Human platelet a-adrenergic receptors and responses during pregnancy: No change except that with differing hematocrit

James M. Roberts, M.D., Vivian Lewis, M.D., Nancy Mize, B.S., Arlene Tsuchiya, M.D., and Jolene Starr, B.S.

San Francisco, California

The possibility that platelet aggregation is altered during pregnancy is controversial. We tested for

alterations of a-adrenergic receptor-induced platelet function during pregnancy compared to that in the

early phase of the menstrual cycle. We found no change in a-adrenergic receptor concentration or in

affinity of the receptor for epinephrine. The potency and sensitivity of epinephrine to inhibit adenylate cyclase, a response mediated by a2-adrenergic receptors, was not different in platelets from the two

groups of subjects. The ability of epinephrine to potentiate aggregation by adenosine diphosphate was significantly increased during pregnancy; however, this was an artifact introduced by lower hematocrits during pregnancy that resulted in an increased citrate concentration in the preparations of platelet-rich

plasma from these women. The difference was eliminated by normalization of citrate concentration. Thus effects of a-adrenergic receptors on platelet function are not different during pregnancy. It is mandatory to recognize artifactual effects of differing hematocrits on platelet aggregation studied in vitro. (AM J 0BSTET

GYNECOL 1986;154:206-10.)

Key words: a-Adrenergic receptors, platelet aggregation, pregnancy, hematocrit

Pregnancy and especially the postpartum period are associated with an increased risk of thromboembolism. 1

While changes in several components of the coagula­tion cascade during pregnancy are well established, 1

there is little agreement on the subject of platelet func­tion. In vitro assays have demonstrated that aggrega­tion is increased2

• 3 or unchanged•· 5 during pregnancy.

This conflicting information can be explained in part by differences in methodology and in part by the use of different agents to initiate aggregation, which may act through different mechanisms.6 Another problem in interpretation of these studies may be the failure to appreciate the artificial increase in sensitivity of platelet aggregation caused by decreased hematocrit. 7

We previously reported that treatment of rabbits with estrogen, which decreases platelet aggregation in this species, also decreases platelet a-adrenergic receptors,8

which are linked to aggregation. Therefore we decided to examine in detail this pathway for platelet aggre­gation in human pregnancy to determine if a-adren-

From the Department of Obstetrics, Gynecology and Reproductive Sciences and the Cardiovascular Research Institute, University of California (San Francisco).

This work was supported by National Institutes of Health Grant HD 16262 and by a grant from the French Hospital Research Faun- 1 dation, San Francisco, California.

ReceivedforpublicationApril12, 1985; revised September 27, 1985; accepted October 14, 1985.

Reprint requests: james M. Roberts, M.D., Department of Obstetrics, Gynecology and Reproductive Sciences, University of California (San Francisco), San Francisco, CA 94143.

206

ergic receptor concentration, agonist affinity for the receptor, and/or linkage of the a-adrenergic receptor to adenylate cyclase were different during gestation. These results were correlated with the ability of epi­nephrine to cause aggregation or potentiate adenosine diphosphate-induced platelet aggregation. We report that platelet a-adrenergic receptor concentration arid agonist affinity are similar in pregnant and nonpreg­nant women as is the ability of epinephrine to inhibit adenylate cyclase. When corrections are made for the decreased hematocrit in pregnant women compared with nonpregnant ones, there is also no change in epi­nephrine-stimulated platelet aggregation.

Material and methods

Materials. Tritium-labeled yohimbine (specific activ­ity 88 Ci/mmol) and tritium-labeled dihydroergocryp­tine (specific activity 46 Ci/mmol) were obtained from New England Nuclear. Phentolamine was a gift from Ciba Pharmaceutical Co. and prostaglandin E, a gift from The Upjohn Company. Other materials were the finest quality commercially available.

Subjects. Women in the studies were nonsmokers who had taken no medications for 2 weeks and had fasted for at least 3 hours before venipuncture. Preg­nant women were between 38 and 40 weeks of gesta­tion, and nonpregnant control subjects were in the first week of the menstrual cycle so that the hormonal en­vironment would be standardized. The follicular phase of the cycle was chosen as a time of low endogenous estrogen levels compared with those of pregnancy since

Volume 154 Number l

animal studies have implicated increased estrogen levels as determinants of platelet adrenergic receptor alterations.

