Cell-type-specific response in host plants to the co-infection by sweet potato viruses

doi:10.21203/rs.3.rs-5025036/v1

Download PDF

Article

Cell-type-specific response in host plants to the co-infection by sweet potato viruses

https://doi.org/10.21203/rs.3.rs-5025036/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Host-virus interactions determine infection outcomes, with cellular heterogeneity playing a critical role in the dynamic interplay between host immune responses and viral evasion strategies. While animal and plant viruses use different mechanisms for cell entry, viral tropism is essential for pathogenesis across both kingdoms. To examine this overarching hypothesis, we studied sweet potato virus disease, the most devastating disease affecting sweet potato (Ipomoea batatas), which involves synergistic co-infection by the aphid-transmitted sweet potato feathery mottle virus and the whitefly-transmitted sweet potato leaf curl virus. By integrating single-cell RNA-Seq profiling with phenotypic assessments, we mapped virus specificity to a particular cell type within the host plant. As a result, we: 1) generated a comprehensive cellular atlas of sweet potato leaves, documenting the transcriptional response of 38,526 cells during co-infection; 2) traced virus-infected cells by examining viral genomic reads in each cell; 3) identified a tissue tropism in mesophyll, suggesting that selectively targeting cells with highly active cellular machinery is a common theme during infection by both plant and animal viruses; and finally 4) identified and characterized VIPE1, an AP2/ERF family transcription factor that contributes to viral resistance in sweet potato. These findings highlight the differential susceptibility and immune responses at both host and virus levels, underscoring the importance of understanding specific cellular and molecular features in developing targeted strategies for managing plant viruses.

Biological sciences/Plant sciences/Plant immunity

Biological sciences/Plant sciences/Plant stress responses

sweet potato

Sweet potato virus disease (SPVD)

sweet potato leaf curl virus (SPLCV)

sweet potato feathery mottle virus (SPFMV)

single-cell RNA sequencing

tropism

cellular atlas

mesophyll

transcription factor

The concept of a pandemic driven by viruses is not exclusive to humans or animals; many plant species face similar threats. Understanding the pathogenesis of viral diseases in plants is crucial for developing effective management strategies. Despite significant efforts to unravel the mechanisms of viral pathogenesis in humans and animals, our understanding of these processes in plants remains limited. Viral cellular tropism, the ability of viruses to infect specific cell types to establish successful infections, is fundamental to pathogenesis in both animals and plants ^1–3. In mammals, viral tropism is primarily determined by specific receptors on host cell surfaces⁴, as seen in the interaction between severe-acute-respiratory-syndrome-related coronavirus 2 ((SARS-CoV-2) and the ACE2 receptor⁵, which is critical for viral infectivity. In addition, other cellular factors, including enzymes and transcription factors (TFs), can influence viral tropism in animals^4,6.

In contrast, plant viruses typically enter through physical wounds or vectors such as insects (e.g., aphids and whiteflies)⁷, bypassing the need for specific cell surface receptors. Initial infection often occurs in cells at the entry site, with subsequent spread facilitated by virus-encoded movement proteins through plasmodesmata⁸. While some plant viruses show a preference for certain cell types^9–11, such as the phloem where they can spread systemically, the underlying molecular mechanisms and the role in pathogenesis of this cellular tropism are not yet fully understood. Recent studies suggest that plant viral tropism is dynamic and may change over time, as seen in the shifting localization of Tomato Golden Mosaic Virus (TGMV, Geminivirus) and Citrus tristeza virus (CTV, Closterovirus) in different plant cell types during infection^3,9,12. TGMV is predominantly associated with phloem cells at early stages of infection but becomes enriched in the nuclei of spongy and palisade cells at later stages, corresponding with the onset of severe leaf curl symptoms^12,13. Similarly, CTV has long been considered a phloem-limit virus, but a recent study discovered its presence in the xylem, where it is also required for viral pathogenicity³. This finding expands the spatial niche of CTV within the vasculature, reinforcing the notion that dynamic viral tropism is crucial for pathogenesis in plants. However, the mechanisms underlying such dynamics and their contribution to plant pathogenesis remain unclear.

Viral cellular tropism leads to the preferential accumulation of viruses in specific cell types, resulting in heterogeneous transcriptomic signatures across different cells. This cellular heterogeneity is influenced by both viral diversity and host cell factors such as metabolism and cell cycle phase, which affect viral replication¹⁴. For example, SARS-CoV-2 prefers residing in pneumocytes, while macrophages drive the inflammatory response ¹⁵. Hepatitis B virus (HBV) prefers infecting marginal zone B cells (MZBs) ¹⁶. Gene ontology (GO) enrichment analysis of the differentially expressed genes in pneumocytes or MZBs suggests heightened protein biosynthesis or DNA replication, indicating that the cellular machineries are highly active in these cells¹⁷. While animal viruses often target cells with active metabolic processes or cellular machinery, similar preferences in plant viruses remain underexplored. For example, the shoot apical meristem shows resistance to viral infection due to the expression of WUSCHEL, a meristem stem cell-specific TF¹⁸. This resistance characteristic has been leveraged in plant tissue culture to generate virus-free plants from infected progenitors. Such TF-mediated immunity is integral to plant defense strategies¹⁹. In addition, plants deploy a diverse array of antiviral defense factors, including RNA silencing suppressors (RSSs)²⁰, resistance (R) genes that produce nucleotide-binding leucine-rich repeat (NLR) proteins, and hormone-related signaling components. These factors can either work alone or interplay with each other, enhancing viral resistance. In some cases, NLRs can recognize RSSs, triggering a hypersensitive response that leads to localized programmed cell death, thereby limiting viral infection^21,22. Growing evidence has also highlighted the critical roles of TFs during viral infection¹⁹. However, the relationship between these factors and viral cellular tropism, as well as the pathogenesis of plant virus disease, remain unclear, limiting our understanding of plant viral pathogenesis.

Plant viruses cause significant crop losses globally, affecting the yield and quality of agricultural products. Sweet potato (Ipomoea batatas), with a global production exceeding 90 million tons, is one of the most important tuber crops. While generally tolerant to most abiotic stresses, sweet potatoes are highly susceptible to sweet potato virus disease (SPVD). This high susceptibility is likely due to the vegetative propagation practice, which easily transmits viruses from one generation to the next. The wide host range of viruses, along with the favorable climates in growing regions for the vectors, facilitate viral transmission. SPVD can cause up to 90% yield losses²³, making sweet potato an important research subject and an ideal model for studying viral infection. The co-infection of Sweet-Potato Leaf Curl Virus (SPLCV) and Sweet-potato Feathery Mottle Virus (SPFMV) exacerbates the SPVD symptoms²⁴. Transcriptomic analysis suggests that SPLCV and SPFMV infection impacts photosynthetic efficiency, abscisic acid biosynthesis and signaling, and carotenoid metabolism, resulting in a less dense distribution of spongy mesophyll cells^24,25. However, little is known about the viral cellular tropism and the details of pathogenesis induced by SPLCV and SPFMV.

Single-cell RNA sequencing (scRNA-seq) is a promising and powerful tool to study the virus-host interactions, allowing for the analysis of gene expression at the level of individual cells^14,26. Comparing the bulk mRNA sequencing, which only shows the average expression of a cell population, the high resolution scRNA-seq enables the identification of viral infection in different cell types, revealing the transcriptomic reprogramming at single cell level. By tracing the viral genome in individual cells, scRNA-seq can potentially explore cellular heterogeneity, elucidating interactions between the virus and the host. This approach has been successfully utilized in animals^14,16,17.

