Clarification of Mammalian Cell Culture Feed Streams Using Depth Filters and Flocculants: A Case Study
0 likes242 views
MilliporeSigma has developed synthetic depth filtration media to enhance midstream clarification efficiency by reducing soluble impurities like host cell proteins and DNA. Tests indicated that these filters maintain human monoclonal antibody recovery while improving impurity clearance, particularly with the millistak+® hc pro x0sp filter. The study suggests that various combined filtration strategies can effectively clarify mammalian cell culture feed streams, benefitting large-scale manufacturing processes.
Clarification of Mammalian Cell Culture Feed Streams Using Depth Filters and Flocculants: A Case Study
- 1. MilliporeSigma.com 0 20 40 60 80 100 120 Pre-Harvest Supernatant pDADMAC 0.05% >40MS Filtrate HCPReductionandHuMAb Recovery(%) HCP Reduction HuMAb Recovery Ramtin Rahbar*,1, Catherine Martel1, Elina Gousseinov2, Jill Lundy2, Pooja Desai1, Ross Martel2 and David Bell1,¶ 1Therapure Biopharma, Inc., 2585 Meadowpine Blvd., Mississauga, Ontario L5N 8H9, Canada and 2MilliporeSigma, 80 Ashby Road, Bedford, MA 01730, USA Introduction MilliporeSigma has developed an all-synthetic depth filtration media (Millistak+® HC Pro D0SP and X0SP) intended to improve the process consistency and efficiency of midstream clarification steps by reducing soluble process impurities such as host cell proteins (HCP) and host cell DNA (HCDNA). These media have been tested in this study and were evaluated against benchmark Millistak+® HC D0HC and X0HC filters. Using a Chinese hamster ovary (CHO)- derived human monoclonal antibody (HuMAbs) culture at VCD: ~1.4 x 107 cells/mL, Viability: ~95% and HuMAb concentration: ~0.9 g/L (Figure 1), both Millistak+® HC Pro D0SP and Millistak+® HC D0HC filters showed no effect on HuMAb concentration; i.e. 100% HuMAb recovery, with a higher HCP clearance when Millistak+® D0HC filter was used (16% for D0HC vs 4% for D0SP, Figure 2). Both filters showed a minimal effect on HCDNA clearance (HCDNA log reduction: -0.3 for D0HC and -0.5 for D0SP, Figure 2). Table 2 shows filter sizing for these filters when used to filter a 500 L batch. The Millistak+® HC Pro X0SP filter increased HCP reduction compared to Millistak+® HC X0HC filter (~91% compared to ~63% respectively), at the cost of HuMAb recovery (Figure 3). HuMAb recovery post Millistak+® HC X0HC was ~94% while the recovery post Millistak+® HC Pro X0SP filter was ~70%. The data suggests that Millistak+® HC Pro X0SP filters are suitable for processes challenged by hard-to- remove HCP impurities during downstream processing. Both Millistak+® HC X0HC and Millistak+® HC Pro X0SP filters were highly efficient in removing HCDNA to ~5 logs (Figure 3B). Table 3 shows filter sizing for these filters when used to filter a 500 L batch. Flocculation is a simple method to pre-clarify high-density CHO cell culture feed streams, allowing for an efficient clarification process that would otherwise not be feasible using traditional depth filters. Here two flocculants were utilized in a dose response study; 1) poly- diallyldimethylammonium chloride (pDADMAC), and 2) Clarisolve® mPAA, modified poly(allyl amine), a polycationic