Background Art Marbling is visible intramuscular fat, which contributes to tenderness and flavour of beef and underpins quality grades in several countries. Higher marbling scores are associated not only with increased consumer satisfaction but also with increased profit to the beef industry. The degree of marbling is affected not only by the environment, but it is also expected to be influenced by many genes, some of which may have a significant individual effect on marbling. Only two genetic effects appear to have a"congenital"effect on marbling, both serving to reduce it drastically, which is not generally desirable. These are mutations at the myostatin gene associated with double muscling and the presence or absence of a Y chromosome (Hanset et al., 1982; Worrell et al. , 1987). No congenital mutations have yet been described that radically increase marbling score in cattle. Higher marbling scores result from feeding cattle excess calories in feedlots, and the differences in individual performance is usually attributed to individual and breed genetic differences.

Marbling score is difficult to measure and hence difficult to improve. Firstly, marbling score is a categoric trait (Sokal and Rohlf, 1981) scored to a visual standard by a trained assessor, so there is a subjective element to it as well. Secondly, it is usually measured in the dead chilled carcass, which makes genetic

improvement difficult. Thirdly, bulls show only trace amounts (Savell et al. , 1986) of marbling, which means that expensive progeny tests are required as surrogates for measuring the animal itself. Proxies for marbling score can be obtained from ultrasound images of scanned muscle (Brethour, 2000) but the correlation to chiller estimates leaves room for significant misclassification.

Marbling score represents an underlying level of intramuscular fat, but measurement of chemically extracted intramuscular fat is even more time consuming and expensive than progeny tests, although there is less of the subjective element to them. Accordingly, there remains a need for a reliable and efficient method for measuring the propensity for marbling in live beasts.

Summary of the Invention The present invention provides a genetic approach to measuring the propensity for marbling. The telomeric part of bovine chromosome 3 was implicated in predicting marbling scores by studying animals of extreme marbling scores and then confirming the results in the wider bovine population. The DNA markers FCGR1A, ILSTS96 and RM19 were found to be associated with the characteristic of marbling. To determine which gene was likely to be causative and to find a diagnostic test that could be used to predict marbling (1) the human comparative map near FCGR1A was examined (2) polymorphisms in candidate genes near FCGR1A were generated by sequencing a panel of individuals and (3) these polymorphisms were tested for consistent genetic effect on marbling in the group of extreme animals previously mentioned. With this approach a plurality of single nucleotide polymorphisms (SNPs) in the RORC (retinoid related orphan receptor C (gamma)- also known as RAR-related orphan receptor gamma, nuclear receptor ROR-gamma and retinoic acid-binding receptor gamma and identified using the alternative symbols NRIF3, RORG, RZRG and TOR) gene that showed consistent

association between genotypes or alleles and marbling were identified. Testing on animals with intramuscular fat measurements available confirms an association with fat percentage.

Accordingly in a first aspect of the present invention there is provided a method for assessing the propensity for marbling in meat derived from an animal, comprising the step of testing the animal for the presence or absence of one or more alleles of the gene encoding the retinoid related orphan receptor C (gamma) (RORC) associated with increased fat deposition in muscle tissue, and/or genetic variation located external to the RORC gene which shows allelic association therewith.

In an embodiment the RORCGH polymorphism may form the basis of the test. The RORCGH polymorphism was detected in the fragment putatively containing the seventh exon and seventh and eighth introns of RORC which has the DNA sequence shown in Table 4.

In the fragment illustrated in Table 4, and allele 1 has adenine at base 693 (SEQ ID NO : 1), while allele 2 has guanine in this position (SEQ ID NO : 2).

Allele 1 is associated with high marbling scores and allele 2 with lower marbling scores. Therefore, an animal possessing allele 1 is more likely to be selected where marbling is desirable.

In a further embodiment the RORCA polymorphism may form the basis of the test. This polymorphism is located in the DNA fragment whose sequence is given in Table 10 and which putatively contains exon 1 of RORC.

Allele 1 (otherwise referred to as allele a) in which base 322 of the fragment illustrated in Table 10 is adenine (SEQ ID N0 : 3) is associated with lower marbling scores and allele 2 (otherwise referred to as allele g), where base 322 of the fragment illustrated in Table 10 is guanine (SEQ ID NO : 4) is associated with higher marbling scores.

In a further embodiment the RORCE polymorphism may form the basis of the test. The DNA sequence for a

fragment putatively containing intron 5 of RORC is given in Table 10, with the alternative allelic forms being set forth in SEQ ID NO : 5 and SEQ ID NO : 6.

In an embodiment there is provided a method of assessing the propensity for marbling in meat derived from animal, comprising the step of testing the animal for the presence or absence of allele 1 of the RORCGH polymorphism and/or allele 2 of the RORCA polymorphism and/or the allele of RORCE associated with increased fat disposition in muscle tissue.

Where there has been a recent reduction in population size for a species, particular haplotypes of individuals will be relatively over-represented. If insufficient time has elapsed to cause allelic association to decay, there will be linkage disequilibrium even for alleles that are far apart. Livestock species such as cattle have been domesticated from a relatively small pool of wild ancestors in recent times, and therefore in these species allelic association is found between alleles that may be remote physically. Thus, it may be expected that regions of genetic variation that are outside the RORC gene will also show allelic association with the polymorphism in the RORC gene described above, and therefore will be suitable genetic markers for marbling.

It has been shown that FCGR1A and RM19 are associated with marbling in different experiments at levels of statistical significance. Since neither association is found in both experiments this suggests a boundary for the region in which markers for marbling may be found.

