Keywords premature - full-term neonates - platelet hypo-responsiveness - ITAM-containing receptors
- development - immune-induced thrombocytopenia
Introduction
Platelets are anuclear haematopoietic cells that play an essential role in haemostasis
and its pathological counterpart thrombosis. Platelets also play critical roles in
other physiological and pathological processes including inflammation,[1 ] infection,[2 ] vascular integrity,[3 ] development[4 ] and cancer metastasis.[5 ] Currently, we have a rudimentary understanding of the role of platelets in many
of these functions and how these processes vary during development and throughout
adulthood.
Thrombopoiesis takes place in multiple sites during development, beginning in the
yolk sac before moving to the liver and finally to the bone marrow and spleen.[6 ]
[7 ]
[8 ] This means that, throughout development, circulating platelets are derived from
more than one haematopoietic site. It is presently unclear to what extent this influences
platelet function.
It is widely acknowledged that platelet reactivity is reduced in neonates, but the
molecular basis underlying this is not known; the degree of hypo-reactivity and the
extent to which this varies between agonists is also uncertain.[9 ]
[10 ]
[11 ]
[12 ] This is partly due to methodological issues related to the low volumes of blood
available for experimentation, with many studies limited to single concentrations
or small panels of agonists.[9 ]
[10 ]
[11 ]
[12 ] One consistent feature is a marked reduction in responsiveness to collagen.[13 ]
[14 ] Collagen activates Src and Syk tyrosine kinases downstream of the glycoprotein VI
(GPVI)-Fc receptor γ-chain (FcRγ) complex, culminating in activation of PLCγ2.[15 ] Collagen also binds to integrin α2β1 which supports adhesion and net binding to
GPVI.[15 ] However, no difference in α2β1 expression between neonatal and adult platelets has
been reported,[10 ]
[13 ] with the reduced response to collagen being attributed to a reduction in phosphoinositide
hydrolysis and Ca2+ mobilization suggesting a defect in collagen signalling.[13 ]
[16 ]
There are also reports of impaired responses to G protein-coupled receptor (GPCR)
agonists, although this has not been seen in all cases, and several mechanisms have
been proposed. The reduced response to adrenaline and thrombin has been attributed
to decreased receptor expression, and in the case of thromboxane A2 mimetic U46619, to defective G protein-coupled activity.[10 ]
[17 ]
[18 ]
Platelets from pre-term infants are also hypo-reactive in comparison to their full-term
counterparts.[14 ]
[19 ] In the western world, more than 10% of babies are pre-term[20 ] and this population displays the highest incidence of intra-ventricular haemorrhage
(IVH). IVH affects up to 25% of infants born with weights of less than 1,500 g and
usually occurs in the first week of life.[11 ] The increase in risk of bleeding coinciding with the time of marked platelet hypo-reactivity
raises the question of whether this contributes to the increase in IVH.[21 ]
[22 ]
In this study, we have assessed the reactivity of human platelets from pre- and full-term
neonates and that of mice platelets during late gestation and in neonates to collagen-related
peptide (CRP) and to the snake venom toxin rhodocytin which activate GPVI and C-type
lectin-like receptor 2 (CLEC-2), respectively.[23 ] Since CLEC-2 signals through a similar pathway to GPVI, a decrease in responsiveness
to CLEC-2 could reflect a reduction in immunotyrosine-based-activation-motif (ITAM)
signalling, rather than a specific loss of response to GPVI. Improving our understanding
of the underlying mechanisms of platelet hypo-reactivity in foetal and neonatal life,
will help to define the contribution of platelet hypo-reactivity in neonatal bleeding
disorders and guide clinical decision making and management of haemostasis and thrombosis
complications especially in the case of pre-term neonates.
Materials and Methods
Materials
The platelet glycoprotein screen assay kit for analysis of human GPIIIa (β3-integrin
subunit), GPIbα and GPIa and the platelet calibrator kit for analysis of GPVI and
CLEC-2 were from Biocytex (Marseille, France). Antibodies for the platelet calibrator
kit were as follows: GPVI-mAb 1G5[24 ]; α-CLEC-2-mAb AYP1[24 ]; α-CD41a*APC; and α-CD62*PE was from BD Biosciences (Madrid, Spain); and Fibrinogen*Alexa
Fluor 488 was from Life Technologies (Madrid, Spain). Protease-activated receptor
1 (PAR-1) peptide (TFLLR) and phorbol 12-myristate 13-acetate (PMA) were from Sigma-Aldrich
(Madrid, Spain), and Human PAR-4 peptide (AYPGKF) was from Alta Biosciences (Birmingham,
UK). Integrilin (eptifibatide) was from Glaxosmithkline (Middlesex, UK). Sodium dodecyl
sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) gels, polyvinylidene difluoride
(PVDF) membranes, peroxidase conjugated secondary antibodies and enhanced chemiluminescence
(ECL) mix were from GE Healthcare (Madrid, Spain). α-PLCγ2 (B-10), α-Syk (4D10) and
anti-Fcγ chain (sc-390222, FcεRIγ [E-12]) were from Santa Cruz (Heidelberg, Germany).
