Adhesion of lymphoid cells to the carboxyl-terminal heparin-binding domains of fibronectin

Experimental Cell Research 181 (1989) 348-361

Adhesion of Lymphoid Cells to the Carboxyl-Terminal Heparin-Binding Domains of Fibronectin

NAN-SHIH LIAO,* JON1 ST. JOHN,* JAMES B. MCCARTHY,? LEO T. FURCHT,? and H. TAK CHEUNG*,’

*Department of Biological Sciences, Illinois State University, Normal, Illinois 61761, and fDepartment of Laboratory Medicine and Pathology, University of Minnesota,

Minneapolis, Minnesota 55455

Previously, we have shown that some lymphoid cell lines adhere to fibronectin (FN)- coated substratum, whereas others do not. In this study, the adhesion of five adherent lymphoid cell lines to different FN domains was examined. These cell lines ranged in their adherence to substratum coated with FN, the cell-binding domain (CBD) fragment, or the heparin-binding domain (HBD) fragments. None of the cell lines adhered to substratum coated with the gelatin-binding domain fragment. Three of the lymphoid cell lines adhered preferentially to HBD over CBD, whereas two other lymphoid cell lines and BHK tibroblasts adhered preferentially to CBD. These results suggest that two distinct adhesive interactions occur between cells and FN and that the pattern of interaction varies among cell types. Using MOPC 315 (which adheres preferentially to HBD) as a cell model to study the cell-HBD interaction, the HBD-promoted adhesion was found to be independent of the RGD sequence and could be inhibited by anti-FN antibodies. Moreover, the MOPC 31%HBD interaction had the following characteristics: (I) adhesion was temperature dependent, (2) presence of divalent cations was necessary, (3) integrity of cellular micro- filaments but not microtubules was required, (4) inhibition of protein synthesis abolished adhesion, (5) pretreatment of cells with trypsin inhibited adhesion, and (6) the adhesion was mediated by the carboxyl-terminal HBD. @ 1989 Academic press, IN.

Fibronectin (FN) is found in plasma, in extracellular matrices, and on the surface of certain cells [I]. Plasma FN is a large glycoprotein containing two similar subunits (A and B chains) with a molecular weight of approximately 220,000 Da each [l , 21. The FN polypeptide functions via multiple specialized intramolecular domains [3], which can be cleaved by proteolytic enzymes into fragments that retain their interactions with collagen, fibrin, heparin, and cells [41.

One of the cell-binding sites, which is contained within a 75- to 120-kDa fragment (generally known as the cell binding domain), can promote cell adhesion and spreading when immobilized on various artificial substrata [5]. From this fragment, a smaller fragment (11.5 kDa), which retains most of the biological activities of the larger fragment, was isolated by monoclonal antibodies directed against a site for cell attachment [6]. Further analysis using small synthetic peptides constructed according to the amino acid sequence of the 11.5 kDa fragment [7] revealed the sequence arginine-glycine-aspartic acid (RGD) [81, which is absent in other domains [9, lo], to mediate some of the cell-adhesion- promoting activity of FN.

’ To whom reprint requests should be addressed.

Copyright 0 1989 by Academic Press, Inc. All rights of reproduction in any form reserved 0014-4827/89 $03.00

348

Lymphoid cell adhesion to heparin-binding domains of FN 349

Even though FN promotes adhesion of many cell types, such as fibroblasts [ll], macrophages [12], neutrophils [13], platelets [14], and keratinocytes [15], little is known about its interaction with lymphocytes. However, lymphocytes could serve as a valuable experimental model to study FN-cell interactions because of their circulatory and migratory properties in uiuo. For example, after leaving the vasculature, lymphocyte subpopulations migrate into organs and localize at specific tissue sites depending on their maturation stages and function- al properties [16, 171. The regulatory mechanism for this tissue-specific and cell- type-specific localization of lymphocytes is unclear. The extracellular matrix, however, could be important in this process, because the composition of matrix components in different tissues and the ability of lymphocyte subpopulations to bind these components could be important factors in their adhesion and motility which, in turn, could determine their tissue positions.

In a previous study, we examined the adhesion of 12 different murine lymphoid cell lines to FN-coated substratum and observed a wide range of adhesiveness [18]. Among these cell lines, the physiology, biochemistry, and regulation of adhesion of a B-cell line, MOPC 315, to FN-coated substratum were examined [18]. In the present study, we further characterized the adhesion of MOPC 315 cells and several other lymphoid cell lines to various FN domains. Our results indicate that MOPC 315 cells, in addition to their binding to the CBD via an RGD- dependent mechanism, bind preferentially to the carboxyl-terminal heparin-bind- ing domain (HBD) by an RGD-independent mechanism. Furthermore, in contrast to MOPC 315 cells, two lymphoid cell lines and BHK fibroblasts bind preferen- tially to the cell binding domain (CBD). These findings suggest that cell-FN interaction is a complex process utilizing different domains of FN and functional- ly distinct receptors.

