Planta (1983)159:512-517
P l a n t a 9 Springer-Verlag 1983
Ultrastrueture of the mesophyll cells of leaves of a catalase-deficient mutant of barley (Hordeum vulgate L.) Mary L. Parker 1 and Peter J. Lea2 1 Plant Breeding Institute, Marls Lane, Trumpington, Cambridge CB2 2LQ, and 2 Department of Biochemistry, Rothamsted Experimental Station, Harpenden, Herts. AL5 2JQ, UK
Abstract. The ultrastructure of mesophyll cells from leaves of a catalase-deficient homozygous mutant of barley (RPr 79/4), which grows poorly in air but normally in carbon-dioxide-enriched air, has been examined and compared with that of the cultivar Maris Mink with normal catalase levels, and with that of the F 1 progeny of the cross RPr 79/4 x Golden Promise with 50% normal catalase levels. In Marls Mink, the F1 progeny, and the mutant in which photorespiration had been suppressed by growing in air enriched to 0.2% CO2, the ultrastructure of the mesophyll cells was typical of young festucoid leaves with the peroxisomes containing thread-like inclusions. In air-grown leaves of the mutant RPr 79/4 which had developed lesions and become shrivelled, all the chloroplasts were irregular in outline, and in some the granal membranes were disrupted into abnormal honeycomb configurations and the plastid envelope was absent. In necrotic tissue, membrane fragments and osmiophilic droplets marked the sites of severely damaged chloroplasts. The peroxisomes contained diffuse tufts of electron-opaque material as well as fibrous strands. Catalase activity, visualised cytochemically by DAB, was located exclusively in the peroxisomes of Maris Mink and the F 1 progeny, but none was found in the mutant grown either in CO2-rich air, or in normal air. The role of catalase in preventing ultrastructural damage by hydrogen peroxide during photorespiration is discussed. Key words: Catalase deficiency - Hordeum (catalase) - Mutant (barley) - Peroxisome - Plastid damage.
riched air which suppresses photorespiration, has been selected from the progeny of sodium-azidetreated seeds of the cultivar Maris Mink (Kendall et al. 1983). The leaves of this mutant line, designated Rothamsted photorespiration mutant RPr 79/4, develop lesions and become white and shrivelled when grown in normal air. Slow growth is maintained by the production of new leaves which in turn become shrivelled. The mutant has slightly lower rates of CO 2 fixation than Maris Mink under a range of different conditions, but no major lesions in either photosynthetic or photorespiratory carbon metabolism (Kendall et al. 1983). Catalase activity in the leaves of the mutant was found to be less than 10% of that of Maris Mink, while the levels of other photorespiratory enzymes within isolated peroxisomes were normal. Heterozygous plants derived from the cross, mutant cultivar Golden Promise, were phenotypically normal and grew well in air, though their catalase activity was halved (Kendall et al. 1983). Peroxisomes are known to be the major site of catalase activity in green leaves (Tolbert et al. 1968; Frederick and Newcomb 1969). The object of the work described in this paper, therefore, was to compare the ultrastructure of leaf mesophyll cells of Maris Mink with that of catalase-deficient mutant lines grown in air and in CO2-enriched air, with particular reference to the peroxisomes and the cytochemical localisation of catalase activity (Frederick and Newcomb 1969; Vigil 1969, 1970). The ultrastructural changes which precede and accompany the formation of lesions in mutant leaves are also described.
Material and methods Introduction Recently, a mutant line of barley which grows poorly in air but normally in carbon-dioxide-enAbbreviation : DAB = 3,3'-diaminobenzidinetetrahydrochloride
Plant material. The young barley (Hordeum vulgare L.) plants used in this ultrastructural study were selected and grown as described by Kendall et al. 1983. Briefly, the lines and treatments consisted of: (1) the parental cultivar Marls Mink grown for 21 d in normal air; (2) the
M.L. Parker and P.J. Lea: Ultrastructure of catalase-deficient barley mutant line (RPr 79/4) with 10% normal catalase levels, grown for 21d in air enriched with CO 2 to give 0.2% CO2; (3) the mutant line grown as in (2) for 14d then transferred to normal CO 2 levels of 0.03% for 7d; and (4) the F 1 progeny of a cross between the mutant and the cultivar Golden Promise, with 50% normal catalase levels, grown in normal air for 21 d.
Electron microscopy. Transverse slices less than 1 mm wide were cut from the midpoint of the lamina of the second leaf of each plant, and fixed for 3 h in 4% glutaraldehyde in 0.05 M cacodylate buffer (pH 7.4) at room temperature. The leaf slices were washed in three changes of 0.1 M cacodylate buffer (pH 7.4) for 30 min and post-fixed overnight in 1% osmium tetroxide in the same buffer. The tissue was dehydrated through a graded ethanol series and embedded in Spurr's resin (Spurr 1969). Sections showing silver-grey interference colours were stained in uranyl acetate and lead citrate. Leaf mesophyll cells were examined and photographed in an AEI 801 electron microscope.
