Cellular Signalling Vol. 3, No. 3, pp. 259-266, 1991. Printed in Great Britain.
0898-6568/91 $3.00+.00 © 1991 Pergamon Press plc
DEFECTIVE PROTEIN KINASE C-MEDIATED ACTIONS IN CYSTIC FIBROSIS NEUTROPHILS ISABELLEGRAFF, ANGI~LESCHRAM-DOUMONTand CLAUDE SZPIRER* Universit6 Libre de Bruxelles, D6partement de Biologie Mol6culaire, Rue des Chevaux, 67, B-1640 Rhode-St-Gen•se, Belgium (Received 19 December 1990; and accepted 4 February 1991)
Abstract--Neutrophils from cystic fibrosis (CF) patients have been shown previously to be defective in their response (fl-glucuronidase exocytosis, NADPH oxidase activation) to the chemotactic peptide FMLP. In this work, we attempted to identify the defective step in this response. We showed that stimulated CF and control neutrophils do not differ in the formation of inositol phosphates. On the other hand, direct stimulation of protein kinase C with phorbol myristate acetate (PMA) revealed a subnormal stimulation of fl-glucuronidase exocytosis in CF neutrophils. Furthermore, retroinhibition exerted by PMA-activated protein kinase C on stimulated inositol phosphates or on fl-glucuronidase exocytosis was marginal or absent in CF neutrophils, whereas it was significant in the case of control neutrophils. Our observations suggest that the CFTR gene is expressed in neutrophils and is involved in protein kinase C-mediated actions. Key words: Cystic fibrosis, neutrophils, inositol phosphates, fl-glucuronidase release, protein kinase C.
INTRODUCTION
protein is structurally related to several proteins involved in transport and to some serinethreonine kinase proto-oncogene products [4, 7]. The role played by the C F T R protein in cellular processes has not been elucidated. CF-associated alterations in Cl- and Na + transports [8] and in mucin sulphation [9] have led to the suggestion that C F T R controls several cellular events [6. 10, l 1]. Neutrophils share with exocrine cells the inositol trisphosphate (IP3) and diacylglycerol (DAG)-dependent activation pathways [12-14]. In the presence of cytochalasin B, neutrophils react to specific stimuli with exocytosis of their granular contents; neutrophils can thus be transformed in exocrine cells. An earlier study from this laboratory [15] showed that neutrophils from CF patients react subnormally to stimulation by the chemotactic peptide F M L P (formyl-Methionyl-Leucyl-Phenylalanine): they exhibit lower fl-glucuronidase (a marker of azurophilic granules) exocytosis and NADPH oxidase-dependent chemiluminescence. Both responses are in part mediated by the IP3/DAG
CYSTIC FIBROSIS(CF) is the most common lethal or semi-lethal hereditary disease among Caucasians. CF is transmitted in an autosomal recessive pattern and is characterized by a general dysfunction of exocrine glands with chronic obstructive pulmonary disease and pancreatic exocrine deficiency as key clinical features. The defective gene, called C F T R (cystic fibrosis transmembrane conductance regulator) was localized to chromosome region 7q31 [1-3] and was cloned recently [4]. Complementation of the CF defect has been achieved in cultured pancreatic or respiratory cells by introduction of full-length C F T R cDNA [5, 6]. The C F T R *Author to whom correspondence should be addressed. Abbreviations: CF---cystic fibrosis; CFTR---cystic fibrosis transmembrane conductance regulator; FMLP--formylmethionyl-leucyl-phenylalanine; DAG---diacylglycerol; PMA--phorbol-myfistate-acetate; Ins--inositol; IP--inositol phosphates; IPt--inositol monophosphates; IP2--inositol bisphosphates; IPa--inositol trisphosphates; IP4--inositol tetrakisphosphates; OPI--glycerophosphoinositol.
259
260
I. GV,AFF et al.
pathway [16-18], but the precise localization of the defect exhibited by the CF neutrophils is unknown. The present paper explores three aspects of the IP3/DAG pathway, comparatively in CF and non-CF neutrophils, i.e.: (a) the formation of inositol phosphates upon F M L P stimulation, (b) the stimulating effect of PMA (phorbolmyristate-acetate, a synthetic D A G analogue which activates directly the protein kinase C [19] on ~-glucuronidase exocytosis, (c) the PMA-induced inhibitory effect of protein kinase C on the subsequent F M L P or N a F (a direct activator of the G-protein coupled with the F M L P receptor; [20]) stimulated IP production and fl-glucuronidase release [21-24].
MATERIALS AND METHODS
lysis of residual erythrocytes, and several washings with normal saline. Unless used for enzyme assays the final cell pellet (95-97% neutrophils; 5-3% eosinophils) was resuspended in 1 ml TC 199 culture medium (pH 7.4) supplemented with 75 pCi of [3H]inositol and incubated at 37°C under an atmosphere of Or/CO: (95 : 5; v/v) for 2.5 h. At the end of this labelling period, 96% of cells were found to be viable by the Trypan Blue exclusion test. After three washings in cold TC199 culture medium, the neutrophils were resuspended in a buffered (pH 7.4) salt solution (NaCi 124 mM; KCI 4 mM; CaCI: 1.3 mM; MgCI2 l mM; NazHPO4 0.64mM; KH2PO4 0.66mM; Hepes 10 mM; glucose 5.6 mM) in the presence of LiCI 10mM and incubated for an equilibration period of 15 min at the end of which cytochalasin B (2.5 #g/ml) was added. The agonists were then added and tested for various lengths of time. Biological activity was stopped by the addition of ice-cold perchioric acid (PCA) 4.5% (v/v) and vigorous vortex mixing.
Materials
Extraction and separation o f ~H]inositol phosphates
Myo-[2-3H]inositol (16.5 Ci/mmoi) was purchased from New England Nuclear (Dupont-NEN, Harcn, Belgium). BSA (Albumin, Bovine fraction V), FMLP (N-Formyl-L-Methionyl-L-Leucyl-L-Phenyl alanine), PMA (phorbol 12-myristate 13-acetate) and 4-methylumlmlliferyl-//-D-glucuronide were purchased from Sigma (St Louis, MO, USA); cytochalasin B from Aldrich (Milwaukee, WI, USA); Lymphoprep from Nyegaard (Oslo, Norway); Plasmagel from Laboratoires Bellon (Neuilly sur Seine, France); TC 199 culture medium from Flow-Laboratories (Irvine, Scotland) and DOWEX AGI-X8 ionexchange resin (formate form, 100-200 mesh) from Bio-Rad (Watford, England). Water-insoluble substances were dissolved in dimethyl sulphoxide (Merck-Darmstadt, Germany) to make appropriate stock solutions giving final concentrations of 0.1% (v/v) DMSO in the reaction mixtures. All other reagents were of the purest grade commercially available.