α-PLCγ2-Y1217 and α-Syk-525/526 were from Cell Signaling Technology (Leiden, The Netherlands).
α-βActin was from Sigma-Aldrich and α-GPVI was from Abcam (Cambridge, UK). Anti-αIIb
antibody 132.1 was a gift from Z.M. Ruggeri (Scripps Clinic, La Jolla, California,
United States). For assays in mice, the antibodies were as follows: The mouse CLEC-2
monoclonal antibody (mAb) 17D9 and α-IgG2b*FITC were purchased from Bio-Rad (Hemel
Hempstead, UK). α-P-Selectin*PE was from Novus Biologicals (Abingdon, UK). α-CD41*APC
and α-CD41*PE were from eBioscience (Hatfield, UK) and BD Pharmigen (Oxford, UK),
respectively. α-CD41*FITC, α-CD42b*FITC, α-CD49b*FITC, α-GPVI*FITC and α-IgG*FITC
were from Emfret (Eibelstadt, Germany). α- GPIbα and immunoglobulin G (IgG) control used for immune depletion were also from
Emfret. α-CLEC-2*FITC was from AbD Serotec (Kidlington, UK). Goat α-Rat*488 secondary
antibody was from Fischer Scientific (Loughborough, UK). Other reagents were from
recognized suppliers and of the highest analytical grade commercially available.
Human Blood Collection
Cord blood (CB) samples were collected from healthy pre-term (26–34 gestational weeks)
and full-term neonates (≥ 37 gestational weeks), admitted to the maternal-fetal unit
of the University Hospital Virgen de la Arrixaca, Murcia, Spain. All neonates were
born from uncomplicated pregnancies and had a normal platelet count (> 150,000 platelets/μL).
Neonates were excluded from the study if mothers had a history of diabetes, hypertension,
pre-eclampsia, active infection, drug or alcohol abuse, had taken aspirin during the
10 days prior to the delivery or there was a family history of abnormal haemostasis
or any congenital disorder. Peripheral blood (PB) samples were collected from the
antecubital veins of healthy adult volunteers who had not taken any medications during
the 10 days prior to the study. All samples (CB and PB) were drawn in to 3.2% buffered
sodium citrate tubes (Vacutainer System; Diagnostica Stago Becton Dickinson, Plymouth,
UK). The study on humans was approved by the Ethics Committee of the Arrixaca University
Hospital and followed the Helsinki Declaration. All adult volunteers and the parents
of neonates provided written informed consent.
Murine Blood Collection
Mouse embryos were dissected from pregnant females after maternal cervical dislocation,
before being decapitated and allowed to bleed into 10 units/mL of heparin in phosphate-buffered
saline (PBS). Neonatal mice were culled via intra-peritoneal injection of Euthatal
(50–100 µL) before being decapitated and allowed to bleed into 10 units/mL of heparin.
Where stated, adult mice were culled as for neonates to enable comparison of results
of platelets prepared in the same way. Otherwise, adult mice were anaesthetised with
isoflurane before CO2 narcosis, descending vena cava isolation and subsequent venepuncture collection.
For repeat blood collection from immune-induced thrombocytopenic mice, the animals
were restrained and blood was withdrawn from the tail vein via needle-prick into 10
units/mL heparin; terminal bleeds were performed under isoflurane/CO2 as described above. All animal work was performed with U.K. Home Office approval
under license PPL70/8286.
Receptors Levels and Platelet Activation in Human Platelets
Platelet receptor expression in adult, pre-term and full-term blood samples was assessed
by flow cytometry using a BD Accuri C6 flow cytometer device (Ann Arbor, Michigan,
United States). The GP Screen and Platelet Calibrator kits (Biocytex), with the appropriate
antibodies as stated above, were used following the manufacturer's instructions.
