Rescue of the first alphanucleorhabdovirus entirely from cloned complementary DNA: An efficient vector for systemic expression of foreign genes in maize and insect vectors

Abstract

Recent reverse genetics technologies have enabled genetic manipulation of plant negative‐strand RNA virus (NSR) genomes. Here, we report construction of an infectious clone for the maize‐infecting Alphanucleorhabdovirus maydis, the first efficient NSR vector for maize. The full‐length infectious clone was established using agrobacterium‐mediated delivery of full‐length maize mosaic virus (MMV) antigenomic RNA and the viral core proteins (nucleoprotein N, phosphoprotein P, and RNA‐directed RNA polymerase L) required for viral transcription and replication into Nicotiana benthamiana. Insertion of intron 2 ST‐LS1 into the viral L gene increased stability of the infectious clone in Escherichia coli and Agrobacterium tumefaciens. To monitor virus infection in vivo, a green fluorescent protein (GFP) gene was inserted in between the N and P gene junctions to generate recombinant MMV‐GFP. Complementary DNA (cDNA) clones of MMV‐wild type (WT) and MMV‐GFP replicated in single cells of agroinfiltrated N. benthamiana. Uniform systemic infection and high GFP expression were observed in maize inoculated with extracts of the infiltrated N. benthamiana leaves. Insect vectors supported virus infection when inoculated via feeding on infected maize or microinjection. Both MMV‐WT and MMV‐GFP were efficiently transmitted to maize by planthopper vectors. The GFP reporter gene was stable in the virus genome and expression remained high over three cycles of transmission in plants and insects. The MMV infectious clone will be a versatile tool for expression of proteins of interest in maize and cross‐kingdom studies of virus replication in plant and insect hosts.

The maize mosaic virus infectious clone enables stable and robust expression of heterologous genes in maize and planthoppers, and is a versatile tool for cross‐kingdom studies of virus replication in plant and insect hosts.

1. INTRODUCTION

Viruses can be used as multifaceted recombinant vectors for virus‐induced gene silencing (VIGS), heterologous protein expression, and genome editing in both dicotyledonous and monocotyledonous plants. However, efficient, stable, and systemic expression of heterologous proteins at high levels by many virus‐based vectors in their hosts is challenging due to genetic instability (reviewed in Willemsen & Zwart, 2019). The Rhabdoviridae is a diverse family of negative‐strand RNA viruses (NSRs) that infect vertebrates, invertebrates, and plants. Plant nonsegmented rhabdoviruses are currently classified into four genera: Cytorhabdovirus, which replicates in the cytoplasm, and three nuclear replicating genera Alphanucleorhabdovirus, Betanucleorhabdovirus, and Gammanucleorhabdovirus (Dietzgen et al., 2020). The genome of plant rhabdoviruses consists of 3′ leader and 5′ trailer sequences and five viral structural protein genes organized in the order of (3′‐nucleoprotein [N], phosphoprotein [P], matrix protein [M], glycoprotein [G], and viral polymerase [L]‐5′) and separated by conserved gene junctions (Jackson & Li, 2016; Walker et al., 2011). The genomic RNA (gRNA) is complementary to its messenger RNA (mRNA); however, it requires a minimal infectious unit consisting of gRNA, N, P, and L proteins for initiation of virus transcription and replication. Unlike animal rhabdoviruses, plant‐adapted rhabdoviruses have at least one additional protein that mediates cell‐to‐cell movement in plant hosts (Mann et al., 2016; Zhou et al., 2019).

Constructing NSR infectious clones continues to be a technical challenge due to large genome sizes, ranging from 10 to 16 kb in size, and NSR genomes are not infectious without the nucleocapsid core of N, P, and L proteins; the full‐length genome and these proteins must be delivered into the same cell. Nevertheless, rhabdoviruses have several features that make them excellent viral vectors: (i) the virions are bacilliform and can encapsidate an expanded genome; (ii) the polar transcriptional mechanism of mRNAs provides regulated expression of cargos in the viral genome; and (iii) the inserts are stable due to their low rate of genome recombination (Finke & Conzelmann, 2005; Jackson & Li, 2016; Roberts & Rose, 1999). The first infectious clones for NSRs were developed for the animal rhabdoviruses, rabies virus (Schnell et al., 1994) and vesicular stomatitis virus (VSV) (Lawson et al., 1995). Cells were cotransfected with antigenomic RNA (agRNA) and viral replication proteins (N, P, and L) with transcription facilitated by bacteriophage T7 polymerase.

In the case of plant NSRs, the path for generating infectious virus from complementary DNA (cDNA) clones required overcoming additional difficulties. The lack of continuous cell cultures for plant NSRs made this technology more challenging, and the relative rigidity of the plant cell wall hinders the entry of multiple plasmids into a single cell to express the full‐length genome and replication proteins required for launching plant NSRs (Walpita & Flick, 2005). Additionally, all plant NSR infectious clones to date require co‐expressing viral suppressors of RNA silencing (VSRs: tomato bushy stunt virus [TBSV] p19, tobacco etch virus [TEV] HcPro or barley stripe mosaic virus [BSMV] γb proteins) that interfere with host RNA silencing to enhance ribonucleoproteins (RNPs) in trans (Ganesan et al., 2013; reviewed in German et al., 2020). Viral sequences are often unstable in Escherichia coli and Agrobacterium tumefaciens due to prokaryotic promoter‐like elements in the viral genomes, and spontaneous insertions or deletions can be a problem (Johansen, 1996; López‐Moya & García, 2000). Rescue of plant rhabdoviruses has been achieved for sonchus yellow net virus (SYNV) in Nicotiana benthamiana (Ganesan et al., 2013; Wang et al., 2015) and barley yellow striate mosaic virus (BYSMV) (Gao et al., 2019) and northern cereal mosaic virus (NCMV) in barley (Fang et al., 2022) by co‐expressing RNPs and VSRs. More recently, reverse genetics were developed for segmented NSRs, including tomato spotted wilt virus (TSWV) (Feng et al., 2020), rose rosette virus (RRV) (Verchot et al., 2020), and a mini‐replicon system for rice stripe virus (RSV) (Feng et al., 2021; Zhang et al., 2021).

