High-resolution mapping of a novel rat blood pressure locus on chromosome 9 to a region containing the Spp2 gene and colocalization of a QTL for bone mass

Abstract

Through linkage analysis of the Dahl salt-sensitive (S) rat and the spontaneously hypertensive rat (SHR), a blood pressure (BP) quantitative trait locus (QTL) was previously located on rat chromosome 9. Subsequent substitution mapping studies of this QTL revealed multiple BP QTLs within the originally identified logarithm of odds plot by linkage analysis. The focus of this study was on a 14.39 Mb region, the distal portion of which remained unmapped in our previous studies. High-resolution substitution mapping for a BP QTL in the setting of a high-salt diet indicated that an SHR-derived congenic segment of 787.9 kb containing the gene secreted phosphoprotein-2 (Spp2) lowered BP and urinary protein excretion. A nonsynonymous G/T polymorphism in the Spp2 gene was detected between the S and S.SHR congenic rats. A survey of 45 strains showed that the T allele was rare, being detected only in some substrains of SHR and WKY. Protein modeling prediction through SWISSPROT indicated that the predicted protein product of this variant was significantly altered. Importantly, in addition to improved cardiovascular and renal function, high salt-fed congenic animals carrying the SHR T variant of Spp2 had significantly lower bone mass and altered bone microarchitecture. Total bone volume and volume of trabecular bone, cortical thickness, and degree of mineralization of cortical bone were all significantly reduced in congenic rats. Our study points to opposing effects of a congenic segment containing the prioritized candidate gene Spp2 on BP and bone mass.

blood pressure (bp) is a complex trait with a strong genetic component. Identification of genomic factors conferring susceptibility to the development of high BP will be important for the development of better clinical strategies to manage hypertension, which is a leading cause for cardiovascular diseases and related mortality. Methods to identify genomic factors on the human genome that cause hypertension are limited to linkage analyses or association studies (30, 35, 39, 41, 67). While linkage analyses have the limitation of identifying only large genomic segments with several candidate genetic elements, association studies have the limitation that they do not necessarily identify inherited variants and also lack the ability to differentiate between causal inherited factors and noncausal linked loci.

Rat genetic models are used as valuable alternatives to understand the genetic underpinnings of complex traits, including BP (2, 8, 12, 14, 15, 28, 29, 43, 47–49, 53, 54, 61). The ability to apply substitution mapping as a strategy beyond linkage or association studies has been advantageous to further dissect and map the loci detected through linkage analyses. In our laboratory, we have applied substitution mapping using congenic strains to locate and fine-map regions of the hypertensive Dahl salt-sensitive (S) rat genome that are causally linked to the alteration in blood pressure (16, 18–20, 26, 27, 34, 37, 45, 52, 55, 57, 58, 64, 65). The background for this current report is one such mapping study on rat chromosome 9 (65). We previously reported a linkage analysis followed by several iterations of substitution mapping that point to multiple regions of rat chromosome 9 as BP QTLs (quantitative trait loci) (17, 34, 65). The purpose of the current study was to construct and characterize a new iteration of congenic strains with shorter introgressed segments developed from a congenic strain [S.SHR(9)x10A, spanning 14.39 Mb], which demonstrated a BP-lowering effect in rats fed with a high salt diet (65). Although a BP QTL was located within a shorter 2.11 Mb region within and toward the p-terminus of the 14.39 Mb segment, the remainder of the 14.39 Mb region remained unmapped, which raises the question of whether there are additional unmapped BP loci within the 14.39 Mb region. The interest in addressing this question was, in part, due to a recent GWAS (genome-wide association study) report on the association of BP and a homologous region of this segment on human chromosome 2q37 for BP (22).

MATERIAL AND METHODS

Animals.