stimulus-responsive flocculation polymer. The dose of each flocculant that resulted in lowest turbidity post centrifugation was selected as the final treating dose (Figure 4A and 5A). Clarisolve® filters, which are designed for higher dirt holding capacity, were used as primary filters, followed by Millistak+® HC X0HC or Millistak+® HC Pro X0SP filters as secondary filters. Harvest material treated with 0.05% pDADMAC polymer and filtered with Clarisolve® 40MS filter showed similar HCP reduction (~16%) to Millistak+® HC D0HC filtrate with minimal (~1%) effect on HuMAb recovery (Figure 4B). The combination, however, showed an efficient HCDNA (~3 logs) clearance (Figure 4C). Harvest material treated with 0.2% Clarisolve® mPAA polymer and filtered with Clarisolve® 60HX filter, on the other hand, showed ~55% reduction of HCP, ~95% HuMAb recovery and ~3 logs HCDNA reduction (Figure 5A and B). The data are intriguing as they are comparable to the results when D0HC > X0HC filters were used (Figure 3A). This means lower numbers of filters (5 vs 13) can be used with similar efficiency, a benefit in large scale manufacturing. When combined with Millistak+® HC X0HC or Millistak+® HC Pro X0SP filters, Millistak+® HC Pro X0SP filter was more efficient in removing HCP (for pDADMAC treated feed: 64% vs 28% reduction and for mPAA treated feed: 92% vs 71% reduction, Figure 6), however, HuMAb recovery showed a lower impact for Millistak+® HC X0HC filters (for pDADMAC treated feed: 84% vs 71% recovery and for mPAA treated feed: 79% vs 68% recovery Figure 6A). Both filters were highly efficient in removing HCDNA to ~4.5-5 log fold (Figure 6B). Tables 4-6 show filter sizing for different combinations of flocculants and secondary filters when used to filter a 500 L batch. Methods A HyPerforma SUB 50 bioreactor was inoculated with CHO- expressing HuMAb at seeding density of ~0.4 x 106 cells/mL and viability of 99% (Figure 1A). The culture was harvested at day 9 post-inoculation with a VCD: ~1.4 x 107 cells/mL and viability of ~95% (Figure 1B and Table 1). Figure 1: A) Viable cell density (VCD), viability (V), and doubling time (Td). B) HuMAb culture had a turbidity ~2817 NTU at day 9 post inoculation; i.e. harvest day. Table 1: Harvest material turbidity, HuMAb, HCP and HCDNA concentration on harvest day. The harvest was either left untreated or treated with different doses of pDADMAC or Clarisolve® mPAA polymers, then filtered using different primary and secondary filters (See Process Flow Diagram (PFD), then analyzed for HuMAb concentration, HCP and HCDNA (See analytical testing plan in PFD). Methods Cont’d Figure 2: Effect of D0HC or D0SP Primary Filters on HCP Reduction and HuMAb Recovery (A) and HCDNA Reduction (B). Table 2: Filter sizing for Millistak+® HC D0HC or Millistak+® HC Pro D0SP filters to filter direct harvest. Figure 3: Effect of Secondary Filters on HCP Reduction and HuMAb Recovery (A) and HCDNA Reduction (B). Table 3: Filter sizing for MilliporeSigma’s Millistak+® X0HC or Millistak+® X0SP filters to filter D0HC or D0SP filtrate respectively. Figure 4: pDADMAC Dose Response Study (A), Effect of Primary Filters ± 0.05% pDADMAC polymer on HCP Reduction and HuMAb Recovery (B) and HCDNA (C). Results Cont’d