Leukocytes of pregnant women with small-for-gestational age neonates have a different phenotypic and metabolic activity from those of women with preeclampsia

Abstract

Objective

Preeclampsia and pregnancies complicated by small-for-gestational age (SGA) neonates share several underlying mechanisms of disease. However, while an exaggerated systemic maternal inflammatory response is regarded as one of the hallmarks of the pathogenesis of preeclampsia, the presence of a similar systemic intra-vascular inflammation in mothers of SGA neonates without hypertension is controversial. The aim of this study was to determine phenotypic and metabolic changes in granulocytes and monocytes of women who develop preeclampsia and those who deliver an SGA neonate, compared to normal pregnant women.

Methods

This cross-sectional study included patients with a normal pregnancy (n=33), preeclampsia (n=33), and SGA without preeclampsia (n=33), matched for gestational age at blood sample collection. Granulocyte and monocyte phenotypes were determined by flow cytometry, using monoclonal antibodies against selective cluster of differentiation (CD) antigens. The panel of antibodies included the following: CD11b, CD14, CD16, CD18, CD49d, CD62L, CD64, CD66b, and HLA-DR. Intracellular reactive oxygen species (iROS) were assessed at the basal state and after stimulation (oxidative burst). Results were reported as mean channel brightness (MCB) or intensity of detected fluorescence. Analysis was conducted with non-parametric statistics. A p-value <0.01 was considered statistically significant.

Results

1) Women who delivered an SGA neonate had a higher MCB of CD11b on granulocytes and monocytes than those with a normal pregnancy (p<0.001 for both); 2) patients with preeclampsia had a lower median MCB of CD62L on granulocytes (p=0.006) and a higher median basal iROS and oxidative burst on monocytes than women with an SGA neonate (p=0.003 and p=0.002, respectively).

Conclusion

Pregnancies complicated by the delivery of an SGA neonate are characterized by a higher activation of maternal peripheral leukocytes than in normal pregnancies, but lower than in pregnancies complicated by preeclampsia.

INTRODUCTION

Preeclampsia and small-for-gestational age fetuses (SGA) represent two of the “Great Obstetrical Syndromes”.[1] These two conditions are the clinical end point of different pathophysiologic mechanisms and are frequently associated with each other, so that the presence of an SGA fetus is also considered as a criterion of the severity of preeclampsia.[2] The known pathological pathways shared by preeclampsia and SGA include the following: 1) abnormal placentation;[3–8] 2) chronic utero-placental ischemia;[9–28] 3) an imbalance between angiogenic and anti-angiogenic factors in maternal blood;[29–51] 4) increased trophoblast apoptosis/necrosis;[52] and 5) an enhanced maternal systemic inflammatory response.[53–79]

Given the several common risk factors, it is unclear why some women will develop a systemic maternal disease with or without fetal involvement and others will have an SGA fetus without maternal hypertension. Friedman et al.[80] have proposed that the difference between preeclampsia and SGA rests on the fact that the excessive maternal response is systemic in the former, while it is limited to the utero-placental compartment in the latter. However, this hypothesis has been challenged by others that reported increased neutrophil[55,64] and endothelial[69] activation in the peripheral blood of women with an SGA fetus. Previous flow cytometry studies have demonstrated that normal pregnancy is associated with changes in the phenotype and the metabolic activity of immune cells consistent with leukocyte activation,[60,81] and that these changes are further accentuated in several pregnancy complications such as pyelonephritis,[81] preterm labor, [82] preterm prelabor rupture of membranes,[83] and preeclampsia.[60,64,66,67] Thus, it is possible that preeclampsia and SGA differ in the severity and the extent of the maternal inflammatory response, as reflected by changes in immunophenotype and metabolic activity of the leukocytes in the innate immune system.

The aims of this study were to examine the phenotypic and metabolic changes in maternal granulocytes and monocytes between the following: 1) women with a normal pregnancy and those who delivered an SGA neonate; and 2) women with preeclampsia and those who delivered an SGA neonate without preeclampsia.

