Maternal Serum Activin A, Inhibin A and Follistatin-Related Proteins across Preeclampsia: Insights into Their Role in Pathogenesis and Prediction

Barrero, Jorge A.; Villamil-Camargo, Laura M.; Imaz, Jose N.; Arciniegas-Villa, Karen; Rubio-Romero, Jorge A.

doi:10.34763/jmotherandchild.20232701.d-23-00002

Figures & Tables

Activin A canonical signalling pathway in the maternal-foetal interface. Activin binding enables receptor activation by acquiring a tetrameric conformation through ALK4 (TβR-I) interaction with one of the following TβR-II: ACVR2A, ACVR2B or TGFβRII [55]. Activated TβR-II phosphorylates the GS domains of ALK4, which then gain high affinity for the receptor-regulated SMAD proteins (SMAD2 and SMAD3) [56]. Facilitated by the auxiliary adaptor protein SARA, SMAD2 and SMAD3 are recruited in proximity to the kinase domains of ALK4 and, upon binding, they are phosphorylated on the SSXSX motifs [87]. Ser/Thr phosphorylation of receptor-SMADs induces dimerisation and exposure of the nuclear translocation region on their MH2 domain and augments their affinity for a Co-SMAD (SMAD-4) [87]. SMAD2/3-SMAD2/3-SMAD4 heterotrimers are then translocated to the nucleus and bind to specific DNA regions (SBE sequence) and transcriptional coactivators or corepressors [88]. Inhibin A binds to TβR-II through its βA subunit but blocks the phosphorylation of ALK4 by its α subunit, which binds to extracellular matrix β-glycans and block receptor tetramerization [88]. Follistatin-related proteins block ligand-receptor complex formation by binding activins. Abbreviations: TβR-I, type I subunit receptor; TβR-II, type II subunit receptor; ALK4, activin-like kinase 4; ACVR2A/B, activin receptor type 2A/B; TGFβRII, transforming growth factor-β receptor type 2; MMP2, matrix metalloproteinase 2. Source: own elaboration.

Changes in maternal serum activin A, inhibin A, follistatin and FSTL3 in preeclamptic pregnancies_

