- Email: [email protected]
Role of cutinolytic enzymes in infection of cucumber by Colletotrichum lagenarium
Physiological and Molecular Plant Pathology (1989) 35, 4 7 5-481
475
Role of cutinolytic enzymes in infection of cucumber by Colletotrichum lagenarium A . M . BONNEN
and R .
j-
HAMMERSCHMIDT+
Department of Rolant and Plant Pathology, Vichigan State University, East Lansing, MI 48824-1312,
.A .S U .
(Accepted for publication August 1989)
The role of cutinolytic enzymes in infection of cucumber by Colletotrichum lagenarium was investigated . Microscopic and visual observations of C. lagenarium penetration and colonization in the presence of organophosphate cutinase inhibitors suggested that cutinolytic enzymes do not have a primary role in penetration . Use of mutants of C. lagenarium supported these observations since only a poor correlation between cutinase production and penetration and disease expression was found, whereas a very good correlation between penetration and disease development was found .
INTRODUCTION
Over the last 15 years, cutinolytic enzymes have been shown to be involved in the penetration of the cuticle by plant pathogenic fungi [6] . Although many fungi produce cutinases, it is still unclear how important these enzymes are for cuticular penetration for a broad range of fungal pathogens . Two lines of evidence have been used to support the role for cutinases in infection . The first is derived from the use of specific organophosphate inhibitors of cutinase to block infection . Application of the potent cutinase inhibitor diisopropylfluorophosphate (DFP) along with spores of Fusarium solani f. sp . pisi or Colletotrichum gloeosporioides prevented infection of the hosts by the pathogens [3, 6, 7] . Similar results were obtained if antibodies raised against the cutinases were included in spore suspensions [3, 6] . The use of mutants of C . gloeosporoides [4] and F. solani f. sp . pisi [2] or natural variants of F. solani f. sp . pisi [7] have provided genetic evidence for the role of cutinase . Research with C. lindemuthianum, C. lagenarium and Pyricularia oryzae, however, has suggested that mechanical penetration is the mechanism of penetration by these fungi [5, 9-13] . This was based on the observation that melanized, and, thus, structurally sound, appressoria are needed for direct penetration . In the case of C. lindemuthianum [12] and P . oryzae [13], DFP had no effect on penetration . Since C . lagenarium has been shown to need melanized appressoria for infection [2, 9-11], and we have demonstrated that this pathogen produces cutinolytic enzymes [1], the following study was carried out to try to determine the role of cutinases in the C . lagenarium-cucumber interactions . t Michigan Agricultural Experiment Station Journal Article No . 13118 . 1 To whom correspondence should be addressed .
0885-5765/89/120475+07 $03 .00/0
© 1989 Academic Press Limited
476
A. M . Bonnen and R . Hammerschmidt
MATERIALS AND METHODS
Culture offungi Colletotrichum lagenarium
(Pass .) E11 . and Halst . race 1 was maintained as previously
described [1] . d-Iutagenesis procedure A C . lagenarium spore
suspension of 1 . 75 x 106 spores ml_ i was exposed to 1 . 0 ° ethyl methane sulfonate in 0 . 05 M sodium phosphate buffer, pH 7 . 0 for 1 . 75 h resulting in 99°,, kill (as calculated from a kill curve) . The reaction was terminated by a 100 fold dilution with sterile distilled H 2 O (dH 2O) . The spore suspension was plated onto PDA containing 0 . 0001 % Terra Coat L-205 (Gustafson Chemical, Co .) (included for maintenance of discrete colonies) and incubated at 21 °C for 6 days . Terra Coat L-205 did not inhibit spore germination . Surviving colonies were transferred to PDA plates (without Terra Coat L-205) . Mutant growth and enzyme assay
Putative mutants were screened for growth on cutin and for p-nitrophenylbutyrate (PNB) hydrolase and cutinase activities . For these assays, each putative mutant was transferred to Nobel water agar and allowed to grow for 3-4 days . Plugs of mycelium (1 mm') from the advancing edge of each colony were transferred to 24 well tissue culture plates (Corning) containing 2 ml of the previously described minimal media plus cutin [1] in each well . Each plate contained two replicates of the wild type and three replicates were made of each plate . After 3 weeks the amount of growth in each well was scored . The culture fluid was assayed for PNB hydrolase and cutinase activity as previously described [1] . Pathogenicity test
The surface of cotyledons from 8-10 day old cucumber plants were inoculated with 6-8 droplets (20 pl per droplet) of a 106 spores ml -' suspension of C . lagenarium spores (wild type or mutant) . Inoculated plants were maintained at 100% for 24 h . When the cuticle was to be by-passed, cotyledons were inoculated by injection of spores via a syringe . Disease ratings were made every 24 h beginning 48 h after inoculation . Determination of_ percentage penetration
Cucumber hypocotyl segments, 2. 0-2 . 5 cm in length excised from 4-6 day old etiolated seedlings, were surface inoculated with 4-8 droplets (approximately 20 µl per droplet) of 10 6 spores ml - ' . Epidermal peels were taken at 48 h and stored in 100 % ethanol . Prior to microscopic examination the peels were rinsed in dH 2 O and stained with 0 . 1 toluidine blue (in 0 . 1 M sodium phosphate buffer, pH 6. 8) . Infection inhibition assays with paraoxon Stock solutions of 2 . 0 mm and 2 . 0
µM of paraoxon (O,O-diethyl-O-p-nitrophenylphosphate) were prepared in dH 2 O and stored at 4 ° C . Procedures for the pathogenicity studies were identical to those previously described except that the spores were suspended in 0 . 05 M sodium phosphate buffer, pH 7 . 0 containing the appropriate concentration of paraoxon .
