An alternative insect pathogenic strategy in an Aspergillus flavus auxotroph

An alternative insect pathogenic strategy in an Aspergillus flavus auxotroph

mycological research 113 (2009) 230–239 journal homepage: An alternative insect pathogenic strategy in an Aspergillus...

791KB Sizes 4 Downloads 29 Views

mycological research 113 (2009) 230–239

journal homepage:

An alternative insect pathogenic strategy in an Aspergillus flavus auxotroph Lisa R. SCULLY, Michael J. BIDOCHKA* Department of Biological Sciences, Brock University, 500 Glenridge Avenue, St. Catharines, Ontario, Canada L2S 3A1

article info


Article history:

In order to study fungal pathogen evolution, we used a model system whereby the oppor-

Received 5 March 2008

tunistic fungus Aspergillus flavus was serially propagated through the insect (Galleria mello-

Received in revised form

nella) larvae, yielding a cysteine/methionine auxotroph of A. flavus with properties of an

10 September 2008

obligate insect pathogen. The auxotroph exhibited insect host restriction but did not

Accepted 10 October 2008

show any difference in virulence when compared with the wild-type (Scully LR, Bidochka

Corresponding Editor:

MJ, 2006. Microbiology 152, 223–232). Here, we report that on 1 % insect cuticle medium and

Richard A. Humber

synthetic Galleria medium, the auxotroph displayed increased extracellular protease production, a virulence factor necessary for insect pathogenesis. In the wild-type strain,


protease production was deregulated during carbon (glucose), nitrogen (nitrate), or

Aspergillus flavus

sulphate deprivation. If all three were present, protease production was vastly reduced.

Cysteine/methionine auxotroph

However, in the cysteine/methionine auxotroph, protease production was deregulated in

Galleria mellonella

complete medium. We suggest that the deficiency in sulphate assimilation in the auxo-

Pathogen evolution

troph resulted in deregulation of protease production. The auxotroph exhibited delayed


germination and slower hyphal growth when compared to the wild-type but there were

Sulfur acquisition

no differences in virulence or cuticle penetration, suggesting a shift in pathogenic strategy


that compensated decreased growth with increased virulence factor (extracellular prote-

Virulence factor

ase) production. We concluded that the biosynthetic deficiency that mediated insect host restriction also increased protease production in the slow-growing auxotroph, resulting in an alternate, more host-specific pathogenic strategy. However, we argue that transmission is not necessarily correlated with virulence as competition bioassays in insect larvae showed that the wild-type generally out-competed the auxotroph by producing the majority of the conidia on the sporulating cadavers. This is one of the few examples that highlight the effect of genome decay on nutrition acquisition, virulence, and transmission in fungal pathogen evolution. ª 2008 The British Mycological Society. Published by Elsevier Ltd. All rights reserved.

Introduction There are many routes by which a pathogen may evolve host specialization (Kitano 2007). Although there are several experimental and evolutionary models for bacterial and viral pathogens there is a paucity of information regarding the

evolution of fungal pathogens (Marques & Carthew 2007; Ochman & Moran 2001; Pallen & Wren 2007). In order to study fungal pathogen evolution, we employed a model pathogen–host system involving the serial propagation of the opportunistic fungus Aspergillus flavus through the insect host Galleria mellonella larvae (Scully & Bidochka 2005). A.

* Corresponding author. Tel.: þ1 905-688-5550x3392. E-mail address: [email protected] 0953-7562/$ – see front matter ª 2008 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.mycres.2008.10.007

An alternative insect pathogenic strategy

flavus and G. mellonella are ideal species for such a model system. A. flavus is an opportunistic pathogen that exhibits a broad host range yet with low virulence (Yu et al. 2005). G. mellonella larvae are susceptible to fungal infection, easily manipulated, and available in large numbers. Furthermore, various aspects of the insect’s innate immune system and the infection response render pathogen studies using insects as hosts applicable to human pathogenic microbes (Scully & Bidochka 2006b). During serial passage, we isolated a strain of A. flavus that demonstrated a severe diminution of conidial production or complete lack of growth on microbiological media, while maintaining adequate conidial production at the culmination of the pathogenic cycle of insects. Although this strain remained pathogenic on insects with no decrease in virulence toward G. mellonella larvae compared with the wild-type, it failed to infect or conidiate sufficiently on plants (Scully & Bidochka 2006a). Essentially, this strain of A. flavus exhibited characteristics of an obligate insect pathogen (Scully & Bidochka 2006a). Biochemical analysis characterized it as a cysteine/methionine auxotroph, deficient in the ability to convert sulphate to sulphite. This led us to hypothesize that this nutrient deficiency, which is overcome during growth on insects, effectively caused a degree of host restriction, which is, in turn, a notable feature of pathogen evolution. We postulated that host restriction due to a biosynthetic deficiency may be an important step of evolution towards obligate pathogenicity as it may provide the physiological backdrop for the selection pressure required to develop further host specialization and increased virulence (Scully & Bidochka 2006a). Here, we report a pleiotrophic effect of cysteine/methionine auxotrophy on virulence factor production. When grown on media that mimic the conditions experienced during insect pathogenesis, the insect-restricted auxotroph exhibited enhanced protease production. Insect-pathogenic fungi normally infect the host by penetrating the insect cuticle thus gaining access to the nutrient-rich haemolymph (Clarkson & Charnley 1996). Extracellular proteases, particularly subtilisin-like proteases, are important virulence factors secreted by facultative insect pathogens, which degrade the proteinaceous insect cuticle, thus facilitating hyphal penetration and insect infection (St Leger 1995). Proteases can also assist in successful pathogen establishment in the haemocoel by cleaving proteins involved in insect immune system-signalling pathways (Griesch & Vilcinskas 1998). We surmised that in the insect-restricted strain, the enhanced protease secretion may have occurred due to cysteine/methionine auxotrophy, as fungi deregulate extracellular proteases in response to carbon, nitrogen, or sulphur starvation (Cohen 1973; Hanson & Marzluf 1975; Srinivasan & Dhar 1990). Here we hypothesize that increased protease production of the auxotroph offsets its slow growth during insect pathogenesis, allowing indistinguishable cuticle penetration and virulence toward G. mellonella larvae. The results highlight a biosynthetic nutrient deficiency as a mechanism of both host restriction and enhanced virulence factor production, signifying the evolution of an alternate pathogenic strategy in a fungal pathogen.