Reference	Population	Protein	Weeks of gestation	PE subtype	Main outcomes
Activin A
Yu et al. (2012) [15]	Severe PE group: 44 women (30.8±5.1 years)Control group: 51 women (29.8±3.2 years)	Activin A	Five intervals:(1) 15⁺⁰–18⁺⁰(2) 19⁺⁰–24⁺⁰(3) 25⁺⁰–30⁺⁰(4) 31⁺⁰–35⁺⁰(5) 36⁺⁰–40⁺⁰	Severe PE	From gestational week 25 until delivery, activin A levels increased significantly in women with severe PE when compared to controls (p<0.05).The difference in maternal serum levels of activin A according to week intervals were:(3) 3.6-fold increase in severe PE (16.6 vs. 4.6 ng/mL),(4) 4.2-fold increase in severe PE (30.1 vs.7.2 ng/mL),(5) 3.5-fold increase in severe PE (34.3 vs. 9.7 ng/mL).
Baumann et al. (2013) [16]	PE group: 46 womenControls: 92 women	Activin A	First trimester	No distinction	Activin A increased significantly in pregnancies with PE when compared to controls (p<0.05).Activin A yielded an AUC of 0.670 in PE prediction.
Lai et al. (2013) [17]	PE group: 50 womenControls: 250 women	Activin A	Two intervals:(1) 11⁺⁰–13⁺⁰(2) 30⁺⁰–33⁺⁰	No distinction	Activin A increased significantly at 30⁺⁰–33⁺⁰ weeks of gestation, but not at 11⁺⁰–13⁺⁰ weeks, in women with PE (1.47 (1.14–2.38) MoM) when compared to controls (0.99 (0.72–1.42) MoM) (p<0.05).Activin A yielded an AUC of 0.722 in PE prediction at 30⁺⁰–33⁺⁰ weeks of gestation.
Tarca et al. (2019) [18]	EO-PE group: 33 women (22 (19.0–25.5) years)Controls: 90 women (24 (21.0–27.8) years)	Activin A	22⁺¹–28⁺⁰	EO-PE	Activin A yielded an AUC of 0.890 in EO-PE prediction.
Hao et al. (2020) [19]	PE group: 20 women (31.8±6.0 years)Controls: 20 women (31.9±4.8 years)	Activin A	Five intervals:(1) 5⁺⁰–9⁺⁰(2) 10⁺⁰–14⁺⁰(3) 15⁺⁰–25⁺⁰(4) 26⁺⁰–33⁺⁰(5) 27⁺⁰–38⁺⁰	No distinction	Activin A begins to rise around 20 weeks of gestation, yet no significant difference was evidenced between women with PE and normal pregnancies.Activin A yielded an AUC of 0.650 in PE prediction at 16⁺⁰–30⁺⁰ weeks of gestation.
Wong et al. (2022) [20]	PE group: 40 women (31.0 [29.3–34.0] years)Control group: 201 women (32.0 [30.0–35.0] years)	Activin A	Two intervals:(1) 27⁺⁰–29⁺⁰(2) 35⁺⁰–37⁺⁰	No distinction	Activin A increased significantly at 35⁺⁰–37⁺⁰ weeks of gestation, but not at 27⁺⁰–29⁺⁰ weeks, in women with PE (706 (2162–8362) pg/mL) when compared to controls (2050 (1018–3723) pg/mL) (p<0.001).Activin A yielded an AUC of 0.710 in PE prediction at 35⁺⁰–37⁺⁰ weeks of gestation.
Activin A and inhibin A
Li et al. (2016) [43]	PE group: 39 women (27.12±3.12 years)Controls: 100 women (28.67±3.51 years)	Inhibin AActivin A	11⁺⁰–13⁺⁰	No distinction	Inhibin A increased significantly in women with PE (1.72±0.02 pg/mL) when compared to controls (1.03±0.06 pg/mL) (p<0.001).Activin A increased significantly in women with PE (1.68±0.38 pg/mL) when compared to controls (1.06±0.42 pg/mL) (p<0.001).
Xu et al. (2017) [44]	Mild PE group: 14 women (30.8±1.0 years)Severe PE group: 17 women (31.5±1.1 years)Controls: 18 women (30.4±0.7 years)Among the 31 women with PE, 13 had EO-PE and 18 had LO-PE	Inhibin AActivin A	Before delivery.Weeks of gestation were not specified.	Mild PESevere PEEO-PELO-PE	Activin A (23.5±2.1 μg/L) increased significantly in women with severe PE when compared to controls and EO-PE (p<0.05). No difference was found with LO-PE.Inhibin A increased significantly in women with severe PE (1.7±0.2 μg/L) and EO-PE (1.9±0.2 μg/L) when compared to controls and mild PE (p<0.001). No difference was found with LO-PE.Both activin A (26.0±2.3 μg/L) and inhibin A (1.9±0.2 μg/L) were significantly higher in EO-PE when compared with women with LO-PE (activin A: 16.3±1.5 μg/L, inhibin A: 1.1±0.1 μg/L) (p<0.001).
Inhibin A
Olsen et al. (2012) [21]	PE group: 459 women, out of which 65 delivered before 34 weeks and 294 delivered after 34 weeks of gestation.Controls: 7,308 women	Inhibin A	Second trimester	EO-PELO-PE	Elevated inhibin A levels were associated with any form of PE (LO-PE: (OR: 3.08; CI 95%: 2.15–4.42) and EO-PE: (OR: 4.17; CI 95%: 2.11–8.27)).The OR increased for each every >2 MoM rise in inhibin A levels.