All procedures involving animals described in this study were approved by the University of Toledo Institutional Animal Care and Use Committee. Dahl salt-sensitive rats (SS/Jr) were from our colony and will be referred to as S. Spontaneously hypertensive rats (SHR/NHsd) were from Harlan Sprague-Dawley (Indianapolis, IN) and will be referred to as SHR. The congenic substrains reported here were derived from S.SHR(9)X10A, which was reported previously from our laboratory (17, 34, 65). In brief, the parental congenic strain was crossed with S to generate a population of F1 rats. The F1 rats were intercrossed to obtain an F2 population, which was genotyped using microsatellite markers throughout the region on rat chromosome 9 from marker D9Rat7 (Chr9:83453664) to D9Mco93 (Chr9:97845199,Rnor 5.0) (Fig. 1A). The recombinant F2 rats with various introgressed regions of the SHR alleles were backcrossed to S, sorted by genotyping, and intercrossed to obtain homozygous congenic animals on the genomic background of the S rat. The substrains obtained are labeled numerically as S.SHR(9) strains 1 through 6.

Fig. 1.

A: schematic map of S.SHR(9) new congenic substrains. The long orange bar at the top with 2 black arrows at both ends denotes the limits of the previously mapped quantitative trait locus (QTL) region as well as the physical limits of the parental strain S.SHR(9)X10A. Below this orange bar, the physical map of rat chromosome 9 is shown along with microsatellite markers and their locations in both versions of the rat genome Rnor 5.0 and Rnor 3.4. The new congenic strains are represented by green or black bars with white open boxes showing regions within which recombination occurred. The color green indicates significant blood pressure (BP)-lowering effect of the introgressed spontaneously hypertensive rat (SHR) segment compared with Dahl salt sensitive (S). Black bars indicate no significant difference in BP between S and congenic strains. S (telemetry: n = 6–10 and n = 20 for tail cuff data); S.SHR(9) strain 1 (n = 7); S.SHR(9) strain 2 (n = 8); S.SHR(9) strain 3 (n = 6); S.SHR(9) strain 4 (n = 6); S.SHR(9) strain 5 (n = 20, tail cuff data); S.SHR(9) strain 6 (n = 6). *P < 0.05 or more; ns, not significant. The orange bar at the bottom denotes the deduced current localization of the BP QTL to 787.9 kb (Rnor 5.0). The gray bars below the bottom orange bar show the homologous regions in mouse and human, with base pair numbers of their location on chromosomes 1 and 2, respectively. Location of Spp2 is shown by the red line. B: radiotelemetry BP measurements of S vs. S.SHR(9) congenic substrains. Top: systolic BP (SBP); middle: diastolic BP (DBP); bottom: mean arterial pressure (MAP). Black, S (telemetry: n = 6–10); red, S.SHR(9) strain 1 (n = 7); yellow-green, S.SHR(9) strain 2 (n = 8); green, S.SHR(9) strain 3 (n = 6); blue, S.SHR(9) strain 4 (n = 6). Note: DBP of strain 4 was highly variable, perhaps suggesting lack of power to detect a difference compared with the S.

Microsatellite markers.

New microsatellite markers were developed from sequences downloaded from the Ensembl database (Rnor 3.4) (http://www.ensembl.org). Sequence information on the newly identified polymorphic markers with the prefix D9Mco is provided in Supplemental Table S1.¹ In brief, primers were designed around dinucleotide or trinucleotide repeats using Primer 3 software (http://bioinfo.ut.ee/primer3-0.4.0/primer3/). Primer pairs were tested for polymorphisms between S and SHR genomic DNA. Those that were polymorphic were used for genotyping F2, backcross, and intercross animals along with DNA from S and SHR as controls.

Genotyping.

Genomic DNA was extracted from tail biopsies using the WizardSV 96 Genomic DNA Purification System (Promega, Madison, WI). PCR-based genotyping with microsatellite markers was performed by techniques described previously (56).

BP measurement.

