Evolutionarily conserved but not indispensable for survival: plastic UV RESISTANCE LOCUS8 functionality in Marchantia polymorpha under laboratory and field conditions

Abstract

The functionality of the UV-B photoreceptor UV RESISTANCE LOCUS8 (UVR8) has been widely demonstrated in angiosperms under controlled conditions, but studies in nature or using bryophytes are scarce. To fill this gap, we compared the UV responses of the model liverwort Marchantia polymorpha and a mutant lacking a functional UVR8, to provide for the first time an evolutionary perspective of UVR8 performance in both controlled and field conditions. Under exposure to UV in the laboratory, UVR8 in wild type plants was essential to induce the expression of genes (CHS1, FNSI, HY5, SIG5, and ELIP2) and the accumulation of metabolites (apigenin and luteolin derivatives, particularly apigenin 7-O-glucuronide) that contributed to conserving photosynthetic performance. In the mutant, gene and metabolite responses were much weaker, occurred only under high UV-B in a UVR8-independent response, and did not confer photoprotection to the photosynthetic apparatus. In a field seasonal cycle, differences between UV treatments and genotypes were attenuated and the UVR8 functionality was less clear, probably due to the low UV-B levels required to trigger plant responses and the interaction of other abiotic factors. Given the similarity of UVR8 functionality in M. polymorpha and Arabidopsis under both controlled and field conditions, we conclude that UVR8 has been conserved across plant evolution.

The UV-B photoreceptor UVR8 stimulates gene expression and flavone accumulation in Marchantia under laboratory and field conditions, enabling plants to maintain photosynthetic performance, and its functionality has been conserved across plant evolution.

Introduction

UV radiation represents a small fraction of the solar spectrum, but has had a crucial impact on plant evolution (Robson et al., 2019; Fürst-Jansen et al., 2020). It is divided into three wavelength bands: UV-A (315–400 nm), UV-B (280–315 nm), and UV-C (100–280 nm). Given that UV-C does not reach the Earth’s surface, it has received less attention than UV-A and UV-B in plant biological studies. Until recently, UV radiation was considered simply as a stressful factor for plants because, in excess, it can damage DNA and the photosynthetic apparatus. However, more recently both UV-B and UV-A have been identified as key plant regulators, participating in multiple morphological and biochemical processes (Jenkins, 2017; Verdaguer et al., 2017; Robson et al., 2019; Pescheck and Bilger, 2020; Podolec et al., 2021; Martínez-Abaigar and Núñez-Olivera, 2022). In particular, UV induces protection mechanisms that attenuate the damaging effects of its exposure, such as DNA repair, antioxidant systems, and the accumulation of UV-absorbing compounds (UVACs). In embryophytes, UVACs are predominantly phenolic in nature and perform crucial functions as both antioxidants and UV screens (Agati et al., 2025). UV radiation concomitantly stimulates the expression of genes encoding key enzymes of the phenolic pathway, such as PHENYLALANINE AMMONIA-LYASE (PAL), CHALCONE SYNTHASE (CHS), and CHALCONE ISOMERASE (CHI) (Pandey et al., 2023). Although UV-A is present in natural sunlight at levels around 30-fold higher than UV-B (Aphalo et al., 2012), accumulation of UVACs can be triggered by both wavelengths, whereas effects on growth and development can be different, reflecting diverse perception mechanisms with multiple photoreceptors (Verdaguer et al., 2017; Rai et al., 2020; Morales et al., 2025).

The discovery of the UV-B photoreceptor UV RESISTANCE LOCUS 8 (UVR8: Jenkins, 2017; Podolec et al., 2021) was a major advance in understanding how photosynthetic organisms from green algae to flowering plants orchestrate UV responses (Tilbrook et al., 2016; Soriano et al., 2018; Han et al., 2019). Under laboratory conditions, UVR8 is involved in the regulation of expression of genes involved in UVAC synthesis and the mitigation of the damaging effects of UV-B on the photosynthetic apparatus (Jenkins, 2009; Davey et al., 2012; Morales et al., 2013; Podolec et al., 2021). However, there are only limited data on its functionality under field conditions. Studies in this regard have been performed by exposing Arabidopsis mutants to different UV treatments at time-scales from hours (Morales et al., 2013; Rai et al., 2020) to days (Findlay and Jenkins, 2016; Coffey et al., 2017) and to months (Stockenhuber et al., 2024). In addition, Neugart et al. (2024) performed a large-scale field experiment in which the wild type and a UVR8-knockout mutant were exposed to UV across a wide range of latitudes. All these studies revealed the significance of UVR8 under natural conditions, particularly in regulating gene expression, rosette size, fruit production, and content of phenolic UVACs.

Most studies on the effects of UV on plants have been performed on angiosperms, but examination of bryophytes (a monophyletic group including hornworts, liverworts, and mosses) can significantly increase our understanding of the evolution of plant UV responses (Martínez-Abaigar and Núñez-Olivera, 2022). Extant bryophytes share ancient common ancestors that were among the first embryophyte lineages to colonize the terrestrial environment, where plants had to face (among other challenges) higher UV levels than in the primitive aquatic environment (Robson et al., 2019). In addition, the simpler structure of bryophytes in comparison with angiosperms might lead to the prioritization of non-structural protection strategies, such as accumulation of UVACs (Martínez-Abaigar and Núñez-Olivera, 2022). In this context, the model liverwort Marchantia polymorpha might be particularly relevant because it accumulates apigenin and luteolin derivatives in response to UV (Soriano et al., 2019; Davies et al., 2020). These UVAC responses are mediated by UVR8, whose presence and functionality has been demonstrated in bryophytes in the laboratory (Clayton et al., 2018; Soriano et al., 2018; Kondou et al., 2019). However, there is no information on bryophyte UVR8 functionality under field conditions. In addition, an evolutionary perspective on this functionality in nature, from bryophytes to angiosperms, is still lacking.

Our aim for this study was to evaluate the role of the UVR8 photoreceptor on the UV responses of M. polymorpha under both controlled and field conditions. We used two different genotypes, a wild type and a UVR8-knockout mutant lacking a functional UVR8 protein, and we applied a wide range of UV treatments, both in the laboratory and (for the first time to our knowledge) in the field, including a seasonal cycle. Given that the potentially different effects of UV-A and UV-B have been little studied in bryophytes, specific UV-A treatments were also included in the experimental design. Overall, our study provides new approaches to understand the evolution of UVR8 functionality in plants and how it interacts with temperature and other environmental factors in nature in the current context of climate change.

Materials and methods

Plant material and general culture conditions

The Marchantia polymorpha subsp. ruderalis accession Takaragaike-1 (referred to as the wild type, WT) was kindly provided by Professor Takayuki Kochi (Kyoto University). We also used a loss-of-function mutant for the UV-B photoreceptor UVR8, uvr8. This is the same male line previously reported by Kondou et al. (2019), who confirmed that its reduced growth under UV-B conditions and attenuated UV-B-induced gene expression are specifically attributable to the loss of the gene. These two genotypes were used for both laboratory and field experiments (Supplementary Fig. S1). For the laboratory experiments, gemmae were sown in half-strength Gamborg’s B5 medium on Petri dishes (Supplementary Fig. S2) in a growth chamber at 22 °C and 70% relative humidity. In the field experiment, gemmae of both genotypes were sown in pressed peat pellets (Jiffy Products International B.V., Norway). To maintain an optimal state of hydration for the plants, an automatic irrigation system and a misting system were used (Supplementary Fig. S2).

