A systematic strategy for large-scale analysis of genotype phenotype correlations: identification of candidate genes involved in African trypanosomiasis

It is increasingly common to combine Microarray and Quantitative Trait Loci data to aid the search for candidate genes responsible for phenotypic variation. Workflows provide a means of systematically processing these large datasets and also represent a framework for the re-use and the explicit declaration of experimental methods. In this article, we highlight the issues facing the manual analysis of microarray and QTL data for the discovery of candidate genes underlying complex phenotypes. We show how automated approaches provide a systematic means to investigate genotype-phenotype correlations. This methodology was applied to a use case of resistance to African trypanosomiasis in the mouse. Pathways represented in the results identified Daxx as one of the candidate genes within the Tir1 QTL region. Subsequent re-sequencing in Daxx identified a deletion of an amino acid, identified in susceptible mouse strains, in the Daxx-p53 protein-binding region. This supports recent experimental evidence that apoptosis could be playing a role in the trypanosomiasis resistance phenotype. Workflows developed in this investigation, including a guide to loading and executing them with example data, are available at http://workflows.mygrid.org.uk/repository/myGrid/PaulFisher/.     

The use of mouse models to unravel genetic architecture of physical activity: a review

The discovery of genetic variants that underlie a complex phenotype is challenging. One possible approach to facilitate this endeavor is to identify quantitative trait loci (QTL) that contribute to the phenotype and consequently unravel the candidate genes within these loci. Each proposed candidate locus contains multiple genes and, therefore, further analysis is required to choose plausible candidate genes. One of such methods is to use comparative genomics in order to narrow down the QTL to a region containing only a few genes. We illustrate this strategy by applying it to genetic findings regarding physical activity (PA) in mice and human. Here, we show that PA is a complex phenotype with a strong biological basis and complex genetic architecture. Furthermore, we provide considerations for the translatability of this phenotype between species. Finally, we review studies which point to candidate genetic regions for PA in humans (genetic association and linkage studies) or use mouse models of PA (QTL studies) and we identify candidate genetic regions that overlap between species. On the basis of a large variety of studies in mice and human, statistical analysis reveals that the number of overlapping regions is not higher than expected on a chance level. We conclude that the discovery of new candidate genes for complex phenotypes, such as PA levels, is hampered by various factors, including genetic background differences, phenotype definition and a wide variety of methodological differences between studies .     

Candidate genes required for embryonic development: a comparative analysis of distal mouse chromosome 14 and human chromosome 13q22

Mice homozygous for the Ednrb(s-1Acrg) deletion arrest at embryonic day 8.5 from defects associated with mesoderm development. To determine the molecular basis of this phenotype, we initiated a positional cloning of the Acrg minimal region. This region was predicted to be gene-poor by several criteria. From comparative analysis with the syntenic human locus at 13q22 and gene prediction program analysis, we found a single cluster of four genes within the 1.4-to 2-Mb contig over the Acrg minimal region that is flanked by a gene desert. We also found 130 highly conserved nonexonic sequences that were distributed over the gene cluster and desert. The four genes encode the TBC (Tre-2, BUB2, CDC16) domain-containing protein KIAA0603, the ubiquitin carboxy-terminal hydrolase L3 (UCHL3), the F-box/PDZ/LIM domain protein LMO7,and a novel gene. On the basis of their expression profile during development, all four genes are candidates for the Ednrb(s-1Acrg) embryonic lethality. Because we determined that a mutant of Uchl3 was viable, three candidate genes remain within the region.     

