WO1988007868A1 - Cytolytic factor - Google Patents

Abstract

The present invention relates to compositions and treatment methods which include or employ a novel macrophage-derived stable cytolytic antitumor factor, termed Cytolytic Factor (CF). Production/release of CF by the macrophage required transcription, translation, glycosylation, and an intact secretory apparatus, as evident from its inhibition by treatment with actinomycin D, cycloheximide, tunicamycin, and monensin, respectively, prior to and during triggering with LPS. CF obtained by culture of BCG-activated macrophages appeared rapidly in the supernatant after triggering. Using the actinomycin-D treated L-929 or EMT-6 targets in a microassay, CF secreted by macrophages cultured in low molecular weight serum components was detected as a -150 kD component on Sephacryl S-200 and was stable at 4°; when the macrophages were instead cultured in lactalbumin hydrolysate, the CF appeared to be unstable at this temperature. CF demonstrated a spectrum of cytotoxic activity against a number of tumor and normal targets in vitro. CF was moderately sensitive to treatment with TLCK and TAME. A rabbit heteroantiserum raised against highly purified necrosin, a product of the murine macrophage cell line J774.1, was extremely effective in neutralizing the biological activity of CF.

Description

CYTOLYTIC FACTOR BACKGROUND OF THE INVENTION

1. Field of the Invention

This invention relates to preparations which are useful in inhibiting the proliferation of human tumor cells and more particularly is directed to proteinaceous factors which are cytolytic for such cells.

2. Description of the Related Art

Cancer is a very wide spread and severe health problem which affects millions of people yearly, resultin in debilitating symptoms and often death. Numerous approaches, often fruitless, have been taken by medical scientists in an attempt to identify substances which may be of some usefulness in slowing or stopping the growth o human tumors. One avenue which has shown some promise is through the stimulation of the afflicted patient's immune system, thereby inducing the patient's immune system to produce substances capable of reducing the growth rate of tumor cells or, hopefully, killing them outright. Unfor¬ tunately, such an approach is often of little use in that the immune systems of cancer patients are either over burdened already or are simply incapable of responding to such immuno-stimulation.

This has led researchers to attempt to identify and purify substances produced by the immune systems of immuno-competent animals which are active in slowing or stopping the growth rate of tumor cells. Included in suc studies have been attempts to identify substances produce by "activated" macrophages, that is, macrophages which have been stimulated by some i muno-stimulating agent to produce tumor-suppressing substances. Indeed, a number o tumor-suppressive substances have been identified from activated macrophages.

Furthermore, numerous in vitro studies demonstrate that coculture of activated macrophages with tumor cells leads .to cytolysis or cytostatis of the tumor cells. Although the cytolytic response has been more extensivel studied, macrophage induced cytostasis is more frequentl observed and may play a role in controlling the develop- ment, progression and spread of tumor cells.

One cytostatic mechanism of activated macrophages i mediated through lesions induced by the effector cell in the target cell mitochondrial electron transport chain (ETC), in particular at Complex I and II. Such lesions result in growth inhibition or, if the damaged cells are not able to conduct adequate levels of glycolysis, death of the target. Kilbourn and co-workers, (1984) J. Immunol., 133:2577, have demonstrated that this cytotoxi mechanism might be accounted for by the secretion of a monokine, termed respiration inhibitory factor (RIF), which appears in large measure to mimic the activated macrophage in the exertion of cytostatic effects on a number of tumor cells.

Studies demonstrating release of intracellular iron by tumor cells provide evidence for a second pathway of macrophage-mediated cytostasis. These studies show a temporal correlation between iron release and inhibition of DNA synthesis in tumor cells cocultured with cytotoxi activated macrophages. Recently, studies by Hibbs et al (1986), J. Clin. Invest., 7j_ϊ790 have demonstrated a mechanism whereby macrophage-induced iron release could affect tumor cell metabolism. With respect to the lytic activity of activated macrophages, evidence has strongly suggested a require for macrophage-target cell contact to initiate the lyti mechanism (Meltzer et al. (1975), Cell Immunol., 17:30- However, under appropriate conditions many laboratories have reported that macrophage-conditioned supernatants also capable of expressing tumor cytotoxicity as well. The candidate mediators include oxygen metabolites (Nat et al. (1979), J. Exp. Med. , 149:100) , arginase (Currie (1978), Nature, 273:758) , thymidine (Stadecker et al. (1977), J. Immunol., 119:1738) , C3a (Ferluga et al. (1978), Clin. Expl. Immunol., _3i^:_>12), tumor necrosis factor (Carswell et al. (1975), Proc. Natl. Acad. Sci. USA., 7j_^:3666>' tumor killing factor (Itoh et al. (1986 J. Biochem., j>9:9), necrosin (Kull et al. (1984) Proc. Natl. Acad. Sci. USA., j3 :7932), respiration inhibitory factor (Kilbourn et al. (1984), J. Immunol., 133: 2577) , neutral proteases (Adams et al. (1980), J. Immunol. , 124:293; Reidarson et al. (1982), J. Natl. Cancer Inst. 9:889), as well as other apparently distinct macromole cules (Mathews (1981), Immunology, _44:135). The role o any of these mediators in the contact-dependent lytic mechanism remains unresolved.

As noted, two groups have previously observed mole cules which appear to depend on protease function for t expression of their cytotoxic function. Adams and coworkers employed the murine BCG-activated peritoneal macrophage as a source of a factor, eventually termed cytolytic protease, which was selectively cytotoxic for tumor cells in vitro (Adams et al. , supra) . This prote had a molecular weight of about 35 kD as determined by molecular sieving, and its lytic activity could be bloc by bovine pancreatic trypsin inhibitor, diisopropyl- fluorophosphate and, notably by serum. Reidarson and coworkers described a macrophage cytotoxin obtained from either alloimmune or inflammatory urine macrophages (Reidarson et al., supra) . MCT had a molecular weight o about 150 kD as determined by molecular sieving, and its cell-lytic activity was susceptible to treatment with

TLCK, TAME, and soybean trypsin inhibitor. Unfortunatel this factor was found to be extremely unstable and exhibited a half-life of only 24 hours when stored at 4° in physiological buffers (Klostergaard et al. (1984), Jrnl. Leukocyte Biol., 35: 229) .

Due to the extremely heterogeneous nature of cancer, and the numerous etiologies which this disease presents, it is unlikely that any one "universal" antitumor agent will be identified. However, certain agents, including various biological and proteineous factors, have been identified which have proved surprisingly efficacious against selected tumors. Thus, there is a continuing ne to identify and isolate agents having anticellular activity against tumor cells in that such agents will likely prove efficacious against one or more types of tumors.

SUMMARY OF THE INVENTION

In light of these and other concerns of the art, it is an overall object of the invention to provide novel antitumor compositions which include a novel cytolytic factor in a substantially purified form which may be employed in the treatment of cancer cells. This factor, designated Cytolytic Factor by the inventor, in its substantially purified form is shown most generally to b a soluble cytolytic protein having a molecular weight of between about 140 and about 160 kilodaltons upon gel exclusion chromatography. As used herein, the term "substantially purified" relates to the fact that the Factor of the invention may be purified away from other biological factors, which other factors may or may not have similar or interfering anticellular activities.

