WO2020234556A1

WO2020234556A1 - Conjugates

Info

Publication number: WO2020234556A1
Application number: PCT/GB2020/050781
Authority: WO
Inventors: Caroline Jane BARELLE; Laura Ann FERGUSON; John Steven; Andrew Justin Radcliffe Porter; Adam Stephen Williams LEACH; Christopher John Scott; Michelle Katherine GREENE; Peter Hugh Robert SMYTH
Original assignee: Elasmogen Ltd
Priority date: 2019-05-17
Filing date: 2020-03-24
Publication date: 2020-11-26
Also published as: GB201906997D0

Abstract

The invention provides a specific antigen binding molecule-substrate conjugate, comprising: (a) a specific antigen binding molecule having an amino acid sequence represented by the formula FW1-CDR1-FW2-HV2-FW3a-HV4-FW3b-CDR3-FW4 in which FW1, FW2, FW3a, FW3b, and FW4 are Framework Regions, CDR1 and CDR3 are Complementarity Determining Regions, and HV2, and HV4 are Hypervariable Regions; and (b) a substrate.

Description

CONJUGATES

Field of Invention

The present invention relates to specific antigen binding molecule-substrate conjugates.

Background

The search for specific, increasingly efficacious, and diversified therapeutic weapons to combat diseases has utilised a myriad of distinct modalities. From the traditional small molecule to incrementally larger biologic pharmaceuticals, for example single binding domains (10-15 kDa) to full IgG (-150 kDa). Single domains currently under investigation as potential therapeutics include a wide variety of distinct protein scaffolds, all with their associated advantages and disadvantages.

Such single domain scaffolds can be derived from an array of proteins from distinct species. The Novel or New antigen receptor (IgNAR) is an approximately 160 kDa homodimeric protein found in the sera of cartilaginous fish^1-3. Each molecule consists of a single N-terminal variable domain (VNAR) and five constant domains (CNAR).

The IgNAR domains are members of the immunoglobulin-superfamily. The VNAR is a tightly folded domain with structural and some sequence similarities to the immunoglobulin and T-cell receptor Variable domains and to cell adhesion molecules and is termed the VNAR by analogy to the N Variable terminal domain of the classical immunoglobulins and T Cell receptors.

The Variable New Antigen Receptors (VNARs) are the smallest (11 kDa) naturally occurring independent binding domains in the vertebrate kingdom ¹'⁴·⁵. They play an integral role in the adaptive immune system in cartilaginous fish and although they are structurally similar to mammalian heavy and light variable chains it has been well documented that they arose from a distinct evolutionary lineage from Immunoglobulins or classical antibodies ⁶. A lack of CDR2 and the addition of two additional loops of diversity (HV2 and 4), the lack of any light-chain partner in their evolutionary history, as well as low percentage sequence homology (around 25%) to mammalian antibody binding sites, confirms their separation from antibodies. Their characteristic protruding paratopes, (encouraged by the presence of disulphide bridges in the vNAR binding site) results in protruding structures often referred to as“canyon-binders”, that predisposes these domains to access and bind epitopes (pockets and clefts in proteins) not normally available to conventional biologies and encourages the selection of highly potent neutralisers specific for enzyme and/or receptor targets ^7-9. For conventional antibodies and many antibody fragments their flat paratopes, made up of 6 loops (3 from a heavy chain and three from a partner light chain), are simply unable to recognise or bind to these recessed epitopes seen by VNARs. Whilst research has shown reformatting of VNAR domains to deliver: multivalent, bi/tri-specific constructs, and serum half-life extension through molecular fusion of VNAR to a second anti-human serum albumin (HSA) VNAR scaffold, or via a more traditional route using IgG (mouse or human) Fc domains³'¹⁰·¹¹ (tailored drug modalities optimised for systemic, site-specific or topical administration) there is no previous publication that shows that monomeric VNAR can be conjugated directly onto the surface of nanoparticles or other biomaterials to deliver high density, functional binders that can recognise target antigens via this binding to cryptic epitopes and/or allosteric binding sites hidden from normal antibodies. These highly stable monomers can be expressed cost-efficiently and at scale in non-mammalian systems and their inherent stability makes them an ideal starting material for chemically controlled (correct orientation) and high-efficiency coating of nanoparticles with functional binding sites. This is in stark contrast to antibodies (or antibody fragments) which are significantly larger (25 - 150 kDa), show reduced tolerance to conjugation chemistries because of their more complex two chain structures and as such are not easily orientated at the

nanoparticle surface. For instance, based purely on size alone, a perfectly spherical nanoparticle possessing a diameter of 200 nm, could facilitate the loading of 7.5-12.9 fold more VNARs when compared to a full IgG antibody.

In therapeutic settings full antibodies have been known to suffer from inappropriate activation of inflammatory responses due to the effector functions of the Fc regions¹² necessitating engineering of constant regions to reduce these functions¹³·¹⁴. This has been demonstrated in targeted liposome formulations where formulations bearing targeting moieties possessing an Fc region were more rapidly cleared due to Fc-mediated uptake into the liver and spleen¹⁵. VNARs are not expected to suffer from similar problems, due to their lack of the IgNAR constant domains (CNAR) the functions of which in their native setting remain relatively unknown¹⁶.

Nanoparticles, i.e. , particles having physical dimensions on the nanometre scale, show great promise in a number of therapeutic areas. Over the last 15-20 years there has been a substantial increase in the study of nanoformulations for both therapeutic and diagnostic applications. With diverse formulations including nanoparticles, liposomes and dendrimers each able to be tailored to suit, the potential uses are considerable. Such delivery systems can shield drugs from degradation before they reach their site of action, which not only increases efficacy by increasing drug concentration at the target site but also minimises off-target side effects¹⁷. Furthermore, through the incorporation of ligands on the corona of a nanoparticle, selective, disease specific, delivery can be enhanced. Not only can such ligands be used to target an encapsulated therapeutic payload to the requisite site but through interaction with receptors and other molecules, targeting moieties can themselves impart biological function. To date, a wide range of ligands have been utilised for this purpose including small molecules¹⁸, carbohydrates¹⁹, peptides²⁰, aptamers²¹ and antibodies²².

Summary of Invention

In a first aspect there is provided a specific antigen binding molecule-substrate conjugate, comprising:

(a) a specific antigen binding molecule having an amino acid sequence represented by the formula FW1-CDR1-FW2-HV2-FW3a-HV4-FW3b-CDR3-FW4 in which FW1 , FW2, FW3a, FW3b, and FW4 are Framework Regions, CDR1 and CDR3 are Complementarity Determining Regions, and HV2, and HV4 are Hypervariable

Regions; and

(b) a substrate.

Preferably, the substrate comprises a biomaterial, a microparticle, or a nanoparticle. The substrate may also comprise a surface, plate, or bead.

Preferably, the specific antigen binding molecule is conjugated to the substrate via a covalent bond. The conjugation may also be via adsorptive or hydrophobic/hydrophilic interaction

The specific antigen binding molecule may also be referred to herein as a vNAR (or VNAR). The specific antigen binding molecule may be humanised.

The specific antigen binding molecule may bind to DLL4, Human serum albumin (HSA), or hen egg lysozyme (HEL).

The specific antigen binding molecule may further comprise an alanine motif represented by the amino acid formula ACA or AACAA, wherein the cysteine residue is available for conjugation to the substrate. Preferably the alanine motif is located at the N-terminus or C-terminus of the specific antigen binding molecule.

Examples of amino acid sequences of specific antigen binding molecules for use in the present invention include specific antigen binding molecules having a sequence selected from those set out in Table 1 .

Table 1 : Examples of specific antigen binding molecules for use in the present invention.

The specific antigen binding molecule may be conjugated to the substrate via any suitable means, including but not limited to maleimide, NHS, VS, or SuFEX.

The substrate may be any suitable material. Preferably, the substrate is a PLGA nanoparticle. In some cases, the conjugate may also include polyethylene glycol (PEG).

