Recombinant clever-1 fragment, method for producing recombinant clever-1 fragment, and uses thereof

Abstract

A recombinant Clever-1 fragment and a method for producing a recombinant Clever-1 fragment, which fragment mimicking a secreted form of Clever-1 (sClever-1) enriched in plasma of cancer patients, and has observed to have T cell-binding activity and is capable of binding to insulin-like growth factor 2 receptor (IGF2R) By its mechanism of action, a recombinant Clever-1 fragment of the present invention is suitable for use in a treatment of inflammation and / or autoimmune diseases.

Description

[0001] RECOMBINANT CLEVER-1 FRAGMENT, METHOD FOR PRODUCING

[0002] RECOMBINANT CLEVER-1 FRAGMENT, AND USES THEREOF

[0003] Technical field

[0004] The present invention relates to a recombinant Clever-1 fragment and a method for producing a recombinant Clever-1 fragment, which fragment mimicking a secreted form of Clever-1 (sClever-1 ) enriched in plasma of cancer patients. The invention also relates to use of a recombinant Clever-1 fragment in a treatment of various diseases and disorders. of the invention

[0005] In recent years, the contribution of scavenger receptors to regulating macrophage responses has attracted significant attention. Scavenger receptors form a vital part of host defence and homeostasis by removing extraneous or modified self- and non-self-macromolecules and acting as coreceptors in the priming of effector immune responses when danger signals are detected. Common lymphatic endothelial and vascular endothelial receptor-1 (Clever-1 , also known as Stabilin-1 and FEEL-1 ) is a multifunctional molecule that contributes to scavenging in a subset of some of the most antiinflammatory macrophages. In these cells, Clever-1 is involved in receptor- mediated endocytosis and recycling, intracellular sorting, and transcytosis of altered and normal self-components. Clever-1 is upregulated when immunosuppression is needed to protect the body. A growing tumour harnesses the immune system for its own development and is enriched with Clever-1 positive macrophages that support the formation of an immunosuppressive tissue microenvironment. Mechanistically, Clever-1 inhibits macrophage pro-inflammatory cytokine secretion and antigen presentation, thereby suppressing anti-tumour immunity mediated by CD8+T cells. While it has been demonstrated that genetic silencing or antibody- mediated blockade of Clever-1 in human monocytes can promote IFNy production in T cell antigen recall assays, it is still not clear whether Clever-1 itself has direct tolerogenic potential or these effects are mediated via an increase in pro-inflammatory monocyte activity during Clever-1 blockade. of the invention

[0006] It has now been found that a secreted form of Clever-1 (sClever-1 ) is enriched in plasma of cancer patients compared to undiagnosed healthy subjects. Specifically, it has now been found that a -200 kDa form of sClever-1 shows increased abundance in cancer patient plasma compared to healthy plasma. This form of sClever-1 is smaller than the full-length Clever-1 protein and is heavily truncated at the C-terminus. Detectable levels of sClever-1 protein are found both from human blood and lymph in microvesicles, exosomes and as soluble protein. At least macrophages and endothelial cells produce sClever- 1 into blood and in elevated amount in cancer patients than in healthy subjects.

[0007] The present disclosure provides a polypeptide mimicking the said form of sClever-1 , and which polypeptide has T cell binding activity and consists of an amino acid sequence of SEQ ID NO: 1 .

[0008] In the present invention, it has been found a fragment mimicking the said form of sClever-1 , which can be recombinantly produced by obtaining a stable recombinant Clever-1 fragment, which shows T cell binding activity, as presented in the Experimental part of the present disclosure.

[0009] Hence, the present disclosure provides a recombinant Clever-1 fragment (also denoted as H1 ), which mimicking sClever-1 enriched in plasma of cancer patients. Further, the present disclosure provides a method for producing the said recombinant Clever-1 fragment, and its use in a treatment of various diseases and disorders.

[0010] The present disclosure further provides an isolated polynucleotide that encodes the polypeptide having T cell binding activity and consists of an amino acid sequence of SEQ ID NO: 1 . Further, the present disclosure provides an isolated polynucleotide that encodes the polypeptide having T cell binding activity and consists of an amino acid sequence of SEQ ID NO: 1 , wherein the isolated polynucleotide has a nucleotide sequence comprising or consisting of a nucleotide sequence that has at least 90 %, identity to SEQ ID NO: 2. In one preferred embodiment according to the present invention, an isolated polynucleotide encoding the polypeptide mimicking sClever-1 and having T cell binding activity and consisting of an amino acid sequence of SEQ ID NO: 1 , has a nucleotide sequence of SEQ ID NO: 2. The polynucleotide according to the present invention is used for producing a recombinant Clever-1 fragment (H1 ). Additionally, the present disclosure provides an expression vector comprising the polynucleotide and a recombinant host comprising said expression vector. Also a method for producing the recombinant Clever-1 fragment having T cell binding activity is disclosed, wherein method comprises cultivating the recombinant host under conditions conducive for production of the recombinant Clever-1 fragment and recovering the recombinant Clever-1 fragment.

[0011] The recombinantly produced Clever-1 fragment (H1 ) according to the present invention comprising or consisting of an amino acid sequence of SEQ ID NO: 1 , preferably a recombinant Clever-1 fragment consisting of an amino acid sequence of SEQ ID NO: 1 . The recombinant Clever-1 fragment according to the present invention mimics a -200 kDa form of sClever-1 that has found to be abundantly enriched in plasma of cancer patients compared to undiagnosed healthy subjects. It has been found that the recombinant Clever- 1 fragment according to the present invention has T cell-binding activity and it has capability to bind to insulin growth factor 2 receptor (IGFR2) on activated T cells via its mannose-6-phosphate (M6P) modification.

[0012] Further, the recombinant Clever-1 fragment according to the present invention comprising or consisting of the amino acid sequence SEQ ID NO: 1 is suitable for use as a medicament. More specifically, the recombinant Clever-1 fragment according to the present invention comprising or consisting of the amino acid sequence SEQ ID NO: 1 is suitable for use in treatment of inflammation and / or autoimmune diseases.

[0013] Lastly, the present disclosure provides a pharmaceutical composition, wherein the composition comprises the recombinant Clever-1 fragment (H1 ) comprising or consisting of the amino acid sequence SEQ ID NO: 1 , and a pharmaceutical acceptable carrier or excipient.

[0014] The objects of the invention are achieved by the products, methods and uses characterized by what is stated in the independent claims. The preferred embodiments of the invention are disclosed in the dependent claims. Other objects, details and advantages of the present invention will become apparent from the following detailed description. The embodiments and advantages mentioned in this description relate, where applicable, both to the product, the method as well as to the uses according to the invention, even though it is not always specifically mentioned.

[0015] Brief description of the drawings

[0016] The invention will be described in more detail with reference to appended drawings, in which

[0017] Figure 1 shows that circulating soluble forms of Clever-1 (sClever-1 ) are enriched in cancer patients. Fig. 1A: Frequency distribution and Fig. 1 B: median of sClever-1 concentration in plasma of healthy donors (n = 32, light grey; the lowest part of the bars in 15-45 ng / mL bars in Fig. 1A), patients with treatment naive breast cancer (BC) (n = 138, dark grey; the central part of the bars in Fig. 1A) and patients with various advanced solid tumors participating in the MATINS trial (n = 193, the upper part of the bars in Fig. 1A). One-way ANOVA with Dunnett’s multiple comparison test. Fig. 1 C: ROC-curves of plasma sClever-1 concentration in BC and MATINS patients. AUC for sClever- 1 was 0.84 (Cl 0.77-0.90) in BC and 0.85 (Cl 0.80-0.90) in MATINS. Specificity for both cohorts at 48.7 ng / mL was 100%. Fig. 1 D: Quantification (ELISA) of sClever-1 in vesicle fractions in KG-1 cell culture supernatant, blood and lymph separated by ultracentrifugation. Samples were normalized to 1 pg / pL of protein. The graphs represent one independent assay with three technical replicates. The lowest part in the bars is microvesicles fraction, the central part is exosomes fraction and the upper part is soluble fraction. Fig. 1 E: Western blot analysis of sClever-1 detected with 3-372 mouse anti-human Clever-1 antibody in vesicle and liquid fractions of plasma and cell culture supernatants of primary human macrophages (M0, no polarization; M1 , IFNy + LPS; M2, IL- 4 + dexamethasone), endothelial cells (HLEC, human lymphatic endothelial cells; HPMEC, human pulmonary microvascular endothelial cells) and cell lines (KG-1 with / without PMA and HEK293). Fig. 1 F: Western blot showing the abundance of a -200 kDa sClever-1 species in healthy donors (n = 3) and MATINS patients (n = 4). # denotes the vesicle fraction. CD63 was used as a loading control. Fig. 1 G: Mass spectrometry analysis of Coomassie stained bands immunoprecipitated from human serum with Bexmarilimab or 9-11 anti- Clever-1 antibodies and relative isotype controls hlgG4 and rlgG2a, respectively. The cartoons show Clever-1 -specific peptide hits mapped to the full-length Clever-1 protein for Bexmarilimab (light grey) and 9-11 (light grey) pulldown. Venn diagrams depict shared peptides identified by Bexmarilimab and 9-11 .

