This is the submitted version of the journal article: Giordano, Andrea; Santo Domingo, Miguel; Quadrana, Leandro; [et al.]. «CRISPR». Journal of Experimental Botany, Vol. 73, Issue 12 (June 2022), p. 4022-4033. DOI 10.1093/jxb/erac148 This version is available at https://ddd.uab.cat/record/259589 under the terms of the license 1 2 CRISPR/Cas9 gene editing uncovers the role of CTR1 and ROS1 in melon 3 fruit ripening and epigenetic regulation 4 Andrea Giordano1*, Miguel Santo Domingo1, Leandro Quadrana3, 5 Marta Pujol1,2, Ana Montserrat Martín-Hernández1,2 and Jordi Garcia-Mas1,2* 1 6 Centre for Research in Agricultural Genomics (CRAG) CSIC-IRTA-UAB-UB, Edifici CRAG, Campus UAB, 7 2 08193 Bellaterra, Barcelona, Spain. 8 Institut de Recerca i Tecnologia Agroalimentàries (IRTA), Edifici CRAG, Campus UAB, 08193 Bellaterra, 9 3 Barcelona, Spain. 10 Institut de Biologie de l’Ecole Normale Supérieure, Paris, France. 11 12 Andrea Giordano: andrea.giordano@cragenomica.es 13 Miguel Santo Domingo: miguel.santodomingo@cragenomica.es 14 Leandro Quadrana: leandro.quadrana@bio.ens.psl.eu 15 Marta Pujol: marta.pujol@irta.cat 16 Ana Montserrat Martín-Hernández: montse.martin@irta.cat 17 Jordi Garcia-Mas: jordi.garcia@irta.cat 18 *Corresponding author: 19 Jordi Garcia-Mas, jordi.garcia@irta.cat, Phone: +34 935636600 20 Andrea Giordano, andrea.giordano@cragenomica.es, Phone: +34 935636600 21 22 Date of submission: 31/01/2022 23 24 Number of tables: 0 25 26 Number of figures: 4 27 28 Word count: 3542 29 30 Number of supplementary Tables: 6 31 32 Number of supplementary Figures: 5 33 34 35 Running title: Gene editing of two key players in melon ripening 36 37 38 39 40 41 42 43 44 45 1 46 Highlight 47 48 CmROS1 and CmCTR1-like CRISPR/Cas9 mutants accelerate fruit ripening in 49 melon. Besides, CmROS1 contributes to maintain the methylation levels along 50 fruit ripening by hypomethylation of ripening genes. 51 52 53 Abstract 54 55 Melon (Cucumis melo L.) has emerged as an alternative model to study fruit 56 ripening due to the coexistence of climacteric and non-climacteric varieties. 57 The previous characterization of a major QTL ETHQV8.1 sufficient to trigger 58 climacteric ripening in a non-climacteric background allowed the identification 59 within the QTL interval of a negative regulator of ripening CmCTR1-like 60 (MELO3C024518), and a putative DNA demethylase CmROS1 61 (MELO3C024516), the orthologue of DML2, a DNA demethylase regulating fruit 62 ripening in tomato. To understand the role of these genes in climacteric 63 ripening, we generated homozygous CRISPR knockout mutants of CmCTR1- 64 like and CmROS1 in a climacteric genetic background. The climacteric behavior 65 was altered in both loss-of-function mutants in two summer seasons with an 66 advanced ethylene production profile compared to the climacteric wild type, 67 suggesting a role of both genes in climacteric ripening in melon. Single cytosine 68 methylome analyses of the CmROS1 knockout mutant revealed DNA 69 methylation changes in the promoter regions of key ripening genes as ACS1, 70 ETR1 and ACO1, and ripening associated-transcription factors as NAC-NOR, 71 RIN and CNR, suggesting the importance of CmROS1-mediated DNA 72 demethylation for triggering fruit ripening in melon. 73 Keywords: fruit ripening, CRISPR, melon, cucurbits 74 Introduction 75 76 During the ripening process, fleshy fruits undergo physiological and metabolic 77 changes affecting color, flavor, firmness, and aroma. These changes are driven 78 by phytohormones and developmental factors and occur in a highly coordinated 79 manner with a direct impact on fruit quality and shelf-life (Giovannoni, 2001). 80 One of the main promoters of fruit ripening is the volatile hormone ethylene. 81 Depending on the involvement of this hormone during ripening, fruits have been 82 traditionally divided into i) climacteric, characterized by an increase in 2 83 respiration and ethylene production at the onset of ripening and ii) non 84 climacteric, presenting low levels of both ethylene production and respiration 85 rate across the process (McMurchie et al., 1972). Dissecting the regulatory 86 network underlying the control of fruit ripening has been a major goal due to its 87 biological significance but also for its commercial value (Giovannoni et al., 2017; 88 Wang et al., 2020). 89 Important advances in the understanding of the molecular mechanisms 90 underlying climacteric fruit ripening have been made in the model species 91 tomato (Giovannoni, 2007). Ripening related mutants allowed the identification 92 of several transcription factors that are upstream regulators of ethylene 93 dependent or independent ripening. Among them RIPENING INHIBITOR (RIN), 94 NON-RIPENING (NOR), and COLORLESS NON-RIPENING (CNR) (Vrebalov 95 et al., 2002; Manning et al., 2006; Giovannoni, 2007). 96 Recent studies demonstrated that DNA methylation levels play an important 97 role at the onset of fruit ripening in tomato (Zhong et al., 2013). Moreover, the 98 DNA methylation dynamics in a climacteric and an ethylene repressed line have 99 been recently studied in melon (Feder et al., 2020). Modulation of DNA 100 methylation levels is governed by DNA methylases and demethylases. The 101 enzymatic removal of methylcytosine in plants is initiated by a family of DNA 102 glycosylases/lyases, including DEMETER (DME), Repressor of silencing 1 103 (ROS1), DEMETER-like2 (DML2) and DEMETER-like3 (DML3), firstly 104 characterized in the model plant Arabidopsis thaliana (Zhu, 2009). In tomato, 105 SlDML2 is induced upon the onset of ripening leading to a global DNA 106 hypomethylation during ripening (Zhong et al., 2013; Liu et al., 2015; Lang et 107 al., 2017). Knockout using CRISPR/Cas9 system and knockdown RNAi mutants 108 in this species revealed that SlDML2 is required for normal fruit ripening by the 109 activation of ripening-induced genes and repression of several ripening- 110 repressed genes (Zhong et al., 2013; Lang et al., 2017). Nonetheless, the 111 tomato model is not universal as different transcriptional positive feedback 112 circuits controlling ripening in climacteric species were identified (Lü et al., 113 2018). 