# Source paper archive - Source title: Molecular and phytochemical investigation of medicinal Rhododendron spp. in the Qinghai-Xizang Plateau: the botanical sources of the traditional Tibetan medicine "Tale" - Author(s): Liu, Q., Xie, S., Huang, X., Liu, X., Wei, Y., Zhang, Z., Zhang, X., Qin, W., Liu, X. - Source landing URL: https://www.biorxiv.org/content/10.64898/2026.11.05.764495v1 - PDF URL: https://www.biorxiv.org/content/10.64898/2026.11.05.764495v1.full.pdf - PDF SHA-256: 23ecab32fa4f567f893efb5ab88eadb88c41a0eaae2faea2280205de315759cd - Converter: pymupdf4llm 0.0.27 - Limits: Automated PDF text extraction; reading order, equations, tables and formatting may be incomplete or incorrect. Images are omitted, and no OCR was performed. Markup and links are neutralized for safe downloading. Consult the original PDF for the authoritative version. --- [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 1 Molecular and phytochemical investigation of medicinal _Rhododendron_ spp. in the 2 Qinghai–Xizang Plateau: the botanical sources of the traditional Tibetan medicine “Tale” 3 4 Qundong Liu [1,2], Siyuan Xie [1], Xiaoxue Huang [1], Xue Liu [1], Yongsheng Wei [1], Zhiwei 5 Zhang [1], Xiaomei Zhang [1], Weihan Qin [1,3,*], Xiang Liu [1,2,*] 6 7 1 Chongqing Key Laboratory for the Utilization and Evaluation of Special Chinese Materia 8 Medica Resources, Chongqing Academy of Chinese Materia Medica, Chongqing, China 9 2 Daba Mountain Medicinal Animals and Plants Chongqing Observation and Research 10 Station, Chengkou, China 11 3 Key Lab. of Tibetan Medicine Resources Conservation and Utilization of Tibet 12 Autonomous Region, Tibet Agriculture & Animal Husbandry University, Nyingchi, China; 13 14 *Corresponding author 15 E-mail: zysliux@ 163. com (X.L.), qinweihan8881@163.com (W.Q.) 16 17 18 19 1 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) # 20 Abstract 21 The Qinghai–Xizang Plateau in China is a hotspot for _Rhododendron_ species, with 22 approximately 400 species distributed across the region. Various _Rhododendron_ species in 23 this area have been used medicinally in traditional Tibetan medicine for a long time and 24 are collectively referred to as “Tale.” Furthermore, “Tale” is classified as “Talegabao” or 25 “Talenabao,” depending on the species. These two types of medicine have similar but 26 distinct properties, and “Talenabao” is usually used as an alternative to “Talegabao.” 27 However, differences in their phytochemical compositions and methods for discriminating 28 between them have not been examined. To solve these problems limited the use of “Tale” 29 medicines, 46 plant specimens belonging to nine _Rhododendron_ species were collected 30 from the Qinghai–Xizang Plateau. Molecular methods based on the internal transcribed 31 spacer 2 (ITS2) region of nuclear DNA and _psb_ A– _trn_ H regions of chloroplast DNA were 32 applied to these samples. In addition, 10 flavonoid compounds were quantitatively 33 determined by HPLC. The results indicated _psb_ A– _trn_ H region successfully distinguished 34 the species used as “Talegabao” or “Talenabao,” whereas ITS2 failed to distinguish them 35 because of its high degree of conservation. All 10 compounds were detected in all samples. 36 However, the species used as “Talegabao” had more than twice the total content of the 10 37 flavonoid compounds observed in the species used as “Talenabao.” Moreover, differences 38 in the contents of certain compounds among species were detected, indicating potential 39 chemical markers for the quality evaluation of “Tale” medicines. “Talegabao” is the first 40 choice among “Tale” medicines. “Talenabao” can be used as an alternative to “Talegabao”; 41 however, the appropriate dose requires further discussion in practical use. The _psb_ A– _trn_ H 2 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 42 region is recommended as a DNA barcode for distinguishing between these two types of 43 medicines. 44 45 Keywords: _Rhododendron_ ; _psb_ A– _trn_ H; molecular identification; flavonoid; active 46 compounds; traditional Tibetan medicine 47 48 49 50 51 52 53 54 55 56 57 58 59 3 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 60 # 61 Introduction 62 The genus _Rhododendron_ (Ericaceae) comprises approximately 1,000 species 63 distributed widely throughout the Northern Hemisphere, particularly in Asia [1]. 64 _Rhododendron_ species are known not only for their ornamental flowers but also for their 65 traditional medicinal uses in many countries [2]. The Qinghai–Xizang Plateau in China is 66 notable for its rich and diverse flora, with approximately 400 _Rhododendron_ species 67 distributed throughout the region, making it a major center of _Rhododendron_ diversity [3]. 68 Owing to their wide distribution and species richness in this area, various _Rhododendron_ 69 species have a long history of medicinal use in traditional Chinese, Tibetan, and Mongolian 70 medicine, among others [4,5]. 71 In traditional Tibetan medicine, _Rhododendron_ species play an important role, with 72 more than 80 species used medicinally. These medicines are called “Dama” ( སྟག་མ ) or “Tale” 73 ( ད་ལིས། ), also known as “Dali,” “Tali,” or “Balu” [4,6]. “Dama” consists of the dried flowers, 74 leaves, and seeds of certain evergreen species with large leaves, including _R. przewalskii_ 75 Maxim., _R. arboreum_ var. _roseum_ Lindl., _R. pingianum_ Fang, _R. coryanum_ Tagg & Forrest, 76 _R. aganniphum_ Balf. f. & Kingdon-Ward, _R. phaeochrysum_ var. _agglutinatum_ (Balf. f. & 77 Forrest) D. F. Chamb., and _R. principis_ var. _vellereum_ Hutch. ex Tagg, which mostly 78 belong to Section _Ponticum_ . “Tale” consists of the dried flowers, leaves, and twigs of 79 certain evergreen species with small leaves and scales on the surface of leaves and twigs. 80 In practice, “Tale” is classified as “White Tale” and “Black Tale.” “White Tale,” locally 4 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 81 called “Talegabao” ( ད་ལིས་དཀར་པོ། ), is characterized by plants with greenish-yellow upper leaf 82 surfaces and yellowish or light-brown lower surfaces, grayish-white branches and twigs, 83 and white, yellowish, or light-pink corollas. More than 17 species are used as “Talegabao,” 84 including _R. anthopogonoides_ Maxim., _R. primuliflorum_ Bur. & Franch., _R. cephalanthum_ 85 Franch., _R. hypenanthum_ Balf.f., _R. anthopogon_ D.Don, _R. mainlingense_ S.H.Huang & 86 R.C.Fang, and _R. micranthum_ Turcz., which mostly belong to Section _Pogonanthum_ . 87 “Black Tale,” locally called “Talenabao” ( ད་ལིས་ནག་པོ། ), is characterized by plants with dark 88 green or gray-green leaves, gray to black branches and twigs, and red, purple, or bluish 89 violet corollas. More than 20 species are used as “Talenabao,” including _R. capitatum_ 90 Maxim., _R. nivale_ Hook.f., _R. thymifolium_ Maxim., _R. intricatum_ Franch., and _R. nitidulum_ 91 Rehder & E.H. Wilson, which mostly belong to the _Rhododendron_ subsection _Lapponica_ 92 [4]. 