UV-B-induced photomorphogenesis in Arabidopsis thaliana


  • Byung Chul Kim,

    1. Boyce Thompson Institute for Plant Research and Section of Genetics and Development, Tower Road, Ithaca, NY 14853, USA, and
    Search for more papers by this author
    • Present address: Hitachi Advanced Research Laboratory, Hatoyama, Saitama, 350–03, Japan.

  • Daniel J. Tennessen,

    1. Department of Floriculture and Ornamental Horticulture, Cornell University, Ithaca, NY 14853, USA
    Search for more papers by this author
    • Present address: Monsanto Company, Ag Sector, St. Louis, MO 63167, USA.

  • Robert L. Last

    1. Boyce Thompson Institute for Plant Research and Section of Genetics and Development, Tower Road, Ithaca, NY 14853, USA, and
    Search for more papers by this author

*For correspondence: Cereon Genomics LLP, One Kendall Square, Building 300, Cambridge, MA 02139, USA (e-mail rob.last@cereon.com).


Relatively little is known about the types of photomorphogenic responses and signal transduction pathways that plants employ in response to ultraviolet-B (UV-B, 290–320 nm) radiation. In wild-type Arabidopsis seedlings, hypocotyl growth inhibition and cotyledon expansion were both reproducibly promoted by continuous UV-B. The fluence rate response of hypocotyl elongation was examined and showed a biphasic response. Whereas photomorphogenic responses were observed at low doses, higher fluences resulted in damage symptoms. In support of our theory that photomorphogenesis, but not damage, occurs at low doses of UV-B, photomorphogenic responses of UV-B sensitive mutants were indistinguishable from wild-type plants at the low dose. This allowed us to examine UV-B-induced photomorphogenesis in photoreceptor deficient plants and constitutive photomorphogenic mutants. The cry1 cryptochrome structural gene mutant, and phytochrome deficient hy1, phyA and phyB mutant seedlings resembled wild-type seedlings, while phyA/phyB double mutants were less sensitive to the photomorphogenic effects of UV-B. These results suggest that either phyA or phyB is required for UV-B-induced photomorphogenesis. The constitutive photomorphogenic mutants cop1 and det1 did not show significant inhibition of hypocotyl growth in response to UV-B, while det2 was strongly affected by UV-B irradiation. This suggests that COP1 and DET1 work downstream of the UV-B signaling pathway.


Because of their growth habit and the requirement of sunlight for photosynthesis, plants are constantly exposed to the damaging effects of ultraviolet-B radiation (UV-B, 290–320 nm). Moreover, stratospheric ozone depletion resulting from air pollution is increasing tropospheric UV-B irradiation (Gleason et al. 1993), prompting many descriptive studies of the effects of UV-B on plants. UV-B radiation exerts opposing effects on plants. Because it is of high energy, UV-B radiation causes DNA and protein damage, lipid peroxidation, and pigment oxidation (reviewed by Jordan 1995). On the other hand, UV-B can provoke plant photomorphogenic responses (Tevini & Teramura 1989). Some of these responses appear to be adaptive. These include hypocotyl growth inhibition (Ballare et al. 1991;Ballare et al. 1995;Lercari et al. 1990), cotyledon curling (Wilson & Greenberg 1993), and rapid induction of expression of a variety of genes, including those involved in synthesis of phenylpropanoid and flavonoid UV sunscreens (Chappell & Hahlbrock 1984;Kubasek et al. 1992;Li et al. 1993). However, the identity of UV-B photoreceptors and downstream signal transduction components is only partially characterized (Jenkins 1997).

Dicotyledonous plants such as Arabidopsis thaliana follow either of two developmental programs after germination. Dark-grown seedlings show the syndrome called skotomorphogenesis, which is manifested by a long hypocotyl, unopened apical hook, unexpanded cotyledons, and lack of pigment accumulation. In contrast, white light-grown seedlings display a shorter hypocotyl, opened apical hook, expanded cotyledons, and pigmentation. Various components needed for photomorphogenesis have been revealed by genetic screens that take advantage of the contrasting morphological differences between light- and dark-grown A. thaliana seedlings (Quail et al. 1995).