MATERIALS AND METHODS Chemicals and reagents. Ethyleneglycol bis-(B-aminoethyi ether)-N,N,N’,N’-tetraacetic acid

(EGTA), cycloheximide, trypsin (type II, from porcine pancreas), colchicine, and cytochalasin B were obtained from Sigma Chemical Co. (St. Louis, MO). Sheep anti-human Fn antiserum was purchased from Miles Scientific (Naperville, IL). Ethylenediaminetetraacetic acid (EDTA) was obtained from Fisher Scientific (Pittsburgh, PA). The hexapeptide glycine-arginine-glycine-aspartic acid-serine-proline (GRGDSP) was a generous gift from Dr. M. D. Pierschbacher, La Jolla Cancer Research Foundation (La Jolla, CA).

Cell lines. All the lymphoid cell lines used in this study were purchased from American Type Culture Collection (Rockville, MD). Their phenotypic characteristics are given in Table 1. They were maintained at 37°C in an air atmosphere containing 5 % CO2 in RPM1 1640 medium supplemented with 5 % Nu serum (Collaborative Research, Lexington, MA) and antibiotics (50 &ml of gentamycin and 50 U/ml of penicillin G, both from United States Biochemical Corp., Cleveland, OH). Cells that were to be used in the adhesion assay were maintained at a density lower than 8~ lo5 cells/ml. Under these conditions, the cell viability was greater than 95 % as determined by eosin Y dye exclusion.

The baby hamster kidney (BHK) libroblastic cell line [24] was a gift from Dr. F. Grinnell, Southwestern Medical School (Dallas, TX) and was grown in RPM1 1640 medium supplemented with 5 % Nu serum and antibiotics at 37°C in an air atmosphere containing 5 % CO?.

Purification ofhuman FN. FN was purified from human plasma using gelatin-affinity chromatogra- phy according to the procedure described by Engvall and Ruoslahti [25]. The gelatin-affinity column was prepared by coupling gelatin (Type I from porcine skin, Sigma Chemical Co.) to cyanogen bromide-activated Sepharose 4B (Sigma Chemical Co.) according to the manufacturer’s procedure. Fresh human plasma was purchased from American Red Cross (Peoria, IL). Phenylmethylsulfonyl

350 Liao et al.

TABLE 1

Characteristics of lymphoid cell lines

Cell line Surface antigens Other characteristics References

MOPC 315 H-2d IgA secreting; dexamethasone sensitive H91 MPCll OUA H-2d IgG2b secreting; ouabaine resistant WI P3X63Ag8 H-2d IgGl secreting; 8-azaguanine (10e4 M) Pll

resistant xc1.5/51 G Non-secreting myeloma; near tetraploid P21 SP2/0-Ag14 H-2d Hybrid does not produce Ig; 8-azaguanine P31

(20 ug/ml) resistant

fluoride (PMSF, Sigma Chemical Co.) was added to the plasma and to all solutions used in the purification to a final concentration of 10e4 M. The plasma was centrifuged at 10,OOOg for 20 min to remove particulate materials and applied to the gelatin-affinity column in a volume that was twice the size of the column bed volume. The column was washed with phosphate-buffered saline (PBS, 0.15 M NaCl, 0.01 M sodium phosphate, pH 7.4) until the absorbance dropred to the baseline level. A solution containing 1 M NaBr (Fisher Scientific) and 0.05 M sodium acetate (Sigma Chemical Co.), pH 5, was then used to elute the FN. FN purified by this procedure was shown to be homogeneous by two closely spaced bands in SDS-polyacrylamide gel electrophoresis (SDS-PAGE), with 10% acrylamide, under reducing conditions. The EW nm for a 1% solution of FN is 13 [26].

Chymotryptic digest of FN. The chymotryptic digest of FN and the purification of the proteolytic fragments were performed according to the procedure described by Pierschbacher et al. [5]. Purified FN (10 mg) was concentrated by precipitation with (NH&SO4 at 50 % saturation at room temperature. The precipitate was collected by centrifugation at 1200s for 30 min and resuspended in a minimal amount of 50% saturated (NH&SO4 solution. The preparation was dialyzed for 12 h against 3 liters of PBS containing 0.02% NaNl at room temperature. The concentration of the dialyzed FN was adjusted within the range of 2 to 8 mg/ml as determined by A 280. The dialyzed FN was digested with type VII a-chymotrypsin (Sigma Chemical Co.) at a concentration of 10 ug enzyme/mg FN at 37°C for 1 h. The digested FN was passed through a gelatin-aftinity column (1 x 10 cm) at a flow rate of 0.5 ml/min. After the nonbound materials were washed wth PBS, the gelatin-binding fragment (GBD) was eluted with 4 M urea in 0.05 M Tris-HCl buffer (pH 7.5). The nonbound fraction was purified by heparin affinity chromatography using Aft&gel heparin (Bio-Rad, Richmond, CA). The non-heparin- bound materials were eluted with PBS, and the bound materials were eluted with 0.5 M NaCl in 50 m&f Tris-HC1 buffer (pH 7). Each fraction was analyzed by 10% SDS-PAGE under reducing conditions. The heparin-binding fraction contained three fragments [5]. The non-heparin-binding, non-gelatin-binding fraction contained a single l20-kDa band, which is the cell-binding fragment [5].

