- Email: [email protected]
Developmental pathways and endothelial to mesenchymal transition in canine myxomatous mitral valve disease
Accepted Manuscript Title: Developmental pathways and endothelial to mesenchymal transition in canine myxomatous mitral valve disease Author: Chi-Chien Lu, Meng-Meng Liu, Michael Clinton, Geoff Culshaw, David J. Argyle, Brendan M. Corcoran PII: DOI: Reference:
S1090-0233(15)00335-4 http://dx.doi.org/doi: 10.1016/j.tvjl.2015.08.011 YTVJL 4593
To appear in:
The Veterinary Journal
Accepted date:
6-8-2015
Please cite this article as: Chi-Chien Lu, Meng-Meng Liu, Michael Clinton, Geoff Culshaw, David J. Argyle, Brendan M. Corcoran, Developmental pathways and endothelial to mesenchymal transition in canine myxomatous mitral valve disease, The Veterinary Journal (2015), http://dx.doi.org/doi: 10.1016/j.tvjl.2015.08.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Developmental pathways and endothelial to mesenchymal transition in canine myxomatous mitral valve disease Chi-Chien Lu a, Meng-Meng Liu a, Michael Clinton b, Geoff Culshaw a, David J. Argyle a, Brendan M. Corcoran a,*
a
Royal (Dick) School of Veterinary Studies and The Roslin Institute, the University of Edinburgh, Easterbush, Roslin, Mid-Lothian, Scotland, EH25 9RG, UK b
The Roslin Institute, the University of Edinburgh, Easterbush, Roslin, Mid-Lothian, Scotland, EH25 9RG, UK
* Corresponding author. Tel.: +44 131 651 9100. E-mail address: [email protected] (B.M. Corcoran).
Page 1 of 23
22 23 24 25 26 27 28 29 30 31 32 33 34 35
Highlights
This is the first report of involvement of development signalling pathways and endothelial to mesenchymal transition in canine myxomatous mitral valve disease (MMVD)
The finding is significant in that it recognises mechanisms previously reported for fibrotic conditions and cancer metastasis, rather than a degenerative condition such as MMVD
Information is provided on biological mechanisms that have therapeutic potential for new drug discovery. Abstract
36
Epithelial to mesenchymal transition (EMT), and the cardiovascular equivalent,
37
endothelial to mesenchymal transition (EndoMT), contribute to a range of chronic
38
degenerative diseases and cancer metastasis. Chronic valvulopathies exhibit some features of
39
EndoMT and activation of developmental signalling pathways, such as osteogenesis and
40
chondrogenesis, expression of cell differentiation markers, basement membrane damage and
41
endothelial transformation. The aim of the present study was to investigate the potential role
42
of developmental mechanisms in canine myxomatous mitral valve disease (MMVD), using a
43
combination of transcriptomic array technology, RT-PCR and immunohistochemistry.
44 45
There was significant differential expression for genes typically associated with
46
valvulogenesis and EndoMT, including markers of inflammation (IL6, IL18 and TLR4),
47
basement membrane disarray (NID1, LAMA2 and CTSS), mesenchymal and endothelial cell
48
differentiation (MYH11 and TAGLN) and EndoMT (ACTA2, SNAI1, CTNNB1, HAS2, CDH5,
49
and NOTCH1), with fold changes from +15.35 (ACTA2) to -5.52 (LAMA2). These changes in
50
gene expression were confirmed using RT-PCR, except for HAS2. In silico analysis identified
51
important gene networks and canonical pathways in MMVD that have associations with
52
development and organogenesis, including inflammation, valve morphogenesis and EMT, as
53
well
as
components
of
the
basement
membrane
and
extra-cellular
matrix.
Page 2 of 23
54
Immunohistochemistry identified changes in expression of hyaluronic acid synthase (Has2),
55
Snai1, α-smooth muscle actin (α-SMA) and VE-cadherin (CDH5), and co-expression of Has2
56
with α-SMA. These research findings strongly suggest involvement of developmental
57
signalling pathways and mechanisms, including EndoMT, in the pathogenesis of canine
58
MMVD.
59 60
Keywords: Endothelial to mesenchymal transition; Myxomatous mitral valve disease;
61
Canine; Transcriptomics
62
Page 3 of 23
63
Introduction
64
Myxomatous mitral valve disease (MMVD) is the most common acquired heart
65
disease affecting dogs (Buchanan, 1977; Borgarelli and Buchanan, 2012). At a structural and
66
cellular level, there is loss of collagen, accumulation of glycosaminoglycans, proliferation of
67
α-smooth muscle actin (α-SMA) and embryonic smooth muscle myosin (Smemb) positive
68
cells adjacent to the valve endothelium, with detachment and loss of endothelial cells (Rabkin
69
et al., 2001; Disatian et al., 2010; Hadian et al., 2010; Han et al., 2010). The histopathological
70
appearance of the diseased valve stroma and some of the cellular changes present, including
71
the appearance of the myxoid stroma itself, increased numbers of α-SMA positive cells,
72
basement membrane damage and changes in the endothelium, are reminiscent of processes
73
involved in valvulogenesis (Black et al., 2005; Han et al., 2010; Hinton and Yutzey, 2011).
74 75
Osteogenesis and chondrogenesis have been identified in aortic and mitral valves,
76
suggesting that developmental signalling pathways can be activated in different chronic
77
valvulopathies (Cheek et al., 2012; Orton et al., 2012; Garside et al., 2013). Osteogenic
78
differentiation capacity has been demonstrated with clonally expanded sheep mitral valvular
79
endothelial cells (VECs), suggesting that a population of multi-lineage progentior cells exist
80
in the valve endothelium, with the potential for mesenchymal differentiation. These may form
81
part of a sub-set of VECs that can replenish valvular interstitial cells (VICs) as a consequence
82
of life-long cell loss and in response to disease (Bischoff and Aikawa 2011; Wyler-Sears et
83
al, 2011). Endothelial-to-mesenchymal transition (EndoMT) has also been shown to be
84
activated by inflammatory cytokines in adult porcine aortic VECs (Mahler et al., 2013).
85 86
EndoMT is involved in valvulogenesis and there are clear examples of a role for
87
EndoMT and epithelial to mesenchymal transition (EMT) in a variety of diseases such as
Page 4 of 23
88
chronic fibrosis of the lung, kidney and heart and in cancer metastasis (Li and Bertram, 2010;
89
Piera-Velazquez et al., 2011; Kovacic et al., 2012). EndoMT is closely regulated by a wide
90
range of signalling pathways and transcription factors, including vascular endothelial growth
91
factor (VEGF) (cell proliferation), NOTCH (induction of EndoMT) and Snail and Slug (VE-
92
cadherin, α-SMA and fibronectin expression) (Armstrong and Bischoff, 2004; Butcher and
93
Markwald, 2007; Hinton and Yutzey, 2011). The hallmark of EndoMT is loss of the VEC
94
marker CD31 and cell-to-cell contact (assessed by CDH5/VE-cadherin expression), and
95
upregulation of α-SMA expression (Bischoff and Aikawa, 2011; Kovacic et al., 2012).
96 97
Members of the transforming growth factor (TGF)-β/bone morphogenic protein
98
(BMP) superfamily also mediate cell differentiation, endothelial cell separation and
99
degradation of the basal lamina, permitting migration of transformed endothelial cells (Boyer
100
et al., 1999; Song et al., 2000; Butcher and Markwald, 2007). BMPs are important in valve
101
development and their expression appears prior to the onset of EndoMT, triggering
102
production of TGF-β, hyaluronic acid synthase 2 (Has2), NOTCH and the transcription
103
factors Snail and Twist (Chen and Massagué, 1999; Ma et al., 2005). The appearance of
104
hyaluronic acid allows invasion and further migration of cells into the cardiac jelly and
105
promotes cell proliferation and differentiation by inducing expression of PI3 kinase and
106
ErbB2/3 (Mjaatvedt et al., 1998; Garside et al., 2013).
107 108
There is evidence of a role for EndoMT/EMT in a range of chronic diseases and some
109
of the stromal and cellular changes seen in MMVD suggest a tissue type similar to that seen
110
in embryonic development of the valve. Therefore, the aim of the present study was to
111
examine the possibility of re-activation and recruitment of developmental processes in canine
Page 5 of 23
112
MMVD, based on analysis of transcriptomic data (Lu et al., 2015), further investigated by
113
real time quantitative PCR and immunohistochemistry.
114 115
Materials and methods
116
Tissue samples
117
All tissue samples were collected with informed owner consent and the study
118
conformed to national (UK) and institutional ethical guidelines for the use of animals and
119
their tissues in research. For immunohistochemistry, myxomatous mitral valve leaflets (n =
120
14) were collected at necropsy from cavalier King Charles Spaniels (CKCS; n = 9) and mixed
121
breed dogs (n = 5) presented to the Hospital for Small Animals, Royal (Dick) School of
122
Veterinary Study, the University of Edinburgh. Control tissue from anterior mitral valve
123
leaflets (n = 10) were collected from young adult dogs that had been euthanased for reasons
124
other than cardiac disease. All samples were immediately fixed in 4% paraformaldehyde.
125
Valves were assessed for the presence of gross lesions by two of the three investigators
126
(CCL, BMC, GC) and classified according to an established scheme (Whitney, 1974). For
127
transcriptomic analysis, valvular tissues from four CKCS and four control dogs had been
128
assessed and the results reported by Lu et al. (2015).
129 130
RNA extraction
131
One hundred milligrams of tissue were minced, snap-frozen in liquid nitrogen and
132
pulverised for 1 min at 2000 oscillations per min in a liquid nitrogen-cooled dismembranator
133
(Mikro-Dismembrator Vessel, Braun Biotech International). One millilitre of TriReagent
134
(Sigma–Aldrich) and 200 µL of chloroform were added, followed by centrifugation (12,000 g
135
for 10 min). RNA extraction and contaminant genomic DNA digestion were carried out using
136
commercially available kits (RNeasy Mini Kit and RNase-Free DNase Set, Qiagen). The
Page 6 of 23
137
RNA purity was analysed by spectrophotometry (NanoDrop 1000, Thermo Scientific;
138
260/280 absorbance ratio ~2) and RNA quality and RNA integrity number (RIN ≥7)
139
determined by electropherogram (Agilent 2100 Bioanalyser). The RNA samples were stored
140
at -80 °C prior to further processing.
141 142
Microarray hybridization and analysis
143
The GeneChip WT Terminal Labelling Kit (Affymetrix), combined with WT
144
Expression Kit (Ambion), were used to generate amplified and biotinylated sense-strand
145
DNA targets for the Affymetrix Canine Gene 1.0 ST Array. The two cycle RNA-cDNA
146
amplification, hybridization, and chip scanning were undertaken by ARK-Genomics (Roslin
147
Institute).