Adult blood and CB samples from both pre-term and full-term neonates were also assessed
for agonist-induced surface P-selectin exposure and fluorescent fibrinogen binding
to αIIbβ3 by flow cytometry using Accuri C6. Briefly, diluted blood was incubated
under static conditions (30 minutes at room temperature) with PBS, as control for
non-stimulated platelets, PAR-1 peptide (25 µM), PAR-4 peptide (250 µM), adenosine
diphosphate (ADP) (25 µM), PMA (100 nM), CRP (5 µg/mL) and rhodocytin (100 nM) in
the presence of α-CD41a*APC, as platelet marker antibody, α-CD62*PE and fibrinogen*Alexa
Fluor 488. Reactions were terminated by 4% paraformaldehyde (PFA) (v/v) and subsequent
10-fold dilution with PBS. For each sample, run up to 10,000 platelets were identified
and analysed by gating events on both CD41a*APC positivity and forward scatter-side
scatter (FSC-SSC). Results were expressed as percentage of positively stained cell
for P-selectin or fibrinogen, as compared with non-stimulated cells.
Assessment of GPVI and CLEC-2 Signalling Pathway in Human Platelets by Immunoblotting
Human adult and full-term neonates' washed platelets in modified Tyrode's buffer were
obtained as previously described.[25 ] For protein phosphorylation studies, washed platelets (6 × 108 /mL) were stimulated under stirring conditions for 180 seconds with CRP (1–10 µg/mL)
or rhodocytin (30–100 nM), using an aggregometer set at 37°C (Aggrecorder II Menarini
Diagnostics, Florence, Italy). Eptifibatide (1 µM) was included to prevent platelet
aggregation and stimulation was stopped by addition of SDS reducing sample buffer.
Proteins in whole cell lysates were separated by SDS-PAGE and transferred to PVDF
membranes by means of semi-dry transfer units. Blots were stepwise incubated with
appropriate primary antibodies and as peroxidase-conjugated secondary antibodies (see
the “Materials” section), and proteins were detected by chemiluminescence.
Gene Expression Analysis of GP6 , CLEC1B , SYK and PLCG2 in Human Platelets
Total ribonucleic acid (RNA) was isolated from ultrapure platelets as we have recently
reported in detail.[26 ] Retrotranscription reaction was performed using 35.2 ng of total RNA, according
to the manufacturer's instructions (SuperScript III First Strand, Thermo Fisher Scientific).
Gene expression was quantified on a LC480 real-time polymerase chain reaction (PCR)
system (Roche Pharma, Basel, Switzerland) using Taqman Premix Ex Taq (Takara Bio Inc.)
and a commercial probe for GP6 (Hs_00212574), FCER1G (Hs_Hs00175408), CLEC1B (Hs_00212925), SYK (Hs_00895377), PLCG2 (Hs01101857_m1) and ACTB (Hs_99999903).
Murine Platelet Receptor Expression and Platelet Activation
To evaluate murine platelet receptor expression, diluted whole blood (10–20 × 109 platelets/L) was stained (30 minutes at room temperature) with both α-CD41*APC and
a fluorescein isothiocyanate (FITC)-labelled antibody (see above) specific to the
target receptor: GPIbα, αIIbβ3, α2β1, GPVI and CLEC-2. Reactions were terminated with
4% PFA solution and 10-fold PBS dilution as above, and platelets acquired and analysed
with the Accuri C6 software. Results were expressed as mean fluorescence of positive
cells. Agonist-induced platelet activation in mice at known pre- and post-natal ages
was evaluated, as in the human studies above, by flow cytometric analysis of fibrinogen-binding
and P-selectin exposure in PBS-diluted murine blood (10–20 × 109 platelets/L). Agonists used in these assays include PBS, as control for non-stimulated
platelets, PAR-4 peptide (50, 100 and 250 µM), CRP (1, 5 and 10 µg/mL) and rhodocytin
(10, 30 and 100 nM). Results were expressed as percentage of murine platelets positively
stained for P-selectin or fibrinogen, as compared with non-stimulated cells.
Murine Recovery from Immune-Mediated Thrombocytopenia
Immune thrombocytopenia was induced in adult mice by injection of 1.5 µg/g α-GPIbα,
which caused a sharp and rapid decline in platelet count.[27 ] Control mice were injected with IgG control. Blood samples were collected before
injection and daily post-injection for analysis of platelet receptor expression and
dose–response assays as described above.
Statistical Analysis
All statistical analysis was performed using GraphPad Prism V7.00 (California, United
States). Differences between neonates and adults in platelet receptor levels, gene
expression levels or platelet functional responses were assessed by t- test or Mann–Whitney U test as appropriated. For murine receptor studies, an ordinary
one-way analysis of variance (ANOVA) was performed with a post hoc Dunnet's multiple
comparisons test. For murine neonatal and post-thrombocytopenia functional assays,
a two-way ANOVA was performed with a Dunnet's multiple comparisons post hoc test.