93 In clinical practice, the two types of “Tale” medicines have similar but somewhat 94 distinct effects. “Talegabao” is described as having effects that clear heat and reduce 95 inflammation; relieve cough and asthma; strengthen the stomach and body, combat aging, 96 and treat disorders caused by rLung (wind), chibpa (bile), and pechen (phlegm), as well as 97 lung diseases, laryngitis with hoarseness, asthma attributed to environmental adaptation 98 issues, bronchitis, emphysema, spleen and stomach deficiency with coldness, indigestion, 99 gastric ptosis, gastric dilation, gastric cancer, hepatosplenomegaly, and edema. It is also 100 applied externally to reduce inflammation and swelling. “Talenabao” is described as 101 having effects that clear heat and reduce inflammation, relieve cough and asthma, 102 strengthen the stomach, disperse swelling, nourish the kidneys and enhance vitality, and 103 combat aging; it is also used to treat lung pain and laryngitis [4]. In practice, local medical 5 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 104 workers consider “Talegabao” to have a better effect and prefer to collect it, particularly _R._ 105 _anthopogonoides_ and _R. primuliflorum_, which are usually the first choice. “Talenabao” is 106 usually used as an alternative. 107 Multiple flavonoid compounds have been isolated from _R. anthopogonoides_, including 108 quercetin, isoquercetin, avicularin, and dihydroquercetin [7]. In addition, owing to the 109 strong scent of these plants, compounds in their essential oils have attracted attention [8]. 110 Pharmacological studies have shown that _R. anthopogonoides_ extracts have expectorant, 111 antitussive, antiasthmatic [4], bactericidal [9], anti-inflammatory [10], smooth muscle– 112 relaxing, and sedative effects. In addition, the essential oil has shown insecticidal activity, 113 suggesting its potential as a biological insecticide [11,12]. However, few studies have 114 reported the chemical composition of other species used as “Tale.” The chemical basis of 115 “Tale” medicines and the differences between “Talegabao” and “Talenabao” remain 116 unclear. Owing to the broad geographic distribution of traditional Tibetan medicine and 117 the use of multiple _Rhododendron_ species, different regions may prefer different species. 118 For example, both _R. anthopogonoides_ and _R. primuliflorum_ are used as “Talegabao”, 119 however, _R. anthopogonoides_ mostly distributed in Qinghai, Gansu provinces and _R._ 120 _primuliflorum_ distributed in Xizang, Yunnan and Sichuan provinces, which leading to 121 confusion regarding the botanical sources of the drugs (Fig 1). Moreover, standards for 122 “Tale” medicines are lacking. Accurate identification and quality assessment have not been 123 adequately established. The wide distribution and morphological similarities of 124 _Rhododendron_ species also complicate their classification, with the taxonomic status of 125 some species remaining debated and nomenclature varying among regions. This situation 126 can lead to confusion regarding the use of “Tale” crude drugs. 6 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 127 128 **Fig 1.** _**Rhododendron**_ **plant specimens and “Tale” medicines.** a, _R. capitatum_ ; b, _R._ 129 _nivale_ ; c, _R. thymifolium_ ; d, _R. intricatum_ ; e, _R. nitidulum_ ; f, locations of _Rhododendron_ 130 plant specimens; g, medicines of “Talenabao”; h, medicines of “Talegabao”; i, _R._ 131 _trichostomum_ ; j, _R. anthopogonoides_ ; k, _R. primuliflorum_ ; l, _R. cephalanthum_ . 132 133 _Rhododendron_ is one of the most taxonomically challenging and species-rich genera 134 for species identification and phylogenetic inference. Molecular methods based on nuclear 135 and chloroplast DNA have been evaluated for discriminating among _Rhododendron_ 136 species. Liu et al. used the internal transcribed spacer 2 (ITS2) region to classify 10 137 _Rhododendron_ species and found that ITS2 sequences successfully discriminated among 138 them [13]. Yan et al. evaluated four DNA barcodes using 531 samples representing 173 139 _Rhododendron_ species. Among the four barcodes, the _psb_ A– _trn_ H region provided the 140 highest species resolution [14]. These findings provided a basis for developing molecular 141 methods to identify _Rhododendron_ species used as “Tale” medicines. 142 In this study, we aimed to discriminate among different “Tale” species and clarify the 143 phytochemical differences between “Tale” medicines. Forty-six plant specimens and nine 144 crude drug samples belonging to nine medicinally used _Rhododendron_ species were 145 collected from the Qinghai–Xizang Plateau for investigation. Molecular methods based on 146 the ITS2 region of nuclear DNA and the _psb_ A– _trn_ H region of chloroplast DNA were 147 applied to the plant specimens. In addition, 10 flavonoid compounds, including 148 dihydromyricetin, hyperoside, isoquercitrin, taxifolin, guaijaverin, avicularin, astragalin, 7 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 149 quercitrin, quercetin, and naringenin, were quantified using high-performance liquid 150 chromatography (HPLC). Variations in compound composition were characterized using 151 multivariate statistical analysis, specifically orthogonal partial least squares discriminant 152 analysis (OPLS-DA). # 153 Materials and Methods 154 ## Plant specimens and drug samples 155 Corolla, leaves, and twigs of 46 plant specimens representing nine _Rhododendron_ 156 species, including _R. anthopogonoides_, _R. primuliflorum_, _R. nivale_, _R. thymifolium_, _R._ 157 _capitatum_, _R. trichostomum_, _R. cephalanthum_, _R. intricatum_, and _R. nitidulum_, were 158 collected from the Qinghai–Xizang Plateau, China (Table 1; Fig 1). In addition, nine crude 159 drug samples were purchased from drug markets in Lhasa, Chengdu, and Xining, China 160 (Table 2). The plant specimens and crude drug samples were morphologically 161 authenticated by the corresponding author. All samples were sealed in plastic bags and 162 stored at room temperature (15°C–25°C) away from sunlight. Voucher specimens were 163 deposited at the Chongqing Academy of Chinese Materia Medica in Chongqing, China. 