Pertussis toxin effects on chemoattractant-induced response heterogeneity in human PMNs utilizing Fluo-3 and flow cytometry

Flow cytometric methods were utilized to determine N-formylpeptide-induced cytosolic calcium levels in human polymorphonuclear leukocytes (PMNs) detected with the calcium indicator Fluo-3. Fluo-3 was readily loaded into PMNs as the acetoxymethyl ester. At room temperature Fluo-3 extrusion was minimal (less than 10%) over a 2 h time period. Flow cytometric histograms yielded symmetric distributions indicating homogeneous labelling of the cells. Stimulation of the cells with N-formyl-met-leu-phe (FMLP) caused homogeneous activation of all cells as indicated by a shift of the fluorescence distribution to higher fluorescence levels while still maintaining a symmetrical distribution. Resting values or FMLP-induced cytosolic calcium levels were similar in cells loaded over a 20-fold range of Fluo-3-acetoxymethyl ester. The effect of graded pertussis toxin (PT) treatment on the calcium response was determined by incubating cells with different concentrations of pertussis toxin for a time period that yielded a range of ADP ribosolation levels inside the cells. When these cells were activated with FMLP, the fluorescence histograms showed that pertussis toxin treatment resulted in a conversion of cells from responders to nonresponders. The responding cells responded with maximum calcium elevations similar to controls. This behavior may reflect heterogeneous insertion of the A-protomer of PT or a very sharp threshold of coupled G-proteins required to transduce the responses.     

Graded G-protein uncoupling by pertussis toxin treatment of human polymorphonuclear leukocytes.

Pertussis toxin (PT) inhibits polymorphonuclear leukocyte (PMN) function by ADP-ribosylating and inactivating guanine nucleotide binding proteins (G-proteins) that transduce activation by chemoattractants such as N-formyl peptides (FP). Studies of PMN activation during the time course of PT treatment yielded these results. 1) Responses were differentiated based on their sensitivity to PT treatment. Suboptimal PT treatment that resulted in 50% inhibition of the FP-induced actin-associated right angle light scatter response resulted in greater than 90% inhibition of oxidant production. Exhaustive PT treatment was required to completely inhibit the right angle light scatter response. This is consistent with previous observations that, relative to oxidant production, actin polymerization requires 100-fold fewer active N-formylpeptide receptors to elicit the response. This differential sensitivity to PT treatment has important implications for studies that use pertussis toxin to determine if the neutrophil G-protein is involved in the signaling of responses. If inhibition of oxidant production is used as the only indicator of the effectiveness of PT treatment, significant cytoskeletal changes may still be activated in these cells. Inhibition of actin polymerization is a much more rigorous indicator of complete G-protein inhibition by PT. 2) Analysis of FP-induced actin polymerization and cytosolic calcium elevation using flow cytometry, which measures individual cell responses, revealed that PT treatment resulted in the conversion of PMN from a responding to a non-responding population. In contrast, in control PMN, submaximal doses of FP caused submaximal stimulation of all the cells. The all-or-none effect of PT may result from heterogeneous insertion of the A-promoter of PT into the cell or it may result from a sharp threshold of coupled G-proteins required to transduce the responses.     

Actin depolymerization and inhibition of capping induced by pentoxifylline in human lymphocytes and neutrophils

Pentoxifylline is used clinically for the treatment of intermittent claudication. It is believed to exert its effect by altering the rheologic properties of blood. The cytoskeleton plays an important role in the maintenance of cell structure and function. In particular, alterations in the state of actin seem to play an important role in cell motility. Therefore, we examined the effect of pentoxifylline on the actin state in human polymorphonuclear leukocytes (PMN) and mononuclear cells. Pentoxifylline (10 mM final concentration) decreased F-actin content in both PMN and mononuclear cells. Pentoxifylline also inhibited concanavalin A-induced capping in PMN and mononuclear cells. Similarly, surface immunoglobulin capping in B lymphocytes was also inhibited. Pretreatment of cells with pertussis toxin did not inhibit pentoxifylline-induced decrease in F-actin, suggesting pentoxifylline does not act through pertussis toxin-sensitive G-proteins. Dibutyryl cyclic AMP failed to show any significant effect on the F-actin content in PMN. Therefore, the effect of pentoxifylline cannot be attributed to changes in cyclic AMP levels. Chemotactic peptide-induced actin polymerization was unaffected in PMN when expressed as percent changes in F-actin. The observations reported here suggest that the rheological effects of pentoxifylline might be due to its effects on the actin state in the cellular elements of the blood. Further studies on the mechanism of action of pentoxifylline on actin state in leukocytes will prove useful in delineating the physiological mechanisms regulating actin state in leukocytes.     

