SFLT 1
SFLT 1
SFLT 1
69–78, 2017
Advanced Access publication on December 28, 2016 doi:10.1093/molehr/gaw077
*Correspondence address. Department of Obstetrics and Gynaecology, Monash University, Monash Medical Centre, 246 Clayton Rd,
Clayton, 3168 Victoria, Australia. E-mail: kirsten.palmer@monash.edu
Submitted on August 10, 2016; resubmitted on November 16, 2016; editorial decision on December 2, 2016; accepted on December 9,
2016
ABSTRACT: Pre-eclampsia is a common obstetric complication globally responsible for a significant burden of maternal and perinatal mor-
bidity and mortality. Central to its pathophysiology is the anti-angiogenic protein, soluble fms-like tyrosine kinase-1 (sFLT-1). sFLT-1 is
released from a range of tissues into the circulation, where it antagonizes the activity of vascular endothelial growth factor and placental
growth factor leading to endothelial dysfunction. It is this widespread endothelial dysfunction that produces the clinical features of pre-
eclampsia including hypertension and proteinuria. There are multiple splice variants of sFLT-1. One, known as sFLT-1 e15a, evolved quite
recently and is only present in humans and higher order primates. This sFLT-1 variant is also the main sFLT-1 secreted from the placenta.
Recent work has shown that sFLT-1 e15a is significantly elevated in the placenta and circulation of women with pre-eclampsia. It is also bio-
logically active, capable of causing endothelial dysfunction and the end-organ dysfunction seen in pre-eclampsia. Indeed, the over-expression
of sFLT-1 e15a in mice recapitulates the pre-eclamptic phenotype in pregnancy. Therefore, here we propose that sFLT-1 e15a may be the
sFLT-1 variant primarily responsible for pre-eclampsia, a uniquely human disease. Furthermore, this placental-specific sFLT-1 variant provides
promise for use as an accurate biomarker in the prediction or diagnosis of pre-eclampsia.
Key words: pre-eclampsia / anti-angiogenic factors / VEGF / placenta / hypoxia / sFLT-1 / splice variant
© The Author 2016. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
70 Palmer et al.
antagonizing the actions of both vascular endothelial growth factor demonstrated an inversely proportional association between the levels
(VEGF) and placental growth factor (PlGF). Multiple splice variants of of sFLT-1 in the serum of pre-eclamptic women and serum concentra-
sFLT-1 have been described (Sela et al., 2008; Heydarian et al., 2009; tions of VEGF and PlGF. They also demonstrated the impact of an
Thomas et al., 2009). These have significantly different tissue distribu- altered angiogenic balance on endothelial cell function. Their work
tions (Jebbink et al., 2011), raising the potential for different physiological showed that human umbilical vein endothelial cells (HUVECs) exposed
and pathological roles. For example, in humans, the main sFLT-1 variant, to pre-eclamptic serum failed to form tubular structures (Maynard
known as sFLT-1 i13, is widely expressed in most tissues, whereas et al., 2003), suggesting the presence of an anti-angiogenic factor within
another variant known as sFLT-1 e15a appears to be almost exclusively the patient serum. Adding exogenous VEGF and PlGF to the pre-
expressed by the placenta (Jebbink et al., 2011). eclamptic serum reversed this effect, further supporting this premise.
Ongoing progression in the understanding of the pathophysiology of This was confirmed, as serum from healthy pregnant women induced
pre-eclampsia is critical to assisting in improving the clinical care for endothelial tube formation, an effect that was inhibited by the addition
affected women. Through better understanding, improved approaches of exogenous sFLT-1. The final confirmation of the involvement of
for disease prediction, diagnosis and therapies can be developed. sFLT-1 in pre-eclampsia came with the adenoviral over-expression of
Lee, 1999). The presence of sFLT-1 in vivo was subsequently con- et al., 2008; Heydarian et al., 2009; Thomas et al., 2009). The signifi-
firmed in the placenta and serum of pregnant women with pre- cance of these different splice variants currently remains unknown.
eclampsia (Clark et al., 1998), where its presence antagonized the Alternative splicing has been found to occur in >80% of genes in
actions of VEGF. the human genome with the majority of splicing events affecting the
VEGF plays a crucial role in angiogenesis and normal endothelial cell coding sequence (Matlin et al., 2005). Splicing events are regulated by
function. Its bioavailability is strictly regulated and alterations in its cir- cis- and trans-elements. The former are sequences within the pre-
culating levels have a negative impact on endothelial cell function. This mRNA that help to direct splicing, such as exon splicing enhancers or
is highlighted by the drastic phenotypes exhibited through the effects silencers (ESEs or ESSs, respectively) and their intronic counterparts
of both under- and over-expression of VEGF on endothelial cell func- (ISEs and ISSs), as well as polyadenylation signals within the sequence.
