Single-cell transcriptome atlas of the human corpus cavernosum

Zhao, LiangYu; Han, Sha; Su, HengChuan; Li, JianYing; Zhi, ErLei; Li, Peng; Yao, ChenCheng; Tian, RuHui; Chen, HuiXing; Chen, HuiRong; Luo, JiaQiang; Shi, ChenKun; Ji, ZhiYong; Hu, JianLin; Wu, Gang; Zhou, WeiDong; Tang, YuXin; Chen, YuZhuo; Lin, GuiTing; Lue, Tom F.; Wu, DengLong; Li, Zheng

doi:10.1038/s41467-022-31950-9

Download PDF

Article
Open access
Published: 25 July 2022

Single-cell transcriptome atlas of the human corpus cavernosum

LiangYu Zhao^1,2^na1,
Sha Han¹^na1,
HengChuan Su³^na1,
JianYing Li¹^na1,
ErLei Zhi¹,
Peng Li¹,
ChenCheng Yao¹,
RuHui Tian¹,
HuiXing Chen¹,
HuiRong Chen¹,
JiaQiang Luo¹,
ChenKun Shi¹,
ZhiYong Ji¹,
JianLin Hu¹,
Gang Wu⁴,
WeiDong Zhou⁴,
YuXin Tang²,
YuZhuo Chen⁵,
GuiTing Lin⁶,
Tom F. Lue⁶,
DengLong Wu³ &
…
Zheng Li ORCID: orcid.org/0000-0003-1070-3466¹

Nature Communications volume 13, Article number: 4302 (2022) Cite this article

8385 Accesses
19 Citations
11 Altmetric
Metrics details

Subjects

Abstract

The corpus cavernosum is the most important structure for penile erection, and its dysfunction causes many physiological and psychological problems. However, its cellular heterogeneity and signalling networks at the molecular level are poorly understood because of limited access to samples. Here, we profile 64,993 human cavernosal single-cell transcriptomes from three males with normal erection and five organic erectile dysfunction patients. Cell communication analysis reveals that cavernosal fibroblasts are central to the paracrine signalling network and regulate microenvironmental homeostasis. Combining with immunohistochemical staining, we reveal the cellular heterogeneity and describe a detailed spatial distribution map for each fibroblast, smooth muscle and endothelial subcluster in the corpus cavernosum. Furthermore, comparative analysis and related functional experiments identify candidate regulatory signalling pathways in the pathological process. Our study provides an insight into the human corpus cavernosum microenvironment and a reference for potential erectile dysfunction therapies.

Sex differences orchestrated by androgens at single-cell resolution

Article 10 April 2024

Spatially organized cellular communities form the developing human heart

Article Open access 13 March 2024

Single-cell multiplex chromatin and RNA interactions in ageing human brain

Article Open access 27 March 2024

Introduction

Sexual reproduction and the consequent relationships of couples are important foundations of human society and culture. Sexual health is thought to reflect overall male health; however, discussing this problem is often considered taboo, so less academic attention than other physiological problems has resulted in few studies on the human corpus cavernosum microenvironment. Through evolution, the human penis has lost structures such as the penis bone. Thus, the penis needs to expand several times in size during sexual intercourse, and the bulbocavernosus reflex during erection rigidity causes the intracorporal pressure to exceed the systolic blood pressure^1,2. During this process, erectile signals are transmitted to the corpora cavernosa (CC) by activation of the parasympathetic nervous system and signal transfer in the sacral erectile centre, regulating the nitric oxide synthase (NOS)/cGMP cascade, subsequently leading to depolarisation followed by smooth muscle relaxation with consecutively increased blood inflow^3,4. The CC is the basic functional unit of the penile erection process. However, its microenvironment’s constituents remain unclear, especially the fibroblasts (FB). These cells comprise about half of the total cavernosal cells, and research on their effect on the normal physiological function of the CC remains inadequate.

Erectile dysfunction (ED) causes physiological and psychological problems for patients and their families. The CC can be considered as a special blood vessel type, and its dysfunction is a harbinger of cardiovascular disease events, indicating that the corpus cavernosum is more susceptible to injury than other blood vessels⁵. Investigational cavernosal injury biomarkers may provide a window for the primary and secondary prevention of ED, and for cardiovascular disease and other vascular disorders⁶. ED’s pathogenesis can be further divided into the stages of functional disorder and organic disturbances. Phosphodiesterase type-5 (PDE5) inhibitors increase intracellular cGMP levels in vascular and corporal SMCs and thus potentiates the NO-mediated vasorelaxant response to sexual arousal^7,8. However, smooth muscle relaxation is only one part of the erectile process. No suitable treatment yet exists for cavernosal structural damage of ED patients at late stages because this depends on a full understanding of the cavernosal microenvironment, including the cell heterogeneity, the regulatory signalling, and the interactions with other cell types at the molecular level. In addition, most previous studies on the CC were based on the rat model, but the human and rat penis structure and erection process show large differences, so the relevant conclusions were sometimes not matched with the observation of human study⁹.

To overcome the above problems, we profiled 64,993 individual penile cavernosal cells from three men with normal erections and five patients with organic ED. Based on these data, we developed a single-cell atlas of the human CC and described the communication network within its microenvironment. Seven major cavernosal cell types, including six FB subclusters, four smooth muscle cell (SMC) subclusters, and four endothelial cell (EC) subclusters were identified with distinct transcriptome profiles. Each subcluster’s heterogeneity, including the spatial location, biological function, and regulatory pathway, was also identified. Furthermore, according to the distinct transcriptome profiles between normal and ED cavernosal cells, we obtained evidence that the Wnt and Hippo signalling pathways regulate the damage and repair of the CC tissue. Collectively, our results provide an in-depth insight into the cavernosal cellular microenvironment and the mechanisms underlying pathogenesis, thereby offering targets for ED prevention and treatment strategies.

Results

Overview of the cellular composition of the human CC

To characterise the CC’s cellular heterogeneity and baseline cellular diversity in a pathological state, we profiled CC cells from three donors with normal erections and five ED patients, including two patients with diabetes mellitus erectile dysfunction (DMED) and three patients with non-diabetic erectile dysfunction (non-DM) (Supplementary Data 1). Thus, 64,993 cells were retained after removing empty droplets, outliers, cell debris, and inferential doublets (Fig. 1a and Supplementary Fig. 1a). All donors ranged in age from 40 to 55, and each donor’s sex hormones were within the normal range (Supplementary Fig. 1a, b). Pathological examination revealed the five ED patients’ cavernosum tissue had obviously undergone fibrosis; the muscle/collagen ratio decreased by 50% in the three non-DM cases and by as much as 70% in the diabetes cases. In addition, the size of the cavernous sinus lumen in ED patients decreased (Supplementary Fig. 1c, d). Cells were further filtered according to the number of genes and reads, and each sample had a similar number of genes expressed (Supplementary Fig. 1e).

**Fig. 1: Global expression profiling of human corpus cavernosum cells in normal male and ED patients by single-cell RNA-seq.**

Seven major clusters were identified in the whole cell population based on the expression of known cell type-specific markers, including ECs, FBs, pericytes (PCs), macrophages (MACs), SMCs, Schwann cells (SWCs), and T cells (Ts). All clusters contained both normal and ED patient cells, but normal and ED CC showed significant heterogeneity in t-SNE plot (Fig. 1b, c). Each cluster had different gene transcription levels (Supplementary Fig. 1f). These seven clusters could further divided into 11 clusters according to the cells whether belong to cavernosal trabecular or the nearby vessels (Supplementary Fig. 1g). t-SNE plot of these cells showed good repeatability within the three normal samples and same type of ED patients (Supplementary Fig. 1h). A total of 6148 differentially expressed genes (DEGs) among the seven clusters under normal state were identified (Supplementary Fig. 1i and Supplementary Data 2). Gene ontology (GO) analysis was performed for the top 30 DEGs of each cluster, the results consistent with the previous understanding of the biological characteristics of these cell types (Fig. 1d). Of note, the FB cluster showed the most powerful signal output capacity among the seven clusters (Fig. 1e and Supplementary Fig. 2b, c) indicated their important regulatory function in the CC microenvironment.

Cell–cell interaction analysis in CC microenvironment under normal and disease state

To investigate the CC microenvironment’s complex signalling networks, we performed unbiased “ligand–receptor”, “extracellular matrix (ECM)–receptor” and “cell–cell contact” interaction analyses by CellChat, and identified 1284, 1461 and 1552 significant interactions in normal, non-DM and DMED CC respectively (Supplementary Fig. 2a, b). Among the seven clusters, FBs showed the strongest outgoing signalling pattern, indicating that FBs were key in regulating the microenvironment’s homeostasis (Supplementary Fig. 2c). Unexpectedly, the MAC, SWC and T clusters showed a typical signal incoming pattern. Among these signalling interactions, we discovered many classical vascular tissue regulatory pathways, such as VEGF, IGF, NOTCH, et al. (Supplementary Fig. 2c)^10,11. And further identified some signals that showed a different pattern between normal and pathological state such as KIT (receptor of SCF), FGF, EGF, et al. (Supplementary Fig. 2d).

Then, RNA-seq, ICC and ELISA experiments were performed to verify whether these ligands were indeed produced by CC FB at the RNA and protein levels. RNA-seq and ICC of FB indicated most of the ligands predicted by CellChat were expressed in FB in vitro (Supplementary Fig. 3a, c), and RNA-seq of EC indicated their paired receptors were expressed in EC (Supplementary Fig. 3b). The ELISA results showed that EC culture consumed most of the ligand molecules, and FB culture increased most of these factors’ concentrations (Supplementary Fig. 3d). These experimental results confirm CellChat’s prediction of the expression patterns of these signal ligands.

