Adapted Murine Sepsis Score: Improving The Research In Experimental Sepsis Mice Model

Objective The Murine Sepsis Score (MSS) is used to assess the severity of sepsis in rats and mice based on observational characteristics. The quantitative variables of blood glucose, body weight and temperature are predictors of severity in experimental models of sepsis. Therefore, our study sought to adapt the MSS with the same variables to indicate earlier the severity of the disease in murine models of the disease. Results Sepsis mice presented hypoglycemia, weight loss, and hypothermia. Therefore, these variables were included in the Adapted Murine Sepsis Score (A-MSS). The A-MASS presented 100% specicity and 87.5% sensibility been able to differentiate the early sepsis symptoms and its severity. The A-MSS allows an early and more complete diagnosis of sepsis in mice and might be considered as a procedure to improve the analysis of systemic sepsis dysfunction in murine experimental models. Score, NaCl:sodium ROC:Receiver Operating Characteristic, MPV:mean platelet volume, MPV/PC:ratio between MPV and platelet count, NLR:neutrophil-lymphocyte ratio, PLR:platelet to lymphocyte ratio, M-CASS:Mouse Clinical Assessment Score for Sepsis, SOFA:Sequential Organ Failure Assessment, CNS:central nervous system, PEPCK:phosphoenolpyruvate carboxykinase, iNOS:inducible nitric oxide synthase, NO:nitric oxide, IL-1β:interleukin-1 beta, TNF-α:tumor necrosis factor alpha, PAF:platelet-activating factor.


Introduction
Sepsis is a systemic infection characterized by multiple organ dysfunction and high mortality rates.
Therefore, studies investigating the pathophysiology in sepsis are required [1,2], and the use of experimental mouse models has been considered a fundamental approach [3].The bacteria-induced sepsis in animal models has similar characteristics found in septic patients [4,5] due to the bacteria membrane endotoxins like lipopolysaccharide (LPS) and lipoteichoic acid (LTA) [6].
One of the leading infectious focus in sepsis is abdominal origin.Staphylococcus aureus and Escherichia coli bacteria are commensals of the mouse gastrointestinal tract and can be found in the feces of these animals [5,7].Therefore, an experimental model of sepsis is peritonitis induction by the intraperitoneal administration of the autogenous fecal solution.From this procedure, sepsis is recognized by an application of Murine Sepsis Score (MSS) that consists of observing and categorizing the level of consciousness, activity, behavior, response to stimuli, respiratory rate, and quality of breathing movements [8].However, it is possible to observe a decreased glycemic levels, core temperature, and body weight loss if we follow up animals for 24 hours [9,10], but these variables have not been included in the MSS protocol [8].
Patients and animals with sepsis may present hypothermia [11,12], hypoglycemia [9,13], body weight loss [14], and hematological alterations [15].Also, early identi cation of sepsis is a well-known, relevant procedure to allow successful treatment of the disease [16].Therefore, herein we proposed an adaptation in the MSS (A-MSS) to provide a score that identi es sepsis in an experimental model in the initial stages.

Animals
Fifteen male C57BL/6 mice aged 90 to 150 days from the Life Sciences Department (DCVida) of the Northwestern Regional University of Rio Grande do Sul State (UNIJUÍ) were used in this study.

Experimental design
The animals were randomly divided into two experimental groups: Control (n = 7) and Sepsis (n = 8).A 20% fecal solution (200 mg/mL) was prepared with fresh stool in 0.9% NaCl, and it was administered at a dose of 5μL/g (1 mg/g, i.p.) in sepsis group, while Control animals received 0.9% NaCl [10].The animals were monitored for blood glucose, blood count, rectal temperature and MSS (Table S1), before (time zero) and at 4, 12, and 24 hours after sepsis induction and were euthanized by decapitation using guillotine.
Sample size was de ned considering the su cient number to nd difference between groups in MSS [10].