Dugoff et al. (2013) [22]	Severe PE group: 26 womenPE group: 56 womenControls: 946 women	Inhibin A	11⁺⁰–14⁺⁰	Severe PE	Inhibin A increased significantly in pregnancies with PE, but not severe PE, when compared to controls (p<0.05).
Boucoiran et al. (2013) [23]	PE group: 34 women (31.5 (26.0–35.0) years)Control group: 584 women (30.0 (27.0–34.0) years)	Inhibin A	Two intervals:(1) 12⁺⁰–18⁺⁰(2) 24⁺⁰–26⁺⁰	No distinction	At weeks 12⁺⁰–18⁺⁰, inhibin A increased significantly in women with PE (1.10 (0.77–1.52) MoM) when compared to controls (0.96 (0.71–1.25) MoM) (p<0.05). In these same weeks, inhibin A yielded an AUC of 0.590 in PE prediction.At weeks 24⁺⁰–26⁺⁰, inhibin A increased significantly in women with PE (1.10 (0.89–1.65) MoM) when compared to controls (0.91 (0.66–1.23) MoM) (p<0.05). In these same weeks, inhibin A yielded an AUC of 0.620 in PE prediction.
Suri et al. (2013) [24]	PE group: 14 womenControls: 42 women	Inhibin A	15⁺⁰–22⁺⁰	No distinction	Inhibin A increased significantly in women with PE (1.37 (0.95–1.62) MoM) when compared to controls (0.96 (0.79–1.33) MoM) (p<0.05).
Park et al. (2014) [25]	PE group: 8 women (33 (25–37) years)Controls: 254 women (33 (20–39) years)	Inhibin A	11⁺⁰–13⁺⁶	No distinction	Inhibin A increased significantly in pregnancies with PE (1.86 (0.82–2.96) MoM) when compared to controls (1.00 (0.39–3.676) MoM) (p<0.05).Inhibin A yielded an AUC of 0.780 in PE prediction.
Giguère et al. (2015) [26]	[1] Preterm PE:Cases: 47 women (29.9±5.5 years)Controls: 94 women (30.3±5.2 years)[2] EO-PE:Cases: 13 women (30.1±6.1 years)Controls: 26 women (30.3±5.4 years)[3] Severe PE:Cases: 68 women (29.7±5.0 years)Controls: 94 women (29.6±4.3 years)	Inhibin A	10⁺⁰–18⁺⁰	Preterm PEEO-PESevere PE	Inhibin A increased significantly in pregnancies with severe PE (267 (176–384) pg/mL) when compared to controls (211 (158–322) pg/mL) (p<0.05), and in preterm PE (1.36 (0.97–1.89) MoM) when compared to controls (1.08 (0.85–1.52) MoM).Inhibin A levels in preterm PE and EO-PE did not change significantly when compared to controls.
Kumer et al. (2016) [27]	PE group: 40 womenControls: 28 women	Inhibin A	25⁺⁰–41⁺⁰	No distinction	Inhibin A increased significantly in pregnancies with PE (1,267± 383 ng/L) when compared to controls (660 ± 395 ng/L) (p<0.001).
Chrelias et al. (2016) [28]	PE group: 98 women (31.1±4.9 years)Control group: 98 women (31.0±5.7 years)	Inhibin A	24 hours before delivery	No distinction	Inhibin A increased significantly in women with PE (1.07±0.83 ng/mL) when compared to controls (0.73±0.35 ng/mL) (p<0.05).PE was highly associated with changes in inhibin A levels (OR: 1.09). The likelihood of developing PE increased by 10% for every rise in inhibin A serum levels by 0.1 ng/mL.
Broumand et al. (2018) [29]	PE group: 14 womenControls: 286 women	Inhibin A	16⁺⁰–20⁺⁰	No distinction	The sensitivity of inhibin A levels ≥1.25 MoM in PE prediction was 83.83%, its specificity 65.30%, positive predictive value 4.67%, negative predictive value 99.48%, positive likelihood 2.41, and negative likelihood 0.24. Its diagnostic accuracy was 65.66%.Inhibin A levels were directly associated with the incidence of PE (p<0.001) and its severity (p<0.05).
Belovic et al. (2019) [30]	104 pregnant women (30.54±4.93 years), of which 6 developed PE.No available data of the control group.	Inhibin A	16⁺⁰–19⁺⁰	No distinction	No significant difference was evidenced in inhibin A levels among women with PE (1.23±0.11 MoM) and normal pregnancies (1.17±0.98 MoM).
Yue et al. (2020) [31]	PE group: 560 women (29.3±2.72 years)Controls: 8333 women (29.0±2.82 years)	Inhibin A	14⁺⁰–20⁺⁰	No distinction	Inhibin A increased significantly in pregnancies with PE (1.34±0.83 MoM) when compared to controls (1.07±0.47 MoM) (p<0.001).Inhibin A was found to be an independent risk factor for preeclampsia (OR: 7.45; CI 95%: 4.69–11.85).
Sharabi-Nov et al. (2020) [32]	PE group: 31 women (33.9 (32.3–35.6) years)Controls: 21 women (34.0 (32.0–35.9) years)	Inhibin A	>24⁺⁰and before delivery	No distinction	Inhibin A increased significantly in pregnancies with PE (2,097 (1,546–2,660) pg/mL) when compared to controls (724 (491–904) pg/mL) (p<0.05).
