Skip to main content

Natural variation in Beauty Mark is associated with UV-based geographical adaptation in Gossypium species



Anthocyanins, a class of specialized metabolites that are ubiquitous among plant species, have attracted a great deal of attention from plant biologists due to their chemical diversity. They confer purple, pink, and blue colors that attract pollinators, protect plants from ultraviolet (UV) radiation, and scavenge reactive oxygen species (ROS) to facilitate plant survival during abiotic stress. In a previous study, we identified Beauty Mark (BM) in Gossypium barbadense as an activator of the anthocyanin biosynthesis pathway; this gene also directly led to the formation of a pollinator-attracting purple spot.


Here, we found that a single nucleotide polymorphism (SNP) (C/T) within the BM coding sequence was responsible for variations in this trait. Transient expression assays of BM from G. barbadense and G. hirsutum in Nicotiana benthamiana using luciferase reporter gene also suggested that SNPs in the coding sequence could be responsible for the absent beauty mark phenotype observed in G. hirsutum. We next demonstrated that the beauty mark and UV floral patterns are associated phenotypes and that UV exposure resulted in increased ROS generation in floral tissues; BM thus contributed to ROS scavenging in G. barbadense and wild cotton plants with flowers containing the beauty mark. Furthermore, a nucleotide diversity analysis and Tajima’s D Test suggested that there have been strong selective sweeps in the GhBM locus during G. hirsutum domestication.


Taken together, these results suggest that cotton species differ in their approaches to absorbing or reflecting UV light and thus exhibit variations in floral anthocyanin biosynthesis to scavenge reactive ROS; furthermore, these traits are related to the geographic distribution of cotton species.


Pollinator-mediated selection is commonly associated with intraspecies variation in flower color, which can be a reliable signal of nectar content and thus directly impact pollinator visiting rates [1]. However, this mechanism cannot account for floral color differences in self-pollinating species. In such cases, indirect selection via biotic and abiotic stresses may be responsible for variations in flower color [2]. Abiotic factors such as temperature, drought stress, and exposure to ultraviolet (UV) radiation influence differences in flower color; pigmentation in flowers can confer fitness advantages under heat and drought stress [3,4,5,6]. Furthermore, compounds that yield floral pigmentation act as antioxidants, protecting against the reactive oxygen species (ROS) produced by UV-B exposure [7]. Such compounds include specialized metabolites such as flavonoids and anthocyanins, which are essential for plant survival under multiple environmental stresses [8]. Anthocyanins are ubiquitously distributed among members of the plant kingdom and have attracted the attention of plant biologists due to their chemical diversity and ecological significance [9, 10]. They endow plant organs with purple, pink, or blue colors, which attract pollinators and seed dispersers [11] and protect plants against UV radiation [12] and pathogen invasion [13]. In addition, antioxidant activity allows anthocyanins to scavenge ROS, facilitating plant survival in response to abiotic stresses [14]. Anthocyanin accumulation in plant organs is recognized as a visible biomarker to indicate past environmental stresses. Plants have developed sophisticated molecular mechanisms to synchronize anthocyanin biosynthesis with growth and development. However, the underlying molecular mechanism by which stressors induce anthocyanin biosynthesis is not clear.

In many plant species, anthocyanin biosynthesis and the associated regulatory mechanisms at the transcriptional level have been thoroughly elucidated [15, 16]. In higher plants, anthocyanin biosynthesis-related gene expression is regulated by the conserved MBW core activation complex, which includes the v-myb avian myeloblastosis viral oncogene homolog (MYB), basic helix-loop-helix (bHLH), and WD40 subunits. R2R3 MYB regulators have been demonstrated to be transcriptional activators of the anthocyanin biosynthetic pathway in many species, such as Arabidopsis [17], petunia [18], tomato [19], grapevine [20], maize [21], potato [22], sweet potato [23], and apples [24]. Some MYB transcription factors (TFs) have also been identified as repressors of the anthocyanin biosynthetic pathway, including MYB1 in strawberries [25], MYB308 and MYB330 in snapdragon [26], MYB10 in apples [27], MYB4a and MYB4b in grapes [28], MYB4 in Arabidopsis [29], MYB27 in petunia [30], and MYBF2 in ginkgo [31]. Petals with large spots and discrete anthocyanin depositions that contrast with the background flower coloration occur in members of the families Liliaceae, Orchidaceae, Asteraceae, Papaveraceae, Fabaceae, Malvaceae, and many others, indicating that this trait has evolved multiple times independently. Numerous field studies have revealed that petal spots are important in mediating interactions with pollinators [1, 32, 33].

Increased plant exposure to UV-B radiation can increase the concentrations of anthocyanins and other specialized metabolites associated with floral pigmentation in both the visible and UV spectra [34]. Biosynthesis of UV light-absorbing compounds (flavonols) may attenuate damage caused by excess irradiance [7, 35]. There is substantial empirical evidence that floral anthocyanins can shield chloroplasts from UV damage [12, 36] by capturing a portion of supernumerary photons that would otherwise strike the chloroplasts, increasing ROS production and therefore ROS-triggered damage [37]. Previous studies in model species such as petunia and snapdragon have clarified the genetic control of some pigment patterns (and the association of such pigments with UV absorbance) [38] and of variation in pigment intensity in different regions of the corolla [30]. Cultivated upland cotton lacks an area of anthocyanin pigmentation at the base (known as a petal spot), but such spots are not uncommon in the so-called primitive cottons or race stocks [39]. For example, G. barbadense and G. arboreum (Asiatic cotton) have petal spots. This characteristic has been identified as a marker and is used by breeders and seed producers [40]. However, general understandings of cotton petal spot development and regulation at the molecular level remain poor.

