Amyloplast to chromoplast conversion in developing ornamental tobacco floral nectaries provides sugar for nectar and antioxidants for protection.

Tobacco floral nectaries undergo changes in form and function. As nectaries change from green to orange, a new pigment is expressed. Analysis demonstrated that it is ß-carotene. Plastids undergo dramatic changes. Early in nectary development, they divide and by stage 9 (S9) they are engorged with starch. About S9, nectaries shift from quiescent anabolism to active catabolism resulting in starch breakdown and production of nectar sugars. Starch is replaced by osmiophilic bodies, which contain needle-like carotenoid crystals. Between S9 and S12, amyloplasts are converted to chromoplasts. Changes in carotenoids and ascorbate were assayed and are expressed at low levels early in development; however, following S9 metabolic shift, syntheses of ß-carotene and ascorbate greatly increase in advance of expression of nectar redox cycle. Transcript analysis for carotenoid and ascorbate biosynthetic pathways showed that these genes are significantly expressed at S6, prior to the S9 metabolic shift. Thus, formation of antioxidants ß-carotene and ascorbate after the metabolic shift is independent of transcriptional regulation. We propose that biosynthesis of these antioxidants is governed by availability of substrate molecules that arise from starch breakdown. These processes and events may be amenable to molecular manipulation to provide a better system for insect attraction, cross pollination, and hybridization.

Tobacco nectarin I. Purification and characterization as a germin-like, manganese superoxide dismutase implicated in the defensE of floral reproductive tissues.

Nectarin I, a protein that accumulates in the nectar of Nicotiana sp. , was determined to contain superoxide dismutase activity by colorimetric and in-gel assays. This activity was found to be remarkably thermostable. Extended incubations at temperatures up to 90 degrees C did not diminish the superoxide dismutase activity of nectarin I. This attribute allowed nectarin I to be purified to homogeneity by heat denaturation of the other nectar proteins. By SDS-polyacrylamide gel electrophoresis, nectarin I appeared as a 29-kDa monomer. If the protein sample was not boiled prior to loading the gel, then nectarin I migrated as 165-kDa oligomeric protein. By matrix-assisted laser desorption/ionization time-of-flight mass spectrometry, the protomer subunit was found to be a 22.5-kDa protein. Purified nectarin I contained 0.5 atoms of manganese/monomer, and the superoxide dismutase activity of nectarin I was not inhibited by either H(2)O(2) or NaCN. Following denaturation, the superoxide dismutase activity was restored after Mn(2+) addition. Addition of Fe(2+), Cu(2+), Zn(2+), and Cu(2+)/Zn(2+) did not restore superoxide dismutase activity. The quaternary structure of the reconstituted enzyme was examined, and only tetrameric and pentameric aggregates were enzymatically active. The reconstituted enzyme was also shown to generate H(2)O(2). Putative nectarin I homologues were found in the nectars of several other plant species.

Fluoroorotic acid-selected Nicotiana plumbaginifolia cell lines with a stable thymine starvation phenotype have lost the thymine-regulated transcriptional program.

We have selected 143 independent Nicotiana plumbaginifolia cell lines that survive in the presence of 5-fluoroorotic acid. These lines show several diverse phenotypes. The majority of these cell lines showed reduced levels of UMP synthase. However, one particular phenotype, which represents 14% of the total independent lines (20 cell lines), showed an unexpected, high level of UMP synthase and was therefore analyzed in detail. The selected cell lines showed no differences with wild-type cells with respect to uptake of orotic acid, affinity of UMP synthase for its substrates, or UMP synthase gene-copy number. Alternative detoxification mechanisms were also excluded. The elevated enzyme activity was correlated with elevated UMP synthase protein levels as well as elevated UMP synthase mRNA levels. In contrast to wild-type cell lines, the fluoroorotic acid-selected cell lines did not respond to thymine or to other biochemicals that affect thymine levels. In addition, there was also a concomitant up-regulation of aspartate transcarbamoylase, however, dihydroorotase and dihydroorotate dehydrogenase are not up-regulated in these cell lines.