For murine post-thrombocytopenia receptor assays, a two-way ANOVA was performed with
a Sidak's multiple comparisons post hoc test. Significance was assumed with a p -value of ≤ 0.05 (*) or ≤ 0.005 (**).
Results
Platelets from Pre-Term and Full-Term Neonates Display Reduced Expression and Function
of GPVI and CLEC-2
We explored the mean platelet volume and expression levels of human platelet receptors
in pre-term neonates (median gestational age of 32.3 [range, 29.8–34.4] weeks), full-term
neonates and adults. The mean platelet volume was similar in all three groups ([Supplementary Fig. S1A ], available in the online version). In contrast, there was a significant (30–35%)
reduction in the levels of GPVI and CLEC-2 in pre-term and full-term neonates versus
adults as shown by flow cytometry ([Fig. 1A ]) and by immunoblotting ([Fig. 1B ]). In agreement with the above findings, quantitative reverse transcription PCR (qRT-PCR)
experiments showed that neonatal platelets displayed a significant reduction in GP6 (60%) and a mild decrease in CLEC1B (30%) messenger RNA (mRNA) levels, compared with adult platelets ([Supplementary Fig. S2A ], [B ], available in the online version). In line with the decrease in GPVI, the expression
of Fcγ chain, which associated with GPVI to form a functional collagen immunoreceptor,[28 ]
[29 ] was also reduced at protein and mRNA level in neonatal platelets ([Supplementary Fig. S3A–C ], available in the online version). In addition, and in concordance with previous
studies,[10 ] pre-term and full-term neonates displayed a mild reduction in expression of integrin
αIIbβ3, but no significant change in the levels of integrin α2β1 and GPIbα ([Fig. 1A ]).
Fig. 1 Flow cytometric assessment of receptor expression and function in platelets from
premature, full-term neonates and adults. Platelet receptor expression was measured
by (A ) flow cytometry using Biocytex kits in human pre-term (white bars; n = 5) and full-term (red bars; n = 8–11) neonates and in adults (blue bars; n = 8–11), and by (B ) western blotting, as described in the “Material and Methods” section. Western blots
images are representative of assays in platelet lysates from different full-term neonates
(n = 6) and adults (n = 6). In activation experiments, platelets from neonates (white and red bars; n = 5 and n = 6, respectively) and adults (blue; n = 6) were activated by collagen-related peptide (CRP) (5 μg/mL), rhodocytin (100
nM), adenosine diphosphate (ADP) (25 µM), Protease-activated receptor (PAR)-1 (25
µM) and PAR-4 (250 µM) peptides, and phorbol 12-myristate 13-acetate (PMA) (100 nM).
The binding of fluorescently labelled fibrinogen (C ) and the P-selectin exposure (D ), was monitored for 30 minutes at room temperature (RT), post-agonist stimulation.
Values are mean plus standard deviation. [* ] and ** denote p ≤ 0.05 and p ≤ 0.005, respectively, versus adults.
Fig. 2 Expression and phosphorylation of Syk and PLCγ2 in platelets from adult and full-term
neonates in response to collagen-related peptide (CRP) and rhodocytin. Human washed
platelets were diluted to 600 × 109 platelets/L and stimulated with different dose of CRP (A , B ; E , F ) and rhodocytin (C , D ) for 5 minutes under stirring conditions. Reactions were stopped by adding one volume
of 5× reducing sample buffer. In panels E and F , identical amounts of total Syk (E ) and PLCγ2 (F ) were loaded in the gel. Samples were developed by sodium dodecyl sulphate-polyacrylamide
gel electrophoresis (SDS-PAGE) and analysed using phospho-specific antibodies. The
images shown are representative of 4 (A –D ) and 3 (E , F ) experiments with independent platelet samples.
Fig. 3 Receptor expression in mouse platelets during development. Platelet surface receptor
expression profiles in mice platelets were assessed by flow cytometry with fluorescently
labelled antibodies, as reported in the “Materials and Methods” section. Results are
expressed as mean plus standard deviation of the median fluorescence of platelet populations
in different mice (gestational and neonatal platelet, n = 12–26; adults platelets n = 6); [* ]
p ≤ 0.05; **p ≤ 0.005
Developmental changes in human platelet reactivity were assessed in flow cytometric
experiments by monitoring fibrinogen-binding and P-selectin expression following platelet
stimulation. Compared with adults, platelets from pre-term and full-term neonates
exhibited a marked reduction in fibrinogen binding in response to CRP, rhodocytin
and PAR-1 (over 50%), along with a moderate decrease for ADP and PMA (∼25%) and no
change in response to PAR-4 ([Fig. 1C ]). Agonist-induced platelet expression of P-selectin, a marker of α-granule release,
was also significantly decreased in pre-term versus adult platelets for CRP, rhodocytin
and PAR-1. In full-term neonates, P-selectin exposure was markedly reduced in response
to CRP and rhodocytin, with only a mild decrease in PAR-1 ([Fig. 1D ]).
