Physiol. Genomics Add DOIs to your references at manuscript stage!
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Physiol. Genomics 36: 79-88, 2009. First published November 4, 2008; doi:10.1152/physiolgenomics.00003.2008
1094-8341/09 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figures and Tables
Right arrow All Versions of this Article:
36/2/79    most recent
00003.2008v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Argyropoulos, G.
Right arrow Articles by Rankinen, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Argyropoulos, G.
Right arrow Articles by Rankinen, T.
Received 3 January 2008; accepted in final form 28 October 2008.
Physiological Genomics 36:79-88 (2009)
1094-8341/09 $8.00 © 2009 American Physiological Society

KIF5B gene sequence variation and response of cardiac stroke volume to regular exercise

George Argyropoulos 1,*, Adrian M. Stütz 2,*, Olha Ilnytska 1, Treva Rice 3, Margarita Teran-Garcia 2, D. C. Rao 3, Claude Bouchard 2 and Tuomo Rankinen 2

1 Energy Balance Laboratory, Pennington Biomedical Research Center, Baton Rouge, Louisiana
2 Human Genomics Laboratory, Pennington Biomedical Research Center, Baton Rouge, Louisiana
3 Division of Biostatistics, Washington University School of Medicine, St. Louis, Missouri


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
A genome-wide linkage scan for endurance training-induced changes in stroke volume detected a quantitative trait locus on chromosome 10p11 in white families of the HERITAGE Family Study. Dense microsatellite mapping narrowed down the linkage region to a 7 Mb area containing 16 known and 14 predicted genes. Association analyses with 90 single nucleotide polymorphisms (SNPs) provided suggestive evidence (P values from 0.03 to 0.06) for association in the kinesin heavy chain (KIF5B) gene locus in the whole cohort. The associations at the KIF5B locus were stronger (P values from 0.001 to 0.008) when the analyses were performed on linkage-informative families only (family-specific logarithm of the odds ratio scores >0.025 at peak linkage location). Resequencing the coding and regulatory regions of KIF5B revealed no new exonic SNPs. However, the putative promoter region was particularly polymorphic, containing eight SNPs with at least 5% minor allele frequency within 1850 bp upstream of the start codon. Functional analyses using promoter haplotype reporter constructs led to the identification of sequence variants that had significant effects on KIF5B promoter activity. Analogous inhibition and overexpression experiments showed that changes in KIF5B expression alter mitochondrial localization and biogenesis in a manner that could affect the ability of the heart to adjust to regular exercise. Our data suggest that KIF5B is a strong candidate gene for the response of stroke volume to regular exercise. Furthermore, training-induced changes in submaximal exercise stroke volume may be due to mitochondrial function and variation in KIF5B expression as determined by functional SNPs in its promoter.

genotype; exercise training; functional studies; mitochondria


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
REGULAR EXERCISE HAS BEEN shown to protect against morbidity and mortality from cardiovascular diseases (CVD), whereas sedentariness is associated with increased risk of CVD. Cardioprotective effects of exercise are mediated by several physiological mechanisms, such as improved plasma lipid and lipoprotein profile, lower blood pressure, improved endothelial function, and enhanced insulin sensitivity. Regular exercise also improves cardiac function. Exercise-trained hearts work more efficiently, i.e., physically active individuals have lower heart rates but greater stroke volume at rest and during submaximal exercise than their sedentary counterparts. Exercise-trained hearts can also adjust better to suddenly increased circulatory demands and tolerate better the demands of heavy physical exertion.

Although regular physical activity improves cardiac function on average, there are marked interindividual differences in exercise training-induced changes in cardiac phenotypes. For instance, in the HERITAGE Family Study, a 20-wk endurance training program resulted in a mean increase of 3.9 ml/beat in stroke volume measured during steady-state exercise at 50 W (SV50) (29). However, the training responses ranged from a decrease of 41 ml/beat to an increase of 45 ml/beat. The strongest predictors of SV50 training response ({Delta}SV50) were baseline level of SV50 and familial aggregation. Baseline SV50 explained ~16% of the variance in {Delta}SV50, while maximal heritability of SV50 adjusted for age, body surface area, sex, and baseline SV50 was 29% (5).

To identify the chromosomal regions that potentially harbor gene(s) affecting SV50 training response, we performed a genome-wide linkage scan using >500 microsatellite markers (18). The strongest evidence of linkage in white HERITAGE families was detected on chromosome 10p11. Here we report the results of positional cloning of the quantitative trait locus (QTL) for {Delta}SV50 on 10p11 and show that the KIF5B gene is strongly associated with SV50 training response. Furthermore, we show that a DNA sequence variant associated with {Delta}SV50 also modifies KIF5B promoter activity. Finally, data are presented showing that inhibition or overexpression of KIF5B in vitro could influence the ability of the cardiac muscle to adapt to regular exercise.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Subjects
The study cohort consists of 483 white subjects (233 men and 250 women) from 99 nuclear families and 259 black subjects (88 men and 171 women) from 105 family units. Complete training response data were available for 450 whites (216 men, 234 women) and 251 blacks (88 men and 163 women). The maximum number of sib-pairs available was 328 and 102 in whites and blacks, respectively. The study design and inclusion criteria have been described previously (8). To be eligible, the individuals were required to be in good health, i.e., free of diabetes, cardiovascular diseases, or other chronic diseases that would prevent their participation in an exercise-training program. Subjects were also required to be sedentary, defined as not having engaged in regular physical activity over the previous 6 mo. Individuals with a resting systolic blood pressure >159 mmHg and/or a diastolic blood pressure >99 mmHg were excluded. The study protocol had been approved by each of the Institutional Review Boards of the HERITAGE Family Study research consortium. Written informed consent was obtained from each participant.

Submaximal Exercise Cardiac Output and Stroke Volume
Before and after the 20-wk training program, each subject completed cycle ergometer (SensorMedics Ergo-Metrics 800S, Yorba Linda, CA) exercise tests conducted on separate days: a maximal exercise test (Max), a submaximal exercise test (Submax), and a submaximal/maximal exercise test (Submax/Max) (26). The Submax test was performed at 50 W and at 60% of the initial maximal oxygen consumption (VO2max), and subjects exercised for 8–12 min at each work rate, with a 4-min period of seated rest between exercise periods. The Submax/Max test was started with the Submax protocol. After exercising at 60% VO2max, subjects also exercised for 3 min at 80% VO2max. The test then progressed to a maximal level of exertion. Heart rate (HR) and cardiac output (Q) were determined twice at 50 W (HR50 and Q50, respectively). The values presented in this paper represent the mean of the responses for the two submaximal tests (i.e., four individual measurements), both before and after training. Q50 was determined using the Collier CO2 rebreathing technique (13), as previously described (28). SV50 was derived by dividing Q50 by HR50 (measured with ECG) at the time of the Q50 determination (i.e., SV50 = Q50/HR50). Q50 and SV50 training responses ({Delta}) were calculated as posttraining values minus pretraining values.

Genotyping
The first step of the fine mapping process was addition of six microsatellite markers on the QTL region to narrow down the target area. The microsatellites were selected from the National Center for Biotechnology Information (NCBI) uniSTS database and they were genotyped using automatic DNA sequencers from LI-COR and genotypes were scored semiautomatically using the software SAGA (11).

