Cellular Signalling Vol. 2, No. 6, pp. 563--568, 1990 0898-6568/90 $3.00+.00 Printed in Great Britain © 1990 Pergamon Press plc

LEUKOTRIENE D 4 INDUCED CALCIUM CHANGES IN U937 CELLS MAY UTILIZE MECHANISMS ADDITIONAL TO INOSITOL PHOSPHATE PRODUCTION THAT ARE PERTUSSIS TOXIN

INSENSITIVE BUT ARE BLOCKED BY PHORBOL MYRISTATE ACETATE

KENNETH POLLOCK* and JUDITH CREBA

Bioscience II, ICI Pharmaceuticals, Mereside, AIderley Park, Cheshire SKI0 4TG, U.K.

(Received 23 March 1990; and accepted 8 June 1990)

Abstraet--DMSO differentiated U937 cells responded to 10 -6 M LTD4, LTB 4 and FMLP with an increase in both InsP formation and [Ca2+]i. FMLP caused a greater rise in InsPs than either LTD 4 or LTB4, which were equivalent. LTD4, however, caused a greater increase in [Ca2+]~ than LTB 4 (4-fold) or FMLP. The FMLP [Ca2+]i and InsP responses were abolished by pertussis toxin (100 ng/ml for 4 h) but were unaffected by PMA (10 -7 M for 3 rain). In contrast, the LTD 4 [Ca2+]i and InsP responses were reduced by only 50% by pertussis toxin, whilst PMA reduced the [Ca2+]~ and InsP responses to LTD 4 by 75 and 30%, respectively. These results suggest that mechanisms additional to InsP formation exist for mediating LTD 4 evoked increases in [Ca2+]i.

Key words: LTD4, U937 cells, Ca 2+, inositol phosphates, pertussis toxin, PMA.

I N T R O D U C T I O N

U937 CELLS are a human monoblastic cell line that have cell-surface receptors for platelet acti- vating factor (PAF), ATP, leukotriene (LT) D 4, LTB 4 and, when differentiated, formyl-Met- Leu-Phe (FMLP) [l]. All of these receptors are coupled to an elevation in the cytosolic free Ca 2 + concentration ([Ca2+]i) [2-5], although with varying degrees of efficacy, presumably through receptor stimulation of phosphoinositidase C (PIC) to generate Ins(l,4,5)P 3. Ins(l,4,5)P 3 on its own can release Ca ̀ '+ from internal stores [6], whilst Ins(1,4,5)P 3 in concert with its phos- phorylated derivative Ins(l,3,4,5)P 4 may control Ca 2+ entry in some circumstances [7]. Studies with non-hydrolysable guanine nucleo- tide analogues and fluro-aluminates have indi-

*Author to whom correspondence should be addressed at: Department of Biological Research (JA3), Rh6ne-Poulenc Ltd, Rainham Road South, Dagenham, Essex RM10 7XS, U.K.

Abbreviations: Ins(l,4,5)P3--inositol- 1,4,5-trisphosphate; Ins(l,3,4,5)P4--inositol- 1,3,4,5-tetrakisphosphate.

563

cated that coupling between receptors and PIC is regulated by guanine necleotide binding (G)-proteins [8] that have been pharma- cologically classified into those that are pertus- sis toxin substrates and those that are not. The predominant G-protein that has been impli- cated in PIC activation in leukocytes is Gi2 [9], a pertussis toxin substrate which is known to interact with FMLP receptors on both HL60 and U937 cells [10, l 1]. As such pertussis toxin treatment of these cell types completely inhibits FMLP stimulated inositol phosphate (InsP) production [12, 13]. However, stimulation of InsP production by ATP [12], LTD 4 and PAF [5] in these cells is only partially inhibited by pertussis toxin. This suggests that additional G-proteins and/or additional mechanisms for controlling InsP production/[Ca2+]i are involved in signal transduction through recep- tors for these agonists.

Here the cause and effect relationship between InsP production and increased [Ca2+]~ was investigated in differentiated U937 cells responding to LTD 4, LTB 4 and FMLP.

