METHODS FOR PREPARING EXTRACELLULAR VESICLES (EV) DEPLETEMEDIA

Abstract

Cell therapy is getting a growing interest in a wide range of indications in human. In many cases, a substantial part of the therapeutic effects relies on cell-secreted factors and the extracellular vesicles (EV) are proposed as a cell-free surrogate for cell therapy. Currently, during the EV production phase, human cells are placed in serum-free media to produce EV, with limited cell survival. Here, the inventors describe a new procedure for GMP-compatible human cells-derived EV wherein Human Platelet Lysate (HPL) is produced from which the EV are removed by tangential-flow-filtration resulting in an EV-depleted HPL. Said EV-depleted HPL may be then uses as a culture medium for the production of EV by cells of interest.

Claims

1. A method for preparing an extracellular vesicle (EV)-depleted medium by removing extracellular vesicles from a medium, comprising the steps of i) filtering the medium by tangential-flow filtration with a filter having a pore size between 100 kDa and 50 nm, a trans-membrane pressure (TMP) between 1 and 6 psi and a shear rate between 2000 and 8000 s.sup.−1 and ii) collecting the permeate after said tangential-flow filtration, wherein the permeate is the (EV)-depleted medium.

2. The method of claim 1 wherein the medium is a serum free medium.

3. The method of claim 1 wherein the medium is a platelet lysate.

4. The method of claim 1 wherein the pore size of is about 100 kDa, about 150 kDa, about 200 kDa, about 250 kDa, about, 300 kDa, about 350 kDa, about 400 kDa, about 500 kDa, about 550 kDa, about 600 kDa, about 650 kDa, about 700, kDa, about 750 kDa, about 750 kDa, about 800 kDa, about 850 kDa, about 900 kDa, about 950 kDa, or about 1000 kDa.

5. The method of claim 1, wherein the pore size is about 500 kDa.

6. The method of claim 1 wherein the filter comprises a hollow fiber module comprising a bundle of filter membranes, each filter membrane being shaped in the form of a hollow tube.

7. The method of claim 1 wherein the TMP is about 1 psi, about 1.5 psi, about 2 psi, about 2.5 psi, about 3 psi, about 3.5 psi, about 4 psi, about 4.5 psi, about 5 psi, about 5.5 psi or about 6 psi is used.

8. The method of claim 1 wherein the TMP is about 2 psi.

9. The method of claim 1 wherein the shear rate of is about 2000 s.sup.−1, about 2500 s.sup.−1, about 3000 s.sup.−1, about 3500 s.sup.−1, about 4000 s.sup.−1, about 4500 s.sup.−1, about 5000 s.sup.−1, about 5500 s.sup.−1, about 6000 s.sup.−1, about 6500 s.sup.−1 about 7000 s.sup.−1, about 7500 s.sup.−1, or about 8000 s.sup.−1.

10. The method of claim 1 wherein the shear rate is about 4000 s.sup.−1.

11. An extracellular vesicles-depleted medium obtainable by the method of claim 1.

12. A method for producing extracellular vesicles from a population of cells comprising the steps of i) preparing an EV-depleted medium by the method of claim 1, ii) culturing the population of cells in a culture medium supplemented by the EV-depleted medium under conditions that allow the production of EV by the population of cells and iii) harvesting the EV that are produced at step ii).

13. The method of claim 12 wherein the population of cells is a population of mesenchymal stem cells.

14. The population of extracellular vesicles obtainable by the method of claim 12.

15. (canceled)

16. The method of claim 3 wherein the platelet lysate is a human platelet lysate.

17. A pharmaceutical composition comprising EV prepared by the method of claim 12, wherein the EV are loaded with or coupled to a therapeutic agent; and a pharmaceutically acceptable carrier.

18. The pharmaceutical composition of claim 17, wherein the therapeutic agent is a small molecule, a protein or a nucleic acid molecule.

