrates an increased ability to interact with its target. By modification of A2M’s bait region sequence, specific proteases can be designated as able to cleave the bait region and trigger this conformational change. Altogether, this can be used to produce proteinaceous fusion constructs of A2M and a biopharmaceutical moiety (e.g., a therapeutic peptide, polypeptide or protein) that function as protease-activated prodrug versions of the biopharmaceutical moiety.

[0161] In one embodiment, the invention provides a proteinaceous prodrug construct, comprising: (a) an alpha-2 -macroglobulin (A2M) protein, and (b) one or more drugs, wherein (i) the A2M protein comprises (1) a bait region with at least one protease cleavage site, and (2) a Receptor Binding Domain (RBD), (ii) the one or more drugs are positioned inside or in the vicinity of the RBD, and (iii) the A2Mprotein shields the one or more drugs and is capable of altering conformation upon proteolytic cleavage of the at least one protease cleavage site, making the one or more drugs accessible.

[0162] In some embodiments, the present invention relates to a proteinaceous fusion construct comprising alpha-2 -macroglobulin (A2M), fused to one or more drugs; or a modified A2M fused to one or more drugs; wherein the one or more drugs are positioned inside or in the vicinity of the RBD domain of A2M.

[0163] In some embodiments, the one or more drugs is positioned inside or in the vicinity of any one of loops 1-4 of the RBD. In some embodiments, the proteinaceous prodrug construct is a fusion protein. In some embodiments, the one or more drugs are positioned inside any one of loops 1-4 of the RBD. In some embodiments, the loop is loop 1. In some embodiments, the loop is loop 2. In some embodiments, the loop is loop 3. In some embodiments, the loop is loop 4. In some embodiments, the loop is modified, in relation to a wildtype loop sequence, by addition, substitution or deletion of one or more amino acids to accommodate the one or more drugs. In some embodiments, the one or more drugs replace one or more amino acids of the loop.

[0164] In one embodiment, the one or more drugs, is inaccessible when the bait region in alpha-2- macroglobulin (A2M) has not been proteolytically cleaved; and the one or more drugs, is accessible when the bait region in alpha-2 -macroglobulin (A2M) has been proteolytically cleaved.

[0165] In one embodiment, the cleavage of the bait region can be effectuated by serine-, cysteine-, aspartic- and / or metalloproteinases.

[0166] The drug can be positioned on different locations within the sequence of the proteinaceous fusion construct.

[0167] The skilled person will be able to recognize the parts of the proteinaceous fusion construct, which originates from A2M. Thus, in embodiments where a drug is inserted into the sequence of A2M, the resulting fusion construct can be seen as a first part of A2M, a drug, and a second part of A2M. In such cases, the skilled person will be able to recognize the first- and the second part of A2M as a complete molecule. Thus, in a particular embodiment, sequence identity of A2M is to be calculated from two separate parts, based on the sequence deriving from A2M, and thus not including the one or more drugs.

[0168] In one embodiment, the A2M molecule is a mammalian A2M molecule or variant thereof, such as a human A2M molecule.

[0169] In one embodiment, the A2M molecule is a human A2M molecule, such as the sequence according to SEQ ID NO: 1, or a variant thereof. In some embodiments, the human A2M molecule is a variant that has been modified as described herein, e.g., the variant may comprise a modified bait region.

[0170] In some embodiments, the A2M molecule has at least about 70% sequence identity to the sequence according to SEQ ID NO: 1. In some embodiments, the A2M molecule has at least about 75% sequence identity to the sequence according to SEQ ID NO: 1. In one embodiment, the A2M molecule has at least about 80% sequence identity to the sequence according to SEQ ID NO: 1. In one embodiment, the A2M molecule has at least about 85% sequence identity to the sequence according toSEQ ID NO: 1. In one embodiment, the A2M molecule has at least about 90% sequence identity to the sequence according to SEQ ID NO: 1.

[0171] In one embodiment, the A2M molecule has at least about 91% sequence identity to the sequence according to SEQ ID NO: 1. In one embodiment, the A2M molecule has at least about 92% sequence identity to the sequence according to SEQ ID NO: 1. In one embodiment, the A2M molecule has at least about 93% sequence identity to the sequence according to SEQ ID NO: 1. In one embodiment, the A2M molecule has at least about 94% sequence identity to the sequence according to SEQ ID NO: 1. In one embodiment, the A2M molecule has at least 95 % sequence identity to the sequence according to SEQ ID NO: 1. In one embodiment, the A2M molecule has at least about 96% sequence identity to the sequence according to SEQ ID NO: 1. In one embodiment, the A2M molecule has at least about 97% sequence identity to the sequence according to SEQ ID NO: 1. In one embodiment, the A2M molecule has at least about 98% sequence identity to the sequence according to SEQ ID NO: 1. In one embodiment, the A2M molecule has at least about 99% sequence identity to the sequence according to SEQ ID NO: 1.

[0172] In further embodiments, the A2M molecule is a human A2M molecule, such as the sequence according to SEQ ID NO: 1, with the proviso that a. the bait region is modified as described above; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0173] In another embodiment, the A2M molecule has at least about 70% sequence identity to the sequence according to SEQ ID NO: 1, with the proviso that a. the bait region is modified as described above; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0174] In one embodiment, the A2M molecule has at least about 80% sequence identity to the sequence according to SEQ ID NO: 1, with the proviso that a. the bait region is modified as described above; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0175] In one embodiment, the A2M molecule has at least about 85% sequence identity to the sequence according to SEQ ID NO: 1, with the proviso thata. the bait region is modified as described above; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0176] In one embodiment, the A2M molecule has at least about 90% (e.g., at least 91%, at least 92%, at least 93% or at least 95%) sequence identity to the sequence according to SEQ ID NO: 1, with the proviso that a. the bait region is modified as described above; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0177] In one embodiment, the A2M molecule has at least about 95% (e.g., at least 96%, at least 97%, at least 98% or about 99%) sequence identity to the sequence according to SEQ ID NO: 1, with the proviso that a. the bait region is modified as described above; and / or b. one or more drugs (e.g., a therapeutic peptide, polypeptide or protein) are inserted into the RBD region, e.g., into loop 2, such as by removing one or more of the residues of loop 2 as described above.

[0178] In one embodiment, the one or more drugs is positioned between 1391 and 1405 in SEQ ID NO: 1. In another embodiment, the one or more drugs is positioned after position 1335 in SEQ ID NO: 1. In another embodiment, the one or more drugs is positioned before position 1474 in SEQ ID NO: 1. In another embodiment, the one or more drugs are positioned between 1391 and 1405 in SEQ ID NO: 1 or after position 1335 but before position 1474 in A2M. In another embodiment, the A2M molecule comprises one or more of the mutations K1393A, K1397A, T654C, and / or T661C.

[0179] K1393A and K1397A remove A2M's interactions with the receptors LRP1 and Grp78, respectively. LRP1 mediates clearance of cleaved A2M, Grp78 induces mitogenic signaling in cells when bound. Both of these receptor interactions are potentially problematic in a drug, as such it can be beneficial to remove these amino acids.

[0180] The T654C and T661C mutations introduce a disulfide which bridges the two disulfide-dimers of A2M, so that the entire A2M tetramer becomes stabilized by disulfide bonds. This prevents the splitting of A2M into its two halves, which can occur during physiological conditions such as inflammation (due to oxidative damage to A2M).

[0181] In an aspect of the invention, the invention relates to a proteinaceous fusion construct comprising alpha-2-macroglobulin (A2M), comprising a bait region with at least one protease cleavage site, said A2M being fused to a peptide drug positioned within residues 1392-1404, 1368-1379, or 1420-1426, of the Receptor Binding Domain (RBD) of A2M. In particular, in such an aspect it may occur that the peptide drug is inaccessible when the bait region in A2M has not been proteolytically cleaved; and the peptide drug, is accessible when the bait region in A2M has been proteolytically cleaved.

[0182] While the foregoing paragraphs describe the positioning of the one or more drugs within the RBD domain and the introduction of disulfide bridges in reference to A2M, a person of skill in the art of proteinaceous prodrug design will appreciate that other CP AMD proteins can take the place of A2M and can identify corresponding residues in these CPAMD proteins to implement the invention (e.g., using the residue numbers provided in Tables 1 and 2 as a guide).The drug

[0183] The proteinaceous prodrug construct can comprise one or more drugs or biopharmaceutical moieties (e.g., a therapeutic peptide, polypeptide or protein).

[0184] In some embodiments, the drug is able to increase or decrease the signal from a receptor upon binding to the receptor.

[0185] In one embodiment, the drug is selected from the group consisting of: an antigen-targeting moiety (e.g., an antibody or an antibody mimetics), a cytokine, the extracellular region of a cell surface receptor, the extracellular region of a cell surface ligand, and a receptor agonist.

[0186] In some embodiments, at least one biopharmaceutical moiety is an antigen-binding moiety (e.g., an antibody or an antigen-binding fragment of an antibody). In one embodiment, the antigen-targeting moiety is selected from the group consisting of: antibody, nanobody, diabody, and single-chain variable fragment. In some embodiments, the antigen-targeting moiety is a single-chain or single-domain antibody (e.g., a nanobody). In some embodiments, the antigen-targeting moiety is a single-chain variable fragment. In some embodiments, the antigen-binding moiety is capable of directing the proteinaceous prodrug construct to a specific tissue, a specific cell type, and / or a specific receptor.

[0187] In another embodiment, the drug is selected from the group consisting of: toxins, enzymes, and conjugates of a protein with small molecule drugs analogous to antibody drug conjugates (ADCs). For example, the protein may contain appropriate sites for small molecule conjugation, for example cysteine residues. In some embodiments, the toxin is selected from bacterially derived anthrax and diphtheria toxins. In particular embodiments, a drug such as a toxin, enzyme or protein conjugated to a smallmolecule drug analog is positioned inside the RBD to render inaccessible until cleavage of the protease cleavage site in the bait region.

[0188] In some embodiments, a proteinaceous prodrug construct is provided that comprises a complement 3- and pregnancy zone protein-like, alpha-2-macroglobulin domain-containing (CPAMD) protein, comprising:(a) a bait region comprising a first protease cleavage site;(b) a Receptor Binding Domain (RBD) comprising: a second protease cleavage site at the N-terminal end of the RBD;a toxin or enzyme positioned inside the RBD; and an antigen-targeting moiety (e.g., an antibody) fused to the C-terminal end of the RBD; wherein:(i) the CPAMD protein shields the toxin or enzyme;(ii) the antigen-targeting moiety (e.g., antibody) is capable of directing the proteinaceous prodrug construct to a specific tissue, a specific cell type, and / or a specific receptor;(iii) the RBD remains bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the second protease cleavage site; and(iv) the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD and thereof making the toxin or enzyme accessible.

[0189] In some embodiments, a proteinaceous prodrug construct is provided that comprises a complement 3- and pregnancy zone protein-like, alpha-2-macroglobulin domain-containing (CPAMD) protein, comprising:(a) a bait region comprising a first protease cleavage site;(b) a Receptor Binding Domain (RBD) comprising an enzyme positioned inside the RBD; and an antigen-targeting moiety (e.g., an antibody) fused to the C-terminal end of the RBD; wherein:(i) the CPAMD protein shields the toxin or enzyme;(ii) the antigen-targeting moiety (e.g., the antibody) is capable of directing the proteinaceous prodrug construct to a specific tissue, a specific cell type, and / or a specific receptor;(iii) the RBD is bound to the CPAMD protein, via non-covalent interactions; and(iv) the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD and thereof making the toxin or enzyme accessible.

[0190] In some embodiments, the antigen-targeting moiety specifically binds to an antigen as an antagonist (e.g., the antigen-targeting moiety is capable of inhibiting the binding of a ligand to its receptor). In some embodiments, the antigen-targeting moiety specifically binds to an antigen as an agonist (e.g., the antigen-targeting moiety is capable of inducing signaling by binding to a receptor).

[0191] In some embodiments, the antigen-targeting moiety specifically binds to an antigen (e.g., a tumor cell surface antigen) selected from the group consisting of BTLA, 0X40, LAG3, NRP1, VEGF, HER2, CEA, CD19, CD20, Amyloid beta, HER3, IGF-1R, MUC1, EpCAM, CD22, VEGFR-2, PSMA, GM-CSF, CXCR4, CD30, CD70, FGFR2, BCMA, CD44, ICAM-1, Notchl, MHC, CD28, IL-1R1, TCR, Notch3, FGFR3, TGF-p, TGFBR1, TGFBR2, CD 109, GITR, CD47, Alpha-synuclein, CD26, LRP1, CD52, IL-4Ra, VAP-1, EPO Receptor, Integrin av, TIM-3, Grp78, LIGHT, TLR2, TLR3, PAR- 2, NRP2, GLP-1 receptor, Hedgehog, and Syndecan 1.

[0192] When developing multi-specific prodrugs (e.g., proteinaceous prodrug construct that comprise at least two antigen-targeting moieties) such as BiTEs, the moiety that does not bind to the immune cell, e.g., the second drug, will typically bind to a tumor cell surface antigen. In some embodiments, the antigen-targeting moiety specifically binds to an antigen selected from the group consisting of NRP1, HER2, CEA, CD19, CD20, DLL3, GPRC5D, gplOO, HER3, IGF-1R, MUC1, EpCAM, CD22, VEGFR-2, PSMA, CD30, CD70, FGFR2, BCMA, CD44, ICAM-1, Notchl, MHC, IL-1R1, MCSP, CD66e, EphA2, Notch3, FGFR3, TGFBR1, TGFBR2, CD109, GITR, CD47, Alpha-synuclein, CD26, LRP1, CD52, IL-4Ra, VAP-1, EPO Receptor, Integrin av, TIM-3, Grp78, LIGHT, TLR2, TLR3, PAR- 2, NRP2, GLP-1 receptor, Hedgehog, Alphafetoprotein, CA-125, ETA, MAGE, and Syndecan 1. The moiety that binds to the immune cell, e.g., the first drug, will typically target a specific immune cell (e.g., a T cell). The moiety that binds to the immune cell is typically positioned inside the RBD to render it inaccessible until cleavage of the protease cleavage site in the bait region. An antigen-targeting moiety (e.g., an antibody) that specifically binds to a target antigen expressed on an immune cell (e.g., a T cell) may be directed against an immune stimulatory molecule, an immune costimulatory molecule, or an immune inhibitory molecule (such as a receptor or ligand). In some embodiments, the target antigen expressed on an immune cell (e.g., a T cell) is selected from the group consisting of CD3, CTLA-4, FcyRIIb, LAG-3, PD-1, PD-L1, PD-L2, TIM-3, TIGIT, CD28, CD27, 0X40, CD137, and ICOS. Accordingly, in some embodiments, a proteinaceous prodrug construct is provided that comprises a CPAMD protein, comprising:(a) a bait region comprising a first protease cleavage site; and(b) a Receptor Binding Domain (RBD) comprising: a second protease cleavage site at the N-terminal end of the RBD, a first antibody positioned inside the RBD, and a second antibody fused to the C-terminal end of the RBD; wherein:(i) the first antibody (e.g., a scFv or a nanobody) specifically binds to a target antigen expressed on an immune cell (e.g., an immune stimulatory receptor such as CD3) and the second antibody (e.g., a scFv or a nanobody) specifically binds to a tumor cell surface antigen;(ii) the CPAMD protein shields the first antibody and the second antibody is unshielded, i.e., accessible;(iii) the RBD remains bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the second protease cleavage site; and(iv) the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD and thereby making the first antibody unshielded, i.e., accessible.

[0193] In some embodiments, the first drug is an antigen-targeting moiety that specifically binds CD3 (e.g., an anti-CD3 antibody) such as CD3-targeting moiety comprised in known BiTEs. Commercially available BiTEs include Blinatumomab (a CD19-directed CD3 T-cell engager), Glofitamab (a CD20- directed CD3 T-cell engager), Mosunetuzumab (a CD20-directed CD3 T-cell engager), Solitomab (an EpCAM-directed CD3 T-cell engager), Talquetamab (a GPRC5D-directed CD3 T-cell engager), Tarlatamab (a DLL3 -directed CD3 T-cell engager), and Tebentafusp (a gplOO-directed CD3 T-cell engager). Without wishing to be bound by any particular theory, the inventors hypothesize that reformatting existing BiTEs as proteinaceous prodrug constructs in accordance with the methods disclosed herein will reduce common side effects such as cytokine release syndrome and / or toxicities such as immune effector cell-associated neurotoxicity syndrome (ICANS).

[0194] Other multi-specific prodrug constructs other than BiTEs that comprise at least two antigentargeting moieties are envisioned herein.

[0195] In one embodiment, an antigen targeting moiety is selected from the group consisting of anti- PD1, anti-PD-Ll, anti-EGFR, ant-CTLA4, anti-CD137, anti-CD3, and anti-TNFa.

[0196] In one embodiment, an antigen-targeting moiety is selected from the group consisting of Atezolizumab, EgAl, Ipilimumab, Nivolumab, KN035, Urelumab, Foralumab, Muromonab, and Adalimumab, or a therapeutically active scFv, fragment or variant thereof comprising one or more CDRs, all three heavy chain CDRs, all three light chain CDRs, all three heavy chain and all three light chain CDRs, a heavy chain variable region, and / or a light chain variable region of any of the foregoing antigen targeting moieties.

[0197] In some embodiments, an antigen-targeting moiety is selected from the group consisting of ANB032, rosnilimab, LY3361237, Encelimab, Cobolimab, Imsidolimab, Dostarlimab, or a therapeutically active scFv, fragment or variant thereof comprising one or more CDRs, all three heavy chain CDRs, all three light chain CDRs, all three heavy chain and all three light chain CDRs, a heavy chain variable region, and / or a light chain variable region of any of the foregoing antigen targeting moieties.

[0198] As outlined above, cytokines may be used as the drug in the present invention. Cytokines are described as a category of small proteins that induce cell signaling.

[0199] In one embodiment, said one or more drugs is a cytokine selected from the group consisting of chemokines, interferons, interleukins, lymphokines and tumor necrosis factors.

[0200] In another embodiment, said one or more drugs is a cytokine selected from the group consisting of IL1, ILlalpha, ILlbeta, IL2, IL3, IL4, IL6, IL7, IL8, IL9, IL10, IL11, IL12, IL13, IL14, IL15, IL16, IL17, 118, IL19, IL20, IL21, IL22, IL23, IL24, IL25, IL26, IL27, IL28, IL29, IL30, IL31, IL32, IL33, IL34, IL35 and IL36.

[0201] In a further embodiment, said one or more drugs is a cytokine selected from the group consisting of IL2, IFN-a, IL-15, IL-21, IL-10, IL-12, IL-17, GM-CSF, TGF-p, CSF-1, insulin, GLP-1, HGH, VEGF, PDGF, BMP, EPO, G-CSF, IL-11, IFN-y, and IFN-p.

[0202] In the preferred embodiment, said one or more drugs is IL2. IL 2 is tested in example 9.

[0203] In one embodiment, the antigen-targeting moiety is an amino acid sequence selected from the group consisting of SEQ ID NO: 27-43. In another embodiment, the antigen-targeting moiety has or comprises an amino acid sequence selected from the group consisting of SEQ ID NO: 27-43. In a further embodiment, the antigen-targeting moiety has an amino acid sequence having at least about 80 % sequence identity, such as at least about 85 %, 90 %, or even about 95 % sequence identity, to a sequence selected from the group consisting of SEQ ID NO: 27-43. If variance is introduced into the antigentargeting moiety, it is preferred that the CDR sequences are not modified.

[0204] In another embodiment, a nucleic acid sequence encoding an antigen-targeting moiety is selected from the group consisting of: SEQ ID NO: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, and 26 or a fragment or variant thereof having at least about 90% sequence identity to any of SEQ ID NO: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, and 26: particularly about 95% identity to SEQ ID NO: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, and 26. In another embodiment, the amino acid sequence is encoded by a nucleic acid sequence selected from the group consisting of: SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26 or a fragment or variant thereof having at least about 90% sequence identity to any of SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26: particularly about 95% identity to SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26.

[0205] In one embodiment, the one or more drugs (e.g., the first and second antibodies) has / have a size of at the most 100 kDa, such as at the most 85 kDa, such as at the most 75 kDa, such as at the most 65 kDa, such as at the most 55 kDa, such as at the most 50 kDa, such as at the most 40 kDa, such as at the most 30 kDa, such as at least 10 kDa.

