Endothelial γ-protocadherins inhibit KLF2 and KLF4 to promote atherosclerosis

Joshi, Divyesh; Coon, Brian G.; Chakraborty, Raja; Deng, Hanqiang; Yang, Ziyu; Babar, Muhammad Usman; Fernandez-Tussy, Pablo; Meredith, Emily; Attanasio, John; Joshi, Nikhil; Traylor, James G.; Orr, Anthony Wayne; Fernandez-Hernando, Carlos; Libreros, Stephania; Schwartz, Martin A.

doi:10.1038/s44161-024-00522-z

Download PDF

Article
Open access
Published: 04 September 2024

Endothelial γ-protocadherins inhibit KLF2 and KLF4 to promote atherosclerosis

Nature Cardiovascular Research volume 3, pages 1035–1048 (2024)Cite this article

2000 Accesses
1 Citations
18 Altmetric
Metrics details

Subjects

Abstract

Atherosclerotic cardiovascular disease (ASCVD) is the leading cause of mortality worldwide. Laminar shear stress from blood flow, sensed by vascular endothelial cells, protects from ASCVD by upregulating the transcription factors KLF2 and KLF4, which induces an anti-inflammatory program that promotes vascular resilience. Here we identify clustered γ-protocadherins as therapeutically targetable, potent KLF2 and KLF4 suppressors whose upregulation contributes to ASCVD. Mechanistic studies show that γ-protocadherin cleavage results in translocation of the conserved intracellular domain to the nucleus where it physically associates with and suppresses signaling by the Notch intracellular domain. γ-Protocadherins are elevated in human ASCVD endothelium; their genetic deletion or antibody blockade protects from ASCVD in mice without detectably compromising host defense against bacterial or viral infection. These results elucidate a fundamental mechanism of vascular inflammation and reveal a method to target the endothelium rather than the immune system as a protective strategy in ASCVD.

Laminar flow inhibits the Hippo/YAP pathway via autophagy and SIRT1-mediated deacetylation against atherosclerosis

Article Open access 21 February 2020

KLF4 recruits SWI/SNF to increase chromatin accessibility and reprogram the endothelial enhancer landscape under laminar shear stress

Article Open access 23 August 2022

Flow-induced reprogramming of endothelial cells in atherosclerosis

Article 24 May 2023

Main

Atherosclerotic cardiovascular disease (ASCVD), which is characterized by fatty plaques within arterial walls, results from converging metabolic, inflammatory and biomechanical factors, including hypertension, hyperlipidemia, smoking and age^1,2. Atherosclerotic plaques form preferentially at curved or branched regions of arteries experiencing low, multidirectional shear stress from blood flow, termed disturbed shear stress (DSS)^1,2. DSS is also the most reliable predictor of plaque erosion^3,4 and plaque vulnerability to rupture⁵. Conversely, straight regions with high unidirectional laminar shear stress (LSS) suppress plaque formation mainly by upregulating the Krüppel-like 2 (KLF2) and 4 (KLF4) transcription factors in endothelial cells (ECs). The two genes, KLF2 and KLF4, which are generally co-regulated in ECs and have partially redundant gene targets and functions, govern approximately 70% of LSS-induced anti-inflammatory and antithrombotic protective genes^{6,7,8,9,10,11,12}. Extensive studies in mice, using both EC-specific knockout (ECKO) and transgenic overexpression, identified KLF2 and KLF4 as potent mediators of resilience against a multiplicity of vascular conditions, including atherosclerosis, pulmonary hypertension and coronavirus disease 2019-mediated vascular dysfunction^12,13. Reduced endothelial KLF2 and KLF4 expression is similarly associated with worsened cardiovascular outcomes in humans. Therefore, restoring high KLF2 and KLF4 expression may protect against a range of inflammatory cardiovascular diseases (CVDs)¹⁴. However, the fundamental mechanisms regulating KLF2 and KLF4 expression are not fully understood.

Toward this goal, we carried out a genome-wide CRISPR knockout (KO) screen using a green fluorescent protein (GFP) reporter driven by the human KLF2 promoter, which identified approximately 300 genes required for shear stress induction of KLF2 (ref. ¹⁵). Systematic analysis of the genes whose CRISPR KO reduced KLF2 (activators) identified a contribution from mitochondrial metabolism that synergizes with the established mechanism, the MEKK2 and MEKK3-MEK5-ERK5 kinase cascade^15,16. Unexpectedly, this screen uncovered an additional 160 genes whose KO increased KLF2 (suppressors). These genes are of great interest both to gain mechanistic insight into KLF2 regulation and as candidate therapeutic targets in vascular inflammatory diseases.

We selected protocadherin γ subfamily A, 9 (PCDHGA9) for further study. This gene is a member of the 22-gene PCDHG subfamily among the clustered protocadherin (cPCDH) family of homophilic adhesion receptors. cPCDHs mediate a wide range of functions in the central nervous system, including adhesion, signaling and cell sorting^17,18. PCDHG has an essential role in neurons and its misexpression is associated with neuronal disorders^19,20; however, little is known about PCDHG in other cell types. In this study, we describe a role for the PCDHG cluster in vasculature inflammation and ASCVD.

Results

CRISPR screen identifies suppressors of KLF2

To identify the mechanisms underlying the mechanoregulation of KLF2, a genome-wide CRISPR screen was performed using a Klf2:GFP reporter expressed in mouse aortic ECs (MAECs) and stimulated with LSS for 18 h (Fig. 1a,b and Extended Data Fig. 1a). In addition to the approximately 300 genes whose CRISPR KO decreased KLF2 (ref. ¹⁵), this screen also identified approximately 160 genes whose KO increased the Klf2:GFP reporter levels (Extended Data Fig. 1b and Supplementary Table 1). Network analysis showed enrichment of pathways and processes that affect vascular and endothelial functions (Fig. 1c), consistent with a central role for KLF2 and KLF4 in ECs.

**Fig. 1: A genome-wide CRISPR screen identifies KLF2 suppressors.**

Reasoning that transmembrane proteins would be readily accessible for therapeutic intervention, we focused on cell surface candidates (Fig. 1d and Supplementary Table 2). Of these, the top eight candidates were functionally verified using independent CRISPR single-guide RNAs (sgRNAs) by assaying for Klf2:GFP induction on LSS and oscillatory shear stress (OSS), commonly used to model in vivo DSS (Fig. 1d, red box). The top candidate, PCDHGA9, was chosen for further study. Examining induction of the Klf2:GFP reporter in the MAEC cell line under both LSS and OSS showed that CRISPR KO of Pcdhga9 increased Klf2:GFP in cells under LSS and did so even more strongly under OSS, where expression is normally low (Fig. 1e). Conversely, overexpression of human PCDHGA9 suppressed reporter expression. Knockdown of Pcdhga9 using small interfering RNA (siRNA) similarly increased Klf2:GFP, which was reversed by overexpression of human protein (Extended Data Fig. 1c–e). Examining endogenous Klf2 in primary human umbilical vein ECs (HUVECs) replicated this behavior, demonstrating conservation across species and EC types (Extended Data Fig. 1f). In ECs under pro-inflammatory OSS, PCDHGA9 depletion also suppressed the induction of the leukocyte adhesion receptor VCAM1 (Fig. 1f) and the adhesion of THP1 monocytes (Fig. 1g). PCDHGA9 thus restrains LSS-dependent induction of anti-inflammatory KLF2 and potentiates OSS-dependent pro-inflammatory activation.

Protocadherin γ gene cluster suppresses KLF2 and KLF4

PCDHGA9 is a member of the clustered protocadherin family, which belongs to the cadherin superfamily of cell–cell adhesion receptors. The clustered protocadherin genes, classified into alpha (a or α), beta (b or β) and gamma (g or γ) clusters, are organized in tandem in the genome. The mouse Pcdhg gene cluster includes 22 genes (Fig. 2a), several of which are expressed in ECs (as shown in this study). Exon 1 is unique, while exons 2–4 are shared by all 22 members, allowing for targeting of the entire Pcdhg cluster by siRNA-mediated silencing (Fig. 2a, region enclosed between the dashed lines). Silencing the Pcdhg gene cluster in HUVECs (Fig. 2b and Extended Data Fig. 2a) strongly increased endogenous Klf2 and Klf4 mRNA levels and suppressed pro-inflammatory E-selectin (Sele) even under OSS (Fig. 2c,d and Extended Data Fig. 2b). Depletion of the Pcdhg gene cluster suppressed adhesion of THP1 monocytes to ECs under pro-inflammatory OSS to the same extent as Pcdhga9 depletion (Fig. 2e). Targeting Pcdhga9 or the entire cluster thus gives indistinguishable results.

**Fig. 2: *Pcdhg* gene cluster suppresses KLF2 and KLF4.**

In addition to LSS, KLF2 and KLF4 are also induced by statins²¹, the cholesterol-lowering drugs used as first-line therapy for patients at risk of ASCVD, an effect believed to contribute to their therapeutic benefits. To test for interactions, control or Pcdhg-depleted HUVECs were treated with lovastatin for 16 h. Pcdhg depletion greatly amplified statin induction of KLF4 (Fig. 2f), showing clear synergy. Conversely, pro-inflammatory tumor necrosis factor (TNF) suppressed KLF4 and induced pro-inflammatory genes, including the leukocyte adhesion receptor VCAM1. Pcdhg depletion alleviated KLF4 suppression by TNF, suppressed TNF-induced VCAM1 expression and subsequent adhesion of THP1 monocytes (Fig. 2g,h). Pcdhg depletion significantly reduced the induction of VCAM1 even at the highest doses of TNF tested (Extended Data Fig. 2c). Targeting Pcdhg thus suppressed EC inflammatory activation in multiple contexts.