19. The pharmaceutical composition of claim 17, wherein the pharmaceutical composition is formulated for topical, parenteral, intravenous, intraarterial, cutaneous, subcutaneous, intratumoral, intramuscular, intracranial, intraorbital, ophthalmic, intraventricular, intrahepatic, intracapsular, intrathecal, intracisternal, intraperitoneal, intranasal, aerosol, suppository, or oral administration.

20. A method of providing a therapeutic agent to a patient in need thereof, comprising administering to the patient a therapeutically effective amount of the pharmaceutical composition of claim 17.

Description

FIGURES

[0032] FIG. 1: Permeate flux expressed as liter/square meter/hour (LMH) as a function of the trans-membrane pressure (psi) following HPL filtration through a 500 kDa pore size hollow fiber filter.

[0033] FIG. 2: Permeate flux expressed as liter/square meter/hour (LMH) as a function of the concentration factor of the retentate compartment following filtration of HPL through 100 kDa (triangle), 500 kDa (diamond) or 50 nm (square) pore size hollow fiber filter.

[0034] FIG. 3: Size exclusion chromatography analysis of the protein content (OD280) of HPL (HPL mix 41), 100 kDa, 500 kDa and 50 nm HPL permeate.

[0035] FIG. 4: Size exclusion chromatography analysis of the protein content (OD280) of 100 kDa, 500 kDa and 50 nm HPL retentate.

[0036] FIG. 5: Nanoparticule Tracking Analysis (NTA) quantification of EV removal from HPL by Tangential Flow Filtration. Intensity versus size representation of the population of EV. Taking into account the dilution of the samples and the presence of EV in the diluent (PBS), the depletion obtained in this case is 99.8%.

[0037] FIG. 6: Number of cells obtained over 3 cycles of EV production of 72 h. Three conditions are compared: Medium, Medium+EV-Free HPL at 5% and Medium+EV-Free HPL at 8%. The results are presented as ratios. The numbers of cells in the EV-Free HPL conditions are expressed relative to the number in the Medium condition at the same time-point.

[0038] FIG. 7: NTA quantification of conditioned medium from MSC incubated for 72 h in α-MEM supplemented with 5% EV-free HPL. Intensity versus size representation of the population of EV produced by MSC in EV-free HPL containing medium.

[0039] FIG. 8: EV concentration in conditioned media obtained over 3 cycles of EV production of 72 h. Three conditions are compared: Medium, Medium+EV-Free HPL at 5% and Medium+EV-Free HPL at 8%.

EXAMPLE 1

[0040] We propose a new protocol to produce EV-depleted media from Human Platelet Lysate (HPL). For this, we decided to adopt the tangential flow filtration (TFF) system proposed by

[0041] Spectrum Laboratories. The R & D version of this device (KrosFlo Research IIi) combined with a hollow fiber filtration system allows ultrafiltration (molecular weight from 1 to 1000 kDa) and/or microfiltration (Sizes from 0.05 to 0.65 microns) of sample volumes ranging from 1 ml to 10 L. Control of continuous feed rate allow the filtration process to operate at constant shear rate. Filtration is fully controlled by three pressure sensors (feed, retentate and permeate sensor). Associated with an automatic backpressure valve, they allow the control of the trans-membrane pressure (TMP) thus ensuring maximum reproducibility of the process. The range of filters proposed allows the separation and the concentration of soluble biomolecules (mainly proteins) contained in different biological fluids (serum, urine, cerebrospinal fluid . . . ) and media conditioned by cultured cells.

[0042] This system, when operated with Spectrum Labs disposable Module-Bag-Tubing (MBT) sets, which are fully assembled and disposable process flow path for TFF, is compatible with regulatory requirements of clinical-grade production units. This gamma-sterile MBT sets are designed for aseptic processing of solution to downstream TFF ultrafiltration. The disposable flow path including the filter, pressure transducers, tubing and fittings completely eliminates the possibility of cross contamination. Last but not least, this process is fully scalable allowing to use hollow-fiber with filtration surface that fit the volume of HPL to be EV-depleted. This allows a cost reduction of research laboratory production.

