Soluble Suppressor Factors Elaborated in Experimental Malignant Ascites

Soluble suppressor factors in the sera of cancer patients inhibit lectin-stimulated lymphocyte proliferation. These factors, derived from human material, preclude easy corroboration by other investigators. To gain a general understanding of soluble suppressor factors and to avoid the necessary restrictions of human experimentation, an animal model was devised. Sprague-Dawley rats were injected ip with the Walker 256 carcinoma. The resultant ascites proved to be a stable, reproducible source of soluble suppressor factors. Ascites inhibited phytohemagglutinin (PHA)- induced blastogenesis of normal splenccytes by 98%. The possibility of a toxic effect was eliminated by vital staining of splenocytes and by examination in a specific lymphotoxin assay. Suppressor activity persisted after heating at 100°C for 40 min. Extraction by lipid solvents revealed that the bulk of suppressor activity resides in the lipid phase. The active fraction of heat-treated ascites passed through an Amicon PM-10 filter. Thin-layer chromatography revealed the presence of prostaglandins E2 and F2.. Tissue culture supematants from short-term cultures derived from tumor-bearing animals revealed suppressor activity from thymus, spleen, and liver cultures (97, 9 1, and 7 I %, respectively). No suppressor activity was detected in cultures of cancer cells. This study has demonstrated in this animal model that prostaglandins play a major role in suppression of lectin-induced blastogenesis. All suppressor factors appear to be host derived. An understanding of the mechanism of release of these suppressor substances may open new avenues in the immunotherapy of cancer.


INTRODUCTION
Soluble factors which suppress lymphocyte proliferation to lectin stimulation in vitro have been demonstrated to be present in the sera of human cancer patients (l-3). However, when these factors have been studied in detail, the physical properties and biologic activity vary among reports. Non-cancer specific, acute phase reactants such as immunoregulatory a-globulin have been found to inhibit lymphocyte proliferation to lectin stimulation in vitro and may be responsible for immunosuppression in tumor-bearing patients (4). A substance with a molecular weight of 50,000 has been isolated from the malignant ascites developing in patients with advanced gastric and ovarian cancers (5,6). Another factor which suppresses normal human lymphocyte function in vitro has been reported to be derived only from tumors (7). These suppressor factors are frequently derived from material obtained from patients in unusual clinical situations which preclude the possibility of easy corroboration by other investigators. Furthermore, it is difficult to separate cancer-related events from the effects of acute and chronic biologic stresses which accompany any human cancer.
To gain a more general understanding of soluble suppressor factors we elected to avoid the necessary restrictions of human experimentation by utilizing an animal model. It was our intent to investigate the elaboration of suppressor factors in a cancer-bearing animal. In this report we describe the initial characterization and partial purification of factors obtained from the ascites of rats injected ip with the Walker 256 carcinoma. These factors inhibit the in vitro proliferation of rat spleen cells to phytohemagglutinin (PHA) stimulation.