Procedures details
Glycemia was measured by distal tail puncture (~5µL) using Optium Xceed® glucometer.Rectal temperature was measured with a digital thermometer, and body weight was measured using a semianalytical balance.For hematological analysis, blood was collected (0, 4, 12 and 24 h) by caudal puncture (10 µl).Samples were diluted 1:3 with 0.9% saline, 1µl of anticoagulant (EDTA).Hematological strains were performed on the slide for differential leukocyte count, stained with a panoptic type [21].

Statistical analysis
Data were analyzed using the Kolmogorov-Smirnov test.The results in which the evaluation was performed over time were veri ed by two-way ANOVA (time x treatment) followed by Bonferroni post-test.

Spearman or Pearson correlation test was performed. The potential diagnosis of A-MSS within four hours after sepsis induction was tested by ROC curve. The cutoff point for A-MSS diagnosis was determined by
observing the most appropriate balance between sensitivity and speci city in the ROC curve [22].The signi cance level of 5% (P <0.05) was considered.

Results
The animals submitted to sepsis presented increased MSS 12 h and 24 h after fecal solution administration (Fig. 1a).Also, animals with sepsis showed a decrease in glycemia, rectal temperature and body weight (Fig. 1b, c, d).
It was possible to verify that sepsis induces alterations in appearance, level of consciousness, activity, stimulus-response, eye aspect and respiratory rate and quality (Figure S1).
Sepsis was able to cause a decrease in absolute and relative platelet count (Figure S2a,b), and increases mean platelet volume (MPV), and the ratio between MPV and platelet count (MPV/PC) (Figure S2c,d).The neutrophil count increased after 12 h, returning to basal levels 24 h after sepsis induction (Figure S3a).At 24 h we found a decrease in leukocyte count, which includes a decrease in lymphocyte and monocyte count (Figure S3b,c,d, respectively).These alterations in immune cells, as well as observed in the platelets, impact in the neutrophil-lymphocyte ratio (NLR), which increased at 12 h together with the platelet to lymphocyte ratio (PLR) in 24 h (Figure S3e,f, respectively).
A negative correlation was observed between MSS and glycemia, body temperature and weight loss, 24 hours after sepsis induction (Fig. 2a-c).Therefore, we proposed the A-MSS (Table 1).
The A-MSS was able to report sepsis within four hours of the sepsis induction (Fig. 2d).When this score was evaluated from the ROC curve in this period, it was found that it can be used as a diagnostic standard (Fig. 2e), from an A-MSS value of 3.5, obtaining a sensitivity of 87.5% and speci city of 100%.
A-MSS measured four hours after sepsis induction showed a negative correlation with the severity marker parameters in experimental sepsis: decrease in lymphocyte count, total leukocytes and platelets (Figure S4a-c).Also, there was a correlation with the increase in NLR, PLR and MPV/PC (Figure S4d-f).

