**2. IGF system in fetal growth and preeclampsia**

#### **2.1 The role of the IGF system in fetal growth**

An increasing wealth of literature including clinical and knockout studies in mice points to the crucial role of IGF axis in correct embryonic and placental development and growth. Regarding the role of IGF-I in fetoplacental growth, clinical studies based on the measurements in cord blood from healthy newborns demonstrated that birth weight is positively correlated with IGF levels and therefore, the levels are low and raised in small-for gestational-age (SGA) infants and large-for gestation-age (LGA) infants retrospectively [Giudice et al., 1995; Osorio et al., 1996; Boyne et al., 2003]. These clinical observations were further confirmed by studies using transgenic mice in which the mutation of the gene encoding either IGF-I or IGF-II resulted in restricted growth [Efstratiadis, 1998]. However, other research groups do not lend support to a relationship between total IGF-I and birth

Recent Insights into the Role of the Insulin-Like Growth Factor Axis in Preeclampsia 151

The evolutionary pattern of expression of IGFs, IGFBPs and IGF-1R genes in the developing human placenta and fetal membranes has been described from as early as 6 weeks' gestation to term in order to detect the exact expression cite of each peptide and the potential cite of its action [Han et al., 1996]. Functionally, IGF axis appears to be involved in many aspects of placental development and metabolism both in uncomplicated and preeclamptic pregnancies though the exact signaling pathways have yet to be determined [Hills et al., 2004]. Studies based on cultured human trophoblast cells and cell lines propose that IGFs promote proliferation, regulate trophoblast migration and the differentiation of cytotrophoblasts into syncytiotrophoblasts and extravillous cells, enhance the proliferation and survival of placental fibroblast, exhibit anti-apoptotic effect and mediate nutrient availability at the fetoplacental unit [Lacey et al., 2002; Smith et al., 2002; Miller et al., 2005]. To date, the majority of published studies carried out to describe the alteration profile of the components of the IGF axis in preeclampsia are based on assays in maternal serum specimens. The extracted results do not always reflect the placental expression of the studied factors as the liver is an additional source of circulating factors in maternal serum or other posttranscriptional and/or posttranslational modifications maybe involved as it has been proposed for FGR placentas [Shin et al., 2003]. Regarding the early stages of gestation, when preeclampsia is in a pre-symptomatic stage, the published results are mainly focused on IGFs, IGFBP-1 and IGFBP-3 and are limited compared to the later stages where there is also discrepancy but to a lesser extent. The comparative analysis of these studies is a challenging task for several reasons such as the conflicting results, the differences in the study design, the sampling weeks, the dissimilar distribution of the severity and the subtypes of preeclampsia analyzed, the selection of the study population, the portion of each factor measured (total, bounded or free form) and the limitations of each study. Differentiations in the sensitivity and specificity of assays applied and the adjustment for confounding factors may also account for some of the variations observed

Several studies have accessed the interesting way in which the trophoblast-derived IGF-II and the decidual-derived IGFBP-I interact in the uterine environment in a paracrine manner and how this modulates trophoblast invasion in preeclampsia [Giudice et al., 1999; Irwin et al., 2001; Shin et al., 2003]. IGF-II m RNA is highly expressed by trophoblast with the greatest concentrations expressed at the invading front even from the 6th week of gestation, indicating a possible role of this peptide in cytotrophoblastic proliferation and the decidualization of the maternal stroma while IGFBP-I is the most plentifully expressed binding protein in the decidualized endometrium under the stimulatory effect of HCG and progesterone [Moy et al., 1996; Fowler et al., 2000]. Interestingly, the mean IGFBP-I m RNA and protein level were significantly elevated in preeclampsia compared with uncomplicated pregnancies in positive correlation with the severity of the disease [Shin et al., 2003]. In line with this observation, Giudice et al indicated the possible involvement of this protein in abnormal placentation and the same research group supported that the decidual derived IGFBP-I may lead to high maternal serum concentrations due to the increased vascular permeability of the leaky capillaries common in preeclampsia [Giudice et al., 1997]. As for the IGF-II, a significant decrease more profound in severe preeclampsia was confirmed both

across studies.

