Lysophosphatidic Acid-Induced Platelet Shape Change Revealed Through LPA_1–5 Receptor-selective Probes and Albumin

Abstract

Lysophosphatidic acid (LPA), a component of mildly-oxidized LDL and the lipid rich core of atherosclerotic plaques, elicits platelet activation. LPA is the ligand of G protein-coupled receptors (GPCR) of the EDG family (LPA_1–3) and the newly identified LPA _4–7 subcluster. LPA₄, LPA₅ and LPA₇ increase cellular cAMP levels that would induce platelet inhibition rather than activation. In the present study we quantified the mRNA levels of the LPA_1–7 GPCR in human platelets and found a rank order LPA₄=LPA₅>LPA₇>LPA₆=LPA₂>>LPA₁>LPA₃. We examined platelet shape change using a panel of LPA receptor subtype-selective agonists and antagonists and compared them with their pharmacological profiles obtained in heterologous LPA_1–5 receptor expression systems. Responses to different natural acyl and alkyl species of LPA, and octyl phosphatidic acid analogs, alpha-substituted phosphonate analogs, N-palmitoyl-tyrosine phosphoric acid, N-palmitoyl-serine phosphoric acid were tested. All of these compounds elicited platelet activation and also inhibited LPA-induced platelet shape change after pre-incubation, suggesting that receptor desensitization is likely responsible for the inhibition of this response. Fatty acid free albumin (10 µM) lacking platelet activity completely inhibited platelet shape change induced by LPA with an IC₅₀ of 1.1 µM but had no effect on the activation of LPA_1,2,3,&5 expressed in endogenously non-LPA-responsive RH7777 cells. However, albumin reduced LPA₄ activation and shifted the dose-response curve to the right. LPA₅ transiently expressed in RH7777 cells showed preference to alkyl-LPA over acyl-LPA that is similar to that in platelets. LPA did not increase cAMP levels in platelets. In conclusion, our results with the pharmacological compounds and albumin demonstrate that LPA does not induce platelet shape change simply through activation of LPA_1–5, and the receptor(s) mediating LPA-induced platelet activation remains elusive.

Introduction

Oxidative modifications of LDL and platelet activation are the central events in the pathogenesis of atherosclerosis and cardiovascular disease [1,2]. Oxidatively modified LDL, present in the circulation or exposed after rupture of atherosclerotic plaques to blood cells, is known to stimulate platelets [3,4]. Platelets, once activated, change their shape, aggregate, and become procoagulant, leading to the formation of a platelet- and fibrin-rich intravascular thrombus, which can precipitate acute ischemic syndromes such as unstable angina, myocardial infarction, and stroke [3,5]. Lysophosphatidic acid (LPA) has been identified as a biologically active lipid in mildly-oxidized LDL (mox-LDL), human atherosclerotic plaques, and supernatant of activated platelets [2,6,7,8,9]. It is also known that serum LPA level is significantly elevated in patients with acute myocardial infarction [10]. Importantly, oxidatively modified LDL such as mox-LDL and the lipid-rich core of human atherosclerotic plaques have been shown to stimulate platelets through the activation of LPA receptors [2,6].

LPA is predominantly associated with serum albumin [11] and acts through various sets of specific G protein-coupled receptors (GPCRs) in an autocrine and paracrine fashion [8,12,13]. The LPA receptor gene products are expressed in most mammalian tissues with spatially and temporally regulated expression patterns. LPA receptors can be divided into two subfamilies. One is composed of three members, LPA₁, LPA₂, LPA₃, belonging to the Endothelial Differentiation Gene (EDG-) subfamily of GPCRs [14]. The second subfamily consists of recently identified LPA receptors: LPA₄ (GPR23), LPA₅ (GPR92), LPA₆ (GPR87), and LPA₇ (P2Y5) [15,16,17,18] which are structurally more related to the purinoreceptor (P2Y) cluster of GPCRs.

Platelets express mRNA for the LPA_1–5 receptors [19,20,21]; the expression of LPA receptors at the protein level is unknown due to the lack of specific antibodies. Although LPA-induced platelet activation is inhibited by the LPA₁ and LPA₃ receptor antagonist dioctylglycerol pyrophosphate (DGPP) (8:0), as well as either N-palmitoyl-serine phosphoric acid (NPSPA) or N-palmitoyl-tyrosine phosphoric acid (NPTPA), both of which activate several LPA receptors [7,22], the LPA responses in platelets (alkyl-LPA being potent than acyl-LPA) are not consistent with the pharmacological properties of LPA₁ and LPA₃ receptors [6,23,24]. Therefore, the receptor(s) through which LPA stimulates platelets remain unidentified. Here we applied a comprehensive set of recently identified LPA receptor agonists and antagonists to evaluate their effects in cells individually expressing LPA_1–5 and compared those with platelet responses to LPA. We have also evaluated the effect of serum albumin that inhibits LPA-induced platelet activation on LPA-GPCR individually expressed in endogenously non-responsive RH7777 cells. The present findings underline the complexity of LPA responsiveness in human platelets revealing a mismatch with the individual pharmacological properties of LPA_1/2/3/4/5.

