A reporter tomato line to track replication of a geminivirus in real time and with cellular resolution

Viruses are obligate intracellular pathogens that manipulate the cells they invade to create an environment conducive to their multiplication and spread. In plants, virus-caused diseases result in severe yield losses. At any given time during infection, only a fraction of cells supports active viral replication, frequently in a tissue-specific manner. It logically follows that, if entire organs (e.g. leaves) are considered when studying virus-induced molecular changes, significant dilution issues and an inability to distinguish cell-autonomous from systemic responses to the infection are faced, which hinder emergence of a clear overview of the impact of the viral invasion. Therefore, methods to specifically distinguish infected cells are required in plant virus research. While some viruses have been engineered to contain a reporter gene in their genomes, for example, encoding a fluorescent protein, to enable monitoring of the infection, these changes are frequently not possible and/or affect the dynamics of the infection.

Geminiviruses belong to a family of plant-infecting viruses with circular, single-stranded (ss) DNA genomes and causal agents of devastating diseases in crops worldwide. Geminiviruses replicate in the nucleus of the infected cell through a combination of rolling-circle replication (RCR) and recombination-dependent replication (RDR), utilising the plant DNA replication machinery and only one viral protein, Rep (for replication-associated). Rep recognises the origin of replication present in the intergenic region (IR) of the viral genome, recruits the necessary plant factors, and introduces a nick in the complementary strand of the double-stranded replicative intermediate; RCR ensues, generating multiple copies of the viral genome, with contribution of RDR (Hanley-Bowdoin et al., 2013).

To monitor the dynamics of the geminiviral infection in space and time with cellular resolution, a simple, visual, and non-destructive method is needed. RCR can be co-opted to enable the Rep-dependent replication of extra-chromosomal replicons (ECRs) containing sequences of interest. In this approach, the sequence of choice is flanked by two direct repeats of the viral IR; when Rep is provided in trans, through transgenic expression or during infection, the replication of this sequence as an ECR leads to the accumulation of the encoded protein to very high levels (Figure 1a; Morilla et al., 2006). Transgenic reporter plants containing this type of construct have been successfully employed to monitor the geminiviral infection and to evaluate virulence in reverse-genetic experiments in model plants (Kato et al., 2020; Krenz et al., 2015; Lozano-Durán et al., 2011; Morilla et al., 2006). However, these approaches are still lacking in crop species.

Figure 1

Open in figure viewerPowerPoint

Generation and characterisation of 2IR-DsRed transgenic tomato plants. (a) 2IR-DsRed construct. IR: TYLCV intergenic region; 35S pro: 35S cauliflower mosaic virus (CaMV) promoter; NosT: NOS terminator. Upon TYLCV infection (‘infected plants’), the viral Rep specifically binds the IR, triggering the generation of DsRed extra-chromosomal replicons (ECRs), which results in an overexpression of the DsRed transgene and accumulation of the fluorescent protein. In the absence of TYLCV (‘non-infected plants’), basal DsRed accumulation is low. (b) Symptoms in hydroponically grown TYLCV-infected 2IR-DsRed plants. Images were taken at 9, 16, and 23 days post-inoculation (dpi). Scale bar = 10 cm. The DsRed signal was observed under a fluorescence stereomicroscope in representative organs both in the absence (Mock) and presence (TYLCV) of the virus. Images from roots and leaves are shown in (c); images from hand-sectioned petioles and stems are shown in (d). Scale bar = 1 mm. (e) Representative photographs of petiole and stem sections under the confocal microscope. Samples were taken at 14 dpi. The arrows indicate phloem cells supporting viral replication. Scale bar = 200 μm. Accumulation of viral strand (VS) (f, g) and complementary strand (CS) (h, i) measured by two-step anchored quantitative PCR (qPCR) (Rodríguez-Negrete et al., 2014); DsRed (j, k), quantified by qPCR. Values are presented relative to the 25S ribosomal DNA interspacer (ITS). Samples were collected from the most apical leaves of two non-infected plants (Mock P1 and P2) and three infected plants of two independent 2IR-DsRed lines (TYLCV, P1, P2 and P3) at 9, 16, and 23 dpi. DNA extraction and qPCR were performed as in Wu et al. (2021). For viral infection assays in (b–k), Agrobacterium carrying the TYLCV infectious clone (OD₆₀₀ = 0.1) was syringe-inoculated in the stem of 2- to 3-week-old plants. Primers are listed in Table S1. (l) Progression of DsRed signal in 2IR-DsRed plants. Three levels of expression, high (red), low (pink), and none (green, leaves; and grey, roots) are indicated. Numbers mark leaf position with respect to the inoculation point.

