Abstract: With a view towards, the genetic improvement of economically important traits, we previously generated a beef cattle F₂ resource population from a Japanese Black and Limousin cross. Here, we report on a genome-wide scan for QTLs affecting muscle fiber diameter and composition traits. The longissimus thoraces muscle samples from 150 F₂ animals were examined to determine muscle fiber type and diameter. At the same time, F₂ animals as well as the F₁ parents and F₀ grandparents were genotyped for 313 informative micro satellite markers that spanned 2.382 cM of the Bovine autosomes. The results of a genome scan using least-squares regression interval mapping provided evidence for a QTL affecting the diameter of the red muscle fiber (<5% chromosome-wide level) at 42 cM on BTA17; a QTL affecting the composition of the red and white muscle fibers (<5% chromosome-wide level) at 0 cM on BTA5 and a QTL affecting the composition of the red muscle fiber (<5% chromosome-wide level) at 106 cM on BTA8. Thus in this study, we detected 3 QTLs for 3 traits. The detected QTLs were either contiguous with or overlapped previously reported QTLs with effects on several traits assumed to be related to muscle histochemical properties.

INTRODUCTION

The histochemical properties of muscle in beef cattle are of considerable interest with regard to their presumed effect on meat quality, for example, meat color, marbling and tenderness. In the muscles that contract in a fast or slow-twitch manner, fueled by energy from oxidative and glycolytic metabolisms, a heterogeneity of myofibers has been recognized (Peter and Starson, 1972, 1990). Myofibers can essentially be divided into 3 major categories, namely, red (correspond to type I, βR, Slow-twitch Oxidative; SO), intermediate (correspond to type IIA, αR, Fast-twitch Oxidative Glycolytic; FOG) and white (correspond to type IIB, αW, Fast-twitch Glycolytic; FG), depending on their histochemical, physiological and biochemical properties (Klont et al., 1998). A significant positive correlation has been demonstrated between beef muscle fiber diameter and shear force (Herring et al., 1965; Crouse et al., 1991). Crouse et al. (1991) have also revealed a negative correlation between sensory tenderness and the percentage of white muscle fiber area. Further, Calkins et al. (1981) reported a negative correlation between αW muscle fiber and marbling and tenderness, but a positive correlation between αR and βR muscle fibers and the same qualities. Whipple et al. (1990) also verified the relationship between muscle fiber traits and beef longissimus tenderness and reported the effect of αR and βR muscle fibers on this trait. Differences in these muscle fiber characteristics have been reported to be related to both growth and breed (Wegner et al., 2000). Moreover, compared with other breeds of cattle, the Japanese Black has been revealed to have a significantly higher percentage distribution and larger diameter of type I myofibers (Gotoh, 2003).

There have been several reports on the Quantitative Trait Loci (QTL) analysis of industrial animals. In particular, there have been several reports on the QTL mapping of muscle histochemical traits in pigs (Malek et al., 2001; Nii et al., 2005; Wimmers et al., 2006; Estellé et al., 2008). To the best of the knowledge, however, there have been no reports to date on the QTL mapping of these muscle properties in cattle.

In this study, we focused on the histochemical properties of muscle in beef cattle and used interval mapping to search for QTLs associated with these traits, using an F₂ resource population derived from a cross between Japanese Black and Limousin breed.

MATERIALS AND METHODS

Animals and phenotypes: The F₂ population used in this study was described in detail in our previous report (Abe et al., 2008). A total 150 animals (78 steers, 72 heifers) were used for muscle fiber trait measurements.

The rib roast blocks of the seventh to 8th rib bones were sampled from these F₂ animals and the longissimus thoracis muscle was excised from the block. A 1x1x1 cm cube was then cut from the center of the rib eye, immediately frozen in liquid nitrogen and stored at -30°C, until used for analysis. Transverse serial sections (8 μm thick) were cut at -18°C using a cryostat microtome (HM505E; Microm International GmpH, Walldorf, Germany). Succinic dehydrogenase activity was localized using nitroblue tetrazolium as a substrate according to the method of Gauthier et al. (1968). Sections were incubated for 30 min at 37°C. The proportion of each fiber type (red, intermediate and white) was obtained by counting >200 fibers in 3 fields of view. The minor axis of each fiber was also measured. All image processing was carried out using WinROOF software (Mitani Corp., Tokyo, Japan).