In various embodiments of the invention, this Cytolytic Factor may be characterized by one or more physical or biological properties which serve to both demonstrate its novelty over previously known biologica factors and further, to demonstrate its usefulness as a important new antitumor agent. For example, when isola in the manner disclosed herein, the Factor is shown to highly stable upon extended storage, for example, it is stable and retains its biological activity at 4^βC in physiological buffers for more than 6 months and, typically, much longer. Moreover, unlike many biologic factors of the prior art, the Cytolytic Factor disclose herein has surprisingly been found to be highly active the presence of serum, as shown by its retention of activity in the presence of 10% fetal calf serum.

In certain other embodiments, the Factor is shown be immunologically cross-reactive with polycional anti- serum raised against the well known antitumor factor

Necrosin. In fact, with certain such sera the Factor i shown to be biologically inhibited in' its presence. However, the Factor is shown to be distinct from Necros by numerous other accepted criteria.

In certain embodiments, the biological activity of the Factor is shown to be labile to treatment with trypsin, or upon heating to about 100°C for 10 minutes. Such properties demonstrate, for example, that the Cytolytic Factor includes a proteinaceous component tha is required for its antitumor action.

In still other embodiments, the Factor is shown to require protease activity for_^ the expression of its biological action. More particularly, the Factor is sh to be a neutral protease, it being active at a neutral Such activity is further demonstrated by sensitivity, generally a dose-dependent sensitivity, of the factor's biological antitumor activity to protease inhibitors su as TLCK or TAME.

With regard to its spectra of antitumor action, Cytolytic Factor is generally active to some extent against a number of tumor cell lines. However, the Fac is found to be surprisingly efficacious against some tumors and tumor cell lines, such as L-929, SVT-2 A375 human melanoma cells, these cells being generally regar in the art as indicative of activity against human tumo

In still other embodiments, the novel factor of th invention is preferably obtained from macrophages, most preferably _in_ vivo or _in vitro "activated" macrophages, derived from a mammalian origin such as a mouse, rabbit, guinea pig, etc., but preferably of murine origin. As used herein, "activated" macrophages refers generally t -macrophages which have been treated with an macrophage activating agent as such regents are generally referred in the art.

The Factor is thus most generally isolatable from media or culture supernatant wherein activated n_ vivo and/or jln_ vitro macrophages have been cultured so as to release the Factor therein. Such media is referred to generally herein as ".conditioned" supernatant in that i has been "conditioned" by virtue of the release of biological factors by the macrophages cultures therein.

In general, preparation of the macrophage conditio supernatant includes the steps of harvesting macrophage from a mammal, incubating the macrophages in an incubat medium and separating the macrophages to provide the resultant conditioned supernatant. Any number of tissu culture media or physiologic buffers known in the art c be used as the incubation medium. Moreover, it has bee found that pre-immunization of the mammal with a macrophage activator will result in the generation of e greater amounts of the factor by the subsequently harvested macrophages.

A number of suitable activators are described here however, in a preferred embodiment Bacillus Calmette Guerin ("BCG") is utilized as the activator. Similarly in a preferred embodiment, the mammal used is a mouse. a further embodiment, prior to being separated from the conditioned medium, the harvested macrophages are incubated in medium to which a triggering agent, for example, bacterial endotoxin, is added. Furthermore, where the macrophages are treated in vitro with a suita triggering agent, the in vivo activation step is not absolutely required.

Therefore, it is an object of the invention is to provide a method for cytolytically inhibiting the proliferation of tumor cells which includes subjecting tumor cells to an effective dose of a preparation which includes the cytolytic factor in substantially purified form. For these purposes, an effective dose is defined be 1 to 2 ul. conditioned by 1 x 10 BCG activated macrophages per ml. This dose of supernatant causes a 5 cell death of 25 x 10³ L-929 cells in 18-24 hours.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGURE 1. Supernatants from LPS-triggered, BCG-activate macrophages cultured in medium supplemented with LMS wer concentrated on a YM-10 membrane, and the concentrates subjected to -molecular sieving on Sephacyl S-200. The fractions were assayed for lytic activity on L-929 and EMT-6 targets, both targets were actinomycin D-treated. The resulting lytic activity detected on each target is shown. Molecular weight markers indicated are blue dextran, immunoglobulin G, hemoglobulin, and cytochrome

FIGURE 2. Adherent peritoneal exudate macrophages were established in 2 cm 2 wells. The macrophages were expose for 4 hr to LPS, PIC, SMDP, tuftsin, or PMA in the dose range from 3-1000 ng/ml, or A23187 in the range of 0.3 t 10 uM. The resulting lytic activity for each sample was determined by bioassay on L-929 targets. The units/ml o CF in test preparations was normalized with respect to that in supernatants from macrophages releasing CF spontaneously, and expressed as a stimulation index (mean _+ S.E. ) .

FIGURE 3. Monolayers of BCG-activated macrophages were est ~~a~~b~^—lished at 25 or 100 x 106• cells/78 cm2, and culture in 30 ml of medium supplemented with LMS or LAH after triggering with 100 ng/ml of LPS. One ml samples of supernatant were harvested at intervals up to 24 hr and were assayed for CF on L-929 targets. The resulting lyt activity (mean _+ S.E.) is shown for each sample. FIGURE 4. Macrophage monolayers were established in 2 wells in medium with LMS. Doses of actinomycin D, cycloheximide, monensin, or tunicamycin ranging from 0. 10 ug/ml, were added to parallel series of wells and incubated at 37° for 30 in before triggering with 100 ng/ml of LPS. Supernatants harvested 4 hr after triggering were dialyzed and assayed for CF on the L-9 target. The activity in units/ml for test cultures was normalized to the value for non-drug treated control cultures and expressed as a percentage (mean _+ S.E.) thereof.

FIGURE 5. Cytolytic Factor was pretreated with various levels of TLCK for 1 hr at 37°; after overnight dialysi against DPBS, TLCK-treated and control cultures were assayed on L-929 cells. Similarly, the Factor was coincubated with various levels of TAME or catalase on 929 cells. Lytic activity in treated CF samples was compared to a control Factor preparation, and expressed a percentage (mean _+ S.E.) thereof.

FIGURE 6. Purified preparations of Cytolytic Factor an necrosin were incubated with various doses of anti- necrosin antiserum at ambient temperature for 1 hr. Th mixtures were then assayed for lytic activity on L-929 cells; these values were normalized to the nonantibody- treated control value and expressed as a percentage thereof.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention, in its most general and overall scope, is directed to methods and compositions cytolytically inhibiting tumor cell growth and 10 proliferation. Compositions of the invention are gener¬ ally defined as including a factor, termed Cytolytic Factor (CF) _r which exhibits very high cytolytic activity against a number of various tumor cell targets and very low, if any, such activity against normal cell targets.

This cytolytic factor is found to be generally soluble and stable in physiologic buffers, for example, phosphate-buffered saline at neutral pH, for long periods of time (greater than six months) when stored at 4°C. As will be appreciated, this allows for ready formulation of the factor into pharmaceutically acceptable vehicles for administration to patients.

The Factor may be characterized in physical terms according to its apparent molecule weight upon gel exclusion chromatography. In general, the factor is fou to exhibit a molecular weight between about 140 and abou 160 kilodaltons, with a peak of activity generally observed at about 150 kilodaltons, when the factor is subjected to chromatography on Sephacryl S-200 (Pharmaci Uppsala, Sweden).