Localisation ofcatalase activity. The procedure is based on the oxidation of 3,3'-diaminobenzidine tetrahydrochloride (DAB) by the peroxidatic action of catalase. The oxidised DAB then interacts with osmium tetroxide to yield an insoluble granular deposit (Vigil 1970). Slices of leaf material from each treatment, fixed in glutaraldehyde and washed as above, were chopped into t-mm 2 pieces and rinsed in 0.1 M 2-amino-2-methyl-l,3propanediol buffer (AMPD; pH 9) for 10 rain. Leaf segments were preincubated in fresh AMPD buffer for 30 rain at 37~ C. They were then transferred to freshly prepared medium containing 20 mg DAB, 10 ml 0.05 M AMPD buffer (pH 9) and 0.2 ml 3% hydrogen peroxide (Frederick and Newcomb 1969). The sections were incubated in capped vials for 1 h at 37~ in the dark, then rinsed in 0.05 M AMPD buffer (pH 9) and washed in 0.05 M cacodylate buffer (pH 7.4). The leaf pieces were post-fixed in osmium tetroxide and embedded in resin as described above. The following procedures served as controls to establish whether catalase alone was responsible for oxidation of DAB. To ensure the penetration of inhibitors, preincubation periods of 1 h at 37~ C were used. (a) Sections were preincubated in AMPD buffer containing 0.02 M 3-amino-l,2,-4-triazole, then transferred to standard DAB medium with 0.02 M aminotriazole. Aminotriazole is known to inhibit catalase activity (Margoliash and Novogrodsky 1958; Frederick and Newcomb 1969) and has been shown to inhibit totally the activity of catalase extracted from normal barley leaves (Kendall et al. /983) (b) Sections were preincubated in AMPD buffer with 0.01 M potassium cyanide, followed by incubation in standard DAB medium also containing 0.0t M KCN. Haem-containing enzymes, including peroxidases are inhibited in the presence of KCN (Vigil 1970) (c) Sections were preincubated in AMPD buffer, and then in standard DAB medium without hydrogen peroxide, but with the addition of 0.002 M sodium pyruvate. This control procedure has been used with plant material to suppress the production of endogenous hydrogen peroxide during incubation in DAB medium (Vigil 1970). All sections, including those of control incubations, were examined unstained in the electron microscope.
Results Mesophyll cells from the second leaf of healthy plants of Maris Mink grown in air are typical of those of young graminaceous leaves. The cells are highly vacuolate, with much of the peripheral cyto-
513
plasm occupied by hemispherical chloroplasts with regular outlines, well-developed granal stacks and some starch (Fig. 1). The peroxisomes (Fig. 1, arrows) are generally located on the vacuolar side of the plastids. While some peroxisomes are in close contact with the outer membrane of plastids, others are not, although they may be associated at a different plane of sectioning. The peroxisomes are readily distinguished from mitochondria, being more or less circular in outline, up to 1.2 ~tm in diameter and bound by a single membrane (Fig. 1). They have a fine granular matrix containing threadlike fibrils, but lack nucleoids or crystalline inclusions. After incubation in complete DAB medium, the coarse reaction product of osmium black is located exclusively within the peroxisomes of Maris Mink (Fig. 1, inset). The DAB medium is known to penetrate poorly into thick tissue slices, giving inconsistent results in deeper-lying cells (Hall and Sexton 1972). By observing longitudinally sectioned leaf material, undamaged peripheral cells could be selected. No staining reaction was found in any of the control treatments. The ultrastructure of the leaves of the F~ progeny of the cross, mutant x Golden Promise (Fig. 2), is identical to that of Maris Mink. A group of peroxisomes and mitochondria together with two plastids is shown in Fig. 2. However, the peroxisomes stain less heavily than those of Maris Mink when incubated in DAB to locate catalase activity (Fig. 2, inset). Again, no reaction product was seen in control DAB treatments. Plants of the mutant RPr 79/4 maintained in an atmosphere enriched with CO 2 to 0.2% CO 2 grew as well as those of Maris Mink and the F 1 progeny. No differences could be detected between the ultrastructure of the mutant kept at high CO 2 levels (Fig. 3) and that of Maris Mink (Fig. 1) and the F1 progeny (Fig. 2). However, no catalase-dependent reaction product was observed anywhere in the cells of the mutant maintained in; high CO 2 (Fig. 3, inset) following incubation of the sections in complete or control DAB media. Leaves of the mutant RPr 79/4 developed necrotic lesions after growing in normal air for 7 d. The ultrastructure of leaves sampled before the onset of shrivelling (Figs. 4-8) is markedly different from that of the mutant maintained at high CO 2 levels. The plastids are less regular in outline and some have narrow projections extending into the cytoplasm (Figs. 6, 8). Some plastids contain peripheral pockets of cytoplasm (Fig. 4, arrows), which sometimes include mitochondria and peroxisomes (not shown). In some cells, disorganised
514
M.L. Parker and P.J. Lea: Ultrastructure of catalase-deficient barley
