PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
J Cell Biochem. Author manuscript; available in PMC 2017 May 1.
Published in final edited form as:
PMCID: PMC4777692
NIHMSID: NIHMS761721

THE DMP1-SOST TRANSGENE INTERACTS WITH AND DOWNREGULATES THE DMP1-CRE TRANSGENE AND THE ROSANOTCH ALLELE

Abstract

Activation of Notch1 in osteocytes of RosaNotch mice, where a loxP-flanked STOP cassette and the Nicd coding sequence were targeted to the reverse orientation splice acceptor (Rosa)26 locus, causes osteopetrosis associated with suppressed Sost expression and enhanced Wnt signaling. To determine whether Sost downregulation mediates the effects of Notch activation in osteocytes, RosaNotch mice were crossed with transgenics expressing Cre recombinase or SOST under the control of the dentin matrix protein (Dmp)1 promoter. Dmp1-SOST transgenics displayed vertebral osteopenia and a modest femoral cancellous and cortical bone phenotype, whereas hemizygous Dmp1-Cre transgenics heterozygous for the RosaNotch allele (Dmp1-Cre;RosaNotch) exhibited osteopetrosis. The phenotype of Notch activation in osteocytes was prevented in Dmp1-Cre;RosaNotch mice hemizygous for the Dmp1-SOST transgene. The effect was associated with downregulated Notch signaling and suppressed Dmp1 and Rosa26 expression. To test whether SOST regulates Notch expression in osteocytes, cortical bone cultures from Dmp1-Cre;RosaNotch mice or from RosaNotch control littermates were exposed to recombinant human SOST. The addition of SOST had only modest effects on Notch target gene mRNA levels and suppressed Dmp1, but not Cre or Rosa26, expression. These findings suggest that prevention of the Dmp1-Cre;RosaNotch skeletal phenotype by Dmp1-SOST is not secondary to SOST expression but to interactions among the Dmp1-SOST and Dmp1-Cre transgenes and the Rosa26 locus. In conclusion, the Dmp1-SOST transgene suppresses the expression of the Dmp1-Cre transgene and of Rosa26.

Keywords: Sost, Cre recombinase, Rosa26 locus, Dmp1, bone remodeling, osteocytes

INTRODUCTION

The four Notch receptors and their classical cognate ligands are single-pass transmembrane proteins that mediate signals between adjacent cells and thereby determine cell differentiation and function [Zanotti and Canalis, 2010; Zanotti and Canalis, 2013]. Upon ligand binding, Notch receptors are subjected to sequential proteolytic cleavages that result in the release of the Notch intracellular domain (Nicd) in the cytoplasm. Subsequently, the Nicd translocates to the nucleus and interacts with the DNA associated protein recombination signal binding protein for immunoglobulin kappa J region. These events result in the displacement of transcriptional repressors by activators of transcription and in the induction of gene expression. Classical target genes of Notch signaling include hairy enhancer of split (Hes)1, 5 and 7 and Hes related with YRPW motif (Hey)1, 2 and like (L) [Kovall, 2008; Nam et al., 2006; Schroeter et al., 1998; Wilson and Kovall, 2006].

Osteocytes are cells of osteoblastic origin that become embedded in the bone matrix as new bone is formed and that communicate with each other and with endosteal lining cells through a canalicular network. Osteocytes translate mechanical forces into intracellular signals that regulate cellular behavior and allow bone to adapt to extracellular stimuli, a process known as mechanotransduction [Kamel et al., 2010; Nakashima et al., 2011; Tatsumi et al., 2007; Turner and Pavalko, 1998; Xiong et al., 2011]. Osteocytes and osteoblasts play different roles in skeletal homeostasis, and activation of regulatory mechanisms shared by osteoblasts and osteocytes may lead to distinct outcomes on bone remodeling. For example, Notch activation in immature and mature osteoblasts causes osteopenia, whereas its preferential activation in osteocytes causes osteopetrosis [Canalis et al., 2013b]. The osteopenia caused by Notch induction in immature and mature osteoblasts is secondary to an inhibitory role of Notch in osteoblast differentiation [Canalis et al., 2013a; Canalis et al., 2013b; Zanotti and Canalis, 2014; Zanotti et al., 2011]. In contrast, the pronounced increase in cancellous bone induced by Notch1 activation in osteocytes is due to a suppression of bone resorption and an increase of cortical bone formation [Canalis et al., 2013a; Canalis et al., 2013b].

Although the mechanisms of Notch1 action in osteoblasts have been subjected to intense study, those determining the effects of Notch activation in osteocytes are less clear. In osteocytes, Notch suppresses Sost expression and, as a consequence, it can enhance Wnt/β-catenin signaling [Canalis, 2013]. In the present study, we sought to determine whether Sost downregulation contributes to the skeletal effects of Notch1 activation in osteocytes. To activate Notch1 preferentially in these cells, RosaNotch mice, where a loxP-flanked STOP cassette was cloned into the reverse orientation splice acceptor (Rosa)26 promoter upstream of the coding sequence of the Notch1 Nicd, were used [Murtaugh et al., 2003]. These mice were crossed with transgenics expressing Cre recombinase under the control of the 9.6 kilobase (kb) fragment of the dentin matrix protein (Dmp)1 promoter (Dmp1-Cre) to generate Dmp1-Cre;RosaNotch mice [Lu et al., 2007]. To determine the contribution of Sost downregulation to the skeletal effects of Notch activation in osteocytes, Dmp1-Cre;RosaNotch mice were crossed with transgenics expressing SOST under the control of the 8 kb fragment of the Dmp1 promoter to obtain Dmp1-Cre;RosaNotch;Dmp1-SOST mice [Tu et al., 2012]. The phenotype of Notch1 activation and SOST overexpression in osteocytes was determined in male and female mice. These studies were complemented by in vitro experiments where osteocyte-enriched cortical bone cultures from Dmp1-Cre;RosaNotch and RosaNotch littermates were exposed to recombinant human SOST.

