Increased Assembly of Clathrin Occurs in Response to Mitogenic Activation of Murine Lymphocytes*

The unassembled (soluble) and assembled (particu- late) pools of clathrin in murine lymphocytes have been separated by centrifugation, and specifically quanti- fied by immunoblotting of cellular extracts with an anticlathrin heavy chain monoclonal antibody. In rest- ing spleen lymphocytes only 25-30% of the total cellular clathrin was found to be present in an assembled form. Upon activation of lymphocytes with B or T cell mitogens (lipopolysaccharide or concanavalin A), the levels of assembled clathrin increased to 60% of the total. These changes in the levels of assembled clathrin were not due to an increase in total cellular clathrin concentration following lymphocyte activation, but rather to changes in the steady state ratio of assembled to unassembled clathrin. The increase in assembled clathrin preceded the expression of transferrin recep- tors, as measured by the cell surface binding of an anti-transferrin receptor monoclonal antibody, and maxi- mal DNA synthesis, indicating that clathrin assembly occurs early after lymphocyte activation and precedes cell division. Immunofluorescence analysis of activated lymphocytes with an anti-clathrin heavy chain mono- clonal antibody revealed a punctuate staining pattern characteristic of coated pits and vesicles. Activated B lymphocytes displayed particularly prominent staining in the perinuclear region compared to T cells, suggesting that clathrin assembly may be important for B cell functions such as immunoglobulin

The unassembled (soluble) and assembled (particulate) pools of clathrin in murine lymphocytes have been separated by centrifugation, and specifically quantified by immunoblotting of cellular extracts with an anticlathrin heavy chain monoclonal antibody. In resting spleen lymphocytes only 25-30% of the total cellular clathrin was found to be present in an assembled form. Upon activation of lymphocytes with B or T cell mitogens (lipopolysaccharide or concanavalin A), the levels of assembled clathrin increased to 60% of the total. These changes in the levels of assembled clathrin were not due to an increase in total cellular clathrin concentration following lymphocyte activation, but rather to changes in the steady state ratio of assembled to unassembled clathrin. The increase in assembled clathrin preceded the expression of transferrin receptors, as measured by the cell surface binding of an antitransferrin receptor monoclonal antibody, and maximal DNA synthesis, indicating that clathrin assembly occurs early after lymphocyte activation and precedes cell division. Immunofluorescence analysis of activated lymphocytes with an anti-clathrin heavy chain monoclonal antibody revealed a punctuate staining pattern characteristic of coated pits and vesicles. Activated B lymphocytes displayed particularly prominent staining in the perinuclear region compared to T cells, suggesting that clathrin assembly may be important for B cell functions such as immunoglobulin synthesis or secretion. These results suggest that in lymphocytes, clathrin assembly is a dynamic process that is triggered by mitogenic stimuli.
The formation of coated pits and vesicles is considered to be an essential step in the internalization of ligands through receptor-mediated endocytosis, and in the process of intracellular transport and protein sorting. The formation of these structures involves the continuous assembly of clathrin and other polypeptide components onto plasma membranes, as well as the continuous disassembly of such components from coated vesicles (reviewed by Brodsky, 1988). These active processes result in the formation of two cellular pools of clathrin, one consisting of soluble triskelions found in cyto-* This work was supported by Grant DK40330 from the National Institutes of Health (to S. C.), a grant from the Eastern Pennsylvania Chapter of the Arthritis Foundation (to J. S. M.), and by a Grant from the Lucille Markey Charitable Trust. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. solic fractions of cells and tissues (Bruder and Widenmann, 1986;Goud et al., 1985;Moore et al., 1987), and another consisting of polymerized triskelions found stably assembled onto cellular membrane fractions. It has been reported that whereas the total amount of clathrin (expressed as the percentage of total cellular protein) is constant among different cell lines, the ratio of assembled to unassembled clathrin varies, being higher in cells that are endocytically or exocytically active (Goud et al., 1985).
One of the most rapid responses of cells to several hormones and growth stimuli is a change in receptor-mediated endocytic activity, resulting in an increased uptake of diverse nutrients and macromolecules. For example, within minutes after stimulation of cells by insulin, insulin-like growth factor I or epidermal growth factor Wiley and Kaplan, 1984), a rapid increase in the cell surface concentration of transferrin receptors and a concomitant stimulation of iron uptake is observed. In addition, an increase in the number of coated pits in the plasma membrane of PC-12 pheochromocytoma cells has been shown to occur within minutes of stimulation by nerve growth factor (Connoly et al., 1984). More delayed effects of growth factors on endocytic processes also occur, such as the induction by interleukin 2 of transferrin receptors and iron uptake in mitogenically activated human lymphocytes (Neckers and Cossman, 1983).