MATERIALS AND METHODS
Growth and collection of the ascites. Sprague-Dawley rats (Harlan Sprague-Dawley, Madison Wise.) were injected ip with 1 ml of a fine mince of subcutaneously grown solid Walker 256 carcinoma (Mason Research Institute, Worcester, Mass). This resulted in the production of a malignant ascites. Ten to fourteen days postinnoculation, the abdominal wall became tense due to distention by fluid. The ascites, serum, and tissues were harvested. Ascites was removed by transabdominal needle aspiration from Halothansanesthetized rats. Blood was obtained by cardiac puncture. The ascites and blood were allowed to stand at 20°C for 1 hr, then centrifuged at 1500 rpm for 10 min, the supematant serum or ascites was aliquotted and stored at -35°C until assayed for suppressor activity.
Primary tissue culture. Tissue cultures were established using thymus, spleen, liver, and ascites tumor cells obtained from sacrificed tumor-bearing rats. The organs were finely minced with iris scissors in a sterile petri dish containing RPM1 1640 medium (G&co, Grand Island, N.Y.) and filtered through a sterile fine mesh gauze, to produce a single-cell suspension. The resulting cell suspension was washed three times with PBS. A primary culture was started by adding IO' cells/ml to a mixture of RPM1 1640 with 20% FCS. The cell pellet obtained from the centrifugation of crude ascites was similarly resuspended in RPM1 1640 with 20% FCS. This cell pellet consisted of cancer cells with some host derived inflammatory cells. The cultures were incubated at 37°C in a mixed atmosphere of 95% 02 and 5% CO,. Tissue culture medium was added as needed. The spent tissue culture medium was collected, centrifuged at 1500 rpm, aliquotted, and stored at -35°C. All cultures were terminated at 96 hr.
Immunosuppressor assay. A proliferation-inhibition assay utilizing normal splenic lymphocytes obtained from healthy Sprague-Dawley rats was employed. The donor animal was anesthetized with Halothane; the spleen was removed and forced through a fine wire screen into a sterile petri dish containing PBS. Red blood cells were lysed by suspending the cells in 1% ammonium chloride. Nonlysed cells were centrifuged at 1500 rpm for 6 min and the supematant discarded. The cell pellet was resuspended in PBS and washed in similar fashion two times. The cells were counted and the dilution adjusted such that 180 ~1 of the cell suspension contained 1 X 1 O6 cells. The final resuspension was accomplished with RPM1 1640 with 10% fetal calf serum and containing an appropriate amount of PHA (2.5-5.0 &ml; Wellcome Research Labs, Beckenham, England).
The assay was conducted in microtiter test plates (Flow Laboratories, Inc., McLean, Va.), with each sample assayed in sextuplicate. Twenty microliters of the sample to be tested was added to each well. The splenic lymphocyte suspension was added (1 X lo6 cells/well) and the plates were incubated at 37°C in an atmosphere 95% 02-5% COz. Maximal response was determined using a control in which the test sample was replaced by 20 ~1 of pooled normal rat serum. Minimal response was determined using the same control but in the absence of added PHA. Ten microliters of tritiated uridine (50 &i/ml; Amersham, Arlington Heights, Ill.) was added to the test plates at 24 hr and the cells were harvested 4 hr later. This test monitored early RNA associated events in blastogenesis. In the majority of experiments 10 ~1 of tritiated thymidine ( 100 &i/ml, Amersham) was used to monitor DNA-associated events of blastogenesis. The thymidine was added at 72 hr following initiation of the culture and the cells were harvested 24 hr later.
Isotope uptake was determined by aspirating the cell suspension using an automated multiple-sample harvester (Flow Laboratories, McLean, Va.) which deposited the contents of the wells on a filter paper sheet. The filter paper sheet was dried for 1 hr at 95"C, cut into squares, which were placed in scintillation vials containing 2 ml of scintillation cocktail, and counted on a liquid scintillation counter (Amersham-Searle, DesPlaine, Ill.). The results of the assay for each sample were obtained and a mean and standard deviation were calculated. The percentage of inhibition was calculated by dividing the counts per minute of the experimental sample by counts per minute of the positive control, subtracting the result from one, and multiplying by 100.
Cytotoxicity assay. Cell cultures of lymphotoxin (LT)-sensitive G929 fibroblasts were established at a density of 100,000 cells/ml in RPM1 1640 with 3% newborn calf serum (NCS, Gibco) and 0.5 &ml mitomycin C (Sigma Chemical Co., St. Louis, MO.). These cultures were incubated overnight at 37°C. Aliquots of test samples were added in triplicate to the cultures. Appropriate controls included RPM1 tissue culture medium alone and a tissue culture supematant known to contain LT. The incubation was continued for an additional 24 hr. The number of viable adherent cells was determined by trypsinizing the monolayer and passing the cell suspension through a Model F Coulter counter (Coulter Electronics, Hialeah, Fla.) by the technique of Spofford et al. (8). A significant decrease in the number of viable cells as compared to the control with RPM1 alone would indicate the presence of LT.
Heat stability and molecular seiving. Crude ascites was heated to 100°C for 40 min and centrifuged at 30,000 rpm for 30 min. The supematant was then assayed for suppressor activity. Heat-treated ascites was then subjected to molecular filtration using Amicon PM-30 and PM-10 membranes under a positive pressure of 30 pounds per square inch of nitrogen.
Chloroform:methanol extraction. The heat-treated ascities was extracted twice with 3 vol of a 3: 1 (v:v) mixture of chloroform and methanol. The combined extract was dried under streaming nitrogen at room temperature and the slight residual aqueous component was completely dried in a vacuum. The residue was dissolved in a minimal volume of ethylacetate:acetic acid:methanol:2,2,4-trimethylpentane:water (110:30:35:10: 100) the solvent AI1 of Green and Samuelsson (9). The sample was then subjected to silica gel thin-layer chromatography as described by Willis (10). Prostaglandins present were tentatively identified by their Rr values, using known prostaglandins of the E and F series as standard markers. When utilized in the proliferation inhibition assay the dried extracts were dissolved in a volume of RPM1 equivalent to the original volume of the extracted sample.

