Main

IVH occurs primarily, but not exclusively, in premature infants. The bleeding originates in the microvasculature of the subependymal germinal matrix and can extend into the lateral ventricles or the brain parenchyma(1, 2). The etiology is incompletely understood but is probably multifactorial. IVH is associated with asphyxia, acidosis, hypoxia and fluctuations in blood pressure(1–4), and the decreasing incidence of IVH is likely related to improved neonatal care addressing some of the factors which contribute to abnormal cerebral blood flow. Other contributing factors have been implicated, including fragility of the capillaries in the germinal matrix and impaired hemostasis. Therapeutic manoeuvres designed to improve hemostasis by increasing levels of plasma clotting factors, increasing platelet counts or inhibiting fibrinolysis have met with variable success to date(5–9). The role that impaired platelet function may play in this context has not been evaluated as information about platelet function in premature infants is limited.

Platelets of full-term neonates are known to be less responsive to physiologic agonists than adult platelets under similar in vitro conditions(14, 15). Neonatal cord blood platelets demonstrate decreased aggregation, secretion and Tx production in response to collagen, epinephrine, and U46619 (a stable analog of TxA2)(14–17). Similarly, these platelets express fewer adhesive receptors, P-selectin and GPIIb/IIIa, on the plasma membrane after activation with thrombin, U46619, and the combination of ADP and epinephrine(18). Corresponding studies on preterm neonatal platelets have not been carried out, although preliminary reports have indicated a more significant abnormality in platelets of premature infants(19, 20).

In the present study, aggregation, secretion, and Tx production in cord blood platelets from preterm infants (<35-wk gestation) have been examined. Furthermore the differences identified in response to U46619, initiated further investigation of TxA2 receptor characteristics in platelets of full-term and preterm neonates.

METHODS

Subjects. Studies were performed on cord venous blood from 35 preterm infants with a mean gestation of 32 ± 3.2 wk, and weight of 1466 ± 1007 g. There were 19 boys, 16 girls, and three set of twins, and in 13 cases delivery was by means of cesarean section. Gestational age was confirmed either by ultrasonography or Dubowitz score within 48 h of delivery. Mothers had no history of bleeding diathesis, alcohol or acetylsalicylic acid ingestion for 2 wk before delivery. Two mothers had received hydralazine, and two had received steroids before delivery. Cord blood platelet counts were 304± 141 × 109/L. Subgroups of this population were used for some studies. Control platelets were obtained from 14 healthy nonsmoking adults with no history of drug or alcohol ingestion. Platelet counts in the adult controls were 381 ± 121 × 109/L. Cord blood platelets from 12 full-term neonates (gestational age ≥ 37 wk; platelet counts 283 ± 58 × 109/L) were also used for comparison in some studies.

Blood collection and preparation. Blood was collected from the umbilical vein immediately after clamping of the cord. 40-60 mL of cord blood were drawn into plastic syringes containing ACD anticoagulant (1.9:8.1; ACD:blood; v:v). Blood from adult controls was obtained by venipuncture and drawn into ACD (1.9:8.1; ACD:blood; v:v). Platelet-rich plasma was prepared by centrifugation of the citrated whole blood at 800 × g for 5 min at 20°C. Platelet counts were carried out on a Coulter counter (model JT2, Coulter Electronics, Hialeah, FL), and the platelets were then washed and resuspended in buffer before study(21). All experiments were conducted within 2 h of blood collection.

Aggregation studies. Platelets were suspended in HEPES-Tyrode's buffer (134 mM NaCl, 12 mM NaHCO3, 2.9 mM KCl, 0.34 mM Na2HPO4, 1 mM MgCl2, 10 mM HEPES, 5 mM dextrose, 0.3% BSA, pH 7.4) at a final concentration of 2 × 108/mL. Aggregation was studied using a Payton dual channel aggregometer (Payton Assoc., Scarborough, ON) at 37°C under stirred conditions. CaCl2 (1 mM) was added to each sample 2 min before the addition of agonist. The agonists used were α-thrombin (1 U/mL), type 1 bovine collagen (2 μg/mL and 10μg/mL), ADP (10 μM), and U46619 (0.1-1 μM). Fibrinogen (1 mg/mL) was added to samples before addition of ADP. Aggregation was monitored for 5 min.

Secretion of serotonin from platelet dense granules. Platelet-rich plasma was preincubated with [14C]serotonin (1.0μCi/mL for 30 min) by the method of Jerushalmy and Zucker(22). The platelets were then washed and resuspended in HEPES-Tyrode's buffer, and samples were incubated in the aggregometer with agonists, as described above. Aggregation was allowed to proceed for 2 min, and the release was terminated by the addition of an equal volume of 0.1% glutaraldehyde. The samples were centrifuged at 2000 × g for 10 min to pellet the platelets, [14C]serotonin content of pellet and releasate was determined, and the percent release was calculated.

