SEARCH

SEARCH BY CITATION

Keywords:

  • arbuscular mycorrhiza (AM);
  • Glomus phylotypes;
  • liverwort;
  • Marchantia foliacea;
  • ribosomal DNA

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • • 
    Microscopic evidence suggests that fungi forming endosymbioses with liverworts in the Marchantiales are arbuscular mycorrhizal (AM) fungi from the Glomeromycota. Polymerase chain reaction amplification of ribosomal sequences confirmed that endophytes of the New Zealand liverwort, Marchantia foliacea, were members of the genus Glomus.
  • • 
    Endophytes from two Glomus rDNA phylotypes were repeatedly isolated from geographically separated liverwort samples.
  • • 
    Multiple phylotypes were present in the same liverwort patch. The colonizing Glomus species exhibited substantial internal transcribed spacer sequence variation within phylotypes.
  • • 
    This work suggests that certain liverwort species may serve as a model for studying DNA sequence variation in colonizing AM phylotypes and specificity in AM–host relationships.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Arbuscular mycorrhizal (AM) fungi in the phylum Glomeromycota (Schüßler et al., 2001) form symbioses with the majority of land plants (Morton & Benny, 1990). They are characterized by an obligate symbiotic life and the formation of intracellular structures (i.e. the arbuscules) and are considered to be one of the most ecologically important groups of fungi. The AM fungi probably originated more than 450 million years ago (Redecker et al., 2000) and arbuscule-like structures have been found in fossils of Aglaophyton, the earliest known vascular land plant (Remy et al., 1994). The coevolution of plants and AM fungi may have been a key factor enabling the first, rootless plants to colonize land (Pirozynski & Malloch, 1975; Simon et al., 1993; Schüßler, 2002).

Extant liverworts and hornworts possibly have features of the earliest plants to colonize land (Edwards et al., 1995; Nickrent et al., 2000; Qiu & Lee, 2000; Wellman et al., 2003). Both groups often form symbioses with fungi (Read et al., 2000). Basidiomycete fungi form associations with leafy hepatics in the Jungermanniales and the thalloid species Aneura pinguis and Cryptothallus mirabilis in the Metzgeriales. The ascomycete Hymenoscyphus ericae is also a common symbiont of the Jungermanniales (Duckett & Read, 1995). Microscopic examinations have suggested that the associations formed by many thalloid liverworts and hornworts are with AM fungi in the phylum Glomeromycota. Arbuscular mycorrhiza fungi have been described colonizing many species of simple thalloid liverworts in the Metzgeriales (Pellia, Fossombronia) and Calobryales (Haplomitrium), as well as complex thalloid species in the Marchantiales (Boullard, 1988; Read et al., 2000; Carafa et al., 2003). Ultrastructural studies have shown that infections of the marchantialian liverwort Conocephalum conicum and the hornwort Phaeoceros laevis were of the AM type (Ligrone, 1988; Ligrone & Lopes, 1989). Schüßler (2000) also showed that the hornwort Anthoceros punctatus was able to form an AM-like symbiosis with two isolates of Glomus claroideum. However, no species of AM fungi associating with liverworts in natural conditions have yet been identified.

With the advent of polymerase chain reaction (PCR) amplification, comparisons of ribosomal RNA (rRNA) gene sequences have been widely adopted as a means for identifying species and strains of fungi. For example, Bidartondo et al. (2003) recently identified the basidiomycete endophyte of the achlorophyllous liverwort Cryptothallus mirabilis and demonstrated that the same fungus, a Tulasnella species, formed ectomycorrhizas on the roots of surrounding trees. Molecular techniques are especially useful for identification and detection of AM fungal endophytes, which are notoriously difficult to identify by morphology alone. AM fungal small subunit (SSU) or large subunit (LSU) rDNA sequences can be placed in context with rDNA sequences from other AM species without specifically comparing the morphology of each fungal isolate. For this reason, we refer to our data as fungal DNA phylotypes.

In this study, we examined arbuscule-forming fungi from Marchantia foliacea, a complex thalloid liverwort found growing in New Zealand, Australia and the Juan Fernandez Islands (Bischler-Causse, 1989). Marchantia foliacea has a wide distribution throughout the North and South Islands of New Zealand, growing mainly in lowland forests where the predominant mycorrhizal association is arbuscular (Cooper, 1976; Johnson, 1977). Polymerase chain reaction-based analyses of nuclear rRNA genes confirmed that M. foliacea endosymbionts were from the Glomeromycota. Of nine proposed families in the phylum Glomeromycota (Schüßler et al., 2001), we detected only species from Glomus group A. We found evidence for repeated relationships between M. foliacea and specific Glomus phylotypes. Some of these phylotypes were also observed in root nodules of New Zealand podocarp trees.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Samples measuring 5 × 5 cm of M. foliacea containing several thalli were collected from wet banks and streamsides at eight sites in New Zealand's South Island (Fig. 1). Samples were kept cool until examination within 2 d of collection.

image

Figure 1. Liverwort collection sites. (a) Collection sites across the South Island of New Zealand. (b) Collection sites within the Te Roto o Wairewa/Lake Forsyth catchment on Banks Peninsula, New Zealand.

