The PcHY5 methylation is associated with anthocyanin biosynthesis and transport in ‘Max Red Bartlett’ and ‘Bartlett’ pears

WEI Wei-lin, JIANG Fu-dong, LIU Hai-nan, SUN Man-yi, LI Qing-yu, CHANG Wen-jing, LI Yuan-jun,LI Jia-ming, WU Jun#

1 College of Horticulture/State Key Laboratory of Crop Genetics & Germplasm Enhancement and Utilization, Nanjing Agricultural University, Nanjing 210095, P.R.China

2 Yantai Academy of Agricultural Sciences, Yantai 264000, P.R.China

3 College of Horticulture and Plant Protection, Henan University of Science and Technology, Luoyang 471023, P.R.China

Abstract The red coloring of pear fruits is mainly caused by anthocyanin accumulation.Red sport, represented by the green pear cultivar ‘Bartlett’ (BL) and the red-skinned derivative ‘Max Red Bartlett’ (MRB), is an ideal material for studying the molecular mechanism of anthocyanin accumulation in pear.Genetic analysis has previously revealed a quantitative trait locus (QTL) associated with red skin color in MRB.However, the key gene in the QTL and the associated regulatory mechanism remain unknown.In the present study, transcriptomic and methylomic analyses were performed using pear skin for comparisons between BL and MRB.These analyses revealed differential PcHY5 DNA methylation levels between the two cultivars; MRB had lower PcHY5 methylation than BL during fruit development, and PcHY5 was more highly expressed in MRB than in BL.These results indicated that PcHY5 is involved in the variations in skin color between BL and MRB.We further used dual luciferase assays to verify that PcHY5 activates the promoters of the anthocyanin biosynthesis and transport genes PcUFGT, PcGST, PcMYB10 and PcMYB114, confirming that PcHY5 not only regulates anthocyanin biosynthesis but also anthocyanin transport.Furthermore, we analyzed a key differentially methylated site between MRB and BL, and found that it was located in an intronic region of PcHY5.The lower methylation levels in this PcHY5 intron in MRB were associated with red fruit color during development, whereas the higher methylation levels at the same site in BL were associated with green fruit color.Based on the differential expression and methylation patterns in PcHY5 and gene functional verification, we hypothesize that PcHY5, which is regulated by methylation levels, affects anthocyanin biosynthesis and transport to cause the variations in skin color between BL and MRB.

Keywords: pear, PcHY5, DNA methylation, anthocyanin, biosynthesis and transport

1.Introduction

Red pears are particularly popular with consumers, so red fruit color is an important goal of pear (Pyrusspp.)breeding.Anthocyanin accumulation is known to cause red peels in pear fruits (Dussiet al.1995).Anthocyanins are natural pigments that are widely distributed among plant organs, giving blue, purple, red, or orange coloration to the fruits, flowers, leaves, and roots (Zhanget al.2020; Ayvazet al.2022).As the major class of plant pigments, anthocyanins not only promote seed dispersal and pollinator attraction but also protect plants from a range of biotic and abiotic stresses (Kovinichet al.2014).As polyphenols, anthocyanins also possess antioxidant activity, so they can prevent cancer, diabetes, obesity,and neurological diseases, among other human disorders(Yanget al.2011; Alappat and Alappat 2020).

Bud sports often occur in fruit plants and serve as an important material for trait improvement and breeding(Huanget al.2023).They are also useful resources for studying traits such as sugar metabolism (Wanget al.2022), fruit size (Buet al.2022), plant height (Zhenget al.2022), seedlessness (Royoet al.2018), and color(Azuma and Kobayashi 2022).The pear cultivar ‘Max Red Bartlett’ (MRB), which produces dark red fruit, is a bud mutation of ‘Bartlett’ (BL) pear (Reimer 1951).Over the past two decades, researchers have studied MRB to determine why it can produce a large amount of anthocyanins in the skin, but a clear explanation has yet to be found.Linkage group (LG) 4 was identified from the‘Abbè Fétel’×MRB cross and was found to be associated with skin color regulation (Dondiniet al.2008).Later,LG4 was confirmed as the candidate region that regulates skin color in MRB (Kumaret al.2019).Furthermore,methylation of thePyruscommunisMYB10promoter may be associated with the green-skin mutation in MRB (Wanget al.2013).However,PcMYB10is located in LG9, not LG4, indicating that multiple genes are likely involved in the red color mutation of MRB.

