[CANCER RESEARCH 50. 7153-7158. November 15. 1990]

In Vitro Migration of Lymphocytes through Collagen Matrix: Arrested Locomotionin Tumor-infiltrating Lymphocytes1

Kelli G. Applegate,2 Charles M. Balch, and Neal R. Pellis3

Department of General Surgery; The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030

ABSTRACT

Antitumor immunity requires (a) extravasation of lymphocytes fromthe blood stream to interstitium, (b) locomotion through extracellularmatrix to the site of the tumor, (<•)effector cell recognition of the tumortarget with cell/cell contact and binding of adhesion receptors, (d) T-cellreceptor binding to histocompatibility and tumor antigens, and (<•)tumor

cell lysis. We hypothesize that the tumor microenvironment inhibitslymphocyte locomotion through extracellular matrix as one mechanismby which tumors may avert host defense. Lymphocyte locomotion wasinvestigated ¡nvitro using a three-dimensional collagen gel model. Freshtumor-infiltrating lymphocytes (TIL) were obtained by enzymatic digestion of melanomas and renal cell carcinoma, and mononuclear cells wereisolated by discontinuous Ficoll-Hypaque gradient. The lymphocyteswere analyzed for motility from a point of origin between basal andoverlay layers of collagen gel. Results showed that TIL migration wasalmost completely inhibited, compared with migration of normal andcancer patient peripheral blood leukocytes and lymphocytes from lymphnodes. Short-term (24-h) exposure of lymphocytes to cytokines duringthe assay in the collagen gel matrix had no effect on locomotor ability.Long-term (19, 30, or 35 days) culture of TIL in 200 units/ml ofinterleukin 2 reinstated locomotor ability. Short-term exposure of any ofthe lymphocyte populations to interleukin I -«.interleukin 1-/3,interleukin2, interleukin 3, interleukin 4, o-interferon, or •v-interferonhad no effect

on migration. Thus, TIL display a uniquely arrested ability to locomotethrough collagen gel. Inhibition of the locomotion of infiltrating effectorcells is possibly a mechanism by which the tumor evades the host immunesystem.

INTRODUCTION

Host immunity to cancer occurs within two compartments:(a) the vascular system and lymphatics and (¿>)the ECM4 of

the tumor microenvironment. For effective host immunity, T-cells home to the site of the tumor, extravasate from thebloodstream to the interstitium, and locomote through ECMin the tumor microenvironment. Subsequently, ancillary adhesion peptides on the effector cells bind to ligands on target cells,and binding of the T-cell receptor to histocompatibility andtumor antigens occurs (1-3). Thus, the ability of lymphocytesto migrate and accumulate at various sites in the body is acritical step in the effector phase of cellular immunity.

Trafficking of lymphocytes through lymphoid organs, lymph,blood, and nonlymphoid organs is nonrandom, as determinedusing fluorescein isothiocyanate-labeled or radiolabeled sensi-

Received 4/16/90; accepted 8/17/90.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' This work was supported by the Gillson-Longenbaugh Foundation.2Surgical oncology fellow supported by National Cancer Institute Training

Grant CA09599-02.3To whom requests for reprints should be addressed, at Department of General

Surgery. The University of Texas M. D. Anderson Cancer Center. 1515 HolcombeBlvd., Box 174, Houston, TX 77030.

4The abbreviations used are: ECM, extracellular matrix: HEV. high endothe-lial venules; TIL. tumor-infiltrating lymphocytes; PBL. peripheral blood leukocytes: LNL, lymph node lymphocytes; IL-2, interleukin 2; HBSS. Hanks' balanced

salt solution; FBS, fetal bovine serum; Ml,,, interleukin In; Ml.;, interleukin1/3; IL-3, interleukin 3; IL-4, interleukin 4; IFN-o, a-interferon; IFN-i, y-interferon: FITC, fluorescein isothiocyanate; IgG, immunoglobulin G.

tized and nonsensitized lymphocytes (4-6). There are threesubpopulations of lymphocytes: (a) cells with preference for theintestinal lymphoid tissues; (b) cells directed to subcutaneouslymph nodes; and (c) cells that migrate preferentially throughthe skin or other peripheral nonlymphoid tissues. This trafficking is directed by specific lymphocyte surface receptors thatbind to HEV, allowing access from the blood stream to lymphnodes (7-9). Despite the proposition that lymphocytes distribution is controlled through cell surface expression of adhesionmolecules, little is known about regulation of lymphocyte locomotion after the cell leaves the confines of vascular or lymphendothelia and enters ECM at the site of inflammation ortumor. Indeed, the impaired movement of lymphocytes throughECM may contribute to failed immunity at the tumor site.