Preparation of platelet-rich plasma lysates and par­ticulate. Venous blood was collected through a large­bore needle without stasis into a l/10 volume of 3.8% sodium citrate and 0.9% sodium chloride (citrate) in plastic syringes. Platelet-rich plasma was prepared by standard techniques.6 Platelet lysates and particulate were prepared by centrifuging platelet-rich plasma at 1200 x g for 15 minutes. Platelets were lysed by freeze­thawing. The resulting lysates were diluted and stored at - 70° C for later use in tritiated dihydroergocryp­tine-binding assays or centrifuged twice at 29,000 X g (with an intermediate wash) to be stored and used for adenylate cyclase assays or for binding assays with tri­tiated yohimbine.

Aggregation of platelet-rich plasma. Platelet aggre­gation was performed on a Payton dual-channel ag­gregometer.6 Aggregating agents used were epineph­rine (l0- 9 to I0- 5 moi/L) dissolved in l mmoi/L hy­drochloric acid and adenosine diphosphate (l0- 9 to

10-5 moi/L) dissolved in double-distilled water. Addi­tion of 25 J.LI of I mmol/L hydrochloric acid had no effect on aggregation. Each aggregating agent was used at several doses to generate a dose-response curve. A new platelet-rich plasma sample was used for each dose of agonist. In addition, a dose-response curve for epi­nephrine potentiation of adenosine diphosphate ag­gregation was generated: 60 seconds after addition of 25 J.LI of epinephrine (l0- 9 to 10-' moi/L) to platelet­rich plasma, 25 J.LI of adenosine diphosphate at a con­centration sufficient to cause 10% reversible aggrega­tion (l0- 7 to I0- 6 moi/L) was added. The time from obtaining of the platelets until their use in the assay and the order of experiments were the same for all samples. In other experiments the citrate concentration in platelet-rich plasma was adjusted to correct for dif­ferent concentrations of citrate in platelet-rich plasma prepared from blood with different hematocrits. 7

Tritiated yohimbine- and tritiated dihydroergo­cryptine-binding studies. Studies of tritiated yohim­bine and dihydroergocryptine binding were performed at 30° C in a total volume of 0.25 ml until equilibrium was reached (30 minutes for yohimbine, 60 minutes for dihydroergocryptine). The incubation mixture con­tained either yohimbine (0.15 to 10 nmoi/L) or dihy­droergocryptine ( 1 to 30 nmoi/L), platelet particulate or lysates (0.15 to 0. 75 mg of protein), 50 mmoi/L Tris, pH 7.4, 80 mmoi/L sodium chloride, 1 mmoi/L ascor­bate, and 0.1 mmol/L hydrochloric acid alone or with added adrenergic antagonists. The mixture also con­tained a final concentration of 2% ethanol to ensure solubility of dihydroergocryptine and yohimbine. This concentration of ethanol had no effect on binding of

a-Adrenergic receptors and responses during pregnancy 207

either radioligand. Incubations were terminated by the addition of 5 ml of ice-cold wash buffer (50 mmol/L Tris, pH 7.4) and immediate filtration over Whatman GFIC filters under vacuum after three 5 ml washes. Filters were dried by increased vacuum, and radioac­tivity was counted in a scintillation counter.

Specific binding was defined as the difference be­tween radioactivity bound in the absence of and that bound in the presence of 10 J.Lmol/L phentolamine and composed 75% to 90% of bound radioactivity for yohimbine and 50% to 60% of that for dihydroergo­cryptine; it was a linear function of added platelet pro­tein over the range of concentrations used in this study. Pissociation constants (Kd) for the radioligand and re­ceptor concentration were determined by a computer­assisted analysis of specific bound versus free radio!i­gand.9 In competition experiments the concentration of competitor that reduced radio ligand binding by 50% (IC50) was determined by computer-assisted curve fit­ting."

