Small chemical molecule CPP promotes angiogenesis in surgically created severe lower limb ischemia and diabetes-induced limb vascular reduction models

Patients with peripheral artery disease (PAD) commonly experience chronic limb-threatening ischemia (CLTI) in the end stage, leading to severe functional impairment of the limbs, amputation, and even death [1]. Among the various disease factors contributing to PAD, diabetes is significantly associated with PAD risk, leading to more severe symptoms and a poorer prognosis. Promoting angiogenesis at ischemic sites and improving blood flow are key to the recovery of limbs from ischemic injury. Angiogenic therapies based on cells and small-molecule drugs are considered important and promising strategies [2]. Cell-based therapies can be classified based on cell source into allogeneic therapies and autologous therapies. Autologous cells offer the advantage of not requiring immunosuppression but may be limited by factors such as cell unavailability and scarcity due to autologous factors. Allogeneic cells, while offering a more abundant cell source, present drawbacks, including immune incompatibility and ethical concerns, which restrict their application. Small chemical molecules offer advantages such as convenience, low cost, ease of synthesis and storage, and rapid action without the need for introducing exogenous genes. They circumvent the high costs associated with cell therapies and potential side effects from allogeneic cell transplantation, holding significant promise and potential in both biomedical research and clinical applications.

Fibroblasts are a cell type with functional and transcriptional heterogeneity and cellular fate plasticity, making them a promising source of seed cells for clinical tissue and organ repair and regeneration [3]. Small molecules have become recognized as crucial tools for the induction of cellular reprogramming. In our laboratory’s previous research, a novel small chemical molecule probe, CPP ((E)-4-(4-(4-(7-(diethylamino)-2-oxo-2H-chromene-3-carbonyl)piperazin-1-yl)styryl)-1-methylpyridin-1-ium iodide), was identified as an inhibitor of prolyl-4-hydroxylase 2 (PHD2) [ 4] . CPP induces the differentiation of cultured human dermal fibroblasts into vascular endothelial cells (VECs) via the PHD2/hypoxia-inducible factor-1α/hairy-related transcription factor 1 (PHD2/HIF1α/HEY1) signaling pathway [ 4, 5] . The differentiated VECs exhibited therapeutic efficacy in treating lower limb ischemia in mice, indicating that CPP holds promise as a therapeutic candidate for limb ischemia [ 5] . In this study, we aimed to investigate the capability of CPP to directly induce angiogenesis in vivo under conditions of vascular injury.

Critical limb ischemia (CLI) occurs at the end stage of PAD. To evaluate the in situ therapeutic effects of CPP on CLI, a CLI model was established using C57BL/6 mice (Hubei Research Center of Laboratory Animals, Wuhan, China). CPP was synthesized and provided by the laboratory of Professor Baoxiang Zhao at Shandong University (Jinan, China). Following surgery, continuous subcutaneous multipoint injections of CPP were administered for 14 d to mimic localized drug treatment ( Figure 1A). Laser speckle blood flow imaging was used to assess lower limb perfusion on days 0, 7, and 14 ( Figure 1B, upper panel). The results showed that by day 7 post-surgery, both CPP concentrations significantly enhanced perfusion in the ischemic limb compared to the control group. From days 7 to 14, perfusion changes in both CPP-treated groups plateaued. Meanwhile, the control group showed increased perfusion. Throughout the treatment period, no significant differences were observed between the 1 and 10 mg/kg/day CPP treatment groups ( Figure 1C,D). On day 14, the capillary density in the skin and muscle was significantly higher in the 1 mg/kg/day CPP group than in the control group. Although the 10 mg/kg/day group had a slightly higher density than the control, the difference was not statistically significant ( Figure 1E–G). In addition, the organ toxicity of 10 mg/kg/day CPP was assessed in the CLI model. No treatment-related abnormal lesions or drug-induced toxic injuries were observed in the heart, liver, lungs, or kidneys compared with mice not treated with CPP ( Supplementary Figure S1). Our findings demonstrate that high-dose CPP retains a favorable safety profile, while low-dose CPP injections rapidly restore blood flow and achieve favorable therapeutic outcomes in the treatment of severe hindlimb ischemia, promoting in situ angiogenesis and perfusion recovery.

Figure 1 .

