Fruits of the SP3H loquat with obvious red pigmentation were collected from the National Loquat Germplasm Bank (Fuzhou, Fujian, China) and used for further analysis. Unpigmented and red-pigmented peels were collected and sliced into small pieces. Three biological replicates were prepared, with ten fruits were included in each replicate. The peels of each replicate were pooled and rapidly frozen in liquid nitrogen and stored at -80°C until use.

Peel samples were subjected to anthocyanin-targeted metabolome analysis to identify anthocyanins accumulated in red-pigmented peels. All procedures of the metabolome analysis were performed by Metware Biotechnology Co., Ltd. (Wuhan, China). Loquat peels were freeze-dried using a vacuum freeze-dryer and then ground into fine powder using a mixer mill. Anthocyanins were extracted and subjected to UPLC-MS/MS analysis as described by Fang et al. (2025). Briefly, 50 mg of powder was dissolved in 0.5 mL of extraction solution (methanol: water: hydrochloric acid=500: 500: 1, V/V/V) and the mixture was vortexed for 5 min. Then the mixture was sonicated for 5 min and centrifuged at 12, 000 g at 4°C for 3 min. The supernatant was collected and the residue was re-extracted as described above. The supernatants were pooled and filtered through a 0.22 μm pore size filter prior to LC-MS/MS analysis. UPLC-MS/MS analysis was conducted using an UPLC-ESI-MS/MS system (UPLC, ExionLC™ AD; MS, Applied Biosystems 6500 Triple Quadrupole). The parameters were as follows: the column employed was a Waters ACQUITY BEH C18 (1.7 μm, 2.1 mm×100 mm); The mobile phase was a mixture of water (0.1% formic acid): methanol(0.1% formic acid); The gradient elution was 95:5 V/V for 6 min, 50:50 V/V for 6 min, 5:95 V/V for 2 min, and then returned to 95:5 V/V for 2 min; The flow rate was 0.35 mL/min, and the temperature was maintained at 40°C; The volume of injected sample was 2 μL.

Extraction of total RNA from loquat peels and library construction were carried out at Beijing BioMarker Technologies (Beijing, China) as described in our previous study (Su et al., 2023). cDNA libraries were sequenced using Illumina NovaSeq 6000. RNA-seq was performed in three biological replicates. The RNA-seq reads have been deposited in the Genome Sequence Archive (GSA) (Chen et al., 2021) of the National Genomics Data Center (Members and Partners, 2024) (https://ngdc.cncb.ac.cn/gsa) and are accessible under PRJCA038065. The obtained raw 150 bp paired-end reads were processed as described previously (Su et al., 2023) and mapped to the genome of ‘Jiefangzhong’ (Su et al., 2021) using HISAT2. DESeq2 was used to calculate gene expression level and identify differentially expressed genes (DEGs). Only genes that met with the criteria of fold change >2 and a false discovery rate (FDR) <0.01 were selected as DEGs. GO and KEGG pathway enrichment analyses was performed as described previously (Su et al., 2023). Prediction of transcription factors was performed using the Transcription Factor Prediction module of PlantTFDB v5.0 (https://planttfdb.gao-lab.org/prediction.php).

Total RNA was extracted from loquat peels using the EZNA Plant RNA Kit (Omega Bio-tek). qRT-PCR was performed as described previously (Zhang et al., 2021). EjActin (Ej00095133) was used as the reference gene. A LightCycler 480 real-time PCR system was employed for qRT-PCR analysis, with three biological and four technical replicates. The primer sequences used in qRT-PCR are listed in Supplementary Table S1 .

cDNA from red pigmented SP3H loquat peels was synthesized using the HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China). The isolation of EjMYB10 coding sequence and insertion into pSAK277 were carried out using the 2×Phanta Max Master Mix (Vazyme, Nanjing, China) and the ClonExpress Ultra One Step Cloning Kit (Vazyme, Nanjing, China), respectively. The promoter of EjF3′H, EjANS and EjUFGT were amplified and cloned into pGreenII 0800-LUC to generate reporter construct as described previously (Zhang et al., 2021) and their sequences are provided in Supplementary Table S2 . All primers used for cloning are listed in Supplementary Table S3 .

Phylogenetic analyses were performed using the neighbor-joining method with 1000 bootstrap replicates by MEGA6. The alignment of amino acid sequences was carried out using ClustalW (https://www.genome.jp/toolsbin/clustalw). ESPript 3.0 (Robert and Gouet, 2014) was employed to shade the results of the multiple sequence alignment. The accession numbers of additional sequences from other species are provided in Supplementary Table S4 .