114 Melon (Cucumis melo L.) has emerged as an alternative model to study fruit 115 ripening since both climacteric (e.g. cantalupensis types as ‘Védrantais’ (VED)) 116 and non-climacteric (e.g. inodorus types as ‘Piel de Sapo’ (PS)) genotypes 3 117 exist. The recent characterization of a major QTL in chromosome 8 of melon, 118 ETHQV8.1, which is sufficient to activate climacteric ripening in a non- 119 climacteric background, allowed the identification of candidate genes related to 120 fruit ripening in a genomic interval of 150 kb that contained 14 annotated genes 121 (Pereira et al., 2020). Some of these genes are highly expressed in fruits and 122 contain multiple non-synonymous polymorphisms distinguishing the climacteric 123 VED from the non-climacteric PS genotype. 124 One of the candidates (CmROS1, MELO3C024516) encodes the homolog of 125 the main DNA demethylase ROS1 in Arabidopsis, which targets mainly 126 transposable element (TE) sequences and regulates some genes involved in 127 pathogen response and epidermal cell organization (Yamamuro et al., 2014; Le 128 et al., 2014). The closest orthologue in tomato, SlDML2 is crucial for the DNA 129 demethylation of fruit ripening genes including ethylene synthesis and signaling 130 (Lang et al., 2017). 131 The other candidate gene is CONSTITUTIVE TRIPLE RESPONSE 1 (CTR1) a 132 serine/threonine kinase (CmCTR1-like, MELO3C024518). This kinase interacts 133 physically with ethylene receptors as a negative regulator of the ethylene signal 134 transduction pathway (Kieber et al., 1993). In the absence of ethylene, CTR1 is 135 activated, preventing the downstream transduction pathway; when ethylene is 136 present, the ethylene receptor terminates the activation of CTR1, leading to the 137 ethylene responses (Binder, 2008). In tomato, the silencing of CTR1 promoted 138 fruit ripening, validating its role as a negative regulator of the ethylene signal 139 transduction pathway (Fu et al., 2005). 140 In this study, we aimed to better understand the role of the two ETHQV8.1- 141 containing candidate genes CmROS1 and CmCTR1-like in fruit ripening by 142 obtaining CRISPR/Cas9-induced loss-of-function mutants in a climacteric melon 143 genotype. Furthermore, we characterized the role of CmROS1 in DNA 144 methylation homeostasis during fruit ripening. 145 Materials and methods 146 CRISPR/Cas9 vector Construction 147 To target CmROS1 (MELO3C024516) two different guide RNAs (gRNA) of 20 148 nucleotides in length separated by 188 bp were designed using Breaking Cas 4 149 tool (https://bioinfogp.cnb.csic.es/tools/breakingcas/) (Table S1). The two 150 oligonucleotides generated for each gRNA were annealed and cloned in the 151 sites BbsI and BsaI into the plasmid p-tandemgRNA. The construct was verified 152 by sequencing and then digested with SpeI and KpnI to release the cassette 153 that was then inserted into the same sites in the pB7-Cas9-TPC-polylinker 154 binary vector. Cloning vectors were kindly provided by Prof. Puchta (KIT, 155 Germany). 156 For CmCTR1-like (MELO3C024518) we used the pEn-CHIMERA vector 157 provided by Prof. Puchta (KIT, Germany) to generate the entry construct. A 158 single gRNA of 20 nucleotides was designed using Breaking Cas tool 159 (https://bioinfogp.cnb.csic.es/tools/breakingcas/) (Table S1). Cloning steps of 160 the gRNA and transfer to the pDe-Cas9 binary vector were performed as 161 previously described (Schiml and Puchta, 2016). 162 Agrobacterium mediated plant transformation 163 164 Agrobacterium tumefaciens (strain AGL-0) cells were transformed with the 165 binary CRISPR/Cas9 constructs. Plant transformation was performed by co- 166 cultivation of the Agrobacterium culture with one-day-old cotyledons of VED as 167 previously described (Castelblanque et al., 2008), except that cotyledons were 168 dissected as in (García-Almodóvar et al., 2017). In brief, seeds were peeled and 169 incubated for one day in germination MS medium. Then, the embryo was 170 removed from the cotyledons and the half proximal part was incubated with the 171 Agrobacterium culture for 20 minutes in the presence of 200 μM 172 acetosyringone. After incubation, Agrobacterium was co-cultured with the 173 explants during three days at 28 ºC in the regeneration medium (Castelblanque 174 et al., 2008) supplemented with 0,5 mg/L 6-bencylaminopurine (BA), 0,1 mg/L 175 Indole-3-acetic acid (IAA) and 200 μM acetosyringone. Every three weeks, calli 176 were cleaned and the green buds were selected and replicated in fresh 177 regeneration media without acetosyringone and supplemented with L- 178 Phosphinothricin (PPT) for selection. Selected transgenic plants containing the 179 bar gene were grown in a growth room under a 12-h light/12-h dark cycle at 28 180 °C. After two to four months, individual transgenic plants were transferred to 181 rooting medium (regeneration medium without hormones). 5 182 183 Detection of mutations 184 Genomic DNA from leaves of in vitro plantlets (T0) and from young leaves of T1 185 and T2 plants was extracted using the CTAB method with some modifications 186 as described in (Pereira et al., 2018). The transgene presence was detected by 187 PCR using specific primers targeting Cas9. Genomic regions flanking gRNA1 188 and gRNA2 of CmROS1 were amplified by PCR using specific primers. For 189 detection of mutations in CmCTR1-like, a region targeting the gRNA was 190 amplified with specific primers. All primers are listed in Table S2. Mutations 191 were detected by sequencing the amplified fragments and identified by double 192 peaks in the sequence chromatograms. Purified PCR products were cloned into 193 p-Blunt II-TOPO vector (Life Technologies) and sequencing of colonies using 194 M13F and M13R primers was performed to confirm the mutations. 195 196 Generation of T2 plants and phenotyping of climacteric ripening traits 197 Ploidy level of T0 plants was evaluated by flow-cytometry analysis and selected 198 T0 plants for each gene were grown under greenhouse conditions (25ºC for 16 199 hours and 22ºC for 8 hours) and self-pollinated. T1 seedlings were screened for 200 the presence of Cas9 by PCR. After segregation, non-transgenic homozygous 201 edited T1 plants were selected and grown under greenhouse conditions to 202 obtain the T2 seeds for the phenotypic assay. 203 Edited T2 CmROS1 (n=8) and CmCTR1-like plants (n=8) were grown 204 randomized under greenhouse conditions (24ºC for 16 hours and 22ºC for 8 205 hours) at Caldes de Montbui (Barcelona) in 2020 and 2021. VED plants were 206 used as a wild type control plant (n=8). Plants were weekly pruned and 207 manually pollinated to obtain one fruit per plant. The harvest date was 208 determined following two criteria: either abscission date, when the fruit abscised 209 from the plant, or 5 days after the formation of the abscission layer when it was 210 not complete. 