DYNAMIC EXPERIMENTAL EVIDENCE FOR ... - Wiley Online Library

7 downloads 15541 Views 8MB Size Report
domain hypothesis of haustorial transport and indicates that each ... t Present address: Department of Biological Sciences, Wye College, Wye, Ashford, Kent ...
iVett)/%fo/. (1987) 107, 541-548



DYNAMIC EXPERIMENTAL EVIDENCE FOR THE PLASMA MEMBRANE ATPase DOMAIN HYPOTHESIS OF HAUSTORIAL TRANSPORT AND FOR IONIC COUPLING OF THE HAUSTORIUM OF ERYSIPHE GRAMINIS TO THE HOST CELL {HORDEUM VULGARE) BY JOHN L. GAY, A. SALZBERG* AND A. M. WOODSf Department of Pure and Applied Biology, Imperial College of Science and Technology, London SW7 2BB, UK {Accepted 13 July 1987) SUMMARY

Barley coleoptile epidermes, prepared in special mounts for microscopic examination, were infected with Erysiphe graminis D.C. ex Merat (powdery mildew) and, after 2 d, used for experiments in which reagents were introduced to the exposed underside. Fluorescein, introduced by means of fluorescein diacetate, fluoresced for periods of over 45 min in the epidermal cytoplasm and for longer in the haustoria which had developed. Cytoplasmic streaming in the epidermal cells continued for over 24 h. The duration of fluorescence was reduced by FCCP, but fusicoccin increased both the duration and intensity. In similar experiments in which the upper surfaces of infected leaves were detached so that the epidermal cells were disrupted and the extrahaustorial membranes were in direct contact with the experimental solutions, the duration of haustorial fluorescence was shorter than as described above in controls and with FCCP. Fusicoccin had no effect. The fluorescence, interpreted as an indicator of intracellular pH, confirms the ATPase domain hypothesis of haustorial transport and indicates that each haustorium is ionically coupled to the cytoplasm of its host cell. Key words: Haustorium, fluorescence, intracellular pH, Erysiphe graminis (powdery mildew). INTRODUCTION

Biotrophy is a mode of nutrition whereby parasites obtain nutrients over prolonged periods from living cells of their hosts. Thus, although the host cells are not killed, they sustain abnormal losses. An hypothesis for its operation in infections by fungal haustoria was first proposed by Spencer-Phillips & Gay (1981) and has subsequently been applied to a wide range of species of pathogens and plants. However, it has not been confirmed by dynamic experiments and the present paper seeks to remedy the deficiency. The hypothesis was founded on cytochemical investigations of plants infected by powdery mildew and rust fungi (Spencer-Phillips & Gay, 1981). It was found that the region of the host plasma membrane invaginated around each haustorium did not exhibit the ATPase activity characteristic of the rest of the plasma membrane, which was indistinguishable from that in uninfected tissues. From this evidence, it was argued that solutes taken up normally by the unaffected part of * Present address: 13 Hatishbi Street, Haifa, Israel. t Present address: Department of Biological Sciences, Wye College, Wye, Ashford, Kent TN25 5AH UK. 0028-646X/87/110541 +08 $03.00/0

© 1987 The New Phytologist

542

J. L . G A Y , A. S A L Z B E R G AND

A. M .

WOODS

the plasma membrane would be off-loaded at the face adjacent to the haustorium because there the host had diminished control of solute retention. After offloading, they would be available for uptake by the parasite. It was further pointed out that the efficiency of the mechanism was likely to be high because the edge of the invaginated region of the plasma membrane was sealed to the haustorial neck (Heath, 1976; Gil & Gay, 1977; Manners & Gay, 1977; Gay & Manners, 1987) so that the cells of host and parasite were coupled in a single system of transport. Very similar systems have now been found in infections involving diverse fungi and plants so that the hypothesis has general applicability (Woods & Gay, 1983; Gay, 1984; Gay & Woods, 1987; Woods & Gay, 1987). However, in each investigation similar cytochemical and electron microscopical methods have been employed, and therefore it is essential to test the hypothesis with dynamic experiments. The technique devised by Bushnell, Dueck & Rowell (1967) in which a monolayer of infected cells (the epidermis) prepared from barley coleoptiles can be observed microscopically over a period of several days offered a unique experimental system for the test. In the present study, it has been employed in conjunction with inhibitors (vanadate and FCCP) and a promoter of proton extrusion (fusicoccin) in short-term studies of fluorescence in host and haustorial cytoplasms. M A T E R I A L S AND M E T H O D S