Anthocyanin biosynthesis involves a series of enzymes, with phenylalanine as the starting material.The anthocyanin synthesis enzymes which have been identified in many plant species can be divided into early biosynthetic genes (EBGs) and late biosynthetic genes (LBGs) (Yanget al.2018).EBGs, which function in the early steps of the pathway, includephenylalanine ammonialyase(PAL),4-coumaryl:CoAligase(4CL),chalconesynthase(CHS),chalconeisomerase(CHI),andflavanone3-hydroxylase(F3H).The LBGs includeflavonoid3′5′-hydroxylase(F3′5′H),dihydroflavonol 4-reductase(DFR),anthocyanidinsynthase(ANS),UDP-3-O-glucosyltransferases(UFGT),rhamnosyl transferase(RT),anthocyaninacyltransferase(AAC),putativeanthocyanintransporter(PAT), andglutathione-S-transferase(GST) (Alappat and Alappat 2020).Anthocyanin biosynthetic gene (ABG) expression is regulated by MBW complexes, which include three separate types of transcription factors (TFs): R2R3-MYB,basic helix-loop-helix (bHLH), and WD40-repeat proteins.The R2R3-MYB proteins, which are the key component of the MBW complex, can bind ABG promoters (Xuet al.2015).In pear, two MYB TFs (MYB10 and MYB114) have been shown to promote anthocyanin accumulation by directly regulating anthocyanin structural gene expression(Yaoet al.2017).

LONGHYPOCOTYL5(HY5), which encodes a basic leucine zipper (bZIP) TF, is a key TF for photomorphogenesis in plants that is light-inducible(Changet al.2008).InArabidopsisthaliana, HY5 promotes anthocyanin biosynthesis by directly regulatingCHS,CHI,F3H,F3′H,DFR, andANS(Jeonget al.2010).Moreover, AtHY5 activatesAtPAP1expression by directly binding to G-boxes and ACE-boxes in theAtPAP1promoter region (Shinet al.2013).PyHY5 can also bind to the G-box motifs in thePyMYB10andPyWD40promoters to enhance the expression of these genes;and in the red pear cultivar ‘Yunhongli No.1’, this leads to increased anthocyanin accumulation (Wanget al.2020).

DNA methylation, which can be stably inherited across generations, is a well-studied chromatin modification found in both animals and plants (Kohler and Springer 2017).DNA methylation plays an important role in plant growth and development, and in regulating the activities of transposable elements (TEs) (Anet al.2022).Three sequence contexts containing cytosine (C)residues are methylated in plants: CpG, CHG, and CHH(Bewicket al.2017).The methylation of DNA regulatory elements may affect TF binding, thus regulating gene expression (Niederhuth and Schmitz 2017).In apple(Malusdomestica), methylation of theMdMYB1promoter negatively regulates anthocyanin biosynthesis (Jianget al.2020).Similarly, methylation of thePcMYB10promoter in pear leads to lowerPcMYB10expression, repressing anthocyanin biosynthesis (Wanget al.2013).Methylation affects not only promoter regions, but also enhancers.For example, inArabidopsis, DNA methylation ofFLOWERING LOCUST(FT) gene enhancers affects the expression levels of the corresponding genes, altering the flowering time (Zicolaet al.2019).An intron of the homeotic geneDEFICIENScontains a long interspersed nuclear element(LINE) retrotransposon related to the rice retrotransposon Karma; while in oil palm (Elaeisguineensis), alternative splicing and premature termination ofEgDEF1are caused by intronic DNA hypomethylation, producing somaclonal variation (Ong-Abdullahet al.2015).

In the present study, a correlation analysis was conducted to compare the transcriptomes and methylomes of BL and red-skinned mutant MRB pears.PcHY5was found to be differentially expressed and differentially methylated between BL and MRB during fruit development.Functional characterization revealed that PcHY5 can activatePcMYB10,PcMYB114,PcUFGT,andPcGST, thus regulating anthocyanin biosynthesis.Based on gene functional verification and the patterns of differentialPcHY5expression and DNA methylation, we hypothesize that the DNA methylation ofPcHY5may be affecting anthocyanin biosynthesis and transport.

