Journal of the Formosan Medical Association (2013) 112, 713e720

Available online at www.sciencedirect.com

journal homepage: www.jfma-onl ine.com

ORIGINAL ARTICLE

Microarray analysis of gene expression ofbone marrow stem cells cocultured withsalivary acinar cells

Chia-Yung Lin a,b, Jhih-Ren Yang a, Sheng-Lin Teng b,Stephanie Tsai b, Min-Huey Chen b,*

aDentistry Department, National Taiwan University Hospital Hsinchu Branch, Hsinchu, Taiwan, ROCbGraduate Institute of Clinical Dentistry, School of Dentistry, National Taiwan University, Taipei,Taiwan, ROC

Received 15 June 2012; received in revised form 2 August 2012; accepted 9 August 2012

KEYWORDSbone marrow stemcells;

salivary acinar cells;bone morphogeneticprotein-6;

transdifferentiation;microarray analysis

* Corresponding author. Graduate InE-mail address: minhueychen@ntu

0929-6646/$ - see front matter Copyrhttp://dx.doi.org/10.1016/j.jfma.201

Background/Purpose: Our previous work has demonstrated that rat bone marrow stem cells(BMSCs) can transdifferentiate into a-amylase-producing cells after coculture with rat subman-dibular gland acinar cells. These transdifferentiated cells may be used for regeneration ofdamaged salivary gland. The purpose of this study was to investigate the global gene expres-sion of rat BMSCs cocultured with rat submandibular gland acinar cells and the factors inducingthis transdifferentiation.Methods: Rat BMSCs were indirectly cocultured with rat submandibular gland acinar cells byusing the double chamber system for 5 and 10 days. The global gene expression of BMSCs duringtransdifferentiation into acinar cells was investigated by microarray analysis.Results: A total of 45,018 probes were used and 41,012 genes were detected. After coculturefor 5 days, 1409 genes were upregulated more than twofold and 1417 genes were downregu-lated more than twofold (p< 0.005). Moreover, after coculture for 10 days, 1356 genes wereupregulated more than twofold and 1231 genes were downregulated more than twofold(p< 0.005). Bone morphogenetic protein (BMP)-6 was one of the top-ranked upregulatedgenes. The hub genes were interleukin-6 and CCAAT/enhancer-binding protein b (CEBPB) inthe early and late response gene groups, respectively.Conclusion: This is believed to be the first study on the global gene expression of rat BMSCscocultured with rat acinar cells. Many genes related to the function of salivary acinar cells suchas those responsible for the production of a-amylase protein were upregulated and many genesrelated to the differentiation of BMSCs into adipocytes and osteoblasts were downregulated. Inaddition, BMP-6 gene was found to be highly upregulated. We proposed that three target

stitute of Clinical Dentistry, School of Dentistry, National Taiwan University, Taipei, Taiwan, ROC..edu.tw (M.-H. Chen).

ight ª 2012, Elsevier Taiwan LLC & Formosan Medical Association. All rights reserved.2.08.006

714 C.-Y. Lin et al.

genes, BMP-6, interleukin-6 and CEBPB, play important roles in the transdifferentiation ofBMSCs into acinar cells, and are worthy of further investigation.Copyright ª 2012, Elsevier Taiwan LLC & Formosan Medical Association. All rights reserved.

Introduction

Adult stem cells may have great therapeutic potential intissue repair and regeneration.1,2 Transdifferentiationoccurs when a committed cell type that already developsalong a certain differentiation pathway changes into a celltype of another cell lineage.3,4 Bone marrow stem cells(BMSCs) can be transdifferentiated into hepatocytes andpancreatic islet cells.5,6 In our previous work, we demon-strated the transdifferentiation of BMSCs into salivaryacinar cells (ACs) and these transdifferentiated ACs do havetherapeutic effects for damaged salivary glands.7,8 Thesignaling molecules and pathway leading to this trans-differentiation are completely unknown, and if and to whatextent established crucial factors are involved is also notyet known.

Transdifferentiation of rat BMSCs into salivary ACs canbe achieved by an in vitro double chamber culture system,7

therefore, rat BMSCs on the lower part of the chamber areeasily collected during the coculture period. The molecularmechanisms provoking this reprogramming are unknown,thus, we analyzed genome-wide gene expression patternsby using an Agilent Gene-Chip (Agilent Technologies, SantaClara, CA, USA) representing 41,012 genes. Thereby, in thisstudy, gene expression of rat BMSCs after 5 or 10 dayscocultured with salivary acinar cells were collected andcompared to those original stem cells without coculture.