Pcdhg endothelial KO protects against atherosclerosis

EC KLF2 and KLF4 are essential for vascular resilience against inflammatory CVDs, including atherosclerosis. Importantly, elevating KLF4 in ECs protects against ASCVD, demonstrating its sufficiency for atheroprotection²². To test if the effects of Pcdhg seen in culture are retained in vivo, Pcdhg^loxP/con3 (loxP-targeted Pcdhg constant exon 3, tagged with GFP) mice^23,24 were crossed with constitutive Cdh5-Cre mice to delete the entire Pcdhg cluster in ECs (ECKO) (Fig. 3a). Homozygous Pcdhg ECKO pups were obtained at the expected Mendelian frequency and were phenotypically normal, showing that Pcdhg expression in ECs, unlike in neurons, is not essential for development^25,26 (Fig. 3b and Extended Data Fig. 3a–c). These mice were thus further analyzed.

The aortic arch contains the athero-resistant greater curvature that is under LSS and the nearby athero-susceptible lesser curvature that is under DSS, offering a well-established model. Examination en face revealed modestly increased KLF4 expression in Pcdhg ECKO ECs in the greater curvature and a markedly larger increase in the lesser curvature compared to controls (Fig. 3c). We next induced hyperlipidemia by injecting adeno-associated virus 8 (AAV8)-PCSK9 and feeding them a high-fat diet (HFD) for 16 weeks (Fig. 3d)^27,28. Pcdhg ECKO mice showed reduced atherosclerotic plaques as observed using Oil Red O staining in aortas from both males and females (n = 6, each sex) (Fig. 3e,f). Examination of the aortic root using hematoxylin & eosin (H&E) and Oil Red O staining showed smaller atherosclerotic plaques with drastically smaller necrotic cores (NCs), thicker fibrous caps (FCs) (Fig. 3g) and greatly reduced macrophage and monocyte content in Pcdhg ECKO mice (Fig. 3h). No differences were observed in blood lipids (triglycerides, cholesterol, high-density lipoprotein cholesterol (HDL-C)) (Extended Data Fig. 3d–f) or body weight (Extended Data Fig. 3g).

Immune host–pathogen defense

Limiting EC inflammatory gene expression could conceivably suppress immune responses via effects on leukocyte trafficking or other mechanisms. To assess immune function, control and Pcdhg ECKO mice were infected with lymphocytic choriomeningitis virus (LCMV) clone 13, which triggers T cell activation. Seven days after infection, spleen cell readouts and function were measured using fluorescence-activated cell sorting (FACS). Control and Pcdhg ECKO mice displayed similar frequency of CD8a⁺ T cells (Fig. 3i) and similar functional capacity after restimulation with an LCMV-specific peptide (GP33) as observed by unaltered GP33 tetramer⁺, granzyme B⁺, interferon-γ (IFNγ)⁺TNF⁺ and IFNγ⁺TNF⁻ populations (Fig. 3j–m). The viral loads, measured as plaque-forming units (PFUs) from kidneys were similar in control and Pcdhg ECKO mice (Fig. 3n). Myeloid compartment (CD11b⁺) cells were similarly identical in frequency and activation status (Extended Data Fig. 4a–c). Thus, Pcdhg ECKO had no detectable effect on host defense against this virus.

We next examined responses to bacterial infection using a self-limited Escherichia coli serotype O6:K2:H1 peritonitis model that allows for assessment of the initiation, peak and resolution phases of inflammation^29,30,31. Pcdhg ECKO mice showed no deficit in bacterial defense; instead, to our surprise, they showed faster bacterial clearance compared to controls, as evidenced by reduced bacterial colony-forming units (CFUs) from peritoneal exudate 12 h after infection, in both males and females (Fig. 3o and Extended Data Fig. 4d,e); 12–24 h after infection marked the peak of inflammation in this study (Fig. 3o and Extended Data Fig. 4d). No CFUs were detected in the blood from either control or Pcdhg ECKO mice, demonstrating no systemic bacterial dissemination (Extended Data Fig. 4f). Bacteria trigger acute inflammation that mobilizes neutrophils from bone marrow (BM) reservoirs to the site of infection, which, owing to their short half-life (approximately 6 h), resolves within several hours. Accordingly, the peak of neutrophil recruitment and the percentage (of CD11b⁺F4/80⁻cells) in the peritoneal exudate were lower in Pcdhg ECKO mice compared to controls (Fig. 3p, Extended Data Fig. 4g,h and Supplementary Fig. 1). The total number of leukocytes followed a similar but nonsignificant trend (Extended Data Fig. 4i). Although Pcdhg ECKO mice showed fewer neutrophils accumulating, peritoneal neutrophils collected at 12 h showed increased per-cell phagocytic clearance of bacteria compared to controls (Fig. 3q and Extended Data Fig. 4j,k). BM neutrophils from uninfected mice showed no difference in phagocytic capacity, ruling out an inherent difference in neutrophil function (Extended Data Fig. 4l). Thus, Pcdhg ECKO mice showed somewhat slower neutrophil recruitment in response to infection, but faster clearance of bacteria associated with greater phagocytosis per neutrophil. The combined data from viral and bacterial challenges showed that Pcdhg ECKO does not detectably limit host immune responses.

Nuclear PCDHG intracellular domain suppresses KLF2 and KLF4

To investigate the molecular mechanism by which PCDGH suppresses KLF2 and KLF4, we performed a structure–function analysis using the alternative splicing of Pcdhg cluster members to define domain boundaries. Each PCDGH protein is organized into an extracellular domain (ECD) containing six cadherin domains involved in homophilic or heterophilic cis and trans interactions, a transmembrane (TM) domain governing its membrane localization and an intracellular domain (ICD) involved in intracellular trafficking, surface delivery, localization and signaling. The ICD is subdivided into a variable C-terminal domain (VCD) and a constant C-terminal domain (CCD). The conserved CCD is encoded by the three 3′ exons shared by all 22 members (Fig. 4a).

**Fig. 4: A physical and functional PCDHG–Notch axis regulates KLF2 and KLF4.**

Overexpression of Pcdhga9 suppressed KLF2 and KLF4 (Fig. 1e and Extended Data Fig. 1e). Therefore, we used this assay to identify functional domains. Full-length, ECD + TM, TM + ICD and CCD-lacking (ΔCCD) mutants (Fig. 4a,b) were expressed in HUVECs, which were treated with LSS and OSS and assayed for KLF4 protein. The mutants were expressed at comparable levels and localization was analyzed using GFP fluorescence (Fig. 4b and Extended Data Fig. 5a,b). While the full-length mutant localized to both cell–cell junctions and intracellular vesicles, ECD + TM and ΔCCD were more junctional, as expected (owing to reduced internalization in the absence of the CCD), while TM + ICD was a mixture of punctate and distributed evenly (Extended Data Fig. 5b). KLF4 immunoblotting revealed that the CCD is essential for KLF2 and KLF4 suppression, whereas the ECD is dispensable (Fig. 4b and Extended Data Fig. 5a). Further experiments thus focused on the CCD.

Unlike classical cadherins, PCDHG members are processed by proteolytic cleavage; this releases the ICD, which due to its nuclear localization signal (NLS) translocates to the nucleus³² (Fig. 4a). To test the role of cleavage and nuclear translocation, FLAG-tagged versions of the CCD and an NLS-deleted CCD mutant (ΔNLS-CCD) were expressed in HUVECs, stimulated with LSS and the KLF4 levels assayed (Fig. 4a). Mutants were expressed at comparable levels and localized as expected, with the full-length mutant present in the secretory system and plasma membrane, the CCD mainly in the nucleus and the ΔNLS-CCD in the cytoplasm (Extended Data Fig. 5c). The CCD suppressed KLF4 whereas the ΔNLS mutant was ineffective (Fig. 4c). Interestingly, the CCD is very highly conserved across species, supporting its critical role (Extended Data Fig. 5d). Thus, PCDHG signals via cleavage and nuclear translocation of its common cytoplasmic regions.