Materials and methods Fungal strains and culture conditions Aspergillus flavus 6982 (Af6982) was obtained from the University of Alberta Microfungus Collection and Herbarium (UAMH). A. flavus 6982conins (Af6982conins) was derived from Af6982 as previously described (Scully & Bidochka 2006a) and was deposited with UAMH (accession number 10904). Strains were maintained on PDA (potato dextrose agar, Difco, Oakville, ON).

Biomass determination Fungi were subcultured onto synthetic Galleria medium (SGM) agar (Dunphy & Webster 1986) or 1 % (w/v) cuticle agar. Powdered cuticles from Locusta sp. (courtesy of R. Meldrum Robertson, Queen’s University, ON), was prepared as described by Bidochka & Khachatourians (1990). SGM is intended to mimic nutritional conditions within the insect haemolymph, whereas cuticle medium is intended to mimic conditions when the fungus penetrates the cuticle. Fungal colonies were photographed after 7 d at 30  C, and biomass was determined as previously described (Scully & Bidochka 2005) as the mean of three replicates.

Germination and hyphal growth rate SGM agar or 1 % cuticle agar plates were inoculated with 1  104 conidia in a volume of 100 ml and incubated at 30  C. At 1 h intervals from 6–12 h following inoculation (4 h intervals from 10–30 h following inoculation for Af6982conins on 1 % cuticle), sections of the agar were removed, stained with lactophenol cotton blue, and examined microscopically. For each time interval, 100 conidia were assessed for percentage germination and average hyphal length. Conidia were considered germinated if they possessed a germ tube visible at 400 magnification. Median germination time (GT50: time to 50 % germination, in hours) was calculated using Probit analysis with statistical significance determined by a lack of overlap between 95 % confidence intervals. Hyphal growth rate was calculated as D log10 (average hyphal length (mm))/D t (h). Three trials were performed.

Enzyme assays in insect medium Flasks of 1 % (w/v) Locusta sp. cuticle (sterilized in dH2O) or SGM liquid medium (25 ml) containing 2  106 conidia ml1 were incubated at 30  C, 250 rev min1 for 18 h. Extracellular filtrates were collected, filtered through 0.2 mm filters, and assayed for subtilisin-like protease activity toward succinyl-(alanyl)2prolyl-phenylalanine p-nitroanilide (Suc-Ala-Ala-Pro-Phe-NA) (Sigma, Oakville, ON), and trypsin-like activity toward N-benzoyl-phenylalanyl-valyl-arginine p-nitroanilide (Bz-Phe-ValArg-NA) (Sigma) (St Leger et al. 1987). For total protease activity, 100 ml of filtrate were incubated with 1 mg hide powder azure (Sigma), 100 ml of 1 M NaOH–glycine buffer (pH 8.5), and 300 ml ddH2O at 37  C, 180 rev min1 for 3.5 h. Absorbance was measured at 630 nm. Assays were performed in triplicate, and analysed using t-tests (a ¼ 0.05).


Enzyme assays in minimal medium Fungi grown in 100 ml YPD broth (0.2 % yeast extract, 1 % peptone, 2 % dextrose) for 3 d, 250 rev min1 at 30  C were collected, washed with sterile dH2O, and ca 2 g wet weight was transferred to 40 ml of one of the following media: minimal medium [MM; 0.036 % (w/v) KH2PO4, 0.093 % Na2HPO4, 0.1 % KCl, 0.06 % MgSO4$7H2O, 0.15 % NaNO3, 1 % dextrose, 0.8 % elastin], MM lacking a carbon source (dextrose; MM-C), MM lacking a nitrogen source (NaNO3; MM-N), MM lacking a sulphur source (MgSO47$H2O; MM-S), and MM lacking MgSO47$H2O but supplemented with 10 mM methionine (MM þ met). The cultures were incubated 18 h, 250 rev min1 at 30  C before the extracellular filtrates were assayed for subtilisin-like, trypsin-like, and total protease activity. Assays were performed in triplicate and independently duplicated.

Isoelectric focusing (IEF) Extracellular filtrates of cultures grown in YPD and transferred to MM-C, MM-N, MM-S, MM, MM þ met, 1 % cuticle, or SGM for 18 h, 250 rev min1 at 30  C were frozen and lyophilized. Residual solids were dissolved in 500 ml ddH2O (4  C), desalted, and concentrated to <50 ml using Nanosep 10 K Omega (Pall, Mississauga, ON) centrifugal devices according to the manufacturer’s instruction. Samples were subjected to analytical IEF (pH 3–10) using Model 111 Mini IEF Cell (BioRad, Mississauga, ON) according to the manufacturer’s protocol. Proteolytic activity was detected by gelatin zymography or enzyme overlay membranes (EOM) impregnated with Suc-(Ala)2-Pro-Phe-7-amino-4-trifluoromethylcoumarin (MP Biomedicals, Solon, OH) as previously described (St Leger et al. 1994; Smith 1984).