477
Role of cutinolytic enzymes in cucumber infection RESULTS
Effect of paraoxon on pathogenecity and growth of C . lagenarium The radial growth of C . lagenarium was unaffected by any
of the concentrations of paraoxon used (0 .02, 0 . 2 and 2 . 0 µM) (Table 1) . Paraoxon also had no effect on the ability of the fungus to cause disease ('Table 1) . At a concentration of 2 . 0 µM symptom development was only slightly behind that observed for the control (no paraoxon) at 4 days and by 6 days was equal to the control . The ability of C. lagenarium to penetrate etiolated cucumber hypocotyls in the presence of paraoxon was also tested ("Table I ) . At the two highest concentrations, 0 . 2 µM and 2 . 0 pM, penetration was I l fect of paraoxon on
in vitro
ABLE
growth of C .
I lagenarium
and mJection
Paraoxon Concentration
growth" 1 nfection ` Penetration" Ire vitro
of cucumber
,11M ; °
0. 02
0.2
2 .0
96 100 81 '41)
95 100 62 (81)
97 100 70 (77)
Values in table are expressed as percent of water control . Radial growth on PDA containing indicated paraoxon concentration . Lesion formation on cucumber cotyledons . Expressed as percentage penetration from apprcscria as compared to control at 48 h alter noculation (value in parentheses is percentage of spores producing appressoria relative to on t rol) . reduced by approximately 30°x, . Percentage appressoria formed in the presence of paraoxon as compared to water was utilized as a control, as well as radial growth, in order to determine the effect of the inhibitor on the general metabolism of the fungus . Although radial growth was unaffected, the number of appressoria produced was significantly reduced (Table 1) . This result indicated that the inhibitor was not specific for the satin degrading enzymes, and as such, any reduction in penetration could not be taken as being strictly due to a loss in cutinolytic activity . The observation that the lowest concentration of paraoxon (0 . 02 1Mj resulted in the greatest reduction in number of appressoria produced was an unexpected but reproducible result .
Generation and characterization of mutants of C . Mutants of C . lagenarium were generated
lagenarium in attempts to help define the role of cutinolytic enzymes in the penetration of cucumber by this fungus . The first screens tested for growth on cutin (in liquid culture medium) as the sole carbon source and levels of PNB hydrolase activity per ml of culture fluid . Individuals exhibiting either increased or decreased levels of hydrolase activity or growth on the eutin substrate, relative to the wild-type, were chosen for further screening . In order to determine if reduced growth on cutin was a result of auxotrophy or of a general inability to metabolize carbon, those colonies exhibiting reduced growth on eutin, relative to the wild-type, were grown on minimal media [1] containing either sucrose or glucose . Those with reduced growth on these media were discarded .
478
A . M . Bonnen and R . Hammerschmidt
The remaining mutants were then checked fhr their cutinolytic activity and pathogenicity . In order to discern between those mutants which could not penetrate and those which were no longer able to colonize the tissue, all isolates unable to cause disease when surface inoculated were infiltrated into the tissue . Any isolates unable to cause disease when infiltrated were discarded . The radial growth of the 13 remaining mutants, on PDA, was found to be equal to or greater than the wild-type, with the exception of M-9, M-24 and M-30, which had 70° ;, the growth of the wild-type ('fable 2) . Each of the mutants was tested for : (I) PNB hydrolase activity ; (2) cutinolytic. activity ; (3) ability to penetrate etiolated cucumber hypocotyls and (4) ability to infect cucumber cotyledons (pathogencity) (Table 2) . All the mutants were able to produce melanized appressoria . Poor TABLE 2
Characteristics of wild type and mutant Colletotrichum lagenarium
Isolate Wild type M-4 M-9 M-11 M-18 M-20 M-24 M-25 M-29 ~1-30 M-31 M-36 M-37 M-42
Growths on cutin
PNB hydrolase activity'
Cutinolytic activity`
+ + +
0. 36 0. 38 0. 14 0-37 0. 24 0. 19 0. 04 0-23 0-53 0-05 0 . 21 0 . 15 0 . 25 0 . 22
7 14 . 5 4.6 9. 1 14. 2 11 . 2 0-5 10. 1 16 1 0. 8 7. 5 7-1 10 . 3 19 . 7
+ + + + + + + + + +
12 .
penetration ° 86 96 79 81 91 92 15 94 15 8 76 3 90 79
Pathogenicity 4-25 4. 75 2.5 3 . 25 3.5 445 0.0 4 . 75 0-0 0.0 4.3 0.0 4 . 75 5-5
- = no growth ; + + = good growth -t AA405 min -t ml Drnin-t ( x 10') 8 h -t ml -t ° percent of appressoria which had penetrated 48 h after inoculation disease scored at 4 days after inoculation : 0 = no symptoms ; 1 = small water soaked lesions at 25 °¢, or fewer inoculation sites ; 2 = water soaking at 25-50°0 of inoculation sites ; 3 = water soaking at all inoculation sites ; 4 = well developed lesions at all sites with necrotic centres ; 5 = lesions all necrotic ; 6 = lesions necrotic and coalesced .
