Mapping of Multiple Quantitative Trait Loci Affecting Bovine Spongiform Encephalopathy

Copyright 2004 by the Genetics Society of AmericaDOI: 10.1534/genetics.104.026401

Mapping of Multiple Quantitative Trait Loci Affecting BovineSpongiform Encephalopathy

Chi Zhang, Dirk-Jan de Koning, Jules Hernandez-Sanchez, Chris S. Haley,John L. Williams and Pamela Wiener1

Roslin Institute (Edinburgh), Roslin, Midlothian EH25 9PS, United Kingdom

Manuscript received January 13, 2004Accepted for publication April 21, 2004

ABSTRACTA whole-genome scan was conducted to map quantitative trait loci (QTL) for BSE resistance or susceptibil-

ity. Cows from four half-sib families were included and 173 microsatellite markers were used to constructa 2835-cM (Kosambi) linkage map covering 29 autosomes and the pseudoautosomal region of the sexchromosome. Interval mapping by linear regression was applied and extended to a multiple-QTL analysisapproach that used identified QTL on other chromosomes as cofactors to increase mapping power. Inthe multiple-QTL analysis, two genome-wide significant QTL (BTA17 and X/Yps) and four genome-widesuggestive QTL (BTA1, 6, 13, and 19) were revealed. The QTL identified here using linkage analysis donot overlap with regions previously identified using TDT analysis. One factor that may explain the disparitybetween the results is that a more extensive data set was used in the present study. Furthermore, methodolog-ical differences between TDT and linkage analyses may affect the power of these approaches.

BOVINE spongiform encephalopathy (BSE) is a ods following experimental challenge of sheep with dif-chronic, degenerative disorder affecting the cen- ferent sources of scrapie and BSE (see Hunter 1999

tral nervous system of cattle (http://www.bse.org.uk). for review). In humans, the sporadic forms of CJD andIt was first described in 1987 (Wells et al. 1987) and vCJD both seem to be associated with certain PrP geneconfirmed to be a transmissible spongiform encepha- polymorphisms (Palmer et al. 1991; Collinge et al.lopathy (TSE) in 1988 (http://www.bseinquiry.gov.uk). 1996b). Although Neibergs et al. (1994) suggested anTSEs also include scrapie in sheep and Creutzfeldt-Jakob increased BSE incidence in families of cattle with adisease (CJD) and Kuru in humans. According to the particular genotype within the octapeptide repeat re-aberrant protein agent hypothesis, a disease-specific iso- gion of the PrP gene, to date, no convincing associationform of the prion protein PrPSc interacts with normal between polymorphisms at the PrP locus and incidencehost PrP, resulting in its conversion to PrPSc. PrPSc is of BSE disease has been demonstrated for cattle, whichresistant to digestion with protease and can be detected may be due partly to the relatively limited variability ofin peripheral organs following infection; it later accu- the bovine PrP gene (Hills et al. 2003). Studies of themulates in the brain as the disease progresses. Clinical progeny of cows that developed BSE have suggestedonset of disease is associated with neuronal cell death that there may be some elements of genetic control ofresulting in the spongiform appearance of the brain susceptibility in cattle as there is an increased risk ofand results in “nervousness,” kicking, abnormal gait, BSE in offspring of BSE-affected cows compared withand pelvic limb ataxia in affected animals (Patterson unaffected controls (Donnelly et al. 1997; Fergusonand Painter 1999). BSE can be experimentally trans- et al. 1997; Wilesmith et al. 1997). However, it has beenmitted across species (Bruce et al. 1994) and a distinct difficult to distinguish between increased susceptibilityvariant of CJD (vCJD) in humans has been linked to and maternal transmission of disease. Intriguingly, re-BSE (Collinge et al. 1996a; Almond and Pattison cent studies of TSE transmission in mice showed that1997). genetic loci other than the PrP gene are likely to be

A major feature of prion diseases is that, although involved in the genetic control of prion diseases (Baronthey are not genetic, susceptibility of individuals to TSE

2002; Moreno et al. 2003). In an earlier study on cattle,diseases is influenced by their genetic makeup and in

we identified three marker loci on bovine chromosomesparticular, their PrP genotype. Three codons in the5, 10, and 20 that were significantly associated with BSEsheep PrP gene, at positions 136, 154, and 171, haveand are not linked to the PrP locus (Hernandez-San-been strongly associated with different incubation peri-chez et al. 2002).

A prion disease can be considered as a continuoustrait if, for example, incubation period is measured or1Corresponding author: Roslin Institute (Edinburgh), Roslin, Midlo-

thian EH25 9PS, United Kingdom. E-mail: [email protected] as a binary trait, if we consider individuals as being either

Genetics 167: 1863–1872 ( August 2004)

1864 C. Zhang et al.

TABLE 1affected or unaffected. A binary trait can be assumedto be controlled by an underlying continuous variable Numbers of BSE-affected and BSE-unaffected(Falconer and Mackay 1996) where the values of the offspring per sire in the studyvariable above a certain threshold level are designatedas affected (1) and those below it are designated as Sire Case Control Totalunaffected (0). Under this assumption, binary data can

1 83 30 113be analyzed for QTL with the same methods used for 2 70 22 93quantitative traits, although power and accuracy may be 3 72 18 90reduced (Xu and Atchley 1996; Kadarmideen et al. 4 43 22 652000). Visscher et al. (1996a) studied QTL for binary

