Spotting method as a high throughput alternative to the conventional spread plating method
Résumé
Background
In vitro time-kill experiments are commonly performed to obtain information about the dynamics of a drug antimicrobial activity. It is
however labor intensive due to the conventional spread plating method used to count bacteria. In this work we compare a spotting
method with the conventional spread plating method.
Methods
Time-kill experiments with 4 strains (2 Pseudomonas aeruginosa, 1 Escherichia coli and 1 Klebsiella pneumoniae), 4 antibiotics
(ceftazidime, ciprofloxacin, colistin and tobramycin), 6 concentrations (0x to 4x MIC) and 6 time-points (0 to 30h) were performed. Each
time-kill was replicated at least 3 times. At each time-point, 3 samples were taken.
Each sample was then diluted serially by factors of 10 from 100 to 10-7. After that 2 plating methods were performed. The conventional
spread plating method consisted in plating 25 μL of 3 selected dilutions into an agar plate (Figure 1a). The spotting method consisted
of plating 1 drop (~1μL) of all dilutions via a 48-pin microplate replicator (Figure 1b).
Variability of each method was estimated by computing the standard deviation of standardized observed bacterial densities for each
method. The difference observed between the two methods was evaluated by performing linear regression and applying the “power
model” [1] statistical test on the estimated slope.
Results
The spotting method was slightly more variable than conventional spread plating (log10(CFU/mL) standard deviation: 0.30 vs 0.13) and
predicted slightly lower bacterial counts (average bias: -0.23 log10(CFU/mL)). However, the linear regression was not significantly
different from the identity line since the 90% confidence interval of the regression slope [0.930–0.953] was within the reference interval
[0.87-1.13] [1]. Comparisons of bacterial densities observed with both methods can be found on figures 2 and 3.
The spotting method was higher throughput since it was possible to perform 40 to 60 time-kill experiments per week while only 8 to 16
time-kill experiments per week were possible with the conventional spread plating method.
Conclusions
The spotting method yields bacterial counting results comparable with the conventional plating method while being significantly
higher throughput.
Domaines
PharmacologieOrigine | Fichiers produits par l'(les) auteur(s) |
---|