correlations were found for the relationship between PNB hydrolase activity and pathogencity (R = 0 . 25) and between cutinolytic activity and pathogenicity (R = 0 . 56) . These results are presented in Table 1 and Fig . I a and 1 b . Two mutants, M-29 and M-36 exhibited very low pathogenicity but had wild type or higher levels of cutinase production, respectively . A highly significant correlation (R = 0 . 94) was found for percent penetration v . pathogenicity (Table 2, Fig . 1 C) . It is important to note that M-29 and M-36 did not exhibit high numbers of penetrations even though they did produce melanized appressoria and high levels of cutinase (Table 1) . A fair correlation (R = 0 . 71) was found between the PNB hydrolase activity v . cutinolytic activity (Table 2, Fig . 2) .
Role of cutinolytic enzymes in cucumber infection
(a)
r
479 b)
20
100
18
E 0.5
ô 16E a 04
∎
v
10 IE
u
40
6
30
4
20
M
c- 0-I t
50
a
i
a 0
r
•
8
Q0.2
70
o 60
Co
v
ô
o
∎
12
0 0.3
2
80
14
â0
(c 1
90
10 0
2
0$
.
1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4 5 6
Pathogenicity
Fin . 1 . Relationship between pathogenicity r . PNB hydrolase activity (a), vs . cutinolytic activity (b) or vs . penetration into host epidermal cells (c) . ( •, mutants ; /, wild type) . Data from Table 2 .
20 18 ô 16 14 ∎ 12
.
CC) 10 E
9
ao 0
8
.
6 4
0 0. 1
0 .3 0.2
0. 5 0.4
PNB hydrolase (40D405 min 1 ml -1 )
2 . Relationship between cutinolytic activity r . PNB hydrolase activity . • = mutants ; wild type . Data from Table 2 .
FIG .
DISCUSSION
Several distinct proteins or protein fractions from C . lagenarium have been shown to contain cutinolytic activity [1] . Whether any are involved in the penetration of cucumber by C. lagenarium is unknown . Studies utilizing a specific inhibitor of cutinase, paraoxon, were undertaken in attempts to determine the role of cutinases in penetration. Paraoxon, at all concentrations tested (0 . 02, 0 . 2 and 2 . 0 µM), failed to affect the radial growth of C . lagenarium on PDA . This presumably indicates normal metabolic functioning of the fungus . However, a reduction in the number of appressoria produced in the presence of paraoxon indicated that the inhibitor was not specific for the cutinolytic enzymes . Thus, the observed reduction in percent penetration of etiolated cucumber hypocotyls could not be attributed solely to the loss of cutin degrading activity . Despite the reduction in penetration efficiency, no reduction in disease was
480
A . M . Bonnen and R . Hammerschmidt
observed . These results are quite different from that reported for F. solani f. sp . pisi in which paraoxon severely reduced infection on pea hypocotyls without affecting the growth of the fungus [7] . This observation was used to support the involvement of cutinolytic enzymes in penetration by F. solani f. sp . pisi . It must now be taken as an indication that cutin degrading enzymes may not be involved (at least not as primary determinants) in penetration by C. lagenarium . Mutants of C. lagenarium were generated also in au attempt to determine the role of cutinase in the penetration of cucumber by this fungus . No isolates from the mutagenesis treatment were found which could not grow at all on cutin as the sole carbon source, although there were individuals with significantly reduced growth on the cutin . Two of these (M-24 and M-30) also had reduced levels of cutin degrading activity (Table 2) . These two were non-pathogenic unless infiltrated into the leaf tissue (thus by-passing penetration of the cuticle) . Two isolates, M-4 and M-42, with increased pathogenicity relative to the wild-type also showed incrased cutinolytic activity . This information might imply that cutinolytic enzymes arc important in penetration ; however, two other isolates were found which had reduced pathogenicity (M-29 and M-36) but with significant to very high levels of cutinolytic activity . These were also pathogenic only when infiltrated into the tissue . Correlation coefficients were determined for pathogenicity versus PNB hydrolase activity, cutinolytic activity and percent penetration for the 13 mutants and the parental wild type isolate. Of these three comparisons, only pathogenicity v . penetration was significantly correlated (R = 0 . 94) . The correlations between pathogenicity and cutinolytic activity and pathogenicity and PNB hydrolase activity could be improved if two of the mutants, M-29 and M-36, were disregarded . These data indicate that pathogenicity of C. lagenarium is not directly proportional to either of the enzyme activities observed, especially to PNB hydrolase . The correlation between cutinolytic activity and PNB hydrolase activity was also low (R = 0 . 71) (Fig . 