564 K. POLLOCK and J. CREBA

M A T E R I A L S A N D M E T H O D S

Materials

The following were obtained as indicated. Cell culture media, FLOW Laboratories, Irvine, U.K.; myo[2-3H]inositol (10-20 Ci/mmol), DuPont NEN, U.K.; Fura-2 acetoxymethyl ester, Molecular Probes, OR, U.S.A.; pertussis toxin, Porton Products, Porton Down, U.K.; FMLP and PMA, Sigma, Poole, U.K.; LTD4 and LTB4, Chiral Organics, University of Reading, Reading, U.K. All other reagents were of analytical grade.

U937 cell culture

U937 cells were routinely grown in medium RPMI-1640 plus 10% v/v FCS. For inositol phos- phate studies cells were reseeded at 5 × 105 cells/ml in medium 199 plus 1% v/v FCS and labelled/differ- entiated for 72 h with 2 #Ci/ml [3H]inositol and 1.25% v/v DMSO as described [13]. These cells were washed and resuspended at 5×106cells/ml in HEPES buffered tyrodes (HBT) (145 mM NaCI; 5 mM KC1; 0.5 mM Na:HPO4; 1 mM M g S O 4 ; 5 mM D-glucose; 10mM HEPES; 0.1% BSA, pH7.35) containing 20 mM LiCI. Aliquots (i ml) of cells were incubated at 37°C in the presence of 1 mM CaCI 2 plus agonists; reactions were quenched with 6% per- chloric acid. [3H]InsPs were separated on Dowex (AG l-X8 200-400 mesh) as described [14].

[Ca 2 + ]~-measurements

[Ca2+]i was measured using the fluorescent Ca 2+ indicator dye Fura-2 [15]. For these studies, cells (106cells/ml) were differentiated for 72 h in RPMI-1640/10% FCS containing 1.25% v/v DMSO. Cells were centrifuged (500g for 10min), resus- pended at 5 x 10 6 cells/ml in HBT and incubated with 2.5 X 10 - 6 M Fura-2-AM for 45 min at 37°C. Loaded cells were washed and resuspended at 5 x 10 6 cells/ml in HBT. Fluorescence (339 nm excitation, 500 nm emission) was monitored from stirred cell suspen- sions using a Kontron SFM 25 fluorimeter. Apparent [Ca2+]~ was calculated as previously described [16]. Resting [Ca:+]i in different experi- ments ranged between 85 and 100 nM.

RESULTS

Changes in [Ca2+]i in U937 cells stimulated by 10 6M or 10-6M F M L P are shown in Fig. 1. In the presence of 1 mM external Ca 2+, LTD 4 caused a rapid spike in [CaZ+]i to over 700nM, which was followed by a slower

secondary rise to 350 nM [Fig. l(a)]. Removal of external Ca 2+ reduced the LTD 4 [Ca2+]i response to a simple spike that returned to baseline within 60 s [Fig. l(b)]. Pertussis toxin treatment reduced the maximum [Ca2+]~ response to LTD 4 by 50_+2% (2_+ S.E.M, n = 3 expts) although the second phase was less, if at all, affected [Fig. l(e)]. LTD4 stimulated Ca 2+ release from internal stores was also partially inhibited by pertussis toxin [Fig. l(f)] although this response returned to baseline more slowly than in control cells. The [Ca2+]i profile in response to F M L P [Fig. I(c)] is quite different to that evoked by LTD4. This agonist caused a rapid increase in [Ca2+]~ to over 500 nM that was more sustained than the LTD 4 response and thereafter returned slowly to resting levels. Removal of external Ca 2+ simply reduced the FMLP evoked [Ca2+]i response in terms of both magnitude (350 nM peak) and duration (75 s) [Fig. l(d)], whilst pertussis toxin treatment completely abolished both the influx and mobil- ization components of the F M L P [Ca2+]i response [Fig. l(g) and (h)]. In contrast to 10-6M LTD 4, 10-~M LTB 4 caused only a small increase in [Ca2+]i that amounted to 24+ 1% (n = 3 experiments) of the increase caused by LTD 4 [see Fig. 2(a)] and was com- pletely abolished in pertussis toxin treated cells [Fig. 2(b)]. Also this LTB 4 response did not desensitize the LTD 4 response as shown in Fig. 2.

Agonist stimulation of InsP production in U937 cells, at a time (30 s) consistent with peak [Ca2+]~ responses, is shown in Fig. 3. In response to LTD 4 there was a concentration dependent accumulation in both the [3H]InsP~ _, and the [3H]InsP3 4 fractions which was maxi- mal at 10 6 M agonist.