These results demonstrate a hypo-sensitivity to GPVI and CLEC-2 agonists, which is
in part explained by a decrease in receptor expression.
Expression and Tyrosine Phosphorylation of Syk and PLCγ2 is Reduced in Neonatal Platelets
To investigate whether ITAM receptor signalling varies during development and contributes
to the defect in platelet activation by GPVI and CLEC-2, human platelets from full-term
neonates and adults were compared for Syk and PLCγ2 expression, and for tyrosine phosphorylation
of these key signalling proteins after stimulation with CRP and rhodocytin. There
was a reduction in the total level of Syk and PLCγ2 proteins in resting neonatal platelets
when compared with adult platelets (Syk: 0.53 ± 0.14 vs. 0.73 ± 0.19 arbitrary units;
PLCγ2: 0.15 ± 0.1 vs. 0.27 ± 0.06 arbitrary units) as measured by western blotting
([Fig. 2 ]). The gene expression level of SYK was also lower in platelets from full-term neonates
versus adult platelets ([Supplementary Fig. S2C ], available in the online version). The level of PLCγ2 mRNA was too low for quantitation.
Stimulation with either CRP or rhodocytin for 5 minutes induced dose-dependent tyrosine
phosphorylation of Syk and PLCγ2, as assessed with phospho-specific antibodies to
the activation site in Syk (residues 525/526)[30 ] and to an established marker of PLCγ2 activation (Tyrosine 1217).[31 ] As illustrated in [Fig. 2 ], phosphorylation of Syk and PLCγ2 was markedly reduced in neonatal platelets at
all concentrations of CRP ([Fig. 2A, B ]) and rhodocytin ([Fig. 2C, D ]). This reduced phosphorylation of Syk and PLCγ2 in neonatal platelets is still appreciable
after normalization for the reduced level of expression of Syk and PLCγ2 ([Fig. 2E, F ]), thus supporting the impairment in the signal transduction pathway common to GPVI
and CLEC-2 in full-term neonates.
Gestational and Neonatal Mice Platelets Display a Hypo-Reactivity to GPVI and CLEC-2
To further assess the influence of the gestational age on the change in platelet reactivity,
we extended our research to mouse platelets (E17.5 to P14.5 and adult). As with human
studies, we first measured the levels of GPVI, CLEC-2 and other major glycoproteins
by flow cytometry. As with human platelets, a mild reduction in the levels of GPVI
and CLEC-2 was observed in gestational and neonatal mice platelets relative to adults
([Fig. 3 ]), which cannot be explained by a change in mean platelet volume ([Supplementary Fig. S1B ], available in the online version).
We performed functional studies on the murine platelets in response to CRP, rhodocytin
and PAR-4 peptide. Foetal (E17.5) and early neonatal (P1.5–P7.5) mice platelets displayed
a significant impairment in fibrinogen binding in response to all concentrations of
CRP ([Fig. 4A ]) and to low and moderate concentrations of rhodocytin ([Fig. 4B ]). By 2 weeks of age (P10.5–P14.5), mice platelets stimulated with high concentrations
of rhodocytin recovered their fibrinogen-binding capacity ([Fig. 4B ]). In contrast, we observed only a mild reduction in fibrinogen binding of murine
platelets aged E17.5 to P14.5 in response to PAR-4 peptide ([Fig. 4C ]). A more severe decrease in response was observed in mice platelets for agonist-induced
expression of P-selectin. Thus, as compared with adults, the surface expression of
the α-granule protein was significantly reduced for all three stimuli in foetal and
neonatal mice platelets, with recovery increasing with age ([Fig. 4D ]–[F ]).
Fig. 4 Agonist-induced activation of mouse platelets during development. Activation of mouse
platelets during development was monitored by flow cytometry for (A –C ) fluorescently labelled fibrinogen binding and (D –F ) expression of P-selectin expression in response to (A , D ) collagen-related peptide (CRP) (2, 5, 10 µg/mL); (B , E ) rhodocytin (10, 30, 100 nM) and (C , F ) protease-activated receptor (PAR)-4 peptide (50, 100, 250 µM). Values are mean plus
standard error of the percentage of positive labelled platelets achieved in 3 to 9
different samples per group. [* ] and ** denote p ≤ 0.05 and p ≤ 0.005, respectively, in comparison to values in adults platelets.