In certain embodiments, the Cytolytic Factor may be characterized according to various biological properties. For example, the factor is found to maintain its cytolyti activity in the presence of serum, thus demonstrating it usefulness for direct administration to patients. Moreover, the factor appears to exert its cytolytic acti by way of a mechanism which involves a "neutral" proteas function. For example, in certain embodiments the Facto is found to be inhibited_, in a dose-dependent fashion by protease inhibitors such as TAME (alpha, N-tosyl-L-argin methyl ester) or TLCK (alpha, N-tosyl-L-lysyl- chloromethylketone) . Interestingly, Cytolytic Factor is immunolo ically cross-reaction with Necrosin, a cytotoxic derived from murine macrophage cell line. Necrosin exists as a holotoxin having two molecular weight forms, 70 and 55 kilodaltons, which are multimers of a 15 kilodolten protein (see, e.g., Kull et al.-, (1984), Proc. Natl. Ac Sci. U.S.A. , j$l_:7932). The molecular or etiological ba for this immunological cross-reactivity is unknown and based on the observation that polyclonal antiseru rais against Necrosin has been found to inhibit the cytolyti activity of the Factor.

In still other embodiments, the Cytolytic Factor i shown to be a protein whose cytolytic activity is dependent upon an intact glycosylation apparatus of the macrophage cells from which it is obtained. For exampl the production and release of Cytolytic Factor by macrophages is found to require transcription, transla¬ tion, glycosylation and an intact secretory apparatus o the producing cell, as is evident from its inhibition following treatment of the macrophages with actinomycin cycloheximide, tunicamycin and monesin, respectively, during culturing of the cells. Moreover, the Factor itself appears to lose its biological activity followin treatment with trypsin or heating to 100°C for 10 minut

In terms of its cytolytic antitumor activity, the Cytolytic Factor of the present invention has proved highly active in killing tumor cells of lines which are well accepted in the art as appropriate to demonstrate such activity. Moreover, the Factor is found to exhibi much less activity against normal cells. For example, certain factor preparations were found to be highly act against the murine L-929 or EMT-6 adeno-carcinoma and various human tumor, while much less active against "normal" cell lines such as normal mouse lung fibroblasts or peritoneal exudate cells.

Various properties, in addition to its molecular weight, serve to illustrate the novelty of the present

Factor as compared to previous tumoricidel or tumoristati factors of biological origin. For example, Cytolytic Factor has no apparent effect on the catalysis of iron- release by tumor cells, and is thus distinguishable from various "iron-releasing" monokines which have been described (Klostergaard et al. (1987), Ly phokine Res, j:14). Moreover, partially purified preparations of Cytolytic Factor exhibit only a minor ability to inhibit the mitochondrial electron transport system, thus demonstrating that Cytolytic Factor is distinguishable from Respiratory Inhibitory Factor (Kilbourn et al., supra. )

The foregoing characteristics further demonstrate that Cytolytic Factor preparations may be obtained in accordance with the invention in a substantially purifie form as determined by the relative absence of iron- releasing monokines and Respiratory Inhibitory Factor, such additional products being often present in super- natants which^" have been "conditioned" with activated macrophages in the manner of the invention.

Preparation of the factor of the present invention generally involves several steps. • The first step involv the preparation of a supernatant derived from cultures o activated macrophages. Although not mandatory, the macrophages to be cultured are generally first primed or "initiated" ij vivo by injecting a mammal (a mouse is preferred, but other, mammals such as guinea pigs, hamsters, rats, rabbits, etc., may be used) with an effective amount of a macrophage activation initiator. BCG is a preferred activator, but other agents known to induce cytotoxic activation of macrophages iτ_ vivo may^ used as well. These compounds include, but are not limited to, Immunotone (American Biotechnology Co.), C. Parvum, supernatant derived from mitogen stimulated mur T-cells, gamma-interferon and muramyl dipeptide. Of course, the amount administered will vary with the particular initiator or activator used, but in general, effective amount is that amount required to produce cytotoxic macrophages, recognizable by established criteria. More particularly, when BCG is used, a dose approximately 2 x 10 colony forming units is the preferred dose. Intraperitoneal administration is preferred.

Any regimen of injection sufficient to induce activation of the macrophages in vivo may be used, however, in a preferred embodiment, the mammals are firs injected with the activator 25 days before the macrophag are harvested, then "boosted" with a second dose 4 days before harvest. Furthermore, it should be appreciated that iri vivo activation is not strictly required where t macrophages are sufficiently activated in vitro as described below.

After the _in_ vivo activation step, if such a step i employed, the macrophages are harvested from the host mammal by peritoneal lavage. This technique generally involves injecting the peritoneal cavity of the animal with a liquid medium, such as a physiologic buffer, massaging the peritoneal area, and draining the peritone exudate from the animal. At this stage, the peritoneal exudate contains a mixture of cell types, including the activated macrophages. The cell mixture may be enriched for the activated macrophages by any of a number of .methods known to separate macrophages from contaminating cell types. These include, but are not limited to: treating the mixture which includes the contaminants wit 5 specific antibodies and complement; and allowing the macrophages to adhere to a solid support such as a tissu culture dish, and washing away non-adherent contaminants In a preferred embodiment of the present invention, the macrophages are selected by culturing them in serum free

10 medium in a plastic tissue culture dish, allowing them t adhere to the plastic, and washing away non-adherent con taminants. This procedure yields a population which is approximately 90% pure. The macrophages may then be recultured in any of a number of suitable culture media.

15 Such media are well known to those skilled in the art.

In general, in order to obtain optimal production o Cytolytic Factor, the macrophages should be treated with triggering agent shortly after they are established in

20 culture. In a preferred embodiment, the triggering agen is endotoxin derived from E. coli, however it will be appreciated that other agents, such as other bacterial endotoxins, muramyl dipeptide, and phorbol mysistate acetate may be used as triggering agents. ϊn fact, wher

25 the macrophages have been sufficiently activated _in vivo this step may be omitted entirely. However, Factor obtained from cells cultured in media containing 0.1% lactalbumin hydrolysate (LAH) was generally found to be highly unstable. Thus, LAH should not be employed as an -. 30 activator or otherwise present in the activation culture media in amounts as high as 0.1%.

Production of macrophage conditioned supernatant is generally completed after four to twelve hours of cultur 35 Since macrophages triggered with endotoxin as described above have been shown to produce the Cytolytic Factor almost immediately after triggering, and since the Fact is detectable in cultures incubated up to at least 24 hrs., shorter or longer' incubation times may be used. However, four to twelve hours is considered optimal. A any rate, at the termination of the selected incubation period, the resultant conditioned supernatant is remove from the culture, for example, by centrifugation, siphoning or filtration. It may then be stored at -20° until needed for further use. The cytolytic action of supernatant can be titered on actinomycin D-treated L-9 cells. It has generally been found that 1-2 ul of conditioned supernatant produced by culturing 1 x 10 macrophages in 1 ml of culture medium is effective to provide a 50% killing of 25 x 10³ L-929 cells in 18-24 hours. Therefore, a unit of cytolytic activity is defi as this amount.

Of course, the advantages of the present invention are realized by identification and separation,of a solu Cytolytic Factor having a relative molecular weight between 140,000 and 160,000 datons from the macrophage conditioned supernatant. The Cytolytic Factor may be identified and separated by any of a number of selectio techniques known to those skilled in the art of separati biological molecules. These techniques include, but are not limited to, selective ultrafiltration, ultracentri- fugation, preparative gel electrophoresis, molecular exclusion chromatography, ion exchange chromatography an the like. However, a preferred method for separating an identifying the monokine entails clarification of the conditioned supernatant by centrifugation, followed by ultrafiltration and subsequent gel exclusion chroma¬ tography. More specifically, the ultrafiltration step involves placing the conditioned supernatant in a stirred cell apparatus preferably pressurized with nitrogen. The apparatus contains a membrane which retains only molecule above a specified molecular weight. Several membranes ar commercially available and may be used for practicing the invention; however, a YM-10 membrane, which retains molecules with a molecular weight greater than 10,000, is preferred. Using this ultrafiltration protocol, the conditioned supernatant containing the Factor is concentrated approximately 20 to 100-fold.