Materials and methods

All animal studies were performed according to a protocolapproved by the Institutional Review Board of NationalTaiwan University Hospital. Rats were purchased from theNational Taiwan University Laboratory Animal Center andkept under clean conventional conditions.

Isolation and culture of salivary ACs

ACs were obtained from 7-day-old Wistar rats using theexplant outgrowth technique.9 The thin fascia layer of thesubmandibular gland was carefully removed to expose thetissue, which was then cut into small fragments about1 mm3 in size. The tissue fragments were then cultured at37�C in a 5% CO2 humidified incubator with M199 culturemedium (BioChrom AG, Berlin, Germany) supplementedwith 20% fetal bovine serum (Biological Industries, BeitHaemek, Israel) and antibiotics/antimycotics (penicillin Gsodium 100 U/mL, streptomycin sulfate 100 mg/mL,amphotericin B 0.25 mg/mL; Gibco, Grand Island, NY, USA).Cells released from the tissues grew to confluence. Atapproximately 90% confluence, tissue fragments wereremoved and subcultured on two 100-mm culture dishes(Corning Inc., Corning, NY, USA). Under these conditions,

cell division and growth continued to yield enough cells andthe second passage of cells was used.

Isolation and culture of BMSCs

BMSCs were harvested from the tibia of 4-week-old Wistarrats. Bone marrow was collected in a syringe with a 21-gauge needle and mixed (1:9, v/v) with a buffer containing0.8% ammonium chloride and 10 mM EDTA (Sigma, St. Louis,MO, USA) for 10 minutes to lyse the red blood cells. After 24hours, the spindle-shaped colonies in the adherent cellswere identified as mesenchymal stem cells. There were stillsome suspended small stem cells,10 therefore, the non-adherent cells in the supernatant were also collected bycentrifugation at 500 g for 5 minutes and plated on 100-mmculture dishes containing Dulbecco’s Modified Eagle’sMedium (BioChrom AG, Berlin, Germany) supplementedwith 20% fetal bovine serum and antibiotics/antimycoticsand cultured at 37�C in a 5% CO2 humidified incubator. Thesmall stem cells adhered to the culture dish after 24e48hours. Both spindle-shaped cells and small stem cells werecollected together as BMSCs.

Identification of ACs and BMSCs

For identification, flow cytometry analysis was performed.ACs were prepared as aliquots containing 5� 105 cells andnonspecific binding sites were blocked with 0.1% Triton X-100 (Sigma) and bovine serum albumin (Sigma) for 30minutes, and then incubated with the primary rabbit antihuman anti-a-amylase antibody for 1 hour (SigmaeAldrich,St. Louis, MO, USA). After washing with phosphate-bufferedsaline and incubation with the secondary sheep anti rabbitIgG antibody conjugated with fluorescein isothiocyanate(FITC; Chemicon International, Temecula, CA, USA) foranother 10 minutes, cells were then fixed in 4% para-formaldehyde and ready for flow cytometry analysis. BMSCswere stained for CD45, CD54 and CD90 surface markers.The adherent cells were detached by trypsinization,centrifuged for 5 minutes at 500 g and resuspended inphosphate-buffered saline. Aliquots containing 5� 105 cellswere incubated with primary antibodies conjugated withFITC (anti-CD45, anti-CD54, and anti-CD90; AbD-Serotec,Oxford, UK) for 15 minutes at 4�C. Cells were fixed in 4%paraformaldehyde and were ready for flow cytometryanalysis. The amount of a-amylase on ACs and each surfacemarker on BMSCs were quantified using the FACSCaliburflow cytometer (Becton Dickinson, Mountain View, CA,USA). As negative controls, cells were incubated only withthe secondary antibody conjugated with FITC and withoutprimary antibody. Fixed ACs and BMSCs were analyzed usingthe FACSCalibur cytometer (Becton Dickinson). For eachsample, 10,000 events were acquired and analyzed using

Microarray analysis of bone marrow stem cells cocultured with acinar cells 715

Cellquest Software (Becton Dickinson). Compensation wasachieved by application of quantitative microbeads (Cali-BRITE; Becton Dickinson) using FACSCopm software. Eventswere acquired using identical conditions and gated toexclude cellular debris from the analysis. Cell surfacemakers expressed on BMSCs and a-amylase expressed onACs were determined by comparison with isotype controlson a histogram plot.