Genotyping: DNA was extracted from blood using automatic extraction equipment (NA1000; Kurabo, Osaka, Japan) and the final DNA concentration was adjusted to 20 ng μL^-1. A genome screen was conducted using microsatellite markers (Kappes et al., 1997; Ihara et al., 2004). Polymerase Chain Reaction (PCR) amplification was performed in a volume of 15 μL containing 20 ng of genomic DNA, 1.67 mM MgCl₂, 6.25 pmol of each primer, 0.2 mM deoxynucleotides (dNTPs) and 0.375 U Taq DNA polymerase (ABgene, Epsom, UK). The annealing temperatures of each marker in the thermocycling steps were optimized in accordance with Ihara et al. (2004). Amplifications were performed under the following conditions: 5 min at 94°C; 30 cycles of 30 sec at 94°C, 30 sec at the annealing temperature and 30 sec at 72°C and a final extension of 7 min at 72°C. After PCR amplification, the reaction products were fractionated in an ABI377 DNA sequencer (Applied Biosystems, Foster City, CA) and fragment analysis was performed using GeneScan and Genotyper software (Applied Biosystems, Foster City, CA).

Linkage analysis and QTL mapping: Linkage maps for the 29 bovine autosomes were constructed using CRI-MAP (Green et al., 1990) and the constructed maps were used for the whole-genome QTL scan. The information content of markers was calculated using the method described by Knott et al. (1998).

The maps were then used for QTL detection on the 29 autosomes using QTL Express software (Seaton et al., 2002). The least-squares regression model (Haley et al., 1994) was used for QTL analysis, including individual sex, breed of the recipients and combinations of seasons and locations as the fixed effects and ages of the recipients as covariates, along with additive and dominance effects for the putative QTLs. Detection of QTLs was based on an F-statistic that was computed from sums of squares explained by the additive and dominance coefficients for the QTL. Significance thresholds for the F-statistic were derived at the chromosome and experiment-wide levels on a single-trait basis using a permutation test, with 1,000 repetitions for each trait. The percentage of F₂ variance explained by the model was calculated as follows:

RESULTS AND DISCUSSION

Phenotype measurement: The longissimus thoracis muscle samples from 150 F₂ individuals were classified into 3 muscle fiber types (red, intermediate and white; Table 1). Significant differences were observed between steers and heifers with respect to the diameters of both intermediate and white muscle fibers (p<0.001), with heifers having larger diameters than steers. Gotoh (2003) reported gender differences in myofiber type distribution between steers and bulls and discussed the possibility of the effect of hormonal regulation differences. Although, the combination of genders differed in the present study, the same influences could still be at work.

Whole-genome QTL analyses of F₂ animals were successfully carried out. No QTLs at experiment-wide level were detected in this study, but 3 chromosome-wide significant QTLs for 3 muscle fiber traits were detected (Table 2).

QTL on BTA17: A QTL for the diameter of red muscle fiber was detected at 42 cM on BTA 17 (Fig. 1). F₂ individuals inheriting the Japanese Black allele had larger red muscle fiber diameters. A QTL for lean-to-fat ratio and marbling score have previously been mapped at 21 cM and 38 cM, respectively, on BTA17 (Casas et al., 2000, 2004). The composition of type I myofiber has been reported to be positively correlated with marbling (Calkins et al., 1981) and the percentage of intramuscular fat content (Gotoh, 2003).

Table 1:	Muscle fiber traits of F2 animals from an inter cross of F1 animals deribed from 2 Japanese Black and Limousin dams
1Mean±SE; 2Total of all the F2 population

Table 2:	Summary of chromosome-wide significant QTL for muscle fiber traits
1All F-statistics inscribed in this column were significant at the 5% chromosome-wide level

Fig. 1:	Plot of the F-ratios from multilocus least square analisis (Haley et al., 1994) of the diameter of red muscle fiber on BTA17. The x-axis indicates the relative position on the linkage map and the y-axis represents the F-ratio. Triangles on the x-axis indicate marker positions. Markers were RM156, BMS2220, MNB-10, BMS941, BM305, TGLA231, MB008, IDVGA-40, BM8125, BM1862, DIK643, BM1233 and MS2263

This suggests that the same latent gene may have an effect on each the QTL for lean-to-fat ratio, marbling score and our QTL for the diameter of red muscle fiber.

A QTL for feed-to-gain ratio has previously been reported at 40 cM on BTA 17 (Nkrumah et al., 2007). Given the fact that the differences in muscle fiber type originate from differences in energetic metabolism, this would appear to be a reasonable results. More interestingly, QTLs for average daily gain and post-weaning average daily gain were also detected at 43.5 and 35.1 cM, respectively, on BTA17 (Nkrumah et al., 2007; Alexander et al., 2007). These results might be summarized as follows. The difference in energetic metabolism of different muscle fiber types has an effect on feed efficiency and this in turn is reflected in average daily gain.

QTL on BTA5: A QTL for the composition of red and white muscle fibers was detected at 0 cM on BTA5 (Fig. 2). In this case, F₂ individuals that inherited the Limousin alleles had a higher percentage of red muscle fiber, whereas animals inheriting the Japanese Black allele had a higher percentage of white muscle fiber (Table 2). This finding is contrary to what we had earlier predicted, because it has previously been reported that the Japanese Black breed tends to develop a higher percentage of type I muscle fiber than the Japanese Red or Holstein breeds (Gotoh, 2003). In addition, there is a positive correlation between the percentage of type I muscle fiber and fat accumulation, whereas the type IIB muscle fiber tends to accumulate less intramuscular fat (Clakins et al., 1981; Gotoh, 2003). Gotoh (2003) deduced that Japanese Black cattle contain more type I myofibers compared to other breeds and thus, they have a greater propensity to accumulate intramuscular fat. Furthermore, approximately, 70% of semitendinosus muscle fiber in 16-months-old pure-bred Limousin bulls was typed as FG muscle fiber, which is equivalent to Type IIB muscle fiber (Jurie et al., 1995). Thus, the observations made in the present study tend to conflict with those made previously. The Limousin allele had an increasing effect on the percentage of red muscle fiber and therefore, if we were to introgress the Limousin allele into the Japanese Black, it may be possible to increase the percentage of red muscle fiber in the latter breed.

A QTL for average daily gain-related traits had been reported around the centromeric region of BTA5. Li et al. (2002) detected QTLs for both average daily gain and pre-weaning average daily gain at 6.05-17.28 cM on BTA5 in their Bos taurus cattle study. Therefore in this QTL, the same latent gene may have affects on both average daily gain and muscle fiber percentage.

QTL on BTA8: A QTL for the composition of red muscle fiber was detected at 106 cM on BTA8 (Fig. 3). F₂ individuals that inherited the Limousin allele had a higher percentage of red muscle fiber, which shown the same tendency as the QTL we detected on BTA5. Davis et al. (2007) reported a QTL for the adhesion properties of cattle longissimus lumborum muscle on BTA8.

Fig. 2:

Plot of the F-ratios from multilocus least square analisis (Haley et al., 1994) of the composition of red and white muscle fiber on BTA5. The x-axis indicates the relative position on the linkage map and the y-axis represents the F-ratio. Triangles on the x-axis indicate marker positions. Markers were BMS1095, BMS610, BP1, RM103, BMS1898, MS2106, CA084, ETH10, MNS-44, BMS1248, BM315, DIK2206, DIK2287, DIK2122, BM733, DIK2035 and BMS597

Fig. 3:	Plot of the F-ratios from multilocus least square analisis (Haley et al., 1994) of the composition of red muscle fiber on BTA8. The x-axis indicates the relative position on the linkage map and the y-axis represents the F-ratio. Triangles on the x-axis indicate marker positions. Markers were BMS1864, INRA097, BMS1591, INRA180, DIK1174, BMS2072, IDVGA-52, BM711, DIK074, DIK1169 and BMS2629

Briefly, adhesion is a measure of how easily a muscle can be pulled apart by a force applied perpendicular to the muscle fibers. Thus, this measurement represents connective tissue toughness, which was reflected by collagen cross-linking. To date, however, there have been no reports that have directly discussed the relationship between connective tissues and muscle fiber types, not only in cattle but also in other mammalian species. Nevertheless, it seems unlikely that adhesion is related to muscle fiber type composition. Furthermore, according to the linkage map constructed by Ihara et al. (2004), the interval between the highest peak of each of these 2 QTLs was estimated to be approximately 30 cM, suggesting them to be different QTLs with different effects.

Presumable relationship with calving ease: The QTLs detected on BTA8 and 17 in this study overlap with the previously reported QTLs for calving ease and percentage difficult birth (CE_PDB) traits (Ashwell et al., 2005). On BTA8, the QTL for CE_PDB was mapped at 116 cM, spanning the makers BMS2847 and BMS2629. The QTL for the composition of red muscle fiber on BTA8 in this study spanned DIK1169 and BMS2629. Consequently, the QTL for red muscle fiber composition covers the entire CE_PDB QTL. Furthermore, the QTL for CE_PDB on BTA17 was mapped at 69 cM and spanned BM305 and URB002. The QTL for the diameter of red muscle fiber on BTA17 in this study spanned MB008 and IDVGA-40. Consequently, the QTL for CE_PDB covers the entire QTL for the diameter of red muscle fiber. The foregoing results were completely unexpected. It is widely understood that Double Muscle (DM) cattle exhibit a high frequency of calving difficulty (Casas et al., 1999) and that they also have a higher percentage of Type IIB muscle fiber (Wegner et al., 2000). However, since the DM animals are an extreme case, we should perhaps not infer that this relationship is typical of all breeds. Nevertheless, the relationship between muscle fiber type and calving ease traits can be readily imaged. Considering the importance of calving ease traits, further studies should be conducted on this matter.

CONCLUSION

In this study, we identified 3 QTLs with effects on 4 muscle fiber traits. These QTLs may be of potential value in marker-assisted introgression; however, further studies will be required in order to establish their reliabilities.

ACKNOWLEDGEMENTS

We thank, the staff of Tokachi and Ohu Stations, National Livestock Breeding Center, for feeding and measuring various traits of the F₂ population.