Two screening approaches have been employed to isolate large collections of mutants with lesions in photomorphogenic control: identification of mutants that are etiolated even when grown in the light, and screening for plants that exhibit photomorphogenic traits in the dark. Mutants of the former class are affected in the red/far red phytochrome photoreceptors and their downstream signaling components, phytochrome chromophore biosynthesis (Koornneef et al. 1980;Nagatani et al. 1993;Parks & Quail 1993;Whitelam et al. 1993), or production of a blue light cryptochrome photoreceptor (Ahmad & Cashmore 1993;Koornneef et al. 1980). In contrast, the constitutively photomorphogenic det/cop/fus loss of function Arabidopsis mutants show photomorphogenic phenotypes in darkness. Many of the highly pleiotropic mutants such as det1 and cop1 presumably encode gene products that function as repressors of the light signaling pathway (Chory et al. 1996;Wei & Deng 1996), while det2 is defective in brassinolide steroid hormone biosynthesis (Fujioka et al. 1997;Li et al. 1996).

In contrast to the red/far red and blue/UV-A regions of the solar spectrum, relatively little is known about the receptors or signal transduction pathways for UV-B light in plants (Jenkins 1997). One of the difficulties associated with UV-B photobiology is that light in this region not only signals photomorphogenic responses, but also causes damage. Overcoming this obstacle requires defining UV-B treatments that induce photomorphogenic responses without causing damage. We tested fluence rate responses of cellular morphogenesis and hypocotyl growth inhibition to define fluence rate conditions under which UV-B induces photomorphogenic responses without causing obvious damage. We then investigated UV-B-induced photomorphogenesis in mutants defective in known photomorphogenic signaling pathways to ask whether these are involved in UV-B responses. We report evidence that the PYHA and PHYB photoreceptors and COP1 and DET1 play roles in UV-B-induced photomorphogenesis.


UV-B can act as a photosignal or harmful radiation depending upon the fluence rate

To obtain information about the fluence rate requirement for UV-B-induced photomorphogenesis, fluence rate response of hypocotyl growth inhibition was investigated with A. thaliana ecotype Landsberg erecta (Ler). A range of UV-B fluence treatments was used to obtain a fluence rate response curve of hypocotyl growth. Although some other wavelengths of light were present, the range in fluences was predominantly UV-B. The results shown in Fig. 1 indicate that this curve has three distinct elements. These include growth inhibition, which increased exponentially in the range of 0.01–0.1 μmol m–2 s–1 (very low fluence rate, VLF), a saturated response in the range of 0.1–1.0 μmol m–2 s–1 (low fluence rate, LF), and further growth retardation above 1.0 μmol m–2 s–1 (high fluence rate, HF). To ask whether the observed hypocotyl response resulted specifically from the change of UV-B intensity, a second fluence rate response curve was constructed starting with 0.2 or 2.2 μmol m–2 s–1 UV-B, and using pyrex glass filters to titrate changes in UV-B fluences (0.02–0.1 and 0.2–1.0 μmol m–2 s–1) as described by Landry et al. (1995). The resultant hypocotyl response curve is very similar to that in Fig. 1, indicating that the simultaneous changes in other spectral regions does not affect the hypocotyl growth inhibition (data not shown).

Figure 1.

UV-B fluence response curve of hypocotyl growth inhibition.

Hypocotyl lengths of 5-day-old wild-type seedlings (Ler) were measured after growth in various fluence rates of UV-B. Each symbol and bar represents mean and standard error of more than 20 hypocotyls. Note that the X-axis is in log scale.

Based on the fluence rate response curve, the LF (0.22 μmol m–2 s–1) and HF (2.1 μmol m–2 s–1) UV-B treatments shown in Fig. 2 were used to investigate the differences in more detail. As a negative control for UV-B (–UV-B), plants were grown under mylar filters which remove UV-B without substantial effects on longer wavelengths (Fig. 2). Significant differences were observed in the morphology of seedlings grown at LF UV-B compared with those exposed to HF UV-B (Figs 3 and 4). This strongly suggests that the observed differences in seedling growth and morphogenesis were caused by UV-B rather than other regions of the spectrum.

Figure 2.

Spectral distribution of UV-B light treatments.

Spectral distributions of low (a) and high (b) fluence rate (LF and HF, respectively) UV-B light treatments measured at 10 nm intervals. Note that the Y-axis is in log scale.

Figure 3.

Effects of UV-B on root growth.

Wild-type Ler seedlings were grown for 5 days under LF (0.22 μmol m–2 s–1) or HF (2.1 μmol m–2 s–1) UV-B (+UV-B) or without UV-B (–UV-B). Data represent mean and standard error for more than 20 roots. The absence of error bars indicates very small standard error.

