After intracerebral inoculation, Theiler's virus induces in its natural host, the mouse, an acute encephalomyelitis followed, in susceptible animals, by chronic inflammation and primary demyelination. Susceptibility to demyelination among strains of laboratory mice is explained by the capacity of the immune system to control viral load during persistence. Also, differences of susceptibility to viral load between the susceptible SJL strain and the resistant B10.S strain are mainly due to two loci, Tmevp2 and Tmevp3, located close to the Ifng locus on chromosome 10. In this article, we show that the Tmevp3 locus controls both mortality during the acute encephalomyelitis and viral load during persistence. Most probably, two genes located in the Tmevp3 interval control these two different phenotypes with efficiencies that depend on the age of the mouse at inoculation. Il22, a member of the IL-10 cytokine family, is a candidate gene for the control of mortality during the acute encephalomyelitis.

---

THEILER'S virus is a picornavirus whose natural host is the mouse. After intracerebral inoculation, Theiler's virus induces an acute encephalomyelitis in all strains of laboratory mice. The virus is cleared from the gray matter of the brain and the spinal cord after 2 weeks regardless of the mouse genotype. Some mice suffer sequelae such as flaccid paralysis of hind legs. In susceptible mice, the virus persists in the white matter of the spinal cord for the lifetime of the animal inducing chronic inflammation and primary demyelination (BRAHIC et al. 2005; LIPTON et al. 2005); spastic paralysis is common. Lesions are very similar to those of the human disease multiple sclerosis. During persistence there is a great variability of viral load, measured as the quantity of viral genomes present in the spinal cord, among strains of laboratory mice (BUREAU et al. 1992). A major locus controlling viral load, the H2-D class I gene, has been characterized on chromosome 17 (AZOULAY et al. 1994; AZOULAY-CAYLA et al. 2000; AUBAGNAC et al. 2001; MENDEZ-FERNANDEZ et al. 2005). However during persistence, the SJL strain is infected at a higher level than the B10.S strain, although both bear the same H2^s haplotype. Two non-H2 loci controlling viral load, Tmevp2 and Tmevp3, have been located on chromosome 10 close to the Ifng locus by studying an (SJL/J x B10.S)F₁ x B10.S backcross and SJL lines congenic for different B10.S genetic intervals of chromosome 10 (BUREAU et al. 1993; BIHL et al. 1999). The B10.S allele of the Tmevp3 locus confers resistance to viral load. The Tmevp3 locus has been finely mapped. The smallest B10.S Tmevp3 interval among those of the different SJL congenic strains is borne by the SJL. Tmevp3^B10S line and is 4.1 Mb long. A cluster of cytokines with at least three genes—Il22, Ifng, and Tmevpg1—is located in the most telomeric part of this interval just telomeric to the Mdm1 gene. This gene, which codes for a putative transcription factor, is amplified in a 3T3 transformed cell line (SNYDER et al. 1988). The Il22 gene, which is a member of the IL-10 cytokine family, is expressed predominantly by memory CD4⁺ Th17 lymphocytes and is involved with innate immunity (CHUNG et al. 2006; LIANG et al. 2006; NAGEM et al. 2006; ZHENG et al. 2007). The Ifng gene has well known antiviral and immunoregulatory activities (VAN DEN BROEK et al. 1995; NOVELLI and CASANOVA 2004; HARARI et al. 2006). The Tmevpg1 gene codes for a noncoding mRNA expressed in nonstimulated CD4⁺, CD8⁺, and NK cells (VIGNEAU et al. 2003).

Surprisingly, study of bone marrow chimeras between SJL.Tmevp3^B10S congenic mice and the parental strains revealed that 13–14-week-old SJL.Tmevp3^B10S mice are infected at a higher level than SJL mice (AUBAGNAC et al. 2002), in contrast to the infection of 3–4-week-old mice in which the SJL.Tmevp3^B10S mice are infected at a lower level than the SJL mice. These opposite phenotypes for mice inoculated at different ages suggest that the effect of the Tmevp3 locus is due to the action of two genes: one responsible for the susceptibility of 3–4-week-old mice, when it bears the SJL allele, and the other responsible for the susceptibility of 13–14-week-old mice, when it bears the B10.S allele. The study of bone marrow chimeras showed that the second gene is expressed in radio-resistant cells. However, in spite of the high number of backcrosses used to develop the SJL.Tmevp3^B10S congenic strain, we cannot formally rule out that another B10.S interval was introgressed into its genome or that an intragenic crossing over occurred. The analysis of a congenic strain developed independently of the SJL.Tmevp3^B10S strain, if it is confirmed that phenotype(s) segregate differently depending on the age of the mouse at the time of inoculation, would exclude this possibility.

In this work, we have developed a new congenic line, B10S.Tmevp3^SJL, with a B10.S background and a Tmevp3^SJL interval. Using the two parental strains and the two congenic lines, B10S.Tmevp3^SJL and SJL.Tmevp3^B10S, the goals of this study were to reduce the interval of the Tmevp3 locus, to define which phenotypes are controlled by the Tmevp3 locus and the number of genes affecting these phenotypes, and, lastly, to find candidate gene(s).

DNA and genotyping:

Eleven segments of genomic DNA from 14 mouse strains (C57BL/6J, FVB/NCrl, SWR/OlaHsd, 129S2/SvPas, DBA/2J, BALBc/ByJ, C3H/HeOuJIco, DBA/1JPas, STF/Pas, NZB/OlaHsd, MAI/Pas, MBT/Pas, SJL/JOrlCrl, and B10.S-H2S/SgMcdJ) have been PCR amplified using 9700 GeneAmp PCR system (Applied Biosystems), cloned in TOPO vector (Invitrogen, Cergy Pontoise, France), and sequenced. These fragments have been amplified from the eight exons of Tmevpg1, intron 1 of Ifng, intron 11 of Mdm1, the intergenic region between Mdm1 and Rap1b, and intron 8 of Mdm2. Sequences from the 11 segments were concatenated in a 3.9-kb sequence that has been used to construct a phylogenetic tree (SCHMIDT et al. 2002; LI 2003). Four polymorphic microsatellites were discovered in the sequence alignment of the Il22 gene and its pseudogene (nos. AJ294727 and AJ294728, respectively). Duplication of this region has been studied in the 14 mouse strains using these four microsatellites and confirmed by amplifying, cloning, and sequencing a region covering the 603-bp deletion found in the promoter and the first exon of the pseudogene. In all cases of duplications, two bands were amplified for the four microsatellites and the 603-bp deletion was recovered. In all the other cases, one band was amplified without recovering the deletion.