Total 268 92 360traits in backcross and F2 populations using linear re-gression (LRG) and a generalized linear model (GLM)and concluded that both LRG and GLM gave similarresults when the errors within each genotype class were

by histopathology. No other information on their dis-normally distributed. Kadarmideen et al. (2000) com-ease progression was available. Samples were collectedpared GLM with LRG to map QTL for binary traits inby the U.K. Veterinary Field Service at the time of slaugh-a multi-family half-sib design and also reported that theyter of BSE suspects. Their sires were commonly used ashad similar power to detect the QTL and gave similarartificial insemination donors and were not known toestimates of QTL location, effects, and variances.have an increased frequency of BSE-affected offspring.In cattle, QTL mapping is commonly performed inControl (unaffected) animals and BSE cases were sam-a collection of paternal half-sib families that are pro-pled at approximately the same time from the sameduced by artificial insemination. Neimann-Sorensenfarms and were sire matched and age matched; hence,and Robertson (1961) and Weller et al. (1990) pro-we make the assumption that both sets of animals wereposed, respectively, the daughter and granddaughterexposed to the same level of infectious agent. However,designs to analyze linkage between a single marker andthere is no way to confirm that they were exposed toa QTL when data are structured in half-sib families. Thethe same degree as the affected animals. None of thedrawback of these methods is that they use informationcontrols subsequently appeared in the BSE case data-from a single marker at a time. Interval mapping, whichbase of the British Ministry of Agriculture. Microsatelliteuses information from multiple markers simultaneously,markers were chosen from two published bovine linkagehas been used by Georges et al. (1995) and Knott etmaps, the International Bovine Reference Panel (IBRP)al. (1996) for analyzing data from half-sib families. demap (http://www.cgd.csiro.au) and the U.S. Meat AnimalKoning et al. (2001) developed a cofactor analysis strat-Research Center map (http://www.marc.usda.gov), toegy for simultaneous analysis of multiple chromosomescover all autosomes (Bos taurus chromosomes BTA1–29)to increase the power and the precision of QTL map-and the pseudoautosomal region of the sex chromo-ping in outbred populations, structured in half-sib fami-somes (BTAX/Yps). A panel of 173 microsatellite mark-lies.ers was genotyped on all samples using an ABI373 DNAIn this study, a genome-wide scan for QTL associatedsequencer. The genotypes of the sires were inferredwith BSE resistance or susceptibility was performed byfrom those of their daughters while the genotypes ofapplying interval mapping and cofactor analysis meth-dams were unknown. For more details about the samplesods in paternal half-sib Holstein cattle families. Theand genotyping see Hernandez-Sanchez et al. (2002).goodness of fit of a two-QTL model was also tested to

Markers and map construction: Marker linkage mapssearch for multiple QTL within the same linkage group.were constructed using Cri-map software (Green et al.The results were compared with previous findings ob-1990) after setting all identifiable genotyping errors totained using transmission-disequilibrium tests (TDT;unknown genotypes. Various Cri-map options were usedHernandez-Sanchez et al. 2002).to determine linkage groups, marker orders, and markerdistances within linkage groups. If the map distance of

MATERIALS AND METHODS two neighboring markers was �100 cM, that linkage groupwas divided into two unlinked groups. Thirty-three linkageSamples and genotyping: Four half-sib Holstein siregroups, covering 2835 cM of the bovine genome acrossfamilies with 360 cows, of which 268 were BSE affectedall autosomes and the pseudoautosomal region of the sexand 92 BSE unaffected, were analyzed (Table 1). Thesechromosomes, were produced. The complete set of mapsfamilies were selected from the same sample of animalscan be viewed as supplementary material (supplement 1as used in the previous study (Hernandez-Sanchez et al.at http://www.genetics.org/supplemental/).2002). Study animals contracted BSE from the environ-

QTL analysis: Linear regression methods were usedment during the United Kingdom’s epidemic of the latefor QTL analysis of the binary data (Visscher et al. 1996a;1980s and 1990s. They were diagnosed by veterinarians

using standard clinical criteria and BSE was confirmed Kadarmideen et al. 2000). Initial QTL mapping was per-

1865QTL Affecting BSE

formed by using the web-based software package QTL linkage groups where significant evidence for a singleQTL was detected by the cofactor analysis. A 1-cM gridExpress (Seaton et al. 2002), which implements thesearch was performed in QTL Express to estimate themulti-marker linear regression method (Knott et al.effects of two QTL at separate positions within the same1996). In short, a conditional probability of inheritinglinkage group simultaneously, examining all possiblea particular haplotype from the sire was inferred frompairs of locations, and to test whether the two-QTL modelthe marker genotypes in all half-sib offspring. Thenexplained significantly more variation than the bestthe phenotypic value (i.e., affected or unaffected) wasQTL from the single-QTL analysis. Cofactors from otherregressed on the probability that a particular QTL allelelinkage groups identified in the cofactor analysis were(associated with susceptibility or resistance) was inher-also fitted. The genetic model wasited from the sire. The genetic model was Yij � �i �

biXij � eij , where Yij is the phenotype of animal j, offspringYij � �i � bilXijl � bimXijm � �

n

k�1

bikXijk � eij ,of sire i ; �i is the mean of sire family i ; bi is the allelesubstitution effect of the QTL within family i ; Xij is the

where the variables are as described above and l and mprobability that animal j inherited the (arbitrarily as-refer to the two putative QTL positions on the chromo-signed) first haplotype of sire i ; and eij is the residualsome. The F - statistics were calculated by comparingeffect. For every linkage group, F - statistics were calcu-mean squares from the two-QTL model with that of thelated by comparing the pooled mean squares obtainedbest one-QTL model with degrees of freedom equal tofrom regression within families to the residual meanS (numerator) and N � S(3 � C) (denominator), wheresquare. Numerator degrees of freedom was S and de-S is the number of sires, C is the number of cofactors,nominator degrees of freedom was N � 2S, where S isand N is the total number of offspring.the number of sires and N is the total number of prog-

