Sugarcane Mosaic Disease: Characteristics, Identification and Control

Mosaic is one of the most important sugarcane diseases, caused by single or compound infection of Sugarcane mosaic virus (SCMV), Sorghum mosaic virus (SrMV), and/or Sugarcane streak mosaic virus (SCSMV). The compound infection of mosaic has become increasingly serious in the last few years. The disease directly affects the photosynthesis and growth of sugarcane, leading to a significant decrease in cane yield and sucrose content, and thus serious economic losses. This review covers four aspects of sugarcane mosaic disease management: first, the current situation of sugarcane mosaic disease and its epidemic characteristics; second, the pathogenicity and genetic diversity of the three viruses; third, the identification methods of mosaic and its pathogen species; and fourth, the prevention and control measures for sugarcane mosaic disease and potential future research focus. The review is expected to provide scientific literature and guidance for the effective prevention and control of mosaic through resistance breeding in sugarcane.


Introduction
Sugarcane (Saccharum spp. hybrids), the most important sugar and energy crop originating in the tropics, is a perennial, high biomass herb ratoon C 4 crop. It is widely cultivated in more than 100 countries or regions in the tropics and subtropics, with a total area of about 27 million hectares. The world annual output is about 1.95 billion tons of fresh cane, which provides nearly 80% of sugar, 60% of bioethanol, and a total economic value of 75 billion U.S. dollars (FAO, 2019, http://www.fao.org/faostat/zh/#data/, accessed on 25 May 2021). Furthermore, the pressed cane juice can be used to produce diesel, jet fuel, and other high-value products [1,2]. Sugarcane by-products can also be used for direct-fired power generation, field fertilizers, and culture substrate for fruit tree seedlings [3,4].
Mosaic is one of the main viral sugarcane diseases. Systemic infection is caused by the virus after it invades sugarcane. The incubation period is generally about 10 d, but can be up to 20-30 d. The disease may even manifest in the second year of infection [5]. The disease was first described in 1892 by Musschenbroek [6] in Java as "yellow stripe disease". Subsequently, it was found in Australia [7], Puerto Rico, the United States [8], and India [9]. In 1920, Brandes identified the disease as a transmissible viral disease that could be transmitted by aphis (Rhopalosiphum maidis Fitch) [10]. Summers et al. [11] speculated that the disease started in New Guinea and was introduced into Java from infected sugarcane, and then further spread to the Americas and other countries [12]. So far, mosaic has been widely discovered in most sugarcane planting regions around the world [13,14].
Before the 1990s, scientists generally agreed that mosaic was caused by the Sugarcane mosaic virus (SCMV). Since then, the Sorghum mosaic virus (SrMV) [15] and the Sugarcane

Transmissions
The primary infectious sources of mosaic mainly include infected plants of sugarcane and other Gramineae hosts. In nature, transmission of SCMV and SrMV is primarily by several aphid vectors including Dactynotus ambrosiae [55], Hysteroneura setariae [56], Longiunguis sacchari [57,58], Rhopalosiphum maidis [10,59], and Toxoptera graminum [60][61][62] in a non-persistent manner. Ants also have indirect transmission effect if they interact actively with aphids in diseased sugarcane fields [5,63]. However, insect borne SCSMV has not yet been detected [53,64,65], although Triticum mosaic virus (TriMV) and Wheat streak mosaic virus (WSMV), which belong to the same genus and share a high sequence similarity with SCSMV, can be transmitted by wheat curl mites (Aceria tosichella Keifer) [16,66]. The three viruses are easily transmitted over a short distance by machines, cutting tools, and juice fluid friction, but transmission over long-distance is mainly through infected materials [12,20]. A diagram of the specific transmission pathway is shown in Figure 2.

