Role of Enzymes in Plant Disease Development!
Plasma Membrane bound Proteases in Roots:
Plant proteases participate in different physiological processes. By unlimited proteolysis, amino acids are recycled for protein synthesis. Defective proteins are removed and metabolism is regulated by inactivation and degradation of definite proteins. By limited proteolysis, proteases regulate activity of enzymes and presence of structural proteins in a more specific and irreversible manner, whereby precursors are converted in their active forms.
So far, proteases bound to plasma membranes have been not described in plants. Previously, we could demonstrate the existence of various plasma membrane associated proteases in tobacco roots with much higher specific activity than detected in plasma membrane vesicles prepared from leaves (diploma thesis by Manuela Eick).
Based on these facts, the current work aims primarily for the purification of root-specific plasma membrane- bound proteases. Our main concern is with proteases, whose activity or gene expression might be influenced by external nitrate.
Pectolytic Enzymes:
Pectolytic enzymes are classified according to their catalytic activity and this is complicated by the fact that the substrate is heterogeneous. There are two mechanisms by which the linear chains of pectate or pectin are broken, hydrolysis and b -elimination, the latter giving rise to oligomers which are 4,5 unsaturated at the non-reducing end.
For pectate and pectin the hydrolytic enzymes are referred to as polygalacturonases and pectinases, respectively whereas for the enzymes that cleaves the b-1, 4 bonds by b elimination the terms are pectate lyases and pectin lyases, respectively.
These names are often prefixed by exo- or endo- according to whether the enzymes attack the polymer chains at the ends giving rise to dimers or in the middle, producing mixtures of oligomers. Pectin is de-esterified by the action of pectin methyl esterase.
Studies of two pectin lyases from Aspergillus niger have shown that they share only 17 per cent sequence homology with pectate lyases but, despite this, some of their structural features, such as amino acid stacks and the asparagine ladder are remarkably similar.
However, the two enzymes diverge markedly in their substrate-binding clefts and by the fact that calcium is required for pectate lyase activity but not for pectin lyase activity. In addition and perhaps because of their enzymic activities, endopolygalacturonases, endo- pectate lyases and endo-pectin lyases kill cells.
The simplest explanation for this is that the osmotically sensitive protoplast is rendered more vulnerable by the digestion of the enzymes’ substrates.
Pectic enzymes are regulated by diverse factors. These include the presence of pectin, growth phase, catabolite repression, plant extracts, temperature, anaerobiosis, iron limitation and nitrogen starvation.
For example – in soft-rot Erwinias, pectate lyase production is increased by plant cell walls, polygalacturonate and oligogalacturonate but repressed by glucose and 4,5 unsaturated digalacturonate concentrations that support bacterial growth.
Evidence has also been obtained in Erwinia chrysanthemi for the role of a cyclic AMP receptor protein in activation of some pectinolysis genes but repression of one of these, the pelA gene, important for pathogenicity.
Scott-Craig and co-workers (1990) found that a gene from the maize pathogen, Cochliobolus carbonum, encoding an endopolygalacturonase was expressed when the fungus was grown on pectin as carbon source but not on sucrose.
In contrast, Leone (1990) found that 11 isolates of Botrytis cinerea required phosphate for production of the polygalacturonase in vitro and gave rise to spreading lesions on French bean and tomato leaves when phosphate and glucose were included with the inoculum droplets.
The induction of pectic enzymes during cell-wall penetration by bacteria is thought to involve the following sequence of events: in response to a basal level of secretion of pectic enzymes acting on the cell walls of the host, oligogalacturonate (particularly dimers) are produced.
The oligogalacturonates are taken up by the cell and metabolized via 5-keto-4- deoxyuronate (DKI), 2, 5-diketo-3-deoxgluconate (DKII) and 2-keto-3-deoxygluconate (KDG) to pyruvate and 3-phosphoglyceraldehyde.
In Erwinia chrysanthemi, DKI, DKII and KDG interact with the repressor of a regulator gene, kdgR which controls the expression of 11 genes involved in pectinolysis. Thus the presence of DKI, DKII and KDG and their interaction with this repressor explains the coordinated expression of these genes.