The single nucleotide polymorphisms (SNPs) were selected from the NCBI dbSNP database. The first set of SNPs were selected before the HapMap reference data were available. Therefore, the SNPs were selected if they were located within 5 kb of the target gene and if they were reported to be polymorphic at least in one population. All SNPs were genotyped using a primer extension method with fluorescence polarization detection (FP-TDI). Changes in fluorescence polarization after excitation of the samples by plane-polarized light were measured using a Victor2 Plate Reader (Perkin Elmer Life Sciences). The allele calling was done using the SNP scorer genotyping software (Perkin Elmer Life Sciences). Details of the polymerase chain reaction (PCR) conditions and primer sequences are available upon request.

Functional Studies With Promoter Haplotypes
In silico KIF5B promoter analysis.
Manual expressed sequence tag (EST) assembly using the NCBI database was complemented by the use of the Database of Transcriptional Start Sites at http://dbtss.hgc.jp/. Analyses of GC-richness and the presence of CpG islands (CGI) were done using http://www.uscnorris.com/cpgislands2/cpg.aspx. Haplotype construction was performed with the Merlin software.

Cloning of human KIF5B haplotype promoter constructs and expression vectors.
In the first series of experiments, 250-bp-long promoter constructs containing the three haplotypes of the SNPs at positions –835 (rs211302) and –809 (rs211301) were PCR amplified and cloned into pGL3basic (Promega, Madison, WI). In the second phase, a 1794-bp human KIF5B promoter fragment (from nucleotides 1866 to 73 upstream of the ATG) was amplified from human genomic DNA with the Expand High Fidelity polymerase (Roche, Indianapolis, IN) and Q-solution from the Taq-pol kit (Qiagen, Valencia, CA) using the PCR primers 5'-CTGACTCGAGGGTTTCTGAAGACAACACTTCCC-3' and 5'-GACGAAGCTTTGAGGGCTTGTGGTCGCGAG-3'. The PCR fragment was cloned into pGL3basic to create KIF5B Haplotype 2 (containing all the common alleles). Most other haplotypes were generated with these primers using DNA from homozygous individuals containing the desired rare alleles as the template, except for haplotype 7, which used the modified upstream primer 5'-CTGACTCGAGGGTTTCTGAAGACAACAGTTCCC-3'. Haplotype 6 was created using the QuikChange II Site-Directed Mutagenesis Kit (Stratagene) and the haplotype 7 containing plasmid as the template with primers 5'-ACGGAAACAGAACCCGGGGTCCATGGTGG-3' and 5'-CCACCATGGACCCCGGGTTCTGTTTCCGT-3'. A plasmid containing the full-length cDNA and additional 5' and 3' DNA sequences of the mouse Kif5b gene was purchased from I.M.A.G.E. (http://image.hudsonalpha.org), and the insert was subcloned into the mammalian expression vector pcDNA3.1 (Invitrogen, Carlsbad, CA). Plasmids were verified by DNA sequencing. All DNAs for transfections were generated with the Endofree-Maxiprep kit (Qiagen) at the same time. Concentrations were measured multiple times with the ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE), showing <3.5% difference between haplotype constructs (phase 2).

Cell culture and reagents.
The mouse myoblast C2C12 cell line was purchased from the American Type Culture Collection (Manassas, VA). Cells were maintained in Dulbecco's modified Eagle's medium with 4 mM L-glutamine, 4,500 mg/l glucose, 1 mM sodium pyruvate, and 1,500 mg/l sodium bicarbonate supplemented with 10% fetal bovine serum (FBS), 100 units/ml penicillin, and 100 µg/ml streptomycin in a humidified atmosphere at 37°C under 5% CO2. The human embryonic kidney cell line HEK-293 (ATCC) was cultured in MEM medium (Invitrogen) supplemented with 10% FBS (Invitrogen) and 1% penicillin-streptomycin. Cells were grown in a 5% CO2 humidified incubator at 37°C and seeded into 48-well plates for HEK-293 and 12-well plates for C2C12 cells, 48 h before use in promoter studies and transfected at 80% confluence.

Transient transfections.
In the first phase of promoter analysis, cells were transfected and normalized as described previously (6, 17). In the second phase, cells were transfected with Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol using 350 ng and 1.2 µg reporter DNA and 40 ng and 400 ng Renilla vector pRL-TK (Promega) for the 48-well and 12-well formats, respectively. After incubation for 25 h, cells were lysed with 1x Passive Lysis buffer (Promega). Firefly and Renilla luciferase activity was measured using the Dual Luciferase Kit (Promega) and a FB12 luminometer (Berthold Detection systems, Oak Ridge, TN). All firefly luciferase data were normalized to equal expression of Renilla luciferase of the same tube. Shown are the means + SE of four independent experiments with each construct in triplicate (n = 12) for each cell line.

For the silencing and overexpression experiments, C2C12 cells (1 x 106) were harvested and electroporated with 2 µg of KIF5B pcDNA3.1 or 0.5 µg, 1 µg, and 2 µg of Kif5b Silencer Predesigned siRNA (targeted to exons 10, 11 of the mouse Kif5b gene) (Ambion, Austin, TX) and 2 µg of green fluorescence protein (GFP) using Nucleofector Kit V (Amaxa Biosystems, Gaithersburg, MD), program B032 according to the manufacturer's protocol. As negative controls, Silencer Negative Control scrambled 5 siRNA (Ambion) or pcDNA3.1 empty vector were used. Cells were cultured for 48 h after transfection and prior to beginning experiments.

Northern analysis.
Two premade blots were purchased (BD Biosciences Clontech, Palo Alto, CA). One blot, termed Cardiovascular (CVD), consisted of cardiac tissues, and the other blot, termed Multiple Tissue Northern (MTN), consisted of multiple human tissues. A probe representing exon 1 was generated by amplification of human genomic DNA; 100 ng of the amplicon was labeled with {alpha}–32P d(CTP) using the Mega-prime labeling kit (Amersham, Piscataway, NJ). Prehybridization and hybridization of the two blots were performed in the Express Hybridization buffer, as prescribed by the manufacturer (BD Biosciences Clontech). Blots were exposed overnight to a Kodak BioMax X2 (Rochester, NY) film.

Western blotting.
Cells were lysed in Triton lysis buffer (20 mM Tris·HCl, pH 8.0, 137 mM NaCl, 2 mM EDTA, 10% glycerol, 1% Triton X-100), containing Complete protease and phosphatase inhibitor mixtures for 20 min on ice. The lysate was centrifuged at 9,000 g for 15 min at 4°C. Protein concentration of the lysate was estimated using the Bio-Rad protein assay reagent (Pierce Biotech., Rockford, IL) with a series dilution. For Western blotting analysis, total protein lysates (30–50 µg/lane) were separated on 10% SDS-polyacrylamide gel electrophoresis and blotted to Immun-Blot polyvinylidene difluoride membranes (Bio-Rad, Hercules, CA). The membranes were incubated first with primary antibody to Kif5b (Ab9097, Abcam) and then with anti-mouse antibody (Santa Cruz Biotechnology, Santa Cruz, CA) as secondary antibody coupled to horseradish peroxidase. For loading control anti-GAPDH (AM4300, Ambion) antibody was used. The signal was detected with Amersham ECL Western Blotting Detection Reagents (GE Healthcare, Buckinghamshire, UK). Western blot images were quantified using Quantity One software (Bio-Rad).

