Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Mapping of interactions and mouse congenic strains identified novel epistatic QTLs controlling the persistence of Salmonella Enteritidis in mice


The host response to infection in humans is multifactorial and involves the complex interaction between two genomes (the host and the pathogen) and the environment. Using an experimental mouse model of chronic infection, we have previously identified the individual effect of three significant and one suggestive quantitative trait loci (QTLs) (Ses1, Ses2, Ses3 and Ses1.1) on Salmonella Enteritidis persistence in target organs of 129S6/SvEvTac mice. Congenic strain construction was performed by transferring each of these QTLs from C57BL/6J onto the 129S6/SvEvTac background, and phenotypic analysis confirmed that Ses1 and Ses1.1 contribute to bacterial clearance. Additional QTLs regulating Salmonella carriage in 129S6/SvEvTac mice were identified using a two-locus epistasis QTL linkage mapping approach conducted separately in females and males. The epistatic model for females included the individual effect of Ses3 and two significant interactions (Ses1D7Mit267 and Ses1DXMit48) accounting for 47% of the total phenotypic variance. The model for males included the individual effect of Ses1.1, three interactions (Ses1D9Mit218, D2Mit197D4Mit2 and D3Mit256D13Mit36) and explained 47% of the phenotypic variance. Our results suggest that the oligogenic nature of Salmonella persistence and epistasis are important constituents of the genetic architecture of the host response to chronic Salmonella infection.


Salmonella are ubiquitous, facultative intracellular Gram-negative bacteria causing human diseases of increasingly important public health concern both in developed and developing countries. Salmonella enterica serovar Typhi (S. Typhi) is the causative agent of typhoid fever and affects 21 million people annually with 200 000 associated deaths. Approximately 5% of infected patients develop a chronic carrier state-excretion of the bacteria for more than 1 year and certain individuals become life-long carriers.1 Salmonella enterica serovar Typhimurium (S. Typhimurium) and Enteritidis (S. Enteritidis) are associated with frequent outbreaks of food-borne gastroenteritis (salmonellosis). There are 1.4 million reported cases of salmonellosis annually in the United States alone, with more than 600 deaths associated with untreated infections. Gastroenteritis caused by Salmonella infection is usually self-limiting; however, approximately 2% of individuals infected with Salmonella will develop chronic arthritis resulting from reactive joint complications, whether or not they were treated with antibiotics.1

The outcome of Salmonella infection depends on the innate and adaptive immune responses of the host and on the strategies developed by Salmonella to interfere with the host immune system to survive and replicate in a hostile environment. Studies in experimental animal models of infection have shown that susceptibility to Salmonella infection is under complex genetic control.2 Most models have used S. Typhimurium, which causes an acute fatal infection in laboratory mice and shares many similarities with human typhoid disease with respect to pathology and host response.3 Several critical innate immune genes, including Nramp1 (Natural resistance macrophage protein 1, renamed Slc11a1), Tlr4 (Toll-like receptor 4), Nos2 (Nitric oxide synthase 2) and Nadph oxidase have been identified using survival rates to infection and bacterial load in the reticuloendothelial system (RES) as phenotype (reviewed in Roy and Malo2). The genetics of chronic Salmonella infection in mouse models has also been studied, but not as extensively. Chronic models of infection have been developed using mutant S. Typhimurium strains in susceptible animals, sublethal infection in resistant mice, or pretreatment of mice with antibiotics.4, 5 The clearance of the bacteria involves B cells, CD4+ and CD8+ T cells and MHC class II genes.6 A number of genetic factors including several cytokines (IFNγ, IL-12, IL-18 and TNFα) have been found to influence the clearance of the bacteria from the RES during the late stage of infection using congenic mice and mutant mouse resources.2, 7

In order to determine the genetic basis for the persistence of Salmonella infection, we have developed a model of chronic Gram-negative infection using a sublethal dose of S. Enteritidis.8 The mouse strain C57BL/6J cleared the bacteria completely from their RES (spleen and liver) by 42 days postinoculation, whereas the 129S6 mice had high bacterial load in the spleen up to 60 days postinfection.8 Despite the presence of bacteria, 129S6 mice did not develop any disease symptoms. C57BL/6J and 129S6 mice carry the same MHC II haplotype but they differ at Nramp1 (C57BL/6J mice carry a nonfunctional form of the gene and 129S6 carry the wild-type allele). Our initial whole-genome scan linkage analysis in 300 segregating (C57BL/6J × 129S6) F2 mice led to the mapping of three highly significant quantitative trait loci (QTLs), located on chromosomes 1, 7 and 15, respectively. All QTLs were associated with disease susceptibility in 129S6 mice, and their estimated effects on bacterial clearance were greater in females. There was significant statistical interaction detected between the loci on chromosomes 1 and 7. Using a model including those three loci, the interaction as well as sex as a covariate, the genetic components explained 32% of the phenotypic variance, suggesting the action of additional unidentified genetic loci contributing to the expression of the phenotype.