Congenic substrains were raised concomitantly with a group of control S rats, weight-matched, and tested for BP compared with the S rats. Rats were weaned at 28–30 days of age and fed a low-salt (0.3% NaCl) Harlan Teklad diet. At 40–42 days of age, male rats were housed one S and one congenic per cage and fed with 2% NaCl diet for the rest of their life. Tail cuff-based BP readings were obtained using the CODA tail-cuff BP system from Kent Scientific. BP was also measured by radiotelemetry. At age 64–66 days, rats housed in the same manner as described prior to the tail-cuff were surgically implanted with C-40 transmitters (Data Sciences International, St. Paul, MN), transferred to a clean cage, and thereafter housed individually. Rats were allowed to recover from surgery for 4 days before the transmitters were turned on for recording BP (20). Rats undergoing the telemetry measurements were S and the S.SHR(9) congenic strains 1, 2, 3, 4, and 6.

RNA isolation and RT-PCR analysis.

Kidneys from S and congenic rats were isolated and homogenized in TRIzol reagent (Invitrogen) for extracting RNA. We used 1 μg of total RNA to obtain cDNA by reverse transcription with SuperScript III kit (Invitrogen). Levels of mRNA expression of Spp2, Trpm8, and Arl4c were analyzed by real-time PCR (Applied Biosystems), and expression levels relative to the expression of rRNA were calculated by the 2^−ΔΔCT method (40).

Urinary protein excretion.

After BP measurement, we measured total protein in urine collected over a 24 h period. In brief, 80-day old male rats were caged individually in metabolic cages (Lab Products, Seaford, DE) with free access to water. Urine was collected in the presence of 0.01% sodium azide, and the urine volume was recorded. Total protein level was determined by Quan Ttest Red Total Protein Assay system (Quantimetrix, Redondo Beach, CA).

Genomic sequence analysis.

Sequence variation data between the two strains, SS/Jr(ICL) and SHR/NHsd(ICL), were extracted from RGD (http://rgd.mcw.edu). For manual sequencing, primers were synthesized by Integrated DNA Technologies (Coralville, IA). PCR products were sequenced through the commercial sequencing service provided by Eurofins MWG Operon (Huntsville, AL).

Microcomputed tomography analysis.

Right tibia bones of 66 to 70 day old male rats, which were weaned on day 28 and fed a high-salt (2% NaCl) diet, were isolated and fixed in 10% buffered formalin. Microcomputed tomography (microCT) analysis was performed using a SCANCO μCT35 (SCANCO Medical, Bassersdorf, Switzerland) equipped with a 20 mm focal spot microfocus X-ray tube. Scans were performed with the following instrument settings: E = 70 KVp, I = 110 uA, increment 7 um, threshold value = 220. Slices of the proximal (n = 340) and the midshaft tibia (n = 57) were used for the trabecular and cortical bone analysis, respectively (25). The analysis of trabecular bone microstructure and the cortical parameters was conducted using Evaluation Program V6.5–1 (Scanco Medical, Bruettisellen, Switzerland) and conformed to recommended guidelines (10).

Protein secondary structure modeling.

Secondary structure of protein was predicted using SWISS-MODEL (http://swissmodel.expasy.org/) (3, 6, 21, 32).

Statistical analyses.

SPSS17 was used for all analyses. Student's t-test was used to compare two groups, whereas one-way ANOVA was used for studies with more than two groups followed by multicomparison carried out by least significant difference.

RESULTS

Positional cloning of a novel BP QTL to 787.9 kb.

