Abstract
Previous studies have shown that the Copenhagen 2331 (COP) and Dark Agouti (DA) rats have significant differences in bone structure and strength despite their similar body mass. Thus, these inbred rat strains may provide a unique resource to identify the genetics underlying the phenotypic variation in bone fragility. A sample of 828 (405 males and 423 females) COP × DA F2 progeny had extensive phenotyping for bone structure measures including cortical bone area and polar moment of inertia at the femur midshaft and total, cortical and trabecular bone areas, for the lumbar vertebra 5 (L5). Bone strength phenotypes included ultimate force, stiffness and work to failure of femur and L5. These skeletal phenotypes were measured using peripheral quantitative computed tomography (pQCT) and mechanical testing. A whole-genome screen was conducted in the F2 rats, using microsatellite markers spaced at approximately 20 cM intervals. Genetic marker maps were generated from the F2 data and used for genome-wide linkage analyses to detect linkage to the bone structure and strength phenotypes. Permutation testing was employed to obtain the thresholds for genome-wide significance (p<0.01). Significant QTL for femur structure and strength were identified on chromosome (Chr) 1 with a maximum LOD score of 33.5; evidence of linkage was found in both the male and female rats. In addition, Chrs 6, 7, 10, 13, 15 and 18 were linked to femur midshaft structure. QTL linked to femur strength were identified on Chrs 5 and 10. For L5 vertebrae, Chrs 2, 16, and 18 harbored QTL for cortical structure and trabecular structure for L5 was linked to Chrs 1, 7, 12, and 18. One female-specific QTL for femur ultimate force was identified on Chr 5, and two male-specific QTL for L5 cortical area were found on Chrs 2 and 18. Our study demonstrates strong evidence of linkage for bone structure and strength to multiple rat chromosomes.
Keywords: QTL, bone strength, bone structure, bone biomechanics, inbred rats
Introduction
Bone strength is a major determinant of osteoporotic fracture. Identification of the genes underlying bone structure and strength will provide important insight regarding the genetics of osteoporosis and fracture risk. To complement ongoing human studies to identify genes contributing to the variation in bone strength we have chosen to employ animal model to identify the genes contributing to important bone phenotypes, such as biomechanical strength. There are several advantages of an animal model, and in particular a rat model for the study of biomechanical strength. First, collection of these phenotypes requires destructive testing of the bone, which is only possible in animal models. Rats also have advantages over mice for these tests because of their larger bone size, allowing more precise structural and biomechanical measurements. Importantly, rats have proven to be a highly predictive animal model of fracture risk in humans [1]. Our previous work has allowed us to identify several inbred rats strains that are similar in body weight but vary considerably in bone strength and structure and therefore are ideal for the identification of genes contributing to the variation in these two traits.
We previously reported results from a genome screen for bone strength and structure phenotypes in a second filial (F2) population of offspring from inbred Fischer 344 (F344) and Lewis (LEW) progenitor strains [2]. Significant quantitative trait loci (QTL) for femoral structure were detected on chromosomes (Chrs) 2, 4, 5, 7 and 15. QTL affecting femoral structure on Chrs 2 and 5 were also found to influence femur strength. QTL on Chrs 2, 10 and 19 were also found to contribute to variability in lumbar vertebral strength.
We conducted a second study using Copenhagen 2331 (COP) and Dark Agouti (DA) rats. We chose COP and DA rats because these two inbred rats differ substantially in skeletal phenotypes of structure and strength despite their similar body mass. Specifically, DA rats have 12% more femoral ultimate force, 59% more femoral work to failure, 49% more femoral cross-sectional moment of inertia, and 12% more lumbar vertebral work to failure than that of COP rats. Therefore, it seems likely that DA rats carry alleles that cause stronger skeletal structure and strength. Because of the different skeletal and structural properties of the COP/DA rats as compared with the F344/Lew rats, it is likely that novel loci will be identified in the COPxDA cross that will provide additional insight regarding the genes and pathways contributing to skeletal structure and strength. In the current study, we conducted a genome-wide screen for the structure and strength phenotypes of femur and lumbar vertebrae. The sample included 828 F2 progeny generated from the cross of COP and DA rats. Our goal of this study was to identify QTL controlling bone structure and strength in this novel intercross.
