Adhesion of B Cell Lines to Endothelial Cells from Human Lymphoid Tissue Modulates Tyrosine Phosphorylation and Endothelial Cell Activation

Reagents, cells, and Abs

Mouse anti-human β₁ integrin (Lia1/2; IgG1) (20) and anti-human α₄ integrin (HP2/1; IgG1) (21) were provided by Dr. Sánchez-Madrid (Servicio de Inmunología, Hospital de la Princesa, Madrid, Spain). Mouse anti-human β₁ integrin (TS2/16.2.1; IgG1) and anti-human CD40 (G28-5; IgG1) were obtained from American Type Culture Collection (Manassas, VA). Rabbit anti-paxillin (H114), anti-pp125^FAK (A17), and anti-ERK2 (C14) Ab were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-ERK1 (MK12) mAb was purchased from Transduction Laboratories (Lexington, KY). Anti-phosphotyrosine Ab (anti-pTyr Ab; 4G10) was furnished by Dr. S. Fischer (Genopole, Evry, France). Anti-Fy6 (clone 2C3), an mAb to Duffy Ag receptor for chemokine (DARC), a high endothelial cell marker (22), was supplied by Dr. J.-P. Cartron (Institut National de la Santé et de la Recherche Médicale, Unité 76, Institut National de Transfusion Sanguine, Paris, France). HRP-conjugated goat anti-rabbit IgG and goat anti-mouse IgG were obtained from Biosys (Compiegne, France). For cross-linking we used a goat anti-mouse IgG F(ab′)₂ from DAKO (Copenhagen, Denmark). Other Abs used are as indicated in Ref. 23 . All other reagents were purchased from Sigma (St. Louis, MO) unless noted otherwise.

HUTEC were obtained from human tonsils and characterized as previously described (see below) (23). In brief, human tonsils obtained after routine tonsillectomy were sequentially rinsed in sterile PBS, 70% ethanol, and sterile PBS, and the covering epithelium was removed. After cutting in small pieces, the tonsils were digested with 400 μg/ml collagenase A and 100 μg/ml DNase type I for 1 h at room temperature. The mixture was sedimented at 1 × g for 2 min, and the supernatant was centrifuged at 250 × g for 10 min. The cells were washed twice in RPMI 1640 (Life Technologies, Grand Island, NY) and plated at a density of 2 × 10⁶ cells/cm² in RPMI 1640 supplemented with 2 mM l-glutamine, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 10% FBS (Life Technologies). FACScan studies of these cells showed a homogeneous population of cells expressing several endothelial cell markers (23), such as intracellular expression of von Willebrand factor, vimentin, and lymphocyte vascular adhesion protein-2, a marker displayed by endothelial cells of lymphoid organs (24). They express CD40 and integrin subunits β₁, α₁, α₄, and α₅. Although 50% of early passage cells express ICAM-1, and 20% express VCAM-1 constitutively, later passages show that the proportion of cells expressing ICAM-1 remains unchanged, while the fraction of cells expressing VCAM-1 decreased to <1% (data not shown). The cells are negative for markers characteristic of follicular dendritic cells (R4/23, KI M4, and CD11b), interdigitant dendritic cells (CD11c), and monocyte/macrophage (CD11b, CD14, and β₂ integrin chain) (for details, see Ref. 23). More recently, we have determined that these cells are positive for DARC, a cell surface marker specific for human high endothelial venules (HEVs; data not shown) (22).

The human B cell lines Ramos and Daudi and the T cell line Jurkat as well as the mouse B cell line A20 were obtained from American Type Culture Collection. The human T cell line JM was a gift from Dr. M. Fellous (Institut Pasteur, Paris, France). TBCL-10 is a cell line obtained in our laboratory from human tonsil B cells transformed with EBV. All these cells were maintained in RPMI 1640 supplemented with 10% FBS, 2 mM l-glutamine, 100 IU/ml of penicillin, and 100 μg/ml of streptomycin.

