CELLULAR

IMMUNOLOGY

137, 5 14-528 (1991)

Actin Polymerization in Murine 6 Lymphocytes Is Stimulated by Cytochalasin D but Not by Anti-immunoglobulin J.A.WILDERANDR.F.ASHMAN* Departments of Microbiology and *Internal Medicine, University of Iowa College of Medicine and Veterans Administration Hospitals, Iowa City, Iowa 52242 Received January 23, 1991: accepted May 30, 1991 One might predict that cytochalasin D, which slows polymerization of actin in solution and which inhibits actin-containing microfilament function in live B lymphocytes, would also prevent actin polymerization in these cells. However, we have used the NBD-Phallacidin flow cytometric assay for F-actin and the DNase I inhibition assay for G-actin to demonstrate that cytochalasin D (at 20 rg/ml and higher) stimulates actin polymerization in murine B lymphocytes within the first 30 set of exposure. A similar response was seen in human neutrophils. Actin polymerization induced in neutrophils by chemotactic peptides has been linked to activation of the polyphosphoinositide-calcium increase-protein kinase C signal transduction pathway. As B lymphocytes also transduce signals using this pathway, we investigated whether cytochalasin D induced actin polymerization by activating this pathway. Cytochalasin D and ionomycin both stimulated a rapid increase in internal calcium (by 1 min) in the B cell which was inhibitable by EGTA, implicating calcium influx. Ionomycin also induced actin polymerization, detectable later, by 10 min. EGTA blocked the ionomycin-induced actin polymerization, but not that induced by cytochalasin D. Cytochalasin D-induced actin polymerization was not associated with detectable hydrolysis of polyphosphoinositides, nor was it inhibited by H7 (a protein kinase C inhibitor) or by HA1004 (an inhibitor of cyclic nucleotide-dependent kinases). Furthermore, anti-immunoglobulin antibodies, which stimulate B lymphocytes through the polyphosphoinositide hydrolysiscalcium increase-protein kinase C pathway, failed to induce actin polymerization in these cells. These antibodies did, however, stimulate the cells to perform activities that involve actin-containing microfilaments. Other primary activators of B lymphocytes (dextran sulfate, PMA, and LPS) and a panel of lymphokines previously shown to enhance B lymphocyte activation (IL-l, IL-2. IL-4, IL-5) were also screened in the F-actin assay and no evidence for actin polymerization was found. We conclude that the actin polymerization response to cytochalasin D in the B cell does not involve the polyphosphoinositide hydrolysis-calcium increase-protein kinase C pathway, nor does it depend on cyclic nucleotide-dependent kinases. Furthermore, our studies failed to provide any evidence that early actin polymerization occurs in murine B lymphocyte activation. o 1991 Academic

Press, Inc.

INTRODUCTION Actin, one of the most abundant and highly conservedproteins in nature, is involved in many cell functions such as motility, secretion, capping of cell surface receptors, and cell division. Actin polymerization, the processof converting monomeric G-actin into polymeric F-actin, accompanies the cell motility response induced in polymorphonuclear cells (PMNs)’ by chemotactic peptides (l-3). Chemotactic peptides, such I Abbreviations used: PMA, phorbol my&ate acetate; PK, protein kinase; PMN, polymorphonuclear leukocyte; ATP, adenosine triphosphate; MFI, mean fluorescence intensity: PPI, polyphosphoinositide; IL, interleukin; BAPTA-AM, 1.2-bis(o-aminophenoxy)ethane-N,N,N’,N’-tetracetic acid acetoxymethylester. 514 0008-8749/91 $3.00 Copyright 0 1991 by Academic Press, Inc. All ngbts of reproduction in any form reserved.

B CELL ACTIN

POLYMERIZATION

515

as f-met-leu-phe (fmlp), also stimulate PMNs to hydrolyze polyphosphoinositides (4) increase their intracellular calcium levels (1, 5, 6) and activate protein kinase C (6). Intracellular second messengerssuch as inositol phosphates and calcium can affect actin equilibrium by interacting with actin-binding proteins (7). In resting B lymphocytes, anti-immunoglobulin (anti-Ig) reagentsstimulate changes in inositol phosphate (8) and calcium levels (9) much like chemotactic peptides do in PMNs. They also elicit a brief burst of motility, as well as capping and endocytosis of surface Ig, functions requiring microfilament contraction (10). The ability of anti-Ig to stimulate actin polymerization in normal mouse B cells, however, has not yet been investigated. Another molecule reported to engagethe same signal transduction pathway as anti-Ig is the fungal metabolite cytochalasin D (11, 12). Widely used as an inhibitor of microfilament function, cytochalasin D, at concentrations above 2.5 pg/ ml, inhibits actin polymerization in solution by binding to and slowing monomer addition to the fast-growing barbed end of actin filaments (13). Surprisingly, in the current study, cytochalasin D actually stimulated actin polymerization in live murine B lymphocytes, whereas anti-Ig did not. The optimal concentration of cytochalasin D wasgreater for actin polymerization than for comitogenic activity with anti-Ig reagents ( 14) or lectins (15) dissociating these two responses. Although we can demonstrate such large dosesof cytochalasin D stimulated calcium influx, cytochalasin D-induced actin polymerization did not depend upon intracellular calcium fluxes or PKC activation. In contrast, ionomycin stimulated actin polymerization by a mechanism requiring calcium influx. On the basis of thesedata, we discuss possible mechanisms for cytochalasin D- and ionomycin-induced actin polymerization. MATERIALS AND METHODS Female (C57BL/6 X DBA/2)Fl (B6D2FI) mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Antibodies F(ab’)2 fractions of goat anti-mouse IgM reagents (GaMIg) were purchased from Tago Immunologicals (Lot No. 420301), Accurate Chemicals (Lot No. E3419), and Immunosearch (Lot No. 04-514-37). F(ab’), rabbit anti-mouse Ig (heavy and light chain specific) was purchased from Cappel (RaMIg, Lot No. 29820). Mouse B Cell Preparation B lymphocytes were purified from mouse splenocyte suspensions by complement lysis with anti-Thy 1.2 (from the HO1 3.3.9 hybridoma), sufficient to render the B cells unresponsive to concanavalin A (Sigma Chemical Co., St. Louis, MO) as measured by tritiated thymidine incorporation (data not shown). Preparation of Human Polymorphonuclear Cells Blood from healthy volunteers was heparinized and mixed with an equal volume of 3% dextran. After red cell sedimentation at room temperature for 18 min, the supernate was centrifuged at 2000 rpm for 10 min. The resuspended cell pellet was layered onto 10 ml Ficoll-Hypaque at 1400 rpm for 40 min using no brake. The

