- Email: [email protected]
Lindane decreases forskolin-stimulated cyclic AMP accumulation but does not modify Gs in rat enterocytes
Cellular Signalling Vol. 5, No. 4, pp. 453-462. 1993. Printed in Great Britain.
0898--6568/93 $6.00 + 0.00 © 1993 Pergamon Press LM
L I N D A N E DECREASES FORSKOLIN-STIMULATED CYCLIC AMP A C C U M U L A T I O N BUT DOES NOT M O D I F Y Gs IN RAT ENTEROCYTES I. CARRERO, N. RODRIGUEZ-HLr~CH~, L. G. GUUARRO, M. N. 1 ~ o o , M. A. I~h~z-At,nARS^NZ and J. C. PRIE'rO* Unidad de Neuroendocrinologia Molecular, Departamento de Bioquimica y Biologia Molecular, Universidad de Alcalfi, E-28871 Alcalfi de Henares, Spain (Received 3 November 1992; and accepted 5 February 1993) Abstract--Treatment of isolated rat enterecytes with the halogenated insecticide lindane (the y-isomer of hexachlorocyclohexane, HCCH) did not modify the general membrane fluidity (as estimated by a fluorescence polarization technique) nor the guanine nucleotide binding regulatory protein Gs (as studied by both ADP-ribosylation of its • subunit by cholera toxin and Gpp[NH]p stimulation of membrane adenylate cyclase activity). However, iindane decreased in a dose-dependent manner the effect of the diterpene forskolin on direct activation of the adenylate cyclase catalytic subunit. After 5 min of cell treatment with 0.5 mM lindane, the maximal stimulatory effect of forskolin (at 100 I~M) decreased by about 50%. There was a certain degree of specificity since 6-HCCH was indeed more potent, whereas dieldrin and endrin (non-lindane related halogenated compounds) behaved as findane, and ~- and /]-HCCH were poorly efficient on the inhibition of forskolin stimulation of adenylate cyclase activity. A similar effect of lindane was observed on receptorstimulated cyclic AMP accumulation by using vasoactive intestinal peptide instead of forskolin. The results on a non-receptor mediated effect of lindane on the adenylate cyclase catalytic subunit itself could be related to: (i) alterations of membrane microdomains surrounding this and other integral proteins which would result in modifications of their activities; and/or (ii) a reciprocal relation between the two main routes of signal transduction so that the activation of protein kinase C (or other Ca2+-dependent protein kinases) by lindane would lead to phosphorylation of the adenylate cyclase catalytic subunit. The simultaneous preincubation of enterocytes with lindane and a tumour-promoting phorbol ester resulted in additive inhibitory effects on forskolin stimulation of cyclic AMP accumulation which is apparently against the suggested involvement of protein kinase C in the mechanism of action of the insecticide. However, it cannot be discarded since this enzyme activity represents a family of isoforms so lindane and the phorbol ester could act on different isoenzymes. Key words: Lindane, Gs protein, forskolin, adenylate cyclase, protein kinase C, rat enterocytes. INTRODUCTION TIlE organochlorine insecticide lindane (the yisomer of hexaclorocyclohexane, ~,-HCCH) is widely used in veterinary and human medicine to control ectoparasites [1]. Lindane gains entry * Author to whom correspondenceshould be addressed. Abbreviations: BSA--bovine s e r u m albumin; DMSO--dimethyl-sulphoxide; DPH--diphenylhexatriene; Gpp[NH]p--guanyl-5'-ylimidodiphosphate;HCCH--hexachlorocyclohexane; IBMX--3-isobutyl-1-methylxanthine; PDD--~-phorbol 12,13-didecanoate; PKC--protein kinaseC; SDS--PAGE--sodiumdodecyisulphat~polyacrylamide gel electrophoresis; TPA--12-O-tetradecanoylphorbol-I3-acetate; VIP--vasoactive intestinal peptide. 453
into the body system as a food toxicant, by inhalation into lungs or by diffusion through the skin [2, 3]. As a highly lipophilic compound, lindane accumulates in the adipose tissue and in the membrane lipid bilayer of all cells [4]. The insecticide is a potent convulsant agent acting on the central nervous system [5] and produces a constellation o f other effects including potential carcinogenicity [6], liver oxidative stress [7], nephrotoxicity [8] and degenerative changes in the reproductive system [9, 10]. The mechanism of action o f lindane remains poorly understood. As it occurs with many other insecticides, biomembranes appear to be
454
I. CARItEROet al.
primary targets of lindane action [4]. Although controversy exists, studies with various model and native membranes indicate that lindane interacts with membrane lipids and accommodates into the membranes without causing perturbation of the general membrane fluidity [11]. However, lindane is able to perturbate discrete membrane domains such as those surrounding integral proteins including receptors and ionic channels [12, 13]. Furthermore, lindane and inositol are structurally related and the insecticide interferes with inositol-phospholipid turnover [14] and increases free intracellular Ca 2+ levels which can result in activation of protein kinase C (PKC) and other protein kinases [15]. It has been demonstrated in different types of cells that PKC activation by tumour-promoting phorbol esters and other agents can regulate signal transduction through the adenylate cyclase pathway [16, 17]. In this context, previous results have shown that lindane impairs the stimulatory activity of fladrenergic agonists on cyclic AMP accumulation in rat prostatic epithelial cells [18] and renal cortical tubules [19]. Lindane also inhibits cyclic AMP accumulation induced by the neuropeptide vasoactive intestinal peptide (VIP) in rat prostatic [20] and intestinal [21] epithelial cells but it does not modify the binding of VIP to the corresponding membrane receptors [21]. These features prompted us to study the possibility of a direct interaction of lindane with the adenylate cyclase pathway of signal transduction. For this purpose, we tested the effect of lindane treatment of isolated rat enterocytes on: (i) guanine nucleotid¢ regulatory protein Gs as assessed by experiments on cholera toxin-catalysed ADP-ribosylation of membranes, and adenylate cyclase stimulation by guanyl-5'-yl imidodiphosphate (Gpp[NH]p); and (ii) the catalytic subunit of adenylate cyclase by means of stimulation with the diterpene forskolin. We extended our analysis in order to study possible effects of lindane on membrane fluidity as estimated by fluorescence polarization measurements. An approach to the possibility of a PKC-mediated effect of lindane
on adenylate cyclase was performed with experiments with phorbol esters.
MATERIALS A N D METHODS Chemicals
Lindane and other organochlorine compounds were purchased from Pierce Scientific (Rockford, IL, U.S.A.). Highly purified VIP was obtained from Prof. V. Mutt (Karolinska Institute, Stockholm, Sweden). Forskolin, 12-O-tetradecanoylphorbol-13acetate (TPA), 4~t-phorbol 12,13-didecanoate (PDD), diphenylhexatriene (DPH), dimethylsulphoxide (DMSO), 3-isobutyl- l-methylxanthine (IBMX), bovine serum albumin (BSA), Gpp[NH]p, ATP, creatine phosphate, creatine phosphokinase, cholera toxin, and protein markers and chemicals for gel electrophoresis were from Sigma Quimica (Alcobendas-Madrid, Spain). Animals and cell membrane preparation
Adult male Wistar rats weighing 180-250 g (aged about 4 months) were used. The animals were decapitated and the jejunum excised quickly and used for enterocyte isolation as described previously [22]. Cell viability was always greater than 90% as estimated by exclusion of trypan blue dye. Basolateral plasma membranes were isolated from enterocytes as described elsewhere [23]. Protein concentration was determined [24] using BSA as a standard. Lindane treatment
In standard conditions, rat enterocytes (about 50 mg wet wt per ml Krebs-Ringer phosphate buffer, pH 7.4) or enterocyte basolateral membranes (1 mg protein per ml buffer) were preincubated for 5 rain at 25°C with lindane (0.5 mM in 0.25% DMSO) or 0.25% DMSO alone and washed. In some experiments, the preincubation mixture included also a phorbol ester (1 ~tM TPA or PDD). Cyclic A M P and adenylate cyclase activity
The stimulation of cyclic AMP accumulation by isolated enterocytes was studied as described previously [25]. Briefly, control or lindane-treated enterocytes (1 mg cell protein/ml) were incubated in standard conditions in the absence or presence of 100nM-100~tM forskolin for 45min at 15°C in 0.5 ml of a medium consisting of 35 mM Tris-HCI buffer (pH 7.5), 50 mM NaCI, 1.4% BSA, 1 mg/ml bacitracin and 0.2 mM IBMX. The reaction was stopped by the addition of 2.5 ml methanol. After
A B C D Mr
x 10-3
-84.0
-36.5 -26.6 -
÷ ÷-
CTX
FIG. 1. Autoradiograph of [32p]ADP-ribosylated membrane proteins of control and lindane-treated rat enterocytes. The cells were incubated with [32p]NAD in the absence and presence of cholera toxin (CTX) as described under Materials and Methods. Lanes A and B correspond to control conditions whereas lanes C and D are those of lindane-treated enteroeytes.
