PROCESS FOR THE PREPARATION OF LIPIDATED PROTEINACEOUS STRUCTURES

Abstract

The present invention relates to a process for the preparation of a conjugate, comprising a biological molecule, an enzymatic tag, a hydrophilic spacer, a linker and a lipophilic moiety using enzymatic coupling.

Claims

1. Process for the preparation of a conjugate, wherein the conjugate comprises a biological molecule, an enzymatic tag, a hydrophilic spacer, a linker and a lipophilic moiety, such process comprises enzymatic coupling of a component comprising the enzymatic tag, the hydrophilic spacer, the linker and the lipophilic moiety with the biological molecule in an aqueous medium and purification of the conjugate.

2. Process according to claim 1, wherein the biological molecule is a polypeptide.

3. Process according to claim 2, wherein the biological molecule is an antigen, a cell adhesion protein such as an integrin or a cadherin, a peptide hormone such as a growth factor, a cytokine such as an interleukin, or a receptor related to any of these molecules, an enzyme, or a natural or artificial antibody or a fragment thereof.

4. Process according to claim 3, wherein the antibody is a monoclonal antibody or a fragment thereof, such as a single-chain variable fragment (scFv), a variable fragment (Fv), or a fragment antigen binding (Fab, Fab′ or F(ab′)2); a camelid or cartilaginous fish-derived heavy-chain only antibody or a fragment thereof, such as a VHH or a vNAR, or wherein the artificial antibody is a DARPin, a adnectine, an anticalin, or an affibody.

5. Process according to claim 3, wherein the biological molecule is a single-domain antibody derived from the variable domain of camelid heavy-chain only antibodies (VHH).

6. Process according to claim 1, wherein the biological molecule prior to its coupling with the component comprising the enzymatic tag, the hydrophilic spacer, the linker, and the lipophilic moiety carries a C-Terminal motif for enzymatic conjugation by transpeptidases, preferably sortase A.

7. Process according to claim 6, wherein the C-terminal motif consists of the amino acid sequence “leucine—proline—X—threonine—glycine” (LPXTG), wherein “X” can be any proteinogenic amino acid.

8. Process according to claim 7, wherein the proteinogenic amino acid present in the LPXTG motif is glycine, alanine, valine, leucine, isoleucine, methionine, phenylalanine, proline, serine, threonine, tyrosine, asparagine, glutamine, aspartic acid, glutamic acid, lysine, arginine or histidine, preferably glutamic acid

9. Process according to claim 1, wherein the enzymatic tag is an aliphatic amine, a single glycine, a peptide sequence with one or more N-terminal glycines, preferably pentaglycine, which is C-terminally linked to the hydrophilic spacer.

10. Process according to claim 1, wherein the hydrophilic spacer is a polyoxy(C.sub.2-C.sub.3)alkylene (e.g. polyethylene glycol or polypropylene glycol), a polysaccharide (e.g. dextran, pullulan, chitosan, hyaluronic acid), a polysialic acid, a polyethyleneimine, preferably polyethylene glycol.

11. Process according to claim 1, wherein the linker is isoglutamine, on the δ-position amide linked with a 3-amino-1,2-propanediol, and on the α-standing amine-function linked to the hydrophilic spacer.

12. Process according to claim 1, wherein the lipophilic moiety is or are one or more, independently of each other, saturated or unsaturated, straight or branched hydrocarbon chains, such as fatty alcohols or fatty acids with chain lengths of 6-30 carbon atoms, or sterols such as cholesterol.

13. Process according to claim 12, wherein the lipophilic moiety comprises two myristyl alcohols each ether linked to the diol-group of the 3-amino-1,2-propanediol.

14. Process according to claim 1, wherein the component comprising an enzymatic tag, a hydrophilic spacer, a linker and a lipophilic moiety (Component A) is DMA-PEG-G5.

15. Process according to claim 1, comprising the steps (a) preparing an aqueous dispersion of a component comprising an enzymatic tag, a hydrophilic spacer, a linker and a lipophilic moiety; (b) adding an enzyme and a biological molecule; (c) incubating the mixture obtained in step (b) to produce the conjugate; (d) purifying the conjugate obtained in step (c).

16. Process according to claim 15, wherein the enzyme in step (b) is a ligase, including sortase, butelase, trypsiligase, subtiligase, peptiligase and omniligase, preferably sortase A.

17. A method for the modification of lipid based drug delivery systems such as solid-lipid nanoparticles, nanoemulsions, micelles or liposomes, preferably liposomes to claim 1.

18. A method for the modification of membranes of living cells, preferably T cells by the conjugate obtained by the process according to claim 1.