5 16

WILDER

AND

ASHMAN

resulting cell pellet was subjected to hypotonic lysis to remove residual red blood cells and resuspended in medium.

Assessmentof F-Actin Content IB lymphocytes, resuspended in RPM1 1640 medium at 5-10 X 106/ml, were prewarmed to 37°C for 15 min before various stimuli were added. At 0.5, 1, 10, and 60 min after the stimuli were added, l-2 X lo6 cells were removed and fixed for 15 min at 2 1“C with 1 ml 3.7% formaldehyde containing 100 pg/ml lysophosphatidylcholine (Sigma Chemical Co.). Then the cells were stained with a I/ 160 final dilution of NBDPhallacidin, a fluorescent dye which specifically binds F-actin but not G-actin (16). Where indicated, the cells were preincubated with 100 PLM H76 or HA 1004 (Calbiothem, La Jolla, CA) for 90 min at 37°C before the addition of stimuli. The cells were routinely stored overnight at 4°C before analysis by flow cytometry.

Assessmentof G-Actin Content The relative G-actin content of B cells was measured using the DNase I inhibition assay originally described by Blikstad et al. ( 17). Briefly, 5 X lo6 cells were incubated for 15 min at 37°C before stimuli were added. At 1, 10, and 60 min after stimulus addition, 4 X lo6 cells were removed and pelleted. The cell pellet was lysed with 100 ~1 of lysis buffer (5 mA4 potassium phosphate, 150 mA4 NaCl, 2 mM MgC& , 0.2 mA4 ATP, 0.2 mM dithioerythritol, 0.5% Triton X, 0.0 1 M phenylmethylsulfonyl fluoride, pH 7.6) and 25 ~1 of the cell lysate was then mixed with 3 ml of calf thymus DNA (40 pug/ml in DNA buffer; Sigma Chemical Co.), 15 ~1 of DNase (99.2 U/ml in DNase buffer; Worthington Biochemical Co., Freehold, NJ), and 15 ~1 of G-actin buffer. DNA buffer consisted of 0.1 M Tris-HCl, 4 rnA4 MgS04, 18 mM CaC&, pH 7.5. DNase buffer consisted of 50 mA4 Tris-HCl, 0.01 M phenylmethylsulfonyl fluoride, 0.3 mM CaC&, pH 7.5. G-actin buffer consisted of 5 rnA4 potassium phosphate, 150 r&l NaCl, 2 mM MgC&, 0.2 mA4 ATP, 0.2 mM dithioerythritol, 0.5% Triton X, pH 7.6. The rise in OD260 as DNase digested the DNA was measured on a scanning spectrophotometer (Hitachi U2000) over a 4-min time span. The slope of the linear portion of the resulting curve was calculated for each sample and compared to the slope generated by a mixture containing only DNA, DNase I, and G-actin buffer. The decreased digestion rate resulting from G-actin in the cell lysate was determined in the following manner: % Inhibition

= 1-

ODz60/min of mixture containing lysate OD260/min of mixture containing no lysate ’

A standard curve was generated at the time of assay by adding a known amount of pure G-actin (Sigma Chemical Co.) to the DNA and DNase mixture.

Measurement of [3Hj Thymidine Incorporation II cells were resuspended at 1 X 106/ml in RPM1 complete medium (RPM1 1640, GIBCO Laboratories), 5% fetal calf serum (KC Biologicals, Kansas City, MO), 0.01 A4 Hepes, 0.1 r&4 nonessential amino acids, 2 r&4 L-glutamine, 0.0 1 mg/ml gentamicin sulfate, 1 mA4 sodium pyruvate, and 5 X 10m5M 2-mercaptoethanol. Aliquots (200 ~1) were incubated in 96-well flat-bottomed plates with stimuli in 5% CO2 at 37°C for 72 h with 0.5 &i[3H]thymidine/well (Amersham, Arlington Heights, IL)