Download figure to PowerPoint

Five samples, from Ross (I, II and III), Hari-Hari and Karamea, were from lowland West Coast rainforest sites of mixed podocarps with a broadleaf understorey. The dominant podocarp species at these lowland sites were Dacrydium cupressinum (Rimu), Dacrycarpus dacrydioides (Kahikatea) and Prumnopitys ferruginea (Miro). The Ross samples were collected from the same 4 m2 bank site on three separate occasions over the course of a year. A further sample from the Otira Valley was collected from a more elevated inland forest, but one still within a New Zealand lowland classification. Four samples from Banks Peninsula (Reynolds Valley, Montgomery Road, Port Levy Road and Kinloch) on the east coast of the South Island were from sites within 10 km of one another in the Te Roto o Wairewa/Lake Forsyth catchment area (Fig. 1b). These liverwort samples were gathered from stream edges bordered by regenerating native broadleaf species. Away from the streams the environment was mostly grazed pasture. A small number of podocarps, especially P. totara, remained in the upper part of the catchment area.

Two Marchantia polymorpha samples (Lincoln and Prebbleton) were collected from a culvert and a plant nursery near Christchurch, where they grew among exotic grasses and herbaceous weeds.

The identity of the Marchantia samples was confirmed by microscopic examination. The Montgomery Road sample was identified as M. pileata, a species similar to M. foliacea but having differences in the shape of teeth on the ventral scale appendage margin (Bischler-Causse, 1989). However, no nucleotide differences between a trnL-F gene sequence from this sample and the M. foliacea cpDNA trnL-F sequence (EMBL accession AJ698942) were observed. For the purposes of this study we treated the M. pileata sample as another M. foliacea sample, leaving the separation of these closely related species for future research.

Microscopy techniques

For light microscopy, thallus pieces were finely sectioned and stained for a minimum of 1 h in lactophenol red. For electron microscopy, thalli were cut into approx. 5 mm sections and fixed overnight in 2% paraformaldehyde, 1% glutaraldehyde, 0.5% tannic acid, in phosphate buffer, pH 7.4. After rinsing in buffer, samples were postfixed in 2% OsO4 and dehydrated in a graded acetone series. Samples were embedded in Spurr resin (Electron Microscopy Sciences, Fort Washington, PA, USA) for transmission electron microscopy (TEM) or transferred to amyl-acetate and critical-point dried for scanning electron microscopy (SEM). For TEM, resin blocks were thin-sectioned with a diamond knife and sequentially stained with 2% aqueous uranyl acetate for 10 min and lead citrate for 5 min. Observations were made with a JEOL-1200EX electron microscope (JEOL, Tokyo, Japan) at 60 kV.

Molecular methods

DNA extracts from the Ross I and Hari-Hari samples were prepared from approx. 5-mm wide midsections of thalli. These were extensively cleaned under a microscope to remove external debris. For subsequent DNA extractions, sections of liverwort thalli, approx. 2 mm wide and up to 5 mm long, were dissected from the central endophyte-containing rib. Care was taken to leave little or no surface thallus tissue. From each liverwort patch, two to five DNA extractions were prepared from separate thalli sections. DNA extractions were carried out according to a protocol from Bernatzky and Tanksley (1986). Thalli samples were homogenized in approx. 500 µl extraction buffer (0.35 m sorbitol, 100 mm Tris pH 8.0, 5 mm ethylenediaminetetraacetic acid (EDTA)). After centrifugation at 6500g for 15 min, the pellet was resuspended in 400 µl cetyltrimethylammonium bromide (CTAB) buffer (20 mm Tris pH 8.0, 20 mm EDTA, 0.8 m NaCl and 2% CTAB) plus 80 µl 5% sarcosyl, and incubated at 65°C for 30 min. This solution was extracted once with 800 µl chloroform: iso-amyl alcohol (24 : 1) and then precipitated with isopropanol. DNA samples were resuspended in 20 µl H2O.

One microlitre of DNA template was used in each standard 25 µl PCR reaction, as described in Bulman & Marshall, 1998). In most instances, dilution of the DNA template was not required. For amplifying DNA fragments greater than 1 kb in length, the Expand Long Template PCR kit (Roche, Auckland, New Zealand) was used. The PCR primers Universal C and Universal F (Taberlet et al., 1991) were used to amplify and directly sequence liverwort cpDNA trnL-F fragments. The PCR primers used for amplification of rDNA sequences are shown in Fig. 2. DNA fragments spanning from the 3′ end of the SSU rRNA gene to the 5′ end of the large subunit (LSU) rRNA gene, and contain the internal transcribed spacer (ITS) spacers 1 and 2 and the 5.8S rRNA gene were amplified with the primers ITS5 (White et al., 1990) and ITS26 (AB28, Howlett et al. 1992). These sequences are hereafter referred to as ITS fragments. The DNA fragments of the near full-length SSU plus the above ITS fragments (hereafter referred to as SSU ± ITS fragments) were amplified with the primers NS1 (White et al., 1990) and ITS26. From each liverwort patch the DNA extract yielding the strongest amplification with ITS PCR primers was chosen for further study.

image

Figure 2. Polymerase chain reaction (PCR) primers and amplified DNA fragments. (a) The position of PCR primers in a ribosomal repeat. PHY SPEC, phylotype specific PCR primers designed in the internal transcribed spacer (ITS) 1 spacer. (b) Directly amplified DNA fragments. (c) DNA fragments amplified by nested PCR. First and second round DNA fragments are indicated by the arrows from (b) to (c), or within (c).

Download figure to PowerPoint

For nested PCRs, 1 µl of a first-round PCR reaction was used as a template in subsequent amplifications. Glomeromycota-specific nested PCR was performed with primers, as outlined in Russell et al. (2002) (Fig. 2), as well as with the primers MNS1 and MNS4 (Table 1), designed from the Ross I Glomus sequence (Fig. 2).

