Cj1386 is an ankyrin-containing protein involved in heme trafficking to catalase in Campylobacter jejuni

Abstract

Campylobacter jejuni, a microaerophilic bacterium, is the most frequent cause of human bacterial gastroenteritis. C. jejuni is exposed to harmful reactive oxygen species (ROS) produced during its own normal metabolic processes and during infection from the host immune system and from host intestinal microbiota. These ROS will damage DNA and proteins and cause peroxidation of lipids. Consequently, identifying ROS defense mechanisms is important for understanding how Campylobacter survives this environmental stress during infection. Construction of a ΔCj1386 isogenic deletion mutant and phenotypic assays led to its discovery as a novel oxidative stress defense gene. The ΔCj1386 mutant has an increased sensitivity toward hydrogen peroxide. The Cj1386 gene is located directly downstream from katA (catalase) in the C. jejuni genome. A ΔkatAΔ Cj1386 double deletion mutant was constructed and exhibited a sensitivity to hydrogen peroxide similar to that seen in the ΔCj1386 and ΔkatA single deletion mutants. This observation suggests that Cj1386 may be involved in the same detoxification pathway as catalase. Despite identical KatA abundances, catalase activity assays showed that the ΔCj1386 mutant had a reduced catalase activity relative to that of wild-type C. jejuni. Heme quantification of KatA protein from the ΔCj1386 mutant revealed a significant decrease in heme concentration. This indicates an important role for Cj1386 in heme trafficking to KatA within C. jejuni. Interestingly, the ΔCj1386 mutant had a reduced ability to colonize the ceca of chicks and was outcompeted by the wild-type strain for colonization of the gastrointestinal tract of neonate piglets. These results indicate an important role for Cj1386 in Campylobacter colonization and pathogenesis.

Fig 1

The katA and Cj1386 genes are independently transcribed. RT-PCR products from RNA extracted from C. jejuni NCTC11168 were analyzed by agarose gel electrophoresis for operon identification analysis. Lanes 1 to 5, RT-PCR from RNA extracted from C. jejuni NCTC11168; lanes 6 and 7, RT-PCR from genomic DNA extracted from C. jejuni NCTC11168. Lanes: L, 1-kb ladder; 1, katA-RT-SE and katA-RT-AS (909-bp band visible, demonstrating presence of katA transcript); 2, Cj1386-RT-SE and Cj1386-RT-AS (349-bp band visible, demonstrating presence of Cj1386 transcript); 3, katA-RTint-SE and Cj1386-RTint-AS (absence of predicted 1,062-bp band, indicating that katA and Cj1386 are not cotranscribed); 4, Cj1384c-RT-SE and katA-RT-AS (absence of predicted 1,560-bp band, indicating that Cj1384c and katA are not cotranscribed); 5, Cj1386-RT-SE and Cj1387-RT-AS (absence of predicted 697-bp band, indicating that Cj1386 and Cj1387c are not cotranscribed); 6, positive-control katA-RTint-SE and Cj1386-RTint-AS (1,062-bp band present); 7, positive-control katA-RT-SE and Cj1386-RT-AS (1,615-bp band present); 8, negative control (no RT added to reaction mixture).

Fig 2

The C. jejuni strain contains a single enzyme with catalase activity, and the ΔCj1386 mutant exhibits a reduced catalase activity. All bacterial strains (C. jejuni NCTC11168 and ΔkatA and ΔCj1386 mutants) were grown in MEMα to mid-log phase (OD₆₀₀ of 0.2) under microaerophilic conditions at 37°C prior to cell extract preparation. When required, wild-type C. jejuni was exposed to 1 mM H₂O₂ for 10 min prior to total protein isolation. Fifty micrograms of whole-cell extract was loaded onto an 8% nondenaturing gel run at 25 mA at 4°C and assayed for catalase activity. BLC, bovine liver catalase (Sigma-Aldrich).

Fig 3

Cj1386 enhances KatA catalase activity. (A) Catalase activities of cytoplasmic and periplasmic cellular protein fractions for wild-type C. jejuni and ΔCj1386, ΔCj1386+Cj1386, ΔkatA, and ΔkatA+katA strains are expressed as μmol of hydrogen peroxide decomposed per minute per mg of protein. Error bars indicate standard errors of the means (n = 4). An asterisk indicates a P value of <0.05 using the Student t test. (B) Cell fractionation control assay using Western blot analysis. Five micrograms of wild-type C. jejuni and ΔCj1386 mutant cytoplasmic and periplasmic protein extracts were loaded into each well and subsequently assessed for cytoplasmic protein contamination by immunoblotting using an anti-PerR antibody.

Fig 4

KatA expression is not affected in the ΔCj1386 mutant, as shown by Western blot analysis of KatA protein content in wild-type C. jejuni; ΔCj1386, ΔkatA, and ΔkatA+katA mutants; and affinity-purified C. jejuni KatA. (A) Five micrograms of protein lysate or 100 ng of purified protein was loaded into each lane followed by immunoblotting using anti-KatA or anti-Fur antiserum. (Top) KatA protein level as detected by anti-KatA antiserum. (Bottom) Loading control of total protein content detected by anti-Fur antiserum. (B) Quantification of KatA protein level in wild-type C. jejuni, ΔCj1386, ΔkatA, and ΔkatA+katA strains. Relative intensity of KatA was determined by quantifying each band from the immunoblot and standardizing it against WT KatA. Error bars represent the standard errors of 3 biological replicates.

Fig 5

KatA immunoprecipitated from the ΔCj1386 mutant has decreased catalase activity relative to wild-type C. jejuni. (A) Immunoprecipitated KatA from wild-type C. jejuni, ΔCj1386, ΔkatA, and ΔCj1386+Cj1386 strains eluted in 50 mM glycine, pH 2.8. Five microliters of each immunoprecipitated sample and 0.5 μg of purified C. jejuni KatA were loaded into each lane and separated on a 10% denaturing SDS-PAGE gel. (B) Catalase activity of KatA immunoprecipitated samples eluted in soft-elution buffer from wild-type C. jejuni, ΔCj1386, ΔkatA, and ΔCj1386+Cj1386 strains. KatA protein concentrations were determined by densitometry from the SDS-PAGE gel (not shown), and 250 μg of KatA was assayed for activity. Catalase activity is expressed as μmol of hydrogen peroxide decomposed per minute per mg of protein. Error bars indicate standard errors of the means (n = 4). An asterisk indicates a P value of <0.05 using the Student t test.

Fig 6

The ΔCj1386 mutant is defective for cecal colonization of chicks. Chicks were grouped into two sets of 10 (WT and ΔCj1386 + Cj1386) and one set of 12 (ΔCj1386). Data points correspond to the levels of colonization of the ceca per chick. The dashed line represents the limit of detection for the assay. Solid bars indicate the median colonization level of bacteria for each strain. The asterisk indicates a P value of <0.05 using a nonparametric Mann-Whitney rank sum test.

Fig 7

ΔCj1386 and ΔkatA mutants are affected for intestinal colonization of piglets. Competitive index in neonate piglets. Each data point represents the competitive index for C. jejuni NCTC11168 and the ΔCj1386 (A) or ΔkatA (B) mutant in five intestinal segments from one piglet. The dashed line represents the ratio at which wild-type C. jejuni and the ΔCj1386 (A) or ΔkatA (B) mutant are colonizing the intestinal segment at similar levels (one is not outcompeting the other). Solid bars represent the medians for each segment.