PCA extracts were alkalinized to pH 8-8.5 by addition of a KOH 0.5 M/Na2B,O7 9 mM/NaOH 3.8 mM/EDTA 1.9 mM solution [13]. After potassium perchlorate precipitation, the supernatants were frozen at -20°C or directly applied to (0.6 x 4.0 cm)AG1-X8 columns preequilibrated with 15 ml of ammonium formate 50 mM/inositol 5 mM. Ins, GPI, IPl, IP2, IP3 and IP4 were eluted in a stepwise procedure, slightly modified from Downes and Michell [26], using for each fraction successive volumes of: 20 ml of (HCOONH4 50 mM/Ins 5 raM), 8 ml of (60 mM HCOONH4/5 mM Na2B407), 36 ml of (150raM HCOONHJ5 mM Na2B407), 28 ml of (400 mM HCOONH4/100 mM HCOOH), 16 ml of (800 mM HCOONHJ100 mM HCOOH) and 16 ml of (1.2 M HCOONHJI00 mM HCOOH). Each fraction (4 ml) was sampled for the determination of radioactivity in Instagel scintillation fluid. Results are expressed as dpm/12x 106 neutrophils. Each experimental condition was tested in two parallel but separate incubations and chromatographic runs, the results of which were averaged. On each experimental day, samples from one CF patient and from one non-CF patient were processed under strictly comparative conditions. The representativeness of the CF sample was checked by its lesser release of fl-glucuronidase in the presence of FMLP [151.
Neutrophils preparation
Heparinized blood (15-30ml) was collected by venipuncture from CF and non-CF patients [H6pital de Braine rAlleud, Belgium (Dr Van Geffel) and H6pital Erasme, (Dr Baran)]. Informed consent was obtained in all cases. PMN were obtained following a procedure adapted from Boyum [25]. Ficoll-metrizoate (Lymphoprep) density gradient centrifugation of heparinized blood was followed by sedimentation of erythrocytes in presence of plasmagel, hypotonic
Incubation conditions and enzyme assays
After washing with normal saline, neutrophils (2x 106cells/ml) were suspended in buffered salt
Defectiveprotein kinase C-mediatedactions in cysticfibrosis neutrophils solution (pH 7.4) (see section on neutrophil preparation), allowed to stay for 30 min at room temperature and then for 15 min at 37°C in a shaking water bath prior to stimulation, according to the various protocols described in the legends and in the results. After incubation, the samples were centrifuged and the clear supernatants assayed for fl-glucuronidase. The enzyme activity was measured fluorimetrically using 4-methylumbelliferyl-fl-D-glucuronide as substrate. Results are expressed as the percentage of total cellular enzyme activity obtained when cells were completely lysed with 0.2% (v/v) Triton X-100 [15, 27].
261
4OO
3OO
| IO0 40
RESULTS CF and non-CF neutrophils do not differ in IP production upon FMLP or N a F stimulation The differential white blood cell counts of both CF and non-CF patients were unremarkable and within accepted limits. As expected [15], in the presence of cytochalasin B, FMLPinduced exocytosis of fl-glucuronidase was significantly lower in CF neutrophils: 32% +S.E.M. 1.6 of total cell content, versus 4 8 % + 2 . 4 in non-CF neutrophils (n = 14; P < 0.001). As preliminary controls, the effects of cytochalasin B and of LiCI on inositol phosphates metabolism of neutrophils were tested. Cytochalasin B (2.5/~g/ml) is routinely added to the incubation medium to transform the neutrophils in exocrine cells, and LiC1 is used to block the IP~ phosphatase [28]. In that way, the IP~ fraction acts as the final collecting pool for the IP 3 originally released from PIP 2 under stimulation. We found that cytochalasin B had no significant effect on the IP metabolism of CF or non-CF neutrophils. LiCI (10 mM) reduced the incorporation of [~H]inositol in the individual inositol phosphates fractions (IP~, IP 2, IPa, IP4) of unstimulated neutrophils (basal state) to two-thirds of the control values; however, in its absence, no significant response to F M L P stimulation could be detected. The basal state of IP fractions in CF and non-CF neutrophils was also compared. Paired data analysis failed to detect any significant
80
120
160
200
240
280
Time aftra' 10-6 M FMLP ~rnulation (~sc)
FIG. 1. IP generation time curves in response to 10-6M FMLP in neutrophils from a non-CF patient. The results correspond to a single experiment representative of at least three separate experiments. CF neutrophils gave identical results.
difference between 13 non-CF and CF neutrophil pairs as regards [3H]inositol labelling o f individual IP fractions (IP~: 75.5-79.3%; IP2: 16.7-20.4%; IP3: 2.3-2.5%; IP4: 0.8-1%), total inositol phosphates (9.000-9.800 dpm/12 x 10 6 cells) and total inositol cellular content (free inositol, glycerophosphoinositol and IP). Action of FMLP. Upon F M L P treatment (15 s to 5 min) the time course of IP radiolabelling showed that stimulation affected the IP3, IP4, IP2 and IP~ fractions successively (Fig, 1). When stimulation is expressed in relative terms (i.e. ratio stimulated state to basal state), it is maximal in the IP3 fraction and occurs within the first 30 s of exposure. When stimulation is expressed in absolute terms (i.e. stimulated state minus basal state) it is maximal in the IP t fraction, at all poststimulatory intervals. Stimulation was already evident after 30 s exposure to 10 -8 M F M L P in the IP 3 fraction and was maximal for all fractions in the presence of 10-6M FMLP. No effect or a slight inhibition was observed at the highest F M L P concentration tested, 10 -4 M. Twenty-one paired observations made on neutrophils from 13 CF and 13 non-CF
262
I. GRxrvet al.
subjects, under strictly matched conditions of FMLP concentration and incubation time, failed to show any significant difference either in total IP accumulation or in the relative contribution of individual IP (IP t, IP 2, IP 3 and IP+) to the total IP pool. Figure 1 is thus representative of either CF or non-CF neutrophils. Generation of individual IP evolved along similar time-curves in the presence of 10-6M FMLP in both groups and qualitatively and quantitatively similar dose-response curves were obtained in the presence of 10 -8 M to 10-4M FMLP. With 10-6M FMLP, the stimulation factor for total IP was 1.36 to 1.40 and the contribution of each IP fraction were: IP,: 77.2-78%; IP2: 14.7-18.4%; IP3: 2.6-3.1%; IP4: 1.0-1.2%. Action of NaF. Incubation of neutrophils with NaF (18 mM for 20 min) stimulated total IP accumulation similarly in both CF and non-CF groups, with unremarkable contribution of individual IP to total IP pool. As no difference could be found between CF and non-CF neutrophils in the IP generation under FMLP or NaF stimulation, the causative step of the subnormal secretory response of CF neutrophils was assumed to lie downstream from the IP generation and some protein kinase C-mediated reactions were thus explored. CF neutrophils show a reduced #-glucuronidase secretion upon PMA stimulation Notwithstanding the fact that PMA is a poor secretagogne for azurophilic granules [29], it was used to test directly the effect of protein kinase C on neutrophil degranulation. PMA was added after 3 min of pretreatment with cytochalasin, at final concentrations of, respectively, 10 -7 M and 5 #g/ml in Hanks' buffer (see Material and Methods) supplemented with 1 mg/ml BSA to lower the basal release. Samples were incubated for various lengths of time (see Fig. 2). In the absence of cytochalasin, no significant secretion ( < 2%) could be detected, either in non-CF neutrophils or in CF neutrophils. The exocytosis of fl-glucuronidase was
,.__----~ o
ti+o
j
l
t
t~
0
,
0
,
2
,
t?
