Proc. Natd. Acad. Sci. USAVol. 88, pp. 11281-11284, December 1991Microbiology

Penetration of hard substrates by a fungus employing enormousturgor pressures

(appressorium/biodeterioration/Magnaporthe gnsea/plant pathogen/rice blast)

RICHARD J. HOWARD*t, MARGARET A. FERRARI*, DAVID H. ROACHt, AND NICHOLAS P. MONEY§*Central Research and Development, and tFibers, The DuPont Company, Wilmington, DE 19880-0402; and §Department of Biochemistry, Colorado StateUniversity, Fort Collins, CO 80523

Communicated by Arthur Kelman, September 20, 1991 (received for review June 27, 1991)

ABSTRACT Many fungal pathogens penetrate plantleaves from a specialized cell called an appressorium. The riceblast pathogen Magnaporthegnsea can also penetrate syntheticsurfaces such as poly(vinyl chloride). Previous experimentshave suggested that penetration requires an elevated appres-sorial turgor pressure. In the present report we have usednonbiodegradable Mylar membranes, exhibiting a range ofsurface hardness, to test the proposition that penetration isdriven by turgor. Reducing appressorial turgor by osmoticstress inhibited penetration of these membranes. The size of theturgor deficit required to inhibit penetration was a function ofthe surface hardness. Penetration of the hardest membraneswas inhibited by small decreases in appressorial turgor, whilepenetration of the softer membranes was sensitive only to largedecreases in turgor. Similarly, penetration of the host surfacewas inhibited in a manner comparable to penetration of thehardest Mylar membranes. Indirect measurements of turgor,obtained through osmotically induced collapse of appressoria,indicated that the infection apparatus can generate turgorpressures in excess of 8.0 MPa (80 bars). We conclude thatpenetration of synthetic membranes, and host epidermal cells,is accomplished by application of the physical force derivedfrom appressorial turgor.

The mechanism of host surface penetration by plant patho-genic fungi has been debated for nearly a century (1-6). Thepotential role of extracellular enzymes, to facilitate perfora-tion of the host cuticle or cell wall during fungal invasion, ispoorly understood (with one exception) due to the complexand ill-defined chemical nature of plant surfaces (7). On theother hand, an essential role for mechanical force during hostsurface penetration has been proposed for the rice blastfungus Magnaporthe grisea (Hebert) Barr (8). This pathogenproduces unicellular infection structures, called appressoria,which adhere tightly to the host surface and produce slenderinfection pegs that pierce the underlying cell wall. The cellwalls of appressoria contain a dense layer of pentaketide-derived melanin whose presence is correlated with a build-upof appressorial turgor pressure (8) and is essential for pene-tration (8, 9). In this study, we have inhibited penetration byexposing appressoria to solutions of high osmotic pressure.This approach was used to reduce the hydrostatic pressure(or turgor) within the infection apparatus and to estimate themagnitude of the turgor involved in penetration. Our resultsoffer unequivocal evidence for an extraordinary mechanicalcomponent of the mechanism by which appressoria pene-trate hard surfaces, but do not exclude a role in host pene-tration for some other factor such as extracellular en-zymes.

MATERIALS AND METHODSOrganism and Growth Conditions. These studies were

conducted with strain 042 (see ref. 8) of M. grisea (Hebert)Barr, telomorph of Pyricularia grisea Sacc. (10). The timecourse of infection-structure development in vitro has beenwell documented and closely resembles development on thehost (11, 12). When harvested and placed on a surface indistilled water, conidia germinate in 1-3 hr to form germtubes. By 4 hr appressoria begin to form and are firmlyattached to the substratum. By 6-8 hr their structure appearscomplete. Much later, from 20 to 240 hr depending upon thesubstrate, a penetration peg emerges from the appressoriumand penetrates the underlying surface.