MATERIALS AND METHODS

Genetically Modified Mice

To express SOST preferentially in osteocytes, Dmp1-SOST transgenic mice in a C57BL/6 genetic background were obtained from C. O’Brien and R. Jilka (University of Arkansas for Medical Sciences, Little Rock, AR) [Rhee et al., 2011; Tu et al., 2012]. These mice were generated by oocyte injection of a transgenic DNA construct containing the 8 kb upstream of the putative transcriptional start site, the first exon, the first intron and 17 base pair (bp) of exon 2 of Dmp1, followed by the human SOST coding sequence [Kalajzic et al., 2004]. Hemizygous Dmp1-SOST mice (Dmp1-SOST) were crossed with wild type C57BL/6 mice to obtain experimental Dmp1-SOST and control wild type littermates.

To induce Notch signaling in vivo, RosaNotch mice in a 129SvJ/C57BL/6 genetic background were used (Jackson Laboratory, Bar Harbor, ME) [Murtaugh et al., 2003; Stanger et al., 2005]. In these mice, the Rosa26 locus was targeted with a DNA construct containing a fragment of the Rosa26 promoter, a STOP cassette flanked by loxP sites and the coding sequence of the Notch1 Nicd. Excision of the STOP cassette by Cre recombination causes expression of Nicd from the Rosa26 locus and subsequent activation of Notch signaling [Buchholz et al., 1996; Sauer and Henderson, 1988]. To activate Notch1 preferentially in osteocytes, homozygous RosaNotch mice were mated with hemizygous Dmp1-Cre transgenics (Dmp1-Cre) in a C57BL/6 background to create Dmp1-Cre;RosaNotch experimental mice and RosaNotch littermate controls. Dmp1-Cre transgenics express Cre under the control of a DNA fragment containing the 9.6 kb upstream of the putative transcriptional start site, the first intron and 17 bp of exon 2 of Dmp1 (Dr. Feng, Texas A&M Health Science Center, Dallas, TX) [Lu et al., 2007].

To determine the skeletal effects of Notch1 activation in the context of SOST overexpression in osteocytes, Dmp1-SOST transgenics were crossed with RosaNotch mice to create Dmp1-SOST hemizygotes, homozygous for the RosaNotch allele. The latter were crossed with homozygous Dmp1-Cre mice to generate an offspring heterozygous for the RosaNotch allele and hemizygous for the Dmp1-Cre transgene, where ~50% of the mice would carry one copy of the Dmp1-SOST transgene (Dmp1-Cre;RosaNotch;Dmp1-SOST) and 50% would not (Dmp1-Cre;RosaNotch). This mating scheme does not allow for a direct comparison between Dmp1-Cre;RosaNotch and control RosaNotch mice in the same litter. Therefore, data obtained from the Dmp1-Cre;RosaNotch mice generated for this study were pooled with data from studies comparing Dmp1-Cre;RosaNotch and RosaNotch littermate mice, collected over a period of ~4 years. The pooled data from Dmp1-Cre;RosaNotch and RosaNotch mice, and data from the Dmp1-Cre;RosaNotch;Dmp1-SOST mice, were subjected to non-parametric statistical testing to determine whether restoration of SOST reverses the skeletal phenotype of Notch1 activation in osteocytes. Genotyping of Dmp1-Cre and Dmp1-SOST transgenes and RosaNotch alleles was carried out by polymerase chain reaction (PCR) in tail DNA extracts in the presence of specific primers (Table 1, all from Integrated DNA Technologies (IDT), Coralville, IA).

Table 1
Primers used for allele identification by PCR.

Animal experiments were approved by the Animal Care and Use Committees of Saint Francis Hospital and Medical Center and of UConn Health.

Microcomputed Tomography (μCT)

Bone microarchitecture of femurs and vertebrae from experimental and control mice was determined using a microcomputed tomography instrument (μCT 40; Scanco Medical AG, Bassersdorf, Switzerland), which was calibrated weekly using a phantom provided by the manufacturer [Bouxsein et al., 2010; Glatt et al., 2007]. The vertebral body of L3 and femurs were scanned in 70% ethanol at high resolution, energy level of 55 peak kV, intensity of 145 μA and integration time of 200 ms. Trabecular bone volume fraction and microarchitecture were evaluated starting ~1.0 mm from the cranial side of the vertebral body, or 1.0 mm proximal from the condyles of the distal femur. For L3 vertebrae and femurs, a total of 160 consecutive slices were acquired at an isotropic voxel dimension of 216 μm3 and a slice thickness of 6 μm and chosen for analysis. Contours were manually drawn every 10 slices a few voxels away from the endocortical boundary to define the region of interest for analysis. The remaining slice contours were iterated automatically. Trabecular regions were assessed for total volume, bone volume, bone volume fraction (bone volume/total volume), trabecular separation, number and thickness, connectivity density, structure model index, and material density using a Gaussian filter (σ = 0.8) and user-defined thresholds [Bouxsein et al., 2010; Glatt et al., 2007]. For analysis of femoral cortical bone, contours were iterated across 100 slices along the cortical shell of the femoral mid-shaft, excluding the marrow cavity. Analyses of bone volume/total volume, cortical thickness, periosteal perimeter, endosteal perimeter, total cross-sectional and cortical bone area, were performed using a Gaussian filter (σ = 0.8) and user-defined thresholds. The terminology and units used are in accordance with guidelines published in the Journal of Bone and Mineral Research [Bouxsein et al., 2010].

Osteocyte-enriched Cultures

Cultures of osteocyte-enriched cells were obtained following a modification of a previously described method [Halleux et al., 2012]. One month old Dmp1-Cre;RosaNotch mice and littermate RosaNotch controls were sacrificed by CO2 inhalation followed by cervical dislocation. Femurs were removed aseptically, the surrounding tissues dissected, the proximal epiphyseal ends excised and the bone marrow removed by centrifugation. The distal epiphysis was excised, and femurs were digested for 20 min at 37°C with bacterial collagenase (Type II collagenase from Clostridium histolyticum, Worthington Biochemical Corp., Lakewood, NJ) pretreated with tosyllisine chloromethyl keton hydrocholride (Calbiochem, La Jolla, CA) and subsequently exposed to EDTA 5 mM (Life Technologies, Grand Island, NY) for 20 min at 37°C. The resulting osteocyte-enriched cortical femurs were cultured individually in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with nonessential amino acids (both from Life Technologies), 100 μg/ml ascorbic acid (Sigma-Aldrich, St. Louis, MO) and 10% fetal bovine serum (Atlanta Biologicals, Inc., Atlanta, GA) for 3 days at 37°C in a humidified 5% CO2 incubator. The effects of SOST on Notch signaling and gene expression in osteocytes were tested with recombinant human SOST (Thermo Scientific, Waltham, MA). SOST was dissolved at a concentration of ~220 μM in phosphate buffered saline and 0.1% bovine serum albumin (Sigma-Aldrich). Osteocyte-enriched cortical femurs were transferred to culture medium without serum for 24 h and exposed to SOST 40 nM for 6 h or to an equivalent dilution of vehicle solution, as a control.