Radiation regimes

Different combinations of lamps and filters were used to generate different treatments in the experiments (Supplementary Table S1). The spectral characteristics of the treatments are shown in Supplementary Fig. S3. Spectral irradiances were measured with a Macam SR9910 spectroradiometer (Macam Photometrics Ltd, Livingstone, UK).

Laboratory culture: UV-B fluence rate assays

Plants were grown for 3 weeks under a low fluence rate (20 µmol m⁻² s⁻¹) of photosynthetically active radiation (PAR). Then, six different UV-B photon fluence rates (0, 0.75, 1.25, 2.50, 6.25, and 10.00 µmol m⁻² s⁻¹) were applied for 4 h using UV-B narrowband lamps (Supplementary Table S1), which had a maximum emission peak at 313 nm. For the 0 µmol m⁻² s⁻¹ treatment, a transparent polymethacrylate filter that completely blocked UV-B radiation was used (Vitroflex 395, Polimer Tecnic, Girona, Spain), whilst a filter blocking UV-C radiation was used for the other treatments (Ultraphan 295, Digefra GmbH, Munich, Germany).

Laboratory culture: responses and acclimation to low and high UV

To compare the responses of the two genotypes to low- and high-UV doses, 16 gemmae per Petri dish were sown in a regular 4×4 grid and placed in a growth room under a 10/14 h light/dark photoperiod. PAR, UV-A, and UV-B radiation treatments were provided using the lamps and filters described in Supplementary Table S1. Two different irradiance conditions of low and high UV (Supplementary Table S2) were applied for 5 weeks, after which the first sampling was performed. The plants were then all cultured for a further 3 weeks under high UV, after which a sampling was performed in order to study the effect of the acclimation to low or high UV. During both the 5 week and 3 week periods, four radiation regimes were imposed using different cut-off filters, as follows. (1) P (PAR alone) using Vitroflex 395, which cut off all UV radiation. (2) PA (PAR+UV-A) using Folex 320 (Folex, GmbH, Dreieich, Germany), which cut off UV-C and UV-B radiation. (3) PB (PAR + UV-B), using Ultraphan 295 and switching on only the PAR and UV-B lamps. (4) PAB (PAR+UV-A+UV-B), using Ultraphan 295, which cut off UV-C radiation. Samples were collected at the end of both cultivation periods.

Field culture: seasonal changes

The experiments were carried out consecutively through winter, spring, and summer in the experimental field of the University of La Rioja (Logroño, northern Spain, 42°27′N, 2°25′W, 373 m elevation). Due to the different climatic conditions of each season, the plants took different times to reach a standard state for harvesting: 100 d for winter, 30 d for spring, and 64 d for summer. In each experiment, the following five different radiation regimes were applied using combinations of filters and lamps (Supplementary Fig. S2; Supplementary Table S1). (1) PAB, using Vitroflex 295 (Polimer Tecnic). Among the regimes used, this was the most similar to the natural solar radiation. (2) Pab (reduced UV-A and UV-B), using Vitroflex 395. (3) PAb (reduced UV-B), using acetate Folex 320. As this filter was flexible, it was complemented with an additional polymethacrylate rigid filter (Vitroflex 295) to prevent tearing of the acetate film by adverse meteorological conditions. (4) PaB (reduced UV-A), using a Vitroflex 395 filter and UV-B lamps, which were switched on for 10 min periods around noon to provide the plants with the same UV-B that they would have received if exposed to ambient sunlight. (5) PAB↑ (enhanced UV-B), using the same filter as in PAB and the same lamps as in PaB but providing 10% higher total UV-B than that received in the PAB treatment by adjusting the time of functioning of the lamps.

The ambient PAR, UV-A, and UV-B irradiances were continuously recorded close to the experimental plot using Skye Quantum SKP 215, SKU 420, and SKU 430 broadband sensors, respectively (Skye Instruments Ltd., Llandrindod Wells, UK). Total doses of the different radiations received by the plants during the period of study are shown in Supplementary Table S3. Temperature data during the culture period were obtained from the Gobierno de La Rioja database (https://www.larioja.org/agricultura/es/informacion-agroclimatica/red-estaciones). The average minimum, mean, and maximum temperatures for each experimental period were: winter, 3.7/7.4/11.1 °C; spring, 11.3/16.5/21.7 °C; and summer, 13.6/19.3/25.0 °C.

Gene expression analysis

For the UV-B fluence rate assays, samples were collected at the end of the 4 h exposure period. In the laboratory culture experiment, samples were collected at midday on the 35th day of culture. In the field culture experiment, samples were harvested at noon on the penultimate day of each culture period. Samples consisted of healthy whole plants pooled together and were immediately frozen in liquid nitrogen after harvest. They were subsequently ground using a TissueLyser II (Qiagen), and total RNA was extracted using a Plant RNeasy Mini Kit (Qiagen). DNase treatment was done using TURBO DNase-free (Ambion) to remove any gDNA contamination. cDNA was then synthesized with SuperScript IV Reverse Transcriptase (ThermoFisher Scientific) using the same amount of RNA for each sample. For qPCR analysis of gene expression, an equal amount of cDNA was used. Transcript levels of each gene studied were measured using a Bio-Rad iCycler and the specific primers listed in Supplementary Table S4, with ACTIN7 as the reference gene. The stable expression of ACTIN7 has previously been demonstrated in M. polymorpha through systematic reference-gene validation analyses (Saint-Marcoux et al., 2015), and it is widely used as a reliable normalization control in molecular studies of this species (Clayton et al., 2018; Soriano et al., 2018). Its stability under the conditions of the present study was further supported by an analysis of its C_T values in the seasonal field experiment, which represented the most environmentally variable part of our study (Supplementary Table S5). We assayed transcripts of genes encoding the flavonoid biosynthesis enzymes CHALCONE SYNTHASE 1 (CHS1) and a putative flavone synthase I (FNSI; see Li et al., 2020); the transcription factor ELONGATED HYPOCOTYL5 (HY5), which regulates many UV-B responses; a photolyase that repairs DNA damage in chloroplasts and mitochondria (CRY-DASH); the multiple-stress responsive plastid sigma factor (SIG5); and an early light-inducible pigment-binding protein (specifically EARLY LIGHT-INDUCED PROTEIN 2, ELIP2).