Most efforts in complex trait analysis have been directed towards mapping individual loci having individual effects on the disease phenotype. However, recent evidence has shown that gene–gene interactions are crucial in the etiology of many complex diseases both in humans and animal models of human diseases.9, 10 Another component of the complexity of the host response to infection can be attributed to sex differences. The objectives of the present study were to validate existing QTLs using congenic mouse strains, to carry out a linkage analysis separately in males and females to allow the detection of sex-specific QTLs, and to scan for two-locus interactions to identify additional genetic modifiers contributing to S. Enteritidis carriage in 129S6 mice. Studying interaction in F2 cross was proven to be an essential and critical step in the genetic study of the host response to chronic Salmonella infection in mice.


Contribution of individual QTLs to Salmonella clearance

Following our original linkage analysis of Salmonella infected (C57BL/6J × 129S6) F2 mice, we identified three significant and one suggestive QTLs that influenced bacterial clearance following a sublethal challenge. These QTLs were located on chromosomes 1 (Ses1 and Ses1.1), 7 (Ses2) and 15 (Ses3).8 To confirm the existence of each QTL in vivo and to study their impact on a Salmonella infection, we constructed a panel of congenic mouse strains carrying chromosomal regions encompassing Ses1.1, Ses1, Ses2 and Ses3. These strains were obtained by introgressing specific chromosomal segments of the C57BL/6J mice (donor strain) onto the background of the recipient 129S6 strain by repeated backcrossing (129S6.B6) (Table 1). A 2-LOD down support interval was used to establish the limits of the genetic interval to be transferred. The Ses1, Ses2, Ses3 and Ses1.1 genomic intervals transferred from C57BL/6J to 129S6 mice are shown in Figure 1. We also constructed double congenic mice for Ses1 and Ses2 to test functionally the statistical interaction between Ses1 and Ses2 described in our previously proposed model.8

Table 1 Description of congenic strains
Figure 1

Genetic map of the (C57BL/6J × 129S6) F2 cross. The genetic map is a sex-average map. The distances are in centimorgan and drawn to scale from MapManager QTXb19. The map was obtained from the genotyping data of Caron et al (2002), where a total of 116 markers were used to genotype the 93 F2 phenotypic extreme mice, merged with data obtained from the genotyping of 141 additional markers on the 300 F2 progeny. The QTLs are indicated by hatched box next to their respective chromosomic location, and the markers involved in our models are marked by clear boxes on the respective chromosomes. The region of chromosome 7 containing TRD is represented by a grey bar.

To assess the effects of each individual QTL on Salmonella clearance, the congenic strains were infected with a sublethal dose of S. Enteritidis as described by us previously, and the results were analyzed separately in male and female mice because of the known sex effect on the phenotype.8 Mice were killed at day 42 and the spleen was collected for CFUs determination. Consistent with previous results, there was a significant difference in the bacterial load of parental strains: C57BL/6J mice had low bacterial counts (0.29±0.29 in females and 1.24±0.42 in males) compared to 129S6 mice that carried an average of 3.17±0.13 (females) and 3.82±0.1 (males) LOG CFUs per gram of spleen (P=0.0000 and 0.0002, respectively) (Figure 2). CFUs counts were 2.6 LOG (males) to 2.9 LOG (females) times lower in C57BL/6J mice compared to 129S6. Female and male congenic mice 129.B6-Ses1.1, 129.B6-Ses1 and 129.B6-Ses1/Ses2 cleared the infection significantly better than the 129S6 parental strain (P values for females and males are respectively: P=0.0077 and 0.0006 for 129.B6-Ses1.1; P=0.0267 and 0.0000 for 129.B6-Ses1 and P=0.0013 and 0.0000 for 129.B6-Ses1/Ses2) (Figure 2). For all mouse strains with the exception of 129.B6-Ses1.1, females had significantly lower bacterial counts in their spleen than did males. We did not detect an effect of Ses2 and Ses3 in the congenics: 129.B6-Ses2 and 129.B6-Ses3 mice had bacterial loads in their spleen comparable to those observed in 129S6 mice, suggesting that the effect measured in 129.B6-Ses1/Ses2 on bacterial clearance is contributed mostly by Ses1. The transfer of resistance alleles detected by linkage analysis was successful and confirms that Ses1 and Ses1.1 contribute to bacterial clearance in both males and females. Ses1 and Ses1.1 have a similar impact on bacterial clearance (1 LOG decrease compared to strain 129S6) in congenic mice.

Figure 2

Salmonella-carrier status of 129S6.B6 congenic mice in the spleen 42 days postinoculation following a sublethal infection with S. Enteritidis. The data are expressed in LOG CFU/g spleen±s.e.m. The black and white columns represent the female and male mice, respectively. Asterisks represent the level of significance in the bacterial load difference between the two strains of mice (*P<0.05), **P<0.01 and ***P<0.001).