B CELL ACTIN

POLYMERIZATION

517

added during the last 6 hr. The cells were harvested with a MASH cell harvester (Skatron, Sterling, VA) onto glassfiber discs using saline and 10%trichloroacetic acid washes.Radioactivity on the discs was measured in a Beckman LS 7500 liquid scintillation counter. Measurement of Polyphosphoinositide Hydrolysis This procedure was performed as originally described by Bijsterbosch et al. (18). Briefly, B cells were washed once in inositol-free RPM1 (GIBCO Laboratories), supplemented with 5% inositol-free fetal calf serum, incubated at 25 X 106/ml at 37°C for 4 hr in the presence of myo-[3H]inositol (1 &i/lo6 cells; Amersham), washed three times in inositol-free RPM1 containing 0.0 1 A4 Hepes, 2 mM L-glutamine, and 0.01 M LiC&, and resuspended at 10 X 106/ml in this buffer. Aliquots of 0.3 ml were prewarmed to 37°C for 30 min before the cells were stimulated in triplicate for 60 min. The stimulation was stopped by adding 1.5 ml CHC13:methanol (1:2) and 0.6 ml HZ0 to each aliquot. An additional 0.5 ml CHClj was added and the cells were stored at -20°C overnight. The following day total inositol phosphateswere extracted, eluted from Dowex ACA-200 with buffer, and counted in a liquid scintillation counter. Measurement of Intracellular Calcium B cells, at 20 X 106/3 ml of RPM1 containing 8 mM EGTA, were incubated for 60 min at 37°C with 3 pLMIndo 1-acetomethoxy ester (19) and washed twice in the same medium. Aliquots of 2 X 1O6in 1 ml were stimulated with various reagentsimmediately before flow cytometric analysis. Where indicated, the Indo 1 loading was preceded by a 30-min incubation with 10 pike 1,2-bis(o-aminophenoxy)ethane-N,N,N’,N’-tetraacetic acid acetoxymethylester (BAPTA-AM; Molecular Probes, Eugene, OR) at 37°C. Measurement of Capping and Endocytosis B cells (5 X 106/ml) were equilibrated for 15 min at 37°C stimulated for 10 or 20 min with 25 pg/ml F(ab’)2 goat anti-mouse IgM (Tago Immunologicals), fixed with 3.7% paraformaldehyde, stained with rabbit anti-goat Ig-FITC (Jackson Laboratories) for 20 min on ice, and washed twice before assayon a Leitz Orthoplan fluorescence microscope. Flow Cytometry F-actin was measured by NBD-Phallacidin fluorescence on a Becton Dickinson dual laser fluorescence-activatedcell sorter (FACS IV, Mountain View, CA) or a Coulter Epics 753 (Hialeah, FL), with 400 mW excitation at 488 nm from a 5-W argon ion laser (Coherent, Palo Alto, CA). Emission was measured with a 525 band pass filter. Forward and orthogonal scatter were used to gate out debris. Data acquired on the FACS IV were collected in listmode on a VAX 1l/750 (Digital Equipment Corp., Maynard, MA) and analyzed with Electric Desk software written by Wayne Moore at Stanford University. Data acquired on the Epics 753 were collected in histogram mode and subsequently analyzed with Coulter’s Easy2 software running on an IBM PC/AT compatible computer. Data are expressedas mean fluorescence intensity on a linear scale. Cells loaded with Indo 1 were excited with 100 mW in the uv range (351 to 364 nm) from a 5-W argon ion laser. The emission spectrum of Indo 1 bound by calcium

518

WILDER

AND

ASHMAN

was measuredusing a 395125nm band passfilter, while the unbound dye was measured using a 525/20 nm band pass filter. Flow cytometry data were collected in listmode anId analyzed using Coulter’s Gateway software. RESULTS Cytochalasin D Stimulates B Cell Actin Polymerization Cytochalasin D stimulated large increases in B lymphocyte F-actin content when used at concentrations of 20 pg/ml or more (Fig. 1). Increases in F-actin occurred early (30 set after stimulus application) and were sustained as long as 60 min (data not shown). At 20 pg/ml cytochalasin D also stimulated decreasesin B cell G-actin content as measured by DNase I inhibition (Table 1). Although Table 1 depicts one representative experiment, the ability of cytochalasin D to lower B cell G-actin content has been confirmed in all of 16 others. The fact that cytochalasin D stimulated a simultaneous increase in F-actin and decreasein G-actin suggeststhat this reagent truly stimulates B cell actin polymerization in living B cells. Cytochalasin D Stimulates Actin Polymerization in Human PA4Ns Cytochalasin D (20 pg/ml) induced an increase in F-actin content in human PMNs as early as that induced by the chemotactic peptide fmlp: an increase of 82 channels in mean fluorescence intensity (MFI) at 1 min. By comparison, 100 ruJ4fmlp induced an MFI increase of 4 1 channels. Induction of actin polymerization by fmlp in PMNs has been previously reported (l-3). Both these reagents also stimulated a decreasein

120 110 100 90 80 70 2 -3

60 50 40 30 20 10 0:

-

is 48

pg/ml Cytochalasin

D

FIG. I. Large doses of cytochalasin D stimulated increases in B cell F-actin content by 1 min. NBDPhallacidin binding was measured to assess the relative F-actin content of B cells stimulated with various doses of cytochalasin D. The data represent the means (k standard deviation) of five experiments. AMFI = m’ean fluorescence intensity of stimulated cells - mean fluorescence intensity of unstimulated cells.

B CELL ACTIN

TABLE Cytochalasin D-Stimulated

1

Decreases in B Cell G-Actin Content

AAU/min”

Treatment DNase only Cytochalasin D (20 &ml) 0 1 min 10 min 60 min

519

POLYMERIZATION

% Inhibition

[G-actin] *

0.098 0.063 0.089 0.089 0.083

35.7 9.2 9.2 15.3

1.24 /lg 0.65 pg 0.65 fig 0.78 pg

’ AAU/min = increase in absorbance units at ODzeO nm per minute. * The concentration of G-actin/ IO6 cells was determined using a standard curve: 0.85 pg G-actin produced an AU/min value of 0.058.

the G-actin content of PMNs: 4.35 pg/106 unstimulated cells, contrasted to 2.23 pg/ lo6 cells after 1 min with cytochalasin D, and 1.43 pg/106 cells with fmlp. Anti-immunoglobulin Reagents Do Not Stimulate B Cell Actin Polymerization Since in the NBD-Phallacidin assayincreases of less than 15-20 channels in fluorescence intensity were rarely reproducible, Table 2 shows B cells stimulated with various dosesof F(alQz goat anti-mouse IgM (GaMIg; Tago Immunologicals) did not change their F-actin content. Three other F(ab’)2 anti-Ig preparations gave similar results (Accurate Chemicals, Immunosearch Laboratories, and Cappel). Anti-Ig also failed to stimulate decreasesin B cell G-actin content as measured by the DNase I inhibition assay: Both unstimulated cells and those stimulated for 1 or 10 min with 6.25 Kg/ml RaMIg had an estimated 1.7 pg G-actin/106 cells. A second

TABLE 2 Anti-immunoglobulin

Did Not Stimulate Increases in B Cell F-Actin Content AMFI” Experiment

I

Experiment 2

Reagent (pg/ml)

30 set

1 min

10 min

30 set

1 min

10 min

F(ab’h GaMIg* 50 25 12.5 6.25 3.12 1.6

+8 -2 -5 -3 -7 -4

+12 -5 -9 -1 +6 -4

+1 -3 -6 +9 +5 -15

-3 +5 +2 -5 f2 -4

+2 +11 +2 -1 +8 n.d.’

