- Email: [email protected]
MicroRNAs expression and their regulatory networks during mesenchymal stem cells differentiation toward osteoblasts
Accepted Manuscript Title: MicroRNAs Expression and their Regulatory Networks During Mesenchymal Stem Cells Differentiation towards Osteoblasts Author: S. Vimalraj N. Selvamurugan PII: DOI: Reference:
S0141-8130(14)00114-7 http://dx.doi.org/doi:10.1016/j.ijbiomac.2014.02.030 BIOMAC 4184
To appear in:
International Journal of Biological Macromolecules
Received date: Revised date: Accepted date:
23-10-2013 18-12-2013 13-2-2014
Please cite this article as: S. VimalrajN. Selvamurugan MicroRNAs Expression and their Regulatory Networks During Mesenchymal Stem Cells Differentiation towards Osteoblasts (2014), http://dx.doi.org/10.1016/j.ijbiomac.2014.02.030 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.
MicroRNAs Expression and their Regulatory Networks During Mesenchymal Stem Cells
us
cr
S. Vimalraj and N. Selvamurugan*
ip t
Differentiation towards Osteoblasts
an
Department of Biotechnology, School of Bioengineering, SRM University, Kattankulathur,
te
d
M
Tamil Nadu. India.
Ac ce p
*To whom correspondence should be made: N. Selvamurugan, Ph. D. Professor
Department of Biotechnology School of Bioengineering SRM University
Kattankulathur 603 203. Tamil Nadu. India. Cell: 91-9940632335 E mail: [email protected] [email protected]
1 Page 1 of 34
Abstract MicroRNAs (miRNAs) are small endogenous non coding RNAs which regulate mRNAs post-transcriptionally. In this study, a selective number of miRNAs was investigated
ip t
for their expression and intracellular regulatory networks involved in differentiation of human mesenchymal stem cells (hMSCs) towards osteoblasts. The expression of miR-424,
cr
miR-106a, miR-148a, let-7i and miR-99a miRNAs was found to be specific in hMSCs; whereas expression of miR-15b, miR-24, miR-130b, miR-30c, and miR-130a miRNAs was
us
found to be specific in differentiated osteoblasts. A bioinformatics approach identified that the MAPK pathway was mostly targeted by hMSCs specific miRNAs; whereas JAK-STAT,
an
p53, Focal adhesion, Gap junction, Ubiquitin mediated proteolysis pathways were targeted by osteblastic specific miRNAs. Altering expression of osteoblast specific miRNA (miR-15b)
M
promoted adipogenesis and myogenesis lineages. Thus, we suggest that miRNAs’ regulatory networks and their target genes might provide an insight of their role during differentiation of
d
hMSCs towards osteoblasts, and alteration in the expression of miRNAs would be a valuable
te
approach for controlling osteoblast differentiation.
Ac ce p
Keywords: MicroRNAs, Mesenchymal Stem Cells, Osteoblasts, miR-15b, ToppCluster
2 Page 2 of 34
1. Introduction The process of skeletal development involves differentiation and maturation of osteoblast, a mesenchymal origin cell type. During this developmental process, several
ip t
signaling pathways like TGF-β, BMP, Wnt, and MAPK are activated. Runx2 and osterix are the major transcription factors involved in osteogenesis and skeletal formation by activating
cr
osteoblast differentiation marker genes [1, 2]. Mesenchymal stem cells (MSCs) have the
us
potential to differentiate into multiple lineages with response to respective transcription factors, for instance PPARγ for adipose, myogenin for myoblast, FGF2 for chondrocyte and
an
Runx2 for osteoblast differentiation. There are several regulatory mechanisms involved in differentiation of MSCs into osteoblasts which may be essential for maintaining sustainable
M
bone mass. Understanding the regulatory mechanisms would also develop new strategies to treat bone related diseases like osteoporosis, osteosclerosis.
d
Osteoblast differentiation and maturation is typically controlled by microRNAs
te
(miRNAs) [3-5]. miRNAs are small endogenous non coding RNA molecules (19-25 nt) that
Ac ce p
regulate mRNAs at posttranscriptional level. A single miRNA can target more than hundred mRNAs, and a single mRNA can also be targeted by multiple miRNAs suggesting that miRNAs might have a wide range of structural and functional networks [6]. They control several physiological process including cell growth, apoptosis, metabolism, hormone signaling, stress adaptation, proliferation and differentiation by direct or indirect regulation [4, 5, 7-9]. During osteoblast differentiation, a significance variation in miRNAs expression profile has been reported between undifferentiated hMSCs and osteo differentiated hMSCs [10-12]. Several miRNAs have been identified during differentiation of hMSCs into osteoblasts by miRNA array profiling but their functional roles are still unexplored [10,11,13] 3 Page 3 of 34
and for some miRNAs alone, limited number of targets have been detected which is not sufficient to go for clinical application of particular miRNA because it may elicit some other undesired physiological functions. Thus, the aim of this study was to identify expression of
ip t
miRNAs during differentiation of hMSCs into osteoblasts, and to reveal their complete regulatory network and target genes by bioinformatics tools. Based on the computational
cr
analysis, miR-15b, an osteoblast specific miRNA was preferably selected for further functional analysis and we identified that the inhibition of miR-15b expression favored
us
towards other cell lineages.
an
2. Materials and Methods
2.1. In vitro culture of hBMSCs and osteoblast differentiation
M
Human bone marrow stromal cells (hBMSCs) was isolated from 27 years old female and cultured according to previously described method [14]. The Institutional review
d
committee in SRM University approved the protocol for isolation of hMSCs. BMSCs were
te
obtained with the written consent from donor. Rat calvarial cells isolation from 2 days old rat
Ac ce p
pups was carried out as described previously [15]. All animal procedures were approved by the University of Madras and the SRM university animal ethical committees. The MG63 cell line, human osteoblast like cells were purchased from NCCS, Pune, India. The cells were induced to osteo differentiation in osteogenic medium [alpha MEM (Cellgro), 15% FBS (Invitrogen), Penicillin-Streptomycin (Invitrogen), 10-4 M L-Ascorbic Acid, 10-8 M Dexamethasone and 1.80 mM Potassium phosphate monobasic (Sigma)]. The medium was changed once in three days. 2.2. miRNA and mRNA expression A list of miRNAs was selected based on the microarray profiling in hMSCs and osteo differentiated hMSCs and based on their functionally important putative targets associated
4 Page 4 of 34
with differentiation of osteoblasts [10, 12, 13, 16]. Putative target prediction online tools such as
microrna
(http://www.microrna.org/microrna/home.do),
TargetScan
(http://www.targetscan.org/) and PicTar (http://pictar.mdc-berlin.de/) were used to find out
ip t
the miRNAs’ potential target genes. The stem loop sequence of miRNA was retrieved from miRNA database (http://www.mirbase.org/) and according to them, primers were designed
cr
(Table 1). The expression of these miRNAs was analyzed by semi quantitative RT-PCR method. Total RNA was isolated from cultured cells with Trizol reagent (Invitrogen)
us
according to the manufacturer’s instructions and was quantitatively evaluated by Qubit 2.0
an
Flurometer (Invitrogen). cDNA was synthesized using GeNei AMV RT-PCR kit according to the manufacturer’s protocol as follows: 1 µg of total RNA, 1µl of oligo(dT) or random
M
hexamer (50µM) and 1µl of dNTP Mixture (10mM) were added and made up to 14µl reverse transcript reaction solution using RNase free dH2O, followed by incubation at 65ºC for 5 min
d
and subjected to fast chill using ice cubes. Then 4µl of 5X First-strand Buffer, 1µl of Reverse
te
Transcriptase (200U/μl) and 1µl of RNase Inhibitor (40U/μl) were added to 14µl reverse transcript reaction solution and this reaction mixture was kept in 30°C for 10 min, 42°C for
Ac ce p
50 min and 95°C for 5 min in thermal cycler. The PCR reaction for each miRNA was performed in triplicate and the products were fractionated using 2.5% agarose gel electrophoresis and. The intensity of bands was measured by densitometry scanning (ImageJ software (http://rsbweb.nih.gov/ij/)) [17] and the miRNAs band intensities were normalized with internal control U6 band intensity. The average of respective band intensity was blotted in the form of bar diagram with standard deviation. Table 1 The real-time PCR analysis was performed in Bio-Rad system using KAPA SYBR FAST qPCR Kits (Kapabiosystem) according to the manufacturer’s instruction. Stem loop sequence of miR-15b primers or mRNA primers and internal control (U6 or GAPDH) 5 Page 5 of 34
primers (Table 1) were used for amplification. The Ct (threshold cycle) values were calculated from amplification curve. ΔCtneg (ΔCt negative control) value of candidate gene of negative control miRNA transfected cells was calculated by subtracting the Ct value of its
ip t
internal control gene. Similarly, ΔCtinh (ΔCt miR-15b inhibitor) value of miR-15b inhibitor transfected cells was calculated by subtracting the Ct value of its internal control gene. The
cr
actual ΔΔCt value was calculated by subtracting ΔCtneg with ΔCtinh and then the ΔΔCt values were converted to fold change over negative control miRNA by raising 2 to the ΔΔCt power.
us
The thermal cycling conditions were as follows: 95ºC for 5 min as initial denaturation,
an
followed by 40 cycles of 95ºC for 30s, 58ºC for 30s and 72ºC for 30s, with a final extension at 72ºC for 5 min.
DIANA-mirPath
is
a
M
2.3. DIANA-mirPath analysis
free
online
bioinformatics
tool
d
(http://diana.cslab.ece.ntua.gr/?sec=home) to find out molecular signaling pathways targeted
te
by single miRNA or combined effect of multiple miRNAs on a given pathway [18]. This tool was used to carry out an enrichment analysis of input datasets by comparing each set of
Ac ce p
targeted genes to all available biological pathways which is provided by Kyoto Encyclopedia of Genes and Genomes (KEGG). Furthermore, this software allocates an overall priority score (enrichment p-value) based on the predicted strength of the miRNA interactions with the molecules of the target pathway. Enrichment analysis of each KEGG signaling pathway was represented by negative natural log of P value (-ln P). The enrichment -lnP values bar graph helps to distinguish the relative importance of each pathway in various dataset. 2.4. ToppCluster analysis ToppCluster
(http://toppcluster.cchmc.org/)
and
Cytoscape
(http://www.cytoscape.org/) were used for biological functional enrichment analysis and preparation of network figures. The miRNAs putative targets were used to generate 6 Page 6 of 34
regulatory network clusters by hierarchical clustering in ToppCluster [19]. The resulted files in the format of xgmml were then subjected to Cytoscape for the preparation of visualisable networks. Regulatory networks of miRNAs in differentiation of osteoblasts were analyzed
ip t
using important signaling pathways, transcription factor, specific markers and antagonistic target of signaling pathways. In the detailed network result, hexagons and boxes were
cr
represented in different size based on their relative importance which means the number of
us
edges joined to particular pathway, gene or miRNA. 2.5. Transient transfection
an
The miR-15b inhibitor (MH10904) and negative control miRNA (4464076) were obtained from Applied Biosystems. miR-15 inhibitor was designed to bind with endogenous
M
miR-15b and thus inhibits its activities in the cells. For transfection, X-treme Gene transfection reagent (Roche) was used in this study. Rat osteoprogenitor cells were seeded in
d
6-well plate with 1 x 105 cells/2 ml growth medium per well on the day before the
te
transfection. The transfection reagent was mixed with 50 nM of miR-15b inhibitor or
Ac ce p
negative control miRNA and transient transfection was carried out according to the manufacturer’s instructions. 2.6. Statistical analysis
The statistical analysis was carried out using one way ANOVA. The significant
difference (P< 0.05) between groups was determined by the student’s t-test. All data were shown as mean ± SD (standard deviation) with n=3. 3. Results and Discussion 3.1. Characterization of osteoblastic differentiation Human MSCs were induced into osteoblastic lineage differentiation by providing osteogenic medium up to 14 days. It was initially confirmed by the determination of 7 Page 7 of 34
increased expression of osteoblastic differentiation marker genes (ALP, type 1 collagen, osteonectin), osteoblast transcription factor (Osterix) (Fig. 1). This supports to the previous evidence that the increased expression of osteoblast marker genes could confirm the
ip t
differentiated status of osteoblasts [20, 21]. hMSCs differentiation into osteoblasts can be strongly induced by several factors that include BMPs, insulin growth factors (IGF) and
cr
osteogenic media containing ascorbic acid, β-glycerol phosphate and dexamethasone. The potential growth factors or osteostimulants have been used in clinical applications for bone
us
regeneration and bone defect treatments and the underling mechanisms of transcript and
an
transcriptome level have also been demonstrated in several studies [1, 2, 22]. However, the miRNA regulatory mechanism via intracellular signaling pathways during MSC
M
differentiation towards osteoblast followed by maturation is still understudied. Figure 1
te
towards osteoblasts
d
3.2. Determination of differentially expressed miRNAs during hMSCs differentiation
Ac ce p
There are evidences which brought forward that differentiation of MSCs into
osteoblasts is tightly regulated by miRNAs [3, 4, 11]. The published data indicated that there is differential expression of miRNAs during osteoblast differentiation [10, 12].