Analysis of UV-absorbing compounds

Analysis of methanol-soluble UV-absorbing compounds (SUVACs) was performed following the method of Soriano et al. (2019) and the bulk level of SUVACs was obtained as the area under the absorbance curve in the interval 280–400 nm (AUC_280–400). The individual phenolic compounds were analysed by ultra-performance liquid chromatography (UPLC) using a Waters Acquity UPLC system following Soriano et al. (2019). For the UV-B fluence rate assays, analysis of SUVACs was performed at 3 d and 7 d after UV-B treatment. In the laboratory culture experiment, samples were collected at midday on the 35th and 56th days of culture (first and second sampling, respectively). Finally, in the field culture experiment, samples were collected on the final day of each treatment.

Chlorophyll fluorescence and photosynthesis measurements

Maximum (F_m) and minimum (F₀) chlorophyll fluorescence values were measured using a portable pulse amplitude modulation fluorometer (PAM2500, Walz). The maximum quantum yield of PSII was then determined as F_v/F_m, where F_v=F_m−F₀. In the laboratory culture experiment, measurements were performed before the lights were switched on (pre-dawn) and at midday, and they were taken after 5 weeks and 8 weeks of culture. In the field experiment, measurements were conducted on the final day of treatment, at pre-dawn and noon. In the field experiment, net assimilation of CO₂ (A_N) was recorded in situ under steady-state conditions using an infrared open gas-exchange system (LI-6400, LI-COR Inc.), with measurements being taken in the morning prior to the final harvest of each season. All measurements were performed under ambient conditions of humidity and temperature, using a saturating photosynthetic photon flux density (PPFD) of 500 μmol m⁻² s⁻¹ and a CO₂ concentration of 500 ppm.

Determination of photosynthetic pigments

Fresh plants (∼250 mg) were collected and immediately frozen in liquid nitrogen. After grinding, pigments were extracted with 100% acetone, and their contents were analysed using an Agilent HP1100 HPLC system and an Agilent photodiode array detector (Soriano et al., 2019). Chlorophylls a and b, β-carotene, and xanthophylls were determined and quantified using commercial standards. The Chl a/b ratio was calculated together with the xanthophyll index: (antheraxanthin+zeaxanthin)/(violaxanthin+antheraxanthin+zeaxanthin).

Statistical analysis

The number of replicates used for each treatment and genotype depended on both the experiment and variable considered. In the fluence rate assays, we used three replicates for gene expression and four or five for UV-absorbing compounds. In the laboratory culture experiment and the field experiment, we used three replicates for all the variables. All the statistical procedures were performed using SPSS 24.0 for Windows. Pairwise differences between the WT and uvr8 were examined using Student’s t-test. Once it was determined that the data met the assumptions of normality (Shapiro–Wilk’s test) and homoscedasticity (Levene’s test), one-way ANOVA was used to compare the effect of the different treatments within each experiment and genotype. In the field experiment, the global effects of season, genotype, and treatment on the different variables were tested using three-way ANOVA. Where significant differences were detected by the ANOVAs, means were then compared using Tukey’s test. Results of the field experiment for the contents of phenolic compounds and F_v/F_m were ordinated through principal component analysis (PCA).

Results

Gene expression responses to UV-B fluence rates in the laboratory

Transcript levels were assayed for plants grown for 3 weeks under low PAR and exposed to UV-B for 4 h, and the gene expression profiles varied according to both the genotype and the UV-B fluence applied (Fig. 1). There were no differences between genotypes under control conditions in the absence of UV-B. Expression of genes encoding CHS1, FNSI, HY5, and SIG5 was significantly higher in the WT than in the uvr8 mutant at all the UV-B fluence rates. For ELIP2, these differences appeared at 2.5 µmol m⁻² s⁻¹ and above. In WT plants, the flavonoid-related transcripts CHS1 and FNSI showed a linear increase across the lower fluence rates (0–2.5 µmol m⁻² s⁻¹) and then reached a saturation level (Supplementary Fig. S4). In the case of HY5 and SIG5, transcript accumulation remained stable across all the UV-B fluence rates. In the uvr8 mutant, a small accumulation of some transcripts (CHS1, FNSI, HY5, ELIP2) was found only at the highest fluence rates. Transcript levels of CRY-DASH were higher in the WT than in the uvr8 mutant at lower fluence rates, whereas similar values were found at higher fluence rates.

Fig. 1.

Effects of different UV-B fluence rates on expression of genes related to physiological responses in the Marchantia polymorpha wild type (WT) and a UVR8-knockout mutant (uvr8). Plants were grown for 3 weeks under low PAR conditions (20 μmol m⁻² s⁻¹) and then exposed for 4 h to different UV-B fluence rates. Gene expression was determined by qPCR using ACTIN7 as the internal reference and is shown relative to zero UV-B for each genotype, the values of which were set as 1. CHS1, CHALCONE SYNTHASE 1; FNSI, putative flavone synthase I; HY5, ELONGATED HYPOCOTYL5; CRY-DASH, photolyase that repairs DNA damage in chloroplasts and mitochondria; ELIP2, EARLY LIGHT-INDUCED PROTEIN 2; SIG5, multiple-stress responsive plastid sigma factor. Data are means (±SE), n=3. Different letters indicate significant differences among fluence rates for the WT (lowercase) and the mutant (uppercase) as determined using one-way ANOVA followed by Tukey’s test (P<0.05). Significant differences between the genotypes at each fluence rate were determined using Student’s t-test: *P<0.05, **P<0.01, ***P<0.001.

Flavone accumulation under different UV-B fluence rates in the laboratory

At 3 d after the 4 h exposure to UV-B irradiation, the content of the major flavone apigenin 7-O-glucuronide showed no difference between the genotypes under control conditions, whereas WT plants showed significantly higher contents than the mutant across the other fluence rates (Fig. 2A). Furthermore, strongly significant and positive linear regressions were obtained between the contents and the transcript levels of HY5, CHS1, and FNSI across the two genotypes (Fig. 2B). A similar relationship was found between apigenin 7-O-glucuronide content and the fluence rate in the WT, but not in the uvr8 mutant (Fig. 2C). Minor flavones showed more variable responses to genotype and fluence than apigenin 7-O-glucuronide (Supplementary Fig. S5).

Fig. 2.

Effects of different UV-B fluence rates on accumulation of apigenin 7-O-glucuronide in the Marchantia polymorpha wild type (WT) and a UVR8-knockout mutant (uvr8). Plants were grown for 3 weeks under low PAR conditions (20 μmol m⁻² s⁻¹), then exposed for 4 h to different UV-B fluence rates, and samples were taken 3 d later for determination using UPLC-MS. (A) Content of apigenin 7-O-glucuronide in the plants. Data are means (±SE), n=4-5. Different letters indicate significant differences among fluence rates for the WT (lowercase) and the mutant (uppercase) as determined using one-way ANOVA followed by Tukey’s test (P<0.05). Significant differences between the genotypes at each fluence rate were determined using Student’s t-test: **P<0.01, ***P<0.001. (B) Regressions between the contents of apigenin 7-O-glucuronide and the transcript levels of CHS1, FNSI, and HY5 (see Fig. 1) across both genotypes. (C) Regressions between the contents of apigenin 7-O-glucuronide and the UV-B fluence rate in the WT and uvr8 mutant. The significance of each regression is indicated: P<0.001.