Quantitative real-time PCR for mtDNA.
To measure mtDNA copy number per nuclear genome, Cytochrome b was used as a marker for mtDNA and β-actin for nuclear DNA as previously described (7, 25). DNA from transfected C2C12 cells was isolated using the DNeasy tissue kit (Qiagen). Total DNA concentration was determined using the NanoDrop ND 3300 spectrophotometer. Sequences for mouse Cytochrome b and β-actin DNA were obtained from GenBank and input for Taqman probe design. The primers and probes used were as follows: Cytocrome b forward: 5'-CCACTTCATCTTACCATTTATTATCGC-3', Cytocrome b reverse: 5'-TTTTATCTGCATCTGAGTTTAATCCTGT-3', Cytocrome b probe - 5'FAM d(AGCAATCGTTCACCTCCTCTTCCTCCAC) BHQ-1-3'; β-actin forward: 5'-CTGCCTGACGGCCAGG-3', β-actin reverse 5'-GGAAAAGAGCCTCAGGGCAT-3', β-actin probe - 5'FAM d(CATCACTATTGGCAACGAGCGGTTCC) BHQ-1-3'. Ten nanograms of DNA in duplicates were mixed with TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, CA), Cytochrome b or β-actin primers, and probe and run on in MicroAmp Optic 384-well Reaction Plates [Applied Biosystems (ABI)] on an ABI PRISM 7700 Sequence Detection system under standard conditions (2 min at 50°C, 10 min at 95°C, then 40 cycles of 15 s at 95°C and 1 min at 60°C). mtDNA per nuclear genome was calculated as the ratio of Cytochrome b DNA level to β-actin DNA level. Experiments were repeated a minimum of three times using quadruplicate samples.

RNA extraction and real-time quantitative PCR.
Total RNA was extracted from transfected C2C12 cells after 48 h of transfection using RNeasy Mini Kit (QIAGEN). RNA concentration and quality were assessed spectrophotometrically at wavelengths 260 and 280 nm on NanoDrop spectrophotometer. Quantitative PCR was performed using the TaqMan one-step RT-PCR core reagents kit [Applied Biosystems (ABI)] in MicroAmp Optic 384-well reaction plates (ABI) on an ABI PRISM 7700 Sequence Detection system under standard conditions (30 min at 48°C, 10 min at 95°C, then 40 cycles of 15 s at 95°C and 1 min at 60°C). We used 20 ng of total DNase-treated RNA in duplicates and each run included a standard curve with five serial dilutions and a nontemplate control. The level of gene expression for each gene was quantified relative to the level of the housekeeping gene cyclophilin using the standard curve method. The primers and probes used were as follows: Kif5b forward: 5'-ATAAGGGACTTGTTAGATGTTTCAAAGAC-3', Kif5b reverse: 5'-TCTGTTGGATTTCCCTTCATCTATG-3' and Kif5b probe: 5'-6-FAM-ACGAAACGTTCTGTGCACCCCTTTACATAG BHQ-1-3', cyclophilin forward: 5'-TAGAGGGCATGGATGTGGTAC-3', cyclophilin reverse: 5'-GCCGGAGTCGACAATGATG-3', and cyclophylin probe: 5'-6-FAM-AGCCGGGACAAGCCACTGAAGGAT-BHQ-1-3'. Experiments were repeated a minimum of three times with quadruplicate samples.

Immunofluorescence microscopy.
Electroporated C2C12 cells were seeded on 12-mm coverslips coated with poly-D-lysine (BD BioCoat; BD Biosciences, Bedford, MA) and incubated for 48 h at 37°C under 5% CO2. Coverslips were washed twice with PBS and fixed for 15 min at room temperature in PBS/3.7% formaldehyde. To detect mitochondria, cells were permeabilized in 0.25% Triton X-100 for 15 min and incubated with 10 nM MitoFluor Red 594 (Invitrogen, Molecular Probes, Eugene, Oregon) for 20 min at room temperature. Immunocytochemistry was performed as follows: cells were washed in PBS and incubated in blocking medium (1% bovine serum albumin and 10% normal goat serum) at room temperature for 1 h. Cells were incubated overnight at 4°C in the presence of an anti-Kif5b antibody (Ab9097, Abcam, Cambridge, MA) (dilution 1:350). This was followed by three washes with PBS, 1 h incubation in the presence of a secondary Alexa Fluor 350 goat anti-mouse IgM antibody (Invitrogen, Carlsbad, CA) in a working dilution 1:100 at room temperature, and again three washes with PBS. Finally slides were mounted in Prolong Gold Antifade Reagent (Invitrogen). Microphotographs of stained cells were taken on 3I Everest Imaging System. Color images were captured with the Photometrics CoolSNAPHQ Monochrome camera under x63 oil magnification.

Data Adjustment
Baseline SV50 and Q50 were adjusted for the effects of sex, age and body surface area (BSA) using stepwise multiple regression (20). Baseline HR50 was adjusted for sex, age, and body mass index (BMI, kg/m2). Training response phenotypes were adjusted also for baseline values of the respective phenotypes. In summary, SV50, Q50, and HR50 phenotypes were regressed on baseline BSA (HR50 on baseline BMI), baseline SV50, Q50, or HR50 (for training responses only) and up to a 3rd degree polynomial in age, separately within race-by-sex-by-generation subgroups. Only significant terms (5% level) were retained (i.e., the model did not need to be saturated). The residuals from this regression (or the raw score if no BSA or BMI or age terms were significant) were then standardized to zero mean and unit variance within each subgroup and constituted the phenotype for the present study.

Statistical Methods
Linkage analyses were performed using a multipoint variance components and regression-based models as implemented in MERLIN (2, 24). Under the variance component (VC) model, a phenotype is influenced by the additive effects of a trait locus (g), a residual familial background modeled as a pseudo-polygenic component (GR), and a residual nonfamilial component (r). The effects of the trait locus and the pseudo-polygenic component on the phenotype represent the locus-specific (h2g) and residual genetic (h2r) heritabilities. The linkage hypothesis is tested by restricting the trait locus heritability to zero. A likelihood ratio test contrasts the null hypothesis (h2g = 0) with the alternative (h2g estimated). The difference in –2 ln L (minus twice the log likelihood) between the null and alternate hypotheses is asymptotically distributed as a 50:50 mixture of a {chi}21 and a point mass at zero, and the P value is 1/2 of that associated with the {chi}2 value with 1 df (2). In regression-based linkage analysis (Reg), siblings who share a greater proportion of alleles identical-by-descent (IBD) at the marker locus will also show a greater resemblance in the phenotype. The phenotypic resemblance of the siblings, modeled as squared sums and squared differences of trait values of the sibling pairs, is linearly regressed on the estimated proportion of alleles that the sibling pair shares IBD at each marker locus (24).

