...reconstruction of follicular remodeling in the human adult...
ARTICLESingle-cell reconstruction of follicular remodeling inthe human adult ovaryX. Fan1,6, M. Bialecka1,6, I. Moustakas 1,2, E. Lam1, V. Torrens-Juaneda 1, N.V. Borggreven3, L. Trouw3,L.A. Louwe4, G.S.K. Pilgram4, H. Mei2, L. van der Westerlaken 4 S.M. Chuva de Sousa Lopes 1,5The ovary is perhaps the most dynamic organ in the human body, only rivaled by the uterus.The molecular mechanisms that regulate follicular growth and regression, ensuring ovariantissue homeostasis, remain elusive. We have performed single-cell RNA-sequencing usinghuman adult ovaries to provide a map of the molecular signature of growing and regressingfollicular populations. We have identified different types of granulosa and theca cells anddetected local production of components of the complement system by (atretic) theca cellsand stromal cells. We also have detected a mixture of adaptive and innate immune cells, aswell as several types of endothelial and smooth muscle cells to aid the remodeling process.Our results highlight the relevance of mapping whole adult organs at the single-cell level andreflect ongoing efforts to map the human body. The association between complement systemand follicular remodeling may provide key insights in reproductive biology and (in)fertility.https://doi.org/10.1038/s41467-019-11036-9 OPEN1Department of Anatomy and Embryology, Leiden University Medical Center, 2333 ZC Leiden, Netherlands. 2Sequencing Analysis Support Core, LeidenUniversity Medical Center, 2333 ZC Leiden, Netherlands. 3Department of Immunohematology and Blood Transfusion, Leiden University Medical Center,2333 ZA Leiden, Netherlands. 4Department of Gynaecology, Division of Reproductive Medicine, Leiden University Medical Center, 2333 ZA Leiden,Netherlands. 5Department for Reproductive Medicine, Ghent University Hospital, 9000 Ghent, Belgium.6These authors contributed equally: X. Fan, M.Bialecka. Correspondence and requests for materials should be addressed to S.M.C.d S.L. (email: lopes@lumc.nl)NATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunications 11234567890():,;Content courtesy of Springer Nature, terms of use apply. Rights reserved In the absence of a pregnancy, both the ovary and the uterusundergo significant monthly remodeling during the entirereproductive period (about 40 years) in healthy women. Fromthe 1 million follicles present in the ovary at birth, only about 500reach the ovulatory phase during the reproductive span of healthywomen, while the rest degenerates1–3. Follicular growth involves:(1) maturation of the oocyte; (2) the extensive cellular pro-liferation and differentiation of the granulosa cells (GC) to sup-port the oocyte (cumulus GC) and allow the accumulation offollicular fluid in the antrum (mural GC); and (3) the generationof a specialized tissue layer of theca cells (TC) from stromal cells,surrounding the follicle, with high vascularization2–4.Follicular degeneration or atresia occurs at any stage offolliculogenesis4,5. As per month only one dominant folliclereaches the ovulatory stage6, it is imperative for ovarianhomeostasis that other growing follicles are efficiently removedby atresia to accommodate following waves of follicular growth.Robust tissue remodeling also occurs monthly with the trans-formation of the ovulatory follicle into the hormone-producingcorpus luteum (formed by lutein GC, lutein TC, and vascu-lature), followed by regression to a corpus albicans2–5.Although the molecular signature of the cumulus GC of thedominant ovulatory follicle is known, due to the accessibility tomaterial from patients using artificial reproductiontechnologies7,8, the continuous process of follicular growth anddegeneration is not well understood in humans. Here, we havethought to identify the somatic cell types and associated signalsthat regulate tissue remodeling in the adult ovary. Under-standing these mechanisms is paramount to pinpoint causes ofinfertility and to develop both treatments and diseasemodels1,9–11.ResultsSample preparation from the inner cortex of adult ovary.Wehave analyzed anonymised ovarian tissue (inner cortex) fromadult women (N=5) undergoing fertility preservation proce-dures (outer cortex is cryopreserved). This included several smallantral follicles (tissue with an outer diameter of about 2–4mmincluding stroma, TC, and a visible follicle of 1–2 mm in dia-meter) and selectable follicles (tissue with an outer diameter ofabout 5–8 mm including stroma, TC, and a visible follicle with2–5 mm of diameter) (Fig. 1a, b). Follicles in (early) stages ofatresia (Fig. 1b) may resemble growing follicles (Fig. 1a) in size,but the cellular organization of mural GC and TC layers differ.Follicles in (early) stages of atresia (Fig. 1c) showed less cellularproliferation and only moderate levels of apoptosis compared togrowing follicles (Fig. 1d). In these (early) atretic follicles, thelayer of TC showed pronounced signs of luteinization (hyper-trophied morphology), whereas the GC detach from the basementmembrane (Fig. 1b, iv and vi)12,13. In later stages of atresia, itremains unclear whether TC differentiate to or are replaced byfibroblasts (Fig. 1b, v)12,13.A total of 31 different tissue samples from the inner cortex ofthose adult ovaries (N=5), including stroma and visible folliclesof different sizes were collected for single-cell sequencing(Supplementary Data 1). After enzymatic dissociation, eachsample was subjected to FACS-sorting to remove dead cells(Fig. 2a) and analyzed by single-cell sequencing using the 10XGenomics platform (56,206 cells) (Supplementary Data 1). Thedata was filtered using quality control parameters as described inR package Seurat14 (Supplementary Fig. 1a). In addition, cellsexpressing high levels ( 6% of total UMIs) of dissociation-relatedgenes15 were excluded from further analysis (SupplementaryFig. 1b; Supplementary Data 1). We retained 20,676 cellsexpressing 2516 (highly variable) genes for further analysis.Single cell clustering and cell type identification. Using aSeurat-based workflow and a correction step to minimize patient-specific effects (using mutual nearest-neighbor method), weclustered the cells and identified 19 clusters. In the same work-flow, we run the non-linear dimensionality reduction algorithmtSNE to visualize the cells in a two-dimensional plot (Fig. 2b;Supplementary Fig. 1c, d). We calculated the top 30 differentiallyexpressed genes (DEGs) from each cluster, filtered for genes withaverage logefold change 0.5, sorted by their adjusted p-values(Wilcoxon rank sum test) (Supplementary Data 2) and plottedone representative DEG per cluster (Fig. 2c). Based on the DEGs,we performed a Gene Ontology (GO) enrichment analysis(Supplementary Data 3) to facilitate cluster identification. Next,we generated a hierarchical clustering using the 50 most variablyexpressed gene means per cluster and distinguished five majorcell types: GC (five clusters), TC and stroma (five clusters),smooth muscle cells (two clusters), endothelial cells (three clus-ters), and immune cells (four clusters) (Fig. 2d). To confirm theidentity of these cell types in the tSNE, we colored the single cellsaccording to the expression levels of several expected
Markergenes: AMH,HSD17B1,SERPINE2,GSTA1 for GC; DCN,LUMfor TC and stroma; TAGLN and RGS5 for smooth muscle cells;VWF and CLDN5 for endothelial cells; and CD53 and CXCR4 forimmune cells (Fig. 2e).To provide a brief characterization of the cells removed fromthe total dataset (56,206 cells), we plotted the retained cells(20,676 cells) in a tSNE that included all cells (SupplementaryFig. 1e–g). Instead of 19 clusters, we obtained 21 clusters, eachcontaining both retained and removed cells. Comparing theDEGs associated with each of the 21 clusters (SupplementaryData 4) with the DEGs associated with the 19 clusters obtainedfrom the retained cells (Supplementary Data 2), we were able tomatch the large majority of the clusters, confirming that the cellsremoved from each cluster corresponded to stressed cells fromeach specific cluster, due to high levels of dissociation-relatedgenes15. From the DEGs of the unmatched clusters, we were ableto identify those extra populations as stroma and endothelial cells,related to the retained stroma and endothelial cell clusters. Wecannot exclude that those correspond to biological relevantpopulations.Vascular remodeling in the adult ovary. Vascular remodeling inthe ovary, supporting the dynamic changes in follicular growthand degeneration, has gained more attention in recent years2–4.We identified three separate clusters (CL) of endothelial cells(CL7, CL9, CL16) expressing markers associated with lymph andblood vascular system (such as PECAM1,CD34,CTGF), but alsoassociated with remodeling and inflammatory response (such asTXNIP,ANGPT2) (Fig. 3a–d). The DEGs of CL7 (such as CCL14,SOCS3,EGFL7) and CL16 (such as CCL21,TFF3) are linked toangiogenesis and lymphatics, respectively, while DEGs of CL9(TM4SF1,NMMT) were more related to regulation of apoptosis(Fig. 3c, d). The clusters of smooth muscle cells (CL14, CL17) alsoshowed features of growth and remodeling: many DEGs of CL17(such as CRYAB,GJA4) were involved in regulation of immuneresponse and apoptosis, whereas DEGs of CL14 (such as ACTA2,PLN,ADIRF, and MYH11) associated with mature smoothmuscle cells (Fig. 3e–g).Molecular and cellular signature of different GC populations.To reveal the dynamic cellular changes that take place duringantral follicle maturation, we focused first on the analysis of theGC in antral follicles (Fig. 4a; Supplementary Fig. 2a, b). The GCof small antral follicles (1–2 mm diameter) clustered pronounc-edly in cluster (CL15) showing WT1high/EGR4high/VCANlow/ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-92NATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunicationsContent courtesy of Springer Nature, terms of use apply. Rights reserved FSTlow expression (Fig. 4b), suggesting that at that stage muraland cumulus GC still have a common progenitor (pGC) sig-nature. In agreement WT1 was expressed in GC, in particular inthe GC forming the corona radiata (Fig. 4c), in contrast to mouseovaries where WT1 marked TC-progenitors16. GC from select-able follicles (2–5 mm diameter) separated into two main celltypes: cumulus GC (VCANhigh/FSThigh/IGFBP2high/HTRA1high/INHBBhigh/IHHhigh) and mural GC (WT1low/EGR4low/KRT18high/CITED2high/LIHPhigh/AKIRIN1high) (Fig. 4b)7,17.Inagreement, pan-KRT immunostaining revealed higher proteinexpression in mural than in cumulus GC (Fig. 4c).Several samples (such as selectable follicle D, but not follicle C)contained GC in CL10 (Fig. 4a, Supplementary Fig. 2a, b). TheGC of CL10 were negative for several GC markers, such as VCANand FST (Fig. 4b), but were also negative for KRT18 (Fig. 4b),similarly to pan-KRT-negative GC in atretic follicles (Fig. 4d).This suggested that CL10 could represent GC in the early stagesof atresia. The GC in CL10 expressed lower levels of GJA1 andCDH2 compared to the other GC clusters (Fig. 5a, b). Lowerlevels of GJA1 have been described in GC of atretic comparedwith healthy follicles in rats18, where it was suggested thatreduced gap junctions, and hence cellular communication, play arole in atresia. Using immunostaining, we confirmed lowerexpression of GJA1 and CDH2 in GC of atretic follicles inhumans (Fig. 5c, d, bottom two rows) compared with growingfollicles (Fig. 5c, d, top two rows).We used two independent methods to analyze the celltrajectories of the GC. Due to the limited number of samplesavailable, the intermediate states are not well represented andhence our conclusions regarding trajectories should be consideredø1.7 mm ø3.2 mm ø3.7 mm ø5.5 mmDDX4 DAPIDDX4 DAPIø2.0 mmø2.0 mmKI67 TUNELTUNEL KI67 DAPI KI67 TUNELTUNEL KI67 DAPIaø1.7 mmø3.2 mmcø0.7 mm ø1.3 mm ø1.6 mm ø1.4 mm ø1.9 mm ø2.0 mmMin Max Min Maxi ii iiidv vi viiiii iiiivvviviiivbFig. 1 Morphology of different follicles in adult ovaries. a,bImmunofluorescence of several healthy (a) and atretic follicles (b) of different sizes present inthe ovaries immunostained for DDX4. Green arrows point to (DDX4-positive) oocytes. Panels in the bottom row show magnifications of the boxed areaswith corresponding roman numbers. Tissue was counterstained with DAPI. Scale bars are 1 mm in two top rows and 100 μm in bottom row (panels i-vii).c,dImmunofluorescence of atretic (c) and healthy follicles (d) of different sizes present in the ovaries analyzed for KI67 and TUNEL. Insets in dshow adifferent area of the follicle showing mural GC with same magnification. Single channel images were converted to an intensity map. Scale bars are 100 μmNATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-9 ARTICLENATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunications 3Content courtesy of Springer Nature, terms of use apply. Rights reserved preliminary. Pseudotime analysis using Monocle 3 alpha, thatplaces the progenitor cell population in the middle of a longertrajectory segment, revealed that pGC (CL15) branched to muralGC (CL11) and mature cumulus GC (CL8 and CL3) (Fig. 6a). Aspseudotime analysis is susceptible to be affected by inter-individual variation, we highlighted cells from two individuals(P7 and P3) showing cells of both in each cluster. The celltrajectories obtained by Monocle 3 alpha were consistent with thecell trajectories obtained using Diffusion maps (Fig. 6b).Several genes not earlier associated with GC include TNNI3,MAGED2,SPINT2 PLA2G16,BEX1,DSP,TSPAN6, and LCMT1(Fig. 6c). Due to the limited amount of samples per individual, wecannot exclude that some genes may represent differentialexpression between individuals. We verified the expression ofSERPINE2AMHGSTA1Granulosa135135HSD17B1024RGS5LUMDCNTAGLNSmooth muscle Theca stroma135135135135CXCR4CLDN5CD53VWFImmune Endothelial135024123135024daeDNAJB9GPRC5ACXCL21PTCH1NKG7HTRA1TCEAL4CCL14INHBBTM4SF18RBP1LRATIL7RLYZACTA2EGR4FABP4GJA4IGKC0123567810111213141516171849ClusterIDc1679124181311151038171402615ClusterIDGranulosa Theca stromaSmoothmuscleEndothelial ImmuneMALAT1B2MHSP90AA1TMSB10CD74SRGNGPR183CYBAHLA-DPA1HLA-DPB1HLA-DRB1HLA-DRAANXA1HLA-ECTGFTM4SF1SPARCL1HLA−BHLA−CHLA−AIGFBP7FOSJUNHSPA1AHSPB1SAT1DNAJA1HSPE1HSPD1APOELUMDCNCCL21TFPIADIRFMYL9ACTA2TAGLNGSTA1TNNI3AMHIFI27RBP1STMN1SERPINE2MAGED2FHL2SOX4NGFRAP1MDKValueColor key0200Count0143625b0123456789101112 131415161718−2502550−50 −25 0 25 50tSNE_1tSNE_20123456789101112131415161718ClusterIDSingle cellsuspensionTissue dissection Live cellsSingle cell RNAseqSSC7AADFig. 