Virus–Host Interactions and Genetic Exchange in Mixed Infections of Tomato Yellow Leaf Curl Virus (TYLCV), Tomato Leaf Curl New Delhi Virus (ToLCNDV), and Tomato Chlorosis Virus (ToCV)

Abstract

Tomato yellow leaf curl virus (TYLCV), tomato leaf curl New Delhi virus (ToLCNDV), and tomato chlorosis virus (ToCV) are emerging viruses that cause significant damage to tomato (Solanum lycopersicum). TYLCV and ToLCNDV are single-stranded DNA viruses from the genus Begomovirus, family Geminiviridae, while ToCV is an RNA virus from the genus Crinivirus (family Closteroviridae). These viruses share overlapping geographic ranges, vectors (the whitefly Bemisia tabaci), and host plants, making mixed infections common. This study investigated interactions between TYLCV and ToLCNDV and between ToLCNDV and ToCV in mixed infections of susceptible and TYLCV-resistant tomato genotypes. We evaluated infection, disease development, trans-replication of genome components, and genetic exchange. Our results showed no significant synergistic or antagonistic interactions, complementation, or interference between the viruses. TYLCV resistance in tomato genotypes remained stable. The DNA-B component of ToLCNDV exhibited impaired functionality and was not complemented by TYLCV. No evidence was found that the crinivirus tomato chlorosis virus (ToCV) enhances ToLCNDV infection, suggesting limited interactions despite shared vectors. Genetic exchange was detected in defective DNA (def-DNA) molecules using high-throughput sequencing (HTS), indicating potential genetic interactions between these viruses. These findings suggest that mixed infections do not pose immediate concerns for increased pathogenicity but highlight the ecological implications of genetic exchange, warranting further study of the evolutionary consequences of such interactions in mixed-virus environments.

1. Introduction

The emergence of plant viruses presents a significant global threat to agriculture, leading to substantial damage in economically important crops [1,2]. Whitefly-transmitted viruses, including begomoviruses (Begomovirus, family Geminiviridae) and criniviruses (Crinivirus, family Closteroviridae), have become more prominent in recent decades [3,4]. Their global spread is largely driven by the expansion of cryptic species within the Bemisia tabaci Genn. vector complex, particularly Middle East Asia Minor 1 and Mediterranean species, which have successfully invaded new ecological niches [5,6,7]. Notable examples include single-stranded DNA begomoviruses like tomato yellow leaf curl virus (TYLCV) and tomato leaf curl New Delhi virus (ToLCNDV), as well as the single-stranded RNA crinivirus tomato chlorosis virus (ToCV), all of which severely impact tomato (Solanum lycopersicum L.) production worldwide [4,8,9,10,11].

Begomoviruses are divided into Old World and New World groups, which differ in their geographical distribution, genome organization, and genetic diversity, with Old World viruses generally showing greater genetic variability.

Begomoviruses are a major constraint for vegetable, root, and fiber crops [4,11]. Most Old World begomoviruses, such as TYLCV, are monopartite, containing a single genomic DNA component. However, bipartite begomoviruses like ToLCNDV, which possess two components (DNA-A and DNA-B), are also spreading in this region [12]. The begomovirus genome, approximately 2.6 to 2.8 kb in size, encodes open reading frames (ORFs) arranged bidirectionally from a non-coding intergenic region [13], which contains key elements for replication and transcription [13,14]. This includes the origin of replication nicked by the replication initiation protein, Rep, to initiate virus DNA replication and iterated DNA motifs (iterons) that are specifically recognized by the iteron-related domain in the N-terminal region of the Rep for effective replication [15].

DNA-A of bipartite begomoviruses encodes proteins essential for virus encapsidation, replication, pathogenesis, and transmission: Rep and replication enhancer protein (REn) (AC1 and AC3 ORFs, respectively), transcription activator protein (TrAP, C2 ORF), AV2 precoat protein (AV2 ORF), coat protein (CP, V1 ORF), and protein AC4 (AC4 ORF). DNA-B encodes movement proteins: the nuclear shuttle protein (NSP) and the movement protein (MP) (BC1 and BV1 ORFs, respectively) [13]. Monopartite begomoviruses have a single genomic component that resembles the DNA-A of bipartite begomoviruses [13]. Recently, additional small proteins that can play relevant biological roles have been described in begomoviruses, which implies that the repertoire of proteins might be expanded [16,17,18]. Notably, defective DNAs (def-DNAs), derived from extensive recombination events, are often associated with infections and may influence virus biology [19].

Managing begomovirus-induced damage in tomato plants has primarily depended on genetic resistance, especially through dominant resistance genes like Ty-1, Ty-3, and Ty-2, which restricts viral infections [20,21]. However, the evolution of begomoviruses, frequently driven by recombination and mixed infections, can undermine these resistance strategies [22]. Mixed infections often lead to synergistic interactions, reassortment [23], or recombination, all of which can modify plant–virus interactions and intensify epidemics [24,25,26]. Recombination has been reported in both closely related [27,28,29] and distantly related [30,31] begomoviruses, including both monopartite and bipartite members [23,32]. Additionally, mixed infections of begomoviruses and criniviruses can occur because they share the whitefly vector and host plant and have been linked to pathological alterations [33,34]. These processes can give rise to novel pathogenic variants with potentially severe consequences [27,30,35,36].

Recently, ToLCNDV isolates have emerged in the Mediterranean Basin, some of which are poorly adapted to tomato plants [37,38]. However, mixed infections involving ToLCNDV and TYLCV have been reported to enhance ToLCNDV’s infection capabilities in tomatoes, highlighting the potential for synergistic interactions [39]. Considering the contrasting biological characteristics of ToLCNDV isolates and their varying adaptability to tomato plants, the broader implications of such interactions remain unclear.

Understanding the mechanisms of recombination and trans-replication between TYLCV and ToLCNDV is essential for real-world crop management because these processes can give rise to novel viral variants with enhanced pathogenicity or broader host ranges. Such variants may overcome existing resistance genes deployed in tomato breeding programs and compromise integrated pest management strategies. Monitoring these genetic interactions is therefore key to anticipating viral emergence and improving crop resilience.

