Intraperitoneal versus intranasal administration of lipopolysaccharide in causing sepsis severity in a murine model: a preliminary comparison

Abstract

Community-acquired respiratory infection is the commonest cause of sepsis presenting to emergency departments. Yet current experimental animal models simulate peritoneal sepsis with intraperitoneal (I.P.) injection of lipopolysaccharide (LPS) as the predominant route. We aimed to compare the progression of organ injury between I.P. LPS and intranasal (I.N.) LPS in order to establish a better endotoxemia murine model of respiratory sepsis. Eight weeks old male BALB/c mice received LPS-Escherichia coli doses at 0.15, 1, 10, 20, 40 and 100 mg per kg body weight (e.g. LPS-10 is a dose of 10 mg/kg body weight). Disease severity was monitored by a modified Mouse Clinical Assessment Score for Sepsis (M-CASS; range 0-21). A M-CASS score ≥ 10 or a weight reduction of ≥ 20%, was used as a criterion for euthanasia. The primary outcome was the survival rate (either no death or no need for euthanasia). The progression of disease was specified as M-CASS, body weight, blood glucose, histopathological changes to lung, liver, spleen, kidney, brain and heart tissues. Survival rate in I.P. LPS-20 mice was 0% (2/3 died; 1/3 euthanized with M-CASS > 10) at 24 h. Survival rate in all doses of I.N. LPS was 100% (20/20; 3-4 per group) at 96 h. 24 h mean M-CASS post-I.P. LPS-10 was 6.4/21 significantly higher than I.N. LPS-10 of 1.7/21 (Unpaired t test, P < 0.05). Organ injury was present at 96 h in the I.P. LPS-10 group: lung (3/3; 100%), spleen (3/3; 100%) and liver (1/3; 33%). At 24 h in the I.P. LPS-20 group, kidney injury was observed in the euthanized mouse. At 96 h in the post-I.N. LPS-20 group, only lung injury was observed in 2/3 (67%) mice (Kruskal-Wallis test with Dunn's, P < 0.01). At 24 h in the post-I.N. LPS-100 group all (4/4) mice had evidence of lung injury. Variable doses of I.N. LPS in mice produced lung injury but did not produce sepsis. Higher doses of I.P. LPS induced multi-organ injury but not respiratory sepsis. Lethal models of respiratory virus, e.g., influenza A, might provide alternative avenues that can be explored in future research.

Keywords: Histology; Intranasal; Intraperitoneal; LPS; Organ injury; Sepsis.

Fig. 1

Survival curve of 8-week-old male Balb/c mice after treated with different LPS doses at 0.15, 1, 10 and 20 mg/kg body weight via intraperitoneal (I.P.) route, and 0.15, 1, 10, 20 and 40 mg/kg body weight via intranasal (I.N.) route over 96 h. Three or four mice for each group

Fig. 2

Measurement of a modified Mouse Clinical Assessment Score for Sepsis (M-CASS) of 8-week-old male Balb/c mice after intraperitoneal (I.P.) or intranasal (I.N.) treatment with 0.9% saline, 0.15 mg/kg LPS, 1 mg/kg LPS, 10 mg/kg LPS and 20 mg/kg LPS over 96 h. Data were presented as mean with standard deviation (SD). Comparison of M-CASS between I.P. LPS and I.N. LPS at the same dose was made by unpaired t test with Welch correction (* P < 0.05; ** P < 0.01; **** P < 0.0001). Three or four mice for each group. Note: A total score of 10 or more, or when one category reaches a 3, was used as a criterion for euthanasia; as for the dead mice, a score of 10 was allocated at the time points after mice died

Fig. 3

Measurement of body weight loss (%) of 8-week-old male Balb/c mice after intraperitoneal (I.P.) treatment with 0.15, 1, 10 and 20 mg/kg LPS over 96 h (a), measurement of body weight loss (%) of mice after intranasal (I.N.) treatment with 0.15, 1, 10, 20 and 40 mg/kg LPS over 96 h, and 100 mg/kg LPS over 24 h (b), comparison of body weight loss (%) of mice between 0 h, 1 h, 4 h and 24 h for I.P. 20 mg/kg LPS stimulation and I.N. 100 mg/kg LPS stimulation, respectively (Kruskal-Wallis test with Dunn’s multiple comparisons test; ns = not significant; * P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001) (c), comparison of body weight loss (%) of mice between I.P. 20 mg/kg LPS stimulation and I.N. 100 mg/kg LPS stimulation at 1 h, 4 h and 24 h (Mann-Whitney test with correction for multiple comparisons using the Holm-Sidak method) (d). Data were presented as median with 95% confidence interval (CI) in Fig. 4a and b, and median with 2.5-97.5 percentile in Fig. 4c and d. Three or four mice for each group. Note: suppositional = A body weight loss of 20 % which has been defined as one humane endpoint was allocated at the time points after mice died

Fig. 4

Measurement of blood glucose level (mg/dl) of 8-week-old male Balb/c mice after intraperitoneal (I.P.) treatment with 10 mg/kg LPS, 20 mg/kg LPS and 0.9% saline over 24 h (a), measurement of blood glucose level (mg/dl) of mice after intranasal (I.N.) treatment with 10 mg/kg LPS, 20 mg/kg LPS, 40 mg/kg LPS, 100 mg/kg LPS and 0.9% saline over 24 h (b), comparison of blood glucose level (mg/dl) of mice between 0 h, 1 h, 4 h and 24 h for I.P. 20 mg/kg LPS stimulation and I.N. 100 mg/kg LPS stimulation, respectively (Kruskal-Wallis test with Dunn’s multiple comparisons test; ns = not significant; * P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001) (c), comparison of blood glucose level (mg/dl) of mice between I.P. 20 mg/kg LPS stimulation and I.N. 100 mg/kg LPS stimulation at 1 h, 4 h and 24 h (Mann-Whitney test with correction for multiple comparisons using the Holm-Sidak method) (d). Data were presented as median with 95% confidence interval (CI). Three or four mice for each group. Note: for the groups in which mice died, blood glucose level of the whole group has been removed from analysis thereafter

Fig. 5

Histology of lung, liver, spleen and kidney from control and lipopolysaccharide (LPS) treated mice via intraperitoneal (I.P.) or intranasal (I.N.) routes at 24 h and 96 h stained with hematoxylin and eosin (H&E). Size bar = 50 µm. Histological scores were achieved by semi-quantitative scoring systems (Additional file 2), and statistical analysis was performed by Kruskal-Wallis test with Dunn’s multiple comparison test (ns = not significant; * P< 0.05; ** P < 0.01; *** P < 0.001; **** P< 0.0001; data were presented as median with interquartile; three or four mice for each group). Note: Sample availability was affected by the death of mice and/or experimental design. Representative graphs from one of 3-4 mice in each group are shown