Materials and methods
Animal breeding
Reciprocal mating of 12 breeding pairs of COP rats with DA rats purchased from Harlan (Indianapolis, IN) were used to generate 190 first filial (F1) offspring. Brother and sister F1 rats were then intercrossed to create a total of 828 (405 males and 423 females) F2 offspring. All rats were group-housed (two per cage) and maintained with ad libitum water and standard rat chow (NIH-31 Mouse/Rat diet 7017, Harlan Teklad, Madison, WI) under constant ambient temperature (21 ± 2° C) in a 12 h light/dark cycle at Indiana University Laboratory Animal Resource Center (LARC). F2 rats were euthanized at the age of 26 weeks and femurs and lumbar vertebrae (L3–L6) were disected. Right femurs and lumbar vertebrae were stored in −20°C freezer for future biomechanical tests. Left femurs were prepared for bone structure measurements by removing all muscle tissue. Spleens were excised and stored at −80°C. All procedures were approved by Indiana University Animal Care and Use Committee and performed in accordance with National Institutes of Health (NIH) guidelines for the care and use of animals in research.
Peripheral quantitative computed tomography (pQCT)
For femur and lumbar 5 vertebrae (L5) pQCT scans, each scan slice was 0.26 mm in thickness with a voxel size of 0.07 mm. The specimen (left femur or L5) were placed in plastic tubes and centered in the gantry of a Norland Stratec XCT Research SA+ pQCT (Stratec Electronics, Pforzheim, Germany). For each left femur, cross-sectional slices were taken through the femoral midshaft and 1mm proximal to growth plate of the distal femur. L5 vertebrae were scanned in cross-section at the mid cross-section of the vertebral body. For each slice, the X-ray source was rotated through 180 degrees of projection. The slice through the femur midshaft included only cortical bone, whereas the distal slice included cortical shell and trabecular bone. The scan slice through the midpoint of L5 included cortical and trabecular bone. Total, cortical cross sectional areas (mm2) and polar moment of inertia (Ip; mm4) for femur midshaft were measured from the pQCT images. For distal femur and L5 scans, total, cortical and trabecular areas were measured. Density thresholds of 500 and 900 were used to differentiate cortical from trabecular bone.
Biomechanical testing
Femur three-point bending test
The frozen right femurs were first brought to room temperature slowly in a saline bath. After cleaning of muscle, the femurs were placed with the anterior side facing up on the lower supports of a three-point bending fixture, where the distal-most support was immediately proximal to the femoral condyles. Load was applied using a material testing machine (Alliance RT/5; MTS Systems Corp., Eden Prairie, MN, USA). The bones were held in place by small (1N) preload and loaded in monotonic axial compression until fracture at a crosshead speed of 20 mm/minute. Load was applied midway between two supports that were 15 mm apart for female rats and 20 mm for male rats. The tests were performed at 37° C.
L5 compression test
L5 vertebrae were dissected from the lumbar spine, and the neural arch was removed from the vertebral bodies by clipping through the base of the pedicles using bone cutters. Then the endplates of the vertebral bodies were removed by clamping the body into a custom chuck that mounts to a low-speed diamond-wafering saw (Isomet; Buehler, Lake Bluff, IL, USA). Parallel, transverse cuts were made at each end of the vertebral body, yielding a specimen that was about 4 mm in caudo-cranial height. The outer dimensions in the ventrodorsal and mediolateral planes were measured to the nearest 0.01 mm using digital calipers (Mitutoyo, Aurora, IL, USA). The vertebrae were mounted between the plates of the same testing machine with a small preload (1N) and loaded in monotonic axial compression until fracture at a crosshead speed of 20 mm/minute. The compression tests were performed at 22° C.
Data collecting
During the biomechanical testing, force-displacement measurements were collected using MTS TestWorks software (version 4.06). Ultimate force (Fu; N), work to failure (U; mJ), and stiffness (S; N/mm) were obtained directly from the force versus displacement curves. Fu reflects the strength of the bone or maximum load that the bone can support before failing; U reflects the total energy the specimen can absorb before fracture; and S reflects the stiffness or maximum slope of the load displacement curve.
DNA isolation and microsatellite marker genotyping
The rat spleen from each rat was used to isolate the genomic DNA using the Puregene DNA isolation kit (Gentra Systems, Minneapolis, MN) according to the manufacturer’s protocol. Microsatellite markers (Research Genetics, Birmingham, AL) which were genetically informative for the COP X DA cross were selected for genotyping in the F2 sample. A total of 93 markers spanning the entire genome (Chromosomes 1–20 and X) were selected for genotyping and yielded an average intermarker distance of 20 cM. Out of these 93 markers, 82 markers are different from those selected for the F344-LEW cross in previous study [2]. The PCR primer sequences were designed for multiplexing using the fluorescent dyes. The QIAGEN Multiplex PCR Kit (Qiagen, Valencia, CA) was used for all multiplex PCR reactions. Allele sizes of the fluorescently labeled PCR products were determined on an ABI 3100 Genetic Analyzer (PE, Applied Biosystem, Foster City, CA) using the program, Genotyper, version 3.6.