M.L. Parker and P.J. Lea: Ultrastructure of catalase-deficient barley
plastids are present (Figs. 4, 5, p). In these, the chloroplast envelope is absent, the stroma is less dense and the granal membranes are disrupted, showing abnormal honeycomb configurations. Osmiophilic lipid globules, up to I gm in diameter, are frequently found in the cytoplasm (Figs. 4-6, 8), probably forming when plastoglobuli and membrane fragments are released from disorganised plastids. Plastid damage may be so extensive that only membrane fragments and small plastoglobuli persist in the cytoplasm (Fig. 6, 8, p). Where necrosis is more advanced (Fig. 8), disrupted plastids (p) are abundant, while within the same cells the plasmalemma, tonoplast, mitochondria and nuclei are intact. The peroxisomes of mutant plants transferred to air, like those of mutant plants maintained in high CO2, show no catalase activity detectable by the DAB reaction. However, in conventionally stained sections, their appearance is quite distinctive. In addition to the granular matrix and threadlike inclusions previously described are diffuse tufted areas, frequently associated with the bounding membrane (Figs. 6, 7, arrows). Nearly all the peroxisome profiles fi'om necrotic tissue contain some of these diffuse areas. Discussion
Young plants of Maris Mink, the F1 progeny of the mutant cross and the mutant grown at high CO 2 levels were phenotypically normal and could not be distinguished on the basis of the ultrastructure of their leaf mesophyll cells. The differences in catalase activity in leaves of these three lines, reported by Kendall et al. 1983, were confirmed at the ultrastructural level using DAB. The reaction product was confined, when present, to the peroxisomes. The inclusion of inhibitors prevented
Fig. 1. Part of a mesophyll cell from a second leaf of Hordeum vulgate cv. Maris Mink showing peroxisomes (arrowed) and mitochondria (rn) adjacent to the plastids. Following incubation m DAB medium (inset), dense reaction product attributable to catalase activity is located in the peroxisomes arrowed, All scale bars in gm Fig. 2. Mesophyll cell from the barley cross, mutant x Golden Promise, showing the normal appearance of peroxisomes. The inset demonstrates the comparatively light DAB reaction product within four peroxisomes Fig. 3. Mesophyll cell from a mutant barley plant maintained at high CO 2 levels showing normal ultrastructure. The inset demonstrates the absence of reaction product due to catalase activity within the peroxisomes
515
any deposition of reaction product, indicating the specificity of the technique for catalase. No reaction product attributable to catalase activity was found in the mutant line growing either at high CO 2 levels or in air, although this line is known to contain approximately 10% of the normal amount of catalase. Presumably the cytochemical test was not sufficiently sensitive to visualise such a low level of activity. The peroxisomes in all lines contained numerous fibrillar inclusions with a distinct substructure, typical of festucoid grasses (Frederick and Newcomb 1971), but lacked crystals and nucleoids. Peroxisomes of the F 1 progeny, and the mutant grown in conditions of high CO 2 which suppress photorespiration were indistinguishable in structure', from those of Maris Mink. Those from the mutant transferred to normal air contained additional diffuse tufts of material. As far as is known, such inclusions have not been previously described, although in detached leaves of wheat, osmiophilic fibrils and deposits develop in the microbodies as the leaf senesces (Mittelheuser and Van Steveninck 1971). Certainly in the barley mutant, peroxisome inclusions developed as a result of exposure to normal air, and subsequently the leaves became necrotic and shrivelled (Kendall et al. 1983). Even cells in which there was no sign of impending plastid degeneration contained peroxisomes with inclusions. The question arises as to the relationship between the appearance of inclusion-containing peroxisomes and the eventual degeneration of plastids in catalase-deficient, air-grown mutant plants All peroxisomes contain enzymes which produce hydrogen peroxide, a compound which is exceptionally destructive in the cell through its effect on membranes (Kaiser 1976). In photorespiring green plants, hydrogen peroxide is generated within the peroxisomes during the oxidation of glycolate to glyoxylate by glycolate oxidase. The large amount of catalase within the same compartment rapidly breaks down the toxic hydrogen peroxide Ibrming oxygen and water. The presence of one dose of the normal allele for catalase production in the F 1 progeny of the cross, mutant x Golden Promise, giving 50% catalase levels, was apparently sufficient to prevent the deleterious build up of hydrogen peroxide, and likewise in the mutant grown at high CO 2 levels, hydrogen peroxide accumulation was prevented by the suppression of the photorespiratory pathway. In contrast, in the mutant grown in normal air, catalase activity of less than 10% was unable in the long term to prevent build up of peroxide with its eventual consequences.
