Comparison of commonly used primer sets for

Soil Biology & Biochemistry 68 (2014) 482e493

Contents lists available at ScienceDirect

Soil Biology & Biochemistry journal homepage: www.elsevier.com/locate/soilbio

Comparison of commonly used primer sets for evaluating arbuscular mycorrhizal fungal communities: Is there a universal solution? c, d Petr Kohout a, b, *, Radka Sudová a, Martina Janousková a, Martina Ctvrtlíková , a, e , Martin Hejda a, Hana Pánková a, Renata Slavíková a, Katerina Stajerová Miroslav Vosátka a, Zuzana Sýkorová a Institute of Botany, Academy of Sciences of the Czech Republic, CZ-252 43 Pr uhonice, Czech Republic Institute of Ecology and Earth Sciences, University of Tartu, EE-50411 Tartu, Estonia , Czech Republic Institute of Botany, Academy of Sciences of the Czech Republic, CZ-379 82 Trebon d Biology Centre, Institute of Hydrobiology, Academy of Sciences of the Czech Republic, CZ-370 05 Ceské Bud ejovice, Czech Republic e Department of Ecology, Faculty of Science, Charles University, CZ-128 01 Prague 2, Czech Republic a

b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 10 October 2012 Received in revised form 23 August 2013 Accepted 31 August 2013 Available online 18 September 2013

Different primer systems have been developed to characterize arbuscular mycorrhizal fungal (AMF) communities; however, a direct comparison of their specificity, potential to describe diversity and representation of different phylogenetic lineages is lacking. Using seven root samples, we compared four routinely used AMF-specific primer systems for nuclear ribosomal DNA covering i) the partial small subunit (SSU), ii) the partial large subunit (LSU), iii) the partial SSU and internal transcribed spacer (ITS; “Redecker”) and iv) the partial SSUeITSepartial LSU region (“Krüger”). In addition, a new primer combination v) covering the ITS2 region (ITS2) was included in the comparison. The “Krüger” primers tended to yield the highest AMF diversity and showed a significantly higher Shannon diversity index than the SSU primers. We found a strong bias towards the Glomeraceae in the LSU and SSU primer systems and differences in the composition of AMF communities based on the “Redecker” primer system. Our results confirm the crucial role of the choice of target rRNA marker region for analysing AMF communities. We also provide evidence that nested-PCR based data can be interpreted semi-quantitatively and that the extent of observed AMF community overdominance largely depends on the choice of primer. Ó 2013 Published by Elsevier Ltd.

Keywords: Glomeromycota Primers rRNA Arbuscular mycorrhizal fungi Diversity 454-Sequencing

1. Introduction Fungi from the phylum Glomeromycota (Schüssler et al., 2001) are an important component of ecosystems because they form arbuscular mycorrhiza, the most widespread type of symbiosis in the plant kingdom. In natural conditions, plants are colonized by communities of arbuscular mycorrhizal fungi (AMF), whose diversity and identity influence the structure and functioning of plant communities (van der Heijden et al., 1998). The development of molecular methods for identifying AMF directly within plant roots has boosted the research of AMF communities (Simon et al., 1992; Helgason et al., 1998; Öpik et al., 2009). In contrast to diversity studies based on morphological determinations of soil-born spores (Johnson et al., 1991), DNA-based

* Corresponding author. Institute of Botany, Academy of Sciences of the Czech Republic, CZ-252 43 Pr uhonice, Czech Republic. Tel.: þ420 728228263. E-mail address: [email protected] (P. Kohout). 0038-0717/$ e see front matter Ó 2013 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.soilbio.2013.08.027

approaches enable the identification of AMF taxa, which directly interact with plants (Clapp et al., 1995; Hempel et al., 2007). Different molecular markers have been described to identify AMF species or to investigate the AMF phylogeny. Most studies routinely use nuclear ribosomal RNA (rRNA) gene sequence markers. Three rRNA regions, individually or in combination, are used as molecular markers: the partial small subunit (SSU) rRNA gene, the internal transcribed spacers (ITS1, 5.8S and ITS2) and the partial large subunit (LSU) rRNA gene. The choice of rRNA region is crucial because rRNA regions differ in their ability to distinguish closely related AMF species (species resolution power) and in the extent to which well-determined sequences are represented in public sequence databases (Stockinger et al., 2010; Schoch et al., 2012). Moreover, the primer systems used for their amplification often discriminate certain AMF lineages or co-amplify DNA from non-target organisms (Stockinger et al., 2010), reflecting the difficulty to develop primers exclusively for AMF. The SSU rRNA gene is the most frequently used molecular marker (e.g. Helgason et al., 1999; Lee et al., 2008; Öpik et al., 2008).

P. Kohout et al. / Soil Biology & Biochemistry 68 (2014) 482e493

The relatively low variability of AMF SSU sequences, compared to the ITS region, makes it possible to align the whole Glomeromycota into a single dataset, which facilitates phylogenetic analyses. Compared to the other primer systems in use, this region also provides semi-quantitative information about AMF communities because it can be amplified in a single-step PCR. However, previously designed primer systems for the SSU rRNA gene may exclude basal lineages of the Glomeromycota (Lumini et al., 2010), or coamplify plant DNA (Alguacil et al., 2011) or non-target fungal groups (Liu et al., 2011). The ITS rRNA region offers large sequence variability within the Glomeromycota and consequently a high discriminative power down to the species level (Stockinger et al., 2010). Currently, most Glomeromycota diversity studies which use the ITS marker employ a system of family-specific primers developed by Redecker (2000) and Redecker et al. (2003). Although it is more labour intensive than systems using a single AMF-universal primer pair, its higher sensitivity for less abundant AMF lineages, which may remain undiscovered by a single primer system on sites dominated by Glomus species (e.g. Hijri et al., 2006), presents a clear advantage. The LSU resolution power at the species level is based on the D2 variable region and is comparable to that of ITS (Stockinger et al., 2010). However, results obtained by the most commonly used FLR4/FLR3 primers (Gollotte et al., 2004) might suffer from a considerable bias towards the Glomeraceae (Gamper et al., 2009). Despite this disadvantage, they are frequently used for analyses of AMF communities (e.g. Bainard et al., 2011; Meadow and Zabinski, 2012), especially by terminal restriction fragment length polymorphism (tRFLP) according to the system proposed by Mummey and Rillig (2007). To overcome the above-mentioned problems and improve molecular species characterization of the Glomeromycota, Krüger et al. (2009) designed a mixed primer set, which amplifies an AMF rRNA fragment of approximately 1500 bp covering the partial SSU, the whole ITS and the partial LSU including the variable D1 and D2 regions. This primer combination, however, has so far been tested only in a very limited number of field studies (Wang et al., 2011; Fahey et al., 2012). Because of the length of the amplified DNA fragment, it is not suitable for next generation sequencing. For 454 sequencing-based studies, Stockinger et al. (2010) therefore recommended a combination of their primer system with nested-PCR amplification of a short variable fragment such as ITS2-partial LSU regions. Still, no suitable AMF-specific primer combination exists for its amplification. Whereas most AMF diversity data are based on cloning and Sanger sequencing of clones, several recent studies have adopted the new 454-sequencing approach (e.g. Öpik et al., 2009; Lumini et al., 2010). This high-throughput technology overcomes the cloning step and provides orders of magnitude more data while saving time, labour and financial costs. The choice of primers and

