METHODS FOR DECONTAMINATING CIRCUITS FOR PRODUCING GLUCOSE POLYMERS AND HYDROLYSATES OF GLUCOSE POLYMERS

Abstract

The present invention concerns a method for determining the impact of a production step or a purification step on the presence or nature of pro-inflammatory contaminating molecules in glucose polymers or the hydrolysates of same by using an in vitro test of inflammatory response using cell lines. It further concerns an optimised method of producing or purifying glucose polymers or the hydrolysates of same comprising an analysis of the pro-inflammatory contaminating molecules in glucose polymers or the hydrolysates of same and the selection of production or purification steps optimised with respect to the presence and nature of the pro-inflammatory contaminating molecules.

Claims

1-13. (canceled)

14. A method for testing the effectiveness of a purification step or purification steps on the presence of pro-inflammatory molecules in a final preparation of glucose polymers or hydrolysates thereof, the method comprising: a) providing an initial preparation of glucose polymers or hydrolysates thereof, the initial preparation containing pro-inflammatory molecules; b) detecting or assaying the amount of pro-inflammatory molecules in the initial preparation of glucose polymers or hydrolysates thereof provided in step a); c) carrying out the purification step or purification steps on the initial preparation of glucose polymers or hydrolysates thereof provided in step a) to produce the final preparation of glucose polymers or hydrolysates thereof; d) detecting or assaying the amount of pro-inflammatory molecules in the final preparation of glucose polymers or hydrolysates thereof; e) comparing the amount of pro-inflammatory molecules in the initial preparation of glucose polymers or hydrolysates thereof detected or assayed in step b) with the amount of pro-inflammatory molecules in the final preparation of glucose polymers or hydrolysates thereof detected or assayed in step d); and f) identifying the production step or production steps or the purification step or purification steps as: effective if the amount of pro-inflammatory molecules in the final preparation of glucose polymers or hydrolysates thereof is lower than the amount of pro-inflammatory molecules in the initial preparation of glucose polymers or hydrolysates thereof, or not effective if the amount of pro-inflammatory molecules in the final preparation of glucose polymers or hydrolysates thereof is not lower than the amount of pro-inflammatory molecules in the initial preparation of glucose polymers or hydrolysates thereof; g) selecting the purification step or purification steps to implement in the method for producing or purifying glucose polymers or hydrolysates thereof; wherein the steps for detecting or assaying the pro-inflammatory molecules in the initial and final preparations of glucose polymers or hydrolysates thereof of steps b) and d) comprise an in vitro inflammatory response test, the test comprising the steps of: i) contacting the initial or the final preparation of glucose polymers or hydrolysates thereof with a cell line expressing a TLR2 receptor and a reporter gene, wherein the transcription of the reporter gene is under the control of the TLR2 signaling pathways, ii) measuring the activity or the signal of the reporter gene of the preparation mentioned in step i), iii) contacting the initial or the final preparation of glucose polymers or hydrolysates thereof with a control line not transfected with an immunity receptor, iv) measuring the activity or the signal of the reporter gene of the preparation mentioned in step iii), and v) verifying that the activity or the signal measured from steps ii) and iv) is not induced by the transcription of the reporter gene via a parasitic mechanism.

15. The method of claim 14, wherein the in vitro inflammatory response test further comprises contacting the initial or the final preparation of glucose polymers or hydrolysates thereof with: a) an MDP- or LPS-sensitized, macrophage-differentiated THP-1 cell line, the pro-inflammatory molecules being detected or assayed by measuring the amount of RANTES or TNF- produced by the cell line; and/or b) a macrophage line transfected with a reporter gene, the transcription of which is under the direct control of the inflammatory signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; and/or c) a cell line expressing the NOD2 receptor and a reporter gene, the transcription of which is under the direct control of the NOD2 signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; and/or d) a cell line expressing the TLR4 receptor and a reporter gene, the transcription of which is under the direct control of the TLR4 signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene.

16. The method of claim 14, wherein the in vitro inflammatory response test comprises contacting the initial or the final preparation of the glucose polymers or hydrolysates thereof with: a) a macrophage line transfected with a reporter gene, the transcription of which is under the direct control of the inflammatory signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; b) a cell line expressing the TLR2 receptor and a reporter gene, the transcription of which is under the direct control of the TLR2 signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; c) a cell line expressing the TLR4 receptor and a reporter gene, the transcription of which is under the direct control of the TLR4 signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; d) a cell line expressing the NOD2 receptor and a reporter gene, the transcription of which is under the direct control of the NOD2 signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; and e) a control line not transfected with an immunity receptor.

17. The method of claim 14, wherein the pro-inflammatory molecules are molecules of bacterial origin.

18. The method of claim 14, wherein the production or purification step or steps is or are chosen from steps of heat treatment, of acidification, of passing over activated carbon, of passing over adsorption resins, of ultrafiltration, of filtration, or of chemical or enzymatic hydrolysis, or combinations thereof.

19. The method of claim 14, wherein the glucose polymers are selected from icodextrin and branched or unbranched maltodextrins, and the glucose polymer hydrolysates are a product of total hydrolysis.

20. The method of claim 14, wherein, before-the steps b) and d) for detecting or assaying pro-inflammatory molecules in the initial and the final preparation of glucose polymers or hydrolysates thereof, the initial and the final preparation of glucose polymers or hydrolysates thereof is prefiltered with a cut-off threshold at 30 kDa to provide a filtrate of the initial and final preparations, and the tests of steps b) and d) are carried out on the filtrate of the initial and final preparations.