Table 4: Filter sizing for Clarisolve® 40MS filters to filter 0.05% pDADMAC Treated Harvest Material. Figure 5: Clarisolve® mPAA polymer Dose Response Study (A), Effect of Primary Filters ± 0.2% Clarisolve® mPAA polymer on HCP Reduction and HuMAb Recovery (B) and HCDNA (C). Table 5: Filter sizing for Clarisolve® 40MS filters to filter 0.2% mPAA Polymer Treated Harvest Material. Figure 6: Effect of Secondary Filters on HCP Reduction and HuMAb Recovery (A) and HCDNA (B) when 0.05% pDADMAC Treated Harvest> 40MS or 0.2% mPAA Treated Harvest > 60HX used as feed. Table 6: Filter sizing for Millistak+® X0HC or X0SP filters to filter 0.05% pDADMAC Treated Harvest > Clarisolve® 40MS filtrate. Summary Table 7 : Sizing and effect of Primary/Secondary Filters ± pDADMAC or Clarisolve® mPAA polymers on HCP Reduction, HuMAb Recovery and HCDNA. This case study shows promising results and provides alternative paths for CHO cell culture clarification, based on the nature of HCP and of the protein of interest. Clarification of mammalian cell culture feed streams using depth filters and flocculants; a case study © 2020 Merck KGaA, Darmstadt, Germany and/or its affiliates. All Rights Reserved. MilliporeSigma, Millipore, Clarisolve, and Millistak+ are trademarks of Merck KGaA, Darmstadt, Germany or its affiliates. All other trademarks are the property of their respective owners. Detailed information on trademarks is available via publicly accessible resources. Lit. No. MS_PS1149EN Ver. 1.0 0% 0.05% 0.1% 0.2% 0.3% 0.4% 0.8% 0% 0.025% 0.05% 0.075% 0.1% 0.2% pDADMAC mPAA Feed Stream Treatment Untreated 0.05% pDADMAC 0.2% mPAA Depth Filtration D0HC Clarisolve® 40MS Clarisolve® 60HX D0SP X0HC X0SP X0HC X0SP X0HC X0SP 1º 2º OR OR OR Analytical Testing Turbidity Pellet Size HuMAb Titre HCP HCDNA A B Test Sample Feed pH Harvest Turbidity (NTU) HuMAb (mg/mL) HCP (mg/mL) HCDNA (µg/mL) Direct harvest 6.8 2817 0.93 0.3 4.5 0 1 2 3 4 Pre-Harvest Supernatant D0HC D0SP logHCDNA(pg/uL) 0 20 40 60 80 100 120 Pre-Harvest Supernatant D0HC D0SP HCPReductionandHuMAbRecovery (%) HCP Reduction HuMAb Recovery A B Test Fluid Test Filter Test Filter Area (cm2) Test Endpoint Pressure (psi) Test Load (L/m2) Test Average Flux (L/m2/hr) Turbidity In (Feed) (NTU) Turbidity Out (Filtrate) (NTU) Batch* Filter Area Minimum (m2) Batch* Suggested Filter Configuration Suggested Configuration Filter Area (m2) Direct Harvest D0HC 23 23 109 126 2817 669 3.10 5 x 1.1m2 5.5 Direct Harvest D0SP 23 8 180 127 2817 13.2 1.83 4 x 0.77m2 3.08 * 500 L 0 20 40 60 80 100 120 HCPReductionandHuMAbRecovery (%) HCP Reduction HuMAb Recovery Pre-Harvest Supernatant D0HC >X0HC D0SP >X0SP A B Results -2 -1 0 1 2 3 4 Pre-Harvest Supernatant D0HC >X0HC D0SP >X0SP logHCDNA(pg/uL) * 500 L Test Fluid Pre- filter >Test Filter Test Filter Area (cm2) Test Endpoint Pressure (psi) Test Load (L/m2) Test Average Flux (L/m2/hr) Turbidity In (Feed) (NTU) Turbidity Out (Filtrate) (NTU) Batch* Filter Area Minimum (m2) Batch* Suggested Filter Configuration Suggested Configuration Filter Area (m2) Direct harvest D0HC >X0HC 5 23.5 45 223 426.0 1.32 5.96 8 x 1.1m2 8.8 Direct harvest D0SP >X0SP 5 12.5 400 236 19.0 2.52 0.73 2 x 1.1m2 2.2 101 7.44 4.86 8.49 17.20 121.00 1 10 100 1000 0 0.025 