In a preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 1 conjugated to a PLGA nanoparticle via maleimide, optionally wherein the conjugate also includes PEG. In another preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 2 conjugated to a PLGA nanoparticle via maleimide, optionally wherein the conjugate also includes PEG. In a preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 3 conjugated to a PLGA nanoparticle via maleimide, optionally wherein the conjugate also includes PEG. In a preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 4 conjugated to a PLGA nanoparticle via maleimide, optionally wherein the conjugate also includes PEG. In a preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 5 conjugated to a PLGA nanoparticle via maleimide, optionally wherein the conjugate also includes PEG. In a preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 6 conjugated to a PLGA nanoparticle via maleimide, optionally wherein the conjugate also includes PEG.

In a yet further preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 1 conjugated to a PLGA nanoparticle via NHS, VS or SuFEX, optionally wherein the conjugate also includes PEG. In another preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 2 conjugated to a PLGA nanoparticle via NHS or SuFEX, optionally wherein the conjugate also includes PEG. In a preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 3 conjugated to a PLGA

nanoparticle via NHS or SuFEX, optionally wherein the conjugate also includes PEG. In a preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 4 conjugated to a PLGA nanoparticle via NHS or SuFEX, optionally wherein the conjugate also includes PEG. In a preferred embodiment, the specific antigen binding molecule-substrate conjugate comprises SEQ ID NO: 5 conjugated to a PLGA nanoparticle via NHS or SuFEX, optionally wherein the conjugate also includes PEG. In a preferred embodiment, the specific antigen binding molecule- substrate conjugate comprises SEQ ID NO: 6 conjugated to a PLGA nanoparticle via NHS or SuFEX, optionally wherein the conjugate also includes PEG. The specific antigen binding molecule may recognise an epitope not otherwise accessible to antigen binding molecules such as antibodies, due to steric hinderance caused by the epitope’s three dimensional structure. Such epitopes may be referred to as a cryptic or hidden epitope.

Preferably, the specific antigen binding molecule binds to DLL4 or HSA.

Preferably, when the substrate is a nanoparticle and the specific antigen binding molecule binds to HSA, binding causes alteration of the composition of the nanoparticle corona

Preferably, when the specific antigen binding molecule-substrate conjugate binds to HSA, the bound HSA is still capable of functional binding to the protein SPARC.

In addition, there are provided herein specific antigen binding molecule-substrate conjugates as described above for use in therapy. Furthermore, there are provided herein specific antigen binding molecule-substrate conjugates for use in the treatment of cancer. A method of treating cancer by administration of the specific antigen binding molecule- conjugates described above, preferably in association with administration of

chemotherapy, preferably albumin-associated chemotherapy, is also contemplated herein.

In another aspect, there is provided herein a specific antigen binding molecule for use in the treatment of cancer. More specifically, there is provided herein a specific antigen binding molecule for use in the treatment of cancer via chemotherapy, preferably albumin- associated chemotherapy. Corresponding methods of treatment of cancer by

administration of the specific antigen binding molecule described, preferably in association with administration of chemotherapy, preferably albumin-associated chemotherapy, are also contemplated herein. The specific antigen binding molecule for use as described in this aspect binds to an epitope on human serum albumin (HSA) that does not prevent HSA from binding to Secreted Protein Acidic and Rich in Cysteine (SPARC). The features of the other aspects of the present invention as described above may be combined with the specific antigen binding molecule for use as described in this aspect. Preferably, the specific antigen binding molecule for use has the sequence of BA11 (SEQ ID NO: 3).

In one embodiment, the invention provides a specific antigen binding molecule-substrate conjugate as described above for use in treatment of cancer wherein the specific antigen binding molecule is an anti-DLL4 specific binding molecule. The anti-DLL4 specific binding molecule may comprise E4. Accordingly, the anti-DLL4 specific binding molecule may comprise SEQ ID NO: 4. Description of Figures

Figure 1: Schematic representation of PLGA-PEG-Mal nano-formulation approach.

Polymeric nanoparticles were formed from a polymer blend of PLG A 502H and PLGA- PEG-Mal (w/w 15%/25%). Nanoparticles were formed using a single emulsion-solvent evaporation approach resulting in polymeric nanoparticles bearing a reactive maleimide group on their surface.

Figure 2. SPR comparison of PLG A-PEG-Mal-10 and PLGA-PEG-Mal-E4

nanoparticles. PLGA-PEG-Mal polymeric nanoparticles were incubated with anti-DLL4 clone 10 vNAR or anti-DLL4 clone E4 vNAR. Post-incubation, (A) the ability of the nanoparticles to bind DLL4 was determined by SPR. Data presented as representative SPR binding sensorgram with corresponding details for relative binding response and (B), varying concentrations of PLGA-PEG-Mal-E4 nanoparticles with data presented as relative binding response.

Figure 3. Modified ELISA comparison of PLG A-PEG-Mal-10 and PLG A-PEG-Mal-E4 nanoparticles and effect of percentage maleimide polymer on binding affinity (A)

Rhodamine 6G loaded PLG A-PEG-Mal-10 and PLGA-PEG-Mal-E4 nanoparticles assessed in terms of their binding affinity to an immobilized DLL4 receptor via a modified ELISA method. Binding was assessed by fluorescence spectroscopy with data expressed as mean ± standard deviation (SD). Assay performed in triplicate. Representative of three independent experiments. Statistical significance was established by two-way ANOVA and Tukey's post-hoc test (****p < 0.0001). To assess the importance of the reactive maleimide group in the conjugation process (B) Rhodamine 6G-loaded PLGA-PEG-Mal polymeric nanoparticles possessing varying proportions of PLGA-PEG-Mal copolymer, were conjugated with anti-DLL4 clone E4 vNAR and binding assessed by modified ELISA. Data expressed as mean ± standard deviation (SD). Assay performed in triplicate.

Representative of two independent experiments Statistical significance was established by two-way ANOVA and Tukey's post-hoc test (**p £ 0.01; ****p £ 0.0001).

Figure 4. Comparison of PLGA-PEG-Mal and PLGA-PEG-NHS nanoparticles conjugated to anti-DLL4 vNARs. Rhodamine 6G loaded PLGA-PEG-Mal and PLGA- PEG-NHS polymeric nanoparticles were incubated with anti-DLL4 clone E4 vNAR. Post incubation, nanoparticles were assessed (A) in terms of binding affinity via modified ELISA method. Data expressed as mean ± standard deviation (SD). Assay performed in triplicate. Representative of three independent experiments. Statistical significance was established by two-way ANOVA and Tukey's post-hoc test (****p < 0.0001). (B) Binding assessment of PLGA-PEG-Mal-E4 and PLGA-PEG-NHS-E4 nanoparticles by SPR. Data presented as relative binding response with corresponding SPR binding sensorgram details. Representative of three independent experiments

Figure 5. Assessment of PLGA-PEG-Mal-E4 nanoparticle binding following anti- DLL4 antibody pre-block. Rhodamine 6G loaded, PLGA-PEG-Mai-E4 nanoparticles assessed in terms of binding affinity via modified ELISA method. Nanoparticle binding was assessed following a 2 h incubation of antigen with an anti-DLL4 antibody. Data expressed as mean ± standard deviation (SD). Assay performed in triplicate.

Representative of two independent experiments. Statistical significance was established by two-way ANOVA and Tukey's post-hoc test (ns p>0.05; ****p £ 0.0001).

Figure 6. Assessment of PLGA-PEG-Mai-E4 nanoparticle binding in biological media. Rhodamine 6G loaded, PLGA-PEG-Mal-E4 nanoparticle binding affinity assessed via a modified ELISA method. Binding was assessed in PBS, 10% FCS/PBS or 10% FCS/DMEM growth media with data expressed as mean ± standard deviation (SD). Assay performed in triplicate. Statistical significance was established by two-way ANOVA and Tukey's post-hoc test (****p < 0.0001).