Quantitative linkage analysis
All F2 genotypic data was reviewed to ensure the expected 1:2:1 Mendelian ratio. The genotypic data from all 828 F2 animals were used to generate recombination-based marker maps using the program MAPMARKER/EXP [3]. The marker order and intermarker distances were compared with those previously published by the Rat Genome Database (RGD) (http://rgd.mcw.edu/).
The following phenotypes were prioritized for analysis: femur and L5 bone areas, femur and L5 biomechanics, and femur Ip. The distribution of each phenotype in the F2 rats was reviewed; all outliers were verified or removed from further analysis.
Multipoint linkage analysis using the program R/qtl [4] was performed using the genotypic and phenotypic data from the 828 F2 animals. The recombination-based marker map generated using the data from the 828 F2 animals was used. Permutation testing was employed to obtain appropriate genome-wide significance levels for the linkage results [5]. To maintain the correlations underlying the phenotypes being analyzed, the phenotypes for each rat were kept together and randomly reassigned (permuted) to another rat in the sample. This phenotypic reassignment was performed for each of the 828 F2 animals to generate a permutated sample. This permutation process was repeated and used to generate 5,000 permutated datasets. Genome-wide linkage analysis was then performed in each of the 5,000 permutated datasets. By including all phenotypes in the permutation testing, genome-wide significance thresholds appropriate for each phenotype were obtained. The maximum LOD score for linkage for each phenotype was recorded in each permutated dataset. By reviewing the distribution of the maximum LOD scores, the 99th (significant) percentile of the maximum genome-wide LOD score across all phenotypes was found to be 4.3. To further characterize the genotypic effects in each QTL region, analysis of variance (ANOVA) was performed using the most significant marker in each QTL.
Sex specificity of each QTL was evaluated following the method described by Solberg et al [6]. This method involves calculating the likelihood of the genotype and phenotype data under two models. The first (full) model contains effects for the QTL, sex, and a QTL-by-sex interaction term, while the second (reduced) model contains terms for QTL and sex; thus the models differ only in the QTL-by-sex interaction effect. At the position of each QTL achieving genome-wide significance in the primary screen (LOD>4.3), a traditional likelihood ratio test (1-df chi-square) was performed using the full and reduced models. This represents a test for sex-specificity of the QTL for the phenotype and chromosomal position considered.
Results
A number of QTL with LOD scores indicating significant linkage (>4.3) for structure and strength were identified in this study. Results are summarized in Table 1 and Table 2, and also presented graphically in Figure 1. Several highly significant QTL for femur structure and strength were identified in the distal region of Chr 1, with LOD scores ranging from 8.5 to 33.5. The most significant statistical evidence was observed on Chr 1 for femur total area, producing a LOD score of 33.5 when analyzing all F2 rats. In the same region of Chr 1, highly significant linkage for other femur structure phenotypes was also detected, including Ip (LOD = 32.6) and femur cortical area (LOD = 18.6). Notably, when either male F2 rats or female F2 were analyzed separately, high LOD scores were still produced for femur Ip and total area QTL in the same region. In addition, significant linkage to this chromosomal region was detected for femur biomechanics (Table 1, Figure 1B). LOD scores of 10.7, 7.1 and 7.1 were obtained for U, S and Fu, respectively. In addition, significant QTL for lumbar vertebrae structure were also present on the distal region of Chr 1 (Table 2, Figure 1C). LOD scores of 9.4 and 7.3 were obtained for L5 total area and trabecular area in all F2 rats, respectively.
Table 1.
Summary of genome screen LOD score for structure and biomechanics of femura
| Chromosome | Structure
|
Biomechanics
|
||||||
|---|---|---|---|---|---|---|---|---|
| LOD (Position in cM)b | LOD (Position in cM) | |||||||
| Phenotype | All | Male | Female | Phenotype | All | Male | Female | |
| 1 | Ip (midshaft) | 32.65 (110) | 17.21 (109) | 19.56 (109) | Ultimate Force | 10.66 (114) | 5.48 (112) | 6.22 (115) |
| Total Area (midshaft) | 33.52 (111) | 17.64 (109) | 19.41 (111) | Stiffness | 7.05 (117) | |||
| Cortical Area (midshaft) | 18.63 (112) | 11.77 (109) | 8.48 (112) | Work to Failure | 7.10 (114) | |||
| 5 | Ultimate Force | 4.64 (84)c | ||||||
| 6 | Ip (midshaft) | 5.05 (20) | ||||||
| Total Area (midshaft) | 5.18 (20) | |||||||
| Cortical Area (midshaft) | 6.79 (16) | 4.68 (16) | ||||||
| 7 | Ip (midshaft) | 7.64 (37) | 4.84(40) | |||||
| Total Area (midshaft) | 6.71 (37) | |||||||
| 10 | Ip (midshaft) | 9.34 (73) | 7.79 (73) | Work to failure | 5.66 (8) | |||
| Total Area (midshaft) | 10.96 (73) | 8.47 (73) | ||||||
| 13 | Ip (midshaft) | 4.61 (41) | ||||||
| Total Area (midshaft) | 4.60 (41) | |||||||
| 15 | Cortical Area (midshaft) | 5.31 (32) | ||||||
| 18 | Ip (midshaft) | 4.59 (29) | ||||||
All LOD score above the significance for the 99th percentile (LOD = 4.3) are reported
Position in cM is based on Rat Genomic Database (RGD) map
Female sex-specific QTL
Table 2.
