Xenia effect of ‘Murcott’ tangor pollen on sugar-acid metabolism and phenolic compounds in Fuju

Abstract

The sweetness of citrus fruits is primarily determined by the concentrations and ratios of sugars and organic acids. The xenia effect, defined as the influence of pollen genotype on fruit phenotype, has been reported in several fruit crops and may affect traits such as fruit shape, sugar–acid balance, ripening time, and seed characteristics. This study aimed to evaluate the effects of Murcott tangor (Citrus reticulata × Citrus sinensis) pollen on sugar-acid metabolism and phenolic composition in Fuju fruit (Citrus reticulata ‘Tangerina’) through controlled cross-pollination. Self-pollinated Fuju fruits were used as controls. Changes in sugars, organic acids, phenolic compounds, and related metabolic gene expression patterns during fruit development were analyzed using metabolite profiling, and quantitative gene expression analysis. The results indicated that Murcott pollen-pollinated Fuju fruits exhibited altered sugar accumulation patterns during the late developmental stage (210–230 days after flowering), with sucrose content increasing by approximately 15% compared with self-pollinated Fuju at harvest. These changes were associated with altered expression patterns of sucrose metabolism-related genes, including sucrose synthase and neutral invertase. Citric acid content showed no significant differences among pollination treatments, suggesting that xenia-associated effects were more evident in sugar accumulation than organic acid accumulation under the conditions tested. These findings indicate that Murcott tangor pollen is associated with changes in sugar accumulation patterns and phenolic profiles in Fuju fruit, providing insights into the potential contribution of pollen genotype to citrus fruit quality development.

Highlights

• •Xenia effect increased sugar accumulation, enhancing Fuju sweetness at ripening.
• •Xenia altered organic acid metabolism enzymes but did not change citric acid levels.
• •Xenia influenced flavonoid accumulation, modulating secondary metabolite metabolism.

1. Introduction

Citrus is among the most widely cultivated fruit crops globally, with China representing one of the largest production regions (Ren et al., 2015). Fruit development in citrus involves dynamic sugar accumulation and organic acid metabolism, and the ratio of sugars to acids at maturity critically determines flavor quality (He et al., 2023; Qiu et al., 2022; Zhou et al., 2018). Improving the sugar–acid ratio is therefore a key goal in citrus production. Besides sugars and organic acids, phenolic compounds represent important secondary metabolites in citrus fruits and contribute to antioxidant capacity and characteristic fruit chemical profiles (Y. Chen et al., 2021; Mateus et al., 2024). However, whether pollen genotype influences phenolic accumulation during citrus fruit development remains poorly understood.

The xenia effect refers to the phenomenon where pollen genotype influences fruit phenotypic traits (including shape, sugar-acid ratio, and maturity period) through potential effects on seed-associated tissues and maternal tissues. In pollination studies of “Piarom” date palm (Phoenix dactylifera L.), the hormones (including indole-3-acetic acid, cytokinins, and gibberellic acid) of pollen showed a positive linear correlation with the hormones of fruits at different fruit growth stages (Shahsavar & Shahhosseini, 2021). Some studies also suggest that the transfer of signaling molecules during pollen tube growth affects fruit performance. During cross-pollination of different citrus cultivars, the number of pollen tubes varies significantly, which further affects seed and fruit characteristics (Alinezhad Jahromi et al., 2019). The xenia effect has been utilized in orchard production to enhance fruit yield and quality in certain crops, including some citrus varieties, through cross-pollination (Trueman et al., 2022). Pollination of grapefruit and ‘Majia’ pomelo with ‘Jinsha’ pomelo pollen resulted in significant differences in soluble sugar and amino acid content in fruits (Zhao et al., 2025). Similarly, Citrus mangshanensis pollination of Huanong Red pummelo has been shown to influence fruit aroma (Zhang et al., 2019).

Fuju, a citrus cultivar native to Fuzhou classified as Citrus reticulata ‘Tangerina’, shows improved fruit quality traits following pollination with Murcott tangor (Citrus reticulata × Citrus sinensis) pollen during late developmental stages. This cross-pollination induces significant changes in sugar and organic acid contents and phenolic compound levels, substantially improving the overall quality of Fuju fruits. However, xenia's regulatory mechanisms on Fuju sugar-acid and phenolic metabolism are unknown.