Quantitative Reverse Transcription-Polymerase Chain Reaction (qRT-PCR)

Total RNA was extracted from calvariae of Dmp1-SOST male mice and wild type littermates of the same sex, tibiae of Dmp1-Cre;RosaNotch and Dmp1-Cre;RosaNotch;Dmp1-SOST littermates of both sexes following the removal of the epiphyses and bone marrow by centrifugation, and osteocyte-rich preparations. Integrity of total RNA was tested by microfluidic electrophoresis on an Experion automated electrophoresis system (Bio-Rad, Hercules, CA), and RNA exhibiting signs of degradation was excluded from further analysis. To measure mRNA levels, equal amounts of RNA were reverse-transcribed using the iScript RT-PCR kit (Bio-Rad), according to manufacturer’s instructions, and amplified in the presence of specific primers (Table 2, all primers from IDT) and iQ SYBR Green Supermix (Bio-Rad) at 60°C for 35 cycles [Nazarenko et al., 2002a; Nazarenko et al., 2002b]. Transcript copy number was estimated by comparison with a serial dilution of cDNA for Axin2 (GE Healthcare Dharmacon, Lafayette, CO), Cre recombinase (Life Technologies), Dmp1, Notch1 and Sost (all from Thermo Scientific), Hey1, Hey2 (both from T. Iso, Los Angeles, CA), HeyL (from D. Srivastava, Dallas, TX), tumor necrosis factor receptor superfamily, member 11b (Tnfrsf11b, encoding for osteoprotegerin) and Wnt1 inducible signaling pathway protein (Wisp1) (both from American Type Tissue Culture Collection (ATCC), Manassas, VA), tumor necrosis factor, member 11 (Tnfsf11, encoding for receptor activator of nuclear factor-κB ligand, Rankl, from Source BioScience, Nottingham, UK), dickkopf-related protein 1 (Dkk1, from C. Niehrs, Heidelberg, Germany) and Rosa26 [Glinka et al., 1998; Iso et al., 2001; Nakagawa et al., 1999]. The latter was generated by cloning a ~200 bp synthetic DNA fragment (from IDT), homologous to expressed sequences from the Rosa26 locus, into pcDNA3.1(−) (Life Technologies) by isothermal single reaction assembly, using commercially available reagents (New England Biolabs, Ipswich, MA) [Gibson et al., 2009]. Expression of the Dmp1-SOST transgene was detected with a forward primer spanning the exon1-exon2 junction of the Dmp1/SOST hybrid transcript and a reverse primer homologous to a 5′ fragment of the SOST coding sequence (Table 2). Dmp1-SOST expression was estimated by comparing to a serial dilution of reverse transcribed mRNA from calvaria of a Dmp1-SOST transgenic mouse. Reactions were carried out in a CFX96 qRT-PCR detection system (Bio-Rad), and fluorescence was monitored during every PCR cycle at the annealing step. Gene expression was corrected for glyceraldehyde-3-phosphate dehydrogenase (Gapdh) or for ribosomal protein L38 (Rpl38) copy number, estimated by comparison with a serial dilution of cDNA for Gapdh (from R. Wu, Ithaca, NY) or Rpl38 (from ATCC) [Tso et al., 1985]. Presence of a single PCR product was documented by melting curve analysis carried out after the final amplification cycle.

Table 2
Primers used for qRT-PCR determinations. GenBank accession number identifies transcript recognized by primer pairs.

Statistical Analysis

Data for body weight, femoral length and indices of bone microarchitecture are expressed as medians, means, 75th and 25th percentile, and maximum and minimum because of the limited number of observations. Normal distribution for these parameters was not assumed, and statistical differences between groups were determined by Mann-Whitney U-test for pairwise comparisons or by Kruskal-Wallis test for multiple comparisons. Data for mRNA expression are displayed as means ± SEM, and statistical differences were determined by Mann-Whitney U-test for pairwise comparisons or by ANOVA with Holm-Sidak post-hoc analysis for multiple comparisons.

RESULTS

Femoral Microarchitecture of Dmp1-SOST Transgenics

To characterize the impact of the preferential SOST overexpression in osteocytes, 1 month old Dmp1-SOST hemizygous (Dmp1-SOST) mice were compared to wild type littermates. Expression of the Dmp1-SOST transgene was confirmed in calvarial bones of Dmp1-SOST male mice, and body weight and femoral length were not appreciably different between male and female transgenics and control wild type littermates of the same sex (Figure 1). In accordance with previous observations, Dmp1-SOST transgenics displayed modest osteopenia in the femoral cancellous bone compartment, and cortical bone architecture was not different from control littermates [Rhee et al., 2011; Tu et al., 2012]. Confirming a previous report, male and female Dmp1-SOST transgenics exhibited a marked decrease of cancellous bone volume in lumbar vertebrae (L3), secondary to reduced trabecular number and thickness (Table 3) [Tu et al., 2012].

Figure 1
In A, qRT-PCR analysis of Dmp1-SOST expression in calvarial total RNA from 1 month old control male mice and Dmp1-SOST (SOST) transgenic littermates of the same sex. Data expressed as Dmp1-SOST copy number, corrected for Rpl38 copy number. Values are ...
Table 3
Femoral and vertebral microarchitecture assessed by μCT of 1 month old male and female Dmp1-SOST transgenics (SOST) and littermate wild-type controls (Control) of the same sex.

Global and Skeletal Phenotype of Preferential Notch1 Activation in Osteocytes and SOST Overexpression

Previous work from this laboratory demonstrated that preferential expression of the Notch1 Nicd in osteocytes in Dmp1-Cre;RosaNotch mice causes a marked increase in cancellous bone due to a suppression of bone resorption [Canalis et al., 2013a; Canalis et al., 2013b]. Notch1 activation in osteocytes downregulated Sost expression and induced Wnt/β-catenin signaling, possibly explaining the osteopetrotic phenotype observed [Canalis, 2013; Canalis et al., 2013a]. To establish the contribution of the downregulation of Sost to the effects of Notch1 activation in osteocytes, the consequences of Notch1 induction were characterized either in the context (Dmp1-Cre;RosaNotch;Dmp1-SOST), or not (Dmp1-Cre;RosaNotch), of SOST overexpression.

