Occurrence and infective potential of Colletotrichum gloeosporioides isolates associated to Citrus limon var Eureka.

A collection of 37 fungi associated to Italian lemon plants with disease symptoms, was obtained. Ten genera including Aspergillus, Alternaria, Nigrospora, Lasiodiplodia, Dothideomycetes, Pleurostoma, Setosphaeria, Penicillium, Fusarium and Colletotrichum were identified by using ITS1-5.8S-ITS2, D1/D2 26S and COX1 loci. The last three genera were abundant on the damaged fruits, being Colletotrichum the more abundant (32.4 %). CaInt2 and CgInt primers support the identity of these isolates as C. gloeosporioides. Variability, inferred by rep-PCR and multilocus sequence analysis shows genetic differences among the C. gloeosporioides isolates. Infective profile evaluated in Colletotrichum isolates shows different leave infection percentages (26 to 60 %). SEM analysis showed mycelium, spores and appressoria on the leaves of selected Colletotrichum isolates. Specifically, the AL-05 and AL-13 isolates showed a high chitin deacetylase activity (CDA) peaking at 1.2 U/mg protein in AL-13. This is the first report on C. gloeosporioides infecting Italian lemon leaves in Mexico.

Introduction

Citrus trees constitute a fruit crop widely cultivated in tropical and subtropical areas around the world [1,2]. Mexico was the fourth place worldwide in lemon exports in 2017 with 729,650 tons. The Italian lemon (Citrus limon var Eureka) is one of the main species produced in Mexico, being the state of Tamaulipas the main producer with around 100,000 tons in 2017 [3]. Citrus crops are severely affected by fungal pathogens, which infect the fruits in pre- and post-harvest stages. Typical pre-harvest diseases are phomosis stem-end rot (Lasiodiplodia theobromae, Diaporthe citri), brown rot (Phytophthora citrophthora), brown spot (Alternaria alternata f. sp. citri) and anthracnose (Colletotrichum gloeosporioides) [1]. Colletotrichum spp. were recently included in the list of the 10 most important plant pathogenic fungi in the world [4]. Infections by species of Colletotrichum can result in several diseases, including anthracnose, fruiting after flowering and postharvest anthracnose [5]. In Mexico, there are reports that C. acutatum and C. gloeosporioides cause diseases in fruits of Mexican lime (Citrus aurantiifolia) [6,7] and grapefruit (Citrus paradisi) [8]. Differentiation among Colletotrichum species responsible for disease epidemics is critical for developing and implementing effective control strategies [9]. Occurrence, diversity and pathogenicity of Colletotrichum spp. associated with Citrus trees have been explored by a multi-locus approach in diverse European orchards and gardens, identifying members of the C. gloeosporioides, C. boninense, and C. acutatum species complexes, and revealed to C. gloeosporioides strains as the most virulent specie on these fruits [10]. Similarly, diverse Colletotrichum isolates were isolated and identified by a multilocus approach on pomegranate fruits in the southeastern United States. These isolates were grouped within the C. acutatum and C. gloeosporioides species complexes with differences in its pathogenic profiles level being more aggressive on inoculated fruits or leaves, respectively [11]. According to the above, this study addresses the analysis of the fungal community associated to different plant tissues with disease symptoms in Italian lemon trees (Citrus limon var. Eureka) cultivated in Tamaulipas state, Mexico. Specifically, genetic variability and infective profile of representative Colletotrichum isolates were analyzed.

Material and methods

Study site and sampling strategy

Fungal isolate collection was established in a Citrus limon var. Eureka orchard in the Tamaulipas state (N 24° 04.161′ WO 98° 55.712′) of Mexico. This citrus orchard has conditions of controlled humidity by irrigation and with a periodical fungal chemical control by fumigation. Samples were collected from trees with visible disease symptoms. In this orchard, the last fumigation was carried out at least three months previous to sampling. Random sampling was used to survey the fungal diversity from persistent calyxes, leaves with brown lesions, and damaged fruits (yellowish immature, broke, or dry and broken fruits) from eleven trees within an area of 4 km2.

Isolation and morphological identification of associated fungi

Fungal isolates were obtained from intracellular (2×2 mm) sections from plant damage tissues (fruits, calyxes, and leaves). These sections were then cut and disinfected with sodium hypochlorite (1% v/v) for 5 min and washed 3 times with sterile distilled water. Then, these were placed on Petri dishes containing Potato-Dextrose Agar (PDA, BD Bioxon, USA) and incubated for 5 days at 29°C. The fungi were subcultured from agar plugs obtained from the edge of the colonies to recover actively growing hyphal tips, to obtain pure isolates. Preliminary morphological identification by sporulation (size and shape of the spores) was used to characterize different fungal species. The selected isolates were kept in PDA at 4°C after its use, and preserved in glycerol 86% (v/v) at -70°C.