2 Transcriptome map of human adult ovaries analyzed. aSchematic representation of human ovarian tissue preparation for single cell transcriptomeanalysis. btSNE cluster map revealing 19 specific clusters representing the major ovarian somatic cell types. cViolin plots showing expression of onerepresentative differential expressed gene for each cluster. dHeatmap and hierarchical clustering based on expression of top 50 most variable genes.etSNE cluster map showing expression of genes characteristic the major ovarian somatic cell types. Red dashed lines give the boundaries of the mainclusters of interestARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-94NATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunicationsContent courtesy of Springer Nature, terms of use apply. Rights reserved TNNI3 in GC and observed a specific cellular localization. Thismay be important for regulating GC shape and function: TNNI3formed a characteristic ring-shaped structure in inner GC, butwas distributed to the baso-lateral cell membrane of the GC onthe basement membrane (Fig. 6d).Molecular and cellular signature of TC and stromal cells. TheTC appeared in three separate TC clusters (CL2, CL5, CL6)(Figs. 2b, 4a). Those included TC present in samples with follicles(Supplementary Fig. 2a, b) and in stromal samples without visiblefollicles, but that could contain TC from follicle walls, corpusluteum or albicans (Supplementary Fig. 2c). TC from (healthy)follicles, such as follicles A, B, and C, were mainly clustered inCL5 (Fig. 4a), characterized by expression of known markersPTH1,APOD,APOC1, and several genes not earlier associatedwith TC, such as WFDC1,MATN2,COLEC11 (Fig. 7a). TC fromthe small antral follicles A and B did not overlap with TC fromselectable follicle C in CL5 (Fig. 4a). To explore differencesbetween these domains inside CL5, we calculated a separate tSNEusing only CL5 cells to further characterize the sub-populationsof TC (Fig. 7b). Combining differential gene expression of theobtained TC sub-clusters (Fig. 7c; Supplementary Data 5) withcell trajectory analysis using two independent methods (Monocle3 alpha and Diffusion maps) (Fig. 7d, e), we concluded that TCfrom the small follicles correspond to common progenitor TC(pTC) that progress to interna TC (inTC) and externa TC (exTC)from selectable follicles. Both CL-T0 and CL-T1 showed a profileACTA2DAPI VWFVWFPECAM1DAPI PECAM1DAPI ACTA2aeEndothelialCD3413PECAM1024CCL21246TFF3135CCL14135SOCS3024EGFL7024CTGF246246TM4SF1TXNIP024ANGPT2135NNMT024CRYABGJA4024135Smooth musclePLNACTA2024135ADIRF MYH11246024bcfdGO:0043068 GO:0010942GO:0043067 GO:0042981R-HSA-8953897: Cellular responses to external stimuliGO:0002444: Myeloid leukocyte-mediated immunityR-HSA-9018519: Estrogen-dependent gene expressionGO:0030029: Actin filament-based processR-HSA-445355: Smooth muscle contractionR-HSA-195258: RHO GTPase effectorsGO:0019221: Cytokine-mediated signaling pathway GO:0097190: Apoptotic signaling pathwayGO:0060333: Interferon-gamma-mediated signaling pathwayhsa04612: Antigen processing and presentationGO:0051101: Regulation of DNA bindingGO:0010631: Epithelial cell migrationR-HSA-109582: Hemostasis GO:0045055: Regulated exocytosisGO:0060838: Lymphatic endothelial cell fate commitmentCL7 (63)CL9 (99)CL16 (107)CL14 (106)CL17 (104)gCL763 50CL999456107CL1642CL14 CL1752106 1041428 301524Cell death and apoptosisFig. 3 Vascular remodeling in the ovaries analyzed. a,bImmunofluorescence of ovarian stroma for PECAM1 a, VWF b, and the respective single channelimages. Slides were counterstained with DAPI. Scale bars are 100 μm. ctSNE cluster maps showing expression of selected endothelial marker genes. Reddashed lines give the boundaries of the endothelial-clusters of interest. dVenn diagram showing the intersection of 200 differential expressed genes(DEGs) of the three endothelial cell clusters (CL7, CL9, and CL16); and three selected enriched terms obtained for the unique DEGs. eImmunofluorescenceof ovarian stroma for ACTA2, and the respective single channel image. Slides were counterstained with DAPI. Scale bars are 100 μm. ftSNE cluster mapshowing expression of selected smooth muscle marker genes. Red dashed lines give the boundaries of the smooth muscle-clusters. gVenn diagramshowing the intersection of 200 differential expressed genes (DEGs) of the two smooth muscle cell clusters (CL14 and CL17) and genes from four-celldeath and apoptosis-related GO terms (GO:0043068 positive regulation of programmed cell death, GO:0010942 positive regulation of cell death,GO:0043067 regulation of programmed cell death, GO:0042981 regulation of apoptotic process); and three selected enriched terms obtained for theunique DEGsNATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-9 ARTICLENATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunications 5Content courtesy of Springer Nature, terms of use apply. Rights reserved of pTC, but cells in CL-T0 showed additional expression of stress-related markers (such as FOS,JUN) (Fig. 7c). Similarly, CL-T2and CL-T4 showed a profile of exTC, but cells in CL-T2 alsoshowed more pronounced expression of stress-related markers(such as FOS,JUN) (Fig. 7c). Interestingly, there were differencesin the trajectories obtained by the two methods regarding thelocalization of CL-T2 (stressed exTC) in the trajectory. Inmonocle, we obtained pTC →stressed pTC +pTC →inTC →exTC →stressed exTC (Fig. 7d). By contrast, in the diffusionmaps the two population of stressed TC (CL-T0 and CL-T2) werein the same trajectory and we observed pTC →stressed pTC →stressed exTC +pTC →inTC →exTC (Fig. 7d).Immunostaining for STAR and ACTA2 was sufficient todistinguish inTC from exTC in growing follicles (Fig. 7f, top tworows). This highlights similarities between TC and the corre-sponding male-equivalent (ACTA2+) peritubular myoid cells inthe testis19. In atretic follicles, STAR expression remained in TC,whereas ACTA2 became more restricted to smooth muscle cells(Fig. 7f, bottom two rows). IFITM3, so far only reported in bovinefollicles20 proved useful in humans to distinguish (IFITM3+)TCfrom (GJA1+) GC at least in growing follicles (Fig. 5c).Surprisingly, we observed in two follicles from one woman (andnot in any of the follicles analyzed from the other four women)that (CDH2−/COLIV+or GJA1−/IFITM3+) TC showedclear protrusions into the cumulus (CDH2+/COLIV−orGJA1+/IFITM3−) GC area (Fig. 5c, d). Due to the low numberof individual women (N=5), it remains unclear how commonthis feature is.TC from (early atretic) follicle D and present in stromalsamples were mainly present in CL2 and CL6 (Fig. 4a,Supplementary Fig. 2c), suggesting that those may representatretic TC. The TC in CL2 and CL6 expressed IFITM3, lowerlevels of COL3A1 and higher levels of FOS and IGFBP5 comparedto the TC in CL5 (Fig. 8a). Using immunostaining, we confirmedexpression of FOS and IGFBP5 in the TC of early atretic follicles,where it colocalized with marker of DNA/RNA damage 8OHdGand STAR (Fig. 8b, c, bottom row) in contrast to TC in growingfollicles (Fig. 8b, c, top row).Most stromal cells from selectable follicle C clustered in CL1(Fig. 3a). Although stromal clusters (CL0, CL1) showed highexpression of GNL3 and ARID5B (Fig. 8d), CL0 expressed highlevels of XBP1 and SELK (Fig. 8e), both involved in endoplasmicreticulum (ER)-stress-induced apoptosis, whereas CL1 expressedhigh levels of GPRC5A and TNFRS12A (Fig. 8f).Immune cells and complement system in the adult ovary.Interestingly, both the TC clusters (CL2, CL5, CL6) and stroma(CL1) showed pronounced expression of several components ofa−2502550−50 −25 0 25 50tSNE_1tSNE_2Rest samplesFoll. 1–2 mm-AFoll. 1–2 mm-BFoll. 2–5 mm-CFoll. 2–5 mm-DatrTC atrGCCumulusGCpGCMuralGCTCbcMin MaxWT1ø1.7 mmWT1 pKRT DDX4 DAPIWT1WT1 pKRT DDX4 DAPIø3.2 mmpKRTpKRTWT1 pKRT DDX4 DAPIWT1ø2.0 mmWT1 pKRT DDX4 DAPIWT1ø2.0 mmpKRTpKRTMin MaxdIHHIGFBP2 INHBBHTRA1Cumulus GC024024024024024024CITED2 LIPHKRT18 AKIRIN1Mural GC pGC1313024FSTWT1 VCANEGR4 pGC vs GC0241313Fig. 4 Divergent populations of granulosa cells in different follicles. aDistribution of single cells from different-sized follicles on the tSNE. Black dashed linesgive the boundaries of several clusters of GC (common progenitor GC (pGC), mural GC, cumulus GC, atretic GC (atrGC), theca cells (TC), and atretic TC(atrTC)). btSNE cluster map showing expression of selected marker genes differentially expressed by GC and pGC (top row), cumulus GC (middle row),and mural GC and pGC (bottom row). Red dashed lines give the boundaries of the GC-clusters of interest. c,dImmunostaining of follicles (ø, diameter)growing (c) and atretic (d) for WT1 and pan-KRT (pKRT). Inset shows mural GC of the same follicle with the same magnification. Single channel imageswere converted to an intensity map. Scale bars are 100 μmARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-96NATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunicationsContent courtesy of Springer Nature, terms of use apply. Rights reserved the complement system, such as C1R, C1S, and C7 (Fig. 9a). Inagreement, gene network in CL1, CL2, CL5, and CL6 alsorevealed an association between complement genes and TC andstromal genes (such as LUM,COL6A1,COL1A1) (Fig. 9b). On theother hand, pGC, mural GC, and other cell populations in theovary expressed CD55 and CD59, known to protect against tar-geting and damage by the complement cascade (Fig. 9a), sug-gesting a concerted mechanism involving the complement systemto allow growth and degeneration of the different follicularcompartments in a timely manner.The complement system has been reported in the follicularfluid of women undergoing in vitro fertilization (IVF) treat-ment21, but has not been associated with physiological tissueremodeling in the human adult ovary. Using immunofluores-cence, we confirmed increasing expression of C1S from healthy todegenerating follicles (Fig. 9c, d), whereas expression of C1Q wasmainly confined to (CD68+) macrophages22 present in in theovary. There was an overabundance of (CD68+) macrophages indegenerated follicles (Fig. 9c, d), suggesting an innate immuneresponse during follicular remodeling. Although most circulatingcomplement proteins (except C1Q and C7) are produced in theliver, local production of complement has been detected in avariety of tissues and cells23. To determine the local production ofcomplement components by ovarian cells, we cultured pieces ofhuman ovarian stroma (N=3) and atretic follicle walls (N=6)for 1 and 5 days and determined the concentration of C1Q andC3 by