211 Ripening-related traits were evaluated as described in Pereira et al. 2020 in two 212 consecutive summer seasons (2020 and 2021). Production of aroma (ARO), 6 213 chlorophyll degradation (CD) and abscission layer formation in the pedicel of 214 the fruit (ABS) were daily evaluated and firmness was measured at harvest 215 time. The visual inspection of melon fruits, attached to the plant, was performed 216 daily, from approximately 20 days after pollination (DAP) until harvest. In 217 addition, individual pictures of the fruits were obtained weekly. ARO, ABS and 218 CD were recorded as 0 = absence and 1 = presence. The aroma production 219 was evaluated every day by smelling the fruits. The firmness of fruit flesh was 220 measured at harvest using a penetrometer (Fruit TestTM, Wagner Instruments), 221 in at least three regions of the fruit (distal, proximal and median), and the mean 222 value was registered. 223 Ethylene production 224 Ethylene production in planta was measured in the 2020 summer season using 225 non-invasive gas chromatography – mass spectrometry (GC-MS) method, as 226 described in (Pereira et al., 2017). The ethylene peak was monitored before 227 ripening from 20 DAP until harvest. The atmosphere of the chamber containing 228 the fruit was measured every day. 229 The ethylene peak was characterized by four traits, measured as described in 230 Pereira et al., 2020: maximum production of ethylene in the peak (ETH), 231 earliness of ethylene production (DAPE), earliness of the ethylene peak 232 (DAPP), and width of ethylene peak (WEP). 233 Epigenomics 234 DNA was extracted from fruit flesh of ROS1-CRISPR-2 and the wild-type VED 235 at different ripening stages (15, 25 and 30 DAP and harvest point) following the 236 CTAB protocol (Doyle and JJ, 1990) adding a purification step using 237 Phenol:Chloroform:Isoamyl alcohol (25:24:1). For each time point, three 238 biological replicates were analysed. Bisulfite conversion, BS-seq libraries and 239 sequencing (paired-end 100 nt reads) were performed by BGI Tech Solutions 240 (Hong Kong). Mapping was performed on melon genome v3.6.1 (Ruggieri et 241 al., 2018) using Bismark v0.14.2 (Krueger and Andrews, 2011) and the 242 parameters: --bowtie2, -N 1, -p 3 (alignment); --ignore 5 --ignore_r2 5 -- 243 ignore_3prime_r2 1 (methylation extractor). Only unique mapping reads were 7 244 retained. The methylKit package v0.9.4 (Akalin et al., 2012) was used to 245 calculate differential methylation in 100 bp non-overlapping windows (DMRs). 246 Significance of calculated differences was determined using Fisher’s exact test 247 and Benjamin-Hochberg (BH) adjustment of p-values (FDR<0.05) and 248 methylation difference cutoffs of 40% for CG, 20% for CHG and 20% for CHH. 249 Differentially methylated windows within 100 bp of each other were merged to 250 form larger DMRs. 100 bp windows with at least six cytosines covered by a 251 minimum of six (CG and CHG) and ten (CHH) reads per comparison were 252 considered. 253 254 Statistical analyses 255 All the statistical analyses and graphical representations were obtained using 256 the software R v3.2.3 (R Core Team, 2020) with the RStudio v1.0.143 interface 257 (RStudio: Integrated development environment for R, 2012). 258 259 Results 260 261 Generation of CRISPR/Cas9 knockout mutants in candidate genes for 262 ETHQV8.1 and inheritance of the editions 263 264 To investigate the role of CmROS1 (MELO3C024516) and CmCTR1-like 265 (MELO3C024518) genes in the fruit ripening process in melon, we knocked 266 them out using the CRISPR/Cas9 gene editing system in a climacteric genetic 267 background (VED). 268 A strategy with two target sites in exon 2 was used for CmROS1 (Fig. 1). We 269 obtained 15% transformation efficiency, recovering in total 59 transgenic rooted 270 plants. From the transgenic plants, almost half of them (46%) were edited. 271 Multiple independent transgenic plants were genotyped by sequencing the 272 genomic DNA spanning both target sites. Most of the editions (75%) occurred in 273 target 1 (gRNA1) whereas only a few editions (25%) were obtained for target 2 274 (gRNA2). Several different insertions and deletions were obtained in T0 plants 275 with biallelic or heterozygous mutations (Fig. S1), with several plants carrying 276 the same mutation (+1 bp). A diploid biallelic line with an insertion of 1 bp and a 277 deletion of 23 bp that were predicted to generate truncated proteins was 278 selected for further work (Fig. 1A). 8 279 The selected biallelic T0 line was self-pollinated to obtain non-transgenic (Cas9 280 free) plants carrying homozygous editions. After segregation, T1 lines 281 homozygous for the 1 bp insertion (ROS1-CRISPR-1) or the 23 bp deletion 282 (ROS1-CRISPR-2) were selected for further study (Fig. 1A and C). 283 A different CRISPR Cas9 strategy was used to target the CmCTR1-like gene. A 284 single target site was selected in exon 6 of CmCTR1-like (Fig. 1B). For this 285 target gene we obtained 12% of transformation efficiency. Transgenic T0 plants 286 were screened for mutations in the target site and 40% were edited showing 287 mainly large or small deletions (Fig. S1). From the edited T0 plants, a biallelic 288 line carrying a 11 bp deletion and a 1 bp insertion was selected and self- 289 pollinated to segregate out the Cas9 transgene. The genetic editions were 290 stably transmitted to T1 plants. After segregation, a homozygous edited line 291 carrying the 11 bp deletion (CTR1-CRISPR-1), which is predicted to generate a 292 premature termination codon, and the homozygous line with 1 bp insertion 293 (CTR1-CRISPR-2), generating a frame shift, were grown under greenhouse 294 conditions for the characterization of fruit ripening related traits (Fig. 1B and C). 295 296 CmROS1 and CmCTR1-like edited plants show altered ripening 297 phenotypes 298 ROS1-CRISPR-1/2 and CTR1-CRISPR-1/2 were evaluated and characterized 299 for ripening related traits in two consecutive summer seasons (2020 and 2021). 