The association tests were conducted in two steps. First, all HERITAGE subjects were included in the analysis. In the second step the families were divided in two groups based on family-specific logarithm of the odds ratio (LOD) scores from the linkage analyses: linkage-positive families were defined as those with family-specific LOD score ≥0.025, whereas linkage-negative families had family-specific LOD scores <0.025. The rationale for dividing the families in two groups based on family-specific LOD scores is that the associations underlying the original linkage signal should be stronger among the linkage-positive families, whereas no associations should be observed among subjects from linkage-negative families.

A {chi}2 test was used to verify whether the observed genotype frequencies were in Hardy-Weinberg equilibrium. The pair-wise linkage disequilibrium (LD) among the SNPs was assessed using the ldmax program available in the GOLD software package (3). Associations between the SNPs and SV50, Q50, and HR50 training responses were analyzed using a VC and likelihood ratio test based procedure in the QTDT software package (1). The total association model of the QTDT software utilizes a variance-components framework to combine phenotypic means model and the estimates of additive genetic, residual genetic, and residual environmental variances from a variance-covariance matrix into a single likelihood model (1). The evidence of association is evaluated by maximizing the likelihoods under two conditions: the null hypothesis (L0) restricts the additive genetic effect of the marker locus to zero (βa = 0), whereas the alternative hypothesis does not impose any restrictions on βa. The quantity of twice the difference of the log likelihoods between the alternative and the null hypotheses {2[ln (L1) – ln (L0)]} is distributed as {chi}2 with 1 df (difference in number of parameters estimated). To test whether an associated SNP explained the observed linkage in part or in full, a combined linkage and association analysis was performed using a special option of QTDT by modeling genetic variance components and allelic transmission deviation simultaneously (1). If the variant is the sole functional polymorphism influencing the trait in this region, then the P value should markedly weaken in the conditional analysis.

Bivariate correlation analysis between adjusted residuals of baseline SV50 and {Delta}SV50 were done using a VC model implemented in the computer program SOLAR version 4.0.7. (4). A maximum likelihood VC method was used to partition the phenotypic variance into two components: additive familial effect (polygenic heritability, h2) and nonfamilial effect (environmental component). The observed covariances between two subjects within a pedigree were compared with the expected values based on the product of their coefficient of relationship (twice the kinship coefficient). The significance of the heritability was tested using a likelihood ratio test comparing the null hypothesis of no genetic effect (h2 = 0) with the alternative hypothesis estimating h2. The bivariate model partitioned the phenotype correlation ({rho}p) into additive genetic ({rho}g) and random environmental ({rho}e) correlations.

Differences in normalized promoter activity data (phase 1) were analyzed by the Student t-test in Excel, whereas normalized promoter data (phase 2) were analyzed using a general linear model. In the latter case, the constructs (haplotypes), the experiments, and the replicates within each experiment were first modeled as main effects. Second, when the construct main effect was statistically significant, post hoc analyses were performed to test for pair-wise differences among constructs.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Linkage Studies
The strongest signal in the original linkage scan for SV50 training response ({Delta}SV50) phenotype was identified on chromosome 10p11 (VC LOD = 1.96; Reg LOD = 2.69) (18). In the first stage of the fine mapping process we genotyped six additional microsatellite markers on the target region. The additional markers slightly increased the linkage signal (VC LOD = 2.19; Reg LOD = 3.04) and defined the target region within 7 Mb between 30 and 37 Mb. Linkage results using the denser marker set on chromosome 10 are summarized in Fig. 1. The latest version of the NCBI sequence map (build 36) contains 42 genes in the target region: 16 of them are known genes, 12 are pseudogenes, and 14 are predicted genes.


Figure 1
View larger version (9K):
[in this window]
[in a new window]

  		Fig. 1. Multipoint linkage of stroke volume during steady-state exercise at 50 W (SV50) training response on chromosome 10 in white families of the HERITAGE Family Study. The results are derived from regression-based linkage method implemented in the MERLIN software package using microsatellite panel that included 6 additional markers on the original quantitative trait locus (QTL) region on 10q11.2. LOD, logarithm of the odds ratio.

 
Association Studies
A total of 90 SNPs located in the known genes were genotyped and tested for associations (Supplemental Table S1). Among all white HERITAGE subjects, SNPs in the kinesin family member 5B (KIF5B; Entrez GeneID: 3799) and integrin beta 1 (ITGB1; Entrez GeneID: 3688) gene loci showed suggestive associations with {Delta}SV50 (0.05 P > 0.03) (Supplemental Table S1). 1 However, among the linkage-positive families (see detailed description of the families below), eight SNPs within the KIF5B locus were associated with {Delta}SV50 (0.05 > P > 0.001), four of them being in absolute LD. On the other hand, none of the ITGB1 SNPs showed associations with {Delta}SV50 (P > 0.05). Among the KIF5B SNPs, the strongest association (P = 0.0012) was seen with a cluster of four markers in complete LD (rs211302, rs211286, rs172431, rs398686). The marker explained 7.3% of the {Delta}SV50 variance among the linkage-positive families (Fig. 2). The same pattern of associations with the KIF5B SNPs was also detected for {Delta}Q50 (0.05 > P > 0.007), whereas {Delta}HR50 was not associated with any of the SNPs in the region. To test if the KIF5B SNPs were causal or in LD with causal variants, conditional linkage analysis was done using two approaches. First, both regression and VC based linkage analysis with microsatellite markers were repeated among the linkage-positive families. The LOD scores from the regression model were 7.40 and 4.94, and from the VC model 5.00 and 3.80 before and after, respectively, adjustment for SNP rs211302. Second, using the QTDT program and the SNP data, we performed linkage analysis with and without simultaneous modeling of associations (linkage vs. linkage + association models). Again, for the SNPs that showed the strongest associations in single SNP tests, the evidence of linkage weakened considerably when also association term was included in the model (Supplementary Table S2). Weakening of the linkage signal in both approaches strongly suggests that the KIF5B variants explain considerable portion of the original linkage signal. None of the three training response phenotypes were associated with any of the SNPs among linkage-negative families (Supplemental Table S1).


Figure 2
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 2. Association between SV50 training response and KIF5B SNP rs211302 (–835) in linkage-positive and linkage-negative subjects.

 
Based on the maximal linkage signal on chromosome 10p11, the white HERITAGE families were divided in two groups: linkage-positive families (37 families with family-specific LOD >0.025 at peak linkage location) and linkage-negative families (the remaining families with family-specific LOD ≤0.025). Neither SNP allele or genotype frequencies nor {Delta}SV50 differed between linkage-positive and linkage-negative subjects. However, the linkage-positive subjects had greater mean SV50, BSA, and VO2max, and lower HR50 than the linkage-negative subjects at baseline (Table 1). The differences were more pronounced among offspring. In addition, the frequency of males and male sib-pairs was higher among the linkage-positive families (Table 1).


View this table:
[in this window]
[in a new window]

  		Table 1. Characteristics of the subjects from linkage-positive and linkage-negative families

 
Bivariate correlation analyses with baseline SV50 and {Delta}SV50 revealed an interesting difference between linkage-positive and -negative subjects (Supplemental Table S3). As expected, phenotypic correlations were zero because the training response trait was adjusted for the baseline SV50 value. However, both genetic and environmental correlations in linkage-positive families were greater than in linkage-negative families (Supplemental Table S3). In addition, univariate maximal heritability of {Delta}SV50 was higher among linkage-positive families, while baseline SV50 heritabilities were similar in linkage-positive and -negative families (Supplemental Table S3).

KIF5B Tissue Expression
To determine the tissue expression profile of the human KIF5B gene, two premade blots containing total RNA from cardiac tissues or from multiple human tissues were used. A predominant band corresponding approximately to 5.5 kb was evident on both blots (Fig. 3). Almost every cardiac tissue expressed KIF5B with perhaps a stronger expression in the adult heart and skeletal muscle. On the CVD blot, a higher band corresponding to ~6 kb was also evident, and it could be due to nonspecific hybridization or to a yet-to-be determined transcript of KIF5B.


Figure 3
View larger version (69K):
[in this window]
[in a new window]

  		Fig. 3. Tissue profile of the human KIF5B by Northern analysis. Two commercial blots (Clontech, Palo Alto, CA) were used: multiple human tissues (MTB) and cardiac (CVD). MTB, multiple tissue Northern blot, Lanes 1–12: brain (whole), heart, skeletal muscle, colon, thymus, spleen, kidney, liver, small intestine, placenta, lung, leukocyte. CVD, cardiovascular blot, Lanes 1–8: fetal heart, adult heart, aorta, apex of the heart, atrium left, atrium right, ventricle left, ventricle right. The KIF5B transcript is indicated with angled arrows.

 
Resequencing of the KIF5B Locus
The KIF5B gene was screened for DNA sequence variations by resequencing the coding (all exons and at least 250 bp of the flanking introns) and regulatory [2 kb of the putative promoter and 3'-untranslated region (UTR) until poly-A signal] regions in 95 individuals from the linkage-positive families. We identified 36 sequence variants (Supplemental Table S4). Two of the SNPs were located in exons (exon 6 +22 T/C -> Val/Ala; exon 16 +19 A/G -> Met/Val), but only one heterozygote was identified for both variants. In addition, one novel SNP was detected in exon 25, an A/G transition located 17 bp after the stop codon. The promoter region of the KIF5B turned out to be quite polymorphic. We identified 10 DNA sequence variants within 1850 bp from the start codon, four of which were novel. Two SNPs that were already included in the dbSNP database (rs211302, rs211300) showed complete pair-wise LD (r2 = 1.0). None of the new sequence variants (i.e., r2 <0.80 with previously genotyped SNPs) were associated with SV50 training response.

KIF5B Promoter Functional Studies
Human EST assembly revealed that KIF5B transcripts (37 ESTs) start within a narrow region between –470 and –416 upstream of the ATG, adding up to 156 bp sequence upstream of the start of the reference transcript NM_004521 (Fig. 4A). Several promoter prediction programs did not reveal a TATA-box within 500 bp but highlighted this area as being very GC-rich. This area lies in the middle of a predicted long CGI spanning 956 bp upstream of the ATG with a GC content of 63.4% and a ratio of observed CpG to expected CpG nucleotides of 0.854 (Fig. 4B).


Figure 4
View larger version (12K):
[in this window]
[in a new window]

  		Fig. 4. KIF5B in silico promoter analysis. A: expressed sequence tag assembly shows that KIF5B transcripts all initialize in a discrete locus ~100 bp upstream of the start of the NCBI reference sequence for the KIF5B mRNA (dotted arrow). The initiation positions are indicated by arrows and the number of transcripts are written above. The position of the single nucleotide polymorphism (SNP) rs1221445 at –444 is highlighted. B: 2000 bp around the KIF5B promoter was analyzed for the presence of CpG islands. CpG dinucleotides are depicted as vertical lines, the identified CpG island is marked with a solid line, and the transcription start site area with an arrow. Numbering of the nucleotides in both panels is in relation to the start codon (i.e., nucleotide preceding ‘A’ of the ATG codon is –1).

 
In a first series of functional experiments, short promoter constructs containing only the main SNP rs211302, which showed significant association with SV50 training response, and its near neighbor rs211301 were assessed for activity. The three naturally occurring haplotype constructs were cloned into a luciferase reporter vector. The constructs were tested in the mouse skeletal muscle C2C12 (Fig. 5A) and the Chinese hamster ovary (CHO) (Fig. 5B) cell lines. The data show an identical pattern for each haplotype in both cell lines, but the effect was more prominent in the CHO cells possibly due to a higher transfection efficiency. Essentially, the rs211302 C allele resulted in lower promoter activity. The same allele was associated with lower {Delta}SV50 in the linkage-positive families (Fig. 2).


Figure 5
View larger version (20K):
[in this window]
[in a new window]

  		Fig. 5. Functional analyses with KIF5B variants rs211302 (–835) and rs211301 (–809) Transient transfection data of KIF5B haplotype promoter reporter constructs in C2C12 (A) and CHO (B) cells.

 
The second phase of experiments aimed to extend these findings by testing the effect of KIF5B promoter SNPs in naturally occurring haplotypes using longer constructs. Initially, KIF5B promoter constructs spanning –850/–73 (not including any TATA box) and –1866/–73 (1794 bp) were tested for promoter activity, and both showed comparable strong activity over the empty vector (data not shown). Therefore, the 1794-bp construct was used in the subsequent experiments. Haplotype analysis of eight SNPs with a minor allele frequency >0.03 in the promoter region in whites revealed seven haplotypes (Fig. 6A). These haplotypes were cloned and tested for promoter activity in two cell lines. Strong promoter activity was found in both the C2C12 and HEK-293 cell lines, 598 and 103 times higher than the empty vector, respectively (Fig. 6B and Supplemental Fig. S1). For both cell lines, a significant difference (P < 0.0001) in promoter activities between haplotype constructs was observed. Compared with haplotype 2, which contains all common alleles, haplotypes 4 and 5, which contain rare alleles at rs211302 and position –603, showed lower promoter activity in C2C12 cells (P = 0.005 and P < 0.0001). The construct for haplotype 3, which is defined by the rare allele at rs12251445, also showed lower promoter activity than the common haplotype (P = 0.02). Haplotypes 6 and 7, sharing the rare allele at position rs12412654, showed similar activities as the common haplotype construct. In HEK-293 cells, the same three haplotypes (3, 4, and 5) and haplotype 1 showed different activity compared with the common haplotype 2 (0.0001 P < 0.0003) (Supplementary Fig. S1). However, unlike in the C2C12 cells, the activity was consistently higher than the common haplotype in the HEK-293 cells. Haplotype 3 in the HEK-293 cells was 25% more active than the common haplotype and highly significantly different (P < 0.0001) compared with all other haplotype constructs.


Figure 6
View larger version (31K):
[in this window]
[in a new window]

  		Fig. 6. Functional analyses of KIF5B promoter haplotypes. A: summary of KIF5B promoter haplotypes listing the positions relative to the ATG, the dSNP rs-number, and the frequency. The rare alleles are boxed. B: transient transfection data of KIF5B haplotype promoter reporter constructs and the empty vector in C2C12 cells. Shown are means ± SE of 4 independent experiments with each construct in triplicates (n = 12). Haplotype constructs that share rare alleles are depicted in the same color, haplotype 2 (open bar) contains all common alleles. Post hoc pair-wise comparison P values are written above the bars. ***P < 0.0001 against all other constructs.