+10 +12 +3 -8 -5 -12

a AMFI = mean fluorescence intensity of NBD-Phallacidin staining of stimulated cells - mean fluorescence intensity of NBD-Phallacidin staining of unstimulated cells. b GaMIg = goat anti-mouse immunoglobulin. c n.d. = not done.

52:0

WILDER

AND

ASHMAN

experiment showed a similar result, Together, these data suggestthat anti-Ig did not stimulate actin polymerization in B cells. Anti-Ig preparations vary in their stimulatory activity, so one possible explanation for the failure of our reagents to stimulate actin polymerization might be that they harppenedto be poor stimulators of B cell activation. This did not prove to be the case. All four anti-Ig reagents used in these studies stimulated polyphosphoinositide (PPI) hydrolysis at 25 pg/ml (stimulation indices ranged from 2.5- to 10.2-fold increases).In two separate experiments, increases in the bound:unbound Indo 1 ratio were 29 and 33%, reflecting an increase in intracellular calcium. DNA synthesis as mleasuredby [3H]thymidine incorporation also increased (stimulation indices ranged from 3.8- to 17-fold). F(ab’)z GaMIg (25 pug/ml;Tago Immunologicals) also stimulated capping and endocytosis of surface Ig as measured by indirect immunofluorescence: after a IO-min incubation at 37°C 60% ofthe cells counted displayed a capped phenotype while 28% of the cells had endocytosed their surface Ig. Both processeswere inhibited by 0.0 1 M sodium azide as expected. Other Molecules with Roles in B Cell Activation Proved Negative in the F-Act& Assay Having found that anti-Ig reagents which stimulated other activation events failed to induce actin polymerization, we tested lipopolysaccharide (LPS, 1 pg/ml (20)), dextran sulfate (10, 30 pg/ml (21)), and phorbol myristate acetate (PMA, 3-100 ng/ ml (22)) aswell as four recombinant lymphokines known to increaseB cell proliferation or differentiation (IL 1: 50, 250, 1000 U/ml; IL 2, IL 4, and IL 5: 5, 10, 50 U/ml (23)) in the NBD-Phallacidin binding assayafter 0.5, 1, or 10 min of stimulation. All of the F-actin values fell in the $25 to -5 range of AMFI, and none of the small increases in F-actin seen were reproducible on repeat testing. Only PMA was tested in the G-actin assay,and it was negative (data not shown). Large Concentrations of Cytochalasin D Stimulate Small Increases in Calcium in B Cells To define the mechanism by which cytochalasin D at concentrations of 20 pg/ml and higher induced actin polymerization in B lymphocytes, we tested whether this dose of cytochalasin D would induce changesin intracellular calcium concentration. In an experiment where the bound/unbound Indo 1 ratio increased from 0.086 (unstimulated cells) to 0.118 after 2.5 min with GaMIg (25 pg/ml, Tago Immunologicals) and 0.373 with ionomycin (0.1 pug/ml),cytochalasin D (20 pg/ml) increased the ratio to 0.102, and this small increasewas reproducible. The increasein intracellular calcium induced by all three reagents was only partially blocked by preloading the cells with 10 PM BAPTA-AM (an intracellular calcium chelator) but totally blocked by 8 mM EGTA with or without BAPTA-AM (data not shown). These results suggestthat calcium influx from the medium is stimulated by cytochalasin D, but do not rule out an earlier internal releasecomponent. Ionomycin Stimulated B Cell Actin Polymerization To test whether changesin intracellular calcium alone might account for the actin polymerization induced by cytochalasin D, B cells were stimulated with various concentrations of ionomycin, a calcium ionophore, and the changesin actin equilibrium

B CELL ACTIN

521

POLYMERIZATION

252015 log a

5O-5 -10 -15 -20 -25 -

0.02 kg/ml

0.1 lonomycln

FIG. 2. Effect of ionomycin on B cell F-actin content. NBD-Phallacidin binding was measured to assess the relative F-actin content of B cells stimulated with various doses of ionomycin. The data represent the mean (+ standard deviation) of three experiments. AMFI = mean fluorescence intensity of unstimulated cells. From left to right, the bars with different shadings represent 30 set, 1 min, 10 min, and 60 min.

were measured. Figure 2 shows that large concentrations (0.5 pg/ml) of ionomycin stimulated decreasesin B cell F-actin content at 30 set and 1 min, and increases at 10 and 60 mitt, as measured by NBD-Phallacidin binding. Lower concentrations of ionomycin failed to stimulate consistent increasesin F-actin content, as indicated by the large standard deviations. Table 3 shows that ionomycin stimulated a reciprocal decreasein G-actin content by 10 and 60 min to match the increase in F-actin at these time points. However, there was no great increase in G-actin content at 1 min to match the decreasein F-actin seen in Fig. 2. The basic relationships in this repre-

TABLE 3 Ionomycin-Stimulated Treatment None Ionomycin 1 min 10 min 60 min

Decreases in B Cell G-Actin Content [G-actin] a 1.19 fig

(0.5 &ml) 1.22 jig 0.95 /lg 0.54 pg

’ The concentration of G-actin/106 cells was determined by measuring changes in absorbance units at 260 nm/min (AAU/min) of a solution containing DNA, DNase, and test cell lysate. These values were compared to a standard curve produced by pure G-actin: 0.85 pg G-a&n produced a AAU/min value of 0.045; 1.36 fig G-actin produced a AAU/min value of 0.03.