Table 1.  Polymerase chain reaction primers designed from Glomus sequences obtained in this work
Primer namePrimer sequence (5′−3′)
MNS1TGCATGTCTAAGTATAAACCATTTATACAGG
MNS4TCCCTAGTCGGCATAGTTTATGGT
MarGR1ACGTTCGAAAAATCATGCAAAATT
OV1R1GAAAAATCTTAGAAACACGTTCGT
OV2R1ACGCGTAAGAAAATCTTAGAAACG
OV3R1TTGAAAAGTAAACCCAACCCAAA
HH1R1AACCTGATGATTAAGACGACCC

Five PCR primers (Table 1) specific for the Glomus phylotypes in M. foliacea were designed within the ITS spacer sequences. These were paired with the NS1, NS7 and ITS5 primers (White et al., 1990) for phylotype-specific PCR.

The PCR products were purified with the High Pure PCR Product Purification Kit (Roche), or were first separated on 1% agarose gels and then purified using the QIAquick Gel Extraction Kit (Qiagen, USA). The PCR products were cloned into the pGEM-T Easy vector (Promega, USA). Insert-containing plasmids were identified by PCR screening of colonies with Sp6 and T7 primers. Near full-length SSU rDNA products were sequenced with the PCR primers used for amplification and internal primers from White et al. (1990) or Table 1.

Phylogenetic techniques

The DNA sequence output was assembled with sequencher (Gene Codes; http://www.genecodes.com). DNA sequences were compared to GenBank 4.1 sequences via blastn (Altschul et al., 1990). Near full-length Glomus SSU rDNA sequences were aligned to a dataset from Schüßler et al. (2001), using the profile alignment function in clustal x (Thompson et al., 1997). Several further sequences from GenBank that gave high blastn matches to the New Zealand Glomus sequences were included in this dataset (see Fig. 5). After initial analyses, the dataset was trimmed of several taxa to hasten further analyses and allow clearer tree presentation. A ‘400 bp’ alignment, from the GLOM1310 primer to the end of the SSU rRNA gene, was created using clustal x. No manual adjustments were made to either alignment. DNA sequences were cropped to exclude the GLOM1310 primer sequence. Because initial DNA sequence comparisons and phylogenies showed that all of the endophyte sequences were from Glomus group A, only sequences from Glomus groups A and B were included in this dataset. The data were analysed with maximum parsimony (MP) and minimum evolution (ME; Saitou & Nei, 1987) (Kimura two-parameter model; Kimura, 1980) using paup*4.0b10 (Swofford, 2002). Initial ME analyses of the 400 bp dataset were extremely slow because of the presence of identical or near-identical sequences (analysis incomplete after 3 d). Consequently, this dataset was trimmed of closely related taxa to allow completion of ME analyses. 1000 bootstraps (Felsenstein, 1985) were performed on both MP and ME trees. For the 400 bp dataset, max trees was set to 500 without adjustment.

image

Figure 5. Phylogenetic tree from near full-length small subunit (SSU) rDNA sequences. The single tree obtained with minimum evolution (K2A model) is shown. Minimum evolution (below) and maximum parsimony (above) bootstrap values of > 70% are shown on each node. Sequences from Marchantia foliacea are shown in bold type. The tree was rooted with Endogone pisiformis. A measure of distance is indicated by the bar at the base of the diagram. The three subclades of Glomus group A (Aa, Ab and Ac) identified in Schwarzott et al. (2001) are indicated by brockets on the right-hand side.

Download figure to PowerPoint

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Microscopy

Below the upper green photosynthetic epidermis, the M. foliacea thallus consisted mainly of highly vacuolated parenchyma cells with a central midrib. Every thallus examined was colonized by aseptate fungal hyphae in parenchyma tissue around the central midrib. Closer examination of stained sections revealed two distinct zones of infection, the lower one lightly stained and the upper one more heavily stained (Fig. 3a). Both SEM and TEM revealed that the lightly stained band of cells contained hyphae travelling longitudinally through the thallus (Fig. 3b,c) with no evidence of arbuscules. In contrast, hyphae in cells of the upper darker staining infection zone were extensively coiled and surrounded by living and senescent arbuscules (Fig. 3d,e). The fine, dichotomously branching hyphae of the arbuscules originated from intracellular trunk hyphae (Fig. 3e).

image

Figure 3. Microscopic examinations of Marchantia foliacea. (a) Cross-section of thallus of M. foliacea, stained with fuchsin red. Two distinct zones of infection surround the area of the central rib. (b) Scanning electron photomicrograph of the lightly staining zone of infection showing cross-sections of intracellular hyphae growing longitudinally through the thallus. (c) Transmission electron photomicrograph showing detail of (b); there is no evidence of arbuscules and the cell appears free of cytoplasm. Fh, fungal hypha. (d) Transmission electron photomicrograph of a cell in the upper infection zone containing large trunk hyphae (Fh) surrounded by much smaller hyphae forming the branches of arbuscules (Ar). As in (c) there is no evidence of living cytoplasm in the cell. (e) Scanning electron photomicrograph of arbuscules in the upper zone of infection.

Download figure to PowerPoint

Infection appeared to be entirely intracellular with hyphae crossing directly through cell walls (Fig. 3b,c). The fungus was separated from the host cytoplasm by a perifungal membrane that was continuous with the host plasmalemma (Fig. 4). All hyphae were aseptate and relatively thin-walled, with trunk hyphae being large and uniform in size. Fungal hyphae colonized the thallus through the smooth rhizoids, which in M. foliacea formed two clumps on either side of the midrib at the base of the growing thallus.

image

Figure 4. Microscopic examinations of Marchantia foliacea. (a) Fungal hyphae (Fh) crossing through the host cell wall. A layer of deposited material (FM) surrounds the hyphae. (b) Nomarski optics of parenchyma cells immediately below the epidermis packed with fungal hyphae.