Addressing the local tumor immune response has been difficult owing to the complex nature of immune cell accumulationat the site of the tumor, the lack of correlation with systemicimmune performance, and the difficulty in developing isolatedin vivo models. We hypothesize that the neoplastic site isunfavorable to lymphocyte locomotion through ECM. To testthis hypothesis a three-dimensional in vitro collagen model wasadapted to assess lymphocyte locomotion. The purpose of thisinvestigation was 2-fold: to determine the effect of microenvironment on the locomotor ability of TIL as compared withcancer patient PBL, normal PBL, and LNL; and to determinea possible biological mechanism of lymphocyte locomotionthrough interstitium by analyzing the effect of lymphokines onin vitro lymphocyte locomotion.

Freshly isolated TIL from melanomas and renal cell carcinoma were analyzed for migration in collagen, either fresh orafter culture in the presence of IL-2 and tumor-associatedantigen. Freshly isolated TIL were significantly inhibited intheir ability to migrate, as compared with PBL and LNL.However, TIL stimulated in culture with IL-2 and antigenmigrated similarly to PBL. When biologically active amountsof cytokine were added directly into the medium of the assay,there was no effect on the distance of random migration ofTIL, TIL cultured in IL-2, PBL, or LNL. In conclusion, TIL

are unique among lymphocytes in their abated migratory ability.Inhibition is not constitutive, since long-term activation by IL-2 and antigen restores the migratory activity but short-termexposure does not improve migration.

MATERIALS AND METHODS

Lymphocytes. Buffy coats from healthy donors were purchased fromthe Regional Gulf Coast Blood Center (Houston, TX). PBL wereisolated using Ficoll-Hypaque gradients by centrifugation at 850 to1000 x g for 30 min and then washed with HBSS 3 times. TIL wereobtained from seven tumor specimens removed from six patients withmetastatic melanoma and one patient with metastatic renal cell carcinoma. Five of the specimens had an adequate number of TIL for theassay. LNL were obtained from three fresh pathological specimens.Two of the lymph nodes were infiltrated with melanoma, and one nodewas uninvolved with tumor in a patient with two of 11 axillary nodespositive for infiltrating melanoma. The tumor or lymph node tissue

7153

Research. on November 17, 2020. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

TIL LOCOMOTION THROUGH COLLAGEN

was minced sterilely and enzymatically digested with 2 mg/ml ofcollagenase D, 0.4 mg/ml of DNase I, and 0.4 mg/ml of hyaluronidasetype V for 1 to 1.5 h in HBSS. All enzymes were obtained from SigmaChemical Co. (St. Louis, MO). The cells in the effluent were thenseparated using Ficoll-Hypaque gradients to obtain the mononuclearcells. Three TIL specimens were cultured with 200 units/ml of IL-2 inAIM-V serum-free medium, or RPMI 1640 medium, with 10% FBS,at 37'C in an atmosphere of 7% carbon dioxide and 93% air for 19,

30, or 35 days in the presence of tumor-associated antigen. Additionally,TXM 141 TIL were cultured in 200 units/ml of IL-3 and in 200 units/ml of IL-2 plus 10 units/ml of IL-4. Limited availability of TILprecluded culturing them in other cytokines.

Cryopreserved PBL from two patients with metastatic melanoma(TXM 56 and TXM 81) were utilized to assess their ability to migratethrough collagen matrix because of limited availability of cancer patientperipheral blood specimens. There were sufficient cells to perform oneassay with TXM 81 and four assays with TXM 56. To ensure thatcryopreservation did not affect lymphocyte locomotor ability, normaldonor PBL were divided into two aliquots, one cryopreserved for 72 hand the other held at 4'C until the assay was performed. The mono-

nuclear cells were collected using Ficoll-Hypaque gradients and cryopreserved at 10" cells/ml in 10% dimethyl sulfoxide and 90% FBS at—70°C.The cells were placed directly at -70°Cfrom room temperature,

and thawing was performed at room temperature. Eighty-five % ofcryopreserved cells were viable as determined by trypan blue exclusion.The cancer patient PBL and normal PBL were assessed for migratoryability in the collagen assay.

Cytokines. The recombinant IL-2 (1.5 x IO6 colony-forming units/mg of protein) was obtained from Hoffmann-LaRoche (Nutley, NJ).The other human recombinant lymphokines were purchased from Gen-zyme (Boston, MA). The specific activity of human recombinant IL-1«was 1000 units/ml (Lot 08731), of IL-lfi was 1000 units/ml (Lots07731 and 06634), of IL-3 was 10* colony-forming units of protein(Lots 04733 and B9263), and of IL-4 was 10* proliferation units/mg

of protein. Human IFN-«and IFN-7 were purchased from Hoffmann-LaRoche. The activity of IFN-a was IO5 units/ml (Lot 3102) and ofIFN-7 was IO5units/ml (Lot N9207AX).