21. The method of claim 17, wherein said pro-inflammatory molecules of bacterial origin are peptidoglycans (PGN), lipopolysaccharides (LPS) lipopeptides, PGN depolymerization products, muramyl dipeptide (MDP), formylated microbial peptides, formyl-Met-Leu-Phe tripeptide (f-MLP) or -glucans.

22. A method for producing or purifying glucose polymers or hydrolysates thereof, the method comprising: a) providing said glucose polymers or hydrolysates thereof; b) detecting or assaying pro-inflammatory molecules in the glucose polymers or hydrolysates thereof provided in step a); c) selecting the step or steps for producing or purifying said glucose polymers or hydrolysates thereof for separating the pro-inflammatory molecules present in the glucose polymers or hydrolysates thereof detected in step b) from said glucose polymers or hydrolysates thereof; d) optionally, carrying out the selected production or purification step or steps on the glucose polymers or hydrolysates thereof provided in step a); and e) optionally, detecting or assaying pro-inflammatory molecules in the glucose polymers or hydrolysates thereof obtained after step d); in which the steps for detecting or assaying the pro-inflammatory molecules in the glucose polymers or hydrolysates thereof of steps b) and e) comprise an in vitro inflammatory response test that comprises: (i) contacting the glucose polymers or hydrolysates thereof with a cell line expressing the TLR2 receptor and a reporter gene, the transcription of which is under the control of the TLR2 signaling pathways, (ii) measuring the activity or the signal of the reporter gene of the preparation mentioned in step i) (iii) contacting the initial or the final preparation of glucose polymers or hydrolysates thereof with a control line and not transfected with an immunity receptor (iv) measuring the activity or the signal of the reporter gene of the preparation mentioned in step iii), and (v) verifying that the activity or the signal measured from steps ii) and iv) is not induced by the transcription of the reporter gene via a parasitic mechanism.

23. The method of claim 22, wherein the in vitro inflammatory response test further comprises bringing the glucose polymers or hydrolysates thereof into contact with: a) an MDP- or LPS-sensitized, macrophage-differentiated THP-1 cell line, the pro-inflammatory molecules being detected or assayed by measuring the amount of RANTES or TNF- produced by the cell line; and/or b) a macrophage line transfected with a reporter gene, the transcription of which is under the direct control of the inflammatory signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; and/or c) a cell line expressing the NOD2 receptor and a reporter gene, the transcription of which is under the direct control of the NOD2 signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; and/or d) a cell line expressing the TLR4 receptor and a reporter gene, the transcription of which is under the direct control of the TLR4 signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter.

24. The method of claim 22, wherein the in vitro inflammatory response test comprises bringing the glucose polymers or hydrolysates thereof into contact with: a) a macrophage line transfected with a reporter gene, the transcription of which is under the direct control of the inflammatory signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; b) a cell line expressing the TLR2 receptor and a reporter gene, the transcription of which is under the direct control of the TLR2 signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; c) a cell line expressing the TLR4 receptor and a reporter gene, the transcription of which is under the direct control of the TLR4 signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; d) a cell line expressing the NOD2 receptor and a reporter gene, the transcription of which is under the direct control of the NOD2 signaling pathways, the pro-inflammatory molecules being detected or assayed by measuring the activity or the signal of the reporter gene; and e) a control line not transfected with an immunity receptor.

25. The method of claim 22, wherein the pro-inflammatory molecules are molecules of bacterial origin.

26. The method of claim 22, wherein the production or purification step or steps is or are chosen from steps of heat treatment, of acidification, of passing over activated carbon, of passing over adsorption resins, of ultrafiltration, of filtration, or of chemical or enzymatic hydrolysis, or combinations thereof.

27. The method of claim 22, wherein the glucose polymers are selected from icodextrin and branched or unbranched maltodextrins, and the glucose polymer hydrolysates are a product of total hydrolysis.

28. The method of claim 22, wherein, before steps b) and e) for detecting or assaying pro-inflammatory molecules in the glucose polymers or hydrolysates thereof, the samples of glucose polymers or hydrolysates thereof are prefiltered with a cut-off threshold at 30 kDa to provide a filtrate of the initial and final preparations, and the tests of steps b) and e) are carried out on the filtrate of the initial and final preparations.

29. The method of claim 25, wherein said pro-inflammatory molecules of bacterial origin are peptidoglycans (PGN), lipopolysaccharides (LPS) lipopeptides, PGN depolymerization products, muramyl dipeptide (MDP), formylated microbial peptides, formyl-Met-Leu-Phe tripeptide (f-MLP) or -glucans.

30. The method of claim 14, wherein said pro-inflammatory molecules are peptidoglycans of bacterial origin.

31. The method of claim 22, wherein said pro-inflammatory molecules are peptidoglycans of bacterial origin.

32. The method of claim 14, wherein the cell line expressing the TLR2 receptor is transfected with the TLR2 receptor gene and the reporter gene.

33. The method of claim 22, wherein the cell line expressing the TLR2 receptor is transfected with the TLR2 receptor gene and the reporter gene.