522

WILDER

AND ASHMAN

sentative experiment have been confirmed in 17 other trials, and not contradicted in any. In no instance did PMA (which synergizes with ionomycin to promote DNA synthesis) alter the F-actin responseto ionomycin (3 experiments). Clearly, ionomycin induced actin polymerization in B cells, but it occurred later than that induced by cytochalasin D. The earlier apparent decreasein F-actin content cannot be interpreted as actin depolymerization becauseG-actin levels in the cells did not change. Ionomycin-Induced, but Not Cytochalasin D-Induced, Actin Polymerization Was Dependent on Calcium B cells were stimulated with ionomycin or cytochalasin D in the presenceof 8 n~J4 EGTA to prevent influx of calcium. Table 4 shows that cytochalasin D stimulated increases in F-actin content in the presence of EGTA, with or without BAPTA-AM, whereas both increases and decreasesin Phallacidin binding induced by ionomycin (seenalso in Fig. 2) were inhibited by EGTA. Table 5 showsthat the reciprocal changes in G-actin content of B cells stimulated with cytochalasin D were resistant to EGTA and BAPTA-AM, whereasthe changesinduced by ionomycin were blocked by EGTA. Together, these data prove that late ionomycin-induced actin polymerization is largely dependent on calcium influx, whereas actin polymerization induced by cytochalasin D is independent of calcium influx. Cytochalasin D Failed to Stimulate Polyphosphoinositide Hydrolysis The releaseand influx of calcium during the activation of B cells (and many other cells) has been attributed to the releaseof inositol phosphates by activation of a polyphosphoinositide-phospholipase C (24). When B cells were stimulated with low (2.5 pg/ml) and high (25 pg/ml) doses of cytochalasin D we detected no increase in PPI hydrolysis above background in both strains of mice tested (Table 6). The normal increase in PPI hydrolysis was seen when the same cells were stimulated with F(ab’)* TABLE 4 Cytochalasin

D, but Not Ionomycin, Stimulates an Increase in B Cell F-Actin Content in the Presence of a Calcium Chelator AMFI”

Cytochalasin D (20 &ml) 30 set 1 min 10 min Ionomycin (0.5 &ml) 30 set 1 min 180min 610min

No EGTA

EGTA (8 M)

BAPTA-AM (10 rw

EGTA + BAPTA-AM

+30 +46 +35

+31 f44 +56

+82 +49 +37

+47 +50 +46

-34 -51 +38 +32

+14 -12 -2 -1

n.d.h n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

n AMFL = mean fluorescence intensity of NBD-Phallacidin staining of stimulated cells - mean fluorescence intensity of NBD-Phallacidin staining of unstimulated cells. b n.d. = not done.

B CELL ACTIN

523

POLYMERIZATION

TABLE 5 Cytochalasin D, but Not Ionomycin, Decreases B Cell G-Actin Content in the Presence of Calcium Chelators [G-actin] a

Treatment Cytochalasin D (20 &ml) 0 1 min 10 min 60 min 0 1 min 10 min 60 min Ionomycin (0.5 &ml) 0 1 min 10 min 60 min

No addition

+8 mM EGTA

1.93 jig 0.96 pg 1.19 jig 1.36 /.rg 0.78 pg 0.70 pg 0.56 pg 0.43 pg

1.59 /.lg 0.79 pg 0.74 /.Lg 0.74 /Lg

1.87 pg 1.80 /.ig 1.51 rg 0.63 pg

1.21 /.lg 1.51 /Lg I .29 fig 1.36 fig

+10 PM BAPTA-AM

0.82 0.34 0.28 0.39

pg pg /.tg pg

’ The concentration of G-actin/106 cells was determined by measuring changes in absorbance units at 260 nm/min of a solution containing DNA, DNase, and test cell lysates. These values were compared to a standard curve using pure G-actin.

RaMIg (Table 6). Apparently, under the conditions we employed, cytochalasin D was able to trigger detectable calcium influx without detectable PPI hydrolysis. Cytochalasin D-Induced Actin Polymerization Did Not Depend on PKC Activation We preincubated B cells with the PKC inhibitor H7 (25) to block any PKC activation which might result from cytochalasin D stimulation and measured the changes in actin equilibrium. HA1004 (26), an inhibitor of cyclic nucleotide-dependent protein kinases, served as a control. As expected, H7 (100 /*M) inhibited B cell DNA synthesis induced by PMA and ionomycin (78 to 96% in three experiments), much better than 100 pm HA1004 (16 to 26%). Neither H7 nor HA 1004 inhibited cytochalasin Dinduced increases in F-actin content (Table 7). Two other experiments gave similar results. Furthermore, neither H7 nor HA1004 blocked the decrease in G-actin content induced by cytochalasin D. In the most complete experiment, G actin content was 1.24 pg/ lo6 unstimulated cells, but after 10 min it decreasedto 0.97 pg/ 1O6cells with cytochalasin D alone, 0.85 pg/ml with cytochalasin D + H7, and 0.88 pg/106 cells with cytochalasin D + HA1004. Two other experiments also showed no consistent inhibition of G-actin changesby theseinhibitors. There was no effect of H7 or HA1004 alone in the G- or F-actin assays(data not shown). Therefore, we conclude that cytochalasin D-induced actin polymerization is independent of PKC (and cyclic nucleotide-dependent protein kinase) activation.