Recently, we cloned Beauty Mark (BM) from G. barbadense using a map-based cloning method. We demonstrated that GbBM activated anthocyanin production in the petal spot region and directly led to the formation of a pollinator-attracting purple spot (the beauty mark) at the base of flower petals. We confirmed that the presence of these spots significantly increased the number of visiting honeybees, and that introducing the GbBM allele from G. barbadense to G. hirsutum improved pollination [41]. In the present study, we found the causal single nucleotide polymorphism (SNP) within the GbBM coding sequence (CDS) that was responsible for beauty mark development; furthermore, the beauty mark was shown to be associated with UV floral patterning. We demonstrated that the co-occurrence of the beauty mark and UV floral patterning protected flower reproductive tissues from UV light by producing antioxidants. Finally, we demonstrated the role of geographic distribution and associated UV light intensity in cotton species evolution.


Co-occurrence of the beauty mark phenotype and UV patterns in Gossypium species

We previously studied phenotypic differences between G. barbadense and G. hirsutum flowers to determine the genuine signals in a signaling relationship between honeybee pollinators and flowers. A dark purple mark at the base of the petals is characteristic of G. barbadense genotypes, whereas most of the cultivated G. hirsutum genotypes do not have a beauty mark. We here studied beauty marks in cultivated cotton G. herbaceum, and wild cotton species, namely G. capitis-viridis, G. anomalum, G. davidsonii, Rozi, G. klotzschianum, G. stocksii, and G. bickii, which had beauty marks at the petal bases of petals. We also studied one species, G. tomentosum, which did not have a beauty mark. The intensity and size of the beauty marks varied between species. Beauty marks and petal UV patterns were previously found to be the most conspicuous traits associated with honeybee visits. We therefore focused on the association between anthocyanin pigmentation in the beauty mark and UV-responsive flavonol compounds. When exposed to UV light, the petals of G. barbadense, G. herbaceum, G. capitis-viridis, G. anomalum, G. davidsonii, Rozi, G. klotzschianum, G. stocksii, and G. bickii significantly darkened, whereas the non-beauty-marked petals of G. hirsutum and G. tomentosum remained relatively bright (Fig. 1A). Consistent with those data, the UV light patterns of the petals differed between species with and without beauty marks (Fig. 1B). UV patterning may lead to an increase in petal temperature in response to UV exposure. We therefore proposed that UV light exposure may lead to differences in temperature and metabolic activities in floral tissues of Gossypium species with and without beauty marks. To test this hypothesis, we measured the petal temperatures of wild and cultivated Gossypium species. The temperatures of G. barbadense and wild cotton petals that contained the beauty mark were ~ 2 °C higher than those of G. hirsutum and G. tomentosum (Fig. 1C).

Fig. 1
figure 1

Effects of co-occurrence of Beauty Mark phenotype and UV light pattern on floral petals of Gossypium species. A Flowers of Gossypium species under visible light and UV light. Scale bar, 2 cm. The data were obtained from three independent replicates. B UV-reflectance data measured from floral petal of Gossypium species. C Temperature data measured from floral petal of Gossypium species. The data in B and C were analyzed by ANOVA one-way comparison followed by LSD test. Different letters above the bars indicate a significant difference at P < 0.05

Beauty Mark decreased cellular H2O2 accumulation in Gossypium petals

Anthocyanins are more efficient than ascorbate and tocopherol at scavenging nearly all types of ROS and free radicals [42]. Increased temperatures generate ROS and free radicals. To cope with these challenges, cotton plants may increase anthocyanin biosynthesis to scavenge ROS. We therefore used 3,3′-diaminobenzidine (DAB) staining to measure ROS (H2O2) in the petals of Gossypium species with and without beauty marks. Strikingly, the petals of beauty mark-containing Gossypium species (both in the beauty mark and non-beauty mark regions) showed significant dark DAB staining. In comparison, the petals of G. hirsutum and G. tomentosum had reduced H2O2 levels (Fig. 2).

Fig. 2
figure 2

Staining of floral petals of Gossypium species with DAB to determine the contents of ROS produced. Scale bar, 3 mm. The data were obtained from three independent replicates

We next assessed the UV reflectance in the petals of G. hirsutum P30B and G. hirsutum × G. barbadense near-isogenic lines (NILs). There were differences in the floral UV patterns but not in visible light reflectance. Moreover, NIL flowers exposed to UV light appeared significantly darker in color compared to P30B petals (Fig. 3A). Consistently, the NIL flowers showed comparatively lower levels of UV reflectance compared to P30B flowers (Fig. 3B). Altered flavonol contents and corresponding differences in UV patterns may impact physical traits (e.g., temperature) and/or physiological traits (e.g., ROS levels) in these plants (Fig. 3C, D). H2O2 levels were measured in the petals by quantifying DAB staining; this showed that NILs containing the beauty mark had considerably higher H2O2 levels compared to G. hirsutum (Fig. 3D). These data further supported the hypotheses that beauty marks and UV floral patterns were associated phenotypes, that UV exposure increased ROS generation in floral tissues, and that plants increased anthocyanin biosynthesis to mitigate the damaging effects of higher ROS levels.

Fig. 3
figure 3

Measurement of H2O2 and antioxidant contents from floral petals of P30B and NIL. A Appearance of the flowers of P30B and NIL in visible light (upper) and UV light (lower). Scale bar, 2 cm. B UV reflectance in the flowers of P30B and NIL. Value is mean S.D. (n = 10, P ≤ 0.05, one-way ANOVA, Tukey’s HSD test). C Flowers petal temperature of P30B and NIL. Value is mean S.D. (n = 10). D Total extractable flower H2O2 contents in P30B and NIL. Value is mean S.D. (n = 3). EF SOD activities and POD activities in the P30B and NIL flowers of beauty marker region (BMR) and non-beauty marker region (NBMR). Values are mean S.D. (n = 3). *P ≤ 0.05, **P ≤ 0.01. Student t-test was used to generate P value

We therefore evaluated the activities of two ROS-scavenging enzymes, superoxide dismutase (SOD) and peroxidase (POD), in P30B and NILs with and without the beauty mark. We found that SOD and POD activities were significantly increased in NIL petals compared to P30B petals (Fig. 3E, F). Moreover, SOD and POD activities were significantly higher in the beauty mark region than in the non-beauty mark region of NIL petals. These results demonstrated that the beauty mark protein GbBM contributed to ROS scavenging in G. barbadense and wild cotton plant flowers containing the beauty mark.