[0206] In one embodiment, the one or more drugs comprise at most 900 amino acids, such as at most 770 amino acids, such as at most 680 amino acids, such as at most 590 amino acids, such as at most 500 amino acids, such as at most 450 amino acids, such as at most 360 amino acids, such as the most 270 amino acids, such as at least 90 amino acids.

[0207] In a further embodiment, the proteinaceous prodrug construct according to invention comprises 1-5 drugs, such as 1-4, such as 1-3, such as 1-2. In a specific embodiment, the proteinaceous prodrug construct according to invention comprises 1 drug.The bait region

[0208] As previously described, the proteinaceous prodrug construct’s conformational change (figure 1) is initiated when a protease cleaves within the exposed and highly susceptible bait region.

[0209] In some embodiment, a proteinaceous prodrug construct in accordance with the invention comprises a CPAMD protein (e.g., A2M) with a modified bait region. In some embodiments, the bait region is modified to change the selection of proteases that are able to cleave it and trigger theconformational change of the CPAMD protein (e.g., A2M). For example, the bait region may be modified to be cleaved by a particular protease or class of proteases (e.g., MMPs such as MMP2).

[0210] To engineer specificity, it may be advantageous to first engineer a bait region that cannot be cleaved by proteases. A bait region that cannot be cleaved by proteases is referred to herein as a "tabula rasa bait region”. For example, to prevent cleavage by proteases, a tabula rasa bait region may comprise an engineered amino acid sequence that is flexible and / or hydrophilic. In some embodiments, the engineered amino acid sequence comprises a sequence of glycine, serine, alanine, threonine, and / or proline residues. In some embodiments, the engineered amino acid sequence replaces all or a portion of a wildtype bait region. In some embodiments, the engineered amino acid sequence is about 15-51 amino acids, such as about 30-40, such as about 31-39, such as about 32-35. In a particular embodiment, the length of the engineered amino acid sequence is about 32-33 amino acids. In some embodiments, the engineered amino acid sequence replaces all of the wildtype bait region and has a length equivalent to the wildtype bait region.

[0211] For instance, to prevent cleavage by proteases, a tabula rasa bait region can be composed of a series of amino acid repeats. The series of amino acid repeats may replace part or all of the native bait region. For example, a tabula rasa bait region can comprise a series of amino acid repeats. An example of a series of three amino acid repeats are Gly-Gly-Ser, Gly-Gly-Gly, Gly-Ser-Gly, Gly-Ser-Ser, Ser- Gly-Gly, Ser-Gly-Ser, Ser-Ser-Gly, Ser-Ser-Ser. Each series of three amino acids are either repeated or combined with each other. For instance, a tabula rasa bait region can comprise one or more amino acid repeats, wherein the repeats are selected from the list consisting of Gly-Gly-Ser, Gly-Gly-Gly, Gly-Ser- Gly, Gly-Ser-Ser, Ser-Gly-Gly, Ser-Gly-Ser, Ser-Ser-Gly and Ser-Ser-Ser. Alternatively, a tabula rasa bait region can comprise of one or more amino acid repeats, wherein the repeats are amino acid triplets comprised by Ser, Gly, and Ala residues. In some instances, a tabula rasa bait region can be comprised of one or more amino acid repeats, wherein the repeats are selected from the list consisting of Gly-Gly- Ser, Gly-Gly-Gly, Gly-Ser-Gly, Gly-Ser-Ser, Ser-Gly-Gly, Ser-Gly-Ser, Ser-Ser-Gly, Ser-Ser-Ser and Ala. A tabula rasa bait region can be comprised of one or more amino acid repeats, wherein the repeats are selected from the list consisting of Gly-Gly-Ser, Gly-Gly-Gly, Gly-Gly-Ala, Gly-Ser-Gly, Gly-Ser- Ser, Gly-Ser-Ala, Gly-Ala-Ser, Gly-Ala-Gly, Gly-Ala-Ala, Ser-Gly-Gly, Ser-Gly-Ser, Ser-Gly-Ala, Ser-Ser-Gly, Ser-Ser-Ser, Ser-Ser-Ala, Ser-Ala-Gly, Ser-Ala-Ser, Ser-Ala-Ala, Ala-Gly-Ser, Ala-Gly- Gly, Ala-Gly-Ala, Ala-Ser-Gly, Ala-Ser-Ser, Ala-Ser-Ala, Ala-Ala-Ser, Ala-Ala-Gly and Ala-Ala-Ala. For example, a tabula rasa bait region consisting of 13 Gly-Gly-Ser repeats can be seen in SEQ ID NO: 124.

[0212] In one embodiment, the bait region comprises 5, such as 7, such as 9, such as 11, such as 13, such as 15, such as 17 repeats. In one particular embodiment, the bait region comprises 13 repeats. In another embodiment, the bait region comprises about 5-17 repeats, such as about 7-15, such as about 9- 13.

[0213] The total length of the bait region can vary between 15 and 51 amino acids. The length of a tabula rasa bait region can be about 15-51 amino acids, such as about 30-40, such as about 31-39, suchas about 32-35. In a particular embodiment, the length of a tabula rasa bait region is about 32-33 amino acids.

[0214] At least one protease cleavage site can be introduced into the tabula rasa bait region. By using such a modified bait region, it is possible to control which proteases are able to cleave and thereby introduce the conformational change to the proteinaceous prodrug construct.Cleavage sites

[0215] In some aspects, the present disclosure provides prodrugs comprising at least two protease cleavage sites. A first protease cleavage site is comprised in the bait region, and is responsible for the unshielding of a shielded biopharmaceutical moiety. A second protease cleavage site can be introduced at the N-terminal end of the CPAMD protein’s RBD domain. The second protease cleavage site enables the release of RBD from the CPAMD protein. When the first protease cleavage site is not cleaved, the RBD domain remains bound to the CPAMD protein via non-covalent interactions. The second protease cleavage site is thus optional in some aspects of the disclosure.

[0216] The skilled person will be able to select the protease cleavage sites to design prodrugs with different purposes. In some embodiments, the first protease cleavage site and the second protease cleavage site are specific towards the same proteases, i.e., the proteases described herein as part of the bait region may be introduced as the second protease cleavage site. For example, both the first protease cleavage site comprised in the bait region and the second protease cleavage site at the N-terminal end of the RBD protein may be capable of being cleaved in vivo by the same protease(s) (e.g., one or more endogenous protease(s) present in a tissue affected by a disease such as a tissue comprising a tumor or a site of inflammation).

[0217] In other embodiments, the second protease cleavage site is different from the first protease cleavage site, such as to make sure that RBD domain comprising the first and second biopharmaceutical moieties is not released until cleavage of the first protease cleavage site (e.g., in a tumor or at a site of inflammation). For example, when a BiTE is developed using a CPAMD protein as described herein, it is desirable to provide a proteinaceous prodrug comprising a second protease cleavage site that is processed before introduction into the circulation of a subject. Premature crosslinking of T cells can be avoided because the first antibody positioned inside the RBD is made accessible only upon cleavage of the first protease cleavage site. In addition, crosslinking can occur only upon release of the RBD comprising both the first and second antibodies from the CPAMD protein. Accordingly, in some embodiments, the first protease cleavage site is capable of being cleaved by an endogenous protease (e.g., an endogenous protease present at the site of a disease such as cancer or an inflammatory disease), whereas the second protease is capable of being cleaved a protease that is not normally present extracellularly or in circulation.

[0218] In particular, preferred protease cleavage sites for the second protease cleavage site may be furin, TEV protease, Enterokinase, or Thrombin, for example, to facilitate an in vitro cleavage of the proteinaceous prodrug construct, before introduction into a living subject. In some embodiments, thesecond protease cleavage site is specific towards furin, TEV protease, Enterokinase, or Thrombin. In a preferred embodiment, the second protease cleavage site is specific towards furin. In another preferred embodiment, the second protease cleavage site is specific towards TEV protease. The specificity towards furin may be achieved by the incorporation of an amino acid sequence according to SEQ ID NO: 225. The specificity towards TEV protease can be achieved by the incorporation of an amino acid sequence according to SEQ ID NO: 226. As provided in the section further down, the second protease cleavage sites may additionally be surrounded by linkers.

[0219] In some embodiments, the first protease cleavage site is specific for an aspartic-, cysteine-, glutamic-, asparagine-, serine-, threonine-, or metalloprotease.

[0220] The second protease cleavage site is inserted upstream (i.e., at the N-terminal end) of the RBD domain. Alternatively, the first part of the RBD domain is replaced by the second protease cleavage site. Upon cleavage of the second protease cleavage site, the RBD domain and the at least one biopharmaceutical moieties are non-covalently bound to the CPAMD protein, until the first protease cleavage site is cleaved and a conformational change occurs resulting in a release of the RBD domain and the at least one biopharmaceutical moiety.

[0221] In some embodiments of the present invention, the second protease cleavage site is inserted in accordance with the start of the RBD domain as indicated in Table 2. The skilled person will appreciate that this position may not strictly need to be adhered to, and insertion + / - 20 amino acids of the N- terminus of the RBD may be acceptable. Thus, in some embodiments of the present invention, the second protease cleavage site is inserted in any position chosen in accordance with the positions as indicated in Table 4. For instance, the insertion site for CPAMD5 (a.k.a. A2M) may be chosen as 1335 according to Table 2; a position of 20 amino acids N-terminal thereto, such as at position 1315 according to Table 4; or a position of 20 amino acids C-terminal thereto, such as at position 1355 according to Table 4. The skilled person will find that + / - 20 amino acids may also mean any position in the approximately 40 amino acid region, such as for instance the insertion site for A2M may be chosen as 1325 or 1345. As seen from Examples 12 and 13, a second protease cleavage site of 11 amino acids can efficiently be inserted directly into the sequence of A2M. In some embodiments of the present invention, the second protease cleavage site is replacing a suitable number of amino acids in accordance with the positions as indicated in Table 4.Table 4. Positions of N-terminal insertion sites of the second protease cleavage site.

[0222] In some embodiments, the second protease cleavage site introduced at the N-terminal end of CPAMD protein’s RBD domain is introduced from residue position 1334 to 1340 of A2M.

[0223] For the proteinaceous prodrug construct to be effective as a medicament, and to control the activity of the proteinaceous prodrug construct, individual protease cleavage sites can be introduced into a tabula rasa bait region. Thus, the skilled person can control which proteases are able to cleave and thereby introduce the conformational change to the proteinaceous prodrug construct.

[0224] The invention is not limited to introducing a single protease cleavage site in the bait region. In some embodiments, the bait region can have a number of cleavage sites, which are cleaved by different proteases. Accordingly, in one embodiment said, bait region comprises one or more protease cleavage sites (e.g., two, three or four protease cleavage sites). In another embodiment, said bait region comprises only one protease cleavage site.

[0225] In some embodiments, the protease cleavage site for use in the proteinaceous prodrugs (and specifically in the bait region) are selected from the group consisting of activated protein C, ADAM 10, ADAM12, ADAM15, ADAM17 / TACE, ADAM9, ADAMDEC1, ADAMTS1, ADAMTS4, ADAMTS5, BACE, BMP-1, Caspase 1, Caspase 10, Caspase 14, Caspase 2, Caspase 3, Caspase 4, Caspase 5, Caspase 6, Caspase 7, Caspase 8, Caspase 9, Cathepsin A, Cathepsin B, Cathepsin C, Cathepsin D, Cathepsin E, Cathepsin G, Cathepsin K, Cathepsin L, Cathepsin S, Cathepsin V / L2, Cathepsin X / Z / P, Chymase, Cruzipain, DESCI, DPP-4, Elastase, FAP, Granzyme B, Guanidinobenzoatase, Hepsin, HtrAl, Neutrophil Elastase, KLK10, KLK11, KLK13, KLK14, KLK4, KLK5, KLK6, KLK7, KLK8, Legumain, Marapsin, Matriptase-2, Meprin, MMP1, MMP8, MMP10, MMP11, MMP12, MMP13, MMP14, MMP15, MMP16, MMP17, MMP19, MMP2, MMP20, MMP23, MMP24, MMP26, MMP27, MMP3, MMP7, MMP8, MMP9, MT-SPl / Matriptase, Neprilysin, NS3 / 4A,Otubain-2, PACE4, Plasmin, PSA, PSMA, Renin, Thrombin, TMPRSS2, TMPRSS3, TMPRSS4, tPA, Tryptase, uPA, ADAM8, FVIIa, FIXa, Furin, Fxa, FXIa, FXIIa, and TAFI.

[0226] In some embodiments, the bait region contains a single cleavable site selected from the group of SEQ ID NO: 96-123. In some embodiments, the bait region contains only one single protease cleavage site which can be cleaved by a matrix metalloprotease (MMP). In a particular embodiment, the bait region comprises one single protease cleavage site which can be cleaved by a protease selected from the group consisting of MMP2, MMP9, MMP14, MMP1, MMP3, MMP13, MMP17, MMP11, MMP8, MMP10, and MMP19.

[0227] In some embodiments, the bait region comprises two cleavage sites. For example, the bait region may comprise exactly two cleavable sites, one of which is cleavable by the group of proteases consisting of MMP2, MMP9, MMP 14, MMP1, MMP3, MMP 13, MMP 17, MMP11, MMP8, MMP 10, and MMP 19, and the other of which is cleavable by the group of proteases consisting of activated protein C, ADAM 10, ADAM 12, ADAM 15, ADAM17 / TACE, ADAM9, ADAMDEC1, ADAMTS1, ADAMTS4, ADAMTS5, BACE, BMP-1, Caspase 1, Caspase 10, Caspase 14, Caspase 2, Caspase 3, Caspase 4, Caspase 5, Caspase 6, Caspase 7, Caspase 8, Caspase 9, Cathepsin A, Cathepsin B, Cathepsin C, Cathepsin D, Cathepsin E, Cathepsin G, Cathepsin K, Cathepsin L, Cathepsin S, Cathepsin V / L2, Cathepsin X / Z / P, Chymase, Cruzipain, DESCI, DPP-4, Elastase, FAP, Granzyme B, Guanidinobenzoatase, Hepsin, HtrAl, Neutrophil Elastase, KLK10, KLK11, KLK13, KLK14, KLK4, KLK5, KLK6, KLK7, KLK8, Lactoferrin, Legumain, Marapsin, Matriptase-2, Meprin, MT- SPI / Matriptase, Neprilysin, NS3 / 4A, Otubain-2, PACE4, Plasmin, PSA, PSMA, Renin, Thrombin, TMPRSS2, TMPRSS3, TMPRSS4, tPA, Tryptase, uPA, ADAM8, FVIIa, FIXa, Furin, Fxa, FXIa, FXIIa and TAFI. In another embodiment, the bait region comprises two cleavable sites selected from the group of SEQ ID NO: 96-123.

[0228] In a further embodiment, the bait region is free from protease cleavage sites recognized by human proteases except MMPs. In some embodiments, the bait region contains one or more (e.g., at least two or three) protease cleavage sites which can be cleaved by one or more (e.g., at least two or three) MMPs. In yet a further embodiment, the bait region is free from protease cleavage sites recognized by human proteases except for a single cleavage site.

[0229] As seen from the examples, the bait region can be highly modified, and the skilled person will be able to select any suitable cleavage site into the bait region, according to the needed specificity. Accordingly, in particular embodiments, a proteinaceous prodrug construct in accordance with the invention comprises a CPAMD protein (e.g., A2M) comprising a modified bait region that can be selectively cleaved by one or more proteases.

[0230] A protease site is “selectively cleavable” when cleavage occurs only or predominantly in the presence of one particular protease. A modified bait region may be engineered to comprise one or more (e.g., at least two or three) cleavage sites, wherein each of the cleavage sites is “selectively cleavable” by a different protease. For example, a modified bait region may be engineered to comprise one or two or three unique recognition sites, each specific for a different protease.

[0231] Exemplary MMP cleavage sites include the A21A, B74, C9 and SI. In a specific embodiment, the bait region comprises one or more (e.g., at least two or three) of the A21A, B74, C9 and / or the SI cleavage sites. Exemplary modified bait regions comprising said cleavage sites can be seen in SEQ ID NO: 126-133. In another specific embodiment, the bait region comprises a lysine, such as in SEQ ID NO: 125.

[0232] In some embodiments, the modified bait region comprises an engineered amino acid sequence that is flexible and / or hydrophilic. In some embodiments, the engineered amino acid sequence comprises a sequence of glycine, serine, alanine, threonine, and / or proline residues. In some embodiments, the engineered amino acid sequence comprises a combination of glycine, serine, and / or alanine residues. In some embodiments, the engineered amino acid sequence replaces a wildtype bait region and has a length equivalent to the wildtype bait region. In one embodiment, the wildtype bait region is replaced by a combination of glycine, serine, and / or alanine residues with an equivalent length to the wildtype bait region.

[0233] Exemplary sequences, where cleavage sites have been inserted into a tabula rasa region can be seen in any of the sequences identified by SEQ ID NO: 125-133. In another embodiment, only a part of the wildtype bait region has been replaced by a tabula rasa region as described herein, such as in SEQ ID NO: 130 where the C-terminal quarter of the wildtype bait region is retained.

[0234] In another embodiment, one or more cleavage sites in the bait region have been replaced by a combination of glycine, serine, and / or alanine residues. In one embodiment, the bait region comprises one or more repeats, such as at least 5, such as at least 6, such as at least 7, such as at least 8. In a particular embodiment, the length of the tabula rasa bait region is at least about 10 repeats.

[0235] In one embodiment, the bait region has a size of about 8 kDa, such as at the most about 5 kDa, such as at the most about 4 kDa, such as at the most about 3 kDa, such as at the most about 2 kDa. In a particular embodiment, the bait region has a size of at the most about 2.5 kDa.

[0236] In one embodiment, the length of the bait region is about 15 to 51 amino acids. In one embodiment, the length of the bait region is about 30-40 amino acids, e.g., about 31-39 amino acids or 32-35 amino acids. In a particular embodiment, the total length of the bait region is about 32-33 amino acids.

[0237] In one embodiment, the bait region comprises an engineered amino acid sequence that is entirely flexible and / or hydrophilic, e.g., a random sequence of glycine, serine, alanine, threonine, and / or proline residues , and optionally one or more protease cleavage sites (e.g., MMP cleavage sites), such that the total length of the bait region, including the repeats and cleavage site(s) is about 15 to 51 amino acids, e.g., about 32-33 amino acids.

[0238] In one embodiment, when a protease cleaves the “bait region”, the protease is trapped inside the proteinaceous prodrug construct.RBD domain

[0239] As explained above, the prodrug is generated by bringing the drug or biopharmaceutical moiety that must be shielded (e.g., a therapeutic peptide, polypeptide or protein) into contact with the RBD domain, such that the folding of the RBD domain, shields the drug, such that the drug is inaccessible. A therapeutic protein may be brought into contact with the RBD domain by inserting it into the RBD domain or by replacing a part of the RBD domain with the therapeutic protein.

[0240] Accordingly, in a typical embodiment of a proteinaceous prodrug construct of the invention, the drug (e.g., a therapeutic peptide, polypeptide or protein) is positioned inside the RBD in such a manner that the CPAMD protein (e.g., A2M) is capable of altering conformation upon proteolytic cleavage of a protease cleavage site comprised within the bait region, thereby making the drug accessible.

[0241] Given the size of the RBD domain, numerous suitable sites exist for insertion into the RBD domain. As visualized by Figure 13, the RBD domain is largely comprised of beta sheets, and as shown in the examples of the present invention, the loops in between individual beta strands are suitable for insertion of the drug. For example, in the native human A2M protein, loop 1 is formed by amino acid residues 1368-1379, loop 2 by amino acid residues 1392-1404, loop 3 by amino acid residues 1420- 1426, and loop 4 by amino acid residues 1450-1457.