TxB2 formation. The production of TxB2, the stable metabolite of TxA2, was quantified by ELISA using specific rabbit antibody to TxB2 as previously described(23, 24). Briefly, after aggregation of washed platelets, as outlined above, the platelets were pelleted, and the supernatant was removed and diluted in PBS-BSA. Aliquots plus anti-TxB2 were added to wells of microtiter plates, previously coated with TxB2-BSA conjugate, and allowed to react for 18 h. The plates were washed, and goat anti-rabbit IgG-alkaline phosphatase conjugate was added. After incubation, the plates were washed again, phosphatase substrate was added, and the color was allowed to develop. The absorbance at 405 nm was read for each well using a microtiter plate reader (Titertek Multiscan, Mississauga, ON).

Flow cytometry analysis of TxA2 receptor. The TxA2 receptor was evaluated by flow cytometry as previously described(16). Briefly, washed platelets were resuspended in NH4Cl-Tris-buffered saline (pH 7.4) and fixed with an equal volume of 2% paraformaldehyde in 0.2 M cacodylate for 1 h. The fixed platelets were incubated with 25 μg/mL of a rabbit polyclonal anti-TxA2 receptor antibody, anti-P2 (a gift from Dr. G. C. Le Breton, Chicago, IL)(25), for 2 h at 25°C. Binding was detected using a biotinylated goat anti-rabbit second antibody (Hyclone Labs, Logan, UT) and fluoresceinavidin (Dimension Lab Inc. Mississauga, ON). Control studies were performed using nonimmune rabbit serum or buffer in place of the primary antibody. Platelet-associated fluorescence was quantified using an EPICS V flow cytometer (Coulter, Hialeah, FL) equipped with an Argon-Ion laser (400 mW, 488 nm). Fluorescence was detected at 525 nm. Platelets were identified by forward angle and orthogonal light scatter. Single parameter, 255-channel log integral green fluorescent histograms were obtained, and the negative and positive fluorescent populations were resolved by comparison with negative controls.

Immunoblotting of TxA2 receptor. Immunoblots were performed as previously described(26, 27). Briefly, platelet proteins were solubilized, and aliquots, each containing 20μg of protein, were separated by 10% SDS-PAGE and transferred to nitrocellulose. The nitrocellulose was blocked with 10% skimmed milk for 1 h at 25°C and incubated with anti-P2 overnight at 4°C. The nitrocellulose was washed with Tris-buffered saline/Tween, incubated with peroxidase-conjugated goat anti-rabbit IgG, and visualized by enhanced chemiluminescence (Amersham, Oakville, ON).

Ligand binding to TxA2 receptor. Equilibrium binding of[3H]U46619 (Dupont NEN, Boston, MA) to washed platelets was performed essentially as described by Kattelman et al.(28). Platelet-rich plasma was incubated with 1 mM acetylsalicylic acid for at least 30 min at 37°C to inhibit endogenous TxA2 formation, then centrifuged at 1100 × g for 10 min, and the platelets were resuspended in HEPES/Tyrode's buffer, pH 7.4(described above), at a final concentration of 3 × 108 platelets/mL. Aliquots of platelet suspension were incubated with 4 nM[3H]U46619 plus appropriate concentrations (0.15 pM to 1 μM) of unlabeled U46619 for 30 min at 37°C. Nonspecific binding was determined using incubations containing a 1000-fold excess of unlabeled U46619. Specific binding was defined as the difference between binding in the absence and presence of excess unlabeled ligand. The reaction was stopped by centrifugation at 7000 × g for 3 min, and the pellet was washed twice and then dissolved in 0.1 mL of formic acid for 30 min at 37°C, before transfer to a scintillation vial with 5 mL of Ecolite scintillation cocktail. Duplicate samples were counted in a Beckman, model LS 9800, liquid scintillation counter (Beckman Instrument Inc., Fullerton, CA). Binding analysis was performed using the computer programs EBDA and LIGAND.

Statistical analysis. Means, standard deviations, standard errors, and tests of significance (t test or one-way ANOVA) were calculated with the computer program Microsoft Excel.

RESULTS

Platelet aggregation. Platelet aggregation in response to several agonists was compared in 35 preterm infants and 10 adult control subjects (Fig. 1A). Aggregation in response to collagen, ADP, thrombin, and U46619 were all significantly decreased in the preterm platelets when compared with those of adult controls (p < 0.05). Comparison of aggregation responses to low concentrations (<1 μM) of U46619 showed that full-term platelets were more responsive than preterm platelets, although less responsive than those of adult controls (Fig. 2).