At 7 d after the UV-B treatment, apigenin 7-O-glucuronide showed similar differences between the genotypes as observed at 3 d, with content increasing with increasing UV-B (Supplementary Fig. S6). Meanwhile, the minor flavones showed clearer differences between the genotypes than were seen at 3 d, with generally higher contents in the WT than in the mutant. The contents of minor flavones significantly and linearly increased with increasing UV-B in the WT plants at 3 d after treatment, but not after 7 d, except for apigenin 7,4′-di-O-glucuronide (Supplementary Fig. S7). Luteolin 7,3′-di-O-glucuronide showed the clearest increases with time in both the WT and mutant plants (Supplementary Table S6).

Responses to low and high UV in the laboratory

We exposed plants to a longer, 5-week period of UV irradiation using two levels of UV irradiance, low and high, in combination with four different radiation regimes (Supplementary Table S2). WT plants showed higher contents of the major flavone apigenin 7-O-glucuronide in those treatments where UV-B was present, irrespective of the irradiance used (Fig. 3). Specifically, under low UV conditions, its content was two-fold higher for the PAB treatment (PAR+UV-A+UV-B) and three-fold higher for the PB treatment (PAR+UV-B) as compared with the P treatment (PAR alone). High UV conditions increased these differences, reaching contents of 40- and 25-fold higher for PAB and PB, respectively. The uvr8 mutant showed a smaller increase than the WT, and only under high UV irradiances, including PA. The remaining flavones, as well as the bulk level of SUVACs, generally showed similar changes to those found for apigenin 7-O-glucuronide in both genotypes. Changes in accumulation of the phenolic compounds were well matched with those found in CHS1 expression: under high UV treatments, transcript levels were higher in both genotypes, and the WT showed significantly higher values than the mutant.

Fig. 3.

Effects of different UV conditions on the phenolic composition and expression of CHALCONE SYNTHASE 1 (CHS1) in the Marchantia polymorpha wild type (WT) and a UVR8-knockout mutant (uvr8). Plants were grown for 5 weeks under either high or low UV-B in combination with four different radiation regimes: P, PAR alone; PA, PAR+UV-A; PB, PAR+UV-B; PAB, PAR+UV-A and UV-B (Supplementary Table S2). The bulk levels of soluble UV-absorbing compounds (SUVACs) were measured by spectrophotometry as the area under the absorbance curve in the interval 280–400 nm (mg^–1 DM), CHS1 relative expression was determined by qPCR and is relative to the values in treatment P for each genotype and UV condition (values set as 1), and contents (ng mg⁻¹ DM) of the six flavones were determined using UPLC-MS. All data are means (±SE), n=3. Different letters indicate significant differences among different treatments in WT+low-UV (lowercase), uvr8+low-UV (uppercase), WT+high-UV (lowercase italics), and uvr8+high-UV (uppercase italics), as determined using one-way ANOVA followed by Tukey’s test (P<0.05). Significant differences between genotypes within each treatment were determined using Student’s t-test: *P<0.05, **P<0.01, ***P<0.001.

Regarding photosynthetic performance, F_v/F_m remained fairly constant in WT plants (Fig. 4). In contrast, it decreased in the uvr8 mutant under several different conditions, under both low UV (in PAB at pre-dawn) and high UV (in PB and PAB at both pre-dawn and noon).

Fig. 4.

Effects of different UV conditions on photosynthetic performance in the Marchantia polymorpha wild type (WT) and a UVR8-knockout mutant (uvr8). Plants were grown for 5 weeks under either high or low UV-B in combination with four different radiation regimes: P, PAR alone; PA, PAR+UV-A; PB, PAR+UV-B; PAB, PAR+UV-A and UV-B (Supplementary Table S2). The maximum quantum yield of PSII (F_v/F_m) was determined (A) pre-dawn and (B) at noon. All data are means (±SE), n=3. Different letters indicate significant differences among different treatments in WT+low-UV (lowercase), uvr8+low-UV (uppercase), WT+high-UV (lowercase italics), and uvr8+high-UV (uppercase italics), as determined using one-way ANOVA followed by Tukey’s test (P<0.05). Significant differences between genotypes within each treatment were determined using Student’s t-test: **P<0.01.

Acclimation to UV radiation in the laboratory

To examine potential effects of acclimation to the low and high UV-B conditions, we next exposed the 5-week-old plants referred to above to a further 3 weeks of high UV (Figs 5, 6). Generally, at the end of this second period, both the WT and the uvr8 plants previously acclimated to low UV continued to have lower flavone contents than plants acclimated to high UV. In the WT exposed to UV-B, the apigenin 7-O-glucuronide content was 10-fold lower in low-UV- than high-UV-acclimated plants, and a similar trend was found for the other flavones. In the mutant, the changes in flavones were similar to those in the WT, but they were mostly independent of the presence of UV in the treatments. In both genotypes, most of the minor flavones increased in content during the 3-week period in plants that were acclimated to high-UV (Supplementary Table S7): this was observed in most treatments, including P. On the other hand, only small changes were observed for the major flavone apigenin 7-O-glucuronide. Consistent with our other experiments, the bulk level of SUVACs showed similar results to those found for the individual flavones, although changes were slighter.

Fig. 5.

Effects of acclimation to low and high UV conditions on photosynthetic performance and the phenolic composition in the Marchantia polymorpha wild type (WT). Plants were first grown for 5 weeks under either high or low UV-B in combination with four different radiation regimes: P, PAR alone; PA, PAR+UV-A; PB, PAR+UV-B; PAB, PAR+UV-A and UV-B. All plants were then grown for a further 3 weeks under high UV-B (Supplementary Table S2). The maximum quantum yield of PSII (F_v/F_m) was determined at pre-dawn. The bulk levels of soluble UV-absorbing compounds (SUVACs) were measured by spectrophotometry as the area under the absorbance curve in the interval 280–400 nm (mg^–1 DM), and contents (ng mg⁻¹ DM) of the six flavones were determined using UPLC-MS. All data are means (±SE), n=3. Different letters indicate significant differences among different treatments for plants previously acclimated to low UV (normal font) and to high UV (italics) as determined using one-way ANOVA followed by Tukey’s test (P<0.05). Significant differences between low- and high-UV-acclimated plants within treatments were determined using Student’s t-test: *P<0.05, **P<0.01, ***P<0.001.

Fig. 6.

Effects of acclimation to low and high UV conditions on photosynthetic performance and the phenolic composition in the Marchantia polymorpha UVR8-knockout mutant (uvr8). Plants were first grown for 5 weeks under either high or low UV-B in combination with four different radiation regimes: P, PAR alone; PA, PAR+UV-A; PB, PAR+UV-B; PAB, PAR+UV-A and UV-B. All plants were then grown for a further 3 weeks under high UV-B (Supplementary Table S2). The maximum quantum yield of PSII (F_v/F_m) was determined at pre-dawn. The bulk levels of soluble UV-absorbing compounds (SUVACs) were measured by spectrophotometry as the area under the absorbance curve in the interval 280–400 nm (mg^–1 DM), and contents (ng mg⁻¹ DM) of the six flavones were determined using UPLC-MS. All data are means (±SE), n=3. Different letters indicate significant differences among different treatments for plants previously acclimated to low UV (normal font) and to high UV (italics) as determined using one-way ANOVA followed by Tukey’s test (P<0.05). Significant differences between low- and high-UV-acclimated plants within treatments were determined using Student’s t-test: *P<0.05, **P<0.01, ***P<0.001.