2.Materials and methods

2.1.Plant materials and growth conditions

‘Bartlett’ (BL) and the red-skinned ‘Max Red Bartlett’(MRB) (Pyruscommunis) fruit samples were harvested at the Yantai Academy of Agricultural Sciences, Yantai,Shandong Province, China at 30, 60, 90, and 120 days after full bloom (DAFB).The sampling included three biological replicates for each cultivar at each developmental stage.After collection, the fruits were peeled and the skins were immediately placed in liquid nitrogen, then stored at -80°C prior to subsequent analysis.

2.2.Anthocyanin extraction and quantification

Anthocyanin contents were determined as described by Yaoet al.(2017).Briefly, each sample was ground in liquid nitrogen and weighed, then incubated in methanol with 0.1% HCl at 4°C for 12 h.Samples were centrifuged for 20 min at 12 000×g and the supernatant was collected.Absorbance (A) values were measured in the supernatant at 530, 620, and 650 nm using an ultraviolet (UV)–visible light spectrophotometer.Anthocyanin content was calculated using the following formula:

Anthocyanin content=[(A530–A620)–0.1×(A650–A620)]/FW where FW is the fresh weight.Three biological replicates were analyzed per cultivar at each developmental stage.

2.3.Library construction and whole-genome bisulfite sequencing analysis (WGBS)

BL and MRB fruits were collected at 30 and 60 DAFB,and DNA was extracted from the skin.For each sample,100 ng of genomic DNA was aliquoted and spiked with 0.5 ng λDNA as a reference.The resulting samples were fragmented to 200–300 bp in sizeviasonication with a Covaris S220.Fragmented DNA samples were treated with bisulfite using the EZ DNA Methylation-Gold Kit(Zymo Research, Orange, CA, USA), and sequencing libraries were constructed by the Novogene Corporation(Beijing, China).Library quality was assessed with the Agilent Bioanalyzer 2100 System (Agilent Technologies,California, USA).Paired-end sequencing was performed on the Illumina platform (Illumina, CA, USA), generating 150-bp paired-end reads.

2.4.RNA sequencing (RNA-seq) and data analysis

BL and MRB fruits were sampled at 30 and 60 DAFB,and the skins were collected.Total RNA isolation,sequencing library construction, and RNA-seq were performed by Novogene (Beijing, China) as described by Liuet al.(2019).Significantly differentially expressed genes (DEGs) were called using thresholds of |log2(fold change)|>1.0 andP≤0.05.Three biological replicates were analyzed for each cultivar.

2.5.Quantitative reverse transcription (qRT)-PCR analysis

BL and MRB fruits were collected at 30, 60, 90, and 120 DAFB.Total RNA was isolated from the skin with a FastPure Plant Total RNA Isolation Kit (Vazyme, Nanjing,China).cDNA was synthesized using the Hifair?III 1st Strand cDNA Synthesis SuperMix for qPCR (gDNA digester plus) (Yeason, Shanghai, China) following the manufacturer’s instructions.qRT-PCR reactions were performed in a total volume of 10 μL, containing 5 μL of 2× SYBR Green master mix (11201ES08, Yeason,Shanghai, China), 1 μL of 10× diluted cDNA, 0.2 μL each forward and reverse primer at 10 μmol L–1, and 3.6 μL ddH2O.Amplification was performed using a LightCycler480 II (Roche, Switzerland).Tubulin(Chenet al.2020) was used as the internal reference gene for expression normalization.Relative gene expression was calculated with the 2–ΔΔCTmethod (Livak and Schmittgen 2001).The analysis included three biological replicates per sample.The primers are shown in Appendix A.

2.6.PcHY5 overexpression in pear

Transient expression assays were performed in pear as described by Liuet al.(2019) usingPyrusbretschneidericv.‘Zaosu’.The coding sequences (CDSs) ofPcHY5were inserted into the pSAK277 vector and controlled by the CaMV 35S promoter.The constructs were then transformed intoAgrobacteriumstrain (GV3101), which was cultured overnight at 30°C.Next, it was diluted to an OD600of 1.0 with infiltration buffer (10 mmol L–1MES,10 mmol L–1MgCl2, and 150 mmol L–1acetosyringone).‘Zaosu’ pear skin was infiltrated at the ripe stage, and the injected fruits were incubated under strong light in an incubator.The fruit phenotypes were analyzed after five days, and the infiltrated skins were frozen in liquid nitrogen and stored at –80°C for further analysis.The primer sequences can be found in Appendix A.

