Perivascular Adipose Tissue Regulates Vascular Function by Targeting Vascular Smooth Muscle Cells

Abstract

Adipose tissues are present at multiple locations in the body. Most blood vessels are surrounded with adipose tissue which is referred to as perivascular adipose tissue (PVAT). Similarly to adipose tissues at other locations, PVAT harbors many types of cells which produce and secrete adipokines and other undetermined factors which locally modulate PVAT metabolism and vascular function. Uncoupling protein-1, which is considered as a brown fat marker, is also expressed in PVAT of rodents and humans. Thus, compared to other adipose tissues in the visceral area, PVAT displays brown-like characteristics. PVAT shows a distinct function in the cardiovascular system compared to adipose tissues in other depots which are not adjacent to the vascular tree. Growing and extensive studies have demonstrated that presence of normal PVAT is required to maintain the vasculature in a functional status. However, excessive accumulation of dysfunctional PVAT leads to vascular disorders, partially through alteration of its secretome which, in turn, affects vascular smooth muscle cells (VSMC) and endothelial cells (EC). In this review, we highlight the crosstalk between PVAT and VSMC and its roles in vascular remodeling and blood pressure regulation.

Graphical abstract

Adipose tissue develops naturally and is required for normal physiology in mammals. It is regarded as the largest endocrine organ which produces and secretes a large number of adipokines, cytokines and chemokines into the blood, affecting the functions of other organs from a distance¹. Adipose tissues at different locations are distinct in morphology and composition. The visceral and omental adipose tissues are pure WAT (white adipose tissue), while the subcutaneous adipose tissue is BeAT (beige adipose tissue) with clusters of adipocytes showing similar characteristics to classical BAT (brown adipose tissue) among the white adipocytes². PVAT (perivascular adipose tissue) shows characteristics of BeAT in human³ and BAT-like in mice⁴. However, PVAT is not always BAT in mice and humans. It depends on the anatomic location and environmental/metabolic context^{5, 6}. Excessive WAT accumulation is associated with CVD (cardiovascular diseases), hypertension, type 2 diabetes, cancers and other diseases, while BeAT and BAT might be negatively associated with CVD^7–10. The association between CVD and obesity was thought to operate through endocrine roles of adipose tissue. However, PVAT (perivascular adipose tissue) is tightly adherent to most blood vessels, including the aorta and arteries such as carotid, coronary, and mesenteric arteries. The paracrine effects of PVAT on blood vessels have not received much attention for a long time until 1991¹¹. Growing data from clinical and experimental animal models indicate that PVAT is engaged in paracrine cross-talk with blood vessels and is involved in the physiological homeostasis and pathological changes of the cardiovascular system^12–15. Other reviews in this series discuss the crosstalk between PVAT and other vascular cells, such as EC (endothelial cells) and immune cells. This review specifically discusses the crosstalk between PVAT and VSMC (vascular smooth muscle cells) and its roles in vascular remodeling and BP (blood pressure) regulation.

1. Developmental origins of PVAT adipocytes

1.1. UCP-1 and PPARγ are key molecules of PVAT adipocytes

The precise process and mechanisms of PVAT development remain unclear. The discovery of unique cell markers provided powerful tools to study the development of adipose tissues, including PVAT. Thoracic PVAT harbors two types of adipose progenitor cells: CD105⁺/CD73⁺/CD45^-/CD140A⁺/CD31^-, and CD105^-/CD73⁺/CD45^-/CD44⁺/CD90⁺/CD29⁺/CD34^-, both able to differentiate into UCP-1 (uncoupling protein-1, brown adipocyte marker) positive adipocytes in vitro^{16, 17}. PPARγ (peroxisome proliferator-activated receptor γ) is a key transcription factor for differentiation of all types of adipocytes. Disruption of PPARγ expression in adipocyte precursors will block maturation of adipose tissues. Consistently, PPARγ deletion in adipocytes results in lipodystrophy in mice¹⁸.

1.2. Adipocytes might originate from precursors in neighboring tissues.

Adipocytes from different locations may have distinct precursors. Albeit limited, experimental evidence documented that adipocytes share the same precursors with neighboring organs. For example, brown adipocytes in the interscapular area may share common progenitors with myoblasts which can be differentiated into either brown adipocytes or skeletal muscle fibroblasts¹⁹. PRDM16 (PR domain containing 16) contributes to adipogenesis as a coregulator of PPARγ. Compared to WAT in other depots, PRDM16 is highly expressed in subcutaneous WAT and BAT. Overexpression of PRDM16 in precursors of myoblasts resulted in a switch of differentiation to brown adipocyte lineage¹⁹.

1.3. PVAT adipocytes and VSMC share common progenitors

In a similar fashion that brown adipocytes in BAT and myoblasts share common progenitors, adipocytes in PVAT may share common progenitors with VSMC in the artery. Primary VSMC with overexpression of PRDM16 cultured in adipocyte differentiation cocktails had increased accumulation of lipid droplets and expression of thermogenic marker genes, suggesting a trans-differentiation to thermogenic adipocytes²⁰. Additionally, SM22α (smooth muscle protein 22 alpha) is first expressed within the dorsal aorta at embryonic day 9.5 and continues to be expressed in VSMC into adulthood²¹. We generated a smooth muscle-specific Cre mouse by knocking in the Cre into the SM22α promoter and deleted floxed PPARγ in VSMC²². We found that PVAT was completely missing throughout the thoracic and abdominal aorta and the mesenteric artery. However, the adipose tissues in other locations were not affected⁴. Together with a recent study that showed that PVAT in mouse starts to develop after birth²³, this phenotype suggests that PVAT adipocytes potentially share the same SM22α⁺ precursors with VSMC.

1.4. NCC and MSC are other sources of PVAT adipocytes

A more precise and recent study identified that adipocytes in PVAT from the aortic arch area belong to distinct populations. Lineage tracing indicated that adipocytes in the anterior PVAT are primarily derived from SM22α⁺ progenitors, whereas lateral PVAT contains both SM22α⁺ and Myf5⁺ (myogenic factor 5) cells²³. An additional study found that a subset of adipocytes in the PVAT at the aortic arch area originates from NCC (neural crest cells). The NCC-derived stromal vascular fraction (SVF) can be differentiated into both brown and white adipocytes²⁴. These studies demonstrated that PVAT adipocytes originate from SM22α⁺ progenitors and NCC. On the other side, PVAT harbors multipotent MSC (mesenchymal stem cells). Single-cell RNA sequencing revealed that stromal cells in thoracic PVAT express genes related to vasculature development²⁵. Furthermore, MSC in PVAT could differentiate to the VSMC lineage²⁶.

