Fine Mapping of “Mini-Muscle,” a Recessive Mutation Causing Reduced Hindlimb Muscle Mass in Mice

Abstract

Prolonged selective breeding of Hsd:ICR mice for high levels of voluntary wheel running has favored an unusual phenotype (mini-muscle [MM]), apparently caused by a single Mendelian recessive allele, in which hindlimb muscle mass is reduced by almost 50%. We recently described the creation and phenotypic characterization of a population suitable for mapping the genomic location of the MM gene. Specifically, we crossed females from a high-runner line fixed for the MM allele with male C57BL/6J. F1 males were then backcrossed to the MM parent females. Backcross (BC) mice exhibited a 50:50 ratio of normal to MM phenotypes. Here, we report on linkage mapping of MM in this BC population to a 2.6335-Mb interval on MMU11. This region harbors ∼100 expressed or predicted genes, many of which have known roles in muscle development and/or function. Identification of the genetic variation that underlies MM could potentially be very important in understanding both normal muscle function and disregulation of muscle physiology leading to disease.

Prolonged selective breeding of Hsd:ICR mice for high levels of voluntary wheel running has favored an unusual phenotype (mini-muscle [MM]), apparently caused by a single Mendelian recessive allele, in which hindlimb muscle mass is reduced by almost 50%. This phenotype was originally observed in 2 of 4 replicate selected lines (termed HR for high runner) and in 1 of 4 control (nonselected) lines (Garland et al. 2002). Analyses of data from the first 22 generations of the selection experiment indicated that the mutant allele was present at a frequency of approximately 7% in the base population (outbred Hsd:ICR mice). Five of the 8 total lines apparently lost the allele by random genetic drift. In one control line, the phenotype, representing homozygotes, was observed at a frequency of 0–10% for the first 22 generations (Garland et al. 2002) and was then apparently lost (T Garland Jr. unpublished observations). The 2 selected lines that have exhibited the phenotype showed an increase in frequency consistent with positive selection (Garland et al. 2002). In one (lab designated line 6), the phenotype remains polymorphic as of generation 50. In the other (lab designated line 3), the mutation apparently had gone to fixation by generation 36 (Syme et al. 2005).

The most characteristic phenotype of MM allele homozygotes is a 50% reduction in mass of the triceps surae muscle complex (Garland et al. 2002) as well as in mass of mixed hindlimb muscle exclusive of the triceps surae (Houle-Leroy et al. 2003). Beyond this, the MM allele has many pleiotropic effects in homozygotes, including a doubling of mass-specific aerobic capacity as compared with wild-type muscle (Houle-Leroy et al. 2003), altered fiber type composition in the gastrocnemius (Guderley et al. 2008), altered muscle contractile performance (Syme et al. 2005), an increase in size of their ventricles, liver, and spleen (Garland et al. 2002; Swallow et al. 2005), and longer and thinner hindlimb bones (Kelly et al. 2006). Many of these effects seem conducive to the support of endurance running (Garland 2003; Guderley et al. 2006; Rezende et al. 2006). To date, clearly deleterious consequences of the MM allele have not been reported.

Although the physiological consequences of MM are becoming well understood, the nature of the underlying mutation has not been characterized. Identification of the MM gene, and how variation within that gene leads to the MM phenotype, would be important for understanding both normal and abnormal muscle development in mammalian species. We recently described the creation and phenotypic characterization of a population suitable for mapping the genomic location of the MM gene (Hannon et al. 2008). We crossed females from the HR line that is fixed for the MM allele with male C57BL/6J. F1 males were then backcrossed to the MM parent females. The HR(B6HRF1) backcross (BC) mice (N = 404) were dissected, and a 50:50 ratio of normal to MM phenotype was observed. In this paper, we report on linkage mapping of MM in this BC population to a 2.6335-Mb interval on MMU11. This region harbors ∼100 expressed or predicted genes, many of which have known roles in muscle development and/or function.