Regarding F_v/F_m, both the WT and the uvr8 mutant acclimated well to low-UV, showing similarly high values in all the treatments independently of the presence of UV at the end of the experiment (Figs 5, 6). In contrast, F_v/F_m decreased in both WT and mutant plants acclimated to high-UV when UV-B was present (PB and PAB treatments), and showed physiological damage that was probably caused by sustained exposure to high UV-B. This decrease, which was stronger in the mutant, occurred despite the fact that flavone contents were higher in plants acclimated to high- than to low-UV. In addition, in the treatments with UV-B radiation (PB and PAB), F_v/F_m was significantly higher in plants acclimated to low- than to high-UV.

Responses to UV treatments in the field

Under field conditions, season, genotype, and treatment significantly influenced most of the measured variables, and several interactions between the three main factors were found (Supplementary Table S8). Contents of apigenin 7-O-glucuronide were higher in the WT than in the uvr8 mutant in all seasons and treatments, and the bulk level of SUVACs showed similar results in almost all the seasons and treatments (Fig. 7). Neither apigenin 7-O-glucuronide nor SUVACs showed clear patterns of variation among the different treatments in any genotype or season, although the highest values were always found in treatments with UV-B. The highest and lowest contents of apigenin 7-O-glucuronide were found in winter and summer, respectively. Changes in other flavones were similar to those found for apigenin 7-O-glucuronide (Supplementary Figs S8–S10). WT plants showed higher contents than the mutant in most treatments, with the differences being more evident in winter than in spring and summer. Regarding the influence of the UV treatments on WT plants, no clear pattern of variation was observed. Nevertheless, in winter, monoglucuronides and diglucuronides seemed to show different changes among the treatments, with monoglucuronides showing lower values in the Pab treatment whilst diglucuronides did not show any difference between treatments. There was generally no apparent influence of the UV treatments on the uvr8 plants.

Fig. 7.

Effect of season and different UV conditions on the phenolic composition in the Marchantia polymorpha wild type (WT) and a UVR8-knockout mutant (uvr8) grown in the field. Across the three seasons, the following UV treatments were applied: PAB, similar to natural sunlight with PAR, UV-A, and UV-B; Pab, reduced UV-A and UV-B; PAb, reduced UV-B; PaB, reduced UV-A; PAB↑, enhanced UV-B (see Methods; Supplementary Table S3). The bulk levels of soluble UV-absorbing compounds (SUVACs) were measured by spectrophotometry as the area under the absorbance curve in the interval 280–400 nm (mg^–1 DM), and the contents of apigenin 7-O-glucuronide (ng mg⁻¹ DM) were determined using UPLC-MS. Data are means (±SE), n=3. Different letters indicate significant differences among treatments for the WT (lowercase) and the mutant (uppercase) as determined using one-way ANOVA followed by Tukey’s test (P<0.05). Significant differences between the genotypes within each treatment were determined using Student’s t-test: **P<0.01, ***P<0.001.

The ratios of luteolins/apigenins and mono-/diglucuronides of WT plants showed differences between seasons (Fig. 8). In winter, the luteolins/apigenins ratio was much lower than in summer, whilst intermediate values were found in spring. This pattern was found in all the UV treatments. The mono-/diglucuronides ratio showed the opposite seasonal pattern, with the highest values in winter and the lowest in summer.

Fig. 8.

Effects of season and different UV conditions on the balance of flavones in the Marchantia polymorpha wild type grown in the field. The ratios of luteolins/apigenins and mono-/diglucuronides in winter, spring, and summer were determined in plants subjected to the following UV treatments: PAB, similar to natural sunlight with PAR, UV-A, and UV-B; Pab, reduced UV-A and UV-B; PAb, reduced UV-B; PaB, reduced UV-A; PAB↑, enhanced UV-B (see Methods; Supplementary Table S3). Data are means (±SE), n=3. Different letters indicate significant differences among seasons for each treatment as determined using one-way ANOVA followed by Tukey’s test (P<0.05).

Regarding variables related to photosynthesis, F_v/F_m was highest in spring and lowest in winter, but it did not show any difference between treatments in any season (Supplementary Fig. S11; Supplementary Table S8). Differences between the two genotypes were small, and only occurred consistently in spring at pre-dawn, with the WT showing higher values than the uvr8 mutant in all treatments. A_N showed more subtle differences between treatments, seasons, and genotypes than F_v/F_m, although there was a certain relationship between the two photosynthetic variables. Differences between treatments were found in spring (with the highest values in PAB, the lowest in PAB↑, and intermediate values in the remaining treatments) and in summer (with higher values in those treatments without UV-B), in both cases only in WT plants. A_N was higher in spring than summer and winter. Differences between the genotypes were rare and, when apparent, showed higher values in the WT. In terms of photosynthetic pigments, Chl a+b, the Chl a/b ratio, and the xanthophyll index did not show clear patterns of variation with season, genotype, or treatment (Supplementary Fig. S12; Supplementary Table S8). Nevertheless, Chl a+b was higher in spring than in summer and winter, and the xanthophyll index was higher in summer than in winter and spring. Gene expression showed a great variability, and none of the genes studied showed any consistent difference between treatments in any genotype (Supplementary Figs S13, S14).

To summarize the effects of the different UV treatments and seasons, the results were ordinated using PCA on the basis of phenolic composition and photosynthetic performance (F_v/F_m) (Fig. 9). The accumulated variance in the three first axes was 91.1% (41.7, 29.2, and 20.2% for axes I, II, and III, respectively). In the plot generated using axes I and II, the contents of all flavones together with SUVACs and F_v/F_m were significant loading factors for the positive part of axis I, while the luteolins/apigenins ratio was significant for its negative part. The two genotypes were clearly separated along this axis, with all uvr8 data-points placed in its negative part and almost all WT points in the positive part. In addition, the WT points showed a bigger separation between treatments than uvr8, especially in summer and, to a lesser extent, in winter. For axis II, samples were separated by the differences in the type of compounds accumulated. Luteolins and the luteolins/apigenins ratio were loading factors for the positive part of the axis, while apigenin 7-O-glucuronide, the mono-/diglucuronides ratio, and F_v/F_m were significant for its negative part. For this axis, samples were organized by season. Summer and winter samples were placed towards the positive and negative parts of this axis, respectively, while spring samples were in between. Again, summer WT points were those with the biggest separation between treatments.

Fig. 9.

