P–375 Intralipid supplementation in women with unexplained recurrent implantation failure and elevated uterine natural killer cell levels - A randomized placebo controlled trial

Study question Does intralipid supplementation in women with unexplained recurrent implantation failure (RIF) with elevated uterine natural killer cell (uNK) levels improve pregnancy outcomes during IVF? Summary answer Intralipid supplementation appears to improve clinical pregnancy rate in women with unexplained RIF with elevated uNK cell levels. What is known already The increased numbers of uNK cells in peri-implantation endometrium have been reported in women with recurrent miscarriage (RM) and RIF after IVF. However, reports are contradictory when it comes to correlation of increased numbers of uNK cells with pregnancy outcome. Current opinion suggests there is a potential for intralipid therapy in improving reproductive outcome, although data on live birth rate is very limited. No studies have assessed the effect of intralipid on IVF outcomes in RIF women based on elevated uNK cells. Identified studies have all used pNK cell testing as preferred diagnostic tool for analysis of NK cell levels. Study design, size, duration A randomized placebo controlled trial was conducted at Division of Reproductive Medicine at tertiary care institute. Thirty women with RIF and fifty fertile controls with age <35 years having regular menstrual cycles and no hormonal treatment in last 3 months were enrolled in the study from January 2019 to December 2020 for uNK cell testing. Randomization was done using random numbers and sealed envelopes. Only the subjects were masked and allocation concealment was done. Participants/materials, setting, methods Subjects included RIF 20–35 years, normal ovarian reserve, unexplained and tubal factors, normal karyotype and normal uterine cavity. Cut off for uNK cells was derived from fertile controls by immunohistochemical staining of CD56+ cells from midluteal endometrial biopsy sample. Subjects with elevated uNK cell levels were randomized during IVF to group A (Intralipid) or group B ( saline ). The infusion was repeated within one week of positive pregnancy test and then every 2 weeks. Main results and the role of chance The mean age and BMI were comparable between fertile control and study group(29.45±3.3 vs 31.17±3.3 years, 22.97±1.89 vs 23.21±2.2 kg/m2 ;p>0.05). The median uNK cell levels was 7%(used as cut off) in fertile controls and 13.5% in RIF. 18 women (60%, 18/30) with RIF who had elevated uNK cell level (>7%) were randomized. Four women were lost to follow up. The median age, BMI, number of previous failed cycles and duration of infertility were comparable between Group A(n = 7) and Group B(n = 7){30(IQR:27–31) vs 33(IQR:30–34)years, 22.7(IQR:21.08–24.4) vs 22.6(IQR:21.37–24.2)kg/m2, 2(IQR:2–3) vs 2(IQR:2–3), 8(IQR:7–8) vs 8(IQR:7–10)years}. The median FSH, AMH and AFC were 5.86(IQR:5.13–7.67)mIU/l, 2.4(IQR:2.16–6.12)ng/ml, 10(IQR:8–12) in Group A which were comparable with Group B {6.2(IQR:4.78–6.5)mIU/l, 4.8(IQR:2.67–6.25)ng/ml, 12(IQR:12–16) }. All patients underwent antagonist protocol. The clinical pregnancy rate was 57.14%(4/7) in group A which was significantly higher as compared to 28.6%(2/7) in group B(p < 0.05). None of the patients reported any side effects due to intralipid. Limitations, reasons for caution The limitation of present study is its small sample size. However, the study is currently recruiting more RIF patients, and these are the interim results of the same. More RCTs with larger sample size are required to assess the efficacy of intralipid in this specific subset of population. Wider implications of the findings: The present study suggests the beneficial effect of intralipid in women with unexplained RIF with elevated uNK cell levels in increasing the chemical and clinical pregnancy rate. However, ongoing pregnancy rate and live birth rate should be investigated further in this subset of population. Trial registration number CTRI/2019/01/017213

Continuous human uterine NK cell differentiation in response to endometrial regeneration and pregnancy