Significance thresholds and confidence intervals:eny. This process was repeated at 1-cM intervals alongSearching for QTL across an entire genome involves aeach linkage group with the maximum F - value indicat-large number of statistical tests for marker-trait associa-ing the most likely position of a QTL.tions. The use of an appropriate significance thresholdA multiple (cofactor) analysis can partly account forfor each test is necessary to keep the number of falsethe variance generated by other segregating QTL and sub-positives to an acceptable level (Lynch and Walshstantially increases both power to detect a QTL and pre-1998). Significance thresholds were determined empiri-cision of estimating the QTL position (Lynch andcally by permutation tests to account for missing geno-Walsh 1998). The strategy adopted was the simultane-types and differences in marker density (Churchillous analysis of multiple chromosomes in an outbredand Doerge 1994).half-sib design (de Koning et al. 2001) where all posi-

Chromosome-wide significance levels (Pchromosome) weretions within a linkage group could be included in theobtained for each linkage group using 10,000 permuta-analysis of any family. Every round of analysis in this pro-tions. Two genome-wide thresholds were set, one forcedure involved two steps. Chromosome positions show-“suggestive” and one for “significant” QTL. For genome-ing significant associations with the trait from the pre-wide suggestive linkage, one false positive is expectedliminary analysis were first selected as cofactors. Then(Lander and Kruglyak 1995). The threshold for thisall linkage groups were analyzed by standard intervallevel for a specific linkage group corresponds to that

mapping including these cofactors as covariates, usingfor a P-value equal to the contribution (R) of that link-

the following genetic model, age group to the total genome length, which was ob-tained by dividing the length of a linkage group by the

Yij � �i � biXij � �n

k�1

bikXijk � eij , total length of the genome. The second threshold, forgenome-wide significant linkage, was set to that corre-

where the variables are as given above, and the sum- sponding to 5% significance (i.e., one false positive ex-mation term reflects the contribution of cofactors, QTL pected in 20 genome scans; Lander and Kruglyakidentified on other chromosomes. Cofactors on the chro- 1995), and it was obtained from the chromosome-widemosome under analysis were omitted (n is the number significance levels using the following Bonferroni cor-of linkage groups minus 1). QTL exceeding 5% chromo- rection (de Koning et al. 1999):some-wide threshold levels were selected as cofactors in

Pgenome � 1 � (1 � Pchromosome)1/R.the next round and this analysis was repeated until nonew QTL were revealed. Cofactors were dropped from Confidence intervals (C.I.) for the location of possiblethe analysis if the corresponding significance level was QTL were constructed by bootstrapping the sampleslower than the threshold calculated at every round for 1000 times (Visscher et al. 1996b). Taking the top andeach linkage group. bottom 2.5% of resampled positional estimates gave the

To distinguish between the presence of one QTL with estimated 95% confidence interval for each QTL.a large effect and two linked QTL with smaller effects, In the two-QTL model, an F - test of two QTL vs. one

QTL was performed for linkage groups harboring sug-a two-dimensional QTL search was carried out for those

1866 C. Zhang et al.

TABLE 2

Distribution of markers, marker intervals, and information content across chromosomes

Length Average marker Contribution InformationBTA Markers (cM) interval (cM) (R) content

1 11 185 16.8 0.065 0.522 6 134 22.3 0.047 0.653 7 147 21.0 0.052 0.584 5 121 24.2 0.043 0.505 12 182 15.2 0.064 0.626 6 126 21.0 0.044 0.637A 3 30 10.0 0.011 0.497B 3 11 3.7 0.004 0.618 8 113 14.1 0.040 0.619 6 102 17.0 0.036 0.6310 8 157 19.6 0.055 0.4811A 2 11 5.5 0.004 0.2811B 2 2 1.0 0.001 0.5712 6 121 20.2 0.043 0.5213 7 151 21.6 0.053 0.4814 5 81 16.2 0.029 0.5715 4 65 16.3 0.023 0.5016 5 66 13.2 0.023 0.6717 10 164 16.4 0.058 0.5118 6 86 14.3 0.030 0.6219 6 104 17.3 0.037 0.5720 6 96 16.0 0.034 0.5721B 4 73 18.3 0.026 0.4522 4 81 20.3 0.029 0.5523 5 77 15.4 0.027 0.5024 4 67 16.8 0.024 0.4725 3 52 17.3 0.018 0.5326 6 84 14.0 0.030 0.5727 3 19 6.3 0.007 0.5428 2 10 5.0 0.004 0.6229 3 29 9.7 0.010 0.68X/Yps 4 88 22.0 0.031 0.36

Total 172 2835 16.5 (average) 1.000 0.55 (average)

Contribution was calculated by dividing the length of a linkage group by the total length. Informationcontent measured as the ratio of the variance of QTL conditional probability to the maximum variance wasaveraged over every centimorgan along a linkage group. Chromosomes 7, 11, and 21 were each divided intotwo linkage groups, 7A/7B, 11A/11B, and 21A/21B (21A contains only one marker and is not shown), becausethe map distance between two adjacent markers on each chromosome was �100 cM.