New quantitative trait loci that regulate wound healing in an intercross progeny from DBA/1J and 129×1/SvJ inbred strains of mice

Wound healing/regeneration mouse models are few, and studies performed have mainly utilized crosses between MRL/MPJ (a good healer) and SJL/J (a poor healer) or MRL/lpr (a good healer) and C57BL/6J (a poor healer). Wound healing is a complex trait with many genes involved in the expression of the phenotype. Based on data from previous studies that common and additional quantitative trait loci (QTL) were identified using different crosses of inbred strains of mice for various complex traits, we hypothesized that a new cross would identify common and additional QTL, unique modes of inheritance, and interacting loci, which are responsible for variation in susceptibility to fast wound healing. In this study, we crossed DBA/1J (DBA, a good healer) and 129/SvJ (129, a poor healer) and performed a genome-wide scan using 492 (DBA×129) F2 mice and 98 markers to identify QTL that regulate wound healing/regeneration. Four QTL on chromosomes 1, 4, 12, and 18 were identified which contributed toward wound healing in F2 mice and accounted for 17.1% of the phenotypic variation in ear punch healing. Surprisingly, locus interactions contributed to 55.7% of the phenotype variation in ear punch healing. In conclusion, we have identified novel QTL and shown that minor interacting loci contribute significantly to wound healing in DBA×129 mice cross. The authors Masinde, Li, and Nguyen contributed equally to this article.     

The Tribolium castaneum ortholog of Sex combs reduced controls dorsal ridge development

Several constitutional chromosomal rearrangements occur on human chromosome 17. Patients who carry constitutional deletions of 17q21.3-q24 exhibit distinct phenotypic features. Within the deletion interval, there is a genomic segment that is bounded by the myeloperoxidase and homeobox B1 genes. This genomic segment is syntenically conserved on mouse chromosome 11 and is bounded by the mouse homologs of the same genes (Mpo and HoxB1). To attain functional information about this syntenic segment in mice, we have generated a 6.9-Mb deletion [Df(11)18], the reciprocal duplication [Dp(11)18] between Mpo and Chad (the chondroadherin gene), and a 1.8-Mb deletion between Chad and HoxB1. Phenotypic analyses of the mutant mouse lines showed that the Dp(11)18/Dp(11)18 genotype was responsible for embryonic or adolescent lethality, whereas the Df(11)18/+ genotype was responsible for heart defects. The cardiovascular phenotype of the Df(11)18/+ fetuses was similar to those of patients who carried the deletions of 17q21.3-q24. Since heart defects were not detectable in Df(11)18/Dp(11)18 mice, the haplo-insufficiency of one or more genes located between Mpo and Chad may be responsible for the abnormal cardiovascular phenotype. Therefore, we have identified a new dosage-sensitive genomic region that may be critical for normal heart development in both mice and humans.     

Genetic linkage analysis of X-ray hypersensitivity in the LEC mutant rat

The LEC rat has been reported to exhibit X-ray hypersensitivity and deficiency in DNA double-strand break (DSB) repair. The present study was performed to map the locus responsible for this phenotype, the xhs (X-ray hypersensitivity), as the first step in identifying the responsible gene. Analysis of the progeny of (BN × LEC)F₁× LEC backcrosses indicated that the X-ray hypersensitive phenotype was controlled by multiple genetic loci in contrast to the results reported previously. Quantitative trait loci (QTL) linkage analysis revealed two responsible loci located on Chromosomes (Chr) 4 and 1. QTL on Chr 4 exhibited very strong linkage to the X-ray hypersensitive phenotype, while QTL on Chr 1 showed weak linkage. The Rad52 locus, mutation of which results in hypersensitivity to ionizing radiation and impairment of DNA DSB repair in yeast, was reported to be located on the synteneic regions of mouse Chr 6 and human Chr 12. However, mapping of the rat Rad52 locus indicated that it was located 23 cM distal to the QTL on Chr 4. Furthermore, none of the radio-sensitivity-related loci mapped previously in the rat chromosome were identical to the QTL on Chrs 4 and 1 in the LEC rat. Thus, it seems that X-ray hypersensitivity in the LEC rat is caused by mutation(s) in as-yet-undefined genes. Received: 14 February 2000 / Accepted: 17 May 2000     

Construction of a BAC Contig for a 3 cM Biologically Significant Region of Mouse Chromosome 1