Immune cell differentiation is critical for adequate tissue-specific immune responses to occur. Here, we studied differentiation of human uterine natural killer cells (uNK cells). These cells reside in a tissue undergoing constant regeneration and represent the major leukocyte population at the maternal-fetal interface. However, their physiological response during the menstrual cycle and in pregnancy remains elusive. By surface proteome and transcriptome analysis as well as using humanized mice, we identify a differentiation pathway of uNK cells in vitro and in vivo with sequential acquisition of killer cell immunoglobulin-like receptors and CD39. uNK cell differentiation occurred continuously in response to the endometrial regeneration and was driven by interleukin-15. Differentiated uNK cells displayed reduced proliferative capacity and immunomodulatory function including enhanced angiogenic capacity. By studying human uterus transplantation and monozygotic twins, we found that the uNK cell niche could be replenished from circulation and that it was under genetic control. Together, our study uncovers a continuous differentiation pathway of human NK cells in the uterus that is coupled to profound functional changes in response to local tissue regeneration and pregnancy.

Involvement of uterine natural killer cells in the natural selection of human embryos at implantation

Decidualizing endometrial stromal cells (EnSC) critically determine the maternal response to an implanting conceptus, triggering either menstruation-like disposal of low-fitness embryos or creating an environment that promotes further development. However, the mechanism that couples maternal recognition of low-quality embryos to tissue breakdown remains poorly understood. Recently, we demonstrated that successful transition of the cycling endometrium to a pregnancy state requires selective elimination of pro-inflammatory senescent decidual cells by activated uterine natural killer (uNK) cells. Here we report that uNK cells express CD44, the canonical hyaluronan (HA) receptor, and demonstrate that high-molecular weight HA (HMWHA) inhibits uNK cell-mediated killing of senescent decidual cells. By contrast, low-molecular weight HA (LMWHA) did not attenuate uNK cell activity in co-culture experiments. Killing of senescent decidual cells by uNK cells was also inhibited upon exposure to medium conditioned by IVF embryos that failed to implant, but not successful embryos. Embryo-mediated inhibition of uNK cell activity was reversed by recombinant hyaluronidase 2 (HYAL2), which hydrolyses HMWHA. We further report a correlation between the levels of HYAL2 secretion by human blastocysts, morphological scores, and implantation potential. Taken together, the data suggest a pivotal role for uNK cells in embryo biosensing and endometrial fate decisions at implantation.

4307 Role of Pre-pregnancy Uterine Natural Killer Cells in Human Embryo Implantation

OBJECTIVES/GOALS: Human placentation requires complex coordination between maternal and fetal cell types but remains incompletely understood. We hypothesize that uterine natural killer (uNK) cells, an immune cell type that increases in abundance during the implantation window, is essential for appropriate implantation and placentation. METHODS/STUDY POPULATION: We plan to examine stromal cell (SC) decidualization, spiral artery remodeling, and EVT invasion, processes vital for early pregnancy establishment, in the presence or absence of secretory phase uNK cells. Fetal extravillous trophoblasts (EVTs) will be isolated from first trimester pregnancy tissue; maternal SCs, endothelial cells (ECs) and uNK cells will be obtained from secretory phase uterine tissue. SCs will be placed in monoculture and coculture with uNK cells and prolactin will be measured to evaluate decidualization. To study EVT invasion, we will utilize our novel “implantation-on-a-chip” device to determine how addition of uNK cells affects EVT migration through a collagen-matrigel matrix. In this system, we will also examine spiral artery remodeling with or without uNK cells via TUNEL staining. RESULTS/ANTICIPATED RESULTS: We anticipate that uNK cell addition to SCs will lead to a significant increase in SC prolactin levels, suggesting a role of uNK cells in endometrial decidualization. In vitro, we expect the addition of uNK cells will increase EC apoptosis and promote EVT invasion. DISCUSSION/SIGNIFICANCE OF IMPACT: Although decidual NK cells are known to participate in placentation, the role of pre-pregnancy uNK cells is unknown. uNK cell involvement in processes important for the earliest stages of pregnancy would provide a potential marker for abnormal placentation and offer avenues for intervention to decrease placentation associated perinatal morbidity.

