Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 May 18.
Published in final edited form as: J Infect Dis. 2009 Feb 1;199(3):355–361. doi: 10.1086/595987

Quantitative Trait Loci Associated with Susceptibility to Escherichia coli Bladder and Kidney Infections in C3H/HeJ Female Mice

Walter J Hopkins 1, Johny Elkahwaji 2, Christina Kendziorski 3, Amy R Moser 4, Paulette M Briggs 5, Kaleigh A Suhs 6
PMCID: PMC2683358  NIHMSID: NIHMS104359  PMID: 19061424

Abstract

Background

The C3H/HeJ strain of mice develop severe bladder and kidney infections after intravesical inoculation with uropathogenic E. coli. This susceptibility is genetically determined, but the specific genes involved have not been completely defined. The objective of the current study was to use quantitative trait locus (QTL) mapping to identify chromosomal sites in C3H/HeJ mice associated with infection susceptibility.

Methods

Female mice from a backcross of C3H/HeJ and (BALB/c × C3H/HeJ)F1 mice were inoculated with E. coli, and the number of E. coli present in the bladder and kidneys was quantified ten days later. Genomic DNA was scanned using microsatellite markers to localize chromosomal segments derived from parental strains. Statistical analyses associated infection phenotypes with chromosomal sites.

Results

A highly significant QTL for bladder infection susceptibility was identified on chromosome 4, and C3H/HeJ alleles at this locus interacted with BALB/c alleles on chromosome 19 to increase infection severity. A significant QTL on chromosome 6 was associated with severe kidney infections.

Conclusions

Increased susceptibility to E. coli bladder and kidney infections in female C3H/HeJ mice is associated with specific chromosomal sites located near genes contributing to host resistance to infection. The results demonstrate the multigenic nature of UTI susceptibility.

Keywords: E. coli, mouse, urinary tract infection, Tlr4, lipopolysaccharide, Toll-like receptor, genetic, QTL, mapping

Introduction

Urinary tract infections are estimated to occur in up to 50% of all women in their lifetimes, and 10-15% of women experience recurrent infections that require ongoing medical care and may lead to pyelonephritis and renal scarring [1-4]. The exact reason for the increased susceptibility of UTI-prone women is largely unknown. A genetic predisposition is suggested by the observations that recurrent UTIs occur in a limited segment of the female population and are seen more often within some families [5,6].

Our evaluation of candidate genes in women with recurrent UTIs did not detect disproportionate frequencies of major histocompatibility complex or red blood cell antigen phenotypes [7]; however, two recent reports have found a genetic basis for asymptomatic bacteriuria in children [8] and susceptibility to acute pyelonephritis [9]. Predisposition to asymptomatic bacteriuria was associated with reduced expression of the Toll-like receptor 4 (Tlr4) on neutrophils and acute pyelonephritis with polymorphisms and mutations in the CXCR1 gene. While these findings have identified correlations between individual genes and infections in different organs, it is likely that, overall, increased susceptibility to UTIs is complex and multigenic.

Animal models of UTI have been valuable in defining host genetic factors contributing to both host resistance and increased susceptibility to infection. In particular, mouse models have shown that genetic background strongly influences an animal’s ability to resolve induced E. coli bladder and kidney infections [10]. Female mice from the BALB/c, DBA, C3H/HeN, and C57Bl/6 inbred strains effectively clear bladder and kidney infections within two weeks after inoculation with uropathogenic E. coli and are resistant to kidney infections. In contrast, C3H/HeJ and C3H/HeOuJ mice develop bladder infections that are initially comparable in intensity to those in resistant strains but cannot be resolved over a two-week period. These two susceptible strains also develop severe, persistent kidney infections. Because each inbred strain is genetically distinct, the differential bladder and kidney infection phenotypes observed in them must be attributed to differences in the unique genetic background of each strain.