MATERIAL AND METHODS

Study Design and population

A cross-sectional study was conducted to compare the phenotypic and metabolic characteristics of peripheral blood granulocytes and monocytes obtained from 99 patients in the following 3 groups: 1) normal pregnant women (n=33); 2) women with preeclampsia (n=33); and 3) women who delivered an SGA neonate without preeclampsia (n=33). Patients were matched for gestational age at blood sampling (within 2 weeks).

Eligible patients were approached at the Detroit Medical Center/Wayne State University in Detroit, Michigan. Biological materials and some results of flow cytometric analysis of patients included in this study have been used for other studies of inflammation in pregnancy complications reported elsewhere. All women provided written informed consent prior to the collection of maternal blood. The collection of and utilization of maternal blood for research purposes was approved by the Institutional Review Boards of Wayne State University and by the Eunice Kennedy Shriver National Institute of Child Health and Human Development, NIH, DHHS.

Clinical definitions

Patients with normal pregnancies met the following criteria: 1) no medical, obstetrical or surgical complication; 2) gestational age ranging from 20 to 41 weeks; and 3) delivery of a term neonate (≥37 weeks), appropriate for gestational age, without complications. Preeclampsia was defined as the onset of hypertension (systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg on at least two occasions, 4 hours to 1 week apart) after 20 weeks of gestation and proteinuria (≥ 300 milligrams in a 24 hour urine collection at least one dipstick measurement ≥ 2+). [2] An SGA neonate was defined by sonographic estimated fetal weight below the 10^th centile for gestational age and confirmed by neonatal birthweight.[84]

Blood samples collection

A sample of peripheral blood was obtained by venipuncture using a syringe, added to an anticoagulant solution (20 μg/ml of the protease inhibitor leupeptin), placed on ice, and transported to the laboratory. The blood was processed and analyzed within 60 minutes of phlebotomy.

Flow cytometry studies

Evaluation of the granulocytes and monocytes surface markers was performed following the methods described by McCarthy et al.[85,86] Upon arrival of the sample to the laboratory, a vital nucleic acid dye -LDS-751- (Molecular Probes, Eugene, OR, USA) was immediately added to the specimen (final concentration was 0.0001%). The objective of this step was to separate leukocytes from anucleated red blood cells. Anti-sera against the cluster of differentiation (CD) markers and blood were mixed in pre-cooled tubes. This study included a panel of 12 tubes containing optimal concentrations of negative isotype control antibodies (IgG1 and IgG2a), CD11b, CD14, CD15, CD16, CD18, CD49d, CD62L, CD64, CD66b, and HLA-DR (Immunotech, Miami, FL, USA), which had been conjugated to the fluorescent dye, fluorescein isothiocyanite (FITC). After 10-minute incubation with the anti-sera, samples were analyzed with a flow cytometer.

Flow cytometry analysis was performed on a Coulter XL-MCL (with an argon-ion 488-nm laser) flow cytometer. FITC was detected at 525 nm and LDS-751 at 620 nm. Red blood cells, which are not labeled with LDS-751, were excluded from analysis. Granulocytes and monocytes were gated according to the characteristic staining pattern. For both cell types, the intensity of fluorescence or mean channel brightness (MCB) was recorded. The surface markers studied and the percentage of positive staining cells for each marker in healthy non-pregnant women are described in Table I.

Table I.

Leukocyte surface antigens analyzed and staining of leukocyte subpopulation with specific monoclonal antibodies with the use of whole-blood flow cytometry

