Functional Segregation of Pancreatic Microcirculation Reveals Sex‐Dependent Microhemodynamic Signatures
Yuan Li, Yingyu Wang, Bing Wang, Qin Ouyang, Weiqi Liu, Xiang Xu, Xu Zhang, Mingming Liu, Ruijuan Xiu
TL;DR
The pancreatic microcirculation shows sex-specific differences in blood flow regulation, influenced by hormones and genetic factors.
Contribution
The study reveals sex-dependent microhemodynamic signatures in the pancreatic exocrine compartment linked to steroid-hormone interactions.
Findings
Male pancreatic blood flow is directly coupled to systemic blood pressure, while females show ERβ-mediated oscillatory patterns.
Sexual dimorphism in microcirculation is anatomically restricted to the exocrine compartment, not the endocrine islets.
Genetic background modulates steroid responsiveness and affects microvascular adaptability.
Abstract
The microcirculation is a determinant of organ function, translating systemic signals into local physiological responses. However, whether the regulation of microhemodynamics is sexually dimorphic within the pancreas has remained unknown. Here, using a multi‐scale approach in healthy mice, we report a sexual dimorphism in pancreatic microhemodynamics that is anatomically restricted to the exocrine compartment, defined by divergent expression of the endothelial marker CD31 and estrogen receptor ERβ, whereas the endocrine islet microvasculature remains conserved between sexes. We demonstrate that microhemodynamic dimorphism is functionally coupled to divergent systemic steroid hormone profiles, a male signature characterized by elevated androgens and steroidogenic precursors, and a female signature dominated by glucocorticoids and estrogen metabolites. The distinction manifests as…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7- —Beijing Municipal Natural Science Foundation10.13039/501100005089
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsNeuropeptides and Animal Physiology · Pancreatic function and diabetes · Hypothalamic control of reproductive hormones
Introduction
1
Sex is a fundamental biological variable that dictates differences in physiology and disease susceptibility (vom Steeg and Klein 2016; Ober et al. 2008). The incidence, progression, and pathophysiology of metabolic diseases exhibit significant sexual dimorphism (Tramunt et al. 2020; Colafella and Denton 2018). Clinically, premenopausal women demonstrate a lower incidence of type 2 diabetes and superior insulin sensitivity compared to age‐matched men. Although these disparities are frequently attributed to the direct action of sex steroids on β‐cells, the vascular component remains a critical, yet overlooked, factor. Given that adequate islet perfusion is a prerequisite for efficient hormone secretion and that the vasculature is a known target of estrogenic signaling, we posit that these systemic metabolic differences are underpinned by distinct microvascular regulatory strategies. Consequently, understanding the pancreas, the central regulator of both glucose homeostasis and digestion, requires looking beyond the endocrine cells alone. Although functional sex differences in pancreatic endocrine output, such as insulin secretion, have been described (Yong et al. 2022; Qadir et al. 2024), the anatomical locus and the upstream regulatory principles governing these sex‐specific phenotypes are poorly defined.
Although the endocrine and exocrine compartments were historically studied in isolation, modern evidence has established the pancreas as a single integrated organ defined by extensive crosstalk (Villaca and Mastracci 2024; Overton and Mastracci 2022; Mostafa et al. 2024), the functional integration of which is physically mediated by the pancreatic microcirculation (Dybala et al. 2020). The islet‐acinar portal system, where blood leaving the islets subsequently perfuses the surrounding acinar tissue, creates a local communication axis of unparalleled hormone concentration (Valente et al. 2024; Mastracci et al. 2023; Slak Rupnik and Hara 2024). The microvascular network is therefore a primary candidate for translating systemic signals into coordinated organ‐wide physiological responses.
Despite its important role as an integrator, the pancreatic microvasculature itself has been largely overlooked as a potential site of primary sexual dimorphism. It remains unknown whether the regulation of pancreatic microhemodynamics differs between males and females, and how such differences might be coupled to the distinct systemic hormonal milieus that characterize the sexes. Addressing this knowledge gap is essential, as it could provide a tissue‐level mechanism to explain how systemic sex differences are translated into organ‐specific function and pathology.
Here, we address this question through a multi‐scale investigation across three distinct mouse strains integrating microcirculatory analysis and targeted steroid metabolomics. Our findings reveal that pancreatic sexual dimorphism is not distributed but is instead anatomically confined to the exocrine compartment. We demonstrate that this localized dimorphism is functionally coupled to fundamentally divergent, sex‐specific systemic steroid hormone strategies, proposing a model in which the exocrine microvasculature acts as a key functional integrator, translating the systemic hormonal state into a sex‐specific local environment that governs overall pancreatic homeostasis.
Materials and Methods
2
Animal Procedures and Metabolic Phenotyping
2.1
All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of the Institute of Microcirculation, CAMS (approval no. CAMS‐IM‐IACUC‐2023‐AE0937). Eight‐week‐old male and female BALB/c, C57BL/6J, and Kunming mice (Vital River Laboratory Animal Technology, Beijing, China) were housed under controlled conditions (12:12 h light–dark cycle, 26°C, 55%–70% humidity) with ad libitum access to food and water. Metabolic function was assessed via intraperitoneal glucose tolerance tests (IGTT) and insulin tolerance tests (ITT). Following a 16 h (IGTT) or 4 h (ITT) fast, mice received an intraperitoneal (i.p.) injection of either glucose (2 g/kg body weight; Merck KGaA) or insulin (0.5 U/kg body weight; Eli Lilly), respectively. Blood glucose concentrations were measured from the tail vein using a glucometer (One Touch UltraEasy) at baseline (0 min) and at 15, 30, 60, 90, and 120 min post‐injection. For terminal biochemical assays, blood was collected from the inferior vena cava and processed to obtain serum. Serum lipase concentrations were subsequently quantified using an automated chemistry analyzer (Atellica Solution; Siemens).
Monitoring of Macro‐Circulation
2.2
Non‐invasive blood pressure (BP) and heart rate (HR) were measured using a computerized tail‐cuff system (BP‐2010A; Softron Biotechnology, Beijing, China). Mice were placed in individual restrainers on a heating pad (TMC‐213) pre‐warmed to 37°C to maintain body temperature and prevent heat‐loss‐induced vasoconstriction. Following a 5‐min acclimation period, systolic, diastolic, and mean arterial pressures were recorded. To ensure data stability and accuracy, measurements were taken in triplicate for each animal and the average value was used for analysis.
