Stabilized Interleukin-2

The invention concerns both stabilizing a preparation of interleukin-2 intended for therapeutic purposes, especially in humans and animals, and a preparation obtained accordingly in the form of aqueous solutions or solid forms.

Lymphokines, in their capacity as hormone-like signal transmitters, carry out a central function in creating and regulating immune responses (Sorg, C. and Schimpel, A., Cellular and Molecular Biology of Lymphokines, Orlando, Academic Press, 1985, and Gillis, S. and Inman, F. P., Contemporary Topics in Molecular Immunology, Vol. 10, The Interleukins, New York & London, Plenum Press, 1985). A series of immunological activities is controlled by one lumphokine, specifically interleukin-2 (IL-2). Within the overall framework of cellular immune response, interleukin-2 affects the release of other lymphokines: IFN-gamma, BCGF, and BCDF (Greene, W. C. and Robb, R. J., Receptors for T-Cell Growth Factor: Structure, Function and Expression on Normal and Neoplastic Cells, Contemporary Topics in Molecular Immunology, vol. 10, 1-34 (1985)).

Human interleukin-2 (also called TCGF) is a glycoprotein produced in small quantities by human peripheral T cells (T lymphocytes) subsequent to stimulation by antigens, mitogens (Mier, J. W. and Gallo, R. C., Purification and Some Characteristics of Human T-Cell Growth Factor from Phytohaemagglutinin-Stimulated Lymphocyte-Conditioned Media, Proc. Natl. Acad. Sci. 77, 6134-38 (1980)), or appropriate monoclonal antibodies (with the "first signal" being activation) (Cantrell, D. A. and Smith, K. A., The Interleukin-2 T-Cell System: A New Growth Model, Science 224, 4655, 1312-1316 (1984)).

Interleukin-2 provides the "second signal" in the immune response (Wagner, H., Hardt, C., Heeg, K., Pfitzenmaier, K., Solbach, W., Bartlet, R., Stockinger, H., and Rollinghof, M., In Vivo and in Vitro Effects of Interleukin-2, Immunol. Rev. 51, 215-36 (1980) and Watson, J. and Mochizuki, D., Interleukin-2: A Class of T-Cell Growth Factors, Immunol. Rev. 51, 257 (1980)) and represents a proliferation-and-regulation factor by means of which neoplastic T cells can be continuously raised in vitro. The effectiveness of interleukin-2 is evident in its in-vivo and in-vitro activation of alloreactive and tumor-reactive T cells, natural and lymphokine-activated killer-cell populations as well as helper, suppressor, and cytotoxic T cells. Human interleukin-2 also stimulates the proliferation of cells that depend on mouse or rat interleukin-2.

Human interleukin-2 (n IL-2) is a protein formed of 133 amino acids. It is hydrophobic. It is glycosylated at position 3 by the amino acid trieonine. According to its microheterogenous glycosylation pattern during post-translational modification, human interleukin-2 exhibits, as its degree of glycosylation drops, at least three molecular species at 17, 16, and 14.5 kD (SDS-Page) (Korner, I. J., Dettmer, R., Kopp, J., Gaestel, M., and Malz, W., Simple Preparative Two-Step Purification of Interleukin-2 from Culture Medium of Lectin-Stimulated Normal Human Lumphocytes, J. Immunol. Meth. 87, 185-191 (1986)).

Activated but non-resting T cells express high- and low-affinity interleukin-2 receptors (Robb, R. J., Greene, W. C., & Rusk, C. M. Low- and High-Affinity Cellular Receptors for Interleukin-2, J. Exp. Med. 160, 1126-46 (1984) and Robb, R. J., Interleukin-2 and its Cell-Surface Receptor, Behring Inst. Mitt. 77, 55-67 (1985)). A maximum number of interleukin-2 receptors will form on normal lymphocytes treated with mitogens within approximately 60 hours, with a total of approximately 20 000 per normal T cell.

The bonding of interleukin-2 in the capacity of peptide hormone to interleukin-2 receptors and its subsequent internalization, finally, triggers cell division (Robb, R. J., Interleukin-2 and its Cell-Surface Receptor, Behring Inst. Mitt. 77, 55-67 (1985)). The signals are transmitted from cell to cell at this point through a cascade of signal communications, which Ca.sup.2+ participates in, and through downstream phosphorylation of the ribosome protein S 6 prior to translation.

The production and purification of human interleukin-2 is described in the literature (Sonneborn, H.-H. and Schwulera, U., Production, Purification, and Properties of Human Interleukin-2 (IL-2), Biotest-Bulletin 1, 3, 222-27 (1982)).

In addition to its use in the laboratory and for diagnostic purposes, interleukin-2 can also be employed therapeutically in humans and animals, to inhibit tumor growth for example. The most recent research indicates that multiple intratumoral injections in the mouse model can lead to deceleration of tumor growth to the extent of complete rejection.