Pcdhg regulates Klf2 and Klf4 via the Notch pathway

To assess the genes and processes regulated by PCDHG in ECs, we performed bulk RNA sequencing (RNA-seq) analysis of control siRNA and Pcdhg siRNA HUVECs under OSS (where PCDHG had the greatest effect). We included Pcdhg + Klf2 and Klf4 triple siRNA to identify Klf2 and Klf4-independent effects (Fig. 4d,e). Upstream regulatory pathway analysis of these Pcdhg-dependent and Klf2 and Klf4-independent differentially expressed genes (DEGs) identified Notch as the most overrepresented pathway (Fig. 4f), confirmed by strong upregulation of known Notch target genes (Fig. 4g). Notch1–4 family transmembrane receptors are critical for EC functions, including determination of arterial identity and vascular stability, with Notch1 being the main isoform. Binding of Notch ligands such as DLL4 triggers stepwise proteolysis of the receptors by members of the ADAM metalloprotease family followed by γ-secretase, releasing the Notch intracellular domain (NICD), which translocates to the nucleus and binds the transcription factor recombination signal binding protein for immunoglobulin kappa J region (RBPJ) to induce target genes³³. Interestingly, LSS both activates Notch and induces KLF2 and KLF4, both of which stabilize and protect vessels against inflammation and atherosclerosis^34,35,36,37, although the effects of Notch and KLF2 and KLF4 have not been linked. We found that Pcdhg depletion increased the levels of the cleaved, activated NICD (Val1744) (Fig. 4h). Notably, the promoters for both human KLF2 and mouse Klf2 and Klf4 contain previously unappreciated canonical Notch-RBPJ binding sites (Extended Data Fig. 5e,f). To test the role of Notch in KLF2 and KLF4 induction by LSS, RBPJ was blocked by the small-molecule inhibitor RBPJ Inhibitor-1 (RIN1) or by RBPJ siRNA, with or without Pcdhg depletion. ECs under LSS were then examined. The increase in KLF4 levels upon Pcdhg siRNA was completely prevented by inhibition of the Notch pathway (Fig. 4i). We conclude that Pcdhg regulates Klf4 via Notch.

To identify proteins that interact with the Pcdhg CCD, we did a proteomic analysis of immunoprecipitates (IPs) from full-length Pcdhga9 and the ΔCCD mutant, looking for binding partners that required the CCD. Notch1 peptides were detected in the IPs of full-length Pcdhga9, but not the ΔCCD mutant or vector alone. Sequence analysis showed that these Notch1 peptides came from the NICD region (Fig. 4j asterisks and Extended Data Fig. 5g), which is consistent with functional effects. The immunoprecipitation and immunoblot analyses showed that in addition to full-length Pcdhg, the CCD fragment but not ΔNLS-CCD physically associated with the NICD (Fig. 4k,l). Finally, a Notch transcriptional activity reporter showed that Pcdhg ICD (PICD) harboring the conserved CCD suppressed NICD transcriptional activity in a dose-dependent manner (Fig. 4m). Together, these results show that PICD physically associates with NICD to suppress Notch signaling and its downstream target Klf2 and Klf4.

PCDHGA9 blocking antibody in experimental atherosclerosis

PCDHG was chosen as a target in part because its cell surface localization makes it amenable to inhibition by antibodies or other cell-impermeant reagents. Toward this goal, we purified murine PCDHGA9 ECD protein, which was used both for generating monoclonal antibodies (mAbs) and for developing a cell adhesion assay (Fig. 5a and Extended Data Fig. 6a). Cells were plated in 96-well microplates coated with PCDHGA9 ECD adhered over time; adhesion to uncoated wells was minimal (Fig. 5b), demonstrating specificity.

Rat hybridomas against PCDHGA9 ECD (Extended Data Fig. 6a,b) were purified and tested in the homophilic adhesion assay³⁸. Antibodies A9, B1 and B4 strongly blocked adhesion (Fig. 5c and Extended Data Fig. 6c). Antibody specificity was verified by immunoblotting using purified ECD-GST or ECD alone (Extended Data Fig. 6d). Next, these function-blocking mAbs were tested for their effect on KLF2 and KLF4 upon LSS, and VCAM1 upon OSS. Blocking antibodies increased flow induction of the Klf2:GFP reporter and decreased VCAM1, normalized to the internal mCherry control (Fig. 5d,e and Extended Data Fig. 6e,f). mAb A9 showed the strongest effect on adhesion to ECD, KLF2 induction upon LSS, VCAM1 induction upon OSS and the highest signal-to-noise ratio in the immunofluorescence (IF) assay (Extended Data Fig. 6g). Hence, mAb A9 was studied further.

A9 was highly specific, as shown by immunoblotting of cell lysates from control versus Pcdhg or Pcdhga9 knockdown MAECs (Extended Data Fig. 6i), with purified ECD as a positive control. Therefore, we examined its efficacy in a model of experimental atherosclerosis in mice. Because A9 is a rat monoclonal antibody, we avoided immune recognition and clearance of the rat IgG by using the PCA ligation model of accelerated atherosclerosis in Apoe^−/− mice where lesions develop within 1 week (Fig. 5f)^28,39,40. Mice were maintained on an HFD for a total of 3 weeks (1 week before ligation and 2 weeks after ligation). The A9 half-life, determined by intraperitoneal injection of 2 μg of isotype control or mAb A9 per mouse and measuring blood plasma levels, was approximately 8 days (Extended Data Fig. 6i). Hence, Apoe^−/− mice on an HFD were subjected to surgery and A9 versus control IgG was injected once per week for 2 weeks as described in the Methods (Fig. 5g). The operated left carotid artery (LCA) was then compared to the unligated right carotid artery (RCA) as an internal control. Treatment with mAb A9 strongly reduced plaque in this acute model of ASCVD, as seen by whole-mount carotid preparations (Fig. 5h, brackets) and verified by staining carotid sections with Oil Red O to detect lipid-rich plaques (Fig. 5i). Reduction in lumen diameter was also prevented by A9 (Extended Data Fig. 6j). No difference was observed in blood lipids (triglycerides and cholesterol) (Extended Data Fig. 6k,l). Taken together, these results identify PCDHG as a new therapeutic target in atherosclerosis.

PCDHG expression in atherosclerosis

Lastly, we analyzed the levels of PCDHG in human arteries. Staining for PCDHG using an antibody against the CCD showed approximately three times higher expression of PCDHG in the endothelium (marked by eRG) from CVD donors compared to healthy, age-matched donors (Fig. 6a). Atherosclerotic mouse arteries also showed elevated PCDHG staining (Extended Data Fig. 7a). Coronary arteries from three asymptomatic older donors stained for PCDHG showed approximately a four times higher signal in the regions of plaque compared to regions of the same artery without evident plaque (Fig. 6b). Antibody specificity was verified with immunoblotting using control and PCDHG siRNA HUVECs (Fig. 2b) and staining of control versus PCDHG ECKO retinas (Extended Data Fig. 7b). PCDHG is thus upregulated in the endothelium from individuals with ASCVD.

**Fig. 6: *PCDHG* in human atherosclerosis.**

Discussion

ECs have a pivotal role in vascular physiology and pathology via functions ranging from regulating vessel diameter to immune responses to nutrient transport. Cell–cell adhesions are a locus of EC signaling and function, including solute transport, leukocyte trafficking, growth control and shear stress signaling. This study is based on a genome-wide CRISPR KO screen that identified approximately 160 genes that suppress KLF2. KLF2 and KLF4 are generally co-regulated in ECs and show a high degree of functional redundancy; indeed, they recognize the identical consensus sequence in gene promoters and enhancers⁴¹. The results obtained for the two were essentially identical in this study. We selected the strongest cell-surface-localized KLF2 suppressor, PCDHGA9, a member of the PCDHG family of adhesion receptors, for further study. Although extensively studied in neuronal systems^{17,19,20,23,25,26,42,43}, only two papers reported roles for another PCDHG family member (PCDHGC3) in the vasculature, with effects on EC tube formation, migration and permeability in vitro, suggestive of a role in angiogenesis^44,45. In this study, depletion of PCDHGA9 or the whole PCDHG cluster showed essentially identical results. We found that PCDHG is a pro-inflammatory gene that suppresses KLF2 and KLF4 through a pathway that involves the release and translocation of the intracellular domain of its protein product to the nucleus where it binds the Notch1 ICD to suppress transcription of Notch target genes. EC deletion of Pcdhg in mice increases KLF2 and KLF4 levels and protects against ASCVD without apparent effects on development or viability. Antibodies that block homophilic PCDHG adhesion also increase KLF2 and KLF4 and limit experimental atherosclerosis. Within the vasculature, Pcdhg is expressed primarily in ECs, with marked increases in atherosclerotic regions and low expression in other cell types.

Analysis of Pcdhg-dependent genes identified the Notch pathway as a major downstream target. Notch1 limits angiogenesis in development and promotes arterial identity and decreases atherosclerosis in adults; thus, it was further investigated. LSS activation of Notch and induction of KLF2 and KLF4 are well established, both promoting vascular stability^34,35,36,37, but an interconnection, to our knowledge, has not been reported. Mechanistic data in this study showed Notch as a direct inducer of KLF2 and KLF4 via consensus sites in their promoters and enhancers, which is suppressed by PCDHG, and which correlates with physical association of the PICD and NICD.

While this study focused on atherosclerosis, vascular inflammation driven by EC activation is a major contributing factor in many diseases, including pulmonary arterial hypertension, venous thrombosis and autoimmune diseases, to name a few^17,18. The CANTOS clinical trial identified the benefits of immune suppression, but these were balanced by increased deaths from infections⁴⁶. We suggest that targeting endothelial inflammatory activation rather than immune mediators offers a path to limiting vascular inflammation without compromising host–pathogen defense. Further studies to fully elucidate the mechanisms of PCDHG expression, activation and contributions to inflammatory disease are thus important directions for future work. Additionally, the remaining 159 hits whose CRISPR KO increased KLF2 levels offer a rich source of new therapeutic targets.

