Functional analysis of apple stem pitting virus coat protein variants

Apple stem pitting virus (ASPV) is the type species of the Foveavirus genus in the Betaflexiviridae family [1]. It possesses a single stranded positive RNA (+ssRNA) genome comprising of approximately 9300 nucleotides (nts), which encodes five open reading frames (ORFs, ORF1-ORF5) as well as the 5′ untranslated region (UTR) and 3’ UTR. ORF1 encodes the viral RNA-dependent RNA polymerase (RdRP), ORF2-ORF4 encode triple gene block proteins (TGBp1-TGBp3) and ORF5 encodes the viral coat (capsid) protein (CP) [2]. ASPV infects several plant species and causes a wide range of symptoms from symptomless to xylem pits, epinasty, decline, vein yellowing, leaf red mottling, pear necrotic spot or fruit stony pits depending on the plant species, the cultivar and the viral strain/isolate [2–5].

It has been shown that when a virus adapts to a new host, variation is primarily manifested as amino acids substitutions, which allows virus entry into the new host efficiently, blocks interactions with host proteins or allows the virus to circumvent immunity in both the new and the old host [6–8]. The RdRP encoded by many RNA viruses are known to be error-prone, and this error-prone replication is thought to be important for viruses to generate a pool of different progeny genomes to adapt to potentially diverse new hosts [9]. Several studies have shown that each ASPV ORF possesses a high degree of genetic variability between/within isolates [4, 5, 10–13], especially in the CP-encoding ORF. Our previous study showed that ASPV CP variants could be divided into three groups (pear group, apple group and Korla pear), which correlated with their isolated hosts [13]. Furthermore, variants from pear isolates could be divided into six subgroups (subgroup A-F), and CP variants from a different subgroup have a different CP size because of amino acid insertions or deletions in the N terminal portion of CP [13]. These previous observations might imply host-driven adaptations have affected genetic diversification of ASPV CP variants.

In addition to mediating encapsidation and protecting the viral genome from degradation, multiple reports have shown that viral CPs play multiple functions, including roles in viral RNA translation, viral RNA replication, viral movement, activation of host immune, RNA binding, virus transmission, symptom development, and viral suppressor of RNA silencing (VSR) [14]. TGBp1 proteins, encoded by several viruses in the genus Potexvirus, have been shown to have VSR activity [15, 16]. Potexvirus and Foveavirus are phylogenetically related viruses, both of which encoded potex-like TGB proteins as viral movement protein [17]. TGBp1 encoded by TGB-encoding viruses were clustered into two major groups by phylogenetic analysis of the NTPase/helicase (which is a conserved domain) sequences of TGBp1, filamentous viruses (genera Potexvirus, Carlavirus, Foveavirus and Allexivirus) and rod-shaped viruses (genera Hordeivirus, Benyvirus, Pomovirus and Pecluvirus) [18]. However, none of those functions have been explored with respect to ASPV: whether TGBp1 encoded by ASPV functional as a VSR or whether the diversity of symptoms induced by ASPV in its natural hosts is due to the genetic diversity of its genome, including the CP. In this study, we selected six CP sequences belonging to different subgroups including HB-HN1–3/subgroup A, HB-HN7–18/subgroup B, YN-MRS-17/subgroup D, HB-HN6–8/subgroup E, HB-HN9–3/subgroup F, LN-AP-1/apple group [13], to test whether differences in CP amino acid sequences could result in changes in symptomatology, VSR activity, and serological reactivity.

In this study, we have shown that ASPV isolates originating from pears with or without symptoms (Fig. 1a) or symptomless apple plants in the same orchard (Wuhan, China) induced different symptoms in N. occidentalis (Fig. 1b). Viral infections can induce symptoms by interacting and interfering with host components [28], and viral CPs have often been implicated in contributing to infectivity, pathogenicity, and symptom development [14, 29]. Our result showing that each PVX-ASPV-CP enhanced PVX symptoms in N. occidentalis implied that ASPV CP also functions as a symptom determinant (Fig. 5). This may be related to its VSR activity as similar increases in symptoms have been observed when VSRs from other viruses are expressed from the PVX genome [30–34]. We do not rule out other possibility, but it seems that distinct characteristics of ASPV CP variants may account for the different symptoms induced by ASPV isolates on N. occidentalis.

First, we have previously shown that CP variant composition of these isolates is complicated and sequences of CP variants from subgroups have great differences [13]. For example, isolate HB-HN9 consists of variants from subgroup E and F, HB-HN7 (subgroup A and B), HB-HN1/HB-HN2 (subgroup A), HB-HN6/HB-HN10 (subgroup E), HB-AP1 (Apple group) [13]. Pervious study showed that viral CP modifications lead to symptom changes, fifteen Pepino mosaic potexvirus (PepMV) isolates that shared a very close similarity in their CP sequences caused similar symptoms, while one strain that has great differences in the CP sequence did not induce symptoms in tomato [35]. The N-terminus of ASPV CP variants belonging to different subgroups have been shown to have high variability in length and sequence [13], thus it is likely that N-terminal of ASPV CP variants is differently involved in interactions with host factors. This is not only the case of ASPV CP, as several reports have shown that differences in N-terminus of viral CPs results in differences in CP functions: The N-terminus of Potato Virus Y (PVY) CP plays a critical role in symptom development and determines the pathogenicity of different PVY isolates [36]. In Arabidopsis Col-0 and WS plants, the systemic infection of Tobacco Etch potyvirus (TEV) was restricted by RTM genes (Restricted TEV Movement, RTM1, RTM2, and RTM3) through interaction with TEV CP, with amino acid changes in the CP N-terminus resulting in resistance-breaking strains of TEV [37, 38]. The N-terminus of the Plum pox potyvirus (PPV) CP is likely important to virus systemic movement, probably via effects on virion assembly and/or stability; modifications in the N-terminal 30 aa region of two PPV CP variants results in systemic infection of N. benthamiana and N. clevelandii by PPV [39]. The molecular basis of this link is not yet clear, although it is possible that differences in CP sequences affect viral transport or virion assembly and/or stability in plant cell.