NSRs represent a significant technical advance for expression of heterologous proteins and RNAs in plants, and an efficient NSR vector for maize has not been developed. Here, we used maize mosaic virus (MMV) as a model to develop the first infectious clone of an alphanucleorhabdovirus. The ability of MMV to infect maize and sorghum, two important crop plants, and the planthopper vector made MMV an attractive candidate for developing an infectious clone and an expression vector. Alphanucleorhabdovirus maydis is a typical member of the genus Alphanucleorhabdovirus and one of the best characterized plant rhabdoviruses (Ammar et al., 2009; Whitfield et al., 2018). MMV is transmitted in a persistent, propagative manner by its vector, Peregrinus maidis. The virus accumulates throughout insect development and has a broad tissue tropism in its vector (Ammar et al., 2009; Barandoc‐Alviar et al., 2016; Yao et al., 2019). The MMV genome consists of 12,170 bases and encodes six proteins in the order 3′‐N‐P‐3‐M‐G‐L‐5′ (Martin & Whitfield, 2019). Cellular localization and interactions of MMV proteins are well conserved between plant and insect cells (Martin & Whitfield, 2018). MMV is a versatile tool for stable expression of heterologous proteins in maize and planthoppers.

2. RESULTS

2.1. Insertion of plant intron in the L gene increased the stability of the MMV infectious clone

During construction of plasmids (pJL‐L and pJL‐MMV‐WT) containing L gene sequences, we observed that initial transformed colonies contained the entire desired fragment by colony PCR with MMV N, P and L gene‐specific primers (Table S1). However, plasmid extracted from overnight bacterial culture produced a low quantity of pJL‐MMV‐WT and pJL‐L plasmid. Moreover, none of the extracted plasmids generated desired fragments when validated with plasmid PCR and gene‐specific primers (Table S1). Growing bacterial cultures with plasmids at lower temperatures (25 and 30°C) also did not solve the viral genome instability in E. coli. Instability of full‐length MMV infectious clones and L gene in E. coli and Agrobacterium could be due to its large size or expression of toxic viral products from the MMV genome. We added an intron between the 3281 and 3282 splice sites (CTGCGGACAG^GTATCGATAT) of the L gene to improve the stability of pJL‐L and pJL‐MMV‐WT constructs and to prevent the expression of the polymerase gene that appeared to be detrimental to bacteria. Plasmid extracted from the overnight grown bacteria retained the full‐length L with the intron (Table S1). These results indicate that the presence of a plant intron in the MMV L gene increased the stability of the plasmid and growth of E. coli and Agrobacterium in liquid cultures at different temperatures for both L by itself and when present in the L of the full‐length viral genome.

2.2. Recovery of infectious MMV from cloned cDNAs in N. benthamiana leaves

To establish a reverse genetics system, NSRs require in vivo reconstitution of active replicase complexes of the N, P, and L proteins (Figure 1a). Thus, a mixture of Agrobacterium cultures harbouring the pJL‐MMV‐WT and supporting plasmids, pTF‐N&P and pJL‐L‐intron, were coinfiltrated into N. benthamiana leaves. No visible symptoms were observed in both infiltrated and systemic leaves at 12 days postinfiltration (dpi) in replicated experiments. Anti‐MMV virion immunoblots revealed viral structural proteins (MMV N, P, M, and G) in infiltrated but not systemic leaves (data not shown). This indicates that wild‐type MMV failed to move systemically in infiltrated N. benthamiana leaves. To monitor virus infection in vivo, we engineered the MMV antigenome with a green fluorescent protein (GFP) gene insertion between the duplicated N/P gene junction (Figure 1b). N. benthamiana leaves were agroinfiltrated with pJL‐MMV‐GFP together with expression constructs of N, P, and L genes. At 14 dpi, GFP expression was observed in individual cells of infiltrated leaves of N. benthamiana (Figure 2a). GFP fluorescence was not observed in the pJL‐MMV‐GFP without expression constructs of N, P, and L (Figure 2a). Immunoblot analysis showed the evidence of viral proteins (MMV N, P, M, and G) and GFP (Figure 2b) in the infiltrated leaves and indicates that they functionally interacted with the antigenome RNA (agRNA) transcripts to rescue recombinant MMV tagged with GFP (MMV‐GFP) in the agroinfiltrated N. benthamiana leaves.

FIGURE 1.

Rescue of infectious maize mosaic virus (MMV) from full‐length cDNA clones in plants. (a) Schematic diagrams of the pJL‐MMV‐WT, pTF‐N&P and pJL‐L‐intron plasmids. The full‐length MMV‐plasmid designed for transcription to yield the MMV antigenome RNA (agRNA) and contains the full‐length MMV cDNA positioned between a truncated CaMV double 35S promoter (2 × 35S) and the hepatitis delta ribozyme (Rz) sequence in pJL89 binary plasmid. Note that the sequences are shown in antigenomic (mRNA) sense. The full‐length cDNA of N and P were inserted between the 2 × 35S and 35S terminator sequences in pTF binary plasmid. The full‐length cDNA of L with plant intron ST‐LS1 inserted between 2 × 35S and 35S terminator sequences in pJL89 binary plasmid. (b) Schematic representation of agroinfiltration with Agrobacterium strains containing pJL‐MMV‐GFP, pTF‐N&P and pJL‐L‐intron plasmids and illustration of the pJL‐MMV‐GFP plasmid construction. The full‐length pJL‐MMV‐GFP contains duplicate N/P gene junctions flanking the GFP gene between the N and P genes of MMV antigenome cDNA. Le, leader; tr, trailer; Ter, terminator; TEV, tobacco etch virus; LB, left border sequence; RB, right border sequence. (c) Illustration of the MMV rescue procedure via Nicotiana benthamiana and transfer to maize and Peregrinus maidis planthoppers. dpi, days postinoculation. Figure 1c: Created with BioRender.com

FIGURE 2.

