Adult male Sprague–Dawley rats (Harlan, Udine, Italy) weighing 320–330 g were used for this study. The experimental protocol was approved by the Committee for Animal Experimentation of the Ministry of Health, Rome, Italy. Animals were treated according to Italian and European Guidelines for Animal Care and Experimentation, DL 116/92, in agreement with the European Communities Council Directive guidelines (86/609/EEC).
The SD rats were initially housed three per cage in a temperature (18–22 °C) and humidity-controlled room under a constant 12-h light/dark cycle and fed with standard rat chow and water ad libitum. After acclimatization, animals were identified from number 1–36 and then, by using the Microsoft Excel software (Lombardia, Italy), assigned to one of the two study groups by a blind operator. Eighteen rats underwent sepsis (induced by cecal ligation and puncture (“CLP group”), and 18 rats (who received laparotomy only) served as controls (“Sham-group”). At the end of the surgical procedure each rat received the same amount of CTX intra-peritoneally (i.p.)(see below). A schematic representation of the experimental design is shown in Fig. 1.
The CLP procedure was performed as previously detailed [12–14]. Briefly, rats anesthetized with sodium pentobarbital (65 mg/kg, i.p.) and positioned on a homoeothermic heating pad in order to maintain body temperature between 36.5 and 37.5 °C, received a 3 cm midline laparotomy on the anterior abdomen. The cecum was exposed and ligated distally to the ileocecal valve, without causing intestinal obstruction, and it was punctured on the anti-mesenteric border with a 16 G (1.65 mm diameter) needle. The cecum was then squeezed to extrude its fecal content and was replaced into the peritoneal cavity; finally, a 20 G cannula was positioned in the peritoneal cavity and secured to the abdominal wall in order to subsequently inject CTX and/or collect peritoneal samples. The anterior peritoneal wall and the skin were closed with 3-0 silk sutures. “Sham” animals underwent laparotomy and peritoneal cannula positioning only. After surgery, all the rats were housed individually for urine collection in metabolic cages in the same environmental conditions.
To document the development of sepsis, the clinical appearance, mortality during the experimental time (6 h) and micro-organism growth in the peritoneal cavity were evaluated. For micro-organism detection and count, samples of peritoneal fluid were collected at the end of the experimental time (6 h) and cultured for the growth of Gram-positive and Gram-negative isolates. Samples were incubated on Mannitol Salt Agar for 24 h at 37 °C for Gram-positive strains and layered on Mac-Conkey Agar III for 24 h at 37 °C for Gram-negative culture. Species identification was determined by the API system (BioMerieux, Inc., Hazelwood, MO), using a panel Gram-positive (GP) card. The effects of sepsis on the status of the GBF, and on CTX elimination were also examined (see below).
The CTX was purchased from Sigma-Aldricht Co. (Saint Louis, Missouri, USA). The antimicrobial solution was prepared daily by dissolving the powder in sterile saline. 100 mg/kg CTX was injected intra-peritoneally through the 20 G abdominal cannula in a volume of 1 mL; the cannula was flushed immediately with 1 mL of saline. The dose of CTX was chosen according to previous experimental protocols in order to mimic the human dose of 1 g [15].
Plasma and lung concentrations were evaluated for the first 6 h after CTX administration. Serial blood samples (drawn from the tail vein) were collected at the following times: before the administration of the antibiotic (baseline sample) and after 1, 2, 4 and 6 h (n = at least four samples each experimental time). The blood was immediately centrifuged at 4000 rpm for 15 min at 4 °C, plasma collected, divided into aliquots and stored at −80 °C until assayed. Lung specimens were also collected at 2, 4 and 6 h after antibiotic treatments to evaluate antibiotic tissue penetration into the lung. Finally, urinary samples were collected from the metabolic cage after the 6 h interval in order to measure CTX elimination.
The CTX concentrations were determined in triplicate by a validated large-plate agar diffusion technique, according to Good Laboratory Practice (GPL) standards, using the Mueller–Hinton Agar (Oxoid, UK) as the culture medium and Escherichia coli K12 as the test organism, with a lower limit of sensitivity of 0.125 mg/L. Standard concentrations were prepared daily in pooled plasma for plasma samples, in saline for the urine and lung samples. The test organism (1 × 106 CFU/mL) was added using the surface layer technique. After homogeneous distribution of the culture, the excess liquid was removed with a pipette. The plates were incubated at 37 °C in air overnight. Best-fit standard curves were obtained by linear regression analysis. The correlation coefficient was no less than 0.99. For all CTX samples, intra-assay precision ranged from 1.5 to 6.8% and the inter-assay precision at a level of 0.75 mg/L ranged from 4.6 to 5.6%. Amicon R 10 K filters (Sigma Aldrich, Milan Italy) were used for the determination of free drug concentration (the filter cut-off being 10.000 nominal molecular weight limit). In short, 0.5 mL of plasma were stratified on the filters and centrifuged at 4000g for 30 min. The filtrate was collected and analyzed with the same bioassay described above.