Surface marker	Function	Monoclonal antibody subtype	Percentage of positive-staining cells*
			Granulocytes	Monocytes
CD11b	Integrin αM sub-unit, binds to ICAM-1	IgG1	60.5 (21.3)	51.4 (28.0)
CD14	Receptor for lipopolysaccharide and its binding protein	IgG2a	2.7 (1.2)	93.2 (4.3)
CD15	Fucosylated carbohydrate structure, recognizes endothelial selectins	IgM	98.6 (1.3)	21.8 (17.1)
CD16	Low affinity receptor for aggregated IgG	IgG1	96.3 (2.6)	13.3 (5.4)
CD18	Integrin β2 sub-unit, mediates migration of leukocytes through vessel wall	IgG1	98.1(3.4)	93.4 (7.1)
CD49d	α4-integrin, binds to VCAM-1 and fibronectin	IgG1	4.5 (2.0)	29.0 (17.8)
CD62L	L-selectin, mediates tethering and rolling of leukocytes	IgG1	78.0 (13.4)	53.5 (22.4)
CD64	High affinity receptor for IgG, mediates release of IL-1, IL-6 and TNF-α	IgG1	5.9 (2.6)	77.8 (21.4)
CD66b	Member of the CEA family, mediates cell-cell interaction	IgG1	70.6 (18.4)	3.5 (1.8)
HLA-DR	Class II MHC Antigen	IgG1	3.2 (1.2)	68.5 (13.6)

ICAM-1: intercellular adhesion molecule-1; VCAM-1: vascular cell adhesion molecule-1; IgG: immunoglobulin G; MHA: major histocompatibility antigen.

^*Granulocyte and monocyte subgroups distinguished by side scatter and LDS-751 fluorescence characteristics. For each subpopulation, the percentage of positive cells is given as mean (SD) for normal non-pregnant women (n=20).

The presence of iROS within granulocytes and monocytes was assessed by determining the basal content, the production in response to a stimulant (i.e., oxidative burst), and the stimulation index (ratio between oxidative burst and basal value of iROS). This was performed using the method described by Himmelfarb et al.[87] Briefly, one milliliter of peripheral blood, which was drawn with a syringe and inserted into a tube containing sodium heparin (10 IU/ml), was placed on ice and transported to the laboratory. The cells were incubated for 15 minutes at 37°C with 2′, 7′dichlorofluorescein diacetate (DCFH-DA), which diffuses across the cell membrane and is trapped within the cell by a deacetylation reaction. When the DCFH-DA is exposed to hydrogen peroxide, it is oxidized to the highly fluorescent 2′, 7′ dichlorofluorescein (DCF).

The oxidative burst was studied by adding 10 μl of N-formyl-methionyl-leucyl-phenylalanine (FMLP) (Sigma, St. Louis MO), dissolved in ethanol to a tube containing 50 μl of blood and DCFH-DA. The contents were gently mixed and incubated for 30 minutes at 37°C. After this period of time, 5 μl of LDS-751 (final concentration 0.0001%) in methanol were added, mixed and incubated for 1 minute at room temperature. The samples were then analyzed immediately on the flow cytometer. DCF was detected at 525 nm, LDS-751 at 620 nm, and phycoerythrin at 575 nm. FITC mean channel density staining was calibrated before analysis with Standard Brite Beads (Beckman Coulter, Miami, FL). A discriminator was set to exclude red blood cells, which do not label with LDS-751. Granulocytes were gated using an LDS-751 versus side scatter histogram, and the monocytes using a CD14-phycoerythrin versus side scatter histogram. The MCB was measured. Ten thousand events, excluding red blood cells, were collected for analysis.

Statistical analysis

Kruskal-Wallis and post-hoc with Mann-Whitney U tests were employed for comparisons of continuous variables. Chi-square tests were utilized for comparison of proportions. The statistical package used was SPSS v.12.0 (SPSS Inc., Chicago, IL, USA). A p-value <0.01 was considered significant.

RESULTS

Patient characteristics

The demographic and clinical characteristics of the study groups are displayed in Table II. There were no significant differences in maternal age, ethnic group distribution, and gestational age at sample collection among the three groups. As expected, the group with preeclampsia had the highest proportion of nulliparous women.

Table II.

Demographic and clinical characteristics of the study groups.