We have been particularly interested in the genetic basis of the increased susceptibility to E. coli bladder and kidney infections in C3H/HeJ and C3H/HeOuJ mice. Mice from the C3H/HeJ strain carry a mutation in the Tlr4 gene that makes them unresponsive to the biological effects of bacterial lipopolysaccharide (LPS) [11], and the increased susceptibility of C3H/HeJ mice to E. coli UTIs has been attributed to the absence of Tlr4-mediated inflammatory responses [12-14]. This model, however, does not account for the observation that C3H/HeOuJ mice, which have a normal Tlr4 gene and LPS responsiveness, develop E. coli bladder and kidney infections equivalent in severity and duration to those of C3H/HeJ mice. Possible explanations for these results are that genes other than Tlr4 play a significant role in UTI susceptibility or a gene closely linked to, and segregating with, Tlr4 decreases host resistance to UTI. While there is a clear relationship between Tlr4 deficiency and severe E. coli bladder and kidney infections in C3H/HeJ mice, a model based on multiple genes is more consistent with other infectious disease models and the infection data from C3H/HeOuJ mice [15]. Thus, the primary objective of the current study is to identify chromosomal sites in C3H/HeJ mice associated with increased susceptibility to induced E. coli UTIs.

One approach to identifying genes associated with a specific disease is to evaluate the effects of candidate genes on one or more phenotypes related to that disease. While this strategy has been applied successfully in demonstrating genetic predisposition to some infectious diseases, it is unlikely that all genes contributing to a phenotype can be determined a priori. In addition, statistical analyses must be adequately corrected for multiple comparisons before any positive results can be viewed with certainty [16]. A preferable, unbiased method of identifying genetic loci contributing to infection susceptibility is by genetic linkage analysis study where a correlation can be determined between a disease phenotype and genotypes at numerous chromosomal sites [17-19]. Quantitative trait locus (QTL) mapping is an established method to map genetic loci that play a significant role in predisposition to, and defense against, infectious diseases in animal models [20-25].

The QTL mapping approach is well-suited to explore the genetic basis of susceptibility to E. coli bladder and kidney infections in C3H/HeJ mice for several reasons. First, our previous studies on the inheritance of UTI susceptibility in C3H/HeJ mice demonstrated that susceptibility to bladder and kidney infections were genetically distinct, recessive traits and that multiple genes likely contributed to the infection phenotypes [26]. Second, the bladder and kidney infection intensities observed in susceptible and resistant mouse strains are quantitative phenotypes with widely separated numerical values. Third, results of QTL mapping can be used in statistical models to discover interacting loci contributing to infection susceptibility. Thus, QTL mapping in this disease model has the potential to identify both the main effect and interacting loci associated with E. coli bladder and kidney infections in C3H/HeJ mice as well as to provide insight into novel genetic pathways not previously suspected of playing a role in disease susceptibility.

The genetic linkage analysis conducted in the current study revealed a highly significant QTL associated with susceptibility to E. coli bladder infection on chromosome 4 near the Tlr4 locus. This QTL interacts with another locus on chromosome 19 to increase the severity of bladder infections. A single, significant QTL associated with susceptibility to E. coli kidney infection was located on chromosome 6 and did not appear to interact with other loci. These results demonstrate the multigenic and complex nature of UTI susceptibility in C3H/HeJ mice.

Materials and Methods

Animals

Male BALB/c mice were purchased from Harlan Sprague Dawley (Indianapolis, IN), and C3H/HeJ females were supplied by Jackson Laboratories (Bar Harbor, ME). Female mice used in this study were bred in our animal facility from a backcross of C3H/HeJ females with (BALB/c × C3H/HeJ)F1 males. Animals were housed in accordance with guidelines of the Association for the Assessment and Accreditation of Laboratory Animal Care, and all protocols were approved by the University of Wisconsin Animal Care and Use Committee.

Infection phenotype determination

Bladder and kidney infections were induced in 154 female backcross mice by intravesical inoculation with uropathogenic E. coli [27,28]. Escherichia coli strain 1677 was grown from frozen stock by two passages in tryptose broth (Difco Laboratories; Detroit, MI), washed with PBS by centrifugation, and resuspended to a concentration of 2 × 1010/mL. Mice were deprived of water for one hour and had urine expressed from their bladders immediately prior to inoculation. Ten microliters of bacterial inoculum were instilled into the bladder by urethral catheterization under isoflurane anesthesia, resulting in a dose of 2 × 108 E. coli per mouse. The animals were allowed to recover from anesthesia and given free access to water one hour later.