Diuresis was collected for each rat (at least four rats each group). The total amount of diuresis in the 0–6 h interval of time was noted. Urine was stored at −20 °C until examination.
Changes of GFB-associated glycocalyx, in particular in the sialic components, was performed by lectin histo-chemistry. At the end of the experimental time, after the pentobarbital overdose, a midline incision was made in the abdomen, and kidney specimens were fixed in Carnoy’s fluid and routinely processed to obtain 6 μm-thick paraffin sections. Maackia amurensis agglutinin (MAA) and Sambucus nigra agglutinin (SNA) digoxigenin (DIG) labeled lectins (Roche Diagnostic, Mannheim, Germany) were used to identify sialic acids linked α2–3 and α2–6 to galactose or galactosamine, respectively. Lectin histo-chemistry was performed as previously described [16, 17]. In short, sections were treated with 20% acetic acid to inhibit the endogenous alkaline phosphatase, then treated with 10% blocking reagent in Total Buffered Saline (TBS) to reduce the background labelling. Afterwards, sections were washed in TBS and rinsed in Buffer 1, then incubated in DIG-labelled lectins diluted in Buffer 1 (1 and 5 μL/mL for SNA and MAA respectively) for 1 h at room temperature. Sections were then rinsed in TBS, incubated with anti-digoxigenin, conjugated with alkaline phosphatase diluted in TBS and washed in TBS. Labelling of the sites containing bound lectin-digoxigenin was obtained by incubating slides with Buffer 2 containing nitroblue tetrazolium (NBT)/X-phosphate (Roche Diagnostics, Mannheim, Germany) [16, 18].
Controls for lectin specificity included pre-incubation of lectins with the corresponding hapten sugars [16, 18]. For evaluation of the lectin-stained location, ten random fields (40× ocular) in each section (five sections for each specimen) were examined using light microscopy. A densitometric analysis was also carried out for lectin reactivity intensity by measuring the average optical density (OD) of region of interest (ROI, 40 µm2 area), using ImageJ National Institute of Health (NIH, Bethesda, MD, USA) software. Measured values were normalized to background [(OD-ODbkg)/ODbkg]. At least eight regions of interest in ten different optical fields were analyzed in each experiment.
For the PD analysis, the CTX concentration data on plasma were used to develop a population model by non- linear mixed effects modelling (using SAS 9.3 PROC NLMIXED).
We considered the following first-order compartment model for these
$$C_{it} = \frac{{Dk_{ei} k_{ai} }}{{Cl_{i} \left( {k_{ai} - k_{ei} } \right)}} \left[ {e^{{ - k_{ei} t}} - e^{{ - k_{ai} t}} } \right] + e_{it}$$
where C
it
is the observed concentration of the i-th subject at time t, D is the dose of ceftriaxone, k
ei
is the elimination rate constant for subject i, k
ai
is the absorption rate constant for subject i, Cl
i
is the clearance for subject i, and e
it
are normal errors. To allow for random variability between subjects, we assumed:
$$Cl_{i} = e^{{\beta_{1} + b_{i1} }}$$
$$k_{ei} = e^{{\beta_{2} + b_{i2} }}$$
$$k_{ai} = e^{{\beta_{3} }}$$
where the βs denote fixed-effects parameters and the bis denote random-effects parameters with an unknown covariance matrix.
To obtain effect estimation and random effect joint distribution the matrix dual quasi-Newton algorithm was used in order to achieve convergence.
Two dosing regimens (100 and 200 mg/kg, corresponding to 1 and 2 g in humans respectively [15]) were simulated using Monte Carlo simulation based on the PK model. Each simulation generated concentration–time profiles from 10,000 rats per dosage regimen. Based on this data, Tfree > MIC and the probability target attainment (PTA) for different multiples of MIC, were estimated graphically. MIC values according to the EUCAST breakpoints were considered, in which CTX susceptibility for Enterobacteriaceae was ≤1 mg/L.
The free concentrations (Cfree) were estimated from the total concentrations (C
tot
) using in vivo binding parameters of CTX and the following equation described by Kodama et al. [19]:
$$C_{{free}} = \frac{1}{2}\left[\begin{gathered}- \left( {nP + ~\frac{1}{{k_{{aff}} }} - ~C_{{tot}} } \right) \hfill \\ + ~\sqrt {\left( {nP + ~\frac{1}{{k_{{aff}} }} - ~C_{{tot}} } \right)^{2} + \frac{{4C_{{tot}} }}{{K_{{aff}} }}} \hfill \end{gathered} \right]$$
The following binding of CTX was used: the total concentration of protein binding sites (nP) 517 μmol l−1 and the binding affinity constant (Kaff) 0.0367 l μmol−1.
In order to assess differences in continuous variables between the sham and CLP groups the t test or the Mann–Whitney test were used. The test choice was made following the Shapiro–Wilk test result for normality distribution of the variable in each group.
The difference between sham and CLP groups in PK parameters was assessed by Welch’s t-test. Statistical significance was set at p value lower than 0.05.