[0018] Figure 2 shows quality control of recombinant Clever-1 fragment H1 . Fig. 2A: The extracellular domain architecture of Clever-1 . FAS1 , fascicl in-l i ke domain; EGF-like, epidermal growth factor -like domain. Fig. 2B: Coomassie-stained SDS-PAGE of H1 under reducing (R) or non-reducing (NR) conditions. H1 is denoted by the arrow. Marker sizes are shown on the left in kDa. Fig 2C: SEC- MALS elution profile of H1 (black line) showing the light scattering-derived absolute MW (solid grey line). The individual protein and glycan components are shown with grey dash and dot lines, respectively. Fig 2D: Experimental and simulated far-UV CD spectrum of H1 (black line) and its AlphaFold2 prediction (grey dash), respectively. The average secondary structure contents are denoted by H (a-helix), S (P-sheet), T (turn), and O (other). Fig. 2E: Thermal unfolding of H1. The fluorescence ratio and aggregation signals are indicated in the top and bottom graphs (black circle), respectively. First differential of the fluorescence signal (grey line) indicates the melting temperature (Tm). Fig 2F: Distribution of H1 particles sizes (black line). The intensity-weighted average hydrodynamic radius (Rh) is indicated by a grey dash. Fig 2G: Baseline-corrected SEC-SAXS elution profile (black line) showing the radius of gyration (Rg, grey circles) of H1 derived from a Guinier approximation (top left). Deconvoluted intensity profile (black circle) with a Bayesian fit indicated by a grey line (top right). The inset shows the Guinier fit (grey line) between 0.65 < qRg < 1.30 limits (black circle). Pair-distance distribution (grey line) with real-space-derived Rg, maximum particle diameter (Dmax), and MW (bottom left). Dimensionless Kratky plot (bottom right) of the intensity data (black circle) and the respective Bayesian fit (grey line). Grey dash reference lines denote the expected peak position of an idealised compact, globular protein.

[0019] Figure 3 shows that Clever-1 binds activated lymphocytes during transient phosphatidylserine exposure. Fig 3A: Flow cytometry plots and Fig 3B: quantification of H1 binding on primary T cells after 48 hours of activation with anti-CD3 / CD28 and IL2 (FCShighgate, light grey; n = 4 donors). Streptavidinphycoerythrin (SA-PE) and biotinylated human lgG4 served as negative controls for H1 staining. Fig 3C: Co-staining of activated T cells with Annexin V (to detect cell surface phosphatidyl serine) and FAM-VAD-FMK (to detect pre-apoptotic events) showing apoptotic cells in the FCS|OWgate. MFI, median fluorescence intensity. One-way ANOVA with Dunnett’s multiple comparison test (B and C). Fig 3D: Pre-incubation of activated T cells with either H1 or Annexin V 30 min prior staining cells with indicated reagents. Fig 3E: Flow cytometry analysis of H1 binding on lymphocytes obtained from healthy donors (HD, n = 7) and breast cancer patients (BC, n = 10). Fig 3F: Analysis of CD45RO expression in CD8- and CD8+ cells binding H1 in comparison to H1 negative cells. Paired Student’s t-test. Fig 3G: Expression of PD-1 in H1 positive and negative cells from HD and BC patients. Paired Student’s t-test.

[0020] * P <0.05, ** P <0.01 , *** P <0.001 , **** P <0.0001 .

[0021] Figure 4 shows that Clever-1 promotes the differentiation of suppressive FoxP3+CD8+T cells. Fig 4A: Proliferation of CD8+T cells after six days in culture with H1 (50 pg / mL) as assessed by CFSE dilution and quantification of cell numbers relative to control (BSA) treated cells (n = 8 donors), The results of control are left bars and recombinant Clever-1 fragment H1 are right bars. Fig 4B: Flow cytometry plots showing CD8+populations from which T cell lineage markers T-bet and FoxP3 were quantified (n = 7 donors). Paired t-test. Fig 4C: Representative Western blot of Lek phosphorylation at different time points after T cell activation with or without H1 . Representative blots from four independent experiments. Fig 4D: Jurkat NFKB reporter activity induced after TCR-ligation with or without H1 . Graph showing four independent experiments performed in triplicates. Fig 4E: Cytokine analysis of primary T cell cultures incubated with or without H1 (n = 10 donors). Significance shown as FDR- adjusted q-values for multiple t-tests. Fig 4F: Schematic of the assay design in which primary T cells were activated with or without H1 for six days and in parallel monocytes from the same donors were differentiated into macrophages. Thereafter, the T cells were washed and incubated with the macrophages for 48h and polarization markers expressed by macrophages (CD64+) were assessed with flow cytometry (n = 6 donors). Paired t-test. * P <0.05, *** P <0.001.

[0022] Figure 5 shows that Clever-1 binds to insulin-like growth factor 2 receptor (IGF2R) via mannose-6-phosphate. Fig 5A: Flow cytometry plots showing H1 and Annexin V binding on Jurkat cells after apoptosis induction (5 pg / mL of puromycin for 3 hours). Fig 5B: STRING database network summary of mass spectrometry proteomics results from pulldowns with biotinylated-H1 from Jurkat cell membranes. Fig 5C: Flow cytometry analysis of cell surface IGF2R expression on ionomycin stimulated primary human T cells (n = 2 donors). Experiment repeated twice. Fig 5D: Western blots of pulldowns with biotinylated-H 1 detected with anti-IGF2R. The primary T cells were stimulated with ionomycin for 30 minutes before performing the pulldowns. Control samples contained beads only. Fig 5E: Octet BLI analysis of the binding of H1 to IGF2R domains 1 -10 and 11 -13. Single cycle kinetics data was fitted using a 1 :1 Langmuir binding model with reference subtraction, to estimate the binding affinity. Fig 5F: Flow cytometry of H1 binding to Jurkat cells in the presence of excess mannose-6-phosphate (M6P) or after alkaline phosphatase (AP) treatment. Fig 5G: Flow cytometry of CRISPR / Cas9 edited Jurkat cells for H1 binding. gRNA against Rosa26 was included as a negative control. Fig 5H: Western blot to detect M6P on endogenous serum sClever-1 with recombinant IGF2R-His-AviTag protein. Biotinylated 9-11 antibody was used to pulldown sClever-1 from either human AB serum, or human AB serum depleted of sClever-1 by pulldown with 9-11 (confirmed via ELISA). Equal loading was confirmed with an anti-rat secondary antibody. Note, due to the non-reducing conditions of the gel, any streptavidin that leached from the Dynabeads would remain in complex with the biotinylated 9-11 antibody, creating a band larger than the anticipated 150 kDa for an lgG2a.

[0023] Figure 6 shows amino acid sequence of SEQ ID NO: 1 according to the present invention.

[0024] Figure 7 shows eight different fragments spanning the Clever-1 protein sequence. The Clever-1 protein has seven fasciclin-like (FAS1 ) domains and nine epidermal growth factor -like (EGF-like) domain. H1 refers to the first half of Clever-1. Dark grey “Y” refers to anti-Clever-1 antibody Bexmarilimab and denotes the binding side of Bexmarilimab on the Clever-1 protein sequence; light grey Y refers to anti-Clever-1 antibody 9-11 , and denotes the binding side of 9-11 on the Clever-1 protein sequence.

[0025] In the present disclosure, the following sequences are presented SEQ ID NO:1 discloses an amino acid sequence of a polypeptide according to the present invention and a recombinant Clever-1 fragment H1 according to an embodiment of the present invention.

[0026] SEQ ID NO: 2 discloses a nucleotide sequence (cDNA) coding the amino acid sequence of a recombinant Clever-1 fragment H1 according to an embodiment of the present invention.

[0027] SEQ ID NO: 3 discloses an amino acid sequence of a recombinant Clever-1 fragment H1 with the insulin signal peptide at C-terminus, and Hiss-tag at the N-terminus according to an embodiment of the present invention.

[0028] SEQ ID NO: 4 discloses a nucleotide sequence coding the amino acid sequence of a recombinant Clever-1 fragment H1 with a Kozak sequence, a start codon and the insulin signal peptide at 5’ end, and Hiss-tag at the 3’ end according to an embodiment of the present invention.

[0029] Detailed description of the invention

[0030] With the context of this disclosure, the term “soluble Clever-1 ”, “a secreted form of sClever-1 ” or “sClever-1” refers to Clever-1 protein present in a soluble form in circulation and / or lymph.

[0031] With the context of this disclosure, the term “treatment” or “treating” refers to complete curing of a disease or disorder, as well as amelioration or alleviation of said disease or disorder.

[0032] With the context of this disclosure, the term “codon” refers to three nucleotides corresponding to a single amino acid. With the context of this specification, the term “synonymous codon” refers to codons that comprise of different nucleotides but encode the same amino acid. With the context of this specification, the term “codon-optimized” refers to the use of synonymous codon changes based on species-specific preferences to improve gene expression and increase the translational efficiency of a gene of interest.

[0033] With the context of this disclosure, the term “cDNA” refers to complementary DNA that has been produced from an RNA via reverse transcription. With the context of this disclosure, the term “expression vector” refers to plasmid that is used for recombinant protein production.

[0034] With the context of this disclosure, the term “recombinant” refers to DNA, proteins, cells and organism that are made by combining genetic material from two different sources.

