Transcriptome and Proteome Analysis of Osteocytes Treated with Nitrogen-Containing Bisphosphonates Nicoletta Bivi,† Jessica Z. Bereszczak,‡ Milena Romanello,† Leo A. H. Zeef,§ Daniela Delneri,§ Franco Quadrifoglio,† Luigi Moro,| Francesco L. Brancia,*,‡ and Gianluca Tell*,† Department of Biomedical Sciences and Technologies, University of Udine, Udine, Italy, Shimadzu Research Laboratory (Europe), Manchester, U.K., Department of Life Science, University of Manchester, U.K., and Center for the Study of Metabolic Bone Diseases, Via Vittorio Veneto 171, Gorizia, Italy 34170 Received July 24, 2008
We combined high-throughput screening of differential mRNAs with mass spectrometric characterization of proteins obtained from osteocytes untreated and treated with Risedronate. Microarray analysis revealed, upon treatment, a marked upregulation of messengers encoding zinc-proteins. MS analysis identified 84 proteins in the osteocytes proteome map. Risedronate affected the expression of 10 proteins, associated with cytoskeleton, stress-response and metabolism. Data validated using gel imaging in combination with the GLaD post digestion isotopic labeling method provide the molecular basis for understanding the role of bisphosphonates as antiapoptotic drugs for osteocytes. Keywords: bisphosphonates • osteocytes • differential proteomics • gene expression • GLaD • isotopic labelling
Introduction Osteoporosis is a pathological condition in which the physiological balance between bone resorption and bone formation is compromised, with the former prevailing over the latter, leading to a thinner and weaker microarchitecture of the bone and getting it more susceptible to fractures.1 Since one of the major triggers of osteoporosis is the physiological decrease of sexual hormones due to aging, it is estimated that 1 in 2 women over 50 years will develop an osteoporosis-related fracture during her lifetime.2 The standard treatment for osteoporosis is the administration of bisphosphonates (BPs), metabolically stable analogues of pyrophosphate. They are characterized by a common P-C-P backbone that determines the affinity for apatite crystals. The types of side chains present in BPs confer to each compound specific chemical and biological properties. In particular, the second generation of BPs, that is, aminobisphosphonates (N-BPs), are derivatives containing an amino group within one of the side chains. N-BPs represent the first choice in the treatment of osteoporosis3 as they can act on both resorption and bone formation during the entire process of bone remodelling. The inhibition of bone resorption is achieved through the induction of apoptosis in osteoclasts, whose decrease in activity and life span3 is due to the direct inhibition of the enzyme farnesyl pyrophosphate synthetase.2 This results in a decrease in post* To whom correspondence should be addressed. Dr. Francesco L. Brancia, Shimadzu Research Laboratory, Wharfside, Trafford Wharf Road, Manchester, M17 1GP, U.K. Tel, 0044 161 8866550; fax, 0044 161 88665501. Prof. Gianluca Tell, Department of Biomedical Sciences and Technologies, University of Udine, Italy. Tel.: +39 0432 494311; fax, +39 0432 494301. † University of Udine. ‡ Shimadzu Research Laboratory (Europe). § University of Manchester. | Center for the Study of Metabolic Bone Diseases. 10.1021/pr8005606 CCC: $40.75
2009 American Chemical Society
translational prenylation of GTP-binding proteins such as Ras and Rac, essential for osteoclast function.4 Moreover, it has been recently established that the decrease in prenylation is followed by the accumulation of the unprenylated forms of these protein species, which can still retain a partial activity, thus, compromising the cell functions.5 N-BPs are also able to exert an opposite, antiapoptotic effect on bone-forming cells at concentrations 4 orders of magnitude lower than those used to induce a pro-apoptotic effect on osteoclasts. As previously shown, N-BPs can prevent osteoblast and osteocyte apoptosis via transient opening of Cx43 hemichannels and ERKs activation.6,7 While the effect on bone resorption is widely accepted as being dependent on protein prenylation inhibition, the effect on bone formation is only partially understood. Recently, our group proposed a novel model to explain the effect of N-BPs on bone forming cells, that involves the nonlytic release of ATP by osteoblasts and the synergy with purinergic signaling, thus, contributing to a mitogenic effect.8 Osteocytes are likely to be derived by terminal