Biochflnica et Biophysica Acre, i 127 (1992) 49-56
© 1992 Elsevier Science Publishers B.V. All rights reserved 0005-2760/92/$05.00
49
BBALIP 53961
Effects of sphingosine, albumin and unsaturated fatty acids on the activation and translocation of phosphatidate phosphohydrolases in rat hepatocytes Antonio Gomez-Mufioz, Essam H. Hamza and David N. Brinoley Department of Biochemistry and Lipid and Lipoprotein Research Group, Unirersityof Alberta, Edtn,vuon (Canada)
(Received 7 January 1992)
Key words: Albumin; Amphiphilic amine; Fatty acid: Plmsphatidate phosphohydrolase; Sphingosine: (Rat hcpamcytc) The activities of two phosphatidate phosphohydrolases were measured in cultured rat hepatocytes incubated with 0.1 mM albumin. The activity, which is inhibited by N-ethylmaleimide (PAP-I) is located in the cytosolic and membrane fractions. PAP-i activity is stimulated by Mg a~ and it can be translocatcd from the cytosol to the membranes by relatively low (0.5-1 raM) concentrations of fatty acids. In addition, higher concentrations (1-3 mM) of fatty acids cause an increase in the total PAP-I activity. Translocation of PAP-I activity in the hepatocytes is preferentially promoted by unsaturated fatty acids (C~s: ~, C~s:2, Cis:3, C20:4 and C20:5), rather than by saturated acids (Ci4:0, Cit,: 0, Cis:0). Increasing the extraceilular concentration of albumin from 30/zM to 1 mM displaces PAP-I activity from the membrane fraction. Sphi,~gosine, but not staurosporine, can inhibit the redistribution of PAP-I activity induced by oleate. The amphiphilic amines, sphingosine, chlorpromazine and propranolol, also decrease membrane-bound PAP-I activity in the absence of fatty acids, but they do not alter, significantly, the activity of the cytosolic PAP-I. In the presence of ! mM oleate, sphingosine, chlorpromazine and propranolol decrease the translocation of PAP-I from the cytosol to the membranes. The phosphohydrolase activity, which is insensitive to N-ethylmaleimide (PAP-2), is specifically located in the plasma membrane (Jamal, Z., Martin, A., Gomez-Mufioz, A. and Brindley, D.N. (1991) J. Biol. Chem. 266, 2988-2996) and it is not stimulated by Mg :+. Saturated fatty acids, albumin, sphingosiac and propranolol have no significant effects on PAP-2 activity. However, chlorpromazine decreases PAP-2 activity by about 14%. Linolenate, arachidonate and eicosapentaenoate at 1 mM also produced small (7-10%) decreases in PAP-2 activity. It is proposed that both PAP-i and PAP-2 activities may be involved in signal transduction, although the main function of PAP-1 seems to be involved in the synthesis of glycerolipids.
Introduction Phosphatidate phosphohydrolase (PAP) has classically been related to the synthesis of triacylglycerols, phosphatidylcholine and phosphatidylethanolamine. This enzyme (PAP-I) requires Mg 2+ for its activity and it is located in the cytosol and endoplasmic reticulum [1]. The cytosolic PAP-1 is thought t o represent an inactive form that becomes functional when it translo-
Correspondence to: D.N. Brindley, Departmen! of Biochemistry and Lipid and Lipoprotein Research Group, 328 Heritage Medical Research Centre, Faculty of Medicine, University of Alberta, Edmonton, Alberta, T6G 2S2 Canada. Abbreviations: C20:4, arachidonate; C2o:s, eicosapentaenoate; Cts:2, linolenate; Ct8:3, linolenate; Ct4:o, myristate; Cu,:a, palmitate; Cis: t, oleate; C,s:0, stearate; BSA, bovine serum albumin; LDH, lactate dehydrogenase; PMA, 4fl-phorbol 12-myristate 13-acetate; PAP, phosphatidate phosphohydrolase.
cates to the endoplasmic reticulum, in response to an increased fatty acid load [1,2]. The ability to translocate is also modulated by cyclic AMP [3], glucagon '[4], or okadaic acid [5] indicating a possible role of a phosphorylation mechanism. More recently PAP activity has been implicated in signal transduction in many cell types, following the agonist-stimulated breakdown of phosphatidylcholine via phospholipase D [6-9]. Phosphatidate can be rapidly formed in plasma membranes by the action of phospholipase D on phosphatidylcholine and then dephosphorylated to diacyiglycerol by PAP activity. It is well known that phosphatidate or lysophosphatidate [7,8,10-16] and diacylglycerol [17] are second messengers. Regulation of PAP activities could therefore control the balance between the signals. A second PAP activity (PAP-2) that does not require Mg 2+, and which is located in plasma membranes, has recently been described [18]. Specific assays using N-ethyimaleimide were developed to distin-
50 guish between the two PAP activities. PAP-I is inhibited by N-ethylmaleimide, whereas PAP-2 is not [18,19]. Amphiphilic amines can inhibit PAP activity [20,21] and more recently, sphingosine (a natural amphiphilic amine) has been shown to have this effect [16,18,22]. Sphingosine inhibits both PAP-I and PAP-2 activities in cell, free systems [18]. An objective of the present work was to determine whether sphingosine could modulate the distribution and activity of the two PAP activities in hepatocytes. In the case of PAP-l, C~8:~ causes both the translocation to the endovlasmic retie.. ulum and an activation of total activity. We, therefore, investigated the effects of sphingosine on both of these effects. Furthermore, we determined the effects of different fatty acids and albumin on the subcellular distribution of PAP-I and on the activity of PAP-2.
Experimental procedures Animals Male Wistar rats (about 200 g) were obtained from Charles River (Quebec, Canada). They were housed in a room that was lit from 08:00 to 20:00 h, and they were fed on Wayne Rodent BIox (Continental Grain, Chicago, IL, USA). The rats were used for the preparation of hepatocytes between 09: 00 and 11 : 00 h. Materials Fatty acids, sphingosine, chlorpromazine hydrochloride, DL.propranolol, 4/3-phorbol 12-myristate 13acetate (PMA), essentially fatty ;~id-frce BSA, L-aphosphatidic acid (from egg lecithin), L-~-phosphatidylcholine (Type V from egg yolk), N.ethyimaleimid¢ (NEM) and Triton X-100 were from Sigma Chemicals, St. Louis, MO, USA. Staurosporine was from Kamiya Biomedical, Thousand Oaks, CA, USA. Tctrahydrolipstatin was a gift from Dr. M.K. Meier, HoffmanLaRoche, Basel, Switzerland. The sources of other reagents are described by Gomez-Mufioz ct al. [23]. Preparation and culture of hepatoo'tes Hepatocytes were prepared and attached to collagen-coated tissue-culture dishes in a modified Leibovitz L-15 medium, containing 25 mM Hepes (pH 7.4), 10% (v/v) newborn-calf serum, 8.3 mM glucose and 5 mM galacto~ [24]. After 45-60 min at 37°C in humidified air, the unattached and non.viable cells were removed by replacing the medium. The incubation was continued for a further 4-5 h, and the medium was then replaced by modified Leibovitz medium, containing 0.1 mM essentially fatty acid-free BSA, in the absence of serum, and the cells were incubated for a further 18-20 h. The medium was then replaced and the BSA t;u:~:t:ntration was changed, as indicated. Additions of fatty acids, albumin, or amphiphiic amines
were made so that the total volume of the medium changed by less than 2%. Fatty acids were prepared in a 20% molar excess of KOH, dissolved by warming and then slowly pipetted, with shaking, into the BSA-containing medium [25,26]. At the times of incubation indicated, the cells were washed once with ice-cold 0.16 M NaCl. Cytosolic and membrane fractions were obtained by lysing the cells with digitonin [24], except that 0.1 mg of digitonin per ml was sonicated in 10 mM Hepes, adjusted to pH 7.4 with HCI, and containing 0.5 mM dithiothreitol, in the absence of sucrose. Samples were stored at -70°C until required for analysis.
Assays for PAP- i and PAP-2 acti~'ities PAP activities were determined by using N-ethylmaleimide to distinguish between the two activities [18,19]. For the assay of PAP-I, the substrate was prepared by sonicating 3.33 mM [3H]phosphatidate (0.4 Ci of ['~H]palmitate/mol)and 2.22 mM phosphatidylcholine in 5.56 mM EGTA and 5.56 mM EDTA. To every 9 vol of this mixture was added 1 vol of 100 mg of fatty acid-poor BSA/ml. PAP-I activity was determined in 0.1 ml of buffer containing: 100 mM Tris/maleate (pH 6.5), 1 mM dithiothreitol, 0.6 mM [3H]phosphatidate, 0.4 mM phosphatidylcholine and 0.2 mg of albumin, I mM EDTA, plus 1 mM EGTA carried through with the substrate, 1.5 mM MgCI 2, 120 ~M tetrahydrolipstatin and up to 200 /.tg of enzyme protein. Parallel incubations were performed after preincubating the enzyme for 10 min with 5 mM Nethylmalcimide. The difference in activity between these two different assays was taken as PAP-1 activity. The remaining N-ethylmaleimide insensitive activity is essentially PAP-2. However, this activity was also measured under optimum conditions [18,19], using an emulsion of potassium phosphatidate that w~:s prepared as above except that phosphatidylcholine was omitted. Each incubation contained in 0.1 mh 100 mM Tris/maleate buffer (pH 6.5), 1 mM dithiothreitol, 0.6 mM [3H]phosphatidate (0.4 Ci/mol), 0.2 mg of albumin, 1 mM EDTA plus I mM EGTA, that were carried through with the substrate, 5 mM N-ethylmaleimide, 0.5% Triton X-100, 120 p,M tetrahydrolipstatin and up to 100/zg of enzyme protein. In all cases the reactions were started by the addition of the substrate and stopped after 30 to 60 min with 2 ml of chloroform/methanol (19: I, v/v) containing 0.08% of olive oil as carrier. The amount of enzyme protein was adjusted so that less than 15% of the phosphatidate was converted to diacylglycerol. The latter was assayed after removing [3H]phosphatidate and [3H]palmitate with alumina [27]. The tetrahydrolipstatin was routinely added as an inhibitor of acylglyccrol iipase to protect against possible degradation of the diacylglycerol product [18,19].
51
Assay of lactate dchydrogenase LDH was measured as described previously [25,28]. The total content of LDH per dish of cells was 3.24 ± 0.1 /.tmol of lactate oxidized/min (means ± S.E. from 14 independent preparations) and 1 /zmol/min is equivalent to 1 unit of activity at 22°C. One dish of cells contained about 1.5 mg of protein, so that results can be readily calculated relative to protein if this is required for comparison with other work.