516
M.L. Parker and P.J. Lea: Ultrastructure of catalase-deficient barley
M.L. Parker and P.J. Lea: Ultrastructure of catalase-deficient barley
In view of the general toxicity of hydrogen peroxide, it is surprising that plastid degeneration in the mutant was apparently so ordered. Peroxide may be released into the cytoplasm either by leakage through the peroxisome membrane, or following the complete destruction of the organelle. In either case, the effect of peroxide on other organelles might be expected to be rather unspecific. In fact, although all the plastids in the mutant transferred to air were irregular in outline and contained cytoplasmic invaginations, at first only one or two lost their outer envelope, releasing membrane fragments and plastoglobuli into the cytoplasm. At the same time, mitochondria, nuclei and the tonoplast membranes apparently remained intact (although they may have been functionally damaged) and only ruptured as necrosis became advanced. Studies on the effects of herbicides containing 3-amino-~,2,4-triazole, which is known to inactivate catalase by binding to the protein (Margoliash and Novogradsky 1960), have led to the suggestion that chlorosis in treated plants may be initiated by the accumulation of hydrogen peroxide in peroxisomes (Feierabend and Schubert 1978). While there are several reports on the ultrastructural effects on plastids of germinating seedlings treated with aminotriazole, only one could be found describing the effect on mature chloroplasts, and that was in maize, a C 4 plant where photorespiration is slow, and catalase activity therefore less important (Guillot-Salomon 1966). Hydrogen peroxide accumulates in ryegrass plastids following application of the herbicide paraquat (Harvey and Fraser 1980). In susceptible lines, the first detectable ultrastructural change was abnormal fusion of thylakoid grana to produce a honeycomb effect similar to that described here for mutant barley maintained in air, although in
Figs. 4-8. Mesophyll cells from mutant barley plants maintained at normal CO 2 levels for 7d. Scale bars in gin. Fig. 4. Note the absence of the chloroplast envelope and the onset of granal disruption in the plastid (p). There are pockets of cytoplasm (small arrows) at the periphery of an intact plastid. Fig. 5. The damaged plastid (p) has no chloroplast envelope, and the internal membranes have assumed unusual configurations. An osmiophilic lipid drop has appeared in the cytoplasm (small arrow). Fig. 6. The site of a severely disrupted plastid (p) is marked by membrane fragments and plastoglobuli. Neighbouring plastids appear normal. The peroxisomes (arrows) contain small areas of dense material. Fig. 7. The tufts of dense material (arrows) within peroxisomes are often associated with the bounding membrane. Fibrils are present in the matrix. Fig. 8. As leaf necrosis proceeds, many more plastids (p) disintegrate. Peroxisomes are difficult to recognise. The organelles arrowed are tentatively identified as peroxisomes. The plasmalemma, tonoplast, mitochondria (m) and nucleus (n) appear intact
517
this case the plastid envelope remained intact. Interestingly, in paraquat-tolerant lines which possess increased levels of catalase and peroxidase, diffuse deposits developed in the chloroplast stroma, while the peroxisomes appeared unaffected. We are totally indebted to the strenuous efforts of Alan Kendall and Janice Turner in carrying out the initial selection of the mutant RPr 79/4 and for maintaining it in a viable state for the last four years. We would also like to thank Drs. S.W.J. Bright, A.J. Keys and B.J. Miflin for many helpful discussions on the proiect of selecting for photorespiratory mutants in barley.
References Feierabend, J., Schubert, B. (1978) Comparative investigation of the action of several chlorosis-inducing herbicides on the biogenesis of chloroplasts and leaf microbodies, plant Physiol. 61, 1017-1022 Frederick, S.E., Newcomb, E.H. (1969) Cytochemical localization of catalase in leaf microbodies (peroxisomes). J. Cell Biol. 43, 343-353 Frederick, S.E., Newcomb, E.H. (1971) Ultrastructure and distribution of microbodies in leaves of grasses with and without COa-photorespiration. Planta 96, 152-174 Guillot-Salomon, T. (1966) Action du 3-amino-l~2,4--triazole sur l'ultrastructure des chloroplastes de Mais. C.R. Acad. Sci. Ser. D 262, 2510-2513 Hall, J.L., Sexton, R. (1972) Cytochemical localization of peroxidase activity in root cells. Planta 108, 103-120 Harvey, B.M.R., Fraser, T.W. (1980) Paraquat tolerant and susceptible perennial ryegrasses: effects of paraquat treatment on carbon dioxide uptake and ultrastructure of photosynthetic cells. Plant Cell Environ. 3, 107-117 1 Kaiser, W. (1976) The effect of hydrogen peroxide on CO2 fixation of isolated intact chloroplasts. Biochim. Biophys. Acta 440, 476-482 Kendall, A.C., Keys, A.J., Turner, J.C., Lea, P.J., Miflin, B.J. (1983) The isolation and characterisation of a catalase-deficient mutant of barley (Hordeum vulgare L.). Planta 159,
505-511 Margoliash, E., Novogrodsky, A. (1958) A study of the inhibition of catalase by 3-amino-l,2,4-triazole. Biochem. J. 68, 468475 Margoliash, E., Novogrodsky, A. (1960) Irreversible reaction of 3-amino-l,2,4-triazole and related inhibit0rs with the protein of catalase. Biochem. J. 74, 339-348 Mittelheuser, C.J., Van Steveninck, R.F.M. (197!) The ultrastructure of wheat leaves I. Changes due to natural senescence and the effects of kinetin and ABA on de~ached leaves incubated in the dark. Protoplasma 73, 239-25~ Spurr, A,R. (1969) A low-viscosity epoxy resin en]bedding medium for electron microscopy. J. Ultrastruct. P~es. 26, 3143 Tolbert, N.E., Oeser, A., Kisaki, T., Hageman, R.H., Yamazaki, R.K. (1968) Peroxisomes from spinach leaves containing enzymes related to glycolate metabolism. Ji Biol. Chem. 243, 5179-5184 Vigil, E.L. (1969) lntracellular localization of catalase (peroxidatic) activity in plant microbodies. J. Histdchem. Cytochem. 17, 425-428 Vigil, E.L. (1970) Cytochemical and developmental c]hanges in microbodies (glyoxysomes) and related organ~lles of castor bean endosperm. J. Cell Biol. 46, 435-454 Received 14 June; accepted 19 Juty 1983