A high-purity interleukin-2 preparation is a prerequisite for injection and infusion. The purer the preparation, however, the more unstable its biological activity.

However, since as previously discussed herein, only small amounts of human interleukin-2 can be obtained from human peripheral T cells, large amounts will have to be stored for a fairly long time in the form of solutions, deep-frozen batches, or lyophilizates of a high-purity interleukin-2 preparation to allow its extended and reliable clinical use as a therapeutic and especially antitumoral agent. Drug authorities for instance require that a therapeutic agent remain stable at least one year, a demand that no known preparation has satisfied as yet.

The activity of high-purity interleukin-2 in the context of therapy, however, decreases when it is stored very long, especially when deep-frozen, thawed, or lyophilized.

The object of the invention is accordingly to provide a high-purity interleukin-2 preparation that will be appropriate for therapeutic purposes, especially in humans, that will remain stable when stored for a long time, and that will continue to exhibit sufficient activity after being deep-frozen, lyophilized, and even thawed several times.

This object is attained in accordance with the invention by using a particular stabilizer as characterized hereinbelow.

It was surprisingly discovered that substances--such as globulins, sugar alcohols, monomeric and polymeric sugars, gelatins, and mixtures thereof--not yet known as stabilizers for interleukin-2, do indeed stabilize it very well. In addition to a series of other known stabilizers--such as for example SH-reducing compounds, glycerin, and polyethylene glycols, albumins, and especially human-serum albumin, are already known as stabilizers for lower-purity interleukin-2 preparations.

It is, however, completely surprising that albumins selected on the basis of a special method of characterization have the capacity to stabilize for long periods even the high-purity interleukin-2 preparations previously described herein as being especially likely to lose their activity unless stabilized. Specially characterized albumins are to be understood herein as those that exhibit the criteria that will be described hereinbelow for albumins that can be used to stabilize interleukin-2 for therapeutic purposes. It has been discovered that high-purity interleukin-2 will, when stabilized with a specially characterized human-serum albumin of this type, retain 85% of its activity when stored for 12 months at 37.degree. C., 90% when stored that long at 22.degree. C., and even 95% for the same period at 4.degree. C. and will accordingly completely comply with the authorities's requirements that a therapeutic preparation remain sterile at least one year. With no stabilizer added, a high-purity interleukin-2 preparation will lose its activity in accordance with its protein concentration.

Even freeze-drying will not stabilize an interleukin-2 preparation unless a stabilizer is added.

Table I provides a comparison of the properties of interleukin-2 preparations for laboratory and for therapeutic purposes.

A high-purity interleukin-2 preparation that is especially appropriate for processing in accordance with the invention is one prepared by the method disclosed in German OS 3 411 184 and U.S. Pat. No. 4,508,833 (corresponding to German P 3 149 360.2).

In this method the interleukin-2 is prepared from human peripheral-blood lymphocytes by adding a phorbol ester, especially .beta.-phorbol-12-myrisate-13 acetate (PMA) and Calcium Ionophor A 23 187 (6S-[6-.alpha.-(2S*,3S*),8.beta.(R*),9.beta.,11-.alpha.]-5-(methylamino)-2 -[[3,9,11-trimethyl-8-[1-methyl-2-oxo-2-(1H-pyrrol-2-yl)-ethyl]-1,7-dioxasp iro-[5.5]-undec-2-yl]-methyl]-4-benzoxazole-carboxylic acid, cf. the brochure Calbiochem-Behring, Doc. No. 8154-1082 Biologics, Antibiotica A 23187) and by employing several chromatographic purification stages. The resulting product has the following characteristics:

Table II lists stabilizers that are appropriate in accordance with the invention along with their stabilization effect on the previously described interleukin-2 preparations.

Another essential criterion besides stabilization activity for the selection of stabilizers for a therapeutically usable interleukin-2 preparation is that they be non-toxic and pharmacologically unobjectionable.

Table II shows that plasma expanders have a stabilizing effect on interleukin-2 and that globulins have an even more powerful one. The best results, however, will be evident in conjunction with the addition of human serum selected in accordance with a special method of characterization or with the addition of selected batches of albumin. Freezing and thawing 5 times resulted in practically no decrease in activity. Furthermore, no stabilizer protein precipitates. The same is true of the lyophilized product, whereby the effect is even more drastic in that the unstabilized control exhibits a residual activity of only 2% subsequent to lyophilization. Human-serum albumin (HSA) turned out to be particularly effective. The preferred human-serum albumin is at least 96% pure, with the percentage of monomers as high as possible and the remaining dimer and polymer share as low as possible. The pH should be between 6.7 and 7.4.