10 6 M LTB 4 caused a response of similar magnitude to 10 6M LTD4; both LTD 4 and LTB 4 have in fact similar dose-response curves for stimulation of InsP production (data not shown) in differentiated U937 cells. However, 10 6 M FMLP caused a larger accumulation of [3H]InsPs, most notably in the [3H]InsP3 4 frac- tion, than either LTD 4 or LTB 4. Pertussis toxin completely abolished the [3H]InsP accumulation

Effect of pertussis toxin and PMA on U937 cell Ca 2+ 565

%

800- (a) (b) (c) (d)

LTD 4 LTD 4 FMLP FMLP

|00--

500--

100--

(e) (f) (g) (h)

Z._ & LTO 4 LTD 4 FMLP FMLP

m

30s

FIG. 1. The effect of pertussis toxin on agonist stimulated Ca 2+ influx and mobilization in U937 cells. Fura-2-1oaded U937 cells were stimulated with 10 -6 M LTD4 or 10 -6 M FMLP as shown. (a)-(d) are control cells, (e)-(h) are pertussis toxin (100 ng/ml for 4 h) cells. (a), (c), (e) and (g) were in the presence of 1 mM Ca2+; (b), (d), (f) and (h) were in the presence of 1 mM EGTA. [Ca2÷]i calibration is shown on the left.

!:

I " 1 4. 04

U I . J

500- -

200--

100-- • •

LTB 4 LTD 4

I 0 -0 10 -6

500-- P.tox.

200--

100--

LTB 4 LTD 4

10-0 10-0

Fie. 2. Inhibi t ion of LTB4 stimulated [Ca2*]~ by pertussis toxin. 10 -6 M LTB4 and l0 -6 M LTD 4 were added sequentially to Fura-2-1oaded U937 cells where indicated in the presence of 1 mM Ca 2+. The upper panel shows control cells, the lower panel pertussis toxin (100 ng/ml for 4 h) treated cells.

caused by FMLP and LTB4, but only inhibited [3H]InsP accumulation in response to LTD 4 by 50-60% (Fig. 3).

Pharmacological activation of protein kinase C (PKC) by phorbol esters has been shown to interfere with receptor stimulation of InsP production and [Ca2+]~ changes in a variety of cell types including HL60 ceils [12]. Pre-treatment of U937 cells with 10 -7 M PMA reduced the complex [Ca2+] i response to LTD 4 [Fig. 4(a)] to a small transient response [Fig. 4(b)]; in fact, peak [Ca2+]i was 889 + 12 nM in control cells and 231 + 27 nM in PMA treated cells (~+S.E.M., n = 3 ) . In contrast, 10-TM PMA had no measurable effect on the peak [Ca2÷]~ response to FMLP; 555 + 50 nM in controls [Fig. 4(c)] vs 572+10 nM after PMA addition [Fig. 4(d)]. The effect of 10 -7 M PMA on InsP production in response to LTD 4 and FMLP is shown in Fig. 5. Although PMA itself had no effect on [3H]InsP levels the increase in [3H]InSPl 2 and [3H]lnsP34 following LTD 4 addition, was attenuated by 35 and 25%, respectively.

5 6 6 K . POLLOCK a n d J . C R E B A

5

~:~. r m c°ntr°l o E 4 " ' r~ ._~ >,

E.~ 2

.g× o 5 ' 1 c o T ~

to? 0 soline --9 --8 --7 --6 --6 --6

t LTD 4 _ ~ LTB 4 FMLP

trol 8

~ 0

5"g 3 ~ x 2

c o 1

~ 0 gollna --9 --8 - 7 - 6 - 6 - 6

L _ _ LED 4 _ J LTB 4 FMLP

FIG. 3. Inhibition of agonist-stimu]ated lnsP produc- tion by pertussis toxin. Control or pertussis toxin (100 ng/m] for 4 h) treated cells were stimulated for 30 s with agonists at the log-molar concentrations indicated. Results are mean values-I- S.E.M. of tripli- cate determinations.

Stimulation of [3H]InsP~ 2 formation by FMLP was only inhibited by 20% by PMA whilst [3H]InsP3 4 was barely affected.