Collagen Gel Assay. Type I rat tail collagen was used to investigatelymphocyte motility in a three-dimensional collagen gel model previously described by Ratner et al. (10). An adaptation of their model wasused in our investigation (Fig. 1). Briefly, 1 ml of cold liquid collagensolution in lOx RPMI 1640 medium was transferred to each 25-mm-diameter tissue culture well and then incubated for 15 min at 37°Cto

polymerize the collagen. Lymphocytes were mixed with cold collagensolution at 2.5 x IO7 cells/100 p\. Twenty ^1 of cell suspension were

transferred to the base layer of collagen gel and incubated IO min toform a 10-mm disk. An overlay of polymerized 3-mm-thick collagenwas precast in 18-mm-diameter culture wells and then placed over thelymphocyte layer. Finally, 0.8 ml of collagen solution provided a sealinglayer. One ml of RPMI 1640 medium with 10% FBS was placed overthe gel, either alone or supplemented with biotherapeutic amounts ofcytokine (i.e., 200 units/ml of IL-2, 200 units/ml of IL-3, 200 units/ml of IL-4, 250 units/ml of IFN-«, or 1000 units/ml of IFN-7).Triplicate or quadruplicate assays were incubated for 22 to 28 h,allowing the lymphocytes to locomote in an omnidirectional array. Theoverlay and basal gel layers were then separated with forceps, fixed in1% formaldehyde solution for 24 h, and stored at room temperature inphosphate-buffered saline containing 0.1% sodium azide. The vertical"leading front" migration distance in micrometers was measured from

the center of the cell disk. The leading front distance is the point atwhich three of the most distantly migrating lymphocytes are simultaneously in focus in one high-power field.

Analysis of Lymphocyte Surface Markers. FITC-conjugated reagents,including anti-CD3, -CDS, and -CD 16, and phycoerythrin-conjugatedreagents, including anti-CD3, -CD4, and -CDS, were purchased fromBeckton Dickinson Co. (Mountain View, CA). FITC- or phycoerythrin-conjugated irrelevant mouse monoclonal IgGl or IgG2a (BecktonDickinson Co.) was used as control monoclonal antibody for two-coloranalysis using flow cytometry. Two of the fresh TIL samples and two

OVERLAY\\SEALINCLAYER/Z/LYMPHOCYTES\

BASA

LAYEI

NON-LOCOMOTORY

LYMPHOCYTES

LOCOMOTORYLYMPHOCYTES

Fig. 1. Collagen gel model. A, A 12-mm tissue culture well is shown in cross-section. The basal layer of type I rat tail collagen was cast first and allowed topolymerize. Lymphocytes were suspended in cold collagen solution at a densityof 2.5 x 107cells/0.1 ml, and 20 ¡n\were transferred to the basal layer. An overlaylayer of collagen was precast in an 18-mm-diameter culture well and transferredto the basal layer overlying the lymphocytes. A scaling layer was polymerized tohold the overlay in place. One ml of medium containing lO^c FBS was placedover the assay and readily diffused through the collagen gel. B, After 24 h ofincubation, migratory lymphocytes had moved into the overlay and basal gellayers. These layers could then be easily separated with forceps. Leading frontmigration distance was measured using an inverted light microscope.

TIL cultured in IL-2 for 15 or 20 days were incubated with monoclonalantibodies at 4°Cfor 30 min, and surface marker expression was

analyzed by flow cytometry (EPICS C; Coulter, Hialeah, FL) as previously described (11).

Data Analysis. The mean migration distance and standard error ofthe mean were calculated, and the means were compared betweengroups, using analysis of variance and the Student-Newman-Keuls test.

RESULTS

Locomotor Ability of Normal and Cancer Patient PBL. TXM56 and TXM 81 PBL obtained from melanoma patients andnormal PBL were assessed for migration in collagen. Theresults after 24-h incubation in the collagen assay are shown inFig. 2. There was no statistically significant difference in thedistance migrated by normal PBL held at 4°C,cryopreserved

normal PBL, or cryopreserved cancer patient PBL. Moreover,there was no statistically significant difference in the meanmigration distance of this cryopreserved PBL population compared with 14 other normal donors tested (Table 1). Thus, thePBL of cancer patients migrate at an apparently normal rate.

Locomotor Ability of LNL. Freshly isolated LNL were obtained from pathological specimens from three patients withmetastatic melanoma. The 24-h mean migration through collagen, as assessed by leading front distance into the overlaylayer of collagen, was 796 ±102 ^m and into the basal layerwas 790 ±44 ¿tm.These values are not statistically significantlydifferent from normal PBL migration (Fig. 3). Thus, lymphocytes from two distinct lymphoid compartments randomly migrate in collagen in vitro without prior stimulation with mitogenor lymphokine. Furthermore, enzymatical digestion of thelymph nodes did not alter the ability of the lymphocytes tomigrate.