Purification ofheparin-bindingfragments. The tryptic/catheptic fragments of FN were prepared as previously described [27] by sequential affinity chromatography over gelatin, heparin, and mono- clonal antibody affinity columns. Three heparin-binding fragments were isolated: (a) a 27-kDa tryptic fragment from the amino-terminus that binds heparin weakly, (b) a 33-kDa trypticicatheptic heparin- binding fragment adjacent to the 30 kDa fibrin-binding fragment at the carboxyl terminus of the A chain, and (c) a 60-kDa fragment from the carboxyl terminus of the B chain.

Adhesion assay. The adhesion of lymphoid cells to FN-coated substratum was determined by an adhesion assay using radioactively labeled cells and FN-coated 20-ml polyethylene scintillation vials (Fisher Scientific) [18]. The labeled cells were prepared by incubating the cells with [‘Hlleucine (40-60 Wmmol, ICN, Irvine, CA) at a concentration of 4 @i/2x 10’ cells/ml for 16 h at 37’C, followed by washing the cells three times with PBS to remove unincorporated radioactivity. The protein-coated substratum was prepared by first coating the polyethylene vials at 37°C for 30 min with FN or other polypeptides in serum-free medium at the concentrations indicated under Results, followed by three washes with PBS. Any uncoated sites were filled by coating the vials with bovine serum albumin (BSA, Sigma Chemical Co.) at a concentration of 2 mg/ml under the same conditions. For control, vials were coated only with BSA. The adhesion assay was carried out by incubating 1 X 10’ labeled cells in each vial at 37°C for 2 h or for different durations as indicated under Results. At the end of

Lymphoid cell adhesion to heparin-binding domains of FN 351

MOPC 315

OFN -HBD WlRKBD

MPCll OUA

30

20

10

0

20

10

0 xc1.5/51 P3X63AG8

60 -

SP2/0

Fig. 1. Adhesion of lymphoid cell lines to FN, HBD, and CBD. The coating concentration in equivalent molar concentration of FN was 15 ug/ml for all three polypeptides. Percentage specific adhesion was determined as described under Materials and Methods. The nonspecific adhesion to BSA for the cell lines was between 3 and 10%.

incubation, the nonadherent cells were removed by washing three times with PBS. The adherent cells were determined by liquid scintillation as counts per minute (cpm). Each determination was per- formed in triplicate, and the results were expressed as means + SD. The following formula was used to express the results:

cpm for adherent cells (Exp)-cpm for adherent

% specific adhesion = cells to BSA (control) x 100. cpm for total cells added to vial

RESULTS

Adhesion of lymphoid cells to FN, CBD, and HBD. In our previous study, we identified five lymphoid cell lines that adhere to FN-coated substratum [18]. In this study, we examined the specific adhesion of these lymphoid cells to substra- tum coated with the CBD, HBD, or GBD fragments. The five cell lines adhered to CBD- and HBD-coated substratum to different extents (as shown in Fig. l), but none adhered to GBD-coated substratum (data not shown). Cell lines MOPC 315, P3X63AG8, and SP2/0-Ag14 exhibited higher adhesion to HBD than to CBD, with percentage specific adhesion to HBD at least twice that to CBD, whereas the cell lines MPCI 1 OUA and XC1.5/51 exhibited higher adhesion to CBD than to HBD. These results suggest that HBD, in addition to CBD, can promote adhesion of lymphoid cell lines. To further study the HBD-lymphoid cell interac- tion, the adhesion of MOPC 315 cells to HBD was characterized.

Adhesion of MOPC 315 cells to FN, CBD, and HBD. To compare the optimal concentration of intact FN, CBD, and HBD in promoting adhesion of MOPC 315 cells, the substratum was coated with different concentrations of intact FN or equivalent molar concentrations of CBD and HBD. As shown in Fig. 2, the

23-898334

352 Liao et al.

0.01 0.10 1 .oo 10.00 100.00 COATING CONCENTRATION

f&g/ml OR MOLAR EQUIVALENT OF FN)

Fig. 2. Adhesion of MOPC 315 cells to FN, HBD, and CBD. The substratum was prepared by coating scintillation vials with different polypeptides at concentrations indicated in the figure. Per- centage specific adhesion was determined as described under Materials and Methods.

adhesion of MOPC 315 cells to substratum coated with FN, CBD, or HBD was dependent on the coating concentrations. At 10 kg/ml, 38,23, and 11% of MOPC 315 cells adhered to intact FN, HBD, and CBD, respectively. With higher coating concentrations, there was no increase in adhesion to FN, but there was an increase in adhesion to HBD and CBD, with maximum adhesion of 36 and 14 % to HBD and CBD, respectively, at 50 @ml. As indicated in previous studies [27] and in the present study (data not shown), using intact FN and fragments labeled with tritium by reductive methylation, the fragments coated the plastic substra- tum at least as well as or better than intact FN. Therefore, the difference in adhesion of MOPC 315 cells to intact FN, HBD, and CBD was not due to the inability of the fragments to coat the substratum. These results demonstrate that HBD, although not as effective as intact FN, is more effective than CBD in promoting the adhesion of MOPC 315 cells.