Molecular identification

Fungal isolates were grown in 100 mL of Potato Dextrose broth for five days at room temperature under rotary shaking at 250 rpm. Cells were harvested by filtration through a piece of filter paper and then were washed with distilled water. Fresh biomass (500 mg) was homogenized in liquid nitrogen and transferred to a 1.5 mL tube containing 1 mL of TEN buffer (100 mMTris-HCl, 50 mM EDTA, 500 mMNaCl, pH 8.0) and vortexed for 1 min. After centrifugation for 10 min at 10.000xg at RT, the pellet was re-suspended in 1 mL of TEN buffer and transferred to a fresh 1.5 mL tube to continue with phenol-chloroform extraction method [12]. ITS1‐5.8S‐ITS2 and D1/D2 LSU regions were amplified using the primers ITS‐1 5´TCCGTAGGTGAACCTGCGG3´ and ITS‐4 5´TCCTCCGCTTATTGATATGC3´ and NL1 5´GCATATCAATAAGCGGGAGGAAAAG3´ and NL4 5´GGTCCGTGTTTCAAGACGG3´, respectively [13,14]. For the amplification of the cytochrome oxidase subunit 1 (COX1) was necessary three primers sets specifically targeted to Penicillium/Aspergillus (Pen-F1 5´GACAAGAAAGGTGATTTTTATCTTC3´ and AspR1 5´GGTAATGATAATAATAATAATACAGCTG3´ [15]), Fusarium (AHyFu-F 5´CTTAGTGGGCCAGGAGTTCAATA3´ and AHyFu-R 5´ACCTCAGGGTGTCCGAAGAAT3´), and diverse fungal genera, including to Colletotrichum, (Pez-F 5´TCAGGRTTAYTAGGWACAGCATTT3´ and Pez-R 5´ACCTCAGGRTGYCCGAAGAAT3´ [16]).

Amplification of the different gene regions was performed in 50 µL of PCR reaction mix containing 5 µL of reaction buffer, 5 units of Taq polymerase (Gibco-BRL, Rockville, MD), 1.5 mM MgCl2, 5µM of each primer, 120 µM dNTPs and 50 µg of template DNA. Amplification was carried out with an initial denaturation of 5 min at 94 ºC, followed by 30 cycles of 1 min at 94 ºC, 45 s at 58 ºC and 60 ºC for D1/D2 26S and ITS1–5.8S–ITS2 regions, respectively, and 1 min at 72 ºC, with a final extension of 5 min at 72 ºC. COX1 amplification (with the different primer sets) was conducted in PCR touchdown conditions with an initial denaturation of 3 min at 94 ºC, followed by 15 cycles of three steps beginning with 30 s at 94 ºC, 45 s at 65 ºC and 1 min at 72 ºC. Alignment temperature dropped 3°C every 3 cycles, ending at 50°C. Finally, an extension step 10 min at 72 ºC was done. Amplicons were confirmed by electrophoresis on 1% agarose gel in TBE buffer (108 gTrisHCl, 55g boric acid, 40 ml of EDTA 0.5mM, pH 8.0 on 1 liter of distilled water). The PCR products were directly sequenced in an automatic sequencer (ABI 377 DNA sequencer, Applied Biosystems, USA) by using the BigDye terminator cycle sequencing ready reaction kit according to the instructions of manufacturer (Applied Biosystems, USA). The nucleotide sequences of each locus were aligned and compared with those at GenBank database. In addition, nucleotide sequences obtained in this study were deposited in the GenBank database, as summarized in Table 1. For each gene, the sequences from different isolates and their closest relatives were aligned. The dendrograms were constructed by an analytical program (MEGA version 10) using the Maximum Likelihood method based on the Tamura-Nei model. The reliability of the clusters was evaluated by bootstrapping with 1000 replicates.

Table 1

Origin and identity of the isolates obtained in different tissues of Italian lemon (Citrus limon var. Eureka)