Transmissions
The primary infectious sources of mosaic mainly include infected plants of sugarcane and other Gramineae hosts. In nature, transmission of SCMV and SrMV is primarily by several aphid vectors including Dactynotus ambrosiae [55], Hysteroneura setariae [56], Longiunguis sacchari [57,58], Rhopalosiphum maidis [10,59], and Toxoptera graminum [60][61][62] in a non-persistent manner. Ants also have indirect transmission effect if they interact actively with aphids in diseased sugarcane fields [5,63]. However, insect borne SCSMV has not yet been detected [53,64,65], although Triticum mosaic virus (TriMV) and Wheat streak mosaic virus (WSMV), which belong to the same genus and share a high sequence similarity with SCSMV, can be transmitted by wheat curl mites (Aceria tosichella Keifer) [16,66]. The three viruses are easily transmitted over a short distance by machines, cutting tools, and juice fluid friction, but transmission over long-distance is mainly through infected materials [12,20]. A diagram of the specific transmission pathway is shown in Figure 2.

Epidemiology
The severity of mosaic disease in sugarcane fields is closely associated with sugarcane variety, infected setts, climatic conditions, and intermediate hosts. Among the six Saccharum species, S. officinarum are highly susceptible, S. sinense, S. spontaneum [67] and S. barber [68] are highly resistant or immune, S. robustum [67,68] are susceptible to mosaic. This disease was also found on S. edule [69], but the resistance to viruses is still uncertain. However, a recent study showed that most of Saccharum species have poor resistance to SCSMV; only three out of eight accessions of S. robustum among all of the 210 tested clones of Saccharum are identified to be resistant [70]. Generally, sugarcane cultivars with more resistant consanguinity also tend to show stronger resistance [65]. Young sugarcane plants are more susceptible than old, mature plants [71]. Drought and less rainfall environments are beneficial to the reproduction and activities of aphids, which promote the spread of mosaic. However, an extremely hot climate is not conducive to disease transmission, leading to slow virus proliferation, less disease symptoms, and less severe disease incidence [5,13]. In general, mosaic often occurs seriously in weedy or intercropping sugarcane fields [35]. High susceptibility of main varieties, relatively high temperatures and less rain, different planting periods in the same region, long-term rotation, and single variety and long-term successive cropping can all lead to a serious occurrence or epidemic of mosaic [72].

Epidemiology
The severity of mosaic disease in sugarcane fields is closely associated with sugarcane variety, infected setts, climatic conditions, and intermediate hosts. Among the six Saccharum species, S. officinarum are highly susceptible, S. sinense, S. spontaneum [67] and S. barber [68] are highly resistant or immune, S. robustum [67,68] are susceptible to mosaic. This disease was also found on S. edule [69], but the resistance to viruses is still uncertain. However, a recent study showed that most of Saccharum species have poor resistance to SCSMV; only three out of eight accessions of S. robustum among all of the 210 tested clones of Saccharum are identified to be resistant [70]. Generally, sugarcane cultivars with more resistant consanguinity also tend to show stronger resistance [65]. Young sugarcane plants are more susceptible than old, mature plants [71]. Drought and less rainfall environments are beneficial to the reproduction and activities of aphids, which promote the spread of mosaic. However, an extremely hot climate is not conducive to disease transmission, leading to slow virus proliferation, less disease symptoms, and less severe disease incidence [5,13]. In general, mosaic often occurs seriously in weedy or intercropping sugarcane fields [35]. High susceptibility of main varieties, relatively high temperatures and less rain, different planting periods in the same region, long-term rotation, and single variety and long-term successive cropping can all lead to a serious occurrence or epidemic of mosaic [72].  [15,73]. SCSMV belongs to Poacevirus, as do TriMV and Caladenia virus A (CalVA) [17].  [15,73]. SCSMV belongs to Poacevirus, as do TriMV and Caladenia virus A (CalVA) [17].

Morphology, Size, and Viability
Similar to all members of the Potyviridae family, the three viruses present nonenveloped, flexuous-filamented viral particles [73]. The basic pathogenic characteristics of the sugarcane mosaic virus is shown in Table 1. Note: "-" means undetermined.

