Investigation of the actin scavenging system in pre-eclampsia.

OBJECTIVES: Cell injury releases actin, the most abundant cell protein. Gelsolin and vitamin D binding protein (VDBP) together depolymerise and clear cell-free actin. Impaired actin clearance is associated with several diseases and correlates with clinical outcome. The actin scavenging system was investigated in pre-eclampsia (PE), a procoagulant and proinflammatory state with placental and vascular damage. STUDY
DESIGN: Plasma gelsolin and actin free VDBP (AFVDBP) were measured in PE (early onset <33weeks; late onset ≥36weeks), matched normal pregnant (normP) and non-pregnant (nonPr) women, using commercially available ELISAs. Longitudinal samples from normP and women who subsequently developed PE were also analysed.
RESULTS: Plasma gelsolin fell during pregnancy (p=0.0002), with a concomitant rise in actin-free VDBP (p<0.001). Gelsolin concentrations were only significantly lower in established PE (p<0.05) when compared to non-pregnant controls.
CONCLUSIONS: We have shown that the components of the actin clearance system, gelsolin and AFVDBP, are altered in normal pregnancy and further changes occur in established PE, suggesting depleted actin clearance in PE. Whether this is a cause or consequence of PE pathophysiology requires further investigation.

Introduction

Pre-eclampsia (PE) affects 3–5% of pregnancies, causing maternal and perinatal mortality or morbidity [1] and lasting health implications for both mother and baby [2]. It is characterised by a maternal systemic inflammatory response, endothelial cell dysfunction and increased coagulation, secondary to disordered placental function [1]. In PE, the physiological activation of coagulation, present in normal pregnancy, is exaggerated with excessive platelet activation and intervillous fibrin deposition in the placenta [3], [4]. As the causative factor(s) originate from the placenta, removal of the placenta remains the only effective treatment.

Many placental derived factors are implicated in the maternal syndrome [5], with anti-angiogenic factors being the best characterised. Alternatives include ‘danger’ molecules released from damaged or dying cells that become toxic in the extracellular milieu, such as actin [6], [7]. Extracellular actin can damage microvascular capillaries, activate platelets and impede clot lysis, all effects associated with the excessive procoagulation evident in PE [4], [8], [9]. An efficient system comprising gelsolin [10], [11] and vitamin D binding protein (VDBP) works to cleave extracellular actin and inhibit repolymerisation, blocking its thrombotic effects [9], [12], [13]. Meanwhile, hypogelsolinemia is common in trauma and inflammatory diseases, with higher gelsolin levels correlating with improved mortality rates [14], [15]. Restoring circulating gelsolin and VDBP levels may therefore avoid more serious complications and reduce mortality rates [16]. Gelsolin is also an attractive candidate for therapy as it modulates several proinflammatory pathways [14].

Cell-free actin levels are increased in normal pregnancy, shown by increased actin-VDBP complex levels [17]. In PE there is potential for a further increase in circulating actin due to increased placental apoptosis and necrosis and activated endothelial and immune cells [18], [19]. It is not known how the components of the actin scavenging system change during normal pregnancy or PE. Therefore, the aim of this pilot study was to determine the circulating levels of gelsolin and actin-free VDBP (AFVDBP; to measure the actin binding capacity of VDBP) in normal pregnancy and PE, compared to non-pregnant women.

Materials and methods

To determine whether PE was associated with changes in circulating levels of plasma AFVDBP and gelsolin, a single blood sample was taken from women with PE (early onset (<33 weeks gestation; n = 10) and late onset PE (≥36 weeks gestation; n = 10)) and matched to normal pregnant women (age (±4 years), parity (0, 1–3, 4+), and gestational age (±13 days)) (normP) and non-pregnant (age and parity) (nonPr) controls (Table 1).

Table 1

Clinical characteristics of participants recruited to a cross-sectional study investigating changes in plasma gelsolin and actin free vitamin D binding protein levels in normal pregnancy and pre-eclampsia (PE). A single plasma sample was collected from women diagnosed with either early onset (<33 weeks; n = 10) or late onset (≥36 weeks; n = 10) PE and matched non pregnant (NonPr; n = 10/group) and normal pregnant (NormP; n = 10/group) women. Data are shown as median (range).

