|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Minireview |
Department of Obstetrics and Gynecology,2 University of Leipzig, 04103 Leipzig, Germany
Department of Cell Biology,3 Histology and Embryology, Medical University of Graz, 8010 Graz, Austria
Department of Molecular Biological-Biochemical Processing Technology,4 Center of Biotechnology and Biomedicine, University of Leipzig, 04103 Leipzig, Germany
Department of Pharmacology,5 Erasmus Medical Center, 3000 Rotterdam, The Netherlands
Department of Cardiology,6 Charité, Campus Benjamin Franklin, 12200 Berlin, Germany
Despite recent research progress, the biology of preeclampsia is still poorly understood and neither effective prediction nor causal therapy have yet emerged. Nevertheless, recent studies have documented new and exciting pathophysiological mechanisms for the origin and development of preeclampsia. These studies provide a more differentiated view on alterations of particular peptide systems with strong impact on angiogenesis and cardiovascular regulation in this pregnancy disorder. With the identification of the antiangiogenic factor soluble fms-like tyrosine kinase 1 and the agonistic autoantibody to the angiotensin II type 1 receptor, two factors have been described with a clear linkage to the development of the disease. This review focuses on the most recent and relevant insights into the biology of preeclampsia and develops hypotheses regarding possible links between the reported aspects of preeclampsia.
placenta, pregnancy, signal transduction, trophoblast
There is growing evidence that an imbalance between factors promoting angiogenesis such as vascular endothelial growth factor (VEGF) or placental growth factor (PGF) and factors antagonizing angiogenesis such as soluble fms-like tyrosine kinase 1 (soluble FLT1) have a fundamental role in the pathogenesis of preeclampsia. In particular, soluble FLT1 has become a strong candidate as a mediator and relevant pathogenic factor of the syndrome. As a splice variant of the VEGF receptor FLT1, this secreted form of FLT1 is a potent inhibitor of VEGF and PGF. In preeclampsia, placental soluble FLT1 is upregulated, leading to dramatically increased soluble FLT1 concentrations in the maternal blood that normalize after delivery. These increased soluble FLT1 concentrations are accompanied by decreased levels of free VEGF and PGF, suggesting that soluble FLT1 binds VEGF and PGF in the maternal circulation and thereby blocks their angiogenic effects [1]. Even though soluble FLT1 was first described by Kendall and Thomas [2] in 1990 and initially characterized in the placenta in 1998 by Clark et al. [3], its biological function still remains to be elucidated. Recently several groups investigated the role of soluble FLT1 in animal models. Maynard et al. [4] demonstrated that administration of soluble FLT1 in a rat model causes elevated blood pressure, proteinuria, and renal changes, the classic lesions of preeclampsia. Moreover, the resulting endothelial dysfunction has been rescued by exogenous VEGF and PGF administration. Furthermore, soluble FLT1 is able to inhibit the microvascular relaxation of rat renal arteries induced by VEGF and PGF [4]. In addition, Hirashima et al. [5] demonstrated that mice with disrupted placental soluble FLT1 show normal placentation with an undisturbed maternal-fetal interface. Also, a number of human studies add considerable weight to the mounting evidence that soluble FLT1 is a significant contributor to preeclampsia. Although there are first data on altered soluble FLT1 concentration in newborns from preeclamptic pregnancies [6], soluble FLT1 plays its major role on the maternal side. Adenovirus-mediated soluble FLT1 transfer also causes preeclamptic symptoms in nonpregnant animals, indicating that the effect is direct and independent of the presence of a placenta [4]. A clinical study showed that maternal soluble FLT1 concentrations remained unchanged until 3336 wk of gestation and afterwards rose until delivery [1]. In preeclamptic patients, several groups have measured excessively high soluble FLT1 concentrations [1, 7, 8] being directly proportional to proteinuria and correlating inversely with platelet count and neonatal weight [9]. Moreover, soluble FLT1 concentrations seem to be higher in the first pregnancy compared to the second pregnancy of a woman, which is in line with the known increased risk for preeclampsia in nulliparous women [10]. A large clinical study demonstrated that increased soluble FLT1 plasma concentrations are detectable about 5 wk before the onset of preeclampsia [1]. The fact that the soluble FLT1 increase antedates the clinical manifestation of the disease might be useful for early clinical screening. However, this excess of soluble FLT1 does not seem to be totally specific to preeclampsia, because several studies observed this soluble FLT1 excess in intrauterine growth restriction as well [7, 11]. In addition, recent data of our group demonstrated that second-trimester women with abnormal uterine perfusion and subsequent pregnancy complications have elevated serum soluble FLT1 concentrations [7]. Thus, a combination of Doppler measurement of uterine perfusion and detection of maternal soluble FLT1 concentrations could be an early and efficient screening test for preeclampsia or pathological pregnancies in general. The measurement of urinary soluble FLT1 and PGF as previously reported would allow an even easier screening [12, 13]. Levine et al. [12], for instance, report significantly reduced urinary PGF concentrations beginning at 2528 wk of gestation, before the onset of preeclampsia. With manifest disease, preeclamptic patients showed a mean PGF concentration of 32 pg/dl, compared to a mean concentration of 234 pg/dl in the control group. This was confirmed by another study that also showed significantly decreased urinary PGF concentrations in preeclamptic women combined with increased soluble FLT1 concentrations. Regarding the differentiation of preeclamptic patients and healthy controls, measurements in this study reach a sensitivity of 88.2% and a specificity of 100% in the ratio log (soluble FLT1 /PGF) [13].