164 Table 1 Plant specimens of _Rhododendron_ Voucher Section Scientific name Collected from No. RP1 Beishan, Huzhu, Qinghai _Rhododendron_ Sect. _Pogonanthum_ (“Talegabao”) _R._ _anthopogonoides_ RP2 Beishan, Huzhu, Qinghai RP3 Shierpan, Yuanfu, Bazha, Huzhu, Qinghai RP4 Wanglvdaban, Xianmi, Menyuan, Qinghai 8 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) RP5 Qunjialinchang, Huangzhong, Qinghai RP6 Jianglalinchang, Guide, Qinghai RP7 Dalijiashan, Xunhua, Qinghai RP8 Linxia, Gansu RP9 Gaba, Tanggu, Linzhou, Xizang RP10 Pangduo, Linzhou, Xizang RP11 Songpan, Linzhou, Xizang RP12 Linzhou, Xizang _R. primuliflorum_ RP13 Habaxueshan, Sanba, Xianggelila, Yunnan RP14 Pujinnima, Deqin, Yunnan RP15 Chitu, Daocheng, Sichuan RP16 Yading airport, Daocheng, Sichuan _R. cephalanthum_ RP17 Echushan, Geka, Daocheng, Sichuan _R. trichostomum_ RP18 Zheduoshan, Kangding, Sichuan RL1 Garang, Guide, Qinghai RL2 Jianglalinchang, Guide, Qinghai RL3 Beishan, Huzhu, Qinghai RL4 Qunjialinchang, Huangzhong, Qinghai _R. capitatum_ RL5 Dalijiashan, Xunhua, Qinghai RL6 Linxia, Qinghai _Rhododendron_ Sect. _Rhododendron_ Subsect. _Lapponica_ (“Talenabao”) _R. nivale_ RL7 Wutailing, Tianzhu, Gansu RL8 Xueshanliang, Songpan, Sichuan RL9 Youxi, Jitang, chaya, Changdu, Xizang RL10 Basu, Changdu, Xizang RL11 Leiwuqi, Changdu, Xizang RL12 Shangka, Leiwuqi, Changdu, Xizang Qialashan, Songpan, Linzhou, Lhasa, RL13 Xizang Baga, Wumatang, Dangxiong, Lhasa, RL14 Xizang 9 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) Lagen, Gongtang, Dangxiong, Lhasa, RL15 Xizang RL16 Sejilashan, Nyingchi, Xizang RL17 Huya, Pingwu, Sichuan RL18 Xinduqiao, Kangding, Sichuan RL19 Gedaliangzi, Danba, Sichuan RL20 Balangshan, Xiaojin, Sichuan RL21 Rangkou, Hongyuan, Sichuan RL22 Nangqian, Yushu, Qinghai RL23 Beishan, Huzhu, Qinghai RL24 Garang, Guide, Qinghai _R. thymifolium_ RL25 Qunjialinchang, Huangzhong, Qinghai RL26 Linxia, Gansu _R. intricatum_ RL27 Xinduqiao, Kangding, Sichuan _R. nitidulum_ RL28 Bamei, Daofu, Sichuan 165 166 Table 2 Crude drug samples of “Tale” Local Local Vouche Derived from name r No. Market r No. _R. anthopogonoides_ _R. primulaeflorum_ D1 Chengdu, Sichuan D2 Xining, Qinghai D3 Xining, Qinghai D4 Xizang D5 Xizang D6 Xizang D7 Lhasa, Xizang D8 Xizang D9 Zhangshu Market, Jiangxi 10 167 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) ## 168 Isolation of total DNA 169 Total DNA was extracted from 30 mg of dried leaf samples using a Plant DNA 170 Isolation Kit (FOREGENE Co., Ltd., Chengdu, China) according to the manufacturer’s 171 instructions, with some modifications. A 1-μL aliquot of the extraction solution from each 172 sample was examined using 1.0% agarose gel electrophoresis. ## 173 PCR amplification and sequence analysis of ITS2 and psb A-– trn H 174 ## regions 175 Thirty-eight plant specimens representing nine species were selected for ITS2 176 sequence analysis. All 46 plant specimens representing the nine species were subjected to 177 _psb_ A– _trn_ H sequence analysis. The ITS2 and _psb_ A– _trn_ H regions were amplified by PCR 178 using two pairs of primers: ITS2_F (5′-ATG CGA TAC TTG GTG TGA AT-3′), ITS2_R 179 (5′-GAC GCT TCT CCA GAC TAC AAT-3′), and _psb_ A– _trn_ H_F (5′-GTT ATG CAT 180 GAA CGT AAT GCT C-3′), and _psb_ A– _trn_ H_R (5′-CGC GCG TGAT GAA TTC ACA 181 ATC C-3′), respectively. A 25-μL PCR mixture contained 2 μL of total DNA as the 182 template, 1 μL of 10 μM each primer, and 12.5 μL of 2× PCR HeroTM Mix (FOREGENE 183 Co., Ltd., Chengdu, China), which contained Foregene Taq Hero, dNTPs, Tris-HCl (pH 184 8.8), KCl, MgCl2, and DNA loading buffer. PCR amplification was performed using an 185 Arhat 96 PCR Thermal Cycler (Monad Biotech Co., Ltd., Suzhou, China). The cycling 186 conditions included an initial hot start at 94°C for 3 min, followed by 35 cycles of 187 denaturation at 94°C for 10 s, annealing at 55°C for 10 s, and extension at 72°C for 20 s, 188 with a final extension at 72°C for 5 min. A 1-μL aliquot of each PCR product was examined 11 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 189 using 1.0% agarose gel electrophoresis. The PCR products were then bidirectionally 190 sequenced by Tsingke Biotech Co., Ltd. (Chongqing, China) using the primers employed 191 for the PCR amplification. Sequencing data were assembled using BioEdit (version 7.0.0), 192 and primer sequences were removed using MEGA (version 10.1.6). Consensus sequences 193 were aligned and compared using Multalin software 194 (http://multalin.toulouse.inra.fr/multalin/) or BioEdit (version 7.0.0). 195 ## Chemicals and reagents 196 Chemical standards of quercitrin, isoquercitrin, hyperoside, quercetin, 197 dihydromyricetin, guaijaverin, taxifolin, astragalin, avicularin, and naringenin were 198 purchased from Beijing Bethealth People Biomedical Technology Co., Ltd. (Beijing, 199 China), with purity ≥98%. HPLC-grade methanol and acetonitrile were purchased from 200 TEDIA (Fairfield, OH, USA), whereas HPLC-grade phosphoric acid was procured from 201 Aladdin (Shanghai, China). Analytical-grade methanol was sourced from Chongqing 202 Chuandong Chemical (Group) Co., Ltd. (Chongqing, China). Ultrapure water (18.2 203 MΩ·cm) was obtained using a Milli-Q Integral 5 water purification system (Millipore, 204 Billerica, MA, USA). 205 ## Preparation of standard and sample solutions for HPLC analysis 206 Standard solutions were prepared using the corresponding reference compounds. 207 Quercetin, isoquercetin, hyperoside, quercetin, dihydromyricetin, guaijaverin, taxifolin, 208 astragalin, avicularin, and naringenin (2.79, 2.69, 2.59, 2.55, 2.57, 2.50, 2.69, 2.87, 2.78, 209 and 2.67 mg, respectively) were accurately weighed and dissolved in 25 mL of methanol. 12 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 210 Subsequently, 1 mL of each standard solution was mixed to obtain concentrations of 11.16, 211 10.76, 10.36, 10.20, 10.28, 10.00, 10.76, 11.48, 11.12, and 10.68 μg·mL [−1] for each 212 compound, respectively. The standard mixture solution was further diluted 2-, 4-, 5-, 8-, 213 and 10-fold. Calibration curves were generated by plotting the peak area against the 214 corresponding concentration of each analyte. 215 _Rhododendron_ specimens and crude drug samples were heated at approximately 50°C 216 until dry, ground into powder, and passed through a 65-mesh (250-μm) sieve. The resulting 217 fine powder was stored at room temperature away from sunlight until analysis. 218 Subsequently, 500 mg of fine powder from each sample and 50 mL of methanol were added 219 to a covered conical flask. The flask was weighed before extraction and then extracted in 220 a water bath at 70°C for 60 min. After the solution was cooled to room temperature, the 221 flask was weighed again, and methanol was added to restore the original weight. The 222 supernatant was collected after static settling and filtered through a 0.22-μm filter (Tianjin 223 Jinteng Experiment Equipment Co., Ltd., Tianjin, China) for HPLC analysis. Each sample 224 was prepared in triplicate. ## 225 HPLC conditions 226 The HPLC system (Agilent, Santa Clara, CA, USA) was equipped with a G7111A 227 quaternary pump, G1322A degasser, G7129A autosampler, G1316A column oven, and 228 G7115A DAD detector. HPLC analysis was conducted using an Agilent 1260 system 229 (Agilent, USA) with a Welch UltimateXB-C18 column (4.6 mm × 250 mm, 5 μm). The 230 mobile phase consisted of solvent A (acetonitrile) and solvent B (0.1% phosphoric acid in 231 water, v/v), with the following gradient elution program: 0–5 min, 15% A; 5–25 min, 30% 13 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 232 A; 25–45 min, 45% A; 45–60 min, 60% A; 60–87 min, 80% A; 87–90 min, 10% A; and 233 90–100 min, 10% A. The flow rate was 1.0 mL/min, and the column temperature was 234 maintained at 30°C. Detection was performed at 254 nm with a 10 μL injection volume. ## 235 HPLC method validation 236 Precision was evaluated by performing six analyses of a standard solution containing 237 10 standard compounds, and the RSD of the peak area for each compound was calculated. 