RESULTS
The addition of 20 ~1 of ascites fluid to 180 ~1 of PHA-stimulated lymphocytes reduced the rate of uptake of tritiated thymidine to or below the level demonstrated by nonstimulated cells. Progressive dilution of the ascites with RPM1 also caused a loss of potency, indicating that an additive factor is present. (Table 1). Cell counts conducted on lymphocytes incubated with ascites fluid under similar conditions showed no loss in cell number or in cells taking up trypan blue. This demonstrated that the inhibitory effect was probably not due to cytotoxicity. This conclusion was conlirmed using the lymphotoxin assay summarized in Table 2. The addition of as much as 500 ~1 of undiluted ascites fluid did not reduce the number of adherent lymphotoxinsensitive cells, indicating no loss of viability. A portion of the suppressor activity is heat stable. Immersion of the ascites fluid in boiling water for 40 min only reduced its potency slightly (Table 1).
Heat-treated ascites was extracted with chloroform:methanol as described. The majority of the activity extracts into the nonpolar phase (Table 3). Thin-layer chromatography indicated that such extracts contained prostaglandins Ez and F2, and other unidentified substances. The active substance in the heat-treated ascites also passed through Amicon PM-30 or PM-10 ultrafilters (Table 4).
To determine the possible source of the active material, tumor, spleen, thymus, and liver cells were grown in short-term cultures as described. Aliquots of the media were then examined in the thymidine assay. The results summarized in Table 5 show that the conditioned media from the tumor cells actually stimulated thymidine uptake while the media from the cultures of spleen, thymus, and liver cells suspensions inhibited blastogenesis.
The data summarized in Table 6 show that the less time-consuming uridine assay gives results equivalent to those of the thymidine assay.

DISCUSSION
We found that ascites obtained from Sprague-Dawley rats injected ip with a Walker 256 carcinoma exerts a powerful suppressor effect in vitro upon proliferation of rat splenocytes to mitogen stimulation. Serial dilution of the ascites resulted in the loss  of the suppressor effect, eliminating the possibility of lack of enhancing factor and demonstrating the presence of suppressor factors. In other words, inhibition of mitogeninduced blastogenesis could be explained by the absence of factors necessary to support blastogenesis or by the elaboration of new factors exerting a biologic suppressor effect.
In the first case, lack of essential factors, could be eliminated, since dilution of ascites would be expected to result in sustained inhibition of blastogenesis. The observation that dilution of ascites returns blastogenesis to levels near those of the positive controls conthms the presence of suppressor factors, whose effects were decrease d upon dilution. This early work gave impetus to further efforts to characterize these factors and to elucidate their site of elaboration. The observation that the uridine assay yielded results comparable to thymidine assay corroborates the findings of other investigators ( 11,12) and supports the use of the uridine assay for the rapid screening of suppressor activity. Both vital staining and lack of LT activity proved that the suppression observed was not the result of a cytotoxic effect upon the responding lymphocytes. Studies were conducted to establish the molecular characteristics of material(s) with suppressor activity. Activity persisted after heating the ascites to 100°C for 40 min. Indeed, this provided an excellent means of removing protein from the ascites while preserving suppressor activity. We also found that active material passed through both Amicon PM-30 and PM-10 filters without loss of activity, indicating that we were dealing with material of molecular weight less than 10,000.
Next we extracted active preparations with chlorofonmmethanol. Biologic activity was found to be extractable in nonpolar lipid solvents. This material was chromatographed revealing the presence of prostaglandins E2 and F2,. Prostaglandins E2 is known to inhibit lymphocyte function (13) and has been shown to be secreted by  suppressor cells during stressful situations (14). The presence of other lipid-soluble suppressor substances, such as slow reacting substances, was not measured. The cultures prepared from tissue obtained from tumor-bearing animals produced evidence that immunosuppressive factors are produced by the host cells rather than by the tumor. Significant immunosuppressive activity was detected in the thymic, splenic, and hepatic cell cultures, but not in cultures of tumor cells. Other workers have shown that these cell suspensions contain populations of suppressor cells capable of secreting prostaglandins ( 15). There are a host of protein and lipoprotein substances in serum capable of suppressing mitogen-induced lymphocyte blastogenesis in vitro (16). These substances are synthesized in the liver and may account for the presence of suppressor activity in the tissue culture supematant of the hepatic cell cultures. This lack of suppressor activity arising from tumor cell cultures suggests that all suppressor activity observed in this model is derived from multiple organ host and tissue sites and is not derived from tumor tissue.
These findings then lead to the question of whether or not unique suppressor factors are elaborated in the presence of cancer cells. If unique cancer-related suppressor factors exist, then their utility in the immunodiagnosis and immunotherapy of human cancer would be great. If these factors are not unique to cancer then the problem of clinical immunosuppression as a result of elaboration of suppressor factors could be viewed as a final common pathway in a variety of ills. Indeed, this seems to be the case as suppressor substances have appeared in malignant ascites as well as ascites produced by serious but non-cancer-related states ( 17). Major trauma ( 18), surgery (19), and serious medical illnesses such as multiple sclerosis (11) are all associated with suppressor factors.  The area of soluble suppressor factors appears more complex than initially perceived. Although prostaglandins were detected in this model, there is strong suspicion that protein-like suppressor substances with molecular weights greater than 10,000 are also present. This speculation arises from preliminary work demonstrating loss of a portion of suppressor activity after heat treatment and a persistance of suppressor activity after dialysis which would be expected to remove prostaglandins. Suppressor activity is detected in material adherent in ion-exchange chromatography, again suggesting the presence of protein-like material. Further investigation will be required to elucidate the precise nature of what appears to be multiple suppressor factors elaborated from multiple sites in a tumor-bearing host.