Methods

Development of BC Mapping Population

As described elsewhere (Hannon et al. 2008), 20 C57BL/6J males (The Jackson Laboratory, Bar Harbor, ME) were harem mated with each of 3 females from HR line 3, which is apparently fixed for the MM mutant allele (Syme et al. 2005). Sixty male F1 mice were weaned at 21 days of age. At approximately 8 weeks of age, they were randomly backcrossed (one dam with one sire) to the HR line 3 dams, with the exception that mother–son and aunt–nephew matings were disallowed. Once dams were visually pregnant, F1 males were removed. The HR(B6HRF1) BC pups were weaned at 21 days, then all line 3 dams and F1 males were dissected following Garland et al. (2002) to determine MM phenotype status (see below).

When BC pups reached approximately 7 weeks of age, they were given 6 days of wheel access (as in the routine selective breeding protocol described by Swallow et al., 1998) and then dissected to determine MM phenotype. For mice in all 4 groups (C57BL/6J males, HR line 3 females, F1 offspring, and BC offspring), plots of mean triceps surae muscle mass against body mass allowed unambiguous identification of MM individuals (Figure 1).

Figure 1.

Relation between triceps surae mass (mean of left and right) and body mass for HR(B6HRF1) BC mice. The MM phenotype, characterized by a nearly 50% reduction in muscle mass, was found in ∼50% of BC mice, supporting a single recessive locus model.

Chromosomal Localization and Fine Mapping of MM

General Methods

Microsatellite marker analysis was performed in 2 overall steps. An initial genome-wide scan was performed with sparse genome coverage using 54 BC mice segregating for MM, to identify the general chromosomal region harboring the MM gene. This was followed by fine mapping with dense marker coverage within the MM region using all BC mice. Microsatellite markers were genotyped as previously described (Allan et al. 2005). Linkage maps were constructed using MAP MANAGER QTX (http://www.mapmanager.org/mmQTX.html). For all linkage estimations, genotypes for MM were inferred based on phenotypic evidence as described above. BC mice with MM were designated mini/mini, whereas mice with no evidence of MM were designated mini/B6.

Sparse Mapping

Because HR has not been previously genotyped for microsatellite markers, and is not fully inbred, we initially tested several markers on each autosome for informativeness in the BC. Two or 3 microsatellite markers (total of 48) across all 19 autosomes were selected for the scan (data not shown). These were genotyped across a subset of 54 BC mice having a 50:50 ratio of MM and normal phenotypes selected across several BC families. A Chi-square test was used to test for significant (P < 0.05) linkage between a microsatellite marker and the mutant gene, by examining deviations from the expectation that 50% of the samples should have a mini/mini marker genotype and 50% should have a mini/B6 genotype.

Fine Mapping

Initially, markers D11Mit108, D11Mit260, D11Mit141, and D11Mit212, all found to be informative in the initial marker testing, were genotyped across the full BC population of 404 mice, and a linkage map was generated for these 4 markers and MM. For further fine mapping, additional markers were tested in the region between 37.0 and 43.8 cM (Table 1). From this test, an additional 3 markers (D11Mit4, D11Mit31, and D11Mit90) were selected and genotyped in the full BC and a new linkage map was generated based on 7 microsatellite markers and MM.

Table 1.