Mice and viral infection:

SJL mice were purchased from Charles River (L'Arbresle, France). SJL.Tmevp3^B10S, B10S.Tmevp3^SJL congenic mice, and B10.S-H2S/SgMcdJ mice were bred at the Pasteur Institute animal facility. The SJL.Tmevp3^B10S congenic strain was named SJL.B10.D10Mit180-D10Mit74 in a previous article (BIHL et al. 1999). B10S.Tmevp3^SJL congenic mice are a new congenic strain stabilized after 10 backcrosses toward the B10.S strain and selection of the progeny with the D10Mit68, D10Mit233, D10Mit180, D10Mit237, and D10Mit271 markers. Its exact name is B10.S.SJL.D10Mit233-D10Mit271. The DA strain of Theiler's virus was produced by transfection of BHK-21 cells with the pTM762 plasmid as described elsewhere (MCALLISTER et al. 1989; MICHIELS et al. 1997). Anesthetized mice were inoculated intracerebrally with 10⁴ phage-forming units (PFU) of the DA strain in 40 µl of phosphate-buffered saline. Mice were killed at 7, 21, and 45 days postinoculation (p.i.).

Bone marrow reconstitution:

Three- to four-week-old mice were irradiated with a ¹³⁷Cs source at 9.5 Gy. Mice were reconstituted with syngeneic or allogeneic bone marrow cells that had been harvested from the tibias and femurs of age- and sex-matched mice. For the reconstitution, we used 2.0 x 10⁶–4.0 x 10⁶ bone marrow cells from the donor strain. Reconstituted and 12–14-week-old control mice were anesthetized and inoculated intracerebrally as described above. The efficiency of the reconstitution was assessed with peripheral blood lymphocytes at the time of death for each mouse reconstituted with allogeneic bone marrow cells using PCR with D10Mit180 marker and/or FACScan and anti-Ly-9.1 antibodies as previously described (AUBAGNAC et al. 2002). The degree of chimerism varied from 75 to 100%. The mean of chimerism is >94% in each of the six groups of mice. We also analyzed complete blood counts and the degree of chimerism on CD3⁺, CD4⁺, CD8⁺, and B220⁺ spleen cells with FACScan for reconstituted mice at 10 weeks postreconstitution and control mice of the same age. No difference was detected for all these parameters between these two groups of mice (supplemental Table 1S at http://www.genetics.org/supplemental/).

RT–PCR assays:

Anesthetized mice were perfused through the left ventricle with phosphate-buffered saline. The spinal cord was dissected out and stored at –80°. Total RNAs were extracted using TRI-REAGENT (Molecular Research Center, Cincinnati) and reverse transcribed as previously described (VIGNEAU et al. 2003). Hprt amplification was used to normalize the efficiency of the reverse transcription. Forward and reverse primers were chosen in different exons of the studied gene. Both PCR products of the cDNA to quantify and of the Hprt cDNA were cloned in Bluescript vector. The insert containing both PCR products was digested from the vector using appropriate restriction enzymes, eluted, quantified by spectrophotometry, and 10-fold diluted from 10⁷ to 10² molecules per microliter. This dilution was used to construct a standard curve in each Taqman RT–PCR assay. This curve allowed the determination of the number of copies of the gene to quantify and that of the Hprt gene in each sample. The square root of the ratio of the two quantities was used in the different figures of this article. Real-time PCR was performed using qPCR MasterMix (Eurogentec, Seraing, Belgium) according to the manufacturer conditions on a 7000 sequence detection system (Applied Biosystems). The Hprt primers were forward CTGGTGAAAAGGACCTCTCG, reverse TGAAGTACTCATTATAGTCAAGGGCA, and the probe Yakima Yellow-TGTTGGATACAGGCCAGACTTTGTTGGAT-DDQ1. The TMEV primers were forward GCCGCTCTTCACACCCAT, reverse AGCAGGGCAGAAAGCATCAC, and the probe 6FAM-CGACGTGGTTGGAGAT-DDQ1. The Il22 primers were forward AGAAGGCTGAAGGAGACAGT, reverse GACATAAACAGCAGGTCCAGTT, and the probe 6FAM-AAAAGCTTGGAGAGAGTGGA-DDQ1. The probe specific for the B10.S allele was 6FAM-CAATCGCCTTGATCTC-DDQ1 and that for SJL allele was Yakima Yellow-CCAATCGCTTTGATCTC-DQQ1. Concentration of the primers and the probe were chosen to optimize the amplification and the detection of each PCR product. Each assay was validated only if the slope of the standard curve is between 3.32 and 3.8. Allelic discrimination of Il22 gene used the allelic discrimination technology develops by Applied Biosystems.

Statistical analysis:

The Statview version 4.5 package was used for all the analysis. Each comparison of mRNA expression among the four groups of mice was performed using nonparametric test of Kruskal-Wallis. If this first test was significant, groups were merged according to the allele at the Tmevp3 locus or the origin of the background, and distributions were compared using the nonparametric test of Mann-Whitney. Interactions between the Tmevp3 locus and the background were tested by comparing the distributions of the parental strains and that of the congenic strains. Mortality between two groups of mice was compared using Fisher's exact test. Survival data were analyzed using the nonparametric Kaplan-Meier model with stratification. Differences were considered significant if the probability P < 0.05 (*). Two higher levels of significance were also used: P < 0.01 (**) and P < 0.001 (***). Nonsignificant P-values <0.1 were also reported.