differentiation of osteoblasts which become trapped within the forming osteoid tissue during bone formation in spaces called lacunae.9 A characteristic morphologic feature of these cells is the expression of dendritic processes which allow them to be in contact with each other and with other cells on the bone surface via gap junctions. Osteocyte processes are contained in small channels called canaliculi, that connect the cell bodycontaining lacunae with each other and that pass through the matrix, forming a cellular network filled with a complex unmineralized fluid.10,11 Osteocytes play a crucial role in the coordination of responses to the external environment, such as mechanical loading and hormonal stimulation. Several evidence point toward the interpretation that osteocytes are the primary chemo- and mechano-sensory cells of the bone12 as they are in Journal of Proteome Research 2009, 8, 1131–1142 1131 Published on Web 02/18/2009
research articles close contact with the fluid that mediates mechanical loading, hormonal stimulation and drugs effects.13–15,7 Osteocytes physiology is only partially understood and the effect of N-BPs on these cells seems to involve several mechanisms; in order to reach a better insight on these effects, we applied the unbiased strategy of studying the consequences of N-BPs treatment on the proteome and on the transcriptome of osteocytes. To achieve this purpose, murine osteocytic cells were used. Because of the difficulty of growing primary cells in vitro, an immortalized mouse osteocyte-like cell line (i.e., murine long bone osteocyte Y4, MLO-Y4) was developed.16 These cells share most of the features with the primary osteocytes in culture (low alkaline phosphatase, high osteocalcin and connexin 43, the presence of dendritic processes, among others), can be easily grown in vitro16 and proved to be highly useful for early studies devoted to the comprehension of the physiology and biochemistry of this cell type. In this study, we exploit a combination of analytical techniques to study the effect of N-BPs treatment on osteocytes at both the gene expression profile and proteomic levels.
Materials and Methods Cell Culture. MLO-Y4 cells were kindly provided by Dr. L. F. Bonewald (Department of Oral Biology, School of Dentistry, University of Missouri, Kansas City, MO). Cells were cultured on collagen-coated plates (rat tail collagen type I, 0.15 mg/ mL) in alpha modified essential medium (R-MEM) supplemented with 5% (v/v) fetal bovine serum (FBS) and 5% (v/v) calf serum (CS), 2 mM glutamine, 100 U/mL penicillin, 100 µg/ mL streptomycin sulfate and cultured at 37 °C in a humidified atmosphere containing 5% CO2. The cells were seeded in 10 cm dishes and grown to about 80% confluence before harvesting. For the stimulations, Risedronate and Alendronate treatments were performed on cells in medium with serum. The bisphosphonates used in this study were provided by Procter and Gamble Pharmaceuticals (Cincinnati, OH). Preparation of Total Cell Extracts. To obtain total protein extracts from MLO-Y4 osteocytes, cells were lysed directly into buffer containing 7 M urea, 2 M thiourea, 2% (w/v) 3-[(3cholamidopropyl)dimethylammonio]-1-propanesulfonic acid, 10 mM DTT, 1% pH 4-7 L IPG Buffer (GE Healthcare, Milan, Italy), 1% (v/v) β-mercaptoethanol and 40 mM Tris-HCl. Homogenates were then centrifuged at 10 000g for 30 min, at 4 °C, and supernatants collected and stored at -80 °C. Protein quantification and sample normalization were performed by SDS-PAGE and subsequent analysis of gel bands (Image Quant, GE Healthcare). Two-Dimensional Polyacrylamide Gel Electrophoresis. Thirty to 50 µg of total cell extracts was loaded onto 13 cm, pH 4-7 L IPG strips (GE Healthcare) in duplicate. IEF was conducted using an IPGPhor II system (GE Healthcare) according to the manufacturer’s instructions. Focused strips were equilibrated with 6 M urea, 26 mM DTT, 4% (w/v) SDS, 30% (v/v) glycerol in 0.1 M Tris-HCl (pH 6.8) for 15 min, followed by 6 M urea, 0.38 M iodoacetamide, 4% (w/v) SDS, 30% (v/v) glycerol, and a dash of bromophenol blue in 0.1 M Tris-HCl, pH 6.8, for 10 min. The equilibrated strips were applied directly to 12% SDS-polyacrylamide gels and separated at 130 V. Gels were fixed and stained by colloidal Coomassie blue.17 Evaluation of Differentially Represented Spots. Gels were scanned with an Image Scanner II apparatus and analyzed by the ImageMaster 2D Platinum software (GE Healthcare). For each sample, three 2-DE gels were performed and then a 1132
