Reduction of endotoxin in polysialic acids

Abstract

The present invention relates to process for reducing the endotoxin content of a sample of fermentation broth containing polysialic acid and endotoxin comprising the sequential steps: (i) adding to the sample a base having a pKa of at least 12 to form a basic solution having a pH of at least 12, incubating the solution for a pre-determined time at a pre-determined temperature; and (ii) recovery of PSA, suitably by (iii) passing the sample through an anion-exchange column whereby polysialic acid is absorbed on the ion exchange resin; (iv) washing the column with one washing buffer, whereby polysialic acid remains absorbed on the ion exchange resin; and (v) eluting the polysialic acid from the column using an elution buffer to provide a product solution of polysialic acid having reduced endotoxin content.

Claims

1. A method for reducing the content of an endotoxin in a sample that contains a polysialic acid (PSA) and the endotoxin, the method comprising: a) adding a base having a pKa of at least 12 to the sample to form a solution having a pH of at least 12; wherein the PSA in the sample has a weight average molecular weight of up to about 100 kDa; b) incubating the sample for a time period ranging from 5 minutes to 24 hours; c) passing the sample through an anion-exchange column; d) washing the column with a wash buffer; and e) recovering the PSA from the anion exchange column, wherein the recovered PSA sample has a reduced endotoxin content.

2. The method of claim 1, wherein the base has a pKa of at least 13 and/or the pH of the formed solution is 13.

3. The method of claim 2, wherein the base has a pKa of at least 14 and/or the pH of the formed solution is 14.

4. The method of claim 1, wherein the time period ranges from 30 minutes to about 6 hours.

5. The method of claim 4, wherein the time period is about 2 hours.

6. The method of claim 1, wherein the PSA includes a poly(2,8-linked sialic acid), a poly(2,9-linked sialic acid), an alternating a poly(2,8-2,9-linked sialic acid), or an oxidized, a reduced, an aminated, or a hydrazide derivative thereof.

7. The method of claim 6, wherein the PSA is colominic acid or an oxidized, a reduced, an aminated, or a hydrazide derivative thereof.

8. The method of claim 1, wherein step (a) is carried out at a temperature ranging from 0 C. to about 60 C.

9. The method of claim 8, wherein the temperature ranges from 2 C. to about 40 C.

10. The method of claim 9, wherein the temperature is about 20 C.

11. The method of claim 1, wherein in step (d) the washing buffer is a low ionic strength washing buffer.

12. The method of claim 11, wherein the low ionic strength washing buffer has a salt concentration of 100 mM or less.

13. The method of claim 1, wherein the reduced endotoxin content is 25 EU/mg or less.

14. The method of claim 13, wherein the reduced endotoxin content is 5 EU/mg or less.

15. The method of claim 14, wherein the reduced endotoxin content is 0.5 EU/mg or less.

16. The method of claim 1, wherein step (a), step (b), step (c), step (d) and/or step (e) are repeated.

17. The method of claim 1, wherein step (c), step (d) and/or step (e) are repeated.

18. The method of claim 1, further comprising neutralizing the sample after step (b).

19. The method of claim 18, wherein the sample is neutralized to a pH of about 7.4.

20. The method of claim 1, wherein the sample recovered in step (e) is volatilized to remove a volatile acid and/or a volatile base.

21. The method of claim 1, wherein the sample recovered in step (e) is treated to form a more highly concentrated PSA sample.

22. The method of claim 1, wherein the sample recovered in step (e) is further purified to remove salt.

23. The method of claim 1, wherein the sample recovered in step (e) is precipitated to reduce polydispersity.

24. The method of claim 1, further comprising passing the sample through a hydrophobic interaction chromatography column, an affinity chromatography column or a size exclusion column before step (a) or after step (e).

25. The method of claim 1, further comprising incubating the sample with a surfactant, a chelating agent, an organic solvent, an oxidase and/or a peroxidase before step (a) or after step (e).