Refining the localization of the Tmevp3 locus:

The smallest Tmevp3^B10S interval among those of the different SJL congenic strains is borne by the SJL.Tmevp3^B10S strain and is 4.1 Mb long (Figure 1). The crossing over defining the telomeric border of this interval has been located at <2 kb from the D10Mit74 marker (Chr10: 117,957,802–117,958,312; UCSC Mouse Feb 2006). The centromeric border is located between the D10Mit233 marker (Chr10: 113,785,306–113,785,435), the last marker with an SJL genotype, and the D10Mit180 marker (Chr10: 117,553,604–117,553,737), the first marker with a B10.S genotype. They are separated by 3.7 Mb in a region poor in polymorphisms between the two parental strains. An extensive search of polymorphic microsatellites and single nucleotide polymorphisms (SNPs) in this region was negative (Table 1). The first marker with a B10.S genotype (Chr10: 117,496,216–117,496,364; UCSC Mouse Feb 2006) most probably defines the centromeric border of the Tmevp3 interval. Interestingly, this region of 400 kb long is rich in polymorphisms between the two parental strains: a polymorphism is found every 100–200 bp. This result suggests that this 400-kb region is most likely to contain the Tmevp3 locus.

View larger version (23K): In this window In a new window Download PPT slide   FIGURE 1.— Genotype of the B10S.Tmevp3SJL and SJL.Tmevp3B10S strains on chromosome 10. Microsatellite markers are listed on the left. Distances are arbitrary and do not reflect physical distances. Solid squares, B10.S genome interval; open squares, SJL genome interval; shaded squares, interval in which the crossing over occurs.

 

View this table: In this window In a new window   TABLE 1 Microsatellite and SNP diversity between the B10S and SJL strains

 
To confirm this hypothesis, the sequence of 11 DNA segments covering the 400-kb region was compared among 14 inbred mouse strains. Eleven were laboratory mouse strains of known susceptibility to Theiler's virus persistent infection, including the B10.S and the SJL strains, and 3 were wild-derived strains (MAI, MBT, STF). A phylogenetic tree of the region was constructed using the TREE-PUZZLE program (Figure 2). The 14 inbred strains are divided in three groups: the first one contains only the STF, a strain from the spretus subspecies; the second one contains 8 strains of laboratory mice and the B10.S strain; and the third one contains 2 wild-derived strains from the Mus musculus musculus subspecies, MAI and MBT, and 2 strains of laboratory mice, SJL and NZB. Therefore, the SJL and NZB strains are the 2 strains of laboratory mice that received the entire 400-kb region from a mouse of the M. m. musculus subspecies. The 9 other strains of laboratory mice received this region from a mouse of the M. m. domesticus subspecies (data not shown). There is a good correlation between the haplotype of this 400-kb region and the susceptibility to persistent infection by Theiler's virus. The SJL strain is the only strain of laboratory mouse among the 11 studied in which susceptibility to viral load is not predicted by the H2 haplotype, suggesting that it is the only strain of laboratory mice bearing a susceptible allele at the Tmevp3 locus. The only exception to the correlation between this prediction and the haplotype at the 400-kb region is the NZB strain that bears the same Tmevp3 haplotype as the SJL strain, but for which susceptibility to the persistent infection is predicted by its H2 haplotype. These results confirmed that the Tmevp3 locus is located within this 400-kb region.

View larger version (4K): In this window In a new window Download PPT slide   FIGURE 2.— Phylogenetic tree of 14 inbred strains of mice. From each 11 strains of laboratory mice and 3 wild-derived strains, MAI, MBT, and STF, sequences from 11 DNA segments covering the 400-kb region were concatenated in a 3.9 kb long sequence. These concanated sequences were aligned according to CLUSTALW program. A phylogenetic tree was constructed using the TREE-PUZZLE program with the Hasegawa–Kishino–Yano model of substitution. Parameters were estimated from the data. The tree was drawn by the DRAWTREE program. Italics, strains of laboratory mice in which viral load is predicted by the H2 haplotype; underlined type, laboratory mouse strain in which viral load is not predicted by the H2 haplotype; boldface type, wild-derived strains.

 

Genetic characteristics of the congenic strains:

A new congenic strain, B10S.Tmevp3^SJL, was constructed for this work. The strain was stabilized after 10 backcrosses toward the B10.S parent and selection of the offspring with the D10Mit68, D10Mit233, D10Mit180, D10Mit237, and D10Mit271 markers. Figure 1 shows the genetic map of the telomeric region of chromosome 10 for the two parental strains, the B10S.Tmevp3^SJL strain and the previously described SJL.Tmevp3^B10S strain. Four hundred thirty-three polymorphic SNPs from the MMDAP panels (MORAN et al. 2006) and 61 polymorphic microsatellites were used to verify that the interval of each congenic strain on chromosome 10 was the only one that had been introgressed during backcrossing. Other than the introgressed interval selected for, no other interval was detected in the genome of the two congenic strains.

Susceptibility to persistent infection of mice inoculated at 3–4 weeks of age:

Three- to four-week-old B10.S, SJL, B10S.Tmevp3^SJL, and SJL.Tmevp3 ^B10S mice were inoculated intracerebrally with 10⁴ PFU of Theiler's virus. The mice were killed 45 days later and viral load in the spinal cord was measured using the Taqman real-time RT–PCR assay. Figure 3 shows the result of the experiment. A significant difference of viral load is detected among the four groups of mice (P < 0.01). The viral load depends on the allele of the Tmevp3 locus (P < 0.001) and not on the background (P = NS, nonsignificant difference) or on an interaction between the two (P = NS). Mice with the Tmevp3^SJL allele were infected at a higher level than mice with the Tmevp3^B10.S allele.