Efsect ofGRGDSP on adhesion. To determine whether the adhesion of MOPC 315 cells to HBD was mediated by RGD-containing fragments in the HBD preparation, the effect of soluble GRGDSP was tested. As shown in Fig. 3, GRGDSP at 100 &ml inhibited the adhesion of MOPC 315 cells to CBD by 40%, inhibited the adhesion to FN by 20%, and had no effect on the adhesion to HBD. Concentrations of GRGDSP as high as 1 mg/ml also had no effect on the adhesion of MOPC 315 cells to HBD (data not shown), although the effect at this concen- tration was greater on FN (45 % inhibition) and on CBD (65 % inhibition) than at 100 ug/ml. This result indicates that the adhesion of MOPC 315 cells to HBD is independent of the RGD-mediated mechanism, whereas the adhesion to CBD is mediated by an RGD-dependent mechanism. In addition, the effect of GRGDSP on the adhesion of BHK fibroblasts to PN,“CBD, and HBD was also tested. As shown in Fig. 3, CBD promoted adhesion of BHK tibroblasts, and the adhesion was inhibited by GRGDSP, as previously reported [28]. However, BHK fibro-

Lymphoid cell adhesion to heparin-binding domains of FN

80 MOPC 315 BHK

0Nedbmlomtfol l-7

OMedlum contml mSoRoDsp (loo w/ml) _

30 CBD R-4 HBD CBD

353

Fig. 3. Effect of cell adhesion peptide GRGDSP on the adhesion of MOPC 315 cells and BHK fibroblasts to FN, HBD, and CBD. The substratum was coated with FN, HBD, or CBD at 5, 10, and 10 t&g/ml, respectively, for the adhesion of MOPC 315 cells; and 1,3, and 3 ug/ml, respectively, for the adhesion of BHK fibroblasts. Percentage specific adhesion was determined as described under Materials and Methods.

blasts also displayed weak adhesion to HBD, which was not inhibited by GRGDSP (Fig. 3).

Znhibition of adhesion by anti-FN antiserum. A polyclonal anti-FN antiserum was employed to determine the nature of the HBD recognition site. Because the anti-FN antiserum was raised against the intact FN molecule, it should inhibit the adhesion of MOPC 315 cells to HBD, if the HBD recognition site is not the result of exposure by proteolytic cleavage but rather is exposed on the intact FN molecule in its native conformation. The intact FN-, HBD-, or CBD-coated substratum was preincubated with anti-FN antiserum or control serum for 30 min before assaying for adhesion in the presence of the antiserum or control serum. As shown in Fig. 4, adhesion of MOPC 315 cells to FN and CBD was completely

4. OControl Serum c

’ T 1

FN

BYAnti-FN Ab 1

I&- L CBD

Fig. 4. Effect of anti-FN antibody on the adhesion of MOPC 315 cells to FN, HBD, and CBD. The substratum was coated with FN, HBD, or CBD at 5, 10, and 10 &ml, respectively, and preincubated with 50 pl of goat anti-human FN serum (1: 4 dilution) or nonimmune goat serum in 2 ml RPM1 1640 medium for 15 min at room temperature before the addition of cells. Percentage specific adhesion was determined as described under Materials and Methods.

354 Liao et al.

60

l -OHED

0 10 20 30 40

TEMPERATURE (“C)

Fig. 5. Effect of temperature on MOPC 315 cell adhesion to FN and HBD. The substratum was coated with FN or HBD at 5 and 10 &ml, respectively. Percentage specific adhesion was determined as described under Materials and Methods.

inhibited by the anti-FN antiserum. The adhesion to HBD was also inhibited, even though not completely. This result suggests that the HBD recognition site for cell adhesion is normally exposed on the intact FN molecule, and the adhesion is the result of a specific interaction. However, the inability of the adhesion to HBD to be completely blocked by the anti-FN antiserum suggests the existence of another distinct interaction between the MOPC 315 cells and HBD. Similarly, it is also possible that the antiserum recognizes fewer epitopes in the HBD as compared to CBD and therefore is unable to completely inhibit the binding to HBD.

Temperature dependence and the requirement of divalent cations in MOPC 315 cell adhesion to FN and HBD. Adhesion of MOPC 315 cells to FN and HBD was determined at different temperatures. As shown in Fig. 5, adhesion to FN

60 -

0.02 0.10 1 .oo 10.00

CONCENTRATION (mM)

Fig. 6. Effect of EDTA and EGTA on MOPC 315 cell adhesion to HBD. The substratum was coated with HBD at 10 t&ml. Percentage specific adhesion was determined as described under Materials and Methods.

Lymphoid cell adhesion to heparin-binding domains of FN 355

0-OColchicine \ l -•Cytochalosin B

\ i

0 0.02 0.10 1 .oo 10.00 100.00

CONCENTRATION (/Ad)

Fig. 7. Effect of colchicine and cyt0chalasin.B on MOPC 315 cell adhesion to HBD. The substratum was coated with HBD at 10 &ml. Percentage specific adhesion was determined as described under Materials and Methods.

and HBD was inhibited at 4°C; however, adhesion to both increased with rising temperature and reached a maximum at 37°C. This result suggests that at 4°C the adhesion was prevented either by the inhibition of energy metabolism or by a decrease in membrane fluidity. The first explanation appears to be correct because adhesion was prevented by inhibiting energy metabolism with 2-deoxy- glucose and sodium azide at 37°C (data not shown). Furthermore, adhesion to HBD required the presence of divalent cations. As shown in Fig. 6, increasing concentrations of both EDTA and EGTA inhibited the adhesion of MOPC 315 cells to HBD, which correlates with the results that MOPC 315 cells did not adhere to HBD in the absence of calcium and magnesium ions (data not shown).