Genome Structure
The genome of SCMV, SrMV, and SCSMV is represented by a positive-sense singlestranded RNA (+ssRNA) of about 10 Kb, consisting of untranslated regions (UTR) at both ends and a single open reading frame (ORF) encoding for a large polyprotein. The viral RNA harbour a genome-linked protein (VPg) at the RNA 5 -terminus and a poly (A) tract at the 3 -terminus [75]. The genome structure of the sugarcane mosaic virus is shown in Figure 3. The polyprotein is processed by the virus-encoded proteases P1-pro, HC-Pro and NIa-Pro into 10 mature functional proteins [73,76]. In addition, SCMV and SrMV encode Microorganisms 2021, 9, 1984 5 of 19 an additional PIPO [77], and SCSMV encodes P3N-PIPO, which are expressed from the P3 ORF through a +2 or +1 frame-coding slippage mechanism, respectively [78,79].
The genome of SCMV, SrMV, and SCSMV is represented by a positive-sense singlestranded RNA (+ssRNA) of about 10 Kb, consisting of untranslated regions (UTR) at both ends and a single open reading frame (ORF) encoding for a large polyprotein. The viral RNA harbour a genome-linked protein (VPg) at the RNA 5′-terminus and a poly (A) tract at the 3′-terminus [75]. The genome structure of the sugarcane mosaic virus is shown in Figure 3. The polyprotein is processed by the virus-encoded proteases P1-pro, HC-Pro and NIa-Pro into 10 mature functional proteins [73,76]. In addition, SCMV and SrMV encode an additional PIPO [77], and SCSMV encodes P3N-PIPO, which are expressed from the P3 ORF through a +2 or +1 frame-coding slippage mechanism, respectively [78,79].  Table 2 describes the main functions of the proteins encoded by viruses of the Potyviridae family. The PIPO and P3N-PIPO mainly affect the movement of the virus between cell-to-cell movement. P3N-PIPO binds to CI to recruit itself into plasmodesmata to promote intercellular movement of the virus [77,80]. Compared with SCMV and SrMV, a highly conserved motif of "Asp-Ala-Gly (DAG)" was absent in the CP sequence of SCSMV, which is necessary for aphid transmission [81,82]. It is worth mentioning that RNA silencing and RNA silencing repressors are mechanisms of defence and counterdefence interactions between host plants and viruses [83]. HC-Pro is the first strong RNA  Table 2 describes the main functions of the proteins encoded by viruses of the Potyviridae family. The PIPO and P3N-PIPO mainly affect the movement of the virus between cell-to-cell movement. P3N-PIPO binds to CI to recruit itself into plasmodesmata to promote intercellular movement of the virus [77,80]. Compared with SCMV and SrMV, a highly conserved motif of "Asp-Ala-Gly (DAG)" was absent in the CP sequence of SCSMV, which is necessary for aphid transmission [81,82]. It is worth mentioning that RNA silencing and RNA silencing repressors are mechanisms of defence and counter-defence interactions between host plants and viruses [83]. HC-Pro is the first strong RNA silencing inhibitor discovered [84], which has multiple targets in the RNA silencing pathway to regulate the accumulation of different siRNAs [85]. Moreover, when it fuses with P1, the expression of P1/HC-Pro and the inhibitory activity is enhanced [86]. The P1 protein of Poacevirus also has a silencing inhibitory function, which is more obvious than HC-Pro. However, when HC-Pro is present, the inhibitory activity of P1 on RNA silencing is reduced [87].