Description

EXAMPLES

Reference Assay for Endotoxin

(1) To perform the assay for endotoxin, The EndosafePTS (Portable Test System) from Charles River Laboratories was used. This is based on the LAL (Limulus Amoebocyte Lysate) assay.

(2) Instrument Operation

(3) All required information was entered into the reader. Once all the test information had been entered, the reader displayed add sample; press enter. The PSA samples were prepared, unless otherwise specified at 1 mg/ml in 20 mM TEA buffer at pH 7.4. 25 L of sample was pipetted into all four sample reservoirs and enter key was pressed on the reader. The pump drew sample aliquots into the test channel and the results were produced in 15-20 minutes. When the test was completed the instrument displayed the endotoxin measurement and the assay acceptance criteria on the screen. The instrument gave the following specifications: Sample EU/mL, sample % CV, spike EU/mL, spike % CV and % spike recovery.

Example 1

Endotoxin Reduction Using Sodium Hydroxide

(4) 6 mg/ml solution of colominic acid contaminated with endotoxin (31 kDa unoxidised) was prepared in 0.5M NaOH Hepes buffer and was incubated at room temperature for 10 minutes. Then 0.5 ml solution was loaded on size-exclusion chromatography desalting column and the fraction collected was discarded. The column was then washed with 2.5 ml HEPES buffer and the fraction was collected followed by collection of another 2 ml fraction with HEPES buffer. The elution fractions collected were then analysed for colominic acid content by resorcinol assay. The fractions containing colominic acid were then pooled together and were analysed for endotoxin content. The samples were tested for degree of deacetylation of the PSA. FIG. 1 shows the Native PAGE of colominic acid in 0.5 M NaOH and 1% SDS HEPES buffer solution.

(5) There was a 53-fold reduction in endotoxin content using NaOH and there was no detectable degree of deacetylation. Nor is any breakdown of CA observed in the PAGE using NaOH and SDS.

Example 2

Reduction of Endotoxin by Anion Exchange after Base

(6) A sample taken directly from E. coli K1 fermentor (1 mg/ml) in 20 mM TEA buffer (pH 7.4) was measured and found to be more than 10.sup.5 EU/mg. Then NaOH was added to the PSA solution to make final normality of 2N NaOH in PSA sample solution and then it was incubated for 2 hours at room temperature with gentle mixing. The pH of the solution was recorded. The HiTrap QFF (1 ml) column was washed with 10 column volume deionised water (pH 7.4) and then column was equilibrated using 10 column volumes of 20 mM TEA. The conductivity of the sample solution was measured and was diluted appropriately with 20 mM TEA buffer to match the conductivity of buffer solution. The sample solution was then loaded on QFF (1 ml column) at the rate of 1 ml/min and the void volume (.about. column volume) was collected separately. 1 ml fractions were collected. The column was then washed with 20 mM TEA and washing fractions of 1 ml each was collected. The sample was then eluted with 1M NaCl in 20 mM TEA and the fractions were collected. Resorcinol assay of the elution samples was performed to calculate the amount of colominic acid present in the samples and the endotoxin content of the pooled elution samples containing colominic acid was determined. The endotoxin level of the product was found to be reduced to 1407 EU/mg, i.e. more than about.71 fold. The recovery of PSA was 91%.

(7) The base treatment and recovery by anion exchange was repeated taking above product and further reduction in endotoxin content was seen, down to less than 300, i.e. a further 5-fold reduction.