Isolates	Identity	28S/26S (LSU)	ITS (ITS1-5.8S-ITS2)	COX1	Tissue origin
AL-01	Colletotrichum gloeosporioides	KC341907.1	KC341946.1	MZ407497	Persistent calyx
Al-02	Colletotrichum gloeosporioides	KC341908.1	KC341947.1	NA	Yellowish immature fruits
AL-03	Colletotrichum gloeosporioides	KC341909.1	KC341948.1	MZ407498	Leaf with brown lesions
AL-04	Colletotrichum gloeosporioides	KC341910.1	KC341949.1	MZ407499	Leaf with brown lesions
AL-05	Colletotrichum gloeosporioides	KC341911.1	KC341950.1	MZ407500	Persistent calyx
AL-06	Colletotrichum gloeosporioides	KC341912.1	KC341951.1	MZ407501	Yellowish immature fruits
AL-08	Colletotrichum gloeosporioides	KC341914.1	KC341953.1	MZ407502	Necrotic leaf
AL-09	Colletotrichum gloeosporioides	KC341915.1	KC341954.1	MZ407503	Persistent calyx
Al-10	Colletotrichum gloeosporioides	KC341916.1	KC341955.1	MZ407504	Persistent calyx
AL-11	Colletotrichum gloeosporioides	KC341917.1	KC341956.1	MZ407505	Leaf with brown lesions
AL-12	Colletotrichum gloeosporioides	KC341918.1	KC341957.1	MZ407506	Leaf with brown lesions
AL-13	Colletotrichum gloeosporioides	KC341919.1	KC341958.1	MZ407507	Persistent calyx
AL-14	Fusarium sp.	KC341920.1	KC341959.1	NA	Persistent calyx
AL-15	Fusarium sp.	KC341921.1	KC341960.1	MZ407508	Yellowish immature fruits
AL-16	Fusarium sp.	KC341922.1	KC341961.1	NA	Broken fruit
AL-17	Fusarium sp.	KC341923.1	KC341962.1	MZ407509	Broken fruit
AL-18	Fusarium sp.	KC341924.1	KC341963.1	NA	Yellowish immature fruits
AL-20	Fusarium sp.	KC341926.1	KC341965.1	NA	Broken fruit
AL-21	Fusarium sp.	KC341927.1	KC341966.1	MZ407510	Dry and broken fruit
AL-22	Fusarium sp.	KC341928.1	KC341967.1	NA	Broken fruit
AL-23	Fusarium sp.	KC341929.1	KC341968.1	MZ407511	Yellowish immature fruits
AL-24	Fusarium sp.	KC341930.1	KC341969.1	NA	Broken fruit
AL-25	Aspergillus niger	KC341931.1	KC341970.1	NA	Leaf with brown lesions
AL-26	Aspergillus niger	KC341932.1	KC341971.1	MZ407512	Leaf with brown lesions
AL-27	Aspergillus niger	KC341933.1	KC341972.1	NA	Leaf with brown lesions
AL-28	Aspergillus niger	KC341934.1	KC341973.1	MZ407513	Leaf with brown lesions
AL-29	Aspergillus niger	KC341935.1	KC341974.1	MZ407514	Leaf with brown lesions
AL-30	Aspergillus niger	KC341936.1	KC341975.1	NA	Dry and broken fruit
AL-31	Penicillium sp.	KC341937.1	NA	NA	Dry and broken fruit
AL-32	Alternaria tenuissima	KC341938.1	KC341976.1	NA	Yellowish immature fruits
AL-33	Alternaria tenuissima	KC341939.1	KC341977.1	NA	Leaf with brown lesions
AL-34	Setosphaeria	KC341940.1a	KC341978.1	NA	Leaf with brown lesions
AL-35	Dothideomycetes	KC341941.1a	KC341979.1	NA	Broken fruit
AL-36	Lasiodiplodia pseudotheobromae	KC341942.1	KC341980.1	NA	Leaf with brown lesions
AL-37	Nigrospora oryzae	KC341943.1	KC341981.1	NA	Dry and broken fruit
AL-38	Penicillium italicum	KC341944.1	KC341982.1	MZ407515b	Leaf with brown lesions
AL-39	Pleurostoma richardsiae	KC341945.1	KC341983.1	NA	Leaf with brown lesions

These isolates received its name according to ITS identity.

This isolate received its name according to COX1 identity. NA: No responsive to gene marker amplification.

Molecular differentiation of the Colletotrichum genus

Specific PCR primes CgInt (5´-GGCCTCCCGCCTCCGGGCGG-3´) and CaInt2 (5´-GGGGAAGCCTCTCGCGG-3´) for C. gloeosporioides and C. acutatum, respectively, were selected. These primers were used in combination with the ITS-4 primer as describes [17] with the Colletotrichum isolates of this study and the reference strains Cg-06 for C. gloeosporioides, and Ca-93 and Ca-96 for C. acutatum, kindly provided by the Laboratory of Plant Biotechnology at CBG–IPN.

Amplification was performed in 25 µL of PCR reaction mix containing 2.5 µL of reaction buffer, 2.5 units of Taq polymerase (Gibco-BRL, Rockville, MD), 1.5 mM MgCl2, 5µM of each primer, 120 µM dNTPs and 50 µg of template DNA. Amplification was carried out with an initial denaturation of 5 min at 94 ºC, followed by 30 cycles of 1.5 min at 94 ºC, 2 min at 55 ºC and 3 min at 72 ºC, with a final extension of 10 min at 72 ºC. Amplicons were visualized by electrophoresis on 1.5 % agarose gel in TAE 1X buffer (40mM Tris, 20mM acetic acid, and 1mM EDTA) using SYBR Green (Invitrogen, USA).

In addition, genetic variability of these Colletotrichum isolates was explored by using the rep-PCR analysis. This methodology was selected as it has proven to be effective in identifying fungal species [18]. Rep-PCR amplification was performed in 25 µL of PCR reaction mix containing 2.5 µL of reaction buffer, 5 units of Taq polymerase (Gibco-BRL, Rockville, MD), 5 μM of each primer REPIR-1 5′IIIICGICGICATCIGGC3′ and REP2-I 5′ICGICTTATCIGGCCTAC3′ [19], 1.5 mM of MgCl2, 120 µM of dNTPs and 50 µg of template DNA. Amplification was carried out with an initial denaturation of 5 min at 94 ºC, followed by 30 cycles of 30 s at 94 ºC, 1 min at 47 ºC and 4 min at 65 ºC, with a final extension of 7 min at 72 ºC. PCR products were separated by electrophoresis in an agarose gel (2 %; w/v) at 80 V during 4 h in TAE 1X buffer (40 mM Tris, 20mM acetic acid, and 1mM EDTA), and visualized using SYBR Green (Invitrogen, USA). Rep-PCR bands were analyzed using the PyElph program with a threshold of 10, and a filter parameter of 3. Absence/presence band information was used to build a UPGMA tree to compare the amplification profile of the Colletotrichum strains selected.

Genetic variability was also supported in the C. gloeosporioides isolates by a concatenated multilocus analysis. For this, the different DNA regions (D1/D2 26S, ITS1–5.8S–ITS2 and COX1) were concatenated using the Bioedit software. Then, a dendrogram was constructed using the MEGA version10 program by using the Maximum Likelihood method based on the Tamura-Nei model. Clusters reliability was evaluated by bootstrapping with 1000 replicates. Concatenated sequences (ITS1–5.8S–ITS2; LUXP01000012.1:2474961-2475497, D1/D2 26S; LUXP01000012.1:2474386-2474948 and COX1; KR349346.1:2552-3040) of the reference C. acutatum strain KC05 were used as outgroup only with aim comparison.

