Search

Bacterial blight resistance gene B5 has received little attention since it was first described in 1950. A near-isogenic line (NIL) of Gossypium hirsutum cotton, Ac B5, was generated in an otherwise bacterial-blight-susceptible ‘Acala 44’ background. The introgressed locus B5 in Ac B5 conferred strong and broad-spectrum resistance to bacterial blight. Segregation patterns of test crosses under Oklahoma field conditions indicated that Ac B5 is likely homozygous for resistance at two loci with partial dominance gene action. In controlled-environment conditions, two of the four copies of B5 were required for effective resistance. Contrary to expectations of gene-for-gene theory, Ac B5 conferred high resistance toward isogenic strains of Xanthomonas citri subsp. malvacearum carrying cloned avirulence genes avrB4, avrb7, avrBIn, avrB101, and avrB102, respectively, and weaker resistance toward the strain carrying cloned avrb6. The hypothesis that each B gene, in the absence of a polygenic complex, triggers sesquiterpenoid phytoalexin production was tested by measurement of cadalene and lacinilene phytoalexins during resistant responses in five NILs carrying different B genes, four other lines carrying multiple resistance genes, as well as susceptible Ac44E. Phytoalexin production was an obvious, but variable, response in all nine resistant lines. Ac B5 accumulated an order of magnitude more of all four phytoalexins than any of the other resistant NILs. Its total levels were comparable to those detected in OK1.2, a highly resistant line that possesses several B genes in a polygenic background.

Near-isogenic lines of upland cotton (Gossypium hirsutum) carrying single, race-specific genes B4, BIn, and b7 for resistance to bacterial blight were used to develop a pyramid of lines with all possible combinations of two and three genes to learn whether the pyramid could achieve broad and high resistance approaching that of L. A. Brinkerhoff's exceptional line Im216. Isogenic strains of Xanthomonas axonopodis pv. malvacearum carrying single avirulence (avr) genes were used to identify plants carrying specific resistance (B) genes. Under field conditions in north-central Oklahoma, pyramid lines exhibited broader resistance to individual races and, consequently, higher resistance to a race mixture. It was predicted that lines carrying two or three B genes would also exhibit higher resistance to race 1, which possesses many avr genes. Although some enhancements were observed, they did not approach the level of resistance of Im216. In a growth chamber, bacterial populations attained by race 1 in and on leaves of the pyramid lines decreased significantly with increasing number of B genes in only one of four experiments. The older lines, Im216 and AcHR, exhibited considerably lower bacterial populations than any of the one-, two-, or three-B-gene lines. A spreading collapse of spray-inoculated AcBIn and AcBInb7 leaves appears to be a defense response (conditioned by BIn) that is out of control.

2,7-Dihydroxycadalene and lacinilene C, sesquiterpenoid phytoalexins that accumulate at infection sites during the hypersensitive resistant response of cotton foliage to Xanthomonas campestris pv. malvacearum, have light-dependent toxicity toward host cells, as well as toward the bacterial pathogen. Adaxial epidermal cells surrounding and sometimes covering infection sites turn red. The red cells exhibited 3-4-fold higher absorption at the photoactivating wavelengths of sunlight than nearby colorless epidermal cells. Red epidermal cells protected underlying palisade mesophyll cells from the toxic effects of 2,7-dihydroxycadalene plus sunlight, indicating a role for epidermal pigments in protecting living cells that surround infection sites from toxic effects of the plant's own phytoalexins. A semi-quantitative survey of UV-absorbing substances extracted from epidermal strips from inoculated and mock-inoculated cotyledons indicated that the principal increase in capacity to absorb the photoactivating wavelengths was due to a red anthocyanin and a yellow flavonol, which were identified as cyanidin-3-O-beta-glucoside and quercetin-3-O-beta-glucoside, respectively.

Approximately 185,000GossypiumEST sequences comprising >94,800,000 nucleotides were amassed from 30 cDNA libraries constructed from a variety of tissues and organs under a range of conditions, including drought stress and pathogen challenges. These libraries were derived from allopolyploid cotton (Gossypium hirsutum; ATand DTgenomes) as well as its two diploid progenitors,Gossypium arboreum(A genome) andGossypium raimondii(D genome). ESTs were assembled using the Program for Assembling and Viewing ESTs (PAVE), resulting in 22,030 contigs and 29,077 singletons (51,107 unigenes). Further comparisons among the singletons and contigs led to recognition of 33,665 exemplar sequences that represent a nonredundant set of putativeGossypiumgenes containing partial or full-length coding regions and usually one or two UTRs. The assembly, along with their UniProt BLASTX hits, GO annotation, and Pfam analysis results, are freely accessible as a public resource for cotton genomics. Because ESTs from diploid and allotetraploidGossypiumwere combined in a single assembly, we were in many cases able to bioinformatically distinguish duplicated genes in allotetraploid cotton and assign them to either the A or D genome. The assembly and associated information provide a framework for future investigation of cotton functional and evolutionary genomics.