Materials and methods

Materials

The LPA receptor agonists and antagonists were purchased either from Avanti Polar Lipids (Alabaster, AL) or synthesized as previously described [25,26]. ADP, apyrase (A-6535), fatty acid free bovine serum albumin (FAF-BSA), acetylsalicylic acid were obtained from Fluka (Taufkirchen, Germany). The enzyme immunoassay kit for cyclic adenosine monophosphate (cAMP) measurement was from Assay Designs, distributed by Biotrend (Cologne, Germany). DNase I, ThermoScript RT-PCR System for First-Strand cDNA Synthesis were purchased from Invitrogen (Carlsbad, CA). RT2 Real-Time SYBR Green/ROX kit was from SuperArray (Frederick, MD).

Agonist and Antagonist preparation

LPA species were either dissolved and diluted in ethanol, or else after ethanol evaporation reconstituted in FAF-BSA buffer (20 mM HEPES, 138 mM NaCl, FAF-BSA 0.25 mM, at a LPA/BSA ratio of 4:1). The pharmacological compounds were dissolved and used in methanol or in FAF-BSA buffer (compound/BSA ratio of 4:1), as described [26].

Preparation of Washed Human Platelets for Activation Studies

Blood was obtained by venipuncture from healthy male and female donors using 3.13% natrium citrate as an anticoagulant. Venous blood was drawn and collected into plastic tubes containing natrium citrate in a 1:10 anticoagulant to blood ratio. All studies involving human subjects were conducted in accordance with the Declaration of Helsinki following the protocols approved by the ethical committee of LMU Munich. The whole blood was centrifuge at 180g for 20 min at 24°C. The supernatant (platelet-rich plasma) was aspirated, taking care not to disturb the blood/plasma interface and PRP was incubated in the presence of acetylsalicylic acid (1mM) and apyrase (0.3 U/mL) for 15 min at 37°C. Subsequently, citric acid (9 mM) and EDTA (5 mM) were added and PRP was centrifuged at 800 g for 20 min at 24°C. The platelets were washed in buffer B (20 mM HEPES, 138 mM NaCl, 2.9 mM KCl, 1 mM MgCl₂, 0.36 mM NaH₂PO₄, 0.3 U /mL apyrase; pH 6.2). The platelet pellet was then washed once with buffer B containing apyrase (0.3U/mL). The final platelet pellet was resuspended at a concentration of 400,000/µL in buffer C (20 mM HEPES, 138 mM NaCl, 2.9 mM KCl, 1 mM MgCl₂, 0.36 mM NaH₂PO₄, 5 mM glucose, 0.6 U/mL apyrase; pH 7.4).

Measurement of Platelet Shape Change

For measurement of shape change, human platelets were isolated and incubated at 37°C with various concentrations of the LPA receptor agonists and antagonists, FAF-BSA or vehicle control before exposure to LPA. Shape change was measured by the decrease in light transmission of the stirred (1100 rpm) platelet suspension as described previously [6]. To analyze agonist effect of the compounds, washed platelets were incubated at 37°C for 2 minutes and stimulated with different concentration of compounds dissolved in methanol or FAF-BSA buffer (at a lipid:BSA molar ratio 4:1). For analysis of the antagonist effect of the compounds, washed platelets were incubated with different concentrations of the compounds at 37°C for 30 minutes or shorter and stimulated with 20nM acyl-LPA 16:0.

Measurement of platelet cAMP

Platelets were pretreated with aspirin, washed, and resuspended in buffer (400,000/µL) containing apyrase as described above. Platelet suspensions were incubated at 37°C while stirring before exposure to LPA (0.02- 40 µM) or iloprost (50 nM) for 1 minute. Levels of cAMP were determined with an enzyme immunoassay kit from Assay Designs GmBH according to the instructions of the manufacturer.

Isolation of mRNA from Human Platelets

Two units of platelet rich plasma (∼505 ml containing ∼57 ml ACD each expired for human use 1 day prior) were put through a Purecell PL high efficiency leukocyte reduction filter for platelets (PL6T from Pall Biomedical Products Company in East Hills, NY). The leukocyte reduced filtrate was then mixed with an equal volume of buffer consisting of 138mM NaCl, 3.3mM NaH₂PO₄, 2.9mM KCl, 1mM MgCl₂, 20mM HEPES, 1mg/ml glucose, containing 15% of 0.8% citric acid, 2.2% Na-citrate2H₂O, 2.45% glucose and 1µM PGE₁ (Sigma-Aldrich, St. Louis, MO). The sample was centrifuged at 3000 × rpm for 10 min at room temperature and the supernatant was carefully removed and discarded. The platelet pellet was transferred to 2ml Eppendorf tubes and 1ml of TRIZol reagent (Invitrogen, Carlsbad.CA) was added. RNA was extracted according to the protocol provided by the manufacturer. The RNA pellet was dissolved in 15 µl DEPC-treated water yielding ∼50 µg per unit of platelets.