Kim et al. (2021) [33]	PE group: 13 women (33.50±6.80 years)Controls: 338 women (33.90±4.0 years)	Inhibin A	14⁺⁰–22⁺⁰	No distinction	No significant difference was evidenced in inhibin A levels among women with PE (1.00 (0.39–2.29) MoM) and normal pregnancies (0.93 (0.06–3.86) MoM).
Keikkala et al. (2021) [34]	EO-PE: 11 women (31.8±4.6 years)LO-PE: 34 women (31.2±5.4 years)Controls: 89 women (31.8±5.1 years)	Inhibin A	Three intervals:(1) 12⁺⁰–14⁺⁰(2) 18⁺⁰–20⁺⁰(3) 26⁺⁰–28⁺⁰	EO-PELO-PE	Inhibin A increased significantly in pregnancies with PE at weeks 18⁺⁰–20⁺⁰ (p<0.05) and 26⁺⁰–28⁺⁰ (p<0.001) when compared to controls.At weeks 26⁺⁰–28⁺⁰, inhibin A yielded an AUC of 0.717 in PE prediction, AUC of 0.889 in EO-PE prediction, and AUC of 0.674 in LO-PE prediction.
Follistatin and activin A
Charkiewicz et al. (2018) [36]	Mild PE group: 21 women (29±5.65 years)Controls: 27 women (28±5.49 years)	FollistatinActivin A	25⁺⁰–40⁺⁰	Mild PE	No significant difference was evidenced in activin A levels among women with PE (801.40±729.47 pg/mL) and normal pregnancies (260.28±212.90 pg/mL).Follistatin decreased significantly in women with PE (20,753.30 ± 3,196.87 pg/mL) when compared to controls (33,290.16±3,217.63 pg/mL) (p<0.05).
Follistatin-related proteins
Han et al. (2014) [40]	PE group: 39 women (30.54±4.06 years), of which 10 had mild PE and 29 had severe PE.Controls: 73 women (29.10±3.40 years).	FSTL3	24⁺⁰–28⁺⁰	Mild PESevere PE	FSTL3 increased significantly in women with PE (26.68±8.20 ng/ml) when compared to controls (21.93±9.02 ng/ml) (p<0.001).No significant difference was evidenced in FSTL3 levels among women with mild PE (29.10±10.29 ng/mL) and severe PE (25.84±7.54 ng/mL).
Garcés et al. (2015) [35]	PE group: 20 women (19.5 (18–26) years)Controls: 28 women (23.5 (19–30) years)	Follistatin	Three periods of gestation:(1) 11⁺⁵–12⁺⁵(2) 24⁺²–24⁺⁶(3) 34⁺²–35⁺⁴	No distinction	Follistatin decreased significantly during late pregnancy (34⁺²–35⁺⁴ weeks of gestation) in women with PE (116.2±50.1 ng/mL) when compared to controls (151.2±36.2) (p<0.05).
Horvath et al. (2016) [42]	PE group: 32 womenControls: 44 women	FSTL3	Third trimester Other gestational periods were not specified.	No distinction	FSTL3 increased significantly during the third trimester in PE pregnancies (25.30 ng/mL) when compared to controls (15.64 ng/mL) (p<0.001) but did not in earlier gestational time periods.
Charkiewicz (2016) [37]	Mild PE group: 12 womenControls: 12 women	Follistatin	25⁺⁰–40⁺⁰	Mild PE	Follistatin decreased significantly in women with mild PE when compared to controls (p<0.05).Follistatin yielded an AUC of 0.771 in mild PE prediction.
Luo and Han (2017) [39]	PE group: 33 women (30.50±4.12 years)Controls: 71 women (29.30±3.35 years)	FSTL3	24⁺⁰–28⁺⁰	No distinction	FSTL3 increased significantly in women with PE (0.55 (0.00–1.34) MoM) when compared to controls (1.21 (0.58–1.97) MoM) (p<0.05).FSTL3 yielded an AUC of 0.706 in PE prediction.
Nevalainen et al. (2017) [38]	EO-PE group: 29 women in the training set (28 (26–33) years) and 42 women in the test set (29 (25–33) years)Controls: 652 women in the training set (30 (26–34) years) and 141 women in the test set (29 (25–33) years)	FSTL3	9⁺⁰–13⁺⁶	EO-PE	No significant difference was evidenced in FSTL3 levels among women with PE and normal pregnancies in both sets.FSTL3 yielded an AUC of 0.750 in the training set, and an AUC of 0.650 in EO-PE prediction.
Purut et al. (2019) [41]	PE group: 13 womenControls: 131 women	FSTL3	11⁺⁰–14⁺⁰	No distinction	No significant difference was evidenced in FSTL3 levels among women with PE (10.85±5.55 pg/mL) and normal pregnancies (11.25±4.86 pg/mL).FSTL3 yielded an AUC of 0.522 in PE prediction.

Maternal Serum Activin A, Inhibin A and Follistatin-Related Proteins across Preeclampsia: Insights into Their Role in Pathogenesis and Prediction

Figures & Tables

Figure 1.

Figure 2.

Changes in maternal serum activin A, inhibin A, follistatin and FSTL3 in preeclamptic pregnancies_

Paradigm

My account