Neutrophil thrombospondin receptors are linked to GTP-binding proteins

The extracellular matrix (ECM) protein thrombospondin (TSP) binds to specific receptors on polymorphonuclear leukocytes (PMNs) and stimulates motility. TSP can also enhance the response of PMNs to the formylated peptide, N-formyl-methionyl-leucyl-phenylalanine (FMLP). Our initial evidence suggesting that PMN TSP receptors were linked to GTP-binding proteins (G-proteins) came from studies using pertussis toxin (PT) and cholera toxin (CT) to inhibit TSP-mediated motility. Both PT and CT inhibited TSP-mediated chemotaxis and substrate-associated random migration. Inhibition was not indirectly caused by a rise in cAMP since neither dibutyryl cAMP (300 μM) nor 8-bromo-cAMP (300 μM) significantly affected TSP-mediated motility. In fact, TSP itself caused a significant rise in intracellular cAMP levels (from 7.2 ± 0.3 to 14.2 ± 0.1 pmol/10⁶ cells). Although we could not test the PT sensitivity of TSP priming for FMLP-mediated chemotaxis (as PT inhibits FMLP-mediated chemotaxis itself), we evaluated the effect of CT on this response. CT completely abolished TSP-dependent priming of FMLP-mediated chemotaxis. Direct evidence for an interaction between TSP receptors and G-proteins was obtained by examining the effect of TSP on α-subunit ADP-ribosylation, GTPase activity, and GTPγS binding. We observed a decrease in the ability of FMLP to stimulate GTPase activity on membranes isolated from PMNs incubated with TSP. Furthermore, the PT-dependent ribosylation of G_iα2,3 stimulated by FMLP was eliminated by TSP treatment. These data indicated that the two receptors share a pool of G-proteins. However, TSP did not block the CT-dependent ribosylation stimulated by FMLP, suggesting that TSP receptors may also interact with a different pool of G_iα2,3. TSP itself significantly (P < 0.005) increased GTP hydrolysis in PMN membranes (to 110.6 ± 2.7% of control values). In addition, GTPγS binding to membranes increased significantly (P < 0.005) following exposure to 10 nM TSP (to 108 ± 1.4% of control values). Conversely, GTP treatment reduced the affinity of TSP for its receptor without altering total binding. These data demonstrate that TSP receptors are linked to G-proteins, a subpopulation of which also associates with FMLP receptors. © 1996 Wiley-Liss, Inc.     

Introduction of NAD decreases fMLP-induced actin polymerization in chicken polymorphonuclear leukocytes--the role of intracellular ADP-ribosylation of actin for cytoskeletal organization.

We reported previously that the arginine-specific ADP-ribosyltransferase in chicken polymorphonuclear leukocytes specifically modified actin, thereby inhibiting actin polymerization in vitro. In the present study, we investigated the effect of ADP-ribosylation on actin polymerization in situ. In the leukocytes, the introduction of NAD inhibited the increase in filamentous actin contents induced by a chemotactic peptide formyl-methionyl-leucyl-phenylalanine, while introduction of NAD together with novobiocin, a specific inhibitor for ADP-ribosyltransferase, did not. These results suggest that ADP-ribosylation regulates the formation of filamentous actin by the covalent modification of the protein in vivo.     

An endogenous inhibitor of the ADP-ribosylation of GTP-binding proteins by pertussis toxin is present in bovine brain

The ADP-ribosylation of GTP-binding proteins (G-proteins) catalyzed by pertussis toxin was inhibited by endogenous inhibitor activity in the membrane extract of bovine brain. Most of the activity appeared in the fractions eluted from a DEAE-Sephacel column by 0.5 M NaCl. The activity was heat-stable and sensitive to pronase K. The results suggest the presence of an endogenous inhibitor of pertussis toxin in bovine brain.     

Binding of IgG containing immune complexes to human neutrophil Fc gamma RII and Fc gamma RIII induces actin polymerization by a pertussis toxin-insensitive transduction pathway.