Indirect coculture system

BMSCs and ACs were cocultured indirectly with the doublechamber system as described previously.7 ACs were seededon the upper inserts and BMSCs were loaded at the lowerwells for culture. The ratio of cell density of BMSCs to ACswas about 1:4. BMSCs were treated with 1% EDTA (Sigma)and resuspended with M199 culture medium at a density of2� 104 cells/well in a 12-well plate (BD Falcon; BD Biosci-ences, Franklin Lakes, NJ, USA). Culture inserts (Costar;Corning), each containing 8� 104 ACs, were placed in eightof the 12 wells; one insert for each well. The other fourwells contained only BMSCs. In eight coculture wells, the 5-day coculture period proceeded in four wells and 10-daycoculture in the other four wells.

Total RNA isolation

Total RNA was extracted from each sample (nZ 4 from 5-day coculture system, nZ 4 from 10-day coculturesystem, and nZ 4 from BMSCs without coculture) with Trizol(Invitrogen, Carlsbad, CA, USA) and purified using RNAeasyColumns (Qiagen GmbH, Hilden, Germany,). RNA wasquantitated by NanoDrop ND-1000 (Thermo Scientific,

Figure 1 Bone marrow stem cells were negative for CD45 (A) andexpression was 70%, and CD90þ expression was 97%. Acinar cells w

Wilmington, DE, USA) and the quantity and integrity wasestablished by resolving on 0.8% formaldehyde agarose gels.

Expression profiling of genes by microarrayexperiments

Microarray experiments were performed by using AgilentRat Genome 4� 44k oligonucleotide arrays. These arrayscovered the entire genome. For labeling reaction, 1 mg RNAfrom original BMSCs, 5-day cocultured BMSCs, and 10-daycocultured BMSCs was used (nZ 4 from original BMSCs,nZ 4 from 5-day cocultured BMSCs, and nZ 4 from 10-daycocultured BMSCs). Labeling was done using the Quick Amplabeling kit (Agilent Technologies) according to the manu-facturer’s protocol. Briefly, using T7 promoter elementcoupled with oligodT primer, cDNA was generated andlabeled cRNA was generated via an in vitro transcriptionreaction using T7 RNA polymerase and cyanine (Cy)3 (fororiginal BMSCs) or Cy5 (original 5-day cocultured, 10-daycocultured BMSCs) CTP. Labeled cRNA (0.825 mg) of therespective sample was used for hybridization in thefollowing combinations: original BMSCs (Cy3) and 5-daycocultured (Cy5), original BMSCs (Cy3) and 10-day cocul-tured (Cy5). Both 5-day and 10-day cocultured BMSCs werelabeled with Cy5 for comparing with original BMSCs labeledwith Cy3.

Cy 3 and Cy5 are cyanine dyes and are used typically fortwo-color detection. Cy3 dyes fluoresce yellowegreen(w550 nm excitation, w570 nm emission), whereas Cy5 isfluorescent in the red region (w650/670 nm) but absorbs inthe orange region (w649 nm). Hybridization was performedfor 17 hours, rotating at a speed of 10 rpm at 65�C ina hybridization oven (Agilent Technologies).

positive for CD54 (B) and CD90 (C). The percentage of CD54þ

ere consistently positive for a-amylase (D).

Figure 2 The intensity plot and MA plot demonstrate the consistency and correction of these array experiments. The real imagesof these two array chips are also shown.

716 C.-Y. Lin et al.

Microarray image and data analysis

Microarray image analysis was done using Feature extrac-tion version 10.5.1.1 (Agilent Technologies) and data anal-ysis was done using Gene Spring version 10 (AgilentTechnologies). The background corrected intensity valueswere used for analysis. Normalization was done usingLOWESS algorithm. Similarly expressed genes were filteredon the basis of standard deviation between two biologicalreplicates with the cutoff < 1. Fold changes were calcu-lated and genes with more than a twofold difference were

Table 1 Noteworthy genes in bone marrow stem cellscocultured with salivary acinar cells were listed. Top rankedgenes indicated the most changed genes in the up-/down-regulated groups.