CPP enhances blood perfusion and angiogenesis in vivo

(A) Chemical structural formula of CPP. (B) Schematic diagram of subcutaneous multi-point injection of vehicle or CPP at the ischemic site and laser speckle blood flow detection in mice with CLI (upper panel). Schematic diagram of intraperitoneal injection of vehicle or CPP to db/m and db/db mice (lower panel). (C) Representative images of lower limb perfusion in CLI mice treated with vehicle or CPP, captured via laser speckle flow imaging at days 0, 7, and 14 (n = 3). (D) Ischemic limb perfusion on days 0, 7, and 14. Perfusion was quantified as the ratio of ischemic to non-ischemic limbs. (E) Representative images of CD31 (1:8000, 28083-1-AP; Proteintech, Wuhan, China) immunohistochemical staining in skin and gastrocnemius muscle from mice with CLI (n = 3). (F) Quantification of skin capillary density in CLI mice. (G) Quantification of gastrocnemius muscle capillary density in CLI mice. (H) Representative images of CD31 immunohistochemical staining in skin and gastrocnemius muscle from db/m and db/db mice (n = 7). (I) Quantification of skin capillary density in db/m and db/db mice. (J) Quantification of gastrocnemius muscle capillary density in db/m and db/db mice. CLI, Critical limb ischemia. *P < 0.05, **P < 0.01, Scale bars: 50 μm (E, H upper panel), 20 μm (E, H lower panel). Data are presented as the mean ± SD.

Given the higher prevalence of CLTI in PAD and considering that CLI often develops from CLTI, early adjuvant intervention rather than end-stage salvage therapy is more likely to improve patient quality of life. The progression of diabetes leads to the dysfunction of VECs and CTLI, and 14-week-old db/db mice (Jiangsu Huachuang Sino PharmaTech, Taizhou, China) with type 2 diabetes were used to investigate whether CPP could induce angiogenesis under conditions of vascular impairment and metabolic dysregulation that are unfavorable for blood vessel formation. Mice were intraperitoneally injected with 1 or 5 mg/kg CPP for 30 days, whereas the remaining groups received vehicle injections only ( Figure 1B, lower panel). The results showed that db/db mice treated with vehicle had reduced capillary density in the skin and gastrocnemius muscle compared to db/m mice ( Figure 1H–J). This indicates that the prolonged effects of type 2 diabetes caused damage and a decrease in capillary numbers, which impaired limb perfusion in these mice. Compared to the db/db group treated with vehicle, treatment with both CPP concentrations effectively promoted an increase in capillary numbers in the mouse skin and gastrocnemius muscle. However, the 5 mg/kg/day CPP dose did not have a stronger inducing effect than the 1 mg/kg/day dose ( Figure 1H–J). CPP effectively induced angiogenesis in the skin and gastrocnemius muscles of diabetic mice under conditions of vascular impairment and an unfavorable microenvironment for angiogenesis, without requiring prior in vitro induction. This approach saves time and reduces cell culture costs associated with in vitro differentiation.

Fibroblasts are heterogeneous cell types that are typically classified into distinct subpopulations. To provide research clues for the possible in vivo induction of fibroblast subclusters into endothelial cell subcluster differentiation pathways by CPP and to further investigate the mechanisms of angiogenesis promotion, single-cell RNA sequencing (scRNA-seq) was performed on skin tissue, and single-nucleus RNA sequencing was conducted on the gastrocnemius muscle, both of which were obtained from db/db mice treated with either vehicle or 5 mg/kg/day CPP for 30 days ( Figure 1B, lower panel). Through dimensionality reduction clustering and marker gene analysis, fibroblasts and endothelial cells were identified in both the skin and the gastrocnemius muscle. In the skin tissue, fibroblasts were classified into eight subclusters, and endothelial cells were classified into four subclusters ( Supplementary Figure S2A). In the gastrocnemius muscle, fibroblasts were classified into nine subclusters, and endothelial cells were classified into six subclusters ( Supplementary Figure S2B). Monocle2 ( https://bioconductor.org/packages/monocle/) was used to simulate the differentiation pathways from fibroblasts to VECs in the vehicle and CPP groups. The results showed altered simulated differentiation pathways from fibroblasts to endothelial cells in the CPP-treated group compared to the vehicle group ( Figure 2A,B and Supplementary Figure S2C–F).

