Case study 3 Pilot-scale clarification of yeast homogenate by continuous in-line PEI flocculation.

In a refinement to the pilot-scale batch PEI flocculation procedure described earlier (section 5.2.3) a continuous flocculation procedure was developed (Fig.5.3) and is

Cell suspension

Fig.5.3

Homogenisation-K3 Continuous "in-line" PEI Flocculation _{Ratio PEI: Homogenate = 0.3:0.7} 1%(w/v) PEI

100% Sat(NH4)2S04 0.02M KH2PO4, p H7

CSA-1 Disk stack Clarification

Packed bed chromatography Schematic layout of the continuous PEI flocculation process for the clarification of

S.cerevisiae cell debris following high pressure homogenisation. Mixing of the homogenate and PEI solutions was performed in a T-piece upstream of the CSA-1 centrifuge.

On

described below. A small-scale chromatography cycle was performed in which clarified material was loaded to 100% breakthrough o f ADH. Thus the total binding capacity o f the hydrophobic matrix for this clarified material could be evaluated. A slightly larger chromatography cycle was also performed in which loading was stopped at a ADH breakthrough level o f 5% to realistically simulate industrial recovery processes operated with high yield. The chromatography runs are described towards the end o f this case study.

5.2.4.1 Preparation of homogenate

Baker’s yeast (280g/L(wet weight), 60L) in buffer (O.IM KH2PO4, pH6.5) was disrupted in a pilot-scale high-pressure homogeniser (Model K3) for 5 discrete passes at 500 bar(g) at a throughput of 280L/h and temperatures maintained below 5°C. 5.2.4.2 Determination of ratio of PEI to homogenate.

A volume o f homogenate was added to an appropriate volume of PEI (l% (w/v) pH6.5) in a l.SmL eppendorf according to Table 5.2. The eppendorfs were mixed by inversion and spun at 12,400g for 15 minutes using a bench top centrifijge (Model M il). The optical density of the supernatant at 650nm was measured using a spectrophotometer (Model 922, Kontron Instruments Ltd., Watford, UK). Assays were performed in duplicate with a reproducibility o f ±5%.

PEI Ratio %tw/vl Homogenate (mL) PEI (mL) 0 0 1.5 0.1 0.15 1.35 0.2 0.3 1.2 0.3 0.45 1.05 0.4 0.6 0.9 0.5 0.75 0.75 0.6 0.9 0.6

Table 5.2. Volume o f 1% (w/v) PEI pH 6.5 necessary fo r addition to yeast homogenate in determining correct mix ratio fo r flocculation.

5.2 4.3 Continuous “in-line” PEI flocculation and centrifugal clarification. Two tanks, one filled with homogenate, the other with PEI (l%(w/v), pH6.5) were connected to two lobe pumps (6mm port diameters, pumping capacity =2.05L/100

rev.; 25NDM pump heads, SSP Pumps Eastbourne, UK) and fed continuously to a mixing T-piece where flocculation of the cell debris occurred.

The appropriate mix ratio of PEI to homogenate was set by differing the individual flowrate on either lobe pump. Check valves between the lobe pumps and the tanks prevented the back flow of liquid from the T-piece. For a final flowrate o f 30L/h and a mix ratio o f PEI to homogenate of 0.3%(w/v) the PEI pump was set to 9.0 L/h while the homogenate pump was set to 21 L/h

The resultant mixture was pumped into the hydro-hermetic feed zone CSA-1 centrifuge for solid-liquid separation. The quality o f the supernatant in the outlet from the CSA-1 centrifuge was continuously monitored by visual inspection through a sight glass in the supernatant outlet pipe. Full discharges o f the bowl were carried out every 150s with flow to the machine stopped.

To ascertain whether any further clarification could be achieved by passing the CSA-1 supernatant through the IP tubular bowl, a rapid laboratory mimic o f the clarification one might expect from the IP was performed on two samples o f the CSA-1 supernatant pool. This small scale mimic was performed as follows:

Two samples of the CSA-1 supernatant pool were spun in a bench top centrifuge (Model J2-M1, Spinco, Beckman Instruments, Pan Alto, California, USA) with a fixed angle rotor (type JA-20.1) at 19,500rpm (49,000g) for 600s. The contents of one of the tubes was adjusted to 0.78M ammonium sulphate prior to centrifugation. Visual inspection was employed to compare the relative amount of sedimented solids in the centrifuge tubes after clarification with a sample o f the unspun CSA-1 supernatant.

Ammonium sulphate (100% saturated solution in O.IM KH2PO4, pH6.5) was added to the CSA-1 supernatant pool to adjust the salt concentration to 0.78M. ADH and protein assays were performed in duplicate at each stage o f the process for the completion o f a mass balance.

5.2.4.4 Total binding capacity of Phenyl Sepharose FF (low sub) for ADH. An X K l6/20 column packed with Phenyl Sepharose FF (low sub) to a final bed height o f 0.15m was equilibrated with 10 volumes of buffer A at a superficial liquid velocity o f 2m/h. Clarified material prepared as described in section 5.2.4, was loaded onto the column at the same flowrate in the same direction as equilibration buffer. Fractions were collected every 120s and the concentration o f ADH in the outlet from the column was assayed on collection of the fraction. Loading proceeded to total breakthrough as defined by the outlet concentration o f ADH in three successive fractions being equal to the inlet concentration. At this point, the flow to the column was switched from supernatant to buffer A and 15 volumes were used to wash the column before the flowrate was decreased to 0.75m/h and a step change to buffer B was applied to elute bound material.

Eluted fraction were pooled and assayed for ADH and protein. The total ADH eluted per unit volume o f matrix corresponded to the total binding capacity (TBC) of the matrix for ADH.

5.2.4.5 Dynamic binding capacity of Phenyl Sepharose FF (low sub) for ADH. The XK50/40 column prepared and equilibrated as previously described (section 5.2.2.3) was loaded with supernatant at 2m/h until the breakthrough o f ADH reached approximately 5% of the inlet concentration. Unbound material was washed from the column in the reverse direction to loading with 7.5 volumes of buffer A. Bound ADH was eluted with a step decrease in salt concentration by applying buffer B to the column at 0.75m/h. The total ADH eluted per unit volume o f matrix corresponded to the dynamic binding capacity (DBC) o f the matrix for ADH at a breakthrough of 5% under the adsorption conditions chosen. The column was regenerated with a CIP cycle as described in section 5.2.2.3.

5.2.5 Case study 4 - Pilot-scale clarification of yeast homogenate by two-cut