Genetic map and regression mapping

A higher density framework genetic map was constructed on the basis of the segregating population of 300 (C57BL/6J × 129S6) F2 progeny. The linkage map consists of 244 microsatellites and three SNP markers distributed along the 19 autosomes and chromosome X. The average intermarker distance was 10.7±0.4 cM. The coverage and resolution of this linkage map provided the basis for the identification of QTLs having small effects on Salmonella clearance and of QTLs undetected in our previous linkage analysis, as well as for the identification of interactions between genomic regions with no or small individual effects. The positions of the markers across the mouse genome are represented to scale in Figure 1. The higher density of markers on the current genetic map allowed us to detect the presence of transmission ratio distortion (TRD) encompassing a 40.7 cM region on chromosome 7, located between the markers D7Mit228 and D7Mit237. In this region of the genome, the C57BL/6J homozygous genotype was found at frequencies ranging from 39 to 41%, which are much higher than the expected 25% in female F2 mice (Figure 1). Interestingly, the observed allele frequencies in this region were very close to the expected 50%, and the frequency of the C57BL/6J homozygotes was increased when compared to the other two genotypes. The presence of TRD can mislead the identification of QTLs particularly when interval mapping techniques are used. To circumvent this problem, linkage analysis was performed using bacterial load in the spleen 42 days postinfection as phenotype using the approach of regression on the markers. Having observed sex–genotype interaction in affecting bacterial clearance previously8 and the presence of TRD in the F2 female mice, the analysis was carried out separately in females (n=148) and males (n=152).

Preliminary single-point linkage analysis in females provided evidence of linkage to chromosome 1 at marker D1Mcg5 (LOD score of 7.59, P=0.0001). This marker is located at the peak maximum LOD score of Ses1 (Figure 1). However Ses2, which maps within the TRD region of chromosome 7, and Ses3 were not detected at this stage. In males, Ses1 showed significant linkage with a peak LOD score of 4.4 (P=0.0082) at D1Mit19. For both QTLs, the C57BL/6J allele is protective and inherited in a recessive way.

Global two-locus epistatic QTL mapping analyses

The experimental data set was reanalyzed to detect pairwise interactions (epistasis) by examining all combinations of marker loci for association with the phenotype. The QTLs and their positions in the genome are listed in Table 2. This analysis allowed us to identify seven new interacting QTLs, designated Ses4 to Ses10, involved in the clearance of Salmonella in female and male F2 mice.

Table 2 Genomic regions affecting Salmonella clearance in (C57BL/6J × 129S6) F2 mice

In females, two interaction terms were detected with D1Mcg5 (Ses1): one highly significant interaction with D7Mit267 (Ses5) and a second one with DXMit48 (Ses4) (Table 3). The marker D7Mit267 is located 20 cM proximal to Ses2 and was given a specific designation (Ses5). Ses4 (Chr X) and Ses5 (Chr 7) had no individual detectable effects (Table 3). In this analysis, Ses1 had no or small individual effect and necessitated the presence of Ses4 and/or Ses5 to express its action on Salmonella clearance (Figure 3). The mice that are homozygous for the C57BL/6J allele at Ses1 and Ses5 (genotype 1.1 on the box plot of Figure 3a) had a bacterial load 15 times lower then mice that are homozygous for the C57BL/6J alleles only at Ses1 (genotypes 1.2 and 1.3 on the box plot of Figure 3a). A similar decrease in bacterial load is observed when mice are homozygous C57BL/6J at Ses1 and Ses4 (genotype 1.1 on the box plot of Figure 3b) compared to mice that are homozygous only at Ses1 (genotype 1.2). Only Ses3 had a small but detectable individual effect on bacterial clearance in females (Table 3). The most parsimonious model for Salmonella clearance in the F2 females included the effect of Ses3 and two highly significant interactions between Ses1 and Ses4 and between Ses1 and Ses5. The contribution of this model to the total phenotypic variance is 47% and the interactions alone accounted for almost three quarters (72%) of this variation.

Table 3 Genetic model of Salmonella Enteritidis clearance in (C57BL/6J × 129S6) F2 mice
Figure 3

Two-way statistical interactions between D1Mcg5 (Ses1) and D7Mit267 (Ses5) (a) and between D1Mcg5 (Ses1) and DXMit48 (Ses4) (b) in female mice as shown by a box-plot graphed in R. The numbers indicate the genotype, where 1 represents C57BL/6J homozygous alleles, 2 is the heterozygous state between the C57BL/6J and 129S6 alleles, and 3 represents homozygous 129S6 alleles. The x-axis represents the different combination of markers for D1Mcg5D7Mit267 and D1Mcg5DXMit48, and the y-axis represents the LOG CFU/g spleen.

