Method for Emulsifying a Triepitope Peptide with Montanide and Kits for Performing the Same

Abstract

The present invention relates to a method for preparing a vaccine formulation with an optimized polypeptide comprising three epitopes and MONTANIDE as adjuvant, for use in anti-cancer immunotherapy for administration to a human subject.

Claims

1. A method of obtaining a water in oil emulsion for administration of a polypeptide of sequence YLQVNSLQTVYLEYRQVPVYLEEITGYL (SEQ ID NO: 1) to a human subject, comprising: (i) obtaining an aqueous antigenic phase by dissolving lyophilized polypeptide of SEQ ID NO: 1 in pure sterile water; (ii) loading one volume of said aqueous antigenic phase and one volume of MONTANIDE ISA 51 into a device comprising two syringes linked by a connector; (iii) pre-emulsifying the formulation by performing at least three cycles of transfer of the whole formulation from one syringe to the other and then back to the first syringe, wherein each cycle lasts at least 6 seconds; and (iv) emulsifying the formulation by performing at least 20 additional transfer cycles, wherein each cycle lasts at most 2 seconds.

2. The method of claim 1, wherein step (i) is performed by hydrating a lyocake of polypeptide of SEQ ID NO: 1 with pure sterile water at a temperature between 10 and 40 C. and waiting until the dissolution is completed.

3. The method of claim 1, wherein the peptide concentration in the aqueous solution is between 2 and 20 mg/ml.

4. The method of claim 1, wherein the (ii) loading is performed by loading one volume of the aqueous antigenic phase into a first syringe and one volume of MONTANIDE ISA 51 into a second syringe, and then bridging the two syringes by the connector.

5. The method of claim 1, wherein the connector is an I connector.

6. The method of claim 1, wherein in the (iii) pre-emulsifying, 3 to 10 cycles are performed.

7. The method of claim 1, wherein in the (iii) pre-emulsifying, each cycles lasts between 6 and 30 seconds.

8. The method of claim 1, wherein in the (iv) emulsifying, between 20 and 40 cycles are performed.

9. The method of claim 1, wherein in the (iv) emulsifying, each cycles lasts between 0.5 and 2 seconds.

10. A kit for obtaining a water in oil emulsion containing the polypeptide of SEQ ID NO: 1 and MONTANIDE ISA 51, wherein said kit comprises at least 1 mg of lyophilized polypeptide of SEQ ID NO: 1, at least 1 ml of MONTANIDE ISA 51, two syringes, an I-connector and a notice of use describing the process according to claim 1.

11. The kit of claim 10, wherein each syringe is a 2 ml syringe.

12. The kit of claim 10, which further comprises a vial of water for injection and/or a vial adapter.

13. The kit of claim 10, which further comprises e sterile needle for vaccinating the patient after obtaining the emulsion.

14. A water in oil emulsion containing the polypeptide of SEQ ID NO: 1 and MONTANIDE ISA 51, wherein the peptide concentration in said emulsion is in the range 1 to 10 mg/ml, and wherein the Dv 50 of the emulsion is below 3 m.

15. The water in oil emulsion of claim 14, wherein the Dv 50 of the emulsion is below 2 m.

Description

FIGURE LEGEND

[0046] FIG. 1: Vx006 in solution. A: in water for injection; B: in PBS buffer; C: in NaCl 0.9% saline buffer.

EXAMPLES

[0047] The examples have been performed using the following materials and methods:

[0048] Peptides. Peptides were synthesized by BACHEM (Basel, Switzerland). Vx006 (SEQ ID NO: 1) has been prepared in ammonium solution prior to lyophilisation.

[0049] MONTANIDE ISA51VG. MONTANIDE ISA51VG was provided by SEPPIC (Castres, France), Batch number T 134201.

[0050] Micelle kun and emulsification devices. A prototype of the Micelle kun (described in EP 2 322 132 A1), was used at SEPPIC (Castres, France). Silicone free syringes of 2 ml are produced and marketed by BBraun (reference 4606701V). I-connector are produced and marketed by didactic GROUP (Etainhus, France) (reference RACDLLFTF).

[0051] Emulsion characteristisation. To characterize the emulsion, two tests are commonly used; the drop test and the measure of the drops size.

[0052] Drop test: The drop test allows checking the type of emulsion: W/O (aqueous phase drops in oily base) or O/W (drops of oil (adjuvant) in aqueous phase). In a beaker with clear water, lay 1 drop of emulsion on the surface of water and mix gently.

[0053] If the emulsion drop stays on the surface, it is a W/O emulsion.

