1. Ecology
Download icon

Algal-fungal symbiosis leads to photosynthetic mycelium

  1. Zhi-Yan Du
  2. Krzysztof Zienkiewicz
  3. Natalie Vande Pol
  4. Nathaniel E Ostrom
  5. Christoph Benning  Is a corresponding author
  6. Gregory M Bonito  Is a corresponding author
  1. Michigan State University, United States
  2. Georg-August-University, Germany
  3. Nicolaus Copernicus University in Toruń, Poland
Research Article
  • Cited 1
  • Views 3,964
  • Annotations
Cite this article as: eLife 2019;8:e47815 doi: 10.7554/eLife.47815

Abstract

Mutualistic interactions between free-living algae and fungi are widespread in nature and are hypothesized to have facilitated the evolution of land plants and lichens. In all known algal-fungal mutualisms, including lichens, algal cells remain external to fungal cells. Here, we report on an algal–fungal interaction in which Nannochloropsis oceanica algal cells become internalized within the hyphae of the fungus Mortierella elongata. This apparent symbiosis begins with close physical contact and nutrient exchange, including carbon and nitrogen transfer between fungal and algal cells as demonstrated by isotope tracer experiments. This mutualism appears to be stable, as both partners remain physiologically active over months of co-cultivation, leading to the eventual internalization of photosynthetic algal cells, which persist to function, grow and divide within fungal hyphae. Nannochloropsis and Mortierella are biotechnologically important species for lipids and biofuel production, with available genomes and molecular tool kits. Based on the current observations, they provide unique opportunities for studying fungal-algal mutualisms including mechanisms leading to endosymbiosis.

https://doi.org/10.7554/eLife.47815.001

eLife digest

Yeast, molds and other fungi are found in most environments across the world. Many of the fungi that live on land today form relationships called symbioses with other microbes. Some of these relationships, like those formed with green algae, are beneficial and involve the exchange carbon, nitrogen and other important nutrients. Algae first evolved in the sea and it has been suggested that symbioses with fungi may have helped some algae to leave the water and to colonize the land more than 500 million years ago.

A fungus called Mortierella elongata grows as a network of filaments in soils and produces large quantities of oils that have various industrial uses. While the details of Mortierella’s life in the wild are still not certain, the fungus is thought to survive by gaining nutrients from decaying matter and it is not known to form any symbioses with algae.

In 2018, however, a team of researchers reported that, when M. elongata was grown in the laboratory with a marine alga known as Nannochloropsis oceanica, the two organisms appeared to form a symbiosis. Both the alga and fungus produce oil, and when grown together the two organisms produced more oil than when the fungus or algal cells were grown alone. However, it was not clear whether the fungus and alga actually benefit from the symbiosis, for example by exchanging nutrients and helping each other to resist stress.

Du et al. – including many of the researchers involved in the earlier work – have now used biochemical techniques to study this relationship in more detail. The experiments found that there was a net flow of carbon from algal cells to the fungus, and a net flow of nitrogen in the opposite direction. When nutrients were scarce, algae and fungi grown in the same containers grew better than algae and fungi grown separately. Further, Mortierella only obtained carbon from living algae that attached to the fungal filaments and not from dead algae. Unexpectedly, further experiments found that when grown together over a period of several weeks or more some of the algal cells entered and lived within the filaments of the fungus. Previously, no algae had ever been seen to inhabit the living filaments of a fungus.

These findings may help researchers to develop improved methods to produce oil from M. elongata and N. oceanica. Furthermore, this partnership provides a convenient new system to study how one organism can live within another and to understand how symbioses between algae and fungi may have first evolved.

https://doi.org/10.7554/eLife.47815.002

Introduction

Mutualistic symbioses are defined as those in which partners interact physically and metabolically in mutually beneficial ways. Mutualisms underlie many evolutionary and ecological innovations including the acquisition of plastids and mitochondria, and evolution of symbiotic mutualisms such as mycorrhizas, lichens and corals (Little et al., 2004; Service, 2011; Tisserant et al., 2013; Spribille et al., 2016). An understanding of the underlying principles that govern microbial mutualisms informs microbial ecology and efforts to engineer synthetic microbiomes for biotechnological applications (Egede et al., 2016).

Terrestrialization of Earth has been associated with lineages of early diverging fungi belonging to the Mucoromycota. However, recent analyses indicate that fungal colonization of land was associated with multiple origins of green algae prior to the origin of embryophytes (Lutzoni et al., 2018). Research indicating that plants were genetically pre-adapted for symbiosis with fungi, has renewed interest in fungal-algal associations (Delaux et al., 2015; Spatafora et al., 2016).

The most well-known mutualisms that exist between algae and fungi are lichens, which were estimated to radiate 480 million years ago (Lutzoni et al., 2018). Lichen symbiosis is adaptive in that it allows mycobiont and photobiont symbionts to survive in habitats and environments that would otherwise be uninhabitable by either species growing alone, such as on a rock outcrop or in a desert crust. Lichenized fungi have been shown to have multiple independent origins in Ascomycota and Basidiomycota, and are themselves meta-organisms that include communities of chlorophyte algae, cyanobacteria, in addition to basidiomyceteous yeasts (Spribille et al., 2016).

Nutrient exchange often underlies mutualisms between photobionts and mycobionts. For example, reciprocal transfer of carbon and nitrogen was shown for synthetic consortia composed of Chlamydomonas reinhardtii and a diverse panel of ascomycete fungi, demonstrating a latent capacity of ascomycetous yeasts and filamentous fungi to interact with algae (Hom and Murray, 2014). In a separate study, the filamentous ascomycetous fungus Alternaria infectoria was demonstrated to provision nitrogen to C. reinhardtii in a long-lived bipartite system, whereby the nitrogen-starved alga responded favorably to the growing fungus (Simon et al., 2017). A non-lichen algal-fungal mutualism was described involving the chytrid fungus Rhizidium phycophilum and the green alga Bracteacoccus providing evidence that early diverging fungi have evolved mutualisms with algae based on solute exchange (Picard et al., 2013). However, in all known examples of fungal-algal symbioses algal cells remain external to fungal hyphae and are not known to enter living fungal cells.

While studying a synthetic co-culture composed of two biotechnologically important oil-producing organisms, the soil fungus Mortierella elongata and the marine alga Nannochloropsis oceanica, we observed an interaction between fungal and algal cells that led to changes in metabolism of both partners (Du et al., 2018a). This biotrophic interaction showed high specificity and resulted in close physical contact of partners, with the eventual incorporation of functional algal cells within fungal mycelium. Here, we describe this apparent symbiosis in detail. We further demonstrate through isotope tracer experiments that bidirectional nutrient exchange underlies the described algal-fungal interactions.

Results

Interaction between N. Oceanica and M. elongata after short-term co-culture

Nannochloropsis oceanica cells flocculated in dense clusters around M. elongata mycelium when they were incubated together (Figure 1A and B). After 6 days co-cultivation, scanning electron microscopy (SEM) revealed a wall-to-wall fungal-algal interface between the organisms grown in co-culture, (Figure 1C) with the morphology of N. oceanica cells differing from those of cells grown in the absence of fungus. Specifically, SEM images showed that N. oceanica cells incubated alone in f/2 medium have a smooth outer layer (Figure 1D and Figure 1—figure supplement 1A), which was fragmented or lacking after co-culture with M. elongata AG77 and fibrous extensions underneath the smooth outer layer were exposed (Figure 1E and Figure 1—figure supplement 1B). While it is possible that the fibrous extensions have been elicited by the contact with the fungus, remnant pieces of the outer coat covering the underlying extensions are evident in our observations (Figure 1E and F and Figure 1—figure supplement 1B). Therefore, it seems likely that these extensions are present underneath the outer smooth layer. In addition, the fibrous extensions appeared to contribute to anchoring the algae to hyphae and irregular tube-like extensions were formed between the two interacting cell types (Figure 1F). Further SEM revealed that the fibrous extensions only were exposed in N. oceanica cells that were in physical contact with live fungal hyphae. N. oceanica cells maintained a smooth outer wall covering (Figure 1—figure supplement 1) when they were co-cultured without physical contact with live fungi, and when they were co-cultured with M. elongata mycelium that had been killed in a 65°C water bath. We demonstrated that a combination of enzymes, including 4% fungal hemicellulase and 2% driselase, could partially digest the outer smooth cell wall layer of N. oceanica and expose the fibrous extensions to mimic the morphological change observed during the physical interaction between live M. elongata and N. oceanica cells (Figure 1—figure supplement 2).