In summary, these studies demonstrated that VSMC, perivascular MSCs and PVAT adipocytes share common precursors. Although the literature discussed above show the contributions of various lineages to PVAT and to VSMC, there are several interesting questions that are still left unaddressed. For example, it is unknown whether PVAT from different lineages behaves any differently from one another; it is unknown whether the plasticity of PVAT lineages is involved in development of CVD such as atherosclerosis, hypertension, aortic aneurysm, etc. It was reported that adipose lineage cells in general, participate in formation of the blood vessels and enhance neovascularization in ischemic muscle in mice²⁷. It is likely that the progenitors in PVAT could be differentiated into VSMC which may contribute to vascular remodeling, as described below.

2. Contributions of signaling from PVAT to VSMC and vascular remodeling

Neointima formation is one of the major features of vascular remodeling, especially under vascular injury conditions such as atherosclerosis or mechanical damage. As described below, the adipocytes, stem cells, immune cells and other components in PVAT, such as EV (extracellular vesicles), can produce and secrete PVAT-derived factors such as adipokines, cytokines and growth factors which regulate vascular remodeling²⁸.

2.1. Perivascular cells and vascular remodeling

The origins of cells in the neointima remains unclear. It was reported that MSC from bone marrow contribute to neointima formation^29–35. Other literature points to contributions of local VSMC to neointima formation. Injurious stimuli may change VSMC from a contractile to a synthetic phenotype, which results in migration and proliferation^36–42. Recently, the contribution of resident progenitors from the adventitia or PVAT to neointima formation has also been explored. Those progenitors could migrate to the injury site where they differentiate towards vascular lineages. A study using a vein graft model in the mouse carotid artery demonstrated this possibility. Transplantation of RFP (red fluorescent protein)-labeled progenitors isolated from PVAT into the adventitial side of the vein graft significantly increased neointima formation in the graft. Interestingly, the RFP signal was found in the neointima area, and the RFP-positive cells in the neointima expressed the VSMC marker ACTA2 (alpha smooth muscle actin), indicating that progenitors in PVAT contributed to neointima formation⁴³.

The characteristics of the progenitors in PVAT change during the aging process. Matrigel implants with embedded periaortic stem cells from young mice placed in the perivascular area of carotid arteries after ligation injury, showed that those cells differentiated into EC or myofibroblasts in the neointima. These cells also differentiated into brown-like adipocytes in the perivascular region. Importantly, co-culture of VSMC with brown adipocytes differentiated from periaortic stem cells from young mice significantly inhibited VSMC proliferation. In contrast, brown adipocytes differentiated from periaortic stem cells of old mice promoted VSMC proliferation. These results suggest that periaortic stem cells from young mice differentiate into brown-like adipocytes and inhibit neointima formation, while aged stem cells lose this capability and promote neointima hyperplasia in the injured artery²⁵.

Additionally, as described above, periaortic PVAT may differentiate from NCC which highly express Wnt1 (wingless-type MMTV integration site family, member 1). Knockout of PPARγ mediated by Wnt1-Cre resulted in delay and dysplasia of PVAT development. Infusion of Ang II (angiotensin II) to these mice did not significantly change the SBP (systolic blood pressure) when compared to that of wild-type mice. However, Ang II infusion markedly aggravated media thickness and collagen accumulation in common carotid arteries but not the aortic arch²⁴.

Thus, in addition to MSC from bone marrow, local VSMC and progenitors from the adventitia, resident progenitors from the PVAT contribute to neointima formation during vascular remodeling. Mechanistically, TGF-β (transforming growth factor-beta) signaling promotes differentiation of PVAT resident progenitors into VSMC since they express Tgfbr2 (transforming growth factor beta receptor 2) and Anxa1 (annexin A1). Treatment of PVAT-derived progenitors with TGFβ promoted differentiation towards VSMC⁴³. The characteristics of those progenitors in PVAT and other signaling pathways which contribute to vascular remodeling remain unexplored. Additionally, the mechanisms underlying the enhanced Ang II-induced vascular remodeling in mice lacking PVAT in the aortic arch region due to deletion of PPARγ in NCC remain unknown. Additionally, the contributions of PVAT to vascular remodeling in disease conditions could be mediated by increased macrophages and T-lymphocytes infiltrating into a proinflammatory PVAT, which is reviewed independently in this series.

2.2. EV in PVAT and vascular remodeling

Most cell types in the cardiovascular system including EC, VSMC, macrophage and cardiomyocytes produce and release EV into the extracellular space. EV enclose biological contents derived from the original cells, and can be classified into exosomes (ranging from 30 to 100 nm), microvesicles/microparticles (ranging from 200 nm to 1 μm) and apoptotic bodies (ranging from 1 to 4 μm) according to their biogenesis and sizes. Exosomes are formed within the endosomal network and released by fusion with the plasma membrane. Microvesicles/microparticles are directly shed from the plasma membrane. Apoptotic bodies are released as blebs from cells undergoing apoptosis. EV are crucial regulators of vascular homeostasis by transferring biological messages including mRNAs, proteins, and noncoding RNAs to neighboring cells as cargo^44,45. For example, EC and VSMC can communicate through the release of EV, and laminar shear stress on EC induce production of EV containing miR-143/145, which inhibits proliferation and migration of VSMC^46–48.

PVAT also secretes EV which may convey paracrine signaling from PVAT to VSMC through microRNAs in EV. Obese mice secrete abundant EV containing microRNAs, which evoke inflammatory responses in PVAT and VSMC phenotypic switch from a contractile to a synthetic phenotype in aorta. VSMC can uptake EV secreted from PVAT and the encapsulated microRNAs, such as miR-221–3p, promote VSMC proliferation and migration⁴⁹. It is unclear which types of cells in PVAT generate EV. Mesenchymal stem cells in adipose tissue generate EV which were found to exert effects on angiogenesis, cell survival and apoptosis, inflammation, tissue regeneration, and reduction of disease pathology⁵⁰. Further studies are needed to examine the origins, characteristics and function of EV in PVAT and their potential as mediators of crosstalk with VSMC and EC.

2.3. PVAT-derived adipokines and vascular remodeling

PVAT-derived mediators such as adipokines, cytokines, ROS (reactive oxygen species) and gaseous compounds might contribute to vascular remodeling. We focus on PVAT-derived adipokines and their roles in vascular remodeling in this review.

Adiponectin is one of the most abundant adipokines normally produced and released by PVAT under physiological or pathophysiological conditions⁵¹. Neointima formation upon intravascular injury was markedly enhanced when PVAT was removed. Interestingly, neointima formation was markedly reduced by local perivascular delivery of recombinant adiponectin or transplantation of subcutaneous adipose tissue to the artery injury area⁵². Accordingly, neointima after injury was increased in adiponectin-deficient mice, and the conditioned medium from subcutaneous fat attenuated VSMC proliferation in response to PDGF-BB (platelet-derived growth factor-BB)⁵².