In this study, we leveraged scRNA-seq to investigate cell heterogeneity and pathogenesis under the coinfection of SPLCV and SPFMV in sweet potato. We analyzed the single-cell transcriptomes of 38,526 sweet potato leaf cells to determine viral localization and gene expression changes. Our dataset spanned all major cell types, enabling the discovery of viral cellular tropism, showing a preferential accumulation of viruses in the palisade mesophyll cell (PMC). We elucidated the spatiotemporal dynamics of plant responses and identified cell-type-specific processes in plant cells infected by SPLCV and SPFMV. We also explored the heterogenous expression of immune receptor genes, TFs, and hormone-related genes across distinct cell types. Our study highlighted the impacts of viral infection on photosynthetic efficiency and identified several PMC-specific TFs affected by viral infection. Overexpression of one such regulator, Virus Infection-related PMC ERF1 (VIPE1), enhanced viral resistance of sweet potato. This work demonstrates how previously unknown molecular events of plant-virus interactions can be uncovered using single-cell-based analyses. The high susceptibility of sweet potato to viral infection provides an excellent system for understanding plant-virus interactions at the single-cell level, ultimately aiding the development of viral-resistant crop varieties. The scRNA-seq dataset generated from this study will be a valuable resource for exploring gene expression patterns in specific cell populations and for supporting in-depth investigations of the molecular mechanisms that govern disease development. This knowledge will be instrumental in developing virus-resistant in sweet potato and other crops by precisely regulating defense-related genes in specific cells, which may minimize growth penalties in the plants.

1. Sweet potato viral disease severely affects growth, development and yield

The sweet potato variety Japanese Hong Yao (JHY) is widely cultivated and marketed due to its excellent taste and quality. However, like many other sweet potato varieties, JHY is easily infected by viruses when grown in the field. Viral infection causes severe yield reduction, with the disease accumulating in subsequent generations due to the asexual propagation practice. We collected the virus-infected JHY plants in the field and created virus-free lines (VF-JHY) by regenerating the shoot from the apical meristem-derived callus of JHY. Apical meristem is free of viruses. We then grew the virus-infected JHY and VF-JHY lines in the greenhouse and field to examine the effect of the virus disease on the plant phenotype and yield. JHY plants exhibited growth inhibition with significant dwarf vines in both field and greenhouse (Fig. 1A). SPVD resulted in approximately 50% yield loss and photosynthesis decline (Fig. 1B) (Table S1A). JHY plants displayed curved leaves, a typical symptom of SPLCV infection (Figure S1A). We then examined the viral infection using classic viral detection methods, PCR and ELISA, which confirmed that JHY plants were infected by SPFMV and SPLCV (Figure S1B and C). This result was further validated by a novel, well-established CRISPR-Cas13a-based fluorescent detection method which leverages the collateral cleavage activity of the Cas13a enzyme^3,44 (Fig. 1C). Starch staining indicated that, compared to leaves from VF-JHY, leaves from JHY were significantly affected by viral infection and exhibited a marked decrease in starch synthesis capacity (Figure S1D).

2. Construction of a sweet potato leaf cell transcriptome atlas during viral infection

To ensure the uniformity of leaf age and their suitability for protoplast extraction, we collected the first pair of expanded leaves from four-week-old plants of JHY and VF-JHY. We isolated protoplasts from the leaves, processed them in duplicates, and conducted single cell partitioning, library construction, and sequencing using the 10x Genomics workflow. After implementing quality control steps and reducing batch effects, our final dataset included 38,526 cells (Fig. 2A). Sequence alignment to the sweet potato genome covered 44,848 genes, representing 69.8% of all protein-coding genes in the sweet potato genome (Table S2A). On average, each cell exhibited 2,688 genes and 7,266 unique molecular identifiers (UMIs). UMIs are unique tags assigned to each molecule of complementary DNA (cDNA) to trace its origin and prevent PCR duplicates (Fig. 2A).

Utilizing the scRNA-seq dataset and a graph-based unsupervised clustering analysis^27,28 (Table S2B), we created a transcriptome atlas comprising 29 cell clusters (Fig. 2B). To assign specific cell types, we explored the leaf cell marker genes from prior Arabidopsis scRNA-seq data and reported sweet potato genes which had specific cellular localizations (Table S2C). This information allowed us to assign each cluster to specific cell types, except for clusters 1 and 26 (Fig. 2B and S2A). Ten of these clusters were identified as mesophyll cells, which were further grouped into palisade mesophyll cells (PMC) and spongy mesophyll cells (SMC) using markers g57924 (IQ-DOMAIN 22) and g17163 (SQUALENE MONOOXYGENESE 6), respectively, known to be specifically expressed in PMC and SMC ²⁹(Fig. 2C). Other clusters were classified as vasculature (VC, 5 clusters), epidermis (EC, 8 clusters), and guard cells (GC, 4 clusters). EC was subsequently divided into abaxial (lower EC, L_EC) and adaxial (upper EC, U_EC). L_EC included clusters 11, 12, 13, and 24, characterized by the highly expressed g30414 (ABNORMAL FLORAL ORGANS), whereas U_EC expressed g59966 (Phytosulfokines1) (Figure S2B). Based on the difference in orthologs between Arabidopsis and sweet potato, we validated the cell assignment results by performing RNA in situ hybridization (RISH) using the probes of g57924, g17163, g29052, and g14823 in the major cell types of PMC, SMC, VC, and GC (Fig. 2C). The results showed that the sweet potato orthologs were highly expressed and localized in the assigned cell types, indicating that our cell assignment was reliable. Genes that exhibited high expression levels in each cell type could be considered as novel markers, potentially useful for future research (Figure S2B and S2C).

Proportions of cells from JHY and VF-JHY were then calculated to evaluate the distribution and influence of the viral infection. The cell proportions varied drastically across clusters (Fig. 2D). For example, cluster 16 and 28 of JHY had a significantly higher number of cells compared with those of VF-JHY. These clusters were assigned as guard cells and epidermis, suggesting that the viral infection may affect epidermal cell development and division. Additionally, we noticed that cluster 1 and 26 cannot be assigned as specific cell types, the differentially expressed genes suggested that these two clusters might be the mixture of several cell types (Fig. 2D).

Protoplasting is known to affect expression of genes related to the wounding response in Arabidopsis³⁰. To mitigate any potential bias introduced by the protoplasting process during sample preparation, we assessed gene expression alterations resulting from protoplast isolation using the homologs of top 500 Arabidopsis protoplasting-induced genes (Table S2E). We found that transcripts induced by protoplasting accounted for less than 1% in each cluster, and this proportion remained consistent across clusters, suggesting that protoplasting did not lead to a redistribution of cell populations (Figure S2D). We also evaluated potential cell damage during protoplasting by examining sequencing reads originating from mitochondrial and chloroplast genes. Less than 10% of total transcripts per cluster were from these organelles (Figure S2D), suggesting that the distinct transcriptomic signatures and the clustering of cell populations were predominantly influenced by cell type and their responses to viral infection. Together, we compiled a single-cell transcriptome atlas that encompasses all major cell types in sweet potato leaf tissues.

3. Viral infection results in cell-type-specific transcriptional reprogramming

Using the transcriptome atlases derived from both JHY and VF-JHY, we identified 4,978 differentially expressed genes (DEGs) that are either upregulated or downregulated in all cell types under viral infection (Table S3A). Interestingly, 4,413 DEGs were found to be unchanged in our bulk RNA-seq analysis using the same JHY and VF-JHY leaf materials (Table S3B). Our findings highlighted the cell-type-specific differential gene expression (Fig. 3A). Additionally, 2,262 DEGs identified in the bulk RNA-seq analysis were not observed in our single-cell-based study (Figure S3A). These observations underscore the significance of performing transcriptomic analyses at both the single-cell and whole-tissue levels to capture a comprehensive view of gene expression dynamics.

