Fv/Fm recovery
Desiccated, field-collected S. caninervis plants from all treatments had very low maximum potential PSII quantum efficiency, Fv/Fm, when initially rehydrated but recovered over eight days in simulated winter conditions in which Fv/Fmincreased from less than 0.1 to 0.81. In unstressed land plants Fv/Fm is nearly constant around 0.83 (Björkman and Demmig, 1987; Proctor, 2001). Often a low Fv/Fm is assumed to indicate stress related to PSII damage primarily attributed to inactivation of the core reaction center D1-protein (Demmig and Björkman, 1987; Csintalan et al., 1999), and thus increasing Fv/Fm is interpreted as repair of PSII as part of the D1 cycle (Melis, 1999). However, because Fv/Fmis a normalized ratio it is important to determine which component is driving Fv/Fm depression or recovery. As Fv = Fm – Fo, Fv/Fm is equivalent to (Fm – Fo)/Fm, the ratio can increase over time (i.e., during recovery) due to increasing Fm or decreasing Fo, or both. Understanding change in these variables over time can provide insight on the underlying biological processes contributing to observed change in Fv/Fm. For example, Fo is high when PSII is damaged (Rintamäkiet al., 1994; Ritchie, 2006; Murchie and Lawson, 2013). An increase in Fv/Fm due to a decrease in Fo with a relatively constant Fm would be strongly indicative of PSII damage and subsequent repair. On the other hand, an increase in Fv/Fmdriven by rising Fm is consistent with a relaxation of NPQ (Müller et al., 2001). This latter scenario is what we observed in S. caninervis recovering from Fv/Fm depression: an increase in Fv/Fm over the recovery period driven by Fm, which suggests relaxation of sustained NPQ rather than repair of damaged or inactivated PSII (Fig. S3).
Different pigment and antioxidant profiles in field-collected and lab-cultured plants
Comparison of the photosynthetic pigment profiles in field-collected and lab-cultured S. caninervis supports this hypothesis of relaxation of sustained NPQ. Zeaxanthin, which is associated with both rapidly reversible (qE) and sustained NPQ mechanisms such as photoinhibitory quenching (qI; Demmig-Adams, 1990; Verhoeven et al., 1996), was more than five times higher in field-collected plants than in lab-cultured plants (Table 4). In fact, the relative VAZ pool was larger in field-collected plants, which is unsurprising as these pigments increase in abundance in high-light environments (Siefermann-Harms, 1985; Demmig-Adams, 1990; Jahns et al., 2009). Zeaxanthin accumulation is associated with sustained NPQ in desiccation-tolerant mosses, specifically accumulating when desiccation occurs in natural light conditions (Verhoeven et al., 2020). The higher levels of zeaxanthin in field-collected plants suggests accumulation due to the desiccation in the natural habitat but not in the laboratory cultures. Similarly, in field-collected plants the total chlorophyll pool was reduced and the ratio of chlorophyll a:b was increased, also consistent with acclimation to high light intensity (Björkman, 1981; Leong and Anderson, 1984; Lindahl et al., 1995). Tocopherol abundance was much higher in field-collected plants than those cultured in the lab, as well (Table 6). Tocopherols are membrane-bound phenolic antioxidants that may be increased due to the higher light intensity and UV exposure in the field site (Delong and Steffen, 1998; Yao et al., 2015) or due to other stresses such as desiccation and freezing that these plants frequently face in their natural habitat (Munné-Bosch, 2005).
Altered Fv/Fm recovery following UV-filtering
Surprisingly, Fv/Fm was not affected in desiccated S. caninervis when natural levels of UV were reduced, but the recovery of Fv/Fm was impaired during at least 192 hours in winter recovery conditions (Fig. 3). In contrast, many plants respond to supplemental UV radiation with reduced Fv/Fm (Bradshaw, 1965; Strid et al., 1990; He et al., 1993; Pukacki & Modrzyński, 1998; Ranjbarfordoei et al., 2011; but see Takács et al., 1999; Csintalan et al., 2001; Lau et al., 2006; Basahi et al., 2014). Furthermore, relative abundance of the xanthophyll zeaxanthin was also increased in UV-filtered plants (Fig. 4), a response also typically seen with UV supplementation (Agrawal et al., 2009). Why should removal of UV radiation, presumably a stressor, result in altered recovery of Fv/Fm and more antioxidant xanthophylls in S. caninervis? One possible explanation for the observed reduction in Fv/Fm recovery is that removal of UV somehow causes an impairment in relaxation of sustained NPQ. As with un-manipulated field-collected plants, the observed Fv/Fm increase over the recovery period for UV-filtered and UV-transmitted plants was driven by an increase in Fm and thus is consistent with relaxation of sustained NPQ. Indeed, the increased abundance of zeaxanthin with removal of UV is consistent with the hypothesis that UV filtering induces a sustained zeaxanthin-related NPQ (Verhoeven et al., 1996). Although zeaxanthin is not required for sustained quenching, its accumulation likely contributes to sustained NPQ when present (Verhoeven et al., 2020).
