Inoculation Pneumonia Caused by Coagulase Negative Staphylococcus

Abstract

Rationale: Although frequently retrieved in tracheal secretions of critically ill patients on mechanical ventilation, the existence of pneumonia caused by coagulase-negative staphylococci (CoNS) remains controversial.

Objective: To assess whether Staphylococcus haemolyticus (S. haemolyticus) inoculated in mice's trachea can infect normal lung parenchyma, increasing concentrations of S. haemolyticus were intratracheally administered in 221 immunocompetent mice.

Methods: Each animal received intratracheally phosphate-buffered saline (PBS) (n = 43) or live (n = 141) or inactivated (n = 37) S. haemolyticus at increasing load: 1.0 × 10⁶, 1.0 × 10⁷, and 1.0 × 10⁸ colony forming units (CFU). Forty-three animals were sacrificed at 12 h and 178 were sacrificed at 36 h; 64 served for post-mortem lung histology, 157 served for pre-mortem bronchoalveolar lavage (BAL) analysis, and 42 served for post-mortem quantitative bacteriology of lung tissue. The distribution of biofilm-associated genes was investigated in the S. haemolyticus strain used in our in vivo experiment as well as among 19 other clinical S. haemolyticus strains collected from hospitals or nursing houses.

Measurements and main results: Intratracheal inoculation of 1.0 × 10⁸ CFU live S. haemolyticus caused macroscopic and histological confluent pneumonia with significant increase in BAL white cell count, tumor necrosis factor-α (TNF-α), and macrophage inflammatory protein (MIP)-2. At 12 h, high concentrations of S. haemolyticus were identified in BAL. At 36 h, lung injury and BAL inflammation were less severe than at 12 h and moderate concentrations of species belonging to the oropharyngeal flora were identified in lung tissue. The inoculation of 1.0 × 10⁶ and 1.0 × 10⁷ CFU live S. haemolyticus caused histologic interstitial pneumonia and moderate BAL inflammation. Similar results were observed after inoculation of inactivated S. haemolyticus. Moreover, biofilm formation was a common phenotype in S. haemolyticus isolates. The low prevalence of the ica operon in our clinical S. haemolyticus strain collection indicated icaA and icaD independent-biofilm formation.

Conclusion: In immunocompetent spontaneously breathing mice, inoculation of S. haemolyticus causes concentration-dependent lung infection that spontaneously recovers over time. icaA and icaD independent biofilm formation is a common phenotype in S. haemolyticus isolates.

Keywords: Staphylococcus haemolyticus; coagulase-negative staphylococci; inoculation pneumonia; macrophage inflammatory protein-2; tumor necrosis factor-α.

FIGURE 1

(A) The detailed antimicrobial resistance profile of S. haemolyticus isolated from the blood culture specimen of a female hospitalized patient with Hodgkin’s lymphoma. (B) The number of colony forming units (CFU) was calculated when optical density value was 0.8. The left panel showed 10^–6 times diluted inoculum, and the right panel showed 10^–7 times diluted inoculum.

FIGURE 2

Pictures of Petri dish showing the heterogeneity of strains growth in bronchoalveolar lavage (BAL) and lung homogenate. (A) Strains collected from BAL at 12 h. (B) Strains collected from lung homogenate at 12 h. (C) Strains collected from BAL and lung homogenate at 36 h. The black arrows showed colonies identified as S. haemolyticus.

FIGURE 3

White cells and neutrophils in the blood (A,C) and BAL (B,D), 36 h following tracheal inoculation of phosphate-buffered saline (PBS), live and inactivated S. haemolyticus. Median values, interquartile 25th–75th range and extreme values (>2 SD) are represented. Blood neutrophils were significantly different between groups: Kruskal–Wallis p = 0.0095; ^∗ indicates p < 0.05 for 10⁶ CFU versus PBS and 10⁷ CFU; BAL White cells were significantly different between groups: Kruskal–Wallis p = 0.0023; ^†indicates p = 0.0013 for 10⁸ CFU versus PBS; BAL neutrophils were significantly different between groups: Kruskal–Wallis p = 0.01; ^††indicates p = 0.0084 for 10⁸ CFU versus PBS.

FIGURE 4

Macroscopic (square windows) and histological lung aspects 12 and 36 h following tracheal inoculation of PBS, and live S. haemolyticus at 10⁶, 10⁷, and 10⁸ CFU. Lung injury score (LIS) is represented as median values and interquartile 25th–75th range and extreme values (>2 SD). Twelve and 36 h after inoculation, LIS was significantly different between groups: Kruskal–Wallis p = 0.0002; ^∗indicates p < 0.05 for 10⁸ CFU versus PBS and 10⁶ CFU; ^†indicates p < 0.05 for 10⁸ CFU versus PBS and 10⁶ CFU. Twelve and 36 h after tracheal inoculation of 10⁶ and 10⁷ CFU, interstitial pneumonia is present, characterized by inflammatory cells infiltrating interalveolar septa and respecting alveolar space and lung architecture. Twelve and 36 h after tracheal inoculation of 10⁸ CFU, confluent pneumonia is present, characterized by inflammatory cells invading alveolar space, hyaline membranes, lung consolidation, and loss of basal lung architecture.