As previously noted, further purification of the Factor can be achieved by a number of techniques. However, in a preferred embodiment, the concentrated supernatant is chromatographed on a gel filtration column Numerous gels suitable for such chromatography are known by those skilled in the art of chromatography. These include, but are not limited to, agarose gels, Sepharose gels, Sephadex gels, Sephacryl gels, and polyacrylamide gels. However, a Sephacryl S200 gel filtration column is preferred. The column may be equilibrated with any suitable equilibration solution or physiological buffer; however, Dulbecco's phosphate buffered saline is pre- ferred. Fractions eluting from the column in the same volume with molecular weight standards ranging from 140 t 160 kilodaltons are a particularly preferred source of t Factor.

The purified Cytolytic Factor may be formulated into a number of preparations suitable for treatment of tumor patients. The monokine may be formulated into such preparations in neutral or salt forms. Pharmaceutically acceptable salts include the acid addition salts (formed with the free amino groups of the monokine)which are formed with inorganic acids such as, for example, hydro chloric or phosphoric acids, or such organic acids as acetic, oxalic, tartaric, andelic, and the like. Salt formed with the free carboxyl groups may also be derive from inorganic bases such as, for example, sodium, pota sium, ammonium, calcium, or ferric hydroxides, and such organic bases as isopropylamine, trimethylamine, 2-ethylamino ethanol, histidine, procaine, and the like

In general, the Factor may be admixed with any vehicle in which it retains function. Of course, selec tion of the proper vehicle will depend on the manner in which the Factor will be used. In general, suitable vehicles include aqueous physiological solutions, salin dextrose, glycerol and the like, or combinations thereo

Obviously, any preparation of the monokine to be u in human tumor therapy must be formulated with a nontox excipient. In theory, any of a number of pharmaceutica excipients may be used, so long as the factor retains i function when it is administered in the presence of suc excipient. The factor may be administered by any of a number of techniques known to those skilled the art of administering biologic compounds. Such techniques in- elude, but are not limited to, intravenous injection, subcutaneous injection, intraperitoneal injection, and continuous infusion. Furthermore, it is contemplated t the factor could be administered in conjunction with ot chemotherapeutic agents known to those skilled in the a

The following examples demonstrate methods for preparing, assaying, and characterizing the Cytolytic Factor. Of course, alternative, and possibly more effective modes for practicing the invention may be determined in the future. The examples are simply mean to demonstrate the best mode for practicing the inventio known to the inventor at the present time.

EXAMPLE I

Preparation of Cytolytic Factor

A. Preparation of BCG-Activated Macrophages

BCG-activated macrophages produced according to the protocol described below were used to study the mechanis of activated macrophage induction of Cytolytic Factor release. They were also used to prepare conditioned supernatant from which the Factor was purified.

1. Animals

Six- to eight-week old male CD-I mice were obtained from the University of Texas, Science Park, Bastrop, TX. Ten- to twelve-week old male CB6F1 mice (Balb/cAnN x C57B1/6N) were provided through the breedin program of the Department of Tumor Biology, The Universi of Texas M.D. Anderson Hospital and Tumor Institute, Houston, TX.

2. Preparation of BCG-Activated Macrophages

Macrophage monolayers were obtained by peri- toneal lavage of mice that had been injected i.p. with 2

10 colony forming units of BCG (Trudeau Institute, Saranac Lake, NY), 25 and 4 days before harvest. The peritoneal exudate cells were suspended in Ca - and

Mg -free phosphate buffered saline (PBS), collected bbyy centrifugation, adjusted to the proper density (1-2 x 10 cells/ml) and then allowed to adhere to plastic for 4 h in serum-free culture medium at 37°C. After this time, non-adherent cells were removed by washing with PBS. Remaining cells were judged to be > 90% macrophages by morphological and functional criteria. These cells wer then utilized either in cocultures with radiolabeled tu cells to determine cytotoxicity, or for supernatant production.

B. Procedures Employed to Characterize the Facto

The following procedures are protocols which have been employed by the inventor to illustrate variou preferred embodiments of the invention.

1. Characterization of the target cell spectrum of Cytolytic Factor in vitro

Cytolytic Factor was routinely assayed in vit employing the murine L-929 target (C3H/An; ATCC) in a modification of the microcytotoxicity assay of Fisch an

Gifford (1983), J. Immunol. Meth., _57:311). Briefly

3 described, 15 x 10 cells in 150 microliters per well o

96-well plate were incubated overnight after' seeding. Fifty microliters of 4 ug/ml of actinomycin-D (Sigma) w added to each well immediately before assay of CF super natants or fractions from biochemical separations.

The Cytolytic Factor preparations were titered on drug-treated targets, and incubation carried out over¬ night. 'Fifty microliters of a 0.02% neutral red soluti were added to each well; after 1-2 hr, the wells were washed with DPBS, and incorporated stain assessed after treatment with citrate-ethanol (0-.1 M citric acid in 7 ethanol-water) . Quantitation of A_ς._Q was achieved using Titertek Multiskan. Units of lytic activity were define as the reciprocal of the dilution of the Cytolytic Facto that caused 50% reduction in dye uptake. In some experiments, cell viability was enumerated by use of the MTT assay Mosmann (1983), J. Immunol. Meth., _6Jj: 55. Uni of lytic activity were defined as the reciprocal of the dilution of Factor that caused 50% reduction in MTT formazan production, measured at 570 nm.

2. Tumor cytotoxicity of BCG-activated macrophages in vitro

In demonstrating the in vitro antitumor activi of activated macrophages, BCG-activated macrophage were seeded at 2.5 x 10 /well in 96-well plates (- .33

2 cm /well). EMT-6 and L-929 targets were labeled as single-cell suspensions with Na~ Cro. (ICN, Irvine, CA) in a 15 ml polypropylene tube (Corning Glass Works,

Corning, NY) following treatment with trypsin-EDTA. 50 uCi 51Cr in approximately 200 ul volume was incubated wi

2-10 x 10 targets for 1 hr. After extensive washing, t targets were added to the macrophages at effector to target ratios of 10-25:1 in a final volume of 200 ul medium (DME-F12; GIBCO, Grand Island, NY) with fetal cal serum.

Parallel experiments were conducted using resident macrophages as effector cells; these macrophages were obtained by an identical lavage procedure on nonimmunize mice. Some cultures of either resident or BCG-activated macrophages were triggered with 250 ng/ml lipopoly- saccharide (LPS; see below) during coculture with tumor cells. After 16-20 hr, released Cr was quantitated in a Beckman Gamma 8000 (Beckman Instruments, Fullterton, CA Total release was determined by counting the appropriat volume of the target cell suspension used for seeding; spontaneous release was measured in cultures of target cells with no macrophages. The presence of LPS did not influence spontaneous release.

The percent cytotoxicity was calculated by the following expression:

ercen Experiment Release - Spontaneous Release cytotoxicity = ^ _

Total Release - Spontaneous Release

The ability of macrophages to produce the Factor i the presence of serum substitutes, the kinetics of its production, and the effect of macrophage density were determined by seeding either 25 x 10 or 100 x 10 macrophages in a 100 mm dish; after adherence, these cel were cultured in medium with LAH (0.1%) or LMS (10%) an triggered with LPS (100 ng/ml). Samples of supernatant were harvested at various ti epoints and were frozen at 20°C until they were bioassayed on L-929 cells (see below) .