**2.3 IGFs** 

weight [Chellakooty et al., 2003]. Surprisingly, increased transcription of IGF-I and IGF-II not essentially associated with a corresponding increase in proteins levels have also been noted in IUGR pregnancies interpreted as a compensatory attempt against the inhibitory effect of either the enhanced level of IGFBPs and/or number of IGF2R or a response to impaired growth [Dalcik et al., 2001; Sheikh et al., 2001]. Alternatively, maternal diabetes seems to result in inverse changes of circulating fetal IGF-I and IGFBP-1 at birth leading to the proposal that a decrease in IGFBP-1 and to a lesser extent an increase in IGFs of cord blood samples may represent an important mechanism that contributes to macrosomia in these pregnancies [Lindsay et al., 2007].

Additionally, extensive data are available on the involvement of the IGFBPs in the abnormal fetal growth and maturity, chiefly highlighting the role of IGFBP-1 & IGFBP-3 in this process. Several research groups support that IGFBP-1 concentrations in amniotic fluid, maternal serum and fetal cord blood are predictive of newborn birth weight in both humans and mice; particularly, a striking inverse correlation between IGFBP-1 levels and fetal size is established although it is not clarified if an elevated level of circulating IGFBP-I is sufficient to cause fetal growth retardation or it depends on IGF-I mediated cell proliferation [Chard 1994; Crossey et al., 2002]. In contrast to IGFBP-1 and IGFBP-2 which correlate inversely with IGF-1 levels and birthweight, IGF-I and IGFBP-3 appear to be regulated in a coordinated manner, and show a significant positive relationship to parameters of fetal growth in both normal and complicated pregnancies despite the narrow reported conflicting results [Fant et al., 1993; Verhaeghe et al., 1993]. However, there are also reports of elevated IGFBP-3 levels in SGA infants and in low-birth weight offspring of mice over expressing the human IGFBP-3 gene [Holmes et al., 1997; Modric et al., 2001]. Also, disrupted IGF receptors' expression can also negatively affect fetal growth as it is demonstrated by a recent study reporting severe fetal growth restriction in two infants with a heterozygous missense mutation in the IGFIR gene [Walenkamp et al., 2006]. On the contrary, IGF-2R is thought to clear the IGF-II from the circulation and this is obviously supported by studies showing that mice lacking the IGF2R have greater birth weights than wild-type littermates [Lau et al., 1994; Efstratiadis, 1998].

#### **2.2 The role of IGF system in preeclampsia**

A central feature of preeclampsia is defective placentation expressed as shallow placental invasion limited to the superficial portion of decidua and abnormal remodeling of the endometrial vasculature due to failure of the conversion of the maternal spiral arteries into vessels of low resistance and high capacitance, as it has been supported by histological analysis of preeclamptic tissue specimens [Goldman-Wohl & Yagel, 2002]. Indirect evidence for impaired placental perfusion in pregnancies destined to develop preeclampsia has been provided by Doppler studies of the uterine arteries which showed increased pulsatility index (PI) from the first trimester of pregnancy [Martin et al., 2001; Poon et al., 2009b]. Immunohistochemical studies in human and animal placentas have meticulously demonstrated the placental synthesis of IGF system components; however, it is not accurate to draw conclusions for the development of human placenta based on animal models as the placenta is one of the very few organs that is unique to each species, both anatomically and functionally.

The evolutionary pattern of expression of IGFs, IGFBPs and IGF-1R genes in the developing human placenta and fetal membranes has been described from as early as 6 weeks' gestation to term in order to detect the exact expression cite of each peptide and the potential cite of its action [Han et al., 1996]. Functionally, IGF axis appears to be involved in many aspects of placental development and metabolism both in uncomplicated and preeclamptic pregnancies though the exact signaling pathways have yet to be determined [Hills et al., 2004]. Studies based on cultured human trophoblast cells and cell lines propose that IGFs promote proliferation, regulate trophoblast migration and the differentiation of cytotrophoblasts into syncytiotrophoblasts and extravillous cells, enhance the proliferation and survival of placental fibroblast, exhibit anti-apoptotic effect and mediate nutrient availability at the fetoplacental unit [Lacey et al., 2002; Smith et al., 2002; Miller et al., 2005].