Organisms Erysiphe graminis D.C. ex Merat f. sp. hordei Em. Marchal, V,, (originally obtained from the Plant Breeding Institute, Cambridge, UK) infecting Hordeum vulgare L. cv. Proctor (R,,) was maintained in a growth cabinet at 20 °C with 12 h light and dark periods. Plants were infected 15 d after sowing by shaking heavily infected plants over them. Sporulating colonies had developed 7 d later and were used to inoculate prepared coleoptiles of H. vulgare L. cv. Atlas (R^). Plants were raised for experiments using a method modified from that used by Professor W. R. Bushnell, University of Minnesota, St Paul, USA. Plant pots (100 mm square) were filled to the ledge with a 1:1 mixture of potting compost (Fisons) and medium coarse sand and watered with 215 ml tap water. Seeds were sown, 20 per pot, and the pots filled with the compost-sand mixture. The surface layer was moistened and tlie pots incubated in a tray in a growth cabinet as described above. On the third day, a 10~^ dilution of a concentrate prepared by mixing 58 g NH^NOg, 58 g KCl and 58 g NH4H2PO4 in 1 1 water was added to the tray at the rate of 125 ml per pot. Subsequently, daily additions of tap water were made to the tray at the rate of 63 ml per pot. Infected coleoptile epidermes The method was essentially as described by Bushnell et al. (1967). The apical 10 to 15 mm segments of coleoptiles were excised from 10 to 12 d old plants and manipulated so that they were clamped between two perforated plastic wafers. Part of the abaxial tissues was removed so that a single layer of epidermal cells was beneath the viewing aperture. The wafers were supported over a shallow chamber containing a bathing solution of Ca(NO3)2 (10 mM), cemented on a microscope slide. The wafers had been pretreated with P T F E (Teflon) so that the solution was restricted to the lower (abaxial) surface of the coleoptile and the adaxial epidermis was bathed only on its newly exposed (interior) surface.

Haustorial transport in Erysiphe graminis

. 543

The epidermes were inoculated with young conidia obtained from colonies from which old spores had been removed on the previous day. About 20 conidia were deposited within each viewing aperture. The cultures were incubated for 2 d in the dark at 20 °C over a saturated solution of ZnSO, in a plastic container This treatment maintained a relative humidity of 90 % and thereby prevented the condensation of moisture. Before experiments were begun the bathing solution was replaced with Ca(NO3)2 (10 mM, pH 5-5) which included fluorescein diacetate (FDA) (50/tM). The latter was prepared from a stock solution (5 mM) in dry acetone which was kept refrigerated. The preparations were observed microscopically immediately after adding the new bathing solution and only used in an experiment if cytoplasmic streaming occurred in the epidermal cells, and their peripheral cytoplasm (groundplasm) and that of the haustoria and mycelia were fluorescent. Each experiment was begun, within 5 min of adding FDA, by replacing the bathing solution with a treatment medium. At this and any subsequent replacement, the wafers were lifted from the support and their underside, especially the slot below the exposed coleoptile epidermis, was thoroughly washed in a stream of the new solution. Leaf epidermes with ruptured host cells {stripped epidermes) In order to assess the contribution by host cells to the experimental results, preliminary experiments were conducted by piercing individual epidermal cells of coleoptiles set up as above. It proved difficult to obtain suflicient numbers for reliable results. As found by Sullivan, Bushnell & Rowell (1974), a considerable amount of damage to the host cytoplasm was necessary to ensure an unambiguous result. Thus, the following procedure was adopted. Barley (cv. Atlas) grown as above, was infected 15 d after sowing and used 4 to 5 d later. The adaxial surface was prepared by making a shallow incision across a half of the reverse side of the leaf and then stripping the adaxial surface by pulling, with fine forceps, across the incision. A medium comprising 2-(N-morphololino) ethanesulphonic acid (MES) (pH 6-1, 10 mM), betaine (0-3 M) (osmoticum), KCl (1 mM), MgClg (1 mM), CaClg (1 mM) and FDA (50 /*M) was added as the stripping progressed and each strip mounted under a coverglass and observed immediately. Germination of conidia Conidia of E. graminis were deposited on Ca(NO3)2 solution and allowed to germinate overnight at 20 °C in the dark. FDA was added for observation. Treatment media Sodium orthovanadate (1 mM), carbonyl cyanide-p-trifluoromethoxyphenylhydrozone (FCCP) (10/*M) or fusicoccin (50/AM) were included in the Ca(NO3)2 (10 mM) or buffer medium, usually with FDA also. After 2 to 5 min, the mixture was replaced by Ca(NO3)2 (10 mM) or buffer which was usually the only constituent of the bathing solution for the remainder of the experimental period. Observations were continued at approximately 5 min intervals for 1 to 2 h and occasionally after 24 h. Microscopy The microscope provided illumination by transmitted light and incident u.v. with filters to detect fluorescence. The filters were: excitation transmitting at A =