238 Repeatability was assessed by analyzing six sample solutions, with variations expressed as 239 RSD. The stability of one sample solution was evaluated by analyzing it at 0, 4, 8, 12, and 240 24 h. The accuracy of the method was assessed using a recovery test. The standard 241 compounds were weighed to approximately 100% of their original amounts and added to 242 the RL13 sample powder. The spike sample was then extracted and analyzed using the 243 aforementioned method. 244 ## Data processing 245 Independent two-sample t-tests or one-way analysis of variance were conducted using 246 SPSS Statistics (version 20; IBM, USA). The data matrix used for OPLS-DA was 247 generated from the HPLC peak areas of the 55 samples, and the data were normalized using 248 the Z-score transformation method, as described in our previous study [15]. SIMCA-P 249 (version 14.1; Umetrics, Umea, Sweden) was used to perform the OPLS-DA. # 250 Results ## 251 Molecular identification of Rhododendron species 14 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 252 Thirty-eight ITS2 sequences and 46 _psb_ A– _trn_ H sequences from nine species were 253 successfully amplified and determined. The 46 _psb_ A– _trn_ H sequences were all 379 bp in 254 length and showed high similarity, with several single-nucleotide differences detected 255 among them. All sequences were classified into six sequence types (Table 3). Among the 256 species in Section _Pogonanthum_ (“Talegabao”), eight _R. anthopogonoides_ specimens 257 showed identical sequences. _R. cephalanthum_ and _R. trichostomum_ showed the same 258 sequence, with guanine at alignment position 242, whereas eight _R. primuliflorum_ 259 specimens showed identical sequences, with thymine at position 187 and guanine at 260 position 242. However, sequences from the “Talenabao” species in Section _Rhododendron_, 261 Subsection _Lapponica_, including seven _R. capitatum_, 12 _R. nivale_, four _R. thymifolium_, 262 one _R. intricatum_, and one _R. nitidulum_, showed the same sequence, with five differences 263 at alignment positions 229 (G), 235 (G), 242 (G), 255 (A), and 262 (G) compared to the 264 sequence of _R. anthopogonoides_ . In addition, two _R. nivale_ sequences (RL13 and RL16) 265 had cytosine at position 105, and one _R. capitatum_ sequence (RL7) had thymine at 266 positions 255 and 262. The sequence alignment results indicated that the _psb_ A– _trn_ H region 267 successfully classified the species used as “Talegabao” and “Talenabao.” Furthermore, 268 among the four species used as “Talegabao,” the _psb_ A– _trn_ H region clearly distinguished 269 _R. anthopogonoides_ and _R. primuliflorum_ from _R. cephalanthum_ and _R. trichostomum_ . 270 Compared with _psb_ A– _trn_ H sequences, ITS2 sequences were more conserved. All 38 271 ITS2 sequences were 428 bp in length. Among them, 27 sequences from 8 species, 272 excluding _R. thymifolium_, were identical (Table 3). Several nucleotide differences were 273 detected among the other seven sequences from _R. anthopogonoides_, _R. capitatum_, and _R._ 274 _nivale_, indicating intraspecific nucleotide variation in individual plants of these species. 15 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 275 The four _R. thymifolium_ specimens had identical sequences. Two positions, 132 and 262, 276 contained thymine and distinguished this species from others. In particular, thymine at 277 position 132 was detected only in _R. thymifolium_, indicating that this position may be 278 species-specific. 279 280 Table 3. _psb_ A– _trn_ H and ITS2 sequences detected in _Rhododendron_ species Local Scientific name Voucher No. name Acession Number _R. anthopogonoides_ RP1~8 `T` `C` `A` `C` `T` `T` `T` PZ896252 Talegabao _R. cephalanthum, R. trichostomum_ RP17, RP18 `*` `*` `*` `*` `G` `*` `*` PZ896253 _R. primuliflorum_ RP9~16 `*` `T` `*` `*` `G` `*` `*` PZ896254 _R. capitatum, R. nivale, R. thymifolium,_ RL1~6, RL8~12, `*` `*` `G` `G` `G` `A` `G` PZ896255 _R. intricatum, R. nitidulum_ RL14~15,RL17~28 Talenabao _R. nivale_ RL13, RL16 `C` `*` `G` `G` `G` `A` `G` PZ896256 _R. capitatum_ RL7 `*` `*` `G` `G` `G` `*` `*` PZ896257 Acession Number Talegabao & _R. anthopogonoides, R. primuliflorum,_ _R. cephalanthum, R. trichostomum, R._ _capitatum, R. nivale, R. intricatum, R._ _nitidulum_ `T` `T` `C` `G` `C` `C` `C` PZ888515 Talenabao RP4, RP10, RP13~18; RL1~7, RL9~10, RL12~19, RL27~28 _R. anthopogonoides_ RP1 `A` `A` `*` `*` `*` `*` `*` PZ888518 Talegabao _R. anthopogonoides_ RP7 `*` `A` `*` `*` `*` `*` `*` PZ888516 _R. capitatum_, _R. nivale_ RL8, 11 `*` `*` `*` `*` `*` `A` `*` PZ888520 _R. nivale_ RL20 `*` `*` `*` `*` `*` `*` `T` PZ888522 Talenabao _R. nivale_ RL21 `*` `*` `T` `*` `*` `A` `*` PZ888519 _R. nivale_ RL22 `*` `*` `*` `A` `*` `*` `*` PZ888517 _R. thymifolium_ RL23~26 `*` `*` `*` `*` `T` `*` `T` PZ888521 281 Numbers above sequence indicate the alignment positions. Asterisk (*) indicates 282 nucleotide is same with that in the first sequence. 16 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 283 ## Quantitative analysis of the 10 compounds in “Tale” 284 The quantitative method was validated by evaluating linearity (R2), precision, stability, 285 repeatability, and recovery (Table 4). High correlation coefficients were observed for all 286 10 analytes (R2 ≥ 0.9990), indicating strong correlation between concentration and peak 287 area within the tested range (1.000–11.48 μg·mL [−1] ). Precision, repeatability, and stability 288 were also validated for each analyte, with relative standard deviations (RSDs) ranging from 289 0.96% to 2.57%, 1.48% to 2.97%, and 0.25% to 2.91%, respectively. Recoveries of the 10 290 analytes ranged from 94.43% to 106.08%, with RSDs ranging from 1.66% to 2.89%. These 291 findings indicate that the established method is sufficiently precise and accurate for 292 quantitative analysis of the 10 compounds in _Rhododendron_ . 