Analysis of volatile compound variations and screening of regulatory factors in the peel of Citrus reticulata ‘Chachi’

【Objective】The peel of Citrus reticulata‘Chachi’, a geographically protected medicinal resource in traditional Chinese medicine, harbors volatile organic compounds (VOCs) that underpin its therapeutic efficacy and aromatic traits. This study systematically characterized the stage-specific VOC dynamics across three critical developmental phases—Geqingpi (GQP, 45 days after flowering), Sihuaqingpi (SHQP, 105 days after flowering), and Chenpi (CRP, 195 days after flowering)—and deciphered the transcriptional regulatory network governing terpenoid biosynthesis.【Methods】Fresh peel were collected from four replicate trees at the Jiangmen City Xinhui District Forestry Science Research Institute with eight fruits per tree (two per cardinal direction) sampled at GQP (May), SHQP (July), and CRP (November). Samples were flash-frozen in liquid nitrogen and stored at -80 ℃. For VOC extraction, samples were ground to 100- mesh powder under liquid nitrogen conditions, and 0.1 g aliquots were extracted with 500 μL methyl tert-butyl ether (containing 0.02% methyl nonanoate as internal stan dard) and 500 μL deionized water via ice-bath ultrasonication (40 min at 4 ℃ ). After centrifugation (12 000×g, 10 min, 4 ℃), supernatants were filtered (0.22-μm PTFE membrane) and analyzed by GCMS (Agilent 8890-5977B) using a HP-5MS UI column (30 m × 0.25 mm × 0.25 μm). The oven program comprised: 40 ℃ held for 3 min, then ramped at 3 ℃ · min- 1 to 160 ℃ and held for 1 min, followed by ramping at 5 ℃·min-1 to 200 ℃ and holding for 1 min, and finally ramping at 8 ℃·min-1 to 240 ℃ and holding for 3 min, with He flow at 1.0 mL·min- 1 , split ratio 50∶1, and injector/ion source/ transfer line temperatures at 250/230/250 ℃. Compounds were identified via NIST17 library matching (similarity＞85%) and quantified using six-point calibration curves (R2 ＞0.988), with intra-day RSD＜ 5.2% and inter-day RSD＜7.8%. Transcriptome sequencing was performed on triplicate samples using Illumina HiSeq 2500 (150- bp paireD- end), generating 45.8- 52.6 million clean reads/sample (Q30＞ 90%, mapping rate 92.3% to C. clementina JZ v1.0). Library construction utilized the NEBNext® Ultra™ II RNA Library Prep Kit, and clean reads were aligned with HISAT2 v2.2.1.0. DEGs were identified by DESeq2 (|log2 FC|＞1, FDR＜0.05), yielding 4, 490 DEGs (CRP vs. GQP). Weighted gene co-expression network analysis (WGCNA) was conducted on 12, 986 genes (soft threshold β=12, scale-free R2 = 0.92), with module-trait correlations evaluated at |r|＞0.8, P＜0.01. Transcription factors (TFs) were predicted by co- expression (Pearson r＞0.8) with GPPS/HDS/HDR and cis- element analysis (PlantCARE), annotated via BLASTP against TAIR10 (E-value＜1e−10), and validated by qRT-PCR (primer efficiency 92%-105%, thermal cycle: 95 ℃ for 3 min, 40 cycles of 95 ℃/10 s, 60 ℃/30 s, melt curve 65-95 ℃ at 0.5 ℃ increments).【Results】GC-MS identified 30 VOCs, with 16 terpenoids constituting 53.2%-76.8% of total abundance. D-Limonene increased significantly from 26.70% ± 1.82% (GQP) to 58.20% ± 3.01% (CRP, P＜0.001), while γ-Terpinene declined from 45.10% ± 2.50% to 18.30 ± 1.42% (ANOVA, Tukey’s HSD, P＜0.01). The alkaloid Methyl N- methylanthranilate dropped from 4.30 ± 0.21% to 0.65% ± 0.08%, alongside stage-specific trends in α-Pinene (decreased from 6.10% to 1.10%), β-Pinene (decreased from 4.70% to 1.40%), and β-Myrcene (increased from 2.00% to 5.10%) (OPLSDA, VIP＞1.5, P＜2.3e−5, permutation test R2 X = 0.98, R2 Y = 0.99, Q2 = 0.97). PCA revealed clear stage separation along PC1 (68.3% variance) and PC2 (19.7% variance), with GQP samples clustering in the positive PC1 region. Transcriptome analysis identified 4, 490 DEGs, with WGCNA clustering genes into seven modules. The MEturquoise module (824 genes) strongly correlated with D-Limonene (r = 0.92, P = 6e−4) and γ-Terpinene (r = −0.89, P = 0.002). Three MEP pathway genes showed stagedependent expression: Geranyl pyrophosphate synthase (GPPS, Ciclev10008936m) upregulated 4.2- fold in CRP (FPKM 10 912.1), 1- Hydroxy- 2- methyl- 2- (E)- butenyl- 4- diphosphate synthase (HDS, Ciclev10027881m) downregulated 3.8-fold (FPKM 8 505.97), and (E)-4-Hydroxy-3-methylbut-2-enyl diphosphate reductase (HDR, Ciclev10029878m) downregulated 5.1- fold (FPKM 12, 843.88). HDS/ HDR expression correlated with γ-Terpinene (r = 0.91-0.94) and α-Pinene (r = 0.89-0.90), while GPPS linked to D-Limonene (r = 0.86, P＜0.001). A total of 37 transcription factors that regulate these three genes were identified, including 14 AP2/ERF (e.g., Ciclev10016982m homologous to AtERF12), 9 WRKY (e.g., Ciclev10007882m homologous to AtWRKY20), 6 MYB, and 4 bHLH. Three TFs— Ciclev10007882m (WRKY), Ciclev10015354m (BPC1), and Ciclev10001893m (ABREBP3)—co-regulated HDS/HDR, binding to W-box (TTGACC) and GCC-box (AGCCGCC) motifs in gene promoters. The interaction network analysis revealed a TF-gene regulatory network comprising 42 nodes and 117 edges, in which Ciclev10007882m exhibited the highest betweenness centrality (0.78).【Conclusion】We constructed a transcriptional regulatory network for terpene biosynthesis in the peel of C. reticulata ‘Chachi’. In brief, stage signals initiate TF activation (WRKY20/BPC1 upregulates GPPS/HDS in GQP; ERF represses HDS/HDR in CRP), which in turn modulates pathway gene expression, accounting for 82.3% of VOC variance (WGCNA module-trait correlation, P＜0.001). The GPPS/HDS/HDR module serves as the core regulatory unit for terpenoid accumulation. In terms of practical application, the identified WRKY20 (encoded by Ciclev10007882m) provides a target for developing functional molecular markers (such as KASP markers based on key SNPs/InDels in its promoter or coding region) to assist in screening breeding materials with high terpenoid content. Additionally, the six key biomarkers combined with near-infrared spectroscopy enable precise maturity assessment and authenticity identification of Chenpi, thereby supporting the standardization of the industrial production chain.