It is possible that UV radiation is a photomorphogenic (Gitz and Liu-Gitz, 2003) or regulatory signal rather than (or in addition to) being a stressor such that the absence of this signal indirectly affects Fv/Fm recovery. For example, UV may induce production of enzymatic antioxidants or phenolics (Cooper-Driver et al., 1998; Clarke and Robinson, 2008; Waterman et al., 2017) that may have roles beyond UV protection, such as in desiccation tolerance (Gitz and Liu-Gitz, 2003; Poulson et al., 2006; Robson et al., 2015). Without these UV-associated responses, desiccation in the field might cause more photo-oxidative stress. In addition to increased VAZ pool size, the relative abundance of tocopherols increased with removal of UV from S. caninervis in the field (Fig. 6 and Table 5), suggestive of increased ROS activity. Tocopherols quench singlet oxygen from the PSII reaction center (Trebst et al., 2002; Trebst, 2003; Krieger-Liszkay, 2005), and α-tocopherol has been shown to confer antioxidant protection to thylakoid membranes in UV-B-exposed spinach plants (Delong and Steffen, 1998). There are a number of stress protection mechanisms mediated by UVR8, the UV-B sensing protein receptor (Singh et al., 2014), many of which could result in slower Fv/Fm recovery and increased antioxidant abundance without UV-induced signaling. In fact, the UV-B response pathway and the photomorphogenesis pathway have substantial overlap (Stanley and Yuan, 2019).
Transcriptomic response to reduced UV
Transcriptomic profiling of the UV-filtered and UV-transmitted plants revealed an altered transcript abundance with genes involved in flavonoid biosynthesis and essential plant function. Commonly located in vacuoles or cell walls, flavonoids are phenolic secondary metabolic compounds important for tolerance to UV (e.g. via antioxidant function or UV absorption) and a variety of other stresses in plants (Cooper-Driver et al., 1998; Graham, 1998; Markham et al., 1998; Grace and Logan, 2000; Wolf et al., 2010). Two of the 19 differentially abundant transcripts, including the most differentially abundant transcript, were for the α-xylosidase 1-like genes, Sc_g01390 and Sc_g08662), which were nearly 1.5 and 0.76 log2-fold higher, respectively, with UV filtering. A glycoside hydrolase, α-xylosidase 1-like may be involved in the breakdown of flavonol glycosides, as glycosylation is necessary for stable flavonoid accumulation in other plants (Luo et al., 2007; Lee et al., 2017). In both rice and Arabidopsis thaliana, abundance of flavonol glycosides including kaempferol and quercetin glycosides increase with UV-B radiation (Graham, 1998; Markham et al., 1998; Veit and Pauli, 1999). Correspondingly, glycosyl hydrolase transcript abundance decreases with UV-B exposure in Artemisia annua, suggesting breakdown inhibition (Pan et al., 2014). In contrast, we found increased abundance of glycoside hydrolase α-xylosidase 1-like transcripts with UV filtering in S. caninervis, suggesting increased glycoside breakdown and reduced glycoside accumulation, which may negatively affect UV tolerance. Similarly, transcripts of the gene Sc_g05612: β-galactosidase 8-like isoform X1 increased with UV filtering and also codes for a glycoside hydrolase and may be involved in inhibition of flavonoid biosynthesis. Importantly in A. thaliana, β-galactosidase has been shown to increase in activity during drought and senescence-induced photoinhibition (Mohapatra et al., 2010; Pandey et al., 2017), a form of photosynthetic downregulation.