Natural variation in the Beauty Mark CDS affected differences in expression between G. barbadense and G. hirsutum

To determine whether SNPs in Beauty Mark were associated with the beauty mark in G. barbadense and other wild cotton species, we retrieved the CDSs of Beauty Mark from 336 G. barbadense [43] and 419 G. hirsutum accessions [44]. Nucleotide sequence alignment revealed no change in the Beauty Mark sequence in the 336 G. barbadense accessions. However, comparison of Beauty Mark between G. barbadense and G. hirsutum revealed four synonymous polymorphisms (SNP 1, 14C ˃ A; SNP 2, 490 T ˃ C; SNP 3, 494C ˃ T; and SNP 4, 696A ˃ T) and one missense polymorphism (SNP 5, 510C ˃ T). SNP5 was associated with the beauty mark phenotype; all accessions bearing the beauty mark contained the T allele, whereas accessions without the beauty mark contained the C allele. This suggested that SNP5 played an important role in controlling petal spot development.

We hypothesized that differences in Beauty Mark expression between G. barbadense and G. hirsutum could be caused by SNP5 in the CDS. We therefore conducted transient expression assays in N. benthamiana using the GbBM and GhBM promoters fused to the LUC reporter gene and GbBM and GhBM driven by the 35S promoter. The results showed that luminescence intensity was repressed in both GbBM and GhBM promoter-driven reporters that expressed the GhBM allele, whereas co-expression of GbBM with either GhBMPromoter-LUC or GbBMPromoter-LUC showed high levels of luminescence (Fig. 4A, B). Next a yeast one-hybrid assay revealed that the GAL4 transcriptional activation domain-GbBM (AD-GbBM) and GhBM (AD-GhBM) fusion protein activates the LacZ reporter gene driven by the promoters both of GbBM and GhBM, respectively (Fig. 4C). Furthermore, chromatin immunoprecipitation with quantitative PCR (ChIP-PCR) assay confirmed that both GbBM and GhBM bound to the promoters fragments of GbBM and GhBM containing the MYB core motifs in the chromatin samples (Fig. 4D, E). It has been demonstrated that several R2R3 MYBs' activity is subject to an auto-regulatory mechanism, with the resultant effect occasionally being repressive. As a component of a negative auto-regulatory loop, Arabidopsis MYB4, whose encoded protein has the ability to bind its own promoter, inhibits transcription [45]. Petunia MYB27 [16], rice OsMYB4 [46], and Arabidopsis AtMYBL2 all exhibit a similar repressive mechanism [16, 47]. Taken together, these data demonstrated that polymorphisms in the CDS could be responsible for the lack of a beauty mark in G. hirsutum. Overall, these results demonstrated that SNP5 was directly associated with the petal spot phenotype in G. barbadense.

Fig. 4
figure 4

GhBM/GbBM directly targeting its own promoter. A, B Transient transcriptional activity analysis of Beauty Mark using luciferase transient expression. Transient expression assays showing that GhBM repress the transcriptional activity of both GhBMPro:LUC and GbBMPro:LUC (A). Representative images of N. benthamiana leaves 72 h after infiltration are shown. The left panel indicates the infiltrated constructs. Scale bar = 0.5 cm. Quantitative analysis of luminescence intensity with different letters indicates significant differences (P ≤ 0.05, one-way ANOVA, Tukey’s HSD test) (B). C AD-BM activates the expression of the LacZ reporter genes driven by the promoters of respective GhBM and GbBM in yeast. Representative data are shown from one of three biological replicates, which yielded similar results. D The regions tested by ChIP assays are shown in the schematic representation. The putative MYB-core elements in the promoter of GhBM/GbBM gene are indicated by black lines. E ChIP assays indicating the association of GbBM/GhBM with several regions in the promoters of GbBM and GhBM. The promoter of GbActin1 was used for normalization. Bars represent means SD of three biological replicates. Student t-test was used to generate P value

SNP5 arose during G. hirsutum selection

To determine whether the nucleotide diversity in G. hirsutum was caused by selection pressure, we conducted Tajima’s D test on GhBM sequences. The value was significant for G. hirsutum (Tajima’s D = 3.599), indicating strong selection on the GhBM locus during G. hirsutum domestication (Fig. 5A and Table S1 and S2). To exclude the potential impact of geography on GhBM diversity, we further examined nucleotide diversity and Tajima’s D-values in the 400-kb region surrounding the GhBM locus and in GbBM in the 419 and 336 G. barbadense and G. hirsutum diversity panels, respectively, because selection may have led to a selective sweep in the flanking region (Fig. 5B, C). The linkage disequilibrium of the 1-Mb fragment containing Beauty Mark was comparatively greater in G. hirsutum than in G. barbadense (Fig. 5D, E). Taken together, these data showed that the T allele at SNP5 of GhBM in G. hirsutum was associated with the non-beauty mark phenotype in G. hirsutum cultivars, whereas the C at that position may be an ancient allele. A common evolutionary trend in plants (even in highly selfing plants) is “selfing syndrome,” in which traits associated with pollinator attraction (floral pigmentation) are lost or greatly reduced. However, it is unclear at present whether such trait reductions are favored by natural selection or result from reduced purifying selection coupled with genetic drift.