Nectarin I is a novel, soluble germin-like protein expressed in the nectar of Nicotiana sp.

We have identified a limited number of proteins secreted into the nectar of tobacco plants. Nectarin I is the most highly expressed nectar protein and has a monomer molecular mass of 29 kDa. The other major nectar proteins are expressed at lower levels and have monomer molecular masses of 41, 54, and 65 kDa respectively. Nectarin I was purified and antiserum was raised against the protein. Under nondenaturing conditions, Nectarin I has an apparent molecular mass of > 120 kDa. The expression of Nectarin I was restricted to nectary tissues and to a much lower level in the ovary. No Nectarin I was found in petals, stems, leaves, or roots or other floral tissues. The expression of Nectarin I was also developmentally regulated. It is expressed in nectary tissues only while nectar is being actively secreted. Subsequently, the N-terminus of purified Nectarin I was sequenced. Sequence identity showed Nectarin I is related to wheat germin. Although hydrogen peroxide is readily detectable in tobacco floral nectar, we were unable to demonstrate any oxalate oxidase activity for Nectarin I. A partial cDNA encoding the mature Nectarin I N-terminus was isolated and used to probe a Nicotiana plumbaginifolia genomic library. The Nectarin I gene was isolated and the translated sequence was consistent with both N-terminal and internal cyanogen bromide-derived amino acid sequence. The gene contains a single 386 nt intron and encodes a mature protein of 197 amino acids.

Arabidopsis thaliana contains a large family of germin-like proteins: characterization of cDNA and genomic sequences encoding 12 unique family members.

We have identified 39 Arabidopsis thaliana ESTs encoding germin-like proteins (GLPs) and have completely sequenced 25 of these cDNAs. Our analysis demonstrates that the Arabidopsis genome contains a gene family with at least 12 GLP genes. Comparisons with other known germins and germin-like proteins indicate that these Arabidopsis GLP subfamilies are unique from wheat germin. All other known GLPs fall into one of these subfamilies. The translated GLPs show approximately 35% amino acid identity with other GLPs outside of their subfamily and significantly higher levels of identity within their respective subfamily. The 3' ends of many of the GLP cDNAs are heterogeneous and several sites of polyadenylation are used. Ten of the GLPs have N-terminal signal sequences and most appear to be exported from the cell. Structurally, the GLPs are predicted to have a high content of beta-pleated sheet. Seven conserved regions of beta-sheet were found in each of the GLP proteins along with alpha-helices located at both N- and C-termini. These same structural elements are also conserved in wheat germin. With one exception, all GLP family members contain at least one N-glycosylation site. All of these sites are conserved in an unstructured loop between beta-1 and beta-2. Genes for two of these GLPs were identified in genomic sequences previously deposited in the GenBank. The GLP3b gene is physically linked to the polyubiquitin 4 gene. The 3' end of the GLP3b mRNA is only 0.5 kb from the ubq4 start of transcription. Analysis of the GLP3b promoter shows the presence of a single putative auxin-response sequence located at -124 to -111 upstream from the 5' end of the GLP3b mRNA. The GLP9 gene was identified in an Arabidopsis contig from Chromosome 4.

Site-specific mutations of conserved residues in the phosphate-binding loop of the Arabidopsis UMP/CMP kinase alter ATP and UMP binding.