Next, we explored the biological functions of the DEGs across cell types. We conducted gene ontology (GO) functional enrichment analysis on the up- and down-regulated DEGs in each cell type of JHY (Table S3C and S3D). Upregulated DEGs were predominantly related to plant specialized metabolism, phytohormone response, photosynthesis, and ribosomal function (Fig. 3B). However, DEGs exhibited distinct functional associations in different cell types. For instance, mesophyll cells, particularly the PMC, showed significant enrichment in photosynthesis and ribosome-related genes. This elevated expression of photosynthesis-related genes in PMC was somewhat unexpected. Both our bulk RNA-seq data (Table S3A and S3B) and previous studies²¹ have suggested that viral infections typically reduce photosynthesis. The high induction of ribosome-related genes is likely due to viruses hijacking host cell machinery for replication and protein production. We further examined the expression levels of the gene sets related to defense response (GO:0002217) and phytohormone signaling, including SA (GO:0009751), JA (GO:0009753), and ABA (GO:0009737) signaling in different cell types. We found that almost all genes associated with hormone signaling and defense response (GO:0002217) were suppressed in PMC but enhanced in epidermal and guard cells (Fig. 3C). A similar pattern of suppression was also detected in the phenylpropanoid metabolism (Fig. 3C), which has been suggested to be associated with antiviral activities ^31,32.

The suppression of SA signaling in PMC also occurs during TYLCV infection³³. The TYLCV C4 protein can translocate from the plasma membranes (PM) to chloroplasts, consequently inhibiting SA biosynthesis and signaling. Sequence alignments showed that the SPLCV C4 shares the N-myristoylation motif and chloroplast signal peptide with TYLCV C4, suggesting a similar translocation capability (Figure S3B). We thus expressed the SPLCV-C4-GFP fusion protein in Nicotiana benthamiana leaves. In the absence of SA (-SA), SPLCV-C4-GFP predominantly presented in PM, whereas in the presence of SA (+ SA), the SPLCV-C4-GFP clearly disappeared from PM while appearing in chloroplasts (Figure S3B). The TYLCV C4 protein antagonizes the calcium-dependent protein kinase 16 (CPK16), a key activator of SA signaling³³. We found that the sweet potato homologs of AtCPK16 were downregulated in JHY PMC (Figure S3C), suggesting that SPLCV-C4 similarly inhibits SA signaling in PMC. In cases where JA signaling is dampened, it has been reported that the suppression of JA signaling by the P1 protein represents a conserved mechanism among potyviruses and several other viruses, as this protein disrupts JA production ³⁴. Given that SPFMV and TuMV belong to the same family, we hypothesize that the P1 protein from SPFMV is responsible for the suppression of JA signaling. However, experimental validation is necessary to confirm this hypothesis.

The involvement of ABA signaling in plant–virus interactions is multifaceted^35,36. Growing evidence suggests that ABA signaling can interplay with NLR-related signaling, enhancing the viral resistance ^37–39. Therefore, we explored the expression of all NLRs identified in the sweet potato genome across various cell types. There are 889 genes encoding putative NLRs⁴⁰ (Table S3F), of which 469 were detectable in our scRNA-seq dataset (Table S3G). Although mesophyll has the highest number of infected cells, the expression of NLRs were relative higher in vasculature cells (Fig. 3D). Next, we built the NLR-related co-expression networks in the vasculature cells (Figure S3D). The vasculature-associated expression of NLRs is consistent with what has been found in the Arabidopsis under fungal infection⁴¹. However, it is not known whether viruses induce similar immune responses similar to the fungal pathogen. This weakened defense response could create a more favorable environment for viral replication and spread, highlighting the complex interplay between different signaling pathways in plant-virus interactions.

4. Spatial heterogeneity of gene expression in correlation to virus tropism

Due to the heterogeneity of viral infections, we employed the Viral-Track method to identify infected cells ¹⁶. Briefly, we traced the number of reads from the SPLCV genome, designating any cell with at least one read count from the SPLCV as infected. Attempts based on the SPFMV genome yielded few reads. Given the co-infection and synergistical effects of SPLCV and SPFMV, we focused on SPLCV. We found that unassigned clusters 1 and 26 (the Unknown) contained a high number of infected cells. This result suggested that the unique gene expression signatures in these clusters were likely due to transcriptome reprogramming in response to viral infection. We compared the DEGs enriched in these clusters and bulk RNA-seq data and found that the DEGs were highly related with GO items such as response to wounding and responsive to JA signaling (Table S4A). Using all 3,995 cells from these two clusters as input, we performed clustering analysis and identified 9 sub-clusters, which were successfully categorized into PMC, SMC, vasculature, and epidermis (Figure S4A). We then merged these cells with other previously classified cell types.

We then mapped these virus-infected cells for downstream analysis (Fig. 4A). Additionally, we reassembled the transcriptome and examined the averages transcripts of SPLCV in each cell (Fig. 4A). No viral reads were detected in VF-JHY. Since the number of viral reads may indicate the relative viral titers, we analyzed the distribution of these reads across major cell types and observed that PMC, SMC, and VC exhibited relatively higher viral loads, suggesting significant infection levels (Fig. 4B). We further validated these findings using RISH with probes for SPLCV and SPFMV on leaves from JHY plants (Fig. 4C). Consistently, PMC showed a dominant positive response to SPLCV as demonstrated by RISH. However, SPFMV exhibited more intense staining than SPLCV, implying higher SPFMV titers (Fig. 4C). This aligns with previous findings that SPLCV infection can synergistically increase SPFMV titers ²⁵.

To elucidate the heterogeneity of viral infection across leaf tissues, we analyzed the proportion of infected cells relative to the total cell counts in JHY and VF-JHY (Fig. 4D, Table S4B). Notably, the percentage of infected cells (17.2–23.8%) was consistent across all cell types, except in VC where approximately 7.9% were infected. Furthermore, the distribution of infected cells in JHY revealed a slight reduction in upper epidermis cells (17.2%), whereas PMC was more prevalent (23.8%). The distribution in lower epidermis cells (20.3%), SMC (20.2%), and GC (19.9%) were comparable. To assess the impact of viral infection on cellular architecture, we conducted detailed morphological examinations of JHY leaf tissues. The SMC appeared more dispersed throughout JHY leaves, and there were notable abnormalities in the development of vascular structures (Fig. 4D). Additionally, JHY epidermis displayed an increased stomatal density compared to VF-JHY, highlighting significant alterations in epidermal structure and function (Figs. 4D and S4E). These findings underscore the profound influence of viral infection on leaf tissue morphology and cellular distribution.

Considering the cell type-specific morphological changes tightly associated with transcriptional reprogramming triggered by the viral infection, we next analyzed transcriptomic signatures across different cell types. We segregated all cells into virus-infected and virus-uninfected groups, including those in the VF-JHY control. Subsequently, we compared the average gene expression levels between these groups to identify DEGs specific to each cell type. We identified 1,615 DEGs in PMC and 306 DEGs in SMC, with 147 DEGs shared between them (Table S4C, Figure S4C). GO functional enrichment analysis revealed that upregulated DEGs in SMC were notably enriched in processes such as isopentenyl diphosphate biosynthesis (GO:0019288), a pathway also prominent in PMC (Table S4D). In contrast, PMC-specific DEGs were significantly enriched in the cytosolic large ribosomal subunit (GO:0022625) and photosynthesis (GO:0015979). PMC was subsequently selected for detailed downstream analysis. Additionally, we observed a significant overlap of 1,239 DEGs between the L_EC and U_EC, suggesting similar transcriptional changes across the epidermis. We thus combined L_EC and U_EC for further analysis that identified 2,027 DEGs after reassessing the transcriptomic differences between infected and uninfected cells (Figure S4C).