In agreement with previous findings, Notch1 Nicd induction resulted in a decrease in weight and femoral length in female but not in male mice [Canalis et al., 2013b]. In contrast to the observations in Dmp1-SOST mice, SOST overexpression in the presence of Notch1 activation resulted in reduced weight and shorter femoral length in male and female mice, so that both parameters were significantly lower in Dmp1-Cre;RosaNotch;Dmp1-SOST mice than in RosaNotch controls (Figure 2). These observations suggest that the dual Notch1 activation and SOST overexpression in osteocytes causes a reduction in skeletal and body size in male and female mice.

Figure 2
Weight and femoral length of male and female 1 month old RosaNotch controls (Control, white boxes), Dmp1-Cre;RosaNotch (Notch, dark grey boxes) and Dmp1-Cre;RosaNotch;Dmp1-SOST mice (Notch;SOST, light grey boxes). RosaNotch and Dmp1-Cre;RosaNotch mice ...

To circumvent the underlying skeletal effects of SOST overexpression in cancellous vertebral bone, which would affect the interpretation of the results, skeletal phenotyping was conducted in femoral cancellous and cortical bone, where the Dmp1-SOST transgene caused only a modest phenotype. Microcomputed analysis of the distal femur and of the femoral mid-shaft confirmed former studies reporting the effects of Notch1 activation in osteocytes [Canalis et al., 2013a; Canalis et al., 2013b]. One month old male and female Dmp1-Cre;RosaNotch mice displayed higher trabecular bone volume/tissue volume than littermate RosaNotch controls of the same sex. This effect was secondary to increased trabecular number and was associated with an increase in connectivity density. Although periosteal perimeter, total area and bone area were greater and endocortical perimeter was smaller in the context of Notch1 activation than under control conditions, cortical bone was porous so that cortical bone volume/total volume decreased in Dmp1-Cre;RosaNotch mice of both sexes (Figures 3 and and4).4). Indices of cancellous femoral bone volume and structure were not different between Dmp1-Cre;RosaNotch;Dmp1-SOST male and female mice and RosaNotch controls of the same sex. A modest increase in structural model index (SMI) was noted in the context of SOST overexpression, indicating that trabeculae in Dmp1-Cre;RosaNotch;Dmp1-SOST mice were rod-like and more fragile. SOST overexpression partially restored indices of cortical bone microarchitecture and restored bone area in male mice, and precluded the effects of Notch1 on endocortical perimeter in male and female mice (Figures 3 and and4).4). These findings demonstrate that the Dmp1-SOST transgene prevents the cancellous bone osteopetrosis induced by Notch1, and lessens the impact of Notch1 activation in osteocytes on the cortical bone compartment.

Figure 3
Femoral microarchitecture assessed by μCT analysis of the distal femoral trabecular compartment (upper panel) and femoral mid-shaft cortical bone (lower panel) of 1 month old control RosaNotch (Control, white boxes), Notch overexpressing Dmp1-Cre;Rosa ...
Figure 4
Femoral microarchitecture assessed by μCT analysis of the distal femoral trabecular compartment (upper panel) and femoral mid-shaft cortical bone (lower panel) of 1 month old control RosaNotch (Control, white boxes), Notch overexpressing Dmp1-Cre ...

Mechanisms of Notch Action in Osteocytes

In previous work, we reported that activation of Notch1 in osteocytes induced Tnfrsf11b and suppressed Sost and Dkk1, which encodes for the Wnt antagonist dickkopf1, possibly explaining the suppression of bone resorption observed in Dmp1-Cre;RosaNotch mice [Canalis et al., 2013a]. To explore the mechanisms whereby SOST overexpression prevents Notch1-induced osteopetrosis, gene expression was analyzed in tibiae from 1 month old Dmp1-Cre;RosaNotch and Dmp1-Cre;RosaNotch;Dmp1-SOST littermates. Because the Dmp1-SOST transgene prevented the skeletal phenotype of Dmp1-Cre;RosaNotch mice both in male and female mice, these studies were carried out in bones from mice of both sexes, and data from gene expression in male and female mice were pooled.

Expression of the Dmp1-SOST transgene was confirmed in Dmp1-Cre;RosaNotch;Dmp1-SOST mice and was associated with downregulation of Tnfrsf11b and Tnfsf11, and no effects of Dmp1-SOST on murine Sost and on Dkk1 expression were noted (Figure 5A–B). The Dmp1-SOST transgene downregulated expression of Axin2 and Wisp1, indicating that the transgene suppressed Wnt/β-catenin signaling in the context of Notch induction in osteocytes. To demonstrate whether the Dmp1-SOST transgene had an effect on the activation of Notch signaling, transcript levels of Notch1 Nicd and of the Notch target genes Hey1, Hey2 and HeyL were measured. Notch1 Nicd and Hey transcripts were detected in bone extracts from both experimental cohorts and were suppressed by 60 to 80% in the context of SOST overexpression (Figure 5C).

Figure 5
qRT-PCR analysis of Dmp1-SOST (A), Tnfrsf11b, Tnfsf11, Sost, Dkk1, Axin2 and Wisp1 (B), Notch1 Nicd (Nicd) Hey1, Hey2, HeyL (C) and Rosa26, Cre and Dmp1 (D) expression in total RNA from tibiae of 1 month old Notch overexpressing Dmp1-Cre;RosaNotch (Notch ...

Levels of Notch1 Nicd transcripts are dictated by the activities of the Rosa26 locus and of the Dmp1-Cre transgene, so that inhibition of Notch signaling by Dmp1-SOST could be secondary to either the suppression of Rosa26 expression or the downregulation of the Dmp1-Cre transgene. Analysis of gene expression revealed that Rosa26, Cre and endogenous Dmp1 mRNA levels were reduced by ~50 to 90% in Dmp1-Cre;RosaNotch;Dmp1-SOST mice compared to Dmp1-Cre;RosaNotch littermates (Figure 5D). Therefore, the reversal of the Notch-induced osteopetrotic phenotype by Dmp1-SOST could be due to the inhibition of Nicd expression from the RosaNotch allele and not due to the reinstitution of SOST levels. The downregulation of Notch1 Nicd was attributable to the suppression of the Rosa26 promoter and to a reduction in Cre recombinase.