Mice were sacrificed ten days after inoculation to assess intensities of bladder and kidney infections. The organs were removed, weighed, and homogenized in sterile PBS, after which the homogenates were serially plated onto Levine’s EMB agar (Difco Laboratories; Detroit, MI). The number of E. coli colony-forming units (CFU) on each plate was determined after overnight incubation at 37° C. The total number of CFUs in each bladder and pair of kidneys was normalized by weight as done previously (26) and was used as the quantitative phenotype for bladder (BLCFU) and kidney (KDCFU) infections, respectively.

Genotype determination

Genomic DNA was prepared from the spleen of each backcross mouse using the Puregene Tissue Kit (Gentra, Minneapolis, MN) and used in PCR reactions to determine the parental genotype at DNA microsatellite markers spaced at a minimum distance of 20 cM on each chromosome. For this study, the 19 autosomal chromosomes were evaluated. The X chromosome was not genotyped since the backcross mice were bred from a female C3H/HeJ parent and a male F1, which would not have recombinations in the X chromosome, making backcross females homozygous for C3H/HeJ alleles and not informative.

The PCR reactions used microsatellite primer pairs (Invitrogen, Carlsbad, CA or Integrated DNA Technologies, Coralville, IA) previously tested to insure size polymorphisms in the products generated from each parental DNA. Cycling conditions were 96°C for 2 min, 30 cycles each of 94°C for 45 sec, 57°C for 45 sec, and 72°C for 60 sec, followed by a final extension at 72°C for 7 min. Size polymorphisms in PCR products were determined by agarose gel electrophoresis.

Statistical analyses

The phenotype distributions were skewed toward higher values, so log-transformed phenotypes were considered. The log-transformation attenuated the effect of outliers on the mean but did not result in a phenotype distribution that approximated normal for the bladder or kidney phenotypes. The effects of square- or cube-root transformations were also tested, and did not result in an approximately normal distribution of phenotypes. For this reason, traditional mapping approaches that assume normally distributed phenotypes could not be used. Instead, we used the non-parametric (NP) mapping approach developed by Kruglyak and Lander [29] and implemented in R/QTL [30] (denoted as R/QTL-NP) to calculate the log10 of the odds (LOD) scores for associations between BLCFU or KDCFU phenotypes and specific chromosomal sites. Permutation tests were performed (10,000 permutations per phenotype) to determine thresholds for suggestive (P < 0.05) and significant (P < 0.01) linkage at the genome level [31].

To assess the presence of interactions among QTL, we first considered a model selection procedure as detailed in Lan et al. [32]. This procedure identifies potential interactions among significant QTL and also allows for the possibility that significant QTL interact with other genome regions that do not show significant main effects. The procedure first identifies putative QTL using a LOD score profile. In the initial step, we do not require that putative QTL be statistically significant; they are defined as those with maximum LOD score (one per chromosome) where LOD scores are calculated using R/QTL-NP. We then consider all possible models allowing for additive effects among the putative QTL and pairwise interactions. The Bayes Information Criterion (BIC) is used to score each model [33]. The BIC balances goodness of model fit with the number of model parameters. The model with the best (lowest) BIC is then identified. We note that this procedure requires that some parametric specification be made. We conducted the analysis twice, once under the assumption of a Gaussian distributed phenotype and then again under the assumption of a Poisson distributed phenotype. The results were robust to these assumptions, and p-values associated with the Gaussian model are reported.

Results

Main effect QTL for bladder and kidney infection susceptibility

A genome scan of backcross mice was conducted using DNA microsatellite size polymorphisms to identify chromosomal segments derived from each parental strain. Statistical analyses described in Methods were used to associate bladder and kidney phenotypes with marker genotype for chromosomes 1 through 19. Figure 1 shows the LOD scores calculated by interval mapping using markers at a density of at least 20 cM on each chromosome. There was a single, highly significant QTL on chromosome 4 at 29.0 cM with a LOD score of 4.91 (P < 0.001). This QTL has been named Becis1 for bladder E. coli infection susceptibility. The LOD scores for all other markers were at or below 1.00.

Figure 1.

Figure 1

Open in a new tab

Genome scan of female backcross mice for chromosomal sites associated with bladder infection susceptibility. A total of 154 mice from a backcross between C3H/HeJ and (BALB/c × C3H/HeJ)F1 were inoculated with E. coli strain 1677 and assayed for E. coli CFU in the bladder (BLCFU) 10 days later. Log10 of the odds (LOD) scores are shown, with thresholds for statistical significance indicated by dashed lines.