Next, we evaluated the effects of these signal ligands on EC functions, which were related to the repair of CC tissue after pathological injury. In the single factor variable experiment, IGF1 treatment had a positive effect on tube formation (Supplementary Fig. 4b, d), vascular bud length (Supplementary Fig. 4g), and cell migration (Supplementary Fig. 4j) of EC, while AGEs (Advanced Glycation End Products) and TGFβ acted as a negative regulator. These results were also verified by orthogonal experiments with various variables (Supplementary Fig. 4c, e, h, k). In addition, supernatant of FB also promoted above functions of EC, however, expect for cell migration, these effects disappeared after FB were treated with AGEs.

Nitric oxide (NO) and endothelin (ET) were two important regulators for the contraction and relaxation of SMC. The ELISA results indicated that most signal molecules even TGFβ and low level of AGEs treatment alone showed a positive effect on endothelial NO releasing (Supplementary Fig. 5a). However, the effect of each signal ligand significantly decreased in the orthogonal experiment (Supplementary Fig. 5b and Supplementary Data 6). Both high and low concentrations of AGEs can effectively stimulate the ET releasing. Except for VEGFA and CXCL12, most of the factors show inhibitory effects on ET releasing, especially the HBEGF has shown a strong inhibitory effect in both single factor stimulation and orthogonal experiment (Supplementary Fig. 5c, d, and Supplementary Data 6). Furthermore, we found high concentrations of AGEs, TGFβ, and Hippo pathway inhibitor XMU-MP-1 led to increased endothelial permeability (Supplementary Fig. 5e). TGFβ also showed a similar effect in the orthogonal experiment (Supplementary Fig. 5f), but there was no statistical difference (Supplementary Data 6). In general, IGF1 and SCF signals positively regulated CC tissue repair-related function. In addition, IGF1, SCF, FGF7 and HBEGF signals were beneficial to the regulatory function of smooth muscle relaxation, while high concentrations of AGEs and TGFβ were harmful to these functions. These signalling network results helped in further understanding the cavernosal microenvironment’s physiological and pathological processes.

Cellular heterogeneity of biological processes and spatial localisation in the FB cluster

To further explore the corpus cavernosal FB (CCFB) and their biological function in a framework of fibroblast lineage, we first combined our CCFB with the fibroblast dataset from 9 different studies including heart¹², skeletal muscle¹³, colon¹⁴, liver¹⁵, kidney¹⁶, skin^17,18 and lung^19,20 under normal and disease state. The similarity between fibroblasts derived from each tissue was compared and we found that CCFB was more similar to that of heart and skeletal muscle (Supplementary Fig. 6c). After referring to the classification criteria of Buechler’s study²¹, we found CCFB mainly consists of three clusters, PI16+, APOC1+ and COMP+ FB (Supplementary Fig. 6a, b). PI16 was predicted to act upstream of or within negative regulation of cell growth involved in cardiac muscle cell development and this cluster was thought as a root in trajectories of fibroblast differentiation²¹. These three clusters could further divide into 6 subclusters based on some CCFB specific markers including APOC1+/PTCHD1+ (FB1), APOC1+ /PPP1R14A+ (FB2), PI16+/FMO2+ (FB3), PI16+/BMP7+ (FB4), COMP+/KERA+ (FB5) and COMP+/MFAP5+ FB (FB6) (Fig. 2a). We identified 212, 151, 71, 429, 254 and 442 DEGs among these six CCFB subclusters under normal state, respectively (Fig. 2b, c and Supplementary Data 8). Then the DEGs between normal and ED state in each CCFB subcluster were also calculated (Supplementary Fig. 7 and Supplementary Data 9). We found that muscle-related GO terms, such as “muscle system process” and “muscle cell proliferation” were enriched in the PI16-FB (FB1 and FB2) subclusters. The FB4 cluster containing KERA, THBS4, SCX and collagens was consistent with a tenogenic signature. Although the general ECM genes were expressed at relatively higher levels in FB3, FB4 and FB5, each collagen type showed different expression patterns in the six FB subclusters. For example, collagen I was highly expressed in FB4 and FB5, while collagen II and collagen IV were highly expressed in FB1 and FB2 (Supplementary Fig. 8a). The collagen expression pattern also differed between normal and ED FB. For example, normal FB showed higher COL1A1 and COL6A3 expression levels but a lower COL4A1 expression level compared with FB from ED patients (Supplementary Fig. 8b, c).

**Fig. 2: Transcription characteristics and heterogeneity among the six FB subclusters.**

We then analysed the spatial localisation of each FB subcluster by immunohistochemical staining (IHC) of the CC sections. We first divided the CC into five major regions: (a) the septum pectiniforme, (b) the region near the septum pectiniforme, (c) the cavernosal trabecular region, (d) the cavernosal artery and (e) the nerve bundles (Supplementary Fig. 9a). PI16 was highly expressed in FB3, FB5 and FB6, and the IHC results showed that part of PI16+ cells existed in the artery region, while other PI16+ cells located between the smooth muscle bundles in the cavernosal trabecular region (Supplementary Fig. 9b). WISP2 showed a similar expression pattern and spatial localisation as PI16, but WISP2 was not expressed in FB6 and no WISP2+ cells were detected outside the artery region (Supplementary Fig. 9c). SRD5A2 was expressed in FB1, FB4 and part of FB3, and the SRD5A2+ cells were located in the region around the septum pectiniforme (Supplementary Fig. 9d). Neurofilament light chain transcript (NEFL) was only expressed in FB4, indicating that FB4 was highly correlated with neurons. The IHC results showed that FB4 mainly existed in large nerve bundles near the deep penis artery and in small nerve bundles within the cavernosal trabecular region (Supplementary Fig. 9e). Most NEFL+ cells were coated by a layer of NGFR+ cells that were identified as the FB6 cluster (Supplementary Fig. 9f).

When compared with ED patients, the proportion of the APOC1+ FB, especially FB1 subcluster in both non-DM and DMED patients was increased (Fig. 2d and Supplementary Fig. 6a, b). We then used ingenuity pathway analysis (IPA) to analyse each subcluster’s cell death and proliferation and found that proliferation- and survival-related terms, such as “cell cycle progression”, “proliferation of epithelial cells”, “cell viability” and “cell survival”, were significantly activated in the FB1 subcluster (Fig. 2e). Regarding heterogeneity in energy metabolism, we found that FB1 expressed a lower level of glycolysis-related genes, while FB1 and FB2 expressed a higher level of triglyceride metabolic-related genes (Fig. 2f).

Regulatory networks involved in phenotypic transformation of FBs

Considering that the FB1 subcluster showed muscle-relate characteristics and its proportion had increased in ED patients, we hypothesised that the FB1 subcluster is associated with excessive contraction and fibrosis of the cavernous sinus. To determine which signal pathway regulate phenotypic transformation among FB subclusters, we built a cell interaction network between the FBs and other CC cells with CellChat. The CellChat result indicated that FB1 mainly received WNT, IL17, NT and VCAM signalling (Fig. 3a). Then, we further identified regulatory pathways in each FB subcluster by IPA according to the downstream DEGs. IPA analysis showed that terms such as “Wnt/β-catenin signalling”, “VEGFs interactions”, “NAD signalling” and “Endothelin-1 signalling” were activated (Fig. 3b). To further explore the key regulators of each FB subcluster, we analysed 1665 human transcription factors and their downstream gene set using GENIE3. We found that factors such as FOS, EGR1 and JUN might specifically play critical roles in the FB1 subcluster (Fig. 3c), and other subclusters also had their own special regulators (Supplementary Fig. 11a). Interestingly, among these transcription factors, JUN and TBX3 are direct targets of Wnt/β-catenin signalling, further demonstrating its importance. Then, we focused on the target genes of Wnt/β-catenin signalling and found that the expression levels of CTNNB1, ATF3, CCND2 and MYC were significantly higher in the ED FB1 subcluster (Fig. 3d), and more JUN and ATF3 positive cells were also detected in ED CC tissues (Supplementary Fig. 10). This evidence suggested that WNT pathway activation may drive the phenotypic transformation between normal and perturbed state.

**Fig. 3: Regulatory signalling of FB phenotypic transformation.**

To investigate this hypothesis, we cultured FBs in vitro with a WNT signalling pathway inhibitor (ICG-001) or activator (SKL2001). The change in β-catenin protein level and its translocation demonstrated the treatment efficiency for WNT signalling pathway (Supplementary Fig. 11b). The cell morphology in both ICG-001-treated and SKL2001-treated groups changed markedly. In the SKL2001-treated group, cells became bigger and a myo-texture appeared, and decreased cell refraction indicated that the FBs were getting thinner (Fig. 3e). The ICG-001-treated group showed an opposite change; the cell area became smaller, the refraction increased, and the internal texture became invisible (Fig. 3e). In addition, a low dose (10 µM) of SKL2001 improved FB proliferation significantly, while ICG-001 treatment severely inhibited FB proliferation (Fig. 3f). To further explore the FB inner changes after activating or inhibiting WNT signalling, we performed transcriptome sequencing of FBs treated with SKL2001 or ICG-001 in vitro. We identified 1711 upregulated and 255 downregulated genes in SKL2001-treated FBs when setting the thresholds of expression fold change >2 and P-value <0.05 (Supplementary Data 10). In addition, the GO terms “muscle contraction” and “steroid biosynthetic process” were found positively and negatively enriched, respectively (Fig. 3g). In ICG-001-treated FBs, 1064 upregulated and 803 downregulated genes were identified, and the “collagen trimer” term was significantly inactive (Supplementary Fig. 11c). The trend of muscle contraction and collagen-related gene upregulation was more significant when comparing the SKL2001 and ICG-001 groups (Supplementary Fig. 11d). It is worthy to note that the upregulated genes in pathway “muscle contraction” not only contained the executor of muscle contraction such as ACTA2 and CACNA1H, but also contained some regulatory signal molecules such as EDN1. Neither the constriction of the supplying artery muscle nor of the CC SMC is conducive to penile erection, so this result may represent a pathological state of CCFB.