19. A method for the modification of surfaces with affinity towards hydrophobic substances, such as hydrophobic polystyrene by the conjugate obtained by the process according to claim 1.

20. A method for the modification of membranes of exosomes by the conjugate obtained by the process according to claim 1.

Description

EXAMPLE 1

[0146] Stock of structure 2, in FIG. 1, further called here DMA-PEG-G5 (25 mg/mL in chloroform) was aliquoted in a HPLC vial. Chloroform was evaporated under a gentle stream of nitrogen creating a thin lipid film, which was hydrated with DPBS (Dulbecco's phosphate buffered saline, Sigma Aldrich, # D1408) pH 7.4 as micellar aqueous lipid dispersion (2 mM). Alternatively, DMA-PEG-G5 was dissolved in DPBS pH 7.4 (5-10 mg/mL) using ultrasound. 400 μL of 50 μM VHH, 25 μM Sortase-A and 1 mM DMA-PEG-G5, which corresponded to a target product mass of 320 μg, were incubated for 4 h at 4° C., until the conjugate was isolated using reversed-phase HPLC (rp-HPLC). Reaction bulk and product were separated utilizing an Aeris Widepore C4 column (3.6 μm particle size, 100 mm length, 2.1 mm diameter, Waters Corporation, Milford, Mass., USA) and a binary gradient pattern (Table).

TABLE-US-00002 TABLE 2 rp-HPLC gradient pattern for separation of reaction bulk and product (A: water with 0.1% trifluoroacetic acid (TFA v/v; B: acetonitrile with 0.05 % TFA v/v) time [min] solvent composition 0   95% A 5.5 45.5% A 15 18.5% A 15.1   5% A 18   5% A 18.1   95% A 20   95% A

[0147] Eluent A was water with 0.1% TFA, eluent B was acetonitrile with 0.05 TFA. Analysis was performed on an Agilent 1110 HPLC system equipped with a degasser, binary pump, temperature controlled autosampler, column oven, diode array detector (DAD) and an analytical fraction collector (AFC), controlled by EZChrom Elite Software (Agilent Technologies, Santa Clara, Calif., USA). Column temperature was set to 30° C., autosampler temperature was 4° C. Standard analytical or isolation injection volume was 5 μL or 25-100 μL, respectively. Flow rate was 0.5 mL/min. Data was recorded using DAD at 214 and 280 nm.

[0148] The conjugate peak was collected either manually or using an automated fraction collector. Collections of single injections were stored on ice, until the eluent mixture was removed using a vacuum centrifuge (RVC 2-33 IR, Martin Christ, Osterode am Harz, Germany; speed: 1500 rpm; temperature: 40° C.; 100 mbar for 10 min, followed by 20 mbar for 20 min and further evaporation at 2 mbar). The resulting pellet was hydrated with water, and protein concentrations were determined by UV spectroscopy (NP80, Implen, Westlake Village, Calif., USA) using extinction coefficients calculated by ExPASy ProtParam web application (https://web.expasy.org/protparam, SIB, Lausanne, Switzerland). Yield was calculated based on the mass and concentration of the recovered protein solution and the target product mass. Purity was determined by rp-HPLC analysis at 214 nm with automated peak detection between 2-15 min and a threshold level obtained from background noise of a water blank.

[0149] To verify the reaction product, the method was transferred to a similar HPLC system equipped with an electrospray ionization mass spectrometer (ESI-MS, amaZon SL, Bruker Corporation, Billerica, Mass., USA). Ion source type was set to ESI with positive polarity. Capillary exit was 140 V, trap drive was set to “94”. The mass range mode was set to enhanced resolution, with a scanning range from 100-2200 m/z. 5 spectra were averaged per run. Masses were calculated using deconvolution of raw spectra.

EXAMPLE 2

[0150] The isolated VHH ENH conjugate from Example 1 or native VHH ENH (100 nM) was spiked to 100 nM eGFP. VHH ENH is known for its ability to increase the intrinsic fluorescence of eGFP upon binding. Increase in fluorescence intensity compared to sole eGFP was measured in a black 96-well plate (Thermo Fisher Scientific, Waltham, Mass., USA) utilizing a Spark Plate Reader (Tecan Group, Männedorf, Switzerland) with excitation and emission set to 485 nm and 535 nm, respectively.