Genetic Diversity and Taxonomy
During evolution, SCMV, SrMV and SCSMV accumulated a rich pool of genetic diversity. Before 2000, Summers [106,107] and Summers et al. [11] divided SCMV into 10 strains and sub-strains according to the disease symptoms of sugarcane varieties CO281, CP29-291 and CP31-294. Tippett and Abbott [108] divided SCMV into five strains, namely, A, B, D, E and F, according to the mosaic symptoms of CP31-294, but two strains, A and H, according to CP31-588. According to serological cross-reaction, the evolutionary relationship and characteristics of CP nucleotide sequence based on the 3'-end sequence of the viral genome, Potyvirus isolates from Australia and the U.S. were divided into 11 SCMV strains, namely, MBD, A, B, D, E, SC, BC, Sabi, ISIS, Bris, and Bund, and three SrMV strains, namely H, I and M [15,43]. Since 2000, the genetic diversity of mosaic viruses isolates from China [109,110], South Africa [111], India [112,113], Mexico [114], Argentina [115] and Thailand [116,117], and others have also been reported based on the nucleotide and/or amino acid sequence variability of CP [81,118], HC-Pro [119] and P1 [79]. To sum up the results of genetic diversity, 216 relatively complete nucleotide sequences of the coat protein genes from102 SCMV isolates from 26 countries, 58 SrMV isolates from five countries, and 56 SCSMV isolates from 11 countries are downloaded from the NCBI database (accessed date 16 May 2021) and analysed using the maximum likelihood method (ML) of MEGA v6.0 software (Raynham MA, USA). The results are depicted in Figure 4.
Plant RNA viruses mutate in a variety of ways, including natural selection and substitution, transversion, deletion, insertion, recombination, reassortment, etc. [120,121]. Gell et al. [122] found that recombination was the main driving force for evolving SCMV subgroup populations. Recombination and strong selection pressure may accelerate the elimination process of deleterious mutations in the SCSMV genome P1 gene [79]. Strong purifying selection has been dominant in Indian SCMV populations, in which the CI and HC-Pro genes are prevalent [123]. He [64] reported that negative selection and genetic drift rather than recombination were the main driving force for the evolution of SrMV and SCSMV in China. In addition, natural selection, gene migration and geographical isolation may also affect the evolution of virus population in different regions [22,81,124].
protein genes from102 SCMV isolates from 26 countries, 58 SrMV isolates from five countries, and 56 SCSMV isolates from 11 countries are downloaded from the NCBI database (accessed date 16 May 2021) and analysed using the maximum likelihood method (ML) of MEGA v6.0 software (Raynham MA, USA). The results are depicted in Figure 4. Plant RNA viruses mutate in a variety of ways, including natural selection and substitution, transversion, deletion, insertion, recombination, reassortment, etc. [120,121]. Gell et al. [122] found that recombination was the main driving force for evolving SCMV subgroup populations. Recombination and strong selection pressure may accelerate the elimination process of deleterious mutations in the SCSMV genome P1 gene [79]. Strong purifying selection has been dominant in Indian SCMV populations, in which the CI and HC-Pro genes are prevalent [123]. He [64] reported that negative selection and genetic drift rather than recombination were the main driving force for the evolution of SrMV and SCSMV in China. In addition, natural selection, gene migration and geographical isolation may also affect the evolution of virus population in different regions [22,81,124].

Visual Observation
Mosaic disease can be identified by visual inspection of sugarcane leaves for symptoms when evaluating germplasm resistance [38,39,125,126]. The method is simple and timesaving. However, it cannot determine the virus types. In addition, it requires a higher level of professional knowledge, skill, and ability to catch sugarcane leaf mosaic symptoms with the naked eye.

Visual Observation
Mosaic disease can be identified by visual inspection of sugarcane leaves for symptoms when evaluating germplasm resistance [38,39,125,126]. The method is simple and timesaving. However, it cannot determine the virus types. In addition, it requires a higher level of professional knowledge, skill, and ability to catch sugarcane leaf mosaic symptoms with the naked eye.

Biological Identification
Traditionally, the type of virus, host range and transmission mode are determined mainly based on the symptoms of the host plant after inoculation with the virus [11,127]. Before 1940, Summers EM artificially inoculated leaves of indicator plants of varieties CP31-294 and Co281 with press juice friction from a symptomatic plant and made a judgment based on whether the leaves showed any mosaic symptoms [11,106,107]. However, it is hard to do, labour intensive, and difficult to determine the specific virus strain. The amount of needed plant material can fill up the greenhouses and strict test conditions are required.

Microscopic Observation
The most prominent pathological feature of plant tissue cells infected by a virus of the Potyviridae family is the production of columnar inclusions. The feature has been used to distinguish among virus species. Edwardson [128] divided cylindrical inclusions into four types: scrolls body and pinwheels (Type I); lamellar aggregates with arm extension and pinwheels (Type II); scrolls body, pinwheels, and lamellar aggregates (Type III); and scrolls body, pinwheels, and short and curved lamellar aggregates (Type IV). Studies have shown that the characteristics of columnar inclusions in SCMV-, SrMV- [129] and SCSMV-infected [53] sugarcane leaf cells were similar to III, I, and II, respectively. However, this method is seldomly used in practice due to the complicated operation procedure of electron microscopy.