Since receptor concentration was low enough to guarantee that free ligand was similar at all competitor concentrations, K, could be calculated as:

where Kd is the dissociation constant for either yo­himbine or dihydroergocryptine, IC50 is the concentra­tion of adrenergic agent required to inhibit specific binding by 50%, and LF is the concentration of either yohimbine or dihydroergocryptine in the particular ex­periment. 10

Adenylate cyclase assay. Assays to determine the activity of adenylate cyclase were performed by a mod­ification of the method of Salomon et al.'' Platelet par­ticulates (20 J.Lg of protein) were added to incubation tubes containing 50 mmoi/L Tris buffer (pH 7.4), 80 mmoi/L sodium chloride, 6 mmoi/L magnesium chlo­ride, 0.2 mmol/L ethyleneglycol-bis(f3-aminoethyl ether)-N,N'-tetraacetic acid, 2 mmoi/L [3-mercaptoeth­anol, 30 mmol/L guanosine triphosphate, 0.1 mg/ml of bovine serum albumin, 10 mmol/L creatine phosphate, 10 units per milliliter of creatine phosphokinase, 0.4 mmoi/L adenosine triphosphate, 1 mmoi/L cyclic adenosine monophosphate, and adenosine triphos­phate labeled with cx[32-P] (1 to 1.5 million cpm). Tri­tiated cyclic adenosine monophosphate (approximately 30,000 cpm) was added to determine assay recovery. Some assay tubes also contained 1 mmoi/L prostaglan­din E" either alone or in combination with w-s to IO-' moi/L epinephrine. In addition, each experiment con­tained one set of tubes that received I0- 5 moi/L pro­pranolol in conjunction with 1 mmoi/L prostaglandin E, and I0- 5 moi/L epinephrine. The tubes were incu­bated 10 minutes at 30° C; the reaction was stopped,

208 Roberts et al.

Table I. Platelet aggregation

Platelet aggregation (mean ± SEM)

January, 1986 Am J Obstet Gynecol

Not citrate adjusted Citrate adjusted

Pregnant I Not pregnant Pregnant I Not pregnant

EC50 of epinephrine (IJ.mol/L) p Value

l.l ± 0.6 (n = 6) 24 ± 14 (n = 6) 2.4 ± 0.36 (n = 5) 10.4 ± 9 (n = 5) >0.1* >0.1*

Epinephrine plus adenosine diphqsphate

0.11 ± 0.03 (n = 5) 0.34 ± 0.08 (n = 5) 0.19 ± 0.09 (n = 5) 0.20 ± 0.08 (n = 5)

p Value <0.011 t >0.40t Adenosine diphosphate

(f.Lmol/L) 1.2 ± 0.55 (n = 6) 1.6 ± 0.7 (n = 4) 3.9 ± 0.9 (n = 6) 2.2 ± 0.6 (n = 5)

p Value >0.20t >0.09t

*The p value was determined by Wilcoxon's rank sums. Value is for a one-tailed test. tThe p value was determined by Student's unpaired t test. Value is for a one-tailed test.

Table II. a-Adrenergic receptor concentration and dissociation constant

Pregnant (mean± SE)

Nonpregnant (mean± SE)

Lysates dihydroergocryptine ~ (nmol/L) Bmax (fmollmg protein)

3.4 ± O.o7 (n = 9) 178 ± 27 (n = 9)

3.0 ± 0.3 (n = 12) 174 ± 20 (n = 12)

Particulate (yohimbine) ~ (nmol/L) Bmax (fmollmg protein) Epinephrine K; (f.Lmol/L)

3.2 ± 2.2 (n = 3) 196 ± 8 (n = 5) 4.3 ± 0.5 (n = 4)

2.6 ± l.l (n = 3) 228 ± 28 (n = 5) 3.3 ± 0.32 (n = 4)

Bmax = Maximal binding. p > 0.05 for all comparisons.

and cyclic adenosine monophosphate labeled with chromic phosphate P 32 was separated from adenosine triphosphate by sequential Dowex and alumina chro­matography. 11 Recovery of cyclic adenosine mono­phosphate was 75% to 85%.

Statisti~al analysis. If variance was similar between groups comparisons were made by unpaired Student's t test. If there was unequal variance a nonparametric comparison (Wilcoxon rank sums) was used. Statisti­cally significant differences were accepted at p < 0.05 for a one-tailed t test, since our hypothesis was that platelet aggregation would be more sensitive to epi­nephrine during pregnancy. Data are presented as the mean± SEM.