34. A method for testing the effectiveness of a production step or production steps or of a purification step or purification steps on the presence of pro-inflammatory molecules in a final preparation of glucose polymers or hydrolysates thereof, the method comprising: a) providing an initial preparation of glucose polymers or hydrolysates thereof, the initial preparation containing pro-inflammatory molecules; b) detecting or assaying the amount of pro-inflammatory molecules in the initial preparation of glucose polymers or hydrolysates thereof provided in step a); c) carrying out the production or production step or production steps or the purification step or purification steps on the initial preparation of glucose polymers or hydrolysates thereof provided in step a) to produce the final preparation of glucose polymers or hydrolysates thereof; d) detecting or assaying the amount of pro-inflammatory molecules in the final preparation of glucose polymers or hydrolysates thereof; e) comparing the amount of pro-inflammatory molecules in the initial preparation of glucose polymers or hydrolysates thereof detected or assayed in step b) with the amount of pro-inflammatory molecules in the final preparation of glucose polymers or hydrolysates thereof detected or assayed in step d); and identifying the production step or production steps or the purification step or purification steps as: effective if the amount of pro-inflammatory molecules in the final preparation of glucose polymers or hydrolysates thereof is lower than the amount of pro-inflammatory molecules in the initial preparation of glucose polymers or hydrolysates thereof, or not effective if the amount of pro-inflammatory molecules in the final preparation of glucose polymers or hydrolysates thereof is not lower than the amount of pro-inflammatory molecules in the initial preparation of glucose polymers or hydrolysates thereof; wherein the steps for detecting or assaying the pro-inflammatory molecules in the initial and final preparations of glucose polymers or hydrolysates thereof of steps b) and d) comprise an in vitro inflammatory response test, the test comprising the steps of: i) contacting the initial or the final preparation of glucose polymers or hydrolysates thereof with a cell line expressing a TLR2 receptor and a reporter gene, wherein the transcription of the reporter gene is under the control of the TLR2 signaling pathways, ii) measuring the activity or the signal of the reporter gene of the preparation mentioned in step i), iii) contacting the initial or the final preparation of glucose polymers or hydrolysates thereof with a control line not transfected with an immunity receptor, iv) measuring the activity or the signal of the reporter gene of the preparation mentioned in step iii), and v) verifying that the activity or the signal measured from steps ii) and iv) is induced by the transfected immunity receptor; wherein before-the steps b) and d) for detecting or assaying pro-inflammatory molecules in the initial and the final preparation of glucose polymers or hydrolysates thereof, the initial and the final preparation of glucose polymers or hydrolysates thereof is prefiltered with a cut-off threshold at 30 kDa to provide a filtrate of the initial and final preparations, and the tests of steps b) and d) are carried out on the filtrate of the initial and final preparations; and wherein said pro-inflammatory molecules are peptidoglycans of bacterial origin.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0200] FIG. 1: Raw-Blue cell responses to standard agonists.

[0201] FIG. 2: HEK-Blue TLR2 cell responses to standard agonists.

[0202] FIG. 3: HEK-Blue TLR4 cell responses to standard agonists.

[0203] FIG. 4: HEK-Blue NOD2 cell responses to standard agonists.

[0204] FIG. 5: HEK-Blue Null cell responses to standard agonists.

[0205] FIG. 6: Raw-Blue cell responses induced by nonfiltered matrices and after passing over 100 kDa and 30 kDa filters.

[0206] FIG. 7: HEK-Blue TLR2 cell responses induced by nonfiltered matrices and after passing over 100 kDa and 30 kDa filters.

[0207] FIG. 8: HEK-Blue TLR4 cell responses induced by nonfiltered matrices and after passing over 100 kDa and 30 kDa filters.

[0208] FIG. 9: HEK-Blue NOD2 cell responses induced by nonfiltered matrices and after passing over 100 kDa and 30 kDa filters.

[0209] FIG. 10: HEK-Blue Null cell responses induced by the matrices.

[0210] FIG. 11: Assessment of the inflammatory activities of the various matrices.

[0211] FIG. 12: Cell responses induced by the matrices after passing over carbon SX+.

[0212] FIG. 13: Cell responses induced by the E3063 matrix after passing over various carbons.

[0213] FIG. 14: Cell responses induced by the E1565 matrix after passing over various carbons.

[0214] FIG. 15: Cell responses induced by the E1242 matrix after passing over various carbons.

[0215] FIG. 16: Cell responses induced by the Lab3943 matrix after passing over various carbons.

[0216] FIG. 17: Cell responses induced by the E1565 matrix after treatment by ultrafiltration on 5 kDa.

[0217] FIG. 18: Cell responses induced by the Lab3943 matrix after treatment by ultrafiltration on 5 kDa.

[0218] FIG. 19: Cell responses induced by the E3063 and E5250 matrices after treatment with Mannaway.

[0219] FIG. 20: Cell responses induced by the E5250 matrix after treatment with SEBflo TL.

[0220] FIG. 21: Cell responses induced by the E3063 matrix after treatment with SEBPro FL100.

[0221] FIG. 22: Cell responses induced by the E1565 matrix after treatment with industrial resins.

EXAMPLES

Example 1: Establishment of the Dose-Response Curves

[0222] The dose-response curves are produced with standard agonist molecules: LPS, PGN, LTA, zymosan and MDP. The Raw-Blue and HEK-Blue TLR2, TLR4, NOD2 and Null lines are incubated with increasing concentrations of agonists, and the cell response is measured by quantifying the SEAP activity (FIGS. 1-5). TNF- is used as a positive control for cell activation: [0223] Raw-Blue line: the cells respond to the major inflammatory molecules that may be present in the glucose polymer matrices and derivatives (PGN, LPS, zymosan, LTA); they in particular have a strong reactivity with respect to PGNs, but do not respond to their depolymerization products (MDP). [0224] HEK-Blue hTLR2 line: strong reactivity with respect to PGNs; the cells respond more weakly to the other TLR2 ligands (LTA, zymosan) and show no reactivity with respect to LPSs and to MDP. [0225] HEK-Blue hTLR4 line: strong reactivity with respect to LPSs; the cells respond very weakly to zymosan and show no reactivity with respect to PGN, LTA and MDP. [0226] HEK-Blue hNOD2 line: strong reactivity with respect to MDP. [0227] HEK-Blue Null2 line: control for absence of cell toxicity.