Principal component analysis of the results obtained from the Marchantia polymorpha wild type (WT) and a UVR8-knockout mutant (uvr8) grown in the field. Across winter, spring, and summer, the following UV treatments were applied: PAB, similar to natural sunlight with PAR, UV-A and UV-B; Pab, reduced UV-A and UV-B; PAb, reduced UV-B; PaB, reduced UV-A; PAB↑, enhanced UV-B (see Methods; Supplementary Table S3). The vertical axis explains 41.7% of the variance and the horizontal one explains 29.2%. Significant loading factors for the positive and negative parts of each axis are shown: *P<0.05, **P<0.01, ***P<0.001. SUVAC, the bulk level of UV-absorbing compounds from the soluble fraction; Lut., luteolin; Apig., apigenin.

Discussion

UVR8-dependent and -independent responses under laboratory conditions: genes and metabolites

The experiments performed under laboratory conditions using UV-B exposures from 4 h to 5–8 weeks showed similar results. In WT plants, expression increased for most of the genes studied (CHS1, FNSI, HY5, SIG5, ELIP2, and less clearly CRY-DASH; Figs 1, 3). In addition, accumulation of flavones was activated, especially for the major component apigenin 7-O-glucuronide but also for other apigenin and luteolin derivatives, as well as the bulk level of SUVACs (Figs 2A, 3, 5). These responses generally depended on the doses/irradiances used (i.e. responses were mainly quantitative), particularly in the case of apigenin 7-O-glucuronide (Fig. 2C; Supplementary Fig. S7). However, in the uvr8 mutant, the increase in gene expression was much weaker, and only occurred under the highest doses used (Figs 1, 3), suggesting a UVR8-independent response, probably as a result of stress. Moreover, where flavone accumulation was observed, it was modest and only found under high UV-B (Figs 2A, 3, 6), and no correlation was found between metabolite contents and the UV-B doses used (Fig. 2C).

Overall, the gene and metabolite responses in M. polymorpha were UVR8-dependent and were modulated by the UV-B level applied. This represented the most common system to manage plant UV-B responses, whilst the responses shown by the uvr8 mutant were UVR8-independent and only operated at high UV-B levels. This duality, and its dependence on the UV-B level, has repeatedly been found in Arabidopsis (Brown and Jenkins, 2008; González Besteiro et al., 2011; Kusano et al., 2011; Morales et al., 2013; Jenkins, 2017; O’Hara et al., 2019), and also in M. polymorpha under controlled conditions (Clayton et al., 2018; Kondou et al., 2019). Thus, the presence of UVR8-dependent and UVR8-independent systems has been evolutionarily conserved from bryophytes to angiosperms. The UVR8-independent system could include non-specific stress-related responses mediated by HY5 and MYB14, activated by the oxidative stress and DNA damage caused by high UV-B (Kondou et al., 2019).

The UV-B-stimulated expression of the genes studied (Figs 1, 3; Supplementary Fig. S4) was mostly expected because they are involved in photoprotection. In particular, the levels of the flavonoid-related transcripts of CHS1 and FNSI in WT plants fitted well to a saturation model. This indicated both a high sensitivity to low UV-B fluence rates (consistent with an early photoprotective transcriptional response) and the existence of a maximum inducible transcriptional capacity under the conditions imposed. CHS and FNSI are clearly induced in many angiosperms (Brown et al., 2005; Kanazawa et al., 2012; Righini et al., 2019; Santin et al., 2020) and in the case of CHS also in bryophytes, including M. polymorpha (Clayton et al., 2018; Kondou et al., 2019), the model moss Physcomitrium patens (Wolf et al., 2010), and other mosses (Liu et al., 2021). However, FNSI has rarely been examined in UV-B-exposed bryophytes, although increased UV-B tolerance has been reported in transgenic Arabidopsis overexpressing FNSI from the moss Pohlia nutans from Antarctica (Wang et al., 2020). HY5 is a key effector of the UVR8 pathway (Brown et al., 2005; Favory et al., 2009), and its induction by UV-B has previously been found in bryophytes, both liverworts and mosses (Clayton et al., 2018; Kondou et al., 2019; Liu et al., 2021). The effect of UV-B on SIG5 has not previously been tested in bryophytes, but its induction in M. polymorpha supports results found in Arabidopsis (Davey et al., 2012). The UV-B induction of ELIP2 in M. polymorpha that we observed confirms previous results in the same species (Clayton et al., 2018), although less clear responses have been reported in the moss Syntrichia canivervis (Liu et al., 2020). CRY-DASH was less clearly stimulated by UV-B, perhaps because it does not contribute significantly to protection from elevated UV-B (Rredhi et al., 2021). Nevertheless, transcript levels were generally higher in the WT than in the uvr8 mutant. This would support the notion that CRY-DASH expression is mediated by UVR8 in M. polymorpha, as occurs in Arabidopsis (Brown et al., 2005), although its induction ceased at high UV-B.

The increase in flavones in response to UV-B (Fig. 2A, Supplementary Fig. S5) was expected because the accumulation of phenolic compounds is probably the best characterized plant response to increased UV-B, both in tracheophytes and bryophytes (Searles et al., 2001; Newsham and Robinson, 2009; Pescheck and Bilger, 2020; Martínez-Abaigar and Núñez-Olivera, 2022). These compounds play a dual role as UV-screens and antioxidants (Agati et al., 2025). In M. polymorpha, flavone accumulation was previously found in plants exposed to UV-B under controlled conditions (Clayton et al., 2018; Soriano et al., 2019). In particular, apigenin 7-O-glucuronide seems to be a good UV-B biomarker (Figs 2, 3), and given the wide geographical distribution of M. polymorpha (Wu et al., 2025), it could be used in the biomonitoring of the UV-B changes associated with stratospheric ozone depletion and climate change (Martínez-Abaigar and Núñez-Olivera, 2022; Barnes et al., 2023). Flavone accumulation was also consistent with the increase in the expression of genes related to the synthesis of phenolic compounds, such as CHS1 and FNSI (Fig. 2B). All these responses were much clearer in WT plants, and were UVR8-dependent, as Clayton et al. (2018) previously found in M. polymorpha under controlled conditions. Given that increased gene expression will normally precede accumulation of metabolites, an appropriate sampling collection program needs to be used to detect both responses. Even with such a program in place, the response of specific compounds could be overlooked due to different response times, which might vary from hours (this study) to days (Clayton et al., 2018), probably depending on the experimental conditions applied and the compound considered; to the biochemical transformation of some compounds into others; and to the thallus age, because young tissues are more UV-responsive than adult tissues (Soriano et al., 2019). Overall, we have demonstrated that a short UV-B exposure is enough to activate the UVR8-mediated induction of several flavones in M. polymorpha in a fluence-dependent manner.

The bulk level of SUVACs showed similar although smaller responses to UV-B than individual flavones (Figs 3, 5, 6), probably because it encompasses a high diversity of individual compounds that might respond to UV-B in different ways. Nevertheless, an increase in SUVACs has also previously been found in M. polymorpha exposed to UV-B (Kondou et al., 2019), and many studies examine SUVACs rather than individual compounds in the context of UV research due to its easy measurement and responsiveness. On the basis of our results, we can recommend the continued use of SUVACs to assess the effects of UV-B on plants.