P l a n t a 9 Springer-Verlag 1983
Ultrastrueture of the mesophyll cells of leaves of a catalase-deficient mutant of barley (Hordeum vulgate L.) Mary L. Parker 1 and Peter J. Lea2 1 Plant Breeding Institute, Marls Lane, Trumpington, Cambridge CB2 2LQ, and 2 Department of Biochemistry, Rothamsted Experimental Station, Harpenden, Herts. AL5 2JQ, UK
Abstract. The ultrastructure of mesophyll cells from leaves of a catalase-deficient homozygous mutant of barley (RPr 79/4), which grows poorly in air but normally in carbon-dioxide-enriched air, has been examined and compared with that of the cultivar Maris Mink with normal catalase levels, and with that of the F 1 progeny of the cross RPr 79/4 x Golden Promise with 50% normal catalase levels. In Marls Mink, the F1 progeny, and the mutant in which photorespiration had been suppressed by growing in air enriched to 0.2% CO2, the ultrastructure of the mesophyll cells was typical of young festucoid leaves with the peroxisomes containing thread-like inclusions. In air-grown leaves of the mutant RPr 79/4 which had developed lesions and become shrivelled, all the chloroplasts were irregular in outline, and in some the granal membranes were disrupted into abnormal honeycomb configurations and the plastid envelope was absent. In necrotic tissue, membrane fragments and osmiophilic droplets marked the sites of severely damaged chloroplasts. The peroxisomes contained diffuse tufts of electron-opaque material as well as fibrous strands. Catalase activity, visualised cytochemically by DAB, was located exclusively in the peroxisomes of Maris Mink and the F 1 progeny, but none was found in the mutant grown either in CO2-rich air, or in normal air. The role of catalase in preventing ultrastructural damage by hydrogen peroxide during photorespiration is discussed. Key words: Catalase deficiency - Hordeum (catalase) - Mutant (barley) - Peroxisome - Plastid damage.
riched air which suppresses photorespiration, has been selected from the progeny of sodium-azidetreated seeds of the cultivar Maris Mink (Kendall et al. 1983). The leaves of this mutant line, designated Rothamsted photorespiration mutant RPr 79/4, develop lesions and become white and shrivelled when grown in normal air. Slow growth is maintained by the production of new leaves which in turn become shrivelled. The mutant has slightly lower rates of CO 2 fixation than Maris Mink under a range of different conditions, but no major lesions in either photosynthetic or photorespiratory carbon metabolism (Kendall et al. 1983). Catalase activity in the leaves of the mutant was found to be less than 10% of that of Maris Mink, while the levels of other photorespiratory enzymes within isolated peroxisomes were normal. Heterozygous plants derived from the cross, mutant cultivar Golden Promise, were phenotypically normal and grew well in air, though their catalase activity was halved (Kendall et al. 1983). Peroxisomes are known to be the major site of catalase activity in green leaves (Tolbert et al. 1968; Frederick and Newcomb 1969). The object of the work described in this paper, therefore, was to compare the ultrastructure of leaf mesophyll cells of Maris Mink with that of catalase-deficient mutant lines grown in air and in CO2-enriched air, with particular reference to the peroxisomes and the cytochemical localisation of catalase activity (Frederick and Newcomb 1969; Vigil 1969, 1970). The ultrastructural changes which precede and accompany the formation of lesions in mutant leaves are also described.
Material and methods Introduction Recently, a mutant line of barley which grows poorly in air but normally in carbon-dioxide-enAbbreviation : DAB = 3,3'-diaminobenzidinetetrahydrochloride
Plant material. The young barley (Hordeum vulgare L.) plants used in this ultrastructural study were selected and grown as described by Kendall et al. 1983. Briefly, the lines and treatments consisted of: (1) the parental cultivar Marls Mink grown for 21 d in normal air; (2) the
M.L. Parker and P.J. Lea: Ultrastructure of catalase-deficient barley mutant line (RPr 79/4) with 10% normal catalase levels, grown for 21d in air enriched with CO 2 to give 0.2% CO2; (3) the mutant line grown as in (2) for 14d then transferred to normal CO 2 levels of 0.03% for 7d; and (4) the F 1 progeny of a cross between the mutant and the cultivar Golden Promise, with 50% normal catalase levels, grown in normal air for 21 d.
Electron microscopy. Transverse slices less than 1 mm wide were cut from the midpoint of the lamina of the second leaf of each plant, and fixed for 3 h in 4% glutaraldehyde in 0.05 M cacodylate buffer (pH 7.4) at room temperature. The leaf slices were washed in three changes of 0.1 M cacodylate buffer (pH 7.4) for 30 min and post-fixed overnight in 1% osmium tetroxide in the same buffer. The tissue was dehydrated through a graded ethanol series and embedded in Spurr's resin (Spurr 1969). Sections showing silver-grey interference colours were stained in uranyl acetate and lead citrate. Leaf mesophyll cells were examined and photographed in an AEI 801 electron microscope.