The observation that receptor-mediated endocytic processes can be regulated by growth factors made it interesting to examine the possibility that the process of clathrin assembly might be affected in response to these stimuli. Lymphocytes are well suited as a model system in which to examine this possibility, in that resting cells express few if any growth factor receptors, and display a low ratio of assembled to unassembled clathrin (Goud et al., 1985). However, exposure of these cells to antigens or mitogens rapidly triggers the expression of various receptors, such as those for insulin, interleukin 2, and transferrin (Neckers and Cossman, 1983;Helderman and Strom, 1979;Kronke et al., 1985). In this study we analyzed the state of clathrin assembly during the mitogenic induction of cellular proliferation in murine lymphocytes.

EXPERIMENTAL PROCEDURES
Cell Isolation and Culture-Lymphocytes were prepared from the spleens of 4-12-week-old female (BALB/c BYJ mice (Jackson Laboratories, Bar Harbor, ME). Spleens were excised under sterile conditions and gently teased in a medium composed of RPMI-1640 supplemented with 10% heat-inactivated fetal calf serum (Hyclone), 10 mM each of L-glutamine, nonessential amino acids, and sodium pyruvate, 25 mM Hepes,' and 50 GM 2-mercaptoethanol (Roman et  al., 1988). Cells were cultured a t a density of 2.5 X lo6 lymphocytes/ ml in 24-well multiwell plates (Corning) a t 37 "C in 5% CO,. Lipopolysaccharide (LPS) or concanavalin A (ConA) (Difco) were added to the cultures at a final concentration of 10 pg/ml. In some experiments, erythrocytes were removed from the spleen cell suspensions by hypotonic shock lysis (Moore and Calkins, 1985). The results of experiments done with erythrocyte-depleted cultures were similar to those obtained when these cells were allowed to remain throughout the incubations, which is consistent with the observation that erythrocytes are practically devoid of clathrin (Goud et al., 1985).
B-and T-lymphocytes were separated by direct panning, using 100-mm polystyrene culture dishes coated with affinity purified goat anti-mouse immunoglobulin (Zymed). Non-adherent T cells were collected, and B cells were recovered by strongly forcing a stream of medium over the plate surface (Mage et al., 1977). Purified B-and Tlymphocytes were then cultured in the medium described above a t a density of 2.5 X lo6 cells/ml with either LPS or ConA a t a concentration of 10 pg/ml. Cell viability was routinely 80-90%, as assayed by trypan blue exclusion.
Cell Fractionation-The separation of the pools of assembled and unassembled clathrin was performed essentially as described by Goud et al. (1985) with some modifications. Lymphocytes (1-5 X 10') were harvested by centrifugation in 15-ml conical tubes, the medium was aspirated, and the cell pellets were lysed by vortexing in 500 pl of a buffer composed of 20 mM MES, pH 6.8, 2.5 mM MgC12, 2.5 mM EGTA, 1 mM 1-10-phenanthroline, 1 mM benzamidine, 1 mM phenylmethylsulfonyl fluoride, 10 pM leupeptin, 15 p M antipain, 20 pg/ml aprotinin, 10 pg/ml chymostatin, and 0.2% Triton X-100 a t 4 "C. The extracts were then centrifuged a t 350,000 X g for 15 min in a Beckman TJ-100 table top ultracentrifuge. The supernatants were collected and used as the source of unassembled clathrin. The cell pellets were resuspended in 500 pl of a solution composed of 20 mM TAPS, pH 9.0, and the protease inhibitors described above, and homogenized by five passes in a 2-ml Potter-Elvehjem tissue grinder. After 20-30 min of incubation on ice, the suspensions were centrifuged and the supernatants used as the source of assembled clathrin. In preliminary experiments, the remaining pellet was dissolved and analyzed for the presence of clathrin by immunoblotting. Virtually no remainingclathrin could be detected in this fraction (data not shown).