Effects of Notch1 Activation and of SOST in Osteocyte-enriched Cultures

The inhibition of Notch signaling and downregulation of Rosa26, Dmp1-Cre and Dmp1 by Dmp1-SOST in vivo could be secondary to SOST overexpression or to a genetic interaction between the Dmp1-SOST and Dmp1-Cre transgenes. To determine whether SOST was responsible for the present observations, osteocyte-enriched cultures from Dmp1-Cre;RosaNotch mice and RosaNotch littermates of both sexes were exposed to recombinant human SOST. Notch1 Nicd, Hey1, Hey2 and HeyL mRNA levels were increased in the context of Notch1 induction, confirming activation of Notch signaling in osteocytes. SOST had no effect on Notch1 Nicd and Hey1 expression either in cultures from Dmp1-Cre;RosaNotch or from RosaNotch mice; and in the context of Notch1 Nicd overexpression, SOST induced Hey2 mRNA levels and caused a modest suppression of HeyL transcripts (Figure 6). These results demonstrate that SOST has modest effects on Notch target gene expression and does not inhibit Notch signaling in osteocytes.

Figure 6
Osteocyte-enriched femoral diaphyses from 1 month old RosaNotch mice (Control, 2 leftmost bars in each panel) or Dmp1-Cre;RosaNotch littermates (Notch, 2 rightmost bars in each panel) of both sexes exposed to vehicle (white bars) or SOST (black bars). ...

In agreement with the known inhibitory effects of SOST on Wnt/β-catenin signaling, SOST suppressed the Wnt target genes Axin2 and Wisp1 expression in osteocyte-enriched cultures (Figure 6). In contrast to the stimulation of Wnt signaling by Notch1 activation in osteocytes in vivo [Canalis et al., 2013a], transcript levels for Axin2 were suppressed in cultures from Dmp1-Cre;RosaNotch mice and Wisp1 expression was not affected by Notch1 activation (Figure 6). These observations indicate that suppression of Sost and enhancement of Wnt/β-catenin signaling by Notch1 activation occurs only in vivo.

A modest induction of Rosa26 transcripts was observed in osteocytes from RosaNotch mice treated with SOST, an effect that did not recapitulate the suppression of Rosa26 expression by Dmp1-SOST transgenics in vivo. Dmp1-Cre transgene expression was documented in cultures from Dmp1-Cre;RosaNotch mice, and the addition of SOST had no effect on Cre mRNA levels. However, in accordance with the in vivo results, SOST suppressed endogenous Dmp1 transcript levels (Figure 6). These findings indicate that SOST does not suppress Notch signaling in osteocytes and does not regulate the expression of Rosa26 or of the Dmp1-Cre transgene in these cells, suggesting that reinstitution of SOST levels was not the event that prevented the osteopetrotic phenotype of Dmp1-Cre;RosaNotch mice.

DISCUSSION

Our findings confirm that preferential Notch1 activation in osteocytes causes an osteopetrotic phenotype that can be explained by an induction of osteoprotegerin leading to an inhibition of bone resorption and decreased cancellous bone remodeling [Canalis et al., 2013a; Canalis et al., 2013b]. The effect on osteoprotegerin may be a direct action of Notch or may be secondary to an enhancement of Wnt signaling due to the suppression of Sost. Wnt signaling induces osteoblastogenesis and inhibits osteoclastogenesis and bone resorption, and the constitutive activation of β-catenin in the osteoblastic and osteoclastic lineages causes osteopetrosis [Glass et al., 2005; Monroe et al., 2012; Wei et al., 2011]. The inhibitory effect of Wnt signaling on bone resorption has been explained by an increase in osteoprotegerin expression by cells of the osteoblastic lineage and by direct effects of Wnt on osteoclast precursors [Albers et al., 2013; Chen and Long, 2013; Sato et al., 2009]. Although an increase in bone volume is often the result of increased bone formation, it also occurs in conditions where suppressed bone resorption and remodeling are the dominant events, such as the inactivation of the receptor activator of nuclear factor-κB and the transgenic overexpression of Tnfrsf11b [Dougall et al., 1999; Li et al., 2000; Ominsky et al., 2009; Simonet et al., 1997].

We postulated that the downregulation of Sost by Notch1 activation in osteocytes could be responsible for the skeletal phenotype of Dmp1-Cre;RosaNotch mice causing an enhancement of Wnt/β-catenin signaling. An attempt to test this hypothesis was made by crossing Dmp1-Cre;RosaNotch mice with mice harboring conditional alleles for cadherin-associated protein beta (Ctnnb)1, which encodes for β-catenin, in order to inactivate β-catenin preferentially in osteocytes [Brault et al., 2001]. This approach however, resulted in a pronounced skeletal phenotype and high mouse lethality (Canalis unpublished observations), findings that are in accordance with previously published work documenting a marked osteopenic phenotype and lethality in the context of the conditional Ctnnb1 inactivation in osteocytes [Kramer et al., 2010]. Therefore, this model could not be used to test the contribution of Wnt/β-catenin signaling to the phenotype of Notch1 activation in osteocytes.

As an alternative approach to the Ctnnb1 deletion, we elected to reverse the phenotype of the Notch1 activation by the reinstitution of sclerostin levels using Dmp1-SOST transgenics, which exhibit only a modest cancellous and cortical bone phenotype at femoral sites [Rhee et al., 2011; Tu et al., 2012]. Although Dmp1-Cre;RosaNotch;Dmp1-SOST mice were smaller and displayed a reduction in femoral length, the Dmp1-SOST transgene prevented the osteopetrotic phenotype of Dmp1-Cre;RosaNotch mice. This may suggest that downregulation of Sost and upregulation of Wnt signaling was responsible for the skeletal phenotype of Notch1 activation in osteocytes. However, gene expression analysis revealed that the Dmp1-SOST transgene suppressed expression of the RosaNotch allele and Dmp1-Cre activity, the former possibly secondary to suppression of the Rosa26 locus. These events were associated with inhibition of Hey1, 2 and L expression, suggesting that suppression of Notch signaling, and not reinstitution of SOST levels, prevented the Notch1 phenotype.