All markers used in microsatellite mapping of MM

Marker	Chromosome	Linkage position (cM)	Physical position (Mb)	Phase of mapping
D1mit318	1	18.5	33.84	Sparse mapping
D1mit33	1	81.6	160.34	Sparse mapping
D1mit128	1	36.9	72.57	Sparse mapping
D2mit297	2	29.0	42.46	Sparse mapping
D2mit171	2	98.0	34.34	Sparse mapping
D3mit203	3	11.2	26.84	Sparse mapping
D3mit147	3	79.4	148.41	Sparse mapping
D4mit171	4	6.3	22.41	Sparse mapping
D4mit139	4	28.6	55.26	Sparse mapping
D4mit71	4	61.9	133.34	Sparse mapping
D5mit126	5	20.0	35.80	Sparse mapping
D5mit240	5	59.0	109.52	Sparse mapping
D6mit119	6	25.5	50.81	Sparse mapping
D6mit31	6	38.5	92.67	Sparse mapping
D6mit217	6	60.5	125.27	Sparse mapping
D7mit227	7	16.0	37.37	Sparse mapping
D7mit321	7	48.5	105.05	Sparse mapping
D8mit64	8	16.0	33.90	Sparse mapping
D8mit215	8	59.0	118.38	Sparse mapping
D9mit205	9	18.0	37.11	Sparse mapping
D9mit165	9	38.0	68.53	Sparse mapping
D9mit212	9	61.0	108.54	Sparse mapping
D10mit126	10	21.0	26.73	Sparse mapping
D10mit70	10	59.0	103.54	Sparse mapping
D11mit108	11	18.0	35.86	Sparse mapping and initial fine mapping
D11mit141	11	27.0	52.96	Initial fine mapping
D11mit260	11	34.4	61.61	Initial fine mapping
D11Mit115	11	37.0	67.45	Tertiary fine mapping
3D	11	—	67.97	Tertiary fine mapping
D11Mit1000a	11	—	68.32	Secondary fine mapping
D11Mit4	11	37.0	68.42	Secondary fine mapping
D11mit60	11	40.0	69.55	Tertiary fine mapping
D11mit30	11	39.8	69.61	Tertiary fine mapping
D11mit29	11	40.0	69.61	Tertiary fine mapping
D11mit31	11	40.0	—	Tertiary fine mapping
1F	11	—	70.09	Tertiary fine mapping
1H	11	—	70.09	Tertiary fine mapping
1A	11	—	70.10	Tertiary fine mapping
1B	11	—	70.10	Tertiary fine mapping
D11Mit90	11	42.0	70.31	Secondary fine mapping
D11mit212	11	50.0	88.67	Sparse mapping and initial fine mapping
D12mit185	12	11.0	27.75	Sparse mapping
D12mit153	12	15.0	35.92	Sparse mapping
D12mit156	12	34.0	80.56	Sparse mapping
D12mit158	12	38.0	83.72	Sparse mapping
D13mit158	13	5.0	49.33	Sparse mapping
D13mit198	13	7.7	35.06	Sparse mapping
D13mit139	13	32.0	51.86	Sparse mapping
D13mit202	13	47.0	91.61	Sparse mapping
D14mit141	14	15.0	47.38	Sparse mapping
D14mit82	14	19.5	53.41	Sparse mapping
D14mit196	14	47.0	—	Sparse mapping
D15mit12	15	4.7	31.61	Sparse mapping
D15mit107	15	49.0	84.22	Sparse mapping
D16mit132	16	3.7	—	Sparse mapping
D16mit71	16	70.7	97.13	Sparse mapping
D17mit175	17	17.7	—	Sparse mapping
D17mit72	17	47.4	79.39	Sparse mapping
D18mit123	18	31.0	56.13	Sparse mapping
D18mit4	18	57.0	84.30	Sparse mapping
D19mit128	19	10.9	17.33	Sparse mapping
D19mit67	19	43.0	44.47	Sparse mapping
D19mit71	19	54.0	59.67	Sparse mapping

^aD11Mit1000 is named D17Mit144 in the Ensembl database.

Subsequently, additional informative markers were genotyped but only in specific BC mice harboring a recombination near the estimated linkage position of MM. New markers were genotyped in the regions between D11Mit260 (34.4 cM, 61.61 Mb) and D11Mit4 (37.0 cM, 68.42 Mb) and between D11Mit31 (40.0 cM) and D11Mit90 (42.0 cM, 70.31 Mb). When no previously published markers were available, primers were designed flanking putative microsatellites identified through BLAST searches in Ensembl and informative markers were genotyped. This process was continued until nearly all information on location of recombinations in the BC was exhausted. A total of 17 newly designed markers were tested, of which 15 amplified well and 5 were informative (Table 1).