DISCUSSION

Stimulus response coupling for activation of U937 cells by FMLP appears to be straight- forward; the receptor couples through Gi2-alpha activating PIC and leading to an increase in [Ca2+]~. Similarly, LTB 4, though less effective than FMLP in these cells, probably acts exclusively through PIC. By contrast the data shown here indicate that lnsP formation cannot fully explain the LTD 4 increase in [Ca2+]~. Three pieces of evidence suggest such a

possibility; (i) LTD 4 caused a greater increase in [Ca2+]~ than FMLP with signigicantly less InsP formation; conversely, LTB 4 increased [Ca2+]ito only 25% of the LTD 4 response, yet both these agonists display similar InsP production dose- response curves. Desensitization [5] and antago- nist studies [17] have shown that these agonists act independently through their own receptors, hence agonist/receptor promiscuity cannot explain the different responses. Also the maxi- mum InsP responses to LTD 4 and LTB 4 are additive (data not shown); (ii) pertussis toxin, at a concentration that promotes complete ADP ribosylation of Gi2-alpha [13], reduced the peak [Ca2+]~ response and InsP formation by 50% and left the secondary [Ca2+]~ response essen- tially intact [Fig. l(a) and (e), Fig. 2]; (iii) PMA almost abolished the LTD 4 [Ca2+]~ response, most notably the secondary rise, [Fig. 4(a) and (b)] whilst reducing InsP formation by only 25-35%. It was surprising that PMA did not significantly affect the responses to FMLP (Figs 4 and 5) as has been previously reported [4, 12, 19] but this may depend on the cell type, the concentration of PMA used and, for InsP stimulation, the time at which the response was measured. In any case, what is demonstrated here is selective inhibition of LTD4 changes in [Ca2+]i by PMA.

It therefore appears that in LTD4 stimulated U937 cells the correlation between InsP forma- tion and increased [Ca2+]i is not straight- forward. The pertussis toxin insensitive components of both these responses to LTD4 [Figs l(e) and (f) and 3] indicates the involve- ment of, at least, an additional G-protein and/or more than one receptor.

Furthermore, the selective inhibition of the LTD 4 [Ca2+]i response by PMA indicates that a Ca 2+ influx mechanism may be operating essen- tially independently of InsP production since the latter response, as mentioned above, is minimally affected whilst this putative Ca 2+ influx process is shut down by protein kinase C activation. Such a mechanism may be impor- tant to other LTD 4 responses that rely on sustained Ca 2+ entry; contraction of airway smooth muscle, for example [18]. Alternatively

Effect of pertussis toxin and PMA on U937 cell Ca 2+ 567

t:

r~ t. t~ d L,J

(a ) (b) 800 --

SO0 --

1 0 0 -

LTO 4

10-6

• • P M A LT 04

10-? I0-6

(c) (d) IOO --

T o e -

• • •

F M L P P M A F M L P

30~s 10 - I 10 -7 ! 0 -8

FIG. 4. The effect of PMA treatment on agonist-stimulated [Ca 2 +]i changes. Fura-2-1oaded cells were stimulated with 10 -6 M LTD 4 or 10 -6 FMLP in the presence of I mM Ca 2+. PMA (10 7 M) was added for 3 min prior to the agonist addition in records (b) and (d).

PMA could specifically interfere with the production of Ins(1,3,4,5)P4 should this mole- cule be responsible for mediating LTD4 stimu- lated Ca :+ entry as has been proposed for other systems [7]. Stimulation of 5-phosphatase acti- vity by PMA, for instance, would reduce Ins(1,3,4,5)P4 levels, perhaps sufficiently to suppress Ca 2÷ influx in response to LTD4. FMLP having produced higher levels of InsPs, (Fig. 3), would be more resistant to such an effect. In our experiments any changes in Ins(1,3,4,5)P4 levels would not necessarily be detected by the fractionation procedure we have employed, hence a more rigorous analysis by HPLC of the InsP isomers formed in these experiments would clarify this possibility.