Maize mosaic virus (MMV) expression of GFP in the Nicotiana benthamiana and maize plants. (a) Foci of green fluorescent protein (GFP) in N. benthamiana leaves after agroinfiltration with Agrobacterium strains containing mock (Agrobacterium alone), MMV‐wild‐type (WT) alone, MMV‐GFP alone, pJL‐GFP, MMV‐WT + NPL, and MMV‐GFP + NPL. GFP foci were photographed at 14 days postinfiltration (dpi) with a fluorescence microscope (bars, 1000 μm). (b) Immunoblot analysis of MMV proteins and GFP in N. benthamiana plants with the anti‐MMV virion and anti‐GFP antibodies, respectively. (c) Maize leaves showing systemic mosaic symptoms 14 dpi with MMV‐WT and MMV‐GFP infected N. benthamiana sap by vascular puncture method. Images of maize leaves expressing GFP in the MMV‐GFP infected plants. GFP foci were photographed at 24 dpi with a fluorescence microscope (bars, 1000 μm). Mock plants inoculated with pJL‐MMV‐GFP alone crude extract did not show any symptoms. (d) Immunoblot analysis of MMV proteins and GFP in maize plants with the anti‐MMV virion and anti‐GFP antibodies, respectively. GelCode blue safe (ThermoFisher Scientific) stained RuBisCO bands served as loading control. M, prestained protein ladder (10–250 kDa). (e) Transmission electron micrograph of thin sections of MMV‐GFP‐infected maize showing particle accumulation in the nucleus. The dotted square box indicates the portion of the image that is magnified in the inset. Bars, 2 μm

2.3. Rescue of MMV‐WT and MMV‐GFP in maize

To test the ability of the MMV infectious clone to infect maize plants, crude extracts obtained from N. benthamiana leaves agroinfiltrated with pJL‐MMV‐WT + NPL or pJL‐MMV‐GFP + NPL were inoculated to maize kernels by vascular puncture inoculation (VPI). After 14 dpi, pJL‐MMV‐GFP plants developed mosaic symptoms on systemic leaves, similar to those observed in wild‐type MMV infections. In addition, MMV‐GFP + NPL resulted in high‐intensity GFP fluorescence that was systematically distributed in the symptomatic young leaves (Figure 2c). Immunoblot analysis confirmed that MMV viral proteins (N, P, M, and G) and GFP accumulated in systemically infected maize leaves (Figure 2d). Transmission electron microscopy analysis of thin sectioned maize cells infected with pJL‐MMV‐GFP revealed abundant bacilliform virus particles in the nucleus (Figure 2e).

2.4. Stable GFP expression by MMV vector in maize and planthoppers following virus passages

We demonstrated that the MMV‐GFP vector could be successfully used to stably express the GFP in passages between various hosts: from N. benthamiana to maize, maize to maize, maize to P. maidis, and P. maidis to maize. Stability of the inserted GFP gene was evaluated in plants and planthoppers following three successive passages by reverse transcription‐polymerase chain reaction (RT‐PCR) using MMV N‐ and GFP‐specific primers (Table S1). Furthermore, GFP fluorescence in the MMV‐infected plants was monitored weekly (Figure 3a,b). MMV‐GFP infected leaves (5L–9L) from the base of the plant were tested for virus and GFP. The result showed presence of MMV N and full‐length GFP bands in all the infected plants (Figure 3c,d). We also observed GFP in the MMV‐GFP infected plant leaves using fluorescence microscopy for the lifetime of the plant (data not shown). When P. maidis adults were injected with MMV‐GFP infected maize crude extract, three of 10 insects in the first passage, four of 10 in the second passage, and four of 10 in the third passage tested positive for MMV and GFP (Figure 3e). Only MMV‐GFP infected insects tested positive for both MMV and GFP. After the 24‐day inoculation access period, the infected P. maidis adults and nymphs showed GFP fluorescence (Figure 4a). Immunoblot analysis confirmed that MMV viral proteins (N, P, M, and G) and GFP accumulated in systemically infected P. maidis adults. The presence of MMV N and full‐length MMV GFP bands in all the three passage generations in both plants and planthoppers showed insert stability in the MMV vector (Figure 3c–f). We observed that purified MMV‐GFP virions injected to P. maidis nymphs showed intense GFP fluorescence in all tested insects at 7 and 12 dpi (Figure 4c). Moreover, we also used frozen MMV‐GFP infected leaf crude extract for injections, which also resulted in successful virus infection in insects and virus transmission.

FIGURE 3.