tion in the kidney. Mice lacking Vegf solely within the podocytes (a vital Trans-elements are cellular factors such as RNA or protein that regu-
component of the glomerular basement membrane) die within late splicing. For example, the SR family of RNA-binding proteins bind
18 hours of birth due to hydrops, renal failure and abnormal glomeruli to ESEs and assist in the assembly of the spliceosome complex (Matlin
(Eremina, 2003). Podocytes with heterozygous Vegf expression are et al., 2005).
Flt-1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
sFlt-1 1 2 3 4 5 6 7 8 9 10 11 12 13
sFlt-1 v4 13 14 i14
Figure 1 Schematic representations of FLT-1 splice variants. The FLT-1 gene encodes the full-length membrane-bound FLT-1 receptor VEGFR-1, as
well as the four soluble alternative sFLT-1 splice variants. All transcripts are identical to exon 13, with the soluble transcripts each identified by a unique
C-terminal region. Of these, sFLT-1 i13, sFLT-1 e15a and sFLT-1 v4 are translated into protein with only the former two being increased in pre-
eclampsia. Reprinted with permissions from Palmer et al. (2015).
14 exons of FLT-1 followed by a 480 nucleotide stretch of intronic body (Rajakumar et al., 2005; Suwaki et al., 2006; Souders et al.,
sequence (Sela et al., 2008). This encodes the unique alternatively 2015). However, with the exception of human brain microvascular
spliced exon 15a followed by a 3′-UTR containing the complete AluSq endothelial cells, it is always expressed to a lesser extent than the
sequence. A polyadenylation signal sequence inserted as a result of the membrane-bound FLT-1 (Jebbink et al., 2011). While this suggests a
Alu sequence likely directs the alternative splicing of exon 15a, which tissue-specific expression profile of FLT-1 transcripts, expression also
encodes a unique polyserine tail (Thomas et al., 2007). The resulting appears to be cell-type specific. In blood vessels, endothelial cells
protein shares 706 amino acids with FLT-1 and has a unique 28 amino appear to exclusively produce the i13 variant, whereas the underlying
acid tail, producing a protein that is 95–135 kDa in size due to varia- vascular smooth muscle exclusively expresses the e15a variant (Sela
tions in the level of glycosylation (Gu et al., 2008). Importantly, sFLT-1 et al., 2008). Furthermore, there is the potential for temporal regula-
e15a is absent from all mammals except humans and higher order pri- tion of expression, as splicing patterns have been shown to alter over
mates, highlighting it as potentially the key variant underlying pre- the course of pregnancy in mice. In murine placenta, Flt-1 predomi-
eclamptic pathophysiology. nates early in gestation before sFlt-1 becomes more highly expressed
from the second trimester onwards (He et al., 1999b). Whether this
holds true in humans remains to be fully characterized; however, it is
sFLT-1 splice variant mRNA possible given the rise in circulating sFLT-1 seen with advancing gesta-
tion (Levine et al., 2004; Palmer et al., 2015).
expression and regulation Within the placenta, sFLT-1 e15a is mainly produced by the syncy-
Both sFLT-1 i13 and sFLT-1 e15a are capable of binding to VEGF and tiotrophoblast and cytotrophoblast, where its mRNA expression is
antagonizing its actions (Kendall et al., 1994; Sela et al., 2008; Palmer 100-fold greater than in placental vascular smooth muscle cells, endo-
et al., 2015). The i13 variant is also known to antagonize PlGF (Kendall thelial cells and macrophages (Thomas et al., 2009). In the setting of
et al., 1994) and it is likely the e15a variant does too, given the preser- pre-eclampsia, all FLT-1 transcripts appear up-regulated (Jebbink et al.,
vation of the ligand-binding site. Interestingly, the e15a variant is the 2011); however, sFLT-1 e15a shows the greatest increase, with levels
predominant isoform in human placenta compared to i13, which is the appearing to correlate with severity of pre-eclampsia (Whitehead
main variant in rhesus monkey placenta (Thomas et al., 2009). et al., 2011).