The ability of a phagocytic stimulus, rabbit IgG anti-BSA/BSA immune complexes, to increase the F-actin content of human polymorphonuclear leukocytes was quantitated by flow cytometry following staining with nitrobenzoxadiazole-phallacidin. A significant rise in F-actin assembly was induced by addition of 5 micrograms/ml immune complex. Concentrations of immune complex of more than 200 micrograms/ml caused a maximal (approximately twofold) increase in F-actin content. After a delay of 5 s, the F-actin levels rose and reached maximum levels by 60 s after adding immune complexes. The twofold elevation in F-actin persisted for up to 60 min. Both anti-Fc gamma RII and anti-Fc gamma RIII mAb blocked immune complex stimulated actin polymerization. Exposure to pertussis toxin failed to affect the rate or extent of immune complex-induced actin polymerization. Cells incubated with immune complexes and then lysed with Triton had an increased number of sites able to nucleate actin polymerization. These findings suggest that immune complex binding to both polymorphonuclear leukocytes Fc gamma RII and Fc gamma RIII is required for actin filament assembly and that the induction of assembly occurs via transduction pathways that differ from those used by chemoattractants. As with adhesion this phagocytic stimulus induces actin assembly by a pertussis toxin insensitive pathway and produces a rise in actin filament content that persists for prolonged periods of time.     

Pertussis toxin catalyzed ADP-ribosylation of a 41 kDa G-protein impairs insulin-stimulated glucose metabolism in BC3H-1 myocytes

In this study, we examined the effects of pertussis toxin (PT) on the ADP-ribosylation of guanine nucleotide binding proteins (G-proteins) and various insulin-stimulated processes in cultured BC3H-1 myocytes. Treatment of intact myocytes with 0.1 microgram/ml PT for 24 hours resulted in the complete ribosylation of a 41 kDa protein. The 41 kDa PT substrate was immunoprecipitated with antibodies directed against a synthetic peptide corresponding to a unique sequence in the alpha subunit of Gi-proteins. PT treatment of intact cells had no effect on insulin receptor binding or internalization. However, PT inhibited insulin-stimulated glucose transport at all insulin-concentrations tested (1-100 ng/ml). Maximally stimulated glucose transport was reduced by 50% +/- 15%. Insulin-stimulated glucose oxidation was also decreased by 31% +/- 8%. The toxin had no significant effect on the basal rates of glucose transport and glucose oxidation. The time course of PT-induced inhibition on glucose transport correlated with the time course of the "in vivo" ADP-ribosylation of the 41 kDa protein. The results suggest that a 41 kDa PT-sensitive G-protein, identical or very similar to Gi, is involved in the regulation of glucose metabolism by insulin in BC3H-1 cells.     

Expression of bacterial cytotoxin genes in mammalian target cells

We have studied the expression of the gene fragments encoding the enzymatically active portion of three bacterial cytotoxins: exotoxin A (ETA) of Pseudomonas aeruginosa, and pertussis toxin (PT) and adenylate cyclase toxin (CYA) of Bordetella pertussis, in sensitive mammalian target cells. Expression of active ETA and CYA was lethal to the producing cells and stable transfectants of Cos-1 cells containing the corresponding genes could not be obtained. The expression of the PTS1 subunit was tolerated by the producing mammalian cells. Since PT is cytotoxic because of ADP-ribosylation of G-proteins, we assume that the endogenously expressed PTS1 may not find the cellular target G proteins or PTS1 alone may not be sufficient for ADP-ribosylation of these proteins in vivo.     

Oscillating actin polymerization/depolymerization responses in human polymorphonuclear leukocytes

Leukotriene B4 and platelet-activating factor induced a rapidly oscillating actin polymerization/depolymerization response in human polymorphonuclear leukocytes. N-Formylpeptides were deficient in the ability to induce these oscillations. Flow cytometric analysis of filamentous actin verified that all cells were synchronously responding in this cyclic manner. The hypothesis was tested that these oscillations were analogous to chemical oscillations, i.e. oscillations of intermediate species in chemical systems that are far from equilibrium (Epstein, I. R., Kustin, K., DeKepper, P., and Orban, M. (1983) Sci. Am. 248, 112). Actin polymerization/depolymerization cycles were terminated by adding receptor antagonist a few seconds after initiation of the response by agonists. Thus the oscillations did not represent chemical oscillations that hypothetically could result from a rapid jump of the intracellular milieu to a state far from equilibrium. Rather, continued occupancy of receptors and/or occupancy of new receptors was required to sustain the oscillations. This suggested that the oscillations resulted from regulated polymerization and depolymerization pathways. In simultaneous measurements of actin-associated right angle light scatter and intracellular calcium, no calcium oscillations were detected. Thus, cycles of actin polymerization/depolymerization were not regulated by calcium oscillations.     