Leptin is another abundant adipokine in adipose tissues, including PVAT. Recombinant leptin or conditioned medium from visceral adipose tissue stimulated VSMC proliferation in vitro⁵³. Hydrosoluble protein growth factor(s) with a molecular mass >100 kDa released from PVAT adipocytes stimulated VSMC proliferation, which is enhanced in leptin receptor-deficient obese Zucker rats⁵⁴. These studies suggested that PVAT-derived leptin influences VSMC and is involved in neointima formation. Indeed, overexpression of leptin in PVAT enhanced cell proliferation and neointima formation in wild-type mice, but not in leptin receptor-deficient mice⁵³. Neointima formation was also enhanced by transplantation of visceral adipose tissue from obese wild-type mice on a high-fat diet to the carotid artery of immune-deficient mice. Interestingly, transplantation of visceral adipose tissue from ob/ob mice could not increase neointima formation, highlighting the importance of PVAT-derived leptin on vascular remodeling⁵³.

It has been well established that obesity is positively correlated with vascular dysfunction, partially due to inflammation and oxidative stress in adipose tissues⁵⁵. However, PVAT mass is markedly increased under the obese condition. Incubation of VSMC with conditioned medium from PVAT of obese rats promoted VSMC phenotypic switch⁵⁶, which was defined as any change in the normal function or structure of the differentiated VSMC⁵⁷. PVAT-derived adipokines might contribute to the VSMC phenotypic switch. Obesity is associated with increased leptin expression in visceral adipose tissue as well as in PVAT⁵³. Pretreatment with a leptin receptor antagonist inhibited obese PVAT-induced VSMC phenotypic switch⁵⁶. Mechanistically, p38 MAPK (p38 mitogen-activated protein kinases) signaling pathway is involved in VSMC phenotypic switch induced by obese PVAT. Leptin receptor antagonist upregulated p38 MAPK phosphorylation which was associated with inhibition of VSMC phenotypic switch⁵⁶.

The roles of the cAMP/ERK/p38 MAPK signaling pathway on PVAT-mediated VSMC proliferation was further evidenced by studies of visfatin, an adipokine preferentially expressed in PVAT, compared with subcutaneous and visceral adipose tissues. Visfatin could act as a nicotinamide phosphoribosyltransferase and biosynthesize NMN (nicotinamide mononucleotide) which mediated a proliferative response in VSMC via ERK (extracellular signal-regulated kinase) and p38 MAPK signaling pathways⁵⁸.

ERK signaling also contributes to VSMC phenotypic switch mediated by the adipokine CTRP 9 (C1q/TNF-related protein 9), which is secreted by adipose tissue. CTRP 9 functions as an adipokine that reduces serum glucose levels and is down-regulated in obese mice⁵⁹. Interestingly, CTRP9 enhanced VSMC differentiation mediated by hypoxia⁶⁰. Systemic delivery of an adenoviral vector expressing CTRP9 or CTRP9 recombinant protein significantly reduced neointima formation⁶¹. CTRP9 could attenuate VSMC proliferation induced by PDGF-BB which was associated with an increase in cAMP levels and a decrease of ERK phosphorylation in VSMCs⁶¹, suggesting that PVAT-derived CTRP9 might inhibit neointima formation via the cAMP-ERK signaling pathway. However, direct evidence is still needed. CTRP9 also significantly restrained VSMC proliferation through suppression of the TGF-β1/ERK1/2 pathway and promoted apoptosis in response to hypoxia⁶², suggesting that the TGF-β1/ERK pathway in VSMC inhibits proliferative signals from PVAT. Additionally, CTRP9 treatment significantly blocked the migratory potential of VSMC by decreasing the expression of the matrix metallopeptidases MMP-2/9. PVAT inflammatory responses significantly increased expression of activated MMP-2 in PVAT which leads to increased TGF-β1 in VSMCs. This, in turn, results in VSMC proliferation⁶².

PVAT-derived resistin might mediate OPN (Osteopontin) expression since a resistin-neutralizing antibody could attenuate obese PVAT-induced OPN expression⁶³. Alternatively, physiologic levels of resistin could induce a shift of VSMC proliferation to apoptosis when VSMC are co-cultured with macrophages⁶⁴. Since resistin in PVAT is mostly co-localized with macrophages, questions arise on the cellular source of local resistin in PVAT and the relative contribution of local and systemic resistin to vascular remodeling. Additionally, homocysteine is a known independent risk factor for CVD. Homocysteine can induce VSMC migration through adipocyte-derived resistin⁶⁵, which was associated with increased PKCε-dependent expression of MMP⁶⁶. Intimal hyperplasia and VSMC dysfunction associated with resistin were also related to PKCε- dependent Nox activation and ROS generation⁶⁷. Co-culture with PVAT from obese mice significantly increased OPN expression via the AP-1 signaling pathway in VSMC.

There are more adipokines secreted from PVAT than those discussed above⁶⁸. However, their role in VSMC growth and physiology and the underlying mechanisms remain to be systematically investigated. Adipokines such as leptin and chemerin also mediate formation of blood vessels through stimulating EC proliferation and migration, which is discussed in another review in this series. Furthermore, inflammatory signals from PVAT regulate the functions of cells in the vascular wall including VSMC, which is specifically discussed in another review in this series.

2.4. PVAT-derived growth factors and vascular remodeling

Growth factors, including thrombospondin-1, serpin-E1, TGF-β, PDGF-BB, VEGF (vascular endothelial growth factor), bFGF (basic fibroblast growth factor), PLGF (placental growth factor), HGF (hepatocyte growth factor) and ILGFBP-3 (insulin-like growth factor-binding protein-3), have well-known effects in stimulation of VSMC proliferation and migration and have been previously reviewed^69–71. PVAT adipocytes express and secrete high amounts of all those growth factors^{72, 73}. The significance of PVAT-derived growth factors on vascular remodeling is highlighted by a study on VEGF from PVAT of type 2 diabetic patients. Conditioned medium from PVAT adipocytes significantly increased expression of VEGF-R (VEGF-receptor) 1 and 2, and VEGF secretion from VSMC, which further induced VSMC proliferation⁷⁴. Further studies are needed to investigate the effects of growth factors from PVAT on VSMC survival and growth.

In summary, these studies have demonstrated that PVAT-derived factors or multiple types of cells in PVAT are actively involved in adjacent vascular remodeling under disease conditions (Figure 1). This crosstalk is critical to normal vascular function as well. However, PVAT produces and secretes a multitude of factors including adipokines, cytokines, growth factors and other factors yet undetermined. The types and mechanisms of these PVAT-derived factors on vascular remodeling remain to be further investigated. Follow up extensive studies will be needed to further understand the contribution of PVAT to vascular remodeling and elucidate the underlying mechanisms.

Figure 1. Crosstalk between PVAT and VSMC in vascular remodeling.

Smooth muscle-like cells in the neointimal area of the vascular wall originate from four sources: 1) Wnt1 positive progenitors in PVAT; 2) PVAT-resident or bone marrow-derived mesenchymal stem cells; 3) local VSMC stimulated by signaling from PVAT adipocytes and mediated by adipokines, growth factors and extracellular vesicles; and 4) adventitia-derived cells. Additionally, Wnt1-positive progenitors in PVAT and VSMC in the vascular wall can be differentiated into brown-like adipocytes in PVAT.

3. PVAT regulates vascular tone by targeting VSMC

PVAT-derived bioactive factors not only contribute to vascular remodeling but are also involved in vascular homeostasis by modulating vascular tone. Next, we summarize the recent progress regarding PVAT-derived factors on vascular tone regulation.

Using in vitro organ bath or wire myograph techniques, the vasoactivities of blood vessel rings from thoracic aorta, abdominal aorta, mesenteric, carotid or coronary arteries with or without PVAT (or by incubation of vessel rings with a conditioned buffer from an intact piece of PVAT or with PVAT extracts) was studied. The presence of PVAT on the vessel rings markedly attenuated the contractile response to norepinephrine (referred to as anti-contractile effect)¹¹. PVAT’s anti-contractile effects were observed as well in response to stimulation with phenylephrine, 5-HT (serotonin) or Ang II. The anti-contractile property of PVAT was further confirmed by incubation of PVAT-free vessel rings with a PVAT-conditioned solution prepared in vitro, which produced a significant relaxation response^{75, 76}.

Anti-contractile effects of PVAT might be mediated by uptake and/or metabolism of the vasoactive amines such as dopamine, norepinephrine, and serotonin in PVAT itself ⁷⁷. PVAT adipocytes have MAO-A/B (monoamine oxidase A/B) and SSAO (semicarbazide-sensitive amine oxidase) which catalyze the metabolism of amines in PVAT⁷⁷. MAO-A/B catalyzes the oxidative deamination of amines⁷⁸, and SSAO is a multi-functional enzyme that promotes the generation of H₂O₂ and NH₃⁷⁹. Indeed, inhibition of MAO and SSAO, or inhibition of NET (norepinephrine transporter) with nisoxetine increased vasocontraction induced by norepinephrine on vessel rings with PVAT⁷⁷. Additionally, the PVAT’s anti-contractile effects are mediated by PVRF (PVAT-derived relaxing factors). Recent studies also demonstrated that PVAT secretes PVCF (PVAT-derived contracting factors), which induce blood vessel contraction^{4, 80–82}. PVRF and PVCF influence local vascular tone through endothelium-dependent or endothelium-independent effects^{83, 84}.

3.1. PVRF regulate anti-contractile effects through NO production by the endothelium

The anti-contractile effect of PVAT was lost in endothelium-denuded rings⁸⁵, suggesting that intact-endothelium is required to induce relaxation of vessel rings. Endothelium-derived NO (nitric oxide), one of the most potent vascular dilators, stimulates guanylate cyclase in VSMCs, resulting in increase of cyclic GMP and activation of signaling cascades to induce blood vessel relaxation. PVAT expresses eNOS and produces NO⁸³. The anti-contractile effects of PVAT on blood vessels might be mediated by PVAT-derived NO, since NOS (nitric oxide synthase) inhibition with L-NAME (Nω-nitro-L-arginine methyl ester) enhanced the phenylephrine-induced contraction in endothelial-denuded rings with PVAT⁸³, indicating that NO might be one of the PVRF. Ang-(1–7) and leptin are PVRF which mediate anti-contractile effects in an endothelium-dependent manner since this effect was prevented by inhibition of NOS or NO scavenger^{85, 86}. The details of endothelium-dependent anti-contractile effects of PVAT are discussed in another review in this series.

3.2. PVRF regulate anti-contractile effects through modulation of VSMC function

Inhibition of NOS attenuated the anti-contractile effects of PVAT in PVAT-intact vessel rings^{83, 84, 87, 88}. Therefore, it is possible that PVRF, such as NO, directly target VSMC to induce vessel relaxation. Indeed, numerous studies documented that PVAT controls vascular tone through potassium channels in VSMC. Blockage of Ca²⁺-activated potassium channels by tetraethylammonium chloride, or inhibition of adenosine triphosphate (ATP)-dependent potassium channels by glibenclamide, or Kv7 (KCNQ) voltage-dependent potassium channels by XE991 in VSMC markedly reduced the anti-contractile effects of PVAT^89–92.

H₂S (hydrogen sulfide) gas is a vasodilator^{93, 94}. Interestingly, expression of cystathionine β-synthase and cystathionine-γ-lyase, the enzymes required for H₂S endogenous production in PVAT⁹⁵ suggests that H₂S is a PVRF. Pre-treatment with propargylglycine, an inhibitor of H₂S production, significantly increased the noradrenaline-induced contraction of vessel rings with PVAT^{93, 96}. ATP-sensitive K⁺ channels in VSMC might mediate the anti-contractile effects of PVAT-derived H₂S^{93, 97, 98}. Ang-(1–7) might act as a hyperpolarizing factor through K(Ca) channels as well to cause relaxation of the blood vessel⁸⁶.

However, these results are based on pharmacological evidence in vitro. The specificity of these drugs in vivo is uncertain. Studies using K⁺ channel knockout mice suggest more complicated mechanisms behind the roles of K⁺ channels in PVAT’s anti-contractile effects. Kcna5^−/− arteries and wild-type arteries incubated with DPO-1 (neomenthyl diphenylphosphine oxide), an inhibitor of the Kv1.5 potassium channel, showed normal vasoconstriction in response to phenylephrine in the presence and absence of PVAT. KV current density and response to inhibition by XE991, a KCNQ channel blocker, were normal in mesenteric artery VSMC isolated from Kcna5^−/− mice⁹¹. Additionally, the anti-contractile effects of PVAT in Kcnq1^−/− mesenteric arteries were comparable to those in wild-type mice. The anti-contractile effects of PVAT were not affected by KV 7.1 channel blockers such as chromanol 293B and HMR1556. VSMCs isolated from Kcnq1^−/− mice exhibited normal peak KV currents. The KV 7.2–5 channel opener retigabine caused similar relaxation in Kcnq1^−/− and wild-type vessels as well. Therefore, KV 7.1 channels were apparently not involved in the control of arterial tone by PVAT⁹⁹.

It is possible that only part of them mediate PVAT’s anti-contractile effects. At least five classes of K⁺ channels including BKCa (large-conductance Ca²⁺-activated K⁺) channels, KCa3.1(intermediate-conductance Ca²⁺-activated K⁺) channels, KV (multiple isoforms of voltage-gated K⁺) channels, KATP channels, KIR (inward-rectifier K⁺) channels, and K2P (members of the two-pore K⁺) channels are expressed in VSMC¹⁰⁰. Only inhibition of BKCa channel or knockout BKCa blocked PVAT-induced anti-contractile effects¹⁰¹. Additionally, PVAT from different locations (i.e., thoracic aorta, abdominal aorta or mesenteric artery) might mediate anti-contractile effects via distinct K⁺ channels.

Whether other mechanisms contribute to PVAT’s anti-contractile effects needs to be further investigated. Thus, for instance, it is unclear whether PVAT-derived adipokines mediate PVAT’s anti-contractile effects through modulation of VSMC function. PVAT-conditioned solution transfer studies documented that PKG (cGMP-dependent protein kinase) was necessary for PVAT’s paracrine effects on smooth muscle and endothelium of blood vessels. The anti-contractile effect of PVAT was not present in arterial rings isolated from PKG^−/− mice, or arterial rings isolated from wild-type mice with inhibition of PKG signaling. Activation of PKG by ANP (atrial natriuretic peptide) rescued the loss of PVAT’s anti-contractile effects on wild-type vessel rings, but not on PKG^−/− vessel rings upon hypoxia stimulation. Interestingly, adiponectin expression in PVAT adipocytes was reduced in PKG^−/− mice, and ANP could not restore the reduced anti-contractile capacity of PVAT from adiponectin^−/− mice. These data suggest that PVAT-derived adiponectin modulates vascular tone via PKG signaling in VSMC¹⁰².

3.3. PVCF induces contractile effects

Apart from anti-contractile effects, PVAT also induces contractile effects on VSMC. Perivascular nerve activation by EFS (electrical field stimulation) elicited a frequency-dependent contractile response in mesenteric artery rings both in the presence and absence of PVAT, but the amplitude of contraction was higher in arterial rings with PVAT than in those without PVAT¹⁰³, suggesting that PVAT can release vasoconstrictors. Interestingly, Ang II treatment of mesenteric artery rings enhanced EFS-induced contraction, while arterial rings with intact PVAT treated with ACE (angiotensin-converting enzyme) inhibitor or AT1R (type 1 angiotensin II receptor) blockers showed reduced vessel contraction induced by EFS¹⁰⁴.

PVAT’s contractile effects were further evidenced by incubation of PVAT-null vessel rings with extracts or conditioned buffer of PVAT. We found that incubation of vessel rings of aorta, mesenteric or carotid artery from mice lacking PVAT with extracts of thoracic or mesenteric PVAT, which were obtained by mincing PVAT into small pieces or homogenizing PVAT, markedly constricted the vessel rings⁴. Pretreatment of the blood vessel rings with AT1R blockers significantly attenuated contraction of the vessel rings induced by PVAT extracts⁸⁰, demonstrating that PVAT-derived Ang II contributed to the contractile effects of PVAT. Additionally, the conditioned buffer of coronary PVAT increased the baseline tension of coronary arteries, and this effect was dependent on the amount of PVAT added to the bath⁸¹. The increase in the basal tone of the coronary artery was markedly augmented by the conditioned buffer from PVAT of obese swine compared to lean swine. Additional proof-of-principle studies demonstrated that PVAT-derived calpastatin constricts arterial rings via the Rho-kinase pathway and an increase in VSMC Ca²⁺ handling via Ca_v1.2 channels and H₂O₂-sensitive K⁺ channels⁸¹.

PVAT can release neurotransmitters to constrict blood vessels as well⁸². PVAT surrounding the rat thoracic aorta contains significant amounts of 5-HT. Fenfluramine increased the secretion of 5-HT and norepinephrine from PVAT. Consistently, the contraction of thoracic aortic rings with PVAT in response to fenfluramine was higher than rings without PVAT. This effect was diminished by prazosin (an α-adrenergic receptor antagonist) or nisoxetine (a norepinephrine transporter inhibitor)⁸². This phenotype was also observed in PVAT from the renal artery. The maximum contraction to the sympathomimetic tyramine was higher in the renal artery with PVAT versus without PVAT, and this effect was also reduced by prazosin and nisoxetine¹⁰⁵.

In addition to 5-HT, the adipokine chemerin and its receptor ChemR23 co-localize with tyrosine hydroxylase, a sympathetic nerve marker, in mesenteric PVAT. Treatment of the vessel rings with intact PVAT with CCX832, an antagonist of ChemR23, or with prazosin blocked chemerin-9 or EFS induced vessel ring contraction^{106, 107}.

These studies demonstrate that calpastatin, Ang II, 5-HT, chemerin and norepinephrine locally-produced in PVAT are potent PVCF which stimulate blood vessel constriction, most likely through their respective receptors on VSMC.

3.4. Potential mechanisms underlying PVAT-induced VSMC contractility

PVAT’s contractile effects might be regulated by oxidases including NADPH (nicotinamide adenine dinucleotide phosphate) oxidase, SOD (superoxide dismutase) and catalase. PVAT expresses the p67phox subunit of the NAD(P)H oxidase, Mn-SOD (superoxide dismutase), and CuZn-SOD^{84, 103}. Stimulation of PVAT with norepinephrine increased Mn-SOD expression, decreased catalase expression, and induced O₂^.- generation in PVAT ¹⁰⁸. Thus, it was hypothesized that mitochondria-derived ROS (reactive oxygen species) in PVAT modulates vascular reactivity¹⁰⁸. Uncoupling mitochondria, as well as removal of H₂O₂, increased the contraction in vessel rings with PVAT present in response to norepinephrine¹⁰⁸. EFS increased O₂^.- generation in isolated PVAT, and this effect was attenuated by NAD(P)H oxidase inhibition. Treatment with NAD(P)H oxidase inhibitors (apocynin and DPI) further attenuated the contraction in response to EFS in the vessel rings with PVAT than in those without PVAT, while exogenous O₂^.- augmented the contractile response to EFS and to phenylephrine in vessel rings without PVAT¹⁰³. Therefore, PVCF mediated ROS generation in PVAT acts as a pivotal signaling molecule regulating VSMC contraction.

3.5. Bioactive factors in PVAT can act as either PVRF or PVCF

EFS elicits an anti-contractile effect on mesenteric arteries with PVAT in an endothelium-independent manner^{109, 110}. Furthermore, sympathetic stimulation in PVAT triggers the release of adiponectin via β3-adrenoceptor activation¹¹⁰. PVAT also acts as a reservoir for noradrenaline, preventing it from reaching the vessel and causing contraction¹¹⁰. Additionally, perivascular sympathetic-sensory interactions have been shown to regulate calcitonin gene-related peptide (CGRP)-mediated vasodilation in rats via the release of methyl palmitate¹⁰⁹. Therefore, in addition to the data discussed above showing that EFS induces VSMC contraction via ROS, stimulation of perivascular sympathetic nerves results in secretion of both PVCF and PVRF.

Actually, the same factor in PVAT can act as either PVRF or PVCF. PVAT-derived H₂S acted as an anti-contractile factor in normotensive Wistar rats. However, in SHR (spontaneously hypertensive rat), the H₂S in PVAT acted as a pro-contractile factor which was associated with the stimulation of perivascular nerves⁹⁶ The pro-contractile effect of H₂S in the arterial wall could represent a pathologic feature. Similar to H₂S, PVAT-derived prostanoids such as PGE2 (prostaglandin E2) and PGI2 (prostacyclin) were responsible for anti-contractile effects of the normal PVAT^{84, 89, 103, 111}. On the other side, PVAT-secreted thromboxane TXB2, PGE2 and PGF2α were responsible for the impairment of PVAT’s anti-contractile effects upon HFD (high fat diet) feeding¹¹². Therefore, H₂S and prostanoids in PVAT have anti-contractile effects under normal conditions, while they induce contractile effects under disease conditions, likely reflecting pathological changes in the target cells. Further studies are needed to distinguish the contributions of the cellular targets, VSMC and EC, on the contractile/anti-contractile effects of these PVAT-secreted bioactive factors.

In summary, the experimental data discussed demonstrate that PVAT produces PVRF such as H₂S, Ang (1–7) and methyl palmitate, which could induce vasodilation by opening K⁺ channels on VSMC. PVAT also releases PVCF such as calpastatin, Ang II, 5-HT, chemerin and norepinephrine which could induce vasoconstriction by, in some cases, still underdetermined mechanisms, albeit most likely mediated by their receptors on VSMC. Thus, those PVAT-derived bioactive factors collectively and coordinately regulate vascular tone by targeting VSMC (Figure 2).

Figure 2. PVAT regulates VSMC dilation and contraction.

PVAT (perivascular adipose tissue) regulates vascular tone through relaxing factors and contractile factors. PVAT adipocytes express CBS (cystathionine β-synthase), CSE (cystathionine γ-lyase) and eNOS (endothelial nitric oxide synthase). CBS and CSE catalyze generation of H₂S (hydrogen sulfide) from cysteine, which induces VSMC (vascular smooth muscle cells) dilation by opening BKCa (calcium activated K⁺-channels) on VSMC, and eNOS-dependent NO production from L-arginine which induces vasodilation mediated by the cGMP signaling pathway. Additionally, EFS (electrical field stimulation) induces PVAT adipocytes to release adiponectin, PGI2 and CGRP which induce vasodilation as well. On the other side, EFS induces PVAT release of 5-HT, dopamine and norepinephrine which cause VSMC contraction thorough adrenergic receptors (AdR) on VSMC. EFS also induces VSMC contraction by stimulating secretion of PGF_2a, chemerin and O₂^.- from PVAT adipocytes. Additionally, PVAT adipocyte-derived calpastatin might induce VSMC contraction mediated by Ca_v1.2 channels on VSMC.

4. PVAT modulates vascular tone and BP in vivo

VSMC are the medium layer and an essential structural component of the blood vessels. VSMC play critical physiological functional roles in the regulation of vascular tone. As described above, in vitro experimental data documented that PVAT-derived factors regulate vascular tone in part by targeting VSMC. Therefore, it was hypothesized that PVAT modulates BP under physiological and/or pathophysiological conditions.

BP exhibits a robust circadian rhythm, which causes a dipping phenotype of BP between light-on and light-off periods in mice. It is unclear regarding the essential and physiological significance of the circadian rhythm and dipping pattern of BP in humans. However, in hypertension, sleep apnea, and even shift work, this balanced rhythm is perturbed^{113, 114}. Changes in the dipping phenotype in particular, which may manifest from dippers to extreme dippers, non-dippers and reverse dippers, have been associated with adverse outocomes^115–117. Lack of nocturnal BP fall (non-dipping) has been shown to be more closely associated with target organ damage and worsened cardiovascular outcome compared with essential hypertension in patients with dipping pattern¹¹⁸. Apart from the central circadian clock in the suprachiasmatic nucleus of the hypothalamus, a local biological clock in VSMC of the blood vessels contributes to regulation of BP circadian rhythm^{58, 119, 120}. It was unknown if PVAT-derived factors target VSMC to regulate circadian rhythm and dipping patterns of BP.

We reported that the BP of mice lacking PVAT generated by deletion of PPARγ in VSMC was comparable with that in mice with intact PVAT during the night period (active phase in mice). However, the BP of mice lacking PVAT was markedly lower during day-time when the mice were in the resting phase, which resulted in an extreme-dipper phenotype^{4, 22}. This phenotype might be caused by a lack of PVAT because the development of adipose tissue in other locations such as subcutaneous, gonadal, para-renal and interscapular was not affected⁴. Notably, another mouse model, in which PPARγ was deleted in VSMC by a transgenic SM22α Cre mouse tool, showed intact PVAT. This mouse model displayed a hypertensive phenotype¹²¹. A dominant negative PPARγ mutation (P467L) associated with hypertension in humans¹²², showed hypertension as well in a mouse model carrying the same mutation in VSMC¹²⁰. Therefore, the hypotensive phenotype in mice lacking PVAT was most likely caused by a lack of PVAT, rather than by deficiency of PPARγ in VSMC.

PVAT regulation of the BP dipper pattern was further documented by another mouse model with Agt (angiotensinogen) deficiency in brown adipocytes⁸⁰. As described above, Ang II is one of the PVCF. We found that UCP1-Cre-mediated deficiency of Agt in brown adipocytes reduced Ang II levels in thoracic PVAT, and decreased BP in the resting phase without affecting BP in the active period, which also caused an extreme-dipper BP⁸⁰. However, the limitation of this mouse model is that Agt was also deleted in brown adipocytes in other locations which may affect the overall secretome from BAT. Currently, we do not have evidence that BAT-derived Ang II systemically affects BP because the plasma levels of Ang II were not reduced in brown adipocyte selective Agt deficient mice⁸⁰. Additionally, we reported that PVAT has a peripheral clock. Knockout of Bmal1 in brown adipocytes of mice also reduced BP during the resting phase and we further documented that Bmal1 transcriptionally regulates Agt expression and Ang II generation in PVAT⁸⁰.