483

PCR conditions, however, remain crucial for obtaining an unbiased picture of the fungal community (Lumini et al., 2010; Tedersoo et al., 2010). Though previous studies point at the limitations of certain specific primer systems in terms of exclusion or discrimination of certain lineages within the Glomeromycota (e.g. Daniell et al., 2001; Gamper et al., 2009), a direct comparison, which would systematically assess their relevance in diversity studies is missing. The aim of our study was therefore to compare AMF communities described by the most commonly used AMF primer systems in a set of fieldcollected root samples of arbuscular mycorrhizal plants. The specific goals of our comparison were the following: i) to evaluate specificity of the selected primer systems to AMF, ii) to determine differences in the detected spectra of AMF taxa, iii) to compare semi-quantitative information about the relative abundance of AMF molecular operational taxonomic units (MOTUs) and iv) to critically assess the suitability of different primer systems for AMF diversity studies. Our comparison was mainly based on the primer system developed by Krüger et al. (2009) because the length of the amplified fragment enables a direct comparison with previously used SSU-ITS-based and LSU-based primer systems (Redecker, 2000; Gollotte et al., 2004). Additionally, we amplified the ITS2 region using a newly proposed primer combination. Unfortunately, the fragment amplified by the primers of Krüger et al. (2009) does not cover the SSU region used in most diversity studies (Helgason et al., 1998, 1999; Lee et al., 2008), thus precluding a direct comparison of the detected taxa between these two systems. We nevertheless included the SSU region in our study to compare the obtained diversity with results attained for other markers, focussing on the representation of the main AMF lineages and relative abundance of AMF taxa. 2. Materials and methods 2.1. Sampling and study sites Six plant species from different locations and biotopes were chosen for this comparative study based on our preliminary results, which indicated that these plants differ in AMF taxon richness and community composition (see Table 1). A single adult plant per each species/biotope combination was sampled and transported to the laboratory, where the root system was washed, cut into pieces, frozen in aliquots of 50e100 mg in Eppendorf tubes and stored at 80 C until use. 2.2. DNA extraction and polymerase chain reactions (PCR) Root samples were ground in liquid nitrogen using a mortar and pestle. DNA was extracted using the DNeasy Plant Mini Kit (Qiagen,

Table 1 List of samples included into the study. Sample code

Host plant

Locality

Biotope

Coordinates

Altitude (a.s.l.)

Sampling time

LIT LOB TAN

Littorella uniflora Lobelia dortmanna Tanacetum vulgare

50 m 50 m 237 m

Leucanthemum ircutianum

Freshwater oligotrophic lake Freshwater oligotrophic lake Species-poor field abandoned for 10 years Species-rich mountain meadow

58 22ʹ53ʺN, 6 06ʹ46ʺE 58 22ʹ53ʺN, 6 06ʹ46ʺE 50 13ʹ18ʺN, 15 52ʹ49ʺE

LEU

Mjåvatn, Rogaland S, Norway Mjåvatn, Rogaland S, Norway Hradec Králové, E Bohemia, Czech Republic Haratice, N Bohemia, Czech Republic

50 41ʹ18ʺN, 15 19ʹ5900 E

490 m

Briza media Brachypodium pinnatum Brachypodium pinnatum

Malesov, N Bohemia, Czech Republic Malesov, N Bohemia, Czech Republic Malesov, N Bohemia, Czech Republic

August 2010 August 2010 November 2010 November 2010 July 2010 July 2010 July 2010

BMG BPG BPF

Seminatural dry grassland Seminatural dry grassland Adjacent field abandoned for about 20 years

0

00

0

00

50 30 02 N, 14 18 55 E 50 300 0200 N, 14 180 5500 E 50 300 0500 N, 14 180 5600 E

242 m 242 m 242 m

484


Fig. 1. Target nuclear ribosomal DNA regions and primers locations. Lines above the rDNA represent products of the first PCRs. Lines below the rDNA represent products of the second PCRs. All of the regions studied are shown on the upper (A) figure except for those amplified by the “Redecker” primers, which are shown on the below figure (B).

Germany) according to the manufacturer’s instructions. DNA was eluted in 50 ml of autoclaved ddH2O. The DNA extracts were diluted 1:10 in autoclaved ddH2O, and 1 ml of each diluted extract was used as a template in each of the five PCR reactions specified below (each reaction was performed in triplicate). i) “SSU”: A fragment of 510e570 bp covering a variable region of the SSU (Fig. 1A) was amplified in a single step using the universal eukaryotic primer NS31 (Simon et al., 1992) in combination with the AMF-specific primer AML2 (Lee et al., 2008). This primer pair was selected based on the work of Liu et al. (2011) and pers. comm. with M. Öpik. ii) “Redecker”: The ITS region and its flanking regions in the SSU and LSU (Fig. 1B) were amplified using the universal eukaryote primers NS5 and ITS4 (White et al., 1990). Seven separate PCR reactions, each specific for a subgroup of the Glomeromycota, were performed using the primer pairs GLOM1310/ITS4i (specific for the Glomeraceae), LETC1677/ ITS4i (specific for the Claroideoglomeraceae), ACAU1661/ ITS4i (specific for the Acaulosporaceae), ARCH1311AB/ITS4i (specific for the Archaeosporales), GIGA1313/GIGA5.8R (specific for the Gigasporaceae), PARA1313/ITS4i (specific for

the Paraglomeraceae) and GLOC1355/ITS4i (specific for the Diversisporaceae) (Redecker, 2000; Redecker et al., 2003, 2007). The length of the amplicons ranged from ca 700 to 1000 bp depending on the primer combination. iii) “LSU”: DNA fragment of 448e505 bp covering the variable D2 region of the LSU (Fig. 1A) was amplified by nested PCR with the general fungal primers LR1/FLR2 (van Tuinen et al., 1998; Trouvelot et al., 1999) in the first amplification step and the AMF-specific primer FLR4 (Gollotte et al., 2004) in combination with the eukaryote primer 250f (Sýkorová et al., 2012) in the second step. The forward primer 250f was preferred to the commonly used primer FLR3 (Gollotte et al., 2004), which is not strictly specific to AMF and also exhibits mismatches with certain AMF groups such as the genera Scutellospora or Paraglomus (Stockinger et al., 2010). iv) “Krüger”: DNA fragment of 1485e1610 bp covering the partial SSU, the whole ITS, and the variable D1 and D2 regions of the LSU (Fig. 1A) was amplified by nested PCR using the primers developed by Krüger et al. (2009). v) “ITS2”: The ITS2 fragment of 290e328 bp (Fig. 1A) was amplified by nested PCR using the AMF-specific primer mixtures SSUmAf and LSUmAr (Krüger et al., 2009) in the

Table 2 Description of PCR conditions for all primer systems. Primer system

SSU

“Redecker”

LSU

“Krüger”

PCR conditionsa 1st PCR

2nd PCRb

0.5 U Taq, 0.2 mM primers, 16 mg BSA 3 m 94 C / 35 (30 s 94 C / 30 s 52 C / 45 s 72 C) / 30 m 72 Cc 0.5 U Taq, 0.2 mM primers, 16 mg BSA 3 m 94 C / 30 (45 s 94 C / 50 s 51 C / 90 s 72 C) / 10 m 72 C 0.5 U Taq, 0.2 mM primers, 16 mg BSA 4 m 94 C / 35 (30 s 94 C / 30 s 58 C / 90 s 72 C) / 10 m 72 C 1 U Taq, 0.5 mM primers, 16 mg BSA