Thinking one step further, the imbalance between proangiogenic and antiangiogenic factors could also be a target for future therapeutic interventions. In first attempts to use an adenovirus-mediated soluble FLT1 gene therapy to inhibit the proliferation of ovarian carcinomas, Mahasreshti et al. [14] demonstrated an effective inhibition of tumor growth and increased survival times in mice with intraperitoneally inoculated ovarian carcinoma cells. Moreover, there are first experiments of a VEGF-induced angiogenic gene therapy in patients with peripheral artery disease. In these patients, the intramuscular injection of human VEGF165-containing DNA caused significant improvement of the clinical symptoms [15]. Thus, knowledge of potential therapeutic applications gained from investigations of tumor angiogenesis and other fields might be useful to target placental malfunction by changing the proportion of angiogenesis to antiangiogenesis in favor of angiogenesis.
At present, it is unclear which factors trigger the excessive upregulation of placental soluble FLT1 production in pregnancies with upcoming pathology (Fig. 1). One factor might be hypoxia, because it has been demonstrated that hypoxia triggers upregulation of soluble FLT1 in trophoblast cells [16]. This sounds conclusive, because the postulated hypoxic condition in the preeclamptic placenta might then contribute the soluble FLT1 release. Furthermore, exposure of normal villous explants to hypoxia increases soluble FLT1 release. Treatment of endothelial cells with conditioned medium from preeclamptic women results in reduced angiogenesis. This effect can be reversed by soluble FLT1 removal via immunoprecipitation [17]. Therefore, soluble FLT1 could be one of the factors released by the ischemic placenta into the maternal circulation causing disseminated endothelial dysfunction and impaired angiogenesis. Not only hypoxia but also other factors are able to trigger soluble FLT1 release. According to Kim et al. [18] angiotensin II (AGT II) induces an increased soluble FLT1 production in human proximal tubule cells. These findings provide interesting links to the renin-angiotensin system, which also interacts with angiogenesis. Furthermore, in accordance with other groups, we observed that AGT II can regulate angiogenesis in vivo depending on which of its receptors (AGTR1 or AGTR2) is activated [19, 20].
|
Preeclamptic patients are characterized by increased vascular responsiveness to AGT II but unaltered circulating AGT II concentrations. Thus, one of the key events in the etiology of preeclampsia may not be the altered stimulation of AGTR1 by AGT II but an altered intracellular signaling of the receptor initiated by modified interaction of AGTR1 with other G protein-coupled receptors (GPRs) as the phenomenon of receptor dimerization/oligomerization [21]. Interestingly, previous studies have shown that AGTR1 can form hetero-oligomeric complexes with certain other GPRs. For example, it has been proposed that dimerization between AGTR1 and the bradykinin B2 receptor results in enhanced function of AGT II, and may underlie much of the hypertension associated with the condition of preeclampsia in pregnant women [22]. Interactions between AGTR1 and AGTR2 have also been demonstrated in vitro, but in such circumstances AGTR2 appears to functionally antagonize AGTR1 [23]. Another GPR was recently introduced by our group as a new candidate that might influence AGTR1 signaling by physical interaction: the GPR MAS1. This receptor is encoded by the Mas proto-oncogene that was initially detected through its tumorigenic activity in in vivo tumor assays [24]. Recent studies suggested a physiological role of MAS1 in the function of AGT II. For example, von Bohlen und Halbach et al. [25] demonstrated an alteration of neuronal AGTR1 signaling after AGT II stimulation in mice with disrupted Mas proto-oncogene [26]. These findings provided the first strong evidence for an in vivo interaction between the GPR MAS1 and AGTR1.