Localisation ofcatalase activity. The procedure is based on the oxidation of 3,3'-diaminobenzidine tetrahydrochloride (DAB) by the peroxidatic action of catalase. The oxidised DAB then interacts with osmium tetroxide to yield an insoluble granular deposit (Vigil 1970). Slices of leaf material from each treatment, fixed in glutaraldehyde and washed as above, were chopped into t-mm 2 pieces and rinsed in 0.1 M 2-amino-2-methyl-l,3propanediol buffer (AMPD; pH 9) for 10 rain. Leaf segments were preincubated in fresh AMPD buffer for 30 rain at 37~ C. They were then transferred to freshly prepared medium containing 20 mg DAB, 10 ml 0.05 M AMPD buffer (pH 9) and 0.2 ml 3% hydrogen peroxide (Frederick and Newcomb 1969). The sections were incubated in capped vials for 1 h at 37~ in the dark, then rinsed in 0.05 M AMPD buffer (pH 9) and washed in 0.05 M cacodylate buffer (pH 7.4). The leaf pieces were post-fixed in osmium tetroxide and embedded in resin as described above. The following procedures served as controls to establish whether catalase alone was responsible for oxidation of DAB. To ensure the penetration of inhibitors, preincubation periods of 1 h at 37~ C were used. (a) Sections were preincubated in AMPD buffer containing 0.02 M 3-amino-l,2,-4-triazole, then transferred to standard DAB medium with 0.02 M aminotriazole. Aminotriazole is known to inhibit catalase activity (Margoliash and Novogrodsky 1958; Frederick and Newcomb 1969) and has been shown to inhibit totally the activity of catalase extracted from normal barley leaves (Kendall et al. /983) (b) Sections were preincubated in AMPD buffer with 0.01 M potassium cyanide, followed by incubation in standard DAB medium also containing 0.0t M KCN. Haem-containing enzymes, including peroxidases are inhibited in the presence of KCN (Vigil 1970) (c) Sections were preincubated in AMPD buffer, and then in standard DAB medium without hydrogen peroxide, but with the addition of 0.002 M sodium pyruvate. This control procedure has been used with plant material to suppress the production of endogenous hydrogen peroxide during incubation in DAB medium (Vigil 1970). All sections, including those of control incubations, were examined unstained in the electron microscope.
Results Mesophyll cells from the second leaf of healthy plants of Maris Mink grown in air are typical of those of young graminaceous leaves. The cells are highly vacuolate, with much of the peripheral cyto-
513
plasm occupied by hemispherical chloroplasts with regular outlines, well-developed granal stacks and some starch (Fig. 1). The peroxisomes (Fig. 1, arrows) are generally located on the vacuolar side of the plastids. While some peroxisomes are in close contact with the outer membrane of plastids, others are not, although they may be associated at a different plane of sectioning. The peroxisomes are readily distinguished from mitochondria, being more or less circular in outline, up to 1.2 ~tm in diameter and bound by a single membrane (Fig. 1). They have a fine granular matrix containing threadlike fibrils, but lack nucleoids or crystalline inclusions. After incubation in complete DAB medium, the coarse reaction product of osmium black is located exclusively within the peroxisomes of Maris Mink (Fig. 1, inset). The DAB medium is known to penetrate poorly into thick tissue slices, giving inconsistent results in deeper-lying cells (Hall and Sexton 1972). By observing longitudinally sectioned leaf material, undamaged peripheral cells could be selected. No staining reaction was found in any of the control treatments. The ultrastructure of the leaves of the F~ progeny of the cross, mutant x Golden Promise (Fig. 2), is identical to that of Maris Mink. A group of peroxisomes and mitochondria together with two plastids is shown in Fig. 2. However, the peroxisomes stain less heavily than those of Maris Mink when incubated in DAB to locate catalase activity (Fig. 2, inset). Again, no reaction product was seen in control DAB treatments. Plants of the mutant RPr 79/4 maintained in an atmosphere enriched with CO 2 to 0.2% CO 2 grew as well as those of Maris Mink and the F 1 progeny. No differences could be detected between the ultrastructure of the mutant kept at high CO 2 levels (Fig. 3) and that of Maris Mink (Fig. 1) and the F1 progeny (Fig. 2). However, no catalase-dependent reaction product was observed anywhere in the cells of the mutant maintained in; high CO 2 (Fig. 3, inset) following incubation of the sections in complete or control DAB media. Leaves of the mutant RPr 79/4 developed necrotic lesions after growing in normal air for 7 d. The ultrastructure of leaves sampled before the onset of shrivelling (Figs. 4-8) is markedly different from that of the mutant maintained at high CO 2 levels. The plastids are less regular in outline and some have narrow projections extending into the cytoplasm (Figs. 6, 8). Some plastids contain peripheral pockets of cytoplasm (Fig. 4, arrows), which sometimes include mitochondria and peroxisomes (not shown). In some cells, disorganised
514
M.L. Parker and P.J. Lea: Ultrastructure of catalase-deficient barley