A single, statistically significant QTL for kidney infection phenotype was identified at 17.0 cM on chromosome 6 by a similar statistical analysis (Figure 2). This QTL had a LOD score of 3.27 (P < 0.01). We shall refer to this locus as Kecis1 for kidney E. coli infection susceptibility. The second highest QTL (not on chromosome 6) was on chromosome 4 at 5.0 cM with a LOD score of 2.10 (P = 0.15).

Figure 2.

Figure 2

Open in a new tab

Genome scan of female backcross mice for chromosomal sites associated with kidney infection susceptibility. A total of 154 mice from a backcross between C3H/HeJ and (BALB/c × C3H/HeJ)F1 were inoculated with E. coli strain 1677 and assayed for E. coli CFU in the kidneys (KDCFU) 10 days later. Log10 of the odds (LOD) scores are shown, with thresholds for statistical significance indicated by dashed lines.

Interacting QTL for bladder and kidney infection susceptibility

Our previous investigations on the genetics of susceptibility to E. coli bladder and kidney infections indicated that each of these recessive traits was determined by more than one gene [26]. The traditional mapping approach used here identified one highly significant QTL for susceptibility to E. coli bladder infection and one QTL for susceptibility to E. coli kidney infection; however, the analysis did not consider the possibility of interacting loci contributing to the infection phenotypes. As a second analysis, we used the BIC approach to evaluate models of interactions between the main effect QTL and other loci.

The BIC procedure confirmed a model for bladder infection susceptibility where the main effect was associated with C3H/HeJ alleles near marker D4Mit84. In addition, the procedure identified a model for bladder infection susceptibility in which there was an interaction between D4Mit84 at 37.7 cM, the marker closest to Becis1, and marker D19Mit69 located at 6.0 cM on chromosome 19. We refer to this locus as I-Becis1. The p-values in the identified regression model for Becis1, I-Becis1, and the interaction term are <0.001, 0.06, and 0.006, respectively. As detailed in Methods, the BIC approach can identify genomic regions that are not significant on their own, but show significant interactions. That was the case for I-Becis1. The data in Table 1 demonstrate the effect of the Becis1 and I-Becis1 interaction on the intensity of E. coli bladder infection in backcross mice.

Table 1.

Bladder infection intensities in mice with different combinations of BALB/c and C3H/HeJ alleles at Becis1 and I-Becis1

D4Mit84 (Becis1)
C3H/C3Ha
 BALB/C3Hb
D19Mit69 (I-Becis1) C3H/C3Ha 5.16 ± 0.44 (174.2) nd = 45 4.81 ± 0.54 (122.7)c n = 41
BALB/C3Hb 6.67 ± 0.69 (788.4) n = 35 3.23 ± 0.38 (25.2) n = 31
Open in a new tab
a

Mice had only C3H/HeJ alleles at the D4Mit84 or D19Mit69 loci.

b

Mice had both BALB/c and C3H/HeJ alleles at the D4Mit84 or D19Mit69 loci.

c

Mean loge-transformed BLCFU +/- SEM (antilog of transformed mean)

d

Number of animals in group. A total of 152 animals were analyzed out of a possible 154, due to two animals missing genotypes at one or more markers considered.

As shown in Table 1, when backcross mice were heterozygous for both BALB/c and C3H/HeJ alleles near microsatellite markers D4Mit84 (Becis1) and D19Mit69 (I-Becis1), their mean (log-transformed) BLCFU phenotype was 3.23. The phenotype increased to 4.81 when D4Mit84 remained heterozygous and the D19Mit69 marker was homozygous for C3H/HeJ alleles, indicating an increase in susceptibility from the effects of C3H/HeJ alleles. These infection intensities were lower than those observed for mice homozygous for C3H/HeJ alleles at both D4Mit84 and D19Mit69 (BLCFU = 5.16).