Accordingly, in a second aspect of the present invention there is provided a method for assessing the propensity for marbling in the meat of cattle comprising the step of testing for the presence or absence of at least one genetic marker characteristic of marbling located on bovine chromosome 3 in a region bounded by the genetic markers FCGR1A and RM19.

Typically the DNA markers are ILSTS96, FCGR1A, RM19 or RME23 or the RORC polymorphism described above but other polymorphisms in the gene and haplotypes between polymorphisms at the gene as well as other adjacent DNA markers could be used.

It will be appreciated that animals may be selected for greater or lesser propensity for marbling using the method of the invention. The presence of one or more alleles associated with increased fat deposition implies a greater propensity for marbling, while the absence of such alleles implies a lesser propensity for marbling. As a leaner meat may be desirable, testing for a lesser propensity for marbling is as much a part of the invention as testing for an increased propensity for marbling.

According to a third aspect of the present invention there is provided a genetic marker for marbling in the meat of an animal which is a polymorphic form of the RORC gene, being the RORCGH, RORCA or RORCE polymorphism.

According to a fourth aspect of the present invention there is provided an isolated nucleic acid comprising the nucleotide sequence set forth in SEQ ID NO : 1, SEQ ID NO : 2, SEQ ID N0 : 3, SEQ ID NO : 4, SEQ ID NO : 5 or SEQ ID NO : 6.

According to a fifth aspect of the present invention there is provided an isolated nucleic acid consisting of the nucleotide sequence set forth in SEQ ID MO : 1, SEQ ID NO : 2, SEQ ID NO : 3, SEQ ID NO : 4, SEQ ID N0 : 5 or SEQ ID N0 : 6.

According to a sixth aspect of the present invention there is provided a method for identifying an animal suitable for fattening to enhance marbling in meat derived from the animal, comprising the steps of: (1) testing the animal for the presence of one or more alleles of the gene encoding the retinoid related orphan receptor C (gamma) (RORC)

associated with increased fat deposition in muscle tissue, and/or genetic variation located external to the RORC gene which shows allelic association therewith ; and (2) selecting animals where one or more of said alleles or said genetic variation is present for fattening.

According to a seventh aspect of the present invention there is provided a method for identifying a cow likely to yield meat showing an increase in marbling, comprising the steps of: (1) testing the cow for the presence of at least one genetic marker characteristic of marbling located on bovine chromosome 3 in the region bounded by the genetic markers FCGR1A and RM19 ; and (2) selecting a cow that shows genetic variation associated with an increase in marbling.

According to an eighth aspect of the present invention there is provided a method for selecting an animal for breeding in order to improve marbling characteristics in a herd, comprising the steps of: (1) testing a plurality of animals for the presence of one or more alleles of the gene encoding the retinoid related orphan receptor C (gamma) (RORC) associated with increased fat deposition in muscle tissue, and/or genetic variation located external to the RORC gene which shows allelic association therewith ; and (2) selecting at least one animal where one or more of said alleles or said genetic variation is present for breeding.

According to an ninth aspect of the present invention there is provided a method for selecting a bull or cow for breeding to improve marbling characteristics in a herd, comprising the steps of: (1) testing a plurality of bulls and/or cows for the presence of at least one genetic marker

characteristic of marbling located on bovine chromosome 3 in a region bounded by the genetic markers FCGR1A and RM19 ; and (2) selecting at least one bull and/or cow that shows genetic variation associated with an increase in marbling for breeding.

Advantageously in order to assess the propensity for fat deposition in an animal and/or to select an animal to yield meat with improved marbling, testing may comprise the steps of: (1) obtaining a biological sample from the animal ; (2) extracting nucleic acids from the sample ; (3) amplifying at least those sections of the extracted nucleic acids which contain one or more genetic markers as described above ; and (4) establishing the presence or absence of alleles associated with increased fat deposition in muscle tissue in the amplified nucleic acids.

Preferably the biological sample is blood, but other biological samples from which nucleic acids can be amplified may be used. For example, hair root samples, cheek scrapings, skin samples and the like may be used.

Typically amplification is performed using the polymerase chain reaction (PCR), but other DNA amplification methods such as the ligase chain reaction are well known in the art and could be used.

Identification of the polymorphisms may be achieved through any suitable method. Techniques that feature primer extension, single strand conformational technology or DER chip base technologies could be used.

According to a tenth aspect of the present invention there is provided an oligonucleotide probe for amplification of a genetic marker as described above.

According to an eleventh aspect of the present invention there is provided a kit for use in assessing the propensity for fat deposition in muscle tissue of an animal and/or selecting an animal of high marbling score,

comprising means for amplifying an allele or genetic variation as described above.

The design of the oligonucleotide primers is well within the skill of the person skilled in the art. All primers that amplify a region of genetic variation associated with increased marbling between FCGR1A and RM19 are envisaged. Any primer that can amplify a particular example of such a genetic variation, such as the polymorphism in the RORC gene described above, may be used, and it will be appreciated that such primers can vary in sequence and length.

As an alternative to analysis of the nucleic acids of the animal, indirect detection of genetic markers through analysis of the expressed protein, or in any other suitable way, may be undertaken. The RORC gene is transcribed into mENA and this mRNA is then translated into protein. Differences in mRNA sequence, abundance and location in different tissues may be used to predict marbling. Moreover, differences in the RORC protein sequence, abundance or location in different tissues might be used to predict marbling. These differences could be detected by methods that do not depend on DNA, such as ELISA assays using monoclonal antibodies to the RORC protein, as one example.