3) indicating, as was found during purification of cutinolytic enzymes from C . lagenarium, that there is riot always a direct relationship between these two activities due to the presence of a cutinase that lacks PNB hydrolase activity [1] . Collelotrichum lagenarium and cucumber as a system to study cutin degrading enzymes and their role in penetration has turned out to be more complicated than first suspected . Until all the enzymes involved are isolated and characterized, it will be difficult to assign a definitive role for them in penetration . However, our initial work suggests their role in infection of cucumber by C . lagenarium is not as essential as that reported for F . solani f. sp . pisi [71 in the penetration of pea and C. gloeosporioides in penetration of papaya [3] . In the case of C. lagenarium, mechanical force may be a much more important factor, if riot the singular factor, involved in cuticular penetration by this fungus . This is supported by evidence from several groups investigating the importance of melanization of the appressoria in mechanical penetration . Nonmelanized appressoria of C . lagenarium, whether the lack of melanization was due to specific chemical inhibition (tricyclazole) or mutation, resulted in a 90-95 ",, reduction in the penetration of nitrocellulose membranes by this fungus [5, 9-11] . Microscopic examination showed that the in vim situation on cotyledons was similar to that on nitrocellulose filters . Thus, these authors concluded that appressorial pigmentation is essential for penetration by C . lagenarium . Recently, this group has suggested a role for
Role of cutinolytic enzymes in cucumber infection
481
cellulase in the infection process [5] . Wolkow et al . [12] and Woloshuk et al . [13] determined that melanin was important in appressoria for Colletotrichum lindemuthianum and Pyricularia oryzae, respectively . As observed in the present study with C . lagenarium, an organophosphate, (in this case DFP) had little effect on the disease caused by C . lindemuthianum and P . oryzae . As a result, these authors discounted the importance of cutinolytic enzymes in penetration and subsequent disease development for both C . lindemuthianum and P . oryzae . Although the question yet remains as to whether cutinolytic enzymes are important for many fungal pathogens in the penetration of their hosts, the evidence accumulating in recent studies suggests, at least for certain direct penetrating fungi, that these enzymes are less important than is mechanical force . It is possible that cutin degrading enyzmes play a minor role in reducing the force necessary for penetration by loosening the cutin . This is an activity which may not be essential or readily detectable .
REFERENCES
I.
A . M . & HAMMERSCHMIDT . R . (1989 ; . C utinolytic enzymes fi-om C5/lctobuluuu Ingcnarinm . Physiological and Molecular Plant Pathology 35, 463--47-1 . 2 . DANTZIC, A. H ., ZUCKERMAN, S . H . & ANDONOV-ROLAND, M . M . (1986) . Isolates of a Fusariumt solam mutant reduced in cutinase activity and virulence . journal of Bacteriology 168, 911--916 . 3 . DICKMAN, M . B ., PATIL, S . S . & KOLATTUKUDY, P . E . (1982) . Purification, characterization and role in infection of an extracellular cutinolytic enzyme from Colletotrichum gloeosporioides Penz . on Carica papaya L . Physiological Plant Pathology 20, 333-347 . 4 . DICKMAN, M . B . & PATIL, S . S . 11986) . Cutinase deficient mutants of Colletotrichum gloeosporioide.s arc nonpathogenic to papaya fruit . Physiological and Molecular Platt Pathology 28. 235 242 . 5 . KATOH, M ., HIROSE, I ., KuBo, Y . . HIKICHI, Y ., KUNOH, H ., I URUSAWA, 1 . & SHIZHIYAMA, J . ( 1988! . Use of mutants to indicate factors prerequisite for penetration of Colletotrichum lagenarium by appressoria . Physiological and Molecular Plant Pathology 32, 177--184 . 6 . KOLATTUKt :DY, P . E . (1987) . Lipid-derived defensive polymers and waxes and their role in plant microbe interactions . In Biochemistry of Plants, Vol . 7, Lipids . Structure and Function, Ed . by P. K . Stumpf pp . 291-314 . Academic Press, New York . 7 . KOI .LER, W ., ALLAN . G . R . & KOLATTUKUDY, P. E . (1982) . Protection of Pisum satiaum from hùsarium ,olani f. sp . pisi by inhibition ofcutinase with organophosphate pesticides . Phytopathology 72, 1425-1430 . 8 . KOLLER, W ., ALLAN, C . R . & KOLATTUKUDY, P . E . (1982) . Role of cutinase and cell wall degrading enzyme in infection of Pisum satiaum by Fusarium solani f. sp . pisi . Phaiological Plant Pathology 20, 47--60 . 9 . Kuso, Y ., SuzuKI, K ., FURUSAWA, I . & YAMAMOTO, M . 1982) . Relation of appressorium pigmentation and penetration of nitrocellulose membranes by Colletotrichum lagenarium . Phytopathology 72, 498-501 . 10 . KuBo . Y ., SuzuKI, K ., FukUSAwA, I . & YAMAMOTO, M . (1982': . Effect of tricyclazole on appressoria) pigmentation and penetration from appressoria oi ' Colletotrichum lagenarium . Phylopalhologc 72, 4198-1200 . 11- S1 ZUKI, K ., Kueo, Y., FUSUSAWA, I ., ISHIDA . N . & VAMAMUTO, M . '1982', . Behavior of colorless appressoria in an albino mutant of Colletotrichum lagenarium . Canadian Journal o/ _Microbiology 28 . 1210-1213 . 12 . AVorxow, P . M ., SISt.ER, H . I) . & Vicn., E. L . (1983 Efiict of inhibitors of melanin biosynthesis on structure and function of appressoria of Colletotrichum lindemnthianunn . Physiological Plant Pathology 23, 55--71 . l3 . WoLOSHUK . C . P ., SISI .ER, H . 1) . & Vtott., E . 1 . . ( 1983 .Action of the antipencu:utt, ti ioyi laud . on appressoria of Pvricularia or) . ae. Physiological Plant Pathology 22 . 245 259 . BONNEN,
475
Role of cutinolytic enzymes in infection of cucumber by Colletotrichum lagenarium A . M . BONNEN
and R .
j-
HAMMERSCHMIDT+
Department of Rolant and Plant Pathology, Vichigan State University, East Lansing, MI 48824-1312,
.A .S U .