Furthermore, in humans sFLT-1 e15a appears to be almost exclusively Placental hypoxia is a key factor in pre-eclamptic pathophysiology.
expressed by the placenta where its expression is ≥600-fold higher Many transcriptional regulators have been shown to be involved in
than in the other organ systems where it is found (Jebbink et al., 2011). increasing the expression of both FLT-1 and sFLT-1 in response to hyp-
Certainly within the placenta, sFLT-1 e15a is the predominant isoform oxia, such as hypoxia inducible factor (Gerber et al., 1997; Nagamatsu
of FLT-1 mRNA, constituting >80% of transcripts compared to sFLT-1 et al., 2004; Tal et al., 2010), growth arrest and DNA damage 45
i13, which constitutes 12%, and membrane-bound FLT-1, which is (Xiong et al., 2009) and early growth response protein 1 (Vidal et al.,
<5% (Jebbink et al., 2011). There are many extra-placental sources of 2000). Similarly, many key molecules in the hypoxic response also
sFLT-1 (He et al., 1999b; Jebbink et al., 2011), which are likely to increase the expression of sFLT-1, such as VEGF (Gerber et al., 1997)
mainly be the i13 isoform as it is widely expressed throughout the and adenosine through the up-regulation of a key enzyme, CD73
Placental-specific sFLT-1 73
(Iriyama et al., 2015). Furthermore, the mechanism controlling the expression, as outlined in Table I. Interestingly, in vitro treatment with
regulation of alternative splicing of FLT-1 also appears to be hypoxia both YC-1 and pravastatin led to a significant reduction solely in sFLT-1
dependent. Recently, it has been shown that a nuclear protein, known e15a mRNA expression with no significant effect seen on the sFLT-1
as jumonji domain containing protein-6 (JMJD6) is a regulator of alter- i13 variant (Brownfoot et al., 2015a,b). Metformin led to a significant
native splicing of the FLT-1 pre-mRNA (Boeckel et al., 2011). Under reduction in sFLT-1 i13 mRNA expression in endothelial cells and sFLT-1
normoxic cellular conditions, JMJD6 is able to interact with a compo- e15a mRNA expression from cytotrophoblast and placental explants
nent of the spliceosome complex, U2AF65, hydroxylating it. (Brownfoot et al., 2016), while treatment with ouabain led to a signifi-
Hydroxylated U2AF65 then directs the splicing machinery to produce cant reduction in both sFLT-1 i13 and e15a variant expression in pla-
full-length FLT-1. Importantly, they were able to show that under hyp- cental explants, though with a trend towards greater repression of the
oxic conditions, there was inhibition of the interaction of JMJD6 and e15a variant (Rana et al., 2014). These findings all support the import-
U2AF65, as JMJD6 requires oxygen for its enzymatic function. Under ance of sFLT-1 in pre-eclampsia and potentially that targeting therapies
these conditions, within endothelial cells, non-hydroxylated U2AF65 that reduce the placental-specific variant is crucial for effectively redu-
directed splicing to produce sFLT-1 (Boeckel et al., 2011). We have cing circulating sFLT-1 in early-onset pre-eclampsia.
Table I Effects of potential pre-eclamptic therapeutics on sFLT-1 variant mRNA expression level in vitro.
Drug Mechanism of action Cells treated Culture Length of [Drug] sFLT-1 i13* sFLT-1
(human) conditions treatment e15a*
(hours)
.............................................................................................................................................................................................
Metformin Inhibition of complex I in the HUVEC 10 ng/ml TNF-α, 24 2 mM <0.05 N/A
(Brownfoot et al., mitochondrial transport chain 20% O2 5 mM <0.01
2016)
Villous CTB 8% O2 48 2 mM N/A <0.00001
5 mM <0.00001
Placental explants 8% O2 72 2 mM N/A <0.05
5 mM <0.0001
Ouabain (Rana et al., Inhibition of HIF-1α through Normal and PET 21% O2 72 500 nM <0.05 <0.05
2014) phosphorylation of HSP27 placental explants
Pravastatin Inhibiting the HMG-CoA HUVEC 1% FCS and 6 200 μmol/l NS <0.001
(Brownfoot et al., reductase pathway 10 ng/ml TNF-α 2000 μmol/l NS <0.0001
2015a)
Early-onset PET 20% O2 48 2000 μmol/l NS <0.05
placental explants
Pre-/post-delivery- 20% O2 48 200 μmol/l NS NS
treated explants 2000 μmol/l NS <0.0001
[Drug], drug concentration; TNF, tumour necrosis factor; NS, not significant; N/A, not available; HIF-1α, hypoxia inducible factor 1α; HSP27, heat-shock protein 27; NO, nitric oxide;
PET, pre-eclampsia; FCS, foetal calf serum; HMG-CoA, 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase and CTB, cytotrophoblast. All cells tested were cultured under standard
conditions at 37°C in 5% CO2 with oxygen conditions as stated.