2.7.Dual-luciferase reporter assay

ThePcHY5CDSs were amplified from BL and MRB.Each gene was then inserted separately into the pSAK277 effector vector under the control of the CaMV 35S promoter.ThePcMYB10,PcMYB114,PcUFGT, andPcGSTpromoters were ligated into the reporter vector pGreenII 0800-LUC.These constructs were transformed into theAgrobacteriumtumefaciensstrain GV3101 (with pSoup).TheAgrobacteriumtransformants were cultured overnight at 30°C, then diluted to an OD600of 1.0 with infiltration buffer (10 mmol L–1MES, 10 mmol L–1MgCl2,and 150 mmol L–1acetosyringone).Agrobacteriumstrains were mixed in a 9:1 ratio of those with thePcHY5vector and those with the vector containing one of thePcMYB10,PcMYB114,PcUFGT, orPcGSTpromoters,and the mixed cultures were infiltrated intoNicotiana benthamianaleaves.At three days after inoculation,firefly luciferase (LUC) and Renilla luciferase (REN)activities were measured following the manufacturer’s instructions (Yeasen, Shanghai, China).These assays included three or four biological replicates per promoter.Primer sequences are shown in Appendix A.

2.8.Bisulfite sequencing PCR (BS-PCR)

Genomic DNA samples purified from BL and MRB fruit skins were converted with sodium bisulfite using the EpiArt DNA Methylation Bisulfite Kit (Vazyme, Nanjing,China) following the manufacturer’s instructions.Bisulfite-converted DNA was amplified in 50 μL reactions containing 1.25 U ofTaqHS DNA polymerase (R007Q,TaKaRa, Tokyo, Japan), 5 μL of 10× PCR Buffer (Mg2+plus), 5 μL of dNTPs (2.5 mmol L–1), 2 μL of primers(10 μmol L–1), and ddH2O to 50 μL.The thermocycling parameters were 95°C for 3 min, followed by 35 cycles of 95°C for 10 s, 55°C for 30 s, and 72°C for 60 s.Amplified fragments were cloned into the T-vector using pMD-19(TaKaRa, Tokyo, Japan).For each sample, seven individual clones were sequenced.These analyses included three biological replicates per sample.The primers are shown in Appendix A.

2.9.Statistical analysis

Statistically significant differences between sample types were assessed with Student’st-test in the SPSS 20.0 package for Windows (SPSS, Inc., Chicago, IL, USA).The mean±standard error was calculated for each set of biological replicates.Differences were considered significant atP<0.05.

3.Results

3.1.Phenotypes and anthocyanin contents in BL and MRB during fruit development

MRB is a red-skinned mutant variety of BL.To uncover the biological basis for the difference in skin color between BL and MRB, skin samples were collected at 30, 60, 90,and 120 DAFB (Fig.1-A).Although both BL and MRB had some red color in the skin during fruit development, MRB had markedly redder skin than BL.The red coloration of both varieties gradually weakened as the fruits developed.The anthocyanin content was higher in MRB compared to BL across all developmental timepoints, although the anthocyanin contents of the MRB and BL skins gradually decreased over time, consistent with phenotypic observations (Fig.1-B).These results indicated that the color changes over time and the differences in skin color between BL and MRB were caused by anthocyanin accumulation.

3.2.WGBS analysis of BL and MRB

To identify the genes potentially associated with differences in skin color, WGBS at 10× coverage was performed on MRB and BL fruit skins collected at 30 and 60 DAFB.These timepoints were selected because they showed the largest differences in anthocyanin content between the two varieties.Pyrusbretschneideriwas used as the reference genome (Wuet al.2013).Analysis of the WGBS results showed that the total methylation level was higher in the BL genome than in the MRB genome at 30 DAFB, but lower in BL than in MRB at 60 DAFB(Fig.2-A).Among the DNA methylcytosine (mC) levels in BL and MRB, the CHH context was the most prevalent type, followed by the CpG context, with the CHG context being the least abundant (Fig.2-B).At two timepoints, the DNA methylcytosine densities were calculated in 400-kb tiled bins for each context; and the mC distributions were similar between BL and MRB in all genes and transposable elements (Appendix B).The average DNA methylation levels in the CpG and CHG contexts remained largely similar between BL and MRB across the genome.However,there were great differences between the two cultivars in CHH methylation in both gene flanking sequences and gene bodies, with lower levels in MRB than in BL at both timepoints (Fig.2-C).These results suggested that changes in DNA methylation levels, especially in the CHH context, may have affected the skin color in BL and MRB.