[0054] If the emulsion disperses in water, it is an O/W emulsion.

[0055] Size of drops: The size of drops in the dispersed phase was checked by laser scattering granulometer. Particle size of emulsions was measured with a Malvern Mastersizer S.

[0056] The average particle size was represented by:

[0057] Dv 50 means: 50% of the volumes of particles have a size lower than the Dv50,

[0058] DV 90 means: 90% of the volumes of particles have a size lower than the Dv90.

Example 1

Emulsification of Placebo Using the SEPPIC Recommended Protocol

[0059] Emulsifying MONTANIDE ISA 51 VG requires a saline aqueous phase. It is known that the use of saline buffer (such as PBS and NaCl 0.9%) allows producing thinner emulsions than pure water. To exemplify this, investors have compared emulsions prepared by five independent manipulators with MONTANIDE ISA 51 VG with either pure water or NaCl 0.9% saline solution. 0.7 ml of placebo solution was loaded into a silicone free syringe and 0.7 ml of MONTANIDE ISA 51 VG into a second syringe. After connection of both syringes onto an I-connector, the classical protocol was applied, i.e., 20 slow cycles followed by 40 rapid cycles. The size of drops in the dispersed phase was checked by laser scattering granulometer. Particle size of emulsions was measured with a Malvern Mastersizer S.

[0060] Table 1 shows that Dv50 and Dv90 are significantly lower with NaCl 0.9% than with water. Dv50 and Dv90 measured with emulsion of NaCl 0.9% with MONTANIDE were 0.5 and 1.1 m in average, compared to 0.8 and 1.7 m in emulsions with water. More importantly, emulsions performed with NaCl 0.9% saline buffer were much more stable than the emulsion with sterile water, since after 24 hours at 4 C., Dv50 and Dv90 measured with emulsions of NaCl 0.9% with MONTANIDE were respectively 1.1 and 1.5 m in average, compared to 3.6 and 8.3 m in emulsions with water. These results are in accordance with the recommendation of SEPPIC, MONTANIDE ISA51VG producer.

TABLE-US-00001 TABLE 1 characteristics of emulsion performed with ISA51VG with pure water or NaCl 0.9% saline buffer MONTANIDE ISA51VG + MONTANIDE ISA51VG + NaCl 0.9% Pure water T0 T24 h T0 T24 h Dv50 Dv90 Dv50 Dv90 Dv50 Dv90 Dv50 Dv90 Manipulator 1 0.7 1.4 0.8 1.4 1.1 1.6 3.6 8.2 (test 1) Manipulator 1 0.3 0.6 0.3 0.7 ND ND (test 2) Manipulator 1 0.4 1.3 0.5 1.1 ND ND (test 3) Manipulator 2 0.9 1.5 0.9 1.5 1.2 1.8 3.6 8.3 Manipulator 3 0.5 1.1 0.5 1.1 1.3 1.8 3.7 8.4 Manipulator 4 0.3 0.8 0.4 1.4 ND ND Manipulator 5 0.3 0.8 2.3 4.4 ND ND (ND: Not Done)

Example 2

Vx006 Solubilisation

[0061] Vx006 was produced by BACHEM (Basel, Switzerland). For clinical trial, Vx006 was produced with a GMP grade, as vials containing 10 mg of lyophilized peptide.

[0062] For vaccinating a patient, 1 ml of solvent is added on the lyocake to resuspend the peptide, 0.7 ml of peptide in solution is then loaded in a silicone free syringe and 0.7 ml of MONTANIDE ISA51VG in a second syringe. Both syringes are then connected to the I-connector for the emulsification protocol.

[0063] In practice, solubilizing a peptide can be quite a challenge, although the following chemical rules exist to predict peptide solubility (disclosed, for example, in Sigma-Aldricht's web site):

[0064] 1. Peptides which are shorter than 5 residues are generally soluble in aqueous media, except in extreme cases where all the residues are very hydrophobic (W, I,

[0065] 8

[0066] 2. Hydrophilic peptides, containing >25% charged residues (E, D, K, R and H) and <25% hydrophobic residues also generally dissolve in aqueous media, provided that the charged residues are fairly distributed throughout the sequence. Peptides are generally purified with 0.1% TFA/water and 0.1% TFA/ACN solvent system. Therefore, if the peptide is dissolved in aqueous solution which is unbuffered or insufficiently buffered, the resulting peptide solution can be acidic. The pH of the solution must be around neutrality before trying other means of solubilization. Both acidic peptides (E+D residues>K+R+H residues) and basic peptides (R+K+H residues>E+D residues) are more soluble at neutral pH than at acidic pH.