,
4
i
i
6
~
T
8
,
,
10
n,
~ 1
r
12
I--
14
16
Time after 1~7 M PMA stimulation (min)
FZG. 2. Time-course of #-glucuronidase secretion after PMA stimulation on non-CF and CF neutrophils. (a, b) Basal secretion in the presence of cytochalasin alone. The results correspond to a single experiment representative of at least three separate pairs of samples.
measured after a 10 min incubation of neutrophils in the presence of PMA. After correction for the background release caused by cytochalasin alone, the net release amounted to 6.2% (S.E.M. 1.2; n = 9) in CF and to 10.3% (S.E.M. 1.1; n = 9) in non-CF neutrophils respectively, with P < 0.025. This result suggests that PMAactivated protein kinase C mediates a lower than normal #-glucuronidase exocytosis in CF neutrophils. CF neutrophils fail to exhibit the retroinhibition exerted in non-CF neutrophils by PMAactivated protein kinase C on FMLP- or NaFstimulated IP formation In the basal state, the addition of PMA alone (l.6x 10-SM) had no significant effect on either total IP accumulation or on the relative contribution of individual IP fraction to the total IP pool (non-CF and CF cells). In agreement with earlier studies [21, 22], we found that pre-incubation of non-CF neutrophils with PMA (l.6x 10-aM for 5 min), before addition of FMLP (10 -6 M, 30 s) effectively curtailed the expected increase in IP
Defective protein kinase C-mediated actions in cystic fibrosis neutrophils accumulation over basal conditions (Table 1 A, columns 1, 2, 5). In sharp contrast, prior incubations of C F neutrophils with P M A failed to block the F M L P stimulation of IP formation
263
so ~_~__ non CF
so .~
(Table 1 B, columns 1, 2, 5) When FMLP was
o
tl
replaced by N a F (18 m M for 20 min), prior incubation of neutrophils with PMA (1.6 x 10 -8 M for 5 min)effectively blocked the effect of N a F stimulation in the non-CF neutrophils but failed to do so in the CF neutrophils (Table 1 A and B, columns 3, 4, 6).
~ ~ so h ~ 20 *~ 10
The protein kinase C-mediated inhibition of exocytosis is marginal in the case of CF neutrophils
2
4 6 8 10 Duril~On Of PMA 10 "7M ~ b a l k ~
12
14 (min)
16
FIG. 3. Inhibition of FMLP-induced fl-glucuronidase exocytosis by preincubation with PMA in non-CF and CF neutrophils. In the absence of PMA, the values of fl-glucuronidase secretion were: 47% (non-CF) and 35% (CF). On the other hand, incubation with cytochalasin alone, PMA alone (incubation of 15 min) or PMA (incubation of 15 rain) and cytochalasin, gave values comprised between 10 and 15%. The results correspond to a single experiment representative of at least three separate pairs of samples.
Finally, the protein kinase C-mediated inhibition o f FMLP-induced fl-glucuronidase secretion was also examined. In a preliminary test, the inhibitory effect of P M A (10 -7 M) on F M L P (10 -6 M)-induced secretion was checked as a function of P M A pretreatment length (0-15 rain). Cytochalasin (5/~g/ml) was added five minutes prior to F M L P . The fl-glucuronidase release was measured 5 min after F M L P addition (Fig. 3). It was found that a 15 min preincubation with P M A gave the deeper inhibition (Fig. 3), as had also been observed by Barrowman et al. [24], and was therefore
selected for the comparison of CF and n o n - C F neutrophils (Table 2). F r o m the results summarized in Table 2, it can be concluded
TABLE 1. EFFECTOF PRIOR PMA n'~CUBATIONON FMLP AND NaF-INDUCED IP FORMATION Stimulation with
A. non-CF X S.E.M. B. CF. X S,E.M.
Ratios
FMLP
PMA + FMLP
NaF
PMA + NaF
2/1
4/3
1
2
3
4
5
6
1.39 0.09
1.00 0.14
1.49 0.08
1.08 0.14
0.72 0.07
0.73 0.05
1.19 0.28
1.20 0.09
1.46 0.09
1.46 0.12
1.01 0.04
1.01 0.04
Columns 1 to 4: the stimulation factor (ratio of IP accumulation under stimulation to that in the basal state) in different experimental conditions (1: FMLP 10 -6 M, 30 s; 2: PMA 1.6 x 10 -8 M 5 min + FMLP 10 -6 M, 30 s; 3: NaF 18mM, 20min; 4: PMA 1.6x 10-SM 5 m i n + N a F 18mM, 20min). Columns 5 and 6: the ratios of columns 2 (FMLP+PMA) to 1 (FMLP) and of columns 4 (NaF+PMA) to 3 (NaF). PMA alone had no effect. Paired t-tests show the effect of a prior PMA incubation to be significantly different in the CF and non-CF groups at the P < 0.01 level.
264
I. Groa~ et al.
TABLE 2. EFFECT OF PRIOR P M A INCUBATION ON FMLPqNDUCED ~LUCURONIDASE RELEASE
Stimulation with
Non-CF (n = 6). X S.E.M. CF (n --- 6). X S.E.M.
Ratios
FMLP
PMA + FMLP
2/1
1
2
3
44.2 2.9
33.8 3.9
0.76 0.04
35.4 1.2
32.4 1.0
0.92 0.02
fl-Glucuronidase release (% of total): column 1 in response to FMLP (10-6M); column 2 in response to FMLP (10-eM) after a 15 min preincubation in the presence of PMA (10-7 M); column 3, the ratios of colunm2 (FMLP+PMA) to 1 (FMLP). PMA by itself has a minor stimulatory effect on/~-glucuronidase exocytosis (see section D and Fig. 3). that this PMA mediated-inhibition is significant in non-CF neutrophils (P < 0.05) whereas it is only marginal in cells from CF patients (P < 0.1). The difference between the ratios observed for non-CF and CF subjects is significant at the level P < 0.01.
DISCUSSION Paired analysis of CF and non-CF neutrophils in the presence of the physiologic agonist FMLP failed to show any significant difference in the acumulation of [3H]IP (IPI, IP2, IP3, IP4 and sum of IP) or in their respective dose-response and time curves. Sample size was sufficient to detect a possible difference between means equal to 1.5 times the SD common to the two groups with a power > 0.95 (Type II error, type I error being kept at the 0.05 conventional level). No such difference was found and it is concluded that the difference in IP generation under FMLP stimulation, if any, between CF and non-CF neutrophils is smaller than 1.5 SD and cannot account for the observed differences (FMLP-induced exocytosis and chemiluminescence) between these cells. The defective step lies presumably downstream from the IP generation.