A panel of six new congenic substrains was developed using the congenic strain S.SHR(9)X10A, reported in our previous study (65). The new panel of congenic substrains had varying sizes of rat chromosome 9 from the SHR introgressed into the genetic background of the S rat (Fig. 1A). The systolic BP of four of the congenic strains 1, 2, 3, and 4 (shown in green in Fig. 1A) was significantly lower than the S (control) background strain (Fig. 1B and Table 1). Systolic BP of the congenic strains 5 and 6 (strains shown in black in Fig. 1A) was not significantly lower than that of the S rats (Table 1). Collectively, these data provide evidence to support the interpretation that the distal portion of the 14.39 Mb introgressed congenic region of S.SHR(9)X10A consists of a new systolic BP QTL, which is located within a 787.9 kb region from 95,117,755-95,905,612 (Rnor 5.0) on chromosome 9 (Fig. 1A). In the context of the previously reported mapping of the BP QTLs on chromosome 9, the newly identified location is distal to the two previously identified BP QTLs, BP QTL1 (mapped to 2.06 Mb) and BP QTL2 (mapped to 2.11 Mb) (Fig. 2). The newly identified systolic QTL region is therefore named BP QTL region 3.

Table 1.

Blood pressure and other hemodynamic parameters of S and congenic strains

Strain	S	Congenic	Effect	P Value
SBP, mmHg
S.SHR(9) strain 1	213 [2.05]	157 [0.82]	−56 [2]	0.001
S.SHR(9) strain 2	201 [7.40]	171 [7.24]	−30 [11]	0.012
S.SHR(9) strain 3	203 [4.17]	164 [5.22]	−39 [10]	0.001
S.SHR(9) strain 4	203 [4.17]	178 [4.17]	−24 [10]	0.023
S.SHR(9) strain 5*	195 [4.72]	201 [2.65]	6 [5]	0.217
S.SHR(9) strain 6	203 [4.17]	193 [11.25]	−10 [10]	0.339
DBP, mmHg
S.SHR(9) strain 1	154 [1.59]	115 [0.67]	−39 [2]	0.001
S.SHR(9) strain 2	152 [6.49]	128 [5.67]	−24 [9]	0.017
S.SHR(9) strain 3	149 [3.55]	126 [2.70]	−24 [11]	0.043
S.SHR(9) strain 4	149 [3.55]	140 [8.15]	−9 [11]	0.404
S.SHR(9) strain 5	n/a
S.SHR(9) strain 6	149 [3.55]	153 [12.29]	3 [11]	0.762
MAP, mmHg
S.SHR(9) strain 1	174 [1.74]	129 [0.72]	−45 [1.88]	0.001
S.SHR(9) strain 2	175 [6.80]	149 [6.40]	−27 [10]	0.013
S.SHR(9) strain 3	175 [3.75]	145 [3.41]	−31 [10]	0.005
S.SHR(9) strain 4	175 [3.75]	159 [6.08]	−17 [10]	0.109
S.SHR(9) strain 5	n/a
S.SHR(9) strain 6	175 [3.75]	172 [11.48]	3[10]	0.76
Pulse Pressure, mmHg
S.SHR(9) strain 1	59.3 [0.55]	42.2 [0.22]	−17 [0.6]	0.001
S.SHR(9) strain 2	50.4 [2.35]	44.1 [2.47]	−6 [3]	0.085
S.SHR(9) strain 3	53.4 [1.90]	38.3 [4.21]	−15 [8]	0.062
S.SHR(9) strain 4	53.4 [1.90]	38.2 [6.67]	−15 [8]	0.06
S.SHR(9) strain 5	n/a
S.SHR(9) strain 6	53.4 [1.90]	40.3 [7.06]	−13 [8]	0.101
Heart Rate, beats/min
S.SHR(9) strain 1	395 [6.21]	364 [6.71]	−31 [9]	0.001
S.SHR(9) strain 2	397 [9.01]	384 [6.18]	−13 [12]	0.259
S.SHR(9) strain 3	394 [5.97]	390 [4.83]	4 [8]	0.681
S.SHR(9) strain 4	394 [5.97]	399 [6.89]	5 [8]	0.509
S.SHR(9) strain 5	n/a
S.SHR(9) strain 6	394 [5.97]	400 [4.24]	6 [8]	0.431

All values are expressed as means ± SE.