Methods

Primary cells, cell lines and cell culture reagents

The PyMT-immortalized MAECs that express the Klf2 promoter reporter (Klf2:GFP MAEC) were described previously¹⁵. MAECs were maintained in complete EC medium (cat. no. M1166, Cell Biologics)⁴⁷. HUVECs obtained from Yale Vascular Biology and Therapeutics Core were pooled from three donors. They were screened for the absence of pathogens, maintained in EGM2 endothelial cell growth medium (cat. no. CC-3162, Lonza) and used at passages 2–5. All cells were routinely screened for Mycoplasma. siRNA transfection was performed with Lipofectamine RNAiMAX (cat. no. 13778150, Thermo Fisher Scientific) in Opti-MEM medium (cat. no. 31985070, Thermo Fisher Scientific) using ON-TARGETplus SMARTpool siRNAs from Horizon Discovery. For NICD transcriptional activity reporting, 12×CSL/RBPJ-d1EGFP (cat. no. 47684, Addgene) was used. For lentiviral transduction, HEK 293T cells were transfected with lentiviral vectors along with pVSV-G (cat. no. 138479, Addgene) and psPAX2 (cat. no. 12260, Addgene) packaging plasmids using Lipofectamine 2000 (cat. no. 11668019, Thermo Fisher Scientific) according to the manufacturer’s instructions. Supernatants were collected 48–96 h after transfection and filtered through a 0.45-μm low-protein binding filter. Primary HUVECs were infected with lentivirus for 24 h, then the medium was replaced with EGM2 medium. For the monocyte adhesion assays, THP1 labeled with CellTracker Deep Red (cat. no. C34565, Thermo Fisher Scientific) were resuspended in Hanks’ Balanced Salt Solution (HBSS) supplemented with 1 mM Ca²⁺, 0.5 mM Mg²⁺ and 0.5% BSA, added to the slides with HUVECs, incubated for 20 min at 37 °C, washed three times in HBSS and fixed with 3.7% formaldehyde. Cells were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) before fluorescence imaging. Drugs used were lovastatin (cat. no. 438185, Sigma-Aldrich), TNF (cat. no. 300-01A, PeproTech) and RIN1 (cat. no. SS3376, Selleckchem). All antibodies were validated in knockdown and KO depletion using IF or immunoblotting. siRNA and sgRNA sequences, antibodies and primers are listed in Supplementary Tables 3–5.

CRISPR library screen and KLF2 suppressor (gain-of-function) phenotype determination

CRISPR library screening, library preparation and next-generation sequencing (NGS) were described previously¹⁵. Briefly, immortalized Klf2:GFP reporter MAECs were infected with a genome-wide CRISPR library of approximately 160,000 CRISPR sgRNAs, treated with 15 dyn cm⁻² LSS for 18 h, FACS-sorted based on Klf2:GFP reporter levels, with subsequent analysis of sgRNAs in the high Klf2:GFP gate by NGS on an HiSeq 2500 system (Illumina). The KLF2 suppressor (gain-of-function) phenotype was determined as done previously for the loss-of function phenotype ranking genes based on the cumulative z-score from the three highest scoring unique sgRNAs listed in Supplementary Table 1 (ref. ¹⁵).

Shear stress stimulation

All shear stress experiments, unless otherwise indicated, were performed in parallel plate flow chambers perfused within a pump and environmental control system as described elsewhere⁴⁸. Briefly, cells were seeded at 70–90% confluency on 10 μg ml⁻¹ fibronectin-coated glass slides for 48–72 h in complete medium. For shear stress stimulation, slides were mounted in custom-made 25 × 55-mm parallel plate shear chambers with 0.5-mm-thick silicone gaskets and stimulated with 15 dyn cm⁻² for LSS or 1 ± 4 dyn cm⁻² for OSS in complete medium. The medium was maintained at 37 °C and 5% CO₂ with a heat gun and humidified bubbler, respectively. An orbital shaker was used for mAb functional testing in vitro (Fig. 6d,e and Extended Data Fig. 7f,g). The well radius was divided into three equal parts, with the outermost part representing the pulsatile LSS region (to measure Klf2:GFP levels) and the innermost part representing the disturbed shear region (to measure VCAM1 levels).

Animals and tissue preparation

All experiments followed Yale Environmental Health and Safety regulations; all animal work was approved by the Yale Institutional Animal Care and Use Committee and Yale Animal Resource Center. Mice were maintained in a light-controlled and temperature-controlled environment with free access to food and water; all efforts were made to minimize animal suffering. Pcdhg^loxP/con3 mice^23,24 and Cdh5-Cre mice⁴⁹ are described elsewhere. Pcdhg^loxP/con3 mice were a gift from J. Lefebvre, University of Toronto, Canada. All mice in this study were on the C57BL/6J background. Pcdhg^loxP/con3 and Cdh5-Cre mice were maintained and bred as heterozygotes. Euthanasia was performed using an overdose of isoflurane inhalation and death was confirmed by subsequent cervical dislocation or by removing vital organs or opening the chest cavity to verify the absence of cardiovascular function. Mice were perfused through the left ventricle with PBS and then 3.7% formaldehyde followed by tissue collection, as described. The heart and spinal column, with the aorta and carotids attached, were removed and fixed under gentle agitation for an additional 24 h at 4 °C, washed three times with PBS and taken for further analysis. For the whole-aorta en face preparation, the isolated aortas were bisected along the lesser curvature; the aortic arch was also bisected through the greater curvature. For the aortic arch segment preparation, the aortic arch was bisected through the greater curvature. Hearts (containing aortic roots) and carotids were allowed to sink in 30% sucrose in PBS overnight at 4 °C, embedded in optimal cutting temperature (O.C.T.) compound (cat. no. 4583, Sakura) and frozen on dry ice for sectioning. Tissue blocks were cut into 8–10-μm sections using a cryostat (Leica); sections were stored at −80 °C until use.

Atherosclerosis and blood lipid analysis

To induce atherosclerosis, murine AAV8-PCSK9 (pAAV/D377Y-mPCSK9; 2 × 10¹¹ PFUs) produced by the Gene Therapy Program Vector Core at the University of Pennsylvania School of Medicine was injected intraperitoneally. Mice were maintained on an HFD (Clinton/Cybulsky high-fat rodent diet with regular casein and 1.25% added cholesterol; cat. no. D12108c, Research Diet) for 16 weeks. Blood samples were collected from mice starved overnight, centrifuged at 8,000g at 4 °C for 10 min; the supernatant (plasma) was separated, and HDL-C was isolated by precipitation of non-HDL-C (Wako Pure Chemicals). Both HDL-C fractions and total plasma were stored at −80 °C. Total plasma cholesterol and triglycerides were measured using kits according to the manufacturer’s instructions (Wako Pure Chemicals).

Tissue analysis

For IF, O.C.T. tissue sections were thawed and washed three times with PBS to remove O.C.T. Cells were fixed with 3.7% formaldehyde in PBS for 15 min at ambient temperature, washed with and stored in PBS. De-identified human specimens were deparaffinized in Histo-Clear (cat. no. HS-200, National Diagnostics). Sections were progressively rehydrated before antigen retrieval for 30 min at 95 °C in 1× antigen retrieval buffer (cat. no. 51699, Dako). Samples were incubated in perm-block buffer (5% donkey serum, 0.2% BSA, 0.3% Triton X-100 in PBS) for 1 h at room temperature, incubated with primary antibodies in perm-block overnight at 4 °C, washed three times in perm-block and then incubated with Alexa Fluor-conjugated secondary antibodies (Thermo Fisher Scientific) at 1:1,000 dilution in perm-block for 1 h at room temperature. Slides were washed three times in perm-block and three times in PBS before mounting in DAPI Fluoromount G (cat. no. 0100-20, Southern Biotech). Images were acquired on a Leica SP8 confocal microscope with the Leica Application Suite software. Confocal stacks were flattened by maximum-intensity z-projection in ImageJ. After background subtraction, the MFI or nuclear intensity (with DAPI mask) was recorded. For aorta Oil Red O staining, the whole aorta was opened longitudinally on a soft-bottomed silica dish, incubated with Oil Red O solution (0.6% Oil Red O in 60% isopropanol) with gentle rocking for 1 h at ambient temperature, washed in 60% isopropanol for 20 min, washed in distilled H₂O three times and mounted on slides with the endothelium side up in O.C.T. compound. Images were acquired with a digital microscopic camera (Leica DFC295, Leica Microsystems). Oil Red O staining on O.C.T. tissue sections was done similarly. Quantitation of Oil Red O⁺ area was done in ImageJ. H&E staining of O.C.T. tissue sections was done by the Yale Research Histology Core using standard techniques. Plaque morphometric and vulnerability analysis was performed as described⁵⁰. Plaque area was determined by Oil Red O⁺ staining. For each plaque, the NC area was defined as a clear area in the plaque that was H&E-free; FC thickness was quantified by selecting the largest NC and measuring the thinnest part of the cap.

LCMV infection, immune function analysis and FACS analysis

Control or Pcdhg ECKO mice were infected with 4 × 10⁶ PFU LCMV clone 13 by intravenous injection, euthanized on day 7 after infection and examined for immune function. Spleens were collected for flow cytometry analysis. For the cytotoxic T lymphocyte assay (restimulation assay), CD8a T cells from the spleen were pulsed with LCMV-specific peptide (GP33) or an irrelevant control peptide (SIINFEKL) and assayed for IFNγ and TNF levels. Kidneys were collected for viral load assessment.