Summary of genome screen LOD score for structure of lumbar 5a
| Chromosome | Phenotype | LOD (Position in cM)b |
||
|---|---|---|---|---|
| All | Male | Female | ||
| 1 | Total Area | 9.38 (107) | 5.70 (107) | |
| Trabecular Area | 7.25 (106) | |||
| 2 | Cortical Area | 10.97 (41) | 9.07 (43)c | |
| 7 | Total Area | 13.48 (54) | 8.25 (58) | 6.58 (50) |
| Trabecular Area | 16.05 (56) | 9.79 (58) | 7.59 (54) | |
| 10 | Total Area | 5.33 (62) | ||
| 12 | Trabecular Area | 5.61 (43) | ||
| 16 | Cortical Area | 5.13 (18) | ||
| 18 | Cortical Area | 8.80 (14) | 8.81 (13)c | |
| Trabecular Area | 4.64 (13) | |||
All LOD score above the significance for the 99th percentile (LOD = 4.3) are reported
Position in cM is based on Rat Genomic Database (RGD) map
Male sex-specific QTL
FIG. 1.
Genome-wide linkage analysis for femur (midshaft) structure (A), femur strength (B) and lumbar vertebrae (L5) (C) structure for rat Chromosomes 1 – 20 (excluding X and Y chromosomes). Multiple LOD scores plotted on the Y axis vs. the relative location along each chromosome on the X axis. The dashed horizontal lines indicate the threshold value for genome-wide significance at p=0.05 (lower line) and p=0.01 (upper line).
In addition to Chr 1, highly significant evidence of linkage for femur structure phenotypes was obtained on Chr 10 (Figure 1A). QTL for femur total area and Ip were detected, with LOD scores of 11.0 for total area and 9.3, respectively, when analyzing all F2 rats. When only female rats were analyzed, LOD scores for Ip and total area were 7.8 and 8.5, respectively. Significant QTL (LOD = 5.3) for lumbar structure (total area) was also found on Chr 10 near the location of the femur structure QTL (Table 2). A significant QTL (LOD = 5.7) for femur biomechanics (U) was also detected on Chr 10, but in a different region (8 cM).
Major QTL for lumbar vertebral structure were identified on Chrs 2, 7 and 18 (Table 2, Figure 1C). Of these, QTL on Chr 7 produced the maximum LOD scores for L5 total area (LOD = 13.5) and trabecular area (LOD = 16.1) when analyzing all F2 rats. On the same chromosome, when male rats were analyzed alone, LOD scores for L5 total area and trabecular area were 8.3 and 9.8, respectively. For female rats, LOD scores for these two L5 structure phenotypes were 6.6 and 7.6, respectively. A maximum LOD score of 11.0 was obtained for L5 cortical area on Chr 2 in all F2 rats. On Chr 18, we observed LOD scores of 8.8 for L5 cortical area and 4.6 for L5 trabecular area.
Lesser QTL for femur structure and strength were observed on Chrs 5, 6, 7, 13, 15 and 18 (Table 1), with LOD scores ranging from 4.6 to 7.6. Lesser L5 QTL were observed on Chrs 10, 12 and 16, with LOD scores ranging from 5.1 to 5.6.
Several QTL were unique to a single skeletal site. QTL on Chrs 5, 6, 13 and 15 were linked to femur phenotypes but showed no significant linkage to vertebral phenotypes. Conversely QTL on Chrs 2, 12, and 15 were linked to vertebral phenotypes with no linkage to femur phenotypes. QTL on Chr 1, 7, 10 and 18 were linked to both femoral and vertebral phenotypes.
Three sex-specific QTL were observed (Figure 2). The QTL for femoral work to failure detected on Chr 5 is female-specific as indicated in Figure 2A. LOD scores for femoral ultimate force were significant in the full sample and the sample of only female rats, but the LOD score was not significant when male rats were analyzed alone. There was a significant p-value for a QTL-by-sex interaction on Chr 5 (p-value 0.01). QTL for L5 cortical area on Chrs 2 and 18 were male-specific. As shown in Figure 2B and 2C, LOD scores for L5 cortical area were significant in the full sample and the sample of only male rats, but LOD score was not significant when analyzing only the female rats. The p-values for QTL-by-sex interaction on Chrs 2 and 18 were 0.05 and 0.01, respectively. The QTL for Ip on Chr 7 showed statistical evidence for sex specificity (p<0.05), but this effect was not observed in the phenotypic mean values for male and female F2 offspring; therefore, we believe this result is a statistical artifact and does not represent a true sex specific effect.
FIG. 2.
Sex-specific linkage for femur ultimate force on Chromosome 5 (A), for lumber 5 (L5) cortical area on chromosome 2 (B) and 18 (C).
Discussion
In this study, we detected very strong evidence for linkage for bone structure and strength on rat Chr 1. Highly significant results with LOD scores for femur Ip and total area both above 30 were obtained on Chr 1. The significant QTL identified in this study explained approximately 29% of the total phenotypic variance for total area among the F2 offspring. The QTL on Chr 1 accounted for half of this effect (14% to the total variance in total area).
QTL detected on Chrs 1 and 10 appear to affect on both structure and strength, providing evidence of linkage for multiple phenotypes at both the femur and lumbar vertebra. In particular, Chr 1 QTL affects many structure and strength phenotypes including femur Ip, femur total area, femur cortical area, lumbar total area, lumbar trabecular area, femur ultimate force, stiffness and work to failure (Tables 1 and 2), suggesting that the gene(s) underlying this QTL may have a general regulatory role on skeletal growth and development. The linkage to multiple skeletal traits suggests a common genetic mechanism in coordinately regulating skeletal traits [7], which forms the genetic foundation for the high correlation among the traits. The d/d alleles in this study were associated with significantly higher values for all structure and strength phenotypes (Table 3), indicating the contribution from DA at this locus consistently improves bone structure.
Table 3.
Genotypic mean values for femur biomechanics and structure
| Phenotype
|
Genotype
|
|||||||
|---|---|---|---|---|---|---|---|---|
| Femur | Unit | Chr | Marker | c/c | c/d | d/d | R2 | ANOVA P-value |
| Total Area (midshaft) | mm2 | 1 | D1Rat69 | 8.77 ± 0.04 | 9.10 ± 0.03 | 9.39 ± 0.04 | 0.136 | <0.0001 |
| 6 | D6Rat42 | 9.00 ± 0.04 | 9.09 ± 0.03 | 9.28 ± 0.04 | 0.029 | <0.0001 | ||
| 7 | D7Rat23 | 9.22 ± 0.04 | 9.14 ± 0.03 | 8.92 ± 0.04 | 0.033 | <0.0001 | ||
| 10 | D10Rat15 | 8.88 ± 0.04 | 9.10 ± 0.03 | 9.28 ± 0.04 | 0.061 | <0.0001 | ||
| 13 | D13Rat86 | 9.00 ± 0.04 | 9.15 ± 0.02 | 9.26 ± 0.04 | 0.026 | <0.0001 | ||
| Cortical Area (midshaft) | mm2 | 1 | D1Rat69 | 5.28 ± 0.03 | 5.43 ± 0.02 | 5.57 ± 0.02 | 0.079 | <0.0001 |
| 6 | D6Rat42 | 5.35 ± 0.02 | 5.43 ± 0.02 | 5.53 ± 0.02 | 0.037 | <0.0001 | ||
| 15 | D15Rat117 | 5.51 ± 0.02 | 5.47 ± 0.02 | 5.37 ± 0.02 | 0.027 | <0.0001 | ||
| Ip (midshaft) | mm4 | 1 | D1Rat69 | 13.21 ± 0.12 | 14.29 ± 0.08 | 15.16 ± 0.12 | 0.142 | <0.0001 |
| 6 | D6Rat42 | 13.99 ± 0.13 | 14.20 ± 0.09 | 14.83 ± 0.13 | 0.022 | <0.0001 | ||
| 7 | D7Rat23 | 14.66 ± 0.13 | 14.39 ± 0.09 | 13.67 ± 0.13 | 0.036 | <0.0001 | ||
| 10 | D10Rat15 | 13.64 ± 0.13 | 14.26 ± 0.09 | 14.81 ± 0.12 | 0.052 | <0.0001 | ||
| 13 | D13Rat86 | 13.93 ± 0.13 | 14.44 ± 0.09 | 14.77 ± 0.13 | 0.027 | <0.0001 | ||
| 18 | D18Rat50 | 13.95 ± 0.12 | 14.18 ± 0.09 | 14.69 ± 0.14 | 0.022 | 0.0002 | ||
| Ultimate Force | N | 1 | D1Rat169 | 128.71 ± 1.31 | 135.72 ± 0.89 | 140.69 ± 1.26 | 0.051 | <0.0001 |
| 5 | D5Arb1 | 133.56 ± 1.30 | 134.00 ± 0.90 | 140.26 ± 1.31 | 0.022 | <0.0001 | ||
| Stiffness | N/mm | 1 | D1Rat169 | 323.94 ± 5.15 | 350.82 ± 3.47 | 361.98 ± 4.94 | 0.036 | <0.0001 |
| Work to failure | mJ | 1 | D1Rat169 | 45.63 ± 0.70 | 49.51 ± 0.47 | 50.43 ± 0.66 | 0.035 | <0.0001 |
| 10 | D10Mit17 | 46.49 ± 0.68 | 48.99 ± 0.47 | 51.40 ± 0.69 | 0.031 | <0.0001 | ||
Values are mean ± SE. c/c homozygous for COP alleles; c/d, heterozygous; d/d homozygous for DA alleles.