Fig. 5
figure 5

Nucleotide diversity and Tajima’ D test depict the selection pressure on GhBM during G.hirsutum domestication. A Nucleotide diversity in a fragment containing GhBM (including 3000 bp upstream, the coding region, and 3000 bp downstream). B, C To investigate whether the fragment containing GhBM was subject to selection pressure, nucleotide diversity (B) and Tajima’s D test (C) and were analyzed based on the total number of polymorphic sites in this fragment (including 200 kb upstream, the coding region, and 200 kb downstream). Gb, G. barbadense. Gh, G. hirsutum. D, E Comparison of linkage disequilibrium in the 1-Mb region around Beauty Mark between G. barbadense (D) and G. hirsutum (E)

Beauty Mark contributed to UV intensity-based geographical adaptation of Gossypium and modern cotton cultivars

Many Gossypium species are taxonomically well understood due to extensive molecular phylogenetic studies [48, 49]. To elucidate the role of the petal spot in wild cotton species and G. barbadense and the absence of the petal spot in G. hirsutism cultivars, we studied the geographic distribution of wild and cultivated cotton species under different UV exposure conditions. Geographically, Gossypium species (genomes A-G and K) were historically distributed across Northwestern, Central, and Western Australia; South and East Africa; Southeast Asia; the Galapagos Islands; Cape Verde Island; Peru; Mexico; and Southern Arizona. In contrast, accessions containing the AD genome (New World cotton) were geographically distributed across Hawaii, Brazil, the Dominican Republic, and the Galapagos Islands [50]. To examine the potential impact of geographical variation in UV radiation on the development of anthocyanin pigments, we used solar UV data (Erythema UV index) from the Tropospheric Emission Monitoring Internet Service for 2002 to 2022 (Fig. 6). The data showed that most of the regions in which cotton is economically important received the highest solar UV radiation levels during the cotton growth season, with UV index values ranging from 12 to 18. In addition, maximum UV index values were recorded in Northwestern, Central, and Western Australia; South and East Africa; Southeast Asia; the Galapagos Islands; Cape Verde Island; Peru; Mexico; and Southern Arizona. It is therefore likely that modern G. hirsutum cultivars, which are widely distributed across the globe, lack the beauty mark phenotype due to differences in petal UV patterns compared to G. barbadense and other wild cotton species that contain the beauty mark. Solar UV radiation may have led to increased anthocyanin pigmentation in cotton during its evolution. Collectively, our results suggest that environments with low latitude, high temperature, high solar UV exposure, and low precipitation may select (either directly or indirectly) for increased petal anthocyanin production, larger UV floral patterns, lower mean UV petal reflectance, and/or smaller flowers. UV floral patterns appear to confer a direct fitness advantage to individuals in high-UV environments by protecting the developing pollen and reproductive tissues from abiotic stress via increased anthocyanin biosynthesis.

Fig. 6
figure 6

Geographic distribution of wild and cultivated Gossypium species and solar UV radiation levels (2002–2022). Ga, G. arboretum. Gh, G. hirsutum. Gt, G. tomentosum. Gb, G. barbadense. Gdar, G. darwinii. Gorai, G. raimondii. Gm, G. mustelinum. Blue triangles represent accessions from G. barbadense and green triangles represent accessions from G. hirsutum


Plants have multiple methods to balance solar energy capture with tolerance mechanisms (which regulate energy consumption, dissipation, and distribution), repair mechanisms, and avoidance mechanisms (which decrease light absorbance to prevent oxidative damage) [12, 51, 52]. Anthocyanins can significantly modify both the quantity and quality of incident light [53, 54] and absorb ultraviolet, blue, and green light [55]. The absorbance of UV-blue light by anthocyanins reduces the amount of light available to other light-absorbing organelles and protects the plant [56,57,58].

The cotton species G. barbadense is characterized by a petal spot (the beauty mark) at the base of each petal, which is caused by a group of anthocyanin-pigmented cells. The cultivated upland cotton genotypes are devoid of this petal spot. We previously identified Beauty Mark, a gene on chromosome A07 that encodes an R2R3-type GbMYB113 transcription factor, as a controller of petal spot development in G. barbadense [41]. In comparing the genomic sequences of MYB113 between several cultivated and wild cotton species, we identified five SNPs in the CDS; SNP5 (C/T) emerged as a candidate cause of petal spot absence in G. hirsutum. Furthermore, we hypothesized that increased UV floral patterns could result in higher temperatures within floral tissues. This phenomenon was clearly observed in the petals of NIL flowers, and led to elevated ROS production and increases in antioxidant activities. Because the beauty mark is located at the base of each petal, we postulated that GbMB expression may also serve to protect the nearby ovary and young reproductive structures through enhanced antioxidant activity. This hypothesis was supported by significantly higher SOD and POD activity in flower petals in NILs compared to G. hirsutum. This mechanism may also partially explain the growth of G. barbadense in geographic regions with high temperatures, high UV intensity, and a long frost-free period. Further, we established the mechanism by which petal spots protect the reproductive tissues of flowers from UV light and the evolutionary role of geographic distribution in the beauty mark phenotypes of wild species and New World cotton; this may be the basis for the apparent evolutionary convergence of red non-photosynthetic pigments, i.e., the beauty mark. We have established a model illustrating that UV-rich environments induce the accumulation of anthocyanin pigmentation in UV-absorbing plant organs to reduce the availability of UV light to reproductive tissues. However, further experiments are required to support this hypothesized protective function of anthocyanins for floral reproductive tissues.

Petal spots exist in ancestral cotton landraces but are absent in most cultivated cotton varieties. Few reports are available that positively correlate the petal spot with fiber yield and quality, indicating that human selection did not cause the non-spotted petal phenotype in G. hirsutum [40]. The reasons for the negative selection of anthocyanin biosynthesis in purple rice and cotton remain unclear [59]. Regulation of anthocyanin and flavonoid biosynthesis is related to plant adaptability to environments with a high incidence of solar UV exposure. Cotton species with responses to UV light that result in higher temperatures require more anthocyanin biosynthesis to scavenge ROS due to the temperature stress. G. hirsutum cultivars showed significant differences in UV patterning and consequently did not require high anthocyanin levels in the floral reproductive tissues and therefore have a broad geographic distribution. Our findings provide novel insights into the contributions of cotton petal spots to the geographic distribution and adaptation of cotton species.


