Lhcb and other nuclear genes for chloroplastic proteins are regulated by several signals. Among them, light and retrograde signals from the plastid itself appear to act through closely related mechanisms. To investigate this interaction, we analysed an Arabidopsis mutant, hy1, deficient in plastidic heme oxygenase. hy1 is defective in phytochrome chromophore biosynthesis, which has other indirect effects on tetrapyrrole metabolism. We generated double mutants between hy1-6.2, genetically a null mutation, and three known gun (genome uncoupled) mutants, defective in retrograde plastid signalling. Recent molecular evidence shows GUN5 to be involved in tetrapyrrole metabolism (N. Mochizuki and J. Chory, manuscript in preparation). We observed hy1gun4 to be semi-albino plants, and hy1gun5 albino lethal, in a high-light-sensitive manner. Both double mutants showed defective greening and chloroplast development, and expressed Lhcb at reduced levels specifically in high light. Their degree of 'genome uncoupling' (Lhcb expression in the absence of functional chloroplasts) was similar to that observed in single mutants. These results can be interpreted as a metabolic (rather than genetic) interaction between HY1 and GUN4 or GUN5, and this in turn supports the involvement of tetrapyrroles as plastid signals. The tetrapyrrole precursor 5-aminolevulinic acid (ALA) inhibited Lhcb expression in hy1. Surprisingly, ALA also rescued photomorphogenesis of hy1. We speculate that either one tetrapyrrole intermediate, which can accumulate anomalously in hy1, or an altered ratio between two intermediates, plays a role as a repressor of Lhcb expression. gun1 did not exacerbate the plastid or Lhcb expression phenotype of hy1. This can be interpreted as a role for gun1 strictly on the same pathway as hy1 or, more likely, as evidence for the existence of at least one separate, non-tetrapyrrole related plastid signal.

SummaryLhcb and other nuclear genes for chloroplastic proteins are regulated by several signals. Among them, light and retrograde signals from the plastid itself appear to act through closely related mechanisms. To investigate this interaction, we analysed an Arabidopsis mutant, hy1, de®cient in plastidic heme oxygenase. hy1 is defective in phytochrome chromophore biosynthesis, which has other indirect effects on tetrapyrrole metabolism. We generated double mutants between hy1-6.2, genetically a null mutation, and three known gun (genome uncoupled) mutants, defective in retrograde plastid signalling. Recent molecular evidence shows GUN5 to be involved in tetrapyrrole metabolism (N. Mochizuki and J. Chory, manuscript in preparation). We observed hy1gun4 to be semi-albino plants, and hy1gun5 albino lethal, in a high-light-sensitive manner. Both double mutants showed defective greening and chloroplast development, and expressed Lhcb at reduced levels speci®cally in high light. Their degree of`genome uncoupling' (Lhcb expression in the absence of functional chloroplasts) was similar to that observed in single mutants. These results can be interpreted as a metabolic (rather than genetic) interaction between HY1 and GUN4 or GUN5, and this in turn supports the involvement of tetrapyrroles as plastid signals. The tetrapyrrole precursor 5-aminolevulinic acid (ALA) inhibited Lhcb expression in hy1. Surprisingly, ALA also rescued photomorphogenesis of hy1. We speculate that either one tetrapyrrole intermediate, which can accumulate anomalously in hy1, or an altered ratio between two intermediates, plays a role as a repressor of Lhcb expression. gun1 did not exacerbate the plastid or Lhcb expression phenotype of hy1. This can be interpreted as a role for gun1 strictly on the same pathway as hy1 or, more likely, as evidence for the existence of at least one separate, non-tetrapyrrole related plastid signal.