Cell lysates, Western blot, and immunoprecipitation studies

To assess the effect of cell adhesion on protein phosphorylation, tonsil endothelial cells were grown to confluence in 35-mm plates (0.8 × 10⁶ cells) for Western blots or in 60-mm plates (1.5 × 10⁶ cells) for immunoprecipitation and coincubated at 4°C over ice with lymphoid cell lines (test cells) for 30 min. After this time, the cocultures were brought to 37°C for varying amounts of time. Test cells were removed by washing with 2 ml ice-cold PBS containing phosphatase inhibitors (10 mM NaF, 5 mM EDTA, 0.4 mM Na₃VO₄, and 10 mM sodium pyrophosphate) and were reserved for analysis. Endothelial cells were washed an additional three or four times with ice-cold PBS until no test cells were detected on the culture plate by light microscopy. For Western blot studies the endothelial cells were lysed directly in the plate with freshly prepared Laemmli sample buffer containing 1 μg/ml leupeptin, 1 μg/ml pepstatin A, 1 mM sodium vanadate, 0.02% aprotinin, and 1 mM PMSF at 100°C. Lysed cells were scraped from the plate, collected in microfuge tubes, and boiled for 5 min. Finally, the lysates were passed several times through a 26-gauge syringe needle and centrifuged for 15 min at 12,000 × g. Proteins were separated by PAGE on 10% gels and transferred to nitrocellulose sheets. Blots were blocked for 1 h at room temperature, incubated with the anti-pTyr Ab overnight at 4°C, and developed using an ECL kit (Amersham, Arlington Heights, IL). To confirm the effect of adhesion on phosphorylation, Ramos cells were preincubated with an anti-α₄ adhesion-blocking mAb at 4°C for 30 min and washed to remove unbound Ab, and the experiment was continued as described above. To determine the effect of protein cross-linking on protein phosphorylation, endothelial cells were preincubated for 30 min at 4°C with 10 μg/ml anti-integrin β₁ or anti-CD40 Abs before incubation for different times (15 s to 30 min) at 37°C with the cross-linking Ab (10 μg/ml). The medium containing unbound Ab was aspirated and discarded, and the adherent endothelial cells were washed twice with ice-cold PBS containing phosphatase inhibitors and treated for analysis as described above.

For the immunoprecipitation studies supernatants of cell lysates were precleared by incubation with protein G-Sepharose coated with normal rabbit serum for 30 min at room temperature. The precleared lysates were incubated with 50 μl protein G-Sepharose coated with primary Ab overnight at 4°C, washed four times with cold immunoprecipitation buffer (1% Triton X-100, 150 mM NaCl, 10 mM Tris-HCl, 1 mM EDTA, 1 mM EGTA, 0.2 mM Na₃VO₄, 0.2 mM PMSF, and 0.5% Nonidet P-40 (pH 7.4)), and boiled for 5 min in 50 μl Laemmli sample buffer. Eluted proteins were separated by SDS-PAGE, immunoblotted with an anti-pTyr mAb, and detected using ECL. Where indicated, immunoblots were stripped by incubation in 0.1 M glycine/HCl, pH 2–3, at room temperature for 15 min, followed by incubation with 1 M NaCl in PBS for 2 min, and were washed twice with 0.1% TBS/Tween 20 for 10 min each time, blocked again as described above, and reimmunostained with the appropriate Ab as indicated.

For quantitation, blots were examined by scanning densitometry using a ScanMaker E3 (Microtek, Arredondo Beach, CA), and data were normalized with NIH Image 1.6 software. Phosphorylated levels of each protein were corrected for protein loading based on the scanning values obtained after reprobing the membranes with the corresponding Ab and were expressed as the level relative to the untreated control value.

RNase protection assay (RPA)

Endothelial cells were grown in P-60 petri dishes to confluence, washed with ice-cold RPMI, and coincubated at 4°C for 30 min with 5 × 10⁶ test cells or mAbs to the β₁ integrin or CD40. After this, cells were incubated with the cross-linking Ab (10 μg/ml) for 1, 3, or 6 h at 37°C. Total RNA was isolated using TRIzol as directed by the manufacturer (Invitrogen, Carlsbad, CA) and was analyzed using the RiboQuant kit (BD PharMingen, San Diego, CA) as directed by the manufacturer. Probes were synthesized using [³²P]UTP with the hCK-4 and hCK-5 template sets from BD PharMingen. Hybridization was performed using 5 μg of RNA from each sample. The protected probes were purified and resolved in 5% SDS-PAGE. Dried gels were exposed to a phosphorimaging screen, and protected fragments were visualized on a Molecular Imager FX (Bio-Rad, Hercules, CA). The identity of each protected fragment was established by analyzing its migration using a standard curve as indicated by the manufacturer. Samples were normalized to the housekeeping genes L32 and GADPH, and protected bands were quantitated by densitometry analysis with the QuantityOne software (Bio-Rad).