300 However, the line CTR1-CRISPR-2 was only characterized in 2021 due to a 301 powdery mildew infection of some replicates in 2020 that prevented its 302 evaluation. Overall, the fruit appearance (shape, weight and colour) of the 303 CRISPR edited lines did not show major differences with the wild-type VED at 304 harvest time and no significant changes were detected in the flesh firmness 305 (Fig. 1C, Table S3). To better characterize the ripening process, we measured 306 ethylene production in planta in 2020 with a non-invasive methodology allowing 307 observing the phenotype of the downstream effects of this hormone. 308 The phenotypic characterization revealed a significant earliness of the 309 climacteric symptoms for all the edited lines showing the same ripening 310 behavior pattern in both years (Fig. 2, Table S3). In 2020, the earliest 311 climacteric symptom was sweet aroma production (EARO), which in the 312 CRISPR edited lines for both genes appeared around two days before VED. 9 313 The initiation of the rind color change, which is attributed to chlorophyll 314 degradation (ECD), was appreciated almost simultaneously with the detection 315 of the abscission layer formation (EALF) and both ripening-related traits arose 316 in both CmROS1 edited lines two days before VED. The CTR1-CRISPR-1 317 edited line exhibited the earliness of the ripening related traits all at the same 318 time, which differed significantly from VED, arising around three days before 319 than VED for ECD and EALF and two days for EARO. 320 During the second summer season, we evaluated all the CRISPR edited lines. 321 In general, the environmental conditions delayed ripening of both VED and 322 mutant plants (around 4-5 days later in 2021). Despite this environmental effect, 323 all CRISPR edited lines displayed significant advances of about three days in 324 the ripening-related traits ECD, EARO and EALF when compared to VED (Fig. 325 2). Moreover, during this year, the line CTR1-CRISPR-2 was evaluated, and 326 the dataset showed the same behavior for both CTR1-like edited lines. ROS1 327 edited lines also showed the same pattern between them. 328 We also monitored fruit ethylene emission daily in 2020 without altering the 329 ripening process (Fig. 2C). The CRISPR edited lines showed a different 330 ethylene production pattern compared to wild-type VED, with both CmROS1 331 edited lines showing the same profile. In CmROS1 mutant lines, ethylene 332 production started two days before the wild-type VED and with an increment of 333 2.7 to 3-fold of ethylene production (Fig. 2C and Table S3). 334 For CmCTR1-like edited lines, ethylene measurements for CTR1-CRISPR-2 335 were not available due to the infection with powdery mildew of some of the 336 replicates of this line at around 20 DAP. The CTR1-CRISPR1 line showed a 337 significant difference in the earliness of ethylene production (DAPE) and 338 earliness of ethylene peak (DAPP). In this line, ethylene was detected around 339 three days in advance of wild-type VED. Similarly, the peak of ethylene 340 production was also advanced three days in CTR1-CRISPR1 compared to wild- 341 type VED. However, this advancement was not accompanied by a significant 342 difference in the maximum quantity of ethylene produced (Fig. 2C and Table 343 S3). Overall, these results demonstrate that both candidate genes are involved 344 in melon fruit ripening. 345 10 346 Characterization of the ROS1-CRISPR and VED methylome at different 347 fruit ripening stages 348 349 To better understand at the molecular level the role of CmROS1 in DNA 350 demethylation and fruit ripening in melon, we generated single-cytosine 351 resolution methylomes by whole genome bisulfite sequencing from fruits of 352 ROS1-CRISPR-2 and the wild-type VED plants at 15, 25 and 30 DAP as well as 353 at harvest (H) point (Fig. 3A). 354 When comparing the global methylation level along ripening in VED, we found 355 that methylation at CG and CHG contexts declines along fruit ripening, showing 356 around 2,000 and 4,000 hypomethylated regions (DMR), versus 300 and 3,000 357 hypermethylated regions in the CG and CHG context, respectively, at harvest 358 time compared to the first stage of ripening (i.e. H vs 15 DAP) (Fig. 3B). 359 Interestingly, these changes were more often associated with promoter and 360 intergenic regions (Fig. 3D). 361 In order to evaluate the role of CmROS1 in the observed DNA methylation 362 dynamics, we compared the methylation level in the three contexts of ROS1- 363 CRISPR-2 and the wild-type VED plants at the same ripening stage (Fig. 3C). In 364 this way, we identified numerous changes in DNA methylation levels for the 365 three sequence contexts. In total (CG, CHG and CHH context together), we 366 found 16,968 hypermethylated DMRs at 15 DAP, 26,497 at 25 DAP, 19,928 at 367 30 DAP and 43,156 at H time relative to VED, while the total hypomethylated 368 DMRs were 23,742 at 15 DAP, 36,813 at 25 DAP, 24,083 at 30 DAP and 369 33,698 at H time. Overall, CRISPR-ROS1 line is associated with 370 hypomethylation of CG and hypermethylation of CHG DMRs (Fig. 3C). 371 To further investigate the targets of ROS1 we focused on the hypermethylated 372 DMRs in the CRISPR-ROS1 line (Table S4). Moreover, in CHH context at H 373 time there are changes in the number of DMRs annotation between VED and 374 the edited line. Among the CHH hypermethylated regions in CRISPR-ROS1 375 compared to VED at H time, 14% are associated with TEs, 46% with intergenic 376 regions, 6% in promoter regions (defined as 1 kb upstream transcriptional start 377 sites), and 33% in genic regions (Fig. 3D). 378 Notably, Gene ontology (GO) enrichment analysis of genes associated with 379 hypermethylated DMRs at H time in the CRISPR-ROS line compared to VED 380 and hypomethylated along ripening in VED context revealed an 11 381 overrepresentation of genes related to response to stress in CRISPR-ROS1 382 compared to VED (Table S5). 383 CmROS1 targets promoter regions of key genes involved in ripening 384 We have further analysed the methylation level of key genes known to 385 participate in the ripening process in the three contexts. Changes were found at 386 different stages of ripening in the promoter region of genes involved in the 387 ethylene biosynthesis or signaling pathway: ACS1 (MELO3C016340.2), ETR1 388 (MELO3C003906.2) and ACO1 (MELO3C014437) as well as in ripening 389 associated-transcription factors: NAC-NOR (MELO3C016540), RIN 390 (MELO3C026300.2) and CNR (MELO3C002618.2) (Fig. 4). 