Time course observation of green fluorescent protein (GFP) expression in maize mosaic virus (MMV)‐GFP infected maize leaves. (a) GFP foci were photographed weekly at 2 to 9 weeks after inoculation (wai) with a fluorescence microscope (bars, 1000 μm). Maize plants inoculated with MMV‐wild‐type (WT) did not show green fluorescence. (b) Reverse transcription (RT)‐PCR analysis of virus and GFP in the MMV‐GFP infected maize plants collected from 2 to 9 wai. (c–f) Stability of GFP sequence carried by MMV vectors in all the three passage generations in both plants and planthoppers. (c) RT‐PCR analysis of maize plants inoculated with Nicotiana benthamiana crude extract harbouring MMV‐GFP derivatives via vascular puncture inoculation (VPI). (d) RT‐PCR analysis of maize plants inoculated with MMV‐GFP infected crude extract via VPI. 5L, 6L, 7L, 8L, and 9L indicate the leaf number that was sampled from the base of the plant. Plant 1 and plant 2 indicate two individual MMV‐GFP infected plants. (e) RT‐PCR analysis of individual Peregrinus maidis planthoppers injected with MMV‐GFP infected maize crude extract. (f) RT‐PCR analysis of MMV‐GFP infected independent maize plants transmitted by P. maidis. Controls, mock (maize, Agrobacterium alone‐inoculated plants; P. maidis, healthy maize crude extract‐injected planthoppers) and MMV‐WT, maize plants infected with MMV‐WT. MMV N and GFP primers were used to detect the virus presence and stability of the GFP insertion. Zmactin and PmeIF1 were used as internal controls for maize and planthoppers, respectively

FIGURE 4.

Maize mosaic virus (MMV) expression of green fluorescent protein (GFP) in planthoppers and virus transmission efficiency. (a) Peregrinus maidis adults and nymphs that fed on MMV‐wild‐type (WT) and MMV‐GFP infected maize plants at 28 days postinfiltration (dpi) were photographed with a stereomicroscope equipped with a GFP filter set. (b) Immunoblot analysis of the expression of MMV and GFP expression in P. maidis adults infected with MMV‐WT and MMV‐GFP. Total proteins were extracted from planthoppers that fed on healthy (mock) or MMV‐WT and MMV‐GFP infected maize plants were analysed by immunoblotting with antibodies against whole MMV virion and GFP antibodies. Each lane represents a pool of 10 insects feeding on MMV‐WT and MMV‐GFP infected plants and the numbers 1 and 2 indicate independent samples for the respective treatments. GelCode blue safe (ThermoFisher scientific) stained protein bands served as loading controls. M, prestained protein ladder (10–250 kDa). (c) MMV‐GFP virions injected P. maidis nymphs and adults at 7 and 12 dpi were photographed with a fluorescence microscope (bars, 1000 μm). (d) Transmission of MMV‐WT and MMV‐GFP to maize plants by P. maidis. Forty nanolitres of MMV‐WT and MMV‐GFP infected plant crude extracts were injected into P. maidis nymphs. Seven days after the incubation period, insects were transferred to 1‐week‐old healthy maize plants for a 2‐week inoculation access period. Each bar represents the mean and standard error (SE) of three independent experimental replicates, and each replicate consists of a group of 20 plants. Asterisks indicate significant difference in transmission between treatments (t test, p < 0.01)

2.5. MMV‐WT and MMV‐GFP virus transmission

Transmission efficiency of MMV‐WT and MMV‐GFP was measured in experiments with P. maidis nymphs that were injected with MMV‐WT or MMV‐GFP crude extract from infected maize plants. For MMV‐WT, after a 14‐day postinoculation access period, 19 out of 20 (95%) plants became infected in the first experiment, whereas 18 out of 20 (90%) and 19 out of 20 (95%) plants developed mosaic symptoms on systemic leaves in the second and third experiments, respectively (Figure 4d). For MMV‐GFP, after a 14‐day postinoculation access period, 12 out of 20 (60%) plants in the first experiment were infected with MMV while 13 out of 20 (65%) and 14 out of 20 (70%) plants in the second and third experiments developed mosaic symptoms on systemic leaves, respectively (Figure 4d). Moreover, MMV‐GFP plants showed delayed systemic symptom development compared to MMV‐WT plants. These results suggest that the addition of GFP to the viral genome had a slight effect on virus replication. Insect vector survival and transmission efficiency were improved by injecting insects with infected maize extracts compared to injection of infected N. benthamiana crude extracts. These results demonstrate that MMV‐GFP rescued in inoculated maize plants by VPI could be transmitted to P. maidis by injection of leaf extracts and recombinant MMV‐GFP is insect transmissible.

2.6. MMV gene expression does not differ between MMV‐WT and MMV‐GFP in planta

We evaluated the transcription levels using reverse transcription‐quantitative PCR (RT‐qPCR) of all MMV genes between MMV‐WT and MMV‐GFP infectious clones to verify the effect of introducing cargo proteins into the MMV infectious clone on the expression of downstream viral genes. For each gene, two sets of primers were used with similar efficiency (Table S1). The results demonstrated no significant effect on the gene expression of downstream genes in the MMV genome when GFP was inserted at the 3′ region of the genome (Figure 1b), as well as no significant difference observed in viral gene expression between MMV‐GFP and MMV‐WT for both primers tested for each gene (p > 0.05; Figure 5). Our results suggest that adding one single gene at the 3′ end of the MMV genome did not affect downstream viral gene expression. However, the effect of multiple insertions and the location of insertions in the MMV genome needs further characterization.

FIGURE 5.

Maize mosaic virus (MMV) gene expression did not differ between MMV‐wild‐type (WT) and MMV‐GFP. Boxplots of relative viral gene expression comparing MMV‐WT and MMV‐GFP using two different pairs of primers for each gene (Table S1) are shown. Analysis was performed 30 days after plant inoculation. Boxplots represent minimum and maximum, median, and interquartile ranges of virus gene expression. Dots represent individual values from each biological replicate (n = 6). No significant difference in gene expression was observed between MMV‐WT and MMV‐GFP (p > 0.05) except for GFP (**p < 0.01, ****p < 0.001), according to the nonparametric Mann–Whitney test. N, nucleocapsid protein; GFP, green fluorescent protein; P, phosphoprotein; 3, movement protein; M, matrix protein; G, glycoprotein; L, polymerase