[0242] In one embodiment, the drug is positioned in the RBD domain of A2M within loop 2 (at a position between residue 1391 and 1405, e.g., between 1392 and 1404, of native human A2M). In one embodiment, the drug is positioned in the RBD domain of A2M by replacing one or more amino acids corresponding to the region formed residues 1391 to 1405, or residues 1392 to 1404, ofthe native human protein. In one embodiment, the drug is positioned in the RBD domain of A2M between amino acids corresponding to residues 1391 to 1405 (e.g., residues 1392 to 1404) of the native human protein. In another embodiment, one or more of the amino acids corresponding to residues 1391, 1392, 1393, 1394, 1395, 1396, 1397, 1398, 1399, 1400, 1401, 1402, 1403, 1404, and / or 1405 ofthe native human protein is / are replaced by the drug. In another embodiment, the drug is positioned after one or more of the amino acids corresponding to residues 1391, 1392, 1393, 1394, 1395, 1396, 1397, 1398, 1399, 1400, 1401, 1402, 1403, or 1404 ofthe native human protein.

[0243] In one embodiment, the drug is positioned in the RBD domain of A2M within loop 1 (at a position between residue 1368-1379 of native human A2M). In one embodiment, one or more of the amino acids corresponding to residues 1368, 1369, 1370, 1371, 1372, 1373, 1374, 1375, 1376, 1377, and / or 1378 of native human A2M is / are replaced by the drug. In another embodiment, the drug is positioned after one or more of the amino acids corresponding to residues 1368, 1369, 1370, 1371, 1372, 1373, 1374, 1375, 1376, 1377, or 1378 of native human A2M.

[0244] In one embodiment, the drug is positioned in the RBD domain of A2M in loop 3 (at a position between residue 1420-1426 of native human A2M). In another embodiment, one or more of the amino acids corresponding to residues 1420, 1421, 1422, 1423, and / or 1424 of native human A2M is / arereplaced by the drug. In another embodiment, the drug is positioned after one or more of amino acids corresponding to the residues 1420, 1421, 1422, 1423, or 1424 of native human A2M.

[0245] In another embodiment, the drug is positioned in the vicinity of the RBD domain of A2M. In one embodiment, the drug is tethered to the C-terminus of A2M’s RBD domain and brought into close proximity of residues 1391-1405 of the RBD domain through specific interactions, such as coiled-coil interactions between alpha helices. In some embodiments, the drug that is tethered to the C-terminus of A2M’s RBD domain is tethered to the amino acid corresponding to residue 1474 of human A2M.

[0246] While the positioning of the one or more drugs within the RBD domain is described in the foregoing paragraphs is reference to A2M, a person of skill in the art of proteinaceous prodrug design will appreciate that other CPAMD proteins can take the place of A2M and can identify corresponding residues in these CPAMD proteins to implement the invention (e.g., using the residue numbers provided in Table 1 as a guide).

[0247] The inventors have found that proteinaceous prodrug constructs in which a drug (e.g., a therapeutic peptide, polypeptide or protein) is positioned within loop 2 or 3 (or loop 4) of the RBD domain of a CPAMD protein (e.g., by replacing one or more residues, or by direct insertion) can be expressed successfully at high levels (see, e.g., the proteinaceous fusion constructs referred herein as “ciRBD” and “miRBD”). For example, insertion of the drug between the amino acids corresponding to residues 1402 and 1403 of native human A2M has been found to be particularly advantageous. Replacing the amino acids corresponding to residues 1393-1395 of native human A2M with the drug may be similarly advantageous.Linkers

[0248] Linkers can be used to insert drugs into the proteinaceous prodrug construct. Any suitable linker may be used. In some embodiments, the linker used to insert drugs into the proteinaceous prodrug construct is a GS linker. In some embodiments, the linker is a (GGGGS)n (SEQ ID NO: 223) or a (GGS)n. In some embodiments, n = 1, 2, 3, 4, 5, or 6.

[0249] In some embodiments, the second protease cleavage site is positioned between one or more linkers, such as two linkers. The linkers may have any suitable length. The linkers may for instance comprise a selection of small non-polar (e.g. glycine, alanine) or polar (e.g. serine or lysine) amino acids. In some embodiments, the one or more linkers is a GS linker. In some embodiments, at least one of the two linkers is a GS linker. In some embodiments, the two linkers are GS linkers.

[0250] The linker(s) may have a length of 1-5 amino acids, e.g., 2, 3, 4 amino acids. Alternatively, the linker(s) may have a length of 5-30 amino acid, e.g., 5-25 amino acids length. For example, the linker(s) may be (a) GS linker(s) having a length of 5-30 amino acids or 5-25 amino acids.

[0251] Exemplary second protease cleavage sites applied in Examples 12 and 13 are for instance inserted using a combination of the above-described linkers. For instance, the furin cleavage site RRRR (SEQ ID NO: 225) was inserted by including an N-terminal linker KASGSS (SEQ ID NO:248) and a single serine residue C-terminal to the furin cleavage site. Thus, one exemplary sequence of the secondprotease cleavage site may be or may comprise the amino acid sequence KASGSSRRRRS (SEQ ID NO: 249).

[0252] The TEV protease cleavage site was inserted by including a single serine residue N-terminal to the TEV protease cleavage site and a GS linker with the amino acid sequence SSGS (SEQ ID NO: 250) C-terminal to the TEV protease cleavage site. Thus, one exemplary sequence of the second protease cleavage site may be or may comprise the amino acid sequence SENLYFQSSGS (SEQ ID NO:251). In some embodiments, the second protease cleavage site is or comprises the amino acid sequence SENLYFQSSGS (SEQ ID NO: 253). In some embodiments, the second protease cleavage site is or comprises the amino acid sequence SGGSENLYFQS (SEQ ID NO: 254). In some embodiments, the second protease cleavage site is or comprises the amino acid sequence SGGGSENLYFQSSGS (SEQ ID NO: 255). In some embodiments, the second protease cleavage site is or comprises the amino acid sequence SGGSGGSGGSGENLYFQSSGS (SEQ ID NO: 256). In some embodiments, the second protease cleavage site is or comprises the amino acid sequence SGGSGGSGGSGENLYFQSSGGSGGS (SEQ ID NO: 257).

[0253] The skilled person will appreciate that the specificity derives from the protease cleavage site sequence, and thus the composition of the linkers surrounding the protease cleavage site may easily be exchanged.Exemplary proteinaceous fusion or prodrug constructs

[0254] As exemplified in the following sequences, the prodrugs of the invention can take many forms. In some embodiments, the proteinaceous prodrug is encoded by a single nucleic acid as a continuous peptide chain. In some embodiments, the RBD comprising at least one biopharmaceutical moiety, which is released from the CPAMD protein, consist of a single protein chain upon proteolytic cleavage of the first protease cleavage site.

[0255] Specific embodiments can for example be some of the proteins as shown in the examples such as a bi-specific prodrug, comprising two antibodies and a TEV protease site, as illustrated in any of SEQ ID NO: 231 - A2M_tevRBD+2xAb, or SEQ ID NO: 235 - A2M_tevRBD+2xAb_2. In another embodiment, the proteinaceous fusion construct has at least about 80% sequence identity, such as at least about 85% sequence identity, about 90% sequence identity, or even about 95% sequence identity, to a sequence selected from the group consisting of: SEQ ID NO: 231, and SEQ ID NO: 235.

[0256] Other specific embodiments of the invention can for example be intermediary proteins, where only the second protease cleavage site has been inserted, such as any of SEQ ID NO: 227 - A2M_furinRBD, or SEQ ID NO: 229 - A2M_tevRBD. As will be obvious to the skilled person, these intermediate products, can be combined with any of the prodrugs as shown in example 1-11.

[0257] In one embodiment, the proteinaceous fusion construct is comprised of an amino acid sequence selected from the group consisting of: SEQ ID NO: 5, SEQ ID NO: 7, SEQ ID NO: 9, SEQ ID NO: 11, SEQ ID NO: 13, SEQ ID NO: 15, and SEQ ID NO: 17, SEQ ID NO: 19, SEQ ID NO: 21, SEQ ID NO: 23, and SEQ ID NO: 25.

[0258] In another embodiment, the proteinaceous fusion construct has at least about 80% sequence identity, such as at least about 85% sequence identity, about 90% sequence identity, or even about 95% sequence identity, to a sequence selected from the group consisting of: SEQ ID NO: 5, SEQ ID NO: 7, SEQ ID NO: 9, SEQ ID NO: 11, SEQ ID NO: 13, SEQ ID NO: 15, and SEQ ID NO: 17, SEQ ID NO: 19, SEQ ID NO: 21, SEQ ID NO: 23, and SEQ ID NO: 25.

[0259] In one embodiment, the amino acid sequence is encoded by a nucleic acid sequence selected from the group consisting of: SEQ ID NO: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, and 26: or a fragment or variant thereof having at least about 90% sequence identity to anyone of SEQ ID NO: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, and 26, particularly about 95% identity to anyone of SEQ ID NO: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, and 26. In one embodiment, the amino acid sequence is encoded by a nucleic acid sequence selected from the group consisting of: SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26: or a fragment or variant thereof having at least SEQ ID NO: about 90% sequence identity to anyone of SEQ ID NO: 6, SEQ ID NO:8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26, particularly about 95% identity to anyone of SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26.Exemplary nucleic acids

[0260] In one aspect, the present invention relates to a nucleic acid encoding a proteinaceous fusion construct according to the invention. Such specific examples can for example be nucleic acids encoding a bi-specific prodrug, comprising 2 antibodies and a TEV protease site, as presented in any of SEQ ID NO: 232 - A2M_tevRBD+2xAb, or SEQ ID NO: 236 - A2M_tevRBD+2xAb_2.

[0261] In the context of the present invention, a particular aspect relates to nucleic acids and vectors comprising components relevant as intermediates when developing prodrugs. Such intermediate is for instance a prodrug, or the nucleic acid encoding the prodrug, comprising a second protease cleavage site, without a biopharmaceutical moiety inserted. Thus, in another aspect, the present invention relates to a nucleic acid or a vector, such as a plasmid, encoding a multimeric, such as tetrameric CP AMD protein or a fragment or variant thereof, such as A2M protein, the CPAMD protein or a fragment or variant thereof comprising a second protease cleavage site at the N-terminal end of the CPAMD protein’s RBD domain.

[0262] Such intermediates may be as otherwise described herein, and thus in some embodiments, the CPAMD protein or a fragment or variant thereof comprises a bait region said bait region comprising at least one first protease cleavage site. In some embodiments, the CPAMD protein or a fragment thereof comprises a Receptor Binding Domain (RBD).

[0263] Specific examples of possible intermediate products are for example provided in anyone of SEQ ID NO: 228, or SEQ ID NO: 230.

[0264] In another embodiment, the nucleic acid sequence is selected from the group consisting of: SEQ ID NO: 232, SEQ ID NO: 236, SEQ ID NO: 228, and SEQ ID NO: 230: or a fragment or variant thereof having at least about 90% sequence identity to anyone of SEQ ID NO: 232, SEQ ID NO: 236, SEQ ID NO: 228, and SEQ ID NO: 230, particularly about 95% identity to anyone of SEQ ID NO: 232, SEQ ID NO: 236, SEQ ID NO: 228, and SEQ ID NO: 230. In another embodiment, the nucleic acid sequence is selected from the group consisting of: SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26: or a fragment or variant thereof having at least SEQ ID NO: about 90% sequence identity to anyone of SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26, particularly about 95% identity to anyone of SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26. In another embodiment, the nucleic acid is a degenerate sequence of any of the nucleic acid sequences, wherein the nucleic acid sequence may show a higher variance, than at the protein level.

[0265] In one embodiment, the nucleic acid according to the invention encodes a proteinaceous fusion construct according to anyone of SEQ ID NO: 5, SEQ ID NO: 7, SEQ ID NO: 9, SEQ ID NO: 11, SEQ ID NO: 13, SEQ ID NO: 15, and SEQ ID NO: 17, SEQ ID NO: 19, SEQ ID NO: 21, SEQ ID NO: 23, or SEQ ID NO: 25.

[0266] In another embodiment, the nucleic acid sequence is selected from the group consisting of: SEQ ID NO: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, and 26: or a fragment or variant thereof having at least about 90% sequence identity to anyone of SEQ ID NO: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, and 26, particularly about 95% identity to anyone of SEQ ID NO: 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, and 26. In another embodiment, the nucleic acid sequence is selected from the group consisting of: SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26: or a fragment or variant thereof having at least SEQ ID NO: about 90% sequence identity to anyone of SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26, particularly about 95% identity to anyone of SEQ ID NO: 6, SEQ ID NO: 8, SEQ ID NO: 10, SEQ ID NO: 12, SEQ ID NO: 14, SEQ ID NO: 16, SEQ ID NO: 18, SEQ ID NO: 20, SEQ ID NO: 22, SEQ ID NO: 24, and SEQ ID NO: 26. In another embodiment, the nucleic acid is a degenerate sequence of any of the nucleic acid sequences, wherein the nucleic acid sequence may show a higher variance, than at the protein level.Vectors

[0267] For a proteinaceous fusion construct to be expressed, the nucleic acid according to the invention is or can be inserted into an expression vector, which is usually a plasmid or a virus designed to control gene expression in a cell. The vector is engineered to contain regulatory sequences that act as enhancersor promotor for an efficient expression of the desired coding sequence carried by the vector. In a nonlimiting example, the use of a naked circular plasmid with the key features necessary for expression, including promotor, coding sequence of interest and polyadenylation signal is provided.

[0268] Further, to enable an easy production, which might take place using E. coli bacteria, the plasmid comprises a selection marker. This enables production in a bacterium with or without using conventional bacterial resistance selection.

[0269] In another embodiment, the present invention relates to a vector comprising the nucleic acid according to the invention.

[0270] In one embodiment, the nucleic acid encoding the proteinaceous fusion construct is operatively linked to a promotor and optionally, additionally regulatory sequences that regulate expression of said nucleic acid.

[0271] In one embodiment, the vector is a eukaryotic expression vector, particularly a mammalian, e.g., a human expression vector. In one embodiment, the vector is selected from the group consisting of plasmids, cosmids, phages, bacterial artificial chromosomes (BAC), phagemids, and Pl -derived artificial chromosomes.

[0272] In one embodiment, the vector is a plasmid. In one embodiment, said plasmid is selected from the group consisting of TA cloning vectors, Gateway cloning vectors, restriction cloning vectors, Topo cloning vectors, pET vector system, and pBAD vector systems.Host cells

[0273] The vector according to the invention is or can be inserted into a host cells for expression of proteinaceous fusion construct according to the invention.

[0274] In one aspect, the present invention relates to a host cell comprising a vector according to the invention.

[0275] The cells can be either prokaryotic, like bacteria, or eukaryotic cells.

[0276] In one embodiment, the host cell is selected from the group consisting of: bacteria and eukaryotes; typically the host cell is a eukaryote.

[0277] In another embodiment, the host cell is yeast.

[0278] In a typical embodiment, the host cell is a mammalian cell, e.g., a CHO (Chinese Hamster) cell.

[0279] In one embodiment, said host cell is human.

[0280] In another embodiment, said host cell is the HEK293 cell line or descends from the HEK293 cell line.Compositions

[0281] A further aspect of the present disclosure relates to a composition, comprising a proteinaceous prodrug construct as described herein. Also provided are compositions that comprise a nucleic acid, a vector or a host cell as described herein.

[0282] In one embodiment, the composition comprises a pharmaceutically acceptable carrier. Such a composition can also be referred to as a pharmaceutical composition.Therapeutic uses

[0283] The proteinaceous fusion construct according to the invention can be used in the treatment of disease. In further aspects, the composition, a nucleic acid, a vector or a host cell as described herein, can be used in the treatment of disease.

[0284] In one aspect, the present invention relates to the proteinaceous prodrug construct, for use in therapy, e.g., as a medicament. In a further aspect, the present invention relates to the composition, a nucleic acid, a vector or a host cell as described herein, for use in therapy, e.g., as a medicament.

[0285] In some embodiments, the proteinaceous prodrug construct according to the invention, is for use in treating a disease or disorder of the nervous system, the eye, the circulatory system, the respiratory system, the digestive system, or the skin. In some embodiments, the disease or disorder is a neoplasm, a blood disorder, a metabolic disorder, an autoimmune disease, an immunodeficiency, or an infectious disease. In some embodiments, the neoplasm is a cancer selected from brain cancer, glioblastoma, lung cancer, colorectal cancer, skin cancer, malignant melanoma, pancreas cancer, bladder cancer, liver cancer, breast cancer, eye cancer, and prostate cancer, the cancer is a haematological cancer, such as selected from the group consisting of multiple myeloma, acute myeloblastic leukemia, chronic myelogenic leukemia, acute lymphoblastic leukemia, and chronic lymphocytic leukemia, or the cancer is malignant melanoma, breast cancer, non-small cell lung cancer, pancreatic cancer, head & neck cancer, liver cancer, sarcoma, and B cell lymphoma. In some embodiments, the the autoimmune disease is selected from arthritis (e.g., rheumatoid arthritis or psoriatic arthritis), multiple sclerosis, systemic lupus erythematosus, and inflammatory bowel disease.

[0286] In one embodiment, the proteinaceous prodrug construct according to the invention, is for use in the treatment of cancer. Accordingly, at least one or more of the protease cleavage sites are specific to a protease expressed by a cancer. In another embodiment, the proteinaceous prodrug construct according to the invention, is for use in the treatment of arthritis. In a further embodiment, the composition, a nucleic acid, a vector or a host cell according to the invention, is for use in the treatment of cancer. In yet a further embodiment, the composition, a nucleic acid, a vector or a host cell according to the invention, is for use in the treatment of arthritis.

[0287] The proteinaceous prodrug construct, the composition, the nucleic acid, the vector or the host cell as described herein can also be used in methods of treatment. Thus, in another aspect, the disclosure relates to a method of treatment, the method comprising administering a therapeutic amount of the proteinaceous prodrug construct, the composition, the nucleic acid, the vector or the host cell as described herein to a subject in need thereof. The subject in need thereof may be a subject suffering from cancer or arthritis.

[0288] In one embodiment, the cancer is a solid tumor. In some embodiments, the cancer is selected from the list consisting of brain cancer, glioblastoma, lung cancer, colorectal cancer, skin cancer,malignant melanoma, pancreas cancer, bladder cancer, liver cancer, breast cancer, eye cancer, and prostate cancer.

[0289] In another embodiment, said cancer is a hematological cancer, such as selected from the group consisting of multiple myeloma, acute myeloblastic leukemia, chronic myelogenous leukemia, acute lymphoblastic leukemia, and chronic lymphocytic leukemia.

[0290] In a further embodiment, said cancer is malignant melanoma, breast cancer, non-small cell lung cancer, pancreatic cancer, head & neck cancer, liver cancer, sarcoma, or B cell lymphoma.

[0291] In some embodiment, a proteinaceous prodrug construct in accordance with the invention comprises a CPAMD protein (e.g., A2M) with a modified bait region. In some embodiments, the bait region is modified to change the selection of proteases that are able to cleave it and trigger the conformational change of the CPAMD protein (e.g., A2M). For example, the bait region is modified to be cleaved by a particular protease or class of proteases (e.g., MMPs such as MMP2). In other embodiments, second protease cleavage site is chosen to induce specificity of the release.

[0292] In one embodiment, the cancer expresses one or more proteases, specific for a cleavage site in the bait region of the CPAMD protein (e.g., A2M). In particular embodiments, a proteinaceous prodrug construct in accordance with the invention comprises a CPAMD protein (e.g., A2M) comprising a modified bait region that can be selectively cleaved by one or more proteases expressed by the cancer.

[0293] In one embodiment, the cancer expresses one or more proteases selected from the list consisting of activated protein C, ADAM 10, ADAM 12, ADAM 15, ADAM17 / TACE, ADAM9, ADAMDEC1, ADAMTS1, ADAMTS4, ADAMTS5, BACE, BMP-1, Caspase 1, Caspase 10, Caspase 14, Caspase 2, Caspase 3, Caspase 4, Caspase 5, Caspase 6, Caspase 7, Caspase 8, Caspase 9, Cathepsin A, Cathepsin B, Cathepsin C, Cathepsin D, Cathepsin E, Cathepsin G, Cathepsin K, Cathepsin L, Cathepsin S, Cathepsin V / L2, Cathepsin X / Z / P, Chymase, Cruzipain, DESCI, DPP-4, Elastase, FAP, Granzyme B, Guanidinobenzoatase, Hepsin, HtrAl, Neutrophil Elastase, KLK10, KLK11, KLK13, KLK14, KLK4, KLK5, KLK6, KLK7, KLK8, Lactoferrin, Legumain, Marapsin, Matriptase-2, Meprin, MMP1, MMP8, MMP10, MMP11, MMP12, MMP13, MMP14, MMP15, MMP16, MMP17, MMP19, MMP2, MMP20, MMP23, MMP24, MMP26, MMP27, MMP3, MMP7, MMP8, MMP9, MT-SPl / Matriptase, Neprilysin, NS3 / 4A, Otubain-2, PACE4, Plasmin, PSA, PSMA, Renin, Thrombin, TMPRSS2, TMPRSS3, TMPRSS4, tPA, Tryptase, uPA, ADAM8, FVIIa, FIXa, Furin, FXa, FXIa, FXIIa, and TAFI.