EXAMPLE 3

[0151] FITC-dextran (fluorescein isothiocyanate) labelled liposomes were prepared as described elsewhere [4]. In brief, a mixture of DPPC (1,2-dipalmitoylphosphatidylcholine) cholesterol, DPPG (1,2-Dipalmitoyl-phosphatidylglycerole) and DMA-PEG-G5 (59.4:34.6:5.0:1.0, molar fractions) was dissolved to 32 mM in methanol and injected via a computer-controlled binary pumping system into a 10 mg/mL FITC (fluoresceine isothiocyanate) dextran solution in DPBS pH 7.4 utilizing a customized T-piece with a 27G needle. The dispersion was purified and concentrated by tangential flow filtration. Lipid concentration was determined by an rp-HPLC method with evaporative light scattering detection described elsewhere [5]. To prepare immunoliposomes, the lipidated VHH ENH or VHH DC13 were added to the liposomal dispersion to 0.25-2 nM VHH per μM phospholipid (PL). The mixture was thoroughly vortexed and incubated at 50° C. for 30 min.

[0152] Murine myeloid-derived suppressor CD11b+Gr-1+ cells (MDSC) were derived from bone marrow-derived NUP-progenitor cells [6]. MDSC were differentiated for four days in complete RPMI (RPMI 1640 medium, Life Technologies, #21875-034, Carlsbad, Calif., USA) supplemented with 10% heat-inactivated fetal bovine serum, 100 U/mL penicillin (Life Technologies, #15140122), 100 μg/mL streptomycin (Life Technologies, #15140122), 1 mM sodium pyruvate (Life Technologies, #11360070), 50 μM 2-mercaptoethanol (Life Technologies, #31350-010), and 1×non-essential amino acids (Life Technologies, #11140-035) supplemented with 20 ng/mL interleukine-6 and 20 ng/mL granulocyte-macrophage colony-stimulating factor (Biolegend, #576304, San Diego, USA). To investigate binding of VHH-modified liposomes, MDSC were incubated with 500 μM of the liposomes (based on total lipid content) for 4 h at 4° C. Cells were washed with FACS buffer (1×PBS+2% heat-inactivated fetal bovine serum) and antibody staining of cells was performed in presence of Fc receptor block (TruStain FcX, BioLegend, #422302) in FACS buffer. SytoxBlue (Thermo Fisher Scientific, S34857) was used for exclusion of dead cells. Liposomes were detected via encapsulated FITC-dextran.

EXAMPLE 4

[0153] T cells were isolated from spleens of C57BU6j mice maintained under specific pathogen-free conditions at the animal facility of the University of Heidelberg and euthanized under the registered protocol T47/16. Spleens were mashed and CD8+ cells isolated after red cell lysis (ACK lysing buffer, #A1049201, Thermo Fisher Scientific) using a mouse CD8a+ T cell isolation kit (#130-104-075, Miltenyi Biotec, Bergisch-Gladbach, Germany) and magnetic cell isolation (LS columns, #130-042-401, Miltenyi Biotec) used according to manufacturer's instructions. Purified CD8+ T cells were stained by 1 nM Cell Tracer Far Red (#C34564, Thermo Fisher Scientific) for 5 min at 35° C. and washed with FACS buffer. Stained T cells (1.65×108 cells/mL) were incubated with 650 nM native or lipidated VHH DC13 and VHH ENH for 1 h at 4° C. Lipidated VHH binding to T cells was detected by an FITC-anti-llama antibody.

[0154] Murine myeloid-derived suppressor CD11b+Gr-1+ cells (MDSC) were derived from bone marrow-derived NUP-progenitor cells [6]. MDSC were differentiated for four days in complete RPMI (RPMI 1640 medium, Life Technologies, #21875-034, Carlsbad, Calif., USA) supplemented with 10 heat-inactivated fetal bovine serum, 100 U/mL penicillin (Life Technologies, #15140122), 100 μg/mL streptomycin (Life Technologies, #15140122), 1 mM sodium pyruvate (Life Technologies, #11360070), 50 μM 2-mercaptoethanol (Life Technologies, #31350-010), and 1×non-essential amino acids (Life Technologies, #11140-035) supplemented with 20 ng/mL interleukine-6 and 20 ng/mL granulocyte-macrophage colony-stimulating factor (Biolegend, #576304, San Diego, USA). To investigate membrane insertion of lipidated VHHs, MDSC (10.sup.7cells/mL) were incubated for 30 min at 4° C. with 500 nM of native or lipidated VHH ENH or VHH DC13. Cells were washed with FACS buffer (1×PBS+2% heat-inactivated fetal bovine serum) and antibody staining of cells was performed in presence of Fc receptor block (TruStain FcX, BioLegend, #422302) in FACS (fluorescence assisted cell sorting) buffer. SytoxBlue (Thermo Fisher Scientific, S34857) was used for exclusion of dead cells. Lipidated VHH inserted into the cell membrane was detected by a FITC-anti-llama antibody (Invitrogen, #A16061). All analyses of cells were performed by flow cytometry. Flow cytometry was performed on a FACSAria II (Beckton, Dickinson and Company, Franklin Lakes, N.J., USA) and results were analyzed by FlowJo (Tree Star, V.10.0.8).

