ARTICLE https://doi.org/10.1038/s41467-021-22933-3 OPEN Aortic disease in Marfan syndrome is caused by overactivation of sGC-PRKG signaling by NO Andrea de la Fuente-Alonso1,2,11, Marta Toral 1,2,11, Alvaro Alfayate1,2,3, María Jesús Ruiz-Rodríguez 1,2, Elena Bonzón-Kulichenko 2,3, Gisela Teixido-Tura 2,4, Sara Martínez-Martínez1,2, María José Méndez-Olivares1,2, Dolores López-Maderuelo1,2, Ileana González-Valdés3, Eusebio Garcia-Izquierdo5, Susana Mingo5, Carlos E. Martín6, Laura Muiño-Mosquera7, Julie De Backer 7, J. Francisco Nistal 2,8, Alberto Forteza6, Arturo Evangelista2,4, Jesús Vázquez 2,3, 1234567890():,; Miguel R. Campanero 2,9,10,12 ✉ & Juan Miguel Redondo 1,2,12 ✉ Thoracic aortic aneurysm, as occurs in Marfan syndrome, is generally asymptomatic until dissection or rupture, requiring surgical intervention as the only available treatment. Here, we show that nitric oxide (NO) signaling dysregulates actin cytoskeleton dynamics in Marfan Syndrome smooth muscle cells and that NO-donors induce Marfan-like aortopathy in wild- type mice, indicating that a marked increase in NO suffices to induce aortopathy. Levels of nitrated proteins are higher in plasma from Marfan patients and mice and in aortic tissue from Marfan mice than in control samples, indicating elevated circulating and tissue NO. Soluble guanylate cyclase and cGMP-dependent protein kinase are both activated in Marfan patients and mice and in wild-type mice treated with NO-donors, as shown by increased plasma cGMP and pVASP-S239 staining in aortic tissue. Marfan aortopathy in mice is reverted by pharmacological inhibition of soluble guanylate cyclase and cGMP-dependent protein kinase and lentiviral-mediated Prkg1 silencing. These findings identify potential bio- markers for monitoring Marfan Syndrome in patients and urge evaluation of cGMP- dependent protein kinase and soluble guanylate cyclase as therapeutic targets. 1 Gene regulation in cardiovascular remodeling and inflammation group, Centro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid, Spain. 2 Centro de Investigación Biomédica en Red de Enfermedades Cardiovasculares (CIBERCV), Madrid, Spain. 3 Cardiovascular Proteomics Laboratoy, CNIC, Madrid, Spain. 4 Servei de Cardiologia, Hospital Vall d’Hebron, Barcelona, Spain. 5 Cardiology Department, Hospital Universitario Puerta de Hierro, Madrid, Spain. 6 Cardiac Surgery Department, Hospital Universitario Puerta de Hierro, Madrid, Spain. 7 Center for Medical Genetics Ghent, Ghent University Hospital, Ghent, Belgium. 8 Cardiovascular Surgery and Department of Physiology and Pharmacology, Hospital Universitario Marqués de Valdecilla, IDIVAL, Facultad de Medicina, Universidad de Cantabria, Santander, Spain. 9 Department of Cancer Biology, Instituto de Investigaciones Biomedicas Alberto Sols, Consejo Superior de Investigaciones Científicas–Universidad Autónoma de Madrid, Madrid, Spain. 10Present address: Centro de Biología Molecular Severo Ochoa, Consejo Superior de Investigaciones Científicas–Universidad Autónoma de Madrid, Madrid, Spain. 11These authors contributed equally: Andrea de la Fuente-Alonso, Marta Toral. 12These authors jointly supervised this work: Miguel R. Campanero, Juan Miguel Redondo. ✉email: mcampanero@cbm.csic.es; jmredondo@cnic.es NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications 1 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 T horacic aortic aneurysm and dissection (TAAD) is a major Results cause of morbidity and mortality in developed countries1. NO signaling pathway activation dysregulates actin cytoskele- TAAD is characterized by progressive vessel dilation ton dynamics in MFS VSMCs. To investigate whether the associated with vascular smooth muscle cell (VSMC) dysfunction NO–sGC–PRKG signaling pathway is activated in MFS, we and destructive extracellular matrix remodeling2,3. TAAD is often measured VASP-S239 phosphorylation (pVASP-S239) as a asymptomatic until the aorta dissects or ruptures. Approximately readout of PRKG activity30 in primary aortic VSMCs from wild- one in four cases of TAAD have a known genetic basis and can be type (WT) mice31,32 and a mouse model of MFS (Fbn1C1039G/+ categorized as either syndromic (showing prominent phenotypic mice)33 cultured in the presence of inhibitors of NOS (L-NAME), features of a systemic connective tissue disorder such as Marfan sGC (ODQ), and PRKG (KT5823) (Fig. 1a). As pathway activa- syndrome [MFS]) or nonsyndromic, with an essentially vascular tion controls, we also treated WT cells with the cGMP analog 8- phenotype3. TAAD and rupture account for >90% of MFS patient Br-cGMP or the NO-donor DetaNONOate (DetaNO). These deaths4,5. MFS is an inherited autosomal dominant disease caused agents induced pVASP-S239, and this event was prevented by the by pathogenic variants of the fibrillin-1-encoding gene FBN1. corresponding inhibitors ODQ and KT5823 (Fig. 1b, c). Fibrillin-1 is a major component of extracellular microfibrils, Untreated MFS VSMCs had higher pVASP-S239 levels than providing a scaffold for elastic-fiber formation and maturation6. WT cells (Fig. 1b, c), and these levels were sharply decreased by Mutant fibrillin-1 disrupts microfibril formation, leading to treatment with ODQ or KT5823 (Fig. 1b, c), strongly suggesting medial degeneration, which destabilizes the aortic wall, rendering that the NO–sGC–PRKG signaling pathway is activated in MFS the aorta vulnerable to hemodynamic injury7,8. Medial degen- VSMCs. eration is characterized by poor alignment of elastin filaments, VASP-S239 phosphorylation impairs actin filament disorganization of lamellar units, VSMC death, and proteoglycan formation34,35, and we therefore investigated whether pVASP- accumulation9. Proteoglycans play essential roles in the pre- S239 regulation by the NO–sGC–PRKG signaling pathway servation of aortic structure and function by regulating elastic negatively correlated with actin filament formation. Phalloidin fiber assembly and smooth muscle cell proliferation10–12. staining showed a sharp decrease of filamentous actin (F-actin) Current therapy for the cardiovascular complications of MFS accumulation in WT VSMCs treated with DetaNO or 8-Br- consists of strict follow-up, life-style advice, medical treatment to cGMP, correlating with low F-actin in untreated MFS cells slow the rate of aortic root dilation, and surgery to prevent dis- (Fig. 2a, b). Moreover, PRKG inhibition by KT5823 in MFS section or rupture13. Pharmacological therapy is usually based on VSMCs markedly increased F-actin formation to levels similar to ß-adrenoceptor blockade14. However, this treatment neither halts those found in control cells (Fig. 2b), supporting the notion that abnormal aortic growth nor prevents aortic dissection or death14. the NO signaling pathway is critical for actin cytoskeleton Moreover, some studies suggest that ß-adrenergic blockers do not dynamics regulation in MFS. even slow the rate of aortic growth in MFS15,16. There is therefore Regulation of actin cytoskeleton dynamics is critical for a need for effective pharmacological strategies to manage TAAD effective cell contraction; moreover, the expression of contractile in MFS patients. protein markers, including alpha-smooth muscle actin (Acta2), MFS is accompanied by impaired aortic contractility17, caused smooth muscle protein 22 alpha (Tagln2), and calponin-1 by as yet unknown molecular mechanisms. In small arteries, (Cnn1), is increased in the aorta of MFS mice relative to WT VSMC relaxation and vasodilatation are induced by nitric oxide mice36. Given these observations, we assessed the role of the (NO), produced in endothelial cells by constitutively expressed NO–sGC–PRKG pathway in regulating the expression of these endothelial NO synthase (eNOS; also called NOS3). NO can also markers. RT-qPCR analysis showed a substantial increase in be produced by constitutively expressed NOS of neuronal origin Acta2, Cnn1, and Tagln2 mRNA in WT VSMCs treated with (nNOS; also called NOS1) or by inducible NOS (iNOS; also called DetaNO or 8-Br-cGMP (Fig. 2c) and also in MFS VSMCs NOS2)18. NOS2 expression is induced in MFS patients and in a (Fig. 2d). Consistent with a role for the NO pathway in regulating mouse model of MFS, and TAAD is reversed in the mouse model the expression of these markers, pharmacological PRKG by pharmacological NOS2 inhibitors, raising the possibility that inhibition decreased their expression to normal levels (Fig. 2d). blocking NOS2 activity could be a promising treatment for These results indicate that some of the enhanced SMC markers TAAD19. However, the mechanisms by which NOS2 contributes characteristic of MFS VSMCs can be reverted by NO–sGC–PRKG to TAAD in MFS remain unclear. Increased NOS2-derived NO targeting. levels stimulate soluble guanylate cyclase (sGC) to generate MFS smooth muscle cells harboring the FBN1C1242Y patho- cGMP, which in turn activates cGMP-dependent protein genic variant show defective deposition of fibrillin-137. However, kinase G (PRKG), which modulates the contractility of resistance we found no notable effect on extracellular fibrillin-1 deposition vessels by regulating actin filament and myosin dynamics20. in WT VSMCs treated with with 8-Br-cGMP (Supplementary NO also regulates numerous physiological processes in a Fig. 1). cGMP-independent manner, including mechanisms based on Together these results strongly suggest that overactivation of S-nitrosylation21,22 and, in the presence or excess reactive oxygen the NO–sGC–PRKG pathway in MFS VSMCs switches their species (ROS), protein tyrosine nitration23. phenotype by dysregulating the contractile machinery, including In response to limited availability of the substrate L-arginine or actin cytoskeleton dynamics, without affecting microfibril the cofactor tetrahydrobiopterin (BH4)24–26, or high ROS formation. levels27, all NOS enzymes generate superoxide anion instead of NO, resulting in oxidative stress. As ROS levels are elevated in the MFS aorta17,28, NOS2 might be responsible for superoxide anion Supraphysiological NO levels induce an MFS-like aortic generation in this scenario. Moreover, ROS activate PRKG by pathology. VSMCs from MFS patients and the Fbn1C1039G/+ direct oxidation29, raising questions about whether TAAD gen- mouse model (MFS mice) express high levels of NOS219, a NOS eration in MFS is mediated by NOS2-derived superoxide or NO enzyme that produces large amounts of NO38. We therefore and whether cGMP or PRKG act as regulators in this process. investigated whether supraphysiological NO levels affect aortic Here, we show that the NO–sGC–PRKG signaling pathway is homeostasis in healthy mice. As an NO source, we used iso- activated in MFS mice and MFS patients, and demonstrate that sorbide mononitrate (ISMN), a long-lasting nitrate that is meta- this pathway mediates aortopathy in a mouse model of MFS. bolized to NO within VSMCs and induces dilation of systemic 2 NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 ARTICLE Fig. 1 Pharmacological inhibition of signaling components of the NO–sGC–PRKG pathway decreases pVASP-S239 induction in VSMCs from MFS mice. a NO signaling components and the targets of pharmacological stimuli (green) and inhibitors (red). b Representative images of pVASP-S239 immunofluorescence (red) and DAPI-stained nuclei (blue). c Quantification of pVASP-S239 immunofluorescence in WT and MFS VSMCs treated with 300 µM L-NAME, 10 µM ODQ, or 1 µM KT5823 for 1 h before stimulation for 5 minutes with 100 µM 8-Br-cGMP or 100 µM DetaNO, as indicated. b, c Five independent cell batches were used for all conditions except for MFS VSMCs (n = 4). Scale bar, 50 μm. Data are shown relative to untreated WT cells as mean ± s.e.m. Each data point denotes the mean value from an independent experiment. Differences were analyzed by one‐way ANOVA with