Other cellular parameters for the production/releas of Cytolytic Factor were examined by incubating macrophages with various doses of actinomycin-D (Sigma) , cycloheximide (Calbiochem-Behring) , tunicamycin (Sigma), or monensin (Sigma) 30 in before and during the four hr of LPS triggering; Cytolytic Factor was quantitated on L 929 targets as described below. The Cytolytic Factor titer in units/ml for drug-treated macrophages was compared to a non-drug-treated control and expressed as percentage thereof.

3. Functional characterization of Cytolytic Facto 5

Protease characteristics of Cytolytic Factor were probed by incubation of Cytolytic Factor, partially purified by conventional molecular sieving on Sephacryl S-200 and HP anion-exchange LC, with TLCK (Sigma)

10 dissolved in 100 M pH 7.4 borate or 50 mM pH 7.4 Hepes buffer; the control was the same buffer alone, in the absence of inhibitor. TLCK levels ranged from 0.2 to 5. mM. Incubation was maintained for 1 hr a 37°. The samples were dialyzed overnight against DPBS and assayed

15 for lytic activity. The Cytolytic Factor titer in units/ml for the TLCK-treated samples was compared to a non-TLCK-treated control and expressed as a percentage thereof. TAME (Sigma) was dissolved in 100 M pH 7.4 borate and coincubated with the Factor on L-929 cells in

20 the lytic assay; final TAME concentrations were 0.56 to 5.0 M. The CF titer in units/ml in the presence of the various levels of TAME was compared to a non-TAME-treate control and expressed as a percentage thereof.

25 Oxygen metabolite-dependent functions of Cytolytic Factor were probed in two ways: by enzymatic degradatio ' of H.O. produced, and by direct measurement of H-O, or 0 released during lysis. Partially-purified Factor was coincubated with catalase (Sigma) during the lytic assay

30 enzyme was dissolved in borate buffer to give a final concentration range of 222-2000 units/ml. The Cytolytic Factor titer in the presence of the various levels of catalase was compared to a non-enzyme-treated control an was expressed as a percentage thereof.

35 Release H₂0₂ and 0₂~ were measured by the microeli method of Pick and Mizel (1981) J. Immunol. Meth., _46_:211). 2 x 10 L-929 cells in the presence of Cytoly Factor or Cytolytic Factor alone, were coincubated with either the H-,0-, or 0_— reaction mixture. The H_0_ reaction mixture was composed of 0.56 M phenol red (Sigma) with 19 units/ml horseradish peroxidase (Sigma) with or without 0.25 ng/ml PMA in Hanks balanced salt solution without phenol red. The 0₂— reaction mixture 1.2 mM ferrictochrome C (Sigma) with or without PMA in Hanks balanced salt solution without phenol red. Absorbance was determined on a microelisa reader (Dynatech, Alexandria, VA) at 600 nm for H₂0₂ (after alkalanization with 10 ul in NaOH) and at 550 nm for 0₂ The reactions were monitored at 1, 4 and 18 hrs of incubation at 37°C.

In experiments designed to evaluate the role of arginase in the Cytolytic Factor-mediated mechanism, arginine (Sigma) was added to the medium of the lytic assay during incubation with CF; the amino acid was dissolved in borate buffer to give a final concentration of 667 ug/ml. The lytic activity of CF against L-929 cells in the presence of DME-F12 with added arginine was compared to the activity of CF in DME-F12 medium alone, and expressed as a percentage thereof.

The effect of protease inhibitors in serum on lytic activity was evaluated by assay of Cytolytic Factor on L 929 targets as in the assay described below, except that the serum-free medium of Neuman-Tytell (GIBCO) was used incubate targets instead of DME-F12 with 10% FCS. The lytic activity in the former medium was compared to the activity in the presence of serum and was expressed as a percentage thereof. The heat stability characteristics of Cytolytic Factor were explored by incubation of Cytolytic Factor i screw-capped vials for the appropriate times in water baths at the appropriate temperatures. After treatment, the tubes were brought to ambient temperature and assaye on L-929 targets as described below. The lytic activity in the heated preparations was expressed as a percentage of the control preparation which had been held at 4°C.

The dependence of lytic activity of Cytolytic Facto on protein structures were examined by incubation of Cytolytic Factor with trypsin (GIBCO) or control buffer for 15 in at 37°C. The reaction was quenched with 10% FCS, and the samples subjected to bioassay. The lytic activity in the enzyme-treated samples was expressed as percentage of the non-treated control.

The effect of Cytolytic Factor on the electron transport chain (ETC) of the mitochrondria of target cel was determined by an assay as follows. Briefly describe EMT-6 targets were seeded at 5 x 10 /100 ul/well in 96- well plates. After overnight incubation, Cytolytic Factor, purified by molecular sieving as described belo was titered on these targets. A stock preparation of supernatant from BCG-activated macrophages was assayed o the targets as a positive control for respiration inhibitory factor (RIF). After 18-24 hr, at 1:10 diluti of 5 mg/ml MTT (Sigma) was added. After incubation for hr at 37°C the MTT-formazan produced by ETC-mediated reduction of MTT was resolubilized with acidified (0.04 HC1) isopropanol with 0.1% SDS. The blue formazan was quantitated at 570 nm in a Titertek multiscan (Flow Laboratories, McClean, VA). Units of RIF activity were calculated as the reciprocal of the dilution from one ml which caused 50% of the maximum reduction in formazan production.

The ability of Cytolytic Factor to mediate releas intracellular iron from tumor cells was determined by

59 prelabeling L-1210 and EMT-6 cells with Fe (ferrous citrate, ICN). Following extensive washing, targets w

3 seeded in microwells at 10 x 10 /well and Cytolytic Fa was titered on these labeled targets. After 18-24 hr,

59 amount of Fe released into the supernatant was deter

59 mined in a gamma counter. Spontaneous Fe release wa determined from supernatants of target cells alone.

The EMT-6 murine adenocarcinoma line (obtained fr Dr. Gabriel Lopez-Berestein, M.D. Anderson Hospital,

Houston, TX), MCA-1 sarcoma (obtained from Dr. Dolph 0

Adams, Duke University Medical Center, Durham, NC), SV and 3T3 fibroblasts (obtained from Dr. Karen M. Miner,

Merck Laboratories, Rahway, NJ) and B16 melanoma were similarly in microcytotoxicity assays, with or without

3 drug treatment; however for the EMT-6 assay, 5 x 10 c

3 were seeded per well, for the SVT2, line 10 x 10 cell

3 were seeded per well, and for the 3T3 line, 25 x 10 c were seeded. L-1210 lymphoblastic leukemia cells were obtained from Dr. Berestein. Normal mouse lung fibrob cultures were established by mincing of aceptically removed lungs of 8 day old mice.

C. Purification of Cytolytic Factor From Macrophage Conditioned Supernatant

The following procedures were performed in order purify the Cytolytic Factor and determine its approxima molecular weight. Murine macrophages were seeded at 2 10 per 100 mm tissue culture dish. After 4 hr of adherence, the macrophages were washed, recultured in 30 ml of DME/F-12 medium, and triggered for 2-6 hrs with 10 ng/ml of endotoxin (bacterial lipopolysaccharide; phenol extracted E. coli serotype 0128:B12; Sigma Chemical Co., St. Louis, MO). The resultant conditioned supernatant w collected by centrifugation and frozen at -20^βC until further use.

