| lulu; will l1lll ll la! in lfllllL Ill ll This is to certify that the thesis entitled Studies on a Factor Chemotactic for Macrophages Elaborated by SAD/2 Fibrosarcoma cells presented by Louise Scnaub Simon has been accepted towards fulfillment of the requirements for PhD , Microbiology & Publ.Health degree m @{( ){ér L404) /’ big/4min»: 5/18/79 Date 0-7639 OVERDUE FINES ARE 25¢ PER DAY _ PER ITEM Return to book drop to r-ove this checkout from your record. STUDIES ON A FACTOR CHEMOTACTIC FOR MACROPHAGES ELABORATED BY SAD/2 FIBROSARCOMA CELLS by Louise Schaub Simon A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1979 ABSTRACT STUDIES ON A FACTOR CHEMOTACTIC FOR MACROPHAGES ELABORATED BY SAD/2 FIBROSARCOMA CELLS By Louise Schaub Simon Preliminary studies with serum-containing supernatant from the SAD/2 fibrosarcoma cell line grown in vitro determined that the super— natant was apparently chemotactic for macrophages when chemotaxis was assessed microscopically. Prior to isolation and purification of this tumor material,a rapid method for the assessment of chemotaxis utiliz— ing 3H-uridine labeled macrophages was developed. This method employed the use of double polycarbonate filters which separated the lower wells containing SH-uridine labeled macrophages in modified blind-well Boyden chambers. After incubation, cells on the lower filters were precipi- tated with trichloroacetic acid, and radioactivity was assessed by liquid scintillation spectrometry. The radioactive method was compared to the microscopic method of quantitating chemotaxis. There was good correllation between microscop- ic and radioactive methods (10% difference in the two methods) when either endotoxin treated serum, a known chemotactant, or the tumor supernatant was used as a chemotactant. Since good correllation existed between the two methods, the radio- active method was used to determine whether the tumor supernatant (T8) was a true chemotactant or a chemokinetic agent. Four sets of chambers were assembled. The first set contained cells suspended in medium in the top well, TS in the lower well. The second set contained cells suspended in T8 in the top well, TS in the lower well. The third set Louise Schaub Simon contained cells suspended in TS in the top well, medium in the lower well. The fourth set (control) contained cells suspended in medium in the tap well, and medium in the lower well. A 5-10 fold increase in radioactivity of filters from the first set of chambers was observed. Radioactivity of the second and third sets of chambers were comparable to controls. The T8 was considered a true chemotactant, since much greater radioactivity on filters of the second and third sets of cham- bers would be anticipated if the TS were merely chemokinetic. Experiments were then begun to purify the chemotactant using the afbrementioned assay to assess chemotactic activity. Methods used to isolate the tumor material labeled with 3H-leucine were gel filtration through Sephacryl $200 and Sepharose 6B and gel electrophoresis. Serum- free supernatants from SAD/2 tumor cells were concentrated and chromato- graphed on Sephacryl $200. Fractions containing chemotactic activity were applied to a Sepharose 68 column, and then applied to gels for electrophoresis. Active fractions from the Sepharose 68 column con- tained two components on gels with and without sodium dodecylsulfate (SDS). The apparent molecular weights of the 2 components were 68,000 and 78,000 on $08 gels. Samples pretreated with anrcaptoethanol prior to $05 gel electrophoresis also contained two bands, which indicated single polypeptide chains. Trypsinized samples resulted in loss of the 2 bands on 808 gels, and the loss of chemotactic activity. Heat treat- ment of active Sepharose 63 fractions at 100 C for 15 minutes also re- sulted in loss of chemotactic activity. Sepharose 68 active fractions were injected intraperitoneally into mice, and differential cell counts of the peritoneal cells were per- formed 48 hr later. These data were compared to cell counts from Louise Schaub Simon control mice injected with phosphate buffered saline. There was a sig- nificant increase in the numbers of macrophages and a significant de- crease in the number of lymphocytes in test animals compared to controls (p < 0.01). From these results it was concluded that the partially puri- fied tumor material caused an inflammatory response in vivo. The poten- tial implications of this tumor material in tumor-bearing animals was discussed. ACKNOWLEDGEMENTS I wish to express my appreciation to my two advisors, Drs. Tobi Jones and Ronald Patterson for their help and moral support during my research. I am especially grateful to them for their assistance in preparation of manuscripts for this thesis. I would also like to thank other members of my committee. Drs. Gary Hooper, and Lee Velicer. Special thanks is extended to Dr. Walter Esselman whose counseling during this research was invaluable. My sincere gratitude is also extended to Drs. Bill Chaney and Myra Jennings for the stimulating conversation and research ideas generated during my contact with them in the laboratory. I would also like to thank Joyce Wildenthal for her assistance in laboratory techni- ques, Debbie Densmore for the preparation of this dissertation and other manuscripts, and H. Stuart Pankratz for his photographic counsel- ing. I am also grateful to Dr. Wayne Smith and James Hollers for the use of modified Boyden chambers. I would also like to acknowledge my husband, Nick, for his assistance with photography and his encourage- ment throughout this research. ii TABLE OF CONTENTS Page Literature Review 1 Manuscript 1 - A rapid method for assessment of a macrophage chemotactant produced by SAD/2 fibrosarcoma cells grown in vitro. 9 Manuscript 2 — Isolation and partial purification of a macrophage chemotactant produced by SAD/2 fibrosarcoma cells grown in vitro. 27 Bibliography 50 iii LIST OF TABLES Manuscript 1 Page I Use of 3H-Uridine Labeled Macrophages to Assess Chemotaxis 22 II Correlation Between Number of Cells per Filter and Radio- activity (CPM) observed per Filter 23 III Determination of Chemotaxis of Peritoneal Macrophages toward Tumor Supernatant 24 Manuscript 2 I Measurement of Inflammatory Response In vivo to Partially Purified Tumor Material 49 iv LIST OF FIGURES Page Manuscript 1 1 Comparison of visual and radioactive methods for quantitation of chemotaxis 26 Manuscript 2 l Chromatography of serum-free tumor supernatant 41 2 SDS gel electrophoresis of crude tumor supernatant and chemotactively active fractions from the Sepharose 68 column 43 3 Protein profile of radiolabeled Sepharose 68 active fractions on $08 gel electrophoresis 45 4 Protein profile of active Sepharose 68 fractions on native gel 47 LITERATURE REVIEW The involvement of macrophages in protection of the host against tumors and tumor growth has received much attention by investigators in recent years. There is an extensive amount of literature on the effects of tumors and tumor cell products on macrophages, and on the effects of macrophages on tumor growth. Since the research reported in this thesis is limited to a study of a tumor product's affect on macrophage function, the literature review will be focused on the affects of tumors and tumor cell products on macrophage fbnction. The second portion of the litera- ture review will be devoted to chemotaxis and the methods used to study chemotaxis. The literature concerning the affect of tumors on macrophage func- ‘tion is filled with contradiction as will become apparent during this literature review. A point to be remembered, however, is that investi- gators have worked with a variety of animal tumors in their research. Since all tumors are not the same, the host response as well as the in vitro response of inflammatory cells may be different depending on the tumor system. The fact that large numbers of macrophages are found within the tumor mass of a variety of human and murine tumors has been established (10,21,22,S4,SS). The function of these macrophages in tumors has slowly become apparent. The suggestion that resident macrophages in tumors prevented metastasis was made by Evans (10) and Eccles G Alexan- der (8). These investigators found that a correlation existed between the number of macrophages within tumors and metastasis of tumors. This suggestion led Wood 6 Gillespie (55) to determine experimentally that when localized tumors were excised, trypsinized, and depleted of 1 macrophages, the tumors became metastatic when injected into syngeneic animals. Resident macrophages within tumors have also been reported to inhibit growth of tumor cells in vitro (11) or to be cytotoxic to tumor cells in vitro (12,39,40). Haskill (15) reported that a popula- tion of cells within tumors, probably of monocyte origin, was involved in antibody dependent cell mediated cytotoxicity of tumor cells in vitro. Investigators have therefore observed large numbers of macrophages within tumors and have fmmd them functional in vitro (11,12,15). These same investigators observed that many tumors continue to grow in spite of macrophage presence in the tumors. Consequently, investigators have be- come interested in the apparent failure of macrophages to control tumor growth. Several investigators have observed that patients with tumors had impaired macrophage functions in vitro (3,47), and that surgical removal of the tumors restored macrophage function (50). These observationslead