All eukaryotic UMP/CMP kinases contain a glycine-rich sequence GGPG(S/A)GK at the N-terminus. This sequence is homologous to the conserved sequence GXXGXGK found in other ATP-binding proteins. To study the role of this conserved sequence in Arabidopsis UMP/CMP kinase, five conserved residues were mutated by site-directed mutagenesis to generate seven mutant enzymes: G21A, G22A, G24A, G26A, K27R, K27M, and K27E. The G21A and G26A mutants were degraded during the purification phase and were thus unable to be purified. Kinetic studies on the other mutants, when compared to studies on the wild-type enzyme, revealed that this sequence is important for ATP binding and enzyme catalysis. All mutants had a decreased kcat/KATPm value. The G22A and G24A mutants had about half of the kcat value of wildtype and 3.9-fold and 3.3-fold increases in KATPm values, respectively. The kcat/KATPm values in the K27M and K27E mutants were changed significantly and decreased by 1000-fold and 2600-fold, respectively. The removal of the terminal positive charge of Lys27 in the K27M and K27E mutants resulted in 20% of the kcat value of wildtype. However, both mutants had a remarkable increase in KATPm value by 241-fold and 552-fold, respectively. Therefore, the positive charge of Lys27 plays an important role on both ATP binding and enzyme catalysis. Interestingly, the results also showed that the mutations that affected ATP binding also had an effect on UMP binding.

Cloning, expression in Escherichia coli, and characterization of Arabidopsis thaliana UMP/CMP kinase.

A cDNA encoding the Arabidopsis thaliana uridine 5'-monophosphate (UMP)/cytidine 5'-monophosphate (CMP) kinase was isolated by complementation of a Saccharomyces cerevisiae ura6 mutant. The deduced amino acid sequence of the plant UMP/CMP kinase has 50% identity with other eukaryotic UMP/CMP kinase proteins. The cDNA was subcloned into pGEX-4T-3 and expressed as a glutathione S-transferase fusion protein in Escherichia coli. Following proteolytic digestion, the plant UMP/CMP kinase was purified and analyzed for its structural and kinetic properties. The mass, N-terminal sequence, and total amino acid composition agreed with the sequence and composition predicted from the cDNA sequence. Kinetic analysis revealed that the UMP/CMP kinase preferentially uses ATP (Michaelis constant [Km] = 29 microM when UMP is the other substrate and Km = 292 microM when CMP is the other substrate) as a phosphate donor. However, both UMP (Km = 153 microM) and CMP (Km = 266 microM) were equally acceptable as the phosphate acceptor. The optimal pH for the enzyme is 6.5. P1, P5-di(adenosine-5') pentaphosphate was found to be a competitive inhibitor of both ATP and UMP.

Uridine 5'-Monophosphate Synthase Is Transcriptionally Regulated by Pyrimidine Levels in Nicotiana plumbaginifolia

To understand the regulation and expression of pyrimidine biosynthesis in plants, we have examined the effect of the metabolic inhibitor 5-fluoroorotic acid (FOA) on uridine-5'-monophosphate synthase (UMPSase) expression in cell cultures of Nicotiana plumbaginifolia. UMPSase is the rate-limiting step of pyrimidine biosynthesis in plants. Addition of FOA causes an up-regulation of UMPSase enzyme activity in cell cultures after a lag phase of several days. Western-blot analysis demonstrated that the up-regulation in enzyme activity was caused by increased expression of the UMPSase protein. Northern-blot analysis demonstrated a higher level of UMPSase mRNA in the FOA-induced tissues than in control tissues. Run-on transcriptional assays showed that the UMPSase gene was transcriptionally activated after FOA treatment. The mechanism of toxicity of FOA is through thymine starvation. We found that addition of thymine abrogated the FOA-mediated up-regulation of UMPSase. In addition, methotrexate and aminopterin, which affect thymine levels by inhibiting dihydrofolate reductase, also up-regulate UMPSase in N. plumbaginifolia cells.

Structural characterization of the active site of Brucella abortus Cu-Zn superoxide dismutase: a 15N and 1H NMR investigation.