Figure 7. Assessment of SUFEX-vNAR conjugation approach. Binding affinity of vNAR conjugated SUFEX and NHS ester functionalized particles were compared by modified ELISA. Equimolar amounts of each functional group were used in the fabrication of the particles. Statistical significance established by one-way ANOVA and Tukey’s post- hoc test, **** p<0.0001, *** p<0.001

Figure 8. Assessment of VS-vNAR conjugation approach E4 vNAR was conjugated to polymeric nanoparticle formulations possessing VS-functionalised polymer or NHS- functionalised polymer. Binding affinity of VNAR conjugated VS and NHS ester functionalized particles were compared by modified ELISA. Equimolar amounts of each functional group were used in the fabrication of the particles. Statistical significance established by one-way ANO VA and Tukey’s post-hoc test, *** p<0.001.

Figure 9. Validation of DLL4 expression in HUVECS by western blot poststimulation with FGF and VEGF. (A) HUVECs were treated with FGF and/or VEGF for 24 h. Following treatment, cells were lysed and DLL4 expression validated by western blot a-tubulin was used as a loading control. (B) Densitometry quantification of western blot DLL4 expression levels Figure 10. Assessment of PLGA-PEG-Mal-E4 nanoparticle uptake by HUVEC cells.

HUVEC cells (pre-stimulated with FGF and VEGF) were treated with either rhodamine 6G loaded PLGA-PEG-Mal or PLGA-PEG-Mal-E4 nanoparticles for 3 h. Following treatment, the cells were washed, fixed and nuclear regions stained with DAPI. Cells were subsequently imaged by confocal microscopy. Four representative images are presented per treatment group. Scale bars = 10 pm.

Figure 11. Cell viability assessment of DLL4 positive HUVEC cells following treatment with camptothecin (CPT) loaded PLGA-PEG-Mal nanoparticles CPT loaded nanoparticles composed of 75% PLGA 502H; 25% PLGA-PEG-Maleimide were produced and assessed in terms of (A) size, PDI, zeta potential and CPT entrapment via

comparison of CPT fluorescent signal to a standard calibration curve. (B) HUVEC cells stimulated with VEGF/FGF were treated with CPT-loaded nanoparticles for 1 h at 4 °C with cell viability assessed 24h post treatment via CellTiter-Glo®. Assay performed in triplicate. Representative of two independent experiments. Statistical significance was established by one-way ANOVA and Tukey's post-hoc test (**p £ 0.01).

Figure 12. Cell viability assessment of DLL4 positive A 549 cells following treatment with camptothecin (CPT) loaded PLGA-PEG-Mal nanoparticles. PLGA-PEG-Mal nanoparticles encapsulating camptothecin (CPT) and functionalised with E4 vNAR were incubated for 10 min with DLL4 positive A549 cells. Cell viability was assessed 24 h post treatment via CellTiter-Glo®. Assay performed in triplicate. Representative of two independent experiments. Statistical significance was established by one-way ANOVA and Tukey's post-hoc test (**p £ 0.01).

Figure 13. vNAR-targeted nanoparticle binding and uptake in cancer cell lines as determined by fluorescent plate reader method. (A) Nanoparticle binding (top row) and uptake (bottom row) of 3 pancreatic cancer cell lines: BXPC-3 (left), MIA PaCa-2 (middle) and PANC-1 (right). Cells were treated with particles conjugated either to E4 or 2V at a range of polymer concentrations up to 1000 pg mL¹ with fluorescence measured via plate reader

(B) Nanoparticle binding (left) and uptake (right) of E4 conjugated or nude nanoparticles in A549 cells

Figure 14. Binding of fluorescently rhodamine 6G loaded vNAR-targeted

nanoparticles to DLL4 positive A 549 cells following E4 vNAR preblock. Binding of either E4-conjugated or nude fluorescent nanoparticles to A549 cells treated with varying concentrations of free E4 (0, 11 or 44 pg mL ¹) prior to nanoparticle treatment. Cells were treated with free E4 vNAR for 30 minutes prior to treatment with nanoparticles (500 pg mL ¹) for 1 h.

Figure 15. Clonogenic potential of ASPC-1 cells following treatment with PLGA- PEG-Mal and PLGA-PEG-Mal-E4 nanoparticles in the presence of free E4 preblock.

ASPC-1 cells were treated for 45 minutes with: PBS (PBS), 250 ng mL¹ of camptothecin (free CPT), E4-conjugated camptothecin-entrapped particles equivalent to 250 ng mL ¹ camptothecin (E4 CPT) or nude camptothecin-entrapped particles equivalent to 250 ng mL ¹ camptothecin (nude CPT).“+E4 preblock’’ signifies cells were treated with 15 pg mL ¹ free E4 for 15 minutes prior to the addition of PBS, free CPT or nanoparticle treatment. Colonies were allowed to develop for 15 days before staining with 0.5% crystal violet solution, colonies were counted manually and data expressed as colonies per well.

Statistical significance was established by unpaired t-test (* p<0.05).

Figure 16: Assessment of PLGA-PEG-Mal and PLGA-PEG-NHS nanoparticles conjugated to anti-HSA vNARs. Rhodamine 6G loaded PLGA-PEG-Mal and PLGA- PEG-NHS polymeric nanoparticles were incubated with anti-HSA clone BA11 vNAR or anti-HSA clone E06 vNAR. Post-incubation, rhodamine 6G loaded, anti-HSA nanoparticles were assessed in terms of their binding affinity to immobilized HSA via a modified ELISA method. Binding was assessed by fluorescence spectroscopy with data expressed as mean ± standard deviation (SD). Assay performed in triplicate. Representative of three independent experiments. Statistical significance was established by two-way ANOVA and Tukey's post-hoc test (ns p>0.05; ****p £ 0.0001).

Figure 17: Assessment of PLGA-PEG-Mal and PLGA-PEG-NHS nanoparticles conjugated to anti-Hen egg lysozyme vNARs. Rhodamine 6G loaded PLGA-PEG-Mal and PLGA-PEG-NHS polymeric nanoparticles were incubated with the humanized anti- Hen egg lysozyme clone 5A1 vNAR. Post-incubation, rhodamine 6G loaded, anti lysozyme nanoparticles were assessed in terms of their binding affinity to immobilized lysozyme via a modified ELISA method. Binding was assessed by fluorescence spectroscopy with data expressed as mean ± standard deviation (SD). Assay performed in triplicate. Representative of three independent experiments. Statistical significance was established by two-way ANOVA and Tukey's post-hoc test (ns p>0.05; ****p £ 0.0001).

Figure 18: Physicochemical characterisation of particles incubated with human serum albumin. Particles were either incubated with 1.2 nmoles/ mg particle HSA dissolved in PBS (+HSA) or PBS alone for 3 hours before washing. Nude, BA 11- conjugated and 2V-conjugated nanoparticles measured. Clockwise from top left: Hydrodynamic diameter as determined by dynamic light scattering (DLS), Polydispersity index (PDi) as determined by DLS, associated protein mass per mg of polymer as determined by micro BCA assay and zeta potential as determined by Phase Analysis Light Scattering (PALS). Statistical significance was established by one-way ANOVA and Tukey's post-hoc test (ns p>0.05; * p <0.05, ***p £ 0.001)

Figure 19 SPARC pulldown Western blot. Western blot against SPARC, gel lanes loaded with particles pelleted after stirring with PANC-1 conditioned complete media. (1) BA11 -conjugated particles coated with HSA prior to conditioned media pulldown, (2)

BA11 -conjugated particles, (3) 5 A7 -conjugated particles coated with HSA prior to conditioned media pulldown, (4) 5 A7 -conjugated particles

Figure 20 Flow cytometry measurement of pancreatic cancer cell count versus fluorescence intensity. After serum starvation for 4 hours MIA PaCa-2 and PANC-1 cells were treated with HSA-preincubated particles at a polymer concentration of 2.5 pg mL ¹ for 45 mins prior to acid-stripping, washing, trypsinization and preparation for flow cytometry analysis. Particle conditions used were BA11 -conjugated and 2V-conjugated. Results are presented as relative to the mean fluorescence intensity of those cells incubated with 2V- conjugated HSA-preincubated particles. Statistical significance established via unpaired t- test (ns p>0.05, *p<0.05).