Journal of Proteome Research • Vol. 8, No. 3, 2009
Bivi et al. 18
comparative analysis was conducted. Protein spots were detected and matched between the different samples; individual spot volume values were obtained according to the program instruction and normalized using the program volume normalization function. Ratios of the different sample normalized volume values for the candidate proteins were compared with each other and a mean relative difference in spot intensity was calculated. Differences in protein expression levels were considered as significant when Student’s t test gave a p-value < 0.05. Western Blotting Analysis. Total cell extracts from MLO-Y4 osteocytes, obtained as described above, were separated on 10% SDS-PAGE. Then, proteins were transferred onto nitrocellulose membranes (Schleicher and Schuell Bioscience, Keene, NH) or PVDF membrane (Schleicher and Schuell) for the detection of H2B. After saturation with 10% (w/v) nonfat dry milk in PBS and 0.1% (w/v) Tween-20, the membranes were incubated with one of the following antibodies: anti-calreticulin polyclonal antibody (PA3-900, Affinity Bioreagents, Golden, CO), anti-HSP-60 monoclonal antibody (Abcam, Cambridge, U.K.), anti-H2B polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA), anti-HSP-70 monoclonal antibody (MA3-007, Affinity Bioreagents). Filters were incubated for 3 h at room temperature with each of them. After three washes with PBS with 0.1% (w/v) Tween-20, the membranes were incubated with anti-rabbit or anti-mouse immunoglobulins conjugated with peroxidase (Sigma-Aldrich, Milan, Italy). After 2 h of incubation at room temperature, the membranes were washed three times with PBS with 0.1% (w/v) Tween-20, and the blots were developed by the enhanced chemiluminescence (ECL) procedure (Pierce Biotechnology, Rockford, IL). The signals from each protein band were normalized against the tubulin content by using the polyclonal anti-tubulin antibody (Sigma-Aldrich). Blots were quantified by Image Quant (GE Healthcare). Cell Viability Assay. An MTT assay was used for growth experiments in a 96-well plate in quadruplicate. The assay allowed the quantification of the cell viability by measuring the mitochondrial function.19 Cells at 80% confluence were harvested with 1× trypsin/EDTA solution and seeded into a 96-well plate at 3 × 103 cells/well and maintained overnight in complete medium. After incubation with Risedronate or Alendronate at concentrations ranging from 10-8 to 10-6 M for 48 and 72 h, a 1/10 volume of MTT solution (4 mg/mL in PBS 1×) was added and incubated for 3 h under 5% CO2/95% air at 37 °C. Then, the supernatant was aspirated, an equal volume of DMSO was added to the cells, and the MTT formazan was dissolved by pipetting. The absorbance was measured on an enzyme-linked immunosorbant assay (ELISA) plate reader (EL808 Ultra Microplate Reader Bio-Tek Instruments, Inc., Winooski, VT) with a test and reference wavelength of 570 and 690 nm, respectively. RNA Extraction. MLO-Y4 cells were seeded on 10 cm plates and treated with 10-7 M Risedronate and Alendronate as described previously. About 7 × 106 cells were then collected and lysed in Trizol (Invitrogen, Milan, Italy) and the suspension was passed through an 18-gauge syringe. After addition of chloroform (0.2 mL/mL Trizol), the cellular debris were removed by centrifugation at 2700g for 15 min at 4 °C. The RNA was precipitated with an equal volume of isopropanol. Microarrays Hybridization and Analysis. Microarrays were used as described in the Affymetrix gene chip technical manual. Briefly, 15 µg of total RNA was converted to cDNA followed by in vitro transcription for linear amplification of transcripts and
research articles
Transcriptome and Proteome Analysis of Osteocytes incorporation of biotinylated CTP and UTP. The cRNA products were fragmented to 200 nucleotides or less, labeled and hybridized for 16 h at 45 °C to Affymetrix mouse genome (mouse430_2) microarrays. The arrays were washed at low and high stringency before staining with streptavidin-phycoerythrin. Fluorescence was amplified using biotinylated anti-streptavidin followed by incubation with streptavidin-phycoerythrin stain. Technical quality control was performed with dChip (V2005) using the default settings. Background correction and quantile normalization were performed using GCRMA in Bioconductor.20 Differential expression analysis was performed with Limma using the functions lmFit and eBayes.21 The analysis was done by comparing arrays from basal sample replicates