An extremely important factor in selecting an appropriate human-serum albumin is that there exist batches that inhibit interleukin-2 activity. Although the cause of this effect is not yet completely understood, it is suspected that some substance of a medicinal or other nature in the donor's cells is responsible. Every batch of human-serum albumin that is to be employed as a stabilizer must accordingly be tested for inhibition of interleukin-2 activity. This is done with a .sup.3 H-thymidine-incorporation test as described hereinbelow with reference to the examples. One or two days of testing is enough to determine whether a batch is inhibiting or not. The terms "appropriate" and "selected" in the instant description and claims are accordingly to be understood as meaning that a batch of human-serum albumin does not inhibit interleukin-2 activity, whereas an "inappropriate" batch is one that does inhibit it.

The criteria for selecting human-serum albumin as a stabilizer for interleukin-2 that is to be used for therapeutic purposes will now be summarized.

It will not inhibit interleukin-2 activity.

It will be non-toxic to two species in vivo.

It will be sterile.

It will be pyrogen free (in the limulus test and in rabbits).

It will have an albumin level of 96-100% (with a monomer content of 97%).

It will be heat-stable for 50 hours at 57.degree. C.

It will not react with blood-group antibodies.

It can be released for infusion.

As will be evident from the examples and from Table III, even 0.1% of a selected batch of human-serum albumin preparation will be enough to optimally stabilize 1-10.sup.6 U/ml of interleukin-2. Although higher concentrations also have a satisfactory stabilizing action, they frequently lead to albumin-dictated side effects. When the concentration of interleukin-2 is higher than 10.sup.6 U/ml, somewhat higher concentrations of stabilizer will be needed. Tests have shown for example that approximately 0.2-1% w/v of albumin is necessary for 10.sup.7 -10.sup.8 U/ml of interleukin-2.

Table III evidences the outstanding stabilizing action of 0.1% of a selected batch of human-serum albumin over a wide range of concentrations of interleukin-2.

The table lists the interleukin-2 activity remaining after 1 week of storage at 4.degree. C. Human interleukin-2 was stored in RPMI in the presence of 50 .mu.g of streptomycin sulfate per ml and 50 IU of penicillin G with and without stabilizers. The control was optimally stabilized and stored at -70.degree. C.

Since albumins can incompletely polymerize or denature during pasteurization it is practical to add an albumin stabilizer to the preparations. .epsilon.-Aminocaprylate and N-acetyltryptophan have proved to be appropriate stabilizers for albumins. Appropriate amounts of these stabilizers range between 0.1 and 0.04 mM, with 0.08 mM preferable.

It has also turned out to be practical for certain purposes to use mixtures of the claimed stabilizers or to use them in mixtures with other known interleukin-2 stabilizers (cf. examples).

Adding either 0.1M of L-lysine, L-tryptophan, L-arginine, up to 1 mM of monovalent or bivalent salts like NaCl, CaCl.sub.2, and MgCl.sub.2, or 5.times.10.sup.-5 M of sequestering agents alone have only a slight effect on the stabilization of interleukin-2 (with a residual activity that is 1-5% higher). The substances, are however, effective, depending on the purpose, in stabilization mixtures. Monosaccharides (e.g. glucose, mannose, and fructose) and disaccharides (sucrose, maltose, and lactose) alone and sugar alcohols alone at a concentration of 0.1M have only a slight effect (with a residual activity that is 3-6% higher). In a stabilizing mixture, however, especially in relation to long-term storage, it is a very good idea to add them, in conjunction with freeze drying for example.

Other additives that have already been described in the literature, like SH-reducing compounds (.beta.-mercaptoethanol or dithiotreitol) protect interleukin-2 from oxidation (at 10.sup.-3 -10.sup.-5 M). Adding 10-50% v/v of glycerin considerably decreases the inactivation of interleukin-2. A definite stabilizing effect is obtained with 0.1-10% w/v of polyethylene glycols (MW 1 500-40 000).

Some examples of stabilizing mixtures and the purposes that they are especially appropriate for will now be described.

Buffer systems consisting of

Tris-HCl

Na.sup.+ phosphate buffer

K.sup.+ phosphate buffer

Na.sup.30 -K.sup.+ phosphate buffer

were also tested for their suitability as stabilizers.

All three of these buffer systems were appropriate in concentrations of 0.1-50 mM. The Na.sup.+ phosphate buffer at a concentration of 10 mM was preferred because it resulted in a pH that was not too far away from that of the physiological environment.

The optimal conditions for a high-purity stabilized interleukin-2 preparation were found to be

Interleukin-2 concentration: 1-10.sup.9 U/ml

Buffer: 10 mM of Na.sup.+ phosphate

NaCl concentration: 0.15M

Osmolarity: 287 m osmol

HSA level in selected batch: 0.1-0.2% (w/v).