The aim of this study was to investigate the effects of Murcott tangor pollen on sugar, organic acid, and phenolic metabolism in Citrus cultivar Fuju during fruit development. We hypothesized that pollen source influences fruit metabolic profiles, with changes in enzyme-related gene expression patterns potentially contributing to these differences. By integrating metabolite analysis, and gene expression profiling, this study provides insights into the association between pollen genotype and citrus fruit quality formation.

2. Research methods

2.1. Experimental materials and treatment methods

Pollen Collection and Pollination Treatment: Flowers of Murcott tangor at the large bud stage were selected. Anthers were carefully excised, air-dried at room temperature for 24 h, and stored at 4 °C until use. At the initial stage of full bloom, flowers from three randomly selected Fuju trees were used for emasculation. The three trees were considered independent biological replicates throughout the experiment. To ensure balanced sampling, flowers within each tree were randomly assigned to different treatment groups: For each tree, 100 flowers were assigned for cross-pollination with Murcott tangor pollen (designated as xenia-affected Fuju, SF group), and another 100 flowers were assigned for self-pollination (designated as self-pollinated Fuju, FG group). Fruits from Murcott tangor trees were collected as the pollen donor cultivar control (M group). For cross-pollination, a sterile cotton swab was used to collect Murcott tangor pollen and evenly apply it to the stigma. Pollinated flowers were immediately bagged for isolation and labeled with identification tags.

Sample Treatment: Fruit samples were collected at eight developmental stages (120, 140, 160, 180, 200, 210, 220, and 230 days after flowering, DAF). At each developmental stage, three fruits per tree were randomly collected. Fruits with uniform size and without mechanical damage were selected from the sun-exposed middle canopy. The three fruits collected from each tree were pooled as one biological replicate, resulting in three biological replicates per treatment group. Pulp tissues were separated under sterile conditions, snap-frozen in liquid nitrogen, and stored at −80 °C until analysis.

Metabolite Profiling by UHPLC-DAD: Metabolite profiling was performed using a UHPLC-DAD system (Ultimate 3000 UHPLC, Thermo Fisher Scientific, Milan, Italy). Metabolites were extracted using 80% methanol at 4 °C, sonicated for 30 min, centrifuged at 12,000 ×g for 10 min, and the supernatants were collected. Quality control (QC) samples were prepared by pooling equal volumes from all samples and injected after every 10 samples to monitor instrument stability. Chromatographic separation used a reverse phase RP-18 column (100 mm × 2.1 mm, 2 μm; Thermo Fisher Scientific), column temperature 30 °C, flow rate 0.3 mL/min, and detection wavelengths were set at 280 and 340 nm. Peak extraction, alignment, and normalization were performed using Progenesis QI software (Nonlinear Dynamics, Newcastle upon Tyne, UK). Multivariate statistical analyses (PCA, PLS-DA, OPLS-DA) were carried out using EZinfo, with model validation by 200-permutation tests and cross-validation to control overfitting.

2.2. Analysis of sugar components

Sugar extraction and derivatization were based on (Zhou et al., 2018), with modifications. Briefly, 2 g of sample was homogenized with 12 mL of 80% ethanol in three portions and extracted in a water bath at 80 °C for 10 min. After centrifugation at 15,000 ×g for 8 min at 4 °C, the supernatant was collected. The precipitate was re-extracted once with 5 mL of 80% ethanol. Supernatants were combined and evaporated to dryness at 80 °C. The residue was redissolved in 4 mL ultrapure water, shaken for 20 min, centrifuged at 5000 ×g for 6 min, and the supernatant collected and stored at −40 °C. Sugars were analyzed using a Hitachi Lachrom NH2 column (5 μm) with a Hitachi L-3300 RI refractometer (Hitachi, Tokyo, Japan). Calibration curves were established with glucose, fructose, sucrose, and mannitol standards (R² > 0.998), with limits of detection 0.05–0.1 mg/mL.

2.3. Qualitative and quantitative analysis of organic acid components

The extraction and derivatization procedures for organic acids were modified from the method described by (Zhou et al., 2018). Briefly, 5 g of pulp was homogenized with 25 mL of 80% ethanol, extracted at 80 °C for 30 min, and centrifuged at 5000 ×g for 15 min. The precipitate was re-extracted twice with a total of 10 mL 80% ethanol. Supernatants were combined and adjusted to 50 mL. Five mL of supernatant was evaporated to dryness, redissolved in 5 mL ultrapure water, centrifuged, and filtered through 0.45 μm membranes. HPLC analysis was performed on a ROA-Organic Acid column (300 × 7.8 mm, Phenomenex, Torrance, CA, USA), mobile phase 0.005 M H₂SO₄, flow rate 0.5 mL/min, column temperature 30 °C, and detection at 210 nm. Citric, tartaric, malic, succinic, and oxalic acids were quantified using internal standards (oxalic acid‑d₂) and calibration curves (R² > 0.995).