[0294] Another aspect relates to a method of treatment, wherein the suitable biopharmaceutical moieties and optionally, proteases for the first- and / or second protease cleavage sites are chosen, the final construct designed thereafter, hence suitably an intermediate nucleic acid, as described above, is chosen. Thus, in another aspect, the invention relates to a method of treating a subject in need thereof, the method comprising:1. providing the nucleic acid or the vector described above as an intermediate;. determining a first biopharmaceutical moiety or determining a first- and a second biopharmaceutical moiety, suitable for treating the subject, optionally determining a suitable first- and / or second protease cleavage site;3. inserting the first-and / or second biopharmaceutical moiety, and the optional first- and / or second protease cleavage site of step 2, into the nucleic acid, such that the nucleic acid or the vector encodes a proteinaceous prodrug according to the invention;4. introducing the nucleic acid or the vector into a host cell;5. growing the host cell under conditions that allows for expression of the proteinaceous fusion construct from the nucleic acid or the vector;6. optionally, purifying the proteinacous produg and / or cleaving the second protease cleavage site; and7. administering the proteinacous produg to the subject in need thereof.

[0295] A similar aspect relates to the nucleic acid administered as a drug, thus in a further aspect the disclosure relates to a method of treating a subject in need thereof, the method comprising:1. providing the nucleic acid or the vector described above as an intermediate;2. determining a first biopharmaceutical moiety or determining a first- and a second biopharmaceutical moiety, suitable for treating the subject, optionally determining a suitable first- and / or second protease cleavage site;3. inserting the first-and / or second biopharmaceutical moiety, and the optional first- and / or second protease cleavage site of step 2, into the nucleic acid, such that the nucleic acid or the vector encodes a proteinaceous prodrug according to the invention;4. administering the nucleic acid to the subject in need thereof.Subject and administration

[0296] The “subject” as described herein comprises humans of all ages, other primates (e.g., cynomolgus monkeys, rhesus monkeys); mammals in general, including commercially relevant mammals such as cattle, pigs, horses, sheep, goats, mink, ferrets, hamsters, cats, dogs; and / or birds. In a typical embodiment, the subjects are humans.

[0297] The term “subject” also includes healthy subjects of the population and, in particular, healthy subjects, who are exposed to pathogens and in need of protection against infection, such as health personnel.

[0298] Further, pathogenic infections caused by a virus of the respiratory system can be particularly serious in elderly and weak patients and patients with chronic or congenital dysfunction of the respiratory system, such as asthma, cystic fibrosis, or chronic obstructive pulmonary disease (COPD).

[0299] Thus, in an embodiment of the present invention, the subject is selected from the group consisting of; humans of all ages, other primates (e.g., cynomolgus monkeys, rhesus monkeys);mammals in general, including commercially relevant mammals, such as cattle, pigs, horses, sheep, goats, mink, ferrets, hamsters, cats and dogs, as well as birds.

[0300] In a particular embodiment, the subject is a human.Method for producing a proteinaceous fusion construct

[0301] Also provided herein are methods of producing the fusion proteins and proteinaceous prodrug constructs of the invention.

[0302] In some aspects, a method for producing a proteinaceous prodrug construct of the invention comprises providing a host cell comprising a nucleic acid encoding the proteinaceous prodrug construct; and culturing the host cell under conditions that allow for expression of the proteinaceous prodrug construct from the nucleic acid. In some embodiments, the proteinaceous prodrug construct is contacted with a protease (e.g., TEV protease) that specifically cleaves a protease cleavage site provided at the N- terminal end of the RBD (e.g., a second protease cleavage site such as a TEV cleavage site).

[0303] In some aspects, a method for producing a proteinaceous prodrug construct of the invention comprises providing the proteinaceous prodrug construct and contacting it with a protease (e.g., TEV protease) that specifically cleaves a protease cleavage site provided at the N-terminal end of the RBD (e.g., a second protease cleavage site such as a TEV cleavage site).

[0304] In some embodiments, the methods for producing a proteinaceous prodrug construct of the invention comprise one or more purification steps, e.g., using chromatography such as size exclusion chromatography.

[0305] In some aspects, the present invention relates to a method for producing a proteinaceous fusion construct according to the invention, the method comprising:1 . providing the nucleic acid encoding the prodrug as described herein, or the vector encoding the prodrug as described herein;2. introducing the nucleic acid or the vector into a host cell;3. growing the host cell under conditions that allows for expression of the proteinaceous fusion construct from the nucleic acid or the vector; and4. optionally, purifying the proteinacous produg and / or cleaving the second protease cleavage site.

[0306] In some aspects, the present invention relates to a method of producing a proteinaceous prodrug, the method comprising:1. providing the nucleic acid or the vector described above as an intermediate;2. determining a first biopharmaceutical moiety or determining a first- and a second biopharmaceutical moiety, optionally determining a suitable first- and / or second protease cleavage site;3. inserting the first-and / or second biopharmaceutical moiety, and the optional first- and / or second protease cleavage site of step 2, into the nucleic acid, such that the nucleic acid or the vector encodes a proteinaceous prodrug according to the invention;4. introducing the nucleic acid or the vector into a host cell;5. growing the host cell under conditions that allows for expression of the proteinaceous fusion construct from the nucleic acid or the vector; and6. optionally, purifying the proteinacous produg and / or cleaving the second protease cleavage site.

[0307] In some embodiments it is particularly relevant to cleave the second protease cleavage site as part of the production method, since the application of multivalent molecules may be harmful, as described previously.

[0308] Accordingly, provided herein is a method of producing such a proteinaceous prodrug, the method comprising:(a) providing a host cell comprising a nucleic acid encoding a proteinaceous fusion construct according to the invention;(b) culturing the host cell under conditions that allow for expression of the proteinaceous prodrug construct encoded by the nucleic acid; and(c) contacting the proteinaceous prodrug construct with a protease that specifically cleaves the second protease cleavage site.Numbered embodimentsThe invention is further described by reference to the following numbered embodiments:1. A proteinaceous prodrug construct comprising a complement 3- and pregnancy zone protein-like, alpha- 2-macroglobulin domain-containing (CP AMD) protein, such as A2M, or a fragment thereof, the CPAMD protein or a fragment thereof comprising:(a) a bait region comprising at least one first protease cleavage site;(b) a Receptor Binding Domain (RBD);(c) a second protease cleavage site introduced at the N-terminal end of the CPAMD protein’s RBD domain, wherein the RBD domain remain bound to the CPAMD protein, via non-covalent interactions, upon cleavage of the second protease cleavage site; and(d) at least one biopharmaceutical moiety positioned C-terminal to the second protease cleavage site, wherein cleavage of the first protease cleavage site results in release of the RBD domain, and thereby the at least one biopharmaceutical moiety, from the CPAMD protein.2. The proteinaceous prodrug construct according to embodiment 1, wherein the at least one biopharmaceutical moiety is a first biopharmaceutical moiety positioned inside the RBD region, and a second biopharmaceutical moiety positioned C-terminally to the CPAMD protein, wherein the CPAMD protein or fragment thereof shields the first biopharmaceutical moiety and the CPAMD protein or fragment thereof is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site,releasing the RBD, and thereby the first- and second biopharmaceutical moiety, from the CPAMD protein, and making the first biopharmaceutical moiety accessible.3. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein at least one of the at least one biopharmaceutical moiety is capable of directing the proteinaceous prodrug construct to a specific tissue, a specific cell type, and / or a specific receptor.4. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein at least one of the at least one biopharmaceutical moiety is capable of directing the proteinaceous prodrug construct to immune cells.5. The proteinaceous prodrug construct according to embodiment 4, wherein the immune cells are NK cells, macrophages, T cells, or dendritic cells.6. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the biopharmaceutical moiety is T cell specific, such as a T cell specific moiety.7. The proteinaceous prodrug construct according to embodiment 6, wherein the T cell specific moiety is specific against a receptor expressed at increased levels on T cells, such as CD3, CD4 and / or CD8, preferably CD3.8. The proteinaceous prodrug construct according to any of embodiments 6-7, wherein the T cell specific moiety is an anti-CD3 moiety.9. The proteinaceous prodrug construct according to any of embodiments 1-5, wherein the biopharmaceutical moiety is NK cell specific, such as a NK cell specific moiety.10. The proteinaceous prodrug construct according to embodiment 9, wherein the NK cell specific moiety is specific against a receptor expressed at increased levels on NK cells, such as CD 16.11. The proteinaceous prodrug construct according to any of embodiments 1-5, wherein the biopharmaceutical moiety is macrophage specific, such as a macrophage specific moiety.12. The proteinaceous prodrug construct according to embodiment 11, wherein the macrophage specific moiety is specific against a receptor or molecule expressed at increased levels on macrophages, such as a SIRPa inhibitory antibody, such as a SIRPa antibody that blocks CD47.13. The proteinaceous prodrug construct according to any of embodiments 1-5, wherein the biopharmaceutical moiety is specific towards dendritic cells, such as a dendritic cell specific moiety, such as a dendritic cell specific moiety specific against a receptor or molecule expressed at increased levels on macrophages, such as the DNGR1 receptor, such as an antibody against the DNGR1 receptor.14. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein at least one of the at least one biopharmaceutical moiety is a drug.15. The proteinaceous prodrug construct according to any of embodiments 9-10, wherein the drug is able to increase or decrease the signal from a receptor upon binding to the receptor.16. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the RBD is a bi-specific drug upon release from the CP AMD protein.17. The proteinaceous prodrug construct according to embodiment 16, wherein the bi-specific drug comprises one biopharmaceutical moiety capable of directing the proteinaceous prodrug construct to immune cells according to any of embodiments 3-13 and a drug according to any of embodiments 14-16.18. The proteinaceous prodrug construct according to embodiment 12, wherein the bi-specific drug comprises at least two drugs according to any of embodiments 14-16.19. The proteinaceous prodrug construct according to any of embodiments 16-17, wherein the bi-specific drug is a Bi-specific T cell engager (BiTE), Bi-specific NK cell engager (BiKE), Bi-specific Macrophage engager (BiME), or a Bi-specific dendritic cell engager (BiDE).20. The proteinaceous prodrug construct according to any of embodiments 16-17, wherein the bi-specific drug is a Bi-specific T cell engager (BiTE).21. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the at least one biopharmaceutical moiety is a protein, such as an antigen-binding fragment of an antibody, such as a ScFv, or such as a single domain antibody.22. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the biopharmaceutical moiety is an antigen-binding fragment of an antibody, such as a single domain antibody, that specifically binds to antigen selected from the group consisting of IL-2, EGFR, PDL-1, PD-1, CTLA- 4, CD3ya, 4-1BB, IL-2Ra, and TNFa.23. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the biopharmaceutical moiety is selected from the group consisting of Atezolizumab, EgAl, Ipilimumab, Nivolumab, KN035, Urelumab, Foralumab, Muromonab, Adalimumab, and therapeutically active antigenbinding fragments or variants of each.24. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the biopharmaceutical moiety is a cytokine, or a therapeutically active fragment or variant thereof, selected from the group consisting of IL1, ILlalpha, ILlbeta, IL2, IL3, IL4, IL6, IL7, IL8, IL9, IL10, IL11, IL12, IL13, IL14, IL15, IL16, IL17, 118, IL19, IL20, IL21, IL22, IL23, IL24, IL25, IL26, IL27, IL28, IL29, IL30, IL31, IL32, IL33, IL34, IL35, IL36, GM-CSF, TGF-p, CSF-1, insulin, GLP-1, HGH, VEGF, PDGF, BMP, EPO, G-CSF, IL-11, IFN-a, IFN-P and IFN-y.25. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the CPAMD protein is selected from the group consisting ofCPAMD1 (a k a. C3), CPAMD2 (a.k.a. C4A), CPAMD3 (a.k.a. C4B), CPAMD4 (a.k.a. C5), CPAMD5 (a k a. A2M), CPAMD6 (a.k.a. PZP), CPAMD7 (a.k.a. CD 109), CPAMD8, CPAMD9 (a.k.a. A2ML1), Ovostatin 1, and Ovostatin 2.26. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the biopharmaceutical moiety is inserted between residues 1402-1403 of A2M.27. The proteinaceous prodrug construct according to any of the preceding embodiments wherein the proteinaceous prodrug is encoded by a single nucleic acid as a continuous peptide chain, and preferably wherein the at least one biopharmaceutical moiety, and RBD, consist of a single protein chain upon proteolytic cleavage of the first protease cleavage site.28. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the second protease cleavage site is positioned between two linkers, such as a linker of 5-30 amino acid length, such as a linker of 5-25 amino acid length such as a GS linker of 5-30 amino acid length, such as a GS linker of 5-25 amino acid length.29. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein first protease cleavage site and the second protease cleavage site is specific towards the same proteases.30. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the second protease cleavage site is specific towards fiirin, TEV, Enterokinase, or Thrombin.31. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the second protease cleavage site introduced at the N-terminal end of CPAMD protein’s RBD domain is introduced from residue position 1334 to 1340 of A2M.32. The proteinaceous prodrug construct according to any of the preceding embodiments, wherein the first protease cleavage site is specific for a serine-, cysteine-, aspartic- and / or metalloproteinase.33. A nucleic acid or a vector, such as a plasmid, encoding a multimeric, such as tetrameric CPAMD protein or a fragment thereof, such as A2M protein, the CPAMD protein or a fragment thereof comprising a second protease cleavage site at the N-terminal end of the CPAMD protein’s RBD domain.34. The nucleic acid according to embodiment 33, wherein the CPAMD protein or a fragment thereof comprises a bait region said bait region comprising at least one first protease cleavage site.35. The nucleic acid according to any of embodiments 33-34, wherein the CPAMD protein or a fragment thereof comprises a Receptor Binding Domain (RBD).36. A nucleic acid encoding the proteinaceous prodrug construct according to any of embodiments 1-32.37. A vector, such as a plasmid, comprising the nucleic acid according to embodiment 36.38. A host cell comprising the nucleic acid according to embodiment 36 or vector according to embodiment 37.39. The host cell according to embodiment 38, wherein the host cell is a bacteria or eukaryote, e.g., a mammalian cell.40. The proteinaceous prodrug construct according to any of embodiments 1-32, the nucleic acid according embodiment 36, the vector according to embodiment 37 or the host cell according to any of embodiments 38-39 for use as a medicament.41. The proteinaceous prodrug construct according to any of embodiments 1-32, the nucleic acid according embodiment 36, the vector according to embodiment 37 or the host cell according to any of embodiments 38-39 for use in the treatment of a disease or disorder of the nervous system, the eye, the circulatory system, the respiratory system, the digestive system, or the skin.42. The proteinaceous prodrug construct according to any of embodiments 1-32, the nucleic acid according embodiment 36, the vector according to embodiment 37 or the host cell according to any of embodiments38-39 for use in the treatment of a neoplasm, a blood disorder, a metabolic disorder, an autoimmune disease, an immunodeficiency, or an infectious disease.43. The proteinaceous prodrug construct, the nucleic acid, the vector or the host cell for use according to embodiment 42, wherein the neoplasm is a cancer selected from brain cancer, glioblastoma, lung cancer, colorectal cancer, skin cancer, malignant melanoma, pancreas cancer, bladder cancer, liver cancer, breast cancer, eye cancer, and prostate cancer, the cancer is a haematological cancer, such as selected from the group consisting of multiple myeloma, acute myeloblastic leukemia, chronic myelogenic leukemia, acute lymphoblastic leukemia, and chronic lymphocytic leukemia, or the cancer is malignant melanoma, breast cancer, non-small cell lung cancer, pancreatic cancer, head & neck cancer, liver cancer, sarcoma, and B cell lymphoma.44. The proteinaceous prodrug construct, the nucleic acid, the vector or the host cell for use according to embodiment 42, wherein the autoimmune disease is selected from arthritis (e.g., rheumatoid arthritis or psoriatic arthritis), multiple sclerosis, systemic lupus erythematosus, and inflammatory bowel disease.45. A method of treating a subject in need thereof, the method comprising:1. providing the nucleic acid or the vector according to any of embodiments 33-35;2. determining a first biopharmaceutical moiety or determining a first- and a second biopharmaceutical moiety, suitable for treating the subject, optionally determining a suitable first- and / or second protease cleavage site;3. inserting the first-and / or second biopharmaceutical moiety, and the optional first- and / or second protease cleavage site of step 2, into the nucleic acid, such that the nucleic acid or the vector encodes a proteinaceous prodrug according to any of embodiments 1-32;4. introducing the nucleic acid or the vector into a host cell;5. growing the host cell under conditions that allows for expression of the proteinaceous fusion construct from the nucleic acid or the vector;6. optionally, purifying the proteinacous produg and / or cleaving the second protease cleavage site; and7. administering the proteinacous produg to the subject in need thereof. [US method of treatment]46. A method of treating a subject in need thereof, the method comprising:1. providing the nucleic acid or the vector according to any of embodiments 33-35;2. determining a first biopharmaceutical moiety or determining a first- and a second biopharmaceutical moiety, suitable for treating the subject, optionally determining a suitable first- and / or second protease cleavage site;3. inserting the first-and / or second biopharmaceutical moiety, and the optional first- and / or second protease cleavage site of step 2, into the nucleic acid, such that the nucleic acid or the vector encodes a proteinaceous prodrug according to any of embodiments 1-32;4. administering the nucleic acid to the subject in need thereof. [US method of treatment]47. A method of producing a proteinaceous prodrug, the method comprising:1. providing the nucleic acid or the vector according to any of embodiments 33-35;2. determining a first biopharmaceutical moiety or determining a first- and a second biopharmaceutical moiety;3. inserting the first- and / or second drug biopharmaceutical moiety into the nucleic acid, such that the nucleic acid or the vector encodes a proteinaceous prodrug according to any of embodiments 1-32;4. introducing the nucleic acid or the vector into a host cell;5. growing the host cell under conditions that allows for expression of the proteinaceous fusion construct from the nucleic acid or the vector; and6. optionally, purifying the proteinacous produg and / or cleaving the second protease cleavage site.48. A method of producing a proteinaceous prodrug, the method comprising:1. providing the nucleic acid according embodiment 36, or the vector according to embodiment 37;2. introducing the nucleic acid or the vector into a host cell;3. growing the host cell under conditions that allows for expression of the proteinaceous fusion construct from the nucleic acid or the vector; and4. optionally, purifying the proteinacous produg and / or cleaving the second protease cleavage site.49. A method of treating or preventing a disease or disorder in a subject in need thereof, wherein the method comprises administering a therapeutically effective amount of the proteinaceous prodrug construct according to any of embodiments 1-32, the nucleic acid according embodiment 36, the vector according to embodiment 37 or the host cell according to any of embodiments 38-39 to the subject.Equivalents

[0309] Although methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, suitable methods and materials are described below. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety. The references cited herein are not admitted to be prior art to the claimed invention. In addition, the materials, methods, and examples are illustrative only and are not intended to be limiting.Examples

[0310] The invention will now be described in further detail in the following non-limiting examples.Example 1 - Overview of the invention

[0311] This shows the overall design and mechanism of the invention, which relates to a technology for producing protease-activated prodrug versions of biopharmaceuticals. With reference to figure 1A, a proteinaceous prodrug construct (1) comprises a CP AMD protein, e.g., human alpha-2 -macroglobulin (A2M) (2), fused to one or more drugs (3) in such a manner that the drugs accessibility is dependent on the conformational state of the CP AMD protein (e.g., A2M) (2). The CP AMD protein (e.g., A2M) (2) is transformed from an initial “native” conformation to an “activated” conformation by one or more proteases (4). The drugs (3) may be genetically fused to the CP AMD protein (e.g., A2M) (2) at a position where it is inaccessible to its therapeutic target in the “native” conformation (I) of the CP AMD protein (e.g., A2M) (2), but is accessible in the CP AMD protein’s (e.g., A2M’s) (2) “activated” conformation (II). In this way, the activity of the one or more drugs (3) is spatially restricted to tissues where one or more proteases (4) that can activate the CP AMD protein (e.g., A2M) (2) are present and proteolysis- competent. As the CP AMD protein (e.g., A2M) (2) can be modified to be activated by one or more designated proteases (4), this technology allows the targeting of drugs to tissues expressing disease- associated proteases, e.g., diseased tissue, thereby potentially improving the efficacy of a drug while minimizing side effects arising from target binding in healthy tissues.Example 2 - Production of proteinaceous fusion constructs of A2M and antibodiesAim

[0312] This data shows the expression and purification of A2M-antibody constructs as correctly folded tetrameric proteins, where A2M assumes a functional native conformation with a thiol ester.Materials and methodsExpression and purification of A2M-antibody fusion constructs

[0313] The nucleotide sequences encoding A2M-antibody fusion constructs and the corresponding amino acid sequences are given (SEQ ID NO: 5-22).