to either Risedronate or Alendronate treated samples. Genelists of differentially expressed genes [295 probesets for Risedronate, 132 for Alendronate (62 shared by both lists)] were created by filtering for probesets with a p-value less than 0.05 and foldchange greater than 1.5. Gene ontology (GO) over-representation analysis was performed using the functional annotation tool of the Database for Annotation, Visualization and Integrated Discovery (DAVID) 2.1 program.22 Quantitative Evaluation of Gene Expression by Q-PCR. RNA was extracted from 106 cells using Trizol reagent. One microgram of RNA was used in a reverse transcription reaction using iScript cDNA synthesis kit (BioRad, Milan, Italy) following the manufacturer’s instructions. Q-PCR was performed using the BioRad icycler and IQ supermix buffer containing DNA polymerase and SYBR Green (BioRad). Three replicate amplification reactions were performed in 96-well plates (BioRad). Each reaction mixture contained 7.5 µL of IQ supermix buffer 5×, 300 nM forward and reverse primers, and 5 µL of cDNA previously diluted 1:20 in a final volume of 15 µL. Primers for the reactions are shown in Table 2. Cycling conditions were as follows: 10 s at 95 °C, 30 s at 60 °C, 10 s at 65 °C, repeated 40 times. Data analysis was performed using the BioRad icycler PCR detection and analysis software, version 3.0 (BioRad). DNA was quantified using the standard curve method with the background subtracted. Known doses of cDNA were used to obtain the standard curve for each gene (amounts between 1 and 625 ng). A melting curve was determined for each sample to detect primer dimers, in which case data were not used. The signal from each candidate gene was normalized for the signal for three housekeeping genes: TATA box binding protein-like 1 (TBPL1), beta-2-microglobulin (B2M) and hydroxymethylbilane synthase (HMBS). Results are expressed as the ratio of target cDNA in the treated samples and target cDNA in the untreated sample (“basal”). Sample Preparation for Mass Spectrometric Analysis. Spots from 2-DE were manually excised from the gel and dehydrated with acetonitrile (ACN). Proteins were then rehydrated in trypsin solution (Sigma Aldrich) and in-gel protein digestion was performed overnight at 37 °C. A portion of the tryptic digests was analyzed by MALDI-ToF MS and online-LC ESIIT-ToF MS in MS/MS mode for protein identification. Digested aliquots were used directly or subjected to a desalting/ concentration step on ZipTipC18 (Millipore Corp., Bedford, MA) before MALDI-ToF MS analysis. From each spot, 0.5 µL of the obtained peptides were spotted on the MALDI target, allowed to dry and then covered with 0.5 µL of the matrix solution
Table 1. GO Categories with p-Value < 0.1 (DAVID (EASE Modified Fisher Exact p-Value))a RIS and ALE intersection (63 probesets)
RIS (310 probesets)
Zinc Ion Binding Zinc Ion Binding transmembrane receptor – protein serine/threonine kinase signaling pathway cation transmembrane transporter activity –
–
– –
–
helicase activity regulation of transcription ((bZIP) transcription factor) mRNA processing chromatin modification ubiquitin cycle regulation of cell motility ATPase activity regulation of progression through mitotic cell cycle microtubule binding –
–
–
–
–
– – – – – –
–
nucleus
ALE (149 probesets)
Zinc Ion Binding transmembrane receptor protein serine/threonine kinase signaling pathway – nucleus (nuclear pore) – regulation of transcription
– – – – – –
microtubule binding enzyme regulator activity manganese ion binding intracellular protein transport
a Closely related GO categories have been summarized into a single term using the Functional Annotation Clustering analysis in DAVID.
R-cyano-4-hydroxycinnamic acid (CHCA) dissolved in ACN 50% (v/v) and 0.1% TFA. Guanidination Labeling Derivatization (GLaD) of Peptides. The tryptic digests from a selected pool of identified proteins were subjected to guanidino labeling derivatization (GLaD) using isotopic variants of O-methyl isourea (OMIU). Tryptic digests from untreated osteocytes were labeled with the ‘light’ version of O-methyl isourea (12C,14N2-OMIU) (Lancaster, Heysham, Lancashire, U.K.). Tryptic digests from bisphosphonate treated osteocytes were labeled with the ‘heavy’ version of O-methylisourea (13C,15N2-OMIU) (in house synthesized23). Derivatization was carried out in 7 M ammonium hydroxide at pH >10, overnight. Termination of the reaction was carried out by addition of a suitable volume of 10% (v/v) TFA until the pH of the reaction mixture was