In this first transcriptome study of the cotton–Xanthomonas campestris pv. malvacearum interaction, clones from a cDNA library were used to identify host genes expressed in upland cotton leaves following inoculation. cDNA was prepared from inoculated and non-inoculated leaves of cotton line Im216, which has exceptionally high, broad and stable resistance to bacterial blight, expressed as a hypersensitive-resistant response. Suppression subtractive hybridization was used to make a normalized cDNA library that is differential between transcripts of inoculated and non-inoculated leaves. More than 2000 clones were generated. They yielded 121 unique non-redundant sequences, consisting of 97 with similarity to sequences submitted to GenBank and 24 without good matches. PCR-amplified cDNAs of the 121 genes, as well as of 13 previously identified genes, were arrayed onto glass slides. These microarrays were used to analyze transcripts in Im216 leaves over the extended period of 8–60 h after inoculation. Microarray analysis indicated that 98% of the genes were significantly up-regulated at one or more of the sampling times. Of these, 63% had sequence similarity to plant genes associated with defense responses, i.e., to genes that function in disease/defense, protein synthesis/turnover, secondary metabolism, signaling, stress/programmed cell death, or code for pathogenesis-related proteins or retrotransposon-like proteins. Seventeen percent of the genes showed highest differential abundance during the first day after inoculation, prior to when the microscopic hypersensitive response is first seen. A majority had their highest differential expression at the latest time observed, 60 h, when healthy cells neighboring dying or dead cells are likely to be the major participants in the response.

Among near-isogenic upland cotton ( Gossypium hirsutum L.) lines, remarkable variation exists in the ability of the leaves to produce a red accessory pigment anthocyanin when challenged by selected isogenic races of the bacterial pathogen, Xanthomonas campestris pv. malvacearum (Smith) Dye. Inoculated areas that developed visible water-soaking that progressed beyond pinpoint-sized dots, a sign of susceptibility, very rarely produced anthocyanin, and on the couple of occasions in which anthocyanin was seen, it was borderline. Inoculated areas that exhibited resistance by developing little or no water-soaking produced various, but obvious, amounts of anthocyanin, with the exception of the cotton line whose only resistance gene is B In . These results indicate that anthocyanin production by cotton leaves in response to an unsuccessful challenge by the bacterium is a bacterial blight resistance response, but is not essential for resistance. The epidermis of the adaxial surface of the leaves that received direct illumination was the tissue involved in anthocyanin production. The subsidiary cells of the stomatal complex were the initial participants in anthocyanin accumulation. A protective role of anthocyanin-containing cells against damage by infection-related reactive oxygen species and light-activated phytoalexins to the healthy tissues surrounding infection sites is suggested.

In cotton, gossypol and related sesquiterpene aldehydes are present in the glands of aerial tissues and in epidermal cells of roots. A cytochrome P450 was found to be expressed in aerial tissues of glanded cotton cultivars, but not or at an extremely low level in the aerial tissues of a glandless cultivar. Its cDNA was then isolated from Gossypium arboreum L. After expression in Saccharomyces cerevisiae, the P450 was found to catalyse the hydroxylation of (+)-delta-cadinene, forming 8-hydroxy-(+)-delta-cadinene. This P450 mono-oxygenase has been classified as CYP706B1, and is the first member of the CYP706 family for which a function has been determined. Sesquiterpene aldehydes and CYP706B1 transcripts were detected in roots of both the glanded and glandless cultivars and in aerial tissues of the glanded cultivar. In suspension cultured cells of G. arboreum, elicitors prepared from the phytopathogenic fungus Verticillium dahliae caused a dramatic induction of CYP706B1 expression. The expression pattern of CYP706B1 and the position at which it hydroxylates (+)-delta-cadinene suggest that it catalyses an early step in gossypol biosynthesis. Southern blotting revealed a single copy of CYP706B1 in the genome of G. arboreum. CYP706B1 holds good potential for manipulation of gossypol levels in cottonseed via genetic engineering.