Results Aggregation studies. We initially examined the dose

of epinephrine that was needed for half maximal stim­ulation of platelet aggregation (EC50). As has been pre­viously reported, the dose responses were quite steep and it was difficult to accurately determine this param­eter.9· 12 However, results suggested an increased sen­sitivity of platelets of pregnant women to the aggre­gatory effect of epinephrine (EC50 = I ± 0.6 JLmol/L in pregnancy and 24.3 ± 14.7 JLmol/L without preg­nancy; p > O.l). To more accurately and reproducibly determine the sensitivity of platelet aggregation to oc-

cupancy of the a-adrenergic receptor we examined the ability of epinephrine to augment the aggregation to threshold doses of adenosine diphosphate.917 With this response as the end point, full dose-response curves for a-adrenergic receptor agonists can be generated. The EC50 of epinephrine for aggregation studies with platelets from pregnant women (0.11 ± 0.03 ~J.mol/L) was significantly less than that with platelets from nonpregnant women (0.34 ± 0.08 ~J.moi/L, p < 0.02) (Table I).

Interestingly, dose-response curves for adenosine di­phosphate-induced aggregation did not demonstrate a different sensitivity to adenosine diphosphate in preg­nancy (EC50 1.2 ± 0.20 ~J.mol/L in pregnancy and 1.6 ± 0. 7 ~J.mol/L without pregnancy). The differences in epinephrine sensitivity could not be explained by differences in platelet count. The average count for pregnant women (267,000 ± 71,000/mm3

) was, as would be predicted, slightly higher than that for non­pregnant women (228,000 ± 87,000/mm3

). Aggrega­tion differences measured with this technique that were due to different platelet concentrations were not de­tectable at platelet counts > 1 OO,OOO/mm3

•12

Hematocrit was significantly different in the two groups (35.2% ± 2.0% in pregnancy and 40% ± 2.5% without pregnancy). Kelton et aU reported that the sensitivity of the platelet aggregation is artifactually in-

Volume 154 Number 1

Table III. Adenylate cyclase activation

Basal cyclic adenosine monophosphate (pmol/mg/min) Fold increase with prostaglandin E (IJ.mol/L) Inhibition of prostaglandin E increase by epinephrine (%) EC50 of epinephrine (IJ.mol/L)

p > 0.05 for all comparisons.

creased at low hematocrits because of lower citrate con­centrations in platelet-rich plasma. To compensate for differences in citrate concentration in platelet-rich plasma we adjusted the sodium citrate concentration in all samples to that present in blood with a hematocrit of 50 (higher than in any of our subjects) drawn at the usual 1 I 10 citrate dilution. Blood was drawn at the usual citrate dilution and hematocrit was determined. So­dium citrate was then added to achieve a final volume of citrate determined by the formula:

( 1 - hematocrit) Volume of blood X ~-------'-

4.5

volume of citrate added

After normalization of sodium citrate concentration the difference in epinephrine-induced aggregation was no longer present. The EC50 for platelet aggregation by epinephrine alone became more similar (2.4 ± 0.36 ~J.mol/L in pregnancy and 10.6 ± 9.8 ~J.mol!L without pregnancy; p > 0.2). There was no significant differ­ence in the EC,0 of epinephrine to potentiate adeno­sine diphosphate-induced aggregation (0.19 ± 0.09 ~J.mol/L in pregnancy and 0.20 ± 0.08 ~J.mol/L without pregnancy; p = 0.46) (Table 1).

Radioligand binding studies. With dihydroergo­cryptine used as the radioligand there was no differ­ence in a-adrenergic receptor concentration in lysates prepared from platelets of pregnant (178 + 82 fmol/ mg of protein) and nonpregnant (174 ± 46 fmol/mg of protein) women. There was no difference in the Kd of dihydroergocryptine (Table II).

