Medial SMCs in Dilated Human Thoracic Aortas Have Reduced NAMPT

To determine whether NAMPT is expressed in SMCs of the human aorta, we evaluated ascending aortas surgically harvested from patients with dilated ascending aortopathy (n=20; mean age 58.4±16.4) and compared the findings with those of nondilated ascending aorta harvested from individuals undergoing heart transplantation or coronary artery bypass surgery (n=16; mean age 68.7±14.1). The mean aortic diameter of diseased aortas was 54.3±9.5 mm and that of control aortas was 31.1±2.8 mm (P<0.0001). Individual patient demographic data, aortic diameter, and aortic valve configuration and function are presented in Online Table I.

Immunostaining revealed widespread NAMPT expression in medial SMCs of the normal caliber aortas, with signal in both cytoplasm and nucleus (Figure 1A). Medial cell NAMPT expression was also evident in dilated aortas. However, the NAMPT signal was more heterogeneous and on average 34% lower (Figure 1B through 1D), as assessed by DAB (3,3′-diaminobenzidine) intensity analysis. Reduced NAMPT expression in the media of dilated aortas was also observed at the transcript level (Figure 1E). Interestingly, there was an inverse relationship between aortic medial NAMPT content and the maximal diameter of the ascending aorta (R²=0.44; P<0.0001). This inverse relationship existed for both the entire patient cohort and only those patients with ascending aortopathy (R²=0.31; P=0.009; Figure 1F) and persisted after adjusting for patient age (P=0.016). These findings implicate a NAMPT-based NAD⁺ biosynthesis pathway within SMCs of the human aorta. Moreover, they suggest that this local biosynthetic machinery is related to aortic medial integrity.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Human ascending aortic dilation is associated with reduced nicotinamide phosphoribosyltransferase (NAMPT). A and B, Photomicrographs of paraformaldehyde-fixed nondilated (A) and dilated (B and C) human aorta sections immunostained for NAMPT and counterstained with hematoxylin. Arrow in zoomed image of (A) depicts nuclear signal, and arrowhead depicts cytoplasmic signal. Arrows in zoomed image of (B) indicate NAMPT-negative cells. Arrow in zoomed image of (C) indicates absent nuclear signal, and arrowhead depicts weak cytoplasmic signal. D. Graph depicting NAMPT content in aortas from patients with control, nondilated aortas (n=16) and dilated aortas (n=20). *P<0.0001 E, Graph depicting NAMPT mRNA abundance in control (n=6) and dilated (n=6) aortas. *P=0.038 F, Inverse relationship between NAMPT content in dilated aortas and aortic diameter (n=20).

Generation of Mice With SMC Nampt Gene Ablation

To determine whether Nampt within SMCs plays a role in aortic health, we generated mice with targeted deletion of Nampt in SMCs, using the smMHC-Cre/eGFP-expressing mouse line²² and Nampt^flox/flox mice.²¹ Evaluation of eGFP in smMHC-Cre/eGFP mice confirmed Cre recombinase in the aorta, as assessed by whole organ microscopy and immunostaining (Online Figure I). Analysis of >200 offspring from a 2-step breeding protocol revealed close to expected Mendelian ratios for the predicted genotypes: 26.7% WT, 29.1% SMC-specific Nampt heterozygous, 20.6% Nampt heterozygous, and 23.5% SMC-Nampt knockout (KO).