Example 3

Reduction of Endotoxin by Anion Exchange after Nonionic Surfactant Treatment

Triton X-100

(8) Endotoxin content of the original sample (1 mg/ml) in 20 mM TEA buffer (pH 7.4) was measured. Colominic acid solution (35 mg/ml contaminated with endotoxin) was prepared in 1% Triton X 100 and was incubated for 2 hours at room temperature. The pH of the solution was measured. The HiTrap QFF column was prepared as in Example 2 and the sample was prepared and loaded as in Example 2. The column was then washed with 20 mM TEA and washing fractions of 1 ml each were collected. The sample was then eluted with 1M NaCl in 20 mM TEA and the fractions were collected. Resorcinol assay of the elution samples was performed to calculate the amount of colominic acid present in the samples and the endotoxin content of the pooled elution samples containing colominic acid was determined. Where the loading of colominic acid was 7.28 mg, the endotoxin content of the starting material was reduced from 4023 to 1511 EU/mg, i.e. about 3 fold. Recovery was 97%.

Example 4

Reduction of Endotoxin by Nonionic Surfactant Treatment (Triton X 114) and Anion Exchange

(9) 1% Triton X 114 solution was prepared and was added to PSA solution derived from fermentation broth by standard centrifugation, lysis, diafiltration ad ultrafiltration in appropriate amount to make the final concentration of 0.5% Triton X 114 and was incubated for 2 hours at room temperature. The pH of the solution was measured. The HiTrap QFF column was prepared, loaded, washed and eluted as in Example 3 and the endotoxin content of the pooled elution samples containing colominic acid was determined. The starting endotoxin level of more than 10.sup.5 EU/mg was reduced to around 2.310.sup.4 EU/mg, i.e. by around 4 fold.

Example 5

Reduction in Endotoxin by Base/Anion Exchange Followed by Nonionic Surfactant and Anion Exchange

(10) Colominic acid solution product of a different fermentation broth was treated by base by the process of Example 2 and then treated with Triton X 114 as below, then anion exchange.

(11) 1% Triton X 114 solution was prepared and was added to PSA solution in appropriate amount to make the final concentration of 0.5% Triton X 114. The solution becomes cloudy at room temperature (25 C.). To make the solution clear, it was kept in ice for 10-15 mins. Again the solution was kept at room temperature for 20 mins to make it cloudy. The cloudy solution was centrifuged and two layers were separated: upper layer containing PSA and lower layer of Triton X 114 containing endotoxin. The upper layer was kept to load on to the QFF column. The HiTrap QFF column was prepared, loaded, washed and eluted as in Example 4.

(12) In this case the level of endotoxin was reduced from over 10.sup.5 to around 4.4 to 10.sup.3 EU/mg in the first base treatment step and further to 8.210.sup.2 EU/mg after the second surfactant step and anion exchange.

Example 6

Reduction of Endotoxin by Nonionic Surfactant Treatment (Tween 80) and Anion Exchange

(13) Colominic acid solution (100 mg/ml contaminated with endotoxin; 2.5 g batch) was prepared in 0.1% Tween 80 and the pH was adjusted to 7.4. The solution was incubated for 30 minutes at room temperature and diluted to 500 ml with water pH 7.4. The HiTrap OFF column (75 ml) was washed with 10 column volume deionised water (pH 7.4) and then the column was equilibrated using 10 column volumes of water pH 7.4. The sample solution was then loaded at the rate of 7 ml/min at room temperature. The loading fractions were collected in fractions of 50 ml falcon or as appropriate. The first 25 ml of loading out was collected separately which accounts for the void volume of the column. Column was then washed with 0.01% Tween 80 in water pH 7.4 (7 ml/min, 4 CV) and washing fractions of 75 ml each were collected. The column was then washed with water pH 7.4 (7 ml/min, 4 CV) and collected the washing fractions of 75 ml or as appropriate. The column was then washed with 150 mM sodium chloride in water pH 7.4 (7 ml/min, 8 CV) and collected the fractions of 75 ml or as appropriate. The sample was then eluted with 500 mM sodium chloride in water pH 7.4 and collected the elution fraction of 75 ml or as appropriate. The samples were collected at room temperature. Resorcinol assay of the elution samples was performed to calculate the amount of colominic acid present in the samples and the endotoxin content of the pooled elution samples containing colominic acid was determined.