Actin assembly in electropermeabilized neutrophils: role of G-proteins

Polymerization of microfilaments, one of the responses triggered in neutrophils by stimuli such as the chemoattractant N-formyl-methionyl-leucyl-phenylalanine (fMLP), involves the conversion of actin from the monomeric to the filamentous form. The exact sequence of events responsible for this conversion remains to be defined, but its susceptibility to inhibition by pertussis toxin provides indirect evidence that GTP-binding proteins (G-proteins) are involved. In this report, electropermeabilized cells were used to obtain more direct evidence of a role for G-proteins in actin assembly. Staining with 7-nitrobenz-2-oxa-1,3-diazole (NBD)-phallacidin and flow cytometry were used to monitor the formation of filamentous actin. GTP-gamma-S, a nonhydrolyzable analogue of GTP and aluminum fluoride, which in combination with GDP can activate G-proteins, stimulated actin assembly in electropermeabilized cells but had only marginal effects on intact cells. fMLP-induced actin polymerization in permeabilized cells was inhibited by pretreatment with GDP-beta-S, an analogue of GDP that stabilizes the inactive form of G-proteins. In contrast, stimulation by phorbol 12-myristate 13-acetate (PMA) was largely unaffected by GDP-3-S. These observations indicate that activation of G-proteins is essential for actin assembly induced by receptor-dependent stimuli such as fMLP. Moreover, GTP-binding proteins do not seem to be required in the late stages of the signalling cascade, i.e. after stimulation of protein kinase C.     

Subtype-specific increase in G-protein alpha-subunit mRNA by interleukin 1 beta

The guanine nucleotide regulatory proteins (G-proteins) which are substrates for ADP-ribosylation by pertussis toxin (alpha i-1, alpha i-2, alpha i-3 and alpha o) transduce a variety of hormonal signals. Endothelial cells express mRNA for three alpha i subtypes although the level of alpha i-1 mRNA is very low. Interleukin 1 beta (IL 1 beta), a pleiotropic inflammatory mediator which stimulates a complex series of responses in human endothelial cells leading to increased coagulation and platelet adhesion, increases expression of one subtype of alpha i (alpha i-2) mRNA in human endothelial cells as determined by Northern blot analysis without affecting the level of mRNA for other alpha-subunits. These studies show that mRNA levels for alpha i subtypes are independently regulated, suggesting that there may be subtype specificity in the cell's requirements for the Gi class of signal-transducing proteins.     

Different types of ADP-ribose protein bonds formed by botulinum C2 toxin, botulinum ADP-ribosyltransferase C3 and pertussis toxin

We attempted to characterize ADP-ribose-amino acid bonds formed by various bacterial toxins. The ADP-ribose-arginine bond formed by botulinum C2 toxin in actin was cleaved with a half-life of about 2 h by treatment with hydroxylamine (0.5 M). In contrast, the ADP-ribose-cysteine bond formed by pertussis toxin in transducin and the ADP-ribose-amino acid linkage formed by botulinum ADP-ribosyltransferase C3 in platelet cytosolic proteins were not affected by hydroxylamine. HgCl2 cleaved the ADP-ribose-amino acid bond formed by pertussis toxin in transducin but not those formed by botulinum C2 toxin or botulinum ADP-ribosyltransferase C3 in actin and platelet cytosolic proteins, respectively. NaOH (0.5 M) cleaved the ADP-ribose-amino acid bonds formed by botulinum C2 toxin and pertussis toxin but not the one formed by botulinum ADP-ribosyltransferase C3. The data indicate that the ADP-ribose bond formed by botulinum ADP-ribosyltransferase C3 differs from those formed by the known bacterial ADP-ribosylating toxins.     

Immunochemical detection of guanine nucleotide binding proteins mono-ADP-ribosylated by bacterial toxins

Rabbits immunized with ADP-ribose chemically conjugated to carrier proteins developed antibodies reactive against guanine nucleotide binding proteins (G proteins) that had been mono-ADP-ribosylated by bacterial toxins. Antibody reactivity on immunoblots was strictly dependent on incubation of substrate proteins with both toxin and NAD and was quantitatively related to the extent of ADP-ribosylation. Gi, Go, and transducin (ADP-ribosylated by pertussis toxin) and elongation factor II (EF-II) (ADP-ribosylated by pseudomonas exotoxin) all reacted with ADP-ribose antibodies. ADP-ribose antibodies detected the ADP-ribosylation of an approximately 40-kilodalton (kDa) membrane protein related to Gi in intact human neutrophils incubated with pertussis toxin and the ADP-ribosylation of an approximately 90-kDa cytosolic protein, presumably EF-II, in intact HUT-102 cells incubated with pseudomonas exotoxin. ADP-ribose antibodies represent a novel tool for the identification and study of G proteins and other substrates for bacterial toxin ADP-ribosylation.     