Expression of results PAP activities are expressed as means ± S.E., unless stated to the contrary, and the values for individual experiments were taken as means trom duplicate dishes, each assayed in duplicate or triplicate. In most experiments changes in relative activities are given because of differences in the absolute PAP activities in difl'e~kent preparations, of hepatocytes. Statistical significance was calculated by using a paired t-test. PAP activities are expressed relative to total LDH activities to compensate for small differences in the numbers of viable hepatocytes per dish and the r,:covery in the homogenates. Measurement of LDH activities also ensured for each condition that none of the compounds used caused lysis of the hepatocytes, since loss of viability in the hepatocytes is accompanied by a leakage of LDH into the medium. Any condition that produced iysis was excluded. Expression of the results relative to the total LDH is superior to presenting the results relative to protein or DNA, since non-viable and damaged cells can remain attached to the culture dishes. Also, some incubations contained relatively high concentrations of albumin, that could have made it
difficult to accurately determine cell protein concentrations. The activities of PAP-I in the cytosolic and membrane-associated compartments were calculated by correcting for the incomplete release of LDH, as previously reported [4]. Results
Effects of fatty acids on the actit'ities and subceihdar distribution of PAP-1 and PAP-2 b~ hepatocytes Rat hepatocytes showed a translocation of PAP-I activity from the cytosol to the tnembranes after incubation with oleate in the presence of 0.1 mM BSA, as expected from previous work [24]. This effect can be detected at a fatty acid concentration greater than 0.5 raM, or a fatty acid:albumin ratio greater than 5 (results not shown). Higher concentrations of oleate (1-3 mM) also increased the total PAP-I activity. At 3 mM oleate there was a 5-fold increase in membraneassociated PAP-I activity and a 2.4-fold increase in total activity. Similar results were obtained when a physiological mixture of fatty acids [29] was used (Table r,). The combination of the translocatior, and activatio, produced a 4.8 + 0.5-fo~d increase (mean + range for two independent experiments) in the membrane-associated PAP-I activity at 3 mM of the fatty acid mixture. Results at 1 mM of the fatty acid mixture are shown in Table I. The effects of seven other fatty acids at 1 mM on PAP-I activity after 30 min of incubation are also shown in Table 1. The saturated fatty acids (Cm4:0, C~c,:0, C~8:0) had relatively little effect on either PAP-I translocation or activation. However, all of the unsatu-
TABLE I Effects oJ'fiatty acid structure on the acti~'ation and subcelhdar distribution of PAP-! and PAP-2 actit'ilies in rat hepatocytes Monolayer cultures of rat hepatocytes were incubated for 30 min at 37°C with 0.1 mM essentially fatty-acid-free BSA and ! m M concentrations of the fatty acids indicated. Results are expressed relative to the control value of each fraction, and they are meanss: S.E. for the number of experiments indicated in parentheses. The absolute values for the cytosolic and microsomal PAP-I activities of controls were 0.75 ±0.20 and 0.49+0.10 nmol of diacylglycerol/min per unit of total LDH. The absolute value for the control PAP-2 activity in four experiments was I.(19+0.16 nmol of diacylglycerol/min per unit of total LDH. The significance of the difference from the appropriate control value is indicated by: * P < (}.(}5; * * P < 0.025; * * * P < ().01; * P < 0.005; ** P < 0.0005. The mixture of fatty acids consisted of Ci4:o, 2%, Ctc,: o, 28%; Ct~:,, 15r/, ; Cl~, i, 34%; Cla:2, 14%; Cix:,a, I% and C20:4, 6% (mol/100 moll and was designed to resemble the fatty acid concentrations in human blood
[2a]. Fatty acid added None (control) Ci4:o Cu,:0 Cts:o Ct~:l Cta:2 CtK:3 C2o:4 C2o:5 Mixed acids
Relative activity (%) cytosolic PAP-I
membrane PAP-I
total PAP-I
1()0 93+2 * 100+5 965:9 58+3** 615:4' 445:9 *** 39+4* 42+9 + 755:6 * * *
100 1(}8+ 6 1135:7 i185: 7 * 195:t:10' 206+20 * * * 297-t-32 * * * 3245:51 ** 3:05:60 ** 1955:31 **
100 995:3 1055:6 11)45:4 1095 4 ** 112± 5 * 135±~II * 1445:i9 * 14(1:t:17 * 118+ 10
total PAP-2 (4) (3) (4) (4) (15) (4) (5) (4) (4) (5}
100 102+2 101)5:4 10152 111{):t:5 98:t:2 93± I ** 905:3 * 925:2 ** 94±5
(4) (3) (4) (4) (9) (4) (4} (4) (4) (5)
52 rated fatty acids caused transiocation of PAP-1 from the cytosolic to the membrane fraction. Total PAP-1 activities were also significantly increased by 1 mM concentrations of C,a:~, Ci8:2, C!8:3- C.,o:4 and C2o:5 (Table I). The increase in total PAP-1 activity caused by C18.~ or Clg:: is small but statistically significant. The enhancements of total PAP-1 activity increase for the C~s series with the degree of unsaturation. When the concentration of BSA was raised to 0.6 raM, the fat~ acids at up to 2 raM, failed to affect total PAP-1 activity or its subcellular distribution (results not shown). The effect of albumin on PAP-1 activity was also measured in the absence of fatty acid additions. Hepatocytcs were maintained for 18-20 h in 30 ~M albumin and then incubated with increasing concentrations of albumin. This resulted in a progressive loss of PAP-I from the membrane to the cytosolic fraction. There was a small, but statistically significant increase (8%) in the total PAP-I activity when the concentration of albumin was increased from 30/zM to 100/zM (Table !I), but this is probably not really appreciable. A further objective of this work was to determine whether the activity of the N-ethylmaleimide insensitive enzyme (PAP-2), which was completely recovered in the membrane fraction, was altered by fatty acid availability. Only the polyunsaturated fatty acids (C ~a:3, C.,0:4 and C.,o:.~) affected PAP-2, causing a relatively small, but significant, decrease in its activity (Table l).
TABLE II
Effect of BSA on the subceihdar distribution of PAP-I aclit'ity Hepa:ocytes were kept in a low albumin concentration (30#M) for 18-20 h before incubating for 30 rain with the concentration of albumin indicated. All incubations contained some BSA in order to ensure the viability of the cells. Results are means ± S.E. for three independent experiments. Total activities are expressed relative to the result obtained at 30 # M albumin, where the absolute value for the total PAP-I activity was 0.61 +0.11 nmol of diacylglycerol/min per unit of total LDH. * P < 0.025, * * P < 0,005. BSA ( # M )
Relative PAP-I activity (%)
membrane-bound (%)
total 100 108± ! I vs. II ** 102 ± 7
(I) (11)
30 100
53 ± 4 50± 2
(III)
600
(IV)
1000
44±2 I vs, III * II vs. III ** 39 ± 8
I0{) :t: 8
Albumin did not cause any significant change on PAP-2 activity (results not shown).
Effects of amphiphilic amines on PAP actit'ities Amphiphilic cations inhibit PAP activity [2,20,21]. More recently the effects of sphingosine, propranolol and chlorpromazine have been tested in vitro using differential assays to distinguish between the two PAP activities [18]. All of these compounds were able to
T A B L E iil
l~]'~t'l of aml)hiphilic amin('s or! th(° aclit'ily and ,~qd~ct'lhlhlr distrihl(tiot! of PAP-I and PAP-2 a('tirities in
rat
htopatot'ytes
Monolaycr cultutes of rat hepat~x,),tes were incubated for 30 rain at 37°C with 0.1 mM e~entiaUy fatty-acid-free BSA and l(10 # M concentrations of sphingosine, ptopranolol, chlorptomazine in the presence or absence of i mM oleate, as indicated. Results are expressed relative to the control value of each fraction and they were means ± S,E, for the number of experiments indicated in parentheses. The absolute values for the ~,~t~olic, mictosomal and total PAP-I activities of controls were 0,79 ± 0,16, (I,43 ± 0.07 and i,22 ± 0.19 nmol of diacylglycerol/min per unit of total LDH, respectively, The ab,,~)lute activity of PAP-2 in control incubations was 0.88 ± 0.09 nmol of diacylglycerol/min per unit of total LDH. The levels of significance are: * P < 0,05: * * P < 0,025: * * * P < 0,0ilS: t p < 0,0005. Condition
Relative aclivity (~;.) cytosolic PAP-I
membrane PAP-I
total PAP-I
total PAP-2
(I) Conl~)i
I(N)
!00
100
(9)
100
(g)
(il}Oleate
100± 5
(9) (9)
qq± 4
94+14
(6)
(V)Chlotpromazine
~)± 5
80±4 (9) ! vs. ill * * * 88±4 (6) lvs. IV ** 79±5 (4) lvs. V * * * 95±4 (7)
100+ 5
(IV)Ptopranolol
183±14 I vs. II * 59± q ! vs. Ill * * * 60± q ivs. IV * * * ~1 ± 12 Ivs, V ** 122+18 !! vs. V! * * * 109±!5 il vs. VII * * * 123±23 ll vs. V l l l * * *
(9)
(lll)Sphing(~ine
59+ 4 I vs, II * 91 ± II
(Vl)Oleate+ sphingosine (Vll)Oleate+ ptopranolol (Vlil)Oleate+ chlorpromazine
84±15 !! vs, VI * 96+10 !i vs. VII * * 98± 8 ll vs. VIII * *
99±4
95±3
(6)
101+4
(3)
8 6 ± 2 (4) ivs. V * * * 100± 2 (7) 93± 4
(6)
9 2 + 3 (4) I vs. VIII *
53
inhibit both PAP-I and PAP-2 activities. However, the interpretation of those results was complicated since the effects depended on substrate presentation [18]. What is more important is the effect of these cations in whole cells. Fig. 1A, B shows that 100/.tM sphingosine counteracts the effect of l mM oleate in translocating PAP-I activity from the cytosol to the membranes. No significant change in the total PAP-1 activity was detected when the hepatocytes were incubated with sphingosine in the presence of Cts: t (Fig. 1C and Table liD. In the absence of fatty acids, sphingosine decreased PAP-1 activity in the membrane fraction (Fig. 1E and Table lit). About 30% of the membrane-bound PAP-l activity was lost {P < 0.05 for three independent experiments) after 5 rain of incubation (Fig. I E). Sphingosine decreased total PAP-l activity (Fig. IF)by 125
140
(A) Cytosol
100 •~
=o
75
so M
25 i
O0
I
50
i
I
100
i
I
150
|
I
200
[Sphingosine](l,zM)
Fig. 2. Effect of sphingosine on the activity and subcellular distribution of PAP-I in rat hepatocytes. Monolayer cultures of rat hepatocytes were incubated for 30 min at 37°Cwith 0.1 mM essentially fatty
acid free BSA and the concentration of sphingosine indicated. The activities of PAP-I in the cytosolic fraction (e). membrane fraction (A) and the total activity (B) are shown. Results are expressed relative to the control value of each fraction and they are means+ S.E.