Effect of colchicine and cytochalasin B on MOPC 315 cell adhesion. The adhesion of fibroblasts to FN-coated substratum was shown previously to require the integrity of microfilaments but not microtubules [29, 301. To investigate whether adhesion of MOPC 315 cells to HBD-coated substratum also requires intact microfilaments, cells were assayed for adhesion in the presence of either cytochalasin B or colchicine. As shown in Fig. 7, cytochalasin B at 10m4 M reduced cell adhesion by 80 %, whereas colchicine had no effect at the concentra- tions tested. The reduction in adhesion to HBD in the presence of cytochalasin B is not due to cell death, because at the end of the adhesion assay, the cell viability remained unchanged. Furthermore, the effect of cytochalasin D (which does not inhibit glucose uptake) was found to be the same as cytochalasin B (data not shown), providing further evidence that the inhibition of adhesion is the result of disruption of the microfilaments. These results- suggest that the adhesion of MOPC 315 cells to HBD requires the integrity of microfilaments but not microtu- bules.

Effect of protein synthesis inhibitors and trypsin treatment on MOPC 315 cell adhesion. Protein synthesis inhibitors, cycloheximide and puromycin, were test- ed for their effects on MOPC 315 cell adhesion to intact FN-, HBD-, or CBD-

356 Liao et al.

50 c Oktedium contml IDXycloheximide

5 z 40-

!+! 2 30- 0 k g 207 %

x lo-

OL li, FN n E

CBD

Fig. 8. Effect of cycloheximide on MOPC 315 cell adhesion to HBD. The substratum was coated with HBD at 10 &nl. Cells were preincubated with cycloheximide (40 ug/rnl) at 37°C for 2 h and tested for adhesion in the presence of cycloheximide. Percentage specific adhesion was determined as described under Materials and Methods.

coated substratum. Cells were treated for 2 h with cycloheximide and puromycin and tested for adhesion in the presence of the inhibitors. Figure 8 illustrates the effect of cycloheximide (the effect of puromycin was essentially the same). Adhesion to FN, HBD, or CBD was inhibited by 83, 64, and 68%, respectively, indicating that protein synthesis is required for adhesion. Because the cell adhe- sion to these substrata could not be completely inhibited, there must also exist an additional cell adhesion mechanism that does not require protein synthesis. In addition, cells were subjected to trypsinization before testing for adhesion to FN-, HBD-, or CBD-coated substratum. As shown in Fig. 9, trypsinization inhibited adhesion of MOPC 315 cells to FN, HBD, or CBD by 70, 65, and 53%, respectively. In contrast, trypsinization only slightly reduced the adhesion of BHK fibroblasts to FN and CBD (<20%), a result similar to that previously

MOPC 315 au8dhJm cLontd oF1Ttypdnlzd

M HBD CBD

r BHK

Othdlum control

FN HBD

Fig. 9. Effect of trypsin treatment on MOPC 31.5 cell adhesion to HBD. The substratum was coated with HBD at 10 &ml. Cells were treated with trypsin (0.5 mg/ml) at 37°C for 10 min, washed, and assayed for adhesion. Percentage specific adhesion was determined as described under Materials and Methods.

Lymphoid cell adhesion to heparin-binding domains of FN 357

$

q60

2 u

zi40 k!

;20

0 10 100

COATING CONCENTRATION (&ml OR MOLAR EQUIVALENT OF FN)

Fig. 10. Adhesion of MOPC 315 cells to purified HBD fragments. The substratum was prepared by coating scintillation vials with diierent polypeptides at concentrations indicated in the figure. Per- centage specific adhesion was determined as described under Materials and Methods.

reported [31], whereas trypsinization inhibited the adhesion of BHK fibroblasts to HBD by 71%.

Adhesion of MOPC 315 cells to purified heparin-binding fragments. The HBD preparation used in the preceding experiments potentially contained three hepa- tin-binding fragments: a 27-kDa fragment from the amino termini of the A and B chains, a 33-kDa heparin-binding fragment adjacent to the 30-kDa fibrin-binding fragment at the carboxyl terminus of the A chain, and a 60-kDa fragment at the carboxyl terminus of the B chain [32]. To determine which HBD fragment(s) was responsible for the adhesion-promoting activity, each of the isolated fragments was tested. The identity and purity of these fragments had been previously established [27]. As shown in Fig. 10, the 33-kDa HBD fragment displayed the highest adhesion-promoting activity; the 66-kDa fragment also promoted adhe- sion but was less active; and the 27-kDa fragment had only slight adhesion- promoting activity at the concentrations tested.

