Method and kit for sample preparation and endotoxin determination

Abstract

The invention relates to a method for preparation of a sample (10) of a formulation (11) for subsequent endotoxin determination, the formulation (11) suspected of comprising an endotoxin, the formulation (11) preferentially being a pharmaceutical formulation. The method comprises the following steps: application of the sample (10) to an endotoxin-free centrifugation column (2) containing a size exclusion chromatography matrix (5) that has been equilibrated with a suitable equilibration buffer (6) and elution of a flow through (15) of the sample by centrifugation, which flow through (15) can then be used for endotoxin determination. The equilibration buffer (6) is selected according to a subsequently used method of endotoxin determination, the equilibration buffer (6) only containing components not interfering with subsequently used method of endotoxin determination. Furthermore, the invention relates to a kit (20) for preparation of a sample (10).

Claims

1. A method for evaluation of Low Endotoxin Recovery (LER) effects in a pharmaceutical formulation and/or for an elaboration of a composition of a suitable equilibration buffer (6), comprising: spiking an undiluted sample of a pharmaceutical formulation with a known activity of a lipopolysaccharide (LPS) standard, testing an aliquot of the spiked sample for endotoxin, if the spiked sample shows a test result corresponding to less than 50% of the known activity of the lipopolysaccharide (LPS) standard spiked into the sample, then applying aliquots of the spiked sample to at least two endotoxin-free centrifugation columns (2), wherein the centrifugation columns (2) containing a size exclusion chromatography matrix (5) and each of the at least two endotoxin-free centrifugation columns (2) having been equilibrated with different equilibration buffers (6); eluting a flow through (15) from each centrifugation columns (2) by centrifugation, which flow through (15) can then be used for endotoxin determination; and selecting the equilibration buffer (6) best suited for lipopolysaccharide recovery for the given pharmaceutical formulation based on the results of the endotoxin determination.

2. The method of claim 1, wherein the results of the endotoxin determination are compared to an aliquot of the non-centrifuged spiked sample and a positive water control.

3. The method of claim 1, further comprising incubating the undiluted sample of the given pharmaceutical formulation spiked with a known activity of the lipopolysaccharide (LPS) standard at a temperature for a time period prior to endotoxin testing step.

4. The method of claim 3, wherein the incubating step is performed between 1° C. and 37° C.

5. The method of claim 3, wherein the incubating step is performed at about 4° C. or wherein the incubating step is performed at a room temperature between 18° C. and 24° C.

6. The method of claim 3, wherein the incubating step is performed for at least 1 h.

7. The method of claim 3, wherein the incubating step is performed for between 24 h and 168 h.

8. The method of claim 1, wherein the size exclusion chromatography matrix (5) has a size exclusion volume or exclusion cut-off within the range of 2000 Dalton to 20000 Dalton.

9. The method of claim 8, wherein the size exclusion chromatography matrix (5) has a size exclusion volume or exclusion cut-off within the range of 4000 Dalton to 7000 Dalton.

10. The method of claim 9, wherein the size exclusion chromatography matrix (5) has a size exclusion volume or exclusion cut-off below 6000 Dalton.

11. The method of claim 1, wherein the size exclusion chromatography matrix (5) is an uncharged, hydrophilic gel matrix or wherein the size exclusion chromatography matrix (5) is a crosslinked polyacrylamide gel matrix, and wherein the size exclusion chromatography matrix (5) has a high mechanical stability at centrifugation forces of up to 1800 g.

12. The method of claim 1, wherein the equilibration buffer (6) comprises a buffer substance, different from the buffer used in a formulation (11) that is to be tested for endotoxin presence and wherein the equilibration buffer (6) comprises at least one bivalent cation.

13. The method claim 12, wherein the equilibration buffer (6) comprises Ca.sup.2+ or Mg.sup.2+ as bivalent cation, or wherein the pH of the equilibration buffer (6) is 6.0 to 8.5.

14. The method of claim 13, wherein the equilibration buffer (6) comprises Ca.sup.2+ or Mg.sup.2+ in a concentration between 1 mM and 100 mM.

15. The method of claim 13, wherein the equilibration buffer (6) comprises Ca.sup.2+ or Mg.sup.2+ in a concentration between 1 and 50 mM and wherein the pH of the equilibration buffer (6) is between 7.0 and 8.0.

16. The method of claim 13, wherein the equilibration buffer (6) comprises Ca.sup.2+ or Mg.sup.2+ in a concentration between 20 and 50 mM and wherein the pH of the equilibration buffer (6) is between 7.0 and 8.0.