Increased expression of extracellular matrix proteins and increased proliferation of fibroblasts can lead to disruption of microenvironmental homeostasis and tissue fibrosis. To further explore the effect of WNT activation and inhibition on the whole cavernous tissue microenvironment, we used SKL2001 and ICG-001 in the CC tissue culture system. The ratio of smooth muscle to collagen was used as a measure of CC tissue fibrosis. As expected, ICG-001 treatment significantly delayed the CC’s autonomic fibrosis process in vitro, while SKL2001 showed an opposite effect (Supplementary Fig. 11e, f). These results suggest that WNT signalling could be a drug target for FB phenotype transformation-induced CC structure damage.

Heterogeneity of muscle contraction regulation and different spatial location in the SMC cluster

The degree of erection depending upon the balance between the inflow and outflow of blood to the penis, and controlled relaxation of the cavernosal smooth muscle is the key process. Because PC cluster also showed general myocyte features, it was combined with SMC cluster and analysed in this part. After re-clustering and dimensionality reduction processes, we found that the SMC could be further divided into four subclusters (Fig. 4a). Even though all four subclusters showed a high level of ACTA2 and MYH11, each subcluster showed a distinctive transcription pattern of some specific actin and myosin proteins (Fig. 4a). In the four SMC subclusters at normal baseline, 531, 383, 345 and 711 DEGs were identified, respectively (Supplementary Data 11). GO analysis showed that the GO term “vasculogenesis” and “artery development” was enriched in the SMC1 and SMC3 subclusters, suggesting their role in angiogenesis (Fig. 4b).

**Fig. 4: The heterogeneity within the SMC cluster.**

To further determine the spatial location information for each SMC subcluster, we chose desmin (DES), a SMC2-specific protein, to mark this cluster with IHC in the corpus cavernosum tissue. Interestingly, DES+ cells were detected in the nearby septum pectiniforme and cavernosal trabecular regions but not in the vascular region, indicating that DES⁺ SMCs were the cavernosal trabecular SMCs (Fig. 4c). Furthermore, THY1 was expressed highly in SMC3 and part of FB, and THY1+ cells were detected around small vessels where fits the location of pericytes (Supplementary Fig. 12a). In addition, we found that LUM, a collagen fibril binding protein that is distributed in interstitial collagenous matrices throughout the body, was highly expressed in all FBs and in SMC4 (Supplementary Fig. 12b). LUM+ cells were widely distributed in the cavernosal trabecular region but not in the small artery (Supplementary Fig. 12b). Combined the above results and their transcript characteristic, we inferred that SMC1 is the cluster of vascular SMCs (VSMC), SMC2 are the cavernosal trabecular SMCs (CCSMC), SMC3 are pericytes (PC) and SMC4 are a type of myofibroblast (MFB).

Continuous contraction of the penis smooth muscle is an important cause of ED, so we then compared the activation and inhibition of various signalling pathways affecting smooth muscle contraction and relaxation in normal male and ED patients. For the contraction pathways, the alpha adrenoceptor showed significant heterogeneity among the four SMC subclusters. ADRA2A was only detected in VSMC and FB, and its expression level was significantly higher in DMED patients than in normal within the MFB cluster. However, ADRA2C, another subtype of alpha adrenoceptor, was highly expressed in CCSMC, and its expression level was also upregulated in DMED patients but not in non-DM patients (Fig. 4d). MYLPF (myosin light chain, phosphorylatable, fast skeletal muscle) and MYLK (myosin light chain kinase) are involved in maintaining continuous smooth muscle contraction, and both were highly expressed in CCSMC, especially in ED patients (Fig. 4d). Other muscle contraction and maintenance pathways, such as the Rho/ROCK pathway, thromboxane signalling, and endothelin signalling, also showed cell type heterogeneity within the SMC cluster. For example, ROCK1 was highly expressed in VSMC, TBXA2R was highly expressed in CCSMC, and EDNRA was highly expressed in PC (Supplementary Fig. 12c–e). It is noteworthy that even though EDN1 is a potent vasoconstrictor and has been reported to promote contraction of smooth muscle in the CC, we found that its receptor EDNRA was downregulated in both non-DM and DMED patients (Supplementary Fig. 12e).

In addition to the classical signalling pathways associated with erections, we used IPA analysis to comprehensively describe the different signalling in CCSMC between normal male and ED patients according to the DEGs (Supplementary Data 12). The intersection of pathways that changed in both non-DM and DMED patients was identified, and “TGF-β signalling”, “ECM–receptor interaction” and “protein targeting” were activated, while “FoxO signalling”, “mineral absorption” and the “Hippo signalling pathway” were significantly inhibited in ED patients (Supplementary Fig. 13a). The Hippo signalling pathway is involved in the pathophysiology of various cardiovascular diseases²². YAP1 (Yes-associated protein), the main transcriptional complex of the Hippo pathway, was detected in VSMC, CCSMC, and MFB (Supplementary Fig. 13b). Furthermore, we found that CTGF, a downstream target gene of the YAP1 complex and a potent skeletal muscle fibrosis promotor²³, was significantly upregulated in all four SMC subclusters of ED patients (Supplementary Fig. 13b). These results were further confirmed by IHC staining of cavernous tissue. In normal male samples, CTGF+ cells were only detected in the interstitial area between smooth muscle bundles, but almost all SMCs in the CC of ED patients were CTGF+ (Supplementary Fig. 13c). The expression of YAP localized in the CCSMC nucleus of ED patients was significantly elevated (and the volume of these nuclei increased), while p-YAP expression levels were decreased (Supplementary Fig. 13d).

Heterogeneity of biological characteristics and spatial location in the EC cluster

ECs play an important role in smooth muscle contraction regulation and vascular barrier structure, and injury to ECs is a major cause of ED, especially DMED²⁴. After re-clustering and tSNE dimensionality reduction, the EC cluster under normal state was further divided into four subclusters (Fig. 5a). A total of 220, 151, 699 and 662 DEGs were identified in each EC subcluster (Fig. 5b and Supplementary Data 13). And the DEGs between normal and ED patients in each EC subcluster were also identified (Supplementary Fig. 15). Based on these DEGs, we performed GO analysis and found that the terms “migration” and “vascular development”, which belong to the basic functions of ECs in the vascular system, were enriched in EC1 (Fig. 5b). However, in the EC4 subcluster, GO terms such as “extracellular matrix organisation”, “response to TGFβ”, and “mesenchyme development” were enriched. Furthermore, a large number of ECM- and myo-related genes, such as COL1A1, COL5A2, LUM, ACTA2, ACTG2 and DES, were upregulated in EC4, indicating that EC4 underwent endothelial–mesenchymal transition (Fig. 5b and Supplementary Data 13). Finally, GO terms such as “response to cytokine”, “regulation of cell death” and “cell adhesion molecules” were enriched in EC2, and some cytokine receptors and adhesion molecules, such as ACKR1, LIFR, LTBR, IL1R1, CD34, SELP, SELE and 12 MHC-II genes, were highly or specifically expressed in EC2 (Fig. 5b and Supplementary Fig. 16a). In summary, EC2 and EC4 were more like ECs in pathological states, and the increased proportion of these two subclusters in ED patients further suggests their relationship with cavernosal lesions.

**Fig. 5: The heterogeneity within the EC cluster.**

Then, we investigated the different spatial localisation of each EC subcluster. The IHC staining result showed that KIT, an EC1 marker gene, was specifically expressed in the cavernosal trabecular EC (CCEC), but not in vascular EC (VEC) (Fig. 5d). However, it should be noted that the mast cells (mast tryptase-positive cells) distributed in the cavernosal trabecular region were also KIT-positive cells and could be distinguished using general endothelial markers such as VWF and CD31. On the contrary, GJA5 was a EC3-specific marker, and GJA5+ ECs were only detected in arteries and their branches, indicating that EC3 are VEC (Fig. 5e). The human corpus cavernosum is supplied directly by arterioles and lacks capillaries. We further investigated the expression of different vascular endothelial marker genes in the four EC subclusters based on previous reports and found that the EC3 subcluster indeed showed transcription characteristic of artery ECs. However, all these marker genes lacked characteristic transcriptional distribution in the EC1 subcluster, suggesting that the CCEC were a unique endothelial subtype (Supplementary Fig. 16b)^25,26. In biopsies of ED corpus cavernosum, we found that the high expression of IGF1 may be related to endothelial injury because the cavernosum tissue of ED patients contained larger and more IGF1-positive regions than that of normal males. In the high IGF1 expression region, VWF+ ECs disappeared and the FBs were exposed, but this phenomenon was not detected in the low IGF1 region. In addition, IGF1 and VWF double-positive cells were mainly found in the adjacent regions with endothelial deficiency (Fig. 5f, g). IGF1 and its receptor signalling pathway-related genes were highly expressed in the EC4 subcluster (Fig. 5h). IGF1 inhibits EC apoptosis and promotes vascular injury repair in various tissues and organs^27,28,29, indicating that an increase in the local IGF1 concentration and the emergence of EC4 were feedback regulations in response to endothelial injury.