to the proposal that tumor cell products inhibit macrophage function. This depression of macrophage function has been measured in a variety of ways. First, tumor cells are reported to suppress macr0phage migra- tion in vivo and macrophage chemotaxis in vitro (25,47,48). Snyderman et al. (47,48) found a low molecular weight component (6-l0,000 MW) ob- tained from the dialysate of tumor cell lysates that inhibited chemotax- is of macrophages in vitro when cells were pretreated with the component. When the lysate was injected into the flanks of mice, it inhibited an inflammatory response to PHA in the peritoneal cavity. Meltzer and Stevenson (26) also found that resident macrophages from tumors had depressed chemotaxis in vitro, but did not attempt to characterize the component which caused this affect. Secondly, inhibition of macrophage phagocytosis has been associated withgtumor growth. Otu et al (33) reported a marked depression of carbon clearance in viva throughout the first 72 hr after tumor implantation in animals. Gallahon and Wood (14) fbund that macrophages isolated from small tumors had in- creased phagocytosis in vitro, but large tumors (>lcm3) had a marked decrease in phagocytic function. Thirdly, impaired macr0phage func- tion has been measured by suppression of macrophage mediated resistance to infection with intracellular parasites (31). These investigators found that animals injected with tumor cells 24 hr prior to Listeria challenge had depressed rate of Listeria clearance from their livers. These same investigators were careful to observe in subsequent work however, that this impaired state was short-lived and was replaced by a state of greatly enhanced antibacterial resistance which closely par- alleled a state of enhanced resistance to a second challenge of the tumor (32) . Depression of macrophage function has been attributed to a soluble product found in the serum in viva (32,33), in tumor cell lysates (36, 47,48), and in supernatants of in vitra grown tumor cells (30,33). Most of these inhibitors have not been characterized, with the exception of the low molecular weight inhibitor obtained by dialysis of tumor cell lysates described by Snyderman and Pike (48). Activation of macrophage function during tumor growth has also been of interest to investigators, since increased macrophage function has been associated with incidence of tumors in humans (24). These investi- gators reported that increased phagocytosis of radiolabeled aggregated serum albumin in viva was correlated with resistance to tumor metastasis. Other researchers working with animal tumor systems have also observed increased macrophage function. Stimulation of phagocytosis was observed in tumor-bearing mice by Meltzer and Stevenson (26). Meltzer, Tucker and Bruer (28) reported that peritoneal macrophages from tumor immunized BCG infected mice had increased chemokinetic response to tumor cells in vitro. Snodgrass et al., (44) and Schuller et al. (41) used pyran acti- vated macrophages to obtain similar results. Blakeslee (2) reported that supernatants from tumor cells grown in vitra activated peritoneal macr0phages to inhibit tumor growth when macrophages were co-cultured with tumor cells. Meltzer et a1. (27) found a low molecular weight (15,000 MW) product in the supernatant of tumor cells grown in vitra that enhanced chemotaxis of normal and ECG infected macrophages. This material was produced in serumecontaining medium. These investigators did not show whether this component was actively produced by the tumor cells or whether it was the result of the action of tumor cells on serum components in the medium. The SAD/2 fibrosarcoma contains a large number of phagocytic cells (21). In addition, peritoneal macrophages from SAD/2 bearing animals were found to have an enhanced cytostasis of tumor cells in vitra (19). Since the tumor has a large number of macrophages, and its growth in viva stimulated peritoneal macrophages, this was a convenient model to study tumor cell/macrophage interaCtion. The research of this investigator has concentrated on purification of a macrophage chemotactant produced by the SAD/2 tumor cells grown in vitra in serum-free medium. A rapid method for assessment of chemo- tactic activity was developed to process large numbers of samples gen- erated during the purification procedure. The second part of this lit- erature review, therefore, will be devoted to recent methods used to assess leukocyte chemotaxis. For more extensive reviews of chemotaxis, the reader is referred to Harris (17) and Wilkinson (53). The word chemotaxis is derived from the Greek roots xuuo’s (juice, liquid) and Tail; (an arrangement, battle array), but has taken on the meaning of movement in biological contexts (53). This movement in leu— kocyte chemotaxis is a directional movement, toward or away from a con- centration gradient. Chemotaxis toward a concentration gradient is important in inflammation. Inflammatory cells can respond to injury, infection, or tumor cells which may be the result of chemical gradients. The function of these inflammatory cells is to aid in repair of tissue, prevent the spread of infection, and possibly control tumor growth. Most of the early research dealing with leukocyte chemotaxis em- ployed implanting of capillary tubes containing test chemicals into animals (53). Observation of cells which migrated into the tubes after a suitable incubation period was used to assess chemotaxis. The study of chemotaxis in viva was modified by employing the skin window technique by Rebuck et a1. (38). With this technique an area on the skin was scrapped, the test substance was applied to the lesion and the site was covered with a sterile coverslip. The cells which migrated into the site in response to the chemotactant attached to the coverslip, and they were stained and counted after an appropriate incubation period. Perillie and Finch (35) reported an improvement of the skin window tech- nique which was quantitatively more accurate. A chamber containing physiological saline was placed over the lesion, and cells migrating into the chamber could be quantitated. Recent research in leukocyte chemotaxis was stimulated by the intro- duction of a new in vitra technique for assessment of chemotaxis (4). 6 This system involved the use of a chamber containing two wells separated by a porous filter. The lower well contained the chemotactic substance and the upper one contained the cells. The cells then migrated through the pores toward a concentration gradient to the bottom surface of the filter. The filters were removed after incubation, and they were exam- ined microscopically. - Various modifications of the Boyden chamber have been used to assess human (52) and murine (46) chemotaxis. These structural modifi- cations essentially involve the same principles that Boyden used with either Millipore filters or polycarbonate filters separating the two wells. The in vitra method of quantitation of chemotaxis originally util- ized Millipore filters in chemotaxis chambers, followed by staining and counting polymorphonuclear leukocytes (PMN) which had migrated through the pores to the opposite side of the filter (4). Other investigators have used the leading front of cells within the pores to assess chemo- taxis (56). This latter technique makes identification of cell type difficult when mixed cell populations are used in chemotaxis systems. Another method used to quantitate PMN chemotaxis was described by Zigmond and Hirsch (56). This method involved placement of the chemo- tactant on a slide, allowing it to dry, and adding a drop of cell sus- pension to the streak. A coverslip was placed over the suspension, and the cells were observed for the formation of lamellipodium (the front of the cell has a broad smooth veil, the rear is marked by a constricted tail). Only cells with lamellipodium directed toward or away from the test line were quantitated. Since mononuclear cells are also involved in inflammation, it was desirable to devise a method for quantitation of mononuclear cell chemo- taxis. Quantitation of mononuclear cell chemotaxis was made possible by the use of larger pores in the Millipore filter and longer incubation periods (20,51). Horwitz and Garrett (18) used a polycarbonate filter which was 12 um thick (Millipore filters are 150 pm thick) and required less time for incubation than did the Millipore filters. These filters also made identification of cell type more accurate as cells were not trapped within pores of the filter. Attempts to improve quantitation of chemotaxis with a rapid radio~ labeling technique have been fairly successful. Gallin et al. (13) used 51Chromium to label granulocytes followed by chemotaxis in Boyden cham- bers containing two filters. The second filter containing 51Cr-labeled cells, was used for quantitation of chemotaxis. Although assessment of chemotaxis was more objective and rapid, the labeling procedure caused clumping and subsequent loss of cells when they were filtered through sterile gauze (13). This technique also has a high spontaneous release of label from the cells which makes the background counts high (1). The use of gngechnetium to label granulocytes (9) and human monocytes (34) 51Chromium. However, this label is not presents an alternate label to readily available to researchers, and a short half-life (5.9 hr) neces- sitates corrections for decay if large numbers of samples are to be counted (9). The agarose technique has also been used as a rapid method of quantitating leukocyte chemotaxis. This technique involved placement of guinea pig tissue explants or exudate cells in plastic petri dishes, and agarose containing serum or gamma globulin was poured over the pre- paration (55). Migration of cells occurred primarily between the agar and petri dish, however, some cells migrated on the surface of the_agar as well. Cutler (6) later used this technique to quantitate guinea pig neutrophil chemotaxis. Nelson et a1. (29) modified this technique by making wells in the agarose, and introducing chemotactant and cells in separate wells. Dif— fusion of chemotactant into the agarose caused a gradient to which human neutrophils and monocytes migrated. Migration of cells was primarily under the agarose. The agarose could be removed, and cells remaining adherent to the petri dish could be quantitated by measuring the distance the cells migrated toward the chemotactant. The agarose technique is reported to be a rapid and easily quantitated method for both neutrophils and monocytes. However, the method is not consistent when murine peri— toneal macrophages are used (Tobi Jones, personal communication). A RAPID METHOD FOR ASSESSMENT OF A MACROPHAGE CHEMOTACTANT PRODUCED BY SAD/2 FIBROSARCOMA CELLS GROWN IN VITRO by Louise Schaub Simon, Ronald Patterson, and Tobi L. Jones INTRODUCTION The role of macrophages in host defense against neoplasia has re- ceived much attention in recent years. Macrophages are found in large numbers in a variety of human (22,54) and animal tumors (8,10,15,16,21). Speculation that macrophages in tumors prevent metastasis (10) is sup- ported by a report that tumors depleted of macrophages became metastatic in viva (55). Macrophages isolated from tumors are reported to be cyto- static or cytotoxic in vitra (12,39,40). Recent research has focused attention on modulation of macrophage function, in particular, macrophage chemotaxis, by tumor cells or tumor products. Snyderman and Pike (36,48) have isolated by dialysis a low molecular weight intracellular tumor product which inhibits macrophage chemotaxis in vitra and macrophage migration in viva. Meltzer et al. (27) have reported enhancement of in vitra macrophage chemotaxis by a product faund in supernatants of tumor cells grown in vitra in serum- containing medium. In this report, we describe an in vitra macrOphage chemotactant produced by fibrosarcoma tumor cells grown in vitra in serum-free medium. As a prelude to isolation and purification of this product we report a rapid method of assessing chemotaxis in vitro which utilizes [5.6-3H]uri- dine labeled peritoneal macrophages. The usefulness of this method which will be used in processing large numbers of samples generated during purification procedures will be discussed. MATERIALS AND METHODS Media and Reagents, Sterile Eagle's minimal essential medium (MEM) was used for collection and preparation of peritoneal exudate cells for 10 11 ‘ chemotaxis. Tumor cells were cultured in CMRL 1066 medium containing 10% decomplemented fetal calf serum (le66). Serum free 1066 (i1066) was used for the generation of supernatants from tumor cells. All media contained 50 U/ml penicillin, 50 pg/ml streptomycin, and was buffered with NaHCOs. Four percent bovine serum albumin [(BSA) Sigma] was prepared in Hank's balanced salt solution and buffered with NaHCOs. All components were obtained from Grand Island Biological Co., Grand Island, NY, and Microbiological Associates, Walkerville, MD. Preparation of Mouse Peritoneal Macrophages for Chemotaxis. Two to six month old male DEA/2 mice (Jackson Labs, Bar Harbor, ME) were used in all experiments. Peritoneal exudate cells were stimulated according to the method of Snyderman and Pike (46) with some modificae tions. Mice were injected intraperitoneally (IP) with 1.5 ml of sterile 9% proteose peptone (Difco Laboratories, Detroit, MI) 3-4 days prior to harvest of peritoneal cells. Mice were sacrificed by cervical disloca- tion, and the peritoneal cavities were lavaged with 5 ml MEM. The peri- toneal exudate cells (PEC) were counted, centrifuged at 163 x g, and resuspended in MEM to a concentration of 107/ml. The cells were labeled with 10 uCi [5,6-3H]uridine (Amersham, 48 Ci/mmol)/107 PEC/ml MEM for 1 hr at 37C in 17 x 100 mm polystyrene tubes. Macrophages were kept in suspension during the labeling process with a small magnetic stiring bar. After labeling, the cells were washed 3 times in 10 volumes of cold MEM, 6 and resuspended in MEM at 4.4 x 10 PEC/ml. An equal volume of 4% BSA was added to the cell suspension to give a final concentration of 2% BSA and 2.2 x 106 PEC/ml. Chemotaxis of Mouse Peritoneal Macrophages. The method of Snyderman and Pike (46) was used for chemotaxis with modifications. Uridine 12 labeled PEC were employed, and two polycarbonate filters (Nucleopore) with Sum pores were placed in blind-well modified Boyden chambers. The lower well was filled with the chemotactant or control medium, and the two polycarbonate filters were placed on top. The t0p well was assembl- ed and filled with 0.2 ml of the cell suspension. The chambers were incubated at 37 C in a humidified atmosphere of 5% CO for 3.5-4 hr 2 prior to preparation of the filters for quantitation of chemotaxis. Evaluation of Chemotaxis with Radiolabeled Cells. After incubation of the Boyden chambers, the cells were carefully removed from the top chamber with a Pasteur pipette and the chambers were disassembled. The top filter was discarded, and the second filter containing only cells which had migrated through the first filter, was used for evaluation of chemotaxis. The cellular material on the filter was precipitated with approximately 3 ml 10% trichloroacetic acid (TCA) by suction filtration. The filters were air dried, placed in vials containing scintillation fluid, and counted by liquid scintillation spectrometry. The fallowing procedure was used to determine that increased CPM observed on chemotaxis filters was due to increased number of chemotax- ing macrophages. Various concentrations of uridine-labeled cells were pipetted onto polycarbonate filters contained in glass petri dishes and incubated at 37 C in humidified 5% CO2 atmosphere for 3.5 hr. Two fil- ters for each concentration were stained in the same manner as described below fur the visual chemotactic assay, and cells were counted micro- scopically at 1000 X. The mean number of cells per field from 20 fields was determined, and the number of cells per filter was calculated based on the area of the filter. Percent adherence was calculated as follows: #cells calculated/filter #cells placed on filter X 100. A parallel set of filters was 13 precipitated with TCA, dried, placed in scintillation fluid, and mean CPM per filter :_standard error was determined. Evaluation of Chemotaxis by Visual Assay. A visual assay for chemo- taxis was done in parallel using the same uridine labeled PEC population as for the radioactive assay. After incubation of Boyden chambers and removal of cells, duplicate lower filters were prepared for visual quan- titation of chemotaxis by fixing in absolute alcohol, staining with Ehrlich hemotoxylin, and rinsing in tap water. Air dried filters were mounted in immersion oil on slides, and examined microscopically at 1000 X. The mean number of macrophages per field present in 20 oil immersion fields on both sides of duplicate filters was determined. Chemotactants. Serum-free culture supernatants from the in vitro line of the SAD/2 methylcholanthrene—induced fibrosarcoma were tested for chemotactic activity. Confluent cultures were trypsinized with 0.1% trypsin (Sigma) in PBS at 37 C. The cells were centrifuged and resus- pended in d1066 at a concentration of 1.5 x 106/ml, and 10 ml were plated in 250 cm3 flasks (Corning). The cultures were incubated in a humidified atmosphere of 5% CO at 37 C for 24 hr, washed 3 times with 2 PBS, and 10 m1 i1066 were placed on the washed cells. Supernatants were collected after an additional 48 hr incubation period. Following centri- fugation at 163 x g to remove detached cells, the supernatants were tested fer chemotactic activity as previously described. Mouse serum was treated with endotoxin [(E, 5211 serotype 0261 B6) Sigma] to generate chemotactic complement components (45) as described by Snyderman and Pike (46). This endotoxin-treated serum (ETS) was included in each experiment as a positive control. The negative con- trol consisted of 11066 in the lower chamber. 14 Mouse embryonic cells used to generate supernatants from normal cells were in the fourth passage of cells harVested from 15-18 day old DEA/2 embryos. Embryonic cells were grown in the same manner as tumor cells to generate supernatants. Determination of True Chemotaxis or Chemokinesis of Tumor Superna- tggtg, A modification of the method of Zigmond and Hirsch (56) was used to determine whether tumor supernatants were chemotactic (directional movement) or chemokinetic (random movement) for macrophages. Four sets of duplicate blind—well Boyden chambers were prepared. The lower wells of the first set contained tumor supernatants (TS) and the upper wells contained cells suspended in i1066 with 2% BSA. .The second set contain- ed TS in the lower wells, and cells suspended in full strength TS in 2% BSA in the upper wells. The third set contained i1066 in the lower wells, and cells suspended in full strength TS containing 2% BSA in the upper wells. The feurth set was the medium control. Upper and lower wells were separated by double polycarbonate filters, and the radioactive method was used to quantitate migration of macrophages as described pre- viously. Parallel sets of chambers containing ETS as chemotactant in lower and upper wells were included in each experiment. Statistical Evaluation. In both methods, chemotactic activity was determined in duplicate chambers. Twenty oil immersion fields were examined in the visual assay and data is reported as mean number of cells per oil immersion field on duplicate filters. The radioactive assay was reported as mean number of counts per minute per filter. Standard error was used as an estimate of variance. RESULTS Preliminary experiments which utilized a standard chemotaxis assay described by Snyderman and Pike (46) determined that culture supernatant from SAD/2 cells grown in vitra in serum-free medium was chemotactic for peritoneal macr0phages. Since a chemotactic assay was to be used during the isolation and purification of the tumor material, we wished to mod— ify the standard chemotaxis assay to facilitate assaying the large num- ber of samples generated in purification procedures. Initial experiments were set up to determine whether uridine label- ed macroPhages could be used for chemotaxis. Radiolabeled PEC were placed in the top well of a modified Boyden chamber, and were separated from the lower well containing tumor supernatant, ETS or supernatants from embryonic fibroblasts by double polycarbonate filters. When lower filters were precipitated with TCA, there was a 2-6 fold increase in CPM observed in chambers containing tumor supernatant or ETS in the lower well when compared to negative controls (Table I). Radioactivity on the second filter of chambers containing supernatants from the fourth passage of embroyonic cells was comparable to radioactivity of control filters. (Data not shown). Determination of Increased Radioactivity Caused by Increased Numbers of Cells/Filter. Experiments were set up to determine whether increased radioactivity observed on filters placed over ETS represented increased numbers of cells. Various concentrations of cells were pipetted onto filters and incubated. Filters were prepared for liquid scintillation counting and for visual quantitation as described in materials and meth- ods. Table II shows the correlation between visual counts of cells on each filter with the radioactivity of parallel filters. The results 15 16 indicate that as the concentration of cells increase, the radioactivity increases proportionally. ,The specific activity of the cells for each concentration was approximately the same, which indicated a homogeneous labeling of cells. The overall mean and standard error (SE) of the specific activity was 0.034 + 0.005 CMP/cell. An overall mean of ad- herence was 53.3 :_8.9%. Comparisdn of Visual and Radioactive Assays With Macrophage Chemo- tactants. Since increased numbers of cells correlated with increased radioactivity on the filters when cells were layered onto the filters, it was then possible to compare visual and radioactive methods of quan- titating chemotaxis. Both tumor supernatants and ETS were used as chemotactants in this comparison. Each sample of chemotactant and control medium was set up in four Boyden chambers. After incubation, two filters were prepared for scintillation counting and two filters were prepared for microscopic observation. The results of this compar- ison in three separate experiments are shown in Fig. I. When either tumor supernatant or ETS was present, there was an increase in observed cell numbers on the lower filter that correlated with an increase in the precipitable CPM. Note that there was migration of macrophages to the lower filter of negative controls in both the visual and radio- active assay, but this migration represented random migration of cells in response to the medium. The migration of macrophages in response to TS and to ETS was greater than the random migration both in the visual and in the radioactive assay. Determination of Chemotaxis or Chemokinesis of Tumor Supernatant. It is apparent from Fig. I that tumor supernatant caused greater mi- gration of macrophages to the lower filter than did control medium. 17 Since it was uncertain whether the tumor material was chemotactic or chemokinetic fur macrophages, experiments were set up to clarify this 'point. Chambers were assembled with TS in the lower well, cells in the , top well; TS in both wells; and TS in the top well in contact with the cells and i1066 in the lower well. Assessment of macrophage migration to the lower filter by the radioactive method (Table III) showed that the greatest increase in radioactivity occurred when cells were in the top well and TS in the lower well. When TS was in contact with the cells and either T8 or medium was in the lower well, the radioactivity observed was comparable to negative controls (cells in the top well, i1066 in the lower well), which indicated random migration. DISCUSSION The purpose of this study was to develop a rapid in vitra chemo- taxis assay for mouse peritoneal macrophages. The method described used [5,6-3H]uridine labeled PEG and a double polycarbonate filter in each blind-well Boyden chamber. Precipitation of the cellular material on the lower filter with TCA allowed assessment of cells which had mi- grated through the first filter while nonechemotaxing PEC on the first filter were discarded.' The technique allowed one to process a large number of samples in a minimum amount of time. Contrary to the lengthy procedure of counting cells on filters, the results of this assay were available immediately after liquid scintillation counting. These data indicate that radiolabeled macrophagescould be used to assess chemotaxis for a known chemotactant (ETS) and it can also be used to test for chemotactic activity of other material such as T8 18 (Table I). When supernatants from the fOurth passage of embryonic fibroblasts were tested for chemotactic activitx.radioactivity of the lower filters were comparable to control filters (Data not shown). It was concluded, therefore, that normal cells with moderate growth rates do not produce a material comparable to the tumor material, or pro- duce it in extremely low quantities not detectable by the assay system described. We showed that increasing cell numbers adherent to the filter caused an increase in the CPM per filter (Table 11). With the excep- tion of the lowest concentration tested, the percent macrophages ob- served at each concentration were similar. The decrease in adherence of macrophages at the lowest concentration was observed in several experiments. Within the upper range of concentrations tested, however, varying the concentration of cells did not alter adherence of cells to the filters. When the data from Table II were graphed on semi-log paper, a straight line graph was obtained. From such a standard curve, one could calculate the number of cells migrating to the second filter in the radioactive chemotaxis assay if desirable. When such determina- tions from several experiments were made, we observed that the radio- active method predicted approximately 10% more cells than were counted in the visual assay. Two reasons could account far this difference. First, only whole cells on the surfaces of the filters which were well spread were counted in the visual assay. No attempt was made to count cells within the pores of the filters. Second, physical disruption of cells occurred in separation of upper and lower filters. Portions of cells could be observed microscopically, but were not counted. It is 19 likely that the radioactive method detected labeled portions of dis- rupted cells remaining on the filter, as well as those cells within the pores. When the visual and radioactive assays for chemotaxis were com- pared (Fig I) using a known chemotactant (ETS) and tumor supernatant (TS), the radioactive method was sensitive to the increased number of cells present on the lower filters of ETS and TS as compared to con— trols. The radioactive method did not distinguish between cell types migrating to the lower filter. However, visual assessment of cells on the lower filter determined that less than 2.5% of the cells observed were polymorphonuclear leukocytes. The remaining cells were all macrophages. Various investigators (20,56) have described counting methods for determining random (chemokinesis) versus directional movement (chemo- taxis) of cells toward a substance. The radioactive method described in this paper was useful in differentiation of chemokinesis and chemo- taxis. The data from Table III established that T8 was a chemotactant for macrophages. If TS merely increased random movement of macrophages, filters from chambers containing cells and TS in the top well and TS in the lower well should contain approximately the same radioactivityas those chambers containing cells in the top well and TS in the lower well. The same would be true for chambers containing cells and TS in the top well and i1066 in the lower well. Since the radioactivity of filters from sets of chambers containing cells and TS in the top wells was comparable to the control, and the radioactivity of filters from chambers with cells in the top well and TS in the lower well was much greater, the TS stimulated directional movement (chemotaxis) of 20 macrophages. The possibility that the tumor material chemotactic for macro- phages was the result of degrading tumor cells in the serum—free medium was minimal. After 48 hr incubation in serum-free environment, the monolayer of cells remained intact. The viability of tumor cells which had detached from the surface of the flask was monitored when the supernatant was collected and centrifuged. Greater than 96% of the pelleted cells were viable based on exclusion of Trypan Blue dye. The tritiated uridine method used to label macrophages in this research has advantages over the use of 51Chromium and 99mTechnetium 51Cr to label to label cells for chemotaxis assays (9,13). Use of cells causes clumping of leukocytes which necessitates an additional filtering through sterile gauze to remove clumps (13). Preparation of additional cells would therefore be necessary to compensate for loss of clumped cells. The inavailability and short half-life (5.9 hr) of 9ngc make this label difficult to work with. Also, corrections have to be made for the decay of the label if large numbers of samples are processed (9). Papierniak et al., (34) also report a high background on filters due to spontaneous release of label from the cells. These problems were not an issue when macrophages were labeled with1fiLuridine. Cell clumping did not occur during the labeling process. The long half life of tritium and the availability and ease of incorporation of 3H... uridine made it an ideal label for macrophages. Any unincorporated label which might be spontaneously released was eliminated either dur- ing the washing procedure or during the TCA precipitation of the filters. The radiolabeling technique described herein has several advan- tages. It can be used to assess mouse peritoneal macrophage 21 chemotaxis objectively, and can be used to distinguish chemokinesis from chemotaxis. It is also very useful in processing large numbers of samples which would be generated in purification procedures of any chemotactant (43)- 22 .2mm.H zuw>fluoeofioma oweno>m mucomonmon ounce H... E... + :52 na NS. u 89.. z: + 32.. e 9m Se H 33 m.n on» u 23 SH n 28 m «A S H «a: m.~ 2 .+. 38 we H 23 N N6. Rm H 2...: N5 2:. H 2:: 3:. u 3% H me you , «we mam you «mam «Honueou pewsoaomxm Honumou no>o Honueou no>o ammonucm ofiom ommohoeH oHom mfixmuoaonu mmomm< op mommamonomz poaeomq ocwoflhaumn we mm: H mamapu< oamfiuoam nopamm\zao ee>pmmpo mHHou euuaua mHHmu mo Hocasz coo: mo Hopasz youfima you eo>ummpo fizauv aua>mpoaomemm ee< nouawm Hon mHHou mo Hooasz noesuom cowpmaouhou HH mam<fi Heywood He>o endomoefi vaom mueomoumou mwmonueohem cw oaao>e 24 A .+. n2 8.: A u. «.2 8.8 a H mm 2.3 S .+. 3.0 m N n. «.2 2.8 a H o: 3.: 3 M 8m 9.: e .+. m8 N N H m2 £43 a: H an». ad a H RN «8: 2: H 82 a uneafihomxm 8:5 08:. m... E. ":03 .858 mambo my + mnmou wk + mauou mambo "Ham: 90% e m N a amassz Sneeze aneumeuogaw moans photos mome:m0Hoez “unconfined mo mwxouoaonu mo mowuemwauouon HHH mam<fi 25 Figure 1. Comparison of visual and radioactive methods for quantita- tion of chemotaxis. 3H.Uridine labeled macrophages were used in a chemotaxis system as descirbed in the text. Microscopic observation of filters was done in parallel with determination of precipitable radioactivity on filters. Data represent the mean for three experi- ments containing duplicate filters :_standard error of the mean. ‘ D Visual § Radioactive 26 3 To. x 2.6 2 . l§§§§§ II. Tl$x§§§§x 0 3 0 2 O GAME \mJJMO mumZDZ wc< TS ETS Figure 1. CONTROL ISOLATION AND PARTIAL PURIFICATION OF A MACROPHAGE CHEMOTACTANT PRODUCED BY SAD/2 FIBROSARCOMA CELLS GROWN IN VITRO by Louise Schaub Simon, Ronald Patterson, and Tobi Jones INTRODUCTION The effect of tumor cell products on macrophage chemotaxis has been the subject of much research in recent years. Several researchers have reported that patients with cancer have depressed monocyte chemo- tactic responses in vitra (3,50). Murine syngeneic neoplasms have also been reported to depress macr0phage migration in viva and chemo- taxis in vitra (36,47,48). These investigators have isolated by dial- ysis a 6—10,000 MW factor from tumor cells which depressed macrophage chemotaxis (36,48). Stimulation of macrophage function by tumor cell products is also reported in the literature. Snodgrass et a1. (44) reported that tumor cells produce a soluble product which stimulates macrophage chemokinesis in vitro. Meltzer et a1. (27) have reported that culture supernatants of several in vitro tumor lines contain a 15,000 MW component which is chemotactic for macrophages. We reported previously (42) that growth medium from the murine SAD/2 fibrosarcoma line contained material chemotactic for macrophages in vitro. Similar chemotactic activity was not obtained from the growth medium of fourth passage mouse embryonic fibroblasts. In the present report, the isolation and partial purification of radiolabeled chemotactic products from culture supernatants of SAD/2 cells is des- cribed. In addition, the ability of the partially purified material to cause an inflammatory response in mice is assessed. MATERIALS AND METHODS Mgdia, Sterile Eagle's minimal essential medium (MEM) was used for collection and radiolabeling of peritoneal exudate cells. Tumor cells were maintained in CMRL 1066 medium containing 10% heat-inactivated 28 29 fetal calf serum (d1066). Serum-free CMRL 1066 (i1066) was used to generate culture supernatants from tumor cells. Leucine-free MEM (LMEM) prepared according to Eagle (7), was used instead of i1066 when the supernatant products of tumor cells were to be labeled with tritiated leucine. All media contained 50 U/ml penicillin, 50 ug/ml streptomycin, and were buffered with NaHCOS. All components were obtained from Grand Island Biological Co., Grand Island, NY, and Microbiological Associates, Walkerville, MD. Cells Used. An in vitra line of the DEA/2 methylcholanthrene induced fibrosarcoma SAD/2 (obtained from Jackson Labs, Bar Harbor, ME) was used as the source of macrophage chemotactant according to Simon et al., (42). Macrgphage Chemotaxis. Chemotactic activity of tumor cell products was measured as previously described (42). Two to six month old male DBA/Z mice (Jackson Labs, Bar Harbor, ME) were used as a source of peritoneal macrophages for all in vitro chemotaxis studies. Peritoneal exudates were induced in these mice by intraperitoneal (1?) injection of protease peptone (Difco Laboratories, Detroit, MI). Mice were sacrificed by cervical dislocation, and exudate cells were col- lected by lavage of the peritoneal cavity with 5 ml MEM. The cells were counted, centrifuged at 163 x g, and resuspended cells were lab— eled with 10 uCi [5,6-3H]uridine (Amersham, 48 Ci/mmol)/107 PEC/ml MEM for 1 hr at 37 C. The cells were washed three times, and resus- pended in MEM containing 2% bovine serum albumin [(BSA) Sigma] at a concentration of 2.2 x 106 PEC/ml. Blind-well modified Boyden chambers were used in this study. Two polycarbonate filters (Nucleopore Corporation, Pleasonton, CA) with 30 5 um pores separated the lower well containing control medium or chemo- tactant from the upper well containing 4.4 x 105 uridine labeled PEC in a 0.2 ml volume. Following incubation of the chambers at 37 C in a humidified atmosphere of 5% CO2 for 3.5-4 hr, chemotaxis of macro- phages was quantitated by trichloroacetic acid (TCA) precipitation of the cells which migrated onto the lower filter followed by liquid scin- tillation spectrometry. Serum-free supernatants collected from SAD/2 tumor cells previously shown to be chemotactic for macrophages (42), and column fractions pre- pared from these supernatants, were used as chemotactants. In all ex- periments, endotoxin (Sigma)-treated mouse serum, prepared as described by Snyderman and Pike (46) was included as a positive control for chemo- taxis. When crude tumor supernatants were tested fer chemotactic acti- vity, the negative control was i1066 placed in the lower well. Bluates from the Sephacryl $200 and Sepharose 68 columns taken prior to appli- cation of samples to the columns were used as the negative controls when fractions from these columns were tested fer chemotactic activity. Radiolabeling of Tumor Supernatants. The chemotactant from SAD/2 cells was labeled in the following manner. Tumor cells were cultured in d1066 for 24 hr, washed witthBS, and 10 ml of LMEM containing 10 uCi/ml L-[4,5-3H]leucine (48 Ci/mmol, Amersham) were placed on the cells. The cells were cultured an additional 48 hr, and the superna- tants were collected and subjected to chromatographic and electro- phoretic procedures as described below. Chromatography_of Tumor Supernatants. SAD/2 culture supernatants collected after 48 hr of incubation were concentrated 25-30 times by ultrafiltration with a UM 20 filter (Amicon Corporation, Lexington, 31 MA). Five ml of the concentrated supernatant were placed on a Sepha- cryl s-2oo colunm (2.5 x 59 cm bed) equilibrated with phosphate buf- fered saline (PBS). Elution buffer was 0.15 m PBS, pH 7.2. Flow rate was 33 ml/hr, and was controlled by a peristaltic pump (Pharmacia). Absorbance was read at 280 nm. Fractions (3.2 ml) were collected and were assayed in duplicate for chemotactic activity. Those fractions which had chemotactic activity were pooled, concentrated by ultrafil- tration 446 fbld and placed on a Sepharose 6B column (2.5 x 63.5 cm bed) equilibrated with 0.15 M PBS. Flow rate was 18.8 ml/hr and was controlled by a Mariott flask (Pharmacia). The absorbance and collec- tion of fractions were the same as for the Sephacryl column. Fractions were assayed in duplicate for chemotactic activity. Fractions from the Sepharose 68 column which contained chemotac- tic activity were applied to a DEAE Sephacel column (0.80 x 11.6 cm bed) which was equilibrated with 0.15 M PBS. Salt elutions were made with 0.15 M, 0.3 M and 0.6 M NaCl in 0.004 M phosphate buffer. Eluates were brought to physiological saline conditions by dilution with 0.004 M phosphate buffer, concentrated to the original volume by ultrafiltra- tion, and assayed for chemotactic activity. Gel Electrophoresis. Ten percent polyacrylamide gels containing 0.1% sodium dodecylsulfate (SDS) were prepared as described by Porzio and Pearson (37). Native gels were prepared in the same manner, ex- cept that SDS was omitted in preparation of the gel. All ingredients for the gels were purchased from Bio Rad Laboratories, Richmond, CA. The buffer used for SDS gel electrophoresis was 200 mM Tris/glycine (pH 8.5) with 0.1% SDS. The same buffer without SDS was used for electrophoresis of native gels. Fifty ml containing approximately 32 10 ug protein (determined as in Lowry et al., 23) of the chemotactively active Sepharose 68 fractions were placed on the gel and allowed to electrophorese at 1 mA per tube for approximately 2 hr. Gels were fixed, stained with Coomassie Brilliant Blue, and destained according to Porzio and Pearson (37). Gels were scanned at 540 nm to obtain protein profiles. When radiolabeled tumor material was electrophoresed, the gels were cut into approximately 1 mm slices with a gel slicer (Bio Rad). Two adjacent slices were placed in 5 ml toluene based scintilla- tion fluid containing 5% NCS tissue solubilizer and 1% 4N NH4OH and incubated at 37 C overnight. Radioactivity was determined by liquid scintillation spectrometry and was expressed as counts per minute (CPM). Marker proteins and their molecular weights used to determine molecular weight of the tumor material on SDS gels were a actinin, from myofibril protein (gift of'M.A. Porzio) 102,000, BSA (Sigma) 68,000, Soybean trypsin inhibitor (Sigma) 22,000, hemoglobin (Schwarz Mann) 15,500, and lysozyme (Sigma) 13,000. Trypsin Treatment of Chemotactant. Sepharose 68 column fractions with chemotactic activity were subjected to 5% insoluble trypsin (Sigma) treatment far 50 min at 37 C. Trypsin-bearing beads were removed by centrifugation, and the supernatant was applied to SDS gels as previous- ly described. This supernatant was also tested for chemotactic activity. In one experiment the Sepharose 68 active fractions were trypsinized overnight at 15 C prior to application to SDS gels. In Viva Inflammatory Response. Groups of five female (DEA/2 x C57Bl/6) Fl mice were injected IP with 0.5 m1 of 4 feld concentrated Sepharose 6B fractions which had chemotactic activity. Groups of five control mice were injected with the same volume of PBS. Forty-eight 33 hours after injection, mice were sacrificed by cervical dislocation, and peritoneal cells from each mouse were collected by lavage of the peritoneal cavity with 5 m1 MEM. The volume obtained from each mouse was recorded and the total number of peritoneal cells obtained from each mouse was calculated from hemocytometer counts. The cells were centrifuged at 163 x g and resuspended in 0.2 ml MEM. Cells were pipetted onto duplicate slides, air dried, fixed in methanol, and stained with Giemsa stain. Differential cell counts on preparations from each animal were determined by counting 100 cells on each slide at 1000 X under oil. Statistical Evaluation. Chemotactic activity was expressed as the mean CPM of duplicate filters :_standard error (SE). For the inflammatory response studies, the mean number of recovered peritoneal cells and the mean number of each cell type observed in the differen- tial counts from control and experimental animals was compared by the Student's t test. RESULTS Chromatggraphy of Tumor Supernatants. Experiments were conducted to determine whether chemotactic material contained in culture super- natant from SAD/2 cells (42) could be purified by gel filtration. Chemotactic activity was retained when crude supernatants were concen- trated with an Amicon UM 20 filter which retains molecules of greater than 20,000 MW. Crude supernatants which had been concentrated were applied to a Sephacryl $92000 column. The protein profile and corres- ponding chemotactic activity of fractions from this column are depicted in Fig. IA. Greatest chemotactic activity was feund in fractions (12- 15) immediately following the void volume. When fractions with greatest 34 chemotactic activity were pooled, concentrated and applied to a Sepha- rose 68 column, two major protein peaks were obtained (Fig. IB). Chem- otactic activity (fractions 15-23) was found to correspond with the second protein peak (Fig. IB). A 4-5 feld increase in radioactivity occurred when Sepharose 68 active fractions were compared to crude supernatant in the chemotaxis system. In order to determine whether the chemotactant was actively syn- thesized, tumor cells were radiolabeled with 3H-leucine and the super- natants were chromatographed on Sephacryl S 200 and Sepharose 63. Those fractions with chemotactic activity were also found to be labeled with 3H-leucine (Fig. IB). Several attempts to purify Sepharose 6B active fractions with DEAE Sephacel resulted in loss of biological activity when 0.1 M 0.3 M and l M NaCl eluates were tested for chemotactic activity. Fractions obtained when a gradient elution (0-1 M NaCl) was used also contained no chemotactic activity. Gel Electrgphoresis. The fractions from the Sepharose 68 column which had greatest activity corresponded to a single protein peak (Fig IB). Since attempts to further purify these fractions on DEAE Sephacel were unsuccessful, the active fractions were analyzed by gel electro- phoresis to determine whether this peak consisted of one or several proteins. SDS gels were used to assess whether the active peak con- tained more than one protein separable by electrical charge. The results of SDS gel electrophoresis of crude tumor supernatant compared with Sepharose 6B fractions (50 ul containing 10 pg protein applied to each gel) which had chemotactic activity are shown in Fig. 2, A and B. Sepharose 6B fractions contained two Coomassie Blue positive 35 bands on SDS gels. Treatment of active Sepharose 68 fractions with 2 mercaptoethanol (2MB) prior to electrophoresis resulted in gels with 2 bands (Fig. 2C). To confirm that these bands were protein, the Sepharose 6B frac- tions which were chemotactively active were trypsinized, and 50 ul were applied to SDS gels. Fig. 2D shows a diminished amount of Coomassie Blue positive bands after 50 min trypsin treatment. When active frac- tions were treated with trypsin overnight prior to application to gels, there were no Coomassie Blue positive bands after staining of gels. Trypsinization of chemotactively active Sepharose 68 fractions also resulted in loss of chemotactic activity when tested in vitra (data not shown). Radiolabeled Sepharose 68 fractions with good chemotactic activity were applied to SDS gel electrophoresis. Results of these experiments (Fig. 3) show that the two radioactive peaks correspond to the two protein peaks found on SDS gel scans. The apparent molecular weights of the proteins in the Sepharose 68 active fractions were 68,000 and 78,000 on SDS gel electrophresis (Fig. 3). A shift of apparent molecular weights to 74,000 and 81,500 respectively occurred when active fractions were pretreated with 2MB befbre application to SDS gels. Native gels contained 2 bands (Fig.4) which supported data (Fig. 3) obtained from SDS gels. Attempts were made to retrieve biological activity from protein bands in native gels. Slices were made of the gels, and they were extracted overnight with PBS while dialyzing against PBS. Extracts tested had no chemotactic activity. Experiments were carried out to determine whether loss of 36 chemotactic activity was due to toxicity of the gel. Freshly prepared gels were sliced and extracted overnight similar to the gels containing protein bands. Extracts from these gels were used as a diluent in the preparation of ETS in place of PBS. This preparation was tested in parallel with ETS containing PBS diluent in a chemotaxis system. The chemotaxis of macrophages toward ETS containing gel-extract diluent was reduced by 70% compared to ETS containing PBS diluent. Inflammatory_Response. The partially purified tumor material was chemotactic fer peritoneal macrophages in vitra, and tests were con- ducted to determine whether it exhibited activity in viva. To deter- mine whehter chemotactically active Sepharose 68 fractions would elicit an inflammatory response in viva, 0.5 ml of the active fractions were injected IP into mice. Total cell yield and differential counts from the peritoneal wash of individual mice were determined (Table 1). Although the total cell number of peritoneal cells obtained from ex- perimental animals was greater than that obtained from control animals, the values were not statistically significant. Results of the differ- ential cell count, however, demonstrated a significant increase in the number of macrophages, and a significant reduction in the number of lymphocytes of experimental animals when compared to controls. The number of basophils, polymorphonuclear neutrophils, and eosinophils was not significantly different in experimental and control mice. DISCUSSION Previous research from this laboratory has determined that SAD/2 tumor cells grown in vitra in serum-free medium produce a material that is chemotactic for peritoneal macrophages in vitro (42). Puri— fication of this material was desirable in order to investigate its 37 biological properties in viva. If tumor cells produce such a material in viva, this could account fer the accumulation of large numbers of macrophages which have been reported in this tumor in viva (21). The methodology used to evaluate chemotaxis using radiolabeled cells was described in a previous paper (42). This technique greatly facilitated the assessment of numerous fractions obtained in gel fil- tration of crude tumor supernatant. Our isolation procedures utilized two gel filtration columns. The Sephacryl $200 column eliminated many low molecular weight components found in the crude supernatant. Sepharose 6B was utilized to separate higher molecular weight components from the proteins associated with chemotactic activity. Serum-free tumor supernatant was used in this research, and made interpretation of results easier for two reasons. First, most of the serum components were eliminated from gel beds making serum contamin- ants in the active fractions less likely. Secondly, without serum in the medium in which the tumor cell products were generated, there was less chance that the chemotactant was a cleavage product of serum gen- erated by the action of tumor cells on the serum. To show that the proteins in chemotactic Sepharose 68 fractions were synthesized by the cells, attempts were made to radiolabel the proteins associated with chemotactic activity. Results of these ex- periments showed that radioactivity was associated with both proteins found in chemotactic fractions (Fig. 18 and Fig 3). We concluded from this data that the tumor material was actively produced by the cells and was not a cleavage product of residual serum components. Identification of which of the two proteins in Sepharose 6B 38 active fractions was the chemotactant was complicated by loss of bio- logical activity when samples were applied to DEAE Sephacel. Sepha- rose 68 active fractions were therefore applied to native gels and retrieval of chemotactic activity from gel slices was attempted. These attempts were unsuccessful because of the toxicity of the gel in our system. We rechromatographed the active fractions on Sepha- rose 68 and applied individual fractions associated with the protein peaks to gel electrophoresis. Two bands were always observed in each individual fraction, which indicated they had similar molecular weights. The molecular weights of the two proteins found in these active fractions were 68,000 and 78,000 based on SDS gel electrophor- esis. Pretreatment of active fractions with ZME followed by SDS gel electrophoresis resulted in a shift of apparent molecular weight to 74,000 and 81,500. The reason fbr this shift is not known at this time but could be due to breaking of intrachain sulfhydryl bonds and subsequent unfolding of the molecule making it more accessible to SDS. The proteins were apparently composed of a single polypeptide chain since no additional bands were observed in gels containing samples treated with ZME (Fig. 2C). Although data from gel electrophoresis would suggest that the tumor material contained only 2 proteins, this does not eliminate the possibility that a small molecular weight species might adhere non- covalently to the larger molecular weight species. This possibility, however, is minimal, since treatment with SDS should dissociate the small molecular weight species and make it apparent as a separate band farther down the gel. This was never observed even when concen— trated crude supernatants or Sepharose 68 fractions were applied to 39 SDS gels for electrophoresis (Fig. 2A and 2B). There is reason to believe one or both of these proteins, ob- tained from the Sepharose 68 column cause chemotaxis of macrophages. First, trypsinization of these active fractions caused disappearance of chemotactic activity. Second, the same trypsinized sample resulted in loss of bands on SDS gels (Fig. 2D). Thus there was a correlation between the loss of biological activity by trypsinization and the dis- appearance of bands on SDS gels. The effect of these partially purified samples on macrophage migration in viva was tested by IP injection into mice. When periton- eal cells (PC) were harvested 48 hr later, a significant increase in the number of macr0phages in the peritoneal cavity had occurred. Al- though there was no significant difference in the total cell number harvested from control and experimental mice, this could be explained by preliminary experiments in our laboratory which indicate that when PC are pretreated with Sepharose 68 fractions, the in vitra adhesive- ness of peritoneal macrophages is increased (Unpublished data). The data in this paper is different from the data of other re- searchers (36,48,49). These researchers indicate that tumor cell pro- ducts depress macrophage chemotaxis. However, they attribute depressed macrophages migration in viva and macrophage chemotaxis in vitra to a low molecular weight (10,000 MW) component extracted from various tumor cell lines. Meltzer et a1. (27) reported that a low molecular weight (15,000 MW) chemotactant from serum-containing tumor supernatants stimulated macrophage migration in vitra. However, these researchers did not determine whether this product was due to alteration of serum components by the tumor cells, or whether it was actively synthesized 40 by the cells (27). Our research indicates that two large molecular weight components (68,000 and 78,000 MW) found in growth medium of tumor cells cause en— hanced macrophage chemotaxis in vitro and an inflammatory response in viva. The enhancement of inflammatory response in viva by tumor cell products could be one ~reason why numerous macrophages are found in this tumor (21). One reason for the differences we observed with other researchers findings is the tumor types. The fibrosarcoma used in this research is different from the tumor lines used by other researchers (27,48). The SAD/2 tumor line produces material which is chemotactic fer macro- phages in vitra and elicits an inflammatory response in viva. It is possible that in viva the tumor cells produce this material and may be one reason why macrophages are selectively drawn into the tumor and remain localized there. The role of macrOphages in tumors was suggested by Eccles and Alexander (8) and by Evans (10) to be control of metasta- sis. Wood and Gillespie (55) have shown that when tumors were depleted of macrophages prior to injection, metastasis of the tumor resulted. Recent research has indicated that a soluble product is involved in cytotropic responses of macrophages to tumor cells in vitra (44). The argument for a macrophage chemotactant produced by tumor cells in viva is made more plausible by these researchers findings, and our own findings presented in this paper. If tumor cells produced a macro- phage chemotactant in viva, this could explain how macrophages are attracted to the tumor site, and why they remain localized within the tumor mass . 41 Figure 1. Chromatograph of Serum—free Tumor Supernatant. A. Protein profile (--) and chemotactic activity (O———0) of fractions obtained when tumor material was applied to the Sephacryl 8200 column. B. Pro- tein profile (-—) chemotactic activity (O———0) and radioactivity of fractions (0———0) obtained when biologically active Sephacryl S200 fractions were applied to the Sepharose 6B columns. Arrow=phenyl red marker. 42 ABSORBANCE (280nm) 5 IO lb 50 2'5 3'5 40 FRACTION Figure l. CPM x 10'3 43 Figure 2. SDS gel electrophoresis of crude tumor supernatant and chemotactively active fractions from the Sepharose 68 column. A. Crude tumor supernatant. B. Sepharose 68 active fractions without 2MB. C. Sepharose 68 active fractions with 2MB. D. Tryp- sinized Sepharose 68 active fractions. E. Myofibril standards. F. Other standards as described in the text. 44 I-.— awn-.— a-.-— . -0 .— ~__ A h ' a ' \ ‘V a ‘4... tho—o Figure 2. -._-.._.—_—-. Fwy... 45 Figure 3. Protein profile of radiolabeled Sepharose 68 active frac- tions on SDS gel electrophoresis. Protein profile (—--9 and radio- activity of Radiolabeled Sepharose 68 active fractions (l———0) on SDS gel electrophoresis. Molecular weight standards were: A. a actin (102,000 MW), B. Bovine serum albumin (68,000 MW), C. Soybean trypsin inhibitor (22,000 MW), D. Hemoglobin (15,500 MW) and lysozyme (13,000 MW). TD=tracking dye. NOILVHOIW BALLV'IBH 46 ABSORBANCE (540 nm) e) «an +0 to m o: as to g CPMXIO'Z Figure 3. 47 Figure 4. Protein profile of active Sepharose 6B fractions on native gel. TD=tracking dye. 48 l a? .— J— Li ‘1 1°. N. -. 0 O O O (“W 0179) BONVBHOSBV Figure 4. RELATIVE MIGRATION .mosam> Heywood ow oopmmaoo can: Ho.o v no .ooueaoo mHHou co“ pom oak» Hfioo come we w mucomohmou mumom 49 o e.o + 0.9 o w.c + m.H mamcaocmhom m.o.u m.o e.o.u m.o o o.o.u m.o maaaaoapsoz ~.o.u e.o e.o.u H.H q.o.u m.o m.o.u o.H meanaomwm m.~.n a.ae w.~.u «.ma m.m.u ~.we 5.4.“ o.aH amouaoogaeaa m.~.u A.om m.~.u H.ma m.~.n 0.0m o.e.n m.ma amowacaowoaz mucsou HHou . mmmpnowooomo H.o.u Hm.o H.o.u H.H H.o.u em.o H.o.u H.H wonemmsmwmwm>< mHouucou oouoomeH mo mHonueou oouoonem mo ucoaumohh N peoEwHomxm H ueoaflnomxm Hamwtpwz Hosea vehemwsa amawmowwa ow comm 2N oncommoh aboumssonmm mo ucoeazmmoz H mam<fi BI BLIOGRAPHY 10. 11. 12. BIBLIOGRAPHY Barth, R.P., o. Singla, and G.Y. Gillespie. 