Figure 1 with 2 supplements see all
Interaction between N.oceanica and M. elongata cells.

(A) Co-cultivation of M. elongata AG77 and N. oceanica (Noc) in flasks for 6 days. Green tissues indicated by the red arrow head are aggregates formed by AG77 mycelium and attached Noc cells. (B) Differential interference contrast micrographs of the green tissues shown in (A). A large number of Noc cells are trapped by AG77 mycelium. (C–F) Scanning electron microscopy images of alga-fungus interaction. (C) Noc cells stick to the fungal mycelium after 6-day co-culture. (D) Noc controls grown in f/2 medium alone have smooth surface. (E) A Noc cell adheres to an AG77 hypha by the outer surface with fibrous extensions, which were exposed after break of the original out layer. Yellow arrows indicate the residues of the out layer. (F) A Noc cell anchored to the AG77 hypha by the fibrous extensions. Red arrows indicate irregular tube-like extensions of the Noc cell wall connected to the surface of fungal cell wall.

https://doi.org/10.7554/eLife.47815.003

Carbon and nitrogen transfer between N. oceanica and M. elongata

To test whether nutrient, that is carbon or nitrogen exchange underlies the interaction between M. elongata and N. oceanica, we conducted a series of tracer experiments using reciprocally 14C- and 15N-labeled algal and fungal partners. For carbon exchange assays, algal cells were labeled with [14C]-sodium bicarbonate and were co-cultivated with actively growing non-labeled fungal hyphae for 1 week in flasks. Conversely, fungal hyphae were grown in either [14C]-glucose- or [14C]-acetate-containing medium for labelling. Labeled fungi were then co-incubated with non-labeled algal cells in flasks that allowed the two organisms to interact physically. Co-cultured algal and fungal cells were separated from each other by cellulase digestion and mesh filtration (Figure 2—figure supplement 1A–E). Algal and fungal cells were collected and analyzed for 14C exchange, separately. Isotope analyses indicated that a significant amount of 14C-carbon was transferred from the alga to the fungus, and nearly 70% of the total transferred 14C-carbon was incorporated into the fungal lipid pool, with the remaining incorporated into free amino acids (FAAs), proteins, soluble compounds, and carbohydrates (Figure 2A, left). Similarly, 14C-carbon transfer was observed from the labeled fungus to its algal recipient (Figure 2A, right). Fractions of algal cells attached to the fungal hyphae acquired more 14C than unattached cells sampled in the supernatant (Figure 2A and Figure 2—figure supplement 1F and G).

Figure 2 with 4 supplements see all
Carbon exchange between N.oceanica and M. elongata AG77.

(A) Carbon (C) transfer from [14C]-sodium bicarbonate (NaHCO3)-labeled N. oceanica (Noc) cells to M. elongata AG77 (Mel AG77, left panel) or from [14C]-glucose-labeled AG77 to Noc cells (right panel) after 7-d co-culture in flasks (with physical contact). Radioactivity of 14C-carbon was determined with a scintillation counter (dpm, radioactive disintegrations per minute) and then normalized to the dry weight of samples (dpm/mg biomass). Free Noc, unbound Noc cells in the supernatant; attached, Noc cells separated from AG77-Noc aggregates by algal cell wall digestion and mesh filtration; FAAs, free amino acids; soluble compounds, supernatant after acetone precipitation of proteins extracted by SDS buffer. Data are presented as the average of three biological replicates with standard deviation (Means ± SD, n = 3). (B) 14C-carbon transfer between Noc and AG77 without physical contact. Algae and fungi were incubated in cell-culture plates with filter-bottom inserts (pore size of 0.4 μm) which separate Noc cells and AG77 mycelium from each other but allow metabolite exchange during co-culture. Error bars indicate SD of three biological replicates (n = 3). (C and D) Relative abundance of 14C-carbon radioactivity in recipient cells compared to 14C-labeled donor cells after 7-d co-culture. (C) AG77 relative to [14C]-NaHCO3-Noc (100%). (D) Noc relative to [14C]-glucose-labeled AG77 (100%). Physical contact, living 14C-labeled cells added to unlabeled cells for co-cultivation in flasks; no contact, samples grown separately in plates with inserts; heat-killed 14C-cells, 14C-labeled Noc or AG77 killed by heat treatment at 65°C for 15 min before the addition to unlabeled cells in flasks. Free, unbound Noc cells in the supernatant; Att, Noc cells attached to AG77 (isolated by algal cell wall digestion and mesh filtration); Total, Noc cells grown separately from AG77 in plates and inserts. Error bars indicate SD of three biological replicates (n = 3). (E) Nitrogen (N) exchange between N. oceanica (Noc) and M. elongata AG77 examined by 15N-labeling experiments. [15N]-potassium nitrate-labeled Noc cells or [15N]-ammonium chloride-labeled AG77 were added to unlabeled AG77 or Noc cells, respectively, for 7-day co-culture in flasks (physical contact) or cell-culture plates with inserts (no physical contact). Algae and fungi were separated and weighed (dry biomass) after the co-culture, and their isotopic composition in Atom% 15N [15N/(15N+14N)100%] and N content (%N) were determined using an elemental analyzer interfaced to an Elementar Isoprime mass spectrometer following standard protocols. The N uptake rate of 15N-Noc-derived N (15N) by AG77 from and that of 15N-AG77-derived N by Noc cells (15N) were calculated based on the Atom% 15N, %N and biomass. C, chloroplast; N, nucleus; Nu, nucleolus; M, mitochondrion; V, vacuole; L, lipid droplet. Values are the average of three biological repeats.

https://doi.org/10.7554/eLife.47815.006

To assess whether a physical interaction is required for carbon exchange between the photosynthetic alga and the putative fungal heterotroph, we used membrane inserts to physically separate reciprocally 14C-labeled algal and fungal partners (Figure 2—figure supplement 2A–C). We observed that the physical contact between the algae and fungus is essential for 14C-carbon transfer to the fungus (Figure 2B and C) but is not necessary for 14C-carbon transfer to the algal cells (Figure 2B and D and Figure 2—figure supplement 2D).

Considering that Mortierella is commonly regarded as a saprotroph that acquires carbon from dead organic matter (Phillips et al., 2014), we tested whether alga-derived carbon obtained by M. elongata was due to the consumption of algal detritus. First, we repeated the 14C-labeling experiment described above using a 65°C water bath to kill 14C-labeled cells prior to algal-fungal reciprocal pairings. We found that M. elongata incorporates only a small amount (1.3%) of 14C-carbon from dead algal cells, compared to 14C-carbon acquired from living algal cells (12.7%) (Figure 2C and Figure 2—figure supplement 2E). In contrast, the algal cells attached to fungal hyphae (Att) and those free in the medium (Free) acquired more 14C-carbon (Att, 2.4%; Free, 15.8%) from dead fungal cells (Figure 2D). The total abundance of 14C-carbon was higher in the free algal cells, because most of the N. oceanica cells in the medium were free and contained a similar amount of 14C-carbon per mg compared to attached cells (Figure 2—figure supplement 2F). Second, we used confocal microscopy and Sytox Green staining to assess whether fungal and algal cells remained alive during co-culture. Over 95% of algal cells were alive during the period of reciprocal co-cultivation with 14C-carbon-labeled cells, and no dead fungal cells were observed (Figure 2—figure supplement 3A–I). Moreover, the micrographs show that the heat treatment was effective in killing algal and fungal cells (Figure 2—figure supplement 3C–E). Together these data indicate that carbon-transfer from the alga to the fungus is dependent upon physical interaction between living partners. In contrast, the algal cells are able to utilize carbon from the fungus grown in the same culture regardless of whether the hyphae are alive, dead or physically connected.

