Growth Modeling of Weissella viridescens by Real-Time

Quantitative PCR (qPCR)

Wiaslan Figueiredo Martins1, Danielle de Sousa Severo1, Fernanda Cunha Marques1

Gláucia Maria Falcão de Aragão1

1
Federal University of Santa Catarina – Department of Chemical Engineering and Food
Engineering – Biochemical Engineering Laboratory. Florianópolis/SC, Brazil.
glaucia.aragao@ufsc.br

The lactic acid bacteria (LAB) Weissella viridescens has been reported as responsible
for spoilage of meat products. In order to identify, quantify and model this bacterium
growth in culture medium, a SYBR-Green Real-Time Quantitative PCR (qPCR)
procedure has been developed having the recN gene as a target. To confirm the
efficiency and sensitivity of the SYBR-Green based assay, it was performed a melting
curve analysis to check the specificity of the amplification reaction during the qPCR.
The growth curve for a W. viridescens ATCC 12706T pure culture enumerated by qPCR
was compared to that enumerated by plate counts. Experiment was conducted in optimal
growth temperature (30 °C) until the stationary phase. Baranyi and Roberts’ model was
fitted to growth data using Matlab® software. The results showed that the primers were
specific for W. viridescens with specific signals in melting temperatures of 81.3 ± 0.1°C
for this species. Standard curves presented efficiency values of 112% and suitable
correlation coefficients (R2> 0.99). Limit of detection is 1.2 log DNA copy number,
corresponding to 3.7 log CFU, a suitable CFU enumeration range for spoilage LAB,
considering that they should be initially present in meat products. The statistically
significant difference (p-value of <0.05) between qPCR and plate count was observed
only in the exponential phase (15 hours of cultivation), corresponding to 7.95 and 7.6
log CFU, respectively. The model presented a good fit (R2> 0.99) for both curves.
According to the statistical indices, curve by qPCR presented values slightly bigger
(accuracy = 1.5917 and RMSE = 0.2780 log CFU) than the curve by plate count
(accuracy = 1.4341 and RMSE = 0.2364 log CFU), except for the Bias value (Bias = 1.0
log CFU) for both. Curve by qPCR presented higher values for growth parameters (µmax
= 0.87 ± 0.17 h-1 and λ = 4.15 ± 0.82 h) than the curve by plate count (µmax = 0.74 ±
0.05 h-1 and λ = 2.09 ± 0.9 h). Indeed, curve enumerated by qPCR and by plate counts
are convergent for W. viridescens at 30 °C. The data presented here shows that the
SYBR-Green was suitable for using in qPCR assay which provided a specific and
efficient method for W. viridescens detection and quantification. These results highlight
the importance to combine molecular techniques with classical and predictive
microbiology, to provide a new tool for studying the meat products spoilage microbiota.