Endothelial CNP-GC-B Signaling in Pulmonary Arterial Hypertension

Key Takeaways

• Endothelial CNP or endothelial GC-B loss was associated with more severe pulmonary hypertension features across mouse models, while smooth muscle GC-B loss did not show the same worsening pattern.
• CNP-53 increased cGMP signaling and was associated with lower hypoxia-linked inflammatory signaling and a shift away from SMAD2/3-dominant endothelial responses.
• Idiopathic PAH endothelial cells showed lower NPPC and NPR2 expression, and exogenous CNP-53 improved rodent readouts with additive signals alongside macitentan or sotatercept.

Investigators in a Nature Communications study found reduced endothelial CNP/GC-B signaling across experimental pulmonary hypertension models and endothelial data from human idiopathic PAH. In MCTp-treated mice, pulmonary Nppc expression fell to about 65% of untreated levels. Mice lacking endothelial CNP or endothelial GC-B then developed more severe vascular remodeling and greater right-heart load under disease stress. Overall, reduced endothelial CNP signaling aligned with a more adverse preclinical pulmonary vascular phenotype.

The study combined monocrotaline pyrrole-induced and hypoxia-induced mouse PH models with SU5416-hypoxia rat and mouse PAH models. It also included cultured human pulmonary arterial endothelial cells, comparator smooth muscle experiments, and public lung scRNA-seq data from idiopathic PAH and controls. Genetic comparisons focused on endothelial CNP conditional knockout, endothelial GC-B conditional knockout, and smooth muscle cell-specific GC-B knockout against matched controls. Core in vivo readouts were right ventricular systolic pressure, RV/(LV + IVS), and pulmonary arterial medial wall thickness. Across these systems, the most consistent worsening followed endothelial disruption of the pathway rather than smooth muscle GC-B loss.

In diseased mouse lungs, pulmonary Nppc and Npr2 expression declined in both the MCTp and hypoxia models. Lowering endothelial CNP/GC-B expression alone did not produce PH in otherwise untreated mice, but it amplified disease severity after MCTp or hypoxia exposure. Endothelial CNP knockout and endothelial GC-B knockout each aggravated RVSP, RV/(LV + IVS), and pulmonary arterial medial wall thickness versus control animals. Smooth muscle cell-specific GC-B knockout did not reproduce the same worsening pattern in either mouse model. Across models, the adverse phenotype tracked with endothelial pathway loss rather than smooth muscle GC-B deletion.

CNP-53 increased urinary cGMP in vivo, supporting pathway activation during the animal experiments. In hypoxic human pulmonary arterial endothelial cells, CNP-53 suppressed EDN1, IL6, CCL2, and TGFB1, and those effects depended on GC-B. The peptide also reduced hypoxia-induced phospho-SMAD2/3 and restored phospho-SMAD1/5/9 balance through PKG, while not altering the hypoxia-induced decrease in BMPR2 expression. Together, these findings linked endothelial CNP signaling to less inflammatory and profibrotic pathway skewing.

Exogenous CNP-53 improved hemodynamic and remodeling endpoints in MCTp, hypoxia, and SU5416-hypoxia rodent PAH models. Its effects were abolished in endothelial GC-B knockout mice, while preserved activity in smooth muscle GC-B knockout mice supported endothelial dependence. In the SuHx mouse model, CNP-53 showed additive benefit with macitentan or sotatercept, particularly for RVSP, while additional RV/(LV + IVS) reduction versus monotherapy was numerical and not statistically significant. Public lung scRNA-seq data showed lower endothelial NPPC and NPR2 expression in idiopathic PAH than in controls, with p=7.9 × 10^-4 and p=0.043. CNP-53 did not affect systemic blood pressure in treatment experiments, and the human expression pattern aligned with a protective preclinical endothelial CNP-GC-B axis.