J. Cell Sci. 17, 43-55 (1975) 43

Printed in Great Britain

THE REACTION OF MITOCHONDRIA IN THE

COLEOPTILES OF RICE (ORYZA SATIVA L.)

WITH DIAMINOBENZIDINE

HELGI OPIK

Department of Botany and Microbiology, University College of Swansea,Singleton Park, Swansea SA2 8PP, Wales, U.K.

SUMMARY

Diaminobenzidine, DAB, was applied to segments of aerobically and anaerobically growncoleoptiles of rice, Oryza sativa L., with the object of studying the location of cytochromeoxidase at the electron-microscope level. A specific staining of mitochondrial cristae and innermembrane was obtained, with no reaction in other organelles; with extended periods ofincubation, the reaction product filled the mitochondria completely. In anaerobically growncoleoptiles, the reaction was much slower and the difference was particularly marked invascular bundle companion cells and parenchyma, which gave the strongest reaction in aerobictissue, but in the anaerobic stained even less than the cortical parenchyma. The reaction wasinhibited by boiling and slowed very much by lowering of the incubation temperature from27 to 4 °C. This indicated the involvement of an enzymic reaction and cyanide inhibitionindicated that a haem enzyme was involved. The catalase inhibitor aminotriazole did notinhibit DAB oxidation. Nevertheless the specificity of the reaction for cytochrome oxidase mustbe questioned, because preheating of the tissue to 60 °C before incubation, which would beexpected to destroy cytochrome oxidase activity, failed to decrease the oxidation, at least inaerobically grown coleoptiles. It is concluded that DAB is oxidized in the rice coleoptile tissueby a cytochrome system, and the development of this system is inhibited by anaerobiosis, butthe oxidation cannot be claimed to represent cytochrome oxidase activity exclusively. Perhapsother autoxidizable, more heat-stable cytochromes participate in the reaction.

INTRODUCTION

The chemical diaminobenzidine, DAB, is used as a histochemical reagent at theelectron-microscope level, since it can be oxidized by cellular systems to an insolubleproduct which spontaneously polymerizes, forming a compound capable of reactingwith osmium tetroxide to give an electron-dense 'osmium black' deposit (Seligman,Karnovsky, Wasserkrug & Hanker, 1968). Production of electron-dense deposits inmitochondria has been considered to result ultimately from the activity of cytochromeoxidase, DAB feeding electrons into the cytochrome chain at the level of cytochromec, which is then reoxidized by cytochrome oxidase (Reith & Schiiler, 1972; Seligman,Shannon, Hoshino & Plapinger, 1973). DAB has also been used to locate sites ofperoxidase activity (Beard & NovikofF, 1969; Lumsden, Oaks & Mills, 1969; Marty,1970) and catalase activity (Bibby & Dodge, 1973).

In a previous study (Opik, 1973), it was found that whilst rice coleoptiles can growappreciably under anaerobic conditions, their capacity for respiratory oxygen uptake,and their cytochrome oxidase activity as estimated by a colour reaction utilizing a

44 H. Opik

j>-amino-diphenylamine reagent, were very considerably reduced by anaerobiosis;highly cristate mitochondria were, however, still present in coleoptiles grown for4-5 days under complete anaerobiosis. It seemed therefore of interest to see whethermitochondria of aerobically and anaerobically grown coleoptiles could be distinguishedin appearance at the electron-microscope level after treatment with a histochemicalreagent for cytochrome oxidase, and it was hoped that DAB would prove a suitablereagent. While differences in staining intensity were to be expected on the basis ofthe results with />-amino-diphenylamine, it was thought that there might also bedifferences in the intramitochondrial distribution of stain. Specific staining of mito-chondrial cristae was obtained and the reaction was slower in mitochondria of anaero-bically grown material. However, while the reaction showed many of the propertiesexpected from a reaction mediated by cytochrome oxidase, it lacked the heat sensi-tivity that would have been expected. This will be discussed below.

