Signal transduction by the erythropoietin receptor: evidence for the activation of phospholipases A2 and C MEREDITH

MASON-GARCIA,

SANDA

CLEJAN,

JEN-SIE

TOU, AND BARBARA

S. BECKMAN

Departments of Pharmacology, Pathology, and Biochemistry, Tulane University School of Medicine, New Orleans, Louisiana 70112

in the actions of Ep. While the presence of both extracellular and intracellular Ca2+ seems to be absolutely required for erythropoiesis (34), the elevation of intracellular Ca2+ by an ionophore in the absence of Ep is not sufficient to induce proliferation of CFU-E in the absence of Ep (23). Ep itself has been shown to evoke an early rise in intracellular Ca2+ in erythroid progenitor cells (33, 35), but the time course of this elevation (minutes rather than seconds after the addition of Ep) again suggests that it may not be an event that is directly coupled to the activation of the Ep receptor. The lipoxygenase metabolites of arachidonic acid (AA) also have been proposed as mediators of the actions of Ep. Snyder and Desforges (45) and Beckman and Nystuen (2) demonstrated that inhibition of lipoxygenase activity in erythroid progenitors inhibited Epinduced proliferation. In another study, Beckman et al. (1) reported a significant increase in the production of the 12-lipoxygenase metabolite, 12-hydroxyeicosatetraenoic acid (12-HETE), within 15 min of the addition of Ep to intact erythroid progenitor cells (the earliest time point that was studied). Further work has shown that the addition of exogenous 12-hydroperoxyeicosatetraenoic acid (12-HPETE; the bioactive precursor of 12-HETE) to CFU-E in culture will significantly increase the formation of hemoglobinized clonal cell colTHE GLYCOPROTEIN HORMONE, erythropoietin (Ep), onies in the absence of Ep (3). acts via specific cell surface receptors on its target erythChanges in protein phosphorylation also have been roid precursors to induce a program of terminal erythroid differentiation. The binding of Ep to its receptor on reported in response to Ep in isolated cell membranes (11) and in intact cells (47), with both phosphorylation the erythroid progenitor cell, the functionally characterreported to occur within 2.5-10 ized colony-forming unit erythroid (CFU-E), initiates a and dephosphorylation min of the addition of Ep. The kinases or phosphatases program of intracellular events that results in both proliferation and differentiation of these cells. However, the that catalyze these events have not been identified. intracellular signaling pathways that are activated by However, the putative structure of the recently cloned Ep receptor appears to lack the intrinsic tyrosine kinase the binding of Ep to its receptor are not well understood; despite many years of research, a clear paradigm for the domain that is found in the receptors of other growth mechanism of action of Ep has not been established (for factors (15), suggesting that the activation of a serinethreonine kinase may be involved. That this may be the reviews, see Refs. 29 and 47). protein kinase Many potential signal transduction systems have been calcium- and phospholipid-dependent (protein kinase C; PKC) is suggested by three lines of investigated. In regard to the activation of adenylate cyclase, while Ep was shown to increase CAMP levels in evidence. First, inhibitors of PKC have been shown to (24) and Ep-inhuman progenitor-derived erythroblasts at 10 min (33) inhibit Ep-induced colony formation (46). Second, Ep has been and in rabbit bone marrow erythroblasts at 20 min (6, duced c-myc transcription shown to activate a PKC that is associated with the 44), no early changes in adenosine 3’,5’-cyclic monophosphate (CAMP) in response to Ep were detected in nucleus of the erythroid progenitor cell (31). Third, Ep CFU-E derived from rat or murine fetal liver (18, 52) or has been shown to rapidly (2.5 min) induce phosphoryin an Ep-dependent murine cell line (51). In the case of lation of a major PKC substrate, ~80, in intact Epresponsive cells (46). cGMP, Ep has been found to stimulate its production, but a significant increase is not seen for 4 or more hours As putative signal transduction pathways for the Ep (17, 52). Thus the cyclic nucleotides appear to play a receptor, both the formation of lipoxygenase metabomodulatory role in erythropoiesis, that is distal to the lites of AA and the activation of PKC share a common activation of the Ep receptor (EpR). Calcium is another requisite event: the activation of a phospholipase. A intracellular signaling molecule that has been implicated phospholipase A2 (PLA2) can directly release an arachi-

Mason-Garcia, Meredith, Sanda Clejan, Jen-Sie Tou, and Barbara S. Beckman. Signal transduction by the erythropoietin receptor: evidence for the activation of phospholipases AZ and C. Am. J. Physiol. 262 (Cell Physiol. 31): C1197-C1203, 1992.-Erythropoietin (Ep) is the peptide growth factor whose actions on the erythroid progenitor cell induce terminal differentiation. However, the intracellular signaling system that is activated by Ep is poorly understood. Our previous studies have implicated the lipoxygenase metabolites of arachidonic acid in the actions of Ep. In this study, we report an early (30 s to 5 min) increase in levels of two lipoxygenase metabolites: leukotriene B4 (LTB,; 3- to 5-fold) and 12-hydroxyeicosatetraenoic acid (12-HETE; 2-fold). These responses were blocked by an antibody to Ep, by lipoxygenase inhibitors, or by 1,6-di[O(carbamoyl)cyclohexanone oxime] hexane (RHC80267), an inhibitor of diacylglycerol (DAG) lipase. RHC 80267 also significantly inhibited Ep-mediated proliferation. Ep induced the release of [3H]arachidonic acid from cellular phospholipids at 5 min and also increased DAG accumulation at 1 min with a maximum increase of 68.2% over control seen at 30 min. No increase in levels of inositol trisphosphate or phosphatidic acid was observed in response to Ep. Taken together, these data suggest that the signal transduction pathway of the Ep receptor includes the activation of phospholipases A2 and C, resulting in the liberation of DAG and arachidonate and the subsequent formation of LTB, and 12-HETE. erythropoiesis; arachidonic acid; lipoxygenase; diacylglycerol

0363-6143/92

$2.00

Copyright

0 1992 the American

Physiological

Society

Cl197

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (137.154.019.149) on January 10, 2019.