1. Polygalacturonases:
Agrobacterium vitis was formerly classified as biovar 3 of A. tumefaciens. Unlike most strains of A. tumefaciens, which have wide host ranges, A. vitis is specific for grapevines. Both tumorigenic and non-tumorigenic strains produce polygalacturonase in culture showing that the gene is chromosomal rather than plasmid borne and the enzyme has been isolated from necrotic lesions on grape roots.
A strain of the bacterium carrying a Tn5 insertion in the polygalacturonase gene pehA multiplied at a reduced rate produced smaller tumours and did not form necrotic lesions. An explanation of the effect on tumorigenicity may lie in the release of galacturonic acid by the enzyme since this monomer enhances the expression of virulence genes (vir genes) in the related A. tumefaciens. These data strongly implicate the polygalacturonase as a virulence factor.
Flego and co-workers (1997) have shown that the pehA polygalacturonase gene of another plant pathogenic bacterium, Erwinia carotovora, is modulated by calcium. Western analysis showed that production of the enzyme was repressed in tobacco plants grown in the presence of 10mM or 30mM CaCl2 and that such plant were more resistant to the bacterium.
Moreover, when pehA expression was put under the control of a calcium insensitive promoter the resistance engendered by elevated calcium levels was lost.
Another aspect of these experiments is that pectate lyases of the same organism are promoted by calcium. The authors explain this paradox by suggesting that polygalacturonase is required early in infection but that once invasion has occurred, its action releases calcium from the plant cell walls. This would repress pehA expression but stimulate the activity of pectate lyases which may be required for the next phase of host colonization.
Cryphonectria parasitica, the fungus that destroyed the American chestnut trees, produced polygalacturonase in culture and caused browning of the inner bark of the tree. A hypovirulent strain produced less polygalacturonase in vivo and smaller cankers on American chestnut than the wild type. Further evidence implicating the polygalacturonase as a virulence factor came from comparisons of the susceptible American chestnut and the more resistant Chinese chestnut.
The latter developed smaller cankers and lower levels of polygalacturonase were detected in them than in the larger cankers on American chestnut. Possibly a polygalacturonase inhibiting protein was responsible for this difference since a protein extract from Chinese chestnuts was 15 times more inhibitory to the fungal polygalacturonase than a similar extract from American chestnuts.
Yao, Conway and Sams (1996) purified a polygalacturonase from rotted cortical tissue of apples infected with Penicillium expansum. Using degenerate primers based on the amino acid sequence of the protein, they amplified a DNA sequence from the fungus, which gave a predicted amino acid sequence that exactly matched that of the purified enzyme.
A cloned 212-bp PCR product hybridized with 1.5-kb RNA molecules extracted from apples rotted by P. expansum but no transcripts were detected in uninfected apples or in fungal mycelium grown on apple pectin as the sole carbon source. These results demonstrated that the polygalacturonase gene was expressed only in the invaded fruit.
Gene disruption experiments have provided more direct evidence of the role of polygalacturonase enzymes in virulence. Shieh and co-workers (1997) showed that disruption of the gene in Aspergillus flavus decreased invasion of cotton bolls by the fungus.
Similarly, ten Have and co-workers (1998) eliminated an endopolygalacturonase (Bcpg1) from Botrytis cinerea by transformation-mediated gene replacement. Although the mutants were still able to infect tomato leaves and fruits as well as apple fruits there was a significant decrease in growth of the lesions beyond the point of inoculation in the three host tissues.
Isshiki and coworkers (2001) demonstrated that a polygalacturonase, Acpgl, was probably the only polygalacturonase in Alternaria citri and that disruption by integration of an internal fragment of the gene reduced virulence for citrus and reduced its ability to macerate potato tissue.
In contrast disruption of the homologous gene in Alternaria alternata rough lemon pathotype which was 99.6 per cent similar in nucleotide sequence had no effect on pathogenicity, possibly owing to the fact that this organism produces a host selective toxin.
2. Pectate Lyases:
The danger of extrapolating results for enzyme activity obtained in vitro to those obtaining in the plant is dramatically demonstrated by work with the pectate lyases of Erwinia chrysanthemi.
A mutant of strain EC 16 was produced by directed deletions or insertions in four pectate lyase genes as well as an exopolygalacturonate lyase and an exo-poly-a-D- polygacturonosidase (products of these last two enzymes are dimers rather than mixtures of oligomers).