gestive QTL detected by the single-QTL model. The probabilities along each linkage group at 1-cM intervalswere collected for every offspring in a family. The meansignificance of the test statistics was determined from aprobability (Pmean) and standard error (SE) for each fam-standard F - table with degrees of freedom as given above.ily were calculated and t - statistics were obtained fromSegregation distortion: The TDT explores segrega-the formula:tion distortion within the affected individuals from a

heterozygous sire, significant distortion being seen asT �

|Pmean � 0.5|SE

.evidence for the presence of a disease-related gene.To compare our results to those from a previous study(Hernandez-Sanchez et al. 2002), we tested for segre- Thresholds for genome-wide significant segregationgation distortion within affected and unaffected individ- distortion were set at 5%, adjusted using the Bonferroniuals for all linkage groups. correction for multiple testing (number of tests � 688 �

Segregation distortion was studied by testing whether 172 markers � 4 families). This corresponds to a tabu-the average probability that a particular QTL allele in- lated P - value of 0.000071. Degrees of freedom were de-

termined by the number of animals in the smallest fam-herited from the sire deviated from 0.5. The inheritance

1867QTL Affecting BSE

TABLE 3

The result of cofactor analysis by fitting five cofactors

Effect (standard error) by familyPosition 95% C.I.

BTA (cM) (cM) F Pchromosome 1 2 3 4

1 125 0–171 3.9 0.0308 �0.18 (0.100) 0.26 (0.096) �0.08 (0.153) �0.43 (0.181)6 60 11–110 4.5 0.0121 0.23 (0.083) �0.14 (0.082) 0.08 (0.084) �0.26 (0.112)13 55 0–106 3.4 0.0447 �0.26 (0.135) �0.27 (0.144) 0.06 (0.130) �0.31 (0.120)17 144 91–164 6.2 0.0015 �0.01 (0.100) 0.20 (0.093) 0.12 (0.110) �0.84 (0.186)19 97 0–104 4.9 0.0057 �0.12 (0.094) �0.04 (0.087) 0.33 (0.091) 0.30 (0.117)X/Yps 58 50–69 6.1 0.0004 0.07 (0.237) 0.67 (0.144) 0.41 (0.145) �0.01 (0.392)

Two QTL, BTA17 and BTAX/Yps , are 5% genome-wide significant. Others are 5% chromosome-wide significant. Significancethreshold were determined by 10,000 permutations.

ily (43 for affecteds, 18 for unaffecteds), making this a 5% genome-wide significance threshold (Pgenome �0.01).The estimated QTL position for the X/Yps linkage groupconservative test.is at 55 cM with a 95% C.I. between 52 and 65 cM. TheF-ratios for all linkage groups can be viewed as the “no

RESULTS cofactor” points in Figure 2.Cofactor analysis was then applied, where the cofac-Genetic map: A total of 173 microsatellite markers

tors were the QTL exceeding the 5% chromosome-widewere used to construct a 2835 cM (Kosambi) geneticsignificance threshold found in the previous round. Thelinkage map of the bovine genome that included allQTL analysis was repeated including the cofactors untilautosomes and the pseudoautosomal region of the sexno new QTL were revealed. By the final round, fivechromosomes (Table 2). In total, 32 linkage groupscofactors in total were fixed and six QTL were obtainedwere analyzed, excluding one, that contained a single(one of the QTL did not feature as a cofactor). Themarker, because interval mapping would not be verynumbers of cofactors fitted in successive rounds werepowerful for such a case. The average marker intervalzero, seven, three, seven, four, six, and five. Two linkageper linkage group ranged from 1 to 24.2 cM with a ge-groups (BTA17 and X/Yps) showed genome-wide sig-nome-wide mean of 16.5 cM. Most of the linkage groupsnificant evidence and four linkage groups (BTA1, 6, 13,were moderately informative with average informationand 19) showed genome-wide suggestive evidence forcontent (Knott et al. 1998) ranging from 0.28 to 0.68.a single QTL (Table 3 and Figure 1). The significantThe average information content at genome level wasQTL on BTAX/Yps was mapped to 58 cM (Pgenome �0.01),0.55. The mapping results confirmed that the X/Ypsnear marker TGLA325, with 95% C.I. between 50–69markers underwent recombination in the sire and there-cM. Families 2 and 3 accounted for most of this effect.fore were located in the pseudoautosomal region of theThe genome-wide significant QTL on BTA17 was atsex chromosomes.marker INRA25, located at 144 cM (Pgenome �0.02), withTen linkage groups were not in agreement with one91–164 cM 95% C.I. Families 2 and 4 explained mostor both of the two published maps. The majority ofof the effect of this QTL.the differences were minor flips between closely linked

To calculate the proportion of variance explained bymarkers. BTA13 was the only chromosome for whichthe joint QTL, we estimated for each of the six QTLthere was a major difference in the orientation of mark-the ratio between the sum of squares explained by theers; DIK93 was found at the centromeric end of themultiple-QTL model and the total sum of squares. Thechromosome in our study and at the telomeric end ofmodels including QTL plus four cofactors (BTA1, 6, 17,the IBRP map (most of the BTA13 markers used in this19, and X) gave an estimate of 0.23 for the proportion ofstudy were not found on the MARC map). Chance mayvariance explained, while the model including QTL plusexplain the minor differences between maps, but it isfive cofactors (BTA13) gave a slightly higher estimatenot clear why the BTA13 maps were so different. As the(0.25).average number of informative meioses per marker was

Compared with the six QTL obtained without cofac-167 in this study and all maps were well supported bytors, one QTL on BTA22 was removed and one newthe Cri-map “FLIPS” option, we used this map ratherQTL on BTA1 was added. Among the remaining fivethan published map positions.QTL, the estimated positions of three QTL, on BTA6,QTL analysis: Six genome-wide suggestive QTL19, and X/Yps, moved 2–5 cM after the inclusion of(Pchromosome � R) were identified using single-QTL map-cofactors, and the F-statistics for individual QTL wereping (BTA6, 13, 17, 19, 22, and X/Yps). The most striking

was on the BTAX/Yps region, which also exceeded the either decreased or increased, with changes ranging