293 294 Table 4. Regression equations, correlation coefficients, linearity ranges, precisions, 295 repeatability, stability, and recovery for the 10 analytes Stability (RSD%) n=6 Repeatability (RSD%) n=6 Calibration Curve Compound Linear Range/ R [2] Equation μg·mL [－] [1] μg·mL [－] [1] Precision (RSD%) n=6 Recovery RSD% (%) n=6 Dihydromyricetin Y= 4.3625X-0.4832 0.9996 1.028 ～ 10.28 2.57 2.91 2.44 104.54 2.89 Hyperoside Y=26.7X-2.1122 0.9998 1.036 ～ 10.36 1.06 0.25 2.97 94.93 2.01 lsoquercitrin Y=26.562X+1.7001 0.9998 1.076 ～ 10.76 1.04 0.88 2.24 106.08 2.75 Taxifolin Y=3.1571X+0.5621 0.9990 1.076 ～ 10.76 0.81 2.21 2.48 104.66 1.69 Guaijaverin Y=23.52X+1.1895 0.9998 1.000 ～ 10.00 0.96 0.41 2.11 95.34 1.66 Avicularin Y=26.846X+1.1688 0.9998 1.112 ～ 11.12 0.98 0.18 1.48 100.47 2.19 Astragalin Y=17.884X+2.285 0.9997 1.148 ～ 11.48 0.98 0.92 2.01 100.86 2.53 Quercitrin Y=28.144X+1.9873 0.9998 1.116 ～ 11.16 0.97 1.3 2.35 95.50 2.74 Quercetin Y=40.472X-2.2232 0.9998 1.020 ～ 10.20 0.97 0.57 2.28 103.58 2.68 17 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) Naringenin Y=5.8574X+0.4178 0.9999 1.068 ～ 10.68 0.97 0.64 2.95 100.42 2.73 296 297 Ten compounds were simultaneously detected in the HPLC chromatograms of all plant 298 specimens and crude drug samples based on their retention times. The relative contents of 299 the 10 compounds were also determined. The contents of all 10 compounds varied among 300 the plant specimens and crude drug samples (Table 5, Fig 2). Dihydromyricetin, hyperoside, 301 isoquercitrin, taxifolin, guaijaverin, avicularin, astragalin, quercitrin, quercetin, and 302 naringenin had contents ranging from 0.163 to 1.106, 0.052 to 4.523, 0.036 to 3.704, 0.273 303 to 3.214, 0.058 to 0.704, 0.023 to 1.232, 0.030 to 1.643, 0.030 to 2.697, 0.057 to 1.524, 304 and 0.106 to 1.610 mg·g [−1], respectively. The total content of the 10 compounds ranged 305 from 1.220 to 3.836 mg·g [−1], corresponding to 0.24%–2.24% of the samples. Clear 306 differences were detected between “Talegabao” and “Talenabao.” In the “Talegabao” 307 samples, the 10 compounds—dihydromyricetin, hyperoside, isoquercitrin, taxifolin, 308 guaijaverin, avicularin, astragalin, quercitrin, quercetin, and naringenin—had average 309 contents of 0.332, 2.016, 0.929, 1.665, 0.386, 0.326, 0.459, 1.029, 0.743, and 0.454 mg·g [−1], 310 respectively. The average total content was 8.340 mg·g [−1], with an average percentage of 311 1.65%. In contrast, the compounds in “Talenabao” had relatively low contents. 312 Dihydromyricetin, hyperoside, isoquercitrin, taxifolin, guaijaverin, avicularin, astragalin, 313 quercitrin, quercetin, and naringenin had average contents of 0.409, 0.723, 0.266, 0.801, 314 0.196, 0.192, 0.277, 0.306, 0.519, and 0.323 mg·g [−1], respectively. The average total content 315 was 4.011 mg·g [−1], with an average percentage of 0.80%, which was less than half that of 316 “Talegabao.” The 10 detected compounds differed in content between the two types of 317 medicines. Except for dihydromyricetin, which had a lower content in “Talegabao,” the 18 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 318 other nine compounds had higher contents in “Talegabao” than in “Talenabao.” In 319 particular, hyperoside, isoquercitrin, taxifolin, guaijaverin, quercitrin, and quercetin had 320 significantly higher contents in “Talegabao,” with all p-values < 0.001, indicating that these 321 six compounds contributed substantially to the difference in total content between the two 322 types of medicine. Among the “Talegabao” samples, the four species _R. anthopogonoides_, 323 _R. primuliflorum_, _R. cephalanthum_, and _R. trichostomum_ showed similar total contents of 324 the 10 compounds, with average values of 8.411, 8.227, 8.948, and 8.167 mg·g [−1], 325 respectively. Several compounds differed in content among species. _R. anthopogonoides_ 326 had significantly lower contents of dihydromyricetin and quercitrin than _R. primuliflorum_, 327 with p-values of 0.0019 and 0.013, respectively. _R. anthopogonoides_ had a significantly 328 higher quercetin content than _R. primuliflorum_, with a p-value of 0.01. In addition, _R._ 329 _cephalanthum_ had high contents of taxifolin and naringenin and low levels of avicularin 330 and astragalin. _R. trichostomum_ had a high content of avicularin and low contents of 331 astragalin and naringenin. Among the “Talenagao” samples, the three species _R. capitatum_, 332 _R. nivale_, and _R. thymifolium_ had average total contents of the 10 compounds of 4.493, 333 3.665, and 4.372 mg·g [−1], respectively, which were similar to the 4.705 mg·g [−1] of the one 334 _R. intricatum_ sample. However, the _R. nitidulum_ sample had a total content of 2.873 mg·g [−1], 335 which was much lower than that of the other four species. Among these five species used 336 as “Talenabao,” _R. nivale_ had lower quercetin and naringenin contents. _R. thymifolium_ had 337 a significantly higher hyperoside content than _R. capitatum_ and _R. nivale_, with p-values of 338 0.000 and 0.004, respectively, and significantly lower contents of guaijaverin and 339 astragalin than _R. capitatum_ and _R. nivale_, with p-values of 0.016 and 0.008 for guaijaverin 340 and 0.020 and 0.012 for astragalin, respectively. In addition, _R. intricatum_ had high 19 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 341 taxifolin and guaijaverin contents and a low avicularin content. However, _R. nitidulum_ had 342 a somewhat different compound profile from the other species, with a high astragalin 343 content and low contents of isoquercitrin, avicularin, quercetin, and naringenin. 344 345 346 Table 5 The contents of ten compounds in “Tale” samples. Avicular in (mg·g [-1] ) Dihydromyrice tin (mg·g [-1] ) Hyperosi de (mg·g [-1] ) lsoquercitr in (mg·g [-1] ) Astragal in (mg·g [-1] ) Quercitri n (mg·g [-1] ) Querceti n (mg·g [-1] ) Naringen in (mg·g [-1] ) Local name [Species] _R._ _anthopogon_ _oides_ _R._ _primulifloru_ _m_ Voch er No. RP1 0.217 4.523 1.015 1.877 0.636 1.232 0.099 0.759 0.958 0.106 11.421 2.24 RP2 0.471 0.924 2.691 1.882 0.283 0.453 1.169 0.059 1.089 0.558 9.578 1.90 RP3 0.265 1.311 0.759 0.398 0.473 0.482 0.102 0.937 0.822 0.706 6.253 1.23 RP4 0.239 1.569 0.364 0.614 0.305 0.506 0.033 1.088 1.253 0.122 6.093 1.21 RP5 0.231 1.157 0.625 2.278 0.354 0.199 0.059 0.033 1.044 0.377 6.356 1.27 RP6 0.274 1.419 0.406 2.691 0.177 0.047 0.074 2.467 0.523 0.110 8.189 1.63 RP7 0.173 2.104 0.532 0.854 0.374 0.809 0.030 0.870 1.524 0.211 7.481 1.48 RP8 0.206 0.387 1.555 1.370 0.351 0.381 0.641 0.043 1.181 0.627 6.741 1.34 D1 0.268 3.951 1.100 1.382 0.620 0.045 1.611 0.062 0.897 0.191 10.128 2.01 D2 0.436 4.068 1.225 1.966 0.704 0.023 1.643 0.069 0.950 0.147 11.231 2.22 D3 0.346 3.704 0.916 1.483 0.547 0.987 0.082 0.541 0.915 0.115 9.636 1.93 D4 0.248 1.224 0.432 2.006 0.137 0.247 0.795 1.499 0.344 0.468 7.400 1.46 D5 0.269 2.493 0.975 2.073 0.506 0.026 0.796 0.049 1.045 0.604 8.835 1.77 RP9 0.263 0.846 3.704 2.074 0.554 0.033 0.092 0.883 0.635 0.331 9.414 1.88 RP10 0.551 1.633 0.720 1.802 0.262 0.034 0.048 1.817 0.324 0.404 7.596 1.50 RP11 0.334 1.024 0.425 2.029 0.219 0.030 0.069 2.052 0.308 0.648 7.138 1.42 RP12 0.543 1.868 0.689 1.000 0.282 0.032 0.042 2.294 0.449 0.327 7.526 1.48 RP13 0.623 1.766 0.583 1.888 0.430 0.072 1.010 0.448 0.199 1.131 8.150 1.62 RP14 0.364 1.551 0.452 0.650 0.233 0.178 1.166 0.930 0.446 0.712 6.682 1.32 RP15 0.265 1.311 0.759 0.398 