A
comparison of miRNA expression between undifferentiated hMSCs and osteo differentiated hMSCs (differentiated osteoblasts) is shown in Fig. 2A. The result showed that mir-424, mir106a, mir-148a, let-7i and mir-99a were expressed only in undifferentiated hMSCs (MSCs specific miRNAs) and mir-15b, mir-24, mir-130b, mir-30c and mir-130a were expressed only in differentiated osteoblasts (osteoblasts specific miRNAs). This result suggested that the differential expression of miRNAs might have different regulatory roles during osteoblast differentiation. Expression of these osteoblast specific miRNAs was further identified in 8 Page 8 of 34
osteo differentiated hMSCs for 7 days and 14 days, MG63 cell line (human differentiated osteoblast like cells), rat calvarial cells and rat osteoblasts (7 days osteo-differentiated calvarial cells) (Fig. 2B). The result confirmed the expression of these miRNAs in osteo cells
ip t
but their expression level was varied between human and rat. To further confirm the expression of miRNAs by real time RT-PCR analysis, mir-15b and mir-130a along with miR-
cr
21 were analyzed (Fig. 2C). The functional importance of these miRNAs (miR-138, miR206, miR-21, miR-204, miR-133, miR-135) in osteoblast differentiation is well documented
us
[3, 4, 23-25]. Among them, miR-21 was used as a known marker for its positive regulation of
an
osteoblast differentiation [24, 25]. As expected, the result showed the increased expression of mir-15b, mir-130a and mir-21 in 14 days osteo-differentiated hMSCs compared to
M
undifferentiated hMSCs (Fig. 2C). Differential expression of these miRNAs in different types of normal human tissues has also been reported in several studies [12, 26-28]. A selected
d
miRNAs’ expression was analyzed in normal human multiple tissues by the miRNA body-
te
map (http://www.mirnabodymap.org/) analysis [29] (Fig. 2D). This showed that there is differential expression of miRNAs between the tissues.
Ac ce p
Figure 2A Figure 2B Figure 2C Figure 2D
3.3. In silico determination of MSCs specific miRNAs regulatory pathways It is evident from the above that a set of miRNAs are expressed in MSCs and another
set of miRNAs are expressed in differentiated osteoblasts. MSCs specific miRNAs are involved in the process of maintaining the stemcellness by suppressing the genes involved in differentiation of osteoblasts or other lineages [3, 12]. Differentiation of MSCs into osteoblasts is regulated by several pathways namely TGF-β, Wnt and MAPK etc. [22]. Utilizing the bioinformatics tool, DIANA mirPath, we analyzed targets of MSCs specific
9 Page 9 of 34
miRNAs in these pathways (Fig. 3A and Table 2). MSCs specific miRNAs namely miR-424, miR-106a, miR-148a, let-7i and miR-99a were predicted to target 28, 28, 12, 25 and 2 genes, respectively in MAPK pathway and the combined effect of these 5 miRNAs have been
ip t
predicted to target 75 genes, in which some of them were overlapped. In the case of TGF-β pathway, 34 genes were predicted to be their targets. In Wnt signaling pathway, these five
cr
hMSCs specific miRNAs have been recorded to target 19, 18, 11, 10 and 3 genes, respectively, and the united 5 miRNAs have been predicted to target 47 genes, out of which
us
some of them were overlapped. Among the hMSC specific miRNAs, miR-106a has been
an
recorded to target more number of genes in these three major pathways whereas miR-99a has relatively less number of targets. MAPK pathway has been identified as most targeted
M
pathway than TGF-β and Wnt pathways. The comparative analysis of hMSC specific miRNAs targeted pathways is based on their p-values which have been shown in Fig. 3A.
d
Figure 3A Table 2
te
Although various in silico analyser of miRNAs function are available, still they
Ac ce p
remain to consider some conception. As by nature in the biological system, a set of miRNAs do function independently or in combination. Hence, the action of multiple miRNA analysis by bioinformatics tools like DIANA-mirPath seems to be valid but it is limited to single type of illustration and is restricted in the visualization and further scrutiny. ToppCluster based coexpression analysis of multiple miRNAs and their putative targets in various pathways altogether are figured out in Fig. 3B. This network observation not only represents the multiple miRNAs and their associated targets and pathways, but also it shows the relative importance of diverse pathways, miRNAs and targeted genes by representing relative size of boxes, circle and hexagons. In ToppCluster, for example network SMAD5 regulation has
10 Page 10 of 34
been shown distinctly in red color lines (Fig. 3B) and miR-424 and miR-106a were recorded to target SMAD5 that participates in TGF-β, Wnt and BMP receptor signaling pathways. Figure 3B
ip t
Based on the above results among the MSC specific miRNAs, comparatively miR106a has been recorded to influence more on the regulation of osteoblast differentiation
cr
(Table 2) suggesting that further investigation and clinical application of this particular miRNA would expect for better outcome. The result also suggested that the combination of
us
four miRNAs except miR-99a would be more effective on the regulation of osteoblast
an
differentiation. These miRNAs functions have also been reported in different disciplines as follow: miR-424 plays important roles in the regulation of angiogenic activity of endothelial
M
cells, cervical cancer and acute myeloid leukaemia [30-32]. miR-106a and miR-148a were reported to regulate several cancerous cells function [33-35]. let-7i was reported as a marker
d
for dilated cardiomyopathy [36].
te
3.4. In silico determination of osteoblast specific miRNAs regulatory pathways Osteoblast specific miRNAs such as miR-15b, miR-24, miR-130b, miR-30c, and
Ac ce p
miR-130a were analyzed using DIANA-mirPath and ToppCluster which are represented in Figs. 4A & 4B, respectively and Table 2. In DIANA-mirPath analysis, these five miRNAs were predicted to target 40, 21, 29 and 16 genes in ubiquitin mediated proteolysis, gap junction, JAK-STAT signaling pathway and p53 signaling pathway, respectively (Fig. 4A and Table 2). Some of the above mentioned genes were targeted by multiple miRNAs. Among these miRNAs, miR-30c was recorded to target maximum number of genes and the minimum number by miR-24 within these molecular pathways. Putative targets of osteoblast specific miRNAs were incorporated to generate biological regulatory network using ToppCluster (Fig. 4B). The network revealed the information about the relationship among the targeted genes to JAK-STAT, p53, Focal adhesion, Gap junction, Ubiquitin mediated 11 Page 11 of 34
proteolysis, FGF, cell cycle signaling pathways, transcription factors, phenotypic markers and genes involved in negative regulation of osteoblast differentiation. For example, the PDGFRA gene is involved in Gap junction, Focal adhesion and PDGF-alpha signaling
ip t
pathways and it is targeted by multiple miRNAs like miR-15b, miR-130b, miR-24 and miR130a shown in Fig. 4B (red colored lines). The role of osteoblast specific miRNAs has also
cr
been reported in various cases such as, miR-15b as biomarker and its important role in endstage renal disease [37], miR-15b and miR-130b as serum biomarker for diagnosing
us
hepatocellular carcinoma [38] and miR-30c and miR-130a as regulators of various cancers
an
[39, 40].
M
Figure 4A Figure 4B
3.5. Functional analysis of an osteoblast specific miR-15b during osteogenesis The computational analysis provides the complete information of selected miRNAs
d
(MSCs or osteoblasts specific) with their regulatory networks during osteoblast
te
differentiation and experimental validation of these miRNAs would narrow down and expects
Ac ce p
the better results. For example, transfection of these set of miRNAs mimic/inhibitor oligonucleotides to MSCs or osteoblasts would maintain stemcellness, promote osteoblast differentiation or inhibit other lineages differentiation. To investigate the functional role of miR-15b (an osteoblast specific miRNA) on osteoblastic differentiation, rat osteoprogenitor cells were transiently transfected with miR-15b inhibitor or negative control miRNA. We determined the lineage commitment contributed by miR-15b. The MSCs’ major connective tissue lineages are adipose, myoblast, chondrocyte and osteoblast which were analyzed by their specific transcription factor genes like PPARγ, myogenin, FGF2 and Runx2, respectively by real time RT-PCR analysis. The result showed that the inhibition of miR-15b expression by miR-15b inhibitor in rat osteoprogenitor cells enhances mRNA expression of
12 Page 12 of 34
myogenin, PPARγ genes; whereas there was no significant change in Runx2 mRNA expression (Fig. 5A). This result suggested that miR-15b could target genes which may be essential for promoting other cell lineages, and inhibition of miR-15b expression would not
ip t
target those genes, thus promoting other cell lineages. There are evidences clearly indicating a correlation of Runx2 with myogenesis and adipogenesis. It has shown that the increased
cr
Runx2 expression suppresses the myogenesis with the help of Smads [41, 42]. Runx2 has been shown to induce transdifferentiation of primary myoblasts towards osteoblasts by
us
suppressing myogenin expression [43] and miR-133 is involved in lineage determination
an
between myoblasts and osteoblasts from MSCs [13]. PPARγ, a member of nuclear hormone receptor gene superfamily is one of the important transcription factors which up regulates
M
adipogenesis and down regulates osteogenesis by acting in MSCs. Expression of PPARγ is recorded to reduce Runx2 [44, 45], and activation of PPARγ is reduced in osteoblast
d
differentiation and bone mass [46]. Vitamin C deficiency reduces bone mass by promoting
te
transition of osteoblasts into adipocytes through increased expression of PPARγ [47]. It has shown that inhibition of PPARγ expression by miR-20a enhances osteoblast differentiation
Ac ce p
by increasing BMP4, Runx2 and osterix [48]. FGF2, a family of polypeptides, is one of the important factors in regulation of bone
formation and remodeling, involved in the promotion of chondrogenesis and adipogenesis, and potentially contributes to maintenance of MSCs characteristics [49-52]. Additionally, FGF2 is recorded to increase adipogenic lineage differentiation by up regulating PPARγ expression in MSCs and adipose derived stem cells (ASCs) and inhibits osteoblastic lineage differentiation [49, 52]. It has also been shown to increase cell growth and decrease matrix mineralization along with ALP, type I collagen, osteocalcin expression in pre-mature human calvarial osteoblasts [53]. FGF2 was identified as one of the predicted target gene of miR-15b (Fig. 5B). When we used miR-15b inhibitor, there was increased mRNA expression of FGF2 13 Page 13 of 34
but there was no change in Runx2 expression (Fig. 5C). Therefore, miR-15b involves not only in osteoblast differentiation but it could also participate in promotion of differentiation of other lineages.
cr
ip t
Figure 5A Figure 5B Figure 5C Overall, we conclude that there is differential expression of miRNAs during MSCs differentiation towards osteoblasts. The combined in silico analyses of hMSCs or osteoblasts
us
specific miRNAs provide a hint to identify possible miRNA intracellular regulatory network and their gene targets which could regulate osteoblast differentiation. The experimental
an
analysis of miR-15b inhibition by antisense oligonucleotide clearly identified the importance of miR-15b in osteoblast differentiation through suppressing other lineage commitment
M
specific transcription factors like PPARγ, myogenin and FGF2. Thus, miRNAs are potential
Ac ce p
Acknowledgements
te
their expression.
d
candidates to turn the possible physiological functions of MSCs or osteoblasts by altering
This work was supported by a grant from the Indian Council of Medical Research (ICMR), India to N. S. (Grant No: 80/10/2010-BMS). References
[1] N. Selvamurugan, S.C. Jefcoat, S. Kwok, R. Kowalewski, J.A. Tamasi, N.C. Partridge, Overexpression of Runx2 directed by the matrix metalloproteinase-13 promoter containing the AP-1 and Runx/RD/Cbfa sites alters bone remodeling in vivo, J Cell Biochem 99 (2006) 545-557. [2] N. Selvamurugan, S. Kwok, A. Vasilov, S.C. Jefcoat, N.C. Partridge, Effects of BMP-2 and pulsed electromagnetic field (PEMF) on rat primary osteoblastic cell proliferation and gene expression. J Orthop Res 25 (2007) 1213-1220. [3] J.B. Lian, G.S. Stein, A.J. van Wijnen, J.L. Stein, M.Q. Hassan, T. Gaur, Y. Zhang, MicroRNA control of bone formation and homeostasis. Nat Rev Endocrinol 8 (2012) 212227.