X

5 m 95 C / 38 (30 s 95 C / 90 s 60 C / 2 m 72 C) ITS2 / 10 m 72 C a b c

0.5 U Taq, 0.2 mM primers, 16 mg BSA 3 m 94 C / 35 (45 s 94 C / 50 s / 30 m 72 C 0.5 U Taq, 0.2 mM primers, 16 mg BSA 4 m 94 C / 35 (30 s 94 C / 30 s / 30 m 72 C 1 U Taq, 0.5 mM primers, 16 mg BSA 5 m 95 C / 35 (30 s 95 C / 90 s / 30 m 72 C 1 U Taq, 0.2 mM primers 4 m 94 C / 35 (30 s 94 C / 30 s / 10 m 72 C

61 C / 90 s 72 C)

59 C / 90 s 72 C)

63 C / 2 m 72 C)

49 C / 90 s 72 C)

Concentrations of other PCR reagents, not mentioned in the table, were the same for all reactions: 2 mM of MgCl2, 1 Taq buffer with KCl and 0.2 mM of each dNTP. The PCR products of the 1st PCR were diluted 1:100 in ddH2O and 1 ml of the dilution was used as template in the 2nd PCR. Cycling parameters (m e minutes, s e seconds).


first PCR followed by the nested amplification of ITS2. Here, the general eukaryote primer ITS4 (White et al., 1990) was used in combination with a newly designed forward fungal specific primer ITS7o (50 -GTG AAT CAT CRA ATY TTT G-30 , modified from Ihrmark et al. (2012)). Specifications of PCR reagents concentrations and cycling parameters for all primer systems are summarized in Table 2. 2.3. Cloning, sequencing and sequence analyses PCR products were checked on 1% agarose gels, positive reactions were pooled (always the three replicates per DNA extract and primer combination), gel-purified with the Zymoclean Gel DNA Recovery Kit (Zymo Research, USA) and cloned using the TOPO-TA Cloning Kit (Invitrogen, USA) as described in the manufacturer’s instructions. Inserts were re-amplified using the vector primers M13fwd and M13rev. Positive PCR products (25-55 clones from each the LSU, SSU, ITS2 and “Krüger” PCRs and 10-15 clones from each of the “Redecker” PCRs) were purified and sequenced in both directions (Macrogen Inc., South Korea). Sequences were edited using the software FinchTv (Geospiza, Inc., USA), and the forward and reverse strands were assembled into a consensus sequence in ChromasPro (Technelysium Pty Ltd, Australia). The glomeromycotan origin of the sequences was tested by BLAST (Altschul et al., 1997). Sequences were aligned to previously published sequences in MAFFT version 6 (http://mafft.cbrc.jp/ alignment/server) using the slow, iterative refinement method (gap opening penalty 1.0, offset value 0.1). The following alignments were prepared: i) SSU, ii) LSU and “Krüger” sequences shortened to the length of the FLR4/250f fragment (referred to as the LSU fragment), iii) “Redecker” and “Krüger” sequences shortened to the ITS1-5.8S-ITS2 regions (referred to as the ITS fragment; 6 alignments were prepared for each of the families/orders of the Glomeraceae, Acaulosporaceae, Archaeosporales, Claroideoglomeraceae, Diversisporaceae and Paraglomeraceae), iv) ITS2, “Redecker” and “Krüger” sequences shortened to the length of the ITS7o/ITS4 fragment (referred to as the ITS2 fragment), v) the whole “Krüger” fragment. Poorly aligned positions and divergent regions were removed from all alignments except for the ITS2 alignment using the software Gblocks Version 0.91b (Castresana, 2000) under less stringent conditions. Phylogenetic trees were obtained by distance analysis using the neighbour joining algorithm in MEGA5 (Tamura et al., 2011) with 1000 bootstrap replicates, the Kimura two-parameter model and the gamma shape parameter equalling 0.5. Sequences were checked manually for possible chimeras, i.e., artificial sequences generated unintentionally during the PCR that typically feature sequence data from two distinct taxa. The conserved regions in the 5.8S gene and at the beginning of LSU are very likely places for a chimeric breakpoint (Nilsson et al., 2010). Sequences which either turned up as singletons in the phylogenetic trees or changed their position in the inferred phylogenies were considered to be potentially chimeric (Jumpponen, 2007). Such sequences were checked manually in the alignments and by blasting each part of the sequence. We excluded chimeric sequences, prepared new alignments and generated a phylogenetic tree in the same way as described above. 2.4. Definition of molecular operational taxonomic units (MOTUs) MOTUs were defined in a conservative manner as consistently separated and well-supported monophyletic groups in the phylogenetic trees and confirmed by their characteristic sequence signatures in the alignments. We avoided splitting the lineages unless

485

there was positive evidence for doing so. Basically, MOTUs were delimited based on the global (“Krüger”) alignment and tree. As the “Krüger” sequences were used as backbone for each of the trimmed datasets, the MOTUs delimited in the “Krüger”, “Redecker”, ITS2 and LSU data sets (LSU þ ITS region) correspond to each other and were designated after the major clade they belonged to, followed by a numerical index (x in the following examples) identifying each MOTU: GLOM-x (Glomeraceae), CLAR-x (Claroideoglomeraceae), DIV-x (Diversisporaceae), ACAU-x (Acaulosporaceae), PARA-x (Paraglomeraceae) and ARCH-x (Archaeosporales). The applied nomenclature of the Glomeromycota follows http://www.amfphylogeny.com and Krüger et al. (2012). To compare the species resolution power between the “Krüger”, LSU and ITS2 fragments, we also delimited the AMF taxa according to each corresponding trimmed alignment and phylogenetic tree. Because there was no overlap between the target SSU region and the “Krüger” fragment (Fig. 1A), MOTUs were defined independently in the SSU data set. Two approaches were employed: i) based on 97% full-length sequence similarity within MOTUs using TOPALi 2.5 (Milne et al., 2009), which represents the most common approach to MOTU delimitation in the SSU (Öpik et al., 2010); and ii) the phylogenetic approach described above for the LSU þ ITS region. Representative sequences from each plant sample/primer combination/MOTU were deposited in the EMBL database under the accession numbers HE775276eHE775437, HE804054e HE804093 and HE674764eHE674814.

2.5. Statistical analyses Rarefaction curves were constructed for each combination of sample and primer system using the analytical formulas in the programme EstimateS 8.2 (Colwell, 2009) in order to compare the primer systems in terms of the observed MOTU richness (Mao Tau) and the number of sequences necessary for a reasonable coverage of AMF diversity. Shannon diversity index and estimated MOTU richness (Chao 2) for each sample were also calculated in EstimateS 8.2. The quantitative aspects of MOTU detection were explored based on frequencies of the MOTUs in each clone library (sample primer system), calculated from the numbers of clones belonging to a MOTU and the total number of clones with AMF sequences. The “Redecker” primer system was not included into these comparisons because its clone libraries were constructed from multiple PCRs and clonings. The frequencies of MOTUs obtained by the “Krüger”, LSU and ITS2 primers were pairwise correlated across all samples by non-parametric correlation using SPSS 15.0 software (SPSS, an IBM Company, USA). We used only those MOTUs which were detected by both compared primer systems. We tested the effect of the primer system on the number of MOTUs using a one-way ANOVA, with the factor primer having 5 levels: SSU, LSU, “Krüger”, “Redecker” and ITS2. The effect of the primer system on Chao 2, Shannon diversity index and relative abundance of the most abundant MOTU was tested in the same way, but the main factor took on only 4 levels (the “Redecker” primer system was omitted from these comparisons because its clone libraries were constructed from multiple PCRs and clonings). The data on the number of MOTUs and Chao 2 were logtransformed prior to the analysis in order to meet the assumption of normally distributed error structure. Post hoc comparisons of the means were made using a Tukey HSD test. Statistical significance was considered at the conventional 5% level (P 0.05). All calculations were performed in R (R Development Core Team., 2011).