The methods of the invention may be used both for the selection of breeding animals and for the selection of unpedigreed animals for entry into feed lots. In the latter case the methods of the invention allow for animals with unsuitable pedigrees to be excluded from feedlots on the basis that grain feeding is unlikely to increase marbling to the same extent as it would in more suitable animals. Alternatively such measurements may allow for determination of the optimum time to reach maximum marbling.

The invention is therefore also concerned, in further aspects, with animals when selected by the method of the invention, their progeny and the use of both

selected animals and their progeny for breeding as well as meat from these animals.

The methods of the invention are applicable to animals including but not limited to cattle and other bovids, including water buffalo and bison, and to other ungulates including sheep, goats, deer and pigs.

Throughout this specification and the claims, the words"comprise","comprises"and"comprising"are used in a non-exclusive sense, except where the context requires otherwise.

It will be clearly understood that, although a number of prior art publications are referred to herein, this reference does not constitute an admission that any of these documents forms part of the common general knowledge in the art, in Australia or in any other country.

Brief Description of the Drawings Preferred embodiments of the present invention will now be described, by way of example only, with reference to the accompanying drawings in which : Fig. 1 is a radiation hybrid map of the relevant portion of bovine chromosome 3; and Fig. 2 is a plot of animal genotype versus marbling score, which shows the association of allele 1 of RORCGH with increased marbling.

Modes for Performing the Invention Materials and Methods DNA The Cattle DNA samples have been described previously (Barendse, 1997 ; Barendse et al. , in press).

In short, a first sample (CS. 254) consists of 177 animals of Angus and Shorthorn ancestry of extreme marbling scores, representing a sample of approximately 2,500 animals for which blood was collected. Animals were

collected from many abattoirs along the east coast of Australia. The animals were each fed for more than 250 days in a feedlot. A second sample (SBEF. 018) consists of 1750 animals of all marbling scores collected over 9 months from the AMH Toowoomba feed lot. They were each fed for less than 250 days in a feedlot. The bovine- hamster radiation hybrid panel has been described (Womack et al. , 1997).

Polymorphisms We screened for polymorphisms in positional candidate genes on chromosome 3 by examining intronic and untranslated DNA preferentially since bovine coding sequence is less variable than that of many other species. DNA sequences were generated by direct sequencing of DNA from a panel of at least 10 animals using the ABI BigDye Terminator DNA sequencing kit following the manufacturer's instructions (Applied Biosystems Incorporated, Foster City, CA). The DNA fragments were separated at the AGRF (Australian Genome Research Facility). DNA sequences were examined for evidence of polymorphisms by inspection.

Bovine DNA sequence did not exist for all genes, and even when it did, in many cases there was insufficient untranslated or intronic sequence to test for polymorphism. To generate intronic sequences, the putative intron/exon boundaries for cattle were inferred from the human and mouse DNA sequence, introns for sequencing were chosen to be less than 2 kbp (kilo base pairs) in length, and DNA primers were designed to amplify and sequence bovine DNA. Bovine DNA sequence was not often available and primers were designed by making a consensus of the human and mouse DNA sequence for the gene and using the mammalian consensus sequence to generate primers.

Assay DNA from all of these animals had been prepared either as in Barendse et al. , (in press) or by taking 100

pi of whole blood and purifying that using the Qiagen kit following the manufacturers instructions. In the Barendse et al. process DNA was extracted from the blood samples using the following method. A ten ml sample of blood (EDTA as anti-coagulant, Becton Dickonson vacutainers) was lysed in 40 mls of chilled buffered water (79 mg Sodium Bicarbonate, 7.6 g Ammonium Chloride per litre) for at least 15 minutes. White cells were collected by centrifugation and the pellet was completely and vigorously resuspended in 4.5 mls of 1 x Phosphate Buffered Saline. Then 0.25 mls of 0.5 M di-Sodium Ethylene Diamine Tetra-acetic Acid (EDTA) (pH 8. 0), 0.25 mls of 10% Sodium Dodecyl Sulphate (SDS) and 50 pi of Proteinase K (Boehringer ; 10 mg/ml in water) were added, the tubes rolled gently to mix and the samples incubated overnight at 37 degrees Celsius. Following this, 2.4 mls of a 5 M Sodium Chloride/0.2% 2-Mercaptoethanol solution was added, the solution mixed on a roller for 5 minutes and then 5 mls of Analytical Grade Chloroform was added.

The samples were mixed gently for at least 60 minutes after which the organic phase was separated from the aqueous phase by centrifugation. The aqueous phase (top) was removed with a large-bore glass pipette and the DNA precipitated with an equal volume of chilled iso-propanol.

The spooled DNA was fished out on a sealed glass pipette and transferred to 70% Ethanol to remove salt overnight.

The spool of DNA was then transferred to 1 x TE (10 mM Tris, lmM EXTA, pH 8.0). This is a modified procedure of Barendse et al. (1993) and of Mullenbach et al. (1989).

DNA was diluted to approimately 50 ng/pl for the working stock, plated into a 96 well format with known, diagnostic blanks and repeated samples and stored at-20 C between uses.

RORCGH was sampled using the MGB Taqman assay (Livak et al., 1995 ; Livak, 2003). The probes are RORCGHSNP1-1 6FAM-5'TCGGACTTCTCTTGCT (SEQ ID NO : 7) and RORCGHSNP1-2 VIC-5'TCGGACTCCTCTTGCT (SEQ ID NO : 8) and the

primers are RORCGHU3 5'CTGACAATGACACAGTCTTTTTTGAA (SEQ ID NO : 9) and RORCGHD3 5'ACGCCCTGGGTCTGGAA (SEQ ID NO : 10) derived from the sequence in Table 4. The Taqman PCR reaction was performed in 10 pi reactions on 50 ng of DNA using the Taqman Universal PCR Mastermix on a 384 well ABI 9700 PCR machine for 40 cycles at an annealing extension temperature of 60 C for 60 seconds. The fluorescent signal was read as an allele discrimination read using the ABI-7900HT Sequence Detection System and analysed using v2.1 of the SDS software. The genotypes were then checked by at least 2 individuals.