Adherence and appressorial formation Conidial adherence was assessed on polystyrene Petri dishes, which mimics the hydrophobic insect cuticle (Butt 1987). Conidia were washed with ddH2O, suspended in 5 ml of 0.0125 % yeast extract (5  104 conidia ml1), and incubated at 30  C on 5.5 cm diam dishes for 16 h without movement. The Petri dishes were prepared in duplicate with one subjected to a brief jet of water. Percentage adherence was determined by comparing the number of adhered conidia in 15 fields of view of the washed and unwashed dishes. Experiments were performed in triplicate. Percentage appressorial formation was assessed on isolated Galleria mellonella (Peterborough Live Bait, Peterborough, ON) larval cuticles. Decapitated larvae were dissected along the ventral midline, body cavity contents were removed, and cuticles were washed in sterile dH2O. The dorsal surfaces of the cuticles were inoculated with 5  103 conidia, incubated at 30  C for 24 h, and then stained with 0.01 % Calcofluor (Fluorescent Brightener 28, Sigma) (Butt 1987) and viewed and photographed with a Leitz Diaplan microscope with an epifluorescence attachment. Fifty germinated conidia were assessed for percentage appressorial formation in each of three replicates. The interior surfaces of the cuticles were also examined microscopically to determine penetration through the cuticle.

L. R. Scully, M. J. Bidochka

Differences in percentage adherence and appressorial formation were determined using Chi-square tests (a ¼ 0.05).

Competition assay Sets of 25 Galleria mellonella larvae were topically infected (5 ml of 1  105 conidia ml1) with various ratios (10:0, 9.4:0.6, 8.8:1.2, 8.1:1.9, 7.3:2.7, 6.5:3.5, 5.5:4.5, 4.4:5.6, 3.1:6.9, 1.7:8.3, 0:10) of Af6982:Af6982conins conidia as previously described (Scully & Bidochka 2005). Insects infected with Aspergillus flavus, which subsequently died, showed mycelial outgrowth and conidiation on the insect surface; essentially the dead insects were mummified by the fungus. Conidia from the mummified insects were extracted into suspension as previously described (Scully & Bidochka 2005). Petri dishes of PDA and Cz (Czapek’s solution agar, Difco) were each inoculated with ca 100 conidia from suspensions from a conidiating insect cadaver. Af6982 colonies appeared on both media, whereas Af6982conins colonies appeared only on PDA (Scully & Bidochka 2006a). After 2 d at 30  C, colony forming units (CFU) were counted and statistically significant differences between the number of CFU on PDA and Cz were determined (t-tests, a ¼ 0.05). A statistically equivalent number of CFUs on PDA and Cz indicated a preponderance of Af6982, whereas a statistically significant difference in the number of CFU on PDA and Cz indicated the presence of both Af6982 and Af6982conins. When a significant difference was obtained, a Chi-square test (a ¼ 0.05) was employed to determine whether the proportion of Af6982 : Af6982conins CFU differed from the ratio in the original inoculum. As controls, suspensions of various ratios (10:0, 8.8:1.2, 6.3:3.7, 3:7, 0:10) of Af6982 : Af6982conins conidia were prepared, and ca 100 conidia of each suspension were plated on PDA and Cz. After 2 d at 30  C, the CFU were counted. Chi-square tests (a ¼ 0.05) were performed to ascertain whether the number of CFU approximated the ratios of the original suspensions.

Results Af6982conins germination and growth is stunted on cuticle medium Af6982 and Af6982conins displayed different characteristic growth patterns on 1 % cuticle agar and SGM agar, both of which mimic the nutritional environment encountered during infection of insects. Although Af6982 grew moderately on 1 % cuticle agar after 7 d, Af6982conins produced only microscopic amounts of growth (Fig 1A). On SGM agar, both strains grew (Fig 1A), but colony biomass after 7 d was significantly less for Af6982conins (0.056  0.01 g) than for Af6982 (0.148  0.008 g; t ¼ 7.145, P ¼ 0.002). Similarly, Af6982 germinated significantly faster than Af6982conins and exhibited a faster hyphal growth rate on 1 % cuticle agar (t ¼ 7.702, P ¼ 0.005; Fig 1B, Table 1). Conversely, there was no difference in the GT50 values on SGM agar. Hyphal growth rates on SGM agar were statistically different (t ¼ 3.667, P ¼ 0.021), but the discrepancy was not as great as on 1 % cuticle agar (Fig 1B, Table 1).

An alternative insect pathogenic strategy


Fig 1 – Hyphal growth and protease production of Af6982 and Af6982conins in 1 % cuticle agar and SGM agar. (A) Colonies grown on agar for 7 d at 30  C. Af6982conins produced only microscopic amounts of growth on 1 % cuticle agar (see inset). (B) Microscopic examination of germinating conidia after 18 or 24 h at 30  C. (C) Subtilisin-like, trypsin-like, or total protease activity in extracellular filtrate of Af6982 (white) or Af6982conins (hatched) grown in 1 % cuticle liquid medium or SGM liquid medium for 18 h at 30  C. Data represent mean values of three trials. Asterisks represent statistically significant differences (t-tests, P < 0.05).