17. The method of claim 1, wherein the equilibration buffer (6) comprises an amphiphilic substance in a concentration below its critical micelle concentration, or equal or below its solubility threshold in the equilibration buffer (6).

18. The method of claim 16, wherein the amphiphilic substance is selected from a group comprising Lauryl alcohol, Tween 20, Polypropylenglycol or SDS.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) In the following passages, the attached figures further illustrate exemplary embodiments of the invention and their advantages. The size ratios of the individual elements in the figures do not necessarily reflect the real size ratios. It is to be understood that in some instances various aspects of the invention may be shown exaggerated or enlarged to facilitate an understanding of the invention.

(2) FIGS. 1 and 2 show the preparation of a column provided by the kit and used for the methods according to the invention.

(3) FIGS. 3 and 4 show the application and treatment of a sample.

(4) FIG. 5 shows the fluorescence intensity in the flow through of a 200 μl sample depending on used bed volume (see experiment 2).

(5) FIG. 6 shows the fluorescence intensity in the flow through of a 300 μl sample depending on used bed volume (see experiment 2).

(6) FIG. 7 shows the components of a first embodiment of a kit according to the invention.

(7) The same or equivalent elements of the invention are designated by identical reference characters. Furthermore, and for the sake of clarity, only the reference characters relevant for describing the respective figure are provided. It should be understood, that the embodiments described are only examples and they are not intended to limit the scope of the disclosure.

DETAILED DESCRIPTION OF THE INVENTION

(8) FIGS. 1 and 2 show the preparation of a column 1 provided by the kit and used for the methods according to the invention. FIGS. 3 and 4 show the application and treatment of a sample 10.

(9) The column 1 is especially an endotoxin-free centrifugation column 2 that can be placed in a suitable centrifugation container 3 like a Sarstedt tube 4 or the like. The column 1 is filled with a size exclusion chromatography matrix 5, especially with Biogel P4 or another suitable gel matrix or resin. The gel matrix 5 is equilibrated with a suitable equilibration buffer 6, especially two to three volumes equilibration buffer 6 are used compared to the bed volume of the gel matrix 5. The equilibration buffer 6 only contain components not interfering with the subsequently used endotoxin testing method. The equilibration buffer preferentially comprises a buffer substance that is different from the buffer used in the formulation that is to be tested for endotoxin and comprises at least one bivalent cation. Especially the equilibration buffer 6 may comprise Ca.sup.2+ and/or Mg.sup.2+ as bivalent cations in a concentration range between 1 mM and 100 mM. The pH value of the equilibration buffer 6 should be around neutral, preferably between 6.0 and 8.5, most preferably between 7.0 and 8.0. Furthermore, the equilibration buffer 6 may comprise an amphiphilic substance in a concentration below its critical micelle concentration or the concentration of the amphiphilic substance should be equal or below its solubility threshold in the equilibration buffer system, the amphiphilic substance stabilizing monomeric LPS molecules during the phase of aggregate formation. Especially the amphiphilic substance is selected from a group comprising Lauryl alcohol, Tween 20 (Polyethylene glycol sorbitan monolaurate), Polypropylenglycol or SDS (sodium dodecyl sulfate). The equilibration buffer 6 may contain: 20 mM Tris/HCl OR 20 mM HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) at pH 7.4 plus 50 mM NaCl and 20 mM to 50 mM Ca.sup.2+ and/or Mg.sup.2+.

(10) For example, a column 1 containing 1.0 ml gel matrix 5 is equilibrated with 2.0 ml to 3.0 ml equilibration buffer 6.

(11) In order to prepare the column 1 for the sample 10 (see FIG. 3), excess equilibration buffer 6* is removed by centrifugation of the column 1 placed within a collection container 3, leaving equilibrated gel matrix 5* within the column 1. This ensures that the sample volume is not increased during the sample preparation described in FIGS. 3 and 4. This equilibration buffer removing centrifugation step is preferentially performed at a centrifugal force of more than 1,500 g, especially at a centrifugal force of about 1,800 g.

(12) According to FIG. 3 a sample 10 of the formulation 11 suspected of comprising an endotoxin or spiked with a known quantity of endotoxin is applied onto the equilibrated gel matrix 5* of the column 1, especially up to 200 μl sample 10 are applied onto a column 1 with 1.0 ml equilibrated gel matrix 5*. The components within the formulation 11 interfering with the endotoxin tests are usually comparatively small molecules with sizes less than 1,500 Dalton. After application of the sample 10 onto the equilibrated gel matrix 5* of the column 1, these small molecules can quickly spread into the equilibrated gel matrix 5*. The endotoxin or LPS is usually present as a high-molecular complex or aggregate, which remains in the exclusion volume of the column 1.

