Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

Histological description of vexans infection on tea leaves ()

Journal: Songklanakarin Journal of Science and Technology ManuscriptFor ID SJST-2017-0350.R2 Review Only Manuscript Type: Original Article

Date Submitted by the Author: 26-May-2018

Complete List of Authors: MOHKTAR, NORSYUHADA; SCHOOL OF BIOLOGICAL SCIENCES; Nagao, Hideyuki; Universiti Sains Malaysia, School of Biological Sciences

Keyword: leaf gall, cell differentiation, mesophyll, phloem, disease severity

For Proof Read only Page 1 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 2 3 Original Article 4 5 6 Histological description of infection on tea leaves 7 8 (Camellia sinensis) 9 10 Norsyuhada Mohktar1, Hideyuki Nagao1 11 12 1School of Biological Sciences, Universiti Sains Malaysia, Pulau Pinang, 11800, Malaysia 13 14 15 * Corresponding author, Email address: [email protected] 16 17 18 19 For Review Only 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 2 of 23

1 2 3 4 5 1 Abstract 6 7 8 2 Exobasidium vexans Massee is a parasitic that causes tea blister blight. Tea blister blight 9 10 3 is characterised by the swelling of the infected spots on the tea leaves. To date, there is limited 11 12 4 information on the cellular alterations that occur during E. vexans infection. The aim of this 13 14 5 study was to observe the histological changes on tea leaves infected by E. vexans. Diseased and 15 16 17 6 healthy leaves were thinsliced by cryostat microtome and examined by light microscopy. 18 19 7 Observations on the thin Forsections revealedReview the mode ofOnly hymenium development in E. vexans in 20 21 8 which the basidia protruded between the epidermal cells. The infection caused hypertrophy as 22 23 24 9 the size of the cells at the infected spots became double the size of the healthy cells. The first 25 26 10 description of venous infections of E. vexans were also described. The proliferation of the fungus 27 28 11 on the vein resulted in the formation of hymenium on both lower and upper sides of the leaves 29 30 31 12 and there was complete disruption of the vascular bundle in the leaves vein. Blister blight disease 32 33 13 on the tea plantation in this study was found to be moderately severe. Morphological and 34 35 14 molecular identification confirmed that the fungus isolated from the symptomatic leaves was E. 36 37 15 vexans. 38 39 40 16 Keywords: Disease severity, leaf galls, cell differentiation, mesophyll, phloem, plant parasite 41 42 43 44 17 1. Introduction 45 46 18 Tea has been one of the most consumed drinks in the world. The drink originates from 47 48 49 19 the leaves of tea plants (Camellia sinensis (L.) Kuntze) that are processed and dried for brewing. 50 51 20 Exobasidium vexans Massee is a pathogenic plant fungus that infects tea plants and causes tea 52 53 21 blister blight. The spread of this disease has brought tremendous loss to tea production industry 54 55 56 22 and it has become a serious threat to all tea plantations across Asia including India, Sri Lanka, 57 58 59 60 For Proof Read only Page 3 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 2 3 23 Japan and Indonesia (Baby, 2002). As the number one tea producer to the global market, India 4 5 6 24 has about 5 million hectares of land dedicated to tea plantation (Boriah, 2002). The country 7 8 25 which had contributed up to 23% of total world tea production suffered a great loss due to tea 9 10 26 blister blight as the pathogen attacks harvestable young shoots resulting in up to 40% yield losses 11 12 27 (Gulati, Gulati, Ravindranath, & Chakrabarty, 1993; Basu, Bera, & Rajan, 2010). In addition, 13 14 15 28 tea blister blight has inflicted a severe crop loss in Sri Lanka in which 33% of yield reduction 16 17 29 was recorded in unprotected areas compared to fields which were sprayed with chemicals (de 18 19 30 Silva, Murugiah, & Saravapavan,For Review 1997). Transgenic Only tea plant that is resistant to E. vexans 20 21 22 31 infection has been developed in which the clone expresses higher amount of chitinase enzyme to 23 24 32 hydrolyse the cell wall of the invading fungus (Jeyaramraja, Pius, Manian, & Nithya Meenakshi, 25 26 33 2005; Jayaswall, Mahajan, Singh, Parmar, Seth et al., 2016). 27 28 29 34 Symptoms of E. vexans infection include the swelling and formation of conspicuous 30 31 35 white hymenium on the infected spots that will eventually necrotise. Several studies have been 32 33 36 done on the histology of plants infected by other Exobasidium species. According to Mims and 34 35 37 Richardson (2007) and Lee, Lee, Shin, Park, and Sieber (2015), the infection of Exobasidium 36 37 38 38 gracile (Shirai) Syd, and P. Syd on Camellia sasanqua Thunb. was characterised by the 39 40 39 formation of a distinct pseudoparenchymatous layer of intercellular hyphae which initially 41 42 40 developed several layers above the lower epidermis. Eventually the lower epidermis layer 43 44 45 41 became sloughed off to give rise to the white hymenium on the underside of the leaf (Mims & 46 47 42 Richardson, 2007; Lee et al., 2015). Meanwhile, Li and Guo (2008) described a different mode 48 49 43 of hymenium development in the infection caused by Exobasidium ovalifoliae Z. Y. Li and L. 50 51 52 44 Guo in which the basidia protruded between epidermal cells to form continuous white hymenium 53 54 45 upon reaching the final stage of infection. Gadd and Loos (1949) reported that the formation of 55 56 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 4 of 23