These results demonstrate that as with human platelets, mouse platelets show impairment
in response to GPVI and CLEC-2 in foetal, neonatal and early post-natal life. In the
case for PAR-4, they also show a marked reduction for the expression of P-selectin,
and a mild decrease in activation of αIIbβ3 in late gestational and neonatal platelets.
Platelets are Hypo-Responsive to GPVI and CLEC-2 following Immune-Induced Thrombocytopenia
We hypothesized that the hypo-responsiveness to GPVI and CLEC-2 may be related to
the need to generate sufficient numbers of platelets to keep pace with the growth
of embryos and neonates. To model this high-pressure platelet production environment
in adult mice, we established an immune-induced thrombocytopenia mice model in which
we assessed the platelet receptor levels and activation by GPVI, CLEC-2 and PAR-4.
We observed that the new platelets generated following immune depletion have increased
size ([Supplementary Fig. S1C ], available in the online version). In concordance with this finding, the level of
αIIbβ3 and α2β1 integrins increased slightly following immune depletion ([Fig. 5 ]). In contrast, we observed no significant changes in the expression of the ITAM
receptors, GPVI and CLEC-2, following immune depletion ([Fig. 5 ]).
Fig. 5 Receptor levels in mice platelets before and after immune depletion. Surface receptor
expression profiles in mice were assessed following platelet immune depletion using
fluorescently labelled antibodies as described in the Supplementary Methods. Data
are shown as mean plus standard deviation of the median fluorescence of the platelet
population achieved in 12 to 24 different samples in each group. ** denotes p ≤ 0.005 versus values in mice treated with a control immunoglobulin G (IgG) antibody
(sham).
The increase in fibrinogen-binding and P-selectin expression in response to CRP was
markedly reduced in the first few days following immune thrombocytopenia ([Fig. 6A, D ]). A similar but less severe pattern was observed in response to low-to-moderate
concentrations of rhodocytin ([Fig. 6B ], [E ]). By contrast, there was no decline in platelet activation induced by PAR-4 following
immune thrombocytopenia ([Fig. 6C ], [F ]).
Fig. 6 Measurement of murine platelet function following immune depletion. Mice were injected
with α-GPIbα antibody to induce immune thrombocytopenia. Blood samples were collected
before platelet depletion and thereafter for up to 9 days. Platelet reactivity was
measured via (A –C ) fluorescent fibrinogen binding and (D –F ) P-selectin exposure in response to increasing doses of (A , D ) collagen-related peptide (CRP), (B , E ) rhodocytin and (C , F ) protease-activated receptor (PAR)-4 peptide. Data are shown as mean plus standard
error of the percentage of positive platelets achieved in 6 to 9 different samples
in each group. ** denotes p ≤ 0.005 versus values found in mice before injection of the α-GPIbα antibody.
Discussion
Several studies have shown that neonatal platelets are hypo-responsive to various
agonists such as ADP, epinephrine, thromboxane analogues, thrombin and collagen. This
is characterized by decreased aggregation, secretion and expression of platelet activation
upon stimulation. There is considerable variation among these studies on the degree
of reduction in platelet reactivity, which most likely relates to differences in procedures
for blood sampling, platelet function testing and limited sample volume.[10 ]
[11 ]
[12 ]
[19 ]
[32 ] Developmentally regulated expression of platelet receptors and receptor-coupled
signal transduction can underlie the different reactivity of neonatal and adult platelets.[11 ]
[12 ]
[32 ] This study focused on the expression and function of the ITAM receptors GPVI and
CLEC-2 during development in human and murine species, and in a murine model of haematopoietic
stress. Hypo-responsiveness of neonatal platelets to collagen has been consistently
seen in previous studies, and is more dramatic than that to GPCR agonists.[9 ]
[10 ]
[11 ]
[12 ]
[13 ]
[14 ]
In agreement with previous reports, we observed developmental regulation of the expression
of major adhesive platelet receptors in human (αIIbβ3; [Fig. 1 ])[10 ]
[12 ]
[19 ] and mice (GPIbα, and αIIbβ3 and α2β1 integrins; [Fig. 3 ]).[33 ] We speculate that a moderate but significantly reduced level of integrin αIIbβ3
in platelets from pre-term and full-term neonates contributes to the mild impairment
in αIIbβ3 activation (i.e. fibrinogen binding) by all agonists. In mice, however,
αIIbβ3 levels in gestational and early post-natal life are similar to that in adults
([Fig. 3 ]), while there is a marked impairment in fibrinogen binding in response to CRP, rhodocytin
and, to a lesser extent, PAR-4. Thus, this defect may be mainly contributed by impairment
in agonist-induced conformational changes and activation of the integrin, which was
not evaluated in this study.