Thus, while there are limitations intrinsic to each model, these results from mouse models demonstrated that paracrine effects of PVAT contribute to circadian BP regulation and may account for the yet not fully understood clinical observations associated with the dipper and non-dipper phenotypes in humans. The CVD outcomes associated with the regulation of BP dipping by PVAT, and the contributions of crosstalk between PVAT and VSMC to this phenotype should be further investigated.

5. Altered PVAT features from BAT-like to WAT-like contributes to obesity-associated hypertension

PVAT is not only involved in BP regulation under physiological conditions but also in hypertension. The results from animal studies indicated that HFD (high-fat diet) feeding significantly increased PVAT mass¹²³, which results in white-like characteristics of PVAT. Notably, the anti-contractile effects of PVAT were attenuated under the obese condition^{87, 124, 125}. The PVAT with changes in features from BAT-like to WAT-like might be associated with hypertension due to loss of PVAT’s anti-contractile effects. When the PVAT was removed from vascular rings, the relaxant responses to vasodilators were no different between rings isolated from obese and lean mice, suggesting that obesity did not impair the intrinsic vascular bed reactivity but rather the regulatory PVAT function¹²⁶. Therefore, it is likely that PVAT’s dysfunction is related to the development of obesity-associated hypertension. Indeed, mesenteric arterial rings incubated with aortic PVAT from rats on a HFD for 6 months demonstrated lower endothelium-dependent relaxation. This effect was absent in mesenteric arterial rings incubated with aortic PVAT from rats on a standard chow diet¹²³. The impaired anti-contractile effects of PVAT upon HFD feeding were not only dependent on the endothelium, but also consequence of reduced NO bioavailability due to L-arginine deficiency and eNOS uncoupling in white-like PVAT^{88, 127–129}. Ex vivo L-arginine treatment and arginase inhibition could reverse obesity-induced vascular dysfunction¹²⁸.

Additionally, obesity and aging cause inflammation in white-like PVAT characterized by infiltration of macrophage and dendritic cells with high expression of inflammatory adipokines and cytokines, including resistin, visfatin, leptin, MCP-1 (monocyte chemotactic protein-1), TNF-α and IL-6^{63, 125, 130, 131}, while the mRNA levels of the anti-inflammatory mediator adiponectin are decreased in obese PVAT. Depletion of dendritic cells improved the ability of PVAT to augment acetylcholine-induced vasorelaxation and anti-contractile activity¹³². On the other side, HFD reduction of the anti-contractile effect of PVAT was associated with reduced adiponectin secretion and AMPK (AMP-activated protein kinase) phosphorylation¹³³. Chronic adiponectin treatment normalized NO-dependent vasorelaxation by increasing eNOS phosphorylation in mesenteric arteries of HFD-fed rats¹³⁴. PVAT from AMPKα1 knockout mice had increased macrophage infiltration and significantly reduced adiponectin secretion¹³³. Co-culture of VSMCs with PVAT adipocytes from rats on an HFD reduced the AMPK phosphorylation and increased mTOR (mammalian target of rapamycin) phosphorylation¹²³. Aortic Rictor is an essential mTORC2 component. HFD feeding reduced Rictor gene expression in PVAT. Increased contraction and impaired dilation were found in thoracic aortic rings with PVAT from adipocyte-specific Rictor deficient mice, and inhibition of iNOS normalized vascular reactivity in aortic rings from Rictor^−/− mice¹³⁵. PVAT’s anti-contractile effect mimicking the obese phenotype was also lost in mice deficient in eosinophils. Eosinophil reconstitution restored PVAT’s anti-contractile effects accompanied by increased NO bioavailability and adiponectin in PVAT¹³⁶.

Inflammation also stimulated generation of O₂^.- and H₂O₂ in PVAT from HFD-fed mice. Vessel rings with intact obese PVAT showed increased contraction to phenylephrine. Inactivation of O₂^.-, dismutation of mitochondrial-derived H₂O₂, or uncoupling of oxidative phosphorylation decreased phenylephrine-induced contraction in vessels with PVAT from HFD-fed mice^{124, 125}. Consistently, SOD [Cu-Zn], peroxiredoxin-1 and adiponectin were reduced in obese humans compared to healthy subjects¹²⁷. Incubation with SOD, catalase or TNF-α attenuated contractility in vessel rings presenting normal PVAT, while incubation of aortic rings containing obese PVAT with anti-TNF-α antibodies or free radical scavengers partially restored the anti-contractile effect of PVAT^{87, 127}. Additional data from genetically modified mice further support the conclusion that inflammation and oxidative stress in PVAT altered the PVAT’s anti-contractile effects under obese conditions. PVAT from obese mice lacking TNF-α receptors in PVAT prevented H₂O₂ generation and vasocontraction¹²⁴. Conversely, the anti-contractile function of PVAT was impaired in mice with PVAT-specific IL-18 deficiency. This was accompanied by decreased MnSOD expression in deformed mitochondria in PVAT and increased PVAT whitening ¹³⁷.

Thus, brown-like PVAT could prevent inflammation and oxidative stress under physiological conditions, while white-like PVAT was accompanied by increases in inflammation and oxidative stress, and decrease of NO bioavailability under obese conditions. The alteration of brown-like PVAT to white-like PVAT might be associated with the development of hypertension under obese conditions.

6. Potential for prevention of hypertension by restoring brown-like PVAT

PVAT displays heterogeneity according to species and locations. The PVAT is brown-like at areas of larger size blood vessels, and is white-like at areas of the smaller size blood vessels. Compared to the same locations, the PVAT of rodents is more brown-like than that of humans. Nevertheless, there are clusters of brown-like adipocytes in PVAT, or BeAT, of both rodents and humans¹³⁸. Obesity is one of the risk factors for primary hypertension. As described above, the PVAT gradually changes into white-like characteristics during development of obesity, which is associated with alterations of PVAT paracrine profiles, including those factors involved in VSMC growth, regulation of vascular tone and BP. Conservation of PVAT’s brown-like features might be a strategy to prevent development of hypertension by maintaining the homeostasis of blood vessels (Figure 3).

Figure 3. Strategy for hypertension prevention through PVAT browning.