[0314] Proteinaceous fusion constructs were expressed in HEK293 FreeStyle cells using a standard transient transfection protocol. Briefly, 25 kDa linear polyethyleneimine (Polysciences) and plasmid DNA were incubated for 10 min in antibiotic-free FreeStyle medium (Thermo Fisher Scientific) at a 4: 1 w / w PEEDNA ratio, then slowly dripped into a culture of cells at a density of 1 million cells per mb, to a final DNA concentration of 1 pg per mb culture. After 4 days, the supernatant was harvested by spinning down the cells at 1500 x g and adding pH 7.4 HEPES to a final concentration of 50 mM.

[0315] Purification of the constructs was performed using an established protocol for purifying A2M. Supernatants were first run through a Zn2+-loaded Chelating HiTrap column (GE Healthcare) and elutedwith 50 mM EDTA, 150 mM NaCl, 100 mM sodium acetate, pH 7.4. The EDTA eluate was dialyzed against 20 mM HEPES at pH 7.4, then loaded onto a HiTrap Q column (GE Healthcare) and eluted by a gradient of 0-400 mM NaCl (with a constant 20 mM HEPES at pH 7.4). Fractions containing A2M were pooled, concentrated by ultrafdtration, and purified by size exclusion chromatography on a Sephacryl S-300 HR (GE Healthcare), using a 20 mM HEPES, 150 mM NaCl, pH 7.4 running buffer (HEPES-buffered saline, HBS).SDS-PAGE and pore limited native PAGE

[0316] Native pore limited PAGE was performed as previously described (36), using homemade gels in TBE buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA) with an acrylamide gradient of 5-10% for A2M analysis and 10-15% for C3 analysis. Pore limited electrophoresis gels were run overnight at 100 V in TBE buffer.

[0317] Denaturing SDS-PAGE was performed using the discontinuous 2 -amino -2 -methyl -1,3- propanediol and glycine buffer system on homemade 5-15% acrylamide gradient gels. Samples were reduced with 25 mM DTT at 95 °C for 5 minutes.Reaction of A2M with methylamine and proteases

[0318] To amino lyze A2M’s thiol ester, methylamine (pH 8) was added to 250 mM and incubated for at 16 hours at 37 °C. To assess the cleavage of A2M by thermolysin, thermolysin was added to a 2.2: 1 mol / mol ratio of protease:A2M and incubated for five minutes at 37 °C. The digestion was then inhibited using EDTA (10 mM, 15 minutes, room temperature).Results

[0319] Proteinaceous fusion constructs were produced with yields of several mg / L in transient HEK293F transfections. After purification, the constructs migrated as native homotetramers in native PAGE (Figure 2A), and as —190 kDa monomeric subunits in reducing, denaturing SDS-PAGE (Figure 2B) The presence of a thiol ester in the constructs was verified by the presence of the characteristic heat-induced fragmentation at the site of the thiol ester, generating the Nt and Ct autolytic fragments visible in SDS-PAGE (Figure 2B-C). These Nt and Ct autolytic fragments disappeared if the thiol ester was aminolyzed with methylamine prior to SDS-PAGE analysis. Proteolytic processing of the bait region was assessed by treatment with thermolysin. The bait regions of the constructs were preferentially cleaved by thermolysin, resulting in the formation of the Nt and Ct cleavage fragments (Figure 2B-C). Bait region cleavage induced a conformational change in the constructs that increased their migration in native PAGE, although not to the same extent as in wildtype A2M (Figure 2A); this is because the exposed antibody fragments increase the electrophoretic resistance experienced by activated A2M-antibody constructs.Conclusion

[0320] Proteinaceous fusion constructs of A2M and antibody scFvs are produced as homotetrameric proteins. In these proteins, the A2M component is functionally normal, as it assumes a native conformation, forms a thiol ester, and is preferentially cleaved in its bait region by proteases.Example 3 - Conformational dependence of binding in biolayer interferometry

[0321] This example shows how the conformational change of the proteinaceous fusion construct is able to control the activity of the drug. In the native state, the drugs are not exposed and thus, inactive. In the active state, the drug is exposed and able to interact with its target.Aim

[0322] To determine the antigen-binding capacity of fusion constructs of A2M and antibodies, in binding experiments using purified antigens that are immobilized to biosensors, and to determine the extent to which this antigen-binding capacity is affected by the conformation of A2M.Materials and methodsProteins used for binding studies

[0323] Antigens to the antibodies under investigation were recombinantly expressed in HEK293F cells using a standard transient transfection protocol (see Example 1). The antigens were expressed with the leader peptide of A2M, N-terminal StrepII tags, and a C-terminal Fc region from human IgGl (uniprot ID P01857, residues 100-330, SEQ ID NO: 40). The residues included for each antigen were as follows, using numbering before removal of the signal peptide:• EGFR (uniprot ID P00533, residues 25-645, SEQ ID NO: 37)• PD-L1 (uniprot ID Q9NZQ7, residues 19-239, SEQ ID NO: 38)• PD-1 (uniprot ID Q15116, residues 26-150 with a Cys93Ser mutation, SEQ ID NO: 39)• CTLA-4 (uniprot ID Pl 6410, residues 36-161, SEQ ID NO: 40)• CD3ye (uniprot ID P09693, residues 23-103 of the y chain followed by a 26-residue glycineserine linker and uniprot ID P07766, residues 23-118 of the 8 chain, SEQ ID NO: 41)• 4-1BB (uniprot ID Q07011, residues 24-186, SEQ ID NO: 42)

[0324] The final sequences of these antigens in fusion as expressed are given both as amino acid and nucleotide sequences (SEQ ID NO: 45-58).

[0325] An additional antigen, TNFa (uniprot P01375, residues 77-233) was expressed as a StrepII- tagged protein but without a C-terminal Fc region (SEQ ID NO: 59, 60). It was also purified by StrepTactin affinity chromatography and size exclusion chromatography on a Superdex 200 Increase to isolate TNFa trimers.

[0326] Supernatants containing the expressed antigens were purified using StrepTactin affinity chromatography (Iba Life Sciences), followed by size exclusion chromatography on a Superdex 200 Increase (GE Healthcare).

[0327] The A2M-antibody fusion constructs were produced as stated in Example 1. Where stated, native A2M-antibodies were purified by affinity depletion of pre-activated A2M; see Example 5 for further details. The amino acid and nucleotide sequences of the A2M-antibodies are given (SEQ ID NO: 5-22).Reaction of A2M-antibodies with methylamine and proteases

[0328] When A2M-antibodies were treated with methylamine, 200 mM of methylamine (pH 8) was added to the A2M-antibody and it was incubated for 16 hours at 37 °C. When A2M-antibodies were treated with thermolysin, thermolysin from Geobacillus stearothermophilus (Sigma-Aldrich) was added to the A2M-antibody at a 2.2: 1 molar ratio of protease:A2M and incubated for 5 minutes at 37 °C, after which point thermolysin was inhibited by the addition of 25 mM EDTA.Biolayer interferometry

[0329] HEPES-buffered saline (HBS; 20 mM HEPES, 150 mM NaCl, pH 7.4) was used as the buffer in all biolayer interferometry experiments. Antigens were immobilized onto anti-human Fc capture biosensors (AHC biosensors; Fortebio) at 30 nM in HBS for 20 minutes. A2M-antibody fusion constructs were then incubated with the antigen-coated biosensors at various concentrations to measure association, followed by measurement of dissociation in HBS. Where stated, A2M-antibody fusion constructs were activated by methylamine or protease treatment, using the same method as in Example 1.Results

[0330] The conformational dependence of antigen binding by eight different A2M-antibodies was assessed by biolayer interferometry. Antigens were expressed as fusion proteins with the human IgGl Fc region, allowing antigens to be immobilized onto the biosensor surface using anti-human Fc capture biosensors in a standardized manner. A2M-antibodies were then allowed to associate with their immobilized antigens, either without any treatment of the A2M-antibody or with the induction of A2M’s conformational change using methylamine aminolysis and / or proteolysis by thermolysin. In some cases, the A2M-antibodies were enriched for the native conformation of A2M by affinity depletion using an antigen (PD-L1) or LRP1 resin, as noted in the figure legend and elaborated in Example 4.

[0331] For all investigated antibodies, antigen binding was strongly dependent on the conformation of A2M (Figure 3A-I), with thermolysin proteolysis of A2M consistently giving the highest binding response. Methylamine treatment produced variable binding responses; in some cases, methylamine induced a binding response similar to proteolysis (Figure 3B, 31), whereas in other cases it produced an intermediate response (Figure 3C, 3F-H) or a neglible response (Figure 3D). When the native conformation of an A2M-antibody had been enriched by affinity depletion, litte to no antigen binding was detected in the native sample without methylamine / proteolysis (Figure 3A-D, 3F).Conclusion

[0332] Antibodies incorporated into A2M fusion constructs retain the ability to bind their cognate antigen. This antigen binding is determined by the conformation of A2M, with little to no antigen binding in the native conformation of A2M. Activation of A2M by proteolytic cleavage greatly increases antigen binding, whereas activation by methylamine treatment varied depending on the A2M-antibody in question.Example 4 - Enrichment of native A2M-antibody constructs by affinity depletion

[0333] This example shows how modification of the proteinaceous fusion construct can be used to control where the drug becomes exposed. Depending on the cleavage site, the drug can only be exposed at the location where proteases recognizing that cleavage site are present. When a specific protease is present and cleaves the cleavage site, the conformation of the proteinaceous fusion construct is changed from “naive” to “active”.Aim

[0334] Recombinantly expressed A2M-antibody fusion constructs are not exclusively produced with A2M in its native conformation; a minor component is produced in a pre-activated state. Here, it is investigated whether this pre-activated component can be removed by affinity depletion using the antibody’s cognate antigen, the activated A2M receptor LRP1, or kappa light chain -binding Protein L.Materials and methodsProteins used

[0335] A2M-antibodies were produced as described in Example 1. Recombinant LRP1 (residues 20- 974, SEQ ID NO: NO 63-64) was produced as a StrepII-tagged fusion protein with the human IgGl Fc region, as described for the antigens in Example 2.Resin preparationA resin coated with LRP1 was prepared using amine reactive chemistry. A total of 200 mg of NHS- activated agarose (Pierce) and 600 pg of recombinant LRP1 in 0.15 M triethylammonium bicarbonate, 0.15 M HEPES, pH 8.3 were mixed on a rotator at room temperature for 2 hours. Following incubation, the resin was washed twice in HBS and the reaction was quenched with 50 mM Tris-HCl, pH 8 for 20 minutes, followed by a final washing step with HBS. A resin coated with PD-L1 was prepared as described for LRP1. Protein L-coated agarose was purchased from Pierce (Thermo Scientific).Affinity depletion

[0336] To deplete pre-activated A2M, A2M-antibody fusion constructs in HBS at up to 2 mg / mL were incubated with resin at room temperature overnight while shaken using a helicopter rotor. For LRP1-based depletion, 10 mM of CaCE were added to the HBS. After overnight incubation, the supernatant was recovered and the resin was regenerated using HBS with 25 mM of EDTA in the case of LRP1, or using acidic elution with a pH 2.7, 10 mM KH2PO4 buffer for PD-L1 and Protein L. The recovered supernatant was tested using biolayer interferometry, as described in Example 2.Results

[0337] A2M-Atezolizumab was incubated with a resin coated with its cognate antigen, PD-L1. A single round of depletion was performed. The binding of A2M-Atezolizumab to PD-L1 before and after this depletion was then assessed using biolayer interferometry. Whereas A2M-Atezolizumab from before and after depletion bound similarly to PD-L1 upon methylamine treatment, antigen binding by the untreated sample after depletion was greatly decreased compared to the untreated sample before depletion, indicating that PD-L1 depletion had enriched the content of A2M-antibodies with inaccessible antibodies (Figure 4A).

[0338] A2M-Ipilimumab, A2M-Nivolumab, and A2M-Urelumab were incubated with a resin coated with LRP1, a receptor that specifically binds to activated A2M but not to native A2M. Three rounds of depletion were performed for each A2M-antibody, after which biolayer interferometry was used to assess their binding to CTLA-4, PD-1, or 4-1BB, respectively (Figure 4B-D). Antigen binding by the untreated A2M-antibodies before LRP1 depletion was approximately 25% of the maximum binding defined by thermolysin activation for all three antibodies. After LRP1 depletion, no antigen binding was detectable in the untreated A2M-Ipilimumab and A2M-Nivolumab samples and very little antigen binding was detectable in the untreated A2M-Urelumab sample, whereas equivalent binding was observed for the thermolysin-activated A2M-antibodies before and after LRP1 depletion. These data show that LRP1 depletion was able to deplete A2M-antibodies in which the antibodies were prematurely capable of antigen binding, further demonstrating that the conformation of A2M and the antibody accessibility are correlated.

[0339] A2M-Ipilimumab was also depleted using a Protein L-coated resin, which specifically binds to the K light chain of human antibodies. Three rounds of depletion were performed. Biolayer interferometry showed that Protein L-based depletion was able to remove antigen binding in the untreated A2M-Ipilimumab sample (Figure 4E).Conclusion

[0340] A2M-antibodies in their native and activated conformations can be distinguished by affinity depletion based on their binding to their antigen, to LRP1, or to Protein L. This binding can be used to remove activated A2M-antibodies and prepare native A2M-antibodies to a higher purity. Binding experiments comparing antigen binding before and after depletion show that the enrichment of native A2M-antibodies leads to minimal or no detectable antigen binding by the native protein, demonstrating that antigen binding by untreated A2M-antibodies is caused by contamination by non-native A2M- antibodies.Example 5 - Investigating immune checkpoint blockade in a cell assay.Aim

[0341] In order to investigate whether A2M-antibodies demonstrate conformation-dependent target binding in a cellular context and retain the biological activity of their parent antibodies, A2M- Atezolizumab was investigated in a PD-1 / PD-L1 blockade bioassay.Materials and methodsProteins used

[0342] A2M-Atezolizumab was expressed and purified as described for A2M-antibody fusion constructs in Example 2. Native A2M-Atezolizumab was enriched using PD-L1 -based affinity depletion, as described in Example 4. Methylamine-treated A2M-Atezolizumab was prepared by 16 hours of incubation with 200 mM of methylamine at 37 °C, followed by desalting back into HBS on a PD-10 column. The Atezolizumab scFv was also expressed in fusion with a human IgGl Fc region, with N-terminal StrepII tags, and this Atezolizumab-hFc was purified using the same protocol as for antigen- hFc fusion constructs described in Example 3, namely StrepTactin affinity chromatography followed by size exclusion chromatography.Cell-based assessment of immune checkpoint blockade

[0343] The ability of A2M-Atezolizumab and Atezolizumab-hFc to block the PD-1 / PD-L1 pathway on human T cells was tested using the PD-1 / PD-L1 Blockade Bioassay developed by Promega. Jurkat cells were cultured in RPMI 1640 medium supplemented with penicillin / streptomycin and 10% fetal bovine serum, while CHO-K1 cells were cultured in DMEM medium supplemented with penicillin / streptomycin and 10% fetal bovine serum. The day before performing the assay, 40* 103CHO- K1 cells per well were seeded onto a 96-well plate. On the day of the assay, medium was removed from the wells and replaced by 40 pL of antibody solution diluted in assay buffer (RPMI 1640 medium with 1% fetal bovine serum) and 40 pL with 50* 103Jurkat cells in assay buffer. The plates were then incubated at 37 °C for 6 hours, after which point 80 pL of Bio-Gio Reagent (Promega) were added to each well and luminescence was measured in a plate reader. Each antibody concentration and controls were tested in triplicate wells. The luminescence signal is given as averaged normalized luminescence for the three wells, with the background (measured from wells which did not receive any antibody) subtracted and the response normalized to the highest measured luminescence from the assay (with background subtracted).Results

[0344] The PD-1 / PD-L1 Blockade bioassay developed by Promega was used to investigate conformation-dependent PD-L1 blocking by A2M-Atezolizumab. This bioassay uses the human Jurkat T cell line expressing human PD-1, as well as a luciferase reporter gene driven by an NFAT responseelement, to represent human T cells. CH0-K1 cells expressing human PD-L1 and an engineered surface protein that activates cognate TCRs in an antigen-independent manner are used to represent PD-L1+target cells. The TCR-activating CHO-Kls would activate the Jurkat cells and induce a NFAT-driven luciferase response, except that this response is inhibited by PD-1 -mediated signaling due to the engagement of PD-1 on the Jurkat cells by PD-L1 on the CH0-K1 cells. If either PD-1 or PD-L1 is blocked by an antibody, the luciferase response is restored.

[0345] A titration series of A2M-Atezolizumab (from 20 pM to 200 nM) in its native conformation and methylamine-treated collapsed conformation was used to block PD-L1 on the surface of CHO-K1 cells. An IgG-resembling construct produced by fusing the Atezolizumab scFv to a human Fc region was included for comparison. A2M-Atezolizumab in both conformations and the Atezolizumab-hFc all produced a concentration-dependent luminescence response (Figure 5). The maximum responses of methylamine-treated A2M-Atezolizumab and Atezolizumab-hFc were similar, whereas a saturated response for native A2M-Atezolizumab was not reached at the highest measured concentration of 200 nM (and its maximum response was therefore assumed to be the same as for methylamine-treated A2M- Atezolizumab). Both methylamine-treated A2M-Atezolizumab and Atezolizumab-hFc demonstrated sub-nanomolar EC50 values of 400 and 80 pM, respectively, whereas native A2M-Atezolizumab had an EC50 value of 216 nM. There was therefore an approximately 500-fold difference in the activity of A2M-Atezolizumab in its native and activated conformations.Conclusion

[0346] A2M-Atezolizumab demonstrated a conformation-dependent ability to block PD-L1 and restore NFKB signaling in PD-1+T cells in a cellular assay of immune checkpoint blockade. This demonstrates that A2M-Atezolizumab shows conformation-dependent binding to cell surface PD-L1 and that it retains the PD-L1 -blocking functionality of the parent Atezolizumab antibody.Example 6 -Modification of the A2M bait region to target specific proteasesAim

[0347] The sequence of A2M’s bait region determines whether it can be cleaved by a given protease, and thereby determines which proteases are able to activate A2M (and be trapped by A2M). The bait region of wildtype A2M can be cleaved by almost all human proteases and it would be advantageous to restrict the bait region’s cleavage to designated proteases, to more specifically target diseased tissues. We first investigated whether the bait region can be replaced with a minimal sequence that does not contain cleavage sites for the majority of human proteases. We then investigated whether protease cleavage sites could be re-introduced into this minimal sequence, with the intent of producing bait region sequences with improved specificity for a single protease or family of proteases (in this example, matrix metalloproteases (MMPs)).Materials and methodsProteins used

[0348] A2M proteins with modified bait region sequences were expressed in HEK293F cells and purified as described for A2M-antibodies in Example 2. The amino acid sequences of these A2M proteins are given in SEQ ID NO: 65-73.