2.4. Qualitative and quantitative analysis of fruit phenolic compounds

Sample Pretreatment: Separated pulp and peel samples were freeze-dried and passed through a 60-mesh sieve. 0.1 g of sample was extracted with 1 mL methanol at room temperature for 10 min, centrifuged at 9000 ×g for 15 min at 4 °C, repeated three times. Supernatants were combined, brought to final volume with methanol, and filtered through a 0.45 μm membrane.

Chromatographic Conditions: A C18 column (250 × 4.0 mm, Agilent Technologies, Santa Clara, CA, USA) was used with flow rate 0.8 mL/min, temperature 30 °C, and gradient elution (A: water/acetic acid 98:2, B: acetonitrile). Compound identification was performed via comparison with authentic standards and UV spectra. Quantification was based on calibration curves of reference compounds (R² > 0.995).

2.5. Quantitative real-time PCR analysis of genes related to sugar, organic acid, and secondary metabolism

The expression levels of genes related to glycolysis, the tricarboxylic acid cycle, and citric acid degradation pathways, as well as genes related to secondary metabolites, were examined using qRT-PCR. Total RNA was extracted using a plant total RNA extraction kit (Tiangen Biotech, Beijing, China), and RNA integrity was verified (RIN >7.0). First-strand cDNA was synthesized from 1 μg of total RNA using a reverse transcription kit (Promega, Madison, WI, USA). Quantitative real-time PCR was subsequently performed using TB Green Premix Ex Taq II (Takara Bio Inc., Dalian, China). The primers used are listed in Supplementary Table 1. Actin was used as the reference gene. Primer efficiency was 90–110%, confirmed by standard curves, with single-peak melt curves to ensure specificity. Relative quantification was calculated using the 2^–ΔΔCt method.

2.6. Statistical analysis

Two-way analysis of variance (ANOVA, pollination × developmental stage) was performed using a mixed-effects model with tree as a random factor. Post hoc multiple comparisons were corrected using Tukey's HSD test. Data were expressed as mean ± SD, and p < 0.05 was considered statistically significant. All analyses were conducted in R 4.3.2.

3. Results

3.1. Sugar composition dynamics in xenia-affected Fuju

The dynamics of sucrose, glucose, and fructose were examined in Murcott tangor (M), Fuju (FG), and xenia-affected Fuju (SF) during fruit development (Fig. 1), measured in days after flowering (DAF). Sugar accumulation followed a sigmoidal pattern and was divided into three developmental stages. Stage I (120–160 DAF): sugars increased gradually, reaching a minor peak at 160 DAF. Murcott tangor exhibited significantly higher sucrose and fructose concentrations than FG and SF at 120–140 DAF (p < 0.05), whereas glucose remained comparable across genotypes. Stage II (160–180 DAF): sugar content plateaued with minimal inter-cultivar differences. Stage III (200–230 DAF): sugar accumulation accelerated, especially in Murcott tangor after 210 DAF. At harvest (230 DAF), Murcott tangor showed 81.3%, 52.0%, and 83.9% higher sucrose, glucose, and fructose contents than FG, respectively. Notably, SF displayed higher sucrose content than FG at 230 DAF (p < 0.05), with a 15% increase (46.4 ± 2.5 mg/g vs. 40.3 ± 0.4 mg/g), indicating that Murcott tangor pollen was associated with increased sucrose accumulation during late fruit maturation.

Fig. 1.

Sugar content in fruits of Fuju (FG), Murcott tangor (M), and xenia-affected Fuju (SF) during development.

(A) Fructose, (B) glucose, and (C) sucrose contents from 120 to 230 days after flowering (DAF). Values are presented as mean ± standard deviation (SD) (n = 3). Statistical analysis was performed using two-way ANOVA followed by Tukey's HSD post hoc test. Different letters indicate significant differences among groups (p < 0.05): a indicates significant difference vs FG; b indicates significant differences vs both FG and SF; c indicates significant difference vs SF.