	Normal pregnancy (n=33)	p	SGA (n=33)	pa	Preeclampsia (n=33)	pb
Age	27 (18–36)	0.69	27 (18–43)	0.13	22 (18–36)	0.23
Nulliparous	7 (21.2%)	0.40	10 (30.3%)	0.05	18 (54.5%)	0.005*
Race: African American	23 (69.7%)	0.67	25 (75.8%)	0.95	26 (78.8%)	0.51
Caucasian	6 (18.2%)		5 (15.2%)		5 (15.2%)
Other	4 (12.1%)		3 (9.1%)		2 (6%)
GA at blood sampling	32.6 (24–39.5)	0.95	32.5 (25–40.2)	0.68	32.5 (25.4–40.1)	0.66
Blood sampling ≥ 37 weeks	11 (33.3%)	0.80	12 (36.4%)	0.80	11 (33.3%)	1.00
GA at delivery	39 (37–41.1)	<0.001*	35 (26.3–40.6)	0.70	34.5 (25.5–40.1)	<0.001*
Birthweight (grams)	3320 (2600–3885)	<0.001*	1840 (420–2530)	0.08	1880 (320–3670)	<0.001*
Birthweight <10th percentile	0	<0.001*	33 (100%)	<0.001*	15 (45.5%)	<0.001*
Birthweight <5th percentile	0	<0.001*	23 (69.7%)	<0.001*	8 (24.2%)	0.003*
Birthweight ratio ±	0.97 (0.8–1.1)	<0.001*	0.66 (0.2–0.8)	0.001*	0.76 (0.4–1.1)	<0.001*

SGA: small-for-gestational age; GA: gestational age

Values are expressed as median (range) or number (%); ± birthweight ratio = birthweight/median birthweight for gestational age.

p: Comparison between normal pregnant women and SGA; p^a : Comparison between SGA and preeclampsia; p^b : Comparison between preeclampsia and normal pregnant women;

^*Statistically significant (p<0.001).

Differences in phenotypes and metabolic activity in leukocytes between women with a small for gestational age neonate and those with a normal pregnancy

Patients with an SGA had a significantly higher median MCB of CD11b, a marker of leukocyte adhesion to vascular endothelium,[88] than normal pregnant women in both granulocytes and monocytes (p<0.001 for each; see Tables III and IV). The median oxidative burst and the stimulation index for both granulocytes and monocytes were higher in patients with SGA than in those with a normal pregnancy, but these trends did not reach the statistical significance threshold set up for this study (p=0.03 for both comparisons in granulocytes and p=0.02 for both comparisons in monocytes, respectively; see Table V). There were no significant differences in surface marker expression, baseline iROS, and oxidative burst in response to FMLP according to the severity of SGA (less than or more than 5^th percentile of birthweight for gestational age) (all p>0.01, data not shown).

Table III.

Mean channel brightness of labeled antibody binding to peripheral blood granulocytes

Surface marker	Normal pregnancy (n=33)	p	SGA (n=33)	pa	Preeclampsia (n=33)	pb
CD11b	4.1 (2.3–8.1)	<0.001*	6.4 (2.2–23.5)	0.17	5.8 (3.3–15.5)	0.001*
CD14	1.3 (0.7–1.8)	0.08	1.2 (0.7–2.5)	0.76	1.1 (0.6–2.8)	0.07
CD15	88.6 (44.4–122.4)	0.74	88.1 (54–127.3)	0.67	85.7 (66.3–116.1)	0.60
CD16	37.1 (0.9–58.8)	0.11	43.5 (0.5–123.3)	0.15	31.4 (4.1–110)	0.51
CD18	8.8 (5.6–12.3)	0.04	8.8 (6.0–30.7)	0.03	8.1 (4.5–16.3)	0.64
CD49d	0.9 (0.6–1.3)	0.03	0.9 (0.7–1.4)	0.70	0.9 (0.7–1.3)	0.01
CD62L	6.3 (4.1–8.1)	0.64	6.4 (2.0–11.3)	0.006*	5.5 (1.6–9.3)	0.002*
CD64	1.9 (1.2–5.7)	0.50	1.8 (1.1–4.3)	0.21	2.1 (1.0–3.3)	0.41
CD66b	4 (2.5–6.6)	0.02	4.3 (3.5–9.3)	0.63	4.6 (2.7–6.5)	0.08
HLA-DR	0.4 (0.4–0.5)	0.03	0.4 (0.3–0.7)	0.92	0.4 (0.3–0.6)	0.03

Values are expressed as median (range)

p: Comparison between normal pregnant women and SGA

p^a: Comparison between SGA and preeclampsia

p^b: Comparison between preeclampsia and normal pregnant women

^*Statistically significant (p<0.01)

Table IV.