MATERIALS AND METHODS

Plant material

Grains of rice, Oryza sativa L. variety Italpatna, were germinated at 27 °C in the dark understerile conditions on filter paper moistened with distilled water as described previously (Opik,1973); anaerobic conditions were obtained by enclosing the seedlings in anaerobic culture jarsin a hydrogen atmosphere. The coleoptiles were grown for 3 or 4 days under aerobic conditions,4 or 5 days under anaerobic (when growth is slower), to give coleoptiles either 8 ± 1 mm or12 ± 1 mm long. The extreme tip 1-2 mm was discarded and from each coleoptile a segment1-1-5 mm long from directly behind the tip was used. Embryonic, ungerminated coleoptileswere dissected out of grains imbibed at 4 °C.

DAB treatment

Except where stated otherwise, the tissue was prefixed for 1 h in 3 % glutaraldehyde in0-05 M phosphate buffer, pH 7-2, containing also o-io M sucrose. Fixation was usually carriedout at room temperature, but when anaerobically grown plants were involved, all sampling andcutting were carried out at 0-4 °C to eliminate the possibility of cytochrome oxidase synthesisas the plants were exposed to air; the specimens were then allowed to warm gradually to roomtemperature in the fixer. The reaction of air-grown coleoptiles with DAB was the same irre-spective of prefixation temperature. After fixation the tissue was washed for 30 min with 4changes of ice-cold buffer with sucrose ('washer solution'), then vacuum infiltrated with thereaction medium and incubated at 27 °C in the dark. The standard reaction mixture contained:DAB as the hydrochloride (Sigma Chemical Co.), 25 mg/ml; phosphate buffer, 005 M;sucrose O-IOM; the final pH after DAB addition was about 6-2 and the solution was veryslightly pinkish in colour. This medium was prepared immediately before use and in the darkwith a photographic safelight, since light causes photo-oxidation of DAB (Hirai, 1971). Con-trols were incubated in medium lacking DAB. The reaction was terminated by replacing theDAB medium with a large volume of ice-cold washer solution. The time of incubation wasvaried from 15 min to overnight to get various data, but was usually 1-2 h.

Application of inhibitors

The glutaraldehyde-prefixed and washed tissue segments were pretreated for 30 min at roomtemperature with potassium cyanide, io"3 M, or 3-amino-i,2,4-triazole, 5 x io~2 M, in controlmedium, before incubation in the DAB reaction mixture containing the same inhibitorconcentrations.

Reaction of mitochondria with diaminobenzidine 45

Heat pretreatments

These were also usually applied after glutaraldehyde prefixation and washing. Boiling wascarried out in distilled water or washer solution for 5 min. For treatment at 60 °C, the tissuewas immersed in heated washer solution for 15 or 30 min.

Electron microscopy

After incubation, the tissue was washed thoroughly with ice-cold washer solution for at least1 h with 4-5 changes; sometimes the washed specimens were stored in 0-05 M cacodylate buffer,pH 7-2, with o-io M sucrose, at 2-4 °C for a few days. The material was then fixed in osmiumtetroxide, 2 % in phosphate buffer, dehydrated in an ethanol series and embedded in Spurrresin. All the solution changes from osmium tetroxide to final resin were carried out on aSakura REM-20 automatic tissue processor.

Sections with silver interference colours were cut on an LKB Ultrotome with glass knivesand either post-stained with 02 % alkaline lead citrate (Venable & Coggeshall, 1965) for 8 min,or left unstained to enhance the contrast between the DAB-stained organelles and the back-ground. An AEI Corinth 275 electron microscope at 60 kV was used for examination. Forcritical comparisons between the reaction intensities of different specimens, all micrographswere taken on the same piece of film and subjected to identical printing procedure.