Quantification of Pancreatic Microhemodynamics by Laser Doppler Flowmetry
2.3
Pancreatic microcirculatory parameters were quantified using a MoorVMS‐LDF2 laser Doppler flowmetry (LDF) system (Moor Instruments, Axminster, UK). Prior to the procedure, mice were acclimatized to the laboratory environment for at least 60 min. Anesthesia was induced and maintained with 2% isoflurane (R510‐22; RWD Life Science, Shenzhen, China) in a 1:1 mixture of medical air and oxygen, delivered via a calibrated vaporizer and a non‐rebreathing circuit (matrix VMR; Midmark Corporation, OH, USA). Body temperature was maintained at 37.0°C ± 0.5°C throughout the procedure using a homeothermic blanket system.
Following induction of surgical anesthesia, a midline laparotomy was performed to expose the pancreas. The pancreas was kept moist with pre‐warmed sterile saline. The LDF probe (VP4; Moor Instruments) was mounted on a micromanipulator and positioned perpendicularly and stably on the surface of the pancreas, ensuring minimal pressure. After a 5‐min stabilization period to allow the signal to equilibrate, microhemodynamic data were continuously recorded for 1 min from three distinct, randomly selected sites on the pancreatic surface. The system recorded three primary parameters: blood perfusion (Flux), erythrocyte concentration (Conc), and blood flow velocity (Velocity). Given that the exocrine acini comprise approximately 98% of the total pancreatic mass (Alvarez Fallas et al. 2021), the recorded signaling is predominantly representative of the exocrine microcirculation. Consequently, we utilized LDF to assess global/exocrine hemodynamics, whereas employing immunohistochemistry to spatially resolve and quantify islet‐specific vascular phenotypes.
Preprocessing and Normalization of Microhemodynamic Signals
2.4
The raw time‐series data from the LDF recordings were subjected to a standardized preprocessing pipeline. Outliers in the signal, defined as data points falling outside 1.5 times the interquartile range (IQR) below the first quartile (Q_1_) or above the third quartile (Q_3_), were adjusted using Winsorization, where each outlier was replaced by the nearest value within the Q_1_–1.5 × IQR, Q_3_ + 1.5 × IQR range. To ensure comparability for subsequent analyses, time‐series data for each recording were standardized to a uniform length. To facilitate comparison across different animals and scales, the processed data were then subjected to min‐max normalization, transforming values to a dimensionless range using the formula: x′ = (x − min)/(max − min), where min and max represent the minimum and maximum values of the microcirculatory data, respectively.
Time‐Frequency Analysis of Microhemodynamic Oscillations
2.5
To resolve the oscillatory components of the pancreatic microhemodynamic signal, a continuous wavelet transform (CWT) was performed on the preprocessed time‐series data by MATLAB (R2025a; MathWorks, MA, USA). The Morlet wavelet was employed as the mother wavelet, owing to its optimal trade‐off between time and frequency resolution for the analysis of non‐stationary biological signals (Addison et al. 2009; Li et al. 2023). The CWT decomposed the signal into time‐scale space using the fundamental operation defined as the inner product:
Here, Wfa,b represents the complex wavelet coefficients, ft is the input signal, and ψt is the mother wavelet function with ψ*t being its complex conjugate. The scale parameter a controls the wavelet's width (related to frequency), whereas the translation parameter b localizes the analysis in time, and the factor 1a ensures energy conservation. The notation .. represents the inner product.
The time‐frequency analysis yielded a three‐dimensional scalogram, permitting the decomposition of blood flow oscillations into frequency intervals attributed to distinct physiological regulators, as defined by the standardized frequency ranges originally proposed by Stefanovska et al. (1999) and subsequently validated through pharmacological interventions (Kvernmo et al. 1998; Kralj and Lenasi 2022). Our analysis focused on the endothelial contributions, which were partitioned into two distinct bands. The NO‐independent band (0.005–0.0095 Hz), which reflects endothelial mechanisms not mediated by nitric oxide and metabolic regulation, and the NO‐dependent band (0.0095–0.04 Hz), which has been demonstrated to be significantly attenuated by NOS inhibition (Kvandal et al. 2006). Our analysis focused on the endothelial contributions, which were partitioned into the NO‐independent (0.005–0.0095 Hz) and NO‐dependent (0.0095–0.04 Hz) frequency amplitudes. For quantitative comparison, the mean magnitude within each of these two endothelial intervals was calculated from the scalogram, serving as a metric to assess the contribution of these specific oscillators to pancreatic microcirculatory regulation.
Histology and Immunohistochemistry
2.6
Pancreatic tissue was processed into 4% paraformaldehyde‐fixed, paraffin‐embedded sections. Following deparaffinization and rehydration, sections were utilized for histological or immunohistochemical analysis. For histology, sections were stained with hematoxylin and eosin (H&E) via standard protocols and examined under a light microscope (Leica DFC450). For immunohistochemistry (IHC), sections underwent heat‐induced antigen retrieval (10 mM citrate buffer, pH 6.0), followed by quenching of endogenous peroxidase activity with 3% hydrogen peroxide and blocking of non‐specific binding with 3% BSA. Sections were subsequently incubated overnight at 4°C with primary antibodies targeting CD31 (1:25; Santa Cruz, sc‐101454), estrogen receptor alpha (ER‐α, 1:25; Abcam, ab32063), and estrogen receptor beta (ER‐β, 1:25; Abcam, ab187291). After incubation with appropriate secondary antibodies, immunoreactivity was visualized using a 3,3′‐diaminobenzidine (DAB) substrate kit (Zhongshan Golden Bridge) and sections were counterstained with hematoxylin. Microvascular density (MVD) was assessed by manual enumeration of CD31‐positive microvessels (six random ×40 fields), whereas ER‐α and ER‐β expression was quantified by measuring the optical density (OD) of the DAB signal using ImageJ software (National Institutes of Health, Bethesda, MD, USA) with the Fiji color deconvolution plugin (Fiji team, https://fiji.sc/). All vessel counting and optical density measurements were performed by an observer (X.Z.) who was blinded to the sex and strain of the animals to minimize potential bias.
