Placental TRPV2 supports trophoblast function and normal fetal development

All human and animal studies were conducted in accordance with relevant ethical regulations. Human placental tissue collection was approved by the institutional ethical and review board of the University Hospital Gasthuisberg Leuven (S60993) and by the relevant U.K. National Health Service Research Ethics Committees for the EVERREST study (15/LO/1488 and 13/LO/1254). Study participants did not receive financial compensation for donation of tissue samples. Animal studies were approved by the Ethical Review Committee for Animal Experiments of KU Leuven, Belgium. Placental samples from caesarean sections (term) or vaginal delivery (elective abortion) were obtained at the ward of the UZ Leuven hospital and were formalin fixed (4% paraformaldehyde (PFA)) and embedded in paraffin. The use of placental tissue was approved by the institutional ethical and review board of the University Hospital Gasthuisberg Leuven, Belgium for the protection of human subjects (S60993). Written informed consent was obtained from all participating subjects. Exclusion criteria included medication use, multiple gestations, severe obesity, substance abuse and pregnancy complications. The tissue samples used for histological and molecular analyses were preserved and are stored at the respective institutions under approved ethical protocols. Availability of remaining material is limited by sample quantity and ethical restrictions associated with human clinical specimens. Requests for access to remaining material should be directed to the corresponding author. The tissue samples used for RNA analysis were fully consumed during the analyses described in this study and are therefore no longer available for distribution. Placentas were collected from pregnant women who were prospectively recruited at University College London Hospital, London, U.K., through their involvement in the EVERREST trial at the University College London Hospital 39 . Exclusion criteria for all groups included multiple pregnancy, maternal age <18 years, fetal structural or karyotypic abnormalities or maternal virus infection. Placental villous tissue was systematically sampled from the placental parenchyma, midway between the umbilical cord insertion and margin within 30 min of birth. Maternal age, body mass index, fetal sex and clinical characteristics are summarized in Table  2 . Women with pregnancies affected by early-onset FGR were defined by ultrasound-assessed estimated fetal weight (EFW) < 600 g and <3rd centile, between 20 and 26 + 6 weeks of gestation. Women with early-onset FGR delivered either preterm (< 37 weeks of gestation) or at term (≥ 37 weeks of gestation). Women with AGA fetuses (10 th –95 th centile) delivered at term. Exclusion criteria included multiple pregnancies, maternal age <18 years, fetal structural or chromosomal abnormalities, preterm prelabour rupture of membranes or maternal bacterial or viral infection. At delivery of liveborn infants, placental villous tissue was sampled from two points midway between the umbilical cord insertion and margin, dissected free of decidua and chorionic plate and stored in RNAlater (Thermofisher). Human subjects were recruited with written informed consent and ethical approval from U.K. National Health Service Research Ethics Committees (London: Hampstead Research Ethics Committee, REC reference 15/LO/1488 and Stanmore Research Ethics Committee, REC reference 13/LO/1254). All mice were kept in a controlled environment (23 °C ± 1.5 °C, 40–60% relative humidity, 14:10 h light-dark cycle). Animals were housed in filter-top cages under conventional conditions, maximum five animals per cage with food and water ad libitum . Mice were bred between 8 and 12 weeks of age and checked daily for the presence of a copulation plug within the vaginal canal. The day of plug detection was dated as embryonic day 0.5 (E0.5). Adult animals were sacrificed by cervical dislocation, whereas fetuses were culled by decapitation. After sacrificing the mother, the uterus was removed and dissected. The fetuses and their placentas were separated using fine forceps and weighed, taking care that no amniotic sac or umbilical cord was present. Tail pieces of each embryo were used for genotyping. Placentas were stored in 4% PFA for histological means, or in RNAlater (Thermofisher) for RNA analysis. The ethical review committee for animal experiments at the KU Leuven (Belgium) approved the use of mice for this study. Trpv2 F /F mice were a kind gift from Katanosaka Y. et al. 13 Exon 4 of Trpv2 is flanked by two loxP sequences (Supplementary Fig.  1a ), the excision of which results in a frameshift, thereby introducing an alternative stop codon and a short protein product without channel structure. Full Trpv2 knockouts ( Trpv2 -/- ) were created by using maternally derived Sox2 driven Cre expression. Maternally inherited Cre recombinase activity will occur regardless of inheriting the Cre transgene 40 , resulting in whole body heterozygote knockout offspring (Supplementary Fig.  1b , Supplementary Fig.  3 ). This approach resulted in Trpv2 heterozygous animals ( Sox2-Cre wt/wt ; Trpv2 +/- ), which were used to generate global Trpv2 -/- and Trpv2 +/+ littermates. In the large breeding colony, only 12 (1%) Trpv2 -/- animals were identified among the 777 born progeny, whereas 436 (55%) Trpv2 +/- , and 219 (28%) Trpv2 +/+ were recovered, and 69 (15%) animals suffered from perinatal lethality and were not available at the time of genotyping. Sox2 -Cre (B6N.Cg-Edil3Tg( Sox2 -Cre)1Amc/J, Strain #014094) mice were purchased from Jackson Laboratory. Epiblast specific Trpv2 knockouts ( Trpv2 EPI -/- ) were generated using paternally derived Sox2 -Cre resulting in cre expression in the epiblast, but not in the extraembryonic tissues. To create Epiblast specific Trpv2 knockout mice, Male Sox2-Cre wt/tg were crossed with female Trpv2 F/F mice. From this cross, male Sox2-Cre wt/tg ;Trpv2 +/- were crossed with female Trpv2 F/F mice to generate Sox2-cre wt/tg ;Trpv2 -/- ( Trpv2 EPI -/- ) and Sox2-Cre wt/wt ;Trpv2 F/+ ( Trpv2 EPI +/+ ) Junctional zone specific Trpv2 knockouts ( Trpv2 JZ -/- ) were created using Tpbpa -Cre males ensuring cre recombinase activity in Tpbpa -positive cells exclusively located in the junctional zone of the placenta and had been on a C57BL/6 J background for >10 generations 18 , 41 . To create Junctional zone specific Trpv2 knockout mice, Male Tpbpa-Cre wt/tg were crossed with female Trpv2 F/F mice. From this cross, male Tpbpa-cre wt/tg ;Trpv2 F/+ were crossed with female Trpv2 F/F mice to generate Tpbpa-cre wt/tg ;Trpv2 F/F ( Trpv2 JZ -/- ) and Tpbpa-cre wt/wt ;Trpv2 F/F or Tpbpa-cre wt/wt ;Trpv2 F/+ ( Trpv2 JZ +/+ ). Syncytiotrophoblast layer II specific Trpv2 knockouts ( Trpv2 SynTII -/- ) were created using Gcm1 -Cre ensuring recombinase activity in Gcm1 -positive cells exclusively located in the SynTII cells of the placenta 23 . Of note, only the paternal Gcm1 -Cre allele was expressed and Gcm1 is mosaically expressed in other organs of the embryo, including the tail that is routinely used for genotyping (source: emouseatlas 42 ) resulting in deletion in the embryo as well, albeit in a nonuniform manner. Gcm1 -Cre (Tg( Gcm1 -Cre)1Chrn/J, Strain #:026625) mice were purchased from The Jackson Laboratory. To create SynTII knockouts, male Gcm1 -Cre wt/tg ; Trpv2 +/- were crossed with female Trpv2 F /F ; whereas the reverse controls cross was made using male Trpv2 F /F and female Gcm1 -Cre wt/tg ; Trpv2 +/- . Rosa mT/mG mice were used to assess transgene activity by mating with Cre mice according to the previously described breeding strategies. This line expresses a cell membrane localized tdTomato (mT) fluorophore in all cell types, whereas cell membrane-localized EGFP fluorophore (mG) is expressed in Cre expressing cells 43 . Placentas of reporter breedings were collected at E14.5 and immediately fixed in PFA for 24 hours. Collected placentas were then rinsed in 1X phosphate-buffered saline (PBS), transferred to 15% sucrose in 1X PBS for another 24 h, where after they were transferred to 30% sucrose in 1X PBS. Placentas were frozen in Tissue-Tek O.C.T (Sakura, Europe) and stored at −80 °C. Sections of 5 µm were cut and mounted with Vectashield (Vetorlabs) containing DAPI. Tail samples were lysed overnight at 55 °C in lysing buffer containing 10 mM Tris-HCl, 1.5 mM MgCl 2 , 50 mM KCl and 120 U/mL proteinase K (Biolabs, Belgium), after which they underwent denaturation for 30 min at 95 °C. Genotyping was done via PCR, using a master mix containing specific primers (Supplementary Table  1 ) and KAPA Mastermix (Biosystems, Belgium). The resulting amplified DNA was loaded on a 2.5% agarose gel containing 5 μg/mL ethidium bromide upon which gel electrophoresis was performed for 35 min at 110 mV. H&E staining was performed as described previously 44 . Briefly, nuclei were stained with Gill’s haematoxylin (Prosan) during 4 min, followed by a few dips in acid alcohol (1% HCl in ethanol), and subsequently saturated lithium carbonate. Cytoplasmic staining was achieved with eosin (Prosan) during 3 min. Masson Trichrome was performed according to manufacturer’s instructions (Sigma-Aldrich, HT15). For alkaline phosphatase stain, tissues were washed in alkaline phosphatase (AP) buffer (0.1 M Tris-HCl pH9.5, 0.1 M NaCl, 0.01 M MgCl2), and stained with NBT/BCIP solution (0.1 M Tris-HCl pH9.5, 0.1 M NaCl, 0.05 M MgCl2, 0.01% sodium deoxycholate, 0.02% NP-40, 337 μg/ml NBT, 175 μg/ml BCIP). Following, the tissues were counterstained with Nuclear Fast Red (Sigma-Aldrich) solution for 8 min. Tissue was then dehydrated and coverslip-mounted using Depex. Double-label Lectin and Cytokeratin staining was performed as adapted from De Clercq et al 22 . Briefly, endogenous peroxidase activity was blocked with hydrogen peroxide (3% in methanol, 10 min, RT) and antigen retrieval was performed using Pepsin (Sigma, 0.04% in preheated 0.01 M HCl, 10 min, 37 °C). Non-specific binding was blocked using blocking buffer (dH2O containing 2% bovine serum albumin (Sigma), 1% skimmed dry milk (Nestlé), 0.1% Tween80 (Merck) and normal goat serum (1/30), 15 min, RT). Sections were incubated with biotinylated lectin (isolectin B4, B-1205, Vector Laboratories; 1:250, 90 min, 37 °C), which was detected using horseradish peroxide (HRP)-conjugated streptavidin (Agilent P0397; 1:800, 30 min, RT) followed by 3’-Diaminobenzidine (Sigma, 10 min, RT). Next, endogenous phosphatases were blocked with 0.2 M HCl (10 min, RT) and sections were incubated with blocking buffer containing normal goat serum ((1/30), 15 min, RT). Sections were then incubated with rabbit anti-pan cytokeratin (1:500, Dako Z0622, 2 h, 37 °C), blocked again for 15 min, and bound antibody was detected with alkaline phosphatase-conjugated goat anti-rabbit (Jackson Immunoresearch 111-055-144; 1:100, 30 min, RT) followed by incubation with Fast Blue (Sigma F3378, 15 min, RT). Slides were washed in water and counterstained with Nuclear Fast Red (Vector laboratories, 5 min) before dehydrating and mounting with glycerine jelly. Omission of primary antibodies served as negative control Placental halves, fixed in 4% PFA and embedded in paraffin wax, were exhaustively sectioned, and systematic random samples were stained using H&E for further stereological analysis by a point grid. Placental volumes of the different layers were calculated by multiplying the point density for each layer with placental weight. Detailed stereological analysis of labyrinth structure was performed on a midline-section double-labelled with lectin and cytokeratin 45 – 47 . Point counting was used to determine the proportion of maternal blood spaces, fetal capillaries, and trophoblast in twenty random fields, and volumes were determined by multiplying the point density with labyrinth