LI Sisi, YANG Rong, LIU Jiajia, HE Kaiming, LIU Sijia

【Objective】This research investigated the variation in tannin and other phenolic compounds in the kernel pellicle of different walnut (Juglans regia L.) cultivars, aiming to elucidate the mechanisms underlying their biosynthesis and accumulation as well as to provide a theoretical foundation for targeted improvement of quality traits related to phenolic composition in walnut.【Methods】Three walnut cultivars exhibiting distinct tannin levels were selected, including low-tannin Bingta No.1 (BT-1), medium-tannin Wen 185 (W-185), and high-tannin Zha 343 (Z-343). Fruits were collected from fiveyear or older, vigorously growing trees under consistent water and fertilizer management at the XPCC (Xinjiang Production and Construction Corps) walnut germplasm repository (40°23′ N, 80°03′ E). Sam-pling occurred at eight critical developmental stages: T1 (June 12), T2 (June 18), T3 (June 28), T4 (July 5), T5 (July 15), T6 (July 25), T7 (August 9), and T8 (August 24). Fresh seed coats were fixed in FAA for paraffin sectioning. Sections were dehydrated through graded ethanol solutions, cleared in xylene, embedded in paraffin, and stained with safranin-fast green. Tannin cells were specifically identified by the purple-red coloration resulting from the safranin-tannin reaction, observed and photographed using optical microscopy. Tannin spatial distribution during development was visualized via imprint staining. Fresh fruit cross-sections were pressed onto a piece of filter paper, which formed purple imprints due to oxidative property of tannins. Total tannin content was determined spectrophotometrically using Folin reagent with gallic acid as standard. Total phenolic content was assayed using the Folin-Ciocalteu method (gallic acid equivalent). Total flavonoid content was determined by the aluminum nitrate colorimetric method (rutin equivalent). Phenolic composition was analyzed by HPLC using an Agilent TC-C18 (2) column (250 mm × 4.6 mm, 5 μm) at 35 ℃. The mobile phase consisted of (A) 50% methanol and (B) 0.2% aqueous acetic acid, with a gradient elution (A: 8% to 40% from 0-40 min, then back to 8%). Flow rate was 0.8 mL·min-1 , and injection volume 10 μL. Detection was at 280 nm. Six phenolic acids (shikimic, gallic, caffeic, vanillic, chlorogenic, and p-coumaric) and four flavonoids (catechin, epicatechin, rutin, and myricetin) were quantified using external standards.【Results】The tannin content in the kernel pellicle of all three cultivars exhibited a pattern of“decline- rise- decline”during fruit development. Significant differences (P＜0.05) were observed in the timing of key fluctuation nodes and peak accumulation intensity. Peak concentrations were 352.75 mg · g-1 , 386.84 mg · g-1 , and 448.99 mg · g-1 for BT-1, W-185 and Z-343. BT-1 accumulated significantly less tannin than W-185 and Z-343 at peak. Imprint staining confirmed tannin deposition was localized specifically within the seed coat, with staining intensity following a“light-dark-light”trend with fruit maturation. Paraffin section revealed that tannin accumulation was primarily associated with specialized tannin cells surrounding the vascular bundles. These cells function as key sites for tannin synthesis, storage, and transport. Cultivars differed markedly in tannin cell abundance and activity. During peak accumulation, W-185 and Z-343 possessed numerous, metabolically active, densely packed tannin cells, whereas BT-1 had fewer such cells. This results agreed with the lower tannin synthesis rate and content observed in BT-1. HPLC analysis revealed shikimic acid as the predominant phenolic acid and epicatechin as the major flavonoid across cultivars, with distinct accumulation dynamics. BT-1 exhibited significantly lower levels of shikimic acid, gallic acid, p-coumaric acid, and epicatechin compared to W-185/Z-343; Shikimic, vanillic, and chlorogenic acids accumulated continuously, while caffeic acid declined steadily. In Z-343, chlorogenic acid content increased markedly at T5 stage (from 28.32 to 76.29 mg · g-1 ); rutin accumulated continuously, peaking at 10.54 mg · g-1 . In BT-1, a sharp transient increase in rutin was observed at T3 stage. Pearson correlation analysis revealed cultivar-specific regulatory networks linking tannins to other phenolics. In BT-1, tannin positively correlated with p-coumaric acid and epicatechin, but negatively correlated with rutin and total flavonoids; in W-185, tannin showed an extremely significant positive correlation with gallic acid (P＜0.01), but negative correlations with caffeic acid and total phenolics; in Z-343, tannin correlated positively only with total phenolics and caffeic acid, but negatively with most phenolic acids. These distinct correlation patterns indicate that variation in partitioning of phenylpropanoid metabolic flux, mediated by differential precursor supply (e.g., shikimic acid) and the balance of competitive metabolites (e.g., rutin vs tannin precursors), is a critical factor driving divergent tannin accumulation strategies among cultivars, and ultimately influences tannin biosynthetic efficiency.【Conclusion】Tannin accumulation in walnut kernel pellicle is co-regulated by cellular structural morphology and phenolic biosynthe-sis pathways. The low tannin phenotype of Bingta No.1 (BT-1) results from: (1) a reduced number of tannin cells, limiting synthesis capacity; (2) diminished accumulation of key precursors like shikimic acid; and (3) competitive metabolic flux towards rutin synthesis, potentially diverting substrates away from the tannin pathway. In contrast, cultivars Wen 185 (W-185) and Zha 343 (Z-343) achieve higher tannin accumulation through more abundant tannin cells and enhanced provision of precursors, such as chlorogenic acid in Z-343 and gallic acid in W-185. The identified differentiation in phenolic metabolism provides a biochemical basis for targeted regulation of tannin synthesis. As an excellent low-tannin germplasm, Bingta No.1 possesses unique metabolic characteristics valuable for dissecting the molecular mechanisms controlling tannin biosynthesis. Future research should employ multi- omics technologies to identify key regulatory genes within these pathways, enabling precision breeding or biotechnological approaches to modulate astringency and improve walnut quality.