Collection area and biological material

Fungal strains were isolated from rhizosphere sediment collected from the mangrove forest at the Tampamachoco lagoon, Tuxpan, Veracruz, Mexico (21° 0.9' N, 97° 47' W). This mangrove forest has been affected in some areas by high salinity concentration due to the construction of levees that blocked of fresh water flow from the Tuxpan River, resulting in areas ranging from 35 ‰ (preserved area), to 65 ‰ (semi-preserved, chronically affected zone) and up 140 % (massive mangrove death; (^{Vovides et al., 2011}). Therefore, five replicate samples were collected from the rhizosphere with a hand auger at 10-15 cm depth and at 30 m intervals among samples. We collected a complete set (15 samples per species) from Rhizophora mangle L. (red mangrove), Avicennia germinans L. (black mangrove) and Laguncularia racemosa (L.) CF Gaertn (white mangrove). Each undrained sample was kept in sterile polyethylene bags and sealing them for transport inside a cooler. Once in the laboratory, the bags were stored at 4 °C for no more than five days.

Isolation of fungal strains

Ten grams from each sample were diluted in a sterile solution of 0.1 % peptone and 0.4 % NaCl. From this initial stock, three successive dilutions were made to reach a concentration of 10^-3. Afterwards, one mL of each dilution was seeded in duplicate in Petri dishes containing marine agar (BD, Difco™ Marine Agar 2216) and malt extract (BD, Difco™ Malt Extract Broth) prepared according to the manufacturer's instructions with 0.2 g.L^-1 of chloramphenicol and incubated at 25 ± 2 °C until the development of fungal colonies was observed. The fungal colonies were then individually purified by transferring mycelium to new Petri dishes with the same medium, until pure cultures were obtained.

Fermentation and preparation of extracts

Fermentation of 500 mL from each of the pure cultures was performed using Wickerham 's medium (^{Kjer et al., 2010}) and then transferred to five 100 mL Erlenmeyer flasks. This medium was incubated in a dark room at 25 ± 2 °C, first at 150 rpm for 7 days and later for 7 days in a static mode. After the incubation stage, the biomass and culture broths were separated by vacuum filtration. Both fractions were lyophilized and the generated residues were extracted with a Chloroform-Methanol (1:1) mix. Subsequently, the extracts were concentrated to dryness using vacuum evaporation. The extracts from each fungal strain thus obtained were then tested for their anti-proliferative activity.

Cell lines and culture

The human solid tumor cell lines A-549, HBL-100, HeLa, SW1573, T-47D and WiDr, donated by Prof. G. J. Peters (VU Medical Center, Amsterdam, The Netherlands), were used in this study. The cells were maintained in 25 cm² culture flasks in RPMI 1640 supplemented with 5 % heat-inactivated fetal calf serum and 2 mM L-glutamine in an incubator at 37 °C, 5 % CO₂ and 95 % air humidity. Exponentially growing cells were trypsinized and re-suspended in an antibiotic-containing medium (100 units penicillin G and 0.1 mg of streptomycin per mL). Single cell suspensions were counted using an Orflow’s MoxiZ automated cell counter (Ketchum, ID) and dilutions were made to give the appropriate cell densities for the inoculation onto 96-well microtiter plates. Based on their doubling times, cells were inoculated in 100 µL per well at 10 000 (A-549, HBL-100, HeLa and SW1573), 15 000 (T-47D), and 20 000 (WiDr) cells per well.

Anti-proliferative activity

Dry extracts were initially dissolved in DMSO at 400 times the desired final maximum test concentration, i.e. 10 mg.mL^-1 and diluted in the culture media until they reached an assay concentration of 250 µg.mL^-1. Control cells were exposed to an equivalent concentration of DMSO but with no extract (0.25% v/v, negative control). The extract (i.e. drug) treatment began on the first day after plating. The extracts were incubated for 48 h and after that cells were precipitated with 25 μL ice-cold TCA (50 % w/v) and fixed for 60 min at 4 ºC. Then the SRB assay was performed (^{Skehan et al., 1990}). The optical density (OD) of each well was measured at 492 nm using BioTek’s PowerWave XS Absorbance Microplate Reader (Winooski, VT). The percentage growth was calculated as the OD difference between the start and end of each treatment level corrected for background OD of the control wells and compared with untreated control cells. The results were expressed as the concentration of extract causing 50 % reduction in the proliferation of cancer cells (GI₅₀). We selected the most bioactive extracts, with GI₅₀ values less than 100 µg.mL^-1.

Strain identification

The identification of the strains was carried out by the observation of reproductive fungi structures under the microscope (Carl Zeiss Microscopy GmbH, Jena, Germany) and employing taxonomic keys (^{Barnett and Hunter, 1998}; ^{Watanabe, 2002}). The mycelial cultures were observed after seven days of incubation at 25 ± 2 °C employing lacto-phenol as a staining medium.