The invention will be further described in the following illustrative examples in conjunction with the appended drawings wherein

FIGS. 1 to 4 are plots showing the results of the respective examples.

The following albumin preparations were tested for their stabilizing action on interleukin-2:

inactivated human serum (HS) containing 70% human-serum albumin,

inactivated fetal-calf serum (FCS) containing 70% bovine-serum albumin,

Batch I: a 96% pure human-serum albumin, whereof 75% was monomer, plus an albumin stabilizer in the form of 0.08 mM of .epsilon.-aminocaprylate and 0.08 mM of N-acetyltryptophan, and

Batch II: a pure human-serum albumin from another donor.

1000 U/ml of human interleukin-2 obtained from peripheral-blood lymphocytes by the method disclosed in German OS 3 411 184 were incubated in RPMI 1640 for 1 to 3 days in a CO.sub.2 incubator (5% CO.sub.2 and more than 95% humidity) at 37.degree. C. with 50 JU of penicillin G/ml, 50 mg of streptomycin sulfate/ml, and one of the aforesaid albumin preparations (HS and FCS 20%, both batches HSA 10%). This was followed by a .sup.3 H-thymidine-incorporation test as described in the brochure "Lumphocult" (Biotest, September 1985) with ConA-blast cells as targets. The control was the current IUIS reference standard (Lymphokine Research 3, 277 [1984]), the value of which is represented by 100% in the figures. The degree of .sup.3 H-thymidine incorporation is considered the indication of interleukin-2 activity.

FIG. 1 illustrates the results of this test. As will be evident from the figure, the stabilizing action of both the human albumin and the fetal-calf serum was practically identical, still almost 100% after 3 days, whereas Batch I of the human-serum albumin exhibited a slight deviation, although still 100% after 3 days.

Batch II of the human-serum albumin turned out to be inappropriate. As will be evident from FIG. 2, whether a human-serum albumin batch is appropriate or not can be determined after 1 to 2 days, and definitely after 3 days.

100 000 U/ml of the interleukin-2 described with reference to Example 1 plus various amounts of human-serum albumin from the two batches described with reference to Example 1 were incubated in 0.01M of PBS (phosphate buffer 0.01M with 0.15M of NaCl) for 24 hours in a CO.sub.2 incubator (5% CO.sub.2, more than 95% humidity) at 37.degree. C. and a pH of 7.4. The interleukin-2 activity was then determined as described with reference to Example 1.

FIG. 2 shows the results.

If, as will be evident from the figure, the charge is appropriate, even 0.1% of albumin will be enough to stabilize up to 10.sup.6 U of interleukin-2/ml 100% optimally, whereas, if the charge is inappropriate, the inhibiting action increases with increasing albumin concentration.

Testing pH dependence

100 000 U/ml of the interleukin-2 from Example 1 were incubated in 0.01M of PBS with 0.1% of human-serum albumin like that in Batch I from Example 1 but without the caprylate and tryptophan in a CO.sub.2 incubator (5% CO.sub.2, more than 95% humidity) for 24 hours at 37.degree. C. and a pH of 7.4. The pH was then adjusted to various values and the activity of the interleukin-2 determined as described with reference to Example 1 after 1 day of storage at 4.degree. C.

FIG. 3 shows the results of this test.

As will be evident from the figure, the stabilized interleukin-2 was very extensively independent of pH. Only in the alkaline range was there a slight decrease in activity.

100 000 U/ml of the interleukin-2 from Example 1 in enough PBS to allow adjusting the pH to 7.4 and 0.1% human-serum albumin with the same characteristics as Batch I of Example 1 were stored for 12 months at the temperatures listed in FIG. 4. The activity was then determined as described with reference to Example 1.

FIG. 4 shows the results of this test.

As will be evident from the figure, it was surprisingly discovered that, in comparison to control stored at -70.degree. C., in heat-sealed sterile ampules or in the form of a lyophilizate plus an appropriate batch of human-serum albumin, no decrease in activity was found at 4.degree. C., although a decrease of about 10% was observable after 1 year at 22.degree. C., which is, however, still within the margin of error for the interleukin-2 test, and that even with storage for 1 year at 37.degree. C. only about a 15% loss in activity could be discovered. At -70.degree. C., -20.degree. C. and +4.degree. C. even after 2 years no loss in activity could be discovered.

No cloudiness or precipitation was discovered in the liquid inside any of the ampules.

It is possible to employ PHA blasts or MLC blasts instead of ConA blasts in the .sup.3 H thymidine-incorporation test in the foregoing research with practically the same results.

It will be understood that the specification and examples are illustrative but not limitative of the present invention and that other embodiments within the spirit and scope of the invention will suggest themselves to those skilled in the art.