455
L i n d a n e effects in r a t enterocytes
centrifugation, aliquots of the supernatant were evaporated for cyclic AMP determination [26]. Adenylate cyclase activity was measured as reported previously [27] with minor modifications. Enterocyte basolateral plasma membranes (0.5 mg protein/mi) were incubated with 1.5mM ATP and 5mM MgSO4, an ATP-regenerating system (7.4mg/ml creatine phosphate and 1 mg/ml creatine kinase), l m M IBMX, l m M EDTA and lnM-0.1mM Gpp[NH]p in 0.1 ml of 25 mM triethanolamine-HC! buffer (pH 7.4) for 30 min at 30°C. The reaction was stopped by heating for 3 min. After refrigeration, 0.3 ml of an alumina slurry (0.25 g/ml in triethanolamine--HC! buffer, pH 7.4) was added. After centrifugation, cyclic AMP was determined as described above.
457
1¢
c
-6 v
! 4
n z
o -_~-/i |
i0~8
101-6
,. 10-4
[Gpp(NH)p] (M)
Cholera toxin-catalysed ADP-ribosylation ADP-ribosylation of enterocyte membranes was performed as described elsewhere [28]. Membranes (4 mg protein/ml) were incubated with activated cholera toxin (50 gg/ml) in 0.1 M phosphate buffer (pH 7.5) containing 5 mM ATP, 50 ~tM GTP, 5 ~tM [32P]NAD+, 2raM EDTA, 5raM MgCI,, and an ATP-regenerating system. After 30 rain at 30°C, 10% (w/v) trichioroacetic acid was added and the mixture was centrifuged and submitted to sodium dodecyisulphate-polyacrylamide gel electrophoresis (SDSPAGE). The resulting gels were then subjected to autoradiography. Fluorescence polarization measurement Control or lindane-treated rat enterocytes were washed, suspended in phosphate-buffered saline (about 0.2 absorbance units at 365 nm), incubated for 30 rain at 37°C with 0.6 pM DPH (dissolved in tetrahydrofuran) and then subjected to polarization analysis [29] using a Perkin-Elmer model LS-3B spectrofluorometer. DPH was excited at 357 nm, and emission was recorded at 430nm. Steady-state fluorescence polarization was computed according to the relationship P = (Ivv--lvh)/(lvv+lvh), where I w and lvh are the fluorescence intensifies observed with the analysing polarizer parallel and perpendicular to the polarizer excitation beam, respectively. The data were corrected for unequal transmission of differently polarized light and for intrinsic fluorescence. RESULTS The possibility that a lindane perturbation o f the general fluid phase of plasma membrane could result in uncoupling of the components of
FIG. 2. Dose-effect curves for Gpp[NH]p-stimulated adenylate cyclase activity in basolateral membranes of rat enterocytes. Membranes were preincubated for 5 rain at 25°C in the absence (Q) or presence (©) of 0.5 mM lindane. The results are given as the means -t-S.E,M. of nine determinations performed in triplicate. There were no statistically significant differences between the two groups.
the adenylate cyclase system o f signal transduction (receptors, G proteins and adenylate cyclase catalytic subunit) appears unlikely. The insecticide did not modify membrane fluidity in isolated rat cnterocytes since the fluorescence polarization value (and hence the membrane microviscosity) was nearly identical in control and lindane-treated cells: 0.179 +0.020 and 0.182+0.021, respectively (means +S.E.M., n = 5). Lindane showed also a lack of effect at the Gs protein level. The autoradiogram presented in Fig. 1 clearly indicates that lindane treatment of isolated enterocytes did not result in any modification of the extent of ADP-ribosylation of the • subunit of Gs protein in the presence of [32P]NAD, since the radioactivity incorporation into the 42,000 Mr membrane protein upon the addition of cholera toxin was very similar to that in control conditions. The same occurred in the study o f the response o f adenylate cyclase activity to Gpp[NH]p (acting via Gs protein) in enterocyte basolateral membranes; the ability of increasing concentrations of this nucleotide
458
I. CAaar~go et al.
lOG .~
75
I= ¢J
~
50
0.
25
g 8 2g
ols
O
-2
i
4
[Lindane] (mH)
Fro. 3. Dose dependence of lindane inhibition of forskolin-stimulated cyclic AMP accumulation in rat enterocytes. The cells were preincubated for 5 rain at 25°C in the absence or presence of increasing lindane concentrations. After washing, the cells were incubated without ([-]) and with 1 (V), 10 (A) or 100 (O) pM forskolin. Values are means of triplicates. to stimulate the enzyme activity was not significantly altered by lindane treatment of membranes (Fig. 2). We tested the effect o f enterocyte treatment with lindane on the subsequent stimulation of cyclic A M P accumulation in the isolated cells by forskolin, a diterpene which is assumed to act directly on the adenylate cyclase catalytic subunit [30]. Lindane inhibited the forskolin-
0
i
i
m [[10-7 I0~-~ I0~-m I0L4 [Forskolin] (H)
FIG. 5. Dose-response curves for forskolin stimulation of cyclic AMP accumulation in rat enterocytes. After cell preincubation for 5 min at 25°C in the absence (Q) or presence (O) of 0.5 mM lindan¢, the enterocytes were incubated with increasing concentrations of forskolin. Values are the means +S.E.M. of eight determinations performed in triplicate. stimulated cyclic A M P levels at submaximal (1-10 ~M) and maximal (100 oM) active concentrations of the diterpcne (Fig. 3). This
12
i0o .! 75 "~
o.
8
o
50
o.
<
4
25
o
o i
i
Time (min)
Fx6. 4. Effect of increasing time-periods on lindane inhibition of forskolin-stimulated cyclic AMP accumulation in rat enterocytes. The cells were preincubated at 25°C in the absence (closed symbols) and presence (open symbols) of 0.5 mM lindane for the indicated times. After washing, the cells were incubated without (squares) or with (circles) 100 I~M forskolin. Values are means of triplicates.
0
I
i
m III0-7 I0'-z I0"~ 10"4 [Forskofin](H)
Fro. 6. Specificity of lindane action. Rat enterocytes were preincubated for 5 min at 25°C in the absence (O) and presence of 0.5 mM lindane (O), ~-HCCH (<>), dieldrin (&), endrin (A), ~t-HCCH (ll) and flHCCH (D). After washing, the cells were incubated with increasing concentrations of forskolin. Values are means of triplicates.