FIGURE 5

Lung histology 36 h following tracheal inoculation of PBS and inactivated S. haemolyticus at increasing load 10⁶, 10⁷, and 10⁸ CFU. LIS is represented as median values and interquartile 25th–75th range and extreme values (>2 SD). LIS is significantly different between groups: Kruskal–Wallis p = 0.0004; ^∗indicates p < 0.05 for 10⁸ CFU versus PBS and 10⁶ CFU. Thirty-six hours after tracheal inoculation of inactivated S. haemolyticus at 10⁶ and 10⁷ CFU, inflammatory cells infiltrate interalveolar septae, respecting alveolar space and lung architecture. The inoculation of 10⁸ CFU of inactivated S. haemolyticus induced invasion of inflammatory cells within alveolar space, hyaline membrane formation, and lung consolidation.

FIGURE 6

Tumor necrosis factor (TNF-α) and monocyte inflammatory protein-2 (MIP-2) in BAL following tracheal inoculation of PBS, and increasing inoculum (10⁶, 10⁷ and 10⁸ CFU) of live and inactivated S. haemolyticus. Data are expressed as median values, interquartile 25th–75th range and extreme values (>2 SD). Twelve hours following tracheal inoculation of live S. haemolyticus, alveolar levels of TNF-α and MIP-2 significantly increased with the inoculum (Kruskal–Wallis p = 0.0002 for TNF-α and p = 0.004 for MIP-2; ^∗ and ‡ indicate p < 0.05 for 10⁸ CFU versus PBS and 10⁶ CFU). Thirty-six hours following inoculation, alveolar levels of TNF-α and MIP-2 remained significantly elevated in mice that received the 10⁸ CFU (Kruskal–Wallis p = 0.0012 for TNF-α and p < 10^–4 for MIP-2; ^†† and ╪╪ indicate p < 0.05 for 10⁸ CFU versus PBS and 10⁶ CFU; ^†indicates p = 0.03 for 10⁷ versus 10⁶ CFU and ╪ indicates p < 0.05 for 10⁷ versus PBS and 10⁶ CFU). However, both cytokine levels were significantly lower than levels measured 12 h following inoculation of 10⁸ CFU S. haemolyticus (p < 10^–4 for TNF-α and MIP2 alveolar levels at 12 h versus 36 h). Thirty-six hours following tracheal inoculation of 10⁸ CFU inactivated S. haemolyticus, BAL TNF-α and MIP-2 significantly increased with the inoculum (Kruskal–Wallis p = 0.0086 for TNF-α and p = 0.0013 for MIP-2; ^ο indicate p < 0.05 for 10⁸ versus 10⁷ and 10⁶ CFU; ^οο indicate p < 0.05 for 10⁸ versus 10⁷ and 10⁶ CFU, and PBS).

FIGURE 7

The total protein concentration in BAL following tracheal inoculation of PBS, and increasing inocula (10⁶, 10⁷, and 10⁸ CFU) of live and inactivated S. haemolyticus remained unchanged. Data are expressed as median values, interquartile 25th–75th range, and extreme values (>2 SD).

FIGURE 8

Bacterial load in BAL and lung tissue homogenate following tracheal inoculation of PBS, and increasing inocula (10⁶, 10⁷, and 10⁸ CFU) of live and inactivated S. haemolyticus. Data are expressed as median values, interquartile 25th–75th range and extreme values (>2 SD). (A) Kruskal–Wallis p = 0.002; ^∗indicates p < 0.05 for 10⁸ CFU versus PBS and 10⁶ CFU; (B) Kruskal–Wallis p = 0.04; ^∗indicates p = 0.046 for 10⁸ CFU versus 10⁶ CFU; (C) Kruskal–Wallis p = 0.003; ^∗indicates p < 0.05 for 10⁸ CFU and 10⁷ CFU versus PBS. (D) No significant differences of bacterial colonies in BALF among the inactivated S. haemolyticus groups.

FIGURE 9

Biofilm formation of CoNS strains and detection of biofilm-associated genes. (A) Data are shown in mean ± SD. Mean level of absorbance of wells was measured spectrophotometrically at 570 nm by a microplate reader after staining by crystal violet (red cross line shows cutoff value = 0.12). (B) Images of wells represent each strain in crystal violet-stained microplates.