Laser capture microdissection and quantitative reverse transcription-polymerase chain reaction revealed that Nampt transcript abundance in thoracic aortic media of SMC-Nampt KO mice was reduced to 13.5% that of control mice (Figure 2A). Immunostaining revealed nuclear and cytoplasmic Nampt protein in medial SMCs of control aortas, with the nuclear signal being stronger (Figure 2B). Nampt was undetectable in the aortic media of SMC-Nampt KO mice in all but rare cells but still detectable in endothelial cells and adventitial cells (Figure 2B). We also determined aortic NAD⁺ content, using an NAD⁺ cycling assay. This revealed a 43.2% reduction in NAD⁺ content in SMC-Nampt compared with control mice (Figure 2C). Thus, Nampt is expressed in SMCs of the mouse aortic media and its depletion in SMCs, obtained using a Cre-lox approach, compromised aortic NAD⁺ homeostasis.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Deletion of nicotinamide phosphoribosyltransferase (Nampt) in smooth muscle cells (SMCs) in mice yields modest aortic dilatation. A, Graph depicting Nampt transcript abundance in the mouse aortic media harvested using laser capture microdissection, measured by quantitative reverse transcription-polymerase chain reaction. Medial tissues from 3 mice for each genotype were studied. *P=0.0034 vs control, †P<0.0001 vs control B, Photomicrographs of thoracic aortic sections from control and SMC-Nampt knockout (KO) mice, immunostained for Nampt. Black arrows depict Nampt expression in SMCs, and red arrows depict Nampt expression in endothelial cells. C, Graph of nicotinamide adenine dinucleotide (NAD⁺) content in acidic extracts of endothelium-denuded aortas of control and SMC-Nampt KO mice. *P=0.024 vs control. D, Sections from the descending thoracic aorta of 12-wk-old mice stained for smooth muscle α-actin (green) and propidium iodide (red). E–H, Graphs showing the number of elastic lamellae per regional cross section (E), medial cell density (F), regional aortic lumen area (G), and medial area (H). *P=0.023 and †P=0.029 vs respective lumen area in control aortas, *P=0.029, †P=0.004, and ‡P=0.027 vs respective medial area in control aortas.

Mice With SMC Nampt Deletion Have Modestly Dilated Thoracic Aortas

SMC-Nampt KO mice were viable and displayed no gross evidence of vascular anomalies. The average mean arterial pressure and heart rate of 8-week-old SMC-Nampt KO mice were not significantly different than those of control mice (Online Figure II). Immunostaining revealed abundant smooth muscle α-actin in medial SMCs, a normal number of lamellar units throughout the aorta, and unaltered medial SMC content (Figure 2E and 2F). Quantification of lumen area of aortas fixed at physiological pressure did, however, reveal modest dilation of the ascending and descending thoracic aortic regions (by 16.7% and 12.4%, Figure 2G). As well, the aortic medial areas of SMC-Nampt KO mice were mildly reduced in the ascending, descending thoracic, and suprarenal regions (by 13.2%, 17.0%, and 24.5%, respectively, Figure 2H). Transcript analysis revealed a modest (27.2%) decrease in mRNA abundance of SM-α-actin (P=0.001) and a 1.4-fold increase in Mmp2 expression (P=0.005) with no change in Timp1 expression (P=0.010).