5Z!4

WILDER

AND

ASHMAN

TABLE 6 Cytochalasin D Does Not Stimulate Polyphosphoinositide

Hydrolysis

in B Cells

No. experiments

Stimulus B,D,F, mice Cytochalasin D 2.5 pg/ml 25.0 pg/ml Cappel RaMIg b 25.0 fig/ml A/J mice Cytochalasin D 2.5 pg/ml 25.0 yg/ml Cappel RaMIgb 25.0 fig/ml

S.1. (SD)”

I 6

1.2 (0.3) 1.3 (0.3)

7

10.9 (2.8)

1 1

0.7 0.7

1

6.4

’ S.I. (SD) = stimulation index (standard deviation); stimulation index = [3H]inositol counts per minute of stimulated cells/[3H]inositol counts per minute of unstimulated cells. Method adapted from Bijsterbosch et al. (18). ’ RaMIg = rabbit anti-mouse immunoglobulin.

Inhibition of Protein Synthesis with Anisomycin Does Not Stimulate Actin Pal.ymerization in B Cells The fact that cytochalasin D at concentrations of 20 pug/ml and higher inhibits protein synthesis (27) led us to test whether protein synthesis inhibition was sufficient to stimulate actin polymerization in B cells. In pilot experiments, anisomycin at 0.2 ~14 inhibited protein synthesis (as measured by [35S]methionine incorporation) to the same extent as 20 pg/ml cytochalasin D. After 30 set and 10 and 60 mitt, 0.2 yM anisomycin failed to induce any increase in B cell F-actin content or decreasein Gactin content. The inhibition of DNase I ranged from 66 to 74% at these times with anisomycin compared to 71% without anisomycin. We are confident, therefore, that the induction of actin polymerization in B cells does not depend on the inhibition of protein synthesis.

TABLE I -

Neither H7 nor HA 1004 Inhibit Cytochalasin D-Induced Increases in B Cell F-Actin Content AMFI” Treatment

0

1 min

10 min

60 min

100 rM H7 100 pLMHA1004 Cytochalasin D (20 pg/ml) +H7 +HA 1004

-2.4 -6.7 +25.0 +50.9 +51.6

+4.3 +0.5 +38.9 +65.4 f38.8

+3.3 -5.9 +31.9 +26.4 +34.2

-0.9 -3.3 +19.9 +12.2 +8.6

a AMFI = mean fluorescence intensity of NBD-PhalIacidin staining of stimulated cells - mean fluorescence intensity of NDB-Phallacidin staining of unstimulated cells.

B CELL ACTIN

525

POLYMERIZATION

Cytochalasin D also Afects Proliferation Table 8 shows that low concentrations of cytochalasin D which failed to induce significant actin polymerization in mouse B cells (Fig. 1) also failed to induce t3H]thymidine incorporation, but neverthelesscostimulated well with F(ab’)2 anti-Ig. Whereas 20 pg/ml stimulated actin polymerization (Fig. l), it strongly inhibited antiIg-induced [3H]thymidine incorporation and this inhibition was not accounted for by the DMSO vehicle (Table 8). DISCUSSION Cytochalasin D, a small molecular weight fungal metabolite, has been reported to potentiate the DNA synthesis induced in B cells by anti-Ig reagents ( 14) lectins ( 15), or calcium ionophores (28). On the other hand, cytochalasin D inhibits protein synthesis (27) and can block actin polymerization by binding to the barbed end of the actin filament, which slows monomer addition to this end ( 13). We found, to our surprise, that large concentrations of cytochalasin D stimulated actin polymerization in living B lymphocytes. Interestingly, these concentrations (>20 pg/ml) are not those which costimulate with anti-immunoglobulin (( 14) and Table 8) or lectins (15) to induce increased DNA synthesis (~2.5 fig/ml). This finding makes it less likely that actin polymerization is related to progression to proliferation. As cytochalasin D inhibits actin polymerization in the test tube, its ability to increase F-actin content and decreaseG-actin content in living B cells represents a puzzling paradox. One attractive explanation is that a signaling pathway may be involved in the induction of actin polymerization in intact cells that would not come into play when cellular actin extracts are treated with cytochalasin D in the test tube. For several reasons, the polyphosphoinositide-calcium increase-PKC activation pathway appeared to be the most likely candidate. First, cytochalasin D has been reported to stimulate both B lymphocyte PPI hydrolysis (11) and increases in intracellular calcium ( 12) within minutes after coming into contact with the cells. Bengtsson et al. (29) have suggestedthat PPI hydrolysis and calcium increasesmay be important in regulating chemotactic peptide-induced actin polymerization in PMN. Other studies, TABLE 8 Effect of Cytochalasin D on Proliferation t3H]Thymidine Cytochalasin D (a/ml) 0

1.25 2.5 5 10 20

Without anti-kg 496 -c II 620 +- 73 750 f 253 414f 70 206 + 49 611 f 171

incorporation

+ SE”

With 6.25 &ml

anti-&

1012+- 177 50172 k 3440 38765 +_2674 19192 + 1733 2601 f 239 99-c 13

’ [SH]Thymidine incorporation was measured after 72 hr, with [‘Hlthymidine present for the last 6 hr. Macromolecular 3H was precipitated by 10% TCA. Controls: LPS, 118,660 + 7938; DMSO equivalent to samples with 20 pg/ml of cytochalasin D, 943 + 92 alone and 2832 + 591 with anti-lg. The same F(ab’)2 goat anti-mouse Ig preparation was used in Table 2 and Table 8.

