The catalase multigene family in Arabidopsis includes three genes encoding individual subunits that associate to form at least six isozymes that are readily resolved by nondenaturing gel electrophoresis. CATl and CAT3map to chromosome 1 , and CAT2maps to chromosome 4. The nucleotide sequences of the three coding regions are 70 to 72% identical. The amino acid sequences of the three catalase subunits are 75 to 84% identical and 87 to 94% similar, considering conservative substitutions. Both the individual isozymes and the individual subunit mRNAs show distinct patterns of spatial (organ-specific) expression. Six isozymes are detected in flowers and leaves and two are seen in roots. Similarly, mRNA abundance of the three genes varies among organs. All three mRNAs are highly expressed in bolts, and CAT2 and CAT3 are highly expressed in leaves.

In Arabidopsis thaliana, catalase is encoded by a small gene family. We have characterized cDNA and genomic clones containing the Arabidopsis catalase gene CAT3, present as a single copy in the nuclear genome. Six introns were identified in the CAT3 coding region and two transcription start sites have been been mapped by primer extension. The deduced amino acid sequence of CAT3 is highly similar to other catalases. CAT3 expression is similar in seedlings germinated and grown either in continuous light or in continuous dark, suggesting that CAT3 expression in seedlings is not light responsive. CAT3 expression is controlled by the circadian clock; in 5-week-old plants grown on a light-dark cycle and then transferred to continuous light, robust oscillations in CAT3 mRNA abundance with circadian period persist for at least five circadian cycles. Interestingly, the peak in CAT3 mRNA abundance occurs in the subjective evening, which is out of phase with expression of the Arabidopsis CAT2 catalase gene, which shows clock-regulated expression gated to the subjective early morning.

Hai H o n g Zhong,all A n d r e w S. Resnick,a,2 M a r t i n Straume,b a n d C. R o b e r t s o n M~C l u n g~.~ a Department of Biological Sciences, Dartmouth College, 6044 Gilman Laboratory, Hanover, New Hampshire 03755 and Metabolism, University of Virginia, Charlottesville, Virginia 22903 National Science Foundation Center for Biological Timing, Department of Interna1 Medicine, Division of Endocrinology Persistent oscillation in constant conditions is a defining characteristic of circadian rhythms. However, in plants transferred into extended dark conditions, circadian rhythms in mRNA abundance commonly damp in amplitude over two or three cycles to a steady state level of relatively constant, low mRNA abundance. In Arabidopsis, catalase CAT3 mRNA oscillations damp rapidly in extended dark conditions, but unlike catalase CAT2 and the chlorophyll a/b binding protein gene CAB, in which the circadian oscillations damp to low steady state mRNA abundance, CAT3 mRNA oscillations damp to high steady state levels of mRNA abundance. Mutational disruption of either phytochrome-or cryptochromemediated light perception prevents damping of the oscillations in CAT3 mRNA abundance and reveals strong circadian oscillations that persist for multiple cycles in extended dark conditions. Damping of CAT3 mRNA oscillations specifically requires phytochrome A but not phytochrome B and also requires the cryptochromel blue light receptor. Therefore, we conclude that synergistic signaling mediated through both phytochrome A and cryptochromel is required for damping of circadian CAT3 mRNA oscillations in extended dark conditions.

In Arabidopsis seedlings germinated and grown in continuous light, CATZ mRNA abundance peaks 1 d after imbibition, consistent with the role of catalase in detoxifying HzOz generated during the 8-oxidation of fatty acids stored in the seed. A second peak of CATZ mRNA abundance, of lower amplitude than the initial peak, appears 6 d after imbibition and may be associated with the development of photosynthetic competence and induction of photorespiration. This second peak in steady-state CATZ mRNA abundance is regulated by light and is not seen in etiolated seedlings. CATZ mRNA accumulation is induced by exposure to high-fluence blue or far-red light but not by red light. In addition, light induction is unaffected by severa1 mutations that block blue light-mediated inhibition of hypocotyl elongation (blul, blu2, blu3, hy4), suggesting phytochrome involvement. When etiolated seedlings are transferred to continuous white light, CATZ mRNA rapidly (within 30 min) accumulates. It is interesting that in these seedlings CAT2 mRNA abundance undergoes pronounced oscillations with a circadian (24 h) periodicity, indicating control by the endogenous circadian clock. No such oscillations are detected in CAT2 mRNA abundance in etiolated seedlings prior to illumination. Control of CATZ expression by the circadian clock is also seen in 5-week-old plants grown in a light-dark cycle and transferred either to continu o u~ dark or to continuous light; in continuous light the circadian oscillations in CATZ mRNA abundance persist for at least five circadian cycles, indicating the robustness of this circadian rhythm.

Circadian rhythms in the abundance of the CAT2 catalase mRNA were not seen in etiolated seedlings but developed upon illumination. These circadian oscillations were preceded by a rapid and transient induction of CAT2 mRNA abundance that varied strikingly according to the timing (circadian phase) of the onset of illumination. This variation oscillated with a circadian periodicity of approximately 28 hr, indicating that the circadian oscillator is running in etiolated seedlings and regulates (gates) the induction of CAT2 by light. Moreover, because we assayed populations of seedlings, we infer that the individual clocks among populations of etiolated seedlings were synchronized before the onset of illumination. What developmental or environmental signals synchronized the clocks among seedlings? Varying the phase of the onset of illumination relative to release from stratification failed to affect the acute induction of CAT2, indicating that the temperature step from 4 to 22 degrees C associated with release from stratification did not reset the circadian clock. However, the acute induction of CAT2 mRNA varied with time after imbibition, demonstrating that imbibition provides a signal capable of resetting the circadian clock and of synchronizing the clocks among populations of seedlings.