486


HE775296 T. vulgare HE775308 B. pinnatum field FR750200 Rhizophagus irregularis 90 HE775298 L. ircutianum HE775322 B. pinnatum meadow GLOM-1 (R. irregularis) 50 HE775299 L. ircutianum FM992381 Rhizophagus irregularis 62 97 HE775335 B. media 73 99 HE775313 B. pinnatum field FM865588 Rhizophagus irregularis FM865608 Rhizophagus irregularis 100 HE775307 B. pinnatum field GLOM-2 (uncultured Rhizophagus) HE775316 B. pinnatum field FR750073 Rhizophagus fasciculatus FN547500 Rhizophagus proliferus 67 FM865538 Rhizophagus clarus FM865559 Rhizophagus intraradices 99 Fi1cons#33 Sclerocystis sinuosa MD126 GLOM-3 (uncultured Rhizophagus) HE775323 B. pinnatum meadow 68 80 100 HE775282 L. dortmanna GLOM-4 (Rhizophagus sp., new MOTU) HE775280 L. dortmanna HE775340 B. media GLOM-6 (Rhizophagus sp., new MOTU) FR750093 Rhizophagus 'cerebriformis' 53 HE775294 T. vulgare 100 GLOM-5 (Rhizophagus sp., new MOTU) 100 HE775295 T. vulgare 64 HE775301 L. ircutianum 81 HE775338 B. media 95 100 HE775302 L. ircutianum GLOM-9 (uncultured Glomeraceae) HE775330 B. pinnatum meadow 87 95 HE775327 B. pinnatum meadow HE775312 B. pinnatum field 99 GLOM-8 (uncultured Glomeraceae) 99 HE775320 B. pinnatum meadow HE775328 B. pinnatum meadow GLOM-7 (Glomeraceae sp., new MOTU) HE775332 B. media 100 100 HE775293 T. vulgare 100 HE775292 T. vulgare GLOM-10 (uncultured Glomeraceae) HE775309 B. pinnatum field 100 HE775331 B. pinnatum meadow GLOM-11 (uncultured Glomus) 82 FR750203 Glomus species W3347 FR750526 Glomus macrocarpum 50 72 HE775281 L. dortmanna 98 HE775287 L. uniflora GLOM-19 (uncultured Glomeraceae) 100 HE775305 L. ircutianum HE775290 L. uniflora 99 GLOM-18 (Glomeraceae sp., new MOTU) 73 99 HE775277 L. dortmanna 99 HE775285 L. uniflora 100 HE775317 B. pinnatum field GLOM-12 (Glomeraceae sp., new MOTU) HE775334 B. media HE775326 B. pinnatum meadow 80 100HE775339 B. media 100 GLOM-17 (uncultured Glomeraceae) HE775333 B. media 98 HE775289 L. uniflora GLOM-15 (Glomeraceae sp., new MOTU) HE775288 L. uniflora 99 98 HE775291 L. uniflora GLOM-16 (uncultured Glomeraceae) 100 HE775284 L. uniflora 76 HE775304 L. ircutianum GLOM-14 (Glomeraceae sp., new MOTU) 51 HE775303 L. ircutianum 57 HE775286 L. uniflora GLOM-13 (Glomeraceae sp., new MOTU) HE775283 L. uniflora 78 57 AY236320 Funneliformis sp. SP101 74 HE775329 B. pinnatum meadow GLOM-20 (Funneliformis xanthium) AJ849467 Funneliformis xanthium FN547477 Funneliformis sp. WUM3 77 56 FN547494 Funneliformis caledonium 100 FM876798 Funneliformis coronatum FN547474 Funneliformis mosseae GLOM-21 (Funneliformis mosseae) 99 HE775306 B. pinnatum field 85 HE775311 B. pinnatum field 100 FN547571 Gigaspora rosea FM876800 Gigaspora sp. W2992 100 FM876834 Scutellospora nodosa 64 FM876839 Scutellospora heterogama FM876832 Pacispora scintillans 100 FM876790 Acaulospora scrobiculata 94 FM876792 Acaulospora sp. WUM18 67 64 FM876822 Kuklospora kentinensis HE775276 L. dortmanna 100 ACAU-1 (Acaulospora sp.) 100 HE775278 L. dortmanna FM876782 Acaulospora laevis FR686948 Diversispora sp. W5257 61 FN547655 Diversispora aurantia 100 AM713413 Diversispora eburnea DIV-3 (Diversisporaceae sp., new MOTU) HE775319 B. pinnatum field 66 HE775321 B. pinnatum meadow DIV-2 (Diversisporaceae sp., new MOTU) 51 80 HE775336 B. media HE775337 B. media 69 FM876814 Diversispora epigaea DIV-1 (Diversispora epigaea) HE775318 B. pinnatum field HE775297 T. vulgare FM876806 Claroideoglomus sp. W3349 56 FR750061Claroideoglomus claroideum 100 74 FM876810 Claroideoglomus luteum CLAR-3 (Claroideoglomus sp., new MOTU) HE775325 B. pinnatum meadow FN547623 Claroideoglomus etunicatum 87 63 88 HE775310 B. pinnatum field CLAR-2 (uncultured Claroideoglomus) HE775315 B. pinnatum field 100 HE775324 B. pinnatum meadow HE775300 L. ircutianum CLAR-1 (Claroideoglomus sp., new MOTU) 85 70 HE775314 B. pinnatum field ARCH-1 (uncultured Archaeosporales) HE775279 L. dortmanna FR750023 Archaeospora schenckii 100 FN547535 Ambispora fennica 58 FN547524 Ambispora appendicula 100 AM295494 Paraglomus laccatum FR750052 Paraglomus brasilianum Fi1cons#39 Paraglomus occultum IA702, AFTOL-ID 844 100 0.05

91

Glomeraceae

Gigasporaceae

Acaulosporaceae

Diversisporaceae

Claroideoglomeraceae

Archaeosporales Paraglomerales


487

Table 3 Overview of the analysed clones obtained by the different primer systems (AMF sequences/non-target sequences/chimeric sequences). In the family-specific primer pairs (“Redecker”), non-target sequences are given including non-target AMF. Hyphen indicates that no PCR product was obtained. No PCR products were obtained using the primer combination GIGA1313/GIGA5.8R. Samples

“Krüger”

LSU

ITS2

SSU

“Redecker” GLOM1310/ITS4i

LETC1677/ITS4i

ACAU1661/ITS4i

ARCH1311AB/ITS4i

PARA1313/ITS4i

GLOC1355/ITS4i

LIT LOB TAN LEU BMG BPG BPF a Total

27/0/9 22/0/11 29/0/4 27/0/8 33/0/4 23/0/8 28/0/2 189/0/46

25/0/1 29/0/0 34/0/0 27/0/0 33/0/0 35/0/0 35/0/0 218/0/1

39/0/0 41/0/0 35/0/0 32/0/0 33/0/0 34/0/0 33/0/0 247/0/0

23/1/0 55/0/0 19/17/0 24/8/0 31/13/0 34/14/0 30/2/0 216/55/0

9/0/0 10/0/0 17/0/0 12/0/0 8/0/0 11/3/0 10/1/0 77/4/0

e e e 10/0/0 11/1/0 8/1/0 5/2/0 34/4/0

e 10/0/0 0/14/0 e e e e 10/14/0

e 13/0/0 0/31/3 0/22/0 e 4/7/0 0/11/0 17/71/3

e e 15/0/0 e e 12/0/0 12/0/0 39/0/0

e 0/16/0 4/6/0 0/14/0 0/11/0 0/12/0 2/9/0 6/68/0

a

Total number of sequences obtained by each primer set.