14 Page 14 of 34
[4] S. Vimalraj, N. Selvamurugan, MicroRNAs: synthesis, gene regulation and osteoblast differentiation. Curr Issues Mol Biol 15 (2012) 7-18. [5] A. Moorthi, S. Vimalraj, C. Avani, Z. He, N.C. Partridge, N. Selvamurugan, Expression of microRNA-30c and its target genes in human osteoblastic cells by nano-bioglass ceramictreatment, Int J Biol Macromol 56 (2013) 181-185.
ip t
[6] M. Selbach, B. Schwanhausser, N. Thierfelder, Z. Fang, R. Khanin, N. Rajewsky, Widespread changes in protein synthesis induced by microRNAs, Nature 455 (2008) 58-63.
cr
[7] N. Bushati, S.M. Cohen, MicroRNA functions, Annu Rev Cell Dev Biol 23 (2007) 175205.
us
[8] V.K. Gangaraju, J. Lin, MicroRNAs: key regulators of stem cells, Nat Rev Mol Cell Biol 10 (2009) 116-125. [9] S. Vimalraj, P.J. Miranda, B. Ramyakrishna, N. Selvamurugan, Regulation of breast cancer and bone metastasis by microRNAs, Dis Markers 35 (2013) 369-387.
M
an
[10] L.A. Goff, S. Boucher, C.L. Ricupero, S. Fenstermacher, M. Swerdel, L.G. Chase, C.C. Adams, J. Chesnut, U. Lakshmipathy, R.P. Hart, Differentiating human multipotent mesenchymal stromal cells regulate microRNAs: prediction of microRNA regulation by PDGF during ssteogenesis, Exp Hematol 36 (2008) 1354-1369. [11] L. Guo, R.C. Zhao, Y. Wu, The role of microRNAs in self-renewal and differentiation of mesenchymal stem cells, Exp Hematol 39 (2011) 608-616.
te
d
[12] J. Gao, T. Yang, J. Han, K. Yan, X. Qiu, Y. Zhou, Q. Fan, B. Ma, microRNA expression during osteogenic differentiation of human multipotent mesenchymal stromal Cells from Bone Marrow, J Cell Biochem 112 (2011) 1844-1856.
Ac ce p
[13] Z. Li, M.Q. Hassan, S. Volinia, A.J. van Wijnen, J.I. Stein, C.M. Croce, J.B. Lian, G.S. Stein, A microRNA signature for a BMP2-induced osteoblast lineage commitment program, Proc Natl Acad Sci USA 105 (2008) 13906-13911. [14] M.F. Pittenger, A.M. Mackay, S.C. Beck, R.K. Jaiswal, R. Douglas, J.D. Mosca, M.A. Moorman, D.W. Simonetti, S. Craig, D.R. Marshak, Multilineage potential of adult human mesenchymal stem cells, Science 284 (1999) 143-147. [15] J.R. Nefussi, M.L. Boy-Lefevre, H. Boulekbache, N. Forest, Mineralization in vitro of matrix formed by osteoblasts isolated by collagenase digestion, Differentiation 29 (1985) 160. [16] T. Eskildsen, H. Taipaleenmäki, J. Stenvang, B.M. Abdallah, N. Ditzel, A.Y. Nossent, M. Bak, S. Kauppinen, M. Kassem, MicroRNA-138 regulates osteogenic differentiation of human stromal (mesenchymal) stem cells in vivo, Proc Natl Acad Sci USA 108 (2011) 61396144. [17] S.M. Hartig, Basic image analysis and manipulation in ImageJ, Curr Protoc Mol Biol (2013) doi: 10.1002/0471142727.mb1415s102. [18] G.L. Papadopoulos, P. Alexiou, M. Maragkakis, M. Reczko, A.G. Hatzigeorgiou, DIANA-mirPath: integrating human and mouse microRNAs in pathways, Bioinformatics 25 (2009) 1991-1993. 15 Page 15 of 34
[19] V. Kaimal, E.E. Bardes, S.C. Tabar, A.G. Jegga, B.J. Aronow, ToppCluster: a multiple gene list feature analyzer for comparative enrichment clustering and network-based dissection of biological systems, Nucleic Acids Res 38 (2010) W96–W102 (Web Server issue).
ip t
[20] S. Kwok, N.C. Partridge, N. Srinivasan, S.V. Nair, N. Selvamurugan, Mitogen activated protein kinase-dependent inhibition of osteocalcin gene expression by transforming growth factor-beta1, J Cell Biochem 106 (2009) 161-169.
cr
[21] D.D. Liu, J.C. Zhang, Q. Zhang, S.X. Wang, M.S. Yang, TGF-β/BMP signaling pathway is involved in cerium-promoted osteogenic differentiation of mesenchymal stem cells, J Cell Biochem 114 (2013) 1105-1114.
us
[22] G. Chen, C. Deng, Y.P. Li, TGF-β and BMP signaling in osteoblast differentiation and bone formation, Int J Biol Sci 8 (2012) 272-288.
an
[23] H. Inose, H. Ochi, A. Kimura, K. Fujita, R. Xu, S. Sato, M. Iwasaki, S. Sunamura, Y. Takeuchi, S. Fukumoto, K. Saito, T. Nakamura, H. Siomi, H. Ito, Y. Arai, K. Shinomiya, S. Takeda, A microRNA regulatory mechanism of osteoblast differentiation, Proc Natl Acad Sci U S A 106 (2009) 20794-9.
M
[24] O. Trohatou, D. Zagoura, V. Bitsika, K.I. Pappa, A. Antsaklis, N.P. Anagnou, M.G. Roubelakis, Sox2 Suppression by miR-21 Governs Human Mesenchymal Stem Cell Properties, Stem Cells Transl Med (2013) doi: 10.5966/sctm.2013-0081.
te
d
[25] N. Yang, G. Wang, C. Hu, Y. Shi, L. Liao, S. Shi, Y. Cai, S. Cheng, X. Wang, Y. Liu, L. Tang, Y. Ding, Y. Jin, Tumor necrosis factor α suppresses the mesenchymal stem cell osteogenesis promoter miR-21 in estrogen deficiency-induced osteoporosis, J Bone Miner Res 28 (2013) 559-73.
Ac ce p
[26] C.J. Guo, Q. Pan, D.G. Li, H. Sun, B.W. Liu, miR-15b and miR-16 are implicated in activation of the rat hepatic stellate cell: An essential role for apoptosis, J Hepatol 50 (2009) 766-78. [27] H. Nishi, K. Ono, Y. Iwanaga, T. Horie, K. Nagao, G. Takemura, M. Kinoshita, Y. Kuwabara, R.T. Mori, K. Hasegawa, T. Kita, T. Kimura, MicroRNA-15b modulates cellular ATP levels and degenerates mitochondria via Arl2 in neonatal rat cardiac myocytes, J Biol Chem 285 (2010) 4920-30. [28] S. Mujahid, H.C. Nielsen, M.V. Volpe, MiR-221 and miR-130a regulate lung airway and vascular development, PLoS One 8 (2013) e55911. [29] P. Mestdagh, S. Lefever, F. Pattyn, D. Ridzon, E. Fredlund, A. Fieuw, M. Ongenaert, J. Vermeulen, A. De Paepe, L. Wong, F. Speleman, C. Chen, J. Vandesompele, The microRNA body map: dissecting microRNA function through integrative genomics, Nucleic Acids Res 39 (2011) e136. [30] A. Chamorro-Jorganes, E. Araldi, L.O. Penalva, D. Sandhu, C. Fernández-Hernando, Y. Suárez, MicroRNA-16 and microRNA-424 regulate cell-autonomous angiogenic functions in endothelial cells via targeting vascular endothelial growth factor receptor-2 and fibroblast growth factor receptor-1, Arterioscler Thromb Vasc Biol 31 (2011) 2595-606.
16 Page 16 of 34
[31] J. Xu, Y. Li, F. Wang, X. Wang, B. Cheng, F. Ye, X. Xie, C. Zhou, W. Lu, Suppressed miR-424 expression via upregulation of target gene Chk1 contributes to the progression of cervical cancer, Oncogene (2012) doi: 10.1038/onc.2012.121. [32] I. Faraoni, S. Laterza, D. Ardiri, C. Ciardi, F. Fazi, F. Lo-Coco, MiR-424 and miR-155 deregulated expression in cytogenetically normal acute myeloid leukaemia: correlation with NPM1 and FLT3 mutation status, J Hematol Oncol 8 (2012) 5-26.
ip t
[33] Y. Jiang, Y. Wu, A.R. Greenlee, J. Wu, Z. Han, X. Li, Y. Zhao, miR-106a-mediated malignant transformation of cells induced by anti-benzo[a]pyrene-trans-7,8-diol-9,10epoxide, Toxicol Sci 119 (2011) 50-60.
us
cr
[34] R. Hummel, D.I. Watson, C. Smith, J. Kist, M.Z. Michael, J. Haier, D.J. Hussey, Mir148a improves response to chemotherapy in sensitive and resistant oesophageal adenocarcinoma and squamous cell carcinoma cells, J Gastrointest Surg 15 (2011) 429-38.
an
[35] Z. Wang, M. Liu, H. Zhu, W. Zhang, S. He, C. Hu, L. Quan, J. Bai, N. Xu, miR-106a is frequently upregulated in gastric cancer and inhibits the extrinsic apoptotic pathway by targeting FAS, Mol Carcinog (2012) doi: 10.1002/mc.21899.
M
[36] M. Satoh, Y. Minami, Y. Takahashi, T. Tabuchi, M. Nakamura, A cellular microRNA, let-7i, is a novel biomarker for clinical outcome in patients with dilated cardiomyopathy, J Card Fail 17 (2011) 923-929.
d
[37] H. Wang, W. Peng, X. Ouyang, Y. Dai, Reduced circulating miR-15b is correlated with phosphate metabolism in patients with end-stage renal disease on maintenance hemodialysis, Ren Fail 34 (2012) 685-690.
te
[38] A.M. Liu, T.J. Yao, W. Wang, K.F. Wong, N.P. Lee, S.T. Fan, R.T. Poon, C. Gao, J.M. Luk, Circulating miR-15b and miR-130b in serum as potential markers for detecting hepatocellular carcinoma: a retrospective cohort study, BMJ Open 8 (2012) 2:e000825.
