Pre-Symptomatic Detection of Viral Infection in Tobacco Leaves Using PAM Fluorometry

Abstract

Chlorophyll fluorescence imaging was used to study potato virus X (PVX) infection of Nicotiana benthamiana. Infection-induced changes in chlorophyll fluorescence parameters (quantum yield of photosystem II photochemistry (ΦPSII) and non-photochemical fluorescence quenching (NPQ)) in the non-inoculated leaf were recorded and compared with the spatial distribution of the virus detected by the fluorescence of GFP associated with the virus. We determined infection-related changes at different points of the light-induced chlorophyll fluorescence kinetics and at different days after inoculation. A slight change in the light-adapted steady-state values of ΦPSII and NPQ was observed in the infected area of the non-inoculated leaf. In contrast to the steady-state parameters, the dynamics of ΦPSII and NPQ caused by the dark–light transition in healthy and infected areas differed significantly starting from the second day after the detection of the virus in a non-inoculated leaf. The coefficients of correlation between chlorophyll fluorescence parameters and virus localization were 0.67 for ΦPSII and 0.76 for NPQ. In general, the results demonstrate the possibility of reliable pre-symptomatic detection of the spread of a viral infection using chlorophyll fluorescence imaging.

Comparison of the Efficiency of Hyperspectral and Pulse Amplitude Modulation Imaging Methods in Pre-Symptomatic Virus Detection in Tobacco Plants

“…PAM fluorimetry seems to be a fairly sensitive method for detecting infectious agents inside plants [19,24,51]. The high sensitivity of PAM for pathogen detection at the pre-symptomatic stage has been demonstrated in our previous work [26]. The current results demonstrate that F v /F m , which reflects the structural integrity of PSII, does not change throughout the experimental period (Figures 7 and 8).…”

“…ΦPSII, qP, NPQ, and qN were recorded both in the light-adapted state and in the transition state-40 s after the actinic light was switched on. The latter, as shown by our previous study, provides the greatest contrast between infected and healthy areas [26]. Images of Fv/Fm, as well as ΦPSII and NPQ in the transition state, along with RGB and fluorescent images, are presented in Figure 7.…”

“…The fluorescence imaging method was used for pre-symptomatic detection of infection development in ginseng (Panax quinquefolius L.) [24], beans (Phaseolus vulgaris L.) [25], and melon (Cucumis melo L.) [19]. Our previous study also demonstrated the efficiency of PAM imaging in detecting potato virus X (PVX) in tobacco (Nicotiana benthamiana) at the pre-symptomatic stage [26].…”

Chlorophyll a Fluorescence: A Method of Biotic Stress Detection

“…Ralstonia solanacearum is a devastating disease for tomato seedling; in a study using CFI, F v /F m values between 0.55 and 0.65 were detected, and as a result, the infection was predicted 2 days before visual symptoms [75]. Likewise, CFI was used for the detection of potato virus X (PVX) in Nicotiana benthamiana plants; the study demonstrated that Φ PSII and NPQ parameters in infected areas were different significantly starting from the second day after the detection of the virus, considering the CFI as reliable pre-symptomatic detection a viral infection [76]. Also, the CFI was able to detect biotic and abiotic stress in tomato plants, as was reported by Moustaka et al [77].…”

Plant disease symptom segmentation in chlorophyll fluorescence imaging with a synthetic dataset

“…The majority of such images were from the first post-inoculation day, when disease severity varied from 0.012% to 0.24%, with a mean of 0.055%. Several studies showed that in pre-symptomatic detection of plant diseases, differences between healthy and diseased plant tissues become statistically significant from the second day after inoculation (Boureau et al, 2002;Grishina et al, 2021). Thus, we proposed to pre-process the dataset in two steps: 1) we applied morphological operations to eliminate corrupted pixels and to uniform lesions if they included small holes; 2) we eliminated images with S≤0.24% .…”