(D) Cytosol
for three independent experiments. The iibsolute values for the cytosolic, membnme, and total PAP-I activities were 1.11±0.24, 0.51±0.15 and 1.62±0.26 nmol of diacylglycerol/min per unit of total LDH, respectively.
120
100
125
100 80 25 --,
60
3O0 (B) Membranes
125
250
100
200
75
Iso
50
100
25
50
0
(E) Membranes
.;
n-
140
140
(C) Total
I
I
I
I
2*0
~ *0
4 0*
(F) Total
120 120 100 100
6%
80 1 IO
i 20
i 30
, 40
60 50 0 Time (min)
1 I0
50
Fig. I. Effect of sphingosine on the activity and subcellular distribution of PAP-I in the presence and in the absence of oleate. Monolayer cultures of rat hepatocytes were incubated at 37°C with 0.1 mM essentially fatty-acid-free BSA for the times indicated in the presence (A, B, C) or the absence (D, E, F) of 1 mM oleatc and in the presence {closed symbols) or the absence (open symbols) of 100 p,M sphingosine. The relative activity of PAP-I in the cytosolic fraction (o. e), membrane fraction ( zx, ,, ) and total activity ( D, B ) is shown. Results are expressed relative to *he control value of each fraction. The values are means±S.E, for three to six independent experiments except for the 5 and 10 min points in A, B and C, where they are means± ranges of two experiments. The absolute values for the cytosolic, membrane and total PAP-I activities in the absence of Cts:l were ~.88±0.05, 0.37±0.09 and 1.25:t:0.04 nmol of diacylglycerol/ndn per unit of total LDH in six independent determinations.
about 15% ( P < 0 . 0 0 5 for three independent experiments) since there was no significant increase in PAP-1 activity in the cytosol. This effect on membrane assftciated PAP-I activity was maximal after 30 min when this was decreased by about 60% (P < 0.005 for five independent experiments). The optimum concentration of sphingosine (100/,tM) was chosen from the results obtained in Fig. 2. The effects of propranolol and chiorpromazine, in the absence and in the presence of C~s:~ are shown in Table !11. Propranolol and chlorpromazine decreased the membrane-bound PAP-I activity, without significantly changing the activity of cytosolic PAP-I, when C~s:~ was absent (Table III). This loss of the membrane-bound PAP-1 resulted in a decrease of the total PAP-1 activity, in the presence of I mM oleate, 100 t~M of chiorpromazine, or propranolol prevented the cytosolic PAP-1 activity from translocating to the membranes but no significant changes in the total enzyme activity were detected. Sphingosine or propranolol (10-200 # M ) f a i l e d to alter PAP-2 activity in the presence, or absence of fatty acids. However, 100 # M chlorpromazine caused a significant decrease in PAP-2 activity by about 14% (Table !11). Sphingosine might change the activities of PAP-1 and PAP-2 by interaction with phosphatidate during the enzyme assay. We determined PAP-I activity in membrane and cytosolic fractions in the presence of 0.8 mM phosphatidate instead of ().6 mM phosphatidate. This did not change the measured PAP-I activity from cells incubated for 2 min or 30 min with sphingosine in the absence, or presence of fatty acids in two
54 independent experiments. Similarly, the effect of chlorpromazine on PAP-2 was not overcome when the activity was determined in the presence of 0.8 mM phosphatidate or 1.0 mM phosphatidate in three independent experiments (results not shown). Concentrations of phosphatidate above 0.8 mM for the assay of PAP-l, or 1,0 mM for the assay of PAP-2, were not used, since these are inhibitory [18]. Furthermore, an increase of 0.2 mM in substrate concentration in the assay should have been sufficient to counteract the effect of any sphingosine carried through with the sample. We also measured PAP-I and PAP-2 with suboptimum substrate concentrations near to the apparent Km of the enzymes, as previously established using the preparation obtained from the hepatocytes. These concentrations were 0.25 mM phosphatidate in the case of PAP-I and 0.05 mM phosphatidate in the case of PAP-2. When PAP-I activity was measured, constant ratios of PA/PC and Mg-'+/PA/PC were maintained in the assay [27]. No change in the apparent Km values of PAP-I or PAP-2 resulted after incubating hepatocytes with sphingosine. The effect of sphingosine was also tested to see whether it could alter the stimulation of PAP-I at relatively high concentrations of oleate. Sphingosine at up to 200/~ M did not significantly change the stimulation of 2.3 fold in the total PAP-I activity caused by 3 mM oleate (results not shown). Sphingosine is an inhibitor of protein kinase C [30] and so we investigated the effects of other protein kinase C inhibitors. Staurosporine (10 nM-!0 /~M), which blocks the stimulation of glycogen phosphorylase by C~..~ in hepatocytes [29], failed to affect PAP-2 activity, or PAP-I activity and its distribution. PMA, a protein kinase C activator, also failed to alter PAP-1 or PAP-2 activities, or their distribution within the cells. The effect of PMA was also tested in the presence of 2 /~M of the Ca-'* ionophore A23187, which synergizes some of its effects [31]. Discussion
Unsaturated fatty acids, but not saturated fatty acids, activate glycogen phosphorylase by a mechanism probably involving protein kinase C [29]. This implies that unsaturated fatty acids could cause cell activation, in addition to providing substrates for lipid metabolism. The present work demonstrated, for the first time, that the unsaturated fatty acids are preferentially able to alter PAP-I activity and promote translocation of PAPI in whole cells (Table l). These translocations of PAP-I were obtained at molar concentrations of unsaturated fatty acids to albumin of > 5:1. However, an increase in the total enzyme activity requires higher fatty acid concentrations [Ref. 24 and Table I]. Saturated fatty acids produced little change in the activity
and translocation of PAP-1. Previous work with subcellular fractions from liver demonstrated that C~6:0 or palmitoyl-CoA translocated PAP-I to the endoplasmic reticulum, but that these effects were less than for CIs:~ or oleoyi-CoA, respectively [32]. The molecular structure of the fatty acids may be relevant for their mechanisms of action. The fluidity of the membranes increases with the degree of unsaturation of the fatty acids and this could then facilitate the functioning of PAP-1. The preferential response to polyunsaturated fatty acids may stimulate triacylglycerol and, therefore, decrease their potential toxic effects. C2,:.~ is as potent as C~8:a and C20:4 in translocating and increasing total PAP-I activity (Table I). C20:s decreases both triacylglyceroi synthesis and secretion from the liver [26,33]. However, it was among the most effective fatty acids in stimulating the secretion of phosphatidylcholine and lysophosphatidyicholine [26]. Therefore, it is possible that C,,:.~ is preferentially incorporated into phosphatidylcholine via diacylglycerol, or via a deacylation/reacylation cycle. In contrast to the present results, C2o:.~ has been claimed to inhibit PAP activity in perfused rat liver [34]. The authors of the latter report also failed to see the translocation of PAP activity from the eytosolic to the microsomal fraction in the presence of oleate. More recently [35], the same group reported that feeding fish oils enriched in C2,:5 decreases phosphatidate hydrolysis involving a possible decrease in PAP activity. However, their assay system was developed for determining PAP activities in lung samples [36], and it may not accurately measure hepatic PAP activities [37]. Failure to see the translocation of PAP activity induced by unsaturated fatty acids may be caused by the use of a drastic homogenization before the separation of cellular fractions [24]. Furthermore, a differential assay should be used to distinguish between the two hepatic PAP activities, if the decrease in hepatic triacylglycerol formation induced by C20:.~ is not caused by the inhibition of PAP-I activity one might expect another key enzyme, such as the acyI-CoA: 1,2.diacylglycerol acyltransfcrase, to be inhibited and this has been reported [33]. The longer-term administration of fish oil might also cause a decreased PAP-! activity [35], but this should be confirmed using the differential assay. One objective of the present work was to assess the ratio of fatty acid to albumin that could cause translocation of PAP, or increases in total activity. These effects arc seen at ratios of C~s: i to albumin of > 5 : 1, and > 10:1, respectively. Cis:~ had a very similar potency to that observed with a physiological mixture of fatty acids (Table I). The serum concentration of albumin in man is normally about 0.6 mM [38,39] and physiologically fatty acid concentrations can rise to about 1.2 mM [40]. However, in severe stress, trauma, hypoxia, toxic conditions, myocardial infarction and