Gossypium (cotton) spp. produce an array of sesquiterpenoid defense compounds, some of which accumulate in pigmof dHG and HG.ent glands and in root epidermis of healthy plants, and others which function as phytoalexins. Of the phytoalexins which accumulate in stem stele of Verticillium wilt-resistant cotton in response to Verticillium dahliae infection, Mace, Stipanovic and Bell (1985) Physiological Plant Pathology, 26, 209, have shown that desoxyhemigossypol (dHG) has the highest antifungal activity. Of the phytoalexins previously observed in foliar tissue of bacterial blight-resistant Gossypoium hirsutum in response to Xanthomonas campestris pv. malvacearum (Xcm) infection [2,7-dihydroxycadalene (DHC), lacinilene C, lacinilene C 7-methyl ether, and 2-hydroxy-7-methoxycadalene], DHC has the highest antibacterial activity. Both groups of phytoalexins have cadinane carbon skeletons, but they differ in the positions of oxygen-containing functional groups. dHG and its oxidation product hemigossypol (HG) have now been identified as part of the foliar resistance response to Xcm. A time course study showed that the bacterial blight-resistant, pigment-glandless G. hirsutum line WbMgl accumulated dHG and HG more quickly than the cadalene and lacinilene phytoalexins and to similar peak amounts (1–5 μmol/g fr. wt). Bioassays on logarithmically growing cultures of Xcm in defined liquid medium in the dark revealed that both dHG and HG have phytoalexin activity toward this pathogen, but are less potent than DHC. Whether dHG and HG contribute to resistance toward the infection by Xcm or play a different role depends on where these phytoalexins accumulate in inoculated cotyledons, which has yet to be determined.

Average cellular phytoalexin concentrations at infection sites were determined inXanthomonas campestrispv.malvacearuminfected cotton leaves to test the hypothesis that sesquiterpenoid phytoalexins play a major role in heritable resistance of cotton to bacterial blight. Bacteriostasis was achieved in leaves of resistant cotton line WbM(0.0) by 5–7 days after inoculation. Between days 3 and 7, phytoalexin amounts per gram of tissue were above basal levels but were far below their peak values. Nevertheless, when the phytoalexin quantities present on day 7 were divided by the volume of water contained in the hypersensitively responding cells defining the scattered infection centres, the concentrations computed were ones previously found to be highly inhibitory inin vitrobioassays. In necrotic cells of susceptible line WbM(4.0), the values were

Time courses of multiplication of an incompatible strain of Xanthomonas campestris pv. malvacearum , accumulation of phytoalexins, and development of phytoalexin-rich, yellow-green fluorescent, hypersensitively necrotic cells were followed in cotton cotyledons. Occasionally, hypersensitively responding palisade cells were seen to pass through a stage in which the yellow-green fluorescence, due to the phytoalexins lacinilene C and its methyl ether, was confined to the cytoplasmic layer surrounding an intact vacuole, but more often, when cells became yellow-green fluorescent, the fluorescence appeared throughout the cell. The period of most rapid increase in fluorescent cell numbers, 48-72 h post-inoculation, was also the period of most rapid increase in phytoalexin content, and the rate of bacterial multiplication began to slow down at this time. Stable 13 C isotope incorporation from a 4 h pulse with [ 13 C 2 ]acetate into the phytoalexin 2,7-dihydroxycadalene and its methyl ether was measured by direct probe electron impact-mass spectrometry to monitor their biosynthesis at 2, 3, 4, and 6 days after inoculation. Highest incorporation rates occurred during the period of most rapid increase in numbers of fluorescent cells, consistent with the hypothesis that these cells achieve a burst of phytoalexin biosynthesis causing them to become fluorescent as they die. In addition, biosynthesis was still active when fluorescent cell numbers approached their maximum and was appreciably above zero 1 and 2 days later, which suggests the possibility that healthy cells neighboring the hypersensitively responding ones contribute to the biosynthesis of phytoalexins in cotton.

Two new sesquiterpenes, the cis- and trans-diastereomers of 7-hydroxycalamenen-2-one, as well as trans-7-hydroxycalamenene, have been identified in extracts from cotyledons of bacterial blight-resistant Gossypium hirsutum harvested during the period of sesquiterpenoid phytoalexin biosynthesis following inoculation with the bacterial pathogen Xanthomonas campestris pv. malvacearum. The cis- and trans-diastereomers were distinguished by NOE correlations predicted from molecular modelling calculations.

Glandless cotton cotyledons stimulated to produce sesquiterpenoid phytoalexins by inoculation with Xanthomonas campestris pv. malvacearum, or by injection of oligogalacturonide elicitors, generated a hydrocarbon that was absent in mock-inoculated or non-inoculated cotyledons. Enzyme preparations from the same cotton cotyledons catalysed cell-free reactions which converted (E, E)-[1-3H]farnesyl pyrophosphate into a predominant tritium-labelled hydrocarbon product. Large-scale cell-free reactions catalysed by enzyme preparations from cotton cotyledons previously inoculated with Xanthomonas campestris pv. malvacearum converted nonradioactive (E, E)-farnesyl pyrophosphate into the hydrocarbon product, which was identified as (+)-δ-cadinene by chiral GC-mass spectrometry. In planta incorporation of tritium into the sesquiterpenoid phytoalexins 2,7-dihydroxycadalene, lacinilene C, lacinilene C7-methyl ether and structurally related terpenoids occurred following injection of [3H] (+)-δ-cadinene into previously inoculated glandless cotton cotyledons. The accumulation of (+)-δ-cadinene in bacteria-inoculated or elicitor-treated cotton cotyledons and the results of the incorporation experiment suggest that (+)-δ-cadinene is an early enzymatic intermediate in the biosynthesis of the sesquiterpenoid phytoalexins by upland cotton.