ELISA. We observed low, but increasing levels of C1Q andC3 by day 5 (Fig. 9e), suggesting that the ovary could contributeto the local production of circulating complement proteins.Finally, it is not surprising that the ovaries analyzed showed apronounced population of CD53high/CXCR4high immune cells(Fig. 2e), including separate clusters for adaptive T lymphocytesand Natural Killer (NK) cells (CL4 and CL12), B lymphocytes(CL18), and innate immune system, such as monocytes andmacrophages (CL13) (Fig. 2c, d). Note that some of the innateimmune cells expressed high levels of CD68 and IFI30, as well ascomplement component C1QA (Fig. 9f, g). B lymphocytesexpressed high levels of JCHAIN and IGKC (Fig. 9h) and Ta024GJA113CDH2Granulosa: healthy43210GJA1CL3 CL8 CL10 CL11 CL15dø1.7 mmø3.2 mmø2.0 mmø2.0 mmCOLIV CDH2 DDX4 DAPIbCDH2 COLIVMin Max43210CDH2CL3 CL8 CL10 CL11 CL15Min MaxcIFITM3 GJA1 ZP3 DAPI IFITM3 GJA1ø1.7 mmø3.2 mmø2.0 mmø2.0 mmAtretic follicle Growing follicleAtretic follicle Growing follicleFig. 5 Granulosa cells in early atretic follicles. atSNE cluster map showing expression of selected genes downregulated in CL10, but not on the othergranulosa cell (GC) clusters. Red dashed lines give the boundaries of expression. bViolin plots showing expression levels of GJA1 and CDH2 in the differentclusters of GC. cImmunostaining of follicles (ø, diameter) growing (top two rows) and atretic (bottom two rows) for IFITM3, GJA1, and ZP3. Inset showsmural GC of the same follicle with same magnification. Single channel images were converted to an intensity map. White dotted line marks the basementmembrane. Scale bars are 100 μm. dImmunostaining of follicles (ø, diameter) growing (top two rows) and atretic (bottom two rows) for CDH2, COLIV,and DDX4. Inset shows mural GC of the same follicle with same magnification. Single channel images were converted to an intensity map. Scale barsare 100 μmNATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-9 ARTICLENATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunications 7Content courtesy of Springer Nature, terms of use apply. Rights reserved lymphocytes (and NK cells) expressed IL7R,LTB,CCL5, andNKG7 (Fig. 9i).DiscussionSeveral studies have used single-cell technology to reveal themolecular signatures of fetal24,25 and adult oocytes26–31 duringhuman oogenesis. Li and colleagues have investigated the mole-cular signature of human ovarian fetal somatic cells, includingGC present in primordial follicles, providing insights in the sig-naling network that takes place during the formation of pri-mordial follicles during fetal life24. From adult ovaries, pools of 10GC isolated from several preantal follicles and cumulus–oocytecomplex26,30 have been analyzed. Differences between humanatretic and growing follicles (oocyte and GC) have not beenreported and other somatic ovarian cells in the adult ovary havenot been characterized at the transcriptional level. Our datasetprovides a first step to fill the gap in knowledge regarding thecharacterization of the somatic cell types present in theadult ovary.It is well accepted that the events that lead to ovulation(including remodeling of extracellular matrix, chemotaxis,microcirculatory vasomotion, formation of the oocyte–cumuluscomplex) are regulated by a cytokine-mediated inflammatoryresponse orchestrated by lymphocytes, granulocytes, and mac-rophages32. Moreover, GC present in ovulatory follicles seem tob−0.050−0.0250.0000.025−0.03 −0.02 −0.01 0.00 0.01DM1DM2−0.050−0.0250.0000.025−0.03 −0.02 −0.01 0.00 0.01Foll. 1–2 mm-A Foll. 1–2 mm-BFoll. 2–5 mm-C Foll. 2–5 mm-D0.00−0.02 0.00−0.02−0.050−0.0250.0000.025−0.050−0.0250.0000.025Patient P7 P30.00.51.01.52.00.5 1.0 1.50.0 1.0 2.0Pseudotime0.5 1.0 1.50.00.51.01.52.0PatientP7 P30.5 1.0 1.50.5 1.0 1.50.00.51.01.52.00.5 1.0 1.5 0.5 1.01.5Foll. 1–2 mm-A Foll. 1–2 mm-B Foll. 2–5 mm-C Foll. 2–5 mm-D0.00.51.01.52.00.5 1.0 1.5Component 2CL11 CL15CL3 CL8Component 1GranulosacTNNI3024MAGED2024SPINT2135PLA2G16024024BEX1024DSP024TSPAN613LCMT1CL11 CL15CL3 CL8pGCMural GCCumulusGCpGCMural GCCumulusGCaTNNI3Inner GC Inner GCBasement GCBasement GCInner GC Inner GCBasement GCø3.2 mmø1.7 mmø2.0 mmTCTC TCBasement GC Basement GCBasement GCTCø2.0 mm Basement GCdTNNI3 AMH DDX4 DAPIBasement GCMin MaxAtretic follicle Growing follicleFig. 6 Cell trajectory analysis and characterization of granulosa cells. a,bAnalysis of cell trajectories of granulosa cells (GC) (CL3, CL8, CL11, CL15) byMonocle aand Diffusion maps b. Individual cells (dots) are colored by cluster, follicle, patient, and pseudotime (Monocle). ctSNE cluster map showingselected genes expressed by GC. Red dashed lines give the boundaries of the GC-clusters. dImmunostaining of follicles (ø, diameter) growing (top tworows) and atretic (bottom two rows) for TNNI3, AMH, and DDX4. Higher magnification of inner and basement GC is shown on the right side. Singlechannel images for TNNI3 converted to an intensity map are shown. White arrowheads depict basement GC, yellow arrowheads depict inner GC. Scalebars are 100 μmARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-98NATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunicationsContent courtesy of Springer Nature, terms of use apply. Rights reserved have properties of innate immune cells33. However, themechanisms that regulate follicular remodeling and ultimatelyregression are much less well understood5,11,34,35. Although it isfeasible that the components of the complement system expressedin stroma and TC have no impact on follicular remodeling, theirlocal production in the human ovary is intriguing. In agreementwith our results, microarray analysis of TC from atretic follicles(3–5 mm diameter) also revealed a prominent up-regulation ofcomponents of the complement system (such as C1R,C1S,C7,SERPING1) when compared to healthy TC in bovine36. In thatstudy, inflammatory response pathways rather than cell deathcharacterized the atretic TC in bovine. The local activation of thecomplement system, contributing to an inflammatory andimmune response as observed in certain organs23, may be alsodACTA2 STARSTAR ACTA2 ZP3 DAPIfø1.7 mmø3.2 mmø2.0 mmø2.0 mmPTH113APOD135WFDC113APOC1024−20−1001020T0T1T2T3T4−20 −10 0 10 20tSNE_1tSNE_2T0T1T2T3T4ClusterIDFOSFOSBJUNJUNBTXNIPRASD1DUSP1NME2APODAPOC1MESTWFDC1MATN2PTCH1CYP11A1GSTA1APOA1STARFDX1C4BPBINHACOL1A1COL1A2COL6A1COL6A2ACTA2NDRG2VMP1FGFR1STAT3COL8A1ALDH1A1−1 0 1T0T1T2T3T4pTC Stress markersinTC exTCCluster IDPercentage of cells 25 50 750Average expressioninTCexTCpTCTheca: healthy Min Maxeabc024MATN2 COLEC11131.01.52.02.50.0 0.5 1.0 1.5 2.0 2.5Pseudotime0.0 2.01.01.52.02.50.0 1.0 2.0 0.0 1.0 2.0 0.0 1.0 2.00.0 1.0 2.0Foll. 1–2 mm-A Foll. 1–2 mm-B Foll. 2–5 mm-C Foll. 2–5 mm-D1.01.52.02.50.0 0.5 1.0 1.5 2.0 2.5Component 1Component 2T0 T1 T2 T3 T41.01.52.02.50.0 0.5 1.0 1.5 2.0 2.5Patient P7 P9 P0 P3str.exTCexTCstr.pTCpTCinTC−0.030.000.030.06−0.075 −0.050 −0.025 0.000 0.025DM1DM20.00−0.050.00−0.05Foll. 1–2 mm-A Foll. 1–2 mm-BFoll. 2–5 mm-C Foll. 2–5 mm-D−0.030.000.030.06−0.030.000.030.06−0.030.000.030.06−0.075 −0.050 −0.025 0.000 0.025Patient P7 P9 P0 P3str.pTCpTCstr.exTCexTC inTCT0 T1 T2 T3 T4Atretic follicle Growing follicleFig. 7 Divergent populations of theca cells in different follicles. atSNE cluster map showing expression of selected theca cells (TC) genes. Red dashed linesgive the boundaries of the expression. btSNE cluster map revealing sub-clusters of CL5 representing ovarian TC types. Black dashed lines give theboundaries of several sub-clusters of TC: common progenitor TC (pTC), externa TC (exTC), and interna TC (inTC). cExpression of marker genes in sub-clusters of CL5. d,eAnalysis of cell trajectories of TC (CL5) by Monocle aand Diffusion maps b. Individual cells (dots) are colored by cluster, follicle,patient, and pseudotime (Monocle). fImmunostaining of follicles (ø, diameter) growing (top two rows) and atretic (bottom two rows) for STAR, ACTA2,and ZP3. Inset shows mural GC of the same follicle with same magnification. Single channel images were converted to an intensity map. White dotted linemarks the basement membrane. Scale bars are 100 μmNATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-9 ARTICLENATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunications 9Content courtesy of Springer Nature, terms of use apply. Rights reserved taking place in the human ovary. This may prove important forthe physiological homeostasis of the ovary, perhaps potentiatingfollicular remodeling and its role deserves to be explored in thefuture.Female infertility can be caused by immune systemdisorders37,38. Increased activation of complement system inperitoneal fluid has been associated with endometriosis-associated infertility39. Moreover, patients with systemic lupuserythematosus, an autoimmune disease in some cases caused by aC1Q-deficiency40,41, showed levels of infertility that are higherthan in the normal population42. Polycystic ovary syndrome hasbeen confirmed as a low-level chronic inflammation impacting onovulation and luteinization43,44. Although female reproductive(dis)function has not been directly linked to the complementsystem, our study has identified the complement system as pos-sible mechanism to regulate homeostasis and tissue remodeling inthe adult ovary.MethodsEthical permission and collection of human material. Ovaries from cancerpatients undergoing elective ovariectomy, prior to cancer treatment, were removedfor the purpose of fertility preservation (cryopreservation). The phase of menstrualcycle was not determined prior to surgery. Signed informed consent was obtainedform all patients to perform research on the anonymized rest material left overfrom the cryopreservation procedure. The research was approved by the MedicalEthical Committee of the Leiden University Medical Center (CME 05/03K/YR).After the outer layer of the ovary (1 mm thick) was collected for cryopr eservationpurposes and the inside of the ovary was fragmented for further analysis (single-cell RNA-sequencing or immunofluorescence).Single-cell dissociation of ovarian tissue. A total of 31 adult ovarian tissuesamples (2–8 mm diameter) containing (whole or parts of) a single visible follicle(1–2mm or 2–5 mm diameter) or without visible follicles were dissociated forsingle cell transcriptomics as previously described45. Briefly, individual tissuesamples from adult ovary were incubated overnight on ice with 1 mg/ml Col-lagenase Type II (Life Technologies) in 0.25% Trypsin-EDTA (Life Technologies).Next, the samples were centrifuged at 160 × gfor 3 min and incubated withAdvanced DMEM/F12+Glutamax (Life Technologies), 1x Insulin-Transferin-Selenium (Life Technologies), 1x Penicillin/Streptomycin (Life Technologies) and27 IU/ml RNase-free DNase I (Qiagen) at 37 °C for 30 min to 2 h. The digestionwas stopped by adding 10% of fetal calf serum (Gibco), followed by a filtration stepthrough a 100 µm strainer (Corning). Samples were centrifuged at 160 × gfor 5min and stored in liquid nitrogen in Bambanker (Nippon Genetics).Fluorescence-activated cell sorting (FACS). Single cells were resuspended inFACS buffer composed of 1% bovine serum albumin (BSA, Life Technologies),2 mM EDTA (Life Technologies) in DPBS without calcium and magnesium (LifeTechnologies) and passed through the (pre-wet) strainer cap of FACS tubes(Corning). Cells were stained with 7-AAD (1:100, BioLegend) 3 min on ice. Livecells were sorted on a BD FACSAria I (
BD Biosciences) equipped with blue laserand 695/40A long pass filter and BD FACSDiva 8.0.1 software and collected in 1%BSA in DMEM/F12 (Life Technologies) with 1x Penicillin/Streptomycin.RNA-sequencing and primary sequencing analysis. The library preparation wasperformed using the Chromium Single Cell 3′Reagent Kit, version 2 (10XGenomics) and sequenced on a HiSeq4000 using a 300 cycles kit (Illumina). Rawsequencing data was processed using Cell Ranger analysis pipeline v2.1.1. Readswere aligned to human genome version GRCh38. For downstream analysis Cell024GNL3 ARID5BStroma135024GPRC5A TNFRSF12AStroma: healthy13024XBP1 SELKStroma: stressed135cFOS 8OHdGcFOS 8OHdG DAPIMin Maxø2.0 mmø1.7 mmIGFBP5 STARIGFBP5 STAR DAPIMin Maxø2.0 mmø1.7 mmbcdefThecaIFITM3024COL3A1135a135135FOS IGFBP5Fig. 8 Characteristics of theca and stromal cells. atSNE cluster map showing expression of selected marker genes healthy and atretic theca cells (TC). Reddashed lines give the boundaries of the TC-clusters of interest. b,cImmunostaining of follicles (ø, diameter) growing (top row) and atretic (bottom row)for cFOS and 8OHdG (b) and for IGFBP5 and STAR c. Inset shows mural GC and TC of the same follicle with same magnification. Single channel imageswere converted to an intensity map. Scale bars are 100 μm. d,ftSNE cluster map showing expression of selected marker genes in the ovarian stromad, healthy stroma eand stressed stroma (f). Red dashed lines give the boundaries of the stromal clusters of interestARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-910 NATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunicationsContent courtesy of Springer Nature, terms of use apply. Rights reserved Ranger output \"filtered gene-barcode”count matrix, containing the expressionprofile of cells with a correctly detected cellular barcode, was used.Secondary sequencing analysis. For further analysis, we adapted a workflow thatmakes use of R package Seurat, v2.2.014. The following parameters were used tofilter good quality cells (and exclude cells with extreme values indicating lowcomplexity, duplets or apoptotic cells): the total number of expressed genes/cell was200 nGenes 2500; the total number of UMIs/cell was 300 nUMIs 15000; andthe percentage of UMIs mapping to mitochondrial genes to total genes was percent.mito 0.1. In addition, cells with more than 6% of UMIs mapping to dissociation-induced genes, as based on literature15, were not further analyzed and cell-cycleeffects were regressed out46. Counts were normalized using the default normal-ization approach of Seurat (Function NormalizeData). Briefly, for each cell, theUMI counts for each gene were divided by the sum of UMI counts for all genes forthat cell. The result was multiplied by a fixed factor (10,000) and logetransformed.To correct for patient-effects (N=5) the mutual nearest neighbor (MNN)method47 from R package scran (v1.10), function fastMNN was used. Input forfastMNN were the 20 principal components calculated in a previous step in theworkflow. The output from fastMNN, corrected for patient-effects, was usedfurther downstream for the calculation of cell clustering and the tSNE plot.fastMNN also calculated the percentage of variance lost from each patient duringC1RacefgihdC1S C7 CFDCD59CD55CFHC1QACD68JCHAINIFI30 IL7RNKG7 CCL5LTBIGKCC3Serpring1ComplementInnate immuneComplementB-lymphocytesT-lymphocytes NK cells420420420420420420420321420420420420420420420420420321NetworksFunctionPathwayPredictedCo-expressionCo-localizationGenetic interactionsPhysical interactionsShared protein domainsComplemant activationHumoral immune responseAdaptive immune responseNone of aboveRegulation of immune effector processC1Q C3Stroma Atretic follicleD1 D50.000.020.040.06Concentration(µg/ml)D1 D50.00.20.40.60.82.0Concentration(µg/ml)Atretic follicle Growing follicleDegenerated follicleAtretic follicle Growing follicleDegenerated follicleC1S C1Q CD68 DAPI C1S C1Q CD68ø1.7 mmø2.0 mmC1Q STAR DAPIø1.7 mmø2.0 mmMin MaxFig. 