Table 1 – Median germination time (GT50: time to 50 % germination, in hours) with 95 % confidence intervals (CI95) and hyphal growth rate of Af6982 and Af6982conins on 1 % cuticle agar or synthetic Galleria medium (SGM) agar at 30  C Strain

Af6982 Af6982conins

1 % Cuticle


GT50a (CI95)

Hyphal growth rate  S.E. (h1)

GT50a (CI95)

Hyphal growth rate  S.E. (h1)

7.9 (7.58, 8.23) 21.25 (20.1, 22.46)

0.15  0.006 0.071  0.009

6.69 (6.49, 6.9) 6.89 (6.59, 7.2)

0.241  0.009 0.197  0.007

Values represent an average of three independent trials. a GT50 was calculated using Probit analysis, and hyphal growth rate was determined as D log10 (average hyphal length (mm))/ D t (h).


L. R. Scully, M. J. Bidochka

Af6982conins protease production enhanced in insect medium During growth in 1 % cuticle liquid medium, Af6982 and Af6982conins secreted proteases, including subtilisin-like and trypsin-like proteases. However, during growth in SGM, protease production by both strains was relatively lower (Fig 1C). In both media, Af6982conins exhibited significantly greater amounts of subtilisin-like protease (cuticle t ¼ 5.683, P < 0.001; SGM t ¼ 2.661, P ¼ 0.029), trypsin-like protease (cuticle t ¼ 3.187, P ¼ 0.007; SGM t ¼ 3.761, P ¼ 0.004), and total protease (cuticle t ¼ 5.464, P < 0.001; SGM t ¼ 4.368, P < 0.001) activity than Af6982 (Fig 1C). There were no differences in the types of proteases produced by Af6982 and Af6982conins in cuticle medium or SGM when compared using IEF (Fig 2). However, Af6982conins showed qualitatively more subtilisin-like protease in 1 % cuticle than Af6982, supporting the data in Fig 1C.

Af6982conins requires methionine to downregulate protease production Af6982 and Af6982conins produced proteases, including subtilisin-like and trypsin-like proteases, when grown in minimal medium lacking carbon (glucose), nitrogen (nitrate), or sulphur (sulphate) sources, but to varying degrees (Fig 3). Whereas Af6982 downregulated protease production when supplied with sulphate in addition to glucose and nitrate (i.e. MM), Af6982conins did not. However, when MM was supplemented with methionine, both Af6982 and Af6982conins downregulated protease production (Fig 3). Furthermore, protease production by Af6982conins in MM without sulphur was similar to protease production in MM, suggesting a deficiency in sulphate assimilation in Af6982conins with a consequent deregulation of protease. Qualitatively, Af6982conins produced more overall protease than Af6982 (Fig 4). No extracellular proteases were constitutively expressed. The downregulation or upregulation of all detectable proteolytic bands were equally affected by catabolite repression suggesting that extracellular protease regulation appears to be a global phenomenon (Fig 4).

Af6982 and Af6982conins demonstrate identical adherence and appressorial formation Aspergillus flavus conidia adhered to and germinated on the surfaces of Petri dishes in the presence of 0.0125 % yeast extract with no significant difference in percentage adherence of Af6982 (16.2  4.2 %) and Af6982conins (20.6  1.8 %; c2 ¼ 3.366; P ¼ 0.067). Furthermore, there was no significant difference in percentage appressorial formation by Af6982 (44  2 %) and Af6982conins (43.4  6.4 %) on larval cuticle (Fig 5A and B; c2 ¼ 0.011; P ¼ 0.918). Both strains penetrated the cuticle and grew through to the interior surface of the cuticle (Fig 5C and D). Cuticle penetration occurred at darkened lesions (Fig 5C and D) indicative of a melanization reaction stemming from the prophenyloxidase defence cascade (Kanost et al. 2004).

Af6982 outcompetes Af6982conins for the insect cadaver Cz and PDA were used to distinguish Af6982 and Af6982conins as Af6982 CFU form on both media, whereas only PDA

Fig 2 – Analytical IEF (pH 3–10) of extracellular proteases from culture filtrates of Af6982 and Af6982conins grown in 1 % cuticle or SGM at 250 rev minL1 at 30  C for 18 h. Gelatin zymography with X-ray film overlay for (A) 30 min and (B) 2 h at 37  C. (C) Enzyme overlay membrane (EOM) impregnated with subtilisin-like protease substrate succinyl-(alanyl)2-prolyl-phenylalanine-7-amino-4-trifluoromethylcoumarin overlay for 3 h at 37  C.

supports Af6892conins CFU (Scully & Bidochka 2006a). When various ratios of Af6982 : Af6982conins conidia were plated on Cz and PDA, there were no statistically significant differences between the ratio of CFU on Cz to the CFU on PDA minus CFU on Cz [CFUCZ: (CFUPDA–CFUCZ)] and the initial ratio plated according to Chi-square tests (10:0, c2 ¼ 0.09, P ¼ 0.764; 8.75:1.25, c2 ¼ 0.044, P ¼ 0.83; 6.25:3.75, c2 ¼ 2.38,

An alternative insect pathogenic strategy

Fig 3 – Extracellular protease activity from culture filtrates of Af6982 (white) and Af6982conins (hatched) grown in minimal medium (MM); MM lacking carbon (MM-C); MM lacking nitrogen (MM-N); MM lacking sulphur (MM-S); and MM without sulphate but supplemented with 10 mM methionine (MM D met) at 250 rev minL1, 30  C for 18 h. Results represent two independent trials of assays repeated in triplicate.