Phenotypes of TIL. The vast majority of TIL in freshlyisolated samples were T-cells (CD3+). RC6 TIL, from a renalcell carcinoma, were 76.5% CD3\ 72% CD4+, 41% CD8+, 8%

7154

Research. on November 17, 2020. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

TIL LOCOMOTION THROUGH COLLAGEN

TXM 81 PBL

TXM 56 PBL

FROZEN NL PBL

NORMAL PBL

400 800 1200

MIGRATION DISTANCE (pm)

Fig. 2. Migration of cancer patient PBL in collagen. TXM 81 and TXM 56are PBL specimens from two patients with metastatic melanoma. These cellswere available cryopreserved in our laboratory. An aliquot of normal PBL wascryopreserved for comparison. There was no statistically significant differencebetween migration of normal PBL, cryopreserved PBL, or cancer patient PBL.There were only enough TXM 81 cells to perform one assay; four assays wereperformed with TXM 56. NL. normal. Columns, mean; bars, SEM.

Table I Migration of lymphocyte populations in the presence ofcytokinesThe migration distances (f<m)for normal PBL, TIL, and TIL cultured in IL-2

in the presence of various cytokines are given.

Migration distance I/MM)

Cytokine Direction Normal PBL TIL Cultured TIL

NoneIL-

loIL-

IßIL-2IL-3IL-4IFN-nIFN--XOver"BasalOverBasalOverBasalOverBasalOverBasalOverBasalOverBasalOverBasal80476366774669874282979178978311174770083087*(14)c93

(14)62(2)38(2)53

(2)59(2)64(12)92(12)74(12)84(12)118(12)106(12)34(1)69(1)663

±28(1)835±36(1)271

±45(5)''276±66(4)"483

±17(1)597±32(1)533±12(1)475±25(1)332±53(4)429±71(3)197±52(3)364±74(3)320±41(4)438±54(3)423±87(1)545

±28(1)493±7(1)477

±39(1)790

±36(3)791±28(3)966

±52(2)907±41(2)946±37(2)917±23(2)1007+ 41(2)893±56 (2)

" Migration into the overlay or basal layer of collagen.* Mean ±SEM.c Numbers in parentheses, number of replicates in each experimental group.*P< 0.001 compared with normal PBL.

Leul6+, and 12% LeuM3+. There were no detectable CD16"

cells present. M9 TIL, from a metastatic melanoma, were 85%CD3+, 44% CD4+, 30% CD8+, and 0% CD16+. After 20 daysof culture in IL-2, M9 TIL were 99% CD3+, 68% CD4+, 30%CD8+, <2% CD16+, 0% Leul6+, and 0% LeuM3+. Based onthe results of two-color analysis, the majority of CD4+ T-cellswere CD4+ CDS', and the majority of CD8+ were CD4~ CD8+.5

TXM 141 TIL, from a metastatic melanoma, were pheno-typed in our laboratory after culture for 15 days in IL-2. TheseTIL were 79% CD3+, 14.3% CD4+, 50.5% CD8+, and 5.2%CD16+. The percentage of false-positive staining on TIL with

control monoclonal antibodies was less than or equal to 1%.Locomotor Ability of TIL. In contrast to the locomotion of

all other lymphoid populations, TIL locomotion through collagen matrix was significantly decreased by more than 75%compared with that of normal PBL, LNL, and cancer patient

5K. Itoh, personal communication.

NLPBL

LNL

200 400 600 800 1000

MIGRATION DISTANCE (urn)Fig. 3. LNL migration in collagen. LNL were obtained from pathological

specimens. There was no significant difference in the distance migrated by thesecells compared with normal PBL. LNL migration data from a representativepatient are shown. Columns, mean; bars, SEM.

NLPBL

TIL TXM 141

TIL RC6

TILM9

200 400 600 800

MIGRATION DISTANCE (.urn)

1000

Fig. 4. TIL versus normal (NL) PBL migration in collagen. Three representative TIL (TIL TXM 141, TIL RC6, and TIL M9) are shown. TIL migrationwas statistically significantly decreased compared with normal PBL. *, statisticallysignificant difference. Columns, mean; bars, SEM.

PBL (P < 0.001) (Fig. 4). It is unlikely that the inhibited TILmigration is due to the enzymatic dissociation procedure because LNL obtained by enzymatic digestion migrated normally.Furthermore, a melanoma TIL specimen was 80% viable after24-h incubation in collagen. Thus their inability to migrate isprobably not due to TIL cell death, and lymphocytes from thetumor microenvironment are functionally different from lymphocytes deployed systemically.