3. DISCUSSION

A reverse genetics system for plant NSRs provides a platform for studying the basic biology of plant rhabdoviruses and delivery of RNAs and proteins to crop plants with the goal of characterizing and improving agronomic traits. In addition, plant virus‐based expression vectors have been used for the expression of vaccine antigens, antibodies, and other therapeutic proteins in planta (LeBlanc et al., 2020). To date, a few maize viruses have been developed for transient gene function studies (reviewed in Kant & Dasgupta, 2019; Mlotshwa et al., 2020; Xie et al., 2021). While important, these systems have limited use for maize researchers due to transgene instability, and severe symptoms in inoculated plants (Benavente et al., 2012; Ding et al., 2006; van der Linde et al., 2011). Improvements in high‐yielding hybrid crops, including maize, through transgenic modification is complex and limited to a few genotypes, requiring multiple generations to produce agronomically viable lines. In contrast, virus delivery technologies could hasten the breeding process by enabling direct modification of advanced breeding line varieties with desirable traits. Here, we developed an efficient NSR viral vector for expressing heterologous proteins in maize.

The creation of infectious clones for plant NSRs continues to be technically challenging, as each system requires a unique suite of “tricks” to make it work (German et al., 2020). In the construction of an MMV infectious clone, we observed plasmid instability due to spontaneous deletions of viral polymerase sequences during the propagation of cDNA clones in E. coli and A. tumefaciens cells. To circumvent the instability of plasmids containing the L gene, we inserted a plant intron into the polymerase gene L. This insertion increased the plasmid stability during propagation in E. coli and Agrobacterium. Introduction of a plant intron into the plant virus infectious clones has proved to be an efficient way to increase stability for viruses belonging to the genera Tobravirus (Ratcliff et al., 2001) and Potyvirus (Johansen, 1996; López‐Moya & García, 2000; Sun et al., 2017; Tran et al., 2019). Alternatively, Feng et al. (2020) used a codon optimization strategy to remove potential splice sites in the TSWV polymerase sequence. However, codon optimization can lead to alterations in protein conformation and function (Mauro & Chappell, 2014). Codon optimization was not required for successful reconstitution of MMV and that may be partly due to the biology of nuclear‐replicating viruses like MMV, whose mRNAs are adapted for nuclear exit similar to host mRNAs. We anticipate the introduction of introns for increasing L gene stability can be adapted to construct other plant rhabdovirus infectious clones. In this study, MMV was successfully rescued in N. benthamiana without VSRs, although VSRs were required for the initial rescue of other plant rhabdoviruses (e.g., SYNV, BYSMV, and NCMV) to increase the accumulation of core proteins and minireplicon derivatives (Fang et al., 2022; Gao et al., 2019; Ma & Li, 2020; Wang et al., 2015).

SYNV, the first plant rhabdovirus infectious clone, systemically infects N. benthamiana and this feature of SYNV biology facilitated development of this virus as a vector (Peng et al., 2021; Wang et al., 2015). In contrast, MMV initially propagates in individual N. benthamiana cells that have received the ectopically expressed viral replicase proteins (N, P, and L) and agRNA but is unable to move cell to cell or systemically, similar to BYSMV (Gao et al., 2019) and NCMV (Fang et al., 2022). GFP was added to the MMV cDNA clone, between the N and P genes with duplicated N/P gene junctions, to track virus infection in plants and insects. We excised GFP‐expressing cells harbouring recombinant MMV virions in N. benthamiana infiltrated leaves and injected cell extract to maize seeds using the VPI method. Furthermore, it is a major challenge to simultaneously deliver multiple essential components of the NSR rescue system, including core replication proteins and the genomic RNA, into maize cells. Inoculation of maize via agroinoculation or biolistic delivery of plasmids was unsuccessful for both MMV‐WT and MMV‐GFP (data not shown). Initial attempts to infect P. maidis by injecting N. benthamiana crude extract harbouring MMV‐WT also failed in transferring infectious virus from N. benthamiana to the insect vector. It is likely that the secondary metabolites in the crude extracts of N. benthamiana were toxic to the injected planthoppers. Hence, we utilized the VPI method to transfer infectious MMV clones from N. benthamiana to maize for the first time and, although we observed a low rate of infection via VPI (Table S2), we could efficiently generate virus‐infected plants via insect transmission once we moved MMV from N. benthamiana to maize. This allowed us to rescue the first alphanucleorhabdovirus entirely from cDNA clones and to introduce the virus to its natural host, maize.

The MMV infectious clone is a tractable system for stable expression of foreign genes in maize and insects over repeated passages. We observed that recombinant MMV could be stored as infected tissue or purified virus at −80°C and revived in maize by VPI or in insects by microinjection. This shows that recombinant MMV can be easily maintained long term. Comparison of gene expression between MMV‐WT and MMV‐GFP revealed that there were no significant differences in relative abundance of MMV transcripts downstream of the GFP insertion in the MMV genome. This finding suggests that, even with the additional open reading frame and gene junction sequences, the expression of MMV genes and/or stability of transcripts remains unchanged. The development of the BYSMV infectious clone was a major step forward for NSR viral vectors of monocots including barley, wheat, and foxtail millet; however, BYSMV distribution in maize leaves appeared patchy (Gao et al., 2019). The MMV clone supported strong expression of GFP throughout the entire maize leaves over 8 weeks postinoculation. The development of an efficient NSR vector for maize overcomes some of the drawbacks associated with positive‐strand RNA virus (+ssRNA)‐based tools that are currently available for monocots such as insertion size limits and insert instability during passages (Bouton et al., 2018; Cheuk & Houde, 2018; Jarugula et al., 2018; Mei et al., 2019; Tatineni et al., 2011). The differences in insert stability are due to the high recombination rates reported for +ssRNA viruses and the low‐to‐no recombination rates observed for NSR genomes (Chare et al., 2003; Patiño‐Galindo et al., 2021). The stability of the insertions in the MMV genome makes it a good candidate viral vector for expression of genes for crop improvement because they can be expressed over a long period of plant growth in the field or greenhouse. Similar stability and expression characteristics for MMV were observed in maize and planthopper vectors.