Cytokine measurements

For the determination of cytokine production, endothelial cells were grown in 24-well plates to confluence. Before the assay, cells were washed three times with RPMI medium and maintained in medium with 2% FBS during the assay. Supernatants were collected 24 h after stimulation. Assays were performed in triplicate. The amount of IL-8 present in the culture supernatants was measured by a capture ELISA kit (BD PharMingen) according to the manufacturer’s instruction.

TNF bioassay

TNF was quantified by evaluating the cytotoxicity of culture supernatants over actinomycin D-treated murine fibroblast L929. For this, L929 cells were plated in a 96-well microtiter plate at 2 × 10⁴ cells/well and incubated overnight at 37°C in 10% CO₂. The next day the medium was removed, and 50 μl test or control supernatants were added to the cells. Then 50 μl of a stock actinomycin D solution (8 μg/ml) was added to the wells, and incubation was continued overnight at 37°C in 10% CO₂. After this the cells were washed and stained with 50 μl of 0.05% Crystal Violet in 20% ethanol for 10 min at room temperature, washed with water, and allowed to dry. To evaluate cell staining, 100 μl methanol was added to each well, and the OD was read on a microtiter plate reader at 595 nm (25). Each assay was performed in triplicate.

Adhesion of B cell lines induces changes in protein tyrosine phosphorylation on endothelial cells

Recently LeBien and co-workers (6, 7) presented data indicating that upon contact, the lymphoblastoid cell line Ramos is capable of inducing changes in the state of phosphorylation of several proteins in human bone marrow stromal cells. Also, we have previously shown that B cell lines as well as tonsil-derived lymphocytes bind to HUTEC via the α₄β₁ integrin (23). Based on these data and to obtain further understanding on the signaling pathways triggered by B cell adhesion on endothelial cells from lymphoid organs, we investigated whether B cell binding can alter the pattern of Tyr phosphorylation of HUTEC. In these experiments endothelial cells stimulated by lymphoid B cell adhesion consistently showed changes in tyrosine phosphorylation of three main bands of 125, 70, and 45 kDa (see below).

To assess whether the observed changes in protein phosphorylation corresponded to modifications in proteins originating from endothelial cells, lymphoid B cells were independently recovered after adhesion and submitted to the same analysis as endothelial cells. None of the changes in tyrosine phosphorylation found in endothelial cells were detected on lysates of the lymphoid B cells that had been in contact with HUTEC (not shown).

Next we investigated whether the proteins affected in tyrosine phosphorylation corresponded to previously described pp125^FAK (125 kDa), paxillin (70 kDa), and ERK1/2 (45 kDa) (7). Tyrosine phosphorylation was analyzed by immunoblot with an anti-pTyr mAb after immunoprecipitation with the corresponding Ab. The same blots were stripped and reprobed with Abs against pp125^FAK, paxillin, and ERK2 as required and were scanned for quantitation (Fig. 1⇓E).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Contact with Daudi cells modifies the phosphorylation state of pp125^FAK, paxillin, and ERK2 in tonsil endothelial cells. Endothelial cells from human tonsils were left untreated or were incubated with Daudi cells for 0.25, 5, and 30 min; collected at each time point, lysed and immunoprecipitated with anti-pp125^FAK (A), anti-paxillin (B), and anti-ERK2 (C) Abs. The immunoprecipitated proteins were visualized by immunoblotting with anti-phosphotyrosine Ab (upper panels). The amount of each protein loaded in each lane was examined by reprobing the membranes with the corresponding Ab (lower panels). The results shown here are representative of three independent experiments. D, Western blot of HUTEC that were untreated or stimulated for 15 min with Ramos cells or with Ramos cells preincubated with the adhesion-blocking mAb HP2/1. The blotted proteins were visualized with anti-phosphotyrosine Ab. E, Quantitation of pp125^FAK, paxillin, and ERK2 tyrosine phosphorylation at each time point in response to contact with Daudi or Ramos cells.

