Main

Angiogenesis, the formation of new blood vessels from preexisting ones, occurs in a wide range of developmental, physiologic, and pathologic processes and is regulated by the action of soluble polypeptide growth factors, which stimulate the proliferation of endothelial cells(1,2). It has been demonstrated that the angiogenic heparin-binding growth factors, bFGF and VEGF (also known as vascular permeability factor)(3,4), induce an angiogenic response via a direct effect on endothelial cells and that by acting in concert they have a potent synergistic effect on the induction of angiogenesis in vitro and suggestively in vivo(5).

bFGF is widely expressed by embryonic tissues, in which it has both mitogenic and morphogenic actions, as well as in some human fetal tissues in early second trimester and in human placenta until term(6–10). Expression of VEGF during embryonic life has been documented in neuroectoderm, choroid plexus, kidney glomeruli, lung alveoli, adrenal cortex, cardiac myocytes, the highly vascularized placenta, and corpus luteum(2,11).

It was recently reported(12) that the indirect angiogenic factor angiogenin rises significantly in neonatal serum soon after birth. This elevation has been attributed to elimination of the placenta, which is known to produce an RNase inhibitor, placental ribonuclease inhibitor (PRI), abolishing, among other actions, the angiogenic property of angiogenin(13).

This study was based on the hypothesis that, in the first days after birth at term, levels of the angiogenic factors are expected to change because the placenta, the main source for angiogenic factors in the prenatal life(14,15), is eliminated, angiogenic demands are considerably decreased (compared with the in utero state), and adaptation to extraurine life and normal function are urgently required. Thus, alterations in the levels of angiogenic factors postnatally might be implicated in other functions besides slowed down angiogenesis. We therefore investigated possible changes in the serum levels of the direct angiogenic factors bFGF and VEGF from birth to early neonatal life, as well as their association with sex, birth weight, and perinatal stress consequent to the mode of delivery.

METHODS

Subjects. Thirty healthy, appropriate for gestational age, full-term infants, born from 30 healthy, nonsmoking mothers with uncomplicated single pregnancies were included in this study after informed consent of their parents. The study was approved by the Hospital Ethics Committee. Sixteen were boys and 14 girls. Mean (±SD) birth weight was 3264.7 ± 337.1 g (range, 2530-3950 g), and mean gestational age was 38.8 ± 1.1 wk (range, 37-40 wk) Twenty infants were born by vaginal delivery and 10 by cesarean section because the mother had a previous cesarean section. Apgar scores of all studied infants were >8 at birth and 5 min afterward. Placentas were in all cases normal in appearance and weight.

Procedure. One milliliter of blood was drawn from a peripheral vein of all 30 infants on d 1 and 4 of life. In 10 cases (seven vaginal and three cesarean section deliveries) blood was also drawn from the mother before delivery as well as from the umbilical cord at delivery. Blood was collected in pyrogen-free tubes, and serum was immediately separated by centrifugation after clotting and was kept frozen at -20°C until assayed. The analysis was performed by enzyme immunoassays, using commercially available kits (human FGF basic Quantikine HS and human VEGF Quantikine, R&D Systems Minneapolis, MN). The minimal detectable concentration for bFGF and VEGF was 0.5 and 9.0 pg/mL, respectively. Intra- and interassay coefficients of variation of the assays were 7.2 and 8.1% for bFGF and 4.8 and 6.6% for VEGF, respectively.

For each blood sample drawn, a whole blood count (including platelet count) was also performed.

Data analysis. Two nonparametric tests, the Wilcoxon and Mann-Whitney U tests, were used in the analysis. Spearman's correlation coefficients were also calculated.

RESULTS

Serum concentrations (pg/mL) of bFGF and VEGF in MS, UC, N1, and N4 are presented in Table 1. In all MS samples, VEGF levels were below the lower limit of detection; however, for this analysis 9.0 pg/mL, the lower limit of detection, was calculated. On the other hand, platelet counts were in all cases within normal ranges (Table 1). The Wilcoxon test for pair differences, relating to the 10 MS, UC, N1, and N4 samples for bFGF and VEGF, showed significantly lower levels in MS than in UC (p = 0.02 and 0.036, respectively), N1 (p = 0.009 and 0.006, respectively), and N4 (p = 0.009 and 0.006, respectively). Levels of bFGF in UC did not differ significantly from those in N1 and N4. In contrast, levels of VEGF in UC differed significantly from those in N1 (p = 0.008) and N4 (p = 0.006). Both factors did not change from N1 to N4 (neither in the subgroup of 10 nor in the total group of 30 neonates).

Table 1 Concentrations of bFGF and VEGF in the sera and platelet counts of blood obtained from mothers (n = 10), umbilical cords (n = 10), and neonates on the first (n = 30) and fourth days (n = 30) of life
Full size table

The Mann-Whitney U test, applied in the analysis of the levels of each angiogenic factor in UC, N1, and N4 samples between boys and girls, as well as between those infants born by vaginal delivery or by elective cesarean section, did not show any statistically significant difference either for bFGF or VEGF. Spearman's correlation coefficient, applied in matched MS, UC, N1, and N4 of either bFGF or VEGF, did not reveal any statistically significant correlation. Also, bFGF serum levels did not correlate with VEGF serum levels in any of the samples (MS, UC, N1, or N4). Finally, neither bFGF nor VEGF serum levels in UC, N1, and N4 correlated with birth weight.