Example 2: Preparation of the Various Distinct Glucose Polymer Matrices and of a Batch of Glucose Polymer Hydrolysate

[0228] As indicated above, the matrices are the following: [0229] 5 glucose polymers, raw materials of icodextrin (before chromatographic fractionation according to the teaching of patent EP 667 356), referenced here E1565, E3063, E1242, E5248 and E5250.

[0230] The preparation of these five polymers is carried out in accordance with the teachings of patent application WO 2012/059685; [0231] a contaminated batch of icodextrin (referenced here E209J) and a standard batch of icodextrin, i.e. control for non-contamination in the cell tests (referenced here P11-11). These batches are prepared according to the teaching of patent EP 667 356, described in detail in Example 1 of patent application WO 2010/125315; [0232] a batch of branched maltodextrin, sold by the applicant company under the brand name NUTRIOSE FB06; [0233] a batch of dextrose monohydrate, prepared so as to be conditioned in an injectable solution, sold by the applicant company under the brand name LYCADEX PF; [0234] a batch of highly-branched, soluble glucose polymers, for peritoneal dialysis, referenced here LAB3943.

[0235] This batch is prepared by double enzymatic treatment with branching enzyme and amyloglucosidase according to Example 2 of patent application WO 2007/099212. [0236] A commercial maltodextrin (Cargill maltodextrin, C*Dry MD 01915, batch 02044770), referenced here Cargill.

Example 3: Analysis of the Cell Responses Induced by the Samples which are Nontreated or after Passing Over 100 kDa or 30 kDa Filter

[0237] The objective of these tests is to determine the pro-inflammatory reactivity and the nature of the contaminants present in the glucose polymer matrices and the batch of glucose polymer hydrolysate.

[0238] The samples according to Example 2 are prepared at 32% (weight/volume) in non-pyrogenic water (for injection).

[0239] The assays of the LPS and PGN levels were carried out prior to the cell tests using the SLP-HS and LAL assays (data presented below):

TABLE-US-00001 Lab Ico P11-11 E1242 E1565 E3063 E5250 3943 E209J Cargill NUTRIOSE LYCADEX PGN SLP- <3 21 2320 16185 4496 1263 393 2478 315 <2 (ng/g) HS LPS LAL <0.3 2.4 38.4 2.4 19.2 153.6 0.6 9.6 >300 <0.15 (EU/g) LPS LAL <0.3 1.2 4.8 1.2 <0.3 153.6 <0.3 <0.3 >300 / (EU/g) modifi

[0240] For the cell tests, the samples are diluted to 1/10 in the cell culture medium (final concentration: 3.2% (w/v)).

[0241] The analyses are carried out on: [0242] Raw-Blue line: any contaminants with high reactivity for PGNs, [0243] HEK-Blue hTLR2 line: high reactivity for PGNs, [0244] HEK-Blue hTLR4 line: high reactivity for LPSs, [0245] HEK-Blue hNOD2 line: MDP and PGN depolymerization products, [0246] HEK-Blue Null2 line: control for absence of cell toxicity.

[0247] The results by cell type are presented in FIGS. 6 to 10.

Raw Cell Responses (FIG. 6):

[0248] With the exception of the Cargill matrix, which gives a response equivalent to that observed in the presence of the noncontamination control P11-11, all the other samples trigger an inflammatory response on contact with the macrophage line. The most reactive matrices are E-3063 (saturation of the cell response), then E-1242 and E-1565.

[0249] The contaminants are essentially molecules of high molecular weight (for example, PGN, zymosan) or capable of forming aggregates (for example, LPS, LTA). Indeed, the filtration at 100 kDa greatly reduced the responses induced by the samples, indicating that this treatment removed them to a large extent. Only the E-1242 and E-5250 matrices still have an activity significantly higher than that of P11-11, indicating that they contain contaminants having a size <100 kDa, probably originating from the degradation of larger contaminants. The filtration at 30 kDa is even more effective since the various samples lose virtually all their pro-inflammatory activity after treatment.

HEK-TLR2 Cell Responses (FIG. 7):

[0250] The results obtained with the HEK-TLR2 cells confirm the previous results.

[0251] The E-3063 matrix induces a saturated response, thereby indicating a very high TLR2 inducer contamination level. The E-1242, E-1565, Lab3943 and Ico-E209J matrices also give high responses, greater than those observed in the Raw cells. This difference is explained by the fact that they are loaded with strong inducers of TLR2 (PGNs or lipopeptides).

[0252] The filtrations at 100 kDa and at 30 kDa neutralize the inflammatory responses induced by these samples, indicating that the contaminants are predominantly high-molecular-weight PGNs. However, the E-3063 and E-1565 matrices still exhibit a significant activity after filtration. These data show that these compounds contain PGN degradation products and/or lipopeptides. Indeed, contrary to PGNs, the other strong inducers of TLR2 have a weight <30 kDa and would therefore still give a cell response after filtration.