To date, the majority of published studies carried out to describe the alteration profile of the components of the IGF axis in preeclampsia are based on assays in maternal serum specimens. The extracted results do not always reflect the placental expression of the studied factors as the liver is an additional source of circulating factors in maternal serum or other posttranscriptional and/or posttranslational modifications maybe involved as it has been proposed for FGR placentas [Shin et al., 2003]. Regarding the early stages of gestation, when preeclampsia is in a pre-symptomatic stage, the published results are mainly focused on IGFs, IGFBP-1 and IGFBP-3 and are limited compared to the later stages where there is also discrepancy but to a lesser extent. The comparative analysis of these studies is a challenging task for several reasons such as the conflicting results, the differences in the study design, the sampling weeks, the dissimilar distribution of the severity and the subtypes of preeclampsia analyzed, the selection of the study population, the portion of each factor measured (total, bounded or free form) and the limitations of each study. Differentiations in the sensitivity and specificity of assays applied and the adjustment for confounding factors may also account for some of the variations observed across studies.

#### **2.3 IGFs**

150 From Preconception to Postpartum

weight [Chellakooty et al., 2003]. Surprisingly, increased transcription of IGF-I and IGF-II not essentially associated with a corresponding increase in proteins levels have also been noted in IUGR pregnancies interpreted as a compensatory attempt against the inhibitory effect of either the enhanced level of IGFBPs and/or number of IGF2R or a response to impaired growth [Dalcik et al., 2001; Sheikh et al., 2001]. Alternatively, maternal diabetes seems to result in inverse changes of circulating fetal IGF-I and IGFBP-1 at birth leading to the proposal that a decrease in IGFBP-1 and to a lesser extent an increase in IGFs of cord blood samples may represent an important mechanism that contributes to macrosomia in

Additionally, extensive data are available on the involvement of the IGFBPs in the abnormal fetal growth and maturity, chiefly highlighting the role of IGFBP-1 & IGFBP-3 in this process. Several research groups support that IGFBP-1 concentrations in amniotic fluid, maternal serum and fetal cord blood are predictive of newborn birth weight in both humans and mice; particularly, a striking inverse correlation between IGFBP-1 levels and fetal size is established although it is not clarified if an elevated level of circulating IGFBP-I is sufficient to cause fetal growth retardation or it depends on IGF-I mediated cell proliferation [Chard 1994; Crossey et al., 2002]. In contrast to IGFBP-1 and IGFBP-2 which correlate inversely with IGF-1 levels and birthweight, IGF-I and IGFBP-3 appear to be regulated in a coordinated manner, and show a significant positive relationship to parameters of fetal growth in both normal and complicated pregnancies despite the narrow reported conflicting results [Fant et al., 1993; Verhaeghe et al., 1993]. However, there are also reports of elevated IGFBP-3 levels in SGA infants and in low-birth weight offspring of mice over expressing the human IGFBP-3 gene [Holmes et al., 1997; Modric et al., 2001]. Also, disrupted IGF receptors' expression can also negatively affect fetal growth as it is demonstrated by a recent study reporting severe fetal growth restriction in two infants with a heterozygous missense mutation in the IGFIR gene [Walenkamp et al., 2006]. On the contrary, IGF-2R is thought to clear the IGF-II from the circulation and this is obviously supported by studies showing that mice lacking the IGF2R have greater birth weights than wild-type littermates [Lau et al.,

A central feature of preeclampsia is defective placentation expressed as shallow placental invasion limited to the superficial portion of decidua and abnormal remodeling of the endometrial vasculature due to failure of the conversion of the maternal spiral arteries into vessels of low resistance and high capacitance, as it has been supported by histological analysis of preeclamptic tissue specimens [Goldman-Wohl & Yagel, 2002]. Indirect evidence for impaired placental perfusion in pregnancies destined to develop preeclampsia has been provided by Doppler studies of the uterine arteries which showed increased pulsatility index (PI) from the first trimester of pregnancy [Martin et al., 2001; Poon et al., 2009b]. Immunohistochemical studies in human and animal placentas have meticulously demonstrated the placental synthesis of IGF system components; however, it is not accurate to draw conclusions for the development of human placenta based on animal models as the placenta is one of the very few organs that is unique to each species, both anatomically and

these pregnancies [Lindsay et al., 2007].

1994; Efstratiadis, 1998].

functionally.