Here, we have developed a new virus vector using the maize alphanucleorhabdovirus, MMV, and demonstrated the successful GFP expression in both plants and insect vector, P. maidis. The major advantage of rhabdoviruses is that they are able to replicate in both plants and insect vectors, providing a convenient and highly efficient approach for transmission of gene editing reagents to multiple hosts through virus infection and insect transmission. Previous studies of MMV transmission biology, molecular interactions between P. maidis and the virus, and genome editing in the vector make this system a good platform for further characterizing rhabdovirus–vector interactions (Barandoc‐Alviar et al., 2016, 2022; Klobasa et al., 2021). Other rhabdoviruses, BYSMV (Gao et al., 2019) and SYNV (Ma et al., 2020), can express multiple guide RNAs and Cas nucleases. Thus far, successful genome editing using rhabdovirus vectors has been only demonstrated in N. benthamiana. With the ease of manipulation of MMV after initial launch, we expect to see rhabdovirus‐based editing in important crop plants in the near future as this is now a possibility for maize. Creation of an MMV infectious clone opens the door to a wide range of applications for expression of foreign genes and guide RNAs in maize and insects, and for genome editing to enhance crop productivity.

4. EXPERIMENTAL PROCEDURES

4.1. Construction of MMV infectious clone and introduction of intron into the L gene

The full‐length MMV (12,170 bp; accession number MK828539) cDNA was de novo synthesized in pcc1‐pbrick plasmid between the cauliflower mosaic virus promoter (CaMV) 35S promoter and the CaMV PolyA terminator (GenScript, Piscataway, New Jersey). To assemble full‐length MMV‐WT from the entire cDNA into pJL89 binary plasmid (Lindbo, 2007), fragments of 12,205 bp and 4137 bp were amplified from pbrick‐MMV and pJL89 using pJL‐MMV F/R and pJL F/R primers, respectively (Table S1). The forward primer contains 17 nucleotide (nt) overhangs complementary to the 3′ end of the double 35S promoter, whereas reverse primer contains 18 nt overhangs complementary to the 5′ end of the self‐cleaving hepatitis delta virus ribozyme (Rz), respectively. The two overlapping fragments were amplified using Q5 high‐fidelity DNA polymerase (NEB) and introduced into the pJL89 vector by one‐step NEBuilder Hifi assembly (NEB) (Figure 1a). To generate pTF‐N&P (plasmid for expression of N and P proteins) (Frame et al., 2002), MMV N and P were amplified using MMV NF/NR and PF/PR primers from cDNAs synthesized from infected‐plant RNA and subcloned into pTF binary plasmid between the double CaMV 35S promoter and nopaline synthase (NOS) terminator using Gateway Clonase enzyme mix (Invitrogen) (Figure 1a). To generate pJL‐L for expression of the polymerase, two overlapping PCR fragments 6936 bp and 4137 bp were amplified from pbrick‐MMV and pJL89 using 35SLF/TerLR and pJL F&R primers, respectively (Figure 1a). Competent cells of E. coli Top 10, DH10B, DH5α (ThermoFisher Scientific) and NEB stable competent E. coli (NEB) were used for cloning steps. Except pTF‐N&P, all the competent cells mentioned above tended to lose the full‐length MMV and L gene plasmids in overnight cultures at three different temperatures (25, 30, and 37°C).

To increase the stability of the plasmid, intron 2 (189 bp) of the light‐inducible gene ST‐LS1 (X04753) from Solanum tuberosum was introduced into the L gene between the 3281 and 3282 splice site (CTGCGGACAG^GTATCGATAT) nucleotides (Table S3). The putative intron splicing sites of wild‐type L gene sequences were predicted by Alternative Splice Site Predictor (ASSP) (Wang & Marin, 2006). To generate pUC‐MMV‐L‐intron (6784 bp), the MMV L gene from 2893 nucleotides‐ST‐LS1‐pJL89 terminator in pUC57 was synthesized de novo by GenScript. To generate pJL‐L‐intron, three overlapping PCR fragments LF1 (2943 bp) and LF2 (3487 bp) and a third fragment pJL89 (LF3, 4137 bp) were amplified from pbrick‐MMV, pUC‐MMV‐L‐intron, and pJL89 plasmids (Figure 1a) with the primer pairs 35SLF/MMVLXSR, LXSF/TerLR, and TerLF/pJLR, respectively. To generate pJL‐MMV‐intron, three overlapping PCR fragments of MF1 (3489 bp), MF2 (9224 bp), and MF3 (4137 bp) were amplified from pUC‐MMV‐L‐intron, pbrick‐MMV, and pJL89 plasmids using primer pairs LXSF/TerLF, pJLMMVF/MMVLXSR, and TerLF/pJLR, respectively. The overlapping fragments were purified and further assembled using NEBuilder Hifi assembly (NEB). The resulting pJL‐L‐intron was cloned in a binary vector pJL89 between the double 35S promoter (2 × 35S) at the 5′ terminus and the hepatitis delta virus ribozyme at the 3′ terminus. All the fragments (MF1, MF2, and MF3) were amplified using Q5 high‐fidelity polymerase (NEB) and introduced into binary plasmids using NEBuilder Hifi assembly (NEB) (Figure 1a). When the E. coli transformants were cultured overnight, pJL‐L and pJL‐MMV‐WT with the addition of the ST‐LS1 intron resulted in a single plasmid with expected characteristics when analysed by PCR and restriction digestion. Primers used for NEBuilder Hifi assembly (NEB) contained 10–15 nt overhangs at the 5′ end that was homologous to the 5′ end of another PCR product, so that the two PCR products could be ligated and circularized. All PCRs were carried out using Q5 high fidelity polymerase (NEB). NEB stable competent E. coli (NEB) was used in these cloning steps. All the primers used in this study are listed in Table S1.