RORCA was sampled using a modified version of the single base primer extension reaction (Syvanen, et al., 1990, ciao and Kwok, 2003). The DNA around the polymorphism was amplified using the primers RORCA322GAU1 5'AAGTAACAGGAGAGCACAGTCAGA (SEQ ID NO : 11) and RORCA322GAD1 5'AAGGAGTCCCTGTGAAGAAGC (SEQ ID NO : 12) in 10 pl reactions on 50 ng of DNA using the Taqman Universal PCR Mastermix on a 384 well ABI 9700 PCR machine for 40 cycles at an annealing extension temperature of 60 C for 60 seconds.

Primers and unincorporated nucleotides were removed by adding 3.0 pl of an enzyme cocktail containing 1.0 unit of Exonuclease I and 0.5 units of Calf intestinal alkaline phosphatase and incubating the reaction for 60 minutes at 37 degrees Celsius. The enzyme cocktail was then denatured by incubating the reaction for 15 minutes at 80 degrees Celsius. To detect the polymorphism by single base extension, 5.0 pi of a single base extension mixture was then added consisting of 0.1 pH of the extension primer RORCA322GAPX2 5'ACACAGCAGCTCCTCACAGAG (SEQ ID NO : 13), 0. 5 units of Thermosequenase (Amersham Biosciences, New Jersey, USA), 0. 6 x Concentrated Thermosequenase reaction buffer, 0.05 pM R6G labelled dideoxycytosine triphosphate (R6G-ddCTP), 0.05 pM R110 labelled dideoxyuridine triphosphate (R110-ddUTP) (Perkin Elmer Life Sciences Inc., USA), 5 pM each of dideoxyadenosine triphosphate (ddATP) and dideoxyguanosine

triphosphate (ddGTP) (Roche Applied Science, Germany). The single base primer extension reaction was performed in an ABI 7900HT Sequence Detector System for 25 cycles of 95 degrees Celsius for 15 secs, 50 degrees Celsius for 5 secs and 60 degrees Celsius for 30 secs. Fluorescence was measured at 60 degrees Celsius through the ROX, JOE and FAM channels, the ROX for the internal standard in the Taqman Universal PCR Mastermix, the JOE for the R6G signal and the FAM for the R110 signal. The quenching of the R6G and R110 signals indicates incorporation into the primer, so homozygotes are scored if the quenching of one signal is 5 times greater than the quenching of the other signal, relative to the ROX constant for each sample. If both signals are quenched and the ratio of the R110 to R6G is between 0.3 and 3.0 then the animal is a heterozygote. Values for relative quenching outside of these ranges are scored as NA (not available). No template controls, in our system, show negative values for quenching. The RORCA polymorphism affects a restriction palindrome GCGC so could also be detected using the restriction endonucleases Hha I or HinPl I with separation on agarose gels.

Analysis The construction of radiation hybrid maps and the examination of radiation hybrid data was performed as described in (Barendse, 1997 and Barendse et al., 2000).

The association between polymorphisms and marbling were performed as previously reported (Barendse, 1997 ; Barendse et al., in press) for the polymorphisms TG5 and CSSM34/ETH10. In short, the genotypes were compared to the marbling scores by analysing contingency tables using log-likelihood statistics or generalised linear models (GLM) based on the Poisson distribution implemented using S-PLUS (Sokal and Rohlf, 1981; McCullagh and Nelder 1989, Anon 1995, Venables and Ripley 2000). The Poisson

distribution was used in the GLM since marbling score is a range of ordinal categories. This distribution accommodates contingency data having cells with low frequencies including zero, and allows the data to be partitioned into tables based on various factors without the need to perform sequential chi-square tests and then attempt to combine the individual chi- squares. The significance of the deviance, D, associated with each term is tested using the Chi- square distribution. Contingency tables were generated of numbers of individuals of each genotype or allele at each marbling score partitioned across factors which had first been shown to be statistically significantly associated with marbling in these data. Some individual contingency tables were analysed using G (adj), the log likelihood statistic with the Williams correction for small sample size (Sokal and Rohlf 1981). In some data sets the allele associated with higher marbling score was determined by comparing the gene frequency in animals at the low extreme of marbling with animals at the high extreme of marbling. DNA markers such as ILSTS96, FCGR1A and RM19 have large numbers of alleles (n) and very large numbers of genotypes (n (n+l)/2). Such large numbers of genotypes, even with the sample sizes found in this study, provide few animals for most genotypes across the entire experiment, with consequently broad standard errors for each genotype. To increase the numbers of animals in each comparison, the associations were performed by comparing alleles to marbling score.

The relative risk of each allele was calculated. The CS. 254 data set was divided into the pedigreed versus the random cattle while the SBEF. 018 represent only the random cattle. In the pedigreed cattle, families were excluded if they showed no variation for the marker and they did not contain animals of high and

low marbling. All random cattle in CS. 254 and SBEF. 018 were analysed. An experiment-wide P value was calculated following Fisher using-2ElnPi where Pi are the P values to be added. This results in a chi-square statistic with twice the number of degrees of freedom of the number of P values (Sokal and Rohlf, 1981). The degree of linkage disequilibrium for alleles at polymorphisms was estimated using D' (Lewontin, 1964).