Lindane effects in rat enterocytes
459
o
2C
I00 .:..~.:x:x::x:
ilili,i i
rxxxxxxx:
,, x
50
"-i:'.~//////2/.:~ '.. e////////..;:~//~
C
!i!!!:ii!iiiill L
TPA
PDD
L÷TPA
Fro. 7. Dose--response curves for VIP stimulation of cyclic AMP accumulation in rat enter~ytes. After cell preincubation for 5 rain at 25°e in the absence (Q) or presence (©) of 0.5 mM lindane, the enterocytes were incubated with increasing VIP concentrations. Values are the means+S.E.M, of eight determinations performed in triplicate.
F~G. 8. Effect of enterocyte preincubation with lindane and/or phorbol esters on the subsequent stimulation of cyclic AMP accumulation by forskolin. The cells were preincubated for 5 min at 25°C in the absence (C) and presence of 0.5 mM lindane, 1 IxM TPA or 1 ltM PDD. After washing, the cells were incubated with 0.1 mM forskolin. The bars represent the means _S.E.M. of six determinations performed in triplicate and they are expressed as a percentage of the control value by considering stimulated minus basal levels. *P<0.05 vs control, Student's t-test.
inhibitory effect depended on lindane concentration and became maximal at 0.3 m M insecticide. Relatively low doses (0.1-0.2mM) o f lindane elicited a small stimulatory response on the basal cyclic A M P levels, a feature that was maintained even after stimulation with forskolin, as indicated by the shoulders observed in Fig. 3. This unpredicted effect may constitute a further action of lindane but it deserves additional studies for a better understanding. In another context, experiments at increasing time-periods showed that forskolin-stimulated cyclic A M P accumulation resulted in maximal inhibition after 5 min exposure to the organochlorine compound (Fig. 4). The effect of lindane on the stimulatory action of increasing forskolin concentrations on cyclic A M P accumulation in rat enterocytes is shown in Fig. 5. Half-maximal cyclic A M P response was seen at about 10 liM forskolin in both control and lindane-treated cells. However, the maximal stimulatory effect (at 10011M forskolin) decreased by 50% after lindane treatment. The specificity o f lindane activity was analysed by using other H C C H
isomers and unrelated organochlorine compounds including dieldrin and endrin (Fig. 6). The 6-HCCH isomer showed the highest inhibitory effect on forskolin-stimulated cyclic A M P accumulation, whereas endrin and dieldrin behaved as lindane, and the ~- and /7-HCCH isomers elicited a poor effect at this level. The inhibitory effect o f lindane on the direct stimulatory action of forskolin on the adenylate cyclase catalytic subunit was also seen in receptor-mediated responses. Concentrationresponse experiments of VIP-stimulated cyclic A M P accumulation in rat enterocytes showed a decrease of VIP efficiency (by 50% at a maximally active, 100 nM dose o f VIP) after cell preincubation with 0.5 mM lindane (Fig. 7). The potency o f the stimulatory agent was again maintained by the insecticide since half-maximal cyclic A M P response was elicited at about 9 nM VIP in both control and lindane-treated enterocytes. The possibility that lindane could stimulate PKC activity leading to phosphorylation of the adenylate cyclase catalytic subunit (and/or other components o f this pathway of signal
[vm] (~)
460
I. CARREROet al.
transduction) was analysed by performing experiments with phorbol esters. Preincubation of rat enterocytes with the tumour-promoting phorbol ester TPA mimicked the inhibitory effect of lindane on forskolin-stimulated adenylate cyclase activity whereas PDD (a phorbol ester lacking in tumour-promoting activity) was inactive at this level. The simultaneous preincubation of the cells with lindane and TPA resulted in a merely additive inhibitory response. DISCUSSION In the present study, we have evaluated the effect of the insecticide hndane on Gs protein and adenylate cyclase catalytic subunit which are components of a major signal transduction pathway. Treatment of isolated enterocytes with lindane did not perturbate the general membrane fluidity (as estimated by DPH fluorescence polarization) nor Gs protein (as studied by both ADP-ribosylation with cholera toxin and Gpp[NH]p stimulation of membrane adenylate cyclase activity). However, the insecticide severely decreased the effect of the diterpene forskolin on direct activation of the adenylate cyclase catalytic subunit. The experiments were performed in standard conditions with 0.5 mM lindane, a concentration that is similar to those used in processes such as the inhibition of inositol phospholipid biosynthesis in pancreatic acini [31], the increase of calcium levels within synaptosomes [15] and the interaction with ~,-aminobutyric acid-activated chloride channels [32]. However, lower concentrations of lindane were needed to inhibit hepatocyte gap junctional intercellular communication [33] or t-butylbicyclophosphorotionate binding to chloride channels in brain membranes [34]. In the present experimental conditions, lindane did not modify the general membrane fluidity in isolated rat enterocytes, in close agreement with previous results obtained in fluid native membranes and model bilayers [11]. However, evidence exists that the insecticide may differently affect especially ordered micro-
domains surrounding integral membrane proteins and thus result in alterations of the corresponding activities [12,13]. Accordingly, it has been inferred that the perturbation of membrane fluidity by lindane may result in uncoupling of receptor-effector systems and be responsible at least in part for previous observations on the inhibitory action of the insecticide on VIP-stimulated cyclic AMP accumulation in rat enterocytes which was not accompanied by a modification in the binding of VIP to its specific receptors [21]. Gs protein transmits a signal to the catalytic adenylate cyclase subunit after the activation of a stimulatory membrane receptor by a specific agonist. Thus, Gs protein plays a pivotal role in the receptor-linked regulation of adenylate cyclase [35]. Experiments on cholera toxin ADP-ribosylation of the • subunit of Gs were performed for the identification of G protein in rat enterocytes, giving evidence that the extent of specific labelling was unaffected by lindane. This suggestion on the lack of effect of lindane on Gs protein was reinforced by the similar pattern of stimulation of adenylate cyclase activity in control or lindane-treated enterocyte basolateral membranes by increasing Gpp[NH]p concentrations. To our knowledge, there are no previous reports on the interaction of lindane with G proteins. Forskolin is a good research tool in the investigation of the adenylate cyclase system because it is a strong activator of this enzyme in membrane and intact cells. At the micromolar concentrations used in the present study, forskolin is likely to act directly on the catalytic subunit of adenylate cyclase and not via the Gs-cyclase complex [30]. The present results show the ability of lindane to inhibit the stimulatory action of forskolin on cyclic AMP accumulation in rat enterocytes due to a decrease in the efficiency of the diterpene (by 50% at a maximally effective forskolin concentration) that could result from a direct interaction of the insecticide with the adenylate cyclase catalytic subunit. It has been shown that forskolin can increase membrane fluidity and may achieve adenylate cyclase activation by
Lindane effectsin rat enterocytes binding to sites either at the polar surface of the enzyme or buried in the membrane bilayer [36]. Hence, it should be thought that the hydrophobic nature of lindane resulted in the displacement of forskolin from those sites with the corresponding decrease in the extent of cyclic AMP stimulation by the diterpene. This possibility appears unlikely since: (i) lindane-treated cells were extensively washed before using forskolin for cyclic AMP stimulation (although some quantity of the insecticide could remain retained by the cells); and (ii)we observed a similar inhibitory effect of lindane on receptorstimulated cyclic AMP accumulation, i.e. the signal transduction system for the neuropeptide VIP. These results are in agreement with previous studies on the regulation of receptormediated cyclic AMP levels by lindane, including stimulation of cyclic AMP accumulation by VIP in rat enterocytes [21], VIP [20] or the #-adrenergic agonist isoproterenol [19] in rat prostatic epithelial cells, and isoproterenol in rat renal cortical tubules [19]. However, our results on the potent inhibition of forskolinstimulated adenylate cyclase activity by lindane are to our knowledge the first observation on a possible direct non-receptor mediated interaction of this halogenated insecticide on the adenylate cyclase catalytic subunit. The marked inhibitory activity of lindane upon forskolin-stimulated cyclic AMP accumulation in rat enterocytes showed a relative degree of specificity as shown by experiments with various organochlorine substances which were powerful (6-HCCH), median (lindane and the non-related compounds dieldrin and endrin) or poor (the non-insecticide ~- and #HCCH isomers) agents at this level. There are some discrepancies in the pattern of specificity described in the present report and in previous studies on receptor-mediated cyclic AMP accumulation [18, 21], Ca 2+ mobilization from intracellular stores [37], potential tumorigenic effects [6] and binding to different epithelia [38]. It appears likely that the interaction of lindane and other organochlorine compounds with different tissues may depend on highly substrate-specific factors.