9 Complement and immune system in the adult ovaries. atSNE cluster map showing expression of selected complement genes. bGene network of C1Sin TC and stroma. The color of the circles represents function and the color of the edges represent networks. cImmunostaining of follicles (ø, diameter)growing (top row), atretic (middle row), and degenerated (bottom row) for C1S, C1Q, and CD68. Single channel images were converted to an intensitymap. Scale bars are 100 μm. dImmunostaining of follicles (ø, diameter) growing (top row), atretic (middle row), and degenerated (bottom row) for C1Qand STAR. Scale bars are 100 μm. eConcentration of secreted C1Q and C3 produced by pieces of human ovarian stroma (N=3 samples) and atreticfollicle walls (N=6 samples) after 1 and 5 days of culture. Median and sample distribution (dots) are shown. ftSNE cluster map showing expression ofC1QA and C3.g–itSNE cluster map showing expression of selected immune marker genes for innate immune cells g, B lymphocytes hand T lymphocytesand NK cells i. Red dashed lines give the boundaries of the specific immune-clustersNATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-9 ARTICLENATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunications 11Content courtesy of Springer Nature, terms of use apply. Rights reserved orthogonalization at each merge step. The proportion of variance lost for eachpatient was reasonably low (P0 =0.16, P2 =0.04, P3 =0.03, P7 =0.08, P9 =0.07).Function FindAllMarkers from R package Seurat performed differentialexpression analysis (paired-wise) between the cells of a cluster and the rest of thecells in the dataset (Supplementary Data 2, 4, 5 for top 30 genes). The list of DEGsper cluster with adjusted p-value 0.01 (Wilcoxon rank sum test) was used for aGO terms enrichment analysis using R packages topGO v2.30.048 and org.Hs.eg.dbv3.5.049. The significance (p-value) of each GO term was estimated using theKolmogorov–Smirnov test (Supplementary Data 3 for top 20 GO terms). FunctionSplitDotPlotGG from R package Seurat have been used to generate the dot plot.For each cluster, the mean expression of all genes was calculated and the 50 mostvariable gene means were selected using R function rowVars, from package genefilter1.60.050. Those were used to generate a heatmap using function heatmap.2 from Rpackage gplots v3.0.151 and R function hclust was used with distance metric set to‘manhattan’and hierarchical clustering the agglomeration method set to ‘complete’.To infer cell trajectories, we used two methods: one implemented in R packagemonocle52 (http://cole-trapnell-lab.github.io/monocle-release/monocle3/) anddiffusion maps53 implemented in R package destiny. Monocle 3 alpha (v2.99.3) wasused to order the cells and infer their trajectory. In this workflow, UMAP, a non-linear dimensionality reduction method, is used. UMAP parameters (n_neighborsand min_dist) values were selected to optimize the representation of cells in thetwo-dimensional UMAP plot. Last, the beginning of pseudotime was selected onthe UMAP plot based on expression of selected markers. Diffusion map plots werecalculated by running the RunDiffusion function of Seurat with default settings.RunDiffusion calls function DiffusionMap from package destiny v2.12.0.Network analysis was generated by using GeneMANIA (http://genemania.org/)on the top 30 DEGs from the clusters of interest. Venn diagrams were generatedwith webtool http://bioinformatics.psb.ugent.be/webtools/Venn/ using the top200DEGs from each cluster and gene enrichment analysis was done with Metascape(http://metascape.org/).Immunofluorescence. Adult ovarian tissue samples (2–8 mm diameter) containingone or several visible follicles (1–2mmand 2–5 mm diameter) were fixed overnight in4% paraformaldehyde at 4 °C, transferred to 70% ethanol and embedded in paraffinusing a using a Shandon Excelsior tissue processor (Thermo Scientific, Altrincham,UK). Paraffin blocks were sect ioned (5 μm thickness) using a RM2065 microtome(Leica Instruments GmbH, Wetzlar, Germany) onto StarFrost slides (WaldemarKnittel). For immunostaining, paraffin sections were deparafinized in xylene (2 × 10min) followed by rehydration though a series of ethanol (100%, 100%, 90%, 80%, 70%)and ending with distilled water at room temperature (RT). For antigen retrieval,sections were treated for 20 min at 98 °C in a microwave (TissueWave 2, ThermoScientific) with 0.01 M sodium citrate buffer (pH 6.0), except for immunostaining withrabbit anti-C1Q or goat anti-C1S that used Tris–EDTA buffer (10mMTris, 1mMEDTA solution, pH 9.0). After cooling down, the slides were rinsed three times withphosphate-buffered saline (PBS) and blocked for 1 h at RT in blocking buffer (1% BSA,0.05% Tween-20 in PBS). Subsequently, sections were incubated at RT overnight withprimary antibodies, followed by 1 h with secondary antibodies, all diluted in blockingbuffer. The primary antibodies used were rabbit anti-KI67 (1:100, ab15580, Abcam),mouse anti-AMH (1:30, MCA2246T, BioRad), rabbit anti-Troponin I (H-170) (1:100,sc15368, Santa Cruz), goat anti-VASA/DDX4 (1:200, AF2030, R D), mouse anti-Cytokeratin (1:100, M351501, DAKO), rabbit anti-Wilm’s Tumor protein (1:100,CA1026-50, Calbiochem), mouse anti-StAR (D-2) (1:100, sc166821, Santa Cruz),rabbit anti-alpha smooth muscle actin/ACTA2 (1:200; ab5694, Abcam), goat anti-ZP3(N-20) (1:100, sc23715, Santa Cruz), mouse anti-Connexin-43/GJA1 (CX-1B1) (1:50,13-8300, Zymed), rabbit anti-Fragilis/IFITM3 (1:200, ab15592, Abcam), rabbit anti-Collagen Type IV (1:50, AB748, Chemicon), rabbit anti-Von Willebrand factor/VWF(1:100, ab6994, Abcam), mouse anti-N-Cadherin/CDH2 (GC-4) (1:100, C3865, Sigma-Aldrich), rabbit anti-PECAM1 (M-20) (1:200, sc1506, Santa Cruz), rabbit anti-c-FOS(1:20, PC38, Calbiochem), mouse anti-8-OHdG (1:1000, sc66036, Santa Cruz), goatanti-IGFBP5 (1:50, AF875, R D), rabbit anti-C1Q (1:400, A0136, DAKO), goat anti-C1S (1:400, A302, Quidel) and mouse anti-CD68 (1:50, M087629-2, DAKO). Thesecondary antibodies used were Alexa Fluor 488 donkey anti-rabbit IgG (1:500, A-21206, Life Technologies), Alexa Fluor 594 donkey anti-mouse IgG (1:500, A-21203,Life Technologies), Alexa Fluor 594 donkey anti-goat IgG (1:500, A11058, LifeTechnologies) and Alexa Fluor 647 donkey anti-goat IgG (1:500, A-21447, LifeTechnologies). Cell death (TUNEL-assay) was detected by In Situ Cell Death DetectionKit (FITC) (11684817910, Sigma-Aldrich) according to the manufacturer’s instruc-tions. Nuclei were stained with 4′,6-diamidino-2-phenyl-indole (DAPI, Life Technol-ogies) and sections mounted using ProLong Gold (Life Technologies).Imaging. Immunostained slides were scanned with Pannoramic 250 Flash IIIdigital scanner (3DHISTECH Ltd., Budapest, Hungary) and representative areaswere selected for imaging using ‘Pannoramic Viewer’(3D HISTECH, Budapest,Hungary) software.Confocal fluorescence images were obtained on a Leica TC SP8 invertedconfocal microscope (Leica) equipped with white light laser and LAS X software(Leica) or an Inverted Leica TC SP5 confocal microscope (Leica) with the LAS AFsoftware (Leica) using a ×40 oil immersion objective (HC PL APO ×40/1.30 OilCS2). Color adjustments was done using Fiji54 and single channel images shownwere converted to an intensity map using Fire lookup table. Figures were assembledusing Adobe Illustrator software (Adobe).C1Q and C3 ELISA on human ovarian tissue. Small pieces (2 × 2 × 2 mm) ofovarian stroma (N=3) and atretic follicle walls (N=6) from one human adult ovarywere cultured individually in 96-wells plates (655180, Cellstar) with 120 µl McCoy’s5A (Modified) Medium (22330-021, Life Technologies) supplemented with 5% fetalbovine serum (FBS, FB1001, Biosera), L-Glutamine (2 mM, 25030081, The rmo Sci-entific) and Penicillin–Streptomycin (50 U/ml, 15070063, Thermo Scientific). After 1and 5 days of culture, medium (30 µl) was collected and stored at −20 °C.The concentration of C1Q and C3 in the culture medium was determined by anin house-made ELISA55. Briefly, Nunc Maxisorp plates (430341, Thermo Scientific)were either coated with a solution (1 µg/ml) of mouse anti-human C1Q56 or rabbitanti-human anti-C3c (A0062, DAKO) diluted in coating buffer (0.1 M Na2CO3, 0.1M NaHCO3) overnight at RT. After three times washing with 0.05% Tween-20 inPBS and blocking with 1% BSA in PBS for 1 h at 37 °C, the culture media diluted inPBT buffer (1% BSA, 0.05% Tween-20 in PBS) was added (1:1 dilution for C1Q and1:12 dilution for C3) to the plates for 1 h at 37 °C. After three times washing with0.05% Tween-20 in PBS, the plates were incubated with either rabbit anti-humanC1q (1:1000, A0136, DAKO) or goat anti-human C3 (1:5000, A213, Complementtechnology), followed by goat anti-rabbit Ig-HRP (1:5000, P0448, DAKO), or rabbitanti-goat Ig-HRP (1:5000, P0449, DAKO), respectively, all diluted in PBT buffer.The enzymatic activity of HRP was measured after incubation with ABTS (A1888-5G, Merck) and H2O2(1072090250, Merck) at absorbance 415 nm on a microplatereader (BioRad). The concentration of C1Q and C3 was determined by comparisonto a dilution standard of normal human serum.Reporting summary. Further information on research design is available inthe Nature Research Reporting Summary linked to this article.Data availabilityRNA-sequencing data are deposited in Gene Expression Omnibus (GEO) with accessionnumber GSE118127 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE118127).Code availabilityTo ensure the reproducibility of the analysis, Conda package (version 4.6.14) managementand environment management system was used (https://conda.io/docs). The workflowused for this analysis is written in R and the source code is available at github, togetherwith the Conda environment profile (https://github.com/johnmous/singleCell).Received: 10 October 2018 Accepted: 13 June 2019References1. de Mello Bianchi, P. H. et al. Review: follicular waves in the human ovary: anew physiological paradigm for novel ovarian stimulation protocols. Reprod.Sci. 17, 1067–1076 (2010).2. Gougeon, A. Regulation of ovarian follicular development in primates: factsand hypotheses. Endocr. Rev. 17, 121–155 (1996).3. McGee, E. A. Hsueh, A. J. Initial and cyclic recruitment of ovarian follicles.Endocr. Rev. 21, 200–214 (2000).4. Gougeon, A. Human ovarian follicular development: from activation of restingfollicles to preovulatory maturation. Ann. Endocrinol. 71, 132–143 (2010).5. Hussein, M. R. Apoptosis in the ovary: molecular mechanisms. Hum. Reprod.Update 11, 162–177 (2005).6. Pache, T. D., Wladimiroff, J. W., de Jong, F. H., Hop, W. C. Fauser, B. C.Growth patterns of nondominant ovarian follicles during the normalmenstrual cycle. Fertil. Steril. 54, 638–642 (1990).7. Burnik Papler, T., Vrtacnik Bokal, E., Maver, A., Kopitar, A. N. Lovrecic, L.Transcriptomic analysis and meta-analysis of human granulosa and cumuluscells. PLoS ONE 10, e0136473 (2015).8. Grondahl, M. L. et al. Specific genes are selectively expressed between cumulusand granulosa cells from individual human pre-ovulatory follicles. Mol. Hum.Reprod. 18, 572–584 (2012).9. Agarwal, A., Aponte-Mellado, A., Premkumar, B. J., Shaman, A. Gupta, S.The effects of oxidative stress on female reproduction: a review. Reprod. Biol.Endocrinol. 10, 49 (2012).10. Pelosi, E., Forabosco, A. Schlessinger, D. Genetics of the ovarian reserve.Front. Genet. 6, 308 (2015).11. Richards, J. S., Ren, Y. A., Candelaria, N., Adams, J. E. Rajkovic, A. Ovarianfollicular theca cell recruitment, differentiation, and impact on fertility: 2017update. Endocr. Rev. 39,1–20 (2018).12. Hsueh, A. J., Billig, H. Tsafriri, A. Ovarian follicle atresia: a hormonallycontrolled apoptotic process. Endocr. Rev. 15, 707–724 (1994).13. Tajima, K., Orisaka, M., Mori, T. Kotsuji, F. Ovarian theca cells in follicularfunction. Reprod. Biomed. Online 15, 591–609 (2007).ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-912 NATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunicationsContent courtesy of Springer Nature, terms of use apply. Rights reserved 14. Butler, A., Hoffman, P., Smibert, P., Papalexi, E. Satija, R. Integrating single-cell transcriptomic data across different conditions, technologies, and species.Nat. Biotechnol. 36, 411–420 (2018).15. van den Brink, S. C. et al. Single-cell sequencing reveals dissociation-inducedgene expression in tissue subpopulations. Nat. Methods 14, 935–936 (2017).16. Liu, C., Peng, J., Matzuk, M. M. Yao, H. H. Lineage specification of ovariantheca cells requires multicellular interactions via oocyte and granulosa cells.Nat. Commun. 