Gossypium (cotton) spp. produce an array of sesquiterpenoid defense compounds, some of which accumulate in pigmof dHG and HG.ent glands and in root epidermis of healthy plants, and others which function as phytoalexins. Of the phytoalexins which accumulate in stem stele of Verticillium wilt-resistant cotton in response to Verticillium dahliae infection, Mace, Stipanovic and Bell (1985) Physiological Plant Pathology, 26, 209, have shown that desoxyhemigossypol (dHG) has the highest antifungal activity. Of the phytoalexins previously observed in foliar tissue of bacterial blight-resistant Gossypoium hirsutum in response to Xanthomonas campestris pv. malvacearum (Xcm) infection [2,7-dihydroxycadalene (DHC), lacinilene C, lacinilene C 7-methyl ether, and 2-hydroxy-7-methoxycadalene], DHC has the highest antibacterial activity. Both groups of phytoalexins have cadinane carbon skeletons, but they differ in the positions of oxygen-containing functional groups. dHG and its oxidation product hemigossypol (HG) have now been identified as part of the foliar resistance response to Xcm. A time course study showed that the bacterial blight-resistant, pigment-glandless G. hirsutum line WbMgl accumulated dHG and HG more quickly than the cadalene and lacinilene phytoalexins and to similar peak amounts (1–5 μmol/g fr. wt). Bioassays on logarithmically growing cultures of Xcm in defined liquid medium in the dark revealed that both dHG and HG have phytoalexin activity toward this pathogen, but are less potent than DHC. Whether dHG and HG contribute to resistance toward the infection by Xcm or play a different role depends on where these phytoalexins accumulate in inoculated cotyledons, which has yet to be determined.

Several lines of evidence indicate that (+)-δ-cadinene-8-hydroxylase (CYP706B1) plays an important role in biosynthesis of gossypol in Gossypium arboreum L. (Luo et al., 2001; Wang et al., 2003). The catalytically active enzyme has been expressed in yeast microsomes. Some microsomal preparations conjugated the hydroxylated (+)-δ-cadinene to a moiety that has not yet been identified. However, when microsomes were treated with n-octyl-β-d-glucoside (OG), a non-ionic detergent, (+)-δ-cadinene was reproducibly converted to the free alcohol, 8-hydroxy-(+)-δ-cadinene. OG had little effect on K(m) and slightly stimulated apparent V(max). Enzymic activity was more than 10-fold more sensitive to inhibition by the N-substituted imidazole clotrimazole than to miconazole. Sesquiterpene olefins (-)-δ-cadinene, (-)-α-cubebene, (-)-α-muurolene, α-humulene, and a mixture of (-)- and (+)-α-copaene were inhibitory to hydroxylation of (+)-δ-cadinene. In addition, (-)-α-cubebene, (-)-α-muurolene, α-humulene, and, to a smaller extent, (-)-δ-cadinene served as alternative substrates for (+)-δ-cadinene-8-hydroxylase and were converted to mono-hydroxylated products. Of the five olefins tested, α-humulene and α-copaene are found in lysigenous glands of cotton (Elzen et al., 1985), which are also the site of gossypol accumulation (Bell et al., 1978; Mace et al., 1976) and the probable site of its biosynthesis.

Average phytoalexin concentration in the hypersensitively necrotic cells of a leaf sample was computed by dividing the sample's phytoalexin content, determined by extraction and HPLC, by the water content of those mesophyll cells that exhibited yellow-green fluorescence and/or brown pigmentation. Mesophyll water content was determined by measuring total tissue water content and multiplying it by the fraction of the tissue water in the palisade and spongy mesophyll, determined by stereological analysis of cross sections. The percentage of the mesophyll cells that were yellow-green fluorescent and/or brown was determined by counting such cells, as well as total cells, in 24 microscope fields per sample. The error (coefficient of variation of the mean) in these determinations of phytoalexin concentration was estimated by propagation of error analysis to be ±20%. This strategy for determining cellular phytoalexin concentrations is applicable to other host/pathogen systems for which there is a means of determining which cells contain phytoalexin.