544

J- L. GAY, A. SALZBERG AND A. M. WOODS

430 to 450 nm; 45° dichroic mirror transmitting and reflecting, respectively, above and below A = 460nm; and barrier transmitting A > 480 nm. For observations with u.v., the field iris was closed until only the area of immediate interest was irradiated and the duration limited to < 10 s on each occasion. Fluorescence of fluorescein

Solutions of sodium fluorescein (1 [IM) were prepared in sodium phosphate buffer (0-1 M) ranging from pH 5*5 to 8-5 at intervals of 0-5 pH and fluorescence was estimated in a fluorospectrophotometer. The emission intensity at A = 510 nm when illuminated at A = 436 nm (slit widths = 1 nm) increased approximately linearly from pH 5 5 to 7*0; above pH 7 it was almost constant.

RESULTS

Coleoptile epidermes Controls. Infected epidermes with Ca(NO3)2 and FDA, but without any other treatment, exhibited identical cytoplasmic streaming in infected and uninfected cells for over 24 h. The cytoplasmic fluorescence of most of these cells continued for 45 min, when it gradually diminished and was not detectable 70 min after removal of FDA [Fig. l(a)]. The fluorescence of haustorial cytoplasm persisted for 45 to 90 min, when it diminished to zero after 75 to 120 min. When FDA (10 /*M) was added to preparations in which fluorescence was no longer detectable, fluorescence recommenced and persisted for a further 25 to 30 min. In the mycelium, fluorescence occurred in the cytoplasm but not in the vacuoles, and continued for at least as long as in the parent haustorium. After 24 h, any fluorescence observable was only present in mycelial cytoplasm. Fluorescence never occurred in superficial fungal structures which were not associated with an haustorium. Orthovanadate treatment. In three separate experiments, the results were identical with those in controls. The treatment periods were for 5, 30 or 60 min but no effect was observed during observations for 1 h after treatment.

FCCP treatment. Four separate experiments gave similar results. Cytoplasmic streaming in the epidermal c^Us ceased in less than 1 min after the reagent was added. Fluorescence of these cells was unaffected until 5 to 10 min after addition when it diminished and was undetectable at 15 to 20 min [Fig. l(c)]. Haustorial fluorescence followed a similar course, but the changes occurred approximately 10 min later than those in the epidermal cells so it was undetectable by 25 to 30 min. None of the effects was reversed following thorough washing with Ca(NO3)2 to remove the inhibitor. To ascertain whether the diminution of fluorescence could be attributed to a deficiency of FDA, the reagent was added after the final wash in some experiments. The concentration of the FDA added was limited to \0 JLLM SO that background fluorescence did not obscure the result. The duration of fluorescence was then extended up to 40 min after the treatment with FCCP. Fusicoccin treatment. Six separate experiments gave similar results. Cytoplasmic streaming in the epidermal cells continued throughout 2 h observation periods. In most instances, the rate was unaffected but on two occasions it slowed

Haustorial transport in Erysiphe graminis

545

(b)

Control

g 055 (c)

(d

o FCCP

1 o± o

(e)

I

J\ (f

Fusicoccin

00

15

30

45

60

90

120

15

30

45

Time (min)