M.L. Parker and P.J. Lea: Ultrastructure of catalase-deficient barley
plastids are present (Figs. 4, 5, p). In these, the chloroplast envelope is absent, the stroma is less dense and the granal membranes are disrupted, showing abnormal honeycomb configurations. Osmiophilic lipid globules, up to I gm in diameter, are frequently found in the cytoplasm (Figs. 4-6, 8), probably forming when plastoglobuli and membrane fragments are released from disorganised plastids. Plastid damage may be so extensive that only membrane fragments and small plastoglobuli persist in the cytoplasm (Fig. 6, 8, p). Where necrosis is more advanced (Fig. 8), disrupted plastids (p) are abundant, while within the same cells the plasmalemma, tonoplast, mitochondria and nuclei are intact. The peroxisomes of mutant plants transferred to air, like those of mutant plants maintained in high CO2, show no catalase activity detectable by the DAB reaction. However, in conventionally stained sections, their appearance is quite distinctive. In addition to the granular matrix and threadlike inclusions previously described are diffuse tufted areas, frequently associated with the bounding membrane (Figs. 6, 7, arrows). Nearly all the peroxisome profiles fi'om necrotic tissue contain some of these diffuse areas. Discussion
Young plants of Maris Mink, the F1 progeny of the mutant cross and the mutant grown at high CO 2 levels were phenotypically normal and could not be distinguished on the basis of the ultrastructure of their leaf mesophyll cells. The differences in catalase activity in leaves of these three lines, reported by Kendall et al. 1983, were confirmed at the ultrastructural level using DAB. The reaction product was confined, when present, to the peroxisomes. The inclusion of inhibitors prevented
Fig. 1. Part of a mesophyll cell from a second leaf of Hordeum vulgate cv. Maris Mink showing peroxisomes (arrowed) and mitochondria (rn) adjacent to the plastids. Following incubation m DAB medium (inset), dense reaction product attributable to catalase activity is located in the peroxisomes arrowed, All scale bars in gm Fig. 2. Mesophyll cell from the barley cross, mutant x Golden Promise, showing the normal appearance of peroxisomes. The inset demonstrates the comparatively light DAB reaction product within four peroxisomes Fig. 3. Mesophyll cell from a mutant barley plant maintained at high CO 2 levels showing normal ultrastructure. The inset demonstrates the absence of reaction product due to catalase activity within the peroxisomes
515
any deposition of reaction product, indicating the specificity of the technique for catalase. No reaction product attributable to catalase activity was found in the mutant line growing either at high CO 2 levels or in air, although this line is known to contain approximately 10% of the normal amount of catalase. Presumably the cytochemical test was not sufficiently sensitive to visualise such a low level of activity. The peroxisomes in all lines contained numerous fibrillar inclusions with a distinct substructure, typical of festucoid grasses (Frederick and Newcomb 1971), but lacked crystals and nucleoids. Peroxisomes of the F 1 progeny, and the mutant grown in conditions of high CO 2 which suppress photorespiration were indistinguishable in structure', from those of Maris Mink. Those from the mutant transferred to normal air contained additional diffuse tufts of material. As far as is known, such inclusions have not been previously described, although in detached leaves of wheat, osmiophilic fibrils and deposits develop in the microbodies as the leaf senesces (Mittelheuser and Van Steveninck 1971). Certainly in the barley mutant, peroxisome inclusions developed as a result of exposure to normal air, and subsequently the leaves became necrotic and shrivelled (Kendall et al. 1983). Even cells in which there was no sign of impending plastid degeneration contained peroxisomes with inclusions. The question arises as to the relationship between the appearance of inclusion-containing peroxisomes and the eventual degeneration of plastids in catalase-deficient, air-grown mutant plants All peroxisomes contain enzymes which produce hydrogen peroxide, a compound which is exceptionally destructive in the cell through its effect on membranes (Kaiser 1976). In photorespiring green plants, hydrogen peroxide is generated within the peroxisomes during the oxidation of glycolate to glyoxylate by glycolate oxidase. The large amount of catalase within the same compartment rapidly breaks down the toxic hydrogen peroxide Ibrming oxygen and water. The presence of one dose of the normal allele for catalase production in the F 1 progeny of the cross, mutant x Golden Promise, giving 50% catalase levels, was apparently sufficient to prevent the deleterious build up of hydrogen peroxide, and likewise in the mutant grown at high CO 2 levels, hydrogen peroxide accumulation was prevented by the suppression of the photorespiratory pathway. In contrast, in the mutant grown in normal air, catalase activity of less than 10% was unable in the long term to prevent build up of peroxide with its eventual consequences.