5216

WILDER

AND

ASHMAN

however, have disputed the role of calcium in this system (6,30-32). Third, we showed that cytochalasin D stimulated actin polymerization in human PMNs as well as in B cells. However, our results make it unlikely that cytochalasin D-induced actin polymerization in B lymphocytes utilizes the PPI-calcium increase-PKC activation pathway. Although doses of cytochalasin D which stimulated actin polymerization also induced increases in intracellular calcium levels, inhibiting these calcium fluxes did not inhibit actin polymerization (Tables 4 and 5). We have also shown that inhibitors of PKC or cyclic nucleotide-dependent protein kinases fail to prevent cytochalasin D-induced Factin increases (Table 7). Furthermore, unlike Van Haelst and Rothstein (11) we were unable to detect any PPI hydrolysis after treating our B cells with either low (2.5 pg/ ml) or high (25 pg/ml) doses of cytochalasin D (Table 6). The reasons for this discrepancy are unclear given that we used very similar techniques to measure the PPI hydrolysis and that we also failed to detect this event in the mouse strain used by Van Haelst and Rothstein (A/J). However a signal is transduced when cytochalasin D encounters intact B cells, the result is conversion of G-actin to F-actin. Interestingly, the chemistry of the cytosol in resting cells actually favors actin polymerization, but it is restricted by actin-binding prloteins which serve to sequester monomers and block filament ends (7). Actin polymerization could result from either an increase in the concentration of free monomer in the cells after release from actin-binding proteins or an increase in free filament pointed ends upon which monomers can add. Both of these requirements could be fulfilled by the severing action of high concentrations of cytochalasin (20 pg/ml = 40 pA4 range) (33). The fact that only high concentrations of cytochalasin stimulate action polymerization is consistent with a role for filament severing in B cell actin polymerization. Anti-Ig reagents stimulate B cells to undergo PPI hydrolysis (8), increases in intracellular calcium (9), PKC activation (24), as well as actin-myosin microfilamentdependent responses such as capping and endocytosis of membrane Ig (34) and a brief motility response (10). We reasoned therefore that these reagents might also stimulate actin polymerization. However, our anti-Ig reagents failed to stimulate either increases in B cell F-actin content (Table 2) or decreases in G-actin content, even though they stimulated PPI hydrolysis, increased intracellular calcium, increased DNA synthesis, and caused capping and endocytosis of surface Ig, as expected. Bourguignon et al. (35) have shown that actin accumulates under the caps which ap:pear in B cells upon anti-immunoglobulin stimulation. That we were able to demonstrate capping and endocytosis of surface immunoglobulin without concomitant actin polymerization suggests that normal resting murine B cells have a sufficient amount of F-actin to accomplish both capping and endocytosis without further actin polymerization. Our data in normal B cells are in agreement with those of Jackman and Burridge (36) who showed that the capping of surface immunoglobulin and concanavalin A receptors on the continuously growing B cell lymphoma line CH 12 also was not accompanied by actin polymerization. The inability of anti-Ig reagents to stimulate actin polymerization, while stimulating PPI hydrolysis and increasing intracellular calcium, suggests that B cell surface Ig diflfers in its signal transduction mechanism from the chemotactic peptide receptor of PMNs. Although both surface immunoglobulin and the chemotactic peptide receptor are coupled to a polyphosphoinositide-phosphodiesterase via a G-protein, the G-protein in PMNs can be ADP-ribosylated by pertussis toxin (4) while that in B cells cannot

B CELL ACTIN

POLYMERIZATION

527

(37). Perhaps this difference in G-proteins determines the differences in response to these two receptors. We found no evidence for an actin polymerization responseto any of a long list of other B cell-activating molecules except for ionomycin, a calcium ionophore (Fig. 2 and Table 3). The ability of another calcium ionophore, A23 187, to stimulate actin polymerization has been shown previously by Howard and Wang (6) and Sha’afi et al. (32) in human and rabbit PMNs, respectively. Similar to the studies conducted by Howard and Wang, the actin polymerization induced by ionomycin was blocked if the rise in intracellular calcium was prevented (Tables 4 and 5). Of interest is an early decreasein Phallacidin binding (Fig. 2) which was dependent on calcium influx (Table 4) but which was not accompanied by any change in G actin (Table 5). One potential explanation would be a direct perturbation of Phallacidin binding to F-actin under conditions of massive calcium influx. Several possibilities exist to explain how large increases in intracellular calcium stimulate actin polymerization. Calcium has been reported to directly stimulate actin polymerization in a cell-free system (38, 39), but these studies were conducted in nonphysiologic buffers. Alternatively calcium may act through calcium-binding proteins such as calmodulin, caldesmon, and gelsolin. Caldesmon, a ubiquitous protein present in lymphocytes, binds to actin filaments in the absence,but not the presence, of a calcium-calmodulin complex (40) and can inhibit the myosin-ATPase activity normally induced by the actin-tropomyosin complex (41). Wallace and Piazza (42) have shown that a platelet actin preparation treated with calcium-calmodulin underwent actin polymerization. The calmodulin did not bind actin directly, but instead worked through a calmodulin-binding protein, possibly caldesmon. Alternatively, ionomycin-induced actin polymerization could be the result of changes in gelsolin, an actin-binding protein which normally blocks the barbed end of filaments. In the presence of micromolar free calcium, gelsolin acquires the ability to sever these filaments. Upon severingthe filaments, gelsolin binds two actin monomers, thus forming a nucleus upon which more monomers can add at the pointed end (43). All of the aforementioned mechanisms for ionomycin-induced actin polymerization require very large increases in intracellular calcium which would not be achieved by either cytochalasin D or antiIg stimulation. Our study fails to support an integral role for actin polymerization in the activation of the B lymphocytes, even though anti-Ig triggers active contraction and redistribution of actin-containing microfilaments. Such polymerization does follow cell contact with cytochalasin D, or high levels of internal calcium, by mechanisms yet to be defined. ACKNOWLEDGMENTS Supported by DHHS Grant AI22630 and a VA Merit grant. The authors gratefully acknowledge the technical assistance of Kaylan Belville and Jon Vandervelde, the secretarial assistance of Nancy Palm, and guidance on the NBD-Phallacidin assay from Dr. David Chambers.

REFERENCES 1. 2. 3. 4. 5.