Mice With SMC Nampt Deletion Are Susceptible to Ang II–Induced Aortic Dissection

To ascertain the response of the Nampt-deficient aorta to disease-associated stress, mice were subjected to continuous delivery of Ang II (1.44 mg/kg/d) by subcutaneous mini-pump. Interestingly, this was associated with a 54% decrease in aortic Nampt transcript abundance (P=0.009) and a concordant trend for NAD⁺ content (P=0.063; Online Figure III). Mean arterial blood pressure increased at 7 days, with no significant differences among the groups (25.3±4.3, 16.5±18.0, and 34.5±22.3 mm Hg for WT control, Nampt heterozygous control, and SMC-Nampt KO, respectively; P=0.285). However, there was striking aortic hemorrhage in SMC-Nampt KO mice, evident microscopically and grossly. Half of the mice displayed aortic hematomas after 7 days, and this increased to 62.5% on day 28. The hemorrhage typically was within the outer 1 or 2 medial layers and could be extensive, with local elastin breakage (Figure 3A). The hematomas did not rupture into the adventitia, as typically seen in Ang II–infused atherosclerotic mice, but instead could be found at multiple sites and dissecting along the length of the aortic media. In contrast, there was no grossly detectable hemorrhage in control mice subjected to Ang II, although microscopic hemorrhage in the ascending aorta was found in 24% of mice. These small bleeds, also noted in other reports,²³ had resolved by 28 days of Ang II delivery (Online Table II). A similar profile was observed for Nampt heterozygous mice (14% microscopic hemorrhage in the ascending aorta). Transcript analysis of whole aortas revealed that SM-α-actin expression was 22.2% lower in Ang II–infused SMC-Nampt KO than Ang II–infused control mice (P=0.0001), and expression of Col1a1 was 37.4% lower (P=0.002). However, the Mmp2/Timp1 ratio did not increase and in fact reduced somewhat (20.9%; P=0.002). Immunostaining for CD45 showed no differences between control and SMC-KO mice before or after Ang II infusion (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Smooth muscle cell (SMC)-nicotinamide phosphoribosyltransferase (Nampt) knockout (KO) mice are susceptible to aortic wall degeneration and dissection. A, Light micrographs of sections of the descending thoracic aorta of a control (left) and SMC-Nampt KO (middle) mouse after 28 d of infusion with angiotensin II (Ang II), stained with hematoxylin and eosin. Right: An adjacent section of the KO aorta stained with elastin trichrome, showing confinement of hemorrhage to the aortic media. B, Photomicrographs of sections of the ascending aorta stained with hematoxylin and eosin from vehicle-infused and Ang II–infused (28 d) control and SMC-Nampt KO mice. C and D, Graphs depicting cell content of the media cross section after Ang II infusion in ascending (C), *P<0.0001 vs control, and descending thoracic (D) aortas *P=0.002 vs control. E, Sections of descending thoracic aorta from control (left) and SMC-Nampt KO (middle and right) mice after 28 d of infusion with Ang II illustrating focal cell loss. Right: An adjacent section to that in the middle and stained with Movat pentachrome. Arrow depicts proteoglycan-rich cell-poor zone. F, Graph depicts the area of focal zones of SMC loss in the thoracic aorta, *P=0.0003 vs control.

Because of a trend toward greater Ang II–induced blood pressure response in SMC-Nampt KO mice, we determined whether blood pressure elevation, in and of itself, might be responsible for the more striking aortic disruption. For this, control mice were subjected to double infusion of phenylephrine and Ang II, which elevated the mean arterial pressure by 37.2±16.4 mm Hg. This increase was associated with only localized hemorrhage confined to the ascending aorta in 25% of mice (n=8), a profile not different than that of control mice subjected to Ang II. Collectively, these findings reveal that aortas in SMC-Nampt KO mice are prone to Ang II–induced disruption by hematoma and dissection, with minor changes in expression of SM-α-actin and type I collagen chains. Furthermore, whereas control mice adapted over 28 days to the disrupting potential of Ang II, SMC-Nampt KO mice remained vulnerable.

Ang II–Induced Cell Loss in Mice With SMC-Nampt KO Mice

Twenty-eight days of Ang II infusion in control mice resulted in a 1.8-fold increase in medial SMC content in the ascending aorta (Figure 3B and 3C) but no change in SMC content in the descending thoracic aorta, confirming a site-specific response that has been previously noted.²⁴ In contrast, in SMC-Nampt KO mice, medial SMC content in the ascending aorta did not increase in response to Ang II and in the descending thoracic aorta SMC content actually fell by 25.2% (Figure 3D). These aberrant responses occurred despite evidence for residual SMC proliferation in Ang II–infused SMC-Nampt KO mice, as indicated by Ki67 immunostaining (Online Figure IV), supporting a turnover profile skewed toward cell loss. This loss of SMCs was evident both as scattered SMC dropout and cell-free foci of proteoglycan-rich matrix (Figure 3E). Cell-poor zones were occasionally observed in control aortas subjected to Ang II but were 70% larger in aortas of SMC-Nampt KO mice (Figure 3F). Thus Nampt-depleted SMCs failed to adaptively fortify the ascending aorta and repopulate damaged regions in the descending aorta.