RESULTS

One difficulty with the DAB was its slow penetration into the tissue. In specimenswith feeble activity, mitochondrial staining could be confined to the first few cellrows in from the cut edges even after several hours of incubation; an attempt to aidpenetration by adding to the medium 5 % dimethylsulphoxide, which has been usedin similar studies (Seligman, Nir & Plapinger, 1971) failed. With very prolongedincubation one can get almost complete staining in a i-mm segment, but this alsoresults in oxidation and precipitation of DAB in the medium, making results un-reliable, as well as producing ' overstaining' of many mitochondria (see below).Sections were therefore always cut so as to include a cut-open edge of the tissue block,and comparisons of staining intensities between different specimens could then bemade at comparable distances from the cut edge. Penetration through the intactepidermis did not seem to occur to an observable degree.

No organelles except the mitochondria stained with DAB, although a brown stainwas visible in cell walls with the light microscope.

The reaction of aerobically grown coleoptiles

The control, unstained mitochondria are shown in Fig. 1. Fig. 2 shows fully stainedmitochondria after i-h incubation in DAB; both the cristae and the mitochondrialenvelope are stained, the electron-dense deposit lying on the sides of the membranesaway from the matrix. The deposition of the DAB polymer began in a few cristae atfirst and the envelope was then devoid of stain (Fig. 3). Serial sectioning confirmedthat there really were very few stained sites in such mitochondria, though the fewcristae that were stained often showed very heavy stain, as the figure shows. On theother hand prolonged incubation in DAB resulted in a filling of the intracristal spaces,then the matrix also gained in electron opacity; granular electron-dense depositssometimes formed on the matrix sides of membranes and finally the mitochondria

46 H. Opik

became so filled with the deposit that all structure was obscured (Fig. 4) and sectioningbecame difficult, the organelles tearing and folding on cutting. There thus is a sequenceof staining: (i) a few cristae stain; (ii) all cristae stain, then the inner membrane of theenvelope stains; (iii) intracristal spaces are filled; and (iv) reaction product appears inthe matrix - 'overstating'.

In any one cell, all the mitochondria have been found to be stained to a comparabledegree, and all cells showed a positive reaction, but the staining was heaviest inmitochondria of vascular bundle parenchyma and phloem companion cells (Fig. 5);sieve tube mitochondria also stained quite strongly.

The sparse cristae in ungerminated coleoptile cell mitochondria gave a strongreaction, and their inner membrane stained also.

When the coleoptile segments were incubated in the DAB without glutaraldehydeprefixation, penetration into the tissue was appreciably slower and the general pre-servation of cellular structure was much worse than in prefixed specimens; the mito-chondria were highly swollen. The prefixation was therefore routinely employed. Anextension of the period of fixation to 12 h at 4 °C did not produce results differentfrom those obtained with the shorter fixation.

Treatment with io~3 M cyanide inhibited the reaction completely, but no inhibitionwas obtained with the catalase inhibitor, 3-amino-i,2,4-triazole at 5 x io~2 M.

Boiling abolished the reaction. Pretreatment at 60 °C for 15 or 30 min, however,far from inhibiting the reaction, enhanced it. When the heat treatment was appliedbefore prefixation in glutaraldehyde, cellular structure was practically disintegrated(heating after prefixation did not cause structural breakdown); yet even then, theswollen and disrupted mitochondria still stained heavily with DAB. This resultbeing rather unexpected, the identical treatment using the same batch of DAB wasapplied to some other tissues: seedling root tips of rice, mung bean {Phaseolus aureusL.) and pea (Pisum sativum variety Meteor); mung bean etiolated hypocotyl and mungbean cotyledon. In all controls, mitochondria stained; heating at 60 °C abolished thestaining in the mung bean root and pea root, but not in the remaining tissues.

Specimens incubated at 4 °C showed practically no reaction after 3 h, by whichtime the controls at 27 °C were heavily stained, indeed overstained.