Profiling the expression and function of oestrogen receptor isoform ER46 in human endometrial tissues and uterine natural killer cells

STUDY QUESTION Does the oestrogen receptor isoform, ER46, contribute to regulation of endometrial function? SUMMARY ANSWER ER46 is expressed in endometrial tissues, is the predominant ER isoform in first trimester decidua and is localised to the cell membrane of uterine natural killer (uNK) cells where activation of ER46 increases cell motility. WHAT IS KNOWN ALREADY Oestrogens acting via their cognate receptors are essential regulators of endometrial function and play key roles in establishment of pregnancy. ER46 is a 46-kDa truncated isoform of full length ERα (ER66, encoded by ESR1) that contains both ligand- and DNA-binding domains. Expression of ER46 in the human endometrium has not been investigated previously. ER46 is located at the cell membrane of peripheral blood leukocytes and mediates rapid responses to oestrogens. uNK cells are a phenotypically distinct (CD56brightCD16−) population of tissue-resident immune cells that regulate vascular remodelling within the endometrium and decidua. We have shown that oestrogens stimulate rapid increases in uNK cell motility. Previous characterisation of uNK cells suggests they are ER66-negative, but expression of ER46 has not been characterised. We hypothesise that uNK cells express ER46 and that rapid responses to oestrogens are mediated via this receptor. STUDY DESIGN, SIZE, DURATION This laboratory-based study used primary human endometrial (n = 24) and decidual tissue biopsies (n = 30) as well as uNK cells which were freshly isolated from first trimester human decidua (n = 18). PARTICIPANTS/MATERIALS, SETTING, METHODS Primary human endometrial and first trimester decidual tissue biopsies were collected using methods approved by the local institutional ethics committee (LREC/05/51104/12 and LREC/10/51402/59). The expression of ERs (ER66, ER46 and ERβ) was assessed by quantitative PCR, western blot and immunohistochemistry. uNK cells were isolated from first-trimester human decidua by magnetic bead sorting. Cell motility of uNK cells was measured by live cell imaging: cells were treated with 17β-oestradiol conjugated to bovine serum albumin (E2-BSA, 10 nM equivalent), the ERβ-selective agonist 2,3-bis(4-hydroxyphenyl)-propionitrile (DPN; 10 nM) or dimethylsulphoxide vehicle control. MAIN RESULTS AND THE ROLE OF CHANCE ER46 was detected in proliferative and secretory phase tissues by western blot and was the predominant ER isoform in first-trimester decidua samples. Immunohistochemistry revealed that ER46 was co-localised with ER66 in cell nuclei during the proliferative phase but detected in both the cytoplasm and cell membrane of stromal cells in the secretory phase and in decidua. Triple immunofluorescence staining of decidua tissues identified expression of ER46 in the cell membrane of CD56-positive uNK cells which were otherwise ER66-negative. Profiling of isolated uNK cells confirmed expression of ER46 by quantitative PCR and western blot and localised ER46 protein to the cell membrane by immunocytochemistry. Functional analysis of isolated uNK cells using live cell imaging demonstrated that activation of ER46 with E2-BSA significantly increased uNK cell motility. LARGE SCALE DATA N/A. LIMITATIONS, REASONS FOR CAUTION Expression pattern in endometrial tissue was only determined using samples from proliferative and secretory phases. Assessment of first trimester decidua samples was from a range of gestational ages, which may have precluded insights into gestation-specific changes in these tissues. Our results are based on in vitro responses of primary human cells and we cannot be certain that similar mechanisms occur in situ. WIDER IMPLICATIONS OF THE FINDINGS E2 is an essential regulator of reproductive competence. This study provides the first evidence for expression of ER46 in the human endometrium and decidua of early pregnancy. We describe a mechanism for regulating the function of human uNK cells via expression of ER46 and demonstrate that selective targeting with E2-BSA regulates uNK cell motility. These novel findings identify a role for ER46 in the human endometrium and provide unique insight into the importance of membrane-initiated signalling in modulating the impact of E2 on uNK cell function in women. Given the importance of uNK cells to regulating vascular remodelling in early pregnancy and the potential for selective targeting of ER46, this may be an attractive future therapeutic target in the treatment of reproductive disorders. STUDY FUNDING/COMPETING INTEREST(S) These studies were supported by Medical Research Council (MRC) Programme Grants G1100356/1 and MR/N024524/1 to PTKS. H.O.D.C. was supported by MRC grant G1002033. The authors declare no competing interests related to the published work.