1868 C. Zhang et al.

Figure 1.—Profile of the F-statisticsfor four genome-wide suggestive QTL(BTA1, 6, 13, 19) and two genome-widesignificant QTL (BTA17 and X/Yps)identified by cofactor analysis. Hori-zontal dashed line indicates genome-wide suggestive threshold and hori-zontal dashed-and-dotted line indicates5% genome-wide significance threshold.Vertical dotted lines mark 95% confi-dence intervals. Where no dotted line isshown, the confidence limit is the endof the linkage group. Markers in paren-theses are at the same position as theirleft adjacent marker.

from �0.1 (BTAX/Yps) to �2.0 (BTA17). Figure 2 shows within affected individuals was found for one or morefamilies on the following chromosomes: BTA1, 5, 10,the development of the maximum F-statistic for each

individual linkage group during the cofactor analysis 13, 14, 17, 19, 21B, 26, and X/Yps. These include fiveof the six chromosomes with significant QTL (excludingprocess.

Comparison of two-QTL vs. one-QTL models pro- BTA6), two of the three chromosomes with significantTDT results (BTA5 and 10; Hernandez-Sanchez et al.vided no support for additional QTL on any linkage

group where genome-wide significant and suggestive 2002), and three others. Significant segregation distor-tion within unaffected individuals was found on BTA1,QTL were found using cofactor analysis.

Test for segregation distortion on chromosomes with 10, 13, 14, 17, 19, 21B, and X/Yps for one or morefamilies. These include the same five chromosomes withsignificant QTL: The TDT analysis of Hernandez-San-

chez et al. (2002) found associations between BSE infec- significant QTL as seen for the affected individuals,one of the three chromosomes detected using TDTtion and loci on chromosomes 5, 10, and 20 (see below).

In the current study, significant segregation distortion (BTA10), and two others (also seen for the affected

1869QTL Affecting BSE

DISCUSSION

In this study, a genome-wide linkage analysis of QTLinvolved in susceptibility or resistance to BSE was con-ducted in a population of cattle consisting of four half-sib families. The QTL analysis, in which five cofactorswere considered, identified four genome-wide “sugges-tive” and two genome-wide significant QTL. There wasevidence for significant segregation distortion of sirealleles on five of the six chromosomes with QTL, but alsoon other chromosomes. The most consistent picturebetween the QTL and segregation distortion analyseswas seen on BTA19 and X/Yps, where segregation distor-

Figure 2.—Development of the F-statistics for 32 linkage tion was in opposite directions in the affected and unaf-groups during cofactor analysis of the first round (no cofac- fected offspring of a family with a significant QTL. Thetor), the second round (seven cofactors), the sixth round (six

positions of the QTL and the maximum level of segrega-cofactors), and the final seventh round (five cofactors). Thetion distortion overlapped for BTAX/Yps.different symbols indicate the maximum F-statistic for the

linkage group during the process of fitting cofactors and re- Candidate genes: Although variation in the PrP geneanalyzing the data. has been associated with differences in susceptibility or

incubation period of TSEs in other species, up to nowpolymorphisms in the bovine PrP gene have not beenassociated with variation in susceptibility to BSE. One

individuals). The direction of the segregation distortion suggestive QTL in this study was identified on BTA13,(i.e., which allele is overrepresented) was different for which harbors the prion gene (PRNP). However, thethe affected and unaffected offspring in a few cases. 95% C.I. for the QTL identified on BTA13 does notThis pattern was seen for family 3 on BTA1, 19, and X, include the region where PRNP has been mapped (be-but not for the six other family/chromosome combina- tween markers HUJ616 and ABS10; Schlapfer et al.tions where there was significant segregation distortion 2000); the distance between that QTL and PRNP is atfor both affected and unaffected offspring. least 74 cM. Because of inconsistencies between our

For the chromosomes where linkage analysis identi- map order and that of the IBRP map for BTA13, linkagefied QTL, the maximum t-statistics for segregation dis- analysis was also run using the subset of markers foundtortion in affected offspring within individual families on the IBRP map and the IBRP positions for this chro-and their positions are shown in Table 4. When the mosome (results not shown). There was still a suggestiveposition of the maximum t-statistic is compared with QTL and its position was again distant from PRNP.that of the maximum F-statistic on the same linkage In mice, polymorphisms in PRNP are known to begroup, there is substantial overlap in some cases but associated with variation in incidence or onset of dis-not others. The greatest overlap in t-statistic and F-ratio ease. It is therefore interesting that in a QTL study inpositions is for the pseudoautosomal X chromosome a mouse model of BSE infection, Manolakou et al.where the maximum statistics were physically close for (2001) also failed to detect any effect closely linked to

PRNP, although a QTL was found on mouse chromo-both families 2 and 3.

TABLE 4

The maximum t-statistics for segregation distortion within affected offspring forlinkage groups where putative QTL were detected

Chromosomal position of maximum t - statistic andMaximum t - statistic maximum F - ratio (cM)

Family 1 Family 2 Family 3 Family 4 Family 1 Family 2 Family 3 Family 4

BTA1 3.01 1.87 7.38 3.71 122 (0)BTA6 2.96 1.02 2.07 2.45BTA13 5.15 1.92 2.02 3.87 54–57 (52)BTA17 6.01 8.61 8.28 3.12 144 (164) 105–106 (144) 112–113 (7)BTA19 13.44 5.03 4.93 1.38 6 (3) 0–23 (105) 0 (100)BTAX 2.25 20.1 5.08 1.92 54 (58) 44–88 (68)

Values in italic were above the 5% genome-wide threshold. For cases of significant segregation distortion, positions of maximumt-statistics are shown with the maximum F-ratio position in parentheses.

