Is endothelial dysfunction a therapeutic target for Peripheral Artery Disease?

Peripheral artery disease (PAD) is a significant global medical problem, affecting ~200 million people, including >8 million Americans, especially in the aging population ¹. PAD causes claudication, ischemic pain and ulcerations, and critical limb ischemia resulting in limb loss ^{2, 3}. It is also associated with an increased risk of heart attack and mortality rates 2-3 fold over the general population ⁴. Despite the alarming statistics, there is a lack of effective non-surgical medical therapies to prevent and treat PAD patients, especially those who are not eligible for invasive revascularization options. Therefore, there is a dire need to develop therapeutics to improve blood circulation to the affected limb before it leads to leg amputation or heart attacks. A significant factor in dictating the fate of clinical outcomes in PAD is the extent to which preexisting collateral arteries remodel, also known as adaptive arteriogenesis, to restore arterial flow in the ischemic limb ⁵. Unfortunately, the molecular mechanisms underlying the adaptive arteriogenesis process are still unclear. In this issue of Circulation Research, Craps et al. report Prdm16 (PR domain containing 16) as a transcription factor specific in the arterial wall, playing an essential role in adaptive arteriogenesis by maintaining the endothelial vasomotor function using a mouse model of PAD ⁶.

Prdm16 is a transcription factor and member of the PR-domain gene family. It is ubiquitously expressed in endothelial cells (ECs) and smooth muscle cells (SMCs) of arteries but not veins, independent of their anatomic location, vessel caliber, artery type, or health status⁶. It plays a role in several cellular processes, including proliferation and differentiation of smooth muscle cells and fate determination of brown adipocytes ^7–9. However, its role in PAD was not known. The authors previously identified PRDM16 as one of 8 transcription factors expressed explicitly in fresh artery endothelial cells ^7–9. In that study, they carried out a genome-wide transcriptomic study using freshly obtained human umbilical artery (HUAEC) and venous endothelial cells (HUVEC) and found eight transcription factors (PRDM16, SOX17, HEY2, MSX1, EMX2, NKX2-3, PRDM16, and TOX2) that are expressed explicitly in arterial ECs ¹⁰. They previously showed that MSX1 expression is induced in the artery endothelial cells during the collateral remodeling in a hindlimb ischemia model using the mouse femoral artery ligation ¹¹. Further mechanistic studies showed that the altered flow condition in ischemic hindlimb increases MSX1 expression, which triggered expression of vascular adhesion molecule-1 (VCAM1) and intercellular adhesion molecule-1(ICAM1), immune cell recruitment, arteriogenic remodeling, and flow recovery in the hindlimb ¹¹. In contrast, the current study reveals a different mechanism by which Prdm16 expression regulates collateral formation by maintaining endothelial function in a flow-independent manner.

In this paper, the authors studied the role of Prdm16, an artery-specific transcription factor, in hindlimb flow recovery. To test the expression of Prdm16 under physiological and pathological conditions, the authors used the mouse hindlimb ischemia (HLI) model and determined the expression of Prdm16. They identified that Prdm16 expression in collateral arteries stayed unchanged for up to 3 days after HLI induction. However, there was a rapid decline in the Prdm16 expression in the ligated side vs. the contralateral unligated control. The Prdm16 expression slowly recovered from day 7 onwards. These results implicate Prdm16 in the pathophysiology of PAD in this mouse model.

To further understand the role of Prdm16 in the pathophysiology of PAD, the authors used a genetic knockdown approach. Homozygotic whole-body deletion of Prdm16 (Prdm16^−/−) causes perinatal respiratory failure and death, while the heterozygous Prdm16 (Prdm16^+/−) mice are viable and fertile. Although Prdm16^+/− mice showed a minor reduction in femoral artery diameter, the baseline vasomotor function and blood pressure were not different from littermate controls. However, Prdm16^+/− mice showed an impaired artery flow recovery upon HLI. While the littermate control (Prdm16^+/+) mice showed spontaneous arterial flow recovery to >80% at d21 following the HLI, this adaptive flow recovery was impaired from d7 onwards in Prdm16^+/− mice. Consistent with the impaired arterial flow recovery result, the Prdm16^+/− mice showed substantial tissue necrosis, fibrosis, and fibro-adipose tissue accumulation in their gastrocnemius muscles compared to the littermate controls.

Since Prdm16 is expressed both in ECs and SMCs in arteries, the authors tested its cell-type-specific effect by developing two additional mouse lines: (1) the tamoxifen-inducible Cdh5 Cre mice to delete Prdm16 in the endothelium (EC-Prdm16^−/−) and (2) the constitutive Sm22α-Cre to delete Prdm16 in the vascular smooth muscle cells (SMC-Prdm16^−/−). EC-Prdm16^−/− mice were treated with tamoxifen either at birth to induce a long-term deletion or in adults 14 days before the HLI study for a short-term deletion. The long-term deletion in EC-Prdm16^−/− mice significantly impaired arterial flow recovery and adaptive arteriogenic remodeling responses upon HLI, compared to the littermate controls. The short-term deletion in EC-Prdm16^−/− mice showed an impaired arterial flow recovery but at a much less potency and delayed effect than the long-term deletion. Interestingly, SMC-specific deletion of Prdm16 in SMC-Prdm16^−/− mice showed no impairment in artery flow recovery upon HLI. These results demonstrate that the EC-specific deletion of Prdm16, but not the SMC-specific deletion, leads to impaired artery flow recovery upon HLI.