(13) If the formulation 11 comprises a detergent in a concentration above its critical micelle concentration (cmc), the detergent molecules tend to form micelles, thereby integrating LPS/endotoxin into the micelles. This micelle incorporated LPS/endotoxin cannot be detected by the known endotoxin testing methods. The detergent molecules are in a dynamic equilibrium between monomers and micelles. If the formulation 11 is applied onto the column 1, the detergent monomers spread into the equilibrated gel matrix 5* and are not further available for micelle formation. As a result, the LPS/endotoxin is released from the micelles during the centrifugation step can now aggregate with other LPS molecules into high-molecular LPS complexes or LPS aggregates. This LPS aggregation especially takes place in the presence of bivalent cations contained in the equilibration buffer 6, which explains the inhibiting properties of bivalent cation chelators on endotoxin testing methods.

(14) Within the centrifugation container 3 and below the column 1 an endotoxin free collection tube 7 is placed. After another centrifugation step the flow through 15 is collected within the collection tube 7—see FIG. 4. The centrifugation step separates the large LPS complex or aggregate from the small components of the formulation 11. Meanwhile the small components remain within the gel matrix now referenced as 5**, the large LPS complex or aggregate can be found after the centrifugation in the flow through 15. The volume of the flow through 15 exactly corresponds to the volume of the applied sample 10. Therefore, no dilution takes place. The flow through 15 contains also the high molecular components of the sample 10, whereby these high molecular components are now buffered in equilibration buffer 6. Especially the flow through 15 comprises endotoxin in its monomeric and/or its re-aggregated form. This flow through 15 can now be tested for endotoxin.

(15) Hereby, it may be provided that the collected flow through 15 is incubated at a certain temperature for a defined time prior to endotoxin testing to allow equilibrium adjustment between monomeric LPS and aggregated LPS. The equilibrium is located far on the side of the re-aggregated LPS. The incubation time in this step is normally less then 1 h at room temperature.

(16) FIG. 7 shows the component of a first embodiment of a sample preparation kit 20 according to the invention. The sample preparation kit 20 comprises an endotoxin-free spin column 1 prepacked with a size exclusion chromatography gel matrix 5, a centrifugation container 3, a standard equilibration buffer 6, an endotoxin free collection tube 7 and a user manual 9. The equilibration buffer 6 is used to equilibrate the gel matrix 5 (see FIG. 1) prior to sample preparation according to the explained method. The user manual 9 contains explanations and experimental protocols regarding the method and the individual steps to be performed. The user manual 9 might also give examples for the preparation of other suitable equilibration buffer compositions and provides explanations and hints for selecting the equilibration buffer most suited for the preferred endotoxin testing method.

(17) Furthermore, the sample preparation kit 20 contains an endotoxin sample 22 to be used as positive control.

(18) Even if in the context of the figures there is generally talk of “schematic” representations and views, this does not in any way mean that the representations of figures should be of secondary importance with regards to the disclosure of the invention. The expert is perfectly capable of obtaining enough information from the schematically and abstractly drawn representations to facilitate his understanding of the invention. The figures and experiments thus enable the expert to derive a better understanding of the invention abstractly expressed in the claims and in the general part of the description based on the precise, explained implementations of the method in accordance with the invention and the precise, explained function of the kit components in accordance with the invention.

EXPERIMENTS

(19) General information: For all experiments, where possible, single use materials have been used. All substances have been selected for low endotoxin content and all formulations used were tested for absence of endotoxin. Glassware was either heat-baked (4 h at 200° C.) or treated with 1M NaOH overnight to remove residual endotoxin.