Download figure to PowerPoint

Molecular analyses

Universal primers used to amplify ITS fragments from the Ross I sample produced a cluster of DNA fragments of 500–700 bp in size and a bigger DNA band of c. 1.5 kb. In the first screen of 500–700 bp clones, DNA was amplified from 19 inserts. Six representatively sized inserts were sequenced. Of these, one was a Glomus sequence and the others were sequences from ascomycetes (three from a Tetracladium-related species, EMBL AJ810901; EMBL AJ810900) and a basidiomycete (EMBL AJ810902). Later, partial sequencing of clones from the c. 1.5 kb ITS fragment showed these were from the liverwort host (EMBL AJ810904).

To obtain a near full-length SSU rDNA sequence, a specific PCR primer (MarGR1) was designed in the ITS1 spacer of this first Glomus sequence (Fig. 2). When MarGR1 was paired with the NS1 primer (White et al., 1990), a c. 1.5 kb DNA fragment was amplified instead of the expected c. 1.8 kb DNA fragment. Comparison of this sequence with GenBank sequences indicated that a deletion had occurred, resulting in the NS1 primer sequence lying in the usual position of the NS3 primer (White et al., 1990) (Fig. 2). Because of its shorter length, this fragment was amplified preferentially over full-length SSU rDNA fragments. To obtain full-length SSU rDNA sequence the PCR primers MNS1 and MNS4 were constructed, just downstream of the usual NS1 and NS3 positions, respectively (Fig. 2). After the removal of 42 nucleotides at the 5′ end of the NS1–MarGR1 product, the MNS1–MNS4 fragment overlapped the NS1–MarGR1 fragment to yield a near full-length SSU rDNA sequence. The resulting contiguous SSU rDNA sequence was alignable to GenBank Glomus sequences without the need for gaps.

The MNS1–ITS2 and NS1–MNS4 products were amplified from a Hari-Hari DNA sample by nested PCR from a primary SSU + ITS PCR reaction (Fig. 2). After direct sequencing, these fragments overlapped by 997 bp without observable differences, yielding a near full-length Glomus SSU rDNA sequence. The DNA sequence from the ITS spacer region was obtained from four Hari-Hari GLOM1310–ITS26 clones because direct sequencing reads from the MNS1–ITS2 fragment degenerated in the ITS1 spacer sequence.

Semi-specific Glomeromycota detection  To search for the presence of other Glomeromycota species in the Ross I and Hari-Hari samples, combinations of semispecific nested PCR, similar to those outlined in Russell et al. (2002) (Fig. 2), were employed. NS1–GLOM1375R, GLOM1310–ITS26 and MNS1–MNS4 DNA fragments were amplified from a primary SSU + ITS amplification. However, when sequenced, all of these amplicons matched the Glomus sequences previously obtained from the Ross I and Hari-Hari samples. We were not successful in amplifying NS1–ARCH1375R or ARCH1311–ITS26 fragments from the primary SSU + ITS amplification. NS1–NS2 fragments were also amplified by nested PCR from NS1–GLOM1375R or NS1–ARCH1375R first-round amplifications. These amplicons matched either the Glomus sequences already obtained from these samples or liverwort rDNA sequence (EMBL AJ698941).

SSU + ITS sequences  New liverwort-endophyte DNA extracts were prepared from thalli trimmed closely to the fungus-containing rib, leaving little exterior tissue. A PCR from trimmed samples produced more discrete ITS and NS7–ITS26 bands on agarose gels than those amplified from the earlier Ross I and Hari-Hari samples. Two randomly chosen ITS and two NS7–ITS26 clones sequenced from a Kinloch extract were all from Glomus sp. SSU + ITS amplifications from trimmed Ross II, Kinloch and Otira valley samples each provided two DNA bands of close to 2 kb, with the yield of the upper band substantially greater than that of the lower band in each case. Both bands were independently gel-purified and cloned. Partial sequence of 12 clones (four from each sample) showed the larger band was from the liverwort host (EMBL AJ698941). Initial sequencing of 12 smaller-band clones (four from each sample) showed 10 to be Glomus sequences, while two sequences from the Otira valley sample were from a ciliate organism (EMBL AJ810903). Complete Glomus SSU + ITS sequences were obtained from two Kinloch clones, two Ross II clones and three Otira valley clones (EMBL AJ699060–70).

At this point in the study we noted that obtaining sequence by cloning of complete SSU + ITS fragments was comparatively tedious and that the GLOM1310–ITS26 (Fig. 2) primer pair amplified DNA fragments from all DNA samples. Consequently, we gathered new ribosomal DNA sequence information from the 3′-end SSU rDNA plus ITS spacer fragments by amplifying and sequencing GLOM1310–ITS26 DNA fragments. Three to four sequences were obtained from each of the Ross II, Karamea, Kinloch, Reynolds Valley and Montgomery Road samples (EMBL AJ716306–AJ716328).

Phylogenetic analysis

The genus Glomus has previously appeared to fall into two major groups: group A (e.g. G. mosseae) and group B (e.g. G. claroideum). The A group in turn could be divided into three subclades: Aa (G. mosseae group), Ab (G. manihotis/G. intraradices group) and Ac (Glomus sp. W3347 alone) (Schüßler et al., 2001; Schwarzott et al., 2001). When Glomeromycota SSU rDNA sequences from M. foliacea were aligned with a dataset from Schüßler et al. (2001), phylogenetic analyses produced similar tree topologies to those from Schwarzott et al. (2001) and Schüßler et al. (2001) (Fig. 5). However, the discrete nature of Glomus group Ab was reduced by the position of the Otira valley and Ross I sequences, and sequences from Bidartondo et al. (2002). These lay at the base of Glomus group A with little bootstrap support for their placement in either Glomus group Aa or Ab. The Hari-Hari and Kinloch sequences were closely related with some support (ME bootstrap of 85%, MP bootstrap < 50%) for monophyly of these sequences with the Ross II sequences. Together, these sequences formed a clade within Glomus group Ab (Fig. 5).