3. Results Altogether we sequenced 1319 clones. After excluding non-AMF and chimeric sequences, we included 1053 good-quality AMF sequences in our phylogenetic analyses (Table 3). Non-AMF amplicons (mainly plant, basidiomycetous and ascomycetous DNA) were detected by the “Redecker” primer system (39% of sequences analysed) and by the SSU primers (20%). The “Redecker” primers GLOC1355/ITS4i and ARCH1311AB/ITS4i (specific for the Diversisporaceae and Archaeosporales, respectively) co-amplified nontarget AMF families Glomeraceae and Paraglomeraceae. The “Krüger”, LSU and ITS2 primers amplified only the Glomeromycota; however, the “Krüger” primers seem to be most prone to the creation of chimeras (20% of sequences analysed). 3.1. AMF diversity Altogether 41 MOTUs were delimited in the LSU þ ITS region amplified by the “Krüger”, LSU, ITS2 and “Redecker” primers, of which 28 belonged to the Glomeraceae, five to the Claroideoglomeraceae, three to the Diversisporaceae, two to the Paraglomeraceae, two to the Archaeosporales and one to the Acaulosporaceae; no representatives of the family Gigasporaceae were detected in the whole study (Figs. 2, S1). Only the “Redecker” family-specific primers were able to detect the representatives of six AMF families. The “Krüger” and ITS2 primers detected members of five families, while only three AMF families were detected by the LSU primers (Fig. 3). In general, the “Krüger” system tended to reveal higher AMF diversity (Table 4). Nevertheless, only the Shannon diversity index differed significantly between the “Krüger” and SSU primer systems (df ¼ 3, F ¼ 3.72, P ¼ 0.025). As concerns the SSU primers, 17 MOTUs were detected using the 97% threshold (14 MOTUs of the Glomeraceae and one of each of the Claroideoglomeraceae, Archaeosporales and Diversisporaceae), while 25 monophyletic groups of AMF MOTUs were defined according to the topology of phylogenetic tree (Fig. 4). According to clustering analyses (TOPALi), the phylogenic definition of MOTUs matched well with the 98% threshold, which divided the SSU sequences into 27 MOTUs (data not shown). Unfortunately, because of the independent MOTU delimitation in the SSU region, the comparison with other primer systems was limited to analyses of

the relative abundance of the dominant MOTUs and a comparison of representation of different AMF families (Fig. 3). Rarefaction analyses suggested that clone-picking intensity was sufficient except for the samples BPG and BPF, which were the richest in AMF species, where the species accumulation curves for the “Krüger” and LSU primers did not reach an asymptote at >20 clones per sample (Fig. S3). 3.2. Quantitative aspects of MOTU detection Comparing semi-quantitative abundances of AMF taxa defined in the LSU þ ITS region, all relevant primer systems detected the same most dominant MOTUs in each sample except for BPG, where the LSU results differed (Table 5). Congruently, the MOTU frequencies obtained by the different primer systems were pairwise significantly correlated, with Spearman’s correlation coefficients (probability levels) as follows: “Krüger” e LSU 0.65 (P < 0.001); “Krüger” e ITS2 0.57 (P < 0.01); LSU e ITS2 0.60 (P < 0.01) (Fig. S4). However, the frequencies of the MOTUs tended to be more homogenous when determined by the “Krüger” primer combination (Table 4), as documented by significantly lower relative abundance of the most abundant MOTUs compared to the SSU primers (DF ¼ 3, F ¼ 3.90, P ¼ 0.021). 3.3. Detection of specific MOTUs by “Krüger”, “Redecker”, LSU and ITS2 primer systems The most common MOTU GLOM-1, which corresponds to the morphologically-defined species Rhizophagus irregularis, was found in all samples from the terrestrial environment by all primer systems. Similarly, the most abundant AMF from the aquatic samples, GLOM-19, was also detected by all primer systems. At least one of the primer systems mostly failed to detect the rest of the AMF (Table 6). Several non-Glomeraceae MOTUs (CLAR-4, ARCH-2, PARA-1 and PARA-2) were detected exclusively by the “Redecker” primer system. On the other hand, this system did not detect several MOTUs of the Glomeraceae family (GLOM-7, GLOM-10, GLOM-13, GLOM-15, GLOM-16) in samples where they were detected by all other primers. Interestingly, certain other MOTUs were consistently (in at least two samples) detected only by one primer system: two MOTUs (DIV-2 and GLOM-12) were detected exclusively by the “Krüger”

Fig. 2. Phylogenetic tree of the Glomeromycota based on a neighbour-joining analysis of the “Krüger” fragment (1055 characters). The numbers above or below the branches denote neighbour-joining bootstrap values from 1000 replications. The tree was rooted by Paraglomus laccatum, P. brasilianum and P. occultum. Sequences obtained in this study are shown in bold and were all obtained with the “Krüger” primers. They are labelled with the database accession number and the samples codes: LIT e Littorella uniflora, LOB e Lobelia dortmanna, TAN e Tanacetum vulgare, LEU e Leucanthemum ircutianum, BMG e Briza media grassland, BPG e Brachypodium pinnatum grassland, BPF e Brachypodium pinnatum field. Vertical lines show the delimitation of MOTUs, rectangles show the delimitation of classes or orders within the Glomeromycota.

488


Redecker

SSU 1

1

LSU 1

1 2

2

1 2

8

2

19

14 5

Krüger

Glomeraceae Claroideoglomeraceae Archaeosporales 1 Diversisporaceae 3 Paraglomeraceae Acaulosporaceae

I TS2

1

1

3

2 1 2

21

17

Fig. 3. Frequencies of MOTUs in glomeromycotan families as yielded by the different primers across samples. Numbers next to the segments indicate the numbers of MOTUs in each family.

primers, GLOM-24 by the LSU primers and GLOM-25 by the ITS2 primers. An overview of MOTUs detected in each root sample by different primer combinations is presented in Table 6.

Table 4 Diversity parameters of the analysed samples as obtained by the different primers. Parameter

System

“Krüger”

LSU

ITS2

SSU

“Redecker”

No. of MOTUs

BMG BPG BPF LIT LOB LEU TAN BMG BPG

7 11 10 5 5 5 4 1.63 2.15

5 7 9 5 4 5 3 1.00 1.29

6 5 7 4 4 3 4 1.41 1.16

5 6 6 2 2 5 4 1.04 1.04

4 6 7 3 4 5 5 n. d. n. d.

Shannon index

BPF LIT LOB LEU TAN BMG BPG BPF

1.92 1.49 1.18 1.47 1.05 10 4 31 16 13 3

1.75 1.28 0.79 0.97 0.43 41 10 4 11 3

1.54 0.60 1.04 1.10 1.31 84 40 11 7

1.39 0.46 0.16 1.22 0.85 50 12 7 94

n. n. n. n. n. n. n. n.

d. d. d. d. d. d. d. d.

a

LIT LOB LEU TAN BMG BPG BPF

50 74 50 40 0.30 0.22 0.41

50 52 51 30 0.48 0.49 0.34

40 40 30 40 0.42 0.50 0.33

20 20 62 41 0.68 0.65 0.4

n. n. n. n. n. n. n.

d. d. d. d. d. d. d.

