Diversity of Colletotrichum spp. on Aristolochia grandiflora: A Case Study in an Italian Botanical Garden

Abstract

During a survey conducted in autumn 2022 and spring 2023, extensive leaf spots were observed on Aristolochia grandiflora plants in the Botanical Garden “Angelo Rambelli” (Viterbo). To preserve the botanical garden collection and avoid disease spread, morphological and molecular identification of the causal agent were carried out. The results revealed three distinct Colletotrichum species belonging to the Colletotrichum boninense and Colletotrichum orchidearum species complexes, which have never been reported together within the same host and, for two of them, never in Italy. These findings may contribute to further extend the state of the art on leaf anthracnose, as well as provide new insights and molecular data for further phylogenetic studies.

1. Introduction

Aristolochia (Kingdom: Plantae; Phylum: Tracheophyta; Class: Magnoliopsida; Order: Piperales; Family: Aristolochiaceae; Genus: Aristolochia L.) is a genus of flowering plants in the Aristolochiaceae family, encompassing around 400 species of herbaceous perennials, undershrubs, or shrubs, many of which are rich in essential oils. These plants are widely distributed across tropical regions of Asia, Africa, and South America, where they thrive in a variety of ecosystems, from tropical rainforests and deciduous forests to more arid environments [1]. Renowned for their distinctive flower shapes, resembling pipes, Dutch shoes, or pelicans, Aristolochia plants have a long history of medicinal use. These plants have been long used to treat snake bites, fevers, and digestive issues [2]. However, several species contain aristolochic acids, compounds known for their toxic and carcinogenic effects, and, for these reasons, their use in herbal medicine has significantly declined in modern times, leaving more room for their ornamental function [3]. Aristolochia grandiflora stands out for its ornamental value, featuring enormous, velvety flowers that can be up to 30 cm in length and that are often mottled or striped in hues of white, purple, and green. Because of their ornamental value and aesthetic appeal, these plants are highly appreciated and requested by botanical gardens worldwide. These institutions not only showcase the unique beauty of species such as A. grandiflora, but they also play a pivotal role in biodiversity preservation [4,5]. Accordingly, frequent monitoring to ascertain the presence of symptoms becomes a crucial aspect of species preservation. Introducing plant species from diverse global regions into a controlled environment significantly raises the risk of non-native phytopathogen spread, and this plant congregation also facilitates the spread of pathogenic agents (or insect pests) to new host plants, such as Aristolochia, that may not be present in their native ranges. Early detection of threats through frequent screening of potential symptoms is, therefore, essential to prevent the spread of infections and ensure the health of native and non-native plants. Thus, regular inspections, proper sanitation, and careful isolation of infected plants can help mitigate the risk of disease outbreaks in botanical gardens [5].

This frequent screening is also applied at the Botanical Garden “Angelo Rambelli”, Tuscia University in Viterbo (Lazio, Italy). During regular inspections of the ornamental plants preserved, severe anthracnose-like symptoms were observed on A. grandiflora leaves.

Anthracnose is caused by fungal pathogens belonging to different groups, including the genus Colletotrichum. Different studies carried out over the years showed the harmfulness of many Colletotrichum species on annual crops of economic relevance, forest, and ornamental plants worldwide [6,7,8,9,10]. Damage can be observed during both pre- and post-harvest stages, and it is concentrated mainly on leaves, stem tubers, and seedlings [10,11,12,13].

Aristolochia plants inspected at the botanical garden showed symptoms extremely close to anthracnose caused by Colletotrichum species. These clues were the main inspiration for this study, which aimed to identify the causal agent of the observed symptoms to adopt precise control strategies and to prevent spread of the disease.

2. Materials and Methods

2.1. Sampling and Fungal Isolation

During two seasonal surveys conducted in autumn 2022 and spring 2023, anthracnose-like symptoms were observed in 70% of A. grandiflora leaves, whose plants are located in the tropical greenhouse of the Botanical Garden “Angelo Rambelli” of Tuscia University in Viterbo, Italy (Figure 1). Leaves were collected with sterile scissors, sealed in plastic bags, and brought to the laboratory for downstream processing.