Locomotor Ability of TIL Cultured in Cytokine. To determineif the inhibition of TIL locomotion is an intrinsic characteristicof this cell population, we next analyzed whether stimulationof TIL in culture would reinstate their locomotor ability.Freshly isolated TIL from three patients with metastatic melanoma were cultured with 200 units/ml of IL-2 for 19, 30, or 35days, the same time period used to culture TIL for immuno-therapy. An attempt was made to culture the TIL in IL-3, butthe cells were not viable. TIL cultured in 200 units/ml of IL-2plus 10 units/ml of IL-4 for 15 days had as many viable cellsas did cells cultured with IL-2 alone. The viable cultured TILwere assessed for migration in the collagen gel assay. After 24-h incubation, there was a significant increase in the leading

7155

Research. on November 17, 2020. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

TIL LOCOMOTION THROUGH COLLAGEN

front migration distance of the stimulated TIL compared withfresh TIL (P < 0.001). Furthermore, the migration of stimulated TIL was comparable to that of normal PBL or LNL (Fig.5). There was no difference between locomotor ability of cellscultured in IL-2 plus IL-4 versus IL-2 alone. Moreover, there

was no significant difference in the locomotor ability of cellscultured in serum-free medium or medium containing FBS.Thus, TIL recover locomotor ability when activated with IL-2in long-term culture in the presence of tumor-associated anti

gen.Effect of Cytokine on Lymphocyte Locomotion. The recovery

of locomotor activity in TIL cultured with IL-2 suggested that(a) IL-2 modulates cell surface structures that interact with thecollagen, (b) proliferation induced by IL-2 favors migratorylymphocytes, thereby selecting the most motile cells, or (c)proliferating cells are more motile than nonproliferating cells.The prospect that modulation of cell surface structures resultsin improved migration was tested by introducing cytokinesdirectly to the assay medium. IL-la, IL-1/3, IL-2, IL-3, IL-4,IFN-a and IFN--y were individually included in the medium of

the assay. The assay was incubated for 24 h, and leading frontmigration into the overlay and into the basal layer was assessed.The migration distances of fresh TIL, TIL stimulated in culturewith IL-2 for 19 days, and normal PBL were not significantlydifferent with this short-term exposure to cytokine (Fig. 6).

Table 1 shows the mean values and standard errors for allexperiments performed. For two of the TIL specimens, therewere only sufficient numbers of cells to perform the assay inquadruplicate without cytokine in the medium. When the migration distances for all TIL were averaged, it appears thatthere is a significant difference in TIL migration with cytokinein the medium. However, in individual experiments the difference was not significant when the assays containing cytokinewere compared with the controls for each experiment. Also, inone experiment the basal layer was damaged, so the number ofreplicates for the migration into the overlay in the no cytokine,IL-2, and IL-4 groups was larger. Only one TIL specimen

NL PBL

CULTURED TILIN IL-2

TIL0 BASE

• OVERLAY

200 400 600 800 1000

MIGRATION DISTANCE (urn)Fig. 5. Migration of fresh and IL-2-cultured TXM 141 TIL. TXM 141 TIL

were cultured in 200 units/ml of IL-2 for 19 days and then placed in the collagenassay. Locomotor ability of the TIL was statistically significantly improved afterculture in IL-2 and was comparable to the migration of normal PBL. NL, normal.Columns, mean; bars, SEM.

400 800 1200 0 400

MIGRATION DISTANCE (urn)

800 1200

IL-4

IL-3

0

IL-2

CONTROL

0 400 800 1200

MIGRATION DISTANCE (¡¡m)

Fig. 6. Effect of cytokine on lymphocyte locomotion. Biotherapeutically important cytokines (IL-ln, IL-lff. IL-2, IL-3. IL-4, IFN-u. and IFN--y) were addedto the medium of the collagen assay to determine the short-term effect of cytokineon locomotion of normal PBL (A), TIL (TXM 141) (A), and TIL cultured inIL-2 (TXM 141) (C). Migration distances for one representative from each of thethree cell populations are shown. There was no significant effect of cytokine onmigration distance during the 24-h exposure in the assay for any of the lymphocytepopulations tested. Columns, mean; bars. SEM.

contained a sufficient number of cells to perform the assay withall seven of the cytokines tested.