Furthermore, a comparison of DEGs across PMC, epidermis (EC), GC, and VC revealed no DEGs common to all four cell types (Fig. 4E). This underscores the distinct transcriptional responses to viral infection in each cell type. For instance, none of the 694 upregulated DEGs identified in PMC appeared in the other three cell types, highlighting the pronounced cell-type-specific gene expression changes during viral infection. Considering the distinct cellular responses associated with different stages of viral infection, we conducted a trajectory analysis to elucidate the progressive gene expression changes occurring throughout the infection process at the single-cell level. This analysis utilized the Monocle method, assigning a pseudotime value to each cell, indicative of its progression within the continuous infection process (Table S4F). Trajectory curves were then generated for JHY and VF-JHY (Fig. 4F). Concurrently, we monitored the distribution of cell numbers along the pseudotime axis and observed a notable shift toward higher pseudotime values in cell populations impacted by the virus (Fig. 4F). By comparing trajectory curves of VF-JHY and JHY, we inferred the spatiotemporal dynamics of plant responses to viral infection. Given the significant influence of viral infection on stomatal density, we analyzed transcriptional signatures along the epidermal trajectory. Genes activated towards the end of trajectory, such as g13640 (JAZ10), were predominantly involved in defense responses (Fig. 4G, Figure S4D). Furthermore, we identified an EC-specific, virus-induced MYB-family transcription factor, g17403, homologous to Arabidopsis MYB10 (Figure S4D). AtMYB16 is known to promote the division of stomatal lineage ground cells, a process that dynamically adapts to environmental changes ⁴², suggesting the potential significance of g17403 in modulating stomatal density in JHY leaves.

5. A mesophyll cell-specific response reduces photosynthesis efficiency under viral infection

Considering the elevation of photosynthesis-related genes in PMC and SMC during viral infection, we sought to comprehend the reasons and biological consequences. We further analyzed the transcriptional signatures in PMC and SMC, identifying virus response DEGs. In PMCs, we discovered 2,380 DEGs, with 1,430 being upregulated and 950 downregulated (Table S5A). After GO enrichment analysis, we found a couple of genes in downregulated DEGs that were related to immune response (Figure S5A). Moreover, virus-infected cells revealed an elevation of photosynthesis-related transcripts in both PMC and SMC (Figs. 5A, S5A). This observation implies that the upregulation of these genes is primarily due to viral infection, rather than being solely due to that fact that mesophyll cells are the main sites of photosynthesis. Specifically, the increased expression of light-harvesting complex (LHC) genes might enhance the capacity for light energy absorption, potentially increasing the risk of ROS production and subsequent photooxidative damage within the chloroplasts. A detailed examination of PMC chloroplast ultrastructure showed that chloroplasts accumulate starch granules during the dark period. Notably, the number of starch granules in VF-JHY was approximately 1.5 times higher than in JHY. During the light period, chloroplasts in JHY were significantly smaller than those in VF-JHY, but featured thicker granum stacks (Fig. 5B), suggesting photooxidative stress which likely damaged the photosystem by generating excessive ROS. We further observed the elevated expression of oxidative stress-related genes in the PMC (Fig. 5C). The photosystem II core complex (C2) associates with two LHCII trimers (S2) to form a C2S2 complex. C2S2 can incorporate two additional LHCII trimers (M2) to form a less stable C2S2M2 supercomplex that modulates light absorption under excess light condition^43,44. Our comparative analysis between JHY and VF-JHY plants showed a higher prevalence of LHCII trimers and C2S2M2 supercomplexes in JHY (Figure S5B), suggesting an increased susceptibility to photooxidative stress in the chloroplasts of JHY PMCs. We also found that the expression of starch biosynthesis pathway genes was not directly affected during the virus infection (Figure S5C). Taken together, the unique elevation of LHCs triggers the photooxidation and photodamage within chloroplasts, reducing the starch biosynthesis efficiency and may ultimately decrease the yield of JHY during viral infection.

In addition to the suppression of SA, JA, and ABA-responsive gene in PMC (Figure S5D), our results found that 59 transcription factors (TFs) were suppressed in PMC as well, 22 of which belong to the AP2/ERF family (Fig. 5D). The sweet potato genome contains 198 AP2/ERFs, known regulators of growth, development, and viral response in plants ^45–47. The suppression of 22 AP2/ERF TFs prompted us to consider their potential roles in response to viral infections. We constructed a PMC trajectory curve that revealed the top 300 genes exhibiting pseudotime-dependent changes to trace the expression of those TFs along with the developmental trajectory (Fig. 5E). Intriguingly, while photosynthesis-associated genes were highly expressed towards the end of the trajectory, key defense-associated genes such as IbWRKY1, JAZ1, and four AP2/ERF TFs were suppressed. These findings suggest that AP2/ERF TFs might play a significant role in the viral defense response.

6. VIPE1-overexpression enhances the virus resistance of sweet potato plants

To substantiate the roles of AP2/ERF TFs in virus responses, we selected g30808, one of the most significantly affected AP2/EFR TFs. As g30808 was suppressed by viral infection in PMC, we named it Virus Infection-related PMC-ERF 1 (VIPE1). We generated VIPE1-overexpressing (OX-VIPE1) sweet potato lines. Under both field and greenhouse conditions, OX-VIPE1 exhibited a significant reduction of biomass and tuber yield compared to the wild type (wt) (Fig. 6A, and Figure S6A). Given that the upregulation of LHCs coincided with the suppression of VIPE1 in PMC of JHY, we thus hypothesized that VIPE1 was involved in regulating the expression of LHCs. We measured the expression of LHC genes in the OX-VIPE1 line and found that the overexpression of VIPE1 dramatically reduced the transcript level of LHCs (Fig. 6B). Next, we examined the viral resistance of the OX-VIPE1 line. Compared to wild type, OX-VIPE1 plants were more resistant to infection of SPLCV and SPFMV (Fig. 6C). This resistance was maintained even grown under the field conditions. Those results suggested that VIPE1 is involved in response to viral infection by SPLCV and SPFMV. However, we noticed that VIPE1 overexpression also reduced the growth, photosynthesis efficiency, and starch biosynthesis capability, while altered chloroplast ultrastructure (Figure S6B and S6C), suggesting that VIPE1 is also a strong repressor of sweet potato plant development. These results demonstrate a significant contribution of VIPE1 to plant virus defense, presumably through limiting the expression of LHCs in PMC.

In this study, we generated a single-cell transcriptome atlas in the sweet potato leaves with ∼39,000 cells under the infection conditions with SPLCV and SPFMV, two emerging and widespread viruses. This atlas encompasses all major cell types in leaves, allowing for the identification of cellular responses specific to each cell type during viral infection. We further investigated viral cellular tropism and underlined the cell heterogeneity by directly tracing the viral reads in individual cells, revealing significant accumulation of viruses and the induction of photooxidative stresses in PMC. Furthermore, we found the suppression of immune response associated with reduced hormone signaling, as well as NLRs- and TF-related defense response in PMC. We identified and characterized the TF VIPE1, showing that overexpression of VIPE1 enhances virus resistance. Our approaches, using virus-infected plants directly sourced from the field and comparing them with the virus-free plants with the same genetic background, capitalized on the unique materials and offered a comparative advantage.