Candidate Gene Sequencing

DNA from C57BL/6J (n = 1), HR Line 3 mice with MM (n = 3), HR Line 6 mice with MM (n = 3), and HR Line 6 mice without MM (n = 3) was used to generate polymerase chain reaction (PCR) products for sequencing of candidate genes within the fine-mapped MM interval. For sox15, overlapping PCR products were amplified encompassing the entire gene plus some flanking sequence (Ensembl coordinates: 69,468,319–69,470,660). For Chrnb1, 11 overlapping PCR products were amplified to encompass 4959 bp of the gene including all 11 exons and part of the 5′ and 3′ untranslated region (UTR) regulatory sequences (Ensembl coordinates 69,609,023–69,609,447; 69,608,428–69,608,582; 69,607,330–69,607,783; 69,606,871–69,607,171; 69,606,147–69,606,445; 69,600,313–69,600,624; 69,599,120–69,599,461; 69,598,235–69,598,730; 69,597,890–69,598,334; 69,597,558–69,598,061; 69,597,174–69,597,674). For rpl26, overlapping PCR products were amplified representing the entire gene (Ensembl coordinates: 68,715,068–68,718,035). All PCR products incorporated 5′ and 3′ M13 tails and were sequenced from both directions through a service provider (Agencourt Bioscience, Beverly, MA).

Results

Sparse Mapping

Two microsatellite markers on MMU11 (D11Mit108, 18 cM, 36 Mb and D11Mit212, 50 cM, 89 Mb) were found to be linked with MM (P<0.01; Table 2) based on genotyping of 54 BC mice. Given that the estimated strength of linkage between MM and each of these markers was very similar (Table 2), the position of MM was presumed to be near the midpoint of the markers at ∼35 cM. No markers on other chromosomes showed significant linkage to MM.

Table 2.

Chi-square values for 2 markers on Chromosome 11 in the genome-wide screen of 54 BC mice

			MM mice		Normal mice
Marker	Position (cM)	Physical location (Mb)	Mini/mini	Mini/B6	Mini/mini	Mini/B6	Chi square
D11Mit108	18	35.86	24	6	4	20	9.48
D11Mit212	50	88.67	23	6	4	21	10.7

Fine Mapping

Results of initial linkage mapping using MM and 4 microsatellite markers genotyped across the full BC population are summarized in Table 3. The estimated positions are given as ranges. MM mapped between markers D11Mit260 (34.4 cM, 61.6 Mb) and D11Mit212 (50 cM, 88.7 Mb). Location of MM was estimated to be 4.45 cM distal to D11Mit260 and 9.83 cM proximal to D11Mit212.

Table 3.

Initial fine mapping linkage map between MM and 4 microsatellite markers

Marker	Linkage distance (cM)	MGD position (cM)	Physical location (Mb)
D11Mit108		18.0	35.86
	5.49
D11Mit141		27.0	52.96
	6.21
D11Mit260		34.4	61.61
	4.45
Mini-Muscle		—	—
	9.83
D11Mit212		50.0	88.67

MGD, mouse genome database.

After genotyping of 3 additional markers (D11Mit4, D11Mit31, and D11Mit90), only one (D11Mit90, 42 cM, 70.32 Mb) revealed new recombinations for linkage mapping. Results of the new mapping including this additional marker are summarized in Table 4 and Figure 2, showing that the region harboring MM had now been effectively reduced to the interval of 61.6–70.3 Mb. The very tight linkage (0.5 cM) between MM and D11Mit90 provided evidence that MM was likely near the distal edge of this interval.

Table 4.

Linkage map after secondary phase of fine mapping

Marker	Linkage distance (cM)	MGD position (cM)	Physical location (Mb)
D11Mit108		18.0	35.86
	5.49
D11Mit141		27.0	52.96
	6.21
D11Mit260		34.4	61.61
	4.53
Mini-Muscle		—	—
	0.51
D11Mit90		42.0	70.31
	9.39
D11Mit212		50.0	88.67

MGD, mouse genome database.

Figure 2.

Initial linkage map of the MM region after genotyping of 6 markers on MMU11 in the full HR(B6HRF1) BC population.

For further fine mapping of the MM mutation, only mice with recombinations between MM and D11Mit260 on the proximal end (n = 18) or D11Mit90 on the distal end (n = 2) were genotyped. Given that no new estimations of linkage would be feasible, only physical locations were used at this point to guide selection and/or design of additional microsatellite markers. New genotyping within this region revealed additional recombinations, helping to further refine the localization of MM (Table 5), with new boundaries of D11Mit115 (67.453 Mb) and the newly designed marker 1F (70.0865). Having exhausted recombinations in this BC population, MM was localized to this 2.6335-Mb region, which harbors ∼100 genes (Table 6; NCBI Build 37.1 [February, 2008]).