Ac ce p
[39] M. Tanic, K. Yanowsky, C. Rodriguez-Antona, R. Andrés, I. Márquez-Rodas, A. Osorio, J. Benitez, B. Martinez-Delgado, Deregulated miRNAs in hereditary breast cancer revealed a role for miR-30c in regulating KRAS oncogene, PLoS One 7 (2012) e38847. [40] L. Yang, N. Li, H. Wang, X. Jia, H. Wang, J. Luo, Altered microRNA expression in cisplatin-resistant ovarian cancer cells and upregulation of miR-130a associated with MDR1/P-glycoprotein-mediated drug resistance, Oncol Rep 28 (2012) 592-600. [41] L.J. Raggatt, L. Qin, J. Tamasi, S.C. Jefcoat Jr, E. Shimizu, N. Selvamurugan, F.Y. Liew, L. Bevelock, J.H. Feyen, N.C. Partridge, Interleukin-18 is regulated by parathyroid hormone and is required for its bone anabolic actions, J Biol Chem 283 (2008) 6790-8. [42] K.S. Lee, H.J. Kim, Q.L. Li, X.Z. Chi, C. Ueta, T. Komori, J.M. Wozney, E.G. Kim, J.Y. Choi, H.M. Ryoo, S.C. Bae, Runx2 is a common target of transforming growth factor beta1 and bone morphogenetic protein 2, and cooperation between Runx2 and Smad5 induces osteoblast-specific gene expression in the pluripotent mesenchymal precursor cell line, 2C12. Mol Cell Biol 20 (2000) 8783-92. [43] C.A. Gersbach, B.A. Byers, G.K. Pavlath, A.J. García, Runx2/Cbfa1 stimulates transdifferentiation of primary skeletal myoblasts into a mineralizing osteoblastic phenotype, Exp Cell Res 300 (2004) 406-17. 17 Page 17 of 34
[44] M.J. Jeon, J.A. Kim, S.H. Kwon, S.W. Kim, K.S. Park, S.W. Park, S.Y. Kim, C.S. Shin, Activation of peroxisome proliferatoractivated receptor-gamma inhibits the Runx2-mediated transcription of osteocalcin in osteoblasts, J Biol Chem 278 (2003) 23270-7. [45] J.B. Lian, G.S. Stein, A. Javed, A.J. van Wijnen, J.L. Stein, M. Montecino, M.Q. Hassan, T. Gaur, C.J. Lengner, D.W. Young, Networks and hubs for the transcriptional control of osteoblastogenesis, Rev Endocr Metab Disord 7 (2006) 1-16.
ip t
[46] A.A. Ali, R.S. Weinstein, S.A. Stewart, A.M. Parfitt, S.C. Manolagas, R.L. Jilka, Rosiglitazone causes bone loss in mice by suppressing osteoblast differentiation and bone formation, Endocrinol 146 (2005) 1226-35.
us
cr
[47] J.K. Park, E.M. Lee, A.Y. Kim, E.J. Lee, C.W. Min, K.K. Kang, M.M. Lee, K.S. Jeong, Vitamin C deficiency accelerates bone loss inducing an increase in PPAR-γ expression in SMP30 knockout mice, Int J Exp Pathol 93(2012):332-40.
an
[48] J.F. Zhang, W.M. Fu, M.L. He, W.D. Xie, Q. Lv, G. Wan, G. Li, H. Wang, G. Lu, X. Hu, S. Jiang, J.N. Li, M.C. Lin, Y.O. Zhang, H.F. Kung, MiRNA-20a promotes osteogenic differentiation of human mesenchymal stem cells by co-regulating BMP signaling, RNA Biol 8 (2011) 829-38.
M
[49] M. Neubauer, C. Fischbach, P. Bauer-Kreisel, E. Lieb, M. Hacker, J. Tessmar, M.B. Schulz, A. Goepferich, T. Blunk, Basic fibroblast growth factor enhances PPARgamma ligand-induced adipogenesis of mesenchymal stem cells, FEBS Lett 577 (2004) 277-283.
d
[50] A.M. Handorf, W.J. Li, Fibroblast growth factor-2 primes human mesenchymal stem cells for enhanced chondrogenesis, PLoS One 6 (2011) e22887.
Ac ce p
te
[51] W.T. Lai, V. Krishnappa, D.G. Phinney, Fibroblast growth factor 2 (Fgf2) inhibits differentiation of mesenchymal stem cells by inducing Twist2 and Spry4, blocking extracellular regulated kinase activation, and altering FGF receptor expression levels, Stem Cells 29 (2011) 1102-11. [52] N. Kakudo, A. Shimotsuma, K. Kusumoto K. Fibroblast growth factor-2 stimulates adipogenic differentiation of human adipose-derived stem cells. Biochem Biophys Res Commun 359 (2007) 239-44. [53] P.J. Marie, F. Debiais, E. Haÿ, Regulation of human cranial osteoblast phenotype by FGF-2, FGFR-2 and BMP-2 signaling, Histol Histopathol 17 (2002) 877-85.
18 Page 18 of 34
Figure legends Figure 1. Human MSCs differentiation towards osteoblasts. hMSCs were induced into osteoblastic differentiation by providing osteogenic medium. The differentiation of osteoblast was confirmed at molecular level by semi quantitative RT-PCR analysis of osteoblast marker genes (ALP, osteonectin, and type 1 collagen) and transcription factor gene, Osterix. The internal controls used in this study are GAPDH and U6 genes.
M
an
us
cr
ip t
Figure 2. Expression of miRNAs during osteoblast differentiation. (A) Expression of MSCs and osteoblasts specific precursor miRNAs analyzed in undifferentiated hMSCs and 14 days osteo-differentiated hMSCs by semi quantitative RT-PCR analysis. * indicates significant up regulation compared to 14 days osteo-differentiated cells and ** indicates significant up regulation compared to undifferentiated hMSC. (B) The osteoblast specific miRNAs (mir15b, mir-24, mir-130b, mir-30c, and mir-130a) expression was analyzed in 7-days osteodifferentiated hMSCs, 14-days osteo-differentiated hMSCs, human differentiated osteoblasts (MG63), rat calvarial cells and rat osteoblasts by semi quantitative RT-PCR analysis. (C) Real time RT-PCR analysis of mir-15b, miR-130a and miR-21 expression in undifferentiated hMSCs and 14 days osteo differentiated hMSCs. *represents significant up regulation compared to hMSCs. (D) Heat map of differential expression of miRNAs in normal human tissues. Differential expression of MSCs and osteoblast specific miRNAs was identified in various normal human tissues by miRNA body map along with the hierarchical cluster analysis. Expression of miRNAs is represented as blue (down regulated), red (up regulated), and white colors (no significant change or absence of data).
Ac ce p
te
d
Figure 3. Putative targets of hMSCs specific miRNAs (miR-424, miR-106a, miR-148a, let-7i and miR-99a) in major signaling pathways involved in osteoblast differentiation. Kyoto Encyclopedia of Genes and Genomes (KEGG) was used to identify the presence of putative targets. Figures were generated from the DIANA-MicroT4.0 (beta version) databases. (A) A combinational effect of MSCs specific miRNAs in various pathways is visible in bar diagram of the -ln P values. (B) The cytoscape graph networks were generated from ToppCluster network analyser. The functional enrichment analysis of regulatory network influenced by MSCs specific miRNAs is shown. Figure 4. Putative targets of osteoblasts specific miRNAs (miR-15b, miR-24, miR-130b, miR-30c, and miR-130a) in major signaling pathways involved in osteoblast differentiation. Kyoto Encyclopedia of Genes and Genomes (KEGG) was used to identify the presence of putative targets. Figures were generated from the DIANA-MicroT4.0 (beta version) databases. (A) A combinational effect of osteoblast specific miRNAs in various pathways is visible in bar diagram of the -ln P values. (B) The cytoscape graph networks were generated from ToppCluster network analyser. The functional enrichment analysis of regulatory network influenced by osteoblasts specific miRNAs is shown. Figure 5. miR-15b inhibitor up regulates expression of PPARγ, myogenin and FGF2. (A) Rat osteoprogenitor cells were transiently transfected with 50 nM of negative control miRNA or miR-15b inhibitor for 7 days. Total RNA was isolated and real time RT-PCR was carried out using the primers for PPARγ, myogenin and Runx2 genes. The fold change of expression of 19 Page 19 of 34
Ac ce p
te
d
M
an
us
cr
ip t
mRNAs was calculated over negative control miRNA. GAPDH gene was used for normalization. (B) The putative target region analysis was performed for FGF2 mRNAs 3’ UTR by miR-15b seed sequence. Interspecies similarity of putative miR-15b binding region with the 3’UTR of FGF2 was shown. (C) hMSCs were transfected with miR-15b inhibitor in normal medium up to 7 days and the real time RT-PCR was performed for FGF2 and Runx2 genes. GAPDH gene was used for normalization. * indicates significant up regulation of mRNA compared to negative control miRNA.
20 Page 20 of 34
ip t
Table(s)
cr
Table 1. Primer sequences. Name of the miRNA/genes
mir-148a let-7i mir-99a
M
mir-15b mir-24
mir-30c
ep te
mir-130a
d
mir-130b
Human Runx2 Rat Runx2
Type 1 collagen
Ac c
Osteonectin Osterix ALP PPARγ Myogenin FGF2
Human GAPDH Rat GAPDH U6
CGAGGGGATACAGCAGCAATTC CCCCACCTTCTACCTTCCCCAC CCTTGGCCATGTAAAAGTGC CCATGGTAATGTAAGAAGTGC GAGGCAAAGTTCTGAGACAC GAGACAAAGTTCTGTAGTGCACT AGTAGTTTGTGCTGTTGGTCGGGT GCAAGGCAGTAGCTTGCGCAGT CCATTGGCATAAACCCGTAGATCCG CAGACCCATAGAAGCGAGCTTGTGC GGCCTTAAAGTACTGTAGCAGC CCTTAAATTTCTAGAGCAGC CTCCGGTGCCTACTGAGCTG CTCCTGTTCCTGCTGAACTG GGCCTGCCCGACACTCTTTC GACCTGACCGATGCCCTTTC TGTGTAAACATCCTACACTCTCAG GAGTAAACAACCCTCTCCCA GCTGGCCAGAGCTCTTTTCACA CACTACACGGCCAATGCCCTTT CAGTTCCCAAGCATTTCATC TCAATATGGTCGCCAAACAG CAAGTGGCCAGGTTCAACGA GTGAAGACCGTTATGGTCAAAGTG TAACCCCCTCCCCAGCCACAAA TTCCTCTTGGCCGTGCGTCA GCAGTGGTGGGTCCGTGGTC CTGTCACGGCTCGGGTGTGC ACTGGCTAGGTGGTGGTCAG GGTAGGGAGCTGGGTTAAGG AGGCAGGATTGACCACGG TGTAGTTCTGCTCATGGA CCTGAAGCTCCAAGAATACCAAA AGAGTTGGGTTTTTTCAGAATAATAAGG TGAAGAGAAGCACCCTGCTCA TCAATGTACTGGATGGCACTG C CAGATTAGCGGACGCGGT CGGATGGGTGTCTCCGC GAGAGACCCCACTTGCTGCCA CTCACACTGCCCCTCCCTGGT GCAGAGGTTGAATGTGAGCA GGAAGAAGTTCCCATCGTCA CTCGCTTCGGCAGCACA AACGCTTCACGAATTTGCGT
us
mir-106a
Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse
an
mir-424
Sequence (5'->3')
Page 21 of 34
ip t
hMSCs specific miRNAs
-ln (p-value)
hsa-let-7i
hsa-miR-99a
Union
No. of genes No. of genes No. of genes No. of genes No. of genes -ln (p-value) -ln (p-value) -ln (p-value) -ln (p-value) -ln (p-value) targeted targeted targeted targeted targeted
28
12.34
28
9.06
TGF-beta pathway (hsa04350)
8
0.74
13
7.35
Wnt pathway (hsa04310)
19
6.38
18
12
1.17
25
12.12
2
1.23
75
18.19
11
10.78
10
5.86
-
-
34
15.7
11
4.41
10
1.83
3
7.48
47
14.69
M
MAPK pathway (hsa04510)
7.35
d ep te
Osteoblast specific miRNAs hsa-miR-15b Name of the pathway and KEGG pathway id No. of genes targeted
12
-ln (p-value)
3.22
Ac c
Ubiquitin mediated proteolysis (hsa04120)
hsa-miR-148a
us
targeted
hsa-miR-106a
an
hsa-miR-424 Name of the pathway and KEGG pathway id No. of genes
cr
Table 2. KEGG Pathways affected by putative targets of miRNAs.
hsa-miR-24
hsa-miR-130b
hsa-miR-30c
hsa-miR-130a
Union
No. of genes No. of genes No. of genes No. of genes No. of genes -ln (p-value) -ln (p-value) -ln (p-value) -ln (p-value) -ln (p-value) targeted targeted targeted targeted targeted
5
0.27
13
5.02
22
14.88
12
4.19
40
13.32
Gap junction (hsa04540)
7
0.92
4
0.35
8
2.03
8
0.81
8
2.25
21
2.36
Jak-STAT signalling pathway (hsa04630)
10
0.77
4
0.02
5
0.62
13
1.38
5
0.51
29
1.39
p53 signalling pathway (hsa04115)
9
5.33
4
1.15
4
1.13
2
0.93
4
0.19
16
2.24
Page 22 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 23 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 24 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 25 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s) 2c
Page 26 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s) 2D
Page 27 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 28 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 29 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 30 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 31 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 32 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s) 5B
Page 33 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s) 5C
Page 34 of 34
S0141-8130(14)00114-7 http://dx.doi.org/doi:10.1016/j.ijbiomac.2014.02.030 BIOMAC 4184
To appear in:
International Journal of Biological Macromolecules
Received date: Revised date: Accepted date:
23-10-2013 18-12-2013 13-2-2014
Please cite this article as: S. VimalrajN. Selvamurugan MicroRNAs Expression and their Regulatory Networks During Mesenchymal Stem Cells Differentiation towards Osteoblasts (2014), http://dx.doi.org/10.1016/j.ijbiomac.2014.02.030 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.
MicroRNAs Expression and their Regulatory Networks During Mesenchymal Stem Cells
us
cr
S. Vimalraj and N. Selvamurugan*
ip t
Differentiation towards Osteoblasts
an
Department of Biotechnology, School of Bioengineering, SRM University, Kattankulathur,
te
d
M
Tamil Nadu. India.
Ac ce p
*To whom correspondence should be made: N. Selvamurugan, Ph. D. Professor
Department of Biotechnology School of Bioengineering SRM University
Kattankulathur 603 203. Tamil Nadu. India. Cell: 91-9940632335 E mail: [email protected] [email protected]
1 Page 1 of 34
Abstract MicroRNAs (miRNAs) are small endogenous non coding RNAs which regulate mRNAs post-transcriptionally. In this study, a selective number of miRNAs was investigated
ip t
for their expression and intracellular regulatory networks involved in differentiation of human mesenchymal stem cells (hMSCs) towards osteoblasts. The expression of miR-424,
cr
miR-106a, miR-148a, let-7i and miR-99a miRNAs was found to be specific in hMSCs; whereas expression of miR-15b, miR-24, miR-130b, miR-30c, and miR-130a miRNAs was
us
found to be specific in differentiated osteoblasts. A bioinformatics approach identified that the MAPK pathway was mostly targeted by hMSCs specific miRNAs; whereas JAK-STAT,
an
p53, Focal adhesion, Gap junction, Ubiquitin mediated proteolysis pathways were targeted by osteblastic specific miRNAs. Altering expression of osteoblast specific miRNA (miR-15b)
M
promoted adipogenesis and myogenesis lineages. Thus, we suggest that miRNAs’ regulatory networks and their target genes might provide an insight of their role during differentiation of
d
hMSCs towards osteoblasts, and alteration in the expression of miRNAs would be a valuable
te
approach for controlling osteoblast differentiation.
Ac ce p
Keywords: MicroRNAs, Mesenchymal Stem Cells, Osteoblasts, miR-15b, ToppCluster
2 Page 2 of 34
1. Introduction The process of skeletal development involves differentiation and maturation of osteoblast, a mesenchymal origin cell type. During this developmental process, several
ip t
signaling pathways like TGF-β, BMP, Wnt, and MAPK are activated. Runx2 and osterix are the major transcription factors involved in osteogenesis and skeletal formation by activating
cr
osteoblast differentiation marker genes [1, 2]. Mesenchymal stem cells (MSCs) have the
us
potential to differentiate into multiple lineages with response to respective transcription factors, for instance PPARγ for adipose, myogenin for myoblast, FGF2 for chondrocyte and
an
Runx2 for osteoblast differentiation. There are several regulatory mechanisms involved in differentiation of MSCs into osteoblasts which may be essential for maintaining sustainable
M
bone mass. Understanding the regulatory mechanisms would also develop new strategies to treat bone related diseases like osteoporosis, osteosclerosis.
d
Osteoblast differentiation and maturation is typically controlled by microRNAs
te
(miRNAs) [3-5]. miRNAs are small endogenous non coding RNA molecules (19-25 nt) that
Ac ce p
regulate mRNAs at posttranscriptional level. A single miRNA can target more than hundred mRNAs, and a single mRNA can also be targeted by multiple miRNAs suggesting that miRNAs might have a wide range of structural and functional networks [6]. They control several physiological process including cell growth, apoptosis, metabolism, hormone signaling, stress adaptation, proliferation and differentiation by direct or indirect regulation [4, 5, 7-9]. During osteoblast differentiation, a significance variation in miRNAs expression profile has been reported between undifferentiated hMSCs and osteo differentiated hMSCs [10-12]. Several miRNAs have been identified during differentiation of hMSCs into osteoblasts by miRNA array profiling but their functional roles are still unexplored [10,11,13] 3 Page 3 of 34
and for some miRNAs alone, limited number of targets have been detected which is not sufficient to go for clinical application of particular miRNA because it may elicit some other undesired physiological functions. Thus, the aim of this study was to identify expression of
ip t
miRNAs during differentiation of hMSCs into osteoblasts, and to reveal their complete regulatory network and target genes by bioinformatics tools. Based on the computational
cr
analysis, miR-15b, an osteoblast specific miRNA was preferably selected for further functional analysis and we identified that the inhibition of miR-15b expression favored
us
towards other cell lineages.
an
2. Materials and Methods
2.1. In vitro culture of hBMSCs and osteoblast differentiation
M
Human bone marrow stromal cells (hBMSCs) was isolated from 27 years old female and cultured according to previously described method [14]. The Institutional review
d
committee in SRM University approved the protocol for isolation of hMSCs. BMSCs were
te
obtained with the written consent from donor. Rat calvarial cells isolation from 2 days old rat
Ac ce p
pups was carried out as described previously [15]. All animal procedures were approved by the University of Madras and the SRM university animal ethical committees. The MG63 cell line, human osteoblast like cells were purchased from NCCS, Pune, India. The cells were induced to osteo differentiation in osteogenic medium [alpha MEM (Cellgro), 15% FBS (Invitrogen), Penicillin-Streptomycin (Invitrogen), 10-4 M L-Ascorbic Acid, 10-8 M Dexamethasone and 1.80 mM Potassium phosphate monobasic (Sigma)]. The medium was changed once in three days. 2.2. miRNA and mRNA expression A list of miRNAs was selected based on the microarray profiling in hMSCs and osteo differentiated hMSCs and based on their functionally important putative targets associated
4 Page 4 of 34
with differentiation of osteoblasts [10, 12, 13, 16]. Putative target prediction online tools such as
microrna
(http://www.microrna.org/microrna/home.do),
TargetScan
(http://www.targetscan.org/) and PicTar (http://pictar.mdc-berlin.de/) were used to find out
ip t
the miRNAs’ potential target genes. The stem loop sequence of miRNA was retrieved from miRNA database (http://www.mirbase.org/) and according to them, primers were designed
cr
(Table 1). The expression of these miRNAs was analyzed by semi quantitative RT-PCR method. Total RNA was isolated from cultured cells with Trizol reagent (Invitrogen)
us
according to the manufacturer’s instructions and was quantitatively evaluated by Qubit 2.0
an
Flurometer (Invitrogen). cDNA was synthesized using GeNei AMV RT-PCR kit according to the manufacturer’s protocol as follows: 1 µg of total RNA, 1µl of oligo(dT) or random
M
hexamer (50µM) and 1µl of dNTP Mixture (10mM) were added and made up to 14µl reverse transcript reaction solution using RNase free dH2O, followed by incubation at 65ºC for 5 min
d
and subjected to fast chill using ice cubes. Then 4µl of 5X First-strand Buffer, 1µl of Reverse
te
Transcriptase (200U/μl) and 1µl of RNase Inhibitor (40U/μl) were added to 14µl reverse transcript reaction solution and this reaction mixture was kept in 30°C for 10 min, 42°C for
Ac ce p
50 min and 95°C for 5 min in thermal cycler. The PCR reaction for each miRNA was performed in triplicate and the products were fractionated using 2.5% agarose gel electrophoresis and. The intensity of bands was measured by densitometry scanning (ImageJ software (http://rsbweb.nih.gov/ij/)) [17] and the miRNAs band intensities were normalized with internal control U6 band intensity. The average of respective band intensity was blotted in the form of bar diagram with standard deviation. Table 1 The real-time PCR analysis was performed in Bio-Rad system using KAPA SYBR FAST qPCR Kits (Kapabiosystem) according to the manufacturer’s instruction. Stem loop sequence of miR-15b primers or mRNA primers and internal control (U6 or GAPDH) 5 Page 5 of 34
primers (Table 1) were used for amplification. The Ct (threshold cycle) values were calculated from amplification curve. ΔCtneg (ΔCt negative control) value of candidate gene of negative control miRNA transfected cells was calculated by subtracting the Ct value of its
ip t
internal control gene. Similarly, ΔCtinh (ΔCt miR-15b inhibitor) value of miR-15b inhibitor transfected cells was calculated by subtracting the Ct value of its internal control gene. The
cr
actual ΔΔCt value was calculated by subtracting ΔCtneg with ΔCtinh and then the ΔΔCt values were converted to fold change over negative control miRNA by raising 2 to the ΔΔCt power.
us
The thermal cycling conditions were as follows: 95ºC for 5 min as initial denaturation,
an
followed by 40 cycles of 95ºC for 30s, 58ºC for 30s and 72ºC for 30s, with a final extension at 72ºC for 5 min.
DIANA-mirPath
is
a
M
2.3. DIANA-mirPath analysis
free
online
bioinformatics
tool
d
(http://diana.cslab.ece.ntua.gr/?sec=home) to find out molecular signaling pathways targeted
te
by single miRNA or combined effect of multiple miRNAs on a given pathway [18]. This tool was used to carry out an enrichment analysis of input datasets by comparing each set of
Ac ce p
targeted genes to all available biological pathways which is provided by Kyoto Encyclopedia of Genes and Genomes (KEGG). Furthermore, this software allocates an overall priority score (enrichment p-value) based on the predicted strength of the miRNA interactions with the molecules of the target pathway. Enrichment analysis of each KEGG signaling pathway was represented by negative natural log of P value (-ln P). The enrichment -lnP values bar graph helps to distinguish the relative importance of each pathway in various dataset. 2.4. ToppCluster analysis ToppCluster
(http://toppcluster.cchmc.org/)
and
Cytoscape
(http://www.cytoscape.org/) were used for biological functional enrichment analysis and preparation of network figures. The miRNAs putative targets were used to generate 6 Page 6 of 34
regulatory network clusters by hierarchical clustering in ToppCluster [19]. The resulted files in the format of xgmml were then subjected to Cytoscape for the preparation of visualisable networks. Regulatory networks of miRNAs in differentiation of osteoblasts were analyzed
ip t
using important signaling pathways, transcription factor, specific markers and antagonistic target of signaling pathways. In the detailed network result, hexagons and boxes were
cr
represented in different size based on their relative importance which means the number of
us
edges joined to particular pathway, gene or miRNA. 2.5. Transient transfection
an
The miR-15b inhibitor (MH10904) and negative control miRNA (4464076) were obtained from Applied Biosystems. miR-15 inhibitor was designed to bind with endogenous
M
miR-15b and thus inhibits its activities in the cells. For transfection, X-treme Gene transfection reagent (Roche) was used in this study. Rat osteoprogenitor cells were seeded in
d
6-well plate with 1 x 105 cells/2 ml growth medium per well on the day before the
te
transfection. The transfection reagent was mixed with 50 nM of miR-15b inhibitor or
Ac ce p
negative control miRNA and transient transfection was carried out according to the manufacturer’s instructions. 2.6. Statistical analysis
The statistical analysis was carried out using one way ANOVA. The significant
difference (P< 0.05) between groups was determined by the student’s t-test. All data were shown as mean ± SD (standard deviation) with n=3. 3. Results and Discussion 3.1. Characterization of osteoblastic differentiation Human MSCs were induced into osteoblastic lineage differentiation by providing osteogenic medium up to 14 days. It was initially confirmed by the determination of 7 Page 7 of 34
increased expression of osteoblastic differentiation marker genes (ALP, type 1 collagen, osteonectin), osteoblast transcription factor (Osterix) (Fig. 1). This supports to the previous evidence that the increased expression of osteoblast marker genes could confirm the
ip t
differentiated status of osteoblasts [20, 21]. hMSCs differentiation into osteoblasts can be strongly induced by several factors that include BMPs, insulin growth factors (IGF) and
cr
osteogenic media containing ascorbic acid, β-glycerol phosphate and dexamethasone. The potential growth factors or osteostimulants have been used in clinical applications for bone
us
regeneration and bone defect treatments and the underling mechanisms of transcript and
an
transcriptome level have also been demonstrated in several studies [1, 2, 22]. However, the miRNA regulatory mechanism via intracellular signaling pathways during MSC
M
differentiation towards osteoblast followed by maturation is still understudied. Figure 1
te
towards osteoblasts
d
3.2. Determination of differentially expressed miRNAs during hMSCs differentiation
Ac ce p
There are evidences which brought forward that differentiation of MSCs into
osteoblasts is tightly regulated by miRNAs [3, 4, 11]. The published data indicated that there is differential expression of miRNAs during osteoblast differentiation [10, 12].
