摘要
This editorial refers to ‘Regulation of Ptbp1-controlled alternative splicing of pyruvate kinase muscle by liver kinase B1 governs vascular smooth muscle cell plasticity in vivo’, by Z. Cai et al., https://doi.org/10.1093/cvr/cvae187. Aortic aneurysm (AA) is a prevalent degenerative disease of the aorta and the 19th leading cause of death globally.1 It is usually asymptomatic until rupture when it has a mortality rate of 80%. There are currently no effective therapies other than open surgery or endovascular repair, and challenges in detection and identification of small aneurysms mean treatment cannot be targeted to lesions prior to rupture or need for surgery. It is critical therefore to understand the trajectory of AA and dissection and to identify novel diagnostic biomarkers and therapeutic targets. Vascular smooth muscle cells (VSMCs) in the medial layer of healthy blood vessels are crucial to maintaining contractility and vascular tone. Under stress or stimulation by pathological factors, VSMCs can undergo phenotypic switching, enhancing proliferation and migration capacity and differentiation into various cell types, including osteoblast-like, macrophage-like, and chondrocyte-like cells. VSMC differentiation and heterogeneity together with extracellular matrix (ECM) remodelling are key processes in the development of AA; however, current models have rarely recapitulated the earliest events and time-course of disease.2 The VSMC-specific, inducible LKB1 knock-out (KO) mouse model presented in the study by Cai et al. closely recapitulates the progression of AA.3 Integrating data from physiological measurements, single-cell analysis, imaging, and genetics, the authors identify a new aspect of this multifaceted disease and provide evidence for liver kinase B1 (LKB1) as a master regulator of VSMC fate and maintenance of vascular homeostasis. LKB1 is known to phosphorylate AMP-activated protein kinase (AMPK) and AMPK-related kinases involved in regulating cell growth, metabolism, migration, and polarity.4 Recently, the importance of LKB1 in cardiovascular disease has been established with LKB1 absence shown to cause myocardial hypertrophy, cardiac dysfunction, and atherosclerosis.5 Interestingly, LKB1 expression was decreased in human abdominal AA (AAA) samples, and this was accompanied by an increase in osteo-chondrocyte and fibroblast markers. VSMC-specific deletion of LKB1 in mice was embryonically lethal using the Tagln-Cre promoter. However, when it was deleted exclusively in VSMCs of adult mice using the tamoxifen-inducible Mhy11-Cre, it caused death within 8 months with 95% of mice spontaneously developing aortic dilation, aortic and/or arterial aneurysms, and rupture. LKB1-KO also increased vascular stiffness, highlighting the critical role of LKB1 in maintaining vascular health. Single-cell RNA sequencing 1-month post–LKB1-KO induction demonstrated a phenotypic switch of VSMCs from a contractile state to an ‘early modulated’ state, associated with loss of contractility markers and increase in inflammatory markers. At later stages, 3.5- and 4.5-month post-induction, there was differentiation into fibroblast-like and chondrocyte-like phenotypes, with significant collagen and proteoglycan deposition, ECM remodelling, and degradation of elastic fibres. At 6 months, there was expression of osteoblast marker, osteopontin, and regions of calcification were detected. The phenotypic changes were triggered by a shift in the metabolic profile towards aerobic glycolysis at the early stage, as LKB1 deficiency decreased oxygen consumption and reduced mitochondrial respiration. Pyruvate kinase M2 (PKM2) was identified as the mediator of this metabolic change. LKB1-KO upregulated polypyrimidine tract binding protein (PTBP1) expression and nuclear localization, triggering alternative splicing to produce PKM2 leading to reduced pyruvate kinase activity through formation of an inactive dimer. This finding was substantiated by treatment of the mice with a PKM2 agonist which rescued VSMC phenotype and protected against aortic dilation. However, this rescue phenotype was not studied in depth, and further analysis is required to determine the extent of rescue by targeting only the PKM2 pathway. This is crucial as PTBP1 itself has been recognized as a master regulator of VSMC alternate splicing.6,7 The authors showed increased nuclear localization of PTBP1 in the LKB1-KO mice, and it is likely that other targets of PTBP1 are also affected and this possibility needs further investigation. This study also opens up avenues for further investigation as there are questions remaining about the role of LKB1 in VSMCs. There have been several mutations identified in LKB1, causing Peutz–Jeghers syndrome and tumorigenesis, and while LBK1 is ubiquitously expressed, cells and tissues may produce different isoforms, influencing conformational state, accessibility of the nuclear localization signal and post-translational modifications which together affect enzymatic activity.8 It would be interesting to investigate the regulation of LKB1 gene expression, subcellular localization, and activity in the healthy vessel wall, as what triggers the loss of LKB1 in AA in aneurysm-prone regions remains to be elucidated. While the transformation of VSMCs was elegantly demonstrated in this study, there was no discussion of cell health or cell numbers despite AA commonly being characterized by a loss of VSMCs in the vessel wall accompanied by increased ECM deposition.2 A recent paper showed LKB1-KO mesenchymal stem cells (MSCs) exhibited marked reduction in survival compared to wild type cells,9 highlighting LKB1 as essential for cell survival under stress. The complex interplay between AA and vascular calcification was also presented. Previous clinical data have shown that the more severe the calcification, the higher the possibility of AA expansion and rupture.10 Intriguingly, in human AAA samples, there was decreased LKB1 expression which occurred concurrently with upregulation of the osteoblast marker runt-related transcription factor (RUNX2) and downregulation of the contractile marker, α-actin raising the possibility of a role of LKB1 in calcification. Indeed, calcification was also observed in LKB1-KO mice 6 months post-tamoxifen induction but not earlier, suggesting chondrocyte-like and fibroblast-like cells may eventually transform into osteoblast-like cells. However, it should be noted that in mice, chondrogenic calcification is much more prevalent than in human lesions so mechanisms may differ. Despite this caveat, a recent study showed that LKB1 expression is decreased under calcifying conditions and that LKB1 can inhibit VSMC calcification by binding mobility group box 1 (HMGB1) under normal conditions to promote its degradation thus preventing HMGB1-induced osteogenic differentiation.11 Complicating this, in MSCs the LKB1/AMPK pathway was activated in response to metformin, an antidiabetic drug, and was associated with increased RUNX2 gene and protein expression and nuclear localisation,9 highlighting the complex role of LBK1 in different cell types and tissues. In summary, this study contributes to the growing literature that implicates VSMC phenotypic plasticity and ECM remodelling in progression and exacerbation of AA (Figure 1). It not only advances understanding of AA progression but uncovers the novel LKB1–PKM2 axis as a potentially target for therapeutic intervention, and this pathway should continue to be investigated. AA progression is driven by switching of contractile VSMCs to an ‘early modulated’ phenotype and differentiation into fibroblast-like, chondrocyte-like, and eventually osteoblast-like cells. This accompanies ECM remodelling, calcification, and vascular stiffening. Changes in VSMC phenotype are regulated by a decrease in LKB1 expression, resulting in translocation of PTBP1, alternative splicing of PKM2, and its subsequent dimerization and inactivation. Cell fate switching is then triggered by a decline in mitochondrial respiration and shift towards aerobic glycolysis. The authors acknowledge the BHF Programme Grant RG/F/21/110064. Figure was made with Biorender.com.