Substrates. We attempted to grow M. grisea on a numberof surfaces of different composition by adding droplets ofaqueous conidial suspensions, as described previously (8).Based on an earlier report (1) two different gold substrateswere tested: 10-gm-thick gold foil (Engelhard Industries,Carteret, NJ), and cellophane membrane (8) sputter-coatedto 100 nm of gold. Films of the following materials were alsotested: poly(methyl methacrylate) (Lucite), poly(tetrafluoro-ethylene) (Teflon; see ref. 11), a neutralized copolymer ofpoly(ethylenemethacrylic acid) (Surlyn), a proprietary filmmade from the Kevlar polymer p-phenyleneterephthalamide,glass, and polyethylene (cut from PolyGloves, VWR Scien-tific). All but the latter two were obtained from Robert R.Matheson, DuPont Central Research and Development.The following six different Mylar films composed of poly-

(ethylene terephthalate) fiber were also tested (Table 1).Amorphous, as-cast, unoriented Mylar (designated sample1), obtained from John R. Barkley, DuPont ElectronicsDepartment, Circleville Research Laboratory, Circleville,OH, was thermally crystallized to generate samples desig-nated 2, 3, 5, and 6. Crystallizations were achieved byannealing at temperatures between the glass transition (67°C)and melting (280°C) temperatures of the polymer. The degreeof crystallinity and the density were controlled by the an-nealing temperature and the duration of heating (13-15).These modified Mylar samples were generated by John C.Coburn, DuPont Electronics, Wilmington, DE. An additionalMylar substrate, obtained from Robert R. Matheson anddesignated sample 4, was not treated thermally. The densitiesof all six Mylar substrates were determined by the density-gradient technique (16). Crystallinities were calculated byassuming amorphous and crystalline densities of 1.335 and1.495 g/cm3, respectively (17, 18). Surface hardness wasdetermined by Vicker's indentation techniques (19) at loadsranging from 2 to 20 g. The resulting impressions were imagedwith differential interference-contrast optics and a low-lightvideo camera and were measured from a video screen.Hardness, expressed in MPa, was calculated as H = P/2a2,

tTo whom reprint requests should be addressed.

11281

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 7,

202

1

Proc. Natl. Acad. Sci. USA 88 (1991)

Table 1. Mylar [poly(ethylene terephthalate)] films, numbered1-6, used as substrates for appressorium formationand penetration

Tmnew, tanneal Density, xtl, Hardness,Sample 0C hr g/cm3 % MPa

1 1.345 6.1 1402 90 50 1.374 24.1 1733 130 70 1.379 27.6 1804 - ND ND 2005 180 50 1.393 36.3 2406 240 50 1.405 43.6 250Sample 1 was heated at various temperatures (T0nw) and for

various times Qna) to yield samples 2, 3, 5, and 6. Sample 4 wasobtained separately. Density, crystallinity (Xtl), and hardness weredetermined as described in the text. ND, not determined.

where P is the load and 2a is the hardness impressiondiagonal.

Ceil Wail Pore Measurement. Pore size of the appressorialwall was estimated with a solute exclusion technique (20, 21).Briefly, appressoria were incubated for 15 min in concen-trated solutions of poly(ethylene glycols) (PEGs) of differentaverage molecular weights and were screened for plasmolysisand cytorrhysis (collapse). Plasmolysis occurs only when thesolute molecule can diffuse through the pores in the cell wall,whereas cytorrhysis is evidence of exclusion. Therefore, thediameter of the pores in the wall is estimated by the size ofthe smallest excluded molecules. These measurements wereconducted on melanized appressoria and on unmelanizedappressoria formed in the presence of tricyclazole (8).Measurement of Appressorial Turgor. We tried unsuccess-

fully to make direct measurements of appressorial turgor witha pressure probe (22, 23). Indirect measurement by the clas-sical incipient-plasmolysis method could not be applied to theappressoria, because of the restricted permeability of theirmelanized walls. Instead, estimates ofturgor were obtained byan incipient-cytorrhysis technique. Conidia were placed inwater droplets on Mylar membranes, where appressoria de-veloped and adhered to the surface within 6-8 hr. After 18, 26,or 46 hr, the droplets of water were replaced with aqueoussolutions ofPEG-8000. Following a 10-min incubation in PEG,the proportion of collapsed, cytorrhyzed appressoria wasdetermined from 100 cells in each concentration of PEG byphase-contrast light microscopy (8). The external osmoticpressure inducing cytorrhysis of 50% of the cells was used asour estimate of appressorial turgor in pure water. The exper-iment was repeated once with similar results.