volume. Staining of SynTI and SynTII cells was performed by antibodies against MCT1 and MCT4 respectively, as reported before 48 . Briefly, antigen retrieval was performed using Citrate (pH = 6, 1 hr, 95 °C). Thereafter, slides were rinsed in 1% Bovine serum albumin (BSA) in Tris (0.01 M), followed by blocking and permeabilization using 5% Normal Donkey Serum (Merck) and 0.3% Triton X-100 in TBS (0.01 M + 1% BSA). Sections were then incubated overnight at 4 °C with chicken anti-MCT1 (1:500, Milipore AB1286I) and rabbit anti-MCT4 (1:100, Milipore AB3314P). Omission of primary antibodies served as negative control. After washing twice with TBS + 0.1%BSA, MCT1 and MCT4 were detected with respectively Donkey anti-chicken conjugated with CY5 (1:150, Jackson immunoresearch AB_2340607) and Donkey anti-rabbit conjugated with CY3 (1:150, Jackson immunoresearch AB_2340607) dissolved in TBS containing 0.1% BSA, 10% Normal Donkey Serum and 0.3% Triton X-100. Nuclei were counter stained with Hoechst (1:2000, Thermo Fisher Scientific 33342). Autofluorescence was quenched by incubation with True Black (1:20 in ethanol 70%, Biotium #23007, 30 sec), where after they were mounted by Fluorescent mounting medium (Agilent, S302380). Fluorescence in situ hybridisation (FISH) was performed following the manufacturer’s instructions of the RNAscope Multiplex Fluorescent v2 kit (ACDbio, BionTech), allowing for detection of single mRNA molecules 49 . Briefly, sections (4 μm) were deparaffinized, treated with hydrogen peroxide, target retrieval agent and protease plus. Sections were incubated for 2 h at 40 °C with specific RNAscope probes. Signal was amplified and detected by sequential treatment of HRP, fluorophore and HRP blocker. Ultimately, DAPI was used to stain the nuclei and slides were mounted with ProLong Gold Antifade Mountant (Thermo Fisher Scientific), Images were taken using the Zeiss Axioscan 7 taking care to use the same exposure and gain setting for each type of staining. Placentas from E16.5 (WT female, KO female, KO male) and whole concepti from E9.5 (WT) were extracted and were then placed into a mold with OCT compound (Vwr chemicals, ref 00411243) on ice. The sample was then frozen in cold isopentane bath placed in dry ice and then further stored at −80 °C. 14 um thick sections were collected with a cryostat (NX70, Thermo Fisher Scientific) and placed on the MERSCOPE standard slide v2.0 (ref 20400117). To enhance the attachment of the tissue to the slide, the sample was equilibrated for 30 min in the cryochamber. Then brought back to RT for 5 min and fixed with 5 ml PFA 4% (w/v) prepared in PBS (Thermo, PN: 043368.9 M) for 15 min at RT. The fixed sample was washed 3x times and incubated for 5 min with 1X PBS. The sample was then stored in 70% ethanol at 4 °C (up to a week). The MERFISH 2.0 sample preparation workflow for non-resistant fresh frozen samples was followed using manufacturer’s recommendation (MERFISH 2.0 Sample Preparation User Guide for Sectioned Tissue Samples; 91600132, rev A), using the MERSCOPE tissue sample preparation kit (Vizgen, PN: 10400194). Briefly, on the first day of the protocol, the sample was first incubated with a MERSCOPE pre-anchoring activator (Vizgen, PN 20300113) overnight at RT. The second day, we proceeded with an RNA anchoring step for 2 h at 37 °C with MERSCOPE RNA anchoring buffer (Vizgen, PN 20300117), then gel embedding of the tissue with the MERSCOPE gel embedding premix (Vizgen, PN 20300004) along with 25 μl of 10% Ammonium persulfate (Sigma, PN: A3678) and 2.5 μl of N,N,N′,N′-Tetramethyl ethylenediamine (Sigma, PN: 1107320100). The tissue was then subjected to an accelerated clearing for non-resistant tissues by incubating the tissue for 2 h at 47 °C with MERSCOPE clearing premix (Vizgen; PN 20300114). Finally, the probe hybridization was performed by incubating the tissue with 100ul of a MERSCOPE custom 300 gene panel V2.0 overnight at 47 °C. On the third day, the sample was washed twice with MERSCOPE formamide wash buffer (Vizgen; PN 20300002) for 30 min at 47 °C before subjecting for an onvernight hybridization with MERSCOPE enhancer probe mix (Vizgen, PN 20300194) at 37 °C. On the last day of the experiment, the sample was prepared for imaging by following the MERSCOPE instrument User Guide (Vizgen protocol, 91600001, rev K). The sample was then washed twice for 20 min with MERSCOPE formamide wash buffer (Vizgen; PN 20300002) at 47 °C, then stained with DAPI and polyT MERSCOPE staining reagent (Vizgen, PN 10400168) before proceeding with the imaging on the instrument using MERSCOPE Standard 300 Gene imaging kit V2.0 (ref 10400168). The rest of the imaging and downstream image processing workflow was done as per manufacturer recommendation in MERSCOPE instrument User Guide (Vizgen protocol, 91600001, rev K). For the tissue quality evaluation, the following quality criteria were evaluated: 1. Tissue integrity for spatial profiling and, 2. Quality of total RNA extracted from the tissue. For determining the tissue integrity post cryopreservation, a standard H&E staining was performed, and the tissue was imaged on a bright field microscope. For estimating the RNA quality, 5x OCT scrolls of 50 um sections were collected in an RNAse/DNAse-free tube (Eppendorf) and either processed immediately for total RNA extraction or stored at −80*C for subsequent extraction. Prior to RNA extraction, the excess OCT from the tissue scrolls were washed away with ice-cold PBS and RNA extractions were performed using standard column extraction method based on the manufacturer recommendations (innuPREP RNA Mini Kit 2.0; 250 rxn; iST Innuscreen GmbH; PN: B45-KS-2040250). Post RNA extraction, the quantity of RNA is estimated by NanoDrop (Thermo Fisher) and the quality of RNA (RIN) was estimated using bioanalyzer (Agilent; Agilent RNA 6000 Pico Kit; PN: 5067-1513). For the all the spatial analysis performed in this work, the RIN value if the tissue was >8. To align MERscope results with histological findings, consecutive fresh frozen slides were immunostained. Slides were fixed with PFA 4% (VWR chemicals) for 10 min immediately after sectioning and stored at −80 °C. Before use, the slides were dried (2 min at 37 °C) and blocked with 5% normal donkey serum and 0.3% Triton X-100 in TBS for 1 h. Lectin (Isolectin GS-IB4 From Griffonia simplicifolia , Alexa Fluor™ 488 Conjugate, Invitrogen) staining was performed at a 1:200 dilution and incubated for 1 hour at 37 °C. For MCT1/MCT4 staining, primary antibodies were incubated overnight at 4 °C: MCT1 (No. AB1286I, anti-monocarboxylate transporter 1 Polyclonal antibody, Millipore) at a 1:500 dilution (1 µg/ml) and MCT4 (No. AB3314P, formerly AB3546P, rabbit anti-monocarboxylate transporter 4 Polyclonal antibody, Millipore) at a 1:100 dilution (10 µg/ml) in TBS. The secondary antibodies, CyTM5 AffiniPure™ Donkey Anti-Rabbit IgG Fluor 650 (711-175-152, Jackson ImmunoResearch) at a 1:150 dilution and Cy™3 AffiniPure® Donkey Anti-Chicken IgY Fluor 570 (703-166-155, Jackson ImmunoResearch) at a 1:150 dilution, were incubated for 1 h at room temperature in TBS with 10% Normal Donkey Serum. Finally, the slides were mounted using DAPI (4′,6-diamidino-2-fenylindool) Fluorescent mounting medium (Abcam) and were stored at 4 °C. MERscope output was analyzed in R using the DESeq2 package. Detected transcripts from regions of interest (ROI) in E16.5 placenta’s were extracted based on coordinates, followed by analysis using DESeq2. Each ROI was treated as a distinct sample and 6 ROIs were extracted per condition. Three conditions were analyzed: wild-type labyrinth (WT), Synb -negative Trpv2 KO labyrinth, assigned Abnormal KO cells (AbC) and Synb -positive Trpv2 KO labyrinth, assiged Normal KO cells (NC). Heatmaps were generated using the pheatmap package. Mouse trophoblast stem cells (TS-Rs26 50 , kind gift of the Rossant lab, Toronto, Canada), were cultured in 30% RPMI 1640 with L-Glutamine (Thermo Fisher Scientific) supplemented with 20% FBS (Pan-Biotech), 1 mM sodium pyruvate (Thermo Fisher Scientific), 1X anti-mycotic Antibiotic (Thermo Fisher Scientific), 50 μM 2-Mercaptoethanol (Gibco), 25 ng/ml basal Fibroblast growth factor (bFGF, Sigma) and 1 μg/ml Heparin (Sigma); and 70% of conditioned medium (CM) from Mouse Embryonic Fibroblasts (MEF). Differentiation was induced by omitting bFGF, Heparin and CM from the media. CHIR99021 (3 μM, Sigma) was used to induce SynTII differentiation 6 . mTSC were cultured in 21% O 2 and 5% CO 2 incubator under humidified conditions at 37 °C, For generation of CRISPR/Cas9-mediated knockout TSCs, guide RNAs (gRNAs) that result in frameshift mutations (Supplementary Fig.  13a ) were designed using a CRISPR design tool ( http://crispor.tefor.net/ ) and checked for high specificity by nucleotide blast searches (Supplementary Table  3 ). Different combinations of gRNA were used to create different Trpv2 knockout clones, following a CRISPR/Cas9 strategy previously described 51 . gRNA sequences were cloned into the Cas9.2A-eGFP plasmid (Plasmid 48138, Addgene) and verified by sequencing. For generation of control clones, non-targeting scrambled gRNAs (Supplementary Table  3 ) were cloned into the same Cas9.2A-eGFP vector. Transfection was carried out with Lipofectamine 2000 (Thermo Fisher Scientific) according to the manufacturer’s protocol. Following transfection, single-cell-derived control and knockout clones were isolated and expanded. Knockout clones were confirmed by genotyping using primers spanning the deleted exon (Supplementary Fig.  13b ), and by reverse transcription followed by qRT-PCR with primers within, and downstream of the deleted exon (Supplementary Fig.  13c ). Additionally, absence of functional TRPV2 was confirmed by calcium imaging experiments (Supplementary Fig.  13d ). hTSC were obtained as previously described 26 , 52 with modifications. Naïve hPSCs (WiCell#WA09) were seeded in order to reach 90% confluency by day 2 on Mitomycin treated (MMC)-MEFs after dissociating in single cell using TryplE (Thermo Fisher Scientific, 12605-101). The next day, the naïve medium was switched to ACE medium consisting of 1:1 DMEM/F12 (Gibco) and Neurobasal medium (Gibco), 0.5% N2-supplement (Gibco), 1% B27-Supplement (Gibco), 0.5% L-Glutamine (Gibco), 0.1 mM β-mercaptoethanol (Sigma), 1x NEAA (Gibco), 1% BSA (Sigma), 1x penicillin-streptomycin (Gibco), supplemented with 1 μM A83-01 (Peprotech, 9094360), 50 ng/ml hEGF (Miltenyi Biotec, 130-097-750), 2 μM CHIR99021 (Axon Medchem, 1386), 10 μM Y-27632 (Tocris, 1254). The medium was changed every two days and supplemented with 10 μM Y-27632 and passaged every 4 days at 1:6 splitting ratio. hTSC were cultured in 5% O 2 and 5% CO 2 incubator under humidified conditions at 37 °C, on laminin-coated (iMatrix 511, Iwai-chem) dishes. For the conversion towards ST, hTSC were seeded on laminin (iMatrix)-coated coverslips, and incubated under normoxic conditions (21% O2 and 5% CO2) in DMEM/F12 containing 0.1 mM β-mercaptoethanol, 0,3% BSA, 1% insulin transferrin selenium-X (Gibco), 4% knockout serum replacement (KSR, Thermo Fisher Scientific), 2 μM Forskolin (Sigma) and 2.5 μM Y-27632 (Tocris) 53 . The medium was replaced at day 3 and cultured for 3 additional days without splitting. Placentas were first homogenized by use of a power homogenizer (Polytron) and the total RNA was extracted using TriPure Isolation Reagent (Roche), followed by incubation at room temperature (RT) for 5 min. Chloroform was added after which samples were incubated at RT for 10 min. Samples were then centrifuged at 12,000 x g at 4 °C for 5 min. Of the resulting layers, the upper layer, containing the RNA, was removed for further processing. After addition of isopropanol and an incubation period of 10 min at RT, RNA was centrifuged at 12,000 x g at 4 °C for 10 min. The pellet was washed in 75% ethanol, centrifuged at 7500 x g at 4 °C for 