Nitrogen is a major macronutrient that limits net primary productivity in terrestrial and aquatic ecosystems, especially for microalgae such as N. oceanica (Howarth et al., 1997; Vieler et al., 2012; Zienkiewicz et al., 2016). To determine whether nitrogen exchange occurs between M. elongata and N. oceanica, we grew algal cells with [15N]-potassium nitrate and the fungus with [15N]-ammonium chloride as the sole nitrogen source. The labeled cells were co-cultivated with unlabeled partners for 1 week, and then were separated and analyzed for 15N. We detected 15N-nitrogen transfer between algal and fungal partners, irrespective of whether they were in physical contact or not (Figure 2E and Figure 2—figure supplement 4). Over twice as much 15N (~1.6 μmol/mg biomass/d) was transferred from the 15N-fungus to the algal recipient, than from the 15N-algal cells to the fungus (~0.7 μmol/mg biomass/day; Figure 2E), demonstrating a net nitrogen benefit for the alga when co-cultivated with the fungus. The N transfer under conditions of no-contact between the algae and fungi is relatively high compared to the experiment allowing physical-contact, possibly due to the differences in the culturing system. The physical-contact culture was grown in 125-mL flask containing 25 ml medium, while the no-contact culture was incubated in the 6-well culture plates with 5 ml medium in each well, which is a denser culture with the two species only separated by a thin membrane.

Nutrient deficiency and benefits of co-cultivation for N. oceanica and M. elongata

To test whether the algae and fungi benefit from the interaction and the exchange of nutrients, we observed growth in macro- and micronutrient-deficient media. During nitrogen or carbon deprivation in f/2 medium, N. oceanica had significantly increased viability when co-cultivated with M. elongata (Figure 3A–C). No impact of micronutrients was detected. Element analysis of the culture supernatant showed an increase in total organic carbon and dissolved nitrogen when the living M. elongata hyphae were incubated alone in f/2 medium (Figure 3D and E). This is indicative of extracellular release of nutrients by the hyphae, and may explain why physical contact is not required for the 14C-carbon transfer from the fungus to the alga. In addition, following 10-day-prolonged incubation in regular f/2 medium N. oceanica cells showed significant higher levels of chlorophyll with the presence of M. elongata compared to algal cells grown alone, suggesting that the co-cultured algae likely had a higher photosynthetic capacity (Figure 3—figure supplement 1A).

Figure 3 with 4 supplements see all
N.oceanica benefits from co-culture with M. elongata.

(A–C) Viability assay of Noc cells and Noc co-cultured with AG77 under nitrogen (-N, (A) and carbon (-C, (B) deprivation. Dead Noc cells were indicated by SYTOX Green staining (green fluorescence). Red, Noc chlorophyll fluorescence. (C) Viability of nutrient-deprived Noc cells increased when co-cultured with two different M. elongata strains, AG77 and NVP64. Results are calculated from 1000 to 5000 cells of five biological repeats with ImageJ. Asterisks indicate significant differences compared to the Noc control as determined by Student’s t test (*p≤0.05, **p≤0.01; Means ± SD, n = 5). (D and E) Total organic C and dissolved N measurements in the buffer of 18-day fungal cultures of M. elongata strains AG77 and NVP64 compared to the f/2 medium control (f/2 con). Fungal cells were removed by 0.22-μm filters. Data are presented as the average of four biological replicates and asterisks indicate significant differences compared to the f/2 medium control as determined by Student’s t test. Means ± SD, n = 4. *p≤0.05, **p≤0.01.

https://doi.org/10.7554/eLife.47815.011

On the other hand, since the viability of M. elongata was not obviously affected following nutrient deprivation (Figure 3—figure supplement 1B–F), the biomass and growth of Mortierella were estimated using a fatty acid biomarker that can be readily quantified by gas chromatography (GC), and light microscopy, respectively. It was not practical to directly determine fungal biomass, because of the difficulty of completely separating algal and fungal cells without lysing cells or losing significant biomass. To address this issue, we used fatty acid profiling of N. oceanica and M. elongata to identify a biomarker, linolenic acid (C18:3), which is a fatty acid that is predominantly present in the fungus (Du et al., 2018a). Thus, we used linolenic acid as a proxy to quantify the fungal biomass taking into account that the linolenic acid composition in the fungal biomass was consistent following the incubation in N-deprived f/2 medium (Figure 3—figure supplement 2A). Due to the tight interaction between the algae and fungi, it is impractical to accurately determine the correlation of the biomarker with fungal biomass under co-culturing conditions. Instead, we used the correlation of C18:3 with biomass of the fungus grown in N-depleted f/2 medium as a proxy for the fungal biomass in the co-cultures. We made the assumption that relative change of C18:3 in co-cultured and free cells were insignificant allowing for an accurate estimate of fungal biomass in both conditions. Linolenic acid was quantified by GC of its fatty acid methyl ester derivative, from which fungal biomass was calculated. The algal biomass was calculated by subtraction of fungal biomass from the total biomass of alga-fungus aggregates. Significant increases in biomass were observed for the co-cultured alga and fungus, but not when the alga or fungus were grown by themselves (Figure 3—figure supplement 2B). Therefore, both partners benefitted in this interaction. M. elongata was able to grow in nutrient-deprived conditions (PBS buffer) in the presence of the algal photobiont, but not when it was incubated by itself in PBS buffer without carbon (Video 1). Thus, both N. oceanica and M. elongata appear to benefit from their interaction and nutrient exchanges.

Video 1
M. elongata AG77 mycelia (~2 days in PDB), N. oceanica (Noc, 5 days in f/2) or AG77-Noc aggregates (7-d co-culture) were washed three times with phosphate-buffered saline (PBS, pH7.0–7.2, Life Technologies) and incubated in flasks containing PBS for ~2 days.

Samples were then transferred to 35-mm-microwell dishes (glass top and bottom, MatTek) containing semisolid medium (PBS supplemented with 0.25% low-gelling-temperature agarose). The growth of samples was recorded by time-lapse photography (every 20 min for 4 days) using a Leica DMi8 inverted microscope with DIC and time-lapse function. Resultant images were used to create the movie with video-editing software (VideoStudio X9, Corel) to compare the growth of AG77 with or without symbiotic algal cells in nutrient-limited PBS buffer. Only hyphae of AG77-Noc aggregates kept growing, indicating that AG77 benefits from the co-culture with Noc cells.

https://doi.org/10.7554/eLife.47815.016

Specificity of fungi hosts in interactions with N. oceanica

Numerous lineages of fungi have evolved to interact with plants and algae. The question arises whether the interaction we observed is unique to Mortierella or, alternatively, if it is conserved across diverse lineages of fungi. We addressed this through a series of interaction experiments pairing N. oceanica with a panel of 20 fungi (Figure 3—figure supplement 3A). These phylogenetically diverse fungal isolates represented three phyla, 9 orders and 13 families of fungi across trophic strategies from plant-associated fungal mutualists to pathogens and included the yeast Saccharomyces cerevisiae, as well as filamentous ascomycetes, basidiomycetes, and mucoromycetes (Bonito et al., 2016). Mortierella elongata showed the most obvious phenotype of flocculating alga, which consisted of algal cells clustered around the fungal mycelium (Figure 3—figure supplement 3B). Aside from a few Mortierella species tested, interactions between the other fungi and the alga were neutral or negative. It is worth noting that N. oceanica cells maintained an intact and smooth outer layer when co-cultured with the negatively interacting fungi such as Clavulina sp. PMI390 and Morchella americana GB760 (Figure 3—figure supplement 4).