Fig. 1⇑A shows that the anti-pp125^FAK Ab immunoprecipitated a protein of 125 kDa. This protein was constitutively phosphorylated on Tyr in unstimulated HUTEC and was rapidly and transiently dephosphorylated (<15 s) after B cell stimulation, followed by rephosphorylation, returning to its basal level after 5 min. Although the degree of dephosphorylation was moderate (50%), it was consistent in multiple experiments.

The cytoskeletal protein paxillin migrates on gel electrophoresis as a 70-kDa band and has usually been coupled with the activation and phosphorylation of pp125^FAK (26, 27). Fig. 1⇑B shows that the mAb against paxillin immunoprecipitates from B cell-stimulated HUTEC a protein with a 70-kDa band on SDS-PAGE. The kinetics of dephosphorylation and rephosphorylation of this protein follow the same pattern as those of the anti-pp125^FAK-immunoprecipitated protein.

Finally, we investigated the possibility that the 45-kDa protein could correspond to ERK1 or ERK2. Immunoprecipitates with anti-ERK1 or anti-ERK2 Abs of endothelial cell lysates obtained after activation for varying times with Daudi B cells showed tyrosine phosphorylation only of ERK2 (Fig. 1⇑C) and not of ERK1 (data not shown). Changes in the state of phosphorylation of ERK2 differ from those of pp125^FAK and paxillin on two aspects. Firstly, ERK2 remains dephosphorylated for a longer period of time, and secondly, its rephosphorylation surpasses the initial basal level with a maximum of Tyr phosphorylation at 30 min returning quickly to its basal degree of phosphorylation (not shown).

Finally, to confirm that the observed changes in phosphorylation were produced by Ramos cells, adhesion to the endothelial cells was blocked by preincubating Ramos cells with the anti-α₄ integrin mAb HP2/1 (23, 28). This Ab blocks adhesion of Ramos and Daudi cell to HUTEC by >90% (data not shown) (23). Fig. 1⇑D reveals that no change in the phosphorylation of endothelial cells was detected when Ramos cell adhesion was blocked (only the 15 min point is shown). Quantitation of these experiments by scanning densitometry (Fig. 1⇑E) confirmed these results. Thus, the observed changes in protein phosphorylation of pp125^FAK, paxillin, and ERK2 clearly correspond to changes produced on endothelial cells upon B cell binding.

Cross-linking of β₁ integrin and CD40 lead to changes in protein tyrosine phosphorylation on endothelial cells

It is known that there is information transfer between regular vascular endothelium and lymphocytes, and the involvement of integrins and other adhesion molecules in this process has been demonstrated. Although integrins show no intrinsic protein kinase activity, their engagement induces tyrosine phosphorylation commonly associated with changes in cell shape, secretion of growth factors, and lymphocyte transmigration (13). Furthermore, recent reports have shown that cross-linking of the β₁ integrin present on NK cells induces changes in tyrosine phosphorylation of pp125^FAK and paxillin (15). Since our preparation of tonsil endothelial cells exhibits high expression of the β₁ integrin chain, we evaluated integrin-mediated signaling events in tonsil endothelial cells by cross-linking β₁ integrin.