5 min and, after removal of ethanol, resuspended in diethylpyrocarbonate (DEPC)-treated RNase-free water (Qiagen). Next, an incubation of 15 min at 60 °C followed and RNA solutions were stored at −80 °C until further use. A nanodrop spectrophotometer (Thermo Fisher Scientific) was used to measure the RNA concentration while the quality of the RNA was measured using Experion RNA StdSens Analysing kit (Bio-Rad). RNA was considered of good quality when having an RIN >7. RNA extraction from cells was performed with RNeasy mini kit according to manufacturers’ instructions (Qiagen). Triplicate cDNA samples (2.5× diluted) from each independent preparation were used in the StepOne PCR system (Applied Biosystems, Life Technology) using specific TaqMan gene expression assays (Supplementary Table  4 ). Housekeeping genes were: Sdha for mTSC, Gapdh and Actb for mouse placentas, GAPDH and TPB for hTSC, and YWHAZ and B2M for human placentas. RNA was isolated from the labyrinth (E14.5 and E16.5) of Trpv2 +/+ and Trpv2 /- fetuses as described before ( n  = 6 placentas/genotype/day from 3-6 litters). RNA was isolated from mTSC as described before ( n  = 3 clone/genotype/differentiation day or method). RNA concentration and purity were determined spectrophotometrically using the Nanodrop ND-8000 (Nanodrop Technologies) and RNA integrity was assessed using a Bioanalyzer 2100 (Agilent). Per sample, an amount of 1 µg of total RNA was used as input fo samples with RIN > 7. Using the Illumina TruSeq® Stranded (Tissue) or Illumina Stranded (cells) mRNA Sample Prep Kit, poly-A containing mRNA molecules were purified from the total RNA input using poly-T oligo-attached magnetic beads. In a reverse transcription reaction using random primers, RNA was converted into first strand cDNA and subsequently converted into double-stranded cDNA in a second strand cDNA synthesis reaction using DNA PolymeraseI and RNAse H. The cDNA fragments were extended with a single ‘A’ base to the 3’ ends of the blunt-ended cDNA fragments after which multiple indexing adapters were ligated introducing different barcodes for each sample. Finally, enrichment PCR was carried out to enrich those DNA fragments that have adapter molecules on both ends and to amplify the amount of DNA in the library. Sequence-libraries of each sample were equimolarly pooled and sequenced on Illumina NovaSeq 6000 (tissues) or Element AVITI (cells) (Single Read 100 bp) at the VIB Nucleomics Core ( https://nucleomicscore.sites.vib.be) . Reads were aligned against GENCODE mouse genome M27 (tissues) or M28 (cells), quantified using STAR (v2.7.10b) 54 and gene and isoform expression levels were estimated using the RSEM 55 software packages (v1.3.1) (tissues) or Salmon 56 software package (v1.10.3) (cells). Differently expressed genes were analysed using DEseq2 package (v1.44) 57 and GO enrichment was performed by the ClusterProfiler package (v4.12.6). Significant overlap was tested by nemates.org. Trophoblast marker genes for gene-set based analysis were obtained from 2 published datasets (see Source Data) 20 , 21 . Heatmaps were generated using the pheatmap package (1.0.13). Downstream analyses and visualisation were performed in R (v4.4.1) using Seurat (v4.4.0 or v5.4.0) and Nebulosa (v1.14.0). The measurement of intracellular Ca 2+ was performed as previously described 44 . Briefly, cells were incubated with 2 µM Fura-2 acetoxymethyl (AM) ester (Bio-Connect) for 30-45 minutes at 37 °C. Fluorescent signals were evoked during alternating illumination at 340 and 380 nm on a Nikon Eclipse Ti fluorescence microscope, and the ratio of the fluorescence signals (F340/F380) was calculated. The standard solution contained (in mM) 150 NaCl, 2 CaCl 2 , 1 MgCl 2 , 10 D-glucose and 10 HEPES (pH 7.4 with NaOH). Functional TRPV2 activity was assessed by the application of 50 µM Δ9-tetrahydrocannabinol (THC, kindly provided by G. Appendino and F. Pollastro). Ionomycin (2 µM, Sigma) was applied at the end of every experiment as a positive control to evaluate cell viability. Calcium concentrations were calculated as described previously 58 . Measurements were performed with NIS software (v5.42.07) and analysed using Igor Prov v6 (WaveMetrics). Cells were considered responders if the calcium influx during agonist application exceeded 100 nM and when the highest value of the derivative of the calcium trace during the application of an activator exceeded at least 3 times the standard deviation of the derivative during basal conditions. Cells were designated as having high basal calcium, when they had a mean basal F340/F380 ratio > mean + 1.5xSD F340/F380 values of hTSC, with mean basal F340/F380 ratios = average over 5 minutes. To evaluate possible differences in basal calcium levels between cells with low versus high TRPV2 mRNA levels, cells were fixed (4% PFA)) immediately after Fura-2-based calcium imaging, followed by RNAscope analysis using a TRPV2 -specific probe. Regions of interest used to extract calcium imaging data were overlayed on the RNAscope images based on the nuclear (DAPI) staining. Cells were considered TRPV2- positive when the mean TRPV2 signal at the cell nucleus exceeded the mean value of non-stained cells by >3 SD. Whole-cell membrane currents were measured with an EPC-10 and the PatchMasterPro Software (HEKA Elektronik, Lambrecht, Germany). The sampling rate was 20 kHz and currents were digitally filtered at 2.9 kHz. For whole-cell recordings the standard extracellular solution contained (in mM): 150 NaCl, 1 MgCl2, 10 HEPES (pH 7.4 with NaOH) and the standard internal solution contained (in mM): 100 CsAsp, 20 CsCl, 10 1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA), 10 HEPES, 1 MgCl2, 4 Na2ATP and 2.93 CaCl2 (pH 7.2 with CsOH). The standard patch pipette resistance was between 2.5 and 4 MΩ when filled with pipette solution. Between 50 to 70% of the series resistance was compensated using the auto-compensation function of PatchMaster, reducing series resistance errors to below 1–5 mV. No compensations for liquid junction potentials were performed. Recordings with estimated voltage clamp errors >10 mV were excluded from analysis. Trpv2 was added to the list of genes expressed in the mouse labyrinth, obtained from the MGI database, and submitted to TissueEnrich online webtool to investigate Tissue-specific gene enrichment analysis 59 . To assess Trpv2 expression during embryonic development, the results from the Encode/screen gene expression profile were adapted to include only embryonic expression. Publicly available single-nucleus RNA-seq data from March et al. 20 and single-cell RNAseq from Jiang et al. 27 were re-analysed using the processed R object, or by replicating the authors analysis pipeline available on GitHub, respectively. All downstream visualizations, including dot plots, feature plots, and co-expression analyses, were performed using R packages: Seurat v4 61 and Nebulosa 62 . Integrated single cell RNAseq dataset (Fig.  6b ), as published by Pham et al. 26 , combining human pluripotent stem cells and cells of a pre-implantation, post-implantation and a Carnegie stage 7 human embryo (18-21 post implantation) 63 – 67 , The integration of all datasets into a human embryo transcriptome atlas was described in detail in a methodology paper by To et al. 68 Datasets for in vitro Human trophoblast stem cells and in vivo trophoblasts of 8 week placentas were downloaded from reproductivecellatlas.org and visualized using Seurat v4. The VCTfusing cluster of the human TSC dataset was subclustered using FindSubCluster(resolution = 0.3), leading to 3 subclusters. Integrated single cell RNAseq dataset (Fig.  6b ), as published by Pham et al. 26 , combining human pluripotent stem cells and cells of a pre-implantation, post-implantation and a Carnegie stage 7 human embryo (18-21 post implantation) 63 – 67 , The integration of all datasets into a human embryo transcriptome atlas was described in detail in a methodology paper by To et al. 68 Datasets for in vitro Human trophoblast stem cells and in vivo trophoblasts of 8 week placentas were downloaded from reproductivecellatlas.org and visualized using Seurat v4. The VCTfusing cluster of the human TSC dataset was subclustered using FindSubCluster(resolution = 0.3), leading to 3 subclusters. For statistical analysis, Graphpad Prism (v10.2.0) was used. Normality was evaluated using the Shapiro-Wilk Normality test. Results were considered statistically significant when p  < 0.05. Of note, equal amount of male and female placentas were considered during analysis. Litter average/genotype were used in the event of multiple fetuses per genotype. No statistical method was used to predetermine sample size. Sample sizes were based on previous studies and practical considerations. No data were excluded from the analyses unless prespecified technical quality criteria were not met. ROUT test for outliers was applied when data had high variability, and data were excluded when identified as outlier. For electrophysiological recordings, experiments with excessive voltage clamp errors were excluded. For RNA-based analyses, samples failing quality thresholds were not included. All mice were assigned a non-explicit ID, which was used for all measurements. Investigators were blinded during data collection or outcome assessment. Animal experiments were not randomized because allocation was determined by genotype within litters. Independent biological replicates and statistical tests used are described in the corresponding figure legends. Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

A constitutive Trpv2 knockout mouse on a C57BL/6 J background was generated using maternal Sox2 -Cre-mediated excision of the floxed Trpv2 allele. Transient exposure to maternally derived Cre recombinase results in efficient deletion of the floxed allele in both embryonic and extraembryonic lineages, even in offspring that does not inherit the Sox2-Cre transgene (Supplementary Fig.  1a, b ) 16 . The resulting Trpv2 heterozygous animals ( Sox2-cre wt/wt ; Trpv2 +/- ) were intercrossed to generate global Trpv2 -/- and Trpv2 +/+ littermates. Consistent with the original Trpv2 knockout report, which described recovery of only 2.5% Trpv2 -/- animals on a C57BL/6 background instead of the expected 25% 14 , a disturbed Mendelian inheritance pattern was observed at birth, together with reduced litter sizes (Table  1 , Fig.  1a and Supplementary Fig.  1c, d ). Accordingly, only 12 of 777 offspring (1.5%) recovered at birth were Trpv2 -/- , and the few surviving knockout pups were non-viable and exhibited markedly reduced birth weights (Fig.  1b ). In contrast, Trpv2 -/- fetuses were recovered at approximately Mendelian frequencies at E18.5 (55/158 fetuses), indicating that loss of Trpv2 does not substantially affect survival during most of gestation. Nevertheless, fetal weight was significantly reduced already at E16.5 (−10%) compared with Trpv2 +/+ and Trpv2 +/- littermates (Supplementary Fig.  1e ), and growth restriction further worsened by E18.5 (−32%). Consequently, whereas Trpv2 +/- fetuses nearly doubled their body weight between E16.5 and E18.5 (+ 199%), Trpv2 -/- littermates gained substantially less weight (+ 152%). Concomitantly, placental growth was reduced at E16.5 and E18.5 (Fig.  1d and Supplementary Fig.  1f ). Moreover, 37% of Trpv2 -/- fetuses at E16.5 and 91% at E18.5 weighted below the 10 th percentile of Trpv2 +/+ fetal weights (Supplementary Fig.  1g, h ). Fetal demise was observed in 25% (14/55) of Trpv2 -/- fetuses at E18.5 (Fig.  1c and Supplementary Fig.  1e ). Thus, although Trpv2 -/- fetuses remained present at near-Mendelian frequencies at E18.5, viability progressively declined towards term. Together, these findings demonstrate that complete loss of Trpv2 causes progressive late-gestation FGR associated with placental growth impairment, fetal demise and profound