Gigaspora siqueirae F.A. de Souza, Barros-Barreto, Magurno, B.T. Goto, sp. nov.
Index Fungorum number: IF 902416; Mycobank number: MB 902416; Facesoffungi number: FoF 15874; Fig. 1
Etymology – in honor of Prof. José Oswaldo Siqueira, collector and keeper of the species, member of the Brazilian Academy of Science, soil microbiologist, for his enormous contribution to mycorrhizal research, especially in Brazil.
Holotype – UFRN Herbarium Fungos 3674; isotype: UFRN Herbarium Fungos 3675, 3676.
Glomerospores (= spores) formed singly in the soil, terminally on a sporogenous cell. Spores globose, (240–)367(–411) µm diam (n=100), subglobose to oblong, 220–420 × 300–460 µm; yellow cream when young, bright yellow to dark yellow at maturity (3B4–2C8, most 2A8–2C8, Fig. 1A–C), dark yellowish brown (5D8–6F8) when older (Fig. 1B); with yellow (1A7–2A8) cytoplasm. Spore wall structure composed of three layers, spore wall layers (swl) 1–3 (Fig. 1E–F): swl1, forming the spore surface, smooth, rigid, hyaline,<1.0–1.5 µm thick; swl2 yellow to dark yellow, laminate (each sublayer 1.0–1.2 µm thick), (7–)13(–28) µm thick (n=54); swl3, a germinal layer, concolorous with and tightly adherent to swl2,<1.0–1.5 µm thick, with many round papillae/warts on the inner surface formed prior to germination (Fig. 1F). Spore wall layers 1 and 2 turn brownish (6D8–6F7) in Melzer’s reagent. Sporogenous cell (Fig. 1D) formed terminally on a coenocytic subtending hypha, becoming sparsely septate at attaining full spore development, globose to subglobose, (30–)49(–68) µm wide. Wall of sporogenous cell (9.0–)12.5(–14.0) µm thick near the spore base, 2.5–7.0 µm thick 30 µm below the spore base, composed of two layers continuous with swl1 and swl2; rarely with a peg-like projection, up to 50 µm long and 2.5–12.5 µm wide. Sporogenous cells loosely associated with spores, found only in ca. 27% of mature spores. Germination by hyphae growing from the papillae of swl3 and penetrating through swl2 and swl1, usually near the sporogenous cell. Auxiliary cells borne in clusters of ca. 10, with spiny projections, sometimes attached to the host plant root surface (Fig. 1I). Mycorrhiza with arbuscules, straight, and coiled hyphae. Intraradical mycelium without vesicles. Arbuscules with a trunk, 5–7 µm wide, and numerous branches narrowing abruptly towards their tips. In carrot root-organ-culture, straight hyphae 5–11 µm wide, coiled hyphae 5–7 µm wide (Fig. 1H–I).
Material examined – Brazil. Boa Esperança municipality, in cultures (multi-spore) on Urochloa decumbens (Stapf) R.D. Webster (J.O. Siqueira, 27 May 2003); deposited in DCS-UFLA (holotype). ISOTYPE deposited at the International Culture Collection of Glomeromycota (CICG, Blumenau, Santa Catarina, Brazil, http://www.furb.br/cicg/index.php?lang=EN) under the accession MGR253. The fungus was originally cultured from a soil sample collected in March 1986 from a coffee (Coffea arabica L.) plantation in Cerrado Bioma, the Brazilian Savanna, in the south of Minas Gerais State in SE Brazil. The locality is around 900 m above sea level and consists of cerrado fragments and coffee crops. The soils are highly weathered, clayey, acidic, low-fertility Dusky Red Latossol (Oxisol), with a pH (in water) of around 5.7, available P (Mehlich) of 2 mg‧dm−3, and Ca+ Mg of 431.0 mg‧dm−3. Mean precipitation and temperature in this area is 1,400 mm and approx. 19 °C, respectively. The trap culture in which the species was first established was set up on 27 Aug. 1987 by sowing seeds of U. decubens grass in a substrate composed of fumigated soil and sand (1:1 v/v) that received spore suspension. The trap cultures were fertilized with nutrient solution lacking P. Subsequently, multi-spore, single species cultures were established using purified spore suspensions poured onto the roots of mycorrhiza-free U. decumbens seedlings at transplanting. The species has not been established in singlespore cultures, but it has been repeatedly subcultured since its original culture in 1987. Four different multi-spore lines have been maintained, with the reference numbers UFLA 175, 178, 179, and 872.
GenBank accession numbers – OQ302571, OQ302575–OQ302576, OQ302579–OQ302583, OQ302587, OQ302590–OQ302591 and OQ680681–OQ680684. The primer combination used for amplification and cloning of the near clomplete SSU and complete ITS region were NS1 5′-GTA GTCATATGCTTGTCTC-3′ and ITS4 5′-TCCTCCGCTTATTGATATGC-3′ (de Souza et al. 2004). For the LSU we used the primer combination LR1 5′-GCATATCAATAAGCGGAGGA-3′- van Tuinen et al, (1998) and FLR4 (5′-TACGTCAACATCCTTAACGAA-3′- Gollotte et al., (2004). The DNA extraction, amplification, cloning and sequencing was carried out by de Souza et al. (2004) for amplicons covering the SSU and ITS regions. Similarly, the LSU amplicons were obtained by the authors.