View larger version (16K): In this window In a new window Download PPT slide   FIGURE 3.— Viral load in the spinal cords of the four groups of mice inoculated at 3–4 weeks of age, at 45 days post-inoculation. SQRT, square root; rectangle, 90% interval of viral load around median; error bar, 95% interval around median. Viral load among the four groups of mice is significantly different (P < 0.01). Effect of the Tmevp3 locus, the background, and the interaction between both factors are shown below. Significant differences between two groups of mice are shown. P = NS, nonsignificant difference; *P < 0.05; **P < 0.01; ***P < 0.001. n, number of mice.

 

Susceptibility of 3–4-week-old mice to acute encephalomyelitis:

When 3–4-week-old B10.S, SJL, B10S.Tmevp3^SJL, and SJL.Tmevp3 ^B10S mice were inoculated intracerebrally with 10⁴ PFU of Theiler's virus (see previous section), some mortality from acute encephalomyelitis was observed in mice with a B10.S background (18 of 125) whereas no mortality occurred in mice with an SJL background (0 of 104). The difference is highly significant (P < 0.001). Interestingly, mortality was significantly higher in male B10.S mice (9 of 30) than in female B10.S mice (1 of 39; P < 0.01). In contrast, there was no difference of mortality between male and female B10.Tmevp3^SJL mice (4 of 26 for males vs. 4 of 30 for females). These results show a clear effect of the B10.S vs. SJL background on mortality from acute encephalomyelitis. They suggest that, on the B10.S background, the Tmevp3 locus, which was identified because of its effect on viral load during persistent infection, also controls the effect of sex on mortality.

Susceptibility of bone marrow chimeras:

Figure 3 confirms previously published results (BIHL et al. 1999) showing that SJL.Tmevp3^B10S mice are less susceptible to the persistent infection than SJL mice when both strains are inoculated at 3–4 weeks of age. In contrast we showed in another article that SJL.Tmevp3^B10S mice are more susceptible to persistent infection than the SJL parents when the mice are inoculated at 13–14 weeks of age (AUBAGNAC et al. 2002). Study of bone marrow chimeras showed that this difference of viral load depends on the Tmevp3 allele of recipient strains with an SJL background. These opposite phenotypes strongly suggest that at least two genes control this phenotype and that at least one of them is located in the Tmevp3 interval. Therefore, we decided to inoculate 13–14-week-old B10.S and B10S.Tmevp3^SJL mice and bone marrow chimeras to look for a different segregation of mortality during the acute encephalomyelitis and/or viral load during persistence. Since in our previous experiments, the Tmevp3 allele of the recipient conferred the phenotype, this new study used the B10.S and B10S.Tmevp3^SJL mice as bone marrow recipients.

Three- to four-week-old B10.S and B10S.Tmevp3^SJL mice were irradiated and reconstituted with bone marrow from B10.S, B10S.Tmevp3^SJL, SJL.Tmevp3^B10S, or SJL mice. Ten weeks later the mice were inoculated with 10⁴ PFU of Theiler's virus and observed for 45 days. Some mortality from early encephalomyelitis was expected, and observed, since the mice had a B10.S genetic background. Surprisingly, more mortality was observed in B10S.Tmevp3^SJL than in B10.S recipient mice (Figure 4). The level of mortality depended also on the nature of the bone marrow donor strain. B10S.Tmevp3^SJL mice reconstituted with bone marrow from the B10.S and SJL strains showed the highest mortality. All surviving mice were killed at 45 days p.i. Survival data (Figure 4) were analyzed according to the nonparametric Kaplan-Meier model. Even when using a stratification strategy to take into account the effect of the donor strain, the effect of the recipient strain on mortality is highly significant (log-rank test = 7.985 [1 d.f.] P = 0.0047). The reconstitution itself did not significantly affect mortality (Table 2). In conclusion, these experiments showed that, on a B10.S background, the Tmevp3^SJL allele increases mortality from early encephalomyelitis in bone marrow chimeras. The Tmevp3 locus must be expressed in radio-resistant cells since the effect was seen in B10S.Tmevp3^SJL recipients.

View larger version (17K): In this window In a new window Download PPT slide   FIGURE 4.— Mortality during the acute encephalomyelitis of bone marrow chimeras. Cumulative survival curve for B10S.Tmevp3SJL recipient strain (A) and B10.S recipient strain (B) are obtained using the Kaplan-Meier model stratified according to the donor strain (circle, B10.S; square, B10S.Tmevp3SJL; triangle, SJL.Tmevp3B10S; diamond, SJL). n, number of mice. The strain before the arrow is the donor strain and that after is the recipient strain.

 

View this table: In this window In a new window   TABLE 2 Mortality during early disease of bone marrow chimeras and control mice

 
Viral load in the spinal cord of bone marrow recipients was measured 45 days p.i. using a Taqman real-time RT–PCR assay (data not shown). Because up to 50% mortality was observed between 10 and 21 days p.i. in some groups and not in others, as described above, and because mortality might be linked to viral load, the results were difficult to interpret. No major effect of the Tmevp3 locus on viral load in survivors was observed. B10S.Tmevp3^SJL recipient mice were infected at a higher level than B10.S recipient mice but the difference was not statistically significant (P = 0.0546). The effect of the donor strain on viral load was also not significant (P = 0.0902). These results contrast with those obtained with mice inoculated at 3–4 weeks of age in which mortality was lower, 15% in both B10.S and B10S.Tmevp3^SJL mice, and the difference of viral load was larger and statistically significant.

In conclusion, the study of mortality from early encephalomyelitis and of viral load during persistent infection showed that the Tmevp3 locus influences both phenotypes. The difference of segregation in the four mouse strains depending both on age and on the phenotype studied (Table 3) proved that more than one gene controls these phenotypes and that at least one of them is located in the Tmevp3 interval. In the simplest model, two genes are located in the Tmevp3 interval. One gene affects mortality from early encephalomyelitis in mice inoculated at 13–14 weeks of age, and the other gene affects viral load during persistent infection in mice inoculated at 3–4 weeks of age. The SJL allele confers susceptibility over the B10.S allele in both cases. Previous experiments have suggested that the Tmevpg1 gene is a candidate gene to control viral load during persistence of mice inoculated at 3–4 weeks of age (VIGNEAU et al. 2003). Finding a candidate gene for the control of mortality during the acute encephalomyelitis in the 400-kb region would substantiate this model.