Fig. 2⇓ shows that cross-linking β₁ integrin on endothelial cells produced changes in the phosphorylation state on the same proteins of 125, 70, and 45 kDa seen during B cell adhesion. However, some differences should be underlined. Firstly, on β₁-activated HUTEC, dephosphorylation is more modest than on B cell-induced activation. Second, rephosphorylation of pp125^FAK and paxillin reached a higher level than the basal state (30 min), persisting for ∼1 h before returning to unstimulated levels (not shown). Another important difference is seen on the kinetics of ERK2 rephosphorylation, since after β₁ integrin cross-linking this protein starts its rephosphorylation earlier (5 min) compared with B cell adhesion (30 min).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Cross-linking of the β₁ integrin chain modifies the phosphorylation state of pp125^FAK, paxillin, and ERK2 in tonsil endothelial cells. Endothelial cells were left untreated or were treated with the anti-β₁ integrin mAb TS2/16.2.1 for 0.25, 5, and 30 min, collected at each time point, and lysed. The lysates were immunoprecipitated with anti-pp125^FAK (A), anti-paxillin (B), and anti-ERK2 (C) Abs and visualized by immunoblotting with anti-phosphotyrosine Ab (upper panels). The amount of each protein loaded in each lane was examined by reprobing the membrane with the corresponding Ab as indicated (lower panels). The results shown here are representative of three independent experiments. D, Quantitation of pp125^FAK, paxillin, and ERK2 tyrosine phosphorylation at each time point in response to contact with Daudi cells.

These changes in protein tyrosine phosphorylation generated after β₁ integrin cross-linking were not detected in control cultures treated with an isotype-matched Ab or with the second Ab alone or in experiments in which the anti-β₁ mAb was not submitted to cross-linking with a second Ab (not shown). Furthermore, to insure that the changes observed in HUTEC phosphorylation were not due to some nonspecific effect of membrane protein cross-linking, we examined the effect of cross-linking of surface MHC class I Ags, a molecule highly expressed in HUTEC. Under this conditions no changes in the state of phosphorylation of the endothelial cells were observed (data not shown).

It has been reported that proinflammatory cytokines can increase CD40 expression, while ligation of CD40 by CD154 (CD40L) on endothelial cells induces the up-regulation of several intercellular adhesion molecules as well as chemokine secretion, thus contributing to the pattern of leukocyte migration and extravasation in inflammation and immunity (29). Since our previous results showed that tonsil-derived endothelial cells express the CD40 molecule in culture (23), we next evaluated the effect of CD40 cross-linking on protein tyrosine phosphorylation of HUTEC by treating these cells with an anti-CD40 mAb (G28-5) for different times. Cross-linking of CD40 also produced time-dependent changes in tyrosine phosphorylation of pp125^FAK, paxillin, and ERK2 (Fig. 3⇓). An initial dephosphorylation (<15 s) was followed by a rephosphorylation at longer times. The kinetics of tyrosine phosphorylation of the 125-, 70-, and 45-kDa proteins were similar to those observed after β₁ integrin cross-linking (Fig. 3⇓). Thus, cross-linking of β₁ integrin or CD40 produced changes in protein phosphorylation similar to those produced by adhesion of B cell lines, although the kinetics were slightly different. In addition to the above-mentioned proteins, cross-linking of CD40 induced after 30 min the strong phosphorylation of an unidentified protein of 90 kDa.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Cross-linking of CD40 modifies the phosphorylation state of pp125^FAK, paxillin, and ERK2 in tonsil endothelial cells. Endothelial cells were left untreated or were treated with the anti-CD40 mAb G28-5 for 0.25, 5, and 30 min; collected at each time point, lysed, and immunoprecipitated with anti-pp125^FAK (A), anti-paxillin (B), and anti-ERK2 (C) Abs. The immunoprecipitated proteins were visualized by immunoblotting with anti-phosphotyrosine Ab (upper panels). The amount of each protein loaded in each lane was examined by reprobing the membrane with the corresponding Ab as indicated (lower panels), and the protein level was used to calculate the relative changes in phosphorylation. The results shown here are representative of three independent experiments. D, Quantitation of pp125^FAK, paxillin, and ERK2 tyrosine phosphorylation at each time point in response to contact with Daudi cells.

B cell (but not T cell) adhesion induces tonsil endothelial cell activation and the production of mRNA for several cytokines