E. coli peritonitis

Methods were performed as described in ref. ³⁰. Briefly, peritonitis was induced in male and female control or Pcdhg ECKO mice by injecting intraperitoneally with live E. coli serotype O6:K2:H1 at 10⁵ CFUs per mouse, in sterile saline. Mice were euthanized and exudates (peritoneal lavage) were collected for flow cytometry at the designated times. For bacterial titers, serially diluted exudates and blood were plated onto lysogeny broth (LB) agar plates and incubated overnight at 37 °C. The next day, LB plates were imaged and CFUs counted using ImageJ. Leukocyte populations were determined by flow cytometry. Peritoneal exudates were stained with surface antibodies: anti-mouse PerCP/Cy5.5 CD45; anti-mouse APC F4/80; anti-mouse APC/Cy7 Ly6G; anti-mouse FITC Ly6C; and anti-mouse PE/Cy7 CD11b for the identification of neutrophils (CD45⁺CD11b⁺F4/80⁻Ly6C⁻Ly6G⁺), monocytes (CD45⁺CD11b⁺F4/80⁻Ly6G⁻Ly6C⁺) and macrophages (CD45⁺CD11b⁺F4/80⁺). For in vivo exudate E. coli phagocytosis, peritoneal leukocytes were collected at 12 h after infection and stained with surface antibodies: anti-mouse PerCP/Cy5.5 CD45; anti-mouse APC F4/80; anti-mouse APC/Cy7 Ly6G; anti-mouse BV421 Ly6C; and anti-mouse PE/Cy7 CD11b for the identification of neutrophils, monocytes and macrophages. Exudates were then permeabilized using the BD Cytoperm Permeabilization Buffer (BD Biosciences) for 15 min according to the manufacturer’s protocol. Exudates were washed twice with BD Perm/Wash buffer (BD Biosciences). Fc receptor-mediated, nonspecific antibody binding was blocked using CD16/CD32 Fc block antibody (cat. no. MA5-29707, Thermo Fisher Scientific) and stained for intracellular E. coli using an FITC-conjugated anti-E. coli antibody (1:50 dilution) (cat. no. GTX40856; GeneTex). Cells were then analyzed using flow cytometry to assess the percentage intracellular E. coli (FITC⁺) in macrophages (CD45⁺CD11b⁺Ly6G⁻Ly6C⁻F4/80⁺), neutrophils (CD45⁺CD11b⁺F4/80⁻Ly6C⁻Ly6G⁺) and monocytes (CD45⁺CD11b⁺Ly6G⁻F4/80⁻Ly6C⁺). For the in situ E. coli phagocytosis assay, BM cells were collected from uninfected mice by flushing with PBS. BM neutrophils were isolated using the EasySep mouse neutrophil enrichment kit (STEMCELL Technologies). BM neutrophils were incubated with BacLight green (cat. no. B35000, Thermo Fisher Scientific) labeled E. coli (1:25 ratio) at 37 °C for 45 min for phagocytosis, washed, fixed and analyzed using flow cytometry. The BD FACS Diva v.9.0 was used for data acquisition and FlowJo v.10.10.0 for analysis.

Cloning and purification of Pcdhga9 ECD

Human Pcdhga9 mutants were generated by cloning PCR-amplified fragments into the pBob-GFP vector (Addgene). For GFP-tagged constructs, fragments were cloned upstream and in-frame with the GFP open reading frame. For FLAG-tagged constructs, GFP was excised and the fragments were cloned with an added C-terminal FLAG-tag using PCR. Mouse Pcdhga9 ECD was cloned in the pCDNA3.1 vector (Invitrogen). Secreted Pcdhga9 ECD-FLAG-GST was purified from the cell culture supernatant from transfected HEK 293T cells by collecting medium at 48 h and 96 h after transfection. Supernatant was spun at 6,000g, at 4 °C for 15 min to remove debris, incubated with 200 µl washed glutathione beads per 40 ml medium O/N at 4 °C with tumbling, washed three times with 0.1% Triton X-100 in PBS, and eluted by adding excess reduced glutathione solution. The concentration of ECD was estimated using Coomassie brilliant blue staining compared to BSA standards.

Generation and validation of mAbs and PCDHGA9 ECD–cell adhesion assay

A total of 1 mg purified PCDHGA9 ECD protein was used for mAb generation in rats (BiCell Scientific). mAbs were purified and concentrated from serum-free hybridoma cultures (BiCell Scientific). Low-adhesion 96-well plates were coated with ECD (10 μg ml⁻¹) for 1 h at ambient temperature, washed three times with 0.1% Triton X-100 in PBS, and blocked with 1% heat-denatured BSA in PBS for 1 h at ambient temperature. These plates were used for testing IgG specificity and cell adhesion to ECD, as described below. For specificity, affinity and the amount of IgG, mAbs were added to the plates for 1 h at ambient temperature with shaking, washed three times with 0.1% Triton X-100 in PBS, incubated with secondary horseradish peroxidase (HRP) antibody (1:5,000 dilution) for 1 h at ambient temperature with shaking, washed three times with 0.1% Triton X-100 in PBS and three times with PBS, followed by the addition of 100 μl 3,3′,5,5′-tetramethylbenzidine incubated for 15–30 min and absorbance measured at 605 nm, followed by the addition of 100 µl 0.1 N HCl and absorbance measured at 450 nm.

For the cell adhesion assays, wells were washed three times with complete EC medium followed by the addition of Klf2:GFP MAECs in 100 μl total EC medium (with isotype control or test mAbs), incubated at 37 °C in the CO₂ incubator for the indicated time and fixed by adding 33 μl of 16% paraformaldehyde directly to the wells. Unadhered cells were removed by turning the plate upside down in a water bath and cells counted based on mCherry fluorescence.

PCA ligation model of accelerated experimental atherosclerosis

Apoe^−/− mice aged 8–10 weeks maintained on an HFD for 1 week were anesthetized with ketamine and xylazine; surgery was performed as described⁵¹. Briefly, three out of four branches of the left common carotid artery (left external carotid, internal carotid and occipital artery) were ligated with sutures, with the superior thyroid artery left intact. Then, 2 μg mAb A9 or isotype control antibody was injected intraperitoneally once every week for 2 weeks. Mice were euthanized with an overdose of isoflurane and perfused through the left ventricle with PBS and then 3.7% formaldehyde. Aortas with carotid arteries were isolated and imaged whole. Carotid arteries were embedded in O.C.T., sectioned at 10 μm, and immunohistochemistry or IF performed as described. The LCA/RCA inner diameter ratio was calculated by measuring the perimeter to avoid interference from changes in vessel morphology during mounting and handling. For testing mAb retention in vivo (half-life), 1 µg of mAbs (100 μl of 0.25 mg ml⁻¹ in saline for a 20-g mouse) were injected intraperitoneally in C57BL/6 mice; 100 μl of blood was collected via the retro-orbital route at the indicated times, centrifuged at 13,000g at 4 °C for 15 min and the plasma was removed and immediately stored at −80 °C. The mAb concentration was determined using a sandwich enzyme-linked immunosorbent assay with immobilized anti-rat IgG (100 ng) as trap and secondary anti-rat IgG HRP antibody for detection, as described above.

Immunoblotting and immunoprecipitation

Cells were collected and lysed in radioimmunoprecipitation assay buffer (Roche) containing 1× Halt Protease Inhibitor Cocktail (cat. no. 78429, Thermo Fisher Scientific) and 1× PhosStop (cat. no. 4906837001, Roche) for 30 min on ice, clarified at 13,000g at 4 °C for 15 min, the supernatant was transferred to new 1.5-ml tubes, 4× loading buffer (250 mM Tris-HCl, pH 6.8, 8% SDS, 40% glycerol, 20% β-mercaptoethanol, 0.008% bromophenol blue) added and the samples were heated to 95 °C for 5 min. Cell lysates were resolved by a 4–15% SDS–polyacrylamide gel electrophoresis (PAGE), transferred to 0.2-μm nitrocellulose membranes, which were blocked with 5% nonfat skimmed milk for 1 h at ambient temperature and incubated with the desired antibodies diluted in 5% BSA using a standard immunoblotting procedure and detection using electrochemiluminescence (Merck Millipore). Images were quantified with ImageJ using densitometry and normalized to GAPDH or tubulin loading controls. For immunoprecipitation, GFP-trap (cat. no. GTA-20, Chromotech) or FLAG M2 (cat. no. A2220, Sigma-Aldrich) agarose was used. Lysates were collected in 25 mM Tris, pH 7.4, 150 mM NaCl, 1% TritonX-100, 1× Halt Protease Inhibitor Cocktail and 1× PhosStop, clarified at 13,000g at 4 °C for 15 min, incubated with antibody-bound beads at 4 °C for 2 h to overnight. Beads were washed three times with lysis buffer at 4 °C and eluted with 2× protein sample buffer (for GFP-trap) or 3× FLAG peptide competition (for FLAG M2) and subjected to SDS–PAGE or mass spectrometry. Uncropped gels, blots and scans provided (Supplementary Fig. 2).