1870 C. Zhang et al.

some 2 where PRNP is located. This could suggest that sire) or because their inclusion biased the estimation ofsegregation ratios in the absence of maternal genotypesin both cattle and mice there is another locus on the(i.e., daughters with homozygous genotypes). Further-same chromosome as PRNP that is involved in BSE re-more, sires homozygous for a marker were also excludedsistance or susceptibility or, alternatively, that in bothfrom the analysis. Using a very stringent significancestudies, shortcomings in the analyses misplace the QTLthreshold, Hernandez-Sanchez et al. (2002) foundregions.three significant markers associated with BSE on chro-Four previous studies that mapped QTL related tomosomes BTA5, 10, and 20. Surprisingly, none of thesethe incubation period of TSE diseases in mouse modelsthree regions are in common with genome-wide sugges-revealed significant or suggestive evidence for QTL ontive or significant QTL found in this study, althoughmouse chromosomes 2, 4–12, 15, and 17–19 (Stephen-significant segregation distortion within affected indi-son et al. 2000; Lloyd et al. 2001; Manolakou et al.viduals was found on two of these chromosomes. When2001; Moreno et al. 2003). Chromosomes 2 and 11markers closely linked to the three significant associa-showed significant evidence of a QTL in two of thetions were tested, only one, on BTA5, showed a signifi-studies. Other chromosomal regions were supported forcant association with disease status under TDT. Noneassociation with TSE susceptibility either by significantof the markers identified using TDT showed the reverseevidence in only one study or by suggestive evidence inpattern of allele transmission in the unaffected animals.more than one study. Regions of the bovine genome

A number of factors could contribute to differences inshowing conservation of synteny with the mouse chro-the results between linkage and TDT analyses. Althoughmosomes were deduced via Mouse Genome Informaticsthe initial data set was the same, the samples used in(http://www.informatics.jax.org/). Several of the chro-the two analyses differed. The TDT analyzes individualmosomes containing regions where QTL were locatedloci separately and requires the sire to be heterozygousin this study (BTA1, 6, 13, 17, 19, and X/Yps) showedfor the marker locus tested. In addition, the absencehomology with mouse chromosomes with putative QTL.of maternal genotypes meant that the paternal alleleDue to the lack of resolution of the bovine map, how-inherited for a particular locus could not be determinedever, it was difficult to align regions where the cattlefor many of the progeny from heterozygous sires andQTL were located relative to the mouse map. A putativeso these also could not be used in the TDT. As a conse-candidate gene, Nf1, falls within the 95% C.I. of a ge-quence, the data set used by Hernandez-Sanchez et al.nome-wide suggestive QTL on BTA19 in the current(2002) for any particular locus included approximatelystudy and also falls within a QTL region on mouse chro-half the data that was used in the linkage analysis. Thismosome 11 described in two independent studies (Ste-would have reduced the statistical power of the TDT

phenson et al. 2000; Lloyd et al. 2001). Neurologicalstudy (Cardon and Bell 2001). To test whether differ-

and behavioral defects have been described in Nf1 ho- ences in results arose from different data sets used inmozygous mutant mice (Costa et al. 2001), suggesting each study, the same restrictions on genotypic data usedthat NF1 could play a role in the neurodegenerative in the TDT analysis were applied to BTA5, 10, and 20,process of BSE. where significant associations were previously identified

Although none of the above-mentioned mouse stud- by TDT (however, in this case, both affected and unaf-ies identified QTL on the sex chromosomes, a sex effect fected individuals were included in the analysis). Theseon incubation period and survival time was noted by data were then analyzed using linkage methods. ResultsManolakou et al. (2001) and Moreno et al. (2003), using the restricted data set still failed to find significantrespectively. Manolakou et al. (2001) were able to dis- effects above the 5% chromosome-wide threshold al-sect maternal effects from those due to the X chromo- though the maximum F-statistics on BTA5 and BTA10some in the F1 generation and concluded that some of moved closer to markers BM315 and INRA107, whichthe sex differences seen in that population could be showed the greatest association under the TDT (resultsexplained by X chromosome differences. The magni- not shown).tude of the BTAX/Yps effect in our study emphasizes Another possible explanation for the differences inthe need for further investigation into the genetic basis results between the two approaches could be that TDTof sex differences in TSE susceptibility. and linkage analysis via multiple regression are very

Comparison with previous study: Hernandez-San- different tests. The TDT examines distortion within af-chez et al. (2002) conducted a genome-wide search for fected individuals for each sire and locus and usesmarkers associated with BSE incidence using a subset counts of segregating alleles as raw data. Linkage analy-of the data used here. In that study, transmission proba- sis measures mean phenotypic differences between ge-bilities of the two sire alleles were assessed for the af- notype groups. Moreover, the power of TDT is affectedfected daughters at each marker. A number of daughter by the recombination rate and the linkage disequilib-genotypes had to be excluded from the analysis either rium (LD) in the population (composed of four fami-because the transmitted sire allele could not be identi- lies), whereas linkage analysis depends on the recombi-

nation rate and is not affected by population-level LD.fied (i.e., daughters with the same genotype as their

1871QTL Affecting BSE

It is known that TDT is more powerful than linkage if the potential advantages of using different methods toexplore the data and the need for independent verifica-there is strong LD in the sample and the marker map

is dense (Risch and Merikangas 1996). Therefore, tion of results.linkage may have had insufficient power to detect the We thank the Biotechnology and Biological Sciences Researchregions detected by TDT for the markers that were Council for additional funding, D. Matthews and J. Wilesmith for

assistance in obtaining samples, and I. Maclean and D. Pomp (Gene-physically close to the genes affecting the trait. Con-Seek) for technical assistance in genotyping. Data for this work wereversely, over larger genomic distances (and thus highercollected through funding from the U.K. Department for Environ-recombination rates), the linkage approach may bement, Food and Rural Affairs (DEFRA, project SE 1744) and the

more successful than TDT (Kolbehdari and Jansen European Commission (project CT97 3311).2003).