Analyses of a 400 bp dataset from the 3′ end of the SSU rDNA yielded phylogenetic topologies largely consistent with those obtained from the entire SSU rDNA dataset (Fig. 6). Sequences from two Banks Peninsula samples (Reynolds and Montgomery) formed a distinctive, well-supported, deep-branching cluster with the Otira valley sequences (hereafter referred to as the Otira valley phylotype). The Kinloch, Karamea and Hari-Hari sequences, plus two sequences from New Zealand podocarp nodules (Russell et al., 2002) formed another clade within Glomus group Ab (hereafter referred to as the Hari-Hari phylotype).

image

Figure 6. Phylogenetic tree from 400 bp small subunit (SSU) rDNA dataset. The first tree of 9990 maximum parsimony (MP) trees is shown. Minimum evolution (below) and maximum parsimony (above) bootstrap values of > 50% are shown at nodes. Sequences from Marchantia foliacea are shown in bold type. Sequences with asterisks are those included in the full-length SSU rDNA dataset (see Fig. 5). A measure of distance is indicated by the bar at the base of the diagram.

Download figure to PowerPoint

The Ross II sequences occupied a more basal position than in the SSU rDNA phylogeny, separate from the Hari-Hari clade. Some differences in the positions of the deep-branching phylotypes were evident; for example, the Otira valley phylotype appeared more closely related to the G. mosseae/Glomus group Aa clade than it did in the near full-length phylogeny. In all cases, the sequences obtained from each DNA extract grouped together in the same clade in the 400 bp phylogeny.

Specific amplification of New Zealand Glomus

Sequence alignment and phylogenetic analyses showed that SSU rDNA sequences from the M. foliacea Glomus phylotypes were each contiguous with divergent ITS sequences. Because of this variation, distinct motifs were evident in the spacer sequences of each phylotype. To specifically detect phylotypes, PCR primers were designed in the motifs – two PCR primers for the Otira valley phylotype and one for the Hari-Hari phylotype. These primers consistently amplified products from DNA extracts known to contain their target Glomus phylotype but not from extracts containing other phylotypes. For example, the primers for the Otira valley phylotype amplified DNA from the Otira valley, Montgomery Road and Reynolds DNA extracts, but not from other sample extracts. By contrast with the shared-motif primers, a fourth primer (OV3R1) was designed within a particularly highly divergent part of the Otira valley c2 ITS sequence, away from motifs shared with other sequences. This primer yielded only weak amplification from the Otira valley DNA extract and no amplification from any other DNA extracts.

The phylotype-specific primers were applied to DNA extracts prepared from new liverwort samples from Ross (Ross III) and Port Levy. The Otira valley Glomus phylotype was present in four out of five DNA extracts from the Ross III sample, and the Hari-Hari phylotype in the other extract. The Hari-Hari phylotype was detected in one of five extracts from the Port Levy liverwort patch. From two extracts (one each from Port Levy and Ross III) we amplified and directly sequenced near full-length SSU rDNA sequences by pairing the phylotype-specific PCR primers with the NS1 primer. These resulting SSU rDNA sequences were included in the overall SSU rDNA phylogenetic dataset. Both sequences grouped in their expected Otira valley and Hari-Hari clades with high bootstrap support (Fig. 5).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Marchantia foliacea forms AM

The fungi infecting M. foliacea have many of the morphological features of AM endophytes. The aseptate hyphae of the M. foliacea endophytes indicated that these were not basidiomycete or ascomycete fungi. The course of the infection in the liverwort had similarities to AM infections in plant roots as the growth of the fungus was extremely limited in the outer tissues and proliferated mainly in the inner parenchyma of the liverwort thallus. In M. foliacea the infection conforms closely to the original description of the Paris type of AM (Gallaud, 1905) as the hyphae are entirely intracellular. Arbuscules formed by the endophytes were also not of the classical branching form seen in linear Arum-type infections, but rather a mass of very fine dichotomously branching hyphae forming a ‘fuzz’ around the main trunk hyphae. The Paris-type is the most common AM growth form in bryophytes, ferns and gymnosperms which mainly lack intercellular spaces in parenchyma tissue (Smith & Smith, 1997).

Some aspects of M. foliacea symbioses, such as infection through the smooth rhizoids, were similar to those in the related hepatic C. conicum (Ligrone & Lopes, 1989). However, there were differences in the cytology of the infection, in particular the formation of two distinct infection zones. Notably, the upper arbuscular zone lies directly beneath the photosynthetic epidermal cells (Fig. 3b) and is therefore close to the source of organic carbon. The invading endophyte possibly establishes and maintains infection by means of longitudinally growing hyphae in the lower zone, while active metabolic exchange occurs in the upper zone.

Identity of M. foliacea endophytes

A single Glomus sequence was identified in the first batch of ITS clones. These clones were generated from untrimmed liverwort samples with PCR primers able to amplify sequences from a broad range of eukaryotes. The remaining clones were from ascomycetes or a basidiomycete and thus were not candidate endophyte sequences. Trimming the outer liverwort thallus from the endophyte-containing rib of subsequent samples resulted in amplification of Glomus sequences with few contaminating sequences. By contrast, we investigated more than 20 fungal ITS clones from several thalli of an endophyte-containing leafy liverwort growing at the Ross site without detecting a single Glomus sequence (J. Russell & S. Bulman, unpublished). In our view, this indicated that the Glomus sequences were from the M. foliacea endophyte, and not from fungi growing on the surface of the liverworts, a major concern in studies of this kind.