^ 1. Explanatory diagranis to show the approximate average levels of fluorescence in infected epidermal cells of barley coleoptiles (P) and haustoria (H) (left) and in haustoria from which the leaf epidermal cytoplasm had been removed (right). The cells were pretreated with fluorescein diacetate which was removed before addition of treatment media as follows: (a), (b) none; (c), (d) f-CCP; (e), (f) fusicoccin. Arrow (c) indicates cessation of cytoplasmic streaming in epidermal cytoplasm. See text for further details of treatment media.

as soon as the toxin was applied. On every occasion, the whole epidermis behaved uniformly. Fluorescence of the epidermal and haustorial cytoplasms always became conspicuously more intense immediately ( < 1 min) after the toxin was applied [Fig. l(e)]. The increased fluorescence persisted for 60 to 90 min, when it diminished and was barely discernable at 2 h. Usually the epidermal and haustorial fluorescence diminished in unison but sometimes the former decayed first. The mycelial cytoplasm was also fluorescent. When additional FDA (10/^M) was added after the final wash, the period of bright fluorescence in the epidermal cytoplasm was extended slightly while it remained bright in the haustoria for over 2 h. Stripped leaf epidermes Controls. Cytoplasmic streaming in the preparations was never detected, confirming that the host cells were ruptured during stripping. Fluorescence occurred in the haustoria and persisted for 20 to 25 min [Fig. l(b)]. Weak fluorescence was resumed when FDA was added after 45 min incubation. When, after 30 min incubation, the medium was replaced by water, the extrahaustorial membrane distended over a 10 min period so that it separated from the haustorial surface. This indicated that it was still intact. FCCP treatment. Haustorial fluorescence diminished immediately following treatment and was not detectable after 4 min [Fig. l(d)]. Fusicoccin treatment. Fluorescence continued at the same intensity and persisted as in controls [Fig. l(f)]. Germinating conidia When conidia were mounted in FDA, the cytoplasm, excluding the vacuoles, exhibited variable fluorescence ranging from bright to zero, which persisted for 30 min. Fusicoccin had no eflFect.

546

J. L. G A Y , A. SALZBERG AND A. M. W O O D S DISCUSSION

Fluorescence in the cytoplasm (ground plasm) of cells treated with fiuorescein diacetate is due to fluorescein liberated by enzymic de-esterification and oxidation of the diester. The plasma membrane is rapidly permeated by FDA, but the anion formed is unable to pass through it and accumulates progressively (Rotman & Papermaster, 1966). Since the ester is not fluorescent, the test is commonly used to assess cell viability (Rotman & Papermaster, 1966; Heslop-Harrison & HeslopHarrison, 1970; Widholm, 1972) and clearly the epidermal and fungal cells, including many of the germinating conidia used as controls in the experiments described here possessed enzymic activity and intact plasma membranes. The persistence of cytoplasmic fluorescence in aerial parts of the fungus for 24 h after infected epidermes were treated with FDA indicates the continued retention of fluorescein by the fungal plasma membrane. At this time, restriction of fluorescein to the superficial mycelium is to be expected in light of the transport of water by haustoria of E. graminis (Bushnell & Gay, 1978) and the open path afforded by the perforate hyphal septa (Martin & Gay, 1983). When epidermal and haustorial fluorescence faded it could invariably be regained by the reintroduction of substrate, indicating that the plasma membranes were still intact. The fluorescence of fluorescein is affected by pH, diminishing with increased concentration of H^ (Rao, 1975 and present results). Thus, the observed changes in cytoplasmic fluorescence may be attributed to changes in intracellular pH. Hydrogen ion concentration also affects the ionization of fluorescein which is uncharged at pH 2. Therefore, as cytoplasmic pH decreases, fluorescein is more likely to leak out of cells, further contributing to the diminution of fluorescence. Sodium orthovanadate had no observable effect on the epidermal or fungal cells. It is therefore deduced that although it is a potent inhibitor of ATPase activity (Simons, 1979) and is efficacious when applied to the cells under investigation after aldehyde fixation (Woods, 1985), it does not permeate the plasma membrane and reach its site of action in living cells. This is in agreement with the statement that the plasma membrane is relatively impermeable to vanadate (Simons, 1979). It was therefore necessary to use FCCP to modify plasma membrane function, the reagent rapidly permeating cell membranes, rendering them permeable to protons and uncoupling ATP production from the respiratory activity of mitochondria (Douce, 1985). FCCP had an immediate effect on cytoplasmic streaming in epidermal cells in accord with a diminished availability of ATP. Fluorescence in both epidermal cells and haustoria diminished very much earlier than in controls. The change in the epidermal cells may be ascribed to a diminished rate of proton extrusion by plasma membrane ATPase starved of ATP and equilibration of pH with that of the external medium (pH 5 5). The almost immediate decrease in the fluorescence of haustoria from which the host cytoplasm had been ruptured may indicate the presence of ATPase in the haustorial plasma membrane also. The immediate increase of fluorescence when infected cells were treated with fusicoccin indicates a rapid increase in the cytoplasmic pH. This is in accord with the specific stimulation of proton extrusion by plant plasma membrane ATPase which has been attributed to this toxin (Marre, 1979). All of its other physiological effects have been ascribed to the extrusion. It is significant that haustorial fluorescence increased and decreased in synchrony with host cell fluorescence.