(Accepted for publication August 1989)
The role of cutinolytic enzymes in infection of cucumber by Colletotrichum lagenarium was investigated . Microscopic and visual observations of C. lagenarium penetration and colonization in the presence of organophosphate cutinase inhibitors suggested that cutinolytic enzymes do not have a primary role in penetration . Use of mutants of C. lagenarium supported these observations since only a poor correlation between cutinase production and penetration and disease expression was found, whereas a very good correlation between penetration and disease development was found .
INTRODUCTION
Over the last 15 years, cutinolytic enzymes have been shown to be involved in the penetration of the cuticle by plant pathogenic fungi [6] . Although many fungi produce cutinases, it is still unclear how important these enzymes are for cuticular penetration for a broad range of fungal pathogens . Two lines of evidence have been used to support the role for cutinases in infection . The first is derived from the use of specific organophosphate inhibitors of cutinase to block infection . Application of the potent cutinase inhibitor diisopropylfluorophosphate (DFP) along with spores of Fusarium solani f. sp . pisi or Colletotrichum gloeosporioides prevented infection of the hosts by the pathogens [3, 6, 7] . Similar results were obtained if antibodies raised against the cutinases were included in spore suspensions [3, 6] . The use of mutants of C . gloeosporoides [4] and F. solani f. sp . pisi [2] or natural variants of F. solani f. sp . pisi [7] have provided genetic evidence for the role of cutinase . Research with C. lindemuthianum, C. lagenarium and Pyricularia oryzae, however, has suggested that mechanical penetration is the mechanism of penetration by these fungi [5, 9-13] . This was based on the observation that melanized, and, thus, structurally sound, appressoria are needed for direct penetration . In the case of C. lindemuthianum [12] and P . oryzae [13], DFP had no effect on penetration . Since C . lagenarium has been shown to need melanized appressoria for infection [2, 9-11], and we have demonstrated that this pathogen produces cutinolytic enzymes [1], the following study was carried out to try to determine the role of cutinases in the C . lagenarium-cucumber interactions . t Michigan Agricultural Experiment Station Journal Article No . 13118 . 1 To whom correspondence should be addressed .
0885-5765/89/120475+07 $03 .00/0
© 1989 Academic Press Limited
476
A. M . Bonnen and R . Hammerschmidt
MATERIALS AND METHODS
Culture offungi Colletotrichum lagenarium
(Pass .) E11 . and Halst . race 1 was maintained as previously
described [1] . d-Iutagenesis procedure A C . lagenarium spore
suspension of 1 . 75 x 106 spores ml_ i was exposed to 1 . 0 ° ethyl methane sulfonate in 0 . 05 M sodium phosphate buffer, pH 7 . 0 for 1 . 75 h resulting in 99°,, kill (as calculated from a kill curve) . The reaction was terminated by a 100 fold dilution with sterile distilled H 2 O (dH 2O) . The spore suspension was plated onto PDA containing 0 . 0001 % Terra Coat L-205 (Gustafson Chemical, Co .) (included for maintenance of discrete colonies) and incubated at 21 °C for 6 days . Terra Coat L-205 did not inhibit spore germination . Surviving colonies were transferred to PDA plates (without Terra Coat L-205) . Mutant growth and enzyme assay
Putative mutants were screened for growth on cutin and for p-nitrophenylbutyrate (PNB) hydrolase and cutinase activities . For these assays, each putative mutant was transferred to Nobel water agar and allowed to grow for 3-4 days . Plugs of mycelium (1 mm') from the advancing edge of each colony were transferred to 24 well tissue culture plates (Corning) containing 2 ml of the previously described minimal media plus cutin [1] in each well . Each plate contained two replicates of the wild type and three replicates were made of each plate . After 3 weeks the amount of growth in each well was scored . The culture fluid was assayed for PNB hydrolase and cutinase activity as previously described [1] . Pathogenicity test
The surface of cotyledons from 8-10 day old cucumber plants were inoculated with 6-8 droplets (20 pl per droplet) of a 106 spores ml -' suspension of C . lagenarium spores (wild type or mutant) . Inoculated plants were maintained at 100% for 24 h . When the cuticle was to be by-passed, cotyledons were inoculated by injection of spores via a syringe . Disease ratings were made every 24 h beginning 48 h after inoculation . Determination of_ percentage penetration
Cucumber hypocotyl segments, 2. 0-2 . 5 cm in length excised from 4-6 day old etiolated seedlings, were surface inoculated with 4-8 droplets (approximately 20 µl per droplet) of 10 6 spores ml - ' . Epidermal peels were taken at 48 h and stored in 100 % ethanol . Prior to microscopic examination the peels were rinsed in dH 2 O and stained with 0 . 1 toluidine blue (in 0 . 1 M sodium phosphate buffer, pH 6. 8) . Infection inhibition assays with paraoxon Stock solutions of 2 . 0 mm and 2 . 0
µM of paraoxon (O,O-diethyl-O-p-nitrophenylphosphate) were prepared in dH 2 O and stored at 4 ° C . Procedures for the pathogenicity studies were identical to those previously described except that the spores were suspended in 0 . 05 M sodium phosphate buffer, pH 7 . 0 containing the appropriate concentration of paraoxon .