In conclusion, we found that the co-occurrence of the beauty mark and UV patterning led to significant increases in petal temperature, which in turn led to higher ROS levels. Furthermore, we demonstrated that there are differences between cotton species in absorbance and reflectance of UV light and that there are consequently variations in anthocyanin biosynthesis in floral tissues to scavenge ROS; these differences were associated with the geographic distribution of cotton species.


Plant materials and cultivation conditions

Two cultivated cotton species, G. barbadense and G. hirsutum, and nine wild cotton species (G. herbaceum, G. capitis-viridis, G. anomalum, G. davidsonii, Rozi, G. klotzschianum, G. stocksii, G. bickii, and G. tomentosum) were used in this study. Flowers were collected from wild cotton species at the National Wild Cotton Nursery in Hainan, China. Near-isogenic lines (NILs) were developed by backcrossing the G. barbadense line HaiR with the G. hirsutum line P30B eight times. UV light, ROS, and antioxidant data were collected from G. hirsutum P30B and NILs. Cotton plants used for temperature analyses and evaluation of UV light reflectance (described below) were collected from the experimental station at the Biotechnology Research Institute (BRI) of the Chinese Academy of Agricultural Sciences (CAAS) in Sanya (18°15′ N, 109° 30′ E). There were 30 cm between plants in each row and 90 cm between rows.

BM sequence analysis

Sequence fragments of ~ 7 kb (from 2 kb upstream to 500 bp downstream of the CDS), as well 400 kb (± 200 kb) region around BM was selected for sequence analysis in 336 G. barbadense and 419 G. hirsutum accessions (Table S2) [43, 44, 60]. SNPs present in this region were extracted using the vcftools from vcf files for G. hirsutum and G. barbadense [61]. Tajima’s D and nucleotide diversity analysis were conducted using vcftools. In additionally, linkage disequilibrium analysis was conducted for 400 kb region of BM using the LD block show [60]. We then analyzed the results to identify patterns and differences in Tajima’s D, nucleotide diversity, and linkage disequilibrium between the G. barbadense and G. hirsutum accessions. The results of nucleotide diversity and Tajima’s D were plotted using ggplot2 [62]. We then sequenced BM in wild cotton species with the beauty mark (G. herbaceum, G. capitis-viridis, G. anomalum, G. davidsonii, Rozi, G. klotzschianum, G. stocksii, and G. bickii) and one species without it (G. tomentosum).

Transient luciferase expression assay

The GhBM and GbBM promoter sequences were amplified from P30B and HaiR, respectively, and cloned into the pGreenII 0800-LUC reporter vector to drive firefly luciferase (LUC) expression. The Renilla luciferase (REN) gene, under the control of the cauliflower mosaic virus 35S promoter in the pGreenII 0800-LUC vector was used as the internal control. The CDSs of GhBM and GbBM were also cloned into the p2GW7 vector under the control of the 35S promoter as effectors. Agrobacterium tumefaciens strain GV3101 was transformed with each of the reporter and effector plasmids; the transformed lines were then co-infiltrated into N. benthamiana leaves. The NightSHADE LB 985 imaging system (Berthold) was used to assess LUC activity 72 h after infiltration.

Yeast one-hybrid assay

Yeast one-hybrid experiment was carried out according to the method described previously [41]. To generate AD-GbBM and AD-GhBM, the full-length GbBM and GhBM were amplified and cloned into the vector pJG4-5 vector (Clontech). To generate GbBMp::LacZ and GhBMp::LacZ reporter constructs, the promoter fragments were amplified and then separately subcloned into the pLacZi2μ vector. Primers were listed in Table S3. Transformants were cultivated on proper drop-out plates containing X-gal for blue color show. Representative data was shown from one of the three biological replicates which yield similar results.

Chromatin immunoprecipitation assays

The G. hirsutum line R15 and G. barbadense line HaiR and their transgenic calli containing the 35S:GbBM-FLAG and 35S:GhBM-FLAG constructs were used for ChIP experiment as described previously [41]. About 2 g of each calli were collected and cross-linked in 1% formaldehyde for 15 min, next neutralized with 0.125 M of glycine for 5 min. After washing with distilled water five times, the samples were ground to power in liquid nitrogen. The chromatin complexes were isolated and sonicated, and then the chromatin complex was immunoprecipitated by anti-FLAG antibody (Abmart, catalog number M20008). The precipitated DNA was analyzed with quantitative real-time PCR. Primers used for ChIP-PCR are listed in Table S3. The promoter of Actin1 was used as negative control. Three independent biological repeats were performed.

H2O2 measurements

H2O2 was extracted from fully opened flowers during peak sunlight hours, as previously described [63]. After extraction, H2O2 was quantified using the Hydrogen Peroxidase Assay Kit (Solarbio, China) following the manufacturer’s instructions. A mass spectrometer was used to measure absorbance at 415 nm.

DAB staining

The in situ detection of hydrogen peroxide was carried out in fully opened flower petals of cultivated and wild cotton species by staining with DAB using the procedure described by [64].

Determination of SOD and POD activities

SOD and POD activities were measured as previously described [63]. Total protein content was measured using a Bradford protein assay kit (Sangon Biotech, Shanghai, China).

Temperature measurements and UV reflectance scoring

Flower images were captured with a Panasonic DMC-LX5GK 24-mm wide camera. UV light was provided by a 253.7 nm UV-emitting lamp. Flowers were directly compared and scored for specific UV patterns. LightScout UV meters (3414F) and LightScout UV sensors (3637I) (Spectrum Technologies USA) with a capacity of 250–400 nm were used to measure UV radiation in an open field with a range of 0–200 μMol/m2/s (± 5%). Three UV sensors were installed at the field research station of the BRI. UV sensor data was collected using WatchDog 2000 series data loggers and analyzed with Spec Pro 9 (Spectrum Technologies USA). Flower temperatures were recorded with a temperature sensor gun (Thermo Fisher Scientific) following the manufacturer’s instructions. Temperatures were consistently measured on sunny days during the same light period.