SMCs Within the Aorta of SMC-Nampt KO Mice Are Susceptible to Stress-Induced Premature Senescence

To determine whether the aberrant SMC responses to Ang II were caused by induction of apoptosis, aortas were evaluated using TUNEL (terminal deoxynucleotide transferase-mediated dUTP nick end labeling) and active caspase-3 immunostaining. Surprisingly, although Ang II–induced apoptosis was evident in adventitial cells, we did not identify apoptosis signals in the thoracic aortic media (ascending or descending) of either control or Nampt-deficient mice (Online Figure V). We next asked whether SMCs instead developed features of senescence, recognizing that in vitro studies have implicated Nampt in slowing SMC aging.⁷ Intravenous infusion and ex vivo immersion in X-Gal solution²⁵ revealed no senescence-associated ß-galactosidase activity in the aorta of control mice subjected to Ang II. Remarkably, however, in SMC-Nampt–deficient mice, there was senescence-associated ß-galactosidase activity throughout the ascending aorta, arch, and proximal descending thoracic aorta after 7 days of Ang II infusion (Figure 4A). Microscopy established that senescence-associated ß-galactosidase activity was confined to the medial layers of the aorta and particularly the outer medial layers (Figure 4B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Nicotinamide phosphoribosyltransferase (Nampt)–deficient smooth muscle cells (SMCs) within the aorta undergo senescence cellular senescence in response to angiotensin II (Ang II) infusion. A, Whole mount images of ascending, arch, and proximal descending thoracic aortas harvested after 7 d of Ang II infusion and stained for senescence-associated β-galactosidase (SA-ßGal) activity (blue). B, Cryosections of the ascending aorta depicting SA-ßGal activity localized to the aortic media. C, Chart showing proportion of aortas (n=16) with SA-ßGal activity. *P=0.0011 D, Micrographs of ascending aorta subjected to 7 d of vehicle or Ang II infusion and immunostained for p16. Arrows depict p16-postive nuclei. E, Graph depicting prevalence of p16-positive nuclei.

As an additional test for cell senescence, we immunostained aortic sections for the cell cycle inhibitor, p16^INK4a. This revealed a 6.1-fold increase in the abundance of p16-expressing cells in the media of the ascending aorta of Ang II–infused SMC-Nampt KO mice relative to Ang II–infused control mice and a 4.7-fold increase in the thoracic aorta (Figure 4E). Therefore, Nampt deficiency impacts the fate of SMCs in the aorta, whereby Ang II sends them into a state of premature senescence.

Nampt-Deficient SMCs Accumulate Oxidized DNA Lesions and Single-Stranded DNA Breaks

To explore why Nampt-ablated SMCs were prone to senescence, we assessed for DNA integrity. We first immunostained for the oxidized nucleoside, 8-oxo-2′-deoxyguanosine (8-oxodG), recognizing that reactive oxygen species can drive cell senescence.²⁶ The proportion of ascending aorta medial SMCs with oxidized DNA was 3-fold higher in SMC-Nampt KO mice than in control mice. After a 7-day infusion of Ang II, DNA oxidation in SMC-Nampt KO aortas was even more prevalent and was 2.5-fold greater than that in Ang II–infused control mice (Figure 5A and 5B). The oxidized DNA profile for SMCs in the thoracic aorta was similar although less striking (Figure 5B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Nicotinamide phosphoribosyltransferase (Nampt) deficiency and susceptibility to oxidative DNA damage. A, Micrographs of ascending aorta 7 d after infusion with angiotensin II (Ang II), depicting oxidative DNA lesions as indicated by 8-oxo-2′-deoxyguanosine (8-oxodG) immunoreactivity. B, Graph showing the percentage of 8-oxodG–positive cells in the ascending and descending thoracic aortas. C, Fluorescent micrographs of fragmented DNA comet tails from mouse aortic smooth muscle cells (SMCs) subjected to vehicle, 100 μmol/L H₂O₂ (1 h), and 10^–7 mol/L Ang II (24 h). D, Graph depicting mean±SEM SMC comet tail moments, relative to control SMCs incubated with vehicle. *P≤0.0001. E and F, Graphs depicting survival of mouse aortic SMCs, 24 h after incubation with H₂O₂ (*P=0.012, †P=0.014, ‡P=0.0001) or methyl methanesulfonate (MMS; *P=0.009, †P<0.0001). G, Graph depicting the survival of mouse aortic SMCs, 24 h after incubation with 600 μmol/L MMS, with and without the addition of 100 μmol/L nicotinamide riboside (NR), *P≤0.0001.