This unconventional method for estimating turgor wasverified by an independent approach. Appressoria bathed insolutions of different osmotic pressure (controlled with PEG)were punctured with a micropipet to determine whether theysupported positive turgor. Those appressoria from whichcytoplasm was expelled after damage to the wall werejudgedto be pressurized. At higher solute concentrations the ap-pressoria did not burst, but leaked cytoplasm slowly throughthe broken wall, consistent with a lack of turgor. These cellswere stabilized against bursting by the PEG. Plots of percentstabilization against external osmotic pressure yielded esti-mates ofturgor similar to those from the incipient-cytorrhysisexperiments (data not shown).

Penetration ofMylars. Penetration was detected with eitherlight microscopy or scanning EM (8, 11). Mylar samples 1-4were sufficiently transparent to allow clear light microscopicvisualization of pegs under intact appressoria. For thedenser, opaque Mylar samples 5 and 6, scanning EM wasused to collect data. Data collected by light microscopy andscanning EM were compared and found equivalent, as fol-lows. Drops of aqueous conidial suspensions were placed onMylar and incubated as above for 9 days. The number of

appressoria forming pegs during incubation was determinedeither by examining intact appressoria by light microscopy orby sonicating the specimens for 5 min, sputter coating themwith gold, and then examining the appressorial remnants withscanning EM (8). This experiment was repeated twice, with4 samples of 100 appressoria counted by each technique foreach experiment (1600 appressoria total). No significantdifference was found between data obtained by the twomethods.

Inhibition of Penetration. Appressoria were allowed toform in water droplets for 12-18 hr on the surface of Mylarmembranes or of detached newly expanded third foliarleaves of rice cultivar M201. The water droplets were thenreplaced with PEG solutions of varying osmotic pressure.Penetration of the Mylars and rice leaves was assessed after13 days and 48 hr, respectively.

Solutions of PEG-8000 (Sigma, lot 49F-0383) were pre-pared by dissolving 2.0-13.0 g in 10.0 ml ofdistilled deionizedwater. An average molecular weight of 8410 was determinedin tetrahydrofuran at 350C by R. E. Davis (DuPont PolymerProducts) and used to calculate molarity (M). The osmoticpressure H of each solution was calculated with the formulaH = a(M) + ,3(M2), where a and j3 were derived byinterpolation from the values given by Money (24).

RESULTSSelection of Substrates. M. grisea formed appressoria on all

surfaces tested except gold foil, gold-coated cellophane, andglass. Teflon was the only surface never penetrated. TheMylar films, however, offered a series of chemically homo-geneous surfaces that were easily modified with respect tohardness (Table 1). We found a linear correlation betweensample density and hardness (y = -2525.32 + 1973.98x, r =0.96). Thus, surfaces with hardnesses in the range of 140-250MPa could be tested for mechanical penetration.

Cell Wail Porosity. Nonmelanized and melanized appres-sorial cell walls were found to differ significantly in porosity.After formation in tricyclazole and a further 17- to 20-hrincubation, 90%o of unmelanized appressoria were plasmo-lyzed in a concentrated solution of PEG-400; PEG-600 in-duced 90%o cytorrhysis. Therefore, the exclusion thresholdlay between PEG-400 and PEG-600, corresponding to a porediameter of 1-2 nm (21). Melanized appressoria were notplasmolyzed with any of the polymers tested. Even PEG-200induced cytorrhysis, indicating that while the walls werefreely permeable to water, the pore size was smaller than thediameter of these hydrated molecules (i.e., <1 nm).

Appressorial Turgor. When appressoria were probed witha micropipet tip, the wall ruptured and cytoplasm wasexpelled violently. In contrast, nonmelanized, tricyclazole-treated appressoria did not burst when impaled with themicropipet. This response of melanized appressoria coupledwith their small size (mean diameter = 9.2 + 0.8 jtm, n = 100,measured from differential interference-contrast video im-ages at a magnification of x4600) precluded direct measure-ment of turgor with a pressure probe. Incipient-cytorrhysisdata showed that turgor increased prior to penetration ofMylar membranes (Fig. 1), rising to more than +8.0 MPa by46 hr.

Penetration ofMylar Substrates. We found that appressoriacould penetrate each of the six different Mylar samples.Penetration was always correlated with the presence of apenetration peg (Fig. 2). In addition, we found that theincubation time required for penetration was directly pro-portional to surface hardness (Fig. 3A).