To determine whether the inhibitory effects of Dmp1-SOST on Dmp1-Cre and Rosa26 expression were due to the presence of the Dmp1-SOST transgene or to an effect of SOST in osteocytes, cortical bone cultures from Dmp1-Cre;RosaNotch and RosaNotch mice were exposed to recombinant human SOST. The validity of this system for the study of Notch and SOST signaling in osteocytes was documented by induction of Notch target genes in cultures from Dmp1-Cre;RosaNotch mice, and by the inhibitory effects of SOST on gene markers of Wnt/β-catenin signaling. It is important to mention that the induction of Wisp1 in bones from Dmp1-Cre;RosaNotch mice in vivo could not be replicated in osteocyte-enriched cultures, indicating that conditions found in vivo are required for induction of Wnt signaling by Notch1 activation in osteocytes [Canalis et al., 2013a]. Although these findings indicate that osteocyte-enriched cultures are not suitable to investigate regulation of Wnt/β-catenin signaling by Notch1, this shortcoming should not alter the interpretation of the direct effects of SOST on Notch signaling and on the activity of the Dmp1-Cre transgene and Rosa26 locus.

Direct addition of SOST to cortical bone cultures did not affect Notch1 Nicd, Rosa26 or Dmp1-Cre expression, suggesting that the suppression of these transcripts in vivo is secondary to genetic interactions among the Dmp1-SOST and Dmp1-Cre transgenes and the RosaNotch allele. SOST increased Hey2 expression and decreased HeyL transcripts modestly and in the context of Notch1 induction, indicating that suppression of Notch signaling by Dmp1-SOST is not the result of a direct effect of SOST in bone cells. These in vitro studies would suggest that the reversal of the skeletal phenotype of Dmp1-Cre;RosaNotch mice is an artifact caused by the Dmp1-SOST transgene in the context of a Dmp1-Cre;RosaNotch genetic composition. The decrease in body weight and femoral length of Dmp1-Cre;RosaNotch;Dmp1-SOST mice is possibly secondary to systemic effects caused by the presence of two Dmp1 driven transgenes, or to the downregulation of Dmp1, since Dmp1 null mice develop rickets, defective bone mineralization and shortening of long bones [Feng et al., 2006].

Similarly to the effect of the Dmp1-SOST transgene, recombinant SOST suppressed Dmp1 transcripts in vitro, but not the Dmp1-Cre transgene. This discrepancy points to transcriptional effects of SOST on the Dmp1 locus, which are mediated by DNA sequences located outside of the 9.6 kb Dmp1 promoter fragment that governs transcription of Cre from the Dmp1-Cre transgene. Post-transcriptional events that increase degradation of the Dmp1 transcripts when bone cells are exposed to SOST also cannot be excluded. It is not possible to determine whether the suppression of Dmp1 in tibiae from Dmp1-Cre;RosaNotch;Dmp1-SOST mice is due to a direct effect of SOST in osteocytes or to a genetic interaction of the Dmp1 locus with the Dmp1-SOST and Dmp1-Cre transgenes. In support of the notion that SOST is indeed responsible for the suppression of Dmp1 in vivo, a recent study demonstrated that administration of a neutralizing SOST antibody to ovariectomized rats increases Dmp1 mRNA levels in tibiae [Stolina et al., 2014]. Dmp1 is required for osteocyte maturation, and Dmp1 null mice exhibit rickets and osteomalacia; consequently, these observations suggest that the detrimental effects of SOST on the skeleton may be mediated in part by a suppression of Dmp1 and subsequent loss of osteocyte function [Feng et al., 2006].

In conclusion, Notch1 overexpression in osteocytes causes osteopetrosis and suppresses Sost expression, and the Dmp1-Cre and Dmp1-SOST transgenics interact in vivo causing artifactual results and phenotypes.

Acknowledgments

The authors thank J. Feng for Dmp1-Cre transgenic mice, C. O’Brien and R. Jilka for the Dmp1-SOST transgenic mice, T. Iso for Hey1 and Hey2 cDNAs, C. Niehrs for Dkk1 cDNA, D. Srivastava for HeyL cDNA, R. Wu for Gapdh cDNA.

Contract grant name: National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS); Contract grant number: AR063049

The abbreviations used are

ATCC
American type culture collection
bp
base pair
BA
bone area
BV/TV
bone volume/total volume
ConnD
connectivity density
Ct.BV/TV
cortical BV/TV
Ctnnb1
cadherin-associated protein beta 1
DMEM
Dulbecco’s modified Eagle’s medium
Dmp1
dentin matrix protein 1
Dkk1
dickkopf-related protein 1
Gapdh
glyceralehyde-3-phosphate dehydrogenase
Hes
hairy enhancer of split
Hey
Hes-related with YRPW motif
kb
kilobase
IDT
integrated DNA technologies
μCT
microcomputed tomography
Nicd
Notch intracellular domain
PCR
polymerase chain reaction
P.endo
endocortical perimeter
P.peri
periosteal perimeter
qRT-PCR
quantitative reverse transcription-PCR
Rankl
receptor activator of nuclear factor kappa B ligand
Rosa26
reverse orientation splice acceptor 26
Rpl38
ribosomal protein L38
SMI
structural model index
TA
total area
Tb.N
trabecular number
Tb.Th
trabecular thickness
Tnfsf11
tumor necrosis factor, member 11
Tnfrsf11b
tumor necrosis factor receptor superfamily 11b
Wisp1
WNT1 inducible signaling pathway protein 1

Footnotes

The authors declare that they have no conflicts of interest with the contents of this article.