In this study, we examined the effects of mixed infections between TYLCV and ToLCNDV in TYLCV-susceptible and TYLCV-resistant tomato plants. Given that begomoviruses and criniviruses frequently co-infect tomato plants due to shared insect vectors, we also assessed whether ToCV could enhance ToLCNDV infection. Additionally, we investigated the potential for genetic exchange between ToLCNDV and TYLCV through reassortment and recombination using high-throughput sequencing (HTS), aiming to enhance our understanding of their ecological and evolutionary dynamics. Our findings contribute to advancing ecosystem-scale analyses of plant–virus interaction networks and informing strategies for managing these emerging viral threats.

2. Materials and Methods

2.1. Plant Material and Experimental Periods

TYLCV-susceptible tomato lines (cv. Moneymaker, cv. Rondeño, and cv. Marmande; IHSM-CSIC seedbank collection) and TYLCV-resistant lines (Ty-1 F1, with heterozygous Ty-1/ty-1 resistance, kindly supplied by M.J. Díez, Universidad Politécnica de Valencia, Spain; and ABL21-5, with homozygous Ty-2/Ty-2 resistance, IHSM-CSIC breeding program) were used. Nicotiana benthamiana D. plants were also included as susceptible controls in the experiments, as they are susceptible to TYLCV, ToLCNDV, and ToCV. Plants were grown in an insect-proof glasshouse under natural lighting with loose temperature control (22–27 °C day, 17–20 °C night).

Mixed-infection experiments were conducted in two independent assays over different periods: TYLCV + ToLCNDV assays were performed from March to June 2016 (Assay 1) and from September to December 2016 (Assay 2). Mixed infections of TYLCV + ToLCNDV DNA-B were conducted from June to July 2017 (Assay 1) and from October to November 2017 (Assay 2). ToCV + ToLCNDV assays were carried out from October to November 2016 (Assay 1) and from April to May 2017 (Assay 2).

2.2. Whiteflies, Virus Isolates, and Virus Inoculation

Virus-free whitefly adult individuals were obtained from a B. tabaci Mediterranean species [6] colony originated from individuals collected in Málaga, Southern Spain, and reared on melon (Cucumis melo L. ANC42, IHSM seedbank collection) plants within insect-proof cages, in an insect-proof glasshouse with loose temperature control (see above) and light supplementation when needed.

Two infectious clones, previously generated from Spanish isolates and described in earlier studies, were used for Agrobacterium tumefaciens-mediated inoculation: the Spanish isolate ES-Alm-Pep-99 of the TYLCV Israel strain (TYLCV-IL, originally described in Israel; GenBank accession number AJ489258) [40] and the Spanish ToLCNDV-ES isolate ES-Al-661-Sq-13 (DNA-A GenBank accession no. KF749223 and DNA-B GenBank accession no. KF749226) [37]. Liquid cultures of A. tumefaciens containing the infectious clones were pelleted and resuspended in agroinoculation buffer (10 mM MgCl₂, 10 mM MES pH 5.8, 2.25 mM acetosyringone) to a final OD₆₀₀ of 1.0. For single infections, inoculation was performed with either the Spanish isolate of TYLCV-IL, ToLCNDV-ES (DNA-A and DNA-B), or ToLCNDV-ES DNA-B alone. In the case of ToLCNDV-ES, the infectious clones containing DNA-A and DNA-B were mixed in a 1:1 volume ratio (v/v) prior to inoculation. For mixed infections, the cultures containing TYLCV-IL and ToLCNDV-ES DNA-A and DNA-B were mixed in equal volumes (1:1:1 ratio, v/v), each adjusted to an OD₆₀₀ of 1.0. Additionally, co-inoculation treatments included TYLCV-IL with ToLCNDV DNA-B alone, mixed in a 1:1 volume ratio (v/v), both adjusted to OD₆₀₀ o of 1.0. One hundred µL of the suspension (for both single and mixed infections) was used to agroinoculate each test plant by stem puncture [41]. Mixed infections were performed to explore potential interactions between the two viruses in the host.

For ToLCNDV-ES and TYLCV-IL co-infection in TYLCV-susceptible, TYLCV-resistant tomato plants and the susceptible control N. benthamiana, two different assays were conducted. In Assay 1, two inoculation strategies were employed: simultaneous and sequential inoculation. In the simultaneous approach, plants were inoculated with both TYLCV-IL and ToLCNDV-ES at the two- to three-leaf growth stage using Agrobacterium tumefaciens-mediated inoculation. For the sequential method, plants were first inoculated with TYLCV-IL at the two- to three-leaf growth stage, followed by ToLCNDV-ES inoculation at the four- to five-leaf growth stage. Assay 2 involved simultaneous co-inoculation of both viruses at the two- to three-leaf growth stage, mirroring the simultaneous method in Assay 1. For the co-inoculations of TYLCV-IL with ToLCNDV DNA-B alone, the two assays were performed following the simultaneous approach. Co-inoculations of ToCV and ToLCNDV (DNA-A and DNA-B) were performed sequentially, with ToCV being inoculated first at the three- to four-leaf stage, followed by ToLCNDV-ES at the five- to six-leaf stage. In all experiments, 10 plants per condition were inoculated, except in the experiment that included separate inoculations with TYLCV-IL, ToLCNDV (DNA-A and DNA-B), and ToLCNDV-ES DNA-B, where only 5 plants were used per condition.