View this table: In this window In a new window   TABLE 3 Phenotypes of the four mouse strains

 

Il22 is a candidate gene for the control of mortality during acute encephalomyelitis:

Il22 is one of the four genes present in the Tmevp3 interval. Because the IL22 cytokine is involved with innate immunity, the Il22 gene is an interesting candidate for the control of mortality during the acute encephalomyelitis. A comparison of the cDNA sequences of the Il22^SJL and Il22^B10.S alleles showed only one nonsynonymous mutation at position three of the protein (I3V), three synonymous mutations, and four polymorphisms in the 3' noncoding sequence (Table 4). None of the 3' noncoding polymorphisms are located in the ATTTA rich region conserved in mammals. Three polymorphisms were also found in the 340 bp long promoter region of the gene suggesting that its expression could be regulated differently in SJL and B10.S mice. It was not possible to study other regulatory elements due to a genomic duplication containing the Il22 gene. This duplication was detected in seven strains of laboratory mice (B10.S, FVB, SWR, C57BL/6, 129/Sv, C3H, DBA/1) using four microsatellites. One of the two Il22 copies is most probably a pseudogene since all seven strains have a 603-bp deletion of its putative promoter and its first exon. No duplication was detected in the seven other mouse strains including the SJL strain.

View this table: In this window In a new window   TABLE 4 Polymorphisms in the Il22 gene

 
The levels of Il22 expression in the spinal cord depend on viral load. His level is high only in a small fraction of mice with a low viral load (supplemental Figure 1S at http://www.genetics.org/supplemental/). Therefore, to compare the levels of Il22 expression in mice with different Il22 alleles, it was necessary that each parental strain and its congenic carried similar viral load (i.e., that there was no effect of the Tmevp3 locus on viral load between the four strains of mice). This condition was met at 21 days p.i. (Figure 5A). At this time, differences of viral load between B10.S, SJL, B10S.Tmevp3^SJL, and SJL.Tmevp3^B10S mice do not depend on the Tmevp3 locus (P = NS) but only on the genetic background (P < 0.001). Mice with an SJL background were infected at higher level than mice with a B10.S background. The level of expression of the Il22 gene was determined in the spinal cord of these mice using Taqman real-time RT–PCR assay (Figure 5B shows the results). The level of Il22 expression is significantly different among the four strains (P < 0.001). It depends on the allele at the Tmevp3 locus (P < 0.001) and not on the genetic background (P = NS) or on an interaction between these two factors (P = NS). The Il22^SJL allele is expressed at higher level than the Il22^B10S allele. Interestingly, the difference of Il22 expression is not limited to the central nervous system as similar results were obtained when comparing the level of Il22 expression in resting CD4⁺ T splenocytes (data not shown).

View larger version (16K): In this window In a new window Download PPT slide   FIGURE 5.— Viral load (A) and expression of Il22 gene (B) in the four groups of mice at 21 days p.i. SQRT, square root; rectangle, 90% interval of viral load around median; error bar, 95% interval around median. Viral load is not significantly different between each parental strain and the congenic strain of the same background. Expression of Il22 gene among the four groups of mice is significantly different (P < 0.01). Effect of the Tmevp3 locus, the background, and the interaction between both factors on Il22 expression are shown below. Significant differences of Il22 expression between two groups of mice are shown. P = NS, nonsignificant difference; *P < 0.05; **P < 0.01; ***P < 0.001. n, number of mice.

 
To confirm these results, the level of Il22 expression was measured in the spinal cord and the thymus of (B10S.Tmevp3^SJL x B10.S)F₁ mice 21 days p.i. using Taqman probes specific for each allele and labeled with different fluorophores. Figure 6A shows the level of expression of the Il22^SJL and the Il22^B10.S alleles measured at the same time and in the same sample. Whereas in thymus both alleles were expressed at the same level, the SJL allele was expressed at higher level than the B10.S allele in the infected spinal cords.

View larger version (16K): In this window In a new window Download PPT slide   FIGURE 6.— Allelic expression of Il22 gene in (B10S.Tmevp3SJL x B10.S)F1 animals at 21 days p.i. (A) and in bone marrow chimeras at 45 days p.i. (B). For each sample, fluorescence (in arbitrary unit) of the SJL allele (x-axis) and that of the B10.S allele (y-axis) are measured using an allelic-specific probe after 40 cycles of amplification by Taqman RT–PCR. The strain before the arrow is the donor strain and the strain after is the recipient strain. Samples of spinal cord at 21 or 45 days p.i. from B10.S and SJL mice with very different level of Il22 expression were chosen as positive controls. Each sample of a B10.S mouse was matched with one of an SJL one according to their level of Il22 expression. In all positive control, the PCR amplification reached a plateau and the specificity of detection of each Il22 allele was always verified.

 
According to the results of the bone marrow chimera experiments described above, the candidate gene must be expressed in radio-resistant cells of the spinal cord. Interestingly, it was reported that the Il22 gene is expressed in uninfected mouse brain (DUMOUTIER et al. 2000a). We confirmed this finding for the spinal cord (Figure 2S). To test if, in bone marrow chimera, the Il22 gene was expressed or not by radio-resistant cells, the expression of both alleles was measured in the spinal cord of bone marrow chimeras 45 days p.i. Figure 6B shows that the Il22^SJL allele was expressed at higher level in the spinal cord of B10S.Tmevp3^SJL recipients of B10.S bone marrow than in the spinal cord of B10.S recipients of B10S.Tmevp3^SJL bone marrow. The converse was true for the Il22^B10.S allele. Therefore, the Il22 allele expressed in the spinal cord of infected bone marrow chimera was predominantly that of the recipient.