[0349] N-terminally StrepII-tagged proMMP2 (uniprot ID P08253, SEQ ID NO: 61-62) was expressed and purified using StrepTactin affinity chromatography and size exclusion chromatography, as described for StrepII-tagged hFc fusion proteins in Example 3. ProMMP2 was activated using 1 mM APMA by incubating for 15 minutes at 37 °C, followed by desalting into HBS with 10 mM CaC12 using a PD-10 column (GE Healthcare).SDS-PAGE and pore limited native page

[0350] Native pore limited PAGE was performed as previously described (36), using homemade gels in TBE buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA) with an acrylamide gradient of 5-10% for A2M analysis and 10-15% for C3 analysis. Pore limited electrophoresis gels were run overnight at 100 V in TBE buffer.Denaturing SDS-PAGE was performed using the discontinuous 2-amino-2 -methyl- 1,3 -propanediol and glycine buffer system on homemade 5-15% acrylamide gradient gels (37). Samples were reduced with 25 mM DTT at 95 °C for 5 minutes.Reaction of A2M with methylamine and proteases

[0351] To amino lyze A2M’s thiol ester, methylamine (pH 8) was added to 250 mM and incubated for at least 45 minutes at 37 °C. To assess the cleavage of A2M by trypsin and LysC, proteases were added to a 2.2: 1 mol / mol ratio of protease:A2M and incubated for five minutes at 37 °C. The digestion was then inhibited using the serine protease inhibitor PMSF (2 mM, 15 minutes, room temperature). To assess the cleavage of A2M by MMP2, MMP2 was added to A2M in HBS with 10 mM CaC12 to a 6: 1 mol / mol ratio of MMP2:A2M, incubated for 15 minutes at 37 °C, and then inhibited using 20 mM EDTA. When cleaving A2M using other human proteases, incubation lasted one hour at 37 °C in HBS with 10 mM CaC12, and PMSF or EDTA was used to inhibit serine proteases and metalloproteases, respectively.Determining A2M’s inhibition of protein substrate cleavage byMMP2

[0352] The inhibition of MMP2 by A2M was investigated using a fluorescently labelled gelatin substrate. 1.4 pmol (7.5 nM) of MMP2 was reacted with with 0-2.7 pmol (0-15 nM) of A2M in 50 mM HEPES, 100 mM NaCl, 5 mM CaCF pH 8 for 15 min at 37 °C. DQ Gelatin From Pig Skin (Invitrogen) was added to a final concentration of 0.1 mg / ml. The fluorescence (excitation at 485 nm and emission at 520 nm) of the unquenched digestion products of DQ gelatin after 10 min at 37 °C were measured in a FLUOstar Omega plate reader (BMG LABTECH). All reactions were performed in triplicates.ResultsBait region substitution with 13 Gly-Gly-Ser triplets produces A2M that is tetrameric, native, and inducible.

[0353] To remove essentially all protease cleavage sites from the bait region and determine the extent to which it tolerates modification, we replaced the 39-residue wildtype A2M bait region sequence with 13 Gly-Gly-Ser repeats, chosen for their solubility and low susceptibility to proteolysis (Figure 6A). The resulting “tabula rasa” (TR) bait region was incorporated into recombinant A2M, resulting in A2M that was predominantly tetrameric and in its native conformation as assessed by native PAGE, with an intact thiol ester apparent from the formation of characteristic heat-induced autolysis products in SDS- PAGE (Figure 6B-C). Upon aminolysis of its thiol ester with methylamine, TR A2M underwent a conformational collapse indistinguishable from that of wildtype A2M, as determined by pore-limited native PAGE; however, TR A2M was not cleaved in its bait region by either trypsin or LysC, and remained in its native conformation despite proteolysis by these proteases outside of its bait region (Figure 6B-C). If a lysine residue was introduced into the TR bait region at position 704, the resulting bait region could be cleaved by both trypsin and LysC, resulting in protease conjugation and A2M’s characteristic conformational change (Figure 6A-C).Identification of an MMP2-cleavable bait region sequence with improved selectivity.

[0354] Four TR bait regions incorporating substrate sequences for human MMP2 were designed (Figure 7A). All four TR-based MMP2 substrate bait regions and the wildtype bait region were cleaved by MMP2, but not the initial TR A2M (Figure 7B-D). Incomplete bait region cleavage and intermediate electrophoretic mobility of A2M:MMP2 complexes was observed both for wildtype A2M and the four MMP2 substrate TR A2Ms (Figure 7C-D).

[0355] We then tested whether the four MMP2 substrate bait regions were cleaved by nine additional human proteases (plasmin, cathepsin G, MMP1, MMP3, MMP8, MMP13, ADAMTS4, ADAMTS5, and ADAMTS13) using reducing SDS-PAGE to assess bait region cleavage and the formation of high-MW conjugation products (Figure 7B). All assessed proteases except ADAMTS13 (which is highly specific towards von Willebrand factor) were able to cleave wildtype A2M, while none were able to cleave TR A2M (Figure 7B). The incorporation of any of the four MMP2 substrate sequences into the TR bait region conveyed cleavage by all tested MMPs, whereas they were differently cleaved by non-MMP proteases; for example, the C9 substrate was the only one containing an arginine residue and was the only A2M to be cleaved by plasmin (Figure 7A-B). The SI substrate was only cleaved by MMPs (Figure 7A-B), and the A2M TR SI protein was therefore selected for further optimization as an A2M with an improved MMP specificity relative to wildtype A2M.The native content of tabula rasa-based A2Ms is improved by shortening the bait region by seven residues or restoring the 10 C-terminal wildtype residues.

[0356] The initial TR A2M proteins were expressed with an increased amount of non-native A2M compared to wildtype A2M (Figure 6A, 7C). To resolve this issue, we tested two altered tabula rasa bait regions where the first, TRA7, was shortened by 7 residues to a total length of 32 residues and the second, TR QRT4, re-introduced the C-terminal quarter of the wildtype bait region (Figure 8A). Both TRA7 and TR QRT4 improved the native content of their resulting A2Ms to that of wildtype A2M (Figure 8B). A position of the SI substrate sequence in the TRA7 bait region was identified that conveyed an efficiency of MMP2 inhibition that was indistinguishable from that of wildtype A2M (Figure 8C), showing that this shortened bait region can produce fully functionally A2M.Conclusion

[0357] The bait region of A2M could be completely replaced by glycine and serine residues without compromising the structure and function of A2M, although a glycine-serine bait region that was shortened to 32 residues was found to give an improved yield of native A2M. The glycine-serine bait region was not cleavable by 10 tested human proteases. Upon incorporation of the SI substrate for MMP2 into the bait region, 5 human MMPs were able to cleave the bait region, while 5 non-MMPs remained unable to cleave. This demonstrates that the glycine-serine bait region can be used as the foundation for making bait regions with an improved specificity to a protease or family of proteases (such as MMPs).Example 7 -Bait region modification of A2M-antibodiesAim of study

[0358] In Example 6, bait region sequences based on the “tabular rasa” (TR) bait region which replaces the wildtype bait region with glycine and serine residues were found to produce A2M proteins which were more specifically cleaved and activated by target proteases. Furthermore, a TR bait region that was shortened by 7 residues (TRA7) to a length of 32 residues was found to convey an increased yield of native A2M, and the placement of the SI substrate for MMP2 into TRA7 at a specific position (TRA7 SI 1703) was found to convey an inhibition of MMP2 that was equivalent to that of the wildtype A2M bait region. Here, we investigated whether A2M-antibodies incorporating TR bait regions with MMP2 substrate sites could be activated by MMP2 in the same manner as A2M-antibodies with wildtype bait regions.Materials and methodsProteins used

[0359] A2M-Atezolizumab with the wildtype bait region (SEQ ID NO: 7-8), the TRA7 SI 1703 bait region (SEQ ID NO: 74-75), or the TRA7 SI 1703 P704 bait region (SEQ ID NO: 76-77) were expressed in HEK293F cells and purified as described for A2M-antibodies in Example 2.

[0360] ProMMP2 was expressed, purified, and activated as described in Example 6.Cleavage of A2M by proteases

[0361] To cleave A2M-antibodies with MMP2, MMP2 was added in HBS with 10 mM CaC12 to a 4: 1 mol / mol ratio of MMP2:A2M, incubated for 15 minutes at 37 °C, and then inhibited using 20 mM EDTA. To cleave A2M-antibodies with thermlysin, thermolysin was added in HBS with 10 mM CaC12 to a 2.2: 1 mol / mol ratio of thermolysin:A2M, incubated for 2 minutes at 37 °C, and then inhibited using 20 mM EDTA.Biolayer interferometry

[0362] Biolayer interferometry was used to investigate the interaction between A2M-Atezolizumab with different bait regions using the method described in Example 3.SDS-PAGE and pore limited native PAGE

[0363] Native pore limited PAGE was performed as previously described (36), using homemade gels in TBE buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA) with an acrylamide gradient of 5-10% for A2M analysis and 10-15% for C3 analysis. Pore limited electrophoresis gels were run overnight at 100 V in TBE buffer.

[0364] Denaturing SDS-PAGE was performed using the discontinuous 2 -amino -2 -methyl -1,3- propanediol and glycine buffer system on homemade 5-15% acrylamide gradient gels (37). Samples were reduced with 25 mM DTT at 95 °C for 5 minutes.Results

[0365] To assess the functionality of A2M-antibodies with engineered bait regions, A2M- Atezolizumab was expressed with either a wildtype bait region, the TRA7 SI 1703 bait region (an optimized MMP2-substrate bait region described in Example 6), or the TRA7 SI 1703 P704 bait region which minimizes the MMP2 cleavage site and prevents cleavage of residue 1703 by serine proteases by adding a P’ l-position proline residue (Figure 9A). All three A2M-Atezolizumab proteins were bait region cleaved when treated with MMP2, as assessed by their conformational change in native PAGE (Figure 9B) and cleavage of the A2M subunit in reducing SDS-PAGE (Figure 9C). These results show that A2M-antibodies with engineered MMP2 -substrate bait regions are cleaved in their bait regions by MMP2.

[0366] Biolayer interferometry was used to investigate the effect of MMP2 cleavage on the A2M- Atezolizumab proteins’ binding to immobilized PD-L1. MMP2 cleavage was found to convey a similar antigen binding to that induced by thermolysin cleavage, for A2M-Atezolizumab with a wildtype bait region (Figure 9D). Both A2M-Atezolizumab proteins with engineered bait regions showed similar antigen binding upon their cleavage with MMP2 (Figure 9D). These results show that bait region cleavage by MMP2 induces antigen binding in A2M-antibodies, both with wildtype bait regions and with engineered MMP2 -substrate bait regions.Conclusion

[0367] Engineered bait regions, as described in Example 5, can be incorporated into A2M-antibodies without disrupting their conformationally dependent antigen binding, and are still preferentially cleaved by target proteases such as MMP2. MMP2 cleavage is able to induce antigen binding in A2M-antibodies with wildtype bait regions or engineered bait regions.Example 8 - Incorporation of the extracellular region of PD1 receptor into A2MAim of study

[0368] Here, we investigate whether the extracellular region of the human PD1 receptor can be incorporated into A2M (in the same manner as antibodies, as shown in previous examples) and whether the resulting A2M-PD1 fusion protein binds to the PD1 receptor’s ligand, PD-L1, in a manner that is dependent on the conformation of A2M.Material and methodsProteins used

[0369] The A2M-PD1 fusion construct was expressed and purified as described for A2M-antibodies in Example 1. The extracellular region of PD 1 that was incorporated into A2M was the same sequence used for testing A2M-nivolumab in Example 2, i.e. uniprot ID Q15116, residues 26-150 with a Cys93Ser mutation. The amino acid and nucleotide sequence of A2M-PD1 is given in SEQ ID NO: 25-26. PD-L1 fused to a human Fc region was prepared as described in Example 2.SDS-PAGE and pore limited native PAGE

[0370] A2M-PD1 was analyzed by reducing SDS-PAGE using the protocol described in Example 2.Reaction of A2M-antibodies with methylamine and proteases

[0371] A2M-PD1 was treated with methylamine or the metalloprotease thermolysin in order to change its conformation, as described in Example 3.Biolayer interferometry

[0372] Biolayer interferometry was used to investigate the binding of A2M-PD1 in its untreated, methylamine-, or thermolysin -treated conformations to PD-L1 immobilized on the surface of biosensors, as described in Example 3.Results

[0373] Using the same fusion strategy that was used to incorporate antibody scFvs and nanobodies into A2M, the extracellular region of human PD1 was incorporated into A2M and the resulting A2M-PD1 was expressed and purified using standard A2M protocols (figure 10A).

[0374] The conformational dependence of PD-L1 binding by A2M-PD1 was then assessed by biolayer interferometry. PD-L1 binding by A2M-PD1 was strongly dependent on the conformation of A2M (figure 10B). A control sample with untreated A2M-PD1 showed a minor degree of binding, as native A2M-PD1 was not purified from non-native A2M-PD1 prior to this experiment, e.g. using LRPl-based depletion as described in Example 3. In contrast, both methylamine- and thermolysin -treated A2M showed greatly enhanced binding responses, and these responses were very similar to each other. When A2M-PD 1 in its native conformation was enriched by three rounds of LRP 1 -based depletion of nonnative A2M-PD1, no binding by untreated A2M-PD1 was detectable.Conclusion

[0375] PD1 could be incorporated into A2M, resulting in functional A2M that was capable of undergoing its typical methylamine- and thermolysin-induced conformational changes and PD1 that was capable of binding to its ligand, PD-L1. Furthermore, the binding of PD1 to PD-L1 was dependent on the conformation of A2M, with no detectable binding of A2M-PD1 in its native conformation to PD- Ll.Example 9 - Incorporation of theIL2 cytokine into A2MAim

[0376] Here, we investigate whether the IL2 cytokine can be incorporated into A2M (in the same manner as antibodies, as shown in previous examples) and whether the resulting A2M-IL2 fusion protein binds to an IL 2 receptor, IL-2Ra, in a manner that is dependent on the conformation of A2M.Material and methodsProteins used

[0377] The A2M-IL2 fusion construct was expressed and purified as described for A2M-antibodies in Example 1. The IL2 cytokine that was incorporated into A2M used the wildtype human sequence (uniprot P60568, residues 21-153). The amino acid and nucleotide sequence of A2M-IL2 is given in SEQ ID NO: 23-24.

[0378] The extracellular region of the IL2 receptor IL-2Ra (uniprot P01589, residues 22-238, with a Cys213Ala mutation, SEQ ID NO: 43) was expressed as a Strep-Tagged human Fc fusion protein (SEQ ID NO: 57-58), as described for other antigens in Example 2, and purified in the same manner as well.SDS-PAGE and pore limited native PAGE

[0379] A2M-IL2 was analyzed by reducing SDS-PAGE using the protocol described in Example 2.Reaction of A2M-antibodies with methylamine and proteases

[0380] A2M-IL2 was treated with methylamine or the metalloprotease thermolysin in order to change its conformation, as described in Example 3.Biolayer interferometry

[0381] Biolayer interferometry was used to investigate the binding of A2M-IL2 in its untreated, methylamine-, or thermolysin-treated conformations to IL-2Ra immobilized on the surface of biosensors, as described in Example 3.Results

[0382] Using the same fusion strategy that was used to incorporate antibody scFvs and nanobodies into A2M, the human cytokine IL2 was incorporated into A2M and the resulting A2M-IL2 was expressed and purified using standard A2M protocols (Figure 11A).

[0383] The conformational dependence of IL-2Ra binding by A2M-IL2 was then assessed by biolayer interferometry. IL-2Ra binding by A2M-IL2 was dependent on the conformation of A2M (Figure 11B). A control sample with untreated A2M-IL2 showed a moderate degree of binding, even after using LRP1- based depletion of non-native A2M-IL2. Cleavage of the A2M bait region with thermolysin conferred an intermediately increased rate of association and saturation of binding, whereas aminolysis of the A2M thiol ester with methylamine conferred an even greater increase in both association rate and binding saturation. The degree of binding (as determined by kobs values) is approximately 10-fold increased by methylamine treatment, compared to A2M-IL2 in its native conformation.Conclusion

[0384] IL2 could be incorporated into A2M, resulting in functional A2M-IL2 that underwent its typical methylamine- and thermolysin-induced conformational changes and IL2 that was capable of binding to the receptor IL-2Ra. Furthermore, this receptor binding by A2M-IL2 was dependent on the conformation of A2M, with increased receptor binding observed upon collapse of the A2M conformation by bait region cleavage or thiol ester aminolysis.Example 10 - Investigating other fusion strategies for the incorporation of biopharmaceutical moieties into A2M.Aim

[0385] The previous examples investigating proteinaceous fusion constructs of A2M and biopharmaceutical moieties use the ciRBD approach, where the biopharmaceutical moieties are inserted between A2M residues 1402 and 1403. Here, we investigate whether target binding that is dependent on the conformation of A2M can be achieved by four other approaches: fusion, iRBD, miRBD, and tRBD.Materials and methodsProteins used

[0386] All fusion constructs of A2M and biopharmaceutical moieties were expressed and purified as described for A2M-antibody fusion constructs in Example 2. No depletion of non-native A2M was performed. The amino acid and nucleotide sequences of A2M-fusion-EgAl (SEQ ID NO: 94-95), A2M- iRBD-EgAl (SEQ ID NO: 84-85), A2M-miRBD-EgAl (SEQ ID NO: 86-87), A2M-miRBD- Atezolizumab (SEQ ID NO: 88-89), A2M-miRBD-KN035 (SEQ ID NO: 90-91), and A2M-tRBD-EgAl (SEQ ID NO: 92-93) are given. EGFR and PD-L1 in fusion with a human FC region (SEQ ID NO: 45- 48) were produced as described in Example 3.Reaction of A2M-antibodies with methylamine and proteases

[0387] A2M-antibodies were reacted with methylamine or proteases as described in Example 3.Biolayer interferometry

[0388] Biolayer interferometry was performed as described in example 3.Results

[0389] To determine whether shielding of biopharmaceutical moieties in A2M could be accomplished by their incorporation into A2M in other ways than the ciRBD approach used previously, four new fusion approaches were tested. In the first, the EgAl nanobody was expressed immediately following the C-terminus of A2M’s RBD domain to produce the A2M-fusion-EgAl protein. A2M-fusion-EgAl did not demonstrate conformational dependence of EgAl’s binding to EGFR (Figure 12A), indicating that not all positions adjacent to the RBD domain are shielded in the native conformation of A2M.

[0390] In the second approach, iRBD, the EgAl nanobody was inserted into the RBD domain replacing A2M residues 1392-1403. The resulting A2M-iRBD-EgAl protein showed a high degree of conformational dependence (Figure 12B), comparable to that of the ciRBD approach (see Example 3). This indicates that the region of A2M’s RBD domain adjacent to residues 1393-1403 is an appropriate site for the incorporation of biopharmaceutical moieties to achieve conformational dependence. Accordingly, the third approach, miRBD, also achieved conformational dependence by incorporation of either the EgAl nanobody, the KN035 nanobody, or the Atezolizumab scFv at a position replacing A2M residues 1393-1395 (Figure 12B-E).

[0391] In the fourth approach, tRBD, coiled-coil interactions were used to bring an incorporated biopharmaceutical moiety (EgAl) into proximity of the RBD residues 1393-1403. The EgAl nanobody was incorporated at a position C-terminal to the RBD domain with an alpha-helix sequence designed to be complementary to A2M residues 1393-1403 at its N-terminus. The alpha-helix sequence is attached to the RBD domain with a 15 -residue linker, in order to permit the alpha-helix to interact with residues 1393-1403. Furthermore, modifications to A2M residues 1393-1403 were made to enhance the designed complementary coiled-coil interactions. The resulting A2M-tRBD-EgAl protein demonstratedconformational dependence of the EgAl / EGFR interaction (Figure 12F), indicating that the coiled-coil interactions were able to bring the EgAl nanobody into proximity with residues 1393-1403 and that this proximity conveyed at least partial shielding of the EgAl nanobody in the native conformation of A2M.Conclusion

[0392] Investigations of the fusion, iRBD, miRBD, and tRBD approaches to producing fusion constructs of A2M and biopharmaceutical moieties showed that positions that are proximal to A2M RBD residues 1393-1403 due to either direct fusion at this site (as seen in the ciRBD, iRBD, and ciRBD approaches) or localization of the moiety to this position through other means (e.g. through coiled-coil interactions, as demonstrated by A2M-tRBD-EgAl) convey conformationally dependent target binding for many different tested biopharmaceutical moieties (16 in total, considering all ciRBD, iRBD, miRBD, and tRBD fusion constructs).Example 11 - Study of insertion sites in the RBD regionAim

[0393] To identify sites for drug insertion in the RBD of A2M which will convey conformationdependent accessibility.Materials and methods

[0394] Figures were prepared using the PyMol Molecular Graphics System software (version 2.3.0).Results

[0395] In the iRBD, miRBD, and ciRBD approaches to creating A2M-based prodrugs, residues 1392 to 1403 of A2M’s RBD domain are either replaced with the drug sequence (as well as N- and C-terminal linkers) or the drug is inserted between residues 1402-1403 without altering any residues of A2M. Residues 1391-1405 or 1392-1404 comprise a loop or linker region between strands of beta-sheet structure, and such loops are well-suited for modification, in contrast to the beta-sheet sequence where modifications are more likely to affect the folding of the domain.