Cl198

SIGNAL

TRANSDUCTION

BY THE

ERYTHROPOIETIN

RECEPTOR

donyl moiety from the sn-2 position of a membrane phospholipid, this being the rate-limiting step in the formation of AA metabolites. The action of a phospholipase C (PLC) results in the formation of a diacylglycerol (DAG); this molecule can serve as the activator of PKC or, via the actions of a DAG lipase, can serve as a further source of AA. The activation of both PLAz and PLC in response to a single agonist has been shown to occur in a number of cell types, including pituitary gonadotrophs (B), macrophages (49), and neutrophils (13). In this study, we have provided evidence linking the activation of PLA2 and PLC to the Ep-mediated proliferation of the erythroid progenitor cell, CFU-E.

in 5 ml a/-MEM. Cells were counted and then resuspended in appropriate density and in appropriate medium for each particular assay. Tritiated thymidine incorporation assay. MFLC prepared as described were resuspended at a concentration of 2 x lo6 cells/ ml in enriched medium (cu-MEM containing 10% fetal bovine serum, 0.1 mM mercaptoethanol). Cells were added to 96-well sterile round-bottom culture plates such that 200,000 cells were present in each well. All agonists and antagonists were prepared in enriched medium at a x20 concentration, and 5 ~1 were added to each well; total incubation volume in each well was brought up to 200 ~1 with enriched medium. When inhibitors were used, they were preincubated with the cells for 15 min at 37°C before addition of Ep or other agonists. After the addition of agonists, plates were incubated for 20 h at 37°C in a humidified atmosphere of 95% air-5% CO, and then pulsed with [3H]thymidine MATERIALS AND METHODS (Amersham, 1 &i/well) and incubated for a further 4 h. The cells were then harvested onto a glass fiber filtermat with a General. Highly purified recombinant human erythropoietin (sp act X60,000 U/mg protein) was obtained from McDonnell Skatron semiautomatic cell harvester. Filters were air dried, and Douglas, St. Louis, MO, and was used exclusively throughout; a individual filter discs were removed and counted in a Packard fresh lyophilized ampoule was reconstituted for each assay. The liquid scintillation counter for 1 min. All experimental condiin each assay. Each assay lipoxygenase inhibitor 3-amino- 1 [ M-(trifluoromethyl)-phenyl] - tions were performed in quadruplicate 2-pyrazoline (BW755c) was a gift of the Wellcome Research contained unstimulated controls and at least two different conLaboratories (Beckenham, Kent, UK). The DAG lipase inhibcentrations of Ep. Radioimmunoassay for LTB4 and 12-HETE levels. MFLC itor 1,6-di[ 0-(carbamoyl)cyclohexanone oxime] hexane (RHC 80267) was a gift of Rorer Central Research (Ft. Washington, prepared as described were suspended in serum-free cu-MEM at PA). Unlabeled AA, A23187, phorbol 12-myristate 13-acetate, a concentration of 5 x lo6 cells/ml. Agonists and antagonists and all chemical reagents were obtained from Sigma (St. Louis, were prepared in cu-MEM in ~100 concentrations; 10 ~1 were added to 1 ml of cell suspension in a 1.5-ml polypropylene MO). Dioctanoylglycerol was obtained from Molecular Probes (Eugene, OR). All reagents were prepared in the solvents recmicrocentrifuge tube. If antagonists were used, the cells were ommended by the manufacturer in the most concentrated stock preincubated with the antagonist for 15 min before the addition solution possible, usually l-100 mM; if ethanol or dimethyl of agonist. Incubations were carried out at room temperature and were stopped by centrifugation in a tabletop high-speed sulfoxide was used, stock concentrations were prepared that centrifuge for 20 s. Supernatant media were collected into clean would result in final dilutions of these solvents, which did not affect the assay system. Stock solutions were stored at the temmicrofuge tubes and stored at -20” C until assayed. Radioimmunoassay kits for LTB* and 12-HETE were obtained from perature recommended by the manufacturer. The polyclonal Advanced Magnetics (Cambridge, MA) and used according to antiserum against recombinant human Ep was prepared in our the manufacturer’s protocol. Assays were counted in a scintillaboratory (30). [5,6,8,9,11,12,14,15-3H]arachidonic acid (76 Ci/ mmol) was obtained from Du Pont-New England Nuclear (Boslation counter, and data reduction was performed on an IBMPC using SECURIA II software from Packard Instruments. ton, MA). Four-parameter logistic curve-fitting algorithms were used Preparation of cells. The cell system we chose for studying the throughout, and standard curve parameters of nonspecific bindactions of Ep is the functionally characterized CFU-E. CFU-E possess the highest number of Ep receptors of the three erything, maximum binding, and slope were monitored from assay to assay. roid progenitor cells that express EpR [burst-forming unit Radioreceptor assay for inositol trisphosphate (IP3). MFLC erythroid (BFU-E), CFU-E, and proerythroblast] and also show the greatest sensitivity to the effects of Ep. CFU-E can be found prepared as described were suspended in serum-free cu-MEM at as an enriched population in murine fetal liver during a specific a concentration of 5 x lo6 cells/ml in 1.5-ml microfuge tubes. The incubation with agonist (x100) was carried out at room developmental period; at days 12-13 of gestation, it is estimated temperature and stopped by adding 200 ~1 of ice-cold 100% that 7080% of fetal liver cells are CFU-E (25). Unpurified murine fetal liver cells (MFLC) thus have a percent purity of trichloroacetic acid. Tubes were immediately placed in an ice bath and incubated on ice for 15 min, after which time they were CFU-E equivalent to that achievable from other sources only centrifuged and the supernatant was collected for IP3 assay. after intensive purification (40). Trichloroacetic acid was removed from the supernatant by Mice used for these experiments were acquired, cared for, and used in accordance with the guidelines approved by the Council extraction with freonltri-n-octylamine (3: 1) and collection of the aqueous top layer for assay. The IP3 radioreceptor assay (Du of the American Physiological Society. Adult male and female Pont-NEN) was performed on these extracts following the manCD-l strain mice (Charles River, Boston, MA) underwent ufacturer’s protocol. Counting and data reduction were pertimed matings. At days 12-13 after mating, the female mice were killed while under ether anesthesia; the peritoneal cavity formed as for the radioimmunoassays. Labeling of cellular phospholipids with tritiated AA. Fetal liver was opened under sterile conditions, and the uterus containing the fetuses was removed and transferred to a sterile petri dish cells were suspended in Hanks’ balanced salt solution at a concontaining a-modified Eagle’s minimum essential medium with centration of 26 x lo6 cells/ml. To 10 ml of the cell suspension acid, 10 &i in 250 ~1 0.9% NaCl glutamine (cu-MEM; GIBCO). The uterus was washed in CY- was added [3H]arachidonic MEM, the amniotic sacs were opened, and the fetuses were containing 0.25 mg fatty acid-free bovine serum albumin (BSA). removed. The livers were gently teased free of the abdominal The cells were then incubated at 37°C for 30 min with constant shaking; at this ratio of cells to labeled AA, individual classes of cavity using fine forceps and placed in a sterile 12-ml roundbottomed plastic centrifuge tube. The cells were gently disagglycerolipids showed similar radioactivities after lo-, 20-, or gregated by sequential passage through 18-, 21-, and 23-gauge 30-min incubations. Free [3H]AA was removed at the end of this period by centrifuging the mixture at 1,000 g for 10 min at hypodermic needles, washed twice in cu-MEM, and resuspended Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (137.154.019.149) on January 10, 2019.