Quantitative PCR of LPA Receptor mRNA in Human Platelets

One µg of total RNA was digested with DNase I and used for the subsequent synthesis of cDNA using the First Strand Synthesis kit as recommended by the manufacturer. The following primer pairs were used: LPA₁: (forward) GTCTTCTGGGCCATTTTCAA and (reverse) TCATAGTCCTCTGGCGAACA; LPA₂: (forward) GGGCCAGTGCTACTACAACG and (reverse) ACCAGCAGATTGGTCAGCA ; LPA₃: (forward) GAAGCTAATGAAGACGGTGATGA and (reverse) AGCAGGAACCACCTTTTCAC ; LPA₄: (forward) TCTGGATCCTAGTCCTCAGTGG and (reverse) CCAGACACGTTTGGAGAAGC ; LPA₅: (forward) CGCCATCTTCCAGATGAAC and (reverse) TAGCGGTCCACGTTGATG ; LPA₆: (forward) AAATCCAGCAGGCAATTCAT and (reverse) CCCTGATGCTCTGGTTATGTT ; LPA₇: (forward) TCTGGCAATTGTCTACCCATT and (reverse) TCAAAGCAGGCTTCTGAGG ; β-actin: (forward) TTCTACAATGAGCTGCGTGTG and (reverse) GGGGTGTTGAAGGTCTCAAA. The primer sets were designed with a melting temperature of 59–61°C. Amplicon size was 50–200 bases. Amplification was performed for 40 cycles at 94ºC/15 sec and 60ºC /60 sec using an ABI Model 7300 Real Time PCR machine (Foster City, CA). Quantitative values were obtained from the threshold cycle value (Ct), which is the point where a significant increase of fluorescence is first detected. The transcript number of human β-actin was quantified as an internal RNA control, and each sample was normalized on the basis of its β-actin content. The relative gene expression level of each gene was then normalized to LPA₁ gene (calibrator). Final results, expressed as N-fold difference in gene expression relative to β-actin and LPA1, termed N, were calculated as: N = 2^{(Ct gene - Ct calibrator)} (http://dorakmt.tripod.com/genetics/realtime.html), where Ct values of the gene and calibrator were determined by subtracting the average Ct value of a target gene from the corresponding Ct value of the β-actin gene.

Measurement of LPA-elicited Ca²⁺transients

Wild type McArdle rat hepatoma (RH7777, from ATCC) cells do not respond to LPA with changes in [Ca²⁺]_i. RH7777 cell lines individually expressing either LPA₁, LPA₂, LPA₃ or LPA₅ receptors and Chinese hamster ovary (CHO) cell line were used to examine agonism and antagonism of the compounds. CHO cells stably expressing either vector or LPA₄ were a kind gift from Dr. Takao Shimizu (University of Tokyo, Tokyo, Japan).

Stable transformants of LPA_1/2/3 receptors; RH7777 cells stably expressing each receptor were plated onto poly-L-lysine (PLL, 0.1 mg/ml)-coated black-wall clear-bottom 96- well plates (Corning Incorporated Life Sciences, Acton, MA) at a density of 5 × 10⁴ cells/well and cultured overnight. The following day, the culture medium was replaced with modified Krebs buffer (120 mM NaCl, 5 mM KCl, 0.62 mM MgSO4, 1.8 mM CaCl2, 10 mM HEPES, 6 mM glucose, pH 7.4), and the cells were serum starved for 6 h. Subsequently, cells were loaded with Fura-2 AM (Invitrogen, Carlsbad, CA) for 35 min in modified Krebs buffer containing 2% (v/v) pluronic acid. Stable transformants of LPA₄; CHO cells stably expressing either vector or LPA₄ were plated non-coated 96 well plates at a density of 4 × 10⁴ cells/ well and cultured overnight. The following day, cells were loaded with Fura-2 AM for 1 h in modified Krebs buffer containing 2% (v/v) pluronic acid and 2.5mM probenecid.

For transient transfection of LPA₅, RH7777 cells in 10 cm dish at a density of 2 × 10⁶ were transfected with 2 µg of plasmid DNA with Effectene (Qiagen, Valencia, CA) according to the manufacturer's instructions for 24 h, then replated onto PLL-coated 96- well plates at a density of 5 × 10⁴ cells/well and cultured overnight. The following day, the culture medium was then replaced with modified Krebs buffer, and the cells were serum starved for 4 h. Subsequently, cells were loaded with Fura-2 AM for 30 min in modified Krebs buffer containing 2% (v/v) pluronic acid, rinsed with Krebs buffer and changes in the intracellular Ca²⁺ concentration were monitored by determining the ratio of emitted light intensities at 520 nm in response to excitation at 340 and 380 nm using FLEXstation II (Molecular Devices, Sunnyvale, CA). Each well was monitored for 80–120s. For testing agonist activity of the compounds, the test compounds were added automatically after 15 s of baseline measurement. To determine antagonist properties, varying concentrations of the compounds were mixed with constant concentration of LPA and responses were monitored. Each test was performed in quadruplicate. CHO cells endogenously express LPA₁, therefore, to access the effect of LPA₄ in CHO cells, the response in vector-transfected cells were subtracted from the response in LPA₄-transfected cells.

Statistical Analysis

Significant difference was determined by the Student’s test at a P value of 0.05. IC₅₀ values were calculated by fitting a sigmoid function to data points by using the nonlinear curve-fitting feature of KaleidaGraph (Synergy Software, Essex Junction, VT).