DISCUSSION

Impaired TIL locomotor capability may be a mechanism ofinhibition of host defense against tumor. TIL isolated fromhuman melanomas and renal cell carcinoma displayed retardedmigration through collagen in vitro compared with normaldonor PBL, LNL, and cancer patient PBL, which migratedsimilarly to normal PBL. The in vitro collagen matrix has beenpreviously used to analyze leukocyte and tumor cell migration(10, 12-16). Collagen fibers are a substantial component of theECM that lymphocytes must traverse to reach tumor. Polymerized collagen provides a three-dimensional and physiologicalenvironment for assessment of cell migration in vitro (12, 17).In contrast to two-dimensional migration on modified plasticsurfaces (18-22), the three-dimensional assay may provide anapproximation of the tissue environment of the tumor. Gelledcollagen is transparent; therefore, migration may be assessedby light microscopy (21). Because of its physical and chemicalsimilarity to in vivo interstitium (23), the gel is compatible withfreshly isolated or cultured lymphocytes. Schor et al. (16)showed that human PBL spontaneously and reproducibly migrate through collagen matrix in a random omnidirectionalpattern. Murine LNL migrate randomly (10, 15, 21), andlocomotion varies with the cell cycle. Migration distance increases in lymphocytes stimulated with concanavalin A (10).Murine splenic lymphocytes cultured 72 to 80 h with recombinant IL-2 have twice the number of motile cells, and they

7156

Research. on November 17, 2020. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

TIL LOCOMOTION THROUGH COLLAGEN

migrate more than twice the distance of uncultured cells (15).In this investigation, culturing human TIL for 19 to 35 daysreinstated migratory ability. The inhibited migration in freshlyisolated TIL suggests that either a suppressor cell or a solublesubstance is present in the tumor microenvironment that mayaffect effector/target cell contact. We believe that the ability ofTIL to migrate through matrix is a critical step in in vivoeffector function.

The immune response to tumor involves a series of lymphocyte actions, including lymphocyte trafficking, homing to thesite of the tumor, extravasation from the bloodstream intointerstitium, migration through ECM, binding of effector celladhesion receptors to target cell surface ligands, activation ofeffector cell via T-cell receptor binding to tumor and histocom-patibility antigens, and tumor cell lysis. Lymphocytes traffic ina nonrandom manner (4-6). Each subpopulation of lymphocytes recirculates by continuous passage from blood throughlymphoid organs and lymph and back to blood. The process oflymphocyte homing is regulated at many levels, and the firstlevel of control is via recognition of, binding to, and extravasation through HEV in peripheral lymph nodes, mucosa of theintestine, or sites of chronic inflammation (24). Binding toHEV is organ specific and dependent on specific homing receptors, such as the homing-associated cell adhesion molecule, theperipheral lymph node homing receptor, or lymphocyte function-associated molecule 1, on the cell surface that recognizemucosal vascular addressin, peripheral lymph node addressin,or both, respectively (25). When a repertoire of clonal lymphocyte specificities circulate, tumor antigen is recognized by aparticular clone with subsequent expansion ofthat clone, whichis antigen dependent (2). It is at this level that matrix may serveas an obstacle to interaction of lymphocytes with tumor.

Once extravasation occurs, it is possible that adhesion receptors such as the very late activation integrin supergene familyof molecules that bind to components of ECM (i.e., collagen,fibronectin, and laminin) are involved in lymphocyte locomotion, but as yet, preliminary data using antibodies to a- and ß-chains suggest that very late activation molecules may not berequired for lymphocyte locomotion.6 One mechanism by which

tumor may avert the host immune system is to inhibit lymphocyte migration, thus preventing the initial binding of effectorto tumor target. Gunji and Gorelik (26) showed that fibrincoagulation induced by murine tumor cells protected them fromdestruction by cytotoxic cells by producing a physical barrier toprevent effector/target contact. In this investigation, we showedthat TIL are markedly inhibited in their ability to locomotethrough collagen in vitro, which is a substantial component ofECM. Since tumor cells are present with the TIL after processing, it is possible that the tumor is suppressing TIL locomotionby its presence in the assay, instead of the prior exposure of theTIL to the tumor causing the suppressive effect.

Analysis of the effect of stimulating TIL on their migrationdetermined that activation of human lymphocytes in culturewith IL-2 improves migration through collagen, as was shownin a murine model (15). Freshly isolated TIL are inactive andhave limited cytotoxic capabilities (27, 28). Long-term (19 to35-day) culture of melanoma TIL in recombinant IL-2 resultsin proliferation of predominantly CDS* T-cells (27-29), which

are specifically cytotoxic for autologous fresh tumor targets. Inour investigation one TIL culture was predominantly CD4+ T-cells, and another was predominantly CDS* T-cells. TIL prop-

' Unpublished data.

agated in culture for this length of time are presently used inimmunotherapy trials (27, 30, 31). To function in the eradication of tumor, cytotoxic cells must be able to home to the tumorsite and then extravasate from the bloodstream into interstitiumand traverse ECM. Studies, although limited, involving homingof these infused lymphocytes have shown that the cells havesome predilection to accumulate at the tumor site (32). Themechanism of human lymphocyte locomotion through ECMfor achievement of effector/target cell contact is unknown.Freshly isolated TIL did not migrate well through collagen.After long-term culture in IL-2, migration was restored. Short-term (24-h) exposure of the cells in the assay to biologicallyimportant amounts of IL-1«,lL-2ß,IL-2, IL-3, IL-4, IFN-a,or IFN-7 did not significantly affect migration of TIL, TILpreviously activated in culture with IL-2, PBL, or LNL. Therefore, it is unlikely that the improved locomotor ability of TILcultured in IL-2 is due to changes in lymphocyte cell surfacemolecule expression that may occur during short-term exposureto cytokine or to the chemokinetic effect of the cytokine. Theimproved locomotor ability of cultured TIL may be due in partto an increase in the number of proliferating cells, because PBLtreated with mitomycin C, an inhibitor of proliferation, did notmigrate as well as untreated cells (data not shown).