SIGNAL

TRANSDUCTION

BY THE

room temperature, and the pellet of labeled cells was washed sequentially with Dulbecco’s phosphate-buffered saline (PBS) containing 0.2% BSA and with PBS alone. The labeled cells were resuspended in 9 ml Hanks’ buffered salt solution and divided into two aliquots. After incubation for 20 min at 37”C, one aliquot was pulsed with Ep (5 U in 500 ~1 PBS) and the other aliquot with PBS alone (control). At intervals of 30 s to 10 min, 1 ml of each cell suspension was withdrawn and immediately delivered to 45ml stoppered glass tubes containing 5 ml methanol and 2.5 ml chloroform. Total lipids were extracted according to the method of Bligh and Dyer (5) and dissolved in chloroform:methanol (2: 1, vol/vol) containing 0.01% butylated hydroxytoluene. Individual phospholipids were resolved by twodimensional thin-layer chromatography (TLC) on silica gel H plates (50). Due to the low levels of phosphatidic acid (PA) in MFLC, a 0.5-pg standard PA phosphorus was applied together with the lipid extract to the TLC plate for localization of PA. Individual phospholipid spots were visualized by exposing the plate to iodine vapor. Phospholipid phosphorus and radioactivity were measured as previously described (50). Because PA mass could not be accurately measured by phosphate assay, the radioactivity of the individual phospholipids was expressed as counts per minute (cpm) per 7 x lo6 cells. The individual lipid spots were scraped from the plate and counted in a Packard liquid scintillation counter for 10 min. Determination of DAG levels. MFLC (5 x 106/ml) were incubated with Ep (0.2 U/ml) for 1, 2, 10,30,40, or 60 min. Cellular lipids were extracted by adding sequentially with vortexing 1 ml chloroform:methanol (1:2, vol/vol), 0.66 ml chloroform, and 0.6 ml deionized water. After a minimum of 5 min, the samples were centrifuged at 2,000 g for 5 min, and the aqueous upper layer was discarded. To the bottom layer was added 1 ml chloroform, and the solution was then filtered through glass wool in a Pasteur pipette. The collected eluate was evaporated under N, at room temperature, redissolved in 0.5 ml chloroform, and applied to a 0.5-ml silicic acid column. Neutral lipid was eluted with 4 ml chloroform, evaporated to dryness, and stored at -70°C under N, (12). Neutral lipid extracts were analyzed using a modification of the method of Hamilton and Comaii (21) as previously described (12). Chromatography was performed using a Waters (Milford, MA) M6000 delivery pump, a U6K injector, a microporosil column (30 cm x 4.0 mm), and a R-401 differential refractometer at half attenuation. The mobile phase was hexane:isopropanol:acetic acid (100: l:O.Ol, vol/vol/vol). Flow rate was 2 ml/min. A volume of 0.15 ml with lo-25 mg DAG was typically injected. DAG eluted as a bimodal peak at 9.6-12.5 min after injection, following the free cholesterol peak. The standard used was 1,2-DAG. The detector output was channeled into a convertor (analog/digital) and stored in a computer. Integration of peaks was performed with a CPLOT system. The peak integral was calibrated against 1,2-DAG standard and calculated in picomoles of DAG per million cells. The method was reproducible with a typical standard deviation of 2 mV at a total of 18-35 mV; recovery was 96-103%. StatisticaL analyses. Statistical analyses were done using Student’s t test for unpaired data or Dunnett’s t test (16) for comparing multiple experimental conditions to a single control. RESULTS

Ep induces a rapid increase in levels of LTB4 and 12HETE in MFLC. The two lipoxygenase metabolites most

strongly implicated in the proliferative response are LTB4 and l%HPETE. If these metabolites mediate the actions of Ep, then treatment of the cells with Ep should induce an increase in the levels of these metabolites. Murine fetal liver cells (5 x lo6 cells/ml) were incubated with Ep for varying periods of time, and the levels of