LIT LOB LEU TAN

0.29 0.55 0.31 0.59

0.56 0.72 0.70 0.88

0.87 0.61 0.39 0.35

0.83 0.96 0.54 0.74

n. n. n. n.

d. d. d. d.

Chao 2

b

FD1

n. d. not determined. a Numbers indicate estimated richness of AMF species in each sample Chao 2 Standard Deviation. b Frequency of the most abundant MOTU.

When comparing the phylogenetic results based on trimmed alignments (LSU; ITS and ITS2) with the whole “Krüger” fragment, we found some differences in the species resolution power between regions. Major differences in species resolution power were observed especially in the less diverse families. According to the LSU as well as the ITS2 tree, all Diversisporaceae sequences clustered within a single MOTU. A similar pattern was observed for the Claroideoglomeraceae, where only two taxa were delimited by the LSU or the ITS2 fragment. Details about other closely related taxa with weak support in the shorter fragments are provided in the corresponding phylogenetic trees (Fig. S1, S2). 4. Discussion 4.1. Technical problems: non-specific amplification and chimeras Some AMF primer systems suffer from co-amplification of other fungal groups or plant DNA. This makes analyses of AMF communities by cloning and sequencing more costly and labour-intensive. It may also be a source of important errors when PCR products serve as basis for fingerprinting approaches such as peak-profile terminal restriction fragment length polymorphism (Dickie and FitzJohn, 2007). Previously, this phenomenon was shown for the combination of SSU primers AM1 (Helgason et al., 1998) with NS31 (Simon et al., 1992; Lumini et al., 2010; Alguacil et al., 2011). Another presumably AMF-specific primer pair AML1/AML2 has been designed in the SSU by Lee et al. (2008) to overcome the problem with amplifying basal lineages of the Glomeromycota. However, this primer system can also amplify a significant proportion of plant DNA (Alguacil et al., 2011). Recently, a combination of both above mentioned SSU primer systems has been proposed, resulting in the primer combination NS31/AML2, optionally used in a nested approach with universal fungal primers (Liu et al., 2011). However, our results show e congruently with Liu et al. (2011) e that even this system to a high extent co-amplifies Asco- and Basidiomycota (ca. 20%). Similarly, a high extent of non-specific amplification (53% in total) was observed using the “Redecker”


primer system (Redecker, 2000), yielding both non-AMF (plant, Asco- and Basidiomycota) and non-target AMF sequences, which corroborates previous reports (e.g. Kovács et al., 2007; Sýkorová et al., 2007). Although the proportion of such non-specific amplifications might be reduced by “hot start” PCR conditions, they persist in the absence of the target taxa, especially in primer combinations for the Archaeosporales, Paraglomeraceae and Acaulosporaceae (Appoloni et al., 2008). Interestingly, none of the tested primer systems, except for “Krüger”, produced chimeric sequences. In contrast to a previous field study adopting the “Krüger” system where only 5% of chimeric sequences was identified using a common program for chimera detection (Wang et al., 2011), our careful manual searching revealed that up to 20% sequences might be chimeric. We therefore suggest that particular attention should be paid to the occurrence of chimeras in all studies using the “Krüger” system because commonly used programmes that check for the presence of chimeras are probably less suitable for the detection of chimeras in the whole fragment. 4.2. AMF diversity as described by the different primer systems Within the LSU þ ITS region, few consistent differences in AMF species richness were found between the tested primer systems except for the “Krüger” primers, which yielded relatively higher diversity parameters in most of the tested samples out of the three directly comparable primer systems (“Krüger”, LSU, ITS2) and also showed significantly higher Shannon diversity indexes than SSU primers. This result supports the conclusions of Krüger et al. (2009) that this primer system amplifies most of the AMF species and offers high resolution power for AMF diversity studies. Nevertheless, according to our best knowledge, our study is only the third one to use the “Krüger” system for describing native AMF communities. More work must therefore be done before a general conclusion can be drawn. Congruently with the study of Stockinger et al. (2010), we have shown that the “Krüger” fragment offers higher species resolution power compared to the other tested regions. Although shorter fragments exhibited a relatively high resolution power for the most diverse family (Glomeraceae), they were weaker in delimitation of species from less abundant families (Claroideoglomeraceae or Diversisporaceae). However, this obvious disadvantage might be overcome by using the phylogenetic reference dataset for the Glomeromycota (Krüger et al., 2012), except for lineages without any described species. The two MOTU delimitation approaches in the SSU, based on the 97% and approx. 98% sequence similarity thresholds, yielded highly divergent numbers of MOTUs (17 and 27, respectively). This must be kept in mind when comparing results from studies in which different thresholds were used for AMF taxa delimitation. For example, Liu et al. (2011) recently showed a high diversity (21 phylotypes) of AMF from the Tibet region; however, because their delimitation level was 98%, we can expect relatively lower richness than if they had applied the most commonly used 97% threshold level. In future studies, the general mixed Yule-coalescent model may offer a new and perhaps more objective method for delimiting fungal taxa in the SSU as well as in other markers (Powell et al., 2011). 4.3. Community composition as described by different primer systems in the LSU þ ITS region Remarkably, the use of the “Redecker” primer system resulted in a different description of AMF community composition compared to the other primer systems. This discrepancy arises from primers

489

addressing less abundant AMF lineages such as the Claroideoglomeraceae or the Paraglomeraceae, which were not detectable with our sequencing effort by other primer combinations. On the other hand, the “Redecker” Glomeraceae-specific primers did not detect numerous MOTUs detected by the other primer systems. This can be attributed to the similar clone numbers sequenced per each family-specific primer pair, which have probably led to undersampling in the Glomerales. This is nevertheless in accordance with the methodologies employed in previous studies using this primer system (e.g. Sýkorová et al., 2007). Taken together, we conclude that results based on the “Redecker” primer system are hardly comparable with the results of studies using one general primer combination for the amplification of all AMF lineages. Contrary to the “Redecker” primers, no other primer system detected any species of Paraglomeraceae in our study, even though members of this family are well amplifiable by “Krüger” (Krüger et al., 2009) as well as SSU primer systems (Lopéz-García et al., 2013). The most probable explanation why this group is more easily detectable by specific primers lies in the naturally low root colonization level by the Paraglomeraceae, even though they can dominate in the surrounding soil (Öpik et al., 2003; Hempel et al., 2007). In other families, such as the Diversisporaceae and Glomeraceae, the “Krüger” system detected almost all MOTUs, with only few exceptions in very rare taxa detected mostly as singletons by the other primer systems. Congruently with previous reports, we found a strong bias towards the Glomeraceae in the LSU primer system, discriminating several Glomeromycota lineages such as the Diversisporaceae (Gamper et al., 2009), Paraglomeraceae, Claroideoglomeraceae and Archaeosporales (Krüger et al., 2009). Since this bias is a result of mismatches in the FLR4 primer (Gamper et al., 2009), it could not be alleviated by the newly proposed primer combination 250f/FLR4 (Sýkorová et al., 2012). The newly designed forward primer ITS7o located in the 5.8S region (derived from the ITS7 primer published by Ihrmark et al. (2012)) amplified the ITS2 region in combination with the ITS4 primer and detected all of the AMF families revealed by the “Krüger” primers. As demonstrated by Stockinger et al. (2010), the ITS or ITS2-LSU region (to the primer LSUmBr) has a high resolution power for AMF taxa and has been therefore suggested as a suitable target fragment for the pyrosequencing approach. However, a suitable forward primer to amplify this region has not been made available because the ITS3 primer largely mismatches with most species from the genera Rhizophagus and Sclerocystis, the Ambisporaceae and some Acaulospora species (Stockinger et al., 2009). The ITS7o primer overcomes these problems and in combination with ITS4 as a reverse primer can therefore be recommended for the second step of a nested PCR approach, following AMF-specific amplification by “Krüger” primers SSUmAf/LSUmAr in the first step. This primer combination might be especially suitable for the new next generation sequencing (NGS) approach e Ion Torrent, which is cheaper than other NGS methods (like 454sequencing or SMRT-sequencing) but offers only shorter fragments with maximal length 400 bp (Glenn, 2011). Alternatively, the AMF-specific primer LSUmBr can be considered a suitable reverse primer to ITS7o, but this combination requires further testing. Although the MOTUs detected in the SSU region could not be directly linked with those obtained by the other primer systems, the comparison of MOTUs representation in different Glomeromycota families showed considerable underestimation of Claroideoglomeraceae, Diversisporaceae or Paraglomeraceae diversity in the SSU clone libraries. Moreover, previous studies described relatively low variability and resolution power of the SSU region to