55 diabetic ketoacidosis fatty acid concentrations can rise to 2 mM or higher [41,42]. Arterial concentrations of fatty acid,,; were reported to rise to about 3 mM after cardiac surgery [43], and during by-pass surgery blood fatty acid concentrations of up to 7 mM have been recorded (G.D. Lopaschuk, personal communication). Therefore, ratios of fatty acid: albumin up to or greater than five can be found in vivo. High ratios of fatty acid:albumin are especially likely to occur in cases of hypoalbuminemia associated with aging, malnutrition and renal disease. increasing extracellular albumin concentrations in the absence of exogenous fatty acids displaced PAP-l activity from the membranes to the cytosol (Table II). Albumin was also able to decrease glycogen phosphorylase activity in hepatoeytes [29]. These inhibitions might be caused by the removal of fatty acids from the plasma membranes by the albumin, which in turn, might cause a decrease in the concentration of the intracellular fatty acid pool. "Fable !! indicates a role for intraceilular fatty acids in regulating PAP-I activity. These results with whole cells complement those in a cell-free system, where albumin prevented the transloeating effects of exogenous fatty acids [44]. Amphiphilic amines inhibit PAP activity [1,20,21]. Sphingosine can modulate PAP-I activation, since, in the presence of fatty acids, it decreases the activity associated with the membranes (Table I11, Figs. 1 and 2). The inhibition of the oleate-induced translocation of PAP-I by sphingosine cannot be explained simply by a stoichometric interaction of the amphiphilic cation with the anionic fatty acid since the ratio of fatty acid:,,,phingosine used was 10:1. The effect of sphingosine is probably caused by an interaction with acidic iipids, e.g., phosphatidate in the endoplasmic reticulum. In the absence of fatty acids, sphingosine, propranolol and chlorpromazine decreased the membrane PAP-! activity without significantly altering the cytosolic form of the enzyme. As a consequence, the total PAP-I activity was also decreased. This effect was not seen in the presence of C~: t. The other effect of very high fatty acid concentrations was to increase total PAP-I activity. Sphingosine at up to 200 pM failed to alter significantly the stimulation of PAP-I activity induced by 3 mM oleate. The activation of glycogen phosphorylase by unsaturated fatty acids probably involved protein kinase C since both sphingosine and staurosporine reversed the effect [29]. This observation is supported by a later report in which the a,/3 and y subspecies of protein kinase C are all activated dramatically by the simultaneous addition of diacylglycerol and the unsaturated fatty acids in the presence of phosphatidylserine [45]. An involvement of protein kinase C on PAP-1 activity is unlikely since staurosporine had no significant effects on the subcellular distribution, or on the total
activity of PAP-1. Furthermore, PMA alone or in combination with the Ca 2÷ ionophore A23187, failed to produce any significant changes on the distributio:-~ or total activity of PAP-I. These results are similar to those of Pollard and Brindley [28] at 1 /~M A23187. However, the latter authors showed a small increase (13%) in PAP activity at 5 ~M A23187. We chose the concentrations of 2/zM A23187 and 25 p,M PMA for the present work since this produced maximum stimulations of glycogen phosphorylase [23]. The differential assays used in this work allowed us to detect a small, but significant, decrease (7-10%) of PAP-2 activity caused by the polyunsaturated fatty acids, Cis:.a, C20:4 , and C20:5 (Table I). The physiological relevance, if any, of this small inhibition of PAP-2 remains to be established. Sphingosine, propranolol, staurosporine, PMA and the Ca 2. ionophore A23187 failed to alter significantly the PAP-2 activity that was extracted from the hepatocytes. However, this does not mean that PAP-2 in hepatocytes and working on endogenous substrate could not be inhibited by sphingosine. Chlorpromazine decreased PAP-2 activity in the presence and absence of Ci8: t. In terms of signal transduction, cytosolic PAP-l could, theoretically, interact with the inner surface of the plasma membrane and translocate when phosphatidate accumulates. Sphingosine inhibited PAP activity in NGI08-15 cells [15] and human neutrophils [22]. Although the authors did not distinguish between PAP-1 and PAP-2 activities, the changes reported were probably mainly on PAP-l activity. Sphingosine inhibits ceil signalling through inhibiting kinase C [30,46,47] and through inhibiting PAP activity [16,18,22]. The present work demonstrates an inhibition of PAP-I activity by preventing the translocation of the cytosolic enzyme to the membranes that is caused by the action of unsaturated fatty acids. The activation and translocation of PAP-1 are believed to protect cells against the toxic accumulation of fatty acids, acyl-CoA esters and phosphatidate. The synthesis of phosphatidate and diacylglycerol de novo may also produce intraeellular messages that cause cell activation [48-50] in addition to those produced by the agonist-stimulated formation of these iipids in the plasma membrane. The present work indicates that sphingosine could also regulate the balance of phosphatidate and diacylglycerol that is formed de novo through its ability to prevent the translocation of PAP-I. it can also acutely inhibit PAP2 activity [18] and, thereby, modify signal transduction.
Acknowledgements We thank the Alberta Heritage Foundation tor Medical Research, the Canadian Medical Research Council and the Juvenile Diabetes Foundation for financial support.
56 References I Brindley, D.N. 119841 Prog. Lipid Res. 23, 115-133. 2 Brindley, D,N. (1988) in Phosphatidate Phosphohydrolase (BrindIcy. D,N., ed.), Vol. 1, pp. 1-77, CRC Series in Enzyme Biology, CRC Press, Boca Raton. 3 Butterwith, S.C., Martin, A. and Brindley, D.N. 119841 Biochem. J. 222, 487-493. 4 Pittner, R.A., Fears, R. and Brindley, D.N, 119851 Biochem. J. 230, 5~-534. 5 Gomez-Mufioz, A., Hatch, G.M., Martin, A., Jamal, Z., Vance, D.E, and Brindley, D,N. 119921 FEBS Lett., 301, 103-105. 6 Bocckino, S.B., Blackmore, P.F., Wilson, P.B. and Exton, J.H. (1987) J, Biol. Chem. 262, 15309-15315, 7 Loffelholz, K. 119891 Biochem. Pharmacol, 38, 1543-1549, 8 Billah, M,M, and Anthes, J,C. (1990) Biochem, J. 269, 281-291, 9 Billah, M.M,, Eekel, S., Mullmann, T.J., Egan, R.W. and Siegel, M,I, (1989) J, Biol. Chem. 264, 17l)69-17077. 10 Rossi, F., Frzeskow, M,, Dellabia. V,, Calretti, P, and Jandini, G. (19901 Biochem, Biophys, Res, Commtin. 168, 32l)-327. 11 Murayama, T. and Ui, M, (19871J. Biol. Chem, .6,,, "~ '~ 5522-5529. 12 I~uldry, S,A,, Bass, D,A,, Cousart, S.L. and McCall, C.E. (19911 J. Biol, Chem, 266, 4173-4179, 13 Van Corven, E,J,, Groenink, A,, Jalink, K,, Eichholtz, T. and Moolenaar, W,H, (19891 Cell 59, 45-54. 14 Tsai, M,H,, Yu, C,L., Wei, F.S, and Stacey, D.W. (1989) Science 2,k~, 522-526, 15 Verheijden, G.F., Verlaan, i., Schlessinger, J. and Moolenaar, W,H. 119911 Cell Regul, 9, 615-620, 16 Lavie, Y,, Piterman, O. and Liscovitch, M. (1990) FEBS Lett. 277, 7-111. 17 Kaibuchi, K., Takai, Y. and Nishizuka, Y. (1981) J, Biol, Chcm, ~6, 7146-7149, 18 Jamal, Z,, Martin, A., Gomez-Mufioz, A. and Brindley, D,N, (1991) J, Biol. Chem. 266, 2988-2996, 19 Martin, A., Gomez-Mufioz, A., Jamal, Z, and Brindley, D,N, (1~!) Moth, Enzymol, 197, 553-563, 211 lk~wley, M,, Ctx~ling, J,, Burditt, S,L. and Brindley, D.N, (19771 Biochem, J, 165, 447-454, 21 Martin, A., Hopewell, R,, Martin-Sanz, P,, Morgan, J,E, and Brindley, D,N, (19861 Biochim. Biophys, Acta 876, 581-591, 22 Mullmann, T J,, Siegel, M,I,, Egan, R,W, and Billah, M,M, (19911 J. Bit)l, Chem, 266, 2(113-2016, 23 Gomez-Mu6t~ A,, Hales, P,, Brindley, D,N, and Sancho, M,J, (1989) Biochcm, J, ~ .6., 417-423, 24 Ca~ales, C,, Mangiapane, E,H, and Brindley, D,N, (19841 Biochem, J, 219, 911-916,
25 Pittner, R.A., Fears, R. and Brindley, D,N. 119851 Biochem, J. 225, 455-462. 26 Graham, A., Zammit, V.A. and Brindley, D.N. (19881 Biochem. J, 249, 727-733. 27 Martin, A., Hales, P. and Brindley, D.N. (1987) Biochem. J. 245, 347-355. 28 Pollard, A,D. and Brindley, D,N, (19841 Biochem, J. 217, 461-469. 29 Gomez-Mufioz, A., Hales, P. and Brindley, D,N, (1991) Biochem. J. 276, 209-215. 30 Rando, R.R. (19881 FASEB J. 2, 2348-2355. 31 van de Werve, G., Proitto, J. and Jeanrenaud, B. 119851 Biochem. J. 231,511-516. 32 Martin-Sanz, P,, Hopewell, R, and Brindley, D.N. (1984) FEBS Left, 175, 284-288. 33 Rustan, A,C., Nossen, J,O,, Christiansen, E,N, and Drevon, C.A. (1988) J, Lipid Res, 29, 1417-1426. 34 Wong, S,-H. and Marsh, J.B. (19881 Metabolism 37, 1177-1181. 35 Halminsky, M.A,, Marsh, J,B. and Harrison, E.H. (1991) J. Nutr. 121, 1554-1561, 36 Walton, P.A, and Po~smayer, F, (1985) Anal. Bioch~m. 151, 479-486. 37 Sturton, R.G. and Brindley, D,N, (1978) Biochem. J. 171,263-266, 38 Halliwell, B, 119881 Biochem, Pharmacol. 37, 569-571. 39 Colli, S., Marinovich, M., Stragliotto, E,, Gall C.L. and Tremoli, E. 11989) Biochem. Pharmacol. 38, 39[19-3912. 411 Specter, A,A., Santos, E,C., Ashbrook, ,I.D. and Fletcher, J.E. 11970) Ann. N.Y Acad, Sci. 226, 247-258. 41 AIberti, K,G,M.M. and Press, C.M. (1982) in Complications of Diabetes (Keen, H, and Jarret, J., eds.), pp. 231-270, Edward Arnold, London, 42 Mueller, H. and Ayres, S,M. 119781 An). J. Cardiol. 42, 363-371. 43 Svensson, S,, Svedjehohn, R. Ekroth, R,, Milocco, !., Nilsson, F., Sabel, K,G, and WilUam-OIsson, G. (1990) J. Thorac. Cardiovasc. Surg, 99, 11163-11173. 44 ttopewell, R., MartinoSanz, P., Martin, A., Saxton, J. and BrindIcy, D.N. 119851 Biochem. J. 232, 485-491. 45 Shinomura, T., Asaoka, Y., Oka, M., Yoshida, K. and Nishizuka, Y, (1991) Proc, Natl. Acad. Sci. USA 88, 5149-4153. 46 ttannun, Y,A. and Bell, R.M. (1989) Science 243, 50{1-507. 47 Merrill, A.H. and Stevens, V.L. (19891 Biochim. Biophys. Acta 1010, 131-139. 4tq Farcsc, R.V., Konda, T.S., Davis, J.S., Standaert, M.L., Pellet, R J, and Cooper, D,R. 119871 Science 236, 586-589. 49 Ishizuka, T., Cooper, D.R,, Hernandez, H., Buckley, D., Standaert, M. and Farese, R.V. 1199111Diabetes 39, 181-190. 50 Rossi, F,, Grzescowiak, M., Della Bianca, V. and Sbartati, A. ( 1991 ) J. Biol. Chem. 266, 8034-8038.