**2.2 The role of IGF system in preeclampsia** 

Several studies have accessed the interesting way in which the trophoblast-derived IGF-II and the decidual-derived IGFBP-I interact in the uterine environment in a paracrine manner and how this modulates trophoblast invasion in preeclampsia [Giudice et al., 1999; Irwin et al., 2001; Shin et al., 2003]. IGF-II m RNA is highly expressed by trophoblast with the greatest concentrations expressed at the invading front even from the 6th week of gestation, indicating a possible role of this peptide in cytotrophoblastic proliferation and the decidualization of the maternal stroma while IGFBP-I is the most plentifully expressed binding protein in the decidualized endometrium under the stimulatory effect of HCG and progesterone [Moy et al., 1996; Fowler et al., 2000]. Interestingly, the mean IGFBP-I m RNA and protein level were significantly elevated in preeclampsia compared with uncomplicated pregnancies in positive correlation with the severity of the disease [Shin et al., 2003]. In line with this observation, Giudice et al indicated the possible involvement of this protein in abnormal placentation and the same research group supported that the decidual derived IGFBP-I may lead to high maternal serum concentrations due to the increased vascular permeability of the leaky capillaries common in preeclampsia [Giudice et al., 1997]. As for the IGF-II, a significant decrease more profound in severe preeclampsia was confirmed both

Recent Insights into the Role of the Insulin-Like Growth Factor Axis in Preeclampsia 153

accelerated proteolysis of IGFBPs [Vatten et al., 2008]. Along with this thesis, a recent study demonstrated elevated second trimester levels of PGH in women, who later developed preeclampsia associated with intrauterine growth restriction leading to greater availability of nutrients for the fetoplacental unit [Papadopoulou et al., 2006]. However, we cannot offer an explanation for the apparent delay from first to second trimester for such a placental

Regarding the period that follows the clinical manifestation of preeclampsia, IGF-I levels seem to markedly decrease in preeclamptic subjects, particularly in the severe subtype of this disorder [Halhali et al., 2000; Ingec et al., 2004; Altinkaynak et al., 2003**]**. Even though it could be the result of the compromised production of placental PGH, relative studies based on a single point-in –time observations in preeclamptic pregnancies complicated further or not by fetal growth restriction are inconsistent and a longitudinal description of the correlation between IGF-I and PGH in different trimesters is lacking [Mittal et al., 2007; Schiessl et al., 2007] In contrast, two different research groups reported no difference between IGF-I values in preeclamptic women and normotensive controls, the one referred to preeclamptic women with late-onset disease and the other recruited women with

A lot of controversy exists regarding the role and the precise mechanisms of action of IGFBP-I at the maternal-fetal interface. On the one hand, IGFBP-I may act as a maternal restrain to trophoblast invasion via its inhibitory effect on mitogenic IGFs in decidual microenvironment [Ritvos et al., 1998]. On the other hand, Gleeson et al. doubt these results showing that IGFBP-I potentiates directly the human trophoblast migration by binding of its RGD domain (Arg-Gly-Asp) to the α5β1 integrin/fibronectin receptor on invading trophoblast leading to activation of mitogen activated protein kinase (MAPK) pathway [Gleeson et al., 2001]. This hypothesis is consistent with the observation that the regulation of α3, α5, β1, β4 integrin subunits (classes of adhesion molecules receptors) in trophoblast cells is altered in preeclamptic pregnancies [Zhou et al., 1993]. In vivo, it is still questionable which pathway of IGFBP-1 action is deregulated resulting in net suppression on trophoblast

A possible source of this inconsistency is the altered post-translational modification of IGFBP-1 in pregnant women [Forbes & Westwood, 2008]. In the circulation of non-pregnant women, IGFBP-1 exists only in its phosphorylated state (p-IGFBP-1) with high affinity for IGFs whereas during pregnancy, IGFBP-1 is extensively dephosphorylated to nonphosphorylated and intermediately phosphorylated isoforms (np-IGFBP-I) with a 6-fold lesser binding affinity for IGF-I and similar affinity for IGF-II [Westwood et al., 1994]. The functional significance of this modification is highlighted by the finding that np-IGFBP-1 enhanced the metabolic actions of IGF-I in nutrient transport in contrast to the inhibitory phosphorylated isoform [Yu et al., 1998]. A more complete interpretation of the data requires consideration of the distribution of the phosphoisoforms of IGFBP-I accompanying

response.