0.473 0.482 0.102 0.937 0.822 0.706 6.253 1.24 RP16 0.346 3.704 0.916 1.767 0.198 0.478 1.481 0.541 0.915 0.115 10.460 2.06 D6 0.319 3.260 1.085 2.278 0.274 0.519 0.944 1.168 0.830 0.372 11.049 2.19 D7 0.415 2.055 0.571 2.131 0.275 0.240 0.051 1.508 0.465 0.815 8.524 1.68 D8 0.213 1.809 0.636 1.776 0.339 0.183 0.031 2.697 0.543 0.344 8.568 1.69 D9 0.446 1.732 0.531 0.905 0.349 0.286 0.155 2.003 0.653 0.309 7.369 1.45 20 Taxifolin SUM (mg·g [-1] ) Percenta ge (%) (mg·g [-1] ) Guaijav erin (mg·g [-1] ) [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) _R._ _cephalanthu_ _m_ _R._ _trichostomu_ _m_ RP17 0.353 1.352 0.806 3.214 0.495 0.047 0.034 0.770 0.353 1.524 8.948 1.78 RP18 0.300 1.693 0.609 2.173 0.580 0.740 0.033 1.270 0.584 0.187 8.167 1.62 Average 0.332 2.016 0.929 1.665 0.386 0.326 0.459 1.029 0.743 0.454 8.340 1.65 _R. capitatum_ _R. nivale_ _R._ _thymifolium_ RL1 0.579 0.713 0.711 0.748 0.097 0.278 0.411 0.321 0.650 1.610 6.119 1.22 RL2 0.592 0.488 0.066 0.662 0.147 0.242 0.042 0.298 0.572 0.215 3.322 0.65 RL3 0.592 0.746 0.086 0.573 0.121 0.397 0.546 0.696 0.744 0.544 5.045 1.00 RL4 0.338 0.482 0.631 0.908 0.121 0.225 0.375 0.359 0.518 0.268 4.225 0.84 RL5 0.163 0.634 0.078 0.606 0.099 0.383 0.479 0.326 0.452 0.213 3.433 0.68 RL6 0.330 0.428 0.386 0.381 0.075 0.184 0.184 0.177 0.303 0.153 2.601 0.52 RL7 0.265 0.612 0.759 0.398 0.473 0.482 0.102 0.937 0.822 0.706 5.554 1.11 RL8 0.505 0.933 0.107 1.744 0.215 0.097 0.614 0.216 0.738 0.472 5.641 1.13 RL9 0.407 0.207 0.078 0.344 0.088 0.025 0.180 0.693 0.082 0.158 2.261 0.45 RL10 0.249 1.259 0.110 0.429 0.245 0.311 0.066 0.128 0.489 0.254 3.539 0.71 RL11 0.989 0.333 0.500 1.594 0.152 0.072 0.354 0.035 0.390 0.147 4.564 0.91 RL12 0.314 0.457 0.045 0.273 0.117 0.297 0.043 0.049 0.404 0.191 2.191 0.43 RL13 0.187 1.654 0.337 0.978 0.407 0.084 0.320 0.072 0.377 0.222 4.638 0.92 RL14 0.529 0.701 0.354 0.854 0.145 0.226 0.677 0.110 0.372 0.328 4.297 0.85 RL15 0.701 0.792 0.750 0.998 0.058 0.264 0.037 0.030 0.437 0.114 4.181 0.83 RL16 1.106 1.168 0.192 2.396 0.176 0.366 0.214 0.063 0.316 0.187 6.186 1.22 RL17 0.362 0.052 0.042 0.278 0.097 0.043 0.057 0.053 0.057 0.178 1.220 0.24 RL18 0.272 0.191 0.140 0.546 0.088 0.033 0.152 0.319 0.149 0.114 2.002 0.40 RL19 0.226 0.510 0.036 0.640 0.185 0.067 0.618 0.138 0.553 0.158 3.131 0.62 RL20 0.450 0.910 0.559 0.386 0.164 0.392 0.503 0.976 0.565 0.255 5.159 1.03 RL21 0.309 0.205 0.519 0.836 0.181 0.167 0.076 0.318 0.242 0.204 3.057 0.60 RL22 0.303 0.938 0.112 1.354 0.177 0.077 0.532 0.123 1.011 0.251 4.878 0.97 RL23 0.264 1.123 0.186 0.632 0.350 0.125 0.087 0.411 0.746 0.447 4.370 0.87 RL24 0.260 1.125 0.125 0.538 0.298 0.162 0.048 0.349 0.674 0.495 4.073 0.81 RL25 0.242 1.326 0.199 0.524 0.324 0.166 0.049 0.578 0.952 0.460 4.819 0.96 RL26 0.300 1.253 0.181 0.522 0.320 0.130 0.055 0.457 0.709 0.300 4.226 0.85 _R._ RL27 0.374 0.452 0.111 1.727 0.332 0.051 0.385 0.244 0.756 0.274 4.705 0.94 _intricatum_ _R. nitidulum_ RL28 0.241 0.556 0.043 0.554 0.227 0.030 0.554 0.084 0.452 0.134 2.873 0.57 Average 0.409 0.723 0.266 0.801 0.196 0.192 0.277 0.306 0.519 0.323 4.011 0.80 347 348 **Fig 2 The contents of ten compounds in “Tale” samples.** 349 21 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 350 OPLS-DA was conducted using HPLC data from all 55 _Rhododendron_ samples. The 351 OPLS-DA score and loading plots are shown in Fig 3. In the score plot, the “Talegabao” 352 samples were distinguished from the “Talenabao” samples. However, samples within each 353 group were not clearly separated, indicating similar chemical profiles (Fig 3a). In the 354 loading plot, the five species used as “Talenabao” clustered together and were associated 355 with a high dihyromyricetin content. The four species used as “Talegabao” clustered in the 356 right part of the plot but were separated from one another. _R. anthopogonoides_ was located 357 in the upper right, distant from the center, and was associated with high quercetin and 358 avicularin content. _R. primuliflorum_ and _R. trichostomum_ were located in the lower left 359 and were associated with high contents of naringenin, quercitrin, and taxifolin. In addition, 360 _R. cephalanthum_ was associated with a high astragalin content (Fig 3b). These OPLS-DA 361 results were consistent with the HPLC analysis. 362 363 **Fig 3 OPLS-DA results of 55 “Tale” samples using the data matrix created by HPLC** 364 **quantitative data.** a, score plot of OPLS-DA results; b, loading plot of OPLS-DA results. 365 # 366 Discussion ## 367 Molecular identification of Rhododendron species 368 In this study, the ITS2 and _psb_ A– _trn_ H regions of all nine species used as the traditional 369 Tibetan medicine “Tale” were amplified and sequenced to distinguish among the species. 22 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 370 The alignment results indicated that the ITS2 region had limited discriminatory power for 371 the _Rhododendron_ species used as “Tale” medicines. In contrast, the _psb_ A– _trn_ H region 372 successfully distinguished the species used as “Talegabao” and “Talenabao” and showed a 373 high capability for discriminating _R. anthopogonoides_ and _R. primuliflorum_ from _R._ 374 _cephalanthum_ and _R. trichostomum_ . These results are consistent with the findings of Yan 375 et al., who reported that _psb_ A– _trn_ H had relatively higher species-discrimination power 376 than ITS sequences and recommended it as a single DNA barcode for _Rhododendron_ 377 species [14]. However, the species-discrimination power of the DNA barcodes evaluated 378 in the present study was relatively low. Neither ITS2 nor _psb_ A– _trn_ H could distinguish all 379 nine species from one another. 380 Interestingly, some studies have reported high discriminatory power of the ITS region 381 for classifying _Rhododendron_ species. Gao Lian-Ming et al. reconstructed a phylogenetic 382 tree based on ITS sequences from 12 species representing two sections of Subgenus 383 _Tsutsusi_ and four species representing three Sections of Subgenus _Pentanthera_ of 384 _Rhododendron_ . The strict-consensus tree topology showed that Subgenus _Tsutsusi_ was 385 monophyletic, and _R. tashiroi_ should be a member of Section _Brachycalyx_, whereas _R._ 386 _tsusiophyllum_ should be a member of Section _Tsutsusi_ . ITS sequences successfully 387 resolved the taxonomic relationships of species in Subgenera _Tsutsusi_ and _Pentanthera_ 388 [16]. Cai Hong-Yue et al. evaluated the relationships among species in the subgenera of 389 _Rhododendron_ and clarified the classification of five native _Rhododendron_ species in 390 Mount Tai based on ITS regions. The results showed that, among the five species from 391 Mount Tai, _R. mariesii_ did not cluster within the branch of Subgenus _Tsutsusi_, whereas the 392 other four species were located in the major branches of their subgenera [17]. 