P ¼ 0.123; 3:7, c2 ¼ 0.46, P ¼ 0.498; 0:10, c2 ¼ 0, P ¼ 1). Therefore, CFUCZ:(CFUPDA–CFUCZ) is a good approximation of Af6982:Af6982conins. Sets of 25 Galleria mellonella larvae were infected with various ratios of Af6982:Af6982conins. A total of 39 insect cadavers that had been infected with various ratios of Af6982:Af6982conins (Table 2) were analysed. Twenty-eight of the cadavers showed no statistically significant difference in CFUCZ: (CFUPDA–CFUCZ) (t-tests, P  0.05), indicating virtually only wild-type conidia. The remaining 11 cadavers exhibited a statistically significant difference in CFUCZ: (CFUPDA–CFUCZ) (t-tests, P < 0.05), indicative of conidia from both strains (Table 2). Ten cadavers demonstrated a statistically significant change in the ratio Af6982:Af6982conins conidia; five had a greater proportion of Af6982 than the initial ratio, whereas five had a greater proportion of Af6982conins than the original inoculum (Table 2). Interestingly, one cadaver initially


Fig 4 – Analytical IEF (pH 3–10) of extracellular proteases from culture filtrates of Af6982 and Af6982conins grown in minimal medium (MM); MM lacking carbon (MM-C); MM lacking nitrogen (MM-N); MM lacking sulphur (MM-S); and MM lacking sulphate but supplemented with 10 mM methionine (MM D met) at 250 rev minL1, 30  C for 18 h. Gelatin zymography with X-ray film overlay for (A) 30 min and (B) 2 h at 37  C. (C) Enzyme overlay membrane (EOM) impregnated with subtilisin-like protease substrate succinyl-(alanyl)2-prolyl-phenylalanine-7-amino-4-trifluoromethylcoumarin overlay for 3 h at 37  C.

infected with a 1.7:8.3 ratio of Af6982:Af6982conins produced only Af6982conins conidia (Table 2).

Discussion In order to study pathogen evolution, we employed a model pathogen–host system involving the serial passage of the opportunistic pathogen Aspergillus flavus through Galleria mellonella larvae (Scully & Bidochka 2005). From this scheme, we isolated a strain of A. flavus (Af6982conins) that exhibited traits of an obligate insect pathogen, specifically a drastic reduction in conidial production on PDA and a restricted host range as it continued to infect insects with no alteration in virulence but failed to infect and/or conidiate sufficiently on plants. This phenotype was associated with cysteine/methionine auxotrophy. We hypothesized that an obligate pathogen might evolve


L. R. Scully, M. J. Bidochka

Fig 5 – Appressoria (ap) differentiated from germinated conidia (co) of (A) Af6982 and (B) Af6982conins on the surface of Galleria mellonella larval cuticle incubated 24 h at 30  C. Emergent hyphae of (C) Af6982 and (D) Af6982conins on the ventral surface of the cuticle following penetration.

when a nutrient deficiency restricts it to a particular host, providing selection pressure for an increase in virulence (Scully & Bidochka 2006a). Here, we report that this cysteine/methionine auxotrophy has the pleiotrophic effect of increasing virulence factor production (i.e. protease deregulation via sulphur limitation), allowing this slow-growing auxotroph to maintain virulence toward insects. These findings signify the development of an alternate pathogenic strategy divergent from that of the wild-type, providing insight into fungal pathogen evolution. Microbial (bacterial and fungal) pathogenesis can be resolved into a series of events. These events include adhesion, invasion of the host, avoidance of host immune responses, dissemination within the host, and transmission from the host. Fungal pathogens of insects, such as Metarhizium anisopliae, infect the host insect by trans-cuticle penetration. Extracellular proteases act as virulence factors; they degrade cuticle proteins and facilitate invasion of the host insect where the fungus gains access to the nutrient rich haemolymph (St Leger 1995). Additionally, proteases assist successful pathogen establishment in the haemocoel by cleaving proteins involved in insect immune system-signalling pathways (Griesch & Vilcinskas 1998), as well as tissues of the haemolymph. Our strains of A. flavus also penetrated the insect cuticle and produced extracellular proteases similar to those of M. anisopliae, particularly the subtilisin-like proteases. On a biomass basis, Af6982conins secreted significantly higher levels of extracellular subtilisin-like, trypsin-like, and total protease than the wild-type in 1 % insect cuticle and SGM. These substrates/ media were intended to mimic, respectively, either the initial stages of insect pathogenesis when the fungus penetrates the cuticle or the latter stages when the fungus grows extensively in the haemolymph (Clarkson & Charnley 1996). Protease overproduction by Af6982conins presumably occurred through protease deregulation as this strain is a cysteine/methionine auxotroph, incapable of assimilating sulphate as a sulphur source (Scully & Bidochka 2006a). Extracellular protease secretion from Neurospora crassa, A. nidulans, and A. flavus is regulated in response to carbon, nitrogen, or sulphur limitation (Cohen 1973; Hanson & Marzluf 1975; Srinivasan & Dhar 1990); starvation of any one of these components results in protease deregulation. In N. crassa, protease production is controlled by three independent regulatory circuits responding to the carbon, nitrogen, or sulphur nutrient levels (Hanson & Marzluf 1975; Dowzer & Kelly 1991; Hynes & Kelly 1977; Fu & Marzluf 1990; Kanaan &