	Age (yrs)	Gestation at sampling (wks+days)	Gestation at delivery (wks+days)	Nulli-parity	Blood pressure (mm Hg) (systolic) (diastolic)	Proteinuria (mg/24 h)	Birthweight (g)
Early onset PE matched group
NonP (n = 10)	32 (21–38)	N/A	N/A	8/10	N/A	N/A	N/A
NormP (n = 10)	33 (23–37)	30+4 (24+2–33+4)	40+5 (39+5–42+0)	8/10	125/80 (120–140) (60–86)	None detected	3613.5 (3069–4002)
PE (n = 10)	30 (21–38)	29+2 (24+4–32+5)	31+2*** (25+4–33+3)	8/10	187.5/120*** (170–215) (110–130)	3803.5 (940–9433)	1245.5*** (490–2009)
Late onset PE matched group
NonP (n = 10)	31.5 (23–36)	N/A	N/A	6/10	N/A	N/A	N/A
NormP (n = 10)	33 (20–36)	37+1 (35+3–38+3)	39+5 (38+5–40+4)	6/10	130/79 (110–137) (70–88)	None detected	3480.5 (3022–3842)
PE (n = 10)	31 (21–39)	37+3 (35+5–38+6)	38+0*** (36+1–38+6)	6/10	166/110*** (148–190) (100–120)	1022 (727–1687)	2911.5** (2241–3592)

NonPr, non-pregnant recruits; NormP, normal pregnancy and PE, pre-eclampsia.

p < 0.01 compared to matched normP.

p < 0.002 compared to matched normP.

Longitudinal samples were collected to determine changes in plasma AFVDBP and gelsolin over the course of normal pregnancy (normP), prior to and subsequent to the onset of PE. These samples were collected from pregnant women recruited to the Oxford Pregnancy Biobank (n = 10) in the first (11–13 weeks gestation), second (20–22 weeks) and third (30–34 weeks) trimesters of pregnancy (Table 2). Blood samples from matched (age and parity) nonPr controls were also collected (n = 10). In those women who subsequently developed PE, 4 developed early onset PE (<33 weeks) and 11 late onset PE (≥35 weeks gestation).

Table 2

Clinical characteristics of matched non pregnant (NonPr), normal pregnant (NormP) and pre-eclamptic (PE) women recruited to a longitudinal study samples investigating changes in plasma gelsolin and actin free vitamin D binding protein levels during normal pregnancy and prior to and following development of pre-eclampsia (PE). Plasma samples were collected from each gestation and once PE had developed. Data are shown as median (range).

	Age (yrs)	Gestation at delivery (wks+days)	Nulli-parity	Blood pressure (mm Hg) (systolic) (diastolic)	Proteinuria (mg/24 h)	Birthweight (g)
Longi-tudinal group
NonP (n = 10)	29 (20–33)	N/A	9/10	N/A	N/A	N/A
NormP (n = 10)	29 (19–34)	40+1 (37–41+5)	9/10	122/76.5 (116–136) (70–91)	None detected	3223 (2530–4000)
PE (n = 10)	29.5 (18–35)	37+2*** (22+3–40+0)	9/10	160/110*** (126–180) (94–128)	2570.5 (586–14645)	2597* (1560–5788)

NonPr, non-pregnant recruits; NormP, normal pregnancy and PE, pre-eclampsia.

p < 0.05 compared to matched normP.

p < 0.002 compared to matched normP.

All pre-eclamptic women were recruited in the hospital following a positive diagnosis. Matched normal pregnant and non-pregnant women were recruited during routine prenatal appointments and from the community. All blood samples were collected into EDTA tubes, plasma isolated and the samples stored at −80 °C until analysis. Due to the potential for cell stress and therefore release of actin, no samples were collected during labour. Normal pregnant women had a healthy singleton pregnancy and no history of chronic illness. Pre-eclampsia was defined as new hypertension (diastolic blood pressure ≥90 mm Hg on two consecutive occasions) and new proteinuria (24 h secretion of ≥500 mg). The Oxfordshire Research Ethics Committee C approved this study (ref. 09/H0606/10) and informed written consent was obtained from all recruits.