Infection assay

Infection assays are performed on detached Italian lemon leaves of approximately 5 to 6 cm in length. The leaves were disinfected in 1% sodium hypochlorite solution for 5 minutes, and then washed three times with distilled water. An agar plug (4 mm in diameter) cut from the mycelium grown in PDA for 5 days was placed on the upper part of the leaves. The experiment was carried out with twenty repetitions per treatment. Inoculated leaves were placed in humid conditions inside Petri dishes with a moistened and sterile filter paper at the bottom of the dish. These were incubated in fluorescent light for 7 days at room temperature [20]. The pathogenic potential was recorded by measuring the diameter of the infection in the damaged area of the leaves. Leaves without inoculation (with and without sterile agar plugs) were used as negative controls. In addition, three isolates from avocado were used as biological controls, which includes two C. acutatum (Ca-93 and Ca-96) and one C. gloeosporioides (Cg-06). Three isolates of this study (Italian lemon) but from different fungal species were also used to compare the pathogenic potential of the Colletotrichum isolates. These were Penicillium digitatum (AL-38) and two Fusarium sp. (AL-20 and AL-21).

Scanning electron microscopy analysis

Some representative Colletotrichum isolates (AL-05, AL-06 and AL-13), according to their infection profile, were selected to further experiments. Leaf infection process of these isolates was observed by Scanning Electron Microscopy (SEM) as follow: Citrus leaves were briefly damaged with a sterile toothpick and inoculated with 20 µL of a concentration of 1×105 conidia/mL of the selected Colletotrichum isolates. SEM was carried out after 6 days of incubation at 29°C. The samples were prepared by cutting square pieces (0.5×0.5 cm) of the damaged leaves with mycelial growth. Samples were fixed with glutaraldehyde solution (5% v/v) in 0.1 M phosphate buffer pH 7.4 for 24 h at 4°C. Then, osmium tetraoxide was added in 0.1 M phosphate buffer pH 7.4 and allowed to react for 2 h. Then, the samples were dehydrated in ethanol series at concentrations of 30 to 100% in 30 min steps. Finally, samples were dried to a critical point, covered with gold, and observed by SEM (JJSM-5900 LV; JEOL, USA).

Chitin deacetylase activity

Two Colletotrichum isolates (AL-05 and AL-13) with differential infection profiles were selected to explore its chitin deacetylase (CDA) activity. These isolates were cultured in glutamic acid broth [21] for 5 days. Culture medium was then collected and homogenized with 0.1 M potassium phosphate buffer pH 6.5 on a 1:1 (v/v) ratio. Finally, the homogenized solution was filtered through using a 0.45 µm filter to recover the crude enzymatic extract. CDA activity was determined using the method reported by [22] with some modifications. In a typical procedure, 50 µL of the crude enzymatic extract were mixed with 100 µL of 50 mM tetraborate buffer pH 8.5 and 100 µL of ethylene-glycol-chitin, the mixture was incubated at 37°C for 30 min. To stop the reaction 250 µL of 5% (w/v) KHSO4 and 250 µL of 5% (w/v) NaNO2 were added. After 15 min 250 µL of N2H6SO3 were mixed, 5 min later other 250 µl of 0.5% (w/v) 3-methyl-2-benzothiazolon hydrazine. The tubes were boiled for 3 min and chilled with cold water. Finally, 250 µL of 0.5 % (w/v) FeCl3 was added. The protein content in the crude extract was determined by Lowry-Peterson assay using serum albumin bovine as standard. The absorbance was read on a DU 650 spectrophotometer (Beckman Coulter, USA) at 650 nm. A unit of CDA was defined as the amount of enzyme required to release 1 μmol of acetate per minute at standard condition. CDA specific activity was expressed as units per mg of soluble protein.