A
comparison of miRNA expression between undifferentiated hMSCs and osteo differentiated hMSCs (differentiated osteoblasts) is shown in Fig. 2A. The result showed that mir-424, mir106a, mir-148a, let-7i and mir-99a were expressed only in undifferentiated hMSCs (MSCs specific miRNAs) and mir-15b, mir-24, mir-130b, mir-30c and mir-130a were expressed only in differentiated osteoblasts (osteoblasts specific miRNAs). This result suggested that the differential expression of miRNAs might have different regulatory roles during osteoblast differentiation. Expression of these osteoblast specific miRNAs was further identified in 8 Page 8 of 34
osteo differentiated hMSCs for 7 days and 14 days, MG63 cell line (human differentiated osteoblast like cells), rat calvarial cells and rat osteoblasts (7 days osteo-differentiated calvarial cells) (Fig. 2B). The result confirmed the expression of these miRNAs in osteo cells
ip t
but their expression level was varied between human and rat. To further confirm the expression of miRNAs by real time RT-PCR analysis, mir-15b and mir-130a along with miR-
cr
21 were analyzed (Fig. 2C). The functional importance of these miRNAs (miR-138, miR206, miR-21, miR-204, miR-133, miR-135) in osteoblast differentiation is well documented
us
[3, 4, 23-25]. Among them, miR-21 was used as a known marker for its positive regulation of
an
osteoblast differentiation [24, 25]. As expected, the result showed the increased expression of mir-15b, mir-130a and mir-21 in 14 days osteo-differentiated hMSCs compared to
M
undifferentiated hMSCs (Fig. 2C). Differential expression of these miRNAs in different types of normal human tissues has also been reported in several studies [12, 26-28]. A selected
d
miRNAs’ expression was analyzed in normal human multiple tissues by the miRNA body-
te
map (http://www.mirnabodymap.org/) analysis [29] (Fig. 2D). This showed that there is differential expression of miRNAs between the tissues.
Ac ce p
Figure 2A Figure 2B Figure 2C Figure 2D
3.3. In silico determination of MSCs specific miRNAs regulatory pathways It is evident from the above that a set of miRNAs are expressed in MSCs and another
set of miRNAs are expressed in differentiated osteoblasts. MSCs specific miRNAs are involved in the process of maintaining the stemcellness by suppressing the genes involved in differentiation of osteoblasts or other lineages [3, 12]. Differentiation of MSCs into osteoblasts is regulated by several pathways namely TGF-β, Wnt and MAPK etc. [22]. Utilizing the bioinformatics tool, DIANA mirPath, we analyzed targets of MSCs specific
9 Page 9 of 34
miRNAs in these pathways (Fig. 3A and Table 2). MSCs specific miRNAs namely miR-424, miR-106a, miR-148a, let-7i and miR-99a were predicted to target 28, 28, 12, 25 and 2 genes, respectively in MAPK pathway and the combined effect of these 5 miRNAs have been
ip t
predicted to target 75 genes, in which some of them were overlapped. In the case of TGF-β pathway, 34 genes were predicted to be their targets. In Wnt signaling pathway, these five
cr
hMSCs specific miRNAs have been recorded to target 19, 18, 11, 10 and 3 genes, respectively, and the united 5 miRNAs have been predicted to target 47 genes, out of which
us
some of them were overlapped. Among the hMSC specific miRNAs, miR-106a has been
an
recorded to target more number of genes in these three major pathways whereas miR-99a has relatively less number of targets. MAPK pathway has been identified as most targeted
M
pathway than TGF-β and Wnt pathways. The comparative analysis of hMSC specific miRNAs targeted pathways is based on their p-values which have been shown in Fig. 3A.
d
Figure 3A Table 2
te
Although various in silico analyser of miRNAs function are available, still they
Ac ce p
remain to consider some conception. As by nature in the biological system, a set of miRNAs do function independently or in combination. Hence, the action of multiple miRNA analysis by bioinformatics tools like DIANA-mirPath seems to be valid but it is limited to single type of illustration and is restricted in the visualization and further scrutiny. ToppCluster based coexpression analysis of multiple miRNAs and their putative targets in various pathways altogether are figured out in Fig. 3B. This network observation not only represents the multiple miRNAs and their associated targets and pathways, but also it shows the relative importance of diverse pathways, miRNAs and targeted genes by representing relative size of boxes, circle and hexagons. In ToppCluster, for example network SMAD5 regulation has
10 Page 10 of 34
been shown distinctly in red color lines (Fig. 3B) and miR-424 and miR-106a were recorded to target SMAD5 that participates in TGF-β, Wnt and BMP receptor signaling pathways. Figure 3B
ip t
Based on the above results among the MSC specific miRNAs, comparatively miR106a has been recorded to influence more on the regulation of osteoblast differentiation
cr
(Table 2) suggesting that further investigation and clinical application of this particular miRNA would expect for better outcome. The result also suggested that the combination of
us
four miRNAs except miR-99a would be more effective on the regulation of osteoblast
an
differentiation. These miRNAs functions have also been reported in different disciplines as follow: miR-424 plays important roles in the regulation of angiogenic activity of endothelial
M
cells, cervical cancer and acute myeloid leukaemia [30-32]. miR-106a and miR-148a were reported to regulate several cancerous cells function [33-35]. let-7i was reported as a marker
d
for dilated cardiomyopathy [36].
te
3.4. In silico determination of osteoblast specific miRNAs regulatory pathways Osteoblast specific miRNAs such as miR-15b, miR-24, miR-130b, miR-30c, and
Ac ce p
miR-130a were analyzed using DIANA-mirPath and ToppCluster which are represented in Figs. 4A & 4B, respectively and Table 2. In DIANA-mirPath analysis, these five miRNAs were predicted to target 40, 21, 29 and 16 genes in ubiquitin mediated proteolysis, gap junction, JAK-STAT signaling pathway and p53 signaling pathway, respectively (Fig. 4A and Table 2). Some of the above mentioned genes were targeted by multiple miRNAs. Among these miRNAs, miR-30c was recorded to target maximum number of genes and the minimum number by miR-24 within these molecular pathways. Putative targets of osteoblast specific miRNAs were incorporated to generate biological regulatory network using ToppCluster (Fig. 4B). The network revealed the information about the relationship among the targeted genes to JAK-STAT, p53, Focal adhesion, Gap junction, Ubiquitin mediated 11 Page 11 of 34
proteolysis, FGF, cell cycle signaling pathways, transcription factors, phenotypic markers and genes involved in negative regulation of osteoblast differentiation. For example, the PDGFRA gene is involved in Gap junction, Focal adhesion and PDGF-alpha signaling
ip t
pathways and it is targeted by multiple miRNAs like miR-15b, miR-130b, miR-24 and miR130a shown in Fig. 4B (red colored lines). The role of osteoblast specific miRNAs has also
cr
been reported in various cases such as, miR-15b as biomarker and its important role in endstage renal disease [37], miR-15b and miR-130b as serum biomarker for diagnosing
us
hepatocellular carcinoma [38] and miR-30c and miR-130a as regulators of various cancers
an
[39, 40].
M
Figure 4A Figure 4B
3.5. Functional analysis of an osteoblast specific miR-15b during osteogenesis The computational analysis provides the complete information of selected miRNAs
d
(MSCs or osteoblasts specific) with their regulatory networks during osteoblast
te
differentiation and experimental validation of these miRNAs would narrow down and expects
Ac ce p
the better results. For example, transfection of these set of miRNAs mimic/inhibitor oligonucleotides to MSCs or osteoblasts would maintain stemcellness, promote osteoblast differentiation or inhibit other lineages differentiation. To investigate the functional role of miR-15b (an osteoblast specific miRNA) on osteoblastic differentiation, rat osteoprogenitor cells were transiently transfected with miR-15b inhibitor or negative control miRNA. We determined the lineage commitment contributed by miR-15b. The MSCs’ major connective tissue lineages are adipose, myoblast, chondrocyte and osteoblast which were analyzed by their specific transcription factor genes like PPARγ, myogenin, FGF2 and Runx2, respectively by real time RT-PCR analysis. The result showed that the inhibition of miR-15b expression by miR-15b inhibitor in rat osteoprogenitor cells enhances mRNA expression of
12 Page 12 of 34
myogenin, PPARγ genes; whereas there was no significant change in Runx2 mRNA expression (Fig. 5A). This result suggested that miR-15b could target genes which may be essential for promoting other cell lineages, and inhibition of miR-15b expression would not
ip t
target those genes, thus promoting other cell lineages. There are evidences clearly indicating a correlation of Runx2 with myogenesis and adipogenesis. It has shown that the increased
cr
Runx2 expression suppresses the myogenesis with the help of Smads [41, 42]. Runx2 has been shown to induce transdifferentiation of primary myoblasts towards osteoblasts by
us
suppressing myogenin expression [43] and miR-133 is involved in lineage determination
an
between myoblasts and osteoblasts from MSCs [13]. PPARγ, a member of nuclear hormone receptor gene superfamily is one of the important transcription factors which up regulates
M
adipogenesis and down regulates osteogenesis by acting in MSCs. Expression of PPARγ is recorded to reduce Runx2 [44, 45], and activation of PPARγ is reduced in osteoblast
d
differentiation and bone mass [46]. Vitamin C deficiency reduces bone mass by promoting
te
transition of osteoblasts into adipocytes through increased expression of PPARγ [47]. It has shown that inhibition of PPARγ expression by miR-20a enhances osteoblast differentiation
Ac ce p
by increasing BMP4, Runx2 and osterix [48]. FGF2, a family of polypeptides, is one of the important factors in regulation of bone
formation and remodeling, involved in the promotion of chondrogenesis and adipogenesis, and potentially contributes to maintenance of MSCs characteristics [49-52]. Additionally, FGF2 is recorded to increase adipogenic lineage differentiation by up regulating PPARγ expression in MSCs and adipose derived stem cells (ASCs) and inhibits osteoblastic lineage differentiation [49, 52]. It has also been shown to increase cell growth and decrease matrix mineralization along with ALP, type I collagen, osteocalcin expression in pre-mature human calvarial osteoblasts [53]. FGF2 was identified as one of the predicted target gene of miR-15b (Fig. 5B). When we used miR-15b inhibitor, there was increased mRNA expression of FGF2 13 Page 13 of 34
but there was no change in Runx2 expression (Fig. 5C). Therefore, miR-15b involves not only in osteoblast differentiation but it could also participate in promotion of differentiation of other lineages.
cr
ip t
Figure 5A Figure 5B Figure 5C Overall, we conclude that there is differential expression of miRNAs during MSCs differentiation towards osteoblasts. The combined in silico analyses of hMSCs or osteoblasts
us
specific miRNAs provide a hint to identify possible miRNA intracellular regulatory network and their gene targets which could regulate osteoblast differentiation. The experimental
an
analysis of miR-15b inhibition by antisense oligonucleotide clearly identified the importance of miR-15b in osteoblast differentiation through suppressing other lineage commitment
M
specific transcription factors like PPARγ, myogenin and FGF2. Thus, miRNAs are potential
Ac ce p
Acknowledgements
te
their expression.
d
candidates to turn the possible physiological functions of MSCs or osteoblasts by altering
This work was supported by a grant from the Indian Council of Medical Research (ICMR), India to N. S. (Grant No: 80/10/2010-BMS). References
[1] N. Selvamurugan, S.C. Jefcoat, S. Kwok, R. Kowalewski, J.A. Tamasi, N.C. Partridge, Overexpression of Runx2 directed by the matrix metalloproteinase-13 promoter containing the AP-1 and Runx/RD/Cbfa sites alters bone remodeling in vivo, J Cell Biochem 99 (2006) 545-557. [2] N. Selvamurugan, S. Kwok, A. Vasilov, S.C. Jefcoat, N.C. Partridge, Effects of BMP-2 and pulsed electromagnetic field (PEMF) on rat primary osteoblastic cell proliferation and gene expression. J Orthop Res 25 (2007) 1213-1220. [3] J.B. Lian, G.S. Stein, A.J. van Wijnen, J.L. Stein, M.Q. Hassan, T. Gaur, Y. Zhang, MicroRNA control of bone formation and homeostasis. Nat Rev Endocrinol 8 (2012) 212227.