The major novel finding of our work is that during foetal and neonatal life platelet
signalling through GPVI and CLEC-2 is impaired in mice and human platelets. While
this investigation was underway, a study has also observed a reduced platelet response
to CRP and rhodocytin in neonatal human platelets.[34 ] In addition to extending these findings to an earlier stage in human development
(pre-term infants) and to development in mice, we also show, for the first time, that
two mechanisms act synergistically to give rise to the impaired response to the two
ITAM receptor agonists. The first one is a reduction in the level of GPVI, its associated
protein Fcγ chain[28 ]
[29 ] and CLEC-2 in neonatal human platelets. This reduction is not accounted by significant
changes in platelet volume ([Supplementary Fig. S1 ], available in the online version), but is associated to changes at the transcriptional
level, as indicated by reduced GP6 , FCER1G and CLEC1B mRNA levels compared with adult platelets ([Supplementary Figs. S2 ] and [S3 ], available in the online version). Consistently, the level of these ITAM receptors
is also reduced in gestational and neonatal murine platelets compared with adult platelets.
In this study, we found no significant changes in platelet volume during mice development
([Supplementary Fig. S1 ], available in the online version), contrarily to previous study.[7 ] This discrepancy is most likely due to analytical differences in blood collection
and the device used to assess platelet volume.
The second mechanism contributing to the hypo-responsiveness of neonatal platelets
to GPVI and CLEC-2 agonists is a regulation in development of the expression of key
proteins in ITAM receptor signalling, such as Syk and PLCγ2, most likely at the transcriptional
level as neonatal platelets display lower SYK mRNA levels. The mechanisms underlying transcriptional differences in genes of the
ITAM receptor pathways remain unknown, and may include potential changes in the activity
of factors involved in the regulation of GP6 , FCER1G , CLEC1B and SYK .
Taken together, these two mechanisms contribute, in a degree that cannot be delineated
from our study, to a significant impairment of signalling downstream of GPVI and CLEC-2,
as reflected by a reduced Syk and PLCγ2 phosphorylation (this study), and the previously
reported decrease in Ca2+ mobilization in neonatal platelets in response to collagen.[16 ]
Moreover, using a mouse model, we show that the responses to GPVI and CLEC-2 are also
sharply reduced in adult mice within the first 3 days of recovery from immune thrombocytopenia,
where a high thrombopoietic pressure would exist ([Fig. 6 ]). This finding is in agreement with a very recent study showing that GPVI-ITAM signalling
is partially inactive in newly formed platelets generated in response to acute thrombocytopenia.[35 ] Thus, it could be speculated that the rapid generation of platelets at the time
of need (i.e. in development and recovery from a marked decrease in platelet count)
may be due to induction of neonatal thrombopoiesis, thus resembling generation of
red cells in response to acute anaemia which occurs through neonatal rather than adult
erythropoiesis.[36 ]
[37 ]
[38 ] However, our current data suggest that the pattern of murine platelet reactivity
following immune depletion in mice is complex and likely influenced by multiple factors.
For instance, we observed an increased P-selectin exposure of days 7 to 9 platelets
after depletion, in response to PAR-4 and rhodocytin. This is unlikely due to an increased
platelet size, which is almost negligible at days 7 to 9 as compared with days 2 to
5 after depletion, and may be related to the rebound in platelet reactivity to these
agonists. In addition, we found a puzzling increase in fibrinogen binding of days
4 to 5 platelets in response to 10 nM rhodocytin. In their recent study, Gupta et
al[35 ] found that newly formed young platelets have increased reactivity to thrombin when
immune depletion is achieved by an anti-GPIbα antibody, but not if this is induced
by anti-αIIbβ3 antibody (day 5). They also found an unaltered response to rhodocytin
(1 μg/mL) at day 5 following immune thrombocytopenia. Further studies are required
to fully clarify the responsiveness of newly formed young platelets released after
immune thrombocytopenia.
In contrast to the study of Baker-Groberg et al showing higher platelet-surface P-selectin
expression in resting neonatal platelets,[34 ] we found that this expression was similar to that in resting adult platelets. The
fact that we used adult venous blood and CB, whereas the previous study assayed capillary
blood,[34 ] may account for the discrepancy. As in humans, in resting conditions, the platelet
surface P-selectin levels were similar in murine neonatal and adult platelets. Moreover,
the level of P-selectin expression was unaffected following immune thrombocytopenia.
