Evaluation of hypothalamic gene expression in mice divergently selected for heat loss

Stephanie R. Wesolowski, Mark F. Allan, Merlyn K. Nielsen and Daniel Pomp

Department of Animal Science, University of Nebraska, Lincoln, Nebraska 68583-0908


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Mouse lines divergently selected for heat loss were evaluated for correlated responses in the hypothalamic transcriptome. High (MH) heat loss mice have ~50% greater heat loss, ~35% less body fat, ~20% greater feed intake, ~100% greater locomotor activity levels, and higher core body temperature compared with low (ML) heat loss mice. We evaluated hypothalamic expression between inbred lines derived from MH and ML lines (IH and IL, respectively) using cDNA microarrays and selected genes previously isolated in a large differential-display PCR experiment. Northern analysis was used to confirm differences, revealing higher hypothalamic mRNA expression of oxytocin (Oxt) and tissue inhibitor of metalloproteinase 2 (Timp-2) in the IH line. Real-time PCR assays were developed for Oxt, Timp-2, and ribosomal protein L3 (Rpl3, previously found to be upregulated in IL) and confirmed differential expression of these genes with potential physiological relevance in energy balance. These results provide information on correlated responses in the transcriptome of mice selected for high and low energy expenditure and reveal new information regarding genetic regulation of energy balance.

obesity; microarrays; real-time polymerase chain reaction


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
ENERGY BALANCE is the difference between energy intake and energy expenditure. Imbalances in either direction can lead to body weight disregulation with potentially serious health consequences. Obesity is an increasingly serious medical problem in today’s society as ~60% of the American population is overweight (15, 35, 38). In domestic livestock species, feed costs are the largest economic inputs in food production systems (10, 24). Improving production efficiency, conserving feed resources, and producing leaner products for consumers are key priorities. Understanding the genetic regulation of energy balance will facilitate development of diagnostic and therapeutic advances to improve human health and enhance efficiency and quality of food products (15, 32, 38).

Energy balance is a complex trait regulated by the combined effects of numerous genes, environmental factors, and multi-way interactions. Several monogenic rodent obesity models produce phenotypes indicative of altered energy metabolism (34). Although several quantitative trait loci (QTL) for traits related to energy balance have been detected, none has been cloned and identified (33). The complex genetic architecture of energy balance remains largely uncharacterized.

To contribute to an understanding of the complexity of energy balance at the polygenic level, we have employed gene expression techniques to find hypothalamic transcriptional differences between two unique mouse lines created by divergent selection based on direct calorimetry measurement of heat loss (29). The high heat loss line (MH) has ~50% higher heat loss, ~20% greater feed intake, ~35% less body fat, ~100% greater locomotor activity, and elevated core body temperature compared with the low line (ML) despite no differences in body weight (25, 27, 29, 30). These lines are unique models to dissect the genetic complexity of energy balance and its related phenotypes on a population basis. The hypothalamus was selected for evaluation because of its role in controlling traits such as feeding behavior, weight gain, and adaptive thermogenesis (14, 22, 36). Our hypothesis was that response to selection for heat loss is mediated, to some extent, through direct and/or correlated responses in the hypothalamic transcriptome.

Specifically, we used replicated cDNA microarray experiments to identify and prioritize candidate genes. Candidate genes were also selected from the literature based on known roles in energy balance regulation and from a previous differential display project that evaluated hypothalamic expression differences between the IH and IL (inbred MH and ML, respectively) lines (1). Confirmation of differential gene expression was accomplished using Northern hybridization, and real-time PCR assays were designed to create sensitive and high-throughput quantification assays.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Mice.
Mice used in this study were from partially inbred lines derived from long-term selection for high (MH) and low (ML) heat loss (29). Briefly, selection was initiated from a composite base population using direct calorimetry measurements of heat loss in 9- to 11-wk males for 16 generations in three separate replicates. Means for heat loss in the MH and ML lines were 179.1 and 107.5 (kcal per kg0.75 per day), respectively, and realized heritability was estimated to be 0.28. Full-sib matings were initiated at generation 21 from replicate 1 of each selection line to develop inbred high (IH) and inbred low (IL) lines. Mice used in this study were from generation 32 and 34 of the IH and IL lines. Mice were weaned at 3 wk, caged by sex, and provided ad libitum access to water and feed (Teklad 8604 rodent chow). Temperature was maintained at 22°C with relative humidity at 35–50%, and a 12:12 light-dark cycle with lights on at 0700 h was employed. All procedures and protocols were approved by the University of Nebraska-Lincoln Institutional Animal Care and Use Committee.

Collection of phenotypic data and tissues.
Heat loss was measured on male mice at 11–12 wk of age using direct, gradient-layer, individual-animal calorimeters (29). Briefly, mice were weighed and placed in calorimeters with 2.25 g feed at ~1700 h. Following a 30-min acclimation period, heat loss measurements were recorded every minute for a 15-h period. Data were adjusted for metabolic body weight and projected to a 1-day basis (kcal per kg0.75 per day). At 14 wk of age, mice were fasted 2–3 h and decapitated at 1000–1100 h after brief exposure to CO2. Hypothalamic samples were immediately dissected and snap frozen in liquid nitrogen as previously described (1).