14 Page 14 of 34
[4] S. Vimalraj, N. Selvamurugan, MicroRNAs: synthesis, gene regulation and osteoblast differentiation. Curr Issues Mol Biol 15 (2012) 7-18. [5] A. Moorthi, S. Vimalraj, C. Avani, Z. He, N.C. Partridge, N. Selvamurugan, Expression of microRNA-30c and its target genes in human osteoblastic cells by nano-bioglass ceramictreatment, Int J Biol Macromol 56 (2013) 181-185.
ip t
[6] M. Selbach, B. Schwanhausser, N. Thierfelder, Z. Fang, R. Khanin, N. Rajewsky, Widespread changes in protein synthesis induced by microRNAs, Nature 455 (2008) 58-63.
cr
[7] N. Bushati, S.M. Cohen, MicroRNA functions, Annu Rev Cell Dev Biol 23 (2007) 175205.
us
[8] V.K. Gangaraju, J. Lin, MicroRNAs: key regulators of stem cells, Nat Rev Mol Cell Biol 10 (2009) 116-125. [9] S. Vimalraj, P.J. Miranda, B. Ramyakrishna, N. Selvamurugan, Regulation of breast cancer and bone metastasis by microRNAs, Dis Markers 35 (2013) 369-387.
M
an
[10] L.A. Goff, S. Boucher, C.L. Ricupero, S. Fenstermacher, M. Swerdel, L.G. Chase, C.C. Adams, J. Chesnut, U. Lakshmipathy, R.P. Hart, Differentiating human multipotent mesenchymal stromal cells regulate microRNAs: prediction of microRNA regulation by PDGF during ssteogenesis, Exp Hematol 36 (2008) 1354-1369. [11] L. Guo, R.C. Zhao, Y. Wu, The role of microRNAs in self-renewal and differentiation of mesenchymal stem cells, Exp Hematol 39 (2011) 608-616.
te
d
[12] J. Gao, T. Yang, J. Han, K. Yan, X. Qiu, Y. Zhou, Q. Fan, B. Ma, microRNA expression during osteogenic differentiation of human multipotent mesenchymal stromal Cells from Bone Marrow, J Cell Biochem 112 (2011) 1844-1856.
Ac ce p
[13] Z. Li, M.Q. Hassan, S. Volinia, A.J. van Wijnen, J.I. Stein, C.M. Croce, J.B. Lian, G.S. Stein, A microRNA signature for a BMP2-induced osteoblast lineage commitment program, Proc Natl Acad Sci USA 105 (2008) 13906-13911. [14] M.F. Pittenger, A.M. Mackay, S.C. Beck, R.K. Jaiswal, R. Douglas, J.D. Mosca, M.A. Moorman, D.W. Simonetti, S. Craig, D.R. Marshak, Multilineage potential of adult human mesenchymal stem cells, Science 284 (1999) 143-147. [15] J.R. Nefussi, M.L. Boy-Lefevre, H. Boulekbache, N. Forest, Mineralization in vitro of matrix formed by osteoblasts isolated by collagenase digestion, Differentiation 29 (1985) 160. [16] T. Eskildsen, H. Taipaleenmäki, J. Stenvang, B.M. Abdallah, N. Ditzel, A.Y. Nossent, M. Bak, S. Kauppinen, M. Kassem, MicroRNA-138 regulates osteogenic differentiation of human stromal (mesenchymal) stem cells in vivo, Proc Natl Acad Sci USA 108 (2011) 61396144. [17] S.M. Hartig, Basic image analysis and manipulation in ImageJ, Curr Protoc Mol Biol (2013) doi: 10.1002/0471142727.mb1415s102. [18] G.L. Papadopoulos, P. Alexiou, M. Maragkakis, M. Reczko, A.G. Hatzigeorgiou, DIANA-mirPath: integrating human and mouse microRNAs in pathways, Bioinformatics 25 (2009) 1991-1993. 15 Page 15 of 34
[19] V. Kaimal, E.E. Bardes, S.C. Tabar, A.G. Jegga, B.J. Aronow, ToppCluster: a multiple gene list feature analyzer for comparative enrichment clustering and network-based dissection of biological systems, Nucleic Acids Res 38 (2010) W96–W102 (Web Server issue).
ip t
[20] S. Kwok, N.C. Partridge, N. Srinivasan, S.V. Nair, N. Selvamurugan, Mitogen activated protein kinase-dependent inhibition of osteocalcin gene expression by transforming growth factor-beta1, J Cell Biochem 106 (2009) 161-169.
cr
[21] D.D. Liu, J.C. Zhang, Q. Zhang, S.X. Wang, M.S. Yang, TGF-β/BMP signaling pathway is involved in cerium-promoted osteogenic differentiation of mesenchymal stem cells, J Cell Biochem 114 (2013) 1105-1114.
us
[22] G. Chen, C. Deng, Y.P. Li, TGF-β and BMP signaling in osteoblast differentiation and bone formation, Int J Biol Sci 8 (2012) 272-288.
an
[23] H. Inose, H. Ochi, A. Kimura, K. Fujita, R. Xu, S. Sato, M. Iwasaki, S. Sunamura, Y. Takeuchi, S. Fukumoto, K. Saito, T. Nakamura, H. Siomi, H. Ito, Y. Arai, K. Shinomiya, S. Takeda, A microRNA regulatory mechanism of osteoblast differentiation, Proc Natl Acad Sci U S A 106 (2009) 20794-9.
M
[24] O. Trohatou, D. Zagoura, V. Bitsika, K.I. Pappa, A. Antsaklis, N.P. Anagnou, M.G. Roubelakis, Sox2 Suppression by miR-21 Governs Human Mesenchymal Stem Cell Properties, Stem Cells Transl Med (2013) doi: 10.5966/sctm.2013-0081.
te
d
[25] N. Yang, G. Wang, C. Hu, Y. Shi, L. Liao, S. Shi, Y. Cai, S. Cheng, X. Wang, Y. Liu, L. Tang, Y. Ding, Y. Jin, Tumor necrosis factor α suppresses the mesenchymal stem cell osteogenesis promoter miR-21 in estrogen deficiency-induced osteoporosis, J Bone Miner Res 28 (2013) 559-73.
Ac ce p
[26] C.J. Guo, Q. Pan, D.G. Li, H. Sun, B.W. Liu, miR-15b and miR-16 are implicated in activation of the rat hepatic stellate cell: An essential role for apoptosis, J Hepatol 50 (2009) 766-78. [27] H. Nishi, K. Ono, Y. Iwanaga, T. Horie, K. Nagao, G. Takemura, M. Kinoshita, Y. Kuwabara, R.T. Mori, K. Hasegawa, T. Kita, T. Kimura, MicroRNA-15b modulates cellular ATP levels and degenerates mitochondria via Arl2 in neonatal rat cardiac myocytes, J Biol Chem 285 (2010) 4920-30. [28] S. Mujahid, H.C. Nielsen, M.V. Volpe, MiR-221 and miR-130a regulate lung airway and vascular development, PLoS One 8 (2013) e55911. [29] P. Mestdagh, S. Lefever, F. Pattyn, D. Ridzon, E. Fredlund, A. Fieuw, M. Ongenaert, J. Vermeulen, A. De Paepe, L. Wong, F. Speleman, C. Chen, J. Vandesompele, The microRNA body map: dissecting microRNA function through integrative genomics, Nucleic Acids Res 39 (2011) e136. [30] A. Chamorro-Jorganes, E. Araldi, L.O. Penalva, D. Sandhu, C. Fernández-Hernando, Y. Suárez, MicroRNA-16 and microRNA-424 regulate cell-autonomous angiogenic functions in endothelial cells via targeting vascular endothelial growth factor receptor-2 and fibroblast growth factor receptor-1, Arterioscler Thromb Vasc Biol 31 (2011) 2595-606.
16 Page 16 of 34
[31] J. Xu, Y. Li, F. Wang, X. Wang, B. Cheng, F. Ye, X. Xie, C. Zhou, W. Lu, Suppressed miR-424 expression via upregulation of target gene Chk1 contributes to the progression of cervical cancer, Oncogene (2012) doi: 10.1038/onc.2012.121. [32] I. Faraoni, S. Laterza, D. Ardiri, C. Ciardi, F. Fazi, F. Lo-Coco, MiR-424 and miR-155 deregulated expression in cytogenetically normal acute myeloid leukaemia: correlation with NPM1 and FLT3 mutation status, J Hematol Oncol 8 (2012) 5-26.
ip t
[33] Y. Jiang, Y. Wu, A.R. Greenlee, J. Wu, Z. Han, X. Li, Y. Zhao, miR-106a-mediated malignant transformation of cells induced by anti-benzo[a]pyrene-trans-7,8-diol-9,10epoxide, Toxicol Sci 119 (2011) 50-60.
us
cr
[34] R. Hummel, D.I. Watson, C. Smith, J. Kist, M.Z. Michael, J. Haier, D.J. Hussey, Mir148a improves response to chemotherapy in sensitive and resistant oesophageal adenocarcinoma and squamous cell carcinoma cells, J Gastrointest Surg 15 (2011) 429-38.
an
[35] Z. Wang, M. Liu, H. Zhu, W. Zhang, S. He, C. Hu, L. Quan, J. Bai, N. Xu, miR-106a is frequently upregulated in gastric cancer and inhibits the extrinsic apoptotic pathway by targeting FAS, Mol Carcinog (2012) doi: 10.1002/mc.21899.
M
[36] M. Satoh, Y. Minami, Y. Takahashi, T. Tabuchi, M. Nakamura, A cellular microRNA, let-7i, is a novel biomarker for clinical outcome in patients with dilated cardiomyopathy, J Card Fail 17 (2011) 923-929.
d
[37] H. Wang, W. Peng, X. Ouyang, Y. Dai, Reduced circulating miR-15b is correlated with phosphate metabolism in patients with end-stage renal disease on maintenance hemodialysis, Ren Fail 34 (2012) 685-690.
te
[38] A.M. Liu, T.J. Yao, W. Wang, K.F. Wong, N.P. Lee, S.T. Fan, R.T. Poon, C. Gao, J.M. Luk, Circulating miR-15b and miR-130b in serum as potential markers for detecting hepatocellular carcinoma: a retrospective cohort study, BMJ Open 8 (2012) 2:e000825.
Ac ce p
[39] M. Tanic, K. Yanowsky, C. Rodriguez-Antona, R. Andrés, I. Márquez-Rodas, A. Osorio, J. Benitez, B. Martinez-Delgado, Deregulated miRNAs in hereditary breast cancer revealed a role for miR-30c in regulating KRAS oncogene, PLoS One 7 (2012) e38847. [40] L. Yang, N. Li, H. Wang, X. Jia, H. Wang, J. Luo, Altered microRNA expression in cisplatin-resistant ovarian cancer cells and upregulation of miR-130a associated with MDR1/P-glycoprotein-mediated drug resistance, Oncol Rep 28 (2012) 592-600. [41] L.J. Raggatt, L. Qin, J. Tamasi, S.C. Jefcoat Jr, E. Shimizu, N. Selvamurugan, F.Y. Liew, L. Bevelock, J.H. Feyen, N.C. Partridge, Interleukin-18 is regulated by parathyroid hormone and is required for its bone anabolic actions, J Biol Chem 283 (2008) 6790-8. [42] K.S. Lee, H.J. Kim, Q.L. Li, X.Z. Chi, C. Ueta, T. Komori, J.M. Wozney, E.G. Kim, J.Y. Choi, H.M. Ryoo, S.C. Bae, Runx2 is a common target of transforming growth factor beta1 and bone morphogenetic protein 2, and cooperation between Runx2 and Smad5 induces osteoblast-specific gene expression in the pluripotent mesenchymal precursor cell line, 2C12. Mol Cell Biol 20 (2000) 8783-92. [43] C.A. Gersbach, B.A. Byers, G.K. Pavlath, A.J. García, Runx2/Cbfa1 stimulates transdifferentiation of primary skeletal myoblasts into a mineralizing osteoblastic phenotype, Exp Cell Res 300 (2004) 406-17. 17 Page 17 of 34
[44] M.J. Jeon, J.A. Kim, S.H. Kwon, S.W. Kim, K.S. Park, S.W. Park, S.Y. Kim, C.S. Shin, Activation of peroxisome proliferatoractivated receptor-gamma inhibits the Runx2-mediated transcription of osteocalcin in osteoblasts, J Biol Chem 278 (2003) 23270-7. [45] J.B. Lian, G.S. Stein, A. Javed, A.J. van Wijnen, J.L. Stein, M. Montecino, M.Q. Hassan, T. Gaur, C.J. Lengner, D.W. Young, Networks and hubs for the transcriptional control of osteoblastogenesis, Rev Endocr Metab Disord 7 (2006) 1-16.
ip t
[46] A.A. Ali, R.S. Weinstein, S.A. Stewart, A.M. Parfitt, S.C. Manolagas, R.L. Jilka, Rosiglitazone causes bone loss in mice by suppressing osteoblast differentiation and bone formation, Endocrinol 146 (2005) 1226-35.
us
cr
[47] J.K. Park, E.M. Lee, A.Y. Kim, E.J. Lee, C.W. Min, K.K. Kang, M.M. Lee, K.S. Jeong, Vitamin C deficiency accelerates bone loss inducing an increase in PPAR-γ expression in SMP30 knockout mice, Int J Exp Pathol 93(2012):332-40.
an
[48] J.F. Zhang, W.M. Fu, M.L. He, W.D. Xie, Q. Lv, G. Wan, G. Li, H. Wang, G. Lu, X. Hu, S. Jiang, J.N. Li, M.C. Lin, Y.O. Zhang, H.F. Kung, MiRNA-20a promotes osteogenic differentiation of human mesenchymal stem cells by co-regulating BMP signaling, RNA Biol 8 (2011) 829-38.
M
[49] M. Neubauer, C. Fischbach, P. Bauer-Kreisel, E. Lieb, M. Hacker, J. Tessmar, M.B. Schulz, A. Goepferich, T. Blunk, Basic fibroblast growth factor enhances PPARgamma ligand-induced adipogenesis of mesenchymal stem cells, FEBS Lett 577 (2004) 277-283.
d
[50] A.M. Handorf, W.J. Li, Fibroblast growth factor-2 primes human mesenchymal stem cells for enhanced chondrogenesis, PLoS One 6 (2011) e22887.
Ac ce p
te
[51] W.T. Lai, V. Krishnappa, D.G. Phinney, Fibroblast growth factor 2 (Fgf2) inhibits differentiation of mesenchymal stem cells by inducing Twist2 and Spry4, blocking extracellular regulated kinase activation, and altering FGF receptor expression levels, Stem Cells 29 (2011) 1102-11. [52] N. Kakudo, A. Shimotsuma, K. Kusumoto K. Fibroblast growth factor-2 stimulates adipogenic differentiation of human adipose-derived stem cells. Biochem Biophys Res Commun 359 (2007) 239-44. [53] P.J. Marie, F. Debiais, E. Haÿ, Regulation of human cranial osteoblast phenotype by FGF-2, FGFR-2 and BMP-2 signaling, Histol Histopathol 17 (2002) 877-85.
18 Page 18 of 34
Figure legends Figure 1. Human MSCs differentiation towards osteoblasts. hMSCs were induced into osteoblastic differentiation by providing osteogenic medium. The differentiation of osteoblast was confirmed at molecular level by semi quantitative RT-PCR analysis of osteoblast marker genes (ALP, osteonectin, and type 1 collagen) and transcription factor gene, Osterix. The internal controls used in this study are GAPDH and U6 genes.
M
an
us
cr
ip t
Figure 2. Expression of miRNAs during osteoblast differentiation. (A) Expression of MSCs and osteoblasts specific precursor miRNAs analyzed in undifferentiated hMSCs and 14 days osteo-differentiated hMSCs by semi quantitative RT-PCR analysis. * indicates significant up regulation compared to 14 days osteo-differentiated cells and ** indicates significant up regulation compared to undifferentiated hMSC. (B) The osteoblast specific miRNAs (mir15b, mir-24, mir-130b, mir-30c, and mir-130a) expression was analyzed in 7-days osteodifferentiated hMSCs, 14-days osteo-differentiated hMSCs, human differentiated osteoblasts (MG63), rat calvarial cells and rat osteoblasts by semi quantitative RT-PCR analysis. (C) Real time RT-PCR analysis of mir-15b, miR-130a and miR-21 expression in undifferentiated hMSCs and 14 days osteo differentiated hMSCs. *represents significant up regulation compared to hMSCs. (D) Heat map of differential expression of miRNAs in normal human tissues. Differential expression of MSCs and osteoblast specific miRNAs was identified in various normal human tissues by miRNA body map along with the hierarchical cluster analysis. Expression of miRNAs is represented as blue (down regulated), red (up regulated), and white colors (no significant change or absence of data).
Ac ce p
te
d
Figure 3. Putative targets of hMSCs specific miRNAs (miR-424, miR-106a, miR-148a, let-7i and miR-99a) in major signaling pathways involved in osteoblast differentiation. Kyoto Encyclopedia of Genes and Genomes (KEGG) was used to identify the presence of putative targets. Figures were generated from the DIANA-MicroT4.0 (beta version) databases. (A) A combinational effect of MSCs specific miRNAs in various pathways is visible in bar diagram of the -ln P values. (B) The cytoscape graph networks were generated from ToppCluster network analyser. The functional enrichment analysis of regulatory network influenced by MSCs specific miRNAs is shown. Figure 4. Putative targets of osteoblasts specific miRNAs (miR-15b, miR-24, miR-130b, miR-30c, and miR-130a) in major signaling pathways involved in osteoblast differentiation. Kyoto Encyclopedia of Genes and Genomes (KEGG) was used to identify the presence of putative targets. Figures were generated from the DIANA-MicroT4.0 (beta version) databases. (A) A combinational effect of osteoblast specific miRNAs in various pathways is visible in bar diagram of the -ln P values. (B) The cytoscape graph networks were generated from ToppCluster network analyser. The functional enrichment analysis of regulatory network influenced by osteoblasts specific miRNAs is shown. Figure 5. miR-15b inhibitor up regulates expression of PPARγ, myogenin and FGF2. (A) Rat osteoprogenitor cells were transiently transfected with 50 nM of negative control miRNA or miR-15b inhibitor for 7 days. Total RNA was isolated and real time RT-PCR was carried out using the primers for PPARγ, myogenin and Runx2 genes. The fold change of expression of 19 Page 19 of 34
Ac ce p
te
d
M
an
us
cr
ip t
mRNAs was calculated over negative control miRNA. GAPDH gene was used for normalization. (B) The putative target region analysis was performed for FGF2 mRNAs 3’ UTR by miR-15b seed sequence. Interspecies similarity of putative miR-15b binding region with the 3’UTR of FGF2 was shown. (C) hMSCs were transfected with miR-15b inhibitor in normal medium up to 7 days and the real time RT-PCR was performed for FGF2 and Runx2 genes. GAPDH gene was used for normalization. * indicates significant up regulation of mRNA compared to negative control miRNA.
20 Page 20 of 34
ip t
Table(s)
cr
Table 1. Primer sequences. Name of the miRNA/genes
mir-148a let-7i mir-99a
M
mir-15b mir-24
mir-30c
ep te
mir-130a
d
mir-130b
Human Runx2 Rat Runx2
Type 1 collagen
Ac c
Osteonectin Osterix ALP PPARγ Myogenin FGF2
Human GAPDH Rat GAPDH U6
CGAGGGGATACAGCAGCAATTC CCCCACCTTCTACCTTCCCCAC CCTTGGCCATGTAAAAGTGC CCATGGTAATGTAAGAAGTGC GAGGCAAAGTTCTGAGACAC GAGACAAAGTTCTGTAGTGCACT AGTAGTTTGTGCTGTTGGTCGGGT GCAAGGCAGTAGCTTGCGCAGT CCATTGGCATAAACCCGTAGATCCG CAGACCCATAGAAGCGAGCTTGTGC GGCCTTAAAGTACTGTAGCAGC CCTTAAATTTCTAGAGCAGC CTCCGGTGCCTACTGAGCTG CTCCTGTTCCTGCTGAACTG GGCCTGCCCGACACTCTTTC GACCTGACCGATGCCCTTTC TGTGTAAACATCCTACACTCTCAG GAGTAAACAACCCTCTCCCA GCTGGCCAGAGCTCTTTTCACA CACTACACGGCCAATGCCCTTT CAGTTCCCAAGCATTTCATC TCAATATGGTCGCCAAACAG CAAGTGGCCAGGTTCAACGA GTGAAGACCGTTATGGTCAAAGTG TAACCCCCTCCCCAGCCACAAA TTCCTCTTGGCCGTGCGTCA GCAGTGGTGGGTCCGTGGTC CTGTCACGGCTCGGGTGTGC ACTGGCTAGGTGGTGGTCAG GGTAGGGAGCTGGGTTAAGG AGGCAGGATTGACCACGG TGTAGTTCTGCTCATGGA CCTGAAGCTCCAAGAATACCAAA AGAGTTGGGTTTTTTCAGAATAATAAGG TGAAGAGAAGCACCCTGCTCA TCAATGTACTGGATGGCACTG C CAGATTAGCGGACGCGGT CGGATGGGTGTCTCCGC GAGAGACCCCACTTGCTGCCA CTCACACTGCCCCTCCCTGGT GCAGAGGTTGAATGTGAGCA GGAAGAAGTTCCCATCGTCA CTCGCTTCGGCAGCACA AACGCTTCACGAATTTGCGT
us
mir-106a
Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse
an
mir-424
Sequence (5'->3')
Page 21 of 34
ip t
hMSCs specific miRNAs
-ln (p-value)
hsa-let-7i
hsa-miR-99a
Union
No. of genes No. of genes No. of genes No. of genes No. of genes -ln (p-value) -ln (p-value) -ln (p-value) -ln (p-value) -ln (p-value) targeted targeted targeted targeted targeted
28
12.34
28
9.06
TGF-beta pathway (hsa04350)
8
0.74
13
7.35
Wnt pathway (hsa04310)
19
6.38
18
12
1.17
25
12.12
2
1.23
75
18.19
11
10.78
10
5.86
-
-
34
15.7
11
4.41
10
1.83
3
7.48
47
14.69
M
MAPK pathway (hsa04510)
7.35
d ep te
Osteoblast specific miRNAs hsa-miR-15b Name of the pathway and KEGG pathway id No. of genes targeted
12
-ln (p-value)
3.22
Ac c
Ubiquitin mediated proteolysis (hsa04120)
hsa-miR-148a
us
targeted
hsa-miR-106a
an
hsa-miR-424 Name of the pathway and KEGG pathway id No. of genes
cr
Table 2. KEGG Pathways affected by putative targets of miRNAs.
hsa-miR-24
hsa-miR-130b
hsa-miR-30c
hsa-miR-130a
Union
No. of genes No. of genes No. of genes No. of genes No. of genes -ln (p-value) -ln (p-value) -ln (p-value) -ln (p-value) -ln (p-value) targeted targeted targeted targeted targeted
5
0.27
13
5.02
22
14.88
12
4.19
40
13.32
Gap junction (hsa04540)
7
0.92
4
0.35
8
2.03
8
0.81
8
2.25
21
2.36
Jak-STAT signalling pathway (hsa04630)
10
0.77
4
0.02
5
0.62
13
1.38
5
0.51
29
1.39
p53 signalling pathway (hsa04115)
9
5.33
4
1.15
4
1.13
2
0.93
4
0.19
16
2.24
Page 22 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 23 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 24 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 25 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s) 2c
Page 26 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s) 2D
Page 27 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 28 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 29 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 30 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 31 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s)
Page 32 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s) 5B
Page 33 of 34
Ac
ce
pt
ed
M
an
us
cr
i
Figure(s) 5C
Page 34 of 34