Decidual NK cell-derived conditioned medium from miscarriages affects endometrial stromal cell decidualisation: endocannabinoid anandamide and tumour necrosis factor-α crosstalk

STUDY QUESTION What are the effects of endocannabinoid anandamide (AEA) in uterine natural killer (unK) cells from miscarriage decidua, regarding their cytokine profile and endometrial stromal cell (ESC) crosstalk? SUMMARY ANSWER uNK-conditioned media from miscarriage samples present high TNF-α levels which inhibit ESC decidualisation. WHAT IS KNOWN ALREADY AEA plasma levels are higher in women who have suffered a miscarriage. Moreover, AEA inhibits ESC proliferation and differentiation, although the levels and impact on the uNK cell cytokine profile at the feto-maternal interface remain elusive. STUDY DESIGN, SIZE, DURATION This laboratory-based study used human primary uNK cells which were isolated from first-trimester decidua (gestational age, 5–12 weeks) derived from 8 women with elective pregnancy termination and 18 women who suffered a miscarriage. PARTICIPANTS/MATERIALS, SETTING, METHODS The first-trimester placental tissues were assayed for AEA levels by UPLC-MS/MS and respective enzymatic profile by western blot. The uNK cells were isolated and maintained in culture. The expression of angiogenic markers in uNK cells was examined by quantitative PCR (qPCR). The uNK-conditioned medium was analysed for IFN-γ, TNF-α and IL-10 production by enzyme-linked immunosorbent assay, and the impact on ESC differentiation was assessed by measuring decidual markers Prl, Igfbp-1 and Fox01 mRNA expression using qPCR. MAIN RESULTS AND THE ROLE OF CHANCE AEA levels were higher in miscarriage decidua compared with decidua from elective terminations. The uNK cell-conditioned medium from the miscarriage samples exhibited high TNF-α levels and interfered with the decidualisation of ESCs. Exacerbated inflammation and elevated TNF-α levels at the feto-maternal interface may trigger AEA signalling pathways that, in turn, may impact decidualisation and the angiogenic ability of uNK cells. LARGE-SCALE DATA N/A. LIMITATIONS, REASONS FOR CAUTION Primary uNK cell responses are based on a simple in vitro model. Thus, in complex microenvironments, such as the feto-maternal interface, the mechanisms may not be exactly the same. Also, the inflammatory events of miscarriage that, in this study, have happened prior to processing of the samples may cause different responses to that observed. In addition, the magnitude of the inflammatory response, required to trigger the AEA pathways that impact decidualisation and the uNK angiogenic ability in vivo, is still unclear. WIDER IMPLICATIONS OF THE FINDINGS The endocannabinoid AEA is a modulator of reproductive competence. AEA not only may contribute to neuroendocrine homeostasis but also can take part in uterine changes occurring during early pregnancy. STUDY FUNDING/COMPETING INTEREST(S) The work was supported by UID/MULTI/04378/2019 with funding from Fundação para a Ciência e a Tecnologia (FCT)/MCTES through national funds and PORTUGAL 2020 Partnership Agreement, NORTE-01-0145-FEDER-000024. S.C. Cunha acknowledges FCT for the IF/01616/2015 contract. There are no conflicts of interest.

Endometrial uNK cell counts do not predict successful implantation in an IVF population

Study question Are uterine natural killer (uNK) cell numbers and their distribution relative to endometrial arterioles altered in women with recurrent implantation failure (RIF) compared to women with embryo implantation success (IS)? Summary answer uNK cell numbers and their distribution relative to endometrial arterioles are not significantly different in women with RIF compared to women in whom embryo implantation occurs successfully following IVF. What is already known uNK cells are regulators of decidual angiogenesis and spiral arteriole remodelling during early pregnancy. Although some studies have shown that uNK cell numbers may be altered in women with RIF, the methods used to measure uNK cell numbers have proven inconsistent, making reproduction of these results difficult. It is unclear, therefore, whether the results reported so far are reproducible. Moreover, it is not known how uNK cell numbers may impact IVF outcomes. Despite the lack of conclusive evidence, uNK cell numbers are often evaluated as a prognostic criterion in women undergoing assisted reproductive procedures. Study design, size, duration Endometrial pipelle biopsies were collected 6–8 days post-LH surge in natural cycles from women with RIF (n = 14), women with IS (n = 11) and women with potential RIF at the time of the study (PRIF; n = 9) from 2013 to 2015. Participants/materials, setting, methods uNK cells (i.e. CD56+ and/or CD16+ phenotypes) and their distribution relative to endometrial arterioles were investigated by standard immunohistochemistry protocols and quantified using Aperio ScanScopeXT images digitized by ImageJ and deconvoluted into binary images for single cell quantification using a Gaussian Blur and Yen algorithm. Main results and the role of chance There was no significant difference in the cell density of CD56+ or CD16+ uNK cells in women with RIF compared to women with IS or PRIF. There was a higher proportion of uNK cells in the distal regions compared to the regions closest to the arterioles in all patient groups. Further, we identified a significant reduction in uNK cell density in women who had a previous pregnancy compared to those who had not, regardless of their current implantation status. Large scale data Not applicable. Limitations, reasons for caution Spiral arterioles could not always be accurately identified by digital image analysis; therefore, all endometrial arterioles were selected and analysed. Patient numbers for the study were low. However, as the clinical phenotypes of each patient were well defined, and endometrial dating was accurately determined by three independent pathologists, differences between patient groups with respect to the uNK numbers and distribution should have been measurable if uNK cell counts were to be useful as a prognostic marker of RIF. Wider implications of the findings Our findings demonstrate that CD56+ and CD16+ uNK cell numbers are not significantly different in women with RIF in a typical cohort of women undergoing IVF. Further, prior pregnancy was associated with a significantly reduced number of uNK cells in both the RIF and IS patient groups, suggestive of a long-term pregnancy induced suppression of uNK cells. Combined, these findings do not support the clinical value of using uNK cell numbers as a prognostic indicator of implantation success with IVF treatment. Study funding/competing interest(s) Funding for this work was provided by Royal Women’s Hospital Foundation. P.P. was supported by an NHMRC Early Career Fellowship [TF 11/14] and W.T.T. was supported by an NHMRC Postgraduate Scholarship [1055814]. The authors do not have any competing interests with this study.

Profiling the expression and function of ER46 in human endometrial tissues and uterine NK cells

Study questionDoes the oestrogen receptor isoform, ER46, contribute to regulation of endometrial function?Summary answerER46 is expressed in endometrial tissues during the proliferative and secretory phases and is the predominant ERα isoform in first trimester decidua. ER46 is abundantly expressed in uterine NK (uNK) cells and localised to the cell membrane. Activation of ER46 regulates the function of human uNK cells by increasing cell motility.What is known alreadyOestrogens acting via their cognate receptors are essential regulators of endometrial function and play key roles in establishment of pregnancy. ER46 is a 46kDa truncated isoform of full length ERα (ER66, encoded by ESR1) that contains both ligand and DNA binding domains. Expression of ER46 in human endometrium has not been investigated previously. ER46 is located at the cell membrane of peripheral blood leukocytes and mediates rapid responses to oestrogens. UNK cells are a phenotypically distinct (CD56brightCD16-) population of tissue-resident immune cells that regulate vascular remodelling within the endometrium and decidua. We have shown that oestrogens stimulate rapid increases in uNK cell motility. Previous characterisation of uNK cells suggests they are ER66-negative but expression of ER46 has not been characterised. We hypothesise that uNK cells express ER46 and that rapid responses to oestrogens are mediated via this receptor.Study design, size, durationThis laboratory-based study used primary human endometrial (n=24) and decidual tissue biopsies (n=30) as well as uNK cells which were freshly isolated from first trimester human decidua (n=18).Participants/materials, setting, methodsPrimary human endometrial and first trimester decidual tissue biopsies were collected using methods approved by the local institutional ethics committee (LREC/05/51104/12 and LREC/10/51402/59). The expression of oestrogen receptors (ER66, ER46 and ERβ) was assessed by qPCR, western blot and immunohistochemistry. Uterine Natural Killer (uNK) cells were isolated from first trimester human decidua by magnetic bead sorting. Cell motility of uNK cells was measured by live cell imaging: cells were treated with oestradiol (E2)-BSA (10nM equivalent), the ERβ-selective agonist 2,3-bis (4-hydroxyphenyl)-propionitrile (DPN; 10nM) or vehicle control (DMSO).Main results and the role of chanceER46 was detected in proliferative and secretory phase tissues and was the predominant ERα isoform in first trimester decidua samples. Immunohistochemistry revealed ER46 was co-localised with ER66 in cell nuclei during the proliferative phase but detected in both the cytoplasm and cell membrane of stromal cells in the secretory phase and in decidua. Triple immunofluorescence staining of decidua tissues identified expression of ER46 in the cell membrane of CD56-positive uNK cells which were otherwise ER66-negative. Profiling of isolated uNK cells confirmed expression ER46 and localised ER46 protein to the cell membrane. Functional analysis of isolated uNK cells using live cell imaging demonstrated that activation of ER46 with E2-BSA significantly increased uNK cell motility.Limitations, reasons for cautionExpression patterns in endometrial tissue was only determined using samples from proliferative and secretory phases. Assessment of first trimester decidua samples was from a range of gestational ages which may have precluded insights into gestation specific changes in these tissues. Our results are based on in vitro responses of primary human cells and we cannot be certain that similar mechanisms occur in situ.Wider implications of the findingsE2 is an essential regulator of reproductive competence. This study provides the first evidence for expression of ER46 in human endometrium and decidua of early pregnancy. We describe a mechanism for regulating the function of human uNK cells via expression of ER46 and demonstrate that selective targeting with E2-BSA regulates uNK cell motility. These novel findings identify a role for ER46 in human endometrium and provide unique insight into the importance of membrane-initiated signalling in modulating the impact of E2 on uNK cell function in women.Study funding/competing interest(s)These studies were supported by MRC Programme Grants G1100356/1 and MR/N024524/1 to PTKS. HODC was supported by MRC grant G1002033.

Uterine spiral artery muscle dedifferentiation

STUDY QUESTION Is vascular smooth muscle cell (VSMC) dedifferentiation a feature of uterine spiral artery (SpA) remodelling in early human pregnancy? SUMMARY ANSWER Remodelling of human uterine SpAs is associated with dedifferentiation of VSMCs and can be induced in vitro by uterine natural killer (uNK) cells and extravillous trophoblast cells (EVTs). WHAT IS KNOWN ALREADY Uterine SpAs undergo profound morphological changes in normal pregnancy with replacement of the musculoelastic arterial wall structure by fibrinoid containing EVTs. The fate of VSMCs in SpA remodelling is unknown; in guinea pig uterine artery VSMCs dedifferentiate, remain in the vessel wall and differentiate after parturition to restore the arterial wall. There is increasing evidence that uNK cells play a role in SpA remodelling. We hypothesized that SpA remodelling in human pregnancy is associated with VSMC dedifferentiation, initiated by uNK cell-derived growth factors. STUDY DESIGN, SIZE, DURATION Formalin fixed, paraffin embedded placental bed biopsies were immunostained for angiogenic growth factor (AGF) receptors and markers of VSMC differentiation. An in vitro model of SpA remodelling using chorionic plate arteries (CPAs) was used to test the effect of different cell types and AGFs on VSMC differentiation. PARTICIPANTS/MATERIALS, SETTING, METHODS Placental bed biopsies were immunostained for vascular endothelial growth factor receptors 1-3 (VEGF-R1, VEGF-R2, VEGF-R3), transforming growth factor beta 1 receptors I and II (TGF-βRI, TGF-βRII), interferon gamma receptors 1 and 2 (IFN-γR1, IFN-γR2), Tie2, α-smooth muscle actin (α-SMA), H-caldesmon (H-Cal), myosin heavy chain (MyHC), osteopontin and smoothelin. Staining intensity was assessed using a modified quickscore. Expression by VSMCs of the AGF receptors was confirmed by laser capture microdissection and real-time RT-PCR of non-remodelled SpAs, after laser removal of the endothelium. As an in vitro model, VSMC differentiation was assessed in CPAs by immunohistochemistry after culture in uNK cell-conditioned medium (CM), EVT-CM, uNK cell/EVT co-culture CM, Ang-1, Ang-2, IFN-γ, VEGF-A and VEGF-C, and after blocking of both Ang-1 and Ang-2 in uNK-CM. MAIN RESULTS AND THE ROLE OF CHANCE SpA VSMC expression of Tie-2 (P = 0.0007), VEGF-R2 (P = 0.005) and osteopontin (P = 0.0001) increased in partially remodelled SpAs compared with non-remodelled SpAs, while expression of contractile VSMC markers was reduced (α-SMA P < 0.0001, H-Cal P = 0.03, MyHC P = 0.03, smoothelin P = 0.0001). In the in vitro CPA model, supernatants from purified uNK cell (H-Cal P < 0.0001, MyHC P = 0.03, α-SMA P = 0.02, osteopontin P = 0.03), EVT (H-Cal P = 0.0006, MyHC P = 0.02, osteopontin P = 0.01) and uNK cell/EVT co-cultures (H-Cal P = 0.001, MyHC P = 0.05, osteopontin P = 0.02) at 12–14 weeks, but not 8–10 weeks, gestational age induced reduced expression of contractile VSMC markers and increased osteopontin expression. Addition of exogenous (10 ng/ml) Ang-1 (P = 0.006) or Ang-2 (P = 0.009) also reduced H-Cal expression in the CPA model. Inhibition of Ang-1 (P = 0.0004) or Ang-2 (P = 0.004) in uNK cell supernatants blocked the ability of uNK cell supernatants to reduce H-Cal expression. LIMITATIONS, REASONS FOR CAUTION This is an in vitro study and the role of uNK cells, Ang-1 and Ang-2 in SpA remodelling in vivo has not yet been shown. WIDER IMPLICATIONS OF THE FINDINGS VSMC dedifferentiation is a feature of early SpA remodelling and uNK cells and EVT play key roles in this process by secretion of Ang-1 and Ang-2. This is one of the first studies to suggest a direct role for Ang-1 and Ang-2 in VSMC biology. STUDY FUNDING/COMPETING INTEREST(S) This work was supported by a grant from British Biotechnology and Biosciences Research Council (BB/E016790/1). The authors have no competing interests to declare.