Figure 2 .

CPP changes the differentiation pathway of fibroblast subclusters into endothelial subclusters in skin and gastrocnemius muscle

Tissue samples for single-cell RNA sequencing were obtained from the skin of db/db mice, and tissue samples for single-nucleus RNA sequencing were obtained from the gastrocnemius muscle. (A) In skin tissue, pseudotime analysis of fibroblast-to-endothelial cell differentiation along subclusters in the vehicle group (left panel) and the CPP group (right panel). (B) In the gastrocnemius muscle, pseudotime analysis of fibroblast-to-endothelial differentiation along subclusters in the vehicle group (left panel) and the CPP group (right panel). (C,D) UMAP plot showing the state distribution (C) and pseudotime trajectory analysis (D) of fibroblast subclusters differentiating into endothelial subclusters in the skin of the CPP group. (E,F) State distribution (E) and pseudotime trajectory analysis (F) of fibroblast subclusters differentiating into endothelial cell subclusters in the gastrocnemius muscle of the CPP group, visualized using a UMAP plot. (G) Heatmap showing four modules into which differentially expressed genes with pseudotime changes in skin were clustered based on similar expression patterns. GO enrichment analysis of BP of genes in Modules. (H) Heatmap of pseudotime-altered differentially expressed genes in gastrocnemius muscle clustered into seven modules based on similar expression patterns, accompanied by BP GO enrichment analysis. GO, gene ontology; BP, biological process.

Further pseudotime analysis of the CPP group within skin tissue was performed using the terminal point of the state 2 branch as the starting point for differentiation ( Figure 2C,D). Pseudotime analysis of the CPP group within the gastrocnemius muscle was performed using the terminal point of the state 3 branch as the starting point for differentiation ( Figure 2E,F). Enrichment analysis of genes involved in the differentiation pathway showed changes in their expression during fibroblast-to-endothelial cell differentiation. Based on the expression patterns, these differentially expressed genes were clustered into several modules, exhibiting dynamic expression changes along the differentiation trajectory. Dynamic gene expression changes related to angiogenesis, cell migration, cell adhesion, actin cytoskeleton organization, ECM organization, and collagen fibril organization were detected in skin and gastrocnemius muscle differentiation pathways ( Figure 2G,H). All of these factors were highly correlated with dynamic angiogenesis. Actin remodeling underpins the transition between endothelial quiescence and angiogenic states, with cell proliferation, migration, matrix adhesion, and remodeling being critically dependent on the actin cytoskeleton [6]. Most mechanical changes during angiogenesis occur within the ECM and require diverse chemical and mechanical cues to be modulated or modulated by ECM stiffness at different angiogenic stages [7].

In skin tissue, consistent with our laboratory’s prior in vitro findings, HIF1α and HEY1 were identified in Module 1 as being upregulated along the differentiation pathway ( Supplementary Figure S2G,H). Module 1 and module 2 showed alterations in genes associated with cell differentiation. The canonical WNT signaling pathway, which was enriched in module 2, is critical for embryonic development, cell proliferation, differentiation, and migration ( Figure 2G). In the gastrocnemius muscle, genes related to glycogen metabolic processes and carbohydrate phosphorylation were upregulated at the end of the differentiation trajectory within Module 4 ( Figure 2H). Among these, hexokinase 2 ( HK2) and 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 ( PFKFB3) were key genes involved in glycolysis ( Supplementary Figure S2I,J). Glycolysis is the primary pathway for ATP generation in endothelial cells. Given the substantial energy demand during angiogenesis, PFKFB3 and HK2 are recognized as crucial genes involved in this process [8]. The substantial energy demands required for both the possible trans‑differentiation process and the subsequent angiogenesis of newly formed endothelial cells are likely fulfilled precisely by the CPP‑induced enhancement of glycolytic flux. Consistent with this view, as a glycolysis regulator, PFKFB3 deficiency inhibits endothelial glycolysis, angiogenesis, and filopodium formation, all of which are essential for endothelial cell motility [9]. FGF signaling regulates vascular and lymphatic development by controlling endothelial metabolism through MYC-dependent regulation of HK2 expression [10]. In Module 5, genes associated with blood vessel remodeling were upregulated ( Figure 2H). Gene enrichment analysis for the gastrocnemius muscle showed distinct patterns compared to skin tissue. Combining these results with the immunohistochemical findings indicating increased vascular density, we hypothesized that fibroblasts in the gastrocnemius muscle may be at a later differentiation stage than those in the skin. Future experiments should shorten the CPP treatment duration to further examine gene expression changes.

