Talaromyces variabilis interferes with Pythium aphanidermatum growth and suppresses Pythium-induced damping-off of cucumbers and tomatoes

Pythium-induced damping-off disease is a major disease limiting cucumber and tomato production in different parts of the world. The current study investigated the efficiency of Talaromyces variabilis and its bioactive metabolites in suppressing Pythium-induced damping-off of cucumbers and tomatoes. T. variabilis inhibited the in vitro growth of P. aphanidermatum in solid and liquid media. In addition, abnormalities in P. aphanidermatum hyphae were observed as a result of T. variabilis. Extracts from T. variabilis induced cellular leakage and suppressed oospore production of P. aphanidermatum. Biochemical analyses of T. variabilis metabolites showed that T. variabilis produces glucanase, cellulase and siderophores, suggesting the contribution of these metabolites in the inhibition of P. aphandermatum growth and in hyphal abnormalities. Treating cucumber seeds with spore and mycelial suspension of T. variabilis isolates led to a significant improvement in the seedling survival of P. aphanidermatum-inoculated seedlings from 18 to 52% (improvement by 34%) for isolate 48 P and from 30–66% (improvement by 36%) for isolate 28 R. Similarly, treating tomato seeds with spore and mycelial suspension of T. variabilis isolates led to a significant improvement in the seedling survival of P. aphanidermatum-inoculated seedlings from 7 to 36% (improvement by 29%) for isolate 28 R and from 20 to 64% (improvement by 44%) for isolate 48 P. Differences in the percent improvement in seedling survival between experiments may be related to difference in the efficacy of the two different isolates or their interaction with the hosts and pathogen. The use of T. variabilis in the biocontrol of Pythium-induced diseases may offer alternatives to the currently used chemical control.


Results
Identification of Talaromyces isolates. The combined ITS, TUB and CMD dataset comprises 18 isolates of Talaromyces. Phylogenetic analysis showed that isolates 48 P and 28 R belong to T. variabilis (Fig. 1).
Antagosnistic effect of T. variabilis isolates against P. aphanidermatum. Both T. variabilis isolates showed antagonistic activity against P. aphanidermatum in PDA medium (Table 1). T. variabilis isolates 48 P and 28 R produced inhibition zones of 8.5 mm and 6.25 mm, respectively.
The second experiment was conducted in PDA plates to observe the antagonistic effect of T. variabilis isolates over time. P. aphanidermatum ceased its growth. However, T. variabilis isolates 48 P and 28 R continued to grow towards P. aphanidermatum and filled the plate after 13.5 days and 15.5 days, respectively ( Table 1).
Effect of culture filtrates of T. variabilis on P. aphanidermatum growth and oospore production. Treating P. aphanidermatum with culture filtrates of T. variabilis isolates led to considerable inhibition in mycelial growth in all the tested concentrations (Fig. 2). T. variabilis isolates fully suppressed the growth of P. aphanidermatum at 75% concentration. However, the growth was reduced at 50% and 25% concentrations (Fig. 2). Furthermore, oospore production by P. aphanidermatum decreased significantly when it was exposed to 20% culture filtrate of T. variabilis (48 P: 8 oospores; 28 R: 9 oospores) compared to the control (56 oospores) ( Table 1). Siderophore production by T. variabilis isolates. Both isolates of T. variabilis produced siderophore in both media. However, King B medium contains the highest amount of siderophore, 24.25 μM for 48 P and 21.85 μM for 28 R compared to Glucose medium, 15.86 μM for 48 P and 13.56 μM for 28 R (Fig. 3).
Effect of T. variabilis on P. aphanidermatum morphology. Both isolates of T. variabilis induced significant abnormalities in general shape, internal content and the tips of main hyphae and hyphal branches of P. aphanidermatum. Isolate 48 P had a greater impact on hyphal morphology compared to isolate 28 R (Fig. 4).
Light microscope examination showed that the general shape of hypha became twisted, bulbous-like, swollen, loss of hyphal content and hyphae have protrusions and narrowings (Fig. 5). Also the internal content became empty or semi-empty, and the tips became wrapped-up, and wavy (Fig. 5).
Furthermore, scanning electron microscope showed similar observations such as shrunken and wavy hyphae, hyphal content exits and hyphae have protrusions and narrowings, as compared to the control which had straight, smooth surface and full hyphae (Fig. 6).