6, 6934 (2015).17. Gaytan, F., Morales, C., Roa, J. Tena-Sempere, M. Changes in keratin 8/18expression in human granulosa cell lineage are associated to cell death/survivalevents: potential implications for the maintenance of the ovarian reserve.Hum. Reprod. 33, 680–689 (2018).18. Wiesen, J. F. Midgley, A. R. Jr. Expression of connexin 43 gap junctionmessenger ribonucleic acid and protein during follicular atresia. Biol. Reprod.50, 336–348 (1994).19. Potter, S. J. DeFalco, T. Role of the testis interstitial compartment inspermatogonial stem cell function. Reproduction 153, R151–R162 (2017).20. Hatzirodos, N. et al. Transcriptome profiling of granulosa cells from bovineovarian follicles during atresia. BMC Genom. 15, 40 (2014).21. Yoo, S. W. et al. Complement factors are secreted in human follicular fluid bygranulosa cells and are possible oocyte maturation factors. J. Obstet. Gynaecol.Res. 39, 522–527 (2013).22. Thielens, N. M., Tedesco, F., Bohlson, S. S., Gaboriaud, C. Tenner, A. J. C1q:A fresh look upon an old molecule. Mol. Immunol. 89,73–83 (2017).23. Li, K., Sacks, S. H. Zhou, W. The relative importance of local and systemiccomplement production in ischaemia, transplantation and other pathologies.Mol. Immunol. 44, 3866–3874 (2007).24. Li, L. et al. Single-cell RNA-Seq analysis maps development of human germlinecells and gonadal niche interactions. Cell Stem Cell 20, 858–873 e854 (2017).25. Vertesy, A. et al. Parental haplotype-specific single-cell transcriptomics revealincomplete epigenetic reprogramming in human female germ cells. Nat.Commun. 9, 1873 (2018).26. Liu, Q. et al. Single-cell analysis of differences in transcriptomic profiles ofoocytes and cumulus cells at GV, MI, MII stages from PCOS patients. Sci. Rep.6, 39638 (2016).27. Ruebel, M. L. et al. Obesity modulates inflammation and lipid metabolismoocyte gene expression: a single-cell transcriptome perspective. J. Clin.Endocrinol. Metab. 102, 2029–2038 (2017).28. Xue, Z. et al. Genetic programs in human and mouse early embryos revealedby single-cell RNA sequencing. Nature 500, 593–597 (2013).29. Yan, L. et al. Single-cell RNA-Seq profiling of human preimplantation embryosand embryonic stem cells. Nat. Struct. Mol. Biol. 20, 1131–1139 (2013).30. Zhang, Y. et al. Transcriptome landscape of human folliculogenesis revealsoocyte and granulosa cell interactions. Mol. Cell 72, 1021–1034 e1024 (2018).31. Zhao, H. et al. Single-cell transcriptomics of human oocytes: environment-driven metabolic competition and compensatory mechanisms during oocytematuration. Antioxid. Redox Signal. 30, 542–559 (2019).32. Field, S. L., Dasgupta, T., Cummings, M. Orsi, N. M. Cytokines in ovarianfolliculogenesis, oocyte maturation and luteinisation. Mol. Reprod. Dev. 81,284–314 (2014).33. Poulsen, L. C. et al. Human granulosa cells function as innate immune cellsexecuting an inflammatory reaction during ovulation: a microarray analysis.Mol. Cell. Endocrinol. 486,34–46 (2019).34. Hummitzsch, K. et al. Stem cells, progenitor cells, and lineage decisions in theovary. Endocr. Rev. 36,65–91 (2015).35. Sanchez, F. Smitz, J. Molecular control of oogenesis. Biochim. Biophys. Acta1822, 1896–1912 (2012).36. Hatzirodos, N., Hummitzsch, K., Irving-Rodgers, H. F. Rodgers, R. J.Transcriptome comparisons identify new cell markers for theca interna andgranulosa cells from small and large antral ovarian follicles. PLoS ONE 10,e0119800 (2015).37. Brazdova, A., Senechal, H., Peltre, G. Poncet, P. Immune aspects of femaleinfertility. Int. J. Fertil. Steril. 10,1–10 (2016).38. Ebrahimi, M. Akbari Asbagh, F. The role of autoimmunity in prematureovarian failure. Iran. J. Reprod. Med. 13, 461–472 (2015).39. Kabut, J., Kondera-Anasz, Z., Sikora, J. Mielczarek-Palacz, A. Levels ofcomplement components iC3b, C3c, C4, and SC5b-9 in peritoneal fluid andserum of infertile women with endometriosis. Fertil. Steril. 88,1298–1303 (2007).40. Grumach, A. S. Kirschfink, M. Are complement deficiencies really rare?Overview on prevalence, clinical importance and modern diagnosticapproach. Mol. Immunol. 61, 110–117 (2014).41. Stegert, M., Bock, M. Trendelenburg, M. Clinical presentation of humanC1q deficiency: How much of a lupus? Mol. Immunol. 67,3–11 (2015).42. Hickman, R. A. Gordon, C. Causes and management of infertility insystemic lupus erythematosus. Rheumatology 50, 1551–1558 (2011).43. Jatzko, B. Ott, J. Circulating inflammatory markers in polycystic ovarysyndrome: a systematic review and meta-analysis. Fertil. Steril. 96, e158 (2011).44. Zafari Zangeneh, F., Naghizadeh, M. M. Masoumi, M. Polycystic ovarysyndrome and circulating inflammatory markers. Int. J. Reprod. Biomed. 15,375–382 (2017).45. Hochane, M. et al. Single-cell transcriptomics reveals gene expressiondynamics of human fetal kidney development. PLoS Biol. 17, e3000152 (2019).46. Kowalczyk, M. S. et al. Single-cell RNA-seq reveals changes in cell cycle anddifferentiation programs upon aging of hematopoietic stem cells. Genome Res.25, 1860–1872 (2015).47. Haghverdi, L., Lun, A. T. L., Morgan, M. D. Marioni, J. C. Batch effects insingle-cell RNA-sequencing data are corrected by matching mutual nearestneighbors. Nat. Biotechnol. 36, 421–427 (2018).48. Alexa, A. Rahnenfuhrer, J. topGO: enrichment analysis for gene ontology. Rpackage version 2.28.0. https://rdrr.io/bioc/topGO/ (2016).49. Carlson, M. Org.Hs.eg.db: Genome wide annotation for human. R packageversion 3.4.1. (2017).50. Gentleman, R., Carey, V., Huber, W. Hahne, F. Genefilter: methods forfiltering genes from high-throughput experiments. R package version 1.60.0.https://rdrr.io/bioc/genefilter/ (2017).51. Warnes G. R., et al. gplots: Various R programming tools for plotting data. Rpackage version 3.0.1. https://CRAN.R-project.org/package=gplots (2016).52. Qiu, X. et al. Reversed graph embedding resolves complex single-celltrajectories. Nat. Methods 14, 979–982 (2017).53. Haghverdi, L., Buttner, M., Wolf, F. A., Buettner, F. Theis, F. J. Diffusionpseudotime robustly reconstructs lineage branching. Nat. Methods 13,845–848 (2016).54. Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis.Nat. Methods 9, 676–682 (2012).55. Lubbers, R. et al. Complement component C1q as serum biomarker to detectactive tuberculosis. Front. Immunol. 9, 2427 (2018).56. van Schaarenburg, R. A. et al. C1q deficiency and neuropsychiatric systemiclupus erythematosus. Front. Immunol. 7, 647 (2016).AcknowledgementsWe would like to acknowledge the patients that contributed with the ovarian material used inthis study, the Leiden Genome Technology Center for sequencing services and A. Mahfouzfor bioinformatics advice. This study was funded by the Europe Research Council Con-solidator Grant OVOGROWTH (ERC-CoG-2016-725722) to E.L., I.M., V.T.J. and S.M.C.d.S.L.; Europe Research Council Consolidator Grant AUTOCOMPLEMENT (ERC-CoG-2016-724517) to N.V.B. and L.T. and the China Scholarship Council (CSC 201706320328) to X.F.Author contributionsM.B., L.T., L.v.d.W. and S.M.C.d.S.L. design the study; L.L., G.S.K.P. and L.v.d.W. iso-lated tissue samples; M.B., N.V.B., E.L., L.T. and X.F. performed experiments; I.M., V.T.J.and H.M. are responsible for the bioinformatics; all authors discussed results, wrote themanuscript and approved the final version.Additional informationSupplementary Information accompanies this paper at https://doi.org/10.1038/s41467-019-11036-9.Competing interests: The authors declare no competing interests.Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/Peer review information: Nature Communications thanks Katsuhiko Hayashi and otheranonymous reviewer(s) for their contribution to the peer review of this work.Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Open Access This article is licensed under a Creative CommonsAttribution 4.0 International License, which permits use, sharing,adaptation, distribution and reproduction in any medium or format, as long as you giveappropriate credit to the original author(s) and the source, provide a link to the CreativeCommons license, and indicate if changes were made. The images or other third partymaterial in this article are included in the article’s Creative Commons license, unlessindicated otherwise in a credit line to the material. If material is not included in thearticle’s Creative Commons license and your intended use is not permitted by statutoryregulation or exceeds the permitted use, you will need to obtain permission directly fromthe copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.© The Author(s) 2019NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-019-11036-9 ARTICLENATURE COMMUNICATIONS | (2019) 10:3164 | https://doi.org/10.1038/s41467-019-11036-9 | www.nature.com/naturecommunications 13Content courtesy of Springer Nature, terms of use apply. Rights reserved 1.2.3.4.5.6.Terms and Conditions Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (\"Springer Nature”). Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (\"Users”), for small-scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. Byaccessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (\"Terms”). For thesepurposes, Springer Nature considers academic use (by researchers and students) to be non-commercial. These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personalsubscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used willapply. We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally withinResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will nototherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission asdetailed in the Privacy Policy. While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users maynot: use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent accesscontrol;use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or isotherwise unlawful;falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature inwriting;use bots or other automated methods to access the content or redirect messagesoverride any security feature or exclusionary protocol; orshare the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journalcontent. In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journalcontent cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or anyother, institutional repository. These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information orcontent on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Naturemay revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved. To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or impliedwith respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,including merchantability or fitness for any particular purpose. Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensedfrom third parties. If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner notexpressly permitted by these Terms, please contact Springer Nature at onlineservice@springernature.com Available via license: CC BYContent may be subject to copyright.Citations (45)References (60)... As above, we determined the fraction of correctly annotated clusters and the AUCRanks 1-5, and we compared these values to those obtained when scALE was applied to the tuning studies. The testing studies are provided in Table 2 [71][72][73][74][75][76][77][78][79]. Table 2. List of human studies that were used to test the cluster annotation algorithm. ...... Table 2. List of human studies that were used to test the cluster annotation algorithm. The columns indicate (1) the tissue from which cells were derived in the study, (2) the title of the published study or dataset, and (3) the reference for the study [71][72][73][74][75][76][77][78][79]. ...... To identify potential novel or poorly characterized markers of well-studied cell types, we compared the mean expression (CP10K) of all genes in the defined cell type of interest to their mean expression in all other cells from our reference dataset (see File S8). The cell types considered here included retinal pigment epithelial cells derived from two independent human studies [30,68] and endothelial cells derived from 31 human studies [26][27][28]30,33,[35][36][37][38][39][40]42,43,[46][47][48][49][50][51][52][53]55,56,62,64,67,[72][73][74][75][76]78,79]. Specifically, we calculated the FC value for each gene as described above in the pan-study method for CDG identification during the cell type annotation algorithm. ...A Literature-Derived Knowledge Graph Augments the Interpretation of Single Cell RNA-seq DatasetsArticleFull-text availableJun 2021 Deeksha DoddahonnaiahPatrick J LenehanTravis K. HughesVenky SoundararajanTechnology to generate single cell RNA-sequencing (scRNA-seq) datasets and tools to annotate them have advanced rapidly in the past several years. Such tools generally rely on existing transcriptomic datasets or curated databases of cell type defining genes, while the application of scalable natural language processing (NLP) methods to enhance analysis workflows has not been adequately explored. Here we deployed an NLP framework to objectively quantify associations between a comprehensive set of over 20,000 human protein-coding genes and over 500 cell type terms across over 26 