490

P. Kohout et al. / Soil Biology & Biochemistry 68 (2014) 482e493 92 AM946888 Uncultured Glomeromycota roots A. alpinus Austria 63 HE775411 B. pinnatum field JN009163 Uncultured Glomeromycota roots E. nutans China 80 HE775437 B. media 100 JN009227 Uncultured Glomeromycota roots China 90 HE775429 B. media 98 JN009483 Uncultured Glomeromycota soil China HE775432 B. media 64 HE775433 B. media HE775412 B. pinnatum field 97 AJ716006 Uncultured Glomeromycota roots A. pseudoplatanus 100 HE775416 B. pinnatum field HE775410 B. pinnatum field AM946883 Uncultured Glomeromycota roots S. varia Austria 58 EF177582 Uncultured Glomeromycota soil HE775434 B. media 91 HE775407 B. pinnatum field FR693424 Uncultured Glomeromycota roots Spain 65 92 HE775436 B. media HE775426 B. pinnatum meadow FN556643 Uncultured Glomeromycota roots UK FJ831550 Uncultured Glomeromycota roots P. falcatus HE775423 B. pinnatum meadow 56 56 FJ875131 Uncultured Glomeromycota roots A. sulcata Portugal

A-11 / MOTU-17 A-10 / MOTU-5 A-7 / MOTU-4 A-12 / MOTU-16 A-13 A-16

HM153424 Glomus iranicum HE775415 B. pinnatum field A-9 HE775395 L. ircutianum 95 AM779228 Uncultured Glomeromycota roots P. lanceolata Austria 79 AB556932 Uncultured Glomeromycota roots S. tosaensis Japan HE775404 B. pinnatum field 59 FN645973 Uncultured Glomeromycota roots R. sphaerocarpa Spain 57 HE775418 B. pinnatum meadow 87 HE775420 B. pinnatum meadow 89 HE775401 L. ircutianum A-14 MOTU-15 GU322394 Uncultured Glomeromycota roots Acer sp. USA 100 GQ140619 Uncultured Glomeromycota roots P. frutescens A-8 / MOTU-6 HE775421 B. pinnatum meadow HE775422 B. pinnatum meadow A-15 98 HE775431 B. media 63 HE775417 B. pinnatum meadow EU123464 Uncultured Glomeromycota roots grass USA 64 98 AM779205 Uncultured Glomeromycota roots P. lanceolata Austria FJ875127 Uncultured Glomeromycota roots A. sulcata Portugal HE775398 L. ircutianum 76 A-5 MOTU-2 HE775382 L. uniflora A-6 64 96 HE775383 L. uniflora 70 AM911126 Uncultured Glomeromycota roots T. repens Ireland AJ563895 Uncultured Glomeromycota roots P. australis Germany A-4 / MOTU-3 98 HE775427 B. pinnatum meadow 98 DQ085216 Uncultured Glomeromycota roots J. procera DQ085218 Uncultured Glomeromycota roots J. procera A-3 99 HE775396 L. ircutianum AM849323 Uncultured Glomeromycota roots H. nobilis Estonia 100 AJ306439 Acaulospora longula Z14005 Acaulospora rugosa 67 AF485889 Glomus hoi A-2 97 69 AJ301857 Glomus sp. W3347 cons#36 Acaulospora sieverdingii 61 HE775428 B. media 94 AJ306442 Acaulospora scrobiculata FR772325 Glomus macrocarpum 81 Z14004 Acaulospora spinosa AF485890 Glomus hoi? 63 FJ009670 Acaulospora mellea MOTU-1 74 HE775386 L. dortmanna FR750213 Acaulospora capsicula A-1 69 FN556630 Uncultured Glomeromycota roots UK cons#40 Acaulospora laevis 94 78 82 HE775385 L. dortmanna Z14006 Acaulospora colombiana 80 87 HE775399 L. ircutianum cons#41 Acaulospora brasiliensis AM779219 Uncultured Glomeromycota roots P. lanceolata Austria cons#34 Redeckera fulva JF414185 Uncultured Glomeromycota thallus P. phyllanthus New Zealand 62 Z14010 Gigaspora gigantea HE775380 L. uniflora 59 75 AJ567844 Gigaspora margarita AM779256 Uncultured Glomeromycota roots Austria 100 X58726 Gigaspora rosea HE775397 L. ircutianum 52 cons#15 Racocetra castanea HE775381 L. uniflora 100 AJ505812 Glomus viscosum? cons#17 Scutellospora spinosissima 99 cons#51 Septoglomus africanum GQ376067 Dentiscutata colliculosa 97 FJ009619 Funneliformis geosporus Z14013 Scutellospora heterogama 60 AJ301858 Funneliformis verruculosus 93 83 AJ619945 Pacispora scintillans Y17653 Funneliformis caledonius 100 FR750375 Pacispora franciscana 94 AJ276086 Funneliformis coronatus cons#27 Diversispora eburnea 55 AJ306438 Funneliformis mosseae 100 cons#26 Diversispora celata AJ315516 Septoglomus constrictum? cons#30 Diversispora aurantia 96 HE775389 T. vulgare A-17 / MOTU-10 cons#29 Otospora bareae 99 HE775392 T. vulgare 50 AJ276077 Diversispora spurca AY919845 Uncultured Glomeromycota roots A. angustifolia Brazil FR686955 Diversispora trimurales 67 HE775406 B. pinnatum field 63 AJ563897 Uncultured Glomeromycota roots P. australis Germany AM946964 Glomus macrocarpum? 93 HE775419 B. pinnatum meadow cons#32 Diversispora epigaea A-18 / MOTU-13 71 90 X86687 Diversispora epigaea HE775414 B. pinnatum field HE775400 L. ircutianum A-20 / MOTU-11 63 HE775388 T. vulgare 97 DQ336465 Uncultured Glomeromycota roots T. chrysantha Ecuador DIV / MOTU-7 HE775390 T. vulgare 100 HE775391 T. vulgare Y17652 Viscospora viscosa? A-19 /MOTU-12 HE775393 T. vulgare 99 cons#69 Claroideoglomus claroideum CLAR / MOTU-8 77 GU353680 Uncultured Glomeromycota soil Italy HE775409 B. pinnatum field AJ133706 Sclerocystis sinuosa 89 FR750220 Claroideoglomus lamellosum 94 97 Y17648 Rhizophagus manihotis Z14008 Claroideoglomus etunicatum 98 cons#61 Rhizophagus clarus AJ276089 Glomus luteum 99 AF213462 Rhizophagus proliferus 100 cons#83 Archaeospora schenckii cons#60 Rhizophagus proliferus cons#82 Archaeospora trappei 54 HE775403 B. pinnatum field A-21 (Rhizophagus intraradices) HE775408 B. pinnatum field EU573738 Uncultured Glomeromycota roots C. sparsiflora cons#59 Rhizophagus intraradices 50 AF131054 Uncultured Glomeromycota roots H. non-scripta 64 AJ418854 Rhizophagus intraradices 65 HQ258988 Uncultured Glomeromycota soil 60 HE775424 B. pinnatum meadow HE775384 L. dortmanna A-22 (Rhizophagus irregularis) ARCH / MOTU-9 57 86 63 HE775413 B. pinnatum field 96 HE775387 L. dortmanna HE775402 L. ircutianum AB365834 Unc. Glomeromycota roots M. oleifera Uganda FR750374 Rhizophagus vesiculiferus cons#79 Ambispora granatensis 58 62 FJ009605 Rhizophagus irregularis MOTU-14 AJ276074 Geosiphon pyriformis cons#57 Rhizophagus irregularis 82 AB047305 Ambispora callosa 74 cons#58 Rhizophagus irregularis 69 79 AJ301861 Ambispora leptoticha Y17640 Rhizophagus fasciculatus DQ396687 Ambispora 98 FJ009617 Rhizophagus irregularis AM400227 Ambispora gerdemannii FR750223 Rhizophagus irregularis HE775435 B. media 100 AJ006799 Paraglomus occultum HE775394 L. ircutianum AM295493 Paraglomus laccatum 60 HE775425 B. pinnatum meadow AJ301862 Paraglomus brasilianum 98 HE775430 B. media 69 HE775405 B. pinnatum field 0.02 cons#64 Glomus indicum