1 2 3 46 hymenium in E. vexans progressed in the similar manner but they only provided an illustration to 4 5 6 47 support their observation. 7 8 48 At the cellular level, the enlargement of the infected spots caused by the infection of 9 10 49 Exobasidium species is either due to hypertrophy or hyperplasia – an increase in the number of 11 12 50 the cells. For example, Graafland (1960) reported that both hypertrophy and hyperplasia 13 14 15 51 occurred in the infection of Exobasidium japonicum Shirai on the leaves of Azaleas. Meanwhile, 16 17 52 the flowers of Lyonia ferruginea (Walter) Nutt. that were infected with Exobasidium ferrugineae 18 19 53 Minnis, A.H. Kennedy &For N.A. Goldberg Review only became Only hypertrophied but this observation was 20 21 22 54 less common in the leaves of the plant (Kennedy, Goldberg, & Minnis, 2012). 23 24 55 Several methods have been developed to measure the severity of fungal diseases on tea 25 26 56 plants. Webster and Park (1956) first evaluated the severity of E. vexans infection based on the 27 28 29 57 disease incidence from a pool of sampled leaves. Meanwhile Balasuriya (2003) assessed blister 30 31 58 blight infection according to the number of blistered lesion on tea leaves. Recently, Sinniah, 32 33 59 Wasantha Kumara, Karunajeewa and Ranatunga (2016) developed an improvised assessment 34 35 60 key to evaluate the severity of E. vexans infection on tea leaves that accounted for both host 36 37 38 61 reactions and the sequential stages of symptom development. 39 40 62 To date, there have been no detailed descriptions on the histological aspects of E. vexans 41 42 63 infection on C. sinensis except the illustration provided by Gadd and Loos (1949). Hence, the 43 44 45 64 aim of this study is to provide a deeper understanding on the histological changes that occurred 46 47 65 to C. sinensis leaf cells upon E. vexans infection. The mode of hymenium development and the 48 49 66 effect of E. vexans infection on the architectural state of the leaf cells were examined. In 50 51 52 67 addition, microscopic descriptions of venous infection were also presented. Disease severity of 53 54 68 E. vexans infection on the tea plantation in this study was assessed and the morphology of 55 56 57 58 59 60 For Proof Read only Page 5 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 2 3 69 basidia and basidiospores of this fungus was described. Finally, a phylogenetic tree was 4 5 6 70 constructed based on the internal transcribed spacer (ITS) gene to assess the position of E. 7 8 71 vexans relative to other Exobasidium species. 9 10 11 12 72 2. Materials and Methods 13 14 2.1 Histology of E. vexans infection 15 73 16 17 18 74 Healthy and infected shoots of tea leaves were collected from MARDI Agrotechnology 19 For Review Only 20 75 Park, Cameron Highlands, Malaysia in August 2016 and preserved in formalin acetic acid 21 22 76 solution. The vein and lamina of both healthy and infected spots were bounded to specimen 23 24 blocks by Jung Tissue Freezing Medium® (Leica Microsystems, Nussloch, Germany). The 25 77 26 27 78 samples were then transversely thinsliced by cryostat microtome (20°C) at 10µm thickness 28 29 79 (Leica CM 1850 UV, Nussloch, Germany). The thin sections were viewed under light 30 31 80 microscope (Olympus BX53, Tokyo, Japan) equipped with a camera (Olympus DP72, Tokyo, 32 33 34 81 Japan). The cell sizes of healthy and infected leaves were measured and statistically analyzed by 35 36 82 Mann WhitneyU Test. 37 38 39 83 2.2 Disease severity 40 41 42 84 Disease severity of E. vexans infection on tea plantation in MARDI Agrotechnology Park, 43 44 45 85 Cameron Highlands, Malaysia was assessed according to the assessment key developed by 46 47 86 Sinniah et al. (2016). Thirty tea shoots showing symptoms of E. vexans infection were collected 48 49 87 and the severity of infection on the third leaf of each shoot was rated according to the following 50 51 88 assessment key. A 0 to 6 point scoring system was described according to the following 52 53 54 89 conditions: (0) No translucent spots (TL) or no disease, (1) all TL show hypersensitive reaction, 55 56 90 (2) TL cover <2% leaf area, (3) Blisters and/ or necrosis covering <5% of total leaf area, or TL 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 6 of 23