461
The observed inhibitory action of lindane on forskolin-stimulated adenylate cyclase activity in rat enterocytes could be related, at least in part, to a connection between the adenylate cyclase and the PKC-CaZ+-inositol triphosphate pathways of signal transduction. Lindane increases free cytosolic Ca 2+ levels [15] which could activate PKC and other protein kinases thereby resulting in phosphorylation of the adenylate cyclase catalytic subunit and/or other cell proteins. Considering the adenylate cyclase system, it would be a selective mechanism depending probably on the nature and/or position of each component in the cell membrane since neither Gs protein (present work) nor membrane receptors such as those for VIP [19] are modified after cell exposure to this halogenated insecticide. Moreover, the inositol-like structure of lindane may be related to the observed effect of this insecticide on the turnover of inositol phospholipids [14] and hence contribute to a connection between the two main routes of signal transduction. However, the present results with simultaneous preincubation of enterocytes with lindane and the tumour-promoting phorbol ester TPA (which activates PKC in a direct manner) indicate an additivity of their inhibitory effects on the subsequent stimulation of cyclic AMP accumulation by forskolin. This feature is at odds with the possibility of a direct interaction between lindane and PKC. However, it should be remembered that PKC represents a family of isozymes [39] with different activation patterns. Thus, it would be possible in the present system for TPA to interact with a PKC isoform, whereas lindane could activate a different one (or even other classes of protein kinases through a Ca~+-mediated mechanism). Clearly, more in-depth studies are required to exclude or accept definitively the role of PKC in this process. Regardless of the mechanism, the present results establish a modulatory effect of lindane and other organochlorine compounds on the adenylate cyclase catalytic subunit itself in the rat enterocyte, which may be the result of alterations of membrane microdomains surrounding this and other integral proteins,
462
I. CARaEROet al.
and/or reciprocal effects between the adenylate cyclase and the PKC-Ca2+-inositol triphosphate pathways of signal transduction. Acknowledgements--This work was supported by grants from the Fondo de Investigaciones Sanitarias de la Seguridad Social (88/904 and 92/1272) and the Direcci6n General de Investigaci6n Cientifica y T6cnica (PM89/96).
REFERENCES I. Wooley D., Zimmer L., Dodge D. and Swanson K. (1985) Neurotoxicology 6, 165-192. 2. Zhu J., Feng Z. and Chen J. (1986) Pestic. Biochem. Physiol. 25, 414-419. 3. Lopez-Aparicio P., Hoyo N. and PerezAlbarsanz M. A. (1988) Pestic. Biochem. Physiol. 31, 109-119. 4. Antunes-Madeira M. C. and Madeira V. M. C. (1985) Biochim. Biophys. Acta 820, 165-172. 5. Tussel J. M., Sufiol C., Gelpi E. and Rodriguez-Farre E. (1987) Arch. Toxicol. 60, 432-437. 6. Wolf G. L., Roberts D. W., Morrisey R. L., Greenman D. L., Allen R. R., Campbell W. L., Bergman H., Nesnow S. and Frith C. H. (1987) Carcinogenesis 8, 1889-1897. 7. Videla L. A., Simizu K., Barros S. B. M. and Junqueira V. B. C. (1989) Cell Biochem. Funct. 7, 179-183. 8. Andrews J. E. and Gray L. E. (1990) Toxicology 60, 99-107. 9. Srinivasan K., Ramesh H. P. and Radhakrishnamurty R. (1988) Bull. Environ. Contain. Toxicol. 41, 531-539. 10. Uphouse L. and Williams J. (1989) Toxicol. Lett. 48, 21-28. I I. Antunes-Madeira M. C. and Madeira V. M. C. 0989) Biochim. Biophys. Acta 982, 161-166. 12. Ogiso T., Iwaki M. and Moil K. (1981) Biochim. Biophys. Acta 649, 325-335. 13. Jones O. T., Froud R. J. and Lee A. G. (1985) Biochim. Biophys. Acta 12, 740-751. 14. Parries G. S. and Hokin-Neaverson M. (1985) J. biol. Chem. 260, 2687-2693. 15. Bondy S. C. and Halsail L. (1988) Neurotoxicology 9, 645-652. 16. Le Goas F., Amiel C. and Friedlander G. (1991) Am. J. Physiol. 261, F587-F592. 17. Ransjo M. and Lerner U. H. (1991) Calc. Tiss. Int. 49, 284-287.
18. Carrero I., Perez-Albarsanz M. A., Carmena M. J. and Prieto J. C. (1990) Pestic. Biochem. Physiol. 38, 197-203. 19. Lopez-Aparicio P., Recio M. N., Prieto J. C., Carmena M. J. and Perez-Albarsanz M. A. (1991) Life Sci. 49, 1141-1154. 20. Carrero I., Perez-Albarsanz M. A., Recio M. N. and Prieto J. C. (1992) Toxic. In Vitro 6, 7-9. 21. Carrero I., Fernandez-Moreno M. D., PerezAibarsanz M. A. and Prieto J. C. (1989) Biochem. biophys. Res. Commun. 159, 1391-1396.
22. Prieto J. C., Laburthe M. and Rosselin G. (1979) Eur. J. Biochem. 96, 229-237. 23. Young G. P., Morton C. L., Rose I. S., Taranto T. M. and Bhathal P. S. (1987) Gut 28, 57-62. 24. Lowry O. H., Rosebrough N. J., Farr A. L. and Randall R. J. (1951) J. biol. Chem. 193, 265-275. 25. Laburtbe M., Prieto J. C., Amiranoff B., Dupont C., Hui Bon Hoa D. and Rosselin G. (1979) Fur. J. Biochem. 96, 239-248. 26. Gilman A. G. (1970) Proc. natn. Acad. Sci. U.S.A. 74, 3424-3428. 27. Houslay M. D. and Gordon L. M. (1983) Curt. Top. Membr. Transp. 18, 179-231. 28. Couvineau A., Amiranoff B. and Laburthe M. (1986) J. biol. Chem. 261, 14,482-14,489. 29. Shinitzky M. and Barenholz Y. (1978) Biochim. Biophys. Acta 515, 367-394. 30. Seamon K. B. and Daly J. W. (1981) J. biol. Chem. 250, 9799-9801. 31. Crouch M. F. and Roberts M. L. (1985) Biochim. Biophys. Acta 844, 149-157. 32. Ogata N., Vogel S. M. and Narahashi T. (1988) FASEB J. 2, 2895-2900. 33. Klaunig J. E., Ruch R. J. and Weghorst C. M. (1990) Toxicol. appl. Pharmac. 102, 553-563. 34. Abalis I. M., Edelfrawi M. E. and Edelfrawi A. T. (1985) Pestic. Biochem. Physiol. 24, 95-102. 35. Spiegel A. M., Backlund P. S., Butrynski J. E., Jones T. L. Z. and Simonds W. F. (1991) Trends Biochem. Sci. 16, 338-340. 36. Whetton A. D., Needham L., Dodd N. J. F., Heyworth C. M. and Houslay M. D. (1983) Biochem. Pharmac. 32, 1601-1608. 37. Kaplan S. S., Zdiarski U. E., Kuhns D. B. and Basford R. E. (1988) Blood 71, 677-683. 38. Brittebo E. B., H6gman P. G. and Brandt I. (1987) Food Chem. Toxicol. 25, 773-780. 39. Blumberg P. M. (1991) Molec. Carcinogen. 4, 339-344.
