DEVICES AND METHODS FOR IMPROVING AND EVALUATING STABILITY OF PUMPED PROTEIN SOLUTIONS IN BIOPROCESSING SYSTEMS

Abstract

The present invention relates to devices and methods for improving and evaluating stability of pumped protein solutions in cross-flow filtration applications. Inter alia, the present invention provides a peristaltic pump for cross-flow filtration having a pump head, wherein the pump head comprises a stepped occlusion plate and at least one pump roller, wherein a tubing is to be arranged between the stepped occlusion plate and the at least one pump roller, wherein the stepped occlusion plate has a specific configuration.

Claims

1.-9. (canceled)

10. A method for evaluating the relative robustness of different proteins in solution, the method comprising the steps of: pumping a protein solution around a cross-flow filtration loop at various flow conditions prior to a filter being fitted to the cross-flow filtration loop, monitoring the rate of protein damage, then, after a given time, or a given pumped volume, or a given level of protein damage, introducing a filter to the loop, subsequently monitoring the level of protein damaged, and assessing whether the filter is removing protein aggregates, has no effect or contributes to protein damage, wherein pumping the protein solution is carried out with a peristaltic pump suitable for cross-flow filtration in protein treatment having a pump head, wherein the pump head comprises a stepped occlusion plate and at least one pump roller, wherein a tubing is to be arranged between the stepped occlusion plate and the at least one pump roller, wherein the stepped occlusion plate comprises: an initial tubing compression lead in section, a constant compression lead in section, a lead in-to-pumping step, a constant full tube compression pumping section, a pumping-to-lead out step, a constant compression lead out section, and a final tubing decompression lead out section, wherein the initial tubing compression lead in section is adapted to provide an initial rapid tubing compression, wherein the constant compression lead in section and the constant compression lead out section are adapted to bring the internal tubing surfaces close to each other but not such that they are touching, wherein the lead in-to-pumping step and the pumping-to-lead out step are adapted to fully pinch the tubing closed by the at least one roller, and wherein the final tubing decompression is adapted to provide a rapid full decompression of the tubing.

11. The method according to claim 10, wherein the peristaltic pump is a sprung occlusion plate peristaltic pump, wherein the pressure of the sprung occlusion plate on the tubing can be adjusted.

12. The method according to claim 10, wherein the peristaltic pump is a sprung roller peristaltic pump, wherein the pressure of the at least one roller on the tubing can be adjusted.

13. The method according to claim 10, wherein a silicon/PTFE composite tubing is arranged between the stepped occlusion plate and the at least one pump roller.

14. A system for determining the relative robustness of different proteins in a solution, the system comprising a reservoir for the solution, the peristaltic pump, a tubing loop of variable configuration and an in-line means of monitoring protein aggregation, wherein the peristaltic pump is a peristaltic pump suitable for cross-flow filtration in protein treatment having a pump head, wherein the pump head comprises a stepped occlusion plate and at least one pump roller, wherein a tubing is to be arranged between the stepped occlusion plate and the at least one pump roller, wherein the stepped occlusion plate comprises: an initial tubing compression lead in section, a constant compression lead in section, a lead in-to-pumping step, a constant full tube compression pumping section, a pumping-to-lead out step, a constant compression lead out section, and a final tubing decompression lead out section, wherein the initial tubing compression lead in section is adapted to provide an initial rapid tubing compression, wherein the constant compression lead in section and the constant compression lead out section are adapted to bring the internal tubing surfaces close to each other but not such that they are touching, wherein the lead in-to-pumping step and the pumping-to-lead out step are adapted to fully pinch the tubing closed by the at least one roller, and wherein the final tubing decompression lead out section is adapted to provide a rapid full decompression of the tubing.

15. The system according to claim 14, wherein the peristaltic pump is a sprung occlusion plate peristaltic pump, wherein the pressure of the sprung occlusion plate on the tubing can be adjusted.

16. The system according to claim 14, wherein the peristaltic pump is a sprung roller peristaltic pump, wherein the pressure of the at least one roller on the tubing can be adjusted.

17. The system according to claim 14, wherein a silicone/PTFE composite tubing is arranged between the stepped occlusion plate and the at least one pump roller.

18. The system according to claim 14, wherein a filter is fitted to the tubing loop.

Description

[0068] FIG. 1 shows a pump head according to an embodiment of the present invention.

[0069] In the following, an embodiment of the pump head (1) of the peristaltic pump as claimed in the present invention is described with reference to FIG. 1.