In viral infection, cells containing the viral genome exhibit distinct transcriptomic signatures compared to uninfected cells. We found that viral infection in sweet potato, which is susceptible to multiple viruses, specifically suppresses defense signaling pathways in mesophyll cells (Fig. 3). These suppressions affect NLRs, hormone, and TF-related signaling pathways. Although little is known about the roles of NLRs and TFs in viral resistance, previous studies have shown that viral proteins can directly inhibit the biosynthesis of SA and JA. For instance, geminiviruses like TYLCV can relocate their C4 proteins from the plasma membrane to the chloroplast to suppress SA biosynthesis and signaling³³. Similarly, the P1 protein of Turnip mosaic virus (TuMV; Potyvirus) facilitates the degradation of the chloroplast signal recognition particle 54 (cpSRP54) through the 26S proteasome pathway, disrupting the localization of ALLENE OXIDE CYCLASE (AOC), a key enzyme in JA biosynthesis, consequently suppressing JA production ³⁴. Given that SPLCV and SPFMV belong to the same families as TYLCV and TuMV, it is plausible that they may similarly suppress SA and JA synthesis, respectively.

We also observed that the virus preferentially resides in cells with active cell machinery and primary metabolic processes, which aligns with findings in the animal and human cells. In mesophyll cells, the elevation of photooxidative stress within chloroplasts may directly damage the photosystem proteins, signaling the host cell machinery to synthesize new proteins to replace the damaged ones through chloroplast-to-nucleus retrograde signaling. Consistently, we found an upregulation of ribosome-related genes in mesophyll cells compared to other cell types (Fig. 3). Mesophyll cells are active in primary metabolism. Glucose and fructose serve as energy sources for viral replication machinery and nourishment for insect vectors, while galactose may be essential for glycoprotein synthesis which is required for capsid formation ⁴⁸. Viral infection often results in reduced chlorophyll content, affecting CO₂ fixation rates, as well as alterations in total soluble sugars, proteins, and starch ^49–51. These nutritional changes induced by geminiviruses favor increased abundance, fecundity, and transmission efficiency of vector whiteflies, thereby facilitating the spread of viruses. Although further empirical data are essential, our findings suggest that interference with defense signals in mesophyll cells promotes viral proliferation. Moreover, the activation of primary metabolic processes enhances sugar production to attract insect vectors, possibly accelerating virus transmission.

A major challenge in studying plant-virus interactions is the uneven distribution of viruses, making it difficult to identify infected cells and dissect the transcriptional reprogramming specific to each cell type. In this study, we traced the number of reads from the SPLCV genome using scRNA-seq data and combined it with the cell-specific morphological changes observed under high-resolution microscopy (Fig. 4). This enabled us to identify genes that are associated with the significant morphological changes across cell types. For example, the increased stomata density under the viral disease, a phenomenon that has been sparsely studied regarding the influences of viral infection on stomata development⁵². We successfully identified an epidermis-specific virus-induced TF, IbMYB16 (Figure S4D), a homolog of Arabidopsis MYB16 and a key regulator of stomata division and development⁵³. Although the impact of viral infection on the stomata density varies across different viruses ^52,54, IbMYB16 may contribute the increased stomata density in our study. However, more studies are needed regarding the effects of viral infection on stomata development.

Our study underscores the importance of TF-associated defense signaling pathways during viral infections. Notably, one of the AP2/ERF TFs, VIPE1, regulates the expression of LHCs and significantly enhances resistance against SPLCV and SPFMV (Fig. 6). The virus may specifically alter the expression of photosynthesis-related genes by interfering with the functions of various TFs, including VIPE1. Subsequently, through retrograde signaling from chloroplasts, this interference further accelerates the activation of the cellular replication machinery, thereby speeding up the virus replication process. While VIPE1 represents a promising target to bolster viral resistance in breeding programs, our results suggest that VIPE1 also inhibits plant growth. This compromise is likely a consequence of growth-defense tradeoff. Another possibility is that VIPE1 is involved in multiple processes and regulated by different factors across different cell types. Therefore, while the 35S promoter-driven whole-plant overexpression of VIPE1 enhances the viral resistance in PMC, it inhibits cell growth in other cell types. A more sophisticated approach, using PMC-specific promoters instead of the 35S promoter to drive VIPE1 expression, may allow for viral resistance with minor growth inhibition.

Our findings strongly suggest that infecting viruses selectively target cells with highly active cellular machinery, which is a critical factor for successful viral invasion and replication. This preference appears to be a fundamental property of viruses, whether in plant or animals, and may represent a universal mechanism across different viruses (Fig. 7). Additionally, SPLCV and SPFMV serve as representative examples of the geminivirus and potyvirus families, respectively. The methodology used in this study provides a framework for investigating plant virus disease in other plant species. Our research addresses a critical challenge in enhancing disease resistance in crops, contributing to the development of more sustainable and resilient agricultural practices.

Plant materials and growth conditions

SPDV infected sweet potato ((cultivar: Japanese Hong Yao (JHY)) plants were harvested from the field and propagated by tissue culture using nodal explants. Virus-free plants (VF-JHY) were regenerated from callus derived from the apical meristem of JHY. The tissue culture plants were maintained in a growth chamber at 25°C, 50% relative humidity (RH), and 16 h light (cool white, fluorescent tubes ~ 50 µmol/m²/s) /8 h dark cycle. JHY and VF-JHY stem cuttings were planted in Wushe Farm (Crop Cultivation and Breeding Base of the Institute of Plant Physiology and Ecology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences). To sample the protoplasting materials, nodal explants from JHY and VF-JHY grown in the greenhouse were collected and generated in the ½ MS agar medium. Four-week-old leaves from the explants were used for measuring physiological, biochemical indexes, chloroplast ultrastructure, and isolating protoplasts.

Detection of Viruses and RNA in situ hybridization

SPLCV (Catalog number: ZK-C7352), SPFMV (Catalog number: ZK-C7201), and SPCSV (Catalog number: ZK-C7351, Shenzhen Zikai) ELISA kits were used to detect viruses following the protocol provided by the manufacturer. The expression patterns of SPLCV and SPFMV in specific cell types was determined using RNA in situ hybridization (RISH). In brief, the fragment from SPLCV or SPFMV were cloned into the pGEM-T easy vector (Promega). RNA probes were synthesized using these vectors and labeled using a DIG RNA Labeling Kit (Roche). The DIG labeled probes were hybridized to paraffin-embedded leaf sections. After treatment with Proteinase K and anti-digoxigenin-AP Fab fragments, color development was conducted using NBT/BCIP (Sigma). The sections were scanned under the microscope lens of the slide scanner (Model: Pannoramic MIDI; Manufacturer: 3DHISTECH (Hungary)) and images were collected using CaseViewer 2.4 software.

Agrobacterium infiltration, SPLCV C4 localization, and detection of chloroplast ultrastructure

The coding sequence SPLCV C4 fused to the green fluorescent protein (GFP) was cloned in pCAMBIA 1302 vector and mobilized to Agrobacterium tumefaciens strain GV3101. A. tumefaciens strain GV3101 carrying the SPLCV C4-GFP was infiltrated into Nicotiana benthamiana leaves. Two days after infiltration, samples were dipped in a solution with a final concentration of 5 mM SA or equal amount of ethanol as control. GFP Fluorescence was detected by a Leica TCS SP8 confocal microscope for GFP (Ex: 488 nm, Em: 500–550 nm) and chlorophyll autofluorescence (Ex: 680 nm, Em: 700–750 nm).