[0396] Furthermore, the orientation of the loop is also critical to achieving a conformationally dependent drug position, as fusion of a drug at the opposite side of the RBD at position 1474 (in the A2M-fusion-EgAl construct) produced an always-accessible drug.

[0397] After a structural evaluation of the RBD domain, three additional loop regions were identified that are suitable for drug insertion, namely the regions comprising residues 1368-1379 (loop 1), 1420- 1426 (loop 3), and 1450-1457 (loop 4), in addition to the empirically tested region comprising residues 1392-1404 (loop 2) (Figure 13). All three loops extend between beta-strands and are oriented in a similar direction to loop 2 (1392-1404), where they face inwards towards the interior of the A2M tetramer.Conclusion

[0398] In addition to the region comprising residues 1392-1404 (loop 2), three additional loops comprising residues 1368-1379 (loop 1), 1420-1426 (loop 3), and 1450-1457 (loop 4) of A2M were identified that are considered usable for replacement by or direct insertion of one or more drugs, in order to convey conformational -dependent binding of their therapeutic target.Example 12 - Production of furin-cleaved A2M with a releasable MG 8 domainAim of study

[0399] We investigated whether the introduction of a furin cleavage site between the CUB and MG8 (aka. RBD) domains of A2M would allow the production of a native A2M protein with a MG8 domain that is released upon A2M’s conformational change.Materials and methodsExpression and purification of A2M-antibody fusion constructs

[0400] A2M with a furin cleavage site (A2M furinRBD) was expressed in HEK293 FreeStyle cells using a standard transient transfection protocol. Briefly, 25 kDa linear polyethyleneimine (Polysciences) and plasmid DNA were incubated for 10 min in antibiotic-free FreeStyle medium (Thermo Fisher Scientific) at a 4: 1 w / w PEI:DNA ratio, then slowly dripped into a culture of cells at a density of 1 million cells per mb, to a final DNA concentration of 1 pg per mb culture. After 5 days, the supernatant was harvested by spinning down the cells at 1500 x g.

[0401] Purification of the constructs was performed using an established protocol for purifying A2M(1, 2). Supernatants were first run through a Zn2+-loaded Chelating HiTrap column (GE Healthcare) and eluted with 50 mM EDTA, 150 mM NaCl, 100 mM sodium acetate, pH 7.4. The EDTA eluate was dialyzed against 20 mM HEPES at pH 7.4, then loaded onto a HiTrap Q column (GE Healthcare) and eluted by a gradient of 0-400 mM NaCl (with a constant 20 mM HEPES at pH 7.4). Fractions containing A2M were pooled, concentrated by ultrafiltration, and purified by size exclusion chromatography on a Sephacryl S-300 HR (GE Healthcare), using a 20 mM HEPES, 150 mM NaCl, pH 7.4 running buffer (HEPES-buffered saline, HBS).SDS-PAGE and pore limited native PAGE

[0402] Native pore limited PAGE was performed as previously described (3), using homemade gels in TBE buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA) with an acrylamide gradient of 5-10%. Pore limited electrophoresis gels were run overnight at 100 V in TBE buffer.

[0403] Denaturing SDS-PAGE was performed using the discontinuous 2 -amino -2 -methyl -1,3- propanediol and glycine buffer system on homemade 5-15% acrylamide gradient gels (4). Samples were reduced with 25 mM DTT at 95 °C for 5 minutes.Reaction of A2Mwith methylamine and proteases

[0404] To amino lyze A2M’s thiol ester, methylamine (pH 8) was added to 250 mM and incubated for at 2 hours at 37 °C. To assess the cleavage of A2M by thermolysin, thermolysin was added to a 2.2: 1 mol / mol ratio of protease:A2M and incubated for five minutes at 37 °C. The digestion was then inhibited using EDTA (10 mM, 15 minutes, room temperature).Size exclusion chromatography

[0405] Analytic size exclusion chromatography was performed using a Superdex 200 Increase (Cytiva Life Sciences) with 20 mM HEPES, 150 mM NaCl, pH 7.4 running buffer and a flow rate of 0.4 mL per minute.Results

[0406] Previous studies have suggested that the MG8 domain (aka. the RBD) of many members of the A2M family forms strong non-covalent interactions with other domains (such as the thiol ester domain) while A2M proteins are in their native conformation, whereas these interactions are no longer present after the aminolysis- or proteolysis-induced conformational change and the MG8 domain remains attached mostly due to the polypeptide backbone. We hypothesized that a furin cleavage could be introduced between the MG8 domain and the domain at its N-terminus (the CUB domain), resulting in the release of the MG8 domain following A2M’s conformational change. Furin is an intracellular protease that processes many immature proteins in the endoplasmatic reticulum: it is highly specific for R / K-X-X-R / K sequences, where X can be any amino acid but is ideally also arginine or lysine (5).

[0407] We produced an A2M (A2M furinRBD) incorporating an ideal furin cleavage site, RRRR, as well as short linkers to provide furin with full access to the cleavage site. A2M furinRBD’s sequence is given in SEQ ID NO: 227 - A2M_furinRBD. The majority of A2M furinRBD was expressed and purified in a native conformation with an intact thiol ester, although a significant population (roughly 25%) failed to form a native conformation and was not cleavable by thermolysin (Figure 15A-B). A2M furinRBD demonstrated a heterogeneous untreated sample in native PAGE, but collapsed normally upon methylamine or thermolysin treatment (Figure 15B). The native gel only shows migration of the larger proteins, thus the MG8 domain released by Thermolysin is not visible on the native gel (figure 2B)

[0408] To investigate whether the MG8 domain of A2M furinRBD was released by A2M’s conformational change, purified A2M furinRBD was separated by size exclusion chromatography, either without any treatment or after methylamine treatment (Figure 16A). After methylamine treatment, the MG8 domain was separated from the remainder of A2M and eluted in its own fraction, whereas in the untreated A2M, the MG8 domain co-eluted with the main A2M peak (Figure 16B). This shows that the MG8 domain is released by the methylamine-induced conformational change of A2M.Conclusion

[0409] Incorporation of a furin cleavage site at the N-terminus of the MG8 domain in recombinant A2M resulted in a mostly native A2M protein that was completely furin processed and could release its MG8 domain following its conformational change.Example 13 - Production of TEV-protease-cleaved A2M with a releasable MG 8 domainAim of study

[0410] We investigated whether the introduction of a tobacco etch (TEV) protease cleavage site between the CUB and MG8 (aka. RBD) domains of A2M would allow the production of a native A2M protein with a MG8 domain that is released upon A2M’s conformational change.Materials and methodsExpression and purification of A2M-antibody fusion constructs

[0411] A2M with a TEV protease cleavage site (tevRBD) was expressed in HEK293 FreeStyle cells using a standard transient transfection protocol. Briefly, 25 kDa linear polyethyleneimine (Polysciences) and plasmid DNA were incubated for 10 min in antibiotic-free FreeStyle medium (Thermo Fisher Scientific) at a 4: 1 w / w PEEDNA ratio, then slowly dripped into a culture of cells at a density of 1 million cells per mb, to a final DNA concentration of 1 pg per mb culture. After 5 days, the supernatant was harvested by spinning down the cells at 1500 x g.

[0412] Purification of the constructs was performed using an established protocol for purifying A2M(1, 2). Supernatants were first run through a Zn2+-loaded Chelating HiTrap column (GE Healthcare) and eluted with 50 mM EDTA, 150 mM NaCl, 100 mM sodium acetate, pH 7.4. The EDTA eluate was dialyzed against 20 mM HEPES at pH 7.4, then loaded onto a HiTrap Q column (GE Healthcare) and eluted by a gradient of 0-400 mM NaCl (with a constant 20 mM HEPES at pH 7.4). Fractions containing A2M were pooled, concentrated by ultrafiltration, and purified by size exclusion chromatography on a Sephacryl S-300 HR (GE Healthcare), using a 20 mM HEPES, 150 mM NaCl, pH 7.4 running buffer (HEPES-buffered saline, HBS).SDS-PAGE and pore limited native PAGE

[0413] Native pore limited PAGE was performed as previously described (36), using homemade gels in TBE buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA) with an acrylamide gradient of 5-10% for A2M analysis. Pore limited electrophoresis gels were run overnight at 100 V in TBE buffer.

[0414] Denaturing SDS-PAGE was performed using the discontinuous 2 -amino -2 -methyl -1,3- propanediol and glycine buffer system on homemade 5-15% acrylamide gradient gels (37). Samples were reduced with 25 mM DTT at 95 °C for 5 minutes.Reaction of A2M with methylamine and proteases

[0415] To amino lyze A2M’s thiol ester, methylamine (pH 8) was added to 250 mM and incubated for at 2 hours at 37 °C. To assess the cleavage of A2M by thermolysin, thermolysin was added to a 2.2: 1mol / mol ratio of protease:A2M and incubated for five minutes at 37 °C. The digestion was then inhibited using EDTA (10 mM, 15 minutes, room temperature). TEV protease cleavage of A2M was performed at room temperature for 24 hours, using a TEV protease stock that was desalted into HBS and did not contain any reducing agent.Size exclusion chromatography

[0416] Analytic size exclusion chromatography was performed using a Superdex 200 Increase (Cytiva Life Sciences) with 20 mM HEPES, 150 mM NaCl, pH 7.4 running buffer and a flow rate of 0.4 mL per minute.Results

[0417] Furin cleavage at the N-terminus of A2M’s MG8 domain functioned as intended, producing a fully cleaved A2M construct that mostly maintained its native conformation and thiol ester, and released its MG8 domain during the conformational change. However, intracellular furin processing cannot easily be controlled and regulated, and heterogeneous A2M furinRBD was produced. This suggests that furin processing during the translation and maturation of A2M is disruptive to the formation of its thiol ester and native conformation. Thus, it could potentially be advantageous to delay the cleavage at the N-terminus of the MG8 domain until after A2M is fully folded and assembled into a native tetramer, e.g. to when it is secreted into the extracellular space and purified. Therefore, we tested whether a cleavage site for the highly specific TEV protease could be used instead of a furin site.

[0418] We produced an A2M (A2M tevRBD) incorporating the TEV cleavage site ENLYFQS (SEQ ID NO: 226) as well as short linkers to provide full access to the cleavage site. A2M tevRBD’s sequence is given in SEQ ID NO: 229 - A2M_tevRBD. A2M tevRBD was expressed and purified as a homogeneously native tetrameric A2M with an intact thiol ester (Figure 17B-C). A2M tevRBD could be cleaved by TEV protease at the expected site, causing the MG8 domain to be noncovalently associated with the remainder of A2M (Figure 17A). After TEV cleavage, A2M tevRBD had a smeared migration pattern in native PAGE, but underwent the normal conformation changes upon methylamine or protease treatment (Figure 17C).

[0419] To investigate whether the MG8 domain of A2M tevRBD was released by A2M’s conformational change, TEV protease-cleaved A2M tevRBD (where the TEV protease was removed by a previous purification step) was separated by size exclusion chromatography, either without any treatment or after methylamine treatment (Figure 18A). After methylamine treatment, the MG8 domain was separated from the remainder of A2M and eluted in two conformations corresponding to two different glycosylation patterns (Figure 18B-C). In the untreated A2M tevRBD, the MG8 domain coeluted with the main A2M peak (Figure 18A-B).

[0420] This shows that the MG8 domain is released by the methylamine-induced conformational change of A2M after cleavage of the TEV protease site.Conclusion

[0421] A TEV protease cleavage site could be used to generate an A2M construct with a MG8 domain that can be released by A2M’s conformational change. Furthermore, the TEV protease site demonstrates a more homogeneous native A2M product compared to the use of the furin cleavage site.Example 14 -Addition of two antibodies to TEV-protease-cleavable A2MAim of study

[0422] We investigated two antibody-derived binding fragments (an scFv and a nanobody) could be simultaneously fused into the A2M tevRBD protein described in Example 13 while still yielding an initially native A2M, and whether TEV protease cleavage of such a protein allowed the release of the MG8 domain and the two antibodies which are fused to it.Materials and methodsExpression and purification of A2M-antibody fusion constructs

[0423] A2M with a TEV protease cleavage site, a C-terminal nanobody specific to EGFR (EgAl), and an anti-CD3 scFv (Foralumab-derived) fused by the ciRBD approach was expressed in HEK293 FreeStyle cells using a standard transient transfection protocol. A similar construct with a modified TEV protease cleavage site, a C-terminal nanobody specific to EGFR (EgAl), and an anti-CD3 scFv (UCHT1 -derived) was also used in this example and expressed / purified in an identical manner. Briefly, 25 kDa linear polyethyleneimine (Polysciences) and plasmid DNA were incubated for 10 min in antibiotic-free FreeStyle medium (Thermo Fisher Scientific) at a 4: 1 w / w PEEDNA ratio, then slowly dripped into a culture of cells at a density of 1 million cells per mb, to a final DNA concentration of 1 pg per mb culture. After 5 days, the supernatant was harvested by spinning down the cells at 1500 x g.

[0424] Purification of the constructs was performed using an established protocol for purifying A2M(1, 2). Supernatants were first run through a Zn2+-loaded Chelating HiTrap column (GE Healthcare) and eluted with 50 mM EDTA, 150 mM NaCl, 100 mM sodium acetate, pH 7.4. The EDTA eluate was dialyzed against 20 mM HEPES at pH 7.4, then loaded onto a HiTrap Q column (GE Healthcare) and eluted by a gradient of 0-400 mM NaCl (with a constant 20 mM HEPES at pH 7.4). Fractions containing A2M were pooled, concentrated by ultrafiltration, and purified by size exclusion chromatography on a Sephacryl S-300 HR (GE Healthcare), using a 20 mM HEPES, 150 mM NaCl, pH 7.4 running buffer (HEPES-buffered saline, HBS).SDS-PAGE and pore limited native PAGE

[0425] Native pore limited PAGE was performed as previously described (36), using homemade gels in TBE buffer (89 mM Tris, 89 mM boric acid, 2 mM EDTA) with an acrylamide gradient of 5-10% for A2M analysis. Pore limited electrophoresis gels were run overnight at 100 V in TBE buffer.

[0426] Denaturing SDS-PAGE was performed using the discontinuous 2 -amino -2 -methyl -1,3- propanediol and glycine buffer system on homemade 5-15% acrylamide gradient gels (37). Samples were reduced with 25 mM DTT at 95 °C for 5 minutes.Reaction of A2M with methylamine and proteases

[0427] To amino lyze A2M’s thiol ester, methylamine (pH 8) was added to 250 mM and incubated for at 2 hours at 37 °C. To assess the cleavage of A2M by thermolysin, thermolysin was added to a 2.2: 1 mol / mol ratio of protease:A2M and incubated for five minutes at 37 °C. The digestion was then inhibited using EDTA (10 mM, 15 minutes, room temperature).Results

[0428] We have previously shown (Examples 2-4) that antibody-derived binding fragments can be fused into an internal position in the MG8 domain (e.g. between residues 1402 and 1403 in the ciRBD approach) to produce prodrug versions of that antibody that require proteolytic activation before the antibody can access its antigen. We have also shown that an antibody can be fused to the C-terminus of the MG8 domain, at which position it can access its antigen independent of the conformational state of A2M (in Example 10 and figure 12A). Here, we investigated whether the fusion of two different antibody fragments at both of these locations, along with the insertion of the TEV protease site at the N-terminus of the MG8 as tested in the tevRBD construct from Example 13, could produce a native A2M protein where the MG8 domain remained releasable by A2M’s conformational change.

[0429] We produced an A2M with the tevRBD sequence combined with an anti-CD3 scFv (derived from Foralumab) in the ciRBD position and an anti-EGFR nanobody (clone EgAl) in the C-terminal position, which we refer to as tevRBD+2xAb in this example. Both of these antibody fragments’ ability to bind their antigens when fused into A2M in these specific sites have been demonstrated previously, in Examples 2-4 and 13. The sequence of tevRBD+2xAb is given in SEQ ID NO: 231 - A2M_tevRBD+2xAb. This protein was produced in a native conformation with an intact thiol ester, as determined by its autolytic product bands in SDS-PAGE (Figure 19A). It assembled into a tetramer with a larger electrophoretic profile than that of wildtype A2M, due to the C-terminal nanobodies which are located outside of A2M (Figure 19B). The conformational change upon methylamine treatment was not apparent in native PAGE (Figure 19B). Thermolysin was able to cleave within the TEV protease site, which contains several hydrophobic residues that are thermolysin substrates (Figure 19A-B). Thus, thermolysin-induced bait region cleavage of tevRBD+2xAb resulted in both a conformational change in A2M and a heterogeneous release of the MG8 domain (and its two attached antibodies), giving 5 distinct bands in native PAGE corresponding to collapsed A2M with 0-4 thermolysin-released MG8 domains, i.e. 0=collapsed A2M, 1= collapsed A2M lacking 1 RBD+2xAb, 2= collapsed A2M lacking 2 RBD+2xAb, etc (Figure 19B). The released RBD+2xAb domains is not visible in the native gel, due to the small size of the proteins.

[0430] We produced an additional A2M with a slightly modified tevRBD sequence combined with another anti-CD3 scFv (derived from UCHT1) in the ciRBD position and an anti-EGFR nanobody (clone EgAl) in the C-terminal position, which we refer to as tevRBD+2xAb_2 in this example. This A2M was initially native, as seen from its thiol ester-mediated autolysis in SDS-PAGE, and remained native after TEV protease cleavage of the tevRBD site (Figure 19C). After changing A2M’s conformation using methylamine, MG8 domain and the two fused antibodies remained associated with A2M in native PAGE prior to TEV protease cleavage. In TEV protease-cleaved tevRBD+2xAb_2, the MG8 domain and the two antibodies were instead released from A2M, allowing its normal conformational collapse to be detected in native PAGE (Figure 19D). The fact that an A2M conformational change is not detected in native PAGE when antibodies are exposed by A2M’s conformational change is elaborated upon in Example 15, and shown to be due to a decrease in electrophoretic mobility of A2M’s collapsed conformation when it is associated with exposed antibodies.Conclusion

[0431] From these results, we conclude that both the ciRBD and C-terminal antibody fusion approaches, as well as the TEV protease site insertion used in tevRBD, can be combined while still producing native A2M. Importantly, the MG8 domain of this construct remains releasable by A2M’s conformational change. A2M with 2 antibodies and a TEV site can be fully processed by TEV protease while maintaining A2M’s native state. Upon induction of A2M’s conformational change, TEV proteasecleaved A2M releases its MG8 domain even if the MG8 domain is fused to two antibodies simultaneously.Example 15 -Release of MG8-antibody domain from A2M-BUE requires protease cleavageAim of study

[0432] Example 14 demonstrated that a TEV protease cleavage site can be used to generate an A2M construct with a MG8 domain that can be released by A2M’s conformational change. In this example, we investigated a further optimized A2M-BiTE construct similar to those tested in Example 14 to confirm that digestion with TEV results in two fragments, the MG8 domain comprising two antibodies and the N-terminal A2M domain, and that the addition of trypsin cuts the A2M domain in the bait region to yield two smaller fragments. A generalized representation of an A2M-BiTE construct is shown in Figure 20.Materials and methods

[0433] A2M with a modified TEV protease cleavage site, a C-terminal antibody that specifically binds a tumor associated antigen, i.e., a nanobody specific to EGFR (EgAl), and an anti-CD3 scFv (derived from Tebentafusp) (SEQ ID NO: 252) fused by the ciRBD approach was expressed in HEK293 FreeStyle cells using a standard transient transfection protocol and purified as described in Example 14.This A2M-BiTE, termed BiTE2-TEV4 (SEQ ID NO: 258) also comprised a TEV cleavage site with additional linkers on both sides, to make it more rapidly cleaved by the TEV protease.