Haustorial transport in Erysiphe graminis

547

Since fungal ATPase activity is unaffected by fusicoccin (Marre, 1979), a conclusion which is confirmed by the absence of any effect on haustoria associated with ruptured host cytoplasms, the increased fluorescence of haustoria in intact cells must be attributed to the host cells. It follows that pH in haustoria is influenced by the rate of proton extrusion by the host. The rapidity of the haustorial response indicates a tight coupling of host cell and haustorium, and this is in accord with the demonstration that the extrahaustorial matrix and haustorial wall constitute a closed intermediate compartment due to the barrier to diffusion between the plasma membranes of host and haustorium at the haustorial neck (neckband) (Gil & Gay, 1977; Manners & Gay, 1977; Gay & Manners, 1987). It was demonstrated that the extrahaustorial domain of the host plasma membrane remained intact after the cytoplasm of the host cell was ruptured. When fusicoccin was brought into direct contact with this membrane, haustorial fluorescence was unchanged. The absence of any detectable effect on this domain of the host plasma membrane is in agreement with the absence from it of ATPase activity as demonstrated by enzyme cytochemistry (Spencer-Phillips & Gay, 1981). It follows that the ionic coupling of the cytoplasms of haustoria and host cells is due to the stimulation by fusicoccin of proton extrusion solely by the walllining region of the host plasma membrane. In conclusion, the results of these experiments are in agreement with the hypothesis of haustorium function based on an investigation of Erysiphe pisi (Spencer-Phillips & Gay, 1981). The results of cytochemical investigations of other combinations of hosts and haustoria (Gay, 1984; Gay & Woods, 1987) including H. vulgare/E. graminis (Woods, 1985) have conformed to the E. pisi model except that ATPase activity at the haustorial plasma membrane was not detected. This exception has not been resolved in the present investigation. However, in either instance, the present results indicate the extrusion of protons by the uninvaginated region of the plasma membrane of an infected cell is accompanied by a compensating flow of protons through the extrahaustorial region and the haustorial plasma membrane. Thus, from an energetic viewpoint, the second and third membranes are both potentially able to effect solute transport against a concentration gradient. This would represent a close parallel of the system which operates in mammalian epithelial cells (Kinne, 1976), except that an Na"^ flux and two rather than three plasma membranes are involved. It would also indicate that the fungus could depend on host pumps for its solute transport.

ACKNOWLEDGEMENTS

J. L. G. is grateful to Professor W. R. Bushnell and Mrs C. Curran for instruction in the preparation of infected coleoptiles, to Dr G. Mellows for a gift of fusicoccin, to Professor J . M . Palmer for valued discussions and to the Agricultural and Food Research Council for financial support. A. S. was supported by the International Association for the Exchange of Students for Technical Experience (United Kingdom) and A.M.W. by the Science and Engineering Research Council. REFERENCES & GAY, J. L. (1978). Accumulation of solutes in relation to the structure and function of haustoria in powdery mildews. In: The Powdery Mildews (Ed. by D. M. Spencer), pp. 183-235. Academic Press, London.

BUSHNELL, W . R .