Protoplasts isolation for scRNA-seq

Protoplasts were isolated from the first pair of expanded leaves of JHY and VF-JHY. After the removal of mid veins, leaf laminas were cut into 0.5-1.0 mm strips with a blade. Approximately 0.3 g of leaves were placed in 20 mL of enzyme solution containing 2% (w/v) cellulose Onozuka R-10, 0.3% (w/v) macerozyme R-10 and 0.1% (v/v) pectinase. The enzymes were dissolved in a solution containing mannitol (0.4M), and 2-(N-morpholino) ethanesulfonic acid (MES; 20mM; pH 5.7–5.8). The digestion was carried out in the dark at 25°C with continuous shaking (45 rpm) for 2 hours. After centrifuge, protoplast viability was checked by trypan blue staining. The samples with over 90% viability were adjusted to approximately 2000 cells per microliter using a hemocytometer. For each biological replicate, at least four technical replicates were prepared, and the top two were selected for further single-cell partitioning.

Single-cell partition and RNA-library construction and raw reads alignment

The protoplast suspensions were diluted to 1500 cells per microliter and immediately processed using a Chromium Single Cell Instrument (10x Genomics, Pleasanton, CA) to create single-cell GEMs (Gel Bead-In Emulsions). Subsequently, scRNA-seq libraries were constructed using the Chromium Single Cell 3' Gel Bead and Library Kit V3.1 (10x Genomics, Pleasanton, CA), following the manufacturer’s instructions. The cDNA amplification was carried out over eleven PCR cycles, followed by another eleven cycles for final library construction. The prepared libraries were sequenced on an Illumina NovaSeq producing paired-end 150 bp reads, with quality assessment conducted using an Agilent 2100 Bioanalyzer.

Reference files for the sweet potato genome were downloaded from the Ipomoea Genome Hub (https://sweetpotao.com) and sweet potato chloroplast genome (NC_026703.1). The “cellranger mkref” function (part of Cellranger (V7.2), 10x Genomics, Pleasanton, CA) was used to build references. Raw reads were aligned against the genome using Cellranger with default settings. The percentage of aligned reads ranged from 80–90% across the samples.

Quality control, doublet detection, cell-cycle regression, and batch effects removal of scRNA-seq data

The downstream analysis involved utilizing modified scripts from Seurat (v5.0.3). Initially, the `Read10x` function was employed to import raw matrix files, followed by the `CreateSeuratObject` function to construct individual Seurat datasets post-Cellranger alignment. Cells with fewer than 200 genes were removed to eliminate low-quality data, and genes that appeared in fewer than five cells were excluded. The data were normalized using the `SCTransform` function and subsequently analyzed with the DoubletFinder workflow to identify doublets in the single cell partitioning experiment. Specifically, `DoubletFinder_v3` was utilized to detect homotypic doublets using parameters such as nExp = round(0.05 * nrow(x)) for the expected number of real doublets, pN = 0.25 for artificial doublets, and pK = 0.09 for neighborhood size. This process calculated the proportion of artificial neighbors per cell to ascertain doublet predictions. Cells labeled as "Singlets" in each library were retained for further analysis. The proportion of read counts aligned to genes induced by chloroplasts and mitochondria was assessed. Cells with less than 5% of transcripts aligning to mitochondrial and less than 10% chloroplast genes were selected for additional analysis. Further quality control included discarding cells expressing fewer than 500 or more than 20,000 genes. Libraries were merged using Seurat, and to address cell cycle heterogeneity's impact on clustering, cell cycle-related genes (G1-S and G2-M phases) identified from (Table S1C) were incorporated using the `CellCycleScoring` function. For further normalization, `SCTransform` was reapplied, regressing out cell cycle genes. Finally, the integrated Seurat dataset was refined using the `RunHarmony` function with the `SCT` assay to correct batch effects across replicates and samples.

Clustering and cell type annotation and gene expression analysis in scRNA-seq

Clustering analysis was executed using `RunPCA` with npc = 30, followed by `RunUMAP` and `RunTSNE`. Cell clusters were then identified using the `FindNeighbors` function with parameters k.param = 10 and dims = 1:30, and the `FindClusters` function at a resolution of 0.8. Gene enrichment within each cluster was assessed using the `FindAllMarkers` function, with parameters logfc.threshold = 0.25, min.pct = 0.1, and min.diff.pct = 0.1. The Wilcoxon Rank Sum test was utilized to identify differentially expressed genes between each cluster and the remaining cells. These genes were cross-referenced with supplemented with Arabidopsis genes from prior studies using Arabidopsis leaf tissue⁴¹. These genes facilitated the annotation of cell types within each cluster. The `FeaturePlot` function was employed to visualize and profile marker gene expression.

For more detailed analysis, sub-clustering of clusters 1 and 26 was conducted by extracting cells and applying `RunUMAP` at a resolution of 0.8 to achieve finer cluster distinctions. The identification of genes enriched in these sub-clusters and their cell type annotations followed the previously described methods. To analyze the expression of potential and known NLRs in sweet potato genome, only those with 500 or more normalized UMI counts across the atlas. The UMI count matrix from Seurat was extracted to determine cell type-specific expression. Expression levels for each gene within each cell type were computed by aggregating expression values for predefined cell groups using the `aggregate_gene_expression` function in Monocle3, with normalization performed by dividing by cell size factors. Expression patterns of photosynthesis-, SA-, ABA and JA-related genes were collected based on the Gene Ontology annotation and investigated by pooling these genes and using them as inputs for the `AddModuleScore` function in Seurat. This approach assessed the relative contribution of these genes to the transcriptome of each cell type.

Viral-Track detection of infected cells

For the viral detection, we use the Viral-track method⁵⁵, the raw reads were mapping to viral reference genomes (SPLCV: FJ176701.1 and SPFMV: NC_001841.1) separately. The cell that contains at least one read count of SPLCV was marked as infected cells.

Trajectory inference and pseudotime analysis

To build a trajectory, cells of the same type were aggregated and processed through a Monocle pipeline. Initially, the UMI count matrix from four major cell types was converted into datasets using the `newCellDataSet` function. This data was then refined to pinpoint genes that were differentially expressed while mitigating batch effects from replicates. From this analysis, the top 2,000 genes showing significant changes (based on q-values) were selected for deeper investigation. To reduce the influence of developmental changes in cell types on the infection-related trajectory, uninfected cells were isolated. Pseudotime for uninfected cells was computed to identify the top 100 significantly altered genes, critical for understanding cell type development and determinant in the infection response. Subsequently, excluding genes identified in the cell type pseudotime analysis, a selected set of genes was used to construct the trajectory that illustrates the dynamics of cellular responses to infection. The `setOrderingFilter` function and the `DDRTree` method were employed to arrange the cells based on the expression patterns of these genes. Pseudotime values for the four cell types—epidermis (upper and lower), guard cells, mesophyll (spongy and palisade), and vasculature—were established.

To identify genes enriched at infection cells, the `FindAllMarkers` function in Seurat was applied with criteria of log2|FC| > 0.25, pct.1 ≥ 0.2, pct.2 < 0.1, and min.diff.pct > 0.2, using the Wilcoxon Rank Sum test to identify differentially expressed genes. All enrichment analyses were conducted using the clusterProfiler package, with thresholds set at a q-value < 0.2 for KEGG metabolic pathways and q-value < 0.05 for GO terms.