perinatal loss. Fig. 1 Labyrinth Trpv2 contributes to fetal growth and survival. Mendelian inheritance pattern, birth weight, and placental and fetal weight at E18.5 of global knockouts ( a–d ) generated by maternal - Sox2- Cre-mediated Trpv2 deletion. Mendelian inheritance pattern, birth weight, and placental and fetal weight at E18.5 of epiblast-specific knockouts ( e–h ) generated by paternal Sox2 -Cre-mediated Trpv2 deletion. Mendelian inheritance pattern, birth weight, and placental and fetal weight at E18.5 of junctional zone-specific knockouts ( i–l ) generated by paternal Tpbpa -Cre-mediated Trpv2 deletion. Data in ( a ), ( e ) and ( i ) are presented as mean ± SD. Box plots in ( b–d ), ( f–h ) and ( j–l ) show the median (centre line); box boundaries indicate the 25 and 75th percentiles; whiskers indicate minimum and maximum values. Data were analysed using one-way ANOVA with Tukey’s multiple-comparison correction ( a , b , e–l ) or Kruskal-Wallis test with Dunn’s multiple-comparison correction ( c ). All tests were two-sided. Exact p -values are provided in the figure panels and Source Data file. n  = 21 litters at E18.5 and n  = 13 litters at term for global knockout experiments; note that Trpv2 −/− birth weights were obtained from non-viable pups ( n  = 4 pups). n  = 5 litters at E18.5 and n  = 5 litters at term for epiblast-specific knockout experiments. n  = 8 litters at E18.5 and n  = 9 litters at term for junctional zone-specific knockout experiments. Table 1 Genotype distribution in mixed litters from Trpv2 +/- x Trpv2 +/- breedings at different developmental stages and in the general breeding colony Genotype Gestational age (days) Sum (litters) Litter size (median) +/+ (%  ±  SD) (exp: 25%) +/- (%  ±  SD) (exp: 50%) -/- (%  ±  SD) (exp: 25%) p -value E10.5 66 (8) 9 13 (20 ± 17) 39 (60 ± 19) 14 (20 ± 14) 0.6262 E12.5 60 (7) 8 16 (27 ± 16) 34 (57 ± 16) 10 (17 ± 9) 0.0304 E14.5 124 (15) 8 36 (30 ± 15) 65 (53 ± 15) 23 (17 ± 15) 0.4428 E16.5 258 (32) 9 65 (25 ± 15) 135 (53 ± 20) 58 (22 ± 14) 0.8036 E18.5 158 (21) 8 33 (21 ± 17) 70 (45 ± 19) 55 (34 ± 19) 0.1523 Term (E21) 45 (7) 6 18 (39 ± 11) 26 (59 + 10) 1 (2 ± 5) p  =  0.0057 Colony 777 (138) 6 219 (28 ± 25) 436 (55 ± 31) 12 (1 ± 5) p  <  0.0001 P-value of mendelian ratios compared to expected numbers analysed with Chi-square test. Mendelian inheritance pattern, birth weight, and placental and fetal weight at E18.5 of global knockouts ( a–d ) generated by maternal - Sox2- Cre-mediated Trpv2 deletion. Mendelian inheritance pattern, birth weight, and placental and fetal weight at E18.5 of epiblast-specific knockouts ( e–h ) generated by paternal Sox2 -Cre-mediated Trpv2 deletion. Mendelian inheritance pattern, birth weight, and placental and fetal weight at E18.5 of junctional zone-specific knockouts ( i–l ) generated by paternal Tpbpa -Cre-mediated Trpv2 deletion. Data in ( a ), ( e ) and ( i ) are presented as mean ± SD. Box plots in ( b–d ), ( f–h ) and ( j–l ) show the median (centre line); box boundaries indicate the 25 and 75th percentiles; whiskers indicate minimum and maximum values. Data were analysed using one-way ANOVA with Tukey’s multiple-comparison correction ( a , b , e–l ) or Kruskal-Wallis test with Dunn’s multiple-comparison correction ( c ). All tests were two-sided. Exact p -values are provided in the figure panels and Source Data file. n  = 21 litters at E18.5 and n  = 13 litters at term for global knockout experiments; note that Trpv2 −/− birth weights were obtained from non-viable pups ( n  = 4 pups). n  = 5 litters at E18.5 and n  = 5 litters at term for epiblast-specific knockout experiments. n  = 8 litters at E18.5 and n  = 9 litters at term for junctional zone-specific knockout experiments. Genotype distribution in mixed litters from Trpv2 +/- x Trpv2 +/- breedings at different developmental stages and in the general breeding colony P-value of mendelian ratios compared to expected numbers analysed with Chi-square test. FGR can be the result of either placental insufficiency or intrinsic failure of the embryo to thrive. According to publicly available datasets, Trpv2 is also expressed in several organs of the mouse embryo, including brain, lung and heart (Supplementary Fig.  2b ), of which abnormal development could also result in fetal compromise. However, the significant Trpv2 expression in the mouse placenta compared to other organs in the embryo or adult was remarkable. Moreover, Trpv2 was identified as specifically enhanced within the mouse placenta, meaning that its expression in the placenta is at least five-fold higher compared to the average levels in all other tissues (Supplementary Fig.  2b ). To further investigate the contribution of embryonic versus placental Trpv2 loss to FGR, an epiblast-specific Trpv2 knockout mouse was generated using paternally inherited Sox2 -Cre 16 , 17 . In this model, Trpv2 was deleted from all epiblast-derived tissues, including the fetal endothelial cells of the placenta, while expression in extraembryonic trophoblast lineages remained intact (Supplementary Figs.  3 and 4a ). This approach allowed for the comparison of Trpv2 EPI -/- and Trpv2 EPI +/+ littermates. Preserving TRPV2 function only in the extraembryonic tissues, and not in the embryo, restored the Mendelian inheritance at birth (Fig.  1e ). Moreover, birth weight, fetal weight and placental weight of Trpv2 EPI -/- were not dissimilar from Trpv2 EPI +/- or Trpv2 EPI +/+ littermates (Supplementary Table  2 and Fig.  1f–h ). Finally, viability and bodyweight at post-natal age of 8 weeks was similar between Trpv2 EPI -/- or Trpv2 EPI +/+ littermates (Supplementary Fig.  4b ). These findings demonstrated that Trpv2 in the extraembryonic trophoblast compartment, but not within the embryo proper, is required for normal placental development and fetal growth. Although Trpv2 expression was shown to be predominantly expressed in the labyrinth layer, sporadic mRNA expression was observed in late gestation in the junctional zone 15 . Thus, Trpv2 was specifically deleted in the junctional zone using Tpbpa -driven Cre expression 18 , 19 , while leaving the expression in the labyrinth intact (Supplementary Figs.  3, 4c ). No differences were observed regarding Mendelian pattern, birth weight, and fetal and placental weight at E18.5 between Trpv2 JZ -/- and Trpv2 JZ +/+ littermates (Supplementary Table  1 and Fig.  1i–l ) suggesting that Trpv2 expression in Tpbpa -positive cells from the junctional zone is not essential for placental development. From these experiments, it can be deduced that Trpv2 within the labyrinth arbitrates developmental outcomes. At E8.5 and E9.5, but not at E7.5, Trpv2 was abundantly expressed in Prl3d1 ⁺ primary parietal trophoblast giant cells (P-TGCs) and in the chorion, while remaining low in the ectoplacental cone (Supplementary Fig.  5 ). Following chorioallantoic attachment, Trpv2 expression was strongly detected in regions of Synb ⁺ SynTII invasion and persisted throughout the emerging labyrinth (Fig.  2a ). To define the spatial organization of Trpv2 expression, MERscope-based spatial transcriptomics was performed at E9.5 and E16.5 using a targeted panel of trophoblast and vascular markers (Supplementary Fig.  6 ; full gene list provided in Source Data ). At E9.5, Trpv2 expression was broadly detected across the chorion and developing labyrinth, including regions adjacent to Epcam ⁺ trophoblast progenitors, indicating its presence in trophoblast compartments prior to overt lineage segregation (Fig.  2b ). No overlap was observed with Pecam1⁺/Cd34 ⁺ fetal endothelial cells, demonstrating that Trpv2 expression was restricted to trophoblast populations (Fig.  2b–d ). Within the developing labyrinth, Trpv2 expression displayed spatial heterogeneity, with signals in Synb ⁺ differentiating SynTII cells, as well as in adjacent regions expressing early SynTII markers such as Slc5a5 and Gcm1 (Fig.  2c and Supplementary Fig.  7a ). This pattern suggests that Trpv2 expression precedes full SynTII maturation and marks early trophoblast states. In contrast, Trpv2 showed minimal overlap with SynTI-associated markers, including Syna and Atp2a3 (Supplementary Fig.  8b ). A subset of Trpv2 ⁺ cells co-expressed Mfsd2a , consistent with early S-TGC-associated regions that later correspond to mature S-TGCs marked by Lepr or Ctsq expression (Supplementary Fig.  8a, c ). These lineage associations were confirmed at later developmental stages by RNAscope (E10.5 and E12.5), showing persistent Trpv2 expression in SynTII ( Synb ⁺) and S-TGC ( Ctsq ⁺) compartments, while remaining absent from SynTI ( Slc16a1 ⁺) cells and fetal endothelium (Supplementary Fig.  9 ). Analysis of published single-cell and single-nucleus RNA sequencing datasets 20 , 21 further supported alignment of Trpv2 expression with SynTII and S-TGC differentiation trajectories (Supplementary Fig.  8c–h ). Together, these data indicate that Trpv2 marks early labyrinth trophoblast populations that give rise to SynTII and S-TGC lineages, while being largely excluded from the SynTI trajectory. Fig. 2 Trpv2 expression in the developing mouse placenta. a Fluorescence in situ hybridization (FISH) of Trpv2 and Synb in E8.5 placentas. Scale bars, 250 µm, 50 µm and 25 µm from left to right. The right-most panel shows Trpv2 signal restricted to cells identified by the indicated marker. Representative images from n  = 3 placentas with similar results. b MERscope spatial transcriptomic analysis of Trpv2 expression in E9.5 placentas. Higher magnification shows co-localization with Synb + Syncytiotrophoblast-II (SynTII) cells (region 1) and expression in adjacent Synb -low areas (arrows) enriched for early SynTII markers Slc5a5 and Gcm1 (region 2), indicating expression prior to full SynTII maturation. Scale bars, 250 µm and 50 µm. Data are representative of n  = 3 biologically independent placentas. c , d FISH of Trpv2 with Synb ( c ) or Pecam1 ( d ) in E10.5 placentas. Scale bars, 250 µm, 50 µm and 25 µm from left to right. The right-most panel shows Trpv2 signal restricted to cells identified by the indicated marker. Representative images from n  = 3 placentas with similar results. a Fluorescence in situ hybridization (FISH) of Trpv2 and Synb in E8.5 placentas. Scale bars, 250 µm, 50 µm and 25 µm from left to right. The right-most panel shows Trpv2 signal restricted to cells identified by the indicated marker. Representative images from n  = 3 placentas with similar results. b MERscope spatial transcriptomic analysis of Trpv2 expression in E9.5 placentas. Higher magnification shows co-localization with Synb + Syncytiotrophoblast-II (SynTII) cells (region 1) and expression in adjacent Synb -low areas (arrows) enriched for early SynTII markers Slc5a5 and Gcm1 (region 2), indicating expression prior to full SynTII maturation. Scale bars, 250 µm and 50 µm. Data are representative of n  = 3 biologically independent placentas. c , d FISH of Trpv2 with Synb ( c ) or Pecam1 ( d ) in E10.5 placentas. Scale bars, 250 µm, 50 µm and 25 µm from left to right. The right-most panel shows Trpv2 signal restricted to cells identified by the indicated marker. Representative images from n  = 3 placentas with similar results. In both Trpv2 -/- and Trpv2 +/+ placentas, the three placental layers were easily discernible (Supplementary Fig.  10a, d ). The labyrinth volume, but not the relative fraction, was smaller in E16.5 Trpv2 -/- placentas (−14%; Supplementary Fig.  10b, c ), and significantly reduced at E18.5 (−26%, Supplementary Fig.  10e, f ), compared to Trpv2 +/+ littermates, while the maternal decidua and junctional zone volumes and densities were unaffected. The morphology of the labyrinth was further analyzed using double-label immunohistochemistry with lectin as fetal endothelial cell marker and cytokeratin as pan-trophoblast cell marker 22 (Supplementary Fig.  10g, h ). Quantification of the