Comparison of real-time PCR and droplet digital PCR for the detection of Xylella fastidiosa in plants

Enora Dupas; Bruno Legendre; Valérie Olivier; Françoise Poliakoff; Charles Manceau; Amandine Cunty

Real-time PCR and ddPCR technology provided equivalent limits of detection for Xf in the 321 following matrices: O. europaea, P. myrtifolia and R. officinalis (Table 1, Table 3 ). The ddPCR 322 technology presented a slightly higher limit of detection of 0.5 log for L. angustifolia. However, 323 a decrease in the limit of detection for Xf of 0.5 log for Q. ilex and the bacterial suspension 324 wereobserved. In the conditions of DNA extraction used for this study, and according to the 325 volume of DNA added to the real-time PCR assay, the theoretical limit of detection should be 326 1x10 2 b/mL for the five plant matrices. These results revealed that L. angustifolia and P. 327 myrtifolia may contain fewer real-time PCR inhibitors than Q. ilex and R. officinalis. Moreover, 328 the limit of detection of the bacterial suspension was 5x10 3 b/mL, meaning that the QuickPick 329 extraction kit may not be 100% efficient to extract the DNA of bacteria in pure culture. 330 331 The 22 naturally infected samples were then analyzed by ddPCR. The presence of Xf was 341 detected in 21 of them, including samples P13 and Q04, with at least two droplets, and a total 342 concentration ranging from 4.41x10 2 cp/mL to 1.61x10 7 cp/mL, confirming the ability of 343 ddPCR to detect Xf in naturally infected samples. As only one positive droplet was detected for 344 sample C01, this sample was considered negative by ddPCR. With the exception of samples 345 L04, P06 and P13, the decrease in the Ct value was correlated with an increase in the quantity 346 of DNA detected by ddPCR (Table 4 ). For each matrix the results obtained by ddPCR and real-347 time PCR were compared and were highly correlated (Figure 3) . 348