Molecular identification

The identity of the fungal strains that showed anti-proliferative activity was confirmed by obtaining ITS-rDNA nuclear sequence data. Genomic DNA was isolated following the protocol proposed by ^{Yu et al. (2011)}, and the nuclear rDNA-ITS regions were amplified using ITS1F (5’-CTTGGTCATTTAGAGGAAGTAA-3’, TM = 55° C) and ITS4 (5’-TCCTCCGCTTATTGATATGC-3’, TM = 53° C) primers by direct PCR technique. The PCR products were purified with a Wizard kit (Pro-Mega®, Madison, WI, USA) following the manufacturer's instructions, and were then sequenced at the Institute of Biotechnology, National Autonomous University of Mexico (UNAM) using an Applied Biosystems (model 391) DNA sequencer.

The identity of the species was confirmed using the search tool Blast in the NCBI GenBank. For phylogenetic analysis, the sequences were aligned using the ClustalW algorithm of MegAlign from MEGA6 software (^{Tamura et al., 2013}). The phylogenetic tree was constructed using the Maximum Likelihood algorithm (K2+G) of the MEGA6 software. The bootstrap analysis was performed with 1000 replications.

Sampling, isolation and identification of fungi

From soil samples collected from mangrove at the Tampamachoco lagoon, 32 fungal strains were isolated, of which 15 were isolated from Rhizophora mangle, 10 strains were obtained from Avicennia germinans, and 7 strains from Laguncularia racemosa. Regarding the degree of salinity of the mangroves, 12 strains were obtained from the area with 35 ‰, 15 from the area with 65 ‰, and 5 from the area with 140 ‰.

The microscopic preparations of reproductive structures were compared with those reported by ^{Barnett and Hunter (1998)} and ^{Watanabe (2002)}. We identified 16 possible genera according to their macroscopic and microscopic characteristics: Fusarium, Alternaria, Phytophthora, Trichoderma, Penicillium, Aspergillus, Sepedonium, Mucor, Acremonium, Nectria, Monacrosporium, Geotrichum, Phomopsis, Blastomyces, Humicola and Talaromyces (Table 1).

Four extracts with anti-proliferative activity were assigned to the genus Aspergillus and two to the genus Fusarium. The morphological characteristics observed for the genus Aspergillus were fast growing cottony white mycelium, with production of black spores from the center to the edges, and microscopically we identified the characteristic aspergilar head composed of a conidiophore with termination in a vesicle covered with phialides, in which the conidia are formed. Fusarium genus was characterized by its purple mycelium, short phialides, conidia added in false heads and abundant microconidia.

Of the isolated strains, 22 % were from the genus Aspergillus, 22 % were from the genus Fusarium, 13 % from genus Penicillium, 6 % from the genus Acremonium, and the remaining genera accounted for 37 %.

Anti-proliferative activity

Of the 32 fungal isolates we obtained from the rhizosphere, 64 extracts (broth and biomass per strain) were prepared and assayed against human cancer cell lines. The chemosensitivity assay indicated that six extracts showed 50% growth inhibition (GI₅₀) to one or more cancer cell lines at concentrations <100 µg.mL^-1 (Table 2). The bioactive fungi were identified as Aspergillus niger (RmS2-2a), A. tubingensis (AgC1-2), A. pulverulentus (AgS4), and Fusarium oxysporum (RmS4-3), which were isolated from the rhizosphere of Avicennia germinans and Rhizophora mangle. The active extracts consisted of three biomass extracts and three culture broths. In broth extracts, GI₅₀ values were <50 µg.mL^-1 in all tested cell lines. In contrast, biomass extracts displayed activity below this threshold only in HeLa cells. GI₅₀ values obtained ranged from 40 to 180 µg.mL^-1 in the biomass extracts and from 6 to 41 µg.mL^-1 in the culture broth (Table 2).

Molecular identification

The isolates RmS4-3, AgS4, AgC1-2 and RmS2-2a were identified morphologically and as belonging to Fusarium and Aspergillus in the case of the last three strains accordingly. A phylogenetic analysis of the four isolates is shown in the phylogenetic tree (Figure 1). Isolate RmS4-3 had a 99% similarity with the Fusarium oxysporum Schltdl (Nectriaceae) ITS sequences deposited in the GeneBank and was located in the F. oxysporum clade (Figure 1), clearly separated from a Gibberella intermediate and G. moniliformis, which is a synonymy of Fusarium verticillioides (Sacc.) Nirenberg. Isolate RmS2-2a shared a sequence similarity of 99% with Aspergillus niger Tiegh (Aspergillaceae), corresponded to the A. niger clade (Figure 1). Isolate AgS4 presented a 99 % similarity with Aspergillus pulverulentus (McAlpine) Wehmer, Bot (Aspergillaceae), corresponded to the A. pulverulentus clade (Figure 1). This species is considered synonymous with A. niger var. pulverulentus (McAlpine) Kozak. Isolate AgC1-2 shared a 100% similarity with A. tubingensis Mosseray (Aspergillaceae), corresponded to the A. tubingensis clade (Figure 1). This last species is considered synonymous with A. niger var. tubingensis (Mosseray) Kozak. Both isolates were obtained from the rhizosphere soil supporting A. germinans in the semi-preserved and preserved zones respectively. Therefore, we conclude that AgS4 and AgC1-2 belong to complex strains of the A. niger species.

Figure 1 Maximum-likelihood phylogenetic tree based on ITS sequences of the fungi extracted from the rhizosphere sediment and reference strains. Numbers at the nodes are bootstrap support values for 1000 replicates.