An N-ethyl-N-nitrosourea (ENU)-induced dominant negative mutation in the JAK3 kinase protects against cerebral malaria

Abstract

Cerebral malaria (CM) is a lethal neurological complication of malaria. We implemented a genome-wide screen in mutagenized mice to identify host proteins involved in CM pathogenesis and whose inhibition may be of therapeutic value. One pedigree (P48) segregated a resistance trait whose CM-protective effect was fully penetrant, mapped to chromosome 8, and identified by genome sequencing as homozygosity for a mis-sense mutation (W81R) in the FERM domain of Janus-associated kinase 3 (Jak3). The causative effect of Jak3(W81R) was verified by complementation testing in Jak3(W81R/-) double heterozygotes that were fully protected against CM. Jak3(W81R) homozygotes showed defects in thymic development with depletion of CD8(+) T cell, B cell, and NK cell compartments, and defective T cell-dependent production of IFN-γ. Adoptive transfer of normal splenocytes abrogates CM resistance in Jak3(W81R) homozygotes, an effect attributed to the CD8(+) T cells. Jak3(W81R) behaves as a dominant negative variant, with significant CM resistance of Jak3(W81R/+) heterozygotes, compared to CM-susceptible Jak3(+/+) and Jak3(+/-) controls. CM resistance in Jak3(W81R/+) heterozygotes occurs in presence of normal T, B and NK cell numbers. These findings highlight the pathological role of CD8(+) T cells and Jak3-dependent IFN-γ-mediated Th1 responses in CM pathogenesis.

Figure 1. ENU-induced mutation that protects mice against P. berghei ANKA-induced cerebral malaria.

(A) Breeding scheme for the production and identification of ENU-induced recessive mutations that convey protection against cerebral malaria (CM). Details of the breeding strategy are described in “Materials and Methods”. G3 and F2 pedigrees were phenotyped for the presence of animals resistant to P. berghei-induced CM. (B) Mice were infected with P. berghei ANKA (10⁶ P. berghei ANKA-parasitized red blood cells, i.v.) and survival was monitored over time for individual pedigree P48 G3s (green and blue lines) derived from independent G2 females and a G1 male, and for susceptible C57BL/6J (B6, red line), and resistant mutant mouse strains bearing loss of function mutations in either the IFN-γ gene KO (IFN-γ KO, black line) or IRF8 (BXH2, brown line). Mice surviving past day 13 post infection were considered to be CM-resistant.

Figure 2. Resistance to cerebral malaria in pedigree 48 maps to the central portion of chromosome 8.

Genome-wide linkage analysis of the CM-resistance trait (survival) was conducted in 44 G3 (15 resistant, 29 susceptible) mice from pedigree 48, using polymorphic markers informative for the B6 and B10 progenitors. (A) LOD score traces identifying significant linkage to chromosome 8 (p = 0.05, genome-wide significance shown as dotted line); the position of informative markers is shown, including rs33080067 and rs32729089 (LOD score ∼5.8). (B) Haplotype analysis of the central portion of chromosome 8 in CM-resistant and CM-susceptible G3 animals from pedigree 48 (A, B6 homozygotes; H, B6/B10 heterozygotes; B, B10 homozygotes) showing exclusion of homozygote B6 haplotypes from the CM-susceptible group. The positions of the markers (in Mb) from the centromere (shown as a grey dot) are shown. (C) The G1 male from pedigree 48 was out-crossed to 129S1/SvlmJ to generate an F2 population (n = 211) that was phenotyped for response to P. berghei ANKA infection. Survival of F2 mice as well as parental 129S1/SvlmJ and C57BL/10J controls is shown.

Figure 3. Phenotypic expression of resistance to cerebral malaria in mice from pedigree 48.