ERYTHROPOIETIN

Cl199

RECEPTOR

LTB4 and 12HETE

in the supernatant medium were determined with specific radioimmunoassays. (12HETE, which is rapidly converted from 12-HPETE, was assayed as a measure of 12-HPETE production.) Ep induced a rapid increase in both LTB4 and 12-HETE levels, which could be seen within 30 s (Fig. 1A). A significant elevation in 12-HETE levels was seen at 10 min and sustained for up to 60 min (Fig. lB), while LTB4 levels were significantly elevated at 2 min but had returned to near baseline values by 60 min, suggesting that the activity of the 5 and 12-lipoxygenase enzymes may be differentially regulated. It should also be noted that levels of 12-HETE produced by these cells were almost three orders of magnitude higher than the LTB4 levels, indicating differences in the relative amounts of the 5 and 12-lipoxygenases as well. The Ep-induced increases in these metabolites could be blocked by preincubation of the cells with the lipoxygenase inhibitor BW 755~ (Fig. 2). The addition of the immunological antagonist anti-Ep to the incubation medium also inhibited the Ep-induced increases in 12HETE and in LTB4 by 80% and 75%, respectively. These

A 120

7c

z 110 u 5 -g

loo

E G ho s !i

60 7cI 2

0

3

4

5

6

6

9

10

11

12

13

14

15

time (win)

B 100

--e-

1

r

LTM

50

90 57 z 0

80

c

70

:

60

0 .-

10

0

5

IO

IS

20

25

30

35

40

45

SO

SS

60

time (min)

Fig. 1. Erythropoietin (Ep) induces a rapid, time-dependent increase in leukotriene B, (LTBJ and 12-hydroxyeicosatetraenoic acid (12HETE) production in murine fetal liver cells. A: in a preliminary experiment, cells (5 x lO”/ml) were incubated with 0.2 U Ep/ml, and supernatant medium was collected for assay at 0, 0.5, 1, 2, 5, and 15 min. A rise in LTB, and 12-HETE was seen by 0.5 min. B: in other experiments, cells (5 x 106/ml) were incubated with Ep (0.2 U/ml), and supernatant was collected for assay at 0, 2, 5, 10, 15, 30, and 60 min. A significant (P c 0.05 vs. control) rise in LTB, was seen at 2 min and sustained through 30 min, decreasing to baseline levels by end of 60-min test period. 12-HETE levels were increased at 2 min, but increase did not become significant (P < 0.05 vs. control) until 10 min; increase continued through 60-min test period. Each point represents mean ~fr SE from 3-8 experiments.

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (137.154.019.149) on January 10, 2019.

Cl200

SIGNAL

TRANSDUCTION

BY THE

A 60 p 3 c 40 0 .I= $# 3o %I z J 10 0 15

15

time (min)

time (min)

Fig. 2. Inhibitor of arachidonic acid metabolism, BW 755c, inhibits Ep-induced increases in LTB, and 12-HETE levels. Cells (5 x lO”/ml) were incubated with or without BW 755~ for 15 min, followed by the addition of Ep (0.2 U/ml). Supernatant medium was collected for assay at 0 min and at 15 min after addition of Ep. A: BW 755~ significantly (*P c 0.01 vs. no inhibitor) inhibited Ep-induced increase in LTB4 levels. Bars, means t SE from 3 experiments. B: BW755c significantly (*P < 0.01 vs. no inhibitor) inhibited Ep-induced increase in 12-HETE levels. Bars, means t SE from 3 experiments. BW 755~ also inhibited basal (time = 0) levels of LTB4 and 12-HETE (data not shown) but was not toxic at this concentration (2).

data support the hypothesis that these metabolites are part of the signal transduction pathway activated by Ep. Ep induces the release of [3H]arachidonic acid from phospholipids in IMFLC. The first regulatory step in the

production of AA metabolites is the liberation of AA from phospholipids. To determine whether Ep enhances AA release from MFLC, phospholipids were labeled by preincubating the cells with [3H]AA. The labeled cells were then incubated further in the absence or presence of Ep. Table 1 illustrates the distribution of [3H]AA radioactivity among different phospholipids in the absence and presence of Ep after a 5-min incubation. With the exception of PA, each phospholipid showed a similar loss of [3H]AA radioactivity. Changes in mass of phospholipids were undetectable by phosphate assay. The average mass of each phospholipid expressed as a percentage of total lipid phosphorus recovered from the TLC plate was 43% phosphatidylcholine (PC), 26.4% phosphatidylethanolamine (PE), 11.4% phosphatidylinositol (PI), 10.2% phosTable 1. Changes in [3H]arachidonic acid distribution in murine fetal liver cell phospholipids in response to erythropoietin Phospholipid

Phosphatidic acid Phosphatidylserine Phosphatidylinositol Phosphatidylcholine Phosphatidylethanolamine

Control

431 6,523 31,361 91,684 48,975

+Ep

(0.1 U)

392 5,420 25,237 72,587 40,662

Stimulated/ Unstimulated

104t5.3 81t4.2 85k2.6 83k5.7 82t4.3

In control column, [3H]arachidonic acid ([3H]AA) is measured in counts/min (cpm) in lipid from 7 x lo6 control (unstimulated) murine fetal liver cells. In +EP column, [3H]AA is measured in cpm in lipid from 7 x lo6 murine fetal liver cells after 5-min stimulation with Ep (0.1 U/ml). Data shown are from 1 representative experiment of a total of 3. In stimulated/unstimulated column, [3H]AA is measured in cpm in Ep-stimulated cells expressed as a percent of [3H]AA cpm in unstimulated cells; mean values & SE; n = 3.