[0035] With the context of this disclosure, the term “host” refers to an organism that is used for recombinant protein production. With the context of this disclosure, the term “mammalian expression system” refers to mammalian cells that are suitable for recombinant protein or peptide production by using expression vectors that harbor the sequence specific for the desired protein to be produced.

[0036] With the context of this disclosure, the term “blood” and “plasma” are used as synonyms. Plasma is the liquid portion of blood.

[0037] A polypeptide according to the present invention consisting of an amino acid sequence of SEQ ID NO: 1 . The polypeptide having an amino acid sequence of SEQ ID NO: 1 has T cell activity and it can bind activated T cells via insulinlike growth factor 2 receptor (IGF2R) via its mannose-6-phosphate modification. In an embodiment according to the present invention, the polypeptide having T cell binding activity, consists of amino acid sequence having at least 90% or more sequence identity to SEQ ID NO: 1 . Further, an isolated polynucleotide encoding said polypeptide is provided. According to an embodiment of the present invention, the isolated polynucleotide has a nucleotide sequence comprising or consisting of SEQ ID NO: 2 or a nucleotide sequence that has at least 90 %, preferably at least 95% or at least 98% or at least 99% identity to SEQ ID NO: 2. The polynucleotide according to the present invention is used for producing a recombinant Clever-1 fragment (H1 ) mimicking sClever-1 enriched in plasma of cancer patients having T cellbinding activity.

[0038] With the context of this disclosure, the recombinant Clever-1 fragment (H1 ) according to the present invention mimics the dominant ~200kDa form of sClever-1 that is enriched in blood of cancer patients. An expression vector comprising the isolated polynucleotide according to the present invention, is used for producing the recombinant Clever-1 fragment (H1 ). The terms “recombinant Clever-1 fragment (H1 )”, “recombinant Clever-1 fragment H1 ” and “H1” refers to same recombinantly produced Clever-1 fragment in this disclosure. In an embodiment according to the present invention, the nucleotide sequence of the polynucleotide within the expression vector is codon-optimized cDNA. In an further embodiment according to the present invention, the nucleotide sequence is cloned into the expression vector between an N-terminal insulin signal peptide and a C-terminal Hiss-tag. Signal peptides are short N-terminal peptides that direct newly synthesized proteins towards the secretory pathway. The use of N-terminal insulin signal peptide enables H1 to be secreted to the cell medium. Additionally, the use of Hiss-tag allows the produced recombinant protein, H1 , to be purified using nickel- charged immobilized metal affinity chromatography (IMAC) resin purification system. A recombinant protein with a Hiss-tag has a high affinity for nickel, whereas most other proteins will either bind with low affinity, or not at all. The nickel-histidine reaction is not dependent on protein secondary structure and thus allows purification of denatured protein. Further, the protein can be renatured while still bound to the nickel resin. According to the present invention, the recombinant protein production is done in a recombinant host comprising the expression vector. The term “recombinant Clever-1 fragment” refers to recombinantly produced Clever-1 fragment. The method according to the present invention for producing a recombinant Clever-1 fragment (H1 ) comprises cultivating the recombinant host under conditions conducive for production of the recombinant Clever-1 fragment (H1 ) and recovering the recombinant Clever-1 fragment (H1 ).

[0039] In an embodiment according to the present invention, the recombinant host is a mammalian expression system. In a further embodiment according to the present invention, the mammalian expression system comprises CHO (Chinese hamster ovary) cells, such as CHOEBNALT85 cells. According to an embodiment of the present invention, the purity of the recombinantly produced H1 may be assessed using Coomassie-stained SDS-PAGE. Additionally, the end-point stability of H1 may be tested before and after three rounds of freezethaw cycles by comparative analytical size-exclusion chromatography (SEC) column. A recombinant Clever-1 fragment (H1 ) according to the present invention comprises or consists of SEQ ID NO: 1 and has T cell-binding activity and is capable of binding to insulin-like growth factor 2 receptor (IGF2R). The binding of the recombinant Clever-1 fragment (H1 ) to T cells takes place via mannose- 6-phosphate. The recombinant Clever-1 fragment (H1 ) mimics the dominant ~200kDa form of sClever-1 that is enriched in blood of cancer patients. Recombinant Clever-1 fragment (H1 ) is smaller fragment than the full-length Clever-1 protein, as H1 is heavily truncated at the C-terminus compared to the full-length Clever-1 protein. As presented in the experimental part, H1 can however selectively bind activated T cells and disrupt T cell receptor signaling leading to impaired Th1 expansion. It has been found and thereafter been able to produce the recombinant Clever-1 fragment (H1 ) which provides the binding to T cells.

[0040] A recombinant Clever-1 fragment (H1 ) comprising or consisting of an amino acid sequence SEQ ID NO: 1 can be used as a medicament. The recombinant Clever-1 fragment (H1 ) may be used in treatment of inflammation and / or autoimmune diseases. The TCR (T-cell receptor) is a complex of integral membrane proteins that participate in the activation of T-cells in response to an antigen. Stimulation of TCR is triggered by MHC (major histocompatibility complex) molecules on cells with the antigen. The mechanism of immunosuppression by H1 results from it disrupting the T-cell receptor activation, which leads to impaired Th1 expansion.

[0041] A pharmaceutical composition according to the present invention comprises a recombinant Clever-1 fragment (H1 ) comprising or consisting of an amino acid sequence SEQ ID NO: 1 , and a pharmaceutical acceptable carrier or excipient. A method according to the present invention for treating a disease or a condition comprises administering a recombinant Clever-1 fragment (H1 ) comprising or consisting of an amino acid sequence SEQ ID NO: 1 to a subject in need thereof. More specifically, a method according to the present invention for treating inflammation and / or autoimmune disease comprises administering a recombinant Clever-1 fragment (H1 ) comprising or consisting of an amino acid sequence SEQ ID NO: 1 to a subject in need thereof.

[0042] The following examples are given to further illustrate the invention without, however, restricting the invention thereto. EXPERIMENTAL

[0043] Examples 1 - 5 below summarize the results according to the present invention. Materials and methods are described more detailed after the examples.

[0044] Example 1. A dominant -200 kDa form of sClever-1 is enriched in blood of cancer patients

[0045] To elucidate the presence of a secreted form of Clever-1 (sClever-1 ) in biological fluids, a specific Time-resolved fluorescence immunoassay (TRFIA) was developed using Bexmarilimab (Bex) and 9-11 anti-Clever-1 antibodies, and optimized for robust sensitivity and stability across different sample material. Next sClever-1 levels were analysed in the plasma of healthy donors (n = 32 samples from 21 donors across 1 to 3 timepoints, EDTA blood), in patients with treatment-naive breast cancer (n = 138, EDTA blood), and in patients with various advanced cancers (n = 193, pre-treatment, Li-He blood) participating in the MATINS (A Study to Evaluate Safety, Tolerability and Preliminary Efficacy of FP-1305 in Cancer Patients; NCT03733990) trial (Figure 1A). A highly significant enrichment of sClever-1 in the plasma of patients with cancer was observed compared to undiagnosed healthy donors (Figure 1 B). In Receiver Operating Characteristic (ROC) analysis the enrichment showed excellent >0.8 performance discriminating cancer patients from healthy donors with 100% specificity for both cohorts at 48.7 ng / mL (Figure 1 C).

[0046] To characterize the secretory profile of sClever-1 for its distribution between extracellular vesicles and a freely soluble form, the culture supernatant was fractioned from Clever-1 -expressing acute myeloid leukemia cells (KG-1 ), human blood and lymph by ultracentrifugation. This approach enabled to determine sClever-1 levels in fractions containing microvesicles, exosomes and soluble proteins. All the studied fractions showed detectable sClever-1 protein in which lymph had the highest content of sClever-1 in the soluble form (Figure 1 D). The release of sClever-1 in the vesicle versus soluble fraction from various primary human cells and cell lines was further assessed by Western blot to identify the cell type producing sClever-1 into blood. Different from the full-length -280 kDa Clever-1 band, the plasma contained a more abundant -200 kDa sClever-1 species, which was observed in all macrophage subsets and endothelial cells as well as PMA-differentiated KG-1 cells (Figure 1 E). Direct comparison of healthy plasma with MATINS patient plasma in the vesicle fraction showed increased abundance of the -200 kDa sClever-1 (Figure 1 F). For validating the Western blot data for sClever-1 a mass spectrometry analysis of each gel band containing serum protein immunoprecipitated via biotinylated anti-Clever-1 antibody (Bexmarilimab and 9-11 , respectively) (Figure 1 G) was performed. Further, an additional mass spectrometry analysis revealed that each species of sClever-1 that was smaller than full-length Clever-1 was heavily truncated at the C-terminus.

[0047] Example 2. Recombinant production and quality control of the Clever-1 fragment H1

[0048] Several fragments spanning the Clever-1 protein sequence were tested (Figure 7). From the eight different sequences, usable quantities of good quality fragments were obtained only from the half-Clever-1 fragment (H1 ), FAS 1 -4 and Clever-1 Q2 proteins. The cancer enriched sClever-1 contains the binding site for antibody 9-11. Therefore, H1 was selected as a recombinant form of sClever-1 , since the two other fragments did not include the protein sequence for antibody 9-11 binding.