Significant upregulated genes Significant downregulated genes

Gene name Gene name

CCL20

CCL6

CxCL1

Expi

GxCL2

BMP6

BMP3

Kcns2

Kcnmb1

ENSRNOT00000040034

Kcnk2

Kcnmb1

Kcns3

AQP9

SLC12A2

Fgl2

Smpd3

Filip1

RT1-Da

Hnmt

Ctrc

Has2

Mepe

AW141646

Bst2

LOC289485

BMP2

Fabp4

Fabp7

top ranked

selected. The IPA software (Ingenuity System) was used toexplore the network between these changed genes.

Results

Identification of cells

Fig. 1 shows phenotypic characterization of the culture-expanded cells by flow cytometric analysis of the surfaceantigens of BMSCs and functional protein expression of ACs.Cells from bone marrow were consistently positive forintercellular adhesion molecule 1 (CD54), Thy-1 (CD90) andnegative for leukocyte common antigen (CD45). The cellsfrom the submandibular glands were consistently positivefor a-amylase. The percentage of CD54þ expression was

Figure 3 The early response, late response, and consistentresponse genes to the cocultured with ACs of BMSCs weredefined.

Microarray analysis of bone marrow stem cells cocultured with acinar cells 717

70%, and CD90þ expression was 97%. Expression of a-amylase was 100%.

Microarray image and data analysis

A total of 45,018 probes were used and 41,012 genes weredetected. After coculture for 5 days, 1409 genes wereupregulated more than twofold and 1417 genes were down-regulated more than twofold (p< 0.005). Moreover, aftercoculture for 10 days, 1356 genes were upregulated morethan twofold and 1231 genes were downregulated more thantwofold (p< 0.005). Fig. 2 shows the intensity plots, MA plotsand array images of 5-day and 10-day coculture comparisonexperiments. Among all these changed genes, top ranked andinterested genes were selected (Table 1). Bone morphoge-netic protein (BMP)-6 was noted to be one of the top-rankedupregulated genes and demonstrated about 14.57- and12.71-fold changes after 5-day and 10-day coculture. Genesrelated to the function of salivary ACs, such as potassiumchannel subfamily K member 2(Kcnk2), solute carrier family12 (sodium/potassium/chloride transporters) member 2

Figure 4 The eligible network of early response genes analyz

(SLC12A2), and aquaporin-9(AQP9), were also noted asupregulated. Genes related to osteogenesis (MEPE), chon-drogenesis (Has2) or adipogenesis (Fabp4) were down-regulated. early response (changed genes at 5-daycocultured), late response (changed genes at 5-day cocul-tured) and consistent response (changed genes at 5-day/10-day cocultured) (Fig. 3) were analyzed by IPA software(Ingenuity System). It was found that the hub gene IL-6 wasone of the early response genes (Fig. 4). The CCAAT/enhancer-binding protein b (CEBPB) gene was found to beone of the late response genes (Fig. 5). The eligible networkof consistent response genes was also analyzed by IPA soft-ware (Fig. 6).

Discussion

In the present study, we tried to profile the global geneexpression in BMSCs cocultured with salivary ACs for 5 daysand 10 days by using microarray analysis and IPA software foreligible network analysis of early response genes, lateresponse genes and consistent response genes. The overall

ed by IPA software. The hub gene interleukin-6 was noted.

Figure 5 The eligible network of late response genes analyzed by IPA software. The hub gene CCAAT/enhancer-binding proteinb was noted.

718 C.-Y. Lin et al.

results from the gene arrays revealed that the expression ofvarious genes with different functions was upregulated anddownregulated, and further study is needed to confirm thegene expression profile. However, the present study providedsome information that is worthy of further study. Forexample, in the group of upregulated genes, chemokineligand, extracellular peptidase inhibitor, and interleukinaccounted the great majority,11e13 but bone morphogeneticproteins were still especially noted. Besides, genes related tothe function of salivary ACs, such as potassium channel(Kcnk2),14e16 sodium channel (SLC12A2),17 and the waterchannel (AQP9).18 In the downregulated group, genes relatedto mesenchymal-preoerty (Ctrc),19 osteogenesis (MEPE),20,21

chondrogenesis (Has2)22 and adipogenesis (Fabp4)23 werenoted, which suggested that the committed differentiationinto osteoblasts/adipocytes of BMSCs was switched toanother direction on account of coculture with ACs.24

