Faculty Bibliography

Transforming growth factor beta 1 (TGF-beta 1), a homodimeric polypeptide (Mr 25,000), derives from inflammatory cells and acts as a chemoattractant for monocytes and fibroblasts. We report here that TGF-beta 1 is also the most potent chemoattractant yet described for human peripheral blood neutrophils. Recombinant TGF-beta 1 elicited dose-dependent directed migration of neutrophils under agarose that was inhibited in the presence of a neutralizing antibody to TGF-beta 1. Maximal chemotaxis was evoked by TGF-beta 1 at femtomolar concentrations, whereas conventional chemoattractants act at nanomolar concentrations: on a molar basis, TGF-beta 1 was 150,000 times more potent than fMet-Leu-Phe. In contrast, TGF-beta 1 provoked neither exocytosis nor the production of superoxide by neutrophils. We further analyzed the mechanism by which TGF-beta 1 elicits chemotaxis (GTPase activity, [Ca2+], and actin polymerization). In contrast to the conventional chemoattractant fMet-Leu-Phe, TGF-beta neither activated classic heterotrimeric guanine nucleotide-binding proteins nor provoked global mobilization of intracellular Ca2+. Chemoattraction by both fMet-Leu-Phe and TGF-beta 1 was inhibited by cycloheximide and actinomycin D. Moreover, chemotaxis in response to TGF-beta 1 was associated with the polymerization of actin. The selectivity and potency of TGF-beta 1 as a chemoattractant suggest that it elicits directed cell migration by means of a pathway that depends not on classic intracellular signals but on protein synthesis

Upon engagement of chemoattractant receptors, neutrophils generate inositol trisphosphate and diacylglycerol (DG) by means of a phosphatidylinositol-specific phospholipase C (PI-PLC) which is regulated by a GTP-binding protein(s). We have previously reported (Reibman, J., H. M. Korchak, L. B. Vosshall, K. A. Haines, A. M. Rich, and G. Weissmann. 1988. J. Biol. Chem. 263:6322-6328) a biphasic rise in DG after exposure of neutrophils to the chemoattractant FMLP: a rapid (less than or equal to 15 s) phase ('triggering') and a slow (greater than or equal to 30 s) phase ('activation'). These derive from distinct intracellular lipid pools. To study the source of rapid and slow DG, we have used a unique probe, protein I, a porin that is the major outer membrane protein of Neisseria gonorrhoeae. Treatment of neutrophils with protein I inhibits exocytosis and homotypic cell adhesion provoked by FMLP without inhibiting assembly of the NADPH oxidase responsible for O2-. generation. DG turnover in PMN labeled with [3H]arachidonate and [14C]glycerol was profoundly altered by protein I. Whereas the rapid peak of DG was only modestly diminished (FMLP vs. FMLP plus protein I = DG labeled with [3H]arachidonic acid (3H-a.a.-DG): 142 +/- 14% SEM vs. 125 +/- 22%; DG labeled with the glycerol backbone with [14C]glycerol (D-14C-G): 125 +/- 10% SEM vs. 107 +/- 8.5% SEM), the slow rise in both 3H-a.a.-DG and D-14C-G was essentially abolished. Moreover, treatment of neutrophils with 4-4'-diisothiocyanatostilbene-2,2'-disulfonic acid (DIDS), which, like protein I, inhibits exocytosis without affecting O2-. generation also inhibited slow DG. However, protein phosphorylation and dephosphorylation (47phox, 66phox) were unaffected in the absence of slow DG. To determine the source of the slow DG, we have analyzed radiolabeled phospholipid (PL) turnover after FMLP +/- protein I (P.I.). Treatment of PMN with FMLP (0.1 microM) resulted in breakdown of phosphatidylcholine (PC), beginning at 30 s, and reaching a nadir at 60 s (3H-PC = 59 +/- 10.2% SEM of resting, 14C-PC = 57 +/- 6.4%). Protein I (0.25 microM) significantly inhibited PC turnover after FMLP ([3H]PC = 95 +/- 5.6% and [14C]PC = 86 +/- 8.4% of resting at 60 s), but failed to alter the metabolism of 3H- or 14C-phosphatidylinositol after FMLP (91 +/- 19.6 and 88 +/- 16.5% vs. 92 +/- 9.2 and 91 +/- 16% at 60 s).(ABSTRACT TRUNCATED AT 400 WORDS)