(20) Materials and Instruments used: USP Reference Standard Endotoxin (RSE) from SigmaAldrich (E. coli 0113 LPS, 10,000 EU/vial); LPS E. coli O55:B5, S. enterica abortus equi, S. enterica typhimorium, P. aeruginosa from SigmaAldrich; Lauryl alcohol and Polypropylenglycol 725 from SigmaAldrich; Bovine albumin, Bovine IgG highest purity from SigmaAldrich, inhouse LPS-depleted twice by use of Hyglos EndoTrap blue affinity matrix according to manufacturer's instructions (>0.1 EU/mg); Kinetic Chromogenic LAL Assay (KCA) from Lonza; EndoZyme, Recombinant Factor C Assay from Hyglos GmbH; PyroDetect-System, Monocyte Activation Test-MAT from Merck-Millipore; Sarsted centrifuge tubes, 15 ml PP, sterile, pyrogen-free; Sarsted reaction tubes, 1.5 ml PP, PCR quality; Qiagen chromatography columns, 1 ml PP; Heraeus Multifuge 3SR+, Swing-out rotor 75006445; and Heidolph Reax Multi, Tube shaker.

(21) Experiment 1: Selection of Suitable Gel Materials

(22) Gel materials used are characterized to have a size exclusion limit below 20,000 Dalton, a generally hydrophilic character and low non-specific interaction capacity. Especially the gel materials are specified as high yield chromatography media. The following gel materials were tested Biogel P2 fine from BioRad, Biogel P4 medium, fine and extra fine from BioRad, Biogel P6 fine and extra fine from BioRad, Biogel P10 fine from BioRad, Sephadex G25 fine and Superdex 30 prep grade, both from GE. The materials were tested for physical stability during centrifugation at 1,800 g. A plus sign (+) was assigned if compression of the gel material was less than 15%, a minus sign (−) was assigned if compression of the gel material was more than 15%. Separation efficiency was tested using with 100 μM 7-amino-4-methyl-coumarin, as described in experiment 2.

(23) TABLE-US-00002 TABLE 2 Suitability of different gel materials for the centrifugation assay Material (fractionation bead range) size Stability Separation comment Biogel P2, fine 45-90 + − High variation, (100-1,800 Da) μm poor separation, (wet) not suitable for the method Biogel P4, 90-180 + − High variation, not medium μm reproducible; not (800-4,000 Da) (wet) suitable for the method Biogel P4, fine 45-90 + + Suitable material (800-4,000 Da) μm for the method (wet) Biogel P4, <45 + + Suitable material extra fine μm for the method (800-4,000 Da) (wet) Biogel P6, 45-90 + + Suitable material fine μm for the method (1,000-6,000 Da) (wet) Biogel P6, <45 + + Suitable material extra fine μm for the method (1,000-6,000 Da) (wet) Biogel P10, fine 45-90 − − Matrix collapsed (1,500- μm during centrifugation, 20,000 Da) (wet) poor separation; not suitable for the method Sephadex G25, 20-80 − − Matrix collapsed fine μm during centrifugation, (100-5,000 Da) (dry) poor separation; not suitable for the method Superdex 30 22-44 + + Suitable material prep grade μm for the method (<10,000 Da) (wet)

(24) Five of the nine tested gel materials showed to be useful for the proposed methods. Inoperative materials either collapse during centrifugation because of low mechanical stability or they show a relatively high variance regarding the fluorescence signal or they leaked too much of the fluorescent dye. For the subsequent experiments Biogel P4 fine was used.

(25) Experiment 2: Gel Bed Volume Versus Separation Efficiency

(26) Optimal bed volume was evaluated using the fluorescent dye 7-Amino-4-methyl-coumarin, furthermore referred to as AMC, for quantification of the separation efficiency. The dye concentration was 100 μM in 20 mM HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid), 50 mM NaCl, 2 mM MgCl.sub.2, pH 7.2. Two test series were performed, one with 200 μl sample volume and one with 300 μl sample volume. The following centrifugation columns were prepared. All steps were performed at room temperature.

(27) Biogel P4 fine was hydrated with an equilibration buffer containing 20 mM HEPES, 50 mM NaCl, 2 mM MgCl.sub.2, pH 7.2 for 30 minutes. The slurry was sucked in a filter funnel and re-suspended with a threefold volume of the equilibration buffer. This procedure was repeated three times. The swelled gel matrix was adjusted to a 50% slurry which was subsequently used to prepare the centrifugation columns with varying bed volumes. The columns were filled using a 5 ml pipette with a disposable tip. Prior to the experiment the excess of equilibration buffer was removed by gravity. The prepared columns were placed in Sarsted tubes in centrifuge buckets. For all the experiments a Heraeus Multifuge 3SR+, Swing-out rotor 75006445 was used. The column was centrifuged with the following settings (see FIGS. 1 and 2).

(28) Acceleration to 1,800 g with profile 4; once the acceleration reached 1,800 g, the centrifuge was immediately stopped.