Bulk RNA-seq data analysis

To investigate transcriptomic alterations in sweet potato during virus infection, we performed bulk RNA-seq using the same materials of JHY and VF-JHY used for scRNA-seq. Samples were processed for total RNA isolation using the RNeasy Plant Kit (Qiagen). The quality of total RNA was assessed using a Bioanalyzer before sequencing. Post-sequencing, the raw reads underwent quality checks and adaptor trimming using Trimmomatic (V0.32). These trimmed reads were then mapped to the sweet potato genome with STAR (2.7.11b), and the resultant data were consolidated into a count matrix via tximport (3.19). For the analysis of differentially expressed genes, sample size factors and normalization were conducted using DESeq2 (1.44.0). To pinpoints sweet potato genes specifically activated during virus infection, we focused on genes exhibiting a log2|FC| > 2 and p-adj < 0.05 in JHY samples.

In parallel, to see the side effect of protoplasting, we selected top 500 differentially expressed genes after protoplasting in a prior Arabidopsis study⁴¹ and blasted their sweet potato orthologs. Expression percentage of those orthologs were examined by “PercentageFeatureSet” function in Seurat package.

Quantification and statistical analysis

Statistical analyses were performed in R (4.3.3) and GraphPad Prism (v10). One-way ANOVA Duncan test was used for multiple groups of observations. Student’s t-test was used to determine the significance for two groups of observations. Details of the statistical analysis can be found in the figure legends.

Sweet potato transformation and analysis of transgenic plants

Transgenic sweet potato lines overexpressing the ERF transcription factor VIPE1 were kindly provided by Dr. Xiaofeng Bian from the Provincial Key Laboratory of Agrobiology, Institute of Food Crops, Jiangsu Academy of Agricultural Sciences. The photosynthesis efficiency of fully expanded fourth leaf of the wild-type (control), JHY, VF-JHY, and OX-VIPE1 plants was measured by assessing chlorophyll fluorescence to determine the photochemical yield (Fv/Fm). Chlorophyll content was measured in the same leaves using a portable chlorophyll meter (SPAD-502, Konica Minolta, Japan). Additionally, expression of genes related to light harvesting complex and SPLCV/SPFMV was measured in control, JHY, VF-JHY, and transgenic plants using RT-qPCR. Sweet potato ubiquitin extension protein (UBI) and SPLCV-AC2, SPFMV genes was used as an internal control in RT-qPCR ⁵⁶.

Funding

The Science and Technology Commission of Shanghai Municipality (22JC1401300),The Shanghai Municipal Afforestation & City Appearance and Environmental Sanitation Administration (G242407, G222411, G222413, and G232405) and small and medium-sized enterprise innovation ability enhancement project of Shandong province Science and Technology (2021TSGC1220) grants to H.W. Crop Research Institute, Guangdong Academy of Agricultural Sciences/Guangdong Provincial Key Laboratory of Crop Genetic Improvement Open Research Fund (NYGC202202). National Natural Science Foundation of China (32472220).

Author contributions

R.Q.L. is involved in the conception of this study, conducted most of the data analysis and wrote the manuscript. W.J.F. and Y.Q.W. performed the experiments. P.S. revised the manuscript and discussed the results. Y.Y. provided VIPE1 overexpression lines. X.G.Z., Y.L.L., and Y.Q.L discussed the results. J.Y. revised the manuscript. H.X.W. and L.Y., designed the study, supervised the project, and revised the manuscript.

Acknowledgements

We would like to thank the Core Facility Centre of CAS Center for Excellence in Molecular Plant Sciences (CFC-CEMPS) for help with preparing TEM samples and taking TEM images. We would like to thank Dr. Bian Xiaofeng from Jiangsu Academy of Agricultural Sciences Department of Crop Breeding for the VIPE1 overexpressing lines. No conflict of interest is declared.

