The specialized reproductive functions of angiosperm pistils are dependent in part upon the regulated activation of numerous genes expressed predominantly in this organ system. To better understand the nature of these pistil-predominant gene products we have analyzed seven cDNA clones isolated from tomato pistils through differential hybridization screening. Six of the seven cDNAs represent sequences previously undescribed in tomato, each having a unique pistil- and/or floral-predominant expression pattern. The putative protein products encoded by six of the cDNAs have been identified by their similarity to sequences in the database of previously sequenced genes, with a seventh sequence having no significant similarity with any previously reported sequence. Three of the putative proteins appear to be targeted to the endomembrane system and include an endo-beta-1,4-glucanase which is expressed exclusively in pistils at early stages of development, and proteins similar in sequence to gamma-thionin and miraculin which are expressed in immature pistils and stamens, and in either sepals or petals, respectively. Two other clones, similar in sequence to each other, were expressed primarily in immature pistils and stamens and encode distinct proteins with similarity to leucine aminopeptidases. An additional clone, which encodes a protein similar in sequence to the enzyme hyoscyamine 6-beta-hydroxylase and to other members of the family of Fe2+/ascorbate-dependent oxidases, was expressed at high levels in pistils, stamens and sepals, and at detectable levels in some vegetative organs. Together, these observations provide new insight into the nature and possible functional roles of genes expressed during reproductive development.

Leucine aminopeptidase (LAP) is induced by wounding and bacterial pathogen infection in tomato. DNA blot analysis of XbaI-digested lambdalap genomic clones demonstrated that LapA1 and LapA2 cDNAs were encoded by two different LapA genes in the tomato genome. The coding and untranslated regions of LapA1 and LapA2 mRNAs shared more than 93% identity. The deduced amino acid sequences of LapA cDNA clones and in vitro translation of LapA1 mRNA indicated that LAP-A was synthesized as a 60-kDa precursor protein. The processing of a 60-kDa preLAP-A into the mature 55-kDa LAP-A was demonstrated in vivo by expression of the full-length LapA1 cDNA in insect cells. Sequencing of a single LAP-A form isolated from a two-dimensional polyacrylamide gel indicated that LAP-A proteins had two different N termini that were separated by two residues. The LAP-A presequence had features similar to chloroplast transit peptides. Comparison of LAP-A levels in chloroplast and total protein extracts from methyl jasmonate-treated leaves indicated that a small proportion of the LAP-A proteins was detected in the plastids. Inspection of the LAP-A presequence indicated the presence of a dibasic protease (Kex2/furin) processing site motif 6-8 residues upstream from the LAP-A N termini. Its potential role in LAP-A precursor biogenesis is discussed.

Wounding of tomato leaves results in the accumulation of an exoprotease called leucine aminopeptidase (LAP-A). While the expression of LapA genes are well characterized, the specificity of the LAP-A enzyme has not been studied. The LAP-A preprotein and mature polypeptide were overexpressed in Escherichia coli. PreLAP-A was not processed and was inactive accumulating in inclusion bodies. In contrast, 55-kDa mature LAP-A subunits assembled into an active, 357-kDa enzyme in E. coli. LAP-A from E. coli cultures was purified to apparent homogeneity and characterized relative to its animal (porcine LAP) and prokaryotic (E. coli PepA) homologues. Similar to the porcine and E. coli enzymes, the tomato LAP-A had high temperature and pH optima. Mn2+ was a strong activator for all three enzymes, while chelators, zinc ion, and the slow-binding aminopeptidase inhibitors (amastatin and bestatin) strongly inhibited activities of all three LAPs. The substrate specificities of porcine, E. coli and tomato LAPs were determined using amino-acid-p-nitroanilide and -beta-naphthylamide substrates. The tomato LAP-A preferentially hydrolyzed substrates with N-terminal Leu, Met and Arg residues. LAP-A had substantially lower levels of activity on other chromogenic substrates. Several differences in substrate specificities for the animal, plant and prokaryotic enzymes were noted.

The wound-induced leucine aminopeptidase (EC 3.4.11.1) genes, LapA1 and LapA2, from tomato (Lycopersicon esculentum Mill.) were isolated and characterized. The genes were organized in a tandem array with approximately 6 kb separating their coding regions. Quantitation of LapA mRNA levels in conjunction with nuclear run-on experiments indicated that LapA genes were primarily under transcriptional control after wounding and infection with Pseudomonas syringae pv. tomato. In contrast, actin genes were down-regulated after pathogen infection. The sequences of the LapA1 and LapA2 5'-flanking regions were determined and several potential regulatory motifs were identified. Ribonuclease protection studies revealed that LapA1 and LapA2 had short 18-bp 5'-untranslated regions (UTR), both genes were expressed after wounding, and LapA1 mRNAs were 3.3-fold more abundant than LapA2 transcripts. While the region surrounding LapA1 was conserved, the 3'-UTRs and 3'-flanking regions of LapA2 had diverged in two inbred tomato lines. The accumulation of LapA mRNAs and of LAP-A (acidic pI), LAP-N (neutral pI) and LAP-related proteins were examined in two monocot and five dicot species. The LAP-N and 66- and 77-kDa LAP-related proteins were detected in healthy and wounded leaves of all plants examined. The LAP-A proteins were only detected in nightshade and their accumulation was distinct from that observed in tomato.

Tomatoes (Lycopersicon esculentum) express two forms of leucine aminopeptidase (LAP-A and LAP-N) and two LAP-like proteins. The relatedness of LAP-N and LAP-A was determined using affinity-purified antibodies to four LAP-A protein domains. Antibodies to epitopes in the most N-terminal region were able to discriminate between LAP-A and LAP-N, whereas antibodies recognizing central and COOH-terminal regions recognized both LAP polypeptides. Two-dimensional immunoblots showed that LAP-N and the LAP-like proteins were detected in all vegetative (leaves, stems, roots, and cotyledons) and reproductive (pistils, sepals, petals, stamens, and floral buds) organs examined, whereas LAP-A exhibited a distinct expression program. LapN was a single-copy gene encoding a rare-class transcript. A full-length LapN cDNA clone was isolated, and the deduced sequence had 77% peptide sequence identity with the wound-induced LAP-A. Comparison of LAP-N with other plant LAPs identified 28 signature residues that classified LAP proteins as LAP-N or LAP-A like. Overexpression of a His(6)-LAP-N fusion protein in Escherichia coli demonstrated distinct differences in His(6)-LAP-N and His(6)-LAP-A activities. Similar to LapA, the LapN RNA encoded a precursor protein with a molecular mass of 60 kD. The 5-kD presequence had features similar to plastid transit peptides, and processing of the LAP-N presequence could generate the mature 55-kD LAP-N. Unlike LapA, the LapN transcript contained a second in-frame ATG, and utilization of this potential initiation codon would yield a 55-kD LAP-N protein. The localization of LAP-N could be controlled by the balance of translational initiation site utilization and LAP-N preprotein processing.