1 2 3 91 covering 215% leaf area, (4) Blisters and/ or necrosis covering 515% of total leaf area or TL 4 5 6 92 covering >15% leaf area, (5) Blisters and/ or necrosis covering 1530% leaf area, and (6) Blisters 7 8 93 and/ or necrosis covering >30% leaf area or stem infections (Sinniah et al., 2016). 9 10 11 94 2.3 Morphological and molecular identification 12 13 14 95 Morphological observation of the basidia and basidiospores of E. vexans was done on fresh 15 16 96 specimens according to Nagao, Akimoto, Kishi, Ezuka and Kakishima (2003a). Hymenia formed 17 18 19 97 on blistered spots were scratchedFor andReview mounted in Shear’s Only solution on a glass slide. The basidia 20 21 98 and basidiospores were observed and measured under light microscope. 22 23 24 99 Molecular identification was done on E. vexans culture that was established according to the 25 26 100 method described by Nagao et al. (2003a). To simplify, the infected spot on a tea leaf was cut 27 28 2 2 29 101 into 2 mm pieces and stuck onto 10 mm agar block on the lid of a petri dish. The hymenium of 30 31 102 the infected spot was set to face the potato dextrose agar (PDA) surface. The petri dish was 32 33 103 incubated at 22°C in the dark, overnight to collect the released basidiospores. Then, the 34 35 36 104 basidiospores were allowed to germinate and E. vexans colony formed after 1 month of 37 38 105 culturing. DNA extraction was adapted from the method by Suyama et al. (1996). An extraction 39 40 106 solution composed of 0.01% SDS, 50mM KCl, 1.5mM MgCl2, 0.01% proteinase K, and 10mM 41 42 43 107 TrisHCL (pH 8.3 at 20 °C) was made. Small clumps of E. vexans colony were added to 20 µL 44 45 108 of the extraction solution and incubated for 60 min at 37°C, followed by heating at 95°C for 10 46 47 109 min. After the incubation process, the extraction solution that contained the extracted DNA was 48 49 110 used as the template for downstream polymerase chain reaction (PCR). 50 51 52 53 111 Two genes were amplified in this experiment. The large subunit (LSU) of the nuclear 54 55 112 ribosomal RNA gene was amplified using the NL1 (5’ GCA TAT CAA TAA GCG GAG GAA 56 57 58 59 60 For Proof Read only Page 7 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 2 3 113 AAG 3’) and NL4 (5’ GGT CCG TGT TTC AAG ACG G3’) primers (O’Donnell 1992, 4 5 6 114 1993). Amplification was done according to the following cycle: Initial denaturation at 94 °C for 7 8 115 2 min; 35 cycles of denaturation at 94 °C for 30 s, primer annealing at 60 °C for 1 min and 9 10 116 extension at 72 °C for 90 s; and a final extension at 72 °C for 10 min. Meanwhile, the ITS gene 11 12 117 was amplified using the ITS1F (5’ CTT GGT CAT TTA GAG GAA GTA A3’) and ITS4 (5’ 13 14 15 118 TCC TCC GCT TAT TGA TAT GC 3’) primers according to the following cycle: Initial 16 17 119 denaturation at 95 °C for 3 min; 35 cycles of denaturation at 95 °C for 30 s, primer annealing at 18 19 120 55 °C for 1 min and extensionFor at Review72 °C for 1 min; andOnly a final extension at 72 °C for 10 min 20 21 22 121 (White, Bruns, Lee & Taylor, 1990; Gardes & Bruns, 1993; Virtudazo, Nakamura & Kakishima, 23 24 122 2001). The PCR mixture to amplify the two genes consisted of 20 mM of each dNTP, 1.5 mM 25 26 123 MgCl2, 4 µL of extracted DNA, and 0.2 µM of each ITS1F – ITS4 and NL1 – NL4 primers. 27 28 29 124 PCR products were sent to a service provider for sequencing. 30 31 32 125 Phylogenetic tree was constructed using the ITS sequence of E. vexans obtained from this 33 34 126 study and other ITS sequences of Exobasidium species available on Genbank (Table 1). The 35 36 127 sequences were aligned and Neighbour Joining tree was constructed using the software MEGA 7 37 38 39 128 (Kumar, Stecher, & Tamura 2016). The stability of clades was evaluated by generating 1000 40 41 129 bootstrap replicates and distance analyses were computed using the Maximum Composite 42 43 130 Likelihood method. Gaps and missing data were completely deleted from the dataset. 44 45 46 131 3. Results 47 48 49 132 3.1 Histology of E. vexans infection on tea leaves 50 51 52 133 Leaves with mature E. vexans infection appeared swollen and the presence of whitevelvety 53 54 134 hymenium was observed in the lower epidermis. Meanwhile, if the infection occurred on the 55 56 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 8 of 23

1 2 3 135 veins, the hymenium could be observed on both lower and upper epidermis (Figure 1A and 1B). 4 5 6 136 The leaf lamina that was close to the infected veins also had the presence of the hymenium in 7 8 137 both lower and upper epidermis. Infection by E. vexans was normally localised and eventually 9 10 138 caused the necrosis of the infected spots. 11 12 13 139 In the case of the hymenium developing on the underside of the leaf (Figure 1B), the 14 15 16 140 structure of the lower epidermis was completely disrupted and filled with masses of slender 17 18 141 hyphae and basidia (Figure 2B). The air cavities that were normally present in the healthy leaf 19 For Review Only 20 142 were filled with networks of intercellular hyphae in the infected leaf (Figure 2B). As the 21 22 infection reaching its final stage, intracellular hyphae could also be observed in the infected leaf 23 143 24 25 144 cells. Mann Whitney U test revealed that the size of the cells in the spongy mesophyll layer was 26 27 145 significantly greater in the infected leaf (Mdn = 38.56 µm) than the healthy leaf (Mdn = 19.23 28 29 146 µm), U = 510, p <0.05. However, the structure of the upper epidermis and the palisade 30 31 32 147 mesophyll of the infected leaf (Figure 2B) appeared similar to the ones in the healthy leaf (Figure 33 34 148 2A). 35 36 37 149 Infection on the vein of the leaf resulted in the proliferation of hymenium on both lower and 38 39 150 upper side of the leaf (Figure 1A and 1B). In this case, both lower and upper epidermis of the 40 41 42 151 leaf were completely absent and filled with exposed, slender hyphae which developed to become 43 44 152 basidia that bore basidiospores (Figure 3A and 3B). The sclerenchyma layer in the infected vein 45 46 153 (Figure 4B) was completely ruptured. Meanwhile, the normally organised and intact structure of 47 48 49 154 the xylem and phloem became disrupted and the intercellular spaces were tightly packed with the 50 51 155 proliferation of intercellular hyphae (Figure 4A and 4B). 52 53 54 156 3.2 Disease severity 55 56 57 58 59 60 For Proof Read only Page 9 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 2 3 157 Eighty percent of the sampled leaves in this study showed apparent symptoms of E. vexans 4 5 6 158 infection which scored ≥ 3 points according to the assessment key by Sinniah et al. (2016). 7 8 159 Meanwhile, 16.7% of the samples scored 6 points as the necrotic spots were observed to cover 9 10 160 >30% of the leaf area (Table 2). The overall disease severity index (DSI) for tea plantation in this 11 12 161 study was 2.07 according to the equation by Sinniah et al. (2016) 13 14 15 16 162 3.3 Morphological identification 17 18 19 163 The hymenium of E. vexansFor consisted Review of clavate to Onlycylindrical basidia (46.9886.42 µm × 45 20 21 164 µm) that generally had 2 sterigmata. Basidiospores (12.8216.07 µm × 3.33.8 µm) were hyaline, 22 23 165 aseptate and ellipsoid. Upon maturation, basidiospores can develop a single septation. 24 25 26 3.4 Molecular identification and phylogenetic tree 27 166 28 29 30 167 Amplification of the LSU and ITS genes of E. vexans yielded PCR products of size 556 31 32 168 bp and 586 bp respectively. The LSU and ITS sequences were deposited to Genbank database 33 34 169 with accession number MG827275 and MG827276 respectively. BLAST search for the LSU 35 36 sequence showed 99% similarity with several E.vexans sequences (AB180380.1, AB262789.1). 37 170 38 39 171 Meanwhile the ITS sequence matched 100% (KC493154.1) and 99% (KY038487.1) with 40 41 172 Exobasidium reticulatum. However, the ITS sequence obtained from this study did not match 42 43 173 with an E. vexans ITS sequence that was readily available in the database (KC442794.1). 44 45 46 47 174 Neighbour Joining tree (NJ) in Figure 5 was constructed based on the ITS sequence 48 49 175 analysis. The NJ tree of Exobasidium species was divided into four clades according to specific 50 51 176 host groups. E. vexans isolated from this study belonged in the Camellia clade together with E. 52 53 177 reticulatum S Ito & Sawada, E. gracile and Exobasidium camelliae Shirai. The separation of this 54 55 56 178 clade from the other clades (Rhododendron, Vaccinium and Symploci) was supported by a high 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 10 of 23