Results

Fungal isolation and identification

A total of 37 isolates of fungi associated with different tissues of Italian lemon plants were obtained. Particularly, samples of fruits with evident symptoms of anthracnose (yellowish immature, broke, or dry and broke fruits), persistent calyxes, and leaves with brown lesions were considered (Table 1). Most of the fungal isolates were identified by the combined analysis of two gene markers (ITS1‐5.8S‐ITS2 and D1/D2 LSU regions; Fig. 1, Fig. 2), although some Fusarium, Penicillium, Aspergillus, and the most of the Colletotrichum isolates were also identified with the COX1 gene (Fig. 3). In general, gene markers allowed a reliable identification of most isolates to the genus level, and most of these exhibited in average a high blast identity (99%) with accessions of their assigned specie (Fig. 1; Table 1). The dendrograms presented a reliable separation at the genus level between the different taxa, with a minimum distance between clades of 0.10. In general, the genera identified were Colletotrichum, Fusarium, Aspergillus, Alternaria, Penicillium, Nigrospora, Lasiodiplodia, Dothideomycetes, Pleurostoma and Setosphaeria. Besides, Colletotrichum, Fusarium and Penicillium were the more abundant genera in the damaged fruits (Table 1). Phylogenetic analysis performed by separate (Fig. 1, Fig. 2, Fig. 3) show some differences in a few isolates, particularly in the AL-34 and AL-38 isolates, which were identified in this study as Setosphaeria spp. and Penicillium italicum, based mainly on their ITS1‐5.8S‐ITS2 and COX1 loci, respectively. Selection of this gene marker to make the final identification was based on this region outperformed that of the D2 region in the identification of fungal species [23]. Although most of the isolates were analyzed with at least two gene regions, in some species, there are few reference sequences uploaded to the Genbank, on which case, consistency was not achieved in the identification of these isolates, such as is the case of the Setosphaeria and Dothideomycetes genera. Importantly, consistency is necessary for the Fusarium isolates, as the phylogenetic analysis of the ITS1-5.8S-ITS2, D1 / D2 LSU and COX1 (Fig. 1, Fig. 2, Fig. 3) regions did not allow a clear assignment of the species. As Fusarium genus was abundant in this study (27 %), and the most frequently occurring on damaged fruit, more efforts are necessary to clarify the specie identity of the Fusarium isolates. Only one isolate, AL-31 (Penicillium sp.) was identified by using a single gene marker (ITS1‐5.8S‐ITS2 region) since it was not responsive to D1/D2 LSU and COX1 (Pez F/R primers) primer amplification. Finally, Colletotrichum genus was the most abundant in this study (32.4 %) with 12 isolates. The sequences obtained from this genus range between 470 to 625 bp, 518 to 519 bp, and 457 to 620 bp for ITS1‐5.8S‐ITS2, D1/D2 LSU and COX1 loci, respectively. Phylogenetic dendrograms closely grouped these isolates (> 0.1 of distance) in a single clade. ITS1‐5.8S‐ITS2 and COX1 loci group these isolates with C. gloeosporioides accessions whereas that D1/D2 LSU dendrogram grouped these with C. acutatum and C. gloeosporioides accessions (Fig. 1, Fig. 2, Fig. 3). All newick trees are provided as supplementary material.

Fig. 1

Phylogenetic dendrogram based on ITS1‐5.8S‐ITS2 regions from the fungal community of Citrus limon var. Eureka. The dendrogram was built using a maximum likelihood method based on the Tamura-Nei model with 1000 bootstrap to infer topology.

Fig. 2

Phylogenetic dendrogram based on D1/D2 LSU regions from the fungal community of Citrus limon var. Eureka. The dendrogram was built using maximum likelihood method based on the Tamura-Nei model with 1000 bootstrap to infer topology.

Fig. 3

Phylogenetic dendrogram based on COX1 gene from the fungal community of Citrus limon var. Eureka. The dendrogram was built using a maximum likelihood method based on the Tamura-Nei model with 1000 bootstrap to infer topology.

Molecular characterization in the Colletotrichum isolates

Colletotrichum isolates of this study were grouped in all the phylogenetic dendrograms, as expected, although they come from different plant tissues, therefore, a subset of seven representative Colletotrichum isolates was selected for further analysis. Genetic variability accessed by rep-PCR and multilocus sequence analysis allowed us to support the genetic differences among the Colletotrichum isolates (Fig. 4).

Fig. 4

Genetic variability of the selected Colletotrichum isolates in the study inferred by rep-PCR analysis a) Left; UPGMA tree from absence/presence band information, right; PCR lanes organized according to UPGMA dendrogram. C. gloeosporioides (AL-05, AL-06, AL-08, AL-09, AL-11, AL-12, AL-13) from Italian lemon. C. gloeosporioides (Cg-06) and C. acutatum (Ca-93 and Ca-96) from avocado. b) Multilocus sequence analysis (D1/D2 26S, ITS1–5.8S–ITS2 and COX1) in the C. gloeosporioides isolates from Italian lemon. Concatenated ITS1–5.8S–ITS2, D1/D2 26S and COX1 sequences of the reference C. acutatum strain KC05 were used as outgroup.

Rep-PCR profile showed major differences when isolates of this study were compared with reference Colletotrichum strains isolated from avocado (Fig. 4A). The Colletotrichum isolates of this study also showed some differences in the rep-PCR profile, although these were minor, being isolate AL-05 the only one that grouped with isolate Cg-06 (Cg) of avocado. Multilocus sequence analysis, considering the concatenated ITS1‐5.8S‐ITS2, D1/D2 LSU and COX1 loci, also showed differences among the Colletotrichum isolates of Italian lemon. As with the rep-PCR analysis, the C. gloeosporioides isolates were grouped differently to C. acutatum (KC05) reference strain (Fig. 4B). Finally, an additional effort for species identification was also conducted on these Colletotrichum isolates. For this, a species-specific PCR primers analysis was carried out. CaInt2 and CgInt primers in combination with the ITS-4 primer were used to discriminate between C. acutatum and C. gloeosporioides species, respectively [17]. Fig. 5A shows the amplification with the CgInt primer in the Colletotrichum isolates selected. In general, all isolates exhibited a single 500 bp band according to the previously reported [17]. Similarly, reference strain Cg-06, isolated from Persea americana, and identified as C. gloeosporioides also exhibited this band. Reference strains Ca-93 and Ca-96, isolated also from P. americana and previously identified as C. acutatum, did not show any amplification. These same Colletotrichum isolates displayed unspecific amplifications when the CaInt2 primer was used, as did reference strain Cg-06. This primer resulted in a specific amplification (~450 bp) in the reference C. acutatum strains (Ca-93 and Ca-96), a size band expected according to the previously reported ([17], Fig. 5B) and confirmed these reference Colletotrichum strains as C. acutatum. According to these results, the Colletotrichum isolates of this study were all identified as C. gloeosporioides.