Intramuscular Fat To test whether the associations between RORCGH and marbling were also reflected in intramuscular fat, DNA samples from cattle in the CRC DNA bank were genotyped at random and a total of 853 with joint intramuscular fat measurements and RORCGH genotypes were obtained. The intramuscular fat phenotypes were analysed using an Analysis of Variance (ANOVA) implemented via the S-PLUS software.

Intramuscular fat was modelled as a dependant variable of the fixed effects of market (Domestic, Korean, Japanese), cohort, finish (feedlot north, feedlot south, pasture north, pasture south), breed and genotype as well as the random effect of sire.

The least square means for each genotype were extracted from the model.

Results A radiation hybrid map was constructed using DNA markers from the genetic linkage map (eg.

Barendse et al. 1997) as well as DNA fragments for several genes expected to map to Bta 3 (Fig. l). The gene PIK4CB maps to the interval between FCGR1A and RM19. LEPR does not map to this interval and is some distance away from these markers. Inspection of the raw scores of the radiation hybrid panel confirms this (Table 1).

FCGR1A and RM19 are associated with marbling in different experiments at suggestive levels of statistical significance (Tables 2,3 and 8). Since neither association is found in both experiments this would place a boundary on the region in which the QTL for marbling is expected to be found, i. e. , between these two markers.

Since PIK4CB occurs between these two markers, positional candidates were sought near PIK4CB using the human gene map (located at genome. ucsc. edu, Genome Browser June 2002 freeze position chrl : 148587065-150677512, 2.1 Mega bp, 41 genes ; cf Genome Browser October 2000 freeze position chrl : 165161690-175169214, 10 Mbp). Polymorphisms for several candidates including PIK4CB were examined for polymorphisms. The candidates are, in order from centromeric to telomeric, cent--FCGR1A-ITGA10B - ( ? PRKAB2 ?)-PIK4CB-RORC--tel. Since RM19 is an anonymous DNA marker, the telomeric comparative boundary for the QTL is not known. The location of PRKAB2 is difficult to specify. In the 2000 release of the human genome sequence (genome. ucsc. edu) it is located next to PIK4CB but in the 2002 release of the human genome sequence it is not located to this interval at all. Polymorphisms for DNA fragments for ITGA10B, PRKAB2, PIK4CB and RORC were sought in cattle. Neither PIK4CB nor PRKAB2 show associations to marbling and so far we have not identified a SEP for ITGA10B.

The RORCGH DNA polymorphism (Table 4) shows statistically significant associations to marbling in both the extreme and general population DNA samples (Table 5 and 6). There are three sets of data for associations involving RORC. In the informative sires (Table 5a) the relative risk of the 1 allele and high marbling is 1.89, indicating that the 1 allele is found more often with higher marbling

scores, but the G (adj) = 2. 949, ldf, P = 0. 08592 is not statistically significant. In the random animals from CS. 254 (Table 5b), the relative risk of the 1 allele and high marbling is 2. 933'and the G (adj) = 5.611, 1 df, P = 0. 01787. In the random animals (Table 7, Fig 2), when the data are collapsed to genotypes versus marbling scores across breeds and days of slaughter, G (adj) = 36. 926, df=8, P = 0.00001. The relative risk for allele 1 and high marbling score is 2.998. However, we know that there are differences in genotype frequency between breed and affects of day of slaughter on marbling score in the random sample. When these are accounted for (Table 6), the association between marbling and genotype has a deviation D = 18. 853, df=8, P = 0.01566. Combining the P values of 0.08592, 0.01787 and 0.01566 for an experiment wide probability using the statistic-2ElnPi gives X26 = 21.28, P < 0.002.

The RORCGH DNA polymorphism (Table 9) shows statistically significant associations to intramuscular fat in the CRC DNA bank samples (F = 4.28, P = 0.014).

The allele 1 is associated with higher fat concentrations and the allele 2 is associated with lower fat concentrations. In particular, difference in least square means between the alternative homozygotes is 0.797 units while the mean for the sample is 3.303 units. The difference between homozygotes is 24 percent of the mean for the trait. The experiment wide probability for the experiment using the statistic-2ElnPi s 28= 25. 54, P < 0.0015.

After the RORCGH polymorphism was found to be associated with marbling, additional effort was expended to obtain other polymorphisms in and around the gene. One was found in intron 1, called RORCA (Table 10). The RORCA polymorphism also shows associations to marbling in the samples tested so far. Like RORCGH, RORCA alleles were not associated with marbling score in the informative

sires of CS. 254 (G (adj) = 0. 04, df = 1, n. s. ). In the random animals from CS. 254 (Table 11) the relative risk of allele 2 and high marbling is 4.64, greater than RORCGH, with G (adj) = 7. 485, df = 1, P = 0.00624. The genotypes are also associated with marbling (Table 11) with G (ad) = 6.977, df = 1, P = 0.03055.

A joint analysis of RORCA and RORCGH was performed on the CS. 254 animals. The joint RORC A-GH genotypes were analysed into haplotypes and compared to marbling scores (Table 12). The haplotypes are significantly associated with marbling G (adj) = 8. 054,2 df, P = 0.01783 and more significant than either set of genotypes. The haplotypes are consistent with the individual polymorphisms, as would be expected. The degree of linkage disequilibrium between the alleles at the A and the GH polymorphism was estimated using D' (Lewontin, 1964) calculated for these samples (Table 13).

Out of a maximum possible value of 1.00, the D'for RORC A-GH is 0.93, which reflects that these two polymorphisms are within the same gene, and that linkage disequilibrium between them is very high. Assuming that low and high marbling can be treated as alleles, the D'was calculated between marbling and RORCA, RORCGH, and the RORCA-GH haplotypes respectively. The values are D'= 0. 57 for RORC A, D'= 0. 24 for RORC GH, and D'= 0. 84 for the RORC A-GH halotypes. This would agree that the haplotypes are a better predictor of marbling than either of the two polymorphisms by themselves.