Two genes associated with oxidoreductase activity, Sc_g11402: probable polyamine oxidase 2 and Sc_g07907: acyl-lipid (9-3)-desaturase-like, significantly increased in abundance with UV filtering in S. caninervis. Polyamine oxidases are involved in ROS homeostasis in A. thaliana, and some are upregulated by drought stress in A. thaliana and the resurrection plant Craterostigma plantagineum (Alcázar et al., 2011; Andronis et al., 2014). Increased abundance of these transcripts with UV-filtering may suggest increased oxidative stress with UV removal.
In addition to being involved in oxidoreductase activity, the fatty acid desaturase Sc_g07907: acyl-lipid (9-3)-desaturase is involved in biosynthesis of fatty acids such as gamma linolenic acid (Sayanova et al., 1997). Linolenic acid is a critical player in maintenance of membrane integrity and functionality of membrane proteins, including photosynthetic machinery proteins (Upchurch, 2008). Along with the second-most differentially abundant transcript (Sc_g07909: omega-6 fatty acid desaturase, chloroplastic), the increased abundance of these two fatty-acid biosynthetic pathway genes suggest biosynthesis or repair of membranes, perhaps chloroplast membranes, with UV-filtering in S. caninervis. In fact, lipid hydroperoxidation of membranes is a major form of ROS damage (Foyer et al., 1994; Alscher et al., 1997; Shigeoka et al., 2002) Together, differential abundance of transcripts involved in oxidative stress and membrane biosynthesis supports our hypothesis that removal of natural levels of UV radiation can lead to oxidative stress in S. caninervis.
Laboratory UV treatment on field-treated samples
In our study, application of an additional UV treatment to field window samples had no significant effect on Fv/Fmof UV-filtered plants over the 192-hour simulated winter recovery period. This result suggests either that the one year of reduced UV in the field was not sufficient to remove previously acquired acclimation or that these plants may have a physiologically constitutive level of protection in this assay. However, the mechanism of protection in the UV-filtered and UV-transmitted plants might have been different, as there were differences in their pigment and antioxidant profiles. For example, zeaxanthin was higher in UV-filtered plants and zeaxanthin has been found to contribute to UV stress protection and UV damage prevention in tobacco plants (Götz et al., 2002). It is also possible that any PSII damage incurred by the UV treatment was repaired in the 30-minute dark acclimation period prior to the first fluorescence measurement. Curiously, UV-transmitted plants had significantly higher Fv/Fm at T24 after the laboratory UV treatment, and UV-filtered plants showed the same pattern, though it was not significant. This result lends further support to the hypothesis that UV exposure has beneficial effects on photosynthetic efficiency in S. caninervis following desiccation, as even a moderate dose of UV applied to these desiccated mosses improved Fv/Fm recovery.
Conclusions
In summary, we find evidence that Mojave Desert S. caninervis plants undergo a sustained form of NPQ that takes days to relax and for efficient photosynthesis to resume in simulated winter conditions. As these plants spend much of the summer season in a dry, quiescent state under extremes in PAR and UV exposure, the 8-day Fv/Fm recovery we observed suggests strong recovery potential which may be mediated by seasonal photoprotective thermal dissipation (Demmig-Adams et al., 2012). Furthermore, reduction of UV radiation from natural sunlight had unexpected and adverse effects on recovery of photosynthetic efficiency in S. caninervis following rehydration. This counterintuitive finding is consistent with photoinhibitory effects from heightened levels of singlet oxygen and other ROS and may be driven by exposure to high visible light in the absence of a UV regulatory signal that likely induces multiple protective responses. Evidence to support this hypothesis includes the three photoprotective response metrics we observed in our UV-filtered plants: significantly higher zeaxanthin and tocopherols —both potential antioxidants – and increased abundance of transcripts associated with oxidative stress. Yet, all field plants in this study had high levels of these antioxidants, which, along with the chlorophyll fluorescence results, suggests they undergo a strong and sustained form of NPQ, which in this system takes as long as eight days post-rehydration before highly efficient photosynthesis can resume. It is difficult to distinguish PSII damage due to ROS in the presence of sustained NPQ and it is possible that UV-reduced plants have higher NPQ. More research is needed to determine to what extent these two processes, ROS damage and sustained NPQ, are contributing to the observed altered recovery of Fv/Fm in UV-reduced plants, and how these factors interact with desiccation in natural populations.