Gelsolin and AFVDBP concentrations were determined using commercially available ELISA kits (Gelsolin: USCN Life Sciences Inc.; VitDBP: Immundiagnostik AG) according to the manufacturers’ instructions.

Kruskal–Wallis test with a Dunn's post hoc test and Mann–Whitney U-test were used to analyse study participants’ clinical characteristics. Friedman test with Dunn's multiple comparison test was used to compare matched ELISA data. For non-parametric longitudinal data a Friedman repeated measures test was used to determine significance of changes over gestation in normP and PE. Statistical analyses were carried out using Prism software. p < 0.05 was considered statistically significant.

Results

Within the study groups, matching criteria did not differ. PE groups had high proteinuria and significantly higher blood pressure (p < 0.002), lower birthweights (p < 0.05–p < 0.002) and shorter gestations (p < 0.002) compared to the normP group (Table 1, Table 2).

In both the early onset and late onset PE groups, plasma AFVDBP levels were significantly higher in normP compared to nonPr (p < 0.01) (Fig. 1Ai). With established early onset PE, plasma AFVDBP levels tended to be lower than normP but not significantly. In both early onset and late onset PE, plasma gelsolin levels were significantly lower than nonPr controls (p < 0.05). Although there was a trend for gelsolin levels to be lower in normP than the nonPr group, this was not statistically significant (Fig. 1Aii).

Fig. 1

(Ai) Plasma actin free vitamin D binding protein (AFVDBP) and (Aii) gelsolin levels in plasma samples from women with early onset (<33 weeks gestation) or late onset (≥36 weeks gestation) PE, and matched normal pregnant (normP) and non-pregnant (nonPr) women (n = 10/group) and (Bi) AFVDBP and (Bii) gelsolin levels throughout pregnancy in women who developed PE and matched normP (n = 10). Blood samples were taken in each trimester of pregnancy (1–3) and at the diagnosis of PE, and compared to non-pregnant (nonPr) matched women. Bars represent median values. a: p < 0.001 and b: p < 0.01 when compared to nonPr.

In normP, AFVDBP was already significantly increased relative to nonPr (p < 0.001) during the first trimester and increased further as pregnancy advanced (p < 0.001). This was also seen in pregnancies subsequently complicated with PE (p < 0.01) but the later rise in AFVDBP was not apparent (Fig. 1Bi). Plasma gelsolin levels fell significantly during normal pregnancy (p = 0.0002), with greater variability in pregnancies that subsequently developed PE (Fig. 1Bii). Third trimester gelsolin levels were significantly lower than nonPr levels in both the normP and PE groups (p < 0.001) (Fig. 1Bii). Finally, plasma gelsolin levels were significantly lower in established PE compared to those measured in the second trimester, prior to the onset of clinical symptoms (Fig. 1Bii).

Comments

We have extended previous observations of raised VDBP during normal pregnancy [20], [21], [22] and shown increased actin binding capacity of VDBP during normal pregnancy. A first trimester rise was evident in both normal pregnancies and pregnancies that subsequently developed PE. This is compatible with VDBP being an acute phase reactant [23] and that normal pregnancy is associated with an early and progressive low grade, systemic inflammatory response, with increased circulating C-reactive protein in the first trimester [24]. AFVDBP levels, however, plateaued in pregnancies where PE later developed, suggesting more actin bound VDBP, caused by increased extracellular actin levels. Alternatively, a reduced placental contribution to circulating VDBP levels in PE may account for the failure of AFVDBP to continue to rise during the second and third trimesters [25]. This is also the first report of falling plasma gelsolin levels during normal and PE pregnancy. Possible mechanisms behind this fall include increased turnover due to clearance of actin complexed gelsolin or cleavage by matrix metalloproteases whereupon gelsolin loses its actin binding capacity [26], [27]. The further decrease in gelsolin and AFVDBP in established PE suggests that decreased gelsolin and AFVDBP is a consequence of PE rather than a cause. A larger longitudinal study is required, however, to determine the time course of the reduction in circulating gelsolin in severe and mild forms of PE. Prophylactic gelsolin/VDBP treatment to reduce the multi-organ damage seen with the most severe forms of PE may still warrant further investigation.