[0070] The pump head (1) as shown in FIG. 1 comprises a stepped occlusion plate (2) and four pump rollers (3). A tubing is to be arranged between the stepped occlusion plate (2) and the at least one pump roller (3). The stepped occlusion plate (2) has an initial tubing compression lead in section (4a), a constant compression lead in section (4b), a lead in-to-pumping step (4c), a constant full tube compression pumping section (4d), a pumping-to-lead out step (4e), a constant compression lead out section (4f), and a final tubing decompression lead out section (4g). The initial tubing compression lead in section (4a) is adapted to provide an initial rapid tubing compression. The constant compression lead in section (4b) and the constant compression lead out section (4f) are adapted to bring the internal tubing surfaces close to each other but not such that they are touching. The lead in-to-pumping step (4c) and the pumping-to-lead out step (4e) are adapted to fully pinch the tubing closed by the at least one roller (3). The final tubing decompression lead out section (4g) is adapted to provide a rapid full decompression of the tubing.

[0071] Summarizing, the present invention as claimed reduces the damage caused to proteins in a peristaltic pump and more generally is expected to reduce damage to sensitive fluids or suspensions. Furthermore, the present invention as claimed reduces the amount of pulsation from a peristaltic pump, wherein pulsation is often considered to be a disadvantage for many processes, reduces particular release, prolongs tubing life, provides a high pump rate stability, protects downstream system elements from excessive pressure, and allows the adjustment of small-scale-cross-flow filtration based systems to mimic larger scale systems in respect to protein damage they cause. Finally, the present invention as claimed provides a system to assess the relative robustness of proteins and the protective effect of carrier fluid composition in a directly relevant model.

[0072] The following specific example is provided for further illustrating the present invention and does not limit the scope of the present invention.

EXAMPLE

[0073] Protein Aggregation Observation

[0074] A significant amount of protein aggregation was observed in a prototype system (Sartorius Stedim Biotech SA) when compared to a standard bench top cross-flow filtration system. Various components of the prototype system were decoupled to test for high shear zones. This test identified an aspect of the feed loop which was responsible for the significant protein aggregation effect.

DETAILED INVESTIGATION

[0075] Materials and Methods

[0076] Test Methodology

[0077] The common test method involved exposing 7 mL protein solution to a certain test condition for a period of time. Samples were then borrowed at regular time intervals for turbidity analysis (indicating insoluble protein aggregation) before returning to the test system. Unless otherwise stated, the common test method conditions were: [0078] Pump speed: 7 mL/min [0079] Transmembrane Pressure (TMP): 1500 mbar [0080] Retentate pressure (Pr) : 2000 mbar [0081] Tubing: Pump head: Bioprene, Flow path: Silicone, 1.6 mm i.d.

[0082] Protein Solutions

[0083] Molecule 1: mAb

[0084] Molecule 2: mAb

[0085] Molecule 3: mAb-dAb

[0086] Protein Aggregation Measurement

[0087] Samples were taken at regular time intervals for turbidity analysis using a spectrophotometer measuring at a wavelength of 600 nm.

[0088] Hardware Setup

[0089] Prototype (Sartorius Stedim Biotech SA)

[0090] Two different peristaltic pump designs were tested: a standard Watson Marlow pump (114DV OEM-pumphead), and a custom, in-house design pump according to the present invention. Backpressure or TMP was generated by an automated proportional valve.

[0091] Bench Scale System

[0092] A standard cross-flow filtration system configuration was used to evaluate relative performance of the Prototype system. A SciLog peristaltic pump was used in combination with a manually controlled pinch valve.

[0093] Results

[0094] Impeller

[0095] Confirmatory test that impeller is not the source of protein aggregation. Minimal change in the observed turbidity of the protein solution with time.

TABLE-US-00001 TABLE 1 10 mL molecule 2 protein solution exposed to 30% impeller power. Molecule 2 Time OD.sub.600 20 mins 0.016 20 mins 0.043

[0096] Benchmark Testing

[0097] Standard configurations were tested in both cross-flow filtration scale systems to benchmark performance (Table 2). Prototype system scale benchmark test reveals a higher rate of protein aggregation when compared to the standard bench scale configuration. At both scales, application of back pressure alleviates the protein aggregation effect. Halving the cross-flow rate also appears to halve the rate at which protein aggregates (Table 3). These factors indicate a direct correlation between mechanical action of the peristaltic pump and the protein aggregation event.