Fig.1 Phenotypes of ‘Bartlett’ (BL) and ‘Max Red Bartlett’ (MRB) fruits during development.A, differences in color between MRB and BL fruits over the course of fruit development.Scale bar=1 cm.B, quantification of anthocyanin contents in the skins of MRB and BL fruits over the course of fruit development.[(A530–A620)–0.1×(A650–A620)]/Fresh weight was considered one anthocyanin unit.Data are presented as the mean±standard error (n=3).＊＊, P<0.01; ＊＊＊, P<0.001 (two-tailed Student’s t-test).

Fig.2 DNA methylation levels in ‘Bartlett’ (BL) and ‘Max Red Bartlett’ (MRB) skins as determined with whole-genome bisulfite sequencing (WGBS) analysis.A, total methylation levels in BL and MRB fruit skins over time.B, proportions of methylation in each sequence context out of the total methylation levels in BL and MRB.C, DNA methylation landscape of CpG, CHG, and CHH (H=A, T, or C) (including gene bodies and ～2-kb upstream and downstream regions).TSS, transcription start sites; TTS,transcription termination sites.

3.3.Identification of candidate genes regulating anthocyanin biosynthesis and integrated transcriptomic analysis

Differentially methylated regions (DMRs) are regarded as epigenetic biomarkers or epialleles.To investigate the differential methylation of candidate ABGs, DMRs were analyzed between BL and MRB.At 30 DAFB, there were a total of 4 018 DMRs between BL and MRB; and 772 DMRs(379 CpG, 255 CHG, and 138 CHH) were hypomethylated in BL compared to MRB, whereas 3 246 DMRs (222 CpG,175 CHG, and 2 849 CHH) were hypermethylated in BL compared to MRB.At 60 DAFB, there were a total of 3 765 DMRs, with 948 hypomethylated (310 CpG, 188 CHG, and 450 CHH) and 2817 hypermethylated (232 CpG, 212 CHG, and 2 373 CHH) in BL compared to MRB(Fig.3-A).Overall, there were more hypermethylated than hypomethylated DMRs in BL compared to MRB,particularly in the CHH context, at both timepoints.These results indicated that differences in skin color between the BL and MRB fruits may have been caused specifically by differential CHH methylation.

To investigate the relationship between gene expression and DNA methylation changes in BL and MRB, we performed RNA-seq analysis on the same samples that were analyzed with WBGS.We identified a total of 8 844 differentially expressed genes (DEGs) at 30 DAFB, comprising 2 300 up-regulated and 6 544 downregulated genes in BL compared to MRB.At 60 DAFB,there were 6 550 DEGs, comprising 4 793 up-regulated and 1 757 down-regulated genes in BL compared to MRB(Appendix C).

We then analyzed the overlap between the DMRs and DEGs (as ‘DMR-mediated DEGs’) in BL and MRB at the two timepoints.There were a total of 924 overlapping DMRs and DEGs among all the DMRs and DEGs at 30 DAFB, and a total of 686 overlapping DMRs and DEGs at 60 DAFB (Fig.3-B).There were 14 candidate genes among the overlapping DMRs and DEGs at the two timepoints.The differential methylation of the nine genes belongs to the CHH context, while the remaining five belong to the CpG context.

SwissProt was used to annotate the 14 candidate genes (Appendix D).One candidate gene,Pbr002622.1(PcHY5), was highly expressed in MRB compared to BL, but had higher DNA methylation levels in an intron in BL than MRB at both 30 and 60 DAFB (Fig.3-C).This gene was then annotated and its function was predicted using the NCBI and TAIR databases, which showed thatPcHY5is homologous toMdHY5(Appendix E).We therefore hypothesized thatPcHY5expression may affect anthocyanin biosynthesis in BL and MRB.

3.4.PcHY5 promotes anthocyanin accumulation by activating anthocyanin biosynthesis and transport genes

To determine whether differences in anthocyanin accumulation between BL and MRB fruit skins were caused by variations inPcHY5expression levels, we quantifiedPcHY5expression in the skins of BL and MRB fruits collected at 30, 60, 90, and 120 DAFBviaqRT-PCR (Fig.4-A).PcHY5was more highly expressed in MRB than in BL throughout fruit development.ThePcHY5expression patterns observed in the RNA-seq data were confirmed with qRT-PCR, which also showed results consistent with the anthocyanin accumulation observed in BL and MRB.We then analyzed the expression levels of selected ABGs.PcUFGTandPcGSTwere expressed at higher levels in MRB than in BL throughout fruit development (Fig.4-A).Similarly,PcMYB10andPcMYB114expression levels were consistent with the observed differences in anthocyanin accumulation and with the differential expression ofPcUFGTandPcGSTbetween BL and MRB.These results indicated that PcHY5 may be a key TF contributing to the differences in skin color between BL and MRB.

To clarify whether PcHY5 is involved in anthocyanin biosynthesis in pear, we constructed an overexpression vector,35S::PcHY5, and transiently expressed the construct in ‘Zaosu’ pears.Fruits transiently overexpressing thegreenfluorescentprotein(GFP)control showed slightly red coloration on the surface,whereas those overexpressingPcHY5showed extremely large areas of dark red coloration (Fig.4-B).Anthocyanin quantification indicated that transientPcHY5overexpression caused increases in the skin anthocyanin content compared to the control(35S::GFP), which was consistent with the observed phenotypes.Further analysis confirmed that the expression levels ofPcHY5were significantly higher in fruits transiently expressing thePcHY5construct compared to those transiently overexpressingGFP(Fig.4-C).Furthermore, the anthocyanin biosynthesis genesPcMYB10,PcMYB114,PcUFGT, andPcGSTwere significantly up-regulated in the skins of fruits overexpressingPcHY5compared to those infiltrated with the35S::GFPconstruct (Fig.4-D).Thus, these results showed that PcHY5 increases anthocyanin biosynthesis,red coloration, and the expression levels ofPcMYB10,PcMYB114,PcUFGT, andPcGST.

Based on the transient overexpression results, we next sought to determine whether PcHY5 could promotePcMYB10,PcMYB114,PcUFGT, orPcGSTexpression.A dual-luciferase experiment demonstrated that PcHY5 did significantly enhance the expression levels ofPcMYB10,PcMYB114,PcUFGT, andPcGST(Fig.4-E).These results therefore indicated that PcHY5 could activate the transcription of anthocyanin biosynthesis genes and an anthocyanin transport gene.

3.5.A PcHY5 intron contains a DMR

Fig.4 PcHY5 promoted anthocyanin biosynthesis and transport.A, expression patterns of genes associated with anthocyanin biosynthesis in ‘Bartlett’ (BL) and ‘Max Red Bartlett’ (MRB).Data are presented as the mean±standard error (SE) (n=3).B,phenotypes of pear fruits overexpressing PcHY5 (right) or a green fluorescent protein (GFP) control (left).C, anthocyanin content in pear fruits overexpressing PcHY5 or the GFP control.[(A530–A620)–0.1×(A650–A620)]/Fresh weight was considered one anthocyanin unit.D, relative expression levels of anthocyanin-related genes in pear as analyzed with quantitative reverse transcription (qRT)-PCR.Data are presented as the mean±SE (n=3).E, detection of anthocyanin biosynthesis gene transcriptional activation activity as assessed with a dual-luciferase reporter assay in tobacco (Nicotiana benthamiana).Data are presented as the mean±SE (n=3 or 4).＊, P<0.05; ＊＊, P<0.01; ＊＊＊, P<0.001 (two-tailed Student’s t-test).

The WGBS results indicated that the key DMR between BL and MRB is within an intronic region ofPcHY5.PcHY5is located on chromosome 15 of the pear genome, and the DMR was found in the 1 000 577–1 000 822-bp region(Fig.5-A).We then performed BS-PCR to determine the methylation levels of the DMR in thePcHY5intron.The BS-PCR amplified region was 448 bp, including a differentially methylated region of 245 bp.This differential methylation region contained 9 CpG, 10 CHG, and 26 CHH sites.The DMR of thePcHY5intron had very low DNA methylation levels in MRB compared to BL at all timepoints from 30 to 120 DAFB.Furthermore, not only did BL have much higher total DNA methylation levels in this region, but it had higher methylation levels in all three contexts (CpG, CHG, and CHH) (Fig.5-B; Appendices F and G).These results showed that a key intronic region inPcHY5was highly methylated in BL compared to MRB.Notably, the BS-PCR results were consistent with the methylomic data for BL and MRB.

We used PlantCARE, a plantcis-acting regulatory element database, to identify potential binding sites in the intron of PcHY5 (Lescotet al.2002).Interestingly,the G-box is a typicalPcHY5TF binding site in pear.We therefore deduced that PcHY5 may regulate the intron(Appendix H).The intron was activated by PcHY5 binding both upstream and downstream of LUC (Appendix I).This result suggested that the intron could be activated by PcHY5, regardless of its comparative location or distance.Thus, thePcHY5intron may function as an enhancer.

4.Discussion

4.1.PcHY5 promotes anthocyanin biosynthesis by activating anthocyanin biosynthesis and transport genes in pear

Fig.5 Bisulfite sequencing analysis of the PcHY5 intronic region.A, visualization of the methylation levels in the PcHY5 intronic region based on whole-genome bisulfite sequencing (WGBS).Black outlining indicates a region with differential methylation levels between the pear cultivars ‘Bartlett’ (BL) and ‘Max Red Bartlett’ (MRB).B, methylation levels in an intronic region of PcHY5.Data are presented as the mean±standard error (n=3).＊, P<0.05; ＊＊, P<0.01; ＊＊＊, P<0.001 (two-tailed Student’s t-test).

HY5is a bZIP TF that can regulate photomorphogenesis,anthocyanin biosynthesis, root development, hormone signalling, and abiotic stress responses by functioning at the center of a transcriptional network hub (Songet al.2008;Gangappa and Botto 2016).InArabidopsis, HY5 binds to many promoters, including those ofCHI,F3H,F3′H,CHS,DFR, andANS(Shinet al.2007; Vandenbusscheet al.2007).In addition to anthocyanin structural genes, HY5 directly binds to MYB gene promoter regions, such as those ofArabidopsisMYB75(Shinet al.2013),Malus×domestica MYB1(Anet al.2017),PyruspyrifoliaMYB10(Baiet al.2019a, b), andSolanumlycopersicumAN1(Qiuet al.2019), in order to promote anthocyanin biosynthesis.PyHY5 also binds to G-box motifs in thePyWD40andPyMYB10promoters, increasing their expression and thus promoting anthocyanin accumulation in pear (Wanget al.2020).In the present study, we found that PcHY5 could promote anthocyanin biosynthesis (Fig.4-B) by increasing the expression levels of anthocyanin structural genes and related TFs (Fig.4-C and D), such asPcUFGT,PcGST,PcMYB10, andPcMYB114.As hypothesized, PcHY5 was shown to bind to thePcUFGT,PcGST,PcMYB10,andPcMYB114promoters and regulate their expression(Fig.4-E).Interestingly, we found that PcHY5 could positively regulatePcGST.Anthocyanin is synthesized on the cytosolic surface of the endoplasmic reticulum, and GSTs are considered to be responsible for anthocyanin transport into the vacuole.MYBs have previously been shown to bindGSTpromoters in order to activate these genes in apple and peach (Jianget al.2019; Zhaoet al.2020).However, the regulatory mechanisms controlling GSTs in plants are still largely unclear, and there have been very few reports of direct GST regulation by HY5.Our findings further indicated that PcHY5 may activatePcUFGT,PcGST,PcMYB10, andPcMYB114through direct binding to their promoters.Therefore, PcHY5 appears to regulate anthocyanin biosynthesis in pear through two mechanisms: activating anthocyanin structural genes and MYB TF genes, and activating an anthocyanin transport gene.Therefore, the biological significance of our study is that activating the anthocyanin structural genes is beneficial for the biosynthesis of anthocyanins.After anthocyanin biosynthesis, the anthocyanins need to be transported by transport proteins into vacuoles for storage.This combination leads to the stable presence of anthocyanins and is beneficial for the coloration of the pear skin, and thus improves fruit quality.

4.2.PcHY5 expression and methylation levels are related to the color mutation in MRB

Red pear mutants are an ideal material in which to study the molecular mechanism underlying color diversity in pear.In previous studies, anthocyanin accumulation was determined to be the primary cause of red coloring in pear (Yaoet al.2017).Similarly, our data showed that the anthocyanin content was higher in MRB than in BL fruits during the developmental stage (Fig.1).PcHY5was identified in an analysis of correlations between DMRs and DEGs in BL and MRB (Figs.2 and 3).HY5 was previously identified as a differentially expressed candidate gene in a transcriptomic analysis of MRB and BL (Daiet al.2022).In this study, we confirmed thatPcHY5was significantly up-regulated in MRB compared to BL (Figs.3 and 4).Taken together, these data suggest that differences inPcHY5expression may be responsible for the variations in skin color between BL and MRB.