Results

Expression of LPA Receptors mRNAs in Human Platelets

We applied real-time PCR to quantify the abundance of LPA_1–7 mRNAs in purified human platelets. We used platelets isolated and pooled from four healthy human donors. The platelets used for mRNA extraction have been depleted of white blood cells and red blood cells using Purecell PL membranes. The purified platelet preparation was stained with May-Grunwald Giemsa stain and non-platelet cells were counted. The preparation contained less than 0.01% white blood cells and was considered highly pure for platelets. The abundance of LPA receptor RNA arbitrarily normalized to LPA₁ is shown in figure 1. The rank order of abundance was LPA₄ = LPA₅ >LPA₇> LPA₆ = LPA₂ >> LPA₁ > LPA₃. These results suggest that LPA receptors of the purinergic cluster represent the most abundant number of transcripts in human platelets.

Figure 1.

Relative abundance of LPA receptor transcripts in purified human platelets determined by quantitative real-time PCR.

The effect of short chain octyl-serinediamide phosphates on platelets and LPA_4,5 receptors

Durgam and colleagues [26] have synthesized and partially characterized analogs of phosphatidic acid (PA) 1–7 (Figure 2). Compounds 2, 3, and 6, 7 had no agonist activity, but potent antagonist effects on LPA₁ and LPA₃ expressed in RH7777 cells; compounds 4 and 5 had agonistic activity on LPA_1–3 , but no antagonistic activity; compound 1 was a mixed LPA₂ agonist/ LPA_1,3 antagonist (Figure 2; Table I). We expanded the characterization of these compounds to human platelets that express very low copies of LPA₁ and LPA₃ (Figure1). The compounds were dissolved in either methanol or FAF-BSA buffer and tested for agonist and LPA-antagonist activity on human platelets. Methanol at the highest concentration tested (0.5% V/V) had no effect on LPA-mediated platelet activation (Figure 3A), whereas the FAF-BSA buffer (5µM) inhibited the LPA-response (Figure 3B). Nonetheless, the agonist activity of the compounds was independent of the type of vehicle used and the maximal efficacy of the drugs (E_max) was not significantly changed (Figure 3A and B; data not shown). Due to the interfering effect of 5 µM FAF-BSA with LPA-induced platelet activation, the compounds were dissolved in methanol for further analysis. In contrast to the results in heterologous expression system, all seven compounds induced platelet activation, and after 30 min incubation, inhibited the LPA-induced platelet shape change in a concentration-dependent manner (Table I). The EC₅₀ values of the compounds were much higher than that of LPA and showed no correlation with their potency established at LPA_1–3. Surprisingly, the LPA₃ selective antagonist compound 7 was maximally active in inducing platelet shape change with a very low EC₅₀ (∼550 nM) even though LPA₃ was the least abundant transcript in platelets (Table I, Figure 1–Figure 3). These data reveal a structure-activity relationship (SAR) in platelets that is not consistent with that of LPA₁, LPA₂, and LPA₃.

Figure 2.

Chemical structure of the compounds used in the present study.

Table I.

Effect of PA 8:0 analogs on intracellular Ca²⁺ transients of LPA_1–5 transfected cells [25] and on platelet shape change

HETEROLOGOUS EXPRESSION SYSTEMS					PLATELETS
PA(8:0) Analogs	LPA receptor	Agonist activity		Antago nist activity	Agonist activity		Antagonist activity
		aEmax %	EC50 µM	IC50 µM	bEmax %	EC50 µM	cImax %	IC50 µM	n
1
LPA1/LPA3	LPA1	NE		0.68
Antagonist	LPA2	58	6.33	NE	69±12	1.77±0.22	76±18	2.92±1.73	3
LPA2	LPA3	NE		0.011
Agonist
2	LPA1	NE		1.58
LPA1/LPA3	LPA2	NE		NE	38±18	5.91±2.62	72±9	1.7±1.5	3
Antagonist	LPA3	NE		0.14
3	LPA1	NE		0.14
LPA1/LPA3	LPA2	NE		NE
Antagonist	LPA3	NE		0.184	69±13	1.79±0.95	91±9	0.55±0.37	3
LPA5	LPA4	NE		NE
Agonist	LPA5	92	NS	NE
4	LPA1	57	3.26	NE
LPA1/LPA3	LPA2	NE		NE	57±4	2.43±1.1	93±4	0.35±0.1	3
Agonist	LPA3	109	0.16	NE
5	LPA1	30.6	0.69	NE
LPA1–3	LPA2	27	5.72	NE
Agonist	LPA3	109	0.003	NE	48±10	0.39±0.11	87±13	0.21±0.16	4
LPA5	LPA4	NE		NE
Agonist	LPA5	97	NS	NE
6	LPA1	NE		NE
select.LPA3	LPA2	NE		NE	68±12	8.4±0.74	80±2	8.2±0.96	3
Antagonist	LPA3	NE		0.414
7	LPA1	NE		NE
select.LPA3	LPA2	NE		NE
Antagonist	LPA3	NE		0.935	110±9	0.55±0.18	95±5	2.6±0.4	3
LPA5	LPA4	NE		NE
Agonist	LPA5	36	NS	NE

^aNote: E_max = maximal efficacy of the drug/maximal efficacy of LPA 18:1; NE= no effect; NS = the response did not saturate at the highest concentration tested (10 µM).