Table 5.

Genotypic matrix for further fine mapping of MM using 20 informative BC mice

	Mouse ID	D11Mit260	D11Mit115	MM	1F	D11Mit90
61.609 Mb	67.453 Mb	—	70.0865 Mb	70.313 Mb
4302	H	R	R	—	R
4482	H	R	R	—	R
4471	H	R	R	—	R
4472	H	R	R	—	R
4081	H	R	R	—	R
4239	—	R	R	—	R
4196	H	R	R	—	R
4238	R	H	H	—	H
4427	R	H	H	—	H
4276	R	H	H	—	H
4269	R	H	H	—	H
4292	R	H	H	—	H
4279	R	H	H	—	H
4152	R	H	H	—	H
4404	R	H	H	—	H
4268	R	H	H	—	H
4197	R	R	H	—	H
4086	H	H	R	—	R
4199	R	—	R	H	H
4344	H	—	H	H	R

H, mini/B6 genotype; R, mini/mini genotype. Relevant recombinations are in bold.

Table 6.

Expressed and predicted genes within the 2.6335 Mb MMU11 region harboring MM

Symbol	Start base pair	Stop base pair	Description
Gas7	67,416,607	67,498,461	Growth arrest specific 7
Rcvrn	67,500,799	67,516,835	Recoverin
Glp2r	67,519,932	67,584,655	Glucagons-like peptide 2 receptor
LOC544792	67,588,110	67,603,400	Similar to germ cell associated 1 (predicted)
Dhrs7c	67,611,790	67,629,507	Dehydrogenase/reductase (family) member 7C
LOC100041899	67,651,191	67,657,281	Hypothetical protein LOC100041899
Usp43	67,668,025	67,735,655	Ubiquitin-specific peptidase 43
Wdr16	67,738,308	67,779,144	WD repeat domain 16
Stx8	67,779,985	68,020,650	Syntaxin 8
Ntn1	68,022,866	68,200,328	Netrin 1
Pik3r5	68,245,627	68,311,348	Phosphoinositide-3-kinase, regulatory subunit 5, p101
BB220380	68,316,521	68366200	Expressed sequence BB220380
BC024997	68,369,688	68,371,747	cDNA sequence BC024997
Ccdc42	68,407,557	68,411,456	Coiled-coil domain containing 42
Myh10	68,505,417	68,630,126	Myosin, heavy polypeptide 10, nonmuscle
Ndel1	68,634,936	68,666,564	Nuclear distribution gene E-like homolog 1 (A. nidulans)
9930039A11Rik	68,702,055	68,708,517	RIKEN cDNA 9930039A11 gene
Rpl26	68,715,068	68,718,036	Ribosomal protein L26
Odf4	68,735,337	68,740,551	Outer dense fiber of sperm tails 4
Arhgef15	68,756,656	68,770,360	Rho guanine nucleotide exchange factor 15
Slc25a35	68,781,633	68,786,017	Solute carrier family 25, member 35
2400006H24Rik	68,785,989	68,787322	RIKEN cDNA 2400006H24 gene
Pfas	68,798,671	68,821,962	Phosphoribosylformylglycinamidine synthase (amidotransferase)
1500010J02Rik	68,829,615	68,849,972	RIKEN cDNA 1500010J02 gene
Aurkb	68,859,145	68,865,164	Aurora kinase B
2310047M10Rik	68,873,277	68,875,078	RIKEN cDNA 2310047M10 gene
Tmem107	68,884,311	68,886,794	Transmembrane protein 107
Vamp2	68,902,030	68,905,883	Vesicle-associated membrane protein 2
Per1	68,912,451	68,923,462	Period homolog 1 (Drosophila)
Hes7	68,933,955	68,936,761	Hairy and enhancer of split 7 (Drosophila)
Aloxe3	68,939,879	68,962,616	Arachidonate lipoxygenase 3
Alox12b	68,970,574	68,983,293	Arachidonate 12-lipoxygenase, 12R type
Alox8	68,997,387	69,011,341	Arachidonate 8-lipoxygenase
Gucy2e	69,036,537	69,050,147	Guanylate cyclase 2e
Cntrob	69,112,998	69,137,277	Centrobin, centrosomal BRCA2-interacting protein
Trappc1	69,137,482	69,139,295	Trafficking protein particle complex 1
Kcnab3	69,139,760	69,146,543	Potassium voltage–gated channel, shaker-related subfamily, beta member 3
A030009H04Rik	69,154,319	69,156,143	RIKEN cDNA A030009H04 gene
Chd3	69,156,775	69,182,928	Chromodomain helicase DNA-binding protein 3
Cyb5d1	69,207,114	69,208,848	Cytochrome b5 domain containing 1
Lsmd1	69,209,293	69,210,173	LSM domain containing 1
Tmem88	69,210,021	69,211,736	Transmembrane protein 88
Jmjd3	69,212,010	69,227,177	Jumonji domain containing 3
Dnahc2	69,234,311	69,362,610	Dynein, axonemal, heavy chain 2
Efnb3	69,367,627	69,373,680	Ephrin B3
Wdr79	69,375,256	69,392,826	WD repeat domain 79
LOC100043030	69,387,322	69,387,686	Similar to ribosomal protein S27 (metallopanstimulin 1)
Trp53	69,393,483	69,405,373	Transformation-related protein 53
Atp1b2	69,413,252	69,419,462	ATPase, Na+/K+ transporting, beta 2 polypeptide
Shbg	69,428,318	69,431,370	Sex hormone–binding globulin
Sat2	69,435,611	69,437,371	Spermidine/spermine N1-acetyl transferase 2
LOC382532	69,444,405	69,444,814	Similar to ribosomal protein S15a
Fxr2	69,446,473	69,466,799	Fragile X mental retardation, autosomal homolog 2
Sox15	69,468,875	69,470,123	SRY-box–containing gene 15
Mpdu1	69,470,206	69,476,144	Mannose-P-dolichol utilization defect 1
Cd68	69,477,873	69,479,564	CD68 antigen
Eif4a1	69,480,441	69,485,817	Eukaryotic translation initiation factor 4A1
Senp3	69,486,620	69,495,348	SUMO/sentrin-specific peptidase 3
Tnfsf13	69,496,079	69,499,056	Tumor necrosis factor (ligand) superfamily, member 13
Tnfsf12-tnfsf13	69,496,079	69,509,499	Tumor necrosis factor (ligand) superfamily, member 12-member 13
Tnfsf12	69,499,850	69,509,256	Tumor necrosis factor (ligand) superfamily, member 12
LOC100041994	69,543,497	69,545,211	Hypothetical protein LOC100041994
Polr2a	69,547,912	69,571,725	Polymerase (RNA) II (DNA directed) polypeptide A
Amac1	69,573,386	69,575,346	Acyl-malonyl–condensing enzyme 1
Zbtb4	69,588,908	69,593,175	Zinc finger and BTB domain containing 4
Chrnb1	69,597,539	69,609,445	Cholinergic receptor, nicotinic, beta polypeptide 1 (muscle)
Fgf11	69,611,704	69,615,127	Fibroblast growth factor 11
Tmem102	69,617,105	69,619,126	Transmembrane protein 102
4933402P03Rik	69,630,068	69,631,942	RIKEN cDNA 4933402P03 gene
1700095G12Rik	69,634,373	69,635,667	RIKEN cDNA 1700095G12 gene
Nlgn2	69,636,625	69,648,351	Neuroligin 2
1810027O10Rik	69,652,027	69,653,060	RIKEN cDNA 1810027O10 gene
Plscr3	69,660,178	69,665,560	Phospholipid scramblase 3
Tnk1	69,664,601	69,672,219	Tyrosine kinase, nonreceptor, 1
LOC432576	69,690,186	69,691,607	Hypothetical LOC432576
Kctd11	69,691,179	69,694,908	Potassium channel tetramerisation domain containing 11
Centb1	69,695,069	69,709,041	Centaurin, beta 1
2810408A11Rik	69,710,860	69,714,488	RIKEN cDNA 2810408A11 gene
0610025P10Rik	69,715,393	69,727,322	RIKEN cDNA 0610025P10 gene
Gps2	69,727,694	69,730,093	G-protein pathway suppressor 2
Eif5a	69,730,224	69,734,843	Eukaryotic translation initiation factor 5A
Ybx2	69,749,401	69,755,101	Y-box protein 2
Slc2a4	69,755,788	69,761,692	Solute carrier family 2 (facilitated glucose transporter), member 4
LOC668442	69,764,300	69,764,986	Similar to Cofilin-1 (Cofilin, nonmuscle isoform)
Cldn7	69,779,003	69,781,380	Claudin 7
Rai12	69,781,726	69,794,012	Retinoic acid induced 12
Dullard	6,9794,670	69,804,103	Dullard homolog (Xenopus laevis)
Gabarap	69,804,872	69,808,448	Gamma-aminobutyric acid receptor–associated protein
Phf23	69,809,279	69,813,418	PHD finger protein 23
Dvl2	69,814,128	69,823,611	Disheveled 2, dsh homolog (Drosophila)
Acadvl	69,823,694	69,828,909	Acyl-Coenzyme A dehydrogenase, very long chain
Dlg4	69,832,366	69,858,270	Discs, large homolog 4 (Drosophila)
Asgr1	69,867,871	69,871,396	Asialoglycoprotein receptor 1
Asgr2	69,906,146	69,919,689	Asialoglycoprotein receptor 2
Mgl2	69,943,859	69,951,044	Macrophage galactose N-acetyl-galactosamine–specific lectin 2
Mgl1	69,980,276	69,984,336	Macrophage galactose N-acetyl-galactosamine–specific lectin 1
Slc16a11	70,027,582	70,029,915	Solute carrier family 16 (monocarboxylic acid transporters), member 11
Slc16a13	70,030,294	70,034,496	Solute carrier family 16 (monocarboxylic acid transporters), member 13
Bcl6b	70,037,630	70,043,223	B-cell CLL/lymphoma 6, member B
0610010K14Rik	70,048,710	70,051,416	RIKEN cDNA 0610010K14 gene
D11Bwg0434e	70,051,628	70,053,348	DNA segment, Chr 11, Brigham and Women's Genetics 0434 expressed
Alox12	70,054,957	70,068,843	Arachidonate 12-lipoxygenase