Long-term co-cultivation leads to internalization of N. oceanica within M. elongata hyphae

Microbial consortia may persist in a stable state, improving each other’s resilience to fluctuating environments and stresses (Brenner et al., 2008). To assess whether the observed interaction between N. oceanica and M. elongata was stable or transient, we carried out a series of long-term incubations (from 1 to 3 months) in which the partners were grown together and nutrients refreshed biweekly. After ~1 month of co-culture, confocal microscopy was used to visualize cells inside the thick aggregates that formed between the alga and the fungus. To delineate cell walls, we used a wheat germ agglutinin conjugate cell wall probe, which binds to N-acetylglucosamine, a component in both the Mortierella and Nannochloropsis cell walls (Javot et al., 2007; Scholz et al., 2014). Microscopic observations indicated the presence of algal cells within fungal hyphae (Figure 4—figure supplement 1A–C and Video 2). Subsequent light and transmission electron microscopy (TEM) were used to further observe this phenomenon, whereby algal cells had been incorporated within hyphae. Differential interference contrast (DIC) microscopy showed the morphology of the ‘green hyphae’ after long-term co-culture, corroborating the presence of intact and presumably functional algal cells attached to the hyphal tip (Figure 4A) and present inside the fungal hyphal cells (Figure 4B–E and Video 3). After long-term co-culture, algae-fungi aggregates became thick and difficult to observe well with light microscopy (Figure 4—figure supplement 1D and E). The viability of M. elongata was demonstrated by transferring M. elongata-N. oceanica aggregates to fresh PDB/2 plates (Figure 4—figure supplement 1F). Additional imaging with TEM was performed to characterize the M. elongata-N. oceanica aggregates. Algal cells were seen outside of fungal cells surrounded by the fungal mycelium (Figure 4—figure supplement 1G–I); however, some algal cells are clearly present within the hyphae (Figure 4F and G and Figure 4—figure supplement 2). Fungal cytoplasmic contents were visible suggesting fungal cells containing algae were alive and functional (Figure 4).

Figure 4 with 6 supplements see all
Intracellular localization of long-term co-cultured N.oceanica within M. elongata AG77 hyphae.

(A–E) DIC images of AG77 ‘green hyphae’ with N. oceanica (Noc) cells inside. The red arrow heads indicate Noc cells at the tip region of the hypha. (B and C) AG77 and Noc co-cultured for ~1 month. (C–E) AG77 and Noc co-cultured over 2 months. (F and G) Transmission electron microscope (TEM) images of increasing magnification showing a cross-section of AG77 mycelium containing a cluster of Noc cells. AG77 and Noc were co-cultured for ~1 month. Red arrowheads indicate the same position. M, mycelium; Nw, Noc cell wall; Mw, Mortierella cell wall; Cy, cytoplasm; V, vacuole C, chloroplast; Mo, Mortierella organelles.

https://doi.org/10.7554/eLife.47815.017
Video 2
Animation of 3D z-stacks of N. oceanica (Noc) or M. elongata AG77-Noc aggregates (35-day co-culture) stained by Wheat Germ Agglutinin Conjugate (WGA) and observed with a confocal laser scanning microscope (FluoView 1000, Olympus).

Green, WGA fluorescence indicates cell wall of Noc and AG77; red, Noc chlorophyll fluorescence.

https://doi.org/10.7554/eLife.47815.024
Video 3
Videos recorded with a Leica DMi8 inverted microscope to show the morphology of green hyphae in AG77-Noc aggregates (co-cultured over 2 months).
https://doi.org/10.7554/eLife.47815.025

While there is no indication that algae are transmitted vertically through fungal reproductive structures, the algal cells remained viable (growing and dividing) during 2 months of co-culture (Video 4). We were not able to capture the exact transitional stage of entry of N. oceanica into hyphae of M. elongata by TEM; however, through DIC and time-lapse microscopy, we repeatedly observed that internalization of algae is preceded by dense aggregation of algal cells around the hyphal tip (Figure 4—figure supplement 3). Dense clusters of algal cells at the tip of a hypha were consistently observed when algal cells were found within fungal hyphae growing in a semisolid medium (Figure 4—figure supplement 4). Furthermore, hyphae proximal from these tips were often green, and the number of algae within these cells increased over time (Figure 4A–E). In fact, trapped algal cells were able to grow and divide within their host (Video 4). To further examine the viability of green hyphae, confocal microscopy with SYTOX Green was carried out in the 1 ~ 2 months alga-fungus aggregates. Exclusion of the dye in this case is primarily an indicator of living fungal hyphae, while persistent green chlorophyll and dividing cells are hallmarks of living algae. The results are consistent with the notion that both fungal host and internalized algae within the hyphae are alive (Figure 4—figure supplement 5). DIC microscopy also confirmed that the algal cells inside green hyphae are surrounded by fungal organelles, especially what appear to be lipid droplets (Figure 4—figure supplement 6 and Video 5).

Video 4
N. oceanica (Noc) cells internalized within a M. elongata AG77 hypha recorded by time-lapse photography (every 10 min for 6 d), showing several Noc cells growing and dividing within the hypha.
https://doi.org/10.7554/eLife.47815.026
Video 5
Light microscope video to show the N. oceanica (Noc) cells inside a M. elongata AG77 hypha.

The green Noc cells are surrounded by fungal organelles such as lipid droplets (blue green/gray) that are presented in living hypha.

https://doi.org/10.7554/eLife.47815.027

Discussion

Here, we show that the alga N. oceanica and the fungus M. elongata have a strong and specific interaction, whereby N. oceanica aggregates along the fungal mycelium during co-incubation. We further demonstrate that in co-culture, living algal and fungal partners establish a tight physical association preceded by the loss of the smooth outer portion of the algal cell wall revealing fibrous extensions, and eventually leading to the incorporation of algal cells within the fungal mycelium. Isotope tracer studies provided evidence for the reciprocal exchange of carbon and nitrogen between living algal and fungal symbionts. Results from isotope tracer and nutrient deficiency studies support our hypothesis that the Mortierella-Nannochloropsis interaction is mutualistic in nature, based upon carbon and nitrogen acquisition and transfer, with potentially adaptive benefits provided to both partners under nutrient-limited conditions. While the apparent fungal-algal symbiosis may conjure the concept of a lichen, it differs in many respects. Mortierella lacks distinct tissue differentiation or hyphal structures (i.e. thallus, haustoria). Moreover, hyphae of Mortierella harbor algal cells intracellularly while lichens maintain algae in a fungal matrix, but external to their cells. In fact, we know of no other examples of fungi that have been shown to host intracellular eukaryotic photobionts.

Endosymbiosis of living eukaryotic cells by fungal hypha is not known from nature nor the lab. However, Geosiphon pyriformis, an early diverged fungal relative of Mortierella and arbuscular mycorrhizal fungi, does form a unique intracellular association with the photosynthetic cyanobacterium Nostoc punctiforme (Mollenhauer et al., 1996). In this fungal-bacterial photobiont symbiosis, the fungus envelops the photosynthetic Nostoc within a specialized swollen multinucleate fungal ‘bladder’ that is morphologically distinct from the rest of the fungal mycelium (Schüßler et al., 1996). Within this bladder, the cyanobacteria are surrounded by a host-derived symbiosome membrane specialized for acquiring photosynthate (Brenner et al., 2008).