Sequencing of cloned DNA products revealed extensive variation in the rDNA sequences of the M. foliacea endophytes. Although contaminating fungi have contributed to some records of AM sequence polymorphism (for example, shown by Pringle et al., 2003; Schüßler et al., 2003), accumulated studies have confirmed high levels of ITS variation in Glomeromycota species/phylotypes (Jansa et al., 2002; but see Bidartondo et al., 2002, for an example of low ITS variation). This polymorphism has recently been attributed to homokaryosis and a relaxation of concerted evolution (Pawlowska & Taylor, 2004). Thus, the pattern of DNA variation seen in the rDNA sequences of the M. foliacea endophytes was consistent with previous studies of glomeromycotan fungi.

Marchantia foliacea endosymbionts were from Glomus group A

All near full-length SSU sequences from the M. foliacea endophytes fell into the Glomus group A clade when placed into a phylogenetic tree with other known AM fungi (Schüßler et al., 2001; Schwarzott et al., 2001). Liverworts may be the closest living relatives of the first plants to colonize land (Nickrent et al., 2000; Qiu & Lee, 2000; Kenrick, 2003), with some of the earliest available plant fossils showing liverwort-like attributes (Edwards et al., 1995; Wellman et al., 2003). However, phylogenetic analysis revealed no sequences from basal Glomeromycota groups such as the Archaeosporaceae or Paraglomeraceae (Morton & Redecker, 2001). The liverwort Glomus sequences all either branched from a deep position in Glomus group A (the Otira valley and Ross I phylotypes), or fell more solidly within Glomus group Ab (the Hari-Hari phylotype).

We think it unlikely that only certain Glomeromycota phylotypes were detected because of the PCR primers used. The ITS5, ITS26 and NS1 primers have amplified DNA from a broad range of eukaryotes in our laboratory (Bulman et al., 2001) and GenBank comparisons suggest that the MNS4 primer is also broadly generic for eukaryotes. The MNS1 primer does not match Archaeospora or Paraglomus sequences but does have similarity to a broad group of Glomeromycota sequences from outside Glomus group A (e.g. Scutellospora and Gigaspora sequences, etc.). The sequences we obtained using these general primers were of the same phylotypes as those obtained with the taxon-specific primers. If other species of AM fungi were present in our samples, we believe we would have detected them.

Specificity of the M. foliacea–AM association

Phylogenetic analyses of the 400 bp dataset provided evidence for a degree of specificity in the association of Glomus spp. with M. foliacea. Several sequences from M. foliacea from new sites grouped robustly with those phylotypes already identified. Sequences from Karamea and Banks Peninsula were identical to, or nearly identical to, those from Otira valley and Hari-Hari. We also demonstrated the repeated occurrence of the Otira valley and Hari-Hari phylotypes in M. foliacea using PCR primers designed in the Glomus ITS spacer sequences. Phylogenetic placement of two near full-length sequences amplified with the specific primers confirmed that the resulting products were from the expected phylotypes. Thus, we can predict that particular Glomus strains are likely to be found in M. foliacea and have tools to specifically detect these strains.

The repeated detection of the same Glomus phylotypes in M. foliacea may be consistent with other molecular studies of Glomeromycota diversity in the wild that have begun to define some structure in AM–plant relationships (Helgason et al., 2002; Vandenkoornhuyse et al., 2002). Notably, Bidartondo et al. (2002) found a restricted set of closely related AM fungi were associated with epiparasitic plants, especially Arachnitis uniflora. Ecological studies showing functional differences between AM phylotypes when paired with particular plant species may also provide evidence of more specificity in these relationships than was previously thought (van der Heijden et al., 1998; Helgason et al., 2002).

An alternative explanation for our findings is that these particular Glomus group A species were the only AM fungi at these eight sites and M. foliacea is colonized by the predominant AM fungi in the environment. In surveys of Glomeromycota spores, both Hall (1977) and Johnson (1977) identified Gigaspora and Acaulaspora spp. as well as a range of Glomus spp. in New Zealand. Ribosomal DNA sequences from Archaespora species have also been detected in the New Zealand forest (Russell et al., 2002) as have sequences from a broader range of Glomus group A phylotypes in podocarp nodules (J. Russell, unpublished). Since studies in temperate woodlands (Helgason et al., 1998) and tropical forests (Husband et al., 2002) have found a wide diversity of glomeromycotan fungi, it would be surprising if New Zealand forests, which are equally botanically diverse, should contain a narrow range of AM fungi. Nevertheless, our observation of the Otira valley and Hari-Hari clades in both podocarp roots and M. foliacea did imply the presence of a distinctive, perhaps predominant, AM microflora in these forests.

Overall, we detected endophyte rDNA sequences from 10 M. foliacea samples using a variety of molecular approaches. We repeatedly detected two Glomus phylotypes from liverwort samples that were collected from widely separated sites featuring different surrounding forest environments. We believe these findings are sufficient to conclude that M. foliacea displays at least a preference for certain AM fungi, but note that they may simply reflect the prevalent Glomeromycota found at each site. To reach a definitive conclusion on AM–M. foliacea specificity, a greater number of endophyte sequences from further M. foliacea samples will be required, as well as identification of Glomeromycota phylotypes on roots of plants at the M. foliacea sites.