In conclusion, the results reported here are congruent with the hypothesis that Il22 is the gene of the Tmevp3 interval, which is responsible for mortality during early encephalomyelitis.

For this work, we developed a new congenic line, B10S.Tmevp3^SJL, which bears the SJL allele of the Tmevp3 locus on a B10.S background. Extensive screening with SNPs did not detect any other SJL interval, besides Tmevp3, in the genome of this strain. The converse is true for the SJL.Tmevp3^B10S strain. Therefore, the phenotypes of the SJL.Tmevp3^B10S and B10S.Tmevp3^SJL strains must result from genes located in, respectively, the D10Mit233–D10Mit74 and D10Mit73–D10Mit151 intervals. For the SJL.Tmevp3^B10S strain, an analysis of polymorphisms between parental strains allowed us to reduce the Tmevp3 interval to 400 kb (Figure 1; http://genome.ucsc.edu/cgi-bin/hgGateway). This region contains on average one polymorphism every 200 bp. The exact size of the region is difficult to determine because it contains a duplication, which has not been sequenced yet in most strains of laboratory mice (DUMOUTIER et al. 2000b; http://genome.ucsc.edu/cgi-bin/hgGateway). However, we cannot rule out that a recent mutation in the centromeric region to the 400-kb Tmevp3 interval is responsible for the Tmevp3 phenotype, although this seems very unlikely. A cluster of cytokines, which covers the most important part of the 400-kb region, contains two genes of cytokine, Ifng and Il22, and Tmevpg1, a gene coding for a noncoding mRNA. The three genes are candidate for the Tmevp3 locus. However, we cannot rule out that the gene that controls mortality during acute encephalomyelitis in mice of B10.S background inoculated at 13–14 weeks of age is located outside the 400-kb region at another location of the D10Mit73–D10Mit151 interval (see Figure 1). Even if it was the case, the gene controlling viral load at 45 days p.i. in 13–14-week-old mice of SJL background would still be located within the 400-kb region.

Our work shows that the Tmevp3 locus affects two distinct phenotypes related to Theiler's virus pathogenesis: mortality during early encephalomyelitis and viral load during persistent infection. The effect of the Tmevp3 locus on mortality during the acute encephalomyelitis was not reported previously because the SJL.Tmevp3^B10S congenic mice bears an SJL background that does not confer susceptibility to the early disease. This complex phenotype (Table 3) is due to the action of more than one gene, one of them being located in the Tmevp3 interval. The simplest model to explain our data is to locate two genes in the Tmevp3 interval. The first gene controls viral load at 45 days p.i. in mice inoculated at 3–4 weeks of age. Its resistant B10.S allele not only decreases viral load during the persistent infection but also increases the difference of mortality between males and females in mice with a B10.S background. We showed previously that the Tmevpg1 gene that codes for a noncoding mRNA is a candidate gene to control viral load of mice inoculated at 3–4 weeks of age (VIGNEAU et al. 2003). Tmevpg1 could control the expression of the Ifng gene whose inactivation increases susceptibility to viral load during persistence (FIETTE et al. 1995; PULLEN et al. 1994; RODRIGUEZ et al. 1995). Interestingly, the activity of the Ifng pathway increases with age (ADKINS et al. 2004; HäRTEL et al. 2005; NEWPORT et al. 2004; SACK et al. 1998). This could explain that the phenotypes we described depend on age. The second gene controls mortality and viral load during the persistent infection of 13–14-week-old mice. Its SJL allele decreases viral load at 45 days p.i. in mice with an SJL background and increases mortality during the acute encephalomyelitis in mice with a B10.S background.

Data from the literature and the study of the Il22 gene are in agreement with this model. Il22 is a good candidate gene for the control of mortality during acute encephalomyelitis in mice inoculated at 13–14 weeks of age, since the Il22^SJL allele is expressed at a higher level than the Il22^B10.S allele in the central nervous system of these mice and mortality is higher in mice with a Tmevp3^SJL allele than in mice with a Tmevp3^B10.S allele. Interestingly, strains whose immune responses control viral load after the acute encephalomyelitis, such as the C57BL/6 and C57BL/10 strains, are also strains that show neurological symptoms and mortality during the early encephalomyelitis. Also, it has been shown that viral clearance from the gray matter during the acute encephalomyelitis is immune mediated (DRESCHER et al. 1999; NJENGA et al. 1997). This suggests that the cost of an efficient immune response, which controls the infection, might be clinical disease and sometimes death. IL22 induces the expression of genes involved with the innate immunity (WOLK et al. 2004). Therefore, it is conceivable that an increased expression of IL22 induces the efficient immune response responsible for the phenotype of mice inoculated at 13–14 weeks of age. However, we cannot rule out a role of the nonsynonymous polymorphism (I3V) present in the gene. In this model, the Il22^B10.S allele is responsible for an increase of viral load in mice with an SJL background and inoculated at 13–14 weeks of age. This late effect for a gene involved in innate immunity would probably be the consequence of an event that occurred during the early encephalomyelitis. The second gene, which controls viral load during persistence in mice inoculated at 3–4 weeks of age, shows the same relationship between susceptibility to acute encephalomyelitis and resistance to viral load during persistence, but in this case it is the B10.S allele that is associated with an efficient immune response. Interestingly, the two parental strains received the 400-kb region from two allopatric subspecies, M. m. domesticus and M. m. musculus, that were separated 700–800 thousand years ago (GUÉNET and BONHOMME 2003). Since these two genes in the Ifng cluster of cytokines might have been differently selected in these two subspecies, admixture of that cluster in strains of laboratory mice would be responsible for the appearance of this complex phenotype. Search for signature of either positive or balancing selection in candidate genes might help to characterize the causal polymorphism.