Transient expression of EjMYB10 was carried out in Nicotiana tabacum young leaves as described previously (Zhang et al., 2021). Basic helix-loop-helix (bHLH) transcription factors that belong to the bHLH2 subgroup (PhAN1/AtTT8) have been suggested as indispensable partners of anthocyanin-promoting R2R3-MYBs (Espley et al., 2007; Gonzalez et al., 2008). These bHLHs are required for anthocyanin biosynthesis in tobacco leaves (Montefiori et al., 2015). Sequence analysis showed that Ej00089823 (EjbHLH3) was predicted to encode a bHLH2 subgroup bHLH protein. EjbHLH3 was highly homologous to apple MdbHLH3 (Xie et al., 2012), peach PpbHLH3 (Ravaglia et al., 2013) and plum PsbHLH3 (Fang et al., 2021a) ( Supplementary Figure S1 ). We failed to clone EjbHLH3, so we chose the plum PsbHLH3 which was sufficient to induce anthocyanin accumulation in tobacco leaves when coinfiltrated with plum MYB10s (Fang et al., 2021a, Fang et al., 2021b) and blueberry VcMYBA (Zhang et al., 2021). Agrobacteria carrying constructs were cultivated and resuspended in the infiltration buffer (10 mM MES, 10 mM MgCl₂, and 100 μM acetosyringone) to an OD₆₀₀ of 0.6 and then infiltrated into tobacco leaves. Photos of transformed tobacco leaves were taken 7 d after infiltration. Anthocyanin in tobacco leaves was extracted and quantified as described by Fang et al. (2025).

pGreenII 0800-LUC vectors carrying the promoter of EjF3′H, EjANS and EjUFGT were transformed into the A. tumefacien strain GV3101 with pSoup. Dual-luciferase assays were performed using 3 to 4-week-old Nicotiana benthamiana leaves. Agrobacteria were grown and resuspended in the infiltration buffer to an OD₆₀₀ of 0.5 and then used for infiltration according to the protocol described above for transient expression assays. Firefly luciferase (LUC) activity detection and image capture were performed 3d after infiltration, as described by Fang et al. (2025). The detection of firefly luciferase and Renilla luciferase activities was carried out using a GloMax Discover instrument (Promega) and the Dual Luciferase Reporter Gene Assay Kit (Yeasen, 11402ES80, China).

Statistical analysis was performed using Student’s t-test (*P < 0.05, **P < 0.01, ***P<0.001 and ****P<0.0001). Data analysis was performed using GraphPad Prism 10.3.1 and TBtools-II (Chen et al., 2023).

Generally, the ripe SP3H loquat fruits are yellow. However, partial red pigmentation was found in the peel of some ‘Shanpaisanhao’ loquat fruits ( Figures 1A–C ). Our previous study showed that anthocyanins are the main pigments accumulated in red loquat peels (Su et al., 2023). Flavonoids metabolomics analysis was performed to investigate the composition of red pigments in the red pigmented areas. In total, 41 flavonoid metabolites, including seven cyanidin, five pelargonidin, nine delphinidin, five petunidin, two peonidin and two malvidin derivatives were identified ( Figure 1D ). 21 of the identified anthocyanins were significantly accumulated in the red-pigmented areas, and cyanidin-3-O-galactoside was the most predominant anthocyanin ( Figure 1D ). In addition, the accumulation of rutin, naringenin-7-O-glucoside and quercetin-3-O-glucoside were also detected in the red pigmented areas ( Figure 1C ). These results suggest that anthocyanin accumulation is responsible for the red pigmentation in the peel of SP3H loquat fruits.

The yellow and red pigmented peels of SP3H loquat fruits were further subjected to RNA-seq analysis to explore the mechanism of anthocyanin biosynthesis. A total of 38.17 Gb clean data was obtained from loquat peel samples ( Supplementary Table S5 ). Over 94% of the obtained reads could be mapped to the genome of ‘Jiefangzhong’ ( Supplementary Table S5 ). In total, 10072 transcripts were detected in analyzed peel samples. These include 2429 putative new genes that have not been predicted in the genome of ‘Jiefangzhong’ and 1624 of them was annotated by databases ( Supplementary Table S6 ). Transcriptome comparison showed that 2821 genes were differentially expressed (1770 upregulated and 1051 downregulated) between yellow and red pigmented areas.

KEGG enrichment analysis of DEGs indicated that 52 DEGs were involved in phenylpropanoid biosynthesis, flavonoid biosynthesis and anthocyanin biosynthesis pathways ( Supplementary Table S7 ). Nine of these genes, including CHS (Ej00014264 and Ej00014465), CHI (Ej00092682, Ej00071798 and newGene15317), F3′H (Ej00025308), ANS (newGene13949) and UFGT (Ej00006885 and Ej00084941) were assigned to the anthocyanin biosynthetic pathway ( Figure 2 ). The expression of these genes was positively correlated (r>0.65) with the cyanidin-3-O-galactoside and cyanidin-3-O-glucoside concentration ( Figure 2 ). Other anthocyanin pathway genes, including 4CL (Ej00002975, Ej00005091, Ej00046783 and newGene1146), CHS (Ej00014720, Ej00070948 and Ej00054582), F3H (Ej00026228), DFR (Ej00081751), were not identified as differentially expressed genes, but their transcription was also enhanced in the red-pigmented peels. Additionally, the expression of a glutathione S-transferase (GST) gene Ej00043900 (annotated as GSTF12) is also positively correlated (r>0.7) with the concentrations of cyanidin-3-O-galactoside and cyanidin-3-O-glucoside ( Figure 2 ). Phylogenetic analysis showed that Ej00043900 fell into a same clade with anthocyanin-related GSTs and was closely related to anthocyanin-related GST Raint from peach ( Supplementary Figure S2 ). These results suggested that anthocyanin biosynthesis in peel of SP3H loquats is regulated by the transcription of these structural genes.