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Whether liverwort–fungi symbioses function in the same manner as mycorrhizae has been a matter of debate (e.g. Read et al., 2000). Not all liverworts or even all species of Marchantiales are colonized by fungi, as shown by the two samples of the cosmopolitan M. polymorpha examined here. Nevertheless, all of the M. foliacea samples we examined contained Glomeromycota endophytes. The colonization of podocarp roots by two of the Glomus phylotypes (Russell et al., 2002; J. Russell, unpublished) strongly suggests that the liverwort endophyte also forms mycorrhizal associations with vascular plants in the New Zealand forest, and that the podocarp trees and the liverwort may share a common symbiont. Whether the liverwort–AMF association is representative of an ancestral symbiosis is open to debate. The phylogenetic placement of the liverwort endophytes in Glomus group A is likely indicative of a secondary acquisition of the association in M. foliacea, and lends weight to the idea that liverworts have repeatedly regained and lost symbioses with fungi. Molecular tools such as those developed in this study will be useful for further studying the nature of the liverwort-AM fungi relationship and for examining the linkages (Duckett & Read, 1995; Turnau et al., 1999), distribution and ecology of this important group of fungi.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

We are indebted to David Glenny for advice on liverwort identification and collection. We thank Ian Hall for advice on New Zealand AMs, Cath Brown for discussions with Ngai Tahu, the Department of Conservation for permission to collect samples and Arthur Schüßler for comments on the manuscript. This research was funded in part by the New Zealand Foundation for Research Science and Technology. AJR was awarded a Fellowship by the University of Canterbury.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990. Basic local alignment search tool. Journal of Molecular Biology 215: 403410.
  • Bernatzky R, Tanksley SD. 1986. Genetics of actin-related sequences in tomato. Theoretical and Applied Genetics 72: 314339.
  • Bidartondo MI, Bruns TD, Weiß M, Sergio C, Read DJ. 2003. Specialised cheating of the ectomycorrhizal symbiosis by an epiparasitic liverwort. Proceedings of the Royal Society of London Series B 270: 835842.
  • Bidartondo MI, Redecker D, Hijri I, Wiemken A, Bruns TD, Dominguez L, Sersic A, Leake JR, Read DJ. 2002. Epiparasitic plants specialised on arbuscular mycorrhizal fungi. Nature 419: 389392.
  • Bischler-Causse H. 1989. Marchantia L. The Asiatic and Oceanic taxa. Bryophytorum Bibliotheca 38: 1318.
  • Boullard B. 1988. Observations on the coevolution of fungi with hepatics. In: PyrozynskiKA, HawksworthDL, eds. Coevolution of fungi with plants and animals. London, UK: Academic Press, 107124.
  • Bulman SR, Marshall JW. 1998. Detection of Spongospora subterranea in potato tuber lesions using the polymerase chain reaction (PCR). Plant Pathology 47: 759766.
  • Bulman SR, Kühn SF, Marshall JW, Schnepf E. 2001. A phylogenetic analysis of the SSU rRNA from members of the Plasmodiophorida and Phagomyxida. Protist 152: 4351.
  • Carafa A, Duckett JG, Ligrone R. 2003. Subterranean gametophytic axes in the primitive liverwort Haplomitrium harbour a unique type of endophytic association with aseptate fungi. New Phytologist 160: 185197.
  • Cooper KM. 1976. A field survey of mycorrhizas in New Zealand ferns. New Zealand Journal of Botany 14: 169181.
  • Duckett JG, Read DJ. 1995. Ericoid mycorrhizas and rhizoid–ascomycete associations in liverworts share the same mycobiont: isolation of the partners and resynthesis of the association in vitro. New Phytologist 129: 439447.
  • Edwards D, Duckett JG, Richardson JB. 1995. Hepatic characters in the earliest land plants. Nature 374: 635636.
  • Felsenstein J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783791.
  • Gallaud I. 1905. Études sur les mycorrhizes endotrophes. Revue Générale de Botanique 17: 748, 66–85, 123–136, 223–239, 313–325, 423–433, 479–496.
  • Hall I. 1977. Species and mycorrhizal infections of New Zealand Endogonaceae. Transactions of the British Mycological Society 68: 341356.
  • Van Der Heijden MGA, Klironomos JN, Ursic M, Moutoglis P, Streitwolf-Engel R, Boller T, Wiemken A, Sanders IR. 1998. Mycorrhizal fungal diversity determines plant biodiversity, ecosystem variability and productivity. Nature 396: 6972.
  • Helgason T, Daniell TJ, Husband R, Fitter AH, Young JPW. 1998. Ploughing up the wood-wide web? Nature 394: 431.
  • Helgason T, Merryweather JW, Denison J, Wilson P, Young JPW, Fitter AH. 2002. Selectivity and functional diversity in arbuscular mycorrhizas of co-occuring fungi and plants from a temperate deciduous woodland. Journal of Ecology 90: 371384.
  • Howlett BJ, Brownlee AG, Guest DI, Adcock GJ, McFadden GI. 1992. The 5S ribosomal RNA gene is linked to large and small subunit ribosomal RNA genes in the oomycetes, Phytophthora vignae, P. cinnamomi, P. megasperma f. sp. glycinae and Saprolegnia ferax. Current Genetics 22: 455461.