[0067] 3. Hydrophobic peptides containing 50% to 75% hydrophobic residues may be insoluble or only partially soluble in aqueous solutions, even if the sequence contains 25% charged residues. It is best to first dissolve these peptides in a minimal amount of stronger solvents such as DMF, acetonitrile, isopropyl alcohol, ethanol, acetic acid, 4-8M GdnHCl or urea, DMSO (if the sequence does not contain C, W or M), and other similar organic solvents, and then slowly add (drop wise) the solution to a stirred aqueous buffer solution. If the resulting peptide solution begins to show turbidity, the solubility limit might have been reached and it will be futile to proceed.

[0068] 4. Very hydrophobic peptides containing >75% hydrophobic residues will generally not dissolve in aqueous solutions. These peptides generally require initial solubilization in very strong solvents such as TFA and formic acid and may precipitate when added into an aqueous buffered solution. The final peptide solution may require a higher concentration of organic solvent or denaturant, which may not be applicable in biological studies involving live cells.

[0069] 5. Peptide sequences containing a very high (>75%) proportion of S, T, E, D, K, R, H, N, Q or Y are capable of forming extensive intermolecular hydrogen bonding network and have a tendency to form gels in concentrated aqueous solutions. These peptides may have to be treated similarly to step #3.

[0070] Since the Vx006 peptide of SEQ ID NO: 1 comprises 15 hydrophobic residues (between 50 and 75%), 4 charged residues (<25%) and 16 (<75%) residues of the list cited in point 5 above, it theoretically needs a treatment as described in the above point 3, but may be partially soluble in water.

[0071] According to SEPPIC recommendations, to the fact that the emulsion is performed for human administration (preventing the use of solvents as DMSO etc.), results presented in Example 1 above and solubility prediction, Vx006 was initially diluted in saline buffer. Two saline solvents were tested, NaCl 0.9% and PBS. In both cases, the peptide in solution was cloudy, the solubilisation was not complete. More importantly and surprisingly, after a few minutes, the solution became viscous and finally a gel was formed, preventing any emulsification of the peptide with MONTANIDE. Moreover, the addition of several solvents developed by Seppic and compatible with injection in humans and emulsion with MONTANIDE were tested during the Vx006 manufacturing development. None of these solvents had any impact on the Vx006 solubilisation.

[0072] After a series of unsuccessful tries, inventors discovered that the peptide was perfectly soluble in water for injection, and that the pH of the obtained solubilised peptide was compatible with injection into humans (pH around 7). The lyocake was hence hydrated with a small volume of pure water (1 to 5 ml), and left at room temperature (15-25 C.) until full dissolution (10 seconds to a few minutes). The vial can be gently agitated to accelerate the dissolution, if necessary. The dissolution can also be performed at cold temperature (0-10 C.). The peptide in solution was totally clear and no gel formation was observed (FIG. 1).

Example 3

Vx006 Emulsification in MONTANIDE

[0073] Vx006 in solution in pure water at a concentration of 10 mg/ml was then emulsified with MONTANIDE ISA51VG. First of all, 0.7 ml of peptide in solution was loaded in a silicone free syringe and 0.7 ml of MONTANIDE ISA51VG in a second syringe. After connection of both syringes onto an I-connector, the typical protocol recommended by SEPPIC was applied, i.e., 20 slow cycles (of 4 seconds) followed by 40 rapid cycles. Four manipulators trained for performing emulsion have applied the same protocol.

[0074] The drop test was then performed to evaluate the emulsions. As shown in table 2, in approximately 25% of the emulsifications, the emulsions were oil in water (O/W) emulsions, instead of water in oil (W/O). Moreover, even if the peptide was diluted 5 times (final concentration 2mg/ml), the result remained the same, showing that the emulsification capabilities of the peptide are not due to peptide concentration but to intrinsic characteristics of the Vx006 peptide. Such variability is not acceptable in the context of clinical trials, since oil in water emulsion do not play the role of adjuvant in the vaccination. Inventors then modified several parameters to define a robust protocol for emulsifying Vx006.