DISCUSSION

In this study, we have presented evidence that the adhesion of MOPC 315 cells to FN-coated substratum is the result of at least two distinct interactions between cells and FN cell recognition sites. The predominant interaction between MOPC 315 cells and FN is the carboxy-terminal HBD by an RGD-independent mechan- ism. This conclusion is supported by the evidence that (1) when the substratum was coated with equivalent molar concentrations of HBD and CBD, the adhesion of MOPC 315 cells to HBD was two to three times greater than to CBD; (2) at optimum substratum coating concentration, the adhesion of cells to HBD equaled that to intact FN, but the adhesion to CBD remained two to three times lower; and (3) the RGD-containing peptide inhibited adhesion to CBD, but not adhesion to HBD. Similar results were obtained with two other lymphoid cell lines, P3X63Ag8 and SP2/0-Ag14. However, the adhesion of the lymphoid cell lines,

358 Liao et al.

MPCll OUA and XC1.5/51, and BHK libroblasts appears to use mainly the CBD-dependent interaction, supplemented by the HBD-dependent interaction. Thus, more than one type of interaction is possible between FN and lymphoid cells in mediating cell adhesion, and the predominant type of interaction is distinct for different types of cells. However, it is likely that multiple interactions between cells and the FN molecule contribute to the overall adhesion of 1 ymphoid cells.

The results also support findings of studies on the ability of HBD to promote adhesion and spreading of other cells. McCarthy et al. [27] reported that both HBD and CBD promote cell adhesion and spreading of B16 melanoma cells, although only the CBD possesses haptotatic activity. They also reported that the cell-adhesion-promoting activity of the HBD was independent of the RGD se- quence and appears to use a receptor that is functionally distinct from that which interacts with the RGD sequence [27]. Humphries et al. [33, 341 also reported an adhesion- and spreading-promoting activity for B16 melanoma cells in the alter- natively spliced IIICS region between the HBD and carboxyl-terminal fibrin- binding domain of the A chain. However, preliminary studies (McCarthy and Furcht, submitted for publication) have shown that a unique adhesion-promoting activity in the HBD is in a type III repeat that is present normally in various forms of FN and is outside the region that is alternatively spliced.

Although cell adhesion can be promoted independently by CBD or HBD, the cooperative interactions between cell-surface structures and different FN do- mains have been firmly established to be necessary for achieving the spreading morphology of certain cells 135, 361. For example, 3T3 fibroblasts, when spread on substratum coated with intact FN, form tight contacts and extensive intracel- lular stress fibers [37]. Cells that spread on substratum coated with either CBD or HBD, however, form only close contacts with actin filaments restricted in their spiky projections and lack intracellular stress fibers [36, 371. These results provide convincing evidence that, in addition to the RGD-mediated FN receptor, other cell-surface molecules are also involved in the interaction with HBD in order to achieve full spreading. Several recent studies suggested cell-surface proteoglycans (PG) as a cell-surface receptor in promoting interactions with HBD. For example, mutant CHO fibroblasts without cell-surface PG are reported to be unable to form F-actin-containing stress fibers as that found in wild-type cells, although the mutant cells are able to adhere to a FN-coated substratum by an RGD-dependent mechanism [38]. In another study, mouse mammary epithelial cells were shown to adhere to the carboxyl-terminal HBD by cell-surface PG, which when isolated and incorporated into liposomes, promotes the adhesion of the liposomes to FN and HBD [39]. In this study, because the MOPC 315 cell adhesion to HBD is sensitive to the inhibition of protein synthesis and to the proteolytic treatment of the cells, the cell-surface recognition structure for HBD is at least partially protein. In addition, preliminary studies demonstrated that neither heparin nor heparan sulfate can inhibit the adhesion of MOPC 3 15 cells to HBD (data not shown) suggesting the possibility of PG as the receptor structure for HBD to be unlikely.

Lymphoid cell adhesion to heparin-binding domains of FN 359

As demonstrated previously for the adhesion of many cell types to FN, MOPC 315 cell adhesion to HBD is energy-dependent and is sensitive to the effect of cytochalasin B. These results suggest that reorganization of the cytoskeleton, an energy-dependent process, is necessary for cell adhesion to HBD. These results provide evidence that lymphoid cell adhesion to HBD is not a result of nonspecif- ic charge interaction between HBD and cell-surface molecules, but is an active adhesive process.

What is the significance of this finding, particularly for lymphocyte migration and localization, processes that are important in the immunological function of these cells? Lymphocyte subpopulations localize in different anatomical sites during their maturation and differentiation, such as the localization of immature T cells in the thymus. Their tissue localization is also related to their functional properties. For example, IgA producing B cells localize in the mucosa of the gut, where secreted IgA are actively transported by mucosal epithelial cells into the intestinal lumen [40]. However, the IgG producing B cells are found primarily in lymph nodes [17], from which the IgG are secreted into the circulation. Further- more, T cells are more abundant in skin epidermis than in mucosal linings 1411 for reasons that are not apparent.