Figure 4.

Effects of UV-B on epidermal cells of hypocotyls and cotyledons.

Cell morphology of Ler hypocotyls (a) and cotyledons (b) of the plants shown in Fig. 3. Seedlings were grown for 5 days under LF (0.22 μmol m–2 s–1) or HF (2.1 μmol m–2 s–1) UV-B (+UV-B) or without UV-B (–UV-B). The size of scale bars is 100 μm.

To examine UV-B effects on root growth, we measured the root length of seedlings following LF and HF treatments. Whereas root growth was not affected in plants under LF of UV-B compared with –UV-B controls, seedlings grown at HF UV-B displayed highly inhibited root growth (Fig. 3). Furthermore, anthocyanin accumulation, which is induced by UV-B and other stresses, was observed in cotyledons and the upper part of hypocotyls of seedlings subjected to HF of UV-B (data not shown).

To investigate UV-B-induced photomorphogenic traits and damage phenotypes at the cellular level, hypocotyl and cotyledon epidermal cells were examined by light microscopy. Hypocotyl cell elongation was inhibited at both LF and HF of UV-B (Fig. 4a), while cotyledon cell expansion was observed only in seedlings at LF of UV-B (Fig. 4b). Root growth and cotyledon cell morphology of A. thaliana seedlings show distinct photomorphogenic phenotypes at low UV-B dose (below 1 μmol m–2 s–1), and damage-related phenotypes at high UV-B dose (above 1 μmol m–2 s–1) under our growth conditions (Figs 3 and 4).

We tested this idea using three classes of mutants previously shown to be hypersensitive to UV-B damage. tt5 is deficient in the accumulation of flavonoid and hydroxycinnamic acid UV-sunscreens (Landry et al. 1995;Li et al. 1993), uvr2–1 is a DNA damage repair mutant deficient in cyclobutylpyrimidine dimer photolyase (Ahmad et al. 1997;Landry et al. 1997), and vtc1 is defective in l-ascorbic acid biosynthesis (Conklin et al. 1996;Conklin et al. 1997). Our working hypothesis is that if LF hypocotyl responses are a result of the damaging effects of UV-B, mutants that are more transparent to UV-B (tt5), or are unable to repair damage products (uvr2 and vtc1), should show increased sensitivity to LF treatment. However, these mutants showed normal hypocotyl growth inhibition at LF (0.22 μmol m–2 s–1) of UV-B (Fig. 5). These results are consistent with the idea that LF UV-B is exerting a photomorphogenic effect, rather than causing cellular damage. Thus, we used this low fluence rate of UV-B for further study of UV-B-induced photomorphogenic signaling.

Figure 5.

Hypocotyl growth of UV-B sensitive mutants under UV-B.

Wild-type (Ler) and Col-0) plants, UV-B sensitive mutants (tt5, uvr2, and vtc1) were grown for 5 days under low fluence rate (0.22 μmol m–2 s–1) of UV-B. Data represent the mean and standard error for more than 20 seedlings.

Either PHYA or PHYB are required for UV-B-induced photomorphogenesis

It is possible that previously identified photoreceptors such as phytochromes and blue light photoreceptors can sense UV-B and signal UV-B responses. To test this hypothesis, we examined UV-B-induced photomorphogenesis with phytochrome defective phyA and phyB mutants, a phytochrome chromophore deficient hy1 mutant, and the cry1 mutant, which is defective in cryptochrome 1 which controls blue light-mediated hypocotyl inhibition. We also examined phyA/phyB double mutant plants because there is evidence on their overlapping functions (Devlin et al. 1996;Reed et al. 1994). Figure 6 shows that LF UV-B-induced hypocotyl growth inhibition is not significantly different in the single mutants compared with the wild type. In contrast, phyA/phyB double mutant seedlings exhibited much longer hypocotyls than the wild-type seedlings under LF UV-B. In addition, only the double mutant seedlings exhibited the skotomorphogenic phenotypes of highly reduced apical hook opening and reduced cotyledon expansion in LF UV-B (data not shown). Furthermore, the double mutant seedlings resembled the wild-type plants grown in darkness or under HF of UV-B (data not shown), reinforcing the idea of distinct LF and HF responses.

Figure 6.

