Lyso-Phosphatidylcholine Is a Signal in the Arbuscular Mycorrhizal Symbiosis

See allHide authors and affiliations

Science  12 Oct 2007:
Vol. 318, Issue 5848, pp. 265-268
DOI: 10.1126/science.1146487


The arbuscular mycorrhizal (AM) symbiosis represents the most widely distributed mutualistic root symbiosis. We report that root extracts of mycorrhizal plants contain a lipophilic signal capable of inducing the phosphate transporter genes StPT3 and StPT4 of potato (Solanum tuberosum L.), genes that are specifically induced in roots colonized by AM fungi. The same signal caused rapid extracellular alkalinization in suspension-cultured tomato (Solanum lycopersicum L.) cells and induction of the mycorrhiza-specific phosphate transporter gene LePT4 in these cells. The active principle was characterized as the lysolipid lyso-phosphatidylcholine (LPC) via a combination of gene expression studies, alkalinization assays in cell cultures, and chromatographic and mass spectrometric analyses. Our results highlight the importance of lysophospholipids as signals in plants and in particular in the AM symbiosis.

The AM symbiosis is thought to have facilitated plant colonization of land more than 400 million years ago, and today 80% of terrestrial plant species are colonized by AM fungi. In the AM symbiosis, fungal hyphae form coiled or ramified structures in root cortical cells of the host plant. Despite intracellular accommodation of the microsymbiont, the symbiotic partners remain separated by their plasma membranes, which thus demarcate the symbiotic interface, that is, the site of bidirectional exchange of compounds including signals and nutrients. Phosphorus (P) is taken up by plants as orthophosphate (Pi) via two pathways: the direct uptake pathway, at the level of the root-soil interface including root hair cells, as opposed to the mycorrhizal uptake pathway, extending from extraradical fungal hyphae to the cortical cells harboring fungal symbiotic structures. A detailed analysis of the contribution of the mycorrhizal Pi uptake pathway indicated that this pathway can dominate Pi supply to plants irrespective of whether plants forming AM symbioses exhibited improved growth and/or total P uptake (1). Mycorrhiza-inducible phosphate transporters have been found in several plant species and are likely to play a prominent role in growth and development of >200,000 plant species forming AM symbioses (24). A 5′ upstream untranslated region of 1.7 kb (herein referred to as StPT3 promoter) of the gene encoding the mycorrhiza-inducible high-affinity Pi transporter StPT3 from potato (5) was recently shown to be activated in different plant species exclusively by fungi of the phylum Glomeromycota (6). This is consistent with a model in which the signaling pathway leading to the induction of mycorrhizal Pi transporters in the plant is evolutionarily conserved. To identify such signal compounds involved in StPT3 gene regulation, we analyzed extracts from mycorrhized roots for their potential to trigger StPT3 promoter activation in a bioassay based on transgenic potato roots carrying the StPT3 promoter-β-glucuronidase (GUS) reporter gene construct (5).

Infiltration of phospholipid (PL) extracts from mycorrhized roots (PL+myc) of plantain (Plantago lanceolata L.) caused StPT3 promoter activation, as evident from positive GUS staining of transgenic roots; corresponding extracts from nonmycorrhized roots (PL–myc) failed to activate the StPT3 promoter (Fig. 1A). GUS staining was absent on infiltration of neutral lipid fractions (NL+myc and NL–myc). Induction of GUS in transgenic potato roots was observed with PL+myc extracts from different plant species, including plantain, potato, and maize (Zea mays L.), but did not develop when PLs from nonmycorrhized roots of these species were applied (table S1). Moreover, PL extracts from AM fungal material, such as dormant spores, germinated mycelium, presymbiotic mycelium, or extraradical hyphae, were inactive (table S1). GUS staining was detected in roots treated with bioactive PL preparations, predominantly in the zone behind the root tip (Fig. 1A), which is the zone of AM fungal colonization. To verify induction of mycorrhiza-inducible phosphate transporter genes, we performed reverse transcription polymerase chain reaction (RT-PCR) with RNA from potato roots treated with PL–myc or PL+myc from plantain roots. Transcripts encoding the two inducible phosphate transporters StPT3 and non-orthologous StPT4 accumulated specifically after infiltration of PL+myc, whereas the level of constitutively expressed and AM-nonresponsive StPT1 (6) transcripts remained constant (Fig. 1B). Overall, these results indicate the presence of signals in PL+myc extracts that can specifically activate mycorrhiza-specific transporter gene expression.

Fig. 1.

Induction of phosphate transporter gene in roots treated with phospholipid extracts from mycorrhized roots. (A) Neutral lipids (NLs) and phospholipids (PLs) from mycorrhized (+myc) and nonmycorrhized (–myc) roots of plantain were applied to potato plants containing a chimeric gene consisting of a GUS reporter gene under the control of the StPT3 promoter. Roots were stained for GUS after 3 hours of treatment with the lipids. Scale bars indicate 100 μm. (B) RT-PCR performed on RNA from potato roots after 3-hour treatment with PL–myc or PL+myc as indicated. RNA from untreated roots (N.T.) or roots treated with solvent (Solv.) were used as controls. Gene-specific primers for StPT1, StPT3, and StPT4 were used. Non–reverse-transcribed RNA samples (–RT) did not give rise to amplification of cDNA, indicating absence of genomic DNA in the PCR templates. Shown is the inverted picture.