Discussion A consistent association has been found between marbling scores and alleles of the SNPs RORCGH, RORCA and RORCE in several independent samples derived from the DNA samples collected for both extreme and random animals. In all samples, for RORCGH the relative risk is positive and greater than 1 for allele 1 and higher marbling scores, and in the

random samples both relative risks are between 2.9 and 3.0. The association between genotypes and intramuscular fat is consistent with that for marbling, and the 1 allele is associated with higher fat levels while the 2'allele is associated with lower fat levels just as the'1'allele is associated with higher marbling scores and the 2 allele with lower marbling scores.

RORCGH sits in the region that contains RM19 and FCGR1A, two DNA markers that are adjacent framework markers on the bovine linkage map (Barendse et al., 1997 ; www. cgd. csiro. au) and both have been associated herein to marbling. PIK4CB maps between these two markers and the orientation of the two maps is such that RORCGH is expected to be closer to RM19 than to FCGR1A.

Additional polymorphisms at RORC support the association between RORC and marbling, while haplotypes of these polymorphisms show greater predictability for marbling than each polymorphism by itself. The high degree of linkage disequilibrium found between the RORC polymorphisms indicates that other polymorphisms in and around the gene are expected to show associations to marbling. This is consistent with the data for RM19 and FCGR1A which also show associations to marbling. DNA markers from this region would thus be expected to be predicting the marbling QTL described here.

RORC is a gene that is highly expressed in skeletal muscle and its function is consistent with a gene that might affect marbling. Firstly, it controls apoptosis in tissues ; mice that have had RORC knocked out show increases in sporadic apoptotic events in tissues (Kurebayashi et al. 2000 ; Sun et al. 2000). One theory for the development of marbling is that muscle cells die and are replaced by fat tissue, as the number of muscle cells is fixed

once the muscle has been formed. Secondly, RORC is a member of the retinoid orphan receptor group of DNA binding proteins and it is sensitive to both thyroid hormones and to vitamin A (Ortiz et al. 1995) and differences in the level of both thyroid hormone and vitamin A have effects on marbling (Torii et al.

1996 ; Mears et al. 2001) Industrial Applicability The present invention is useful in allowing selection and breeding of animals with improved fat deposition characteristics, particularly high marbling scores.

TABLES Table 1. Raw RH clone positives and negative for genes and markers on Bta3. NAME CHROM RHVECTOR RM19 Chr3 000000011000000100001001001010010110000100000010001001000000 000000000000011000000000 010000 LEP-R Chr3 000000000000000100001001000000000100000010000000001000000000 020000010000100000001000 001001 FCGR1A Chr3 010000 010000 ILSTS96 Chr3 000000000000000101101001001010000110000000000010001000100010 000000000000010000110000 100000 RME23 Chr3 000000000000000101101001001010000110000100000000001001100010 000000000000000000000000 010000 OSG Chr3 000000000000010100101001001010000110000101000000001101100210 000000000000100000010000 000010 NAK Chr3 010000000000010100101001001010000210000101000000001101100020 000002090000000000000000 000010 TGLA76 Chr3 900000000001010100001001000000000100000010000000001000100100 010000010000100000001002 001001 CD3Z Chr3 010001 010001 TGLA127 Chr3 000010000000000100100001000010010001010110000000100000100010 000000110000000000100000 001000 PIK4CB Chr3 000100000000000100101001001010000110000100090010021001100010 000009090000019000909000 010090

Table 2 Association between alleles at FCGR1 and marbling. marbling scores L H allele fl 51 49 f2 1 10 f3 25 23 f4 1 4 G (adj) = 8. 92, df = 3, P = 0.03 Table 3. The association between RM19 and marbling in extreme cattle Analysis of Deviance Table Poisson model Response: score Terms added sequentially (first to last) Df Deviance Resid. Df Resid. Dev Pr (Chi) NULL 719 9632.082 breed 2 67.839 717 9564.243 0.000000000 day 27 554.148 690 9010.095 0.000000000 allele 3 5250.913 687 3759.182 0.000000000 marbling 4 2394.364 683 1364.818 0.000000000 breed: day 6 38.777 677 1326. 041 0.000000792 breed: allele 6 102.031 671 1224.010 0.000000000 breed: marbling 8 42.563 663 1181.447 0.000001062 day: allele 81 309.135 582 872.312 0.000000000 day: marbling 108 449.277 474 423.034 0.000000000 allele: marbling 12 29.308 462 393.726 0.003544895

Table 4. DNA sequence of RORCGH fragment showing the location of the single nucleotide polymorphism.

Table 5 Association between RORCGH alleles and marbling a) Informative Sires marbling scores L H allele rl 19 30 r2 43 36 G (adj) = 2. 949,1 df, P=0.08592 b) Random Steers marbling scores L H allele rl 30 33 r2 24 9 G (adj) = 5. 611,1 df, P = 0.01787 Table 6. The association between RORCGH genotype and marbling in extreme cattle Analysis of Deviance Table Poisson model Response: score Terms added sequentially (first to last) Df Deviance Resid. Df Resid. Dev Pr (Chi) NULL 614 4067.342 breed 2 119.600 612 3947.742 0. 00000000 day 27 389.255 585 3558.487 0.00000000 marbling 4 1309.838 581 2248.650 0. 00000000 genotype 2 757.370 579 1491.279 0.00000000 breed: day 11 221.733 568 1269.546 0. 00000000 breed: marbling 8 33.419 560 1236.127 0.00005174 breed: genotype 4 527.442 556 708.685 0.00000000 day: marbling 108 235.986 448 472.699 0.00000000 day: genotype 54 155. 225 394 317.474 0.00000000 marbling: genotype 8 18.853 386 298.620 0.01566433

Table 7. Lumped marbling scores and genotypes for the extreme cattle data set. marbling scores ml m2 m3 m4 m5 genotype rll 55 426 418 121 26 rl2 39 236 158 45 7 r22 7 84 39 25 0 G (adj) = 36.926, df = 8, P = 0. 00001 Table 8. DNA markers along chromosome 3 tested for association to marbling in the extreme sample.