Penetration of the hardest surfaces was inhibited by smallincreases in external osmotic pressure, compared with thoserequired to inhibit penetration of softer Mylars (Fig. 3B).Only 20 of 100 appressoria could still penetrate the hardestMylar (sample 6, hardness 250 MPa) after an increase in

11282 Microbiology: Howard et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 7,

202

1

Proc. NatL. Acad. Sci. USA 88 (1991) 11283

~"80 -Jlh26hr

o -0- ~~46hr60-

40-

c320-

00 2 4 6 8 10

Extracellular osmotic pressure, MPa

FIG. 1. Appressoria build turgor pressure during incubation.Appressoria were formed in water on Mylar and incubated for 18, 26,or 46 hr. The water was then replaced with various concentrations ofPEG-8000, providing extracellular osmotic pressures that rangedfrom 0 to more than 10 MPa. After 10 min in PEG the proportion ofcytorrhyzed appressoria was determined from 100 cells at each PEGconcentration with phase-contrast light microscopy.

external osmotic pressure of only 0.5 MPa. On Mylar sample5 (hardness, 240 MPa), 60 penetrations occurred after thesame change in osmotic pressure. On each ofthe softer Mylarsamples (200 to 140 MPa), a 0.5-MPa increase in externalosmotic pressure had no significant effect on the number ofappressoria that penetrated the underlying surface.

Penetration of Rice Leaf. After 48 hr of incubation in PEGsolutions, penetration of rice was significantly reduced orinhibited compared with controls. Penetration was reduced801% by a solution with an osmotic pressure of about 0.5 MPaand more than 95% by a solution with an osmotic pressure of2.5 MPa. Penetration was completely inhibited by an extra-cellular osmotic pressure of about 6 MPa. The experimentwas repeated three times with similar results.

DISCUSSIONOur aim was to establish whether there was a relationshipbetween elevated turgor and penetration, not to discount thepossible involvement of enzymes during penetration of the

FIG. 2. Scanning electron micrograph documenting surface pen-etration of Mylar sample 5. A site of penetration was exposed bysonicating the Mylar, shearing the attached appressorium, leaving aring of attached cell wall surrounding the appressorium pore (8).Within the pore a hole in the Mylar is visible (arrow), produced bya penetration peg (see ref. 11) that was removed during sonication.(Bar = 1 Am.)

'100-

.40-4 4

42

ctI bJ I I 'T II

60

Extracellular osmotic pressure, MPa

FIG. 3. Appressoria were allowed to form in water, on the sixdifferent Mylar samples described in Table 1. Each data pointrepresents the observation of 100 appressoria. Both experimentswere repeated with similar results. (A) Aflter various times ofincubation the number of penetrations was determined for eachMylar sample. The softer the surface (the lower the sample number),the earlier it was penetrated. (B) After 18 hr.of incubation the waterwas replaced with various concentrations of PEG-8000. We foundthat the harder the Mylar (the higher the sample number), the lowerthe osmotic pressure required to reduce or-inhibit penetration.

host. However, for our purposes it was necessary to elimi-nate any possible involvement of chemical modification of atest surface (e.g., by the activity of extraceliular enzymes).Therefore, we chose model substrates that are thought to beno~nbiodegradabl'e. Additional important requisites for modelsubstrates included uniform integrity, hardness that could bevaried without altering chemical composition, and the abilityto induce. infectio~n-structure development .by M. grisea.

Since the classic 1895 report by Miyoshi (1), gold foil hasbeen considered by some plant pathologists as a suitable inertsurface with which to study the penetration process. Wefound that gold foils~were very fragile, and thus uniformintegrity was always questionable. More important, how-ever, was our finding that the M. grisea strains we testedfailed to form appressoria on gold foil or gold-coated cello-phane membranes.Numerous synthetic polymeric materials, unavailable to

Miyoshi, provide strong, virtually inert surfaces, resistant tobiodegradation." For example, M. grisea forms appressoriaand can penetrate poly(vinyl formal) (Formvar) membranes(25), copolymeric coverslips of poly(vinyl acetate) and poly-(vinyl chloride) (8), as well as polyethylene and the variouspolymers of ethylenemet~lacrylic acid tested here. Glass didnot normally induce appressoria, and Teflon was neverpenetrated. From these various substances it might be pos-

Microbiology: Howard et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 7,

202

1

Proc. Nati. Acad. Sci. USA 88 (1991)

sible to assemble a series of surfaces exhibiting a wide rangeofhardness. However, such a heterogeneous collection couldnot serve as a suitable model system.