*All statistically significant changes refer to a decrease in sFLT-1 variant mRNA expression in experimental groups compared to control treated cells. A t-test was used for statistical
analysis in all studies.
74 Palmer et al.
particular sFLT-1 splice variant of interest. These antibodies, devel- conditions of placental malfunction, such as foetal growth restriction
oped in the non-commercial setting, enable the assessment of how the (Palmer et al., 2016a), although this requires further investigation.
translated sFLT-1 proteins function and their expression in both health
and disease (see Fig. 2).
We have developed an ELISA that specifically detects the sFLT-1
Implications for clinical practice
e15a variant (Palmer et al., 2015). This involved developing a peptide The discovery of multiple sFLT-1 splice variants provides the potential
sequence from the unique tail of the variant protein (Fig. 2; boxed to better identify disorders of placentation through targeting strategies
sequence). This peptide was then coupled to a highly immunogenic towards the placental-specific variant, sFLT-1 e15a.
protein for animal immunization. Once an appropriate antibody The use of both anti-angiogenic and angiogenic factors as biomar-
response had occurred, polyclonal antibodies were isolated from the kers for the prediction or diagnosis of pre-eclampsia was identified as
serum and purified for testing. Importantly, both the antibody and the soon as their importance in the disease process was realized (Hertig
subsequent sFLT-1 e15a ELISA that was developed are specific to this et al., 2004). Subsequent work has shown that total sFLT-1, particu-
isoform. Importantly, an independent group has also successfully devel- larly as a ratio with PlGF, appears able to diagnose and even predict
Figure 2 Alignment of the amino acid sequence for human FLT-1, sFLT-1 i13 and sFLT-1 e15a. The amino acid sequence for all proteins is shown
from amino acid 650: sequence prior to amino acid 650 shares 100% homology between the three proteins. The unique C-terminal tails are highlighted
for sFLT-1 i13 and sFLT-1 e15a (grey). Antibodies that have been developed are indicated with the underlined sequences indicating those used by
Souders et al. (2015) and the boxed sequence by Palmer et al. (2015). The sFLT-1 e15a polyserine tail continues for a further 11 serine residues.
Placental-specific sFLT-1 75
premise is strongly supported by the fact that those developing term (#1050765 to S.T. #1062418 to T.J.K.). K.R.P. received salary support
pre-eclampsia have an average 3-fold increase in circulating sFLT-1 from the Department of Obstetrics and Gynaecology, Monash
levels compared to normal term controls, whereas a 43-fold increase University.
in serum sFLT-1 is seen in early-onset disease compared to preterm
controls (Wikstrom et al., 2007). However, this remains an area Conflict of interest
requiring further investigation as studies performed to date on late-
onset pre-eclampsia were limited by a small sample size. Furthermore, None declared.
research in this area is currently limited by the absence of any com-
mercially available assays. References
Ahmad S, Ahmed A. Elevated placental soluble vascular endothelial growth
Discussion factor receptor-1 inhibits angiogenesis in preeclampsia. Circ Res 2004;
95:884–891.
Pre-eclampsia remains a significant burden to women and their health- Andersen LB, Dechend R, Jorgensen JS, Luef BM, Nielsen J, Barington T,
inhibition and renal thrombotic microangiopathy. N Engl J Med 2008; factor (PGF) and ameliorates preeclampsia in a mouse model. Proc Natl
358:1129–1136. Acad Sci U S A 2011;108:1451–1455.
Eremina V SM, Haigh J, Nagy A, Lajoie G, Ferrara N, Gerber HP, Kikkawa Y, Leung DW, Cachianes G, Kuang WJ, Goeddel DV, Ferrara N. Vascular
Miner JH, Quaggin SE. Glomerular-specific alterations of VEGF-A expres- endothelial growth factor is a secreted angiogenic mitogen. Science
sion lead to distinct congenital and acquired renal diseases. J Clin Invest 1989;246:1306–1309.
2003;111:707–716. Levine RJ, Maynard SE, Qian C, Lim K-H, England LJ, Yu KF, Schisterman
Gerber HP, Condorelli F, Park J, Ferrara N. Differential transcriptional EF, Thadhani R, Sachs BP, Epstein FH et al. Circulating angiogenic factors
regulation of the two vascular endothelial growth factor receptor genes. and the risk of preeclampsia. N Engl J Med 2004;350:672–683.