References

  • Albers J, Keller J, Baranowsky A, Beil FT, Catala-Lehnen P, Schulze J, Amling M, Schinke T. Canonical Wnt signaling inhibits osteoclastogenesis independent of osteoprotegerin. J Cell Biol. 2013;200:537–549. [PMC free article] [PubMed]
  • Bouxsein ML, Boyd SK, Christiansen BA, Guldberg RE, Jepsen KJ, Muller R. Guidelines for assessment of bone microstructure in rodents using micro-computed tomography. J Bone Miner Res. 2010;25:1468–1486. [PubMed]
  • Brault V, Moore R, Kutsch S, Ishibashi M, Rowitch DH, McMahon AP, Sommer L, Boussadia O, Kemler R. Inactivation of the beta-catenin gene by Wnt1-Cre-mediated deletion results in dramatic brain malformation and failure of craniofacial development. Development. 2001;128:1253–1264. [PubMed]
  • Buchholz F, Ringrose L, Angrand PO, Rossi F, Stewart AF. Different thermostabilities of FLP and Cre recombinases: implications for applied site-specific recombination. Nucleic Acids Res. 1996;24:4256–4262. [PMC free article] [PubMed]
  • Canalis E. Wnt signalling in osteoporosis: mechanisms and novel therapeutic approaches. Nat Rev Endocrinol. 2013;9:575–583. [PubMed]
  • Canalis E, Adams DJ, Boskey A, Parker K, Kranz L, Zanotti S. Notch Signaling in Osteocytes Differentially Regulates Cancellous and Cortical Bone Remodeling. J Biol Chem. 2013a;288:25614–25625. [PMC free article] [PubMed]
  • Canalis E, Parker K, Feng JQ, Zanotti S. Osteoblast Lineage-specific Effects of Notch Activation in the Skeleton. Endocrinology. 2013b;154:623–634. [PubMed]
  • Chen J, Long F. beta-catenin promotes bone formation and suppresses bone resorption in postnatal growing mice. J Bone Miner Res. 2013;28:1160–1169. [PMC free article] [PubMed]
  • Dougall WC, Glaccum M, Charrier K, Rohrbach K, Brasel K, De ST, Daro E, Smith J, Tometsko ME, Maliszewski CR, Armstrong A, Shen V, Bain S, Cosman D, Anderson D, Morrissey PJ, Peschon JJ, Schuh J. RANK is essential for osteoclast and lymph node development. Genes Dev. 1999;13:2412–2424. [PubMed]
  • Feng JQ, Ward LM, Liu S, Lu Y, Xie Y, Yuan B, Yu X, Rauch F, Davis SI, Zhang S, Rios H, Drezner MK, Quarles LD, Bonewald LF, White KE. Loss of DMP1 causes rickets and osteomalacia and identifies a role for osteocytes in mineral metabolism. Nat Genet. 2006;38:1310–1315. [PMC free article] [PubMed]
  • Gibson DG, Young L, Chuang RY, Venter JC, Hutchison CA, 3rd, Smith HO. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods. 2009;6:343–5. [PubMed]
  • Glass DA, Bialek P, Ahn JD, Starbuck M, Patel MS, Clevers H, Taketo MM, Long F, McMahon AP, Lang RA, Karsenty G. Canonical Wnt signaling in differentiated osteoblasts controls osteoclast differentiation. Dev Cell. 2005;8:751–764. [PubMed]
  • Glatt V, Canalis E, Stadmeyer L, Bouxsein ML. Age-Related Changes in Trabecular Architecture Differ in Female and Male C57BL/6J Mice. J Bone Miner Res. 2007;22:1197–1207. [PubMed]
  • Glinka A, Wu W, Delius H, Monaghan AP, Blumenstock C, Niehrs C. Dickkopf-1 is a member of a new family of secreted proteins and functions in head induction. Nature. 1998;391:357–362. [PubMed]
  • Halleux C, Kramer I, Allard C, Kneissel M. Isolation of mouse osteocytes using cell fractionation for gene expression analysis. Methods Mol Biol. 2012;816:55–66. [PubMed]
  • Iso T, Sartorelli V, Chung G, Shichinohe T, Kedes L, Hamamori Y. HERP, a new primary target of Notch regulated by ligand binding. Mol Cell Biol. 2001;21:6071–6079. [PMC free article] [PubMed]
  • Kalajzic I, Braut A, Guo D, Jiang X, Kronenberg MS, Mina M, Harris MA, Harris SE, Rowe DW. Dentin matrix protein 1 expression during osteoblastic differentiation, generation of an osteocyte GFP-transgene. Bone. 2004;35:74–82. [PubMed]
  • Kamel MA, Picconi JL, Lara-Castillo N, Johnson ML. Activation of beta-catenin signaling in MLO-Y4 osteocytic cells versus 2T3 osteoblastic cells by fluid flow shear stress and PGE2: Implications for the study of mechanosensation in bone. Bone. 2010;47:872–881. [PMC free article] [PubMed]
  • Kovall RA. More complicated than it looks: assembly of Notch pathway transcription complexes. Oncogene. 2008;27:5099–5109. [PubMed]
  • Kramer I, Halleux C, Keller H, Pegurri M, Gooi JH, Weber PB, Feng JQ, Bonewald LF, Kneissel M. Osteocyte Wnt/beta-catenin signaling is required for normal bone homeostasis. Mol Cell Biol. 2010;30:3071–3085. [PMC free article] [PubMed]
  • Li J, Sarosi I, Yan XQ, Morony S, Capparelli C, Tan HL, McCabe S, Elliott R, Scully S, Van G, Kaufman S, Juan SC, Sun Y, Tarpley J, Martin L, Christensen K, McCabe J, Kostenuik P, Hsu H, Fletcher F, Dunstan CR, Lacey DL, Boyle WJ. RANK is the intrinsic hematopoietic cell surface receptor that controls osteoclastogenesis and regulation of bone mass and calcium metabolism. Proc Natl Acad Sci U S A. 2000;97:1566–1571. [PubMed]
  • Lu Y, Xie Y, Zhang S, Dusevich V, Bonewald LF, Feng JQ. DMP1-targeted Cre expression in odontoblasts and osteocytes. J Dent Res. 2007;86:320–325. [PubMed]
  • Monroe DG, McGee-Lawrence ME, Oursler MJ, Westendorf JJ. Update on Wnt signaling in bone cell biology and bone disease. Gene. 2012;492:1–18. [PMC free article] [PubMed]
  • Murtaugh LC, Stanger BZ, Kwan KM, Melton DA. Notch signaling controls multiple steps of pancreatic differentiation. Proc Natl Acad Sci U S A. 2003;100:14920–14925. [PubMed]