One QTL and genes and phenotypes have been localized in the region between 92 cM and 95cM of mouse chromosome 1. The QTL locus contributes to approximately 40% of the variation of the peak bone density between C57BL/6J (B6) and CAST/EiJ (CAST) strains. Other loci located in this chromosomal region include a neural tube defect mutant loop-tail (Lp), a lymphocyte-stimulating determinant (Lsd), and the Transgelin 2 (Tagln 2). The human chromosome region homologous to this region is 1q21-23, which also contains a QTL locus for high bone mineral density (BMD). Furthermore, it has been reported that this region may have duplicated several times in the mouse genome. Therefore, genomic sequencing of this region will provide important information for mouse genome structure, for positional cloning of mouse genes, and for the study of human homologous genes. In order to provide a suitable template for genomic sequencing by the NIH-sponsored genomic centers, we have constructed a BAC contig of this region using the RPCI-23 library. We have also identified the currently available mouse genomic sequences localized in our BAC contig. Further analysis of these sequences and BAC clones indicated a high frequency of repetitive sequences within this chromosomal area. This region also contains L1 retrotransposon sequences, providing a potential mechanism for the repetitive sequences described in the literature.     

Identification of a novel inflammation-protective locus in the Fischer rat

Inbred LEW/N rats are relatively susceptible, while histocompatible inbred F344/N rats are relatively resistant to development of a wide variety of inflammatory diseases in response to a range of pro-inflammatory stimuli. In a LEW/N vs. F344/N F2 intercross, we identified a quantitative trait locus (QTL) on Chr 10 that protects in a dominant fashion against the exudate volume component of innate inflammation in the F344/N rat, as well as a suggestive QTL on Chr 2 near the Fibrinogen cluster region. The exudate volume linkage region on Chr 10 may be similar to one of the multiple regions found to link to inflammatory arthritis phenotypes in other crosses. The suggestive linkage on Chr 2 has not been previously reported and does not seem to contribute to this phenotype in the same manner as the QTL on Chr 10. These findings are consistent with the hypothesis that the innate exudate volume trait is a sub-phenotype of more complex inflammatory phenotypes, such as arthritis, and genes within the Chr 10 linkage region could account for differences in this non-specific acute phase component of the inflammatory response. Since the rat Chr 10 exudate volume linkage region we have identified is syntenic with a region of human Chr 17 that has been shown to link to a variety of autoimmune/inflammatory diseases, including insulin-dependent diabetes mellitus, multiple sclerosis, and psoriasis, identification of genes within this linkage region will shed light on genes relevant to the earliest inflammatory component and to susceptibility and resistance to such human autoimmune/inflammatory diseases. Received: 4 August 1998 / Accepted: 4 December 1998     

Mapping of a Chromosome 12 Region Associated with Airway Hyperresponsiveness in a Recombinant Congenic Mouse Strain and Selection of Potential Candidate Genes by Expression and Sequence Variation Analyses

In a previous study we determined that BcA86 mice, a strain belonging to a panel of AcB/BcA recombinant congenic strains, have an airway responsiveness phenotype resembling mice from the airway hyperresponsive A/J strain. The majority of the BcA86 genome is however from the hyporesponsive C57BL/6J strain. The aim of this study was to identify candidate regions and genes associated with airway hyperresponsiveness (AHR) by quantitative trait locus (QTL) analysis using the BcA86 strain. Airway responsiveness of 205 F2 mice generated from backcrossing BcA86 strain to C57BL/6J strain was measured and used for QTL analysis to identify genomic regions in linkage with AHR. Consomic mice for the QTL containing chromosomes were phenotyped to study the contribution of each chromosome to lung responsiveness. Candidate genes within the QTL were selected based on expression differences in mRNA from whole lungs, and the presence of coding non-synonymous mutations that were predicted to have a functional effect by amino acid substitution prediction tools. One QTL for AHR was identified on Chromosome 12 with its 95% confidence interval ranging from 54.6 to 82.6 Mbp and a maximum LOD score of 5.11 (p = 3.68×10⁻³). We confirmed that the genotype of mouse Chromosome 12 is an important determinant of lung responsiveness using a Chromosome 12 substitution strain. Mice with an A/J Chromosome 12 on a C57BL/6J background have an AHR phenotype similar to hyperresponsive strains A/J and BcA86. Within the QTL, genes with deleterious coding variants, such as Foxa1, and genes with expression differences, such as Mettl21d and Snapc1, were selected as possible candidates for the AHR phenotype. Overall, through QTL analysis of a recombinant congenic strain, microarray analysis and coding variant analysis we identified Chromosome 12 and three potential candidate genes to be in linkage with airway responsiveness.     