1974. Use of gngc as a radioisotopic label to study the migratory patterns of normal and neoplastic cells. J. Nucl. Med. 15:654-661. Blakeslee, D. 1977. Enhanced non5pecific colony inhibition of mouse tumor cells in vitra by macrophages in the presence of a culture tumor factor. J. Reticulo. Soc. 23:447-458. Boetcher D.A. and E.J. Leonard. 1974.Abnormal monocyte chemotactic response in cancer patients. J. Natl. Cancer Inst. 52: 1091- 1099. Boyden, S. 1962. The chemotactic effect of mixtures of antibody and antigen on polymorphonuclear leucocytes. J. Exp. Med. 115: 453-466. Carpenter, R.R., P.B. Barsales, and R.P. Ganchan. 1968. Antigen- induced inhibition of cell migration in agar gel, plasma clot, and liquid media. J. Reticulo. Soc. 5:472-483. Cutler, J.E. 1974. A simple in vitra method for studies on chemo- taxis. Proc. Soc. Exp. Biol. Med. 147:471-474. Eagle, H. 1959. Amino acid metabolism in mammalian cell cultures. Science. 130:432-437. Eccles, S.A. and P. Alexander. 1974. Macrophages content of tumors in relation to metastatic spread and host immune reaction. Nature. 250:667-669. English, D. and T.A. McPherson. 1977. Chemotaxis Radioassays: a simplified, quantitative method using Technetium-99m radiocolloid labeled granulocytes. Am. J. Hematol. 3:245-256. Evans, R. 1972. Macrophages in syngeneic animal tumors. Trans- plantation 14:458-473. Evans, R. 1973. Macrophages and the tumor-bearing host. Br. J. Cancer 28, Suppl. 1:19-25. Evans, R. and P. Alexander. 1972. Mechanism of immunologically specific killing of tumor cells by macr0phages. Nature 236:168-170. 50 13. 14. 15. 16. l7. 18. 19. 20. 21. 22. 23. 24. 25. 51 Gallin, J.I., R.A. Clark, and H.R. Kimball. 1973. Granulocyte chemotaxis: an improved in vitro assay employing Cr-labeled granulocytes. J. Immunol. 110:223-240. Gollahon, R.A. and G.W. Wood. 1977. Characterization of the tumor associated host cellular response to primary 3-methyl- cholanthrene induced murine fibrosarcomas. Proc. Amer. Assoc. Cancer Res. 18:144. Haskill, J.S. 1977. ADCC effector cells in a murine adenocar- cinoma. 1. Evidence for blood-borne bone-marrow-derived mono- cytes. Int. J. Cancer 20:432-440. Haskill, J.S., J.W. Proctor, and Y. Yamamura. 1975. Host re- sponses within solid tumors. 1. Monocytic effector cells within rat sarcomas. J. Natl. Cancer Inst. 54:387-392. Harris, H. 1954. Role of chemotaxis in inflammation. Physiol. Rev. 34:529-562. Horwitz, D.A. and M.A. Garrett. 1971. Use of leucocyte chemo- taxis in vitra to assay mediators generated by immune reactions. 1. Quantitation of mononuclear and polymorphonuclear leucocyte chemotaxis with polycarbonate (Nucleopore) filters. J. Immunol. 106:649-655. Jones, T.L. 1977. Alteration in macrophage function during tumor growth. Fed. Proc. 36:1256. Keller, H.U. and E. Sorkin. 1967. Studies on chemotaxis VI. Specific chemotaxis in rabbit polymorphonuclear leucocytes and mononuclear cells. Int. Arch. Allergy 31:575-586. Kerbel, R.S. and H.F. Pross. 1976. Fc receptor-bearing cells as a reliable marker for quantitation of host lymphoreticular infil- tration of progressively growing solid tumors. Int. J. Cancer 18:432-438. Kerbel, R.S., H.F. Pross, and A. Leibovitz. 1977. Analysis of established human carcinoma cell lines for lymphoreticular-asso- ciated membrane receptors. Int. J. Cancer 20:673-679. Lowry, O.H., N.J. Rosenbrough, A.L. Farr, and R.J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. Magarey, C.J. and M. Baum. 1970. Reticulo-endothelial activity in humans with cancer. Br. J. Surg. 57:748-752. Meltzer, M.S. and M.M. Stevenson. 1977. Macrophage function in tumor-bearing mice: Tumoricidal and chemotactic responses of macro- phages activated by infection with Mycobacterium bovis, strain BCG. J. Immunol. 118:2176-2181. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 52 Meltzer, M.S. and M.M. Stevenson. 1978. Macrophage function in tumor-bearing mice: Dissociation of phagocytic and chemotactic responsiveness. Cellular Immunol. 35:99-111. Meltzer, M.S., M.M. Stevenson, and J. Leonard. 1977. Characteri- zation of macrophage chemotaxins in tumor cell cultures and compar- ison with lymphocyte-derived chemotactic factors. Cancer Res. 37: 721-725. Meltzer, M.S., R.W. Tucker, and A.C. Breuer. 1975. Interaction of ECG-activated macrophages with neoplastic and nonneoplastic cell lines in vitra: Cinemicrographic analysis. Cellular Immunol. 17:30-42. Nelson, R.D., P.G. Quie, and R.L. Simmons. 1975. Chemotaxis under agarose: a new and simple method for measuring chemotaxis and spon- taneous migration of human polymorphonuclear leucocytes and mono- cytes. J. Immunol. 115:1650-1656. Normann, S.J. and E. Sorkin. 1977. Inhibition of macrophage chem- otaxis by neoplastic and other rapidly proliferating cells in vitra. Cancer Res. 37:705-711. North, R.J., D.P. Kirstein, and R.L. Tuttle. 1976. Subversion of host defense mechanisms by murine tumors 1. A circulating factor that suppresses macrophage-mediated resistance to infection. J. Exp. Med. 143:559-573. North, R.J., D.P. Kirstein, and R.L. Tuttle. 1976. Subversion of host defense mechanisms by murine tumors. II. Counter influence of concomitant antitumor immunity. J. Exp. Med. 143: 574-584. Otu, A.A., R.J. Russell, P.G. Wilkinson, and R.G. White. 1977. Alterations of mononuclear phagocyte function induced by Lewis lung carcinoma in C57Bl mice. Br. J. Cancer 36:330-340. Papierniak, C.K., R.E. Bourey, R.R. Kretschmer, S.P. Gotoff, and L.G. Colombetti. 1976. Technetium-99m labeling of human monocytes fer chemotactic studies. J. Nuc. Med. 17:988-992. Perillie, P.E. and S.C. Finch. 1964. Quantitative studies of the local exudative cellular reaction in acute leukemia. J. Clin. In- vest. 43:425-430. Pike, M.C. and R. Snyderman. 1976. Depression of macrophage function by a factor produced by neoplasms: a mechanism for abroga- tion of immune surveillance. J. Immunol. 117:1243-1249. Porzio, M.A. and A.M. Pearson. 1977. Improved resolution of myo- fibrillar proteins with sodium dodecyl sulfate-polyacrylamide gel electr0phoresis. Biochim. Biophys. Acta 490:27-34. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 53 Rebuck, J.W., H.I. Coffman, G.B. Blum, and C.L. Earth. 1964. A structural study of reticulum cell and monocyte production with quantitation of lymphocytic modulation of nonmultiplicative type to histocytes. Ann. N.Y. Acad. Sci. 113:595-611. Russell, S.W., W.F. Doe, and A.T. McIntosh. 1977. Functional characterization of a stable, noncytolytic stage of macrophage activation in tumors. J. Exp. Med. 146:1511-1520. Russell, S.W., G.Y. Gillespie, and A.T. McIntosh. 1977. Inflam- matory cells in solid murine neoplasms. III. Cytotoxicity medi- ated in vitra by macrophages recovered from disaggregated regres- sing Maloney sarcomas. J. Immunol. 118:1574-1579. Schuller, G.B., M.J. Snodgrass, and A.M. Kaplan. 1978. Macrophage motility and chemotaxis in response to culture fluids from tumor cell lines. Fed. Proc. 37:1697. Simon, L., R. Patterson, and T.L. Jones. 1979. A rapid method fer assessment of a macrophage chemotactant produced by SAD/2 fibrosarcoma cells grown in vitra. Submitted for publication. Simon, L., R. Patterson, and T.L. Jones. 1979. Isolation and partial purification of a macrophage chemotactant produced by SAD/2 fibrosarcoma cells grown in vitra. Submitted for publica- tion. Snodgrass, M.T., T.M. Harris, and A.M. Kaplan. 1978. Chemokin- etic response of activated macrOphages to soluble products of neo- plastic cells. Cancer Res. 38:2925-2529. Snyderman, R., M.C. Pike, D. McCarley, and L. Lang. 1975. Quanti- fication of mouse macrophage chemotaxis in vitra: Role of C5 for the production of chemotactic activity. Infect. Immun. 11:488-492. Snyderman, R. and M. Pike. 1976. Chemotaxis of mononuclear cells. In "In vitro methods in cell-mediated and tumor immunity," eds. B.R. Bloom and J.R. David, Acad. Pres, N.Y., pg. 651-675. Snyderman, R. and M.C. Pike. 1976. Defective macrophage migra- tion produced by neoplasma: Identificationtrfan inhibitor of mac- rophage chemotaxis. In "The macrophage in Neoplasia," ed. M.A. Fink. Acad. Press, N.Y. pg. 49-65. Snyderman, R. and M.C. Pike. 1976. An inhibitor of macrophage chemotaxis produced by neoplasms. Science 192:370-372. Snyderman, R., M.C. Pike, B.L. Blaylock, and P. Weinstein. 1976. Effects of neoplasms an inflammation: Depression of macrophage accumulation after tumor implantation. J. Immunol. 116:585-589. 50. 51. 52. 53. 54. 55. 56. 54 Snyderman, R., H.F. Seigler, and L. Meadows. 1977. Abnormalities of monocyte chemotaxis in patients with melanoma: Effects of immunotherapy and tumor removal. J. Natl. Cancer Inst. 58:37-41. Ward, P.A. 1968. Chemotaxis of mononuclear cells. J. Exp. Med. 128:1201-1221. Ward, P.A., G.C. Cochrane, and H.J. Muller-Eberhard. 1965. The role of serum complement in chemotaxis of leucocytes in vitra. J. Exp. Med. 122:327-347. Wilkinson, P.G. 1974. "Chemotaxis and Inflammation". Churchill Livingston, Edinburgh and London. Wood, G.W. and R.A. Gollahon. 1977. Detection and quantitation of macr0phage infiltration into primary human tumors with the use of cell-surface markers. J. Natl. Cancer Inst. 59:1081-1086. Wood, G.W. and G.Y. Gillespie. 1975. Studies on the role of macrophages in regulation of growth and metastasis of murine chemically induced fibrosarcomas. Int. J. Cancer 16:1022-1029. Zigmond, S.H. and J.G. Hirsch. 1973. Leucocyte locomotion and chemotaxis. New methods for evaluation, and demonstration of a cell-derived chemotactic factor. J. Exp. Med. 137:387-410. "IIIIIIIII'IIIIIIIIII