In males, one significant QTL with individual effect was detected on chromosome 1 at D1Mit123 (Ses1.1) (Figure 1 and Table 3). There were also three significant interaction terms identified: between Ses1 and D9Mit218 (Ses6), between D2Mit192 and D4Mit2 (Ses7 and Ses8) and between D3Mit256 (Ses9) and D13Mit36 (Ses10) (Figure 1 and Table 3). As observed in females, the impact of Ses1 on bacterial clearance was seen only in the context of epistasis and all QTLs involved in the interaction terms had no detectable individual effects. For every QTL, the C57BL/6J allele was protective and inherited in a recessive way, except for Ses7 that was inherited dominantly. The epistasis model for Salmonella clearance in males included one QTL with individual effect (Ses1.1) and three significant interactions (Ses1–Ses6, Ses7–Ses8 and Ses9–Ses10). This model explained 47% of the phenotypic variance. In this case, interactions alone explain two-thirds (66%) of this variation.


Development of clinical disease in humans after exposure to microorganisms is a multifactorial process under host and environmental control.11, 12 The host genetic architecture of susceptibility to Salmonella infection and persistence is complex and its understanding requires the identification of the causative genes and the knowledge of how these genes interact functionally. To find the genes that control Salmonella persistence, different mouse models of chronic Salmonella infection have been developed to test the biological impact of candidate genes and to perform QTL analyses.7, 8, 13 The experimental strategy we used to study the complex inheritance of Salmonella persistence in mice was based on mapping studies using phenotypic variation to identify QTLs. For this purpose, we have developed a model of persistent Salmonella infection based on the inoculation of a sublethal dose of S. Enteritidis in inbred laboratory strains 129S6 and C57BL/6J. 129S6 mice were previously shown to be extremely resistant to infection with S. Typhimurium compared to all other classical inbred strains including C57BL/6J.2 The extreme susceptibility of C57BL/6J mice to infection with S. Typhimurium has been attributed to the presence of a nonfunctional allele at Nramp1.14, 15 In contrast to S. Typhimurum, S. Enteritidis does not cause clinical disease in C57BL/6J and 129S6 mice. The infection is associated with a transitory splenomegaly in both strains and C57BL/6J mice are able to clear completely the bacteria within 7 weeks postinoculation. In 129S6 mice, S. Enteritidis persist predominantly in the spleen and to a lesser extent in the mesenteric lymph nodes (MLN) for a period extending 10 weeks without causing significant histological changes.

We previously detected by linkage analysis of a segregating C57BL/6J × 129S6 F2 progeny four QTLs conferring bacterial susceptibility on chromosomes 1 (Ses1 and Ses1.1), 7 (Ses2) and 15 (Ses3).8 We constructed congenic strains to provide support for a functional role of these QTLs by introgressing the specific chromosomal regions from C57BL/6J mice onto a 129S6 background. The use of congenic mice for fine mapping of Salmonella-resistance loci has several advantages including repeat phenotyping of genetically identical animals and dissection of a particular haplotype on a defined genetic background. On a fixed genetic background as seen in congenic 129S6.B6-Ses1, an individual effect of Ses1 was detected in both males and females although the impact of Ses1 was more evidenced in females. Replacing the QTL Ses1 reduced the bacterial load by 19-fold in females and 4-fold in males. Ses1.1 was also involved in the clearance of S. Enteritidis in spleen in both males and in females: there was a 11-fold reduction in bacterial load in females and a 18-fold reduction in males. The transfer of the Ses2 interval had no impact on bacterial clearance, which could be explained by the fact that our initial interval mapping analysis did not identify the correct position of the QTL on chromosome 7 because of the presence of strong TRD (see below). The contribution of Ses3 on bacterial clearance was not detected in the spleen of congenic mice. This could be explained by the small individual effect of Ses3 (2%) to the phenotypic variance. We can conclude that the congenic mouse strains we have developed for the chromosome 1 QTLs provide biological evidence for the role of specific Ses1 and Ses1.1 alleles in bacterial clearance both in females and males.

We undertook a global search for epistatic QTLs that mediate their effects on bacterial clearance through interaction with other genes, based on the fact that our previous linkage analysis explained only one-third of the phenotypic variance. In recent years, the importance of mapping epistatic QTL to increase the understanding of the molecular mechanisms behind complex diseases has been stressed16 and several methods for mapping epistatic QTL have been proposed.17, 18 To carry out this type of analysis, we have increased the density of markers and genotyped all 300 segregating F2 progeny. The analysis was stratified by sex because our experiments provided clear evidence of sex-specific QTL differences. The higher density map was essential to detect TRD on chromosome 7 and identify new interacting QTLs. TRD may have implication for the mapping of loci for complex disease and may lead to misinterpretation of the data. The reanalysis of our data set confirmed that Ses1 plays a major role in Salmonella persistence both in male and female mice. In the novel proposed models, Ses1 has no individual effect, but exerts its effect on the phenotype through its interaction with Ses4 (DXMit48) and/or Ses5 (D7Mit267) in females and with Ses7 (D9Mit218) in males. The best model for bacterial clearance in females involved the individual effect of Ses3 and interactions of Ses1 with Ses4 and Ses5. In males, the genetic inheritance of bacterial clearance involves one locus with individual effect (Ses1.1) and three interactions (Ses1–Ses6, Ses7–Ses8 and Ses9–Ses10). In these new models, the proportion of the phenotypic variance explained by the QTL model has increased (47% in both males and females). In addition, the majority of the explained variance can be attributed to loci interactions (72% in females and 66% in males), showing the importance of accounting for epistasis in order to identify the mechanisms underlying a complex phenotype such as bacterial persistence.