Availability of data and materials

All data generated or analyzed during this study are included in this published article, its supplementary information files, and publicly available repositories. Quantitative data generated in this study are presented in Additional file 2.



Basic helix-loop-helix


Beauty mark


Coding sequence




Firefly luciferase


MYB–bHLH–WD repeat protein


V-myb avian myeloblastosis viral oncogene homolog


Near-isogenic lines


Reactive oxygen species


Quantitative reverse transcription polymerase chain reaction




Renilla luciferase


Single nucleotide polymorphism


Superoxide dismutase




  1. Kantsa R, Raguso R, Dyer A, Sgardelis S, Olesen J, Petanidou T. Community-wide integration of floral colour and scent in a Mediterranean scrubland. Nat Ecol Evol. 2017;1:1502–10.

    Article  PubMed  Google Scholar 

  2. Vaidya P, McDurmon A, Mattoon E, Keefe M, Carley L, Lee C, et al. Ecological causes and consequences of flower color polymorphism in a self-pollinating plant (Boechera stricta). New Phytol. 2018;218(1):380–92.

    Article  CAS  PubMed  Google Scholar 

  3. Schemske DW, Bierzychudek P. Perspective, evolution of flower color in the desert annual Linanthus parryae, Wright revisited. Evolution. 2001;55(7):1269–82.

    CAS  PubMed  Google Scholar 

  4. Warren J, Mackenzie S. Why are all color combinations not equally represented as flower-color polymorphisms? New Phytol. 2001;151(1):237–41.

    Article  PubMed  Google Scholar 

  5. Coberly L, Rausher M. Analysis of a chalcone synthase mutant in Ipomoea purpurea reveals a novel function for flavonoids, amelioration of heat stress. Mol Ecol. 2003;12(5):1113–24.

    Article  CAS  PubMed  Google Scholar 

  6. Arista M, Talavera M, Berjano R, Ortiz PL. Abiotic factors may explain the geographical distribution of flower color morphs and the maintenance of color polymorphism in the scarlet pimpernel. J Ecol. 2013;101(6):1613–22.

    Article  Google Scholar 

  7. Agati G, Tattini M. Multiple functional roles of flavonoids in photoprotection. New Phytol. 2010;186(4):786–93.

    Article  CAS  PubMed  Google Scholar 

  8. Dixon RA, Paiva NL. Stress-induced phenylpropanoid metabolism. Plant Cell. 1995;7:1085–97.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Saito K, Yonekura-Sakakibara K, Nakabayashi R, Higashi Y, Yamazaki M, Tohge T, et al. The flavonoid biosynthetic pathway in Arabidopsis: structural and genetic diversity. Plant Physiol Biochem. 2013;72:21–34.

    Article  CAS  PubMed  Google Scholar 

  10. Hopkins R, Rausher MD. Identification of two genes causing reinforcement in the Texas wildflower Phlox drummondii. Nature. 2011;469:411–4.

    Article  CAS  PubMed  Google Scholar 

  11. Winkel-Shirley B. Flavonoid biosynthesis. A colorful model for genetics, biochemistry, cell biology, and biotechnology. Plant Physiol. 2001;126(2):485–93.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  12. Steyn W, Wand S, Holcroft D, Jacobs G. Anthocyanins in vegetative tissues: a proposed unified function in photoprotection. New Phytol. 2002;155(3):349–61.

    Article  CAS  PubMed  Google Scholar 

  13. Liang J, He J. Protective role of anthocyanins in plants under low nitrogen stress. Biochem. 2018;498(4):946–53.

    CAS  Google Scholar 

  14. Nakabayashi R, Yonekura-Sakakibara K, Urano K, Suzuki M, Yamada Y, Nishizawa T, et al. Enhancement of oxidative and drought tolerance in Arabidopsis by over accumulation of antioxidant flavonoids. Plant J. 2014;77:367–79.

    Article  CAS  PubMed  Google Scholar 

  15. Dixon RA, Liu C, Jun J. Metabolic engineering of anthocyanins and condensed tannins in plants. Curr Opin Biotechnol. 2013;24:329–35.

    Article  CAS  PubMed  Google Scholar 

  16. Albert NW, Davies KM, Lewis DH, Zhang H, Montefiori M, Brendolise C, et al. A conserved network of transcriptional activators and repressors regulates anthocyanin pigmentation in eudicots. Plant Cell. 2014;26:962–80.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Gonzalez A, Zhao M, Leavitt JM, Lloyd AM. Regulation of the anthocyanin biosynthetic pathway by the TTG1/bHLH/Myb transcriptional complex in Arabidopsis seedlings. Plant J. 2008;53:814–27.

    Article  CAS  PubMed  Google Scholar 

  18. Quattrocchio F, Wing J, Van der Woude K, Souer E, de Vetten N, Mol J, et al. Molecular analysis of the anthocyanin2 gene of petunia and its role in the evolution of flower color. Plant Cell. 1999;11:1433–44.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Mathews H, Clendennen SK, Caldwell C, Liu XL, Connors K, Matheis N. Activation tagging in tomato identifies a transcriptional regulator of anthocyanin biosynthesis, modification, and transport. Plant Cell. 2003;15(8):1689–703.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Deluc L, Bogs J, Walker AR, Ferrier T, Decendit A, Merillon JM, et al. The transcription factor VvMYB5b contributes to the regulation of anthocyanin and proanthocyanidin biosynthesis in developing grape berries. Plant Physiol. 2008;147(4):2041–53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Paz-Ares J, Ghosal D, Wienand U, Peterson P, Saedler H. The regulatory c1 locus of Zea mays encodes a protein with homology to myb proto-oncogene products and with structural similarities to transcriptional activators. EMBO J. 1987;6(12):3553–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Zhang Y, Cheng S, Jong D, Griffiths H, Halitschke R, De Jong W. The potato R locus codes for dihydroflavonol 4-reductase. Theor App Genet. 2009;119(5):931–7.