^bE_max= (maximal shape change induced by drug/ shape change induced by 20 nM acyl-LPA 16:0)x100. Shape change induced by 20–25 nM LPA was 89±11 % (mean SD, n=22) of maximal, and set to 100 %. Values are mean ± SD from different experiments with different platelet donors. Inhibitory activity was tested 30 min after addition of the phospholipids, when platelets had reached again their discoid shape (Figure 3A, right tracings).

^cI_max= maximal inhibition of LPA-induced shape change tested by 20 µM of PA analogs.

Figure 3. Effect of LPA₃ antagonist (7) on platelet.

Suspensions of washed human platelets were incubated with solvent or 20 µM 7 dissolved A, in methanol or B, in albumin buffer (lipid:BSA 4:1) for 30 min before exposure to acyl-LPA 16:0 (20 nM). Shape change was recorded as decrease of light transmission. Tracings shown are representative for 3 experiments.

Since the compound 4 an apparent LPA₁/LPA₃ agonist, was found to be a weak agonist, and had a higher antagonistic than agonistic potency (IC₅₀ 0.35±0.1 µM; EC₅₀ 2.43±1.1 µM) (Figure 2; Table I), further experiments were carried out with this compound to distinguish whether platelet inhibition by 4 was due to LPA receptor desensitization or antagonism. Concentrations of 4 that did not induce shape change were chosen, and incubated with platelets for various times (5 sec, 5 min, 15 min) before addition of LPA. No inhibition was found after 5 sec pre-incubation but shape change was inhibited progressively at longer pre-incubation times, consistent with a mechanism of receptor desensitization rather than a LPA receptor antagonistic effect (Figure 4).

Figure 4. The inhibitory effect of a LPA_1/3 agonist (compound 4) on platelets is pre-incubation time dependent.

A concentration of 0.2 µM compound 4 that itself did not induce shape change was incubated for different time periods before exposure to acyl-LPA 16:0 (20nM).

As the compounds 3, 5, 7 induced platelet shape change with higher potency, we also determined their action on LPA₄ and LPA₅ receptors. All compounds had no effect on LPA₄ activation and were weak agonists of LPA₅ (Table I). These data indicate that LPA₅ could be involved in LPA-induced platelet shape change.

The effect of N-palmitoyl amino phosphoric acids on LPA responses in platelets and heterologously expressed LPA_1–5

We and others have shown previously that compounds 10 NPSPA and 11 NPTPA (Figure 2) effectively induce platelet shape change and aggregation; however, after preincubation inhibit platelet responses to LPA [6,27,28]. NPSPA was slightly more potent than NPTPA, the EC₅₀ and IC₅₀ of these phospholipids were in a similar range (between 0.2 and 0.3 µM) [6]. These compounds were reported not to inhibit the EDG family LPA receptors [22]. Here we tested both compounds for their effects on LPA_1–5 (Table II). NPSPA was a weak agonist of the EDG family LPA receptors, had no effect on LPA₄ and was a partial agonist of LPA₅. Also, NPTPA was inactive on LPA₄ and in the micromolar range was partial agonist of LPA₅. In contrast to NPSPA, NPTPA weakly inhibited LPA₁ with a K_i of 533 nM and had no agonist or antagonist effect on LPA₂, LPA₃. Like in the case of the PA analogs, we could not draw a clear correlation between their pharmacological profile and the expression of the LPA receptor found in platelets.

Table II.

Effect of NPSPA NPTPA on LPA responses of heterologously expressed LPA_1–5

Compound	LPA1	LPA2	LPA3	LPA4	LPA5
NPSPA	Partial agonist (Emax= 63%)	Partial agonist (Emax= 46%)	Partial agonist (Emax= 81%)	No effect	Partial agonist (Emax= 23%)
	EC50= 0.74 µM	EC50= 11.9 µM	EC50= 0.73 µM		EC50= 0.42 µM
NPTPA	Antagonist	No effect	No effect	No effect	Partial agonist (Emax= 72.0%)
	Ki= 0.53 µM				EC50 - NS

Note: E_max = (maximal efficacy of compound/maximal efficacy of LPA 18:1)×100

EC₅₀ = concentration eliciting half maximal activation

K_i = inhibitory binding constant

NS= the response did not saturate at the highest concentration tested

The effect of LPA₄ receptor agonist and antagonist compounds on platelets and LPA₅

Jiang et al. reported the initial characterization of a series of α-substituted phosphonate analogs of LPA and the first two compounds 8 and 9 (Figure 2) with LPA₄ receptor agonist and antagonist properties, respectively [25]. Here we have expanded the characterization of the compounds to include LPA₅ and also assessed their actions in platelets. Both compounds 8 and 9 were weak agonists of LPA₅ yielding 59% and 45% activation of the maximum response to LPA 18:1 at 10 µM, respectively (Table III). Both α-substituted phosphonate analogs activated platelets and inhibited LPA-induced platelet shape change after 30 min pre-incubation. Interestingly, these two phosphonate analogs were more active than the PA analogs as shown by their lower EC₅₀ values (Table III). These data indicate that these analogs are not specific for LPA₄ and also suggest that not LPA₄ but LPA₅ could be involved in LPA-elicited platelet shape change. We have also examined the agonist properties of alkyl-LPAs relative to acyl-LPAs at LPA₄ expressed heterologously in CHO cells and found that the alkyl-ether-analogs were weaker agonists than the acyl-analogs with an equal number of hydrocarbons (Table III).