**2.4 IGFBPs 2.4.1 IGFBP-1** 

invasion.

preeclampsia of all subtypes [Lewitt et al., 1998].

preeclampsia in each trimester of pregnancy.

in m RNA and protein expression in preeclamptic placentas, by different research groups [Giudice et al., 1999; Shin et al., 2003]. Published data indicate that IGF-II stimulates proliferation and migration of cytotrophoblasts, regulates nutrient exchange but is not implicated in placental cell differentiation so the decreased levels of this factor may alter the invasive nature of cytotrophoblasts resulting in placental dysfunction [Shin et al., 2003]. On the contrary, Gratton et al. observed increased IGF-II m RNA abundance in the intermediate trophoblast surrounding placental infracts suggesting a role in placental repair or remodeling and decreased IGFBP-1 m RNA levels [Gratton et al., 2002]. Besides, in an invivo analysis of term placentas of gestations complicated by acute or chronic hypoxic ischemia, IGF-2 m RNA levels were elevated only in chronic hypoxia in contrast to IGF-I and PGH m RNA levels which did not significantly change, maybe as a part of a local metabolic adjustment [Trollmann et al., 2007].

IGF-II level in maternal serum is less investigated and up to now it seems not to significantly vary between preeclamptic and normotensive women [Giudice et al., 1997; Lewitt et al., 1998]. A possible explanation is that it does not share exactly the same regulation pattern with IGF-I as it is not primarily GH-dependent [Giudice et al., 1997]. On the contrary, there is a remarkable inconsistency among published results about the concentration of IGF-I in pregnancies destined to be preeclamptic. The only prospective cohort study that measured free rather than total IGF-I and IGFBP-1 levels and made adjustment for several confounding factors such as gestational age at blood collection, maternal age, parity, and prepregnacy adiposity, reported lower concentrations of these factors apparent in first trimester [Ning et al., 2004]. The attenuation of IGF-1 synthesis may result from impaired PGH synthesis as a consequence of the compromised placentation observed in preeclampsia [Ning et al., 2004]. Although several cross-sectional and longitudinal studies have found a positive correlation between serum PGH and IGF-1 values in pregnancies with fetoplacental unit disorders, our recent results do not provide support to the upper speculation as lower IGF-I levels but no alteration in PGH levels were observed in 11-13 weeks of gestation in pregnancies subsequently complicated by preeclampsia of either early or late onset [Sifakis et al., 2010; Sifakis et al. 2011a]. To elucidate this matter, it would be of special interest to investigate the range of the IGF-I levels in parallel with the change of PGH production throughout the disease transition from a latent preclinical stage to the clinically manifested preeclampsia.

Given that IGF-I acts positively on insulin sensitivity and the tissue availability of IGF-I is a significant determinant of insulin sensitivity in patients with essential hypertension, it could be presumed that decreased circulating levels of this factor would be responsible, at least in part, for insulin resistance in preeclampsia [Kocyigit et al., 2004]. However, Bartha et al. supported that increased insulin resistance is associated with gestational hypertension rather than preeclampsia and that there is no significant correlation between insulin sensitivity index and serum IGF-I levels [Bartha et al., 2002]. Also, there is no obvious explanation for the findings of other studies which recruited women before the clinical manifestation of the disease and reported that the maternal serum levels of IGF-I were either increased or not significantly altered [Hubinette et al., 2003; Vatten et al., 2008]. Based on a similar prospective approach, an increase in IGF-I from the first to second trimester and not within each trimester was associated with higher risk of preterm preeclampsia possibly interpreted as a compensatory mechanism to preserve fetal growth or/and a possible accelerated proteolysis of IGFBPs [Vatten et al., 2008]. Along with this thesis, a recent study demonstrated elevated second trimester levels of PGH in women, who later developed preeclampsia associated with intrauterine growth restriction leading to greater availability of nutrients for the fetoplacental unit [Papadopoulou et al., 2006]. However, we cannot offer an explanation for the apparent delay from first to second trimester for such a placental response.

Regarding the period that follows the clinical manifestation of preeclampsia, IGF-I levels seem to markedly decrease in preeclamptic subjects, particularly in the severe subtype of this disorder [Halhali et al., 2000; Ingec et al., 2004; Altinkaynak et al., 2003**]**. Even though it could be the result of the compromised production of placental PGH, relative studies based on a single point-in –time observations in preeclamptic pregnancies complicated further or not by fetal growth restriction are inconsistent and a longitudinal description of the correlation between IGF-I and PGH in different trimesters is lacking [Mittal et al., 2007; Schiessl et al., 2007] In contrast, two different research groups reported no difference between IGF-I values in preeclamptic women and normotensive controls, the one referred to preeclamptic women with late-onset disease and the other recruited women with preeclampsia of all subtypes [Lewitt et al., 1998].