For ToCV inoculations, B. tabaci-mediated transmission was used, along with the Spanish isolate Pl-1-2 obtained from a naturally infected tomato plant collected in 1997 in Málaga and maintained at the IHSM on Moneymaker tomato by periodic transmission [42]. This method was chosen because, unlike TYLCV and ToLCNDV, no infectious clone of ToCV is available, and therefore vector-mediated inoculation remains the most biologically relevant and technically feasible approach to achieve consistent ToCV infection. Viruliferous whiteflies were obtained by mass feeding of virus-free B. tabaci adults (48 h acquisition access period, within insect-proof cages) on Moneymaker tomato plants infected with ToCV 30 days before they were used for virus acquisition. Test plants were inoculated by using clip-on cages, with one cage containing 30 viruliferous whiteflies per test plant for a 48 h inoculation access period (IAP). Following the IAP, plants were maintained in an insect-proof glasshouse (see above) and sprayed with insecticide combinations to eliminate whitefly eggs, larval stages, and adults. Treatment routines alternated insecticides with different modes of action to prevent the generation of insecticide resistances, using either pyriproxyfen or spiromesifen combined with either flupyradifurone, spirotetramat, or acetamiprid. Plants were inoculated at the two- to six-leaf growth stage. Mock-inoculated control plants were obtained following the same inoculation procedures but using virus-free whiteflies or virus-free A. tumefaciens.

2.3. Symptom Evaluation and Virus Detection by Tissue Blot Hybridization

TYLCV infection was evaluated using a 0–5 arbitrary symptom severity rating scale adapted from Friedmann et al. [43] as follows: 0 = asymptomatic; 1 = slight yellowing and mosaic on top leaves, no curling; 2 = mild yellowing, mosaic and/or slight curling on top leaves; 3 = moderate yellowing and/or moderate curling on top leaves; 4 = severe yellowing, blistering, and/or severe curling with some leaf size reduction on top leaves of the main stem and/or one branch; 5 = severe yellowing, blistering, very severe curling, leaf deformation and size reduction, and stunting (internode shortening) on top leaves of the main stem and/or one branch. Moreover, virus infection was analyzed by tissue blot hybridization. Fresh cross sections of petioles from the youngest newly emerged leaves of test plants were tissue-blotted onto positively charged nylon membranes (Roche Diagnostics GmbH, Mannheim, Germany). Tissue prints were performed at different time points depending on the experiment: 14, 21, 28, 38, and 48 days post-inoculation (dpi). The final sampling point varied according to the specific viral combination under study and was determined based on the time at which no further differences in hybridization signals were observed relative to the previous time point. In TYLCV and ToLCNDV co-infections, samples were collected up to 48 dpi; in TYLCV and ToLCNDV DNA-B co-inoculations, up to 28 dpi; and in ToCV and ToLCNDV assays, up to 38 dpi. The selected final time point in each case reflects the latest time at which an increase in hybridization signal was observed. Nucleic acids were UV-cross-linked and hybridized with a specific probe. For TYLCV-IL detection, a DIG-labelled DNA probe specific to TYLCV-IL was used [27]. The presence of ToLCNDV-ES DNA-A and DNA-B components was analyzed using a DIG-labelled DNA probe based on the common region and DIG-labelled DNA probes specific to each viral component [37]. ToCV presence was analyzed with a DIG-labelled RNA probe specific to ToCV [44].

2.4. Detection of Recombination Events in Mixed Infections by HTS

To evaluate potential recombination events between TYLCV-IL and ToLCNDV-ES by HTS (Supplementary Figure S1), the youngest apical leaves of co-infected plants confirmed positive for both viruses at 48 dpi were sampled at 90 and 110 dpi and pooled. Total DNA was extracted using a CTAB-based purification method [45] from 5 Rondeño tomato plants (TYLCV-IL + ToLCNDV-ES-Tom sample) and 10 N. benthamiana plants (TYLCV-IL + ToLCNDV-ES-Nb sample). A control sample consisting of a combination of N. benthamiana plants with single infections of TYLCV-IL (5 plants) and ToLCNDV-ES (5 plants) (TYLCV-IL/ToLCNDV-ES-Nb combined sample) was also included. Biological replicates were not included, as the samples were pooled plants. This pooling strategy reflected the variability in mixed infections across the plant populations, enabling the detection of recombination events but not statistical comparisons of viral accumulation.

Viral loads were quantified by qPCR using 2 μL of DNA extract (10 ng/μL), 10 μL of 2× SsoFast EvaGreen supermix (Bio-Rad, Hercules, CA, USA), 1 μL each of 10 μM forward and reverse primers, and 6 μL of ultrapure H₂O, in a final volume of 20 μL per reaction. The qPCR conditions consisted of an initial denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C for 10 s and 60 °C for 15 s on a LightCycler^® 96 system (Roche). Primer pairs used for TYLCV and ToLCNDV quantification were custom-designed using the Primer-Blast tool (NCBI). For TYLCV, the primers were TYLCV_F_2237 (5′-TTGGCAGATTGCTGACCTCC-3′; positions 2237–2256 of AJ489258) and TYLCV_R_2414 (5′-ACCAACGGTTCTTCGACCTG-3′; positions 2395–2414 of AJ489258). For ToLCNDV-ES DNA-A, the primers were ToLCNDV-A_F_1733 (5′-TTGCTTTGCCAGTCACGTTG-3′; positions 1733–1752 of KF749223) and ToLCNDV-A_R_1906 (5′-CAAAACAATGTGGGCGCGTT-3′; positions 1906–1925 of KF749223). For ToLCNDV-ES DNA-B, the primers were ToLCNDV-B_F_2149 (5′-CGACACACCCAATGACACGA-3′; positions 2130–2149 of KF749226) and ToLCNDV-B_R_2333 (5′-TCGTAGAAAAATGAGTGCCGT-3′; positions 2313–2333 of KF749226). Standard curves for TYLCV-IL were generated using a pBSK construct containing a monomeric TYLCV-IL (ES-Alm-Pep-99 isolate; GenBank AJ489258) described by Morilla et al. [46]. For the ToLCNDV-ES standard curve, a pBSK dimer of DNA-A (KF749223) and a pBSK monomer of DNA-B (KF749226) from the ES-Al-661-Sq-13 isolate described by Fortes et al. [47] were employed. Samples and standards were analyzed in triplicate.