  • Nakagawa O, Nakagawa M, Richardson JA, Olson EN, Srivastava D. HRT1, HRT2, and HRT3: a new subclass of bHLH transcription factors marking specific cardiac, somitic, and pharyngeal arch segments. Dev Biol. 1999;216:72–84. [PubMed]
  • Nakashima T, Hayashi M, Fukunaga T, Kurata K, Oh-Hora M, Feng JQ, Bonewald LF, Kodama T, Wutz A, Wagner EF, Penninger JM, Takayanagi H. Evidence for osteocyte regulation of bone homeostasis through RANKL expression. Nat Med. 2011;17:1231–1234. [PubMed]
  • Nam Y, Sliz P, Song L, Aster JC, Blacklow SC. Structural basis for cooperativity in recruitment of MAML coactivators to Notch transcription complexes. Cell. 2006;124:973–983. [PubMed]
  • Nazarenko I, Lowe B, Darfler M, Ikonomi P, Schuster D, Rashtchian A. Multiplex quantitative PCR using self-quenched primers labeled with a single fluorophore. Nucleic Acids Res. 2002a;30:e37. [PMC free article] [PubMed]
  • Nazarenko I, Pires R, Lowe B, Obaidy M, Rashtchian A. Effect of primary and secondary structure of oligodeoxyribonucleotides on the fluorescent properties of conjugated dyes. Nucleic Acids Res. 2002b;30:2089–2195. [PMC free article] [PubMed]
  • Ominsky MS, Stolina M, Li X, Corbin TJ, Asuncion FJ, Barrero M, Niu QT, Dwyer D, Adamu S, Warmington KS, Grisanti M, Tan HL, Ke HZ, Simonet WS, Kostenuik PJ. One year of transgenic overexpression of osteoprotegerin in rats suppressed bone resorption and increased vertebral bone volume, density, and strength. J Bone Miner Res. 2009;24:1234–1246. [PubMed]
  • Rhee Y, Allen MR, Condon K, Lezcano V, Ronda AC, Galli C, Olivos N, Passeri G, O’Brien CA, Bivi N, Plotkin LI, Bellido T. PTH receptor signaling in osteocytes governs periosteal bone formation and intra-cortical remodeling. J Bone Miner Res. 2011;26:1035–1046. [PMC free article] [PubMed]
  • Sato MM, Nakashima A, Nashimoto M, Yawaka Y, Tamura M. Bone morphogenetic protein-2 enhances Wnt/beta-catenin signaling-induced osteoprotegerin expression. Genes Cells. 2009;14:141–153. [PubMed]
  • Sauer B, Henderson N. Site-specific DNA recombination in mammalian cells by the Cre recombinase of bacteriophage P1. Proc Natl Acad Sci U S A. 1988;85:5166–5170. [PubMed]
  • Schroeter EH, Kisslinger JA, Kopan R. Notch-1 signalling requires ligand-induced proteolytic release of intracellular domain. Nature. 1998;393:382–386. [PubMed]
  • Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang MS, Luthy R, Nguyen HQ, Wooden S, Bennett L, Boone T, Shimamoto G, DeRose M, Elliott R, Colombero A, Tan HL, Trail G, Sullivan J, Davy E, Bucay N, Renshaw-Gegg L, Hughes TM, Hill D, Pattison W, Campbell P, Sander S, Van G, Tarpley J, Derby P, Lee R, Boyle WJ. Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell. 1997;89:309–319. [PubMed]
  • Stanger BZ, Datar R, Murtaugh LC, Melton DA. Direct regulation of intestinal fate by Notch. Proc Natl Acad Sci U S A. 2005;102:12443–12448. [PubMed]
  • Stolina M, Dwyer D, Niu QT, Villasenor KS, Kurimoto P, Grisanti M, Han CY, Liu M, Li X, Ominsky MS, Ke HZ, Kostenuik PJ. Temporal changes in systemic and local expression of bone turnover markers during six months of sclerostin antibody administration to ovariectomized rats. Bone. 2014;67:305–13. [PubMed]
  • Tatsumi S, Ishii K, Amizuka N, Li M, Kobayashi T, Kohno K, Ito M, Takeshita S, Ikeda K. Targeted ablation of osteocytes induces osteoporosis with defective mechanotransduction. Cell Metab. 2007;5:464–475. [PubMed]
  • Tso JY, Sun XH, Kao TH, Reece KS, Wu R. Isolation and characterization of rat and human glyceraldehyde-3-phosphate dehydrogenase cDNAs: genomic complexity and molecular evolution of the gene. Nucleic Acids Res. 1985;13:2485–2502. [PMC free article] [PubMed]
  • Tu X, Rhee Y, Condon KW, Bivi N, Allen MR, Dwyer D, Stolina M, Turner CH, Robling AG, Plotkin LI, Bellido T. Sost downregulation and local Wnt signaling are required for the osteogenic response to mechanical loading. Bone. 2012;50:209–217. [PMC free article] [PubMed]
  • Turner CH, Pavalko FM. Mechanotransduction and functional response of the skeleton to physical stress: the mechanisms and mechanics of bone adaptation. J Orthop Sci. 1998;3:346–355. [PubMed]
  • Wei W, Zeve D, Suh JM, Wang X, Du Y, Zerwekh JE, Dechow PC, Graff JM, Wan Y. Biphasic and dosage-dependent regulation of osteoclastogenesis by beta-catenin. Mol Cell Biol. 2011;31:4706–4719. [PMC free article] [PubMed]
  • Wilson JJ, Kovall RA. Crystal structure of the CSL-Notch-Mastermind ternary complex bound to DNA. Cell. 2006;124:985–996. [PubMed]
  • Xiong J, Onal M, Jilka RL, Weinstein RS, Manolagas SC, O’Brien CA. Matrix-embedded cells control osteoclast formation. Nat Med. 2011;17:1235–1241. [PMC free article] [PubMed]
  • Zanotti S, Canalis E. Notch and the Skeleton. Mol Cell Biol. 2010;30:886–896. [PMC free article] [PubMed]
  • Zanotti S, Canalis E. Notch signaling in skeletal health and disease. Eur J Endocrinol. 2013;168:R95–R103. [PMC free article] [PubMed]
  • Zanotti S, Canalis E. Notch1 and Notch2 expression in osteoblast precursors regulates femoral microarchitecture. Bone. 2014;62:22–28. [PMC free article] [PubMed]
  • Zanotti S, Smerdel-Ramoya A, Canalis E. Reciprocal regulation of notch and nuclear factor of activated T-cells (NFAT)c1 transactivation in osteoblasts. J Biol Chem. 2011;286:4576–4588. [PMC free article] [PubMed]