labyrinth’s cellular subtypes revealed a significantly reduced fraction of fetal capillaries at E16.5 in Trpv2 -/- placentas compared to Trpv2 +/+ littermates (−21%), along with a significantly increased fraction of trophoblasts (+ 43%, Supplementary Fig.  10i ). After accounting for labyrinth volume, Trpv2 -/- placentas were shown to have significantly reduced fetal capillaries volume (−28%), associated with a numerical increase in trophoblast volume (+ 48%, p  = 0.058) at E16.5 (Supplementary Fig.  10j ). Moreover, the labyrinth architecture of Trpv2 -/- placentas was severely disrupted, characterized by regions of densely packed cytokeratin-positive trophoblast cells that completely lack fetal vasculature (Fig.  3a ). At E16.5 and E18.5, this phenotype was fully penetrant, with cellular malformations consistently observed in all Trpv2 -/- samples but absent in Trpv2 +/+ controls, and displaying a random distribution across the labyrinth layer. The extent of the defect was variable, covering between 10 and 45% of the labyrinth, with an average around 25% at all the examined stages of gestation (Supplementary Fig.  10k, l ). Importantly, no labyrinth malformation was observed in Trpv2 -deficient animals reconstituted with a functionally normal placenta ( Trpv2 EPI -/- ) (Supplementary Fig.  10m, n ), suggesting that a primary vascular defect of the fetal capillaries is unlikely. Fig. 3 SynTII cell differentiation is affected in Trpv2 -/- placentas. a Lectin-cytokeratin double staining at E16.5 showing normal (insert 1) and abnormal (insert 2) labyrinth morphology in Trpv2 −/− placentas. Scale bar = 400 μm and 50 μm. Representative images from n  = 12–13 placentas per genotype from 8 independent litters. Similar results were observed in all Trpv2 −/− placentas analysed. b Immunohistochemistry of MCT1 and MCT4, representing differentiated SynTI and SynTII cells, respectively. Arrows indicate SynTI cells with altered morphology. Scale bar = 250 μm and 20 μm. Staining was performed on consecutive sections. Representative images from n  = 5–6 placentas per genotype from 6 independent litters with similar results. c Enhanced volcano plot of differentially expressed genes identified by bulk RNA-sequencing of E14.5 labyrinth tissue ( Trpv2 +/+ versus Trpv2 −/− ; 111 downregulated and 56 upregulated genes). Differential expression analysis was performed using DESeq2 with Benjamini-Hochberg correction for multiple testing. n  = 6 placentas per genotype (3 male and 3 female) from 5 independent litters. Marker genes for different labyrinth trophoblast populations are indicated. d qRT-PCR validation of selected differentially expressed genes identified by RNA-sequencing. Data are presented as mean values ± SEM. Statistical analysis was performed using two-sided Student’s t tests on ΔCt values. n  = 6 placentas per genotype (3 male and 3 female) from 5 independent litters. Exact p-values are also provided in the Source Data file. e Cnetplot showing the association between differentially expressed genes and enriched Gene Ontology biological processes. f Overrepresentation analysis showing enrichment of SynTII marker genes among downregulated genes. Statistical significance was determined by hypergeometric testing with correction for multiple testing. g MERscope spatial transcriptomic overview of E16.5 placental sections from wild-type (WT) and Trpv2 -deficient placentas (KO1, KO2). Shown are Bdkrb2 (junctional zone), Pecam1 (fetal endothelial cells), and Vtcn1 (SynTII), illustrating overall placental architecture and organisation of major placental compartments used for subsequent region-based analyses. Scale bar = 1 mm h MERscope spatial transcriptomic overview illustrating region-based annotation of labyrinth tissue used for downstream analyses; wild-type (WT) labyrinth regions and two classes of knockout labyrinth regions: morphologically normal regions (NC) and abnormal regions lacking fetal capillaries (AbC). Scale bar = 1 mm i Heatmap of region-based pseudo-bulk expression profiles derived from MERscope analysis comparing WT, NC and AbC labyrinth regions. Samples and genes are ordered by unsupervised hierarchical clustering. Colour scale indicates relative expression (scaled z-score). j Representative MERscope expression maps of SynTI-associated markers ( Atp2a3, Tgfa and Syna ) and the SynTII precursor-associated marker Slc5a5 in WT and Trpv2 -deficient placentas at E16.5, illustrating the spatial distribution of trophoblast lineage markers in normal and abnormal labyrinth regions. Scale bars = 1 mm. a Lectin-cytokeratin double staining at E16.5 showing normal (insert 1) and abnormal (insert 2) labyrinth morphology in Trpv2 −/− placentas. Scale bar = 400 μm and 50 μm. Representative images from n  = 12–13 placentas per genotype from 8 independent litters. Similar results were observed in all Trpv2 −/− placentas analysed. b Immunohistochemistry of MCT1 and MCT4, representing differentiated SynTI and SynTII cells, respectively. Arrows indicate SynTI cells with altered morphology. Scale bar = 250 μm and 20 μm. Staining was performed on consecutive sections. Representative images from n  = 5–6 placentas per genotype from 6 independent litters with similar results. c Enhanced volcano plot of differentially expressed genes identified by bulk RNA-sequencing of E14.5 labyrinth tissue ( Trpv2 +/+ versus Trpv2 −/− ; 111 downregulated and 56 upregulated genes). Differential expression analysis was performed using DESeq2 with Benjamini-Hochberg correction for multiple testing. n  = 6 placentas per genotype (3 male and 3 female) from 5 independent litters. Marker genes for different labyrinth trophoblast populations are indicated. d qRT-PCR validation of selected differentially expressed genes identified by RNA-sequencing. Data are presented as mean values ± SEM. Statistical analysis was performed using two-sided Student’s t tests on ΔCt values. n  = 6 placentas per genotype (3 male and 3 female) from 5 independent litters. Exact p-values are also provided in the Source Data file. e Cnetplot showing the association between differentially expressed genes and enriched Gene Ontology biological processes. f Overrepresentation analysis showing enrichment of SynTII marker genes among downregulated genes. Statistical significance was determined by hypergeometric testing with correction for multiple testing. g MERscope spatial transcriptomic overview of E16.5 placental sections from wild-type (WT) and Trpv2 -deficient placentas (KO1, KO2). Shown are Bdkrb2 (junctional zone), Pecam1 (fetal endothelial cells), and Vtcn1 (SynTII), illustrating overall placental architecture and organisation of major placental compartments used for subsequent region-based analyses. Scale bar = 1 mm h MERscope spatial transcriptomic overview illustrating region-based annotation of labyrinth tissue used for downstream analyses; wild-type (WT) labyrinth regions and two classes of knockout labyrinth regions: morphologically normal regions (NC) and abnormal regions lacking fetal capillaries (AbC). Scale bar = 1 mm i Heatmap of region-based pseudo-bulk expression profiles derived from MERscope analysis comparing WT, NC and AbC labyrinth regions. Samples and genes are ordered by unsupervised hierarchical clustering. Colour scale indicates relative expression (scaled z-score). j Representative MERscope expression maps of SynTI-associated markers ( Atp2a3, Tgfa and Syna ) and the SynTII precursor-associated marker Slc5a5 in WT and Trpv2 -deficient placentas at E16.5, illustrating the spatial distribution of trophoblast lineage markers in normal and abnormal labyrinth regions. Scale bars = 1 mm. While the regions that lack fetal capillaries retained cytokeratin expression, the cells did not originate from the junctional zone, given the absence of Tpbpa expression (Supplementary Fig.  11a–c ). Moreover, intense alkaline phosphatase reactivity was observed, which localizes between SynTI and S-TGC and in the maternal blood spaces, and identified these cells as labyrinth trophoblasts (Supplementary Fig.  11d ). Finally, Masson’s Trichrome staining did not reveal any apparent fibrosis in these areas (Supplementary Fig.  11e ). Interestingly, the lack of fetal capillaries in malformed areas was concomitant with the absence of SynTII cells (Fig.  3a, b and Supplementary Fig.  12a ), as the SynTII marker Slc16a3 (MCT4) was not detected, whereas Slc16a1 + (MCT1) SynTI and Ctsq + S-TGC were present (Supplementary Fig.  12b ). On consecutive slices, the lack of MCT4, but not MCT1, at the protein level confirmed the absence of SynTII cells in malformed areas (Fig.  3b ), an observation that was never detected in Trpv2 +/+ littermate placentas (Supplementary Fig.  12c, d ). Interestingly, MCT1-positive SynTI cells in areas lacking SynTII presented with an altered morphology resembling uni-nucleated cells, rather than thin, stretched syncytium cells, indicating that SynTI cells might have impaired fusion without affecting their formation. To further identify the molecular signatures associated with the phenotypic differences observed between Trpv2 -/- and Trpv2 +/+ , the transcriptome of micro-dissected labyrinth samples at E14.5 was analyzed, when labyrinth malformation was present, but placental and fetal weight were not yet affected. Differentially expressed genes (DEGs) that were significantly downregulated included genes that are specifically expressed in SynTII cells ( Gcm1, Synb, Slc16a3, Vegfa, Apela, Gcgr) and fetal endothelial cells ( Pecam1, Wnt2b, Meox2, Vcam1 ), while genes expressed in SynTII precursors ( Lgr5 ), S-TGC ( Ctsq) , SynTI cells ( Syna, Dlx3) or non-cell specific labyrinth genes ( Esx1, Cebpb, Nr6a1) (Fig.  3c ) were unaffected (Supplementary Data  1 ). Decreased expression of SynTII and endothelial genes was confirmed by qRT-PCR (Fig.  3d ). Gene ontology analysis of downregulated DEGs in Trpv2 -/- placentas revealed enrichment of biological processes related to branching morphogenesis, cell fusion, syncytium formation and angiogenesis (Fig.  3e ). Moreover, SynTII marker genes were significantly overrepresented in downregulated DEGs (Fig.  3f ). Beyond downregulation of SynTII-associated genes, bulk RNA-sequencing of E14.5 labyrinth revealed up-regulation of genes associated with immune- and interferon-related GO terms (Supplementary Fig.  13a, b and Supplementary Data  1 ). Inspection of the underlying gene sets demonstrated that these terms were driven predominantly by interferon-responsive and stress-associated regulators rather than canonical immune cell identity genes. Similar findings were observed at E16.5 with respect to reduced expression of SynTII markers (Supplementary Fig.  14 and Supplementary Data  1 ). However, in contrast to E14.5, upregulated genes at E16.5 did not show significant enrichment for immune-related pathways. Together, the decrease in SynTII and fetal endothelial markers supports the histological observation of reduced SynTII and endothelial cells in malformed areas. To spatially contextualize the transcriptional changes observed in histology and bulk RNA‑sequencing, region‑based MERscope spatial transcriptomic analysis was performed at E16.5 using a targeted gene panel enriched for trophoblast and vascular markers (Fig.  3g–j and Supplementary Fig.  15 ). Consecutive tissue sections confirmed the correspondence between histological abnormalities, altered SynTII markers, and spatial transcriptomic profiles (Supplementary Fig.  15a, b ). MERscope signal was uniformly detected across the placental section, excluding regional assay failure as an explanation for transcriptional differences (Supplementary Fig.  15c ). Annotation of canonical placental compartments verified accurate spatial assignment of trophoblast, vascular, and maternal tissues (Supplementary Fig.  15d ). Given the multinucleated syncytial nature of trophoblast lineages, analyses