To investigate the upstream regulatory signalling in each EC subcluster, we further described the signalling network between the four EC subclusters and the other corpus cavernosum cells. The CellChat result showed that “BMP”, “IL6” and “KIT” signalling were enriched in the EC1 subcluster. Immune-related signalling pathways, such as “CCL”, “LIFR” and “LIGHT (TNFSF14-TNFRSF14/LTBR)” signalling, were activated in EC2. In addition, “WNT”, “NOTCH” and “IFN-1” signalling and “IGF”, “EGF” and “FGF” signalling were dominant in EC3 and EC4, respectively (Fig. 5i).

Mitochondrial dysfunction and apoptosis are major pathological features in the cavernosal trabecular endothelium of ED patients

Similar to other vascular systems, the different EC subclusters showed an obvious heterogeneous susceptibility for injury and aging³⁰. The cell function results of the IPA analysis showed that the EC3 and EC4 subclusters had a more active cell proliferation and migration ability, while the IPA term “cell death” was activated in the EC2 subcluster, suggesting that the CCEC had a higher susceptibility to injury (Supplementary Fig. 16c). This hypothesis was further validated by IHC staining. We found that the cavernosal endothelium of both normal and ED patients had good endothelial integrity and endothelial marker expression in the arteries and their branches; however, decreased expression of endothelial markers and many endothelial exfoliation regions were detected in ED samples, especially in DMED patients (Supplementary Fig. 16d).

To get a more intuitive impression, scanning electron microscopy (SEM) was used to detect endothelial changes in normal and ED CC. In normal tissue, the endothelial surface was smooth and the EC were arranged in a neatly manner with narrow gaps. However, in both non-DM and DMED tissue, the cavernosal trabecular endothelium was obviously wrinkled, and particles with diameters of 50–100 nm were densely adhered to its surface. Some regions of the endothelium were detached with the collagen fibres inside exposed. In DMED tissue, the ECs became obviously swollen, forming a papillary bulge on the surface (Fig. 6a). In addition, the results of γH2AX immunofluorescence staining showed that the ECs of ED cavernosal trabeculae underwent a significant increased apoptosis (Fig. 6b).

**Fig. 6: Pathological features of endothelial injury in ED CC.**

Oxidative phosphorylation dysfunction, which induces oxidative stress and decreases metabolism, is an important cause of endothelial injury in ED CC³¹. In the IPA analysis, we found that NCOA7, SFPQ, MCL1, TXN and SOD2, which belong to the “cell death in response to oxidative stress” pathway, were significantly upregulated in ED EC, and no genes belonging to this pathway were downregulated (Fig. 6c). Furthermore, the “oxidative phosphorylation” pathway was significantly inhibited in the EC3 subcluster of ED patients (Fig. 6d). To explore the cause of this energy metabolism disorder, a transmission electron microscope (TEM) was used to detect the shape and structural changes of mitochondria in the cavernosal trabecular ECs. In ED ECs, the mitochondria were markedly swollen, characterised by increased mitochondrial size and decreased density under TEM. In addition, the number of mitochondria per EC decreased in the non-DM ED patients (Fig. 6e). Integrated analysis of genes in the glycolysis and oxidative phosphorylation pathways showed that the glycolysis gene set was significantly upregulated in DMED ECs, while the oxidative phosphorylation gene set was significantly downregulated in both non-DM and DMED patients (Fig. 6f and Supplementary Fig. 16e). In summary, cell apoptosis and mitochondrial dysfunction are the main pathological features of endothelial injury in ED CC.

YAP signalling is involved in endothelial injury of DMED CC

In addition to EC injury, fibrosis is another important pathological feature in DMED CC. EC has been proved as one of major sources to the emergence of fibroblasts³². We found that many fibrogenic signals (Supplementary Data 8), such as “integrin signalling”, “TGF-β signalling” and “renin–angiotensin signalling” in CCEC of ED patients were activated^33,34,35, while negative modulating pathways of fibrosis, such as “HIPPO signalling” and “PPAR signalling”, were inhibited (Fig. 6d)^36,37. Furthermore, the signalling network analysis based on the DEGs of the EC4 subcluster also indicated that “TGFB3” and “AGT (angiotensinogen)”, extracellular signalling factors belonging to “TGF-β signalling” and “renin–angiotensin signalling”, were upregulated. In the intracellular region, their downstream transcription factors such as “TEADs” and “SMAD3” were also activated (Supplementary Fig. 17a). YAP is the most important downstream nuclear effector of the Hippo signalling pathway, and can bind to TEADs or SMAD3 to regulate downstream gene expression, and then induce fibrosis and endothelial dysfunction³⁸. From the IHC staining, we found that the expression level of nuclear-localised YAP protein was significantly increased and phosphorylated YAP was decreased in ED EC, suggesting that the Hippo pathway is indeed inhibited (Fig. 6g and Supplementary Fig. 17b).

Since abnormal inactivation of the Hippo pathway is associated with endothelial injury in ED corpus cavernosum, we further verified the effects of Hippo activation or inhibition on ECs in vitro. After isolation and purification, cavernosal trabecular ECs were treated with Peptide 17 (a YAP-TEAD inhibitor) or XMU-MP-1 (a MST1/2 inhibitor that can activate downstream YAP protein). We found that neither of the two drugs alone obviously altered the morphology and biomarker expression (Supplementary Fig. 8c). We therefore further tested their effects in the presence of TGF-β, the most common inducer of fibrosis. We found that ECs treated with TGF-β became larger in size, became dispersed rather than close together, and the expression of endothelial markers decreased. This process was inhibited and promoted by Peptide 17 and XMU-MP-1 treatment, respectively (Fig. 6h). In addition, TGF-β and XMU-MP-1 treatment also increased permeability of CCEC (Supplementary Fig. 5e). This result suggested that a YAP inhibitor could alleviate the injury to ECs induced by TGF-β. Next, we used CC tissue culture experiment which theoretically contains all the cellular components and signalling networks in the microenvironment, to the maximumly extent possibility of simulating the effects of these factors on CCEC and CC tissue in vivo. We found that TGF-β, Angiotensin II and XMU-MP-1 significantly promoted the progression of tissue fibrosis, while Peptide 17 showed the opposite effect. Two other YAP inhibitors, Verteporfin and TED-347, could also alleviate fibrosis of the CC tissue; however, both their short-term and long-term efficacies were inferior to that of Peptide 17 (Supplementary Fig. 17d).

Discussion

The CC has been extensively studied in model organisms, such as mice and rats. However, our knowledge of the cellular heterogeneity and communication in the human CC at the molecular level and at a single-cell resolution are still limited. Here, we used scRNA-seq to produce a single-cell transcriptomic atlas of CC tissue from normal males and ED patients. These data provide insights into the cellular composition, biological features, and regulatory signalling networks of the CC, to give a better understanding of the human CC microenvironment and its dynamic changes in ED patients.

In this study, we identified seven major cell subsets in eight human CC samples. Most of these cell subsets have been reported to play crucial roles in physiology and diseases^9,39,40. Furthermore, we found the FB cluster, the dominant cavernosal cell subset, as a major signalling output source in CC microenvironment rather than simply forming the scaffold for CC tissue by expressing the extracellular matrix protein. They produce VEGF, BMP and IGF signalling to regulate ECs, and produce NGF signalling to regulate neuron and Schwann cells. The FBs are also involved in the response to injury; for example, they secrete large amounts of IGF1 protein in regions with endothelial deficiency to promote injury repair. This evidence suggests that the FBs have more functions than simply extracellular matrix formation. In addition, the spatial heterogeneity within the FBs suggests that they may be derived from different precursor cells and play different roles in the formation of the cavernosum structure. To clarify this hypothesis, it would be necessary to collect samples at different ages (from the foetus to post-puberty) for single-cell sequencing. Furthermore, the change of proportion of each FB subcluster in ED corpus cavernosum suggests that abnormal phenotypic transformation within the FB cluster may destroy the homeostasis of the microenvironment. The standardized naming of FB subcluster across tissues referenced Buechler’s study, however, considering that CC was a specific type of vascular structure, additional dataset of heart, muscle, skin (contain small vessels) and kidney were compared in this study, and a cluster (APOC1+ FB) enriched in CC, heart, and muscle was identified. The apparent increase in APOC1+/PTCHD1+ FB (FB1) cluster in ED patients indicated their relevance to the CC tissue injury or repair. This cluster also expressed high level of muscle-related genes; it may be related to the CC sinusoids dilatation limitation that similar to restrictive cardiomyopathy⁴¹. However, the similarity between CC and heart highlights the current challenges of defining CC FB subsets by only transcriptomic technologies, for example, the FB4 cluster expressed a high level of SCX, while Scx+ cells have been identified as important in progenitors for cardiac valve morphogenesis⁴². Further spatial transcriptome combined with multiomics analysis would solve many limitations left in this part. In addition, we identified some signalling pathways that may play an important role in this process. For example, WNT/β-catenin signalling promotes the fibroblast-to-myofibroblast transition during cardiac, pulmonary, and renal fibrosis^43,44,45, and it was activated in the FB1 subcluster. In this study, FB treated with WNT signalling inhibitor did not showed a classical myofibroblast phenotype: although most muscle-contraction- and ECM-related proteins were upregulated, some typical myofibroblast upregulated genes such as COL1A1 and LOX were downregulated. However, the delaying effect of ICG-001 on the fibrosis of CC tissue ex vivo still suggested that the WNT pathway may serve as a potential drug target for inhibiting cavernous fibrosis in ED patients.