Poly(ethylene terephthalate) fiber can resist biological at-tack during 7 years of exposure to sea water (for review, seeref. 26). Therefore, we think Mylar films provided an inertbarrierlfor our tests, which never exceeded a duration of 2weeks. Infection structures form readily on the surface ofMylar, which has also been shown to be penetrated byappressoria (27). In addition, and most important, the surfacehardness of Mylar can be manipulated and measured easily.Therefore, the series ofMylar films detailed in Table 1 fulfillsthe requirements for model substrates set forth above and canserve our efforts to investigate the relationship betweenpenetration and turgor pressure. With reference to Table 1,it should be noted that the Vicker's hardness of each Mylaris a measure of bulk compressibility and cannot be equated,in absolute terms, with the fungal pressure required forpenetration.

Consistent with our previous work (8), melanization re-duced the porosity of the appressorial wall to less than 1 nm.By comparison, pore diameters for the hyphal and sporangialwalls of the aquatic fungus Achlya bivexualis have beenreported as 2-3 nm and 2-4 nm, respectively (21).

Genetic evidence has shown that melanization is necessaryfor penetration (8, 9). We now ascribe this requirement formelanin to its effect upon permeability of appressorium cellwall. 'A previous report (11) showed that during the rise inturgor prior to penetration, cytoplasmic glycogen rosettesvirtually disappeared. We suggest t -atthe reduced porosityof the melanized cell wall blocks efflux of some glycogenmetabolite, supporting the substantial increase in turgorpressure.The bursting response of probed cells provided dramatic

confirmation that melanized appressoria were highly pres-surized structures (8), but it also prevented direct measure-ment of appressorial turgor. However, we are confident thatour estimates of turgor derived by the incipient-cytorrhysistechnique are valid. As indicated in Fig. l, appressoria buildextremely high pressures, higher than previously known forbiological systems. In plant and fungal cells pressures only upto 1 MPa have been reported (23, 28).A logical a priori assumption is that mechanical penetration

would proceed as a function of time. Fig. 3A clearly indicatesa consistent correlation between the length of incubationrequired before penetration of the six different Mylar films,as dictated by the hardness of each surface. Apparently,mechanical penetration of each surface occurred only aftersufficient turgor pressure was established.These findings are consistent with the proposition that

there is a direct relationship between appressorial turgorpressure and the ability to penetrate an underlying substrate.This is confirmed by data presented in Fig. 3B, wheremanipulation of appressorial turgor is shown to effect thepenetration of Mylar.

Similar experiments conducted with host leaves as a sub-strate indicate that the surface hardness of Mylar sample 6might be comparable to that of rice leaves. Because of theconsistent reduction in rice leaf penetration efficiency as aconsequence of decreasing appressorial turgor, we concludethat host penetration was also dependent upon the forceapplied by turgor. However, rice leaves were always pene-trated much faster than were any of the Mylar substrates.About 75% of appressoria on a host surface form infectionhyphae within epidermal cells by 48 hr (29). If the rice leafsurface is very hard, as suggested above, and if the mecha-nism of penetration is the same as that for Mylar, our results

(Fig. 3A) predict that penetration would require much morethan 48 hr. We conclude that the force applied by appressorialturgor is essential for the penetration of both Mylar mem-branes and rice leaves. In addition, some other process,perhaps enzymatic modification of the substrate, acceleratespenetration of the host.

We thank Dr. Richard K. Quisenberry, who first suggested Mylarfilms for our use. We also thank Dr. John C. Coburn, who generatedthe annealed Mylars, and Drs. Daniel P. O'Keefe and Robert l.Matheson for helpful discussions. We are grateful to Dr. Charles F.Koerting, Michael A. Picolleili, and John F. Pedrick for technicalassistance and to Timothy M. Bourett for helpful comments through-out this study.