Mean channel brightness of labeled antibody binding to peripheral blood monocytes

Surface marker	Normal pregnant (n=33)	p	SGA (n=33)	pa	Preeclampsia (n=33)	pb
CD11b	5.7 (3.4–9.3)	<0.001*	8.6 (3.7–15.2)	0.10	7.4 (3.9–22.5)	0.002*
CD14	44.7 (33.8–58.9)	0.13	46.8 (30.3–64.2)	0.14	45.4 (32.8–62.4)	0.88
CD15	2.1 (1.1–4.0)	0.06	1.7 (0.9–3.8)	0.17	2.1 (0.8–10.2)	0.80
CD16	0.9 (0.6–1.7)	0.61	0.9 (0.3–2.1)	0.92	0.9 (0.6–2.8)	0.53
CD18	16.4 (12.3–22.4)	0.04	17.9 (12.6–28.4)	0.04	16.1 (9.6–24.3)	0.95
CD49d	3.5 (2.2–5.3)	0.22	3.2 (2.0–4.4)	0.56	3.1 (2.2–5.2)	0.06
CD62L	6.2 (3.9–9.3)	0.02	7.1 (3.0–12.8)	0.08	6.4 (2.5–10.7)	0.51
CD64	13.3 (8.8–23.6)	0.30	12.5 (5.9–24.8)	0.52	12.7 (6.7–19.3)	0.62
CD66b	0.7 (0.5–1.4)	0.80	0.8 (0.3–1.1)	0.17	0.8 (0.6–1.5)	0.04
HLA-DR	9.4 (4.1–20.6)	0.15	11 (3.5–29.3)	0.97	10.8 (3.5–22.1)	0.34

Values are expressed as median (range)

p: Comparison between normal pregnant women and SGA

p^a: Comparison between SGA and preeclampsia

p^b: Comparison between preeclampsia and normal pregnant women

^*Statistically significant (p<0.01)

Table V.

Mean channel brightness of intracellular DCFH-DA activated iROS at baseline and after stimulation with FMLP (oxidative burst) and stimulation index (ratio of iROS after stimulation over base line level) in peripheral blood.

	Normal pregnant (n=33)	p	SGA (n=33)	pa	Preeclampsia (n=33)	pb
Granulocytes
Basal	4.5 (1.5–11.7)	0.17	5.7 (2.1–38.3)	0.33	6.4 (2.3–13.9)	0.01
Oxidative burst	18.6 (3.7–34.4)	0.03	21.3 (5.8–186.8)	0.02	31.2(9.4–66.8)	<0.001*
Stimulation index	3.6 (1.4–5.5)	0.03	4.1 (2.1–5.9)	0.03	4.4 (1.6–9.6)	<0.001*
Monocytes
Basal	4.6 (2.4–9.9)	0.08	4.9 (2.5–12.4)	0.003*	7.8 (2.9–12.2)	<0.001*
Oxidative burst	4.6 (2.5–11.4)	0.02	6.4 (2.9–22.0)	0.002*	10.7 (4.1–23.1)	<0.001*
Stimulation index	1.3 (0.7–2.1)	0.02	1.4 (0.7–2.1)	0.10	1.5 (0.9–2.5)	<0.001*

Values are expressed as median (range)

p: Comparison between normal pregnant women and SGA

p^a: Comparison between SGA and preeclampsia

p^b: Comparison between preeclampsia and normal pregnant women

^*Statistically significant (p<0.01)

Differences in phenotypes and metabolic activity in leukocytes between patients with preeclampsia and women with a small for gestational age neonate

Women with preeclampsia had a lower median MCB of CD62L on granulocytes than those with an SGA neonate (p=0.006; see Table III). Expression of the surface antigen CD62L is known to be down-regulated during leukocyte activation.[89] The median MCB for basal iROS and the response to FMLP in monocytes were significantly higher in patients with preeclampsia than that of those with an SGA neonate (p=0.003 and p=0.002, respectively; see Table V). Similarly, the median MCB for oxidative burst in granulocytes and the stimulation index were higher in mothers with preeclampsia than that of those with an SGA; however, the difference did not reach the statistical significance threshold set for this study (p=0.02 and p=0.03, respectively; see Table V).