391 Notably, the promoter region of ACS1 appeared hypomethylated on the three 392 sequence contexts in the CRISPR-ROS1 line compared to VED in all the time 393 points studied along ripening. Furthermore, hypomethylation of the ACO1 394 promoter (CG and CHH context) was observed at 25 and 30 DAP and the ETR1 395 promoter region (CHG and CHH context) at 30 DAP. In contrast, CHG 396 hypermethylation of NAC-NOR was found from the earliest stage until 30 DAP 397 in the mutant and was hypomethylated at 30 DAP in the CHH context. For the 398 other two transcription factors, we observed CHH hypomethylation of RIN and 399 CNR promoter regions at H time. These results suggest that CmROS1 plays a 400 role in the complex modulation of DNA methylation levels of promoter regions of 401 important genes involved in ripening. 402 403 Discussion 404 Advances in genome editing have been obtained applying the CRISPR/Cas 405 technology in several plant species. However, among the Cucubitaceae family 406 studies were only reported in watermelon for herbicide resistance (Tian et al., 407 2016, 2018) and cucumber for virus resistance (Chandrasekaran et al., 2016). 408 More recently, edited plantlets with a disruption of a visual reporter gene 409 (CmPDS), which could not be carried to the next generation, were generated in 410 melon using CRISPR/Cas9 (Hooghvorst et al., 2019). To our knowledge, 411 hereby we report for the first time the generation of melon knockout mutants for 412 an agronomic important trait such as fruit ripening and the inheritance of the 413 introduced mutations to the following generations using CRISPR/Cas9. 12 414 Melon is considered a recalcitrant species for genetic transformation. In this 415 study, we obtained on average 15% transgenic plants and from these, 40% and 416 46% of them were successfully edited plants for our target genes CmROS1 and 417 CmCTR1-like using either two or one gRNA strategy, respectively. The edited 418 plants carried several types of editions nearby the protospacer adjacent motif 419 (PAM) sequence of the target gRNA. As reported for other species (Feng et al., 420 2014), biallelic edited plants were obtained (70% of the edited plants), 421 suggesting early editions during developmental stages. 422 In accordance with the mutations induced by Non-homologous end Joining 423 pathway, the sequence analysis of the edited lines revealed that the most 424 frequent editions were insertions and deletions with more than one independent 425 event exhibiting the same edition. All the gRNA used here successfully induced 426 mutations in the target genes. However, editions in CmROS1 were mainly 427 obtained in gRNA1 suggesting a higher edition efficiency for this gRNA. In 428 addition, in contrast to the observations reported by Hooghvorst et. al., base 429 pair substitutions were not obtained for any of the genes targeted in this study. 430 The improvements in the transformation protocol of melon allowed setting up an 431 efficient method to obtain transgenic plants in a recalcitrant species and hence, 432 increased the chances to obtain edited plants. On the other hand, according to 433 our results, the efficiency of the gRNA determines the rate of success of edited 434 lines on the target genes used in this work. Thus, testing the efficiency of the 435 gRNA before transformation could be a key step for gene editing in this species. 436 Improving fruit quality and shelf life has been one of the main challenges for 437 agriculture. During the last decades, advances in understanding the ripening 438 process were approached by conventional breeding and genetic engineering 439 tools. For instance, CRISPR knockout mutants in tomato have proved the 440 importance of master ripening regulator genes (Ito et al., 2015). 441 More recent studies showed that epigenetic regulation plays a key role in fruit 442 ripening with both hypermethylated and hypomethylated loci for several species 443 (Lü et al., 2018). The balance of global DNA methylation/demethylation is 444 altered during fruit ripening and these alterations are governed by DNA 445 demethylases. In tomato, more than 200 promoters of ripening-related genes, 446 including master regulators, ethylene related genes, fruit softening, and 13 447 carotenoids synthesis genes, are regulated by DNA demethylation at the onset 448 of ripening (Zhong et al., 2013). 449 In Arabidopsis, the protein repressor of silencing 1 (ROS1), which belongs to 450 the subfamily of bifunctional 5-methylcytosine DNA glycosylases/lyases, has 451 been characterized as the main sporophytic DNA demethylase (Gong et al., 452 2002). In tomato, there are four genes (SlDML1, SlDML2, SlDML3 and SlDML4) 453 encoding putative DNA demethylases, being SlDML2 the closest ortholog to 454 Arabidopsis ROS1 gene. Furthermore, SlDML2 expression is highly correlated 455 with fruit ripening (Zhong et al., 2013; Liu et al., 2015). In melon, we have 456 identified four putative ROS1 homologues (MELO3C024516, MELO3C021451, 457 MELO3C002241 and MELO3C009432) (Fig. S2). The gene MELO3C024516 458 locates in the previously identified ripening QTL interval ETHQV8.1 (Table S6) 459 and therefore was edited in this study. The CRISPR-ROS1/2 lines, carrying 460 loss-of-function homozygous alleles of MELO3C024516, showed an advance in 461 climacteric ripening compared to the wild type, suggesting a role of this gene in 462 the complex regulation of climacteric ripening in melon. Interestingly, RNA-seq 463 expression analysis of several fruit ripening stages in wild type climacteric VED 464 shows that the four putative ROS1 genes have a similar expression profile 465 along ripening (Fig. S3), suggesting that more than one DNA demethylase may 466 be involved in this process. Moreover, a recent study showed hypomethylation 467 of ethylene induced genes at 30 DAP in a climacteric variety, suggesting the 468 important role of DNA demethylases during melon ripening (Feder et al., 2020). 469 Unlike in tomato, in which the CRISPR SlDML2 mutant showed an inhibitory 470 effect on fruit ripening (Lang et al., 2017), the CmROS1 knockout melon fruit 471 ripens ahead of the wild type VED. 472 Our methylome analysis of the climacteric variety VED showed an overall 473 demethylation in CG and CHG context along fruit ripening, similar to what has 474 been reported in tomato (Liu et al., 2015; Lang et al., 2017), orange (Huang et 475 al., 2019) and strawberry (Cheng et al., 2018). 