Example 7

Endotoxin Removal by Hydrophobic Interaction Chromatography

(14) 2M Ammonium sulphate in deionised water/deionised water was used as the loading buffer. 500 g of Colominic Acid (7 kDa produced in Reference Example 1) was dissolved in 500 L of loading buffer and the solution was loaded on the HIC columns specified in the Table below. The columns were then incubated for one hour. Elution samples were collected and resorcinol assay was performed to estimate the amount of colominic acid in the elution samples. Elution samples containing colominic acid were pooled and analysed for endotoxin content.

(15) TABLE-US-00001 Original Final Colominic Endotoxin Endotoxin Column Loading Acid Content Content Used Buffer (g) (EU/mg) (EU/mg) Butyl FF 2M (NH.sub.4).sub.2SO.sub.4 336.95 3070 211 Phenyl FF 2M (NH.sub.4).sub.2SO.sub.4 353.80 3070 148 Octyl FF Deionised Water 379.49 3070 214

(16) Endotoxin content was reduced by up to 14.5 fold. The order of Endotoxin reduction by different columns were in order: Phenyl>Butyl.gtoreq.Octyl.

Example 8

Endotoxin Removal Using Anionic Surfactant (Sodium Dodecyl Sulphate)

(17) 6 mg/ml solution of colominic Acid (31 KDa contaminated with endotoxin) was prepared in 1% SDS HEPES Buffer and was incubated for 1 hour at 37 C. Then 0.5 ml solution was loaded on Pd10 column and the fraction collected was discarded. The column was then washed with 2.5 ml HEPES buffer and the fractions were collected. The column was then eluted with 2 ml HEPES buffer. The elution fractions collected were then analysed for colominic acid content by resorcinol assay. The fractions containing colominic acid were then pooled together and were analysed for endotoxin content.

(18) Endotoxin content was reduced from >10 exp 6 EU/mg to 741.6 EU/mg. The recovery of the PSA was 84%.

Reference Example 1

Fractionation Using Anion Exchange

(19) A new prepacked column (1000 ml; Q Sepharose FF, GE Healthcare) was prepared and the preservative was washed with three column volumes of deionised water, and then with 3 column volumes of wash buffer at a flow rate of 50 ml/min. The pump tubing was filled with start buffer (triethanolamine buffer, pH 7.4; 20 mM) and the column was connected to the pump and a few drops of start buffer were applied to the top of the column to avoid introducing air into the column. Colominic acid solution derived from E. coli fermentation broth contaminated with endotoxin was prepared in triethanolamine buffer and the pH of the solution was adjusted to 7.4. The sample (colominic acid (CA) was obtained from Marukin, Japan) 750 ml in wash buffer) was then applied to column at the rate 50 ml/min, followed by washing of the column with another 750 ml of wash buffer. The column was then washed with 1500 ml washing buffer. The bound CA was eluted with 1500 ml of different elution buffers from 100 mM NaCl to 475 mM NaCl collecting washings from each run and transferring them in respective containers. All residual CA and other residues were removed with 1500 ml 1 M NaCl and the washings were collected. The column was then regenerated with 3 column volumes of wash buffer. The column was then stored in 20% ethanol at RT. The large chain length samples (removed by the high salt eluent) were then concentrated to minimum volume in 250 ml concentrators (Vivacell, Vivascience) under 4 bar pressure at 4 C. The concentrates were washed four times with distilled water (pH 7.4 adjusted with NaOH). The small chain length samples were also concentrated in 50 ml concentrators (Vivaflow, Vivascience) under pressure at 4 C. The concentrate was washed four times with distilled water. Samples were assayed for colominic acid content by resorcinol assay. The samples were then analyzed for endotoxin content.

(20) The results show that the endotoxin level was reduced from a value of 1.6.times.10 exp 5 EU/mg to 3070 EU/mg. There was almost a 5 fold reduction in the endotoxin content.

