Infiltrating Human Pulmonary Tumors Depression of Natural Killer Cytotoxic Activity in Lymphocytes

We assessed natural killer activity of lymphocytes present at the site of human tumors to determine the local effects of the tumor on immune function. Lymphocytes were extracted from human pulmonary tumors of varying histological types. In addi tion to tumor-infiltrating lymphocytes (TIL), peripheral blood lym phocytes from the same patients were prepared. Using a Michaelis-Menten kinetic model and titration of K562 targets in a 61Cr release assay, TIL exhibited a marked depression of maxi mal lytic capacity (V™«) when compared to autologous peripheral blood lymphocytes. In a single cell lysis and binding assay to assess the proportion of target-binding lymphocytes and targetlysing binders, both TIL and peripheral blood lymphocytes had equivalent numbers of lymphocytes binding target cells and an equivalent number of target-binding cells that could mediate cytolysis. Analysis of lymphocyte subsets was performed using mouse monoclonal antibodies. The TIL population expressed markers found on natural killer cells, including HNK-1 and B73.1. Thus, natural killer cells are present at the tumor site, show lytic capability, but appear to be unable to recycle for multiple lytic events. INTRODUCTION We have demonstrated previously that the TIL2 from human pulmonary tumors exhibit marked depression of NK cytotoxic activity compared to autologous PBL (10). However, the mech anism for this depression has not been fully elucidated. In this study, several techniques have been utilized, including the single cell lysis and binding assay and the 51Crrelease assay, to further characterize the nature of depressed NK activity among TIL. The single cell assay prevents effector cell recycling, although it does permit one to estimate the number of effector cells participating in target binding and lysis. In contrast, the 51Cr release test permits recycling of effectors against multiple targets. Used together, they can define the parameters of lytic action by cells. In addition, we sought to determine whether NK cells were actually present at the tumor site or were preferentially excluded by the tumor. Using murine monoclonal antibodies with defined specificity against human lymphocyte cell surface markers, we have determined the relative proportions of NK and other cell populations among TIL and PBL. MATERIALS AND METHODS Specimens. Surgically resected specimens of human pulmonary tu mors were obtained on the day of operation. Tumor specimen weights 1Supported in part by USPHS Grant CA12582. 2The abbreviations used are: TIL, tumor-infiltrating lymphocytes; NK, natural killer; PBL, peripheral blood lymphocytes; PCS, fetal calf serum; IFN, interferon. Received April 16,1984; accepted September 21,1984. ranged from 0.5 to 20.6 g (average, 6.6 g). The majority of tumors were primary adenocarcinomas or squamous cell carcinomas. Also included were metastatic tumors (one melanoma and one osteogenic sarcoma). Preparation of TIL. TIL were purified as described previously (9). Extraneous normal and necrotic tissues were removed from the fresh tumor. The specimens were then minced with scissors in a Retri dish in the presence of 3 to 5 ml of RPMI-1640 (Flow Laboratories, Rockville, MD) supplemented with 10% PCS, 10 mw 4-(2-hydroxyethyl)-1-p¡perazineethanesulfonic acid buffer, and antibiotics. Cell suspensions were allowed to settle 2 to 3 min to remove debris before the supernatant was collected. Additional extraction of cells was performed by pressing residual tumor chunks in medium with a syringe plunger, and the cell suspension was again collected. After washing 3 times, the cells were placed on Ficoll :Hypaque for discontinuous density gradient separation. Dead cells, erythrocytes, granulocytes, and many tumor cells pelleted to the bottom, and the cells at the interface consisting of lymphocytes, monocytes, macrophages, and some tumor cells were collected. The cells at the gradient interface were washed twice in RPMI-1640. While the lymphocyte preparation can be utilized at this point, a significant proportion of macrophage contam ination remains. Macrophages among the TIL could produce many factors such as prostaglandins, IFN, or other mediators which could greatly affect in vitro analysis of TIL function. To obtain a macrophage-depleted lymphocyte population, the cells were resuspended in 2 ml of carbonyl iron (Sigma Chemical Co., St. Louis, MO) at 5 mg/ml in 10% FCS-containing medium. Macrophages, which phagocytose the carbonyl iron, were hence made heavier compared to the remaining mononuclear cell population. The mixture was then incubated in a 37°shaking water bath for 30 min with manual resuspension every 5 min. The mixture was centrifuged for 10 min at 2000 x g and resuspended in 25 ml of Hanks' balanced salt solution supplemented with 1% Ficoll (Sigma), 1% PCS, 4-(2-hydroxyethyl)-1-piperazineethanesulfonicacid buffer, and antibiotics. After allow ing iron particles to settle for 2 to 3 min, the supernatant was collected. Density gradient cell separation was performed on a Wescor Model 6000 CelSep (Wescor, Inc., Logan, UT). A stock solution of 10% Ficoll (Sigma) was made in Hanks' balanced salt solution (Flow Laboratories, Rockville, MD) supplemented with 1% PCS, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid buffer, and antibiotics. A continuous 2 to 4% gradient was loaded into the bottom port with the CelSep in the tilted ("up") position. A Masterflex roller pump with Cole-Parmer head (Model 701420) (Cole-Parmer, Chicago, IL) and a gradient maker were used. The rate of flow was approximately 50 ml/min, and the total volume was 850 ml. An additional 150 ml of 10% Ficoll cushion were added. The cell suspension in 1% Ficoll was then carefully loaded onto the Ficoll gradient through the top port of the gradient chamber at 5 to 10 ml/min. This was followed by 75 ml of Hanks' balanced salt solution as an overlay also at 5 to 10 ml/min. An equivalent volume of cushion solution was therefore displaced from the separation chamber. The separation chamber was placed in a horizontal position. After incubation for 1 hr, the chamber was then retilted to its starting position, and 50-ml fractions were collected from the bottom port. The fractions were centrifuged at 2000 x g for 10 min and resuspended in 10% FCScontaining medium. Cell number and viability were determined by trypan blue dye exclusion. Nonspecific esterase staining as described by Koski CANCER RESEARCH VOL. 45 JANUARY 1985