    Article  CAS  Google Scholar 

  23. Chu H, Jeong JC, Kim WJ, Chung DM, Jeon HK, Ahn YO, et al. Expression of the sweet potato R2R3-type IbMYB1a gene induces anthocyanin accumulation in Arabidopsis. Physiol Planta. 2013;148(2):189–99.

    Article  CAS  Google Scholar 

  24. Espley RV, Hellens RP, Putterill J, Stevenson DE, Kutty-Amma S, Allan AC. Red colouration in apple fruit is due to the activity of the MYB transcription factor, MdMYB10. Plant J. 2007;49(3):414–27.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Aharoni A, De Vos C, Wein M, Sun Z, Greco R, Kroon A, et al. The strawberry FaMYB1 transcription factor suppresses anthocyanin and flavonol accumulation in transgenic tobacco. Plant J. 2001;28(3):319–32.

    Article  CAS  PubMed  Google Scholar 

  26. Tamagnone L, Merida A, Parr A, Mackay S, Culianez-Macia FA, Roberts K, et al. The AmMYB308 and AmMYB330 transcription factors from Antirrhinum regulate phenylpropanoid and lignin biosynthesis in transgenic tobacco. Plant Cell. 1998;10(2):135–54.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Kui LW, Micheletti D, Palmer J, Volz R, Lozano L, Espley R, et al. High temperature reduces apple fruit color via modulation of the anthocyanin regulatory complex. Plant Cell Environ. 2011;34(7):1176–90.

    Article  CAS  Google Scholar 

  28. Cavallini E, Matus JT, Finezzo L, Zenoni S, Loyola R, Guzzo F, et al. The phenylpropanoid pathway is controlled at different branches by a set of R2R3-MYB C2 repressors in grapevine. Plant Physiol. 2015;167(4):1448–70.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Jin H, Cominelli E, Bailey P, Parr A, Mehrtens F, Jones J, et al. Transcriptional repression by AtMYB4 controls production of UV-protecting sunscreens in Arabidopsis. EMBO J. 2000;19(22):6150–61.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Albert N, Lewis D, Zhang H, Schwinn K, Jameson P, Davies K. Members of an R2R3-MYB transcription factor family in Petunia are developmentally and environmentally regulated to control complex floral and vegetative pigmentation patterning. Plant J. 2011;65(5):771–84.

    Article  CAS  PubMed  Google Scholar 

  31. Xu F, Ning Y, Zhang W, Liao Y, Li L, Cheng H, et al. An R2R3-MYB transcription factor as a negative regulator of the flavonoid biosynthesis pathway in Ginkgo biloba. Funct Integ Genom. 2014;14(1):177–89.

    Article  Google Scholar 

  32. Van Kleunen M, Nänni I, Donaldson JS, Manning JC. The role of beetle marks and flower colour on visitation by monkey beetles (Hopliini) in the greater cape floral region, South Africa. Ann Bot. 2007;100:1483–9.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Schiestl FP, Johnson SD. Pollinator-mediated evolution of floral signals. Trends Ecol Evol. 2013;28(5):307–15.

    Article  PubMed  Google Scholar 

  34. Koski M, Ashman T. Macro-evolutionary patterns of ultraviolet floral pigmentation explained by geography and associated bioclimatic factors. New Phytol. 2016;211(2):708–18.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Pollastri S, Tattini M. Flavonols: old compounds for old roles. Ann Bot. 2011;108:1225–33.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Close D, Beadle C. The ecophysiology of foliar anthocyanin. Bot Rev. 2003;69(2):149–61.

    Article  Google Scholar 

  37. Nishio JN. Why are higher plants green? Evolution of the higher plant photosynthetic pigment complement. Plant Cell Environ. 2000;23(6):539–48.

    Article  CAS  Google Scholar 

  38. Sheehan H, Michel M, Ulrich K, Korinna E, Alexandre D, et al. MYB-FL controls gain and loss of floral UV absorbance, a key trait affecting pollinator preference and reproductive isolation. Nat Genet. 2016;48(2):159–69.

    Article  CAS  PubMed  Google Scholar 

  39. Fryxell P. Taxonomy and germplasm resources. In: Kohel RJ, Lewis CF, editors. Cotton. Madison: Am. Soc. Agron.; 1984. p. 27–58.

    Google Scholar 

  40. Ahuja S, Dhayal L. Comparative characteristics and gene action in three petal-spotted mutants of Gossypium hirsutum. J Genetic. 2007;86(1):81–4.

    Article  CAS  Google Scholar 

  41. Abid MA, Wei Y, Meng Z, Wang Y, Ye Y, Wang Y, et al. Increasing floral visitation and hybrid seed production mediated by beauty mark in Gossypium hirsutum. Plant Biotechnol J. 2022;20(7):1274–84.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Rice-Evans C, Miller N, Paganga G. Antioxidant properties of phenolic compounds. Trends Plant Sci. 1997;2(11):152–9.

    Article  Google Scholar 

  43. Zhao N, Wang W, Grover CE, Jiang K, Pan Z, Guo B, et al. Genomic and GWAS analyses demonstrate phylogenomic relationships of Gossypium barbadense in China and selection for fibre length, lint percentage and Fusarium wilt resistance. Plant Biotechnol J. 2022;20:691–710.

    Article  CAS  PubMed  Google Scholar 

  44. Ma Z, He S, Wang X, Sun J, Zhang Y, Zhang G, et al. Resequencing a core collection of upland cotton identifies genomic variation and loci influencing fiber quality and yield. Nat Genet. 2018;50:803–13.

    Article  CAS  PubMed  Google Scholar 

  45. Zhao J, Zhang W, Zhao Y, Gong X, Guo L, Zhu G, et al. SAD2, an importin β-like protein, is required for UV-B response in Arabidopsis by mediating MYB4 nuclear trafficking. Plant Cell. 2007;19(11):3805–18.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Baldoni E, Genga A, Medici A, Coraggio I, Locatelli F. The OsMyb4 gene family: stress response and transcriptional auto-regulation mechanisms. Biol Plant. 2013;57(4):691–700.