The reaction of mitochondria in anaerobically grown coleoptiles

The characteristics of the reaction were as observed in aerobically grown plants;the staining again started in a few cristae only, with gradual increase in the numberof cristae stained. In any one cell, all mitochondria were stained to the same degree.In one respect, however, the anaerobically grown material reacted differently fromthe aerobic: heating at 60 °C for 30 min strongly decreased the reaction in mito-chondria of anaerobically grown plants, though the staining was not completelyabolished.

A detailed comparison was carried out with aerobic and anaerobic coleoptiles about8 mm long and an incubation period of 1 h in DAB; trje time was chosen as givinggood strong stain with the aerobic tissue without overstating of the matrix. In thecut-open cells at the very edge of the section, staining was strong in both aerobically

Reaction of mitochondria with diaminobenzidine \-j

and anaerobically grown plants. In the aerobic material, strong staining extended forat least 7-8 cells (about 400 /mi) into the tissue (Fig. 6). In the anaerobic material,some cells of the first intact row were still reasonably well stained, but already in thisregion cells with unstained mitochondria could be found. There was some variationbetween individual specimens: in some there was no stain visible after the first oneor two cell rows; in some, a feeble stain did extend to the 7-8-cell level, as shown inFig. 8. The cristae here are more dilated than in the aerobically grown material; thisprobably reflects fixation damage. Tissue preservation tended to be worse in theanaerobic material; the optimal fixation procedure, employing simultaneous glutar-aldehyde-osmium fixation (Opik, 1973) could not be used because of the present needfor glutaraldehyde prefixation. The differences in crista configuration hampered thecomparison between staining intensity in aerobic and anaerobic material to someextent, but it was nevertheless clear that in anaerobically grown coleoptiles the DABreaction was much weaker. Whereas in aerobically grown tissue the vascular bundlecell mitochondria stained more strongly than those of the cortical parenchyma, thevascular tissue cells of the anaerobically grown coleoptiles were even less reactive thanground parenchyma at the same distance from a cut edge (Fig. 7). Comparison ofFigs. 5 and 7 also suggests that there are fewer mitochondria per cell profile in theanaerobic vascular bundles. Rough counts on the rather limited numbers of bundleparenchyma and companion cells which were included in sections as complete longi-tudinal profiles indicated that in the anaerobic cells, the number of mitochondrialprofiles per cell section was about one-third of that in the aerobic.

DISCUSSION

Several of the present observations agree with those of other authors. The penetra-tion of DAB is reported to be limited; with Pisum sativum root cells, penetration wasnot quite complete even into 30-/<m-thick sections (Hall & Sexton, 1972); Reith &Schiiler (1972) found a marked decrease in reaction of rat liver beyond 20 /tm of theedge. The extent of penetration is of course dependent on the time of incubation.Appearance of the reaction product first in the intracristal spaces, then on the innerenvelope membrane and finally on the matrix side of the cristae, has been reportedbefore (Barnard, Afzelius & Lindberg, 1971), as has the gradual filling up of intra-cristal spaces with time (Reith & Schiiler, 1972). The feature that in 'weakly stained'mitochondria there is a distinct deposit on a few cristae rather than a weak overalldeposit was recorded by Chesnoy (1974) in the female gametophyte of a conifer. Thisuneven staining is a curious phenomenon. It might be postulated that staining pro-ceeds from, say, the outer necks of the cristae inwards and, if only a small part of eachcrista is stained at first, only a few such areas might be included in any one section; butthis possibility was excluded by examining serial sections. Perhaps the polymerizationof oxidized DAB proceeds from centres analogous to nuclei of crystallization. It ishard to believe that a few cristae are so much more reactive than all the rest.