Since dihydroergocryptine binds to sites other than a-adrenergic receptors and nonselectively to both a,­and u2-adrenergic receptors, "·' 5 an increase in a 2-ad­renergic receptors might be masked by a decrease of either a,-adrenergic receptors or other unknown sites. Therefore we also examined the binding of yohimbine, an a-adrenergic receptor antagonist that binds selec­tively and with high affinity to ~-adrenergic recep­tors. 13 In our initial experiments receptor concentration was determined in lysates because we were concerned that recovery of platelet particulate might be influenced by pregnancy. To eliminate the possibility that proteo­lysis might be greater in platelet 1ysates of pregnant women and thus falsely lower receptor concentration, we examined binding of yohimbine to washed platelet

a-Adrenergic receptors and responses during pregnancy 209

Pregnant

7.4 ± 1 (n = 5) 20.8 ± 1.4 (n = 5) 27.6 ± 2.6 (n = 5)

2.7 ± 0.5 (n = 5)

Nonpregnant

6.9 ± 0.8 (n = 5) 25.3 ± 1.7 (n = 5) 30.2 ± 2.6 (n = 5)

1.9 ± 0.4 (n = 5)

particulates. Because preliminary experiments indi­cated no difference in the Kd of yohimbine for platelets during pregnancy (Kd = 3.2 ± 2.2 in pregnancy and 2.6 ± 1.1 nmol/L without pregnancy), we used a sat­urating concentration of yohimbine (24 nmol/L) with I0- 5 mol/L phentolamine to determine nonspecific binding, for estimation of receptor concentration. The estimate of receptor concentration by this method demonstrated no difference between platelets of preg­nant (196 ± 20 fmol/mg of protein) and nonpregnant (228 ± 63 fmol/mg of protein) women (Table II). In addition, experiments examining the ability of epi­nephrine to compete for yohimbine binding to platelet particulation showed that the affinity of epinephrine for the a 2-adrenergic receptor was not different in platelets from pregnant and nonpregnant women (K, 4.3 ± 0.3 ~J.mol/L in pregnancy and 3.3 ± 0.32 ~J.mol/L without pregnancy).

Adenylate cyclase inhibition. To probe the linkage of a 2-adrenergic receptor occupancy to platelet aggre­gation we examined the ability of epinephrine to inhibit adenylate cyclase'5 that has been stimulated by pros­taglandin E in platelet particulate from pregnant and nonpregnant women. Basal and prostaglandin £-stim­ulated adenylate cyclase activity was similar in both groups (Table III). In addition neither the EC,0 of epi­nephrine (2.7 ± 0.5 ~J.mol/L in pregnancy and 1.9 ± 0.4 ~J.mol/L without pregnancy) nor the maxi­mum decrease of adenylate cyclase by epinephrine (27.6% ± 2.6% in pregnancy and 30.2% ± 2.6% with­out pregnancy) was different (Table III). Addition of propranolol did not potentiate the effect of epineph­rine (data not shown).

Comment Our initial examination of platelet aggregation with­

out correction for hematocrit supported our hypothesis that platelet a-adrenergic receptor response and per­haps number of receptors were affected by estrogen as previously described for these receptors in rabbits."· 16

There was an increased sensitivity to epinephrine, de­termined as the EC50 of epinephrine to augment aden­osine diphosphate response, without a generalized in­creased sensitivity since adenosine diphosphate alone induced aggregation with a similar sensitivity in both groups.

210 Roberts et al.

We were therefore surprised to find that the platelet a-adrenergic receptor concentrations of pregnant and nonpregnant women were not different. This was the case with two a-adrenergic radioligands and whether determined in lysates or particulates. There was also no change in the affinity of epinephrine for a 2-adren­ergic receptors.

The linkage of a 2-adrenergic receptor occupancy to subsequent response also did not appear different since adenylate cyclase inhibition by epinephrine was also similar in the two groups. We could not reconcile these similarities in a-adrenergic receptor concentration and function with our original finding of an increase in epinephrine-induced platelet aggregation in preg­nancy. Since platelet function in vitro may be influenced by several factors, we reexamined our assay techniques.