RNA extraction.
Total RNA was isolated from hypothalamus samples using TRIzol LS reagent (Invitrogen Life Technologies, Carlsbad, CA). Pools of hypothalamus (4–5 samples) or a single hypothalamus were extracted following the manufacturer’s protocol. Single hypothalamic extractions were performed in the presence of 5 µl of glycogen (20 mg/ml) using phase-lock gel tubes (Eppendorf, Westbury, NY). Aliquots of RNA samples were loaded in agarose gels to visualize quality and integrity. RNA was quantified with a fluorometer (model TD-700; Turner Designs, Sunnyvale, CA) or plate reader (Wallac Victor2; Perkin-Elmer, Boston, MA) using the RiboGreen RNA quantitation kit (Molecular Probes, Eugene, OR).

Microarrays.
Two separate hybridizations were performed using the Mouse GEM 2.08 cDNA microarray (Incyte, St. Louis, MO). Twenty-five hypothalamic tissue samples collected from male mice in each line with the most extreme heat loss measurements were randomly allocated to six groups within line. Tissue samples within each group were pooled, extracted, and checked for quality and quantity as described above. Pooled RNA samples were mixed within each line to yield two pools per line of total RNA. Samples were purified for mRNA using two passes through Oligotex mRNA isolation columns (Qiagen, Valencia, CA). Aliquots (600 ng) of each pooled sample were sent to Incyte for array experiments. In the first experiment (replicate 1), the IH pooled sample was labeled with Cy3 dye and the IL pooled sample with Cy5. In replicate 2, different IH and IL pools were used, and dye labels were reversed. The array contained 9,596 sequence-verified cDNA clones. The ratio of the two fluorescent intensities, adjusted using controls, provided a quantitative measurement of gene expression (GemTools Software, v. 2.4, Incyte Genomics). Genes with significantly and suggestively different expression levels were defined to be those with absolute balanced differential expression (ABDE) ratios greater than or equal to 2.0 and 1.6, respectively.

Candidate genes.
Numerous genes in various pathways have been implicated in the regulation of energy balance based on identification of loci responsible for single-gene obesity phenotypes, gene expression, and transgenic model analyses (16, 34, 36). Genes that have been specifically shown to have altered hypothalamic mRNA expression resulting in phenotypic differences related to energy balance were selected as candidates for evaluation for hypothalamic transcriptional differences between the IH and IL lines. These genes included beacon (Beac) (5), neuropeptide Y (Npy) (36), metallothionein 1 (Mt1) (4), metallothionein 2 (Mt2) (4), Vgf nerve growth factor (Vgf) (34), and carboxypeptidase E (Cpe) (16). Genes were also selected from a differential-display PCR experiment that evaluated differences in hypothalamic gene expression between the IH and IL lines (1). Differentially displayed fragments from this project had already been cloned and sequenced (GenBank accession numbers AW358547AW358553 and BQ135237BQ135326). Genes selected from this analysis for use in our study included glutamine synthetase (Glns) and ribosomal protein S11 (Rps11), which had greater expression in the IL line, and ribosomal protein S10 (Rps10) and tissue inhibitor of metalloproteinase 2 (Timp-2), which had greater expression in the IH line.

Northern hybridization.
Northern blots were made with four pools of RNA samples (two per line). Five or 10 µg of total RNA was used following standard procedures (1). Probes for Northern hybridization were prepared for genes of interest using the Megaprime DNA random labeling kit (Amersham). PCR products for each target gene were amplified using primers designed from available GenBank sequences (Table 1) and cDNA prepared from pooled hypothalamic RNA samples. Products were cloned and sequenced to verify gene sequence specificity (University of Nebraska Core Sequencing Facility). For genes identified from the previous differential-display PCR experiment, fragments of interest were amplified from the cloned vector using the respective primers that generated the product. Briefly, cDNA was synthesized using Superscript II reverse transcriptase (Invitrogen) in the presence of RNasin (Promega, Madison, WI) following the manufacturer’s protocols. PCR products were purified using QiaQuick PCR purification kit (Qiagen), and 12.5 ng of purified product was used in a 25-µl labeling reaction following the manufacturer’s protocol. Hybridizations were done according to standard procedures (1). Evaluation of Northern results was accomplished using images of autoradiographs and densitometric values obtained using the Hitachi CCD Bio System and GeneTools Software (Hitachi Genetic Systems, Alameda, CA) or PhosphorImager System SF (Molecular Dynamics). Membranes were stripped with 0.1% SDS at 95°C and reprobed with glyceraldehyde phosphate dehydrogenase (Gapdh) for standardization.


View this table:
[in this window]
[in a new window]
 
Table 1. Primer sequences used to generate probes for genes evaluated with Northern hybridizations

 
Real-time PCR.
Real-time Taqman PCR assays were designed and optimized for quantification of oxytocin (Oxt) and ribosomal protein L3 (Rpl3) gene expression. Reagents and conditions previously reported for the tissue inhibitor of metalloproteinase 2 (Timp-2) gene were used (44). RNA samples from single hypothalamic extractions (500 ng) were treated with DNase I (Invitrogen) to remove any contaminating genomic DNA. Reactions contained 1x DNase I buffer and 100 U DNase I in a 10-µl volume. Reverse transcription was accomplished in a 20-µl volume using 200 ng RNA from DNase treatment, 1 x RT buffer, 5.5 mM MgCl2, 500 µM dNTP mixture, 2.5 µM oligo-d(T)16 primer, 8 U RNase inhibitor, and 25 U MultiScribe enzyme (Applied Biosystems, Foster City, CA). PCR quality checks for cDNA were performed using primers for D4Mit190 microsatellite marker to ensure no genomic DNA was present and primers for growth hormone releasing hormone (Ghrh) to ensure hypothalamic transcripts were extracted.