148 149
For differential gene expression analysis, all the raw Affymetrix CEL files were
150
imported into Partek Genomic Suite version 6.6 software (Partek). Robust Multi-array
151
Average (RMA) was used for data normalisation and final summarisation, and Student’s t test
152
was used to identify differentially expressed genes comparing the two groups (cases and
153
controls). Network pathway analysis, including biological functions identification, was
154
undertaken using the Ingenuity Pathways Analysis (IPA) database. 1 Normalised signal
155
intensity data for whole probe sets were also produced by Affymetrix Expression Console.
156 157
Real time quantitative PCR
158
Six RNA samples (three CKCS affected with MMVD and three controls), that had
159
been analysed by microarray, were assessed by quantitative real-time PCR. Assays were
160
designed for four candidate genes (determined by signal intensity), six differentially
1
See: http://www.ingenuity.com/products/ipa (accessed 6 August 2015)
Page 7 of 23
161
expressed genes, and one reference gene (MRPS25) that had previously been validated for
162
normalising real-time quantitative PCR data generated from canine heart valve tissue (Yang
163
et al., 2012) (see Appendix: Supplementary Table S1). Primers (MWG Biotech) and labelled
164
probes (Roche Diagnostics) were synthesised and real-time quantitative PCR assays were
165
performed in triplicate in 96 well plates (LightCycler 480; Roche Diagnostics) with one no-
166
template control for each sample. Reaction volumes consisted of 5 μL LightCycler 480
167
Probes Master (Roche Diagnostics) (See Appendix: Supplementary Table S1 and
168
Supplementary File 1). Real-time quantitative PCR data were analysed using LightCycler 480
169
Basic Software (Roche Diagnostics). The reference gene (MRPS25) was used to normalise
170
the relative expression of the target genes, using the comparative Ct (∆∆Ct method) (Yang et
171
al., 2012).
172 173
Immunohistochemistry and immunofluorescence
174
Primary antibodies were against α-SMA (Sigma-Aldrich), vascular endothelial (VE)-
175
cadherin (CDH5) (Abcam), Snail (Santa Cruz Biotechnology), and hyaluronic acid synthase
176
2 (Has2) (Abcam) (see Appendix: Supplementary Table S2). Immunoreactivity was
177
visualised either using a standard peroxidase method for immunohistochemistry, with
178
NovaRed as the chromagen (ABC Elite Kit, Vector Laboratories) or using either
179
AlexaFluor488 or AlexaFluor568 conjugated secondary antibodies (Invitrogen) for
180
immunofluorescence.
181 182
For paraformaldehyde-fixed tissues (n = 14 MMVD cases and n = 10 controls)
183
antigen retrieval was undertaken by heating in citrate buffer (0.01 M, pH 6.0) for 5 min at
184
120 °C using a HISTO5 Rapid Microwave Histoprocessor (Milestone). For the peroxidase
185
technique, sections were incubated with 1% hydrogen peroxide in phosphate-buffered saline
Page 8 of 23
186
(PBS, pH 7.4) for 10 min and blocked with 10% goat serum diluted in PBS supplemented
187
with 0.5% Tween 20 (PBST, Vector Laboratories) for 30 min at room temperature. Sections
188
were then incubated for 60 min at room temperature with the primary antibodies, rinsed in
189
PBST, incubated for 30 min at room temperature with the biotinylated secondary antibody
190
(goat anti-rabbit IgG, Vector Laboratories), counterstained with haematoxylin, dehydrated
191
through graded ethanols and xylene and mounted in a xylene-based medium (DePex, Gurr-
192
BDH Chemicals). For immunofluorescence, the protocol was similar but without addition of
193
hydrogen peroxide, and the secondary antibody was applied for 60 min at room temperature
194
in a dark humid chamber. Slides were washed in PBS (three times for 5 min each), mounted
195
and nuclei counterstained with DAPI (Prolong Gold, Life Technologies).
196 197
Image capture and statistical analysis
198
Images were captured using light microscopy with fluorescent filters (Leica-DMLB).
199
Semi-quantitative analysis of staining was graded as: 0, no staining; 1, mild (<29% of cells);
200
2, marked (30–59% of cells); 3, diffuse and strong (>60% of cells). Qualitative and semi-
201
quantitative assessment was carried out on the NovaRed stained sections (Has2). Fluorescent
202
double-labelled sections (α-SMA/CDH5, α-SMA/Has2, and α-SMA/Snai1) were examined
203
qualitatively.
204 205
Results
206
Validation of microarray data with quantitative PCR
207
The results of microarray analysis, previously undertaken with these samples (Lu et
208
al., 2015) was confirmed by real-time quantitative PCR assays for ACTA2, CDH5, NOTCH1,
209
SNAI1, and HAS2 (Fig. 1). There was good agreement between the microarray findings
Page 9 of 23
210
(signal intensity) and statistically significant differences between normal and disease groups,
211
with the exception of HAS2.
212 213
Transcriptional profiling identifies a potential role for EndoMT in MMVD
214
In silico analysis of microarray data identified networks, mainly associated with
215
inflammation and maintanance of basement membrane integrity (Table 1). Expression of the
216
genes IL6, IL18 and TLR4 were significantly upregulated. Basement membrane disruption
217
was suggested by down-regulation of genes associated with basement membrane components
218
(NID1, LAMA2) and increased expression of the basement membrane cleavage enzyme,
219
cathepsin S (CTSS). The differentially expressed genes also included myofibroblastic
220
differentiation transcripts such as MYH11 and TAGLN. Furthermore, normalised signal
221
intensity for each probe set revealed greater mRNA expression of genes implicated in
222
EndoMT including, ACTA2, SNAI1, CTNNB1, and HAS2, whereas there was lowered
223
expression of CDH5, which favours endothelial cell migration. There was differential
224
expression of NOTCH signalling transcripts, including NOTCH1, RBPJ, NFATc1 and
225
NFATc2.
226 227
Increase in HAS2 expression in myxomatous mitral valves
228
In normal mitral valves, immunoreactivity for Has2 was observed in endothelial cells,
229
predominantly in the atrialis, and along the whole leaflet length, with sparse expression in the
230
spongiosa (Fig. 2A). In the myxomatous mitral valve distal zone, Has2 expression was
231
diffuse and significantly greater than in normal valves (P <0.001; Fig. 2B). Staining was also
232
found in the mid-zone of MMVD valves, but was not significantly different to that seen in
233
normal valves (Fig. 2). Co-expression of Has2 and α-SMA was identified in small numbers
Page 10 of 23
234
of myofibroblasts in MMVD and in endothelial cells in those regions where there was no
235
sub-endothelial myofibroblast accumulation (Fig. 3).
236 237
Snai1 expression in myofibroblast clusters in myxomatous mitral valves
238
In normal mitral valves, expression of Snail was minimal, and was localised to the
239
endothelium (Fig. 4A). In myxomatous mitral valves, cells expressing Snai1 were identified
240
in the cell dense layer at the tip of the leaflet and in the stroma of myxomatous areas (Fig.
241
4B–D). The cells were round in shape and distinctly different from the adjacent spindle-
242
shaped α-SMA positive cells. Furthermore, the Snai1 positive cells found in myofibroblast
243
clusters in the myxomatous mitral valves were predominantly α-SMA negative, with
244
relatively few cells weakly co-expressing α-SMA.
245 246
VE-cadherin (CDH5) and α-SMA are co-expressed in normal and diseased mitral valves
247
VE-cadherin was clearly expressed in the endothelium of both normal (Fig. 5A) and
248
myxomatous (Fig. 5B) mitral valves, with some of these cells co-expressing α-SMA (Fig.
249
5C). Low level expression of VE-cadherin was observed in interstitial cells in the spongiosa
250
layer of normal mitral valves, which was more pronounced in myxomatous regions of
251
diseased valves (Fig. 5). Individual α-SMA positive cells in normal mitral valves usually lay
252
underneath and in close proximity to VE-cadherin expressing endothelial cells, whereas in
253
myxomatous mitral valves, α-SMA positive cells aggregated in multiple layers.
254 255
Discussion
256
The results of this study provide evidence of developmental signalling pathways and
257
EndoMT in canine MMVD, by examining gene expression of ACTA2 (α-SMA), CTNNB1 (β-
258
cateninin), CDH5 (VE-cadherin), NOTCH1, SNAI1 and HAS2. Other genes identified by
Page 11 of 23
259
microarray analysis implicate other biological mechanisms in the pathogenesis of MMVD,
260
including the inflammatory response, basement membrane disruption and mesenchymal
261
differentiation. In heart valves from dogs affected with MMVD, there was increased
262
expression of these genes, with the exception of NOTCH1, and a reduction in CDH5. Apart
263
from down-regulation of NOTCH1 expression, all other changes support a potential
264
contribution of EndoMT and endothelial migration in the pathogenesis of MMVD.
265 266
Increased expression of Has2 in the myxoid stroma and close to the endothelium of
267
diseased valves is reminiscent of the events that drive EndoMT in valvulogenesis (Camenisch
268
et al., 2000; Bakkers et al., 2004; Lagendijk et al., 2013). Furthermore, co-expression of Has2
269
with α-SMA in some of the endothelial and interstitial cells implies mesenchymal cell
270
differentiation is present in diseased valves (Klewer et al., 2006). While VE-cadherin is
271
clearly expressed in both normal and diseased valve endothelium, suggesting cell-to-cell
272
contact has not been lost, there was increased expression of Snai1 in the endothelium of
273
MMVD valves and a reduction in CDH5 gene expression. Snai1 is the down-stream
274
transcription factor of the Notch-TGF-β signalling pathway, important for EndoMT, and
275
which controls expression of endothelial cell markers, thereby allowing EndoMT to occur
276
(Barrallo-Gimeno and Nieto, 2005; Kokudo et al., 2008). Furthermore, in valvulogenesis,
277
Snai1 increases expression of mesenchymal cell markers, such as α-SMA (Barrallo-Gimeno
278
and Nieto, 2005).
279 280
Studies of canine MMVD to date have examined cellular and molecular events, where
281
there is end-stage disease present. Evidence has emerged from several studies, suggesting that
282
myxomatous degeneration in the valve results from activation of a quiescent VIC population
283
in the valve stroma to an activated myofibroblast phenotype, and that these cells then
Page 12 of 23
284
contribute to inappropriate remodelling of the valve (Liu et al., 2007). The trigger for this
285
might be endothelial activation and/or damage, which then results in activation of the TGF-
286
β/5-hydroxytryptamine signalling pathway, enabling transformation to a VIC phenotype (Ng
287
et al., 2004; Aupperle et al., 2008; Aupperle and Disatian, 2012; Geirsson et al., 2012).
288
EndoMT suggests an alternative pathway, whereby the α-SMA positive cells, seen in
289
MMVD, are not exclusively a transformed VIC population, but derived from the valve
290
endothelial cells.
291 292
Expression of Notch promotes non-invasive EndoMT in a Snail1 and TGF-β2–
293
dependent manner, but for invasive EndoMT, BMP2 secretion from the adjacent myocardium
294
is required, inducing EndoMT by maintaining Snai1/2 activity in conjunction with Notch
295
(Rabkin et al., 2001; Garside et al., 2013). Surprisingly, Notch expression was greater in
296
normal mitral valves, but downstream targets of Snai1, such as ACTA2, SM22 and CDH5
297
were still affected. These findings suggest there may be a NOTCH-independent Snai1
298
signalling pathway in MMVD, and normal NOTCH expression might play a crucial role in
299
maintaining a healthy mitral valve. In normal canine mitral valves there was strong
300
expression of Snai1, but only seen in the endothelial cells, which might represent routine
301
endothelial repair mechanisms operating through EndoMT. In myxomatous mitral valves,
302
Snail positive cells were found within the myofibroblast clusters at the leaflet surface and in
303
the stroma. These Snail positive cells were mostly α-SMA negative and showed a distinct
304
round shaped morphology, compared with the adjacent spindle-shaped myofibroblasts. These
305
Snail positive/α-SMA negative cells might be trans-differentiated cells through partial
306
EndoMT (acquiring a migratory phenotype, without expressing α-SMA) and originating from
307
surface endothelial cells (Craig et al., 2010).
308
Page 13 of 23
309
Has2 is highly expressed during heart development and is involved in endocardial
310
cushion expansion (Camenisch et al., 2000; Ng et al., 2004; Aupperle et al., 2010;
311
MacGrogan et al., 2011). In the adult heart, hyaluronan is mainly found in aortic and mitral
312
valve leaflets and appears to contribute to normal valve remodelling, but changes in
313
hyaluronan synthesis have been noted in human myxomatous mitral valves suggesting a role
314
in disease (Bakkers et al., 2004; Toole, 2004; Lagendijk et al., 2011). Hyaluronan can induce
315
EndoMT, and increase cell migration through ErbB2 phosphorylation effects on β-catenin
316
expression (Maroski et al., 2011; Rodriguez et al., 2011). In normal canine mitral valves,
317
Has2 expression was mainly seen in the endothelium, at the atrial and ventricular sides. In the
318
spongiosa of normal mitral valves, a few spindle-shape interstitial cells (α-SMA negative)
319
expressed Has2, possibly identifying a pro-migratory phenotype of VICs (Camenisch et al.,
320
2001; Ma et al., 2005; Bourguignon et al., 2007).
321 322
Conclusions
323
Evidence of EndoMT and other developmental pathways was identified in normal and
324
myxomatous mitral valves, with those valves from dogs affected with MMVD showing
325
increased gene expression of ACTA2 (α-SMA), CDH5 (VE-cadherin), SNAI1 and HAS2, but
326
down-regulation of CDH5. Co-expression of VE-cadherin and α-SMA in normal (endothelial
327
cells) and myxomatous mitral valves (stroma cells) suggests that some myofibroblasts may
328
originate via EndoMT. EndoMT suggests an additional contibuting factor to the disease
329
pathogenesis, whereby the α-SMA positive cells seen in MMVD partially originate from the
330
VIC population and partially by differentiation and migration of endothelial cells.
331 332
Acknowledgements
Page 14 of 23
333
C-C Lu was in receipt of a Charles Darwin Scholarship, University of Edinburgh, and
334
support was obtained from the cavalier King Charles spaniel Club of England and the
335
Taiwanese Government. Preliminary results were presented as an oral presentation at the
336
European College of Veterinary Internal Medicine Conference, Mainz, 4-6 September 2014.
337 338 339 340
Conflict of interest statement None of the authors of this paper has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.
341 342 343 344
Appendix: Supplementary material Supplementary data associated with this article can be found in the online version at doi: setters please insert doi number
345 346
References
347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367
Armstrong, E.J., Bischoff, J., 2004. Heart valve development: Endothelial cell signaling and differentiation. Circulation Research 95, 459–470. Aupperle, H., Disatian, S., 2012. Pathology, protein expression and signaling in myxomatous mitral valve degeneration: Comparison of dogs and humans. Journal of Veterinary Cardiology 14, 59–71. Aupperle, H., März, I., Thielebein, J., Dinges, G., Schoon, H., 2010. Histomorphological findings and expression of matrix metalloproteinases and their tissue specific inhibitors (TIMPs) in normal tricuspid valves and in chronic tricuspid valvular disease in dogs. The Veterinary Journal 183, 176–183. Aupperle, H., Thielebein, J., Kiefer, B., Marz, I., Dinges, G., Schoon, H., 2009. An immunohistochemical study of the role of matrix metalloproteinases and their tissue inhibitors in chronic mitral valvular disease (valvular endocardiosis) in dogs. The Veterinary Journal 180, 88–94. Bakkers, J., Kramer, C., Pothof, J., Quaedvlieg, N.E.M., Spaink, H.P., Hammerschmidt, M., 2004. Has2 is required upstream of Rac1 to govern dorsal migration of lateral cells during zebrafish gastrulation. Development 131, 525–537.
Page 15 of 23
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417
Barrallo-Gimeno, A., Nieto, M.A., 2005. The Snail genes as inducers of cell movement and survival: Implications in development and cancer. Development 132, 3151–3161. Bischoff, J., Aikawa, E. 2011. Progenitor cells confer plasticity to cardiac valve endothelium. Journal of Cardiovascular Translational Research 4, 710–719. Black, A., French, A.T., Dukes-McEwan, J., Corcoran, B.M., 2005. Ultrastructural morphologic evaluation of the phenotype of valvular interstitial cells in dogs with myxomatous degeneration of the mitral valve. American Journal of Veterinary Research 66, 1408–1414. Borgarelli, M., Buchanan, J.W., 2012. Historical review, epidemiology and natural history of degenerative mitral valve disease. Journal of Veterinary Cardiology 14, 93–101. Bourguignon, L.Y.W., Peyrollier, K., Gilad, E., Brightman, A., 2007. Hyaluronan-CD44 interaction with neural wiskott-aldrich syndrome protein (N-WASP) promotes actin polymerization and ErbB2 activation leading to β-catenin nuclear translocation, transcriptional up-regulation, and cell migration in ovarian tumor cells. Journal of Biological Chemistry 282, 1265–1280. Boyer, A.S., Ayerinskas, I.I., Vincent, E.B., McKinney, L.A., Weeks, D.L., Runyan, R.B., 1999. TGFbeta2 and TGFbeta3 have separate and sequential activities during epithelialmesenchymal cell transformation in the embryonic heart. Developmental Biology 208, 530–545 Buchanan, J.W., 1977. Chronic valvular disease (endocardiosis) in dogs. Advances in Veterinary Science and Comparative Medicine 21, 75–106. Butcher, J.T., Markwald, R.R., 2007. Valvulogenesis: The moving target. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 362, 1489– 1503. Camenisch, T.D., Biesterfeldt, J., Brehm-Gibson, T., Bradley, J., McDonald, J.A., 2001. Regulation of cardiac cushion development by hyaluronan. Experimental and Clinical Cardiology 6, 4–10. Camenisch, T.D., Spicer, A.P., Brehm-Gibson, T., Biesterfeldt, J., Augustine, M.L., Calabro, A., McDonald, J.A., 2000. Disruption of hyaluronan synthase-2 abrogates normal cardiac morphogenesis and hyaluronan-mediated transformation of epithelium to mesenchyme. The Journal of Clinical Investigation 106, 349–360. Cheek, J.D., Wirrig, E.E., Alfieri, C.M., James, J.F., Yutzey, K.E., 2012. Differential activation of valvulogenic, chondrogenic, and osteogenic pathways in mouse models of myxomatous and calcific aortic valve disease. Journal of Molecular and Cellular Cardiology 52, 689–700. Chen, Y.-G., Massagué, J., 1999. Smad1 recognition and activation by the ALK1 group of transforming growth factor-β family receptors. Journal of Biological Chemistry 274, 3672–3677.
Page 16 of 23
418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467
Craig, E.A., Austin, A.F., Vaillancourt, R.R., Barnett, J.V, Camenisch, T.D., 2010. TGFβ2mediated production of hyaluronan is important for the induction of epicardial cell differentiation and invasion. Experimental Cell Research 316, 3397–3405. Disatian, S., Lacerda, C., Orton, E.C., 2010. Tryptophan hydroxylase 1 expression is increased in phenotype-altered canine and human degenerative myxomatous mitral valves. The Journal of Heart Valve Disease 19, 71–78. Garside, V.C., Chang, A.C., Karsan, A., Hoodless, P.A., 2013. Co-ordinating Notch, BMP, and TGF-β signaling during heart valve development. Cellular and Molecular Life Sciences 70, 2899–2917. Geirsson, A., Singh, M., Ali, R., Abbas, H., Li, W., Sanchez, J.A, Tellides, G., 2012. Modulation of transforming growth factor-β signaling and extracellular matrix production in myxomatous mitral valves by angiotensin II receptor blockers. Circulation 126, S189–197. Hadian, M., Corcoran, B.M., Bradshaw, J.P., 2010. Molecular changes in fibrillar collagen in myxomatous mitral valve disease. Cardiovascular Pathology 19, 141–148. Han, R.I., Black, A., Culshaw, G., French, A.T., and Corcoran, B.M., 2010. Structural and cellular changes in canine myxomatous mitral valve disease: An image analysis study. The Journal of Heart Valve Disease 19, 60–70. Hinton, R.B., and Yutzey, K.E., 2011. Heart valve structure and function in development and disease. Annual Review of Physiology 73, 29–46. Klewer, S.E., Yatskievych, T., Pogreba, K., Stevens, M.V, Antin, P.B., Camenisch, T.D., 2006. Has2 expression in heart forming regions is independent of BMP signaling. Gene Expression Patterns 6, 462–70. Kokudo, T., Suzuki, Y., Yoshimatsu, Y., Yamazaki, T., Watabe, T., Miyazono, K., 2008. Snail is required for TGFbeta-induced endothelial-mesenchymal transition of embryonic stem cell-derived endothelial cells. Journal of Cell Science 121, 3317–24. Kovacic, J.C., Mercader, N., Torres, M., Boehm, M., Fuster, V., 2012. Epithelial-tomesenchymal and endothelial-to-mesenchymal transition: From cardiovascular development to disease. Circulation 125, 1795–808. Lagendijk, A.K., Goumans, M.J., Burkhard, S.B., Bakkers, J., 2011. MicroRNA-23 restricts cardiac valve formation by inhibiting Has2 and extracellular hyaluronic acid production. Circulation Research 109, 649–57. Lagendijk, A.K., Szabó, A., Merks, R.M.H., Bakkers, J., 2013. Hyaluronan: A critical regulator of endothelial-to-mesenchymal transition during cardiac valve formation. Trends in Cardiovascular Medicine 23, 135–42. Li, J., and Bertram, J.F., 2010. Review: Endothelial-myofibroblast transition, a new player in diabetic renal fibrosis. Nephrology 15, 507–12.
Page 17 of 23
468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
Liu, A.C., Joag, V.R., Gotlieb, A.I., 2007. The emerging role of valve interstitial cell phenotypes in regulating heart valve pathobiology. The American Journal of Pathology 171, 1407–18. Lu, C-C., Liu, M.-M., Culshaw, G., Clinton, M., Argyle, D.J., Corcoran, B.M., 2015. Gene network and canonical pathway analysis in canine myxomatous mitral valve disease: A microarray study. The Veterinary Journal 204, 23–31. Ma, L., Lu, M.-F., Schwartz, R.J., Martin, J.F., 2005. Bmp2 is essential for cardiac cushion epithelial-mesenchymal transition and myocardial patterning. Development 132, 5601– 5611. MacGrogan, D., Luna-Zurita, L., de la Pompa, J.L., 2011. Notch signaling in cardiac valve development and disease. Birth Defects Research. Part A, Clinical and Molecular Teratology 91, 449–59. Mahler, G.J., Farrar, E.J., Butcher, J.T., 2013. Inflammatory cytokines promote mesenchymal transformation in embryonic and adult valve endothelial cells. Arteriosclerosis, Thrombosis and Vascular Biology 33, 121–130. Maroski, J., Vorderwülbecke, B.J., Fiedorowicz, K., Da Silva-Azevedo, L., Siegel, G., Marki, A., Zakrzewicz, A., 2011. Shear stress increases endothelial hyaluronan synthase 2 and hyaluronan synthesis especially in regard to an atheroprotective flow profile. Experimental Physiology 96, 977–986. Mjaatvedt, C.H., Yamamura, H., Capehart, A.A., Turner, D., Markwald, R.R., 1998. The cspg2 gene, disrupted in the hdf mutant, is required for right vardiac chamber and endocardial cushion formation. Developmental Biology 202, 56–66. Ng, C.M., Cheng, A., Myers, L.A., Martinez-murillo, F., Jie, C., Bedja, D., Dietz, H.C., 2004. TGF- β-dependent pathogenesis of mitral valve prolapse in a mouse model of Marfan syndrome. Journal of Clinical Investigation, 114, 1586-1592. Orton, E.C., Lacerda, C.M.R., Maclea, H.B., 2012. Signaling pathways in mitral valve degeneration. Journal of Veterinary Cardiology 14, 7–17. Piera-Velazquez, S., Li, Z., Jimenez, S.A., 2011. Role of endothelial-mesenchymal transition (EndoMT) in the pathogenesis of fibrotic disorders. The American Journal of Pathology 179, 1074–80. Rabkin, E., Aikawa, M., Stone, J.R., Fukumoto, Y., Libby, P., Schoen, F.J., 2001. Activated interstitial myofibroblasts express catabolic enzymes and mediate matrix remodeling in myxomatous heart valves. Circulation 104, 2525–2532. Rodriguez, K.J., Piechura, L.M., Masters, K.S., 2011. Regulation of valvular interstitial cell phenotype and function by hyaluronic acid in 2-D and 3-D culture environments. Matrix Biology 30, 70–82.
Page 18 of 23
516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533
Song, W., Jackson, K., McGuire, P.G., 2000. Degradation of type IV collagen by matrix metalloproteinases is an important step in the epithelial-mesenchymal transformation of the endocardial cushions. Developmental Biology 227, 606–617. Toole, B.P., 2004. Hyaluronan: From extracellular glue to pericellular cue. Nature Reviews. Cancer 4, 528–39. Whitney, J.C., 1974. Observations on the effect of age on the severity of heart valve lesions in the dog. Journal of Small Animal Practice 15, 511–522. Wylie-Sears, J., Aikawa, E., Levine, R.A., Yang, J-H., Bischoff, J., 2011. Mitral valve endothelial cells with osteogenic differentiation potential. Arteriosclerosis, Thrombosis and Vascular Biology 31, 598–607. Yang C.H., Culshaw G.J., Liu M.M., Lu C-C., French A.T., Clements D.N., Corcoran B.M., 2012. Canine tissue-specific expression of multiple small leucine rich proteoglycans. The Veterinary Journal 193, 374–80.
Page 19 of 23
534
Figure legends
535 536
Fig. 1. Real time quantitative PCR analysis of candidate genes. (A) Relative gene expression
537
of ACTA2, CDH5, NOTCH1, SNAI1, and HAS2 was determined by real-time quantitative
538
PCR using the comparative ∆∆Ct method. (B) Signal intensity of ACTA2, NOTCH1, CDH5,
539
SNAI1 and HAS2 for myxomatous and normal mitral valves. Bars represent mean SD of n =
540
6 samples. *P <0.05, ***P <0.001, NS, not significant.
541 542
Fig. 2. Representative immunostaining for Has2. (A) In the mid-zone of normal mitral valves,
543
Has2 was expressed by interstitial cells in the spongiosa and endothelial cells at the atrial and
544
ventricular surface, with a minimal amount expressed in the fibrosa layer. In the mid-zone of
545
valves affected by myxomatous degeneration, Has2 expression was similar to the normal
546
mitral valve. In the distal region of normal valves, Has2 was expressed by endothelial cells at
547
the valve surface and a few interstitial cells, whereas in a similar area of myxomatous valves
548
Has2 was strongly expressed by most interstitial cells. Hematoxylin counterstain. Scale bar =
549
50 µm. (B) Semi-quantitative analysis showed Has2 expression to be significantly greater in
550
the distal portion of the myxomatous mitral valves compared with normal valves. There was
551
no difference in Has2 expression in the mid-zone of the myxomatous and normal mitral
552
valves. Mean SD; Grade = staining 0 (none) to 3 (diffuse, strong >60% cells); NS, not
553
significant; **P < 0.001.
554 555
Fig. 3. Photomicrographs of immunofluorescence staining for α-SMA (green; upper panels)
556
and Has2 (red; middle panels). Merged images are shown in lower panels; DAPI nuclear
557
counterstaining (blue). (A) Atrial endothelial cells of a normal mitral valve co-expressing
558
Has2 and α-SMA (arrows). Scale bar = 25 µm (B) Atrial endothelial cells of a normal mitral
Page 20 of 23
559
valve positive for Has2 or α-SMA but with no evidence of co-expression. Scale bar = 25 µm
560
(C) Co-expression (arrow) of Has2 and α-SMA in a small number of cells in the
561
myofibroblast clusters in the distal zone of a myxomatous mitral valve. Scale bar = 50 µm.
562
(D) Co-expression of Has2 and α-SMA in an elongated and spindle-shaped cell (arrow), near
563
the mitral valve surface of a myxomatous mitral valve. Scale bar = 25 µm.
564 565
Fig. 4. Photomicrographs of immunofluorescence staining for α-SMA (green; upper panels),
566
and Snai1 (red; middle panel). Merged images are shown in lower panels; DAPI nuclear
567
counterstaining (blue). (A) Snai1 is expressed by relatively few endothelial cells and α-SMA
568
by occasional cells (faint staining) in the valve stroma (arrow) of a normal mitral valve, but
569
no evidence of co-expression. Scale bar = 50 µm. (B) Spindle shaped α-SMA positive
570
myofibroblasts and more round Snai1 positive cells in the stroma of a myxomatous valve,
571
with only occasional cells showing co-expression (arrow). Scale bar = 25 µm. (C and D)
572
Myofibroblast clusters close to the endothelium with Snai1 positive cells mainly distributed
573
along the endothelium and a few Snai1 positive cells in the stroma. There are occasional cells
574
showing co-expression in the endothelium (arrow). Scale bar = 25 µm.
575 576
Fig.
577
Immunohistochemistry for VE-cadherin in a normal mitral valve showing endothelial cells
578
positively stained with occasional weakly positive cells in the stroma. Scale bar = 25 µm. (B)
579
Immunohistochemistry for VE-cadherin in a myxomatous mitral valve showing endothelial
580
and several stromal interstitial cells positively stained. Scale bar = 25 µm. (C)
581
Immunofluorescence staining for VE-cadherin (red; upper panel) and α-SMA (green; middle
582
panel). Merged image shown in lower panel; DAPI nuclear counterstaining (blue). VE-
5.
Photomicrographs
of
immunoreactivity
for
VE-cadherin
(CDH5).
(A)
Page 21 of 23
583
cadherin and α-SMA positive endothelial cells were observed on the atrial side, with
584
evidence of co-expression (arrows and insert lower panel). Scale bar = 25 µm.
585 586 587 588 589 590
Table 1. Microarray analysis identified differentially expressed genes associated with inflammation, maintainance of basement membrane integrity and Notch signalling. Category
Gene (symbol)
Inflammatory
Interleukin 6 (IL6) Interleukin 18 (IL18) Toll-like receptor 4 (TLR4)
Basement membrane
Nidogen 1 (NID1) Laminin 2 (LAMA2) Cathepsin S (CTSS)
Mesenchymal Myosin heavy chain 11 differentiation (MYH11) (myofibroblastic) Transgelin (TAGLN) Smooth muscle actin α (ACTA2) Smooth muscle actin γ (ACTG2) SNAI1 Vascular endothelial cadherin (CDH5) Bone morphogenetic protein 6 (BMP6) β-catenin (CTNNB1) Hyaluronic acid synthase 2, (HAS2) Notch signalling
591 592 593
a b
NOTCH1 RBPJ NFATc1 NFATc2
Signal intensity b Control MMVD 21 176 97 667 371 826
Foldchange a 7.86 6.92 2.34
P value
-3.08 -5.15 2.26
0.0004 0.0011 0.0001
1645 365 1405
467 69 3310
4.93
0.005
80
438
2.39 10.66a
0.0042 0.027
1635 329
3994 1857
15.35
0.0031
73
1473
4.80a -2.59a
0.045 0.029
162 1630
419 877
2.38
0.0006
293
755
1.38 2.18
0.0037 0.0001
1866 417
2642 934
-2.33 1.60 -1.87 -2.13
0.0001 0.0034 0.0036 0.0066
545 1401 908 329
247 2312 517 159
0.0048 0.0029 0.0007
Fold-change determined by real time quantitative PCR. Signal intensity represents the raw microarray data.
594
Page 22 of 23
595
Page 23 of 23
S1090-0233(15)00335-4 http://dx.doi.org/doi: 10.1016/j.tvjl.2015.08.011 YTVJL 4593
To appear in:
The Veterinary Journal
Accepted date:
6-8-2015
Please cite this article as: Chi-Chien Lu, Meng-Meng Liu, Michael Clinton, Geoff Culshaw, David J. Argyle, Brendan M. Corcoran, Developmental pathways and endothelial to mesenchymal transition in canine myxomatous mitral valve disease, The Veterinary Journal (2015), http://dx.doi.org/doi: 10.1016/j.tvjl.2015.08.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Developmental pathways and endothelial to mesenchymal transition in canine myxomatous mitral valve disease Chi-Chien Lu a, Meng-Meng Liu a, Michael Clinton b, Geoff Culshaw a, David J. Argyle a, Brendan M. Corcoran a,*
a
Royal (Dick) School of Veterinary Studies and The Roslin Institute, the University of Edinburgh, Easterbush, Roslin, Mid-Lothian, Scotland, EH25 9RG, UK b
The Roslin Institute, the University of Edinburgh, Easterbush, Roslin, Mid-Lothian, Scotland, EH25 9RG, UK
* Corresponding author. Tel.: +44 131 651 9100. E-mail address: [email protected] (B.M. Corcoran).
Page 1 of 23
22 23 24 25 26 27 28 29 30 31 32 33 34 35
Highlights
This is the first report of involvement of development signalling pathways and endothelial to mesenchymal transition in canine myxomatous mitral valve disease (MMVD)
The finding is significant in that it recognises mechanisms previously reported for fibrotic conditions and cancer metastasis, rather than a degenerative condition such as MMVD
Information is provided on biological mechanisms that have therapeutic potential for new drug discovery. Abstract
36
Epithelial to mesenchymal transition (EMT), and the cardiovascular equivalent,
37
endothelial to mesenchymal transition (EndoMT), contribute to a range of chronic
38
degenerative diseases and cancer metastasis. Chronic valvulopathies exhibit some features of
39
EndoMT and activation of developmental signalling pathways, such as osteogenesis and
40
chondrogenesis, expression of cell differentiation markers, basement membrane damage and
41
endothelial transformation. The aim of the present study was to investigate the potential role
42
of developmental mechanisms in canine myxomatous mitral valve disease (MMVD), using a
43
combination of transcriptomic array technology, RT-PCR and immunohistochemistry.
44 45
There was significant differential expression for genes typically associated with
46
valvulogenesis and EndoMT, including markers of inflammation (IL6, IL18 and TLR4),
47
basement membrane disarray (NID1, LAMA2 and CTSS), mesenchymal and endothelial cell
48
differentiation (MYH11 and TAGLN) and EndoMT (ACTA2, SNAI1, CTNNB1, HAS2, CDH5,
49
and NOTCH1), with fold changes from +15.35 (ACTA2) to -5.52 (LAMA2). These changes in
50
gene expression were confirmed using RT-PCR, except for HAS2. In silico analysis identified
51
important gene networks and canonical pathways in MMVD that have associations with
52
development and organogenesis, including inflammation, valve morphogenesis and EMT, as
53
well
as
components
of
the
basement
membrane
and
extra-cellular
matrix.
Page 2 of 23
54
Immunohistochemistry identified changes in expression of hyaluronic acid synthase (Has2),
55
Snai1, α-smooth muscle actin (α-SMA) and VE-cadherin (CDH5), and co-expression of Has2
56
with α-SMA. These research findings strongly suggest involvement of developmental
57
signalling pathways and mechanisms, including EndoMT, in the pathogenesis of canine
58
MMVD.
59 60
Keywords: Endothelial to mesenchymal transition; Myxomatous mitral valve disease;
61
Canine; Transcriptomics
62
Page 3 of 23
63
Introduction
64
Myxomatous mitral valve disease (MMVD) is the most common acquired heart
65
disease affecting dogs (Buchanan, 1977; Borgarelli and Buchanan, 2012). At a structural and
66
cellular level, there is loss of collagen, accumulation of glycosaminoglycans, proliferation of
67
α-smooth muscle actin (α-SMA) and embryonic smooth muscle myosin (Smemb) positive
68
cells adjacent to the valve endothelium, with detachment and loss of endothelial cells (Rabkin
69
et al., 2001; Disatian et al., 2010; Hadian et al., 2010; Han et al., 2010). The histopathological
70
appearance of the diseased valve stroma and some of the cellular changes present, including
71
the appearance of the myxoid stroma itself, increased numbers of α-SMA positive cells,
72
basement membrane damage and changes in the endothelium, are reminiscent of processes
73
involved in valvulogenesis (Black et al., 2005; Han et al., 2010; Hinton and Yutzey, 2011).
74 75
Osteogenesis and chondrogenesis have been identified in aortic and mitral valves,
76
suggesting that developmental signalling pathways can be activated in different chronic
77
valvulopathies (Cheek et al., 2012; Orton et al., 2012; Garside et al., 2013). Osteogenic
78
differentiation capacity has been demonstrated with clonally expanded sheep mitral valvular
79
endothelial cells (VECs), suggesting that a population of multi-lineage progentior cells exist
80
in the valve endothelium, with the potential for mesenchymal differentiation. These may form
81
part of a sub-set of VECs that can replenish valvular interstitial cells (VICs) as a consequence
82
of life-long cell loss and in response to disease (Bischoff and Aikawa 2011; Wyler-Sears et
83
al, 2011). Endothelial-to-mesenchymal transition (EndoMT) has also been shown to be
84
activated by inflammatory cytokines in adult porcine aortic VECs (Mahler et al., 2013).
85 86
EndoMT is involved in valvulogenesis and there are clear examples of a role for
87
EndoMT and epithelial to mesenchymal transition (EMT) in a variety of diseases such as
Page 4 of 23
88
chronic fibrosis of the lung, kidney and heart and in cancer metastasis (Li and Bertram, 2010;
89
Piera-Velazquez et al., 2011; Kovacic et al., 2012). EndoMT is closely regulated by a wide
90
range of signalling pathways and transcription factors, including vascular endothelial growth
91
factor (VEGF) (cell proliferation), NOTCH (induction of EndoMT) and Snail and Slug (VE-
92
cadherin, α-SMA and fibronectin expression) (Armstrong and Bischoff, 2004; Butcher and
93
Markwald, 2007; Hinton and Yutzey, 2011). The hallmark of EndoMT is loss of the VEC
94
marker CD31 and cell-to-cell contact (assessed by CDH5/VE-cadherin expression), and
95
upregulation of α-SMA expression (Bischoff and Aikawa, 2011; Kovacic et al., 2012).
96 97
Members of the transforming growth factor (TGF)-β/bone morphogenic protein
98
(BMP) superfamily also mediate cell differentiation, endothelial cell separation and
99
degradation of the basal lamina, permitting migration of transformed endothelial cells (Boyer
100
et al., 1999; Song et al., 2000; Butcher and Markwald, 2007). BMPs are important in valve
101
development and their expression appears prior to the onset of EndoMT, triggering
102
production of TGF-β, hyaluronic acid synthase 2 (Has2), NOTCH and the transcription
103
factors Snail and Twist (Chen and Massagué, 1999; Ma et al., 2005). The appearance of
104
hyaluronic acid allows invasion and further migration of cells into the cardiac jelly and
105
promotes cell proliferation and differentiation by inducing expression of PI3 kinase and
106
ErbB2/3 (Mjaatvedt et al., 1998; Garside et al., 2013).
107 108
There is evidence of a role for EndoMT/EMT in a range of chronic diseases and some
109
of the stromal and cellular changes seen in MMVD suggest a tissue type similar to that seen
110
in embryonic development of the valve. Therefore, the aim of the present study was to
111
examine the possibility of re-activation and recruitment of developmental processes in canine
Page 5 of 23
112
MMVD, based on analysis of transcriptomic data (Lu et al., 2015), further investigated by
113
real time quantitative PCR and immunohistochemistry.
114 115
Materials and methods
116
Tissue samples
117
All tissue samples were collected with informed owner consent and the study
118
conformed to national (UK) and institutional ethical guidelines for the use of animals and
119
their tissues in research. For immunohistochemistry, myxomatous mitral valve leaflets (n =
120
14) were collected at necropsy from cavalier King Charles Spaniels (CKCS; n = 9) and mixed
121
breed dogs (n = 5) presented to the Hospital for Small Animals, Royal (Dick) School of
122
Veterinary Study, the University of Edinburgh. Control tissue from anterior mitral valve
123
leaflets (n = 10) were collected from young adult dogs that had been euthanased for reasons
124
other than cardiac disease. All samples were immediately fixed in 4% paraformaldehyde.
125
Valves were assessed for the presence of gross lesions by two of the three investigators
126
(CCL, BMC, GC) and classified according to an established scheme (Whitney, 1974). For
127
transcriptomic analysis, valvular tissues from four CKCS and four control dogs had been
128
assessed and the results reported by Lu et al. (2015).
129 130
RNA extraction
131
One hundred milligrams of tissue were minced, snap-frozen in liquid nitrogen and
132
pulverised for 1 min at 2000 oscillations per min in a liquid nitrogen-cooled dismembranator
133
(Mikro-Dismembrator Vessel, Braun Biotech International). One millilitre of TriReagent
134
(Sigma–Aldrich) and 200 µL of chloroform were added, followed by centrifugation (12,000 g
135
for 10 min). RNA extraction and contaminant genomic DNA digestion were carried out using
136
commercially available kits (RNeasy Mini Kit and RNase-Free DNase Set, Qiagen). The
Page 6 of 23
137
RNA purity was analysed by spectrophotometry (NanoDrop 1000, Thermo Scientific;
138
260/280 absorbance ratio ~2) and RNA quality and RNA integrity number (RIN ≥7)
139
determined by electropherogram (Agilent 2100 Bioanalyser). The RNA samples were stored
140
at -80 °C prior to further processing.
141 142
Microarray hybridization and analysis
143
The GeneChip WT Terminal Labelling Kit (Affymetrix), combined with WT
144
Expression Kit (Ambion), were used to generate amplified and biotinylated sense-strand
145
DNA targets for the Affymetrix Canine Gene 1.0 ST Array. The two cycle RNA-cDNA
146
amplification, hybridization, and chip scanning were undertaken by ARK-Genomics (Roslin
147
Institute).
148 149
For differential gene expression analysis, all the raw Affymetrix CEL files were
150
imported into Partek Genomic Suite version 6.6 software (Partek). Robust Multi-array
151
Average (RMA) was used for data normalisation and final summarisation, and Student’s t test
152
was used to identify differentially expressed genes comparing the two groups (cases and
153
controls). Network pathway analysis, including biological functions identification, was
154
undertaken using the Ingenuity Pathways Analysis (IPA) database. 1 Normalised signal
155
intensity data for whole probe sets were also produced by Affymetrix Expression Console.
156 157
Real time quantitative PCR
158
Six RNA samples (three CKCS affected with MMVD and three controls), that had
159
been analysed by microarray, were assessed by quantitative real-time PCR. Assays were
160
designed for four candidate genes (determined by signal intensity), six differentially
1
See: http://www.ingenuity.com/products/ipa (accessed 6 August 2015)
Page 7 of 23
161
expressed genes, and one reference gene (MRPS25) that had previously been validated for
162
normalising real-time quantitative PCR data generated from canine heart valve tissue (Yang
163
et al., 2012) (see Appendix: Supplementary Table S1). Primers (MWG Biotech) and labelled
164
probes (Roche Diagnostics) were synthesised and real-time quantitative PCR assays were
165
performed in triplicate in 96 well plates (LightCycler 480; Roche Diagnostics) with one no-
166
template control for each sample. Reaction volumes consisted of 5 μL LightCycler 480
167
Probes Master (Roche Diagnostics) (See Appendix: Supplementary Table S1 and
168
Supplementary File 1). Real-time quantitative PCR data were analysed using LightCycler 480
169
Basic Software (Roche Diagnostics). The reference gene (MRPS25) was used to normalise
170
the relative expression of the target genes, using the comparative Ct (∆∆Ct method) (Yang et
171
al., 2012).
172 173
Immunohistochemistry and immunofluorescence
174
Primary antibodies were against α-SMA (Sigma-Aldrich), vascular endothelial (VE)-
175
cadherin (CDH5) (Abcam), Snail (Santa Cruz Biotechnology), and hyaluronic acid synthase
176
2 (Has2) (Abcam) (see Appendix: Supplementary Table S2). Immunoreactivity was
177
visualised either using a standard peroxidase method for immunohistochemistry, with
178
NovaRed as the chromagen (ABC Elite Kit, Vector Laboratories) or using either
179
AlexaFluor488 or AlexaFluor568 conjugated secondary antibodies (Invitrogen) for
180
immunofluorescence.
181 182
For paraformaldehyde-fixed tissues (n = 14 MMVD cases and n = 10 controls)
183
antigen retrieval was undertaken by heating in citrate buffer (0.01 M, pH 6.0) for 5 min at
184
120 °C using a HISTO5 Rapid Microwave Histoprocessor (Milestone). For the peroxidase
185
technique, sections were incubated with 1% hydrogen peroxide in phosphate-buffered saline
Page 8 of 23
186
(PBS, pH 7.4) for 10 min and blocked with 10% goat serum diluted in PBS supplemented
187
with 0.5% Tween 20 (PBST, Vector Laboratories) for 30 min at room temperature. Sections
188
were then incubated for 60 min at room temperature with the primary antibodies, rinsed in
189
PBST, incubated for 30 min at room temperature with the biotinylated secondary antibody
190
(goat anti-rabbit IgG, Vector Laboratories), counterstained with haematoxylin, dehydrated
191
through graded ethanols and xylene and mounted in a xylene-based medium (DePex, Gurr-
192
BDH Chemicals). For immunofluorescence, the protocol was similar but without addition of
193
hydrogen peroxide, and the secondary antibody was applied for 60 min at room temperature
194
in a dark humid chamber. Slides were washed in PBS (three times for 5 min each), mounted
195
and nuclei counterstained with DAPI (Prolong Gold, Life Technologies).
196 197
Image capture and statistical analysis
198
Images were captured using light microscopy with fluorescent filters (Leica-DMLB).
199
Semi-quantitative analysis of staining was graded as: 0, no staining; 1, mild (<29% of cells);
200
2, marked (30–59% of cells); 3, diffuse and strong (>60% of cells). Qualitative and semi-
201
quantitative assessment was carried out on the NovaRed stained sections (Has2). Fluorescent
202
double-labelled sections (α-SMA/CDH5, α-SMA/Has2, and α-SMA/Snai1) were examined
203
qualitatively.
204 205
Results
206
Validation of microarray data with quantitative PCR
207
The results of microarray analysis, previously undertaken with these samples (Lu et
208
al., 2015) was confirmed by real-time quantitative PCR assays for ACTA2, CDH5, NOTCH1,
209
SNAI1, and HAS2 (Fig. 1). There was good agreement between the microarray findings
Page 9 of 23
210
(signal intensity) and statistically significant differences between normal and disease groups,
211
with the exception of HAS2.
212 213
Transcriptional profiling identifies a potential role for EndoMT in MMVD
214
In silico analysis of microarray data identified networks, mainly associated with
215
inflammation and maintanance of basement membrane integrity (Table 1). Expression of the
216
genes IL6, IL18 and TLR4 were significantly upregulated. Basement membrane disruption
217
was suggested by down-regulation of genes associated with basement membrane components
218
(NID1, LAMA2) and increased expression of the basement membrane cleavage enzyme,
219
cathepsin S (CTSS). The differentially expressed genes also included myofibroblastic
220
differentiation transcripts such as MYH11 and TAGLN. Furthermore, normalised signal
221
intensity for each probe set revealed greater mRNA expression of genes implicated in
222
EndoMT including, ACTA2, SNAI1, CTNNB1, and HAS2, whereas there was lowered
223
expression of CDH5, which favours endothelial cell migration. There was differential
224
expression of NOTCH signalling transcripts, including NOTCH1, RBPJ, NFATc1 and
225
NFATc2.
226 227
Increase in HAS2 expression in myxomatous mitral valves
228
In normal mitral valves, immunoreactivity for Has2 was observed in endothelial cells,
229
predominantly in the atrialis, and along the whole leaflet length, with sparse expression in the
230
spongiosa (Fig. 2A). In the myxomatous mitral valve distal zone, Has2 expression was
231
diffuse and significantly greater than in normal valves (P <0.001; Fig. 2B). Staining was also
232
found in the mid-zone of MMVD valves, but was not significantly different to that seen in
233
normal valves (Fig. 2). Co-expression of Has2 and α-SMA was identified in small numbers
Page 10 of 23
234
of myofibroblasts in MMVD and in endothelial cells in those regions where there was no
235
sub-endothelial myofibroblast accumulation (Fig. 3).
236 237
Snai1 expression in myofibroblast clusters in myxomatous mitral valves
238
In normal mitral valves, expression of Snail was minimal, and was localised to the
239
endothelium (Fig. 4A). In myxomatous mitral valves, cells expressing Snai1 were identified
240
in the cell dense layer at the tip of the leaflet and in the stroma of myxomatous areas (Fig.
241
4B–D). The cells were round in shape and distinctly different from the adjacent spindle-
242
shaped α-SMA positive cells. Furthermore, the Snai1 positive cells found in myofibroblast
243
clusters in the myxomatous mitral valves were predominantly α-SMA negative, with
244
relatively few cells weakly co-expressing α-SMA.
245 246
VE-cadherin (CDH5) and α-SMA are co-expressed in normal and diseased mitral valves
247
VE-cadherin was clearly expressed in the endothelium of both normal (Fig. 5A) and
248
myxomatous (Fig. 5B) mitral valves, with some of these cells co-expressing α-SMA (Fig.
249
5C). Low level expression of VE-cadherin was observed in interstitial cells in the spongiosa
250
layer of normal mitral valves, which was more pronounced in myxomatous regions of
251
diseased valves (Fig. 5). Individual α-SMA positive cells in normal mitral valves usually lay
252
underneath and in close proximity to VE-cadherin expressing endothelial cells, whereas in
253
myxomatous mitral valves, α-SMA positive cells aggregated in multiple layers.
254 255
Discussion
256
The results of this study provide evidence of developmental signalling pathways and
257
EndoMT in canine MMVD, by examining gene expression of ACTA2 (α-SMA), CTNNB1 (β-
258
cateninin), CDH5 (VE-cadherin), NOTCH1, SNAI1 and HAS2. Other genes identified by
Page 11 of 23
259
microarray analysis implicate other biological mechanisms in the pathogenesis of MMVD,
260
including the inflammatory response, basement membrane disruption and mesenchymal
261
differentiation. In heart valves from dogs affected with MMVD, there was increased
262
expression of these genes, with the exception of NOTCH1, and a reduction in CDH5. Apart
263
from down-regulation of NOTCH1 expression, all other changes support a potential
264
contribution of EndoMT and endothelial migration in the pathogenesis of MMVD.
265 266
Increased expression of Has2 in the myxoid stroma and close to the endothelium of
267
diseased valves is reminiscent of the events that drive EndoMT in valvulogenesis (Camenisch
268
et al., 2000; Bakkers et al., 2004; Lagendijk et al., 2013). Furthermore, co-expression of Has2
269
with α-SMA in some of the endothelial and interstitial cells implies mesenchymal cell
270
differentiation is present in diseased valves (Klewer et al., 2006). While VE-cadherin is
271
clearly expressed in both normal and diseased valve endothelium, suggesting cell-to-cell
272
contact has not been lost, there was increased expression of Snai1 in the endothelium of
273
MMVD valves and a reduction in CDH5 gene expression. Snai1 is the down-stream
274
transcription factor of the Notch-TGF-β signalling pathway, important for EndoMT, and
275
which controls expression of endothelial cell markers, thereby allowing EndoMT to occur
276
(Barrallo-Gimeno and Nieto, 2005; Kokudo et al., 2008). Furthermore, in valvulogenesis,
277
Snai1 increases expression of mesenchymal cell markers, such as α-SMA (Barrallo-Gimeno
278
and Nieto, 2005).
279 280
Studies of canine MMVD to date have examined cellular and molecular events, where
281
there is end-stage disease present. Evidence has emerged from several studies, suggesting that
282
myxomatous degeneration in the valve results from activation of a quiescent VIC population
283
in the valve stroma to an activated myofibroblast phenotype, and that these cells then
Page 12 of 23
284
contribute to inappropriate remodelling of the valve (Liu et al., 2007). The trigger for this
285
might be endothelial activation and/or damage, which then results in activation of the TGF-
286
β/5-hydroxytryptamine signalling pathway, enabling transformation to a VIC phenotype (Ng
287
et al., 2004; Aupperle et al., 2008; Aupperle and Disatian, 2012; Geirsson et al., 2012).
288
EndoMT suggests an alternative pathway, whereby the α-SMA positive cells, seen in
289
MMVD, are not exclusively a transformed VIC population, but derived from the valve
290
endothelial cells.
291 292
Expression of Notch promotes non-invasive EndoMT in a Snail1 and TGF-β2–
293
dependent manner, but for invasive EndoMT, BMP2 secretion from the adjacent myocardium
294
is required, inducing EndoMT by maintaining Snai1/2 activity in conjunction with Notch
295
(Rabkin et al., 2001; Garside et al., 2013). Surprisingly, Notch expression was greater in
296
normal mitral valves, but downstream targets of Snai1, such as ACTA2, SM22 and CDH5
297
were still affected. These findings suggest there may be a NOTCH-independent Snai1
298
signalling pathway in MMVD, and normal NOTCH expression might play a crucial role in
299
maintaining a healthy mitral valve. In normal canine mitral valves there was strong
300
expression of Snai1, but only seen in the endothelial cells, which might represent routine
301
endothelial repair mechanisms operating through EndoMT. In myxomatous mitral valves,
302
Snail positive cells were found within the myofibroblast clusters at the leaflet surface and in
303
the stroma. These Snail positive cells were mostly α-SMA negative and showed a distinct
304
round shaped morphology, compared with the adjacent spindle-shaped myofibroblasts. These
305
Snail positive/α-SMA negative cells might be trans-differentiated cells through partial
306
EndoMT (acquiring a migratory phenotype, without expressing α-SMA) and originating from
307
surface endothelial cells (Craig et al., 2010).
308
Page 13 of 23
309
Has2 is highly expressed during heart development and is involved in endocardial
310
cushion expansion (Camenisch et al., 2000; Ng et al., 2004; Aupperle et al., 2010;
311
MacGrogan et al., 2011). In the adult heart, hyaluronan is mainly found in aortic and mitral
312
valve leaflets and appears to contribute to normal valve remodelling, but changes in
313
hyaluronan synthesis have been noted in human myxomatous mitral valves suggesting a role
314
in disease (Bakkers et al., 2004; Toole, 2004; Lagendijk et al., 2011). Hyaluronan can induce
315
EndoMT, and increase cell migration through ErbB2 phosphorylation effects on β-catenin
316
expression (Maroski et al., 2011; Rodriguez et al., 2011). In normal canine mitral valves,
317
Has2 expression was mainly seen in the endothelium, at the atrial and ventricular sides. In the
318
spongiosa of normal mitral valves, a few spindle-shape interstitial cells (α-SMA negative)
319
expressed Has2, possibly identifying a pro-migratory phenotype of VICs (Camenisch et al.,
320
2001; Ma et al., 2005; Bourguignon et al., 2007).
321 322
Conclusions
323
Evidence of EndoMT and other developmental pathways was identified in normal and
324
myxomatous mitral valves, with those valves from dogs affected with MMVD showing
325
increased gene expression of ACTA2 (α-SMA), CDH5 (VE-cadherin), SNAI1 and HAS2, but
326
down-regulation of CDH5. Co-expression of VE-cadherin and α-SMA in normal (endothelial
327
cells) and myxomatous mitral valves (stroma cells) suggests that some myofibroblasts may
328
originate via EndoMT. EndoMT suggests an additional contibuting factor to the disease
329
pathogenesis, whereby the α-SMA positive cells seen in MMVD partially originate from the
330
VIC population and partially by differentiation and migration of endothelial cells.
331 332
Acknowledgements
Page 14 of 23
333
C-C Lu was in receipt of a Charles Darwin Scholarship, University of Edinburgh, and
334
support was obtained from the cavalier King Charles spaniel Club of England and the
335
Taiwanese Government. Preliminary results were presented as an oral presentation at the
336
European College of Veterinary Internal Medicine Conference, Mainz, 4-6 September 2014.
337 338 339 340
Conflict of interest statement None of the authors of this paper has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.
341 342 343 344
Appendix: Supplementary material Supplementary data associated with this article can be found in the online version at doi: setters please insert doi number
345 346
References
347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367
Armstrong, E.J., Bischoff, J., 2004. Heart valve development: Endothelial cell signaling and differentiation. Circulation Research 95, 459–470. Aupperle, H., Disatian, S., 2012. Pathology, protein expression and signaling in myxomatous mitral valve degeneration: Comparison of dogs and humans. Journal of Veterinary Cardiology 14, 59–71. Aupperle, H., März, I., Thielebein, J., Dinges, G., Schoon, H., 2010. Histomorphological findings and expression of matrix metalloproteinases and their tissue specific inhibitors (TIMPs) in normal tricuspid valves and in chronic tricuspid valvular disease in dogs. The Veterinary Journal 183, 176–183. Aupperle, H., Thielebein, J., Kiefer, B., Marz, I., Dinges, G., Schoon, H., 2009. An immunohistochemical study of the role of matrix metalloproteinases and their tissue inhibitors in chronic mitral valvular disease (valvular endocardiosis) in dogs. The Veterinary Journal 180, 88–94. Bakkers, J., Kramer, C., Pothof, J., Quaedvlieg, N.E.M., Spaink, H.P., Hammerschmidt, M., 2004. Has2 is required upstream of Rac1 to govern dorsal migration of lateral cells during zebrafish gastrulation. Development 131, 525–537.
Page 15 of 23
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417
Barrallo-Gimeno, A., Nieto, M.A., 2005. The Snail genes as inducers of cell movement and survival: Implications in development and cancer. Development 132, 3151–3161. Bischoff, J., Aikawa, E. 2011. Progenitor cells confer plasticity to cardiac valve endothelium. Journal of Cardiovascular Translational Research 4, 710–719. Black, A., French, A.T., Dukes-McEwan, J., Corcoran, B.M., 2005. Ultrastructural morphologic evaluation of the phenotype of valvular interstitial cells in dogs with myxomatous degeneration of the mitral valve. American Journal of Veterinary Research 66, 1408–1414. Borgarelli, M., Buchanan, J.W., 2012. Historical review, epidemiology and natural history of degenerative mitral valve disease. Journal of Veterinary Cardiology 14, 93–101. Bourguignon, L.Y.W., Peyrollier, K., Gilad, E., Brightman, A., 2007. Hyaluronan-CD44 interaction with neural wiskott-aldrich syndrome protein (N-WASP) promotes actin polymerization and ErbB2 activation leading to β-catenin nuclear translocation, transcriptional up-regulation, and cell migration in ovarian tumor cells. Journal of Biological Chemistry 282, 1265–1280. Boyer, A.S., Ayerinskas, I.I., Vincent, E.B., McKinney, L.A., Weeks, D.L., Runyan, R.B., 1999. TGFbeta2 and TGFbeta3 have separate and sequential activities during epithelialmesenchymal cell transformation in the embryonic heart. Developmental Biology 208, 530–545 Buchanan, J.W., 1977. Chronic valvular disease (endocardiosis) in dogs. Advances in Veterinary Science and Comparative Medicine 21, 75–106. Butcher, J.T., Markwald, R.R., 2007. Valvulogenesis: The moving target. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 362, 1489– 1503. Camenisch, T.D., Biesterfeldt, J., Brehm-Gibson, T., Bradley, J., McDonald, J.A., 2001. Regulation of cardiac cushion development by hyaluronan. Experimental and Clinical Cardiology 6, 4–10. Camenisch, T.D., Spicer, A.P., Brehm-Gibson, T., Biesterfeldt, J., Augustine, M.L., Calabro, A., McDonald, J.A., 2000. Disruption of hyaluronan synthase-2 abrogates normal cardiac morphogenesis and hyaluronan-mediated transformation of epithelium to mesenchyme. The Journal of Clinical Investigation 106, 349–360. Cheek, J.D., Wirrig, E.E., Alfieri, C.M., James, J.F., Yutzey, K.E., 2012. Differential activation of valvulogenic, chondrogenic, and osteogenic pathways in mouse models of myxomatous and calcific aortic valve disease. Journal of Molecular and Cellular Cardiology 52, 689–700. Chen, Y.-G., Massagué, J., 1999. Smad1 recognition and activation by the ALK1 group of transforming growth factor-β family receptors. Journal of Biological Chemistry 274, 3672–3677.
Page 16 of 23
418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467
Craig, E.A., Austin, A.F., Vaillancourt, R.R., Barnett, J.V, Camenisch, T.D., 2010. TGFβ2mediated production of hyaluronan is important for the induction of epicardial cell differentiation and invasion. Experimental Cell Research 316, 3397–3405. Disatian, S., Lacerda, C., Orton, E.C., 2010. Tryptophan hydroxylase 1 expression is increased in phenotype-altered canine and human degenerative myxomatous mitral valves. The Journal of Heart Valve Disease 19, 71–78. Garside, V.C., Chang, A.C., Karsan, A., Hoodless, P.A., 2013. Co-ordinating Notch, BMP, and TGF-β signaling during heart valve development. Cellular and Molecular Life Sciences 70, 2899–2917. Geirsson, A., Singh, M., Ali, R., Abbas, H., Li, W., Sanchez, J.A, Tellides, G., 2012. Modulation of transforming growth factor-β signaling and extracellular matrix production in myxomatous mitral valves by angiotensin II receptor blockers. Circulation 126, S189–197. Hadian, M., Corcoran, B.M., Bradshaw, J.P., 2010. Molecular changes in fibrillar collagen in myxomatous mitral valve disease. Cardiovascular Pathology 19, 141–148. Han, R.I., Black, A., Culshaw, G., French, A.T., and Corcoran, B.M., 2010. Structural and cellular changes in canine myxomatous mitral valve disease: An image analysis study. The Journal of Heart Valve Disease 19, 60–70. Hinton, R.B., and Yutzey, K.E., 2011. Heart valve structure and function in development and disease. Annual Review of Physiology 73, 29–46. Klewer, S.E., Yatskievych, T., Pogreba, K., Stevens, M.V, Antin, P.B., Camenisch, T.D., 2006. Has2 expression in heart forming regions is independent of BMP signaling. Gene Expression Patterns 6, 462–70. Kokudo, T., Suzuki, Y., Yoshimatsu, Y., Yamazaki, T., Watabe, T., Miyazono, K., 2008. Snail is required for TGFbeta-induced endothelial-mesenchymal transition of embryonic stem cell-derived endothelial cells. Journal of Cell Science 121, 3317–24. Kovacic, J.C., Mercader, N., Torres, M., Boehm, M., Fuster, V., 2012. Epithelial-tomesenchymal and endothelial-to-mesenchymal transition: From cardiovascular development to disease. Circulation 125, 1795–808. Lagendijk, A.K., Goumans, M.J., Burkhard, S.B., Bakkers, J., 2011. MicroRNA-23 restricts cardiac valve formation by inhibiting Has2 and extracellular hyaluronic acid production. Circulation Research 109, 649–57. Lagendijk, A.K., Szabó, A., Merks, R.M.H., Bakkers, J., 2013. Hyaluronan: A critical regulator of endothelial-to-mesenchymal transition during cardiac valve formation. Trends in Cardiovascular Medicine 23, 135–42. Li, J., and Bertram, J.F., 2010. Review: Endothelial-myofibroblast transition, a new player in diabetic renal fibrosis. Nephrology 15, 507–12.
Page 17 of 23
468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515
Liu, A.C., Joag, V.R., Gotlieb, A.I., 2007. The emerging role of valve interstitial cell phenotypes in regulating heart valve pathobiology. The American Journal of Pathology 171, 1407–18. Lu, C-C., Liu, M.-M., Culshaw, G., Clinton, M., Argyle, D.J., Corcoran, B.M., 2015. Gene network and canonical pathway analysis in canine myxomatous mitral valve disease: A microarray study. The Veterinary Journal 204, 23–31. Ma, L., Lu, M.-F., Schwartz, R.J., Martin, J.F., 2005. Bmp2 is essential for cardiac cushion epithelial-mesenchymal transition and myocardial patterning. Development 132, 5601– 5611. MacGrogan, D., Luna-Zurita, L., de la Pompa, J.L., 2011. Notch signaling in cardiac valve development and disease. Birth Defects Research. Part A, Clinical and Molecular Teratology 91, 449–59. Mahler, G.J., Farrar, E.J., Butcher, J.T., 2013. Inflammatory cytokines promote mesenchymal transformation in embryonic and adult valve endothelial cells. Arteriosclerosis, Thrombosis and Vascular Biology 33, 121–130. Maroski, J., Vorderwülbecke, B.J., Fiedorowicz, K., Da Silva-Azevedo, L., Siegel, G., Marki, A., Zakrzewicz, A., 2011. Shear stress increases endothelial hyaluronan synthase 2 and hyaluronan synthesis especially in regard to an atheroprotective flow profile. Experimental Physiology 96, 977–986. Mjaatvedt, C.H., Yamamura, H., Capehart, A.A., Turner, D., Markwald, R.R., 1998. The cspg2 gene, disrupted in the hdf mutant, is required for right vardiac chamber and endocardial cushion formation. Developmental Biology 202, 56–66. Ng, C.M., Cheng, A., Myers, L.A., Martinez-murillo, F., Jie, C., Bedja, D., Dietz, H.C., 2004. TGF- β-dependent pathogenesis of mitral valve prolapse in a mouse model of Marfan syndrome. Journal of Clinical Investigation, 114, 1586-1592. Orton, E.C., Lacerda, C.M.R., Maclea, H.B., 2012. Signaling pathways in mitral valve degeneration. Journal of Veterinary Cardiology 14, 7–17. Piera-Velazquez, S., Li, Z., Jimenez, S.A., 2011. Role of endothelial-mesenchymal transition (EndoMT) in the pathogenesis of fibrotic disorders. The American Journal of Pathology 179, 1074–80. Rabkin, E., Aikawa, M., Stone, J.R., Fukumoto, Y., Libby, P., Schoen, F.J., 2001. Activated interstitial myofibroblasts express catabolic enzymes and mediate matrix remodeling in myxomatous heart valves. Circulation 104, 2525–2532. Rodriguez, K.J., Piechura, L.M., Masters, K.S., 2011. Regulation of valvular interstitial cell phenotype and function by hyaluronic acid in 2-D and 3-D culture environments. Matrix Biology 30, 70–82.
Page 18 of 23
516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533
Song, W., Jackson, K., McGuire, P.G., 2000. Degradation of type IV collagen by matrix metalloproteinases is an important step in the epithelial-mesenchymal transformation of the endocardial cushions. Developmental Biology 227, 606–617. Toole, B.P., 2004. Hyaluronan: From extracellular glue to pericellular cue. Nature Reviews. Cancer 4, 528–39. Whitney, J.C., 1974. Observations on the effect of age on the severity of heart valve lesions in the dog. Journal of Small Animal Practice 15, 511–522. Wylie-Sears, J., Aikawa, E., Levine, R.A., Yang, J-H., Bischoff, J., 2011. Mitral valve endothelial cells with osteogenic differentiation potential. Arteriosclerosis, Thrombosis and Vascular Biology 31, 598–607. Yang C.H., Culshaw G.J., Liu M.M., Lu C-C., French A.T., Clements D.N., Corcoran B.M., 2012. Canine tissue-specific expression of multiple small leucine rich proteoglycans. The Veterinary Journal 193, 374–80.
Page 19 of 23
534
Figure legends
535 536
Fig. 1. Real time quantitative PCR analysis of candidate genes. (A) Relative gene expression
537
of ACTA2, CDH5, NOTCH1, SNAI1, and HAS2 was determined by real-time quantitative
538
PCR using the comparative ∆∆Ct method. (B) Signal intensity of ACTA2, NOTCH1, CDH5,
539
SNAI1 and HAS2 for myxomatous and normal mitral valves. Bars represent mean SD of n =
540
6 samples. *P <0.05, ***P <0.001, NS, not significant.
541 542
Fig. 2. Representative immunostaining for Has2. (A) In the mid-zone of normal mitral valves,
543
Has2 was expressed by interstitial cells in the spongiosa and endothelial cells at the atrial and
544
ventricular surface, with a minimal amount expressed in the fibrosa layer. In the mid-zone of
545
valves affected by myxomatous degeneration, Has2 expression was similar to the normal
546
mitral valve. In the distal region of normal valves, Has2 was expressed by endothelial cells at
547
the valve surface and a few interstitial cells, whereas in a similar area of myxomatous valves
548
Has2 was strongly expressed by most interstitial cells. Hematoxylin counterstain. Scale bar =
549
50 µm. (B) Semi-quantitative analysis showed Has2 expression to be significantly greater in
550
the distal portion of the myxomatous mitral valves compared with normal valves. There was
551
no difference in Has2 expression in the mid-zone of the myxomatous and normal mitral
552
valves. Mean SD; Grade = staining 0 (none) to 3 (diffuse, strong >60% cells); NS, not
553
significant; **P < 0.001.
554 555
Fig. 3. Photomicrographs of immunofluorescence staining for α-SMA (green; upper panels)
556
and Has2 (red; middle panels). Merged images are shown in lower panels; DAPI nuclear
557
counterstaining (blue). (A) Atrial endothelial cells of a normal mitral valve co-expressing
558
Has2 and α-SMA (arrows). Scale bar = 25 µm (B) Atrial endothelial cells of a normal mitral
Page 20 of 23
559
valve positive for Has2 or α-SMA but with no evidence of co-expression. Scale bar = 25 µm
560
(C) Co-expression (arrow) of Has2 and α-SMA in a small number of cells in the
561
myofibroblast clusters in the distal zone of a myxomatous mitral valve. Scale bar = 50 µm.
562
(D) Co-expression of Has2 and α-SMA in an elongated and spindle-shaped cell (arrow), near
563
the mitral valve surface of a myxomatous mitral valve. Scale bar = 25 µm.
564 565
Fig. 4. Photomicrographs of immunofluorescence staining for α-SMA (green; upper panels),
566
and Snai1 (red; middle panel). Merged images are shown in lower panels; DAPI nuclear
567
counterstaining (blue). (A) Snai1 is expressed by relatively few endothelial cells and α-SMA
568
by occasional cells (faint staining) in the valve stroma (arrow) of a normal mitral valve, but
569
no evidence of co-expression. Scale bar = 50 µm. (B) Spindle shaped α-SMA positive
570
myofibroblasts and more round Snai1 positive cells in the stroma of a myxomatous valve,
571
with only occasional cells showing co-expression (arrow). Scale bar = 25 µm. (C and D)
572
Myofibroblast clusters close to the endothelium with Snai1 positive cells mainly distributed
573
along the endothelium and a few Snai1 positive cells in the stroma. There are occasional cells
574
showing co-expression in the endothelium (arrow). Scale bar = 25 µm.
575 576
Fig.
577
Immunohistochemistry for VE-cadherin in a normal mitral valve showing endothelial cells
578
positively stained with occasional weakly positive cells in the stroma. Scale bar = 25 µm. (B)
579
Immunohistochemistry for VE-cadherin in a myxomatous mitral valve showing endothelial
580
and several stromal interstitial cells positively stained. Scale bar = 25 µm. (C)
581
Immunofluorescence staining for VE-cadherin (red; upper panel) and α-SMA (green; middle
582
panel). Merged image shown in lower panel; DAPI nuclear counterstaining (blue). VE-
5.
Photomicrographs
of
immunoreactivity
for
VE-cadherin
(CDH5).
(A)
Page 21 of 23
583
cadherin and α-SMA positive endothelial cells were observed on the atrial side, with
584
evidence of co-expression (arrows and insert lower panel). Scale bar = 25 µm.
585 586 587 588 589 590
Table 1. Microarray analysis identified differentially expressed genes associated with inflammation, maintainance of basement membrane integrity and Notch signalling. Category
Gene (symbol)
Inflammatory
Interleukin 6 (IL6) Interleukin 18 (IL18) Toll-like receptor 4 (TLR4)
Basement membrane
Nidogen 1 (NID1) Laminin 2 (LAMA2) Cathepsin S (CTSS)
Mesenchymal Myosin heavy chain 11 differentiation (MYH11) (myofibroblastic) Transgelin (TAGLN) Smooth muscle actin α (ACTA2) Smooth muscle actin γ (ACTG2) SNAI1 Vascular endothelial cadherin (CDH5) Bone morphogenetic protein 6 (BMP6) β-catenin (CTNNB1) Hyaluronic acid synthase 2, (HAS2) Notch signalling
591 592 593
a b
NOTCH1 RBPJ NFATc1 NFATc2
Signal intensity b Control MMVD 21 176 97 667 371 826
Foldchange a 7.86 6.92 2.34
P value
-3.08 -5.15 2.26
0.0004 0.0011 0.0001
1645 365 1405
467 69 3310
4.93
0.005
80
438
2.39 10.66a
0.0042 0.027
1635 329
3994 1857
15.35
0.0031
73
1473
4.80a -2.59a
0.045 0.029
162 1630
419 877
2.38
0.0006
293
755
1.38 2.18
0.0037 0.0001
1866 417
2642 934
-2.33 1.60 -1.87 -2.13
0.0001 0.0034 0.0036 0.0066
545 1401 908 329
247 2312 517 159
0.0048 0.0029 0.0007
Fold-change determined by real time quantitative PCR. Signal intensity represents the raw microarray data.
594
Page 22 of 23
595
Page 23 of 23