Table III.

Effect of acyl and alkyl LPAs and α-substituted phosphonate analogs of LPA on intracellular Ca²⁺ transients of LPA_1–5 transfected cells and platelet shape change

HETEROLOGOUS EXPRESSION SYSTEMS					PLATELET
Analogs	LPA receptor	Agonist activity		Antago nist activity	Agonist activity		Antagonist activity
		aEmax %	EC50 µM	IC50 µM	bEmax %	EC50 nM	cImax %	IC50 nM	n
Acyl LPA 18:1	LPA1	100	0.130
	LPA2	100	0.003
	LPA3	100	0.081		100	9±2.9			4
	LPA4	100	0.245
	LPA5	100	0.015
Alkyl LPA 18:1	LPA1	100	1.445
	LPA2	108	0.100
	LPA3	63	dNS		100	1.85±0.7			4
	LPA4	98	0.303
	LPA5	97	0.002
8	LPA1	NE		PA
LPA4,5	LPA2	NE		2.59
Agonist	LPA3	NE		2.56	78±9	280±40	81±5	900±110	3
LPA1–3	LPA4	29	5.4	NE
Antagonist	LPA5	59	dNS
9	LPA1	NE		1.5
LPA1–4	LPA2	NE		1.42
Antagonist	LPA3	NE		1.16	73±9	810±370	80±9	300±150	3
LPA5	LPA4	NE		0.27
Agonist	LPA5	45	dNS

^aNote: E_max = (maximal efficacy of compound/maximal efficacy of LPA 18:1)x100; NE= no effect; PA=partial antagonist with 39±3.2% inhibition of 200 nM LPA response at the highest concentration (30µM) tested.

^bE_max=(maximal shape change induced by drug/ shape change induced by 20 nM acyl-LPA 16:0)×100. Values are mean ± SD from different experiments with different platelet donors. Inhibitory activity was tested 30 min after addition of the phospholipids.

^cI_max= maximal inhibition of LPA-induced shape change tested by 10 µM of analogs.

^dNS = the response did not saturate at 10 µM the highest concentration tested preventing the determination of EC₅₀

LPA₅ Receptor and Platelet Activation Shows Similar SAR

Recently GPR92 has been identified as a new LPA receptor (LPA₅) and we have been able to test the SAR of this receptor in RH7777 using the Ca²⁺ mobilization assay [16]. In platelets and in LPA₅–transfected RH7777 cells a similar rank order of activation was found: alkyl-LPA 18:1 ≥ alkyl-LPA 16:0 > acyl-LPA 18:1 >> alkyl-LPA 18:0 (Table IV; Figure 5; data not shown). This striking similarity in the SAR argues for the possible involvement of LPA₅ receptor in mediating LPA-induced platelet shape change.

Table IV.

Activities of different molecular species of LPA

LPA species	Platelets	LPA5 transiently transfected RH7777 cells
	EC50 nM	EC50 nM
acyl-LPA 18:1	9 ±2.9	15.2±5.4
alkyl-LPA 16:0	4.6±0.98	4.1±2.1
alkyl-LPA 18:0	20.17±17.5	69.9±21.4
alkyl-LPA 18:1	1.85±0.7	2.1±0.9

Note: EC₅₀ values were calculated from concentration-response curves induced by various LPA species in platelets (n=4) and LPA₅ receptor transiently transfected RH7777 cells (mean ± SD).

Figure 5. Dose-response curves induced by various LPA species.

Suspensions of washed human platelets were stirred at 37°C for 2 minutes and then exposed to increasing concentrations of LPA species. Shape change was measured by the decrease in light transmission in a LABOR aggregometer^®. Values represent the mean ± SD from 4 different experiments with different platelet donors.

The effect of FAF-BSA on LPA-induced platelet shape change

In serum, LPA is predominantly associated with serum albumin [11]. In order to investigate the effect of FAF-BSA on LPA-induced platelet shape change, we determined the dose-response curves for acyl-LPA 16:0 dissolved in ethanol or FAF-BSA buffer. As shown in Figure 6A, the presence of FAF-BSA (molar LPA: BSA ratio, 4:1) had only a small inhibitory effect on platelet activation induced by LPA, which was not significant (p>0.05). However, pre-incubation of platelets with FAF-BSA at a concentration of 2.5 µM shifted the dose-response curve of LPA to the right (Figure 6B). LPA (1 µM) could overcome the inhibition of FAF-BSA. FAF-BSA inhibited platelet shape change regardless of whether acyl- or alkyl-LPA was used (data not shown).

