FILTRATION

Abstract

The present invention provides method of removing particles from a feed fluid, the method comprising: passing the fluid through a first filtration medium having a thickness of from 5 to 20 μm, wherein passing the feed fluid through the first filtration medium provides a particle removal probability log10 reduction value (LRV) of greater than or equal to 1 for particles having a diameter of from about 10 to about 40 nm and a particle removal probability log10 reduction value (LRV) of greater than or equal to 3 for particles greater than about 40 nm in diameter; and passing the fluid through a second filtration medium having a thickness of from 20 to 70 μm (e.g. 20 to 45 μm) 20 to 45 pm, wherein passing the feed fluid through the second filtration medium provides a particle removal probability log10 reduction value (LRV) of greater than or equal to 3 for particles having a diameter of from about 10 to about 40 nm and a particle removal probability log10 reduction value (LRV) of greater than or equal to 3 for particles having a diameter of greater than or equal to about 40 nm; so as to retain at least a portion of the particles on each medium to produce a filtrate containing a lower concentration of the particles than the feed fluid.

Claims

1. A method of removing particles from a feed fluid, the method comprising: passing the fluid through a first filtration medium having a thickness of from 5 to 20 μm, wherein passing the feed fluid through the first filtration medium provides a particle removal probability log10 reduction value (LRV) of greater than or equal to 1 for particles having a diameter of from about 10 to about 40 nm and a particle removal probability log10 reduction value (LRV) of greater than or equal to 3 for particles greater than about 40 nm in diameter; and passing the fluid through a second filtration medium having a thickness of from 20 to 70 μm (e.g. 20 to 45 μm), wherein passing the feed fluid through the second filtration medium provides a particle removal probability log10 reduction value (LRV) of greater than or equal to 3 for particles having a diameter of from about 10 to about 40 nm and a particle removal probability log10 reduction value (LRV) of greater than or equal to 3 for particles having a diameter of greater than or equal to about 40 nm; so as to retain at least a portion of the particles on each medium to produce a filtrate containing a lower concentration of the particles than the feed fluid.

2. The method according to claim 1, wherein the first filtration medium has a pore size distribution such that the modal pore diameter is in the range of from 10 to 30 nm (preferably 10 to 25 nm) and wherein the second filtration medium has a pore size distribution such that the modal pore diameter is in the range of from 10 and 25 nm.

3. The method according to claim 1, wherein the filtration media comprise cellulose fibres.

4. The method according to claim 3, wherein the cellulose fibres comprise elementary fibrils of a diameter of greater than or equal to about 10 nm.

5. The method according to claim 1, wherein the particles are selected from aggregates, high molecular weight protein impurities, unfolded or misfolded proteins.

6. The method according to claim 1, wherein the particles are selected from proteins such as soluble and insoluble protein aggregates, high molecular weight protein impurities, unfolded or misfolded proteins, and/or protein prion particles.

7. The method according to claim 1, wherein die particles comprise microorganisms, such as viruses.

8. The method according to claim 1, wherein the particles have a diameter of greater than, or equal to, about 10 nm.

9. The method according to claim 1 wherein passing the teed fluid through the first filtration medium provides a particle removal probability log10 reduction value (JRV) of greater than or equal to 2, for particles having a diameter of from about 10 to about 40 nm.

10. The method according to claim 1 wherein passing the feed fluid through the first filtration medium provides a particles removal pro liability log10 reduction value (LRV) of greater than or equal to 4 for particles greater than, or equal to, about 40 nm in diameter.

11. The method according to claim 1 wherein passing the feed fluid through the second filtration medium provides a particle removal probability log10 reduction value (LRV) of greater than or equal to 4 for particles having a diameter of from about 10 to about 40 nm.

12. The method according to claim 1, wherein passing the feed fluid through the second filtration medium provides a particle removal probability log10 reduction value (LRV) of greater than or equal to 4 for parricles having a diameter of greater than, or equal to, about 40 nm.

13. The method according to claim 1, wherein the fluid is passed through the filtration media under a pressure differential of approximately 3 to 600 kPa.

14. The method according to claim 1, wherein the first and second medium are independently interchangeable.

15. The method according to claim 1, wherein the method comprises the step of passing the feed fluid through at least one pre-filtration membrane prior to passing through the first filtration medium, wherein the at least one pre-filtration membrane has a pore size distribution such that the modal pore diameter is greater than, or equal to, about 100 μm.