To prepare DNA for HTS (Figure 1), circular viral DNA was selectively amplified from 1 μL of total DNA extract using rolling circle amplification (RCA) with φ29 DNA polymerase (TempliPhi kit, GE Healthcare, Chicago, IL, USA), following the manufacturer’s instructions. RCA products were used directly for library preparation. Reactions were performed in duplicate and pooled to minimize random errors introduced by φ29 DNA polymerase. Sequencing was carried out at Macrogen (Seoul, Republic of Korea) using equimolar sequencing libraries prepared with a TruSeq DNA PCR-free kit (550 bp insert; Illumina, San Diego, CA, USA). Libraries were normalized to 4 nM by dilution with 10 mM Tris-HCl (pH 8.5) and sequenced as paired-end 2 × 300 bp reads on an Illumina MiSeq platform.

Raw reads were preprocessed using QuasiFlow software (version 1.0) [48] (Figure 1) with default settings to remove low-quality, ambiguous, and low-complexity sequences, as well as linkers, adapters, and host-derived sequences. Only reads aligning to viral reference sequences were retained. Sequences shorter than 95 nucleotides or with a PHRED quality score below 26 were discarded.

Cleaned reads were aligned to an artificial viral reference constructed by concatenating the genomic sequences of individual viral components or segments. TYLCV-IL (GenBank AJ489258), ToLCNDV-ES DNA-A (GenBank KF749223), and ToLCNDV-ES DNA-B (GenBank KF749226) were used as reference sequences, joining them end to end to form a single continuous sequence. This strategy was used to achieve the mapping of sequencing reads in the mixed infection, as the software can only process one reference sequence. The Virema Module in QuasiFlow was used to annotate potential recombination events in DNA molecules. Circular representations were generated using CanvasXpress, with annotations added in Inkscape [49].

To validate QuasiFlow’s ability to reliably detect recombinants, we analyzed an RCA-amplified DNA sample (Mm-A) from a previous study on Moneymaker tomato plants co-infected with the Mild strain of tomato yellow leaf curl virus (TYLCV-Mld) and tomato yellow leaf curl Sardinia virus (TYLCSV). This sample was characterized by Garcia-Andres et al. [50], who identified recombinant genome types R1 to R4 between these two viruses. Since only RCA products were available for this sample, DNA quantification was not performed. For analysis, a reference sequence was constructed by concatenating the complete genomic sequences of TYLCV-Mld (GenBank AF071228) and TYLCSV (GenBank Z25751).

2.5. Cloning and Sequencing of def-DNA and Recombinant Molecules

RCA products were obtained from the pooled N. benthamiana samples with mixed infections of TYLCV-IL and ToLCNDV-ES (Supplementary Figure S2). These RCA products were digested with SacI or SalI restriction endonucleases, selected based on the presence of restriction sites identified in the predominant def-DNA molecules detected during the HTS analysis. DNA fragments ranging from 1 to 1.5 kb (SacI digestion) and 0.8 to 1.5 kb (SalI digestion) were purified from agarose gels and ligated into the cloning sites of the pBluescript SK vector (Stratagene, La Jolla, CA, USA). The cloned inserts were sequenced by Macrogen Inc. (Seoul, Republic of Korea) using M13 forward and reverse primers.

The presence of cloned def-DNA molecules was validated by PCR using non-overlapping abutting primers [51] targeting junction/recombinant sites (Supplementary Table S1). Amplification was performed in a T100 Thermal Cycler (Bio-Rad, Hercules, CA, USA) in a 20 μL reaction volume containing NH4 Reaction Buffer (Meridian Bioscience Inc., Cincinnati, OH, USA), 1.5 mM MgCl₂, 0.2 mM of each dNTP, 0.2 μM of each primer, 1.5 U of Biotaq DNA Polymerase (Meridian Bioscience Inc.), and 10 ng of total DNA extract as template. The cycling parameters were as follows: initial denaturation at 95 °C for 5 min, followed by 34 cycles of denaturation at 95 °C for 30 s, annealing at 60 °C (for primer pair MA-2854/MA-2855), 55 °C (for primer pairs MA-2856/MA-2857 and MA-2858/MA-2859), or 65 °C (for primer pair MA-2860/MA2861) for 30 s, and extension at 72 °C for 1 min. A final extension was performed at 72 °C for 5 min. The amplified fragments were cloned into a pGEM-T Easy Vector system (Promega Biotech Ibérica, Madrid, Spain) and sequenced using M13 forward and reverse primers.

2.6. Statistical Analyses

Statistical analyses were performed using IBM SPSS version 29.0 (IBM, Armonk, NY, USA). TYLCV symptom severity levels were analyzed via one-way ANOVA followed by Tukey-b post hoc tests. Binary response data (0 = non-infected, 1 = infected) were subjected to generalized linear model (GzLM) analyses using Logit as the link function and a binomial distribution with multiple comparisons performed using LSD tests.

3. Results

3.1. ToLCNDV-ES and TYLCV-IL Co-Infection in TYLCV-Susceptible, TYLCV-Resistant Tomato Plants and N. benthamiana

We examined the interactions between TYLCV-IL and ToLCNDV-ES in both TYLCV-susceptible and TYLCV-resistant tomato genotypes, as well as in N. benthamiana plants.

Consistent with previous findings [29], TYLCV-IL effectively infected TYLCV-susceptible tomato plants, inducing characteristic symptoms (Figure 2 and Figure 3). As already observed for the TYLCV-resistant genotypes used, TYLCV-IL was not detected in plants carrying the Ty-2 resistance gene, suggesting it failed to infect, while it did infect those with the Ty-1 gene although inducing only mild symptoms (Figure 2 and Figure 3). During sequential inoculation in Assay 1, fewer Rondeño plants were infected by TYLCV-IL. However, overall, there were no significant differences in infection rates (GzLM, LSD test, p < 0.05) or symptom severity (ANOVA, Tukey-b test, p < 0.05) between the non-infected and the infected condition across the tested genotypes (Figure 2 and Figure 3).