^*Blood pressure (BP) data reported in this table were collected by the telemetry method except for S.SHR(9) strain 5, which is tail-cuff data. S, Dahl salt sensitive; SHR, spontaneously hypertensive rat; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; n/a, not available. The following S.SHR (9) congenic substrains were concomitantly raised and tested together for BP along with S rats (n = 20 tail cuff, n = 6–10 for telemetry): S.SHR(9) strain 1(n = 7); S.SHR(9) strain 2(n = 8); S.SHR(9) strain 3(n = 6), S.SHR(9) strain 4(n = 6) and S.SHR(9) strain 6(n = 6); S.SHR(9) strain 5 (n = 20).

Fig. 2.

Comprehensive substitution mapping results on chromosome 9 (RNO9) for BP QTL obtained using S.SHR(9) congenic strains. At the top is the logarithm of odds (LOD) plot obtained for BP from linkage analysis of an F2 (S × SHR) population (17). The x-axis used is the cytogenetic map of RNO9. The SHR allele at this QTL had a BP-lowering effect. The green and black horizontal bars shown below the LOD plot represent the introgressed segments of S.SHR(9) congenic strains. References for the studies reporting the strains are Toland et al. 2008 (65) and Kumarasamy et al. 2011 (34) and the current study. The strains being reported in the current study are numbered strain 1–6, and the corresponding strain names designated in the laboratory are given alongside these strains with the prefix S.SHR(9)x10Ax6x. The end markers and their locations on RNO9 in Rnor 5.0 are listed for the congenic strains. Green indicates that the BP of the congenic strain was lower than that of the S, and black indicates that the BP of the congenic strain was not significantly different from the BP of the S. Multiple locations mapped thus far from the substitution mapping of the initial single BP QTL identified by linkage analysis are highlighted by the vertical purple bars ending in red and labeled in yellow highlight. Note: *BP QTL 2 region was redefined from a previous publication (24); the QTL region was defined here based on the BP effect alone (regardless of the urinary protein excretion effect).

Rat genome version 5.0 (Rnor 5.0) (http://www.ensembl.org) indicates that the only complete annotation within the 787.9 kb region defined above is that of the protein-coding gene Spp2 (Fig. 1A). Exonic sequencing revealed a single G/T polymorphism between the S and SHR alleles within exon 4 of Spp2 (Fig. 3A). This variation translates to a substitution at amino acid 125 within the polypeptide chain of Spp2 from Alanine in S rats to Serine in SHR. As shown in Fig. 3B, as a result of the SHR allelic variant of Spp2, the predicted protein structure of Spp2 is significantly altered. Prominent among the structural alterations are the alpha helices, the locations and conformations of which are altered in the SHR compared with S (circled in red in Fig. 3B). Interestingly, the rat genome sequence data available for the location of the Spp2 G/T polymorphism at Chr9:87301858 (Rnor 3.4), Chr9:95192812 (Rnor 5.0), showed that 45 strains had the G allele, whereas SHR from three different sources [SHR/NHsd (ICL), SHR/Olalpcv (KNAW), SHR/Olalpcv (ICL)] and one of the WKY strains, WKY/Gcrc (ICL), were the only strains that had the T allele (http://www.rgd.mcw.edu, Fig. 3C). The origin of the SHR used in the present study (SHR/NHsd) is one of those strains carrying the T allele.

Fig. 3.

A: sequencing results from the PCR products for both S and SHR at the location Chr9:87301858 (Rnor 3.4), Chr9:95192812 (Rnor 5.0). B: secondary structure of Spp2 in S and SHR rats. Protein modeling prediction was done by SWISS-MODEL (http://swissmodel.expasy.org). C: list of nucleic acid at the location Chr9:87301858 (Rnor 3.4) of rat strains listed on the Rat genome database (RGD website: http://rgd.mcw.edu/).

Urinary protein excretion.

