Red light-transmittance bagging promotes carotenoid accumulation of grapefruit during ripening

Carotenoids are lipid-soluble pigments vital for plant growth, photosynthesis, photoprotection, and development, and they provide colors and nutritional value to many fruits and vegetables. In humans, dietary carotenoids contribute to vitamin A and are linked to reduced risk of chronic diseases. Grapefruit (Citrus paradisi) is rich in carotenoids and economically important; red grapefruit is particularly valued for its color and flavor. Light is known to regulate carotenoid biosynthesis and degradation, yet the mechanisms by which light quality during on-tree development shapes carotenoid metabolism in grapefruit remain incompletely understood. This study investigates how altering incident light via light-transmittance bagging (red, blue, white vs. natural light) during ripening affects carotenoid accumulation and the underlying gene regulatory networks, aiming to improve fruit coloration and nutritional quality and to elucidate the transcriptional control of carotenoid metabolism in response to light.

Prior work shows that light irradiation modulates carotenoid biosynthesis and catabolism, influencing carotenoid concentration and composition in various fruits (e.g., tomato, pepper, kumquat, mandarin, watermelon, sweet orange). Light deprivation generally reduces carotenoid accumulation, whereas specific light qualities (e.g., red or blue LEDs) can enhance carotenoids by upregulating carotenogenic genes. In tomato, TFs such as PIF1, RIN, and FUL homologs regulate carotenoid pathway genes (PSY, Z-ISO, CRTISO, LCYB). In citrus, MADS-box TFs (e.g., CsMADS6/56) modulate PSY, PDS, LCYB1/2. In papaya, bHLH and NAC TFs influence lycopene and downstream carotenoids. However, the effects of light-transmittance bagging during on-tree grapefruit ripening and the associated transcriptional networks had not been fully elucidated.

Plant material: 'Houyat' grapefruit grown at the Citrus Research Germplasm Repository, Citrus Research Institute, Chinese Academy of Agricultural Sciences (Chongqing, China). Trees were of similar age and managed uniformly. Experimental design: Fruits were covered on-tree for 40 days with different light-transmitting plastic bags to alter solar spectra: red-light-transmitting (RL), blue-light-transmitting (BL; peak 478 nm), white-light-transmitting (WL), and no bagging as control (CK). Light transmission properties of bags were characterized (Supplementary Fig. 4). Sampling: Uniform fruits were collected at two ripening stages, 184 days after blossom (DAB184, maturation) and 220 days after blossom (DAB220, fully ripe). Each fruit constituted a biological replicate; three biological replicates per treatment per time point. After measuring basic parameters, juice vesicles were frozen in liquid nitrogen and stored at −80 °C. Phenotyping: Total soluble solids (TSS), titratable acidity (TA), and color index (CCI; calculated as C = (L*−a*/b*)) were measured using standard methods and calibrated instruments (HunterLab colorimeter; Aago Pro-110). Carotenoid extraction and HPLC: Carotenoids were extracted with methanol/acetone/n-hexane (50:25:5, v/v), filtered, and re-dissolved in methanol/acetonitrile (2:1, v/v). HPLC (Waters system) with a reversed-phase column (250 × 4.5 mm; 5 µm) and multistep gradient using methanol, ethanol, and aqueous phases was used for identification/quantification by comparison with standards; contents expressed as mg/g fresh weight. RNA-seq: Total RNA quality assessed by Agilent 2100 Bioanalyzer and NanoDrop. Libraries prepared (TruSeq DM RNA Library Prep Kit V2) for CK, BL, WL at both stages (and RL in comparative analyses), sequenced on MGISEQ-2000 (BGI). Reads were cleaned (SQXcheck v1.0, TrimGalore v0.6.4), aligned to reference (base accession AP009080) using HISAT2 v2.1.0, and annotated with plant transcriptome databases. Expression quantification and DEGs: Expression estimated as FPKM using RSEM. DEGs screened using a Poisson-based approach (BCG2 algorithm) with thresholds p ≤ 0.05 and |log2FC| ≥ 1. WGCNA: Genes with FPKM > 1 and high variance (12,801 unigenes; top 75%) were used to construct weighted gene co-expression networks (WGCNA v1.69+) in R (v4.1.17), relating modules to carotenoid phenotypes (p ≤ 0.05). Module networks visualized in Cytoscape v3.7.2. TF analysis: Candidate TFs within carotenoid-associated modules were identified; phylogenetic analyses were performed with MEGA7 (ClustalW alignment; neighbor-joining; 1000 bootstrap replicates). qRT-PCR: Selected carotenogenic genes and TFs were validated by qRT-PCR (Bio-Rad instrument; SYBR Premix Ex Taq II). Actin (Citrus sinensis act-s7: LOC102753790) was the reference; cycling: 95 °C 1 min; 40 cycles of 95 °C 20 s, 58 °C 20 s, 72 °C 30 s; 2−ΔΔCt for relative expression. Statistics: Three biological replicates; data shown as mean ± SD with individual points. One-way ANOVA with post-hoc SNK; correlations by Origin 2018.