Data retrieved from NCBI Build 37.1 (February, 2008).

Candidate Gene Sequencing

Three genes within the 2.6335-Mb region harboring MM were selected for sequencing based on having well-characterized roles in muscle development. For sox15, no genetic variation was observed across the entire gene sequence between any of the samples. For chrnb1, 2 single nucleotide polymorphisms and a 3-bp deletion (GGA/-) were observed between C57BL/6J and HR mice within the 3′ UTR (data not shown). However, none of the sequence variants found within chrnb1 correlated with the MM phenotype within HR mice. For rpl26, many polymorphisms were found between C57BL/6J and HR mice (data not shown), but again, none of this genetic variation segregated in a manner that provided evidence for being causal for MM.

Discussion

Mapping and subsequent identification of spontaneous mutations with major phenotypic effect in mice have played a significant role in furthering our understanding of many physiological processes. These include growth (e.g., growth hormone–releasing hormone receptor; Godfrey et al. 1993), obesity (e.g., leptin; Zhang et al. 1994), and reproduction (e.g., gonadotropin-releasing hormone; Mason et al. 1986). Spontaneous mutations in genes regulating muscle development have also been identified. For example, the “compact” mutation causing hypermuscularity in mice (Varga et al. 1997) was subsequently identified as a deletion in myostatin (mstn) (Szabó et al. 1998), a finding that was driven by the discovery that muscles of myostatin knockout mice weigh 2–3 times more than those of wild-type animals (McPherron et al. 1997). A second example is the “high-growth” mutation, which was first observed because it caused a 30–50% increase in postweaning growth rate of mice (Bradford and Famula 1984). The underlying mutation was later identified as a ∼500-kb deletion of a region containing the gene cytokine signaling 2 (Socs2), a member of a family of regulators of cytokine signal transduction (Horvat and Medrano 2001).

In this study, we have mapped the recessive MM mutation, which causes most hindlimb muscles to have markedly reduced mass, to a 2.6335-Mb interval on MMU11. Effects of this MM mutation were originally observed in 2 of 4 replicate lines during the course of long-term selective breeding for high levels of voluntary wheel running and in 1 of 4 control (nonselected) lines (Garland et al. 2002). Analyses of data from the first 22 generations of the selection experiment indicated that the mutant allele was present at a frequency of approximately 7% in the base population (outbred Hsd:ICR mice). It is intriguing that both the compact and high-growth mutations were also originally discovered during the course of long-term selection experiments, for increased carcass protein content (Varga et al. 1997) and rapid weight gain (Bradford and Famula 1984), respectively. It is not known if the compact mutation was segregating in the base population prior to selection. For high growth, the mutation arose spontaneously after more than 20 generations of selection. The appearance of major mutations affecting a trait under selection has been documented many other times, as discussed by Varga et al. (1997). The strong selection intensity placed on quantitative traits during artificial selection experiments is well suited to take advantage of preexisting genetic variation in the base population or newly created genetic variation during the selection experiment, arising due to spontaneous mutations with major phenotypic effect.

Although our mapping efforts have succeeded in localizing MM to a reasonably small genomic interval, the goal of identifying the genetic variation responsible for the MM phenotype remains a formidable challenge. Mouse chromosome 11 is gene rich, with an average of 14 genes per Mb compared with the genome-wide average of 6.5. However, the 2.6335-Mb MM interval on MMU11 is extremely dense with genes, harboring 102 expressed and predicted genes (NCBI Build 37.1; February, 2008). Furthermore, many of the genes within this region represent viable candidates based on known physiological roles in muscle development. For example, SRY-box containing gene 15 (sox15) plays a role determining the early myogenic cell lineages during mouse skeletal muscle development (Lee et al. 2004). Mutations within cholinergic receptor, nicotinic, beta polypeptide 1 (chrnb1) have been shown to cause congenital myasthenic syndromes in humans (Quiram et al. 1999), as have mutations within the epsilon subunit (chrne), which falls just ∼500 kb distal to the MM region. Ribosomal protein L26 (rpl26) is very strongly upregulated (∼35-fold increase) in satellite cells during muscle regeneration (Kardon G, personal communication). Although we found no causal sequence variation within these 3 genes, additional polymorphisms could exist in regions outside of those in our sequencing design.

In addition to these 3 candidate genes, many of the other ∼100 genes in the MM region are expressed in muscle, and the Acadvl, Myh10, Slc2a4, and Trp53 genes all have effects on muscle when genetically modified (Mouse Genome Informatics database; February, 2008). Many of the genes in this region still have unknown function and may be involved in muscle development or regulation. Further, there are several microRNA genes (miR-324-3P, miR-497, miR-195) in the MM region that target genes with phenoconnectivity to muscle physiology. Finally, the genetic variation underlying MM could be acting via trans-regulatory effects on other genes with physiological roles in muscle development, including excellent candidates that are just ∼500 kb proximal to the MM interval (e.g., myosin, heavy polypeptide 3, skeletal muscle, embryonic; myh3). Disruption of myh3 results in reduced muscle mass similar to what is observed in MM (Acakpo-Satchivi et al. 1997; Allen et al. 2001) and a missense mutation in this gene results in autosomal dominant myopathy in humans (Martinsson et al. 2000). A polymorphism within the MM region could have regulatory effects on this closely adjacent gene or on unlinked genes that regulate muscle.

In summary, a recessive MM mutation causing reduced hindlimb muscle mass in mice has been mapped to a 2.6335-Mb interval on MMU11 containing ∼100 genes, many of which have known roles in muscle development. Identification of the genetic variation that underlies MM could potentially be very important in understanding both normal muscle function and disregulation of muscle physiology leading to disease. However, finding the DNA polymorphisms that cause MM will be very challenging given the gene density and complexity of the mapped region.

Funding

United States National Science Foundation (IOB-0543429 to T.G.); the National Institutes of Health (DK076050 to D.P.).