TABLE-US-00002 TABLE 2 Emulsification of Vx006 in MONTANIDE ISA 51 with the classical protocol. Emulsion characteristics Manipulator Test 1 Test 2 Test 3 Test 4 M1 (10 mg/ml) W/O W/O W/O M2 (10 mg/ml) W/O O/W W/O M3 (10 mg/ml) W/O W/O O/W M4 (10 mg/ml) O/W W/O O/W M1 (2 mg/ml) W/O W/O W/O O/W M3 (2 mg/ml) W/O W/O O/W M4 (2 mg/ml) O/W O/W W/O W/O

[0075] To avoid the manipulator bias, next tests were performed using a Micelle kun, which is an automatized apparatus conceived to perform controlled emulsion processes (EP 2 322 132 A1). A prototype of the device was provided by SEPPIC. Three slow-cycle speeds were tested, 0 which corresponds to 12 s for one cycle, 5 which corresponds to 6 s for one cycle and 10, which is 4 s per cycle. 20 slow cycles and 40 rapid cycles were then applied (Micelle kun performs the rapid cycles at the constant speed of 0.7 s per cycle). Finally, both peptide concentrations 10 mg/ml and 2 mg/ml were tested. Results in table 3 show that Micelle kun was always able to produce W/O emulsions at speeds 0 and 5 but not at 10, showing the importance of applying very slow cycles during the pre-emulsifying step. It is important to note that the supplier of the Micelle kun initially envisaged to produce the Micelle kun with a unique low speed, but SEPPIC requested that the low speed can vary and reach cycles as long as 12 seconds, in order to ensure that it could be used to emulsify all peptides, including particular peptides such as Vx006. Indeed, the specialists from SEPPIC never worked with any peptide as sensitive to the parameters of the pre-emulsification step as the peptide of SEQ ID NO: 1.

TABLE-US-00003 TABLE 3 Emulsification of Vx006 in MONTANIDE ISA 51 with various slow cycle speeds using the micelle kun. Peptide concentration Slow cycle speed Drop test 2 mg/ml 0 W/O 2 mg/ml 0 W/O 2 mg/ml 0 W/O 2 mg/ml 0 W/O 10 mg/ml 0 W/O 10 mg/ml 0 W/O 10 mg/ml 5 W/O 10 mg/ml 5 W/O 10 mg/ml 5 W/O 10 mg/ml 5 W/O 10 mg/ml 5 W/O 2 mg/ml 10 O/W

[0076] To simplify the preparation of the vaccine in the hospital, the inventors decided to determine the minimum number of slow cycles necessary to obtain a W/O emulsion. Using the Micelle kun, the number of slow cycles was decreased. The protocol that allows to obtain each time W/O emulsions of Vx006 and MONTANIDE ISA 51 VG with the Micelle kun is: 5 cycles at very low speed (speed 0) followed by 40 cycles at high speed (table 4). It is to be noted that only 3 slow cycles can be sufficient, when these cycles are performed manually at a very low speed (at least 10-15 seconds per cycle).

TABLE-US-00004 TABLE 4 determination of the minimal number of slow cycles Number of slow Peptide Slow cycle cycle + number of concentration speed rapid cycle Drop test 2 mg/ml 0 0 + 60 O/W 2 mg/ml 0 3 + 40 W/O 2 mg/ml 0 5 + 40 W/O 2 mg/ml 0 10 + 40 W/O 2 mg/ml 0 20 + 40 W/O 2 mg/ml 10 3 + 40 O/W 2 mg/ml 10 3 + 40 W/O 10 mg/ml 0 3 + 40 O/W 10 mg/ml 0 3 + 40 W/O 10 mg/ml 0 5 + 40 W/O 10 mg/ml 0 10 + 40 W/O 10 mg/ml 0 20 + 40 W/O 10 mg/ml 5 5 + 40 W/O

[0077] Finally, to test the robustness of the defined protocol, three manipulators applied 5 very slow cycles and 40 rapid cycles. Table 5 shows that 100% of the emulsions were W/O, thin and stable at 24 hours. Interestingly, the emulsions with Vx006 are more stable than those without the peptide (compare the data of Table 5 below with those of Table 1).

TABLE-US-00005 TABLE 5 robustness of the protocol 5 + 40, performed manually MONTANIDE ISA51VG + Vx006 Emulsion T0 T24h characteristic Dv50 Dv90 Dv50 Dv90 Manipulator 1 (10 mg/ml) W/O 1.2 1.8 3.0 4.8 Protocol 5 + 40 Manipulator 2 (10 mg/ml) W/O 1.2 1.7 2.9 4.5 Protocol 5 + 40 Manipulator 3 (10 mg/ml) W/O 1.3 1.6 3.0 4.8 Protocol 5 + 40

REFERENCES

[0078] Ascarateil and Koziol: New equipment optimized for emulsified vaccine preparation in the hospital. Journal for ImmunoTherapy of Cancer 2013 1(Suppl 1):P196.