Characterization of cytosolic pertussis toxin-sensitive GTP-binding protein in mastocytoma P-815 cells

We have characterized a soluble pertussis toxin (PT)-sensitive GTP-binding protein (G-protein) present in mouse mastocytoma P-815 cells. 65% of total ADP-ribosylation of PT substrate having a molecular mass of 40 kDa on SDS-polyacrylamide gel electrophoresis in cell homogenate was detected in the supernatant after centrifugation at 100,000 x g for 90 min. [32P]ADP-ribosylation of cytosolic PT substrate was significantly enhanced on the addition of exogenous beta gamma complex. The molecular mass of the cytosolic PT substrate was estimated to be about 80 kDa on an Ultrogel AcA 44 column, but the beta gamma complex was not detected in the cytosol by using the anti-beta gamma complex antibody. Furthermore, the cytosolic PT substrate was found to have some unique properties: [35S]GTP gamma S binding was not inhibited by GDP and [32P]ADP-ribosylation was not affected by GTP gamma S treatment. Only after the cytosolic PT substrate had been mixed with exogenous beta gamma complex, did it copurify with exogenous beta gamma complex by several column chromatographies including an Octyl-Sepharose CL-4B column. The PT substrate was identified as Gi2 alpha by Western blot analysis and peptide mapping with S. aureus V8 protease. These results suggest that Gi2 alpha without beta gamma complex exists with an apparent molecular mass of about 80 kDa in the cytosolic fraction of P-815 cells.     

Characterization of the ADP-ribosylation of actin by Clostridium botulinum C2 toxin and Clostridium perfringens iota toxin

Clostridium botulinum C2 toxin and Clostridium perfringens iota toxin belong to a novel family of actin ADP-ribosylating toxins. ADP-ribosylation of actin inhibits actin polymerization and G-actin-associated ATPase activity. The ADP-form of actin is ADP-ribosylated at a higher rate than actin with bound ATP. ADP-ribosylation of actin is reversible, a reaction, which is accompanied by reconstitution of actin ATPase activity.     

Cholera toxin treatment produces down-regulation of the alpha-subunit of the stimulatory guanine-nucleotide-binding protein (Gs).

The effect of activation of the alpha-subunit(s) of the stimulatory guanine-nucleotide-binding protein, Gs, on levels of this polypeptide(s) associated with the plasma membrane of L6 skeletal myoblasts was ascertained. Incubation of these cells with cholera toxin led to a time- and concentration-dependent 'down-regulation' of both 44 and 42 kDa forms of Gs alpha as assessed by immunoblotting with an anti-peptide antiserum (CS1) able to identify the extreme C-terminus of Gs. The effect of cholera toxin was specific for Gs; levels of Gi alpha in membranes of cholera toxin-treated cells were not different from untreated cells. Down-regulation of Gs was absolutely dependent upon prior ADP-ribosylation, and hence activation of Gs and was not mimicked by other agents which elevate intracellular levels of cyclic AMP. Pretreatment with pertussis toxin, which catalyses ADP-ribosylation of Gi but not of Gs, did not down-regulate either Gi or Gs, demonstrating that covalent modification by ADP-ribosylation is alone not a signal for removal of G-proteins from the plasma membrane.     

Pertussis toxin-sensitive G-proteins are not involved in activation of T-lymphocytes.

Multiple effects of pertussis toxin (PT) on Jurkat T-cells can be distinguished on the basis of their dose-response and their kinetics. High concentrations of PT deliver to cells an activating signal resulting in a rapid rise in [Ca2+]i followed by IL-2 synthesis. This activation is accompanied (within 2 h) by a down-regulation of the CD3/TCR complex from the cell surface. Cells then become refractory towards stimulation by CD3 mAb or PHA. All these effects, referred to as 'mitogenic effects', present the same dose-response curves with an EC50 of 0.5 micrograms/ml. Short term effects (PT-induced Ca2+ movements, down-regulation of CD3/TCR complex and inhibition of PHA and CD3-induced Ca2+ signal) are observed under conditions where no PT-induced ADP-ribosylation can be detected. In contrast, ADP-ribosylation of the 40,000 alpha-subunit of G-proteins requires a sustained (18 h) incubation of intact cells in the presence of low concentration (EC50 = 0.3 ng/ml) of PT. Dose-response curves for PT-dependent ADP-ribosylation and mitogenic effects are separated by three orders of magnitude. Covalent modification of G-protein has no effect on CD3-induced increase in [Ca2+]i and IL-2 synthesis induced by a combination of phorbol ester and either CD3 mAb, PHA or calcium ionophore. These data indicate that transduction of the mitogenic signal does not involve a PT-sensitive G-protein. Furthermore, inhibition of mitogenic signals following PT treatment results from a PT-induced activation leading to a down-regulation of the CD3/T cell receptor complex.     