Consistent with the above examples, we next showed that in the absence of ASPV, YFP-CPs variants have different abilities to form aggregates in the cytoplasm of infiltrated N. benthamiana cells at 72 hpi, with variant LN-AP1–1 forming the most aggregates, followed by HB-HN6–8. Viral proteins form aggregates in the cytoplasm and nuclei of infected cells [40]. However, the role of viral proteins aggregates remains dynamic: viruses may use aggregates to concentrate host and viral proteins in one place in the cell to facilitate replication, assembly and movement. Alternatively, aggregates may form part of an innate cellular immunity response that recognizes virus components and targets them for storage and degradation [40]. A previous study showed that Tomato yellow leaf curl virus (TYLCV) CP forms aggregates of increasing size during viral infection, suggested a functional role for aggregates during the development of infection [41]. If the same were true for ASPV CP aggregates, we would infer that the CP variant (LN-AP1–1), which has a stronger aggregate formation propensity, could recruit more viral and host components required for viral replication and assembly in the same infection time. As a result, viral isolates that harbor this kind of CP variant would replicate efficiently during early infection stages and more virus progeny would appear in infected plant cells. Unfortunately, this could not be further explored due to a lack of ASPV infectious clone, which needs to be done in future work. Since the CP of each ASPV isolate contained two or more variants in pear or apple plants, and different variants have different aggregation propensity, which then probably results in different infection efficiency. This result implies that the different symptoms induced by ASPV isolates in N. occidentalis may be affected by different CPs variants due to the intrinsic propensity to aggregate. Our study of different ASPV CP variants aggregation propensity in plant cells provides new insights into virus-host interactions. RNA silencing is one of the major documented antiviral mechanism that inhibits RNA viruses and RNA viruses encode VSRs as a strategy to defend against RNA silencing. However, in contrast to Potexvirus encoded TGBp1, our results indicate that TGBp1 encoded by ASPV, which is a member of Foveavirus, is neither a local nor a systemic silencing suppressor (Additional file 2: Figure S1). Interestingly, ASPV CP, when expressed in PVX vector, suppressed local silencing in N. benthamiana (Fig. 4) and induced more serious symptoms in N. occidentalis compared to PVX (wt) (Fig. 5). Several lines of evidence support our result, firstly, proteins homologous to TGB1 encoded by members of different viral genera do not necessarily have the same functions. For example, TGBp1 encoded by Barley stripe mosaic virus (BSMV, genus Hordeivirus) or Peanut clump virus (PCV, genus Pecluvirus) does not function as a VSR, instead, γb of BSWV and p15 encoded by PCV supressed RNA silencing [42, 43]; Potato virus M (PVM), a member of the genus Carlavirus, employs a dual strategy to block host antiviral silencing, TGBp1 showed suppressor activity only on systemic silencing, whereas a cysteine-rich protein (CRP) encoded by PVM inhibited both local and systemic silencing [44]; even in the genus Potexviruse, the ability of TGBp1 encoded by potexviruses to suppress RNA silencing varies significantly in N. benthamiana: PVX TGBp1 (p25) is one of the well-characterized VSR, supressed systemic silencing but not local silencing [26]; TGBp1 encoded Tulip virus X (TVX) showed very weak (almost no) local silencing ability [16], wherein TGBp1 of Plantago asiatica mosaic virus (PlAMV), Asparagus virus 3 (AV3), and White clover mosaic virus (WClMV) have very strong local and systemic silencing suppressor activity [16]. Secondly, the CP encoded by several plant viruses have been shown to be a VSR, for example, Cowpea mosaic virus (CPMV), Tomato ringspot virus (ToRSV), Cocksfoot mottle virus (CfMV) [45–47]. Likewise, in the case of the Potexvirus PepMV, in addition to the TGB1 protein, the CP also showed strong VSR activity [48]. Third, many VSRs have been identified as pathogenicity determinants or symptom determinants [49] and consistent with this, we showed that the ASPV CP is also a symptom determinant (Fig. 5).

In addition, we showed that RNA silencing suppressing ability of six ASPV CP variants that differed greatly at their N termini, turned out to be the same (Fig. 4). This is not a surprise, since all identified VSRs are diverse in sequence and structure, and act via different molecular mechanisms [50]. How ASPV CPs variants maintain their conserved VSR activity remains unknown in this study, although it is possible that the conserved C-terminal of ASPV CP mediates its VSR ability [13]. The sequencing of the pear (Pyrus bretschneideri Rehd.) genome [51] and the identification of some RNA silencing components in pear provide a great opportunity to further study the antiviral role of pear RNA silencing components [52] and how ASPV CP functions in counteracting the RNA silencing pathway.

Different levels of serological reactivity among the tested six CPs with the three antibodies (Fig. 2) underlined the necessity of producing different antibodies according to different ASPV isolates. ASPV diagnosis in the field, at least in China, would benefit from our work. Also, considering the 3′ terminal region of the ASPV CP is relatively conserved, in the future, it would be more useful to raise antibodies against the more conserved part of ASPV CP.

Taken together, our results indicate that the ASPV CP has VSR activity, which is conserved among the different CP variants. However, CP variants have different abilities to aggregate in N. benthamiana, probably due to amino acid differences in the N terminus, which may also result in different symptoms induced by ASPV isolates in Nicotiana. occidental. Until now, little has been known about the molecular defense and counter defense mechanisms between ASPV and its host, however, our study provides insights that can be used to further explore this topic.