Skip to main content

Table 3 Problems when performing experiments with Brucella in the mouse model and general recommendations

From: What have we learned from brucellosis in the mouse model?

Experimental variables Problems Recommendations
Mouse breed - Different susceptibility to Brucella infections. - Heterogeneity of infection. - Eight to ten weeks old (20 g) female BALB/c.- For BALB/c, n = 5 gives homogeneous spleen counts. For outbred breeds (e.g. CD1) the number per group should be increased (n ≥ 6) and the body weight factor taken into account (see Vaccination). For the recommended weight at the Mouse Phenome Database [169].
Target organ - Inconsistent infection in some organs. - Count CFU in spleens (consistently colonized in infected animals; longer persistence than in liver or other organs) after determining the individual organ weight.
Brucella wild-type challenging strain and virulence controls - Attenuation by inappropriate storage and/or handling. - Species, biovar and reference strain differences. - Use only reference strains (B. abortus 2308 or 544, B. melitensis 16 M or H38, B. ovis PA, B. suis 1330). Aliquots should be maintained in cryoprotective solution at–80°C (or below), or freeze-dried at 4°C.- Stocks seeded on agar plates and then cultured only once on new plates to obtain final inoculi.- Plates checked for contaminants and freed from condensation or syneresis water.- Make sure that reference strains reproduce typical virulence patterns (see The Brucella strains: replication patterns and related effects)
Attenuated Brucella strains and virulence control - Over-attenuation by inappropriate storage and/or handling. - Inappropriate infectious dose. - Lack of appropriate controls. - Competing events in superinfection protocols - See above for storage and inoculum preparation.- Typical multiplication (acute phase) and persistence (chronic phase) patterns in spleens should be assessed.- Use adequate virulent controls (see above, Wild-type strains). In genetic manipulation experiments, consider appropriateness of complemented strains and controls for unrelated attenuation cause by in vitro manipulations.- Avoid using protocols in which mixtures of virulent and attenuated Brucella strains are used as an attempt to determine the relative virulence of the former in relation to the latter bacteria.
Brucella infectious dose and route - Not optimized for the purpose of the experiment. - Dose not adequate to the route. - Animal handling during inoculation.Intrinsic problems in some routes. - Use PBS pH 6.85 for preparing the inoculum.- In a preliminary dose–response assay, determine the optimal dose/route (see Route of the infection) for each Brucella species, biovar and reference strain (e.g. B. abortus 2308 vs. 544 or B. melitensis 16 M vs. H38) and also for each mouse breed. For outbred mice, doses must also be adapted to the body weight/age (Physiopathology).- The intraperitoneal route is recommended for most purposes. To avoid intra-intestinal inoculation, displace the intestines and inoculate in the lateral of the abdomen (not in the linea alba). Intravenous inoculation in tail is exceedingly difficult and poses repeatability problems, particularly in small breeds. Oral and aerosol routes are not recommended (see Route of the infection).
Vaccination/attenuated strains dose - Inappropriate dose and route. - Absence of controls - For classical smooth vaccines, follow the OIE protocol (1 × 105 CFU/subcutaneously).- To evaluate B. melitensis and B. abortus vaccines, always include Rev1 or S19 reference strain controls, respectively.
Time intervals for virulence studies Not meaningful. - For screening, analyze two times corresponding to the multiplication phase and the persistence (e.g. 2 and 8 weeks post-infection). For definite results, test four times (e.g. 2, 4, 6, and 8 weeks post-infection).
Assessment of vaccine efficacy - Challenge strain. - Time intervals. - Differentiation of vaccine and challenge strains. - Use fully virulent reference strains and a control reference vaccine strain (see above).- Challenge 4 weeks after vaccination (see Vaccination).- Whenever possible, the challenge strain should carry identifiable marker(s).
CFU determination - Limit of detection not optimized. - Expression of results (CFU/organ vs. CFU/weight) - Homogenize the organ in 1:9 (weight:volume) PBS and plate 100 μL by triplicate of each dilution (limit of detection of this method = 3.3 CFU/mL of dilution, corresponding to less than 5 CFU/spleen)- Express the results as log CFU/organ and report spleen weights separately (inflammation varies depending on the Brucella strain and among mouse strains).
Evaluation of immune response - Presence of antigens and bacteria when performing ex vivo experiments. - Lack of correlation between transcripts and protein immune mediators - Lack of sensitivity- Lack of specificity - Inappropriate plotting of data - Procure APC from non-infected mice to avoid dragging antigen or bacteria. Whenever possible perform direct assays (e.g. flow cytometry, microscopy, cell protein extraction, and serum detection).- Contrast the results obtained with indirect methods with those generated by direct methods.- Use times of maximum expression of cell types or immune mediators.- Consider the “Mackaness effect”.- Include a saturating control that could reveal the real magnitude of the response. Whenever possible, avoid expressing data in relative numbers or “fold responses” and procure the inclusion of absolute values.
Assessment of depletion of cells and immune factors - Inefficient depletion - Check antibody concentration, reactivity, dose and time intervals of administration. Consider that depletion seldom last more than 8 days due to neutralization by generation of anti-antibodies
Statistical analysis - Inappropriate normalization and statistical tests - Outlier values - Transform logarithmically the individual number of CFU/spleen, calculate the mean Log10 CFU/spleen, and compare means by the Fisher’s Protected Least Significant Differences test (PLSD), using a maximum of 4 groups per comparison (including reference or wild-type strain and, for protection studies, both the reference vaccine and placebo control groups). The RT50 calculations should be performed in the freely available statistical program at [170]. - If controls do not give expected values, the assay should be discarded (do not remove outliers).