548

J. L. GAY, A. SALZBERG AND A. M. W O O D S

J. & ROWELL, J. B. (1967). Living haustoria and hyphae of Erysiphe graminis f. sp. hordei with intact and partly dissected host cells of Hordeum vulgare. Canadian Journal of Botany, 45, 1719-1732. DOUCE, R. (1985). Mitochondria in Higher Plants. Structure, Function and Biogenesis. Academic Press, London. GAY, J. L . (1984). Mechanisms of biotrophy in fungal pathogens. In: Plant Diseases: Infection, Damage and Loss (Ed. by R. K. S. Wood & G. J. Jellis), pp. 49-59. Blackwell Scientific Publications, Oxford. GAY, J. L . & MANNERS, J. M. (1987). Permeability of the haustorium-host interface in powdery mildews. Physiological and Molecular Plant Pathology, 30, 389-399. GAY, J. L . & WOODS, A. M. (1987). Induced modifications in the plasma membranes of infected cells. In: Fungal Infection of Plants (Ed. by G. F. Pegg & P. G. Ayres), pp. 79-91. Cambridge University Press, London. GIL, F . & GAY, J. L. (1977). Ultrastructural and physiological properties of the host interfacial components of haustoria of Erysiphe pisi in vivo and in vitro. Physiological Plant Pathology, 10, 1-12. HEATH, M . C . (1976). Ultrastructural and functional similarity of the haustorial neckband of rust fungi and the Casparian strip of vascular plants. Canadian Journal of Botany, 54, 2484-2489. HESLOP-HARRISON, J. & HESLOP-HARRISON, Y. (1970). Evaluation of pollen viability by enzymatically induced fluorescence; intracellular hydrolysis of fluorescein diacetate. Stain Technology, 45, 115-120. KiNNE, R. (1976). Properties of the glucose transport system in the renal brush border membrane. In: Current Topics in Membranes and Transport, vol. 8 (Ed. by F. Bonner & A. Kleinzeller), pp. 209-267. Academic Press, London. MANNERS, J. M. & GAY, J. L. (1977). The morphology of haustorial comlexes isolated from apple, barley, beet and vine infected with powdery mildews. Physiological Plant Pathology, 11, 261-266. MARRE, E . (1979). Fusicoccin: a tool in plant physiology. Annual Review of Plant Physiology, 30, 273-288. MARTIN, M . & GAY, J. L. (1983). Ultrastructure of conidium development in Erysiphe pisi. Canadian Journal of Botany, 61, 2472-2495. RAO, C . N . R. (1975). Ultra-violet and Visible Spectroscopy \ Chemical Applications, 3rd Edn. Butterworth, London. ROTMAN, B. & PAPERMASTER, B. W . (1966). Membrane properties of living mammalian cells as studied by enzymatic hydrolysis of fluorogenic esters. Proceedings of the National' Academy of Sciences, U.S.A., 55, 134-141. SIMONS, T . J. B. (1979). Vanadate — a new tool for biologists. Nature, 281, 337-338. SPENCER-PHILLIPS, P. T. N. & GAY, J. L. (1981). Domains of ATPase in plasma membranes and transport through infected plant cells. New Phytologist, 89, 393^00. SULLIVAN, T . P., BUSHNELL, W . R. & ROWELL, J. B. (1974). Relations between haustoria of Erysiphe graminis and host cytoplasm in cells opened by microsurgery. Canadian Journal of Botany, 52, 987-998. WiDHOLM, J. M. (1972). The use of fluorescein diacetate and phenosafranine for determining the viability of cultured plant cells. Stain Technology, 47, 189-194. WOODS, A. M. (1985). Ultrastructural and cytochemical studies of higher plant-fungal interfaces, with special reference to biotrophy. Ph.D. thesis. University of London. WOODS, A. M. & GAY, J. L. (1983). Evidence for a neckband delimiting structural and physiological regions of the host plasmalemma associated w.th haustoria of Albugo Candida. Physiological Plant Pathology, 23, 73-88. WOODS, A. M. & GAY, J. L. (1987). The interface between haustoria of Puccinia poarum (monokaryon) and Tussilago farfara. Physiological and Molecular Plant Pathology, 30, 167-185.

BUSHNELL, W . R., DUECK,