Individual species differ with respect to the genes in which mutations cause variations in color.For example, a rearrangement in the upstream regulatory region ofMYB10is associated with anthocyanin accumulation in apple flesh(Espleyet al.2009).A 14-nucleotide deletion in the coding region ofPpBBX24causes the characteristic red skin of the pear cultivar ‘Zaosu Red’, which is derived from ‘Zaosu’(Ouet al.2020).InCapsicumannuum, a nonsense mutation inCaHY5that leads to a truncated protein product is associated with the near-absence of anthocyanin accumulation in the hypocotyl, and this mutation (a C to T conversion) is located in the second exon of the mutant e1898 (Chenet al.2022).Such previous results suggest thatPcHY5may have played a key role in the mutation that produced MRB from BL.However, no mutation in the HY5 DNA sequence has been identified between MRB and BL,in either the promoter or the gene body.

In addition to mutations in DNA sequences, bud sports may arise from epigenetic changes.For example, the high methylation levels in theMdMYB10promoter are likely causal for the loss-of-color mutation in ‘BLO’ apple (El-Sharkawyet al.2015).Interestingly, the methylation levels inPcMYB10may have caused the formation of a greenskinned sport from MRB (Wanget al.2013).However,this DMR was not identified in the present study, perhaps becausePcMYB10promoter methylation is unrelated to the color differences between MRB and BL.In this study, we found that the key DMR between MRB and BL was located in aPcHY5intron.Methylation levels in thePcHY5intron were lower in MRB than in BL during fruit development,andPcHY5was up-regulated in MRB.These results suggested that hypomethylation of thisPcHY5intron may increase its expression.The regulation of gene expression through the methylation of promoter regions has been well studied; however, the effects of gene body methylation on gene expression are comparatively understudied (Kimet al.2018).Introns are non-coding sequences in a gene that are involved in important biological processes, such as alternative splicing, transcriptional regulation, and nuclear export (Gaoet al.2021).Enhancers arecis-regulatory DNA sequences that can be bound by TFs to activate transcription, irrespective of their location, distance, or orientation relative to the genes that they regulate (Benetatos and Vartholomatos 2018).Enhancer regions have many distinct characteristics, such as low DNA methylation levels(Shlyuevaet al.2014), and high methylation in enhancers can alter the expression patterns of their associated genes(Clermontet al.2016; Ordonezet al.2019).

We found that the intronic region ofPcHY5may contain an enhancer region that can be regulated by PcHY5(Appendix H and I).In this study, we discovered that PcHY5 can regulate its own expression level through the intron region that has differential methylation levels, which suggests that the expression ofPcHY5may be regulated by PcHY5 itself.Enhancers typically promote gene expression (Weberet al.2016), and intronic enhancers may regulate cognate genes (Smemoet al.2014).DNase I sequencing inArabidopsishas shown that intronic enhancers are widely distributed throughout the genome and that they regulate the expression of cognate genes(Menget al.2021).In addition, the transcription factorHY5can be regulated by a number of different transcription factors to regulate anthocyanin biosynthesis, including BBX, MYC, and HY5 itself (Jobet al.2018; Chakrabortyet al.2019; Songet al.2020).Studies have shown that HY5 requires cofactors such as BBX to precisely regulate biological responses throughout various stages of plant development (Baiet al.2019a, b; Burschet al.2020).These studies indicate thatHY5can be regulated by many factors, therefore, the regulatory effect of PcHY5 on the anthocyanin biosynthesis in pear will require further study.

5.Conclusion

Our findings revealed thatPcHY5affects anthocyanin biosynthesis and transport in pear, andPcHY5, which is regulated by methylation, may be responsible for variations in skin color between BL and MRB.Thus, our research provides novel evidence thatPcHY5methylation is associated with variations in skin color between BL and MRB.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (31820103012), the earmarked fund for China Agriculture Research System (CARS-28),and the earmarked fund for Jiangsu Agricultural Industry Technology System, China (JATS[2022]454).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on https://doi.org/10.1016/j.jia.2023.07.017