The mutant did not cause pitting of semi-solid pectate agar medium, a standard test of pectolytic activity in bacteria, but still macerated leaves of chrysanthemum, although less actively than the parent strain.
This result was explained by the finding that the mutant produced a second set of enzymes in planta and also in minimal media containing chrysanthemum extracts or cell walls as the sole carbon source but not in minimal media containing pectate. Sterile preparations of the enzymes macerated chrysanthemum leaf tissue.
Greater success has attended experiments designed to assess the role of pectate lyase in the fungus, Colletotrichum gloeosporioides. Yakoby and co-workers (2001) disrupted the pe1B gene by homologous recombination and found that decay of avocado fruits was reduced by 36-45 per cent.
Also, Rogers and coworkers (2000) showed that disruption of both pelA and pelD of Nectria haematococca drastically reduced virulence of the fungus for pea epicotyls but that disruption of one or other of the genes had no effect. Virulence of the double mutant could be restored by complementation with the pelD gene or addition of either of the purified enzymes PLA or PLD.
3. Cellulases and Xylanases:
Both cellulose and arabinoxylan are depolymerized by b-1, 4-glycanases, usually termed cellulases and xylanases, respectively. Genes encoding these enzymes have conserved regions, which are common to both and occur in discrete domains.
It is probable that both types of enzyme arose from progenitor sequences by gene duplication, mutation and domain shuffling. Some show mixed specificity, hydrolysing not only the b-1, 4 bonds of cellulose but also those of xylan, chitin and related substrates.
Before the cellulose chains that make up microfibrils can be depolymerized, the microfibrils must be rendered amorphous and it is thought that this is achieved by free- radical attack. After amorphogenesis the cellulose chains are readily depolymerized by cellulases which may attack the b-1, 4 linkages within the chains, endocellulases, or at the ends, exocellulases. These normally yield the dimer, cellobiose, which in turn is split by exocellobiohydrolases and b-glucosidases.
Walker, Reeves and Salmond (1994) obtained evidence for the role of a cellulase in the pathogenicity Erwinia carotovora subsp. carotovora for potato. Using ethylmethylsulfonate (EMS) as a mutagen, they obtained a mutant which was incapable of degrading carboxymethylcellulose and which was significantly impaired in its ability to macerate potato tissue.
Western analysis with polyclonal antibodies against the cellulase, CelV, showed that neither the enzyme nor a truncated version of it was produced by the mutant. Complementation of the mutant by cosmids containing celV restored synthesis and secretion of the enzyme as well as the ability to macerate potato tuber tissue.
A similar, clear-cut story for the role of cellulases in fungal infections of plants does not seem to have been established. Sposat, Ahn and Walton (1995) disrupted the gene CEL1 in the maize pathogen, Cochliobolus carbonuni but pathogenicity was not affected, presumably owing to the ability of the fungus to secrete other cellulases and, indeed, these were detected in culture filtrates of the fungus grown on cellulose or cell walls of maize.
Since arabinoxylans are such a dominant feature of the cell walls of monocotyledonous plants, it is not surprising that production of xylanases is a prominent feature of their pathogens.
For example – Septoria nodorum, a pathogen of wheat produces more xylanase-degrading enzymes than pectinases and strains of Erwinia chrysanthemi which infect maize secrete more xylanase activity than those that are pathogens of dicotyledonous plants.
Once again, it has been difficult to prove their involvement in pathogenicity or virulence. Wu, S.C. and co-workers (1997) deleted two xylanase genes of the rice blast fungus, Magnaporthe grisea.
The double mutant only accumulated half the mycelial mass of the wild type when grown on rice cell walls or xylan as sole carbon source but strains carrying mutations at one or other or both loci were as virulent as the parent.
Analysis of the culture filtrate of the double mutant revealed four further xylanases. As with enzymes that depolymerize pectin, xylanases are also capable of killing plant cells.
4. Enzymes Involved in the Degradation of Lignin:
The complexity of lignin has hampered elucidation of its degradation. Among the best- studied systems are those from two white rot fungi, Phanerochaete chrysosporium and Trametes versicolour.
Both produce lignin peroxidase, a heme peroxidase which causes the oxidative depolymerization of lignin, but other components are also important such as manganese-dependent peroxidases, laccases, H2O2, generating enzymes, veratryl alcohol, lignin and manganese.