We did not detect an individual or epistatic effect of Ses2 in the reanalysis of our data. This could be explained by the fact that the region of chromosome 7 encompassing Ses2 presents a strong TRD in favour of C57BL/6J alleles in F2 females. Of interest, TRD has been associated with the alleles at disease loci in a number of human diseases19 and the association of TRD with only one sex of offspring may be a hallmark of defective imprinting at both the autosomal locus and an X-chromosome locus.19, 20, 21 In fact, this central region of chromosome 7 comprises a cluster of imprinted genes.22, 23 An unexpected finding was that Ses1.1 contributes to the phenotype in both male and female congenic mice. The difficulty of detecting phenotypic effects of Ses1.1 in F2 female mice by mapping may be related to the limited sample size. In fact, the sample size must be large enough to detect variation at both Ses1.1 and Ses1, which are estimated to be 17.4 cM apart. However, evidence for both Ses1 and Ses1.1 was obtained from the congenic strain analyses.

We previously reported strong evidence suggesting that Nramp1 is the gene underlying Ses1.8 Nramp1 is known to control the exponential growth rate of taxonomically unrelated pathogens, including S. Typhimurium, in mice during the early phase of an acute infection.24 During Salmonella infection, phagocytes ingest the bacteria into a phagosome, which matures by sequential fusions with a series of endosomal and lysosomal compartments and results in the formation of phagolysosomes that possess antimicrobial properties. To survive within host cells, Salmonella evade this process and affect the maturation process of the phagosome by generating a unique compartment termed Salmonella containing vacuole (SCV).25 During Salmonella infection, Nramp1 is recruited to the SCV and functions as a pH-dependent manganese (Mn2+) and iron (Fe2+) pump.26 In addition, Nramp1 affects SCV maturation by promoting the acquisition of mannose 6 phosphate receptor (M6PR) a protein known to regulate the delivery of a subset of lysosomal enzymes from the trans-Golgi network to the prelysosomal compartment, thereby facilitating bacterial killing.27, 28 Iron depletion in primary macrophages from Nramp1-deficient mice restored phagosome maturation, suggesting that Nramp1 counteracts the ability of Salmonella to arrest phagosome maturation through depletion of iron and other cations.29

In our model of chronic infection, congenic mice carrying a nonfunctional allele at Nramp1 (129.B6-Ses1) and 129S6-Nramp1 deficient mice had a more efficient bacterial clearance rate than 129S6 mice. Our results strongly suggest that Nramp1 interacts with other genes to influence the outcome of a chronic Salmonella infection. We searched for candidate genes that are located on the chromosomal regions adjacent to interacting loci D7Mit267 (18.5 Mb), DXMit48 (40.4 Mb) and D9Mit218 (9.0 Mb). We have identified several genes for which a role in immunity to Salmonella infection is known or may be envisioned including Nfkbib (nuclear factor of kappa light chain gene enhancer in B-cells inhibitor beta), CD22, Rog (repressor of Gata) and Hamp (hepcidin) on chromosome 7, Elf4 (ets domain transcription factor 4), Il13ra1 (interleukin 13 receptor alpha 1) and Lamp2 (lysosomal membrane glycoprotein 2) on chromosome X and Casp1 (caspase 1) and Ilf3 (interleukin enhancer binding factor 3) on chromosome 9. One of these genes (Hamp) retained more our attention because of its role in restricting the availability of iron to microbes,30 Hamp play a central role in iron metabolism as a potent negative regulator of iron absorption and mobilization.31, 32 Liver Hamp mRNA levels are upregulated after iron overload,33 exposure to LPS33 or bacterial infections.34 Hamp-induced sequestration of iron in macrophages, which has an impact on phagocytosis, and IFNg regulated production of nitric oxide and TNFα.35 The biological interaction between Nramp1 and Hamp needs to be confirmed but it is tempting to speculate that the absence of Nramp1 may promote the activity of Hamp.

We can conclude from this study that the combination of a higher density linkage map together with modelling gene–gene interactions and stratification by sex permitted us to propose novel models of inheritance of Salmonella persistence for males and females. The analysis by sex showed clearly that Ses1 had a greater impact on the phenotypic variance in the context of interaction with other sex-specific loci. The analysis also detected the presence of TRD and led to the rectification of the position of the QTL on chromosome 7. The identification of gene interactions not only provided a more comprehensive understanding of the complex inheritance of bacterial persistence in chronic Salmonella infection but will also be crucial for further creation of congenic and double congenic mice, which will then be instrumental to finely resolve QTLs, confirm their association with the phenotype and identify the causative genes.