During the hypersensitive response of resistant cotton leaves to attempted bacterial infection and during the period of phytoalexin accumulation, mesophyll cells closest to bacterial colonies developed yellow-green autofluorescence and, later, brown pigmentation. Susceptible leaves, which accumulate less of the phytoalexins, developed fewer fluorescent cells. Filter sets designed for fluorescence-microscope observation of fluorescein are fortuitously selective for the autofluorescent cells. Videodensitometry showed that all such cells exhibited substantially less fluorescence after mild extraction. The two phytoalexins which are green fluorescent accounted for 75% of the green fluorescent material extracted from inoculated resistant leaves. It was concluded that all yellow-green fluorescent cells contain diffusible phytoalexins.

A technique was developed to inoculate uniformly and gently the internal phyllosphere from the upper surface of cotton leaves with the phytopathogenic bacterium Xanthomonas campestris pv. malvacearum. The inoculum consisted of 2 to 3 × 107 CFU/ml in CaCO3-saturated sterile distilled water containing 0.02%, vol/vol, of the wetting agent Silwet L-77. A custom-made inoculation apparatus was employed to immerse a circular area of the adaxial surface of a leaf in inoculum for 90 s. This resulted in uniform, passive entry of bacteria into the substomatal chambers, producing an endophytic bacterial population of 2 × 104 CFU/cm2. Microscopic signs of infection were visible 48 to 72 h after inoculation. In susceptible leaves, uniformly distributed water-soaked spots were observed 7 to 8 days after inoculation. When the technique was used on resistant leaves, the autofluorescence that is characteristic of hypersensitively necrotic cells developed in the guard cells and palisade cells lining substomatal chambers, but not in the underlying spongy mesophyll.

The development and genetic characterization of four near-isogenic lines (NILs) of cotton (Gossypium hirsutum) is described herein. Each line contains a single, but different, gene for resistance to bacterial blight caused by Xanthomonas campestris pv. malvacearum. The lines were derived using at least six backcrosses to the susceptible recurrent parent ‘Acala 44’, followed by single plant-progeny row selection for uniformity. The NILs are homozygous for the B2, B4, BIn, or b7 genes and are designated as AcB2, AcB4, AcBIn, and Acb7, respectively. In the ‘Acala 44’ background, B2, B4, and BIn are partially dominant genes; b7 is partially recessive. Relative strengths of resistance conferred by those genes toward race 1 of the pathogen were B4˜b7>BIn˜B2. B4, BIn, and b7 each conferred resistance toward X. campestris pv. malvacearum carrying a single avirulence gene, whereas B2 was less specific.

Two sesquiterpenoid naphthols, 2,7-dihydroxycadalene and 2-hydroxy-7-methoxycadalene, and their optically active oxidation products, lacinilene C and lacinilene C 7-methyl ether, have been isolated from cotton leaves and cotyledons after inoculation with incompatible bacteria. All four sesquiterpenoids accumulated to amounts one to two orders of magnitude higher in leaves and cotyledons infiltrated with incompatible bacteria than in those inoculated with compatible bacteria or sterile medium. In cotton plants inoculated withXanthomonas campestris pvmalvacearum in the field in August, sesquiterpenoid levels in incompatible leaves rose shortly before bacterial multiplication stopped, continued to rise for several days thereafter, and then fell. In leaves of a congenic compatible line, only the relatively insoluble, non-inhibitory 2-hydroxy-7-methoxycadalene accumulated to a level comparable to that in the incompatible line. However, all four sesquiterpenoids accumulated to moderate levels in unioculated leaves of both lines by 30 October. Water-solubilities and inhibitory activities towardX. campestris pvmalvacearum indicate that, with the exception of 2-hydroxy-7-methoxycadalene, these sesquiterpenoids can form aqueous solutions capable of inhibiting the bacterial pathogen.

In an effort to expand the Gossypium hirsutum L. (cotton) expressed sequence tag (EST) database, ESTs representing a variety of tissues and treatments were sequenced. Assembly of these sequences with ESTs already in the EST database (dbEST, GenBank) identified 9675 cotton sequences not present in GenBank. Statistical analysis of a subset of these ESTs identified genes likely differentially expressed in stems, cotyledons, and drought-stressed tissues. Annotation of the differentially expressed cDNAs tentatively identified genes involved in lignin metabolism, starch biosynthesis and stress response, consistent with pathways likely to be active in the tissues under investigation. Simple sequence repeats (SSRs) were identified among these ESTs, and an inexpensive method was developed to screen genomic DNA for the presence of these SSRs. At least 69 SSRs potentially useful in mapping were identified. Selected amplified SSRs were isolated and sequenced. The sequences corresponded to the EST containing the SSRs, confirming that these SSRs will potentially map the gene represented by the EST. The ESTs containing SSRs were annotated to help identify the genes that may be mapped using these markers.Key words: drought stress, gene annotation, gene mapping, tentative consensus sequence (TC), Xanthomonas campestris.