ERYTHROPOIETIN

RECEPTOR

phatidylserine (PS), and 8.94% sphingomyelin. Although PC and PE contain more [3H]AA than PI, PI exhibited a higher specific radioactivity (cpm/pg phospholipid phosphorus) than PC and PE. PS had the lowest specific radioactivity. These findings suggest that Ep enhances the expression of PLAZ activity in MFLC and that PC, PE, and PI provide AA for the synthesis of LTB4 and 12-HETE by MFLC in response to EP. Further, Ep did not induce any change in PA radioactivity, suggesting that phospholipase D activation is not a major signal transduction pathway of Ep in these cells. Ep induces an increase in DAG in MFLC. While the activation of a PLAz is characterized by the release of AA from membrane phospholipids, the activation of a PLC is followed by the cleavage of DAG from phospholipids. To determine if MFLC show increased levels of DAG in response to Ep, cells were incubated in the presence of Ep for varying amounts of time, and the changes in amount of cellular DAG were determined using high-performance liquid chromatographic analysis. When data from several experiments were analyzed as percent increase over control levels of DAG (Table 2), a 10.4% rise in DAG was seen following 1 min of incubation with Ep, with the apparent maximal increase (68.2%) occurring at 30 min after the addition of Ep. These data strongly support the hypothesis that the activation of a PLC is an early event in the intracellular signaling pathway of Ep. Ep does not induce a rapid increase in IP3 in MFLC.

When PI bisphosphate is the substrate for the actions of PLC, the release of DAG is accompanied by the rapid (2-10 s) generation of a second intracellular messenger, IP3. The action of IP3 in releasing Ca2+ from intracellular stores results in increased levels of Ca2+ within the cell, facilitating the activation of the calcium- and phospholipid-dependent protein kinase, PKC. To determine if this PLC-mediated pathway is activated by Ep, MFLC were incubated with Ep and the reaction was terminated by the addition of trichloroacetic acid at 0,2,5, 10, 20, or 30 s or 1,2, or 4 min. After extraction, the amount of IP3 was determined using a sensitive and specific radioreceptor assay. No significant increase in IP3 levels above baseline (time = 0) was seen at any of these time points (Table 3), although significant decreases were seen. These data suggest that the PLC that is activated by the binding of Ep to Ep receptor in MFLC is not the PI-specific PLC. Ep may act via a pathway that requires the sequential Table 2. Effect of erythropoietin on diacylglycerol accumulation in murine fetal liver cells Time,

0 1 2 10 30 40 60

min

Diacylglycerol,

pmol/106

16.9t0.7 18.6t0.9 19.3t0.9 21.9tl.l* 28.4+1.1t 25.5+1.2? 18.9

Murine fetal liver cells (5 x for indicated times. Values are in duplicate except for 60-min single experiment performed tP < 0.001 vs. 0 time.

cells

Percent

Increase

10.4t3.2 14.1t3.3 29.5t4.1 68.2t9.8 :Fg-7.

107) were incubated with Ep (0.2 U/ml) means t SE of 3 experiments performed time point, which represents results of a in duplicate. *P < 0.05 vs. 0 time;

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (137.154.019.149) on January 10, 2019.

SIGNAL

TRANSDUCTION

BY THE

Inositol

Trisphosphate,

pmol/106

0.327t0.038 0.102kO.O43* 5s 0.205kO.085 10 s 0.247kO.110 20 s 0.365t0.112 30 s 0.150t0.033* 1 min 0.155~0.054* 2 min 0.423t0.069 4 min 0.196t0.012 Murine fetal liver cells (5 x 106/ml) were incubated with Ep (0.2 U/ml), and cells were collected for assay at indicated time points. Results shown are means t SE from 3 experiments. *P c 0.05 vs. basal levels using Student’s t test.

Table 4. Effect of inhibitor of diacylglycerol lipase (RHC 80267) on erythropoietin-induced thymidine incorporation in murine fetal liver cells

Control, Ep only RHC 80267

RHC 80267, sults are means

[3H] thymidine Incorporation, %control

Concentration

0.2 1 5 10

U/ml PM PM PM

lOOk5 73.3&2.1* 69.8&5.4* 69.5t8.8*

1,6-di[O-(carbamoyl)cyclohexanone & SE from 4-7 experiments.

oximelhexane.

Re-

*P < 0.05 vs. control.

Table 5. Effect of an inhibitor of diacylglycerol lipase (RHC 80267) on erythropoietin-induced elevations in LTB4 and 12-HETE in murine fetal liver cells Time

LTB4,

pg/106

cells

12-HETE,

Cl201

cells

OS 2s

Inhibitor

RECEPTOR

These data support the hypothesis that activation of the PLC-DAG lipase pathway by Ep may be an important precursor event for the subsequent activation of the lipoxygenase pathway in MFLC.

Table 3. Time course of erythropoietin-induced changes in inositol trisphosphate levels in murine fetal liver cells Time

ERYTHROPOIETIN

ng/106

cells

0 min 50.2k8.07 13.8k2.1 15 min 83.0k31.4 18.8k2.9 -RHC 80267 14.222.4 +RHC 80267 51.7t6.4 Values are means t SE (n = 2-5); 15-min time points, 15 min after Ep was added. LTB4, leukotriene B,; 12-HETE, 12-hydroxyeicosatetraenoic acid.

actions of PLC and DAG lipase. While the release of DAG

from phospholipid by PLC may result in the activation of PKC, DAG may itself be cleaved by DAG lipase and thus serve as a source of AA. To determine if DAG lipase activity is required for the Ep-mediated proliferative response in MFLC, cells were pretreated with an inhibitor of DAG lipase, RHC 80267 (48). RHC 80267 was found to induce a significant and dose-dependent inhibition of Ep-induced [3H]thymidine incorporation at 1, 5, and 10 PM (Table 4), but it did not significantly affect proliferation in control cells ( [ 3H] thymidine incorporation at 10 PM RHC 80267 = 91.9% t 5.15; n = 3). To determine if AA derived from the sequential actions of a PLC and DAG lipase can act as substrate for the 5and 12-lipoxygenases, the ability of RHC 80267 to inhibit Ep-induced LTB4 and 12-HETE production in MFLC was tested. At a concentration of 5 PM, this compound was found to inhibit Ep-induced elevations of both LTB4 and 12-HETE at 15 min (Table 5), although this inhibition was not statistically significant.