Transduction of the chemotactic signal to the actin cytoskeleton of Dictyostelium discoideum

Dictyostelium discoideum amebae chemotax toward folate during vegetative growth and toward extracellular cAMP during the aggregation phase that follows starvation. Stimulation of starving amebae with extracellular cAMP leads to both actin polymerization and pseudopod extension (Hall et al., 1988, J. Cell. Biochem. 37, 285-299). We have identified an actin nucleation activity (NA) from starving amebae that is regulated by cAMP receptors and controls actin polymerization (Hall et al., 1989, J. Cell Biol., in press). We show here that NA from vegetative cells is also regulated by chemotactic receptors for folate. Our studies indicate that NA is an essential effector in control of the actin cytoskeleton by chemotactic receptors. Guided by a recently proposed model for signal transduction from the cAMP receptor (Snaar-Jagalska et al., 1988, Dev. Genet. 9, 215-225), we investigated which of three signaling pathways activates the NA effector. Treatment of whole cells with a commercial pertussis toxin preparation (PT) inhibited cAMP-stimulated NA. However, endotoxin contamination of the PT appears to account for this effect. The synag7 mutation and caffeine treatment do not inhibit activation of NA by cAMP. Thus, neither activation of adenylate cyclase nor a G protein sensitive to PT treatment of whole cells is necessary for the NA response. Actin nucleation activity stimulated with folate is normal in vegetative fgdA cells. However, cAMP suppresses rather than activates NA in starving fgdA cells. This indicates that the components of the actin nucleation effector are present and that a pathway regulating the inhibitor(s) of nucleation remains functional in starving fgdA cells. The locus of the fgdA defect, a G protein implicated in phospholipase C activation, is directly or indirectly responsible for transduction of the stimulatory chemotactic signal from cAMP receptors to the nucleation effector in Dictyostelium.     

Responses of pertussis toxin-treated microvascular endothelial cells to transforming growth factor beta 1. No evidence for pertussis-sensitive G-protein involvement in TGF-beta signal transduction.

Responses of bovine adrenal capillary endothelial cells (BACE) on treatment with transforming growth factor beta 1 (TGF-beta 1) have been characterized and tested for sensitivity to inactivation of pertussis toxin-sensitive G-proteins. TGF-beta 1 elicited growth inhibition, monolayer remodeling, elevation of steady state mRNA levels for collagen type 1 (alpha 1(1) and alpha 2(1)) and TGF-beta 1, and inhibition of p34cdc2 histone H1 kinase activity in BACE cells. Pertussis toxin treatment enhanced both inhibition of BACE cell [3H]methylthymidine uptake and remodeling of BACE monolayers by TGF-beta 1. These findings contrast with studies of mink lung epithelial cells, in which TGF-beta 1 growth inhibition has been shown to be pertussis-sensitive. Further investigation revealed that pertussis toxin treatment of BACE cells had no effect on TGF-beta 1-stimulated elevation of steady state mRNA levels for collagen type 1 (alpha 1(1) or alpha 2(1)) or for TGF-beta 1. Analysis of p34cdc2 activity in BACE cells revealed potent inhibition of p34cdc2 histone H1 kinase activity by TGF-beta 1. Pertussis toxin treatment also abolished the increase in p34cdc2 activity, however, precluding the determination of the pertussis toxin sensitivity of this response to TGF-beta 1. Consistent with suppression of p34cdc2 activation, pertussis toxin also caused substantial inhibition of mitogen-stimulated BACE cell [3H]methylthymidine uptake. It is concluded that TGF-beta 1 signal transduction in this cell type does not involve G-proteins of the pertussis toxin-sensitive class and that, in view of its potent effects on DNA synthesis and p34cdc2 activation, the use of pertussis toxin to determine G-protein involvement in cytokine signalling pathways should be approached with caution.