Prokaryotic Cu-Zn superoxide dismutases (SODs) are rare and poorly characterized compared to their eukaryotic counterparts. To better characterize the structure of the prokaryotic enzyme, an NMR investigation of Brucella abortus Cu-Zn SOD in the reduced form was undertaken. The enzyme studied was a recombinant form, expressed in Escherichia coli. The enzyme initially lacked a full complement of Cu and Zn ion. After demetallation and remetallation with a stoichiometric amount of Cu and Zn ion, the specific activity of the recombinant B. abortus Cu-Zn SOD was comparable to the specific activity of the bovine enzyme. The 15N and 1H resonances of seven active site histidine imidazole rings were assigned using two-dimensional NMR methods. A self-consistent set of nuclear Overhauser effects between imidazole ring protons was observed, which was in agreement with the predictions of a model based on the X-ray crystallographic structure of the oxidized bovine enzyme (Tainer, J.A., Getzoff, E. D., Beem, K. M., Richardson, J.S., & Richardson, D.C. (1982) J. Mol. Biol. 160, 181-217). These observations strongly suggest that the structure of the active site of the prokaryotic enzyme is similar to that of the eukaryotic enzyme. Differences in the observed and predicted nuclear Overhauser effects could be ascribed to differences in the oxidation state of the Cu ion (Cu(I) in the reduced B. abortus enzyme and Cu(II) in the oxidized bovine enzyme), as much as they could to the different origins of the enzymes. The NMR data were also compared to a similar 1H NMR study of the human enzyme (Bertini, I., Capozzi, F., Luchinat, C., Piccioli, M., & Viezzoli, M. S. (1991) Eur. J. Biochem. 197, 691-697). The pattern of nuclear Overhauser effects and the chemical shifts of corresponding resonances were very similar in 1H NMR spectra of the human and B. abortus enzymes. Significant differences in the chemical shifts or exchange behavior of a few resonances indicated differences in the environments of several histidines in the active sites of reduced B. abortus and human Cu-Zn SODs. This is consistent with the presence of a number of insertions and deletions in the loop regions that make up the active site as indicated by amino acid sequence alignment studies. The tautomeric and protonation states of the active site histidines were also determined in this study, and the results were in agreement with previous studies. The resonances of nitrogen atoms coordinated to metal ions were found to fall between those of protonated and unprotonated nitrogens on histidine imidazoles.

Systemic induction of a potato pin2 promoter by wounding, methyl jasmonate, and abscisic acid in transgenic rice plants.

To address the question whether common signal(s) and transduction pathways are used to mediate a systemic wound response in monocot and dicot plants, a fusion of the potato proteinase inhibitor II gene (pin2) promoter and the bacterial beta-glucuronidase gene (Gus)-coding region was introduced into rice. In transgenic rice plants, the expression of the pin2-Gus fusion gene displays a systemic wound response, although the expression level is relatively low. Incorporation of the first intron from the rice actin 1 gene (Act1) into the 5'-untranslated region of the pin2-Gus construct results in high-level, systemically wound-inducible expression of the modified construct in transgenic rice plants. Histochemical analysis shows that this high-level, wound-inducible expression is associated with the vascular tissue in both leaves and roots. Furthermore, the expression of the pin2-Act1 intron-Gus fusion gene in transgenic rice plants can be systemically induced by both methyl jasmonate (MJ) and the phytohormone abscisic acid (ABA). These results suggest that the signal(s) mediating the observed systemic wound response and certain steps of the transduction pathways are conserved between dicot and monocot plants. Transient expression assays show that the pin2-Act1 intron-Gus construct is also actively expressed in transformed cells and tissues of several other monocot plants. Thus, the wound-inducible pin2 promoter in combination with the rice Act1 intron 1 might be used as an efficient regulator for foreign gene expression in transgenic monocot plants.

Construction of a UMP synthase expression cassette from Dictyostelium discoideum.