The DAB specifically stains the rice mitochondrial cristae and inner membrane -the sites where cytochrome oxidase is considered to be located. The question of

48 H. Opik

specificity of the reaction for cytochrome oxidase must, however, be carefully con-sidered. Some authors have obtained good evidence for the mediation of cytochromeoxidase in the oxidation of DAB. Reith & Schiiler (1972) found that in isolated ratliver mitochondria DAB stimulated oxygen uptake and caused the mitochondria toassume the orthodox configuration; cyanide and azide inhibited both the stimulationof oxygen uptake and the configurational changes, and the addition of cytochrome cenhanced the rate of reaction with DAB. In a number of other instances, however,the claim that the deposition of the DAB oxidation product results from the activityof cytochrome oxidase seems to rest largely on the basis of the inhibition of thereaction with cyanide and azide - e.g. Ekes (1971), Nir, Poljakoff-Mayber & Klein(1970) and Seligman et al. (1968). Other authors have been more cautious aboutclaiming that cytochrome oxidase is definitely or exclusively involved in the productionof the cristal stain. Lumsden et al. (1969) suggest merely that the observed reaction isdue to ' cytochromes'; Marty (1970) says it is due to cytochrome c; Novikoff &Goldfisher (1969) state that the reaction on the cristae is probably due to cytochrome cand that on the envelope to cytochrome b5. Childs (1973), though finding that thereaction is sensitive to cyanide, to 30 min heating at 56 °C and to a lesser degree toazide, still states that one 'cannot identify the enzymes in the mitochondria whichinduce the DAB oxidation'.

In the rice coleoptiles, the abolition of reactivity by boiling suggests an enzymicreaction; cyanide inhibition points to the involvement of a haem enzyme and the lackof requirement for hydrogen peroxide and insensitivity towards aminotriazole (acatalase inhibitor) eliminate the actions of peroxidase and catalase respectively. Onthese observations coupled with the location of the stain one might think that thereaction is mediated by cytochrome oxidase. But the failure of the 60 °C pretreatmentto inhibit in aerobic coleoptiles, at least, is difficult to reconcile with cytochromeoxidase action, for this enzyme is highly heat-labile. (See e.g. Burstone, i960; Perner,1952.) In the rice coleoptiles, the reaction with />-amino-diphenylamine is very heat-sensitive (Opik, 1973) in contrast to the reaction with DAB. The enhancement of theDAB reaction by the heat treatment can be explained by facilitated penetration of thechemical due to some damage to cellular membranes; penetration of the chemical isclearly a limiting factor in the reaction. The observation that a 12-h fixation withglutaraldehyde does not impair the reaction with DAB is also disturbing, for suchlong fixation is said to inactivate cytochrome oxidase (Hirai, 1971).

Oxidized DAB can bind firmly to haem proteins, including cytochromes, this bind-ing being inhibited by cyanide (Hirai, 1971). Non-enzymic autoxidation of DAB ispromoted by light, and exposure of the solution to light was therefore strictly avoided;nevertheless the faintly pinkish colour of the DAB as obtained indicated that a smallamount of oxidized DAB was present: Hirai (1971) states that pure unoxidized DABis white. This opens up the possibility that the reaction observed in the rice coleoptilesis due to the reaction of some mitochondrial cytochrome(s) with contaminatingautoxidized DAB rather than to cytochrome oxidase activity. In one experiment, thepink colour from the solution was removed with activated charcoal, but this seemed torender the medium unstable, precipitation of oxidized DAB occurring very quickly,

Reaction of mitochondria with diaminobenzidine 49

and the contamination had to be accepted. Yet it is difficult to accept that all the DABreaction results from a purely non-enzymic binding of pre-oxidized DAB to mito-chondrial cytochromes. The very severe inhibition of the reaction at low temperaturesuggests continuous oxidation of DAB rather than adsorption of pre-oxidized materialwhich should still occur at 4 °C. One must also consider the amount of deposit thatcan be accumulated. Even supposing that there is enough oxidized DAB present inthe solution, or that (at 27 °C) DAB is oxidized by some heat-stable system in thetissue and diffuses to the mitochondria (diffusion of DAB oxidation products has beenclaimed - Novikoff, Novikoff, Quintana & Davis, 1972), can there be so many bindingsites in the mitochondria as to allow them to get packed full of the product ? Thelarge amount that can be accumulated with extended incubation periods stronglysuggests the continuous formation of oxidized DAB in situ in the mitochondria.Indirect evidence against the reaction being with contaminating oxidized DAB comesfrom the finding that the same batch of DAB gave a heat-sensitive reaction withmung bean and pea root tips. On the other hand, mung bean hypocotyl and cotyledonand rice root showed a heat-stable reaction, so that this is not unique to the ricecoleoptiles. Of other authors who have used DAB, Childs (1973) found the reactionsensitive to 30 min at 56 °C in the amoeba Hartmannella culbertsoni and Hirai (1971)reports the reaction in rat tissues to be destroyed by 60 °C treatment, but mostworkers have not published data on the effects of heat.