DNA markers Chromosome Association Interaction ILSTS96 3 19 cM 0.367 0.156 FCGR1A 3 24 cM 0.442 0.765 RM19 3 29 cM 0.004** 0.635 RME23 3 31 cM 0.178 0. 127 Distances between markers are determined from the genetic linkage map at www. cgd. csiro. au (note differences to the radiation hybrid map) Table 9. The association between RORCGH genotype and intramuscular fat in CRC DNA bank cattle.

Df Sum of Sq Mean Sq F Value Pr (F) market 2 873.2516 436. 6258 256.5661 0.00000000 cohort 10 476. 7448 47. 6745 28.0140 0.00000000 finish 3 341. 7584 113. 9195 66.9403 0.00000000 breed 5 164.6585 32. 9317 19.3510 0.00000000 sireid 133 385.9119 2.9016 1.7050 0.00002120 rorcgh 2 14.5717 7.2858 4.2812 0.01432476 Residuals 511 869.6230 1.7018 genotype mean rll rl2 r22 3.30323 0.315436 0.166389-0. 418182

Table 10. DNA sequence of RORCA fragment and RORCE fragment showing the location of the single nucleotide polymorphism.

Table 11. Association between RORCA genotypes or alleles and marbling Sample-random steers marbling scores Low High Genotypes rall 5 0 ral2 7 4 ra22 13 16 Gadj = 6. 98,2 df, P = 0.03055 marbling scores Low High Alleles ral 17 4 ra2 33 36 Gadj = 7. 485, 1 df, P = 0.00624 Table 12. Haplotypes between RORCA and RORCGH associated with marbling Samples-random steers marbling scores Low High Haplotypes ral-r2 10 1 ral-rl 1 0 (excluded from the analysis) ra2-r2 4 4 ra2-rl 25 29 Gadj = 8. 054,2 df, P = 0.01783

Table 13. Degree of linkage disequilibrium between RORCA, RORCGH and marbling.

I. Linkage disequilibrium between RORCA and RORCGH RORCGH rl r2 RORCA ral 2 48 ra2 150 60 D'= 0. 93 II. Linkage disequilibrium between RORCA and marbling marbling Low High RORCA ral 17 4 ra2 33 36 D'= 0. 57 III. Linkage disequilibrium between RORCGH and marbling marbling Low High RORCGH rl 30 33 r2 22 9 D'= 0. 24 IV. Linkage disequilibrium between RORCA-GH haplotypes and marbling <BR> <BR> <BR> <BR> <BR> <BR> <BR> marbling<BR> Low High RORCA-GH ral-r2 10 1 ra2-rl 25 29 D'= 0. 84

References The contents of the following documents are incorporated herein by reference: Anon. 1995. S-PLUS Guide to Statistical and Mathematical Analysis. MathSoft Inc. , Seattle, Washington.

Barendse, W. , Armitage, S. M. , Ryan, A. M. , Moore, S. S. , Clayton, D. , Georges, M., Womack, J. E. , and Hetzel, J. 1993. A genetic map of DNA loci on bovine chromosome 1. Genomics 18,602--608.

Barendse, W. 1997. Assessing lipid metabolism.

Patent W09923248 Patent US6383751 (Patent Application PCT/AU98/00882).

Barendse, W., Vaiman, D., Kemp, S. J., Sugimoto, Y., Armitage, S. M., Williams, J. L., Sun, H. S. , Eggen, A., Agaba, M., Aleyasin, S. A., Band, M., Bishop, M. D. , Buitkamp, J. , Byrne, K., Colins, F. , Cooper, L., Coppettiers, W. , Denys, B. , Drinkwater, R. D. , Easterday, <BR> <BR> K. , Elduque, C. , Ennis, S. , Erhardt, G. , Ferretti, L., Flavin, N. , Gao, Q. , Georges, M. , Gurung, R. , Harlizius, B. , Hawkins, G. , Hetzel, J. , Hirano, T. , Hulme, D., Jorgensen, C. , Kessler, M. , Kirkpatrick, B. W., Konfortov, B. , Kostia, S. , Kuhn, C. , Lenstra, J. A. , Leveziel, H., Lewin, H. A. , Leyhe, B. , Li, L. , Martin Burriel, I., McGraw, R. A. , Miller, J. R. , Moody, D. E. , Moore, S. S., Nakane, S., Nijman, I. J. , Olsaker, I., Pomp, D. , Rando, A. , Ron, M. , Shalom, A. , Teale, A. J. , Thieven, U., Urquhart, B. G. D. , Vage, D-I., Van de Weghe, A., Varvio, S., Velmala, R., Vilkki, J. , Weikard, R. , Woodside, C., Womack, J. E., Lanotti, M., and Saragoza, P. 1997. A medium-density linkage map of the bovine genome.

Mammalian Genome 8 21--28.

Barendse, W., Armitage, S. M. , Aleyasin, A. and Womack, J. E. 2000. Differences between the radiation hybrid and genetic linkage maps of bovine chromosome 5 resolved with a quasi-phylogenetic method of analysis.