0898--6568/93 $6.00 + 0.00 © 1993 Pergamon Press LM
L I N D A N E DECREASES FORSKOLIN-STIMULATED CYCLIC AMP A C C U M U L A T I O N BUT DOES NOT M O D I F Y Gs IN RAT ENTEROCYTES I. CARRERO, N. RODRIGUEZ-HLr~CH~, L. G. GUUARRO, M. N. 1 ~ o o , M. A. I~h~z-At,nARS^NZ and J. C. PRIE'rO* Unidad de Neuroendocrinologia Molecular, Departamento de Bioquimica y Biologia Molecular, Universidad de Alcalfi, E-28871 Alcalfi de Henares, Spain (Received 3 November 1992; and accepted 5 February 1993) Abstract--Treatment of isolated rat enterecytes with the halogenated insecticide lindane (the y-isomer of hexachlorocyclohexane, HCCH) did not modify the general membrane fluidity (as estimated by a fluorescence polarization technique) nor the guanine nucleotide binding regulatory protein Gs (as studied by both ADP-ribosylation of its • subunit by cholera toxin and Gpp[NH]p stimulation of membrane adenylate cyclase activity). However, iindane decreased in a dose-dependent manner the effect of the diterpene forskolin on direct activation of the adenylate cyclase catalytic subunit. After 5 min of cell treatment with 0.5 mM lindane, the maximal stimulatory effect of forskolin (at 100 I~M) decreased by about 50%. There was a certain degree of specificity since 6-HCCH was indeed more potent, whereas dieldrin and endrin (non-lindane related halogenated compounds) behaved as findane, and ~- and /]-HCCH were poorly efficient on the inhibition of forskolin stimulation of adenylate cyclase activity. A similar effect of lindane was observed on receptorstimulated cyclic AMP accumulation by using vasoactive intestinal peptide instead of forskolin. The results on a non-receptor mediated effect of lindane on the adenylate cyclase catalytic subunit itself could be related to: (i) alterations of membrane microdomains surrounding this and other integral proteins which would result in modifications of their activities; and/or (ii) a reciprocal relation between the two main routes of signal transduction so that the activation of protein kinase C (or other Ca2+-dependent protein kinases) by lindane would lead to phosphorylation of the adenylate cyclase catalytic subunit. The simultaneous preincubation of enterocytes with lindane and a tumour-promoting phorbol ester resulted in additive inhibitory effects on forskolin stimulation of cyclic AMP accumulation which is apparently against the suggested involvement of protein kinase C in the mechanism of action of the insecticide. However, it cannot be discarded since this enzyme activity represents a family of isoforms so lindane and the phorbol ester could act on different isoenzymes. Key words: Lindane, Gs protein, forskolin, adenylate cyclase, protein kinase C, rat enterocytes. INTRODUCTION TIlE organochlorine insecticide lindane (the yisomer of hexaclorocyclohexane, ~,-HCCH) is widely used in veterinary and human medicine to control ectoparasites [1]. Lindane gains entry * Author to whom correspondenceshould be addressed. Abbreviations: BSA--bovine s e r u m albumin; DMSO--dimethyl-sulphoxide; DPH--diphenylhexatriene; Gpp[NH]p--guanyl-5'-ylimidodiphosphate;HCCH--hexachlorocyclohexane; IBMX--3-isobutyl-1-methylxanthine; PDD--~-phorbol 12,13-didecanoate; PKC--protein kinaseC; SDS--PAGE--sodiumdodecyisulphat~polyacrylamide gel electrophoresis; TPA--12-O-tetradecanoylphorbol-I3-acetate; VIP--vasoactive intestinal peptide. 453
into the body system as a food toxicant, by inhalation into lungs or by diffusion through the skin [2, 3]. As a highly lipophilic compound, lindane accumulates in the adipose tissue and in the membrane lipid bilayer of all cells [4]. The insecticide is a potent convulsant agent acting on the central nervous system [5] and produces a constellation o f other effects including potential carcinogenicity [6], liver oxidative stress [7], nephrotoxicity [8] and degenerative changes in the reproductive system [9, 10]. The mechanism of action o f lindane remains poorly understood. As it occurs with many other insecticides, biomembranes appear to be
454
I. CARItEROet al.
primary targets of lindane action [4]. Although controversy exists, studies with various model and native membranes indicate that lindane interacts with membrane lipids and accommodates into the membranes without causing perturbation of the general membrane fluidity [11]. However, lindane is able to perturbate discrete membrane domains such as those surrounding integral proteins including receptors and ionic channels [12, 13]. Furthermore, lindane and inositol are structurally related and the insecticide interferes with inositol-phospholipid turnover [14] and increases free intracellular Ca 2+ levels which can result in activation of protein kinase C (PKC) and other protein kinases [15]. It has been demonstrated in different types of cells that PKC activation by tumour-promoting phorbol esters and other agents can regulate signal transduction through the adenylate cyclase pathway [16, 17]. In this context, previous results have shown that lindane impairs the stimulatory activity of fladrenergic agonists on cyclic AMP accumulation in rat prostatic epithelial cells [18] and renal cortical tubules [19]. Lindane also inhibits cyclic AMP accumulation induced by the neuropeptide vasoactive intestinal peptide (VIP) in rat prostatic [20] and intestinal [21] epithelial cells but it does not modify the binding of VIP to the corresponding membrane receptors [21]. These features prompted us to study the possibility of a direct interaction of lindane with the adenylate cyclase pathway of signal transduction. For this purpose, we tested the effect of lindane treatment of isolated rat enterocytes on: (i) guanine nucleotid¢ regulatory protein Gs as assessed by experiments on cholera toxin-catalysed ADP-ribosylation of membranes, and adenylate cyclase stimulation by guanyl-5'-yl imidodiphosphate (Gpp[NH]p); and (ii) the catalytic subunit of adenylate cyclase by means of stimulation with the diterpene forskolin. We extended our analysis in order to study possible effects of lindane on membrane fluidity as estimated by fluorescence polarization measurements. An approach to the possibility of a PKC-mediated effect of lindane
on adenylate cyclase was performed with experiments with phorbol esters.
MATERIALS A N D METHODS Chemicals
Lindane and other organochlorine compounds were purchased from Pierce Scientific (Rockford, IL, U.S.A.). Highly purified VIP was obtained from Prof. V. Mutt (Karolinska Institute, Stockholm, Sweden). Forskolin, 12-O-tetradecanoylphorbol-13acetate (TPA), 4~t-phorbol 12,13-didecanoate (PDD), diphenylhexatriene (DPH), dimethylsulphoxide (DMSO), 3-isobutyl- l-methylxanthine (IBMX), bovine serum albumin (BSA), Gpp[NH]p, ATP, creatine phosphate, creatine phosphokinase, cholera toxin, and protein markers and chemicals for gel electrophoresis were from Sigma Quimica (Alcobendas-Madrid, Spain). Animals and cell membrane preparation
Adult male Wistar rats weighing 180-250 g (aged about 4 months) were used. The animals were decapitated and the jejunum excised quickly and used for enterocyte isolation as described previously [22]. Cell viability was always greater than 90% as estimated by exclusion of trypan blue dye. Basolateral plasma membranes were isolated from enterocytes as described elsewhere [23]. Protein concentration was determined [24] using BSA as a standard. Lindane treatment
In standard conditions, rat enterocytes (about 50 mg wet wt per ml Krebs-Ringer phosphate buffer, pH 7.4) or enterocyte basolateral membranes (1 mg protein per ml buffer) were preincubated for 5 rain at 25°C with lindane (0.5 mM in 0.25% DMSO) or 0.25% DMSO alone and washed. In some experiments, the preincubation mixture included also a phorbol ester (1 ~tM TPA or PDD). Cyclic A M P and adenylate cyclase activity
The stimulation of cyclic AMP accumulation by isolated enterocytes was studied as described previously [25]. Briefly, control or lindane-treated enterocytes (1 mg cell protein/ml) were incubated in standard conditions in the absence or presence of 100nM-100~tM forskolin for 45min at 15°C in 0.5 ml of a medium consisting of 35 mM Tris-HCI buffer (pH 7.5), 50 mM NaCI, 1.4% BSA, 1 mg/ml bacitracin and 0.2 mM IBMX. The reaction was stopped by the addition of 2.5 ml methanol. After
A B C D Mr
x 10-3
-84.0
-36.5 -26.6 -
÷ ÷-
CTX
FIG. 1. Autoradiograph of [32p]ADP-ribosylated membrane proteins of control and lindane-treated rat enterocytes. The cells were incubated with [32p]NAD in the absence and presence of cholera toxin (CTX) as described under Materials and Methods. Lanes A and B correspond to control conditions whereas lanes C and D are those of lindane-treated enteroeytes.