Hirai (1974) obtained evidence for a cytochrome c peroxidase reacting with DABin mitochondria of Tetrahymena pyriformis; this reaction was destroyed by a 10-minpretreatment at 50 °C, was insensitive to cyanide, and proceeded only in the presenceof hydrogen peroxide, while cytochrome oxidase activity in T. pyriformis was veryweak. It seems that several mitochondrial systems capable of DAB oxidation exist.

How then is the reaction of the rice coleoptiles to be interpreted in the light of theabove discussion ? There is no evidence for participation of catalase (aminotriazolefails to inhibit), nor peroxidase (no requirement for hydrogen peroxide), nor is therice coleoptile reaction at all similar to the cytochrome c peroxidase reaction ofTetrahymena (Hirai, 1974), which is highly heat-sensitive and unaffected by cyanide,as well as requiring hydrogen peroxide. Perhaps the final oxidation is mediated byautoxidizable cytochromes not dependent on cytochrome oxidase; plant tissues areknown to contain autoxidizable cytochromes of the b type. Not all cytochromes arehighly heat-labile; cytochrome c can withstand heating to above 100 °C (Levitt,1972). On the data available, the DAB oxidation system of the rice coleoptiles seemsto be a cytochrome system but not necessarily using cytochrome oxidase as the (sole)terminal oxidase; it is even possible that cytochrome oxidase acts in unheated tissuebut an alternative 'shunt' operates after heating.

Because of the doubt about the precise system responsible for DAB oxidation, onecannot with complete certainty equate the DAB-oxidizing activity of the mito-chondria with their normal oxidative activity during respiration. A close correlationseems quite likely, however. The highest DAB-oxidizing activity, is, for instance,found in the vascular bundles of the (aerobically grown) coleoptiles, and vasculartissue is generally known to have a higher respiration rate than ground parenchyma.

4 C E L 17

50 H. Opik

The DAB-oxidizing system develops to a lesser extent under anaerobiosis, thoughit is not completely suppressed. A quantitative comparison of the intensity of the DABreaction of different tissues is not feasible with the present method, so that one cannotcompare the degree of suppression by anaerobiosis of the development of the DAB-oxidizing system and the previously investigated ^>-amino-diphenylamine oxidizingsystem, for which the evidence for cytochrome oxidase participation is good (Opik,1973), and for which quantitative measurements were obtained. The greater heat-sensitivity of the DAB-oxidizing system in the anaerobic tissue can be interpreted asindicating a greater proportion of the activity being due to cytochrome oxidase.

One of the reasons for applying the DAB to the coleoptiles was to see whether allthe cristae in anaerobically developed mitochondria would stain equally. Duringgermination there is an increase in mitochondrial cristae, the organelles in ungermi-nated coleoptiles having only very few cristae, which do stain with DAB. It was wishedto see if during anaerobic germination the original active cristae are retained with newinactive ones being added, but no evidence for this was obtained - admittedly overallstaining of all cristae is only achieved after a stage when only few cristae stain, but thishappens equally in aerobically grown coleoptiles. There is no difference in the intra-mitochondrial distribution of the oxidizing system in aerobic and anaerobic material.