Our attempts to capture the biogenesis of N. oceanica internalization within M. elongata through DIC and time-lapse microscopy show that algal internalization is preceded by dense aggregates of algal cells close to the fungal hyphal tip (Figure 4—figure supplement 3A and B). The hyphal tip is the actively growing region of the fungal colony, and a point of growth, membrane endocytosis, and cell wall construction (Steinberg, 2007). Aggregates of algal cells surrounding hyphal tips were also frequently observed in the long-term co-culture of the fungus and the alga (Figure 4—figure supplement 3C–F), and dense clusters of algal cells forming at the tip of a hypha were observed when partners were grown in a semisolid medium (Figure 4—figure supplement 4). Hyphae proximal from these tips were often green, and the number of algae within the cells increased over time (Figure 4A–E). Not only do algal cells enter the fungal hyphae, but their plastids were intact with all indications of being healthy and photosynthetically active. Further, Nannochloropsis cells continued to grow and multiply within the fungal mycelium (Video 4) and both fungal host and algae inside of green hyphae remained alive after 2 months co-culture (Figure 4—figure supplements 5 and 6 and Video 5). The living fungal hyphae exclude SYTOX green invalidating this assay for the internalized algae as they may never see the dye. However, the algal cells inside the hypha showed autofluorescence of chlorophyll, consistent with assembled photosynthetic membranes inside chloroplasts that are usually observed in living cells but not in dead disintegrating cells. In addition, it should be noted that instantly killed algal cells may still have chlorophyll autofluorescence as indicated by the white arrow heads in Figure 4—figure supplement 5C. Thus, a better indicator for living algal cells in the hyphae is the fact that they are actively dividing. Ultimately, we hypothesize that the hyphal tip is the initial point of entry for the algal cells into the hyphae, as the hyphal tips are the least differentiated cell features in a mycelial network, and the site of continuous plasma membrane recycling and cell wall generation.

Irrespective of whether the species studied here are ancient fungal-algal symbionts, or whether our findings demonstrate a latent capacity for intricate fungal-algal interactions, Mortierella fungi do share habitat and ecological niches with other Chromealveolate and algae. Although we have studied these organisms in a synthetic co-culture, given the global distribution of M. elongata and N. oceanica, it is plausible that these organisms (or their relatives) interact naturally, such as in marine tidal zones, estuaries or other ecotones.

Many endosymbionts are acquired laterally, such as Symbiodinium photobionts of coral (Mies et al., 2017). Other endosymbionts are heritable, such as the bacterial endosymbiont Glomeribacter gigasporarum, which is transmitted in spores of mycorrhizal fungi (Bianciotto et al., 2004). Given the relative size of Nannochloropsis cells (2–3 μm) to those of sporangiospores (2–3 μm), chlamydospores (23–30 μm) and zygospores (25–38 μm) of M. elongata, it is possible that internalized algae are transmitted vertically to fungal progeny. However, we have no evidence for heritability at this time.

Fungi are ubiquitous and function as root symbionts for a majority of land plant lineages and are regarded as having an essential role in the terrestrialization of Earth (Field et al., 2016). Recent phylogenomic analyses of Fungi resolve Mucoromycota as the earliest monophyletic lineage of plant-associated fungi (Spatafora et al., 2016), which include Glomeromycotina (arbuscular mycorrhizal fungi), Mucoromycotina (sugar-fungi), and Mortierellomycotina (soil molds). One trait that characterizes fungi in the Mucoromycota is their ability to form intimate cross-kingdom intracellular symbioses, which may be facilitated by the lack of regular septate cross walls between hyphal cells. These fungi form mycorrhizal symbiosis with extant early diverging plant lineages including hornworts, liverworts and mosses (Redecker et al., 2000; Spatafora et al., 2016). This radiation of Mucoromycota involved the loss of flagella in this lineage of Fungi and was contemporaneous with at least two radiations of green algae prior to the emergence of terrestrial embryophytes (Lutzoni et al., 2018). We show here that Mortierella that belong to this lineage also interact with the microalgae N. oceanica. This finding enables future studies on cross-kingdom signaling, and genetic and metabolic factors underlying this symbiosis.

Mortierella ecology has evaded mycologists for centuries. These fungi are commonly isolated from soils and plant roots (Summerbell, 2005). They have also been isolated from strata directly under green macroalgae in Antarctica (Furbino et al., 2014). Mortierella is common within cryptobiotic desert crusts (along with bacteria, algae, and other fungi) (Bates et al., 2010), and M. elongata has even been detected in association with red algae in alpine snow packs (Brown and Jumpponen, 2014). Although commonly regarded as soil saprotrophs in the literature, our results demonstrate that at least some of these fungi are also involved in biotrophic mutualisms. Although N. oceanica is not a charophyte, the closest algal relative to land plants identified to date, our study shows that an early diverging fungus is adapted to form a biotrophic mutualism with Chromalveolate alga, indicating that algal-fungal mutualisms may be more ancient and diverse than previously recognized.

Conclusions

Through stable- and radio-isotope-tracer experiments, metabolic analysis and microscopy, we report that the globally distributed early-diverging terrestrial fungus M. elongata interacts intimately with the marine alga N. oceanica in a mutualism that leads to the incorporation of intact living algal cells within fungal hyphae. This symbiosis appears to be based upon an exchange of carbon and nitrogen between the cells. M. elongata is the first taxon in the Kingdom Fungi that has been shown to internalize actively photosynthesizing eukaryotic cells.

Materials and methods

Key resources table
Reagent type
(species) or
resource
DesignationSource or
reference
IdentifiersAdditional
information
Strain (Nannochloropsis oceanica CCMP1779)NocProvasoli-Guillard National Center for Culture of Marine PhytoplanktonCCAP211/46Kuwait Institute for Scientific Research
Strain (Mortierella elongata AG77)AG77/Mel AG77Uehling et al., 2017North Carolina
Strain (Mortierella elongata NVP64)NVP64Uehling et al., 2017Michigan
Strain (Mycoavidus cysteinexigens)M. cysteinexigensUehling et al., 2017
Commercial assay or kitSYTOX GreenThermo Fisher ScientificR37168
Commercial assay or kitBODIPY 493/503Thermo Fisher ScientificD3922
Commercial assay or kitWheat Germ Agglutinin Conjugate Alexa Fluor 488Thermo Fisher ScientificW11261
Commercial assay or kitresin Epon/Araldite mixtureElectron Microscopy Sciences13940
Chemical compound, drug[14C]-sodium bicarbonateAmerican Radiolabeled ChemicalsARC 0138 C-1 mCi
Chemical compound, drug[14C]-D-glucoseMoravek BiochemicalsMC144W
Chemical compound, drug[14C]-sodium acetateAmerican Radiolabeled ChemicalsARC 0101A
Chemical compound, drug[15N]-ammonium chlorideSigma-Aldrich299251
Software, algorithmVideoStudio X9VideoStudioX9

Strains and growth conditions

Request a detailed protocol

The marine alga Nannochloropsis oceanica CCMP1779 was obtained from the Provasoli-Guillard National Center for Culture of Marine Phytoplankton and incubated as previously described (Vieler et al., 2012). In brief, N. oceanica cells were grown in f/2 medium containing 2.5 mM NaNO3, 0.036 mM NaH2PO4, 0.106 mM Na2SiO3, 0.012 mM FeCl3, 0.012 mM Na2EDTA, 0.039 μM CuSO4, 0.026 μM Na2MoO4, 0.077 μM ZnSO4, 0.042 μM CoCl2, 0.91 μM MnCl2, 0.3 μM thiamine HCl/vitamin B1, 2.05 nM biotin, 0.37 nM cyanocobalamin/vitamin B12, and 20 mM sodium bicarbonate and 15 mM Tris buffer (pH 7.6) to prevent carbon limitation. The cultures were incubated in flasks under continuous light (~80 μmol/m2/s) at 22°C with agitation (100 rpm). Log-phase algal cultures (1 ~ 3×107 cells/mL) were used for co-cultivation with fungi. Cell size and density of algal cultures were determined using a Z2 Coulter Counter (Beckman). Mortierella elongata AG77 and NVP64 isolates were made from soil samples collected in North Carolina (AG77) and Michigan (NVP64), USA. M. elongata AG77 and NVP64 are known to contain an endosymbiotic bacterium, Mycoavidus cysteinexigens, and were cleared of this endosymbiont through a series of antibiotic treatments as previously described (Partida-Martinez and Hertweck, 2007; Uehling et al., 2017). The resultant Mycoavidus-free strains were used for the co-cultivation with N. oceanica. Other fungal strains used in this study were obtained from the fungal culture suppliers and isolated from sporocarps, soils, and from healthy surface-sterilized Populus roots obtained from the Plant-Microbial Interfaces project (Bonito et al., 2016). Fungi were incubated in flasks containing PDB medium (12 g/L potato dextrose broth and 5 g/L yeast extract, pH5.3) at room temperature (RT, ~22°C).