1. Miyoshi, M. (1895) Jahrb. Wiss. Bot. 28, 269-289.2. Aist, J. R. (1976) in Encyclopedia of Plant Physiology, eds.

Heitefuss, R. & Williams, P. H. (Springer, New York), Vol. 4,pp. 197-221.

3. Dickinson, S. (1960) in PlantPathology:An Advanced Treatise,eds. Dickinson, S. & Horsfall, J. G. (Academic, New York),pp. 203-232.

4. van den Ende, G. & Linskens, H. F. (1974) Annu. Rev.Phytopathol. 12, 247-258.

5. Wolf, F. A. & Wolf, F. T. (1947) The Fungi (Wiley, NewYork), Vol. 2, p. 538.

6. Wood, R. K. S. (1960) in Plant Pathology: An Advanced Trea-tise, eds. Dickinson, S. & Horsfall, J. G. (Academic, NewYork), Vol. 2, pp. 233-272.

7. Kolattukudy, P. (1985) Annu. Rev. Phytopathol. 23, 223-250.8. Howard, R. J. & Ferrari, M. A. (1989) Exp. Mycol. 13, 403-

418.9. Chumley, F. G. & Valent, B. (1990) Mol. Plant-Microbe In-

teract. 3, 135-143.10. Rossman, A. Y., Howard, R. J. & Valent, B. (1990) Mycologia

82, 509-512.11. Bourett, T. M. & Howard, R. J. (1990) Can. J. Bot. 68,

329-342.12. Hamer, J. E., Howard, R. J., Chumley, F. G. & Valent, B.

(1988) Science 239, 288-290.13. flers, K. H. & Breuer, H. (1963) J. Coll. Sci. 18, 1-31.14. Mehta, A., Gaur, U. & Wunderlich, B. (1978) J. Polym. Sci.

Polym. Phys. Ed. 16, 289-296.15. Van Antwerpen, F. & Van Krevelen, D. W. (1972) J. Polym.

Sci. Polym. Phys. Ed. 10, 2423-2435.16. American Society for Testing and Materials (1990) Annu. Book

ASTM Stand. 8.01, 458-463.17. Starkweather, H. W., Jr., Zoller, P. & Jones, G. A. (1983) J.

Polym. Sci. Polym. Phys. Ed. 21, 295-299.18. Mehta, A., Gaur, U. & Wunderlich, B. (1978) J. Polym. Sci.

Polym. Phys. Ed. 16, 289-296.19. Shaw, M. C. (1966) in Mechanical Behavior ofMaterials, eds.

McClintock, F. A. & Argon, A. (Addison-Wesley, Reading,MAI), pp. 443-448.

20. Carpita, N. C., Sabularse, D., Montezinos, D. & Delmer, D. P.(1979) Science 205, 1144-1147.

21. Money, N. P. (1990) Exp. Mycol. 14, 234-242.22. Hfisken, D., Steudle, E. & Zimmermann, U. (1978) Plant

Physiol. 61, 158-163.23. Money, N. P. (199Q) Exp. Mycol. 14, 416-425.24. Money, N. P. (1989) Plant Physiol. 91, 766-769.25. Woloshuk, C. P., Sisler, H. D. & Vigil, E. L. (1983) Physiol.

Plant Pathol. 22, 245-259.26. Potts, J. E. (1984) in Kirk-Othmer Encyclopedia of Chemical

Technology, ed. Eckroth, D. (Wiley, New York), Suppl. Vol.,pp. 626-668.

27. Howard, R. J., Bourett, T. M. & Ferrai, M. A. (1991) inElectron Microscopy ofPlant Pathogens, eds. Meadgen, K. &Lesemnann, D.-E. (Springer, New York), pp. 251-264.

28. Cram, W. J. (1976) in Encyclopedia ofPlant Physiology, eds.Lfittge, U. & Pitman, M. G. (Springer, New York), Vol. 2A,pp. 283-316.

29. Heath, M. C., Valent, B., Howard, R. J. & Chumley, F. G.(1900) Can. J. Bot. 68, 1627-1637.

11284 Microbiology: Howard et al.

Dow

nloa

ded

by g

uest

on

Nov

embe

r 7,

202

1