Figure 21 HUVEC tube morphology is altered upon treatment with anti-DLL4 VNAR conjugated NPs (A) Representative DLL4 Western blot of lysates derived from HUVECs cultured under varying oxygen conditions (5, 10 or 21% oxygen), beta-actin used as loading control (B) quantification of total mesh area from the below phase contrast images using Angiogenesis Analyzer plugin. Data pooled from 3 independent experiments and normalized to untreated control. Statistical significance determined by two-way ANOVA (

** p<0.01, **** p<0.0001). (C) representative 10x phase-contrast images of HUVECs taken 6 hours after seeding with no treatment (untreated), E4 NPs ( 100 or 500 nM of particle-bound VNAR) or a matched polymer concentration of nude NP. (D) quantification of total mesh area from the below phase contrast images using Angiogenesis Analyzer plugin. Data pooled from 3 independent experiments and normalized to untreated control. Statistical significance determined by two-way ANOVA (**** p<0.0001). (E) representative 10x phase-contrast images of HUVECs taken 6 hours after seeding with no treatment (untreated), E4 NPs (100 nM of particle-bound VNAR) or free E4 (100 nM). Detailed Description

The present invention provides specific antigen binding molecule-substrate conjugates. In particular, the present invention provides specific antigen binding molecules having an amino acid sequence represented by the formula FW1-CDR1-FW2-HV2-FW3a-HV4- FW3b-CDR3-FW4, in which FW1 , FW2, FW3a, FW3b, and FW4 are Framework Regions, CDR1 and CDR3 are Complementarity Determining Regions, and HV2, and HV4 are Hypervariable Regions, conjugated to a substrate.

The specific antigen binding molecule may also be referred to herein as a vNAR.

Examples of substrates include, but are not limited to, biomaterials, microparticles, and nanoparticles. Additional substrates including drug delivery devices, implantable medical devices, prosthetic implants, supports or the like are also contemplated herein. An implantable medical device can be formed of plastics material, metal or another biocompatible and/or biostable material as known in the art. Furthermore, the substrate can be a quantum dot or the like.

The substrate provides a surface for presentation or display of the vNAR. The substrate thus acts as a scaffold to arrange the vNAR on the surface of a device, for example a nanoparticle or microparticle or a biomaterial, which can be further used to deliver a therapeutic to a cell.

It will be appreciated that the vNARs are provided on the substrate in such a way that they are available to bind their cognate molecule/receptor. This may include the presentation of vNARs in manner permitting the binding and activation and/or modulation of a receptor function. The conjugation chemistry adopted to enable vNAR presentation provides vNAR molecules on the substrate at an appropriate concentration density to enable binding of the cognate antigen.

The conjugation may be via any suitable means. Preferably, the conjugation is via a covalent bond. Other forms of attachment are also possible and are expressly

contemplated herein.

As will be understood by those of skill in the art, the term“microparticle” refers to particles with a greatest cross-sectional width less than 1000 micrometres and which is greater than or equal to 1 micrometre. Such a microparticle may be particularly effective for use in the treatment of diseases of the lung. It will also be understood that the term“nanoparticle” refers to particles with physical dimensions on the nanometre scale, i.e., with a greatest cross-sectional width less than 1000 nanometres and which is greater than or equal to 1 nanometre. Nanoparticles may be referred to herein by the abbreviation NP. Typically, nanoparticles will have a diameter of from 50 nm to 500nm. More typically, nanoparticles will have a diameter of from 100 nm to 300 nm. However, it will be appreciated that in any given nanoparticle sample there may be significant variation in physical size and shape.

Nanoparticles have been used in a number of therapeutic applications, typically as drug delivery vehicles. However, they must be targeted to the cells intended for treatment in order to avoid off-target effects. More efficient targeting allows for lower doses, which in turn reduces unwanted side effects.

A large variety of nano-formulations have already been approved for clinical use in a number of indications, for example a 2016 study identified 51 FDA-approved

nanomedicines and 77 products in clinical trials²³. The vast majority of these are untargeted formulations, for example approved formulations in the field of systemically administered anti-neoplastic drug delivery includes Abraxane (albumin-bound paclitaxel), Doxil (liposomal doxorubicin) and Onivyde (liposomal irinotecan). Phase 3 trials of these formulations have shown either reduced toxicity as a monotherapy in direct comparison to free drug (Doxil) or improved survival when combined with existing drug regimens (Abraxane and Onivyde), where presumably the combination of free drug with existing drug regimens would lead to treatment-ending toxicity. In addition to the aforementioned shielding of drug from the general biological environment, a major cause of this reduced toxicity is the absence of toxic solvents such as Cremophor-EL²⁴ in order to solubilise hydrophobic chemotherapy drugs for systemic administration. Additionally NPs are used in a variety of other clinical settings, for example in imaging, (iron oxide NPs for MR imaging have been clinically approved since the mid-1990s²⁵), tissue engineering (e.g. hydroxyapatite NPs used as bone substitutes²⁶) and against infectious disease (e.g. Abelcet and AmBisome²⁶).

The presentation of the vNAR on the surface of a nanoparticle can provide for an increased uptake of the nanoparticle by a cell. The presentation of the vNAR on the surface of a nanoparticle can enable binding to a cognate receptor and in turn alter biological function.

The material from which the microparticle or nanoparticle is formed may be selected from a range of options. One such option is polymeric poly(lactic-co-glycolic acid) (PLGA). The composition of the nanoparticle may also include polyethylene glycol (PEG). Any suitable material as known in the art to form a microparticle or nanoparticle may be suitably utilised. For example gold, polystyrene, biodegradable polymers, liposomes, alginate, chitosan, albumin-drug complexes and quantum dots.

Liposomal nanoparticles are known delivery vehicles which can encapsulate therapeutic payloads and can display ligands on their surface. Suitable liposome formulations would be known to those in the art and appropriate chemistry to attach ligands (in the present invention a vNAR) to the surface of such liposomes is known.

Suitably, nanoparticles or microparticles can be parenterally administered. After parenteral administration, nanoparticles can selectively accumulate in particular tissues or body locations. Therefore, nanoparticle conjugates as described herein can deliver a

therapeutic payload to the cell or tissue. Nanoparticles can access diseased tissue through an enhanced permeability and retention effect. A pharmaceutical composition of the invention is formulated to be compatible with its intended route of administration.

Examples of routes of administration include parenteral, e.g. intravenous, intradermal, subcutaneous, oral (for example inhalation), transdermal (topical), transmucosal, and rectal administration.

The delivery system, microparticle, nanoparticle etc., may be a polymeric particle, in particular a particle may be formed from a biodegradable polyester such as poly(lactide) (PLA), poly(glycolide)(PGA), poly(butyl cyanoacrylate) (PBCA), or N-(2- hydroxypropyl)methacrylamide (HPMA) copolymers and poly(lactic-co-glycolic acid) (PLGA), which have been used in pharmaceutical and biomedical applications. Suitable polymers for use as a substrate, wherein the substrate can form a nanoparticle or a medical device will be known in the art. A polymeric particle can be formed of poly(lactic- co-glycolic acid) (PLGA). A polymeric particle may also incorporate polyethylene glycol (PEG). Additionally, a polymeric nanoparticle may incorporate a second polymer or copolymer surfactant or coating in addition to the first PLGA polymer or copolymer of the particle. This second polymer or copolymer surfactant or coating may bear terminal functional groups such as, maleimide, N-hydroxysuccinimide (NHS), allowing ligand (vNAR) conjugation. Suitably the second polymer or copolymer can be branched or linear. VNARs may be conjugated to the surface specifically to bind specific biomolecules in circulation, thereby controlling the corona which forms on the nanoparticle surface. The corona is defined as those biomolecules which associate with the previously‘pristine’ nanoparticle surface when the particle enters a biological environment. The nanoparticle may encapsulate one or more therapeutic agents. In particular, the nanoparticle may encapsulate one or more cytotoxic agents, including but not limited to camptothecin.