Molecular Detection
PCR-based molecular techniques have been developed and widely used in the detection of the mosaic virus, such as direct-binding polymerase chain reaction (DB-PCR), reverse transcription PCR (RT-PCR), immunocapture RT-PCR (IC-RT-PCR), duplex immunocapture RT-PCR (D-IC-RT-PCR), one-step multiplex RT-PCR, real-time quantitative RT-PCR (qRT-PCR) and loop-mediated isothermal amplification (LAMP) [83]. These molecular detection techniques have a high sensitivity and specificity and can identify virus species rapidly and accurately. Yang and Mirkov [135] used DNA restriction enzymes to digest RT-PCR fragments to produce SCMV-or SrMV-specific RFLP patterns without the need of using identification host plants. Xie et al. [136] established a one-step quadruplex RT-PCR method to simultaneously detect four viruses, namely, SrMV, SCMV, SCSMV and Sugarcane yellow leaf virus (SCYLV). Smith and Velde [137] were able to detect SCMV by RT-PCR in samples that were diluted 10,000-fold. Fu et al. [52] developed a qRT-PCR method for the detection of SCSMV, which was 100 times more sensitive than conventional RT-PCR. Hema et al. [138] reported that IC-RT-PCR was 5,000 times more sensitive than DBIA and 10 times more sensitive than DB-PCR to detect SCSMV. Chen et al. [139] demonstrated that the sensitivity of detecting SrMV by IC-RT-PCR 100-times higher than that of indirect ELISA and 1000-times higher than that of Dot-ELISA and therefore recommended IC-RT-PCR was more suitable for large volumes of samples. Subba et al. [140] developed a D-IC-RT-PCR method that combined both serological and molecular methods to simultaneously detect and distinguish SCMV and SCSMV and was more sensitive than DAC-ELISA. In addition, a new reverse transcription-LAMP (RT-LAMP) technology was also applied to detect SCMV and SrMV with a lower sensitivity than RT-PCR and qRT-PCR [141]. The primer sequence information of different molecular techniques for detecting sugarcane mosaic viruses is shown in Table 3.

Exploitation and Utilization of Resistant Germplasm
Selection and rational distribution of disease-resistant varieties are the most economic and effective prevention and control measures against these viruses. The genetic base of modern sugarcane cultivars is narrow, with about 80% from S. officinarum, 10-15% from S. spontaneum, and 5-10% from recombinant chromosomes [142]. During the long evolutionary process, the Saccharum and its related genera have formed extremely rich and valuable germplasm resources [68,143], which contain large numbers of disease resistance genes [144]. In recent years, some Saccharum hybrid clones with high resistance to mosaic disease have been identified by natural infection or artificial inoculation, such as SP70-1143, IACSP95-2078 [145], SWSwm1 [146], YG34, YG55 [147], ROC16 [38], GT03-2309, and LC03-1137 [39]. The current artificial inoculation methods are shown in Table 4.

Acceleration of Molecular Breeding
Molecular marker-assisted breeding and genetic engineering improvements have helped promote the development of resistant varieties. Molecular marker-assisted breeding is an effective method to accelerate the breeding process of multi-resistant varieties. However, the technology is limited by the lack of markers closely linked to disease resistance. Sugarcane (2 n = 12 x = 100−130; genome size =~10 Gb) is a highly complex autopolyploid and aneuploid crop, and a complete reference genome of modern sugarcane cultivated species is still lacking up to now [154,155]. The development of molecular markers associated with economic target traits is an extremely slow process for sugarcane. Previous studies on the development of molecular markers for mosaic resistance were mainly focus on corn (2 n = 2 x = 20; genome size =~2300 Mb) [156][157][158][159][160][161][162][163][164]. However, these studies may provide a good reference for developing molecular markers and related gene mining on sugarcane. In addition, genetic engineering is an effective way to obtain disease-resistant sugarcane varieties [165]. The most mature strategy is CP gene-mediated transfer since the first report of the introduction of the Tobacco mosaic virus (TMV) CP gene into tobacco in 1986 [166]. Smith et al. [167] used a gene gun to bombard a SCMV-CP gene into sugarcane meristem and obtained chimeric transformed plants. Subsequently, Joyce et al. [168], Ingelbrecht et al. [169], and Sooknandan et al. [170] also succeeded in obtaining transgenic sugarcane plants by using the same method. On the other hand, an RNA interference technology that targets sugarcane virus encoded RNA silencing inhibitors has also been successfully used to develop highly effective disease-resistant transgenic sugarcane plants [171][172][173][174][175]. Field experiments on transgenic sugarcane have shown improved resistance to mosaic disease with significantly increased yield and sucrose content [171,[176][177][178]. However, none of the transgenic sugarcane has yet been applied in field production due to regulations.