[0253] The E-5250 and E-5248 matrices and the NUTRIOSE give weak responses, of strength equivalent to that observed with the Raw cells, suggesting that these three samples contain weak inducers of TLR2 (for example, zymosan, (3-glucans or LTA).

[0254] As previously, the Cargill matrix and the LYCADEX do not trigger responses, indicating the absence of TLR2-inducing contaminants.

HEK-TLR4 Cell Responses (FIG. 8):

[0255] The E1565, E3063, Lab3943, Cargill and NUTRIOSE matrices trigger a response of medium strength in the HEK-TLR4 cells, thereby confirming the presence of LPS. The filtrations partly reduce the responses of the cells, which can be explained by the fact that LPS can form aggregates, and that only the nonaggregated molecules were removed.

[0256] The E1242, IcoE209J, E5248, E5250 and LYCADEX matrices do not trigger a significant response, indicating that the LPS levels are below the thresholds capable of triggering an inflammatory response. The first two matrices give an inflammatory response in the Raw cells, which correlates with a strong reactivity in the HEK-TLR2 cells. These data indicate that these two matrices are essentially contaminated with PGNs. The E5248 and E5250 matrices and the LYCADEX also trigger an inflammatory response in the Raw cells. However, they are only barely or not at all active with respect to TLR2, thereby showing the presence of contaminants other than PGNs and LPSs in these three samples.

HEK-NOD2 Cell Responses (FIG. 9):

[0257] The HEK-NOD2 cells respond to all the samples, but only the E-1565, Lab3943 and Cargill matrices are strongly loaded with inducers of NOD2. This receptor reacts to the final product of PGN depolymerization (MDP), but also to the low-molecular-weight degradation products thereof. Consequently, the strength of the responses observed indicates that the samples are and/or were contaminated with PGN having undergone a more or less advanced degradation process. As expected, the filtrations at 100 and 30 kDa do not have a significant effect on the response of the cells, since the compounds in question (MDP and degraded PGNs) are of small size.

HEK-Null Cell Responses (FIG. 10):

[0258] Finally, the HEK-Null cells do not give significant responses in the presence of the various samples, proof that the reactivities observed in the other cell lines are not linked to a toxic effect, but indeed to a response of inflammatory type.

Assessment by Sample (FIG. 11):

[0259] E1242: medium-strength inflammatory activity linked to a strong contamination with slightly degraded PGNs. [0260] E1565: medium-strength inflammatory activity linked to a strong contamination with partially degraded PGNs and to the presence of LPSs. [0261] E3063: high-strength inflammatory activity linked to a very strong contamination with slightly degraded PGNs and to traces of LPSs. [0262] E5248: low-strength inflammatory activity linked to a weak contamination with PGNs and to the presence of inflammatory molecules other than PGNs and LPSs. [0263] E5250: low-strength inflammatory activity linked to a weak contamination with PGNs and to the presence of inflammatory molecules other than PGNs and LPSs. [0264] Lab3943: medium-strength inflammatory activity linked to medium contaminations with partially degraded PGNs and with LPSs. [0265] IcoE209J: medium-strength inflammatory activity linked to a strong contamination with slightly degraded PGN. [0266] Cargill: absence of detectable inflammatory activity, but presence of PGN degradation products. [0267] NUTRIOSE: medium-strength inflammatory activity linked to traces of partially degraded PGNs and to a medium contamination with LPSs. [0268] LYCADEX: low-strength inflammatory activity linked to a weak contamination with PGN degradation products and to the presence of inflammatory molecules other than PGNs and LPSs.

Example 4: Effect of the Treatments by Passing Over Activated Carbons

[0269] In a first series of experiments, all the samples were subjected to two successive treatments with the same carbon (1% of NORIT SX+ carbon, at 80 C. for 1 h).

[0270] FIG. 12 presents the results obtained by cell line.

[0271] The first carbon treatment drastically decreases the capacity of the E1242, E-565, E3063 and IcoE209J samples to trigger an inflammatory response in the Raw and HEK-TLR2 cells. In any event, the second treatment further improves the removal of the molecules responsible for the inflammatory response. The effect of the treatment is much less marked for the E5248, E5250, and Lab3943 samples and the NUTRIOSE. These data indicate that the treatment with the SX+ carbon would be effective for removing the barely degraded PGNs, thus reducing the inflammatory activity of the matrices for which these contaminants are predominant.

[0272] For the HEK-TLR4 cells, the response is very reduced for the E1565, Lab3943 and NUTRIOSE matrices, which are the most contaminated with LPSs. However, the effect is not as marked as for the responses attributed to PGNs, showing kinetics specific to LPS.

[0273] The treatments with the SX+ carbon have little effect on the response of the HEK-NOD2 cells, thereby indicating that the PGN degradation products are not correctly removed. Finally, the HEK-Null cells are not reactive with respect to the samples treated, excluding any toxic effect associated with the carbon.

[0274] The differences observed after treatment by passing over activated carbon were expected, since the previous tests show that the samples contain contaminants of different molecular nature. An important piece of information provided by this experiment is the demonstration of a difference in effectiveness according to the size of the contaminants. Thus, the treatment with SX+ carbon would have a more marked effect on the removal of a certain category of contaminants, in particular high-molecular-weight PGNs.

[0275] In order to confirm this hypothesis, several carbons of different porosity were tested for their effectiveness in decontaminating the E3063, E1565, E1242 and Lab3943 matrices.