Flt-1, but not Flk-1/KDR, is up-regulated by hypoxia. J Biol Chem 1997; Levine RJ, Qian C, Maynard SE, Yu KF, Epstein FH, Karumanchi SA. Serum
272:23659–23667. sFlt1 concentration during preeclampsia and mid trimester blood pressure
Gilbert JS, Babcock SA, Granger JP. Hypertension produced by reduced in healthy nulliparous women. Am J Obstet Gynecol 2006;194:1034–1041.
uterine perfusion in pregnant rats is associated with increased soluble Li C, Raikwar NS, Santillan MK, Santillan DA, Thomas CP. Aspirin inhibits
fms-like tyrosine kinase-1 expression. Hypertension 2007;50:1142–1147. expression of sFLT1 from human cytotrophoblasts induced by hypoxia,
Glusker P, Recht L, Lane B. Reversible posterior leukoencephalopathy syn- via cyclo-oxygenase 1. Placenta 2015;36:446–453.
Palmer KR, Kaitu’u-Lino TJ, Hastie R, Hannan NJ, Ye L, Binder N, Cannon P, tyrosine kinase gene (flt) closely related to the fms family. Oncogene 1990;
Tuohey L, Johns TG, Shub A et al. Placental-specific sFLT-1 e15a protein 5:519–524.
is increased in preeclampsia, antagonizes vascular endothelial growth fac- Souders CA, Maynard SE, Yan J, Wang Y, Boatright NK, Sedan J, Balyozian D,
tor signaling, and has antiangiogenic activity. Hypertension 2015;66: Cheslock PS, Molrine DC, Simas TA. Circulating levels of sFlt1 splice var-
1251–1259. iants as predictive markers for the development of preeclampsia. Int J Mol
Palmer KR, Tong S, Tuohey L, Cannon P, Ye L, Hannan NJ, Brownfoot FC, Sci 2015;16:12436–12453.
Illanes SE, Kaitu’u-Lino TJ. Jumonji domain containing protein 6 is Steegers EA, von Dadelszen P, Duvekot JJ, Pijnenborg R. Pre-eclampsia.
decreased in human preeclamptic placentas and regulates sFLT-1 splice Lancet 2010;376:631–644.
variant production. Biol Reprod 2016b;94:59. Sugimoto H, Hamano Y, Charytan D, Cosgrove D, Kieran M, Sudhakar A,
Park JE, Chen HH, Winer J, Houck KA, Ferrara N. Placenta growth factor. Kalluri R. Neutralization of circulating vascular endothelial growth factor
Potentiation of vascular endothelial growth factor bioactivity, in vitro (VEGF) by anti-VEGF antibodies and soluble VEGF receptor 1 (sFlt-1)
and in vivo, and high affinity binding to Flt-1 but not to Flk-1/KDR. J Biol induces proteinuria. J Biol Chem 2003;278:12605–12608.
Chem 1994;269:25646–25654. Suwaki N, Masuyama H, Nakatsukasa H, Masumoto A, Sumida Y,
Park M, Lee ST. The fourth immunoglobulin-like loop in the extracellular Takamoto N, Hiramatrsu Y. Hypoadiponectinemia and circulating angio-
Whitehead CL, Palmer KR, Nilsson U, Gao Y, Saglam B, Lappas M, kinase-1 in early-onset and late-onset preeclampsia. Obstet Gynecol
Tong S. Placental expression of a novel primate-specific splice variant of 2007;109:1368–1374. 1310.1097/1301.AOG.0000264552.000028543
sFlt-1 is upregulated in pregnancies complicated by severe early onset 6.a0000264551.
pre-eclampsia. BJOG 2011;118:1268–1271. Xiong Y, Liebermann DA, Tront JS, Holtzman EJ, Huang Y,
WHO World Health Report: Making Every Mother and Child Count. Geneva: Hoffman B, Geifman-Holtzman O. Gadd45a stress signaling regu-
World Health Organisation, 2005. lates sFlt-1 expression in preeclampsia. J Cell Physiol 2009;220:
Wiesmann C, Fuh G, Christinger HW, Eigenbrot C, Wells JA, de Vos AM. 632–639.
Crystal structure at 1.7 A resolution of VEGF in complex with domain 2 Zeisler H, Llurba E, Chantraine F, Vatish M, Staff AC, Sennstrom M,
of the Flt-1 receptor. Cell 1997;91:695–704. Olovsson M, Brennecke SP, Stepan H, Allegranza D et al. Predictive
Wikstrom A-K, Larsson A, Eriksson UJ, Nash P, Nordén-Lindeberg S, value of the sFlt-1:PlGF ratio in women with suspected preeclampsia.
Olovsson M. Placental growth factor and soluble FMS-like tyrosine N Engl J Med 2016;374:13–22.