0.05 0.075 0.1 0.2 SupernatantLog Turbidity(NTU) pDADMAC % Concentration in the Sample 0 1 2 3 4 Pre-Harvest Supernatant pDADMAC 0.05% >40MS logHCDNA(pg/uL) A B C Test Fluid Pre-filter >Test Filter Test Filter Area (cm2) Test Endpoint Pressure (psi) Test Load (L/m2) Test Average Flux (L/m2/hr) Turbidity In (Feed) (NTU) Turbidity Out (Filtrate) (NTU) Batch* Filter Area Minimum (m2) Batch* Suggested Filter Configuration Suggested Configuration Filter Area (m2) 0.05% pDADMAC treated Harvest Clarisolve® 40MS 23 16 221 156 2355 5.6 2.14 5 x 1.1m2 2.75 * Presenter ¶ Corresponding Author; dbell@therapurebio.com 101 28.3 15.0 14.2 14.2 78.2 49.0 1 10 100 1000 0 0.05 0.1 0.2 0.3 0.4 0.8 SupernatantLog Turbidity(NTU) mPAA % Concentration in the Sample 0 1 2 3 4 Pre-Harvest Supernatant mPAA 0.2% >60HX logHCDNA(pg/uL) 0 20 40 60 80 100 120 Pre-Harvest Supernatant mPAA 0.2%>60HX Filtrate HCPReductionandHuMAb Recovery(%) HCP Reduction HuMAb Recovery A B C Test Fluid Pre-filter >Test Filter Test Filter Area (cm2) Test Endpoint Pressure (psi) Test Load (L/m2) Test Average Flux (L/m2/hr) Turbidity In (Feed) (NTU) Turbidity Out (Filtrate) (NTU) Batch* Filter Area Minimum (m2) Batch* Suggested Filter Configuration Suggested Configuration Filter Area (m2) 0.2% mPAA treated Harvest Clarisolve® 60HX 23 25 204 168 6390 3.3 2.19 5 x 1.1m2 2.75 0 20 40 60 80 100 120 HCPReductionandHuMAb Recovery(%) HCP Reduction HuMAb Recovery -2 -1 0 1 2 3 4 logHCDNA(pg/uL) 1. Pre-Harvest Supernatant, 2. pDADMAC polymer 0.05% > Clarisolve® 40MS filter >Millistak+® X0HC filter, 3. pDADMAC polymer 0.05% > Clarisolve® 40MS filter > Millistak+® X0SP filter 4. mPAA polymer 0.2% > Clarisolve® 60HX filter > Millistak+® X0HC filter, 5. mPAA polymer 0.2% > Clarisolve® 60HX filter> Millistak+® X0SP filter 1 2 3 4 5 1 2 3 4 5 Test Fluid Pre-filter >Test Filter Test Filter Area (cm2) Test Endpoint Pressure (psi) Test Load (L/m2) Test Average Flux (L/m2/hr) Turbidity In (Feed) (NTU) Turbidity Out (Filtrate) (NTU) Batch* Filter Area Minimum (m2) Batch* Suggested Filter Configuration Suggested Configuration Filter Area (m2) 0.05% pDADMAC treated harvest Clarisolve® 40MS >Millistak+® X0HC 5 7.8 400 238 15.4 1.31 0.76 1 x 1.1m2 1.1 0.05% pDADMAC treated harvest Clarisolve® 40MS >Millistak+® X0SP 5 6.8 400 240 15.4 1.24 0.43 1 x 1.1m2 1.1 0.2% mPAA treated Harvest Clarisolve® 60HX >Millistak+® X0HC 5 7.2 400 238 3.5 0.84 0.72 1 x 1.1m2 1.1 0.2% mPAA treated Harvest Clarisolve® 60HX >Millistak+® X0SP 5 5.7 400 240 3.5 0.87 0.88 1 x 1.1m2 1.1 A B Filter Train HCP Reduction % HuMAb Recovery % HCDNA Log Fold Reduction Batch* Suggested Filter Configuration Suggested Configuration Filter Area (m2) D0HC 19 100 -0.34 5 x 1.1m2 5.5 D0HC >X0HC 63 94 -4.71 8 x 1.1m2 8.8 D0SP 4 100 -0.49 4 x 0.77m2 3.08 D0SP >X0SP 91 70 -5.15 2 x 1.1m2 2.2 pDADMAC 0.05% >40MS 16 99 -3.49 5 x 1.1m2 2.75 pDADMAC 0.05% >40MS >X0HC 28 84 -4.87 1 x 1.1m2 1.1 pDADMAC 0.05% >40MS >X0SP 64 72 -4.49 1 x 1.1m2 1.1 mPAA 0.2% >60HX 55 95 -3.23 5 x 1.1m2 2.75 mPAA 0.2% >60HX >X0HC 71 79 -4.41 1 x 1.1m2 1.1 mPAA 0.2% >60HX >X0SP 92 68 -4.03 1 x 1.1m2 1.1 * 500 L * 500 L * 500 L * 500 L The life science business of Merck KGaA, Darmstadt, Germany operates as MilliporeSigma in the U.S. and Canada.