Brown-like adipocytes in PVAT produce more PVAT-derived relaxing factors (PVRF), while white-like adipocytes produce more PVAT-derived contraction factors (PVCF). PVRF mediate anti-contractile effects of PVAT and reduce BP (blood pressure), while PVCF account for contractile effects of PVAT and increase blood pressure. Induction or maintenance of PVAT browning by drugs, but not cold stimuli, can be an efficient strategy to prevent hypertension development.

Multiple strategies including cold stimuli or growth factors such as FGF21 (fibroblast growth factor 21)¹³⁹, ANP (atrial natriuretic peptide)¹⁴⁰ and BMP (bone morphogenetic proteins) ¹⁴¹ are able to convert white-like WAT into BeAT through a “browning” process. It is hypothesized that browning is beneficial to prevent obesity and associated CVD. Therefore, restoring the normal beige features of human PVAT might reverse or prevent development of vascular disorders¹⁴². It is unknown whether the whitening-like process of PVAT under obese conditions results in elevation of BP. However, it is reasonable to hypothesize that restoring PVAT to brown-like characteristics could be a potential strategy for hypertension treatment.

Even though cold acclimation is the strongest stimuli to induce WAT browning process, unfortunately, cold significantly increases BP and heart rate and leads to more cardiac events¹⁴³. Therefore, strategies other than cold stimuli are required to safely induce WAT or PVAT browning. Indeed, mitoNEET is a mitochondrial membrane protein that is regulated by thermogenic genes such as PGC1 (peroxisome proliferator-activated receptor-gamma coactivator 1). We reported that overexpression of mitoNEET in brown adipocytes, including PVAT adipocytes, significantly prevented arterial stiffness¹⁴⁴ and atherosclerosis¹⁴⁵. The Crataegus sp. extract WS® 1442 is an herbal compound that is beneficial in supporting cardiovascular function in heart failure patients¹⁴⁶. Treatment of obese mice with WS® 1442 normalized vascular function without changing fat mass. The effect was associated with enhanced acetylation and phosphorylation levels of eNOS in PVAT¹⁴⁷. Further studies are needed to test the effects of these factors on the PVAT browning process and the readouts on BP regulation, especially under hypertension associated with obesity. Further studies are needed to elucidate the underlying mechanisms for these observations and develop targeted therapeutics.

PERSPECTIVES AND CONCLUSIONS

There is no doubt that endocrine roles of adipose tissues, through adipokines, chemo/cytokines, hormones and other yet unknown factors, contribute systemically to many aspects of the physiology of the cardiovascular system. The dysfunction of adipose tissues, such as obesity, is one of the major risk factors for CVD. In that regard, PVAT is unique because it is adjacent and intimately integrated within the blood vessel wall and the significance of PVAT in development and prevention of CVD should not be ignored. PVAT is involved in all aspects of vascular physiology and pathophysiology. As a component of the vasculature, development and existence of PVAT is critical to maintenance of the vasculature in a normal functional status, partially through paracrine factors. Dysfunctional PVAT, such as obese PVAT or aging PVAT, secretes disease-promoting factors which cause abnormal changes towards various pathologies in the underlying layers of the blood vessels, the so-called “outside-to-inside” paradigm of PVAT effects on vascular pathologies. To comprehensively understand the vascular physiology and pathophysiology, in addition to understanding PVAT itself, more extensive and deeper research on the paracrine signaling from PVAT and its cross-talk with the underlying vascular cells, including EC, VSMC and fibroblasts under physiological and pathological conditions, is urgently needed. Adipocytes in PVAT are beige adipocytes which show mixed features of both brown and white adipocytes. Brown adipocytes are believed to promote energy expenditure and promote beneficial effects on the vascular system. There is evidence showing that PVAT is able to transform towards a fat tissue with increased white features during development of obesity and aging. Thus, reversing the white features of PVAT to brown features or maintaining PVAT beige features might be a crucial strategy to maintain a healthy vasculature. Additionally, PVAT secretes both PVCF and PVRF, especially in response to the same stimuli, suggesting that PVCF and PVRF precisely and concurrently control the blood vessel tone together with the systemic neurohumoral fluid system. The balance of PVCF and PVRF in the regulation of BP dipper pattern should be given more attention in order to develop novel strategies to maintain the physiological dipper pattern in order to prevent exacerbated adverse outcomes and CVD development. Finally, even though this review focuses on the crosstalk between PVAT and VSMC and its roles in vascular remodeling and BP regulation, the paracrine secretomes of PVAT and their roles in development of CVD should be systematically and extensively studied.

Highlights:

• PVAT is a unique adipose tissue and is involved in all the aspects of vascular physiology and pathophysiology.

• Existence of PVAT is critical to maintenance of the vasculature in a normal functional status, partially through paracrine factors. Dysfunctional PVAT secretes disease-promoting factors which cause abnormal changes towards various pathologies in the underlying layers of the blood vessels.

• Reversing the white features of PVAT to brown characteristics or maintaining PVAT beige features might be a crucial strategy to maintain a healthy vasculature.

• PVAT secretes both PVCF and PVRF, which precisely and concurrently control the blood vessel tone together with the systemic neurohumoral fluid system.

• Extensive and deeper research on the paracrine signaling from PVAT and its cross talk with the underlying vascular cells, including EC, VSMC and fibroblasts under physiological and pathological conditions, is urgently needed.

NONSTANDARD ABBREVIATIONS AND ACRONYMS:

PVAT

(perivascular adipose tissue)

WAT

(white adipose tissue)

BeAT

(beige adipose tissue)

BAT

(brown adipose tissue)

CVD

(cardiovascular diseases)

EC

(endothelial cells)

VSMC

(vascular smooth muscle cells)

BP

(blood pressure)

UCP-1

(uncoupling protein-1)

PPARγ

(peroxisome proliferator-activated receptor γ)

PRDM16

(PR domain containing 16)

SM22α

(smooth muscle protein 22 alpha)

NCC

(neural crest cells)

MSCs

(mesenchymal stem cells)

EV

(extracellular vesicles)

TGFβ

(transforming growth factor beta)

Ang II

(angiotensin II)

PDGF-BB

(platelet-derived growth factor-BB)

p38 MAPK

(p38 mitogen-activated protein kinases)

CTRP 9

(C1q/TNF-related protein 9)

ERK

(extracellular signal–regulated kinase)

PVRF

(PVAT-derived relaxing factors)

PVCF

(PVAT-derived contracting factors)

NO

(nitric oxide)

NOS

(nitric oxide synthase)

BKCa

(large-conductance Ca²⁺-activated K⁺)

EFS

(electrical field stimulation)

SOD

(superoxide dismutase)

CBS

(cystathionine β-synthase)

CSE

(cystathionine γ-lyase)

H₂S

(hydrogen sulfide)

HFD

(high-fat diet)