Materials and methods


Inbred mouse strains 12956/SvEvTac and C57BL/6J, initially obtained from Taconic (Germantown, NY, USA) and The Jackson Laboratories (Bar Harbor, ME, USA), respectively, were bred and maintained in our animal facilities under conditions specified by the Canadian Council on Animal Care. In total, 14 breeding F1 mouse pairs were set up and produced 300 F2 (C57BL/6J × 129S6) hybrids.

Congenic strains

Congenic mice for Ses1, Ses2, Ses3 and Ses1.1 and the double congenics Ses1–Ses2 were produced by creating F1 mice from the 129S6 and C57BL/6J inbred mice then performing five to six successive backcrosses to the 129S6 parental strain. The congenic strains were produced in the absence of functional phenotype analysis by using marker-assisted genotyping to identify and maintain the targeted introgressed segments. At the N2 generation, we selected mice providing multiple chromosomal segments from the donor strains as progenitor for multiple regions. We have performed whole-genome scans at the N2 and N3 generations with 194 fluorescently labelled markers spaced at approximately 9 cM intervals to select mice with the least overall heterozygosity for subsequent backcrossing. Mice from the N4 generation were genotyped with 70 markers and those showing no detectable C57BL/6J donor background were selected for further breeding. Homozygous founders were established by brother–sister mating of N5–6 mice.

Salmonella infection

The preparation of the Salmonella stock and inoculum were as described by us previously.36 F2 progeny, N5 or N6 congenic mice, 129S6 and C57BL/6J controls were infected with 1000 CFUs of S. Enteritidis strain 3b (Dr William Kay, University of Victoria) intravenously and killed by carbon dioxide asphyxiation 42 days postinoculation. The bacterial load from the spleens was evaluated by counting colony-forming units (CFUs) as described by us previously.8 The LOG of the number of CFUs per gram of spleen (log10 CFU/g) was analyzed as a quantitative trait.


A total of 244 microsatellite markers and three SNPs informative for the C57BL/6J and 129S6 parental strains (The Jackson Laboratories, Bar Harbor, ME, USA; Center for Inherited Disease Research,; Applied Biosystems, Foster City, CA, USA; Research Genetics, Hunstville, AL, USA) were used to cover the genome at approximately 10.7 cM interval. Genotypes for the simple sequence length polymorphisms (SSLPs) were determined by a standard polymerase chain reaction (PCR)-based technique using either an end-labelled (γ-33P)-ATP primer, followed by separation on denaturing polyacrylamide gels8 or microsatellite markers tagged with fluorochromes (Research Genetics, Hunstville, AL, USA; Applied Biosystems, Foster City, CA, USA; Invitrogen Corporation, Burlington, ON, Canada) followed by capillary separation using an ABI prism 3700 instrument (Applied Biosystems, Foster City, CA, USA). Genotypes were analyzed using the GeneScan® V3.5 and Genotyper® V3.6 softwares (Applied Biosystems, Foster City, CA, USA) (McGill University and Génome Québec Innovation Centre, Montréal, PQ, Canada).

Statistical analysis

Owing to the known sex difference in the expression of the phenotype, the statistical analyses were carried out separately in F2 males and females using the statistical package ‘R’.37 An exploratory analysis consisting of a test of Mendelian segregation of the markers in one- and two-locus genome-wide scans was performed using the library ‘R/qtl’.38 The method of ‘marker regression’ was used because departure from Mendelian segregation was detected on chromosome 7 in females, one of the chromosomes of interest. The suggestive results of the one- and two-locus scan were analyzed graphically and the significance of the results obtained from the two-locus scan was reassessed by resampling. For the loci that remained significant, a design matrix of factors, taking into consideration the mode of inheritance, was constructed. The library ‘leaps’39 was used to select the best model with the Bayesian Information Criterion40 as the main indicator of parsimony. Significance of the selected models was assessed by resampling.41 For congenics data, statistical analysis to test the differences in bacterial counts was carried out using the two-sample t-test.


  1. 1

    CDC. Center for Disease Control and Prevention Report. Division of bacterial and mycotic diseases. Disease information,, Atlanta, Giorgia, December 2003.

  2. 2

    Roy MF, Malo D . Genetic regulation of host responses to Salmonella infection in mice. Genes Immun 2002; 3: 381–393.

    CAS  Article  Google Scholar 

  3. 3

    O'Brien AD, Scher I, Formal SB . Effect of silica on the innate resistance of inbred mice to Salmonella typhimurium infection. Infect Immun 1979; 25: 513–520.

    CAS  PubMed  PubMed Central  Google Scholar 

  4. 4

    Nauciel C, Vilde F, Ronco E . Host response to infection with a temperature-sensitive mutant of Salmonella typhimurium in a susceptible and a resistant strain of mice. Infect Immun 1985; 49: 523–527.

    CAS  PubMed  PubMed Central  Google Scholar 

  5. 5

    Barthel M, Hapfelmeier S, Quintanilla-Martinez L et al. Pretreatment of mice with streptomycin provides a Salmonella enterica serovar Typhimurium colitis model that allows analysis of both pathogen and host. Infect Immun 2003; 71: 2839–2858.