ToLCNDV-ES exhibited limited ability to infect TYLCV-susceptible tomato genotypes, with systemic infection observed only in a few Rondeño plants, in addition to what is observed in N. benthamiana plants (Figure 3). Simultaneous inoculation with TYLCV-IL slightly increased ToLCNDV-ES infection rates, significantly in the Rondeño genotype (GzLM, LSD test, p < 0.05), suggesting potential functional complementation by TYLCV-IL. All plants testing positive for ToLCNDV-ES were also positive for TYLCV-IL, indicating mixed infections. Plants singly infected with ToLCNDV-ES remained asymptomatic, whereas mixed infections exhibited symptoms similar to those caused by TYLCV-IL alone (Figure 2). Interestingly, ToLCNDV-ES DNA-B was detected less frequently than DNA-A in infected plants.

In TYLCV-resistant tomato plants, ToLCNDV-ES was undetectable regardless of whether inoculation was single or mixed, indicating that there was no complementation of ToLCNDV by TYLCV-IL (plants with the Ty-1 gene) and no overcoming of resistance (plants with the Ty-2 gene) in TYLCV-resistant tomato plants inoculated with TYLCV-IL and ToLCNDV-ES. Conversely, N. benthamiana plants were readily infected by both viruses, either singly or in combination, with no significant differences in symptom expression between single and mixed infections. However, single ToLCNDV-ES infections in N. benthamiana resulted in milder symptoms (Figure 2).

Overall, our findings suggest no significant synergistic interactions between TYLCV-IL and ToLCNDV-ES, except for a slight enhancement of ToLCNDV-ES infection by TYLCV-IL in one TYLCV-susceptible tomato line.

3.2. Absence of Heterologous Trans-Replication of ToLCNDV-ES DNA-B by TYLCV-IL

Previous research has indicated that monopartite begomoviruses can trans-replicate heterologous DNA-B components, potentially modulating plant–virus interactions [36,52,53]. To assess this, we conducted two independent assays involving N. benthamiana and TYLCV-susceptible tomato genotypes, with two assays performed in Rondeño and a single assay in Moneymaker. Our results (Figure 4) show effective TYLCV-IL infections in both plant species regardless of whether they were singly inoculated or co-inoculated with ToLCNDV-ES. However, we did not detect the maintenance of ToLCNDV-ES DNA-B in any co-inoculated plants. Furthermore, the characteristic symptoms induced by TYLCV-IL were consistent across both singly inoculated and co-inoculated plants. These findings suggest that TYLCV-IL does not trans-replicate ToLCNDV-ES DNA-B. A comparative analysis of the iterons in the intergenic region and iteron-related domain of Rep between TYLCV-IL and ToLCNDV-ES revealed significant differences, which may account for the lack of trans-replication observed (Supplementary Figure S3).

3.3. Lack of Enhancement of ToLCNDV-ES Infection in Tomato Plants by the Crinivirus ToCV

Some studies have suggested that ToCV may facilitate the infection of distantly related viruses in tomato plants [44]. We therefore investigated whether ToCV could enhance ToLCNDV-ES infection in tomato plants, as mixed infections of these viruses are common due to shared insect vectors. To test this, we performed two independent co-inoculation assays with ToCV and ToLCNDV-ES (Figure 5). Efficient ToCV infections were achieved in both singly and co-inoculated tomato plants, accompanied by characteristic symptoms [10]. However, ToLCNDV-ES rarely infected tomato plants when inoculated alone and did not cause evident symptoms. Co-inoculation with ToCV or pretreatment with virus-free B. tabaci whiteflies did not significantly improve ToLCNDV-ES infection rates in any of the tested tomato genotypes (GzLM, LSD test, p < 0.05). In N. benthamiana plants, both ToCV and ToLCNDV-ES resulted in effective infections when singly or co-inoculated, but no significant increase in symptoms was observed in mixed-infected plants. These results suggest that ToCV does not enhance ToLCNDV-ES infection in tomato plants.

3.4. Recombination Events Between TYLCV-IL and ToLCNDV-ES in Mixed-Infected Plants

To explore the occurrence of recombination events between TYLCV-IL and ToLCNDV-ES in mixed infections, HTS data were analyzed using QuasiFlow software (version 1.0). Sequencing depths for bulk samples from Rondeño tomato (TYLCV-IL + ToLCNDV-ES-Tom co-infected sample) and N. benthamiana (TYLCV-IL + ToLCNDV-ES-Nb co-infected sample) plants with mixed infections ranged from 4.63 × 10⁶ to 7.41 × 10⁶ filtered reads (Supplementary Table S2). For comparison, similar sequencing depths were obtained for the bulk sample from N. benthamiana plants with single infections of TYLCV-IL or ToLCNDV-ES combined prior to DNA extraction (TYLCV-IL/ToLCNDV-ES-Nb-combined sample) and a control sample from tomato plants co-infected with TYLCV-Mld and TYLCSV (Mm-A sample). These datasets were analyzed for intermolecular and intramolecular recombination events between the two viruses.

Analysis of reads from the TYLCV-Mld + TYLCSV Mm-A sample using QuasiFlow detected intermolecular recombination events equivalent to those identified by García-Andrés et al. [50] (Supplementary Figure S4), confirming the reliability of this method for detecting bona fide recombinants. Additionally, the HTS/QuasiFlow strategy identified new recombination events not previously reported.

In the tomato sample with mixed infections (TYLCV-IL + ToLCNDV-ES-Tom), lower sequence coverage for ToLCNDV-ES DNA-B was observed compared to the N. benthamiana sample (TYLCV-IL + ToLCNDV-ES-Nb) (Figure 6a). This suggests that ToLCNDV-ES DNA-B was less abundant in the tomato plant extract. Supporting this, hybridization results (Figure 3) showed a lower accumulation of ToLCNDV-ES DNA-B compared to DNA-A in tomato plants, as fewer DNA-B-positive plants were detected than DNA-A-positive ones. Viral DNA quantification by qPCR confirmed these observations (Figure 6b). In the mixed-infected tomato sample, TYLCV was more abundant than ToLCNDV-ES DNA-A, with ToLCNDV-ES DNA-B being the least prevalent. Conversely, in N. benthamiana with mixed infections, all viral components were present at higher levels, with TYLCV being the most abundant followed by ToLCNDV-ES DNA-B and DNA-A. In the N. benthamiana combined sample with single infections, TYLCV accumulated to higher levels than ToLCNDV-ES DNA-A and DNA-B. These results suggest that in tomato plants, ToLCNDV-ES DNA-B is less represented compared to TYLCV and DNA-A, a pattern not observed in N. benthamiana.