The authors next studied the mechanism underlying the impaired artery flow recovery in EC-Prdm16^−/− mice compared to Prdm16^+/− and littermate control mice. In response to HLI, hindlimb arteries undergo adaptive structural remodeling mediated by immune cell recruitment and luminal and medial expansion ¹². Surprisingly, they found that the structural adaptive arterial remodeling was not affected in Prdm16^+/− and EC-Prdm16^−/− mice compared to the littermate controls. These findings suggest that Prdm16 in ECs regulates arterial flow recovery independent of the adaptive arteriogenic remodeling process.

They next tested if the endothelial vasomotor function is responsible for the impaired arterial flow recovery in EC-Prdm16^−/− mice. For this study, they measured vascular relaxation responses in the femoral arteries obtained from the ligated and non-ligated hindlimbs. They found that endothelial-knockout of Prdm16 induced a significant impairment in the endothelial-dependent relaxation response, but not the EC-independent relaxation response. Interestingly, the femoral arteries from the ligated and unligated Prdm16^+/− mice showed no evidence of impaired endothelial-dependent relaxation response compared to the littermate controls. The discrepancy between the Prdm16^+/− and EC-Prdm16^−/− may be due to a gene-dose effect, but the exact reason needs further studies. Collectively, these results indicate that endothelial Prdm16 deficiency causes the early onset of endothelial dysfunction in response to HLI challenge, although it shows no effect under basal conditions. One implication of these findings is that Prdm16 is required to maintain endothelial function and is crucial for arterial flow recovery in response to ischemia. However, a significant loss of endothelial Prdm16 under certain pathological conditions or due to genetic mutations could lead to endothelial dysfunction, resulting in a poor arterial recovery in PAD settings. At present, what is unknown is whether Prdm16 expression level is severely lost under any condition relevant to PAD patients.

To further identify the mechanisms by which endothelial dysfunction occurs in EC-Prdm16^−/− mice, the authors performed a single-cell (sc) RNA sequencing analysis using the femoral arteries from the ligated and unligated mice. The single-cell data showed 181 genes in EC-Prdm16^−/− mice and 217 genes in EC-Prdm16^+/+ mice that were differentially expressed in a ligation- and Prdm16-dependent manner. Additional gene ontology analyses revealed that the pathways related to Ca²⁺ homeostasis and regulation of eNOS, NO production, and redox status were the most affected ones, responsible for the endothelial dysfunction in EC-Prdm16^−/−mice. To further validate the single-cell results, they overexpressed Prdm16 in HUVECs and found that it increases Ca²⁺ storage. They performed additional studies to investigate further the involvement of Prdm16 in NO bioavailability focusing on the eNOS and NADPH oxidases (Nox) in mouse hindlimbs upon HLI. While eNOS expression did not differ in EC-Prdm16^−/− mice, Nox2 and Nox4 expression were increased. This suggests a possibility that reactive oxygen species produced from Nox2/4 could lead to eNOS uncoupling ^13–15. While it is highly plausible, these results need to be further validated both in animal and patient samples.

Together, the Prdm16^+/−, SMC-Prdm16^−/−, and EC-Prdm16^−/− mice models showed that Prdm16 in artery endothelial cells play a crucial role in maintaining endothelial function and arterial flow recovery. EC-specific loss of Prdm16 leads to endothelial dysfunction without altering the structural arterial remodeling responses in response to hindlimb ischemia, resulting in prevention of adequate arterial flow recovery in PAD condition. The single-cell RNA sequencing data suggests that the pathways involved in the regulation of eNOS production, NO bioavailability, reactive oxygen species, and Ca²⁺ homeostasis as potential mechanisms responsible for the endothelial dysfunction induced in EC-Prdm16^−/− mice.

This study generates many exciting ideas to understand the fundamental mechanisms and potential therapeutic targets for PAD. It is crucial to understand how Prdm16 regulates endothelial dysfunction. Although the single-cell RNA sequencing data is an excellent place to formulate hypotheses, much remains to be investigated and validated to establish a precise mechanism. The clinical relevance of the role of Prdm16 in PAD is currently completed unknown. Is the loss of Prdm16 a potential prognostic marker of poor outcome of PAD in humans? Are there any pathological conditions or risk factors of PAD that regulate Prdm16 expression? Clinical studies involving human PAD patients should be used to answer these questions. The novel role of Prmd16 illustrated in the current study can lead to novel therapeutic opportunities for PAD. Mechanistically, only inducing collateral growth or angiogenesis per se may be insufficient to recover from an ischemic insult. Therefore, combined therapeutic approaches aimed at improving capillary angiogenic network and collateral arterial network expansion could be complemented with preservation of endothelial vasomotor function by targeting Prmd16 could provide additional benefits to PAD patients where invasive revascularization options are not viable. Collectively, these insights will further facilitate the development of new therapeutics for PAD.