BMP-6 was one of the top-ranked upregulated genes, andits expression was increased 14.57-fold and 12.71-fold after5-day and 10-day coculture. BMP-6 is a member of thetransforming growth factor-b superfamily, and it appears toplay various biological roles in developing tissues. The TGF-b superfamily of 50 structurally related cytokines includesfactors important in cell growth and differentiation.25,26

Apart from TGF-b, members of this multigene familyinclude members of the activin/inhibin subunit gene family,

members of the decapentaplegic-Vg-related gene family,including BMPs. TGF-b family members signal through het-erodimeric complexes of type-I and type-II transmembranousserine/threonine kinase receptors.27 Downstream signalingis transduced by unique proteins of the Smad gene family.28

During organogenesis, TGF-b family members regulate cellgrowth, differentiation, and tissue regeneration.26

BMP-6 is specifically expressed in suprabasal layers ofthe epidermis, and it has been shown to inhibit cell divi-sion, to promote terminal epithelial differentiation.29 Arecent study has demonstrated that BMP-6 expression insalivary glands is uniquely associated with AC differentia-tion and suggests that BMP-6 may play a role in salivarygland function.30 In our results, consistent upregulation ofBMP-6 in BMSCs after coculture with ACs suggests that BMP-6 is related to transdifferentiation of BMSCs. In our study,BMP-6 was highly upregulated 14-fold and 12-fold after 5days and 10 days coculture, respectively. Cytokines wereoriginally identified as genes critically involved in theregulation of osteogenic differentiation, and BMP-6 is nowconsidered to have critical functions in many develop-mental processes.31 Multiple components of the BMPsignaling cascade, including SMADs, HIVEP2, CEBPB, andRUNX, are involved in an array of cellular processes.Although these signaling mediators are considered to haveredundant activities, the influence of BMP activation can

Figure 6 The eligible network of consistent response genes analyzed by IPA software.

Microarray analysis of bone marrow stem cells cocultured with acinar cells 719

have widely distinct outcomes on a particular celldepending on the cellular context.32 Moreover, we notedthat interleukin (IL)-6 was the hub gene in the earlyresponse group (Fig. 4). A previous study has shown that IL-6 is induced by BMP-6 and stimulates cell growth anddifferentiation, and a more recent study has shown that theBMP-6/IL-6 loop promotes cells transdifferentiation.33

Mechanistically, BMP-6 derived from prostate cancer cellsincreases the expression of IL-6 in macrophages, and IL-6induces the neuroendocrine differentiation of prostatecancer cells. IL-6 induced by BMP-6 drives this trans-differentiation process. When this feedback loop is dis-rupted with blockade of BMP-6 or IL-6, transdifferentiationis no longer observed.33

We also noted the expression of the hub gene CEBPB inthe late response group (Fig. 5). A previous study has shownthat CEBPB is critical for macrophage functioning, and hasbeen shown to bind to the IL-6 gene.34,35 A more recentstudy has demonstrated that the conversion from ACs toa ductal phenotype, and from pancreatic to hepatocyte orductal cells, is also dependent upon the expression ofCEBPB.36 Besides, CEBPB is induced by a BMP-6-mediatedpathway to affect cell growth.37 Another recent study hasreported the consistent upregulation of CEBPB with IL-6

during tissue regeneration.38,39 From the present studyusing microarray analysis and a previous study, we suggestthat three target genes, BMP-6, IL-6 and CEBPB, might playimportant roles during the transdifferentiation of BMSCsinto salivary ACs, and are worthy of further investigation.Further study is necessary to explore detailed mechanismsand to investigate interactions between cells and dynamicchanges of gene expression, to understand the possiblepathway of signal transduction.

Acknowledgments

The authors would to thank the core facility from theResearch Center of National Taiwan University Hospital,Hsin-Chu Branch of National Taiwan University Hospital,and funding from National Taiwan University Hospital.

References

1. Wagers AJ, Sherwood RI, Christensen JL, Weissman IL. Littleevidence for developmental plasticity of adult hematopoieticstem cells. Science 2002;297:2256e9.