DISCUSSION

The information gained from these studies makes it clear that lipid molecules produced subsequent to the activation of phospholipases represent an important group of intracellular signaling molecules in erythropoiesis: phospholipids provide AA either directly or via a DAG intermediate; AA is further metabolized to LTB4 or 12-HPETE; and DAG may provide AA and/or serve as an activator of PKC. The actions and interactions of these signaling pathways may thus serve to regulate erythropoiesis. We propose that the binding of Ep to its cell surface receptor activates both PLA2 and a PLC; the activation of these enzymes then causes the release from membrane phospholipids of AA or DAG. DAG can be acted on by DAG lipase to release arachidonate or can activate PKC. The formation of the AA metabolites LTB* and 12HPETE, along with the activation of PKC, can increase intracellular calcium and intracellular pH, possibly by activating ion channels and transporters in the cell membrane. This increase in both intracellular calcium and pH is a requisite event for the initiation of DNA synthesis (43) in proliferation. The phospholipases AZ are a family of enzymes that exist in a variety of tissues; their activity can be positively modulated by different mechanisms, including increases in intracellular calcium levels, PKC-mediated inactivation of the endogenous PLAz inhibitor lipocortin, or the actions of guanine nucleotide-binding proteins (for a review, see Ref. 9). The possibility that PLA2 is directly activated by the interaction between the Ep receptor and a guanine nucleotide binding protein (G protein) is the most intriguing. While the predicted conformation of the Ep receptor differs from that of the classical G-pro tein-coupled receptor (1 vs. 7 predicted transmembrane domains), another nonclassical growth factor receptor, that of plateletderived growth factor, has been reported to be coupled to an initial early release of AA, which precedes an increase in calcium or DAG (19). Furthermore, Miller et al. (33) have recently demonstrated that, in human bone marrowderived erythroblasts, the Ep-mediated rise in intracellular Ca2+ can be blocked by pretreatmen t of the cells with pertussis toxin. Some preliminary experiments from our laboratory (data not shown) indicate that pertussis toxin but not cholera toxin may inhibit Ep -induced LTB* production (12-HETE levels were not measured). In another experiment, pertussis toxin also completely abolished Ep-induced [ 3H] thymidine incorporation. Along with earlier studies that show an enhancement of Ep-induced colony formation by nonhydrolyzable analogues of GTP (GTPTS) (B. Beckman, unpublished observations), these data suggest that activation of PLA2 by Ep in erythroid progenitor cells may be via a receptor-G protein interaction. Like PLA2, PLC also represents a family of enzymes;

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (137.154.019.149) on January 10, 2019.

Cl202

SIGNAL

TRANSDUCTION

BY THE

unlike the phospholipases AZ, however, the phospholipases C show substrate specificity among their subtypes. The best known PLC is the PI-specific PLC, which causes the release of two intracellular signaling molecules (DAG and IP3) from PI. As the MFLC do not show the early rise in IP3, which is characteristic of the activation of this enzyme (Table 3), it is more likely that the activation of a PC- or PE-specific PLC is involved. For example, the activation of a PE-specific PLC has been reported to be an early event in the mitogenic response to prolactin in NB2 lymphoma cells (20). The existence of a PC-specific PLC has been reported in a number of cell types (for a review, see Ref. 4), including cells of the hematopoietic system. Hormonally activated PC-specific phospholipases C have also been reported; such a PLC is activated by bombesin in 3T3 cells (143 and by carbachol and bradykinin in PC12 cells (22). These enzymes also seem to lack a requirement for increased intracellular calcium for activation and have been suggested as candidates for activation by receptor-coupled, pertussis toxinsensitive G proteins. The lack of a change in PA in response to Ep suggests that the source of DAG is not via the actions of a phospholipase D or via de novo synthesis of DAG, both of which pathways use PA as an intermediate for the formation of DAG (28). The inhibitory effects of the DAG lipase inhibitor (RHC 80267) in MFLC, along with the presence of phospholipids that are known in other systems to contain a high percentage of diacyl linkages, suggest that the release of AA in our cells is mediated in part by the activation of a PLC/DAG lipase pathway. This pathway has been described in a number of cell types, including thrombin-stimulated platelets (41) and acetylcholinestimulated chromaffin cells (42). DAG has also been demonstrated to be the specific source of precursor AA for the production of 12-HPETE in adrenal glomerulosa cells (39). In the MFLC, inhibition of DAG lipase inhibits production of this AA metabolite as well as the Slipoxygenase metabolite LTB,+ This finding suggests that the DAG lipase pathway is a possible source of AA in erythroid progenitor cells, in addition to the direct liberation of AA from phospholipids by PLA2. The regulation of Ep-induced proliferation in erythroid precursor cells is thus seen to be mediated or modulated by several signaling pathways acting in concert: PLA2 and PLC, AA and its lipoxygenase metabolites, and diacylglycerol-activated PKC. The combined actions of these molecules result in the initiation of DNA synthesis and increased proliferation. Taken together, the findings reported here support the hypothesis that the actions of Ep are mediated by the Ep-induced generation of intracellular lipid molecules. This work was supported by American Cancer Society Grant CH-345 (to B. S. Beckman), private funds (S. Clejan), and a grant from the American Heart Association of Louisiana (J.-S. Tou). M. Mason-Garcia was a National Science Foundation Predoctoral Fellow (RCD-8758119). Address for reprint requests: B. S. Beckman, Dept. of Pharmacology, Tulane Univ. School of Medicine, 1430 Tulane Ave., New Orleans, LA 7n113

ERYTHROPOIETIN Received

17 September

RECEPTOR 1991; accepted

in final

form

6 December

1991.