    CAS  Article  Google Scholar 

  6. 6

    Young D, Hussell T, Dougan G . Chronic bacterial infections: living with unwanted guests. Nat Immunol 2002; 3: 1026–1032.

    CAS  Article  Google Scholar 

  7. 7

    Monack DM, Bouley DM, Falkow S . Salmonella typhimurium persists within macrophage. J Exp Med 2004; 199: 231–241.

    CAS  Article  Google Scholar 

  8. 8

    Caron J, Loredo-Osti JC, Laroche L, Skamene E, Morgan K, Malo D . Identification of genetic loci controlling bacterial clearance in experimental Salmonella enteritidis infection: an unexpected role of Nramp1 (Slc11a1) in the persistence of infection in mice. Genes Immun 2002; 3: 196–204.

    CAS  Article  Google Scholar 

  9. 9

    Kim JH, Sen S, Avery CS et al. Genetic analysis of a new mouse model for non-insulin-dependent diabetes. Genomics 2001; 74: 273–286.

    CAS  Article  Google Scholar 

  10. 10

    Xu J, Langefeld CD, Zheng SL et al. Interaction effect of PTEN and CDKN1B chromosomal regions on prostate cancer linkage. Hum Genet 2004; 115: 255–262.

    CAS  PubMed  PubMed Central  Google Scholar 

  11. 11

    Buer J, Balling R . Mice, microbes and models of infection. Nat Rev Genet 2003; 4: 195–205.

    CAS  Article  Google Scholar 

  12. 12

    Casanova JL, Abel L . The human model: a genetic dissection of immunity to infection in natural conditions. Nat Rev Immunol 2004; 4: 55–66.

    CAS  Article  Google Scholar 

  13. 13

    Sukupolvi S, Edelstein A, Rhen M, Normark SJ, Pfeifer JD . Development of a murine model of chronic Salmonella infection. Infect Immun 1997; 65: 838–842.

    CAS  PubMed  PubMed Central  Google Scholar 

  14. 14

    Malo D, Vogan K, Vidal S et al. Haplotype mapping and sequence analysis of the mouse Nramp gene predict susceptibility to infection with intracellular parasites. Genomics 1994; 23: 51–61.

    CAS  Article  Google Scholar 

  15. 15

    Vidal S, Tremblay ML, Govoni G et al. The Ity/Lsh/Bcg locus: natural resistance to infection with intracellular parasites is abrogated by disruption of the Nramp1 gene. J Exp Med 1995; 182: 655–666.

    CAS  Article  Google Scholar 

  16. 16

    Carlborg O, Haley CS . Epistasis: too often neglected in complex trait studies? Nat Rev Genet 2004; 5: 618–625.

    CAS  Article  Google Scholar 

  17. 17

    Kao CH, Zeng ZB, Teasdale RD . Multiple interval mapping for quantitative trait loci. Genetics 1999; 152: 1203–1216.

    CAS  PubMed  PubMed Central  Google Scholar 

  18. 18

    Pociot F, Karlsen AE, Pedersen CB, Aalund M, Nerup J . Novel analytical methods applied to type 1 diabetes genome-scan data. Am J Hum Genet 2004; 74: 647–660.

    CAS  Article  Google Scholar 

  19. 19

    Naumova AK, Leppert M, Barker DF, Morgan K, Sapienza C . Parental origin-dependent, male offspring-specific transmission-ratio distortion at loci on the human X chromosome. Am J Hum Genet 1998; 62: 1493–1499.

    CAS  Article  Google Scholar 

  20. 20

    Naumova A, Sapienza C . The genetics of retinoblastoma, revisited. Am J Hum Genet 1994; 54: 264–273.

    CAS  Article  Google Scholar 

  21. 21

    Naumova AK, Greenwood CM, Morgan K . Imprinting and deviation from Mendelian transmission ratios. Genome 2001; 44: 311–320.

    CAS  Article  Google Scholar 

  22. 22

    Croteau S, Andrade MF, Huang F, Greenwood CM, Morgan K, Naumova AK . Inheritance patterns of maternal alleles in imprinted regions of the mouse genome at different stages of development. Mamm Genome 2002; 13: 24–29.

    CAS  Article  Google Scholar 

  23. 23

    Cattanach BM, Beechey CV, Peters J . Interactions between imprinting effects in the mouse. Genetics 2004; 168: 397–413.

    CAS  Article  Google Scholar 

  24. 24

    Vidal SM, Pinner E, Lepage P, Gauthier S, Gros P . Natural resistance to intracellular infections: Nramp1 encodes a membrane phosphoglycoprotein absent in macrophages from susceptible (Nramp1 D169) mouse strains. J Immunol 1996; 157: 3559–3568.

    CAS  Google Scholar 

  25. 25

    Gruenheid S, Finlay BB . Microbial pathogenesis and cytoskeletal function. Nature 2003; 422: 775–781.

    CAS  Article  Google Scholar 

  26. 26

    Forbes JR, Gros P . Iron, manganese, and cobalt transport by Nramp1 (Slc11a1) and Nramp2 (Slc11a2) expressed at the plasma membrane. Blood 2003; 102: 1884–1892.