(29) Deceleration with profile 9

(30) Temperature setting 20° C.

(31) The liquid flow-through of excess equilibration buffer was removed from the Sarsted tubes and small reaction vials were inserted to collect the flow through in the next step (see FIGS. 3 and 4). To each of the columns 200 μl dye solution (series 1) or 300 μl dye solution (series 2) was applied. The buckets were placed in the centrifuge again and the centrifugation was repeated using the same setting as for the first centrifugation step removing excess equilibration buffer. The flow through of the second centrifugation was collected and analysed for volume and fluorescence intensity using a microplate fluorescence reader BioTek Flx800, Ex 380 nm/Em 440 nm and applying 100 μl of the flow through for testing.

(32) The results of this experiment are shown in table 3 and displayed in FIGS. 5 and 6, wherein FIG. 5 shows the fluorescence intensity in the flow-through of a 200 μl sample dependent on bed volume from 0.6 ml to 1.4 ml and FIG. 6 shows the fluorescence intensity in the flow-through of a 300 μl sample dependent on bed volume ranging from 0.6 ml to 1.8 ml.

(33) TABLE-US-00003 TABLE 3 Different column volumes tested with 200 μl and 300 μl sample volumes and the determined volumes of centrifugation eluate/flow through. bed volume/sample volume recovered 200 μl Sample 300 μl Sample 1 0.6 ml/202 μl 0.6 ml/302 μl 2 0.8 ml/201 μl 0.9 ml/308 μl 3 1.0 ml/204 μl 1.2 ml/310 μl 4 1.2 ml/203 μl 1.5 ml/312 μl 5 1.4 ml/205 μl 1.8 ml/314 μl

(34) It can be seen, that the separation efficiency for the 200 μl AMC solution is >99% at a bed volume of 1 ml and for 300 μl AMC solution a bed volume of 1.5 ml is sufficient to reduce the concentration by a factor of 100. Therefore, when applying a 200 μl sample, the bed volume of the gel matrix should preferentially be at least 1.0 ml and when applying a 300 μl sample, the bed volume of the gel matrix should preferentially be at least 1.5 ml.

(35) Experiment 3: LPS Recovery Versus Gel Bed Volume

(36) The following serial dilutions of LPS 055 were prepared in 20 mM HEPES, 50 mM NaCl, 2 mM MgCl.sub.2, pH 7.2 in endotoxin-free glass vials: 20 EU/ml, 4 EU/ml, 0.8 EU/ml, 0.16 EU/ml and 0.032 EU/ml. Between each dilution step the solution was vortexed for 2 minutes at 1,400 rpm using a Heidolph Reax Multi tube shaker. Centrifugation columns according to the procedure of experiment 2 were prepared with a bed volume of 1.5 ml. 300 μl of each dilution was applied and the column was centrifuged at 1,800 g. The flow-through fractions were collected in endotoxin-free plastic cups. For each dilution the flow-through of three columns was pooled (giving 900 μl) and analysed.

(37) The centrifuged samples (50 μl sample+50 μl endotoxin-free water per determination) were analysed in triplicates with three different endotoxin detection methods. Especially Kinetic chromogenic (KCA) LAL from Charles River, EndoZyme from Hyglos GmbH (recombinant Factor C-based fluorescent assay) and, Monocyte Activation Test (MAT) with frozen blood and IL-1B as readout, Pyrodetect System, Merck-Millipore were used. As a comparison and to calculate the recovery, the non-centrifuged dilutions are also analysed. All tests were carried out according to the manufacturer's instruction using the LPS standards supplied with the respective test kits or bought as separate items from the same manufacturer.

(38) Table 4 lists the recovery of LPS 055 after centrifugation through a column filled with 1.5 ml Biogel P4 fine.

(39) TABLE-US-00004 TABLE 4 Recovery of LPS O55 LAL EndoZyme ® MAT   20 EU/ML 89 ± 12% (92 ± 13%) invalid spike invalid spike control control    4 EU/ml 95 ± 8%  87 ± 18% invalid spike control  0.8 EU/ml 92 ± 7%  89 ± 11% 83 ± 26%  0.16 EU/ml 82 ± 13% 95 ± 9%  80 ± 41% 0.032 EU/ml 86 ± 14% 73 ± 15% below LoD* Average 89% recovery 87% recovery 82% recovery *Limit of Detection

(40) In summary, the three tests provide an overall LPS recovery between 82% and 89% showing that LPS can be recovered by centrifugation through the gel material in good yield. For interpretation one should have in mind that according to the acceptance criteria for endotoxin assays results are valid between 50% and 200% of the nominal value. Yields could probably further be increased by optimizing the composition of the gel equilibration buffer. The results of the Monocyte Activation Test are incomplete because of the very narrow dynamic range of the assay and the lower sensitivity of this cell-based assay compared to the other two tests.