Several mechanisms have been proposed to explain the phenomenon of selec- tive lymphocyte migration and localization. These include the lymphocyte cell- surface “homing” receptor that can bind selectively to endothelial cells of lymphoid organs located at different anatomical regions [42, 431 and the lympho- cyte function-associated-l (LFA-1) and other related cell-surface antigens [44, 451. The LFA-1 cell-surface molecule promotes interaction between lymphocytes and other tissue cells, such as epithelial cells and fibroblasts, which possess the corresponding cell-surface ligand ICAM- 1 [46, 471. These cell-surface receptors appear to be crucial in the distribution of lymphocytes from the circulation to various tissues [43, 451. The factors, however, that determine the cell-type- specific tissue localization remain largely unknown. It is conceivable that the interactions between lymphocytes and extracellular matrix components, such as FN, direct cell migration and localization. Two possible regulatory mechanisms could exist through the selective adhesive interaction of lymphoid cells and FN.

The first regulatory mechanism of cell localization could involve the differen- tial expression, either qualitatively or quantitatively, of the cell-surface FN receptors; the absence of functional receptors at certain stages of the cell’s life cycle would release the cells from a tissue environment that is enriched with FN in the interstitial matrix. Examples of such a regulatory mechanism have been demonstrated in the maturation and the release of reticulocytes from bone marrow into the peripheral circulation [48-501, in the homing of hemopoietic precursor cells to the embryonic thymus [51], and in the interaction of hemopoiet- ic cells with bone narrow matrix [52]. Such a mechanism is also likely to be involved in the regulation of lymphocyte localization. We have observed pre- viously that in a survey of 12 different lymphoid cells lines there is a wide range of adhesiveness of these cells to FN-coated substratum [18]. Some of these cell lines are as adhesive as fibroblasts, whereas some are nonadherent to FN-coated

360 Liao et al.

substratum. This difference in interaction with FN has also been found in thymocytes [53], activated T cells [54], and bone marrow pre-B lymphocytes [55]. In these cell populations, there exist distinct subpopulations of FN-adherent and FN-nonadherent cells. The second regulatory mechanism on lymphocyte migra- tion and localization by cell-FN interaction could be the cell type-specific inter- action with different FN cell recognition sites. Evidence in this study supports this hypothesis, and a recent study using B-lymphoid cell lines that differ in their stage of differentiation also indicates the cell-specific interaction with HBD and CBD [55]. To provide further evidence in support of this hypothesis, studies are under way in our laboratory to determine the interaction between normal lym- phocyte populations and different FN domains.

The incorporation of FN in the interstitial matrix, depending on the other interacting matrix components, could result in different molecular conformations because of the apparent flexibility of the molecule as indicated by ultrastructural studies [2]. In some conformations, the CBD might be more available than HBD, and in other conformations, the HBD might be more available for cell binding to occur. The ability of lymphoid cells and other cells to interact with specific FN domains suggests that the regulation of cell migration and localization by FN is a complex process in rho.

We thank Ms. D. L. Xia for her technical assistance. This research was supported in part by Grant AI20343 from the National Institute of Allergy and Infectious Diseases and by an Organized Research Grant from Illinois State University,

REFERENCES 1. Akiyama, S. K., and Yamada, K. M. (1987) Adu. Enzymol. 59, l-57. 2. Erickson, H. P., Carrell, N., and McDonagh, J. (1981) J. Cell Biol. 91, 673-678. 3. Ruoslahti, E., Hayman, E. G., Engvall, E., Cothran, W. C., and Butler, W. (1981) J. Biol. Chem.

256, 7277-7281. 4. Hynes, R. O., and Yamada, K. M. (1982) J. Cell Biol. 95, 369-377. 5. Pierschbacher, M. D., Hayman, E. G., and Ruoslahti, E. (1981) Cell 26, 259-267. 6. Pierschbacher, M. D., Ruoslahti, E., Sundelin, J., Lind, P., and Peterson, P. A. (1982) .I. Biol.

Chem. 257,9593-9597. 7. Pierschbacher, M. D., Hayman, E. G., and Ruoslahti, E. (1983) Proc. Null. Acad. Sci. USA 80,

1224-1227. 8. Pierschbacher, M. D., and Ruoslahti, E. (1984) Nature (London) 309, 3G33. 9. Schwarzbauer, J. E., Tamkun, J. W., Lemischka, I. R., and Hynes, R. 0. (1983) Cell35,421-431.

10. Skorstengaard, K., Jensen, M. S., Sahl, P., Petersen, T. E., and Magnusson, S. (1986) Eur. J. Biochem. 161, 441-453.

11. Grinnell, F., Hays, D. G., and Minter, D. (1977) Exp. Cell Res. 110, 175-190. 12. Bevilacqua, M. P., Amrani, D., Mosesson, M. W., and Bianco, C. (1981) J. Exp. Med. 153,

42-60. 13. Marino, J. A., Pensky, J., Gulp, L. A., and Spagnuolo, P. J. (1985) J. Lab. Clin. Med. 105,

725-730. 14. Koteliansky, V. E., Leytin, V. L., Svirdov, D. D., Repin, V. S., and Smirnov, V. N. (1981) Fed.