In previous work, undifferentiated suspension-cultured plant cells proved useful to identify and characterize signals derived from both plant pathogens and symbionts (7, 8). The cells respond to these signals with rapid, characteristic changes in ion fluxes across their plasma membrane, which result in a change in the extracellular pH. We used suspension-cultured tomato cells to study their response to root lipid (L) and PL extracts. Rapid extracellular alkalinization was observed upon treatment of cells with L+myc and PL+myc extracts, whereas equivalent extracts from nonmycorrhized roots failed to evoke a rise in extracellular pH (fig. S1 and Fig. 2A). Alkalinization was dose-dependent and saturable (fig. S2) and resembled the response of these cells to microbial elicitors, as exemplified for the bacterial pathogen-associated molecular pattern (PAMP) flg22 (fig. S1 and Fig. 2A). However, the L+myc and PL+myc preparations did not induce Ca influx in these cells (fig. S3A), nor did the PL+myc and PL–myc preparations induce production of reactive oxygen species (ROS) in plant leaves (fig. S3B), two other responses typically observed after treatment with known elicitors like flg22. Instead, tomato cells responded with the accumulation of transcripts encoding LePT4 when treated with PL+myc but not with PL–myc (Fig. 2B). LePT4 is a phosphate transporter from tomato that is normally induced by AM fungal colonization of cortex cells and is orthologous to StPT4 (9). This induction of a normally mycorrhiza-specific gene prompted us to further use medium alkalinization in cultured tomato cells as a rapid and reliable assay toward the identification of the “signal” present in the PL+myc extract. By using various separation techniques coupled to mass spectrometry, we identified lyso-phosphatidylcholine (LPC) as a candidate bioactive compound. To prove the bioactivity of LPC directly, we tested commercially available phosphatidylcholine from soybean (PCsb). This preparation did not provoke any response in the cell culture (Fig. 2A). However, when PCsb was hydrolyzed with phospholipase A2 (PLA2) to release LPC, the preparation induced a strong alkalinization response (Fig. 2A), indicating that LPC might be the active signal. Indeed, commercially available LPC from egg yolk (LPCey) and soybean (LPCsb) induced alkalinization in a saturable and dose-dependent way (Fig. 2A and fig. S2B) but did not induce other responses typically observed after flg22 treatment, including production of ROS and ethylene in plant leaves (fig. S3, B and C).

Fig. 2.

Effects of lipid extracts on suspension-cultured tomato cells. (A) Shift in extracellular pH after 10 min of treatment or at the pH maximum reached within ∼35 min of treatment with PL extracts from mycorrhized (PL+myc) or non-mycorrhized (PL–myc) roots, defined PLs or flg22 as indicated. PL extracts were applied at 15 μl/ml of suspension; defined PLs, at 100 μM; and flg22, at 1 μM. Results show means and standard deviations of three replicates. Controls label indicates treatment with solvent (methanol:water, 1:9); PCsbA2, PCsb treated with PLA2 (from bovine brain). (B) RT-PCR performed on RNA from tomato cell culture after application of PL–myc and PL+myc from plantain roots. Gene-specific primer pairs for LePT1 and LePT4 were used. Non–reverse-transcribed RNA samples (–RT) were used to test absence of genomic DNA in the RNA samples. Shown is the inverted picture.

Based on these findings, PLs from mycorrhized and nonmycorrhized roots were analyzed by matrix-assisted laser desorption ionization/time of flight mass spectrometry (MALDI-TOF-MS). This revealed the presence of distinct signals that specifically occurred in the PL+myc fractions. Molecular ions at m/z = 494.3, 496.3, 518.3, 520.3, 522.3, and 524.3, indicative of LPC containing C16:1, C16:0, C18:3, C18:2, C18:1, and C18:0 fatty acids at sn-1 position, were detected (Fig. 3, A and B). Moreover, after treatment of PC with PLA2, releasing the fatty acid at position sn-2 of PC (10), molecular ions with the same masses were detected (Fig. 3C). This pattern of molecular ions was also observed in LPCsb (Fig. 3D). Nanoflow reverse-phase liquid chromatography (LC) separation and quantification of LPC revealed that the relative amounts of monounsaturated LPC species 16:1 and 18:1 in PL+myc extracts were higher than those present in LPCsb (fig. S4). LPC concentration in PL+myc was estimated to be 4 μM, whereas LPC amounts in PL–myc accounted for 0.5% of the amount in PL+myc. Thus, our analyses indicated that LPC might be the bioactive ingredient in PL+myc to induce mycorrhiza-specific phosphate transporter gene expression. Next, externally applied LPC was shown to be the signal inducing phosphate transporter genes in nonmycorrhized roots directly. As compared with solvent controls, StPT3 and StPT4 transcripts accumulated in potato roots that had been infiltrated with LPC (Fig. 4A). In contrast, transcript concentrations of the AM-nonresponsive StPT1 gene (9) remained constant (Fig. 4A). We then looked at the GUS response in transgenic potato roots carrying the StPT3 promoter–GUS chimeric gene. Similar to the activity seen in roots infiltrated with PL+myc (Fig. 1A), GUS activity in roots was stimulated by LPCsb and synthetic LPC with the fatty acid C18:1 (Fig. 4, B and C). In addition, no increase in GUS staining was observed when transgenic roots were treated with lyso-phosphatidylethanolamine (LPE) (Fig. 4C). This result suggested selectivity for the phosphocholine headgroup at sn-3 position.