were performed at the level of spatially defined regions rather than individual cells (Supplementary Fig.  15e ). Principal component analysis and region‑based differential expression confirmed that wild‑type, morphologically normal knockout regions, and abnormal knockout regions represent distinct transcriptional states (Supplementary Fig.  15f, g ). Unbiased clustering of regional pseudo-bulk expression profiles revealed that abnormal labyrinth regions in Trpv2 ‑deficient placentas represent transcriptionally coherent states, distinct from both wild‑type and morphologically normal knockout regions (Fig.  3i ). These regions exhibited coordinated loss of mature SynTII and exchange‑associated programs, together with enrichment of alternative trophoblast transcriptional programs, including SynTI‑associated ( Syna ) and precursor‑associated signatures ( Slc5a5, Tgfa ) (Fig.  3j ). In addition, genes specifically enriched in placental pericytes, including Acta2, Ace2, Tagln , and Gpc3 , were decreased in both normal and abnormal areas of knockout placentas, suggesting a reduction in the pericyte signature and potentially reduced vascular support (Supplementary Fig.  16 ). Canonical immune cell markers were sparse and not enriched in abnormal regions, indicating absence of immune cell infiltration (Supplementary Fig.  13d ). CRISPR/Cas9-mediated ablation of Trpv2 in mTSCs was used to further determine its trophoblast-specific functions (Fig.  4a and Supplementary Fig.  17a, b ). Efficient knockout of Trpv2 was verified by decreased mRNA expression, as well as decreased functional expression shown by the lack of Tetrahydrocannabinol (THC)-induced TRPV2-mediated calcium influx or currents in Trpv2 -KO mTSC (Fig.  4b–d and Supplementary Fig.  17c ). Stemness of mTSC was unaffected in Trpv2 -KO mutants as they appeared to self-renew and expand for over 10 passages, and evidenced by similar expression of stem markers  Eomes , Elf5 and Cdx2 (Fig.  4f ) . Upon induction of differentiation, Trpv2 -KO mTSCs showed a reduced induction of selected SynTII - associated markers, including Gcm1 and its downstream target Synb , while expression of markers associated with other trophoblast lineages was less affected (Fig.  4e, f ). Activation of canonical Wnt signaling by the addition of CHIR99021 forced the differentiation specifically towards the SynTII lineage, whereas markers for other trophoblast subtypes were repressed (Supplementary Fig.  18a, b ). A similar defect was observed in Trpv2 -KO mTSCs, demonstrated by decreased expression of Gcm1 , and to a lesser extent Synb , (Supplementary Fig.  18c ). To further assess the transcriptional programs underlying SynTII differentiation, bulk RNA sequencing was performed on WT and Trpv2 -KO mTSCs after 3 days of differentiation under standard and CHIR99021-treated conditions (Supplementary Fig.  18d and Supplementary Data  1 ). Under standard differentiation conditions, transcriptional differences between WT and Trpv2 -KO cells were relatively limited. In contrast, CHIR99021 treatment induced a robust and coordinated activation of SynTII-associated transcriptional programs in WT mTSCs, whereas this response was markedly attenuated in Trpv2 -KO cells (Fig.  4g ). Together, these data indicate that TRPV2 is not required for trophoblast stem cell maintenance but is necessary for efficient activation of SynTII differentiation programs, particularly under conditions that promote SynTII-lineage specification. Fig. 4 SynTII differentiation is impaired in Trpv2 -KO mouse trophoblast stem cells. a Representative images of scrambled control (Scr) and Trpv2 -KO mTSCs, acquired using a 4x objective. b qRT-PCR using primers spanning exon 5–6 (within the targeted region) and exon 13–14 (downstream of the targeted region). Data are presented as mean values ± SEM. Two-sided Student’s t test on ΔCt values. n  = 5 independent clones per genotype. c Representative Fura-2AM calcium imaging traces from Scr and Trpv2 -KO mTSCs differentiated for 6 days. Cells were stimulated with THC (50 μM) as a TRPV2 agonist and ionomycin (2 μM) as a positive control. d Percentage of THC-responsive cells (263/493 Scr cells and 6/783 Trpv2 -KO cells, N  = 4 measurements). Data are presented as mean values ± SEM and were analysed using two-sided Fisher’s exact test. e Schematic overview of mTSC differentiation pathways and marker genes. Created in BioRender. Leuven, L. (2026) https://BioRender.com/vayiu8w . f Analysis of Trpv2 -KO mutants under self-renewal conditions (TSC) or following differentiation for 3 days (3D) or 6 days (6D). Data are presented as mean values ± SEM. Two-way ANOVA with Šidák correction for multiple comparisons on ΔCt values. n  = 5–10 independent clones per genotype. Exact sample sizes for each gene and condition are provided in the Source Data file. Exact p -values are also provided in the Source Data file. g Heatmap showing gene set-based transcriptional activity across Scr and Trpv2 -KO mTSCs under stem-cell conditions, 3-day differentiation, and 3-day CHIR99021-treated conditions, together with magnification of SynTII precursor and SynTII marker genes. Colour scale indicates relative expression (scaled z-score). Bulk RNA-sequencing was performed on n  = 3 independent clones per genotype and condition. a Representative images of scrambled control (Scr) and Trpv2 -KO mTSCs, acquired using a 4x objective. b qRT-PCR using primers spanning exon 5–6 (within the targeted region) and exon 13–14 (downstream of the targeted region). Data are presented as mean values ± SEM. Two-sided Student’s t test on ΔCt values. n  = 5 independent clones per genotype. c Representative Fura-2AM calcium imaging traces from Scr and Trpv2 -KO mTSCs differentiated for 6 days. Cells were stimulated with THC (50 μM) as a TRPV2 agonist and ionomycin (2 μM) as a positive control. d Percentage of THC-responsive cells (263/493 Scr cells and 6/783 Trpv2 -KO cells, N  = 4 measurements). Data are presented as mean values ± SEM and were analysed using two-sided Fisher’s exact test. e Schematic overview of mTSC differentiation pathways and marker genes. Created in BioRender. Leuven, L. (2026) https://BioRender.com/vayiu8w . f Analysis of Trpv2 -KO mutants under self-renewal conditions (TSC) or following differentiation for 3 days (3D) or 6 days (6D). Data are presented as mean values ± SEM. Two-way ANOVA with Šidák correction for multiple comparisons on ΔCt values. n  = 5–10 independent clones per genotype. Exact sample sizes for each gene and condition are provided in the Source Data file. Exact p -values are also provided in the Source Data file. g Heatmap showing gene set-based transcriptional activity across Scr and Trpv2 -KO mTSCs under stem-cell conditions, 3-day differentiation, and 3-day CHIR99021-treated conditions, together with magnification of SynTII precursor and SynTII marker genes. Colour scale indicates relative expression (scaled z-score). Bulk RNA-sequencing was performed on n  = 3 independent clones per genotype and condition. To further elucidate its importance in SynTII cells, Trpv2 was conditionally deleted using the Gcm1 -Cre deleter strain 23 , 24 , which, in the placenta specifically targets SynTII cells (Supplementary Fig.  19a ). Trpv2 was shown to be expressed in differentiating Gcm1 + SynTII cells (Supplementary Fig.  19b ). Deleting Trpv2 in SynTII cells partially recapitulated the phenotype observed in global Trpv2 deficient mice. While mortality was not affected in Trpv2 SynTII -/- animals, reduced body weight was observed in Trpv2 SynTII -/- fetuses at E18.5 and at birth compared to Trpv2 SynTII +/+ littermates (−8% and −27%, respectively, Fig.  5a, b ). Heterozygote conditional knockouts ( Trpv2 SynTII +/- ) did not significantly differ from Trpv2 SynTII +/+ littermates (Supplementary Fig.  19d ). Fetal or placental weight of Trpv2 SynTII -/- at E16.5 was not altered (Fig.  5b, c ). Trpv2 SynTII -/- specimens at E18.5 revealed similar labyrinth malformation as the constitutive Trpv2 knockout mouse (Fig.  5d, e ). Likewise, areas lacking fetal capillaries were also devoid of SynTII cells, while SynTI marker–positive cells were present but did not exhibit the characteristic elongated and flattened morphology associated with fully differentiated syncytiotrophoblast cells. (Supplementary Fig.  19g ). The Gcm1 transgene is imprinted and expressed only when inherited from the male 20 . As such, the reciprocal cross using male Trpv2 F /F x female Gcm1 -Cre wt/tg ; Trpv2 +/- did not result in decreased birth weight in Gcm1 -Cre wt/tg ; Trpv2 F /- fetuses (Supplementary Fig.  19e ). Together, these findings indicate that deleting Trpv2 specifically from Gcm1 + SynTII cells is sufficient to cause FGR, but at a lower penetrance and expressivity as the constitutive Trpv2 knockout mouse, so that it does not compromise embryo survival. Fig. 5 Phenotype of Trpv2 SynTII-/- placentas. a Mendelian inheritance pattern of litters obtained from Gcm1 -Cre wt/tg ; Trpv2 +/- crossed with female Trpv2 F/F mice. Data are shown as litter means ± SEM. b Fetal weight of Trpv2 SynTII-/- and Trpv2 SynTII+/+ fetuses at E16.5, E18.5 and birth. Box plots show the median (centre line); box boundaries indicate the 25th and 75th percentiles; whiskers indicate minimum and maximum values. Statistical analysis was performed using two-way ANOVA with Holm-Šídák’s multiple comparisons test. n  = 11 litters at E16.5, n  = 15 litters at E18.5 and n  = 9 litters at term. c Placental weight of Trpv2 SynTII-/- and Trpv2 SynTII+/+ fetuses at E16.5 and E18.5. Box plots show the median (centre line); box boundaries indicate the 25th and 75th percentiles; whiskers indicate minimum and maximum values. Two-sided unpaired Student’s t test. n  = 11 litters at E16.5 and n  = 15 litters at E18.5. d Lectin-cytokeratin double staining of E18.5 Trpv2 SynTII-/- placentas showing regions lacking fetal capillaries. Scale bars, 500 μm and 50 μm. Representative images from n  = 15 Trpv2 SynTII-/- placentas and n  = 10 control placentas collected from 13 independent litters. Similar results were observed in all affected Trpv2 SynTII-/- placentas analysed. e Quantification of fetal capillary-void areas within the labyrinth at E18.5. Data were analysed using a two-sided unpaired Student’s t test. n  = 10 control placentas and n  = 15 Trpv2 SynTII-/- placentas from 13 independent litters. Exact p -values are also provided in the Source Data file. a Mendelian inheritance pattern of litters obtained from Gcm1 -Cre wt/tg ; Trpv2 +/- crossed with female Trpv2 F/F mice. Data are shown as litter means ± SEM. b Fetal weight of Trpv2 SynTII-/- and Trpv2 SynTII+/+ fetuses at E16.5, E18.5 and birth. Box plots show the median (centre line); box boundaries indicate the 25th and 75th percentiles; whiskers indicate minimum and maximum values. Statistical analysis was performed using two-way ANOVA with Holm-Šídák’s multiple comparisons test. n  = 11 litters at E16.5, n  = 15 litters at E18.5 and n  = 9 litters at term. c Placental weight of Trpv2 SynTII-/- and Trpv2 SynTII+/+ fetuses at E16.5 and E18.5. Box plots show the median (centre line); box boundaries indicate the 25th and 75th percentiles; whiskers indicate minimum and maximum values. Two-sided unpaired Student’s t test. n  = 11 litters at E16.5 and n  = 15 litters at E18.5. d Lectin-cytokeratin double staining of E18.5 Trpv2 SynTII-/- placentas showing regions lacking fetal capillaries. Scale bars, 500 μm and 50 μm. Representative images from n  = 15 Trpv2 SynTII-/- placentas and n  = 10 control placentas collected from 13 independent litters. Similar results were observed in all affected Trpv2 SynTII-/- placentas analysed. e Quantification of fetal capillary-void areas within the labyrinth at E18.5. Data were analysed using a two-sided unpaired Student’s t test. n  = 10 control placentas and n  = 15 Trpv2 SynTII-/- placentas from 