The nanoparticle may be conjugated to the specific antigen binding molecule via a free cysteine residue in the amino acid sequence of the specific antigen binding molecule. The free cysteine may be located anywhere in the amino acid sequence, including at either the N or C termini.

Preferably, the specific antigen binding molecule is conjugated to the nanoparticle via ethyl(dimethylaminopropyl) carboiimide/N-hydroxysuccinimide (EDC/NHS) chemistry, maleimide, vinyl sulfone (VS), or sulfonyl fluoride exchange (SuFEX).

Introduction of a single cysteine residue at either the N- or C-termini is highly amenable for bioconjugation both due to the privileged position of the residue (i.e. it is not buried within the VNAR framework and it does not have an intramolecular disulfide bonding partner) and the unique reactivity of cysteine among the natural amino acids: the thiol sidechain has the lowest pKa of the nucleophilic amino acid residues, e.g. thiol pKa is ~8, amino and hydroxyl pKa are ~9. The thiol functional group is“softer” (I.e. a higher-energy HOMO) than corresponding hydroxyls or amines, this leads to a selectivity for other“soft” electrophiles (those with correspondingly higher-energy LUMOs) in order to maximise orbital interaction.

One potential limitation of these cysteine mediated conjugation approaches however is that, cysteines, when located terminally, have been reported to be susceptible to decarboxylation and degradation²⁷. Additionally, histidine residues have been reported to catalyse degradation of amide bonds in recombinant proteins through coordination of various metal ions and promotion of either hydrolysis²⁸·²⁹ or radical-mediated redox reactions³⁰.

For the purposes of this work, such degradation would in turn minimise the efficiency of the vNAR to nanoparticle conjugation. In an attempt to circumvent this, the cysteine residue in the vNAR clones to be used for conjugation was located prior to a His-tag region and placed within an‘alanine motif. Thus the cysteine residue was flanked by either one (ACA) or two (AACAA) alanine residues on either side. Alanine was selected in this instance due to the relatively unreactive nature of its methyl side chain. This conformation was expected to reduce decarboxylation of the cysteine whilst maintaining its availability for reaction with the functional grouping on the nanoparticle corona. The afore mentioned“soft” electrophiles include various tt-systems, particularly those with electron-deficient positions which will readily react with thiols in the absence of catalyst at room temperature. Examples include vinyl sulfones and maleimides, which are both widely-known and utilised reactive partners for thiol bioconjugation³¹ and used here for vNAR conjugation to NPs. Synthesis of monodisperse vinyl sulfone functionalized polymers is possible through RAFT polymerization e.g. of styrene or organic acids such as lactic acid, followed by aminolysis of the residual terminal RAFT chain transfer agent and trapping of the resulting thiol with divinyl sulfone³²·³³. Polymers such as these can then be formed into nanoparticles displaying surface vinyl sulfone groups, for example herein a previously described vinyl sulfone functionalised PLA-P(OEGMA) polymer³⁴ was incorporated into nanoparticles alongside PLGA and VNAR conjugated to the resulting nanoparticles. VNARs can therefore be used alongside this facile method of functionalised polymer synthesis, potentially aiding the future manufacturability of VNAR-nanoparticle conjugates.

Unexpectedly a recently rediscovered reaction, sulfonyl fluoride exchange or SuFEx³⁵, was also found to be of utility in conjugating VNAR to NPs, displaying greater binding affinity to target protein than NHS-conjugated NPs (NHS chemistry being a known non- site-selective conjugation chemistry). This suggests that SuFEx conjugation may occur through the free C-terminal cysteine of the VNAR under the reaction conditions forming the thiosulfonate ester, whereas it might be expected that the much more

thermodynamically stable sulfonamide adduct would be formed. The observed behaviour may be attributable to the pH of the conjugation reaction (pH 5). At this pH Lys residues will be protonated, leaving only the free Cys residue able to act as a nucleophile. The rate constant of hydrolysis of the thiosulfonate ester adduct is also likely to be lower at this lower pH (by analogy to published data on the pH dependence³⁵ of thiosulfinate ester hydrolysis³⁶) allowing the VNAR nanoparticle conjugate product to accumulate. SuFEx chemistry is particularly sensitive to any factors that stabilise the expulsion of the fluoride ion from the sulfonyl fluoride group, e.g. acidic aqueous conditions, the ability of VNARs to remain stable under these conditions makes them a good partner for SuFEx conjugation.

A particularly attractive property of SuFEx-functionalised polymers for conjugation of nanoparticles to VNARs is the high stability of the sulfonyl fluoride group to an extensive range of conditions³⁵, raising the possibility of synthesising functionalised nanoparticles and performing VNAR conjugation with a significant delay, allowing more flexibility in manufacturability. It should also be noted conjugation via nucleophilic Cys is orthogonal to other highly efficient conjugation reactions such as, e.g. azide-alkyne cycloaddition either copper-catalysed (CuAAC) or strain-promoted (SPAAC), and therefore VNAR conjugation to particles via Cys may be used alongside these complementary conjugation

approaches.

Thiols also present the possibility of radical-mediated bioconjugation e.g. in the thiol-ene and thiol-yne reactions. Both of which require some form of radical initiator, either thermal or photochemical³¹. Radical reactions are normally avoided for bioconjugation due to fears about biomolecule stability under the reaction conditions. VNARs may be suitably stable in the presence of radicals, UV light etc. to enable these reactions to be used.

The specific antigen binding molecules described herein have an amino acid structure as follows:

FW 1 -C D R 1 - FW2- H V2- FW3a- H V4- FW3b-C D R3- FW4 in which FW1 , FW2, FW3a, FW3b, and FW4 are Framework Regions, CDR1 and CDR3 are Complementarity Determining Regions, and HV2, and HV4 are Hypervariable

Regions.

VNARs appear to possess a range of properties which would make them ideal for use as nanoparticle targeting moieties.

Compared to a full IgG (-150 kDa) the small size of a VNAR (12 kDa) makes them particularly susceptible to removal from systemic circulation by glomerular filtration, severely limiting their serum half-life¹⁶. Therefore, increasing the effective size of the VNAR by conjugating it to a functional partner, such as a nanoparticle, should aid in negating VNAR removal before it reaches its site of action. Furthermore, the small size of the VNAR allows higher relative loading onto the nanoparticle surface when compared to larger molecules such as a full IgG, in turn resulting in a smaller increase in nanoparticle diameter³⁷.

It has been demonstrated that greater surface ligand density can be achieved by using binding fragments instead of full binders such as antibodies, for example in the context of biosensors³⁸ (wherein an anti-gliadin fragment coated surface achieved a coverage of 2.72 pmol / cm² and the full antibody-coated surface 0.85 pmol / cm²) . In addition to their size and high binding affinity, vNARs are also amenable to nanoparticle conjugation due to their inherent stability. This is of benefit both during the conjugation process and post-conjugation, maximising the opportunity for the nanoparticle-VNAR conjugate to reach its target with vNAR functionality intact.

Without wishing to be bound by any one particular theory, the present inventors believe that due to the smaller size and less complex structure, vNARs are less prone to issues such as steric hindrance and more amenable to site-specific and correctly orientated conjugation. As a result, higher density of vNARs on biomaterial/nanoparticle surfaces is achievable.

An important feature of the procedure adopted herein is the ability to conjugate both native and humanised vNAR clones to a nanoparticle corona. Following conjugation, both native and humanised vNARs are shown to retain their ability to bind their cognate antigen. This functionality, is demonstrated using native and humanized binders of HSA (E06 and BA11 respectively) and a humanized binding clone (5A7), targeting HEL. Indeed the ability to utilise humanised vNAR clones in this way increases the possibility of using vNAR-targeted nanoparticles in a therapeutic setting.