The obvious upregulation of anthocyanin biosynthetic genes CHS, CHI, F3′H and ANS and anthocyanin transportation gene GSTF12 in red-pigmented peels suggests that anthocyanin accumulation in the SP3H loquat peel was transcriptionally regulated. To identify candidate transcriptional regulators that regulate anthocyanin accumulation, the sequences of loquat genes were submitted to PlantTFDB v5.0 to predict transcription factors. In total, 2365 genes were identified to encode transcriptional factor. We further analyzed the expression patterns of the identified transcription factors, and the results showed that 208 of them were differentially expressed (with maximum FPKM values ≥ 5, Supplementary Table S8 ). Among these differentially expressed transcription factors, 42 were positively correlated (r > 0.9) and 47 were negatively correlated (r|< -0.9) were positively and negatively correlated with the concentrations of cyanidin-3-O-galactoside and cyanidin-3-O-glucoside or structural genes, respectively ( Figure 3 ).

The results of RNA-Seq were validated by analyzing the expression profiles of eight candidate anthocyanin-related genes, including six anthocyanin biosynthetic genes (CHS, CHI, F3′H, ANS and UFGT), one GST gene, and one MYB, using qRT-PCR. The results showed that all analyzed genes were upregulated in the red-pigmented peel ( Figure 4 ).

MYBs serve as crucial regulators of anthocyanin biosynthesis. Our results demonstrated that four MYBs (Ej00002033, Ej00027194, Ej00029887 and newGene_15405) exhibited an expression profile that was positively correlated with anthocyanin concentration and the expression pattern of structural genes of anthocyanin biosynthesis pathway ( Figure 3 ). Phylogenetic analysis showed that Ej00002033 (EjMYB10) encodes a MYB that fell into the same clade with MYB10s from other rosaceous species and was closely related to apple MdMYB110a ( Figure 5A ; Supplementary Table S ). Multiple sequence alignment revealed EjMYB10 was highly homologous to MdMYB110a and a conserved R2R3 domain as well as the ‘ANDV’ and SG6 motifs, which are characteristic of anthocyanin-promoting MYBs ( Figure 5B ). However, the C-terminal of EjMYB10 protein is shorter than that of MdMYB10, PyMYB10, PsMYB10.1, and PsMYB10.2 ( Figure 5B ).

Transient expression in tobacco leaves was employed to validate the function of EjMYB10 in anthocyanin biosynthesis. Our results showed that faint red coloration was observed in leaves co-infiltrated with EjMYB10 and PsbHLH3 eight days after transformation. However, no red pigmentation was observed when EjMYB10 or PsbHLH3 was infiltrated alone ( Figure 6A ). Quantification of anthocyanins in tobacco leaves revealed that only a small amount of anthocyanin was accumulated in leaves infiltrated with EjMYB10 and PsbHLH3 ( Figure 6B ). In contrast, no anthocyanin was detected in leaves infiltrated with empty vector, EjMYB10 or PsbHLH3 ( Figure 6B ). These results suggest that EjMYB10 functions as a weak anthocyanin activator. To confirm this, the anthocyanin-promoting activity of EjMYB10 was further compared with strong activator PsMYB10.1 (Fang et al., 2021a) and weak activator PsMYB10.2 (Fang et al., 2021b) from plum. Coinfiltration of PsMYB10.1 with PsbHLH3 led to obvious red pigmentation at infiltration sites three days after transformation, and strong anthocyanin accumulation was observed eight days after transformation ( Figure 6C ). However, the infiltration of PsMYB10.2 and PsbHLH3 or EjMYB10 and PsbHLH3 resulted in only weak red pigmentation in the leaves ( Figure 6C ). Our results showed that the anthocyanin content in tobacco leaves infiltrated with EjMYB10 and PsbHLH3 was comparable to that in leaves infiltrated with PsMYB10.2 and PsbHLH3. However, it was significantly lower than the content in leaves infiltrated with PsMYB10.1 and PsbHLH3 ( Figure 6D ). Additionally, promoter structure analysis revealed that the promoters of the anthocyanin biosynthetic genes EjF3′H (Ej00025308), EjANS (newGene 13949) and EjUFGT (Ej00006885) contain multiple potential MYB binding sites ( Figure 6E ). Dual-luciferase assays were carried out to further verify whether EjMYB10 could interact with the promoters of these anthocyanin biosynthetic genes. Infiltration of EjMYB10 was able to enhance the transcriptional activity of EjF3′H, EjANS and EjUFGT ( Figures 6F–M ). These results suggest that EjMYB10 acts as a weak anthocyanin activator.