It is likely that population-wide distortion in markeralleles caused by factors other than BSE is present in

LITERATURE CITEDcattle populations that are subjected to intensive breed-Almond, G., and J. Pattison, 1997 Human BSE. Nature 389: 437–ing selection. This distortion could be detected by TDT,

438.giving misleading results. This is consistent with the fact Baron, T., 2002 Mouse models of prion disease transmission. Trendsthat TDT found no significant effects when comparing Mol. Med. 8: 495–500.

Bruce, M., A. Chree, I. McConnell, J. Foster, G. Pearson et al.,cases vs. controls by testing interactions in the transmis-1994 Transmission of bovine spongiform encephalopathy andsion of alleles (Hernandez-Sanchez et al. 2002) while scrapie to mice: strain variation and the species barrier. Philos.

one would have expected different alleles to be overrep- Trans. R. Soc. Lond. B Biol. Sci. 343: 405–411.Cardon, L. R., and J. I. Bell, 2001 Association study designs forresented in the affected and unaffected classes. A similar

complex diseases. Nat. Rev. Genet. 2: 91–99.pattern was seen in our survey of segregation distortion Churchill, G. A., and R. W. Doerge, 1994 Empirical thresholdacross the genome, although there were a few cases values for quantitative trait mapping. Genetics 138: 963–971.

Collinge, J., K. C. L. Sidle, J. Meads, J. Ironside and A. F. Hill,where the direction of the distortion differed for the1996a Molecular analysis of prion strain variation and the aetiol-affected and the unaffected classes. On the other hand, ogy of ‘new variant’ CJD. Nature 383: 685–690.

the linkage profiles using the same reduced data set Collinge, J., J. Beck, T. Campbell, K. Estibeiro and R. G. Will,1996b Prion protein gene analysis in new variant cases of Creutz-as used for TDT (as mentioned above) reached theirfeldt-Jakob disease. Lancet 348: 56.maximum at the positions where TDT found significant

Costa, R. M., T. Yang, D. P. Huynh, S. M. Pulst, D. H. Viskochil eteffects, although the results from the linkage analysis al., 2001 Learning deficits, but normal development and tumor

predisposition, in mice lacking exon 23a of Nf1. Nat. Genet. 27:did not exceed the threshold values. Hernandez-San-399–405.chez et al. (2002) noted that TDT could not be used

de Koning, D. J., L. L. Janss, A. P. Rattink, P. A. van Oers, B. J.to test the X/Yps chromosome because of the real segre- de Vries et al., 1999 Detection of quantitative trait loci for

backfat thickness and intramuscular fat content in pigs (Susgation distortion of this region caused by linkage to Xscrofa). Genetics 152: 1679–1690.combined with the fact that all samples were from fe-

de Koning, D. J., N. F. Schulman, K. Elo, S. Moisio, R. Kinos etmales (i.e., it is expected that alleles from the sire’s X al., 2001 Mapping of multiple quantitative trait loci by simplechromosome-linked pseudoautosomal segment will be regression in half-sib designs. J. Anim. Sci. 79: 616–622.

Donnelly, C. A., N. M. Ferguson, A. C. Ghani, J. W. Wilesmithoverrepresented in his daughters relative to alleles fromand R. M. Anderson, 1997 Analysis of dam-calf pairs of BSEhis Y chromosome-linked pseudoautosomal segment). cases: confirmation of a maternal risk enhancement. Proc. R.

Our preliminary exploration indicated that linkage Soc. Lond. B Biol. Sci. 264: 1647–1656.Falconer, D. S., and T. F. C. Mackay, 1996 Introduction to Quantita-analysis is robust to segregation distortion of this kind

tive Genetics, Ed. 4. Addison-Wesley Longman, Harlow, UK.(see supplement 2 at http://www.genetics.org/supple Ferguson, N. M., C. A. Donnelly, M. E. J. Woolhouse and R. M.mental/). Anderson, 1997 A genetic interpretation of heightened risk of

BSE in offspring of affected dams. Proc. R. Soc. Lond. B Biol.In summary, both analyses may have detected actualSci. 264: 1445–1455.effects and thus provide complementary evidence for Georges, M., D. Nielsen, M. Mackinnon, A. Mishra, R. Okimoto

loci involved in susceptibility or resistance. The fact that et al., 1995 Mapping quantitative trait loci controlling milk pro-duction in dairy cattle by exploiting progeny testing. GeneticsQTL and cofactors identified in the linkage analysis139: 907–920.explained only a quarter of the variation allows for the Green, P., K. Falls and S. Crooks, 1990 Cri-map, Version 2.4.

possibility that other regions, such as those previously Washington University School of Medicine, St. Louis.Hernandez-Sanchez, J., D. Waddington, P. Wiener, C. S. Haleyidentified by Hernandez-Sanchez et al. (2002), could

and J. L. Williams, 2002 Genome-wide search for markers asso-also be important. Additional work to explore the re-ciated with bovine spongiform encephalopathy. Mamm. Genome

gions identified by this and the previous study will be 13: 164–168.Hills, D., J. Schlaepfer, S. Comincini, I. Maclean, G. Dolf et al.,required to confirm these results.