Marzluf 1993; Hanson & Marzluf 1973). Likewise, Af6982 and Af6982conins produced protease under carbon, nitrogen, or sulphur-limiting conditions. In conjunction with glucose and nitrate, either sulphate or methionine repressed protease production by Af6982, whereas only methionine sufficed for Af6982conins. We suggest that the cysteine/methionine auxotroph, Af6982conins, is starved for sulphate, which results in deregulated protease production in 1 % insect cuticle or SGM. In entomopathogenic fungi, protease production responds to carbon and nitrogen deprivation experienced during germination and growth on the insect cuticle surface (Paterson et al. 1994; St Leger et al. 1988; Screen et al. 1997; Screen et al. 1998; St Leger et al. 1991). Secreted proteases, particularly the subtilisin-like proteases, act as virulence factors by degrading cuticle proteins, allowing the fungus to infect, colonize, and subsequently kill the insect (St Leger 1995). Indeed, conidial virulence is correlated concomitantly with both nutrient deprivation (low endogenous C:N ratio) and greater amounts of protease transcripts (Shah et al. 2005). Because insect infection by A. flavus and Metarhizium anisopliae is analogous (Clarkson & Charnley 1996; Kumar et al. 2004), this link between starvation, protease production, and insect pathogenesis presumably also occurs in A. flavus and suggests that enhanced protease secretion due to auxotrophy has implications regarding the evolution of pathogenesis. The cysteine/methionine auxotroph showed depressed rates of germination and hyphal growth on a variety of media including insect cuticle and SGM. This is not unexpected, as, by definition, auxotrophs display stunted growth in culture medium containing suboptimal amounts of the required supplement(s). Nevertheless, Af6982conins remained as virulent as the wild-type (Scully & Bidochka 2006a). We suggest that the higher levels of secreted protease compensate for the stunted growth of Af6982conins. The inverse correlation between growth and protease production in Af6982 and Af6982conins indicates the evolution of alternate pathogenic strategies of equal virulence (Fig 6). As cuticle penetration is achieved by both mechanical pressure and proteolytic action (Goettel et al. 1989), Af6982 may rely more heavily on mechanical pressure afforded by its greater mycelial biomass, whereas Af6982conins may depend on protease deregulation (Fig 6). Consequently, there were no differences in cuticle penetration by Af6982 and Af6982conins. There were also no differences in adherence or appressorium formation, factors known to affect insect pathogenesis (Butt et al. 1995). Establishment within the insect haemocoel may be mediated

An alternative insect pathogenic strategy


Table 2 – Competition assays between Af6982 and Af6982conins on Galleria mellonella larvae Initial ratio (Af6982:Af6982conins)

No. of insects examineda

No. of insects: With Af6982 conidia onlyb

With Af6982 and Af6982conins conidiac

10:0 9.4:0.6 8.8:1.2

2 5 4

2 5 2

0 0 2

8.1:1.9 7.3:2.7 6.5:3.5 5.5:4.5 4.4:5.6

4 4 4 4 8

3 3 3 4 5

1 1 1 0 3

3.1:6.9 1.7:8.3

2 4

1 2

1 2






Final ratio (Af6982:Af6982conins)

NA NA 162:122 119:65 175:47 124:107 153:121 NA 109:87 90:53 126:49 125:62 100:107 0:272 0:313 0:228 0:195 0:235

NA NA 5.7:4.3d 6.5:3.5d 7.9:2.1 5.4:4.6d 5.6:4.4d NA 5.6:4.4d 6.3:3.7d 7.2:2.8d 6.7:3.3d 4.8:5.2d 0:10d 0:10 0:10 0:10 0:10

a Insect excluded from analysis either had not died 10 d post-infection or failed to produce conidia. b No statistically significant difference in number of colony-forming units (CFU) on potato dextrose agar (PDA) and Czapek’s solution agar (Cz) plates (t-test, a < 0.05). c Statistically significant difference in number of CFU on PDA and Cz plates (t-test, a < 0.05). d Statistically significant change in ratio (Chi-square test, a < 0.05).

primarily by the higher growth rate of Af6982, whereas protease deregulation in the nutrient-rich haemolymph may compensate for the slow growth of Af6982conins by degrading tissues of the haemolymph and compromising the immune response (Griesch & Vilcinskas 1998). These two different strategies resulted in equal virulence (Fig 6). Although Af6982 and Af6982conins were equally virulent, we observed differences in colonization of the insect cadaver. Competition assays heavily favoured outgrowth and conidiation on the insect cadaver by the wild-type, probably because Af6982 produced ten-times more conidia than Af6982conins on

Fig 6 – Diagram showing the development of an alternative pathogenic strategy toward Galleria mellonella larvae by the Aspergillus flavus cysteine/methionine auxotroph Af6982conins. Whereas the wild-type strain Af6982 had a higher relative growth rate, the cysteine/methionine auxotroph exhibited greater protease production, resulting in equal virulence between the strains. As mechanical pressure afforded by fungal biomass, as well as protease secretion, are important in insect cuticle penetration, the shift in these parameters due to a nutrient deficiency in Af6982conins constitutes an alternate pathogenic strategy.