Figure 1
figure 1

(A) Percent maximum aggregation of washed platelet suspensions in response to 1 U/mL thrombin, 10 μM ADP, 2 μg/mL collagen, and 1 μM U46619. Each column represents the mean ± 1 SD of 35 preterm infants or 10 adult control subjects. The asterisk identifies a difference between the means significant at p < 0.05.(B) Percent secretion of [14C]serotonin from washed platelet suspensions in response to 1 U/mL thrombin, 2 μg/mL collagen, and 1 μM U46619. Each column represents the mean ± 1 SD of 35 preterm infants or 10 adult control subjects. The asterisk identifies a difference between the means significant at p < 0.01.

Full size image
Figure 2
figure 2

Aggregation patterns for washed platelets from adults(A) and full-term (F) and preterm (P) infants in response to low concentrations of U46619 (0.1-1.0 μM). This figure is representative of five similar experiments.

Full size image

Serotonin secretion. The release of [14C]serotonin was studied in the same 35 preterm infants and 10 adult control subjects described above for aggregation studies (Fig. 1B). Secretion from preterm platelets was significantly decreased in response to collagen and U46619 (p < 0.01). In contrast, comparison of serotonin secretion in full-term and preterm platelets in response to 1 μM U46619 showed no difference (25 ± 7% versus 24 ± 5%).

TxB2 formation. Quantification of TxB2 production in response to thrombin, collagen, and U46619 was performed on 22 preterm infants and 14 adult controls (Table 1). There was no significant difference in the amount of TxB2 produced in response to thrombin (1U/mL). In contrast, the preterm platelets produced significantly less TxB2 in response to collagen (10 μg/mL) than the adult controls. Preterm platelets also synthesized less TxB2 in response to collagen than that previously reported for full-term neonatal platelets (30.2± 5.5 ng/mL versus 52.7 ± 12.6 ng/mL; p < 0.05)(16).

Table 1 TxB2 production by preterm and adult platelets
Full size table

Flow cytometry analysis of TxA2 receptor. The binding of the anti-TxA2 receptor antibody, anti-P2, was studied in five preterm and five full-term infants and five adult control subjects (Fig. 3). The mean channel fluorescence was not different among the three groups: preterm (84 ± 21), full-term (70 ± 16), adult (69 ± 16). Similarly, the percent positive cells did not differ among the groups and ranged from 58 to 88%.

Figure 3
figure 3

Flow cytometric analysis of anti-TxA2 receptor antibody (anti-P2) binding to adult (thin line), full-term (dotted line) and preterm platelets (thick line). Representative fluorescence histogram recorded after incubation with primary antibody, biotinylated second antibody, and fluorescein avidin. A negative control with buffer replacing the primary antibody is also shown. The vertical line separates negative and positive fluorescent populations. For each sample 1 × 104 platelets were counted. Percent positive cells: adult (83%), full-term (80%), and preterm (87%).

Full size image

TxA2 receptor immunoblotting. Immunoblot analysis was carried out on platelet membranes isolated from preterm and full-term infants and adults, using anti-P2. The antibody detected a single band of 55 kD, as previously reported by Borg et al.(25), which appeared identical in both size and quantity in all three groups (Fig. 4).

Figure 4
figure 4

Immunoblot of TxA2 receptor using anti-P2 antibody. Aliquots (20 μg) of solubilized platelet protein from preterm(A), full-term (B), and adult (C) platelets were separated on a 10% SDS-PAGE gel and transferred to nitrocellulose. The TxA2 receptor was detected as a single 55-kD band.

Full size image

Radioligand binding assays. Ligand binding studies were performed on platelet membrane preparations from 10 preterm and 10 full-term infants and 10 adult control subjects using [3H]U46619. Scatchard analysis yielded a single class of binding sites, with Kd andBmax values within the previously reported range(28, 29) (Table 2). Although there was a suggestion of fewer binding sites on preterm platelets, one-way ANOVA did not identify significant differences among the groups. There was also no significant difference among the Kd values of the three groups.

Table 2 Platelet TxA2 receptor characteristics
Full size table

DISCUSSION

Hemostasis in premature infants appears to be more significantly compromised than that in full-term neonates. Premature infants have lower concentrations of coagulation proteins(30, 31), poor regulation of fibrinolysis(32), and previous preliminary reports suggest a significant impairment of platelet activation(19, 20). These factors contribute to an increased risk of bleeding, in general, and perhaps to IVH in particular.