Analysis using QuasiFlow revealed no intermolecular recombination events between TYLCV-IL and ToLCNDV-ES in the tomato sample (Figure 7a). However, intramolecular recombination events were observed in TYLCV-IL, suggesting the presence of putative def-DNAs. In N. benthamiana plants with mixed infections, both TYLCV-IL and ToLCNDV-ES DNA-B exhibited intramolecular recombination events (Figure 7b), indicating the formation of putative def-DNAs. No recombination events were detected for ToLCNDV-ES DNA-A, consistent with findings from single-infection samples (Figure 7c). Interestingly, intermolecular recombination events were identified between TYLCV-IL and ToLCNDV DNA-B but not with ToLCNDV DNA-A (Figure 7b).

A higher percentage of reads with recombination events was observed in N. benthamiana samples compared to the tomato sample (Supplementary Table S2). Specifically, the tomato sample with mixed infections exhibited approximately 5.92-fold fewer rearrangements than the mixed-infection N. benthamiana sample. Intermediate levels of recombination were detected in the N. benthamiana single-infection sample. Notably, an intermolecular recombination event between ToLCNDV DNA-B and TYLCV-IL in the N. benthamiana mixed-infection sample was supported by 655 reads (indicated by red arrows in Figure 7b), reinforcing the possibility of genetic exchange and the formation of recombinant def-DNA.

To confirm the presence of the putative defective molecules detected in silico, def-DNAs were cloned from RCA products using single-site restriction enzyme digestion followed by cloning and Sanger sequencing (Supplementary Figure S2). Intramolecular recombination events frequently detected in the control N. benthamiana mixed-infection sample were further examined (Supplementary Table S3). As shown in Figure 8a, def-DNA molecules for TYLCV-IL exhibited deletions between the V2 and Rep ORFs. Similarly, def-DNA molecules for ToLCNDV-ES DNA-B showed deletions encompassing significant portions of the BV1 and BC1 ORFs (Figure 8b). Sequences were deposited in GenBank under the following accession numbers: OR699191 to OR699196 for def-DNA molecules of TYLCV-IL, and OR699199 to OR699202, OR699204 to OR699210 for def-DNA molecules of ToLCNDV-ES DNA-B.

The presence of these def-DNAs was confirmed in a number of cases by PCR amplification directly from the DNA extract (Supplementary Figure S2) using non-overlapping abutting primers followed by cloning and sequencing of the amplified fragment (sequences marked with asterisks in Figure 8a,b). Sequence data are available under the following GenBank accession numbers: OR880353 and OR880354 for def-DNA molecules of TYLCV-IL, OR880355 and OR909915 for def-DNA molecules of ToLCNDV-ES DNA-B. Notably, all the def-DNA characterized here are about half the size or smaller than the full genomic DNA, retaining the intergenic/common region necessary for rolling circle replication, suggesting potential replication using viral machinery.

During the cloning and sequencing of def-DNAs from the RCA product, two distinctive recombinant molecules were identified. One def-DNA of ToLCNDV-ES DNA B, measuring 879 base pairs (bp), incorporated a fragment from the C1/Rep ORF of TYLCV-IL (designated as To-TYc2 in Figure 9a), corresponding to the intermolecular recombination event detected via QuasiFlow analysis (red arrows in Figure 7b) and supported by 655 reads. This finding suggests genetic exchange between ToLCNDV-ES and TYLCV-IL. Additionally, a def-DNA of ToLCNDV-ES DNA B, measuring 1306 bp, incorporating a sequence from the N. benthamiana host was detected (designated as To-Nb-c6 in Figure 9b). Further analysis of the HTS data identified approximately 22,000 reads supporting this rearrangement. The N. benthamiana sequence corresponded to a 361-nucleotide fragment located in an intergenic region on scaffold Niben101Scf14297 (positions 176,849–177,209) of the N. benthamiana genome (v1.0.1 “Scaffolds + NrContigs”, database available at https://solgenomics.net/. Sequences were deposited in GenBank under the following accession numbers: OR699190 (ToLCNDV-ES DNA-B-TYLCV-IL recombinant def-DNA) and OR699211 (ToLCNDV-ES DNA-B-N. benthamiana recombinant def-DNA). The presence of both def-DNAs was confirmed in the bulk DNA extract by PCR amplification with non-overlapping abutting primers followed by cloning and sequencing (Supplementary Figure S2). Sequence data are available under the following GenBank accession numbers: OR880356 for ToLCNDV-ES DNA-B-TYLCV-IL recombinant def-DNA and OR880357 for ToLCNDV-ES DNA-B-N. benthamiana recombinant def-DNA.

Although these recombinant def-DNA molecules were detected in N. benthamiana, their formation highlights the potential for similar rearrangements to occur in tomato, especially in mixed infections or under conditions that alter viral replication dynamics. The absence of such molecules in tomato may reflect lower viral accumulation or a more restrictive cellular environment rather than a fundamental inability of the viruses to recombine. Thus, N. benthamiana provides a useful model to explore the patterns of recombination that could be relevant to natural tomato infections.

4. Discussion

Understanding mixed virus infections in plants is essential due to their potential to alter disease dynamics, expand host range, and impact resistance stability. These interactions can be synergistic, where one virus enhances the replication or effects of another, or antagonistic, where one virus suppresses the other, leading to unpredictable outcomes in plant health and disease management [25,30,35,39,44].