In vitro the MAS1 receptor inhibits AGTR1, and in some tissues this oligomeric interaction may represent a natural state for these receptors in vivo [27]. Together with the dimerization demonstrated in bioluminescence resonance energy transfer studies, our findings in native tissues suggest that the MAS1 receptor can act in vivo as a functional antagonist of AGTR1 because of formation of a hetero-oligomeric complex. Although in Mas transfected mammalian cells MAS1 was not activated by AGT II, the AGTR1-mediated and AGT II-induced production of inositol phosphates and mobilization of intracellular Ca2+ was diminished by 50% following coexpression of Mas, despite a concomitant increase in AGT II binding capacity [27]. It seems likely that these collective findings indicate a novel property of GPR signaling: the direct regulation of signal transduction via one GPR by the physical presence of another. Furthermore, the absence of the "brake" function of MAS1 could result in an increased reactivity of AGTR1 under AGT II stimulation, as observed in preeclamptic patients. Although to date there are no data on MAS1 regulation in vascular and trophoblast cells under preeclamptic conditions, we also identified MAS1, beside its interaction with AGTR1, as an endogenous receptor for the angiotensin metabolite AGT-(17) [28]. Both the heptapeptide and the peptidase ACE2, responsible for the generation of AGT-(17), are strongly regulated in pregnancy [29] and altered in preeclampsia [30].
Further findings support the key role of the renin-angiotensin system in the etiology of preeclampsia as well. In 1999, Wallukat et al. [31] identified an agonistic autoimmune antibody against AGTR1 (AGTR1-AA) in preeclamptic patients but not in healthy pregnancies or those with essential hypertension. As AGTR1 stimulation by this agonistic AGTR1-AA leads in vitro to reduced trophoblast invasiveness (a typical characteristic of preeclampsia), a possible causality of AGTR1-AA for preeclampsia has been postulated [32]. Moreover, the antibody induces Ca++ release in vascular smooth muscle cells and could therefore mediate the vascular alterations in preeclampsia [33]. Because the cause for AGTR1-AA production during pregnancy is still unclear, while AGTR1 is expressed independently of pregnancy in a variety of organs, the question of whether there might be a link between receptor heterodimerization and AGTR1-AA production arose. One of the reasons for AGTR1 antigen presentation could be the altered physical interaction of AGTR1 with other receptors as model-like, demonstrated above, for AGTR1 and MAS1. As indicated in Figure 2A, the monomeric or homodimeric AGTR1 (normal pregnancy) could interact with another receptor early in the development of preeclampsia, or another intracellular factor (factor Y) could bind to AGTR1 under these conditions. This physical interaction alters the conformation of AGTR1 and therefore exposes the 7-amino acid epitope, identified by Wallukat et al. [31], for immunological response. This conformational change allows AGTR1 stimulation by AGTR1-AA and subsequent excessive pathologic signaling. Also, the inverted model lacking such a cofactor, as demonstrated in Figure 2B, would lead to the same final result. Whereas during a normal pregnancy the receptor interacts with another receptor or factor X, this mutual effect is lacking in the pathological situation, and thus the receptor conformation again changes, with the fatal consequences described above. What data in recent research may support our hypothesized factors X or Y as AGTR1 modulators? Besides the above discussed intramembrane interaction of AGTR1 with other G protein-coupled receptors, these possible factors could be circulating substances, intracellular proteins, or biochemical modifications. For instance, it has been demonstrated that high glucose concentrations or plasma insulin levels have AGTR1-modulating properties [34, 35]. Moreover, the cytoplasmic AGT II type I receptor-associated protein directly interacts with AGTR1 in a cell-type-specific manner and thus can modulate AGTR1 behavior [36, 37]. Nevertheless, the factors could also represent a simple chemical receptor modulation. In particular, N-glycosylation is able to alter targeting, affinity, and quality control of the human AGTR1 [38]. Although proof for the causality of these interactions/modifications in epitope presentation is still missing, this hypothesis may add to existing theories of auto-antibody generation such as molecular mimicry or vascular damage.
|
Because both AGTR1-AA and soluble FLT1 are promising candidate factors triggering the early pathophysiological changes of later preeclamptic features, a possible link between AGTR1-AA and soluble FLT1 can be hypothesized. As mentioned in the first section, AGT II stimulates soluble FLT1 [18]. Hence, the agonistic antibody detectable in the second trimester of pregnancy, as discovered by our group recently [39], could be directly responsible for the development of reduced trophoblast invasiveness and then could indirectly (by inducing soluble FLT1) promote other characteristics of preeclampsia such as renal dysfunction. Because both circulating factors cause endothelial dysfunction, the parallel expression of both could have additive effects in damaging the endothelium. Thus, although decisive evidence that one of the two factors is causative for preeclampsia has not yet been provided, it may be hypothesized that concerted regulation is needed to mediate the initial pathology for preeclampsia. Consequently, clinical studies need to investigate the correlation between AGTR1-AA and soluble FLT1. However, the complicated detection of the agonistic antibody by a bioassay using neonatal rat cardiomyocytes excludes a high number of screened pregnancies. Thus, new detection methods have to be developed to make AGTR1-AA screening easier and faster.