Previous studies have demonstrated hyporeactivity in the platelets of full-term newborns to a variety of physiologic agents. This is manifest as decreased aggregation and secretion(10–13, 16, 17), decreased synthesis of Tx(14–17), decreased calcium mobilization(16, 33), and decreased expression of activation markers on the platelet surface(18). All of these observations point to deficiencies in signal transduction pathways in the neonatal platelet. This is not unique to the platelet, as many activation and metabolic pathways are immature in the neonate. For example, the generation of thrombin is an activation pathway affected at several points resulting in delayed, and decreased thrombin activity(30). It is possible that the signal transduction pathways involved in neonatal platelet activation are affected at multiple points and that these deficiencies are more marked in the platelets of premature infants. The results of the present studies support this suggestion.

Aggregation and secretion responses of preterm platelets to type 1 collagen were markedly decreased compared with those of adult control subjects, similar to previous observations in full-term neonatal platelets(16). However, in contrast to full-term neonates, the responses to thrombin, ADP, and U46619 in preterm platelets were also significantly decreased in comparison to those of adult controls (Fig. 1). Collagen-induced Tx production by preterm platelets was decreased compared with that of adult controls(Table 1), and this was more marked than that previously reported for full-term platelets(16). In full-term platelets, this effect on Tx production has been attributed to decreased cyclooxygenase activity(14, 15). As collagen-induced platelet activation is dependent on endogenously generated TxA2(34), poor TxA2 production in neonatal platelets would contribute significantly to the poor response to collagen.

The decreased responsiveness of preterm platelets to agonists other than collagen clearly indicates that the hyporeactivity of preterm platelets may be heterologous in nature and results from mechanisms in addition to defective TxA2 synthesis. Of particular significance are the effects of the stable Tx mimetic, U46619. Activated platelets synthesize and release TxA2, which is important for platelet recruitment to the forming aggregate in vivo. TxA2 (and U46619) bind to a specific G protein-coupled receptor on the platelet membrane, which subsequently activates the phospholipase C cascade(35). As preterm platelets aggregated poorly in response to U46619 it was important to determine the etiology of this defect. Significantly impaired signal transduction through the TxA2 receptor has previously been identified in rare patients with hemostatic defects(36–38).

The TxA2 receptor in adult, full-term, and preterm platelets was examined both by use of an anti-TxA2 receptor antibody(anti-P2), and by determining the binding characteristics of[3H]U46619. Immunoblotting with anti-P2 identified similar levels of an identical 55-kD protein band, and flow cytometry produced similar profiles, in all three groups of platelets (Figs. 3 and 4). These data reflect similar levels of anti-P2 antibody-binding sites per cell in each group of platelets.

Ligand binding studies were also performed to evaluate the number and affinity of the TxA2 receptors on the platelets. It is known that TxA2 receptor numbers can vary under certain conditions, including up-regulation by testosterone(39), and after myocardial ischemia(40, 41). TxA2 receptor gene transcription can be modulated by treatment of megakaryocyte cell lines with phorbol esters(42). It was therefore possible that TxA2 receptor expression might differ in neonatal platelets because of factors affecting regulation of gene transcription in neonatal megakaryocytes. However, no significant differences were found in either the binding affinity or number of binding sites on preterm, full-term or adult platelets, using[3H]U46619 (Table 2).

As platelets of preterm infants demonstrate normal binding characteristics of the TxA2 receptor itself, the defect in U46619 response may lie in postreceptor signal transduction. The receptor is linked to the phospholipase C cascade which produces the important second messengers, inositol trisphosphate and diacylglycerol. These, in turn, stimulate the release of intracellular calcium stores and activate protein kinase C, required for platelet activation. Decreased activity of any component of the postreceptor signal transduction pathway could result in the unresponsiveness to TxA2 observed in the platelets of preterm infants.

The preterm platelets also demonstrated impaired aggregation in response to both thrombin and ADP. Thrombin shares many of the postreceptor mechanisms with TxA2, including phospholipase C activation and the associated intracellular events. In contrast, in human adults, ADP-induced platelet aggregation is associated with the elevation of intracellular calcium in the absence of phospholipase C activity(43). Therefore it is possible that the deficient response is due to an event common to TxA2-, ADP-, and thrombin-induced activation, such as those events regulated by an elevation in the cytosolic free calcium levels(33).

In summary cord blood platelets from preterm infants show impaired aggregation and secretion responses to agonists including thrombin, collagen, ADP, and TxA2 analogs. They produce less TxA2 in response to collagen than either adult or full-term platelets. The TxA2 receptor binding characteristics (affinity and number of sites) are comparable to adult values in both full-term and preterm platelets. The poor response to TxA2 analogs may therefore be due to TxA2 receptor-coupled events. It is likely that the hyporesponsiveness of the premature platelet to agonists, such as collagen and low dose thrombin, is significantly influenced by the defects in both TxA2 synthesis and response.