[0276] Contrary to the E3063 and E1242 matrices which are predominantly contaminated with PGNs of large size (strong TLR2 response), the E1565 and Lab3943 matrices contain partially degraded PGNs (TLR2 and NOD2 responses) and LPSs (TLR4 response).

[0277] The optimized treatment conditions are the following: carbon at 0.5%, pH adjusted to 4.5, incubation for 1 h at 80 C. After treatment, the samples are filtered through 0.22 m filter and then used in the cell tests.

[0278] The results for the E3063 matrix are presented in FIG. 13.

[0279] The absence of response in the HEK-Null cells confirms that none of the carbons exhibit cell toxicity.

[0280] All the carbons tested are effective for drastically reducing the Raw and HEK-TLR2 cell responses, thereby attesting to their effectiveness for removing the large PGNs, which are the main contaminants of this matrix.

[0281] It is noted that the new carbons are equivalent to or more effective than the SX+.

[0282] Thus, the samples treated with L4S, L3S, ENO-PC, C-extra USP and SX2 induce cell responses close to the background noise in the HEK-TLR2 cells, thereby suggesting that these carbons are more effective for removing PGNs.

[0283] Regarding the less effective carbons, the filtration on 30 kDa and 100 kDa reduces the residual responses after treatment, which is proof that they were due essentially to traces of unremoved PGNs.

[0284] The same results are observed in the Raw cells, with the exception of L4S, which appears to be less effective, and A-Supra-Eur, which, on the contrary, is found to be more effective than in the HEK-TLR2 cells. The latter carbon could therefore have a broader spectrum of action and remove the other molecules, such as LPSs.

[0285] Even though the carbons significantly reduce the reactivity of the HEK-NOD2 and HEK-TLR4 cells with respect to the E3063 matrix, it is difficult to distinguish differences in effectiveness between the treatments because of the small size of the responses induced by this matrix in the two cell types.

[0286] The results for the E1565 matrix are presented in FIG. 14.

[0287] As for the E3063 matrix, the same carbons are effective for reducing the HEK-TLR2 and Raw cell responses induced by E1565. However, a few minor differences can be noted, probably due to differences in sizes and therefore in properties of the PGNs.

[0288] Virtually the entire HEK-NOD2 cell response is clearly due to the presence of PGN degradation products, since the filtration at 30 kDa retains the contaminants only very slightly or not at all. On the other hand, the carbons are not very effective for reducing the response of these cells. Indeed, the decrease in the response caused by the reduction in the load of contaminants of MDP type does not exceed 50% of the maximum response of the cells.

[0289] Finally, the carbons have a medium effect on the TLR4 response. With the exception of SX+ which is very effective for removing all the forms of LPS, the reduction caused by the treatment with other carbons does not exceed 50% of the response induced by the same sample which has not been treated. However, it can be noted that the L4S and A-Supra-Eur carbons, and to a lesser extent the C-extra USP and ENO-PC carbons, are more effective for reducing the responses induced by molecules of size <100 kDa, thereby suggesting that these carbons preferentially act on nonaggregated LPSs.

[0290] The results for the E1242 matrix are presented in FIG. 15.

[0291] All the carbons tested are effective for reducing the responses induced by the E1242 matrix in the Raw and HEK-TLR2 cells. The responses obtained, which are close or equal to the background noise, are identical before and after filtration on 30 kDa and 100 kDa, which is proof that the large molecules corresponding to PGN have been removed.

[0292] The E1242 matrix is very weakly contaminated with LPS. However, it can be noted that ENO-PC and A-Supra-Eur are more effective than the other carbons for decreasing the HEK-TLR4 cell response to the level of the background noise. This observation confirms that these two carbons have a broad spectrum of action and are effective for removing molecules other than PGNs, such as LPSs.

[0293] Finally, the carbons are not effective for removing PGN degradation products, with the exception of ENO-PC and SX2, which reduce the HEK-NOD2 cell response by approximately 50%.

[0294] The results for the Lab3943 matrix are presented in FIG. 16.

[0295] This matrix is contaminated with a broad spectrum of different molecules. As expected, all the carbons reduce the responses induced in the Raw cells, but with different effectivenesses. It can be noted that SX+, CSA, L4S and, to a lesser extent, A-Supra-Eur are the most effective, thereby confirming that these carbons have a broad spectrum of action. For the HEK-TLR2 cells, the carbons are all found to be effective for removing PGNs, but the residual responses remain identical before and after filtration, indicating that the PGN degradation products are more difficult to remove.

[0296] The behavior of the carbons in the removal of LPSs is also variable. Thus, the SX+, ENO-PC, L4S and A-Supra-Eur carbons are still the most effective for reducing the HEK-TLR4 cell response. Finally, only the ENO-PC, C-Extra-USP and SX2 carbons are found to be relatively active for strongly reducing the HEK-NOD2 cell response, which is proof that they are effective for removing the PGN degradation products present in this matrix. [0297] C-extra-USP and SX2: effective for removing PGNs and the degradation products thereof. [0298] A-Supra-Eur: broad spectrum with higher effectiveness for high-molecular-weight molecules (for example: aggregated LPSs and PGNs). [0299] ENO-PC: broad spectrum with higher effectiveness for molecules having a molecular weight <100 kDa (for example: LPSs and PGN degradation products). [0300] other carbons: spectra of action and effectiveness at most equivalent to those of the SX+ carbon.

Example 5: Effect of a Treatment by Ultrafiltration on 5 kDa

[0301] The objective of the treatment by ultrafiltration is to reduce, or even remove, the contamination with molecules of small size, so as to counter their participation in the triggering of an inflammatory response, whether it is via a direct effect or via a phenomenon of synergy with other contaminating molecules.