We have prepared a DNA cassette containing the UMP synthase (UMPS)-encoding gene (PYR5-6) from Dictyostelium discoideum. This gene contains no introns and can be used for expression of the UMPS protein. Due to the high percentage of AT in the flanking regions, useful restriction sites were absent, therefore the PYR5-6 was subcloned as three separate parts, manipulated, and religated to make a full-length clone. After reconstructing the coding region, we examined its functionality by introducing this gene under the control of the yeast GAL1 promoter into several uracil-requiring mutants of Saccharomyces cerevisiae. These studies demonstrated that the reconstructed PYR5-6 gene was functional and could complement independent ura3 and ura5 mutations in yeast.

Isolation and Characterization of UMP Synthase Mutants from Haploid Cell Suspensions of Nicotiana tabacum.

Uridine 5'-monophosphate (UMP) synthase mutants of tobacco have been produced from haploid cell-suspension cultures of a transgenic Nicotiana tabacum line, Tr25. The mutants were induced by incubating the suspension-cultured cells with 1 mmN-nitroso-N-methylurea for either 5 or 12 hours. Twenty mutant calli were isolated on selection medium containing 20 milligrams per liter of 5-fluoroorotic acid. Of those tested, most had reduced regeneration capacity. Characterization of UMP synthase activities in the isolated calli showed that UMP synthase activity varied from 8 to nearly 100% of the wild-type activity. The growth of the calli on the media containing different levels of 5-fluoroorotic acid correlated with decreasing UMP synthase activity. Because the UMP synthase enzyme has two separate enzymic activities (orotate phosphoribosyl transferase and orotidine-5'-monophosphate decarboxylase), several mutants were further characterized to determine how the mutations affected each of the two enzymic activities. In each case, the enzymic activity affected was the orotate phosphoribosyl transferase and not the orotidine-5'-monophosphate decarboxylase. The wound-inducible phenotype of the Tr25 plants as measured by the activation of the pin2-CAT gene remained unchanged by introduction of the UMP synthase mutations.

Wounding Nicotiana tabacum Leaves Causes a Decline in Endogenous Indole-3-Acetic Acid.

We have previously observed that auxin can act as a repressor of the wound-inducible activation of a chimeric potato proteinase inhibitor II-CAT chimeric gene (pin2-CAT) in transgenic tobacco (Nicotiana tobacum) callus and in whole plants. Therefore, this study was designed to examine endogenous levels of indole-3-acetic acid (IAA) in plant tissues both before and after wounding. Endogenous IAA was measured in whole plant tissues by gas chromatography-mass spectrometry using an isotope dilution technique. (13)C-Labeled IAA was used as an internal standard. The endogenous levels of IAA declined two- to threefold within 6 hours after a wound. The kinetics of auxin decline are consistent with the kinetics of activation of the pin2-CAT construction in the foliage of transgenic tobacco.

A yeast mitochondrial leader peptide functions in vivo as a dual targeting signal for both chloroplasts and mitochondria.

A fusion protein was expressed in transgenic tobacco and yeast cells to examine the functional conservation of mechanisms for importing precursor proteins from the cytosol into mitochondria and chloroplasts. The test protein consisted of the mitochondrial leader peptide from the yeast precursor to cytochrome oxidase subunit Va (prC5) fused to the reporter protein chloramphenicol acetyltransferase. This protein, denoted prC5/CAT, was transported into the mitochondrial interior in yeast and tobacco cells. In both organisms, the mitochondrial form of prC5/CAT was smaller than the primary translation product, suggesting that proteolytic processing occurred during the transport process. prC5/CAT also was translocated into chloroplasts in vivo, accumulating to approximately the same levels as in plant mitochondria. However, accumulation of prC5/CAT in chloroplasts relative to mitochondria varied with the conditions under which plants were grown. The chloroplast form of prC5/CAT also appeared to have been proteolytically processed, yielding a mature protein of the same apparent size as that seen in mitochondria of either tobacco or yeast. Chloramphenicol acetyltransferase lacking a mitochondrial targeting peptide did not associate with either chloroplasts or mitochondria. The results demonstrated that in plant cells a single leader peptide can interact functionally with the protein translocation systems of both chloroplasts and mitochondria, and raised the possibility that certain native proteins might be shared between these two organelles.