455
L i n d a n e effects in r a t enterocytes
centrifugation, aliquots of the supernatant were evaporated for cyclic AMP determination [26]. Adenylate cyclase activity was measured as reported previously [27] with minor modifications. Enterocyte basolateral plasma membranes (0.5 mg protein/mi) were incubated with 1.5mM ATP and 5mM MgSO4, an ATP-regenerating system (7.4mg/ml creatine phosphate and 1 mg/ml creatine kinase), l m M IBMX, l m M EDTA and lnM-0.1mM Gpp[NH]p in 0.1 ml of 25 mM triethanolamine-HC! buffer (pH 7.4) for 30 min at 30°C. The reaction was stopped by heating for 3 min. After refrigeration, 0.3 ml of an alumina slurry (0.25 g/ml in triethanolamine--HC! buffer, pH 7.4) was added. After centrifugation, cyclic AMP was determined as described above.
457
1¢
c
-6 v
! 4
n z
o -_~-/i |
i0~8
101-6
,. 10-4
[Gpp(NH)p] (M)
Cholera toxin-catalysed ADP-ribosylation ADP-ribosylation of enterocyte membranes was performed as described elsewhere [28]. Membranes (4 mg protein/ml) were incubated with activated cholera toxin (50 gg/ml) in 0.1 M phosphate buffer (pH 7.5) containing 5 mM ATP, 50 ~tM GTP, 5 ~tM [32P]NAD+, 2raM EDTA, 5raM MgCI,, and an ATP-regenerating system. After 30 rain at 30°C, 10% (w/v) trichioroacetic acid was added and the mixture was centrifuged and submitted to sodium dodecyisulphate-polyacrylamide gel electrophoresis (SDSPAGE). The resulting gels were then subjected to autoradiography. Fluorescence polarization measurement Control or lindane-treated rat enterocytes were washed, suspended in phosphate-buffered saline (about 0.2 absorbance units at 365 nm), incubated for 30 rain at 37°C with 0.6 pM DPH (dissolved in tetrahydrofuran) and then subjected to polarization analysis [29] using a Perkin-Elmer model LS-3B spectrofluorometer. DPH was excited at 357 nm, and emission was recorded at 430nm. Steady-state fluorescence polarization was computed according to the relationship P = (Ivv--lvh)/(lvv+lvh), where I w and lvh are the fluorescence intensifies observed with the analysing polarizer parallel and perpendicular to the polarizer excitation beam, respectively. The data were corrected for unequal transmission of differently polarized light and for intrinsic fluorescence. RESULTS The possibility that a lindane perturbation o f the general fluid phase of plasma membrane could result in uncoupling of the components of
FIG. 2. Dose-effect curves for Gpp[NH]p-stimulated adenylate cyclase activity in basolateral membranes of rat enterocytes. Membranes were preincubated for 5 rain at 25°C in the absence (Q) or presence (©) of 0.5 mM lindane. The results are given as the means -t-S.E,M. of nine determinations performed in triplicate. There were no statistically significant differences between the two groups.
the adenylate cyclase system o f signal transduction (receptors, G proteins and adenylate cyclase catalytic subunit) appears unlikely. The insecticide did not modify membrane fluidity in isolated rat cnterocytes since the fluorescence polarization value (and hence the membrane microviscosity) was nearly identical in control and lindane-treated cells: 0.179 +0.020 and 0.182+0.021, respectively (means +S.E.M., n = 5). Lindane showed also a lack of effect at the Gs protein level. The autoradiogram presented in Fig. 1 clearly indicates that lindane treatment of isolated enterocytes did not result in any modification of the extent of ADP-ribosylation of the • subunit of Gs protein in the presence of [32P]NAD, since the radioactivity incorporation into the 42,000 Mr membrane protein upon the addition of cholera toxin was very similar to that in control conditions. The same occurred in the study o f the response o f adenylate cyclase activity to Gpp[NH]p (acting via Gs protein) in enterocyte basolateral membranes; the ability of increasing concentrations of this nucleotide
458
I. CAaar~go et al.
lOG .~
75
I= ¢J
~
50
0.
25
g 8 2g
ols
O
-2
i
4
[Lindane] (mH)
Fro. 3. Dose dependence of lindane inhibition of forskolin-stimulated cyclic AMP accumulation in rat enterocytes. The cells were preincubated for 5 rain at 25°C in the absence or presence of increasing lindane concentrations. After washing, the cells were incubated without ([-]) and with 1 (V), 10 (A) or 100 (O) pM forskolin. Values are means of triplicates. to stimulate the enzyme activity was not significantly altered by lindane treatment of membranes (Fig. 2). We tested the effect o f enterocyte treatment with lindane on the subsequent stimulation of cyclic A M P accumulation in the isolated cells by forskolin, a diterpene which is assumed to act directly on the adenylate cyclase catalytic subunit [30]. Lindane inhibited the forskolin-
0
i
i
m [[10-7 I0~-~ I0~-m I0L4 [Forskolin] (H)
FIG. 5. Dose-response curves for forskolin stimulation of cyclic AMP accumulation in rat enterocytes. After cell preincubation for 5 min at 25°C in the absence (Q) or presence (O) of 0.5 mM lindan¢, the enterocytes were incubated with increasing concentrations of forskolin. Values are the means +S.E.M. of eight determinations performed in triplicate. stimulated cyclic A M P levels at submaximal (1-10 ~M) and maximal (100 oM) active concentrations of the diterpcne (Fig. 3). This
12
i0o .! 75 "~
o.
8
o
50
o.
<
4
25
o
o i
i
Time (min)
Fx6. 4. Effect of increasing time-periods on lindane inhibition of forskolin-stimulated cyclic AMP accumulation in rat enterocytes. The cells were preincubated at 25°C in the absence (closed symbols) and presence (open symbols) of 0.5 mM lindane for the indicated times. After washing, the cells were incubated without (squares) or with (circles) 100 I~M forskolin. Values are means of triplicates.
0
I
i
m III0-7 I0'-z I0"~ 10"4 [Forskofin](H)
Fro. 6. Specificity of lindane action. Rat enterocytes were preincubated for 5 min at 25°C in the absence (O) and presence of 0.5 mM lindane (O), ~-HCCH (<>), dieldrin (&), endrin (A), ~t-HCCH (ll) and flHCCH (D). After washing, the cells were incubated with increasing concentrations of forskolin. Values are means of triplicates.
Lindane effects in rat enterocytes
459
o
2C
I00 .:..~.:x:x::x:
ilili,i i
rxxxxxxx:
,, x
50
"-i:'.~//////2/.:~ '.. e////////..;:~//~
C
!i!!!:ii!iiiill L
TPA
PDD
L÷TPA
Fro. 7. Dose--response curves for VIP stimulation of cyclic AMP accumulation in rat enter~ytes. After cell preincubation for 5 rain at 25°e in the absence (Q) or presence (©) of 0.5 mM lindane, the enterocytes were incubated with increasing VIP concentrations. Values are the means+S.E.M, of eight determinations performed in triplicate.