UVR8-mediated responses protect photosynthetic performance in the laboratory

Both UV-B and UV-A wavelengths can damage PSII components, reducing photochemical efficiency and causing a decrease in F_v/F_m (Verdaguer et al., 2017; Pescheck and Bilger, 2020). In WT plants, the induction of genes and flavones contributed to maintaining high F_v/F_m levels (Figs 4, 5), indicating reduced photoinhibition and conserved photosynthetic performance and physiological vitality (Maxwell and Johnson, 2000). F_v/F_m consistently decreased only when plants were exposed to sustained high UV-B, and this decrease was stronger in the uvr8 mutant than in the WT (Figs 4–6). Thus, UVR8-mediated processes were essential to confer photoprotection to the photosynthetic apparatus, allowing the acclimation of M. polymorpha to most of the UV-B conditions applied in our study. This is consistent with previous findings in Arabidopsis (Davey et al., 2012). In bryophytes, F_v/F_m can increase, decrease, or remain unchanged in response to UV radiation, depending on the experimental conditions used (Martínez-Abaigar and Núñez-Olivera, 2022). Specifically, in M. polymorpha, F_v/F_m has never been shown to decrease in response to different controlled UV treatments (Soriano et al., 2019). Taken together with our results, this suggests that M. polymorpha is a UV-B tolerant species within bryophytes, probably due to the induction of protective genes and metabolites. In particular, flavones have been highlighted as efficient UV-B screens and scavengers of reactive oxygen species (Agati et al., 2025).

UVR8 functionality under field conditions

Different field UV treatments were applied to both WT and uvr8 plants along a seasonal cycle (Figs 7–9; Supplementary Figs S8–S14). To our knowledge, this is the first seasonal study performed on UVR8 in a bryophyte and only the second in plants, after that conducted on Arabidopsis by Coffey et al. (2017). The variability of the responses was higher than under controlled conditions, as has previously been observed in the few (non-seasonal) field studies undertaken on UVR8 functionality in plants, all of them performed on Arabidopsis (Morales et al., 2013; Findlay and Jenkins, 2016; Pescheck and Bilger, 2019; Rai et al., 2020; Neugart et al., 2024; Stockenhuber et al., 2024). The increased variability was due to the complex interactions between genotype, UV treatments, and the changing environmental factors across the seasons, which resulted in attenuated differences between genotypes and treatments.

The two genotypes were clearly separated in the PCA, mainly due to their different phenolic compositions (Fig. 9). The contents of apigenin 7-O-glucuronide (the most UV-B-responsive compound across all the experiments) were consistently higher in the WT than in the uvr8 mutant, regardless of the season or the UV treatment applied (Fig. 7). Other flavones and SUVACs showed similar results, although less consistent because of a higher variability (Supplementary Figs S8–S10). Thus, UVR8 was essential for flavone induction in M. polymorpha under field conditions, as previously observed in Arabidopsis for other phenolic compounds and the bulk level of SUVACs (Morales et al., 2013; Coffey et al., 2017; Neugart et al., 2024). These effects are qualitative rather than quantitative, because they are more dependent on the presence of UVR8 than on the doses of UV-B received, in both M. polymorpha and Arabidopsis (Neugart et al., 2024). F_v/F_m (and to a lesser extent A_N) was significantly higher in the WT than the uvr8 mutant (Supplementary Fig. S11), probably due to the higher photoprotection provided by UVR8-mediated responses. Thus, although both F_v/F_m and A_N can negatively be affected by UV radiation through damage to PSII and Rubisco together with alterations in stomatal conductance (Verdaguer et al., 2017; Pescheck and Bilger, 2020), the accumulation of flavones efficiently protected the photosynthetic apparatus due to their double role as UV screens and antioxidants. No genotype influence was found on chlorophylls or xanthophylls (Supplementary Fig. S12). There are not many comparable data in the literature regarding photosynthetic variables, but Coffey et al. (2017) reported similar results for F_v/F_m in Arabidopsis. Gene expression did not show genotype differences (Supplementary Figs S13, S14). Opposite results have been reported by Morales et al. (2013) and Rai et al. (2020) in field-grown Arabidopsis, where shorter UV exposures (6–36 h) could have facilitated the detection of gene induction.

The different treatments were not clearly separated in the PCA (Fig. 9), due to the relative similarity between their phenolic compositions, which in turn was probably due to the presence of some natural and/or artificial UV-B in all treatments. Indeed, our laboratory experiments demonstrated that a low UV-B dose was enough to induce flavone accumulation, and even a high UV-A dose can lead to the same result in Arabidopsis (Rai et al., 2020). Nevertheless, the highest values of apigenin 7-O-glucuronide and SUVACs were generally found in treatments with UV-B, which highlights its role in the induction of phenolic compounds in the field. Small differences between treatments were also found in other variables, such as gene expression, F_v/F_m, A_N, and photosynthetic pigments. Comparable data for these variables are scarce in the literature, but similar findings have been reported for F_v/F_m in Arabidopsis (Coffey et al., 2017). Small differences between treatments could also be due to the interaction of other environmental factors that might mask UV effects, such as temperature (Coffey et al., 2017; Coffey and Jansen, 2019; Pescheck and Bilger, 2019). In addition, photosynthetic pigments might be more responsive to PAR than UV in M. polymorpha (Soriano et al., 2019).

Seasons were clearly separated in the PCA (Fig. 9). In particular, flavone contents and SUVACs were higher in winter than in summer. This is consistent with previous results obtained in Arabidopsis for SUVACs (Coffey et al., 2017; Coffey and Jansen, 2019) and quercetin (Pescheck and Bilger, 2019), and ties in with the well-known activation of flavonoid biosynthesis by cold (Agati et al., 2025). In M. polymorpha, this cold-mediated accumulation was UVR8-independent because we found a similar seasonal accumulation in both WT and uvr8 plants. However, a latitudinal transect performed with Arabidopsis showed that the impact of UVR8 on flavonoid accumulation increased at higher temperatures (Neugart et al., 2024). Overall, temperature strongly modulates UV effects under field conditions, including the dynamics of phenolic compounds and also UVR8 function (Findlay and Jenkins, 2016; Neugart et al., 2024), but the specific regulatory role of UVR8 under different temperatures is not yet completely understood.