Chloramphenicol Acetyl Transferase (CAT) Protein Is Expressed in Transgenic Tobacco in Field Tests following Attack by Insects.

The expression of chloramphenical acetyl transferase (CAT) protein driven by the wound-inducible promoter from the proteinase inhibitor II K (pin2) gene was examined in whole tobacco (Nicotiana tabacum L.) plants under field conditions. Mechanical wounding of the field-grown leaves caused an accumulation of CAT protein in these leaves which begins several hours after wounding and continues to accumulate for about 36 hours. When sections of leaves were assayed for accumulation of CAT protein following wounding, the CAT protein was found to accumulate in the apical portions of the leaves. When endogenous insects attacked the leaves of transgenic plants grown in the field, the plants responded by inducing CAT protein. The mesophyll cells of the leaf were the site of expression of the CAT protein rather than the mid-vein or major veins within the leaf blade, indicating that the wound-inducible pin2 promoter specifically directs the synthesis of novel genes in tissues preferentially consumed by larval insects.

Auxin Levels Regulate the Expression of a Wound-Inducible Proteinase Inhibitor II-Chloramphenicol Acetyl Transferase Gene Fusion in Vitro and in Vivo.

Proteinase inhibitor genes are expressed in solanaceous and leguminous plants following wounding of the foliage by mechanical methods. Previous studies have shown that a cloned proteinase inhibitor II-chloramphenicol acetyl transferase (pin2-CAT) chimeric gene is regulated in a wound-inducible manner in transgenic plants. In this study, we analyzed transgenic plant tissues for expression of the pin2-CAT gene in response to various plant hormones. We found that CAT activity was induced in tobacco (Nicotiana tabacum) callus incubated in the absence of any plant growth regulators. Addition of growth regulators to the medium thus permitted us to measure the effects of these substances on the activity of the pin2-CAT gene construction. Cytokinin (BAP) and ethylene (ethophon) even at low concentrations stimulated the expression of CAT activity by 25 to 50%. Abscisic acid at concentrations up to 4.4 x 10(-5) molar had no effect upon CAT activity, but increasing auxin (naphthalene acetic acid) levels completely inhibited the synthesis of CAT protein. Gibberellic acid had little effect except at very high concentration (2.9 x 10(5) molar). The kinetics of activation of the pin2-CAT gene were quite long (5 to 7 days) when unwounded calli were plated on media lacking auxin. This effect was documented for calli derived from several transformed plants, containing the full, chimeric pin2-CAT (pRT45) gene. In addition, calli from tissues transformed with wild-type vectors or from several plants transformed with pRT50 (a noninducible derivative of pRT45) were not induced by plating on media lacking auxin. Other naturally occurring and synthetic auxins had similar effects to naphthalene acetic acid in inhibiting the induction of the chimeric gene fusion. Finally, leaf discs from transformed plants were induced by incubation in MS liquid medium in the presence and absence of naphthalene acetic acid. NAA was also effective in down regulating the chimeric gene in whole plant tissues.

Functional analysis of the 3' control region of the potato wound-inducible proteinase inhibitor II gene.

Proteinase inhibitor genes are expressed strongly in specific plant tissues under both developmental and environmental regulation. We have studied the role of the 3' control region of the potato proteinase inhibitor II gene (PI-II) that is inducible in leaves in response to herbivore attacks or other severe wounding. Comparison of the terminator from the PI-II gene with two different terminators from the 6b and 7 genes, driven by a common PI-II promoter-cat fusion molecule, indicated that the PI-II terminator provided the most efficient expression of cat. The PI-II terminator also caused a significantly elevated cat gene expression driven by the cauliflower mosaic virus 35S promoter. The increase in the level of expression is probably not due to the presence of an enhancer element in the PI-II terminator region, but to cis-acting elements involved in mRNA processing or stability. Both transient and stable transformation analyses of the deletion mutants in the 3'-flanking sequence indicated that about a 100-base pair DNA fragment surrounding the polyadenylation site is essential for the efficient gene expression. This region seems to consist of several regulatory elements, including the conserved sequence, CGTGTCTT, which is located 9 bases downstream from the polyadenylation site. The elements appear to contribute to the increased stability of mRNAs containing the PI-II terminator.