million biomedical documents. The resultant gene-cell type associations (GCAs) are significantly stronger between a curated set of matched cell type-marker pairs than the complementary set of mismatched pairs (Mann Whitney p = 6.15 × 10−76, r = 0.24; cohen’s D = 2.6). Building on this, we developed an augmented annotation algorithm (single cell Annotation via Literature Encoding, or scALE) that leverages GCAs to categorize cell clusters identified in scRNA-seq datasets, and we tested its ability to predict the cellular identity of 133 clusters from nine datasets of human breast, colon, heart, joint, ovary, prostate, skin, and small intestine tissues. With the optimized settings, the true cellular identity matched the top prediction in 59% of tested clusters and was present among the top five predictions for 91% of clusters. scALE slightly outperformed an existing method for reference data driven automated cluster annotation, and we demonstrate that integration of scALE can meaningfully improve the annotations derived from such methods. Further, contextualization of differential expression analyses with these GCAs highlights poorly characterized markers of well-studied cell types, such as CLIC6 and DNASE1L3 in retinal pigment epithelial cells and endothelial cells, respectively. Taken together, this study illustrates for the first time how the systematic application of a literature-derived knowledge graph can expedite and enhance the annotation and interpretation of scRNA-seq data.ViewShow abstract... Another single-cell RNA sequencing study of the inner cortex of adult ovaries was recently performed [39]. As a result, different populations of GCs were distinguished based on specific gene expression. ...... Mural GCs expressed high levels of KRT18 (keratin 18), CITED2 (CBP/p300-interacting transactivator 2) and AKIRIN1 and low levels of WT1 and EGR4. In contrast, GCs from atretic follicles did not express VCAN, FST or KRT18 and expressed lower levels of GJA1 (gap junction protein alpha 1) and CDH2 (cadherin 2) compared to other clusters of GCs [39]. However, other cells with stem-like properties were found to reside in the ovary. ...Human Granulosa Cells—Stemness Properties, Molecular Cross-Talk and Follicular AngiogenesisArticleFull-text availableJun 2021 Claudia Dompe Magdalena Kulus Katarzyna Stefańska Bartosz KempistyThe ovarian follicle is the basic functional unit of the ovary, comprising theca cells and granulosa cells (GCs). Two different types of GCs, mural GCs and cumulus cells (CCs), serve different functions during folliculogenesis. Mural GCs produce oestrogen during the follicular phase and progesterone after ovulation, while CCs surround the oocyte tightly and form the cumulus oophurus and corona radiata inner cell layer. CCs are also engaged in bi-directional metabolite exchange with the oocyte, as they form gap-junctions, which are crucial for both the oocyte’s proper maturation and GC proliferation. However, the function of both GCs and CCs is dependent on proper follicular angiogenesis. Aside from participating in complex molecular interplay with the oocyte, the ovarian follicular cells exhibit stem-like properties, characteristic of mesenchymal stem cells (MSCs). Both GCs and CCs remain under the influence of various miRNAs, and some of them may contribute to polycystic ovary syndrome (PCOS) or premature ovarian insufficiency (POI) occurrence. Considering increasing female fertility problems worldwide, it is of interest to develop new strategies enhancing assisted reproductive techniques. Therefore, it is important to carefully consider GCs as ovarian stem cells in terms of the cellular features and molecular pathways involved in their development and interactions as well as outline their possible application in translational medicine.ViewShow abstract... This angiogenic mechanism not only allows to grow new vessels via duplication (intussusceptive microvascular growth) but also to remodel the vascular tree via arborization (intussusceptive arborization) and pruning (intussusceptive pruning), contributing to the control of the vascular tree geometry [49]. Angiogenesis in healthy adult organs is rare, with a few exceptions (skeletal muscle angiogenesis during physical exercise; endometrial angiogenesis during epithelial regeneration [50,51]). However, upon injury, adult ECs can rapidly grow new vessels via re-activation of developmental angiogenic programs (SA or IA). ...Endothelial cell plasticity at the single-cell levelArticleFull-text availableMay 2021AngiogenesisAlessandra PasutLisa M. BeckerAnne CuypersPeter CarmelietThe vascular endothelium is characterized by a remarkable level of plasticity, which is the driving force not only of physiological repair/remodeling of adult tissues but also of pathological angiogenesis. The resulting heterogeneity of endothelial cells (ECs) makes targeting the endothelium challenging, no less because many EC phenotypes are yet to be identified and functionally inventorized. Efforts to map the vasculature at the single-cell level have been instrumental to capture the diversity of EC types and states at a remarkable depth in both normal and pathological states. Here, we discuss new EC subtypes and functions emerging from recent single-cell studies in health and disease. Interestingly, such studies revealed distinct metabolic gene signatures in different EC phenotypes, which deserve further consideration for therapy. We highlight how this metabolic targeting strategy could potentially be used to promote (for tissue repair) or block (in tumor) angiogenesis in a tissue or even vascular bed-specific manner.ViewShow abstract... Studies of region-specific epididymal proteins showed that certain cell types were able to express different classifications of genes, which contribute to the different physiologic functions of the segments 2,14 . With the development of single-cell RNA sequencing (scRNAseq), a number of organs have been analyzed in mammals [15][16][17] , including male and female reproductive organs such as the testis [18][19][20] and ovary 21,22 . The spatio-temporal repertoire of epididymal cells and their gene expression in the epididymis are still less characterized. ...Spatio-temporal landscape of mouse epididymal cells and specific mitochondria-rich segments defined by large-scale single-cell RNA-seqArticleFull-text availableMay 2021Jianwu Shi Ellis Kin Lam FokPengyuan Dai Hao ChenSpermatozoa acquire their fertilizing ability and forward motility during epididymal transit, suggesting the importance of the epididymis. Although the cell atlas of the epididymis was reported recently, the heterogeneity of the cells and the gene expression profile in the epididymal tube are still largely unknown. Considering single-cell RNA sequencing results, we thoroughly studied the cell composition, spatio-temporal differences in differentially expressed genes (DEGs) in epididymal segments and mitochondria throughout the epididymis with sufficient cell numbers. In total, 40,623 cells were detected and further clustered into 8 identified cell populations. Focused analyses revealed the subpopulations of principal cells, basal cells, clear/narrow cells, and halo/T cells. Notably, two subtypes of principal cells, the Prc7 and Prc8 subpopulations were enriched as stereocilia-like cells according to GO analysis. Further analysis demonstrated the spatially specific pattern of the DEGs in each cell cluster. Unexpectedly, the abundance of mitochondria and mitochondrial transcription (MT) was found to be higher in the corpus and cauda epididymis than in the caput epididymis by scRNA-seq, immunostaining, and qPCR validation. In addition, the spatio-temporal profile of the DEGs from the P42 and P56 epididymis, including transiting spermatozoa, was depicted. Overall, our study presented the single-cell transcriptome atlas of the mouse epididymis and revealed the novel distribution pattern of mitochondria and key genes that may be linked to sperm functionalities in the first wave and subsequent wave of sperm, providing a roadmap to be emulated in efforts to achieve sperm maturation regulation in the epididymis.ViewShow abstract... It has been speculated many times that oogonial stem cells ( cortical reserve ) may exist, but this hypothesis remains controversial and heavily debated. Cells expressing extracellular DEADbox polypeptide 4 (ecDDX4) have been advanced as stem cell candidates, but recent single-cell omics identified these ecDDX4+ cells as perivascular cells (Fan et al., 2019;Wagner et al., 2020). To date, the Zuckerman axiom that a fixed number of oocytes is present and available throughout a woman s lifetime still stands (Zuckerman, 1951). ...Organoids of the Female Reproductive Tract: Innovative Tools to Study Desired to Unwelcome ProcessesArticleFull-text availableApr 2021 Ruben HeremansZiga Jan Dirk Timmerman Hugo VankelecomThe pelviperineal organs of the female reproductive tract form an essential cornerstone of human procreation. The system comprises the ectodermal external genitalia, the Müllerian upper-vaginal, cervical, endometrial and oviductal derivatives, and the endodermal ovaries. Each of these organs presents with a unique course of biological development as well as of malignant degeneration. For many decades, various preclinical in vitro models have been employed to study female reproductive organ (patho-)biology, however, facing important shortcomings of limited expandability, loss of representativeness and inadequate translatability to the clinic. The recent emergence of 3D organoid models has propelled the field forward by generating powerful research tools that in vitro replicate healthy as well as diseased human tissues and are amenable to state-of-the-art experimental interventions. Here, we in detail review organoid modeling of the different female reproductive organs from healthy and tumorigenic backgrounds, and project perspectives for both scientists and clinicians.ViewShow abstractMonocyte perturbation modulates the ovarian response to an immune challengeArticleJul 2021MOL CELL ENDOCRINOL Simin Younesi Sarah J Spencer Luba SominskyOur recent findings indicate that an acute depletion of monocytes has no sustained effects on ovarian follicle health. Here, we utilised a Cx3cr1-Dtr transgenic Wistar rat model to transiently deplete monocytes and investigated the impact of an acute immune challenge by lipopolysaccharide (LPS) on ovarian follicle health and ovulatory capacity relative to wt once the monocytes had repopulated. Monocyte depletion and repopulation exacerbated the effects of LPS in several domains. As such, monocyte perturbation decreased the numbers of secondary follicles in those challenged with LPS. Monocyte perturbation was also associated with reduced antral follicle numbers and circulating luteinising hormone (LH) levels, as well as potential changes in ovarian sensitivity to LH, exacerbated by LPS. These data suggest that monocyte depletion and repopulation induce a transient suppression of ovulatory capacity in response to a subsequent immune challenge, but this is likely to be restored once the pro-inflammatory environment is resolved.ViewShow abstractThe making of an ovarian nicheArticleJul 2021SCIENCELin YangHuck-Hui NgViewTemporal transcriptomic landscape of postnatal mouse ovaries reveals dynamic gene signatures associated with ovarian agingArticleFull-text availableJun 2021HUM MOL GENET Zixue ZhouXi YangYuncheng Pan Yanhua WuThe ovary is the most important organ for maintaining female reproductive health, but it fails before most other organs. Aging-associated alterations in gene expression patterns in mammalian ovaries remain largely unknown. In this study, the transcriptomic landscape of postnatal mouse ovaries over the reproductive lifespan was investigated using bulk RNA sequencing in C57BL/6 mice. Gene expression dynamics revealed that the lifespan of postnatal mouse ovaries comprised four sequential stages, during which 2517 genes were identified as differentially enriched. Notably, the DNA repair pathway was found to make a considerable and specific contribution to the process of ovarian aging. Temporal gene expression patterns were dissected to identify differences in gene expression trajectories over the lifespan. In addition to DNA repair, distinct biological functions (including hypoxia response, epigenetic modification, fertilization, mitochondrial function, etc.) were overrepresented in particular clusters. Association studies were further performed to explore the relationships between known genes responsible for ovarian function and differentially expressed genes identified in this work. We found that the causative genes of human premature ovarian insufficiency were specifically enriched in distinct gene clusters. Taken together, our findings reveal a comprehensive transcriptomic landscape of the mouse ovary over the lifespan, providing insights into the molecular mechanisms underlying mammalian ovarian aging and supporting future etiological studies of aging-associated ovarian disorders.ViewShow abstractDecoding dynamic epigenetic landscapes in human oocytes using single-cell multi-omics sequencingArticleApr 2021Rui YanChan