Real-time PCR reactions were performed in a 25-µl volume using 12.5 µl of 2 x Universal Master Mix without AmpErase UNG (Applied Biosystems), appropriate amounts of primers and probe, and 1 ng cDNA. Reactions were amplified in the ABI Prism 7700 sequence detector system (Applied Biosystems). PCR conditions were 50°C for 5 min, followed by 40 cycles with denaturation at 95°C for 15 s and a combined annealing and extension step at 60°C for 1 min. Primer and probe sequences were designed using Primer Express software (version 1.5a, Applied Biosystems) and GenBank sequences GI:200167 and GI:6113383537 for Oxt and Rpl3, respectively (Table 2). Amplicon specificity was verified by sequence analysis (University of Nebraska Core Sequencing Facility). Target gene probes were synthesized by Applied Biosystems with the reporter dye FAM attached to the 5' end. The Rodent Gapdh Control Reagents (Applied Biosystems) with VIC reporter dye were used to amplify Gapdh to evaluate and adjust for sample loading differences.


View this table:
[in this window]
[in a new window]
 
Table 2. Oligonucleotide sequences for target genes in real-time PCR assays

 
Optimal monoplex reaction conditions for Oxt, Rpl3, and Timp-2 were achieved with 200 nM of forward and reverse primers with 100 nM probe. Standard Gapdh concentrations of 100 nM forward and reverse primers and 200 nM probe were used. Duplex reactions for each target gene (Oxt, Rpl3, Timp-2) and Gapdh were optimized. For Oxt and Rpl3 assays, 150 nM of forward and reverse primers and 100 nM of probe were used with 75 nM Gapdh forward primer, 100 nM Gapdh reverse primer, and 200 nM Gapdh probe. For Timp-2, 200 nM of forward and reverse primers and 100 nM probe were duplexed with 50 nM Gapdh forward, 75 nM Gapdh reverse, and 200 nM Gapdh probe. To make accurate comparisons among samples and adjust target gene data for the control gene (Gapdh), the efficiency of the two PCR reactions must be similar. Dilutions of cDNA were run in triplicate using the optimal monoplex primer and probe concentrations. The average threshold cycle (CT) value for each dilution was plotted against the log of the input cDNA amount to evaluate linearity of the assay. To find the relative efficiency, the difference between average target gene CT and average control gene CT ({Delta}CT) for each dilution was plotted against the log input cDNA amounts. A slope with an absolute value <0.1 indicates that the efficiencies of the target gene and control gene reactions are similar. Deviations from this slope indicate that assays have different PCR efficiencies and that comparisons among samples would be unreliable.

For monoplex assays, eight males from each line with the most extreme heat loss measurements were used, and each sample was run in triplicate. Duplex assays for each target gene were run in duplicate, and assays were repeated. In the first run, 16 males and 4 females from each line were evaluated. In the second run, the same samples were used with 4 additional male samples per line for a total of 20 male and 4 female samples. The duplex data were pooled and analyzed.

Real-Time PCR data analysis.
Several methods were used to evaluate data from monoplex and duplex assays. The comparative CT method expresses data relative to a calibrator sample (18). Instead of using a calibrator sample, we expressed differences relative to the average {Delta}CT for all samples analyzed in each assay. The final normalized value for each sample is calculated with the expression 2-{Delta}{Delta}CT. Normalized values were calculated by dividing the target gene CT values by the control gene CT values. Covariate-adjusted values were obtained using control gene CT values as covariates for each target gene sample CT value in an analysis of variance. Data were analyzed using the GLM procedure of the SAS statistical package (SAS Institute, Cary, NC). The model included selection line as a fixed effect.

Gene mapping.
The map locations for all genes were identified by searching the Mouse Genome Database (http://www.informatics.jax.org) or LocusLink (http://www.ncbi.nlm.nih.gov/locuslink). The genetic map location of the Beacon gene (5) was unknown, so we mapped it using the T31 Mouse Radiation Hybrid Panel (http://www.jax.org/resources/documents/cmdata/rhmap/). The forward primer 5'-CACGATTTTTAAGGACCACG-3' and reverse primer 5'-AGCTCCAGGTTCATCCCATC-3' were used in a standard PCR reaction with 2.5 mM MgCl2 and 0.2 µM of each primer. Reactions were run with an annealing temperature of 60°C. PCR products were detected by gel electrophoresis, and data were submitted to the Jackson Laboratory Mouse Radiation Hybrid Database for analysis (http://www.jax.org/resources/documents/cmdata/rhmap/rhsubmit.html).


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Microarray experiments.
The microarray experiments revealed several potential candidate genes with differential hypothalamic gene expression between the IH and IL heat loss lines. There were 23 significantly different genes with ABDE values of 2.0 or greater in either replicate 1 or replicate 2 (Table 3). Only four of these significant genes had differential expression in the same direction in the other replicate. These genes included oxytocin (Oxt), integral membrane protein 2 (Itm2), and kinesin heavy chain member 1A (Kif1a), which had higher expression in the IH line, and fibroblast growth factor 1 (Fgf1), which had higher expression in the IL line. There were 109 suggestively significant genes with ABDE between 1.6 and 2.0 in either replicate 1 or replicate 2 (data not shown). Of these, only six displayed differential expression in the same direction in the other replicate (Table 3).