Wound-inducible expression of a potato inhibitor II-chloramphenicol acetyltransferase gene fusion in transgenic tobacco plants.

A potato inhibitor II gene (IIK) was isolated from a library of potato genes in lambda bacteriophage. An 8-kilobase-pair (kbp) insert was identified using a tomato inhibitor II cDNA as a hybridization probe, and a 2.6-kbp fragment containing the gene was subcloned into the plasmid pUC13 and characterized. The nucleotide sequence of the isolated gene exhibited 87% identity with the wound-inducible tomato inhibitor II cDNA sequence. The amino acid sequence of inhibitor IIK, deduced from the potato gene, exhibited 84% identity with the tomato inhibitor II protein. A 1000-bp restriction fragment from the 5' flanking region of the gene was fused to the open reading frame of the chloramphenicol acetyltransferase (CAT) gene. This fusion was terminated in two ways: (i) with a terminator sequence from the potato inhibitor II gene and (ii) with a terminator from the 6b gene of Ti plasmid pTiA6. These chimeric genes were transferred into tobacco cells via a binary Ti vector system, and transgenic plants were regenerated. The CAT gene was expressed in leaves of transformed plants in response to wounding when fused with the inhibitor IIK promoter and terminator regions. The chimeric gene containing the 6b terminator did not express CAT in response to wounding. The wound-inducible expression of CAT activity was systemic and was induced in tissues distal to the wounded tissues. The time course of wound induction of CAT activity in transgenic tobacco leaves is similar to that found for wound-inducible inhibitor I and II mRNAs in tomato leaves. These results demonstrate that sequences necessary and sufficient for wound inducibility are present within approximately 1000 bp of the control regions of the inhibitor IIK genes and that wound-inducible components of tobacco leaf cells can regulate these sequences.

Molecular characterization of a wound-inducible inhibitor I gene from potato and the processing of its mRNA and protein.

Genomic blotting of restriction fragments of Russet Burbank DNA indicated that at least 6 copies of Inhibitor I are present in the tetraploid potato genome. A library of potato genes in bacteriophage λ was screened for the presence of Inhibitor I genes using a wound-inducible tomato Inhibitor I cDNA as a hybridization probe. One phage with an insert of 13.1 kb was isolated that hybridized most strongly with the probe. A 4.2 kb Eco RI fragment containing the gene was isolated from the clone and 2.2 kb region was sequenced that included about 800 bp of both the 5' and 3' regions. The gene contained two introns of 479 and 417 bp respectively, and the splice junctions were typical of other eukaryotic genes. Putative TATAA and CAAT boxes were identified. The nucleotide sequence, when compared with a wound-inducible tomato Inhibitor I cDNA, exhibited over 90% identity. The gene codes for a prepro-Inhibitor I protein of 96 amino acids. The putative pre-sequence of 19 amino acids, differs in only one residue from that of tomato Inhibitor I. The potato pro-sequence, however, is lacking a tetrapeptide that is found in the tomato pro-sequence in the region of pro-peptide processing. This deletion, together with a substitution of a Gln for a Leu (4 residues toward the N terminus) provides an explanation for the differences at the N-termini between tomato and potato Inhibitor I natural proteins by providing different processing sites in the two pro-inhibitors. Thus, amino acid sequence differences between the N termini of tomato and potato Inhibitor I are easily explained by the mutational events. The different proposed pro-processing sites of the tomato and potato inhibitors suggest that a processing protease may be present in the vacuole with a specificity for Asn-X and Gln-X bonds.