The contraction and relaxation of smooth muscle is the key event in the erection process^46,47. Unlike rats, human smooth muscle bundles are larger and occupy most of the space in the cavernosal trabecular region, rather than being banded in the subendothelial region⁴⁰. Because of the absence of the penile bone, the human cavernosal smooth muscles are subjected to tremendous pressure compared with those in other tissues, usually higher than systolic blood pressure in the rigidity phase of an erection, to maintain erection and complete insertion². We observed a significant difference in desmin protein expression in the cavernosal trabecular region. Desmin filaments form the intermediate filament system of muscle cells where they play important roles in maintaining mechanical integrity and elasticity. Desmin mutations lead to entangled filaments and impaired tensile properties^48,49. These results suggest that the CCSMC is a special kind of smooth muscle rich in desmin intermediate filaments, which may be the structural basis of its ability to withstand the drastic changes in length and pressure during the cavernous sinus filling process. In addition, there were significant differences in the distribution of some classic signalling receptors that regulate SMC contraction and relaxation among SMC subclusters, such as adrenergic receptors. This heterogeneity may help the development of therapeutic drugs that only target cavernosal trabecular smooth muscle, to reduce the multiple target effects of the current drugs that are commonly used, such as headaches, gastrointestinal complaints, and hypotension^50,51.

The number of ECs was comparable to that of FBs, accounting for about 40% of the total cavernosal cells. Even though the heterogeneity of ECs among different tissues or within one organ has been reported^25,52,53, whether human cavernosal ECs are also heterogeneous has remained unclear because of limited access to adult human tissue. In this study, we identified four EC subclusters, by comparing the sequencing data from multi-tissue-derived ECs²⁵, we found that the expression profile of CCEC was significantly different from that of arteries or veins, especially regarding the lack of GAP junction protein GJA5, suggesting a unique mechanism of signal transduction and endothelial junctions. Endothelial injury is the most significant pathological change of corpus cavernosum disorder, which may occur in the early stage of ED²⁴. We found that the CCEC were more susceptible than the adjacent vascular ECs (EC3), providing a histological explanation for the clinical observation that ED shares risk factors with cardiovascular disease and could be a harbinger of cardiovascular events. However, the molecular mechanism underlying the differing susceptibilities still requires further exploration. The endothelial fibroblasts (EC4) only existed in the region of endothelial shedding and high IGF levels, indicating that they may be involved in endothelium repair under the regulation of paracrine IGF signalling.

In ED samples, we observed a regulatory pattern of activated TGF-β signalling and inhibited Hippo signalling in both the SMC and EC clusters. The crosstalk between Hippo/YAP and TGF-β/SMAD3 signalling extensively affects the fibrosis process of various tissues and organs^36,38,54. We found that Peptide 17 or XMU-MP-1 treatment could relieve or accelerate the endothelial injury and cavernosal fibrosis induced by TGF-β treatment in vitro, indicating that Hippo/YAP could be another candidate drug target for organic lesions of the corpus cavernosum.

Our study reveals the human corpus cavernosum’s transcriptome landscape and cellular heterogeneity and offers insights into the cellular foundation of normal erection and ED knowledge. Supported by verification at the protein level and cell function experiments, we also highlight some regulatory signalling pathways and therapeutic targets.

Methods

Experimental models and subject details

The experiments performed in this study were approved by the Ethics Committee of Shanghai General Hospital (License No. 2021SQ259). The Office of Academic Research and the Ethics Committee of Shanghai General Hospital have authorised this study and confirmed the study design and conduct complied with all relevant regulations regarding the use of human study participants and was conducted in accordance with the criteria set by the Declaration of Helsinki. For CC samples, all donors signed their consent after being fully informed of the goal and characteristics of our study. We have obtained all donors’ consent to publish information that identifies individuals including sex, age, BMI and the hospital name. Three normal tissue samples were obtained from the tumour margin of penile carcinoma resection, all patients reported good stimulated erections and early morning erections. The five ED CC tissues were obtained from biopsy samples in artificial cavernous body implantation, all the ED patients were diagnosed with organic ED rather than psychological ED by the nocturnal penile tumescence (NPT) and intracavernosal injection (ICI) tests. The DMED patients had an at least 10-year history of type 1 diabetes and had good blood glucose control before surgery. The detailed clinical information of these patients could be found in Supplementary Data 1.

Histological examination

Fresh CC tissues obtained from surgery were washed with PBS twice to remove blood and immediately fixed in 4% paraformaldehyde for 12 h at 4 °C, then embedded in paraffin and sectioned. Before staining, the tissue sections were dewaxed in xylene, rehydrated using a gradient series of ethanol solutions, and washed in distilled water. For Masson staining, the sections were stained in turn with haematoxylin, aniline blue, and ponceau. For Sirius red staining, the sections were stained with Sirius red and haematoxylin. Then, sections were dehydrated using increasing concentrations of ethanol and xylene, and allowed to dry before applying neutral resin to the coverslips. Images were captured with a Nikon Eclipse Ti-S fluorescence microscope (Nikon). The images were measured and analysed with Adobe Photoshop CC 2018 software. In brief, the muscle/collagen ratio was represented by the area and identity of the red region (colour code = #0068b7; colour range = 160%) and the blue region (colour code = #b26c91; colour range = 160%), respectively.

Immunohistochemical staining and fluorescence intensity measure

Sections were performed with rehydration, antigen retrieval, and blocking, and then incubated with appropriate primary antibodies (Supplementary Data 15) at 4 °C overnight. Sections were further incubated with secondary antibody for 2 h at room temperature. Nuclei were labelled with Hoechst33342 by incubating tissue sections for 15 min⁵⁵. Images were captured with an OLYMPUS IX83 confocal microscope.

Fluorescence intensity measurements were processed with Adobe Photoshop CC 2018 software. For the fluorescence intensity of β-catenin and YAP, the nuclear regions were selected, for phosphorylated YAP, CD31 and VWF, the whole cell regions were selected, and then the single staining of the target protein was converted to greyscale. After that, the fluorescence intensity of the chosen regions was recorded.

Isolation of single CC cells

Fresh CC tissue samples of about 1 × 1-cm size were first immersed in 15 ml PBS with vigorous shaking to remove residual blood cells. Then, tissues were cut into pieces of 1 × 1-mm size and enzymatically digested with 2.5 mg/ml collagenase type I (17104-019, Gibco), 4 mg/ml collagenase type IV (17018029, Gibco), 0.1 mg/ml neutral protease, and 2 mg/ml DNase I (#AMPD1, Sigma-Aldrich) at 37 °C for 40 min. Subsequently, the cell suspension was filtered through a 40-mm nylon mesh, and the cells were sorted by MACS with a Dead Cell Removal Kit (130-090-101, Miltenyi Biotec) to remove dead cells. The cells were re-suspended in 0.05% BSA/PBS buffer before 10× Genomics library preparation.

Single-cell RNA-seq library preparation and sequencing

Single-cell RNA-seq libraries were prepared using the Chromium Single Cell 3′ Library & Gel Bead Kit v3 (PN-1000094, 10× Genomics) according to the manufacturer’s instructions. Final libraries were sequenced on an Illumina NovaSeq 6000. The raw sequencing reads were processed by Cell Ranger (v.3.1.0) with the default parameters. The reference genome was GRh38-1.2.0.

Quality control and sample integration

The gene–cell matrix of each sample was used to create a Seurat object with the Seurat package in R. Cells were further filtered according to the following threshold parameters: the total number of expressed genes, 800–7000; total UMI count, 0–30,000; and proportion of mitochondrial genes expressed, <10%. Batch correction was performed using the IntegrateData function in the Seurat package according to the “package manual [https://satijalab.org/seurat/v3.1/pbmc3k_tutorial.html]”.

Cell identification and clustering analysis

The merged Seurat objects were scaled and analysed by principal component analysis (PCA). The first 20 PCs were also used to get clustering and perform t-distributed stochastic neighbour embedding (tSNE) dimensionality reduction. The FindClusters function in Seurat package with the resolution parameter set as 0.5 was used to cluster the cells. For further analysis of each cluster, we isolated them and performed the above two steps again to get subcluster information.

For each major cluster, the EC cluster was identified by high expression of PECAM1 and VWF; the FB cluster was identified by expression of LUM, COL1A1 and PDGFRA; the PC cluster was identified by high expression of RGS5 and KCNJ8, but low expression of MYOCD; the SMC cluster was identified by high expression of ACTA2 and MYH11; the SWC cluster was identified by high expression of S100B and MPZ; the MAC cluster was identified by high expression of CD163 and CD68; and the T cluster was identified by high expression of CD3D and CD3E.