To develop a recombinant MMV vector for foreign gene expression in plants, a duplicated N/P gene junction sequence along with the enhanced green fluorescent protein (eGFP) coding sequence were inserted into the pJL‐MMV‐WT plasmid between the N and P genes to generate pJL‐MMV‐GFP. The complete coding region of GFP followed by the N/P gene junction sequence was synthesized by GenScript. To facilitate subcloning, a Bsu36I (ThermoFisher Scientific) site was introduced at either ends of the GFP and N/P gene junction clone. Both pJL‐MMV‐WT and pUC57‐GFP‐N/P plasmids were digested by Bsu36I and ligated to generate pJL‐MMV‐GFP (Figure 1b). In this configuration, GFP mRNA synthesis is initiated immediately after termination of the upstream N protein mRNA synthesis by the duplicated N/P gene junction and is followed by P mRNA synthesis that is directed by the native N/P gene junction. Clones with correctly oriented GFP‐N/P inserts were verified by MMV1355F and MMVPR primers and sequencing. All constructs were confirmed by sequencing and transformed into A. tumefaciens GV3101.

4.2. Agrobacterium infiltration

Virus infection of N. benthamiana was achieved by Agrobacterium‐mediated transient gene expression of infectious constructs from the T‐DNA of binary plasmids pJL89 and pTF. A. tumefaciens GV3101 with pJL‐MMV‐WT, pJL‐MMV‐GFP, pJL‐L plasmids (kanamycin, 50 μg/ml), and pTF‐N&P plasmid (spectinomycin, 50 μg/ml) were individually grown in Luria broth containing rifampicin (50 μg/ml) and gentamicin (50 μg/ml). Cells grown overnight were harvested by centrifugation and resuspended in the infiltration buffer and incubated for 3 h at room temperature in the presence of 100 μM acetosyringone, as described in Kanakala et al. (2019). After incubation, equal volumes of Agrobacterium cultures harbouring the pJL‐MMV‐WT or pJL‐MMV‐GFP alone and pJL‐MMV‐WT or pJL‐MMV‐GFP + NPL were coinfiltrated into the N. benthamiana with a final OD₆₀₀ of 0.5 for each construct.

4.3. Vascular puncture inoculation, crude sap injections, and insect transmission studies

N. benthamiana leaf extracts infiltrated with pJL‐MMV‐WT and pJL‐MMV‐GFP plasmids were homogenized in extraction buffer (10 mM potassium buffer; pH 7) and centrifuged at 12,000 × g for 5 min at 4°C (Louie, 1995). Then 5 μl of crude supernatant was vascular puncture inoculated using a five‐pin tattoo needle into the overnight soaked maize kernels at 30°C (Figure 1c). Then the VPI seeds were incubated in a humid tray at 30°C for 2 days before sowing. These plants were subsequently observed for systemic symptom expression or image analysis. The symptomatic plant leaves showing mosaic symptoms and GFP expression were homogenized in extraction buffer (100 mM Tris–HCl, 10 mM Mg [CH₃COO]₂, 1 mM MnCl₂, and 40 mM Na₂SO₃, pH 8.4) and centrifuged at 12,000 × g for 10 min at 4°C (Jackson & Wagner, 1998). Then, 40 nl of the crude extract supernatants was injected into the anaesthetized P. maidis (nymphs) using a Nanoinject III Programmable Nanoliter injector (Drummond Scientific Co.) (Figure 1c). In three independent experiments, the surviving injected nymphs were maintained on the healthy maize seedlings for a 7‐day incubation period and then transferred to healthy maize seedlings for a 2‐week inoculation access period (Figure 1c). Adult insects injected with extraction buffer alone (about 30 per replicate) were used for control experiments. Symptoms in plants and GFP expression in systemic leaves and insects were monitored weekly and documented.

4.4. Electron microscopy and image acquisition

Systemically infected leaf tissues were fixed in 2.5% glutaraldehyde and 1% osmium tetroxide (both in 100 mM sodium cacodylate buffer, pH 7.0) as previously described by Kong et al. (2014). Following ethanol dehydration, the fixed tissues were then embedded in Spurr's resin (Spurr, 1969) as instructed by the manufacturer (Sigma‐Aldrich). Ultrathin sections (90 nm) were cut with a Diatome diamond knife from the embedded tissues using an EM U7 ultramicrotome (Leica). The thin sections were then stained with 4% uranyl acetate for 20 min followed by lead citrate for 30 s. The stained sections were imaged on a ThermoFisher Talos F200X transmission electron microscope operated at 80 kV to determine virion structure in the infected cells.

The fluorescence of plant leaves and insects was observed with the objective of 4X using the BioTek slide reader and Image software v. 3.04. Images were captured using default settings. Leaves agroinfiltrated with pJL89 alone were used as a negative control.

4.5. RNA extraction, cDNA synthesis, RT‐PCR, and quantitative real‐time PCR

To assess reporter gene stability, multiple successive passage experiments were performed in the following order: (1) N. benthamiana to maize, (2) maize to maize, (3) maize to P. maidis, and (4) P. maidis to maize. During N. benthamiana to maize and maize to maize passages, we collected infected leaves from 2 to 9 weeks after inoculation for further analysis. To determine the passage of maize to P. maidis, MMV‐GFP infected crude sap was injected into insects and collected 2 weeks after release on maize plants for further analysis. To determine virus passage from P. maidis to maize, we collected infected maize leaves for further analysis. Leaves of vascular puncture‐inoculated maize plants and crude sap injected adult P. maidis were harvested for RNA extraction using the RNeasy Plant Mini kit (Qiagen) and TRIzol reagent (ThermoFisher Scientific) according to the manufacturer's recommendations, respectively. For virus detection and gene expression quantification, an aliquot containing 1.5 μg of total RNA was used for the first‐strand cDNA synthesis using the Verso cDNA Synthesis Kit with RT‐enhancer (ThermoFisher Scientific) to remove residual genomic DNA, following the manufacturer's instructions. After first‐strand cDNA synthesis, primer pairs MMV N F&R, GFP F&R were used to detect the presence of MMV by PCR. Zea mays actin and P. maidis eIF1 genes were used as an internal control with primer pairs Zmactin F&R and PmeIF1 F&R, respectively.