Steroid Hormones Target Metabolomics
2.7
After centrifugation, the serum were subjected to targeted metabolomic analysis employing a liquid chromatography–tandem mass spectrometry (LC–MS) platform (QTRAP 6500+ SCIEX; APExBIO Technology LLC, Shanghai, China). Chromatographic resolution was achieved using a binary mobile phase system on a Phenomenex Kinetics C18 column (2.1 mm internal diameter × 100 mm length, 1.7 μm particle size). The mobile phase composition included: (A) acetonitrile‐aqueous solution (30:70, v/v) containing 0.04% acetic acid, and (B) a mixture of acetonitrile‐isopropanol (50:50, v/v) supplemented with 0.04% acetic acid. A temperature‐controlled column compartment (40°C) operated at a constant flow rate of 0.35 mL/min, with 5 μL sample injections.
The gradient program initiated with 95% phase A (5% B) for 1 min, followed by a linear transition to 10% A (90% B) over 9 min (1–10 min). This ratio persisted until 12.5 min, after which rapid column re‐equilibration (95% A) occurred from 12.6 to 15 min. Mass detection was conducted in positive ESI mode with optimized source parameters: ion spray voltage 5500 V, source temperature 550°C, and curtain gas at 35 psi. Multiple reaction monitoring transitions were executed on a triple quadrupole analyzer, with declustering potentials and collision energies individually optimized and calibrated for maximal sensitivity.
Multivariate pattern recognition through principal component analysis (PCA) evaluated global metabolic disparities among sample groups. Univariate statistical comparisons employed Student's t‐test for pairwise analyses and one‐way ANOVA for multi‐group assessments, maintaining original metabolite concentration data integrity throughout computational processing.
Statistical Analysis
2.8
All quantitative data were subjected to statistical analysis using GraphPad Prism (version 10.2.0; GraphPad Software, CA, USA) and R (version 4.4.3; http://www.R‐project.org) via the RStudio (version 2024.12.1+563; Posit Software, MA, USA). Data are presented as mean ± standard deviation (SD). The statistical significance threshold was set at p < 0.05. Prior to comparative analysis, datasets were assessed for normality using the Shapiro–Wilk test and for homogeneity of variances using Levene's test. Then unpaired t‐test was performed between male and female groups. For comparisons between two groups where data did not meet the assumptions for parametric testing, the non‐parametric Mann–Whitney U‐test was employed. To assess the linear relationships between circulating steroid hormone concentrations and pancreatic microhemodynamic parameters, Pearson correlation analysis was performed. Correlation matrices were generated for each sex and strain combination to resolve interdependencies. The resulting Pearson correlation coefficient (r) and corresponding p‐values were calculated to determine the strength and significance of associations.
Results
3
Characterization of Systemic Metabolic and Macrovascular Phenotypes
3.1
To establish the systemic characterization for our investigation of pancreatic microcirculation, we first characterized the basal metabolic profiles of male and female mice across BALB/c, C57BL/6J, and Kunming strains. A fundamental sexual dimorphism was observed in body weight, with male mice exhibiting significantly higher body mass than their female counterparts across all three strains (all p < 0.0001, Figure 1A). In the assessment of glucose homeostasis, a strain‐specific sex difference in fasting blood glucose was identified exclusively in C57BL/6J mice, where males displayed higher glucose levels than females (p < 0.001). No significant sex‐based differences in fasting glucose were detected in either BALB/c or Kunming mice, and serum lipase levels were also comparable between sexes in BALB/c mice (p > 0.05, Figure 1B,C). It is noted that all fasting glucose values remained within the normal physiological range. Dynamic testing of glucose metabolism revealed further layers of sexual dimorphism. Although IGTT showed no significant differences in the AUC between sexes in any strain. In contrast, ITT demonstrated a consistent sexual dimorphism in systemic insulin action. Specifically, male mice from all three strains exhibited a significantly larger AUC during the ITT compared to their female counterparts (p < 0.05, Figure 1D), a finding indicative of a relatively lower insulin sensitivity in males of these strains.
*Systemic metabolic and macrocirculatory characterization among BALB/c, C57BL/6J, and Kunming mice. (A) Body weight in male and female mice across BALB/c, C57BL/6J, and Kunming strains. (B) Levels of lipase in serum. (C) Fasting blood glucose levels after a 6‐h fast. The dotted line represents the hyperglycemic threshold 11.1 mmol/L (200 mg/dL). (D) Comparations of IGTT and ITT between male and female mice. (E–H) Quantification of macrovascular parameters including HR (E), SBP (F), DBP (G), and MAP (H). DBP, diastolic blood pressure; HR, heart rate; IGTT, intraperitoneal glucose tolerance test; ITT, intraperitoneal insulin tolerance test; MAP, mean arterial pressure; SBP, systolic blood pressure. n = 6 per group. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ***p < 0.0001.
Macrovascular profiling also revealed a strain‐dependent sexual dimorphism. HR dimorphism was directionally opposite depending on the genetic background; female BALB/c mice displayed a higher HR than males, whereas male Kunming mice exhibited a higher HR than females (both p < 0.01, Figure 1E). No sex difference in HR was observed in C57BL/6J mice. For blood pressure, a significant sexual dimorphism was limited to SBP in the BALB/c strain, where males had higher SBP than females (p < 0.05, Figure 1F). In contrast, no significant sex‐based differences were found for SBP in the other two strains, nor for DBP or MAP in any of the strains examined (all p > 0.05, Figure 1G,H). Collectively, these data demonstrate that the fundamental metabolic and macrovascular physiology of mice is characterized by a sexual dimorphism.
Distinct Sexual Dimorphism in Pancreatic Microhemodynamics
3.2
To investigate the fundamental characteristics of pancreatic microcirculation, we quantified key microhemodynamic parameters, including Flux, Conc, and Velocity, in both male and female mice across three distinct genetic backgrounds, namely BALB/c, C57BL/6J, and Kunming. Our analysis revealed no significant sex‐dependent differences in either pancreatic microvascular Flux or Velocity within any of the strains examined (all p > 0.05, Figure 2A,B).
*Pancreatic microhemodynamics exhibit parameter‐specific sexual dimorphism and sex‐divergent coupling to systemic circulation. (A) Representative three‐dimensional scatter plots illustrating the distribution of pancreatic microhemodynamic parameters, blood perfusion (Flux), erythrocyte concentration (Conc), and erythrocyte velocity (Velocity), in male and female mice. (B) Quantification of mean pancreatic microvascular Flux, Conc, and Velocity across sexes. (C) Correlation network analysis between macrocirculatory parameters (HR, SBP, DBP, MAP) and pancreatic microhemodynamic parameters (Flux, Conc, Velocity) in male (left) and female (right) mice. Line width is proportional to the absolute value of Pearson's correlation coefficient (r). HR, heart rate; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure. n = 6 per group. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, **p < 0.001.
However, a sexual dimorphism was identified in erythrocyte concentration. Within the BALB/c strain, male mice exhibited a significantly higher Conc compared to their female counterparts (p < 0.01, Figure 2B). The differences in Conc between sexes were strain‐dependent, as it was not observed in either C57BL/6J or Kunming mice, suggesting that genetic background modulates the manifestation of this sex‐specific trait.
Sex‐Divergent Correlation Between Systemic and Pancreatic Microcirculation
3.3
We then sought to determine whether pancreatic microhemodynamics were coupled with systemic blood pressure. Correlation analysis revealed a sex‐exclusive relationship. In male mice, a significant positive correlation was established between DBP and pancreatic microvascular Flux (r = 0.478, p < 0.05), suggesting that in males, systemic vascular tone is directly associated with pancreatic tissue perfusion (Figure 2C). In contrast, no significant correlations were found between any of the measured pancreatic microcirculatory parameters and systemic blood pressure metrics in female mice across all strains (p > 0.05). HR in males was negatively correlated with the NO‐dependent magnitude of Flux (r = −0.563, p < 0.05) (Figure S1). In female mice, a distinct set of correlations was identified. The NO‐dependent magnitude of Flux was negatively correlated with DBP (r = −0.587, p < 0.05). In addition, the NO‐independent magnitude of Conc was positively correlated with both DBP (r = 0.651, p < 0.01) and MAP (r = 0.534, p < 0.05). Collectively, these findings indicate that the functional coupling between systemic circulation and pancreatic microhemodynamics is fundamentally different between the sexes (Figure 2C, Figure S1).
Sexual Dimorphism Revealed in Microhemodynamic Oscillation
3.4
To further elucidate the underlying regulatory mechanisms, we performed wavelet transform analysis to decompose the microcirculatory signals into their constituent oscillatory components, focusing on NO‐dependent and NO‐independent endothelial activities (Figure 3A,B). Analysis of endothelial oscillatory magnitudes revealed sexual dimorphism. In the BALB/c strain, males exhibited significantly greater NO‐dependent endothelial magnitudes than females for all three parameters of Flux, Conc, and Velocity (all p < 0.01). Within the NO‐independent band, BALB/c male‐dominant dimorphism was limited to the Conc parameter (p < 0.01), with no sex differences observed for Flux or Velocity. An opposing pattern of sexual dimorphism was observed in C57BL/6J mice. Here, females displayed significantly higher NO‐dependent magnitudes for Flux and Conc relative to males (both p < 0.01), whereas the Velocity parameter showed no difference. No sex‐based differences were detected in C57BL/6J strain for any parameter within the NO‐independent band (all p > 0.05). In Kunming mice, the sexual dimorphism was less extensive and isolated to the NO‐independent endothelial oscillation of Flux, where males revealed a higher magnitude compared to females (p < 0.001). The oscillatory magnitudes for Conc and Velocity did not differ significantly between sexes across either the NO‐dependent or NO‐independent endothelial magnitudes (all p > 0.05, Figure 3C).
*Wavelet analysis reveals sexual dimorphism in the oscillatory regulation of pancreatic microhemodynamics. (A) Representative time‐frequency spectral scalograms generated by continuous wavelet transform of pancreatic microhemodynamic signals from male and female mice. The x‐axis represents time, the y‐axis represents frequency, and the color intensity corresponds to the normalized power of the oscillatory magnitude. The regions corresponding to NO‐independent (0.005–0.0095 Hz) and NO‐dependent (0.0095–0.04 Hz) endothelial activity are highlighted in blue and purple. (B) Corresponding three‐dimensional visualizations of the spectral scalograms shown in (A), plotting time, frequency, and spectral amplitude to illustrate the oscillatory landscape. (C) Quantification of mean magnitude for NO‐dependent (0.0095–0.04 Hz) and NO‐independent (0.005–0.0095 Hz) endothelial oscillations for microvascular Flux, Conc, and Velocity. n = 6 per group. Data are presented as mean ± SD. **p < 0.01, **p < 0.001.
Absence of Sexual Dimorphism in the Endocrine Pancreas at the Structural and Molecular Level
3.5
To investigate the structural and molecular supports of the observed functional differences, we performed histological and immunohistochemical analyses on the endocrine and exocrine compartments of the pancreas. Within the islets of Langerhans (endocrine), quantitative analysis of HE stained sections revealed that the microvessel density was comparable between male and female mice across all three genetic strains (BALB/c, C57BL/6J, and Kunming) (all p > 0.05, Figure 4A,B). Furthermore, immunohistochemical assessment of the endothelial marker CD31 showed no significant sex‐dependent differences in expression within the islets of any strain (all p > 0.05, Figure 4C,D). Similarly, the expression levels of ERα and ERβ were also found to be statistically comparable between sexes, irrespective of the genetic strains (all p > 0.05, Figure 4E–H).
The endocrine pancreas is structurally and molecularly conserved between sexes. (A) Representative micrograph of HE‐stained pancreatic sections. The upper panel provides a low‐magnification overview of an islet of Langerhans. Scale bar = 200 μm; the lower panel presents a high‐magnification view of the same islet. Scale bar = 100 μm. (B) Quantification of microvessel density within islets of Langerhans reveals no significant differences between male and female mice across all three strains. (C, D) Immunohistochemical micrographs (C) and corresponding quantification (D) of the endothelial cell marker CD31 show comparable expression in islets between sexes. (E–H) Representative immunohistochemical micrographs of estrogen receptor α (ERα; E) and estrogen receptor β (ERβ; G) with corresponding quantification of expression (F, H) reveal no sexual dimorphism within the islets. Scale bar, 200 μm. ERα, estrogen receptors α; ERβ, estrogen receptors β; HE, hematoxylin and eosin. n = 6 per group. Data are presented as mean ± SD.
Despite the absence of sexual dimorphism in these baseline metrics, a subsequent correlation analysis revealed that these structural and molecular features were functionally coupled with microhemodynamic oscillations in a distinctly sex‐specific manner. In male mice, the NO‐independent magnitude of Flux showed a significant negative correlation with both microvessel number (r = −0.484, p < 0.05) and CD31 expression (r = −0.563, p < 0.05). Moreover, the NO‐dependent magnitudes of Flux (r = −0.557, p < 0.05) and Velocity (r = −0.475, p < 0.05) were negatively associated with ERβ expression. In female mice, a different correlational pattern was observed, where CD31 expression was negatively correlated with both the NO‐dependent (r = −0.533, p < 0.05) and NO‐independent (r = −0.610, p < 0.01) magnitudes of Flux, with no other significant associations found (Figure S2A,B). Collectively, these data indicate that while the endocrine pancreas does not exhibit sexual dimorphism in microvessel density or in the expression of these key receptors, the functional integration of these elements with pancreatic microhemodynamics is fundamentally sex‐dependent.
Sexual Dimorphism Is Confined to the Exocrine Pancreas
3.6
In contrast to the endocrine compartment, the exocrine pancreas emerged as a primary site of sexual dimorphism. Although the microvessel density in the exocrine tissue, assessed by HE staining, showed no significant differences between sexes in any strain (all p > 0.05, Figure 5A,B). However, the expression of molecular markers revealed sex‐specific patterns. Immunohistochemical analysis for CD31 uncovered a divergent sexual dimorphism in endothelial protein expression that was dependent on genetic background. Specifically, male mice exhibited significantly higher CD31 expression than females in both BALB/c and Kunming strains (both p < 0.05, Figure 5C,D). The pattern was inverted in C57BL/6J mice, where females displayed significantly higher CD31 expression than their male counterparts (p < 0.01, Figure 5C,D).
*The exocrine pancreas is a primary locus of molecular sexual dimorphism. (A) Representative micrograph of HE‐stained sections of the exocrine pancreas. The upper panel displays the general morphology of acinar and ductal structures. Scale bar = 200 μm; the lower panel provides a high‐magnification view that specifically highlights an interacinar microvessel. Scale bar = 100 μm. (B) Quantification of microvessel density in the exocrine compartment shows no significant difference between sexes. (C, D) Representative immunohistochemical micrographs (C) and corresponding quantification (D) of the endothelial marker CD31. (E–H) Representative immunohistochemical micrographs of estrogen receptor α (ERα; E) and estrogen receptor β (ERβ; G) with corresponding quantification (F, H). Scale bar = 200 μm. ERα, estrogen receptors α; ERβ, estrogen receptors β; HE, hematoxylin and eosin. n = 6 per group. Data are presented as mean ± SD. *p < 0.05, *p < 0.01.
The observed dimorphism extended to the expression of estrogen receptors. For ERα, a significant sex difference was exclusively observed in the C57BL/6J strain, with females showing higher expression levels than males (p < 0.01, Figure 5E,F); no such dimorphism was present in BALB/c or Kunming mice. The sexual dimorphism in ERβ expression paralleled the pattern observed for CD31. Male mice demonstrated elevated ERβ expression in the BALB/c and Kunming strains, whereas females showed higher expression in the C57BL/6J strain (all p < 0.05, Figure 5G,H).
To determine the functional relevance of these expression differences, we correlated them with exocrine microhemodynamic parameters. It was revealed that both shared and sex‐specific functional couplings. In male mice, CD31 expression was positively correlated with the NO‐dependent magnitude of Conc (r = 0.470, p < 0.05) and the NO‐independent magnitude of Velocity (r = 0.498, p < 0.05). A partially overlapping pattern was seen in females, where the NO‐dependent magnitude of Conc was also positively correlated with CD31 (r = 0.503, p < 0.05). However, a distinct negative correlation emerged in females, where the NO‐independent magnitude of Flux was inversely associated with the expression of both CD31 (r = −0.586, p < 0.05) and ERβ (r = −0.528, p < 0.05) (Figure S2C,D).
Taken together, these findings demonstrate that sexual dimorphism in the pancreatic tissue architecture is anatomically compartmentalized, being absent in the endocrine islets but within the exocrine component, that might modulate by genetic background and is evident in the parallel, sex‐specific regulation of the endothelial marker CD31 and the estrogen receptor ERβ.
Steroid Profiles Exhibit Strain‐Modulated Sexual Dimorphism
3.7
To investigate the hormonal basis for the observed sex‐specific phenotypes, we performed targeted metabolomics on the circulating steroidome. PCA revealed a sexual dimorphism, with distinct separation between male and female metabolic profiles within the BALB/c and Kunming strains, indicating that sex is a primary determinant of the overall steroid landscape (Figure 6A). Hierarchical clustering and volcano plot analysis further illustrated the metabolites driving this dimorphism (Figure 6B,C). A consistent pattern emerged where male mice, particularly in the BALB/c and Kunming strains, exhibited a steroid profile dominated by androgens (testosterone, dihydrotestosterone, androstenedione) and key upstream precursors in the steroidogenic pathway (cholesterol, lathosterol, pregnenolone). Conversely, female mice in these strains revealed a profile characterized by significantly higher levels of glucocorticoids (corticosterone, 11‐dehydrocorticosterone, cortisone) and estrogen metabolites (2‐methoxy‐estrone). Although the C57BL/6J strain also showed sex‐based separation, the specific pattern of dimorphism differed, highlighting a strong interaction between sex and genetic background in shaping the steroidome (Figure 6B,C, Figures S3–S8).
Circulating steroid hormone profiles are sexually dimorphic. (A) PCA scores plot of circulating steroid hormone profiles, demonstrating clear separation between male and female mice. Each dot represented an individual mouse, and the Hotelling's T‐squared ellipse corresponded to the 95% confidence interval for each group. (B) Heatmap with hierarchical clustering of differentially abundant steroid metabolites. Rows correspond to metabolites and columns to individual samples. Color intensity indicates z‐scored relative abundance, highlighting distinct male and female metabolic signatures. (C) Volcano plots comparing male and female steroid profiles across the three strains. The x‐axis represents log₂ (fold change) and the y‐axis represents −log10 (p‐value). 11‐DHC, 11‐dehydrocorticosterone; 2‐MeOE1, 2‐methoxy‐estrone; 5α‐PD, 5alpha‐pregnane‐3,20; 5β‐A, 5beta‐androsterone; 7‐HCO, 7‐hydroxy‐cholesten‐3‐one; 7‐KC, 7‐ketocholesterol; 7α,25‐diOHC, 7α,25‐dihydroxycholester; AD, androstenedione; CHOL, cholesterol; COR, cortisone; CORT, corticosterone; DOC, deoxycorticosterone; FC, fold change; LATH, lathosterol; PC1, the first principal component; PC2, the second principal component; PCA, principal component analysis; PREG, pregnenolone; T, testosterone. n = 6 per group.
Quantitative analysis of key metabolites also revealed both conserved and strain‐specific aspects of the dimorphism (Figure 7A). Across all three strains, female mice consistently displayed lower circulating levels of the cholesterol precursor lathosterol and the oxysterol 7α,25‐dihydroxycholesterol, alongside higher levels of the glucocorticoid 11‐dehydrocorticosterone compared to their male counterparts. Conversely, the patterns of sexual dimorphism for other key steroids, including cholesterol, pregnenolone, corticosterone, and deoxycorticosterone, were strain‐dependent.
*Divergent steroid hormone metabolites are functionally coupled to sex‐specific pancreatic microhemodynamics. (A) Quantification of key circulating steroid hormones and precursors revealing conserved and sexual dimorphism. (B) Sex‐stratified correlation network analysis linking steroid metabolites to pancreatic microhemodynamics. The colored nodes represented the correlations among key metabolites, whereas the lines indicated the associations between steroid hormones and microcirculatory profiles. Line width is proportional to the absolute value of Pearson's correlation coefficient (r), with the red and green lines representing significant negative and positive correlations, respectively. (C) KEGG pathway bar plot. KEGG “Count” revealed the number of differential metabolites enriched in each pathway. The red color indicates up‐regulated differential metabolites, whereas the blue color indicates down‐regulated metabolites. (D) Summary of KEGG pathway enrichment results. KEGG, Kyoto Encyclopedia of Genes and Genomes (http://www.kegg.jp). 11‐DHC, 11‐dehydrocorticosterone; 24‐HYCL, 24‐hydroxycholesterol; 7‐HCO, 7‐hydroxy‐cholesten‐3‐one; 7α, 25‐diOHC, 7α, 25‐dihydroxycholester; CHOL, cholesterol; CORT, corticosterone; DOC, deoxycorticosterone; KEGG, Kyoto Encyclopedia of Genes and Genomes (http://www.kegg.jp); LATH, lathosterol; PREG, pregnenolone. n = 6 per group. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ***p < 0.0001.
Sex‐Specific Coupling of the Steroidome With Pancreatic Microcirculatory Function
3.8
To determine if these dimorphic steroid profiles were functionally linked to the previously observed pancreatic microcirculatory phenotypes, we performed a correlation analysis. Our data revealed sex‐specific correlational patterns between the steroidome and pancreatic microhemodynamic oscillations. For lathosterol, a significant negative correlation was observed in male mice with the NO‐dependent endothelial signals of Flux (r = −0.547, p < 0.05), Conc (r = −0.473, p < 0.05), and Velocity (r = −0.491, p < 0.05). In females, the association was limited to a negative correlation with the NO‐independent signal of Flux (r = −0.583, p < 0.05). A sexual dimorphism was also evident for corticosterone. In males, corticosterone was positively associated with the oscillatory magnitudes of Conc (NO‐dependent: r = 0.650, p < 0.01; NO‐independent: r = 0.520, p < 0.05) and Velocity (NO‐dependent: r = 0.483, p < 0.05; NO‐independent: r = 0.508, p < 0.05). In contrast, the effect in females was restricted to a strong positive correlation with the NO‐independent magnitude of Flux (r = 0.719, p < 0.001). Cholesterol also displayed sex‐specific associations, showing a negative correlation with the NO‐dependent magnitude of Velocity in males (r = −0.502, p < 0.05), whereas in females, it was negatively correlated with the NO‐dependent (r = −0.478, p < 0.05) and NO‐independent (r = −0.610, p < 0.01) magnitudes of Flux. A shared regulatory pattern was observed for 11‐dehydrocorticosterone, which was positively correlated with the NO‐independent magnitude of Flux in both male (r = 0.625, p < 0.01) and female (r = 0.504, p < 0.05) mice. Deoxycorticosterone was negatively associated with the NO‐dependent magnitude of Flux in males (r = −0.489, p < 0.05), whereas no significant association was found in females. Collectively, these data establish that the sexual dimorphism extends beyond steroid concentrations to include their functional coupling with pancreatic microhemodynamics (Figure 7B).
Divergent Metabolic Pathway Enrichment Underlies Steroid‐Related Sexual Dimorphism
3.9
To elucidate the systemic biological processes underlying these dimorphic steroid profiles, we conducted KEGG pathway enrichment analysis. Although pathways related to lipid metabolism and the endocrine system were predictably enriched in both sexes (Figure 7C), the specific subordinate pathways were divergent. Compared with male counterparts, the female‐dominant differential metabolites were significantly enriched in pathways central to steroid hormone production including primary bile acid biosynthesis, steroid hormone biosynthesis, and aldosterone synthesis and secretion (Figure 7D). Collectively, these metabolomic data reveal a sexual dimorphism in the circulating steroidome, which is modulated by genetic background.
Discussion
4
The present study identifies the exocrine pancreas as a major site of sexual dimorphism within pancreatic microcirculation, a phenomenon driven by divergent systemic steroid hormone strategies and manifested within the microvasculature. Although sex differences in islet β‐cell function and insulin secretion are well‐documented (Brownrigg et al. 2023), the surrounding exocrine tissue has, until now, been largely overlooked as a microhemodynamic, hormonally‐responsive environment that could contribute to sex‐specific metabolic phenotypes. The anatomical and functional partitioning challenges the islet‐centric view of pancreatic microcirculatory sex differences and suggests that the microcirculation acts as a conduit, translating systemic hormonal signals into localized, sex‐specific physiological responses that integrate the entire organ.
Our data establish a distinction between the pancreatic compartments. The absence of sexual dimorphism in islet microvessel density and CD31, ERα/β expression indicates a conserved architecture, likely reflecting the stringent homeostatic requirements of glycemic control (Evans et al. 2002). In contrast, the exocrine tissue emerged as a hormonally responsive environment. The observed co‐regulation of the endothelial marker CD31 (Chow et al. 2010) and ERβ in the exocrine compartment is informative. The preferential recruitment of ERβ aligns with its specialized role in regulating resistance‐level microvessels, in contrast to the conduit‐vessel predominance of ERα. Mechanistically, ERβ is uniquely coupled to sustained eNOS activation (Hamada et al. 2006) and endothelium‐derived hyperpolarization (Arnal et al. 2010), signaling pathways that are essential for generating the low‐frequency vasomotor oscillations observed in the present study. Such evidence supports the interpretation that the exocrine endothelium preferentially engages ERβ‐dependent signaling to fine‐tune local perfusion pressure and buffer against systemic hemodynamic fluctuations. However, these findings indicate that sexual dimorphism appears to be modulated by genetic background rather than uniformly expressed across strains. Opposing patterns of microvascular regulation observed across strains suggest that genetic background serves as a modifier of sex steroid signaling. Although direct strain‐specific quantification of ERα/ERβ density was not performed, prior studies in inbred models have reported genetic background‐dependent differences in ER expression patterns, steroidogenic enzyme activity, and endothelial signaling efficiency, providing a plausible mechanistic framework for the inversion of dimorphic responses observed. The observation that perfusion maintenance is governed by the concentration parameter rather than flow velocity points to a mechanism involving functional capillary reserve recruitment, as opposed to simple upstream arteriolar vasodilation. Functional recruitment of capillaries parallels the structural preservation of CD31^+^ vessels, suggesting that metabolic resilience depends on maximizing the effective surface area available for exchange. The functionally adaptive phenotype is characterized by the capacity to maintain a high level of functional capillary density, thereby limiting microvascular rarefaction, and may therefore reflect a genetically determined capacity to sustain ERβ‐coupled microvascular signaling in a sex‐dependent manner, suggesting that the capacity of the exocrine microcirculation to maintain density and function buffers the metabolic risks associated with the male sex. Collectively, these strain‐dependent contrasts support the interpretation that successful microvascular remodeling may confer metabolic resilience, whereas failure to mount this adaptive response may reflect a biological substrate underlying increased susceptibility to metabolic disease.
The structural dimorphism has direct functional consequences for pancreatic perfusion and crosstalk (Rizk et al. 2023). The pancreas is unique in that its microcirculation forms an integrated network, with the islet‐acinar portal system ensuring that blood exiting the islets (Jansson and Carlsson 2019; Dybala and Hara 2021), rich in hormones, subsequently perfuses the surrounding acinar tissue. Given the islet‐acinar portal arrangement, the exocrine microvasculature acts as the downstream regulator of islet outflow. We propose that the robust NO‐dependent oscillations observed in females actively lower downstream resistance, facilitating the rapid “washout kinetics” necessary for pulsatile insulin delivery. Conversely, the lack of such active vasomotion in males may create a perfusion‐function mismatch, where hormonal export is dampened by a passive, high‐resistance environment despite conserved islet vascular density. Endothelial dysfunction is a known antecedent to metabolic disease (Kopaliani et al. 2024), and our work suggests the exocrine microvasculature may contribute to early susceptibility. Our findings add another layer to this model, indicating that the regulation of the microvascular network itself is sexually dimorphic. The distinct oscillatory dynamics and differential reliance on systemic blood pressure between sexes imply that the nature of this exocrine‐endocrine circulatory coupling is sex‐dependent (Song et al. 2020; Connelly et al. 2022; Alfie et al. 1995). Spectral decomposition of microvascular flow revealed a different sex‐specific signature within the NO‐dependent frequency interval (0.0095–0.04 Hz). The oscillatory component reflects an active, protective mode of endothelial vasomotion. Enhanced oscillatory power in females is proposed to arise from estrogen‐dependent upregulation of the eNOS pathway, consistent with the established role of estrogen in increasing NO bioavailability and modulating low‐frequency oscillations of vascular tone (SenthilKumar et al. 2026; Dhungana et al. 2025). These oscillatory divergences were evident even in strains like Kunming where mean Flux values showed no sexual dimorphism, indicating that regulatory mechanisms diverge before gross perfusion deficits manifest, highlighting microvascular vasomotion as a far more sensitive biomarker for early endothelial dysfunction than standard static flowmetry, capable of unmasking sex‐specific alterations that total flow volume misses. The presence of stable NO‐dependent oscillations in females indicates a dynamic and responsive microvasculature capable of sustaining capillary perfusion and limiting vascular stasis. By contrast, lower NO‐dependent oscillatory power in males is consistent with a reduced endothelial vasomotor reserve, which may indicate increased susceptibility to hemodynamic stress.
From a translational perspective, our findings suggest an explanation for the documented sexual dimorphism in the incidence and severity of other major pancreatic pathologies such as acute pancreatitis (Drake et al. 2021). Clinical data consistently show that men have a higher incidence of pancreatitis and suffer from more severe outcomes (Weissman et al. 2023). Our results support a model in which the androgen‐dominant steroid profile and distinct microvascular regulation in males render the exocrine pancreas more susceptible to insults (Kapoor et al. 2005; Xu et al. 2019), thereby lowering the threshold for acinar cell injury and inflammation. Considering microcirculatory failure is an early event in pathogenesis, the lack of protective NO‐dependent oscillatory mechanisms may underlie the increased severity observed in males (Widaeus et al. 2025; Bahadoran et al. 2023). Furthermore, since exocrine pancreatic inflammation is a potent systemic stressor known to exacerbate endothelial dysfunction and cardiovascular risk (Desai et al. 2023), our work reveals the pancreatic microvasculature as an active contributor to a sex‐specific pathological feedback loop that connects local pancreatic injury to systemic cardiometabolic decline. Our view highlights the potential for developing sex‐stratified therapeutic strategies that aim to restore the unique microvascular homeostasis of the exocrine pancreas to mitigate both local damage and systemic consequences.
Targeted metabolomics identified the circulating steroidome as a systemic driver of the dimorphism (Girel et al. 2024; Zhang et al. 2023; Storbeck et al. 2019). Although absolute hormone concentrations differed markedly between sexes, analysis of product‐to‐precursor ratios (Figure S9) suggested that conversion efficiency varied with genetic background, indicating strain‐dependent regulatory set‐points rather than a uniform synthesis or clearance pattern. The composition of these circulating pools was associated with the distinct vascular oscillatory profiles identified in our study. The estrogen‐dominant profile in females paralleled enhanced NO‐dependent vasomotion, whereas the androgen‐dominant signature in males aligned with greater dependence on systemic pressure and structural constraints (Robert 2023; Willemars et al. 2022). This physiological divergence may help contextualize clinical observations such as earlier onset of type 2 diabetes in men (Navarro et al. 2015; Liu and Sun 2018; Allan 2014). A hormonally influenced exocrine microenvironment, characterized by altered perfusion dynamics, could secondarily affect downstream islet function (Raeisi‐Dehkordi et al. 2024; Milionis et al. 2024; Morford and Mauvais‐Jarvis 2016), providing a tissue‐level interface between the male sex and increased T2D susceptibility.
The present study is a characterization of physiological phenomena and establishes functional correlations between the steroidome and pancreatic microhemodynamics. Although we postulate that sex steroids drive the observed dimorphism, direct mechanistic evidence is lacking. The approach to definitively attribute these phenotypes to specific hormones would involve surgical gonadectomy followed by controlled hormone replacement (e.g., add‐back of testosterone, estradiol, or corticosterone (Seale et al. 2004; Haupt et al. 2020)). Our molecular analysis, whereas revealing, was not exhaustive. The focus on estrogen receptors was hypothesis‐driven, but the steroidomics data revealed significant dimorphism in androgens and glucocorticoids. Consequently, future work is necessary to establish causality and dissect the molecular pathways to confirm the proposed regulatory axes. Furthermore, surgical gonadectomy with controlled hormone replacement would be required to definitively attribute the observed phenotypes to specific steroids. Longitudinal studies in preclinical models of metabolic disease are essential to determine how these basal sex differences evolve during pathogenesis and influence disease outcomes.
Conclusion
5
Our study repositions the exocrine pancreas as a key determinant of sexual dimorphism in pancreatic physiology. We propose a model where divergent systemic steroid strategies are translated via a dimorphically regulated microvasculature into a sex‐specific exocrine environment. The microcirculation acts as the functional conduit that integrates the exocrine and endocrine compartments, suggesting that the health of the exocrine tissue is a sex‐dependent factor in maintaining overall pancreatic homeostasis.
Author Contributions
M.L. designed the experiments. Y.L., Y.W., Q.O., W.L. and X.X. performed all experiments. Y.L., B.W. and X.Z. analyzed the data. M.L. and Y.L. drafted and revised the manuscript. R.X. helped to revise the manuscript. M.L. conceived and supervised the project. All authors discussed the results and commented on the manuscript.
Funding
This work was supported by the Beijing Municipal Natural Science Foundation grant (7252093).
Ethics Statement
All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of the Institute of Microcirculation, CAMS (approval no. CAMS‐IM‐IACUC‐2023‐AE0937).
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Figure S1: cph470130‐sup‐0001‐FiguresS1‐S9.pdf. Figure S2: cph470130‐sup‐0001‐FiguresS1‐S9.pdf. Figure S3: cph470130‐sup‐0001‐FiguresS1‐S9.pdf. Figure S4: cph470130‐sup‐0001‐FiguresS1‐S9.pdf. Figure S5: cph470130‐sup‐0001‐FiguresS1‐S9.pdf. Figure S6: cph470130‐sup‐0001‐FiguresS1‐S9.pdf. Figure S7: cph470130‐sup‐0001‐FiguresS1‐S9.pdf. Figure S8: cph470130‐sup‐0001‐FiguresS1‐S9.pdf. Figure S9: cph470130‐sup‐0001‐FiguresS1‐S9.pdf.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Addison, P. , J. Walker , and R. Guido . 2009. “Time‐Frequency Analysis of Biosignals.” IEEE Engineering in Medicine and Biology Magazine 28, no. 5: 14–29. 10.1109/MEMB.2009.934244.19775952 · doi ↗ · pubmed ↗
- 2Alfie, J. , G. D. Waisman , C. R. Galarza , et al. 1995. “Relationship Between Systemic Hemodynamics and Ambulatory Blood Pressure Level Are Sex Dependent.” Hypertension 26, no. 6 Pt 2: 1195–1199. 10.1161/01.hyp.26.6.1195.7498995 · doi ↗ · pubmed ↗
- 3Allan, C. A. 2014. “Sex Steroids and Glucose Metabolism.” Asian Journal of Andrology 16, no. 2: 232–238. 10.4103/1008-682X.122589.24457840 PMC 3955332 · doi ↗ · pubmed ↗
- 4Alvarez Fallas, M. E. , S. Pedraza‐Arevalo , A.‐M. Cujba , et al. 2021. “Stem/Progenitor Cells in Normal Physiology and Disease of the Pancreas.” Molecular and Cellular Endocrinology 538: 111459. 10.1016/j.mce.2021.111459.34543699 PMC 8573583 · doi ↗ · pubmed ↗
- 5Arnal, J. F. , C. Fontaine , A. Billon‐Gales , et al. 2010. “Estrogen Receptors and Endothelium.” Arteriosclerosis, Thrombosis, and Vascular Biology 30, no. 8: 1506–1512. 10.1161/ATVBAHA.109.191221.20631350 · doi ↗ · pubmed ↗
- 6Bahadoran, Z. , P. Mirmiran , K. Kashfi , and A. Ghasemi . 2023. “Vascular Nitric Oxide Resistance in Type 2 Diabetes.” Cell Death & Disease 14, no. 7: 410. 10.1038/s 41419-023-05935-5.37433795 PMC 10336063 · doi ↗ · pubmed ↗
- 7Brownrigg, G. P. , Y. H. Xia , C. M. J. Chu , et al. 2023. “Sex Differences in Islet Stress Responses Support Female Beta Cell Resilience.” Molecular Metabolism 69: 101678. 10.1016/j.molmet.2023.101678.36690328 PMC 9971554 · doi ↗ · pubmed ↗
- 8Chow, R. W. , D. J. Handelsman , and M. K. Ng . 2010. “Minireview: Rapid Actions of Sex Steroids in the Endothelium.” Endocrinology 151, no. 6: 2411–2422. 10.1210/en.2009-1456.20392826 · doi ↗ · pubmed ↗