Fig. 5

Discrimination between C. gloeosporioides and C. acutatum isolates by using species-specific primers. a) 1.5% agarose gel showing the specific amplification of the C. gloeosporioides strains with the CgInt and ITS4 primers, and b) 1.5% agarose gel showing the specific amplification of the C. acutatum strains with the CaInt2 primers and ITS4.

Phenotypic characterization of the Colletotrichum isolates

Molecular characterization showed above suggested that C. gloeosporioides isolates presented some genetic differences, particularly the AL-05 isolate, although all these isolates were grouped within the different loci. Therefore, a general phenotypic characterization was conducted for these isolates, which showed similar morphological traits such as round-edged orange mycelium and conidia between 4.5 µm and 14.3 µm in size (data no shown). Additionally, and considering that many of these isolates came from fruits with anthracnosis apparent symptoms, an infection analysis on Italian lemon leaves was conducted in the same subset of isolates molecularly analyzed. For this analysis, we also included representative Fusarium and Penicillium isolates, which were also abundant on these damage fruits (Table 1). In general, all the C. gloeosporioides isolates exhibited a basal pathogenic potential, characterized by leaf necrosis (Fig. 6). Radial growth rate, infection diameter and infection percentages are presented in Table 2. All C. gloeosporioides and one Fusarium tested strains exhibited an infective potential that ranged between 4.4 and 6.3 mm for infection diameter and showed a similar radial growth rate (~250 μm/h), showing no statistically significant differences amongst the C. gloeosporioides isolates. The leaf´s infection percentage varied between the different isolates from 26 to 60 %. Reference Colletotrichum strains isolated from avocado were also used for this infection analysis. In general, Colletotrichum isolates from Italian lemon were more infective on leaves of this citrus tree, in relation to reference strains. Only the C. gloeosporioides strain (Cg-06) of avocado showed infective potential (27 %) on leaves of Italian lemon. Interestingly, both C. acutatum strains (Ca-93 and Ca-96) were unable to infect the Italian lemon leaves. This might suggest that exists a difference between the infection mechanisms between the C. acutatum and C. gloeosporioides species. Infection assay conducted with the representative P. italicum (AL-38) isolate showed null infection on Italian lemon leaves, whereas Fusarium isolates behaved very differently. Particularly, Fusarium (AL-21) isolate showed a high infection percentage (67 %) on these detached Italian lemon leaves. The above indicates that other types of fungi can contribute to the disease symptoms in these citrus trees. More effort must be conducted to clarify the specie identity and the pathogenic potential of the Fusarium isolates in combination with the Colletotrichum species on the citrus trees.

Fig. 6

Infective potential of the C. gloeosporioides isolates and other fungal abundant genera on leaves of the Italian lemon (Citrus limon var. Eureka). a) AL-09, b) AL-08, c) AL-06, d) AL-12, e) AL-13, f) AL-05, g) AL-11, h) Cg- 06, i) Ca-93, j) Ca-96, k) AL-38, l) AL-21, m) AL-20, and n) Negative control.

Table 2

Infective potential characterization for selected fungal isolates from Italian lemon (Citrus limon var. Eureka) and avocado (Persea americana) strains

Isolate	Identity	Origen	Radial growth rate (μm/h)	Infection фa (mm)	Infected leaves (%)
AL-09	C. gloesporoides	Citrus limon	258	5.4±0.7b	53
AL-08	C. gloesporoides	Citrus limon	261	6.3±0.8b	27
AL-06	C. gloesporoides	Citrus limon	232	5.3±0.6b	40
AL-12	C. gloesporoides	Citrus limon	255	5.2±0.3b	30
AL-13	C. gloesporoides	Citrus limon	241	4.9±0.2b	60
AL-05	C. gloesporoides	Citrus limon	269	4.7±0.2b	47
AL-11	C. gloesporoides	Citrus limon	248	4.4±0.2b	33
Cg-06	C. gloesporoides	Persea americana	168	4.6±0.2b	27
AL-20	Fusarium sp.	Citrus limon	251	5.3±0.4b	17
AL-21	Fusarium sp.	Citrus limon	252	4.4±0.2b	67
Ca-93	C. acutatum	Persea americana	130	0	0
Ca-96	C. acutatum	Persea americana	164	0	0
AL-38	Penicillium digitatum	Citrus limon	140	0	0
Negative control				0	0

Mean ± standard error.

No significant differences (ANOVA One-Way, P=0.005) in isolates with infective potential.

According to infection assays, two Colletotrichum isolates (AL-05 and AL-06) with average infection profiles (~40 %) and the highly infective AL-13 isolate (60 %) were selected to conducted further experiments. Infection colonization documented by SEM analysis shows the mycelial growth for all these Colletotrichum isolates at 6 days on surface wound of the lemon leaves (Fig. 7). Isolates AL-05 and AL-06 were less aggressive (sparse mycelia formation, Fig. 7a and c), as was previous inferred in the infection analysis (Table 2). Isolate AL-13 was the more aggressive on leaves, presenting abundant sporulation and formation of appressoria structures (Fig. 7b and d). Finally, isolates AL-13 and AL-05 were selected to analyze their CDA activity, to assess if this correlated with their infective profile (Fig. 8). Results shows that both isolates presented CDA activity, being isolate AL-13 the one with a higher activity peaking to 1.2 U/mg protein after 120 h of incubation, according to the previous SEM and infection profile results. A different profile for CDA activity was observed for isolate AL-05, as it had its peak of activity (0.96 U/mg protein) at 24 h of culture, corresponding to the beginning of the growth phase.