1 2 3 179 NJ bootstrap value of 84%. Nonetheless, the ‘E. vexans KC442794.1’ sequence did not belong in 4 5 6 180 the Camellia clade, instead it was found to be clustered with the outgroup of this NJ tree 7 8 181 (Tilletiopsis washingtonensis). 9 10 11 182 4. Discussion 12 13 14 183 Observations on the mode of hymenium development during E. vexans infection on C. 15 16 184 sinensis were coherent with the illustration provided by Gadd and Loos (1949) in which basidia 17 18 19 185 protruded through the intercellularFor Review spaces of the lower Only epidermis. This finding is similar to the 20 21 186 mode of hymenium development in the majority of other Exobasidium species including E. 22 23 187 ovalifoliae and Exobasidium tengchongense Z.Y. Li and L. Guo (Li & Guo, 2008). Meanwhile, 24 25 26 188 Wolf and Wolf (1952) reported that the hymenium of E. gracile developed differently than the 27 28 189 rest of Exobasidium species including E. vexans in which the hymenium layer first formed 29 30 190 beneath the lower epidermis and became exposed when the epidermal layer were sloughed off. 31 32 33 191 This observation was further confirmed by Mims and Richardson (2007) and Lee et al. (2015). 34 35 36 192 In the case of hymenium formation only on the lower side of the leaf, the morphology of 37 38 193 the palisade layer remained cylindrical indicating that the infection did not disrupt the cellular 39 40 194 differentiation. Similarly, Li and Guo (2009) also reported the clear differentiation of palisade 41 42 43 195 and mesophyll cells in the leaf of C. sinensis infected by Exobasidium yunnanense Z.Y. Li and 44 45 196 L. Guo. This observation was different from the infection of E. gracile and E. ovalifoliae on C. 46 47 197 sasanqua and Lyonia ovalifolia var. elliptica (Siebold & Zuccarini) HandelMazzetti 48 49 198 respectively, in which the palisade layers in the diseased leaves were not able to differentiate 50 51 52 199 properly (Mims & Richardson, 2007; Li & Guo, 2008). 53 54 55 56 57 58 59 60 For Proof Read only Page 11 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 2 3 200 In addition, the prominent hypertrophied condition primarily in the spongy mesophyll 4 5 6 201 layer of E. vexans infected leaf was similar to the infection of several other Exobasidium species 7 8 202 on their respective hosts (Gómez P & KisimovaHorovitz, 1998; Chlebicki & Chlebická, 2007; 9 10 203 Kennedy et al., 2012). However, the effect of Exobasidium infection was not only limited to 11 12 204 hypertrophy as Akai (1938) reported that the infection of E. camelliae on Camellia japonica L. 13 14 15 205 resulted in both hypertrophy and hyperplasia of the leaf cells. Several researchers also reported 16 17 206 the occurrence of both hypertrophy and hyperplasia in the infection of other Exobasidium species 18 19 207 on the leaf (Graafland, 1960;For Li & Review Guo, 2008; Lee et Onlyal., 2015) and different plant organs such 20 21 22 208 as flower (Kennedy et al., 2012) and fruit (Nagao, Ogawa, Sato, & Kakishima, 2003b). 23 24 25 209 Gadd and Loos (1949) described that E. vexans could invade C. sinensis by forming 26 27 210 appressoria through either lower or upper sides of the leaf but eventually, the white hymenium 28 29 211 still proliferated on the lower side of the leaf regardless of the entry side. In contrast, the 30 31 32 212 observation in this study showed that the infection of E. vexans could result in the formation of 33 34 213 hymenium on both the lower and upper sides of the leaf primarily on the leaf midrib and vein. 35 36 214 Venous infection of E. vexans could relatively enhance the spread of the infection to other 37 38 39 215 tissues. The sclerenchyma and phloem components of the vascular bundle in the infected vein 40 41 216 were completely collapsed and filled with intracellular hyphae which were similar to the 42 43 217 histological appearance of strawberry petiole infected with Colletotrichum fragariae 44 45 (Milholland, 1982). Meanwhile, the cellular structure of the xylem became disorganised and the 46 218 47 48 219 intercellular spaces were filled with E. vexans hyphae. Based on the descriptions of the different 49 50 220 symptoms of Exobasidium infection provided by Nannfeldt (1981), E. vexans infection could be 51 52 221 categorised as circumscribed and monocarpic in which the fungus usually shows localised 53 54 55 222 infection spots and dies after sporulation. 56 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 12 of 23