(41) Experiment 4: Separation of Inhibitory Substances from Endotoxin

(42) A set of inhibitory substances or physical conditions, especially low and high pH values, was selected to prove the principle capacity of the method to eliminate interfering components from formulations. All formulations listed below were spiked with approximately 25 EU/ml LPS E. coli O113. 300 μl of the spiked formulation was either processed according to the methods described herein via a centrifugation column (1.5 ml bed volume, equilibrated with 20 mM HEPES, 50 mM NaCl, 2 mM MgCl.sub.2, pH 7.2) or analysed directly using the kinetic chromogenic (KCA) LAL from Charles River. Conditions for preparation, centrifugation and assay are as described in the previous experiments. 80 μl sample plus 20 μl endotoxin-free water was applied in the test (triplicate determination). 20 mM Acetat, 50 mM NaCl, 2 mM MgCl.sub.2, pH 4.0 100 mM Na Borat, 50 mM NaCl, 2 mM MgCl.sub.2, pH 9.0 20 mM HEPES, 50 mM NaCl, 2 mM MgCl.sub.2, 5% Ethanol, pH 7.2 20 mM HEPES, 50 mM NaCl, 2 mM MgCl2, 5% DMSO, pH 7.2 20 mM HEPES, 50 mM NaCl, 2 mM MgCl2, 0.5% SDS, pH 7.2 20 mM HEPES, 50 mM NaCl, 2 mM MgCl2, 0.05% Tween 20, pH 7.2 20 mM HEPES, 50 mM NaCl, 2 mM MgCl2, 0.05% Tween 20, 20 mM Citrat 20 mM HEPES, 50 mM NaCl, 2 mM MgCl2, 2 mM EDTA, pH 7.2 20 mM HEPES, 50 mM NaCl, 2 mM MgCl2, 20 mM Citrat, pH 7.2 20 mM HEPES, 50 mM NaCl, 2 mM MgCl2, 1 mM Benzamidine, pH 7.2 20 mM HEPES, 50 mM NaCl, 2 mM MgCl2, 2 mM PMSF, pH 7.2 20 mM HEPES, 50 mM NaCl, 2 mM MgCl2, 1 mM Chloramphenicol, pH 7.2

(43) Table 5 list the results of the LAL endotoxin determination

(44) TABLE-US-00005 TABLE 5 LAL results with centrifugation (method according without Additive to invention) centrifugation No Additive 18.85 ± 1.44 EU/ml 20.71 ± 1.28 EU/ml Acetate, pH 4.0 17.54 ± 2.34 EU/ml Invalid spike control Borate, pH 9.0 13.06 ± 1.41 EU/ml  3.39 ± 2.87 EU/ml 5% Ethanol 18.09 ± 1.89 EU/ml Invalid spike control 5% DMSO 16.99 ± 0.83 EU/ml Invalid spike control 0.5% SDS 18.17 ± 2.06 EU/ml  0.38 ± 2.05 EU/ml 0.05% Tween 20 17.07 ± 2.68 EU/ml Invalid spike control 20 mM Citrate, 14.76 ± 0.97 EU/ml Invalid spike control 0.05% Tween 20 2 mM EDTA 18.20 ± 1.58 EU/ml Invalid spike control 20 mM Citrate 15.66 ± 1.73 EU/ml Invalid spike control 1 mM 16.43 ± 1.86 EU/ml Invalid spike control Benzamidine 2 mM PMSF 12.75 ± 2.31 EU/ml Invalid spike control 1 mM 19.31 ± 1.95 EU/ml  4.87 ± 2.62 EU/ml Chloramphenicol

(45) The actual amount of the spike was determined to be 20.71 EU/ml. Direct analysis of the formulations with LAL provided mostly invalid results for the spike control or a significant under-determination. After centrifugation, no invalid results occurred and the recovery has been between 12.75 EU/ml (worst) and 19.31 EU/ml (best). This corresponds to 62% (worst) and 93% (best). These results show that adverse effects caused by different chemicals or extreme pH could be substantially reduced by centrifuging the sample through a gel filtration matrix having the capacity to withhold small molecules to very large extend and to let high molecular weight entities to pass the column without significant loss.