Fig. 7

SEM images of wounded leaves of Citrus limon var. Eureka with representative C. gloeosporioides isolates showing the mycelial growth at 6 days, and spores and appressoria structures. a) AL-06, b) AL-13, c) AL-05, and d) details of AL-13.

Fig. 8

Determination of the CDA of representative C. gloeosporioides; AL-13 (a) and AL-05 (b).

Discussion

Fungal endophytic community identified in the current study is similar to the observed in other citrus trees [24]. For example, Penicillium and Dothideomycetes were often isolated from Citrus reticulata cv. Siyahoo [25], and abundant Alternaria isolates were recovered from Citrus sinensis [26]. Some of these fungi have also been related to different citrus diseases. Species of the genera Dothiorella, Diplodia and Lasiodiplodia were the most widespread and abundant in commercial citrus orchards of Algeria, being some Lasiodiplodia species the most aggressive on citrus shoots highlighting its importance as canker and fruit dieback agents [27]. Similarly, some genera as Penicillium, Aspergillus, Alternaria, Lasiodiplodia, Fusarium and Colletotrichum have also been associated to skin deterioration of fruits of Citrus reticulate [28]. The last two genera (Fusarium and Colletotrichum) were the most abundant on Italian lemon in the current study. In the current study, Fusarium isolates could only be identified at the genus level, even when three loci were analyzed. In this sense, several gene markers have been proposed to resolve closely-related Fusarium species and some recent studies point to the use of TEF-1α to support the identification of Fusarium species [29]. Therefore, more investigation must be conducted to resolve the species identity of these abundant isolates in the Italian lemon. According to [30] Fusarium reads were the most abundant in the acid Mexican lime (Citrus aurantifolia) as inferred by a high throughput llumina-MiSeq analysis, which also evidenced the presence of other genera such as Alternaria, Aspergillus and Penicillium. Isolates of Colletotrichum have been widely reported as causal agents of anthracnosis in citrus trees [31]. In the current study, many Colletotrichum isolates were obtained from persistent calyxes, which remain attached to the peduncle after the young fruit drop, and from damage fruits [32]. In general, these Colletotrichum isolates were phylogenetically grouped with accessions of C. acutatum and C. gloeosporioides. Identification of Colletotrichum species is critical for developing and implementing effective control strategies [9]. The occurrence and pathogenicity of C. acutatum and C. gloeosporioides species have also been previously detected in diverse citrus trees in Mexico [6], [7], [8] and in the southeastern of the United States [11]. Therefore, a further identification by using specie-specific primers was conducted. Particularly, CgInt/ITS4 primer has been useful to distingue to the specie C. gloeosporioides from others Colletotrichum species [17,33]. This analysis, combined with the phylogenetic dendrograms, allowed the identification of Colletotrichum isolates of this study as C. gloeosporioides, which supports the previous evidence that this specie is associated with the anthracnose disease in citrus crops [1]. The Colletotrichum isolates in this study were closely related, as inferred by using three loci markers. Therefore, the genetic and pathogenic diversity of these isolates were evaluated by rep-PCR profiles and multilocus sequence analysis in a sub-set of our isolates, and comparing to other Colletotrichum strains obtained from avocado. Regarding, to rep-PCR profiles, similar results have been obtained by using different minisatellite or repeated sequences, comparing C. gloeosporioides isolates from almond and avocado, being the banding patterns more diverse for the isolates of this last plant [34]. Similar results, but using arbitrarily-primed PCR, on Colletotrichum isolates allowed a differential grouping of C. gloeosporioides as compared to C. capsica isolates in papaya [33]. Other molecular based fingerprinting techniques, such as RAPD and rDNA- RFLP have also been used on C. gloeosporioides isolates from cashew, differentiating them according to their geographical origin, but not being related with their pathogenicity [35].

Regarding multilocus sequence analysis, the Colletotrichum strain diversity has also been characterized in pomegranate fruits, where a multilocus sequence analysis allowed the separation of the isolates into two major species complexes; C. acutatum and C. gloeosporioides [11]. These authors used a different genes combination to the one used in our study, including the locus: ITS, GAPDH, CHS-1, and TUB2 [11]. In another study, the authors included seven genomic loci (ITS, GAPDH, ACT, CAL, CHS-1, HIS3 and TUB2) that allowed them to identify strains of three major Colletotrichum species complexes on orange fruits, identifying C. acutatum on citrus in Europe [10]. In the current study, a mitochondrial DNA (COX1) locus in combination with two nuclear loci (ITS1‐5.8S‐ITS2 and D1/D2 LSU) were utilized. In general terms, mitochondrial DNA (mtDNA) mutate more rapidly than nuclear DNA, due to a higher copy number, a greater exposure to reactive oxygen species, and a less efficient DNA repair system [36]. In addition, a previous study, using mitochondrial DNA on the evolution of C. gloeosporioides demonstrated that mitogenome sequence can complement prevailing nuclear markers in improving species delimitation accuracy [37]. Similarly, in the current study, the mitochondrial DNA locus (i.e. COX1) used as an individual dendrogram or in a multilocus analysis supported the analysis of the genetic delimitation and variability in the C. gloeosporioides strains.