F~G. 8. Effect of enterocyte preincubation with lindane and/or phorbol esters on the subsequent stimulation of cyclic AMP accumulation by forskolin. The cells were preincubated for 5 min at 25°C in the absence (C) and presence of 0.5 mM lindane, 1 IxM TPA or 1 ltM PDD. After washing, the cells were incubated with 0.1 mM forskolin. The bars represent the means _S.E.M. of six determinations performed in triplicate and they are expressed as a percentage of the control value by considering stimulated minus basal levels. *P<0.05 vs control, Student's t-test.
inhibitory effect depended on lindane concentration and became maximal at 0.3 m M insecticide. Relatively low doses (0.1-0.2mM) o f lindane elicited a small stimulatory response on the basal cyclic A M P levels, a feature that was maintained even after stimulation with forskolin, as indicated by the shoulders observed in Fig. 3. This unpredicted effect may constitute a further action of lindane but it deserves additional studies for a better understanding. In another context, experiments at increasing time-periods showed that forskolin-stimulated cyclic A M P accumulation resulted in maximal inhibition after 5 min exposure to the organochlorine compound (Fig. 4). The effect of lindane on the stimulatory action of increasing forskolin concentrations on cyclic A M P accumulation in rat enterocytes is shown in Fig. 5. Half-maximal cyclic A M P response was seen at about 10 liM forskolin in both control and lindane-treated cells. However, the maximal stimulatory effect (at 10011M forskolin) decreased by 50% after lindane treatment. The specificity o f lindane activity was analysed by using other H C C H
isomers and unrelated organochlorine compounds including dieldrin and endrin (Fig. 6). The 6-HCCH isomer showed the highest inhibitory effect on forskolin-stimulated cyclic A M P accumulation, whereas endrin and dieldrin behaved as lindane, and the ~- and /7-HCCH isomers elicited a poor effect at this level. The inhibitory effect o f lindane on the direct stimulatory action of forskolin on the adenylate cyclase catalytic subunit was also seen in receptor-mediated responses. Concentrationresponse experiments of VIP-stimulated cyclic A M P accumulation in rat enterocytes showed a decrease of VIP efficiency (by 50% at a maximally active, 100 nM dose o f VIP) after cell preincubation with 0.5 mM lindane (Fig. 7). The potency o f the stimulatory agent was again maintained by the insecticide since half-maximal cyclic A M P response was elicited at about 9 nM VIP in both control and lindane-treated enterocytes. The possibility that lindane could stimulate PKC activity leading to phosphorylation of the adenylate cyclase catalytic subunit (and/or other components o f this pathway of signal
[vm] (~)
460
I. CARREROet al.
transduction) was analysed by performing experiments with phorbol esters. Preincubation of rat enterocytes with the tumour-promoting phorbol ester TPA mimicked the inhibitory effect of lindane on forskolin-stimulated adenylate cyclase activity whereas PDD (a phorbol ester lacking in tumour-promoting activity) was inactive at this level. The simultaneous preincubation of the cells with lindane and TPA resulted in a merely additive inhibitory response. DISCUSSION In the present study, we have evaluated the effect of the insecticide hndane on Gs protein and adenylate cyclase catalytic subunit which are components of a major signal transduction pathway. Treatment of isolated enterocytes with lindane did not perturbate the general membrane fluidity (as estimated by DPH fluorescence polarization) nor Gs protein (as studied by both ADP-ribosylation with cholera toxin and Gpp[NH]p stimulation of membrane adenylate cyclase activity). However, the insecticide severely decreased the effect of the diterpene forskolin on direct activation of the adenylate cyclase catalytic subunit. The experiments were performed in standard conditions with 0.5 mM lindane, a concentration that is similar to those used in processes such as the inhibition of inositol phospholipid biosynthesis in pancreatic acini [31], the increase of calcium levels within synaptosomes [15] and the interaction with ~,-aminobutyric acid-activated chloride channels [32]. However, lower concentrations of lindane were needed to inhibit hepatocyte gap junctional intercellular communication [33] or t-butylbicyclophosphorotionate binding to chloride channels in brain membranes [34]. In the present experimental conditions, lindane did not modify the general membrane fluidity in isolated rat enterocytes, in close agreement with previous results obtained in fluid native membranes and model bilayers [11]. However, evidence exists that the insecticide may differently affect especially ordered micro-
domains surrounding integral membrane proteins and thus result in alterations of the corresponding activities [12,13]. Accordingly, it has been inferred that the perturbation of membrane fluidity by lindane may result in uncoupling of receptor-effector systems and be responsible at least in part for previous observations on the inhibitory action of the insecticide on VIP-stimulated cyclic AMP accumulation in rat enterocytes which was not accompanied by a modification in the binding of VIP to its specific receptors [21]. Gs protein transmits a signal to the catalytic adenylate cyclase subunit after the activation of a stimulatory membrane receptor by a specific agonist. Thus, Gs protein plays a pivotal role in the receptor-linked regulation of adenylate cyclase [35]. Experiments on cholera toxin ADP-ribosylation of the • subunit of Gs were performed for the identification of G protein in rat enterocytes, giving evidence that the extent of specific labelling was unaffected by lindane. This suggestion on the lack of effect of lindane on Gs protein was reinforced by the similar pattern of stimulation of adenylate cyclase activity in control or lindane-treated enterocyte basolateral membranes by increasing Gpp[NH]p concentrations. To our knowledge, there are no previous reports on the interaction of lindane with G proteins. Forskolin is a good research tool in the investigation of the adenylate cyclase system because it is a strong activator of this enzyme in membrane and intact cells. At the micromolar concentrations used in the present study, forskolin is likely to act directly on the catalytic subunit of adenylate cyclase and not via the Gs-cyclase complex [30]. The present results show the ability of lindane to inhibit the stimulatory action of forskolin on cyclic AMP accumulation in rat enterocytes due to a decrease in the efficiency of the diterpene (by 50% at a maximally effective forskolin concentration) that could result from a direct interaction of the insecticide with the adenylate cyclase catalytic subunit. It has been shown that forskolin can increase membrane fluidity and may achieve adenylate cyclase activation by
Lindane effectsin rat enterocytes binding to sites either at the polar surface of the enzyme or buried in the membrane bilayer [36]. Hence, it should be thought that the hydrophobic nature of lindane resulted in the displacement of forskolin from those sites with the corresponding decrease in the extent of cyclic AMP stimulation by the diterpene. This possibility appears unlikely since: (i) lindane-treated cells were extensively washed before using forskolin for cyclic AMP stimulation (although some quantity of the insecticide could remain retained by the cells); and (ii)we observed a similar inhibitory effect of lindane on receptorstimulated cyclic AMP accumulation, i.e. the signal transduction system for the neuropeptide VIP. These results are in agreement with previous studies on the regulation of receptormediated cyclic AMP levels by lindane, including stimulation of cyclic AMP accumulation by VIP in rat enterocytes [21], VIP [20] or the #-adrenergic agonist isoproterenol [19] in rat prostatic epithelial cells, and isoproterenol in rat renal cortical tubules [19]. However, our results on the potent inhibition of forskolinstimulated adenylate cyclase activity by lindane are to our knowledge the first observation on a possible direct non-receptor mediated interaction of this halogenated insecticide on the adenylate cyclase catalytic subunit. The marked inhibitory activity of lindane upon forskolin-stimulated cyclic AMP accumulation in rat enterocytes showed a relative degree of specificity as shown by experiments with various organochlorine substances which were powerful (6-HCCH), median (lindane and the non-related compounds dieldrin and endrin) or poor (the non-insecticide ~- and #HCCH isomers) agents at this level. There are some discrepancies in the pattern of specificity described in the present report and in previous studies on receptor-mediated cyclic AMP accumulation [18, 21], Ca 2+ mobilization from intracellular stores [37], potential tumorigenic effects [6] and binding to different epithelia [38]. It appears likely that the interaction of lindane and other organochlorine compounds with different tissues may depend on highly substrate-specific factors.