We showed that, upon PMA activation, CF neutrophils release lesser amounts of ~-glucuronidase secretion than non-CF cells. Moreover, the inhibition exerted by PMA pretreatment on the FMLP induced release is only marginal in CF neutrophils. In addition, our observations demonstrate that in CF neutrophils, the retroinbibition exerted by PMA-activated protein kinase C on FMLP- and NaF-stimulated generation of IP fails to materialize. This is in contrast with the response observed in non-CF neutrophils which provides positive evidence of the inhibitory effect commonly exerted by PMA on IP accumulation and granular enzymes secretion [21-24], the mechanism of which remains, however, unknown. In conclusion, the CF primary defect in neutrophils apparently lies downstream from the IP generation, more probably in the immediate vicinity of protein kinase C. Hermelin et al. [30] also suggested that, in CF fibroblasts, the primary defect lies at the level of a regulatory protein or at the level of a phosphorylation cascade (defective substrate or enzyme). Our results could be explained in two ways: either they are secondary effects of the disease (in other words, a non-representative subpopulation of neutrophils not expressing the CFTR gene might be particularly abundant in CF
Defective protein kinase C-mediated actions in cystic fibrosis neutrophils patients), or they reflect the defective activity of the C F T R gene in CF neutrophils. It seems highly unlikely that the defective C F T R gene would cause the accumulation o f a particular subpopulation of neutrophils not expressing the C F T R gene, but deficient in exocytosis and protein kinase C activities. We thus suggest that the C F T R protein is active in neutrophils (or in their precursors) and that in these cells, exists the target, either directly or indirectly, of protein kinase C and participates in the retroinhibition effects exerted by this kinase. That the C F T R protein is the target of protein kinase C (and of protein kinase A) has already been demonstrated in epithelial cells [31, 32]. In addition, the C F T R protein has many sites suitable for phosphorylation by protein kinase C. Moreover, the most frequent D N A mutation observed in the C F T R gene manifests itself at the amino acid 508 o f the polypeptide chain, at the level of an ATPbinding site and in the immediate vicinity of a protein kinase C-phosphorylation site. Our results suggest that protein kinase C might also exert CFTR-mediated retroinhibition effects and it would be interesting to test this possibility in epithelial cells. Expression of the C F T R gene in neutrophils is perhaps not as surprising as it might seem. Indeed, like typical exocrine epithelial cells, neutrophils are secretory cells. It has been suggested that the C F T R protein is a transporter (rather than an ion channel) which binds and transports a molecule that regulates chloride channels [7, 10, 11, 33]. C F T R protein may play a similar role in neutrophils. Acknowledgements--We are grateful to Dr R. VAr~ GEFFEL and Dr D. BARAN for kindly providing us with samples of CF blood. This work was supported by the National Fund for Scientific Research (FNRS, Belgium). I.G. was an IRSIA grantee. C.S. is Senior Research Associate of the FNRS.
REFERENCES 1. Knowlton R., Cohen-Haguenauer O., Van Cong N., Frezal J., Brown V., Barker D., Braman J., Schumm J., Tsui L.-C., Buchwald M. and Donis-Keller H. (1985) Nature 318, 380-382.
265
2. Wainwright B., Scambler P., Sehmidtke J., Watson E., Law H.-Y., Farrall M., Cooke H., Eiberg H. and Williamson R. (1985) Nature 318, 384-385. 3. White R., Woodward S., Leppert M., O'Connell P., Hoff M., Herbst J., Lalouel J.-M., Dean M. and Vande Woude G. (1985) Nature 318, 382-384. 4. Riordan J., Rommens J., Kerem B.-S., Alon N., Rozmahel R., Grzelczak Z., Zielenski J., Lok S., Plavsic N., Chou J.-L., Drumm M., lannuzzi M., Collins F. and Tsui L.-C. (1989) Science 245, 1066-1073. 5. Rich D., Anderson M., Gregory R., Cheng S., Paul S., Jefferson D., MacCann J., Klinger K., Smith A. and Welsh M. (1990) Nature 347, 358-363. 6. Drumm M., Pope H., Cliff W., Rommens J., Marvin S., Tsui L.-C., Collins F., Frizzell R. and Wilson J. (1990) Cell 62, 1227-1233. 7. Hyde S., Emsley P., Hartshorn M., Mimmack M., Gileadi U., Pearce S., Gallagher M., Gill D., Hubbard R. and Higgins C. (1990) Nature 346, 362-365. 8. Quinton P. (1990) FASEB J. 4, 2709-2717. 9. Cheng P., Boat T., Cranfiil K., Yankaskas J. and Boucher R. (1989) J. clin. Invest. 84, 68-72. 10. MacPherson M. and Dormer R. (1988) TIBS 13, 10-13. 11. Welsh M. (1988) lOth Internam. CF Congress, Sydney, Australia, March 1988, Excerpta Medica, Asia Pacific Congress Series, 74. 12. Bradford P. and Rubin R. (1985). Molec. Pharmac. 27, 74-78. 13. Bradford P. and Rubin R. (1985) FEBS Lett. 183, 317-320. 14. Bradford P. and Rubin R. (1986) J. biol. Chem. 261, 15,644-15,647. 15. Kemp T., Schram-Doumont A., Van Geffel R., Kram R. and Szpirer C. (1986) Pediatr. Res. 20, 520-526. 16. White J., Huang J., Naccache P., Becker E. and Sha'afi R. (1984) J. bioL Chem. 259, 8605-861 I. 17. Hayakawa T., Suzuki K., Suzuki S., Andrew P. and Babior B. (1986) J. biol. Chem. 261, 9109-9115. 18. Wright C. and Hoffman M. (1987) Biochem. biophys. Res. Commun. 142, 53-62. 19. Castagna M. (1987) Biol. Cell 59, 3-14. 20. Magnusson M., Halldorson H., Kjeld M. and Thorgeirsson G. (1989) Biochem. J. 264, 703-711. 21. Kato H., Ishitoya J. and Takenawa T. (1986) Biochem. biophys. Res. Commun. 139, 1272-1278. 22. Tyagi S., Tamura M., Burnham D. and Lambeth J. (1988) J. biol. Chem. 263, 13,191-13,198.
266
I. GRAFFet al.
23. Naccache P., Molski T., Borgeat P., White J. and Sha'afi R. (1985) J. biol. Chem. 260, 2125-2131. 24. Barrowman M., Cockcroft S. and Gomperts B. (1986) FEBS Lett. 201, 137-142. 25. B6yum A. (1986) Scand. J. clin. Invest. 22 (suppl. 97): 77-89. 26. Downes C. and Michell P. (1981) Biochem. J. 198, 133-140. 27. Dewald B., Bretz U. and Baggiolini M. (1982) J. clin. Invest. 70, 518-525. 28. Hallcher L. and Sherman W. (1980) J. biol. Chem. 255, 10,896-10,901.
29. Smith R., Justen J. and Sam L. (1988) Biochem. biophys. Res. Commun. 152, 1497-1503. 30. Hermelin B., Cherqui G., Bertrand F., Wicek D., Paul A., Garcia I. and Picard J. (1988) FEBS Left. 229, 161-166. 31. Schoumaker R., Shoemaker R., Haim D., Tallant E., Wallace R. and Frizzell R. (1987) Nature 330, 752-754. 32. Hwang T., Lu L., Zeitlin P., Gruenert D., Huganir R. and Guggino W. (1989) Science 244, 1351-1353. 33. Ringe D. and Petsko G. (1990) Nature 346, 312-313.