In conclusion, we postulate that two genes in the Tmevp3 interval are responsible for the phenotypes that we studied. Their efficiency depends on the age of the mouse at the time of inoculation. For both genes, the allele inducing an efficient immune response increases mortality during the acute encephalomyelitis and decreases viral load during persistence. But for one of the genes, the SJL allele is efficient and for the second gene, it is the B10.S allele. These two genes affect predominantly one of the two phenotypes, susceptibility to mortality during the acute encephalomyelitis or to viral load during persistence. This level of dissection of complex traits can be performed only in animal models at present. Characterizing such loci at the molecular level in mouse models will help to develop strategy that can be applied to susceptibility to infections of humans. The frequent organization of immune genes in clusters suggests that such a level of complexity may be common in human diseases. Il22, a member of the IL-10 family, could be an interesting candidate gene for several infectious diseases of mouse and humans (MISSE et al. 2007; ZHENG et al. 2007).

We thank M. Szatanik and J.-L. Guénet for the gift of genomic DNA, S. Becq for technical help in cytometry analysis, R. Cheynier, J.-L. Guénet, X. Montagutelli, and P. Vidalain for helpful discussions. This work was supported by the Institut Pasteur Fondation, the Association pour la Recherche sur la Sclérose en Plaques, the Centre National de la Recherche Scientifique. M.M. is a recipient of a scholarship from the Association pour la Recherche sur la Sclérose en Plaques.

¹ Present address: Unité de Recherche Oncogenèse et Virologie Moléculaire, Institut Pasteur, 75724 Paris Cedex 15, France.

² Present address: Unité INSERM U561, Hôpital Saint Vincent de Paul, 75014 Paris, France.

ADKINS, B., C. LECLERC and S. MARSHALL-CLARKE, 2004 Neonatal adaptative immunity comes of age. Nat Rev. Immunol. 4: 553–563.[CrossRef][Medline]

AUBAGNAC, S., M. BRAHIC and J. F. BUREAU, 2001 Viral load increases in SJL/J mice persistently infected by Theiler's virus after inactivation of the ß2m gene. J. Virol. 75: 7723–7726.[Abstract/Free Full Text]

AUBAGNAC, S., M. BRAHIC and J.-F. BUREAU, 2002 Bone marrow chimeras reveal non-H-2 hematopoietic control of susceptibility to Theiler's virus persistent infection. J. Virol. 76: 5807–5812.[Abstract/Free Full Text]

AZOULAY, A., M. BRAHIC and J.-F. BUREAU, 1994 FVB mice transgenic for the H-2D^b gene become resistant to persistent infection by Theiler's virus. J. Virol. 68: 4049–4052.[Abstract/Free Full Text]

AZOULAY-CAYLA, A., S. DETHLEFS, B. PéRARNAU, E. L. LARSSON-SCIARD, F. A. LEMONNIER et al., 2000 H-2D^b–/– mice are susceptible to persistent infection by Theiler's virus. J. Virol. 74: 5470–5476.[Abstract/Free Full Text]

BIHL, F., M. BRAHIC and J.-F. BUREAU, 1999 Two loci, Tmevp2 and Tmevp3, located on the telomeric region of chromosome 10, control the persistence of Theiler's virus in the central nervous system. Genetics 152: 385–392.[Abstract/Free Full Text]

BRAHIC, M., J. F. BUREAU and T. MICHIELS, 2005 The genetics of the persistent infection and demyelinating disease caused by Theiler's virus. Annu. Rev. Microbiol. 59: 279–298.[CrossRef][Medline]

BUREAU, J.-F., X. MONTAGUTELLI, S. LEFEBVRE, J.-L. GUÉNET, M. PLA et al., 1992 The interaction of two groups of murine genes determines the persistence of Theiler's virus in the central nervous system. J. Virol. 66: 4698–4704.[Abstract/Free Full Text]

BUREAU, J.-F., X. MONTAGUTELLI, F. BIHL, S. LEFEBVRE, J.-L. GUÉNET et al., 1993 Mapping loci influencing the persistence of Theiler's virus in the murine central nervous system. Nat. Genet. 5: 87–91.[CrossRef][Medline]

CHUNG, Y., X. YANG, S. H. CHANG, L. MA, Q. TIAN et al., 2006 Expression and regulation of IL-22 in the IL-17-producing CD4+ T lymphocytes. Cell Res. 16: 902–907.[CrossRef][Medline]

DRESCHER, K. M., P. D. MURRAY, C. S. DAVID, L. R. PEASE and M. RODRIGUEZ, 1999 CNS cell populations are protected from virus-induced pathology by distinct arms of the immune system. Brain Pathol. 9: 21–31.[Medline]

DUMOUTIER, L., J. LOUAHED and J. C. RENAULD, 2000a Cloning and characterization of IL-10 related T cell-derived inducible factor (IL-TIF), a novel cytokine structurally related to IL-10 and inducible by IL-9. J. Immunol. 164: 1814–1819.[Abstract/Free Full Text]

DUMOUTIER, L., E. VAN ROOST, G. AMEYE, L. MICHAUX and J.-C. RENAULD, 2000b IL-TIF/IL-22: genomic organization and mapping of the human and mouse genes. Genes Immun. 1: 488–494.[CrossRef][Medline]

FIETTE, L., C. AUBERT, U. MULLER, S. HUANG, M. AGUET et al., 1995 Theiler's virus infection of 129Sv mice that lack the interferon alpha/beta or interferon gamma receptors. J. Exp. Med. 181: 2069–2076.[Abstract/Free Full Text]

GUÉNET, J.-L., and F. BONHOMME, 2003 Wild mice: an ever-increasing contribution to a popular mammalian model. Trends Genet. 19: 24–31.[CrossRef][Medline]

HARARI, A., V. DUTOIT, C. CELLERAI, P. A. BART, R. A. DU PASQUIER et al., 2006 Functional signatures of protective antiviral T-cell immunity in human virus infections. Immunol. Rev. 211: 236–254.[CrossRef][Medline]

HäRTEL, C., N. ADAM, T. STRUNK, P. TEMMING, M. MüLLER-STEINHARDT et al., 2005 Cytokine responses correlate differentially with age in infancy and early childhood. Clin. Exp. Immunol. 142: 446–453.[Medline]