GuDi YouFan GuoDeveloping female human germ cells undergo genome-wide epigenetic reprogramming, but de novo DNA methylation dynamics and their interplay with chromatin states and transcriptional activation in developing oocytes is poorly understood. Here, we developed a single-cell multi-omics sequencing method, scChaRM-seq, that enables simultaneous profiling of the DNA methylome, transcriptome, and chromatin accessibility in single human oocytes and ovarian somatic cells. We observed a global increase in DNA methylation during human oocyte growth that correlates with chromatin accessibility, whereas increases of DNA methylation at specific features were associated with active transcription. Integrated analyses of multi-omics data from humans and mice revealed species-specific gene expression, and promoter accessibility contributes to gene body methylation programs. Alu elements retained low DNA methylation levels and high accessibility in early growing oocytes and were located near developmental genes in humans and mice. Together, these findings show how scChaRM-seq can provide insight into DNA methylation pattern establishment.ViewShow abstractMolecular make-up of the human adult ovaryArticleApr 2021Xueying FAN Susana M Chuva de Sousa LopesFunctional ovarian cells are essential for human fertility. In the adult ovary, different cell types ensure ovary homeostasis, enable hormonal production and support oocyte maturation. Hence, the ovary is a complex and highly dynamic organ composed of a great diversity of cell types, many still uncharacterized. The use of single-cell RNA sequencing technologies on human ovarian tissue is starting to unravel the molecular signature of the cells present in the ovary, highlighting dramatic changes in gene expression during follicular growth and regression. This knowledge will ultimately provide insights in female fertility and associated-reproductive diseases and will allow the optimization of humans-based disease models and in vitro gametogenesis protocols.ViewShow abstractShow moreSingle-cell transcriptomics reveals gene expression dynamics of human fetal kidney developmentArticleFull-text availableFeb 2019 Mazène Hochane Patrick Robert Van den Berg Xueying Fan Stefan SemrauThe current understanding of mammalian kidney development is largely based on mouse models. Recent landmark studies revealed pervasive differences in renal embryogenesis between mouse and human. The scarcity of detailed gene expression data in humans therefore hampers a thorough understanding of human kidney development and the possible developmental origin of kidney diseases. In this paper, we present a single-cell transcriptomics study of the human fetal kidney. We identified 22 cell types and a host of marker genes. Comparison of samples from different developmental ages revealed continuous gene expression changes in podocytes. To demonstrate the usefulness of our data set, we explored the heterogeneity of the nephrogenic niche, localized podocyte precursors, and confirmed disease-associated marker genes. With close to 18,000 renal cells from five different developmental ages, this study provides a rich resource for the elucidation of human kidney development, easily accessible through an interactive web application.ViewShow abstractTranscriptome Landscape of Human Folliculogenesis Reveals Oocyte and Granulosa Cell InteractionsArticleFull-text availableNov 2018MOL CELLYaoyao Zhang Zhiqiang YanQingyuan QinLiying YanThe dynamic transcriptional regulation and interactions of human germlines and surrounding somatic cells during folliculogenesis remain unknown. Using RNA sequencing (RNA-seq) analysis of human oocytes and corresponding granulosa cells (GCs) spanning five follicular stages, we revealed unique features in transcriptional machinery, transcription factor networks, and reciprocal interactions in human oocytes and GCs that displayed developmental-stage-specific expression patterns. Notably, we identified specific gene signatures of two cell types in particular developmental stage that may reflect developmental competency and ovarian reserve. Additionally, we uncovered key pathways that may concert germline-somatic interactions and drive the transition of primordial-to-primary follicle, which represents follicle activation. Thus, our work provides key insights into the crucial features of the transcriptional regulation in the stepwise folliculogenesis and offers important clues for improving follicle recruitment in vivo and restoring fully competent oocytes in vitro.ViewShow abstractComplement Component C1q as Serum Biomarker to Detect Active TuberculosisArticleFull-text availableOct 2018 Rosalie Lubbers Jayne S Sutherland Delia Goletti Leendert A TrouwBackground: Tuberculosis (TB) remains a major threat to global health. Currently, diagnosis of active TB is hampered by the lack of specific biomarkers that discriminate active TB disease from other (lung) diseases or latent TB infection (LTBI). Integrated human gene expression results have shown that genes encoding complement components, in particular different C1q chains, were expressed at higher levels in active TB compared to LTBI.Methods: C1q protein levels were determined using ELISA in sera from patients, from geographically distinct populations, with active TB, LTBI as well as disease controls.Results: Serum levels of C1q were increased in active TB compared to LTBI in four independent cohorts with an AUC of 0.77 [0.70; 0.83]. After 6 months of TB treatment, levels of C1q were similar to those of endemic controls, indicating an association with disease rather than individual genetic predisposition. Importantly, C1q levels in sera of TB patients were significantly higher as compared to patients with sarcoidosis or pneumonia, clinically important differential diagnoses. Moreover, exposure to other mycobacteria, such as Mycobacterium leprae (leprosy patients) or BCG (vaccinees) did not result in elevated levels of serum C1q. In agreement with the human data, in non-human primates challenged with Mycobacterium tuberculosis, increased serum C1q levels were detected in animals that developed progressive disease, not in those that controlled the infection.Conclusions: In summary, C1q levels are elevated in patients with active TB compared to LTBI in four independent cohorts. Furthermore, C1q levels from patients with TB were also elevated compared to patients with sarcoidosis, leprosy and pneumonia. Additionally, also in NHP we observed increased C1q levels in animals with active progressive TB, both in serum and in broncho-alveolar lavage. Therefore, we propose that the addition of C1q to current biomarker panels may provide added value in the diagnosis of active TB.ViewShow abstractParental haplotype-specific single-cell transcriptomics reveal incomplete epigenetic reprogramming in human female germ cellsArticleFull-text availableMay 2018Ábel VértesyWibowo Arindrarto Matthias S. Roost Susana M Chuva de Sousa LopesIn contrast to mouse, human female germ cells develop asynchronously. Germ cells transition to meiosis, erase genomic imprints, and reactivate the X chromosome. It is unknown if these events all appear asynchronously, and how they relate to each other. Here we combine exome sequencing of human fetal and maternal tissues with single-cell RNA-sequencing of five donors. We reconstruct full parental haplotypes and quantify changes in parental allele-specific expression, genome-wide. First we distinguish primordial germ cells (PGC), pre-meiotic, and meiotic transcriptional stages. Next we demonstrate that germ cells from various stages monoallelically express imprinted genes and confirm this by methylation patterns. Finally, we show that roughly 30% of the PGCs are still reactivating their inactive X chromosome and that this is related to transcriptional stage rather than fetal age. Altogether, we uncover the complexity and cell-to-cell heterogeneity of transcriptional and epigenetic remodeling in female human germ cells.ViewShow abstractIntegrating single-cell transcriptomic data across different conditions, technologies, and speciesArticleFull-text availableMay 2018Nat BiotechnolAndrew ButlerPaul Hoffman Peter SmibertRahul SatijaComputational single-cell RNA-seq (scRNA-seq) methods have been successfully applied to experiments representing a single condition, technology, or species to discover and define cellular phenotypes. However, identifying subpopulations of cells that are present across multiple data sets remains challenging. Here, we introduce an analytical strategy for integrating scRNA-seq data sets based on common sources of variation, enabling the identification of shared populations across data sets and downstream comparative analysis. We apply this approach, implemented in our R toolkit Seurat (http://satijalab.org/seurat/), to align scRNA-seq data sets of peripheral blood mononuclear cells under resting and stimulated conditions, hematopoietic progenitors sequenced using two profiling technologies, and pancreatic cell atlases generated from human and mouse islets. In each case, we learn distinct or transitional cell states jointly across data sets, while boosting statistical power through integrated analysis. Our approach facilitates general comparisons of scRNA-seq data sets, potentially deepening our understanding of how distinct cell states respond to perturbation, disease, and evolution.ViewShow abstractBatch effects in single-cell RNA-sequencing data are corrected by matching mutual nearest neighborsArticleFull-text availableMay 2018Nat BiotechnolLaleh Haghverdi Aaron T L Lun Michael D MorganJohn C. MarioniLarge-scale single-cell RNA sequencing (scRNA-seq) data sets that are produced in different laboratories and at different times contain batch effects that may compromise the integration and interpretation of the data. Existing scRNA-seq analysis methods incorrectly assume that the composition of cell populations is either known or identical across batches. We present a strategy for batch correction based on the detection of mutual nearest neighbors (MNNs) in the high-dimensional expression space. Our approach does not rely on predefined or equal population compositions across batches; instead, it requires only that a subset of the population be shared between batches. We demonstrate the superiority of our approach compared with existing methods by using both simulated and real scRNA-seq data sets. Using multiple droplet-based scRNA-seq data sets, we demonstrate that our MNN batch-effect-correction method can be scaled to large numbers of cells.ViewShow abstractSingle-Cell Transcriptomics of Human Oocytes: Environment-Driven Metabolic Competition and Compensatory Mechanisms During Oocyte MaturationArticleFull-text availableFeb 2018ANTIOXID REDOX SIGNHongcui ZhaoTianjie Li Yue Zhao Jie QiaoAim: The mechanisms coordinating maturation with an environment-driven metabolic shift, a critical step in determining the developmental potential of human in vitro matured (IVM) oocytes, remains to be elucidated. Here, we explored the key genes regulating human oocyte maturation using single-cell RNA sequencing and illuminated the compensatory mechanism from a metabolic perspective by analyzing gene expression.Results: Three key genes that encode CoA-related enzymes were screened from the RNA sequencing data. Two of them, ACAT1 and HADHA, were closely related to the regulation of substrate production in the Krebs cycle. Dysfunction of the Krebs cycle was induced by decreases in the activity of specific enzymes. Further, the activator of these enzymes, the calcium concentration, was also decreased because of the failure of influx of exogenous calcium. Although release of endogenous calcium from the endoplasmic reticulum and mitochondria met the requirement for maturation, excessive release resulted in aneuploidy and developmental incompetence. High nicotinamide nucleotide transhydrogenase expression induced NADPH dehydrogenation to compensate for the NADH shortage resulting from the dysfunction of the Krebs cycle. Importantly, high NADP+ levels activated DPYD to enhance the repair of DNA double-strand breaks to maintain euploidy.Innovation: The present study shows for the first time that exposure to the in vitro environment can lead to the decline of energy metabolism in human oocytes during maturation but that a compensatory action maintains their developmental competence.Conclusions: In vitro maturation of human oocytes is mediated through a cascade of competing and compensatory actions driven by genes encoding enzymes.ViewShow abstractPolycystic ovary syndrome and circulating inflammatory markersArticleFull-text availableJun 2017IRAN J REPROD MED Farideh Zangeneh Mohammad Mehdi NaghizadehMasoumeh MasoumiBackgroundHuman and experimental studies suggest that the sympathetic regulatory drive in the ovary may be unbalanced (hyperactivity) in polycystic ovary syndrome (PCOS). Dysfunctional secretion of interleukin (IL) -1 (α β) or related cytokines may thus be related to abnormal ovulation and luteinization.ObjectiveThe aim of this study was the evaluation of cytokines’ pattern in PCOS women and discussion about the explanation of cross-talk between