© 1992 Elsevier Science Publishers B.V. All rights reserved 0005-2760/92/$05.00
49
BBALIP 53961
Effects of sphingosine, albumin and unsaturated fatty acids on the activation and translocation of phosphatidate phosphohydrolases in rat hepatocytes Antonio Gomez-Mufioz, Essam H. Hamza and David N. Brinoley Department of Biochemistry and Lipid and Lipoprotein Research Group, Unirersityof Alberta, Edtn,vuon (Canada)
(Received 7 January 1992)
Key words: Albumin; Amphiphilic amine; Fatty acid: Plmsphatidate phosphohydrolase; Sphingosine: (Rat hcpamcytc) The activities of two phosphatidate phosphohydrolases were measured in cultured rat hepatocytes incubated with 0.1 mM albumin. The activity, which is inhibited by N-ethylmaleimide (PAP-I) is located in the cytosolic and membrane fractions. PAP-i activity is stimulated by Mg a~ and it can be translocatcd from the cytosol to the membranes by relatively low (0.5-1 raM) concentrations of fatty acids. In addition, higher concentrations (1-3 mM) of fatty acids cause an increase in the total PAP-I activity. Translocation of PAP-I activity in the hepatocytes is preferentially promoted by unsaturated fatty acids (C~s: ~, C~s:2, Cis:3, C20:4 and C20:5), rather than by saturated acids (Ci4:0, Cit,: 0, Cis:0). Increasing the extraceilular concentration of albumin from 30/zM to 1 mM displaces PAP-I activity from the membrane fraction. Sphi,~gosine, but not staurosporine, can inhibit the redistribution of PAP-I activity induced by oleate. The amphiphilic amines, sphingosine, chlorpromazine and propranolol, also decrease membrane-bound PAP-I activity in the absence of fatty acids, but they do not alter, significantly, the activity of the cytosolic PAP-I. In the presence of ! mM oleate, sphingosine, chlorpromazine and propranolol decrease the translocation of PAP-I from the cytosol to the membranes. The phosphohydrolase activity, which is insensitive to N-ethylmaleimide (PAP-2), is specifically located in the plasma membrane (Jamal, Z., Martin, A., Gomez-Mufioz, A. and Brindley, D.N. (1991) J. Biol. Chem. 266, 2988-2996) and it is not stimulated by Mg :+. Saturated fatty acids, albumin, sphingosiac and propranolol have no significant effects on PAP-2 activity. However, chlorpromazine decreases PAP-2 activity by about 14%. Linolenate, arachidonate and eicosapentaenoate at 1 mM also produced small (7-10%) decreases in PAP-2 activity. It is proposed that both PAP-i and PAP-2 activities may be involved in signal transduction, although the main function of PAP-1 seems to be involved in the synthesis of glycerolipids.
Introduction Phosphatidate phosphohydrolase (PAP) has classically been related to the synthesis of triacylglycerols, phosphatidylcholine and phosphatidylethanolamine. This enzyme (PAP-I) requires Mg 2+ for its activity and it is located in the cytosol and endoplasmic reticulum [1]. The cytosolic PAP-1 is thought t o represent an inactive form that becomes functional when it translo-
Correspondence to: D.N. Brindley, Departmen! of Biochemistry and Lipid and Lipoprotein Research Group, 328 Heritage Medical Research Centre, Faculty of Medicine, University of Alberta, Edmonton, Alberta, T6G 2S2 Canada. Abbreviations: C20:4, arachidonate; C2o:s, eicosapentaenoate; Cts:2, linolenate; Ct8:3, linolenate; Ct4:o, myristate; Cu,:a, palmitate; Cis: t, oleate; C,s:0, stearate; BSA, bovine serum albumin; LDH, lactate dehydrogenase; PMA, 4fl-phorbol 12-myristate 13-acetate; PAP, phosphatidate phosphohydrolase.
cates to the endoplasmic reticulum, in response to an increased fatty acid load [1,2]. The ability to translocate is also modulated by cyclic AMP [3], glucagon '[4], or okadaic acid [5] indicating a possible role of a phosphorylation mechanism. More recently PAP activity has been implicated in signal transduction in many cell types, following the agonist-stimulated breakdown of phosphatidylcholine via phospholipase D [6-9]. Phosphatidate can be rapidly formed in plasma membranes by the action of phospholipase D on phosphatidylcholine and then dephosphorylated to diacyiglycerol by PAP activity. It is well known that phosphatidate or lysophosphatidate [7,8,10-16] and diacylglycerol [17] are second messengers. Regulation of PAP activities could therefore control the balance between the signals. A second PAP activity (PAP-2) that does not require Mg 2+, and which is located in plasma membranes, has recently been described [18]. Specific assays using N-ethyimaleimide were developed to distin-
50 guish between the two PAP activities. PAP-I is inhibited by N-ethylmaleimide, whereas PAP-2 is not [18,19]. Amphiphilic amines can inhibit PAP activity [20,21] and more recently, sphingosine (a natural amphiphilic amine) has been shown to have this effect [16,18,22]. Sphingosine inhibits both PAP-I and PAP-2 activities in cell, free systems [18]. An objective of the present work was to determine whether sphingosine could modulate the distribution and activity of the two PAP activities in hepatocytes. In the case of PAP-l, C~8:~ causes both the translocation to the endovlasmic retie.. ulum and an activation of total activity. We, therefore, investigated the effects of sphingosine on both of these effects. Furthermore, we determined the effects of different fatty acids and albumin on the subcellular distribution of PAP-I and on the activity of PAP-2.
Experimental procedures Animals Male Wistar rats (about 200 g) were obtained from Charles River (Quebec, Canada). They were housed in a room that was lit from 08:00 to 20:00 h, and they were fed on Wayne Rodent BIox (Continental Grain, Chicago, IL, USA). The rats were used for the preparation of hepatocytes between 09: 00 and 11 : 00 h. Materials Fatty acids, sphingosine, chlorpromazine hydrochloride, DL.propranolol, 4/3-phorbol 12-myristate 13acetate (PMA), essentially fatty ;~id-frce BSA, L-aphosphatidic acid (from egg lecithin), L-~-phosphatidylcholine (Type V from egg yolk), N.ethyimaleimid¢ (NEM) and Triton X-100 were from Sigma Chemicals, St. Louis, MO, USA. Staurosporine was from Kamiya Biomedical, Thousand Oaks, CA, USA. Tctrahydrolipstatin was a gift from Dr. M.K. Meier, HoffmanLaRoche, Basel, Switzerland. The sources of other reagents are described by Gomez-Mufioz ct al. [23]. Preparation and culture of hepatoo'tes Hepatocytes were prepared and attached to collagen-coated tissue-culture dishes in a modified Leibovitz L-15 medium, containing 25 mM Hepes (pH 7.4), 10% (v/v) newborn-calf serum, 8.3 mM glucose and 5 mM galacto~ [24]. After 45-60 min at 37°C in humidified air, the unattached and non.viable cells were removed by replacing the medium. The incubation was continued for a further 4-5 h, and the medium was then replaced by modified Leibovitz medium, containing 0.1 mM essentially fatty acid-free BSA, in the absence of serum, and the cells were incubated for a further 18-20 h. The medium was then replaced and the BSA t;u:~:t:ntration was changed, as indicated. Additions of fatty acids, albumin, or amphiphiic amines
were made so that the total volume of the medium changed by less than 2%. Fatty acids were prepared in a 20% molar excess of KOH, dissolved by warming and then slowly pipetted, with shaking, into the BSA-containing medium [25,26]. At the times of incubation indicated, the cells were washed once with ice-cold 0.16 M NaCl. Cytosolic and membrane fractions were obtained by lysing the cells with digitonin [24], except that 0.1 mg of digitonin per ml was sonicated in 10 mM Hepes, adjusted to pH 7.4 with HCI, and containing 0.5 mM dithiothreitol, in the absence of sucrose. Samples were stored at -70°C until required for analysis.
Assays for PAP- i and PAP-2 acti~'ities PAP activities were determined by using N-ethylmaleimide to distinguish between the two activities [18,19]. For the assay of PAP-I, the substrate was prepared by sonicating 3.33 mM [3H]phosphatidate (0.4 Ci of ['~H]palmitate/mol)and 2.22 mM phosphatidylcholine in 5.56 mM EGTA and 5.56 mM EDTA. To every 9 vol of this mixture was added 1 vol of 100 mg of fatty acid-poor BSA/ml. PAP-I activity was determined in 0.1 ml of buffer containing: 100 mM Tris/maleate (pH 6.5), 1 mM dithiothreitol, 0.6 mM [3H]phosphatidate, 0.4 mM phosphatidylcholine and 0.2 mg of albumin, I mM EDTA, plus 1 mM EGTA carried through with the substrate, 1.5 mM MgCI 2, 120 ~M tetrahydrolipstatin and up to 200 /.tg of enzyme protein. Parallel incubations were performed after preincubating the enzyme for 10 min with 5 mM Nethylmalcimide. The difference in activity between these two different assays was taken as PAP-1 activity. The remaining N-ethylmaleimide insensitive activity is essentially PAP-2. However, this activity was also measured under optimum conditions [18,19], using an emulsion of potassium phosphatidate that w~:s prepared as above except that phosphatidylcholine was omitted. Each incubation contained in 0.1 mh 100 mM Tris/maleate buffer (pH 6.5), 1 mM dithiothreitol, 0.6 mM [3H]phosphatidate (0.4 Ci/mol), 0.2 mg of albumin, 1 mM EDTA plus I mM EGTA, that were carried through with the substrate, 5 mM N-ethylmaleimide, 0.5% Triton X-100, 120 p,M tetrahydrolipstatin and up to 100/zg of enzyme protein. In all cases the reactions were started by the addition of the substrate and stopped after 30 to 60 min with 2 ml of chloroform/methanol (19: I, v/v) containing 0.08% of olive oil as carrier. The amount of enzyme protein was adjusted so that less than 15% of the phosphatidate was converted to diacylglycerol. The latter was assayed after removing [3H]phosphatidate and [3H]palmitate with alumina [27]. The tetrahydrolipstatin was routinely added as an inhibitor of acylglyccrol iipase to protect against possible degradation of the diacylglycerol product [18,19].
51
Assay of lactate dchydrogenase LDH was measured as described previously [25,28]. The total content of LDH per dish of cells was 3.24 ± 0.1 /.tmol of lactate oxidized/min (means ± S.E. from 14 independent preparations) and 1 /zmol/min is equivalent to 1 unit of activity at 22°C. One dish of cells contained about 1.5 mg of protein, so that results can be readily calculated relative to protein if this is required for comparison with other work.