LI, K. B., 2003 ClustalW-MPI: ClustalW analysis using distributed and parallel computing. Bioinformatics 19: 1585–1586.[Abstract/Free Full Text]

LIANG, S. P., X.-Y. TAN, D. P. LUXENBERG, R. KARIM, K. DUNUSSI-JOANNOPOULOS et al., 2006 Interleukin (IL)-22and IL-17 are coexpressed by Th17 cells and cooperatively enhance expression of antimicrobial peptides. J. Exp. Med. 203: 2271–2279.[Abstract/Free Full Text]

LIPTON, H. L., A. S. KUMAR and M. TROTTIER, 2005 Theiler's virus persistence in the central nervous system of mice is associated with continuous viral replication and a difference in outcome of infection of infiltrating macrophages versus oligodendrocytes. Virus Res. 111: 214–223.[CrossRef][Medline]

MCALLISTER, A., F. TANGY, C. AUBERT and M. BRAHIC, 1989 Molecular cloning of the complete genome of Theiler's virus, strain DA, and production of infectious transcripts. Microb. Pathog. 7: 381–388.[CrossRef][Medline]

MENDEZ-FERNANDEZ, Y. V., M. J. HANSEN, M. RODRIGUEZ and L. R. PEASE, 2005 Anatomical and cellular requirements for the activation and migration of virus-specific CD8+ T cells to the brain during Theiler's virus infection. J. Virol. 79: 3063–3070.[Abstract/Free Full Text]

MICHIELS, T., V. DEJONG, R. RODRIGUS and C. SHAW-JACKSON, 1997 Protein 2A is not required for Theiler's virus replication. J. Virol. 71: 9549–9556.[Abstract]

MISSE, D., H. YSSEL, D. TRABATTONI, C. OBLET, S. LO CAPUTO et al., 2007 IL-22 participates in an innate anti-HIV1 host-resistance network through acute-phase protein induction. J. Immunol. 178: 407–415.[Abstract/Free Full Text]

MORAN, J. L., A. D. BOLTON, P. V. TRAN, A. BROWN, N. D. DWYER et al., 2006 Utilization of a whole genome SNP panel for efficient genetic mapping in the mouse. Genome Res. 16: 436–440.[Abstract/Free Full Text]

NAGEM, R. A., J. R. FERREIRA JUNIOR, L. DUMOUTIER, J. C. RENAULD and I. POLIKARPOV, 2006 Interleukin-22 and its crystal structure. Vitam. Horm. 74: 77–103.[Medline]

NEWPORT, M. J., T. GOETGHEBUER, H. A. WEISS, H. WHITTLE, C. A. SIEGRIST et al., 2004 Genetic regulation of immune responses to vaccines in early life. Genes Immun. 5: 122–129.[CrossRef][Medline]

NJENGA, M. K., K. ASAKURA, S. F. HUNTER, P. WETTSTEIN, L. R. PEASE et al., 1997 The immune system preferentially clears Theiler's virus from the gray matter of the central nervous system. J. Virol. 71: 8592–8601.[Abstract]

NOVELLI, F., and J.-L. CASANOVA, 2004 The role of IL-12, IL-23 and IFN-gamma in immunity to viruses. Cytokine Growth Factor Rev. 15: 367–377.[CrossRef][Medline]

PULLEN, L. C., S. D. MILLER, M. C. DAL CANTO, P. H. VAN DER MEIDE and B. S. KIM, 1994 Alteration in the level of interferon- results in acceleration of Theiler's virus-induced demyelinating disease. J. Neuroimmunol. 55: 143–152.[CrossRef][Medline]

RODRIGUEZ, M., K. PAVELKO and R. L. COFFMAN, 1995 Gamma interferon is critical for resistance to Theiler's virus-induced demyelination. J. Virol. 69: 7286–7290.[Abstract]

SACK, U., U. BURKHARDT, M. BORTE, H. SCHADLICH, K. BERG et al., 1998 Age-dependent levels of select immunological mediators in sera of healthy children. Clin. Diagn. Lab. Immunol. 5: 28–32.[Medline]

SCHMIDT, H. A., K. STRIMMER, M. VINGRON and A. VON HAESELER, 2002 TREE-PUZZLE: maximum likelihoood phylogenetic analysis using quartets and parallel computing. Bioinformatics 18: 502–504.[Abstract/Free Full Text]

SNYDER, L. C., S. P. TRUSKO, N. FREEMAN, J. R. ESHLEMAN, S. S. FAKHARZADEH et al., 1988 A gene amplified in a transformed mouse cell line undergoes complex transcriptional processing and encodes a nuclear protein. J. Biol. Chem. 263: 17150–17158.[Abstract/Free Full Text]

VAN DEN BROEK, M. F., U. MULLER, S. HUANG, R. M. ZINKERNAGEL and M. AGUET, 1995 Immune defence in mice lacking type I and/or type II interferon receptors. Immunol. Rev. 148: 5–18.[CrossRef][Medline]

VIGNEAU, S., P.-S. ROHRLICH, M. BRAHIC and J.-F. BUREAU, 2003 TmevpgI, a candidate gene for the control of Theiler's virus persistence, could be implicated in the regulation of gamma interferon. J. Virol. 77: 5632–5638.[Abstract/Free Full Text]

WOLK, K., S. KUNZ, E. WITTE, M. FRIEDRICH, K. ASADULLAH et al., 2004 IL-22 increases the innate immunity of tissues. Immunity 21: 251–254.

ZHENG, Y., D. M. DANILENKO, P. VALDEZ, I. KASMAN, J. EASTHAM-ANDERSON et al., 2007 Interleukin-22, a T(H)17 cytokine, mediates IL-23-induced dermal inflammation and acanthosis. Nature 445: 648–651.[CrossRef][Medline]

Communicating editor: N. A. JENKINS

International Access Link

Online ISS 1091-6490
Copyright 2008 by the Genetics Society of America
phone: 412-268-1812   fax: 412-268-1813   email: genetics-gsa{at}andrew.cmu.edu