[0049] To investigate the function of the cancer enriched form of sClever-1 , a recombinant Clever-1 fragment H1 mimicking the -200 kDa sClever-1 was produced (Figure 2A). The production method of the recombinant Clever-1 fragment H1 is disclosed more detailed in the materials and methods.

[0050] Seven days after transfection the secreted protein was subsequently extracted and purified using Ni2+-affinity and size-exclusion chromatography to >95% purity (Figure 2B). A recombinant Clever-1 fragment H1 elutes as a broad, homogeneous monomeric peak with a minor, highly oligomeric fraction present (Figure 2C). The absolute molecular weight MW of H1 agreed well with the sequence-derived molar mass and further suggested the presence of 18 ± 2% glycosylation by mass. The purified protein was properly folded with a similar secondary structure composition as predicted by AlphaFold2 (Figure 2D). H1 was found to be stable and well-behaved (Figure 2E). The broad elution profile of H1 can be explained by the glycan content and a high order of inter-domain flexibility as assessed by dynamic light scattering (DLS) (Figure 2F), sizeexclusion chromatography and small-angle X-ray scattering (SEC-SAXS) (Figure 2G).

[0051] Example 3. Recombinant Clever-1 fragment H1 binds activated T cells upon phosphatidylserine exposure

[0052] The binding of recombinant Clever-1 fragment H1 on freshly isolated peripheral T cells enriched by CD3 negative selection was investigated. In nonactivated T cells positive binding was observed in a FCS|OWpopulation most likely representing apoptotic cells (Figure 3A). When the lymphocytes were activated via T cell receptor (TCR) engagement a FSChighpopulation became H1 positive, which was more prominent than H1 binding to the FSC|OWpopulation in both non-activated and activated states (Figure 3B). To verify that the FCShighcells were not apoptotic, they were stained with Annexin V and FAM-VAD-FMK, which is a caspase inhibitor and because of its carboxyfluorescein can be used to detect early apoptotic cells. As expected, the FSC|OWH1+cells produced a high signal for both markers, which was not observed in the FSChighH1+cells (Figure 3C). Externalization of PS is not restricted to apoptotic cells and can occur upon cellular activation where elevated cytoplasmic Ca2+followed by induction of phospholipid scramblase accelerates phospholipid movement. As ionomycin is known to increase intracellular Ca2+levels, it was tested whether short term activation of T cells with ionomycin can increase H1 binding. Already at five minutes H1 binding increased and peaked around 30 minutes resembling similar kinetics as seen with Annexin V (Figure 3D). To get insight on the cell populations possibly affected by sClever-1 in blood H1 binding was analyzed on peripheral lymphocytes collected from healthy donors and breast cancer patients. H1 binding was observed to be similar between healthy donors (median 4.7 range 22.4%) and breast cancer patients (median 5.7 range 35.5%) (Figure 3E), where H1 positive signal was mostly observed on CD45RO+ memory cells in both CD8 positive and negative cell populations (Figure 3F). The H1 + cells had significantly higher expression of PD-1 compared to cells not binding H1 indicative of selective binding of sClever-1 on activated memory T cells (Figure 3G). Example 4. Recombinant Clever-1 fragment H1 disrupts TCR activation and leads to impaired Th1 expansion

[0053] It was investigated whether recombinant Clever-1 fragment H1 binding on T cells has functional consequences during their activation and proliferation. In the presence of H1 the CD8+T cells reached four generations during the six- day incubation period. Hence, no dramatic effect on proliferation compared to control treated cells was observed (Figure 4A). When looking at their differentiation based on lineage determining transcription factors T-bet and FoxP3, the H1 treated cells had a significantly higher percentage of FoxP3+cells compared to control cells (Figure 4B). Since suboptimal TCR activation has been reported to increase upregulation of FoxP3 on CD8+T cells the early signalling events downstream of TCR were assessed. By binding to T cells H1 was able to impair the phosphorylation of Lek at Y394 (Figure 4C). Similarly, in a Jurkat NFKB reporter cell line, the presence of H1 tended to reduce reporter activity (Figure 4D). FoxP3 is known to suppress the function of NFAT and NFKB, and this leads to suppression of the expression of many genes including IL2 and effector T cell cytokines. Indeed, cytokine profiling of primary T cell cultures showed a significant reduction in IL10, IL6, IL10 and IL17 when incubated with H1 (Figure 4E). These cytokines were not directly indicative of inducing any specific lineage differentiation of the activated T cells, thus it was functionally tested how they polarized autologous primary human macrophages. The H1 cultured T cells promoted a polarization change in macrophages seen as a significant downregulation of HLA-DR and CD40, whereas Clever-1 expression on macrophages was upregulated suggesting a more immunosuppressive phenotype (Figure 4F).

[0054] Example 5. Recombinant Clever-1 fragment H1 binds IGF2R on T cells via mannose-6-phosphate

[0055] To identify the counter receptor for sClever-1 on T cells, it was first screened recombinant Clever-1 fragment H1 binding on a HEK293 cell-based Retrogenix platform consisting of 6449 human plasma membrane proteins, tethered human secreted proteins and 397 heterodimer plasma membrane protein complexes. H1 binding on Jurkat cells was tested to utilize them in affinity pull-down assays. H1 showed very good binding to all Jurkat cells even without prior TCR-stimulation. The binding was not induced by apoptosis since puromycin treated cells did not bind H1 (Figure 5A). Mass spectrometry analysis (LC-ESI-MS / MS) of proteins pulled down with H1 revealed several hits involved in T cell activation such as IGF2R, TRFC, Zap70, CCT8, CCT6A (Figure 5B). IGF2R was investigated in more detail as it was not included in the Retrogenix assay nor would endogenous IGFR2 normally be displayed upon the cell surface of HEK293 cells. A similar upregulation of IGF2R on the membrane of primary human T cells was observed after ionomycin stimulation as was observed for Annexin V in primary human T cells (Figure 5C). Pulldown assays with Jurkat and primary human T cells confirmed the interaction of H1 with IGF2R (Figure 5D). Using recombinant IGF2R in Octet it was demonstrated that H1 bound directly to IGF2R in domains 1 -10 but not in domains 11 -13 known to bind IGF II (Figure 5E). Since IGF2R is a cationindependent mannose-6-phosphate (M6P) receptor it was tested whether an excess amount of M6P or removing M6P from H1 by alkaline phosphate treatment would abolish H1 binding to IGF2R. On Octet, H1 binding to IGF2R was completely abolished with M6P manipulation. Also, on Jurkat cells H1 binding was decreased with either an excess amount of free M6P or removal of M6P from H1 (Figure 5F). To fully demonstrate H1 binding on T cells via IGF2R, either IGF2R or TRFC was knocked out from Jurkat cells by CRISPR / Cas9. A -30% reduction was achieved in the expression levels of both proteins and this was able to decrease H1 binding by -50% to IGF2R CRISPR cells (Figure 5G). It is interesting to note that gene silencing of TRFC downregulated IGF2R and therefore reduced also binding of H1 to TRFC CRISPR cells (Figure 5G). To confirm the interaction of endogenous sClever- 1 with IGF2R, Clever-1 pull-downs were performed from serum and a recombinant IGF2R containing an AviTag was used as detection reagent. Positive signal was observed with IGF2R when sClever-1 was detected, and was clearly reduced when the pull-downs were performed with Clever-1 depleted serum (Figure 5H).

[0056] Materials and methods

[0057] Human samples

[0058] Blood samples were obtained from treatment naive breast cancer patients prior and three weeks after undergoing mastectomy. Patients having a tumor nodule exceeding 2 cm in diameter were included in the study. Neoadjuvant therapy was not used. The EDTA-plasma was separated by centrifugation (2,000g for 10 min) within two-hours of withdrawal and stored at -70°C for later use. Peripheral blood mononuclear cells (PBMC) were isolated from the remaining EDTA-blood by Ficoll-Paque centrifugation and stored in freezing buffer (RPMI, 10% FCS, 1 % glutamine, penicillin / streptomycin and 10% DMSO) at -150°C for later use. Lymph was collected aseptically (needle aspiration) via postoperative lymphatic leakage. The lymph was cleared by centrifugation (2,000g for 10 min) within two-hours of withdrawal and stored at -70°C for later use. Serum samples from lung carcinoma (n = 16) and melanoma (n = 42) patients treated with anti-PD-1 or anti-PD-L1 were obtained from Auria Biobank (Turku, Finland) with a dataset including information on dosing frequency, patient time on immunotherapy, blood cell counts and other treatments during or after immunotherapy (chemotherapy, radiotherapy and operations).

[0059] MATINS (A Study to Evaluate Safety, Tolerability and Preliminary Efficacy of FP-1305 in Cancer Patients; NCT03733990) pre-treatment blood samples were collected per study protocol and centrifuged (as above) within 24 hours of withdrawal to obtain heparin plasma. The full study details can be found in Rannikko et al. (2023), “Bexmarilimab-induced macrophage activation leads to treatment benefit in solid tumors: The phase l / ll first-in-human MATINS trial”. Cell Rep Med 4.