491

Table 5 Frequency ranks of the MOTUs determined by the various AMF-general primers in the LSU þ ITS region (K e “Krüger”, L e LSU, I e ITS2) in each sample (only those MOTUs are included that were found by all three primer combinations in at least one sample). LIT K ARCH-1 GLOM-1 GLOM-5 GLOM-7 GLOM-9 GLOM-10 GLOM-13 GLOM-15 GLOM-16 GLOM-18 GLOM-19

LOB L

I

TAN

K

L

I

2

2

3

1 2 3

1 3 2

1

3

2

2 1

3 1

BMG

BPG

BPF

L

I

K

L

I

K

L

I

K

L

I

K

L

I

3 2

3 2

2 3

1

1

1

1

1

1

1

2

2

1

1

1

2

2

2

2 3

2 3

2 3

3 1

1 3

3 1 2

2

2

1 1 4 3

LEU

K

1

1

2 1

species level in genera such as Ambispora, Diversispora and Scutellospora (de Souza et al., 2004; Walker et al., 2007; Gamper et al., 2009). 4.4. Quantification of abundances of AMF taxa For a long time, researchers have debated the suitability of quantitative information obtained from nested-PCR methods. In their quantitative comparison of one-step PCRs, Öpik et al. (2009) showed that the proportions of individual AMF taxa were highly similar when the same sample was amplified in independent PCRs and subsequently processed by 454-sequencing or cloning followed by Sanger sequencing. However, our study is the first which compares semi-quantitative results obtained by nested-PCR techniques using three different primer systems on two target regions of rRNA. The results were largely comparable for all primer systems, which strongly justifies consideration of this method for interpretations of semi-quantitative abundances. On the other hand, we have to keep in mind that these results may not mirror real proportions in fungal biomass (Alkan et al., 2004) or the activity of AMF (Vandenkoornhuyse et al., 2007). According to quantitative results derived mainly from the SSU rRNA region, Dumbrell et al. (2010) concluded that the structure of AMF communities is usually overdominated by a single taxon. Although we observed similar overdominance in our samples, the abundance of the most dominant AMF MOTU was significantly lower when detected by the “Krüger” primers compared to the SSU primers. This discrepancy indicates possible bias introduced by the rRNA region and primer choice to quantitative interpretation of molecular data. 4.5. Is there a most suitable rRNA region for describing AMF communities? A growing number of databases collect nucleotide sequences and supporting metadata. Although the most commonly used International Nucleotide Sequence Database (INSD) contains a huge amount of sequence data, it also contains non-negligible proportion of erroneous sequences (Nilsson et al., 2006) and lacks a lot of supporting information. Curated databases, where source data are annotated by expert scientists, represent a

suitable alternative to the INSD (Abarenkov et al., 2010a; Öpik et al., 2010; Tedersoo et al., 2011). These databases contain an approximately ten times lower amount of erroneous sequences than the INSD, thus representing the most reliable recent workbenches for environmental MOTU delimitation. In relation to the ever increasing use of high throughput sequencing technologies, the importance of well annotated fungal sequence databases will become even more pronounced. Nowadays, there are two well annotated databases for Glomeromycota sequences, the first for the SSU region (MaarjAM database; Öpik et al., 2010) and second one for the ITS sequences (UNITE; Abarenkov et al., 2010b). In addition to well annotated sequences, these databases also contain all available metadata related to each sequence, which is a prerequisite for meta-analyses studies. In contrast to SSU or ITS regions, studies based on LSU region have to rely only on sequences deposited in the INSD, which is highly discriminative. An important aspect when choosing the target rRNA region for AMF community studies is also the abundance of comparable sequences in public nucleotide databases. In total, 7718 ITS and 3539 SSU sequences with affinities to the Glomeromycota are available in the UNITE and MaarjAM databases, respectively (valid for February 2012). Similarly, the INSD contains 5026 LSU AMF sequences (Kivlin et al., 2011; status on 15 March 2010). As the ITS fragment is the most widely used fungal barcode region (Schoch et al., 2012), a growing number of fungal ITS sequences is expected to land in public databases in the near future in connection with a wider application of the 454 sequencing approach. General ITS fungal primers do not discriminate well against the Glomeromycota (Stockinger et al., 2010; Kohout et al., 2012) so the number of AMF sequences is expected to expand more rapidly in the ITS region compared to other regions. As a result, the ITS region may provide a more robust dataset for AMF meta-analyses and global studies in the future. Moreover, the ITS region also has one specific characteristic, which might potentially be useful for studying active soil components. Compared to the SSU or LSU rRNA regions, ITS is cleaved from the precursor rRNA molecules a short time after transcription (Anderson and Parkin, 2007). Therefore, such a rapid turnover of ITS region affords great opportunity to study active AMF communities by RNA-based approaches, as has recently been shown for the Ascomycota and Basidiomycota (Baldrian et al., 2012).

Fig. 4. Phylogenetic tree of the Glomeromycota based on a neighbour-joining analysis of the SSU fragment (489 characters). The numbers above or below branches denote neighbour-joining bootstrap values from 1000 replications. The tree was rooted by Paraglomus laccatum, P. brasilianum and P. occultum. Sequences obtained in this study are shown in bold and were all obtained with the SSU primers. They are labelled in the same way as for Fig. 2. Rectangles show the delimitation of MOTUs based on the phylogenetic approach, labels “MOTUs-x” show the delimitation of MOTUs based on the 97% full-length sequence similarity using TOPALi 2.5.

492


Table 6 Overview on the MOTUs detected by the various primers in the LSU þ ITS region (K e “Krüger”, L e LSU, I e ITS2, R e “Redecker”). Shading indicates frequencies of the MOTUs in the corresponding clone library, in four classes: 50% (darkest shading); 20%; 10%;