Expression of results PAP activities are expressed as means ± S.E., unless stated to the contrary, and the values for individual experiments were taken as means trom duplicate dishes, each assayed in duplicate or triplicate. In most experiments changes in relative activities are given because of differences in the absolute PAP activities in difl'e~kent preparations, of hepatocytes. Statistical significance was calculated by using a paired t-test. PAP activities are expressed relative to total LDH activities to compensate for small differences in the numbers of viable hepatocytes per dish and the r,:covery in the homogenates. Measurement of LDH activities also ensured for each condition that none of the compounds used caused lysis of the hepatocytes, since loss of viability in the hepatocytes is accompanied by a leakage of LDH into the medium. Any condition that produced iysis was excluded. Expression of the results relative to the total LDH is superior to presenting the results relative to protein or DNA, since non-viable and damaged cells can remain attached to the culture dishes. Also, some incubations contained relatively high concentrations of albumin, that could have made it
difficult to accurately determine cell protein concentrations. The activities of PAP-I in the cytosolic and membrane-associated compartments were calculated by correcting for the incomplete release of LDH, as previously reported [4]. Results
Effects of fatty acids on the actit'ities and subceihdar distribution of PAP-1 and PAP-2 b~ hepatocytes Rat hepatocytes showed a translocation of PAP-I activity from the cytosol to the tnembranes after incubation with oleate in the presence of 0.1 mM BSA, as expected from previous work [24]. This effect can be detected at a fatty acid concentration greater than 0.5 raM, or a fatty acid:albumin ratio greater than 5 (results not shown). Higher concentrations of oleate (1-3 mM) also increased the total PAP-I activity. At 3 mM oleate there was a 5-fold increase in membraneassociated PAP-I activity and a 2.4-fold increase in total activity. Similar results were obtained when a physiological mixture of fatty acids [29] was used (Table r,). The combination of the translocatior, and activatio, produced a 4.8 + 0.5-fo~d increase (mean + range for two independent experiments) in the membrane-associated PAP-I activity at 3 mM of the fatty acid mixture. Results at 1 mM of the fatty acid mixture are shown in Table I. The effects of seven other fatty acids at 1 mM on PAP-I activity after 30 min of incubation are also shown in Table 1. The saturated fatty acids (Cm4:0, C~c,:0, C~8:0) had relatively little effect on either PAP-I translocation or activation. However, all of the unsatu-
TABLE I Effects oJ'fiatty acid structure on the acti~'ation and subcelhdar distribution of PAP-! and PAP-2 actit'ilies in rat hepatocytes Monolayer cultures of rat hepatocytes were incubated for 30 min at 37°C with 0.1 mM essentially fatty-acid-free BSA and ! m M concentrations of the fatty acids indicated. Results are expressed relative to the control value of each fraction, and they are meanss: S.E. for the number of experiments indicated in parentheses. The absolute values for the cytosolic and microsomal PAP-I activities of controls were 0.75 ±0.20 and 0.49+0.10 nmol of diacylglycerol/min per unit of total LDH. The absolute value for the control PAP-2 activity in four experiments was I.(19+0.16 nmol of diacylglycerol/min per unit of total LDH. The significance of the difference from the appropriate control value is indicated by: * P < (}.(}5; * * P < 0.025; * * * P < ().01; * P < 0.005; ** P < 0.0005. The mixture of fatty acids consisted of Ci4:o, 2%, Ctc,: o, 28%; Ct~:,, 15r/, ; Cl~, i, 34%; Cla:2, 14%; Cix:,a, I% and C20:4, 6% (mol/100 moll and was designed to resemble the fatty acid concentrations in human blood
[2a]. Fatty acid added None (control) Ci4:o Cu,:0 Cts:o Ct~:l Cta:2 CtK:3 C2o:4 C2o:5 Mixed acids
Relative activity (%) cytosolic PAP-I
membrane PAP-I
total PAP-I
1()0 93+2 * 100+5 965:9 58+3** 615:4' 445:9 *** 39+4* 42+9 + 755:6 * * *
100 1(}8+ 6 1135:7 i185: 7 * 195:t:10' 206+20 * * * 297-t-32 * * * 3245:51 ** 3:05:60 ** 1955:31 **
100 995:3 1055:6 11)45:4 1095 4 ** 112± 5 * 135±~II * 1445:i9 * 14(1:t:17 * 118+ 10
total PAP-2 (4) (3) (4) (4) (15) (4) (5) (4) (4) (5}
100 102+2 101)5:4 10152 111{):t:5 98:t:2 93± I ** 905:3 * 925:2 ** 94±5
(4) (3) (4) (4) (9) (4) (4} (4) (4) (5)
52 rated fatty acids caused transiocation of PAP-1 from the cytosolic to the membrane fraction. Total PAP-1 activities were also significantly increased by 1 mM concentrations of C,a:~, Ci8:2, C!8:3- C.,o:4 and C2o:5 (Table I). The increase in total PAP-1 activity caused by C18.~ or Clg:: is small but statistically significant. The enhancements of total PAP-1 activity increase for the C~s series with the degree of unsaturation. When the concentration of BSA was raised to 0.6 raM, the fat~ acids at up to 2 raM, failed to affect total PAP-1 activity or its subcellular distribution (results not shown). The effect of albumin on PAP-1 activity was also measured in the absence of fatty acid additions. Hepatocytcs were maintained for 18-20 h in 30 ~M albumin and then incubated with increasing concentrations of albumin. This resulted in a progressive loss of PAP-I from the membrane to the cytosolic fraction. There was a small, but statistically significant increase (8%) in the total PAP-I activity when the concentration of albumin was increased from 30/zM to 100/zM (Table !I), but this is probably not really appreciable. A further objective of this work was to determine whether the activity of the N-ethylmaleimide insensitive enzyme (PAP-2), which was completely recovered in the membrane fraction, was altered by fatty acid availability. Only the polyunsaturated fatty acids (C ~a:3, C.,0:4 and C.,o:.~) affected PAP-2, causing a relatively small, but significant, decrease in its activity (Table l).
TABLE II
Effect of BSA on the subceihdar distribution of PAP-I aclit'ity Hepa:ocytes were kept in a low albumin concentration (30#M) for 18-20 h before incubating for 30 rain with the concentration of albumin indicated. All incubations contained some BSA in order to ensure the viability of the cells. Results are means ± S.E. for three independent experiments. Total activities are expressed relative to the result obtained at 30 # M albumin, where the absolute value for the total PAP-I activity was 0.61 +0.11 nmol of diacylglycerol/min per unit of total LDH. * P < 0.025, * * P < 0,005. BSA ( # M )
Relative PAP-I activity (%)
membrane-bound (%)
total 100 108± ! I vs. II ** 102 ± 7
(I) (11)
30 100
53 ± 4 50± 2
(III)
600
(IV)
1000
44±2 I vs, III * II vs. III ** 39 ± 8
I0{) :t: 8
Albumin did not cause any significant change on PAP-2 activity (results not shown).
Effects of amphiphilic amines on PAP actit'ities Amphiphilic cations inhibit PAP activity [2,20,21]. More recently the effects of sphingosine, propranolol and chlorpromazine have been tested in vitro using differential assays to distinguish between the two PAP activities [18]. All of these compounds were able to
T A B L E iil
l~]'~t'l of aml)hiphilic amin('s or! th(° aclit'ily and ,~qd~ct'lhlhlr distrihl(tiot! of PAP-I and PAP-2 a('tirities in
rat
htopatot'ytes
Monolaycr cultutes of rat hepat~x,),tes were incubated for 30 rain at 37°C with 0.1 mM e~entiaUy fatty-acid-free BSA and l(10 # M concentrations of sphingosine, ptopranolol, chlorptomazine in the presence or absence of i mM oleate, as indicated. Results are expressed relative to the control value of each fraction and they were means ± S,E, for the number of experiments indicated in parentheses. The absolute values for the ~,~t~olic, mictosomal and total PAP-I activities of controls were 0,79 ± 0,16, (I,43 ± 0.07 and i,22 ± 0.19 nmol of diacylglycerol/min per unit of total LDH, respectively, The ab,,~)lute activity of PAP-2 in control incubations was 0.88 ± 0.09 nmol of diacylglycerol/min per unit of total LDH. The levels of significance are: * P < 0,05: * * P < 0,025: * * * P < 0,0ilS: t p < 0,0005. Condition
Relative aclivity (~;.) cytosolic PAP-I
membrane PAP-I
total PAP-I
total PAP-2
(I) Conl~)i
I(N)
!00
100
(9)
100
(g)
(il}Oleate
100± 5
(9) (9)
qq± 4
94+14
(6)
(V)Chlotpromazine
~)± 5
80±4 (9) ! vs. ill * * * 88±4 (6) lvs. IV ** 79±5 (4) lvs. V * * * 95±4 (7)
100+ 5
(IV)Ptopranolol
183±14 I vs. II * 59± q ! vs. Ill * * * 60± q ivs. IV * * * ~1 ± 12 Ivs, V ** 122+18 !! vs. V! * * * 109±!5 il vs. VII * * * 123±23 ll vs. V l l l * * *
(9)
(lll)Sphing(~ine
59+ 4 I vs, II * 91 ± II
(Vl)Oleate+ sphingosine (Vll)Oleate+ ptopranolol (Vlil)Oleate+ chlorpromazine
84±15 !! vs, VI * 96+10 !i vs. VII * * 98± 8 ll vs. VIII * *
99±4
95±3
(6)
101+4
(3)
8 6 ± 2 (4) ivs. V * * * 100± 2 (7) 93± 4
(6)
9 2 + 3 (4) I vs. VIII *
53
inhibit both PAP-I and PAP-2 activities. However, the interpretation of those results was complicated since the effects depended on substrate presentation [18]. What is more important is the effect of these cations in whole cells. Fig. 1A, B shows that 100/.tM sphingosine counteracts the effect of l mM oleate in translocating PAP-I activity from the cytosol to the membranes. No significant change in the total PAP-1 activity was detected when the hepatocytes were incubated with sphingosine in the presence of Cts: t (Fig. 1C and Table liD. In the absence of fatty acids, sphingosine decreased PAP-1 activity in the membrane fraction (Fig. 1E and Table lit). About 30% of the membrane-bound PAP-l activity was lost {P < 0.05 for three independent experiments) after 5 rain of incubation (Fig. I E). Sphingosine decreased total PAP-l activity (Fig. IF)by 125
140
(A) Cytosol
100 •~
=o
75
so M
25 i
O0
I
50
i
I
100
i
I
150
|
I
200
[Sphingosine](l,zM)
Fig. 2. Effect of sphingosine on the activity and subcellular distribution of PAP-I in rat hepatocytes. Monolayer cultures of rat hepatocytes were incubated for 30 min at 37°Cwith 0.1 mM essentially fatty
acid free BSA and the concentration of sphingosine indicated. The activities of PAP-I in the cytosolic fraction (e). membrane fraction (A) and the total activity (B) are shown. Results are expressed relative to the control value of each fraction and they are means+ S.E.