Application of Virus-Free Plantlets
The application of virus-free plantlets through tissue culture can not only eliminate or slow down the incidence of mosaic disease, but also increase the sucrose content by more than 0.5% and the cane yield by 20-40% [179]. In one case, the yield increased by more than 100% [180]. The common methods of detoxification include heat treatment, stem buds tissue culture detoxification, and ultra-low temperature therapy. SCMV elimination can be achieved by cultivation at 52 • C, 57.3 • C, 57.3 • C, once every other day, 20 min each time, without harming cane buds [179]. SCMV can also be removed by cryo-treating micro shoot tips (~3 mm) [181] or by inserting 4~8 mm axillary bud explants into MS regeneration medium supplemented with a 25 mg/L of ribavirin [182]. Hot water treatment for 10 min at 55°C was not effective for SCSMV elimination [183]. Detoxification by stem buds tissue culture could completely eliminate SCSMV [184,185]. A combination of heat treatment and axillary shoot tip culture had a better effect on virus removal [186].
To determine the technical specifications for the production and testing of sugarcane virus-free plantlets, leaves were collected from the virus-free plantlets produced by the heat treatment and axillary shoot tip culture technology in 2015, followed by tracking and PCR testing the leaves of 120 putative virus-free plantlets collected from different companies at different ecological demonstration sites in China [187]. The PCR results showed that SCMV was not detected in 100% of the samples, the rate of SrMV detection was greater than 35%, and the rate of SCSMV detection was more than 50%. Therefore, it was concluded that SrMV and SCSMV were more difficult to be eliminated from infected sugarcane plants than SCMV. Although the virus-free effect of sugarcane stalk can be improved by appropriately increasing treatment temperature and time (30 min at 59 • C), the germination rate, however, was significantly reduced to only about 20% and the water temperature was difficult to control accurately in an industrial setting (unpublished).

Strengthen Cultivation and Control
Good field management is another effective way to enhance plant resistance and reduce the spread of viruses. Specific measures include: (1) avoiding planting virus host crops around or in the sugarcane fields [188]; (2) the timely removal of infected plants and weeds; (3) improvement in soil structure, rational fertilization, and irrigation to promote plant growth and improve disease resistance [189,190]; (4) chemical and biological control of aphids [63]; and (5) fortifying cropping systems and paying attention to rotation with non-host crops, such as soybeans, sweet potatoes and peanuts [191].

Perspective
The cultivation and planting of resistant varieties are the most economic and effective methods to control sugarcane infecting viruses, and growers are most likely to adopt this technology. However, due to the diversity of pathogens, the highly complex genome of sugarcane, the wide segregation of traits among hybrid progenies, and the extremely low probability of excellent gene aggregation, sugarcane cross-breeding may have to rely on a huge population, for example, 1 to 1.2 million of plantlets to be planted annually in China. Therefore, it is very difficult to select good varieties that have both commercial value and mosaic resistance. In addition, field evaluation tests of virus-free plantlets from integrated detoxification technology in China have shown that the "virus-free" effect is not ideal, even if the plantlets were produced by a more effective method of combining heat treatment and axillary bud or shoot tip culture. In view of all these facts, the development of molecular markers associated with disease resistance genes should be the most effective way to breed sugarcane varieties resistant to different viruses. Sugarcane propagates asexually and F 1 populations can be used for correlation analysis between genotypes and phenotypic traits. There are also successful cases on genetic engineering via gene gun bombardment or gene editing to improve disease resistance in sugarcane. Successfully edited alleles can be up to 49 copies simultaneously targeting the chlorophyll content gene, which draws a new blueprint for transforming the disease resistance pathway in sugarcane. We hope that the present review can provide scientific references and thoughts for the effective prevention and control of viruses in sugarcane.