ARTICLE IN PRESS Research in Veterinary Science ■■ (2015) ■■–■■

Contents lists available at ScienceDirect

Research in Veterinary Science j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / r v s c

Alpha-smooth muscle actin and serotonin receptors 2A and 2B in dogs with myxomatous mitral valve disease S.E. Cremer a, S.G. Moesgaard b, C.E. Rasmussen b, N.E. Zois c, T. Falk d, M.J. Reimann a, S. Cirera e, H. Aupperle f, M.A. Oyama g, L.H. Olsen a,* a

Department of Veterinary Disease Biology, University of Copenhagen, Frederiksberg, Denmark Novo Nordisk A/S, Maaloev, Denmark c Department of Clinical Biochemistry, Copenhagen University Hospital Rigshospitalet, Copenhagen, Denmark d Din Veterinär, Helsingborg, Sweden e Department of Veterinary Clinical and Animal Sciences, University of Copenhagen, Frederiksberg, Denmark f Laboklin, Bad Kissingen, Germany g Department of Clinical Studies, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA, USA b

A R T I C L E

I N F O

Article history: Received 4 September 2014 Accepted 14 March 2015 Keywords: Heart valve disease CKCS 5-HT2AR 5-HT2BR MMVD Dog

A B S T R A C T

Canine Myxomatous mitral valve disease (MMVD) is an age-related disease. Serotonin (5-HT) is implicated in the pathogenesis as locally-produced or platelet-derived. Involvement of the 5-HT2A receptor (R) and 5-HT2BR in the induction of myxomatous-mediating valvular myofibroblasts (MF) has been suggested. In an age-matched population of dogs with non-clinical and clinical MMVD, the objectives were to investigate (1) gene expression of 5-HT2AR and 5-HT2BR, (2) protein expression and spatial relationship of 5-HT2AR, 5-HT2BR and MF in the mitral valve (MV) and the cardiac anterior papillary muscle (AP) and (3) serum 5-HT concentrations. Gene expression of 5-HT2BR was significantly higher in MV and AP among dogs with clinical MMVD. This was not found for 5-HT2BR protein expression, though association of 5-HT2BR with myxomatous pathology and co-localization of 5-HT2BR and MF in MV and AP support a functional relationship, perhaps perpetuation of clinical MMVD. 5-HT2AR-expression and serum 5-HT showed no differences between groups. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction Myxomatous mitral valve disease (MMVD) is a common cause of morbidity and mortality among dogs (Buchanan, 1977; Detweiler et al., 1961; Egenvall et al., 2006; Haggstrom et al., 2009). The exact molecular mechanisms involved in the initiation and progression of the disease are not fully understood but serotonin (5hydroxytryptamine, 5-HT) is suggested to be involved in the pathogenesis (Orton et al., 2012; Oyama and Levy, 2010). Tissue studies from dogs have shown increased expression of 5-HT2BR in myxomatous mitral valves (Disatian and Orton, 2009; Oyama and Chittur, 2006), and tryptophan hydroxylase-1 (TPH1), the ratelimiting enzyme in 5-HT synthesis, was increased in early as well as late stage disease (Disatian et al., 2010). It has moreover been suggested that 5-HT, via transforming growth factor-beta (TGF-β), induces transformation of valvular interstitial cells (VIC) into active myofibroblasts (Connolly et al., 2009; Disatian and Orton, 2009;

* Corresponding author. Department of Veterinary Disease Biology, Faculty of Health and Medical Sciences, University of Copenhagen, Ridebanevej 9, 1870 Frederiksberg C, Denmark. Tel.: +45 35333179; fax: +45 35332755. E-mail address: [email protected] (L.H. Olsen).

Hutcheson et al., 2012). Activated myofibroblast are commonly characterized by the presence of alpha-smooth muscle actin (α-SMA) and are believed to be main mediators of myxomatous degeneration (Liu et al., 2007; Rabkin et al., 2001). Both the 5-HT2AR and 5-HT2BR have been linked to VIC transformation (Connolly et al., 2009; Disatian et al., 2010) and a hypothesis of increased valvular 5-HT synthesis leading to VIC transformation via the 5-HT2BR and TGF-β has been suggested (Disatian and Orton, 2009; Orton et al., 2012). In humans, the 5-HT2BR is strongly implicated in druginduced myxomatous valvulopathy (Rothman et al., 2000) and 50% of people suffering from metastatic serotonin-producing tumors show evidence of valvular disease with similarities to canine MMVD (Fox and Khattar, 2004; Gustafsson et al., 2008). In dogs, serum 5-HT has moreover been associated with MMVD-severity (Ljungvall et al., 2013). As MMVD progresses myocardial remodeling takes place (Haggstrom et al., 2009). It has been suggested that myocardial changes of MMVD, like fibrosis and fibromuscular narrowing of intramural coronary arteries (“small vessel disease”), might relate to the same primary disease process as in the mitral valve (Burke et al., 1997; Detweiler, 1989; Falk et al., 2006; Morales et al., 1992). Another marker of myocardial remodeling, cardiac Troponin-I, has been associated with MMVD severity in dogs (Ljungvall et al., 2010) and

http://dx.doi.org/10.1016/j.rvsc.2015.03.020 0034-5288/© 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: S.E. Cremer, et al., Alpha-smooth muscle actin and serotonin receptors 2A and 2B in dogs with myxomatous mitral valve disease, Research in Veterinary Science (2015), doi: 10.1016/j.rvsc.2015.03.020

ARTICLE IN PRESS S.E. Cremer et al./Research in Veterinary Science ■■ (2015) ■■–■■

2

with post-mortem changes of fibrosis and intramural arterial narrowing (lumen area ratio, LAR) (Falk et al., 2013). The etiology of myocardial changes and remodeling in MMVD is poorly understood, and it is unknown to what extent these changes are primary, secondary or age-related. Until recently, no studies have looked at myocardial 5-HT signaling in MMVD. However, new data show increased myocardial 5-HT concentrations in dogs with MMVD compared to dogs without heart disease or heart disease unrelated to MMVD (Cremer et al., 2014a). In addition, 5-HT induced activation of myocardial fibroblasts has been documented in vitro, a response that could be antagonized by 5-HT2AR antagonism (Yabanoglu et al., 2009). MMVD is a highly age-related disease (Haggstrom et al., 1992; Olsen et al., 1999; Whitney, 1974), but no studies have addressed 5-HT markers in groups of age-matched dogs. In humans, the proportion of α-SMA positive cells has been found to decrease with age, whereas in pigs the picture is the opposite (Connell et al., 2012; Stephens and Grande-Allen, 2007). No studies have addressed this question in dogs. However, the influence of age is a common and unavoidable factor in the comparison of early and late-stage disease (Connell et al., 2012). In order to minimize the influence of age, the current study involves only older age-matched dogs. The study analyzes RNA transcription and protein expression of 5-HT2AR and 5-HT 2B R in the mitral valve and myocardium. Moreover, the protein expression of 5-HT 2A R and 5-HT 2B R is described immunohistochemically in relation to α-SMA positive cells and histopathology. Lastly, serum 5-HT concentration is analyzed. 2. Materials and methods 2.1. Animals and groups Privately owned dogs, associated with the University of Copenhagen (UCPH) through an echocardiographic research database, were recruited upon time of elective euthanasia. The diagnosis of MMVD was based on echocardiographic findings and the dogs were grouped into a control group, an asymptomatic MMVD group and a symptomatic MMVD group according to the heart disease classification system of the American College of Veterinary Internal Medicine (ACVIM). In brief, the control group consisted of non-predisposed dogs as well as predisposed dogs without clinical heart disease and no presence of murmur (ACVIM group A). The asymptomatic MMVD group included dogs with presence of heart murmur but without structural remodeling of the heart, defined as left atrial to aortic root ratio (LA/Ao) of maximum 1.5 (ACVIM group B1). No asymptomatic dogs had an LA/Ao >1.5 and, accordingly, group B2 dogs were not represented in the study. The symptomatic MMVD group included dogs in clinical heart failure (ACVIM groups C and D). Other cardiac disease and renal failure were exclusion criteria. Prior to euthanasia and upon owner consent, dogs underwent clinical examination, blood sampling and echocardiographic examination. All echocardiographic examinations and blood samplings were performed within 4 months of euthanasia. One or more dogs have been included in previous studies (Cremer et al., 2014a; Moesgaard et al., 2014; Rasmussen et al., 2011, 2012, 2014; Zois et al., 2012a, 2012b, 2013). 2.2. Sample collection Blood was collected from the jugular vein. Serum was obtained and frozen at −80 °C until later 5-HT enzyme immunoassay (EIA) analyses. The heart was collected post-mortem and for quantitative real-time PCR (qPCR), the posterior mitral valve (MV) leaflet and anterior papillary muscle (AP) of the left ventricle were placed in RNA-later at 5 °C for 24 hours and then stored at −20 °C. For

immunohistochemistry (IHC) and histopathology, the anterior MV leaflet and AP of left ventricle were placed in formalin for 24–72 hours and then paraffin-embedded. In five dogs, the anterior mitral valve leaflet was sampled for another study (Cremer et al., 2014a) and not available for IHC and the posterior leaflet was sampled instead.

2.3. Echocardiography Echocardiographic examinations were performed by one observer (LHO) according to standard recommendations (Thomas et al., 1993) and as previously described (Reimann et al., 2014), using a Vivid i or a Vivid 7 Pro ultrasound echocardiographic units (GE Medical Systems) with 3S and 5S transducers. Echocardiographic parameters measured included mitral regurgitation (MR), LA and Ao internal dimensions, end-diastolic interventricular septum (IVSd), end-systolic interventricular septum (IVSs), end-diastolic left ventricular internal (LVIDd) and end-systolic left ventricular internal (LVIDs) dimensions. All recordings were measured by the same observer (LHO), using EchoPAC software (PC Version 112, GE Medical Systems). Mitral regurgitation was assessed relative to the size of the left atrium and rounded to the nearest 5% (Pedersen et al., 1999) and LA/Ao was calculated (Häggström et al., 1994). Systolic and diastolic left ventricular diameters (LVIDs and LVIDd, respectively) were indexed to body weight accordingly (Cornell et al., 2004): iLVIDs = LVIDs/body weight (BW)0.315 and iLVIDd = LVIDd/BW0.294 and fractional shortening (FS) was calculated using LVIDd and LVIDs (Lombard, 1984).

2.4. Histopathology and immunohistochemistry Four micrometer thickness sections were prepared and stained with hematoxylin (HE) and modified Picrosirius red stain for valvular histopathology and with HE and Masson trichrome for myocardial fibrosis and intramural coronary narrowing (LAR). Valvular and myocardial histopathological examinations were each performed by one observer (HA and TF, respectively) as previously described (Aupperle et al., 2009a; Falk et al., 2006). Alpha-SMA, 5-HT2AR and 5-HT2BR IHC staining was performed as described previously (Cremer et al., 2015a). Briefly, slides were deparaffinized, rehydrated and rinsed with Tris-buffered saline (TBS) with 0.01% Tween-20 prior to staining. TBS was used following all incubation steps. Antigen retrieval was performed using high-pH retrieval (Dako Target Retrieval Solution, pH9) in 95 °C steaming chamber. Hydrogen peroxide (0.03%) blocked endogenous peroxidase activity and was followed by primary antibody incubation. All antibodies were mouse anti-human monoclonal and were: α-SMA antibody (α-Actin (1A4): sc-32251), 5-HT2BR antibody (SR-2B (6B1), sc-135568) and 5-HT2AR antibody (SR-2A (A-4), sc-166775) (all Santa Cruz Biotechnology, Santa Cruz, CA, USA). Antibodies were diluted with Antibody Diluent with Background Reducing Components (Dako, Carpenteria, CA, USA) yielding final titers of 1:600 (αSMA), 1:400 (5-HT2AR) and 1:50 (5-HT2BR). Universal negative control mouse (DAKO, Carpenteria, CA, USA) was used to ensure antibody specificity. Anti-mouse peroxidase-labeled polymer (EnVision+ System-HRP (DAB); DAKO, Carpenteria, CA, USA) was linked to the primary antibody and 3,3′-diaminobenzidine (DAB) chromogen solution in substrate buffer (pH 7.5) visualized peroxidase activity. Sections were counterstained with Mayer’s hematoxylin (DAKO, Carpenteria, CA, USA) and lastly, dehydrated, dried and mounted. Assessments of staining were (1) graded subjectively as: none (no positive cells), mild (50% positive cells) (Cremer et al., 2015a), and (2) described according to spatial staining pattern.

Please cite this article in press as: S.E. Cremer, et al., Alpha-smooth muscle actin and serotonin receptors 2A and 2B in dogs with myxomatous mitral valve disease, Research in Veterinary Science (2015), doi: 10.1016/j.rvsc.2015.03.020

ARTICLE IN PRESS S.E. Cremer et al./Research in Veterinary Science ■■ (2015) ■■–■■

2.5. RNA isolation, cDNA synthesis, primer design and qPCR RNA isolation, cDNA synthesis, primer design and qPCR were performed as previously described (Cremer et al., 2015a). In brief, approximately 50 mg of tissue (MV or AP) was used for RNA extraction (RNeasy fibrous tissue mini kit, Qiagen-Nordic Denmark) and RNA quantity and quality were checked (NanoDrop 1000 machine, Thermoscientific), accepting a 260/280 ratio of 1.8–2.2. Integrity was assessed (ExperionTM system, Biorad using Eukaryote Total RNA RNA StdSens kit) accepting a RNA quality index (RQI) above 6.8. RNA transcription to cDNA (Biometra T Gradient) was done in duplicate according to manufacturer’s recommendations (Promega, M1705). Briefly, 1 μg total RNA was added to 5× M-MLV RT buffer (5 μL), dNTP 10 mM (1.3 μL), Random hexamer primer 2 μg/μL (0.2 μL), Oligo (dt) 0.5 μg/μL (0.4 μL), RNase inhibitor (0.8 μL), M-MLV RT enzyme (1.0 μL) (all reagents from Promega, M1705) following manufacturer’s recommendations. The reverse transcription reaction was performed on a Biometra T Gradient machine using the following thermal protocol: 25 °C for 10 min, 42 °C for 60 min, 95 °C for 5 min, then cooled to 4 °C (Cremer et al., 2015a). cDNA samples were diluted 1:10 and stored at −80 °C. Primers were designed for 5-HT2AR and 5-HT2BR using Primer 3 software (http://primer3.ut.ee/) and spanned over an intron. Reference genes (glyceraldehyde-3-phosphate (G3P), ribosomal protein L32 (RPL32), hypoxanthine phosphoribosyltransferase 1 (HPRT1) and TATA-box binding protein (TBP)) were selected based on high expression stability in canine tissue (Brinkhof et al., 2006; Peters et al., 2007). Cq, melting curve and primer efficiency were assessed and efficiencies between 85 and 110% were accepted. Table 1 lists all primer characteristics. Mastermix for the qPCR reaction was mixed with cDNA and run in duplicates, accepting a standard deviation (SD) of maximum 0.4. In brief, the reaction was performed in a total volume of 10 μL; 1 μL dH2O, 5 μL SYBR Green (Roche, 4887352001), 1 μL 10 μM forward primer, 1 μL 10 μM reverse primer, 2 μL of 1:10 diluted cDNA were mixed in white PCR plates (LightCycler480 Multiwell Plate 96, Roche, 04729692-001). The samples were run on LightCycler480 using the following thermal profile: 45 cycles of 95 °C for 5 min, 60 °C for 10 s, 72 °C for 20 s and a melting curve analysis (55–95 °C) at the end

Table 1 Primer characteristics. Gene Symbol

Oligo Sequence (5′→3′)

5-HT2AR F: CACCATAGCCGATTCAACTC R: GTTATCGTCGGCAAGTAGGC 5-HT2BR F: GTGGCTGATCTGCTAGTTGG R: CAGAAATGGCACAGAGATGC G3P* F: TGTCCCCACCCCCAATGTATC R: CTCCGATGCCTGCTTCACTACCTT RPL32* F: TGGTTACAGGAGCAACAAGAAA R: GCACATCAGCAGCACTTCA HPRT1* F: CACTGGGAAAACAATGCAGA R: ACAAAGTCAGGTTTATAGCCAACA TBP* F: CTATTTCTTGGTGTGCATGAGG R: CCTCGGCATTCAGTCTTTTC

Amplicon Tm (°C) Efficiency (%) Length 157

60

103

159

63

105

100

58

105

100

87

100

123

86

95

96

85

97

Primer characteristics of target genes and reference genes (*). Reference genes were selected based on high expression stability in canine tissue (Brinkhof et al., 2006; Peters et al., 2007). 5-HTR: serotonin receptor, G3P: glyceraldehyde-3-phosphate, HPRT1: hypoxanthine phosphoribosyltransferase 1, RPL32: ribosomal protein L32, TBP: TATA box binding protein, Tm: melting temperature.

3

of the last cycle (Cremer et al., 2015a). Results of gene expression were processed using GenEx software. 2.6. Enzyme immunoassay Human EIA 5-HT research assay (DEE8900, Demeditec Diagnostics GmbH, Germany) measured serum 5-HT concentration according to manufacturer’s instructions. The assay was validated for use in dogs in regards to coefficient of variation (CV). The intra-assay CV was assessed from 10 repeated measurements of four dog samples within the same assay. The inter-assay CV was evaluated on samples from two dogs measured in three separate assays. 2.7. Statistical analysis Due to the small group sizes, continuous variables were analyzed using non-parametrical Kruskal–Wallis test of Wilcoxon rank sums scores and categorical variables by means of Fisher’s exact test. Post hoc group-wise analyses using Fisher’s exact test were performed when the overall p-value was statistical significant and Bonferroni correction was applied. SAS software (version 9.3) was used for all statistical analysis and a p-value below 0.05 was considered significant. Results are listed as median and interquartile range. 3. Results 3.1. Animals and groups Out of 27 dogs, five were excluded from the study. One was excluded due to renal failure and three were excluded due to technical problems in tissue collection. One dog could not be assigned to a group, as the echocardiographic examination was incomplete. Table 2 lists histopathological assessment of MV and AP, Whitney grading of MV, descriptive statistics, echocardiographic parameters and ausculatory findings. Cavalier King Charles Spaniels were represented with 17/22 (77%), but with equal distribution among groups. The remaining five dogs were Bichon, Boston terrier, Poodle, DanishSwedish farm dog and one mixed-breed dog. Of these, three were located to group 1 and two were located to group 3. A total of 11 dogs received one or more of the following medications: furosemide (n = 4), pimobendan (n = 4), benazepril (n = 3), carprofen (n = 2), meloxicam (n = 2), spironolactone (n = 1), hydralazine (n = 1), diazepam (n = 1), penicillin (n = 1) and unspecified (cardiac) drug (n = 1). Reduction in group numbers (Table 2) was caused by inadequate tissue quality of the final slides. 3.2. qPCR of 5-HT2AR and 5-HT2BR In both the MV and the AP, gene expression of 5-HT2BR was significantly different among groups (both overall p = 0.02) (Fig. 1). In the MV, 5-HT2BR was significantly higher in the symptomatic MMVD group versus the control group (p = 0.03). In AP, 5-HT2BR expression was significantly higher in the symptomatic MMVD group versus the asymptomatic group (p = 0.03). For two dogs in the symptomatic group, tissue for PCR analysis was not available. 3.3. Immunohistochemistry of α-SMA 5-HT2AR and 5-HT2BR Table 3 lists the group-wise results for 5-HT2BR and α-SMA in MV and AP. There were no group-differences for either marker in MV or AP, but the markers showed different cellular staining

Please cite this article in press as: S.E. Cremer, et al., Alpha-smooth muscle actin and serotonin receptors 2A and 2B in dogs with myxomatous mitral valve disease, Research in Veterinary Science (2015), doi: 10.1016/j.rvsc.2015.03.020

ARTICLE IN PRESS S.E. Cremer et al./Research in Veterinary Science ■■ (2015) ■■–■■

4

Table 2 Group summary.

Histopathology (none/min./mild/mod./sev.) Whitney(0/1/2/3/4) Myocardial fibrosis(0/0.5/1/1.5/2/2.5/3) Myocardial LAR Age(years) Weight(kg) Gender(M/F) CKCS/non-CKCS Murmur(0/1/2/3/4/5/6) Medication(Y/N) MR(%) LA/Ao(mm) IVSd(mm) IVSs(mm) LVIDd(mm) LVIDs(mm) iLVIDd iLVIDs FS(%)

Control group (n = 6)

Asymptomatic MMVD (n = 7)

Symptomatic MMVD (n = 9)

N (final)

0/0/2/1/2 0/6/0/0/0B1, C 2/1/3/0/0/0/0 0.28(0.25–0.36) 9(7–11) 9.2(8.4–9.5) 2/4 3/3 6/0/0/0/0/0/0B1, C 0/6C 15 (0–25)C 1.20 (1.20–1.30)C 7.20 (7.20–7.60) 9.10 (9.00–9.20) 25.50 (24.20–27.20)C 19.10 (18.40–19.80)C 9.40 (9.24–10.13)C 13.46 (12.73–14.23)C 0.80 (0.70–1.10)C

1/1/0/2/2 0/1/2/1/1A 0/0/6/1/0/0/0C 0.32(0.29–0.34) 10(8–13) 8.7(8.2–10.6) 3/4 7/0 0/1/2/4/0/0/0A, C 2/5C 60 (15–100)C 1.30 (1.20–1.50)C 7.80 (7.00–8.40) 10.80 (10.40–11.60) 28.80 (26.20–37.80)C 18.80 (16.40–24.00)C 9.51 (8.30–10.39)C 15.08 (14.16–18.59)C 1.40 (0.90–1.80)C

0/2/0/2/5 0/0/0/5/2A 1/1/1/0/3/1/1B1 0.20(0.19–0.31) 11(10–13) 8.1(7.9–9.7) 7/2 7/2 0/0/0/0/2/5/2A, B1 9/0A, B1 100 (100-100)A, B1 2.00 (1.80–2.60)A, B1 7.60 (6.80–9.00) 10.00 (8.40–11.40) 42.00 (39.40–47.00)A, B1 25.40 (23.20–28.40)A, B1 13.00 (12.11–13.26)A, B1 22.87 (21.30–24.10)A, B1 2.10 (1.60–2.30)A, B1

20 18 21 21 22 22 22 22 22 22 22 22 22 22 22 22 22 22 22

The table illustrates the results from MMVD dogs grouped according to ACVIM grouping criteria. It includes valvular pathology including histopathology and Whitney score, myocardial (anterior papillary muscle) fibrosis, myocardial intramural narrowing (LAR), descriptive characteristics and echocardiographic variables including mitral regurgitation (MR), ratio of left atrium to aortic root (LA/Ao), end-diastolic interventricular septum (IVSd), end-systolic interventricular septum (IVSs), end-diastolic left ventricular internal dimension (LVIDd), end-systolic left ventricular internal dimension (LVIDs), indexed LVIDd (iLVIDd), indexed LVIDs (iLVIDs) and fractional shortening (FS). Continuous variables are listed as medians and interquartile ranges. Within each row, superscripts indicate which groups are statistically significantly different. Superscripts that are not italicized indicate that the respective p is