 
Supplement Table S5 characterizes the individual SNPs used in the haplotype construction in regard to predicted changes in transcription factor (TF) binding of the rare allele using four different programs with stringent criteria (Alibaba, Matinsepector, Consite, and Algen Promo) as well as changes in nucleotide composition. Consistent functional consequences of rare alleles are predicted for position –603 and for rs12251445, which is located in the region where transcription starts (Fig. 4A).

Molecular Experiments
To evaluate the functional effects of KIF5B on mitochondrial localization and biogenesis, we used two approaches: 1) siRNA against endogenous Kif5b in undifferentiated muscle C2C12 cells, and 2) overexpression of Kif5b in undifferentiated muscle C2C12 cells.

Use of antisense probes against Kif5b in undifferentiated C2C12 cells resulted in a significant reduction of endogenous expression of Kif5b (Fig. 7). This reduction of Kif5b was initially visualized by fluorescent microscopy using GFP to assess the number of cells being transfected (Fig. 7A). It is noticeable that the transfected cells showed reduced Kif5b protein levels and diminished mitochondrial staining. The mitochondria in the transfected cells also showed perinuclear accumulation. This was not the case when scrambled antisense primers were used as a control (Fig. 7B). To quantify this effect, Kif5b expression was measured at the mRNA (Fig. 7C) and protein (Fig. 7D) levels. In both instances, Kif5b expression was significantly reduced.


Figure 7
View larger version (58K):
[in this window]
[in a new window]

  		Fig. 7. Effects of kif5b inhibition on mitochondrial content of C2C12 cells. A: immunostaining of cells transfected with negative control scrambled RNA probes (scrRNA), shown in bright green by green fluorescence protein (GFP), had no effect on kif5b expression nor mitochondrial staining. B: cells transfected with antisense siRNA primers against mouse kif5b, shown in bright green by GFP display reduced Kif5b expression and mitochondrial staining. C: siRNA against kif5b led to significant reduction of kif5b mRNA and protein levels (D). E: Ki5fb inhibition with siRNA reduced mitochondrial biogenesis as measured by the decrease of mitochondrial cytochrome b in a dose-dependent manner. *P < 0.05; ***P < 0.001.

 
The experimental approach was extended to determine the opposite effect of Kif5b silencing on mitochondria by overexpressing Kif5b in transiently transfect undifferentiated C2C12 cells. Overexpression of Kif5b was detected in the transfected cells by immunocytochemistry and cotransfection with a GFP construct (Fig. 8A). The illustrated transfected cell showed an increase of Kif5b protein that was accompanied by an increase of mitochondrial content as reflected by the increased signal (Fig. 8A). This was not the case in the transfected cells as identified by GFP when the empty vector (pcDNA3.1) was used as a negative control (Fig. 8B). As expected, mitochondrial content was not altered in the transfected cells. Overexpression of Kif5b was also confirmed at the mRNA (Fig. 8C) as well as the protein (Fig. 8D) levels.


Figure 8
View larger version (47K):
[in this window]
[in a new window]

  		Fig. 8. Effects of kif5b overexpression on mitochondrial content of C2C12 cells. A: immunostaining of cells transfected with the negative control empty vector pcDNA3.1 had no effect on kif5b expression and mitochondrial content. Transfected cells are shown in bright green by GFP. B: Kif5b expression and mitochondrial staining were significantly increased in cells transfected with the kif5b expression construct, identified by the GFP. C: overexpression of kif5b led to significant increase of kif5b mRNA and protein levels (D). E: transient transfection of cells with the kif5b expression construct increased mitochondrial biogenesis as measured by the increase of mitochondrial cytochrome b adjusted by nuclear DNA. *P < 0.05; **P < 0.01.

 
In addition to the immunocytochemistry, the effects of Kif5b inhibition or overexpression on mitochondrial biogenesis were confirmed quantitatively by measuring the DNA content of cytochrome b, which is encoded by mitochondrial DNA (mtDNA). We found that siRNA against Kif5b resulted in significant reduction of cytochrome b DNA content in a dose-dependent fashion. In the opposite experiment, overexpression of Kif5b by transient transfection of cells with the pcDNA3.1-Kif5b construct resulted in significant increase of cytochrome b DNA content.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
The data presented here show that the KIF5B locus is associated with endurance training-induced changes in submaximal exercise stroke volume. The association was particularly strong among families with positive evidence of linkage within the QTL on chromosome 10p11. Resequencing of the KIF5B showed that the gene locus is characterized by strong LD clusters that span the gene. Moreover, the promoter region of KIF5B was highly polymorphic, and in vitro studies indicated that the promoter variants affect the KIF5B promoter activity. Functional analysis of molecular and biochemical properties of KIF5B showed that the endurance training-induced changes may be due to the effects of KIF5B on mitochondrial biogenesis. Indeed, overexpression and gene silencing studies showed that KIF5B plays a role in determining location and number of mitochondria in cell models.

We focused on the association analyses on those families that contributed the most to the linkage signal. Interestingly, there were no differences in genotype frequencies or SV50 training response between linkage-positive and linkage-negative families. However, the linkage-positive subjects had higher SV50 and lower HR50 at baseline than the linkage-negative subjects. Furthermore, maximal heritability of SV50 training response as well as genetic and environmental correlations between baseline SV50 and SV50 training response were greater in linkage-positive subjects compared with linkage-negative subjects. The greater baseline level of SV50 as well as the positive genetic correlation between baseline and (baseline-adjusted) training response suggests that the linkage-positive families form a subgroup with a large submaximal exercise stroke volume already in the sedentary state but who are still able to improve their stroke volume values with regular exercise.

Resequencing of the KIF5B revealed that the gene locus is characterized by large LD clusters. The largest cluster contained 12 sequence variants and reaches from the 5'-UTR (rs211302) to the 3-UTR covering 45 kb of genomic DNA. The HapMap Caucasian data revealed an additional 24 SNPs belonging in the same cluster and expanded the cluster 20 kb downstream of KIF5B. Thus, the combined HERITAGE and HapMap data contained 36 SNPs over 65 kb region in very strong LD (r2 ≥ 0.9, D' = 1.0). Interestingly, the SNPs that showed the strongest associations in the linkage-positive families belong to this LD cluster. Our functional analyses showed that SNP rs211302 (–835) modified the promoter activity providing a potential mechanism by which the SNP may affect KIF5B expression levels. However, considering the size of the LD cluster, we cannot rule out the possibility that other variants of the cluster may also have functional properties (e.g., alternative splicing, mRNA stability, etc.).

Kinesins are a superfamily of microtubule-based molecular motors that play an integral role in the movement of various cellular organelles and vesicles. Conventional kinesins are heterotetrameres composed of two heavy chains and two light chains. The motor activity is produced by the heavy chain, while the exact contributions of light and heavy chains on the cargo binding are still unclear. There is evidence that cargo is bound by the light chain alone, by the heavy chain alone, and by an interaction between the light and heavy chain (15). KIF5B encodes a common conserved isoform of kinesin heavy chain that is expressed in various tissues, including cardiac myocytes. KIF5B-containing kinesin motors have been shown to participate in transport of Kv1 K+ channels and active zone-components essential for presynaptic assembly in axons (10, 21), transport of mitochondria in neurons and other cell types (14, 19), and movement of glucose transporter 4 (Glut4) to the cell membrane after insulin stimulation in adipocytes and skeletal muscle cells (23).

This study identified and functionally tested the KIF5B promoter. We found no evidence of 5' noncoding exons. In silico analysis indicated that the promoter initializes from a discrete 50 bp region and does not consist of a TATA box but is part of a prominent CGI. CGI promoters have been associated with housekeeping genes for several decades. More recently, the distinction of all four possible combinations between CGI and TATA containing promoters was established and revealed that between 50–60% of all promoters are CGI+/TATA– promoters (22). Depending on the class of promoter, conclusions about tissue specificity and gene function can be deducted. For the CGI+/TATA– subclass of promoters, widespread tissue expression and involvement in intracellular transport as an overrepresented Gene Ontology function was predicted (22). Both the Northern blot data presented here and the known functions of KIF5B (10, 14, 23) support this finding.

The promoter activity of transiently transfected KIF5B promoter reporter constructs was between 103- and 773-fold higher than the empty vector in the cell lines tested. Compared with activity of >150 promoters analyzed in studies using the same vector and one of the same cell lines, the KIF5B promoter activity belongs in the upper range of these promoters (12, 27).

Statistically significant differences in promoter activity among haplotypes were found. An independent effect of SNP rs211302 on promoter activity was found with constructs specifically testing this variant. The minor allele resulted in lower promoter activity, which is concordant with the observed association of the same allele with a lower SV50 training response. Additionally, the importance of rs211302 was confirmed in haplotype constructs. Haplotypes with rare alleles at rs211302 (–835) and position –603 showed consistent differences in promoter activity compared with the common haplotype in both cell lines. In C2C12 cells, they resulted in lower promoter activity, which is concordant with the observed association between the minor allele of rs211302 and a lower SV50 training response. Additionally, the haplotype defined by the SNP rs12251445 at position –444 had different promoter activity compared with the common haplotype in both cell lines, reaching 25% in HEK-293 cells. In silico analysis revealed that all programs that were used to predict changes in transcription factor binding highlighted this position and that at –603 as the most likely to be affected. Both SNPs change the GC content in or near the critical region where transcription is thought to initiate. This may be especially important for the CGI+/TATA– promoter of KIF5B as this kind of promoter is driven by its GC-richness. However, neither SNP at position –603 nor rs12251445 was associated with the SV50 training response. Several SNPs in the KIF5B promoter influence the GC content or create an additional CpG dinucleotide that may be accessible for methylation and thereby offer other mechanisms independent of transcription factor binding to affect promoter activity. In fact, according to a recent study, only 33% of functional SNPs lie in a known TF consensus sequence (9).

Most of the remaining significant differences among haplotypes were <10% and thus close to the experimental error level. We took great care to minimize the factors contributing to unwanted variation. Factors such as batch sample processing to control for equal DNA purity/condition, multiple DNA concentration measurements to guarantee equal starting amounts, internal transfection control to normalize well-to-well transfection efficiency variation, and experimental design (4 independent experiments with each construct in triplicate) to reduce random observations were enforced. Additionally, haplotypes of similar composition behaved similarly compared with the common haplotype, highlighting a consistent pattern. Finally, only haplotypes of SNPs in the proximal but not the distal promoter region showed statistically significant differences in promoter activity compared with the common allele haplotype in both cell lines. This is in line with a recent study, in which an inverse relationship between the likelihood of an SNP's functionality and the distance to the core promoter was found (9). Nevertheless, it remains to be seen if the small differences reported here result in biologically meaningful consequences.

The overexpression of KIF5B in heart muscle and its various compartments, coupled with its known functional properties, make KIF5B a compelling candidate for endurance training-induced changes in submaximal exercise stroke volume. We used an in vitro muscle cell line model (C2C12) to evaluate further its functional significance. Silencing of Kif5b with siRNA resulted in reduction of endogenous expression of Kif5b, which was accompanied by reduction of mitochondrial fluorescence in the transfected cells. siRNA-transfected cells also displayed perinuclear accumulation of mitochondria, which has previously been reported. In contrast, overexpression of Kif5b in C2C12 cells led to an increase of mitochondrial fluorescence. It has been shown that exercise training leads to mitochondrial biogenesis by altering the rates of PGC-1{alpha} and Tfam expression, modifying mitochondrial fission and fusion mechanisms and influencing the import of nuclear-derived gene products into the mitochondrion (16). We therefore measured cytochrome B by semiquantitative real-time PCR to evaluate the effects of Kif5b on mitochondrial biogenesis. We found that siRNA-treated cells had a reduced amount of mitochondria while cells that were transfected by the Kif5b overexpression construct had an increased amount of mitochondria. These data show that endurance training-induced changes in submaximal exercise stroke volume could be associated with changes in cardiac mitochondrial functions as a result of KIF5B expression levels.

Linkage analysis and subsequent association studies have identified KIF5B as a strong candidate gene influencing the response of stroke volume to regular exercise. Functional SNPs in the promoter (especially rs211302 at position –835), which could further affect the expression levels of KIF5B in heart muscle, were identified. Our in vitro data support the hypothesis that exercise training-induced changes in submaximal exercise stroke volume may be due to cardiac tissue mitochondrial function and variation in KIF5B expression as determined by functional SNPs in its promoter.


   GRANTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
The HERITAGE Family Study is supported by National Heart, Lung, and Blood Institute Grant HL-45670 (Claude Bouchard, PI). Claude Bouchard is partially supported by the George A. Bray Chair in Nutrition.


   FOOTNOTES
 
Address for reprint requests and other correspondence: T. Rankinen, Human Genomics Laboratory, Pennington Biomedical Research Center, 6400 Perkins Rd., Baton Rouge, LA 70808-4124 (e-mail: rankint{at}pbrc.edu).

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

* G. Argyropoulos and A. M. Stütz contributed equally to this work. Back

1 The online version of this article contains supplemental material. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Abecasis GR, Cardon LR, Cookson WO. A general test of association for quantitative traits in nuclear families. Am J Hum Genet 66: 279–292, 2000.[CrossRef][Web of Science][Medline]
  2. Abecasis GR, Cherny SS, Cookson WO, Cardon LR. Merlin–rapid analysis of dense genetic maps using sparse gene flow trees. Nat Genet 30: 97–101, 2002.[CrossRef][Web of Science][Medline]
  3. Abecasis GR, Cookson WO. GOLD–graphical overview of linkage disequilibrium. Bioinformatics 16: 182–183, 2000.[Abstract/Free Full Text]
  4. Almasy L, Blangero J. Multipoint quantitative-trait linkage analysis in general pedigrees. Am J Hum Genet 62: 1198–1211, 1998.[CrossRef][Web of Science][Medline]
  5. An P, Rice T, Gagnon J, Leon AS, Skinner JS, Bouchard C, Rao DC, Wilmore JH. Familial aggregation of stroke volume and cardiac output during submaximal exercise: The HERITAGE Family Study. Int J Sports Med 21: 566–572, 2000.[CrossRef][Web of Science][Medline]
  6. Bai F, Rankinen T, Charbonneau C, Belsham DD, Rao DC, Bouchard C, Argyropoulos G. Functional dimorphism of two hAgRP promoter SNPs in linkage disequilibrium. J Med Genet 41: 350–353, 2004.[Abstract/Free Full Text]
  7. Berthiaume J, Wallace KB. Perfluorooctanoate, perflourooctanesulfonate, and N-ethyl perfluorooctanesulfonamido ethanol; peroxisome proliferation and mitochondrial biogenesis. Toxicol Lett 129: 23–32, 2002.[CrossRef][Web of Science][Medline]
  8. Bouchard C, Leon AS, Rao DC, Skinner JS, Wilmore JH, Gagnon J. The HERITAGE family study. Aims, design, and measurement protocol. Med Sci Sports Exerc 27: 721–729, 1995.[Web of Science][Medline]
  9. Buckland PR, Hoogendoorn B, Coleman SL, Guy CA, Smith SK, O'Donovan MC. Strong bias in the location of functional promoter polymorphisms. Hum Mutat 26: 214–223, 2005.[CrossRef][Web of Science][Medline]
  10. Cai Q, Pan PY, Sheng ZH. Syntabulin-kinesin-1 family member 5B-mediated axonal transport contributes to activity-dependent presynaptic assembly. J Neurosci 27: 7284–7296, 2007.[Abstract/Free Full Text]
  11. Chagnon YC, Borecki IB, Perusse L, Roy S, Lacaille M, Chagnon M, Ho-Kim MA, Rice T, Province MA, Rao DC, Bouchard C. Genome-wide search for genes related to the fat-free body mass in the Quebec family study. Metabolism 49: 203–207, 2000.[CrossRef][Web of Science][Medline]
  12. Coleman SL, Buckland PR, Hoogendoorn B, Guy C, Smith K, O'Donovan MC. Experimental analysis of the annotation of promoters in the public database. Hum Mol Genet 11: 1817–1821, 2002.[Abstract/Free Full Text]
  13. Collier CR. Determination of mixed venous CO2 tensions by rebreathing. J Appl Physiol 9: 25–29, 1956.[Abstract/Free Full Text]
  14. Glater EE, Megeath LJ, Stowers RS, Schwarz TL. Axonal transport of mitochondria requires milton to recruit kinesin heavy chain and is light chain independent. J Cell Biol 173: 545–557, 2006.[Abstract/Free Full Text]
  15. Goldstein LS, Philp AV. The road less traveled: emerging principles of kinesin motor utilization. Annu Rev Cell Dev Biol 15: 141–183, 1999.[CrossRef][Web of Science][Medline]
  16. Joseph AM, Pilegaard H, Litvintsev A, Leick L, Hood DA. Control of gene expression and mitochondrial biogenesis in the muscular adaptation to endurance exercise. Essays Biochem 42: 13–29, 2006.[Web of Science][Medline]
  17. Mayfield DK, Brown AM, Page GP, Garvey WT, Shriver MD, Argyropoulos G. A role for the Agouti-Related Protein promoter in obesity and type 2 diabetes. Biochem Biophys Res Commun 287: 568–573, 2001.[CrossRef][Web of Science][Medline]
  18. Rankinen T, An P, Perusse L, Rice T, Chagnon YC, Gagnon J, Leon AS, Skinner JS, Wilmore JH, Rao DC, Bouchard C. Genome-wide linkage scan for exercise stroke volume and cardiac output in the HERITAGE Family Study. Physiol Genomics 10: 57–62, 2002.[Abstract/Free Full Text]
  19. Rice SE, Gelfand VI. Paradigm lost: milton connects kinesin heavy chain to miro on mitochondria. J Cell Biol 173: 459–461, 2006.[Abstract/Free Full Text]
  20. Rice T, Borecki IB, Bouchard C, Rao DC. Commingling analysis of regional fat distribution measures: the Quebec family study. Int J Obes Relat Metab Disord 16: 831–844, 1992.[Web of Science][Medline]
  21. Rivera J, Chu PJ, Lewis TL Jr, Arnold DB. The role of Kif5B in axonal localization of Kv1 K(+) channels. Eur J Neurosci 25: 136–146, 2007.[Web of Science][Medline]
  22. Schug J, Schuller WP, Kappen C, Salbaum JM, Bucan M, Stoeckert CJ Jr. Promoter features related to tissue specificity as measured by Shannon entropy. Genome Biol 6: R33, 2005.[CrossRef][Medline]
  23. Semiz S, Park JG, Nicoloro SM, Furcinitti P, Zhang C, Chawla A, Leszyk J, Czech MP. Conventional kinesin KIF5B mediates insulin-stimulated GLUT4 movements on microtubules. EMBO J 22: 2387–2399, 2003.[CrossRef][Web of Science][Medline]
  24. Sham PC, Purcell S, Cherny SS, Abecasis GR. Powerful regression-based quantitative-trait linkage analysis of general pedigrees. Am J Hum Genet 71: 238–253, 2002.[CrossRef][Web of Science][Medline]
  25. Shen X, Zheng S, Thongboonkerd V, Xu M, Pierce WM Jr, Klein JB, Epstein PN. Cardiac mitochondrial damage and biogenesis in a chronic model of type 1 diabetes. Am J Physiol Endocrinol Metab 287: E896–E905, 2004.[Abstract/Free Full Text]
  26. Skinner JS, Wilmore KM, Krasnoff JB, Jaskolski A, Jaskolska A, Gagnon J, Province MA, Leon AS, Rao DC, Wilmore JH, Bouchard C. Adaptation to a standardized training program and changes in fitness in a large, heterogeneous population: the HERITAGE Family Study. Med Sci Sports Exerc 32: 157–161, 2000.[Web of Science][Medline]
  27. Trinklein ND, Aldred SJ, Saldanha AJ, Myers RM. Identification and functional analysis of human transcriptional promoters. Genome Res 13: 308–312, 2003.[Abstract/Free Full Text]
  28. Wilmore JH, Farrell PA, Norton AC, Coté RW 3rd, Coyle EF, Ewy GA, Temkin LP, Billing JE. An automated, indirect assessment of cardiac output during rest and exercise. J Appl Physiol 52: 1493–1497, 1982.[Abstract/Free Full Text]
  29. Wilmore JH, Stanforth PR, Gagnon J, Rice T, Mandel S, Leon AS, Rao DC, Skinner JS, Bouchard C. Cardiac output and stroke volume changes with endurance training: The HERITAGE Family Study. Med Sci Sports Exerc 33: 99–106, 2001.[CrossRef][Web of Science][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figures and Tables
Right arrow All Versions of this Article:
36/2/79    most recent
00003.2008v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Argyropoulos, G.
Right arrow Articles by Rankinen, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Argyropoulos, G.
Right arrow Articles by Rankinen, T.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2009 by the American Physiological Society.