In our study, TYLCV-IL and ToLCNDV-ES did not exhibit significant synergistic or antagonistic interactions in tomato plants. Both viruses behaved similarly in susceptible and resistant tomato genotypes, whether in single or mixed infections, except for a slight enhancement of ToLCNDV-ES infection by TYLCV-IL in Rondeño tomato plants, aligning with previous findings [39]. In that study, ToLCNDV-ES, which shows low adaptation to tomato, was found in natural and experimental mixed infections with TYLCV in tomato plants, with an infection ratio of up to 50%. In our experiments, susceptible tomato plants infected with ToLCNDV-ES alone remained asymptomatic. In contrast, mixed infections displayed symptoms similar to those caused by TYLCV-IL alone. This contrasts with other studies in which tomato plants co-infected with TYLCV and other begomoviruses, such as tomato leaf curl Sinaloa virus and tomato yellow mottle virus, developed more severe symptoms than those infected with either virus alone [54]. Although synergistic interactions between begomoviruses [30,34,55], or with other viruses [56,57], have been reported, no synergism was observed between TYLCV-IL and ToLCNDV-ES in our study, in contrast to recent reports [39]. Moreover, resistance-breaking effects observed in mixed infections involving other begomoviruses [25] were not found here. For instance, in chili (Capsicum annuum), synergistic interactions between multiple begomovirus genomic components were shown to break resistance in cultivars previously considered resistant due to suppression of host defense responses [25]. In contrast, the TYLCV resistance provided by the Ty-1 and Ty-2 genes in tomato plants [58] remained effective under mixed infection with TYLCV-IL and ToLCNDV-ES. Therefore, these resistance genes appear to be stable even under the combined challenge posed by both viruses.

Except for the Indian isolate of ToLCNDV in which DNA-A alone can infect tomato, both DNA-A and DNA-B were required for systemic infection of tomato plants, with essential movement functions located in DNA-B [47,59,60]. However, isolates of the ToLCNDV-ES strain that spread to the Mediterranean Basin exhibited impaired ability to infect tomatoes [37,39,61]. Our evidence supports the constraint for DNA-B accumulation in tomatoes, as observed by others [39]. According to our HTS and DNA quantification results, a lower accumulation of ToLCNDV-ES DNA-B occurred in mixed-infected tomato plants than in the more permissive host, N. benthamiana. This likely results in a movement defect that can compromise the pathogenicity of this virus in tomatoes [46]. As DNA was extracted from pooled apical leaves collected at 90 and 110 dpi from five tomato plants and ten N. benthamiana plants, no statistical comparisons between individual replicates were performed. Therefore, the observed differences in DNA-B accumulation between hosts represent general trends across pooled samples rather than statistically validated differences at the individual plant level. Notably, the interaction of ToLCNDV-ES with TYLCV-IL did not fully complement the restrictions of the former virus to infect tomato plants, even though movement functions and modulators of host defense response are present in TYLCV [62,63]. As replication of the ToLCNDV-ES isolate is not impeded [37], other virus–host interaction aspects might be impaired, constraining the infectivity of this virus in tomato plants. The virus coat protein has been shown to be crucial for ToLCNDV pathogenic adaptation in tomatoes and is essential for its infectivity [64]. However, the mechanisms associated with the restriction of ToLCNDV-ES in tomato plants and the existence of host-specific factors for effective complementation of DNA-B by DNA-A require further studies.

The presence of ToCV is a cause for concern, as it is an emerging RNA plant virus worldwide [65] that has been involved in synergistic interactions [44] and shares the B. tabaci vector and tomato host plant with the begomoviruses TYLCV-IL and ToLCNDV-ES. Therefore, these viruses can co-exist in tomato plants, as already observed for TYLCV [66] and other tomato begomoviruses [67]. Synergism between DNA and RNA viruses has been reported [68], which can alter plant–virus interactions and compromise resistance to TYLCV in tomato plants [69]. However, although co-infections of ToCV with TYLCV result in more aggressive disease severity in susceptible tomatoes [33,34] and can enhance the breakdown of resistance provided by the Ty-1 gene, we observed no effect on ToLCNDV-ES interaction with tomato plants. Therefore, the modulation of host defenses by ToCV [70] does not seem to assist ToLCNDV-ES in infecting tomato plants, which could increase the chances of interaction with TYLCV-IL during natural epidemics.

Evidence here and elsewhere shows that, even at low efficiency, ToLCNDV-ES can infect tomatoes and that mixed infection with TYLCV-IL may occur in tomato plants [37,39] and probably in other common hosts. TYLCV-IL is a monopartite begomovirus, and trans-replication of the DNA-B of ToLCNDV-ES might occur in mixed infections with unpredictable phytopathological consequences. In fact, the acquisition of heterologous DNA-Bs by monopartite begomoviruses occurs and can lead to more severe damage to infected plants [36,53,71], even involving ToLCNDV [72]. It should be noted that Rep–iteron compatibility plays a key role in determining efficient replication of geminiviruses [15] and effective reassortment among begomovirus genomic components [36]. Notable differences were observed in the iteron and Rep domains of TYLCV-IL and ToLCNDV-ES (Supplementary Figure S3). Nevertheless, reassortment has been reported despite differences, as observed for the DNA-B of the begomovirus tomato leaf curl Gujarat virus when associated with heterologous DNA-As [30]. Our results support that TYLCV-IL is unable to maintain the DNA-B of ToLCNDV-ES, suggesting that the Rep–iteron differences can constrain trans-replication, as previously observed [55]. This is relevant in terms of the interaction of TYLCV with tomato plants and crop protection because more intense damage in infected plants has been associated with heterologous trans-replication of DNA-B of ToLCNDV by monopartite begomoviruses [36,73]. The observed inability of TYLCV-IL to maintain ToLCNDV-ES DNA-B in the more permissive virus host N. benthamiana [74,75] suggests that strong heterologous recognition constraints exist. These findings indicate that TYLCV-IL is unlikely to acquire the DNA-B component of ToLCNDV-ES, which would have potentially facilitated its evolution towards a bipartite genome structure, enabling a more rapid adaptation to changing environments [53]. However, the presence of def-DNAs and recombination events within mixed infections could suggest that alternative evolutionary paths are possible. The emergence of new viral variants, facilitated by such recombination, could pose a challenge to resistance management strategies. As these recombination events may allow viruses to overcome host resistance, it is crucial to develop adaptive resistance strategies that not only monitor current virus strains but also take into account potential recombinant variants that could threaten the stability of resistance genes such as Ty-1 and Ty-2. The presence of def-DNAs, for example, could promote the development of new virulent variants capable of bypassing existing resistance mechanisms, emphasizing the need for more dynamic management approaches. Moreover, early detection of new viral strains and recombinant variants through advanced diagnostic tools is essential for managing viral resistance effectively. By continuously monitoring the emergence of such variants in field conditions, crop protection strategies could be adapted to prevent the breakdown of resistance genes in susceptible tomato varieties.