REFERENCES 1. Beckman, B. S., M. Mason-Garcia, L. Nystuen, L. King, and J. W. Fisher. The action of erythropoietin is mediated by lipoxygenase metabolites in murine fetal liver cells. Biochem. Biophys. Res. Commun. 147: 392-398, 1987. 2. Beckman, B., and L. Nystuen. Comparative effects of inhibitors of arachidonic acid metabolism on erythropoiesis. Prostaglandins Leukotrienes Essent. Fatty Acids 31: 23-26, 1988. 3. Beckman, B. S., and I. Seferynska. Comparative effects of inhibitors of arachidonic acid metabolism on erythropoiesis. Exp. HematoZ. 17: 309-312, 1989. 4. Billah, M. M., and J. C. Anthes. The regulation and cellular functions of phosphatidylcholine hydrolysis. Biochem. J. 269: 281-291, 1990. 5. Bligh, E. G., and W. J. Dyer. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37: 911-917, 1959. 6. Bonanou-Tzedaki, S. A., M. S. Setchenska, and H. R. V. Arnstein. Stimulation of the adenylate cyclase activity of rabbit bone marrow immature erythroblasts by erythropoietin and haemin. Eur. J. Biochem. 155: 365-370, 1986. 7. Chabbot, H., and M. C. Cabot. Phorbol diesters inhibit enzymatic hydrolysis of diacylglycerols in vitro. Proc. N&Z. Acad. Sci. USA 83: 3126-3130, 1986. 8. Chang, J. P., R. 0. Morgan, and K. J. Catt. Dependence of secretory responses to gonadotropin-releasing hormone on diacylglycerol metabolism. J. Biol. Chem. 263: 18614-18620, 1988. 9. Chang, J., J. H. Musser, and H. McGregor. Phospholipase AZ: function and pharmacological regulation. Biochem. PharmacoZ. 36: 2429-2436, 1987. 10. Chilton, F. H. Potential phospholipid source(s) of arachidonate used for the synthesis of leukotrienes by the human neutrophil. Biochem. J. 258: 327-333, 1989. 11. Choi, H.-S., S. C. Bailey, K. A. Donahue, G. J. Vanasse, and A. J. Sytkowski. Purification and characterization of the erythropoietin-sensitive membrane phosphoprotein, pp43. J. Biol. Chem. 265: 4143-4148,199O. 12. Clejan, S., T. A. Krulwich, K. R. Mondrus, and D. Seto-Young. Membrane lipid composition of obligately and facultatively alkalophilic strains of Bacillus spp. J. BacterioZ. 168: 334-340, 1986. 13. Cockcroft, S., and J. Stutchfield. The receptors for ATP and fMetLeuPhe are independently coupled to phospholipases C and A2 via G protein(s). Biochem. J. 263: 715-723, 1989. 14. Cook, S. J., S. Palmer, R. Plevin, and M. J. 0. Wakelam. Mass measurement of inositol 1,4,5-triphosphate and sn-1,2-diacylglycerol in bombesin-stimulated Swiss 3T3 mouse fibroblasts. Biochem. J. 265: 617-620, 1990. 15. D’Andrea, A. D., H. F. Lodish, and G. G. Wong. Expression cloning of the murine erythropoietin receptor. Cell 57: 277-285, 1989. 16. Dunnett, C. W. New tables for multiple comparisons with a control. Biometrics 20: 482-491, 1964. 17. Graber, S. E., J. D. Bomboy, W. D. Salmon, and S. B. Krantz. Effect of erythropoietin preparations on cyclic AMP and cyclic GMP levels in rat fetal liver cell cultures. J. Lab. Clin. Med. 90:162-170,1977. 18. Graber, S. E., M. Carrillo, and S. B. Krantz. Lack of effect of erythropoietin on cyclic adenosine-3’,5’-monophosphate levels in rat fetal liver cells. J. Lab. CZin. Med. 83: 288-295, 1974. J. H., J. V. Bonventre, and R. A. Nemenoff. Iden19. Gronich, tification and characterization of a hormonally regulated form of phospholipase AZ in rat renal mesangial cells. J. Biol. Chem. 263: 16645-16651, 1988. M. M., and M. E. Costlow. Phosphatidylethanolamine 20. Hafez, turnover is an early event in the response of NB2 lymphoma cells to prolactin. Exp. Cell Res. 184: 37-43, 1989. 21. Hamilton, J. G., and K. Comaii. Rapid separation of neutral lipids, free fatty acids and polar lipids using prepacked silica SepPak columns. Lipids 23: 1146-1149, 1988. 22. Horwitz, J. Carbachol and bradykinin increase the production of diacylglycerol from sources other than inositol-containing phospholipids in PC12 cells. J. Neurochem. 54: 983-991, 1990.

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (137.154.019.149) on January 10, 2019.