The regulation of enzymes that degrade lignin and their role in pathogenicity does not seem to have attracted a great deal of attention. Work in vitro has shown that the heme peroxidase of Phanerochaete chrysosporium is triggered by nitrogen starvation and lignolytic activity is enhanced by the lignin degradation product, veratryl alcohol.
More recently, tryptophan was found to be highly stimulatory to lignin peroxidase activity in cultures of four white rot fungi. Dutton and Evans (1996) have suggested that in white rot fungi, oxalate acts as a potential electron donor for lignin-peroxidase reduction and chelates manganese. It is proposed that the latter activity would allow dissolution of the manganese cation from a manganese-enzyme complex causing the stimulation of extracellular manganese peroxidase activity.
Schultz and Nicholas (2000) have pointed out that a satisfactory explanation as to why angiosperm sapwood but not heartwood or gymnosperm wood is usually degraded by white- rot fungi has never been given.
They suggest that the phenolics present in lignin and heartwood extractives act as free radical scavengers and may retard white-rot fungi, whereas these fungi may rapidly colonize angiosperm sapwood as it has a relatively low free-phenolic content.
5. Proteases:
Despite the fact that proteins may make up to 15 per cent of the cell wall of some plants, proteases produced by pathogens and their role as pathogenicity or virulence factors has received little attention. Movahedi and Heale (1990a, b) found an aspartic protease secreted by Botrytis cinerea both in culture and in infected carrots.
When spores of B. cinerea were treated with pepstatin, a specific inhibitor of aspartic protease, there was a marked reduction in symptoms, not only when carrot was challenged but also strawberry, raspberry, cabbage and broad bean. Pagel and Heitefuss (1990) showed that several degradative enzymes appeared sequentially in potato tubers infected by Erwinia carotovora subsp. atroseptica including a protease which was detected 19 h after inoculation.
Dow and co-workers (1990) demonstrated that Xanthomonas campestris pv. campestris produced two proteases in culture, a serine protease and a zinc- requiring protease and these accounted for almost all the proteolytic activity of the wild-type organism.
A mutant that lacked both proteases was less virulent than the wild type when introduced into the cut vein endings of turnip leaves. Ball and co-workers (1991) obtained genetic evidence for the requirement of an extracellular protease in the pathogenicity of the fungus Pyrenopeziza brassicae for oilseed rape.
An ultraviolet-induced mutant which was non-pathogenic and also deficient in extracellular protease production was transformed with clones from a genomic library of P. brassicae. Both pathogenicity and protease activity were restored by a transformant with a single cosmid insert.
More recently, Carlile and co-workers (2000) have reviewed the production of proteases by plant-pathogenic fungi and bacteria and have reported a cell-wall degrading trypsin, SNP1, produced during infection by Stagnospora nodorum (¼ Septoria nodorum).
In this thorough piece of work, these workers showed that the fungus produced a protease when grown on wheat cell walls as the sole source of carbon and nitrogen, purified it and demonstrated that the pure protein degraded wheat cell walls, releasing hydroxyproline- rich proteins. This activity was inhibited by the trypsin inhibitors, aprotinin and leupeptin.
Sequencing of the N-terminal region of the protein showed that it shared 89.5 per cent identity with a trypsin protease from Fusarium oxysporum and 84.2 per cent identity with ALP1 from Cochliobolus carbonum.
The presence of the protease in planta was demonstrated by the presence of protease activity which co-eluted from a cation exchange resin with SNP1, by expression of the encoding gene in Northern blots and by the expression of green fluorescent protein when this was fused to the SNP1 promoter. It will be interesting to see what effect the deletion of this gene, as intended by the authors, has on the virulence of the fungus.
6. Membranlytic Enzymes:
In animal pathology, enzymes that catabolize lecithin are recognized as toxins but there have been few comparable studies in plant pathology. Tariq and Jeffries (1987), on the basis of cytochemistry, invoked the presence of lipolytic enzymes in the penetration of bean leaf tissues by Sclerotinia sclerotiorum.
However, more attention has been paid to the alteration of membranes in connection with plant defence as part of the hypersensitive response and to membrane dysfunction caused by toxins.