720 C.-Y. Lin et al.

2. Horwitz EM. Stem cell plasticity: the growing potential ofcellular therapy. Arch Med Res 2003;34:600e6.

3. Park SR, Oreffo RO, Triffitt JT. Interconversion potential ofcloned human marrow adipocytes in vitro. Bone 1999;24:549e54.

4. Schiller PC, D’Ippolito G, Brambilla R, Roos BA, Howard GA.Inhibition of gap-junctional communication induces the trans-differentiation of osteoblasts to an adipocytic phenotypein vitro. J Biol Chem 2001;276:14133e8.

5. Ayatollahi M, Soleimani M, Tabei SZ, Kabir Salmani M. Hep-atogenic differentiation of mesenchymal stem cells induced byinsulin like growth factor-I.World J Stem Cells 2011;3:113e21.

6. Zanini C, Bruno S, Mandili G, Baci D, Cerutti F, Cenacchi G,et al. Differentiation of mesenchymal stem cells derived frompancreatic islets and bone marrow into islet-like cell pheno-type. PLoS One 2011;6:e28175.

7. Lin CY, Lee BS, Liao CC, Cheng WJ, Chang FM, Chen MH.Transdifferentiation of bone marrow stem cells into acinarcells using a double chamber system. J Formos Med Assoc 2007;106:1e7.

8. Lin CY, Chang FH, Chen CY, Huang CY, Hu FC, Huang WK, et al.Cell therapy for salivary gland regeneration. J Dent Res 2011;90:341e6.

9. Chen MH, Chen YJ, Liao CC, Chan YH, Lin CY, Chen RS, et al.Formation of salivary acinar cell spheroids in vitro abovea polyvinyl alcohol-coated surface. J Biomed Mater Res A 2009;90:1066e72.

10. Chen MH, Lin S, Hsieh CH, Lee HS, Chiang H, Jiang CC. Identifi-cation and initial characterization of small cells in adult carti-lage and bone marrow. J Formos Med Assoc 2004;103:264e73.

11. Gibbs RA, Weinstock GM, Metzker ML, Muzny DM, Sodergren EJ,Scherer S, et al. Genome sequence of the Brown Norway ratyields insights into mammalian evolution. Nature 2004;428:493e521.

12. Saha RN, Wissink EM, Bailey ER, Zhao M, Fargo DC, Hwang JY,et al. Rapid activity-induced transcription of Arc and otherIEGs relies on poised RNA polymerase II. Nat Neurosci 2011;14:848e56.

13. Saar K, Beck A, Bihoreau MT, Bimey E, Brocklebank D, Chen Y,et al. SNP and haplotype mapping for genetic analysis in therat. Nat Genet 2008;40:560e6.

14. Salinas M, Duprat F, Heurteaux C, Hugnot JP, Lazdunski M. Newmodulatory alpha subunits for mammalian Shab Kþ channels. JBiol Chem 1997;272:24371e9.

15. Chang H, Ma YG, Wang YY, Song Z, Li Q, Yang N, et al. Highglucose alters apoptosis and proliferation in HEK293 cells byinhibition of cloned BK Ca channel. J Cell Physiol 2011;226:1660e75.

16. Bockenhauer D, Zilberberg N, Goldstein SA. KCNK2: reversibleconversion of a hippocampal potassium leak into a voltage-dependent channel. Nat Neurosci 2001;4:486e91.

17. Wang Z, Bildin VN, Yang H, Capo-Aponte JE, Yang Y,Reinach PS. Dependence of corneal epithelial cell proliferationon modulation of interactions between ERK1/2 and NKCC1.Cell Physiol Biochem 2011;28:703e14.

18. Damiano A, Zotta E, Goldstein J, Reisin I, Ibarra C. Waterchannel proteins AQP3 and AQP9 are present in syncytio-trophoblast of human termplacenta.Placenta 2001;22:776e81.

19. Wang H, Yuan D, Xu R, Chi CW. Purification, cDNA cloning, andrecombinant expression of chymotrypsin C from porcinepancreas. Acta Biochim Biophys Sin 2011;43:568e75.

20. Wei X, Liu L, Zhou X, Zhang F, Ling J. The effect of matrixextracellular phosphoglycoprotein and its downstreamosteogenesis-related gene expression on the proliferation and

differentiation of human dental pulp cells. J Endod 2012;38:330e8.