13 independent litters. Exact p -values are also provided in the Source Data file. Chorionic villi of the human placenta are covered by a continuous layer of syncytiotrophoblast (ST) cells, established through fusion of the underlying (dis)continuous layer of cytotrophoblast cells. Within the core of the villi, fetal capillaries are found interspersed with mesenchymal cells. In first trimester and term human placentas, TRPV2 mRNA was consistently found in the mesenchymal core of the villi, along with some sporadic signal in CGA + syncytiotrophoblast cells (Fig.  6a and Supplementary Fig.  20a, b ). Apart from this definitive syncytium, which is responsible for nutrient transfer and hormone production, a different type of ST exists during embryo implantation, namely the primitive syncytium. It is the first trophoblast subtype to differentiate from the trophectoderm, and presents as an invasive, multinucleated structure necessary to mediate embryo implantation 25 . Investigating the primitive syncytium from human samples is, however, greatly confounded by ethical and practical considerations, and prompted the use of in vitro models. In an integrated single cell RNAseq dataset 26 , combining human pluripotent stem cells and cells of a pre-implantation, post-implantation and a Carnegie stage 7 human embryo (18-21 post implantation), TRPV2 was highly expressed in ‘late trophoblast cells (TB.late)’, ‘pre-syncytium cells (Pre.ST)’, and ‘syncytiotrophoblasts (ST)’ (Fig.  6b ). Fig. 6 TRPV2 expression and function in human trophoblast cells. a Fluorescence in situ hybridization (FISH) on first-trimester and term human placental tissues showing TRPV2 expression. Arrows indicate syncytiotrophoblast (ST) cells. CDH1 , cytotrophoblast marker; CGA , syncytiotrophoblast marker. Scale bars = 50 µm. Representative images from n  = 2 independent placentas with similar results. b UMAP projection and FeaturePlot of an integrated single-cell RNA-sequencing dataset 26 combining human pluripotent stem cells and cells from pre-implantation embryos, post-implantation embryos and Carnegie stage 7 human embryos. Colour scale indicates average expression. c qRT-PCR analysis of hTSC cultured under stem cell conditions and during syncytiotrophoblast differentiation. Data are presented as mean ± SEM. n  = 3 independent differentiations. Statistical analysis was performed using two-way ANOVA with Dunnett’s multiple-comparison correction (two-sided). Exact p -values are provided in the Source Data file. d FISH on hTSC differentiated towards syncytiotrophoblasts for 1 day. TRPV2 signal was observed in the cytoplasm and nuclei (arrow). SDC1 , syncytiotrophoblast marker. Scale bar = 100 µm. i shows a magnified multinucleated syncytiotrophoblast; scale bar = 50 µm. Representative images from n  = 3 independent differentiations with similar results. e Representative calcium imaging experiments showing calcium levels during hTSC differentiation upon THC application. f THC-induced calcium responses in differentiating trophoblast cells. Data are presented as mean ± SEM. Statistical analysis was performed using the One-way Anova using Tukey’s multiple comparisons test. Exact p -values are provided in the Source Data file. g Mean basal F340/F380 ratios during 5 min of measurement in hTSC and differentiating syncytiotrophoblasts. h Percentage of cells classified as having elevated basal calcium levels. Data in ( g ) and ( h ) are presented as mean ± SEM. n  = 14–18 measurements obtained from N  = 3 independent differentiations. Statistical analysis was performed using one-way ANOVA with Tukey’s multiple-comparison correction (two-sided). Exact p -values are provided in the Source Data file. i Comparison of basal calcium levels between TRPV2 -positive and TRPV2- negative cells following consecutive RNAscope analysis on the same coverslips. Colour scale indicates F340/F380 ratio. Scale bar = 100 µm. Each data point represents a single cell identified by nuclear staining and matched to a corresponding calcium-imaging region of interest. Analysis was performed on one coverslip per differentiation condition and included approximately 300–400 individual cells per condition. Statistical analysis was performed using a two-sided Mann-Whitney test. Exact p -values are provided in the figure and Source Data file. j UMAP projection and DotPlot analysis of publicly available single-cell RNA-sequencing data from in vitro hTSC cultures with sub-analysis of VCT-fusing cells 27 . Colour scale indicates average expression. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001. a Fluorescence in situ hybridization (FISH) on first-trimester and term human placental tissues showing TRPV2 expression. Arrows indicate syncytiotrophoblast (ST) cells. CDH1 , cytotrophoblast marker; CGA , syncytiotrophoblast marker. Scale bars = 50 µm. Representative images from n  = 2 independent placentas with similar results. b UMAP projection and FeaturePlot of an integrated single-cell RNA-sequencing dataset 26 combining human pluripotent stem cells and cells from pre-implantation embryos, post-implantation embryos and Carnegie stage 7 human embryos. Colour scale indicates average expression. c qRT-PCR analysis of hTSC cultured under stem cell conditions and during syncytiotrophoblast differentiation. Data are presented as mean ± SEM. n  = 3 independent differentiations. Statistical analysis was performed using two-way ANOVA with Dunnett’s multiple-comparison correction (two-sided). Exact p -values are provided in the Source Data file. d FISH on hTSC differentiated towards syncytiotrophoblasts for 1 day. TRPV2 signal was observed in the cytoplasm and nuclei (arrow). SDC1 , syncytiotrophoblast marker. Scale bar = 100 µm. i shows a magnified multinucleated syncytiotrophoblast; scale bar = 50 µm. Representative images from n  = 3 independent differentiations with similar results. e Representative calcium imaging experiments showing calcium levels during hTSC differentiation upon THC application. f THC-induced calcium responses in differentiating trophoblast cells. Data are presented as mean ± SEM. Statistical analysis was performed using the One-way Anova using Tukey’s multiple comparisons test. Exact p -values are provided in the Source Data file. g Mean basal F340/F380 ratios during 5 min of measurement in hTSC and differentiating syncytiotrophoblasts. h Percentage of cells classified as having elevated basal calcium levels. Data in ( g ) and ( h ) are presented as mean ± SEM. n  = 14–18 measurements obtained from N  = 3 independent differentiations. Statistical analysis was performed using one-way ANOVA with Tukey’s multiple-comparison correction (two-sided). Exact p -values are provided in the Source Data file. i Comparison of basal calcium levels between TRPV2 -positive and TRPV2- negative cells following consecutive RNAscope analysis on the same coverslips. Colour scale indicates F340/F380 ratio. Scale bar = 100 µm. Each data point represents a single cell identified by nuclear staining and matched to a corresponding calcium-imaging region of interest. Analysis was performed on one coverslip per differentiation condition and included approximately 300–400 individual cells per condition. Statistical analysis was performed using a two-sided Mann-Whitney test. Exact p -values are provided in the figure and Source Data file. j UMAP projection and DotPlot analysis of publicly available single-cell RNA-sequencing data from in vitro hTSC cultures with sub-analysis of VCT-fusing cells 27 . Colour scale indicates average expression. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001. During in vitro differentiation of human trophoblast stem cells (hTSC) towards ST, TRPV2 was significantly upregulated, in line with other ST genes including CGB3 and ERVW-1 (Fig.  6c ). Importantly, FISH experiments further revealed that the increased expression was due to strong upregulation in ST ( CGA + , SDC1 + ) cells, which could be identified by their altered morphology. This upregulation was already evident 24 h after differentiation, and further increased over time (Fig.  6d and Supplementary Fig.  20c ). THC application induced calcium influx and in- and outwardly rectifying currents in line with TRPV2 functionality (Fig.  6e, f and Supplementary Fig.  21a–c ). Interestingly, live imaging of calcium signals using Fura2-AM revealed increased mean basal F340/F380 levels in both 1 and 3 days differentiated ST, so that the percentage of cells with higher basal F340/F380 levels was significantly increased (Fig.  6g, h ). Importantly, mean basal F340/F380 levels were significantly higher in TRPV2 positive cells (Fig.  6i ). Fusion is an important aspect of ST differentiation. Subanalysis of an in vitro hTSC scRNA-seq dataset 27 revealed that TRPV2 was abundantly present in a subset of fusing villous trophoblast cells (VCT-fusing_0) that are characterized by markers of the primitive ST ( SP6 + , ID2 + ) (Fig.  6j and Supplementary Fig.  20d ). Interestingly, marker genes of this subcluster are also enriched in the VCT-fusing cluster of an in vivo scRNA-seq dataset of the fetal-maternal interface at 8 weeks post conception 27 (Supplementary Fig.  20e ). In contrast, VCT-fusing_1 cells were not defined by TRPV2 and correspond to in vivo definitive syncytium cells (SCT cluster, ERVW-1 + , PLAC4 + , CGB1 + ). Finally, TRPV2 expression was quantified in human placentas collected at term (T) from appropriate for gestational age (AGA) or early-onset FGR (FGR diagnosed <32 week’s gestation) pregnancies, or from preterm (PT, <37 week’s gestation) early-onset FGR pregnancies. Maternal age, pre-pregnancy body mass index (BMI), the proportional distributions of fetal sex and delivery mode were similar in AGA and FGR groups, irrespective of gestational age at FGR onset and at delivery (Table  2 ). As expected, women in the PT, early-onset FGR delivered earlier than the term groups (Supplementary Fig.  21c ), and birth weight is lower in FGR groups (Supplementary Fig.  21d ). While TRPV2 expression was reduced in a subset of FGR cases, particularly those delivered preterm, no difference was observed between term FGR samples and gestational age–matched controls (Supplementary Fig.  21e ). Interpretation of these findings is limited by differences in gestational age at delivery and the small number of early-onset cases available for analysis. Moreover, the heterogeneity in clinical characteristics across samples precludes definitive conclusions regarding a direct association between TRPV2 expression and FGR. These findings highlight a possibly conserved role of TRPV2 in the placenta and fetal growth outcomes in humans. Table 2 Demographic characteristics of human study participants with Appropriately for Gestational Age (AGA) and Fetal Growth Restriction (FGR) pregnancies Term born AGA N = 15 Term born early-onset FGR N = 4 Preterm born early-onset FGR N = 18 p-value Maternal age 36(31–39) 29(24–35) 35.5(31–38,5) 0.3088 1 Maternal BMI (kg/m 2 ) 22.25 (20.2–28.8) 23 (21.5–32) 24(23–27) 0.2727 2 Mode of birth (% vaginal births) 4(27%) 2(50%) 1(6%) 0.0613 3 Fetal sex (% of males) 10 (66%) 1(25%) 7(41%) 0.2544 3 Continuous variables are given as median (interquartile range) and compared by one-way ANOVA 1 with Dunnett’s multiple comparison post-hoc or by Kruskal–Wallis test 2 with Dunn’s multiple comparison post-hoc. Categorical variables are given as number (% of n) and compared by chi-squared test 3 . Demographic characteristics of human study participants with Appropriately for Gestational Age (AGA) and Fetal Growth Restriction (FGR) pregnancies Continuous variables are given as median (interquartile range) and compared by one-way ANOVA 1 with Dunnett’s multiple comparison post-hoc or by Kruskal–Wallis test 2 with Dunn’s multiple comparison post-hoc. Categorical variables are given as number (% of n) and compared by chi-squared test 3 .