    Article  CAS  Google Scholar 

  47. Dubos C, Le Gourrierec J, Baudry A, Huep G, Lanet E, Debeaujon I, et al. MYBL2 is a new regulator of flavonoid biosynthesis in Arabidopsis thaliana. Plant J. 2008;55(6):940–53.

    Article  CAS  PubMed  Google Scholar 

  48. Wendel J, Cronn RC. Polyploidy and the evolutionary history of cotton. Adv Agron. 2003;78:139–86.

    Article  Google Scholar 

  49. Wendel J, Grover CE. Taxonomy and evolution of the cotton genus. In: Fang D, Percy R, editors. Cotton, Agronomy Monograph 24. Madison: ASA-CSSA-SSSA; 2015. p. 25–44.

    Google Scholar 

  50. Wang K, Wendel J, Hua J. Designations for individual genomes and chromosomes in Gossypium. J Cotton Res. 2018;1:3.

    Article  Google Scholar 

  51. Demmig-Adams B, Adams WW. Photoprotection and other responses of plants to high light stress. Annu Rev Plant Physiol Plant Mol Biol. 1992;43:599–626.

    Article  CAS  Google Scholar 

  52. Niyogi KK. Photoprotection revisited: genetic and molecular approaches. Annu Rev Plant Physiol Plant Mol Biol. 1999;50:333–59.

    Article  CAS  PubMed  Google Scholar 

  53. Krol M, Gray G, Hurry V, Öquist G, Malek L, Huner N. Low-temperature stress and photoperiod effect an increased tolerance to photoinhibition in Pinus banksiana seedlings. Can J Bot. 1995;73:1119–27.

    Article  CAS  Google Scholar 

  54. Ntefidou M, Manetas Y. Optical properties of hairs during the early growth stages of leaf development in Platanus orientalis. Aust J Plant Physiol. 1996;23(4):535–8.

    Google Scholar 

  55. McClure JW. Physiology and functions of flavonoids. In: Harborne JB, Mabry TJ, Mabry H, editors. The Flavonoids. London: Chapman & Hall Ltd; 1975. p. 970–1055.

    Chapter  Google Scholar 

  56. Pietrini F, Massacci A. Leaf anthocyanin content changes in Zea mays L. grown at low temperature: significance for the relationship between quantum yield of PS II and the apparent quantum yield of CO2 assimilation. Photosynth Res. 1998;58:213–9.

    Article  CAS  Google Scholar 

  57. Smillie RM, Hetherington SE. Photoabatement by anthocyanin shields photosynthetic systems from light stress. Photosynthetica. 1999;36:451–63.

    Article  CAS  Google Scholar 

  58. Neill S, Gould KS. Optical properties of leaves in relation to anthocyanin concentration and distribution. Can J Bot. 1999;77(12):1777–82.

    Article  Google Scholar 

  59. Zheng J, Wu H, Zhu H, Huang C, Liu C, Chang Y, et al. Determining factors, regulation system, and domestication of anthocyanin biosynthesis in rice leaves. New Phytol. 2019;223(2):705–21.

    Article  CAS  PubMed  Google Scholar 

  60. Dong S, He W, Ji J, Zhang C, Guo Y, Yang T. LDBlockShow: a fast and convenient tool for visualizing linkage disequilibrium and haplotype blocks based on variant call format files. Brief Bioinform. 2021;22(4):1–6.

    Article  CAS  Google Scholar 

  61. Danecek P, Auton A, Abecasis G, Albers C, Banks E, DePristo M, et al. The variant call format and VCFtools. Bioinformatics. 2011;27(15):2156–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  62. Ginestet C. ggplot2: elegant graphics for data analysis. J R Stat Soc B. 2011;174:245–6.

    Article  Google Scholar 

  63. Wang Y, Liang C, Meng Z, Li Y, Abid MA, Askari M, et al. Leveraging Atriplex hortensis choline monooxygenase to improve chilling tolerance in cotton. Environ Exp Bot. 2019;162:364–73.

    Article  CAS  Google Scholar 

  64. Daudi A, O’Brien JA. Detection of hydrogen peroxide by DAB staining in Arabidopsis leaves. Bio Protoc. 2012;2(18):e263.

    Article  PubMed  Google Scholar 

Download references


We thank Prof. Fang Liu and Prof. Kunbo Wang (Institute of Cotton Research, CAAS, Anyang, China) for providing the wild cotton species flowers.


This study was supported by the National Natural Science Foundation of China (32072115 and 31771850) and the Agricultural Science and Technology Innovation Program of CAAS.

Author information

Authors and Affiliations



M.A.A., R.Z., and C.L. conceived the project and designed the experiments. M.A.A., Q.Z., M.A., H.H., Z.M., Y.Wang., S.G., and C.L. prepared, integrated, and performed the experiments. M.A.A., Q.Z., and C.L. performed the yeast one-hybrid assay experiment, ChIP-PCR experiment, and gene expression analysis. M.A.A., Q.Z., M.A., Y.Wei., R.Z., and C.L. performed and reviewed data analysis M.A.A., M.A., and C.L. prepared figures for publication. M.A.A., M.A., and C.L. prepared the original draft. M.A.A. and C.L. reviewed and edited the draft. R.Z. and C.L. provided resources and funding. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Rui Zhang or Chengzhen Liang.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1:

Table S1. Tajima’s D test for GhBM and GbBM sequences. Table S2. Cotton accessions used for sequence analysis. Table S3. Sequences of primers used in this study.

Additional file 2.

Includes individual values for quantitative data depicted in Figs. 1, 3 and 4.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Abid, M.A., Zhou, Q., Abbas, M. et al. Natural variation in Beauty Mark is associated with UV-based geographical adaptation in Gossypium species. BMC Biol 21, 106 (2023).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:


  • Gossypium
  • Beauty Mark
  • UV light
  • Anthocyanin
  • Geographical adaptation