With the aim to facilitate tracking of the viral invasion in vivo and in real time in a crop severely affected by a geminivirus-caused disease, we have generated transgenic reporter Solanum lycopersicum (tomato, cv. Moneymaker) plants containing a construct comprising two direct repeats of the IR of the tomato-infecting tomato yellow leaf curl virus (TYLCV) flanking a cassette to express the Discosoma red fluorescent protein (DsRed) (Figure 1a), named 2IR-DsRed (described in Pérez-Padilla et al., 2020). During infection, the TYLCV Rep protein specifically recognises the IRs flanking the cassette and triggers the production of ECRs, leading to the high accumulation of the fluorescent protein (Figure 1a). Therefore, induction of DsRed accumulation directly correlates with Rep activity and hence viral replication, allowing monitoring of the development of infection throughout the plant in a simple visual manner.

Next, we used these plants to follow the TYLCV infection using DsRed accumulation as a proxy for viral replication. Two independent transgenic lines were grown in hydroponic conditions and agroinoculated with a TYLCV infectious clone at 2 weeks post-germination, and DsRed fluorescence was monitored in both aerial parts and roots at 9, 16, and 23 days post-inoculation (dpi). The typical reduction in growth was observed in infected plants (Figure 1b). Strong DsRed fluorescence was detected in TYLCV-infected plants associated to the vasculature in leaves, stems, petioles and roots (Figure 1c,d), while the background fluorescence in mock-inoculated plants disappeared; the progression of DsRed fluorescence, indicating viral replication, is summarised in Figure 1l. Notably, active viral replication can be observed both in aerial parts and in the root at all three sampled times, with a peak between 2 and 3 weeks (Figure 1c; Figure S1). Observation of petiole cross-sections and stem cross- and longitudinal sections under the confocal microscope reveals that single phloem cells support viral replication (Figure 1e; Figure S2). Importantly, visual DsRed fluorescence correlates with viral DNA accumulation (both viral strand, VS, and complementary strand, CS, only present in the replicative intermediate) as well as with the accumulation of DsRed-containing ECRs (Figure 1f–k). Therefore, 2IR-DsRed tomato plants are a reliable tool to track replication of TYLCV in tomato, its natural host, in real time, in a non-destructive and cost-effective manner, and with cellular resolution, which could be used to evaluate how different factors, abiotic or biotic, affect the viral infection in mid- or high-throughput approaches. In addition, the presence of a fluorescent marker would enable isolation of infected cells supporting active replication specifically, by fluorescence-assisted cell sorting (FACS) or laser-assisted microdissection, for molecular analyses. Virus-induced gene silencing (VIGS), a reverse-genetics method applicable to tomato plants, can also be used in combination with 2IR-DsRed reporter plants for the identification of plant genes required for the viral invasion in a simple, medium-throughput, and inexpensive manner, as previously demonstrated for a similar system in Nicotiana benthamiana (Lozano-Durán et al., 2011).

Considering the results presented here, we believe that 2IR-DsRed plants offer an affordable system to monitor TYLCV infection in tomato in a fast and simple manner, enabling screens for factors impacting this process and the identification and isolation of infected cells, and therefore represent a valuable tool for plant virus research. The characterisation of these lines has also shed light on the dynamics of active TYLCV replication in different organs during infection and uncovered sustained viral replication in tomato roots.