For co-culturing N. oceanica and fungi, fungal mycelia were briefly blended into small pieces (0.5 to 2 cm) using a sterilized blender. After a 24-hr recovery in PDB medium, fungal cells were collected by centrifugation (3000 g for 3 min), washed twice with f/2 medium and resuspended in ~15 mL f/2 medium. A portion of fungal mass (3–4 mL) was used for the calculation of dry biomass: 1 mL was transferred and filtered through pre-dried and pre-weighed Whatman GF/C filters and dried overnight at 80°C. A similar method was used for the measurement of algal biomass. About a 3:1 ratio of fungal:algal biomass was used for co-cultivation on a shaker (~60 rpm) under continuous light (~80 μmol/m2/s) at RT. After 18-day co-culture, the shaker was turned off to allow free settling of the algal and fungal cells overnight. The supernatant was removed and the same volume of fresh f/2 medium containing 10% PDB was added to the culture. After that, the alga-fungus co-culture was refreshed biweekly with f/2 medium supplemented with 10% PDB.

Nutrient deprivation of the co-culture was performed according to a published protocol for N. oceanica (Vieler et al., 2012). Mid-log-phase N. oceanica cells (~1×107 cells/mL) grown in f/2 media (25 mL) were harvested by centrifugation and washed twice with nutrient-deficient f/2 media [without carbon (-C), nitrogen (-N) or phosphorus (-P)] and resuspended in 25 mL nutrient-deficient f/2 media, respectively. AG77 mycelia grown in PDB medium were washed twice with the nutrient-deficient f/2 and added into respective N. oceanica cultures for co-cultivation. To block air exchange, the flasks of -C cultures were carefully sealed with Parafilm M over aluminum foil wrap. Cell viabilities were analyzed by confocal microscopy after 10-day co-culture of -N and 20 days of -C.

Light microscopy

Request a detailed protocol

Interaction and symbiosis between the alga and the fungus were examined with an inverted microscope with differential interference contrast (DIC) and time-lapse modules (DMi8, Leica). DIC images were taken from the alga-fungus aggregates after short- (6 days) and long-term (over 1 month) co-cultivation. To characterize the algal endosymbiosis in the fungus, DIC and time-lapse photography were performed after long-term co-culture of the alga and fungus (from 1 to 3 months). For viewing alga-fungus aggregates grown in flasks, the samples were transferred to 35-mm-microwell dishes (glass top and bottom, MatTek) and embedded in a thin layer of semisolid f/2 medium supplemented with 10% PDB and 0.25% low-gelling-temperature agarose (Sigma-Aldrich) to immobilize the cells. The morphology of green hyphae (AG77 hyphae containing intracellular N. oceanica cells) was recorded in DIC micrographs, as well as real-time videos that showed four groups of green hyphae (Video 3). Videos were assembled side by side in Video 3 using video-editing software VideoStudio X9 (Corel). To investigate the establishment of algal cells living inside fungal hypha, randomly selected alga-fungus aggregates were sub-cultured from 35-day co-cultures in 35-mm-microwell dishes containing semisolid f/2 medium with 10% PDB and 0.25% agarose and observed directly in 35-mm-microwell dishes containing semisolid f/2 medium (Figure 4—figure supplement 3; Figure 4—figure supplement 4) and through time-lapse photographs that were combined together with the software VideoStudio to create Video 4.

Scanning electron microscopy

Request a detailed protocol

SEM was performed at the Center for Advanced Microscopy of Michigan State University (CAM, MSU) to investigate the physical interaction between N. oceanica and M. elongata. Alga-fungus aggregates from 6-day co-cultures of N. oceanica and fungal strains were used for interaction analysis, including M. elongata AG77, NVP64 and Clavulina PMI390 and Morchella Americana GB760, which do not have interaction phenotype when co-cultured with algae. N. oceanica cells grown alone in f/2 medium were used as a control. We also observed N. oceanica cells co-cultured with 65°C-killed AG77 mycelium, and algal cells from the supernatant of living M. elongata-N. oceanica co-cultures that were unattached from fungal-algal aggregates. To mimic the exposed fibrous extensions of N. oceanica cells following physical interaction with M. elongata, different enzymes were tested to digest the out layer of algal cell wall. N. oceanica cells were washed with PBS buffer and incubated with different combination of enzymes in PBS buffer at RT for 3 hr: 4% hemicellulase (mixture of glycolytic enzymes such as xylanase and mananase, Sigma-Aldrich); 2% driselase (mixture of carbohydrolases including laminarinase, xylanase, and cellulase, Sigma-Aldrich); 4% hemicellulase and 2% driselase; 1% chitinase (Sigma-Aldrich); 1% lysing enzymes (mixture of glucanase, protease, and chitinase, Sigma-Aldrich). The samples were fixed in 4% (v/v) glutaraldehyde solution and dried in a critical point dryer (Model 010, Balzers Union). After drying, the samples were mounted on aluminum stubs using high vacuum carbon tabs (SPI Supplies) and coated with osmium using a NEOC-AT osmium coater (Meiwafosis). Processed tissues were examined using a JSM-7500F scanning electron microscope (Japan Electron Optics Laboratories).

Nutrient exchange

Request a detailed protocol

Light microscopy and SEM showed a close physical interaction between N. oceanica and M. elongata that led us to examine whether there is metabolite exchange between N. oceanica and M. elongata by isotope labeling and chasing experiments with carbon and nitrogen (14C and 15N), two of the most important nutrients for N. oceanica and M. elongata. 14C assays were performed according to published protocols with modifications (Li et al., 2012). 20 μL of [14C]-sodium bicarbonate (1 mCi/mL, 56 mCi/mmol, American Radiolabeled Chemicals) was added to 20 mL of early log-phase culture of N. oceanica (~2×106 cells/mL) and incubated for 5 days when the 14C incorporation reached ~40%. The 14C-labeled N. oceanica cells were harvested by centrifugation (4000 g for 10 min) and washed three times with f/2 medium. The supernatant of the last wash was analyzed in Bio-Safe II counting cocktail (Research Products International) using a scintillation counter (PerkinElmer 1450 Microbeta Trilux LSC), to confirm that 14C-labeling medium was washed off. The pellet of 14C-labeled N. oceanica was resuspended in 20 mL f/2 medium. Subsequently, non-labeled M. elongata AG77 mycelia (~3 x algal biomass, intact cells without blending) grown in PDB medium were washed twice with f/2 medium and added to the 20 mL 14C-labeled algal culture for 7-day co-cultivation. Alga-fungus aggregates were then harvested by PW200-48 mesh (Accu-Mesh, first filtration) and NITEX 03-25/14 mesh (mesh opening 25 μm, SEFAR, second filtration). Algal cells in the flow through were collected by centrifugation (4000 g for 10 min) and kept as the first part of 14C-labeled alga control. Alga-fungus aggregates were intensively washed in 50-mL conical centrifuge tubes containing 40 mL of f/2 medium using a bench vortex mixer (~1500 rpm, 15 min). Fungal mycelia were collected with NITEX 03-25/14 mesh; algal cells in the flow through were harvested by centrifugation and stored as the second fraction of 14C-labeled alga control. Mesh-harvested fungal mycelia (with obviously reduced the number of algal cells attached) were placed in microcentrifuge tubes containing 300 μL of PBS buffer (pH 5.0) supplemented with 4% hemicellulase and 2% driselase for overnight incubation at 37°C to digest the algal cell walls as previously described (Chen et al., 2008). After cell-wall digestion, 700 μL of f/2 medium were added and the algal cells were separated from hyphae by vortexing for 15 min. The hyphae were collected by NITEX 03-25/14 mesh, and the flow-through containing algal cells was kept as the last fraction of alga control. The fungal hyphae were washed three times with f/2 medium and then used for biomass and radioactivity measurements. The three fractions of 14C-labeled alga controls were combined together for further analyses. Half of the algal and fungal samples were dried and weighed for biomass and the rest was used for 14C measurements. The 14C radioactivity of each sample was normalized to the respective dry biomass. To examine cross contamination after alga-fungus isolation, non-radioactive samples were processed the same way and analyzed by light microscopy (Figure 2—figure supplement 1A–C) and PCR using primers specific for the N. oceanica gene encoding Aureochrome 4 (AUREO4), a blue light-responsive transcription factor that is unique in photosynthetic stramenopiles such as N. oceanica (Figure 2—figure supplement 1D): Aureo4pro F+ (5’-AGAGGAGCCATGGTAGGAC-3’) and Aureo4 DNAD R- (5’-TCGTTCCACGCGCTGGG-3’), and primers specific for M. elongata genes encoding translation elongation factor EF1α and RNA polymerase RPB1 (Figure 2—figure supplement 1E): EF1αF (5’-CTTGCCACCCTTGCCATCG-3’) and EF1αR (5’-AACGTCGTCGTTATCGGACAC-3’), RPB1F (5’-TCACGWCCTCCCATGGCGT-3’) and RPB1R (5’-AAGGAGGGTCGTCTTCGTGG-3’).