23 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 393 However, the species examined in these studies were mainly distributed at lower 394 elevations. The species in the present study were collected from the Qinghai–Xizang 395 Plateau at an average altitude of more than 4,000 m, similar to the species examined by 396 Yan et al., which were collected from the Himalaya–Hengduan Mountains. Highland 397 species appear to show lower levels of species discrimination, with shared sequence 398 identities across species and less distinct species boundaries. A similar phenomenon has 399 also been reported in highland species of subsection Ledum in North America [18]. Yan et 400 al. proposed that these closely related highland species may have diverged morphologically 401 in response to strong selection pressures associated with different habitats while 402 undergoing relatively little genetic divergence, possibly because of long generation times 403 and relatively recent speciation and diversification. In addition, incongruent signals 404 between nuclear and plastid gene regions are common in these species and may result from 405 natural hybridization and gene introgression. These factors may collectively contribute to 406 the limited effectiveness of DNA barcoding in distinguishing these species [14]. 407 To address this issue in highland _Rhododendron_ species, Fu et al. introduced a 408 genome-skimming method based on the chloroplast genome to distinguish _Rhododendron_ 409 species from the Himalayan–Hengduan Mountains. Compared with the 33% of species 410 distinguished using ITS + matK + rbcL + _trn_ H– _psb_ A, 55% of species were distinguished 411 using plastid genome and nrDNA sequences [19]. Given the limited taxonomic information 412 available for these highland _Rhododendron_ species, the _psb_ A– _trn_ H region currently 413 showed the highest species-discrimination power among the regions evaluated in the 414 present study and has potential for use in distinguishing the traditional Tibetan medicine 415 “Tale.” With increasing attention to the medicinal, environmental, and horticultural value 24 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 416 of these _Rhododendron_ species, further exploration of new DNA barcode regions and other 417 methods for species discrimination is warranted. 418 ## Quantitative analysis of the 10 compounds in “Tale” 419 In this study, we conducted a phytochemical investigation of nine _Rhododendron_ 420 species used as traditional Tibetan medicine “Tale.” An HPLC method was used to 421 simultaneously detect 10 flavonoid compounds, and the contents of these compounds were 422 determined in the nine _Rhododendron_ species. All 10 compounds were detected in the plant 423 specimens and crude drug samples, indicating a high degree of phytochemical similarity 424 among the samples. The nine species used as “Tale” medicine are reasonable. The species 425 used as “Tale” medicine differed primarily in the contents of these compounds. The total 426 content of the 10 flavonoid compounds in “Talegabao” was more than twice that in 427 “Talenabao,” indicating a substantial difference between the two types of medicines. 428 Flavonoid compounds have been associated with various pharmacological activities, 429 including antioxidant, anti-inflammatory, antimicrobial, antitumor, hypoglycemic, 430 hepatoprotective, neuroprotective, and cardioprotective effects [20]. As described in the 431 Introduction, “Talegabao” was described as the first choice among “Tale” medicines, 432 whereas “Talenabao” is usually used as an alternative. Our results provide phytochemical 433 evidence relevant to the use of these medicines. 434 Among the 10 compounds, nine had higher contents in the species used as “Talegabao,” 435 and only dihydromyricetin had a higher content in the species used as “Talenabao.” 436 Furthermore, the contents of individual compounds varied among species. Different 437 flavonoid compounds have different properties [21]. These differences may contribute to 25 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 438 differences in the effects of the two types of medicine. Although the content of certain 439 compounds differed among species, samples within each medicine type had similar total 440 contents and phytochemical profiles, and the OPLS-DA results indicated that species 441 within each type were not clearly distinguishable. These findings support the use of 442 multiple species as sources of “Tale” medicines. “Talenabao” can be used as a substitute 443 for “Talegabao.” Among the species analyzed as “Talegabao” or “Talenabao,” certain 444 compounds showed distinct contents among species, providing a basis for further 445 standardization of “Tale” medicines. 446 Few studies on the chemical composition of _Rhododendron_ species used as “Tale” 447 have been reported. To our knowledge, this study is the first conducted to compare 448 _Rhododendron_ species used as “Tale” based on molecular and phytochemical analyses. 449 However, several issues related to the standardization of “Tale” medicines during their 450 production and use remain to be addressed. Multiple species are used as “Tale” medicines, 451 and the contents of individual compounds vary among these species. Differences were 452 observed not only among species but also within species, with some individuals showing 453 distinct compound contents. However, the factors influencing the accumulation of these 454 compounds remain unknown. Although several factors influencing the accumulation of 455 flavonoid compounds have been examined, Maxime assessed the effect of anthropogenic 456 disturbances on the chemical properties of _R. groenlandicum_ in the territories of three 457 Indigenous communities. The results showed that flavonoid concentrations varied mainly 458 by territory, whereas disturbance type had a smaller effect [22]. The mechanisms 459 underlying flavonoid accumulation remain unclear. 