16. A kit-of-parts comprising: a first filtration medium having a thickness of from 5 to 20 μm, wherein the first filtration medium provides a particle removal probability log10 reduction value (LRV) of greater than or equal to 1 for particles having a diameter of from about 10 to about 40 nm and a particle removal probability log10 reduction value (LRV) of greater than or equal to 3 for particles greater than or equal to about 40 nm in diameter; and a second filtration medium having a thickness of from 20 to 70 μm (e.g. 20 to 45 μm), wherein the second filtration medium provides a particle removal probability log10 reduction value (LRV) of greater than or equal to 3 for particles having a diameter of from about 10 to about 40 nm and a particle removal probability log10 reduction value (LRV) of greater than or equal to 3 for particles having a diameter of greater than about 40 nm.

17. The kit-of-parts according to claim 16, wherein the first filtration medium has a pore size distribution such that the modal pore diameter is in the range of from 10 to 30 nm (preferably troth 10 to 25 nm) and wherein the second filtration medium has a pore sire distribution such that the modal pore diameter is in the range or from 10 to 25 nm.

18. The kit-of-parts according to claim 16, wherein the filtration media comprise cellulose fibres.

19. The kit-of-parts according to claim 18, wherein the cellulose fibres comprise elementary fibrils of a diameter of greater than about 10 nm.

20. The kit-of-parts according to claim 16, wherein the first and second media of the kit are independently interchangeable.

21. The use of a kit-of-parts according to claim 16 for removing particles from a feed fluid.

Description

DESCRIPTION OF THE FIGURES

[0088] Embodiments of the present invention will now be described with reference to the following drawings:

[0089] FIG. 1 is a graph showing filter flux using a 10 g/L HSA solution in a two-step filtration process with 11 μm and 22 μm nanocellulose based filters at 1 bar.

[0090] FIG. 2 is a graph showing the comparison of second step filtration with 33 μm nanocellulose based filters at 1 bar and 3 bar.

[0091] FIG. 3 is a graph showing the dynamic light scattering (DLS) analysis of HSA solution representing feed, pre-filtrate 11 μm, and filtrate 22 μm at 1 bar.

[0092] FIG. 4 is three graphs, showing the results of size exclusion gel chromatography (ÄKTA) of 10 mg/ml HSA feed, pre-filtrate 11 μm, and filtrate 22 μm at 1 bar.

[0093] FIG. 5 is a graph showing the LRV of ΦX174 phage-spiked 10 mg/ml HSA filtrate with 22 μm filter at 3 bar.

[0094] FIG. 6 is a graph showing filter flux using a 50 g/L HAS solution in a two-step filtration process with 11 μm and 22 μm nanocellulose based filters at 1 bar.

EXAMPLES

[0095] The present invention will be further described by reference to the following non-limiting examples.

Example 1

[0096] The invention can be illustrated by one of the most common plasma-derived human proteins, i.e. human serum albumin (HSA). HSA has many functions in the body, among which binding and transport properties of various hydrophobic substances is among most important functions. HSA is also an important supplement for cell culture media especially in cell therapies. It can be used for cryopreservation of cell therapies. Therefore, it is critical that human serum albumin is virus- and other pathogen free for these applications. Albumin is a single polypeptide chain of 585 amino acids with molecular weight around 65-67 kDa. The structure is tightly coiled due to numerous thiolic bonds. It is well established that HSA may undergo extensive aggregation due to disulfhydryl binding and various hydrophobic interactions. For this reason, HSA is a suitable model protein to demonstrate the invention.

[0097] A commercially available HSA sample (200 mg/ml) was purchased from a local pharmacy store. The sample was diluted with phosphate buffer solution (PBS) pH 7.4 to 10 g/L concentration.

[0098] The cladophora cellulose dispersion was prepared by passing the starting cellulose material through a high-pressure microfluidizer (Microfluidics, MA, USA; LM20) to disperse the cellulose fibre bundles into individual nanofibres. The dispersion was passed 3 times through a 200 μm grid chamber and 1 time through a 100 μm grid chamber under a pressure of 1800 bar.

[0099] The filters were prepared as previously described (Manukyan et al., J Mem Sci, Vol. 572, 2019, 464-474). The diluted dispersion was drained through a medium (Durapore; 0.65 μm DVPP; Merck Millipore, MA, USA) using a vacuum filtration setup (Advantec, Japan) until a cellulose cake was formed on top of the medium. The wet cake was then removed and dried at the desired temperature and time depending on the type of filter using a hot press (Carver, IN, USA; 4122CE). For the preparation of 11 μm thick pre-filters, 50 ml of 1 mg/ml nanocellulose suspension was used, and the nanocellulose filtrate cake was dried at 140° C. using a hot-press for 40 min. For the preparation of 22 μm thick virus removal filter, 100 ml of 1 mg/mL of nanocellulose suspension was used, and the nanocellulose filtrate cake was dried at 80° C. using a hot-press for 24 h. The dry filters were removed, cut into 47 mm diameter discs.

[0100] Filtration Setup

[0101] An Advantec KST-47 (Japan) filter holder was used. A general purpose filter paper disc (47 mm in diameter, Munktell) was placed beneath the nanocellulose filter as support. The rate of flow was monitored gravimetrically by collecting the outflowing liquid on an analytic balance (Mettler Toledo, Switzerland), connected to LabX software (Version 2.5, Mettler Toledo, Switzerland) at 20 second intervals.

[0102] Pre-Filtration

[0103] 11 μm filters were used in a first filtration step. The removal was validated on single sheet filters. Feed solution was 10 g/L HSA diluted in PBS and adjusted to pH 7.4. The filters were wetted with PBS prior to filtration. Filtrations through the 11 μm filters were carried out at 1 bar. Due to rapid fouling, for each filtration through the 11 μm filters around 25 mL was passed through each 11 μm filter corresponding to 14-15 L/m.sup.2 load volume (FIG. 1). The solution was then passed through a 22 μm filter and little fouling was observed. The permeate fractions were collected, mixed together, and stored at 4° C. before usage.

[0104] The same procedure was carried out for a feed solution of 50 g/L HAS. Around 10 ml was passed through each 11 μm filter corresponding to about 7 L/m.sup.2 load volume. The solution was then passed through a 22 μm filter and little fouling was observed (FIG. 6).

[0105] Filter Flux and Fouling Behaviour

[0106] FIGS. 1, 2, and 6 show the flux and fouling behaviour of different filters at different pressures or different protein concentrations. FIG. 1 shows that the 11 μm filter rapidly fouls when HSA 10 g/L solution is passed through it at 1 bar. Second filtration at 1 bar does not result in filter fouling and the flux is steady. FIG. 2 shows that the 22 μm filter does not foul even when it is operated at 3 bar, using an 11 μm pre-filtered solution. No fouling was observed at both pressures. FIG. 6 shows that when the concentration of HSA is increased to 50 g/L the fouling is even more rapid in the 11 μm filter. However, second filtration of the same solution through the 22 μm filter does not result in filter fouling, and the flux is stable throughout the experiment. The protein recovery after second filtration was high, indicating that only minor protein loss during two-step filtration.

[0107] For a 1% solution of HSA, 87% of product was obtained after filtering through the 11 μm thick filter and 85% of product was recovered through the following step of filtration through the 22 μm. For a 5% solution of HSA the recoveries were 94 and 93% respectively.

[0108] Monitoring of Particle Size Distribution and Removal of Aggregates

[0109] Dynamic Light Scattering

[0110] Dynamic light scattering (DLS) was used to assess particle size distribution of 10 g/L HSA solution in PBS pH 7.4 with Nano ZS instrument (Malvern, UK), see FIG. 3.

[0111] SEC-ÄKTA chromatography Protein Purification of HSA-PBS solution using Size Exclusion Chromatography (ÄKTA START) instrument. Selected chromatographic column (Mw 40-20,000 kDa; HiPrep 26/60 Sephacryl S-500HR, GE, Uppsala, Sweden) was used with a flow velocity of 1 ml/min. The column was equilibrated with 0.5 cV of PBS buffer and then 3.2 ml 1 wt. %. The purified solution was collected using peak fractionating in 10 ml falcon tubes and the fractions were collected when absorbance was ≤5 mAU, see FIG. 4.

