Staphylococcus aureus-dependent septic arthritis in murine knee joints: local immune response and beneficial effects of vaccination

Abstract

Staphylococcus aureus is the major cause of human septic arthritis and osteomyelitis, which deserve special attention due to their rapid evolution and resistance to treatment. The progression of the disease depends on both bacterial presence in situ and uncontrolled disruptive immune response, which is responsible for chronic disease. Articular and bone infections are often the result of blood bacteremia, with the knees and hips being the most frequently infected joints showing the worst clinical outcome. We report the development of a hematogenous model of septic arthritis in murine knees, which progresses from an acute to a chronic phase, similarly to what occurs in humans. Characterization of the local and systemic inflammatory and immune responses following bacterial infection brought to light specific signatures of disease. Immunization of mice with the vaccine formulation we have recently described (4C-Staph), induced a strong antibody response and specific CD4+ effector memory T cells, and resulted in reduced bacterial load in the knee joints, a milder general inflammatory state and protection against bacterial-mediated cellular toxicity. Possible correlates of protection are finally proposed, which might contribute to the development of an effective vaccine for human use.

Figure 1. Intravenous injection of S. aureus in mice results in a rapid infection of bones and joints.

(A) 2D Ventral pictures of a representative CD1 mouse intravenously infected with S. aureus Xen36 bioluminescent strain from day 1 (d1) to day 7 (d7) p.i. (B) Correlation analysis between CFU counts and BLI intensity at day 7 p.i. Spearman’s r value and significance are reported. (C) Quantification of the bioluminescent (BLI) signals belonging to the knee joint areas of infected animals from day 0 (naïve mice, before infection) to day 7. Median values of 5 animals at each time point are reported together with interquartile ranges. (D) 3D analysis of BLI signals from animals infected with Xen36 strain. A ventral picture of 1 animal out of 5 at day 2 p.i. is shown. In the panel on the right, an enlargement of the left hind paw is depicted. (E) Confocal microscopy image of a whole knee from one mouse infected with S. aureus Newman strain 7 days after injection. One out of 5 mice is shown.

Figure 2. Mice intravenously infected with S. aureus develop arthrosynovitis and osteomyelitis, which evolve from an acute to a chronic state.

(A) Histopathological analysis of the knees of infected mice. The knees from 4 to 6 animals/time point from 3 independent experiments were scored for severity of the lesions (Severity) and for progression of the infection (stage) both as to bones and joints analysis and mean +/− SEM were shown. (B) Representative hematoxilin-eosin-stained slide for 1 animal out of 4 to 6 at each time point. Time 0 is representative of naïve mice. Legend: F = femur, T = tibia, * = bone destruction/abscesses, # = fibroplasia/collagenous tissue, arrows = details of inflammatory cell accumulations. Scale bar: days 0, 3, 7, 200 μm; days 14, 30, 90, 1 mm; (C) Immunohistochemical staining of an abscess (Absc.) from an infected bone. S. aureus (SA, red specific staining, blue counterstaining) and neutrophils (N, blue specific staining, red counterstaining) were stained and a negative control (CTL) sample is shown in the third panel. Scale bar: 100 μm.

Figure 3. Infected animals are only temporarily able to control S. aureus dependent disease.

(A–C) Log₁₀ CFU counts recovered from the knee joints (CFU/organ, A), kidneys (B, CFU/organ) and blood (CFU/ml, C) of mice intravenously infected with S. aureus Newman strain from day 3 to day 90 p.i. Medians and interquartile ranges of 5 to15 animals/time point from 2/3 independent experiments are shown. ldv = lower detectable value. (D) Survival curve of mice intravenously infected with Newman strain. (E,F) IgM and IgG titration against S. aureus alpha-hemolysin_H35L (Hla_H35L) in sera (E) and knee joint washes (F) of infected mice from day 0 (naïve animals) to day 90. Graphs report medians and interquartile ranges of Mean Fluorescence Intensity (MFI) signals of 6 to 8 mice at each time point and 2 independent experiments. In all the graphs reported above, the vertical dotted grey lines indicate the edges of the 3 proposed stages of infection.

Figure 4. Rapid proinflammatory response and cellular recruitment follow S. aureus infiltration in the knee joints.