DISCUSSION

Principal findings of this study

1) Women with SGA neonates had a significantly higher expression of CD11b in both granulocytes and monocytes than normal pregnant women; and 2) women with preeclampsia had a significantly lower expression of CD62L on granulocytes and a significantly higher baseline iROS production and oxidative burst on monocytes than women with an SGA neonate.

Immunophenotypic and metabolic changes of leukocytes in pregnancy

It has been proposed that pregnancy is physiologically accompanied not only by the relative suppression of the adaptive limb of immunity, which would be necessary in order to sustain the fetal allograft, but also by the activation of the innate immune system.[90] A good example for this modification is the experimental Shwartzman reaction, which demonstrated that in pregnant animals monocytes are primed, as a single injection of lipopolysaccharide (LPS) is sufficient to induce disseminated intravascular coagulation and renal cortical necrosis, in contrast to non-pregnant animals in which two doses of LPS are required.[91–93] In support of this hypothesis, peripheral leukocytes of women with normal pregnancies show significant changes in antigen expression and iROS concentration, which are akin to those observed in septic non-pregnant patients.[60] Moreover, the expression of CD62L and CD11b as well as other leukocyte surface antigens (such as CD64, CD66b, CD15, CD14) have been found to be further altered in several pregnancy complications such as acute maternal infections,[81] spontaneous preterm labor,[82] preterm prelabor rupture of membranes[83] and preeclampsia,[60,64,66] suggesting that all these obstetrical syndromes are characterized by different degrees of a systemic intra-vascular inflammatory response.

Immunophenotypic changes of leukocytes in women with a small for gestational age fetus

The finding of an increased expression of CD11b in granulocytes and monocytes of patients with an SGA fetus is consistent with a previous report by Sabatier et al.[64] CD11b is a subunit of the heterodimeric integrin α_Mβ_2, also known as macrophage-1 antigen (Mac-1) or complement receptor 3 (CR3). The expression of CD11b is directly involved in leukocyte firm adhesion to the endothelium through interaction with the intercellular adhesion molecule 1 (ICAM-1), while the presence of the β₂ subunit (CD18) is necessary to promote leukocyte migration.[94] The expression of CD11b and that of the CD18/CD11b complex is considered a sensitive marker of leukocyte activation in vivo.[88] Since this change was observed in granulocytes as well as monocytes in women with an SGA fetus, it is suggestive of increased leukocyte adherence to the vascular endothelium in this pregnancy complication. In support of our finding, Johnston et al.[55] reported an increased concentration of neutrophil elastase, a protein released by activated neutrophils, in the plasma of women with an SGA fetus. Collectively, these data suggest that neutrophils of women with an SGA fetus differ from those of women with normal pregnancies in their phenotypic characteristics, which are associated with endothelial adherence and degranulation.

SGA and preeclampsia are associated with different immunophenotypes and metabolic activity of maternal leukocytes

Preeclampsia is associated with phenotypic and metabolic changes of granulocytes and monocytes in the maternal circulation. Indeed, a previous study by our group[66] reported that granulocytes from women with preeclampsia had striking phenotypic and metabolic changes compared to those of women with a normal pregnancy. In particular, granulocytes of women who developed preeclampsia had a significantly higher median MCB of CD11b, lower MCB of CD62L and increased oxidative burst, while their monocytes had a significantly higher median MCB for CD11b, basal iROS concentration and oxidative burst than women with a normal pregnancy.[66] Here, the expression of CD62L in granulocytes of women with preeclampsia was significantly lower than that of women with an SGA neonate. In contrast, the basal iROS production and oxidative burst of monocytes of patients with preeclampsia were higher than those of women with SGA fetuses. These findings are novel and suggest that leukocyte activation in women with SGA fetuses is quantitatively lower and qualitatively different than that of patients with preeclampsia.