0898-6568/91 $3.00+.00 © 1991 Pergamon Press plc
DEFECTIVE PROTEIN KINASE C-MEDIATED ACTIONS IN CYSTIC FIBROSIS NEUTROPHILS ISABELLEGRAFF, ANGI~LESCHRAM-DOUMONTand CLAUDE SZPIRER* Universit6 Libre de Bruxelles, D6partement de Biologie Mol6culaire, Rue des Chevaux, 67, B-1640 Rhode-St-Gen•se, Belgium (Received 19 December 1990; and accepted 4 February 1991)
Abstract--Neutrophils from cystic fibrosis (CF) patients have been shown previously to be defective in their response (fl-glucuronidase exocytosis, NADPH oxidase activation) to the chemotactic peptide FMLP. In this work, we attempted to identify the defective step in this response. We showed that stimulated CF and control neutrophils do not differ in the formation of inositol phosphates. On the other hand, direct stimulation of protein kinase C with phorbol myristate acetate (PMA) revealed a subnormal stimulation of fl-glucuronidase exocytosis in CF neutrophils. Furthermore, retroinhibition exerted by PMA-activated protein kinase C on stimulated inositol phosphates or on fl-glucuronidase exocytosis was marginal or absent in CF neutrophils, whereas it was significant in the case of control neutrophils. Our observations suggest that the CFTR gene is expressed in neutrophils and is involved in protein kinase C-mediated actions. Key words: Cystic fibrosis, neutrophils, inositol phosphates, fl-glucuronidase release, protein kinase C.
INTRODUCTION
protein is structurally related to several proteins involved in transport and to some serinethreonine kinase proto-oncogene products [4, 7]. The role played by the C F T R protein in cellular processes has not been elucidated. CF-associated alterations in Cl- and Na + transports [8] and in mucin sulphation [9] have led to the suggestion that C F T R controls several cellular events [6. 10, l 1]. Neutrophils share with exocrine cells the inositol trisphosphate (IP3) and diacylglycerol (DAG)-dependent activation pathways [12-14]. In the presence of cytochalasin B, neutrophils react to specific stimuli with exocytosis of their granular contents; neutrophils can thus be transformed in exocrine cells. An earlier study from this laboratory [15] showed that neutrophils from CF patients react subnormally to stimulation by the chemotactic peptide F M L P (formyl-Methionyl-Leucyl-Phenylalanine): they exhibit lower fl-glucuronidase (a marker of azurophilic granules) exocytosis and NADPH oxidase-dependent chemiluminescence. Both responses are in part mediated by the IP3/DAG
CYSTIC FIBROSIS(CF) is the most common lethal or semi-lethal hereditary disease among Caucasians. CF is transmitted in an autosomal recessive pattern and is characterized by a general dysfunction of exocrine glands with chronic obstructive pulmonary disease and pancreatic exocrine deficiency as key clinical features. The defective gene, called C F T R (cystic fibrosis transmembrane conductance regulator) was localized to chromosome region 7q31 [1-3] and was cloned recently [4]. Complementation of the CF defect has been achieved in cultured pancreatic or respiratory cells by introduction of full-length C F T R cDNA [5, 6]. The C F T R *Author to whom correspondence should be addressed. Abbreviations: CF---cystic fibrosis; CFTR---cystic fibrosis transmembrane conductance regulator; FMLP--formylmethionyl-leucyl-phenylalanine; DAG---diacylglycerol; PMA--phorbol-myfistate-acetate; Ins--inositol; IP--inositol phosphates; IPt--inositol monophosphates; IP2--inositol bisphosphates; IPa--inositol trisphosphates; IP4--inositol tetrakisphosphates; OPI--glycerophosphoinositol.
259
260
I. GV,AFF et al.
pathway [16-18], but the precise localization of the defect exhibited by the CF neutrophils is unknown. The present paper explores three aspects of the IP3/DAG pathway, comparatively in CF and non-CF neutrophils, i.e.: (a) the formation of inositol phosphates upon F M L P stimulation, (b) the stimulating effect of PMA (phorbolmyristate-acetate, a synthetic D A G analogue which activates directly the protein kinase C [19] on ~-glucuronidase exocytosis, (c) the PMA-induced inhibitory effect of protein kinase C on the subsequent F M L P or N a F (a direct activator of the G-protein coupled with the F M L P receptor; [20]) stimulated IP production and fl-glucuronidase release [21-24].
MATERIALS AND METHODS
lysis of residual erythrocytes, and several washings with normal saline. Unless used for enzyme assays the final cell pellet (95-97% neutrophils; 5-3% eosinophils) was resuspended in 1 ml TC 199 culture medium (pH 7.4) supplemented with 75 pCi of [3H]inositol and incubated at 37°C under an atmosphere of Or/CO: (95 : 5; v/v) for 2.5 h. At the end of this labelling period, 96% of cells were found to be viable by the Trypan Blue exclusion test. After three washings in cold TC199 culture medium, the neutrophils were resuspended in a buffered (pH 7.4) salt solution (NaCi 124 mM; KCI 4 mM; CaCI: 1.3 mM; MgCI2 l mM; NazHPO4 0.64mM; KH2PO4 0.66mM; Hepes 10 mM; glucose 5.6 mM) in the presence of LiCI 10mM and incubated for an equilibration period of 15 min at the end of which cytochalasin B (2.5 #g/ml) was added. The agonists were then added and tested for various lengths of time. Biological activity was stopped by the addition of ice-cold perchioric acid (PCA) 4.5% (v/v) and vigorous vortex mixing.
Materials
Extraction and separation o f ~H]inositol phosphates
Myo-[2-3H]inositol (16.5 Ci/mmoi) was purchased from New England Nuclear (Dupont-NEN, Harcn, Belgium). BSA (Albumin, Bovine fraction V), FMLP (N-Formyl-L-Methionyl-L-Leucyl-L-Phenyl alanine), PMA (phorbol 12-myristate 13-acetate) and 4-methylumlmlliferyl-//-D-glucuronide were purchased from Sigma (St Louis, MO, USA); cytochalasin B from Aldrich (Milwaukee, WI, USA); Lymphoprep from Nyegaard (Oslo, Norway); Plasmagel from Laboratoires Bellon (Neuilly sur Seine, France); TC 199 culture medium from Flow-Laboratories (Irvine, Scotland) and DOWEX AGI-X8 ionexchange resin (formate form, 100-200 mesh) from Bio-Rad (Watford, England). Water-insoluble substances were dissolved in dimethyl sulphoxide (Merck-Darmstadt, Germany) to make appropriate stock solutions giving final concentrations of 0.1% (v/v) DMSO in the reaction mixtures. All other reagents were of the purest grade commercially available.