RNA isolation, sequencing and quantitative PCR

Total RNA was extracted from cells with the RNeasy Plus Mini Kit (cat. no. 74136, QIAGEN) according to the manufacturer’s instructions. RNA was quantified using NanoDrop; RNA integrity was measured with an Agilent Bioanalyzer. Samples were subjected to RNA-seq using an Illumina NovaSeq 6000 (HiSeq paired-end, 100 bp). The base calling data from the sequencer were transferred into FASTQ files, using the bcl2fastq2 conversion software v.2.20 (Illumina). PartekFlow (a start-to-finish software analysis solution for NGS data applications) was used to determine DEGs. For the quantitative PCR with reverse transcription (RT–qPCR) analysis, reverse transcription was performed with the iScript Reverse Transcription Supermix for RT–qPCR (Bio-Rad Laboratories). RT–qPCR was performed with the SsoAdvanced Universal SYBR Green Supermix (Bio-Rad Laboratories). The expression of target genes was normalized to GAPDH. The RT–qPCR primers are listed in Supplementary Table 5.

Quantification, statistics and reproducibility

ImageJ v.1.51 (National Institutes of Health) was used for morphometric analysis. Graph preparation and statistical analysis was performed using Prism 10.0 (GraphPad Software). Unless otherwise indicated, all experiments were repeated at least three times, as described in the figure legends. Data were considered normally distributed and statistical significance was performed using a two-tailed Student t-test for two-group comparisons, a one-way ANOVA with Tukey post hoc analysis or a two-way ANOVA with Bonferroni correction for multiple comparisons, as described in the figure legends. Data are presented as the mean ± s.e.m.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The RNA-seq data were deposited in the Gene Expression Omnibus under accession no. GSE272071. Data generated or analyzed during this study are freely available.

References

Baeyens, N. et al. Defective fluid shear stress mechanotransduction mediates hereditary hemorrhagic telangiectasia. J. Cell Biol. 214, 807–816 (2016).
Article CAS PubMed PubMed Central Google Scholar
Kwak, B. R. et al. Biomechanical factors in atherosclerosis: mechanisms and clinical implications. Eur. Heart J. 35, 3013–3020 (2014).
Article CAS PubMed PubMed Central Google Scholar
Andreou, I. et al. How do we prevent the vulnerable atherosclerotic plaque from rupturing? Insights from in vivo assessments of plaque, vascular remodeling, and local endothelial shear stress. J. Cardiovasc. Pharmacol. Ther. 20, 261–275 (2015).
Article PubMed Google Scholar
Papafaklis, M. I. et al. Effect of the local hemodynamic environment on the de novo development and progression of eccentric coronary atherosclerosis in humans: insights from PREDICTION. Atherosclerosis 240, 205–211 (2015).
Article CAS PubMed Google Scholar
Siasos, G. et al. The role of shear stress in coronary artery disease. Curr. Top. Med. Chem. 23, 2132–2157 (2023).
Article CAS PubMed Google Scholar
Fledderus, J. O. et al. KLF2 primes the antioxidant transcription factor Nrf2 for activation in endothelial cells. Arterioscler. Thromb. Vasc. Biol. 28, 1339–1346 (2008).
Article CAS PubMed Google Scholar
Hahn, C. & Schwartz, M. A. Mechanotransduction in vascular physiology and atherogenesis. Nat. Rev. Mol. Cell Biol. 10, 53–62 (2009).
Article CAS PubMed PubMed Central Google Scholar
Nayak, L., Lin, Z. & Jain, M. K. ‘Go with the flow’: how Kruppel-like factor 2 regulates the vasoprotective effects of shear stress. Antioxid. Redox Signal. 15, 1449–1461 (2011).
Article CAS PubMed PubMed Central Google Scholar
Nigro, P., Abe, J.-I. & Berk, B. C. Flow shear stress and atherosclerosis: a matter of site specificity. Antioxid. Redox Signal. 15, 1405–1414 (2011).
Article CAS PubMed PubMed Central Google Scholar
Novodvorsky, P. & Chico, T. J. A. The role of the transcription factor KLF2 in vascular development and disease. Prog. Mol. Biol. Transl. Sci. 124, 155–188 (2014).
Article CAS PubMed Google Scholar
SenBanerjee, S. et al. KLF2 Is a novel transcriptional regulator of endothelial proinflammatory activation. J. Exp. Med. 199, 1305–1315 (2004).
Article CAS PubMed PubMed Central Google Scholar
Sweet, D. R., Fan, L., Hsieh, P. N. & Jain, M. K. Krüppel-like factors in vascular inflammation: mechanistic insights and therapeutic potential. Front. Cardiovasc. Med. 5, 6 (2018).
Article PubMed PubMed Central Google Scholar
Xu, S. et al. The zinc finger transcription factor, KLF2, protects against COVID-19 associated endothelial dysfunction. Signal. Transduct. Target. Ther. 6, 266 (2021).
Article CAS PubMed PubMed Central Google Scholar
Dabravolski, S. A. et al. The role of KLF2 in the regulation of atherosclerosis development and potential use of KLF2-targeted therapy. Biomedicines 10, 254 (2022).
Article CAS PubMed PubMed Central Google Scholar
Coon, B. G. et al. A mitochondrial contribution to anti-inflammatory shear stress signaling in vascular endothelial cells. J. Cell Biol. 221, e202109144 (2022).
Article PubMed PubMed Central Google Scholar
Abe, J. & Berk, B. C. Novel mechanisms of endothelial mechanotransduction. Arterioscler. Thromb. Vasc. Biol. 34, 2378–2386 (2014).
Article CAS PubMed PubMed Central Google Scholar
Chen, W. V. & Maniatis, T. Clustered protocadherins. Development 140, 3297–3302 (2013).
Article CAS PubMed PubMed Central Google Scholar
Pancho, A., Aerts, T., Mitsogiannis, M. D. & Seuntjens, E. Protocadherins at the crossroad of signaling pathways. Front. Mol. Neurosci. 13, 117 (2020).
Article CAS PubMed PubMed Central Google Scholar
Flaherty, E. & Maniatis, T. The role of clustered protocadherins in neurodevelopment and neuropsychiatric diseases. Curr. Opin. Genet. Dev. 65, 144–150 (2020).
Article CAS PubMed PubMed Central Google Scholar
Jia, Z. & Wu, Q. Clustered protocadherins emerge as novel susceptibility loci for mental disorders. Front. Neurosci. 14, 587819 (2020).
Article PubMed PubMed Central Google Scholar
Sen-Banerjee, S. et al. Kruppel-like factor 2 as a novel mediator of statin effects in endothelial cells. Circulation 112, 720–726 (2005).
Article CAS PubMed Google Scholar
Zhou, G. et al. Endothelial Kruppel-like factor 4 protects against atherothrombosis in mice. J. Clin. Invest. 122, 4727–4731 (2012).
Article CAS PubMed PubMed Central Google Scholar
Lefebvre, J. L., Zhang, Y., Meister, M., Wang, X. & Sanes, J. R. γ-Protocadherins regulate neuronal survival but are dispensable for circuit formation in retina. Development 135, 4141–4151 (2008).
Article CAS PubMed Google Scholar
Prasad, T., Wang, X., Gray, P. A. & Weiner, J. A. A differential developmental pattern of spinal interneuron apoptosis during synaptogenesis: insights from genetic analyses of the protocadherin-γ gene cluster. Development 135, 4153–4164 (2008).
Article CAS PubMed Google Scholar
Wang, X. et al. Gamma protocadherins are required for survival of spinal interneurons. Neuron 36, 843–854 (2002).
Article CAS PubMed Google Scholar
Weiner, J. A., Wang, X., Tapia, J. C. & Sanes, J. R. Gamma protocadherins are required for synaptic development in the spinal cord. Proc. Natl Acad. Sci. USA 102, 8–14 (2005).
Article CAS PubMed Google Scholar
Bjørklund, M. M. et al. Induction of atherosclerosis in mice and hamsters without germline genetic engineering. Circ. Res. 114, 1684–1689 (2014).
Article PubMed Google Scholar
Ilyas, I. et al. Mouse models of atherosclerosis in translational research. Trends Pharmacol. Sci. 43, 920–939 (2022).
Article CAS PubMed Google Scholar
Chiang, N. et al. Infection regulates pro-resolving mediators that lower antibiotic requirements. Nature 484, 524–528 (2012).
Article CAS PubMed PubMed Central Google Scholar
Libreros, S., Nshimiyimana, R., Lee, B. & Serhan, C. N. Infectious neutrophil deployment is regulated by resolvin D4. Blood 142, 589–606 (2023).
Article CAS PubMed PubMed Central Google Scholar
Serhan, C. N. Pro-resolving lipid mediators are leads for resolution physiology. Nature 510, 92–101 (2014).
Article CAS PubMed PubMed Central Google Scholar
Haas, I. G., Frank, M., Véron, N. & Kemler, R. Presenilin-dependent processing and nuclear function of γ-protocadherins. J. Biol. Chem. 280, 9313–9319 (2005).
Article CAS PubMed Google Scholar
Bray, S. J. Notch signalling: a simple pathway becomes complex. Nat. Rev. Mol. Cell Biol. 7, 678–689 (2006).
Article CAS PubMed Google Scholar
Fang, J. S. et al. Shear-induced Notch-Cx37-p27 axis arrests endothelial cell cycle to enable arterial specification. Nat. Commun. 8, 2149 (2017).
Article PubMed PubMed Central Google Scholar
Kong, P. et al. Inflammation and atherosclerosis: signaling pathways and therapeutic intervention. Signal. Transduct. Target. Ther. 7, 131 (2022).
Article CAS PubMed PubMed Central Google Scholar
Mack, J. J. et al. NOTCH1 is a mechanosensor in adult arteries. Nat. Commun. 8, 1620 (2017).
Article PubMed PubMed Central Google Scholar
Polacheck, W. J. et al. A non-canonical Notch complex regulates adherens junctions and vascular barrier function. Nature 552, 258–262 (2017).
Article CAS PubMed PubMed Central Google Scholar
Park, K. S., Schecterson, L. & Gumbiner, B. M. Enhanced endothelial barrier function by monoclonal antibody activation of vascular endothelial cadherin. Am. J. Physiol. Heart Circ. Physiol. 320, H1403–H1410 (2021).
Article CAS PubMed PubMed Central Google Scholar
Kumar, S., Kang, D. W., Rezvan, A. & Jo, H. Accelerated atherosclerosis development in C57Bl6 mice by overexpressing AAV-mediated PCSK9 and partial carotid ligation. Lab. Invest. 97, 935–945 (2017).
Article CAS PubMed PubMed Central Google Scholar
Nam, D. et al. Partial carotid ligation is a model of acutely induced disturbed flow, leading to rapid endothelial dysfunction and atherosclerosis. Am. J. Physiol. Heart Circ. Physiol. 297, H1535–H1543 (2009).
Article CAS PubMed PubMed Central Google Scholar
Niu, N., Xu, S., Xu, Y., Little, P. J. & Jin, Z.-G. Targeting mechanosensitive transcription factors in atherosclerosis. Trends Pharmacol. Sci. 40, 253–266 (2019).
Article CAS PubMed PubMed Central Google Scholar
Lefebvre, J. L., Kostadinov, D., Chen, W. V., Maniatis, T. & Sanes, J. R. Protocadherins mediate dendritic self-avoidance in the mammalian nervous system. Nature 488, 517–521 (2012).
Article CAS PubMed PubMed Central Google Scholar
Rubinstein, R. et al. Molecular logic of neuronal self-recognition through protocadherin domain interactions. Cell 163, 629–642 (2015).
Article CAS PubMed PubMed Central Google Scholar
Gabbert, L., Dilling, C., Meybohm, P. & Burek, M. Deletion of protocadherin gamma C3 induces phenotypic and functional changes in brain microvascular endothelial cells in vitro. Front. Pharmacol. 11, 590144 (2020).
Article CAS PubMed PubMed Central Google Scholar
Kaupp, V., Blecharz-Lang, K. G., Dilling, C., Meybohm, P. & Burek, M. Protocadherin gamma C3: a new player in regulating vascular barrier function. Neural Regen. Res. 18, 68–73 (2023).
Article CAS PubMed Google Scholar
Ridker, P. M. et al. Antiinflammatory therapy with canakinumab for atherosclerotic disease. N. Engl. J. Med. 377, 1119–1131 (2017).
Article CAS PubMed Google Scholar
Ni, C. W., Kumar, S., Ankeny, C. J. & Jo, H. Development of immortalized mouse aortic endothelial cell lines. Vasc. Cell 6, 7 (2014).
Article PubMed PubMed Central Google Scholar
Conway, D. E. et al. VE-cadherin phosphorylation regulates endothelial fluid shear stress responses through the polarity protein LGN. Curr. Biol. 27, 2219–2225 (2017); erratum 27, 2727 (2017).
Article CAS PubMed PubMed Central Google Scholar
Alva, J. A. et al. VE-Cadherin-Cre-recombinase transgenic mouse: a tool for lineage analysis and gene deletion in endothelial cells. Dev. Dyn. 235, 759–767 (2006).
Article CAS PubMed Google Scholar
Seimon, T. A. et al. Macrophage deficiency of p38α MAPK promotes apoptosis and plaque necrosis in advanced atherosclerotic lesions in mice. J. Clin. Invest. 119, 886–898 (2009).
CAS PubMed PubMed Central Google Scholar
Budatha, M., Zhang, J. & Schwartz, M. A. Fibronectin-mediated inflammatory signaling through integrin α5 in vascular remodeling. J. Am. Heart Assoc. 10, e021160 (2021).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We thank J. Lefebvre (University of Toronto) for the Pcdhg^fcon3 mice, J. Zhang (Yale University Animal Core) for the PCA ligation surgery, D. Zhao (Yale Center for Genome Analysis) for the analysis of the RNA-seq data, Schwartz laboratory members for the extensive discussions, the Yale Center for Genome Analysis for the RNA-seq, Yale Keck Oligo Synthesis Resource and DNA Sequencing Core, and the Harvard Center for mass spectrometry. This work was supported by a National Institutes of Health grant no. R01 HL75092, Leducq Trans-Atlantic Network Grant 18CVD03 and a Health Research Consortium grant from Fundación Obra Social La Caixa (AtheroConvergence, no. HR20-00075) to M.A.S. and an American Heart Association grant no. 23POST1026109 to D.J.