    CAS  Article  Google Scholar 

  27. 27

    Gruenheid S, Pinner E, Desjardins M, Gros P . Natural resistance to infection with intracellular pathogens: the Nramp1 protein is recruited to the membrane of the phagosome. J Exp Med 1997; 185: 717–730.

    CAS  Article  Google Scholar 

  28. 28

    Cuellar-Mata P, Jabado N, Liu J et al. Nramp1 modifies the fusion of Salmonella typhimurium-containing vacuoles with cellular endomembranes in macrophages. J Biol Chem 2002; 277: 2258–2265.

    CAS  Article  Google Scholar 

  29. 29

    Jabado N, Cuellar-Mata P, Grinstein S, Gros P . Iron chelators modulate the fusogenic properties of Salmonella-containing phagosomes. Proc Natl Acad Sci USA 2003; 100: 6127–6132.

    CAS  Article  Google Scholar 

  30. 30

    Singh PK, Parsek MR, Greenberg EP, Welsh MJ . A component of innate immunity prevents bacterial biofilm development. Nature 2002; 417: 552–555.

    CAS  Article  Google Scholar 

  31. 31

    Nicolas G, Bennoun M, Devaux I et al. Lack of hepcidin gene expression and severe tissue iron overload in upstream stimulatory factor 2 (USF2) knockout mice. Proc Natl Acad Sci USA 2001; 98: 8780–8785.

    CAS  Article  Google Scholar 

  32. 32

    Nicolas G, Viatte L, Bennoun M, Beaumont C, Kahn A, Vaulont S . Hepcidin, a new iron regulatory peptide. Blood Cells Mol Dis 2002; 29: 327–335.

    CAS  Article  Google Scholar 

  33. 33

    Pigeon C, Ilyin G, Courselaud B et al. A new mouse liver-specific gene, encoding a protein homologous to human antimicrobial peptide hepcidin, is overexpressed during iron overload. J Biol Chem 2001; 276: 7811–7819.

    CAS  Article  Google Scholar 

  34. 34

    Shike H, Lauth X, Westerman ME et al. Bass hepcidin is a novel antimicrobial peptide induced by bacterial challenge. Eur J Biochem 2002; 269: 2232–2237.

    CAS  Article  Google Scholar 

  35. 35

    Ashrafian H . Hepcidin: the missing link between hemochromatosis and infections. Infect Immun 2003; 71: 6693–6700.

    CAS  Article  Google Scholar 

  36. 36

    Sebastiani G, Olien L, Gauthier S et al. Mapping of genetic modulators of natural resistance to infection with Salmonella typhimurium in wild-derived mice. Genomics 1998; 47: 180–186.

    CAS  Article  Google Scholar 

  37. 37

    Ihaka RGR . R: a language for data analysis and graphics. J Comput Graph Stat 1996; 5: 299–314.

    Google Scholar 

  38. 38

    Broman KW, Wu H, Sen S, Churchill GA . R/qtl: QTL mapping in experimental crosses. Bioinformatics 2003; 19: 889–890.

    CAS  Article  Google Scholar 

  39. 39

    Miller A . Subset Selection in Regression. Chapman and Hall: New York, 2002.

    Book  Google Scholar 

  40. 40

    Scharz G . Estimating the dimension of a model. Ann Stat 1978; 6: 461–464.

    Article  Google Scholar 

  41. 41

    Davison ACHD . Bootstrap Methods and Their Application. 4th edn. Cambridge University Press: Cambridge, 1997.

    Book  Google Scholar 

Download references


We thank Dr William Kay for providing S Enteritidis isolates and Rosalie Wilkinson and Line Larivière for their technical expertise. We also thank Dr Thomas Hudson, Dr Andrei Verner and Geneviève Geneau at the McGill University and Génome Québec Innovation Centre for their technical help and the use of their facility and equipment to perform the fluorescent genotyping work. We thank Dr Silvia Vidal for helpful discussions and critical review of the manuscript. This work was supported by grants from the Canadian Institutes of Health Research (CIHR), the Howard Hughes Medical Institute (HHMI, Infectious Diseases and Parasitology Program), the Canadian Genetic Diseases Network and the Mathematics of Information Technology and Complex System Network (Networks of Centres of Excellence Program). JC is the recipient of a CIHR fellowship. JCL-O is a CIHR Strategic Training Fellow in Infectious Diseases and Autoimmunity. DM is a scholar of CIHR and an International Research Scholar of the HHMI.

Author information



Corresponding author

Correspondence to D Malo.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Caron, J., Loredo-Osti, J., Morgan, K. et al. Mapping of interactions and mouse congenic strains identified novel epistatic QTLs controlling the persistence of Salmonella Enteritidis in mice. Genes Immun 6, 500–508 (2005).

Download citation


  • QTL mapping
  • epistasis
  • congenic strains
  • Salmonella

Further reading


Quick links