Numerous studies show that activated endothelium produces a variety of factors, including cytokines and growth factors that promote lymphocyte extravasation or attract leukocytes to inflammatory sites (30, 31, 32). To evaluate the effect of cell adhesion on chemokine production by endothelial cells, we coincubated these cells with several cell lines and examined by RPA the mRNA levels of several cytokines. In these experiments we compared untreated HUTEC with cells pretreated with either the human B cell line Daudi, the tonsillar B cell line TBCL-10, the mouse B cell line A20, and human T cell lines, Jurkat and JM. As a positive control for the assay we used HUTEC treated with TNF. The TNF treatment of HUTEC showed protected RNA fragments for several of the cytokines present in the templates (data not shown), confirming previous data indicating that TNF induces a variety of chemokines on the endothelium. All RPAs were quantitated by comparison with the L-32 or GAPDH housekeeping genes (not shown). Interestingly, untreated HUTEC showed a low constitutive level of MCP-1 mRNA, which increased to a maximum 3 h after a change of culture medium (Fig. 4⇓A). On the other hand, the level of the mRNA for this chemokine was markedly increased by the incubation of HUTEC with human B cell lines Daudi and TBCL-10. Maximum levels were obtained after 3 h of coculture, and this was maintained after 6 h of coincubation. In contrast, compared with the untreated controls at 1, 3, and 6 h, human T cell lines and the mouse B cell line A20 do not affect the level of MCP-1 mRNA over the basal levels. Also, we observed that the level of IL-8 mRNA increased after incubation of HUTEC with human B cell lines (Fig. 4⇓A). As shown in Fig. 4⇓A, endothelial cells treated with TBCL-10, in contrast to other human B cell lines tested, induce messages for other cytokines besides MCP-1 and IL-8. Since the pattern of cytokines induced by this cell line was very similar to that observed for TNF, we investigated whether TBCL-10 cells produce TNF or induce HUTEC to produce it. In fact, we demonstrated that TBCL-10 produces a detectable amount of TNF in culture, while Daudi or Ramos cells do not (data not shown). Considering this last point, we found that after 1 h of coculture Daudi cells induced in HUTEC a low level production of MIP-1β mRNA, which remained unaltered for 3 h, disappearing after 6 h of cell contact (Fig. 4⇓A). We further evaluated whether cross-linking of β₁ integrin or CD40 also produced an increase in the mRNA levels of MCP-1 and IL-8. No changes in mRNA levels for these chemokines were observed when HUTEC were cross-linked for different time periods (data not shown) even though this treatment produces patterns of tyrosine phosphorylation similar to B cell adhesion. Thus, adhesion of B cell lines, but not of T cell lines, or cross-linking of β₁ integrin or CD40 induces the production of mRNA for MCP-1 and IL-8 as well as MIP-1β. Moreover, this effect is species specific, since a mouse B cell line is unable to induce the same effect. Since endothelial cells are a source of cytokines involved in hemopoiesis, we used another set of template for RPA assays to investigate this point. As shown in Fig. 4⇓B, adhesion of the B cell Daudi to HUTEC induced an increase in IL-6 and M-CSF mRNA that reached a maximum level after 3 h of incubation with Daudi cells and returned to its basal level after 6 h of activation. On the other hand, cross-linking of β₁ integrin or CD40 produced a rapid increase in IL-6 mRNA with a maximum at 1 h and a return to basal levels at 3 h of coculture.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Expression of human chemokine mRNAs in tonsil endothelial cells induced by B cell contact and membrane cross-linking. Endothelial cells were left untreated (Unt) or were coincubated with the B cell lines Daudi, TBCL-10, and A20 and the T cell lines Jurkat and JM for 1, 3, and 6 h (A) or with anti-β₁ integrin and anti-CD40 mAbs, followed by cross-linking or were coincubated with Daudi cells for 1, 3, and 6 h (B). Total RNA was extracted at each time point, and RPA was performed using hCK-5 (A) or hCK-4 (B) template sets (BD PharMingen) containing the cytokines indicated.

Finally, to verify that induction of chemokine message correlates with secretion, we used a capture ELISA kit to measure the production of IL-8 induced by the adhesion of Ramos and Daudi cells. HUTEC produced 2.76 ± 0.18 and 9.36 ± 0.89 ng/ml of IL-8 after 24 h of culture with Ramos or Daudi cells, respectively, compared with untreated HUTEC that produced 0.75 ± 0.01 ng/ml of IL-8. Supernatants of Daudi and Ramos cells obtained after 24 h in culture were negative with regard to IL-8 secretion (not shown).

Taken together, these results support the idea that coculture of HUTEC with lymphoid B cells causes the functional activation of the tonsil-derived endothelial cells.