View this table:
[in this window]
[in a new window]
 
Table 3. Genes identified from microarray experiments with significant expression differences between the high (IH) and low (IL) heat loss lines

 
Northern hybridization analysis of candidate genes.
Candidate genes from the microarray experiments, previous differential display project, and literature were tested for confirmation of expression differences using Northern hybridization (Table 4). This analysis confirmed that Oxt and Timp-2 are differentially expressed in the hypothalamus of mice from the heat loss selection lines. Both genes had greater relative mRNA abundance in the IH line, consistent with results from the microarrays and, in the case of Timp-2, our previous differential display experiment (Figs. 1 and 2). None of the remaining candidate genes evaluated, including those selected from the literature and those selected from our previous differential display experiment, exhibited differential expression between the selection lines based on the sensitivity of Northern analyses (Fig. 3).


View this table:
[in this window]
[in a new window]
 
Table 4. Summary of genes evaluated for expression differences with Northern blotting and their respective chromosomal map positions

 


View larger version (71K):
[in this window]
[in a new window]
 
Fig. 1. Microarray expression results for oxytocin (Oxt) and tissue inhibitor of metalloproteinase 2 (Timp-2) genes. Oxt had an absolute balanced differential expression (ABDE) ratio of 2.6 in replicate 1 and 1.3 in replicate 2, with both indicating greater expression in the IH line. Timp-2 had an ABDE of 1.0 in replicate 1 and 2.0 in replicate 2, indicating higher expression in the IH line. The color scale represents relative levels of expression. IH and IL, inbred lines derived from high heat loss mice and low heat loss mice, respectively.

 


View larger version (80K):
[in this window]
[in a new window]
 
Fig. 2. Northern hybridization confirmation of differential hypothalamic gene expression for Oxt (A) and Timp-2 (B) between the heat loss selection lines. Upper transcripts represent Gapdh and lower represents either Oxt or Timp-2. The left two lanes in each blot contain different pooled IL samples, and the right two lanes contain pooled IH samples.

 


View larger version (48K):
[in this window]
[in a new window]
 
Fig. 3. Northern hybridization results for candidate genes. No differences were detected in hypothalamic mRNA expression between the IH and IL line. The left two lanes in each blot contain different pooled IL samples, and the right two lanes contain pooled IH samples.

 
Map positions.
A summary of the map positions for each gene is provided in Table 4. Using radiation hybrid mapping, we localized the position of Beac at 59.1 cR proximal of D9Mit59 and 37.6 cR distal to D9Mit223, with highest linkage (LOD 21.4) to the Pin1 gene. Based on the linkage map location of Pin1, Beac appears to be located at ~4 cM on MMU9. The map locations of two genes, Glns and Rps10, are currently ambiguous due to the existence of pseudogenes. The location of Hp1bp3 also remains unidentified.

Real-time PCR.
Sensitive, high-throughput, real-time PCR assays were designed, optimized, and run for Oxt, Timp-2, and Rpl3. The Rpl3 gene was previously shown to be upregulated in the hypothalamus and brown adipose tissue of IL mice (1, 2). Each assay was evaluated and found to be linear over six dilutions (Fig. 4A). Validation experiments were performed to determine the amplification efficiencies for each target gene relative to Gapdh. The absolute values of the slope for each comparison were less than 0.1, indicating equal efficiencies (Fig. 4B). Thus the Gapdh assay can be used to accurately adjust target gene data.



View larger version (20K):
[in this window]
[in a new window]
 
Fig. 4. Validation of real-time PCR Taqman assays. A: linear dynamic range and efficiency of real-time PCR assays. Dilutions of a pooled cDNA sample were run in triplicate at 5, 2, 1, 0.5, 0.2, and 0.1 ng amounts. The average threshold cycle (CT) value at each dilution was plotted against the log of the input cDNA amount. A slope of -3.3 represents ~100% PCR efficiency. B: relative efficiency of real-time PCR assays for each target gene compared with Gapdh control gene. Dilutions of cDNA were run for both target and control gene assays. At each dilution, the average {Delta}CT (difference between average target gene CT and average control gene CT) was plotted against the log of the input cDNA amount. A slope with an absolute value of less than 0.1 indicates equal PCR efficiencies.

 
Monoplex and duplex assays were optimized to verify differential expression between mice from the IH and IL lines (Table 5). The IH line had 0.27-fold higher expression of Oxt (P = 0.08) and 0.3-fold lower expression of Rpl3 (P = 0.005) compared with the IL line. The difference between the IH and IL lines for Timp-2 mRNA expression was only 0.07-fold, with greater expression in the high heat loss line. Similar relative mRNA expression differences were found using the normalization and covariate analyses. In the latter analyses, a higher value represents lower expression since a higher CT value indicates lower gene expression. These results confirmed the direction of differential expression of these genes compared with the microarray, differential display, and Northern blot results.