To compare gene expression between MMV‐WT and MMV‐GFP in planta, single maize plants were inoculated by a single viruliferous, brachypterous P. maidis male from an age‐calibrated colony according to Yao et al. (2019). Thirty days after inoculation, the second youngest maize leaf was collected from symptomatic plants and 50 mg of tissue was subjected for RNA extractions and used for cDNA synthesis. Primers were designed for all MMV genes using the PrimerQuest Tool software from Integrated DNA Technology (IDT) using as template the full‐length MMV genome (GenBank accession number MK828539), with default settings. A total of 34 pairs of primers were first tested for efficiency by using 5‐fold serial dilutions of cDNA synthesized using oligo‐dT from total RNA obtained from MMV‐GFP infected maize plants as template for RT‐qPCR (Table S1). Efficiency was calculated based on the standard curve method. Primer pairs with efficiencies between 90% and 110% and showing a single peak based on the melting curve were selected for further analyses. The highest difference in efficiency between selected primers pairs for a given gene was 6% (Table S1). In a final volume of 10 μl, the qPCRs contained 4 μl of 7‐fold diluted cDNA, 1 μl of mixed primers containing (5 M of each primer [forward and reverse]), and 5 μl of iTaq Universal SYBR Green Supermix (Bio‐Rad). The qPCR cycles consisted of an initial denaturing step at 95°C for 1 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min, followed by melting curve analyses. All reactions were performed in duplicate. Expression levels for each gene were quantified by the comparative cycle threshold method (2^−∆∆Ct) (Livak & Schmittgen, 2001) using two primers pairs for each gene and the membrane protein (MEP) as internal reference gene (Manoli et al., 2012) (Table S1). Statistical analyses were performed using GraphPad Prism software v. 9.2.0. To compare differences in gene expression between MMV‐WT and MMV‐GFP the expression data were log₁₀ transformed and subjected to the nonparametric Mann–Whitney test.

4.6. Immunoblot analysis of viral proteins

For immunoblot analysis, proteins were separated on 4%–15% gradient Mini‐PROTEAN TGX Precast gels (Bio‐Rad). Immunoblot analysis of the plant and insect protein extracts was performed according to the procedures described previously (Caplan et al., 2008). Proteins were transferred to nitrocellulose membranes using the Trans‐Blot Turbo Transfer system (Bio‐Rad). Membranes were blocked using 5% nonfat dry milk for 1 h then washed in three times in Tris‐buffered saline containing 0.1% Tween‐20 (TBS‐T). Transferred membranes were incubated with anti‐MMV virion (made against purified MMV virus), this antibody recognizes the main structural proteins of the virus (N, P, M, and G when used at 1:2000 dilution), anti‐GFP conjugated with horseradish peroxidase (1:2000 dilution; Invitrogen) or anti‐β‐actin monoclonal antibody (BA3R) conjugated with horseradish peroxidase (1:2000 dilution; Invitrogen) with gentle shaking overnight at 4°C to detect proteins. The antibody‐incubated membranes were washed three times with TBS‐T before adding secondary antibody goat anti‐rabbit (for MMV protein detection) IgG‐conjugated alkaline phosphatase for 2 h at room temperature with gentle shaking, then washed three times with TBS‐T before addition of SuperSignal West Pico PLUS chemiluminescent substrate (ThermoFisher Scientific) following the manufacturer's instructions. Quantification of band signal intensity was performed with the iBright Imaging system (CL1000; ThermoFisher Scientific).

CONFLICT OF INTEREST

The authors have submitted a patent related to negative‐strand RNA virus vectors and received funding from industry sponsors for related research.

Supporting information

Figure S1 Transmission electron micrographs of thin sections of MMV‐GFP infected maize showing MMV‐GFP virions particles accumulated in the nucleus and cytoplasm of the image. White arrows indicate virions in the cytoplasm. Bars, 2 μm

Figure S2 Microscopy images of green fluorescent protein (GFP) expressed from MMV‐GFP infections in the Peregrinus maidis nymphs and adults were microinjected with MMV‐GFP purified virions. Insects were photographed with a fluorescence microscope at 7 and 12 days postinjection. Bars, 1000 μm

Table S1 List of primers used in this study

Table S2 Stable GFP expression by the MMV vector in Nicotiana benthamiana, maize, and planthoppers following virus passages

Table S3 The predicted intron splicing sites of MMV L gene. Alternative Splice Site Predictor (ASSP): MMV L splice site prediction. Sequence: MMV L; sequence length 5769 bp, acceptor site cutoff: 2.2 and donor site cutoff: 4.5

Kanakala, S. , Xavier, C.A.D. , Martin, K.M. , Tran, H.H. , Redinbaugh, M.G. & Whitfield, A.E. (2023) Rescue of the first alphanucleorhabdovirus entirely from cloned complementary DNA: An efficient vector for systemic expression of foreign genes in maize and insect vectors. Molecular Plant Pathology, 24, 788–800. Available from: 10.1111/mpp.13273

DATA AVAILABILITY STATEMENT

Data supporting the findings of this study are available within the paper and its supplements.