I NTRODUCTIQNPkytochrome is one of the most fasrinsring; proteins in plants. A great many striking photomorphoyenctie proccssex arc mediated by it (see [l]). Phytechromc is well charactcrizcd by biochemical and immunological techniques and by methods OF molecular biology, Dcspirc this fact, the mechanism of action of phytochrome, the molecular links between its conformational alterations and the ensuing biochemical and morphogenetic phenomena are still unknown, Many sophisticated experiments have not been able to unveil the secret of light-induced signal-transduction by phytochromc. Here we report on findings which we assume could be evidence of the mode of action and the phylogeny of this light-sensing protein. MATERIALS AND METHODS I. Prepurct~ion of clones uttd scqrrctlcitrgLibraries of cDNA were constructed in the Xgti 1 expression vector. Phages giving a positive response with a mo~wA~nal antibody (Z-3Bl

A C-terminal section of phytochromes turned out to share sequence homologies with the full length of the transmitter modules (about 250 amino acids) of bacterial sensor proteins. Coinciding hydrophobic clusters within the homologous domains imply that the overall folding of the two different types of peptides is similar. Hence, phytochromes appear to possess the structural prerequisites to transmit signals in a way bacterial sensor proteins do. The bacterial sensor proteins are known to be environmental stimuli-regulated kinases belonging to two-component systems. After sensing a stimulus by the N-terminal part of the sensor protein, conformational alterations confer the signal to its (mostly) C-terminal transmitter module which in turn is transitionally autophosphorylated at a conserved histidine. From the histidine the phosphate is transferred to the receiver module of a system-specific regulator protein which eventually acts on transcription or enzyme activity. The histidine is not conserved in phytochromes. Instead, a conserved tyrosine is found spatially very close to the histidine position. This tyrosine might play the role of histidine, and kinase function might be associated with this part of phytochrome. In spite of this divergence, the structural similarities point to a common evolutionary origin of the phytochrome and bacterial modules.

Gymnosperms possess two to four phytochrome types which apparently are the result of successive gene duplications in the genomes of their common ancestors. Phytochromes are nuclear-encoded proteins whose genes, contrary to chloroplast, mitochondrion, and rRNA genes, have hitherto rarely been used to examine gymnosperm phylogenies. Since the individual phytochrome gene types implied phylogenies that were not completely congruent to one another, conflicting branching orders were sorted by the number of gene lineages present in a taxon. The Gnetatae (two gene types) branched at the base of all gymnosperms, a position supported by bootstrap sampling (distance and character state trees, maximum likelihood). The Gnetatae were followed by Ginkgo, Cycadatae, and Pinaceae (three gene types) and the remaining conifers (four gene types). Therefore, in phytochrome trees, the most ancient branch of the conifers (Pinatae) seems to be the Pinaceae. The next split appears to have separated Araucariaceae plus Podocarpaceae from the Taxaceae/Taxodiaceae/Cupressaceae group. Structural arrangements in the plastid genomes (Raubeson and Jansen 1992) corroborate the finding that there is no close connection between Pinaceae and Gnetatae as suggested by some publications. The analyses are based on 60 phytochrome genes (579 positions in an alignment of PCR fragments) from 28 species. According to rough divergence time estimates, the last common ancestor of gymnosperms and angiosperms is likely to have existed in the Carboniferous.

A protein consisting of 60 kDa subunits (As-P60) was isolated from etiolated oat seedlings (Avena sativa L.) and characterized as avenacosidase, a beta-glucosidase that belongs to a preformed defence system of oat against fungal infection. The enzyme is highly aggregated; it consists of 300-350 kDa aggregates and multimers thereof. Dissociation by freezing/thawing leads to complete loss of enzyme activity. The specificity of the enzyme was investigated with para-nitrophenyl derivatives which serve as substrates, in decreasing order beta-fucoside, beta-glucoside, beta-galactoside, beta-xyloside. The corresponding orthonitrophenyl glycosides are less well accepted. No hydrolysis was found with alpha-glycosides and beta-thioglucoside. An anti-As-P60 antiserum was prepared and used for isolation of a cDNA clone coding for As-P60. A presequence of 55 amino acid residues was deduced from comparison of the cDNA sequence with the N-terminal sequence determined by Edman degradation of the mature protein. The presequence has the characteristics of a stroma-directing signal peptide; localization of As-P60 in plastids of oat seedlings was confirmed by western blotting. The amino acid sequence revealed significant homology (> 39% sequence identity) to beta-glucosidases that are constituents of a defence mechanism in dicotyledonous plants. 34% sequence identity was even found with mammalian and bacterial beta-glucosidases of the BGA family. Avenacosidase extends the occurrence of this family of beta-glucosidases to monocotyledonous plants.

The phytochromes, a class of plant light‐sensing pigments, are a gene family with a long, complex evolutionary history. Angiosperms each have five or more phytochromes (designated A to E in Arabidopsis) with distinct functions as light receptors and only moderate sequence identities for different types within a species. The long‐term challenge taken up here is to trace the origin and function of the various motifs within the angiosperm phytochromes through gymnosperm phytochromes (types N, O and P) and lower plant phytochromes, sometimes reaching even to bacterial progenitor molecules. Particularly intriguing are the findings of homology of a C‐terminal region of phytochromes with bacterial transmitter modules and of a large N‐terminal region with a protein encoded by a gene from the cyanobacterum Synechocystis. Phylogenetic analysis helps to answer general questions such as the times of divergence of mono‐ and dicotyledons, of groups of gymnosperms or of ferns. Phytochrome sequences suggest (1) that mono‐ and dicotyledons became separated 150‐200 million years earlier than indicated by the fossil record and (2) that Ginkgo and Cycas have been separated unexpectedly late from the lineage giving rise to the Pinidae. (3) The status of Psilotum as a close relative of the primeval vascular plants is not supported. Phytochrome gene sequences additionally reveal that (4) moss and fern phytochromes have erratically acquired C‐termini which, though kinase‐like, are different from the common ones and that (5) introns have been lost, gained or shifted in position from algae to angiosperms. Phytochromes promise to be a rich source of phylogenetic information into the future as more sequences and functional data emerge, not least from studies of lower plants.