3.2. Citric acid metabolism patterns in xenia-affected Fuju

Citric acid, the main organic acid in citrus, peaked at 120 DAF and gradually decreased until harvest (230 DAF) (Fig. 2). The largest reduction occurred between 160 and 180 DAF. Small variations were observed among the three cultivars, but differences were not statistically significant, suggesting that pollination treatment was not associated with significant changes in citric acid content under the conditions tested.

Fig. 2.

Citric acid content in fruits of Fuju (FG), Murcott tangor (M), and xenia-affected Fuju (SF) from 120 to 230 days after flowering (DAF). Values are means ± SD (n = 3). No significant differences were detected among groups at any time point (two-way ANOVA, Tukey's HSD, p > 0.05).

3.3. Expression patterns of sugar metabolism-related genes in xenia-affected Fuju

Sucrose accumulation is associated with the coordinated expression of sucrose metabolism-related genes, including sucrose synthase (SS) and neutral invertase (NPI). During fruit development, SS and NPI mRNA expression exhibited distinct temporal and tissue-specific patterns (Fig. 3). In the peel, SS expression peaked at 180 DAF, with SF showing significantly higher levels than FG but lower than M (p < 0.05). In the pulp, SS expression reached its maximum at 200 DAF, where SF was lower than FG but higher than M (p < 0.05). For NPI, expression in the peel peaked at 180 DAF with no significant differences among groups. In the pulp, peak expression occurred at different developmental stages across cultivars, with FG showing the highest expression between 160 and 180 DAF. After the peak stages, transcript levels declined across all groups. These findings suggest that Murcott pollen-associated changes in sucrose accumulation coincided with altered expression patterns of SS and NPI during fruit development.

Fig. 3.

Expression patterns of sucrose metabolism-related genes in peel and pulp of Fuju (FG), Murcott tangor (M), and xenia-affected Fuju (SF) from 120 days to 230 days after flowering (DAF). (A) Sucrose Synthase (SS) expression in peel. (B) SS expression in pulp. (C) Neutral Invertase (NPI) expression in peel. (D) NPI expression in pulp. Values are presented as mean ± standard deviation (SD) (n = 3). Statistical analysis was performed using two-way ANOVA followed by Tukey's HSD post hoc test. Different letters indicate significant differences among groups (p < 0.05): a indicates significant difference vs FG; b indicates significant differences vs both FG and SF; c indicates significant difference vs SF.

3.4. Expression patterns of organic acid metabolism-related genes in xenia-affected Fuju

Citrate Synthase (CS) expression (Fig. 4A) peaked at 180 DAF in FG and at 200 DAF in SF, with SF significantly higher than FG and M at 200 DAF. Phosphoenolpyruvate Carboxylase (PEPC) expression (Fig. 4B) showed two peaks: at 160–180 DAF, FG was highest, whereas at 200 DAF, SF surpassed FG. Aconitase (ACO) expression (Fig. 4C) peaked at 200 DAF, with SF levels exceeding FG. ATP-Citrate Lyase (ATP-CL) expression (Fig. 4D) peaked at 180 DAF in FG, but SF exhibited higher expression at 200 DAF. After peak stages, expression of all genes declined in each genotype.

Fig. 4.

Expression patterns of organic acid metabolism-related genes in fruits of Fuju (FG), Murcott tangor (M), and xenia-affected Fuju (SF) during development. (A) Citrate Synthase (CS), (B) Phosphoenolpyruvate Carboxylase (PEPC), (C) Aconitase (ACO), and (D) ATP-Citrate Lyase (ATP-CL) mRNA expression levels from 120 to 230 days after flowering (DAF). Values are presented as mean ± standard deviation (SD) (n = 3). Statistical analysis was performed using two-way ANOVA followed by Tukey's HSD post hoc test. Different letters indicate significant differences among groups (p < 0.05): a indicates significant difference vs FG; b indicates significant differences vs both FG and SF; c indicates significant difference vs SF.

3.5. Flavonoid biosynthetic gene expression in xenia-affected Fuju

The variation of phenolic components in FG and SF was systematically quantified (Table 1). Two-way ANOVA revealed significant effects of developmental stage on most compounds, as well as significant genotype × stage interactions for multiple flavonoids. SF exhibited higher initial concentrations of several flavonoids at 120 DAF, including vitexin rhamnoside, hesperidin, and benzoic acid (p < 0.05). However, SF generally showed a more pronounced decline during development, resulting in similar or lower levels compared with FG at later stages. Hesperidin accounted for over 66% of total flavonoids, followed by vitexin rhamnoside (6.2–13%), synephrine (6.7–7.4%), and nobiletin (4.3–8%). Synephrine levels were consistently higher in FG than in SF at most time points, with significant differences observed at 180, 210, and 220 DAF (p < 0.05).