Tyler KL (1999) In: Granoff A, Webster RG (eds) Encyclopedia of Virology (Second Edition). Elsevier, pp 1175–1184
Zaitlin M, Palukaitis P (2000) Advances in understanding plant viruses and virus diseases. Annu Rev Phytopathol 38:117–143
Sun YD, Folimonova SY (2022) Location matters: from changing a presumption about the Citrus tristeza virus tissue tropism to understanding the stem pitting disease. New Phytol 233:631–638
McFadden G, Mohamed MR, Rahman MM, Bartee E (2009) Cytokine determinants of viral tropism. Nat Rev Immunol 9:645–655. https://doi.org/10.1038/nri2623
Hoffmann M et al (2020) SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor. cell 181, 271–280. e278
Kafai NM, Diamond MS, Fox JM (2022) Distinct cellular tropism and immune responses to alphavirus infection. Annu Rev Immunol 40:615–649
Kappagantu M, Collum TD, Dardick C, Culver JN (2020) Viral hacks of the plant vasculature: the role of phloem alterations in systemic virus infection. Annual Rev Virol 7:351–370
Tilsner J et al (2013) Replication and trafficking of a plant virus are coupled at the entrances of plasmodesmata. J Cell Biol 201:981–995
Bendix C, Lewis JD (2018) The enemy within: phloem-limited pathogens. Mol Plant Pathol 19:238–254. https://doi.org/10.1111/mpp.12526
Price W (1966) Flexuous rods in phloem cells of lime plants infected with citrus tristeza virus. Virology 29:285–294
Karyeija RF, Kreuze JF, Gibson RW, Valkonen JPT (2000) Synergistic Interactions of a Potyvirus and a Phloem-Limited Crinivirus in Sweet Potato Plants. Virology 269:26–36. https://doi.org/https://doi.org/10.1006/viro.1999.0169
Morra MR, Petty IT (2000) Tissue specificity of geminivirus infection is genetically determined. Plant Cell 12:2259–2270. https://doi.org/10.1105/tpc.12.11.2259
Qin Y, Petty IT (2001) Genetic analysis of bipartite geminivirus tissue tropism. Virology 291:311–323
Cristinelli S, Ciuffi A (2018) The use of single-cell RNA-Seq to understand virus-host interactions. Curr Opin Virol 29:39–50. https://doi.org/10.1016/j.coviro.2018.03.001
Speranza E et al (2021) Single-cell RNA sequencing reveals SARS-CoV-2 infection dynamics in lungs of African green monkeys. Sci Transl Med 13:eabe8146. https://doi.org/10.1126/scitranslmed.abe8146
Bost P et al (2020) Host-Viral Infection Maps Reveal Signatures of Severe COVID-19 Patients. Cell 181:1475–1488e1412. https://doi.org/10.1016/j.cell.2020.05.006
Wyler E et al (2019) Single-cell RNA-sequencing of herpes simplex virus 1-infected cells connects NRF2 activation to an antiviral program. Nat Commun 10:4878. https://doi.org/10.1038/s41467-019-12894-z
Wu H et al (2020) WUSCHEL triggers innate antiviral immunity in plant stem cells. Science 370:227–231. https://doi.org/doi:10.1126/science.abb7360
Viswanath KK et al (2023) The role of plant transcription factors in the fight against plant viruses. Int J Mol Sci 24:8433
El-Sappah AH et al (2021) Comprehensive mechanism of gene silencing and its role in plant growth and development. Front Plant Sci 12:705249
Bednarek R, David M, Fuentes S, Kreuze J, Fei Z (2021) Transcriptome analysis provides insights into the responses of sweet potato to sweet potato virus disease (SPVD). Virus Res 295:198293
Zhang K et al (2020) The spread and transmission of sweet potato virus disease (SPVD) and its effect on the gene expression profile in sweet potato. Plants 9:492
Valverde RA, Clark CA, Valkonen JP (2007) Viruses and virus disease complexes of sweetpotato. Plant Viruses 1:116–126
Tibiri EB et al (2020) Characterization of virus species associated with sweetpotato virus diseases in Burkina Faso. Plant Pathol 69:1003–1017. https://doi.org/10.1111/ppa.13190
Hou F et al (2020) Sweetpotato leaf curl virus decreased storage root yield and quality of edible sweetpotato in China. Agron J 112:3948–3962. https://doi.org/https://doi.org/10.1002/agj2.20025
Zhu J, Moreno-Pérez A, Coaker G (2023) Understanding plant pathogen interactions using spatial and single-cell technologies. Commun Biology 6:814
van der Maaten L, Hinton G (2008) Visualizing Data using t-SNE. J Mach Learn Res 9:2579–2605
McInnes L, Healy J, Melville J, Umap (2018) Uniform manifold approximation and projection for dimension reduction. arXiv preprint arXiv:1802.03426
Procko C et al (2022) Leaf cell-specific and single-cell transcriptional profiling reveals a role for the palisade layer in UV light protection. Plant Cell 34:3261–3279. https://doi.org/10.1093/plcell/koac167
Meyer Y et al (1993) Gene expression in mesophyll protoplasts. Morphogenesis plants: Mol approaches, 221–236
Abdelkhalek A et al (2022) Defense Responses and Metabolic Changes Involving Phenylpropanoid Pathway and PR Genes in Squash (Cucurbita pepo L.) following Cucumber mosaic virus Infection. Plants (Basel) 11. https://doi.org/10.3390/plants11151908
Ahmad N, Xu Y, Zang F, Li D, Liu Z (2024) The evolutionary trajectories of specialized metabolites towards antiviral defense system in plants. Mol Hortic 4:2
Medina-Puche L et al (2020) A defense pathway linking plasma membrane and chloroplasts and co-opted by pathogens. Cell 182:1109–1124
Ji M et al (2021) Turnip mosaic virus P1 suppresses JA biosynthesis by degrading cpSRP54 that delivers AOCs onto the thylakoid membrane to facilitate viral infection. PLoS Pathog 17:e1010108. https://doi.org/10.1371/journal.ppat.1010108
Alazem M, Lin N-S (2020) Interplay between ABA signaling and RNA silencing in plant viral resistance. Curr Opin Virol 42:1–7
Zhao S, Li Y (2021) Current understanding of the interplays between host hormones and plant viral infections. PLoS Pathog 17:e1009242
Huang S et al (2024) A plant NLR receptor employs ABA central regulator PP2C-SnRK2 to activate antiviral immunity. Nat Commun 15:3205
Qi S et al (2022) A new NLR gene for resistance to tomato spotted wilt virus in tomato (Solanum lycopersicum). Theor Appl Genet 135:1493–1509
Alazem M, Widyasari K, Kim K-H (2019) An avirulent strain of soybean mosaic virus reverses the defensive effect of abscisic acid in a susceptible soybean cultivar. Viruses 11:879
Parada Rojas CH, Quesada-Ocampo LM (2021) Uncovering the NLR Family of Disease Resistance Genes in Cultivated Sweetpotato and Wild Relatives. Postharvest Pathology: Next Generation Solutions Reducing Losses Enhancing Saf, 41–61
Tang B, Feng L, Hulin MT, Ding P, Ma W Cell-type-specific responses to fungal infection in plants revealed by single-cell transcriptomics. Cell Host Microbe https://doi.org/10.1016/j.chom.2023.08.019
Lee LR, Bergmann DC (2019) The plant stomatal lineage at a glance. J Cell Sci 132:jcs228551
Kouřil R, Wientjes E, Bultema JB, Croce R, Boekema EJ (2013) High-light vs. low-light: effect of light acclimation on photosystem II composition and organization in Arabidopsis thaliana. Biochim et Biophys Acta (BBA)-Bioenergetics 1827:411–419
Nath K et al (2013) Mobilization of photosystem II-light harvesting complex II supercomplexes during high light illumination and state transitions. Rapid Communication Photoscience 2:18–23
Huang Y et al (2016) AP2/ERF transcription factors involved in response to tomato yellow leaf curly virus in tomato. Plant Genome 9, plantgenome2015.2009.0082
Ma N et al (2024) Plant disease resistance outputs regulated by AP2/ERF transcription factor family. Stress Biology 4:2
Paul P et al (2020) Mutually regulated AP2/ERF gene clusters modulate biosynthesis of specialized metabolites in plants. Plant Physiol 182:840–856
Mahmoudabadi G, Milo R, Phillips R (2017) Energetic cost of building a virus. Proc Natl Acad Sci U S A 114:E4324–e4333. https://doi.org/10.1073/pnas.1701670114
Bhattacharyya D et al (2015) A geminivirus betasatellite damages the structural and functional integrity of chloroplasts leading to symptom formation and inhibition of photosynthesis. J Exp Bot 66:5881–5895. https://doi.org/10.1093/jxb/erv299
Kim JY et al (2021) Distinct identities of leaf phloem cells revealed by single cell transcriptomics. Plant Cell 33:511–530. https://doi.org/10.1093/plcell/koaa060
Kushwaha NK, Mansi, Sahu PP, Prasad M, Chakrabroty S (2019) Chilli leaf curl virus infection downregulates the expression of the genes encoding chloroplast proteins and stress-related proteins. Physiol Mol Biol Plants 25:1185–1196. https://doi.org/10.1007/s12298-019-00693-1
Murray RR, Emblow MSM, Hetherington AM, Foster GD (2016) Plant virus infections control stomatal development. Sci Rep 6:34507. https://doi.org/10.1038/srep34507
Yang S-L, Tran N, Tsai M-Y, Ho C-M (2022) K. Misregulation of MYB16 expression causes stomatal cluster formation by disrupting polarity during asymmetric cell divisions. Plant Cell 34:455–476
Robert-Seilaniantz A, Grant M, Jones JD (2011) Hormone crosstalk in plant disease and defense: more than just jasmonate-salicylate antagonism. Annu Rev Phytopathol 49:317–343. https://doi.org/10.1146/annurev-phyto-073009-114447
Kutnjak D et al (2021) A Primer on the Analysis of High-Throughput Sequencing Data for Detection of Plant Viruses. Microorganisms 9:841. https://doi.org/10.3390/microorganisms9040841
Park S-C et al (2012) Stable Internal Reference Genes for the Normalization of Real-Time PCR in Different Sweetpotato Cultivars Subjected to Abiotic Stress Conditions. PLoS ONE 7:e51502. https://doi.org/10.1371/journal.pone.0051502

There is NO Competing Interest.

supplementarydatasets.zip
Table_S1, Table_S2, Table_S3, Table_S4, Table_S5, Table_S6
SuppFigures0829.pptx

Download PDF

Version 1

posted

You are reading this latest preprint version

Cell-type-specific response in host plants to the co-infection by sweet potato viruses

Status:

Version 1

Abstract

Figures

Introduction

Results

1. Sweet potato viral disease severely affects growth, development and yield

2. Construction of a sweet potato leaf cell transcriptome atlas during viral infection

3. Viral infection results in cell-type-specific transcriptional reprogramming

4. Spatial heterogeneity of gene expression in correlation to virus tropism

5. A mesophyll cell-specific response reduces photosynthesis efficiency under viral infection

Discussion

Methods

Plant materials and growth conditions

Protoplasts isolation for scRNA-seq

Single-cell partition and RNA-library construction and raw reads alignment

Quality control, doublet detection, cell-cycle regression, and batch effects removal of scRNA-seq data

Clustering and cell type annotation and gene expression analysis in scRNA-seq

Viral-Track detection of infected cells

Trajectory inference and pseudotime analysis

Bulk RNA-seq data analysis

Quantification and statistical analysis

Sweet potato transformation and analysis of transgenic plants

Declarations

Funding

Author contributions

Acknowledgements

References

Additional Declarations

Supplementary Files

Status:

Version 1