In addition to the BP measurements, urinary protein excretion (UPE) of the congenic strains was examined. Compared with the S rats, all of the four strains that had lower BP (strains 1–4 shown in Fig. 1A) also had significantly lower amounts of protein in their urine (Fig. 4). S.SHR(9) strain 5 and S.SHR(9) strain 6, which did not show a BP-lowering effect compared with S, also did not differ in their urinary protein content compared with S (Fig. 4).

Fig. 4.

Total urinary protein excretion (UPE) of the S (n = 8) S.SHR(9) congenic substrains strain 1 (n = 26), strain 2 (n = 12), strain 3 (n = 6), strain 4 (n = 6), strain 5 (n = 8), strain 6 (n = 6). Values obtained from S.SHR(9) congenic substrains were normalized to that of S rats. Green indicates lower UPE than S; black indicates UPE of the strain not statistically significant compared with S. Errors bars represent SE; *P < 0.05.

Bone analysis.

To evaluate the structure of bones, tibia from S and S.SHR(9) strain 3 rats were examined by microCT imaging. As seen in Fig. 5, A–F, significant morphological differences were observed between the tibia of S and S.SHR(9) strain 3 rats, indicating that there was less bone mineralized tissue in the S.SHR(9) strain 3 rats. Quantitatively, S.SHR(9) strain 3 had less of both the trabecular bone (Fig. 5G) and the cortical bone (Fig. 5H) mass compared with S strain. The ratio of bone volume to total volume, trabecular thickness, and connectivity density were lower in the S.SHR(9) strain 3 compared with S (Fig. 5G). Similarly, there was a significant reduction of cortical bone area and tissue mineral density of cortical bone in the S.SHR(9) strain 3 compared with S rats (Fig. 5, H and I, Supplemental Tables S2 and S3). The space occupied by the growth plate was larger in S.SHR(9) strain 3, suggesting the delayed process of endochondral mineralization.

Fig. 5.

Alteration in bone mass and structure associated with Spp2 polymorphism in S (n = 3) and S.SHR (9) strain 3 rats (n = 3). Rendered microcomputed tomography (mCT) images of the tibia from S (A, B, C) and S.SHR(9) strain 3 (D, E, F) rats. Longitudinal 0.6 mm sections of proximal tibia (A, D) taken along medial plane, whole proximal tibia (B, E), and 0.6 mm cross-sections of tibia metaphysis taken 2.1 mm down from the epiphysis edge (C, F). Bar on mCT renderings represents 1 mm. Trabecular bone morphometric parameters of proximal tibia from S and S.SHR(9) strain 3 rats (G), cortical bone parameters at tibia midshaft (H), mineralization gradient of the proximal tibia measured over the 10 mm length starting at articular surface (I). Black triangles, S strain; gray circles, S.SHR(9) strain 3. BV/TV, bone volume/total volume; Tb.Th, trabecular thickness; Conn. D, connectivity density; Ct.Ar, cortical bone area; TMD, tissue mineral density; HA, hydroxyapatite. Errors bars represent SD; *P < 0.05.

DISCUSSION

This study has resulted in identifying a new BP QTL on chromosome 9. The shortest and most definitive region for the location of this new BP QTL is the introgressed segment to which the BP-lowering effect of the SHR alleles on the genomic background of the S rat can be traced. This is a 2.09 Mb region spanned by the congenic strain S.SHR(9) strain 4 (Fig. 1A). Furthermore, by using the differential congenic segment approach, we can pinpoint the most likely location of the BP QTL to a 787.9 kb region by eliminating the overlapping segments of S.SHR(9) strain 4 with the two strains that did not demonstrate the BP-lowering effect, i.e., S.SHR(9) strain 5 and S.SHR(9) strain 6. This prioritized region contains the distal portion of the gene Trpm8 and the complete annotation of the gene Spp2. The gene Arl4c is in the near vicinity of the QTL region. All three genes, Spp2, Trpm8, and Arl4c, were evaluated for nonsynonymous variants, and only Spp2 had one, and renal gene expression was not statistically significantly differential for any of the genes except for an increased expression of Spp2 in the congenic strain 4 compared with S (Fig. 6). These data lead to the prioritization of Spp2 as a candidate gene for the newly identified BP QTL. It should be emphasized, however, that despite the prioritization of the Spp2 locus for further study, all the variants within the entire congenic segment, irrespective of whether they are within coding genes or within noncoding regions, remain as candidates for both the BP and the bone mass phenotypes.