Mammalian Genome 11, 369-372.

Barendse, W. , Bunch, R. , Thomas, M. , Armitage,

S. , Baud, S. and Donaldson, N. The TG5 Thyroglobulin gene test for a marbling QTL evaluated in feed lot cattle.

Australian Journal of Experimental Agriculture (in press).

Brethour, J. R. 2000. Using serial ultrasound measures to generate models of marbling and backfat thickness changes in feedlot cattle. Journal of Animal Science 78,2055--2061.

Grobet, L. , Martin, L. J. R. , Poncelet, D., Pirottin, D., Browers, B. , Riquet, J. , Schoeberlein, A., Dunner, S. , Menissier, F. , Massabanda, J. , Fries, R., Hanset, R. and Georges, M. 1997. A deletion in the bovine myostatin gene causes the double-muscled phenotype in cattle. Nature Genetics 17,71-74.

Hanset, R., Michaux, C. , Dessy-Doize, C. and Burtonboy, G. 1982. Studies on the 7th rib cut in double muscled and conventional cattle. Anatomical, histological and biochemical aspects in Muscle Hypertrophy of Genetic Origin and its use to Improve Beef Production'eds King, J. W. B. and Menissier, F. Martinus Nijhoff Publishers, The Hague.

He, Y. -W., Deftos, M. L. , Ojala, E. W. and Bevan, M. J. 1998. RORgt, a novel isoform of an orphan receptor, negatively regulates Fas ligand expression and IL-2 production in T cells. Immunity 9,797--806.

Hirose, T. , Smith, R. J. and Jetten, A. M. 1994.

RORg : the third member of ROR/RZR orphan receptor subfamily that is highly expressed in skeletal muscle.

Biochemical and Biophysical Research Communications 205, 1976--1983.

Kurebayashi, S. , Ueda, E., akaue, H. Patelr D. D., Medvedev, A., Zhang, F. and Jetten, A. M. 2000.

Retinoid-related orphan receptor g (RORg) is essential for lymphoid organogenesis and controls apoptosis during thymopoiesis. Proceedings of the National Academy of Sciences (USA) 97,10132--10137.

Lewontin RC 1964. The interaction of selection

and linkage. I. General considerations ; heterotic models.

Genetics 49,49-67.

Livak, K. J. 2003. SNP genotyping by the 5'- nuclease reaction in Methods in Molecular Biology, vol.

212: Single Nucleotide Polymorphisms Methods and Protocols ed P. -Y. Kwok, ppl29--147, Humana Press, Totowa New Jersey.

Livak, K. J. , Flood, S. A. J. , Marmaro, J. , Giusti, W. , and Deetz, K. 1995. Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization. PCR Methods and Applications 5,357--362.

McCullagh P and Nelder JA (1989). Generalized linear models, Second Edition, Chapman and Hall, London.

Mears, G. J. , Mir, P. S. , Bailey, D. R. C. and Jones, S. D. M. 2001. Effect of Wagyu genetics on marbling, backfat and circulating hormones in cattle. Canadian Journal of Animal Science. 81,65--73.

Medvedev, A. , Chistokhina, A. , Hirose, T. and Jetten, A. M. 1997. Genomic structure and chromosomal mapping of the nuclear orphan receptor RORg (RORC) gene.

Genomics 46,93-102.

Mullenbach, R. , Lagoda, P. J. L. , and Welter, C.

1989. An efficient salt-chloroform extraction of DNA from blood and tissues. Trends in Genetics 5,391.

Ortiz, M. A. , Piedrafita, F. J. , Pfahl, M. and Maki, R. 1995. TOR: a new orphan receptor expressed in the thymus that can modulate retinoid and thyroid hormone signals. Molecular Endocrinology 9,1679--1691.

Savell JW, Cross HR and Smith GC. (1986).

Percent ether extractable fat and moisture content of beef longissimus muscle as related to USDA marbling score.

Journal of Food Science 51,838 and 840.

Sokal RR and Rohlf FJ. (1981). Biometry.

Second Edition, W. H. Freeman & Co. , San Francisco.

Sun, Z., Unutmaz, D. , Zou, Y. -R., Sunshine, M. J., Pierani, A. , Brenner-Morton, S. , Mebius, R. E. and Littman,

D. R. 2000. Requirement for RORg in thymocyte survival and lymphoid organ development. Science 288,2369--2372.

Syvanen, A.-C., Aalto-Setala, K. , Harju, L., Kontula, K. and Soderlund, H. 1990. A primer guided nucleotide incorporation assay in the genotyping of aplolipoprotein E. Genomics 8,684--692.

Torii, S. , Matsui, T and Yano, H. 1996.

Development of intramuscular fat in Wagyu beef cattle depends on adipogenic substances present in serum. Animal Science 63,73--78.

Venables WN and Ripley BD. (2000). Modern Applied Statistics with S-PLUS. Third Edition. Springer Verlag New York Inc.

Womack, J. E., Johnson, J. S., Owens, E. K., Rexroad, C. E. III., Schlapfer, J. , and Yang, Y. -P. 1997.

A whole-genome radiation hybrid panel for bovine gene mapping. Mammalian Genome 8,854-856.

Worrell, M. A. , Clanton, D. C. and Calkins, C. R.

1987. Effect of weight at castration on steer performance in the feedlot. Journal of Animal Science 64,343--347.

Woolf B. (1955). On estimating the relation between blood group and disease. Annals of Human Genetics 19,251--253.

Xiao, M. and Kwok, P. -Y. 2003. DNA analysis by fluorescence quenching detection. Genome Research 13, 932--939.