EXAMPLE 5

[0155] CD11b+Gr-1+ cells from example 4 were incubated with 100 μg/mL eGFP at 4° C. for 30 min to detect binding of eGFP to lipidated VHH ENH being inserted into the cell membrane. Flow cytometry was performed on a FACSAria II (Beckton, Dickinson and Company, Franklin Lakes, N.J., USA) for fluorescence of eGFP, results were analyzed by FlowJo (Tree Star, V.10.0.8).

EXAMPLE 6

[0156] For cell-cell interaction experiments, T cells were isolated from spleens of C57BL/6j mice maintained under specific pathogen-free conditions at the animal facility of the University of Heidelberg and euthanized under the registered protocol T47/16. Spleens were mashed and CD8.sup.+ cells isolated after red cell lysis (ACK lysing buffer, #A1049201, Thermo Fisher Scientific) using a mouse CD8a.sup.+ T cell isolation kit (#130-104-075, Miltenyi Biotec, Bergisch-Gladbach, Germany) and magnetic cell isolation (LS columns, #130-042-401, Miltenyi Biotec) used according to manufacturer's instructions. Purified CD8.sup.+ T cells were stained by 1 nM Cell Tracer Far Red (# C34564, Thermo Fisher Scientific) for 5 min at 35° C. and washed with FACS buffer. Stained T cells (1.65×10.sup.8 cells/mL) were incubated with 650 nM native or lipidated VHH DC13 and VHH ENH for 1 h at 4° C. VHH-labeled T cells were washed twice with FACS buffer and 3×10.sup.7 T cells were incubated with 1.1×10.sup.7 MDSC (obtained as described in example 4) for 1 h at 4° C. T cells and MDSC were loaded on LS columns for magnetic bead isolation of CD8.sup.+ T cells and co-purification of MDSC bound to T cells. Eluted cells were stained by anti-CD11b-Brilliant Violet 605 and anti-Gr-1-FITC (#101237 and #108405, Biolegend) and analyzed by flow cytometry as described above.

EXAMPLE 7

[0157] Native or lipidated VHH ENH (batch #3 of example 1) were mixed with eGFP in DPBS pH 7.4 (1×, Sigma, # D1408) in different variations in a ThermoFisher Polysorp 96-well plate (# Nunc 475094). The variations included: PBS, PBS with eGFP (50 nM), PBS with eGFP (50 nM)+native VHH ENH (50 nM), PBS with eGFP (50 nM)+lipidated VHH ENH (5-50 nM). The plates were incubated at 60 rpm, 37° C. for 1.5 h in an orbital shaker. Afterwards, the plate was centrifugated for 1 min@300 g to collect all liquid at the bottom of the well. The plate was then measured using TecanReader Spark (ThermoFischer) with an excitation of 485 nm and emission of 525 nm. Afterwards, the liquid in the wells was exchanged for 5 times. After each the 1., 2., 3. and 5. exchange step, fluorescence was measured.

[0158] Surprisingly, it has been found that a compound composed of an enzymatic tag, an hydrophilic spacer, a linker and a lipophilic moiety (Structure 2 in FIG. 1) can be dissolved in aqueous buffers up to 10 mg/mL (4 mM) (Example 1).

[0159] After preparation of a mixture of 1 mM of Structure 2, 25 μM of the transpeptidase sortase A and 50 μM of an LPETG-modified singe-domain antibody (Example 1) (Structure 1), an efficient lipidation of structure 1 was observed after 4 h after analysis by reversed phase HPLC (FIG. 3, exemplary chromatogram for VHH ENH utilized as structure 1).

[0160] Additionally, no aggregation or precipitation of the lipidated product was observed during the reaction as analyzed via visual inspection (Example 1).

[0161] Mass spectrometry confirmed the expected molecular masses for two different single-domain antibodies (VHH ENH (FIG. 4), VHH DC13 (FIG. 5), (Example 1).

[0162] The reaction bulk was purified by collecting the column effluent of the described reversed phase HPLC method of the “lipidated VHH” peak in a glass vial. The eluent mixture composed of water, acetonitrile and trifluoroacetic acid was removed by vacuum centrifugation. The so obtained pellet was could be dissolved in water to ˜1 mg/mL protein content (Example 1).