1 2 3 223 Assessment of the disease severity on the tea plantation in this study showed that E. 4 5 6 224 vexans infection almost consistently ends with the necrosis of the infected spots or stems as 7 8 225 shown by the high percentage of scores ≥ 3. Some of the resistant tea strains infected by E. 9 10 226 vexans would show hypersensitive reactions or smaller blisters with less sporulation (Sinniah et 11 12 227 al., 2016). Histological observation in this study also supported this result as E. vexans infection 13 14 15 228 on tea leaves led to the formation of hymenium to allow for sporulation and eventually, the 16 17 229 symptomatic spots would necrotise. Susceptibility of the specific tea strain towards blister blight 18 19 230 infection plays an importantFor role in Review the scoring of the diseaseOnly severity. The overall DSI score for 20 21 22 231 tea plantation in this study was 2.07, in which it could be considered as moderately resistant as 23 24 232 compared to a highly resistant tea strain that was reported to have a DSI score of 0.6 (Sinniah et 25 26 233 al., 2016). However, it is crucial to note that the weather condition during the sampling period 27 28 29 234 also plays a huge role in the disease development, thus the overall DSI score might increase as 30 31 235 the weather becomes more favorable for E. vexans proliferation. For example, the monsoon 32 33 236 season between October to March in Malaysia could worsen the disease as a combination of high 34 35 237 relative humidity and less sunshine will result in severe blister blight infection (Homburg, 1955; 36 37 38 238 Reitsma & Van Emden, 1950). 39 40 41 239 Extensive studies on the phylogenetics of Exobasidium species have been done by using 42 43 240 different genes including ITS and LSU (Blanz & Doring, 1995; Begerow, Bauer & Oberwinkler, 44 45 2002; Piatek, Lutz & Welton, 2012). Nonetheless, E. vexans has never been included in any of 46 241 47 48 242 the analyses involving ITS gene as its ITS sequence is not available in Genbank except for 49 50 243 KC442794.1. Based on the ITS sequence obtained in this current study, E. vexans is clustered 51 52 244 with other Exobasidium species that are parasitic to their respective hosts belonging in the 53 54 55 245 Camellia clade. This result is coherent with the analysis by Blanz and Doring (1995) in which 56 57 58 59 60 For Proof Read only Page 13 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 2 3 246 Exobasidium species that are highly host specific cospeciate with their respective hosts. E. 4 5 6 247 vexans was found to be in the same subclade with E. reticulatum as these fungi are the 7 8 248 pathogens of tea plants. Nonetheless, E. reticulatum infection on tea plant causes tea net blight; a 9 10 249 disease with distinctive symptoms than tea blister blight. The surface of the infected tea leaves 11 12 250 appeared to be covered by networks of hymenium, but the infected spots are not swollen which 13 14 15 251 is the prominent characteristics of E. vexans infection instead (Chen and Chen, 1982). In 16 17 252 addition, the symptoms of tea net blight by E. reticulatum has never been observed in tea 18 19 253 plantations in Malaysia. OnlyFor ITS Reviewgene was used to construct Only the Neighbour Joining tree in this 20 21 22 254 study as the LSU sequence was found to not be able to resolve the phylogenetic relationship of 23 24 255 Exobasidium species as reported by Piatek et al. (2012). 25 26 27 256 Interestingly, the E. vexans ITS sequence KC442794.1 that is available in the Genbank 28 29 257 database was clustered with the outgroup of the Neighbour Joining tree (T. washingtonensis). 30 31 32 258 Further analysis of this sequence revealed that it was 99% similar to Tilletiopsis pallascens with 33 34 259 100% query coverage. Hence, it is important to point out that the KC442794.1 sequence could be 35 36 260 contaminated by cooccuring Tilletiopsis species that are common plant endophytes. The 37 38 39 261 contamination of Tilletiopsis is common when a voucher specimen is used in the DNA extraction 40 41 262 of Exobasidium species as reported by several other researchers (Begerow et al., 2002; 42 43 263 Boekhout, 1995; Peatek et al., 2012). Thus, this current study eliminated the possibility of 44 45 contamination by isolating the basidiospore and culturing E. vexans on artificial media though 46 264 47 48 265 the culturing of Exobasidium species was known to be difficult (Begerow et al., 2002). In 49 50 266 addition, the sequencing of two genes (LSU and ITS) and thorough description of infection 51 52 267 histology further confirmed that the fungus isolated in this study was indeed E. vexans. 53 54 55 56 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 14 of 23