516
M.L. Parker and P.J. Lea: Ultrastructure of catalase-deficient barley
M.L. Parker and P.J. Lea: Ultrastructure of catalase-deficient barley
In view of the general toxicity of hydrogen peroxide, it is surprising that plastid degeneration in the mutant was apparently so ordered. Peroxide may be released into the cytoplasm either by leakage through the peroxisome membrane, or following the complete destruction of the organelle. In either case, the effect of peroxide on other organelles might be expected to be rather unspecific. In fact, although all the plastids in the mutant transferred to air were irregular in outline and contained cytoplasmic invaginations, at first only one or two lost their outer envelope, releasing membrane fragments and plastoglobuli into the cytoplasm. At the same time, mitochondria, nuclei and the tonoplast membranes apparently remained intact (although they may have been functionally damaged) and only ruptured as necrosis became advanced. Studies on the effects of herbicides containing 3-amino-~,2,4-triazole, which is known to inactivate catalase by binding to the protein (Margoliash and Novogradsky 1960), have led to the suggestion that chlorosis in treated plants may be initiated by the accumulation of hydrogen peroxide in peroxisomes (Feierabend and Schubert 1978). While there are several reports on the ultrastructural effects on plastids of germinating seedlings treated with aminotriazole, only one could be found describing the effect on mature chloroplasts, and that was in maize, a C 4 plant where photorespiration is slow, and catalase activity therefore less important (Guillot-Salomon 1966). Hydrogen peroxide accumulates in ryegrass plastids following application of the herbicide paraquat (Harvey and Fraser 1980). In susceptible lines, the first detectable ultrastructural change was abnormal fusion of thylakoid grana to produce a honeycomb effect similar to that described here for mutant barley maintained in air, although in
Figs. 4-8. Mesophyll cells from mutant barley plants maintained at normal CO 2 levels for 7d. Scale bars in gin. Fig. 4. Note the absence of the chloroplast envelope and the onset of granal disruption in the plastid (p). There are pockets of cytoplasm (small arrows) at the periphery of an intact plastid. Fig. 5. The damaged plastid (p) has no chloroplast envelope, and the internal membranes have assumed unusual configurations. An osmiophilic lipid drop has appeared in the cytoplasm (small arrow). Fig. 6. The site of a severely disrupted plastid (p) is marked by membrane fragments and plastoglobuli. Neighbouring plastids appear normal. The peroxisomes (arrows) contain small areas of dense material. Fig. 7. The tufts of dense material (arrows) within peroxisomes are often associated with the bounding membrane. Fibrils are present in the matrix. Fig. 8. As leaf necrosis proceeds, many more plastids (p) disintegrate. Peroxisomes are difficult to recognise. The organelles arrowed are tentatively identified as peroxisomes. The plasmalemma, tonoplast, mitochondria (m) and nucleus (n) appear intact
517
this case the plastid envelope remained intact. Interestingly, in paraquat-tolerant lines which possess increased levels of catalase and peroxidase, diffuse deposits developed in the chloroplast stroma, while the peroxisomes appeared unaffected. We are totally indebted to the strenuous efforts of Alan Kendall and Janice Turner in carrying out the initial selection of the mutant RPr 79/4 and for maintaining it in a viable state for the last four years. We would also like to thank Drs. S.W.J. Bright, A.J. Keys and B.J. Miflin for many helpful discussions on the proiect of selecting for photorespiratory mutants in barley.
References Feierabend, J., Schubert, B. (1978) Comparative investigation of the action of several chlorosis-inducing herbicides on the biogenesis of chloroplasts and leaf microbodies, plant Physiol. 61, 1017-1022 Frederick, S.E., Newcomb, E.H. (1969) Cytochemical localization of catalase in leaf microbodies (peroxisomes). J. Cell Biol. 43, 343-353 Frederick, S.E., Newcomb, E.H. (1971) Ultrastructure and distribution of microbodies in leaves of grasses with and without COa-photorespiration. Planta 96, 152-174 Guillot-Salomon, T. (1966) Action du 3-amino-l~2,4--triazole sur l'ultrastructure des chloroplastes de Mais. C.R. Acad. Sci. Ser. D 262, 2510-2513 Hall, J.L., Sexton, R. (1972) Cytochemical localization of peroxidase activity in root cells. Planta 108, 103-120 Harvey, B.M.R., Fraser, T.W. (1980) Paraquat tolerant and susceptible perennial ryegrasses: effects of paraquat treatment on carbon dioxide uptake and ultrastructure of photosynthetic cells. Plant Cell Environ. 3, 107-117 1 Kaiser, W. (1976) The effect of hydrogen peroxide on CO2 fixation of isolated intact chloroplasts. Biochim. Biophys. Acta 440, 476-482 Kendall, A.C., Keys, A.J., Turner, J.C., Lea, P.J., Miflin, B.J. (1983) The isolation and characterisation of a catalase-deficient mutant of barley (Hordeum vulgare L.). Planta 159,
505-511 Margoliash, E., Novogrodsky, A. (1958) A study of the inhibition of catalase by 3-amino-l,2,4-triazole. Biochem. J. 68, 468475 Margoliash, E., Novogrodsky, A. (1960) Irreversible reaction of 3-amino-l,2,4-triazole and related inhibit0rs with the protein of catalase. Biochem. J. 74, 339-348 Mittelheuser, C.J., Van Steveninck, R.F.M. (197!) The ultrastructure of wheat leaves I. Changes due to natural senescence and the effects of kinetin and ABA on de~ached leaves incubated in the dark. Protoplasma 73, 239-25~ Spurr, A,R. (1969) A low-viscosity epoxy resin en]bedding medium for electron microscopy. J. Ultrastruct. P~es. 26, 3143 Tolbert, N.E., Oeser, A., Kisaki, T., Hageman, R.H., Yamazaki, R.K. (1968) Peroxisomes from spinach leaves containing enzymes related to glycolate metabolism. Ji Biol. Chem. 243, 5179-5184 Vigil, E.L. (1969) lntracellular localization of catalase (peroxidatic) activity in plant microbodies. J. Histdchem. Cytochem. 17, 425-428 Vigil, E.L. (1970) Cytochemical and developmental c]hanges in microbodies (glyoxysomes) and related organ~lles of castor bean endosperm. J. Cell Biol. 46, 435-454 Received 14 June; accepted 19 Juty 1983