It is well known that the non-antibody, VNAR paratope is pre-disposed to bind into pockets and grooves in proteins (canyon-binders) and that these cryptic epitopes are not available to antibodies with planar shaped binding sites made up of six CDR loops, 3 from the heavy and 3 from the light chain variable regions. The humanized anti-HEL VNAR binds into the pocket-shaped, active site of its enzyme target (reducing enzyme activity) a region of the protein not seen by antibodies specific for the same antigen. The anti-DLL4 VNAR binder also recognises a different epitope than the current clinical candidate antibody to this same angiogenesis target.

It was not obvious in these studies if specific, cryptic VNAR binding profiles (anti-HEL and anti-DLL4 in particular) would be retained when VNAR were conjugated to a bio-materials surface because of steric-hindrance caused by the conjugation process itself or because of the high density of binders achieved due to the VNARs small size and stable nature. Figures 3 (DLL4), 16 (HSA and hHSA) and 17 (hHEL), confirm that high affinity and specific binding was retained by all the VNAR and humanised variants of VNAR that were tested. A further technical effect of the specific antigen binding molecule-substrate conjugates of the invention is the provision of an enhanced biological effect compared to the specific antigen binding molecule alone. The specific antigen binding molecule alone may alternatively be termed the unconjugated specific antigen binding molecule or the free specific antigen binding molecule. Without being bound by theory, this is thought to arise from conjugation to a substrate displaying specific antigen binding molecules, and VNARs in particular, in a manner in which they are optimally orientated to bind their cognate epitope. This effect may be especially apparent wherein the specific antigen binding molecule binds to a target with a function mediated by receptor clustering. For example, this effect is demonstrated in Example 5 for the anti-DLL4 VNAR, E4.

Examples

Example 1 - Formulation of vNAR-conjugated PLGA-PEG Nanoparticles

In this current study we wished to utilize maleimide coupling of vNARs to the surface of polymeric PLGA NPs. Through the incorporation of a PLGA-PEG-Mal co-polymer within the nanoparticle, conjugation to a free cysteine in the vNAR fragments would be possible. Using an oil-in-water single emulsion evaporation approach, we successfully generated NPs, which upon assessment where found to be approximately 200 nm in diameter with a slightly negative zeta potential of -3 mV in PBS (Figure 1). Furthermore, the low polydispersity (0.1) suggested that a uniform monodisperse population of particles had been generated by this strategy.

To facilitate the conjugation of the vNAR to the maleimide moieties on the NP, the purified vNARs were incubated with the particles under gentle agitation for 2 h, allowing up to a 50% conjugation efficiency equating to 5 pg protein/mg formulation. To assess the ability of these conjugates to recognize DLL4, SPR analysis was carried out using hDLL4-Fc protein immobilized onto a carboxyl methylated dextran chip. Using equalized amounts of particles, we first demonstrated that nude control particles had no binding response towards the immobilized DLL4 protein, whereas the vNAR E4 and 10 conjugated NPs both elicited a positive relative response. Despite comparable levels of the vNARs conjugated to the nanoparticles, there was a 7 fold higher response with the E4 clone. To ensure that we were not simply measuring binding of vNAR aggregates in our

formulations, we also prepared conjugation controls wherein the vNAR was subjected to the same coupling procedure in the absence of nanoparticles. The lack of response in these samples confirmed that the binding observed with the vNAR-NPs was dependent on its association with the particle (Figure 2A). In a subsequent experiment, analysis of increasing concentrations of the vNAR E4 NP conjugates, revealed an eventual saturation of the chip (Figure 2B).

Analysis of vNAR-NP binding to DLL4 was further confirmed by modified ELISA, using maxisorb microtiter plates coated with DLL4-Fc protein. In these experiments, the incorporation of rhodamine 6G in the particle afforded fluorescent measurement of NP binding. Results from this analysis confirmed the earlier SPR findings, demonstrating that at equivalent concentrations, the E4 conjugated particles yielded superior binding to the 10 clone (Figure 3A).

To rule out the possibility that the E4 vNAR was simply adsorbed to the PLGA NP and not covalently attached, we then explored the binding characteristics of nanoparticles employing different proportions of the maleimide functionalized PLGA-PEG-Mal copolymer. Incubations with these varying formulations revealed that incorporation of increasing amounts of the maleimide functionalized copolymer yielded increasing binding (Figure 3B). Importantly, when 0% of the PLGA-PEG-Mal was used in the formulation, binding of the particle to the DLL4 coated plates was directly comparable to controls. This demonstrates that the binding of these particles to DLL4 is dependent on the site-specific, maleimide-cysteine conjugation of the vNAR to the particles.

The specificity of the conjugation approach was further demonstrated through comparison with NHS ester chemistry, wherein conjugation to the nanoparticle can occur at lysine residues throughout the vNAR structure. Whilst the coupling efficiency of the vNAR to the nanoparticle was similar using both chemistries, vastly superior antigen binding was observed with the maleimide chemistry. DLL4 binding, as determined via modified ELISA and SPR (Figure 4A-B), highlights that the site-specific nature of the maleimide conjugation approach plays a critical role in antigen binding; indicating improved paratope accessibility due to controlled orientation of vNAR fragments on the nanoparticle surface.

To further illustrate the specificity of the binding interaction between the anti-DLL4 vNAR targeted nanoparticles and their cognate antigen, modified ELISA binding was assessed following blockade of a portion of the available DLL4 binding sites with an anti-DLL4 antibody. Partial inhibition of vNAR nanoparticle binding highlights the specificity of the binding interaction between the targeted vNAR nanoparticle and its conjugate antigen (DLL4) (Figure 5).

In order to assess the suitability of the vNAR-nanoparticle conjugates for use in a biological setting, antigen binding was assessed via modified ELISA in the presence of biological media. Encouragingly, significantly enhanced binding of vNAR-targeted nanoparticles was maintained over non-targeted nude nanoparticles despite the presence of the biological media, indicating that the nanoparticle-vNAR conjugates could be successfully employed cell-based settings (Figure 6).

EXAMPLE 2 - ADDITIONAL VNAR CONJUGATION CHEMISTRIES: SUFEX AND VINYL SULFONE

To illustrate the suitability of the free C-terminal cysteine of the engineered vNAR for conjugation using additional conjugation chemistries techniques, sulfonyl fluoride exchange (SuFEx) and vinyl sulfone (VS) functional groups were investigated. SuFEx is a recently rediscovered reaction which has potential to be a“click” reaction. Whilst used in organic synthesis and materials chemistry there are few examples of its use in

bioconjugation. In contrast VS is an established cysteine-selective functional group, however maleimide is more commonly used for bioconjugation.

The SuFEx functional group was incorporated into a semi-telechelic PLGA-PEG-SH polymer purchased from PolySciTech. The VS functional group was introduced into a PLA POEGMA block polymer synthesized by ROP RAFT polymerization³⁴.

Conjugation of the E4 vNAR to nanoparticle formulations was evaluated using mBCA and modified ELISA binding assays. Both conjugation chemistries were compared to the known non-cysteine-selective functional group N-hydroxy succinimide (NHS) ester and were found to give particle formulations with superior binding to DLL4. This is potentially due to site-selective conjugation to the single accessible cysteine of the engineered vNAR and subsequent optimal orientation of the paratope. Taken together, these results illustrate the potential for using alternative conjugation techniques for producing vNAR- targeted nanoparticles (Figure 7 & 8).

EXAMPLE 3 - DLL4-TARGETED CELL BASED ASSAYS

Next we wished to determine if the selective binding of the vNAR-targeted NPs could have cellular activity and be used to actively target encapsulated payloads to cells. Firstly, we examined HUVEC cells, which are known to express DLL4 in hypoxic conditions.