Author information

Authors and Affiliations

Yale Cardiovascular Research Center, Section of Cardiovascular Medicine, Department of Internal Medicine, School of Medicine, Yale University, New Haven, CT, USA
Divyesh Joshi, Brian G. Coon, Raja Chakraborty, Hanqiang Deng, Emily Meredith & Martin A. Schwartz
Vascular Biology and Therapeutics Program, Yale University, New Haven, CT, USA
Ziyu Yang, Muhammad Usman Babar, Pablo Fernandez-Tussy, Carlos Fernandez-Hernando & Stephania Libreros
Department of Pathology, Yale University, New Haven, CT, USA
Ziyu Yang, Muhammad Usman Babar & Stephania Libreros
Department of Immunobiology, Yale University, New Haven, CT, USA
John Attanasio & Nikhil Joshi
Department of Pathology and Translational Pathobiology, LSU Health Shreveport, Shreveport, LA, USA
James G. Traylor Jr. & Anthony Wayne Orr
Department of Cell Biology, Yale University, New Haven, CT, USA
Martin A. Schwartz
Department of Biomedical Engineering, Yale University, New Haven, CT, USA
Martin A. Schwartz

Authors

Divyesh Joshi
View author publications
You can also search for this author in PubMed Google Scholar
Brian G. Coon
View author publications
You can also search for this author in PubMed Google Scholar
Raja Chakraborty
View author publications
You can also search for this author in PubMed Google Scholar
Hanqiang Deng
View author publications
You can also search for this author in PubMed Google Scholar
Ziyu Yang
View author publications
You can also search for this author in PubMed Google Scholar
Muhammad Usman Babar
View author publications
You can also search for this author in PubMed Google Scholar
Pablo Fernandez-Tussy
View author publications
You can also search for this author in PubMed Google Scholar
Emily Meredith
View author publications
You can also search for this author in PubMed Google Scholar
John Attanasio
View author publications
You can also search for this author in PubMed Google Scholar
Nikhil Joshi
View author publications
You can also search for this author in PubMed Google Scholar
James G. Traylor Jr.
View author publications
You can also search for this author in PubMed Google Scholar
Anthony Wayne Orr
View author publications
You can also search for this author in PubMed Google Scholar
Carlos Fernandez-Hernando
View author publications
You can also search for this author in PubMed Google Scholar
Stephania Libreros
View author publications
You can also search for this author in PubMed Google Scholar
Martin A. Schwartz
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

D.J. performed most of the experiments, analyzed the data and prepared the figures. B.G.C. performed the CRISPR screen and analysis. D.J. and R.C. performed and analyzed the PCA ligation experiments. D.J., H.D. and E.M. performed the antibody injection experiments in mice. H.D. and P.F.-T. performed the plasma lipid analysis experiments. J.A. performed and analyzed the LCMV infection experiments. N.J. supervised the LCMV infection experiments. D.J., Z.Y., M.U.B. and S.L. performed the E. coli peritonitis experiments. D.J. and S.L. analyzed the peritonitis experiments. S.L. supervised the peritonitis experiments. J.G.T. and A.W.O. provided the human ASCVD samples. C.F.-H. supervised the lipid analysis experiments. M.A.S. conceived and supervised the project. D.J. and M.A.S. acquired the funding for the project and wrote the manuscript.

Corresponding author

Correspondence to Martin A. Schwartz.

Ethics declarations

Competing interests

M.A.S. and D.J. are listed as inventors in a US Provisional Patent Application no. 63/621,466 filed by Yale University for inhibitors of Pcdhga9. The other authors declare no competing interests.

Peer review

Peer review information

Nature Cardiovascular Research thanks Mukesh Jain and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Validation of Klf2 reporter.