Discussion
We showed that the inclusion of variables easily measurables as glycemia, temperature, and body weight in the MSS may improve the research in sepsis mice model.Our proposal of A-MSS represents a sum of observations, and together with the established MSS, the above-mentioned variables might be considered as a new score for the evaluation of sepsis in experimental models (Table 1).Sepsis is a complex disease that requires a complex form to diagnose it successfully.[25].The A-MSS added parameters to the MSS, which allowed an indirect assessment of the cardiovascular (temperature) and metabolic systems (glycemia and body weight), which are essential for the prognosis in sepsis in animals and humans.In humans, the diagnosis is made by measures of neurological, cardiovascular, respiratory, renal, hepatic, and platelet dysfunction, by the reproducible Sequential Organ Failure Assessment (SOFA) score.[26].We believe that A-MSS approximates the clinical variables evaluated in SOFA for the early diagnosis of sepsis in animal models.
The central nervous system dysfunction is mainly characterized by septic encephalopathy, followed by autonomic failure [27] [27].These dysfunctions cause tissue damage and impair brain function [28].In our study, these neurological effects were re ected in the impairment of the level of consciousness and activity in the Sepsis group (Figure S1a,c).The neurological symptoms can be identi ed in the rst hours by MSS [11], but using the A-MSS allows the assessment of autonomic insu ciency evaluating the temperature (Fig. 1c).
Neurohypophysial dysfunction promotes decreased hepatic gluconeogenesis-promoting adrenocortical hormones and liver and muscle glycogenolysis, causing a decrease in glycemia [29].Furthermore, LPS is able to decrease the activity of hepatic and renal enzyme phosphoenolpyruvate carboxykinase (PEPCK) promoting gluconeogenesis, and hypoglycemia [30], as observed in Fig. 1b.
Sepsis impairs neuroendocrine regulation by neuronal dysfunction impairing the secretion of vasopressor hormones [31], accompanied by the desensitization of receptors for vasoconstriction [32].Also, LPS and cytokines induce the nitric oxide production by macrophages, neutrophils and monocytes [32].Decreased cardiac output and blood pressure may be associated with hypothermia (Fig. 1c) [11,37], reinforcing the need to measure body temperature in sepsis mice models.
A catabolic condition leads to a reduction in muscle mass [38] is related to increased hospitalization period and mortality [14].This metabolic effect is also veri ed in an experimental sepsis model [9,10,39], and the results found in our study (Fig. 1d) may be associated with muscle atrophy and lipid catabolism [40].Also, leukocytes release interleukin-1 beta, which has a direct effect on appetite inhibition and food intake, re ecting in body weight loss [41].
The platelet function disorders play a crucial role in the pathophysiology of sepsis, with prognostic accuracy [42].LPS and cytokine endotoxemia (TNF-α, tumor necrosis factor-alpha, IL-8, IL-15) stimulate endothelium, monocytes, neutrophils and basophils to secrete platelet-activating factor (PAF) [42].PAF triggers platelet aggregation with the formation of microthrombi, which, together with leukocyte-induced hemophagocytosis [43], may result in decreased total platelet count (thrombocytopenia) and its relative rate [44].Regarding MPV, it is already known that the 24 h period is insu cient for the increase in MPV to indicate the severity in patients with sepsis [45], as observed in our study (Figure S2c).It has been proposed that MPV indicates severity in just after 72 hours [46] and that the evaluation of MPV is insu cient to predict the worsening of sepsis due to peritonitis with gram-negative bacteria [15].The dysfunction in the platelet count can also be observed by analyzing the ratios between the MPV and platelet count (MPV/CP), representing the risk of microthrombi formation [47].The platelet is a predictor of worsening of sepsis, speci cally when it accuses the systemic infection with gram-negative bacteria [15], as we observed in our study (Figure S2d).
Immunosuppression observed by decreasing lymphocyte and monocytes leads to a decrease in total leukocytes (Figure S3b,c,d).On the other hand, the bone marrow has a reservoir of neutrophils that are released into the circulation to combat infectious in the peritoneal cavity, as veri ed 12 h after sepsis induction [48].Neutrophils are immune cells of the rst line of defense against bacterial infection and may suffer exacerbated apoptosis in severe sepsis [15].
Studies suggest the application of NLR and PLR as in ammatory biomarkers [15].PLR and NLR elevation is a prognostic marker of lethality in patients with peritonitis [15], as well as in an animal models [49], similarly to Figure S3e,f.Also, platelet and immune biomarkers may indicate etiological agents [15].Thus, the A-MSS proposed in our study was able to indicate sepsis in mice just after four hours, correlated with the 24-hour values of the biomarkers mentioned above (Figure S3a-f).

Conclusion
The A-MSS allows an early and more complete diagnosis of sepsis in mice and might be considered as a procedure to improve the analysis of systemic sepsis dysfunction in mice.

Limitations
Our study did not use laboratory tests for sepsis evaluations such as bilirubin, creatinine or lactate.