Emerging Roles of Antiangiogenic and Angiogenic Proteins in Pathogenesis and Prediction of Preeclampsia
Hypertensive disorders are a major cause of maternal and fetal death, especially in developing nations, with the pregnancy-specific disease preeclampsia-eclampsia responsible for most of this mortality. Preeclampsia, recognized by ancient Greeks, was only sporadically investigated until late last century, current literature still labeling this disorder a “disease of theories.” Thus, not surprisingly, we have started the new millennium unable to predict or prevent preeclampsia, and management strategies remain but supportive, with pregnancy termination still the only definitive “therapy.” This may be about to change, as focused research, mainly in the past decade, has resulted in dramatic progress regarding pathogenesis of disease manifestations and risk assessment, while investigators have begun designing and testing definitive therapy.1 There are still a myriad of uncertainties and unanswered questions to address, but here we summarize the extent of 1 new and exciting chapter in preeclampsia research: the roles of angiogenic and antiangiogenic factors, to which an article in this issue by Rana and colleagues,2 contributes.
A landmark report published in 2003 set the stage for all that followed.3 Dr Karumanchi’s laboratory, previously involved in renal cancer research, changed its focus to the kidney in preeclampsia. Using microarray chip technology, they probed gene expression profiles in placentas of women with preeclampsia and those with uncomplicated pregnancies. Of course many differences were noted, but one that caught the investigators’ attention was upregulation of the mRNA for the antiangiogenic protein soluble fms-like tyrosine kinase 1 (abbreviated as sFlt-1 or sVEGFR-1) in placentas from women with preeclampsia. Aware that de novo hypertension and proteinuria were 2 disturbing side effects of certain antiangiogenic compounds tested in cancer patients, these investigators wisely chose sFlt-1, the soluble receptor for vascular endothelial and placental growth factors (VEGF and PlGF, respectively), for further study. Their subsequent experiments and clinical protocols are a textbook example of how to proceed from patient to bench and back to patient with exciting findings during each trip (shall we call it ping pong or reverse “translation”?). These studies also illustrate the importance of maintaining faculty with skills to perform state of-the-art basic science embedded in clinical departments (currently a vanishing species, especially in surgically oriented specialties!).
First, the investigators hypothesized that sFlt-1 would enter the maternal circulation in excess, resulting in increased maternal levels that reduce circulating free (unbound) VEGF and PlGF. Indeed, they observed higher circulating sFlt-1 concentrations and lower free VEGF and PlGF levels in preeclampsia, results quickly confirmed in other reports (reviewed by Levine et al4). Back at the bench, the investigators showed that serum from women with preeclampsia inhibited tube formation (an assay relating to angiogenesis) and arteriolar vasodilatation in vitro, both observations reversed by VEGF. Next, adenovirus overexpression of sFlt-1 in pregnant rats produced 3 clinical features associated with preeclampsia: hypertension, proteinuria, and glomerular endotheliosis, the renal lesion that characterizes the disease in humans.1,5
The above observations, however, accounted for but 2 disease phenotypes, there being in fact other gravid animal models characterized by hypertension and proteinuria, some said to also manifest glomerular endotheliosis6 Preeclampsia, however, is a systemic disorder whose pathology involves other organ systems, most notably blood, liver, and brain. The investigators then searched for additional antiangiogenic factors, rapidly identifying soluble endoglin (sEng), a circulating receptor for transforming growth factor β1 (TGF-β1), whose concentrations were also increased in the placenta and maternal circulation during preeclampsia.7 Of interest, this second finding linked an antiangiogenic factor to another area of preeclampsia research, NO,2 as sEng impairs transforming growth factor β–binding to the endothelium receptor, and should thus decrease endothelial NO synthase (eNOS) mediated vasodilatation. The investigators again showed that sEng, like sFlt-1, could explain why sera from women with preeclampsia inhibit both angiogenesis and arteriolar vasodilatation in vitro, further demonstrating that sEng increased vascular permeability in mice. But even more exciting was that the simultaneous overexpression of sFlt-1 and sEng in gravid rats resulted in severe hypertension, marked proteinuria, liver enzyme abnormalities, and circulating schistocytes, an animal model simulating most of the protean manifestations of severe preeclampsia in humans.
To sum up, the studies described above did not identify the etiology of preeclampsia, but the data did explain the pathogenesis of key phenotypic features, including some related to its severest forms. In essence, the bench had provided for a logical approach to assess prediction, and eventually prevention or treatment. Indeed, commencing in 2004, the results in a number of reports (reviewed by Levine et al4) suggest that measurements of circulating or urinary proteins associated with the angiogenic state (blood, sFlt-1; sEng, PlGF; urine, PlGF) reliably predict preeclampsia. The most striking data were those of Levine et al4 who combined sEng levels and the ratio sFlt-1:PlGF and succeeded in predicting preeclampsia (particularly severe disease) ≈10 weeks before the onset of clinical signs and symptoms, with odd ratios ≈30 in some instances. A note of caution, however, is that studies, to date, have been largely retrospective and cross sectional, and in those reporting the most striking results the sera analyzed had been stored for ≈10 years.4 Thus, prospective observational studies are critically in order, and several are now in progress, including a World Health Organization protocol that aims at recruiting 10 000 participants.