[(3)H](+)-delta-Cadinene and its 8-hydroxy derivative, prepared from (1RS)-[1-(3)H]FPP by the action of one and two recombinant enzymes, respectively, were infiltrated into cotyledons of bacterial blight-resistant cotton plants as they biosynthesized sesquiterpene phytoalexins in response to infection by Xanthomonas campestris pv. malvacearum. Following both treatments, tritium appeared in the HPLC fraction that contained hemigossypol. Hemigossypol was isolated from the cotyledons that had been treated with [(3)H](+)-8-hydroxy-delta-cadinene and was trimethylsilylated and purified. In two experiments, specific radioactivity of the hemigossypol derivative indicated that 5% and 10%, respectively, of the [(3)H](+)-8-hydroxy-delta-cadinene had been converted to hemigossypol.

Chemical degradation of 2,7-dihydroxycadalene and 2-hydroxy-7-methoxycadalene biosynthesized in cotton cotyledons from a mixture of [2- 14 C]mevalonolactone and [5- 3 H]mevalonolactone has revealed that formation of each of these cadalene sesquiterpenoids involves a hydrogen transfer to the methine carbon of the isopropyl side chain. The transfer may occur as a 1,3-hydride shift following cyclization of the farnesyl precursor to a 10-membered ring cation.

Publisher Summary This chapter illustrates how microscopy has been and may be applied to examine plant-bacterium interactions; emphasis is on the localization of responses in plant cells. Some important components of the plant's response are introduced with reference to well-studied examples in the chapter. The general methods described for examination of interactions by electron microscopy are applicable to most plant-pathogen combinations. Cellular Responses in Lettuce Challenged by P. syringae pv. phaseolicola : wild-type strains and hrp mutants' focuses on work which has allowed good definition of cell wall alterations occurring during resistance responses to non-pathogenic hrp mutants and demonstrated that wall alteration does not always require activation of the HR. This chapter concludes with comments on recent developments in understanding of the mechanisms of signal exchange between bacteria and plants. The role of the plant cell wall as a barrier to the establishment of successful parasitism is emphasized.

A sesquiterpene cyclase whose activity is induced in a glandless, bacterial blight-resistant line of cotton (Gossypium hirsutum L.) catalyses the conversion of (E,E)-farnesyl diphosphate to (+)-delta-cadinene. This enzyme was purified by a combination of salt-induced phase separation, hydroxylapatite fractionation, hydrophobic interaction and strong anion-exchange chromatography, and denaturing polyacrylamide gel electrophoresis, followed by renaturation with Tween 80. The purified enzyme has a molecular weight of 64-65 kDa, and exhibited a single silver-staining band following electrophoresis in analytical denaturing polyacrylamide gels. Amino acid sequences of three tryptic peptides from the enzyme have been determined and are similar to known sequences in other terpene cyclases from plants.

Of the many sesquiterpenoids that have been identified from cotton (Gossypium)1,2,four accumulate in leaves and cotyledons in response to infection by Xanthomonas campestris pv. malvacearum (Xcm), the causal agent of bacterial blight.3,4 They are 2,7-dihydroxycadalene (1), 2-hydroxy-7-methoxycadalene (2), and their yellow-green fluorescent oxidation products lacinilene C (3) and lacinilene C 7-methyl ether (4). Compounds 1, 3, and 4 have antibacterial activity in aqueous solution and hence are phytoalexins.3,4 In contrast, 2 is virtually insoluble in water, and its saturated aqueous solution is not inhibitory to Xcm.4

Clusters of mesophyll cells in cotton cotyledons at the sites of infection by Xanthomonas campestris pv. malvacearum developed a yellow-green autofluorescence which was spectrally similar to that of the phytoalexin lacinilene C (LC). Cells were isolated from infected, resistant cotyledons and separated into two fractions using a fluorescence-activated cell sorter: one-third into a more-fluorescent (F) fraction and two-thirds into a less-fluorescent (LF) fraction. F fractions had more than five times as high a percentage of cells with brightly yellow-green fluorescence as LF fractions. F fractions also contained more LC, its precursor 2,7-dihydroxycadalene (DHC), and their 7-methyl ethers per cell than LF fractions, but in general not five times as much, indicating that less-fluorescent cells in each fraction contained some phytoalexin. From the phytoalexin contents of F and LF sorted-cell fractions and the numbers of brightly fluorescent and less-fluorescent cells in each fraction, values were calculated for the phytoalexin content of an average brightly fluorescent cell and of a less-fluorescent cell. A brightly fluorescent cell contained 10–25 times as much LC as a less-fluorescent cell did, and 40 times as much DHC, the most toxic of the four related compounds. More than 90% of the DHC is estimated to be concentrated in and around fluorescent cells at infection centres.