Biocontrol potential of T. variabilis isolates against damping-off diseases of cucumber and tomato. T. variabilis isolates (48 P and 28 R) did not cause any significantly harmful effects on the length, fresh weight and dry weight of cucumber and tomato seedlings (Tables 2, 3). Both T. variabilis isolates 48 P and 28 R had considerable biocontrol efficacy against damping-off disease in cucumber and tomato (Table 3). Treating cucumber seeds with spores and mycelial suspension of T. variabilis isolate 48 P and P. aphanidermatum led to significant improvement in seedlings survival (51.78%) compared to the control (17.86%). Similar results were observed by T. variabilis isolate 28 R with significant improvement in cucumber survival (66.07%) compared to the control (30.36%) ( Table 3). www.nature.com/scientificreports www.nature.com/scientificreports/ Similarly, treating tomato seeds with spore and mycelial suspension of 48 P and P. aphanidermatum led to significant improvement in seedling survival (64.28%) compared to the control (19.64%). Also, 28 R significantly improved tomato survival (35.71%) compared to the control (7.14%).

Discussion
Two strains of T. variabilis 48 P and 28 R were isolated from Rhazya stricta and Zygophyllum coccineum, respectively. The two plants are native to Oman. Species of Talaromyces are known endophytes and found on a wide range of plants such as Potentilla fulgens 24 , Dactylis glomerata 25 and Aloe vera 26 . The current study proved that T. variabilis isolates were not pathogenic to cucumbers and tomatoes plants. This result was compatible with the definition of endophytes as fungi which colonize the stems and leaves of plants without causing any symptoms of disease 27 .
Our study clearly demonstrated that T. variabilis isolates produced inhibition zones against P. aphanidermatum on PDA media. Many other fungi such as Trichoderma spp. 28 and bacteria such as Pseudomonas fluorescens 29 are known to be antagosnistic against harmful pathogens such as Aspergillus flavus, Fusarium moniliforme and Rhizoctonia solani. The inhibition activity is mainly due to their ability to secrete bioactive compounds that inhibit plant pathogens 30 .
The culture filtrates of T. variabilis isolates were effective against P. aphanidermatum growth in liquid media. They significantly decreased P. aphanidermatum dry weight at 25% and 50% concentrations, and fully suppressed its growth at 75% concentration. Previous studies showed that Streptomyces hydrogenans culture filtrates inhibited the growth of Alternaria Brassicicol 31 . Oospore production was greatly decreased in the presence of culture filtrates of 48 P and 28 R. Similar observation was made by 16 , where Aspergillus terreus affected spore production by P. aphanidermatum.
Our data showed that both Talaromyces isolates produce cellulase enzyme. Cellulase enzyme can be produced by several fungal genera such as Trichoderma 32 , Aspergillus 33,34 and Talaromyces 35 . Fungi and bacteria that secrete hydrolytic enzyme have biocontrol ability against plant pathogen. For example chitinase, glucanase and protease enzymes produced by Trichoderma harzianum are antagonistic against some fungi 36 . Chitinase and β-1,3-glucanase enzymes produced by Clonostachys rosea f. catenulata were responsible for efficient biocontrol against fungal plant pathogens 37 . About 18% of the P. aphanidermatum cell wall consists of cellulose 38 . The  www.nature.com/scientificreports www.nature.com/scientificreports/ efficacy of Talaromyces isolates as biocontrol agents against damping-off disease may be in-part due to cellulose production by Talaromyces isolates.
Loss in integrity of P. aphanidermatum cells was observed due to T. variabilis isolates 48 P and 28 R culture filtrates. Consistently, the antifungal metabolites produced by Sporothrix flocculosa led to cellular leakage in several phytopathogens 39 . Another study by Zhao, et al. 40 showed that Streptomyces bikiniensis causes cellular leakage in Fusarium oxysporum.