[0043] Using optimized ultrafiltration parameters (filter surface, shear force, TMP and membrane pore size (or MW Cut-off)), we found this system suitable to separate the vesicular component of HPL (vesicles larger than 30 nm) from the soluble proteins part.

[0044] Shear force: This force, provided by the circulation rate and applied tangentially to the filtration membrane permanently sweeps any un-filtrated material from the membrane surface thus preventing clogging. It is calculated as: (8×Velocity (m.Math.s.sup.−1))/Fiber Internal Diameter (m) with sec−1 units.

[0045] A circulation rate (feed rate) that provides an intermediate shear force, between 4,000 and 8,000 sec.sup.−1, is a good starting point for processing low fouling streams. However, for feed streams containing fragile components such as extracellular vesicles that may be damaged by high circulation rates or high temperatures, shear forces between 2,000 to 4,000 sec.sup.−1 are recommended, therefore 4000 sec.sup.−1 was preferred in this study.

[0046] Trans Membrane Pressure (TMP): Using pressure transducer, the measurement of feeding pressure (P.sub.feed), retentate pressure (P.sub.retentate) and permeate pressure (P.sub.permeate) allow the continuous measurement of the trans-membrane pressure (TMP) using the following equation: TMP=((P.sub.feed+P.sub.retentate)/2)−P.sub.permeate. The use of a backpressure valve that pinch the retentate tubing allows TMP to be automatically and permanently adjusted to a preset value. Processed liquid flux (permeate flux), normalized to the filter surface and expressed as L/m.sup.2/h (LMH) will typically increase as a function of TMP. However, depending on the circulation rate, the flux improvement may become asymptotic as TMP increase because of compaction of the macromolecules that create a “gel layer resistance”.

[0047] TMP between 1 and 6 psi were tested on the initial filtration phase of HPL (up to 2 fold concentrations of the retentate compartment) through a 500 kDa pore size hollow fiber filter operating at a shear force of 4000 sec−1 (FIG. 1).

[0048] All TMP tested can be used in these conditions but a TMP of 2 psi was preferred because it gives the highest permeate flux. This confirms that when filtrating complex solutions such as HPL or serum, limited TMP should be used to favor filtration efficiency. Thus, controlling the permeate backpressure (or permeate flux rate) may reduce the tendency of the membrane to foul in the initial steps of the concentration, providing an overall higher average flux rate.

[0049] Membrane pore size: Then, the capacity to deliver EV-free HPL was evaluated using hollow fiber filters of 3 different pore size (100 kDa, 500 kDa and 50 nm). This test was performed with shear force of 4000 s.sup.−1 and TMP of 2 psi. FIG. 2 shows the permeate flux as a function of the concentration factor (CF) of the retentate up to 10. This CF allows the production of a filtrated HPL volume of 90% of the initial HPL. In any case, permeate flux decreases rapidly in the initial filtration phase of TFF. Evolution of permeate flux of 500 kDa and 50 nm filters are very similar with values of 19.75 and 20.49 L/m.sup.2/h on the overall filtration process. Filtration through 100 kDa filter was nearly 50% less efficient with a mean LMH of 11.53.

[0050] TFF through both 500 kDa and 50 nm pore size hollow fiber filters were more efficient than 100 kDa in term of filtration rate.

[0051] HPL samples and their different TFF permeate were further analyzed by size exclusion chromatography on Superose 6 increase chromatography column, connected to an FPLC AKTA from GE-Healthcare. Protein content of the column eluate was monitored online with a spectrophotometer through its optic deviation (OD) at 280 nm. As shown on FIG. 4, elution profile of 500 kDa and 50 nm permeate are very similar to that of the HPL source except the higher molecular weight proteins or protein complexes eluted before 15 min.

[0052] These high molecular weight components are retained in the retentate fraction (FIG. 4). The 100 kDa filter retains a great part of the protein content of the HPL since only half of the main protein, human serum albumin, was found in the permeate (FIG. 3). As a consequence, its amount in the retentate fraction was very high (FIG. 4).