Fourth and fifth leaves of susceptible (Ac 44) and highly resistant (Im 216) cotton plants were inoculated with a suspension of Xanthomonas campestris pv. malvacearum. Tissue responses were studied by means of electron microscopy. Half the plants were maintained in continuous dark after inoculation, the other half in 14 h light/10 h dark. Fibrillar structures enveloped the bacteria in leaf tissue of Im 216 plants under both conditions. Enveloping structures also formed in Ac 44 leaves which had been kept in continuous dark, but not in leaves maintained in a light/dark cycle. At 2 days post-inoculation, envelopes that formed in Ac 44 which had been kept in the dark ruptured, as they do in cotton lines of intermediate resistance under a light/dark cycle. Plasma membrane, cytoplasm and chloroplast degeneration occurred more rapidly than in Ac 44 kept under the light/dark cycle. Wall appositions were frequent in Im 216 in continuous dark and rare in the other interactions. Although the ultrastructural reactions of Ac 44 and Im 216 plants maintained in the dark resembled the hypersensitive response, bacteria multiplied to levels typically found in susceptible plants maintained in a light/dark cycle. The initial envelopment of bacteria and rapid tissue degeneration observed under such conditions were evidently not sufficient to cause inhibition of bacterial growth. Sesquiterpenoid phytoalexins accumulated as previously observed in Im 216 under the light/dark cycle; however, only low levels of these compounds were present in Im 216 held in continuous dark or in Ac 44 maintained under either light regime. When coinoculated with X. campestris pv. malvacearum into Im 216, the saprophyte Pseudomonas fluorescens also multiplied in continuous dark to a level typical of a compatible pathogen. Thus the mechanisms which inhibit a phytopathogen and a saprophyte failed to function during the disruptive changes that occurred in continuous dark following inoculation with X. campestris pv. malvacearum.

Upon irradiation with light (300–750 nm), 2,7-dihydroxycadalene (DHC), a phytoalexin produced by resistant cotton lines in response to bacterial infection, inactivated virions of cauliflower mosaic virus (CaMV). The treatment destroyed infectivity observed either as local lesions on inoculated turnip leaves or as systemic infection of turnip plants. Neither irradiation alone nor DHC alone destroyed infectivity. Irradiation of a DHC-CaMV DNA solution resulted in numerous nicks in the CaMV DNA. The nicked DNA was, however, as infectious as the untreated DNA. Irradiation of a suspension of CaMV virions in a DHC solution greatly decreased the electrophoretic mobility of 39 and 43 kD coat protein fragments, but not of a 37 kD fragment. The treatment also resulted in a reduction of the extractability of the DNA and in the co-migration of DNA and protein in alkaline agarose gel electrophoresis. We suggest that the inactivation of CaMV virions by irradiated DHC is due to the cross-linking of the nucleic acid binding domain of the coat protein to the viral DNA.

Infiltration of immune leaves of Gossypium hirsutum L. with 102 to 106 cells per ml of Xanthomonas malvacearum (E. F. Sm.) Dows. did not produce confluent necrosis. However, infiltration with 2 to 6 × 105 bacteria per ml, followed by reinfiltration with water 1 week later, led to rapid confluent necrosis of the leaf. When the first infiltration was with only 2 × 103 bacteria per ml, the second infiltration produced brown flecks, nearly equal in number to the bacteria originally introduced. Histological examination revealed bacteria in intercellular spaces adjacent to most of the brown host cells comprising the flecks. These findings suggest that the second vacuum infiltration dispersed bacterial colonies that had grown from the original inoculum and that brown flecks and total leaf necrosis were host responses to different densities of dispersed bacteria. In immune leaves infiltrated with 2 × 103 bacteria per ml, the bacterial population increased 2000-fold and leveled off. Following reinfiltration with water, the bacterial population increased a further 10- to 25-fold. When the second infiltration was with 2 × 103 bacteria per ml, the second inoculum multiplied at a rate similar to rates observed in previously untreated leaves. The growth studies indicate that immune cotton leaves produce a small bacteriostatic zone around each bacterial colony.

In leaves of susceptible upland cotton (Gossypium hirsutum L.), Xanthomonas malvacearum (E. F. Sm.) Dows. produced expanding water-soaked areas, whereas in leaves of an immune cotton line, small clusters of dark brown host cells developed. The number of water-soaked areas or of brown cell clusters was proportional to the number of bacteria administered, when the initial bacterial population density was within the range 20 to 2500 bacteria per cm2 of leaf area. In paradermal sections of infected immune leaves, the brown host cells stained intensely with Azure B. Bacteria were found in the intercellular spaces adjacent to 64% of these darkly stained cell clusters and adjacent to only 7% of symptomless host cells. Compact masses of bacteria were observed on the surfaces of some darkly stained host cells. We conclude that in both susceptible and immune cotton leaves infection can be initiated by a single cell of X. malvacearum which grows into a colony. Bacterial growth was logarithmic in susceptible and immune leaves, but stopped at lower population densities in immune leaves. Any initial bacterial population density less than 850 cells per cm2 in immune leaves gave a growth yield of approximately 2000 per initial bacterium. This suggests that the immune leaf responds locally to each bacterial colony.