• Red light-transmittance (RL) bagging significantly increased total carotenoid content by 62% compared with natural light (control) during grapefruit ripening. Specific carotenoids (β-carotene, ε-carotene, lycopene, cis-lycopene) increased under light-transmitting treatments, with RL showing the strongest promotion. - WGCNA identified 12 co-expression modules from 12,801 genes; the ‘blue’ module (4,832 genes; 37.7%) positively correlated with total carotenoid content (r = 0.69, p = 0.0002), while the ‘turquoise’ module negatively correlated (r = −0.68, p = 0.003). Neither module associated significantly with 9-cis-ABA (p > 0.05). - Transcriptomic analysis revealed RL-responsive differential expression of carotenogenic structural genes. Upregulated under RL: GGPPS2, PDS, Z-ISO, ZDS2/3, CRTISO3, CHYB; downregulated: CYP97A (β-ring hydroxylase), ZEP2, CCD1-2. - Candidate transcription factors involved in RL-mediated carotenoid regulation included bHLH128, NAC2-like/21/72, MYB-like, AGL11/AGL61, ERF023/062, WRKY20, and SBP-like-17/73. Many of these TFs showed coordinated expression with carotenogenic genes within the ‘blue’ (positive) and ‘turquoise’ (negative) modules. - qRT-PCR validation supported RNA-seq trends for 22 key genes, confirming RL-induced transcriptional modulation throughout the carotenoid pathway and supporting the proposed regulatory network.

The study addressed how light quality during on-tree ripening regulates carotenoid metabolism in grapefruit. RL-transmittance bagging robustly enhanced total and specific carotenoids, aligning with increases in color quality metrics. Systems-level transcriptomics (DEG and WGCNA) linked carotenoid phenotypes to distinct co-expression modules: a positively correlated ‘blue’ module and a negatively correlated ‘turquoise’ module. Within these, RL modulated core carotenoid biosynthetic genes (e.g., GGPPS2, PDS, Z-ISO, ZDS, CRTISO3) and repressed carotenoid degradation/epoxidation genes (ZEP2, CCD1-2), consistent with elevated carotenoid accumulation. The identification of TF families (WRKY, MYB, MADS, bHLH, NAC, ERF, SBP) as hubs suggests transcriptional control is a major driver of light-responsive carotenoid flux. The findings extend previous observations in other species by demonstrating that spectral modification via bagging in field conditions can steer carotenoid biosynthesis in grapefruit. Practically, RL-transmittance bagging emerges as a feasible strategy to improve fruit coloration and nutritional value. Conceptually, the proposed regulatory network integrates RL perception with TF-mediated control of carotenogenic genes, offering targets for breeding and postharvest interventions.

This work demonstrates that red light-transmittance bagging during on-tree ripening significantly enhances carotenoid accumulation in grapefruit, increasing total carotenoids by 62% versus natural light. Network and transcriptome analyses identified carotenoid-associated co-expression modules and candidate TFs (bHLH, NAC, MYB, MADS, ERF, WRKY, SBP) that likely coordinate RL-responsive regulation of key carotenogenic genes (GGPPS2, PDS, Z-ISO, ZDS, CRTISO3, CHYB) and suppressors of carotenoid accumulation (ZEP2, CCD1-2). The proposed model provides a mechanistic framework for light-mediated control of carotenoid metabolism in citrus and supports RL bagging as a practical approach to improve fruit coloration and nutritional quality. Future research should functionally validate candidate TFs and target genes (e.g., via gene editing or transgenics), dissect upstream photoreceptor signaling, and assess genotype- and environment-specific responses to optimize light-based cultivation practices.

• The study was conducted on a single grapefruit cultivar and location, which may limit generalizability across genotypes and environments. - Transcriptomic associations and co-expression do not establish causality; functional validation of candidate TFs and structural genes was not performed within this study. - Sample sizes were limited to three biological replicates per condition, which, while standard, may constrain detection of subtle effects. - Potential confounding effects of microclimate changes induced by bagging (temperature, humidity) were not explicitly quantified and could contribute to observed differences.