Immunoblotting-Aliquots of the assembled and unassembled clathrin pools were boiled for 2 min in electrophoresis sample buffer (Laemmli, 1970) a t a final concentration of 2% sodium dodecyl sulfate and 10 mM dithiothreitol, separated on 6% polyacrylamide gels, and electrophoretically transferred onto nitrocellulose paper (Towbin et al., 1979). After blocking in 5% bovine serum albumin, the blots were incubated overnight with anti-clathrin heavy chain monoclonal antibody Chc 5.9 (Boehringer Mannheim) a t a final concentration of 2.5 pglml. The blots were then washed and developed with a polyclonal goat anti-mouse immunoglobulin coupled to alkaline phosphatase (Promega). Color development was performed according to the manufacturers instructions. The intensity of the colored bands was quantified using a LKB Ultroscan XL laser densitometer. Standard curves were constructed using purified clathrin from rat liver coated vesicles prepared essentially as described by Campbell et al. (1984).
Anti-transferrin Receptor IgG Binding-Anti-mouse transferrin receptor monoclonal antibody was purified by batch extraction of culture supernatants of TIB 219 hybridomas (ATCC), using 40-pm Bakerbond ABx resin (Baker). Lymphocytes were harvested, and resuspended to a final concentration of 5 X lo5 cells/ml in ice-cold Krebs-Ringer buffered with Hepes, pH 7.4, supplemented with 3% bovine serum albumin. Aliquots of the cell suspensions (100 pl) were incubated with 20 pg/ml anti-transferrin receptor monoclonal antibody for 2 h a t 5 "C. The cells were washed twice by resuspension in 1 ml of ice-cold buffer and centrifugation for 1 min in an Eppendorf microfuge, and incubated with 10 pg/ml of a mouse anti-rat polyclonal antibody (Boehringer Mannheim) for 30 min. After two washes, 0.2 pCi of lZ5I-protein A (Du Pont-New England Nuclear) were added, and incubations were continued for 30 min. At this time, cells were washed three times, transferred to clean microfuge tubes, centrifuged, and the radioactivity associated to the pellets was quantified by ycounting.
Tritiated Thymidine Incorporation-Lymphocytes were cultured a t a density of 2.5 X lo6 cells/ml in 96-well multiwell dishes. Mitogens were added a t a final concentration of 10 pg/ml, and at the indicated times of culture, 1 pCi of [methyl-'Hlthymidine (2 Ci/mmol, Du Pont-New England Nuclear) was added to each well. After 8 h, cells were harvested onto filter strips and washed with saline, 5% trichloroacetic acid, and absolute methanol. The filter discs were dried and radioactivity associated was measured in a 8-scintillation counter.
Immunofluorescence Microscopy-Purified B-or T-lymphocytes were cytocentrifuged onto microslides and fixed with acetone for 30 min at -20 "C. The excess acetone was allowed to evaporate a t room temperature, and the cells were overlaid with fetal calf serum supplemented with 100 pg/ml of a polyvalent goat anti-mouse immunoglobulin (Zymed) to block nonspecific binding. The slide was washed briefly in phosphate-buffered saline, and the preparation was overlaid with 10 pg/ml of monoclonal antibody Chc 5.9. After 60 min of incubation, the slide was washed three times with phosphate-buffered saline, and incubated with 5 pg/ml of an affinity purified goat antimouse polyclonal antibody coupled to fluorescein isothiocyanate (TAGO). After 30 min of incubation a t 37 "C, the slides were washed in five changes of phosphate-buffered saline over a 45-min period a t room temperature. The preparation was overlaid with a solution of 90% glycerol, 10% Tris-buffered saline, and examined on a Zeiss standard WL microscope equipped with a filter for fluorescein. No specific fluorescence staining was observed in experiments where the anti-clathrin antibody was omitted.

RESULTS AND DISCUSSION
Clathrin Assembly in Resting or Mitogen-stimulated Murine Splenic Lymphocytes-To quantify specifically the amount of clathrin present in lymphocyte extracts, an immunoblotting assay using a monoclonal antibody specific for the heavy chain of clathrin (Chc 5.9, Boehringer Mannheim) was employed. This antibody can detect clathrin heavy chain present in crude cell and tissue extracts from diverse species (Bruder and Widenmann, 1986). An immunoblot of rat liver-coated vesicles probed with antibody Chc 5.9 followed by a goat antimouse immunoglobulin coupled to alkaline phosphatase is shown in Fig. 1. Intense staining of a polypeptide of approximately 180 kDa, corresponding to the molecular weight of clathrin heavy chain, was detected. Quantification of the intensity of the colored band by laser densitometry revealed Immunoblotting of clathrin with monoclonal antibody Chc 5.9. Rat liver-coated vesicles were electrophoresed on a 6% polyacrylamide gel and transferred onto nitrocellulose paper. Immunoblotting was performed with monoclonal antibody Chc 5.9 followed by a goat anti-mouse polyclonal antibody coupled to alkaline phosphatase. The positions of prestained molecular weight markers on the nitrocellulose blot are indicated. The intensity of the colored bands was analyzed by laser densitometry, and the absorbance values plotted against the coated vesicle protein concentration run in each lane. that this technique was sensitive to small amounts of clathrin, and linear over a broad concentration range.