In order to further characterize the Factor in term of spectrum of activity and molecular weight and to demonstrate its further purification, the conditioned supernatant was subjected to various in vitro character izations and molecular weight fractionization. For molecular weight determinations, the supernatants were thawed, clarified by centrifugation, and concentrated on YM-10 membrane in a stirred-cell apparatus. The concentrates were subjected to molecular sieving on Sephacryl S-200 (Pharmacia, Uppsala, Sweden) in a 2.5 x cm column (BioRad, Richmond, CA) equilibrated with DPBS (Dulbecco's Phosphate Buffered Saline). Two ml fraction were collected at a flow rate of 30 cm/hr. Molecular weight markers were blue dextran, immunoglobulin G, hemoglobin, and phenol red. Fractions were assayed for lytic activity as disclosed above in Section' B.l.

FIGURE 1 is a chromatogra which illustrates the elution profile of macrophage conditioned supernatant applied to Sephacryl S200 in the foregoing manner. Each fraction was assayed on both L-929 and EMT-6 targets. A will be appreciated from FIGURE -1, a single cytolytic species of•approximately 150 kD was detected by bioassay on either type of target cell; the EMT-6 target appeared to be about 20-fold more sensitive to the Factor than th L-929 target, when both were treated with actinomycin D. No species in the 40-50 kD range was detected. D. Alternative Triggering Agents for Cytolytic Factor Production

The ability of various agents to trigger th production of Cytolytic Factor _in vitro by macrophages activated i__ vivo with BCG was examined. Alternative triggers for Factor production examined included polyinosinic-polycytidylic acid (PIC; P-L Biochemicals

Inc., Milwaukee, WI), tuftsin (Calbiochem-Behring, San Diego, CA) , stearoyl-mura yl dipeptide (SMDP; Syntex

Laboratories, Palo Alto, CA) , 4-B-phorbol, 12-B-myrist

13-a-acetate (PMA; Sigma), and the calcium ionophore

5 A23187 (Sigma). Macrophages established at 5-10 x 10

2 cm / ml were incubated for four hr with these agents i medium with 0.1% LAH and supernatants were assayed for cytotoxic activity on L-929 targets as described. The lytic activity in these supernatants in units/ml was normalized to the activity found in supernatants of macrophages without triggering, and expressed as a stimulation index.

As will be appreciated from the data displayed in Figure 2, both SMDP and LPS were effective triggers fo release, showing a dose-dependent response from 10-100 ng/ml, with a maximum stimulation of 7-9 fold compared spontaneous release at the highest dose tested. At th lower dose range tested (3 ng/ml), PMA proved to be capable of triggering CF release to somewhat lower lev (about 4-fold). PIC, tuftsin and the calcium ionophor A23187 were not effective triggers at the dose employe for these experiments. E. Direct Tumor Cytotoxicity in Coculture of Macrophages Subjected to Different Priming and Triggering Signals

⁵¹Cr-labeled EMT-6 and L-929 targets were coincubated with resident or BCG-activated macrophages, with or without endotoxin-triggering, for 16-20 hr. Analysis of isotope release (Table 1) indicated that resident macrophages were incapable of significant lysis (up to 7%) of either EMT-6 or L-929 targets, unless puls with endotoxin; under these conditions, the level of lys of L-929 targets rose to 15-25%. BCG-macrophages, in th absence of LPS exerted a similar level of cytotoxicity compared to LPS triggered resident macrophages; however, upon triggering with endotoxin, the macrophages demonstrated - 20% lysis of EMT-6 targets, and - 70% lysis of L-929 targets. Thus, these experiments indicat the need for priming iτ_ vivo and triggering i^ vitro for the highest expression of macrophage tumor cytotoxicity.

TABLE 1

Direct Tumor Cytotoxicity of Macrophages in CoCultur

Target -Percen Macrophageb LPS^C Cells Cr Rel

Resident _, EMT-6 6.9 + 0.

Resident + EMT-6 6.3 __ 0.

BCG - EMT-6 14.3 + 0.

BCG + EMT-6 23.8 + 0.

Resident - L-929 7.3 + 0.

Resident + L-929 26.1 + 0.

BCG - L-929 24.1 + 0.

BCG + L-929 68.1 + 0.

^a 16 hr assay ^b 250, 000/.33 cm^{2 c} 250 ng/ml ^d Effector:Target Ratio - 25:1 ^e Mean + SEM ^f Effector:Target Ratio - 8:1

F. Tumor Cytotoxicity Mediated by Supernatants from Macrophages subjected to Different Priming and Triggering Signals

Monolayers of resident or BCG-activated macrophages, exposed or not exposed to endotoxin triggering, were allowed to condition medium containing LMS for 18 hr. The harvested supernatants were assayed for cytotoxicity on actinomycin-D treated EMT-6 targets, In Table II the results are seen.

Table 2

Tumor Cytotoxicity of Supernatants from Macrophages units/ml

Resident 319 +

Resident 32,900 + 6,

Primed.e 91 + Primed 107, 000 + 47,3

Primed & Boosted 165

Primed & Boosted 127, 000 + 13,8

18 hr. supernatant production in the presence of LMS. ^bl x 10⁶/2cm²/l ml ^c 100 ng/ml added to macrophage cultures. Mean _+ SEM; 18 hr. assay on actinomycin D-treated EMT targets. e Mice received 2 x 107 colony forming units of BCG i.p

25 days before lavage. f Mice received 2 x 107 colony forming units of BCG i.p

25 and 4 days before lavage.

As assayed on actinomycin D treated EMT-6 cells .

A similar dependence on macrophage priming and triggering signals for supernatant cytotoxicity was observed as for direct tumor cytotoxicity. Resident macrophages were incapable of producing a cytolytic mediator unless triggered, in which case, significant supernatant cytotoxicity was observed (- 33,000 units/ml) Macrophages obtained from mice which had only received a single priming injection of BCG also showed a strong dependence on endotoxin triggering for supernatant cytotoxicity, under which conditions high levels of a cytotoxic mediator were detected (- 107,000 units/ml). BCG-primed mice further received an irv vivo boost of BCG, these macrophages displayed the greatest endotoxin- dependent production of the factor (- 127,000 units/ml). This, then formulated the basis for the production of conditioned supernatant in these studies. _

G. Kinetics of and Effects of Macrophage Density and Serum Substitutes on CF Production/Release Following Triggering

Peritoneal macrophages from BCG-i mune mice were established as- adherent monolayers of either 25 x 1 or 100 x 10 6 cells on 78 cm2 dishes. The cells were triggered with 100 ng/ml of LPS and allowed to release

Cytolytic Factor in DME-F12 medium supplemented with LMS (10% v/v) or LAH (0.1%). Release was monitored by removing an aliquot of supernatant at various time point and subjecting it to bioassay on L-929 cells. The activity in lytic units/ml for each sample are shown in

Figure 3.

Cytolytic Factor was released rapidly after triggering macrophages with LPS, reaching peak levels in 4-8 hr. On a per-cell basis, Cytolytic Factor productio was superior with the lower macrophage density (25 x 1

2 cells/78 cm ). Media supplemented with LMS appeared t support higher levels of Cytolytic Factor production b the macrophage at either density tested, than media wi LAH. This is in part likely due to the instability of factor generated in the medium containing LAH; in contrast, Cytolytic Factor released in medium with LMS quite stable and amenable to further biochemical manipulation.