21. Machado JP, Johnson WE, O’Brien SJ, Vasconcelos V,Antunes A. Adaptive evolution of the matrix extracellularphosphoglycoprotein in mammals. BMC Evol Biol 2011;11:342.

22. Li Y, Toole BP, Dealy CN, Kosher RA. Hyaluronan in limbmorphogenesis. Dev Biol 2007;305:411e20.

23. Han R, Kitlinska JB, Munday WR, Gallicano GI, Zukowska Z.Stress hormone epinephrine enhances adipogenesis in murineembryonic stem cells by up-regulating the neuropeptide ysystem. PLoS One 2012;7:e36609.

24. Schilling T, Noth U, Klein-Hitpass L, Jakob F, Schutze N. Plas-ticity in adipogenesis and osteogenesis of human mesenchymalstem cells. Mol Cell Endocrinol 2007;271:1e17.

25. Kingsley DM. The TGF-beta superfamily: new members, newreceptors, and new genetic tests of function in differentorganisms. Gene Dev 1994;8:133e46.

26. Hogan BL. Bone morphogenetic proteins in development. CurrOpin Genet Dev 1996;6:432e8.

27. Massague J. TGFbeta signaling: receptors, transducers, andMad proteins. Cell 1996;85:947e50.

28. Heldin CH, Miyazono K, ten Dijke P. TGF-beta signalling fromcell membrane to nucleus through SMAD proteins. Nature 1997;390:465e71.

29. Gitelman SE, Kobrin MS, Ye JQ, Lopez AR, Lee A, Derynck R.Recombinant Vgr-1/BMP-6-expressing tumors induce fibrosisand endochondral bone formation in vivo. J Cell Biol 1994;126:1595e609.

30. Heikinheimo KA, Laine MA, Ritvos OV, Voutilainen RJ,Hogan BL, Leivo IV. Bone morphogenetic protein-6 is a markerof serous acinar cell differentiation in normal and neoplastichuman salivary gland. Cancer Res 1999;59:5815e21.

31. Blanco Calvo M, Bolos Fernandez V, Medina Villaamil V, Apar-icio Gallego G, Diaz Prado S, Grande Pulido E. Biology of BMPsignalling and cancer. Clin Transl Oncol 2009;11:126e37.

32. Blitz IL, Cho KW. Finding partners: how BMPs select theirtargets. Dev Dyn 2009;238:1321e31.

33. Lee GT, Kwon SJ, Lee JH, Jeon SS, Janq KT, Choi HY, et al.Macrophages induce neuroendocrine differentiation of pros-tate cancer cells via BMP6eIL6 loop. Prostate Mar 3:doi:http://dx.doi.org/1002/pros.21369.

34. Cao Z, Umek RM, McKnight SL. Regulated expression of threeC/EBP isoforms during adipose conversion of 3T3-L1 cells.Genes Dev 1991;5:1538e52.

35. Ruffell D, Mourkioti F, Gambardella A, Kirstetter P, Lopex RG,Rosenthal N, et al. A CREB-C/EBPbeta cascade induces M2macrophage-specific gene expression and promotes muscleinjury repair. Proc Natl Acad Sci USA 2009;106:17475e80.

36. Al-Adsani A, Burke ZD, Eberhard D, Lawrence KL, Shen CN,Rustqi AK, et al. Dexamethasone treatment induces thereprogramming of pancreatic acinar cells to hepatocytes andductal cells. PLoS One 2010;5:e13650.

37. Yin Q, Wang X, Fewell C, Cameron J, Zhu H, Baddoo M, et al.MicroRNA miR-155 inhibits bone morphogenetic protein (BMP)signaling and BMP-mediated EpsteineBarr virus reactivation. JVirol 2010;84:6318e27.

38. Krogmann A, Staiger K, Haas C, Gommer N, Peter A, Heni M,et al. Inflammatory response of human coronary artery endo-thelial cells to saturated long-chain fatty acids. Microvasc Res2011;81:52e9.

39. Hua F, Wang J, Ishrat T, Wei W, Atif F, Sayeed I, et al. Genomicprofile of Toll-like receptor pathways in traumatically brain-injured mice: effect of exogenous progesterone. J Neuro-inflammation 2011;8:42.