Dunnett’s post-hoc test (p-values are shown). Source data are provided in the Source Data file. NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications 3 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 Fig. 2 The NO–sGC–PRKG pathway modulates the contractility phenotype of VSMCs. a, b Representative images of F-actin staining (red) and DAPI- stained nuclei (blue) and F-actin quantification in a WT VSMCs treated as indicated for 5 min (n = 4 independent cell batches per condition) and b WT and MFS VSMCs untreated or treated with KT5823 for 24 h (n = 5 independent cell batches per condition). Scale bar, 50 μm. Data are shown relative to untreated WT cells as mean ± s.e.m. Each data point denotes the mean value from an independent experiment. Differences were analyzed by one‐way ANOVA with Tukey’s post-hoc test (p-values are shown). c, d RT-qPCR analysis of Acta2, Cnn1, and Tagln2 mRNA expression in c WT VSMCs treated as indicated for 4 h and d WT or MFS VSMCs treated as indicated for 24 h (n = 4 independent cell batches per group in c and n = 5 independent cell batches per group in d). mRNA amounts were normalized to Gapdh expression (mean ± s.e.m.). Each data point denotes the mean value from an independent experiment. Differences were analyzed by one‐way ANOVA followed by Dunnett’s post-hoc test (p-values are shown). Source data are provided in the Source Data file. and coronary vascular beds39. We performed in vivo longitudinal induced a dose-dependent decrease in systolic blood pressure dose-response studies to characterize the effects of infusing ISMN (BP) (Fig. 3b) and dilation of both the ascending aorta (AsAo) for 7 days on the aortic phenotype of WT C57BL/6 mice (Fig. 3a). and the abdominal aorta (AbAo) (Fig. 3c, d). Aortic diameters at Since there were no previous reports on the effect of ISMN on the highest dose were similar to those in MFS mice (Fig. 3e). aortic phenotype, and the normal clinical dose of ISMN used to ISMN doses above 50 mg/kg/day did not induce larger aortic treat chronic stable angina pectoris ranges from 0.5 to 4 mg/kg/ dilations or BP drops (Supplementary Fig. 2), and this dose was day, we tested several doses from 1 to 50 mg/kg/day. ISMN therefore used for further experiments. 4 NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 ARTICLE To determine if other NO donors had a similar effect on BP proteoglycan accumulation and elastic-fiber fragmentation and and aortic dilation, we treated WT mice with DetaNO. Unlike disorganization as early as 7 days after treatment initiation, as ISMN, DetaNO is unstable at 37 °C, and we therefore used shown by Alcian Blue and modified Verhoeff elastic-Van Gieson osmotic minipumps to infuse it for just 2 days. Like ISMN, (EVG) staining, respectively (Fig. 3f, g). After 28 days, these DetaNO decreased systolic BP (Supplementary Fig. 3a) and features were more pronounced and were comparable to those in induced marked AsAo and AbAo dilations (Supplementary age-matched MFS mice (Fig. 3f, g). Similar elastic-fiber alterations Fig. 3b-c). were found upon treatment of WT mice with DetaNO FBN1 pathogenic variants cause not only aortic dilation, but (Supplementary Fig. 3d-3e). These results thus indicate that also medial degeneration characterized by elastic-fiber fragmen- supraphysiological levels of NO are sufficient to induce MFS-like tation and disarray and proteoglycan accumulation in the aorta, a aortopathy in WT mice. feature that might predispose to dissection10. Histological analysis In humans, sustained high plasma nitrate concentrations of the AsAo40 in ISMN-infused WT mice revealed both induce a clinically relevant decay of the vasodilatory effect, NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications 5 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 Fig. 3 ISMN induces an MFS-like aortopathy in WT mice. a Experimental design. 12-week-old C57BL/6 mice were treated for 7 days (d) with the NO- donor ISMN at 1, 2.5, 5, 10, 25, or 50 mg/kg/day by osmotic minipump infusion. Mice were monitored by Eco-BP (ultrasound and BP analysis) at the indicated times (empty triangles). b Systolic BP during ISMN infusion. c Maximal AsAo diameter (left) and AbAo diameter (right) during ISMN infusion. Data in b and c are mean ± s.e.m; n = 5 per treated group, n = 3 for control group. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 (versus control at each time point) by repeated-measurements two-way ANOVA with Bonferroni’s post-hoc test. d Representative ultrasound images (orange dashed lines delineate the lumen boundary and yellow dashed lines mark the lumen diameter) and e quantification of end-of-experiment maximal AsAo and AbAo diameter in untreated WT mice (−) (n = 8), MFS mice (n = 7), and WT mice treated with 50 mg/kg/day ISMN for 7 d (n = 5). Each data point denotes an individual mouse, whereas histograms show mean ± s.e.m. f Representative staining with hematoxylin and eosin (HE) (left), Alcian blue (middle), and elastic van Gieson (EVG) (right) in the AsAo from 9 to 12-week-old untreated WT mice (−), 6 MFS mice, and WT mice treated with 50 mg/kg/day ISMN for 7 d (n = 12 mice) or 28 d (n = 6 mice). Yellow arrowheads indicate elastin breaks. Scale bar, 50 μm. g Quantification of elastin breaks in AsAo sections from 9 to 12-week-old untreated WT mice (−), 6 MFS mice, and WT mice treated with 50 mg/kg/day ISMN for 7 d (n = 12 mice) or 28 d (n = 6 mice). Each data point denotes an individual mouse, whereas histograms denote mean ± s.e.m. e, g Differences were analyzed by one-way ANOVA with Dunnett’s post-hoc test (p-values are shown). Source data are provided in the Source Data file. known as nitrate tolerance41. To investigate if continuous ISMN DetaNO-treated WT mice. Mass spectrometry analysis of plasma delivery in mice loses efficacy over the long term, we performed a protein digests quantified 43 nitrated peptide species identified in longitudinal study in which ISMN (50 mg/kg/day) was infused for more than 1 MS/MS spectrum from 19 nitrated proteins that 28 days (Supplementary Fig. 4a). In line with the results shown in included 47 nitrated sites (NS) (Supplementary Data 1 and 2). Of Fig. 3, ISMN sharply decreased systolic BP and increased AsAo the NS, 88% correspond to nitro-Tyr and 12% to nitro-Trp. and AbAo diameters after 1 day of infusion; moreover, these The cumulative distribution of nitro-protein quantifications in effects were sustained over the remaining days of treatment plasma from NO-donor-treated mice was significantly shifted (Supplementary Fig. 4b-4c), ruling out a tolerance phenomenon. toward positive values in relation to other plasma proteins (Fig. 6a), indicating that NO treatment produced a generalized increase in nitrated protein abundance. In contrast, the cumula- The sGC–PRKG signaling pathway is activated in MFS mice tive distribution of all remaining plasma proteins closely and patients. We next investigated whether the elevated NOS2 followed the expected null-hypothesis distribution (Fig. 6a), expression found in MFS patients and mice results in activation demonstrating the accuracy of the model and that the nitration of the sGC–PRKG signaling pathway. sGC generates cGMP, and profile was not biased due to the quantitative analysis applied. intracellular cGMP pools are in a dynamic steady-state relation- Interestingly, the relative abundance of nitrated proteins was also ship with plasma cGMP levels42,43. We found substantially ele- elevated in plasma from MFS mice (Fig. 6a), demonstrating vated plasma cGMP in both MFS mice and ISMN-treated WT an increased nitration in these mice that can be clearly detected mice (Fig. 4a), strongly suggesting that sGC is indeed activated in in plasma. MFS mice. Similar plasma cGMP increases were found upon To determine if increased aortic NO levels in MFS patients is DetaNO treatment (Fig. 4b). As a readout of PRKG activity, we also indicated by elevated plasma-protein nitration, we performed determined pVASP-S239 levels in aortic tissue. pVASP-S239 a quantitative proteomics analysis in plasma samples from 30 levels were very high in the aortas of MFS mice or WT mice healthy donors and 23 MFS patients, identifying 40 nitrated treated with ISMN for 7 or 28 days (Fig. 4c, d and Supplementary peptide species from 18 nitrated proteins, many of which were Fig. 5). The increase in plasma cGMP and aortic pVASP-S239 also found in the analysis of mouse plasma (Supplementary was not due to increased sGC or Prkg mRNA or protein Data 3 and 4). Among 43 NS identified in these peptides, 66% expression in MFS (Supplementary Fig. 6), suggesting that sGC correspond to nitro-Tyr and 34% to nitro-Trp. Reproducing the and Prkg are activated in MFS by post-translational mechanisms. results obtained in mice, plasma from MFS patients showed a To determine if the NO–sGC–PRKG pathway is activated in generalized elevation in the abundance of nitrated proteins MFS patients, we measured serum or plasma cGMP and PRKG relative to controls (Fig. 6b), suggesting that a highly nitrated activity in aortic tissue from MFS patients. Plasma or serum plasma profile may be a signature of MFS and strongly cGMP from three independent MFS patient cohorts was supporting the notion that NO levels are increased in MFS. To markedly higher than in healthy donors (Fig. 5a). PRKG activity provide a set of candidate nitrated peptides as biomarkers for was also higher in aortic sections from two MFS patient cohorts clinical diagnosis or prognosis, we selected the nitrated peptides than in sections from multiorgan transplant donors. Immuno- that were significantly upregulated in MFS patients (Fig. 6c). The histochemistry and immunofluorescence revealed higher pVASP- quantitative values of these peptides were combined to obtain a S239 levels in MFS samples (Fig. 5b, c and Supplementary Fig. 7). nitrated plasma index (NPI), which provides a measure of the Together, these data strongly suggest that the sGC–PRKG increase in nitration. Mean NPI was clearly increased in the MFS pathway is activated in MFS patients. population, with most MFS patients remaining above the mean of the healthy controls (Fig. 6d). MFS mice and patients have elevated levels of protein nitration We additionally searched for aortic tissue proteins whose in aortic tissue and plasma. The lipophilic properties of NO nitration levels could be affected by NO–sGC–PRKG pathway allow it to diffuse from the vessel wall to the vessel lumen23,44, activation. Quantitative proteomics analysis in pooled aorta where it can potentially trigger NO-derived modifications of samples from 12 WT (n = 6 pools) and 12 MFS (n = 6 pools) plasma proteins. These include Cys S-nitrosylation, a labile mice identified 124 nitrated peptide species from 104 nitro- modification that cannot be directly detected unless a stable proteins, which included 147 NS (Supplementary Data 5). Of derivative is generated45, and nitration, a stable modification that these NS, 75% correspond to nitro-Tyr and 25% to nitro-Trp. takes place under pro-oxidant conditions46. We carried out a MFS mice showed a consistent and significant upregulation of 24 high-throughput proteomics analysis to explore the relative aortic nitro-proteins (Fig. 6e and Supplementary Fig. 8). This nitration levels of plasma proteins in seven MFS mice and seven analysis revealed a sharp increase of Acta2 nitration as well as WT littermates. As a positive control, we also included 6 substantial increases in the nitration of seven additional 6 NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 ARTICLE Fig. 4 The sGC–PRKG pathway is activated in MFS and in NO-donor-treated mice. a Plasma cGMP in 12–13-week-old untreated WT mice (−) (n = 14), MFS mice (n = 13), and WT mice treated with 50 mg/kg/day ISMN for 7 d (n = 20). b Plasma cGMP in untreated WT mice (Control) (n = 4) and WT mice treated with 5 mg/kg/day DetaNO for 2 d (n = 10). c Representative images of pVASP-S239 immunofluorescence (red), elastin autofluorescence (green), and DAPI-stained nuclei (blue). d Quantification of pVASP-S239 immunofluorescence in aortic sections from 12-week-old untreated WT mice (−) (n = 9), MFS mice (n = 9), and WT mice treated with 50 mg/kg/day ISMN for 7 d (n = 10) or 28 d (n = 5). IgG staining served as a negative control. Scale bar, 50 µm. a, b, d Data are shown relative to untreated WT mice as mean ± s.e.m. Data points denote individual mice. Differences were analyzed by a, d one-way ANOVA with Dunnett’s post-hoc test or b unpaired two-tailed t-test (p-values are shown). Source data are provided in the Source Data file. cytoskeletal and extracellular matrix proteins that might play a Daily i.p. treatment of MFS mice with the PRKG inhibitor role in contractility regulation (Fig. 6f). KT5823 (2 µmol/kg/day) (Fig. 7g) had similar effects on Prkg activity (Fig. 7h, i), AsAo and AbAo dilation, and BP (Fig. 7j, k), suggesting that sGC and PRKG mediate aortic disease in MFS. The NO–sGC–PRKG pathway mediates aortopathy in MFS. However, no reversion of medial degeneration was observed after Activation of the NO–sGC–PRKG signaling pathway in MFS 7 days of KT5823 treatment (Supplementary Fig. 11a), suggesting patients and mice suggested its involvement in MFS aortopathy. that tissue regeneration might require a longer period of Prkg However, increased plasma-protein nitration in MFS patients and suppression. mice is also consistent with a cGMP-independent contribution of To test this hypothesis, we silenced Prkg expression in the aortas NO to aortic disease. To ascertain whether sGC–PRKG activation of adult MFS mice. There are 2 Prkg isoforms: Prkg1, which is plays a causal role in MFS aortopathy, we treated MFS mice with ubiquitously expressed but particularly abundant in vascular pharmacological inhibitors of these enzymes. Daily intraper- smooth muscle cells; and Prkg2, which is mainly expressed in itoneal (i.p.) administration of the sGC inhibitor ODQ (20 mg/ non-vascular tissues47. We therefore screened candidate shRNAs kg/day) to 14-week-old MFS mice (Fig. 7a) markedly decreased specific for Prkg1 in cultured VSMCs, identifying shPrkg1-A and plasma cGMP (Fig. 7b), inhibited Prkg activity determined by shPrkg1-B as having high silencing capacity (Fig. 8a). The shRNA- pVASP-S239 aortic staining (Fig. 7c, d), and completely reversed encoding lentivirus also encoded green fluorescent protein (GFP) to AsAo and AbAo dilation after 7 days without inducing arterial facilitate assessment of transduction efficiency48,49. In vitro hypertension (Fig. 7e, f). Histological analysis of AsAo cross- transduction of MFS VSMCs with control shRNA (shScr), sections after 21 days of treatment showed that pharmacological shPrkg1-A, or shPrkg1-B lentivirus yielded similar GFP mRNA sGC inhibition partially reverted elastic-fiber fragmentation and expression levels, and Prkg1 expression was specifically silenced by decreased aortic wall thickness (Supplementary Fig. 11). shPrkg1-A and shPrkg1-B lentiviruses (Supplementary Fig. 12a). NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications 7 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 Fig. 5 The sGC–PRKG pathway is activated in aortas of MFS patients. a Plasma (Cohorts 1 and 2) and serum (Cohort 3) cGMP in three independent cohorts including 24 healthy donors and 38 MFS patients (Cohort 1); 11 healthy donors and 8 MFS patients (Cohort 2); and 11 healthy donors and 33 MFS patients (Cohort 3). Data are mean ± s.e.m. Each data point denotes an individual. b Representative medial layer images and quantification of pVASP-S239 immunohistochemistry in aortic cross-sections of human samples from 3 control donors and 9 MFS patients. Scale bar, 50 μm. IgG staining served as a negative control. Data are shown relative to healthy donors as mean ± s.e.m. Each data point denotes an individual. c Representative medial layer images and quantification of pVASP-S239 immunofluorescence (red) and DAPI-stained nuclei (blue) in sections from 5 control donors and 11 MFS patients. Scale bars, 50 μm. IgG staining served as a negative control. Data are shown relative to healthy donors as mean ± s.e.m. Each data point denotes an individual. a–c Differences were analyzed by unpaired two-tailed t-test with Welch’s correction (p-values are shown). Source data are provided in the Source Data file. Consistent with the results obtained upon pharmacological Prkg to generate extracellular fibrillin-1 fibers (Supplementary Fig. 12c). inhibition, Prkg1 knockdown decreased the expression of the Intrajugular delivery of lentivirus encoding shPrkg1-A or shPrkg1-B contractility markers Acta2, Tagln2, and Cnn1 in MFS cells to into MFS mice (Fig. 8b) yielded efficient transduction of all aortic normal levels (Supplementary Fig. 12a) while increasing actin fiber layers, determined by GFP immunostaining of aortic sections formation to normal levels (Supplementary Fig. 12b). Prkg1 4 weeks after lentivirus delivery (Fig. 8c). Prkg1 protein expression knockdown did not, however, restore the capacity of MFS VSMCs was almost undetectable in aortic samples from these mice (Fig. 8c, 8 NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 ARTICLE d). Consistent with the results obtained with the PRKG inhibitor, Discussion Prkg1 silencing in the aortas of MFS mice markedly decreased Our results demonstrate that the NO–sGC–PRKG signaling pVASP-S239 aortic staining (Fig. 8e, f) and reversed AsAo and pathway mediates aortopathy in a mouse model of MFS and is AbAo dilation (Fig. 8g). Prkg1 knockdown did not raise BP activated in MFS mice and MFS patients. These findings identify above normotensive values in untreated littermate controls (Fig. 8h). potential targets for intervention in human MFS, as well as cir- Histological analysis of aortic cross-sections showed that culating activation markers of this pathway that might be useful Prkg1 silencing for 28 days led to an almost complete reversal of for MFS disease monitoring and clinical follow-up. medial degeneration, determined by the thickness of the aortic wall Our previous work in MFS mice showed that NOS2, whose (Fig. 8i, j) and the regression of elastic-fiber fragmentation expression is induced in VSMCs from MFS mice and patients, is and disarray (Fig. 8i, k). Together, these results urge the an important mediator of medial degeneration and aortic dila- evaluation of sGC and PRKG as potential targets for therapeutic tion, key features of MFS aortopathy that are efficiently reverted intervention. by NOS2 inhibition19. NOS2 activity generates much larger NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications 9 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 Fig. 6 Quantitative proteomics shows that increased protein nitration is a signature of MFS in mice and humans. Quantitative proteomics analysis of plasma samples from a untreated WT mice (n = 7), MFS mice (n = 7), and DetaNO-treated WT mice (n = 6) and b MFS patients (n = 23) and healthy donors (n = 30). Data were analyzed using the SanXoT package. Theoretical, normal distribution; Nitro-proteins, cumulative distributions of indicated nitrated protein values corrected to the values for nonmodified peptides from the same protein; All-proteins, cumulative distributions of nonmodified peptides. Data are expressed in standardized log2-ratios (Zq) relative to a untreated WT or b healthy donors (Control). Differences were analyzed by two- tailed Kolmogorov-Smirnov test (p-values are shown; n.s., not significant). Representative annotated fragmentation spectra for 2 mouse nitro-peptides are provided in Supplementary Fig. 9. All MS/MS spectra of identified tryptic nitro-peptides are provided in the Data availability section. c List of nitrated peptides showing significantly higher abundance in MFS patients. The heatmap shows standardized log2-ratios (Zpq) and statistical significance calculated using limma analysis. The annotated fragmentation spectra of these 7 nitro-peptides are provided in Supplementary Fig. 10. d Nitrated plasma index, defined as the weighted mean of the nitro-peptides listed in c, for 23 MFS patients and 30 healthy donors (Control). Each data point denotes an individual, boxes enclose the interquartile range (IQR), the line in the box shows the centre (median), and whiskers extend 1.5 times above and below the IQR. Differences were analyzed by unpaired two-tailed Student’s t-test. e Quantitative proteomics analysis of pooled aortic samples from 12 untreated WT mice (n = 6 pools) and 12 MFS mice (n = 6 pools). Nitrated protein distributions are shown as in a and b. Differences were analyzed by two-tailed Kolmogorov- Smirnov test. f Barplots showing standardized log2-ratios (Zq) integrated by condition (untreated WT and MFS) for significantly upregulated nitro-proteins related to the cytoskeleton: thrombospondin type-1 domain-containing protein 4 (THSD4), coronin 1-C (CORO1C), latent-transforming growth factor beta 2 and 4 (LTBP2, LTBP4), actin aortic smooth muscle (ACTA2), tensin 3 (TNS3), dynein heavy chain 10 (DNAH10), and tubulin beta 4 chain (TUBB4B). Differences were analyzed by limma statistical analysis (exact p-values are shown). Source data are provided in the Source Data file. amounts of NO than NOS3 or NOS138, and therefore its signaling pathways44,54. NOS2-mediated ROS overproduction induction in VSMCs might lead to the sustained production of can trigger a positive feedback loop via oxidation of the NOS high NO levels. This may in turn activate a number of signaling cofactor BH455, further uncoupling NOS2 and increasing pathways, with the major candidate being the sGC–PRKG path- superoxide production. In this scenario, high NO and superoxide way. However, NO can also signal through sGC–PRKG-inde- anion levels would generate peroxynitrite, leading to nitrosative pendent pathways, including those involving nitrosative stress44. stress accompanied by Tyr/Trp-nitration and S-nytrosylation of Our results indicate that supraphysiological NO levels, gener- proteins and oxidative damage to other biomolecules18. Oxidative ated by administration of ISMN or DetaNO, suffice to trigger an stress has been linked to syndromic and nonsyndromic familial MFS-like aortopathy in WT mice. These NO donors induce dose- aortic diseases28,56; moreover, in MFS mice redox stress is asso- dependent aortic dilation, elastic-fiber fragmentation, and a drop ciated with NOX4 (NADPH oxidase 4), which is upregulated in in BP, as well as recapitulating the sGC–PRKG signaling activa- the aortas of MFS patients and whose deficiency protects MFS tion seen in MFS. The association of increased NO–sGC–PRKG mice from elastic-fiber fragmentation and aortic dilation28. signaling with aortopathy warrants consideration of the potential Because NO diffuses from the vascular wall to the lumen, the detrimental effects of drugs in clinical use that elicit chronic sti- rise in nitrated proteins in MFS mice and patients indicates mulation of this pathway. This applies not only to NO donors, increased NO production in MFS and suggests that, despite the but also to drugs that impair cGMP degradation, including presence of high ROS levels, NOS2 generates high amounts of phosphodiesterase type 5 (PDE5) inhibitors. NO donors have NO in MFS mice and patients. These data are also consistent with been extensively used to treat angina pectoris and heart failure, a recent report showing significant elevation of NO-derived and PDE5 inhibitors to treat pulmonary hypertension and erectile metabolites such as nitrites in the plasma of MFS patients57. dysfunction, with, to our knowledge, no reported pathological Nitration on Tyr and Trp is considered a stable NO-derived consequences related to aortic homeostasis or dilation. Based on post-translational modification58, and only a few studies have accumulated clinical experience, we do not anticipate that reported these modifications in plasma59–61. Given the low patients treated with these agents will develop harmful aortic abundance of nitration, its detection in complex samples such as dilation. Nevertheless, our results with high NO-donor doses in plasma is challenging with currently available technology, and we mice suggest that these patients may be at risk of mild aortic were able to detect Tyr/Trp-nitration only in some abundant dilation. Our results therefore indicate the need of caution with proteins. Nevertheless, our data show that nitration is markedly drugs that chronically activate the NO–sGC–PRKG signaling upregulated in MFS mice and that MFS mice and NO-treated WT pathway, since they may be detrimental to aortic homeostasis. mice have similar nitration profiles, further supporting the notion These drugs include advanced sGC stimulators such as Vericiguat that excessive NO production underlies MFS. Moreover, given the (BAY 1021189) and Riociguat (BAY 63–2521), the latter having upregulation of plasma-protein nitration in MFS patients, our received FDA approval for pulmonary hypertension and heart data also suggest that a highly nitrated plasma profile may be a failure50,51, as well as novel short-acting sGC activators that lack signature of MFS, warranting future assessment of this post- hypotensive side effects, such as TY-55002, a candidate for the translational modification as a biomarker of the disease. treatment of patients with acute decompensated heart failure52. Other markers of MFS identified in our study include cGMP, Seemingly in conflict with our results, increased NO produc- whose plasma levels are substantially upregulated in MFS mice tion by constitutively active Nos3 has been reported to exert a and three MFS patient cohorts. Circulating cGMP appears to be a beneficial effect on the aorta in MFS mice53. It should be noted, good indicator of sGC activity that can be easily determined in however, that NO levels generated by Nos3 are almost 1000-fold patient plasma or serum. Indeed, elevated cGMP has been lower than those generated by Nos238. Given that NOS2 inhibi- reported in other diseases, including congestive heart failure and tors regress aortic disease in MFS mice19, our data suggest that several cancers62–64. the overproduction of NO by NOS2 underlies MFS aortopahy Given that pVASP-S239 appears to be a good marker of PRKG through sGC–PRKG pathway activation. It therefore seems that activity30, its increase in the aortas of MFS patients strongly Nos2 activity might overcome the potential positive effects of suggests enhanced PRKG activity in this tissue. This idea is Nos3 in MFS aorta. supported by the marked drop in pVASP-S239 in MFS VSMCs When uncoupled, NOS2 can lead to overproduction of treated with the PRKG inhibitor KT5823 and in the aortas of superoxide anion and peroxynitrite, which activate many MFS mice treated with ODQ, KT5823, or lentivirus encoding 10 NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 ARTICLE Prkg1-shRNA. These findings suggest that analysis of pVASP- Trp-nitrated proteins could help to monitor the course of the S239 in TAAD tissue samples obtained during elective or emer- disease or the efficacy of future treatments in clinical trials; gency aortic root surgery could be used to assess the extent of however, future studies involving a larger clinical sample will be PRKG activation in these tissues. needed to support the prognostic potential of plasma cGMP and Aortic dissection is the major cause of morbidity and mortality protein nitration. It is important to interpret changes in the in MFS. Unfortunately, dissection often occurs when the aortic identified markers with caution because our comparison of cir- diameter is below the recommended threshold for elective pro- culating markers was limited to MFS patients and healthy donors. phylactic surgery65,66. This is also the case for type B dissections Upregulated levels of the markers we describe here are likely a in MFS patients with prior prophylactic aortic surgery67. The feature of other diseases involving NO pathway activation, such identification of biomarkers that predict the risk of dissection as infection, inflammation, allergy, or sepsis68,69. Although the would help in surgical decision-making. Plasma cGMP and Tyr/ identity of nitrated plasma proteins may differ depending on the NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications 11 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 Fig. 7 Pharmacological inhibition of sGC or PRKG reverts aortopathy in Marfan syndrome. a Experimental design. 14-week-old MFS mice were treated daily for 21 days with 20mg/kg/day ODQ. Longitudinal ultrasound and BP analysis (Eco-BP) was performed at the indicated times (empty triangles). b Plasma cGMP at 21 d (n = 8 WT mice, n = 10 MFS mice, and n = 9 ODQ-treated MFS mice). c Representative pVASP-S239 immunofluorescence (red) in mouse aortic sections. Yellow dashed lines delineate the lumen boundary. IgG staining served as a negative control. Scale bar, 50 µm. d Quantification of pVASP-S239 immunofluorescence in aortic sections from 10 untreated WT mice (−), 10 MFS mice, and 9 MFS mice treated with 20 mg/kg/day ODQ for 21 d. IgG staining served as a negative control. Scale bar, 50 µm. b, d Data are mean ± s.e.m. Each data point denotes an individual mouse. Differences were analyzed by one-way ANOVA with Tukey’s post-hoc test (p-values are shown). e, f Systolic BP (e) and maximal AsAo and AbAo (f) diameter at the indicated times (n = 5 mice each for untreated and ODQ-treated MFS groups; n = 7 mice for WT group). Data are mean ± s.e.m. **P < 0.01, ***P < 0.001, and ****P < 0.0001 (versus WT mice); ##P < 0.01, ###P < 0.001, and ####P < 0.0001 (versus untreated MFS) by repeated-measurements two-way ANOVA with Tukey’s post-hoc test. g Experimental design. 14-week-old MFS mice were treated daily for 7 days with 2 µmol/kg/day KT5823 and monitored for aortic dilation and BP before treatment and 3 d and 7 d post-treatment. h Representative pVASP-S239 immunofluorescence (red) in mouse aortic sections. Yellow dashed lines delineate the lumen boundary. IgG staining served as a negative control. Scale bar, 50 µm. i Quantification of pVASP- S239 immunofluorescence in aortic sections from untreated WT mice (−) (n = 5), MFS mice (n = 5), and MFS mice treated with 2 µmol /kg/day KT5823 for 7 d (n = 7). Data are shown relative to untreated WT mice as mean ± s.e.m. Each data point denotes an individual mouse. Differences were analyzed by one-way ANOVA with Tukey’s post-hoc test (p-values are shown). j, k Systolic BP (j) and maximal AsAo and AbAo (k) diameter at the indicated times (n = 10 for untreated WT group; n = 8 per each MFS group). Data are mean ± s.e.m. ***P < 0.001 and ****P < 0.0001 (versus WT), #P < 0.05 and ##P < 0.01 (versus untreated MFS) by repeated-measurements two-way ANOVA with Tukey’s post-hoc test. Source are provided in the Source Data file. pathology, these markers would in principle be more suitable for patients at risk of aortic aneurysm should perhaps be revisited. monitoring or predicting disease course than for diagnosis. Again, Our data showing VASP-mediated modulation of actin fiber future studies in larger clinical cohorts will be important in formation suggest that PRKG regulation of contractility might be determining whether pharmacological treatment can modify the regulated not only through the myosin components of the acto- profile of nitrated proteins and the levels of cGMP. The presence myosin cytoskeleton, but also through actin. Since actin Tyr of increased plasma cGMP and protein nitration in patients and nitration impairs actin polymerization dynamics74, our finding MFS mice indicates that medical treatment does not induce this that numerous Acta2 Tyr residues are nitrated in the aortas of increase, but it will be important to confirm that medication does MFS mice reveals an additional mechanism for NO-mediated not affect the protein nitration profile. Regardless of their regulation of actomyosin cytoskeleton dynamics tuning and for potential as biomarkers in familial aortopathies, cGMP and contractility dysregulation. In this scenario, the induction of protein nitration appear to be useful tools for determining if this contractility markers might be seen as an adaptative response to a pathway is also activated in other TAADs. deficient capacity of the aorta for contraction. Tissue levels of pVASP-S239 could also be used to determine if Our results suggest that the control of NO levels and signaling PRKG activation is a signature of other TAADs. Recent findings in the aorta is critical for VSMC homeostasis and implicate show that a gain-of-function mutation in the human PRKG1 gene abnormally high NO–sGC–PRKG signaling as the underlying generates a constitutively activated kinase and predisposes to cause of aortopathy in MFS mice and patients. PRKG can be TAA in affected families; moreover, the equivalent activating activated by superoxide anion; however, our data showing that mutation in heterozygous mice (Prkg1R177Q/+ mice) triggers an sGC pharmacological inhibition reduces pVASP-S239 in vitro age-dependent aortic dilation56,70,71. In affected families, the and in vivo strongly suggest that PRKG activation in MFS is increased PRKG1 activity stimulates myosin regulatory light dependent on sGC activity. Furthermore, the regression of aortic chain phosphatase, which can alter VSMC contractility71. It is dilation in MFS mice treated with ODQ supports a causal role for thus clear that PRKG activation can mediate aortopathy in syn- sGC activation in the disease. Similarly, the reversion of aortic dromic and nonsyndromic familial TAAD. It remains to be seen dilation and medial degeneration in MFS mice treated with PRKG whether sGC–PRKG pathway activation also mediates disease in inhibitors or aortic Prkg1 knockdown also strongly suggests that syndromic diseases other than MFS or in nonsyndromic TAADs MFS is mediated by PRKG activation. not caused by PRKG1 mutation. Although prophylactic surgery has increased the lifespan of An essential role of NO–sGC–PRKG signaling in aortopathy MFS patients, there are currently no pharmacological treatments does not exclude roles in the disease for NOS2-derived signals able to arrest aortic growth or prevent dissections14,75. There is independent of this pathway. Indeed, oxidants such as NOS2- therefore an urgent need to develop effective therapies. Our data derived superoxide anion can also activate PRKG29, and signals lay the basis for exploring the implication of NO–sGC–PRKG downstream of PRKG involving oxidative stress and JNK acti- signaling in other forms of TAAD and determining the potential vation have been shown to mediate aortopathy in Prkg1R177Q/+ of sGC or PKRG inhibition in the treatment of patients with MFS mice, in which antioxidants prevent age-dependent aortic or other forms of TAAD. dilation56. However, the action of PRKG appears to be mediated mainly through its regulation of contractility via myosin light Methods chain phosphatase activation and decreased calcium influx in Animal procedures. Animal procedures and experiments complied with all rele- VSMCs44,72. The drop in cellular calcium levels might contribute vant ethical regulations, were approved by the CNIC Ethics Committee and the to aortic disease, as suggested by the deleterious effect of calcium Madrid regional authorities (ref. PROEX 80/16) and conformed to EU Directive channel blockers (CCBs) in MFS mice and by the increased risk 2010/63EU and Recommendation 2007/526/EC regarding the protection of ani- mals used for experimental and other scientific purposes, enforced in Spanish law of aortic dissection and need for aortic surgery in MFS patients under Real Decreto 1201/2005. Overall mouse health was assessed by daily treated with CCBs compared with patients receiving other anti- inspection for signs of discomfort, weight loss, or changes in behavior, mobility, hypertensive drugs73. Indeed, the aortic damage caused by CCBs and feeding or drinking habits. Mice were housed in a pathogen-free animal facility in MFS might be attributable to its cooperation with PRKG under a 12 h light/dark cycle at constant temperature and humidity, and fed standard rodent chow and water ad libitum. Fbn1C1039G/+ mice33, which harbor a activation in depleting cellular calcium stocks. Guideline recom- mutation in the Fbn1 gene, were obtained from Jackson Laboratories (JAX mice mendations for CCBs as an alternative blood pressure regulator in stock #012885). This strain had been previously backcrossed to the C57BL/6 12 NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 ARTICLE background for more than nine generations. All mice were genotyped by tail- (KT5823), both from Focus Biomolecules (Plymouth Meeting, PA, USA), were sample PCR using the following primers: 5′-CTCATCATTTTTGGCCAGTTG-3′ dissolved in dimethylsulfoxide (DMSO), diluted in 0.9% NaCl (DMSO content was and 5′−GCACTTGATGCACATTCACA-3′. Wild-type (WT) mice were on the <0.5%), and administered i.p. at 20 mg/kg/day for 21 days and 5 mg/kg/day for C57BL/6 background, and 12–15-week-old males and females were used for all 7 days, respectively. Control mice received the corresponding intraperitoneal experiments. Mice were treated with isosorbide mononitrate (ISMN; MedChem vehicle. Express; Monmouth Junction, NJ, USA) or DetaNONOate (DetaNO; Enzo Life Sciences; Farmingdale, NY, USA) using subcutaneous osmotic minipumps (Alzet Blood pressure measurements. Arterial blood pressure (BP) was measured in Corp; Cupertino, CA, USA). The specific inhibitors of sGC (ODQ) and PRKG mouse tails using the automated BP-2000 Blood Pressure Analysis System and NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications 13 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 Fig. 8 Prkg1 silencing reverts aortopathy in Marfan syndrome. a Representative immunoblot analysis (n = 2 independent experiments) of Prkg1 expression in WT DetaNO-treated VSMCs transduced with lentivirus (LVi) encoding control shRNA (shScr) or Prkg1-specific shRNAs A, B, and C. Uncropped blots in Supplementary Fig. 13. b Experimental design; 14-week-old MFS mice were inoculated through the jugular vein with LVi encoding GFP and either a control shRNA (shScr) or Prkg1-specific shRNAs A and B. Mice were monitored for aortic dilation and BP prior to treatment and at the indicated times. c Representative images of GFP immunohistochemistry (top) and Prkg1 immunofluorescence (bottom). Top and bottom yellow dashed lines delineate the lumen and the adventitia boundary, respectively. IgG staining of AsAo sections from shScr-transduced MFS mice served as a negative control. Scale bar, 50 µm. d Quantification of Prkg1 immunofluorescence in AsAo sections of control WT and shScr-, and shPrkg1-B-transduced MFS mice (n = 9 in WT mice and MFS mice infected with shPrkg1-B, n = 10 in MFS mice infected with shScr or Prkg1-A). Data are shown relative to WT mice as mean ± s.e.m. Each data point denotes an individual mouse. e Representative pVASP-S239 immunofluorescence (red) in mouse aortic sections. Yellow dashed lines delineate the lumen boundary. IgG staining served as a negative control. Scale bar, 50 µm. f Quantification of pVASP-S239 immunofluorescence in aortic sections from 9 WT, 8 MFS shScr, 10 MFS shPrkg1-A, and 9 MFS shPrkg1-B mice. Data are shown relative to uninfected WT mice as mean ± s.e.m. Each data point denotes an individual mouse. d, f Differences were analyzed by one-way ANOVA with Tukey’s post-hoc test (p-values are shown). g, h Maximal AsAo and AbAo (g) diameter and systolic BP (h) at the indicated times (n = 10 per group). Data are mean ± s.e.m. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 versus shScr by repeated-measurements two-way ANOVA with Tukey’s post-hoc test. i Representative images showing staining with hematoxylin and eosin (HE) and elastic van Gieson (EVG) in AsAo of the indicated mice (n = 9–10 mice per group). Yellow arrowheads indicate elastin breaks. Scale bar, 50 μm. j Wall thickness and k elastin breaks in AsAo sections from the mouse cohorts shown in i. Data are mean ± s.e.m. Each data point denotes an individual mouse. j, k Differences were analyzed by one-way ANOVA with Tukey’s post-hoc test (p-values are shown). Source data are provided in the Source Data file. software (Visitech Systems, Apex, NC, USA)32,76. BP measurements were recorded Lentivirus production and infection. A lentiviral plasmid encoding GFP and in mice located in a tail-cuff restrainer over a warmed surface (37 °C). Mice shRNA targeting mouse Prkg1 was engineered by cloning the following shRNAs underwent a training period, with BP monitored every day for 7 days. After that, into the pH1-DUAL lentiviral vector: shPrkg1-A (sense) 5′- GATCCCCTTGCTTT baseline BP was measured 1 day before the beginning of the treatment in each GCTCTGATTATACTCGAGTATAATCAGAGCAA AGCAAGGTTTTTTGC-3′; mouse cohort. Fifteen consecutive systolic BP measurements were made, and the shPrkg1-A (antisense) 5′- GGCCGCAAAAAACCTTGC TTTGCTCTGATTATAC last 10 readings per mouse were recorded, tested for outliers using the Chauvenet’s TCGAGTATAATCAGAGCAAAGCAAGGg-3′; shPrkg1-B (sense) 5′- GATCCCC criterion and averaged. BP monitoring was repeated during experiments as indi- GGAGAATCTCATCCTAGATCTCGAGATCTAGGATGAGA TTCTCCGGTTT cated in the Figure legends. TTGC-3′; shPrkg1-B (antisense) 5′- GGCCGCAAAAACCGGAGAA TCTCATCC TAGATCTCGGATCTAGGATGAGATTCTCCGGG. Pseudo-typed lentiviruses were produced by transient calcium phosphate transfection of HEK-293T cells with In vivo ultrasound imaging. Images of the aorta were obtained in isoflurane- a plasmid encoding GFP and Prkg1-specific shRNA. Viruses were concentrated sedated mice (2% isoflurane) by high-frequency ultrasound with a VEVO 2100 from culture supernatant by ultracentrifugation (2 h at 121,896xg; Ultraclear echography device (Fujifilm-VisualSonics, Toronto, Canada) with a transducer that Tubes; SW28 rotor and Optima L-100 XP Ultracentrifuge; Beckman, Brea, CA, provides 30-micron axial resolution. Maximal internal aortic diameters were USA), suspended in cold sterile PBS, and titrated by infection of Jurkat cells. Jurkat measured at systole using VEVO 2100 software, version 1.5.0 (VisualSonics). cells were seeded in a 96-well plate and infected with the desired lentiviral dilution Measurements were taken before treatment initiation to determine the baseline (1:10-1:100000). After 48 h, cells were centrifuged 5 min at 1800 × g and resus- diameters and were repeated several times during the experiment. pended in 200 microL of ice-cold PBS with 1:1000 propidium iodide. Transduction efficiency (% of GFP-expressing cells) and cell death (propidium iodide incor- poration) were quantified by flow cytometry, as illustrated in Supplementary Fig. 14. Flow cytometry data were collected using a Canto 3L HTS cytometer and Cell procedures. Primary mouse vascular smooth muscle cells (VSMCs) were the BD FACSDiva Software Version 6.1.3, and analyzed using FlowJo isolated from 4–6-week-old mice and grown as described31,32. Briefly, tissue was 10.7.1 software. digested with a solution of collagenase and elastase until a single-cell suspension VSMCs were infected (multiplicity of infection = 10) overnight at 37 °C in was obtained. Cells were then cultured at 37 °C, 5% CO2 in growth medium growth medium. Medium was then replaced with fresh growth medium, and cells (Dulbecco’s modified Eagle’s medium [DMEM)] supplemented with 20% fetal were cultured for 7 additional days and then processed for mRNA expression bovine serum [FBS]). All experiments were performed during passages 3–7. analysis and cell immunostaining or stimulated for 5 min with DetaNO for VSMCs were FBS-starved 36 h before stimulation experiments. The HEK-293T immunoblot assays. For in vivo transduction experiments, animals were (CRL-1573) and Jurkat (Clone E6-1, TIB-152) cell lines, required for high-titer anesthetized (with ketamine and xylazine), and a small incision was made to lentivirus production and lentivirus titration, respectively, were purchased from expose the right jugular vein49. Virus solution (100 μl, 109 particles/ml in PBS) was ATCC. All cells were mycoplasma-negative. injected directly into the right jugular vein. Transduction efficiency was analyzed in VSMCs were incubated with 100µM DetaNO or 100 µM 8-bromo-guanosine aortic samples by GFP immunohistochemistry and PRKG immunofluorescence. 3',5'-cyclic monophosphate (8-Br-cGMP; Biolog; Bremen, Germany). In some experiments, the NOS inhibitor NG-nitro-L-arginine methyl ester (L-NAME, 300 µM; Sigma–Aldrich; St. Louis, MO, USA), ODQ (10 µM), or KT5823 (1 µM), were added to VSMCs 1 h before the addition of DetaNO or 8-Br-cGMP. In another set cGMP immunoassay. cGMP was measured in mouse and human plasma by of experiments, MFS VSMCs were incubated for 24 h with KT5823. competitive enzyme immunoassay (KGE003, R&D Systems; Minneapolis, MN, For cell immunostaining, cells were fixed with 4% paraformaldehyde for 10 min USA). To obtain mouse plasma, blood was extracted after sacrifice by CO2 inha- and permeabilized with 0.3% Triton X-100 in PBS for 30 min. Samples were lation using the cardiac puncture method, collected in EDTA tubes, and cen- incubated overnight with anti-phospho-VASP (1:50, sc-101439; Santa Cruz trifuged for 15 min at 13,000 × g. Plasma and serum samples from MFS patients Biotechnology; Santa Cruz, CA, USA) or rabbit anti-fibrillin-1 polyclonal and healthy blood donors were obtained and processed following standard oper- antibody77 (1:2000, pAb9543; kindly donated by Dr. L. Sakai). The secondary ating procedures. antibody was AlexaFluor568-conjugated goat anti-mouse (1:500; A-11031; Molecular Probes; Carlsbad, CA, USA). To determine F-actin formation, cells were stained with Texas Red-X-conjugated phalloidin (1:2000, T7471; Thermo Fisher Histology. After sacrifice of mice by CO2 inhalation, aortas were perfused with Scientific, Bremen, Germany) in at least three independent batches, with three saline, isolated, fixed in 10% formalin overnight at 4 °C, and paraffin embedded. technical triplicates per condition and batch and at least three images captured per Paraffin cross-sections (5 μm) from fixed ascending aortas (AsAo) were prepared triplicate. Images were acquired at 1024 × 1024 pixels, 8 bits, using a Confocal TCS for immunochemistry or immunofluorescence, or stained with hematoxylin and Leica SP5 microscope (Leica Microsystems GmbH; Wetzlar, Germany) fitted with a eosin, Alcian blue, or modified Verhoeff elastic-Van Gieson (EVG) kit ×40 oil-immersion objective and Leica LAS-AF V2.7.3. acquisition software. All (Sigma–Aldrich). Images were acquired under a Leica DM2500 microscope fitted images were processed for presentation with Photoshop (Adobe) according to the with a ×40 HCX PL Fluotar objective or with a scanner NanoZoomer-2. guidelines of this journal and analyzed with ImageJ software (version 1.52a, NIH, ORSC110730. Leica Application Suite V3.5.0 acquisition software and NDP.view 2 http://rsb.info.nih/ij/). Staining was quantified after setting an intensity threshold V2.7.43 software were used, respectively. Elastic lamina breaks, defined as inter- to include only specific signals. For intracellular staining, total intensity was ruptions in elastic fibers, were counted in the entire medial layer of six non- relativized to cell area. For extracellular staining of Fbn1, the signal was calculated consecutive cross-sections per mouse, using 4–12 mice per experiment. The mean as the stained area. Each condition’s value was relativized to the average staining of number of breaks was calculated. The exact number of mice per group is indicated the experiment. in the figure legends. 14 NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 ARTICLE For immunofluorescence, deparaffinized sections were rehydrated, boiled 3 min Proteomics. Nitration of plasma and ascending aorta proteins was assessed by to retrieve antigens in 10 mM citrate buffer containing 0.05% Tween-20, pH6, and high-throughput multiplexed isobaric labeling analysis in control and DetaNO- blocked for 1 h with 10% goat serum plus 2% BSA in PBS. Samples were incubated treated C57BL/6 WT mice and MFS mice. In the plasma experiment, seven control, with the following antibodies for immunohistochemistry or immunofluorescence: six DetaNO-treated, and seven MFS animals were used, whereas in the aorta mouse monoclonal anti-pVASP (1:50; sc-101439, Santa Cruz Biotechnology, Santa experiment each group of WT and MFS animals consisted of six pools of two Cruz, CA, USA), rabbit polyclonal anti-p-VASP (1:25; SAB4300129, biological replicates (a total of 24 animals). In both experiments, samples were Sigma–Aldrich, St. Louis, MO, USA), rabbit polyclonal anti-PRKG (1:50, ADI- analyzed in two TMT 10-plex batches, each containing three or four biological KAP-PK005-F, Enzo Life Sciences), and rabbit polyclonal anti-GFP (1:100, replicates from each experimental condition. In the aorta experiment, two channels A11122, Invitrogen; Carlsbad, CA, USA). Specificity was determined by were reserved for internal reference standards created by pooling the WT peptide substituting the primary antibody with unrelated IgG (Santa Cruz) at the same samples. Proteins were extracted from AsAo (50 mg per pool) by tissue homo- dilutions as the antigen-specific antibodies. For immunohistochemistry, color was genization with ceramic beads (MagNa Lyser Green Beads, Roche, Germany) in developed with DAB (Vector Laboratories; Burlingame, CA, USA), and sections extraction buffer (100 mM Tris-HCl pH 7.4, 1 mM EDTA, 4% SDS). Plasma were counterstained with hematoxylin and mounted in DPX (Casa Álvarez, proteins (200 µg per animal) were obtained from about 4 µl plasma, which was Madrid, Spain). For immunofluorescence, secondary antibodies were polyclonal mixed with the same volume of 50 mM Tris, 2% SDS, and 50 mM DTT. In both Alexa-Fluor-647-conjugated goat anti-rabbit or polyclonal Alexa-Fluor-647- experiments, proteins were denatured by boiling for 5 min and subjected to filter- conjugated chicken anti-mouse (1:500, Molecular Probes). Sections were mounted aided digestion (Nanosep Centrifugal Devices with Omega Membrane-10K, with DAPI in Citifluor AF4 mounting medium (Aname; Madrid, Spain). For PALL)78. Briefly, urea was added to each sample to bring SDS concentration down immunofluorescence and immunohistochemistry in mouse and human aortic to 0.06%. Samples were then transferred to the filter and centrifuged at 14,000 × g tissue, antibody staining was quantified in the medial layer of three nonconsecutive for 10 min. Cys residues were blocked with 50 mM iodoacetamide for 1 h at room whole cross-sections per aorta, using 4–12 aortas per condition. The rare temperature in the dark. After three washes with 100 µl 50 mM ammonium histological preparations excluded were those lacking a whole aortic cross-section. bicarbonate pH 8.8, proteins were digested with trypsin (1:30 trypsin:protein, The whole aortic cross-section was screened before taking representative pictures Promega) overnight at 37 °C. After protein digestion, peptides were eluted from the (minimum of three). Immunostaining was technically validated with samples from filter in two steps, first with 40 µl 0.5M ammonium bicarbonate and then with 50 µl the same experimental unit in three independent experiments. Images were 0.5M NaCl. Peptides were acidified with trifluoroacetic acid to a final concentration acquired at 1024 × 1024 pixels, 8 bits, using a Confocal TCS Leica SP5 microscope of 1%, desalted on C18 Oasis HLB cartridges (Waters, Milford, MA, USA), and (Leica Microsystems GmbH; Wetzlar, Germany) fitted with a 40x oil-immersion vacuum-dried. Peptides were TMT-labeled according to the manufacturer’s objective. instructions and desalted on OASIS extraction cartridges. Plasma peptides were All images were processed for presentation with Photoshop (Adobe) according separated into four fractions with increasing amounts of acetonitrile (Fr1, 15%; Fr2, to the guidelines of this journal and analyzed with ImageJ software (version 1.52a, 17.5%; Fr3, 20%; and Fr4, 50%), whereas the aorta peptides where separated into NIH, http://rsb.info.nih/ij/). five fractions (Fr1, 12.5%; Fr2, 15%; Fr3, 17.5%; Fr4, 20%; and Fr5, 50%) using the high pH reversed-phase peptide fractionation kit (Thermo Fisher Scientific, Bre- men, Germany), and dried-down before MS analysis. Real-time and quantitative PCR. Total RNA was isolated from VSMCs or from Quantitative proteomics using multiplexed isobaric labeling was also used to aortas by homogenization with TRIzol (Invitrogen; Carlsbad, CA, USA). Total analyze nitration of plasma proteins from 30 healthy donors and 23 MFS patients. RNA (2 μg) was reverse-transcribed at 37 °C for 50 min in a 20-μl reaction mix Samples were analyzed in seven TMT batches. Each TMT batch contained plasma containing 200 U Moloney murine leukemia virus (MMLV) reverse transcriptase samples from each of the two experimental conditions and two channels were (Promega; Madison, Wi, USA), 100 ng random primers and 40 U RNase Inhibitor reserved for reference internal standards created by pooling the control peptide (Invitrogen). The sequences of the sense and antisense primers used for amplifi- samples. Protein digestion and TMT labeling and fractionation of peptides were cation are described in Supplementary Table 1. qPCR reactions were performed in performed as described above, using five fractions with increasing amounts of triplicate with SYBR master mix (Promega) according to manufacturer guidelines. acetonitrile. Probe specificity was assessed in a post-amplification melting-curve analysis. For Each fraction of the labeled peptide samples was analyzed using a Proxeon Easy each reaction, only one melting-temperature (Tm) peak was produced. The nano-flow HPLC system (Thermo Fisher Scientific) coupled via a nanoelectrospray amount of target mRNA in samples was estimated by the 2−ΔCT relative quanti- ion source (Thermo Fisher Scientific) to either a Q-Exactive HF mass spectrometer fication method, using Gapdh for normalization. Fold ratios were calculated rela- (Termo Fisher Scientific) or an Orbitrap Fusion mass spectrometer (Thermo Fisher tive to mRNA expression levels from control animals. qPCR data were analyzed Scientific). A C18-based reverse phase separation was used, with a 2-cm trap using the Bio-Rad CFX Manager 3.1 software and further analysis was performed column and a 50-cm analytical column (75 µm I.D, 2 µm particle size, Acclaim in Excel software. PepMap RSLC, 100 C18; Thermo Fisher Scientific). Peptides were loaded in buffer A (0.1% formic acid in water (v/v)) and eluted with an acetonitrile gradient consisting of 0–30% buffer B (80% acetonitrile, 0.1% formic acid) for 300 min and Immunoblot analysis. Samples from mouse aortas were isolated, frozen in liquid 50–90% B for 3 min at a flow rate of 200 nl/min. Mass spectra were acquired in a nitrogen and then homogenized (MagNA lyzer, Roche). VSMCs were washed with data-dependent manner, with an automatic switch between MS and MS/MS using ice-cold PBS and lysed in ice-cold total lysis buffer (20 mM Tris‐HCl, pH 7.5; 1% a top-15 acquisiton method with dynamic exclusion. MS spectra were acquired Triton X‐100; 50 mM NaF; 5 mM MgCl2; 500 mM NaCl; 10 mM EDTA) supple- with a mass range of 400–1500 m/z and 120,000 FT resolution. HCD fragmentation mented with protease, phosphatase, and kinase inhibitors (100 μM benzamidin, 1 was performed at 30% normalized collision energy, and MS/MS spectra were μg/ml leupeptin, 1 μg/ml pepstatin, 1 μg/ml aprotinin, 1 μM ditiotreitol, 1 mM analyzed at 30,000 resolution in the orbitrap. Proteins were identified with the PMSF, and 3 mM EGTA). Proteins were separated under reducing conditions on SEQUEST HT algorithm integrated in Proteome Discoverer 2.1 (Thermo Fisher SDS–polyacrylamide gels, transferred to nitrocellulose membranes, and detected Scientific) against a mouse or a human database. The mouse reference proteome with rabbit polyclonal anti-PRKG1 (1:1000, ADI-KAP-PK005-F, Enzo Life Sci- database (UniProtKB/Swiss-Prot and TrEMBL December_2016) was ences), anti-Gapdh (1:10,000; ab8245 Abcam, Cambridge, UK), rabbit polyclonal supplemented with 116 CRAP proteins (Global Proteome Machine) anti-GUCY1A3 (1:1000, 12605-1-AP; Proteintech, Rosemont, IL, USA), rabbit (50915 sequences) and was concatenated with the inverted database constructed polyclonal anti-GUCY1B3 GUCY1A3 (1:500, 19011-1-AP; Proteintech), and from the same target databases (101830 sequences in total) using DecoyPYrat79. mouse monoclonal anti-α-Tubulin (1:40,000, T 6074; Sigma–Aldrich, St. Louis, The human reference proteome database (UniProtKB/Swiss-Prot July_2018) was MO, USA). Bound antibodies were detected with enhanced-chemiluminescence concatenated with the corresponding inverted database following the same (ECL) detection reagent (Millipore, Darmstadt, Germany). All uncropped blots are approach (81632 sequences in total). For database searching, parameters were presented in Supplementary Fig. 13. selected as follows: trypsin digestion with two maximum missed cleavage sites, precursor mass tolerance of 2 Da, and a fragment mass tolerance of 30 ppm. Met Human samples. The study complies with all relevant ethical regulations and was oxidation (15.994915 Da), Trp-nitration (44.985078 Da), and Tyr nitration approved by the Research Ethics Committee of Cantabria (ref. 27/2013), the Ethics (44.985078 Da) were set as variable modifications. Fixed modifications were Committee of Ghent University Hospital (B65020111160), and the Ethics Com- 57.021464 Da in Cys and 229.162932 Da in Lys and in the peptide N-terminal mittee of Instituto de Salud Carlos III (CEI PI91_2018-v2-Enmienda_2019). Aorta, position. The false discovery rate (FDR) was calculated based on the search of plasma and serum samples from patients with Marfan Syndrome were obtained results against the corresponding decoy database using the refined method80 with from biobanks of Hospital Puerta de Hierro, Hospital Vall D’Hebron, and Ghent an additional filter for precursor mass tolerance of 15 ppm81 and estimation of the University Hospital. Tissue samples from MFS patients were obtained during corrected Xcorr82. An FDR of 1% was used as the criterion for peptide elective or emergency surgery for aortic root replacement. Age- and sex-matched identification. ascending aorta samples used as controls were obtained anonymously from mul- Quantitative information from TMT reporter intensities was integrated from tiorgan transplant donors. During preparation of the heart for transplantation, the spectrum level to the peptide level and then to the protein level based on the excess AsAo tissue was trimmed and harvested for the study. Tissues were WSPP model83,84 using the GIA integration algorithm85. Quantitative results were immediately fixed, kept at room temperature for 48 h, and embedded in paraffin. expressed as log2(Ai/C), where Ai is the intensity of the TMT reporter of the Age- and sex-matched control plasma samples were obtained from healthy corresponding sample i and C is the mean intensity of the TMT reporters from the volunteers. Informed consent was obtained from all human participants or their controls (for mouse plasma samples) or the mean value of the two internal families. Patient clinical data were retrieved while maintaining anonymity. reference standards (for human plasma or mouse aorta samples). The algorithm NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications 15 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 was modified to incorporate the quantitative values of modified peptides as part of 2. Lindsay, M. E. & Dietz, H. C. Lessons on the pathogenesis of aneurysm from the automated workflow85–87. Hence, quantitative peptide values are referred to the heritable conditions. Nature 473, 308–316 (2011). weighted averages of the nonmodified peptides from the same protein and are 3. Milewicz, D. M. et al. Genetic basis of thoracic aortic aneurysms and therefore not affected by protein abundance changes. This statistical model dissections: focus on smooth muscle cell contractile dysfunction. Annu. Rev. accurately describes the error distribution of abundance changes for both Genomics Hum. Genet. 9, 283–302 (2008). nonmodified and modified peptides in null-hypothesis experiments86,87. Peptide 4. Dietz, H. C., Loeys, B., Carta, L. & Ramirez, F. Recent progress towards a quantifications from the animals in the same group were further integrated to molecular understanding of Marfan syndrome. Am. J. Med. Genet. C. Semin. obtain an averaged value per condition using the GIA algorithm83. Nitro-protein Med. Genet. 139C, 4–9 (2005). abundance is calculated as a weighted average of all quantified nitrated peptides, 5. Judge, D. P. & Dietz, H. C. Marfan’s syndrome. Lancet 366, 1965–1976 (2005). which as commented above, are referred to the weighted averages of the 6. Canadas, V., Vilacosta, I., Bruna, I. & Fuster, V. Marfan syndrome. Part 1: nonmodified peptides of the same protein. Relative changes in peptide and protein pathophysiology and diagnosis. Nat. Rev. Cardiol. 7, 256–265 (2010). abundance (log2-ratios) were expressed in standardized units (Zpq and Zq)83. 7. Bellini, C. et al. Comparison of 10 murine models reveals a distinct Cumulative distributions from two conditions were judged statistically different biomechanical phenotype in thoracic aortic aneurysms. J. R. Soc. Interface 14, according to the two-tailed Kolmogorov-Smirnov test. For human samples, 1–8 (2017). abundance differences of nitrated peptides between healthy individuals and MFS 8. Humphrey, J. D. & Tellides, G. Central artery stiffness and thoracic patients were analyzed using limma88. Plasma-protein nitration in each patient was aortopathy. Am. J. Physiol. Heart Circ. Physiol. 316, H169–H182 (2019). measured as the nitrated plasma index (NPI), defined as the weighted mean of 9. Yuan, S. M. & Jing, H. Cystic medial necrosis: pathological findings and seven most upregulated nitro-peptides. The statistical difference between mean clinical implications. Rev. Bras. Cir. Cardiovasc. 26, 107–115 (2011). NPI values for healthy donors and MFS patients was calculated by unpaired two- 10. Roccabianca, S., Ateshian, G. A. & Humphrey, J. D. Biomechanical roles of tailed Student’s t-test. For mouse aorta samples, abundance differences in nitro- proteins between untreated WT and MFS mice were also analyzed by limma medial pooling of glycosaminoglycans in thoracic aortic dissection. Biomech. workflow. All tryptic nitrated sequences were annotated in silico by extracting Model Mechanobiol. 13, 13–25 (2014). information from the following databases: Swiss-prot, TrEMBL, Nucleotide 11. Shen, Y. H., Lu, H. S., LeMaire, S. A. & Daugherty, A. Unfolding the story of database and PDB. The assignation of leucine and isoleucine residues in nitrated proteoglycan accumulation in thoracic aortic aneurysm and dissection. peptides to the corresponding protein sequences was checked manually. Arterioscler. Thromb. Vasc. Biol. 39, 1899–1901 (2019). 12. Wight, T. N. A role for proteoglycans in vascular disease. Matrix Biol. 71-72, 396–420 (2018). Statistical analysis. Statistical comparisons were carried out with GraphPad 13. Milewicz, D. M., Prakash, S. K. & Ramirez, F. Therapeutics targeting drivers of Prism software 7.05. Appropriate tests were chosen according to the data dis- thoracic aortic aneurysms and acute aortic dissections: insights from tribution. Differences were analyzed by the Student or the Welch t-test (depending predisposing genes and mouse models. Annu. Rev. Med. 68, 51–67 (2017). on the variance of the tested conditions after F-test); one-way, two-way or 14. Gersony, D. R., McClaughlin, M. A., Jin, Z. & Gersony, W. M. The effect of repeated-measurement two-way analysis of variance (ANOVA); and Bonferroni, beta-blocker therapy on clinical outcome in patients with Marfan’s syndrome: Sidack, Tukey, Dunnett, or Newman post-hoc tests (experiments with ≥3 groups), a meta-analysis. Int. J. Cardiol. 114, 303–308 (2007). as appropriate (variance equality was checked by the Bartlett test). The D´Agos- 15. Selamet Tierney, E. S. et al. Beta-blocker therapy does not alter the rate of tino-Pearson and Shapiro-Wilk normality tests were applied before assuming aortic root dilation in pediatric patients with Marfan syndrome. J. Pediatr. normality distribution of the data. Statistical methodology corresponding to pro- teomics analysis is described above. Statistical significance was assigned at *P < 150, 77–82 (2007). 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. 16. Yetman, A. T., Bornemeier, R. A. & McCrindle, B. W. Usefulness of enalapril Sample size was chosen empirically based on our previous experience in the versus propranolol or atenolol for prevention of aortic dilation in patients with calculation of experimental variability. The initial estimation of sample size was the Marfan syndrome. Am. J. Cardiol. 95, 1125–1127 (2005). made by using software Gpower3.1.94. After this estimation, the initial number of 17. Yang, H. H., van Breemen, C. & Chung, A. W. Vasomotor dysfunction in the animals was reduced based on the results obtained from previous experiments. No thoracic aorta of Marfan syndrome is associated with accumulation of data were excluded. The numbers of animals used are described in the oxidative stress. Vascul. Pharmacol. 52, 37–45 (2010). corresponding figure legends. All experiments were performed with at least three 18. Forstermann, U. & Sessa, W. C. Nitric oxide synthases: regulation and biological replicates, and all attempts on replication were successful. No function. Eur. Heart J. 33, 829–837 (2012). 837a-837d. randomization was performed to allocate animals into experimental groups, and 19. Oller, J. et al. Nitric oxide mediates aortic disease in mice deficient in the investigators were not blinded to group allocation during in vivo experiments. metalloprotease Adamts1 and in a mouse model of Marfan syndrome. Nat. Experimental groups were balanced in terms of animal age, weight and AsAo/ Med. 23, 200–212 (2017). AbAo basal diameter. Male mice were used throughout the manuscript, except for 20. Lincoln, T. M., Wu, X., Sellak, H., Dey, N. & Choi, C. S. Regulation of vascular experiments corresponding to Fig. 7, where male and female mice were used. smooth muscle cell phenotype by cyclic GMP and cyclic GMP-dependent Animals were genotyped before experiments, and they were all caged together and protein kinase. Front. Biosci. 11, 356–367 (2006). treated in the same way. 21. Lima, B., Forrester, M. T., Hess, D. T. & Stamler, J. S. S-nitrosylation in cardiovascular signaling. Circ. Res. 106, 633–646 (2010). Reporting summary. Further information on research design is available in the Nature 22. Wolhuter, K. et al. Evidence against stable protein S-nitrosylation as a Research Reporting Summary linked to this article. widespread mechanism of post-translational regulation. Mol. Cell 69, 438–450 (2018). 23. Radi, R. Protein tyrosine nitration: biochemical mechanisms and structural Data availability basis of functional effects. Acc. Chem. Res. 46, 550–559 (2013). Source data are provided with this paper for all figures. The mass spectrometry raw data 24. Alkaitis, M. S. & Crabtree, M. J. Recoupling the cardiac nitric oxide synthases: that support the findings of this study, including nitrated peptide MS/MS spectra tetrahydrobiopterin synthesis and recycling. Curr. Heart Fail Rep. 9, 200–210 (Supplementary Data 2 and 4), mouse reference proteome database December 2016 and (2012). human reference proteome database July 2018, are publicly available in the Peptide Atlas 25. Crabtree, M. J. & Channon, K. M. Synthesis and recycling of repository (http://www.peptideatlas.org/PASS/PASS01528). The following databases tetrahydrobiopterin in endothelial function and vascular disease. Nitric Oxide were used: UniProtKB, Swiss-Prot and TrEMBL (https://www.uniprot.org/uniprot/), 25, 81–88 (2011). Nucleotide database (https://www.ncbi.nlm.nih.gov/nucleotide/), Protein Data Bank, 26. Luo, S., Lei, H., Qin, H. & Xia, Y. Molecular mechanisms of endothelial NO PDB (https://www.rcsb.org/). Other relevant datasets generated and/or analyzed during synthase uncoupling. Curr. Pharm. Des. 20, 3548–3553 (2014). the current study are available from the corresponding authors on reasonable 27. Otani, H. The role of nitric oxide in myocardial repair and remodeling. request. Source data are provided with this paper. Antioxid. Redox Signal 11, 1913–1928 (2009). 28. Jimenez-Altayo, F. et al. Redox stress in Marfan syndrome: dissecting the role of the NADPH oxidase NOX4 in aortic aneurysm. Free Radic. Biol. Med. 118, Received: 21 May 2020; Accepted: 1 April 2021; 44–58 (2018). 29. Burgoyne, J. R. et al. Cysteine redox sensor in PKGIa enables oxidant-induced activation. Science 317, 1393–1397 (2007). 30. Kim, M. et al. GLP-1 receptor activation and Epac2 link atrial natriuretic peptide secretion to control of blood pressure. Nat. Med. 19, 567–575 (2013). 31. Mendez-Barbero, N. et al. A major role for RCAN1 in atherosclerosis References progression. EMBO Mol. Med. 5, 1901–1917 (2013). 1. Ruddy, J. M., Jones, J. A. & Ikonomidis, J. S. Pathophysiology of thoracic aortic 32. Villahoz, S. et al. Conditional deletion of Rcan1 predisposes to hypertension- aneurysm (TAA): is it not one uniform aorta? Role of embryologic origin. mediated intramural hematoma and subsequent aneurysm and aortic rupture. Prog. Cardiovasc. Dis. 56, 68–73 (2013). Nat. Commun. 9, 4795 (2018). 16 NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 ARTICLE 33. Judge, D. P. et al. Evidence for a critical contribution of haploinsufficiency in 64. Chawla, R. K. et al. Elevated plasma and urinary guanosine 3’:5’- the complex pathogenesis of Marfan syndrome. J. Clin. Invest. 114, 172–181 monophosphate and increased production rate in patients with neoplastic (2004). diseases. Cancer Res. 40, 3915–3920 (1980). 34. Benz, P. M. et al. Differential VASP phosphorylation controls remodeling of 65. Gott, V. L. et al. Replacement of the aortic root in patients with Marfan’s the actin cytoskeleton. J. Cell Sci. 122, 3954–3965 (2009). syndrome. N. Engl. J. Med. 340, 1307–1313 (1999). 35. Doppler, H. & Storz, P. Regulation of VASP by phosphorylation: 66. Iskandar, Z., Mordi, I., Lang, C. C., Huang, J. T. J. & Choy, A. M. Biomarkers consequences for cell migration. Cell Adh Migr. 7, 482–486 (2013). of aortopathy in Marfan syndrome. Cardiol. Rev. 28, 92–97 (2020). 36. Dale, M. et al. Premature aortic smooth muscle cell differentiation contributes 67. den Hartog, A. W. et al. The risk for type B aortic dissection in Marfan to matrix dysregulation in Marfan Syndrome. PLoS ONE 12, e0186603 (2017). syndrome. J. Am. Coll. Cardiol. 65, 246–254 (2015). 37. Granata, A. et al. An iPSC-derived vascular model of Marfan syndrome identifies 68. Bae, E. H., Kim, I. J., Ma, S. K., Lee, J. U. & Kim, S. W. Altered regulation of key mediators of smooth muscle cell death. Nat. Genet. 49, 97–109 (2017). renal nitric oxide and atrial natriuretic Peptide systems in lipopolysaccharide- 38. Pfeilschifter, J., Eberhardt, W. & Beck, K. F. Regulation of gene expression by induced kidney injury. Korean J. Physiol. Pharmacol. 15, 273–277 (2011). nitric oxide. Pflug. Arch. 442, 479–486 (2001). 69. Sanyal, A. et al. Interplay between obesity-induced inflammation and cGMP 39. Tarkin, J. M. & Kaski, J. C. Vasodilator therapy: nitrates and nicorandil. signaling in white adipose tissue. Cell Rep. 18, 225–236 (2017). Cardiovasc. Drugs Ther. 30, 367–378 (2016). 70. Evangelista, A. Bicuspid aortic valve and aortic root disease. Curr. Cardiol. 40. Oller, J. et al. C/EBPbeta and nuclear factor of activated T cells differentially Rep. 13, 234–241 (2011). regulate Adamts-1 induction by stimuli associated with vascular remodeling. 71. Guo, D. C. et al. Recurrent gain-of-function mutation in PRKG1 causes Mol. Cell Biol. 35, 3409–3422 (2015). thoracic aortic aneurysms and acute aortic dissections. Am. J. Hum. Genet. 93, 41. Parker, J. O., Farrell, B., Lahey, K. A. & Moe, G. Effect of intervals between 398–404 (2013). doses on the development of tolerance to isosorbide dinitrate. N. Engl. J. Med. 72. Burgoyne, J. R. & Eaton, P. Oxidant sensing by protein kinases a and g enables 316, 1440–1444 (1987). integration of cell redox state with phosphoregulation. Sensors (Basel) 10, 42. Broadus, A. E., Kaminsky, N. I., Hardman, J. G., Sutherland, E. W. & Liddle, 2731–2751 (2010). G. W. Kinetic parameters and renal clearances of plasma adenosine 3’,5’- 73. Doyle J. J. et al. A deleterious gene-by-environment interaction imposed by monophosphate and guanosine 3’,5’-monophosphate in man. J. Clin. Invest. calcium channel blockers in Marfan syndrome. elife 4, 1–18 (2015). 49, 2222–2236 (1970). 74. Aslan, M. Functional consequences of actin nitration: in vitro and in disease 43. Honma, M. & Ui, M. Plasma cyclic GMP: response to cholinergic agents. Eur. states. Amino Acids 42, 65–74 (2012). J. Pharmacol. 47, 1–10 (1978). 75. Milleron, O. et al. Marfan Sartan: a randomized, double-blind, placebo- 44. Farah, C., Michel, L. Y. M. & Balligand, J. L. Nitric oxide signalling in controlled trial. Eur. Heart J. 36, 2160–2166 (2015). cardiovascular health and disease. Nat. Rev. Cardiol. 15, 292–316 (2018). 76. Martinez-Martinez, S. et al. Cardiomyocyte calcineurin is required for the 45. Kohr, M. J. et al. Measurement of S-nitrosylation occupancy in the onset and progression of cardiac hypertrophy and fibrosis in adult mice. FEBS myocardium with cysteine-reactive tandem mass tags: short communication. J. 286, 46–65 (2019). Circ. Res. 111, 1308–1312 (2012). 77. Charbonneau, N. L. et al. In vivo studies of mutant fibrillin-1 microfibrils. J. 46. Patel, R. P., Levonen, A., Crawford, J. H. & Darley-Usmar, V. M. Mechanisms Biol. Chem. 285, 24943–24955 (2010). of the pro- and anti-oxidant actions of nitric oxide in atherosclerosis. 78. Wisniewski, J. R., Zougman, A., Nagaraj, N. & Mann, M. Universal sample Cardiovasc. Res. 47, 465–474 (2000). preparation method for proteome analysis. Nat. Methods 6, 359–362 (2009). 47. Hofmann, F., Feil, R., Kleppisch, T. & Schlossmann, J. Function of cGMP- 79. Wright, J. C. & Choudhary, J. S. DecoyPyrat: Fast non-redundant hybrid dependent protein kinases as revealed by gene deletion. Physiol. Rev. 86, 1–23 decoy sequence generation for large scale proteomics. J. Proteomics Bioinform. (2006). 9, 176–180 (2016). 48. Alfranca, A., Campanero, M. R. & Redondo, J. M. New methods for disease 80. Navarro, P. & Vazquez, J. A refined method to calculate false discovery rates for modeling using lentiviral vectors. Trends Mol. Med. 24, 825–837 (2018). peptide identification using decoy databases. J. Proteome Res. 8, 1792–1796 (2009). 49. Esteban, V. et al. Regulator of calcineurin 1 mediates pathological vascular 81. Bonzon-Kulichenko, E., Garcia-Marques, F., Trevisan-Herraz, M. & Vazquez, wall remodeling. J. Exp. Med. 208, 2125–2139 (2011). J. Revisiting peptide identification by high-accuracy mass spectrometry: 50. Armstrong P. W. et al. Vericiguat in patients with heart failure and reduced problems associated with the use of narrow mass precursor windows. J. ejection fraction. N. Engl. J. Med. 382, 1883–1893 (2020). Proteome Res. 14, 700–710 (2015). 51. Rubin, L. J. et al. Riociguat for the treatment of pulmonary arterial 82. Martinez-Bartolome, S. et al. Properties of average score distributions of hypertension: a long-term extension study (PATENT-2). Eur. Respir. J. 45, SEQUEST: the probability ratio method. Mol. Cell Proteomics 7, 1135–1145 (2008). 1303–1313 (2015). 83. Navarro, P. et al. General statistical framework for quantitative proteomics by 52. Sawabe, T. et al. A novel soluble guanylate cyclase activator with reduced risk stable isotope labeling. J. Proteome Res. 13, 1234–1247 (2014). of hypotension by short-acting vasodilation. Pharmacol. Res. Perspect. 7, 84. Garcia-Marques, F. et al. A novel systems-biology algorithm for the analysis of e00463 (2019). coordinated protein responses using quantitative proteomics. Mol. Cell 53. Sellers, S. L. et al. Inhibition of Marfan syndrome aortic root dilation by Proteomics 15, 1740–1760 (2016). Losartan: role of angiotensin II receptor type 1-independent activation of 85. Trevisan-Herraz, M. et al. SanXoT: a modular and versatile package for the endothelial function. Am. J. Pathol. 188, 574–585 (2018). quantitative analysis of high-throughput proteomics experiments. 54. Kraehling, J. R. & Sessa, W. C. Contemporary approaches to modulating the Bioinformatics 35, 1594–1596 (2019). nitric oxide-cGMP pathway in cardiovascular disease. Circ. Res. 120, 86. Bagwan, N. et al. Comprehensive quantification of the modified proteome 1174–1182 (2017). reveals oxidative heart damage in mitochondrial heteroplasmy. Cell Rep. 23, 55. Chen, W. et al. Peroxynitrite induces destruction of the tetrahydrobiopterin 3685–3697 (2018). and heme in endothelial nitric oxide synthase: transition from reversible to 87. Bonzon-Kulichenko, E. et al. Improved integrative analysis of the thiol redox irreversible enzyme inhibition. Biochemistry 49, 3129–3137 (2010). proteome using filter-aided sample preparation. J. Proteomics 214, 103624 (2020). 56. Schwaerzer, G. K. et al. Aortic pathology from protein kinase G activation is 88. Ritchie, M. E. et al. limma powers differential expression analyses for RNA- prevented by an antioxidant vitamin B12 analog. Nat. Commun. 10, 3533 (2019). sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015). 57. Lomeli, O. et al. The evaluation of flow-mediated vasodilation in the brachial artery correlates with endothelial dysfunction evaluated by nitric oxide synthase metabolites in Marfan syndrome patients. Front. Physiol. 9, 965 (2018). Acknowledgements 58. Heinrich, T. A. et al. Biological nitric oxide signalling: chemistry and We thank L. Sakai for reagents; S. Bartlett for English language editing; S. Lamas and terminology. Br. J. Pharmacol. 169, 1417–1429 (2013). M.A. Hurlé for critical reading of the manuscript and continuous advice; J. Oller for 59. Hamilton, R. T. et al. LDL protein nitration: implication for LDL protein suggestions and help at the beginning of this work; C. Salas for help organizing clinical unfolding. Arch. Biochem. Biophys. 479, 1–14 (2008). database and patient samples; N. Martín-Cofrares and V. Labrador for advice on confocal 60. Parastatidis, I. et al. Increased protein nitration burden in the atherosclerotic imaging and immunohistochemistry experiments; the CNIC Facilities of histology, lesions and plasma of apolipoprotein A-I deficient mice. Circ. Res. 101, proteomics, and advanced imaging; and R. Magni, A.I. Torralbo, and A. Colmenar for 368–376 (2007). excellent technical support and advice. The CNIC is supported by the Spanish Ministerio 61. Piroddi, M. et al. Plasma nitroproteome of kidney disease patients. Amino de Ciencia e Innovación and the Pro-CNIC Foundation and is a Severo Ochoa Center of Acids 40, 653–667 (2011). Excellence (SEV-2015-0505). The project leading to these results has received funding 62. Ogawa, K. et al. Increased plasma cyclic nucleotide concentrations in from “La Caixa” Banking Foundation under project codes HR18-00068 (to M.R.C. and congestive heart failure. Br. Heart J. 52, 524–529 (1984). J.M.R.) and HR17-00247 (to J.V.); Spanish Ministerio de Ciencia e Innovación grants 63. Reginauld, S. H. et al. Differential regulation of ANP and BNP in acute RTI2018-099246-B-I00 (MICIU/AEI/FEDER, UE) and SAF2015-636333R (MINECO/ decompensated heart failure: deficiency of ANP. JACC Heart Fail 7, 891–898 FEDER, UE) to J.M.R., SAF2017-88881R (MINECO/AEI/FEDER, UE) to M.R.C., and (2019). BIO2015-67580-P and PGC2018-097019-B-I00 to J.V.; the Comunidad de Madrid NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications 17 ARTICLE NATURE COMMUNICATIONS | https://doi.org/10.1038/s41467-021-22933-3 through the European Social Fund (ESF)-financed program AORTASANA-CM (B2017/ Correspondence and requests for materials should be addressed to M.R.C. or J.M.R. BMD-3676) to J.M.R., M.R.C., and A.F.; the Instituto de Salud Carlos III (CIBER-CV Peer review information Nature Communications thanks B. Timothy Baxter, Franz CB16/11/00264 and CB16/11/00277; PRB3-IPT17/0019-ProteoRed to J.V.; and PI17/ Hofmann, Xianquan Zhan and the other, anonymous, reviewer(s) for their contribution 00381, with cofinancing from the European Regional Development Fund, to G.T.-T.); to the peer review of this work. Peer reviewer reports are available. Fundacio La Marato TV3 (20151330 to J.M.R. and A.E., and 122/C/2015 to J.V.); The Marfan Foundation USA Faculty grant 2017 MRF/1701 (to J.M.R.); and Spanish Min- isterio de Ciencia e Innovación fellowships FPU (17/05866), FPI (BES-2016-077649), and Reprints and permission information is available at http://www.nature.com/reprints Sara Borrell (CD18/00028), to A.d.l.F., M.J.R.-R., and M.T., respectively. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Author contributions M.R.C. and J.M.R. conceived the study; M.T., A.d.l.F.-A., M.R.C., and J.M.R. designed the study and analyzed the data; A.d.l.F.-A. and M.T. performed most of the experiments, Open Access This article is licensed under a Creative Commons with contributions from S.M.-M., A.A., M.J.R.-R., M.J.M.-O., and D.L.-M.; A.A., E.B.-K., Attribution 4.0 International License, which permits use, sharing, I.G.-V., performed the proteomics and bioinformatics analysis under the supervision of adaptation, distribution and reproduction in any medium or format, as long as you give J.V., who also provided ideas for the project; C.E.M., E.G.-I., S.M., A.F., L.M.-M., J.D.B., J.F.N., G.T.-T., and A.E. provided human tissue samples and ideas for the project; M.R.C. appropriate credit to the original author(s) and the source, provide a link to the Creative and J.M.R. wrote the manuscript with contributions from A.d.l.F.-A., M.T., E.B.-K., A.A., Commons license, and indicate if changes were made. The images or other third party and J.V. All authors read and approved the manuscript. material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory Competing interests regulation or exceeds the permitted use, you will need to obtain permission directly from The authors declare no competing interests. the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/. Additional information Supplementary information The online version contains supplementary material © The Author(s) 2021 available at https://doi.org/10.1038/s41467-021-22933-3. 18 NATURE COMMUNICATIONS | (2021)12:2628 | https://doi.org/10.1038/s41467-021-22933-3 | www.nature.com/naturecommunications