Echinoid mutants exhibit neurogenic phenotypes and show synergistic interactions with the Notch signaling pathway

During neurogenesis in Drosophila, groups of ectodermal cells are endowed with the capacity to become neuronal precursors. The Notch signaling pathway is required to limit the neuronal potential to a single cell within each group. Loss of genes of the Notch signaling pathway results in a neurogenic phenotype: hyperplasia of the nervous system accompanied by a parallel loss of epidermis. Echinoid (Ed), a cell membrane associated Immunoglobulin C2-type protein, has previously been shown to be a negative regulator of the EGFR pathway during eye and wing vein development. Using in situ hybridization and antibody staining of whole-mount embryos, we show that Ed has a dynamic expression pattern during embryogenesis. Embryonic lethal alleles of ed reveal a role of Ed in restricting neurogenic potential during embryonic neurogenesis, and result in a phenotype similar to that of loss-of-function mutations of Notch signaling pathway genes. In this process Ed interacts closely with the Notch signaling pathway. Loss of ed suppresses the loss of neuronal elements caused by ectopic activation of the Notch signaling pathway. Using a temperature-sensitive allele of ed we show, furthermore, that Ed is required to suppress sensory bristles and for proper wing vein specification during adult development. In these processes also, ed acts in close concert with genes of the Notch signaling pathway. Thus the extra wing vein phenotype of ed is enhanced upon reduction of Delta (Dl) or Enhancer of split [E(spl)] proteins. Overexpression of the membrane-tethered extracellular region of Ed results in a dominant-negative phenotype. This phenotype is suppressed by overexpression of E(spl)m7 and enhanced by overexpression of Dl. Our work establishes a role of Ed during embryonic nervous system development, as well as adult sensory bristle specification and shows that Ed interacts synergistically with the Notch signaling pathway.     

Naturally segregating quantitative trait loci affecting wing shape of Drosophila melanogaster

Variation in vein position and wing shape of Drosophila melanogaster depends on many genes. In the following, we report the results of a QTL analysis of wing shape in D. melanogaster. We identified QTL responsible for natural variation for wing shape and analyzed their interactions with developmental genetic signaling pathways important for vein positioning. The QTL analysis indicated that the total number of QTL segregating in this population is likely to be very large. The locations of putative QTL identified in this study were compared to those identified in previous studies and, while there is more correspondence across studies than expected by chance on the third chromosome, the studies appear to have identified different QTL. Using a complementation design, we tested for interactions among these QTL with the Hedgehog and Decapentaplegic signaling pathways, which are important for the development and position of vein pairs L3-L4 and L2-L5. Three QTL showed strong interactions with these two pathways, supporting the hypothesis that these QTL are involved in these pathways. Naturally segregating variation can therefore act through known signaling pathways to produce variation in vein position.     