Regarding infection potential of the isolates, all C. gloeosporioides isolates analyzed exhibited a basal pathogenicity, being the leaf´s infection percentage the trait that more varied amongst the different isolates. In general terms, Colletotrichum isolates from Italian lemon were, as expected, more infective on this plant in relation to reference strains isolated from avocado. Pathogenic profile differences among Colletotrichum isolates according to their host plant have been documented in several studies. For example, C. gloeosporioides isolates from avocado and almond were compared in artificial inoculations, showing that avocado isolates produced various lesions on avocado and almond fruits, whereas the almond isolates infected these fruits at a lower rate [34]. Other study with C. acutatum and C. gloeosporioides isolates from olive fruits, showed that these isolates also were able to infect other plants, as strawberry and lupin plants, revealing their cross-infection potential [38]. In this study, the infection assay was also conducted on some representative Fusarium isolates, which showed different infection profiles. This fungus was abundant among the isolations of this study. Fusarium abundance on citrus trees has been observed by using high throughput sequencing analysis in the roots of acid Mexican lime (Citrus aurantifolia) with 46 to 95 % of the total reads [30]. Several fungal pathogens, including Colletotrichum and Fusarium isolates were collected in infected peel section of diseased fruits in Citrus reticulate Blanco (kinnow). Based on physical fruit appearance, authors classified the pathogenic damage in this fruits, suggesting that Colletotrichum isolates attack the fruits and produce both pre-harvest and post-harvest problems, whereas other fungal genera, including Fusarium, produce damage as a consequence of physical stress by invading the citrus peel, either directly or through already present lesions [28]. More effort must be conducted to clarify the pathogenic potential of the Fusarium isolates in combination with the Colletotrichum species on Italian lemon.

Infection assays conducted in this study also shows differences among C. gloeosporioides isolates from Italian lemon or avocado with regarding to the reference C. acutatum strains. This last reference strain did not show any infection potential on the detached lemon leaves. Wound colonization of some selected C. gloeosporioides isolates, initially observed in the leaves with necrotic areas, was documented by SEM analysis. Colletotrichum species often exhibits a hemibiotrophic colonization strategy, which is characterized by a relatively short biotrophic phase followed of the development of necrotic lesions [39]. SEM analysis of this study also documented the presence of spores and appressoria structures on wounded leaves. Appressoria developed from germinating spores are involved with the cuticle penetration [40]. This results support the reliable identification of these Colletotrichum species as C. gloeosporioides, since it has been recently documented that C. acutatum did not colonize citrus leaves, with or without injury [41]. Finally, CDA activity was identified in two selected Colletotrichum isolates. CDA has been previously identified from crude extracts in C. gloeosporioides [42]. Our results indicate that the isolates analyzed exhibited their maximal CDA specific activities after 120 h of incubation. Specifically, the less aggressive AL-05 showed an activity peak at 24 h of culture, corresponding to the beginning of the growth phase, which is in agreement with the report of [42] with C. gloeosporioides in submerged culture. The more aggressive AL-13 isolate exhibited a gradual increase on CDA activity during the culture. The differences in behavior might be associated to the aggressiveness of the isolate regarding the host modulation on fungal growth, related to pH and production of antagonistic molecules [42]. Moreover, it is worth to note that the production with AL-05 and AL-13 was significantly higher than that found by [43] with C. gloeosporioides at pH 6 and 96 h of submerged culture.

Conclusion

In the present study, 37 fungal isolates were identified on damaged tissues of Italian lemon trees. These including to the genera Aspergillus, Alternaria, Nigrospora, Lasiodiplodia, Dothideomycetes, Pleurostoma, Setosphaeria, Penicillium, Fusarium and Colletotrichum. Particularity, these last three genera were abundant on the damage fruits, being the Colletotrichum isolates the more abundant with 32.4 % of isolates. In general, ITS1‐5.8S‐ITS2, D1/D2 LSU and COX1 loci were suitable to identify to specie level the most of these isolates, although genera as Fusarium and Dothideomycetes might require additional molecular analysis to support their species identity. Specifically, Colletotrichum isolates identified for further specific-PCR as C. gloeosporioides, were closely grouped in the phylogenetic dendrograms, hence their genetic variability was also assessed by rep-PCR and multilocus sequence analysis, including some reference Colletotrichum isolates. Colletotrichum isolates from Italian lemon exhibited a genetic variability and different rep-PCR profile regarding to reference strains, and isolate AL-05 was the more variable. Infection values on lemon leaves ranged between 26 to 60 %. Although, a Fusarium (AL-21) isolate also shows a high infection percentage. Colletotrichum isolates from Italian lemon were more infective on leaves of this plant, in relation to reference strains. Infection colonization in some Colletotrichum isolates with different infection profile was accessed by SEM analysis showing mycelial growth, spores and appressoria structures on the wounded leaves. Infective mechanism was preliminary analyzed in the two more contrasting (AL-05 and AL-13) isolates by CDA activity showing activity in both isolates, being the aggressive AL-13 isolate the one that displayed the highest CDA activity. This is the first report on C. gloeosporioides infecting Italian lemon leaves.

Funding

This study was supported by the Instituto Politécnico Nacional (SIP-IPN) projects 2021-0289 and 20210975.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper