Identification of human lymphocyte-derived lymphotoxins with binding and cell-lytic activity on NK-sensitive cell lines in vitro

Supernatants obtained from lectin-restimulated, preactivated, human peripheral blood lymphocytes rapidly released (5-24 hr) high levels of lymphotoxin (LT) activity in vitro. Peripheral blood lymphocytes were preactivated by coculturing with either fetal calf serum or with allogeneic continuous B-cell lines (LCCL) which were treated with mitomycin C. These supernatants contained a population of L-929 cell-lytic LT forms which also selectively bind to the NK-sensitive K-562 cell. However, lytic LT forms for L-929 cells from cPBL and LCCL cultures did not bind to the NK-sensitive MOLT-4 or NK-resistant Raji cells. Additional studies reveal these supernatants contain a second set of LT forms which have cell-binding and cell-lytic activity detectable on MOLT-4 and K-562 cells in a 12 to 18 hr 51Cr-release assay. Cell-lytic form(s) for the MOLT-4 and K-562 cells were not stable for more than a week at -20 degrees C. These findings indicate that materials with LT activity are heterogeneous with respect to their capacity to recognize common and discrete cell-surface components on different types of target cells in vitro.


INTRODUCTION
Lymphotoxins (LT) are cytotostatic and cytotoxic glycoproteins produced in vitro by activated lymphocytes from various animal species including humans.The LT molecules within each animal species examined are heterogeneous with respect to size and charge (1).Immunological and biochemical studies in both the human, murine, and guinea pig LT systems have determined that certain but not all of the various molecular weight LT classes are an interrelated family of molecules (2,3).Further biochemical analysis of the interrelated human LT components has revealed that one of two possible situations exists.Either the various LT forms comprise a subunit system of toxins, or the lower molecular weight (M,) lytic components result from proteolytic cleavage of the higher A4, LT forms (4,5).The smaller il4, forms or classes of LT, (Y, ,8, and y (<90,000 Da) can be dissociated from the alpha heavy (cyn) (140-l 60,000 Da) or complex (Cx) (>200,000 Da) M, classes of LT by changes in ionic strength or chromatography on molecular sieving columns (4,5).The interrelated M, LT classes appear to have different functional capacities.The low M, forms are lytic on select target cells in vitro, i.e., the L-929 cell (a murine fibroblast cell line); however, the larger human and murine A4, classes are unstable but more lytically active on a wide variety of cells (6,7).Additional findings in both the human and murine LT systems indicate that the Cx and also human an forms released by lectin-stimulated immune lymphocytes may be associated with Ig-like component(s) (6)(7)(8).These receptors are not classical immunoglobins and appear to originate from T lymphocytes.
Functional studies with human ffH (7) and Cx (6) LT forms obtained from lectinactivated nonimmune cells indicated that they are lytic in vitro for an NK-sensitive human cell line, K-562 (9,10).Since these lymphocytes were polyclonally stimulated with lectins, the possibility exists that these forms are produced by the subpopulation of lymphocytes which mediates NK.The present studies examine the possibility that populations of human lymphocytes which have been activated in vitro to contain high levels of "NK-like effecters" ( 11,12) produce LT forms which bind selectively to the K-562.In these studies, human lymphocytes are preactivated by coculturing human peripheral blood lymphocytes with fetal calf serum (cPBL) or mitomycin-C treated continuous lines (LCCL) for 5 days in vitro, and then stimulating these preactivated cells with concanavalin A (Con A).The supematants obtained from the preactivated lymphocytes after lectin stimulation were then used to determine if cellbinding forms and cell-lytic forms are the same or if they represent discrete populations of LT molecules.
Lymphocyte cultures and supernatants.Lymphocytes were obtained from defibrinated peripheral blood by the Ficoll-Hypaque technique of Boyum (13).Interface mononuclear cells were carefully removed and washed three times with serum-free RPM1 1640 (RPMI-0%).These cells were preactivated by coculturing in either the presence of FCS (cPBL) or with mitomycin C (Sigma Chemical Co., St. Louis, MO.)treated continuous cell lines.The cPBL cultures were established at a density of 2 X lo6 cells/ml in RPMI-10% and incubated at 37°C.The LCCL cultures were established by coculturing with mitomycin C-treated lymphoblastoid cells at a ratio of 20: 1 and at a density of 2 X lo6 cells/ml in RPMI-10% at 37°C.After 5-7 days, the viable lymphocytes from the cPBL and LCCL cultures were stimulated with lo-20 CLg/ml Con A (Sigma) at a density of 2 X lo6 cells/ml in RPMI-10% for 5, 24, or 48 hr.The supematants were then collected by centrifugation and tested directly or stored at -20°C until use.
Lymphotoxin assays.The details of these methods have been reported previously (14).Briefly, lo5 L cells in 1 .O ml were established as monolayers in screw-capped tubes in RPMI-3% containing 0.5 pg/ml mitomycin C.After a 24-hr incubation at 37"C, serial dilutions of LT-containing or control media were added to duplicate cultures, and they were incubated for 24 hr at 37°C.The cell number was then enumerated on a Model F Coulter counter.A unit of LT activity is expressed as the amount of LT necessary to lyse 50,000 L-cell targets.Units of LT activity in a supernatant are obtained by taking the reciprocal of the dilution of LT necessary to kill 50% of the target L cells.
"O-Release assay.Human cell lines K-562, MOLT-4, and Raji were used as targets in these assays.All reactions were carried out in round-bottom microcytotoxicity plates (Flow Laboratories, Inglewood, Calif.) containing a total volume of 0.12 ml.Radiolabeling of target cells was accomplished by addition of 100 ~1 of "Na chromate, 100 &I (305 mCi/pg, New England Nuclear Corp., Boston, Mass.), to 5 X lo6 cells in 1 ml RPMI-0%.After l-2 hr at 37°C treated cells were washed three times by alternate sedimentation and resuspension (SOOg for 5 min) with cold RPMI-10%.Cells ( lo4 in 0.02 ml) were added to microtiter wells containing 0.05 ml of sample or control media plus 0.05 ml of RPMI-10% and were incubated at 37°C for 18 hr.The release of "Cr label was measured by uptake of cell-free supematant with Titerteks supematant collection system (Flow Laboratories, Inglewood, Calif.) and quantitated in an automated gamma counter (Beckman Instruments, Fullerton, Calif.).The total "Cr releasable was determined by the addition of 0.1 ml of 3% (w/v) sodium dodecyl sulfate solution (Sigma), which represented 90-95% of the total counts.Spontaneous 5'Cr release was l-2%/hr from the target cells.
Target cell destruction was determined using the formula % lysis = experimental release -spontaneous release total release -spontaneous release x 100.
The data are represented as the mean of 3-6 experimental values + the standard deviation.
Formalin jixation of the adsorbing target cell.Cells were treated with 1 ml of 0.1% Formalin in phosphate-buffered saline (PBS), pH 7.2, at 4°C.After 30 min these cells were washed four times by alternate centrifugation and resuspension with 15ml vol of RPMI-0% and recounted on a hemocytometer.
Adsorption assay.Adsorbing cells were washed twice in RPMI-0% and resuspended in test or control supernatants at a concentration of 25-50 X lo6 cells/ml.After 20 min at 4°C the cells were sedimented by centrifugation (800g for 5 min), and absorbed and unabsorbed supernatants were tested for units of cytotoxic activity on L-929 cells or on "Cr-labeled targets.The amount of LT activity in these supematants was determined as described for the LT assay.For some samples, the percentage adsorption of LT activity was calculated using the formula % adsorption of LT activity = (U before adsorption) -(U after adsorption) x loo (U before adsorption) For the "Q-release assay, the percentage adsorption of lytic activity was calculated as % adsorption of lysis = 1 -% "Cr-release for adsorbed sample % 5'Cr-release for unadsorbed sample x 100.

Removal of LT Activity for L-929 Cells by Incubation with K-542 Cells
Data shown in Table 1 indicate that supematants obtained after 5 or 24 hr of lectin stimulation from cPBL or LCCL cultures contain LT activity detected on a Supematants were collected from preactivated lymphoid cells after 5 or 24 hr of lectin stimulation.b Units of LT activity were determined on L-929 cells as described under Materials and Methods.' Supematant samples were incubated with 2.5 X IO' K-562 cells/ml for 20 min at 4°C.The cells were removed by centrifugation and the supematant was tested for LT activity on L-929 cells.The percentage removed by the absorbing cells is calculated as described under Materials and Methods.d ND, not done.
L-929 cells which is removed by incubation with K-562 cells at 4°C.While all of the data are not shown, we found adsorption with K-562 cells measured 10 or more units/ml of LT in 41 out of 54 different supematants.The percentage LT removal in these 4 1 supematants ranged from 12 to 90%.Lymphocytes obtained at various times from a single donor appear to differ in their capacity to release LT forms which are removed by K-562 cells (Table 1, Expts. 3 and 4, 5 hr; Expts.5 and 6, 24 hr).One sample containing 35 U (units/ml of LT activity) was divided into six aliquots and tested in parallel to examine the degree of variation in a single supematant.The values obtained for the levels of activity removed were 16-20 U or 54 f 5%.In addition, an aliquot of this supematant was stored at -20°C and tested after 30 days; 14 U or 64% of the activity was removed out of 22 total units.that removal of the majority of lytic activity occurs in less than 1 min, and the remainder of binding material is removed within 5 to 10 min.

Specificity of Binding
Supematants from cPBL and LCCL cultures were employed to determine whether binding of LT is related to the NK sensitivity of the adsorbing cell.Several additional FIG. 2. Kinetics of removal of LT activity from a supematant by incubation with K-562 cells.The vertical axis measures percentage LT activity removed.A supematant from cPBL was collected after 5 hr of lectin stimulation.Aliquots were incubated with K-562 cells at concentrations of 25 X l66 cells/ml for various intervals of time and tested for LT activity on L929 cells.The removal of LT units was calculated as described under Materials and Methods.The original sample contained 5 1 units of LT/ml.Open circles, untreated K-562 cells; closed circles, formaldehyde-treated K-562 cells.cell lines with varying degrees of sensitivity to human NK effecters were tested for their ability to remove lytic activity from these supernatants.The results of these studies are shown in Table 2. Supernatants adsorbed with MOLT-4 (NK-sensitive) and Raji (NK-resistant) cells exhibited low levels of binding (<lo units of activity), while K-562 removed from 23 to 46 units of lytic activity.There does not appear to be a direct relationship between the levels of L-cell lytic activity bound and the NK sensitivity of these cell lines.

Lysis of Allogeneic Target Cells by Lectin-Induced Supernatants from Preactivated Human Lymphocytes
Supernatants from LCCL and cPBL were obtained after 5 hr lectin stimulation and immediately tested for their ability to induce lysis of L-929 in the standard LT assay and of MOLT-4 and K-562 cells in a "Cr-release assay (Table 3).The LT activity detected on L-929 cells in these supernatants varies from 12 to 72 units/ml.The data presented in Table 1 indicate that five out of seven supernatants from cPBL, when tested on MOLT-4 cells, result in significant levels of lysis (2 l-63% "Cr release) after 18 hr.While not shown, kinetic studies of cell lysis induced by these supernatants revealed that it became detectable at 6-8 hr and increased thereafter.At longer incubation times (greater than 18 hr), increased cell lysis was observed but also a high spontaneous value of 5'Cr release was seen.Three of the seven supernatants tested were also lytically active on K-562 target cells ( 1 l-3 1% "Cr release).The level of lytic activity detectable on MOLT-4 appears to correlate well with the levels of LT activity (lysis of L-929 cells) in these cPBL supernatants.Of those supernatants which contained > 15 U LT activity (Expts.l-7, Table 3) four out of five are lytic for MOLT-4 cells, and two out of five are toxic for K-562 cells.When lower levels of LT activity are obtained (Expts.5 and 6, Table l), MOLT-4 targets are not lysed, but in one experiment, K-562 targets were lysed (Expt.5, Table 3).
The results of similar experiments utilizing 5-hr supernatants from lectin-stimulated LCCL cultures are shown in Table 3; four of five experiments gave significant lytic activity (20-43% "Cr release) on MOLT-4 target cells (Expts.8, 9, 11, and 12).In contrast, only one of five supernatants tested was lytically active on K-562 targets (Expt.9, Table 3; 31% "Cr release).' Supematants from cPBL or LCCLderived PBL were collected after 5 or 24 hr of lectin stimulation and tested for adsorption to various untreated cell lines as described under the LT-adsorption assay.The cell concentrations used for the adsorption studies were 50 X IO6 cells/ml of supematant.
b The stimulating cell line used in the LCCL was WI-L2.a The cPBL and LCCL were cultured in RPMI-10% for 5-7 days and then stimulated with Con A for 5 hr as described under Materials and Methods.Supematants were collected and tested for (1) LT activity detected on L-929 cells and (2) lysis of "Cr-labeled MOLT-4 and K-562 target cells in 18 hr.
'Different supematants were tested in each experiment.Nine different blood donors were used, two donors were used multiple times, one donor was used for Expts.3, 6, and 9; and the other donors were used for Expts. 2 and 8.