Constitutive Trpv2 -deficient mice display reduced birth weight and perinatal lethality, phenotypes that have been reported across different models 14 , 28 but remain mechanistically unexplained. Even more, the severity of this phenotype has prompted the use of different genetic backgrounds and breeding strategies to mitigate lethality, yet its underlying cause has remained unclear. This variability highlights the strong background dependency of Trpv2 knockout phenotypes, a factor that has complicated interpretation of its developmental role in previous studies. TRPV2 has established roles in multiple tissues, including the brain 10 , 11 , the immune system 12 and the heart 13 . However, these functions do not readily account for the early developmental timing and nature of the phenotype observed in global Trpv2 -deficient animals. Tissue-specific deletion of Trpv2 in cardiomyocytes demonstrates an essential role in adult cardiac structure and function 13 , but does not address its contribution during embryonic development. In this study, the growth restriction and lethality phenotype was recapitulated using a Sox2 -Cre–mediated deletion strategy affecting both embryonic and extraembryonic compartments on a C57BL/6 J background. However, by functionally separating these contributions, TRPV2 activity in the trophoblast compartment, rather than in the embryo proper, emerged as essential for normal development, establishing a primary placental origin for the observed phenotype. This reinterpretation suggests that previously reported phenotypes in global Trpv2 -deficient models may, at least in part, arise from compromised placental function with secondary consequences for embryonic development. As such, recent studies highlight the placenta as a significant contributor of developmental heart disorders 29 . At the cellular level, Trpv2 expression was enriched within labyrinth trophoblast populations and spaned early differentiation states that give rise to SynTII and S-TGC lineages, while being largely excluded from the SynTI trajectory. Spatial transcriptomic analyses further indicated that Trpv2 was expressed not only in mature SynTII cells but also in precursor populations, suggesting a role during lineage progression rather than exclusively in terminal differentiation. This positioning is consistent with emerging models of trophoblast development, in which lineage specification is driven by progressive activation of transcriptional programs along differentiation trajectories rather than discrete binary fate decisions 30 . Loss of Trpv2 resulted in a striking disruption of labyrinth architecture, characterized by focal regions lacking both fetal capillaries and SynTII cells. These regions retained trophoblast identity but failed to establish a functional exchange surface, highlighting the central role of SynTII cells in orchestrating placental branching and vascularization. Notably, SynTI marker-positive cells were still present in these areas but failed to acquire the characteristic elongated and flattened syncytial morphology, consistent with previous reports  stating that SynTI fusion depends on interaction with SynTII cells 23 . Together, these observations suggest that impaired SynTII formation represents a primary defect, with secondary consequences for SynTI maturation and vascular organization. The absence of this phenotype in epiblast-specific mutants argues against a primary vascular defect and instead points towards defective SynTII formation as a primary event. Moreover, consistent with this interpretation, conditional deletion of Trpv2 specifically in Gcm1 ⁺ SynTII cells recapitulated key aspects of the global knockout phenotype, including reduced fetal growth and the presence of abnormal labyrinth regions lacking both SynTII cells and fetal capillaries. However, this phenotype was less severe and did not result in complete embryonic lethality, indicating that Trpv2 function in Gcm1 +  SynTII cells is necessary but not solely sufficient to account for the full spectrum of defects observed in constitutive knockouts. This difference in phenotypic severity likely reflects either strain‑dependent differences in Trpv2 knockout severity or the broader expression of Trpv2 across trophoblast differentiation trajectories, including precursor populations and additional lineages such as S-TGCs, suggesting that TRPV2 contributes to placental development at multiple stages of trophoblast differentiation. In addition, incomplete recombination may have attenuated the phenotype observed in Trpv2 SynTII-/- placentas, as Cre-mediated recombination efficiency can be influenced by the level and duration of Cre expression, as well as by accessibility of the Trpv2 locus. At the molecular level, both in vivo and in vitroanalyses converge on a consistent reduction in SynTII-associated transcriptional programs upon Trpv2 loss. The transcription factor Gcm1 , a central regulator of SynTII differentiation, was reproducibly downregulated across multiple experimental modalities, supporting the idea that TRPV2 acts upstream of, or in coordination with, the regulatory network controlling SynTII specification. Notably, individual marker analyses in vitro revealed variability and modest effect sizes under standard differentiation conditions, indicating that Trpv2 loss does not completely block differentiation. Instead, bulk RNA sequencing and gene set-based analyses demonstrate that TRPV2 is required for the efficient and coordinated activation of a broader SynTII transcriptional program, particularly under conditions that actively promote this lineage, such as Wnt pathway activation by CHIR99021. This distinction between partial marker-level effects and robust program-level alterations highlights the importance of considering lineage specification as a coordinated transcriptional process rather than relying on individual markers alone. Moreover, interpretation of trophoblast differentiation studies performed in vitro also requires caution. Under standard differentiation conditions, mTSC cultures generate heterogeneous trophoblast populations, making robust assessment of lineage-specific processes such as syncytial fusion challenging. In addition, the coordinated interactions between SynTI and SynTII cells that are required for normal labyrinth formation in vivo are unlikely to be fully recapitulated in current in vitro systems. Future studies using more refined lineage-directed differentiation approaches may help distinguish direct effects of TRPV2 on trophoblast fate specification from secondary effects resulting from disrupted trophoblast cell-cell interactions. This may be particularly important given the regional nature of the placental defects observed in vivo, suggesting that local cellular interactions contribute to the phenotypic outcome following Trpv2 loss. Our findings also highlight both similarities and differences between mouse and human placentation. In human placenta, TRPV2 expression is predominantly observed within the villous core, but is also detected in subsets of cells with syncytium-like morphology lacking CGA expression, suggesting that TRPV2 may be expressed prior to terminal ST differentiation. This is consistent with our findings in mouse and in vitro systems and supports the idea that TRPV2 marks early stages of syncytial lineage development rather than exclusively mature ST cells. Furthermore, in differentiating ST, a significant correlation was observed between the expression of TRPV2 and increased basal calcium levels, which has been shown to be a prerequisite for fusion 31 – 33 . In line with this, TRPV2 expression was specifically present in fusing cells in scRNA-seq datasets of both in vitro ST cultures and in vivo first term placental samples. Finally, decreased TRPV2 expression was observed in early-onset fetal growth restriction, particularly in preterm cases, although interpretation of these data is limited by gestational age differences and clinical heterogeneity. These findings are therefore considered exploratory and do not allow for making causal conclusions. Together, this study identifies TRPV2 as a key regulator of trophoblast differentiation and placental morphogenesis. By promoting the activation of SynTII transcriptional programs across trophoblast differentiation stages, TRPV2 supports the establishment of a functional maternal–fetal exchange interface. Global Trpv2 -/- mice suffer from FGR and lethality, which has been independently observed before in different Trpv2 -/- mouse strains 14 , 28 and more recently in rats 34 , but has never been explained so far. This study identifies an important role for TRPV2 during mouse placental development and provides evidence for functional TRPV2 expression in human trophoblast cells. Notably, several TRP channels, including TRPM6 and TRPM7, have been implicated in early development and trophoblast biology 35 – 37 , but their roles are either masked by early embryonic lethality or primarily affect placental function rather than lineage formation. In contrast, TRPV2 appears to regulate SynTII development itself, suggesting a distinct role in placental morphogenesis. The selective vulnerability of the SynTII lineage to Trpv2 deletion remains intriguing. Although TRPV2 expression is highly enriched in SynTII cells, expression is also detected in other trophoblast populations, including primary and sinusoidal trophoblast giant cells, which appear comparatively less affected by Trpv2 loss. Furthermore, processes commonly associated with TRPV2 activity, such as calcium-dependent signalling, cytoskeletal remodelling and membrane dynamics, are not known to be uniquely required for SynTII differentiation. Interestingly, selective defects affecting discrete trophoblast lineages have also been reported following perturbation of other broadly expressed regulators 38 . Whether SynTII cells possess a particular dependence on TRPV2-mediated calcium signalling, or whether TRPV2 interfaces with lineage-specific regulatory networks centered on GCM1, remains to be established.

Size at birth is decisive in determining life expectancy, as it affects mortality and morbidity in adults 1 , 2 . Up to 10% of newborns suffer from fetal growth restriction (FGR), a disorder that is defined as a pathological smallness, often associated with an estimated fetal weight below the 10 th percentile for gestational age and signs of fetal compromise 3 . While abnormal placental growth or function is a major cause of FGR 4 , 5 , the molecular players required for normal placentation are poorly understood. In both human and rodent placentas, establishment of an efficient maternal-fetal exchange interface depends on the coordinated development of trophoblast and fetal vascular compartments. This process involves progressive invagination of fetal blood vessels into the chorionic trophoblast layer, generating the highly branched exchange structures that characterize the mature placenta. The foundation of this structure is established by syncytiotrophoblast II (SynTII) cells in mice and primitive syncytiotrophoblast (ST) cells in humans, both of which are multinucleated trophoblast populations 6 , 7 . Reciprocal interactions between trophoblast cells and the developing fetal vasculature are essential for proper villous branching, trophoblast differentiation and vascularization, yet the signalling pathways regulating this intercellular crosstalk remain incompletely understood. Calcium ions are important secondary messengers and are well suited to decode and integrate informational cues from both chemical and mechanical cellular microenvironments. Transient Receptor Potential (TRP) channels are a superfamily of cation channels that mostly reside in the plasma membrane and function as sensors that transduce environmental signals into a cellular response 8 . TRPV2 is a calcium-permeable channel that is ubiquitously expressed in both neuronal and non-neuronal tissues 9 . TRPV2 function is reported to contribute to neuronal outgrowth 10 , 11 , immune response 12 and cardiac function 13 . Mice lacking TRPV2 are born small for gestational age associated with an increased perinatal lethality rate, although neonatal respiratory function, suckling and postnatal growth were not affected 14 . Yet, none of the previously allocated functions of TRPV2 can explain how its loss results in adverse fetal outcomes. Recently, we identified TRPV2 as the most prominent TRP channel in the mouse placenta, with enrichment in the transport region of the placenta, also known as the labyrinth zone, and sporadic expression in several trophoblast giant cells (TGC) and in the endocrine junctional zone 15 . Nevertheless, its functional role in early placental development has not been addressed. In this study, TRPV2 is identified as a previously unrecognized regulator of placental development, acting within trophoblast lineages to support formation of the maternal-fetal exchange interface and thereby contributing to fetal growth and viability. Loss of Trpv2 in the placenta, but not in the embryo, results in severe defects in placental morphology. Trpv2 is primarily expressed in SynTII cells and their precursors, and CRISPR/Cas9-mediated deletion of Trpv2 in mouse trophoblast stem cells (mTSC) impairs SynTII differentiation. In human systems, TRPV2 is functionally expressed in trophoblast cells and enriched in trophoblast populations undergoing syncytial differentiation.

Source Data Source Data