There was no demonstrable effect on the phenotype when the D4Mit84 marker was changed from heterozygous for C3H/HeJ and BALB/c alleles (BLCFU = 4.81) to C3H/HeJ homozygous (BLCFU = 5.16) while the D19Mit69 locus remained homozygous for C3H/HeJ alleles. There was, however, a large change in mean BLCFU when the D19Mit69 marker was changed from homozygous C3H/HeJ to heterozygous for both BALB/c and C3H/HeJ alleles and the D4Mit84 marker remained homozygous C3H/HeJ (BLCFU increased from 5.16 to 6.67, respectively). The implication of these results is that the levels of bladder infection seen in C3H/HeJ mice can be further increased by the addition of BALB/c alleles at the I-Becis1 locus on chromosome 19.

Discussion

The objective of the study reported here was to perform a genetic linkage analysis of a cross between C3H/HeJ and BALB/c mice to localize genes associated with the increased susceptibility of C3H/HeJ female mice to induced E. coli bladder and kidney infections. Our previous analysis of bladder and kidney infection phenotypes of a backcross between (BALB/c × C3H/HeJ)F1 and C3H/HeJ mice demonstrated that susceptibility to E. coli infection of either organ was multigenic; however, that study could not identify the specific genetic loci involved. We have now identified specific QTLs linked to severe E. coli bladder and kidney infections in C3H/HeJ female mice.

Increased susceptibility to bladder infections by uropathogenic E. coli is associated with a single, strong QTL, Becis1, at 29.0 cM on chromosome 4. At this time it is not possible to specify exactly which gene in the vicinity of Becis1 directly promotes intense, protracted colonization of the bladder because the markers used for screening were spaced at approximately 20 cM intervals along the chromosome. Nevertheless, it is noteworthy that the Tlr4 gene is located at 33.0 cM on chromosome 4 and may be a candidate for Becis1 within the limitations of the analysis performed here. Mice from the C3H/HeJ strain carry a point mutation in the Tlr4 gene that makes them unresponsive to the biological effects of E. coli LPS [11-13], and the lowered resistance of these mice to E. coli UTI has been attributed to their inability to develop local inflammatory responses initiated by the interaction of LPS with Tlr4 receptors on bladder epithelial cells [14]. While Tlr4 plays a central role in this model of host resistance to infection, we have also demonstrated that C3H/HeOuJ female mice are highly susceptible to both bladder and kidney infections caused by E. coli, even though the C3H/HeOuJ mice have a normal Tlr4 gene and are responsive to LPS from E. coli and other gram-negative bacteria [15].

Because the C3H/HeOuJ and C3H/HeJ strains were derived from a common ancestor but likely diverged genetically while bred separately for several decades, it is conceivable that alleles of one or more genes closely linked to the Tlr4 locus in both strains, rather than Tlr4 alone, may be responsible for the severe bladder infections induced by E. coli. The importance of Tlr4 versus another gene, or genes, on chromosome 4 could potentially be determined by analyzing additional animals bred from the current cross, screening a similar cross using C3H/HeOuJ mice, or sequencing chromosome 4 from Becis1 through Tlr4 to identify nucleotide variations.

In our previous studies of bladder infection phenotypes of backcross mice, the Wright-Castle analysis concluded that multiple genes contributed to the high susceptibility phenotype [26]. The current results provide experimental evidence supporting this prediction. Statistical modeling of the main effect QTL (Becis1) using the nearest microsatellite marker (D4Mit84) on chromosome 4 revealed an interacting QTL (I-Becis1) near marker D19Mit69 on chromosome19. Interestingly, the greatest increase in bladder infection intensity was seen in backcross mice that were homozygous C3H/HeJ at D4Mit84 and heterozygous at D19Mit69. One implication of these results is that the severity of bladder infection can be increased by the addition of BALB/c alleles at I-Becis1, suggesting the presence in BALB/c mice of previously unrecognized susceptibility-associated alleles that are only observed in combination with the C3H/HeJ alleles on chromosome 4.