[0302] The experiments were carried out on the E1565 and Lab3943 matrices, which are both contaminated with partially degraded PGNs (TLR2 and NOD2 responses) and LPSs (TLR4 response).

[0303] The filtration on 5 kDa was carried out at an average flow rate of 25 ml/min. The filtrate flow rates are respectively 55 ml/h for E1565 and 65 ml/h for Lab3943.

[0304] In order to verify the effectiveness of the ultrafiltration, the cell responses were first measured using samples originating from the retentate and filtrate fractions recovered after passage of the starting solution (100 ml).

[0305] The ultrafiltration was then carried out in closed circuit with continuous injection of the retentate into the starting sample. In order to compensate for the loss of liquid due to the removal of the filtrate, the volume of the sample was continuously adjusted to the starting volume by adding water for injection. In this case, the cell tests were carried out using specimens taken from the sample after 1 h, 2 h and 3 h of ultrafiltration.

[0306] The results for the E1565 matrix are presented in FIG. 17.

[0307] The responses induced by the retentate fractions remain similar to those observed for the nonfiltered samples in the four cell tests. However, a significant inflammatory response is observed in response to the filtrate fractions in the Raw and HEK-NOD2 cells. These data are compatible with the cut-off threshold of the filter (5 kDa), which allows the PGN depolymerization products to pass, but not the PGNs and the LPS, especially if the latter are in the form of aggregates. On the other hand, the absence of a decrease in inflammatory response in the retentates indicates that a single passage through the filter is ineffective for reducing the inflammatory reactivity of the matrix. This is because the division between the retentate/filtrate fractions is 25 to 1, which is insufficient to remove the small inflammatory molecules.

[0308] The continuous ultrafiltration is effective for decreasing the response induced in the HEK-NOD2 cells, which was predictable, but also in the Raw and HEK-TLR2 cells. Given the contamination of E1565 with PGNs and LPS, the decrease in response in the latter two cell types is certainly linked to a reduction in the synergistic activity of the small inflammatory molecules.

[0309] The results for the Lab3943 matrix are presented in FIG. 18.

[0310] As expected, an inflammatory activity associated with MDP is found in the filtrate (HEK-NOD2). A significant response is also observed in the HEK-TLR4 cells. This result shows that the LPS is in a structural configuration that is less aggregated than that present in the E1565 matrix, thereby allowing it to pass through the filter.

[0311] As for E1565, the continuous filtration is effective for decreasing the PGN depolymerization product load, which is visible through a clear reduction in the HEK-NOD2 cell response (direct effect) and through a smaller but significant decrease in the reactivity of the Raw and HEK-TLR2 cells (synergistic action).

Example 6: Effect of the Enzymatic Treatments

[0312] The objective of these tests is to test the capacity of industrial enzymes to decrease the pro-inflammatory reactivity of the contaminants present in glucose polymer matrices.

[0313] The samples are prepared at 32% (weight/volume) and treated in the presence of the enzymes according to the conditions described hereinafter. After treatment, the enzymes are deactivated by heating, and the solutions are filtered through a sterile 0.22 m filter and then used in the cell tests.

[0314] Three industrial enzymatic preparations were tested for their capacity to reduce the contaminant load: [0315] Mannaway: incubation at 0.4% (vol/vol), pH 10, 50 C., for 4 h and 24 h. [0316] SEBflo TL: incubation at 0.35 mg/g of matrix, 50 C., pH 5, from 30 min to 24 h. [0317] SEBPro FL100: incubation at 4% (vol/vol), pH 3, 55 C., from 1 h to 24 h.

[0318] The two matrices chosen for the tests are: E3063 (strong contamination with slightly degraded PGNs and traces of LPS) and E5250 (weak PGN contamination and presence of inflammatory molecules other than PGNs and LPSs).

[0319] The effects of the treatments of the two matrices with Mannaway are presented in FIG. 19.

[0320] The addition of the enzymatic solution to the P-11.11 matrix (noncontamination control) induces no inflammatory response in the Raw, HEK-TLR2 or HEK-TLR4 cells. A slight increase is observed in the HEK-NOD2 cells, suggesting the presence of PGN degradation products in trace amounts. However, these results show that the enzyme used under the conditions in the experiment does not introduce major pro-inflammatory contaminants.

[0321] The tests carried out on the E3063 matrix show that the enzyme probably has a weak lytic action on PGNs, since a decrease in the responses in the Raw and HEK-TLR2 cells is observed. The treatment causes an increase in the HEK-NOD2 cell response, which is compatible with a partial degradation of PGNs. On the other hand, an increase in the HEK-TLR4 cell response is also observed. Given that the enzymatic preparation does not introduce any contamination, the appearance of LPS is probably due to a release of this contaminant from the matrix itself.

[0322] The enzymatic treatment of the E5250 matrix induces an increase in the inflammatory response in the four cell types. Originally, this matrix triggered weak inflammatory responses. Consequently, the increase in the TLR2 and TLR4 responses suggests that the enzyme released contaminants of PGN and LPS type from the matrix.

[0323] Mannaway is commonly used as an agent for clarifying food-processing products. The results obtained suggest that the enzyme probably dissociated macrocomplexes (bacterial debris) which were normally removed by the step of filtration through a 0.22 m filter. By solubilizing these inflammatory molecules, the enzyme made them accessible for inducing responses in the cell tests.