Differentially expressed gene calculation and gene enrichment analysis

The Seurat function FindAllMarkers (test.use = wilcox; min.pct = 0.1; logfc.threshold = 0.25) was used to identify differentially expressed genes (DEGs) based on the normalised UMI count. Unless otherwise noted, the DEGs in each selected subcluster were calculated based on comparison between that subcluster and the rest of the dataset. GO analysis was performed using the “WebGestalt [website http://www.webgestalt.org]”, the Over-Representation Analysis (ORA) or Gene Set Enrichment Analysis (GSEA) was chosen as Method of Interest, and only Biological Process was chosen in Functional Database. The pathway analysis was performed using Ingenuity Pathway Analysis (IPA) software based on the log₂ (FC) and P-values of the DEGs. The other gene sets used in this study are listed in Supplementary Data 15.

Ligand–receptor interaction and transcription factor network construction

The CellChat package was used for ligand–receptor interaction analysis. The cell–gene matrix was divided according to the six major clusters or five major combined with the subclusters of the other major cluster. The “secreted signalling”, “ECM–receptor” and “cell–cell contact” paired datasets were chosen to analyse the cell communication.

For the transcription factor (TF) regulation network analysis, a total of 1469 human TFs in AnimalTFDB and a cell–gene matrix were taken as input for the GENIE3 package. In the output regulator–target table, only pairs with weights greater than 0.1 were retained and the first 2–6 TFs were displayed further.

CC tissue culture experiment

Fresh CC tissue samples of about 1 × 1-cm size were first immersed in 15 ml PBS with vigorous shaking to remove residual blood cells. Then, tissues were cut into pieces of 1 × 1-cm size and cultured in 2% FBS/DMEM with SKL2001/ICG-001 (for the results of Supplementary Fig. 11f) or ANG II, TGFβ, Peptide 17, Verteporfin, TED-437 and XMU-MP-1 (for the results of Supplementary Fig. 17d). The culture systems were than place in a place on a shaking table in an incubator, 5% CO₂, 37 °C. A quarter of the tissue was collected for Masson or H&E staining on the 3rd and 7th day.

Isolation and culturing of FBs and ECs from the CC

The method of tissue digestion was the same as that of the single-cell sample preparation, the difference was that the digestion time for ECs was 20 min, while that of FB cells was about 40 min. The digested single-cell suspensions were washed and re-suspended in 10% FBS/DMED (Gibco) and EGM-2 (CC-3202; Lonza) medium for FB and EC culturing, respectively. Because of the strong proliferation capacity of FBs, FB cells would eventually dominate in DMEM medium, and the purity was more than 95% after passage. For ECs, cell cloning began at about day 5 of culture, then fibroblast inhibitors were added for 7–10 days. When isolated colonies began to fuse with each other, the EC colonies were marked and digested using a clonal cylinder (C7983-50EA; Sigma-Aldrich), and the EC suspension was further purified using CD31 magnetic beads (130-091-935; Miltenyi).

Endothelial cell functional experiments for validating the prediction of CellChat analysis

Tube formation: EC cells were re-suspended in each group of medium supplemented (Supplement Table 6) at a concentration of 200/μl, and then seeded on µ-Slide (81506, ibidi) with pre-added 10 μl Matrigel (256230, Corning). After 2 hours, EC cells were fixed with PFA, and then incubated with FITC-phalloidin (G1028-50UL, Servicebio) for 30 min. Images were captured with an OLYMPUS IX83 confocal microscope.

Vascular bud length: EC were seeded in AggreWell (34411, STEMCELL) at a concentration of 3 × 10⁵ per well, after 12 h, the cell mass were re-suspended and seeded on µ-Slide (81506, ibidi) with pre-added 10 μl Matrigel (256230, Corning) for 6 h. Next, cell supernatant was replaced with conditioned medium (Supplement Table 6) for 24 h and fixed. The ratio of cell mass diameter to the distance from the centre of the cell mass to the farthest end of the vascular bud in each group was recorded and counted.

Endothelial permeability: The EC were inoculated into the upper chamber of Transwell at a concentration of 2 × 10⁵ per well, and then 700 μl EGM-2 (CC-3202; Lonza) medium was added into the lower well. After cell adherence, the supernatant of the upper well was replaced with conditioned medium of each group (Supplement Table 6), and cultured for 24 h. Then 70-kda TRITC-dextran (T1162, Sigma-Aldrich) with a final concentration of 2 mg/ml was added into the upper well, 5 h later, the fluorescence intensity at 590 nm of the liquid in the lower well was detected.

RNA-seq of FB in vitro

1 × 10⁵ FB cells were seeded in a six-well microplate containing 2 ml 10% FBS/DMEM medium per well. After 24 h, the medium was replaced by 2% FBS/DMED containing 10 UM of ICG-001 (S2662, Selleck), 10 UM of SKL2001 (S8334, Selleck), or an equal volume of DMSO. The medium was changed every other day during culturing. After 7 days, all cells were harvested with TRIzol (15596026; Thermo Fisher) for mRNA sequencing. Library construction and sequencing were performed by Sinotech Genomics Co., Ltd (Shanghai, China). Total RNA was isolated using RNeasy mini kit (74904; Qiagen). The poly(A)-containing mRNA molecules were purified using poly(T)-oligo-attached magnetic beads. Then the mRNA underwent fragmentation, reverse transcription, purification, and enrichment. Clusters were generated by cBot with the library diluted to 10 pM and then were sequenced on the Illumina NovaSeq 6000 (Illumina, USA). Paired-end sequence files (fastq) were mapped to the reference genome (GRh38-1.2.0) using HISAT2. Differential expression analysis for mRNA was performed using the R package edgeR. Differentially expressed RNAs with |log₂(FC)| value >1 and q value <0.05, considered significantly modulated, were retained for volcano plot and GO analysis.

Quantitative real-time PCR

For quantitative real-time PCR, total RNA was isolated using the TRIzol and chloroform liberation method. First-strand cDNA synthesis was done using the First Strand cDNA Synthesis Kit (K1651; Thermo Fisher). Quantitative PCR reactions were performed using Power SYBR^® Green PCR Master Mix (4367659; Applied Biosystems) and a 7500 Fast Real-Time PCR System (4484643, Applied Biosystems). The primer pairs used are listed in Supplementary Fig. 9. PCR products were qualified by the comparative Ct (threshold cycle) method and the copy number for each transcript was expressed relative to that of the housekeeping gene ACTB. The software used for collecting qRT-PCR data was Bio-Rad CFX manager 3.1.

Statistics and reproducibility

Statistical significance was calculated by GraphPad Prism 8 software with Tukey’s multiple comparisons test of ANOVA (for the result in Supplementary Fig. 3d, Supplementary Fig. 4b, c, g, j, Supplementary Fig. 5a, c, e, Fig. 3e, Supplementary Fig. 10, Supplementary Fig. 11b, f, Figs. 5g, 6b, e, g and Supplementary Fig. 17d, unpaired, two-tailed) and ordinary one-way ANOVA (Supplementary Fig. 1b, d). The confidence interval was 95%. Results were considered significant at P < 0.05. Statistical analysis between one group and the rest of the scRNA-seq data (for the result in Supplementary Data 3–5, 8, 11 and 13) were made by the statistical analysis was made by Wilcoxon (Mann–Whitney) rank-sum test; two-tailed, with the Seurat package in R, and the confidence interval is 95%. Statistical parameters are reported in the respective figures and figure legends. Masson and IHC staining for each disease type (for the result in Figs. 4c, 5d–f, 6b, g, h; Supplementary Figs. 1b, 3c, 8c, 9a–f, 12b, 13c, d, 16d, 17c) in this study were based on at least two biological repeats. Scanning electron microscope and transmission electron microscope image (Fig. 6a, e) were based on two biological repeats.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

All scRNA-seq and bulk RNA-seq raw data generated for this study have been deposited in the National Omics Data Encyclopedia database repository “OEP002406”, Genome Sequence Archive “PRJCA009769” and GEO database “GSE206528”. The cell–gene matrix has been uploaded in “GitHub [https://github.com/zlyingithub/single-cell-sequence-of-human-corpora-cavernosa]”. All other relevant data supporting the key findings of this study are available within the article and its Supplementary Information files or from the corresponding author upon reasonable request. Source data are provided with this paper.

Code availability

All code associated with this manuscript and the gene annotation list have been uploaded to “GitHub [https://github.com/zlyingithub/single-cell-sequence-of-human-corpora-cavernosa]” and “Zenodo [https://zenodo.org/record/6596754#.YqNdy3ZBwxY]”.