Stability of Supernatant Lytic Activity for MOLT-4 Target Cells to Storage
Several supematants were tested on MOLT-4 target cells immediately after production, stored at -2O"C, and retested for lytic activity after various time intervals (Table 4).These studies utilized 5-hr supernatants from both cPBL-and LCCL-treated cells.Six of seven supematants were active on MOLT-4 cells by Day 1; only three y Supematants were generated from cPBL or LCCL-derived PBL (RPM1 1788 stimulator cells) by stimulating with Con A for 5 hr.These were tested for lysis of 5'Cr-labeled MOLT-4 target cells in an 18-hr assay.Separate ahquots were stored at -20°C and tested after various intervals.had detectable levels of lytic activity after storage for 2-7 days.The supematant with the highest level of lytic activity after storage (sample 2, Table 4) was not lytically active on MOLT-4 cells when tested on Day 10.However, additional experiments revealed that freeze-thawing of fresh supematants does not affect lytic activity.
The Relationship between Cell-Lytic Forms for MOLT-4, K-562 Cells, and K-562 Binding LT-Lytic Forms for L-929 Cells These supematants contained LT-lytic forms for L cells which bind to K-562 cells.Experiments were designed to examine the relationship between these binding K-562 forms and lytic forms for MOLT-4 and K-562 cells.Aliquots of each culture supernatant were tested for levels of LT activity and K-562-binding forms on L-929 targets.The results shown in Table 5 demonstrate that all five supematants tested had lytic activity for MOLT-4 and L-929 cells, and only two of five supematants contained significant levels of lytic activity for the K-562 cells (Expts. 1 and 2).Yet supematants 3 and 4 possessed good LT activity and K-562-binding forms, with no apparent lytic effect on K-562 targets.The same supematants were also tested for lytic activity on MOLT-4 and K-562 cells.This observation indicates that LT forms detected on K-562 and L-929 cells may not necessarily be the same materials.
Adsorption of Supernatants with K-562, MOLT-I, and Raji Cell Lines Removes Lytic Activity for MOLT-4 and K-562 Cells Supematants from 5-hr Con A-stimulated cPBL and LCCL cultures were adsorbed with MOLT-4, K-562, or Raji and then tested for lytic activity on MOLT-4 cells in an 18-hr "Cr-release assay.The levels of lysis of MOLT-4 cells by the adsorbed aliquots of each sample were compared to the level of lysis by the unadsorbed sample.The results of these adsorption studies are shown in Table 6.Four of the six supernatants tested had lytic forms which were equally removed by adsorption with any LT activity bound to K-562 cells" ' Supematants were generated from cPBL or LCCL-derived PBL by stimulation with Con A for 5 hr.Each supematant was tested for (1) lysis of 5'Cr-labe1ed MOLT-4 target cells in 18 hr (2) lysis of 5'Crlabeled K-562 target cells in 18 hr, and (3) LT activity against L-929 cells.In addition, separate aliquots were adsorbed on K-562 cells at 50 X 106 cells/ml of supematant, and then tested on L-929 cells for LT activity.Cells used for adsorption in Expts.12 hr, and levels of cytotoxicity climb higher after 18 hr.The degree of lysis may actually continue to increase after this time interval, but later time points were not examined because of the high levels of spontaneous "Cr release from the target cells after 18 hr.An examination of the stability of supernatant lytic activity from MOLT-4 and K-562 cells revealed that, for most supernatants, these lytic materials are not stable to storage at -20°C for more than a few weeks.However, L-929 cell-lytic activity in these same supernatants can be recovered after storage under the same conditions for very long periods.
Experiments were designed to determine if the L-929 cell-lytic forms which bind to K-562 cells are related to the lytic activity detectable on "Cr-labeled MOLT-4 or K-562 target cells.The results of these studies indicate that there is no correlation between the presence in a supernatant of MOLT-4 or K-562 cell-lytic LT activity and the presence of K-562-binding forms of LT detected on L-929 cells.These supernatants clearly contain multiple LT forms with different cell-binding and cell-lytic capacity.Clearly, K-562 and, to a lesser extent, MOLT-4 express receptors for a certain population of forms which lyse the L-929 cell.In contrast, the Raji does not express any receptor(s) for L-929 cell-lytic forms.However, MOLT-4, K-562, and Raji all possess receptors for LT forms which have cell-lytic activity in descending order on the same cells.These studies support the concept that LT forms are heterogeneous, both in their binding and in cell-lytic capacity.It should be noted that a low level of L-929 and MOLT-4 cell-lytic activity did not bind to any target.
The cPBL-and the LCCL-prestimulated lymphocytes have been shown by in vitro techniques to contain several types of nonspecific, "NK-like," and anomalous killer effector cells ( 11,12,17); the LCCL also generates T lymphocytes which are immune to the allogeneic stimulator cells (18).The present studies employ a short secondary polyclonal lectin stimulus which induces different populations of the preactivated lymphocytes to rapidly release LT from preformed pools into the supernatant (15).The presence of LT forms with different cell-binding and lytic activity in these supernatants raises the possibility they may be of polyclonal origin, and studies with defined lines of different types of effector cell are underway in attempts to answer this interesting question.These experiments reveal that supematant-induced lysis of MOLT-4 cells, as shown from lysis of L-929 cells, can now also be divided into at least two steps: (a) a cell-binding phase and (b) the actual process of cell lysis.Moreover, these studies also indicate that a cell may be resistant to lysis, but still express receptors for these cell toxins.It is tempting to speculate that the reason why some supernatants have high cell-lytic activity and others low activity, is that binding forms may be more lytically active than nonbinding forms.It is important to note here, as mentioned previously, that no single target cell, for example, the L-929 cell, is uniformly sensitive to all LT forms, and different targets are essential to detect different forms of LT activity ( 19).
FIG. 1. Removal of LT activity from a supematant by incubation with increasing numbers of K-562 cells.A supematant from cPBL was collected after 5 hr of lectin stimulation, One-milliliter aliquots were incubated with various numbers of K-562 for 20 min at 4°C and then tested for LT activity on L929 cells.The percentage removal of LT activity was calculated as described under Materials and Methods.The original sample contained 35 units of LT/ml.Open circles, untreated K-562 cells; closed circles, formaldehyde-treated K-562 cells.The error bars indicate the range of values obtained for four separate aliquots.

TABLE 1
Removal of LT Activity in Supematants from Lectin-Stimulated cPBL and LCCL by Incubation with K-562 Cells

TABLE 2
Adsorption of Supernatants from cPBL and LCCL-Derived PBL with Cell Lines with DifferentSensitivities to Human NK Effector Cells"

TABLE 3
Capacity of Supematants from Lectin-Stimulated cPBL and LCCL to Cause Lysis of Various Target Cell Lines"

TABLE 4
Stability of Supematant Lytic Activity for MOLT-4 Target Cells to Storage at -20°C"

TABLE 5 The
Relationship of Supematant Lytic Activity for MOLT-4 and K-562 Target Cells to the Presence of LT Forms That Bind to K-562 Cells"