The pools of assembled and unassembled clathrin from resting lymphocytes or from cells that had been stimulated for 48 h with B cell specific (LPS) or T cell specific (ConA) mitogens were separated by centrifugation and aliquots of the cell extracts were analyzed by immunoblotting. In unstimulated cells, clathrin concentration in the unassembled pool was found to be 2-%fold higher than in the assembled pool (Fig. 2). Upon stimulation of lymphocytes with mitogens the proportion of clathrin in the assembled pool increased significantly. Analysis of the time course of this effect revealed a significant increase in assembled clathrin by 24 h after exposure to LPS or ConA. By 48 h of incubation the proportion of assembled clathrin increased to approximately 60% of the total. The effect of LPS appeared to slightly precede the effect of ConA, being almost maximal after 24 h.
We investigated whether the changes in the ratio of assembled to unassembled clathrin were accompanied by an increase in the total cellular clathrin concentration. In control cells, the concentration of clathrin was approximately 300 ng/106 cells, equivalent to approximately 3 X IO5 triskelions/cell, similar to that previously determined by enzyme-linked immunoadsorption assays (Goud et al., 1985;Doxey et al., 1987). No significant differences in clathrin concentration (assembled plus unassembled) were detected between control and activated cells during the first 48 h after stimulation with LPS or ConA. After 72 h of incubation with LPS, but not with ConA, the abundance of cellular clathrin increased to approximately 600 ng/106 cells, suggesting that increased synthesis of this protein may be required for one or several B-lymphocyte specific functions. These results suggest that the increase in the levels of assembled clathrin observed upon mitogenic stimulation do not result from the induction of clathrin synthesis, but rather from a change in the steady state ratio of assembled to unassembled clathrin.
Kinetics of Chthrin Assembly, Thymidine Incorporation, and Transferrin Receptor Expression-Various growth factors and macromolecules which are internalized through receptormediated endocytosis are required for B-and T-lymphocyte proliferation (Iscove and Melchers, 1978;Mendelsohn et al., 1983). If the observed increase in clathrin assembly is required to mediate this process, it should precede or parallel maximal cell proliferation. Fig. 3, left panel, shows that under our conditions, stimulation by ConA of thymidine incorporation into DNA was detectable after 24 h, maximal at 48 h, and decreased thereafter. The mitogenic effect of LPS was slightly weaker, being detectable only after 48 h. A comparison with the data in Fig. 2 indicates that clathrin assembly parallels (in the case of ConA stimulation), or precedes (in the case of LPS stimulation), the maximal effect of these mitogens on cell proliferation, consistent with the notion that clathrin assembly is triggered a t early stages of lymphocyte activation preceding cell division.
The molecular mechanisms involved in mediating the assembly of clathrin onto membranes are not known. However, it has recently been proposed that the cytoplasmic domain of the transferrin receptor may serve as an assembly site for coated pit formation at the plasma membrane (Iacoppeta et al., 1988). Thus, the induction of transferrin receptors in mitogenically activated lymphocytes might trigger clathrin assembly in these cells. T o explore this possibility, we compared the kinetics of cell surface transferrin receptor expression to the kinetics of clathrin assembly in response to LPS or ConA. Fig. 3, right panel, shows that specific binding of an anti-transferrin receptor monoclonal antibody (Lesley et al., 1984) to the cell surface of ConA-stimulated cells was observed after 24 h, was maximal a t 48 h, and remained high through 72 h of culture. Thus, clathrin assembly closely parallels transferrin receptor expression in ConA-stimulated cells, consistent with the notion that this receptor may be an assembly site for clathrin. Anti-transferrin receptor antibody binding to the cell surface of LPS-stimulated cells was detected after 48 h, but was only 20% of that detected after ConA stimulation (Fig. 3, right panel). However, clathrin assembly in response to LPS was almost maximal after 24 h and quantitatively similar to that observed in response to ConA (Fig. 2). This kinetic and quantitative lack of correlation between clathrin assembly and transferrin receptor expression during B-lymphocyte activation suggests that at least in these cells, additional factors besides the cytoplasmic domain of the transferrin receptor may be involved in clathrin assembly.