To identify the symptom’s causal agent through fungal isolation, ten symptomatic leaves were rinsed with distilled water, sterilized with 70% ethanol for 30 s, rinsed twice with sterile water, and dried under a laminar flow. Leaf pieces of 3–4 mm were excised from the transition zone between diseased and healthy tissue, placed on Potato Dextrose Agar plates (PDA), and incubated at 25 °C. After 7 days of incubation, the obtained fungal colonies were further purified and classified in three different morphotypes according to their morphological traits. Among a total of twenty, eight representative isolates were selected and maintained in pure culture for further analysis.

2.2. DNA Extraction, PCR Amplification, and Sequencing

Genomic DNA was extracted from 100 mg of fresh mycelium from each of the twenty obtained isolates using the NucleoSpin^® Plant II Midi DNA extraction kit (Macherey-Nagel, Dueren, Germany), following the manufacturer’s instructions, and stored at −20 °C.

Molecular characterization was initially carried out by PCR amplification and Sanger sequencing of the ribosomal internal transcribed spacer (ITS). Then, eight isolates were selected and five additional genes were amplified and sequenced for a more detailed phylogenetic analysis: β-tubulin (TUB2), actin (ACT), partial chitin synthase (CHS-1), histone 3 (HIS3), and a 200 bp intron of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) regions. The sequence of the primers used and the size of the amplicons are reported in Table S1. Forty nanograms of template DNA were incubated with 1x GoTaq Green MasterMix (Promega Corporation, Madison, WI, USA) and 0.5 μM of both forward and reverse primers in a final volume of 25 μL. PCR reactions were carried out according to [14]. The thermal cycle consisted of an initial step of 3 min at 94 °C, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing for 30 s at 55 °C for the ITS regions using the ITS1-ITS4 primers, 30 s at 51 °C for the TUB2 using primers T1 and BT2b, 30 s at 54 °C for the actin gene with ACT-512F and ACT-738R primers, 30 s at 58 °C for the CHS gene using CHS-79F and CHS-354R primers, 30 s at 57 °C for the HIS3 gene using CYLH-3F and CYLH-3R primers, followed by 30 s of elongation at 72 °C and a final extension at 72 °C for 5 min. An aliquot of the amplified products was visualized on a 1.2% agarose gel and the rest sent to Eurofins Genomics (Konstanz, Germany) for Sanger sequencing.

2.3. Identification and Multi-Locus Phylogenetic Analysis

The electropherograms obtained from Sanger sequencing were visually inspected with FinchTV v.1.4 (available at https://digitalworldbiology.com/FinchTV, accessed on 4 October 2023). A BLASTn search with the obtained ITS sequences suggested that the isolates belonged to C. boninense and C. orchidearum species complexes. ITS sequences of Colletotrichum isolates belonging to the same species complexes according to [10,12,13] and whose sequences were available on the GenBank database (Table S2) were used for phylogenetic analysis. Three strains of C. gloeosporioides (CBS 112999, Col-41, and Col-69) were included as outgroups. The sequences were adjusted using UGENE v48.1 [15], concatenated, and further aligned with MUSCLE v3.8.31 (Edgar 2004). The alignment file was then used as input to build a maximum likelihood (ML) phylogenetic tree using RAxML-HPC v8.2.12 [16], and set with a GTRCATI algorithm as a substitution model and 1000 bootstraps. The trees were visualized using FigTree v1.4.4 (available at http://tree.bio.ed.ac.uk/software/figtree/, accessed on 4 October 2024) and further edited with Inkscape v0.92 (available at https://inkscape.org, accessed on 4 October 2024).

Based on these results, a phylogenetic analysis was carried out using the six selected loci of eight representative isolates. Two separated phylogenetic trees were built as described previously, one for the C. orchidearum species complex and one with the isolates belonging to the C. boninense species complex.

2.4. Morphological Characterization and Growth Rates

2.4.1. Morphological Characterization

The eight selected isolates were maintained in pure culture by transferring single hyphae to plates containing various substrates: standard PDA and two media specific for morphological characterization, Synthetic Nutrient-poor Agar medium (SNA, [17]), and Oatmeal Agar (OA, [18]). These plates were then incubated at 25 °C and 100% humidity in the dark. After ten days, the following morphological characteristics were observed: colony shape and color, conidiomata, conidial shape and size, hyphae, and appressoria. The mean lengths and widths of 30 randomly selected conidia from each isolate were measured using 40× magnification in a microscope (Leitz, Wetzlar, Germany).