In contrast to the results for QTLs associated with severe E. coli bladder infection, increased susceptibility to E. coli kidney infections has thus far been linked to a single QTL in backcross mice, Kecis1 on chromosome 6 at 17.0 cM. This QTL on chromosome 6 is located near genes coding for variable regions of the β chain of the T cell antigen receptor at 20.5 cM. By analyzing phenotype and genotype data from additional mice bred by the current backcross mating scheme and by increasing the marker density around Kecis1, it may be possible to determine if Kecis1 resides among the T cell receptor genes. If this is the case, there would be a strong indication that host defense against E. coli kidney infection is largely dependent on T cell recognition of antigens on uropathogenic E. coli and subsequent T cell responses to those antigens. Additional studies could investigate a potential immune deficiency of C3H/HeJ mice to specific E. coli antigens and the relative importance of their CD4+ and CD8+ T cell responses to E. coli. It can also be noted that genes for immunoglobulin kappa chain variable regions are located on chromosome 6 at 30.0 cm; however, the current scan did not place any QTLs in this region even with a microsatellite marker at 34.0 cM. Thus, host resistance to ascending E. coli kidney infection appears to be more dependent on T-cell participation in immune responses than antibody production by B cells.

The multigenic nature of increased susceptibility to E. coli kidney infections in C3H/HeJ mice was implied from our previous studies [26], but statistical modeling of the current data did not detect any QTLs interacting with Kecis1. There were, however, distinct LOD scores peaks below the statistical significance level on chromosomes 1, 4, and 9. It is interesting to note that Tlr5-/- mice have increased susceptibility to E. coli UTI [34] and the antimicrobial peptide, cathelicidin, plays a role in resistance to UTI in humans and mice [35]. The Tlr5 and cathelicidin genes are located on mouse chromosomes 1 at 98.0 cM and 9 at 61.0 cM, respectively, in the vicinity of the LOD score peaks for these chromosomes. The increased LOD score for kidney infections on chromosome 4 was at approximately 10 cM, which is not close to Becis1 or Tlr4. Whether alleles at any of these chromosomal sites contribute significantly to increased kidney infection susceptibility could be determined by analysis of additional animals bred by the current backcross mating scheme.

We can draw several important conclusions from the current mapping study. One of the most significant is that QTL analysis has been successful in localizing chromosomal sites where allelic differences in genes present in C3H/HeJ and BALB/c mice strongly affect susceptibility to E. coli bladder and kidney infections. These two distinct QTLs are located on different chromosomes, indicating that susceptibility is complex and determined by polymorphisms in alleles of multiple host genes. Furthermore, we now have strong evidence that host factors affecting E. coli colonization and defense mechanisms are clearly different for the bladder and kidney. This finding may have been anticipated because the bladder is subject to mucosal colonization by bacteria and protected by mucosal defense mechanisms, whereas the kidney is a more vascularized organ where systemic immune responses may be more effective. Based on these results, it will be possible to propose and test new hypotheses for genetic predisposition and host resistance/susceptibility to bladder and kidney infection in both animal models and UTI-susceptible patients.

Acknowledgments

This work was supported by grants DK44378 and DK61574 from the National Institutes of Health.

Footnotes

Portions of this work were presented at the 2003 American Urological Association Annual Meeting in Chicago, IL.

Conflict of interest statements:
  • Walter J. Hopkins - No conflict
  • Johny Elkahwaji - No conflict
  • Christina M. Kendziorski - No conflict
  • Amy R. Moser - No conflict
  • Paulette M. Briggs - No conflict
  • Kaleigh A. Suhs - No conflict

Contributor Information

Walter J. Hopkins, Department of Urology, University of Wisconsin School of Medicine and Public Health

Johny Elkahwaji, Department of Surgery, University of Nebraska Medical Center.

Christina Kendziorski, Department of Biostatistics and Medical Informatics, University of Wisconsin School of Medicine and Public Health.

Amy R. Moser, Department of Human Oncology, University of Wisconsin School of Medicine and Public Health

Paulette M. Briggs, Department of Urology, University of Wisconsin School of Medicine and Public Health

Kaleigh A. Suhs, Department of Urology, University of Wisconsin School of Medicine and Public Health