insect cadavers (Scully & Bidochka 2006a). Theories on natural selection state that transmission is not necessarily correlated with virulence (Ebert & Herre 1996). For example, in insects coinfected with A. flavus and M. anisopliae, A. flavus is the more frequent colonizer of the cadavers, even though M. anisopliae is the more virulent facultative pathogen. A. flavus is the faster grower and outcompetes M. anisopliae in mummifying the insect cadaver (Hughes & Boomsma 2004). Although Af6982 was more competitive with respect to enveloping the insect cadavaer, Af6982conins might be able to survive and to evolve in nature, assuming that it successfully locates appropriate niches. Successful establishment in nature will depend on a variety of ecological factors (Berns & Rager 2000). Af6982conins may also survive as an insect pathogen simply by chance, as the emergent pathogen population following host infection is influenced by random, chaotic events in addition to selection (Scully & Bidochka 2006c). In fact, it is presumably by chance that this auxotroph was initially isolated on an infected insect larva (Scully & Bidochka 2006a). Using a fungal–insect model system, we isolated an A. flavus cysteine/methionine auxotroph that exhibited host restriction and increased virulence factor (protease) production. This strain adopted a divergent, more host-specific, but equally virulent, pathogenic strategy compared with the opportunistic, wild-type strain (Fig 6). As host restriction and virulence factor production are notable aspects of pathogen evolution, the results highlight nutrient acquisition as one of the driving forces behind the adaptation of infectious microbes. In correlation, several microbial pathogens harness a starvation response to induce virulence gene expression in reaction to nutritionally limited environments encountered


in their hosts in their quest for nutrients (Dozet et al. 2006; Gaynor et al. 2005). Moreover, there are several examples in bacteria where an increase in virulence factor was observed due to a genetic or metabolic loss (Maurelli 2007; Moore et al. 2004; Wren 2003). However, this is one of only a few studies that report the relationship between a metabolic loss and virulence in a fungus emphasizing the global role of genome decay in the evolution of increased virulence in prokaryotic, as well as eukaryotic (i.e. fungal), pathogens.

Acknowledgements This research was supported by a discovery grant from the National Sciences and Engineering Research Council of Canada (NSERC) to M.J.B., and an NSERC postgraduate scholarship to L.S.


Berns DS, Rager B, 2000. Emerging infectious diseases: a cause for concern. Israel Medical Association Journal 2: 919–923. Bidochka MJ, Khachatourians GG, 1990. Identification of Beauveria bassiana extracellular protease as a virulence factor in pathogenicity toward the migratory grasshopper, Melanoplus sanguinipes. Journal of Invertebrate Pathology 56: 362–370. Butt TM, 1987. A fluorescence microscopy method for the rapid localization of fungal spores and penetration sites on insect cuticle. Journal of Invertebrate Pathology 50: 72–74. Butt TM, Ibrahim L, Clark SJ, Beckett A, 1995. The germination behaviour of Metarhizium anisopliae on the surface of aphid and flea beetle cuticles. Mycological Research 99: 945–950. Clarkson JM, Charnley AK, 1996. New insights into the mechanisms of fungal pathogenesis in insects. Trends in Microbiology 4: 197–203. Cohen BL, 1973. Regulation of intracellular and extracellular neutral and alkaline proteases in Aspergillus nidulans. Journal of General Microbiology 79: 311–320. Dowzer CEA, Kelly JM, 1991. Analysis of the creA gene, a regulator of carbon catabolite repression in Aspergillus nidulans. Molecular and Cellular Biology 11: 5701–5709. Dozet M, Boigegrain R, Delrue R, Hallez R, Ouahrani-Bettache S, Danese I, Letesson J, De Bolle X, Kohler S, 2006. The stringent response mediator Rsh is required for Brucella melitensis and Brucella suis virulence, and for expression of the type IV secretion system virB. Cellular Microbiology 8: 1791–1802. Dunphy GB, Webster JM, 1986. Influence of the Mexican strain of Steinernema feltiae and its associated bacterium Xenorhabdus nematophilus on Galleria mellonella. Journal of Parasitology 72: 130–135. Ebert D, Herre EA, 1996. The evolution of parasitic diseases. Parasitology Today 12: 96–101. Fu Y, Marzluf GA, 1990. nit-2, the major positive-acting nitrogen regulatory gene of Neurospora crassa, encodes a sequencespecific DNA-binding protein. Proceedings of the National Academy of Sciences of the United States of America 87: 5331–5335. Gaynor EC, Wells DH, MacKichan JK, Falkow S, 2005. The Campylobacter jejuni stringent response controls specific stress survival and virulence-associated phenotypes. Molecular Microbiology 56: 8–27. Goettel MS, St Leger RJ, Rizzo NW, Staples RC, Roberts DW, 1989. Ultrastructural localization of a cuticle-degrading protease produced by the entomopathogenic fungus Metarhizium anisopliae during penetration of host (Manduca sexta) cuticle. Journal of General Microbiology 135: 2233–2239.