Notes – Gigaspora siqueirae is mainly distinguished by its unique phylogeny, which accommodated its sequences in a fully supported clade, inside a wider polyphyletic clade hosting all sequenced Gigaspora, Intraornatospora and Paradentiscutata species (Fig. 2). Despite its phylogenetic placement, we decide to retain the new species as Gigaspora, based on the morphological features of spores, convincingly resembling those of the genus. Potential emendations inside of the polyphyletic clade will be discussed once additional molecular data will be available for the Intraornatospora and Paradenticutata genera. Morphologically, G. siqueirae is indistinguishable from G. gigantea. Both species produce spores of similar size, the thickness of their spore wall overlaps, and, most importantly, the spore color—“bright yellow”—of these two species come from the pigmentation of the cytoplasm rather than the spore wall. That characteristic was diagnostic to differentiate G. gigantea from all previously described Gigaspora species. Interestingly, the origin of distinctly coloured spores with a colourless or at most lightly coloured spore wall from the pigmentation of the spore cytoplasm has so far been identified only in one other species of the Glomeromycota, i.e., Desertispora omaniana (Symanczik, Błaszk. & Al-Yahya’ei) Symanczik, Błaszk., Kozłowska & Al-Yahya’ei (Symanczik et al. 2018), originally described as Diversispora omaniana Symanczik, Błaszk. & Al-Yahya’ei (Symanczik et al. 2014). Phylogenetically G. gigantea behaves differently, forming a sister relationship with G. albida and G. candida as in the tree depicted in Fig. 2. The relationship of G. gigantea with G. albida and G. candida and the distinctiveness of G. siqueirae was also evident from a PCR-Denaturing Gradient Gel Electrophoresis (DGGE) profiling of inter- and intraspecies sequence heterogeneity of the V9 region of the 18S rRNA gene of 51 of Gigaspora strains, which included G. albida (6), G. candida (1), G. decipiens (2), G. gigantea (9), G. margarita (11), G.ramisporophora (1),G. rosea (19) and G. siqueirae (1), represented by the strain UFLA872 and one Gigaspora sp. (TW1-1) (de Souza et al. 2004).
Figure 1 – Gigaspora siqueirae (UFRN-Fungos 3674, holotype). a Glomerospores with bright yellow colour typical of the new species obtained on a plant host culture. b Mature dark yellowish brown glomerospores. c Glomerospores with bright yellow colour produced root-organ-culture. d Bulbous sporogenous cell with two wall layers attached to spore. e–f Spore wall with three layers (swl1–3). f Germ hypha (gh) developed from a wart formed on the lower surface of swl 3 and emerged from swl 2 and 1. g–i Glomerospore, intraradical hypha and auxiliary cells with spiny ornamentation produced by pigmented extraradical hyphae colonizing carrot rootorgan-culture. a‒b. Spores in water. c‒f. Spores in PVLG. g‒i. Mycorrhizal structures obtained from in vitro cultures. Scale bars: a, c, g=200 μm, b,h=100 μm, d=20 μm, e–f=5 μm, i=50 μm
Figure 2 – Phylogram generated from Maximum Likelihood (ML) and Bayesian Inference (BI) analyses displaying the phylogenetic relationships between the Gigaspora siqueirae (in bold, blue) and members of the Gigasporaceae and Intraornatosporaceae. The family Dentiscutataceae, as sister to the clade Gigasporaceae–Intraornatosporaceae, was used to root the consensus tree. The trees were inferred using a dataset that includes concatenated sequences divided into five partitions (18S: 1–1786, ITS1: 1787–1883, 5.8S: 1884-2042, ITS2: 2043–2251, 28S: 2252–3047). For species in the Intraornatosporaceae only partial 28S sequences were available. Overall, the dataset included 15 species in six genera represented by 55 sequences. Sequence alignment was performed using MAFFT 7.243 (Katoh et al. 2019), strategy E-INS-i. In both ML and BI analyses, GTR+I+G was chosen as a nucleotide substitution model for each nucleotide partition (Abadi et al. 2019). The ML tree was estimated using RAxML-NG 1.0.1 (Kozlov et al. 2019), with a maximum likelihood/1000 bootstrapping run, and ML estimated proportion of invariable sites and base frequencies. In the BI analysis, four Markov chains were run over ten million generations in MrBayes 3.2 (Ronquist et al. 2012), sampling every 500 generations, with a burn-in at 30% of sampled trees. All parameters of the convergence diagnostic (Potential Scale Reduction Factor) indicated that convergence was obtained (Ronquist et al. 2012; Miller et al. 2015). The tree topology obtained from the ML analysis was identical to that generated in the BI analysis. Support values and posterior probabilities greater than 60% and 0.95, respectively, are indicated above or below the nodes. The bar indicates 0.005 expected change per site per branch