In addition to QTL for bone strength on Chr 1, QTL were also detected on Chrs 5 and 10 for femoral ultimate force and work to failure, respectively. The QTL linked to femoral strength on Chr 10 is at a different location from the QTL for bone structure on the same chromosome, suggesting that this QTL influences bone strength independent of influences on bone structure. The QTL on Chr 5 was only linked to femur strength, not structure, and found only in female rats. Furthermore, several QTL appear to be femur-specific. For instance, QTL detected on Chrs 13 and 15 were only linked to femur structure phenotypes.
Highly significant QTL for rat lumbar vertebral structure were observed in this study for the first time. The QTL for lumbar vertebrae structure were distributed widely across the genome, indicating that genetic regulation of the lumbar vertebrae structure is just as complex as that of BMD and other skeletal traits. Significant linkage for L5 structure was found on 7 chromosomes, exhibiting LOD scores ranging from 4.6 to 16. Of those, QTL linked to L5 structure, but not femoral structure, were detected on Chrs 2, 12 and 16. Among those being detected, QTL on Chr 7 is particularly worth noting because it not only exhibited the maximum LOD score, but it is homologous to chromosome 12q21 in human, linked to spine BMD [9]. For lumbar structure, d/d allele caused significantly lower genotypic mean values on Chrs 2, 7, 10, 12 and 16, but the opposite trend was observed on Chr 1.
The observed sex-specific incidence of fractures suggests that skeletal genetic regulation is influenced by gender. Studies in human [10, 11] and mice [12] have uncovered several QTL exerting gender-specific effects on femoral structure and BMD. Sex-specific QTL analysis performed in this study revealed several sex-specific QTL: one female-specific QTL for femur work to failure on Chr 5, two male-specific QTL for L5 cortical area on Chrs 2 and 18. The sex-specific allelic effects for those QTL were clearly observed for genotypic mean values of these phenotypes. Interestingly, the syntenic region for female-specific work to failure in rat coincides with the whole body and hip BMD on 1p36 in human [13, 14]. The male-specific QTL for L5 cortical area on Chr 2 share linkage homology with human chromosome 3q24–25 for pelvic axis length, midfemur, femur head and femur shaft width [15, 16], as well as with chromosome 13q14 for trochanter BMD [17].
Several QTL identified in this study appear to co-localize with QTL identified in our previous study using F344 and LEW rats. For instance, both studies found QTL for femoral structure on Chr 7 and QTL for femur strength on Chr 1. However, the highly significant QTL for femoral and vertebral structure observed on Chr 1 for COP and DA rats was not observed in the F344 and LEW cross. Likewise, new QTL were observed on Chrs 6, 10, 13, and 18 for femoral structure and Chrs 7, 12, 16 and 18 for L5 structure. We conclude that the current study identified numerous new QTL that expand our knowledge of skeletal genetics.
The ultimate goal of animal studies is to lead toward the identification of human genes that contribute to skeletal traits. In this regard, some structure and strength QTL detected in this study are concordant with bone QTL previously observed in humans. For instance, the QTL region for femur Ip on Chr 1 is homologous to the region of human Chr 6q25–27, which is linked to spine and trochanter BMD [18]. Syntenic regions corresponding to the QTL regions for femur total area on Chrs 1, 6, 7 and 10 were also linked to wrist bone size on 9q21, forearm and hip BMD on 2p21–25, spine BMD on 22q13, and hip BMD on17p11–12 in humans [9, 14, 19]. Given the destructive nature of bone strength testing and the lack of statistical power in many human genome-wide linkage studies, it is not surprising that many QTL detected in this study have not been identified in human studies. However, the results from our study may still provide useful information about the genetics of osteoporosis and fracture in humans.
In summary, we have identified multiple QTL associated with bone strength and structure in inbred COP and DA rats, many of which are distinct from those that were previously identified. In particular, QTL we found on Chr 1 in this study not only exhibited a strong association with bone strength and structure, but also shared homologous regions with human loci for bone phenotypes. Further study of these QTL regions offers a good chance to identify the candidate genes that contribute to structure and strength.
Table 4.
Genotypic mean values for lumbar 5 (L5) structure
| Phenotype
|
Genotype
|
|||||||
|---|---|---|---|---|---|---|---|---|
| L5 | Unit | Chr | Marker | c/c | c/d | d/d | R2 | ANOVA P-value |
| Total Area | mm2 | 1 | D1Rat69 | 6.22 ± 0.04 | 6.38 ± 0.03 | 6.54 ± 0.04 | 0.048 | <0.0001 |
| 7 | D7Rat23 | 6.54 ± 0.04 | 6.41 ± 0.02 | 6.16 ± 0.04 | 0.061 | <0.0001 | ||
| 10 | D10Rat124 | 6.38 ± 0.04 | 6.45 ± 0.03 | 6.22 ± 0.04 | 0.031 | <0.0001 | ||
| Cortical Area | mm2 | 2 | D2Rat280 | 2.11 ± 0.02 | 2.03 ± 0.01 | 1.93 ± 0.02 | 0.060 | <0.0001 |
| 16 | D16Rat60 | 2.09 ± 0.02 | 2.01 ± 0.01 | 1.98 ± 0.02 | 0.025 | <0.0001 | ||
| 18 | D18Rat50 | 2.10 ± 0.02 | 2.00 ± 0.01 | 1.97 ± 0.02 | 0.039 | <0.0001 | ||
| Trabecular Area | mm2 | 1 | D1Rat69 | 3.53 ± 0.03 | 3.65 ± 0.02 | 3.77 ± 0.03 | 0.038 | <0.0001 |
| 7 | D7Rat78 | 3.79 ± 0.03 | 3.64 ± 0.02 | 3.46 ± 0.03 | 0.072 | <0.0001 | ||
| 12 | D12Rat30 | 3.72 ± 0.03 | 3.71 ± 0.02 | 3.54 ± 0.03 | 0.031 | <0.0001 | ||
| 18 | D18Rat65 | 3.57 ± 0.03 | 3.65 ± 0.02 | 3.74 ± 0.03 | 0.020 | 0.0003 | ||
Values are mean ± SE. c/c homozygous for COP alleles; c/d, heterozygous; d/d homozygous for DA alleles.
Acknowledgments
This work was supported by the US National Institutes of Health by the following grants: RO1AR047822 and P01AG018397.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Wronski TJ, Walsh CC, Ignaszewski LA. Histologic evidence for osteopenia and increased bone turnover in ovariectomized rats. Bone. 1986;7:119–23. doi: 10.1016/8756-3282(86)90683-6. [DOI] [PubMed] [Google Scholar]
- 2.Alam I, Sun Q, Liu L, Koller DL, Fishburn T, Carr LG, et al. Whole-genome scan for linkage to bone strength and structure in inbred Fischer 344 and Lewis rats. J Bone Miner Res. 2005;20:1589–96. doi: 10.1359/JBMR.050512. [DOI] [PubMed] [Google Scholar]
- 3.Lander ES, Green P, Abrahamson J, Barlow A, Daly MJ, Lincoln SE, et al. MAPMAKER: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations. Genomics. 1987;1:174–81. doi: 10.1016/0888-7543(87)90010-3. [DOI] [PubMed] [Google Scholar]
- 4.Broman KW, Wu H, Sen S, Churchill GA. R/qtl: QTL mapping in experimental crosses. Bioinformatics. 2003;19:889–90. doi: 10.1093/bioinformatics/btg112. [DOI] [PubMed] [Google Scholar]
- 5.Doerge RW, Churchill GA. Permutation tests for multiple loci affecting a quantitative character. Genetics. 1996;142:285–94. doi: 10.1093/genetics/142.1.285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Solberg LC, Baum AE, Ahmadiyeh N, Shimomura K, Li R, Turek FW, et al. Sex- and lineage-specific inheritance of depression-like behavior in the rat. Mamm Genome. 2004;15:648–62. doi: 10.1007/s00335-004-2326-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Li X, Masinde G, Gu W, Wergedal J, Mohan S, Baylink DJ. Genetic dissection of femur breaking strength in a large population (MRL/MpJ × SJL/J) of F2 Mice: single QTL effects, epistasis, and pleiotropy. Genomics. 2002;79:734–40. doi: 10.1006/geno.2002.6760. [DOI] [PubMed] [Google Scholar]
- 8.Koller DL, Alam I, Sun Q, Liu L, Fishburn T, Carr LG, et al. Genome screen for bone mineral density phenotypes in Fisher 344 and Lewis rat strains. Mamm Genome. 2005;16:578–86. doi: 10.1007/s00335-004-2459-0. [DOI] [PubMed] [Google Scholar]
- 9.Deng HW, Xu FH, Huang QY, Shen H, Deng H, Conway T, et al. A whole-genome linkage scan suggests several genomic regions potentially containing quantitative trait Loci for osteoporosis. J Clin Endocrinol Metab. 2002;87:5151–9. doi: 10.1210/jc.2002-020474. [DOI] [PubMed] [Google Scholar]
- 10.Duncan EL, Cardon LR, Sinsheimer JS, Wass JA, Brown MA. Site and gender specificity of inheritance of bone mineral density. J Bone Miner Res. 2003;18:1531–8. doi: 10.1359/jbmr.2003.18.8.1531. [DOI] [PubMed] [Google Scholar]
- 11.Van Pottelbergh I, Goemaere S, Zmierczak H, De Bacquer D, Kaufman JM. Deficient acquisition of bone during maturation underlies idiopathic osteoporosis in men: evidence from a three-generation family study. J Bone Miner Res. 2003;18:303–11. doi: 10.1359/jbmr.2003.18.2.303. [DOI] [PubMed] [Google Scholar]
- 12.Orwoll ES, Belknap JK, Klein RF. Gender specificity in the genetic determinants of peak bone mass. J Bone Miner Res. 2001;16:1962–71. doi: 10.1359/jbmr.2001.16.11.1962. [DOI] [PubMed] [Google Scholar]
- 13.Wilson SG, Reed PW, Bansal A, Chiano M, Lindersson M, Langdown M, et al. Comparison of genome screens for two independent cohorts provides replication of suggestive linkage of bone mineral density to 3p21 and 1p36. Am J Hum Genet. 2003;72:144–55. doi: 10.1086/345819. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Devoto M, Shimoya K, Caminis J, Ott J, Tenenhouse A, Whyte MP, et al. First-stage autosomal genome screen in extended pedigrees suggests genes predisposing to low bone mineral density on chromosomes 1p, 2p and 4q. Eur J Hum Genet. 1998;6:151–7. doi: 10.1038/sj.ejhg.5200169. [DOI] [PubMed] [Google Scholar]
- 15.Koller DL, Liu G, Econs MJ, Hui SL, Morin PA, Joslyn G, et al. Genome screen for quantitative trait loci underlying normal variation in femoral structure. J Bone Miner Res. 2001;16:985–91. doi: 10.1359/jbmr.2001.16.6.985. [DOI] [PubMed] [Google Scholar]
- 16.Koller DL, White KE, Liu G, Hui SL, Conneally PM, Johnston CC, et al. Linkage of structure at the proximal femur to chromosomes 3, 7, 8, and 19. J Bone Miner Res. 2003;18:1057–65. doi: 10.1359/jbmr.2003.18.6.1057. [DOI] [PubMed] [Google Scholar]
- 17.Kammerer CM, Schneider JL, Cole SA, Hixson JE, Samollow PB, O’Connell JR, et al. Quantitative trait loci on chromosomes 2p, 4p, and 13q influence bone mineral density of the forearm and hip in Mexican Americans. J Bone Miner Res. 2003;18:2245–52. doi: 10.1359/jbmr.2003.18.12.2245. [DOI] [PubMed] [Google Scholar]
- 18.Duncan EL, Brown MA, Sinsheimer J, Bell J, Carr AJ, Wordsworth BP, et al. Suggestive linkage of the parathyroid receptor type 1 to osteoporosis. J Bone Miner Res. 1999;14:1993–9. doi: 10.1359/jbmr.1999.14.12.1993. [DOI] [PubMed] [Google Scholar]
- 19.Deng HW, Xu FH, Liu YZ, Shen H, Deng H, Huang QY, et al. A whole-genome linkage scan suggests several genomic regions potentially containing QTLs underlying the variation of stature. Am J Med Genet. 2002;113:29–39. doi: 10.1002/ajmg.10742. [DOI] [PubMed] [Google Scholar]