Isolated algal and fungal cells were frozen in liquid nitrogen and ground into fine powders with steel beads and TissueLyser II (QIAGEN), followed by lipid extraction in 1.2 mL chloroform:methanol (2:1, v/v) by vortexing for 20 min. After addition of double-distilled water (ddH2O, 100 μL), the samples were briefly vortexed and then centrifuged at 15,000 g for 10 min. The organic phase was collected for total lipids. One mL of 80% methanol (v/v) was added to the water phase and cell lysis to extract free amino acids (FAAs). After centrifugation at 20,000 g for 5 min, the supernatant was kept as total FAAs and the pellet was air-dried; 200 μL of SDS buffer (200 mM Tris-HCl, 250 mM NaCl, 25 mM EDTA, 1% SDS, pH7.5) was added to the pellet with incubation at 42°C for 15 min. After centrifugation at 10,000 g for 10 min, while the pellet was kept for carbohydrate analyses, the supernatant (~200 μL) was collected for further protein precipitation (−20°C, 1 hr) with the addition of 800 μL cold acetone. After the 1 hr precipitation, total proteins (pellet) and soluble compounds (supernatant) were separated by centrifugation at 20,000 g for 15 min. The pellet of total proteins was resuspended in 200 μL of SDS buffer for scintillation counting. The pellet of carbohydrates was air-dried, resuspended in 200 μL ethanol, transferred to a glass tube with Teflon-liner screw cap, and then dissolved in 2 to 4 mL of 60% sulfuric acid (v/v) according to described protocols (Velichkov, 1992; Scholz et al., 2014). As needed, vortexing and incubation at 50°C were performed. Total lipids and soluble compounds were counted in 3 mL of xylene-based 4a20 counting cocktail (Research Products International), whereas total FAAs, proteins and carbohydrates were counted in 3 mL of Bio-Safe II counting cocktail. 14C radioactivity of the samples (dpm, radioactive disintegrations per minute) was normalized to their dry weight (dpm/mg).

To examine carbon transfer from the fungus to the alga, 200 μL of 0.1 mCi/mL [14C]-D-glucose (268 mCi/mmol, Moravek Biochemicals) or 100 μL of 1 mCi/mL [14C]-sodium acetate (55 mCi/mmol, American Radiolabeled Chemicals) were added to 20 mL of M. elongata AG77 grown in modified Melin-Norkrans medium [MMN, 2.5 g/L D-glucose, 0.25 g/L (NH4)2HPO4, 0.5 g/L KH2PO4, 0.15 g/L MgSO4, 0.05 g/L CaCl2]. After 5-d 14C-labeling, fungal mycelia were harvested and washed three times with f/2 medium. The supernatant of the last wash was confirmed to be free of 14C by scintillation counting. 14C-labeled hyphae were added to 20 mL of N. oceanica culture for 7-day co-culture. Alga-fungus aggregates were harvested using PW200-48 and NITEX 03-25/14 meshes. Algal cells in the flow-through were harvested and washed twice with f/2 medium by centrifugation and kept as free N. oceanica (unbound algal cells). The remaining steps of sample preparation and 14C measurement were performed as described above.

To test whether physical contact is necessary for the carbon exchange between N. oceanica and M. elongata, 14C-label experiments were carried out using standard six-well cell culture plates (5 mL medium of each well) with inserts that have a bottom composed of hydrophilic polytetrafluoroethylene membrane filters (pore size of 0.4 μm, Millipore) to grow the alga and fungus together, which allows metabolite exchange but no physical contact. 14C-labeling was performed in the same way as described above. For alga-fungus co-culture, 14C-labeled algal cells (or fungal hyphae) were added in either plate wells or cell culture inserts while respective hyphae (or algal cells) were grown separately in the inserts or plate wells to examine cross contamination (Figure 2—figure supplement 2A). After 7-day co-culture, algal and fungal cells grown in the insert-plate system were easily separated by moving the insert to an adjacent clean well (Figure 2—figure supplement 2B and C). Samples were then processed following the protocol described above (without the steps of mesh filtration and cell-wall digestion).

Considering that Mortierella fungi are saprotrophic (Phillips et al., 2014), we performed 14C-label experiments using heat-killed 14C-cells to test whether the alga and fungus utilize 14C from dead cells. Briefly, 14C-labeled algal or fungal cells were washed three times with f/2 medium and incubated in a water bath at 65°C for 15 min, which killed the cells without causing significant cell lyses. Heat-killed 14C-algal cells (or fungal hyphae) were co-cultivated with unlabeled hyphae (or algal cells) for 7 days in flasks. Subsequently, the algal and fungal cells were separated by cell-wall digestion and mesh filtration, and 14C radioactivity of the samples was measured by scintillation counting as described above.

Nitrogen is another major nutrient for N. oceanica (Vieler et al., 2012; Zienkiewicz et al., 2016) and Mortierella (Thornton, 1956). Nitrogen exchange between N. oceanica and M. elongata was tested by 15N-labeling and chasing experiments using isotope ratio mass spectrometry. For 15N labeling of algal and fungal cells, N. oceanica cells were inoculated and grown in 200 mL of 15N-f/2 medium containing ~5% of [15N]-potassium nitrate [15N/(15N+14N), mol/mol], while M. elongata mycelia were inoculated and incubated in 2 L of 15N-MMN medium containing ~5% of [15N]-ammonium chloride for two weeks. The algal culture was maintained in log phase by the addition of fresh 15N-f/2 medium into a larger volume. Eventually, 15N-N. oceanica cells from a 4 L culture and 15N-M. elongata mycelium from a 2-L culture were harvested by centrifugation, with a portion of each sample kept as 15N-labeled controls. The remainder of the samples was added to unlabeled cells in flasks (with physical contact) or 6-well-culture plates with inserts (no physical contact) for 7-day co-cultivation. Algal and fungal cells were separated after the co-culture as described above. Samples were washed three times with ddH2O. Fungal mycelia were homogenized in a TissueLyser II (QIAGEN) using steel beads. The algal and fungal samples were then acidified with 1.5 to 3 mL of 1 N HCl, dried in beakers at 37°C and weighed for biomass. Isotopic composition of the samples [Atom% 15N, 15N/(15N+14N)100%] and N content (%N) were determined using a Eurovector (EuroEA3000) elemental analyzer interfaced with an Elementar Isoprime mass spectrometer following a standard protocol (Fry, 2007). The N uptake rates (μmol N/mg biomass/d) of 15N-N. oceanica cells from the medium (medium-N, isotope dilution) and that of AG77 from 15N-N. oceanica-derived N (15N) were calculated based on the Atom% 15N, %N and biomass following a published protocol (Ostrom et al., 2016). The N uptake rates of 15N-AG77 from the medium and that of recipient N. oceanica from 15N-AG77-derived N (15N) were calculated in the same way.