DISCUSSION

The results of this study indicate that serum levels of bFGF and VEGF are significantly lower in MS before delivery than in UC, N1, or N4. Furthermore, the levels of bFGF in UC do not differ significantly from those in N1 and N4. This fact does not apply to the other factor, VEGF, the levels of which rise considerably in N1, thus differing significantly from levels in UC. Moreover, the levels of both factors do not show a statistically significant change from d 1 to 4 of life. No correlations in matched maternal, fetal, or neonatal serum samples for both factors are found.

Previous studies have demonstrated that in adult life, bFGF, although present within pituitary, brain, adrenal, and ovary(16–18), is nor normally found in either serum or plasma(19–21). This fact has been partly attributed to the lack of a signal sequence domain in the bFGF precursor protein that is necessary for secretion from the cell via the endoplasmic reticulum(22) and to the avid binding of any bFGF released from cells to sulfated glycosaminoglycans within the extracellular matrix(23). Gauthier et al.(19) have suggested an ongoing enzymatic destruction of bFGF in the normal adult circulation. Nevertheless, Hill et al.(14) have reported the presence-in soluble form-of bFGF in the maternal and fetal circulations in levels as high as the ones encountered in patients with pituitary tumors(21), parathyroid adenomas, or insulinomas(14). The placenta is considered the predominant source of maternal circulating bFGF, and its appearance in pregnancy may partly reflect the disappearance of a destructive protease(14). On the other hand, the placenta, together with the fetal arterial and venous vascular endothelium, is responsible for the soluble bFGF in the fetal circulation(8). This study, applying a high-sensitivity bioassay, detected bFGF concentrations in the maternal circulation, however, in levels significantly lower than in the fetus or neonate, a finding possibly explained by the diversity of sources of bFGF production in the fetus. The lack of significant differences between UC, N1, and N4 samples could be accounted for by the assumption that extensive angiogenesis, and thus bFGF production, has slowed down before birth at term.

With respect to VEGF, in the present study, maternal levels of this factor were below the limit of detection. A similar finding has been previously reported(24). This fact could be attributed to the in vivo high-affinity binding of VEGF with the natural soluble receptor (Flt-1), and consequent inhibition of its actions(25). Nevertheless, the refinement of the immunoassay to detect all VEGF isoforms (of 206, 189, 165, 145, and 121 amino acids) and not only VEGF165 might alter the results. On the other hand, the contribution of placenta growth factor to the measured VEGF levels cannot be ruled out, because the VEGF165/placenta growth factor heterodimer exhibits approximately 20% cross-reactivity over the range of the immunoassay.

The higher VEGF levels in UC than in MS samples might be explained by the fact that, at term, fetally derived macrophages within the placental villi are still expressing VEGF. The role of the latter is maintenance of the endothelium or regulation of vascular permeability rather than angiogenesis, which has diminished by that period(15). Moreover, higher levels of this factor in early neonatal life might reflect the adaptation demands to extrauterine life of tissues expressing VEGF mRNA and not undergoing obvious angiogenesis, such as epithelial cells of lung alveoli and cardiac myocytes. In these tissues, VEGF might also play a role in regulating microvascular permeability(26,11) or maintaining the differentiated state of microvessels, which otherwise might undergo involution. In this respect VEGF is primarily considered a regulator of normal function(11). On the other hand, it should be taken into consideration that at delivery, either vaginally or by elective cesarean section, a clinically inapparent, transient hypoxia may occur, which nevertheless could possibly up-regulate VEGF levels soon after birth(27). Finally, it should be stressed that in this study platelet counts were in all cases within normal ranges. It has been recently shown(28) that platelets contain VEGF, which is released on their activation during coagulation. Thus, elevated VEGF levels in this study cannot be attributed to elevated platelet counts.

The lack of correlation in matched and fetal samples for bFGF as well as VEGF could suggest distinct tissue sources(14). In this study the levels of neither bFGF nor VEGF depended on birth weight. The latter is in accordance with a previous report(29), in which bFGF levels in term maternal and cord serum of normal subjects-in contrast to diabetic ones-did not correlate with fetal size. Although some authors state that sex differences in the levels of other angiogenic factors exist in a variety of mouse tissues(30), and Alarid et al.(31) suggest that bFGF plays a role in the development of the male reproductive tract, this study failed in documenting dependence of bFGF and VEGF serum levels on sex. The mode of delivery does not seem to play a role, taking into account the fact that the infants studied were healthy full-term infants with high Apgar scores, in those delivered either vaginally or by elective cesarean section.

Finally, this study, in contrast to a previous study comprising children and adolescents(32), did not demonstrate a correlation between the two angiogenic factors in vivo. Nevertheless, a possible in vivo synergism of these two factors cannot be excluded.

In conclusion, healthy, full-term, appropriate for gestational age infants have significantly higher serum levels of bFGF and VEGF in early neonatal life than corresponding maternal levels. Considering bFGF, neonatal levels do not differ significantly from fetal ones, possibly reflecting the diminished angiogenesis rate in extrauterine life, which already has started in utero at term. In contrast, levels of VEGF rise significantly in early neonatal life, a finding possibly signifying adaptation demands (VEGF is also considered a regulator of normal function) as well as a clinically inapparent transient hypoxia related to the birth process.