It is interesting that the vascular bundles, which stain heaviest in aerobically growncoleoptiles, stain so feebly in the anaerobic. The vascular bundle cells of the anaerobiccoleoptiles also contain fewer mitochondria than the aerobic - at least in the case of theparticular age and region of coleoptile here compared. In more advanced coleoptilespreviously examined, and where the counts were made on cells of the extreme apex,there was no lowering under anaerobiosis of the number of mitochondrial profiles percell cross-section in ground parenchyma, and though there was some indication offewer mitochondria in vascular bundles under anaerobiosis, the difference was lessthan in the younger plants now examined (Opik, 1973, and unpublished).

This work was supported by the Science Research Council (Grant number, B/RG/0803/2).I should like to thank Mrs M. Metcalfe and Miss S. Davies for help with specimen preparation,and Mr K. Jones for photographic work. The rice was kindly supplied by the United NationsFood and Agriculture Organization, Rome.

REFERENCES

BARNARD, T., AFZELIUS, B. A. & LINDBERG, O. (1971). A cytochemical investigation into thedistribution of cytochrome oxidase activity within the mitochondria of brown adipose tissuefrom the prenatal ra t . J . Ultrastruct. Res. 34, 544-566.

BEARD, M. E. & NOVIKOFF, A. B. (1969). Distribution of peroxisomes (microbodies) in thenephron of the rat. J. Cell Biol. 42, 501-518.

BIBBY, B. T. & DODGE, J. D. (1973). The ultrastructure and cytochemistry of microbodies indinoflagellates. Planta 112, 7-16.

BURSTONE, M. S. (i960). Histochemical demonstration of cytochrome oxidase with new aminereagents. J. Histochem. Cytochem. 8, 63-70.

CHESNOY, L. (1974). Evolution de l'activite respiratoire des mitochondries au cours de lamaturation du gamete femelle du Biota orientalis (Cupressacees). C. r. hebd. Stone. Acad.Sci., Paris 278, 727-730.

CHILDS, G. E. (1973). Diaminobenzidine reactivity of peroxisomes and mitochondria in aparasitic ameba, Hartmannella culbertsoni. J. Histocliem. Cytochem. 21, 26-33.

Reaction of mitochondria with diaminobenzidine 51

EKES, M. (1971). The use of diaminobenzidine (DAB) for the histochemical demonstration ofcytochrome oxidase activity in unfixed plant tissues. Histochemie 27, 103-108.

HALL, J. L. & SEXTON, R. (1972). Cytochemical localization of peroxidase activity in root cells.Planta 108, 103-120.

HIRAI, K. (1971). Comparison between 3,3'-diaminobenzidine and auto-oxidized 3,3'-diamino-benzidine in the cytochemical demonstration of oxidative enzymes. J. Histochem. Cytochem.19, 434-442.

HIRAI, K. (1974). Distribution of peroxidase activity in Tetrahymena pyriformis mitochondria.J. Histochem. Cytochem. 22, 189-202.

LEVITT, J. (1972). Responses of Plants to Environmental Stresses. New York and London:Academic Press.

LUMSDEN, R. D., OAKS, J. A. & MILLS, R. R. (1969). Mitochondrial oxidation of diaminobenzi-dine and its relationship to the cytochemical localization of tapeworm peroxidase. J. Parasit.55. 1119-1133-

MARTY, F. (1970). Les peroxysomes (microbodies) des laticiferes d'Euphorbia characias L.J. Microscopie 9, 923—948.

NIR, I., POLJAKOFF-MAYBER, A. & KLEIN, S. (1970). The effect of water stress on mitochondriaof root cells. A biochemical and cytochemical study. PL Physiol., Lancaster 45, 173-177.

NOVIKOFF, A. B. & GOLDFISCHER, S. (1969). Visualization of peroxisomes (microbodies) andmitochondria with diaminobenzidine. ,7. Histochem. Cytochem. 17, 675-680.