H. Cellular Processes Involved in Macrophage Production/Release of CF

Macrophage monolayers were incubated in LAH- containing medium with actinomycin-D, cycloheximide, monensin, or tunicamycin over a dose range from 0.1 to ugt/ml. After 30 min of preincubation, LPS (100 ng/ml) was added and supernatants were collected after four hr After overnight dialysis against DPBS, each sample was assayed for CF activity on L-929 targets; the lytic activity was normalized with respect to that produced b LPS-triggered macrophages in the absence of drug pretre ment, an expressed as a percentage thereof (Figure 4).

At the lowest level tested (0.1 ug/ml), actinomyci caused > 50% inhibition of CF production/release; with log 10 higher dose, inhibition was > 95%. Consistent w this dependence on RNA synthesis, cycloheximide pretrea ment of macrophages caused a similar pattern of inhibit of Cytolytic Factor production/release. Active secreto processes appear to be involved in Cytolytic Factor release, as monensin pretreatment of macrophages caused dose-dependent inhibition (about 50% inhibition at 2 ug/ml). Tunicamycin treatment of macrophages before triggering caused a similar pattern of inhibition (50% inhibition at - 2.5 ug/ml) . This suggests that either a properly N-glycosylated Factor is required for secretion or that glycosylation is required or expression of biological activity.

I. Functional Characterization of CF

The possible role of protease-dependent functions and hydrogen peroxide in the Factor-mediated tumoricidal mechanism was explored with selected inhibitors of serine proteases and with catalase. CF wa pretreated with TLCK over a dose range before assay on L 929 cells, or coincubated on L-929 targets with TAME or catalase. The lytic activity of a Cytolytic Factor preparation after these manipulations was normalized to that fo the Cytolytic Factor preparation without manipulation, and expressed as a percentage thereof (Figure 5) .

Both the irreversible protease inhibitor, TLCK, and the reversible inhibitor, TAME, blocked Factor-mediated lysis of L-929 targets, in a dose-dependent, TAME, block Cytolytic Factor-mediated lysis of L-929 targets, in a dose-dependent manner; 50% inhibition was achieved with 2mM TLCK and with - 1 M TAME. In contrast, the partial inhibition (- 10-15%) observed with catalase did not sho a detectable dose response; increasing the enzyme level from - 200 to - 2000 unit/ml caused no increase in inhibition.

Cytolytic Factor was significantly labile when treated with trypsin or upon heating at 100° (Table 3). Arginine addition had no apparent effect on lytic activity. Lytic activity of Cytolytic Factor on L-929 targets was essentially identical whether serum protease inhibitors (e.g., alpha--macroglobulin) were excluded not.

TABLE 3

Effects of Various Treatments on Cytolytic Factor Activi

Percentage of Control Treatment Lytic Activity (mean + S.E.

Control³ 100 _+ 9.6

Trypsin 5.3 _+ 1.2

Heat^c

37* 92.3 + 5.6

100° < 1% Arginine^d 110.4 + 15.^4

Neuman-Tytell^e 95.3 _+ 13.0

^a Aliquots of a standard CF preparation,partially purifi by conventional molecular sieving and HP anion-exchang LC, kept at 4°C until bioassay.

0.25% trypsin treatment for 15 min. at 37°, followed by addition of FCS to quench trypsinization. ^c 10 min duration of treatment.

667 ug/ml in DME-F12 medium coincubated on L-929 cells with CF. ^e CF assayed on L-929 targets cultured with Neuman-Tytel medium instead of DME-F12 with 10% FCS. I. Characterization of Anticellular Effects of

Cytolytic Factor on Allogeneic Targets in vi

Cytolytic Factor obtained from BCG-activated macrophages from CD-I mice was assayed for cytolytic/cytostatic effects on these target cell lines _in vitro: L-929, EMT-6, MCA-1, B16F1, L-1210, SVT2, an 3T3. A single Cytolytic Factpr preparation, partially purified by molecular sieving was used for all these cytotoxicity studies summarized in Table 4. The L-929 target was sensitive to the Factor in a 24 hr assay. B cytostatic and cytolytic (nonstaining with neutral red) effects were evident. Pretreatment of the L-929 cells with actinomycin-D caused growth inhibition of both control and Factor-treated targets, and therefore allow evaluation solely of the cytolytic effects of the Facto Drug treatment clearly caused marked sensitization of t L-929 target to Cytolytic Factor (about a 70-fold incre in lytic activity).

The EMT-6 target demonstrated marked resistance to the Factor; targets treated with the highest dose emplo were > 99% viable as determined by neutral red staining However, actinomycin-D treatment of these targets rende them even more sensitive to CF (3-4-fold) than the drug treated L-929 cell. The MCA-1 target showed a similar, but less pronounced, ability to be sensitized by drug treatment. The B16F1 melanoma showed significant resistance to the Factpr, as did the L-1210 target, whereas the SVT2 target was sensitive. Although the

Factor had no apparent effect on normal mouse lung fibr blasts, the 3T3 fibroblast line showed some sensitivity a 40 hr assay. Cytolytic Factor treatment of normal peritoneal exudate cells from BCG-immune mice had no effect in 48 hr on the viability of either adherent ₍polymorphonuclear leukocytes and macrophages⁾ or nonadherent (lymphocytes) cells.

TABLE 4

Cytotoxic Effect of Cytolytic Factor on Allogeneic Tumor and Normal Targets In Vitro

Target Lytic Units/ml

L- 929³ 128 L L--992299"^b 8,970

EMT-6³ << 27

EMT-6 27,200

MCA-1 ^C 78

MCA-l^d 205 B B1166FF1^C* < 20

L-121 0^β « 40

SVT2^f 775

3T3^f 177

Normal mouse lung fibroglasts^c < 15

^a 24-hr cytotoxicity assay; neutral red stain.

24 hr cytotoxicity assay; targets treated with 1 ug/ of actinomycin D; neutral red stain. ^c 48 hr cytotoxicity assay; neutral red stain.

48 hr cytotoxicity assay; targets treated with 0.3 ng/ml; of actinomycin D; neutral red stain. e 24 hr 51Cr release assay.

40 hr cytotoxicity assay; MTT stain. A preparation of Cytolytic Factor, adjusted to a 4 titer of about 10 lytic units/ml when assayed on actinomycin D treated L-929 targets, was assayed for RIF activity on EMT-6 targets using a microcolormetric assay based on MTT reduction by the ETC. .A preparation of conditioned supernatant (CS) from LPS triggered BCG- activated macrophages from CD-I mice cultured in LMS for 18 hr was used as one positive control; a preparation of RIF obtained by molecular sieving of CS on S-200 was used as another positive control. In Table 5 are seen the results. Clearly, compared to the CS, the purified Cytolytic Factor demonstrates only a minor ability to inhibit ETC-dependent reduction of MTT, but a higher cell-lytic potency. In contrast, the partially purified RIF is capable of strongly inhibiting MTT-formazan production, but has a low cell-lytic activity.

41 TABLE 5

Effect of Cytolytic Factor on Electron Transport Chain-Mediated Endogenous Substrate Oxidation

Inhibition of

Lytic MTT-For azan

Preparation Units/ml Production (units/ml

CF^C - 10, .000 30 + 10

CS^d - 800 1, 000 + 75

RIF^e — 300 3,000 + 500

Assayed on actinomycin D-treated L-929 targets.

Assayed on EMT-6 targets.

Purified by molecular sieving of CS of Sephacryl S-20

Conditioned supernatant obtained by culturing BCG- activated macrophages for 18 hr in LMS with LPS- triggering.