In both the CLI model and the db/db mouse model, higher doses of CPP did not demonstrate superior therapeutic efficacy compared to the lower dose of 1 mg/kg/day in key functional indicators, including the blood perfusion ratio and capillary density. We propose that 1 mg/kg/day CPP is sufficient to fully occupy or activate its target, achieving a maximal pharmacodynamic response. Consequently, increasing the dose does not lead to enhanced therapeutic effects. These findings indicate that CPP is a highly efficient small molecule capable of maximally activating pro-angiogenic signaling at relatively low concentrations. Together with the organ toxicity studies, this implies that CPP may possess an improved safety profile and a wider therapeutic window. Any pro-angiogenic factor carries the theoretical risk of stimulating blood vessel growth in non-target tissues, and CPP is no exception. To minimize the potential impact of CPP on non-treated areas and increase its effective concentration at the ischemic site, local injection at a low concentration is recommended. Our studies involving localized CPP injection have also demonstrated its efficacy in restoring blood flow. The effects of CPP-induced angiogenesis on various tissues and organs warrant further investigation. Compared to 14 d of CPP administration, prolonged continuous CPP treatment helped maintain the newly formed blood vessels. However, critical indicators for evaluating the therapeutic effectiveness of CPP in treating CLI—such as the maturity of the neovessels, their functional capacity for blood transport, and the stability of these newly formed vessels after CPP withdrawal—remain important subjects for future research. Another limitation of this study is the exclusive use of male db/db mice. This choice was made to control key variables, such as glycemic variability, during the initial proof-of-concept stage. Future studies are warranted to systematically evaluate the efficacy and safety of CPP in female diabetic models to determine the generalizability of the therapeutic effects.

Based on scRNA-seq analysis, we hypothesized that CPP promotes therapeutic angiogenesis through a multi-layered transcriptional program. This program represents a coordinated response that functionally links intracellular mechanisms to tissue-level repair. Genetic changes in pathways related to cellular differentiation, such as the WNT signaling pathway, may drive fibroblast differentiation, offering clues for further in vivo studies on differentiation. Changes in actin cytoskeleton organization seem to provide the necessary physical forces for the angiogenic process. The regulation of genes involved in ECM organization remodels the pericellular microenvironment by loosening the matrix to facilitate cell invasion and lumen formation and by depositing and aligning new matrix components to provide mechanical guidance and stability for nascent vascular networks. The upregulation of glycolysis-related genes supplies the essential bioenergetic foundation, fueling both the differentiation process of endothelial cells and their subsequent migratory activities. Upregulation of blood vessel remodeling genes likely contributes to the stabilization and functional maturation of newly formed capillaries. Collectively, scRNA-seq of the skin and gastrocnemius muscle further modeled the potential CPP-induced differentiation trajectory of fibroblasts toward endothelial cells, indicating that CPP may promote angiogenesis by orchestrating dynamic changes in genes that regulate cell fate, cytoskeletal dynamics, matrix interactions, and metabolic enhancement, ultimately leading to a significant increase in capillary density. This hypothesis awaits further experimental validation.

In summary, we discovered that the small chemical molecule CPP induced angiogenesis in situ and enhanced blood perfusion in ischemic limbs, providing a new approach for restoring vascular function in the treatment of ischemic vascular diseases. scRNA-seq of the skin and gastrocnemius muscle supported a hypothetical model providing a potential multi-layered regulatory network during angiogenesis, offering multiple clues for further investigation of the mechanisms underlying CPP-mediated angiogenesis.

Supplementary Data

Supplementary data are available at Acta Biochimica et Biophysica Sinica online.

COMPETING INTERESTS

The authors declare that they have no conflict of interest.

Funding Statement

This work was supported by the grants from the National Natural Science Foundation of China (No. 32170744) and the Natural Science Foundation of Shandong Province (No. ZR2019ZD36).