The immunophenotypic differences in the leukocytes of patients with preeclampsia and those with an SGA fetus are mainly in the degree of CD62L expression by granulocytes. This surface marker, also known as L-selectin, is constitutively expressed on the surface of most leukocytes. The interactions of CD62L with its endothelial ligands are involved in the initial capture and subsequent rolling of leukocytes along the blood vessels that perfuse sites of inflammation. Rolling allows leukocytes to sense pro-inflammatory cytokines and chemokines produced by activated endothelial cells and is a prerequisite step for firm adhesion to endothelium mediated by the expression of integrins such us CD18/CD11b, and migration into the interstitial tissue.[95] L-selectin is rapidly shed from leukocytes upon activation,[96] through the cleavage of the extracellular domain by the A Disintegrin And Metalloprotease-17 (ADAM-17).[89] Therefore, the down-regulation of CD62L is consistent with leukocyte activation although it can also be caused by other factors such as exposure to hyper- or hypotonic media and shear stress.[89]

The greater production of ROS found in preeclampsia is particularly intriguing since oxidative stress is likely to play a role in determining or amplifying endothelial dysfunction;[97–103] however, the causal-effect relationship has not been clarified, and attempts to prevent preeclampsia or its complications by supplementation of antioxidant during pregnancy have led to disappointing results.[104–107]

Trophoblasts have been found to inhibit, in a contact-dependent manner, the production and releasing of ROS by neutrophils;[108] this effect is mediated by the inactivation of the hexose monophosphate shunt which, in turn, is necessary to the generation of NAD(P)H and the production of superoxide.[108] Of interest, our group has demonstrated that patients with preeclampsia have a greater proportion of neutrophils with high NAD(P)H production than normal pregnant women, both in basal condition and after stimulation with LPS.[109] This finding indicates that the trophoblasts of patients with preeclampsia fail to block the production of ROS by activated leukocytes. It is possible that differences in the properties of the trophoblasts between patients with preeclampsia and those who have an SGA fetus are the cause of the diversity in the basal iROS concentration and the oxidative burst that we have observed in the present study. However, differences in the ability to deactivate leukocytes between trophoblasts from women with preeclampsia and women delivering an SGA neonate have not been investigated.

SGA and preeclampsia are associated with a different profile and intensity of maternal systemic response

Our results suggest that pregnancies with SGA fetuses are characterized by a higher activation of maternal peripheral leukocytes than normal pregnancies but lower than in pregnancies complicated by preeclampsia. As discussed below, such a conclusion is consistent with previous reports regarding other aspects of the maternal systemic response, which appears to be lower or different in pregnancies complicated by an SGA fetus than in those with preeclampsia. During a systemic inflammatory response, there is an increase in pro-inflammatory cytokines, activation of the coagulation cascade and angiogenesis.[110] These three inter-dependent events are part of the process by which the organism reacts to injury and threat. Evidence of the dysregulation of these mechanisms in preeclampsia and in SGA is the following:

Changes in the markers of inflammation in preeclampsia and SGA

Differences in the circulating maternal concentration of many pro-inflammatory cytokines have been reported in patients with preeclampsia in comparison to women with an SGA fetus. Maternal tumor necrosis factor (TNF)-α and its soluble receptor sTNF-R1 are elevated in patients with preeclampsia,[69,78] but TNF-α has been found either to be increased[111] or decreased[112] in mothers of SGA fetuses. In comparison to normal pregnant women, patients with preeclampsia,[69,77] but not women with an SGA neonate,[69,111] had higher circulating interleukin (IL)-6 and IL-8 concentrations. The median plasma concentration of the inflammatory molecule Pentraxin 3 is significantly increased in pregnancies complicated with preeclampsia but not in those complicated with intrauterine growth restriction.[73] The complement split products profile also differs between the two syndromes, as higher concentrations of C5a have been found in preeclampsia, while lower concentrations of C4a were associated with SGA, in comparison to normal pregnancy.[113]