(A) Cytokine profiles measured in knee joint washes of infected animals (pg/ml). Only cytokines which significantly changed with respect to the negative control (not infected animals, time 0) at least at one time point are reported here. (B) Phenotypic characterization of cells recruited into the knee joints of not infected (time 0) and infected mice after bacterial injection. The total number of each cell sub-population/wash is reported here. In both cases, samples were collected from 6 to 14 mice/time point from 3 independent experiments and data are reported as median values and interquartile ranges. Statistical analysis was performed using a two-tailed Mann-Whitney U-test comparing single time points against time 0.

Figure 5. S. aureus systemic infection modulates host immune response in the blood.

(A) Cytokine profiles measured in the sera of infected animals (pg/ml). Only cytokines which significantly changed with respect to time 0 at least at one time point are shown. (B) Phenotypic characterization of immune cells circulating in the blood of not infected (time 0) and infected mice throughout the observation period. Data were expressed as cell number/ml. In both cases, samples were collected from 6 to 14 mice/time point from 3 independent experiments and data were shown as median values and interquartile ranges. Statistical analysis was performed using a two-tailed Mann-Whitney U-test comparing single time points against controls.

Figure 6. Immunization with the alum/4C-Staph protein combination results in reduction of bacterial burden in the knee joints of infected mice.

(A–C) Log₁₀ CFU counts of knee joint washes (A–C) and homogenized kidneys (B) of mice infected with Newman (A-B) or Xen36 strain (C). Mice were either immunized with alum/4C-Staph or alum alone (A–C). Single dots represent CFU counts of single animals from 2 independent experiments, the black horizontal line indicates median values while the dotted gray line is representative for the lower detectable value of the assay (ldv). (D) Normalized bioluminescent intensity (BLI) from knees of CD1 mice immunized with alum/4C-Staph or alum alone at days 1 and 7 p.i. with Xen36 strain. Columns and error bars represent medians and upper interquartile ranges, respectively. In both cases, the two-tailed Mann-Whitney U-test was always used to assess significance.

Figure 7. Antibodies induced by vaccination with alum/4C-Staph formulation reduce CFU counts in the joints of infected animals.

(A,B) Antibody titration (Mean Fluorescent Intensity) against single combo antigens (A) and cytokine measurement (B) in knee joint washes of immunized mice 7 days after infection. Samples were collected from 9 to 10 animals/group from 2 independent experiments. Median and upper interquartile ranges are reported. (C) Log₁₀ CFU counts of knee joint washes collected from passively immunized mice 7 days p.i. Single dots represent single animals form 2 independent experiments, black lines report median values for each group and the grey dotted line shows the lower detectable value (ldv). For all the data reported above, the two-tailed Mann-Whitney U-test was used to assess significance.

Figure 8. Immunization with alum/4C-Staph formulation results in partial restoring of B cell numbers in the knee joints of immunized mice.

(A,B) Percentage of live and dead staining (A) and cellular recruitment analysis (B) in knee joint washes of immunized mice 7 days after infection. In (B), fold change of single cell sub-populations as compared to naïve mice is reported. Legend: N = neutrophils, Mo = monocytes; MΦ = macrophages; DC = dendritic cells; CD4+ = CD4+ T lymphocytes; CD8+ = CD8+ T lymphocytes; B = B lymphocytes. Medians and upper interquartile ranges are depicted from 3 independent experiments and 13–15 mice/group. The two-tailed Mann-Whitney U-test was used to assess significance.

Figure 9. Knee joint recruitment of specific CD4+ Tem cells follows immunization with alum/4C-Staph combination.

(A,B) IL-7Rα+CD62L- (CD4+Tem cells, A) and IL-7Rα+CD62L+ (CD4+ Tcm cells, B) percentage of CD4+/CD44^high T cells recruited into the knees and circulating in the blood of immunized mice after intravenous infection with S. aureus Newman strain. Five to 9 mice (knee joints) or 11 to 12 mice (blood) from 2/3 independent experiments were analyzed and median values together with interquartile ranges are shown in the box-whiskers graphs. The two-tailed Mann-Whitney U-test was used to assess significance.