(D) Cytosol
for three independent experiments. The iibsolute values for the cytosolic, membnme, and total PAP-I activities were 1.11±0.24, 0.51±0.15 and 1.62±0.26 nmol of diacylglycerol/min per unit of total LDH, respectively.
120
100
125
100 80 25 --,
60
3O0 (B) Membranes
125
250
100
200
75
Iso
50
100
25
50
0
(E) Membranes
.;
n-
140
140
(C) Total
I
I
I
I
2*0
~ *0
4 0*
(F) Total
120 120 100 100
6%
80 1 IO
i 20
i 30
, 40
60 50 0 Time (min)
1 I0
50
Fig. I. Effect of sphingosine on the activity and subcellular distribution of PAP-I in the presence and in the absence of oleate. Monolayer cultures of rat hepatocytes were incubated at 37°C with 0.1 mM essentially fatty-acid-free BSA for the times indicated in the presence (A, B, C) or the absence (D, E, F) of 1 mM oleatc and in the presence {closed symbols) or the absence (open symbols) of 100 p,M sphingosine. The relative activity of PAP-I in the cytosolic fraction (o. e), membrane fraction ( zx, ,, ) and total activity ( D, B ) is shown. Results are expressed relative to *he control value of each fraction. The values are means±S.E, for three to six independent experiments except for the 5 and 10 min points in A, B and C, where they are means± ranges of two experiments. The absolute values for the cytosolic, membrane and total PAP-I activities in the absence of Cts:l were ~.88±0.05, 0.37±0.09 and 1.25:t:0.04 nmol of diacylglycerol/ndn per unit of total LDH in six independent determinations.
about 15% ( P < 0 . 0 0 5 for three independent experiments) since there was no significant increase in PAP-1 activity in the cytosol. This effect on membrane assftciated PAP-I activity was maximal after 30 min when this was decreased by about 60% (P < 0.005 for five independent experiments). The optimum concentration of sphingosine (100/,tM) was chosen from the results obtained in Fig. 2. The effects of propranolol and chiorpromazine, in the absence and in the presence of C~s:~ are shown in Table !11. Propranolol and chlorpromazine decreased the membrane-bound PAP-I activity, without significantly changing the activity of cytosolic PAP-I, when C~s:~ was absent (Table III). This loss of the membrane-bound PAP-1 resulted in a decrease of the total PAP-1 activity, in the presence of I mM oleate, 100 t~M of chiorpromazine, or propranolol prevented the cytosolic PAP-1 activity from translocating to the membranes but no significant changes in the total enzyme activity were detected. Sphingosine or propranolol (10-200 # M ) f a i l e d to alter PAP-2 activity in the presence, or absence of fatty acids. However, 100 # M chlorpromazine caused a significant decrease in PAP-2 activity by about 14% (Table !11). Sphingosine might change the activities of PAP-1 and PAP-2 by interaction with phosphatidate during the enzyme assay. We determined PAP-I activity in membrane and cytosolic fractions in the presence of 0.8 mM phosphatidate instead of ().6 mM phosphatidate. This did not change the measured PAP-I activity from cells incubated for 2 min or 30 min with sphingosine in the absence, or presence of fatty acids in two
54 independent experiments. Similarly, the effect of chlorpromazine on PAP-2 was not overcome when the activity was determined in the presence of 0.8 mM phosphatidate or 1.0 mM phosphatidate in three independent experiments (results not shown). Concentrations of phosphatidate above 0.8 mM for the assay of PAP-l, or 1,0 mM for the assay of PAP-2, were not used, since these are inhibitory [18]. Furthermore, an increase of 0.2 mM in substrate concentration in the assay should have been sufficient to counteract the effect of any sphingosine carried through with the sample. We also measured PAP-I and PAP-2 with suboptimum substrate concentrations near to the apparent Km of the enzymes, as previously established using the preparation obtained from the hepatocytes. These concentrations were 0.25 mM phosphatidate in the case of PAP-I and 0.05 mM phosphatidate in the case of PAP-2. When PAP-I activity was measured, constant ratios of PA/PC and Mg-'+/PA/PC were maintained in the assay [27]. No change in the apparent Km values of PAP-I or PAP-2 resulted after incubating hepatocytes with sphingosine. The effect of sphingosine was also tested to see whether it could alter the stimulation of PAP-I at relatively high concentrations of oleate. Sphingosine at up to 200/~ M did not significantly change the stimulation of 2.3 fold in the total PAP-I activity caused by 3 mM oleate (results not shown). Sphingosine is an inhibitor of protein kinase C [30] and so we investigated the effects of other protein kinase C inhibitors. Staurosporine (10 nM-!0 /~M), which blocks the stimulation of glycogen phosphorylase by C~..~ in hepatocytes [29], failed to affect PAP-2 activity, or PAP-I activity and its distribution. PMA, a protein kinase C activator, also failed to alter PAP-1 or PAP-2 activities, or their distribution within the cells. The effect of PMA was also tested in the presence of 2 /~M of the Ca-'* ionophore A23187, which synergizes some of its effects [31]. Discussion
Unsaturated fatty acids, but not saturated fatty acids, activate glycogen phosphorylase by a mechanism probably involving protein kinase C [29]. This implies that unsaturated fatty acids could cause cell activation, in addition to providing substrates for lipid metabolism. The present work demonstrated, for the first time, that the unsaturated fatty acids are preferentially able to alter PAP-I activity and promote translocation of PAPI in whole cells (Table l). These translocations of PAP-I were obtained at molar concentrations of unsaturated fatty acids to albumin of > 5:1. However, an increase in the total enzyme activity requires higher fatty acid concentrations [Ref. 24 and Table I]. Saturated fatty acids produced little change in the activity
and translocation of PAP-1. Previous work with subcellular fractions from liver demonstrated that C~6:0 or palmitoyl-CoA translocated PAP-I to the endoplasmic reticulum, but that these effects were less than for CIs:~ or oleoyi-CoA, respectively [32]. The molecular structure of the fatty acids may be relevant for their mechanisms of action. The fluidity of the membranes increases with the degree of unsaturation of the fatty acids and this could then facilitate the functioning of PAP-1. The preferential response to polyunsaturated fatty acids may stimulate triacylglycerol and, therefore, decrease their potential toxic effects. C2,:.~ is as potent as C~8:a and C20:4 in translocating and increasing total PAP-I activity (Table I). C20:s decreases both triacylglyceroi synthesis and secretion from the liver [26,33]. However, it was among the most effective fatty acids in stimulating the secretion of phosphatidylcholine and lysophosphatidyicholine [26]. Therefore, it is possible that C,,:.~ is preferentially incorporated into phosphatidylcholine via diacylglycerol, or via a deacylation/reacylation cycle. In contrast to the present results, C2o:.~ has been claimed to inhibit PAP activity in perfused rat liver [34]. The authors of the latter report also failed to see the translocation of PAP activity from the eytosolic to the microsomal fraction in the presence of oleate. More recently [35], the same group reported that feeding fish oils enriched in C2,:5 decreases phosphatidate hydrolysis involving a possible decrease in PAP activity. However, their assay system was developed for determining PAP activities in lung samples [36], and it may not accurately measure hepatic PAP activities [37]. Failure to see the translocation of PAP activity induced by unsaturated fatty acids may be caused by the use of a drastic homogenization before the separation of cellular fractions [24]. Furthermore, a differential assay should be used to distinguish between the two hepatic PAP activities, if the decrease in hepatic triacylglycerol formation induced by C20:.~ is not caused by the inhibition of PAP-I activity one might expect another key enzyme, such as the acyI-CoA: 1,2.diacylglycerol acyltransfcrase, to be inhibited and this has been reported [33]. The longer-term administration of fish oil might also cause a decreased PAP-! activity [35], but this should be confirmed using the differential assay. One objective of the present work was to assess the ratio of fatty acid to albumin that could cause translocation of PAP, or increases in total activity. These effects arc seen at ratios of C~s: i to albumin of > 5 : 1, and > 10:1, respectively. Cis:~ had a very similar potency to that observed with a physiological mixture of fatty acids (Table I). The serum concentration of albumin in man is normally about 0.6 mM [38,39] and physiologically fatty acid concentrations can rise to about 1.2 mM [40]. However, in severe stress, trauma, hypoxia, toxic conditions, myocardial infarction and