(46) Experiment 5: Reconstitution of Endotoxin from Protein Containing Formulation.

(47) Two formulations, each containing a detergent, a chelating substance, especially a buffer, a salt and a protein were spiked with endotoxin. Hereby, RSE=Reference Standard Endotoxin was used. The formulations were incubated in glass vials for 10 days at room temperature of about 20° C. to 24° C. and protected from light.

(48) TABLE-US-00006 TABLE 6 Composition of Formulations used in experiment 5 Component Formulation 1 Formulation 2 Buffer/Chelator 20 mM Citrate pH 7.4 25 mM Phosphate pH 7.0 Detergent 0.05% Tween 20 0.05% Tween 80 Salt 50 mM NaCl 60 mM NaCl Protein 3.125 mg/ml BSA 5 mg/ml bovine IgG E.coli LPS 50 EU/ml 50 EU/ml O113 (RSE)

(49) After 10 days the spiked formulations were measured directly using kinetic chromogenic (KCA) LAL from Charles River in dilutions 1:4 and 1:10. Additionally, the same samples (300 μl) were processed using 1.5 ml centrifugation column equilibrated in 20 mM HEPES, 100 mM NaCl, 50 mM MgCl.sub.2, pH 7.4. All samples were applied and centrifuged immediately. After centrifugation the eluates were incubated for 60 minutes at room temperature. Dilutions were prepared right before testing, using endotoxin-free water, glass vials and 2 minutes of intensive mixing at about 1,400 rpm.

(50) TABLE-US-00007 TABLE 7 Recovery of LPS hidden in formulations according to table 6 without centrifugation (method according to invention) with centrifugation 1:4 dilution 1:10 dilution 1:4 dilution 1:10 dilution (recovery) (recovery) (recovery) (recovery) Formulation 0.34 EU/ml 0.18 EU/ml 7.37 EU/ml 2.15 (EU/ml) 1 (2.72%) (3.60%%) (58.96%) (43.00%) Formulation 0.07 EU/ml <0.05 EU/ml 3.63 EU/ml 1.22 (EU/ml) 2 (0.56%) (<1.00%) (29.04%) (24.4%)

(51) Two formulations with the capacity to hide spiked LPS from detection by a kinetic chromogenic LAL assay have been evaluated to determine whether separation of the chelator and detergent from the formulation can improve recovery of LPS from the sample. An one hour incubation of the centrifuged sample was included to allow re-association of the LPS complexes. The results listed in table 7 clearly show that centrifugation of the sample according to the method described by the invention provided a substantial improvement of recovery. For the formulation containing citrate, Tween 20 and BSA (bovine serum albumin) the improvement was approximately 22 fold (1:4 dilution) and approximately 12 fold (1:10 dilution), respectively. For the formulation containing phosphate, Tween 80 and IgG the recovery was increased approximately 52 fold (1:4 dilution) and at least 24 fold for the 1:10 dilution. Recovery level within the acceptance criteria of the LAL assay (50-200%) have not yet been achieved in this experiment. However, optimization of an incubation step on the column and an incubation step after centrifugation, especially in view of association kinetics of the LPS as well as optimizing the composition of the equilibration buffer could provide further improvements. This effort must be spent when the method is applied to real pharmaceutical formulations, as it is expected that the optimal equilibration buffer composition will be specific to some extent for different drug formulations and different drug substances.

(52) Experiment 6: Reconstitution of Various Types of LPS

(53) Formulation 2 of experiment 5 (25 mM Phosphate, 0.05% Tween 80, 60 mM NaCl, 5 mg/ml bovine IgG, pH 7.0) or endotoxin-free water containing 0.5 mM MgCl.sub.2 was spiked with about 50 EU/ml LPS from 4 different origins.

(54) 1) Escherichia coli O56:B5

(55) 2) Salmonella abortus equi

(56) 3) Salmonella enterica typhimorium

(57) 4) Pseudomonas aeruginosa

(58) Spiked samples were incubated for 10 days at room temperature (20-24° C.). Two aliquots (200 μl) of each sample were processed, essentially as described in experiment 5. Columns with a bed volume of 1.2 ml were used. The two aliquots of a sample were pooled right after centrifugation and incubated for 1 hour. The spiked water sample was not centrifuged but used to normalize the results for the different LPS materials (taken as 100%). Before endotoxin determinations using the kinetic chromogenic LAL assay (KCA), samples were diluted 1:4 with endotoxin-free water and vortexed for about 2 minutes.