Our study showed the production of glucanase enzyme by 48 P and 28 R isolates. Several previous studies also documented a role of extracellular enzymes in biocontrol of pathogens. Examples include cellulases produced by Lysinibacillus sphaericus 19 and chitinases and glucanases produced by Trichoderma species 41 . Our results showed that siderophores were produced by 48 P and 28 R isolates in King B and glucose media. Numerous fungi and bacteria could produce siderophores that are effective against pathogens. Siderophores produced by Rhizobium meliloti led to inhibition of Macrophomina phaseolina 42 . Moreover, cucumbers damping-off disease was controlled by Aspergillus terreus which was able to produce siderophores 16 . These siderophores may deprive the pathogen of iron, thus limiting essential nutrient 43 . www.nature.com/scientificreports www.nature.com/scientificreports/ The morphology of P. aphanidermatum hyphae were significantly affected by 48 P and 28 R isolates, showing several abnormalities. Getha and Vikineswary 44 showed the following abnormalities in Fusarium oxysporum hypha due to antagonistic influence of Streptomyces violaceusniger: swelling, distortion and excessive branching of hyphae, thickened with bulbous-like formation along the ends. Another study by Halo, et al. 16 showed abnormalities in hyphae of P. aphanidermatum such as shrunken, semi empty and empty content and wrapped up ends under the effect of Aspergillus terreus.
Our study confirmed the efficient role of 48 P and 28 R isolates against cucumbers damping-off and tomatoes damping-off diseases. A previous study by Sivan, et al. 45 showed the suppression P. aphanidermatum by Trichoderma harzianum. Similarly, Gliocladium catenulatum inhibited cucumbers damping-off and root rot diseases caused by P. aphanidermatum 46 . Also, damping-off of tomatoes disease caused by P. aphanidermatum was inhibited by Trichoderma viride and Pseudomonas fluorescens biocontrol agents 47 . Furthermore, endophytic  www.nature.com/scientificreports www.nature.com/scientificreports/ actinomycetes were able to supress pathogenic activities of P. aphanidermatum because they produce glucanase enzyme 48 .
Our study is the first comprehensive report on the efficacy of T. variabilis isolates and their byproducts on P. aphanidermatum and Pythium damping-off of cucumbers and tomatoes.
The efficacy of these endophytes in suppressing P. aphanidermatum in the in vitro and in vivo tests through multiple mechanisms suggests that they may be effective against other harmful phytopathogens, including Pythium species that cause diseases in other plants. Also, using these endophytes provide an efficient alternative to the use of synthetic chemicals because P. aphanidermatum is less likely to develop resistance against these antagonistic fungi because they have multiple modes of action.

Materials and Methods
Talaromyces isolates. Rhazya stricta and Zygophyllum coccineum plants from desert sites in the Sultanate of Oman were selected for the isolation of endophytic fungi. The collections of samples was in May-August 2016 from Adam, 150 km from Muscat, the capital area of Oman. The method of Larran, et al. 49 was followed for endophytic fungi isolation, as described by Halo et al. 16 .
Two fungal isolates (Talaromyces) were identified using sequences of three genes: the internal transcribed spacer region of the ribosomal RNA (ITS), beta-tubulin (TUB) and Calmodulin (CMD). The three genes were amplified using ITS1/ITS4, BT2A/BT2B and CMD5/CMD6 primers, respectively, using previously described reaction mixtures and conditions [50][51][52] . MEGA V.6 was used for sequence alignment 53  Antagosnistic effect of T. variabilis isolates against P. aphanidermatum. The antagosnistic acivity of T. variabilis isolates 48 P and 28 R against P. aphanidermatum was investigated using fresh cultures of P. aphanidermatum and T. variabilis as explained in our previous work 54 , using dual culture assay. There were four replicates in each experiment.
Isolates 48 P and 28 R were grown on PDB media in an incubator shaker at 28 °C for ten days to produce effective metabolites. The culture filtrates were obtained by centrifugation at 10,000 g, filtered through 0.2 µm Minisart filters, transferred to conical flasks and stored at 4 °C for further experiments.
Three concentrations of culture filtrates (75%, 50% and 25%) were used to study their effect on P. aphanidermatum growth in PDB media while the control included only PDB media. The concentration 75% consisted of 75% filtrate and 25% PDB, and so on for the other concentrations. A 3-mm diameter disk of P. aphanidermatum was added to each flask. Flasks were then kept in an incubator shaker at 28 °C and 120 rpm for 7 days. Finally, the liquid was disposed and the mycelium of P. aphanidermatum was dried in an oven at 65 °C for 24 h. The dry weights of the treatments and control were measured using three replicates for each isolate.
Effect of culture filtrates of T. variabilis on extracellular conductivity and oospore production by P. aphanidermatum. The leakage of cellular components from mycelium of P. aphanidermatum was studied using 5 mg of dried mycelium obtained from the liquid culture. 10 ml of T. variabilis culture filtrates was added to dried mycelium and centrifuged. Extracellular conductivity of the culture was measured after 24 h 16,31 .
The effect of T. variabilis on oospore production of P. aphanidermatum was studied using V8 agar medium with 20% of T. variabilis culture filtrate. There were three replicates per treatment. The V8 agar medium without T. variabilis culture filtrate served as control 16 .
Glucanase activity of T. variabilis culture filtrates. Glucanase production by T. variabilis isolates was detected using a protocol described by Jackson, et al. 55 . Samples were analyzed spectrophotometrically at 400 nm using an ELISA spectrophotometer. The final enzyme activity was calculated as per Jackson, et al. 55 .
Cellulase activity. Two T. variabilis strains were cultivated in basal medium of Mandels and Weber (1969) 56 supplemented with 1% cellulose. The flasks were incubated in a rotary shaker at 200 rpm at 28 °C for 10 days. Culture filtrates were obtained through centrifugation and used fresh. Total cellulase activities of fungal strains were determined as described by Mandel and Sternberg (1976) 57 . Siderophore production by T. variabilis. Based on our previous experiment 16 , the highest concentrations of siderophore were obtained using King B medium and glucose medium. The media were inoculated with  www.nature.com/scientificreports www.nature.com/scientificreports/ disks from fresh PDA cultures of 48 P and 28 R and kept in an incubator shaker at 120 rpm at 28 °C for 7 days. The supernatants were obtained by centrifugation followed by filtration through 0.2 µM Minisart filters. ELISA spectrophotometer at 400 nm was used to detect siderophores concentrations using molar extinction coefficient ε = 20000 58,59 . The experiment had six replicates.

Effect of T. variabilis on morphology of P. aphanidermatum. The antagonistic activity of T. variabilis
isolates (48 P and 28 R) was checked against P. aphanidermatum in vitro as detailed in Halo, et al. 16 . The morphological study focused on main hyphae and their tips. 50 main and hypha tips of P. aphanidermatum near the inhibition zone were examined using a light microscope and scanning electron microscope (SEM, INSTUMENT JSM-5600). These hyphae were compared with P. aphanidermatum hyphae in PDA control plate. All hyphae were stained with cotton blue.
Biocontrol potential of T. variabilis isolates against damping-off diseases of cucumber and tomato. The effect of T. variabilis isolates (48 P and 28 R) on Pythium damping-off of cucumber and tomato was tested using three treatments and one control. 48 P and 28 R isolates, used in the experiment, were grown in PDB media in an incubator shaker at 28 °C and 120 rpm for 15 days to obtain fungal suspension (spore/mycelial). Also, this test used fresh cultures of P. aphanidermatum which were grown in PDA for 3 days at 28 °C. Seven sterilised seeds of cucumber or tomato were sown in each pot. Pots (12-cm in diameter) were autoclaved once. However, soil used in the pots was autoclaved twice.
Four replicate pots were kept for each treatment or control. The control pots were irrigated with 25 ml PDB media; the pots in first treatment were inoculated with 25 ml of fungal suspension containing spore and/or mycelia of 48 P or 28 R; the second treatment was treated with full plate of fresh PDA culture of P. aphanidermatum, 2 cm below soil surface; the pots in the third treatment was treated with full plate of fresh PDA culture of P. aphanidermatum and 25 ml of 48 P or 28 R fungal suspension 60 . The experiment was conducted at 28 °C for three weeks at 12-14 hr day length. After that, shoot length, fresh weight and dry weight were determined. Also seedlings survival rate was calculated by dividing the number of surviving seedlings by 7 (total seed sown) and then multiplying them by 100. This experiment was executed twice.
Statistical analysis. Data were analysed by IBM SPSS Statistics 24.0 using Chi-Square test for morphological study to compare performance of the isolates to the control. The Poisson test was used to compare oospore production by P. aphanidermatum treated with each of the culture filtrates. Independent sample t-test, One-way ANOVA and Duncan's Multiple Range Test were used to compare means of the different treatments. Each test is explained in the caption of each figure/table in the results section.

Data Availability
All data underlying this publication are available in the manuscript.