  • Husband R, Herre EA, Turner SL, Gallery R, Young JPW. 2002. Molecular diversity of arbuscular mycorrhizal fungi and patterns of host association over time and space in a tropical forest. Molecular Ecology 11: 26692678.
  • Jansa J, Mozafar A, Banke S, McDonald BA, Frossard E. 2002. Intra- and intersporal diversity of ITS rDNA sequences in Glomus intraradices assessed by cloning and sequencing, and by SSCP analysis. Mycological Research 6: 670681.
  • Johnson PN. 1977. Mycorrhizal Endogonaceae in a New Zealand forest. New Phytologist 78: 161170.
  • Kenrick P. 2003. Fishing for the first plants. Nature 425: 248249.
  • Kimura M. 1980. A simple model for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16: 111120.
  • Ligrone R. 1988. Ultrastructure of a fungal endophyte in Phaeoceros laevis (L. Prosc. (Anthocerophyta)). Botanical Gazette 149: 92100.
  • Ligrone R, Lopes C. 1989. Cytology and development of a mycorrhiza-like infection in the gametophyte of Conocephalum conicum (L.) Dum. (Marchantiales, Hepatophyta). New Phytologist 111: 423433.
  • Morton JB, Benny GL. 1990. Revised classification of arbuscular mycorrhizal fungi (Zygomycetes): a new order, Glomales, two new suborders, Glomineae and Gigasporineae, and two new families, Acaulosporaceae and Gigasporaceae, with emendation of Glomaceae. Mycotaxon 37: 471491.
  • Morton JB, Redecker D. 2001. Two new families of Glomales, Archaeosporaceae and Paraglomaceae, with two new genera Archaeospora and Paraglomus, based on concordant molecular and morphological characters. Mycologia 93: 181195.
  • Nickrent DL, Parkinson CL, Palmer JD, Duff RJ. 2000. Multigene phylogeny of land plants with special reference to bryophytes and the earliest land plants. Molecular Biology and Evolution 17: 18851895.
  • Pawlowska TE, Taylor JW. 2004. Organisation of genetic variation in individuals of arbuscular mycorrhizal fungi. Nature 427: 733737.
  • Pirozynski KA, Malloch DW. 1975. The origin of land plants: a matter of mycotrophism. Biosystems 6: 153164.
  • Pringle A, Moncalvo J-M, Vilgalys R. 2003. Revisiting the rDNA sequence diversity of a natural population of the arbuscular mycorrhizal fungus Acaulospora colossica. Mycorrhiza 13: 227231.
  • Qiu YL, Lee J. 2000. Transition to a land flora: a molecular phylogenetic perspective. Journal of Phycology 36: 799802.
  • Read DJ, Duckett JG, Francis R, Ligrone R, Russell A. 2000. Symbiotic fungal associations in ‘lower’ land plants. Philosophical Transactions of the Royal Society of London Series B 355: 815831.
  • Redecker D, Kodner R, Graham LE. 2000. Glomalean fungi from the Ordovician. Science 289: 19201921.
  • Remy W, Taylor TN, Hass H, Kerp H. 1994. Four hundred-million-year-old vesicular arbuscular mycorrhizae. Proceedings of the National Academy of Sciences of the USA. 91: 1184111843.
  • Russell AJ, Bidartondo MI, Butterfield BG. 2002. The root nodules of the Podocarpaceae harbour arbuscular mycorrhizal fungi. New Phytologist 156: 283295.
  • Saitou N, Nei M. 1987. The neighbour-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4: 406425.
  • Schüßler A. 2000. Glomus clarioideum forms an arbuscular mycorrhiza-like symbiosis with the hornwort Anthoceros punctatus. Mycorrhiza 10: 1521.
  • Schüßler A. 2002. Molecular phylogeny, taxonomy, and evolution of Geosiphon pyriformis and arbuscular mycorrhizal fungi. Plant and Soil 244: 7583.
  • Schüßler A, Schwarzott D, Walker C. 2001. A new fungal phylum, the Glomeromycota: phylogeny and evolution. Mycological Research 105: 14131421.
  • Schüßler A, Schwarzott D, Walker C. 2003. Glomeromycota rRNA genes – the diversity of myths? Mycorrhiza 13: 233236.
  • Schwarzott D, Walker C, Schüßler A. 2001. Glomus, the largest genus of the arbuscular mycorrhizal fungi (Glomales) is nonmonophyletic. Molecular Phylogenetics and Evolution 21: 190197.
  • Simon L, Bousquet J, Lévesque RC, Lalonde M. 1993. Origin and diversification of endomycorrhizal fungi and coincidence with vascular land plants. Nature 363: 6769.
  • Smith FA, Smith SE. 1997. Structural diversity in (vesicular) -arbuscular mycorrhizal symbioses. New Phytologist 137: 373388.
  • Swofford DL. 2002. paup*. phylogenetic analysis using parsimony. (*and other methods), version 4. Sunderland,. MA, USA: Sinauer Associates.
  • Taberlet P, Geilly L, Pautou G, Bouvet J. 1991. Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology Reporter 17: 11051109.
  • Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG. 1997. The clustalx windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 24: 48764882.
  • Turnau K, Ronikier M, Unrug J. 1999. Role of mycorrhizal links between plants in establishment of liverwort thalli in natural habitats. Acta Societatia Botanicorum Poloniae 68: 6368.
  • Vandenkoornhuyse P, Husband R, Daniell TJ, Watson IJ, Duck JM, Fitter AH, Young JP. 2002. Arbuscular mycorrhizal community composition associated with two plant species in a grassland ecosystem. Molecular Ecology 11: 15551564.
  • Wellman CH, Osterloff PL, Mohiuddin U. 2003. Fragments of the earliest land plants. Nature 425: 282285.
  • White TJ, Bruns T, Lee S, Taylor J. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: InnesMA, GelfandDH, SninskyJJ, WhiteTJ, eds. PCR protocols, a guide to methods and applications. San Diego, CA, USA: Academic Press, 315322.