2.4.2. Mycelium Growth and Data Analysis

Four millimeters diameter plugs (∅) of each Colletotrichum isolate, grown on PDA at 25 °C for one week, were transferred to the center of four new 9 cm diameter (∅) per each PDA, SNA, and OA media plates. After seven days at 25 °C and a relative humidity of 100% RH, the fungus did not reach the border of the dishes, and the mycelium radial growth rate was measured with a ruler (±1 mm) in all four directions from the center of the plug, according to the protocols of Drais et al. [19] and Brugneti et al. [6].

The collected data were imported within the R environment (v.4.2.3), where the isolate and the substrate were considered as factors, while the plate and the orthogonal directions of the measured radii were considered as a random variable. Three different analyses were carried out: the first one to identify differences between the three Collettothricum species regardless of the media, the second one to check the differences among media regardless of the species, and the third one to analyze the mycelium growth differences among the different species on the three different media. Before each analysis, the normality of the dataset was checked through a Shapiro–Wilk test using the shapiro.test() function and through a visual inspection of the Quantile–Quantile (Q–Q) plot using the qqmath() function within the lattice R package. In case the dataset was not normally distributed, as for the first and second analyses, the best expression to transform the dataset was chosen using the bestNormalize() function within the bestNormalize R package.

After transformation, for the first and second analyses, the dataset was analyzed through a Linear Model (LM), considering the radii as numeric values, the species and the substrate as independent variables, and the plate and the orthogonal direction (N-W-S-E) as random variables. For the third analysis, instead, the measured radii were considered as numerical values, the substrate as a factor, and the plate and the orthogonal direction as a random effect. Calculations were carried out using the lmer() function within the lme4 R package. The linear model was followed by a Bonferroni post hoc test (=0.05), carried out through the emmeans() function within the R package emmeans, the pairs() function within the R package multcompView, and the cld() function within the R package multcomp.

The script and the dataset to fully reproduce the results of the present work are publicly available at https://github.com/lucaros1190/Colletotrichum-VTBGarden (accessed on 3 August 2024).

2.5. Pathogenicity Tests

Pathogenicity tests were conducted following the method described by Moral et al. [20] on healthy, lesion-free leaves collected from healthy plants. For the inoculum, a conidial suspension (10⁵ conidia/mL) of each isolate was prepared in sterile distilled water. Detached asymptomatic A. grandiflora leaves were surface sterilized by immersion in 3% NaClO for 2 min, washed twice with sterile-distilled water, and air dried on sterile filter paper under a laminar flow hood. Ten leaves for each isolate were then inoculated with 15 μL of conidial suspension at four different sites on the leaf surface. Control leaves were inoculated with sterile distilled water using the same method. The inoculated leaves were covered with a plastic box to maintain high relative humidity, and incubated in the dark at 25 °C.

An additional pathogenicity test was conducted by inoculating a mixture of the three Colletotrichum species together. One isolate for each Colletotrichum species was chosen to prepare a pooled conidia suspension (10⁵ conidia/mL), following the method used by Garcia-Lopez (2023) [21], and inoculation was carried out as described previously. Symptoms were observed 7 days after inoculation, and to fulfill Koch’s postulates, the three Colletotrichum pathogens were re-isolated from symptomatic leaves, and their identity was confirmed both morphologically and molecularly.

3. Results

3.1. Fungal Isolates Identification

A total of twenty isolates were obtained from ten symptomatic leaves of A. grandiflora and, based on their morphological traits, recognized as Colletotrichum spp. The BLASTn results of the twenty ITS sequences on the GenBank database indicated that the isolates belonged to two specific complexes. The isolates C2, C3, C4, C5, C6, C9, C12, C13, C14, C16, and C19 belonged to the Orchidearum species complex and were grouped in the C. cattleyicola cluster (Figure S1), while the isolates C1, C7, C8, C10, C11, C15, C17, C18, and C20 clustered within the Boninense species complex, and the isolates clustered in the C. karstii and C. boninense group (Figure S2). Among those, eight representative isolates were chosen for phylogenetic analysis and for a detailed morphological characterization based on colony shape, color, growth rate, and conidial morphology, and for phylogenetic analysis.