477
Role of cutinolytic enzymes in cucumber infection RESULTS
Effect of paraoxon on pathogenecity and growth of C . lagenarium The radial growth of C . lagenarium was unaffected by any
of the concentrations of paraoxon used (0 .02, 0 . 2 and 2 . 0 µM) (Table 1) . Paraoxon also had no effect on the ability of the fungus to cause disease ('Table 1) . At a concentration of 2 . 0 µM symptom development was only slightly behind that observed for the control (no paraoxon) at 4 days and by 6 days was equal to the control . The ability of C. lagenarium to penetrate etiolated cucumber hypocotyls in the presence of paraoxon was also tested ("Table I ) . At the two highest concentrations, 0 . 2 µM and 2 . 0 pM, penetration was I l fect of paraoxon on
in vitro
ABLE
growth of C .
I lagenarium
and mJection
Paraoxon Concentration
growth" 1 nfection ` Penetration" Ire vitro
of cucumber
,11M ; °
0. 02
0.2
2 .0
96 100 81 '41)
95 100 62 (81)
97 100 70 (77)
Values in table are expressed as percent of water control . Radial growth on PDA containing indicated paraoxon concentration . Lesion formation on cucumber cotyledons . Expressed as percentage penetration from apprcscria as compared to control at 48 h alter noculation (value in parentheses is percentage of spores producing appressoria relative to on t rol) . reduced by approximately 30°x, . Percentage appressoria formed in the presence of paraoxon as compared to water was utilized as a control, as well as radial growth, in order to determine the effect of the inhibitor on the general metabolism of the fungus . Although radial growth was unaffected, the number of appressoria produced was significantly reduced (Table 1) . This result indicated that the inhibitor was not specific for the satin degrading enzymes, and as such, any reduction in penetration could not be taken as being strictly due to a loss in cutinolytic activity . The observation that the lowest concentration of paraoxon (0 . 02 1Mj resulted in the greatest reduction in number of appressoria produced was an unexpected but reproducible result .
Generation and characterization of mutants of C . Mutants of C . lagenarium were generated
lagenarium in attempts to help define the role of cutinolytic enzymes in the penetration of cucumber by this fungus . The first screens tested for growth on cutin (in liquid culture medium) as the sole carbon source and levels of PNB hydrolase activity per ml of culture fluid . Individuals exhibiting either increased or decreased levels of hydrolase activity or growth on the eutin substrate, relative to the wild-type, were chosen for further screening . In order to determine if reduced growth on cutin was a result of auxotrophy or of a general inability to metabolize carbon, those colonies exhibiting reduced growth on eutin, relative to the wild-type, were grown on minimal media [1] containing either sucrose or glucose . Those with reduced growth on these media were discarded .
478
A . M . Bonnen and R . Hammerschmidt
The remaining mutants were then checked fhr their cutinolytic activity and pathogenicity . In order to discern between those mutants which could not penetrate and those which were no longer able to colonize the tissue, all isolates unable to cause disease when surface inoculated were infiltrated into the tissue . Any isolates unable to cause disease when infiltrated were discarded . The radial growth of the 13 remaining mutants, on PDA, was found to be equal to or greater than the wild-type, with the exception of M-9, M-24 and M-30, which had 70° ;, the growth of the wild-type ('fable 2) . Each of the mutants was tested for : (I) PNB hydrolase activity ; (2) cutinolytic. activity ; (3) ability to penetrate etiolated cucumber hypocotyls and (4) ability to infect cucumber cotyledons (pathogencity) (Table 2) . All the mutants were able to produce melanized appressoria . Poor TABLE 2
Characteristics of wild type and mutant Colletotrichum lagenarium
Isolate Wild type M-4 M-9 M-11 M-18 M-20 M-24 M-25 M-29 ~1-30 M-31 M-36 M-37 M-42
Growths on cutin
PNB hydrolase activity'
Cutinolytic activity`
+ + +
0. 36 0. 38 0. 14 0-37 0. 24 0. 19 0. 04 0-23 0-53 0-05 0 . 21 0 . 15 0 . 25 0 . 22
7 14 . 5 4.6 9. 1 14. 2 11 . 2 0-5 10. 1 16 1 0. 8 7. 5 7-1 10 . 3 19 . 7
+ + + + + + + + + +
12 .
penetration ° 86 96 79 81 91 92 15 94 15 8 76 3 90 79
Pathogenicity 4-25 4. 75 2.5 3 . 25 3.5 445 0.0 4 . 75 0-0 0.0 4.3 0.0 4 . 75 5-5
- = no growth ; + + = good growth -t AA405 min -t ml Drnin-t ( x 10') 8 h -t ml -t ° percent of appressoria which had penetrated 48 h after inoculation disease scored at 4 days after inoculation : 0 = no symptoms ; 1 = small water soaked lesions at 25 °¢, or fewer inoculation sites ; 2 = water soaking at 25-50°0 of inoculation sites ; 3 = water soaking at all inoculation sites ; 4 = well developed lesions at all sites with necrotic centres ; 5 = lesions all necrotic ; 6 = lesions necrotic and coalesced .
correlations were found for the relationship between PNB hydrolase activity and pathogencity (R = 0 . 25) and between cutinolytic activity and pathogenicity (R = 0 . 56) . These results are presented in Table 1 and Fig . I a and 1 b . Two mutants, M-29 and M-36 exhibited very low pathogenicity but had wild type or higher levels of cutinase production, respectively . A highly significant correlation (R = 0 . 94) was found for percent penetration v . pathogenicity (Table 2, Fig . 1 C) . It is important to note that M-29 and M-36 did not exhibit high numbers of penetrations even though they did produce melanized appressoria and high levels of cutinase (Table 1) . A fair correlation (R = 0 . 71) was found between the PNB hydrolase activity v . cutinolytic activity (Table 2, Fig . 2) .