References

Shafik, A., Shafik, I., El Sibai, O. & Shafik, A. A. On the pathogenesis of penile venous leakage: role of the tunica albuginea. BMC Urol. 7, 14 (2007).
Article PubMed PubMed Central Google Scholar
Shafik, A., el-Sibai, O. & Shafik, A. A. Magnetic stimulation of the cavernous nerve for the treatment of erectile dysfunction in humans. Int. J. Impot. Res. 12, 137–141 (2000).
Article CAS PubMed Google Scholar
Ding, Y. Q., Takada, M., Kaneko, T. & Mizuno, N. Colocalization of vasoactive intestinal polypeptide and nitric oxide in penis-innervating neurons in the major pelvic ganglion of the rat. Neurosci. Res. 22, 129–131 (1995).
Article CAS PubMed Google Scholar
Alwaal, A., Zaid, U. B., Lin, C. S. & Lue, T. F. Stem cell treatment of erectile dysfunction. Adv. Drug Deliv. Rev. 82–83, 137–144 (2015).
Article PubMed CAS Google Scholar
Terentes-Printzios, D, Ioakeimidis, N., Rokkas, K. & Vlachopoulos, C. Interactions between erectile dysfunction, cardiovascular disease and cardiovascular drugs. Nat. Rev. Cardiol. https://doi.org/10.1038/s41569-021-00593-6 (2021).
Miner, M. M. Erectile dysfunction: a harbinger or consequence: does its detection lead to a window of curability? J. Androl. 32, 125–134 (2011).
Article PubMed Google Scholar
Ballard, S. A. et al. Effects of sildenafil on the relaxation of human corpus cavernosum tissue in vitro and on the activities of cyclic nucleotide phosphodiesterase isozymes. J. Urol. 159, 2164–2171 (1998).
Article CAS PubMed Google Scholar
Carson, C. C. & Lue, T. F. Phosphodiesterase type 5 inhibitors for erectile dysfunction. BJU Int. 96, 257–280 (2005).
Article CAS PubMed Google Scholar
Chung, E., De Young, L. & Brock, G. B. Investigative models in erectile dysfunction: a state-of-the-art review of current animal models. J. Sex. Med. 8, 3291–3305 (2011).
Article PubMed Google Scholar
Zhao, L. et al. VEGFC/VEGFR3 signaling regulates mouse spermatogonial cell proliferation via the activation of AKT/MAPK and cyclin D1 pathway and mediates the apoptosis by affecting caspase 3/9 and Bcl-2. Cell Cycle 17, 225–239 (2018).
Article CAS PubMed PubMed Central Google Scholar
Zhao, L. et al. Intra-seminiferous tubular injection of vascular endothelial growth factor C sustained-release ultrafine particles: a novel method for improving the regeneration of spermatogenesis after chemotherapy. J. Biomed. Nanotechnol. 15, 2376–2392 (2019).
Article CAS PubMed Google Scholar
Rao, M. et al. Resolving the intertwining of inflammation and fibrosis in human heart failure at single-cell level. Basic Res. Cardiol. 116, 55 (2021).
Article PubMed Google Scholar
De Micheli, A. J., Spector, J. A., Elemento, O. & Cosgrove, B. D. A reference single-cell transcriptomic atlas of human skeletal muscle tissue reveals bifurcated muscle stem cell populations. Skelet. Muscle 10, 19 (2020).
Article PubMed PubMed Central CAS Google Scholar
Kinchen, J. et al. Structural remodeling of the human colonic mesenchyme in inflammatory Bowel disease. Cell 175, 372–86 e17 (2018).
Article CAS PubMed PubMed Central Google Scholar
Ramachandran, P. et al. Resolving the fibrotic niche of human liver cirrhosis at single-cell level. Nature 575, 512–518 (2019).
Article ADS CAS PubMed PubMed Central Google Scholar
Kuppe, C. et al. Decoding myofibroblast origins in human kidney fibrosis. Nature 589, 281–286 (2021).
ADS CAS PubMed Google Scholar
Deng, C. C. et al. Single-cell RNA-seq reveals fibroblast heterogeneity and increased mesenchymal fibroblasts in human fibrotic skin diseases. Nat. Commun. 12, 3709 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Theocharidis, G. et al. Single cell transcriptomic landscape of diabetic foot ulcers. Nat. Commun. 13, 181 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Reyfman, P. A. et al. Single-cell transcriptomic analysis of human lung provides insights into the pathobiology of pulmonary fibrosis. Am. J. Resp. Crit. Care Med. 199, 1517–1536 (2019).
Article CAS PubMed PubMed Central Google Scholar
Habermann, A. C. et al. Single-cell RNA sequencing reveals profibrotic roles of distinct epithelial and mesenchymal lineages in pulmonary fibrosis. Sci. Adv. 6, eaba1972 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Buechler, M. B. et al. Cross-tissue organization of the fibroblast lineage. Nature 593, 575–579 (2021).
Article ADS CAS PubMed Google Scholar
Kimura, M. et al. Homeobox A4 suppresses vascular remodeling by repressing YAP/TEAD transcriptional activity. EMBO Rep. 21, e48389 (2020).
Article CAS PubMed PubMed Central Google Scholar
Rebolledo, D. L., Acuna, M. J. & Brandan, E. Role of matricellular CCN proteins in skeletal muscle: focus on CCN2/CTGF and its regulation by vasoactive peptides. Int. J. Mol. Sci. 22, https://doi.org/10.3390/ijms22105234 (2021).
Castela, Â. & Costa, C. Molecular mechanisms associated with diabetic endothelial-erectile dysfunction. Nat. Rev. Urol. 13, 266–274 (2016).
Article CAS PubMed Google Scholar
Kalucka, J. et al. Single-cell transcriptome atlas of murine endothelial cells. Cell 180, 764–79 e20 (2020).
Article CAS PubMed Google Scholar
Wang, K. et al. Single-cell transcriptomic analysis of somatosensory neurons uncovers temporal development of neuropathic pain. Cell Res. 31, 904–918 (2021).
Article CAS PubMed Google Scholar
Li, H. et al. Initiation of PI3K/AKT pathway by IGF-1 decreases spinal cord injury-induced endothelial apoptosis and microvascular damage. Life Sci. 263, 118572 (2020).
Article CAS PubMed Google Scholar
Ahluwalia, A., Jones, M. K., Hoa, N. & Tarnawski, A. S. NGF protects endothelial cells from indomethacin-induced injury through activation of mitochondria and upregulation of IGF-1. Cell. Signal. 40, 22–29 (2017).
Article CAS PubMed Google Scholar
Zhang, J. et al. Ginsenoside F1 promotes angiogenesis by activating the IGF-1/IGF1R pathway. Pharmacol. Res. 144, 292–305 (2019).
Article CAS PubMed Google Scholar
Dieffenbach, P. B., Aravamudhan, A., Fredenburgh, L. E. & Tschumperlin, D. J. The mechanobiology of vascular remodeling in the aging lung. Physiology https://doi.org/10.1152/physiol.00019.2021 (2021).
Huang, T. et al. Structural and functional abnormalities of penile cavernous endothelial cells result in erectile dysfunction at experimental autoimmune prostatitis rat. J. Inflamm. 16, 20 (2019).
Article CAS Google Scholar
Zeisberg, E. M., Potenta, S. E., Sugimoto, H., Zeisberg, M. & Kalluri, R. Fibroblasts in kidney fibrosis emerge via endothelial-to-mesenchymal transition. J. Am. Soc. Nephrology: JASN 19, 2282–2287 (2008).
Article PubMed Central Google Scholar
Wang, W. et al. Integrin β3 mediates the endothelial-to-mesenchymal transition via the notch pathway. Cell. Physiol. Biochem. 49, 985 (2018).
Article CAS PubMed Google Scholar
Lin, F. Y. et al. A novel relative high-density lipoprotein index to predict the structural changes in high-density lipoprotein and its ability to inhibit endothelial-mesenchymal transition. Int. J. Mol. Sci. 22, https://doi.org/10.3390/ijms22105210 (2021).
Lin, J. R. et al. Suppression of endothelial-to-mesenchymal transition by SIRT (sirtuin) 3 alleviated the development of hypertensive renal injury. Hypertension 72, 350–360 (2018).
Article CAS PubMed Google Scholar
Li, J. et al. YAP-induced endothelial-mesenchymal transition in oral submucous fibrosis. J. Dent. Res. 98, 920–929 (2019).
Article CAS PubMed Google Scholar
Wei, W. Y. et al. Pioglitazone alleviates cardiac fibrosis and inhibits endothelial to mesenchymal transition induced by pressure overload. Cell. Physiol. Biochem. 45, 26–36 (2018).
Article CAS PubMed Google Scholar
Savorani, C. et al. A dual role of YAP in driving TGFbeta-mediated endothelial-to-mesenchymal transition. J. Cell Sci. 134, https://doi.org/10.1242/jcs.251371 (2021).
Hennefarth, M. R. et al. Physicochemical and biochemical spatiotemporal maps of a mouse penis. J. Biomech. 101, 109637 (2020).
Article PubMed Google Scholar
Lin, G. et al. Novel therapeutic approach for neurogenic erectile dysfunction: effect of neurotrophic tyrosine kinase receptor type 1 monoclonal antibody. Eur. Urol. 67, 716–726 (2015).
Article CAS PubMed Google Scholar
Frangogiannis, N. G. Cardiac fibrosis. Cardiovasc. Res. 117, 1450–1488 (2021).
Article CAS PubMed Google Scholar
Levay, A. K. et al. Scleraxis is required for cell lineage differentiation and extracellular matrix remodeling during murine heart valve formation in vivo. Circulation Res. 103, 948–956 (2008).
Article CAS PubMed Google Scholar
Działo, E. et al. WNT/β-catenin signaling promotes TGF-β-mediated activation of human cardiac fibroblasts by enhancing IL-11 production. Int. J. Mol. Sci. 22, https://doi.org/10.3390/ijms221810072 (2021).
Kadota, T. et al. Human bronchial epithelial cell-derived extracellular vesicle therapy for pulmonary fibrosis via inhibition of TGF-β-WNT crosstalk. J. Extracell. Vesicles 10, e12124 (2021).
Article CAS PubMed PubMed Central Google Scholar
Guo, Y. et al. Inhibition of proliferation-linked signaling cascades with atractylenolide I reduces myofibroblastic phenotype and renal fibrosis. Biochem. Pharmacol. 183, 114344 (2021).
Article CAS PubMed Google Scholar
He, W. et al. Alterations in the phosphodiesterase type 5 pathway and oxidative stress correlate with erectile function in spontaneously hypertensive rats. J. Cell. Mol. Med. 24, 14280–14292 (2020).
Article CAS PubMed PubMed Central Google Scholar
Hassanin, A. M. & Abdel-Hamid, A. Z. Cavernous smooth muscles: innovative potential therapies are promising for an unrevealed clinical diagnosis. Int. Urol. Nephrol. 52, 205–217 (2020).
Article CAS PubMed Google Scholar
Kreplak, L. & Bär, H. Severe myopathy mutations modify the nanomechanics of desmin intermediate filaments. J. Mol. Biol. 385, 1043–1051 (2009).
Article CAS PubMed Google Scholar
Kiss, B., Röhlich, P. & Kellermayer, M. S. Structure and elasticity of desmin protofibrils explored with scanning force microscopy. J. Mol. Recognit.: JMR 24, 1095–1104 (2011).
Article PubMed Google Scholar
Bai, W. J. et al. An open-label, multicenter, randomized, crossover study comparing sildenafil citrate and tadalafil for treating erectile dysfunction in Chinese men naïve to phosphodiesterase 5 inhibitor therapy. Asian J. Androl. 17, 61–67 (2015).
Article CAS PubMed Google Scholar
Montorsi, F. et al. A randomized, double-blind, placebo-controlled, parallel study to assess the efficacy and safety of once-a-day tadalafil in men with erectile dysfunction who are naïve to PDE5 inhibitors. J. Sex. Med. 8, 2617–2624 (2011).
Article CAS PubMed Google Scholar
Kalluri, A. S. et al. Single-cell analysis of the normal mouse aorta reveals functionally distinct endothelial cell populations. Circulation 140, 147–163 (2019).
Article CAS PubMed PubMed Central Google Scholar
Li, Q. et al. Single-cell transcriptome profiling reveals vascular endothelial cell heterogeneity in human skin. Theranostics 11, 6461–6476 (2021).
Article CAS PubMed PubMed Central Google Scholar
Qin, Z., Xia, W., Fisher, G. J., Voorhees, J. J. & Quan, T. YAP/TAZ regulates TGF-β/Smad3 signaling by induction of Smad7 via AP-1 in human skin dermal fibroblasts. Cell Commun. Signal 16, 18, https://doi.org/10.1186/s12964-018-0232-3 (2018).
Article CAS PubMed PubMed Central Google Scholar
Zhao, L. et al. Single-cell analysis of developing and azoospermia human testicles reveals central role of Sertoli cells. Nat. Commun. 11, https://doi.org/10.1038/s41467-020-19414-4 (2020).