One question that perplexed investigators was the following: The most plausible hypothesis regarding the initiating cause of preeclampsia, and thus the stimulus for excess production of antiangiogenic proteins, is the presence of abnormal placentation. This process takes place between gestational week 11 to 18, but neither sFlt-1 nor sEng concentrations appear elevated during this period (though decrements in PlGF have been noted). Here is where the study by Rana et al2 may provide insight. These authors show that between week 11 to 13 and 17 to 20 sFlt-1 levels remain essentially stable, and circulating sEng actually decreases during uncomplicated gestations, whereas both values increase in women destined to develop preeclampsia. These findings, as well as data of others,8 imply that stimuli for sFlt-1 and sEng overproduction may well begin during placentation, though more serial data would be needed to verify this. Incidentally, Rana et al2 assayed samples stored from 2 trials, one completed 10 years ago, the other with samples obtained many years later, and the similar values in the control groups finally provide data on the stability of the long stored samples.
Vattan et al,8 and now Rana and colleagues,2 suggest that changes over time in circulating antiangiogenic and angiogenic protein levels can be used to identify patients who are at risk for developing preeclampsia. Obviously, large studies are required to determine whether these approaches can achieve the likelihood ratios required for a useful prediction test. Recall that tests with high false-positive rates are potentially harmful, especially when they lead to unnecessary interventions, whereas an excessive false-negative rate may lead the clinician to a false sense of security. There are two potential pitfalls to consider in the quest toward acceptable likelihood ratios. First, an “antiangiogenic state” may underlie other pregnancy complications such as growth restriction, and second, smaller increments in sFlt1 and/or sEng levels may cause preeclampsia phenotypes in patients with “prepared” endothelium such as women with diabetes, and/or chronic hypertension.
Now that we may have reconciled concerns that stimuli for antiangiogenic overproduction were not associated with abnormal placentation, what might they be? Space precludes discussing the “usual suspects” like relative hypoxemia including genes stimulated by the hypoxemic environment, as well as inflammation, including cytokines and antioxidants. Newer candidates as mediators of the excess antiangiogenic state are the presence of agonistic autoantibodies to the angiotensin I receptor,9 abnormalities in the immune system involving natural killer cells in the decidua,10,11 and perhaps thrombin.12 Time will declare some of these or other candidates as winners, but it is pleasing to see a disorder neglected through the ages, and a “disease of theories” in contemporaneous texts, receiving the scrutiny of such good scientists.
Sources of Funding
This work was supported, in part, by the Division of Intramural Research of the National Institutes of Health.
The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.
Rana S, Karumanchi SA, Levine RJ, Venkatesha S, Rauh-Hain JA, Tamez H, Thadhani R. Sequential changes in antiangiogenic factors in early pregnancy and risk of developing preeclampsia. Hypertension. 2007; 50: 137–142.
Maynard SE, Min JY, Merchan J, Lim KH, Li J, Mondal S, Liberman TA, Morgan JP, Selke FW, Stillman IE, Epstein FH, Sukhatne VP, Kuramanchi SA. Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest. 2003; 111: 649–658.
Moran P, Lindheimer MD, Davison JM. The renal response to preeclampsia. Sem Nephrol. 2004; 24: 586–593.
Ventkatesha S, Torporsian M, Lam C, Hanai JI, Mammoto T, Kim YM, Bdolah Y, Kim KH, Yuan HT, Liberman TA, Stillman IE, Roberts D, D’Amore PA, Epstein FH, Solke FW, Romero R, Sukhatne VP, Letatrte M, Karumanchi SA. Soluble endoglin contributes to the pathogenesis of preeclampsia. Nature Med. 2006; 12: 642–649. Corrig, p 682.
Vatten LJ, Eskild A, Nilsen TL, Jeansson S, Jenum PA, Staff AC. Changes in circulating level of angiogenic factors from first to second trimester as predictors of preeclampsia. Am J Obstet Gynecol. 2007; 196: 239.e1–e6.
Zhou CC, Abbasi S, Xia L, Clare-Day M, Ramin S, Gilstrap LC, Kellems RE, Xia Y. Upregulation of placental soluble fms-like tyrosine kinase 1 (sFlt-1) by AT1 receptor agonistic antibodies in preeclampsia. Hypertens Preg. 2006; 25: 38.
Hanna J, Goldman-Wohl D, Hamani Y, Avraham I, Greenfield C, Nathanson-Yaron S, Cohen-Daniel L, Arnon TI, Manaster I, Gazit R, Yutkin V, Benharroch D, Porgadon A, Kishet E, Yagel S, Mendelboin O, Decidual NK. Cells regulate key development processes at the human fetal-maternal interface. Nature Med. 2006; 12: 1065–1074.
Moffet A, Hiby SF. How does the maternal immune system contribute the development of preeclampsia. Placenta. 2007; 21: S51–S56.