[0053] The amount and size of the EV in the HPL as well as in the different permeate fractions were determined by Nanoparticle Tracking analysis (NTA) using the NS300 apparatus from MALVERN-PANALYTICAL. HPL contained a very high amount of EV, i-e 9.28 10.sup.10 EV/ml that was decreased by 96.9%, 98.6% and 98.2% using 100 kDa, 500 kDa and 50 nm filters respectively.

[0054] Taking into account all these results, the 500 kDa pore size filter was chosen for the production of EV-free HPL because it retains as much EV as the 50 nm filters with equal filtration rate and similar composition of the permeate fraction (EV-free fraction). Finally, the choice of the 500 kDa instead of the 50 nm filter has also been directed according to our hypothesis that it could produce a safer product, devoid of small size virus.

[0055] Conclusion:

[0056] Optimal conditions for EV-free HPL production were set to, [0057] Shear force: should be between 2000 and 8000 s.sup.−1 but 4000 s.sup.−1 is preferred [0058] TMP: should be between 1 and 6 psi but 2 psi is preferred. [0059] Filter pore size should be between 100 kDa and 50 nm but 500 kDa pore size filter is preferred.

EXAMPLE 2

[0060] Using the parameters determined at EXAMPLE 1, another HPL batch was used to produce 1 L of EV-free HPL. Starting from a 1.1 L solution of undiluted HPL, it took about 4 h with a 155 cm.sup.2 hollow-fiber filter to produce 1 L of 99.8% EV-depleted HPL as determined by NTA analysis (FIG. 5).

[0061] Increasing the filtration surface to 1600 cm.sup.2 would allow the production of 10 L of EV-depleted HPL in just the same time (4h) (enough for 100 L of culture medium containing 10% of EV-depleted HPL). In this configuration, EV are retained in the retentate compartment and the permeate constitutes the EV-depleted HPL. The permeate is sterile and free of any bacteria, mycoplasma and virus. Stock of EV-free HPL can be stored frozen (−20 to −80° C.) and used in addition of any culture media, for many different cell types and at various concentration since EV-depletion of HPL occurs before dilution in the culture medium.

EXAMPLE 3

[0062] Mesenchymal stromal cells (MSCs) are multipotent cells found in a large number of adult tissues. Many studies highlighted their ability to participate in the repair of damaged tissues. They exert immunomodulatory, anti-apoptotic, pro-angiogenic, growth support of stem or progenitor cells, anti-fibrotic or chemoattractant effects by secreting a wide range of bioactive molecules. Therefore we used bone marrow MSCs to produce EV with EV-depleted media as prepared in EXAMPLES 1 and 2.

[0063] To test the validity of using EV-free HPL-containing medium instead of basal medium without HPL or serum, we first examined the ability of this supplement to sustained MSC culture for at least 3 periods of 3 days. MSC are first amplified in their standard culture media (α-MEM supplemented with 5% HPL) and then they are “rinsed” in the presence of medium alone or medium supplemented with EV-free HPL. The cells are then placed in medium supplemented or not with EV-Free HPL for a first secretory phase of 72 h. At 72 h the culture medium is recovered for the EV quantification. A sample of cells is harvested for counting. The rest of the cells are replaced in the presence of the same culture conditions again for 72 hours. We can thus perform several cycles of production of EV.

[0064] Thus, we can show for example on human Mesenchymal Stromal Cells (MSCs), adherent stem cells derived from the human bone marrow, that the presence of EV-Free HPL allows maintaining them longer in culture compared to medium without HPL, thereby increasing the number of EV production cycles (FIG. 6).

[0065] We observe that the Medium containing EV-Free HPL (5 or 8%) preserves more cell viability than the basal Medium condition, on 3 consecutive EV production cycles. Moreover, pictures of cultured cells in the different conditions and all along the incubation time show that the presence of EV-free HPL preserves cell morphology whereas medium without HPL do not (Data not shown).

[0066] We reported on FIG. 7 the NTA quantification of EV contained in the medium either before (T 0 h) or after 72 h of MSC secretion. The results indicate that in the presence of 5% EV-free HPL, the amount of EV increases from 1.7 107/ml to 2.6 108/ml. Thus EV secreted by MSC could be calculated to represent 93.5% of total EV in this particular incubation. It means that EV secreted by MSCs represent 93.5% of the total amount of EV observed at 72 h.

[0067] NTA analyses of all the incubation media from the experiments described above are reported on FIG. 8. We observed that EV-free HPL (either 5 or 8%) allow a sustained production of EV all along the three 72 h incubation periods. EV production in medium without HPL is lower and decreases substantially during the third period. The amount of EV produced among the three consecutive cycles of secretion were respectively of 2.27 109, 5.59 109 and 6.14 109 EV/ml for HPL-free, 5% and 8% EV-free HPL containing medium. Thus, in these experiments, a single run of MSCs amplification would produce 2.4 and 2.7 more EV when incubated in the presence of respectively 5 and 8% EV-free HPL compared to HPL-free medium.

[0068] At the end of the process, EV from EV-Free HPL containing conditioned medium can be isolated and concentrated by any technical approach such as ultracentrifugation, ultrafiltration, size exclusion chromatography, precipitation (PEG or Antibodies) . . . .

[0069] Conclusion:

[0070] Therefore our preparation process of EV-Free HPL allows the production of large amounts of EV from various human cells, compatible with clinical use in different therapeutic applications. This process can be extended to any EV-containing animal media such as sera that are used to promote cell survival and/or proliferation from different animal species. It is compatible with the production of large volumes of conditioned media, including in bioreactors, allowing the large-scale production of therapeutic EV for both human and veterinary applications.

EXAMPLE 4

[0071] Ultracentrifugation

[0072] A volume of 2 ml of fetal bovine serum (FBS) on the one hand and of human platelet Lysate (hPL) on the other hand, both pure, are centrifuged at 120,000 g for 18 hours at 22° C. (TL100 optima max XP, rotor MLS50, k factor=159). Dilution of Serum or Platelet Lysate for EV depletion of by ultracentrifugation was recommended in MISEV 2018 guidelines (J. Extracell. Vesicle. 2018, Vol 7). The effect of 1:10 dilution of both additives in αMEM culture medium on the efficiency of the UC step was also evaluated

[0073] After these ultracentrifugations, 75% of the initial volume, i.e. 1.5 ml of the supernatants is recovered. EV present before and after centrifugation (supernatants) were quantified by NTA following suitable dilutions.

[0074] Tangential Flow Filtration

[0075] TFF of both FBS and hPL were performed exactly as described in this patent. For comparison with the UC protocol, EV depletion by TFF of pure and 1:10 dilution of both additives was analyzed. EV contained FBS and haply before TFF and in the TFF filtrate fraction were quantified by NTA following suitable dilutions.

[0076] Results

[0077] The amount of EV remaining in the supernatant following UC or in the filtrate following TFF was express as a percentage of EV initially present before UC or TFF. Results are reported in the Table below as mean+/−SE of the number of determinations (n) in each condition. Pure FBS and Pure hPL refer to undiluted FBS and hPL. 10% FBS and 10% hPL refer to 1:10 dilution of FBS and hPL in αMEM medium.

TABLE-US-00001 Percentage of remaining Percentage of remaining EV after UC EV after TFF Mean +/− SE n Mean +/− SE n Pure FBS 57.3 +/− 7.1 2 1.7 +/− 1.5 11 αMEM 10% 27.1 +/− 3.5 3 3.0 +/− 1.4 2 FBS Pure hPL 71.8 +/− 9.4 3 0.6 +/− 0.6 5 αMEM 10% 46.5 1 0.5 1 hPL

[0078] Conclusion

[0079] TFF is far more efficient than UC in depleting both FBS and hPL from endogenous EV. Dilution of both FBS and hPL only marginally increase the efficiency of depletion by UC whereas depletion by TFF remains maximal.

REFERENCES

[0080] Throughout this application, various references describe the state of the art to which this invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure.