Cellular Responses of the Late Blight Pathogen Phytophthora infestans to Cyclic Lipopeptide Surfactants and Their Dependence on G Proteins

Abstract

Oomycete pathogens cause major yield losses for many crop plants, and their control depends heavily on agrochemicals. Cyclic lipopeptides (CLPs) were recently discovered as a new class of natural compounds with strong activities against oomycetes. The CLP massetolide A (Mass A), produced by Pseudomonas fluorescens, has zoosporicidal activity, induces systemic resistance, and reduces late blight in tomato. To gain further insight into the modes of action of CLPs, the effects of Mass A on pore formation, mycelial growth, sporangium formation, and zoospore behavior were investigated, as was the involvement of G proteins in the sensitivity of Phytophthora infestans to Mass A. The results showed that Mass A induced the formation of transmembrane pores with an estimated size of between 1.2 and 1.8 nm. Dose-response experiments revealed that zoospores were the most sensitive to Mass A, followed by mycelium and cysts. Mass A significantly reduced sporangium formation and caused increased branching and swelling of hyphae. At relatively low concentrations, Mass A induced encystment of zoospores. It had no effect on the chemotactic response of zoospores but did adversely affect zoospore autoaggregation. A loss-of-function transformant of P. infestans lacking the G-protein α subunit was more sensitive to Mass A, whereas a gain-of-function transformant required a higher Mass A concentration to interfere with zoospore aggregation. Results indicate that Mass A disturbs various developmental stages in the life cycle of P. infestans and suggest that the cellular responses of P. infestans to this CLP are, in part, dependent on G-protein signaling.

Oomycetes cause devastating diseases of plants and animals. They are fungal look-alikes that grow as mycelium and propagate via spores but evolved independently from fungi (23). Among the plant pathogens are over 80 Phytophthora species, with the late blight pathogen Phytophthora infestans being the most renowned (12, 16). Late blight control relies heavily on fungicides that contain copper-, tin-, phenylamide-, or cyanocetamide-oximes as active ingredients. Public concerns about the adverse effects of these fungicides on food safety and the environment have led to an increased demand for novel control strategies, preferably based on natural products. In recent years, the destructive effects of cyclic lipopeptides (CLPs) on zoospores of oomycete plant pathogens have attracted considerable attention (8, 9, 33). CLPs are produced by a variety of bacterial genera including Bacillus and Pseudomonas (13, 29, 30, 33, 37). They are composed of a fatty acid tail linked to an oligopeptide, which is cyclized by a lactone ring between two amino acids in the peptide chain. Based on the length and composition of the fatty acid as well as the number, type, and configuration (L-D form) of the amino acids in the peptide moiety, their activity may change (29, 30, 33). CLPs can be chemically produced, and via structural or genetic modifications, their physicochemical properties and antimicrobial activities can be altered (1). Pseudomonas fluorescens strain SS101 produces nine cyclic lipopeptide surfactants, with massetolide A (Mass A) being the main cyclic lipopeptide (7, 9). The others are derivatives of Mass A differing in the amino acid compositions of the peptide ring (7).

One of the main modes of action of natural and synthetic CLPs is interference with the membrane integrity of the target organism, leading to pore formation and cytolysis (4, 5, 18, 19, 27, 34). For example, the CLPs Mass A and viscosin, produced by P. fluorescens strains SS101 and SBW25, respectively, act on membranes of zoospores of plant-pathogenic oomycetes, including Pythium and Phytophthora species, and this leads to the complete elimination of these propagules within 1 min of exposure (6-9). The destructive effects of Mass A on zoospores may explain, at least in part, the activity of P. fluorescens SS101 against Pythium root rot of flower bulb crops (6, 9) and tomato late blight caused by P. infestans (44). However, CLPs not only act on zoospores but may also inhibit mycelial growth of oomycetes and fungi (29, 33). The observations that several CLPs induce systemic resistance in plants against fungal and oomycete pathogens including P. infestans (31, 44) further emphasize their versatile activities and potential for crop protection. To explore and exploit the use of CLPs as a novel supplementary strategy for disease control, fundamental knowledge on their modes of action and the cellular responses of target oomycete pathogens is required.

In this study, we investigated the response of P. infestans to the CLP Mass A produced by P. fluorescens SS101. We examined various growth stages of P. infestans and performed dose-response experiments to determine the effects of Mass A on mycelial growth, sporangium formation, cyst germination, and zoospore behavior, including chemotaxis, autoaggregation, and encystment. We also examined the involvement of the α subunit of the heterotrimeric G protein in the cellular responses of P. infestans to Mass A. The results show differential sensitivities of the various growth stages to Mass A and suggest that G-protein signaling plays a role in mediating the response.

MATERIALS AND METHODS

Microorganisms and growth conditions.

P. fluorescens strain SS101 was grown on Pseudomonas agar plates (Difco, France) at 25°C for 48 h unless otherwise indicated. P. infestans wild-type strain 88069 and its transgenic derivatives gs2 and R2 were routinely grown at 18°C in the dark on rye agar medium supplemented with 2% sucrose. The P. infestans gpa1 (Pigpa1) gain-of-function mutant strain R2 and the Pigpa1-silenced mutant strain gs2 are transformants of wild-type strain 88069 that carry extra copies of a mutated form (R177H) of the G-protein α subunit gene Pigpa1 (21). In R2, the transgene is expressed, leading to the production of a constitutive active form of PiGPA1. In gs2, the Pigpa1 gene is silenced, resulting in a strain that lacks the G-protein α subunit (21). To obtain zoospores, full-grown plates (9-cm diameter) were flooded with 20 ml of sterile distilled water, and hyphae were fully submerged with a sterile glass spreader. Flooded plates were placed in the cold (4°C) for 1 to 2 h, after which the spore suspension was gently filtered over a 22-μm-mesh nylon membrane to remove sporangia. Zoospore density was determined microscopically at a ×100 magnification and adjusted to a final concentration of 10⁵ swimming zoospores ml⁻¹. Encysted zoospores were obtained by vigorously vortexing the zoospore suspensions for 1 min; encystment was checked microscopically at a ×100 magnification.

Extraction and purification of the cyclic lipopeptide surfactant Mass A.

The CLP Mass A (molecular mass, 1,139 Da) was extracted as described previously (7, 9). Reverse-phase high-performance liquid chromatography (HPLC) analysis, performed as described previously by de Souza et al. (9) and de Bruijn et al. (8), showed that Mass A makes up 65 to 70% of the partially purified extract based on peak areas at 206 nm; the other 30 to 35% consists of Mass A derivatives (8, 44). To obtain pure Mass A, the partially purified extract was fractionated by reverse-phase HPLC. The eluent was removed with a rotary evaporator (Büchi) in vacuo, and the identity and purity of Mass A were confirmed by liquid chromatography-mass spectrometry and nuclear magnetic resonance analyses as described previously (7, 9). Purified Mass A was used in all experiments except in assays to determine the effect on mycelial growth and sporangium formation; these assays required large amounts of Mass A, and therefore, we used the partially purified Mass A described previously by Tran et al. (44).

Zoosporicidal activity of Mass A.

Purified Mass A (1-mg ml⁻¹ stock solution) was added to zoospores and mixed to give final concentrations ranging from 0.5 to 1,000 μg ml⁻¹. Mass A was mixed on a glass slide with an equal volume of a zoospore suspension of P. infestans 88069. The behavior of the zoospores was determined microscopically at a ×100 magnification. Each data point is the average of data from two independent assays.

Erythrocyte assays.

The erythrocyte assay was performed according to a protocol described previously by Rainey et al. (34). Heparinated horse erythrocytes (0.2 ml) were added to 4 volumes of phosphate-buffered saline (PBS) (10 mM sodium phosphate, 0.9% NaCl [pH 7.0]), vortex mixed for 10 s, and centrifuged (14,000 × g at 4°C for 10 s). Five subsequent washings were performed before final suspension in 100 ml PBS. All assays were performed at room temperature (RT). Erythrocytes were mixed with PBS to give a final volume of 1 ml and a final optical density at 600 nm (OD₆₀₀) of between 0.3 and 0.4. Purified Mass A (1-mg ml⁻¹ stock solution) was added to the erythrocytes and mixed to give final concentrations ranging from 0.5 to 10 μg ml⁻¹. The rate of lysis, expressed as ΔOD₆₀₀ per minute, was monitored by use of a spectrophotometer (Biophotometer; Eppendorf, Hamburg, Germany). Each data point is the average of data from three independent assays, each containing four replicates per treatment.

Osmotic protection assays.

Osmotic protection assays were performed as described previously by Hutchison et al. (19). Erythrocytes were suspended in PBS (pH 7.0) as described above. Blood cells (2 μl) were mixed with PBS containing 50 mM of a sugar or polyethylene glycol to give a final volume of 1 ml and a final OD₆₀₀ of 0.3 to 0.4. The osmotic protectants used were glucose (0.42 nm), sucrose (0.54 nm), raffinose (0.66 nm), and polyethylene glycol with molecular weights of 1,000 (0.92 nm), 1,500 (1.2 nm), 3,350 (1.8 nm), and 4,000 (2.0 nm); the hydrodynamic radius of each osmolite in nm, taken from data reported previously by Lo Cantore et al. (26), is shown in parentheses after each compound. The cells were incubated for 5 min at RT before the addition of purified Mass A to a final concentration of 5 μg ml⁻¹. The rate of lysis was measured as described above. Each data point is the average of data from three independent assays, each containing four replicates per treatment.

Effect of Mass A on mycelial growth, biomass, and sporangium formation.

The effect of Mass A was studied by growing P. infestans cultures on clarified rye-sucrose agar (CRSA) (pH 7.0). Sterilized growth media were cooled down to 55°C and amended with Mass A to final concentrations of 0, 25, 50, 100, 250, and 500 μg ml⁻¹; each plate contained 20 ml of growth medium. A plug (0.8 cm²) with P. infestans mycelium, excised from the periphery of full-grown CRSA plates, was placed into the center of the Mass A-amended growth medium and incubated in the dark at 18°C. Radial mycelium growth was measured with an electronic ruler after 4, 7, 9, and 11 days. For each treatment, five replicates were used, and the assays were performed three times. The effect of Mass A on the morphology of the hyphae of P. infestans was determined microscopically at different time points with an inverted microscope (Axiovert10; Zeiss). To evaluate the effect of Mass A on sporangium formation, four mycelial plugs (0.8 cm² each) were taken from the edge of 11-day-old cultures of P. infestans grown on CRSA or one-fifth-strength potato dextrose agar amended with different concentrations of Mass A. Each mycelial plug was transferred into a sterile 1.5-ml Eppendorf tube containing 1 ml of Isotone II electrolytic buffer (Coulter Electronic, Inc.) and vortexed vigorously for 2 min to release the sporangia. The numbers of sporangia in 3-μl droplets were determined microscopically at a ×100 magnification.

The effects of Mass A on biomass of P. infestans were studied in clarified rye-sucrose liquid medium; final concentrations of Mass A were 0, 5, 10, 25, 50, 100, and 250 μg ml⁻¹. A plug (0.8 cm²) was transferred to 5-cm-diameter petri dishes containing 5 ml of the growth medium. After incubating the still cultures at 18°C for 11 days, the mycelium was collected by centrifugation and blotted dry on a Whatmann filter prior to determining the fresh weight; after drying at 65°C for 12 h, the mycelium dry weight was determined. The obtained weights were corrected for the weight (fresh and dry) of the agar plugs used to transfer P. infestans mycelium. Liquid cultures were also used to determine the effect of Mass A on sporangium formation. After 11 days of incubation, the cultures were transferred into 50-ml tubes. The mycelium was harvested by centrifugation and resuspended in 5 ml of Isotone II electrolytic buffer. The mycelial suspension was cooled on ice and homogenized by use of a Polytron mixer (Kinematica Gmbh, Luzern, Switzerland) for 2 min with the speed set at 3. The density of the sporangia was determined microscopically as described above. Combined with the fresh and dry weights of the harvested mycelium, sporangium formation per mg fresh and dry weight was calculated.

Effect of Mass A on chemotaxis, autoaggregation, and encystment of swimming zoospores.

The effect of Mass A on the chemotaxis of zoospores was studied as described previously by Latijnhouwers et al. (21). Instead of glass capillaries filled with water agar with and without the chemoattractant glutamic acid, we used two small agar plugs (0.78 cm²), one consisting of 1% (wt/vol) water agar and the other consisting of 1% water agar supplemented with glutamic acid (50 mM). The plugs were transferred to the bottom of a 3-cm-diameter petri dish and subsequently submersed with 5 ml of a suspension containing 10⁵ zoospores per ml. Mass A was added to the zoospore suspension to obtain final concentrations of 0, 0.5, 1.0, 2.5, 5.0, and 10.0 μg ml⁻¹. Movement toward and accumulation of the zoospores on the water agar plugs were scored microscopically at a ×50 magnification over a time course of 5 to 10 min. Autoaggregation assays were performed using 3-cm-diameter petri dishes as described previously by Latijnhouwers et al. (21). The Mass A concentrations used in these assays were the same as those used in the chemotaxis assays. Five to ten minutes after incubation at RT (∼20°C), zoospore autoaggregation was visually scored and recorded with a digital camera. For each of the zoospore suspensions exposed to different concentrations of Mass A, the percentage of swimming and encysted zoospores was determined microscopically after 30 min of incubation at RT (∼20°C).

Effect of Mass A on germination of encysted zoospores and germ tube length.

Encysted zoospores were exposed to Mass A concentrations of 0, 25, 50, 100, 250, and 500 μg ml⁻¹. The experiment was performed using 96-well plates (Greiner). One hundred microliters of a suspension of encysted zoospores was transferred to each well, and Mass A was added to obtain the final concentrations mentioned above. Plates were incubated at 25°C, and samples (20 μl) were taken after 60, 90, and 120 min and kept on ice prior to microscopic analysis. The percentage of germinated cysts was determined using 3-μl droplets at a ×100 magnification. Germ tube length was determined at 25°C for 180 min. Therefore, five 3-μl droplets were taken, and the lengths of 10 to 15 randomly selected germ tubes were measured at a ×100 magnification with an ocular micrometer.

Expression analysis of Pigpa1.

P. infestans mycelium grown on rye-sucrose agar for 7 to 9 days in the dark at 18°C (21) was frozen in liquid N₂ and stored at −80°C. RNA was isolated with Trizol reagent (Invitrogen), followed by DNase I (GE Healthcare) treatment. One microgram of RNA was used for cDNA synthesis with Superscript III (Invitrogen) according to the manufacturer's protocol. For quantitative PCR (Q-PCR) conducted with the 7300SDS system from Applied Biosystems, the SYBR green core kit (Eurogentec) with a final concentration of 3.5 mM MgCl₂ was used according to the manufacturer's protocol. The primer concentrations were optimized (400 nM final concentration for each primer), and a dissociation curve was performed to check the specificity of the primers. The primers used for the Q-PCR were as follows: 5′-TGCGTCCACAAACCGTATGA-3′ (forward) and 5′-TTGTTGTTGATGATCTCGTCGAA-3′ (reverse) for Pigpa1 and 5′-TACCACCATGTACCCGGGCATTG-3′ (forward) and 5′-CGACAGCGACGACTGGATGG-3′ (reverse) for the housekeeping gene actA. The cycle where the SYBR green fluorescence crosses a manually set threshold cycle (C_T) was used to determine transcript levels. For each gene, the threshold was fixed based on the exponential segment of the PCR curve. The C_T value for Pigpa1 was corrected for actA as follows: ΔC_T = C_T(Pigpa1) − C_T(actA). The relative quantification (RQ) values were calculated by the following formula: RQ = 2^{−[ΔCT(transformant) − ΔCT(wild type)]}. If there is no difference in transcript levels in the transformant and the wild type, then RQ equals 1 and log RQ equals 0. Q-PCR analysis was performed in duplicate (technical replicates) upon three independent RNA isolations (biological replicates). Statistically significant differences were determined for log-transformed RQ values by analysis of variance (P < 0.05) followed by Bonferroni post hoc multiple comparisons (SAS Institute, Inc., Cary, NC).

Statistical analysis.

After certifying normal distribution and homogeneity of variances, data were analyzed by analysis of variance followed by Tukey's studentized range test and the Student-Newman-Keuls test (SAS Institute, Inc., Cary, NC). All the assays described in this study were performed at least twice, and representative data are shown. For all assays, treatments were replicated three to five times. Percentages of swimming and encysted zoospores, cyst germination, and zoospore lysis were arcsin transformed prior to statistical analysis.

RESULTS

Mass A-induced cell lysis is due to pore formation.

Dose-response experiments performed previously by de Souza et al. (9) with partially purified Mass A showed that the lysis of P. infestans zoospores occurred within 1 min at concentrations equal or higher than the critical micelle concentration (CMC) of 25 μg ml⁻¹. The same result was obtained in this study with HPLC-purified Mass A. When zoospores were exposed to lower concentrations of Mass A and for 15 min instead of 2 min, 8 to 10% and 33 to 47% of the zoospores disintegrated at Mass A concentrations of 5 and 10 μg ml⁻¹, respectively. This result shows that Mass A also has zoosporicidal activity at concentrations below the CMC, although more time is required to evoke lysis.

For several CLPs including tolaasin, syringomycin, and white line-inducing principle, it has been found that they partition into membranes and form pores that lead to cell lysis (19, 26, 34). This was determined in so-called osmotic protection assays in which the lysis of plasma membranes of erythrocytes is prevented by molecules of specific molecular sizes that physically block the pores, thereby obstructing the ion fluxes that lead to cell lysis. Accordingly, to determine the pore-forming capacity of Mass A, osmotic protection assays were performed with zoospores of P. infestans. In the absence of Mass A, however, several of the osmotic protectants appeared to induce the encystment of zoospores, thereby rendering them insensitive to lysis by Mass A. As an alternative, we used erythrocytes and demonstrated that Mass A has hemolytic activity at a concentration of 4 μg ml⁻¹; most cells were lysed immediately upon the addition of 7 μg ml⁻¹ Mass A (Fig. 1A). The osmotic protection assays showed that none of the osmoprotectants, except polyethylene glycol (T and U), was able to prevent the Mass A-induced lysis of erythrocytes (Fig. 1B). The addition of polyethylene glycol with viscometric radii of >1.2 nm fully protected erythrocytes from the lytic effects of Mass A, whereas polyethylene glycol with a viscometric radius of 0.89 nm and sugars with viscometric radii of 0.66 nm or less had no protective effect. Based on these results, the mean channel radius of pores induced in erythrocyte membranes by Mass A was estimated to be between 1.2 and 1.8 nm.

FIG. 1.

(A) Rate of lysis of heparinated horse erythrocytes at increasing concentrations of purified Mass A. (B) Effect of different colloid osmotic protectants with various viscometric radii on hemolysis by Mass A. In the experiments described in B, the final concentration of purified Mass A was 5 μg ml⁻¹, and the osmotic protectants are abbreviated as follows: G, glucose; S, sucrose; R, raffinose; P, polyethylene glycol (molecular weight, 1,000); Q, polyethylene glycol (molecular weight, 1,500); T, polyethylene glycol (molecular weight, 3,350); U, polyethylene glycol (molecular weight, 4,000). C is a control containing only Mass A. Data are the means of data from four individual measurements, and error bars represent the standard errors of the means.

Mass A affects autoaggregation and encystment of zoospores but not chemotaxis.

Previously, Latijnhouwers et al. (21) demonstrated that P. infestans zoospores can autoaggregate. When we transferred a zoospore suspension (≥2 × 10⁵ zoospores ml⁻¹) to a petri dish, aggregates of swimming zoospores were visible within 2 min (Fig. 2A). However, when zoospores of P. infestans were exposed to different Mass A concentrations, autoaggregation of the zoospores occurred only at low concentrations of Mass A but not at concentrations of 5 μg ml⁻¹ and higher (Fig. 2A). The interference of Mass A with autoaggregation was further investigated microscopically by determining the percentage of swimming and encysted zoospores. With increasing concentrations of Mass A, the number of encysted zoospores significantly increased (Fig. 2B). Especially at concentrations of 2.5, 5, and 10 μg ml⁻¹, there was a significant induction of zoospore encystment in wild-type P. infestans strain 88069 (Fig. 2B). Mass A did not affect the chemotactic response of swimming zoospores toward the chemoattractant glutamic acid.

FIG. 2.

Effect of increasing concentrations of Mass A on autoaggregation of zoospores (A) and zoospore encystment of Phytophthora infestans strain 88069 (B). (A) White aggregates of zoospores are visible within 2 to 3 min. (B) The y axis shows the percentage of encysted zoospores in a zoospore suspension exposed to Mass A for 30 min at RT. Mean values from five replicates are given, and error bars represent the standard errors of the means.

Mass A reduces germ tube growth of cysts.

Germination of encysted zoospores was not adversely affected by Mass A except at a relatively high concentration; at 500 μg ml⁻¹, a small but statistically significant (P < 0.05) reduction was observed (see Fig. S1A in the supplemental material). However, Mass A significantly reduced subsequent outgrowth of the germ tube from the encysted zoospores: the germ tube length decreased with increasing concentrations of Mass A and was significantly (P < 0.05) reduced by approximately 10 to 40% at Mass A concentrations of 100 μg ml⁻¹ and higher (see Fig. S1B in the supplemental material).

Mass A inhibits mycelial growth and reduces sporangium formation.

On CRSA medium, the mycelial growth of P. infestans strain 88069 was significantly (P < 0.05) inhibited at Mass A concentrations higher than 100 μg ml⁻¹. Also, in liquid CRSA medium, growth was significantly inhibited already at a Mass A concentration of 25 μg ml⁻¹ (Fig. 3A): the dry weight of mycelial biomass decreased with increasing concentrations of Mass A (Fig. 3B).

FIG. 3.

Effects of Mass A on mycelial growth and sporangium formation of Phytophthora infestans strain 88069. P. infestans was grown for 11 days (A to D) or 9 days (E) in clarified rye sucrose liquid medium amended with increasing concentrations of Mass A. (A) Representative examples of the growth response of P. infestans to Mass A. The numbers above each of the pictures refer to the Mass A concentrations in μg ml⁻¹. (B and C) Effect of Mass A on mycelium dry weight (B) and sporangium formation (C). In C, sporangium numbers are expressed per unit biomass (dry weight). In B and C, means of data for five replicates are shown, and error bars represent the standard errors of the means. (D) Representative microscopic pictures (×100 magnification) of hyphae of P. infestans grown in medium without Mass A (control) and with Mass A at a concentration of 500 μg ml⁻¹. In the presence of Mass A, the hyphae show increased branching. (E) Representative microscopic pictures (×200 magnification) of hyphae of P. infestans grown in medium without (control) and with Mass A at a concentration of 100 μg ml⁻¹. In the presence of Mass A, the mycelium exhibits hyphal swelling.

The effect of Mass A on sporangium formation by P. infestans was determined after 11 days of incubation. In liquid CRSA medium, the number of sporangia formed per mg of dry weight decreased significantly with increasing concentrations of Mass A (Fig. 3C). Microscopic analysis further revealed that Mass A caused an increased branching of hyphae (Fig. 3D). In contrast to thin, elongated, and extended mycelial growth in the control, Mass A caused hyphal swellings when exposed to 100 μg ml⁻¹ Mass A and higher concentrations (Fig. 3E).

The α subunit of the heterotrimeric G protein affects the sensitivity of P. infestans to Mass A.

To determine if heterotrimeric G proteins play a role in the response of P. infestans to the CLP Mass A, experiments were conducted with Pigpa1-silenced mutant strain gs2, which lacks the G-protein α subunit GPA1, and gain-of-function mutant strain R2, which expresses not only endogenous gpa1 but also a constitutively active form of the G-protein α subunit gene gpa1 (21). Gene expression analyses performed in this study showed relatively high transcript levels of Pigpa1 in the mycelium of mutant strain R2 (i.e., a threefold increase relative to that in the parental strain). Since the transgene encodes a constitutively active form of the G-protein α subunit, we anticipate that mutant strain R2 has higher levels of the Gα subunit locked in its active, GTP-bound state.

In the Pigpa1-silenced mutant strain gs2, mycelium growth was significantly (P < 0.05) inhibited at a Mass A concentration of 50 μg ml⁻¹, whereas the growth of the gain-of-function mutant strain R2 was inhibited at a concentration of 250 μg ml⁻¹ (Fig. 4). In Pigpa1-silenced mutants, zoospore autoaggregation is impaired (21), and the addition of Mass A did not affect this impairment (Table 1). However, the induction of zoospore encystment required a twofold-higher concentration for gs2 than for parental strain 88069. Also, for gain-of-function mutant strain R2, a twofold-higher concentration of Mass A was needed to induce zoospore encystment and to disturb zoospore autoaggregation (Table 1). At a Mass A concentration of 5 μg ml⁻¹, 35 to 40% of the zoospores of R2 were still swimming, compared to only 5 to 10% of the zoospores of strain 88069. This higher percentage of swimming zoospores may explain why zoospores of R2 can still aggregate at a Mass A concentration of 5 μg ml⁻¹.

FIG. 4.

Effect of the cyclic lipopeptide Mass A on mycelial growth of Phytophthora infestans strain 88069 (⧫), Pigpa1-silenced mutant strain gs2 (▪), and PiGPA1 gain-of-function mutant strain R2 (▴). The parental strain and mutant strains were grown on CRSA medium amended with different concentrations of Mass A. Radial mycelial growth was determined after 11 days, and growth at the different Mass A concentrations is expressed relative to the control (set at 100%). Mean values of data from three independent experiments each containing five replicates are given; error bars represent the standard errors of the means. An asterisk indicates significant (P < 0.05) differences between the response of parental strain 88069 and Pigpa1-silenced mutant strain gs2 or Pigpa1 gain-of-function mutant strain R2.

TABLE 1.

Effect of the cyclic lipopeptide Mass A on zoospores of wild-type Phytophthora infestans strain 88069, mutant strain gs2, and mutant strain R2

P. infestans strain	Lowest concn of Mass A (μg ml−1)a
	Occurrence of:		Disturbance of zoospore autoaggregation
	Zoospore lysis	Zoospore encystment
88069 (wild type)	25	2.5	5.0
Pigpa1-silenced mutant gs2	25	5.0	—b
Gain-of-function mutant R2c	25	5.0	10.0

^aIndicated are the lowest concentrations of Mass A (in μg ml⁻¹) at which zoospore encystment and zoospore lysis occur and at which zoospore aggregation is disturbed.

^bPigpa1-silenced mutant strain gs2 shows no autoaggregation or chemotaxis due to the absence of the G-protein α subunit (see reference 21).

^cIn the absence of Mass A, gain-of-function mutant strain R2 behaves the same as wild-type strain 88069.

DISCUSSION

To date, most studies focusing on the sensitivity of plant-pathogenic fungi and oomycetes to novel bioactive metabolites take into account only one particular stage in the life cycle of the pathogen, usually hyphal growth (11). Life cycles of pathogens, however, are more complex and may comprise different types of infectious propagules. Phytophthora species can infect plant tissue by means of sporangia and zoospores. An understanding of the variation in sensitivities of different infectious propagules to specific bioactive compounds will give more insight into their potential efficacy or limitation. In this study, we investigated the response of P. infestans to the CLP Mass A, which is produced by the antagonistic bacterium P. fluorescens strain SS101. The results showed that cysts are the most tolerant to Mass A, followed by mycelium and swimming zoospores. The most striking effect of Mass A on P. infestans is the zoosporicidal activity (9). This study showed that zoospore lysis also occurred at Mass A concentrations below the CMC and that Mass A induces the formation of transmembrane pores with an estimated radius of approximately 1.2 to 1.8 nm, a pore size similar to that reported recently for the structurally related white line-inducing principle (26). A second significant effect of Mass A observed in this study was the substantial reduction in sporangium formation in in vitro cultures. This finding is in line with observations described previously by Tran et al. (44), who measured sporangium formation in late blight lesions on tomato leaves treated with the Mass A-producing strain P. fluorescens SS101; they found significantly less sporangia per lesion area in treated than in nontreated late blight lesions (44). Given the importance of both zoospores and sporangia in the infection process, the zoosporicidal activity as well as the substantial reduction in sporangium formation by Mass A may impact the onset and epidemic progress of late blight disease.

Microscopic analysis further revealed that the exposure of P. infestans to Mass A led to increased hyphal branching and induced the encystment of zoospores at concentrations below the CMC. Similar phenomena were described previously by Thrane et al. (42, 43), who exposed the oomycete Pythium ultimum and the fungus Rhizoctonia solani to viscosinamide. Those authors further hypothesized that the increased amount of branching may be the result of an increased Ca²⁺ influx due to ion channel formation by viscosinamide (43). An increased cytosolic Ca²⁺ concentration was also proposed to be an explanation for the increased amount of branching in Neurospora crassa cells exposed to the ionophore A23187 (36).

In addition to the zoosporicidal activity, Mass A was shown to adversely affect the autoaggregation of zoospores at relatively low concentrations. The aggregation of zoospores of oomycetes occurs rapidly and appears to be a spontaneous event requiring no apparent exogenous stimulus. This intriguing phenomenon was previously reported for Pythium species (35), Phytophthora drechsleri (32), Phytophthora palmivora (20), P. infestans (21), and Achlya species (41). Autoaggregation may enhance the inoculum potential at the infection site and could provide protection against other harmful microorganisms or metabolites, thereby increasing the success of host penetration and infection. Carlile (3) previously demonstrated that zoospores do not utilize external nutrients until germination but exploit internal energy reserves for motility until a suitable host is encountered. Reid et al. (35) previously suggested that autoaggregation is a survival mechanism when internal energy reserves become low, permitting the regermination of some zoospores in the population by the remobilization of resources of other zoospores in the aggregate. The results of our study show that Mass A interferes with zoospore autoaggregation by inducing the encystment of zoospores and may therefore adversely affect survival or plant colonization by P. infestans and other oomycete pathogens.

The observation that a transformant of P. infestans lacking the G-protein α subunit GPA1 was significantly more sensitive to Mass A suggests that G-protein signaling is involved in the response of the late blight pathogen to this CLP. In eukaryotes, heterotrimeric G proteins, composed of α, β, and γ subunits, are key regulators in many signaling pathways. They are linked to heptahelical transmembrane receptors (2, 14, 17, 21, 25, 28, 38, 45), which, upon ligand perception, induce the dissociation of the heterotrimeric G protein into an α subunit and a βγ dimer. This leads to the activation of regulatory cascades and changes in gene expression, cellular functions, and metabolism (10, 17, 21, 25, 45). Unlike most eukaryotes, Phytophthora species possess only one Gα subunit gene and one Gβ subunit gene, which are differentially expressed in the different stages of the life cycles of these pathogens (17, 24). In P. infestans, Pigpa1 is expressed in sporangia, zoospores, cysts, and germinated cysts, with the highest level of expression in sporangia (24). Also, the Gβ subunit gene Pigpb1 has the highest expression level in sporangia (23), and mutants that lack the Gβ subunit can no longer sporulate (22). In mutants that lack the Gα subunit, such as the P. infestans gs2 mutant strain used in this study and Phytophthora sojae gpa1-silenced mutants described previously by Hua et al. (17), zoospore motility and chemotaxis are severely disturbed, and pathogenicity is also reduced or lost (17, 21). Unlike mutants that lack the Gα subunit, Gα gain-of-function mutants show no aberrant phenotypes and have retained the ability to infect potato (21).

In mutant strain gs2, where there is no Pigpa1 expression and, thus, no Gα subunit in any stage, mycelium growth is inhibited at lower Mass A concentrations than in the wild type, and this mutant is thus more sensitive to Mass A. In contrast, gain-of-function mutant strain R2 with the Gα subunit locked in its active, GTP-bound state shows the opposite effect and is less sensitive to Mass A. When assuming that the higher Mass A sensitivity of gs2 is due to the absence of the Gα subunit PiGPA1, the lower Mass A sensitivity of R2 is likely due to a higher availability of an activated form of PiGPA1. Interestingly, for mammalian cells, it was also reported that CLPs can interfere with canonical signaling pathways by inhibiting the activity of a G-protein α subunit (39, 40). Takasaki et al. (39) previously showed that YM-254890, a cyclic depsipeptide of Chromobacterium, is a selective inhibitor of Gα_q/11, one of the several mammalian Gα subunits, and the target of this compound is the exchange of GDP for GTP in Gα_q/11 activation. Whether a similar function could be assigned to Mass A is as yet unknown, but our findings fit in the model that Mass A targets the GDP-for-GTP exchange. In the gain-of-function mutant, there is no need for the exchange of GDP for GTP on the Gα subunit because the point mutation has locked the Gα subunit into its active, GTP-bound state. If Mass A, similarly to YM-254890, inhibits the exchange step of GDP for GTP, then in Gα gain-of-function mutant strain R2, there is no or less target for Mass A, explaining the reduced sensitivity for Mass A. Validation of the model will require more in-depth analyses using a variety of G-protein mutants and should be supported by extensive biochemical studies. To further unravel the modes of action of Mass A on P. infestans and cellular responses, we are currently performing genome-wide expression profiling by microarray analysis (15).