PCA extracts were alkalinized to pH 8-8.5 by addition of a KOH 0.5 M/Na2B,O7 9 mM/NaOH 3.8 mM/EDTA 1.9 mM solution [13]. After potassium perchlorate precipitation, the supernatants were frozen at -20°C or directly applied to (0.6 x 4.0 cm)AG1-X8 columns preequilibrated with 15 ml of ammonium formate 50 mM/inositol 5 mM. Ins, GPI, IPl, IP2, IP3 and IP4 were eluted in a stepwise procedure, slightly modified from Downes and Michell [26], using for each fraction successive volumes of: 20 ml of (HCOONH4 50 mM/Ins 5 raM), 8 ml of (60 mM HCOONH4/5 mM Na2B407), 36 ml of (150raM HCOONHJ5 mM Na2B407), 28 ml of (400 mM HCOONH4/100 mM HCOOH), 16 ml of (800 mM HCOONHJ100 mM HCOOH) and 16 ml of (1.2 M HCOONHJI00 mM HCOOH). Each fraction (4 ml) was sampled for the determination of radioactivity in Instagel scintillation fluid. Results are expressed as dpm/12x 106 neutrophils. Each experimental condition was tested in two parallel but separate incubations and chromatographic runs, the results of which were averaged. On each experimental day, samples from one CF patient and from one non-CF patient were processed under strictly comparative conditions. The representativeness of the CF sample was checked by its lesser release of fl-glucuronidase in the presence of FMLP [151.
Neutrophils preparation
Heparinized blood (15-30ml) was collected by venipuncture from CF and non-CF patients [H6pital de Braine rAlleud, Belgium (Dr Van Geffel) and H6pital Erasme, (Dr Baran)]. Informed consent was obtained in all cases. PMN were obtained following a procedure adapted from Boyum [25]. Ficoll-metrizoate (Lymphoprep) density gradient centrifugation of heparinized blood was followed by sedimentation of erythrocytes in presence of plasmagel, hypotonic
Incubation conditions and enzyme assays
After washing with normal saline, neutrophils (2x 106cells/ml) were suspended in buffered salt
Defectiveprotein kinase C-mediatedactions in cysticfibrosis neutrophils solution (pH 7.4) (see section on neutrophil preparation), allowed to stay for 30 min at room temperature and then for 15 min at 37°C in a shaking water bath prior to stimulation, according to the various protocols described in the legends and in the results. After incubation, the samples were centrifuged and the clear supernatants assayed for fl-glucuronidase. The enzyme activity was measured fluorimetrically using 4-methylumbelliferyl-fl-D-glucuronide as substrate. Results are expressed as the percentage of total cellular enzyme activity obtained when cells were completely lysed with 0.2% (v/v) Triton X-100 [15, 27].
261
4OO
3OO
| IO0 40
RESULTS CF and non-CF neutrophils do not differ in IP production upon FMLP or N a F stimulation The differential white blood cell counts of both CF and non-CF patients were unremarkable and within accepted limits. As expected [15], in the presence of cytochalasin B, FMLPinduced exocytosis of fl-glucuronidase was significantly lower in CF neutrophils: 32% +S.E.M. 1.6 of total cell content, versus 4 8 % + 2 . 4 in non-CF neutrophils (n = 14; P < 0.001). As preliminary controls, the effects of cytochalasin B and of LiCI on inositol phosphates metabolism of neutrophils were tested. Cytochalasin B (2.5/~g/ml) is routinely added to the incubation medium to transform the neutrophils in exocrine cells, and LiC1 is used to block the IP~ phosphatase [28]. In that way, the IP~ fraction acts as the final collecting pool for the IP 3 originally released from PIP 2 under stimulation. We found that cytochalasin B had no significant effect on the IP metabolism of CF or non-CF neutrophils. LiCI (10 mM) reduced the incorporation of [~H]inositol in the individual inositol phosphates fractions (IP~, IP 2, IPa, IP4) of unstimulated neutrophils (basal state) to two-thirds of the control values; however, in its absence, no significant response to F M L P stimulation could be detected. The basal state of IP fractions in CF and non-CF neutrophils was also compared. Paired data analysis failed to detect any significant
80
120
160
200
240
280
Time aftra' 10-6 M FMLP ~rnulation (~sc)
FIG. 1. IP generation time curves in response to 10-6M FMLP in neutrophils from a non-CF patient. The results correspond to a single experiment representative of at least three separate experiments. CF neutrophils gave identical results.
difference between 13 non-CF and CF neutrophil pairs as regards [3H]inositol labelling o f individual IP fractions (IP~: 75.5-79.3%; IP2: 16.7-20.4%; IP3: 2.3-2.5%; IP4: 0.8-1%), total inositol phosphates (9.000-9.800 dpm/12 x 10 6 cells) and total inositol cellular content (free inositol, glycerophosphoinositol and IP). Action of FMLP. Upon F M L P treatment (15 s to 5 min) the time course of IP radiolabelling showed that stimulation affected the IP3, IP4, IP2 and IP~ fractions successively (Fig, 1). When stimulation is expressed in relative terms (i.e. ratio stimulated state to basal state), it is maximal in the IP3 fraction and occurs within the first 30 s of exposure. When stimulation is expressed in absolute terms (i.e. stimulated state minus basal state) it is maximal in the IP t fraction, at all poststimulatory intervals. Stimulation was already evident after 30 s exposure to 10 -8 M F M L P in the IP 3 fraction and was maximal for all fractions in the presence of 10-6M FMLP. No effect or a slight inhibition was observed at the highest F M L P concentration tested, 10 -4 M. Twenty-one paired observations made on neutrophils from 13 CF and 13 non-CF
262
I. GRxrvet al.
subjects, under strictly matched conditions of FMLP concentration and incubation time, failed to show any significant difference either in total IP accumulation or in the relative contribution of individual IP (IP t, IP 2, IP 3 and IP+) to the total IP pool. Figure 1 is thus representative of either CF or non-CF neutrophils. Generation of individual IP evolved along similar time-curves in the presence of 10-6M FMLP in both groups and qualitatively and quantitatively similar dose-response curves were obtained in the presence of 10 -8 M to 10-4M FMLP. With 10-6M FMLP, the stimulation factor for total IP was 1.36 to 1.40 and the contribution of each IP fraction were: IP,: 77.2-78%; IP2: 14.7-18.4%; IP3: 2.6-3.1%; IP4: 1.0-1.2%. Action of NaF. Incubation of neutrophils with NaF (18 mM for 20 min) stimulated total IP accumulation similarly in both CF and non-CF groups, with unremarkable contribution of individual IP to total IP pool. As no difference could be found between CF and non-CF neutrophils in the IP generation under FMLP or NaF stimulation, the causative step of the subnormal secretory response of CF neutrophils was assumed to lie downstream from the IP generation and some protein kinase C-mediated reactions were thus explored. CF neutrophils show a reduced #-glucuronidase secretion upon PMA stimulation Notwithstanding the fact that PMA is a poor secretagogne for azurophilic granules [29], it was used to test directly the effect of protein kinase C on neutrophil degranulation. PMA was added after 3 min of pretreatment with cytochalasin, at final concentrations of, respectively, 10 -7 M and 5 #g/ml in Hanks' buffer (see Material and Methods) supplemented with 1 mg/ml BSA to lower the basal release. Samples were incubated for various lengths of time (see Fig. 2). In the absence of cytochalasin, no significant secretion ( < 2%) could be detected, either in non-CF neutrophils or in CF neutrophils. The exocytosis of fl-glucuronidase was
,.__----~ o
ti+o
j
l
t
t~
0
,
0
,
2
,
t?