Studies on the role of microtubule integrity in stimulus-response coupling in neutrophils have generated contradictory data. To determine the role of microtubule integrity in stimulus-response coupling elicited by two different mechanisms, i.e., engagement of the Fc receptors (FcR gamma II, FcR gamma III) or engagement of the receptor for FMLP, we utilized colchicine (10 microM), which reduces pericentriolar microtubules to 29% of control, and compared its effect with that of nocodazole (50 microM) and lumicolchicine (10 microM). We now demonstrate that treatment of neutrophils with colchicine but not lumicolchicine, inhibits degranulation elicited by engagement of Fc receptors but augments degranulation in response to FMLP. In contrast to the ligand-specific effect of microtubule-disruption on degranulation, superoxide anion production (assembly of the NADPH oxidase) is unaffected by colchicine regardless of the ligand. To determine whether intact microtubules were required for responses elicited by ligation of Fc gamma RII(CD32) or Fc gamma RIII(CD16), mAb directed against these receptors were employed. Treatment of neutrophils with mAb KuFc79 directed against Fc gamma RII(CD32) or mAb 3G8 directed against Fc gamma RIII(CD16) inhibited degranulation of neutrophils elicited by immune complexes (IC). In contrast, removal of most of Fc gamma RIII by phosphatidylinositol-specific phospholipase C did not significantly alter degranulation in response to IC. We conclude that degranulation elicited by IC results from ligation of both Fc gamma RII and phosphatidylinositol-specific phospholipase C-insensitive Fc gamma RIII. The importance of microtubule integrity on the generation of intracellular signals was also examined. Degranulation of neutrophils proceeds via pertussis toxin-sensitive and insensitive pathways; treatment of cells with colchicine did not augment the action of pertussis toxin. Stimulation of neutrophils by chemoattractants results in a biphasic increase in 1,2-sn-diacylglycerol; a rapid increase ('triggering') secondary to the action of a phosphatidylinositol-specific phospholipase C, and a late increase ('activation') secondary to the action of a phosphatidylcholine-specific phospholipase C. Treatment of cells with colchicine altered the production of both [3H]-arachidonic acid-diacylglycerol and diacyl[14C]glycerol in parallel to its effect on degranulation. These studies indicate that the requirement of intact microtubules for degranulation is ligand-specific. Furthermore, assembly of the respiratory burst oxidase does not require intact microtubules. Microtubules most likely alter the cycling of specific receptors or the generation of specific intracellular signals required for stimulus-response coupling in the course of degranulation. Intact microtubules are not uniformly required for the discharge of granule contents during exocytosis

Cocaine and its derivatives blunted responses of neutrophils (cell/cell aggregation, up-regulation of the receptor for C3bi (CR3, CD11b/CD18), generation of superoxide anion (O2-) and degranulation to various stimuli. The order of potency of these agents was the same as that for local anesthesia: tetracaine greater than bupivacaine greater than cocaine greater than lidocaine. Neutrophil aggregation elicited by the chemoattractant FMLP (10(-7) M) was inhibited by cocaine (10 mM) to 13.6 +/- 6% of control (p less than 0.002); the IC50 was approximately 4 mM. Cocaine and the other local anesthetics not only inhibited the upregulation of CR3 and O2- generation, but also blocked degranulation of cytochalasin B-treated cells. Cocaine (10 mM) reduced beta-glucuronidase and lysozyme secretion to 4.3 +/- 0.7 and 13 +/- 2.2% controls, respectively; its IC50 was 4 mM. Local anesthetics added after ligand/receptor engagement (FMLP) interrupted aggregation and halted generation of O2-. Moreover, local anesthetics rapidly inhibited aggregation, O2- generation, and degranulation elicited by PMA (1 microgram/ml) or the Ca ionophore A23187 (10 microM): the effects of cocaine could therefore not be attributed to unique actions at the FMLP receptor. Peak levels of intracellular Ca2+ ([Ca]i) at 5 to 10 s, and levels of [Ca]i 120 s after FMLP in Fura 2-loaded cells were significantly lower in cells treated with lidocaine, findings that could be explained by enhanced 45Ca2+ efflux from neutrophils. In cells loaded with bis(carboxyethyl)carboxyfluorescine (pH indicator) local anesthetics failed to affect the initial FMLP-induced (0 to 15 s) drop of pHi but inhibited the later (120 s) realkalinization of the cytosol (lidocaine, bupivacaine). Most remarkably, autoradiographs of SDS gels prepared from stimulated, 32P-labeled neutrophils treated with local anesthetics showed no difference from resting cells, either with respect to patterns of phosphorylation and dephosphorylation or their kinetics. Labeling of a 47-kDa protein, a component of the reduced nicotinamide-adenine dinucleotide phosphate-oxidase system, was unchanged. The effects of local anesthetics, which blunt neutrophil responses without affecting protein phosphorylation, suggest that protein phosphorylation is an insufficient signal for neutrophil activation. Inasmuch as cocaine and its derivatives affect cell functions at sites distal to activation of protein kinase C, these agents should prove useful in uncoupling protein phosphorylation from functional responses

We sought to determine whether the activation event which renders the CD11/CD18 leukocyte integrin/Leu-CAM glycoproteins capable of promoting cell to cell adhesion was associated with the induced posttranslational modification of phosphorylation. In neutrophils, two species of alpha-chains, a predominant CD11b 165-kDa subunit and a minor 150-kDa CD11c subunit were found to be constitutively phosphorylated. However, the 95-kDa CD18 common beta-chain was not phosphorylated in resting cells but became strongly phosphorylated in cells incubated with PMA. The beta-chain was phosphorylated in a dose-related manner within 1 min of the addition of PMA, reached maximal intensity between 4 to 10 min, and remained fully phosphorylated at 30 min. The similarities of the kinetics of homotypic aggregation induced by PMA to the time course of phosphorylation suggest that phosporylation may be relevant to at least this type of Leu-CAM-dependent adhesion. In contrast, in the presence of FMLP which induces aggregation with different kinetics than PMA, no phosphorylation of the common beta-chain was observed over a time interval of from 30 s to 10 min further emphasizing the apparent differences in the two modes of activation to an adhesive state. The phosphorylated species on neutrophils were readily detected by immunoprecipitation with each CD18 mAb and most but not all CD11b mAb which otherwise precipitated 125I-labeled CD11b species suggesting that the CD11b alpha-chain labelled with 32P may differ slightly from the 125I-labeled species in terms of its recognition by certain CD11b mAb. In mononuclear cells, similar constitutive phosphorylation of the CD11a and CD11c alpha-chains was observed that remained unchanged in the presence of either FMLP or PMA. As was demonstrated in neutrophils, phosphorylation of the CD18 beta-chains of mononuclear cells was not constitutive but was induced in the presence of PMA and not FMLP. Taken together these data suggest the existence of specific recognition sites on beta-chains for a regulatory kinase-phosphatase system

Upon activation neutrophils release reactive oxygen intermediates such as superoxide anion (O2-) which are potent mediators of inflammation. Various agents elicit different responses; N-formylmethionylleucylphenylalanine (fMLP) (0.1 microM) provokes brisk generation of superoxide anion; leukotriene B4 (LTB4, 0.1 microM) is a poor stimulus. In contrast, phorbol myristate acetate (PMA, 1.6 microM) acting directly via protein kinase C is a potent stimulus for O2-. We compared the kinetics of appearance of various 'second messengers' with the capacity of these ligands to elicit O2- generation. Kinetic analysis showed a two-phase response to membrane ligands; both an 'early' (less than or equal to 15 s) and a 'late' (greater than 15 s) increase in [3H]- and [14C]diacylglycerol (DG) was noted in response to fMLP. In contrast, LTB4 elicited only a rapid early increase in DG. The rise in DG evoked by PMA was late. Cytochalasin B increased the late phase of DG labeling elicited by all agonists. Moreover, comparison of increases in [3H]DG versus those of [14C]DG at early and late time points suggested that DG was not formed exclusively from the hydrolysis of polyphosphoinositides. Early increments of DG were also accompanied by addition of plasma membrane (ultrastructural morphometry); the ratio of surface perimeter to area increased rapidly (10 s) and persisted (60 s) in response to fMLP. Increments were more gradual in response to PMA. Kinetic analysis of protein phosphorylation was compared to the early and late increments of DG labeling. A 47,000 Mr protein was phosphorylated with kinetics consistent with the production of O2- and DG in response to fMLP (early and late) and PMA (late). In contrast, LTB4 provoked only early phosphorylation of this protein. The temporal pattern of the formation of diacylglycerol and the phosphorylation of proteins describe a dual signal. The data suggest that neutrophils require not only 'triggering' (the rapid generation of a signal) but also 'activation' (the maintenance of a signal) to sustain responses