26 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 460 Multiple monoterpenoid and sesquiterpenoid compounds have been isolated from the 461 essential oils of _Rhododendron_ species, and these essential oils have shown multiple 462 properties [23]. In addition, the toxic constituents of _Rhododendron_ species warrant 463 attention [2,24]. Future standardization studies should consider additional species and 464 evaluation indicators, including pharmacological studies, essential oils, and toxic 465 compounds. # 466 Conclusions 467 In this study, 46 plant specimens and nine crude drug samples belonging to nine 468 _Rhododendron_ species were collected from the Qinghai–Xizang Plateau. Molecular 469 methods based on the ITS2 region of nuclear DNA and the _psb_ A– _trn_ H region of 470 chloroplast DNA were applied to these plant specimens. In addition, 10 flavonoid 471 compounds were quantitatively determined by HPLC. The molecular analysis revealed that 472 the _psb_ A– _trn_ H region successfully distinguished the species used as “Talegabao” or 473 “Talenabao”; however, ITS2 had limited discriminatory power because of its high degree 474 of conservation. The HPLC analysis showed that all 10 compounds were detected in all 475 samples. The species used as “Talegabao” had more than twice the total content of the 10 476 flavonoid compounds observed in the species used as “Talenabao.” In addition, differences 477 in the contents of certain compounds were detected among species, which may provide 478 potential chemical markers for the quality evaluation of “Tale” medicines. According to 479 the results, “Talegabao” is the first choice among “Tale” medicines, whereas “Talenabao” 480 can be used as an alternative to “Talegabao”; however, the appropriate dose requires further 481 discussion in clinical practice. The _psb_ A– _trn_ H region may serve as a DNA barcode for 27 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 482 distinguishing the species used as “Talegabao” and “Talenabao.” 483 # 484 References 485 1. Hussain A, Azam S, Maqsood R, Anwar R, Akash MSH, Hussain H, et al. Chemistry, 486 biosynthesis, and theranostics of antioxidant flavonoids and polyphenolics of genus 487 _Rhododendron_ : an overview. Naunyn-Schmiedeberg’s Arch Pharmacol. 2025;398: 488 1171–1214. doi:10.1007/s00210-024-03428-6 489 2. Popescu R, Kopp B. The genus _Rhododendron_ : An ethnopharmacological and 490 toxicological review. Journal of Ethnopharmacology. 2013;147: 42–62. 491 doi:10.1016/j.jep.2013.02.022 492 3. Wu Z, Raven PH, Hong D. Flora of China. Science Press & Missouri Botanical Garden 493 Press; 2005. 494 4. Yang Y. Zang yao zhi [= Dictionary of traditional Tibetan medicine]. Qinghai People’s 495 Publishing House; 1991. 496 5. Jiang Q, Bai Y. Research progress of Mongolian drug Rhododendron and plants of the 497 genus _Rhododendron_ . 2014;29: 68–70. doi:10.14045/j.cnki.15-1220.2014.01.015 498 6. Wang S, Lu Y, Zhang K, Wang S, Zhang S, Su H, et al. The chemical constituents and 499 anti-complement activity of seven _Rhododendron_ species in Tibetan medicine. 500 Molecules. 2026;31: 2257. doi:10.3390/molecules31132257 28 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 501 7. Zhuoji Z, Li L, Wang X, Jixian C, Gazangcuo, Gengzangjia. Research advances on 502 chemical constituents, pharmacological effects, and clinical applications of Liexiang 503 Dujuan ( _Rhododendron anthopogonoides_ Maxim.) in Tibetan medicine. 2005;31: 100– 504 107. doi:10.13862/j.cn43-1446/r.2025.08.016 505 8. Zhang J, Ma J, Ynag Y, Yao J, Huang L, Gao L, et al. Analysis of essential oil in 506 _Rhododendron anthopogoncides_ . 2003;34: 304–305. 507 9. J H, X S, L D, X Y, B L, Y W, et al. Chemical constituents, antibacterial, acaricidal 508 and anti-inflammatory activities of the essential oils from four _Rhododendron_ species. 509 Frontiers in veterinary science. 2022;9. doi:10.3389/fvets.2022.882060 510 10. Q S, Tt L, Ym W, Xy S, C L, Jy L, et al. Meroterpenoids with diverse structures and 511 anti-inflammatory activities from _Rhododendron anthopogonoides_ . Phytochemistry. 512 2020;180. doi:10.1016/j.phytochem.2020.112524 513 11. Yan L, Hu F, Wu J, Han F. The biological activity of essential oil in _Rhododendron_ 514 _anthopogonoides_ Maxim and _Artemisia dubia_ Wall. ex Bess .to grassland caterpillar, 515 _Gynaephora_ _menyuanenis_ ( Lepidoptera :Lymantriidae). 2009;18: 58–63. 516 doi:10.3969/j.issn.1004-1389.2009.05.013 517 12. K Y, Yx Z, Cf W, Ss D, Zw D, Qz L, et al. Toxicity of _Rhododendron anthopogonoides_ 518 essential oil and its constituent compounds towards _Sitophilus zeamais_ . Molecules 519 (Basel, Switzerland). 2011;16. doi:10.3390/molecules16097320 520 13. Liu Y, Zhang L, Chen S, Chen K. Identification of 10 medicinally used _Rhododendron_ 29 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 521 species based on ITS2 sequences. 2011;34: 1342–1345. doi:10.13863/j.issn1001 522 4454.2011.09.010 523 14. Yan L, Liu J, Möller M, Zhang L, Zhang X, Li D, et al. DNA barcoding of 524 _Rhododendron_ (Ericaceae), the largest Chinese plant genus in biodiversity hotspots of 525 the Himalaya–Hengduan Mountains. Molecular Ecology Resources. 2015;15: 932– 526 944. doi:10.1111/1755-0998.12353 527 15. Liu Q, Yu H, Dong Y, Quan W, Su Z, Li L. Quality evaluation of Lonicerae Flos 528 produced in Southwest China based on HPLC analysis and antioxidant activity. 529 Molecules. 2024;29: 2560. doi:10.3390/molecules29112560 530 16. Gao L, Yang J, Zhang C, Li D. Phylogenetic relationship of subgenus _Tsutsusi_ 531 ( _Rhododendron_ ) based on ITS sequences. 2002;24: 313–320. 532 17. Cai H, Xing S, He R, Zhou B, Zhao F. Phylogenetic relationship analysis of genus 533 _Rhododendron_ (Ericaceae) from Mountain Tai based on ITS2 sequences and 534 morphological traits. 2017;53: 1489–1498. doi:10.13592/j.cnki.ppj.2017.0002 535 18. Iunusova DR, Polezhaeva MA, Marchuk EA. Confusing boundaries of the labrador tea 536 species: dispersal history explains the lack of clear species structure. Annals of Botany. 537 2024;134: 959–974. doi:10.1093/aob/mcae130 538 19. Fu C, Mo Z, Yang J, Cai J, Ye L, Zou J, et al. Testing genome skimming for species 539 discrimination in the large and taxonomically difficult genus _Rhododendron_ . 540 Molecular Ecology Resources. 2022;22: 404–414. doi:10.1111/1755-0998.13479 30 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/) 541 20. M V, L R. Phytochemical profiling and biological activities of _Rhododendron_ subsect. 542 _Ledum_ : discovering the medicinal potential of labrador tea species in the northern 543 hemisphere. Plants (Basel, Switzerland). 2024;13. doi:10.3390/plants13060901 544 21. Wang H, Li Y, Shen W, Guo G, Jiang Z, Cen Y, et al. Studies on antiproliferative 545 effect of flavones compounds Isolated from Yao herb medicines. 2007;30: 980–984. 546 doi:10.13863/j.issn1001-4454.2007.08.005 547 22. Thomas M, Lamara M, Asselin H, Fenton NJ. Effects of industrial disturbances on the 548 flavonoid concentration of _Rhododendron groenlandicum_ . Botany. 2023;101: 343–356. 549 doi:10.1139/cjb-2022-0136 550 23. Y S, P W, J L, Y H, X W. Chemical composition of essential oils from three 551 _Rhododendron_ species and their repellent, insecticidal and fumigant activities. 552 Chemistry & biodiversity. 2022;19. doi:10.1002/cbdv.202200740 553 24. Q L, Jx H, Xm Z, Wh X. Traditional uses, phytochemistry, pharmacology, toxicology, 554 and quality control of _Rhododendron dauricum_ L. leaves: A comprehensive review. 555 Journal of ethnopharmacology. 2023;305. doi:10.1016/j.jep.2022.116085 556 31 [bioRxiv preprint doi: https://doi.org/10.64898/2026.11.05.764495; this version posted November 11, 2026. The copyright holder for this preprint](https://doi.org/10.64898/2026.11.05.764495) (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made [available under a CC-BY 4.0 International license.](http://creativecommons.org/licenses/by/4.0/)