Interspecies recombination is a common phenomenon in mixed begomovirus infections, significantly influencing virus evolution, host adaptation, and emergence [31,76]. Such recombination events can occur between monopartite and bipartite begomoviruses [23,32]. In this study, HTS data were analyzed with the QuasiFlow pipeline, and recombination junctions were detected using the Virema Module. This allowed the reconstruction of a broad set of putative recombinant def-DNA molecules, including low-abundance forms. The diversity of these recombination patterns is illustrated in Figure 7. We focused our analysis on the subset of def-DNAs that were experimentally validated through RCA amplification, molecular cloning, and Sanger sequencing. These selected molecules represent well-supported recombinant forms but do not encompass the full diversity detected in silico. Although our HTS analysis did not reveal any full-length recombinant genomes in mixed infections of TYLCV-IL and ToLCNDV-ES, this observation should be interpreted with caution. The limited number of plants analyzed and the use of pooled samples may have masked low-frequency recombinant genomes present in individual plants. Moreover, only a subset of recombinant def-DNA molecules was validated and characterized by cloning and sequencing, so structurally complex or rare recombination events may have gone undetected. Further studies involving individual plant sequencing and deeper HTS analysis could help uncover such events and provide a more comprehensive view of the recombination landscape. Interestingly, we identified numerous DNA rearrangement events indicative of def-DNAs in both single and mixed infections of TYLCV-IL and ToLCNDV-ES. Def-DNAs are subgenomic molecules derived from the parent virus through deletions or rearrangements and are known to occur in begomovirus-infected plants [19,77,78]. Similar def-DNAs have been associated with TYLCV-IL infections in previous studies [79]. Interestingly, these recombination events appeared more frequently in N. benthamiana plants than in tomato plants, suggesting potential differences in virus–host interactions and their impact on genomic rearrangements across different host species and infection scenarios. This aligns with the known permissiveness of N. benthamiana to virus infections [75] and the association between high viral replication rates and increased def-DNA production [77]. Indeed, geminivirus def-DNA production appears to be influenced by the host species, with larger amounts observed in the more permissive N. benthamiana plants [19,79].

The def-DNAs cloned here had approximately half the length of native molecules and contained partial, rearranged fragments with genomic portions of TYLCV-IL and ToLCNDV-ES DNA-B, consistent with previous reports [80]. Surprisingly, we detected no events supporting def-DNAs derived from ToLCNDV-ES DNA-A. In other bipartite begomoviruses, a higher generation of DNA-B-specific def-DNAs was described, which was linked to structural differences between the components [19]. All detected def-DNAs retained genome regions required for viral DNA recognition and replication [14], suggesting they could be maintained during infection. Def-DNAs can be encapsidated, spread within the host plant, and transmitted by insect vectors, thereby persisting in epidemics [79,81]. Moreover, the presence of defective genomes can modulate plant–virus interactions by affecting the onset and persistence of viral infections and disease severity, as observed in both RNA and DNA viruses [19,77,81,82,83].

Interestingly, we detected genetic exchanges between TYLCV-IL and ToLCNDV-ES within def-DNA molecules, supporting the possibility of genetic transfer via these molecules. The biological implications of such genetic exchanges are unpredictable, but they may confer adaptive advantages to the viruses. Additionally, we identified a def-DNA suggesting genetic transfer between the plant host and ToLCNDV-ES, providing further evidence of begomoviruses’ propensity to acquire host genetic material within def-DNA [81]. This phenomenon positions def-DNAs as potential vectors for horizontal gene transfer between plants [79]. Recombinant def-DNA molecules may have significant evolutionary consequences in natural field conditions, as they could act as vehicles for genetic exchange among viruses, host plants, and insect vectors. These molecules may facilitate the acquisition of host genetic material, as seen in some begomovirus infections, thus contributing to viral adaptation to changing environments and new host species. The presence of def-DNA molecules might enhance the persistence of the virus in the field by enabling the spread of defective viral genomes that could still be transmitted by insect vectors, further complicating virus–host dynamics. Additionally, def-DNAs could modulate viral pathogenicity, affecting disease severity and persistence in plant populations. These factors may lead to the emergence of more virulent or resistant viral variants, particularly in areas where multiple begomoviruses coexist and interact.

5. Conclusions

While mixed virus infections can potentially alter plant–virus interactions and lead to resistance breakdown, our study observed no significant changes in pathogenicity resulting from the interaction between TYLCV-IL and ToLCNDV-ES. Furthermore, we did not detect the emergence of recombinant or pseudorecombinant viruses with novel pathogenic characteristics from these mixed infections. However, our findings provide evidence that genetic exchanges between TYLCV-IL and the DNA-B of ToLCNDV-ES can occur within def-DNA molecules of virus populations present in infected plants. The consequences of these genetic exchanges on virus–host interactions, virus ecology, and disease epidemics warrant further investigation. Future studies should also focus on variants better adapted to tomato plants, which appear to be emerging in the ToLCNDV-ES population in the Mediterranean region, to gain a broader understanding of the potential interactions between TYCLV-IL and ToLCNDV-ES.