NOVIKOFF, A. B., NOVIKOFT, P. M., QUINTANA, N. & DAVIS, C. (1972). Diffusion artifacts in3,3'-diaminobenzidine cytochemistry. J. Histochem. Cytochem. 20, 745-749.

OPIK, H. (1973). Effect of anaerobiosis on respiratory rate, cytochrome oxidase activity andmitochondrial structures in coleoptiles of rice (Oryza sativa L.). J. Cell Sci. 12, 725-739.

PERNER, E. S. (1952). Zellphysiologische und zytologische Untersuchungen iiber den Nachweisund die Lokalisation der Cytochrom-Oxydase in ylWium-Epidermiszellen. Biol. Zbl. 71,43-O9-

REITH, A. & SCHOLER, B. (1972). Demonstration of cytochrome oxidase activity with diamino-benzidine. A biochemical and electron microscope study. J. Histochem. Cytochem. 20, 583-589-

SELIGMAN, A. M., KARNOVSKY, M. J., WASSERKRUG, H. & HANKER, J. S. (1968). Nondroplet

ultrastructural demonstration of cytochrome oxidase activity with a polymerizing osmio-philic reagent, diaminobenzidine (DAB).J. Cell Biol. 38, 1-14.

SELIGMAN, A. M., NIR, I. & PLAPINGER, R. E. (1971). An osmiophilic tetrazolium salt (DS-NTB) for ultrastructural demonstration of dehydrogenase activity. J. Histochem. Cytochem.19, 273-285.

SELIGMAN, A. M., SHANNON, W. A. Jr., HOSHINO, Y. & PLAPINGER, P. E. (1973). Some im-

portant principles in 3,3'-diaminobenzidine ultrastructure cytochemistry. J. Histochem.Cytochem. 21, 756-758.

VENABLE, J. H. & COGGESHALL, R. (1965). A simplified lead citrate stain for use in electronmicroscopy. J. Cell Biol. 25, 407-408.

{Received \\June 1974)

4-2

52 H. Opik

All the figures are from material post-stained with lead citrate.Fig. i. Mitochondria in an aerobically grown coleoptile cell: control without DABtreatment, x 44500.Fig. 2. Mitochondria in an aerobically grown coleoptile vascular bundle cell incubatedfor 1 h in DAB; cristae and inner membrane are heavily stained, x44500.Fig. 3. Mitochondria, m, in an aerobically grown coleoptile cell, with DAB stainingstarting in a few cristae only. The rest of the mitochondrial membranes appearunstained; compare their electron density with membranes of the plastid, p, and theplasmalemma, pi. x 44500.Fig. 4. 'Overstained' mitochondrion in an aerobically grown coleoptile cell incubatedfor 2 h in DAB. No membrane structure can be discerned through the heavy depositof DAB polymer; the lines in the mitochondrion are folds formed on sectioning, notcristae. p, plastid. x 44500.

Reaction of mitochondria with diaminobenzidine 53

m

54 H. Opik

Figs. 5-8 are all from material treated for 1 h in DAB.Fig. 5. Vascular bundle cells of an aerobically grown coleoptile, about 6 cell rows

from cut edge. There is heavy mitochondrial staining in both the bundle parenchymace\\pc and the companion cell, cc; the mitochondria of the companion cell seem to besomewhat swollen, x 18000.

Fig. 6. Aerobically grown coleoptile; mitochondria in cortex parenchyma cells 7-8rows from cut edge. All cristae are stained, x 44500.

Fig. 7. Vascular bundle cells of an anaerobically grown coleoptile; these cells arenot more than 3 cell rows away from the cut edge, but there is no staining of mito-chondria, st, sieve tube; the cell marked c possibly corresponds to a companion cell;these cells do not differentiate to the same degree under anaerobic conditions asduring aerobic growth, x 18000.

Fig. 8. Anaerobically grown coleoptile; mitochondria in cortex parenchyma cells7-8 rows from cut edge. Only a few cristae are stained (arrows), x 44500.

Reaction of mitochondria with diaminobenzidine 55

8