(59) TABLE-US-00008 TABLE 8 Recovery of various LPS types recovery (normalized) w/o with LPS source Spiked water centrifugation centrifugation E. coli O56:B5 100%  0.64% 53.04% (48.36 EU/ml) S. abortus equi 100%  2.79% 82.72% (60.11 EU/ml) S. enterica typhimorium 100% 0.094% 37.49% (32.78 EU/ml) P. aeroginosa 100%  0.33% 29.78% (43.50 EU/ml)

(60) Different LPS types were solubilized in a matrix containing phosphate, tween 80 and IgG and incubated for 10 days at room temperature. Again, recovery of LPS was significantly reduced when using KCA as a detection method without sample preparation (between 0.094% and 2.79% compared to the control). After applying the gel filtration centrifugation step, recovery was between 29.78% and 82.72% for the different LPS types.

(61) Experiment 7: Evaluation of Additives to the Equilibration Buffer for Improving Reconstitution of Solubilized LPS

(62) A formulation as described in experiment 5 was prepared under same conditions (Formulation 1), spiked with 50 EU/ml LPS 0113 and incubated for 10 days at room temperature (20-24° C.) to solubilize the spiked LPS.

(63) Different variants of the equilibration buffer were used to prepare three centrifugation columns for each formulation.

(64) TABLE-US-00009 TABLE 9 Composition of used equilibration buffers Buffer constituents Var 1 Var 2 Var 3 Var 4 Var 5 Var 6 20 mM HEPES, pH 7.4 + + + + + + 3.125 mg/ml BSA + + + + + + 100 mM NaCl + + + + + + 50 mM MgCl2 + + + + + + 10 μM Lauryl alcohol − + − − − − 120 μM Tween 20 − − + − − − 10 μM Tween 20 − − − + − − 10 μM Polypropylenglycol − − − − + 725 10 μM SDS − − − − − +

(65) After removing the excess of equilibration buffer by centrifugation at 1,800 g, triplicates of each variant were applied to the columns. The samples were incubated on the columns for 5 minutes at room temperature prior to the second centrifugation. The eluates of the centrifugation were incubated for 60 minutes at room temperature. Afterwards a 1:4 dilution of each sample was prepared in glass vials with 2 minutes of intensive vortexing (1,400 rpm). Samples were tested using the LAL assay (KCA).

(66) TABLE-US-00010 TABLE 10 Improvement of recovery applying additives to the equilibration buffer: Endotoxin value Formulation Variant (EU/ml and % recovery) Variant 1 (Control) 6.07 EU/ml (48.56% recovery) Variant 2 (10 μM Lauryl alcohol) 8.61 EU/ml (68.88% recovery) Variant 3 (120 μM Tween 20) 0.63 EU/ml (0.48% recovery) Variant 4 (10 μM Tween 20) 9.25 EU/ml (74.00% recovery) Variant 5 (10 μM Polypropylenglycol 725) 7.80 EU/ml (62.40% recovery) Variant 6 (10 μM SDS) 6.83 EU/ml (54.64% recovery)

(67) Selected amphiphilic substances were added to the equilibration buffer in order to show, that reconstitution of LPS to complexes during or after the separation by gel filtration can be improved by supplementation of the equilibration buffer. Compared to variant 1 (control), the addition of lauryl alcohol (Var.2), Tween 20 (Var.4), Polypropylenglycol 725 (Var.5) and SDS (Var.6) when used well below their critical micelle concentration (cmc), provided equal or increased LPS recovery. Only Tween 20 (Var. 3) used at twice the critical micelle concentration (cmc) provided very low recovery compared to the control. This experiment is indicative that further principles, besides high magnesium ion concentration, when added to the equilibration buffer could improve the overall recovery of the process.

(68) The invention has been described with reference to preferred embodiments. To the expert it is also conceivable, however, to make changes and modifications without leaving the scope of protection of the appended claims.

LIST OF REFERENCE NUMBERS

(69) 1 column 2 endotoxin-free centrifugation column 3 centrifugation container 4 Sarstedt tube 5 size exclusion chromatography matrix; gel matrix 5* equilibrated gel matrix 5** gel matrix buffered with the low molecular weight components of the formulation 6 equilibration buffer 6* excess equilibration buffer 7 collection tube 9 user manual 10 sample 11 formulation 15 flow through 20 sample preparation kit 22 endotoxin sample