In conclusion, the tumor microenvironment has severaldown-regulatory effects on the infiltrating immune cells, andour study shows that one mechanism may be inhibition oflymphocyte locomotion through ECM. In immunotherapymodels, simultaneous i.v. infusion of lymphokines with stimulated TIL may have no immediate effect on the infiltratingability of the lymphocytes, whether they are TIL already at thesite of the tumor or ex v/vo-stimulated TIL used in adoptiveimmunotherapy. Future immunotherapeutic strategies againstcancer may address in vivo stimulation of lymphocyte locomotion to enhance host immunity.

ACKNOWLEDGMENTS

We thank Dr. Stuart Ratner of the Michigan Cancer Foundation,Detroit, for his counsel in the performance of the migration assay, andDr. Marie Salmerón, Dr. Paul Mansfield, Dr. Merrick Ross, Dr. KyogoItoh, and Dr. Ronald Merrell for their contributions of cells or tumortissue.

REFERENCES

1. Haynes, B. F., Telen. M. J.. Hale. L. P., and Denning, S. M. CD44—amolecule involved in leukocyte adherence and T-cell activation. Immunol.Today. 10: 423-428, 1989.

2. Makgoba. M. W.. Sanders, M. E.. and Shaw. S. The CD2-LFA-3 and LFA-1-ICAM pathways: revelance to T-cell recognition. Immunol. Today. 10:417-422. 1989.

3. Denning, S. M., Le. P. T., Singer. K. H.. and Haynes, B. F. Antibodiesagainst the CD44 p80. lymphocyte homing receptor molecule augmenthuman peripheral blood T-cell activation. J. Immunol., ¡44:7-15, 1990.

4. Chin. G. W.. and Hay. J. B. A comparison of lymphocyte migration throughintestinal lymph nodes, subcutaneous lymph nodes, and chronic inflammatory sites of sheep. Gastroenterology, 79: 1231-1242. 1980.

5. Issekutz. T. B.. Chin. G. W.. and Hay. J. B. Lymphocyte traffic throughgranulomas: differences in the recovery of indium-111 labeled lymphocytesin afferent and efferent lymph. Cell. Immunol.. 54: 79-86. 1980.

6. Issekutz. T. B.. Chin. G. W.. and Hay. J. B. Lymphocyte traffic throughchronic inflammatory lesions: differential migration versus differential retention. Clin. Exp. Immunol.. 45: 604-614. 1981.

7. Chin. Y. H.. Carey, G. D.. and Woodruff. J. F. Lymphocyte recognition oflymph node high endothclium. IV. Cell surface structure mediating entryinto lymph nodes. J. Immunol.. 129: 1911-1915. 1982.

8. ( ..i!Iain ni. W. M., Weissman. I. L., and Butcher. E. C. A cell-surface moleculeinvolved in organ-specific homing of lymphocytes. Nature (Lond.). 304:30-34. 1983.

9. Jung. T. M., Gallatin. W. M.. Weissman. I. L., and Dailey. M. O. Down-

7157

Research. on November 17, 2020. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

TIL LOCOMOTION THROUGH COLLAGEN

regulation of homing receptors after T-cell activation. J. Immunol., 141:4110-4117, 1988.

10. Ratner. S., Jasti, R. K., and Heppner, G. H. Motility of murine lymphocytesduring transit through cell cycle. Analysis by a new in vitro assay. J. Immunol..140: 583-588, 1988.

11. Itoh, K.. Platsoucas, C. D., and Baldi. C. M. Autologous tumor-specificcytotoxic T lymphocytes in the infiltrate of human metastatic melanomas.Activation by interleukin-2 and autologous tumor cells, and involvement ofthe T-cell receptor. J. Exp. Med., 168: 1419-1441, 1988.

12. Elsdale, T., and Bard, J. Collagen substrata for studies on cell behavior. J.Cell Biol., 54: 626-637, 1972.

13. Schor, S. L. Cell proliferation and migration on collagen substrate in vitro.J. Cell Sci., 41: 159-175, 1980.

14. Ratner, S., and Heppner, G. H. Mechanisms of lymphocyte traffic in neoplasm. Anticancer Res., 6: 475-482, 1986.

15. Ratner, S., and Heppner, G. H. Motility and tumoricidal activity of interleu-kin-2 stimulated lymphocytes. Cancer Res., 48: 3374-3380. 1988.

