INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

IN HUMANS, GENETIC DETERMINANTS fundamental to ventilatory control have been explored in twin studies (e.g., Refs. 10, 21) and among geographically isolated populations (e.g., Refs. 6, 30). As an alternative strategy, the genetic control of breathing has also been studied by using differential ventilatory characteristics among inbred mammals (23-25). This approach has exploited the outcome of inbreeding in which animals within a strain are genetically identical and homozygotic at all loci throughout the genome (4, 14, 22, 23). By employing simple sequence repeat polymorphisms (i.e., DNA microsatellite marker size variation) that differ between inbred strains, meaningful genotype-phenotype associations have been established, particularly concerning various aspects of lung function and structure (5, 14). The relevance of this strategy to human genetic determinants of ventilatory control is contingent on the potential homology between mouse and human genomes (4, 11-14, 22).

Previous investigations in our laboratory have shown that baseline breathing patterns differed between the C3H/HeJ (C3) and C57BL/6J (B6) inbred mouse strains; the C3 pattern was slow and deep relative to the rapid and shallow pattern of the B6 strain (25). Inspiratory timing (TI) at baseline was the primary component to explain this phenotypic variance, and the strain difference in TI (i.e., 75-80 ms) determined the magnitude of the trait controlled by genetic factors. In the B6C3F₁/J (F₁) offspring, TI was significantly different from both parental responses, representing an intermediate phenotype. This genetic model has been described in greater detail by examining the segregation of alleles among the two backcross and intercross progeny. In addition, recombinant inbred (RI) strains derived from B6 and C3 progenitors (12 BXH RI strains) were phenotyped relative to the parental strains. Individual data and group means for the segregant offspring classes and each BXH RI strain have been previously reported (25). On the basis of the random assortment of parental alleles, ventilatory phenotypes of C3 and B6 progenitors were conserved and randomly distributed among the segregant progeny as well as the BXH RI strains. The inheritance pattern of C3 and B6 progenitors and the segregation of BXH RI phenotypes suggested that strain differences in respiratory timing (i.e., both TI and breathing frequency) were controlled by genetic determinants in as few as two loci (25).

The purpose of the present study was to position the gene(s) controlling differential inspiratory timing at baseline in this genetic model. The BXH RI strain distribution pattern (SDP) was used to initiate linkage analysis by comparing the SDP for TI at baseline to a library of 330 BXH RI SDPs. Mouse chromosome 3 was emphasized to initiate our exploration of important genotype-phenotype associations on the basis of the significant concordance between the phenotypic SDP for baseline TI and the genotypic SDP for DNA marker, D3Mit7 (22, 28). To further investigate this hypothesis by using quantitative trait loci (QTL) analyses, the present study assembled a high-density linkage map by genotyping DNA samples from B6C3F₂ (F₂) segregant progeny that were previously characterized for baseline ventilatory traits (11-13, 25). In this genetic model of ventilatory control mechanisms, the results of linkage analysis support the hypothesis that a candidate genomic region on mouse chromosome 3 encompasses putative QTL that determine a considerable proportion of the phenotypic variation in TI at baseline.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals. Reproductively mature male and female F₁ offspring and the 12 available BXH RI strains were purchased from Jackson Laboratories (Bar Harbor, ME). Intercross progeny (i.e., F₂; n = 69 mice) derived from F₁ progenitors were generated in the animal facilities at Johns Hopkins University. The BXH RI strains were propagated by inbreeding (i.e., brother-sister matings) randomly selected F₂ progeny, and, therefore, represented stable segregant genotypes derived from C3 and B6 progenitors. All animals were weaned within 4-5 wk, and water and chow (Agway Pro-Lab RMH 1000) were provided ad libitum. Of the 69 F₂ male progeny included in segregation analysis, a sample of 52 individuals was subsequently genotyped to establish significant genotype-phenotype associations. In addition, genomic DNA was obtained from the 12 BXH RI strains and genotyped as described below.