Colletotrichum karstii Y.L. Yang, Zuo Y. Liu, K.D. Hyde and L. Cai, Cryptogamie Mycologie 32: 241. 2011 [22] Figure 2: The C1 isolate on PDA produced gray and pale-orange aerial mycelium and appeared orange/reddish from the bottom, with orange conidial masses submerged in the mycelium. Conidiogenous cells were cylindrical, hyaline, straight with rounded ends, with contents appearing granular, and 12–18 × 5.5–7.5 μm in size. Ascospores were allantois and pear-shaped, inequilateral, and often straight on the inner side, had rounded apices, tapered towards base, and were 12–19 × 3.5–7 μm. Colonies on OA were flat with an entire margin, buff to rosy buff to pale salmon, covered with orange to off-white conidiomata, lacking aerial mycelium, reverse buff, and orange to rosy. Colonies on SNA were flat with an entire margin and hyaline.

Colletotrichum cattleyicola Damm and Toy Sato, MycoBank MB824220 [12], Figure 3: All five isolates (C2, C3, C4, C5, and C19) formed gray-white aerial mycelium and appeared gray to yellow from the bottom on PDA. Conidiophores were cylindrical and hyaline. Conidia were in yellowish white masses, 14–21 × 3.5–7 μm, one-celled, smooth-walled, hyaline, cylindrical, straight, and rounded at the ends. Sexual morphs were not observed. Colonies on OA were flat with an entire margin, covered with whitish-gray aerial mycelium, with the same colors on the reverse, and abundant black conidiomata on the surface of the medium. Conidial appeared in whitish masses. Colonies on SNA were flat with entire margins, and had agar medium partially covered by hyaline, whitish slightly floccose mycelium, and the same colors on the bottom.

Colletotrichum boninense Moriwaki, Toy, Sato and Tsukib., Mycoscience 44(1): 48. 2003 [23], Figure 4: Two isolates (C7 and C10) showed an aerial mycelium that was felty to light gray, abundantly slimy, and had orange conidial masses, with the same colors on the bottom of the PDA medium. Conidia were 14–19 × 3.5–6.5 μm, hyaline, smooth-walled, aseptate, straight, and cylindrical, with a round apex and round base with two big polar guttules. Ascomata (perithecia) were observed after 4 weeks of culture on PDA, varying from subglobose to pyriform, and were glabrous and medium brown. Asci were cylindrical-clavate, 40–58 × 11–17 µm, eight-spored, and with a rounded or slightly tapered, slightly curved apex. Colonies on OA were flat with an entire margin; the surface was covered with felty white, rosy buff, or very pale orange aerial mycelium; the center aerial mycelium was white to gray; they were reverse, light brown or pale orange. Colonies on SNA were flat with a slightly undulate margin, and were hyaline with a felty white aerial mycelium on filter paper.

3.2. Phylogenetic Analysis

For a deeper taxonomic classification, five additional genes (GAPDH, CHS-1, HIS3, ACT, and TUB2) were amplified, Sanger sequenced, and all the related sequences deposited on the NCBI Genbank database (Table S3). The sequences were concatenated and aligned to other Colletotrichum isolates belonging to the C. boninense species complex or the C. orchidearum species complex, respectively. In both ML phylogenetic trees (Figure 5 and Figure 6), each isolate clustered within its own expected group: the C1 isolate among C. karstii; C7 and C10 isolates among C. boninense; and C2, C3, C4, C5, and C19 among C. cattleyicola. Both trees were rooted on the C. gloeosporioides species complex. The robustness of each clustering is further supported by the high bootstrap values indicated on the trees.

3.3. Mycelium Growth over Species and Substrates

Overall, the species showed a different mycelium growth rate, independent from the substrate: the isolate C. boninense was slower, while C. cattleyicola was the fastest (Figure S3). All species were statistically different from each other according to the LM analysis, all with a p < 0.0001. Statistical differences were assessed on the substrates as well (Figure S4), with no distinction among the species: mycelium extension was significantly different between the PDA and OA substrates (LM, df = 376, t = 6.319, p < 0.0001) and the SNA and OA substrates (LM, df = 376, t = 4.749, p < 0.0001). No statistical differences were observed between PDA and SNA substrates (LM, df = 376, t = −1.570, p = 0.3516).