Figure 6. Effect of albumin on LPA-mediated shape change.

A, Dose-response curves comparing the effect of LPA dissolved in ethanol (dotted line) and in albumin buffer (LPA:BSA 4:1) (solid line) on platelet shape change (mean SD, n=3). B, BSA inhibits platelet shape change in a competitive manner. Platelets were incubated with BSA (2.5 µM) (solid line) for 1 min before exposure to increasing concentration of LPA (in albumin buffer), n=3. C, BSA inhibits LPA-induced platelet shape change in a dose-dependent manner. Washed platelets were incubated with different concentrations of BSA for 1 min before exposure to 20nM of LPA (in albumin buffer), n=6. D, BSA reduced LPA-initiated platelet shape change in a time-independent manner. Incubation of platelets was performed with BSA (2.5 µM) for different time periods before exposure to 20nM of LPA (in albumin buffer), n=3. Values represent the mean ± SD. *<0.05 vs. control.

To quantify the inhibitory effect of albumin on LPA-induced platelet shape change, we pre-incubated platelets with increasing concentrations of FAF-BSA (0.1–5µM) before addition of 20 nM acyl-LPA 18:1 (dissolved in FAF-BSA buffer). As shown in Figure 6C, FAF-BSA dose-dependently decreased the LPA-elicited shape change response which was complete at 5µM concentration. The calculated IC₅₀ of FAF-BSA was 1.15±0.37 µM (mean SD, n=6). Furthermore, we investigated whether the inhibitory effect of FAF-BSA required pre-incubation with the platelets. As shown in figure 6D, inhibition of LPA-induced shape change by FAF-BSA was independent of the duration of pre-incubation time. This suggests that the inhibitory effect might be due to albumin’s avidity for LPA or due to albumin modulation of platelet LPA receptors.

The effect of FAF-BSA on LPA_1–5 receptor activation

To evaluate the effect of albumin on the LPA receptors and draw a parallel between the activation of platelet LPA receptors and that of LPA_1–5, we tested the effect of albumin on these heterologously expressed receptors. In contrast to the FAF-BSA-inhibition of LPA-induced platelet activation, the pre-incubation of FAF-BSA (15µM) for 5 min did not inhibit activation of LPA₁, LPA₂, LPA₃ and LPA₅ receptors in the heterologous expression system (Figure 7). However, LPA₄ activation was inhibited by 40% in the presence of 15µM FAF-BSA and the dose-response curve of LPA was shifted to the right (Figure 7).

Figure 7. Effect of albumin on LPA_1–5 activation in heterologous expression systems.

The effect of BSA (15µM, 5min pre-incubation) on LPA receptor activation induced by LPA was examined by measuring the transient increase of intracellular [Ca²⁺] in Fura-2-loaded RH7777 cells expressing human LPA_1–3 and LPA₅, or CHO cells expressing LPA₄ (mean ± SD).

The effect of LPA on cAMP levels in platelet

It has been shown that LPA₄ and LPA₅ activation in addition to eliciting Ca²⁺ transients induces an increase of intracellular cAMP [16,29,30]. In the present study we investigated whether LPA might also induce the activation of G_s that leads to an increase of cAMP levels in platelets. Levels of cAMP in unstimulated platelets were 5.83±1.26 pmol/10⁸ platelets (Figure 8). In platelets treated with a large concentration range of acyl-LPA 18:1 (0.02 – 40 µM) and for different times (30 s, 1 min, 5 min, 15 min) no significant change in cAMP was found (p>0.05) (Figure 8; data not shown). The prostacyclin-analog, iloprost (50 nM), elicited a robust increase in cAMP (Figure 8). Also acyl-LPA 16:0 and alkyl-LPA 16:0 did not affect platelet cAMP levels (data not shown). Previously we have found that LPA added subsequently to iloprost did not change the cAMP levels stimulated by this prostaglandin analog [31]. Therefore, platelet LPA receptors are unlikely to couple to G_i and G_s.

Figure 8. Effect of LPA on cAMP levels in platelet.

Suspensions of washed human platelets were at 37°C for 2 minutes and were exposed to increasing concentration of LPA (0.02 µM – 40 µM) or iloprost (50nM) for 1 minute. Levels of cAMP were determined with an enzyme immunoassay kit. Values represent the mean ± SD from different experiments with different platelet donors (n=3).

Discussion

LPA accumulating in atherosclerotic plaque and oxidatively modified LDL activates human platelets [2,6,7]. Although, alkyl glycerophosphate, commonly known as alkyl-LPA, is active at subnanomolar concentrations making it a very potent platelet activating mediator, the receptor(s) mediating this effect remain unknown [6]. The LPA response in platelets shows some distinct features – preference to alkyl-LPA over acyl-LPA, sensitivity to albumin, SAR of inhibitors – that could not be reconciled with the EDG family of LPA receptors. The EDG family shows clear preference for acyl forms of LPA and little inhibition by albumin, although LPA₃ shows some sensitivity to this protein carrier [32]. In light of the recent discovery of a new subcluster of LPA receptors with high degree of sequence homology to the purinergic GPCR, in the present paper we extended a comprehensive analysis of SAR to LPA_1–5. The expression of some receptors of the purinergic cluster has been recently reported in human platelets [20]. While this work was nearing completion, LPA₆ and LPA₇ were reported [17,18]. Our initial attempts to obtain Ca²⁺ transients from LPA₆ and LPA₇ transfected RH7777 cells were unsuccessful so we have omitted it from our functional assays. Nevertheless, we found LPA₆ and LPA₇ transcripts in human platelets at a level similar to that of LPA₂. The role of LPA₆ or LPA₇ and other not yet reported LPA receptors will have to be addressed in future studies.