1 2 3 268 In conclusion, the findings from this study confirmed and expanded the observation 4 5 6 269 provided by Gadd and Loos (1949). Several new observations were made including the 7 8 270 occurrence of venous infection of E. vexans on C. sinensis. Phylogenetic analysis of the ITS gene 9 10 271 revealed that E. vexans belonged in the same clade with other Exobasidium species whose host 11 12 272 plants are in the Camellia. For future works, it would be interesting to study whether the 13 14 15 273 infection on vascular bundles could affect the infection of E. vexans to other parts of the plant as 16 17 274 the infection on the petioles and internodes of C. sinensis shoots were common. 18 19 For Review Only 20 275 Acknowledgement 21 22 23 276 Authors would like to thank the USM 2017/2018 Fellowship Scheme for the support 24 25 26 277 provided throughout the course of the study. 27 28 29 30 278 References 31 32 279 Akai, S. (1938). Studies on the pathological anatomy of the hypertrophied buds of Camellia 33 280 japonica caused by Exobasidium camilliae. The Botanical Megazine 53, 118124. 34 35 281 Baby, U. I. (2002). An overview of blister blight disease of tea and its control. Journal of 36 282 Plantation Crops 30, 112. 37 283 Balasuriya, A. (2003). Seventy five years of research and development in plant Pathology. In: 38 284 Modder,W.W.D. (Ed.), Twentieth Century Tea Research in Sri Lanka. (pp. 165203). 39 285 Talawakelle, Sri Lanka: Tea Research Institute of Sri Lanka. 40 286 Basu, M. A., Bera, B., & Rajan, A. (2010). Tea statistics: global scenario. Inc. J. Tea Sci 8, 121 41 287 124. 42 43 288 Begerow, D., Bauer, R., & Oberwinkler, F. (2002). The : an evolutionary 44 289 hypothesis. Mycological Progress 1(2), 187199. 45 290 Blanz, P., & Doring, H. (1995). Taxonomic relationships in the genus Exobasidium 46 291 (Basidiomycetes) based on ribosomal DNA analysis. Studies in Mycology 38, 119127. 47 292 Boekhout, T. (1995). Molecular systematics of some yeastlike anamorphs belonging to the 48 293 Ustilaginales and Tilletiales. Studies in Mycology 38, 175183. 49 294 Boriah, G. (2002). Tea, India: world's largest consumer. The Hindu Survey of Indian Agricuture, 50 51 295 125128. 52 296 Chen, T. M., & Chen, S. F. (1982). Diseases of Tea and Their Control in the People's Republic 53 297 of China. Plant Disease 66(10), 961965. 54 298 Chlebicki, A., & Chlebická, M. (2007). A new fungus, Exobasidium gomezi, from Chilean 55 299 Patagonia. Nova Hedwigia 85, 145149. 56 57 58 59 60 For Proof Read only Page 15 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 2 3 300 de Silva, R. L., Murugiah, S., & Saravapavan, T. V. (1997). Losses of tea crops caused by 4 301 Exobasidium vexans. Tea Quart 43, 140146. 5 6 302 Gadd, C. H., & Loos, C. A. (1949). The fungus Exobasidium Vexans. Tea Quart 20, 5461. 7 303 Gardes, M., & Bruns, T. D. (1993). ITS primers with enhanced specificity for basidiomycetes – 8 304 application to the identification of mycorrhizae and rusts. Molecular Ecology 2, 113–118. 9 305 Gómez P, L. D., & KisimovaHorovitz, L. (1998). Basidiomycetes of Costa Rica. New 10 306 Exobasidium species () and records of Cryptobasidiales. Revista de 11 307 Biología Tropical 46, 10811093. 12 308 Graafland, W. (1960). The parasitism of Exobasidium japonicum Shir. on Azalea. Acta Botanica 13 14 309 Neerlandica 9, 347379. 15 310 Gulati, A., Gulati, A., Ravindranath, S. D., & Chakrabarty, D. N. (1993). Economic yield losses 16 311 caused by Exobasidium vexans in tea plantations. Ind Phytopathol 46, 155159. 17 312 Homburg, K. (1955). Enige resultatan van de blister blight bestrijding proeven in het Patuhase. 18 313 Bergcultures 24,169185. 19 314 Jayaswall, K., Mahajan, P.,For Singh, ReviewG., Parmar, R., Seth, Only R., Raina, A., . . . Sharma, R. K. (2016). 20 21 315 Transcriptome Analysis Reveals Candidate Genes involved in Blister Blight defense in 22 316 Tea (Camellia sinensis (L) Kuntze). Scientific Reports 6, 30412. 23 317 Jeyaramraja, P. R., Pius, P. K., Manian, S., & Nithya Meenakshi, S. (2005). Certain factors 24 318 associated with blister blight resistance in Camellia sinensis (L.) O. Kuntze. 25 319 Physiological and Molecular Plant Pathology 67, 291295. 26 320 Kennedy, A. H., Goldberg, N. A., & Minnis, A. M. (2012). Exobasidium ferrugineae sp. nov., 27 321 associated with hypertrophied flowers of Lyonia ferruginea in the southeastern USA. 28 29 322 Mycotaxon 120, 451460. 30 323 Kumar, S., Stecher, G., & Tamura, K. (2016). MEGA7: molecular evolutionary genetics analysis 31 324 version 7.0 for bigger datasets. Molecular Biology and Evolution 33(7), 1870–1874. 32 325 Lee, C. K., Lee, S. H., Shin, H. D., Park, J. H., & Sieber, T. (2015). First report of Exobasidium 33 326 gracile causing hypertrophied leaves of Camellia sasanqua in South Korea. Forest 34 327 Pathology 45, 258261. 35 328 Li, Z., & Guo, L. (2008). Two new species of Exobasidium (Exobasidiales) from China. 36 37 329 Mycotaxon 104, 331336. 38 330 Li, Z., & Guo, L. (2009). Two new species and a new Chinese record of Exobasidium. 39 331 Mycotaxon 108, 479484. 40 332 Milholland, R. D. (1982). Histopathology of strawberry infected with Colletotrichum fragariae. 41 333 Cytology and Histology 72, 14341439. 42 334 Mims, C. W., & Richardson, E. A. (2007). Light and electron microscopc observation of the 43 44 335 infection of Camellia sasanqua by the fungus Exobasidium camilliae var. gracilis. Can. 45 336 J. Bot. 85, 175183. 46 337 Nagao, H., Akimoto, M., Kishi, K., Ezuka, A., & Kakishima, M. (2003a). Exobasidium dubium 47 338 and E miyabei sp. nov causing Exobasidium leaf blisters on Rhododendron spp. in Japan. 48 339 Mycoscience 44, 19. 49 340 Nagao, H., Ogawa, S., Sato, T., & Kakishima, M. (2003b). Exobasidium symploci-japonicae 50 341 var. carpogenum var. nov. causing Exobasidium fruit deformation on Symplocos lucida 51 52 342 in Japan. Mycoscience 44, 369375. 53 343 Nannfeldt, J. A. (1981). Exobasidium, a taxonomic reassessment applied to the European 54 344 species. Symb. Bot. Ups 23, 172. 55 56 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 16 of 23