476 The asymmetry in the relative number of statistically significant hypermethylated 477 and hypomethylated DMRs between VED and ROS1-CRISPR-2, lead to an 478 overall hypomethylation of CG and hypermethylation of CHG in the ROS1 479 CRISPR mutant. Both hyper and hypo- methylated loci were also reported in 480 Arabidopsis ros1 mutants (Penterman et al., 2007). Furthermore, 14 481 hypomethylation levels in the promoter regions of key ripening genes (e.g. 482 ACS1, ETR1, ACO1) are in agreement with the phenotype displayed by the 483 CmROS1 CRISPR lines. The expression level of these genes and other DMRs 484 involved in ethylene signaling and ripening needs to be further studied to 485 provide insights into ripening regulation in melon. In addition, in our study, 486 genes related to biotic stress response were also hypomethylated in ROS1 vs 487 VED at harvest, suggesting a possible role of this DNA demethylase in stress- 488 response genes, as reported for ROS1, DML2, DML3 in response to biotic 489 stress in Arabidopsis (Le et al., 2014; Halter et al., 2021). 490 Both mutant lines of CTR1-CRISPR promote fruit ripening in melon in 491 agreement to the phenotype described when silencing LeCTR1 in tomato fruits 492 (Fu et al., 2005) and the previously described role of CTR1 as a negative 493 regulator of ethylene signaling in other species (Binder, 2008). This second 494 candidate gene of the QTL ETHQV8.1 (Table S6) is closely related to CTR1 in 495 other species (Fig. S4) and is differentially expressed at harvest between a non- 496 climacteric and a climacteric variety (Fig. S5). Our study shows that CmCTR1- 497 like plays also an important role in the ripening process as a negative regulator 498 affecting the initiation of the ripening process but without affecting other 499 important traits such as firmness. 500 To our knowledge, this is the first time that the CRISPR technology has been 501 implemented on genes involved in agronomically important traits in melon. The 502 implementation of this technology in this species and the inheritance of the 503 editions to the following generations is of high interest and a valuable resource 504 not only for researchers but also for breeders. We have functionally validated 505 two genes involved in the complex regulation of fruit ripening and studied in 506 depth the role of the DNA demethylase ROS1 in fruit ripening. However, as 507 mutants for both candidate genes CmROS1 and CmCTR1-like showed an 508 altered ripening phenotype, further studies are needed to identify which of them 509 is the candidate for ETHQV8.1. 510 511 Supplementary Data 512 Supplementary Tables: 15 513 Supplementary Table 1: List of gRNAs to target CmROS1 and CmCTR1-like 514 genes. 515 Supplementary Table 2: List of primers to detect Cas9 and mutations in 516 CRISPR lines. 517 Supplementary Table 3: Climacteric ripening related traits in two consecutive 518 summer seasons. 519 Supplementary Table 4: DMRs hypermethylated in ROS line compared to VED 520 at harvest time 521 Supplementary Table 5: GO of target genes at harvest time 522 Supplementary Table 6: List of potential candidate genes for ETHQV8.1. 523 Supplementary Figures: 524 Supplementary Figure 1: Editions obtained for ROS1 and CmCTR1-like in T0 525 melon plants. 526 Supplementary Figure 2: Phylogenetic tree of the ROS1 homologous proteins in 527 Arabidopsis, tomato and melon. 528 Supplementary Figure 3: RNASeq dataset of ROS1 melon homologues in the 529 climacteric genotype VED and the non-climacteric genotype ¨Piel de sapo 530 (PS)¨. 531 Supplementary Figure 4: Phylogenetic tree of CTR1 homologous proteins in 532 several plant species 533 Supplementary Figure 5: RNASeq dataset of gene CmCTR1-like in the 534 climacteric genotype VED and the non-climacteric genotype ¨Piel de sapo (PS)¨ 535 along ripening stages. 536 537 Acknowledgements 538 We acknowledge F. Garcia and E. del Blanco for technical support, and K. G. 539 Alexiou for data management. 540 541 Author contributions 542 543 Conceptualization: AG, MP and JGM. Investigation and Methodology: AG, MSD 544 and LQ. Writing-Original draft preparation: AG. Writing-Review and editing: AG, 545 MP, MM, JGM, LQ. Supervision: JGM and MM. All authors read and approved 546 the final manuscript. 16 547 548 Data availability 549 550 All available data can be found within manuscript and Supplementary Materials, 551 with further enquiries being directed to the corresponding authors. Bisulfite 552 sequencing data were deposited to European Nucleotide Archive (ENA: 553 https://www.ebi.ac.uk/ena/browser/home) with the project accession number 554 PRJEB51881. 555 556 Conflict of interest 557 The authors have no conflicts to declare. 558 559 Funding 560 This work was supported by Grant RTI2018-097665-B-C2, funded by 561 MCIN/AEI/ 10.13039/501100011033 and by “ERDF A way of making Europe”, 562 Severo Ochoa Programme for Centres of Excellence in R&D CEX2019‐000902 563 funded by MCIN/AEI/ 10.13039/501100011033 and the CERCA 564 Programme/Generalitat de Catalunya to J.G.-M. AG was supported by the 565 European Union’s Horizon 2020 research and innovation programme under 566 Marie Skłodowska-Curie (grant agreement No 793090). MSD was supported by 567 a FPI grant from the Spanish Ministry of Economy and Competitiveness BES- 568 2017-079956 funded by MCIN/AEI/ 10.13039/501100011033 and by “ESF 569 Investing in your future”. Work in the Quadrana group is supported by the 570 European Research Council (ERC) under the European Union’s Horizon 2020 571 research and innovation program (grant agreement No. 948674). 572 References Akalin A, Kormaksson M, Li S, Garrett-Bakelman FE, Figueroa ME, Melnick A, Mason CE. 2012. MethylKit: a comprehensive R package for the analysis of genome- wide DNA methylation profiles. Genome Biology 13, 1–9. Binder BM. 2008. The ethylene receptors: Complex perception for a simple gas. Plant Science 175, 8–17. Castelblanque L, Marfa V, Claveria E, I M, L P-G, Dolcet-sanjuan R. 2008. Improving the genetic transformation efficiency of Cucumis melo subsp. melo “Piel de Sapo” via Agrobacterium. Transformation, 627–632. Chandrasekaran J, Brumin M, Wolf D, Leibman D, Klap C, Pearlsman M, Sherman A, Arazi T, Gal-On A. 2016. Development of broad virus resistance in non- transgenic cucumber using CRISPR/Cas9 technology. Molecular Plant Pathology, 1–14. 17 Cheng J, Niu Q, Zhang B, Chen K, Yang R, Zhu JK, Zhang Y, Lang Z. 2018. Downregulation of RdDM during strawberry fruit ripening. Genome Biology 19. Doyle, JJ. 1990. Isolation of plant DNA from fresh tissue. Focus 12, 13–15. Feder A, Jiao C, Galpaz N, et al. 2020. Melon ethylene-mediated transcriptome and methylome dynamics provide insights to volatile production. bioRxiv. doi: 10.1101/2020.01.28.923284 [PrePrint] Feng Z, Mao Y, Xu N, et al. 2014. Multigeneration analysis reveals the inheritance, specificity, and patterns of CRISPR/Cas-induced gene modifications in Arabidopsis. Proceedings of the National Academy of Sciences 111, 4632–4637. Fu DQ, Zhu BZ, Zhu HL, Jiang WB, Luo YB. 2005. Virus-induced gene silencing in tomato fruit. Plant Journal 43, 299–308. García-Almodóvar RC, Gosalvez B, Aranda MA, Burgos L. 2017. Production of transgenic diploid Cucumis melo plants. Plant Cell, Tissue and Organ Culture (PCTOC) 2017 130:2 130, 323–333. Giovannoni J. 2001. Molecular biology of fruit maturation and ripening. Annual review of plant physiology and plant molecular biology 52, 725–749. Giovannoni JJ. 2007. Fruit ripening mutants yield insights into ripening control. Current Opinion in Plant Biology 10, 283–289. Giovannoni J, Nguyen C, Ampofo B, Zhong S, Fei Z. 2017. The Epigenome and Transcriptional Dynamics of Fruit Ripening. https://doi.org/10.1146/annurev-arplant- 042916-040906 68, 61–84. Gong Z, Morales-Ruiz T, Ariza RR, Roldán-Arjona T, David L, Zhu J-K. 2002. ROS1, a Repressor of Transcriptional Gene Silencing in Arabidopsis, Encodes a DNA Glycosylase/Lyase. Cell 111, 803–814. Halter T, Wang J, Amesefe D, Lastrucci E, Charvin M, Rastogi MS, Navarro L. 2021. The arabidopsis active demethylase ros1 cis-regulates defense genes by erasing dna methylation at promoter-regulatory regions. eLife 10, 1–62. Hooghvorst I, López-Cristoffanini C, Nogués S. 2019. Efficient knockout of phytoene desaturase gene using CRISPR/Cas9 in melon. Scientific Reports 2019 9:1 9, 1–7. Huang H, Liu R, Niu Q, Tang K, Zhang B, Zhang H, Chen K, Zhu JK, Lang Z. 2019. Global increase in DNA methylation during orange fruit development and ripening. Proceedings of the National Academy of Sciences of the United States of America 116, 1430–1436. Ito Y, Nishizawa-Yokoi A, Endo M, Mikami M, Toki S. 2015. CRISPR/Cas9- mediated mutagenesis of the RIN locus that regulates tomato fruit ripening. Biochemical and Biophysical Research Communications 467, 76–82. Kieber JJ, Rothenberg M, Roman G, Feldmann KA, Ecker JR. 1993. CTR1, a negative regulator of the ethylene response pathway in arabidopsis, encodes a member of the Raf family of protein kinases. Cell 72, 427–441. Krueger F, Andrews SR. 2011. Bismark: a flexible aligner and methylation caller for Bisulfite-Seq applications. Bioinformatics 27, 1571–1572. Lang Z, Wang Y, Tang K, Tang D, Datsenka T, Cheng J, Zhang Y, Handa AK, Zhu JK. 2017. Critical roles of DNA demethylation in the activation of ripening- induced genes and inhibition of ripening-repressed genes in tomato fruit. Proceedings of the National Academy of Sciences of the United States of America 114, E4511–E4519. Le T-N, Schumann U, Smith NA, et al. 2014. DNA demethylases target promoter transposable elements to positively regulate stress responsive genes in Arabidopsis. Genome Biology 2014 15:9 15, 1–18. 18 Liu R, How-Kit A, Stammitti L, et al. 2015. A DEMETER-like DNA demethylase governs tomato fruit ripening. Proceedings of the National Academy of Sciences of the United States of America 112, 10804–10809. Lü P, Yu S, Zhu N, et al. 2018. Genome encode analyses reveal the basis of convergent evolution of fleshy fruit ripening. Nature Plants 4, 784–791. Manning K, Tör M, Poole M, Hong Y, Thompson AJ, King GJ, Giovannoni JJ, Seymour GB. 2006. A naturally occurring epigenetic mutation in a gene encoding an SBP-box transcription factor inhibits tomato fruit ripening. Nature Genetics 2006 38:8 38, 948–952. McMurchie EJ, McGlasson WB, Eaks IL. 1972. Treatment of fruit with propylene gives information about the biogenesis of ethylene. Nature 237, 235–6. Penterman J, Zilberman D, Jin HH, Ballinger T, Henikoff S, Fischer RL. 2007. DNA demethylation in the Arabidopsis genome. Proceedings of the National Academy of Sciences of the United States of America 104, 6752–6757. Pereira L, Pujol M, Garcia-mas J, Phillips MA. 2017. Non-invasive quantification of ethylene in attached fruit headspace at 1 p . p . b . by gas chromatography – mass spectrometry, 172–183. Pereira L, Ruggieri V, Pérez S, Alexiou KG, Fernández M, Jahrmann T, Pujol M, Garcia-Mas J. 2018. QTL mapping of melon fruit quality traits using a high-density GBS-based genetic map. BMC Plant Biology 18, 1–17. Pereira L, Santo Domingo M, Ruggieri V, Argyris J, Phillips MA, Zhao G,, et al. 2020. Genetic dissection of climacteric fruit ripening in a melon population segregating for ripening behavior. Horticulture Research 7, 1–18. R Core Team. 2020. A Language and Environment for Statistical Computing. RStudio: Integrated development environment for R. 2012. Boston, MA, USA: RStudio Inc. Ruggieri V, Alexiou KG, Morata J, et al. 2018. An improved assembly and annotation of the melon (Cucumis melo L.) reference genome. Scientific reports 8. Schiml S, Puchta H. 2016. Revolutionizing plant biology: multiple ways of genome engineering by CRISPR / Cas. Plant Methods, 1–9. Tian S, Jiang L, Cui X, et al. 2018. Engineering herbicide-resistant watermelon variety through CRISPR/Cas9-mediated base-editing. Plant Cell Reports 2018 37:9 37, 1353– 1356. Tian S, Jiang L, Gao Q, et al. 2016. Efficient CRISPR/Cas9-based gene knockout in watermelon. Plant Cell Reports 2016 36:3 36, 399–406. Vrebalov J, Ruezinsky D, Padmanabhan V, White R, Medrano D, Drake R, Schuch W, Giovannoni J. 2002. A MADS-box gene necessary for fruit ripening at the tomato ripening-inhibitor (rin) locus. Science (296, 343–346. Wang R, Angenent GC, Seymour G, de Maagd RA. 2020. Revisiting the Role of Master Regulators in Tomato Ripening. Trends in Plant Science 25, 291–301. Yamamuro C, Miki D, Zheng Z, Ma J, Wang J, Yang Z, Dong J, Zhu J-K. 2014. Overproduction of stomatal lineage cells in Arabidopsis mutants defective in active DNA demethylation. Nature Communications 2014 5:1 5, 1–7. Zhong S, Fei Z, Chen YR, et al. 2013. Single-base resolution methylomes of tomato fruit development reveal epigenome modifications associated with ripening. Nature Biotechnology 31, 154–159. Zhu JK. 2009. Active DNA demethylation mediated by DNA glycosylases. Annual Review of Genetics 43, 143–166. 