Western blot analysis of HUVEC cells demonstrated weak DLL4 positivity. It has previously been shown that DLL4 expression can be stimulated in endothelial cells with FGF and VEGF.³⁹ We therefore stimulated the cells with these pro-angiogenic growth factors for 24 h, demonstrating that only the combination of both produced a synergistic upregulation of DLL4 (Figure 9). To analyze the effect of targeting this up-regulated DLL4 expression, cellular uptake of the vNAR-targeted nanoparticles was assessed. HUVEC cells were incubated with either nude or E4 targeted PLGA-PEG-Mal nanoparticles, loaded with rhodamine, for 30 min at 4 °C. This approach has been well characterised to discriminate between particle absorption versus active uptake using in vitro studies.⁴⁰ Unbound nanoparticles were then removed by washing before 3 h incubation at 37 °C to enable the cellular uptake of surface bound nanoparticles. Greater uptake of the E4-functionalized nanoparticles when compared to the nude nanoparticles highlights the impact of the anti-DLL4 vNAR in facilitating increased nanoparticle internalization (Figure 10).

Next, we prepared our nanoparticles encapsulating the cytotoxic agent camptothecin (CPT) (Figure 1 1A) and incubated these with the stimulated HUVECs for 1 h at 4 °C, before washing and replacing with fresh media. Measurement of cell viability at 24 h highlighted a significant decrease in cell viability for the vNAR-targeted nanoparticles over non-targeted (nude) controls (Figure 1 1 B). This indicated that the targeting of this protein could be exploited to provide active targeting of the drug-loaded nanoparticles to these cells.

We also examined the ability of the vNAR-targeted CPT loaded nanoparticles to target DLL4 positive A549 cells. These cells were treated as before with vNAR-targeted, CPT- loaded, nanoconjugates. A highly significant cytotoxic effect (Figure 12) in excess of non- targeted control CPT loaded NPs was apparent 24 h post-treatment.

Furthering this work, DLL4 expression and the subsequent potential application of anti- DLL4, vNAR-targeted nanoparticles were investigated in a panel of pancreatic cell lines. DLL4 overexpression in peritumoural vessels of pancreatic adenocarcinoma is well documented. DLL4 blockade has been shown to cause non-productive sprouting of vasculature and is a potential mode of therapy in this context. By using anti-DLL4 vNARs as targeting agents for nanoparticles there is an opportunity to both provide DLL4 blockade but also enhance accumulation of nanoparticles at the tumour site.

Accordingly, several pancreatic cancer cell lines were screened for DLL4 expression by flow immunophenotyping with a patented anti-DLL4 antibody, YW26.82: MIA PaCa-2 and Panc-1 cells were found to be DLL4^hi9h, and ASPC-1 were DLL4^|0W. Again, A549s, a non small cell lung cancer cell line were screened, this cell line displayed two populations, one of which was DLL4^hi9h and the other DLL4^|0W .

Rhodamine 6G-encapsulated PLGA-PEG-Mal and PLGA-PEG-Mal-vNAR nanoparticles were subsequently prepared as previously described with binding and uptake of E4 conjugated and nude particles assessed by fluorescent plate reader and flow cytometry methods. A549, BXPC-3, MIA PaCa-2 and Panc-1 cell lines all displayed greater uptake of E4 conjugated particles than either nude or 2V (isotype control vNAR) conjugated particles as quantified by fluorescent plate reader.

Binding of E4-conjugated nanoparticles to MIA PaCa-2 and Panc-1 cells was shown to correlate with surface expression level of DLL4, providing circumstantial evidence that binding is DLL4 mediated (Figure 13). However, the greater binding of E4 nanoparticles to Panc-1 over MIA PaCa-2 cells does not seem to translate to greater nanoparticle uptake, possibly due to endocytic differences between the two cell lines.

The importance of DLL4 in mediating binding of E4-conjugated particles was further evidenced by preblocking cells with free E4 vNAR (Figure 14). This should reduce free DLL4 on the cell surface and concomitantly reduce DLL4 dependent nanoparticle binding. As anticipated, binding of E4-conjugated particles to A549 cells reduced as the dose of free E4 preblock was increased.

Functional assessment of this preblock was also performed by clonogenic assay in ASPC- 1 , where blocking of DLL4 via preincubation with free E4 decreased toxicity of E4- conjugated CPT-encapsulated particles indicating decreased DLL4-mediated uptake (Figure 15).

EXAMPLE 4 - Additional vNAR-nanoparticle conjugates

In order to demonstrate the robustness and specificity of the maleimide-thiol conjugation approach, vNARs targeting a range of other targets (not DLL4) were investigated. Two vNAR clones (E06 and humanized BA11) targeting human serum albumin (HSA) and an anti-hen egg lysozyme vNAR (5A7) were generated and engineered to bear a c-terminal cysteine group to permit site-specific, orientation-controlled, conjugation to the reactive maleimide group on the nanoparticle surface (Figure 16-17). Once again, it was decided that PLGA-PEG-NHS nanoparticles would serve as a suitable control of non-site-specific conjugation approaches. Despite the similarity in the amount of vNAR conjugated, binding response was notably more pronounced with the maleimide conjugation chemistry for all of the vNARs investigated. This consistency presents the maleimide conjugation approach detailed as a robust platform for producing PLGA nanoparticles bearing any vNAR possessing a c-terminal cysteine.

Further physicochemical and functional characterisation of PLGA-PEG-Mal-BA11 particles was subsequently carried out. BA11 particles and nude/2V negative control particles were incubated with HSA dissolved in PBS or PBS alone and assessed in terms of particle diameter and polydispersity by dynamic light scattering and nanoparticle tracking analysis (Figure 18). After HSA or mock incubation particles were washed twice. This involved centrifugation to form a particle pellet, discarding the supernatant and subsequent resuspension of the particle pellet by sonication in PBS. Incubation time in Figure 18 was maximised (3 hours) to permit equilibration of the nanoparticle corona before washing and characterisation. microBCA assays indicated greater association of HSA with BA 11 -conjugated particles than control particles (Figure 18). There was no significant difference in conjugation of BA11 and 2 V vNAR to PLGA-PEG-Mal particles, however after HSA incubation and particle washing there was significantly more protein associated with the BA11 particles, confirming the HSA binding ability of BA11.

Despite the abundance of endogenous albumin in vivo, it is a putative targeting moiety for nanomedicines. Tumours fulfil a requirement for energy/amino acids through enhanced albumin uptake, this is in part due to overexpression of Secreted Protein Acidic and Rich in Cysteine (SPARC). This matricellular protein is secreted into the tumour

microenvironment where it accumulates through interactions with collagen. SPARC also binds albumin and seems to participate in an albumin shuttle mechanism whereby albumin is brought into cells. Given the possible significance of SPARC in determining the efficacy of albumin-associated chemotherapies, the particle formulations characterised above were assessed for their ability to bind SPARC. First MIA PaCa-2 and Panc-1 cells were screened for SPARC expression, with the former found to be SPARC-negative and the latter SPARC-positive. BA11 and 5A7 vNARs were incubated with Panc-1 conditioned media rich in SPARC, before the particles were pelleted, washed and analysed by Western blot against SPARC. BA11 particles were found to associate with SPARC, precoating with HSA of BA11 particles was found to increase SPARC association. 5A7 particles showed negligible association with SPARC in either HSA-precoated or uncoated conditions (Figure 19). This evidence demonstrates the unexpected finding that HSA is able to simultaneously bind BA11 and SPARC, raising the possibility that BA11- conjugated particles may be used to bind HSA and then bind SPARC in environments where SPARC is abundant (i.e. many cancer microenvironments), potentially increasing nanoparticle accumulation there.

Following on from this result, uptake of rhodamine 6G encapsulated BA11 and 2V particles incubated with HSA was studied in SPARC positive Panc-1 cells and SPARC negative MIA PaCa-2 cells (Figure 20). In the Panc-1 cells there was greater uptake of BA11 than 2V particles, however in MIA PaCa-2 the uptake of the 2 formulations was the same, potentially due to the expression of SPARC in the Panc-1 model favouring the uptake of the HSA-coated BA11 particles.