Immunofluorescence Analysis of Clathrin in Purified Band T-lymphocytes-To further understand the nature of the changes in clathrin assembly that occur during lymphocyte activation, indirect immunofluorescence microscopy of clathrin was performed. Fig. 4 (top panels) shows examples of the immunofluorescence micrographs obtained from isolated Bor T-lymphocytes cultured for 72 h in the presence of LPS or ConA, respectively. In both cell types a punctuate fluorescence pattern extending from the perinuclear region to the cell periphery was detected. This staining pattern is characteristic of clathrin-coated pits and vesicles observed in fibroblasts and other cell types (Willingham et al., 1981;Bruder and Widenmann, 1986). Interestingly, many activated B-cells displayed intense staining in the juxtanuclear region. In contrast to activated lymphocytes, resting B-or T-cells (Fig. 4, lower panels) displayed diffuse, weak staining with the anti-clathrin monoclonal antibody.
These results are consistent with the biochemical results shown in Fig. 2, which indicate that in resting lymphocytes the majority of the clathrin is in a cytoplasmic unassembled form, which would be difficult to detect by immunofluorescence microscopy. In mitogenically activated cells, clathrin is assembled and concentrated in coated pits or vesicles and is thus more readily detectable by this procedure. The enhanced fluorescence in the juxtanuclear region of activated B-cells possibly corresponds to the Golgi apparatus. Although the function of clathrin in the Golgi is not completely defined (Doxey et al., 1987;Wheland et al., 1982;Orci et al., 1986), it

FIG. 4.
Immunofluorescence analysis of the distribution of clathrin in Band T-lymphocytes. Lymphocytes were obtained from spleens of RALH/c mice. Band T-cells were separated by panning, using 100-mm culture dishes coated with rabbit anti-mouse Ig. Purified Bor T-lymphocytes were cultured at a density of 2.5 X lo6 cells/ml in the absence (lower panels) or presence of LPS or ConA, respectively. After 72 h of culture, cells were centrifuged onto microslides, fixed in acetone, and stained with anti-clathrin heavy chain monoclonal antibody Chc 5.9 followed by an affinity purified goat anti-mouse Ig coupled to fluorescein isothiocyanate. Note the bright staining in the juxtanuclear region of mitogen-stimulated Bcells (upper left), and the punctuate staining that extends from the perinuclear region to the cell membrane in activated T-cells (upper right). Very weak, diffuse staining was detected in non-activated lymphocytes (lower panels) ( X 630). Clathrin Assembly during Lymphocyte Activation is possible that in these cells it may be important for antibody synthesis and secretion (Kinnon and Owen, 1983).
In addition to clathrin, coated pits and vesicles are composed of assembly polypeptides (Zaremba and Keen, 1983;Robinson and Pearse, 1986;Ahle et al., 1988) which coassemble with clathrin in in vitro systems (Moore et al., 1987), and promote the formation of clathrin baskets (Zaremba and Keen, 1983). The HA-I assembly polypeptides appear to be associated specifically with clathrin-coated membranes of the Golgi, whereas the HA-I1 peptides are associated with coated pits and endocytic vesicles (Robinson and Pearse, 1986;Ahle et al., 1988), suggesting that clathrin assembly onto specific membrane structures is directed by these polypeptides. Thus, clathrin assembly in lymphocytes may be due to the expression of these polypeptides, and the differences in the pattern of clathrin assembly between Band T-lymphocytes might be explained by differences in the levels of expression of the two classes of assembly proteins.
In summary, the results presented in this paper indicate that mitogenic activation of resting lymphocytes is accompanied by an increase in the levels of assembled clathrin. This increase is not due to an increased cellular clathrin concentration, but rather to a change in the ratio of unassembled to assembled clathrin. These data support the notion that the assembly and disassembly of clathrin-coated membranes is a dynamic process, which can be rapidly modulated by physiological stimuli (Salisbury et al., 1980;Takemura et al., 1986). In the resting lymphocyte, rapid recruitment of clathrin from an unassembled pool after mitogenic or antigenic stimulation may enable the cell to rapidly internalize growth factors and macromolecules necessary for proliferation. Further studies in these cells may provide insight into the importance of clathrin assembly and receptor endocytosis in lymphocyte activation, and into the biochemical mechanisms that regulate clathrin assembly in intact cells.