L. R. Scully, M. J. Bidochka

Griesch J, Vilcinskas A, 1998. Proteases released by entomopathogenic fungi impair phagocytic activity, attachment, and spreading of plasmatocytes isolated from haemolymph of the greater wax moth Galleria mellonella. Biocontrol Science and Technology 8: 517–531. Hanson MA, Marzluf GA, 1973. Regulation of a sulfur-controlled protease in Neurospora crassa. Journal of Bacteriology 116: 785–789. Hanson MA, Marzluf GA, 1975. Control of the synthesis of a single enzyme by multiple regulatory circuits in Neurospora crassa. Proceedings of the National Academy of Sciences of the United States of America 72: 1240–1244. Hughes WHO, Boomsma JJ, 2004. Let your enemy do the work: within-host interactions between two fungal parasites of leafcutting ants. Proceeding of the Royal Society of London, Series B 271: S104–S106. Hynes MJ, Kelly JM, 1977. Pleiotropic mutants of Aspergillus nidulans altered in carbon metabolism. Molecular and General Genetics 150: 193–204. Kanaan MN, Marzluf GA, 1993. The positive-acting sulfur regulatory protein CYS3 of Neurospora crassa: nuclear localization, autogenous control, and regions required for transcriptional activation. Molecular and General Genetics 239: 334–344. Kanost MR, Jiang H, Yu X, 2004. Innate immune responses of a lepidopteran insect, Manduca sexta. Immunological Reviews 198: 97–105. Kitano H, 2007. Biological robustness in complex host–pathogen systems. Progress in Drug Research 64: 241–263. Kumar V, Singh GP, Babu AM, 2004. Surface ultrastructural studies on the germination, penetration and conidial development of Aspergillus flavus link: fries infecting silkworm, Bombyx mori Linn. Mycopathologia 157: 127–135. Marques JT, Carthew RW, 2007. A call to arms: coevolution of animal viruses and host innate immune responses. Trends in Genetics 23: 359–364. Maurelli AT, 2007. Black holes, antivirulence genes, and gene inactivation in the evolution of bacterial pathogens. FEMS Microbiology Letters 267: 1–8. Moore RA, Reckseidler-Zenteno S, Kim H, Nierman W, Yu Y, Tuanyok A, Warawa J, DeShazer D, Woods DE, 2004. Contribution of gene loss to the pathogenic evolution of Burkholderia pseudomallei & Burkholderia mallei. Infection and Immunity 72: 4172–4187. Ochman H, Moran NA, 2001. Genes lost and genes found: evolution of bacterial pathogenesis and symbiosis. Science 292: 1096–1099. Pallen MJ, Wren BW, 2007. Bacterial pathogenomics. Nature 449: 835–842. Paterson IC, Charnley AK, Cooper RM, Clarkson JM, 1994. Partial characterization of specific inducers of a cuticle-degrading protease from the insect pathogenic fungus Metarhizium anisopliae. Microbiology 140: 3153–3159. Screen S, Bailey A, Charnley K, Cooper R, Clarkson J, 1997. Carbon regulation of the cuticle-degrading enzyme PR1 from Metarhizium anisopliae may involve a trans-acting DNA-binding protein CRR1, a functional equivalent of the Aspergillus nidulans CREA protein. Current Genetics 31: 511–518. Screen S, Bailey A, Charnley K, Cooper T, Clarkson J, 1998. Isolation of a nitrogen response regulator gene (nrr1) from Metarhizium anisopliae. Gene 221: 17–24. Scully LR, Bidochka MJ, 2005. Serial passage of the opportunistic pathogen Aspergillus flavus through an insect host yields decreased saprobic capacity. Canadian Journal of Microbiology 51: 185–189. Scully LR, Bidochka MJ, 2006a. A cysteine/methionine auxotroph of the opportunistic fungus Aspergillus flavus is associated with host-range restriction: a model for emerging diseases. Microbiology 152: 223–232.

An alternative insect pathogenic strategy

Scully LR, Bidochka MJ, 2006b. Developing insect models for the study of current and emerging human pathogens. FEMS Microbiology Letters 263: 1–9. Scully LR, Bidochka MJ, 2006c. The host acts as a genetic bottleneck during serial infections: an insect–fungal model system. Current Genetics 50: 335–345. Shah FA, Wang CS, Butt TM, 2005. Nutrition influences growth and virulence of the insect-pathogenic fungus Metarhizium anisopliae. FEMS Microbiology Letters 251: 259–266. Smith RE, 1984. Identification of protease isozymes after analytical isoelectric focusing using fluorogenic substrates impregnated into cellulose membranes. Journal of Histochemistry and Cytochemistry 32: 1265–1274. Srinivasan M, Dhar SC, 1990. Factors influencing extracellular protease synthesis in an Aspergillus flavus isolate. Acta Microbiologica Hungarica 37: 15–23. St Leger RJ, 1995. The role of cuticle-degrading proteases in fungal pathogenesis of insects. Canadian Journal of Botany 73: S1119–S1125. St Leger RJ, Charnley AK, Cooper RM, 1987. Characterization of cuticle-degrading proteases produced by the entomopathogen


Metarhizium anisopliae. Archives of Biochemistry and Biophysics 253: 221–232. St Leger RJ, Durrands PD, Cooper RM, Charnley AK, 1988. Regulation of production of proteolytic enzymes by the entomopathogenic fungus Metarhizium anisopliae. Archives of Microbiology 150: 413–416. St Leger RJ, Staples RC, Roberts DW, 1991. Changes in translatable mRNA species associated with nutrient deprivation and protease synthesis in Metarhizium anisopliae. Journal of General Microbiology 137: 807–815. St Leger RJ, Bidochka MJ, Roberts DW, 1994. Isoforms of the cuticle-degrading Pr1 proteinase and production of a metalloproteinase by Metarhizium anisopliae. Archives of Biochemistry and Biophysics 313: 1–7. Wren B, 2003. The Yersiniae d a model genus to study the rapid evolution of bacterial pathogens. Nature Reviews 1: 55–64. Yu J, Cleveland TE, Nierman WC, Bennett JW, 2005. Aspergillus flavus genomics: gateway to human and animal health, food safety, and crop resistance to disease. Revista iberoamericana de micologı´a: o´rgano de la Asociacio´n Espan˜ola de Especialistas en Micologı´a 22: 194–202.