Eur. Biochem. Sot. 123, 59-62. 15. Clark, R. A. F., Folkvord, J. M., and Wertz, R. L. (1985) J. Invesr. Dermatol. 84, 378-383. 16. Parrott, D. V. M., and de Sousa, M. A. B. (1966) Nature (London) 212, 1316-1317. 17. Parrott, D. V. M., and Wilkinson, P. C. (1981) Prog. Allergy 28, 193-284. 18. Liao, N.-S., St. John, J., Du, Z. J., and Cheung, H. T. (1987) Exp. Cell Res. 171, 306-320. 19. Eisen, H. N., Simms, E. S., and Potter, M. (1968) Biochemistry 7, 4126-4136. 20. Gefter, M. L., Margulies, D. H., and Scharff, M. D. (1977) Somatic Cell Genet. 3, 231-236.

Lymphoid cell adhesion to heparin-binding domains of FN 361

21. Kohler, G., and Milstein, C. (1975) Nature (London) 265, 495-497. 22. Ralph, P. (1973) .I. Immunol. 110, 1470-1475. 23. Shulman, M., Wilde, C. D., and Kohler, G. (1978) Nature (London) 276, 269-270. 24. Grinnell, F. (1980) .I. Cell Biol. 86, 104-112. 25. Engvall, E., and Ruoslahti, E. (1977) Z+t. J. Cancer 20, l-5. 26. Ruoslahti, E., Hayman, E. G., Pierschbacher, M., and Engvall, E. (1982) in Methods in Enzymo-

logy (Cunningham, L. W., and Frederiksen, D. W., Eds.), Vol. 82, pp. 803-831, Academic Press, New York.

27. McCarthy, J. B., Hagen, S. T., and Furcht, L. T. (1986) J. Cell Biol. 102, 179-188. 28. Yamada, K. M., and Kennedy, D. W. (1984) J. Cell Biol. 99, 29-36. 29. Ah, I. U., Mautner, V. M., Lanza, R. P., and Hynes, R. 0. (1977) Cell 11, 115-126. 30. Hynes, R. O., Destree, A. T., and Wagner, D. D. (1982) Cold Spring Harbor Symp. Quant. Biol.

46, 659-670. 31. Oppenheimer-Marks, N., and Grinnell, F. (1984) Exp. Cell Res. 152, 467-475. 32. Smith, D. E., and Furcht, L. T. (1982) J. Biol. Chem. 257, 6518-6523. 33. Humphries, M. J., Akiyama, S. K., Komoriya, A., Olden, K., and Yamada, K. M. (1986) J. Cell

Biol. 103, 2637-2648. 34. Humphries, M. J., Komoriya, A., Akiyama, S. K., Olden, K., and Yamada, K. M. (1987) J. Biol.

Chem. 262, 6886-6892. 35. Laterra, J., Siebert, J. E., and Culp, L. A. (1983) J. Cell Biol. 96, 112-123. 36. Woods, A., Couchman, J. R., Johansson, S., and Hook, M. (1986) Eur. Mol. Biol. Organ. J. 5,

665670. 37. Lark, M. W., Laterra, J., and Culp, L. A. (1985) Fed. Proc. 44, 394-403. 38. LeBaron, R. G., Esko, J. L., Woods, A., Johansson, S., and Hook, M. (1988) J. Cell Biol. 106,

945-952. 39. Saunders, S., and Bernfield, M. (1988) J. Cell Biol. 106, 423-430. 40. Parrott, D. V. M. (1976) Clin. Gastroenterol. 5, 211-228. 41. Cahill, R. N. P., Poskitt, D. C., Frost, H., and Tmka, Z. (1977) J. Exp. Med. 145, 42OA28. 42. Gallatin, W. M., Weissman, I. L., and Butcher, E. C. (1983) Nature (London) 304, 30-34. 43. Woodruff, J. J., Clarke, L. M., and Chin, Y. H. (1987) Annu. Rev. Zmmunol. 5, 201-222. 44. Martz, E. (1987) Human Immunol. 18, 3-37. 45. Springer, T. A., Dustin, M. L., Kishimoto, T. K., and Marlin, S. D. (1987) Annu. Rev. Immunol.

5, 223-252. 46. Dustin, M. L., Rothlein, R., Bhan, A. K., Dinarello, C. A., and Springer, T. A. (1986) J.

Immunol. 137, 245-254. 47. Marlin, S. D., and Springer, T. A. (1987) Cell 51, 813-819. 48. Patel, V. P., and Lodish, H. F. (1984) Science 224, 996-998. 49. Patel, V. P., and Lodish, H. F. (1986) J. Cell Biol. 102, 449-456. 50. Patel, V. P., Ciechanover, A., Platt, O., and Lodish, H. F. (1985) Proc. Natl. Acad. Sci. USA 82,

440444. 51. Savagner, P., Imhof, B. A., Yamada, K. M., and Thiery, J.-P. (1986) J. Cell Biol. 103,2715-2728. 52. Giancotti, F. G., Comoglio, P. M., and Tarone, G. (1986) J. Cell Biol. 103, 429A37. 53. Cardarelli, P. M., and Pierschbacher, M. D. (1986) Proc. Natl. Acad. Sci. USA 83, 2647-2651. 54. Kurki, P., Vartio, T., and Virtanen, I. (1987) Stand. J. Zmmunol. 26, 645652. 55. Bemardi, P., Patel, V. P., and Lodish, H. F. (1987) J. Cell Biol. 105, 489-498.

Received May 18, 1988 Revised version received October 14, 1988

Printed in Sweden