Rao, K. M. K., and Varani, J., J. Immunol. 129, 1605, 1982. Varani, J., Wass, J. A., and Rao, K. M. K., J. Natl. Cancer Inst. 70, 805, 1983. Wallace, P. J., Wersto, R. P., Packman, C. H., and Lichtmann, M. A., J. Cell&o/. 99, 1060, 1984. Dillon, S. B., Murray, J. J., Uhing, R. J., and Snyderman, R., J. Cell Biochem. 35, 345, 1987. White, J. R., Naccache, P. H., Molski, T. F. P., Borgeat, P., and Sha’afi, R. I., Biochem. Biophys. Res. Commun. 113,44, 1983.

528

WILDER AND ASHMAN

6. Howard, T. H., and Wang, D., J. Clin. Invest. 79, 1359, 1987. 7. Stossel,T. P., Chaponnier, C., Ezell, R. M., Hartwig, J. H., Janmey, P. A., Kwiatkowski, D. J., Lind, S. E., Smith, D. B., Southwick, F. S., Yin, H. L., and Zaner, K. S., Annu. Rev. Cell Biol. 1, 353, 1985. 8. Coggeshall, K. M., and Cambier, J. C., J. Zmmunol. 133, 3382, 1984. 9. Braun, J., Sha’afi, R. I., and Unanue, E. R., J. Cell Biol. 82, 755, 1979.

10. Schreiner, G. F., and Unanue, E. R., In “Advances in Immunology” (W. H. Taliaferro, and J. H. Humphrey, Eds.), Vol. 24, pp. 37-165. Academic Press,San Diego, 1976. 11. Van Haelst, C., and Rothstein, T. L., J. Zmmunol. 140, 1256, 1988. 12. Baeker, T. R., Simons, E. R., and Rothstein, T. L., J. Zmmunol. 138,2691, 1987. 13. Bonder, E. M., and Mooseker, M. S., J. Cell Biol. 102, 282, 1986. 14. Rothstein, T., J. Zmmunol. 135, 106, 1985. 15. Hoffman, R., Ferguson, R., and Simmons, R. L., J. Zmmunol. 118, 1472, 1977. 16. Barak, L. S., and Yocum, R. R., Anal. Biochem. 110, 3 1, 198 1. 17. Blikstad, I., Markey, F., Carlsson, L., Pearson, T., and Lindberg, U., Cell 15, 935, 1978. 18. Bijsterbosch, M. K., Meade, C. J., Turner, G. A., and Klaus, G. G. B., Cell41, 999, 1985. 19. Grynkiewicz, G., Poenie, M., and Tsien, R. Y., J. Biol. Chem. 260, 3440, 1985. 20. Morrison, D. C., and Ulevitch, R. J., Am. J. Pathol. 93, 527, 1978. 21. Swain, S. L., Howard, M., Kappler, J., Marrack, P., Watson, J., Booth, R., Wetzel, G. D., and Dutton, R. W., J. Exp. Med. 58, 822, 1983. 22. Davis, L., and Lipsky, P. E., J. Zmmunol. 136, 2946, 1985. 23. O’Garra, A., Umland, S., DeFranco, T., and Christiansen, J., Zmmunol. Today 9, 45, 1988. 24. Ransom, J. T., and Cambier, J. C., J. Zmmunol. 136, 66, 1986. 25. Hidaka, H., Inagaki, M., Kawamoto, S., and Sasaki, Y., Biochemistry 23, 5036, 1984. 26. Asano, T., and Hidaka, H., J. Pharmacol. Exp. Ther. 231, 141, 1984. 27 Omelles, D. A., Fey, E. G., and Penman, S., Mol. Cell. Biol. 6, 1650, 1986. 28 Rothstein, T. L., J. Zmmunol. 136, 749, 1986. 29 Bengtsson,T., Stendahl, O., and Andersson, T., Eur. J. Cell Biol. 42, 338, 1986. 30. Zigmond, S. H., Slonczewski, J. L., Wilde, M. W., and Carson, M., CellMotil. C.vtoskel.9, 184, 1988. 3 1, Carson, M., Weber, A., and Zigmond, S. H., J. Cell Biol. 103, 2707, 1986. 32. Sha’afi, R. I., Shefcyk, J., Yassin, R., Molski, T. F. P., Volpi, M., Naccache,P. H., White, J. R., Feinstein, M. B., and Becker, E. L., J. Cell Biol. 102, 1459, 1986. 33. Urbanik, E., and Ware, B. R., Arch. Biochem. Biophys. 269, 181, 1989. 34. Unanue, E. R., Perkins, W. D., and Karnovsky, M. J., J. Zmmunol. 108, 569, 1972. 35. Bourguignon, L. Y. W., Tokuyasu, K. T., and Singer, S. J., J. Cell Phys. 95, 239, 1978. 36. Jackman, W. T., and Burridge, K., Cell Motil. Cytoskel. 12, 23, 1989. 37. Klaus, G. G. B., Hamett, M. M., and Rigley, K. P., In “Proceedings, 2nd International Conference on Lymphocyte Activation,” pp. 95-100. Plenum Press,New York 1989. 38. Tellam, R., Biochemistry 24, 4455, 1985. 39. Dancker, P., and Low, F., Biochim. Biophys. Acta 484, 169, 1977. 40. Sobue, K., Kanda, K., Tanaka, T., and Ueki, N., J. Cell. Biochem. 37, 3 17, 1988. 41. Pritchard, K., and Moody, C. J., Cell Calcium 7, 309, 1986. 42. Wallace, R. W., and Piazza, G. A., In “Methods in Enzymology” (Means, A. R., and Conn, P. M., Eds.), Vol. 139, pp. 846-857. Academic Press,San Diego, 1987. 43. Janmey, P. A., Iida, K., Yin, H. L., and Stossel,T. P., J. Biol. Chem. 262, 12,228, 1987.