We next asked whether Nampt-deficient SMCs accumulated overt breaks in the DNA. For this, we established a culture system that enabled us to time the ablation of Nampt and evaluate different DNA break pressures. SMCs were cultured from aortas of Nampt^flox/flox mice expressing Cre recombinase under the control of a tamoxifen-inducible promoter (Cre-ERT2). We then evaluated electrophoretic migration of cleaved DNA out of harvested nucleoids (Comet assay) after delivery of vehicle or hydroxytamoxifen. The latter induced robust Nampt knockdown (Online Figure IV). Interestingly, there was evidence for cleaved DNA in Nampt-ablated SMCs under baseline culture conditions (Figure 5C and 5D). Incubation with H₂O₂ (100 μmol/L, 1 hour) to induce single-strand DNA breaks yielded more striking DNA fragment tails, and these were substantially more prominent in Nampt-ablated SMCs. Incubation with Ang II (10^–7 mol/L, 24 hours) yielded a modest DNA fragment signal that was significantly greater in Nampt-ablated SMCs (Figure 5C and 5D).

To further assess the response to DNA break–inducing agents, SMCs were tracked by video microscopy. This revealed a striking death response in Nampt-depleted SMCs, with cell rounding, anoikis, and cessation of cell membrane activity (Figure 5E; Online Figure VII). Incubating SMCs with methyl methanesulfonate, another single-strand DNA-breaking reagent, led to similarly worse survival for Nampt-ablated SMCs (Figure 5F; Online Figure VII). Notably, incubating SMCs with nicotinamide riboside before methyl methanesulfonate inhibited the death response (Figure 5G). Together, these studies reveal that Nampt protects SMCs from accumulating both oxidized DNA lesions and a toxic burden of single-strand DNA breaks.