[0324] The effects of the treatments of the E5250 matrix with SEBflo TL are presented in FIG. 20.

[0325] The addition of the enzymatic solution to the P-11.11 matrix induces no inflammatory response in the four cell types, which is proof that the enzyme used under the conditions of the experiment does not introduce contaminants.

[0326] The treatment of the E5250 matrix induces a slight decrease in the inflammatory responses in the Raw and HEK-TLR2 cells. The enzyme is described essentially for its beta-glucanase properties. However, the decrease in the cell responses observed is accompanied by an increase in the HEK-NOD2 cell response. These data indicate that the enzymatic preparation also contains an activity capable of degrading PGNs.

[0327] In parallel to the appearance of the PGN degradation products, a slight increase in the TLR4 response is observed in the presence of the enzyme, even though the latter is not contaminated. As previously, the enzyme probably released inflammatory contaminants, but to a lesser degree than what is observed with Mannaway.

[0328] The effects of the treatments of the E3063 matrix with SEBPro FL100 are presented in FIG. 21.

[0329] The addition of the enzymatic preparation to the P-11.11 matrix induces a strong inflammatory response in the HEK-TLR4 cells, and also weak but significant responses in the Raw and HEK-TLR2 cells. These data indicate that the enzyme is contaminated with LPS and traces of PGN.

[0330] The addition of the enzyme to the E3063 matrix induces a slight decrease in the TLR2 response, which is accompanied by an increase in the HEK-NOD2 cell response. These data indicate that SEBPro FL100 has a weak lytic action on PGNs. However, its use would necessarily require a prior decontamination step in order to remove the LPS.

Example 7: Effect of the Treatments by Passing Over Resins

[0331] The objective of these tests is to test the capacity of industrial resins to retain the contaminants present in glucose polymer matrices and, consequently, to reduce the pro-inflammatory reactivity of these matrices.

[0332] The tests were carried out with the E1565 matrix (solubilized at 32% weight/volume in sterile water), since it is contaminated with the various types of pro-inflammatory molecules that may be found in production circuits (TLR2, TLR4 and NOD2 responses).

[0333] For the experiments, the solution to be decontaminated was continuously eluted on a column containing 20 ml of each resin (bed volume). The cell tests for inflammatory reactivity were carried out using the solution before it was passed over the column (contamination control), and then on the samples recovered after passing 4 volumes (passage 5) and 10 volumes (passage 11) of solution. This procedure made it possible to verify whether the presence of the glucose polymer did not cause a resin saturation phenomenon.

[0334] The results are presented in FIG. 22.

[0335] The passing of the solution over the various resins causes a decrease in the inflammatory reactivity of the E1565 matrix on contact with the Raw cells. With the exception of FPX66, the decrease in response reaches at least 50% for the other resins. The reduction even reaches 70% after passing over the SD2 resin, thereby indicating that this resin is the most effective for removing the contaminating molecules with pro-inflammatory activity contained in the E1565 matrix. In any event, no significant difference is observed between passages 5 and 11, thereby excluding any substrate saturation phenomenon.

[0336] The reactivity of the HEK-TLR2 cells with respect to the E5250 matrix is not significantly modified after passing over the various resins, thereby indicating that these treatments are ineffective for reducing the contaminated PGN load.

[0337] The MN-100 and XAD-1600 resins are, for their part, found to be very effective for reducing the HEK-TLR4 responses with respect to the E1565 matrix. These data indicate that these two resins have a high capacity for retaining molecules of LPS type. Conversely, the other resins are barely, or even totally, ineffective for retaining this contaminant.

[0338] Finally, the various resins moderately reduce the HEK-NOD2 cell responses with respect to the matrix, with the exception of FPX66, which is totally ineffective. However, the effect observed remains weak, demonstrating that the resins have a weak capacity for retaining PGN degradation products.

[0339] Besides the presence of traces of LPS, the E1565 matrix is strongly contaminated with PGN and with degradation products thereof. The latter have little inflammatory reactivity per se; on the other hand, they are capable of acting in synergy with the other inflammatory molecules that interact with TLRs, such as PGNs and LPSs, and of exacerbating the overall immune response.

[0340] The tests carried out in this example show a significant decrease in the reactivity of the Raw cells after passing over the various resins. This decrease in overall inflammatory response is not subsequent to a retention of PGNs, since the passing over resins does not modify the TLR2 responses.

[0341] Only two resins (MN-100, XAD-1600) out of seven are clearly effective for reducing the TLR4 response. It can therefore be deduced therefrom that the decrease in inflammatory response observed in the Raw cells is at least partially subsequent to the retention of LPS after passing the matrix over these two resins.

[0342] With the exception of FPX66, all the resins moderately retain the PGN degradation products. Given the impact of these small molecules on the exacerbation of inflammatory responses, these results indicate that the major effect of the resins is to remove the synergistic effects associated with these small molecules.

[0343] As for the FPX66 matrix, its effect is probably linked to the removal of inflammatory molecules other than LPS, and PGNs and degradation products thereof. This hypothesis is compatible with the fact that this resin is the least effective for reducing the overall inflammatory response.

[0344] As a whole, these data indicate that treatment of the matrices with certain judiciously selected industrial resins can prove to be effective for removing inflammatory contaminants other than PGNs, but also reducing the effects of synergy observed between these molecules. Example 4 of the present study showed that some industrial carbons are particularly effective for removing PGNs. Consequently, a procedure combining the two types of treatment should make it possible to target the various families of contaminants and to provide matrices free of inflammatory reactivity.