View this table:
[in this window]
[in a new window]
 
Table 5. Results of real-time PCR (Taqman) gene expression quantification assays

 
For the duplex assays, the difference in expression between the lines was 0.17-fold for Oxt (P = 0.05) and 0.23-fold for Rpl3 (P < 0.0001) with greater expression in the IH and IL lines, respectively. For Oxt, the covariate analysis detected greater expression in the IH line (P < 0.0001) and the normalized data showed greater expression in IH line, but this difference was not significant. The normalized and covariate adjusted analyses also indicated greater Rpl3 expression in the IL line (P = 0.0003 and P = 0.0008, respectively). Timp-2 showed higher relative mRNA abundance in the IH line, but a significant difference was detected only when using the covariate Gapdh adjustment (P = 0.04).

The three methods used to normalize target gene data with control gene data produced similar results in terms of the direction and relative magnitude of differential gene expression. The results of the monoplex and duplex assays for each gene showed similar expression differences, indicating that the duplex reactions can quantify mRNA expression levels for these genes as accurately as the monoplex assays. Furthermore, to ensure that each assay accurately measured mRNA expression levels and not residual genomic DNA that could remain after DNase I treatment, DNase I-treated RNA samples without reverse transcription were amplified for each gene assay. Only faint amplification was detected in some samples. The CT values for samples with amplification were all greater than 36.0, which is ~10 CT values higher than the range for each cDNA sample and consequently would have no measurable effect on gene expression measurements.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
The hypothalamus has gained attention for its involvement in regulating energy metabolism via its ability to receive and integrate signals from peripheral sites and produce a variety of signaling molecules in response to changes in energy status to ultimately maintain energy homeostasis (14, 36, 41). We hypothesized that changes in hypothalamic gene expression exist between lines divergently selected for heat loss as correlated responses to selection and that these changes can account for some of the phenotypic differences in traits related to energy balance between the high and low heat loss lines.

The microarray experiments identified potential candidate genes that could be involved in regulating energy balance and contribute to a portion of the phenotypic differences that are observed between the high and low heat loss selection lines. Surprisingly, only a few genes were found with significant differential expression between the lines according to the microarrays. Even fewer genes were differentially expressed as judged by real-time PCR or Northern blotting. The quantitative expression differences for Oxt, Rpl3, and Timp-2 measured with real-time PCR assays were significant but relatively small, with less than 0.5-fold changes observed between the IH and IL lines.

For complex polygenic traits such as energy balance, it is hypothesized that there are numerous genes with small to moderate effects that contribute to variation in the trait. QTL studies have been conducted to identify genomic regions controlling complex traits, and numerous such loci have been localized for traits involved in determining body weight and fatness. Most of these have individually relatively small and subtle influences on phenotype. Consequently, even though small differences in gene expression in the hypothalamus were found for these selection lines, the differences may be biologically significant. These results suggest that mRNA expression differences for Oxt, Rpl3, and Timp-2 are correlated responses to selection for heat loss. The relatively small differences in expression phenotypes may be indicative of the nature of selection response for highly complex traits. It is very possible that biologically relevant differences in expression exist that we were unable to detect with the methods used in this study. Significant microarray replication and much more sensitive statistical analyses (e.g., 43) are likely required to robustly evaluate gene expression underlying polygenic traits. In addition, genes regulating energy balance may be mediated in ways other than changes in mRNA levels at the temporal and spatial coordinates used in this study.

Of primary interest is to understand the biology and physiology of the genes with expression differences and correlate hypothalamic gene expression differences with phenotypic changes in heat loss and other related traits in the IH and IL selection lines. One possibility is that hypothalamic mRNA expression differences reported here may be directly responsible for the phenotypic differences and that selection for heat loss acted directly on allelic variation within or near these loci. Alternatively, the differences in gene expression for Oxt, Timp-2, and Rpl3 may be secondary, caused by pleiotropic effects of other loci upon which selection acted, and that Oxt, Timp-2, and Rpl3 themselves do not contain heritable genetic variation with relevance to energy balance phenotypes. And third, expression differences may be the result of accumulation of genetic drift during the selection and subsequent inbreeding processes. One way to test these alternative possibilities is to utilize measurement of gene expression phenotypes in a mapping population. Determination of the nature of QTL (cis or trans) that regulate expression phenotypes will help partially reveal the nature of selection response. The real-time assays developed here will assist in facilitating such a population-based analysis. Furthermore, substantial functional studies of Oxt, Timp-2, and Rpl3 are required to further elucidate their role in energy balance regulation.

Oxytocin is one of the neurohypophysial peptides with its major site of expression in the hypothalamus. It shares high sequence homology with the other neurohypophysial peptide, vasopressin, and care was used in this effort to design oligonucleotides for real-time PCR that are specific for Oxt only. Oxytocinergic neurons project throughout the central nervous system, and oxytocin is expressed in peripheral tissues such as uterus, placenta, amnion, corpus luteum, testis, and heart (12, 37). Oxytocin is classically known for its involvement in lactation and parturition (31, 45). Recent studies have suggested other physiological functions of oxytocin. Oxytocin receptors are present in a variety of tissues, including adipose tissue, supporting existence of possible peripheral metabolic functions (9, 12). In the mouse, elevated Oxt results in increased grooming, social memory, and male aggression, and decreased anxiety (12). Oxt mutant null mice show reduced aggressive behavior (6, 42). In male rats, a social defeat experience resulted in elevated circulating Oxt levels (7).

Mice from the IH line have higher activity levels (27), and the higher Oxt mRNA levels found in this study are consistent with reports associating elevated Oxt expression with increased behavioral activities. Oxt is expressed in the paraventricular nucleus and referred to as a "satiety hormone" since both peripherally and centrally administered Oxt reduces feed intake (3, 12, 36). Oxytocin has also been shown to be involved in thermoregulation by increasing body temperature in rabbits and colonic temperatures in mice (17, 20). Consistent with these reports, the IH line, derived from the MH line, has higher core body temperature and higher Oxt expression (27).

The tissue inhibitors of metalloproteinases (TIMPs) represent a family of at least four secreted proteins that inhibit activity of matrix metalloproteinases (MMPs) and play pivotal roles in controlling rate of extracellular matrix (ECM) metabolism (8). A balance between MMPs, TIMPs, and other factors is needed for the breakdown of ECM in processes such as embryonic development, morphogenesis, reproduction, and tissue resorption and remodeling (39). Involvement of TIMPs in obesity has been suggested because obesity is associated with extensive reorganization of adipose tissue involving adipogenesis, angiogenesis, and remodeling of the ECM (8, 19). In particular, stromal-vascular cells isolated from fat depots of mice fed a standard diet had higher Timp-2 mRNA expression compared with cells isolated from mice fed a high-fat diet (19). The potential relationship with TIMPs, MMPs, and TNF-{alpha} has also been evaluated in cases of obesity (11, 19, 21). Reports indicate that TIMPs have growth-promoting and mitogenic activity in various cell types (13, 39). However, overexpression of Timp-2 reduces tumor cell growth and inhibits fibroblast growth factor-induced human endothelial cell growth (28). Timp-2 can also suppress growth factor responsiveness by interfering with the activation of tyrosine kinase-type growth factor receptors and its ability to block mitogenic signaling (39). Moreover, no TIMP receptors have been identified, suggesting that TIMPs may act as decoys for various signaling molecules (39).

In relation to the mRNA expression differences in the heat loss selection lines, the role of Timp-2 is unclear. Timp-2 could be involved in regulation of adipose tissue development, and lower expression in the IL line results in increased adipose tissue mass, which is consistent with results presented by Maquoi et al. (19). It is also possible that Timp-2 acts as a signaling molecule in metabolic or growth pathways.

Higher levels of Rpl3 mRNA have been found in polygenic (1) and monogenic (40) mouse models of obesity. Rpl3 is part of the 60S subunit of the ribosome in eukaryotes and acts as the center channel through which the new peptide emerges. It is speculated that Rpl3 may also have alternate roles. Studies in yeast show a shared transcription factor for Rpl3 and some genes involved in the glycolytic pathway (23). The increased expression of genes in the glycolytic pathway could lead to an increase in the rate of glycolysis leading to increased fat deposition and decreased energy consumption. Allan et al. (1) found increased Rpl3 mRNA levels in multiple tissues within the IL line. Rpl3 may act as a trans-acting transcriptional regulator for genes responsible for decreased heat loss or as a repressor for genes involved in increased heat loss (1).

Map positions of the genes that were evaluated in this study were reviewed to see whether any of the genes with expression differences were located near previously identified QTL for energy balance related traits. If a candidate gene with differential expression maps near a QTL, then the gene becomes a positional candidate and could be considered to be a direct response to selection. Interestingly, Oxt maps near a heat loss QTL on mouse chromosome 2, Hlq2, which was identified in an F2 population derived from the high heat loss selection line and C57BL/6J line (26). Further fine mapping QTL efforts are in progress in this region to determine whether Oxt indeed represents a predisposition locus with cis-acting effects on gene expression.

Summary.
We identified and confirmed the differential expression of Oxt and Timp-2 in inbred lines derived from the high and low heat loss selection lines. Our results indicate that both genes have higher relative mRNA expression levels in the IH high heat loss line. Moreover, we optimized and validated real-time PCR assays for the relative quantification of Oxt, Timp-2, and Rpl3 mRNA expression levels in mice. In these assays, Oxt and Timp-2 had greater expression in the IH line, and Rpl3 had greater expression in the IL line. These results suggest that expression differences in levels of mRNA of those genes are correlated responses to selection for heat loss. To further investigate the mechanisms and loci responsible for controlling the expression of these genes in the heat loss lines, we plan to employ a QTL analysis using a large F2 population derived from a cross between the IH and IL lines and to use the real-time PCR assays to assay expression phenotypic data on individuals in this F2 population. Identification of QTL for the regulation of relevant gene expression changes will facilitate an understanding of the correlation between the predisposition and physiology of heat loss and energy balance and contribute to a better understanding of the complex genetic architecture of polygenic traits.


    ACKNOWLEDGMENTS
 
We thank Sara Olberding, Giovani Bertani, Alex Caetano, and Kari Elo for assistance with data collection, and we thank Jeryl Hauptman for experimental animal care. We also thank Dylan Edwards and his laboratory for provision of Timp-2 primer and probe sequences.

This work is published as paper no. 13926 of the journal series, Nebraska Agricultural Experiment Station.


    FOOTNOTES
 
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

Address for reprint requests and other correspondence: D. Pomp, Dept. of Animal Science, Univ. of Nebraska, ANS A218h, Lincoln, NE 68583-0908 (E-mail: dpomp{at}unl.edu).

10.1152/physiolgenomics.00184.2002.


    References
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Allan MF, Nielsen MK, and Pomp D. Gene expression in hypothalamus and brown adipose tissue of mice divergently selected for heat loss. Physiol Genomics 3: 149–156, 2000.[Abstract/Free Full Text]
  2. Allan MF, Potts JK, and Pomp D. Comparative mapping of RPL3, a gene overexpressed in multiple obesity models. Anim Biotechnol 12: 167–171, 2001.[ISI][Medline]
  3. Arletti R, Benelli A, and Bertolini A. Influence of oxytocin on feeding behavior in the rat. Peptides 10: 89–93, 1998.
  4. Beattie JH, Wood AM, Newman AM, Bremmer I, Choo KHA, Michalska AE, Duncan JS, and Trayhurn P. Obesity and hyperleptinemia in metallothionein (-I and -II) null mice. Proc Natl Acad Sci USA 95: 358–363, 1998.[Abstract/Free Full Text]
  5. Collier GR, McMillan JS, Windmill K, Walder K, Tenne-Brown J, de Silva A, Trevaskis J, Jones S, Morton GJ, Augert G, Civitarese A, and Zimmet PZ. A novel gene involved in the regulation of energy balance. Diabetes 49: 1766–1771, 2000.[Abstract]
  6. DeVries AC, Young WS, and Nelson RJ. Reduced aggressive behavior in mice with targeted disruption of the oxytocin gene. Neuroendocrinology 9: 363–368, 1997.
  7. Ebner K, Wotjak CT, Landgraf R, and Engelmann M. A single social defeat experience selectively stimulates the release of oxytocin, but not vasopressin, within the septal brain area of male rats. Brain Res 872: 87–92, 2000.[ISI][Medline]
  8. Edwards DR, Beaudry PP, Laing TD, Kowal V, Leco KJ, Leco PA, and Lim MS. The roles of tissue inhibitors of metalloproteinases in tissue remodeling and cell growth. Int J Obes 3: S9–S15, 2001.
  9. Egan JJ, Saltis J, Wek SA, Simpson IA, and Londos C. Insulin, oxytocin, and vasopressin stimulate protein kinase C activity in adipocyte plasma membranes. Proc Natl Acad Sci USA 87: 1052–1056, 1990.[Abstract]
  10. Ferrell CL and Jenkins TG. Cow type and nutritional environment: nutritional aspects. J Anim Sci 61: 725–741, 1985.[ISI][Medline]
  11. Gearing AJH, Beckett P, Christodouiou M, Churchill M, Clements J, Davidson AH, Drummond AH, Galloway WA, Gilbert R, Gordon GL, Leber TM, Mangan M, Miller K, Nayee P, Owen K, Patel S, Thomas W, Wells G, Wood LM, and Woolley K. Processing of tumor necrosis factor-{alpha} by metalloproteinases. Nature 370: 555–557, 1994.[ISI][Medline]
  12. Gimpl G and Fahrenholz F. The oxytocin receptor system: structure, function, and regulation. Physiol Rev 81: 629–683, 2001.[Abstract/Free Full Text]
  13. Hayakawa T, Yamashita K, Ohuchi E, and Shinagawa A. Cell growth-promoting activity of tissue inhibitor of metalloproteinase-2 (Timp-2). J Cell Sci 107: 2373–2379, 1994.[Abstract/Free Full Text]
  14. Kalra SP, Dube MG, Pu S, Xu B, Horvarth TL, and Kalra PS. Interacting appetite-regulating pathways in the hypothalamic regulation of body weight. Endocr Rev 20: 68–100, 1999.[Abstract/Free Full Text]
  15. Kopelman PG. Obesity as a medical problem. Nature 404: 635–643, 2000.[ISI][Medline]
  16. Leibel RL, Chung WK, and Chua SC. The molecular genetics of rodent single gene obesities. J Biol Chem 272: 31937–31940, 1997.[Free Full Text]
  17. Lipton JM and Glyn JR. Central administration of peptides alters thermoregulation in the rabbit. Peptides 1: 15–18, 1980.[ISI][Medline]
  18. Livak KJ and Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2-{Delta}{Delta}CT method. Methods 25: 402–408, 2001.[ISI][Medline]
  19. Maquoi E, Munaut C, Colige A, Collen D, and Lijnen HR. Modulation of adipose tissue expression of murine matrix metalloproteinasese and their tissue inhibitors with obesity. Diabetes 51: 1093–1101, 2002.[Abstract/Free Full Text]
  20. Mason GA, Caldwell JD, Stanley DA, Hatley OL, Prange AJ, and Pedersen CA. Interactive effects of intracisternal oxytocin and other centrally active substrates on colonic temperature of mice. Regul Pept 14: 253–260, 1986.[ISI][Medline]
  21. McGeshan GM, Beecher JD, Bast RC, Boyer Cm Champion B, Connolly KM, Conway JG, Furdon P, Karp S, Kidao S, McElroy AB, Nichols J, Pryzwansky KM, Schoenen F, Sekut L, Truesdale A, Verghese M, Warner J, and Ways JP. Regulation of tumor necrosis factor-{alpha} processing by a metalloproteinase inhibitor. Nature 370: 558–561, 1994.[ISI][Medline]
  22. Mercer JG and Speakman JR. Hypothalamic neuropeptide mechanisms for regulating energy balance from rodent models to human obesity. Neurosci Biobehav Rev 25: 101–116, 2001.[ISI][Medline]
  23. Mizuta K, Tsujii R, Warner JR, and Nishiyama M. The c-terminal silencing domain of Rap1p is essential for the repression of ribosomal protein genes in response to defect in the secretory pathway. Nucleic Acids Res 26: 1063–1069, 1998.[Abstract/Free Full Text]
  24. Montano-Bermudez M, Nielsen MK, and Deutscher GH. Energy requirements for maintenance of crossbred beef cattle with different genetic potential for milk. J Anim Sci 68: 2279–2288, 1990.[Abstract/Free Full Text]
  25. Moody DE, Pomp D, and Nielsen MK. Variability in metabolic rate, feed intake, and fatness among selection and inbred lines of mice. Genet Res 70: 225–235, 1997.[ISI][Medline]
  26. Moody DE, Pomp D, Nielsen MK, and Van Vleck LD. Identification of quantitative trait loci influencing traits related in energy balance in selection and inbred lines of mice. Genetics 152: 699–711, 1999.[Abstract/Free Full Text]
  27. Mousel MR, Stroup WW, and Nielsen MK. Locomotor activity, core body temperature, and circadian rhythms in mice selected for high or low heat loss. J Anim Sci 79: 861–868, 2001.[Abstract/Free Full Text]
  28. Nagase H and Woessner JF. Matrix metalloproteinases. J Biol Chem 274: 21491–21494, 1999.[Free Full Text]
  29. Nielsen MK, Jones LD, Freking BA, and DeShazer JA. Divergent selection for heat loss in mice. I. Selection applied and direct response through fifteen generations. J Anim Sci 75: 1461–1468, 1997.[Abstract/Free Full Text]
  30. Nielsen MK, Freking BA, Jones LD, Nelson SM, Vorderstrasse TL, and Hussey BA. Divergent selection for heat loss in mice. II. Correlated responses in feed intake, body mass, body composition, and number born through fifteen generations. J Anim Sci 75: 1469–1476, 1997.[Abstract/Free Full Text]
  31. Nishimori K, Young LJ, Guo Q, Wang Z, Insel TR, and Matzuk MM. Oxytocin is required for nursing but is not essential for parturition or reproductive behavior. Proc Natl Acad Sci USA 93: 11699–11704, 1996.[Abstract/Free Full Text]
  32. Pomp D and Nielsen MK. Quantitative genetics of energy balance: lessons from animal models. Obes Res 7: 106–110, 1999.[Abstract]
  33. Rankinen T, Perusse L, Weisnagel SJ, Snyder ES, Chagnon YC, and Bouchard C. The human obesity gene map: the 2001 update. Obes Res 3: 196–243, 2002.
  34. Robinson SW, Dinulescu DM, and Cone RD. Genetic models of obesity and energy balance in the mouse. Annu Rev Genet 34: 687–745, 2000.[ISI][Medline]
  35. Satcher D. Overweight and obesity threaten U.S. health gains. [Online]. US Dept. of Health and Human Services: Office of the Surgeon General. http://www.surgeongeneral.gov/news/pressreleases/pr_obesity.htm [13 December 2001].
  36. Schwartz MW, Woods SC, Porte D, Seeley RJ, and Baskin DG. Central nervous system control of food intake. Nature 404: 661–671, 2000.[ISI][Medline]
  37. Shojo H and Kaneko Y. Characterization and expression of oxytocin and the oxytocin receptor system. Mol Genet Metab 71: 552–558, 2000.[ISI][Medline]
  38. Spiegelman BM and Flier JS. Obesity and the regulation of energy balance. Cell 104: 531–543, 2001.[ISI][Medline]
  39. Sternlicht MD and Werb Z. How matrix metalloproteinases regulate cell behavior. Annu Rev Cell Dev Biol 17: 463–516, 2001.[ISI][Medline]
  40. Vincent D, Piper M, Gammeltoft S, Maratos-Flier E, and Kahn CR. Alterations of skeletal muscle gene expression of ob/ob mice mRNA differential display. Diabetes 47: 1451–1458, 1998.[Abstract]
  41. Williams G, Bing C, Cai XJ, Harrold JA, King PJ, and Liu XH. The hypothalamus and the control of energy homeostasis: different circuits, different purposes. Physiol Behav 74: 683–701, 2001.[ISI][Medline]
  42. Winslow JT, Hearn EF, Ferguson J, Young LJ, Matzuk MM, and Insel TR. Infant vocalization, adult aggression, and fear behavior of an oxytocin null mutant mouse. Horm Behav 37: 145–155, 2000.[ISI][Medline]
  43. Wolfinger RD, Gibson G, Wolfinger ED, Bennett L, Hamadeh H, Bushel P, Afshari C, and Paules RS. Assessing gene significance from cDNA microarray expression data via mixed models. J Comput Biol 8: 625–637, 2001.[ISI][Medline]
  44. Young DA, Phillips BW, Lundy C, Nuttall RK, Hogan A, Schultz GA, Lecos KJ, Clark IM, and Edwards DR. Identification of an initiator-like element essential for the expression of the tissue inhibitor of metalloproteinases-4 (Timp-4) gene. Biochem J 364: 89–99, 2002.[ISI][Medline]
  45. Zingg HH and Lefebvre DL. Oxytocin and vasopressin gene expression during gestation and lactation. Mol Brain Res 4: 1–6, 1988.[ISI]