An in-depth analysis of the dataset provided interesting results on the different responses of the species to the substrates (Figure 7). The highest and lowest mycelium extensions of C. boninense were observed on the OA and PDA substrates, respectively. Overall, all growth was statistically different in all the substrates (LM, df = 89: OA-PDA, t = 7.685, p < 0.0001; OA-SNA, t = 4.229, p = 0.0002; PDA-SNA, t = −3.456, p = 0.0025). The highest and lowest mycelium extension rates of C. karstii, instead, were observed on the OA and SNA substrates, respectively. Statistical differences in growth were assessed between the OA and PDA substrates (LM, df = 45, t = 6.070, and p < 0.0001), and the OA and SNA substrates (LM, df = 45, t = 6.193, and p < 0.0001). No differences in growth on the substrates PDA and SNA were observed (LM, df = 45, t = 0.123, and p = 1). A similar scenario was observed for C. cattleyicola: statistical differences in mycelium growth were assessed between the OA and PDA substrates (LM, df = 230, t = 5.841, and p < 0.0001) and the OA and SNA substrates (LM, df = 230, t = 5.817, and p < 0.0001), while there were no differences between the PDA and SNA substrates (LM, df = 230, t = −0.024, and p = 1). C. cattleyicola grew faster in the OA and slower in the PDA substrate.

3.4. Pathogenicity Tests

Pathogenicity tests were carried out on ten healthy A. grandiflora leaves for each of the eight Colletotrichum isolates. Seven days post-inoculation, all leaves inoculated with one of the Colletotrichum isolates showed necrotic spots, more or less extended depending on the isolate and the species (Figure 8). Abundant acervuli formation was observed on the necrotic tissue (Figure 2J, Figure 3L, and Figure 4J), while control leaves remained healthy (Figure 8). The same results and same symptoms were obtained from the pathogenicity test carried out with the pooled inoculation (Figure S5). Nonetheless, the morphology of the fungal colonies re-isolated from the symptomatic leaves was close to the original isolates used for inoculation, satisfying Koch’s postulates.

4. Discussion

The Colletotrichum genus is among the top ten fungal genera encompassing 340 currently recognized species, grouped in 20 different species complexes, and able to infect 3400 different host species. In this study, we report the coexistence of three Colletotrichum species (C. karstii, C. boninense, and C. cattleyicola) on a single host plant, the ornamental tropical plant Aristolochia grandiflora. These isolates belong to two distinct Colletotrichum species complexes, the boninense species complex (C. karstii and C. boninense isolates) and orchidearum species complex (C. cattleyicola), respectively. The simultaneous coexistence of different Colletotrichum species within the same host is common, and usually there is one species that is predominant over the others [9,20].

From a morphological point of view, the identified species showed a high similarity with the ex-type isolates described in previous studies [12]. However, relying solely on morphological traits or on molecular identification through ITS region is not sufficient for a proper identification, especially for complex fungal genera such as Colletotrichum or Fusarium [24,25,26]. Conversely, a multi-locus phylogenetic analysis based on a combination of six loci (ITS, GAPDH, CHS-1, HIS3, ACT, and TUB2) provided a more reliable classification of the isolated species, showing that the isolates formed three well-supported clades with known species.

C. cattleyicola has been previously reported only on Cattleya spp. [12]. The only sequences available on the GenBank database refer to one isolate from Belgium from Cattleya sp. roots (CBS 170.49), and one isolate from Cattleya sp. from Japan, Mie Prefecture (MAFF 238321). Thus, our study represents the first report of this species in Italy and on Aristolochia plants, expanding the host range of this still poorly known species.

Conversely, species belonging to the C. boninense complex, C. karstii, and C. boninense, are known to infect a large range of hosts from numerous different plant families [9,11]. Although C. karstii have been repeatedly reported in several species in Italy [9,27,28,29], to date, there are no reports of C. boninense s.s. in any hosts.

The genus Colletotrichum has been reported as an endophyte of Aristolochia by Zhi-jun et al. [30] and in Brazil in 2017 (C. gloeosporioides, GenBank accession numbers MF076612–MF076615, unpublished data) but without causing any visible disease symptoms, while Tekade and Mohod [31] observed leaf blight caused by C. dematium on A. bracteata leaves.

5. Conclusions

To the best of our knowledge, this study represents the first report of C. karstii, C. boninense, and C. cattleyicola together within the same host, A. grandiflora, as well as the first report of C. boninense and C. cattleyicola in Italy. Our findings underline even more the importance of botanical gardens not only for plant species preservation but also to better understand the harmfulness of autochthonous and alien pathogens and pests on different host species. As shown by this study, frequent screening of plants can be extremely helpful to gather novel information that can be further helpful for plant nurseries and ornamental plant sellers.