Role of cutinolytic enzymes in cucumber infection
(a)
r
479 b)
20
100
18
E 0.5
ô 16E a 04
∎
v
10 IE
u
40
6
30
4
20
M
c- 0-I t
50
a
i
a 0
r
•
8
Q0.2
70
o 60
Co
v
ô
o
∎
12
0 0.3
2
80
14
â0
(c 1
90
10 0
2
0$
.
1 2 3 4 5 6
1 2 3 4 5 6
1 2 3 4 5 6
Pathogenicity
Fin . 1 . Relationship between pathogenicity r . PNB hydrolase activity (a), vs . cutinolytic activity (b) or vs . penetration into host epidermal cells (c) . ( •, mutants ; /, wild type) . Data from Table 2 .
20 18 ô 16 14 ∎ 12
.
CC) 10 E
9
ao 0
8
.
6 4
0 0. 1
0 .3 0.2
0. 5 0.4
PNB hydrolase (40D405 min 1 ml -1 )
2 . Relationship between cutinolytic activity r . PNB hydrolase activity . • = mutants ; wild type . Data from Table 2 .
FIG .
DISCUSSION
Several distinct proteins or protein fractions from C . lagenarium have been shown to contain cutinolytic activity [1] . Whether any are involved in the penetration of cucumber by C. lagenarium is unknown . Studies utilizing a specific inhibitor of cutinase, paraoxon, were undertaken in attempts to determine the role of cutinases in penetration. Paraoxon, at all concentrations tested (0 . 02, 0 . 2 and 2 . 0 µM), failed to affect the radial growth of C . lagenarium on PDA . This presumably indicates normal metabolic functioning of the fungus . However, a reduction in the number of appressoria produced in the presence of paraoxon indicated that the inhibitor was not specific for the cutinolytic enzymes . Thus, the observed reduction in percent penetration of etiolated cucumber hypocotyls could not be attributed solely to the loss of cutin degrading activity . Despite the reduction in penetration efficiency, no reduction in disease was
480
A . M . Bonnen and R . Hammerschmidt
observed . These results are quite different from that reported for F. solani f. sp . pisi in which paraoxon severely reduced infection on pea hypocotyls without affecting the growth of the fungus [7] . This observation was used to support the involvement of cutinolytic enzymes in penetration by F. solani f. sp . pisi . It must now be taken as an indication that cutin degrading enzymes may not be involved (at least not as primary determinants) in penetration by C. lagenarium . Mutants of C. lagenarium were generated also in au attempt to determine the role of cutinase in the penetration of cucumber by this fungus . No isolates from the mutagenesis treatment were found which could not grow at all on cutin as the sole carbon source, although there were individuals with significantly reduced growth on the cutin . Two of these (M-24 and M-30) also had reduced levels of cutin degrading activity (Table 2) . These two were non-pathogenic unless infiltrated into the leaf tissue (thus by-passing penetration of the cuticle) . Two isolates, M-4 and M-42, with increased pathogenicity relative to the wild-type also showed incrased cutinolytic activity . This information might imply that cutinolytic enzymes arc important in penetration ; however, two other isolates were found which had reduced pathogenicity (M-29 and M-36) but with significant to very high levels of cutinolytic activity . These were also pathogenic only when infiltrated into the tissue . Correlation coefficients were determined for pathogenicity versus PNB hydrolase activity, cutinolytic activity and percent penetration for the 13 mutants and the parental wild type isolate. Of these three comparisons, only pathogenicity v . penetration was significantly correlated (R = 0 . 94) . The correlations between pathogenicity and cutinolytic activity and pathogenicity and PNB hydrolase activity could be improved if two of the mutants, M-29 and M-36, were disregarded . These data indicate that pathogenicity of C. lagenarium is not directly proportional to either of the enzyme activities observed, especially to PNB hydrolase . The correlation between cutinolytic activity and PNB hydrolase activity was also low (R = 0 . 71) (Fig . 3) indicating, as was found during purification of cutinolytic enzymes from C . lagenarium, that there is riot always a direct relationship between these two activities due to the presence of a cutinase that lacks PNB hydrolase activity [1] . Collelotrichum lagenarium and cucumber as a system to study cutin degrading enzymes and their role in penetration has turned out to be more complicated than first suspected . Until all the enzymes involved are isolated and characterized, it will be difficult to assign a definitive role for them in penetration . However, our initial work suggests their role in infection of cucumber by C . lagenarium is not as essential as that reported for F . solani f. sp . pisi [71 in the penetration of pea and C. gloeosporioides in penetration of papaya [3] . In the case of C. lagenarium, mechanical force may be a much more important