19 573 Figure Legends: 574 Figure 1: Schematic representation of the target sites for CRISPR/Cas9 and 575 selected CRISPR edited lines. (A) Position of the gRNA target sites (red 576 triangle) and selected mutations in the CRISPR lines for CmROS1 (B) Position 577 of the gRNA target sites (red triangle) and selected mutations in the CRISPR 578 lines for CmCTR1-like (C) Fruit phenotype at harvest time of the wild type VED 579 and CRISPR edited lines. 580 581 Figure 2: Evaluation of climacteric ripening associated traits in CRISPR edited 582 lines and VED (in two consecutive years) and ethylene emission rates. (A) 583 Earliness of chlorophyll degradation (ECD), Earliness of production of aroma 584 (EARO) and Earliness of abscission layer formation (EALF) in 2020 (B) ECD, 585 EARO and EALF in 2021. Means followed by different letters differ significantly 586 (T-test, p<0.05) (C) Ethylene production in attached fruits from 25 days after 587 pollination (DAP) until harvest in 2020. 588 589 Figure 3: General methylation and DMR regions at different ripening stages 590 (15, 25, 30 DAP and (H) Harvest point) of VED and CRISPR-ROS1 line. (A) 591 Fruit ripening stages (B) number of DMRs along ripening in VED (C) number of 592 DMRs in VED vs CRISPR-ROS1 at the same time point of ripening (D) DMRs 593 annotation in VED along ripening (E) DMRs annotation in VED vs CRISPR- 594 ROS1 at the same time point of ripening. DMRs were detected using Fisher’s 595 exact test and applying a Benjamin-Hochberg (BH) adjusted statistical threshold 596 (FDR 0.05). Only DMRs with DNA methylation changes of 40%, 20% and 20% 597 for CG, CHG and CHH context, respectively were consider. 598 Figure 4: DNA methylation levels of ethylene related genes and ripening associated transcription factors for VED and CRISPR-ROS1 at different fruit ripening development stages in the three contexts (A) ACO1 in CHH context at 25 (left) and 30 DAP (right) (B) ACS1 in CHG context at 25 (left) and 30 DAP (right) (C) ETR1 in CHG context at 30 DAP (D) ripening associated transcription factors in CHH context: NAC-NOR at 30 DAP, RIN at Harvest (H) point and CNN at Harvest (H) point. 20 A ROS1(MELO3C024516) 5´ 3´ gRNA1 AACATCTAGTCCAGCTAGTCAGGGAGCTGCCAATTCTCAAGAAAACGTGAGAACTAAAAGGAAATATGTCCGAAAGAATGCAGCCAACAAATCCCTGGAAAACCCTCTTGAACCAG GAACTCTTGATCCAGTCACCCCAGCTGGTTCTTTGGAAAATTCAAGAGGTCTAAGGACATACACTCGAAAGCGTGGCGTCAACATAATAGAAACAGGCGCATCGACTGATATGG gRNA2 VED AACCAAGAACATCTAGTCCAGCTAGTCAGG ROS1-CRISPR-1 AACCAAGAACATCTAGTCCAGCTATGTCAGG (+1bp) ROS1-CRISPR-2 AACC-----------------------AGG (-23bp) B CTR1-like(MELO3C024518) 5´ 3´ CCATATAGTGTACAGTCGACCCC gRNA1 VED TTCTCCATATAGTGTACAGTCGACCCC CTR1-CRISPR-1 TTCT-----------ACAGTCGACCCC (-11bp) CTR1-CRISPR-2 TTCTCCATATAAGTGTACAGTCGACCCC (+1bp) C VED ROS1-CRISPR-1 ROS1-CRISPR-2 CTR1-CRISPR-1 CTR1-CRISPR-2 Figure 1: Schematic representation of the target sites for CRISPR/Cas9 and selected CRISPR edited lines. (A) Position of the gRNA target sites (red triangle) and selected mutations in the CRISPR lines for CmROS1(B) Position of the gRNA target sites (red triangle) and selected mutations in the CRISPR lines for CmCTR1-like (C) Fruit phenotype at harvest time of the wild type VED and CRISPR edited lines. A Summer season 2020 B Summer season 2021 C Figure 2: Evaluation of climacteric ripening associated traits in CRISPR edited lines and VED (in two consecutive years) and ethylene emission rates. (A) Earliness of chlorophyll degradation (ECD), Earliness of production of aroma (EARO) and Earliness of abscission layer formation (EALF) in 2020 (B) ECD, EARO and EALF in 2021. Means followed by different letters differ significantly (Tukey test, p<0.05) (C) Ethylene production in attached fruits from 25 days after pollination (DAP) until harvest in 2020. 15 25 30 H A VED ROS1-CRISPR VED CRISPR-ROS1 B C hypermethylation hypermethylation Number of DMRs Number of DMRs hypomethylation hypomethylation D CG CHG CHH hypermethylation hypermethylation hypomethylation hypermethylation hypomethylation hypomethylation E CG CHG CHH hypermethylation hypermethylation hypomethylation hypermethylation hypomethylation hypomethylation Figure 3: General methylation and DMR regions at different ripening stages (15, 25, 30 DAP and (H) Harvest point) of VED and CRISPR-ROS1 line. (A) Fruit ripening stages (B) number of DMRs along ripening in VED (C) number of DMRs in VED vs CRISPR- ROS1 at the same time point of ripening (D) DMRs annotation in VED along ripening (E) DMRs annotation in VED vs CRISPR-ROS1 at the same time point of ripening. DMRs were detected using Fisher’s exact test and applying a Benjamin-Hochberg (BH) adjusted statistical threshold (FDR 0.05). Only DMRs with DNA methylation changes of 40%, 20% and 20% for CG, CHG and CHH context, respectively were consider. Ethylene related genes A ACO1 25DAP ACO1 30DAP promoter region promoter region CHH methylation level CHH methylation level 100 100 VED VED 0 0 100 100 CRISPR-ROS1 CRISPR-ROS1 0 0 2,410,000 2,412,500 2,412,500 2,410,000 genome position (bp) genome position (bp) B promoter region 25DAP ACS1 promoter region 30DAP ACS1 CHG methylation level 100 100 CHG methylation level VED VED 0 0 100 100 CRISPR-ROS1 23,752,000 0 23,755,000 CRISPR-ROS1 0 23,752,000 23,755,000 genome position (bp) genome position (bp) C ETR1 30DAP promoter region CHG methylation level 100 VED 0 100 CRISPR-ROS1 0 20,325,000 20,330,000 genome position (bp) Ripening associated TF 30DAP H point D NAC-NOR promoter region CNR promoter region 100 100 CHH methylation level CHH methylation level VED VED 0 0 100 100 CRISPR-ROS1 CRISPR-ROS1 0 0 27,665,000 27,667,500 22,700,000 22,702,500 genome position (bp) genome position (bp) H point RIN promoter region CHH methylation level 100 VED 0 100 CRISPR-ROS1 0 12,325,000 12,330,000 genome position (bp) Figure 4: DNA methylation levels of ethylene related genes and ripening associated transcription factors for VED and CRISPR-ROS1 at different fruit ripening development stages in the three contexts (A)ACO1 in CHH context at 25 (left) and 30 DAP (right) (B) ACS1 in CHG context at 25 (left) and 30 DAP (right) (C) ETR1 in CHG context at 30 DAP (D) ripening associated transcription factors in CHH context: NAC-NOR at 30 DAP, RIN at H (Harvest) point and CNN at H (Harvest)point.