G3 and F2 mice homozygote (P48/P48) or heterozygote (P48/+) for the B6-derived mutant central chromosome 8 were identified by genotyping, and were subjected to several analyses, along with parental C57BL/6J, C57BL/10J and 129S1/SvlmJ controls. (A) Macroscopic examination of thymus from control and mutants showing severely atrophied thymus in homozygote mutants (representative of 5 mice per group). (B) FACS density plots of different cell populations in thymus (top), spleen (middle) and bone marrow (bottom) stained for CD4, CD8, CD19, and CD117. The position of the different cell lineages in the scatter plots are identified at the extreme right panel and their numbers are expressed as a percentage (± SE; n = 5 mice per group) of total cells in this tissue. (C) Flow cytometric analysis of immune cell lineage composition expressed as the absolute number (mean ± SD; n = 4–6 mice per group) of CD4⁺ and CD8⁺ single positive (SP), CD4⁺CD8⁺ double positive (DP), B cells (CD19⁺), granulocytes (GR; Gr1⁺), hematopoietic stem cells (HSC; lineage⁻CD117⁺) and NK cells from 10⁵ cells from spleen, thymus and bone marrow. Asterisks (one-way Anova test with Bonferroni post-test) identify significant differences between experimental animals and the C57BL/6J controls: *p<0.05; ** p<0.01; *** p<0.001. Data are representative of two independent experiments.

Figure 4. A W81R mutation in the FERM domain of Jak3 causes resistance to cerebral malaria in pedigree 48.

(A) Schematic representation of the Jak3 protein, showing the 7 Jak Homology (JH) structural domains, and associated functional annotation (below). The position of the W81R mis-sense mutation discovered in pedigree 48 is shown. (B) Multiple sequence alignment of the amino terminal portion of Jak3 surrounding W81 shows high conservation across Jak3 relatives (the corresponding species is identified).

Figure 5. Resistance to cerebral malaria in Jak3^W81R/− compound heterozygotes.

Jak3^W81R homozygotes were crossed to a mouse line bearing a null Jak3 allele (Jak3^−/−) to create the Jak3^W81R/+ compound heterozygotes. Single heterozygotes (Jak3^W81R/+, Jak3^+/−; Fig. 5A), homozygotes (Jak3^W81R, Jak3^−/−; Fig. 5B), compound heterozygotes (Jak3^W81R/−, gray line; Fig. 5A) and C57BL/10J controls were infected with 10⁶ P. berghei ANKA-parasitized RBCs and monitored for survival. All surviving mice were sacrificed on day 15 post-infection (experimental end-point). The data are representative of 2 independent experiments.

Figure 6. Cell transfer experiments to induce susceptibility to cerebral malaria in Jak3^W81R mice.

Wild type C57BL/10J (B10) mice, Jak3^W81R homozygote mutants (Fig. 6A), or Jak3^W81R homozygote mutants having received the indicated cell populations (20 million total splenocytes or 5 million each of the indicated cell population; 2 hrs prior to infection); (Fig. 6A, 6B, 6C) were infected i.v. with 10⁶ P. berghei ANKA-parasitized RBCs, and survival from infection was monitored. Untreated wild type B10 and Jak3^W81R mutants were used as susceptible and resistant controls, respectively. All surviving mice were sacrificed on day 15 post-infection (experimental end-point). Data represent 2 independent experiments.

Figure 7. The Jak3^W81R mutation confers susceptibility to infection with different bacterial pathogens.

(A) Control (C57BL/6J) and Jak3^W81R homozygote mutants were infected with 5×10⁴ colony-forming units (CFUs) of Mycobacterium bovis (BCG), and 6 weeks later, mice were sacrificed and the degree of infection was assessed by determination of spleen CFUs (left) and splenomegaly (spleen index; right). (B) Jak3^W81R homozygotes and B6 controls were infected by aerosol inoculation of 50 CFUs/lung of Mycobacterium tuberculosis H37Rv, and monitored for survival. (C) Jak3^W81R homozygotes, Jak3^W81R/+ heterozygotes and B6 controls were infected by oral gavage with 3×10⁸ CFUs of Citrobacter rodentium strain DBS100. Mice were monitored for survival for 30 days post infection. Data represent two independent experiments.