Reference Example 2

Affinity ChromatographyDetoxi Gel Column Purification

(21) The Detoxi gel columns were regenerated using pyrogen free solutions to prevent introducing any endotoxin into the sample. All the solutions were degassed before applying to the column to prevent air bubbles from clogging the column and reducing the flow. Detoxi-gel endotoxin removing gel may be used at least 10 times without loss of activity. All the solutions and gel were equilibrated to room temperature before use.

(22) The gels were degassed by placing slurry in the bottom of a suction filter flask with a magnetic stirrer. While the slurry was stirred an aspirator was used to create a vacuum within the flask. The gel was degassed for approximately 15 minutes. The appropriate sized column was packed with degassed slurry and the gel was allowed to settle down for 30 minutes. The gels were regenerated by washing with five column volumes of 1% sodium deoxycholate, followed by 3-5 column volumes of a pyrogen free water to remove excess. The gels were regenerated before each use following the same procedure. The sample was then applied to the column. Aliquots of pyrogen free buffer or water were added and the flow through was collected. The sample emerged from the column after the void volume collection was completed (94% of the bed volume). For greater efficiency, the bottom and top caps were replaced after sample has entered the gel bed. The column was incubated for at least one hour, the top and bottom caps were removed sequentially. Pyrogen free buffer or water was then added to collect the samples. Caution was taken to prevent sample contamination from dust or dirty glassware subsequent to endotoxin removal in all these experiments. Samples were then frozen and stored. The columns were regenerated following the same procedure as above to remove any bound endotoxin and were stored in 25% ethanol at 2-8 C.

(23) In the first example, the sample is the product of a process as disclosed in Example 3 of WO2008/012525, of colominic acid fractionated by the techniques described in WO2006/016161 conjugated to GCSF. The conjugate prior to affinity gel treatment had an endotoxin level of 438 EU/mg, and after treatment the level was reduced to 10.5 EU/mg. Endotoxin content of PSA-protein conjugates was reduced up to 35 times using this affinity chromatography material.

(24) In a second example the sample was neat colominic acid used under the following conditions:

(25) TABLE-US-00002 PSA (mg/ml) Amount [Resorcinol of PSA Formulation assay] (mg) Vol pH PSA 19.3 KDa 0.1749 0.1574 0.9 ml 7.4 (Marukin) In Hepes buffer (20 mM Hepes, 150 mM NaCl; pH 7.4)

(26) Endotoxin content was reduced to 4.2 EU/mg from original value of 16,000 EU/mg. The recovery was more than >90%.

(27) In a third exemplification the starting material was the solution produced in Reference Example 1, under the following conditions. The results are shown below:

(28) TABLE-US-00003 PSA (mg/ml) Amount Endotoxin [Resorcinol of PSA Content Formulation assay] (mg) Volume (EU/mg) CA 7 kDa 0.374 1.87 5 ml 66.04 CA 7 kDa 0.160 0.8 5 ml 178.75

(29) Original endotoxin content in the sample was 3070 EU/mg. The reduction in the endotoxin content was up to 47 fold.

CONCLUSION

(30) Endotoxin-specific affinity chromatography may be used to remove endotoxin from polysialic acid and conjugates thereof, i.e. we have shown that it is possible to select conditions under which endotoxin binds to the column while PSA does not bind but can instead be recovered in a convenient form and with very low endotoxin level. The step may thus be useful to treat PSA products of base (or surfactant) treatment.

Reference Example 3

Endotoxin Removal Using Affinity Column (Cellufine)

(31) This Example is similar to Reference Example 2 but uses a different endotoxin-removing affinity column. Cellufine ET clean column (S beads) was regenerated by washing with 5 column volumes of 0.2M NaOH, 2M NaCl and then with endotoxin free water. The column was then equilibrated with 5 column volumes of suitable endotoxin free buffer (HEPES Buffer). The colominic acid solution produced in Reference Example 1 in HEPES buffer (1 mg/ml) was then applied to the column at a flow rate of 0.1-0.2 ml/min. at 21 C. The total amount loaded was 218 g. The column was then incubated for one hour and then the elutions were collected with HEPES buffer. Resorcinol assay of the elution samples was performed to calculate the amount of colominic acid present in the samples and the endotoxin content of the pooled elution samples containing colominic acid was determined.

