Modification of Actin Fibers Changes the Electrical Phenotype of Cardiac Myofibroblasts

Background: Slow conduction and ectopic activity are major determinants of cardiac arrhythmogenesis. Both of these conditions can be elicited by myofibroblasts (MFBs) following establishment of heterocellular gap junctional coupling with cardiomyocytes. MFBs appear during structural remodeling of the heart and are characterized by the expression of α-smooth muscle actin (α-SMA) containing stress fibers. In this study, we investigated whether pharmacological interference with the actin cytoskeleton affects myofibroblast arrhythmogeneicity. Methods: Experiments were performed with patterned growth strands of neonatal rat ventricular cardiomyocytes coated with cardiac MFBs. Impulse conduction velocity (θ) and maximal upstroke velocities of propagated action potentials (dV/dtmax), expressed as % action potential amplitude change (%APA) per ms, were measured optically using voltage sensitive dyes. Actin was destabilized by latrunculin B (LtB) and cytochalasin D and stabilized with jasplakinolide. Data are given as mean ± S.D. (n = 5-22). Single cell electrophysiology was assessed using standard patch-clamp techniques. Results: As revealed by immunocytochemistry, exposure of MFBs to LtB (0.01-10 μmol/L) profoundly disrupted stress fibers which led to drastic changes in cell morphology with MFBs assuming an astrocyte-like shape. In control cardiomyocyte strands (no MFB coat), LtB had negligible effects on θ and dV/dtmax. In contrast, LtB applied to MFB-coated strands increased θ dose-dependently from 197 ± 35 mm/s to 344 ± 26 mm/s and dV/dtmax from 38 ± 5 to 78 ± 3% APA/ms, i.e., to values virtually identical to those of cardiomyocyte control strands (339 ± 24 mm/s; 77 ± 3% APA/ms). Highly similar results were obtained when exposing the preparations to cytochalasin D. In contrast, stabilization of actin with increasing concentrations of jasplakinolide exerted no significant effects on impulse conduction characteristics in MFB-coated strands. Whole-cell patch-clamp experiments showed that LtB hyperpolarized MFBs from -25 mV to -50 mV, thus limiting their depolarizing effect on cardiomyocytes which was shown before to cause arrhythmogenic slow conduction and ectopic activity. Conclusion: Pharmacological interference with the actin cytoskeleton of cardiac MFBs affects their electrophysiological phenotype to such an extent that they loose their detrimental effects on cardiomyocyte electrophysiology. This result might form a basis for the development of therapeutic strategies aimed at limiting the arrhythmogenic potential of MFBs.

Bacterial phenotype variants in group B streptococcal toxic shock syndrome

We conducted genetic and functional analyses of isolates from a patient with group B streptococcal (GBS) necrotizing fasciitis and toxic shock syndrome. Tissue cultures simultaneously showed colonies with high hemolysis (HH) and low hemolysis (LH). Conversely, the HH and LH variants exhibited low capsule (LC) and high capsule (HC) expression, respectively. Molecular analysis demonstrated that the 2 GBS variants were of the same clonal origin. Genetic analysis found a 3-bp deletion in the covR gene of the HH/LC variant. Functionally, this isolate was associated with an increased growth rate in vitro and with higher interleukin-8 induction. However, in whole blood, opsonophagocytic and intracellular killing assays, the LH/HC phenotype demonstrated higher resistance to host phagocytic killing. In a murine model, LH/HC resulted in higher levels of bacteremia and increased host mortality rate. These findings demonstrate differences in GBS isolates of the same clonal origin but varying phenotypes.

STAR splicing mutations cause the severe phenotype of lipoid congenital adrenal hyperplasia: insights from a novel splice mutation and review of reported cases

OBJECTIVE The steroidogenic acute regulatory protein (StAR) transports cholesterol to the mitochondria for steroidogenesis. Loss of StAR function causes lipoid congenital adrenal hyperplasia (LCAH) which is characterized by impaired synthesis of adrenal and gonadal steroids causing adrenal insufficiency, 46,XY disorder of sex development (DSD) and failure of pubertal development. Partial loss of StAR activity may cause adrenal insufficiency only. PATIENT A newborn girl was admitted for mild dehydration, hyponatremia, hyperkalemia and hypoglycaemia and had normal external female genitalia without hyperpigmentation. Plasma cortisol, 17OH-progesterone, DHEA-S, androstendione and aldosterone were low, while ACTH and plasma renin activity were elevated, consistent with the diagnosis of primary adrenal insufficiency. Imaging showed normal adrenals, and cytogenetics revealed a 46,XX karyotype. She was treated with fluids, hydrocortisone and fludrocortisone. DESIGN, METHODS AND RESULTS Genetic studies revealed a novel homozygous STAR mutation in the 3' acceptor splice site of intron 4, c.466-1G>A (IVS4-1G>A). To test whether this mutation would affect splicing, we performed a minigene experiment with a plasmid construct containing wild-type or mutant StAR gDNA of exons-introns 4-6 in COS-1 cells. The splicing was assessed on total RNA using RT-PCR for STAR cDNAs. The mutant STAR minigene skipped exon 5 completely and changed the reading frame. Thus, it is predicted to produce an aberrant and shorter protein (p.V156GfsX19). Computational analysis revealed that this mutant protein lacks wild-type exons 5-7 which are essential for StAR-cholesterol interaction. CONCLUSIONS STAR c.466-1A skips exon 5 and causes a dramatic change in the C-terminal sequence of the protein, which is essential for StAR-cholesterol interaction. This splicing mutation is a loss-of-function mutation explaining the severe phenotype of our patient. Thus far, all reported splicing mutations of STAR cause a severe impairment of protein function and phenotype.

LQTS-associated mutation A257G in α1-syntrophin interacts with the intragenic variant P74L to modify its biophysical phenotype.

The SNTA1-encoded α1-syntrophin (SNTA1) missense mutation, p.A257G, causes long QT syndrome (LQTS) by pathogenic accentuation of Nav1.5's sodium current (I Na). Subsequently, we found p.A257G in combination with the SNTA1 polymorphism, p.P74L in 4 victims of sudden infant death syndrome (SIDS) as well as in 3 adult controls. We hypothesized that p.P74L-SNTA1 could functionally modify the pathogenic phenotype of p.A257G-SNTA1, thus explaining its occurrence in non-LQTS populations. The SNTA1 variants p.P74L, p.A257G, and the combination variant p.P74L/p.A257G were engineered using PCR-based overlap-extension and were co-expressed heterologously with SCN5A in HEK293 cells. I Na was recorded using the whole-cell method. Compared to wild-type (WT), the significant increase in peak I Na and window current found with p.A257G was reversed by the intragenic variant p.P74L (p.P74L/p.A257G). These results report for the first time the intragenic rescue of an LQT-associated SNTA1 mutation when found in combination with the SNTA1 polymorphism p.P74L, suggesting an ever-increasing picture of complexity in terms of genetic risk stratification for arrhythmia.

Unique mixed phenotype and unexpected functional effect revealed by novel compound heterozygosity mutations involving SCN5A.

BACKGROUND Functional characterization of mutations involving the SCN5A-encoded cardiac sodium channel has established the pathogenic mechanisms for type 3 long QT syndrome and type 1 Brugada syndrome and has provided key insights into the physiological importance of essential structure-function domains. OBJECTIVE This study sought to present the clinical and biophysical phenotypes discerned from compound heterozygosity mutations in SCN5A on different alleles in a toddler diagnosed with QT prolongation and fever-induced ventricular arrhythmias. METHODS A 22-month-old boy presented emergently with fever and refractory ventricular tachycardia. Despite restoration of sinus rhythm, the infant sustained profound neurological injury and died. Using polymerase chain reaction, denaturing high-performance liquid chromatography, and direct DNA sequencing, comprehensive open-reading frame/splice mutational analysis of the 12 known long QT syndrome susceptibility genes was performed. RESULTS The infant had 2 SCN5A mutations: a maternally inherited N-terminal frame shift/deletion (R34fs/60) and a paternally inherited missense mutation, R1195H. The mutations were engineered by site-directed mutagenesis and heterologously expressed transiently in HEK293 cells. As expected, the frame-shifted and prematurely truncated peptide, SCN5A-R34fs/60, showed no current. SCN5A-R1195H had normal peak and late current but abnormal voltage-dependent gating parameters. Surprisingly, co-expression of SCN5A-R34fs/60 with SCN5A-R1195H elicited a significant increase in late sodium current, whereas co-expression of SCN5A-WT with SCN5A-R34fs/60 did not. CONCLUSIONS A severe clinical phenotype characterized by fever-induced monomorphic ventricular tachycardia and QT interval prolongation emerged in a toddler with compound heterozygosity involving SCN5A: R34fs/60, and R1195H. Unexpectedly, the 94-amino-acid fusion peptide derived from the R34fs/60 mutation accentuated the late sodium current of R1195H-containing Na(V)1.5 channels in vitro.

Cardiac phenotype of Duchenne Muscular Dystrophy: Insights from cellular studies

Dilated cardiomyopathy is a serious and almost inevitable complication of Duchenne Muscular Dystrophy, a devastating and fatal disease of skeletal muscle resulting from the lack of functional dystrophin, a protein linking the cytoskeleton to the extracellular matrix. Ultimately, it leads to congestive heart failure and arrhythmias resulting from both cardiac muscle fibrosis and impaired function of the remaining cardiomyocytes. Here we summarize findings obtained in several laboratories, focusing on cellular mechanisms that result in degradation of cardiac functions in dystrophy. This article is part of a Special Issue entitled "Calcium Signaling in Heart".

Possible role of Cdx2 in the serrated pathway of colorectal cancer characterized by BRAF mutation, high-level CpG Island methylator phenotype and mismatch repair-deficiency

Colorectal cancer is a heterogeneous disease at the histomorphological, clinical and molecular level. Approximately 20% of cases may progress through the "serrated" pathway characterized by BRAF mutation and high-level CpG Island Methylator Phenotype (CIMP). A large subgroup are additionally microsatellite instable (MSI) and demonstrate significant loss of tumor suppressor Cdx2. The aim of this study is to determine the specificity of Cdx2 protein expression and CpG promoter hypermethylation for BRAF(V600E) and high-level CIMP in colorectal cancer. Cdx2, Mlh1, Msh2, Msh6, and Pms2 were analyzed by immunohistochemistry using a multi-punch tissue microarray (TMA; n = 220 patients). KRAS and BRAF(V600E) mutation analysis, CDX2 methylation and CIMP were investigated. Loss of Cdx2 was correlated with larger tumor size (P = 0.0154), right-sided location (P = 0.0014), higher tumor grade (P < 0.0001), more advanced pT (P = 0.0234) and lymphatic invasion (P = 0.0351). Specificity was 100% for mismatch repair (MMR)-deficiency (P < 0.0001), 92.2% (P < 0.0001) for BRAF(V600E) and 91.8% for CIMP-high. Combined analysis of BRAF(V600E) /CIMP identified Cdx2 loss as sensitive (80%) and specific (91.5%) for mutation/high status. These results were validated on eight well-established colorectal cancer cell lines. CDX2 methylation correlated with BRAF(V600E) (P = 0.0184) and with Cdx2 protein loss (P = 0.0028). These results seem to indicate that Cdx2 may play a role in the serrated pathway to colorectal cancer as underlined by strong relationships with BRAF(V600E) , CIMP-high and MMR-deficiency. Whether this protein can only be used as a "surrogate" marker, or is functionally involved in the progression of these tumors remains to be elucidated.

Loss of Raf-1 kinase inhibitor protein (RKIP) is strongly associated with high-grade tumor budding and correlates with an aggressive phenotype in pancreatic ductal adenocarcinoma (PDAC)

BACKGROUND Raf-1 kinase inhibitor protein (RKIP) has emerged as a significant metastatic suppressor in a variety of human cancers and is known to inhibit Ras/Raf/MEK/ERK signaling. By suppressing the activation of the NFkB/SNAIL circuit, RKIP can regulate the induction of epithelial-mesenchymal transition (EMT). The aim of this study was to evaluate RKIP expression and to determine its association with clinicopathological features, including EMT in form of tumor budding in pancreatic ductal adenocarcinoma (PDAC). METHODS Staining for RKIP was performed on a multipunch Tissue Microarray (TMA) of 114 well-characterized PDACs with clinico-pathological, follow-up and adjuvant therapy information. RKIP-expression was assessed separately in the main tumor body and in the tumor buds. Another 3 TMAs containing normal pancreatic tissue, precursor lesions (Pancreatic Intraepithelial Neoplasia, PanINs) and matched lymph node metastases were stained in parallel. Cut-off values were calculated by receiver operating characteristic (ROC) curve analysis. RESULTS We found a significant progressive loss of RKIP expression between normal pancreatic ductal epithelia (average: 74%), precursor lesions (PanINs; average: 37%), PDAC (average 20%) and lymph node metastases (average 8%, p<0.0001). RKIP expression was significantly lower in tumor buds (average: 6%) compared to the main tumor body (average 20%; p<0.005). RKIP loss in the tumor body was marginally associated with advanced T-stage (p=0.0599) as well as high-grade peritumoral (p=0.0048) and intratumoral budding (p=0.0373). RKIP loss in the buds showed a clear association with advanced T stage (p=0.0089). CONCLUSIONS The progressive loss of RKIP seems to play a major role in the neoplastic transformation of pancreas, correlates with aggressive features in PDAC and is associated with the presence of EMT in form of tumor budding.

Predictive oncology: a review of multidisciplinary, multiscale in silico modeling linking phenotype, morphology and growth.

Empirical evidence and theoretical studies suggest that the phenotype, i.e., cellular- and molecular-scale dynamics, including proliferation rate and adhesiveness due to microenvironmental factors and gene expression that govern tumor growth and invasiveness, also determine gross tumor-scale morphology. It has been difficult to quantify the relative effect of these links on disease progression and prognosis using conventional clinical and experimental methods and observables. As a result, successful individualized treatment of highly malignant and invasive cancers, such as glioblastoma, via surgical resection and chemotherapy cannot be offered and outcomes are generally poor. What is needed is a deterministic, quantifiable method to enable understanding of the connections between phenotype and tumor morphology. Here, we critically assess advantages and disadvantages of recent computational modeling efforts (e.g., continuum, discrete, and cellular automata models) that have pursued this understanding. Based on this assessment, we review a multiscale, i.e., from the molecular to the gross tumor scale, mathematical and computational "first-principle" approach based on mass conservation and other physical laws, such as employed in reaction-diffusion systems. Model variables describe known characteristics of tumor behavior, and parameters and functional relationships across scales are informed from in vitro, in vivo and ex vivo biology. We review the feasibility of this methodology that, once coupled to tumor imaging and tumor biopsy or cell culture data, should enable prediction of tumor growth and therapy outcome through quantification of the relation between the underlying dynamics and morphological characteristics. In particular, morphologic stability analysis of this mathematical model reveals that tumor cell patterning at the tumor-host interface is regulated by cell proliferation, adhesion and other phenotypic characteristics: histopathology information of tumor boundary can be inputted to the mathematical model and used as a phenotype-diagnostic tool to predict collective and individual tumor cell invasion of surrounding tissue. This approach further provides a means to deterministically test effects of novel and hypothetical therapy strategies on tumor behavior.

Conversion of myoblasts to physiologically active neuronal phenotype.

Repressor element 1 (RE1)-silencing transcription factor (REST)/neuron-restrictive silencer factor (NRSF) can repress several terminal neuronal differentiation genes by binding to a specific DNA sequence (RE1/neuron-restrictive silencer element [NRSE]) present in their regulatory regions. REST-VP16 binds to the same RE1/NRSE, but activates these REST/NRSF target genes. However, it is unclear whether REST-VP16 expression is sufficient to cause formation of functional neurons either from neural stem cells or from heterologous stem cells. Here we show that the expression of REST-VP16 in myoblasts grown under muscle differentiation conditions blocked entry into the muscle differentiation pathway, countered endogenous REST/NRSF-dependent repression, activated the REST/NRSF target genes, and, surprisingly, activated other neuronal differentiation genes and converted the myoblasts to a physiologically active neuronal phenotype. Furthermore, in vitro differentiated neurons produced by REST-VP16-expressing myoblasts, when injected into mouse brain, survived, incorporated into the normal brain, and did not form tumors. This is the first instance in which myoblasts were converted to a neuronal phenotype. Our results suggest that direct activation of REST/NRSF target genes with a single transgene, REST-VP16, is sufficient to activate other terminal neuronal differentiation genes and to override the muscle differentiation pathways, and they suggest that this approach provides an efficient way of triggering neuronal differentiation in myoblasts and possibly other stem cells.

FoxM1B regulates NEDD4-1 expression, leading to cellular transformation and full malignant phenotype in immortalized human astrocytes.

Our recent studies have shown that the FoxM1B transcription factor is overexpressed in human glioma tissues and that the level of its expression correlates directly with glioma grade. However, whether FoxM1B plays a role in the early development of glioma (i.e., in transformation) is unknown. In this study, we found that the FoxM1B molecule causes cellular transformation and tumor formation in normal human astrocytes (NHA) immortalized by p53 and pRB inhibition. Moreover, brain tumors that arose from intracranial injection of FoxM1B-expressing immortalized NHAs displayed glioblastoma multiforme (GBM) phenotypes, suggesting that FoxM1B overexpression in immortalized NHAs not only transforms the cells but also leads to GBM formation. Mechanistically, our results showed that overexpression of FoxM1B upregulated NEDD4-1, an E3 ligase that mediates the degradation and downregulation of phosphatase and tensin homologue (PTEN) in multiple cell lines. Decreased PTEN in turn resulted in the hyperactivation of Akt, which led to phosphorylation and cytoplasmic retention of FoxO3a. Blocking Akt activation with phosphoinositide 3-kinase/Akt inhibitors inhibited the FoxM1B-induced transformation of immortalized NHAs. Furthermore, overexpression of FoxM1B in immortalized NHAs increased the expression of survivin, cyclin D1, and cyclin E, which are important molecules for tumor growth. Collectively, these results indicate that overexpression of FoxM1B, in cooperation with p53 and pRB inhibition in NHA cells, promotes astrocyte transformation and GBM formation through multiple mechanisms.

Involvement of human chromosome 17 in the control of the metastatic phenotype in melanoma

Fusion of nonmetastatic murine melanoma K1735 C19H cells with metastatic human melanoma A375 C15N cells resulted in a hybrid (termed H7) which was highly metastatic in a nude mouse model. The H7 hybrid retained chromosome 17 as the sole intact human chromosome in the cell. A lung bioassay showed that the K1735 C19H cells were present in the lungs of nude mice with s.c. tumors, yet at 6-weeks after tumor resection, no cells remained in the lung and therefore did not form lung metastases. Examination of various phenotypic properties such as in vivo and in vitro growth demonstrated that phenotypically the H7 hybrid was most like the K1735 C19H cell line except for its metastatic ability. In contrast the H7 hybrid cells containing single or multiple copies of human chromosome 17 with a point mutation at codon 249 (arg-gly) of the p53 gene, readily formed lung metastases. A plasmid containing the human p53 from the H7 hybrid and four other contructs with mutations at codon 143 (val-arg), 175 (arg-his), 249 (arg-ser) and 273 (arg-his) were transfected into K1735 C19H cells. K1735 C19H cells expressing human p53 genes with mutations at codons 249, both the arg-ser mutation and the mutation from the H7 hybrid and 273 produced significantly more lung metastases.^ In vitro assays demonstrated that responses to various cytotoxic and DNA damaging agents varied with the presence of mutant p53 and with the type of agent used. When cultured in mouse lung-conditioned medium, the K1735 C19H cell line was growth-inhibited, while cells containing a mutant human p53 (either on the whole chromosome 17, as in the H7 hybrid cells or from a stably transfected construct) were growth stimulated. Western blot analysis of lung-conditioned media derived from either 6-month or 15-month old mice has detected high levels of soluble Fas ligand in the medium from older animals. Comparison of the levels of Fas receptor on the K1735 C19H cell line and the H7 hybrid were almost identical, but 50% of the K1735 C19H cells were killed in the presence of anti-Fas antibody as opposed to 7% of the H7 hybrid cells. The growth-inhibitory effects of the lung-conditioned medium on the K1735 C19H cells were abrogated by coculture with Fas-Fc, which competes with the Fas ligand for receptor binding. Growth-inhibition of the K1735 C19H was 54% when cultured in 60 $\mu$g/0.2 ml lung-conditioned medium and a control Fc, with only 9% inhibition in 60 $\mu$g/0.2 ml lung-conditioned medium and Fas-Fc. Growth of the H7 cells and K1735 C19H cells transfected with various mutant human p53 genes were unchanged by the presence of either the control Fc or the Fas-Fc. This indicates that the presence of human chromosome 17, and mutant p53 in part protects the cells from Fas:Fas ligand induced apoptosis, and allows the growth of lung metastases. ^

Effects of butyrate on the colon cancer cell phenotype-chemosensitization and upregulation of proteins implicated in cancer progression

Colon cancer is the second leading cause of cancer mortality in the U.S. Surgery is the only truly effective human colon cancer (HCC) therapy due to marked intrinsic drug resistance. The inefficacy of therapies developed for metastatic HCC suggests that advances in colon cancer chemoprevention and chemotherapy will be needed to reduce HCC mortality. The dietary fiber metabolite butyrate (NaB) is a candidate cancer chemopreventive agent that inhibits growth, promotes differentiation and stimulates apoptosis of HCC cells. Epidemiological and experimental studies suggest that dietary fiber protects against the development of HCC, however, recent large prospective trials have not found significant protection. ^ The first central hypothesis of this dissertation project is that the diversity of phenotypic changes induced by NaB in HCC cells includes molecular alterations that oppose its chemopreventive action and thereby limit its efficacy. We investigated the effect of NaB on the expression/activity of epidermal growth factor receptor (EGFR) and cyclooxygenase-2 (COX-2) in HCC HT29 cells. NaB treatment induced a 13-fold increase in EGFR expression in concert with its chemopreventive action in vitro, i.e., induction of growth suppression and G1 arrest, apoptosis and a differentiated phenotype. NaB-induced EGFR was active based on multiple lines of evidence. The EGFR was: (1) heavily phosphorylated at Tyrosine (P-Tyr); (2) associated with the cytoskeleton; (3) localized at the cell surface, and activated in response to EGF; and (4) NaB treatment of the cells induced activation of the EGFR effector Erk1/2. NaB treatment also induced a 7-fold increase in COX-2 expression. The NaB-induced COX-2 was active based on significantly increased PGE2 production. ^ The second central hypothesis is that NaB treatment would render HCC cells more chemosensitive to chemotherapy agents based on the increased apoptotic index induced by NaB. NaB treatment chemosensitized HT29 cells to 5-FU and doxorubicin, despite increases in the expression of P-glycoprotein and a related drug resistance protein (MRP). ^ These results raise the intriguing possibility that the chemopreventive effects of fiber may require concomitant treatment with EGFR and/or COX-2 inhibitors. Similarly, NaB may be a rational drug to combine with existing chemotherapeutic agents for the management of advanced HCC patients. ^

Phenotype of capillaries in skeletal muscle of nNOS-knockout mice

Because neuronal nitric oxide synthase (nNOS) has a well-known impact on arteriolar blood flow in skeletal muscle, we compared the ultrastructure and the hemodynamics of/in the ensuing capillaries in the extensor digitorum longus (EDL) muscle of male nNOS-knockout (KO) mice and wild-type (WT) littermates. The capillary-to-fiber (C/F) ratio (-9.1%) was lower (P ≤ 0.05) in the nNOS-KO mice than in the WT mice, whereas the mean cross-sectional fiber area (-7.8%) and the capillary density (-3.1%) varied only nonsignificantly (P > 0.05). Morphometrical estimation of the area occupied by the capillaries as well as the volume and surface densities of the subcellular compartments differed nonsignificantly (P > 0.05) between the two strains. Intravital microscopy revealed neither the capillary diameter (+3% in nNOS-KO mice vs. WT mice) nor the mean velocity of red blood cells in EDL muscle (+25% in nNOS-KO mice vs. WT mice) to significantly vary (P > 0.05) between the two strains. The calculated shear stress in the capillaries was likewise nonsignificantly different (3.8 ± 2.2 dyn/cm² in nNOS-KO mice and 2.1 ± 2.2 dyn/cm² in WT mice; P > 0.05). The mRNA levels of vascular endothelial growth factor (VEGF)-A were lower in the EDL muscle of nNOS-KO mice than in the WT littermates (-37%; P ≤ 0.05), whereas mRNA levels of VEGF receptor-2 (VEGFR-2) (-11%), hypoxia inducible factor-1α (+9%), fibroblast growth factor-2 (-14%), and thrombospondin-1 (-10%) differed nonsignificantly (P > 0.05). Our findings support the contention that VEGF-A mRNA expression and C/F-ratio but not the ultrastructure or the hemodynamics of/in capillaries in skeletal muscle at basal conditions depend on the expression of nNOS.