,
4
i
i
6
~
T
8
,
,
10
n,
~ 1
r
12
I--
14
16
Time after 1~7 M PMA stimulation (min)
FZG. 2. Time-course of #-glucuronidase secretion after PMA stimulation on non-CF and CF neutrophils. (a, b) Basal secretion in the presence of cytochalasin alone. The results correspond to a single experiment representative of at least three separate pairs of samples.
measured after a 10 min incubation of neutrophils in the presence of PMA. After correction for the background release caused by cytochalasin alone, the net release amounted to 6.2% (S.E.M. 1.2; n = 9) in CF and to 10.3% (S.E.M. 1.1; n = 9) in non-CF neutrophils respectively, with P < 0.025. This result suggests that PMAactivated protein kinase C mediates a lower than normal #-glucuronidase exocytosis in CF neutrophils. CF neutrophils fail to exhibit the retroinhibition exerted in non-CF neutrophils by PMAactivated protein kinase C on FMLP- or NaFstimulated IP formation In the basal state, the addition of PMA alone (l.6x 10-SM) had no significant effect on either total IP accumulation or on the relative contribution of individual IP fraction to the total IP pool (non-CF and CF cells). In agreement with earlier studies [21, 22], we found that pre-incubation of non-CF neutrophils with PMA (l.6x 10-aM for 5 min), before addition of FMLP (10 -6 M, 30 s) effectively curtailed the expected increase in IP
Defective protein kinase C-mediated actions in cystic fibrosis neutrophils accumulation over basal conditions (Table 1 A, columns 1, 2, 5). In sharp contrast, prior incubations of C F neutrophils with P M A failed to block the F M L P stimulation of IP formation
263
so ~_~__ non CF
so .~
(Table 1 B, columns 1, 2, 5) When FMLP was
o
tl
replaced by N a F (18 m M for 20 min), prior incubation of neutrophils with PMA (1.6 x 10 -8 M for 5 min)effectively blocked the effect of N a F stimulation in the non-CF neutrophils but failed to do so in the CF neutrophils (Table 1 A and B, columns 3, 4, 6).
~ ~ so h ~ 20 *~ 10
The protein kinase C-mediated inhibition of exocytosis is marginal in the case of CF neutrophils
2
4 6 8 10 Duril~On Of PMA 10 "7M ~ b a l k ~
12
14 (min)
16
FIG. 3. Inhibition of FMLP-induced fl-glucuronidase exocytosis by preincubation with PMA in non-CF and CF neutrophils. In the absence of PMA, the values of fl-glucuronidase secretion were: 47% (non-CF) and 35% (CF). On the other hand, incubation with cytochalasin alone, PMA alone (incubation of 15 min) or PMA (incubation of 15 rain) and cytochalasin, gave values comprised between 10 and 15%. The results correspond to a single experiment representative of at least three separate pairs of samples.
Finally, the protein kinase C-mediated inhibition o f FMLP-induced fl-glucuronidase secretion was also examined. In a preliminary test, the inhibitory effect of P M A (10 -7 M) on F M L P (10 -6 M)-induced secretion was checked as a function of P M A pretreatment length (0-15 rain). Cytochalasin (5/~g/ml) was added five minutes prior to F M L P . The fl-glucuronidase release was measured 5 min after F M L P addition (Fig. 3). It was found that a 15 min preincubation with P M A gave the deeper inhibition (Fig. 3), as had also been observed by Barrowman et al. [24], and was therefore
selected for the comparison of CF and n o n - C F neutrophils (Table 2). F r o m the results summarized in Table 2, it can be concluded
TABLE 1. EFFECTOF PRIOR PMA n'~CUBATIONON FMLP AND NaF-INDUCED IP FORMATION Stimulation with
A. non-CF X S.E.M. B. CF. X S,E.M.
Ratios
FMLP
PMA + FMLP
NaF
PMA + NaF
2/1
4/3
1
2
3
4
5
6
1.39 0.09
1.00 0.14
1.49 0.08
1.08 0.14
0.72 0.07
0.73 0.05
1.19 0.28
1.20 0.09
1.46 0.09
1.46 0.12
1.01 0.04
1.01 0.04
Columns 1 to 4: the stimulation factor (ratio of IP accumulation under stimulation to that in the basal state) in different experimental conditions (1: FMLP 10 -6 M, 30 s; 2: PMA 1.6 x 10 -8 M 5 min + FMLP 10 -6 M, 30 s; 3: NaF 18mM, 20min; 4: PMA 1.6x 10-SM 5 m i n + N a F 18mM, 20min). Columns 5 and 6: the ratios of columns 2 (FMLP+PMA) to 1 (FMLP) and of columns 4 (NaF+PMA) to 3 (NaF). PMA alone had no effect. Paired t-tests show the effect of a prior PMA incubation to be significantly different in the CF and non-CF groups at the P < 0.01 level.
264
I. Groa~ et al.
TABLE 2. EFFECT OF PRIOR P M A INCUBATION ON FMLPqNDUCED ~LUCURONIDASE RELEASE
Stimulation with
Non-CF (n = 6). X S.E.M. CF (n --- 6). X S.E.M.
Ratios
FMLP
PMA + FMLP
2/1
1
2
3
44.2 2.9
33.8 3.9
0.76 0.04
35.4 1.2
32.4 1.0
0.92 0.02
fl-Glucuronidase release (% of total): column 1 in response to FMLP (10-6M); column 2 in response to FMLP (10-eM) after a 15 min preincubation in the presence of PMA (10-7 M); column 3, the ratios of colunm2 (FMLP+PMA) to 1 (FMLP). PMA by itself has a minor stimulatory effect on/~-glucuronidase exocytosis (see section D and Fig. 3). that this PMA mediated-inhibition is significant in non-CF neutrophils (P < 0.05) whereas it is only marginal in cells from CF patients (P < 0.1). The difference between the ratios observed for non-CF and CF subjects is significant at the level P < 0.01.
DISCUSSION Paired analysis of CF and non-CF neutrophils in the presence of the physiologic agonist FMLP failed to show any significant difference in the acumulation of [3H]IP (IPI, IP2, IP3, IP4 and sum of IP) or in their respective dose-response and time curves. Sample size was sufficient to detect a possible difference between means equal to 1.5 times the SD common to the two groups with a power > 0.95 (Type II error, type I error being kept at the 0.05 conventional level). No such difference was found and it is concluded that the difference in IP generation under FMLP stimulation, if any, between CF and non-CF neutrophils is smaller than 1.5 SD and cannot account for the observed differences (FMLP-induced exocytosis and chemiluminescence) between these cells. The defective step lies presumably downstream from the IP generation.