Hypocotyl growth inhibition of photoreceptor deficient mutants under UV-B.

Wild-type (Ler) plants, phytochrome deficient mutants (hy1, phyA, phyB, and phyA/phyB) and a cryptochrome blue light photoreceptor deficient cry1 mutant were grown for 5 days under low fluence rate (0.22 μmol m–2 s–1) of UV-B. Data represent mean and standard error for more than 20 seedlings.

The pleiotropic constitutive photomorphogenic cop1 and det1 mutants do not respond to UV-B

To investigate the responses of the downstream mutants to LF UV-B-induced photomorphogenesis, as shown in Fig. 7, we examined UV-B-induced hypocotyl growth inhibition in the signal transduction mutants det1, det2 and cop1. The hypocotyl length of dark grown det1 and cop1 seedlings is not significantly different from that grown under UV-B. In contrast, hypocotyl growth of det2 seedlings was inhibited by LF of UV-B. These data suggest that DET1 and COP1 are required for LF UV-B-induced photomorphogenesis, while the lack of brassinolide in det2 seedlings does not affect UV-B-induced photomorphogenesis.

Figure 7.

UV-B mediated hypocotyl growth inhibition of constitutive photomorphogenic mutants.

The constitutive photomorphogenic cop1, det1 and det2 mutants, and their parental wild-type Col-0 plants were grown under low fluence rate (0.22 μmol m–2 s–1) UV-B. Data represent mean and standard error for more than 20 seedlings. The absence of error bars indicates very small standard error.


It is widely accepted that photomorphogenesis in plants is mediated by at least three different classes of photoreceptors that specifically sense different wavelengths of light: red/far red sensing phytochromes, blue/UV-A photoreceptors, and UV-B photoreceptors. Hypocotyl growth inhibition appears to be influenced by all three regions of the spectrum, and fluence response data for A. thaliana hypocotyl growth inhibition have been used to construct an action spectrum (Goto et al. 1993). In contrast to our increasingly sophisticated understanding of blue and red/far red signaling pathways, we know very little about UV-B mediated response pathways (Jenkins 1997).

Although UV-B can damage plants, it is proposed to convey photomorphogenic information (Tevini & Teramura 1989), and there are several reports that UV-B-induced hypocotyl growth inhibition is not necessarily a damage response, but a true photomorphogenic effect (Ballare et al. 1991;Ballare et al. 1995;Lercari et al. 1990). We examined this question directly by measuring the dose response of hypocotyl inhibition by UV-B. Based on our fluence rate response curve, we propose that the complex hypocotyl growth inhibition observed in response to UV-B (Fig. 1) includes photomorphogenic and stress components. It is interesting that the biphasic nature of the hypocotyl UV-B growth curve reported here (Fig. 1) is reminiscent of that seen for red light-mediated elongation growth inhibition in oat and corn tissues (Blaauw et al. 1968;Mandoli & Briggs 1981;Vanderhoef et al. 1979) known to be mediated by the phytochrome system. The nature of biphasivity in red-light mediated inhibition is more likely based on contributions of different members of the phytochrome family than resulting from stress. Supporting this concept are reports that phytochrome A and B differentially mediate anthocyanin biosynthesis (Kerckhoffs et al. 1997) and seed germination (Shinomura et al. 1996) in a fluence rate dependent manner.

Although our data implicate phytochrome in UV-B photomorphogenesis (Fig. 6, discussed later in this section), the biphasic nature of the growth inhibition in response to UV-B (Fig. 1) can be explained as photomorphogenesis at low fluence rates, and stress-induced responses at high fluence rates. This is unlike results reported by Young et al. (1992) of fluence response curves for hypocotyl elongation because they used UV-A, blue, red, and far-red light treatments, and focused only on hy6, which may be a leaky mutant. We specifically examined UV-B and used a variety of types of photomorphogenic mutants, including a phyA phyB double mutant, to test whether LF and HF UV-B provoke distinct responses.