[0112] To explain the observed enhanced throughput capacity in the second filtration step as compared to the first one, DLS and ÄKTA chromatography were performed. FIG. 3 shows the results of intensity particle size distributions for samples presented in FIG. 1. It is seen that the second peak in the feed solution, that is stretching in the region between 30 and 200 nm, is not detectable after filtration through 11 μm and 22 μm filters.

[0113] Size exclusion gel chromatography with ÄKTA confirmed the removal of trace quantities of large molecular weight fractions from the feed solution following filtration through 11 μm filter.

[0114] Virus Removal Filtration

[0115] To further illustrate the invention and to confirm that 22 μm filter retains high virus removal capacity after prefiltration, PFU tests were performed. FIG. 5 shows the results of ΦX174 bacteriophage (28 nm) from 10 g/L HSA solution.

[0116] 22 μm and 11 μm filters were used for virus removal studies using single sheet filters. Prior to virus removal filtration, the virus-spiked feed solution was pre-filtered through a 0.2 μm filter (VWR). A highly purified ΦX174 stock solution was spiked at 0.1% into the pre-filtrated 10 mg/ml HSA in PBS, adjusted to specific pH. Virus stability was controlled by a hold sample taken from the pre-filtered spiked feed solution. The filters were wetted with PBS prior to filtration.

[0117] The PFU assay was used as previously described (Manukyan et al., J Mem Sci, Vol. 572, 2019, 464-474). Escherichia coli bacteriophage ΦX174 (ATCC 13706-B1 ™) Escherichia coli bacteriophage PR772 (ATCC® BAA769B1 ™) and the host bacteria Escherichia coli (Migula) Castellani and Chalmers (E. coli) (ATCC 13706) were obtained from ATCC (Manassas, Va., USA). The titer of bacteriophages was determined by plaque forming units (PFU) assay. The feed and permeate samples were serially diluted in Luria-Bertani medium (LBM) (1% tryptone, 0.5% yeast extract, and 1% NaCl in deionized water), and 100 μl of diluted bacteriophage was mixed with 200 μl of E. coli stock. The resulting suspension was mixed with 1 ml of melted soft agar and poured on the surface of prepared hard agar plate (55×15 mm) and incubated at 37° C. for 5 hours. The feed titer was adjusted to about 10.sup.5 to 10.sup.6 bacteriophages ml.sup.−1. The limit of detection, i.e. 0.7 PFU ml.sup.−1, of the current experimental design refers to 5 bacteriophages ml.sup.−1, corresponding to a single detectable plaque in one of the plates for non-diluted duplicate samples, assuming that at the detection limit each plaque is produced by one bacteriophage. The virus retention was expressed as logo reduction value (LRV).

Example 2

[0118] Nominal Pore Size of Filters

[0119] ΦX174 bacteriophage (28 nm) and PR772 phage (70 nm) were used as a model monodispersed probes and virus removal quantification using 11 and 33 μm filters. The tests were carried at 1 bar pressure using 50 mL of PBS corresponding to 26 L/m.sup.2 load volume. Permeate samples and hold samples were collected and stored at 4° C. before PFU assay. FIG. 5 shows the LRV value obtained after filtering through the second (33 μm) filter.

[0120] Plaque Forming Units (PFU) and Log10 Reduction Value (LRV) The PFU assay was used as previously described (Manukyan et al., J Mem Sci, Vol. 572, 2019, 464-474). The titer of ΦX174 bacteriophage was determined by plaque forming units (PFU) assay. The feed and permeate samples were serially diluted in Luria-Bertani medium (LBM) (1% tryptone, 0.5% yeast extract, and 1% NaCl in deionized water), and 100 μl of diluted bacteriophage was mixed with 200 μl of E. coli stock. The resulting suspension was mixed with 1 ml of melted soft agar and poured on the surface of prepared hard agar plate (55×15 mm) and incubated at 37° C. for 5 hours. The feed titer was adjusted to about 10.sup.5 to 10.sup.6 bacteriophages ml.sup.−1. The limit of detection, i.e. 0.7 PFU ml.sup.−1, of the current experimental design refers to 5 bacteriophages ml.sup.−1, corresponding to a single detectable plaque in one of the plates for non-diluted duplicate samples, assuming that at the detection limit each plaque is produced by one bacteriophage. The virus retention was expressed as logo reduction value (LRV).