Fig. 6.

Renal gene expression of Spp2, Arl4c, and Trpm8 mRNA expression. RNA from kidney homogenates of S (n = 3) and S.SHR(9) congenic strain 4 (n = 3) were reverse-transcribed and quantitated by real-time PCR as described in material and methods. Bars represent the fold change of Spp2, Arl4c, and Trpm8 mRNA of the congenic strain relative to S. Errors bars represent SE; *P < 0.05.

While we were mapping for the BP QTL reported in this study, a GWAS reported the association of a region on human chromosome 2q37 to systolic BP responses to cold pressor test, which was also previously demonstrated by the same group to be associated with salt sensitivity (13, 44). This region on human chromosome 2q37 is homologous to the rat BP QTL reported in the current study and near the human TRPM8 and SPP2 loci. Besides the homology, an interesting feature related to salt sensitivity in our study is that the BP QTL reported is mapped between a salt-sensitive hypertensive strain and a relatively salt-insensitive hypertensive rat, the SHR.

The candidate gene, Secreted phosphoprotein-2 (Spp2), also known as secreted phosphoprotein-24 kDa (Spp24), has not been studied in the context of cardiovascular physiology but has been studied in the context of skeletal physiology (9) and in normal bone growth and turnover (11, 50, 51). Spp2 is a bone matrix protein that is produced primarily in the liver and delivered to other tissues for function, especially the bone (1, 11, 24). The function of this protein in the normal bone environment is not clearly delineated. In a study of dizygous twins in the UK, Wilson et al. (69) reported a QTL on human 2q33-37 for bone quality. In this study, SPP2 is a candidate gene, albeit one among a large number of candidate genes. Furthermore, Hsu et al. (23) conducted a GWAS for bone mineral density in three populations and identified a SNP, rs12151790, with SPP2 as the nearest gene listed on 2q37.1. Thus it was intriguing for us to test the possibility that the region that we were mapping for BP could also be a QTL for bone quality. Therefore, we conducted a comparative microCT study of the tibia from high salt-fed S and congenic strain S.SHR(9) strain 3. The choice of S.SHR(9) strain 3 was entirely based on availability of the strain at the time. The data obtained suggest that bone quality indeed mapped to the introgressed 2.48 Mb segment of S.SHR(9) strain 3. Due to the fact that we did not conduct microCT studies on bone quality in the S.SHR(9) strains 5 and 6 (due to cost as a factor), further studies will be required to conclude that the 787.9 kb region prioritized as a BP QTL is also colocalized with a QTL for bone quality. However, given the knowledge on the function of Spp2 in bone physiology, the human mapping data for both BP and bone quality, and the rat mapping data encompassing the Spp2 locus, further consideration of Spp2 as a pleiotropic locus for BP and bone quality is reasonable. We emphasize, however, that the defined congenic segment for BP is shorter than the overlapping segment for bone quality.

The rat Spp2 gene comprises eight exons and seven introns, spanning ∼27 kb DNA. The ATG start codon is located in the 1st exon and the TAA stop codon in the penultimate exon, the final exon coding exclusively for the 3′-untranslated region. The cystatin-like domain of Spp2 is encoded by the first four exons of the gene, while the noncystatin-like region is encoded by exons 5–7 (5). The cystatin domain in Spp2 does not contain the entire consensus sequence required for cystatin activity (33), and Spp2 does not inhibit the activities of cathepsin B or K (66). The SHR allelic variant of Spp2 is located within exon 4 and is predicted to cause structural alterations within the protein product of Spp2. Because Spp2 lacks cystatin activity, it is possible that the likely functional consequence as a result of this predicted structural alteration on the skeletal as well as the cardiovascular system is through other mechanisms. Potential functional clues are reported through findings of binding of Spp2 to cytokines of the transforming growth factor-β/bone morphogenetic protein (TGF-β/BMP) superfamily and activation of intracellular signaling pathways (62, 71). Our data on lowering of BP and UPE in the rats with the SHR alleles of Spp2, coupled with the report of Spp2 binding to cytokines of the TGF-β/BMP superfamily, may suggest the possibility for alterations in a common mechanism involving the members of the TGF-β/BMP family of proteins as mediators of Spp2 function to reduce BP in the cardiovascular system and UPE in kidney and to promote bone loss.

The congenic segment contains alleles from the SHR. There are several reports indicating that SHR, compared with the WKY, has lower bone mass, increased bone turnover, lower bone mineral density, reduced osteogenic markers of bone density (36), and altered vitamin D metabolism (4, 31, 46, 59, 68, 70). A contrasting report on older (20 mo old) SHR suggests that the SHR has higher trabecular bone fraction and microstructure (38). In any case, our results on the bone mass of the S being reduced by introgression of the SHR alleles supports the majority of the reports on the poor quality of the bones from the SHR and points to the RNO9 congenic segment as a candidate region harboring SHR alleles, contributing to the observed inferior bone quality of the SHR.

The idea that the Spp2 locus influencing bone morphology could also be related to BP actually has an interesting parallel in the human syndrome of autosomal dominant hypertension with brachydactyly originally described in a Turkish family (7, 60). This syndrome has recently been shown to be due to missense mutations at the PD3EA locus coding for phosphodiesterase 3A (42, 63), which hydrolyzes cGMP and cAMP and which has prominent effects in heart, vascular smooth muscle, oocytes, and platelets. Moreover, phosphodiesterase 3A is expressed in the developing limbs of mice, suggesting a role in chondrogenesis (42). The fact that six separate kindreds with mutations in the same region of phosphodiesterase 3A causing a gain of function all show hypertension and brachydactyly (42) indicates that the PD3EA locus has pleiotropic effects influencing BP and bone structure. Thus, the association of disparate properties such as BP and bone structure does make sense, once the underlying mechanisms involving cell-signaling pathways were discovered. By extension it is reasonable to speculate that Spp2 might also support dual functions with respect to BP and bone density.

In summary, this study expands our understanding of the genetic underpinnings of a genomic region encompassing the gene Spp2. An interesting and important detail obtained through our study is that the SHR alleles within the S.SHR(9) strain 3 congenic segment had a beneficial effect of lowering BP but had a detrimental effect on bone health. This dichotomy is especially important in the clinical context and serves as a feature to be noted for considering Spp2 as a candidate gene for either hypertension or bone health. Overall, our study serves as a necessary foundation to explore whether Spp2 functions as a pleiotropic gene in cardiovascular and bone health or the possibility that there are closely linked loci that exert independent effects on BP and bone health.

GRANTS

This work was supported by National Heart Lung and Blood Institute Grants HL-076709, HL-112641, and HL-020176 to B. Joe.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Y.N., S.K., H.W., X.C., B.M., P.J.C., and B.L.-C. performed experiments; Y.N., S.K., P.J.C., B.L.-C., and B.J. analyzed data; Y.N., S.K., P.J.C., B.L.-C., and B.J. interpreted results of experiments; Y.N., S.K., P.J.C., B.L.-C., and B.J. prepared figures; Y.N., S.K., P.J.C., B.L.-C., and B.J. edited and revised manuscript; Y.N., S.K., H.W., X.C., B.M., P.J.C., B.L.-C., and B.J. approved final version of manuscript; B.J. conception and design of research; B.J. drafted manuscript.