Table 1.

Variation of phenolic components in Fuju (FG) and xenia-affected Fuju (SF) during fruit development.

Compound	Group	120 DAF	180 DAF	200 DAF	210 DAF	220 DAF	230 DAF	p_Group	p_Time	p_G×T
Gallic acid	FG	3.02 ± 0.02	3.15 ± 0.78	ND	3.22 ± 0.27	3.15 ± 0.78	ND	0.718	<0.001	0.727
	SF	3.51 ± 0.19	3.05 ± 0.55	ND	3.46 ± 2.50	2.38 ± 0.49	ND
Synephrine	FG	2078 ± 68	1379 ± 77	2057 ± 358	1730 ± 221	1379 ± 77	1120 ± 18	0.262	<0.001	<0.001
	SF	2148 ± 180	1859 ± 90a	1205 ± 52	1268 ± 45a	1297 ± 35	1502 ± 159
Chlorogenic acid	FG	24.4 ± 21.1	33.0 ± 3.7	26.3 ± 3.2	34.1 ± 1.2	33.0 ± 3.7	34.5 ± 1.1	0.783	0.408	0.042
	SF	40.9 ± 1.6	35.6 ± 18.3	8.9 ± 15.4	29.5 ± 4.1	22.3 ± 19.3a	32.8 ± 1.2
Protocatechuic acid	FG	117 ± 1	130 ± 11	105 ± 10	52.4 ± 44.5	130 ± 11	131 ± 13	0.145	<0.001	<0.001
	SF	162 ± 13a	123 ± 13	166 ± 43	103 ± 8a	94.2 ± 0.5a	82.6 ± 9.4a
p-Coumaric acid	FG	39.0 ± 2.0	29.7 ± 1.9	48.9 ± 6.7	46.4 ± 2.5	29.7 ± 1.9	21.1 ± 25.8	0.893	0.03	0.402
	SF	38.3 ± 5.4	27.5 ± 4.3	39.7 ± 8.3a	48.6 ± 5.6	26.0 ± 2.8	46.4 ± 4.0
Saintiaoside	FG	46.2 ± 11.1	25.9 ± 11.3	29.2 ± 15.8	7.58 ± 3.05	25.9 ± 11.3	18.6 ± 1.4	0.043	<0.001	0.001
	SF	58.9 ± 5.6a	19.4 ± 5.8	14.0 ± 3.5	13.4 ± 6.7	10.3 ± 9.5	13.1 ± 0.5a
Vitexin rhamnoside	FG	2984 ± 343	2763 ± 307	2332 ± 266	2267 ± 161	2763 ± 307	1456 ± 73	0.001	<0.001	<0.001
	SF	4018 ± 266a	2580 ± 255	3361 ± 480a	1928 ± 11	2247 ± 102	2459 ± 241a
Rutin	FG	346 ± 98	355 ± 32	173 ± 8	273 ± 50	355 ± 32	241 ± 114	0.001	<0.001	<0.001
	SF	23.8 ± 0.5a	149 ± 137a	153 ± 32	195 ± 23	317 ± 36	270 ± 12
Benzoic acid	FG	234 ± 24	277 ± 15	370 ± 15	228 ± 4	277 ± 15	361 ± 62	0.617	0.045	0.355
	SF	379 ± 3a	234 ± 20a	427 ± 7a	264 ± 6a	344 ± 254	287 ± 32
Ruronarin	FG	428 ± 2	361 ± 16	551 ± 24	254 ± 15	361 ± 16	376 ± 33	0.085	<0.001	<0.001