For many agonists, including LTD4, a cause and effect relationship has been proposed for stimulation of InsP formation and increases in

[Ca2+]i simply on the basis that the agonist does both. Clearly, other mechanisms for controlling Ca 2÷ mobilization exist, as has recently been suggested for cholecystokinin analogues [20] and for histamine [21]. Also, whilst Ca 2+ entry relating to InsP formation remains speculative, other second messenger operated Ca 2÷ chan- nels, or true receptor operated Ca :+ channels, could be involved in the LTD 4 response. In a recent paper it has been shown that LTE 4, acting as a partial agonist on the LTD 4 receptor, can raise [Ca2+]i in U937 cells with no detectable increase in Ins(1,4,5)P3 [22].

Collectively these data indicate that some receptors, including those for LTD 4, either utilize the InsP signalling pathway to regulate Ca 2+ in a more complex fashion then previously suspected or they must utilize additional Ca 2+ translocation systems. In addition, the role, if

568 K. POLLOCK and J. CREaA

7

° ° ' ~ " 6 ~ control o .

~ s ~ PMA

g g 3 E ~

.~- ×

_c b ~

LTD 4 FMLP

"~ 10

~ :" 6

.~ ×

sol]ne LTD 4 FMLP

FIG. 5. The effect of PMA on agonJst-stimulated InsP production. Control or PMA (10 -7 M) treated ce]ls were stimulated for 30s with 10-~M LTD. or 10-6M FMLP as indicated. Results are mean values_+ S.E.M. of quadruplicate determinations.

any, o f Gi2 or other G-proteins in the LTD4 receptor transduction systems remains to be investigated.

R E F E R E N C E S

1. Harris P. and Ralph P. (1985) J. Leukocyte Biol. 37, 407-422.

2. Maudsley D. J. and Morris A. G. (1987) Immunology 61, 189-194.

3. Ward S. J. and Westwick J. (1988) Br. J. Pharmac. 93, 769-774.

4. Sakano T., Fujie, A., Hamasaki T., Harada Y., Taniguchi H. and Ueda K. (1988) Cell. lmmunol. 111, 390-397.

5. Winkler J. D., Sarau H. M., Foley J. J., Mong S. and Crooke S. T. (1988) J. Pharmac. exp. Ther. 246, 204-210.

6. Berridge M. J. and Irvine R. F. (1984) Nature 312, 315-321.

7. Berridge M. J. and Irvine R. F. (1989) Nature 341, 197-205.

8. Cockcroft S. and Stuchfield J. (1988) Phil. Trans. R. Soc. Lond. 320, 247-265.

9. Backlund P. S., Aksamit R. R., Unson C. G., Kahn R., Goldsmith P., Spiegal A. M. and Milligan G. (1988) Biochemistry 27, 2040-2046.

10. Gierschick P. and Jakobs K. H. (1987) FEBS Lett. 224, 219-223.

11. Pollock W. K., Creba J., Mitchell F. and Milligan G. (1989) In Biology o f Cellular Transducing Signals, IXth Washington Spring Symposium Abs. 387.

12. Dubyak G. R., Cowen D. S. and Meuller L. M. (1988) J. biol. Chem. 263, 18,108-18,117.

13. Pollock K., Creba J., Mitchell F. and Milligan G. (1990) Biochim. biophys. Acta 1051, 71-77.

14. Berridge M. J., Downes C. P. and Hanley M. R. (1982) Biochem. J. 206, 587-595.

15. Grynkiewicz G., Poenie M. and Tsien R. Y. (1985) J. biol. Chem. 260, 3440-3450.

16. Pollock W. K., Wreggett K. A. and lrvine R. F. (1988) Biochem. J. '256, 371-376.

17. Sarau H., Mong S., Foley J., Wu H. and Crooke S. T. (1987) J. biol. Chem. 262, 4034-4041.

18. Mong S., Hoffman K., Wu H. and Crooke S. T. (1986) Molec. Pharmac. 31, 35-41.

19. Cockcroft S. and Stutchfield J. (1989) Biochem. J. 263, 715-723.

20. Saluja A. K., Powers R. E. and Steer M. L. (1989) Biochem. biophys. Res. Commun. 164, 8-13.

21. Mitsuhashi M., Mitsuhashi T. and Payen D. G. (1989) J. biol. Chem. 264, 18,356-18,362.

22. Saussey D. L., Sarau H. M., Foley J. J., Mong S. and Crooke S. T. (1989) J. biol. Chem. 264, 19,845-19,855.