Platelet aggregation in vitro is altered by many subtle variations in methodology. The time from platelet re­moval to assay can alter responses. In all of our ex­periments we carefully controlled time from drawing of blood to assay and the order of experiments. Platelet count affects platelet aggregation but only at platelet concentrations substantially less than were present in these subjects. 12

Kelton et aF have pointed out the importance of adjusting the sodium citrate concentration in platelet­rich plasma to eliminate differences brought about by use of a fixed concentration of sodium citrate with blood samples with different hematocrits. Sodium ci­trate does not enter cells, thus lower hematocrit results in a lower concentration of sodium citrate in platelet­rich plasma. Since sodium citrate is a calcium chelator, platelet-rich plasma from blood samples with lower he­matocrits (and higher free calcium concentrations), as in our samples from pregnant women, will have en­hanced aggregation. Interestingly, in the study of Kel­ton et al. as in our own, the influence is greater on epinephrine than on adenosine diphosphate-induced aggregation. We found that the elimination of differ­ences in sodium citrate concentration eliminated dif­ferences of epinephrine sensitivity of platelets. Thus similar a-adrenergic receptor concentrations, agonist affinity, and coupling to adenylate cyclase in the two groups are compatible with the similar sensitivity to epinephrine when the confounding variable of a dif­ferent sodium citrate concentration is controlled.

This variable was not controlled in any of the pre­vious comparisons of platelet aggregation in pregnant and nonpregnant women;-s which accounts in part for

January, 1986 Am J Obstet Gynecol

the conflicting results. Kelton et aJ.7 pointed out that many reported differences in platelet aggregation be­tween sexes and with disease are invalidated by failure to correct for different hematocrits. This confounding variable must be taken into account when the differ­ences in platelet aggregation are reported in pathologic pregnancies.

REFERENCES I. Bonnar J, McNichol GP, Douglas AS. Coagulation and

fibrinolytic mechanisms during and after normal child­birth. Br Med J 1970;2:200.

2. Sagi A, Creter D, Goldman], Djaldetri M. Platelet func­tions before, during and after labor. Acta Haematol 1981;65:67.

3. Briel RC, Lippert T. Platelet sensitivity to prostacyclin in pregnancy and uterine cancer. Eur J Obstet Gynaecol Re­prod Bioll981;12:19.

4. LeComite F. Influences of pregnancy and oral contracep­tives on platelets in relation to coagulation and aggrega­tion. Thromb Diath Haemorrh 1973;29:510.

5. Zahn M. Platelet function in pregnancy. Clin Res 1969;17:603.

6. Vargaftig BB, Chignard M, Benveniste J. Present concepts on the mechanisms of platelet aggregation. Biochem Pharmacol 1981 ;30:263.

7. Kelton G, Powersk P, Julian J, et al. Sex-related differ­ences in platelet aggregation-influence of the hemato­crit. Blood 1980;56:38.

8. RobertsJM, Goldfien R, Tsuchiya A, Goldfien A, Insel P. Estrogen treatment decreases alpha adrenergic binding sites on rabbit platelets. Endocrinology 1979;104:722.

9. Murlas C, Nadel J, Roberts JM. The muscarinic cholin­ergic receptors of airway smooth muscle. J Appl Physiol 1982;52:1084.

10. Cheng Y, Prusoff WH. Relationship between the inhibi­tion constant (K1) and the concentration of inhibitor which causes 50 percent inhibition (1-50) of an enzymatic reac­tion. Biochem Pharmacol 1973;22:3099.

II. Salomon Y, Londos C, Rodbell M. A highly sensitive ad­enylate cyclase assay. Anal Biochem 1974;58:541.

12. Roper P, Drewwinko B, Hasler D, Johnston D, Hester J, Freirlich E. Effects of time platelet concentration and sex on the human platelet aggregation response. Am J Clin Pathol 1979;71 :263.

13. Hoffman BB, DeLean A, Wood CL, Schocken D, Lef­kowitz R. Alpha-adrenergic receptors subtypes-quanti­tative assessment by ligand binding. Life Sci 1979;24: 1739.

14. Cronin MJ, RobertsJM, Weiner R. Dopamine and dihy­droergocryptine binding to the anterior pituitary and other brain areas of the rat and sheep. Endocrinology 1978;103:302.

15. Jakobs KH, Saw W, Schultz G. Reduction of adenylate cyclase activity in lysates of human platelets by the alpha­adrenergic component of epinephrine. J Cyclic Nucleo­tide Res 1976;2:381.

16. Roberts JM, Insel P, Goldfien R, Goldfien A. u2 Adre­noreceptors but not 13 adrenoreceptors increase in rabbit uterus with estrogen. Nature 1977;270:624.