Nampt-Deficient SMCs Have Impaired Double-Strand DNA Break Repair

Double-strand DNA breaks are a particularly stressful form of DNA damage and potent driver of cell senescence.²⁷ Using the above culture system, we assessed the potential for accumulating double-strand breaks by subjecting SMCs to irradiation (10 Gy) and immunostaining for phosphorylated histone (γ-H2AX). Interestingly, DNA damage foci were found in low abundance in Nampt-ablated SMCs before irradiation. On irradiation, damage foci accumulated and were significantly more abundant in Nampt-ablated SMCs. This difference was evident after 15 minutes but also after 24 hours, at which time control SMCs showed near-complete resolution of the breaks (Figure 6A and 6B). Incubation of Nampt-ablated SMCs with Ang II (24 hours) also increased double-strand DNA damage foci, and this effect was still evident 24 hours after Ang II washout (Online Figure VIIIa).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Reduced nicotinamide phosphoribosyltransferase (Nampt), poly(ADP-ribose) polymerase-1 (PARP1) inactivation, and double-strand DNA damage in mouse and human aortas. A, Fluorescent micrographs depicting the response of mouse aortic smooth muscle cells (SMCs) to irradiation (10 Gy). Cells were immunostained for γ-H2AX (green) and nuclei were labeled with Hoechst 33258 dye (blue). B, Graph showing the cumulative pixel intensity per cell of γ-H2AX signal. C, Micrographs showing γ-H2AX–positive foci (green) in nuclei in the aortic media of mice subjected to a 7-d infusion of Ang II. Nuclei were counterstained with propidium iodide (red). D, Graph depicting the proportion of γ-H2AX–positive cells in the aortic media of mice subjected to a 7-d infusion of angiotensin II (Ang II), *P=0.044 vs control, ascending aorta; *P=0.022 vs control, thoracic aorta. E, Western blots depicting poly(ADP-ribose) (PAR) moieties, Parp1, and tubulin content in rat aortic SMCs incubated with FK866 (24 h) and subjected to irradiation (50 Gy). F, Immunofluorescence images showing the presence of PAR moieties (green) in control and Nampt-knockout (KO) SMCs subjected to 1 mmol/L H₂O₂ for 15 min. Nuclei were counterstained with DAPI (4′,6-diamidino-2-phenylindole; red). G, Graph depicting cumulative pixel intensity per cell of PAR signal, *P<0.0001 vs KO. H, Prevalence of medial SMCs in aortas with γ-H2AX signal (top, *P=0.047) and intensity of γ-H2AX signal (bottom, *P=0.021). I, Confocal photomicrographs illustrating reciprocal relationship between nicotinamide phosphoribosyltransferase (NAMPT; green) and phosphorylated histone (γ-H2AX) positivity (red) in human ascending aorta. Nuclei were counterstained with DAPI. J, Cellular DNA damage signal, stratified by cellular NAMPT content. *P=0.034 vs upper NAMPT tertile, †P=0.013 vs mid NAMPT tertile.

We next assessed whether this form of DNA damage could be found in vivo. There were no γ-H2AX–positive SMCs in the aortic media of mice subjected to vehicle infusion. Rare (<0.5%) γ-H2AX–positive cells were found in control mice subjected to Ang II infusion (Figure 6C and 6D). In contrast, γ-H2AX–positive SMCs were consistently detected in both ascending (4.0%) and descending (3.8%) thoracic aortas of Ang II–infused SMC-Nampt mice (Figure 6D).

PARP Activity Is Impaired in Nampt-Deficient SMCs and Its Inhibition In Vivo Promotes Aortic SMC DNA Damage and Senescence

PARP1 has roles in repairing multiple types of DNA damage.⁹ To determine whether PARP1 activity was impacted by the disrupted NAD⁺ metabolism, we assessed PAR assembly by Western blot analysis after irradiation of rat aortic SMCs. Irradiation-induced PAR formation was markedly suppressed in SMCs incubated with the Nampt inhibitor, FK866 (100 nmol/L; Figure 6E). The abundance of PARP1 itself did not change on Nampt inhibition. We also immunostained mouse aortic SMCs for PAR. Abundant nuclear PAR signal emerged 15 minutes after control SMCs were exposed to H₂O₂. However, there was barely detectable nuclear PAR in Nampt-ablated SMCs subjected to H₂O₂ (Figure 6F and 6G). In the presence of nicotinamide riboside, PAR assembly in irradiated Nampt-ablated SMCs was restored. Nuclear PAR also accumulated on 24 hours of Ang II exposure. As seen with H₂O₂, the Ang II–induced PAR response was not evident in Nampt-ablated SMCs but was restored in the presence of nicotinamide riboside (Online Figure VIIIb).

In addition, in vivo inhibition of PARP by intrapertoneal injections of olaparib produced SMC defects in control mice subjected to Ang II that were similar to those in Ang II–infused SMC-Nampt KO mice. The proportion of p16-expressing medial SMCs increased by 2.7-fold and 3.7-fold in the ascending and descending thoracic aorta, respectively (P=0.002; P=0.023), and the proportion of SMCs positive for 8-oxodG increased by 2.2-fold and 1.9-fold (P=0.0003; P=0.003; Online Figure IXA and IXB). As well, γ-H2AX–positive SMCs were identified in the media of Ang II–infused mice that received olaparib and in a range similar to that of Ang II–infused SMC-Nampt KO mice (Online Figure IXC).

Collectively, these data indicate the presence of a NAD⁺–PARP1–DNA repair axis in aortic SMCs and the dependence of this cascade on Nampt.

Media of Dilated Human Thoracic Aortas Is Populated by SMCs With DNA Strand Breaks and Low NAMPT

Given the compromised genomic integrity in the mouse models, we asked whether DNA integrity was compromised in SMCs within the human ascending aorta. This was of interest because although double-strand DNA breaks have been found in atherosclerotic lesions²⁸; the media of the nonatherosclerotic aorta is considered a more quiescent vascular setting. Remarkably, γ-H2AX–immunostaining revealed that 25% of normal caliber aortas contained unrepaired DNA strand breaks in the medial layer cells. On average, 2.4% of medial cells displayed DNA damage foci (range 0–10.1%). Even more striking was that 90% of aortopathy samples had evidence of unrepaired DNA strand breaks, with an average of 28.7% SMCs with DNA damage foci (range: 7.3–60.0% of cells, Figure 6H). To more precisely define the relationship between NAMPT expression and unrepaired DNA damage, we double immunolabeled aortas for NAMPT and γ-H2AX (Figure 6I). NAMPT content was then quantified in a total of 524 individual SMCs, which were separated into tertiles based on NAMPT expression. SMCs in the lowest tertile had >5-fold greater DNA damage signal than those in the mid and highest NAMPT tertiles (Figure 6J).

Hypermethylation of the NAMPT Promoter in Dilated Human Thoracic Aortas

Given the linkages between low NAMPT and DNA damage, we assessed what might underlie the low NAMPT in human thoracic aortopathy. Epigenetic control of genes relevant to aortopathy has recently been described.²⁹ To determine whether this might be the case for NAMPT, we first assessed whether the expression profile of NAMPT in human aortic SMCs was preserved between in vivo and culture environments. This revealed that NAMPT transcript abundance in early passage SMCs correlated with NAMPT abundance in the corresponding aortic media (R²=0.72; P=0.004; Figure 7A). Furthermore, NAMPT transcript abundance in SMCs derived from both normal and dilated aortas was stable over 3 serial subcultures, and differences between patient SMCs were maintained (Figure 7B). Bioinformatic assessment (http://genome.ucsc.edu/) predicted a ≈1300-base CpG island surrounding the NAMPT transcriptional start site. Methylation analysis using selective digestion-based polymerase chain reaction revealed little to no methylation of NAMPT promoter DNA from nondilated human aortas. However, 13.0±4.1% of input DNA from dilated aortas was methylated (P=0.044; Figure 7C). Furthermore, in SMCs cultured from the ascending aorta, the level of promoter methylation inversely correlated with NAMPT mRNA abundance in the corresponding SMCs (R²=0.028; P=0.024). This also revealed that the NAMPT promoter in SMCs from patients with dilated aortas was hypermethylated relative to that of SMCs cultured from control aortas (P=0.004; Figure 7D and 7E).

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Hypermethylation of the nicotinamide phosphoribosyltransferase (NAMPT) promoter in the aortic media and cultured smooth muscle cells (SMCs) of patients with dilated ascending aortopathy. A, Relationship between NAMPT transcript abundance in human SMCs cultured from the ascending aorta and NAMPT content in situ in the corresponding aortic media, as assessed by immunostaining. B, NAMPT transcript abundance in early passage human SMCs derived from control and dilated patients measured over early serial cell subcultures, showing stability. C, Methylation of the NAMPT promoter in the media of control, nondilated (n=3) and dilated (n=6) aortas, *P=0.044 vs control. D, Relationship between methylation of the NAMPT promoter in primary early passage (1–4) human SMCs and NAMPT transcript abundance in the respective SMCs. E, Methylation of the NAMPT promoter in early passage human SMCs derived from the ascending aorta of patients with nondilated (n=8) and dilated (n=10) aortas, *P=0.004 vs control.