DNA isolation. High-molecular-weight DNA was isolated from kidney tissue for linkage analysis (18). Sample tissue was gently homogenized by using a dounce homogenizer in an isotonic, high-pH solution that included Nonidet P-40. To release DNA from nuclei, membranes were lysed in a sodium dodecyl sulfate-proteinase K solution. After incubation at 42°C for 3 h, DNA was separated from other cellular components by extraction by using a phenol-chloroform-isoamyl alcohol solution. DNA was precipitated with ice-cold 95% ethanol and 0.15 M potassium chloride. Sample DNA was spooled onto a glass rod, rinsed with 70% ethanol, and resuspended and stored in Tris-EDTA buffer (pH = 8.0). The purity of the sample DNA was determined by using spectrophotometry and calculating the ratio of optical densities (OD) at wavelengths of 260 and 280 nm (a target OD₂₆₀-to-OD₂₈₀ ratio of 1.8). The quantity of DNA was measured by using the OD₂₆₀ reading at 1 unit being equivalent to 50 µg/ml.

PCR amplification. Primers surrounding DNA markers (n = 21) known to be polymorphic for B6 and C3 progenitors were purchased from Research Genetics (Huntsville, AL) and used to amplify DNA from F₂ animals and the BXH RI strains. One primer from each pair was end labeled with -³²P by using the following reaction conditions: 5.4 µl sterile H₂O; 3.6 µl 5× kinase buffer [with a final concentration of 0.25 M Tris (pH = 9.0), 0.05 M MgCl₂, 0.05 M dithiothreitol, 0.25 mg/ml bovine serum albumin]; 4.3 µl primer (10 µM); 0.7 µl T4 polynucleotide kinase (7 units); and 4 µl [-³²P]ATP (specific activity = 6,000 Ci/mmol). The solution was incubated at 37°C for 1 h and diluted with 26 µl sterile H₂O followed by column purification to remove unincorporated -³²P.

Linkage analysis. A multipoint analysis was performed to obtain likely map positions for each baseline ventilatory trait. The method of likelihood ratios was used to establish linkage by comparing the probability that the observed data would arise under one hypothesis (i.e., linkage with 10% recombination between two markers) relative to the probability that it would arise under an alternative hypothesis (i.e., nonlinkage). The ratio of these probabilities, or the log-likelihood, was reported as the log of the odds ratio (LOD value). To further refine the map assignment, a high-density linkage map was generated by identifying polymorphic markers separated by interval distances of 1-5 centimorgans. A centimorgan is the relative distance between any two loci, where 1 centimorgan is equivalent to a recombination fraction of 1%. Flanking markers on mouse chromosome 3 were included to establish the boundaries of the linkage profile.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

BXH RI strain distribution pattern. Figure 1 illustrates the BXH RI SDPs for microsatellite markers in the region between 22.0 and 36.9 centimorgans from the centromere for mouse chromosome 3. At each locus, the average TI (mean ± SE) for the H (i.e., C3 alleles) and B (i.e., B6 alleles) strains is reported, and the probability (P value) resulting from a means comparison between strains is also indicated. At markers D3Mit355 and D3Mit7, the average TI was significantly (t-value_df=10 = 5.5, where df is degrees of freedom; P < 0.0003) different between the H and B strains, supporting a preliminary map assignment for a QTL that determined differential inspiratory timing at baseline. Therefore, a suggestive QTL was assigned to a genomic region 25.1-26.4 centimorgans on mouse chromosome 3. A significant (t-value_df=10 = 3.3; P < 0.008) strain differences was also detected for markers D3Mit121 and D3Mit52. The average TI for markers surrounding this candidate genomic region was not significantly (P > 0.05) different between H and B strains. Moreover, there were no other significant associations as the result of comparing the SDP for TI to 330 SDPs in the BXH RI library.

View larger version (21K): [in this window] [in a new window]   Fig. 1.   BXH recombinant inbred (RI) strain distribution pattern (SDP) for a candidate genomic region of mouse chromosome 3. Progenitor alleles for C3 and B6 strains are indicated by H and B, respectively, and recombination events between 2 markers are underlined. At each locus, average (mean ± SE) inspiratory timing (TI) at baseline for H and B strains was computed from phenotypic data reported elsewhere (25). Two-tailed t-tests were performed to determine P values indicative of significant genotype-phenotype associations. Results suggest that a major genetic determinant of TI at baseline is linked to mouse chromosome 3 in a genomic region positioned in vicinity of D3Mit355 and D3Mit7, or between 25.1 and 26.4 centimorgans (cM) from centromere. NS, not significant. P > 0.05.

Linkage analysis. In Fig. 2, a histogram of the sample frequency response distribution (n = 52 F₂ progeny) used in linkage analysis is compared with the model frequency response distribution (n = 69 F₂ progeny) used to establish a two-gene mode of inheritance for baseline TI (25). The sample distribution is not statistically different (² = 1.54, P > 0.80) from the model distribution.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Histogram of response distribution for TI at baseline. Sample of 52 B6C3F2 (F2) progeny was used to assemble genotype-phenotype associations for linkage analysis. Sample frequency response distribution used in linkage analysis is not statistically different (2 = 1.54, P > 0.80) from model frequency response distribution (n = 69 F2) used in segregation analysis (25).

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Log-likelihood (LOD value) for TI at baseline as a function of relative position of DNA markers in a linkage map for mouse chromosome 3. At peak (LOD = 3.3), a putative quantitative trait locus for TI at baseline is located 25.0 centimorgans from centromere or in proximity to D3Mit119. Potential candidate genes are also depicted: Fgf2, basic fibroblast growth factor; Es27, serum cholinesterase; Gria2, glutamate receptor ionotropic DL--amino-3-hydroxy-5-methylisoxazole-propionic acid 2; Glrb, glycine receptor -subunit.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Evidence from the present study suggests that genetic determinants on mouse chromosome 3 significantly control differential baseline TI between C3 and B6 parental strains. Two approaches to linkage analysis were used in this study. The first approach involved the differential ventilatory responses of 12 BXH RI strains and the assembly of a SDP for TI relative to the parental strains. These data indicate that there is a single major gene located in the vicinity of D3Mit355 and D3Mit7 that determines a significant proportion of the variance in baseline TI between C3 and B6 progenitors. Several influential factors must be considered in drawing this conclusion. First, the P values reported in Fig. 1 for D3Mit355 and D3Mit7 approach but do not attain the recommended P value of 2 × 10⁵ for RI strains, as described by Lander and Schork (13). Second, because the average response of the H strains (167 ± 8 ms) is less than the C3 parental response (188 ± 3 ms) and the average response of the B strains (130 ± 3 ms) is greater than the B6 parental response (111 ± 2 ms), the results further indicate the influence of other major or minor genetic determinants. With these factors in mind, these data suggest that a single major gene is linked to a genomic region between 25.1 and 26.4 centimorgans relative to the centromere on mouse chromosome 3.

A second approach to linkage analysis was accomplished by using an independent sample of F₂ intercross progeny. Microsatellite markers were selected at a high density surrounding D3Mit355 and D3Mit7 to construct a linkage map. The remaining segment of chromosome 3 was also analyzed for other putative QTL or to initiate an exclusion map. At a peak LOD value of 3.3, a putative QTL was established 25.0 centimorgans from the centromere or in close proximity to D3Mit119. Genetic regulation of TI at this locus was estimated to be 25% of the differential response among the F₂ progeny and represented as much as 26 ms of the 80 ms thought to be genetically determined between progenitors. In addition, a second suggestive QTL was established in proximity to D3Mit137 at 35.0 centimorgans from the centromere. By using Mapmaker/QTL, further analyses were performed to ascertain whether the two QTL acted independently to regulate TI. The log-likelihood of a two-QTL model was computed to be less than the sum of the log-likelihoods of a one-QTL model, suggesting that the genetic control of differential TI at baseline was mediated by one putative QTL on mouse chromosome 3. A genome-wide search is being pursued to identify additional putative QTL that may regulate TI at baseline. Taken together, the evidence from both approaches to linkage analyses (i.e., by using the BXH RI SDP library and by employing F₂ progeny to compute LOD values) strongly supports a map assignment for a putative QTL located 25.0 centimorgans from the centromere on mouse chromosome 3. Whether other genetic determinants in this region confer an influence on baseline TI requires further study.

A broader candidate genomic region between D3Mit240 and D3Mit231 cannot be excluded for future consideration on the basis of the results of the present study. Furthermore, there are many potential coding domains and novel genes that may reside within this genomic region. After this region was explored for candidate genes that have biological plausibility (i.e., both mechanical and neural) in baseline ventilatory control mechanisms, several candidate genes emerge from the recognized 35-40 genes in this region (29). For example, the gene Fgf2 (at 19.9 centimorgans) encoding basic fibroblast growth factor may have played a role in determining pulmonary mechanical differences between C3 and B6 progenitors (2, 19). The similarity in time course between basic fibroblast growth factor expression and the development of epithelial basement membrane components has been noted in rat lung (19). Moreover, the allelic forms of Fgf2 have been shown to be polymorphic between C3 and B6 progenitors. Exploratory studies performed in our laboratory suggest that the lung volume of C3 mice was significantly greater relative to B6 mice at airway pressures between 0 and 30 cmH₂O (26). These observations taken together suggest that parental strain differences in the mechanical behavior of the lung may be genetically determined by altering lung structure. As a consequence, mechanisms apart from neural factors involved in the control of TI at baseline may play an important role in this genetic model. The significance of genetic determinants that influence lung growth and development has been proposed by other investigators employing morphometric studies of inbred mice strains (8).

Ventilatory control in many mammalian systems is dependent on cholinergic mechanisms. In a murine system, Burton et al. (3) demonstrated the importance of brain stem acetylcholine turnover in modulating spontaneous ventilation, suggesting that perturbations in cholinergic transmission may contribute to disturbances in ventilatory output. With respect to the present study, isoforms of serum cholinesterase, Es27 and Es26, are encoded by genes located on mouse chromosome 3 at 24.1 and 30.1 centimorgans from the centromere, respectively (29). Furthermore, a genetic deficiency in several cholinergic enzymes including acetylcholinesterase (AChE) in various regions of the brain has been observed in the B6 strain (1). It is unknown whether AChE deficiency in B6 mice is mapped to mouse chromosome 3; however, one hypothesis suggests that AChE deficiency in B6 relative to C3 mice may be associated with a greater cholinergic stimulation manifested by a reduction in TI at baseline.

A family of genes exists in this candidate genomic region that encodes neuroreceptors (9, 29), including the glycine receptor, -subunit (Glrb at 36.0 centimorgans), and the glutamate receptor ionotropic, DL--amino-3-hydroxy-5-methylisoxazole-propionic acid 2 (Gria2 at 33.6 centimorgans). These neuroreceptor systems have been shown to be potentially involved in the integration of breathing (17, 20). Schmid and associates (20) demonstrated that the inhibitory effects of glycine play an important role in the transition from inspiration to expiration. Like Fgf2, allelic forms of Glrb have been shown to be polymorphic between C3 and B6 parental strains. One plausible hypothesis suggests that the rapid breathing pattern characteristic of the B6 strain involves a greater glycinergic inhibition during inspiration.

It is noteworthy to mention the synteny between human and mouse genomes observed for regions of mouse chromosome 3. Specifically, Fgf2, Gria2, and Glrb are mapped within a 10-centimorgan region of mouse chromosome 3 and demonstrate homology to respective genes (FGF2, GRIA2, and GLRB) mapped to a restricted region (i.e., 4q25 to 4q33) of human chromosome 4 (16). On the basis of this synteny, baseline ventilatory control differences in humans may originate from genetic factors located on human chromosome 4. It is unknown whether the candidate genes encoding cholinesterase isoforms on mouse chromosome 3 have synteny to the human genome.

In summary, linkage analysis studies suggest that a putative QTL controlling differential TI at baseline is mapped to mouse chromosome 3. Studies of two independent segregant offspring classes established the preliminary map assignment for this putative QTL to a position 25.0 centimorgans from the centromere or closely linked to D3Mit119. In proximity to the putative QTL were relevant candidate genes including Fgf2 and Glrb, which have synteny to genes located on human chromosome 4. The genetic mechanism may involve a neural component (e.g., cholinergic or glycinergic mechanism) regulating the differential TI at baseline, or, as an alternative hypothesis, the genetic regulation of TI may involve factors (e.g., basic fibroblast growth factor) conferring strain differences in lung structure and mechanics. An improved understanding of the genetic control of baseline ventilation may be facilitated by investigating the genetic determinants that influence lung growth and development.