2003 Sequence variation in the bovine and ovine PRNP genes.Conclusions: Six putative QTL, two genome-wide sig-Anim. Genet. 34: 183–190.

nificant and four genome-wide suggestive QTL, were Hunter, N., 1999 Molecular biology and genetics of bovine spongi-form encephalopathy, pp. 229–246 in The Genetics of Cattle, editedfound associated with BSE incidence using linkage analy-by R. Fries and A. Ruvinsky. CAB International, Wallingford,sis. Lack of correspondence with previous results obtainedUK.

using a TDT approach highlights differences between Kadarmideen, H., L. G. Janss and J. C. M. Dekkers, 2000 Powerof quantitative trait locus mapping for polygenic binary traitslinkage and association methodologies and emphasizes

1872 C. Zhang et al.

using generalized and regression interval mapping in multi-family mozygous prion protein genotype predisposes to sporadic Crentz-feldt-Jakob disease. Nature 352: 340–342.half-sib designs. Genet. Res. 76: 305–317.

Knott, S. A., J. M. Elsen and C. S. Haley, 1996 Methods for multi- Patterson, W. J., and M. J. Painter, 1999 Bovine spongiform en-cephalopathy and new variant Creutzfeldt-Jakob disease: an over-ple-marker mapping of quantitative trait loci in half-sib popula-

tions. Theor. Appl. Genet. 93: 71–80. view. Commun. Dis. Public Health 2: 5–13.Risch, N., and K. Merikangas, 1996 The future of genetic studiesKnott, S. A., L. Marklund, C. S. Haley, K. Andersson, W. Davies

et al., 1998 Multiple marker mapping of quantitative trait loci of complex human diseases. Science 273: 1516–1517.Schlapfer, J., N. Stahlberger-Saitbekova, J. Kuffer and G. Dolf,in a cross between outbred wild boar and Large White pigs.

Genetics 149: 1069–1080. 2000 Genetic mapping of the prion protein gene (PRNP) onbovine chromosome 13. J. Anim. Breed. Genet. 117: 211–216.Kolbehdari, D., and G. B. Jansen, 2003 Simulation of transmission

disequilibrium tests for QTL detection in outbred livestock popu- Seaton, G., C. S. Haley, S. A. Knott, M. Kearsey and P. M. Visscher,2002 QTL Express: user-friendly software to map quantitativelations. Book of Abstracts of the 54th Annual Meeting of the

European Association for Animal Production, 31 August–2 Sep- trait loci in outbred populations. Bioinformatics 18: 339–340.Stephenson, D. A., K. Chiotti, C. Ebeling, D. Groth, S. J. DeAr-tember, Rome.

Lander, E. S., and L. Kruglyak, 1995 Genetic dissection of complex mond et al., 2000 Quantitative trait loci affecting prion incuba-tion time in mice. Genomics 69: 47–53.traits: guidelines for interpreting and reporting linkage results.

Nat. Genet. 11: 241–247. Visscher, P. M., C. S. Haley and S. A. Knott, 1996a MappingQTLs for binary traits in backcross and F2 populations. Genet.Lloyd, S. E., O. N. Onwuazor, J. A. Beck, G. Mallinson, M. Farrall

et al., 2001 Identification of multiple quantitative trait loci Res. 68: 55–63.Visscher, P. M., R. Thompson and C. S. Haley, 1996b Confidencelinked to prion disease incubation period in mice. Proc. Natl.

Acad. Sci. USA 98: 6279–6283. intervals in QTL mapping by bootstrapping. Genetics 143: 1013–1020.Lynch, M., and B. Walsh, 1998 Genetics and Analysis of Quantitative

Traits. Sinauer Associates, Sunderland, MA. Weller, J. I., Y. Kashi and M. Soller, 1990 Power of daughter andgranddaughter designs for determining linkage between markerManolakou, K., J. Beaton, I. McConnell, C. Farquar, J. Manson

et al., 2001 Genetic and environmental factors modify bovine loci and quantitative trait loci in dairy cattle. J. Dairy Sci. 73:2525–2537.spongiform encephalopathy incubation period in mice. Proc.

Natl. Acad. Sci. USA 98: 7402–7407. Wells, G. A. H., A. C. Scott, C. T. Johnson, R. F. Gunning, R. D.Moreno, C. R., F. Lantier, I. Lantier, P. Sarradin and J.-M. Elsen, Hancock et al., 1987 A novel progressive spongiform encepha-

2003 Detection of new quantitative trait loci for susceptibility lopathy in cattle. Vet. Rec. 121: 419–420.to transmissible spongiform encephalopathies in mice. Genetics Wilesmith, J. W., G. A. H. Wells, J. Ryan, D. Gravier-Widen and M.165: 2085–2091. Simmons, 1997 A cohort study to examine maternally-associated

Neibergs, H. L., A. M. Ryan, J. E. Womack, R. L. Spooner and J. L. risk factors for bovine spongiform encephalopathy. Vet. Rec. 141:Williams, 1994 Polymorphism analysis of the prion gene in 239–243.BSE-affected and unaffected cattle. Anim. Genet. 25: 313–317. Xu, S., and W. R. Atchley, 1996 Mapping quantitative trait loci

Neimann-Sorensen, A., and A. Robertson, 1961 The association for complex binary diseases using line crosses. Genetics 143:between blood groups and several production characteristics in 1417–1424.three Danish cattle breeds. Acta Agric. Scand. 11: 163–196.

Palmer, M. S., A. J. Dryden, J. T. Hughes and J. Collinge, 1991 Ho- Communicating editor: J. B. Walsh