Notably, while in humans total P-selectin content is similar between adult and neonatal
platelets,[39 ] in mice it has been suggested that P-selectin expression is developmentally regulated.[40 ] Thus, the marked reduction in P-selectin secretion induced by CRP and rhodocytin
in gestational and early neonatal mice platelets that we have observed may be mediated
by a combination of reduced activation of their ITAM receptors and by developmental
up-regulation of the α-granule protein. In contrast, the impaired P-selectin response
to PAR-4 may be due solely to up-regulation in development, as PAR-4-induced activation
of fibrinogen binding is similar in gestational, neonatal and adult mice.
The functional significance of this decrease in ITAM receptor expression and signalling
in platelet activation during ontogeny is currently unclear. The foetus has little
threat of trauma during development other than the birth process, and GPVI has a minor
role in haemostasis.[41 ] However, platelets are exposed to collagen during vasculogenesis and angiogenesis
and a reduced responsiveness to GPVI may help to avoid unwanted thrombosis. In this
context, it has recently been reported that infusion of adult platelets into mice
at E14.5 leads to the rapid formation of occlusive thrombi,[42 ] and that transfusion of human adult platelets into neonatal blood leads to a hyper-coagulable
profile.[43 ]
To unravel the physiological relevance of reduced CLEC-2 levels during the first stages
of the development, where it plays a key role in blood–lymphatic separation, further
research will be required. Podoplanin expression changes during ontogeny[44 ]: in the first stages, podoplanin is highly and widely expressed, but over time its
expression becomes restricted.[45 ]
[46 ] Importantly, podoplanin is strongly expressed in the villous stroma of the placenta,[47 ] and during angiogenesis of the foetal vessels, platelets could be exposed to the
high amount of placenta podoplanin. Thus, reduction of platelet levels of CLEC-2 could
be a compensatory mechanism to avoid excessive platelet activation during development
and to safeguard neonatal haemostasis.
Our current results are of potential clinical significance, especially as premature
human babies have an increased risk of IVH.[11 ] The pathogenesis of IVH in pre-term infants is complex and multifactorial, and the
contribution of platelet hypo-reactivity is still unclear.[48 ] However, the poorly developed germinal matrix vessels in premature infants have
been established as a contributing factor in the aetiology of IVH.[32 ]
[49 ] Recent studies in murine embryos have demonstrated that deficiency of CLEC-2 or
podoplanin, its endogenous ligand, is associated with impaired angiogenesis and severe
brain haemorrhaging.[4 ] During development, CLEC-2 induces platelet activation and subsequent formation
of platelet aggregates through integrin αIIbβ3, to maintain the integrity of the nascence
blood vessels in the brain. Thus, the reduced CLEC-2 and integrin αIIbβ3 levels in
neonatal platelets that we report may contribute to the increase risk of IVH. Moreover,
apart from considering body weight and surface on the newborns, new insights into
the mechanisms of hypo-reactivity in neonatal platelets could aid the adjustment of
anti-platelet therapy in neonates for the prevention of arterial thrombosis in conditions
such as congenital heart disease, assist devices, prosthetic valves or systemic-to-pulmonary
shunt implants.[50 ]
There are several limitations of this study. The number of neonatal samples is relatively
low due to limited availability and logistical difficulties. The use of CB as a source
of neonatal platelets might not be ideal, but it is not otherwise possible to collect
the volumes that were needed for all assays. Importantly, however, it has been reported
that neonatal platelets from the cord and PB are similarly hypo-reactive.[19 ] In mice, the small sample volumes have limited the number of functional studies
that could be performed.
In conclusion, in this study we show that platelet expression and signalling of GPVI
and CLEC-2 are reduced in neonates of mice and humans. This hypo-reactivity of ITAM
receptors may prevent unwanted platelet activation during vascular development, but
an excessive ITAM signalling impairment could contribute to the increase in risk of
IVH in neonates.
What is known about this topic?
Neonatal platelets display degrees of hypo-responsiveness to most platelet agonists,
but most dramatically to collagen that activates platelets downstream of the ITAM
containing receptor GPVI.
Platelets also contain the hemi-ITAM receptor CLEC-2, which plays a critical role
in blood–lymphatic separation in development.
The potential contribution of developmental changes in ITAM receptor's expression
or signalling in neonatal platelets' hypo-responsiveness is unknown.
What does this paper add?
Pre-term and full-term neonates display mildly reduced platelet expression of ITAM
receptors GPVI and CLEC-2, accounted for at the transcriptional level. They also show
a pronounced ablated activation downstream of these receptors reflected in impaired
integrin αIIbβ3 activation, P-selectin secretion and Syk and PLCγ2 phosphorylation.
Mouse platelets are also hypo-responsive to GPVI and CLEC-2 from late gestation to
2 weeks of age, and following immune thrombocytopenia.
Our study provides a mechanistic explanation for the hypo-responsiveness of neonatal
platelets to collagen.