To solve this problem, Rothermel et al. [40] developed a novel functional cardiomyocyte-based microarray for detecting AGTR1 autoimmune antibody. It was shown that this neonatal rat cardiomyocyte model expressed all functional components of the rennin-angiotensin system and could serve as a suitable biological sensor system to investigate AGTR1-mediated dysfunctions. Thus, electrophysiologically active neonatal rat cardiomyocytes were coupled to a multielectrode array and used as a biological sensor that serves as a highly sensitive, reliable, and reproducible cardiomyocyte based monitoring system. This cell-based multielectrode array consisting of 60 substrate integrated microelectrodes was successfully validated 1) to determine the influence of parameters, e.g., cell adhesion and temperature, on electrophysiological activity and contractibility, 2) to determine the time-dependent stability of contraction rates, 3) to determine the dose-dependent activity of AGT II for the evaluation of the biosensor and its sensitivity, 4) to determine specific effects of AGT II via the AGTR1 antagonist losartan, and (v) to demonstrate the bioelectronic monitoring of AGTR1-AA in sera of preeclamptic pregnant women.
Currently, the two circulating factors AGTR1-AA and soluble FLT1 are discussed as promising candidates with a potential role in the pathophysiology of preeclampsia. Our review attempts to show possible causes for the generation of AGTR1-AA as well as for the excessive soluble FLT1 increase during preeclampsia and to link the two factors that might be mutually regulated. A reasonable pathophysiological connection between these two axes would open a new perspective in describing the causality and biology of preeclampsia. It may be hypothesized that both factors influence each other and that their concertive expression is needed to mediate the preeclamptic phenotype. To investigate the possible correlation between AGTR1-AA and soluble FLT1, parallel measurements of both factors are needed in pregnancies with manifest preeclampsia and especially in second-trimester (or even first-trimester) pregnancies with pathological uterine perfusion, because those pregnancies are at high risk of developing preeclampsia later on.
1 Correspondence: Holger Stepan, Department of Obstetrics and Gynecology, Philipp-Rosenthalstr. 55, 04103 Leipzig, Germany. FAX: 49 341 9723599; holger.stepan{at}medizin.uni-leipzig.de ![]()
Received: 30 July 2005.
First decision: 23 August 2005.
Accepted: 17 January 2006.
This article has been cited by other articles:
![]() |
F. Collino, B. Bussolati, E. Gerbaudo, L. Marozio, S. Pelissetto, C. Benedetto, and G. Camussi Preeclamptic sera induce nephrin shedding from podocytes through endothelin-1 release by endothelial glomerular cells Am J Physiol Renal Physiol, May 1, 2008; 294(5): F1185 - F1194. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. GULLER, Y. Y. MA, H.-H. FU, G. KRIKUN, V. M. ABRAHAMS, and G. MOR The Placental Syncytium and the Pathophysiology of Preeclampsia and Intrauterine Growth Restriction: A Novel Assay to Assess Syncytial Protein Expression Ann. N.Y. Acad. Sci., April 1, 2008; 1127(1): 129 - 133. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. P. Thomas, J. I. Andrews, and K. Z. Liu Intronic polyadenylation signal sequences and alternate splicing generate human soluble Flt1 variants and regulate the abundance of soluble Flt1 in the placenta FASEB J, December 1, 2007; 21(14): 3885 - 3895. [Abstract] [Full Text] [PDF] |
||||
![]() |
A.-K. Wikstrom, A. Larsson, U. J. Eriksson, P. Nash, S. Norden-Lindeberg, and M. Olovsson Placental Growth Factor and Soluble FMS-Like Tyrosine Kinase-1 in Early-Onset and Late-Onset Preeclampsia Obstet. Gynecol., June 1, 2007; 109(6): 1368 - 1374. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y.-L. Dong, D. M. Reddy, K. E. Green, M. S. Chauhan, H.-Q. Wang, M. Nagamani, G. D.V. Hankins, and C. Yallampalli Calcitonin Gene-Related Peptide (CALCA) Is a Proangiogenic Growth Factor in the Human Placental Development Biol Reprod, May 1, 2007; 76(5): 892 - 899. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. C. Zhou, S. Ahmad, T. Mi, L. Xia, S. Abbasi, P. W. Hewett, C. Sun, A. Ahmed, R. E. Kellems, and Y. Xia Angiotensin II Induces Soluble fms-Like Tyrosine Kinase-1 Release via Calcineurin Signaling Pathway in Pregnancy Circ. Res., January 5, 2007; 100(1): 88 - 95. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |