Effects of a global Rab27a null mutation on murine PVAT and cardiovascular function

Abstract

Background:

RAB27A, a modulator of secretion, is expressed within vessels and perivascular adipose tissue (PVAT). We hypothesized that loss of RAB27A would alter cardiovascular function.

Methods: Body weight of Rab27a^ash mice was measured from 2 to 18 months of age, along with glucose resorption at 6 and 12 months of age and glucose sensitivity at 18 months of age. Body weight and cellular and molecular features of PVAT and aortic tissue were examined in a novel C57BL/6J Rab27a null strain. Analyses included morphometric quantification and proteomic analyses. Wire myography measured vasoreactivity, and echocardiography measured cardiac function. Comparisons across ages and genotypes were evaluated via two-way ANOVA with multiple comparison testing. Significance for myography was determined via four-parameter non-linear regression testing.

Results:

Genome-wide association data linked rare human RAB27A variants with body mass index and glucose handling. Changes in glucose tolerance were observed in Rab27a^ash male mice at 18 months of age. In WT and Rab27a null male mice, body weight, adipocyte lipid area, and aortic area increased with age. In female mice, only body weight increased with age, independent of RAB27A presence. Protein signatures from male Rab27a null mice suggested greater associations with cardiovascular and metabolic phenotypes compared to female tissues. Wire myography results showed Rab27a null males exhibited increased vasoconstriction and reduced vasodilation at 8 weeks of age. Rab27a null females exhibited increased vasoconstriction and vasodilation at 20 weeks of age. Consistent with these vascular changes, male Rab27a null mice experienced age-related cardiomyopathy, with severe differences observed by 21 weeks of age.

Conclusion:

Global RAB27A loss impacted PVAT and thoracic aorta proteomic signatures, altered vasocontractile responses, and decreased the left ventricle ejection fraction in mice.

Graphical Abstract

Introduction

Cardiovascular disease remains a leading cause of non-communicable disease mortality worldwide¹. Multiple factors contribute to vascular disease pathology and progression including aging, obesity, and diabetes^1,2. These factors often cause a dysregulation of paracrine and endocrine signaling, which impacts the cardiovascular system^3–5. The vascular microenvironment is comprised of perivascular adipose tissue (PVAT), vascular smooth muscle cells (VSMC), and endothelial cells⁵. Changes in communication between PVAT and the underlying vessel wall can result in vascular stiffening, plaque formation, and the modulation of inflammatory responses^6,7. Defining the mechanisms that regulate communication within this niche, and thereby impact thoracic aorta physiology, is necessary for understanding the pathophysiology of vascular disease.

One potential target is RAB27A, which mediates endosome trafficking to the plasma membrane and the secretion of signaling factors^8,9,10. While the importance of RAB27A in vascular health is poorly understood, its expression within the thoracic aorta microenvironment suggests it may have a role in vascular physiology. Loss of functional RAB27A in humans causes type II Griscelli syndrome, which is classically associated with hypopigmentation and dysregulated immune responses^11,12. Changes in RAB27A expression also impacts the progression of multiple human diseases^13,14. Previously, we showed that RAB27A is present in human PVAT preadipocytes, and that its suppression decreased adipocyte differentiation and lipid storage¹⁵. Although important for endothelial cell secretion, RAB27A has not been well studied in the context of vascular physiology¹⁶. Due to the importance of intercellular and cross-tissue communication within the thoracic aorta microenvironment, and the importance of RAB27A for intercellular communication, we hypothesized that RAB27A expression is necessary for maintaining cardiovascular homeostasis and function^17,18.

Utilizing two mouse models, we provide evidence that the mutation of Rab27a, resulting in a global loss of RAB27A, altered metabolic and thoracic aorta reactivity phenotypes. With the Rab27a^ash strain, age-associated changes in glucose sensitivity were observed in male mice. No such changes were observed in Rab27a^ash females, suggesting sex-dependent responses to the Rab27a mutation. Changes in the proteomic profiles of PVAT and the thoracic aorta in our novel Rab27a global null strain strongly suggest cardiovascular and metabolic phenotypes, and here we provide evidence that global loss of RAB27A affects thoracic aorta vasoreactivity and cardiac function in mice. These differences in reactivity occurred without changes in body weight, adipose tissue lipid accumulation, or thoracic aorta morphology. Future studies will determine the importance of targeted RAB27A loss within this niche, as well as identify the secreted signaling molecules that mediate these effects.

Methods

Please see the Supplemental File for additional methodological details. Reference the Major Resources Tables in the Supplemental File for details related to reagents and mouse strains.

Human Tissue Samples.

Deidentified human tissues were collected from patients undergoing cardiovascular-related surgical procedures under an Institutional Review Board-approved protocol.

Rab27a^ash Strain.

All animal experiments were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Experiments for this strain were performed at The Jackson Laboratory after approval by their Institutional Animal Care and Use Committee. Mice were maintained at a 12-hour (6AM-6PM) light-dark cycle, fed a rodent diet ad libitum, and provided acidified water in a pathogen-free room. Mice were genotyped according to a standardized protocol established by The Jackson Laboratory.

Rab27a Global and Conditional Null Strain:

Development, characterization, and use of Rab27a alleles and additional mouse strains were approved by the Institutional Animal Care and Use Committee of MaineHealth Institute for Research. Mice were group-housed in an Association for Assessment and Accreditation of Laboratory Animal Care-approved mouse barrier facility and provided food and water ad libitum with a 12-hour light-dark cycle. Genomic DNA was isolated from mouse tissue samples. Isolated DNA was used for genotyping via polymerase chain reaction (PCR) targeting exon 4 of Rab27a. The resulting amplicon from a Rab27a null mouse was 164 base pairs (bp), while the resulting amplicon of a wildtype (WT) mouse was 853bp. Gel results were imaged utilizing a Fujifilm imager (LAS-4000).

Glucose Sensitivity and Resorption.

Glucose was measured in mouse urine using a Beckman AU680 chemistry analyzer. A glucose tolerance test was performed on mice that were fasted for 6 hours by injecting 2mg of glucose per gram of body weight (Sigma, Cat. #G8769) intraperitoneally. Two drops of blood from the tail vein were collected at the time of injection and at 15, 30, 60, 120, and 180 minutes post administration. Glucose concentrations were measured in duplicate at each time using NovaStatStrip Xpress (Nova Biomedical).

Tissue Collection.

Mice were anesthetized with isoflurane and euthanized via cervical dislocation. Collected tissues were frozen at −80°C or fixed with 10% formalin overnight at 4°C. Fixed samples were then processed for paraffin embedding, sectioning, and routine histology or immunostaining.

Immunoblot.

Total protein was isolated in the presence of radioimmunoprecipitation assay buffer supplemented with protease/phosphatase inhibitor (CST, Cat. #5872). Frozen tissues received an additional sonication step. Samples were loaded onto 12% SDS-PAGE gels (Bio-Rad, Cat. #1610175) with equal amounts of protein per lane. Proteins were transferred to PVDF membranes (Bio-Rad, Cat. #1704275) utilizing a Bio-Rad Trans-Blot Turbo system. Post blocking, membranes were incubated with a designated antibody in 5% milk (PBS-T) at 4°C while rocking (Major Resource Table 4). Membranes were washed and incubated with the appropriate IgG-HRP secondary in 5% milk (PBS-T) at room temperature for 1 hour. Blots were imaged utilizing Immobilon Forte Western HRP Substrate (Millipore, Cat. #WBLUF0500) and the Azure C600.

Tissue Morphometry.

Changes in lipid area proportions within PVAT and brown adipose tissue (BAT) were determined as previously described¹⁹. For subcutaneous inguinal white adipose tissue (iWAT), tissue area was selected through tracing and then utilized to create a mask. The mask was then applied as described in the original protocol¹⁹. Thoracic aorta sections were imaged in segments and stitched together utilizing Photoshop (Adobe Inc., 2022). Total area was determined by outlining the external vessel wall while the lumenal area was determined by outlining the internal vessel wall (FIJI). Medial area was determined by subtracting the lumenal area from the total area.

Mass Spectrometry:

A minimum of 20mg of iWAT and BAT was utilized from three mice for each experimental group. Due to limited tissue size, PVAT and thoracic aorta samples were pooled by tissue type for each sex and genotype. For each pooled sample, a minimum of 8mg of male tissue and 4.5mg of female tissue was submitted. Male and female tissues were processed independently as described^20,21. The mass spectrometer was operated using data dependent analysis (DDA) to create a spectral ion library. A Sequential Window Acquisition of all Theoretical spectra (SWATH) was implemented for relative quantification²². All data acquired in DDA mode used a high-resolution MS scan prior to MS/MS analysis using collision-induced dissociation. For SWATH acquisition, a ToF MS scan with an accumulation time of 96 milliseconds was followed by 100 variable-width scan windows from 350 to 1500m/z. Accumulation time was 90 milliseconds, and cycle time was 9.1 seconds. Identical chromatography parameters were used for SWATH and DDA analysis. To create a peptide ion library, DDA data were searched against the mouse UniProt database (37,201 proteins) using ProteinPilot software (Version 5.0.2, Sciex) running the Paragon algorithm. Peptides used for protein identification met a minimum confidence threshold of 99%. Spectral alignment and targeted extraction of SWATH data were performed with the SWATH Processing Micro App in PeakView (Version 2.2.0, Sciex) using the reference DDA ion library. Peptide filters were set according to the following parameters: number of peptides per protein = 6; number of transitions per peptide = 5; peptide confidence threshold = 95%; FDR <5%; exclude modified peptides = yes. Settings for extracted ion chromatograms were set according to the following parameters: extracted ion chromatogram extraction window (min) = 5 and extracted ion chromatogram width (ppm) = 75. Data were imported into MarkerView (Version 1.2.1, Sciex), at which point intensity ratios between different analyses were normalized according to most-likely ratios²³. Experimental groups were compared by principal component analysis and significance was evaluated using the Fisher’s modification of Student’s t test²⁴. The male proteomic data included 3,960 detected proteins with 1,833 quantified. The female proteomic data included 3,260 detected proteins with 2,416 quantified.

STRING Analysis.

Comparisons of experimental group SWATH results were completed utilizing STRING bioinformatics analysis^25,26. Results for the KEGG and Mammalian Phenotype Ontology (Monarch) databases were filtered for a minimum strength of 1, with P≤0.05. The remaining enrichment terms, and directionality of targets within those terms, were visualized utilizing R and the ggplot2 and ComplexHeatmap packages.

Immunofluorescence.

Tissue sections were incubated at 4°C overnight with a primary antibody or appropriate IgG control at matching concentrations in PBS + 2% BSA. Post wash, samples were incubated at room temperature for 1 hour with Alexa-Fluor conjugated secondary antibodies in staining solution. Sections were washed and treated with the TrueVIEW Autofluorescence Quenching Kit (Vector Laboratories, Cat. #SP-8400). Slides were washed and mounted with DAPI mounting media (Vector Laboratories, Cat. #H-2000). Sections were imaged with a Leica TCS SP8 laser scanning confocal microscope, a 63x/1.40 NA oil objective, and the Leica SX software. A single z-plane image was captured for one randomly selected tissue sample (n=1) per experimental group. Expression is reported as relative fluorescence intensity/nuclei to the age- and sex- matched WT control.

Wire Myography.

Murine thoracic aortae were excised at 2mm in length, below the ascending aortic arch, and mounted into a wire myograph organ bath (Danish Myography Technologies A420 and 620M, Denmark). Aortae were treated with increasing concentrations of phenylephrine from 2nM-10μM (Sigma-Aldrich, Cat. #P6126). The maximum contractile force was recorded for each dose. Vascular relaxation was measured by pre-contracting vessels to 50–80% of peak contraction with phenylephrine, and then treating vessels with increasing concentrations of acetylcholine from 2nM-10μM (Sigma-Aldrich, Cat. #A2661). The minimum contractile force was recorded for each dose. Vessel integrity was evaluated by responsiveness to 100mM potassium chloride (KCl). The peak contraction was recorded at 8 minutes post addition. Data were recorded using LabChart software (ADInstruments) and the DTM myograph plugin.

Echocardiography.

Rab27a null and littermate WT control mice underwent cardiac function and morphology assessment by echocardiography at 8, 16, and 20 weeks of age. Mice were under isoflurane anesthesia (2%) and placed supine on a heated animal stage with integrated ECG leads. Hair was removed using Nair hair removal. Short axis M-mode images were acquired on a Vevo2100 (VisualSonics Systems, Toronto, ON, Canada) with a 40 MHz solid-state transducer using the papillary muscles as orientation. Data were analyzed with the heart function package of VevoLab (version 5.6.0).

RT-qPCR.

Male Rab27a null and WT mice were euthanized and perfused with PBS at 20 weeks of age. Tissues were incubated with Trizol (Invitrogen, Cat. #15596018) and RNA isolated using the Qiagen RNeasy Mini Kit (Qiagen, Cat. #74104). Primers were obtained through IDT (Major Resources Table 3) and reconstituted in IDTE buffer (pH=7.5). A Bio-Rad CFX Connect 96 Real-Time was used to complete qPCR reactions. Transcript expression of the selected genes was calculated relative to β-actin expression and normalized to the WT control for each tissue type.

Quantification of Total Protein.

Cells were incubated in 5mL of phenol-free DMEM with low glucose and HyClone L-glutamine (Cytiva, Cat. #SH30034.01), supplemented with 1x antibiotic/antimycotic solution for 24 hours at 37°C and 5% CO₂. Conditioned media was stored at −20°C. Live cells were counted using a hemocytometer after staining with trypan blue. Protein concentrations were quantified using the Pierce BCA assay (Thermo Fisher, Cat. #23225).

Extracellular Vesicle Isolation and Analysis.

Plasma samples were collected from male Rab27a null and WT mice at 20 weeks of age via submandibular bleed into K2EDTA coated collection vials (BD Microtainer^®, Cat. #VT365974). Frozen samples were thawed, and extracellular vesicles were isolated using ultra-filtration and size exclusion chromatography. Analysis of particle size and concentration was completed using a ZetaView dual laser system (Particle Metrix). Samples were analyzed under the following settings: Sensitivity = 80, Shutter Speed = 100, Temperature = 25°C, Video Recording Speed = Medium.

Statistical Analysis.

Significant differences in Rab27a^ash body weight, glucose resorption, and GTT data were determined via two-way ANOVA with a Tukey’s multiple comparison test. Significant differences in body weight, tissue lipid area, and vascular morphology data between Rab27a global null groups was evaluated via two-way ANOVA with a Tukey’s multiple comparison test. For the myography data, thoracic aortae identified as outliers based on the peak KCl contractile response using a ROUT test were removed from analyses. Significant differences between myography agonist responses were determined with a variable slope (four parameters) test, which detected and eliminated outliers without weighting. Mice with 50% or more of data points calculated as outliers were removed. Curves were compared through an extra-sum-of-squares F test with a 95% confidence interval. The EC₅₀ of each mouse and group averages were calculated and significant differences were determined via two-way ANOVA with a Tukey’s multiple comparison test. Significant differences in total protein/cell for in vitro experiments were evaluated via Welch’s t test. Significant differences between gene expression data were evaluated via Welch’s t test by target gene and tissue type. For all comparisons, significance was defined as a P<0.05. Analyses were completed utilizing Prism (GraphPad, Version 9.5).

Results

Associated with human metabolic phenotypes, RAB27A is expressed in the human and mouse thoracic aorta.

Mutations in the RAB27A gene can result in the rare human type II Griscelli syndrome, altered insulin secretion, and impaired blood coagulation^16,27–29. To further understand potential associations between RAB27A and human disease, we queried for associations between RAB27A variants and multiple human metabolic disease phenotypes using the Common Metabolic Disease (CMD) Knowledge Portal³⁰. From this extensive database of publicly available genome-wide association studies, human body mass index (BMI) was predicted to have the greatest correlation with common RAB27A variants (Fig. 1A)³¹. Meanwhile, the data indicated that rare RAB27A variants significantly associated with increased BMI and glucose sensitivity (Fig. 1B)³¹.

Figure 1. RAB27A is associated with metabolic phenotypes and is localized in human and mouse vasculature.

(A) Display of metabolic disease phenotypes with the greatest association to common RAB27A variants. Data are available through the CMD Knowledge Portal³¹. BMI = body mass index (P = 0.0003), AFxBMI = atrial fibrillation-SNP x BMI interaction (P = 0.003), and CRP = plasma C-reactive protein (P = 0.003). (B) Display of the two phenotypes significantly associated with rare RAB27A variants: BMI (P = 0.027) and 2-hour glucose levels (P = 0.04). Y-intercepts denote significance thresholds set by the CMD Knowledge Portal³¹. (C) Body weights of Rab27a^ash +/+, ash/+, ash/ash males from 2–18 months of age. (D) Urine glucose concentration from Rab27a^ash +/+, ash/+, ash/ash males at 6 or 12 months of age. (E) Glucose tolerance test of Rab27a^ash +/+, ash/+, ash/ash males at 18 months of age. (F) Body weight of Rab27a^ash +/+, ash/+, ash/ash females from 2–18 months of age. (G) Urine glucose concentration from Rab27a^ash +/+, ash/+, ash/ash females at 6 and 12 months of age. (H) Glucose tolerance trends of Rab27a^ash +/+, ash/+, and ash/ash females at 18 months of age. Bars display the mean ± SEM. Significance was determined via a two-way ANOVA with a multiple comparison test. WT vs. ash/ash: ****, P<0.0001 **, P<0.01, and *, P<0.05. ash/+ vs. ash/ash: ###, P<0.001; #, P<0.05. (I) RAB27A protein in human PVAT isolated from two donors (lanes 1–2), primary human VSMC (lanes 3–4) and primary human endothelial cells (lanes 5–6). Perilipin 1 (PLIN1) – adipose tissue marker, smooth muscle actin alpha (ACTA) – VSMC marker, and platelet endothelial cell adhesion molecule (PECAM1) - endothelial cell marker^55–57. A non-specific band was detected in the first human PVAT sample, and the upper band in the endothelial cell samples are PECAM1. A representative Ponceau S-stained blot is provided for visualizing total protein transfer. (J) Immunofluorescence images of RAB27A in WT mouse PVAT and thoracic aorta. Scale bar = 20μm. EC = endothelial cell, L = lumen.

Utilizing the Rab27a^ash strain, male and female mice were examined for differences in body weight and glucose handling compared to controls (Major Resource Table 1)³². Male Rab27a^ash homozygous (ash/ash) mice gained weight comparably to wildtype (+/+, WT) and heterozygous (ash/+) controls up to 18 months of age (Fig. 1C). However, altered glucose resorption was observed at 12 months when comparing ash/ash males to WT controls (Fig. 1D). Additionally, while the WT males showed increased glucose resorption between 6 and 12 months of age, no significant changes between ages were observed in the ash/+ and ash/ash males. At 18 months of age, ash/ash males exhibited altered glucose sensitivity when compared to age-matched WT mice (Fig. 1E). These data complement the findings of a prior study which showed increased glucose sensitivity at 15 weeks of age²⁸. Expanding on this foundational work, female ash/ash mice were also examined and exhibited no changes in body weight compared to controls (Fig. 1F). Female mice showed no evidence of altered glucose sensitivity and handling associated with genotype (Fig. 1G–H). These data suggest that this Rab27a mutation exhibits age- and sex-specific metabolic effects.

These data and previous studies suggest a role for Rab27a in metabolic health, adipocyte physiology, and endothelial cell function^15,16. With metabolic health being a risk factor for cardiovascular disease, we confirmed the presence of RAB27A in human and murine PVAT, and vascular smooth muscle cells (VSMC), and vascular endothelial cells within the thoracic aorta (Fig. 1I, J)³³. Since RAB27A is expressed throughout the thoracic aorta microenvironment and considering the potential impact on human metabolic disease phenotypes, we sought to define its importance in cardiovascular physiology.

Establishment of novel Rab27a global and conditional null strains.

The Rab27a^ash mouse from a colony of C3H/HeSN mice at The Jackson Laboratory was the first recorded mouse strain with a Rab27a null mutation^32,34. While other strains exist, we present a novel strain on a C57BL/6J background^35,36. Utilizing CRISPR-Cas9 technology, we created novel Rab27a global and conditional null strains by targeting exon 4, which is the first conserved exon among known Rab27a transcripts (Suppl. Fig. 1A). In vitro validation of the designed sgRNA was performed to identify selectivity in genomic targeting of these regions within the Rab27a sequence prior to microinjection (Suppl. Fig. 1B).

Single-cell C57BL/6J embryos were injected with a CRISPR-Cas9 complex targeting exon 4 of Rab27a for the insertion of loxP sequences. Zygotes were transferred to the oviducts of pseudo-pregnant Swiss Webster surrogate mothers and carried to term. A mosaic pup was used as a founder and backcrossed to C57BL/6J mice to establish the Rab27a global null strain (Fig. 2A–C, Major Resource Table 2). These backcrosses yielded a pup containing a homozygous deletion of exon 4 via non-homologous end-joining, as confirmed through genotyping and sequencing (Suppl. Fig. 2A). The null allele resulted in the characteristic ashen coat color (C57BL/6J-Rab27a^em10Llw, Rab27a null, Fig. 2A)³². Immunofluorescence staining of RAB27A was performed in sections from thoracic aortae and PVAT, and demonstrated complete loss of RAB27A in endothelial cells, VSMCs, and PVAT (Fig. 2D). Immunoblot analysis confirmed the loss of RAB27A in multiple tissues of the Rab27a null mice (Suppl. Fig. 2B).

Figure 2. Validation of Rab27a mutant alleles.

(A) Rab27a global null mouse with ashen coat color next to a WT littermate. (B) Schematic showing PCR products as green bars for both WT (top) and Rab27a global null (bottom) alleles. Created with Biorender.com. (C) Representative genotyping of Rab27a global null, heterozygous (Het), and WT mice compared to the C57BL/6J control (B6). (D) Immunofluorescence images of RAB27A localization in PVAT, VSMC, and endothelial cells of the thoracic aorta from WT or Rab27a null male mice. Scale bar = 20μm. (E) Restriction digest of PCR products from the Rab27a^fl/fl strain showing presence of loxP sites. Samples were incubated in the digestion buffer with or without Pcil. Female = female parent, Male = male parent, 1 = offspring of female and male parents. (F) Image of the Rab27a^fl/fl;Sox-2 Cre mouse. (G) Genotyping of the litter shown in (F), where pup 93 is the genotyping result for the Rab27a^fl/fl:Sox-2 Cre pup. Top panel demonstrates complete deletion of exon 4 in pup 93. Bottom panel shows the Cre transgene in three offspring, including pup 93.

A second founder, containing intact loxP sites, was also backcrossed to the C57BL/6J strain to establish the conditional Rab27a null strain (C57BL/6J-Rab27a^em24Llw, Rab27a^fl/fl) (Fig. 2E, Major Resource Table 2). These Rab27a^fl/fl mice were crossed with Sox2-Cre mice (B6.Cg-Edil3^{Tg(Sox2-cre)1Amc}/J, The Jackson Laboratory) on a C57BL/6J background for global deletion of the conditional allele (Major Resource Table 1). Resulting Rab27a^fl/fl;Sox2-Cre pups displayed the expected ashen coat phenotype with the homozygous null genotype (Fig. 2F–G). To characterize the effects of complete deletion of Rab27a within the vascular microenvironment, the Rab27a global null strain was used in this study.

Global loss of RAB27A had no effect on body weight, lipid area, or vessel area.

To characterize our novel Rab27a null strain, the body weights of male and female mice were recorded at 8 weeks or 20 weeks of age. As observed in the Rab27a^ash strain, male and female Rab27a null mice had similar body weights compared to WT controls at matched ages (Fig. 3A, H). The Rab27a null mice also gained weight with age like their genotype controls. Due to the suggested role of RAB27A in lipid accumulation and metabolism, PVAT, brown adipose tissue (BAT), and inguinal white adipose tissue (iWAT) were examined for changes in lipid area¹⁹. Altered adipose morphology often indicates changes in adipose signaling and metabolic activity^7,15,37. Furthermore, due to the relatively small size of the PVAT and BAT depots, it was possible that changes in their physiology would not be detected by total body mass³⁸. However, no significant differences in lipid area proportions were found when comparing these age-matched tissues from Rab27a null and WT mice (Fig. 3B–D and 3I–K). Interestingly, while the male WT, PVAT showed increase lipid area with aging, the Rab27a null PVAT did not. The proportion of lipid area in adipose tissue increased with age for PVAT (P = 0.003), BAT (P = 0.008), and iWAT (P = 0.009) in male mice. No changes were detected in lipid area after 8 weeks of age in female mice. These findings are reflected in the representative histological images (Suppl. Fig. 3).

Figure 3. Global loss of RAB27A does not affect adipose lipid area or thoracic aorta vessel area.

(A) Body weights of WT and Rab27a null males at 8 or 20 weeks of age (n = 8–11/group). Quantification of percent lipid area within (B) PVAT, (C) BAT, and (D) iWAT from male Rab27a null and WT mice at both ages (n = 7/group). Quantification of WT and Rab27a null male thoracic aortae for (E) total, (F) medial, and (G) lumenal area (n = 5/group). (H) Body weights of WT and Rab27a null females at 8 or 20 weeks of age (n = 9–10/group). Quantification of percent lipid area within (I) PVAT, (J) BAT, and (K) iWAT from female Rab27a null and WT mice at both ages (n = 5/group). Quantification of WT and Rab27a null female thoracic aortae for (L) total, (M) medial, and (N) lumenal area (n = 5/group). Bars display the mean and SEM. A two-way ANOVA with a multiple comparisons test was applied to evaluate statistical significance. *, P<0.05. ****, P<0.0001.

Having observed RAB27A expression in thoracic aorta VSMCs and endothelial cells, the areas of the thoracic aorta, media and lumen were quantified. Global loss of RAB27A had no effect on total aortic area for either sex (Fig. 3E–G, 3L–N and Suppl. Fig. 4). However, aging of the male groups was significantly associated with increased thoracic aorta total (P = 0.0035), medial (P = 0.004), and lumenal (P = 0.01) areas. These findings are similar to what has been previously described in humans and C57BL/6 mice^5,39. Tissues from female mice showed no age-associated changes in vessel area when comparing 8-week-old to 20-week-old mice³⁹. These data further support that aging differentially affects male and female vascular phenotypes.

Global RAB27A loss was associated with cardiovascular system phenotypes in male PVAT and thoracic aorta.

To our knowledge, no prior studies have examined whether RAB27A is important for cardiovascular physiology. Previous studies had focused on defining the role of RAB27A in secretion and the immune response^10,27–29. It is well established that intercellular communication via secreted factors is critical for the maintenance of cardiovascular physiology^17,18. Therefore, our aim was to determine whether the proteomic profiles of male Rab27a null PVAT and thoracic aorta associated with cardiovascular phenotypes. Due to the known influence of sex and age on adipose and vascular physiology, murine male and female tissue samples from two age groups were utilized^2,5,39,40.

The proteomic profiles of PVAT (Fig. 4A, C) and thoracic aortae (Fig. 4F, H) from male mice showed changes in quantified protein levels at 8 and 20 weeks old. Results are based on the male data set with 1,834 quantified proteins of the 3,960 that were detected. Quantified proteins that met arbitrary thresholding criteria were associated with extracellular matrix, contractile, and metabolic functions^41–43. Phenotype enrichment analysis of the male PVAT (Fig. 4B, D, Suppl. Table 1) and thoracic aorta (Fig. 4G, I, Suppl. Table 2) tissue profiles provided additional evidence for an enriched association with cardiovascular and metabolic phenotypes. These phenotypes included cardiac hypertrophy, abnormal aortic wall morphology, and abnormal cardiac muscle relaxation. Of the proteins that contributed to the cardiovascular system phenotype category, and that were quantified for both ages, the majority trended towards reduced expression within the Rab27a null PVAT (Fig. 4E) and thoracic aorta (Fig. 4J). From each experimental group, a single male PVAT and thoracic aorta cross-section (n = 1/group) was randomly selected for immunofluorescence staining to evaluate if the proteomic trends could be confirmed via a secondary method. The expression of two contractile proteins, MYH6 and ACTA, confirmed the directional changes in expression for all groups, except for MYH6 expression in the 20-week-old male thoracic aorta (Suppl. Fig. 6).

Figure 4. Global RAB27A loss associated with cardiovascular, metabolic, and adipose phenotypes in male PVAT and thoracic aorta tissue.

Volcano plots display the proteomic profiles of pooled (n = 3) Rab27a null vs. pooled (n = 3) WT male (A, C) PVAT and (F, H) thoracic aorta tissues at 8 and 20 weeks old. Proteins that met the arbitrary significance threshold (P≤0.05) are plotted in magenta. Proteins that also met an arbitrary fold change threshold (FC≥20%) were utilized for phenotype enrichment analysis. Bar plots display phenotype enrichment results for (B, D) PVAT and (G, I) thoracic aorta samples. Bar length indicates the number of phenotypic terms associated with these parent category terms. Bar width indicates the proportion of total proteins associated with each parent term. Colors are coded as follows: cardiovascular system = red; homeostasis and metabolism = purple; adipose-specific = blue; other = gray. Heat plots display directional log(FC) of proteins that contributed to the cardiovascular phenotype and were quantified at both ages in the (E) PVAT and (J) thoracic aorta tissues, independent of the arbitrary significance threshold. Lollipop charts display the terms that populated the phenotype categories above, based on the -log(FDR), associated strength, and observed gene count of each term for PVAT (K and L) and thoracic aorta (M and N). Dot size represents the number of targets associated with each term. Dot color represents the strength of the association. Bar colors are coded as follows: cardiac-associated = red; metabolic-associated = purple; adipose-associated = blue; other = gray. Number of detected proteins = 3,960, number of quantified proteins = 1,834.

The Monarch database was used for phenotype enrichment analysis to connect phenotypes with the absence of RAB27A in PVAT (Fig. 4K–L) and thoracic aorta (Fig. 4M–N) via STRING²⁶. Recurring phenotypical annotations included cardiovascular anatomical and functional pathology, even in PVAT. Additionally, changes in the thoracic aortae with RAB27A loss at 20 weeks of age included annotations related to lipid droplets, which is interesting given the ability of VSMCs to accumulate lipid under pathological conditions⁴⁴. Male BAT and iWAT were used as depot controls and exhibited unique proteomic profiles compared to male PVAT. These tissues exhibited reduced cardiovascular system phenotype associations (Suppl. Fig. 7). However, pathway analysis still showed an enrichment for cardiac muscle contraction and fatty acid metabolism (Suppl. Tables 3–4). These data provided evidence that adipose depots may respond uniquely to global RAB27A loss.

Global RAB27A loss impacted female thoracic aortae and adipose proteomic profiles uniquely from male tissues.

The same workflow and thresholding criteria were used to analyze the female proteomic data. Results are based on the female data set with 2,416 quantified proteins of the 3,260 that were detected. Female proteomic profiles suggested variability between the two ages for both PVAT (Fig. 5A, B) and thoracic aorta (Fig. 5F, H). Similar to the male PVAT and thoracic aorta data, some of the quantified proteins with the greatest FC were associated with contractile functions^42,45. Phenotype enrichment analysis did not provide as strong of an association with cardiovascular or metabolic phenotypes for female PVAT (Fig. 5C, J and Suppl. Table 5) or thoracic aorta (Fig. 5F, H, K, L, and Suppl. Table 6) compared to the male tissues. For female PVAT, cardiovascular system was still the top category, while homeostasis and metabolism was the top category for the female thoracic aorta (Fig 5D, I). However, the number of terms associated with each category was greatly reduced compared to male profiles.

Figure 5. Global loss of RAB27A in female PVAT and thoracic aorta exhibited limited associations with cardiovascular, metabolic, and adipose phenotypes.

Volcano plots display the proteomic profiles of pooled Rab27a null (n = 3) vs. pooled WT (n = 3) female (A, B) PVAT and (E, G) thoracic aorta tissues at 8 and 20 weeks old. Proteins that met the arbitrary significance threshold (P≤0.05) are plotted in magenta. Proteins that also met an arbitrary FC≥20% were utilized for phenotype enrichment analysis. Bar plots display phenotype enrichment results for (C) PVAT and (F, H) thoracic aorta samples. Bar length indicates the number of phenotypic terms associated with these parent category terms. Bar width indicates the proportion of total proteins associated with each parent term. Colors are coded as follows: cardiovascular system = red; homeostasis and metabolism = purple; adipose-specific = blue; other = gray. Heat plots display directional log(FC) of proteins that contributed to the cardiovascular phenotype and were quantified at both ages in the (D) PVAT and (I) thoracic aorta tissues, independent of the arbitrary significance threshold. Lollipop charts display the terms that populated the phenotype categories above, based on the −log(FDR), associated strength, and observed gene count of each term for (J) PVAT and (K, L) thoracic aorta comparisons. Dot size represents the number of targets associated with each term. Dot color represents the strength of the association. Bar colors are coded as follows: cardiac-associated = red; metabolic-associated = purple; adipose-associated = blue; other = gray. Number of detected proteins = 3,260, number of quantified proteins = 2,416.

Monarch analysis identified some cardiac phenotypes associated with both PVAT and thoracic aortae in female Rab27a null mice (Fig. 5J–L). Meanwhile, the female Rab27a null BAT showed the greatest enrichment for cardiovascular phenotypes at 20 weeks of age (Suppl. Fig. 7). Pathway analysis further supported that the proteomic profiles of the female tissues were not strongly associated with cardiovascular phenotypes during global loss of RAB27A (Suppl. Table 7–8). Most of the proteins associated with a cardiovascular system phenotype in the PVAT (Fig. 5D) or thoracic aorta (Fig. 5I) at both ages, exhibited a decreased trend of expression with the loss of RAB27A. A single PVAT and thoracic aorta sample from each of the female experimental groups was quantified for MYH6 and ACTA fluorescence intensity (n=1/group) (Suppl. Fig. 6). These data exhibited divergent trends from the female proteomic data and the male quantifications.

Despite the role of RAB27A in endosome trafficking and secretion, our proteomic data showed no evidence for altered trafficking pathways the Rab27a null tissues¹⁰. To examine whether altered secretion phenotypes were present during global RAB27A loss, we evaluated male plasma for changes in circulating extracellular vesicle populations. These data suggest that while the size of circulating extracellular vesicles remained unchanged, the concentration of these particles have a greater variance in the Rab27a nulls compared to WT controls (Suppl. Fig. 9A–B). Interestingly, there appeared to be a separation into two populations of Rab27a null mice – those with relatively low circulating plasma exosome concentrations, and those with much higher concentrations compared to the average concentrations in WT mice (Suppl. Fig. 9B). We are currently pursuing this observation and addressing the question of whether exosomal cargo is altered with loss of RAB27A. Additionally, using primary PVAT adipocyte progenitor cells isolated from female mice, there was evidence of increased protein secretion with RAB27A loss (Suppl. Fig. 9C–D). However, no changes were observed in the expression of Rab family members in male PVAT or thoracic aorta (Suppl. Fig. 9E, F). These observations suggest that this change in protein secretion may be due to Rab-independent compensatory mechanisms.

Global RAB27A loss increased thoracic aorta contractile response to pharmacological stimulants in 8-week-old male mice.

To evaluate whether global RAB27A loss impacted cardiovascular physiology, thoracic aortae were examined for changes in vasoreactivity. Using wire myography, we observed increased vascular contractile and reduced vascular dilative responses when comparing the 8-week-old Rab27a null or WT male mice to their genotype matched, 20-week-old counterparts (Fig. 6A1, B1). These findings were expected as vasoreactivity is an indicator of vascular stiffness with aging for human and murine vessels³⁹. Thoracic aortae from 8-week-old male Rab27a null mice were significantly more contractile when stimulated with increasing doses of phenylephrine compared to age-matched WT controls (Fig. 6A2). Increased contractility was not observed in the 20-week-old Rab27a null male mice when compared to age-matched WT controls (Fig. 6A3). Additionally, only 8-week-old Rab27a null male mice exhibited reduced vasodilation when treated with acetylcholine (Fig. 6B2, B3). Vessel viability and sensitivity to both chemical agonists were largely unaffected for these comparison groups (Suppl. Fig 8A–C). Of the relevant comparisons, the only significant difference was observed when comparing the acetylcholine EC₅₀ of the 8-week-old Rab27a null group to the WT mice at both ages (Suppl. Fig. 8C). Since vascular health is dependent upon both contraction and relaxation, these data suggest that global loss of RAB27A induces an age-associated vasoreactive phenotype in male mice.

Figure 6. Rab27a null mice exhibit age- and sex-specific cardiovascular phenotypes.

Thoracic aorta segments from male and female WT and Rab27a null mice at 8 or 20 weeks old were examined for changes in vasoconstriction and vasodilation. The genotype-matched 8-week-old and 20-week-old male thoracic aortae (A1) contractile and (B1) dilative responses were compared. The male Rab27a null vs. WT thoracic aortae contractile responses at (A2) 8 weeks old and (A3) 20 weeks old were compared. Results from comparing the male Rab27a null vs. WT thoracic aortae dilative responses at (B2) 8 weeks old and (B3) 20 weeks old are shown. The genotype-matched 8-week-old and 20-week-old female thoracic aortae (C1) contractile and (D1) dilative responses were compared. The female Rab27a null vs. WT thoracic aortae contractile responses at (C2) 8 weeks old and (C3) 20 weeks old were compared. Results from comparing data from the female Rab27a null vs. WT thoracic aortae dilative responses at (D2) 8 weeks old and (D3) 20 weeks old are shown. Symbols display the mean and bars display SEM. Significance was determined via four-parameter non-linear regression test. *, P<0.05. **, P<0.01 ***<0.001. ****, P<0.0001. Cardiac function was analyzed in WT or Rab27a male mice. Representative H&E images from transverse sections through the left ventricle (E, scale bar = 40μm) and higher magnification view of longitudinal sections showing sarcomeric organization (F, scale bar = 20μm) from wild type or Rab27a null mice at 21 weeks of age. Boxed areas on left panels in (F) are magnified on the right to show sarcomeric structure (arrows).

Global loss of RAB27A increased thoracic aorta contractile and dilative response to pharmacological stimulants in 20-week-old female mice.

Due to known differences in the prevalence of cardiovascular disease and other risk factors between human males and females, female mice of matching genotypes and age groups were also examined for changes in thoracic aorta vasoreactivity². In contrast to the male cohorts, WT females showed reduced vasoconstriction in the older cohort (Fig. 6C1). Meanwhile, Rab27a null females showed no significant change in vasoconstriction with aging (Fig. 6C1). 8-week-old Rab27a null and WT female mice exhibited statistically similar contractile profiles, while the 20-week-old Rab27a null females retained contractility compared to the age-matched controls (Fig. 6C2, C3). No significant difference in vasodilative capacity was observed with aging for either genotype in response to acetylcholine (Fig. 6D1). In contrast to males, 8-week-old Rab27a null females showed no change in vasodilation, while the 20-week-old Rab27a null females exhibited increased vasodilation (Fig. 6D2, D3). Female thoracic aortae exhibited no changes in either agonist sensitivity or viability across all ages and genotypes (Suppl. Fig. 9D–F).

Global RAB27A loss reduced cardiac function in 21-week-old male mice.

The male proteomic analysis suggested that global RAB27A loss was associated with cardiac contractile dysfunction (Fig. 4K–N). Therefore, a longitudinal echocardiography study was performed using male Rab27a null vs. WT mice at 8, 16, and 21 weeks of age. Of the parameters examined, Rab27a null males exhibited decreased thickness of the left ventricle (LV) posterior wall at the 8-week timepoint (Table 1). At 16 weeks old, the Rab27a null males exhibited a ~17% decrease in stroke volume. By 21 weeks of age, the Rab27a null mice experienced a ~19% reduction in ejection fraction and ~25% reduction in fractional shortening, which demonstrates weakened myocyte contraction and overall LV dysfunction⁴⁶. Representative left ventricular histological images provide further evidence of altered cardiac structure in the Rab27a null males compared to WT controls. In longitudinal sections (Fig. 6E), cardiomyocytes from Rab27a null mice displayed a wavy morphology, compared to the straight cardiomyocyte pattern in the WT heart. Figure 6F shows the disorganization of the sarcomere structure in Rab27a null mice compared to the clear striation in WT hearts. In particular, the A-band appears thicker in Rab267a null hearts, suggesting disorganization of the myosin bundles.

Table 1. Echocardiography in age-matched WT and Rab27a null male mice.

WT and Rab27a null male mice at 8, 16, or 21 weeks of age were analyzed by ultrasound for cardiac measurements (n = 4/group). AW = anterior wall, d = diastolic, ID = internal diameter, LV = left ventricle, PW = posterior wall, s = systolic. The means ± SE are reported. Significance was determined via a two-way ANOVA with Bonferroni post-hoc multiple comparison test. Bolded boxes display comparisons with a P<0.02. Rab27a null values with asterisks (*) indicate trends towards significant differences (P = 0.06–0.075) from WT controls.

		8 weeks		16 weeks		21 weeks
Parameter	Unit	WT	Rab27a null	WT	Rab27a null	WT	Rab27a null
LVAW;d	mm	0.82 ± 0.032	0.78 ± 0.077	0.83 ± 0.025	0.81 ± 0.052	0.93 ± 0.078	0.7 ± 0.082 *
LVAW;s	mm	0.99 ± 0.086	1.09 ± 0.078	1.11 ± 0.082	1.04 ± 0.061	1.23 ± 0.084	1.04 ± 0.108
LVID;d	mm	3.74 ± 0.307	3.74 ± 0.067	4.09 ± 0.162	3.79 ± 0.154	3.9 ± 0.136	3.95 ± 0.089
LVID;s	mm	2.6 ± 0.356	2.63 ± 0.114	2.48 ± 0.264	2.59 ± 0.178	2.43 ± 0.205	2.88 ± 0.112 *
LVPW;d	mm	0.9 ± 0.116	0.67 ± 0.073	0.82 ± 0.028	0.83 ± 0.114	0.9 ± 0.065	0.69 ± 0.096 *
LVPW;s	mm	1.07 ± 0.098	0.87 ± 0.036	1.23 ± 0.068	1.12 ± 0.089	1.21 ± 0.191	0.87 ± 0.109 *
Heart Rate	BPM	468 ± 34	459 ± 20	474 ± 25	479 ± 19	482.6 ± 13	495 ± 16
Volume;s	μL	26.5 ± 6.1	25.4 ± 3.8	23.6 ± 4.9	21.7 ± 4.2	20.4 ± 3.6	30.4 ± 3.6
Volume;d	μL	65.7 ± 10.4	63.4 ± 3.0	77.3 ± 7.8	65.919 ± 1.5	67.3 ± 4.2	70.0 ± 3.1
Stroke Volume	μL	39.1 ± 4.5	38.0 ± 1.4	53.7 ± 4.1	44.3 ± 3.0	46.8 ± 2.4	39.5 ± 2.0 *
Ejection Fraction	%	60.6 ± 3.3	60.3 ± 4.3	70.1 ± 4.1	67.4 ± 5.7	70.0 ± 4.1	56.7 ± 3.5
Fractional Shortening	%	31.9 ± 2.2	31.9 ± 3.1	39.5 ± 3.4	37.3 ± 4.2	39.3 ± 3.3	29.4 ± 2.3
Cardiac Output	mL/min	18 ± 1.4	17.5 ± 1.3	25.6 ± 2.8	21.3 ± 2.2	22.6 ± 0.8	19.5 ± 0.9
Body weight	g	24.7 ± 1.9	24.8 ± 1	31.6 ± 3.1	32.3 ± 1.1	34.9 ± 4.5	36.2 ± 1.3

Discussion

There has been growing interest to understand the importance of RAB27A in human health due to its role in human disease^13,14,27. An endosomal trafficking protein, RAB27A mediates intracellular trafficking and the secretion of bioactive molecules^9,10. Loss of functional RAB27A impacts insulin secretion and the physiology of PVAT adipocyte progenitor cells^15,28. Metabolic health, adipocyte physiology, and intercellular signaling are critical factors that impact cardiovascular health^17,33,38. With RAB27A expressed throughout the thoracic aorta niche, we aimed to characterize whether global RAB27A loss altered PVAT, thoracic aorta, and cardiac phenotypes.

This study utilized two mouse models to explore the effects of global RAB27A loss on metabolic and thoracic aorta phenotypes. The first model was the Rab27a^ash strain, which carries a point mutation in Rab27a, causing the global loss of functional RAB27A on a C3H/HeSn background³². The second model was the novel Rab27a global null strain, which resulted from the targeted deletion of exon 4 on a C57BL/6J background. For reasons not yet defined, the Rab27a null strain was able to be maintained by homozygous mating, while the Rab27a^ash strain was maintained via heterozygous pairings. These differences in breeding viabilities suggested that some effects of global RAB27A loss may depend on genetic background. This is supported by a previously published study where the Rab27a^ash strain was backcrossed to a C57BL/6J background. Unlike the original Rab27a^ash strain, the Rab27a^ash on the C57BL/6J background exhibited normal clotting abilities²⁹. Additionally, while neurological phenotypes were not examined in the Rab27a global null strain, other studies suggest that such phenotypes are associated with the MyoVa mutant strains and not Rab27a mutant strains^27,29,34.

Metabolic consequences of Rab27a loss -

Data from the Rab27a^ash strain confirmed that global RAB27A loss impacted glucose sensitivity and resorption in a sex- and age-dependent manner. A study by Kasai et al. (2005), provided evidence that glucose-stimulation of RAB27A was required for proper insulin secretion from β cells²⁸. Our data provide support for the human genome-wide association studies data that RAB27A may impact human glucose signaling. However, the inclusion of females in this study provided evidence that RAB27A may induce sex-specific effects on metabolic phenotypes. Furthermore, glucose handling appears to be dependent on homozygous RAB27A loss, as no changes were observed between WT and ash/+ mice. While a prior study showed that loss of RAB27A reduced lipid accumulation in PVAT adipocyte progenitor cells, no morphological changes were detected for any of the adipose depots examined in the Rab27a null mice¹⁵. These findings could be due to the different research models used and the timing of RAB27A loss during development. However, male Rab27a null data suggested lipid area increased with age while the female data did not. These trends align with another study that showed sex-dependent differences in C57BL/6J mouse susceptibility to diet-induced obesity⁴⁷.

Vascular phenotype in Rab27a null mice -

Data from the Rab27a null strain showed that global loss of RAB27A impacted the contractile phenotype of the thoracic aorta. Like the morphological analyses, proteomic and wire myography data showed sex- and age-specific effects. While the aim of this study was not to define the underlying mechanism, we provide evidence that PVAT adipocyte progenitor cells isolated from Rab27a null females exhibited increased protein secretion. Additionally, no evidence for compensatory expression of other Rab family members was observed in male Rab27a null PVAT and thoracic aorta tissues. Together, these data suggest that a global loss of RAB27A may be inducing the activation of alternative intracellular trafficking pathways to maintain cellular signaling.

Age associated cardiomyopathy in Rab27a null mice –

Our longitudinal study of the cardiac contractile function in Rab27a null mice confirmed systolic dysfunction and supports the findings in our proteomic analysis of significantly lower Myh6 expression. Already with the small number of male mice investigated, we detected significant contractile changes at 16 weeks of age, with exacerbation at 21 weeks of age. The mouse heart predominantly expresses Myh6 (alpha-Myh,⁴⁸) and changes in the amount of alpha-Myh expression or mutations in this gene are associated with human and murine dilated cardiomyopathy^49,50. We did not detect a compensatory increased expression of Myh7 (beta-Myh), but also observed decreases in tropomyosin and troponin I, T, and C, as well as titin (see supplemental proteomic data set). Histological analysis showed gross cardiomyocyte disorganization and sarcomeric disarray with thickening of the sarcomeric striations in the Rab27a null mice. This finding indicates that cross bridge cycling in the Rab27a null mice is potentially less efficient compared to WT mice and contributes to the age associated decline in contractile function. Our findings are, to our knowledge, the first to describe cardiac contractile dysfunction in a mouse model with loss of an endosomal trafficking protein. There are only two cases reported in the literature linking Griscelli syndrome to cardiac dysfunction in humans⁵¹. Further work will elucidate how loss of Rab27a contributes to the downregulation of contractile protein expression in the mouse heart.

With the increasing prevalence of cardiovascular disease, establishing a better understanding of how intercellular and tissue-tissue communication impacts disease progression is imperative^38,52. Although this work provides foundational evidence for a role of RAB27A in cardiovascular physiology, further work is needed to identify the underlying mechanisms. Targeting RAB27A loss to specific cell types will also be critical to determine whether a specific cell population is responsible for the phenotypes described here. Having shown sex-specific effects of RAB27A loss on metabolic, vascular reactivity and cardiac contractile phenotypes, future work will also need to examine the molecular differences between male and female Rab27a null mice. Immunoblot analysis of male and female WT mice suggested differences in basal RAB27A expression in the iWAT, but not PVAT, thoracic aorta, or BAT (Suppl. Fig. 9G–J). It is likely that sex hormones may be major effectors of these described differences^53,54. Both our novel Rab27a global and conditional null strains presented here will be utilized to address these aims in future studies.

Highlights.

• Rare RAB27A variants associate with BMI and glucose sensitivity phenotypes.

• Homozygous Rab27a^ash males experience altered glucose sensitivity with aging.

• A new Rab27a null mouse strain was generated on a C57BL/6J background.

• A novel Rab27a conditional floxed allele strain was developed for future cell-selective studies.

• Proteomic comparisons of Rab27a null vs. WT PVAT and thoracic aorta tissue demonstrate changes in proteins related to cardiovascular and metabolic function.

• Rab27a null mice display cardiovascular defects including age-related changes in vasoreactivity and cardiomyopathy.

Abbreviations

Ash

ashen

BAT

brown adipose tissue

BMI

body mass index

Bp

base pairs

DDA

data dependent analysis

iWAT

inguinal white adipose tissue

KCl

potassium chloride

LV

left ventricle

PCR

polymerase chain reaction

PVAT

perivascular adipose tissue

SWATH

sequential window acquisition of all theoretical spectra

VSMC

vascular smooth muscle cells

WT

wild type