Changes in hemostatic activity and thrombin generation in preeclampsia and SGA

Both obstetrical syndromes are associated with increased in vivo thrombin generation in comparison to normal pregnancy.[114–122] However, the concentration and the activity of specific components of the coagulation cascade differ between women with preeclampsia and those with an SGA fetus. Indeed, maternal plasma tissue factor, the major activator of the coagulation cascade, is higher in women with preeclampsia and lower in mothers of an SGA fetus compared with normal pregnancies;[79] moreover, the ratio between maternal plasma tissue factor and its natural inhibitor was significantly lower among patients with preeclampsia than in those with normal pregnancy as well as in those who delivered an SGA neonate.[79] Monocyte activation is associated with secretion of tissue factor in the maternal circulation[123–129] and this process is mediated by the production of the ROS superoxide anion (O₂⁻) and the activation of the transcription factor nuclear factor-kappa B (NF-κB).[129–132] It is possible that the loss of inhibition of leukocyte ROS production by the trophoblasts of women with preeclampsia[109] can also account, at least in part, for the high plasma TF concentration. This process may not be so prominent in pregnancies with SGA fetuses.

Angiogenic and anti-angiogenic state in preeclampsia and SGA

Both preeclampsia and SGA are characterized by an anti-angiogenic state, indicated by decreased concentration of the pro-angiogenic factor placental growth factor (PlGF)[30,32,34–36,41,44,47] and increased concentrations of the anti-angiogenic factors soluble vascular endothelial growth factor receptor (sVEGFR)-1[29,30,32–36,41,42,44] and soluble endoglin (sEng).[31,37–39,41,44,46,48–50] However, each obstetrical syndrome seems to have a unique pattern of angiogenic and anti-angiogenic factors during the course of pregnancy. The following solid body of evidence support this view: 1) women destined to deliver an SGA neonate have a higher plasma concentration of sEng than controls since the first trimester, while similar changes occur only after 23 weeks of gestation in women who develop preterm preeclampsia and after 30 weeks in those who develop term preeclampsia;[41] 2) maternal plasma PlGF concentrations are lower than in normal pregnancy, both in women with SGA neonates and in those with preeclampsia, throughout gestation;[30,41] however, those destined to develop preterm preeclampsia have an earlier decline of maternal plasma PlGF concentrations than those destined to develop preeclampsia at term or to deliver an SGA neonate;[41] and 3) an increase in maternal plasma sVEGFR-1 between the first and the second trimester is associated with an increased risk to develop preterm preeclampsia but not SGA without hypertension.[40,44]

Strengths and limitations of the study

The main strength of this study lies in the sample size (99 subjects), which is the largest among flow cytometry studies performed in pregnancies complicated by SGA or preeclampsia.[60,76,125,133] However, two potential limitations should be considered: first, the cross-sectional design of the study does not allow us to demonstrate neither a temporal nor a causal relationship between the observed changes in the immunophenotype and metabolic characteristics of maternal leukocytes and these pregnancy complications. The second limitation depends on the adoption of a low p-value for statistical significance (<0.01). Such a strict criterion increases the possibility of a type II error, potentially preventing us to demonstrate some existing phenotypic and metabolic differences between the study groups. However, this approach was chosen to correct for multiple comparisons, and thereby it actually reduces the risk of type I error (reporting that a difference exist while, in fact, there is no difference).

In summary, abnormal placentation has been proposed to play a major role in the activation of the maternal intra-vascular systemic inflammatory response through the release of soluble factors (i.e. inflammatory cytokines,[134–137] anti-angiogenic factors,[29,138,139] and free-radical species[19,100,140–142]) or particles (syncytiotrophoblast microparticles[143,144]) into the maternal circulation. While the histological findings in the placentae of pregnancies complicated by preeclampsia or SGA are similar,[4,6] the pattern of the maternal systemic inflammatory response differs. Our findings of a diverse profile of leukocyte activation, along with the discrepancies in other features of systemic inflammatory response, support the hypothesis that SGA and preeclampsia are two distinct entities.