We showed that, upon PMA activation, CF neutrophils release lesser amounts of ~-glucuronidase secretion than non-CF cells. Moreover, the inhibition exerted by PMA pretreatment on the FMLP induced release is only marginal in CF neutrophils. In addition, our observations demonstrate that in CF neutrophils, the retroinbibition exerted by PMA-activated protein kinase C on FMLP- and NaF-stimulated generation of IP fails to materialize. This is in contrast with the response observed in non-CF neutrophils which provides positive evidence of the inhibitory effect commonly exerted by PMA on IP accumulation and granular enzymes secretion [21-24], the mechanism of which remains, however, unknown. In conclusion, the CF primary defect in neutrophils apparently lies downstream from the IP generation, more probably in the immediate vicinity of protein kinase C. Hermelin et al. [30] also suggested that, in CF fibroblasts, the primary defect lies at the level of a regulatory protein or at the level of a phosphorylation cascade (defective substrate or enzyme). Our results could be explained in two ways: either they are secondary effects of the disease (in other words, a non-representative subpopulation of neutrophils not expressing the CFTR gene might be particularly abundant in CF
Defective protein kinase C-mediated actions in cystic fibrosis neutrophils patients), or they reflect the defective activity of the C F T R gene in CF neutrophils. It seems highly unlikely that the defective C F T R gene would cause the accumulation o f a particular subpopulation of neutrophils not expressing the C F T R gene, but deficient in exocytosis and protein kinase C activities. We thus suggest that the C F T R protein is active in neutrophils (or in their precursors) and that in these cells, exists the target, either directly or indirectly, of protein kinase C and participates in the retroinhibition effects exerted by this kinase. That the C F T R protein is the target of protein kinase C (and of protein kinase A) has already been demonstrated in epithelial cells [31, 32]. In addition, the C F T R protein has many sites suitable for phosphorylation by protein kinase C. Moreover, the most frequent D N A mutation observed in the C F T R gene manifests itself at the amino acid 508 o f the polypeptide chain, at the level of an ATPbinding site and in the immediate vicinity of a protein kinase C-phosphorylation site. Our results suggest that protein kinase C might also exert CFTR-mediated retroinhibition effects and it would be interesting to test this possibility in epithelial cells. Expression of the C F T R gene in neutrophils is perhaps not as surprising as it might seem. Indeed, like typical exocrine epithelial cells, neutrophils are secretory cells. It has been suggested that the C F T R protein is a transporter (rather than an ion channel) which binds and transports a molecule that regulates chloride channels [7, 10, 11, 33]. C F T R protein may play a similar role in neutrophils. Acknowledgements--We are grateful to Dr R. VAr~ GEFFEL and Dr D. BARAN for kindly providing us with samples of CF blood. This work was supported by the National Fund for Scientific Research (FNRS, Belgium). I.G. was an IRSIA grantee. C.S. is Senior Research Associate of the FNRS.
REFERENCES 1. Knowlton R., Cohen-Haguenauer O., Van Cong N., Frezal J., Brown V., Barker D., Braman J., Schumm J., Tsui L.-C., Buchwald M. and Donis-Keller H. (1985) Nature 318, 380-382.
265
2. Wainwright B., Scambler P., Sehmidtke J., Watson E., Law H.-Y., Farrall M., Cooke H., Eiberg H. and Williamson R. (1985) Nature 318, 384-385. 3. White R., Woodward S., Leppert M., O'Connell P., Hoff M., Herbst J., Lalouel J.-M., Dean M. and Vande Woude G. (1985) Nature 318, 382-384. 4. Riordan J., Rommens J., Kerem B.-S., Alon N., Rozmahel R., Grzelczak Z., Zielenski J., Lok S., Plavsic N., Chou J.-L., Drumm M., lannuzzi M., Collins F. and Tsui L.-C. (1989) Science 245, 1066-1073. 5. Rich D., Anderson M., Gregory R., Cheng S., Paul S., Jefferson D., MacCann J., Klinger K., Smith A. and Welsh M. (1990) Nature 347, 358-363. 6. Drumm M., Pope H., Cliff W., Rommens J., Marvin S., Tsui L.-C., Collins F., Frizzell R. and Wilson J. (1990) Cell 62, 1227-1233. 7. Hyde S., Emsley P., Hartshorn M., Mimmack M., Gileadi U., Pearce S., Gallagher M., Gill D., Hubbard R. and Higgins C. (1990) Nature 346, 362-365. 8. Quinton P. (1990) FASEB J. 4, 2709-2717. 9. Cheng P., Boat T., Cranfiil K., Yankaskas J. and Boucher R. (1989) J. clin. Invest. 84, 68-72. 10. MacPherson M. and Dormer R. (1988) TIBS 13, 10-13. 11. Welsh M. (1988) lOth Internam. CF Congress, Sydney, Australia, March 1988, Excerpta Medica, Asia Pacific Congress Series, 74. 12. Bradford P. and Rubin R. (1985). Molec. Pharmac. 27, 74-78. 13. Bradford P. and Rubin R. (1985) FEBS Lett. 183, 317-320. 14. Bradford P. and Rubin R. (1986) J. biol. Chem. 261, 15,644-15,647. 15. Kemp T., Schram-Doumont A., Van Geffel R., Kram R. and Szpirer C. (1986) Pediatr. Res. 20, 520-526. 16. White J., Huang J., Naccache P., Becker E. and Sha'afi R. (1984) J. bioL Chem. 259, 8605-861 I. 17. Hayakawa T., Suzuki K., Suzuki S., Andrew P. and Babior B. (1986) J. biol. Chem. 261, 9109-9115. 18. Wright C. and Hoffman M. (1987) Biochem. biophys. Res. Commun. 142, 53-62. 19. Castagna M. (1987) Biol. Cell 59, 3-14. 20. Magnusson M., Halldorson H., Kjeld M. and Thorgeirsson G. (1989) Biochem. J. 264, 703-711. 21. Kato H., Ishitoya J. and Takenawa T. (1986) Biochem. biophys. Res. Commun. 139, 1272-1278. 22. Tyagi S., Tamura M., Burnham D. and Lambeth J. (1988) J. biol. Chem. 263, 13,191-13,198.
266
I. GRAFFet al.
23. Naccache P., Molski T., Borgeat P., White J. and Sha'afi R. (1985) J. biol. Chem. 260, 2125-2131. 24. Barrowman M., Cockcroft S. and Gomperts B. (1986) FEBS Lett. 201, 137-142. 25. B6yum A. (1986) Scand. J. clin. Invest. 22 (suppl. 97): 77-89. 26. Downes C. and Michell P. (1981) Biochem. J. 198, 133-140. 27. Dewald B., Bretz U. and Baggiolini M. (1982) J. clin. Invest. 70, 518-525. 28. Hallcher L. and Sherman W. (1980) J. biol. Chem. 255, 10,896-10,901.
29. Smith R., Justen J. and Sam L. (1988) Biochem. biophys. Res. Commun. 152, 1497-1503. 30. Hermelin B., Cherqui G., Bertrand F., Wicek D., Paul A., Garcia I. and Picard J. (1988) FEBS Left. 229, 161-166. 31. Schoumaker R., Shoemaker R., Haim D., Tallant E., Wallace R. and Frizzell R. (1987) Nature 330, 752-754. 32. Hwang T., Lu L., Zeitlin P., Gruenert D., Huganir R. and Guggino W. (1989) Science 244, 1351-1353. 33. Ringe D. and Petsko G. (1990) Nature 346, 312-313.