1 2 3 345 O’Donnell, K., L. (1992). Ribosomal DNA internal transcribed spacers are highly divergent in 4 346 the phytopathogenic ascomycete Fusarium sambucinum (Gibberella pulicaris). Current 5 6 347 Genetics 22, 213–220. 7 348 O’Donnell, K., L. (1993). Fusarium and its near relatives. In D. R. Reynolds, J. W. Taylor 8 349 (Eds.), The fungal holomorph: mitotic, meiotic and pleomorphic speciation in fungal 9 350 systematics (pp. 225233). Wallingford: CAB International. 10 351 Reitsma, J., & Van Emden, J. H. (1949). De bladpokkenziekte van de thee I. Bergcultures 12, 11 352 218. 12 353 Piątek, M., Lutz, M., & Welton, P. (2012). Exobasidium darwinii, a new Hawaiian species 13 14 354 infecting endemic Vaccinium reticulatum in Haleakala National Park. Mycological 15 355 progress 11(2), 361371. 16 356 Sinniah, G. D., Wasantha Kumara, K. L., Karunajeewa, D. G. N. P., & Ranatunga, M. A. B. 17 357 (2016). Development of an assessment key and techniques for field screening of tea 18 358 (Camellia sinensis L.) cultivars for resistance to blister blight. Crop Protection 79, 143 19 359 149. For Review Only 20 21 360 Suyama, Y., Kawamuro, K., Kinoshita, I., Yoshimura, K., Tsumura, Y., & Takahara, H. (1996). 22 361 DNA sequence from a fossil pollen of Abies spp. from Pleistocene peat. Genes & Genetic 23 362 Systems 71, 145149. 24 363 Virtudazo, E. V., Nakamura, H., & Kakishima, M. (2001). Phylogenetic analysis of sugarcane 25 364 rusts based on sequences of ITS, 5.8 S rDNA and D1/D2 regions of LSU rDNA. Journal 26 365 of General Plant Pathology 67(1), 2836. 27 366 Webster, B.N., & Park, P.O. (1956). Developments in blister blight control. I. Introduction to the 28 29 367 1955 series of blister blight control experiments. Tea Quart 27, 36. 30 368 White, T. J., Bruns, T., Lee, S., & Taylor, J. (1990). Amplification and direct sequencing of 31 369 fungal ribosomal RNA genes for phylogenetics. PCR protocols: a guide to methods and 32 370 applications, 18(1), 315322. 33 371 Wolf, F. T., & Wolf, F. A. (1952). Pathology of Camellia leaves infected by Exobasidium 34 372 camilliae var. gracilis Shirai. Phytopahthology 42, 147149. 35 373 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For Proof Read only Page 17 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Review Only 20 21 22 23 24 25 26 27 28 29 30 31 1 32 2 33 34 35 3 Figure 1. . vexans infection on C. sinensis leaf. A. Upper side of the leaf with white and 36 37 4 velvety hymenium (arrow) formed on the vein. B. Lower side of the same leaf in (A) with 38 39 5 hymenium (arrow) formed on both of the vein and the leaf lamina. Bar = 5 mm 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 18 of 23

2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 1 19 2 Figure 2. Transverse Forsection of Review healthy and infected Only leaf of C. sinensis. A. Healthy leaf 20 21 3 with normally differentiated palisade (PM) and spongy mesophyll (SM) layer. Lower 22 23 4 epidermis (LE) and air cavities are present and intact. Bar = 20 µm. B. Infected leaf with 24 25 26 5 normally differentiated palisade mesophyll but hypertrophied spongy mesophyll cells that 27 28 6 were double the size of healthy spongy mesophyll. Lower epidermis were absent and 29 30 7 basidia were protruding between the cells (arrowhead). Bar = 50 µm. 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 8 47 9 Figure 3. Transverse section of healthy and infected vein of C. sinensis. A. Healthy vein 48 49 50 10 with intact and normally organized sclerenchyma (S), xylem (X) and phloem (P). B. 51 52 11 Infected vein with hymenium (arrowhead) proliferation on upper epidermis (UE) and lower 53 54 55 56 57 58 59 60 For Proof Read only Page 19 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

3 1 2 3 4 1 epidermis (LE). The structure of sclerenchyma, xylem and phloem were disrupted. Spongy 5 6 7 2 mesophyll (SM), palisade mesophyll (PM). Bar = 100 µm. 8 9 10 3 11 12 13 14 15 16 17 18 19 For Review Only 20 21 22 23 24 25 26 4 27 28 29 5 Figure 4. Enlarged images of healthy and infected vascular bundle of C. sinensis in Figure 30 31 6 3. A. Intact and normally organized sclerenchyma (S), xylem (X) and phloem (P). B. 32 33 34 7 Infected vascular bundle with collapsed sclerenchyma (S) and phloem (P) with intercellular 35 36 8 hyphae (arrow) present in the xylem (X). Bar = 20 µm. 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 20 of 23