References

  • 1.Kunin CM. Urinary tract infections in females. Clin Infect Dis. 1994;18:1–12. doi: 10.1093/clinids/18.1.1. [DOI] [PubMed] [Google Scholar]
  • 2.Johnson JR, Stamm WE. Urinary tract infections in women: Diagnosis and treatment. Ann Intern Med. 1989;111:906–17. doi: 10.7326/0003-4819-111-11-906. [DOI] [PubMed] [Google Scholar]
  • 3.Foxman B. Recurring urinary tract infections: incidence and risk factors. Am J Public Health. 1990;80:331–3. doi: 10.2105/ajph.80.3.331. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Foxman B, Brown P. Epidemiology of urinary tract infections: transmission and risk factors, incidence, and costs. Infect Dis Clin North Amer. 2003;17:227–41. doi: 10.1016/s0891-5520(03)00005-9. [DOI] [PubMed] [Google Scholar]
  • 5.Hopkins WJ, Uehling DT, Wargowski DS. Evaluation of a familial predisposition to recurrent urinary tract infections in women. Am J Med Genet. 1999;83:422–4. [PubMed] [Google Scholar]
  • 6.Lundstedt AC, Leijonhufvud I, Ragnarsdottir B, et al. Inherited susceptibility to acute pyelonephritis: a family study of urinary tract infection. J Infect Dis. 2007;195:1227–34. doi: 10.1086/512620. [DOI] [PubMed] [Google Scholar]
  • 7.Hopkins WJ, Heisey DM, Lorentzen DF, Uehling DT. A comparative study of major histocompatibility and red blood cell antigen phenotypes as risk factors for recurrent urinary tract infections in women. J Infect Dis. 1998;177:1296–1301. doi: 10.1086/515283. [DOI] [PubMed] [Google Scholar]
  • 8.Ragnarsdottir B, Samuelsson M, Gustafsson MCU, Leijonhufvud I, Karpman D, Svanborg C. Reduced toll-like receptor 4 expression in children with asymptomatic bacteriuria. J Infect Dis. 2007;196:475–84. doi: 10.1086/518893. [DOI] [PubMed] [Google Scholar]
  • 9.Lundstedt AC, McCarthy S, Gustafsson MCU, et al. A genetic basis of susceptibility to acute pyelonephritis. PloS ONE. 2(9):e825. doi: 10.1371/journal.pone.0000825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Hopkins WJ, Gendron-Fitzpatrick A, Balish E, Uehling DT. Time course and host responses to Escherichia coli urinary tract infection in genetically distinct mouse strains. Infect Immun. 1998;66:2798–2802. doi: 10.1128/iai.66.6.2798-2802.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Beutler B, Poltorak A. The sole gateway to endotoxin response: how lps was identified as Tlr4, and its role in innate immunity. Drug Metab Dispos. 2001;29:474–8. [PubMed] [Google Scholar]
  • 12.Hagberg L, Hull R, Hull S, McGhee JR, Michalek SM, Svanborg-Eden C. Difference in susceptibility to gram-negative urinary tract infection between C3H/HeJ and C3H/HeN mice. Infect Immun. 1984;46:839–44. doi: 10.1128/iai.46.3.839-844.1984. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Shahin RD, Engberg I, Hagberg L, Svanborg-Eden C. Neutrophil recruitment and bacterial clearance correlated with LPS responsiveness in local gram-negative infection. J Immunol. 1987;138:3475–80. [PubMed] [Google Scholar]
  • 14.Haraoka M, Hang L, Frendeus B, Godaly G, Burdick M, Strieter R, Svanborg C. Neutrophil recruitment and resistance to urinary tract infection. J Infect Dis. 1999;180:1220–9. doi: 10.1086/315006. [DOI] [PubMed] [Google Scholar]
  • 15.Hopkins WJ, Gendron-Fitzpatrick A, McCarthy DO, Haine JE, Uehling DT. Lipopolysaccharide-responder and non-responder C3H mouse strains are equally susceptible to an induced Escherichia coli urinary tract infection. Infect Immun. 1996;64:1369–72. doi: 10.1128/iai.64.4.1369-1372.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Marquet S, Schurr E. Genetics of susceptibility to infectious diseases: tuberculosis and leprosy as examples. Drug Metab Dispos. 2001;29:479–83. [PubMed] [Google Scholar]
  • 17.Silver LM. Mapping in the Mouse: an overview. In: Silver LM, editor. Mouse Genetics: Concepts and Applications. New York: Oxford University Press; 1995. pp. 133–58. [Google Scholar]