The main finding of this study is that the antagonists that have been developed to the EDG family LPA_1–3 receptors and to LPA₄ act as agonists for platelet activation. Unexpectedly, some of them, for example the selective LPA₃ receptor antagonist octyl serine diamide compound 7, were more effective in inducing platelet shape change than inhibiting LPA₃. This is in sharp contrast with our earlier analysis of the SAR of platelet responses in which we proposed that LPA₁ and LPA₃ activation triggers LPA-induced platelet activation [6]. The present observations supported by the low abundance of LPA₁ and LPA₃ transcripts in human platelets necessitates the revision of this working hypothesis and argues against that the LPA₁ or LPA₃ receptors mediate LPA-induced platelet shape change. Although knockout mice are available to these two receptor subtypes, direct analysis of their platelet responses is not possible because rodent platelets do not respond to LPA [33].

By comparing the biological effect of three molecular LPA species, it has been previously shown that alkyl-LPA species are more potent at inducing platelet activation than acyl-LPA [6,34]. The present study confirms in part these observations. Our experiments demonstrated that the LPA₅ receptor had pharmacological properties similar to LPA-induced platelet activation arguing for the possible involvement of LPA₅ receptor in mediating platelet shape change. However, it has been shown that LPA₅ activation induces an increase of intracellular cAMP in LPA₅ transfected HeLa HF1 and B103 cells [16,29], and an increase of cAMP mediates platelet inhibition. Similarly, LPA₇ has been shown to couple to the elevation of cAMP [18]. We did not observe an increase of cAMP upon LPA stimulation of platelets. Very recently, Pamuklar and colleagues have noted that platelets of ∼20% of individuals failed to aggregate to LPA and showed an increase of platelet cAMP to LPA (1µM), which correlated with a higher expression of LPA₄ transcripts as compared to the LPA responsive platelet donors. These authors suggested that LPA stimulation of a receptor coupled to cAMP increase could prevent LPA induced platelet aggregation in the non-responsive donors [21]. We also observed previously donor-dependent variations of LPA-induced platelet aggregation in blood and PRP. However, LPA-induced platelet shape change in blood and washed cell suspensions was donor-independent [31]. Our results showing no measurable changes in cAMP and platelet shape change upon LPA-stimulation do not support this hypothesis.

Tokumura et al. have shown that albumin can dose-dependently inhibit LPA-induced platelet aggregation [34]. However, its effect on LPA-induced platelet shape change has not been examined systematically. In our study, we found that FAF-BSA inhibited platelet shape change induced by LPA bound to BSA in a competitive manner and independently of incubation time. It is unlikely that this inhibitory effect can be explained by the property of BSA to bind LPA, since platelet shape change was induced by LPA in complex with BSA at a molar ratio of 4:1. Albumin binds LPA with stoichiometry of 3 moles LPA to 1 mole of BSA [11,35] and we showed that LPA in complex with albumin (ratio LPA:BSA 4:1) activated platelets with similar potency as LPA dissolved in 70 % ethanol (final concentration < 0.07%). Therefore, it seems more plausible that albumin modulates LPA receptors. Indeed, Hama et al. showed that the inhibitory effect of albumin occurs in a receptor specific manner, and Sf9 cells expressing the LPA₃ receptor are more sensitive to BSA than LPA₁ and LPA₂ expressing cells [32]. However, we found that albumin had no effect on LPA₃ receptor activation in transfected RH7777 cells, which might be explained by the differences in the heterologous expression systems used. The lack of effect of albumin on LPA₁, LPA₂, LPA₃ and LPA₅ receptor activation in heterologous expression system further support our conclusion that these receptors cannot simply account for eliciting LPA-induced platelet shape change.

We also found that activation of LPA₄ receptor was significantly inhibited by a low concentration of albumin (15µM), raising the possibility that LPA₄ could play a role in LPA-mediated platelet activation. This would be supported by our Q-PCR data and those by others showing that LPA₄ and LPA₅ transcripts are much more abundant in human platelets than LPA₁ and LPA₃ transcripts [20,21]. However, our experiments with the compound 9, a LPA₄ antagonist, which stimulated platelets, argue against an involvement of LPA₄ in LPA-induced platelet activation.

In summary, the present results provide a comprehensive analysis of the pharmacological properties of the LPA-induced platelet shape changes, and indicate that LPA activates platelets not simply by either one of the LPA_1–5 receptors, but perhaps through a new LPA receptor such as LPA₆, LPA₇ or a yet unidentified GPCR. They point to the need of further studies on this pathophysiologically important question.