	SF	650 ± 47	461 ± 62a	353 ± 54a	253 ± 3	365 ± 70	379 ± 15
Naringin	FG	33.0 ± 6.3	1.63 ± 1.03	1.68 ± 0.35	23.2 ± 1.1	1.63 ± 1.03	20.7 ± 0.3	0.018	<0.001	<0.001
	SF	52.5 ± 5.2a	24.7 ± 23.3	32.2 ± 6.8a	1.81 ± 1.57a	17.6 ± 15.4a	32.2 ± 1.6a
Hesperidin	FG	24908 ± 471	21583 ± 419	18890 ± 18	16708 ± 646	21583 ± 419	17297 ± 1204	0.004	<0.001	<0.001
	SF	28424 ± 605a	20890 ± 2171	17213 ± 534	15260 ± 104a	13178 ± 92a	16537 ± 698
Diosmin	FG	153 ± 1	147 ± 11	114 ± 10	146 ± 10	147 ± 11	127 ± 7	0.003	<0.001	<0.001
	SF	ND	103 ± 20	92.6 ± 82.7	153 ± 4	38.1 ± 33.1a	103 ± 21
Neohesperidin	FG	119 ± 8	42.4 ± 61.8	30.6 ± 3.9	144 ± 61	42.4 ± 61.8	8.08 ± 1.97	0.356	0.101	0.008
	SF	15.7 ± 2.3a	117 ± 59	77.6 ± 101	21.9 ± 12.2	106 ± 67	38.1 ± 8.7a
Quercetin	FG	15.0 ± 0.2	4.97 ± 2.77	3.42 ± 2.36	3.20 ± 0.96	4.97 ± 2.77	64.1 ± 6.2	0.001	<0.001	<0.001
	SF	5.35 ± 0.47a	96.8 ± 7.4a	30.7 ± 51.6	7.62 ± 5.38	1.47 ± 0.94	2.60 ± 1.69a
Kaempferol	FG	59.8 ± 2.3	1.41 ± 0.39	2.88 ± 0.99	12.8 ± 13.6	1.41 ± 0.39	11.5 ± 14.3	0.007	<0.001	0.001
	SF	44.3 ± 2.8	42.6 ± 13.5a	31.1 ± 26.0a	11.7 ± 18.9	12.8 ± 10.7a	24.6 ± 1.8a
Naringenin	FG	12.7 ± 2.4	13.7 ± 3.7	16.5 ± 1.9	12.6 ± 3.1	13.7 ± 3.7	16.5 ± 10.2	<0.001	0.103	0.081
	SF	8.67 ± 0.26	9.23 ± 2.05	7.35 ± 5.10a	4.70 ± 3.43a	6.97 ± 0.82a	5.89 ± 0.56
Nobiletin	FG	1873 ± 45	1661 ± 34	1396 ± 21	1980 ± 78	1661 ± 34	1299 ± 97	<0.001	<0.001	<0.001
	SF	1647 ± 55a	1313 ± 183	1684 ± 25a	1306 ± 11a	999 ± 15a	1459 ± 48a
Acacetin	FG	46.7 ± 1.2	61.1 ± 7.4	49.4 ± 1.0	46.8 ± 1.8	61.1 ± 7.4	68.3 ± 4.9	0.102	0.065	0.016
	SF	59.2 ± 14.2	42.6 ± 6.0	40.2 ± 12.2	34.3 ± 0.3a	54.6 ± 3.3	42.6 ± 4.3a
Tangeretin	FG	805 ± 28	880 ± 39	844 ± 13	875 ± 23	880 ± 39	619 ± 22	0.045	<0.001	<0.001
	SF	854 ± 151	674 ± 62a	930 ± 36a	668 ± 9a	533 ± 6a	740 ± 34a