Fig. 3.

Positive-ion MALDI-TOF mass spectra for LPC in PL fraction originating from mycorrhized (A) and nonmycorrhized (B) roots after chromatography. MALDI spectrum of PCsbA2 (C) and of commercially available LPC (D). Numbers indicate the peak position (m/z) of the protonated form of different LPC molecular species. Species 1 is LPC 16:1; 2, LPC 16:0; 3, LPC 18:3; 4, LPC 18:2; 5, LPC 18:1; and 6, LPC 18:0.

Fig. 4.

Effect of LPC on phosphate transporters in potato roots. (A) RT-PCR with RNA from potato roots treated for 3 hours with LPCsb or the solvent as control. Gene-specific primers for StPT1, StPT3, and StPT4 were used. Non–reverse-transcribed RNA samples (–RT) did not give rise to amplification of cDNA, indicating absence of genomic DNA in the PCR templates. Shown is the inverted picture. (B) StPT3 promoter–driven expression of GUS reporter gene in transgenic potato roots after treatment with a solvent control (methanol:water, 1:9), LPCsb, or synthetic LPC with defined fatty acid 18:1, as indicated. Scale bars, 100 μm. (C) Percentage of root tips with GUS staining after 3 hours of treatment with a solvent control, LPCsb, LPC 18:1, and lyso-phosphatidylethanolamine (LPE). Compounds were applied at 100 μM. Bars show means of three independent experiments with each n > 20 root tips, and error bars indicate standard deviations of the three replicates.

Apparently, the signaling pathway to activate the mycorrhiza-specific phosphate transporters has its origin in the PL PC, a major component of membranes of plants and, probably, also of the AM fungus. However, PC is not active in itself. It gains activity only after treatment with PLA2 (Fig. 2A and Fig. 3C), suggesting an important role of this enzyme in the establishment of the mycorrhizal Pi uptake pathway. Whether the production of the LPC signal involves PLA2 and PC from plants, fungus, or both remains to be explored further. Several PLA2s have been identified in plants, and all are secretory proteins (11); their regulation and substrate specificities are unknown. This might hint at an extracellular production of the LPC signal. However, in the AM symbiosis, the phosphate transporter genes appear to be activated quite specifically in the arbuscule-containing cells (5, 6). Therefore, the LPC signal might be generated more specifically in the arbuscule-containing cells. LPCs are highly mobile within intact cells, and LPC is therefore a good candidate for a cytoplasmic messenger that transduces signals to activate downstream processes and gene expression in the nucleus.

PLA2-dependent activation of the defense responses involving LPC as a putative signal has been reported in plants (12, 13). Interestingly, comparative transcriptomics in rice revealed a set of genes that was similarly expressed in associations with symbiotic and pathogenic fungi, revealing a conserved response to fungal colonization (14). Together with the findings of this report, this suggests commonalities between the signaling pathways in (i) wounding (15) and pathogen attack and (ii) cellular colonization of cortical cells with AM fungi. We observed similarities but also differences in the responses of tomato cells treated with mycorrhizal LPC and with the bacterial PAMP flg22. Further work is required to study the overlap between signaling pathways in response to AM fungi, to herbivore attack, and to PAMPs in plant innate immunity.

It is well documented that LPC is bioactive in mammals. An increasing amount of evidence has suggested that LPC is involved in the activation of inflammatory responses (16, 17). Perception of LPC can occur through binding to G protein–coupled receptors (18). Furthermore, possible roles of LPC as functional vaccine (17) or second messenger have been discussed, underlining the functional importance of LPC in cellular responses.

In plants, PLA2 and its lysolipid product LPC have been reported to be involved in numerous cellular processes (19), including cytoplasmic acidification, which is accompanied by extracellular medium alkalinization, with the latter leading to alterations in gene expression (20). So far, experimental evidence for a participation of PLs in plant signaling and alteration of gene expression has only been demonstrated for phosphatidic acid, which is produced through the activity of phospholipase D (21). Therefore, further research on LPC generation and signaling can hopefully tell us more about the evolution of response regulation in plants and mammals, including that in the development of the AM symbiosis.

Supporting Online Material

Materials and Methods

Figs. S1 to S5

Table S1

References and Notes

View Abstract

Navigate This Article