16. Schor. S. L., Allen. T. D., and Winn, B. Lymphocyte migration into three-dimensional collagen matrices: a quantitative study. J. Cell Biol., 96:1089-1096, 1983.

17. Allen, T. D., Schor, S. L., and Schor, A. M. An ultrastructure review ofcollagen gels. A model system of cell-matrix, cell-basement membrane, andcell-cell interactions. Scanning Electron Microsc.. /: 375-390, 1984.

18. Loring, J.. Glimelius. B., and Weston, J. A. Extracellular matrix materialsinfluence quail neural crest cell differentiation in vitro. Dev. Biol., 90:165-174, 1982.

19. Hall, G. H., Parson, D. A., and Bissell, M. J. Lumen formation by epithelialcell lines in response to collagen overlay: a morphogenetic model in culture.Proc. Nati. Acad. Sci. USA, 79: 4672-4676, 1982.

20. Kramer, R. H. Extracellular matrix interactions with the apical surface ofvascular endothelial cells. J. Cell Sci., 76: 1-16, 1985.

21. Shields, J. M., Haston, W., and Wilkinson, P. C. Invasion of collagen gelsby mouse lymphoid cells. Immunology. 51: 259-268, 1984.

22. Hasten, W. S., Shields, J. M.. and Wilkinson, P. C. Lymphocyte locomotion

and attachment on two-dimensional surfaces and in three-dimensional matrices. J. Cell Biol., 92: 747-752, 1982.

23. O'Neill, G. J., and Parrott, D. M. V. Locomotion of human lymphoid cells.I. Effect of culture and Con A on T- and non-T-lymphocytes. Cell. Immunol¿Î:257-267, 1977.

24. Butcher, E. C. Cellular and molecular mechanisms that direct lymphocytetraffic. Am. J. Pathol., 136: 3-11. 1990.

25. Shaw, S., Luce, G. E. G., Quiñones.R., Gress, R. E., Springer, T. A., andSanders, M. E. Two antigen-independent adhesion pathways used by humancytotoxic T-cell clones. Nature (Lond.), 323: 262-264. 1986.

26. Gunji, Y., and Gorelik, E. Role of fibrin coagulation in protection of murinetumor cells from destruction by cytotoxic cells. Cancer Res., 48: 5216-52211988.

27. Belldegrun, A., Kasid, A., Uppenkamp, M., Topalian, S. L., and Rosenberg,S. A. Human tumor infiltrating lymphocytes. Analysis of lymphokine mRNAexpression and revelance to cancer immunotherapy. J. Immunol.. 142:4520-4526. 1989.

28. Itoh, K., Tilden, A. B., and Balch, C. M. Interleukin-2 activation of cytotoxicT-lymphocytes infiltrating into human metastatic melanomas. Cancer Res.,46:3011-3017, 1986.

29. Okada, Y., Yahata, G., Takeuchi. S., Seidoh, T., and Tanaka, K. A correlationbetween the expression of CDS antigen and specific cytotoxicity of tumor-infiltrating lymphocytes. Jpn. J. Cancer Res.. 80: 249-256, 1989.

30. Yahata, G., Ã’kada, Y.. Honda, S., Tanaka, K., Tokunaga, A.. Takahashi, T.,Seidou, T., and Takeuchi, S. Fundamental and clinical aspects of adoptiveimmunotherapy with tumor-infiltrating lymphocytes. Can To KagakuRyoho, 16: 1474-1482. 1989.

31. Topalian. S. L., Solomon, D., Avis, F. P., Chang. A. E., Freerksen, D. L.,Linehan, W. M., Lotze. M. T., Robertson, C. N., Seipp, C. A., Simon, P., etal. Immunotherapy of patients with advanced cancer using tumor-infiltratinglymphocytes and recombinant interleukin-2: a pilot study. J. Clin. Oncol., 6:839-853, 1988.

32. Rosenberg. S. A. Adoptive immunotherapy of cancer using lymphokineactivated killer cells and recombinant interleukin-2. Important Adv. Oncol.,55-91, 1986.

7158

Research. on November 17, 2020. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from

1990;50:7153-7158. Cancer Res   Kelli G. Applegate, Charles M. Balch and Neal R. Pellis  Arrested Locomotion in Tumor-infiltrating Lymphocytes

Migration of Lymphocytes through Collagen Matrix:In Vitro

  Updated version

  http://cancerres.aacrjournals.org/content/50/22/7153

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/50/22/7153To request permission to re-use all or part of this article, use this link

Research. on November 17, 2020. © 1990 American Association for Cancercancerres.aacrjournals.org Downloaded from