We have found the following experiment to be an interesting way of giving students practise with quantitative colorimetric assays. They repeat an observation that was historically important in the discovery of catabolite repression. The experiment can be performed in two 3.5 hour sessions. During the first session, Escherichia coli is cultured in liquid medium containing both glucose and lactose. The students measure culture turbidity as a function of time and observe the diauxic growth pattern (Fig 1A) that was first reported by Monod and Audureau. 1' 2 The sequential utilization of the two sugars I led to the discovery that during growth on preferred carbon sources such as glucose, many enzymes that function in the catabolism of other carbon sources are repressed. This control mechanism was named 'catabolite repression' and has been found to be mediated by 3', 5' cyclic AMP. 3, 4, s Our students confirm this sequential utilization of the two sugars. They periodically withdraw samples of the culture, remove bacteria by membrane filtration, and freeze the filtrates. During the second laboratory session they determine the concentrations of glucose and lactose in the filtrates by means of a coupled glucose oxidase-peroxidase assay for glucose and the anthrone assay for hexose. The glucose assay iUustrates the high specificity characteristic of enzyme assays. Since the anthrone reaction is positive for both glucose and lactose, the contribution by glucose to the anthroue result must be calculated and subtracted to obtain the concentration of lactose.

Two physically and genetically distinct forms of ribosephosphate isomerase have been identified in Escherichia coli K12. The constitutive ribosephosphate isomerase A has a Km for ribose 5-phosphate (4.4 +/- 0.5 mM) six times greater than that of the inducible ribosephosphate isomerase B (0.83 +/- 0.13 mM). Treatment of the enzymes with 1.25 mM iodoacetate resulted in 100% loss of activity for ribosephosphate isomerase B, whereas ribosephosphate isomerase A was unaffected. Various cellular metabolites were tested and found to be without significant effect on either enzyme. The two enzymes could be separated by filtration on Sephadex G75 superfine and their apparent molecular weights were 45000 for ribosephosphate isomerase A and 32000-34000 for ribosephosphate isomerase B. Under certain conditions the two enzymes showed different patterns of heat inactivation but the results with ribosephosphate isomerase A varied in an unusual way with the protein concentration. Ribosephosphate isomerase B was formed inducibly in a mutant lacking ribosephosphate isomerase A but there was no evidence for the production of ribosephosphate isomerase B in wild-type cells. The formation of ribosephosphate isomerase B was not a consequence of the ribosephosphate isomerase B mutation, since strains could be constructed which formed both enzymes constitutively in the anticipated amounts. The ribosephosphate isomerase formed by a secondary mutant obtained from a ribosephosphate-isomerase-A-negative strain was identified as ribosephosphate isomerase B on the basis of its Km, elution profile from Sephadex G75, inhibition of iodoacetate, and heat inactivation. The ribosephosphate isomerases of another Escherichia coli K12 strain, X289, were investigated, since their properties were reported to be different from many of these described here for ribosephosphate isomerases A and B. In our hands strain X289 contained two ribosephosphate isomerases apparently identical to ribosephosphate isomerases A and B. The evidence to date suggests that ribosephosphate isomerase A catalyses the formation of ribose 5-phosphate from ribulose 5-phosphate and also participates in the reverse reaction during ribose and adenosine catabolism. The normal physiological role of the inducible ribosephosphate isomerase B is still uncertain.

This study employed cotton lines developed by breeding genes for resistance to bacterial blight into the susceptible background of cultivar Ac 44. Leaves of Ac 44 and of lines containing resistance genes B2, B3, b7, or BN, or a combination of these plus a polygenic complex were each inoculated with six strains of Xanthomonas campestris pv. malvacearum. The strains included two field isolates, one from California and one from Oklahoma, neither of which was virulent against any of the four resistance genes, as judged by visible host reaction. Also included were spontaneous mutants derived from those isolates. Three of the mutants were virulent against B3 and b7 and one was virulent against BN. Final bacterial population densities were significantly lower in incompatible cultivar/race combinations than in compatible combinations (analysis of variance followed by the protected least squares difference test: P

Incorporation of [1,2-13C2]acetate into the cotton phytoalexin 2,7-dihydroxycadalene by bacteria-inoculated cotton seedling has established a folding pattern for its carbon skeleton which differs from that deduced for the other cotton phytoalexin gossypol by earlier 14C studies.