TABLE-US-00002 TABLE 2 Molecule 2 protein solution exposed to the different cross-flow filtration scale systems. Volumes and cross-flow rates are scaled linearly in order to maintain pump action:protein molecule ratio. Molecule 2 System Prototype system Benchscale 7 mL protein 35 mL protein 7 mL/min cross-flow 35 mL/min cross-flow rate Bioprene in rate Norprene in the pump head the pump head Test No 2 bar No 2 bar Condition backpressure backpressure backpressure backpressure 15 mins 0.244 0.198 0.053 0.052 30 mins 0.444 0.411 0.109 0.094 45 mins 0.646 0.596 0.161 0.136 60 mins 0.832 0.793 0.211 0.174

TABLE-US-00003 TABLE 3 Molecule 2 protein solution exposed to different cross-flow rates in the prototype system. Molecule 2 7 mL protein Bioprene in the pump head No backpressure Test 7 mL/min cross- 3.5 mL/min cross- Condition flow rate flowrate 15 mins 0.244 30 mins 0.444 45 mins 0.646 0.351 60 mins 0.832 0.461 90 mins 0.652

[0098] Effect of Pump Design

[0099] An in-house, custom designed peristaltic pump according to the present invention was investigated to determine if the protein aggregation effect could be influenced by redesign of the pump. Table 4 shows that the redesigned pump effectively reduces the rate of protein aggregation. Decreasing the force applied at each roller pass is also shown to reduce protein aggregation (Table 5).

TABLE-US-00004 TABLE 4 Molecule 2 protein solution exposed to two different peristaltic pump designs. Molecule 2 Test Watson In-house designed, Condition Marlow pump custom pump 15 mins 0.244 0.152 30 mins 0.444 0.327 45 mins 0.646 0.493 60 mins 0.832 0.652

TABLE-US-00005 TABLE 5 Effect of pressure investigated by reducing the force applied by the back plate onto the peristaltic pump. Molecule 2 In- house designed, custom pump Test Pump roller Pump roller Condition force 5 bar force 2.5 bar 15 mins 0.042 0.05 30 mins 0.087 0.078 45 mins 0.134 0.107 60 mins 0.174 0.135

[0100] Effect of Tubing Material

[0101] Different tubing materials were tested to determine their effect on protein aggregation (Table 6). In the first instance, a buffer solution (no protein) was tested to rule out the fact that the turbidity measurements were as a result of particle release from the tubing. Comparing 6 (iv) with 6 (ii), there is a clear positive effect on the rate of protein aggregation with tubing material in the flow path. Silicone tubing in the pump head drastically reduces the increase in turbidity with time for the protein sample. Similar levels of protein aggregation occur when comparing either 6 (v) with 6 (iv) or 6 (ii) with 6 (iii), illustrating that the protein aggregation effect is as a direct result of the bioprene tubing in the pump head itself, rather than any exposure to this material when not in a high mechanical force environment.

TABLE-US-00006 TABLE 6 Effect of different tubing materials on the protein aggregation. Molecule N/A 2 In- house designed, custom pump No back pressure ii i Standard setup iii iv V Test Buffer Pump head: Bioprene Pump head: Bioprene Pump head: Silicone Pump head: Silicone Condition only Flow path: Silicone Flow path: Bioprene Flow path: Silicone Flow path: Bioprene 15 mins 0.001 0.152 0.157 0.048 0.04 30 mins 0.006 0.327 0.3 0.095 0.081 45 mins 0.012 0.493 0.145 0.122 60 mins 0.02 0.652 0.175

[0102] Confirmation in Multiple Molecules

[0103] This protein aggregation effect was subsequently confirmed in two different molecules (Table 7). Results here confirm that the combination of an in-house designed peristaltic pump according to the present invention with a silicone based tubing drastically reduce the protein aggregation effect.

TABLE-US-00007 TABLE 7 Two different molecules were tested to ensure that protein aggregation effects that had been demonstrated were not molecule specific. Molecule 1 3 In-house In-house Watson designed, Watson designed, Marlow custom pump Marlow custom pump Test Pump head: Pump head: Pump head: Pump head: Condition Bioprene Silicone Bioprene Silicone 15 mins 0.112 0.02 0.16 0.077 30 mins 0.286 0.06 0.348 0.121 45 mins 0.474 0.030 0.545 0.163 60 mins 0.645 0.042 0.717 0.207

[0104] Product Solution

[0105] Silicone tubing has a short mechanical life so is therefore unsuitable for extended periods of time in a peristaltic pump. STA-Pure (platinum-cured silicone rubber, reinforced with expanded polytetrafluoroethylene (ePTFE)) was investigated as a suitable replacement (Table 8). For both molecules investigated, it appears that the positive effect that the silicone tubing chemical composition has on protein aggregation is amplified by the mechanical effect of the reinforced tube structure.

TABLE-US-00008 TABLE 8 Effect of two different silicone tubes with different mechanical properties investigated with two different molecules. Molecule 2 3 In- house designed, custom pump Pump Pump Pump Pump Test head: head: head: head: Condition TuFlux STA-Pure TuFlux STA-Pure 15 mins 0.055 0.015 0.103 0.063 30 mins 0.089 0.021 0.149 0.064 45 mins 0.122 0.026 0.194 0.069 60 mins 0.156 0.03 0.239 0.076