two super systems: sympathetic and immune systems and explanation sympatho-excitation and relationship with interleukins.Materials and MethodsIn this study, 171 PCOS women aged between 20-40 years were studied the. Their body mass index was 28. The patients were divided into two groups: study group (n=85, PCOS women) and control group (n=86 normal women). The blood sample was obtained on the 3rd day of menstruation cycle. IL-17, IL-1α, IL-1β, and TNF-α concentrations were determined in both groups.ResultsThe median serum level of IL-1α in the PCOS group was higher than the control group (293.3 and 8.0, respectively, p 0.001). Also, the median serum level of IL-1β was higher than the control group (5.9 and 3.1 respectively). But the median serum of level IL-17 in women with PCOS was significantly lower than the control group (p 0.001).ConclusionOur results confirm that PCOS is a low-level chronic inflammation.ViewShow abstractHuman granulosa cells function as innate immune cells executing an inflammatory reaction during ovulation: A microarray analysisArticleFeb 2019MOL CELL ENDOCRINOLLiv la Cour PoulsenAnne Lis Mikkelsen EnglundMarie Louise Muff WissingMarie Louise GrøndahlOvulation has been compared to a local inflammatory reaction. We performed an in silico study on a unique, PCR validated, transcriptome microarray study to evaluate if known inflammatory mechanisms operate during ovulation. The granulosa cells were obtained in paired samples at two different time points during ovulation (just before and 36 hours after ovulation induction) from nine women receiving fertility treatment. A total of 259 genes related to inflammation became significantly upregulated during ovulation (2–80 fold, p 0.05), while specific leukocyte markers were absent. The genes and pathway analysis indicated NF-KB-, MAPK- and JAK/STAT signalling (p 1.0E-10) as the major pathways involved in danger recognition and cytokine signalling to initiate inflammation. Upregulated genes further encoded enzymes in eicosanoid production, chemo-attractants, coagulation factors, cell proliferation factors involved in tissue repair, and anti-inflammatory factors to resolve the inflammation again. We conclude that granulosa cells, without involvement from the innate immune system, can orchestrate ovulation as a complete sterile inflammatory reaction.ViewShow abstractChanges in keratin 8/18 expression in human granulosa cell lineage are associated to cell death/survival events: Potential implications for the maintenance of the ovarian reserveArticleFeb 2018Hum ReprodF GaytanC Morales Juan Roa Manuel Tena-SempereStudy question: Is keratin 8/18 (K8/K18) expression linked to cell death/survival events in the human granulosa cell lineage?Summary answer: A close association exists between changes in K8/K18 expression and cell death/survival events along the human granulosa cell lineage lifespan.What is known already: In addition to their structural and mechanical functions, K8/K18 play essential roles regulating cell death, survival and differentiation in several non-gonadal epithelial tissues. Transfection of the granulosa-like tumor KGN cells with siRNA to interfere KRT8 and KRT18 expression increases FAS-mediated apoptosis, while an inverse association between K8/K18 expression and cell death has been found in the bovine antral follicles and corpus luteum. Yet, only fragmentary and inconclusive information exists regarding K8/K18 expression in the human ovary.Study design, size, duration: Expression of K8/K18 was assessed by immunohistochemistry at different stages of the granulosa cell lineage, from flattened granulosa cells in primordial follicles to fully luteinized granulosa-lutein cells in the corpus luteum (including corpus luteum of pregnancy).Participants/materials, setting, methods: Immunohistochemical detection of K8/K18 was conducted in 40 archival ovarian samples from women aged 17-39 years. K8/K18 expression was analyzed at the different stages of follicle development and corpus luteum lifespan. The proportions of primordial follicles showing all K8/K18-positive, all K8/K18 negative, or a mixture of K8/K18 negative and positive granulosa cells were quantified in 18 ovaries, divided into three age groups: ≤ 25 years (N = 6), 26-30 (N = 6) and 31-36 (N = 6) years. A total number of 1793 primordial, 750 transitional and 140 primary follicles were scored.Main results and the role of chance: A close association was found between changes in K8/K18 expression and cell death/cell survival events in the human granulosa cell lineage. Large secondary and early antral follicles (most of them undergoing atresia) and regressing corpora lutea displayed low/absent K8/K18 expression. Conversely, early growing and some large antral follicles, functional menstrual corpora lutea, as well as life-extended corpus luteum of pregnancy, in which cell death was scarce, showed high K8/K18 expression. Three sub-populations of primordial follicles were observed with respect to the presence of K8/K18 in their flattened granulosa cells, ranging from primordial follicles showing only positive granulosa cells [P0(+)], to others with a mixture of positive and negative cells [P0(+/-)] or follicles with only negative cells [P0(-)]. Significant age-related changes were found in the proportions of the different primordial follicle types. In relation to age, a positive correlation was found for P0(+) primordial follicles (R2= 0.7883, N = 18; P 0.001), while negative correlations were found for P0(+/-) (R2= 0.6853, N = 18; P 0.001) and P0(-) (R2= 0.6725, N = 18; P 0.001) follicles. Furthermore, an age-related shift towards greater keratin expression was found in P0(+/-) follicles (χ2 = 19.07, P 0.05).Large scale data: Limitations reasons for caution: This is a descriptive study. Hence, a cause-and-effect relationship between K8/K18 expression and cell death/survival cannot be directly established.Wider implications of the findings: This study describes, for the first time, the existence of sub-populations of primordial follicles on the basis of K8/K18 expression in granulosa cells, and that their proportions change with age. While a progressive increase in K8/K18 expression cannot be ruled out, our data are consistent with the hypothesis that primordial follicles expressing low levels of K8/K18 are preferentially ablated by follicle attrition, while primordial follicles showing high K8/K18 levels are those predominantly recruited into the growing pool. This suggests that K8/K18 expression could constitute a novel factor regulating primordial follicle death/survival, and raises the possibility that alterations of K8/K18 expression could be involved in the accelerated depletion of the ovarian reserve leading to premature ovarian insufficiency.Study funding/competing interest(s): This work was supported by Grants BFU2011-025021 and BFU2014-57581-P (Ministerio de Economía y Competitividad, Spain; co-funded with EU funds from FEDER Program); project PIE14-00005 (Flexi-Met, Instituto de Salud Carlos III, Ministerio de Sanidad, Spain); Projects P08-CVI-03788 and P12-FQM-01943 (Junta de Andalucía, Spain); and EU research contract DEER FP7-ENV-2007-1. CIBER Fisiopatología de la Obesidad y Nutrición is an initiative of Instituto de Salud Carlos III. The authors have nothing to disclose in relation to the contents of this study.ViewShow abstractShow moreAdvertisementRecommendationsDiscover moreProjectinflammation Wei Shi Hussain MusaddiqXimei Wu[...] Xueying FanView projectProjectModelling mouse and human embryonic development in vitro with stem cells Susanne C Van den Brink Katharina Sonnen Vincent van Batenburg[...] Ana M Pereira DaoudModelling mouse and human embryonic post-implantation development in vitro with stem cells View projectArticleReproductive biology of the mottled electric ray, Torpedo sinuspersici Olfer, 1831 (Pisces: Torpedin...January 2014 · Indian Journal of FisheriesK.V.L. ShrikanyaK. SujathaAmong the four species of electric rays of the genus Torpedo recorded from the catches of Visakhapatnam, north Andhra region (lat 17 degrees 01 N to 19 degrees 22 long 83 degrees 23 E to 85 degrees 14 E), Torpedo sinuspersici Olfer, 1831 is the most commonly occurring species in trawl and trammel net bycatches. The present paper deals with various aspects of reproductive biology viz., length ... [Show full abstract] at first maturity, size at birth, sex ratio, gestation period and fecundity of the mottled electric ray represented in the catches of Visakhapatnam. The study is based on 200 specimens measuring 118 - 500 mm TL that include 105 males and 95 females. The mode of reproduction in this ray is ovoviviparous. The two ovaries and the uteri are functional. Gestation period is estimated as 6-8 months approximately. Length at first sexual maturity is 325 mm TL for females and 300 mm TL for males. Embryo counts in uteri of pregnant females range from 8-16. Apparently low fecundity indicates that it is potentially vulnerable to overfishing and bycatch rates should be monitored closely.Read moreArticleFull-text availableThe effect of super-oxidized water on the tissues of uterus and ovary: An experimental rat studyJanuary 2017 · Eastern Journal of Medicine Abbas Aras Erbil Karaman Numan Cim[...]Ö. YılmazSuper-oxidized solutions are known to be potent disinfectants for external surfaces and also for wound care. There are limited data about the use of superoxidized water in the intraperitoneal organs. The aim of the present study was to evalaute its effect on the uterus and ovary when applied via intraperitoneal infusion in a rat model. Thirty Wistar-Albino rats weighing 250-300 g were randomly ... [Show full abstract] divided into three groups (10 rats/group). Group1(control group) rats received single dose of 10 mg/kg saline solution intraperitoneally. Group 2(single dose group) rats received single dose of 10 mg/kg pH-neutral SOW intraperitoneally. Group 3(multiple doses group) rats received multiple doses of 10 mg/kg pH-neutral SOW intraperitoneall at first, third and fifth days. All animals were sacrificed at one week after infusion. The macro- and microscopic histopathological examinations were performed for each rat. All rats remained healthy follow up of one week. The macroscopic examinations of the three groups showed no significant differences. No toxicity findings were found in three groups. The microscopic examinations revealed active endometial glandular structures in uterus and functional follicules at different stages of maturation in ovary. There were no significant differences with regards to the microscopic findings between three groups. Intraperitoneal infusion of pH-neutral SOW does not result in any significant toxicity and complications on the tissues of uterus and ovary. © 2017, Yuzuncu Yil Universitesi Tip Fakultesi. All rights reserved.View full-textArticleEffect of Tannation on Human Pituitary Gonadotropin as Determined by General Gonadotropin AssaysNovember 1965 · EndocrinologyA ALBERTE RosembergTannate complexes of human pituitary gonadotropin (HPG) of urinary origin were assayed by 3 general gonadotropin assay methods-the rat ovarian and uterine weight methods and the mouse uterine weight method. Assays were also performed by injecting single doses · of tannated and untannated HPG and determining the magnitude of the response at daily intervals. It was found that tannation did not ... [Show full abstract] alter the potency of HPG as measured in these 3 assay systems. The results obtained with single-dose assays confirmed those obtained by conventional assays. (Endocrinology 77: 766,1965Read moreArticleGiant pelvic epidermoid cyst: a rare observationMay 2003 · Gynécologie Obstétrique Fertilité Hafid HachiA RegraguiA BougutabS BenjellouneEpidermoid cyst is a frequent benign tumor. It gives rise to destruction of adjoining tissues, chiefly in the skull. The pelvic epidermoid cyst is rare. We report the second case of giant pelvic epidermoid cyst with a 25-year-old girl. Ultrasound exploration and computed tomography evoked an ovarian origin. The surgery discovers an epidermoid cyst under the peritoneum. Ovaries and uterus were ... [Show full abstract] undamaged. We realized an evacuation of cyst, which contained lamellas of keratin. Histology allowed confirmation of diagnosis. The patient is always observed after 18 months.Read moreDiscover the world s researchJoin ResearchGate to find the people and research you need to help your work.Join for free ResearchGate iOS AppGet it from the App Store now.InstallKeep up with your stats and moreAccess scientific knowledge from anywhere orDiscover by subject areaRecruit researchersJoin for freeLoginEmail Tip: Most researchers use their institutional email address as their ResearchGate loginPasswordForgot password? Keep me logged inLog inorContinue with GoogleWelcome back! Please log in.Email · HintTip: Most researchers use their institutional email address as their ResearchGate loginPasswordForgot password? Keep me logged inLog inorContinue with GoogleNo account? Sign upCompanyAbout usNewsCareersSupportHelp CenterBusiness solutionsAdvertisingRecruiting© 2008-2021 ResearchGate GmbH. All rights reserved.TermsPrivacyCopyrightImprint