55 diabetic ketoacidosis fatty acid concentrations can rise to 2 mM or higher [41,42]. Arterial concentrations of fatty acid,,; were reported to rise to about 3 mM after cardiac surgery [43], and during by-pass surgery blood fatty acid concentrations of up to 7 mM have been recorded (G.D. Lopaschuk, personal communication). Therefore, ratios of fatty acid: albumin up to or greater than five can be found in vivo. High ratios of fatty acid:albumin are especially likely to occur in cases of hypoalbuminemia associated with aging, malnutrition and renal disease. increasing extracellular albumin concentrations in the absence of exogenous fatty acids displaced PAP-l activity from the membranes to the cytosol (Table II). Albumin was also able to decrease glycogen phosphorylase activity in hepatoeytes [29]. These inhibitions might be caused by the removal of fatty acids from the plasma membranes by the albumin, which in turn, might cause a decrease in the concentration of the intracellular fatty acid pool. "Fable !! indicates a role for intraceilular fatty acids in regulating PAP-I activity. These results with whole cells complement those in a cell-free system, where albumin prevented the transloeating effects of exogenous fatty acids [44]. Amphiphilic amines inhibit PAP activity [1,20,21]. Sphingosine can modulate PAP-I activation, since, in the presence of fatty acids, it decreases the activity associated with the membranes (Table I11, Figs. 1 and 2). The inhibition of the oleate-induced translocation of PAP-I by sphingosine cannot be explained simply by a stoichometric interaction of the amphiphilic cation with the anionic fatty acid since the ratio of fatty acid:,,,phingosine used was 10:1. The effect of sphingosine is probably caused by an interaction with acidic iipids, e.g., phosphatidate in the endoplasmic reticulum. In the absence of fatty acids, sphingosine, propranolol and chlorpromazine decreased the membrane PAP-! activity without significantly altering the cytosolic form of the enzyme. As a consequence, the total PAP-I activity was also decreased. This effect was not seen in the presence of C~: t. The other effect of very high fatty acid concentrations was to increase total PAP-I activity. Sphingosine at up to 200 pM failed to alter significantly the stimulation of PAP-I activity induced by 3 mM oleate. The activation of glycogen phosphorylase by unsaturated fatty acids probably involved protein kinase C since both sphingosine and staurosporine reversed the effect [29]. This observation is supported by a later report in which the a,/3 and y subspecies of protein kinase C are all activated dramatically by the simultaneous addition of diacylglycerol and the unsaturated fatty acids in the presence of phosphatidylserine [45]. An involvement of protein kinase C on PAP-1 activity is unlikely since staurosporine had no significant effects on the subcellular distribution, or on the total
activity of PAP-1. Furthermore, PMA alone or in combination with the Ca 2÷ ionophore A23187, failed to produce any significant changes on the distributio:-~ or total activity of PAP-I. These results are similar to those of Pollard and Brindley [28] at 1 /~M A23187. However, the latter authors showed a small increase (13%) in PAP activity at 5 ~M A23187. We chose the concentrations of 2/zM A23187 and 25 p,M PMA for the present work since this produced maximum stimulations of glycogen phosphorylase [23]. The differential assays used in this work allowed us to detect a small, but significant, decrease (7-10%) of PAP-2 activity caused by the polyunsaturated fatty acids, Cis:.a, C20:4 , and C20:5 (Table I). The physiological relevance, if any, of this small inhibition of PAP-2 remains to be established. Sphingosine, propranolol, staurosporine, PMA and the Ca 2. ionophore A23187 failed to alter significantly the PAP-2 activity that was extracted from the hepatocytes. However, this does not mean that PAP-2 in hepatocytes and working on endogenous substrate could not be inhibited by sphingosine. Chlorpromazine decreased PAP-2 activity in the presence and absence of Ci8: t. In terms of signal transduction, cytosolic PAP-l could, theoretically, interact with the inner surface of the plasma membrane and translocate when phosphatidate accumulates. Sphingosine inhibited PAP activity in NGI08-15 cells [15] and human neutrophils [22]. Although the authors did not distinguish between PAP-1 and PAP-2 activities, the changes reported were probably mainly on PAP-l activity. Sphingosine inhibits ceil signalling through inhibiting kinase C [30,46,47] and through inhibiting PAP activity [16,18,22]. The present work demonstrates an inhibition of PAP-I activity by preventing the translocation of the cytosolic enzyme to the membranes that is caused by the action of unsaturated fatty acids. The activation and translocation of PAP-1 are believed to protect cells against the toxic accumulation of fatty acids, acyl-CoA esters and phosphatidate. The synthesis of phosphatidate and diacylglycerol de novo may also produce intraeellular messages that cause cell activation [48-50] in addition to those produced by the agonist-stimulated formation of these iipids in the plasma membrane. The present work indicates that sphingosine could also regulate the balance of phosphatidate and diacylglycerol that is formed de novo through its ability to prevent the translocation of PAP-I. it can also acutely inhibit PAP2 activity [18] and, thereby, modify signal transduction.
Acknowledgements We thank the Alberta Heritage Foundation tor Medical Research, the Canadian Medical Research Council and the Juvenile Diabetes Foundation for financial support.
56 References I Brindley, D.N. 119841 Prog. Lipid Res. 23, 115-133. 2 Brindley, D,N. (1988) in Phosphatidate Phosphohydrolase (BrindIcy. D,N., ed.), Vol. 1, pp. 1-77, CRC Series in Enzyme Biology, CRC Press, Boca Raton. 3 Butterwith, S.C., Martin, A. and Brindley, D.N. 119841 Biochem. J. 222, 487-493. 4 Pittner, R.A., Fears, R. and Brindley, D.N, 119851 Biochem. J. 230, 5~-534. 5 Gomez-Mufioz, A., Hatch, G.M., Martin, A., Jamal, Z., Vance, D.E, and Brindley, D,N. 119921 FEBS Lett., 301, 103-105. 6 Bocckino, S.B., Blackmore, P.F., Wilson, P.B. and Exton, J.H. (1987) J, Biol. Chem. 262, 15309-15315, 7 Loffelholz, K. 119891 Biochem. Pharmacol, 38, 1543-1549, 8 Billah, M,M, and Anthes, J,C. (1990) Biochem, J. 269, 281-291, 9 Billah, M.M,, Eekel, S., Mullmann, T.J., Egan, R.W. and Siegel, M,I, (1989) J, Biol. Chem. 264, 17l)69-17077. 10 Rossi, F., Frzeskow, M,, Dellabia. V,, Calretti, P, and Jandini, G. (19901 Biochem, Biophys, Res, Commtin. 168, 32l)-327. 11 Murayama, T. and Ui, M, (19871J. Biol. Chem, .6,,, "~ '~ 5522-5529. 12 I~uldry, S,A,, Bass, D,A,, Cousart, S.L. and McCall, C.E. (19911 J. Biol, Chem, 266, 4173-4179, 13 Van Corven, E,J,, Groenink, A,, Jalink, K,, Eichholtz, T. and Moolenaar, W,H, (19891 Cell 59, 45-54. 14 Tsai, M,H,, Yu, C,L., Wei, F.S, and Stacey, D.W. (1989) Science 2,k~, 522-526, 15 Verheijden, G.F., Verlaan, i., Schlessinger, J. and Moolenaar, W,H. 119911 Cell Regul, 9, 615-620, 16 Lavie, Y,, Piterman, O. and Liscovitch, M. (1990) FEBS Lett. 277, 7-111. 17 Kaibuchi, K., Takai, Y. and Nishizuka, Y. (1981) J, Biol, Chcm, ~6, 7146-7149, 18 Jamal, Z,, Martin, A., Gomez-Mufioz, A. and Brindley, D,N, (1991) J, Biol. Chem. 266, 2988-2996, 19 Martin, A., Gomez-Mufioz, A., Jamal, Z, and Brindley, D,N, (1~!) Moth, Enzymol, 197, 553-563, 211 lk~wley, M,, Ctx~ling, J,, Burditt, S,L. and Brindley, D.N, (19771 Biochem, J, 165, 447-454, 21 Martin, A., Hopewell, R,, Martin-Sanz, P,, Morgan, J,E, and Brindley, D,N, (19861 Biochim. Biophys, Acta 876, 581-591, 22 Mullmann, T J,, Siegel, M,I,, Egan, R,W, and Billah, M,M, (19911 J. Bit)l, Chem, 266, 2(113-2016, 23 Gomez-Mu6t~ A,, Hales, P,, Brindley, D,N, and Sancho, M,J, (1989) Biochcm, J, ~ .6., 417-423, 24 Ca~ales, C,, Mangiapane, E,H, and Brindley, D,N, (19841 Biochem, J, 219, 911-916,
25 Pittner, R.A., Fears, R. and Brindley, D,N. 119851 Biochem, J. 225, 455-462. 26 Graham, A., Zammit, V.A. and Brindley, D.N. (19881 Biochem. J, 249, 727-733. 27 Martin, A., Hales, P. and Brindley, D.N. (1987) Biochem. J. 245, 347-355. 28 Pollard, A,D. and Brindley, D,N, (19841 Biochem, J. 217, 461-469. 29 Gomez-Mufioz, A., Hales, P. and Brindley, D,N, (1991) Biochem. J. 276, 209-215. 30 Rando, R.R. (19881 FASEB J. 2, 2348-2355. 31 van de Werve, G., Proitto, J. and Jeanrenaud, B. 119851 Biochem. J. 231,511-516. 32 Martin-Sanz, P,, Hopewell, R, and Brindley, D.N. (1984) FEBS Left, 175, 284-288. 33 Rustan, A,C., Nossen, J,O,, Christiansen, E,N, and Drevon, C.A. (1988) J, Lipid Res, 29, 1417-1426. 34 Wong, S,-H. and Marsh, J.B. (19881 Metabolism 37, 1177-1181. 35 Halminsky, M.A,, Marsh, J,B. and Harrison, E.H. (1991) J. Nutr. 121, 1554-1561, 36 Walton, P.A, and Po~smayer, F, (1985) Anal. Bioch~m. 151, 479-486. 37 Sturton, R.G. and Brindley, D,N, (1978) Biochem. J. 171,263-266, 38 Halliwell, B, 119881 Biochem, Pharmacol. 37, 569-571. 39 Colli, S., Marinovich, M., Stragliotto, E,, Gall C.L. and Tremoli, E. 11989) Biochem. Pharmacol. 38, 39[19-3912. 411 Specter, A,A., Santos, E,C., Ashbrook, ,I.D. and Fletcher, J.E. 11970) Ann. N.Y Acad, Sci. 226, 247-258. 41 AIberti, K,G,M.M. and Press, C.M. (1982) in Complications of Diabetes (Keen, H, and Jarret, J., eds.), pp. 231-270, Edward Arnold, London, 42 Mueller, H. and Ayres, S,M. 119781 An). J. Cardiol. 42, 363-371. 43 Svensson, S,, Svedjehohn, R. Ekroth, R,, Milocco, !., Nilsson, F., Sabel, K,G, and WilUam-OIsson, G. (1990) J. Thorac. Cardiovasc. Surg, 99, 11163-11173. 44 ttopewell, R., MartinoSanz, P., Martin, A., Saxton, J. and BrindIcy, D.N. 119851 Biochem. J. 232, 485-491. 45 Shinomura, T., Asaoka, Y., Oka, M., Yoshida, K. and Nishizuka, Y, (1991) Proc, Natl. Acad. Sci. USA 88, 5149-4153. 46 ttannun, Y,A. and Bell, R.M. (1989) Science 243, 50{1-507. 47 Merrill, A.H. and Stevens, V.L. (19891 Biochim. Biophys. Acta 1010, 131-139. 4tq Farcsc, R.V., Konda, T.S., Davis, J.S., Standaert, M.L., Pellet, R J, and Cooper, D,R. 119871 Science 236, 586-589. 49 Ishizuka, T., Cooper, D.R,, Hernandez, H., Buckley, D., Standaert, M. and Farese, R.V. 1199111Diabetes 39, 181-190. 50 Rossi, F,, Grzescowiak, M., Della Bianca, V. and Sbartati, A. ( 1991 ) J. Biol. Chem. 266, 8034-8038.