Purified by molecular sieving of CS on Sephacryl S-20

The same preparation of Cytolytic Factor was used t investigate its ability to mediate loss of intracellular iron in lytically resistant L-1210 and EMT-6 targets. No release of 59Fe from these targets above spontaneous release in response to Cytolytic Factor was detectable

(Table 6); in contrast, iron was released upon coculture of the targets with activated macrophages.

43 TABLE 6

Release of Intracellular Iron from EMT-6 and L-1210 Target Cells Mediated by

Activated Macrophages and Cytolytic Factor

Target Cocultured Incubated __□ Percent Cell with Macrophages' with CF^{D ay}Fe Releas

EMT-6 47.1 + 1.2 EMT-6 0.8 + 0.4

L-1210 42.1 + 1.8 L-1210 -1.2 + 2.4

a 250,000 BCG activated macrophages per 0.33 cm2 well; triggered with 100 ng/ml LPS; cultured for 22 hr.

250 units of CF partially purified by molecular sievi ^c Calculated as

Experim• ental Release - Spontaneous Release x

Total Release - Spontaneous Release

44

J. Anti-necrosin antiserum neutralization of cytolytic factor

Necrosin and a rabbit antiserum raised again the electrophqretically homogenous toxin (Kull et al., supra) were obtained from Dr. Frederick Kull, The Wellco Research Laboratories, Research Triangle Park, NC. The antiserum was diluted 1:100 in PBS so that 1 ul would neutralize the lytic activity of 5 units of necrosin. From this stock, serial two-fold dilutions were made fro 1:40 to 1:640. To these diluents were added samples of stocks of either necrosin or Cytolytic Factor purified b molecular sieving. These stocks were adjusted to give a concentration of 1000-5000 units/ml as assayed on actinomycin D-treated EMT-6 cells. After 1 hr at room temperature, the antiserum-toxin mixtures were titered o EMT-6 targets and assayed for lytic activity 18-24 hr later. The percent neutralization of lytic activity of necrosin and Cytolytic Factor by the anti-necrosin antiserum was calculated as follows:

Percent neutralization = [1- Lytic Activity in the Presence of Antiserum (units/ml) Lytic Activity in the Absence of Antiserum (units/ml)

Normal rabbit serum was without effect at these levels i this assay.

Typical results are shown in Figure 6. The result indicated that such antisera was fully capable of neutralizing the activity of the Factor and necrosin; th antiserum dose response curve reflects the titratable nature of this response. K. Role of CF in Direct Nonspecific Tumor Cytotoxicity Mediated by BCG-Activated Macrophages

The anti-necrosin antiserum, which was shown to cross-react with the Cytolytic Factor produced by BCG- activated macrophages (Figure 6), was used to probe the possible role of the Factor in mediating direct nonspecific tumor cytotoxicity of macrophages in coculture. The results are seen in Table 7. Under the coculture conditions, the antiserum could significantly but not completely, block cytotoxicity expressed by activated macrophages against L-929 targets. The inhibition observed was dependent on the antiserum dose employed. The highest level of inhibition achieved brought the cytotoxic activity of LPS-triggered macrophages close to that observed for non-triggered macrophages.

46 TABLE 7

Inhibition of BCG-Activated Macrophage Nonspecific Tumor Cytotoxicity In Vitro with Anti-Toxin Antiserum^a

Effector Cells LPS Percent Cytotoxicity

Macrophages 27.1 3.7"

Macrophages 70.4 + 3.6

Macrophages + _f 1:1,000 Antiserum 33.8 + 2.9 (52.0)'

Macrophages + 1: 5, 000 Antiserum 54.6 _+ 3.1 (22.4)

Macrophages + 1:25,000 Antiserum 62.7 + 1.0 (10.9)

Macrophages + 1:125,000 Antiserum 63.1 + 1.0 (10.4)

Rabbit heteroantiserum against homogenous necrosin. Based on release of Cr in a 16 hr. assay using 20 x 10 L-929 targets/well in 96-well plate. ^c 250,000 BCG-activated macrophages/well in 96-well plat Mean +_ SEM; Percent release = Exp. Rel. - Spont. Rel

Total Rel. - Spont. Rel ^e 200 ng/ml Final dilution of anti-necrosin antiserum in microwell

^ Percent inhibition of endotoxin-triggered macrophage cytotoxicity of targets. 47

As noted above, the present invention has been disclosed in terms of examples considered by the inven to be the preferred embodiments for practicing the inv tion. However, they are in no way meant to be the onl modes for practicing this invention. For example, al- though the conditioned supernatant is fractioned by molecular exclusion chromatography, other methods of fractioning the supernatant are contemplated. For example, other fractionation methods which employ a molecular sizing or molecular weight fractionation can employed. Such separation might therefore involve gel electrophoresis, ultracentrifugation or any other sepa tion technique based on differences in molecular size weight. Similarly , although the macrophages employed herein are murine macrophages, there in no reason why other mammalian macrophages cannot be similarly employe Additionally, although the macrophages of the present invention are initiated using BCG and activated using E coli bacterial endotoxin, there is no reason why other initiators such as lymphokines and other activators suc as other bacterial endotoxins, cannot be utilized to ga the advantages of the present invention. These and all other changes should be considered within the scope of appended claims.

Claims

WHAT IS CLAIMED IS:

1. A composition comprising a substantially purified soluble cytolytic protein having a molecular weight of between about 140 and about 160 kilodaltons upon gel exclusion chromatography, said protein exhibiting biological properties which include stability of cytolyt activity for at least six months when stored at 4° in a physiological buffer and retention of said activity in t presence of 5% fetal calf serum.

2. The composition of claim 1 wherein said properties further include immunological reactivity with polyclonal antisera raised against necrosin.

3. The composition of claim 1 wherein said properties further include lability of cytolytic activity upon treatment with trysin, or heating to 100°C for 10 minute

4. The composition of claim 1 wherein said properties further include protease activity at neutral pH.

5. The composition of claim 1 wherein said properties further include cytolytic activity against L-929 cells.

6. ^" The- composition of claim 1 wherein said properties further include sensitivity of cytolytic activity to TLC or TAME.

1: The composition of claim 1 wherein said protein is obtained by a process which includes the steps of:

(a) preparing a macrophage conditioned supernatan

(b) ,fract.ionating the conditioned supernatant to " provide a fraction which includes the protein a substantially purified form; and

(c)' collecting the fraction.

8. The composition of claim 7 wherein step (a) comprises:

(a) innoculating a mammal with an activator;

(b) harvesting macrophages from the innoculated mammal;

(c) incubating the macrophages with medium containing a triggering agent; and

(d) separating the macrophages from the resultant conditioned supernatant.

9. The composition of claim 8, wherein the activator BCG, the triggering agent is bacterial endotoxin and th mammal is a mouse.

10. The composition of claim 7 wherein said supernatan is fractionated into protein fractions of selected 50 molecular weights to provide a fraction which includes protein in a substantially purified form.

11. The composition of claim 10 wherein said fraction selected by assaying fractions .for cytolytic activity against human tumor cells.

12. The composition of claim 8 wherein said activator selected from the group of initiators consisting of BCG, Immunotone, C. parvum, gamma interferon, supernatant derived from mitogen-stimulated T-cells,»and SMDP.

13. The composition of claim 8 wherein said triggering agent is selected from the group consisting of endotoxin SMDP, LPS and PMA.

14. A method for inhibiting the proliferation of tumor cells comprising subjecting said cells to an effective dose of a composition according to claim 1.