Values are means ± SD (n = 3). Two-way ANOVA with Tukey's HSD post-hoc test was applied to evaluate the effects of group (FG vs. SF), time (DAF), and their interaction. p-values for group (p_Group), time (p_Time), and group × time interaction (p_G × T) are shown. Superscript a within the same row indicates a significant difference between SF and FG at the corresponding time point (p < 0.05). DAF: days after flowering.

To further investigate potential molecular associations, we analyzed the expression patterns of chalcone synthase (CHS) and chalcone isomerase (CHI) (Fig. 5). In the peel, CHS expression showed stage-specific peaks, with M exhibiting the highest level at 120 DAF (p < 0.05), while FG showed the highest expression at 200 DAF. In the pulp, CHS expression peaked in FG at 120 DAF and in SF at 160 DAF. At 220–230 DAF, SF consistently showed the highest CHS expression in both peel and pulp. CHI expression in the peel of SF remained higher than FG at later stages (200–230 DAF). Similarly, in the pulp, SF showed significantly higher CHI expression than FG at 210 DAF (p < 0.05).

Fig. 5.

Expression patterns of flavonoid biosynthesis-related genes in peel and pulp of Fuju (FG), Murcott tangor (M), and xenia-affected Fuju (SF) from 120 days to 230 days after flowering (DAF). (A) Chalcone synthase (CHS) expression in peel. (B) CHS expression in pulp. (C) Chalcone isomerase (CHI) expression in peel. (D) CHI expression in pulp. Values are presented as mean ± standard deviation (SD) (n = 3). Statistical analysis was performed using two-way ANOVA followed by Tukey's HSD post hoc test. Different letters indicate significant differences among groups (p < 0.05): a indicates significant difference vs FG; b indicates significant differences vs both FG and SF; c indicates significant difference vs SF.

4. Discussion

The xenia effect refers to the phenomenon in which the pollen genotype directly influences the development and phenotypic characteristics of seeds and fruits. This effect can modify not only seed size and number but also fruit set, ripening period, and flavor (Kitagawa, 2024). Fuju, a citrus cultivar native to Fujian Province, has recently suffered from issues such as small fruit size and sour taste due to varietal degradation. In this study, Fuju pollinated with Murcott tangor was associated with higher sucrose and fructose levels during the ripening stage (210–230 DAF), indicating that Murcott tangor pollen-associated changes were mainly observed during the ripening stage, particularly in sugar accumulation patterns between 210 and 230 DAF. Sucrose transported from leaves is metabolized in the fruit, primarily regulated by SS and NPI. SS catalyzes both sucrose synthesis and cleavage, whereas NPI mediates sucrose hydrolysis into glucose and fructose (Li et al., 2023). The observed changes in sucrose content coincided with altered expression patterns of SS and NPI, suggesting that these genes may participate in pollen-associated differences in sugar metabolism.

Although xenia did not significantly alter citric acid content, gene expression analysis revealed stage-specific patterns. Citric acid accumulation involves CS, PEPC, ACO, and other enzymes (Fan et al., 2022). PEPC expression in xenia-affected Fuju mirrored that of paternal Murcott tangor during ripening, while ACO expression shifted from maternal to paternal patterns after 200 DAF. The expression patterns of PEPC and ACO differed among pollination treatments during late fruit development, suggesting that Murcott pollen may be associated with temporal variation in organic acid metabolism-related pathways.

During 120 to 240 days after flowering (DAF), the content of most phenolic compounds (e.g., gallic acid, chlorogenic acid, hesperidin) in the two citrus cultivars FG and SF showed an overall downward trend, with only some substances fluctuating transiently in the mid-term (210–230 DAF). In SF, the levels of ferulic acid and naringin were relatively higher in SF than in FG, while the content of some phenolics (e.g., p-coumaric acid, santinatoside) exceeded that in FG. These phenolics are known to have antioxidant and other bioactivities in vitro (Bajgai et al., 2024; F. Chen et al., 2024; Pyrzynska, 2022), though the present study did not directly assess health effects. These differences suggest that pollen source may influence secondary metabolite profiles, which could have implications for the functional quality of citrus fruits.

Once desirable traits are achieved via xenia, maintaining them in subsequent generations remains challenging. At present, artificial pollination is mostly limited to laboratory research or production of a small number of high-quality fruits. However, pollen xenia is an important research tool, which can produce different fruit traits under the same genetic background. This is helpful to reveal the formation mechanism of fruit traits and screen key molecular targets. In the future, molecular breeding can be used to cultivate citrus varieties with excellent agronomic traits.

The current study targeted key genes involved in sugar and organic acid metabolism, providing a preliminary understanding of the molecular basis of xenia in Fuju. However, the scope of gene expression analysis was limited, and the regulatory network remains incompletely characterized. Future research could further elucidate the genetic regulatory network of xenia-mediated sugar and acid accumulation through transcriptomics, bulked segregant analysis (BSA), and hormone assays, identifying key genes to provide insights into the metabolic basis of Fuju fruit quality variation. Overall, Murcott tangor pollen was associated with altered sugar accumulation patterns and differences in selected phenolic profiles in Fuju fruit, particularly during late fruit development. These findings provide preliminary evidence that pollen genotype may contribute to metabolic variation in citrus fruit, while the underlying mechanisms require further investigation.

CRediT authorship contribution statement

Yuan Chen: Writing – original draft, Methodology, Conceptualization. Zhiqiang Lin: Writing – review & editing, Resources. Wenquan Yu: Writing – review & editing, Resources, Project administration, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data that has been used is confidential.