To determine whether the cause of hypocotyl growth inhibition in LF UV-B is different than in HF UV-B, we measured root growth and cotyledon cell expansion. Root growth was unaffected in LF UV-B (Fig. 3), although photomorphogenic hypocotyl growth inhibition occurred (Fig. 1). In contrast, root growth was inhibited under HF UV-B (Fig. 3) and we conclude that this was a stress response to the damaging effects of UV-B. Stronger evidence is that the typically photomorphogenic cotyledon cell expansion in response to white light was evident in response to LF UV-B (Fig. 4). The typical photomorphogenic combination of seedling cotyledon cell expansion and hypocotyl growth are evident in LF UV-B (Fig. 4), leading us to propose that these responses are truly photomorphogenic. In contrast, HF UV-B inhibited the photomorphogenic expansion of cotyledons (Fig. 4b) and inhibited root growth (Fig. 3) and we conclude that these, along with the further hypocotyl growth inhibition that occurred under HF UV-B (Fig. 1) are stress responses. In further support that LF UV-B is photomorphogenic and not stressful, hypocotyl response in UV-B stress sensitive mutants were identical to wild type (Fig. 5).

We used the results obtained with wild-type plants to examine the LF hypocotyl growth inhibition in well-characterized photomorphogenic mutants of Arabidopsis (Fig. 6). These results suggest a role for phytochrome in this signaling pathway. Interestingly, while the phyA/phyB double mutant was insensitive to LF UV-B, the phyA and phyB single mutants and hy1 chromophore mutant had wild-type responses.

These results are consistent with two sets of published observations on Arabidopsis phytochrome mutants. First, our data suggest that phyA and phyB are functionally redundant for the LF UV-B hypocotyl inhibition response, and it is necessary to ablate both the light labile phyA and light stable phyB proteins. This is consistent with observations that red light-mediated cotyledon expansion and cab gene expression are normal in phyA or phyB single mutants, but reduced in the double mutant (Reed et al. 1994).

The results demonstrating that the hy1 chromophore mutant has a wild-type UV-B-induced hypocotyl inhibition response are consistent with the previous observation that this is a leaky mutant, which accumulates a small amount of functional phytochromes (Parks & Quail 1991). In contrast to our results, normal UV-B photomorphogenic responses were reported for the lh mutant of cucumber, which is deficient in light stable phytochrome (Ballare et al. 1995), and the phytochrome chromophore biosynthesis aurea mutant of tomato (Lercari et al. 1990). Based upon the results described above, we hypothesize that the lack of altered UV-B photoresponses in these mutants is due to residual phytochrome activity.

Taken together, these results and other published work suggest that the LF UV-B-induced hypocotyl response is mediated by more than one class of phytochrome, and only a small amount of the active photoreceptor is needed for a wild-type response. The requirement for phytochrome in LF UV-B photomorphogenesis (Fig. 6) may result from its role as a photoreceptor. In fact, phytochrome is known to absorb light in the UV-B region of the spectrum. Alternatively, phytochrome may be required for co-action with a UV-B photoreceptor, in which case phytochrome would be required as an accessory component (Mohr 1994).

Fig. 6 shows that the cry1 mutation does not affect UV-B-induced hypocotyl growth inhibition. These data suggest that CRY1, a blue light photoreceptor (Ahmad & Cashmore 1993), has no significant role in UV-B signal transduction leading to the hypocotyl response. These results are consistent with the observation that, while CRY1 mediates the UV-A/blue light induction of chalcone synthase (Jenkins 1997), cry1 mutant plants have normal UV-B-induced chalcone synthase gene induction (Fuglevand et al. 1996). Pharmacological experiments in Arabidopsis also point to distinct UV-A and UV-B signal transduction pathways for chalcone synthase gene induction (Fuglevand et al. 1996). Batschauer et al. (1996) reported that both single and double mutants of phytochrome are similar to the wild type in UV-A-mediated induction of CHS, indicating that there is a distinct UV-A photoreceptor. This is different from our results, which argue for a phytochrome requirement for UV-B-induced photomorphogenesis.

An interesting complexity to our understanding of blue and UV light signaling was recently introduced by the identification of A. thaliana CRY2, encoding a cryptochrome homologue (Lin et al. 1996). More recently, Lin et al. (1998) demonstrated that CRY1 and CRY2 have overlapping functions in blue light-mediated inhibition of hypocotyl growth: CRY1 predominates at high fluence rates, while CRY2 is functional at low fluence rates. While CRY1 was not implicated in the responses reported here (Fig. 6), it remains to be seen whether CRY2 has a role in UV-B signal perception or if it might work by co-action with phytochrome.