SIGNAL

TRANSDUCTION

BY THE

23. Imagawa, S., B. R. Smith, R. Palmer-Cracker, and H. F. Bunn. The effect of recombinant erythropoietin on intracellular free calcium in erythropoietin-responsive cells. Blood 73: 14521457, 1989. 24. Jenis, D. M., C. S. Johnson, and P. Furmanski. Effects of inhibitors and activators of protein kinase C on late erythroid progenitor (CFU-e) colony formation in vitro. Int. J. CeZZ Cloning 7: 190-202, 1989. 25. Johnson, G. R., and D. C. Barker. Erythroid progenitor cells and stimulating factors during murine embryonic and fetal development. EXJJ. HematoZ. 13: 200-208, 1985. 26. Landschultz, K. T., A. N. Noyes, 0. Rogers, and S. H. Boyer. Erythropoietin receptors on murine erythroid colonyforming units: natural history. BZood 73: 1476-1486, 1989. 27. Lee, T.-S., .K. A. Saltsman, H. Ohashi, and G. L. King. Activation of protein kinase C by elevation of glucose concentration: proposal for a mechanism in the development of diabetic vascular complications. Proc. NutZ. Acad. Sci. USA 86: 5141-5145, 1989. 28. Martin, T. W., R. B. Wysolmerski, and D. Lagunoff. Phosphatidylcholine metabolism in endothelial cells: evidence for phospholipase A and a novel Ca2+ -independent phospholipase C. Biochim. Biophys. Acta 917: 296-307, 1987. 29. Mason-Garcia, M., and B. S. Beckman. Signal transduction in erythropoiesis. FASEB J. 5: 2958-2964, 1991. 30. Mason-Garcia, M., B. S. Beckman, J. Brookins, J. Powell, W. Lanham, S. Blaisdell, L. Keay, and J. W. Fisher. Development of a new radioimmunoassay for erythropoietin using recombinant erythropoietin. Kidney Int. 38: 969-975, 1990. 31. Mason-Garcia, M., C. L. Weill, and B. S. Beckman. Rapid activation by erythropoietin of protein kinase C in the nuclei of erythroid progenitor cells. Biochem. Biophys. Res. Commun. 168: 490-497, 1990. 32. McIntyre, T. M., S. L. Reinhold, S. M. Prescott, and G. A. Zimmerman. Protein kinase C activity appears to be required for the synthesis of platelet-activating factor and leukotriene B, by human neutrophils. J. BioZ. Chem. 262: 15370-15376, 1987. 33. Miller, B. A., K. Foster, J. D. Robishaw, C. F. Whitfield, L. Bell, and J. U. Cheung. Role of pertussis toxin-sensitive guanosine triphosphate-binding proteins in the response of erythroblasts to erythropoietin. BZood 77: 486-492, 1991. 34. Misiti, J., and J. L. Spivak. Erythropoiesis in vitro: role of calcium. J. CZin. Inuest. 64: 1573-1579, 1979. 35. Mladenovic, J., and N. E. Kay, Jr. Erythropoietin induces rapid increases in intracellular free calcium in human bone marrow cells. J. Lab. CZin. Med. 112: 23-27, 1988. 36. Moroney, M. A., R. A. Forder, F. Carey, and J. R. S. Hoult. Differential regulation of 5-lipoxygenase and cycle-oxygenase pathways of arachidonate metabolism in rat peritoneal leukocytes. Br. J. Pharmacol. 101: 128-132, 1990. 37. Nakadate, T. The mechanism of skin tumor promotion caused by phorbol esters: possible involvement of arachidonic acid

ERYTHROPOIETIN

38.

39.

40.

41.

42.

43. 44.

45.

46.

47. 48.

49.

50. 51.

52.

RECEPTOR

Cl203

cascade/lipoxygenase, protein kinase C and calcium/calmodulin systems. Jpn. J. PharmacoZ. 49: l-9, 1989. Naor, Z. Further characterization of protein kinase C subspecies in the hypothalamo-pituitary axis: differential activation by phorbol esters. Endocrinology 126: 1521-1526, 1990. Natarajan, R., N. Stern, and J. Nadler. Diacylglycerol provides arachidonic acid for lipoxygenase products that mediate angiotensin II-induced aldosterone synthesis. Biochem. Biophys. Res. Commun. 156: 717-724, 1988. Nijhof, W., and P. K. Wierenga. Isolation and characterization of the erythroid progenitor cell: CFUe. J. CeZZ BioZ. 96: 386392, 1983. Prescott, S. M., and P. W. Majerus. Characterization of l,2diacylglycerol hydrolysis in human platelets: demonstration of an arachidonoyl-monoacylglycerol intermediate. J. BioZ. Chem. 258: 764-769, 1983. Rindlisbacher, B., M. A. Sidler, L. E. Galatioto, and P. Zahler. Arachidonic acid liberated by diacylglycerol lipase is essential for the release mechanism in chromaffin cells from bovine adrenal medulla. J. Neurochem. 54: 1247-1252, 1990. Rozengurt, E. Early signals in the mitogenic response. Science Wash. DC 234: 161-166, 1986. Setchenska, M. S., S. A. Bonanou-Tzedaki, and H. R. V. Arnstein. Independent activation of adenylate cyclase by erythropoietin and isoprenaline. MOL. Cell. Endocrinol. 56: 199-204, 1988. Snyder, D. S., and J. F. Desforges. Lipoxygenase metabolites of arachidonic acid modulate hematopoiesis. BZood 67: 1675-1679, 1986. Spangler, R., S. C. Bailey, and A. J. Sytkowski. Erythropoietin increases c-myc mRNA by a protein kinase C-dependent pathway. J. BioZ. Chem. 266: 681-684, 1991. Spivak, J. L. The mechanism of action of erythropoietin. Int. J. CeZZ Cloning 4: 139-166, 1986. Sutherland, C. A., and D. Amin. Relative activities of rat and dog platelet phospholipase A2 and diglyceride lipase. J. BioZ. Chem. 257: 14006-14010, 1982. Tanaka, Y., F. Amano, H. Kishi, M. Nishijima, and Y. Akamatsu. Degradation of arachidonyl phospholipids catalyzed by two phospholipases A, and phospholipase C in a lipopolysaccharide-treated macrophage cell line, RAW 264.7. Arch. Biochem. Biophys. 272: 210-218, 1989. TOU, J.-S. Acylation of docosahexaenoic acid into phospholipids by intact human neutrophils. Lipids 21: 324-327, 1986. Tsuda, H., T. Sawada, M. Sakaguchi, M. Kawakita, and K. Takatsuki. Mode of action of erythropoietin (Epo) in an Epo-dependent murine cell line. I. Involvement of adenosine 3’,5’cyclic monophosphate not as a second messenger but as a regulator of cell growth. Exp. HematoZ. 17: 211-217, 1989. White, L., J. W. Fisher, and W. J. George. Role of erythropoietin and cyclic nucleotides in erythroid cell proliferation in fetal liver. Exp. Hematol. 8: 168-181, 1980.

Downloaded from www.physiology.org/journal/ajpcell by ${individualUser.givenNames} ${individualUser.surname} (137.154.019.149) on January 10, 2019.