[0060] Time-resolved fluorescence immunoassay (TRFIA)

[0061] A TRFIA was generated using two different anti-Clever-1 antibodies (9-11 and biotinylated Bexmarilimab), and a europium-labelled streptavidin-based detection method to analyze sClever-1 in human samples. Briefly, Nunc Maxisorp white assay plates were coated with 9-11 (InVivo Biotech) diluted at 10 pg / mL in 0.1 M NaHCO3 pH 9.6. The plate was washed six times with 0.1 % tween in PBS and blocked with 1 % milk powder and 1 % gelatin in PBS for 1 h at RT. Thereafter, the plasma samples were diluted 1 :20 in PBS and incubated for 1 h at RT. After six washes, the biotinylated Bex was added for 1 h RT. The plate was washed again and Eu-Labeled Streptavidin diluted 1000-fold in assay buffer (Ready-for-use Tris-HCI Buffered NaCI solution) (pH 7.8) was applied to the wells for 30 min at RT. Following the incubation, the plate was washed and developed with Enhancement Solution and measured with Viktor Nivo (all from Perkin Elmer). For measuring mouse sClever-1 from serum the biotinylated Bexmarilimab was replaced by biotinylated mStab1.26 antibody (InVivo Biotech).

[0062] Due to the unavailability of a recombinant Clever-1 standard when running the cancer cohorts against healthy controls, the standard curve was produced using a dilution series on human lymph from a healthy donor.

[0063] By the production of H1 and validation of its binding to bexmarilimab, the standard curve was created with sClever-1 depleted plasma spiked with different concentrations of H1 . The absolute levels of sClever-1 in the plasma of patients with cancer and healthy controls were calculated by running assay control samples containing low, medium and high sClever-1 with the lymph standard in parallel with the H1 standard. The stability of sClever-1 was tested and showed that the signal remained stable for at least five freeze- thaw cycles (-70°, n = 7) in the tested matrices.

[0064] Primary cells

[0065] Peripheral blood mononuclear cells were separated by Ficoll-Paque PLUS (GE) density gradient centrifugation from which monocytes were enriched using CD14 magnetic beads (Human Monocyte Isolation Kit human, Miltenyi) and T cells using Human T Cell Isolation Kit (negative selection, Stemcell). The monocytes were differentiated into macrophages by a seven-day incubation in IMDM (Gibco) supplemented with 10% FCS (Merck), penicillin / streptomycin (P / S) (Thermo Fisher) and 50 ng / mL of M-CSF (Biolegend) with one medium change. At this point the differentiated cells are referred to as M0. The macrophages were then polarized to M1 by first adding 20 ng / mL of recombinant IFNy for 24h and then 100 ng / mL of LPS (Invivogen) for another 24h. For M2 polarization, the macrophages were incubated with recombinant human IL-4 (Peprotech) and 100 nM of dexamethasone (Merck) for 48h.

[0066] Human lymphatic endothelial cells (HLEC, No.2500) were purchased from ScienCell and cultured in Endothelial Cell Medium (ECM, No.1001 ). Human pulmonary microvascular endothelial cells (HPMEC, C-12281 ) were purchased from PromoCell and cultured in Endothelial Cell Growth Medium (ECGM, #C-22020) with supplement mix (#C-39225, both from PromoCell). Cell lines

[0067] KG-1 (CCL-246, ATCC) were cultured in IMDM supplemented with 20% FCS and P / S. To induce differentiation into macrophages they were incubated in 100 ng / mL of phorbol-myristate acetate for 72h. The other cell lines used are listed with their specific culture medium: HEK293T (CRL-1573, ATCC; DMEM + 1 % FCS); Jurkat (ATCC), Jurkat-Dual™, Jurkat-Lucia™ TCR-h-PD-1 , Raji- APC-hPD-L1 (both from InvivoGen; IMDM, 2 mM L-glutamine, 25 mM HEPES, 10% FCS, P / S, 100 pg / mL Normocin).

[0068] Separation of sClever-1 fractions by ultracentrifugation

[0069] KG-1 acute myelogenous leukemia cells were used as a positive control for high Clever-1 expression. Culture supernatant from 50-70% confluent KG-1 cells was collected. Blood and lymph were first diluted with an equal volume of PBS (1 :1 ) due to their viscosity. All samples were then centrifuged by the optimized routine proposed by Thery et al., without the filtration process for exosomes purification. Briefly, the samples were first centrifuged at 300g for 10 min at 4°C using Sorvall RT6000B refrigerated centrifuge for the removal of cells. Then cells were discarded, and the supernatant was collected and centrifuged at 2,000g for 10 min at 4°C using Sorvall RT6000B refrigerated centrifuge for the removal of cell debris. The supernatant was then collected and centrifuged at 20,000g for 30 min at 4°C using rotor SS-34, Sorvall RC5C centrifuge. The MVs pellet was then collected and suspended in 100-200 pL of PBS and stored at -20°C. For purifying the exosomes, the supernatant was transferred to ultracentrifuge tubes and then centrifuged at 100,000g for 70 min at 4°C using Beckman coulter, Optima L-90K ultracentrifuge. The exosome fraction was collected and the remaining supernatant was considered to contain free soluble Clever-1 .

[0070] Isolation of vesicle fractions

[0071] Plasma samples were pre-cleared by a two-step centrifugation, first at 2,000g for 20 min to remove cells and next at 10,000g for 20 min to remove debris. The plasma was diluted in PBS (1 :1 ), mixed with Exosome Precipitation Reagent (from plasma, Invitrogen) and vortexed until the solution was homogenous. The samples were incubated at RT for 10 min and centrifuged at 10000g for 5 min. Cell conditioned media was pre-cleared by centrifugation at 2,000g for 30 min and a 0.5 volume of Exosome Precipitation Reagent (from cell culture media, Invitrogen) was added and vortexed to a homogenous solution. The samples were then incubated 4°C o / n and centrifuged at 10,000g for 1 h at 4°C. Cold PBS was added and the samples were centrifuged at 14,000g for 5 min at 4°C. The supernatant was collected (liquid fraction) and the pellets were resuspended in PBS (vesicle fraction).

[0072] Western blot

[0073] T cells were pelleted and resuspended in lysis buffer (10 mM Tris-HCI [pH 7.4], 5 mM NaF, 1.0% Triton X, 1 mM EDTA, 1 mM PMSF) supplemented with EDTA-free complete Protease Inhibitor Cocktail (Roche). The cells were lysed at 4°C with mixing for 60 min and centrifuged at 14,000g and 4°C for 20 min. The supernatant was collected and Laemmli sample buffer (reducing) was added to the samples. The samples were run on a 4-20% FastGene PAGE gel (Nippon Genetics Europe). The gel was transferred to a PVDF membrane using the Trans Blot Turbo (Bio-Rad) protein transfer system. The membrane was blocked with 5% BSA in 0.1 % Tween-TBS (Tris-buffered saline) and incubated sequentially with anti-pLck (cloneY394; R&D), anti-Lck (Cell Signaling) and anti-GAPDH (clone 6C5, HyTest) with in between stripping (1.5% Glycine, 0.1 % SDS and 1 % Tween 20, pH 2.2, 15 min at RT). The membrane was detected with ECL™ (Cytiva).

[0074] IGF2R was detected using rabbit polyclonal anti-IGF2R antibody (20253-1 -AP, Proteintech). For detection of mannose-6-phosphate and sClever-1 , the samples were denatured at 70°C for 10 minutes but not reduced mannose-6- phosphate was detected by incubation with 10 pg / mL recombinant human IGF2R-His-AviTag protein (IGF-HM12RB, Kactus Bio) followed by 1 pg / mL HRP-conjugated rabbit anti-AviTag polyclonal antibody (orb475166, Biorbyt). To detect sClever-1 the membrane was incubated with 1 pg / mL rat lgG2a monoclonal anti-human Clever-1 antibody 9-11 (InVivo Biotech) or mouse antihuman Clever-1 antibody 4G9 (Santa Cruz Biotechnology). When necessary, HRP was directly conjugated to the primary detection antibody, as in the case of anti-Clever-1 -HRP (9-11 ), by using a HRP Conjugation Kit (ab102890, AbCam). For primary macrophage and EV blots, TFRC NBP2-34602 (Novus), CD206 12981 S (Cell Signaling), CD63 sc-5275 (Santa Cruz) Pulldown of sClever-1 from serum

[0075] Human lgG4, bexmarilimab (both from Abzena), rat lgG2a isotype control (BioXcell) and anti-Clever-1 9-11 (InVivo Biotech) antibodies were diluted to 1 mg / mL in PBS (100 pL) and biotinylated with NHS-PEG4-biotin (EZ-Link™, No-Weigh™ Format, Thermo) at a calculated ratio of 6 biotin molecules per molecule of antibody, according to the manufacturer’s instructions. Following the biotinylation reaction the excess biotin was removed by overnight dialysis in PBS in Slide-a-Lyzer MINI Dialysis Devices 10 kDa MWCO (Thermo).

[0076] Human AB serum (Millipore) was diluted 1 / 10 to a final volume of 200 mL in ice-cold PBS + 0.1 % Tween 20. This was precleared by end-over-end mixing at 4°C overnight with 10 mg pre-washed Dynabeads M-280 Streptavidin (Invitrogen), divided equally into four 50 mL Falcon tubes. The beads were pelleted at 3000g and 4°C for 10 min, held in place with a magnet, and the precleared supernatant removed and divided into four separate 50 mL Falcon tubes on ice. Fifty micrograms of each biotinylated antibody were added separately to each tube of precleared supernatant. The biotinylated antibodies were mixed with the precleared supernatants by end-over-end mixing at 4°C for 8 h. Following this 2.5 mg of pre-washed Dynabeads M-280 Streptavidin were added to each tube to capture the complexes of biotinylated antibodies and ligands, and incubation was continued with end-over-end mixing at 4°C overnight. The following day the beads were pelleted at 3000g and 4°C for 10 min and the supernatant (“flow-through”) removed and frozen. The pelleted magnetic beads were resuspended in 1 .5 mL of ice-cold PBS + 0.1 % Tween 20, mixed thoroughly by vortexing then pulled-down using a DynaMag-2 magnetic rack (Invitrogen) and thoroughly aspirated. This washing procedure was repeated for a total of four times to remove non-specifically bound proteins from the beads. Finally, the fully aspirated beads were eluted in 40 pL SDS- PAGE denaturing sample buffer without reducing agent, pulled-down with the magnet, aspirated, and then the elution repeated in a second 40 pL volume of SDS-PAGE denaturing sample buffer without reducing agent. The 80 pL of eluate from each sample was pooled and heated at 70°C for 10 min prior to loading on an SDS-PAGE gel for Coomassie staining and isolation of specific bands for mass-spec proteomics analysis. For this purpose, 50 pL of sample was loaded per well of a 4-20% Mini-PROTEAN TGX precast gel (BioRad) and run to completion at 100 V. Spectra Multicolor High Range Protein Ladder (Thermo) was included as molecular weight standard. Following electrophoresis, the gel was washed in ddH20 and then stained in GelCode Blue Safe Protein Stain (Thermo) according to the manufacturer’s instructions. Following de-staining in ddH2O the gel was photographed before and after the excision of specific molecular weight bands of interest using a clean scalpel and collection tube. The collected bands were sent for mass-spec proteomics analysis. A repeat of the experiment for western blotting was also performed to confirm the presence and approximate molecular weight of Clever-1 pulled down. In a variation of this method, the magnetic Dynabeads used for the pulldown were washed 4 times in PBS with 0.1 % Tween 20, then three times in PBS (to remove residual detergent) and sent for mass spectrometry and proteomics analysis of bound protein without any prior separation via SDS- PAGE. When necessary to detect mannose-6-phosphate on protein that was pulled down, the PBS was supplemented with 5 mM NaF, 1 mM EDTA, 1 mM PMSF and EDTA-free complete Protease Inhibitor Cocktail (Roche).

[0077] Expression, purification and modification of recombinant Clever-1 fragment Hl

[0078] The codon-optimised cDNA of human Clever-1 (UniProt ID: Q9NY15) between residues 26-1255 was cloned into a pQMCF-1.2 expression vector between an N-terminal insulin signal peptide and a C-terminal Hiss-tag to create a 133.6 kDa soluble construct; recombinant Clever-1 fragment H1. The recombinant expression and purification of H1 were performed by Icosagen Cell Factory. In summary, H1 was transiently expressed in CHOEBNALT85 cells using the QMCF technology (Silla et al., 2005). Briefly, the cells were grown in CHO TF (Xell AG) media and cells were chemically transfected with R007 in 250 mL, at a cell density of 2 x 106cells / mL at 37 degrees C. The following day after transfection the cell media was supplemented with ActiPro. Three days after transfection the cell culture temperature was decreased to 30 degrees C. Four days after transfection ActiPro supplements were added . The secreted protein was captured on Ni2+Sepharose affinity resin (Cytiva) and washed with high- salt buffer before elution with 0.3 M imidazole. A final polishing step was performed by size-exclusion chromatography (SEC) on a Superdex 200 Increase 10 / 300 GL (GE Healthcare) column using a 20 mM HEPES pH 8.0, 150 mM buffer NaCI, 2 mM CaCl2, and 5% glycerol buffer. This buffer (subsequently called modified HBS) is used for further experiments unless stated otherwise. The purity was assessed using Coomassie-stained SDS- PAGE, while end-point stability was tested before and after three rounds of freeze-thaw cycles by comparative analytical SEC on a BioSuite 250 4.6 / 300 column (Waters). Aliquots of H1 were snap-frozen for long-term -70°C storage and sterile-filtered using a 0.2 pm membrane before subsequent use.

[0079] To biotinylate H1 protein, “No Weigh” NHS-PEG4-Biotin (Thermo) was used to label H1 at a molar ratio of 6:1. The biotinylated protein was then dialyzed at 4°C in HBS to remove excess biotin. The dialyzed, biotinylated protein was filter-sterilized and aliquoted at -70°C. To measure how many molecules of biotin were conjugated per molecule of H1 , a Pierce Biotin Quantitation Kit (Thermo) was used according to the manufacturer’s instructions. There were 3.8755 biotin molecules on average per molecule of H1. To determine what percentage of H1 molecules were biotinylated, a non-reducing denaturing SDS-PAGE gel was run with unheated non-reduced samples of H1 , streptavidin, and H1 precomplexed with streptavidin, alongside Spectra Multicolor High Range Protein Ladder (Thermo). The gel was stained with GelCode Blue (Thermo), and the molecular weight-shift of complexed H1 analysed. More than 95% of H1 was biotinylated.

[0080] To dephosphorylate H1 , the protein was digested with Fast AP Alkaline Phosphatase enzyme (Thermo) by diluting the protein to a final concentration of 0.1 mg / mL in a final volume of 100 pL of 1x reaction buffer with 10 units of Fast AP enzyme, followed by incubation at 37°C overnight. To detect dephosphorylation, the loss of binding to recombinant IGF2R protein (Kactus Bio) on Octet was confirmed. To confirm integrity of the protein, the concentration was checked by NanoDrop and a sample run on SDS-PAGE and stained with GelCode Blue (Thermo).

[0081] The absolute molecular mass of recombinant Clever-1 fragment H1 was determined using size-exclusion chromatography-coupled multi-angle light scattering (SEC-MALS). 150 pg of H1 was injected on a Superdex 200 Increase 10 / 300 GL (GE Healthcare) column and eluted at RT using 0.75 mL / min flow rate in modified HBS without glycerol. The light scattering signal was recorded using a miniDAWN TREOS II (Wyatt Technology) detector. Online concentration was measured with an Optilab T-rEX (Wyatt Technology) differential refractometer. Bovine serum albumin (Sigma-Aldrich) was used for calibration. The glycosylation level was derived using conjugate analysis in the ASTRA software (Wyatt Technology) assuming refractive index increment values of 0.185 and 0.156 mL / g for the protein and the glycan component, respectively.

[0082] Hydrodynamic radius of the monomeric H1 was assessed using dynamic light scattering (DLS) on a Zetasizer Nano ZS (Malvern) system. Data acquisition was carried out at RT with a protein concentration of 0.2 mg / mL. Six measurements were averaged to determine the intensity size distribution.

[0083] Circular dichroism spectroscopy (CD)

[0084] CD spectra between 180-250 nm were recorded on a Chirascan CD spectrophotometer (Applied Photophysics) at RT using a 0.1 mm path-length quartz cuvette. The measurement of recombinant Clever-1 fragment (H1 ) at a 1 mg / mL concentration was carried out in a buffer containing 20 mM sodium phosphate pH 8.0 and 150 mM NaF. Six scans were averaged and smoothed. As a reference, five representative models of H1 were prepared using AlphaFold2 (Mirdita et al., 2022). Simulated CD spectra of the models were prepared and averaged using the PDBMD2CD server (Drew et al., 2020). The CD spectra were deconvolved (Micsonai et al., Nucleic Acids Res 2022) and classified (Micsonai et al., Front Mol Biosci 2022) using the BeStSel server.

[0085] Differential

[0086] The stability of recombinant Clever-1 fragment H1 was characterised by following its thermal unfolding at 0.2 mg / mL using a Prometheus NT.48 (NanoTemper) system between 20-95°C with a ramping speed of 1 °C / min. The measurement was carried out in triplicates.

[0087] Small-angle X-ray scattering (SAXS)

[0088] SEC-SAXS measurements were carried out on the CoSAXS beamline at the MAX IV synchrotron. 250 pg of recombinant Clever-1 fragment H1 was applied on a Superdex 200 Increase 10 / 300 GL (GE Healthcare) column and resolved at RT with a 0.6 mL / min flow rate. SAXS data were recorded between scattering vectors 0.003-0.334 A’1. The significant components in the elution data were reduced using regularized alternating least squares (REGALS) method (Meisburger et al., 2021 ) and analysed with BioXTAS RAW (Hopkins et.al., 2017). The MW was derived from the whole q range using the Porod invariant method (Piiadov et al., 2019).

[0089] T cell activation

[0090] T cell density was adjusted to 1 .0 x 106cells / mL in RPM1 1640 medium (Sigma) supplemented with 10% FCS, 100 U / mL penicillin, 100 pg / mL streptomycin, 55 pM 2-ME and 2 mM L-glutamine. The T cells were plated in 96-well plates (Sarstedt) at 1.0 x 105cells per well and activated either by ionomycin stimulation at 1 pg / mL (Sigma) or plate-coated anti-CD3 at 10 pg / mL (clone OKT3; Invitrogen) and anti-CD28 at 10 pg / mL (clone 28.2; Invitrogen) with IL- 2 (Miltenyi Biotech) at 20 ng / mL H1 or BSA control was added upon activation at 50 pg / mL for the whole duration of the experiment.

[0091] Flow cytometry

[0092] T cells were activated with CD3 / CD28+IL-2 for 48 h or ionomycin (up to 120 minutes) and thereafter incubated with 50 pg / mL of biotin-H1 (Icosagen) or biotin-lgG4 (Abzena) at 37°C for 30 min in PBS supplemented with 0.1 % BSA followed by streptavidin-PE (5 pg / mL) (BD). When co-staining for Annexin V- FITC (10 pg / mL, BD) and FAM-VAD-FMK (1 :250, AAT Bioquest) was performed, a Annexin V Binding Buffer containing 10 mM HEPES pH 7.4, 140 mM NaCI, 2.5 mM CaCI2 and 0.1 % BSA was used. Recombinant Annexin V (10 pg / mL, BD) or H1 (50 pg / mL) at 37°C for 30 min was added before staining to test competitive binding to PS.

[0093] For measuring T cell proliferation, the cells were pre-labelled with 1 pM Vybrant™ CFDA SE dye (CFSE; Invitrogen) according to the manufacturer’s protocol and stained with anti-CD8-APC (clone RPA-T8, BD) after six days. For intracellular staining of lineage determining transcription factors, the cells were first surface stained with anti-CD8-BV510 (clone SK1 , BD), fixed and permeabilized using FoxP3 / Transcription Factor Staining Buffer Kit (Thermo Fisher) according to the manufacturer’s protocol and stained with anti-FoxP3- PE (clone PCH101 , Invitrogen) and anti-T-bet-BV412 (clone04-46, BD). For assessing macrophage polarization after autologous T cell coculture cell surface staining was performed with the following antibodies: anti-CD64-PE (clone 10.1 , BD), anti-HLA-DR-PE-Cy7 (cloneG46-6, BD), anti-CD163-BV711 (clone GH1 / 61 , BD), anti-CD40-BV510 (clone 5C3, BD), anti-Clever-1 -Alexa Fluor 647 (clone 9-11 , in-house conjugated) or isotype control antibodies with the same fluorochromes.

[0094] For surface staining of Jurkat cells, the cells were pelleted and resuspended in ice-cold Annexin V Binding Buffer plus 100 pg / mL Kiovig and stained with biotinylated H1 as described for primary T cells. Following this, Annexin V- FITC was added for a 5-minute incubation at room temperature, where applicable. Thereafter, all washing, and incubation steps were performed on ice, in ice-cold Epics 1 Buffer (PBS, 2% FCS, 0.1 % azide). Alexa Fluor 647 labelled streptavidin (Life Technologies) was used at 2 pg / mL, to detect H1 surface staining. Cells were finally fixed in 0.5% paraformaldehyde PBS before flow cytometry. For intracellular staining of Jurkat cells, the cells were first fixed in 4% paraformaldehyde in PBS, then permeabilized with PBS + 0.2% Tween 20. Thereafter, 0.1 % Tween 20 was included in all steps to maintain permeability. Anti-TFRC (NBP2-34602; Novus) and anti-IGF2R (20253-1 -AP; Proteintech) were used to detect knockdown efficiency. The cells were analysed with LSRFortessa (BD) and data processed with FlowJo software v.10.7.1 (TreeStar).

[0095] Frozen aliquots of PBMCs from healthy donors and breast cancer patients were pre-stained with fixable viability dye eFluor 450 (Thermo) and thereafter with anti-IGF2R + goat anti-rabbit AF488 (Invitrogen), biotinylated H1 + PE- streptavidin (#349023, BD), anti-CD8-APC (clone RPA-T8, BD), anti- CD45RO-APC-H7 (clone UCHL1 , BD) and anti-PD-1 -PerCP-Cy5.5 (EH12.1 , BD).

[0096] Jurkat-Dual assay

[0097] Jurkat-Dual cells were plated 0.4 x 106cells / well in 96-well plates in 200 pL of cell culture medium supplemented with anti-CD3 and anti-CD28 as described above. Recombinant Clever-1 fragment H1 or BSA control was added upon activation at 50 pg / mL for 18 hours. NFKB activation was analysed by measuring the luciferase reporter from the cell culture supernatant with QUANTI-Luc reagent (InvivoGen) according to the manufacturer’s instruction using Tecan Infinite (Tecan).

[0098] Primary T cell culture supernatants were analyzed for cytokine content six days after activation with Bio-Plex Pro™ Human Cytokine 27-Plex Assay (#M500KCAF0Y, BioRad) according to the manufacturer’s protocol.

[0099] Retrogenix Cell Microarray assay

[0100] To discover recombinant Clever-1 fragment H1 binding ligands, biotin-H 1 was tested in the Retrogenix Cell Microarray assay consisting of 6449 human plasma membrane, tethered human secreted proteins (using an epitope tag and tethered single-chain variable fragment [scFv] anchor technology) and 397 heterodimer plasma membrane protein complexes (Charles River Discovery). For library screening 10 pg / mL of biotin-H1 was screened in duplicate for binding against fixed HEK293 cells using both the sequential method (test protein added to slides, washed and then AlexaFluor647 Streptavidin detection reagent added) and the pre-incubation method (test protein pre-incubated with AlexaFluor647 Streptavidin at a 4:1 molar ratio before addition to slides). Specific hits were confirmed by staining un-fixed and fixed cells and further validated by flow cytometry.

[0101] Pulldown with biotinylated H1 in Jurkat cells and primary T cells

[0102] Method 1 (with crosslinker): Six million Jurkat cells were resuspended in 0.6 mL ice-cold Annexin V Binding Buffer with or without 50 pg / mL biotinylated H1 protein. The cells were then placed in a 37_,OC cell culture incubator for 30 minutes for H1 to bind to the cell outer membranes. The cells were then pelleted at 300g and washed twice with ice-cold PBS to remove unbound H1 . The aspirated cell pellets were resuspended in freshly prepared 3.5 mM BS3 crosslinker (Pierce, No-Weigh Format) in 4°C PBS, then incubated for 30 minutes at room temperature for crosslinking. Twenty millimolar Tris pH 7.6 was added for 15 minutes to quench the crosslinker. The cells were then pelleted at 400g, aspirated, and washed twice with ice-cold PBS to remove residual unreacted crosslinker. The washed and pelleted cells were placed on ice and lysed in 1 mL ice-cold Lysis Buffer containing protease inhibitors. Following 15 minutes of lysis the cells were centrifuged at 4°C and 10,000g for 5 minutes to remove nuclei. The clarified supernatants were incubated with magnetic M280 Streptavidin Dynabeads (Invitrogen) overnight at 4°C with mixing, for pulldown of biotinylated H1 complexes to occur. Following the pulldown, the Dynabeads were pelleted and washed 4x in 1 mL ice-cold Lysis Buffer, then pelleted and washed 3x in TBS to remove residual detergent. The aspirated beads were then sent for mass spec and proteomics analysis to identify and quantify the proteins that had been co-precipitated with biotinylated H1 .

[0103] Method 2 (with purified plasma membranes): Fifty million Jurkat cells were resuspended in 1.2 mL ice-cold Annexin V Binding Buffer with or without 50 pg / mL biotinylated H1 protein. After H1 binding the aspirated cell pellets were resuspended in 0.5 mL ice-cold PBS premixed 50:50 with hypotonic Buffer A from a Minute Plasma Membrane Protein Isolation Kit (Invent Biotechnologies) to which 1 mM PMSF, 5 mM NaF, Roche complete EDTA-free Protease Inhibitor Cocktail had been added. The cell suspensions were incubated on ice for 5 minutes to sensitize the cells to rupturing by shear forces. The cell suspensions were then passed once through a Minute Plasma Membrane Protein Isolation Kit Filter Cartridge at 16,000g for 30 seconds to rupture the cells. Thereafter the isolation of plasma cell membranes and other separate cell fractions (nuclei, cytosol and organelles) were performed according to the manufacturer’s instructions. The final yield of plasma membrane protein was measured by micro-DS Protein Assay (Bio-Rad) and found to be between 55- 57 pg for each sample. The purified plasma membrane pellets were solubilized in 0.5 mL ice-cold Lysis Buffer and streptavidin Dynabead pulldowns were performed as previously.

[0104] Mass-spectrometry (MS), proteomics analysis and data filtering

[0105] Samples were on-bead digested according to standard protocol at the Turku Proteomics facility. Protein samples, bound on magnetic streptavidin beads, were denatured with 8 M urea in 50 mM Tris-HCI, pH 8. Proteins were reduced with 10 mM D,L-dithiotreitol (in 50 mM Tris-HCI, pH 8), and alkylated with 40 mM iodoacetamide (in 50 mM Tris-HCI, pH 8). Samples were digested overnight with sequencing grade modified trypsin (Promega). Peptide samples were desalted using Sep-Pak tC 18 well plate (Waters) and peptides were dried in a vacuum centrifuge. Digested peptides were dissolved in 11 pL of 0.1 % formic acid. From each digested peptide sample 5 pL was injected for analysis. The LC-ESI-MS / MS analysis was performed on a nanoflow HPLC system (Easy-nLC1000, Thermo Fisher Scientific) coupled to the Q Exactive HF mass spectrometer (Thermo Fisher Scientific, Bremen, Germany) equipped with a nano-electrospray ionization source. Peptides were first loaded on a trapping column and subsequently separated inline on a 15 cm C18 column (75 pm x 15 cm, ReproSil-Pur 3 pm 120 A C18-AQ, Dr. Maisch HPLC GmbH, Ammerbuch-Entringen, Germany). The mobile phase consisted of water with 0.1 % formic acid (solvent A) or acetonitrile / water (80:20 [v / v]) with 0.1 % formic acid (solvent B). A 40 min long step gradient (from 5% to 21 % of solvent B in 17 min, from 21 % to 36% of solvent B in 13 min, and from 36% to 100% of solvent B in 5 min, followed by a 5 min wash stage with 100% of solvent B) was used to eluate peptides.

[0106] MS data was acquired automatically by using Thermo Xcalibur 4.1 software (Thermo Fisher Scientific). An information dependent acquisition method consisted of an Orbitrap MS survey scan of mass range 350-1750 m / z with resolution of 120,000, a target value of 3,000,000, and maximal injection time of 100 ms. The 15 most intense peptide ions were selected for MS2 fragmentation with resolution of 15,000, a target value of 50,000, and injection time of 200 ms. Data files were searched for protein identification using Proteome Discoverer 3.0 software (Thermo Fisher Scientific) connected to an in-house server running the Mascot 2.8.2 software (Matrix Science). Variable modifications included where applicable were mass shift of hydrolyzed BS3 and the mass shift of Tris-quenched BS3. Data was searched against a SwissProt database (version 2022_03) with taxonomy filter Homo sapiens.

[0107] For each experiment + control “pair”, the results of the experiment pulldown with biotinylated H1 were compared to the results of the control pulldown without any biotinylated H1. Any hits where the Score Mascot for the experiment was less than threefold greater than the Score Mascot for the control were filtered out to remove noise, such as for e.g., keratin, albumin, etc. Unused filters added for qualitative assessment were Retrogenix array score and CRAPome (https: / / reprint-apms.org) score. The hits were ranked according to number of peptides for a given protein. Any hit protein with less than two unique peptides was deprioritized. The results of the pulldown with crosslinking were compared to the results for the pulldown with noncrosslinked purified plasma membranes. STRING database (https: / / string- db.org) node analysis was performed to look for patterns in the results. Octet

[0108] Measurement of protein interactions via biolayer interferometry (BLI) was performed using a Fortebio Octet RED384 in 8-channel mode and 96-well plate format. All assays were performed at 1000 rpm, 25°C, and in 200 pL Kinetics Buffer (10 mM HEPES pH 7.4, 150 mM NaCI supplemented with 0.1 % BSA and 0.02% Tween 20). The assay definition included baseline (5 min), loading (5 min), association (3 min), dissociation (30 min). The biosensors used were streptavidin (Sartorius). The biotinylated ligands loaded onto the streptavidin biosensors at 10 pg / mL were either biotinylated H1 protein or biotinylated IGF2R domains 1 -10 protein (Kactus Bio), as applicable. The nonbiotinylated solution analyte proteins used in binding assays were either H1 protein or IGF2R domains 11 -13 (R&D Systems), as applicable. All experiments included a negative control reference biosensor for subtraction. Single cycle kinetics data was fitted using TraceDrawer software according to a 1 :1 Langmuir model.

[0109] CRISPR / Cas9

[0110] Jurkat cells were electroporated with pooled Edit-R Predesigned sgRNAs targeting IGF2R (SG-010601-01 , SG-010601 -02, SG-010601 -03) or TFRC (SG-003941-01 , SG-003941 -02, SG-003941 -03; all from Dharmacon) complexed with TrueCut Cas9 Protein v2 using the SE Cell Line kit and 4D- Nucleofector X Unit (Lonza) according to manufacturer's protocol. Primary human monocytes were positively enriched from buffy coat PBMCs with CD14 Microbeads and LS columns (Miltenyi Biotec). TrueCut Cas9 Protein v2 and TrueGuide Synthetic single guide (sg)RNAs (both from Thermo Fisher) were complexed at RT and delivered into purified monocytes by electroporation as described (Freund et al., 2020). Subsequently, the monocytes were differentiated into macrophages as described above. sgRosa26 (A35525) targeting the murine Rosa26 locus was used as a negative control. Cited references

[0111] Drew ED, Janes RW. PDBMD2CD: providing predicted protein circular dichroism spectra from multiple molecular dynamics-generated protein structures. Nucleic Acids Res 2020;48:W17-W24.

[0112] Freund EC, Lock JY, Oh J, Maculins T, Delamarre L, Bohlen CJ, et al. Efficient gene knockout in primary human and murine myeloid cells by non-viral delivery of CRISPR-Cas9. J Exp Med 2020; 217.

[0113] Hopkins JB, Gillilan RE, Skou S. : improvements to a free open-source program for small-angle X-ray scattering data reduction and analysis. J Appl Crystallogr 2017; 50:1545-53.

[0114] Meisburger SP, Xu D, Ando N. : a general method to deconvolve X-ray scattering data from evolving mixtures. ILICrJ 2021 ; 8:225-37.

[0115] Micsonai A, Moussong E, Wien F, Boros E, Vadaszi H, Murvai N, et al. BeStSel: webserver for secondary structure and fold prediction for protein CD spectroscopy. Nucleic Acids Res 2022; 50:W90-W8.

[0116] Micsonai A, Moussong E, Murvai N, Tantos A, Toke O, Refregiers M, et al. Disordered-Ordered Protein Binary Classification by Circular Dichroism Spectroscopy. Front Mol Biosci 2022; 9:863141.

[0117] Mirdita M, Schutze K, Moriwaki Y, Heo L, Ovchinnikov S, Steinegger M. ColabFold: making protein folding accessible to all. Nat Methods 2022; 19:679- 82.

[0118] Piiadov V, Ares de Araujo E, Oliveira Neto M, Craievich AF, Polikarpov I. SAXSMoW 2.0: Online calculator of the molecular weight of proteins in dilute solution from experimental SAXS data measured on a relative scale. Protein Sci 2019; 28:454-63.

[0119] Rannikko, J.H., Verlingue, L., de Miguel, M., Pasanen, A., Robbrecht, D., Skytta, T., livanainen, S., Shetty, S., Ma, Y.T., Graham, D.M., et al. (2023). Bexmarilimab-induced macrophage activation leads to treatment benefit in solid tumors: The phase l / ll first-in-human MATINS trial. Cell Rep Med 4, article no: 101307.

[0120] Silla, T., Haal, L, Geimanen, J., Janikson, K., Abroi, A., llstav, E., and llstav, M. (2005). Episomal maintenance of plasmids with hybrid origins in mouse cells. J Virol 79, 15277-15288.

[0121] Thery, C., Amigorena, S., Raposo, G., and Clayton, A. (2006). Isolation and characterization of exosomes from cell culture supernatants and biological fluids. Curr Protoc Cell Biol Chapter 3, Unit 3.22.

Claims

Claims:1 . A polypeptide having T cell binding activity, consisting of an amino acid sequence of SEQ ID NO: 1 .

2. An isolated polynucleotide encoding the polypeptide of claim 1 .

3. The isolated polynucleotide according to claim 2, wherein it has a nucleotide sequence comprising or consisting of a nucleotide sequence having at least 90 %, preferably at least 95% identity to SEQ ID NO: 2.

4. Use of the polynucleotide according to claim 2 or 3 for producing a recombinant Clever-1 fragment having T cell-binding activity.

5. An expression vector comprising the polynucleotide according to claim 2 or 3.

6. The expression vector according to claim 5, wherein the nucleotide sequence is codon-optimized cDNA.

7. The expression vector according to claim 5 or 6, wherein the nucleotide sequence is cloned into the expression vector between an N-terminal insulin signal peptide and a C-terminal Hiss-tag.

8. A recombinant host comprising the expression vector according to any one of the preceding claims 5-7.

9. A method for producing a recombinant Clever-1 fragment having T cell binding activity, the method comprises cultivating the recombinant host according to claim 8 under conditions conducive for production of the recombinant Clever-1 fragment, and recovering the recombinant Clever-1 fragment.

10. The method according to claim 9, wherein the recombinant host is mammalian expression system.11 . The method according to claim 10, wherein the mammalian expression system comprises CHO cells.

12. A recombinant Clever-1 fragment comprising an amino acid sequence of SEQ ID NO: 1.

13. The recombinant Clever-1 fragment according to claim 12, wherein the fragment is produced by the method according to any one of the preceding claims 9 -11.

14. A recombinant Clever-1 fragment comprising an amino acid sequence SEQ ID NO: 1 for use as a medicament.

15. A recombinant Clever-1 fragment comprising an amino acid sequence SEQ ID NO: 1 for use in a treatment of inflammation and / or autoimmune diseases.

16. A pharmaceutical composition, wherein the composition comprises recombinant Clever-1 fragment comprising an amino acid sequence SEQ ID NO: 1 , and a pharmaceutical acceptable carrier or excipient.

Images