The highly pleiotropic constitutively photomorphogenic (cop) or de-etiolated (det) mutants manifest a variety of de-etiolated phenotypes even when germinated in complete darkness. Hypocotyl growth of det1 and cop1 seedlings is indifferent to LF UV-B, while those of det2 exhibited LF UV-B photomorphogenesis (Fig. 7). This implicates DET1 and COP1 in the LF UV-B responses reported here, and is consistent with the idea that DET1 and COP1 repress multiple light signaling pathways in darkness (Chory et al. 1996;Wei & Deng 1996). Based upon published reports and our data (Fig. 7) we conclude that these molecules are needed for the repression of UV-B-induced photomorphogenesis in darkness. In contrast to the repressor mutants, the brassinolide deficient det2 mutant retained sensitivity to UV-B. Therefore, DET2 is not necessary for the LF UV-B-mediated photomorphogenesis described in this paper.

Although our results suggest that PHYA, PHYB, COP1 and DET1 are involved in UV-B-induced photomorphogenesis, more studies are needed to further elucidate the photoreceptor(s) and downstream signal transduction elements of UV-B-induced photomorphogenesis. Our ability to define conditions that discriminate between damage effects and photomorphogenesis should facilitate isolation of mutants defective in UV-B signal transduction. This would allow identification of other UV-B sensing and response components in Arabidopsis.

Experimental procedures

Plant materials and growth conditions

Seeds of phyA-201, det1–1 and det2–1 mutant lines were obtained from the Arabidopsis Biological Resource Center (ABRC, Ohio State University, USA). The hy1, phyB-1 and cry1–1 mutant lines were obtained from the Nottingham Arabidopsis Stock Centre (NASC, Nottingham, UK). The cop1–6, uvr2 and vtc1 (formerly soz1) mutant seeds were provided by Drs Xing-Wang Deng, Laurie Landry and Patricia Conklin, respectively. The phyA/phyB double mutant line was obtained from a cross between the phyA-201 and phyB-1 homozygous mutants, and the genotype was confirmed by test crosses to each homozygous single mutant line.

Seeds were surface-sterilized with commercial bleach, washed five times with sterile distilled water, and sown on 0.75% agar without salts or sugars. The plates were then placed at 4°C for 2 days. After the cold treatment, seeds were illuminated with white light for 2 h and germinated in darkness for 2 days at room temperature. After germination, seedlings were grown in darkness or continuous UV-B at 22°C for 3 days.

Light sources

White light was supplied by cool white fluorescent tubes (CW1500; General Electric, USA). UV-B light was supplied by UV-B fluorescent tubes (F40UVB; Philips, New Jersey, USA) filtered through 3 mm-thick Pyrex glass (Landry et al. 1997). Under these experimental conditions, there were also changes in spectral properties outside the UV-B range. Pyrex glass filters were used to vary the UV-B fluence rate without affecting other spectral regions (Landry et al. 1995). UV-B fluence rates were varied by changing the number of Pyrex filters, and by altering the UV-B bulb output with a rheostat. To remove the wavelengths in the UV-B region (–UV-B controls), UV-B light was filtered through two layers of 0.13 mm-thick mylar (AIN Plastics, Mt. Vernon, NY, USA). Spectral energy distributions of UV-B light sources (Fig. 2) were measured with a spectroradiometer (Model OL 752; Optronics Laboratory, Florida, USA) calibrated with an OL 752–150 calibration module and OL752–10 spectral irradiance standard.

Hypocotyl and root length measurements

Hypocotyl and root length measurements were performed in triplicate, using >20 seedlings for each measurement. All experiments were repeated at least four times with similar results. To minimize experimental variation, the seedlings of all mutant lines under direct comparison were grown on the same Petri plates. The length was measured with a 0.5 mm scale ruler.

Light microscopy

Epidermal imprints of hypocotyls and cotyledons were prepared using 3% molten agar solution as described by Mathur & Koncz (1997). The images were then viewed in an inverted light microscope (Olympus, Japan), photographed on color print film, and the prints were rephotographed on black and white print film (Kodak, USA).


We thank Dr Xing-Wang Deng for providing cop1–6 seeds, Dr Laurie Landry for the photolyase deficient uvr2–1 seeds, Dr Patricia Conklin for the ascorbate deficient vtc1 mutant, and ABRC and NASC for providing mutant seeds. This research was supported by National Science Foundation Presidential Young Investigator Award DMB 90–58134 (to R.L.L.) and a postdoctoral fellowship from the Korean Science Foundation (to B.K.).