[0163] The purity of the so obtained lipidated single-domain antibodies was analysed via reversed phase HPLC combined with UV detection at 214 nm. It revealed purities above >95% (based on the UV area) for following batches (FIG. 6, Table 1) (Example 1).

TABLE-US-00003 TABLE 1 purity conjugate [area %] yield VHH ENH lot #1 96% 50% VHH ENH lot #2 97% 52% VHH ENH lot #3 95% 27% VHH DC13 lot #1 97% 60%

[0164] The yield was calculated based on the mass and concentration of the recovered protein solution and the target product mass (320 μg protein per batch). Good yields >50% were obtained, except for VHH ENH lot #3, which was prepared without cooling the column effluent (Example 1).

[0165] The biological activity of the via reversed phase HPLC purified conjugates was analyzed by the ability of the lipidated VHH ENH to increase the fluorescence of eGFP upon binding (Example 2) [1].

[0166] Three different batches were incubated with with eGFP and the fluorescence intensity at 485 nm excitation and 535 nm emission was compared to that of native VHH ENH. The data revealed no losses of binding in lot #1 and lot #2 (FIG. 7) (Example 2). Lo t#3 was prepared without cooling the column effluent and showed slightly decreased fluorescence enhancement of eGFP (Example 2).

[0167] The lipidated and isolated single-domain antibodies VHH ENH and VHH DC13 were incubated with FITC-labelled liposomal drug delivery systems (Example 3) in different ratios of VHH to phospholipid. The via this post-insertion process modified liposomes were then incubated with cultured murine CD11b+Gr-1+ cells. VHH DC13 binds to the cellular surface receptor CD11b, and VHH DC13-modified liposomes showed a clear cellular association with the CD11b+Gr-1+ cells during flow cytometry analysis (FIG. 8, shows data for VHH DC13 surface density on the liposomes of 2 nM/μM phospholipid, percentage in each dot plot indicates number of positive cell in the marked gate. FSC: forward scatter) (Example 3).

[0168] When the liposomes were incubated with different ratios of lipidated VHH DC13 to the phospholipid concentration, an optimum of binding on the cells was observed for 0.5 nM VHH DC13 per μM phospholipid (FIG. 8) (Example 3).

[0169] To assess whether the lipidated VHHs can be used for cell membrane remodelling, CD11b+Gr-1+ cells or T cells were incubated with lipidated or native VHH ENH or VHH DC13 (Example 4). After washing of cells, VHH presence on the cellular surface was confirmed by flow cytometry after staining the cells with an FITC-anti-llama antibody (Example 4, FIG. 10). Lipidated VHHs clearly increase the fluorescence signal obtained from the cells in flow cytometry (Example 4). If VHH DC13 was incubated with CD11b+ cells, no differences between lipidated and non-lipidated form was detected, since the direct binding of VHH DC13 to CD11b overrides the hydrophobic insertion effect (Example 4).

[0170] To assess whether the lipidated VHHs present on the cellular surface are also accessible for soluble antigens, CD11b+Gr-1+ cells were incubated with lipidated or native VHH ENH or VHH DC13 (Example 5). After cell washing, the cells were incubated with the for VHH ENH corresponding antigen eGFP. Flow cytometry revealed a selective capturing of eGFP by cells which had been treated with the lipidated VHH ENH (Example 5, FIG. 11).

[0171] To assess whether the lipidated VHHs present on the cellular surface are also able to promote a cell-cell interaction (Example 6), isolated CD8+ T cells were incubated with lipidated or native VHH ENH or VHH DC13. Afterwards, the so modified cells were incubated with CD11b+Gr-1+ cells. This cell mixture was firstly separated by magnetic bead assisted cell sorting specific for the antigen CD8. Afterwards, CD8.sup.+ retentate was stained for CD11b and Gr-1 and analyzed by flow cytometry (FIG. 14) (Example 6). Cellular populations pre-incubated with lipidated VHH DC13 caused higher signals for MDSC in the CD8-positive column retentate, indicating a VHH promoted cellular interaction between MDSC and T cells (FIG. 13) (Example 6).

[0172] To assess whether lipidated VHHs can be immobilized on hydrophobic surfaces such as hydrophobic polystyrene, lipidated or native VHH ENH were incubated in presence of 50 nM eGFP (the corresponding antigen of VHH ENH) in a Polysorp® 96-well plate (Example 7). After several washing steps, a significant coating and retention eGFP was observed in the wells for 25 nM and 50 nM coating (FIG. 14). The data indicates usability of the lipidated VHHs for antigen or cell capturing or immobilization on cells (Example 7). It could also be used for purification of proteins.