4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 For Review Only 20 21 22 23 24 25 26 1 27 2 Figure 5. Neighbour Joining tree of E. vexans relative to other Exobasidium species 28 29 3 constructed based on the ITS sequences with Tilletiopsis washingtonensis as the outgroup. 30 31 32 4 Values shown at each node are obtained from 1000 bootstrap replicates, meanwhile the 33 34 5 evolutionary distances were computed using the Maximum Composite Likelihood method. 35 36 6 The sum of branch length is 1.0190 and a total of 420 nucleotides were used in the final 37 38 39 7 dataset. The letters represent the clustering of Exobasidium species according to the host 40 41 8 group. R = Rhododendron, V = Vaccinium, S = Symploci, C = Camellia 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For Proof Read only Page 21 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 2 3 4 5 Table 1. List of Exobasidium species used to construct the phylogenetic tree with their 6 respective hosts, isolate/ strain, locations, Genbank accession numbers and sources 7 8 Species Isolate/ Host Location Source Genbank 9 strain ITS 10 accession 11 number 12 E. vexans MC32016 C. sinensis Cameron This study MG827276 13 Highlands 14 Malaysia 15 16 E. vexans GZZCB C. sinensis China Genbank, KC442794.1 17 submitted by L. 18 Peng in 2013 19 E. gracile CBFor ReviewCamellia China Only Genbank, HQ398622.1 20 (Shirai) Syd. oleifera submitted by L. 21 & P. Syd Abel. Peng in 2010 22 23 E. camelliae MAFF Camellia Japan Genbank, AB180317.1 24 Shirai 238578 japonica L. submitted by H. 25 Nagao in 2004 26 27 E. IFO30393 C. sinensis Japan Nagao et. al in AB180369.1 28 reticulatum 2004 29 S. Ito & 30 Sawada 31 E. MAFF Rhododendr Japan Genbank, AB180375.1 32 nobeyamense 239439 on wadanum submitted by H. 33 Nagao & Makino Nagao in 2004 34 Ezuka 35 E. MAFF R. x Japan Genbank, AB180357.1 36 cylindrosporu 238663 mucronatum submitted by H. 37 m Ezuka Nagao in 2004 38 39 E. japonicum MAFF Rhododendr Japan Genbank, AB180326.1 40 Shirai 238591 on obtusum submitted by H. 41 var. Nagao in 2004 42 kaempferi 43 Planch. 44 E. MAFF Rhododendr Japan Genbank, AB180342.1 45 woronchinii 238610 on submitted by H. 46 Nagao brachycarpu Nagao in 2004 47 m D. Don ex 48 G. Don 49 E. otanianum MAFF Rhododendr Japan Genbank, AB180345.1 50 Ezuka 238613 on reticulatu submitted by H. 51 52 emend. m D.Don ex Nagao in 2004 53 Nagao G.Don 54 f. glabrescen 55 56 57 58 59 60 For Proof Read only Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR Page 22 of 23

1 2 3 4 5 s (Nakai & 6 H.Hara) 7 T.Yamaz. 8 E. MAFF Rhododendr Japan Genbank, AB180336.1 9 shiraianum 238602 on submitted by H. 10 Henn. degronianu Nagao in 2004 11 12 m Carrière 13 E. Voucher - England Genbank, EU784219.1 14 rhododendrii RBG Kew submitted by 15 (Fuckel) C.E. K(M)1024 P.M. Brock in 16 Cramer 94 2009 17 E. symploci MAFF Symplocos Japan Genbank, AB180351.1 18 japonicae 238620 lucida submitted by H. 19 Kusano & For Review(Thunb.) OnlyNagao in 2004 20 Tokubuchi Siebold & 21 var. Zucc. 22 carpogenum 23 24 Nagao & S. 25 Ogawa 26 27 E. symploci- MAFF Symplocos Japan Genbank, AB180678.1 28 japonicae 238811 lucida submitted by H. 29 Kusano & (Thunb.) Nagao in 2004 30 Tokubuchi Siebold & 31 Zucc. 32 E. kishianum MAFF Vaccinium Japan Genbank, AB180353.1 33 Nagao & 238623 hirtum submitted by H. 34 Ezuka Thunb. Nagao in 2004 35 var. 36 37 pubescens 38 (Koidz.) 39 T. Yamaz. 40 E. rostruspii CNJ2-3 Vaccinium USA Genbank, KR262425.1 41 Nannf. macrocarpo submitted by J. 42 n Ait. E. Stewart in 43 2015 44 E. bisporum IFO9942 Eubotryoide Japan Genbank, AB180364.1 45 Sawada ex s grayana submitted by H. 46 Ezuka (Maxim.) H. Nagao in 2004 47 Hara var. 48 49 glabra 50 (Komatsu) 51 H. Hara 52 E. vacinii MAFF Vaccinium Japan Genbank, AB180362.1 53 (Fuckel) 238668 vitis-idaea submitted by H. 54 Woronin L. Nagao in 2004 55 56 57 58 59 60 For Proof Read only Page 23 of 23 Songklanakarin Journal of Science and Technology SJST-2017-0350.R2 MOHKTAR

1 2 3 4 5 E. MAFF Vaccinium Japan Genbank, AB180347.1 6 inconspicuum 238616 hirtum submitted by H. 7 Nagao & Thunb. var. Nagao in 2004 8 Ezuka pubescens 9 (Koidz.) 10 T. Yamaz 11 12 Tilletiopsis PB 383 contaminant Germany Genbank, DQ025483.1 13 washingtonen 18S from an submitted by H. 14 sis Nyland Exobasidium Doering in 2005 15 infection on 16 Vaccinium 17 uliginosum 18 L. 19 For Review Only 20 21 22 23 Table 2. Number of symptomatic host leaves as rated according to the blister blight severity 24 assessment key by Sinniah et al. (2016). TL = Translucent spots 25 26 Disease severity score Number of leaves 27 0 (No TL or no disease) 1 28 1 29 (all TL show hypersensitive reaction) 2 30 2 (TL cover <2% leaf area) 4 31 3 (Blisters and/ or necrosis covering <5% of total leaf 7 32 area, or TL covering 2-15% leaf area) 33 4 (Blisters and/ or necrosis covering 5-15% of total leaf 4 34 area or TL covering >15% leaf area) 35 5 (Blisters and/ or necrosis covering 15-30% leaf area) 7 36 6 (Blisters and/ or necrosis covering >30% leaf area or 5 37 stem infections) 38

39 40

41 42

43 44

45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For Proof Read only