Seasonal changes occurred not only in flavone contents but also in the proportions of the different flavones, as occurs in Arabidopsis for different flavonols and their patterns of glycosylation (Coffey and Jansen, 2019). In our study, the luteolins/apigenins ratio was higher in summer than in winter, whereas the opposite pattern was found for the mono-/diglucuronides ratio (Fig. 8). High PAR and UV-B can trigger changes in these ratios, resulting in the synthesis of compounds with higher antioxidant capacities: luteolins over apigenins, and mono- over diglucuronides (Heim et al., 2002; Agati et al., 2025). Thus, M. polymorpha might use different strategies to increase antioxidant capacity in winter (increasing mono- over diglucuronides) than in summer (increasing luteolins over apigenins). The specific factors potentially responsible for these strategies are unknown, but would mainly be PAR, UV, and temperature, because water deficit was avoided by irrigation. In addition, each compound might specifically respond to the different combinations of environmental factors. Interestingly, the increase in the luteolins/apigenins ratio in response to UV-B under field conditions corresponds well with the increase in the quercetin/kaempferol ratio reported in Petunia cultivated outdoors (Ryan et al., 2002). In both cases, the ratio of di- to mono-hydroxylated compounds increased, either in flavones or flavonols, which are probably the two most UV-B-reactive groups of phenolic compounds (Agati et al., 2025), thus increasing photoprotection. A similar increase in the quercetin/kaempferol ratio in Arabidopsis has been attributed to differential UVR8 regulation of their synthesis (Neugart et al., 2024). It could be speculated that the same occurs for the synthesis of luteolins and apigenins in Marchantia.

The highest values of F_v/F_m and A_N were found in spring, the intermediate season between the two extremes of winter and summer (Supplementary Fig. S11); however, photosynthetic pigments and gene expression were less sensitive and did not show seasonal changes (Supplementary Figs S12–S14). Arabidopsis shows less variable seasonal values of F_v/F_m than M. polymorpha, with lower values only in winter (Coffey et al., 2017), but this can be explained by the differences in species and lineage, and in the specific climatic conditions (oceanic versus mediterranean) between the studies. Overall, our field results suggest an integrated network of multi-stress pathways that conferred plants with a certain plasticity to cope with natural conditions but did not mask the UVR8 effects. Thus, UVR8 is not indispensable for M. polymorpha survival in nature, but contributes to a higher photosynthetic performance in WT than in mutant plants.

UV-B or UV-A?

In the field, the influence of UV-A was only evident when combined with UV-B (Figs 7–9; Supplementary Figs S8–S14). In the laboratory, most of the observed changes were due to UV-B radiation while UV-A responses were more modest, being mainly conditioned by the irradiance and by the presence of UV-B, as occurred in the field (Figs 3, 5, 6). The influence of UV-A irradiance was clear in that some flavones increased under high- but not low-UV conditions in both WT and uvr8 plants (Fig. 3). This suggests that M. polymorpha is reactive to UV-A through UVR8-dependent and -independent systems. In the first case, short UV-A wavelengths could induce UVR8-mediated responses, as occurs in Arabidopsis (Rai et al., 2020). In the second case, the action of cryptochromes could support our results, because they are involved in the UV-A induction of CHS and phenolic compounds in Arabidopsis (Verdaguer et al., 2017; Robson et al., 2019; Rai et al., 2020).

We found that UV-B radiation induced much stronger (and sometimes damaging) responses in genes, flavones, and F_v/F_m, especially in combination with UV-A at high irradiances. This suggests a synergistic effect, although both specific and synergistic effects of UV-B and UV-A have previously been reported in M. polymorpha (Soriano et al., 2019). More research is needed to differentiate the effects of UV-A and UV-B in bryophytes, as well as the photoreceptors involved. In addition, it should be emphasized that our results and conclusions are directly related to the experimental conditions applied, and that using different levels and proportions of UV-A and UV-B bands would most likely have produced different results.

UVR8 evolution

Overall, M. polymorpha has developed a coordinated system of genetic, metabolic, and photosynthetic responses to UV-B radiation, which are mainly regulated by UVR8 and ultimately protect its photosynthetic machinery. The UVR8-mediated responses that we found in M. polymorpha are similar to those reported in angiosperms and, thus, our study confirms that UVR8 functionality has been evolutionarily conserved from bryophytes to angiosperms, as evidenced not only under controlled conditions (Clayton et al., 2018) but also in nature. This is not surprising because both the model moss Physcomitrium patens and the model liverwort M. polymorpha itself express functional UVR8 photoreceptor proteins similar to Arabidopsis (Soriano et al., 2018). In addition, other plant photoreceptors have also been conserved during plant evolution, although maintaining particular properties in specific lineages (Lariguet and Dunand, 2005). UVR8 is an ancient photoreceptor that is present in different lineages of green algae, such as chlorophytes, zygnematophytes, and charophytes (Tilbrook et al., 2016; Jenkins, 2017; Han et al., 2019; Mishra et al., 2022; Busch et al., 2024). It can be inferred that UVR8 played a key role in the water-to-land transition of photosynthetic organisms because it facilitated photoprotection under ambient levels of UV-B.

Supplementary data

The following supplementary data are available at JXB online.

Fig. S1. Experimental design of the study.

Fig. S2. Images illustrating field and laboratory experiments.

Fig. S3. Spectral characteristics of the treatments in the experiments.

Fig. S4. Relative transcript levels of CHS1 and FNSI versus UV-B fluence rates in WT plants.

Fig. S5. Flavone contents versus UV-B fluence rates after 3 d of culture.

Fig. S6. Flavone contents versus UV-B fluence rates after 7 d of culture.

Fig. S7. Regressions between flavone contents and the UV-B fluence rates in WT plants.

Fig. S8. Flavone contents in the winter field experiment under the five UV treatments applied.

Fig. S9. Flavone contents in the spring field experiment under the five UV treatments applied.

Fig. S10. Flavone contents in the summer field experiment under the five UV treatments applied.

Fig. S11. F_v/F_m and net assimilation under the five UV treatments applied in the seasonal field study.

Fig. S12. Photosynthetic pigment variables under the five UV treatments applied in the seasonal field study.

Fig. S13. Gene expression in WT plants under the five UV treatments applied in the seasonal field study.

Fig. S14. Gene expression in MpUVR8-KO plants under the five UV treatments applied in the seasonal field study.

Table S1. Details of lamps and filters used in the different experiments.

Table S2. Experimental design, sampling, and radiation conditions used in the laboratory experiments.

Table S3. Total doses of PAR, UV-A, and UV-B received by plants under the five UV treatments applied in the seasonal field study.

Table S4. Primers used for qPCR.

Table S5. ANOVA of C_T values of the reference gene ACTIN7 across the seasonal field experiments.

Table S6. Differences in flavone contents between the two different harvest times in the fluences experiment.

Table S7. Differences in flavone contents between the two different harvest times in the acclimation experiment.

Table S8. Results of three-way ANOVA for global effects of season, genotype, and treatment in the field experiment.

Author contributions

GS, LM, M-ÁD-C-A, and RT-L-H: investigation, data analysis, visualization; GS: writing—original draft; GIJ and YK: writing—review & editing; JM-A and EN-O: conceptualization, supervision, data analysis, funding acquisition, writing—review & editing. All the authors revised and approved the manuscript.

Funding

This work was supported by the Ministerio de Ciencia, Innovación y Universidades of Spain – Agencia Estatal de Investigación – Fondo Europeo de Desarrollo Regional (MCIU/AEI/10.13039/501100011033/FEDER, EU) project no. PID2023-150695NB-I00 to JM-A and EN-O.

Data availability

The data that support the findings of this study are available at Zenodo https://doi.org/10.5281/zenodo.19825875 (Soriano et al., 2026).