461
The observed inhibitory action of lindane on forskolin-stimulated adenylate cyclase activity in rat enterocytes could be related, at least in part, to a connection between the adenylate cyclase and the PKC-CaZ+-inositol triphosphate pathways of signal transduction. Lindane increases free cytosolic Ca 2+ levels [15] which could activate PKC and other protein kinases thereby resulting in phosphorylation of the adenylate cyclase catalytic subunit and/or other cell proteins. Considering the adenylate cyclase system, it would be a selective mechanism depending probably on the nature and/or position of each component in the cell membrane since neither Gs protein (present work) nor membrane receptors such as those for VIP [19] are modified after cell exposure to this halogenated insecticide. Moreover, the inositol-like structure of lindane may be related to the observed effect of this insecticide on the turnover of inositol phospholipids [14] and hence contribute to a connection between the two main routes of signal transduction. However, the present results with simultaneous preincubation of enterocytes with lindane and the tumour-promoting phorbol ester TPA (which activates PKC in a direct manner) indicate an additivity of their inhibitory effects on the subsequent stimulation of cyclic AMP accumulation by forskolin. This feature is at odds with the possibility of a direct interaction between lindane and PKC. However, it should be remembered that PKC represents a family of isozymes [39] with different activation patterns. Thus, it would be possible in the present system for TPA to interact with a PKC isoform, whereas lindane could activate a different one (or even other classes of protein kinases through a Ca~+-mediated mechanism). Clearly, more in-depth studies are required to exclude or accept definitively the role of PKC in this process. Regardless of the mechanism, the present results establish a modulatory effect of lindane and other organochlorine compounds on the adenylate cyclase catalytic subunit itself in the rat enterocyte, which may be the result of alterations of membrane microdomains surrounding this and other integral proteins,
462
I. CARaEROet al.
and/or reciprocal effects between the adenylate cyclase and the PKC-Ca2+-inositol triphosphate pathways of signal transduction. Acknowledgements--This work was supported by grants from the Fondo de Investigaciones Sanitarias de la Seguridad Social (88/904 and 92/1272) and the Direcci6n General de Investigaci6n Cientifica y T6cnica (PM89/96).
REFERENCES I. Wooley D., Zimmer L., Dodge D. and Swanson K. (1985) Neurotoxicology 6, 165-192. 2. Zhu J., Feng Z. and Chen J. (1986) Pestic. Biochem. Physiol. 25, 414-419. 3. Lopez-Aparicio P., Hoyo N. and PerezAlbarsanz M. A. (1988) Pestic. Biochem. Physiol. 31, 109-119. 4. Antunes-Madeira M. C. and Madeira V. M. C. (1985) Biochim. Biophys. Acta 820, 165-172. 5. Tussel J. M., Sufiol C., Gelpi E. and Rodriguez-Farre E. (1987) Arch. Toxicol. 60, 432-437. 6. Wolf G. L., Roberts D. W., Morrisey R. L., Greenman D. L., Allen R. R., Campbell W. L., Bergman H., Nesnow S. and Frith C. H. (1987) Carcinogenesis 8, 1889-1897. 7. Videla L. A., Simizu K., Barros S. B. M. and Junqueira V. B. C. (1989) Cell Biochem. Funct. 7, 179-183. 8. Andrews J. E. and Gray L. E. (1990) Toxicology 60, 99-107. 9. Srinivasan K., Ramesh H. P. and Radhakrishnamurty R. (1988) Bull. Environ. Contain. Toxicol. 41, 531-539. 10. Uphouse L. and Williams J. (1989) Toxicol. Lett. 48, 21-28. I I. Antunes-Madeira M. C. and Madeira V. M. C. 0989) Biochim. Biophys. Acta 982, 161-166. 12. Ogiso T., Iwaki M. and Moil K. (1981) Biochim. Biophys. Acta 649, 325-335. 13. Jones O. T., Froud R. J. and Lee A. G. (1985) Biochim. Biophys. Acta 12, 740-751. 14. Parries G. S. and Hokin-Neaverson M. (1985) J. biol. Chem. 260, 2687-2693. 15. Bondy S. C. and Halsail L. (1988) Neurotoxicology 9, 645-652. 16. Le Goas F., Amiel C. and Friedlander G. (1991) Am. J. Physiol. 261, F587-F592. 17. Ransjo M. and Lerner U. H. (1991) Calc. Tiss. Int. 49, 284-287.
18. Carrero I., Perez-Albarsanz M. A., Carmena M. J. and Prieto J. C. (1990) Pestic. Biochem. Physiol. 38, 197-203. 19. Lopez-Aparicio P., Recio M. N., Prieto J. C., Carmena M. J. and Perez-Albarsanz M. A. (1991) Life Sci. 49, 1141-1154. 20. Carrero I., Perez-Albarsanz M. A., Recio M. N. and Prieto J. C. (1992) Toxic. In Vitro 6, 7-9. 21. Carrero I., Fernandez-Moreno M. D., PerezAibarsanz M. A. and Prieto J. C. (1989) Biochem. biophys. Res. Commun. 159, 1391-1396.
22. Prieto J. C., Laburthe M. and Rosselin G. (1979) Eur. J. Biochem. 96, 229-237. 23. Young G. P., Morton C. L., Rose I. S., Taranto T. M. and Bhathal P. S. (1987) Gut 28, 57-62. 24. Lowry O. H., Rosebrough N. J., Farr A. L. and Randall R. J. (1951) J. biol. Chem. 193, 265-275. 25. Laburtbe M., Prieto J. C., Amiranoff B., Dupont C., Hui Bon Hoa D. and Rosselin G. (1979) Fur. J. Biochem. 96, 239-248. 26. Gilman A. G. (1970) Proc. natn. Acad. Sci. U.S.A. 74, 3424-3428. 27. Houslay M. D. and Gordon L. M. (1983) Curt. Top. Membr. Transp. 18, 179-231. 28. Couvineau A., Amiranoff B. and Laburthe M. (1986) J. biol. Chem. 261, 14,482-14,489. 29. Shinitzky M. and Barenholz Y. (1978) Biochim. Biophys. Acta 515, 367-394. 30. Seamon K. B. and Daly J. W. (1981) J. biol. Chem. 250, 9799-9801. 31. Crouch M. F. and Roberts M. L. (1985) Biochim. Biophys. Acta 844, 149-157. 32. Ogata N., Vogel S. M. and Narahashi T. (1988) FASEB J. 2, 2895-2900. 33. Klaunig J. E., Ruch R. J. and Weghorst C. M. (1990) Toxicol. appl. Pharmac. 102, 553-563. 34. Abalis I. M., Edelfrawi M. E. and Edelfrawi A. T. (1985) Pestic. Biochem. Physiol. 24, 95-102. 35. Spiegel A. M., Backlund P. S., Butrynski J. E., Jones T. L. Z. and Simonds W. F. (1991) Trends Biochem. Sci. 16, 338-340. 36. Whetton A. D., Needham L., Dodd N. J. F., Heyworth C. M. and Houslay M. D. (1983) Biochem. Pharmac. 32, 1601-1608. 37. Kaplan S. S., Zdiarski U. E., Kuhns D. B. and Basford R. E. (1988) Blood 71, 677-683. 38. Brittebo E. B., H6gman P. G. and Brandt I. (1987) Food Chem. Toxicol. 25, 773-780. 39. Blumberg P. M. (1991) Molec. Carcinogen. 4, 339-344.