EXAMPLE 5 - HUVEC tubulogenesis assay

10 pL of growth factor reduced Matrigel (Corning) was added to the lower well of a 15-well tissue-culture treated Angiogenesis m-slide (Ibidi) and allowed to polymerise for 2 hours. 3000 HUVECs at around 70% confluency and passage <12 were then seeded in 50 pL of treatment containing media on top of the Matrigel disc in the upper chamber of each well. Networks of tubes were allowed to form for 5-6 hours and imaged via phase-contrast microscopy (EVOS, 10x).

Given the endogenous function of DLL4 in angiogenesis it would be expected that DLL4 blockade by VNAR NPs would impact angiogenic signalling. Bearing this in mind, we investigated the effect of E4-conjugated particles in a tubulogenesis assay using

HUVECs. In addition to growth factor stimulation, DLL4 expression is upregulated under hypoxic conditions. Such conditions are frequently found in the tumour microenvironment, and were therefore used here to replicate physiologically relevant enhancement of DLL4 expression (Figure 21 A). Following treatment with E4-conjugated NPs there is a marked difference in HUVEC tube phenotype (Figure 21 B & C). Quantification revealed no significant difference in branching interval between E4 NP treatments at 100 and 500 nM of particle-bound VNAR, indicating that anti-DLL4 effect of the E4 NPs was saturating under the assay conditions at 100 nM . Free E4 VNAR was also compared to particle bound VNAR at 100 nM (Figure 21 D & E).

Equimolar concentrations of conjugated and free anti-DLL4 vNAR were assessed in terms of their ability to modulate HUVEC tubulogenesis. It was noted that the conjugated anti- DLL4 vNAR had a more pronounced effect on HUVEC tube morphology than when unbound. This discovery highlights an additional benefit of utilizing polymeric

nanoparticle scaffolds to display vNARs in a manner in which they are optimally orientated to bind their cognate epitope. This approach could be beneficial for a range of other vNARs against biological targets, particularly in instances where receptor clustering is required to elicit a functional effect.

References

1. Greenberg et al. Nature. 1995;374:168-173. 2. Dooley H et al. Mol Immunol. 2003;40:25-33.

3. Muller MR et al. MAbs. 2012;4(6):673-685.

4. Barelle C., Gill D.S. CK. Shark Novel Antigen Receptors - The Next Generation of Biologic Therapeutics? In: Pharmaceutical Biotechnology. Advances in

Experimental Medicine and Biology. 655th ed.; 2009.

5. Barelle C, Porter A. Antibodies. 2015;4:240-258.

6. Flajnik MF, Kasahara M. Nat Rev Genet. 2009; 11 :47.

7. Nuttall SD et al. FEBS Lett. 2002;516(1-3):80-86.

8. Stanfield RL et al. Science (80- ). 2004;305(5691): 1770-1774.

9. Nuttall SD et al. Eur J Biochem. 2003;270(17):3543-3554.

10. Ubah OC, Steven J, Kovaleva M, et al. Front Immunol. 2017;8: 1780.

1 1. Kovaleva M et al. Front Immunol. 2017;8: 1 121.

12. Wang X et al. Protein Cell. 2018;9(1):63-73.

13. Leabman MK et al. MAbs. 2013;5(6):896-903.

14. Vafa O, Gilliland GL, Brezski RJ, et al. Methods. 2014;65(1): 114-126.

15. Cheng WWK, Allen TM. J Control Release. 2008;126:50-58.

16. Zielonka S et al. MAbs. 2015;7(1): 15-25.

17. Jain S, Doshi AS, Iyer AK, Amiji MM. J Drug Target. 2013;21 (10):888-903.

18. Goren D et al. Clin Cancer Res. 2000;6(May): 1949-1957.

19. Chen WC et al. Blood. 2010; 1 15(23):4778-4787.

20. Shan D, Li J, Cai P, et al. Drug Deliv Transl Res. 2015;5: 15-26.

21. Farokhzad OC, Cheng J, Teply BA, et al. PNAS. 2006; 103(16):6315-6320.

22. Schmid D et al. Am Soc Gene Cell Ther. 2014;22(12):2083-2092.

23. Bobo D et al. Pharm Res. 2016;33(10):2373-2387.

24. Gelderblom H et al. Eur J Cancer. 2001 ;37(13): 1590-1598.

25. Corot C, Robert P, Idee J-M, Port M. Adv Drug Deliv Rev. 2006;58(14): 1471-1504.

26. Ventola CL. Pharmacy & Therapeutics. 2017;42(12):742-755.

27. Morais M, Nunes JPM, Karu K, et al. Org Biomol Chem. 2017; 15:2947-2952.

28. Rajkovic S et al.. Dalt Trans. 2009;(39):8370-8377.

29. Milovic NM, Kostic* NM. J Am Chem Soc. 2002; 124(17):4759-4769.

30. Andberg M, Jantti J, Heilimo S, et al. Protein Sci. 2007; 16(8): 1751-1761.

31. Stenzel MH. ACS Macro Lett. 2013;2(1): 14-18.

32. Grover GN et al. Macromolecules. 2009;42(20):7657-7663.

33. Alconcel SNS et al. Aust J Chem. 2009;62(11): 1496-1500.

34. Themistou E, Battaglia G, Armes SP. Polym Chem. 2014;5(4): 1405-1417.

35. Dong J et al. Angew Chemie Int Ed. 2014;53(36):9430-9448.

36. Nagy P, Ashby MT. Chem Res Toxicol. 2007;20(9): 1364-1372.

Claims

1. A specific antigen binding molecule-substrate conjugate, comprising:

(a) a specific antigen binding molecule having an amino acid sequence represented by the formula FW1-CDR1-FW2-HV2-FW3a-HV4-FW3b-CDR3-FW4 in which FW1 , FW2, FW3a, FW3b, and FW4 are Framework Regions, CDR1 and CDR3 are Complementarity Determining Regions, and HV2, and HV4 are Hypervariable Regions; and

(b) a substrate.

2. The specific antigen molecule-substrate conjugate of claim 1 , wherein the substrate comprises a biomaterial, a microparticle, or a nanoparticle.

3. The specific antigen molecule-substrate conjugate of either claim 1 or claim 2, wherein the specific antigen binding molecule has been humanised.

4. The specific antigen binding molecule-substrate conjugate of any one of claims 1 to 3, wherein the specific antigen binding molecule further comprises an alanine motif represented by the amino acid formula ACA or AACAA, wherein the cysteine residue is available for conjugation to the substrate.

5. The specific antigen binding molecule-substrate conjugate of claim 4, wherein the alanine motif is located at the N-terminus or C-terminus of the specific antigen binding molecule.

6. The specific antigen binding molecule-substrate conjugate of any one of claims 1 to 5, wherein the specific antigen binding molecule has an amino acid sequence represented by any one of SEQ ID NOS 1 to 6.

7. The specific antigen binding molecule-substrate conjugate of any one of claims 1 to 6, wherein the specific antigen binding molecule is conjugated to the substrate via maleimide, NHS, VS, or SuFEX.

8. The specific antigen binding molecule-substrate conjugate of any one of claims 1 to 7, wherein the substrate is a PLGA nanoparticle.

9. The specific antigen binding molecule-substrate conjugate of any one of claims 1 to 8, wherein the specific antigen binding molecule recognises a cryptic or hidden epitope.

10. The specific antigen binding molecule-substrate conjugate of any one of claims 1 to 9, wherein the specific antigen binding molecule binds to DLL4 or HSA.

1 1. The specific antigen binding molecule-substrate conjugate of any one of claims 1 to 10, wherein substrate is a nanoparticle, further wherein the specific antigen binding molecule binds to HSA and binding causes alteration of the composition of the

nanoparticle corona.

12. The specific antigen binding molecule-substrate conjugate of any one of claims 1 to 1 1 , wherein the specific antigen binding molecule-substrate conjugate binds to HSA and HSA is still capable of functional binding to the protein SPARC.

13. The specific antigen binding molecule-substrate conjugate of any one of claims 1 to 12 for use in therapy.