(a) Immunoblot validation of LSS-mediated induction of Klf2:GFP in reporter MAECs; reporter is also sensitive to statin treatment, another potent inducer of Klf2 used as a positive control (N = 3 independent experiments). (b) Rank-ordered candidate suppressors from the CRISPR screen to identify Klf2 modifiers (z < 4 gray, z > 4 blue), with validated cell-surface exposed candidates which are amenable to function neutralization marked in red, and known positive controls, Ccm2 and Pdcd10, marked in black. (c) qRT-PCR validation of siRNA mediated knockdown of Pcdhga9 in MAECs (Pcdhga9 si) (N = 3 independent experiments). (d) Immunoblot validation of human Pcdhga9-FLAG overexpression (Hs Pcdhga9 OE) in MAECs (N = 3 independent experiments). (e) Control si and Pcdhga9 si Klf2 reporter MAECs were exposed to static (St), LSS, or oscillatory shear stress (OSS) for 16 h and immunoblotted for Klf2:GFP (N = 3 independent experiments). (f) qRT-PCR for endogenous Klf2 levels in Control si and Pcdhga9 si in Human Umbilical Vein Endothelial Cells (HUVECs) exposed to St or OSS for 16 h (N = 3 independent experiments). Values are means ± SEM. Statistical analysis used one-way ANOVA.

Extended Data Fig. 2 Protocadherin gamma (Pcdhg) gene cluster promotes inflammatory signaling.

(a) Validation of siRNA mediated knockdown of Pcdhg gene cluster in HUVECs using two different siRNAs targeting the common region in the 3′ end, by qRT-PCR for Pcdhgc3, the highest expressed Pcdhg member in HUVECs (N = 3 independent experiments). (b) qRT-PCR for the OSS induced pro-inflammatory marker E-selectin (Sele) in control and Pcdhg depleted HUVECs exposed to St, LSS and OSS for 16 h (N = 3 independent experiments). (c) Immunoblot for VCAM1 in Control si or Pcdhg si HUVECs, after treatment with indicated doses of TNFα for 16 h (N = 3 independent experiments). Graph: quantitation of VCAM1 normalized to Tubulin loading control. Values are means ± SEM. Statistical analysis used one-way ANOVA.

Extended Data Fig. 3 Validation and blood lipid analysis of Pcdhg ECKO mouse.

(a) Generation of Pcdhg endothelial knockout (ECKO) by crossing Pcdhg^fcon3 with Cdh5Cre, and confirmation by genotyping PCR. wt: wild type for Pcdhg allele; flox: Pcdhg floxed; Cre: Cdh5Cre (N > 10 animals). (b, c) Analysis of progeny genotype showing no significant deviation from Mendelian ratio as tested by Chi-squared analysis. (d–g) Plasma triglycerides, cholesterol, HDL-C and body weights of male and female Control or Pcdhg ECKO mice injected with pCSK9-Adeno Associated Virus 8 (AAV8) and maintained on High Fat Diet (HFD) for 16 weeks, starved overnight before analysis (N = 5). Values are means ± SEM. Statistical analysis was carried out using one-way ANOVA.

Extended Data Fig. 4 Pcdhg ECKO does not affect immune function.

(a-c) Mice were infected with Lymphocytic Choriomeningitis Virus (LCMV) Clone 13 and spleen myeloid cell frequency measured on Day 7 post infection (N = 6 animals). Graphs: (a) percentages of live CD11b+ or CD11c + CD11b- non T/B-cells; (b) monocyte (Ly6c + Ly6g-) and neutrophil/granulocyte (Ly6c + Ly6g + ) population of the CD11b+ cells; (c) subsets of CD11b+ cells, as indicated. (d-i) Immune response in Control or Pcdhg ECKO to intraperitoneally injected E. coli by measuring CFUs (d) and representative LB agar plates showing CFUs in males (e), CFUs in blood (f), flow cytometry quantification of neutrophils (CD45⁺CD11b⁺F4/80⁻Ly6C⁻Ly6G⁺) and monocytes (CD45⁺CD11b⁺F4/80⁻Ly6G⁻Ly6C⁺), and macrophages (CD45⁺CD11b⁺F4/80⁺) numbers (g, h) and cell numbers in peritoneal exudate (i) (N = 3). (j, k) Phagocytic function in the peritoneal exudates at 12 h post infection (N = 3). (l) Phagocytic function of bone marrow neutrophils from uninfected Control or Pcdhg ECKO (j) (N = 4). Values are means ± SEM. Statistical analysis was carried out using unpaired two-tailed Student’s t-test (a-c, j, l) or two-way ANOVA (d, i, h).

Extended Data Fig. 5 Conserved nuclear ICD region of Pcdhg associates with Notch ICD and is necessary and sufficient for function.

(a) Pcdhg mutants were expressed in HUVECs which were treated with Static, LSS or OSS and immunoblotted for GFP, Klf4 and GAPDH (N = 3 independent experiments). (b) GFP was imaged in HUVECs expressing the above mutants (N = 3 independent experiments). (c) HUVECs expressing additional mutants (C-terminal FLAG-tagged) were immunostained for FLAG and counterstained with DAPI to mark nuclei (N = 3 independent experiments). (d) Multiple sequence alignment to test domain homology domains, using Pcdhga9 as an example, showing near-complete conservation in CCDs (highlighted in yellow). Percent conservation shown in the box. (e) RBPJ-Notch DNA-binding consensus motif. (f) RBPJ-Notch binding motifs in mouse and human Klf2 and Klf4 promoters. (g) NICD peptides detected in proteomic analysis of the IPs of full but not the ΔCCD Pcdhg mutant. Scale bar: (b, c) 20 μm.

Extended Data Fig. 6 Pcdhga9 blocking antibody generation and validation.

(a) Pcdhga9 ECD (ECD-FLAG-TEV-GST) protein run on a 10% Polyacrylamide SDS gel and visualized using Imperial Protein stain (Thermo Scientific) (N = 3 independent experiments). (b) ELISA for the 24 selected high affinity monoclonal antibody (mAb) from clones labeled as A1-12 and B1-12 using ECD alone (GST cleaved off using TEV protease). (c) Adhesion of MAECs to ECD in the presence of Isotype control or mAbs A9, B1 and B4 (N = 8 images across 4 independent experiments). Graph: quantitation of percent total cells adhered to ECD. (d) Immunoblot with mAbs A9, B1, B4 shows detection of both ECD-FLAG-TEV-GST and ECD-FLAG (N = 3 independent experiments). (e, f) Klf2:GFP reporter MAECs tested for Klf2:GFP reporter expression after 16 h LSS (N = 12 images across 3 independent experiments) (e) or immunostained for VCAM1 when exposed to OSS for 16 h OSS (N = 7-8 images across 3 independent experiments) (f), in the presence of Isotype control or mAbs A9, B1 and B4 as indicated. Graphs: quantitation of Klf2:GFP and VCAM1 levels normalized to mCherry internal control. (g) Immunofluorescence with mAbs A9, B1, B4 showing highest signal from A9. (h) Immunoblot of Pcdhg knockdown cells with mAb A9. Purified ECD used as positive control. (i) Antibody half-life in vivo. A single dose of 1 μg of Isotype control or mAb A9 antibody was administered IP, and plasma levels of antibody were measured using an ELISA as described in Methods. mAb levels at 3 h post injection were considered 100%. (j) LCA and RCA sections were stained for smooth muscle specific Acta2 (SMA) for marking plaque neointima (N = 6). Graph: quantitation of the LCA to RCA inner diameter. (k, l) Plasma triglycerides and cholesterol in Isotype control or mAb A9 injected male and female Apoe^−/− mice on HFD from Fig. 5h, i (N = 6). Values are means ± SEM. Statistical analysis was carried out using unpaired two-tailed Student’s t-test (j) or one-way ANOVA (c, e, f, k, l). Scale bar: (c) 100 μm, (g) 20 μm, (j) 100 μm.

Extended Data Fig. 7 Pcdhg level correlates with atherosclerosis.

(a) Mouse carotids from the Partial Carotid Artery (PCA) Ligation model of accelerated atherosclerosis stained for Pcdhg and counter stained with DAPI to mark nuclei (N = 3 animals). RCA: Right Carotid Artery (control), LCA: Left Carotid Artery (atherosclerotic plaque. Graph: quantitation of Pcdhg staining intensity. (b) Commercial Pcdhg antibody and mAb A9 staining of retinal vasculature from Control and Pcdhg ECKO mouse (N = 3 animals). mAb A9 is specific to mouse Pcdhga9 and Pcdhg antibody targets the conserved region in 22 Pcdhg genes, also conserved between mouse and human. Values are means ± SEM. Statistical analysis was carried out using unpaired two-tailed Student’s t-test. Scale bar: (a) 100 μm, (b) 10 μm.

Supplementary information

Supplementary Information

Supplementary Figs. 1 and 2 and Tables 3–5.

Reporting Summary

Supplementary Tables

Table 1. Candidate Klf2 suppressors. Table 2. Cell-surface-expressed candidate Klf2 suppressors.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Joshi, D., Coon, B.G., Chakraborty, R. et al. Endothelial γ-protocadherins inhibit KLF2 and KLF4 to promote atherosclerosis. Nat Cardiovasc Res 3, 1035–1048 (2024). https://doi.org/10.1038/s44161-024-00522-z

Download citation

Received: 16 May 2024
Accepted: 18 July 2024
Published: 04 September 2024
Issue Date: September 2024
DOI: https://doi.org/10.1038/s44161-024-00522-z

This article is cited by

Saving KLF2/4 from γ-protocadherin to reduce vascular inflammation and atherosclerosis
- Christian Park
- Kyung In Baek
- Hanjoong Jo
Nature Cardiovascular Research (2024)