(32) The endotoxin level was reduced from 3070 to 75 EU/mg i.e. amount 40-fold.

(33) The method was repeated using the GCSF-PSA conjugate used as the starting material for Reference Example 2 part 1, reducing the endotoxin level from 438 to 12.4 EU/mg.

(34) We conclude that the Cellufine column is suitable for removing endotoxin from PSA.

Reference Example 4

Removal of Endotoxin Affinity Column (EndoTrap Red)

(35) This example uses another affinity column for endotoxin adsorption. Endotrap Red column was regenerated and equilibrated using deionised water pH 7.4. 10 ml of 50 mg/ml colominic acid solution contamination with endotoxin was prepared. Two columns were kept in series. The sample solution was loaded onto the column and the void volume was collected (.about. of the column volume). Then remaining loading solution was collected. Then the column was washed with 6 column volume deionised water pH 7.4 and the fractions of 1 ml each was collected. Resorcinol assay for all the fractions was performed to calculate the amount of colominic acid present and the endotoxin content of the samples containing colominic acid was determined.

(36) The results for 16 kDa colominic acid show a reduction in endotoxin level from 564 to 6 EU/mg, a reduction of around 90 fold.

(37) The procedure was also used for an endotoxin-contaminated insulin-PSA conjugate produced by the method described in WO2008/012528. The endotoxin content was reduced from 111 to 12.5 EU/mg, a 9-fold reduction.

(38) These examples show that another column is useful for removal of endotoxin from PSA and conjugates thereof.

Reference Example 5

(39) Removal of Endotoxin Through Affinity Column Using HEPES Buffer

(40) Endotrap Red column (1 ml column volume) was regenerated with regeneration Buffer Red provided with the column. The column was then equilibrated using 5 column volumes of suitable endotoxin free 20 mM HEPES Buffer. The endotoxin-contaminated PSA solution in HEPES buffer (750 g/ml) was then applied to the column at a flow rate of 0.1-0.2 ml/min at 21 C. Then the column was eluted using 0.3 ml of 20 mM HEPES buffer. Resorcinol/protein assay of the elution samples was performed to calculate the amount of colominic acid present in the samples and the endotoxin content of the pooled eluted samples containing colominic acid was determined. The results show that the endotoxin level may be reduced by around 2 to 7 fold, and that this is affected by the load of PSA used (lower loads on the column give better reductions) at loading levels at which 100 PSA recovery is achieved.

Reference Example

Removal of Endotoxin, Proteins, DNA and Cell Debris Through Anion Exchange Column Using 30% IPA

(41) Fermentation broth recovered by centrifugation, supernatant recovery, diafiltration and ultrafiltration, with a colominic acid concentration of 35 mg/ml was prepared and its pH was measured. The HiTrap OFF column was washed with 10 column volume deionised water (pH 7.4) and column was equilibrated using 10 column volumes of 20 mM TEA. The conductivity of the sample solution was measured and was diluted appropriately with 20 mM TEA buffer to match the conductivity of buffer solution. The sample solution was then loaded at the rate of 1 ml/min fractions were collected separately. Then 1 ml of fractions was collected. Column was then washed with 30% IPA and washing fractions of 1 ml each was collected. The sample was then eluted with 1M NaCl in 20 mM TEA and the fractions were collected. Resorcinol assay of the elution samples was performed to calculate the amount of colominic acid present in the samples and the endotoxin content of the pooled elution samples containing colominic acid was determined. The initial endotoxin content or the broth extract was 3.210.sup.4 EU/mg and the final content was 1.810.sup.3 EU/mg, i.e. the endotoxin content was reduced by 18 fold.