  • 18.Lengeling A, Pfeffer K, Balling R. The battle of two genomes: genetics of bacterial/host interactions in mice. Mamm Genome. 2001;12:261–71. doi: 10.1007/s003350040001. [DOI] [PubMed] [Google Scholar]
  • 19.Churchill GA, Doerge RW. Emperical threshold values for quantitative trait mapping. Genetics. 1994;138:963–71. doi: 10.1093/genetics/138.3.963. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Dietrich WF. Using mouse genetics to understand infectious diseases. Genome Res. 2001;11:325–31. doi: 10.1101/gr.173101. [DOI] [PubMed] [Google Scholar]
  • 21.Georges M. QTL mapping to QTL cloning: mice to the rescue. Genome Res. 1997;7:663–5. doi: 10.1101/gr.7.7.663. [DOI] [PubMed] [Google Scholar]
  • 22.Mitos LM, Cardon LR, Fortin A, et al. Genetic control of susceptibility to infection with Mycobacterium tuberculosis in mice. Gene and Immun. 2000;1:467–77. doi: 10.1038/sj.gene.6363712. [DOI] [PubMed] [Google Scholar]
  • 23.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;47:180–6. doi: 10.1006/geno.1997.5116. [DOI] [PubMed] [Google Scholar]
  • 24.Roper RJ, Weis JJ, McCracken BA, et al. Genetic control of susceptibility to experimental Lyme arthritis is polygenic and exhibits consistent linkage to multiple loci on chromosome 5 in four independent mouse crosses. Genes and Immun. 2001;2:388–97. doi: 10.1038/sj.gene.6363801. [DOI] [PubMed] [Google Scholar]
  • 25.Weis JJ, McCracken BA, Ma Y, et al. Identification of quantitative trait loci governing arthritis severity and humoral responses in the murine model of Lyme disease. J Immunol. 1999;162:948–56. [PubMed] [Google Scholar]
  • 26.Hopkins WJ, Elkahwaji JE, Heisey DM, Ott CJ. Inheritance of susceptibility to induced Escherichia coli bladder and kidney infections in female C3H/HeJ mice. J Infect Dis. 2003;187:418–23. doi: 10.1086/367963. [DOI] [PubMed] [Google Scholar]
  • 27.Hopkins WJ. Mouse Model of Ascending Urinary Tract Infection. In: Zak O, Sande MA, editors. Handbook of Animal Models of Infection. London: Academic Press; 1999. pp. 435–9. [Google Scholar]
  • 28.Hopkins WJ, Hall JA, Conway BP, Uehling DT. Induction of urinary tract infection by intraurethral inoculation with Escherichia coli: Refining the murine model. J Infect Dis. 1995;171:462–5. doi: 10.1093/infdis/171.2.462. [DOI] [PubMed] [Google Scholar]
  • 29.Kruglyak L, Lander ES. A nonparametric approach for mapping quantitative trait loci. Genetics. 1995;139:1421–8. doi: 10.1093/genetics/139.3.1421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Broman KW, Wu H, Sen S, Churchill GA. R/qtl: QTL mapping in experimental crosses. Bioinformatics. 2003;19:889–90. doi: 10.1093/bioinformatics/btg112. [DOI] [PubMed] [Google Scholar]
  • 31.Churchill GA, Doerge RW. Empirical threshold values for quantitative trait mapping. Genetics. 1994;138:663–71. doi: 10.1093/genetics/138.3.963. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Lan H, Kendziorski CM, Shepel LA, Haag JD, Newton MA, Gould MN. Genetic loci controlling breast cancer susceptibility in the Wistar-Kyoto rat. Genetics. 2001;157:331–9. doi: 10.1093/genetics/157.1.331. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Schwarz G. Estimating the dimension of a model. Ann Stat. 1978;6:461–4. [Google Scholar]
  • 34.Andersen-Nissen E, Hawn TR, Smith KD, et al. Tlr5-/- mice are more susceptible to Escherichia coli urinary tract infection. J Immunol. 2007;178:4717–20. doi: 10.4049/jimmunol.178.8.4717. [DOI] [PubMed] [Google Scholar]
  • 35.Chomek M, Slamova Z, Bergman P, et al. The antimicrobial peptide cathelicidin protects the urinary tract against invasive bacterial infection. Nature Med. 2006;12:636–41. doi: 10.1038/nm1407. [DOI] [PubMed] [Google Scholar]

RESOURCES