[0434] The A2M-BiTE prodrug was treated with TEV protease at a molar ratio of 4: 1 (TEV:A2M- BiTE) or HBS followed by treatment with serial dilutions of Trypsin (TR) for 10 minutes, Methylamine (MA) for 45 minutes or HEPES-buffered saline (HBS; 20 mM HEPES, 150 mM NaCl, pH 7.4) for 45 minutes at 37°C. The trypsin reaction was stopped with 2 mM phenylmethylsulfonyl fluoride (PMSF) for 15 minutes at RT.

[0435] The samples were then analyzed by pore-limited and SDS-PAGE gel analysis, performed as described in previous examples.Results

[0436] In this example, we explored the possibility of providing a BiTE construct that cross-links a cell expressing a tumor-associated antigen (TAA) with a T cell via an anti-TAA antibody and an anti-CD3 antibody. The anti-CD3 antibody was positioned inside the MG8 domain so that it becomes accessible only in the tumor, the aim being to reduce side effects that result from the BiTE -dependent activation of T cells outside the tumor tissue.

[0437] The C-terminal anti-TAA antibody is freely accessible and can target the BiTE construct to cells expressing the TAA. We used trypsin as a model for a cancer-specific protease that the BiTE construct would encounter in the tumor microenvironment. The provided construct can be pre -cleaved with TEV, but releases the MG8 domain with the two antibodies only in the tumor microenvironment when it is cleaved by the cancer-specific protease and alters its conformation. Thus, the anti-CD3 antibody becomes accessible only upon cleavage the cancer-specific protease.

[0438] After digest with TEV, the MG8-BiTE domain comprising anti-CD3-MG8-anti-EGFR can be released by the conformation change (induced with MA or trypsin), giving an “empty” A2M that migrates the same as unmodified A2M. These results are shown in Figure 21 A, which shows a porelimited 4-15% gel.

[0439] After digest with TEV, followed by Trypsin digest, we observed three distinct bands representing the N-terminal A2M domain (annotated as the BR-cleaved fragment), a fragment comprising the MG7-CUB2 domains (annotated as the BR-cleaved fragment w / o (“without”) MG8), and the MG8 domain that is connected to the EGFR nanobody and the anti-CD3 scFv (annotated as MG8+2xAb), as shown in Figure 2 IB. Faint bait region cleavage products are visible even before trypsin treatment, as a small amount of A2M is bait region-cleaved by proteases present during HEK293F expression.Conclusion

[0440] This example demonstrates that the MG8 domain, to which the two antibodies of choice are fused, can be released once the A2M-BiTE has been subjected to selected protease activity, e.g., TEV- protease, followed by conformational change.Example 16 -A2M-BITE activates T cells in a protease cleavage-dependent mannerAim of study

[0441] As explained in Example 15 and shown in Figure 20B, the anti-CD3 scFv of the A2M-BiTE is shielded by A2M until conformational change is induced by cleavage of the first cleavage site, e.g., by trypsin. Trypsin serves as a model for a protease present in the tumor microenvironment. In this example, we investigate whether the anti-CD3-scFv in TEV-cleaved A2M-BiTEs can bind to the T cell receptor expressed on Jurkat cells and induce Erk 1 / 2 signaling in a trypsin-dependent manner.Materials and methods

[0442] For the proteolytic TEV cleavage of the BiTE2-TEV4 (SEQ ID NO: 258) and BiTE3-TEV5 (SEQ ID NO: 259) construct, 200-400 pg of purified BiTE construct were digested with TEV protease at a molar ratio of 8: 1 (TEV / A2M) in HBS buffer containing 0.5 mM EDTA and 50 pg / ml Leupeptin in a final volume of 300 pl. After incubating for 16 hours at 24°C, 200 pl HBS were added, and the sample was filtered with a 45 pm filter. The sample was loaded on a HiPrep 16 / 60 Sephacryl S-300 (Cytiva #17116701) and the column was run with 0.5 ml / min HBS with 1 ml fractions were collected. The fractions from 42 ml to 47 ml were pooled concentrated to 200-400 pl using Amicon Ultra-15 30K (Millipore UFC903024) and stored on ice until further use.

[0443] TEV-cleaved BiTE2-TEV4 (SEQ ID NO: 258) or BiTE3-TEV5 (SEQ ID NO: 259) was treated with Trypsin (TR) at a molar ratio of 2: 1 (TR:BiTE) or HBS (0) and incubated for 10 minutes at 37°C. The reaction was stopped with RPMI containing 10% FBS and the samples were kept on ice. The trypsin-treated (TR) and control (0) A2M BiTE sample were serially diluted in RPMI 10% FBS and incubated in the presence of 150,000 fluorescently labelled (with Carboxyfluorescein succinimidyl ester; CFSE) target cells (CHO or HCT116 cells) and 300,000 effector cells (Jurkat) per well for 5 minutes at 37°C in a final volume of 200 pl.

[0444] The cells were fixed by adding 100 pl 4% PFA to a final concentration of 1.3% paraformaldehyde for 10 minutes at room temperature. The fixed cells were centrifuged at 250 x g for 5 minutes at 24°C, the supernatant was removed and the cell pellet resuspended in 500 pl ice cold 95% methanol with vortexing (2500rpm) and incubated for 15 minutes on ice. After centrifugation at 500g for 5 minutes at room temperature, the supernatant was removed. The cells were washed 3 -times with 1 ml PBS 0.1% BSA and then resuspended in 100 pl staining media and transferred to a round bottom 96-well plate. 0.05 pl Fluor647-conjugated anti-phospho Erkl / 2 antibody (Biolegend #675503) was added per sample and incubated for 30 minutes at RT. The cells were washed twice with PBS and resuspended in 200 pl 1% PFA for 10 minutes at RT for fixation. The samples were centrifuged and resuspended in 200 pl staining media and stored at 4°C until they were analyzed using a Novocyte Quanteon flow cytometer (Agilent).Results

[0445] We tested whether a bispecific anti-CD3, anti-EGFR moiety released from TEV-cleaved A2M would induce activation of effector Jurkat cells in the presence of EGFR+target cells (HCT116), as determined by flow cytometry-based measurement of pERKl / 2 levels. This was tested using two different A2M-BiTEs; BiTE2-TEV4 (SEQ ID NO: 258, as described in Example 15) and BiTE3-TEV5 (SEQ ID NO: 259), which has extended linkers around the TEV protease cleavage site to make it more accessible and incorporates the Lysl393Ala and Lysl397Ala mutations that prevent A2M’s binding to the LRP1 and Grp78 receptors. The bispecific moiety was released from the A2M-BiTEs by proteolytic cleavage of their bait regions with trypsin. The data are summarized in Figures 22A and 22B. EC50 values were calculated from the data shown in these figures and are summarized in Table 5.

[0446] For BiTE2-TEV4, ERK1 / 2 phosphorylation was 30.8-fold higher after treatment with Trypsin than without, as shown in Table 5. For BiTE3-TEV5, which had longer linkers, ERK1 / 2 phosphorylation was 61.5 -fold higher after treatment with Trypsin than without, as shown in Table 5.Table 5.

[0447] T cell activation by a BiTE antibody construct is dependent on the presence of target cells expressing the tumor associated antigen (TAA) bound by one of the antibodies of the construct. To confirm that ERK1 / 2 phosphorylation, and thus T cell activation, was strictly dependent on the presence of TAA-expressing target cells, we repeated the experiment with wild-type CHO cells (CHO-wt) and CHO cells expressing EGFR (CHO-EGFR). The data are summarized in Figure 22C.

[0448] The CHO cell experiment confirmed that T cell activation was dependent on both treatment of BiTE2-TEV4 with Trypsin and presence of target cells expressing the TAA (EGFR). No ERK1 / 2 phosphorylation was observed upon incubation with CHO-wt cells, regardless of whether or not BiTE2- TEV4 was trypsin-cleaved. In contrast, an increase in pERKl / 2 was observed when CHO-EGFR were used and BiTE2-TEV4 was trypsin-cleaved (but not if it was untreated). This result shows that T cell activation by BiTE2-TEV4 is dependent on both TAA-expressing target cells and the conformational change of A2M.Conclusion

[0449] This example demonstrates that T cell activation by A2M-BiTEs is dependent on A2M altering conformation upon proteolytic cleavage of the protease cleavage site located in the bait region and release of the TEV-cleaved MG8 domain comprising the anti-CD3 scFv. The data confirm that the anti- CD3 scFv is masked by A2M prior to cleavage of the protease cleavage site in the bait region. The results further indicate that T cell activation is dependent on the presence of target cells expressing the TAA bound by the antibody fused to the C-terminus of the MG8 domain.

[0450] We envision that the protease cleavage site in the bait region is modified to be selectively cleaved by one or more proteases expressed in the tumor environment, such that cleavage of this site in the A2M-BiTE occurs predominantly in the tumor, limiting T cell activation and reducing side effects associated with conventional BiTEs.Example 17 - Protease-activated A2M-BiTEs induce antigen-specific cytotoxicityAim of study

[0451] In this example, we investigated whether the A2M-BiTE construct generated in Example 15, i.e., BiTE2-TEV4 (SEQ ID NO: 258), could induce specific killing of TAA-expressing target cells after cleavage of both their bait regions and the TEV cleavage site using thermolysin.Materials and methods

[0452] A2M-BiTEs were either pre-treated (“activated”) with thermolysin as described in Example 14 or TEV as described in Example 16, respectively, or left untreated (“non-activated”). TEV-activated A2M-BiTEs were purified by size exclusion chromatography (SEC) as described in Example 16. To change A2M’s conformation and render the anti-CD3 scFv accessible, the A2M-BiTEs were incubated with either Trypsin or MA to induce cleavage of the protease site located in the bait region.

[0453] PBMC s were purified from fresh buffy coats using Histopaque 1077 (Sigma-Aldrich) according to the manufacturer’s instruction and frozen aliquots were stored in liquid nitrogen. PBMCs were thawed and activated with 50 ng / ml OKT3 and 100 lU / ml IL-2. After 3 days, the medium was changed to RPMI 10% FBS with 100 IU of IL2 and cells were used for the toxicity assay between day 7 and 14. For the cytotoxicity assay, 100 pl of the target cells (HCT116, CHO-wt or CHO-EGFR) in 100 pl medium were seeded into 96 well plates and incubated for 4h in a humified atmosphere at 37°C 5% CO2. After 4h, 2* 105PBMC effector cells in 50 pl RPMI 10% FBS and 50 pl of the BiTE samples were added. The plates were centrifuged for 2 minutes at 250 g and incubated for 16 hours in a humified atmosphere at 37°C 5% CO2. The PBMCs were washed away by three washes with 200 pl of PBS. 100 pl ofDMEM 10% FBS with 10 mM Resazurinwas added to each well and the plates were incubated in a humified atmosphere at 37°C 5% CO2 for 3 hours. The percentage of live cells was determined by measuring the reduction of the resazurin fluorescence (at 540 nm extinction and 590 nm emission).Results

[0454] Wild-type CHO cells (CHO-wt) do not express any EGFR on their surface. They can be induced to express EGFR by transient transfection with an EGFR expression vector to yield EGFR-expressing CHO cells (CHO-EGFR). CHO-wt and CHO-EGFR therefore provide a suitable model to investigate whether A2M-BiTEs targeting EGFR are capable of inducing cytotoxicity in an antigen-dependent manner. We therefore performed a first set of experiments using either CHO-wt or CHO-EGFR as target cells to study whether thermolysin-activated or non-activated A2M BiTEs targeting EGFR, here, BiTE2-TEV4 (SEQ ID NO: 258) can induce a cytotoxic T cell response when using PBMCs as effectorcells. Thermolysin cleaves both the TEV site (enabling the release of the BiTE moiety from A2M) and the A2M bait region (which triggers the conformation change and release of the BiTE). As shown in Figure 23A, CHO-EGFR were killed by effector cells only in the presence of the activated EGFR- targeting A2M-BiTE, indicating BiTE activity is dependent on the release of the antibody-containing MG8 domain from A2M. Non-activated A2M-BiTE targeting EGFR had no cytotoxic effect although the anti-EGFR antibody is exposed at the C-terminus. As expected, CHO-wt were not killed, regardless of whether the A2M-BiTE was activated or not, indicating that T cell activation by the A2M-BiTE was dependent on the presence of cells expressing the tumor-associated antigen, here, EGFR.

[0455] The human colon carcinoma cell line HCT116 has previously been used as a model system for EGFR-expressing colorectal cancers. Hence, it represents a therapeutically more relevant target cell to study EGFR-targ eting A2M-BiTEs. We therefore performed a second set of experiments using HCT116 cells to test whether a TEV-activated or non-activated A2M-BiTEs targeting EGFR, here, BiTE3-TEV5 (SEQ ID NO: 259) were capable of inducing a cytotoxic T cell response against these cells. As shown in Figure 23B, HCT116 cells were killed effectively by PBMC-derived cytotoxic T cells in the presence of the TEV-activated BiTE3-TEV5, with increased potency observed for MA -treated or Trypsin-cleaved BiTE3-TEV5 compared to when it was untreated. These results indicate that the conformational change of the A2M-BiTE, which exposes the anti-CD3-scFv and releases the MG8 domain from the N-terminal A2M domain, is required to induce an effective cytotoxic T cell response.Conclusion

[0456] From these results, we conclude that an activated A2M-BiTE targeting a TAA is capable of inducing a cytotoxic T cell response that is specific to TAA-expressing cells. In other words, T cell activation is dependent on the presence of cells expressing the TAA targeted by the A2M-BiTE. Moreover, BiTE-induced cytotoxicity is dependent on the conformational change of the A2M and release of antibody-containing MG8 domain.

Claims

Claims1. A proteinaceous prodrug construct comprising a complement 3- and pregnancy zone protein-like, alpha- 2-macroglobulin domain-containing (CPAMD) protein, comprising:(a) a bait region comprising a first protease cleavage site; and(b) a Receptor Binding Domain (RBD) comprising: a second protease cleavage site at the N-terminal end of the RBD, a first drug positioned inside the RBD, and a second drug fused to the C-terminal end of the RBD; wherein:(i) the CPAMD protein shields the first drug and the second drug is accessible; and(ii) the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD and thereby making the first drug accessible.

2. The proteinaceous prodrug construct of claim 1, wherein the first protease cleavage site is different from the second protease cleavage, or the first cleavage site and the second cleavage site are the same.

3. The proteinaceous prodrug construct of claim 2, wherein the second protease cleavage site is different and capable of specifically being cleaved by furin, TEV, Enterokinase, or Thrombin.

4. The proteinaceous prodrug construct of any one of the preceding claims, wherein the second protease cleavage site is introduced at one or more amino acids corresponding to residues 1334 to 1340 of human A2M.

5. The proteinaceous prodrug construct of any of the preceding claims, wherein the second protease cleavage site is positioned between two linkers, optionally wherein the linkers are 5-30 amino acid length, optionally wherein the linkers are GS linkers.

6. A proteinaceous prodrug construct comprising a complement 3- and pregnancy zone protein-like, alpha- 2-macroglobulin domain-containing (CPAMD) protein, comprising:(a) a bait region comprising a first protease cleavage site; and(b) a Receptor Binding Domain (RBD) comprising a first drug positioned inside the RBD; and a second drug fused to the C-terminal end of the RBD; wherein:(i) the CPAMD protein shields the first drug and the second drug is accessible;(ii) the RBD is bound to the CPAMD protein, via non-covalent interactions; and(iii) the CPAMD protein is capable of altering conformation upon proteolytic cleavage of the first protease cleavage site, releasing the RBD and thereby making the first drug accessible.

7. The proteinaceous prodrug construct of any one of the preceding claims, wherein the first protease cleavage site is: cleaved specifically by a protease expressed by a cancer; and / or specific for a serine-, cysteine-, aspartic- and / or metalloproteinase.

8. The proteinaceous prodrug construct of any one of the preceding claims, wherein the CPAMD protein is human alpha-2 -macroglobulin (A2M), or a functional homolog thereof, e.g., a mammalian A2M.

9. The proteinaceous prodrug construct of any one of the preceding claims, wherein the first drug is inserted into loop 2 of the RBD, corresponding to amino acid residues 1391-1405 of human A2M, optionally the first drug is inserted between amino acids corresponding to residues 1402-1403 of human A2M or replaces one or more of the amino acids corresponding to residues 1391, 1392, 1393, 1394, 1395, 1396, 1397, 1398, 1399, 1400, 1401, 1402, 1403, 1404, and / or 1405 of human A2M.

10. The proteinaceous prodrug construct of any one of the preceding claims, wherein the second drug is fused to the amino acid corresponding to residue 1474 of human A2M.

11. The proteinaceous prodrug of any one of the preceding claims, wherein the first drug and / or the second drug is an antigen-targeting moiety such as an antibody, e.g., a single-chain or a single-domain antibody.

12. The proteinaceous prodrug construct according to any of the preceding claims, wherein the first and / or the second drug is / are capable of directing the proteinaceous prodrug construct to a specific tissue, a specific cell type, and / or a specific receptor.

13. The proteinaceous prodrug construct according to any one of the preceding claims, wherein the first drug is capable of directing the proteinaceous prodrug construct to an immune cell, e.g., an NK cell, a macrophage, a T cell, or a dendritic cell.

14. The proteinaceous prodrug construct of claim 12 or 13, wherein the first drug is(i) a T cell specific moiety, e.g., targeting a receptor expressed at increased levels on T cells such as CD3, CD4 and / or CD8, optionally wherein the T cell specific moiety is an anti-CD3 antibody.(ii) an NK cell specific moiety, e.g., targeting a receptor expressed at increased levels on NK cells such as CD16, optionally wherein the NK cell specific moiety is an anti-CD16 antibody.(iii) a macrophage specific moiety, e.g., targeting a receptor or molecule expressed at increased levels on macrophages such as a SIRPa or DNGR1, optionally wherein the macrophage specific moiety is an inhibitory anti-SIRPa antibody that blocks CD47 or an antibody against the DNGR1 receptor.

15. The proteinaceous prodrug construct according to any one of claims 1-12, wherein the first drug is a toxin, an enzyme, a protein conjugate comprising a small-molecule drug (e.g., a cytotoxin), a cytokine, an extracellular region of a cell surface receptor, an extracellular region of a cell surface ligand, or a receptor agonist.

16. The proteinaceous prodrug construct according to any one of the preceding claims, wherein the second drug is an antigen-targeting moiety that specifically binds to bind to a tumor cell surface antigen.

17. A nucleic acid, e.g., a vector such as a plasmid, encoding the proteinaceous prodrug construct according to any one of claims 1-5 or 7-16.

18. A host cell comprising the nucleic acid according to claim 17.

19. A pharmaceutical composition comprising the proteinaceous prodrug construct of any one of claims 1- 16, or the nucleic acid of claim 17, and a pharmaceutically acceptable excipient.

20. The proteinaceous prodrug construct according to any one of claims 1-16, or the nucleic acid of claim 17 for use as a medicament, e.g., for use in a method of treating cancer.

21. A method of treating or preventing a disease or disorder in a subject in need thereof, wherein the method comprises administering a therapeutically effective amount of the proteinaceous prodrug construct according to any one of claims 1-16, or the nucleic acid of claim 17 to the subject.

22. A method of producing the proteinaceous prodrug of any one of claims 1-5 and 7-16, the method comprising:(a) providing a host cell comprising the nucleic acid of claim 17; and(b) culturing the host cell under conditions that allow for expression of the proteinaceous prodrug construct from the nucleic acid.

23. A method of producing the proteinaceous prodrug of claim 6, the method comprising:(a) providing a host cell comprising the nucleic acid of claim 17;(b) culturing the host cell under conditions that allow for expression of the proteinaceous prodrug construct encoded by the nucleic acid; and(c) contacting the proteinaceous prodrug construct with a protease that specifically cleaves the second protease cleavage site.

24. A method of producing a proteinaceous prodrug, the method comprising:(a) contacting the proteinaceous prodrug of claim 6 with a protease that specifically cleaves the second protease cleavage site; and(b) purifying the proteinaceous prodrug to remove the protease.

25. The method of claim 24, wherein the purification step comprises chromatography, e.g., size exclusion chromatography .

Images