Download references

Acknowledgements

This work was supported by grants from Shanghai Scientific and Technological Project (21410760300, Z.L.); China Postdoctoral Science Foundation (2021M703747, L.Z.); GuangDong Basic and Applied Basic Research Foundation (2021A1515111109, L.Z.); National Natural Science Foundation of China (82171597, Z.L.; 81871215, Z.L.; and 82071636, Y.T.); the open funds of Guangdong Provincial Key Lab of Biomedical Imaging (GPKLBI202104, L.Z.). We wish to thank the Sinotech Genomics CO. LTD. Shanghai for their support of RNA-seq library preparation. We thank Dr. Jianming Zeng and his bioinformatics team for their reference codes. We thank Catherine Perfect, MA (Cantab), from Liwen Bianji (Edanz) (www.liwenbianji.cn/), for editing the English text of a draft of this manuscript.

Author information

These authors contributed equally: LiangYu Zhao, Sha Han, HengChuan Su, JianYing Li.

Authors and Affiliations

Department of Andrology, the Center for Men’s Health, Urologic Medical Center, Shanghai Key Laboratory of Reproductive Medicine, Shanghai General Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, 200080, China
LiangYu Zhao, Sha Han, JianYing Li, ErLei Zhi, Peng Li, ChenCheng Yao, RuHui Tian, HuiXing Chen, HuiRong Chen, JiaQiang Luo, ChenKun Shi, ZhiYong Ji, JianLin Hu & Zheng Li
Department of Urology, Department of Interventional Medicine, Guangdong Provincial Key Laboratory of Biomedical Imaging, The Fifth Affiliated Hospital, Sun Yat-sen University, Zhuhai, Guangdong Province, 519000, China
LiangYu Zhao & YuXin Tang
Department of Urology, Fudan University Shanghai Cancer Center, Department of Oncology, Shanghai Medical College, Fudan University, Shanghai, 200032, China
HengChuan Su & DengLong Wu
Department of Urology, Tongji Hospital, School of Medicine, Tongji University, Shanghai, 200065, China
Gang Wu & WeiDong Zhou
Department of Ultrasound, The Fifth Affiliated Hospital, Sun Yat-sen University, Zhuhai, Guangdong Province, 519000, China
YuZhuo Chen
Knuppe Molecular Urology Laboratory, Department of Urology, School of Medicine, University of California, San Francisco, CA, USA
GuiTing Lin & Tom F. Lue

Authors

LiangYu Zhao
View author publications
You can also search for this author in PubMed Google Scholar
Sha Han
View author publications
You can also search for this author in PubMed Google Scholar
HengChuan Su
View author publications
You can also search for this author in PubMed Google Scholar
JianYing Li
View author publications
You can also search for this author in PubMed Google Scholar
ErLei Zhi
View author publications
You can also search for this author in PubMed Google Scholar
Peng Li
View author publications
You can also search for this author in PubMed Google Scholar
ChenCheng Yao
View author publications
You can also search for this author in PubMed Google Scholar
RuHui Tian
View author publications
You can also search for this author in PubMed Google Scholar
HuiXing Chen
View author publications
You can also search for this author in PubMed Google Scholar
HuiRong Chen
View author publications
You can also search for this author in PubMed Google Scholar
JiaQiang Luo
View author publications
You can also search for this author in PubMed Google Scholar
ChenKun Shi
View author publications
You can also search for this author in PubMed Google Scholar
ZhiYong Ji
View author publications
You can also search for this author in PubMed Google Scholar
JianLin Hu
View author publications
You can also search for this author in PubMed Google Scholar
Gang Wu
View author publications
You can also search for this author in PubMed Google Scholar
WeiDong Zhou
View author publications
You can also search for this author in PubMed Google Scholar
YuXin Tang
View author publications
You can also search for this author in PubMed Google Scholar
YuZhuo Chen
View author publications
You can also search for this author in PubMed Google Scholar
GuiTing Lin
View author publications
You can also search for this author in PubMed Google Scholar
Tom F. Lue
View author publications
You can also search for this author in PubMed Google Scholar
DengLong Wu
View author publications
You can also search for this author in PubMed Google Scholar
Zheng Li
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

L.Z., Z.L., D.W., J.L. and T.L. designed the experiments. H.S., E.Z., P.L., C.Y., R.T., H.C. (HuiXing Chen), J.H. and H.C. (Huirong Chen) collected the samples and clinical information. Z.L., J.L. and C.S. performed immunostaining. Z.J., G.W., W.Z., and L.Z. isolated cells for scRNA-seq. Z.L., J.L. and S.H. performed cell culture experiments. L.Z. and Y.T. performed bioinformatics analyses. G.L. interpreted data. Y.C. drew the Schematic illustration. L.Z., Z.L. and T.L. wrote the manuscript. All authors revised and approved the manuscript.

Corresponding authors

Correspondence to Tom F. Lue, DengLong Wu or Zheng Li.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Description of additional Supplementary File

Supplementary Dataset 1

Supplementary Dataset 2

Supplementary Dataset 3

Supplementary Dataset 4

Supplementary Dataset 5

Supplementary Dataset 6

Supplementary Dataset 7

Supplementary Dataset 8

Supplementary Dataset 9

Supplementary Dataset 10

Supplementary Dataset 11

Supplementary Dataset 12

Supplementary Dataset 13

Supplementary Dataset 14

Supplementary Dataset 15

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Zhao, L., Han, S., Su, H. et al. Single-cell transcriptome atlas of the human corpus cavernosum. Nat Commun 13, 4302 (2022). https://doi.org/10.1038/s41467-022-31950-9

Download citation

Received: 11 November 2021
Accepted: 11 July 2022
Published: 25 July 2022
DOI: https://doi.org/10.1038/s41467-022-31950-9

This article is cited by

Estrogenic endocrine disruptor exposure directly impacts erectile function
- Samuel M. Cripps
- Sarah A. Marshall
- Andrew J. Pask
Communications Biology (2024)
Single-cell RNA sequencing reveals critical modulators of extracellular matrix of penile cavernous cells in erectile dysfunction
- Chao Luo
- Yaqian Peng
- Anyang Wei
Scientific Reports (2024)
The role of mechano-regulated YAP/TAZ in erectile dysfunction
- Mintao Ji
- Dongsheng Chen
- Lei Chang
Nature Communications (2023)
Comprehensive analysis of biological landscape of oxidative stress-related genes in diabetic erectile dysfunction
- Qingjun Meng
- Yinwei Chen
- Penghui Yuan
International Journal of Impotence Research (2023)
Vsig4+ resident single-Kupffer cells improve hepatic inflammation and fibrosis in NASH
- Hui-Yi Li
- Seng-Wang Fu
- Ming-Yi Xu
Inflammation Research (2023)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.