Rearranged gene order between pig and human in a QTL region on SSC 7

On porcine Chromosome 7, the region surrounding the MHC region contains QTL influencing many traits including growth, back fat thickness, and carcass composition. Towards the identification of the responsible gene(s), this article describes an increase of density of the radiated hybrid map of SSC 7 in the q11-q14 region and the comparative analysis of gene order on the porcine RH map and human genome assembly. Adding 24 new genes in this region, we were able to build a framework map that fills in gaps on the previous maps. The new software Carthagene was used to build a robust framework in this region. Comparative analysis of human and porcine maps revealed a global conservation of gene order and of distances between genes. A rearranged fragment of around 3.7 Mb was, however, found in the pig approximately 20 Mb upstream from the expected location on the basis of the human map. This rearrangement, found by RH mapping on the IMpRH 7.000 rads panel, has been confirmed by two-color FISH and by mapping on the high resolution IMNpRH2 12.000 rads panel. The rearranged fragment contains two microsatellites found at the most likely QTL location in the INRA QTL experiment. It also contains the BMP5 gene, which, together with CLPS, could be considered as a possible candidate.     

Genetic mapping of two QTL from the wild tomato Solanum pimpinellifolium L. controlling resistance against two-spotted spider mite (Tetranychus urticae Koch)

A novel source of resistance to two-spotted spider mite (Tetranychus urticae Koch) was found in Solanum pimpinellifolium L. accession TO-937 and thereby a potential source of desirable traits that could be introduced into new tomato varieties. This resistance was found to be controlled by a major locus modulated by minor loci of unknown location in the genome of this wild tomato. We first applied a bulked segregant analysis (BSA) approach in an F₄ population as a method for rapidly identifying a genomic region of 17 cM on chromosome 2, flanked by two simple sequence repeat markers, harboring Rtu2.1, one of the major QTL involved in the spider mite resistance. A population of 169 recombinant inbred lines was also evaluated for spider mite infestation and a highly saturated genetic map was developed from this population. QTL mapping corroborated that chromosome 2 harbored the Rtu2.1 QTL in the same region that our previous BSA findings pointed out, but an even more robust QTL was found in the telomeric region of this chromosome. This QTL, we termed Rtu2.2, had a LOD score of 15.43 and accounted for more than 30 % of the variance of two-spotted spider mite resistance. Several candidate genes involved in trichome formation, synthesis of trichomes exudates and plant defense signaling have been sequenced. However, either the lack of polymorphisms between the parental lines or their map position, away from the QTL, led to their rejection as candidate genes responsible for the two-spotted spider mite resistance. The Rtu2 QTL not only serve as a valuable target for marker-assisted selection of new spider mite-resistant tomato varieties, but also as a starting point for a better understanding of the molecular genetic functions underlying the resistance to this pest.     

There is more to tomato fruit colour than candidate carotenoid genes

Determining gene sequences responsible for complex phenotypes has remained a major objective in modern biology. The candidate gene approach is attempting to link, through mapping analysis, sequences that have a known functional role in the measured phenotype with quantitative trait loci (QTL) that are responsible for the studied variation. To explore the potential of the candidate approach for complex traits we conducted a mapping analysis of QTL for the intensity of the red colour of the tomato fruit (mainly lycopene) and for probes associated with the well-characterized carotenoid biosynthesis pathway. Seventy-five tomato introgression lines (ILs), each containing a single homozygous RFLP-defined chromosome segment from the green-fruited species Lycopersicon pennellii delimited 107 marker-defined mapping bins. Three of the bins resolved known qualitative colour mutations for yellow (r) and orange (B and Del) fruits resulting from variation in specific carotenoid biosynthesis genes. Based on trials in different environments, 16 QTL that modified the intensity of the red colour of ripe fruit were assigned to bins. Candidate sequences associated with the carotenoid biosynthesis pathway were mapped to 23 loci. Only five of the QTL co-segregated with the same bins that contained candidate genes - a number that is expected by chance alone. Furthermore, similar map location of a QTL and a candidate is far from a direct causative relationship between a gene and a phenotype. This study highlights the wealth and complexity of the variation present in the genus Lycopersicon that could be employed for basic research and genetic improvement of fruit colour in tomato.