factor, if riot the singular factor, involved in cuticular penetration by this fungus . This is supported by evidence from several groups investigating the importance of melanization of the appressoria in mechanical penetration . Nonmelanized appressoria of C . lagenarium, whether the lack of melanization was due to specific chemical inhibition (tricyclazole) or mutation, resulted in a 90-95 ",, reduction in the penetration of nitrocellulose membranes by this fungus [5, 9-11] . Microscopic examination showed that the in vim situation on cotyledons was similar to that on nitrocellulose filters . Thus, these authors concluded that appressorial pigmentation is essential for penetration by C . lagenarium . Recently, this group has suggested a role for
Role of cutinolytic enzymes in cucumber infection
481
cellulase in the infection process [5] . Wolkow et al . [12] and Woloshuk et al . [13] determined that melanin was important in appressoria for Colletotrichum lindemuthianum and Pyricularia oryzae, respectively . As observed in the present study with C . lagenarium, an organophosphate, (in this case DFP) had little effect on the disease caused by C . lindemuthianum and P . oryzae . As a result, these authors discounted the importance of cutinolytic enzymes in penetration and subsequent disease development for both C . lindemuthianum and P . oryzae . Although the question yet remains as to whether cutinolytic enzymes are important for many fungal pathogens in the penetration of their hosts, the evidence accumulating in recent studies suggests, at least for certain direct penetrating fungi, that these enzymes are less important than is mechanical force . It is possible that cutin degrading enyzmes play a minor role in reducing the force necessary for penetration by loosening the cutin . This is an activity which may not be essential or readily detectable .
REFERENCES
I.
A . M . & HAMMERSCHMIDT . R . (1989 ; . C utinolytic enzymes fi-om C5/lctobuluuu Ingcnarinm . Physiological and Molecular Plant Pathology 35, 463--47-1 . 2 . DANTZIC, A. H ., ZUCKERMAN, S . H . & ANDONOV-ROLAND, M . M . (1986) . Isolates of a Fusariumt solam mutant reduced in cutinase activity and virulence . journal of Bacteriology 168, 911--916 . 3 . DICKMAN, M . B ., PATIL, S . S . & KOLATTUKUDY, P . E . (1982) . Purification, characterization and role in infection of an extracellular cutinolytic enzyme from Colletotrichum gloeosporioides Penz . on Carica papaya L . Physiological Plant Pathology 20, 333-347 . 4 . DICKMAN, M . B . & PATIL, S . S . 11986) . Cutinase deficient mutants of Colletotrichum gloeosporioide.s arc nonpathogenic to papaya fruit . Physiological and Molecular Platt Pathology 28. 235 242 . 5 . KATOH, M ., HIROSE, I ., KuBo, Y . . HIKICHI, Y ., KUNOH, H ., I URUSAWA, 1 . & SHIZHIYAMA, J . ( 1988! . Use of mutants to indicate factors prerequisite for penetration of Colletotrichum lagenarium by appressoria . Physiological and Molecular Plant Pathology 32, 177--184 . 6 . KOLATTUKt :DY, P . E . (1987) . Lipid-derived defensive polymers and waxes and their role in plant microbe interactions . In Biochemistry of Plants, Vol . 7, Lipids . Structure and Function, Ed . by P. K . Stumpf pp . 291-314 . Academic Press, New York . 7 . KOI .LER, W ., ALLAN . G . R . & KOLATTUKUDY, P. E . (1982) . Protection of Pisum satiaum from hùsarium ,olani f. sp . pisi by inhibition ofcutinase with organophosphate pesticides . Phytopathology 72, 1425-1430 . 8 . KOLLER, W ., ALLAN, C . R . & KOLATTUKUDY, P . E . (1982) . Role of cutinase and cell wall degrading enzyme in infection of Pisum satiaum by Fusarium solani f. sp . pisi . Phaiological Plant Pathology 20, 47--60 . 9 . Kuso, Y ., SuzuKI, K ., FURUSAWA, I . & YAMAMOTO, M . 1982) . Relation of appressorium pigmentation and penetration of nitrocellulose membranes by Colletotrichum lagenarium . Phytopathology 72, 498-501 . 10 . KuBo . Y ., SuzuKI, K ., FukUSAwA, I . & YAMAMOTO, M . (1982': . Effect of tricyclazole on appressoria) pigmentation and penetration from appressoria oi ' Colletotrichum lagenarium . Phylopalhologc 72, 4198-1200 . 11- S1 ZUKI, K ., Kueo, Y., FUSUSAWA, I ., ISHIDA . N . & VAMAMUTO, M . '1982', . Behavior of colorless appressoria in an albino mutant of Colletotrichum lagenarium . Canadian Journal o/ _Microbiology 28 . 1210-1213 . 12 . AVorxow, P . M ., SISt.ER, H . I) . & Vicn., E. L . (1983 Efiict of inhibitors of melanin biosynthesis on structure and function of appressoria of Colletotrichum lindemnthianunn . Physiological Plant Pathology 23, 55--71 . l3 . WoLOSHUK . C . P ., SISI .ER, H . 1) . & Vtott., E . 1 . . ( 1983 .Action of the antipencu:utt, ti ioyi laud . on appressoria of Pvricularia or) . ae. Physiological Plant Pathology 22 . 245 259 . BONNEN,