Confocal microscopy

Request a detailed protocol

Viability of N. oceanica and M. elongata cells during their co-culture was determined by confocal microscopy using a confocal laser scanning microscope (FluoView 1000, Olympus) at CAM, MSU. SYTOX Green nucleic acid stain (Molecular Probes, Thermo Fisher Scientific), a green-fluorescent nuclear and chromosome counterstain impermeant to live cells, was used to indicate dead cells following a published protocol (Tsai et al., 2014). Briefly, 1 μL of 5 mM SYTOX Green was added to 1 mL of cell culture and incubated for 5 min at RT in the dark. Samples were washed twice with f/2 medium before observation (SYTOX Green, 488 nm excitation, 510 to 530 nm emission; chlorophyll, 559 nm excitation, 655 to 755 nm emission). Viability of N. oceanica cells was calculated using ImageJ software. Cell viability was analyzed during the alga-fungus co-culture in flasks containing f/2 medium (1, 4 and 7 days, Figure 2—figure supplement 3F–I) to investigate whether the cells were alive or dead during the 7-day co-culture of 14C- and 15N-labeling experiments. Viability of N. oceanica cells co-cultivated with M. elongata AG77 and NVP64 under nutrient deprivation (-N and -C) was tested to show whether N. oceanica benefits from the co-culture with Mortierella fungi. Viability of M. elongata AG77 was analyzed during the 30-day incubation in f/2 medium (9, 18 and 30 days) to check whether the cells were alive or dead (Figure 3—figure supplement 1) when the culture media were collected using 0.22 μm Millipore filters after 18-day incubation for nutrient analyses (total organic C and dissolved N). Viability of green hyphae containing algal cells was analyzed in the randomly selected alga-fungus aggregates after 1–2 months co-culture.

Localization of N. oceanica cells in alga-fungus aggregates was investigated by cell-wall staining using Wheat Germ Agglutinin Conjugate Alexa Fluor 488 (WGA, Thermo Fisher Scientific) following the manufacturer’s instruction (Figure 4—figure supplement 1A–C). In brief, alga-fungus aggregates were collected by centrifugation and washed once with PBS buffer (pH7.2), followed by addition of 5 μg/mL WGA and incubation at 37°C for 10 min. Samples were washed twice with f/2 medium and observed under a FluoView 1000 microscope (WGA, 488 nm excitation, 510 to 530 nm emission; chlorophyll, 559 nm excitation, 655 to 755 nm emission).

M. elongata AG77 hyphae contain many lipid droplets visible by light microscopy. To confirm the distribution of lipid droplets in hyphae grown in single and co-culture, confocal microscopy was carried out using BODIPY 493/503 (Thermo Fisher Scientific), a lipophilic probe for lipid droplets (488 nm excitation, 510 to 530 nm emission).

Carbon and nitrogen measurements

Request a detailed protocol

Total organic C (TOC) and dissolved N (TDN) in the media of Mortierella cultures were measured with a TOC-Vcph carbon analyzer with total nitrogen module (TNM-1) and ASI-V autosampler (Shimadzu) at Kellogg Biological Station, MSU. M. elongata strains AG77 and NVP64 were incubated for 18 days in flasks containing 25 mL of f/2 medium. Fungal tissues were removed by filtration with 0.22 micron filters (Millipore) and the flow-through was subject to TOC and TDN analyses following published protocols (Heinlein, 2013; Lennon et al., 2013).

Chlorophyll assay

Request a detailed protocol

Chlorophyll measurement was performed as previously described (Du et al., 2018b). In brief, N. oceanica cells were incubated in f/2 medium until they reached stationary-phase (0 day control), and the cells were further incubated for 10 days within the same medium or with the addition of about three-times biomass of f/2-washed and blot-dried AG77 mycelium. Algal cells were collected from 1 ml culture of N. oceanica controls and unbound cells from alga-fungus co-culture by centrifugation. Chlorophyll of the algal cells was extracted by 900 μl of acetone:DMSO (3:2, v/v) for 20 min with agitation at RT, and then measured with a spectrophotometer (Uvikon 930, Kontron).

Fatty acid analysis and biomass calculation

Request a detailed protocol

Lipid extraction and fatty acid analysis were performed following a published protocol (Du et al., 2018a). Linolenic acid (C18:3) was used as a biomarker, as it is present in M. elongata AG77 but not in N. oceanica cells and its abundance in total biomass was steady following the incubation in N-deprived f/2 medium. Thus, C18:3 was quantified by gas chromatography of its methyl ester derivative and used for the calculation of fungal biomass in dense alga-fungus aggregates, when it was not feasible to physically separate algal and fungal cells without significant loss of biomass or cellular lysis. Briefly, alga-fungus aggregates were collected with mesh filtration and total lipid was extracted with methanol/chloroform/88% formic acid (1:2:0.1 by volume) and washed with 0.5 vol of 1 M KCl and 0.2 M H3PO4. After phase separation by centrifugation (3000 g for 3 min), total lipids were collected for the preparation of fatty acid methyl esters by transesterification and analysis by gas chromatography. The remaining cell lysate were dried at 80°C overnight to provide the nonlipid biomass. Total dry biomass of alga-fungus aggregates was obtained by combining the lipid and non-lipid parts. Fungal biomass within alga-fungus aggregates was quantified using the C18:3 content-based calculation. Algal biomass in aggregates was determined by subtracting fungal biomass from the total biomass.

Phylogeny of fungal strains

Request a detailed protocol

DNA was extracted from fungal isolates by placing a small amount of mycelium into 20 μL of extraction solution (Sigma-Aldrich) and heating at 95°C for 10 min, after which 60 μL of bovine serum albumin (BSA, 3%) was added to the lysate and PCR was employed to directly amplify the nuclear-encoded ribosomal RNA genes (rDNA): ITS (internal transcribed spacer) with the primers ITS1f (5’-CTTGGTCATTTAGAGGAAGTAA-3’) and ITS4 (5’-TCCTCCGCTTATTGATATGC-3’), and 28S rDNA with primers LROR (5’-ACCCGCTGAACTTAAGC-3’) and LR3 (5’-CCGTGTTTCAAGACGGG-3’) following a published PCR protocol (Bonito et al., 2016). Amplicons were sequenced with an ABI3730XL automated sequencer (Applied Biosystems). The resultant sequences were identified by BLAST in the NCBI nucleotide database (Altschul et al., 1990), and by sequence alignment in MUSCLE (Edgar, 2004). Unalignable regions were excluded in Mesquite (Maddison and Maddison, 2009). Phylogenetic relationships among isolates were inferred with PAUP* (Swofford, 2002) using the neighbor joining optimization criterion and were visualized with FigTree (Rambaut, 2007). The alignment of sequences used in this study has been deposited in TreeBase (#20243).

Transmission electron microscopy.

Request a detailed protocol

TEM was performed at CAM, MSU using N. oceanica and Mortierella aggregates co-cultured for ~1 month. Randomly collected alga-fungus aggregates were fixed overnight at 4°C in sodium cacodylate buffer (50 mM, pH 7.2) supplemented with 2.5% (v/v) glutaraldehyde. The fixed samples were washed three times with sodium cacodylate buffer, post-fixed in 1% OsO4 (v/v) for 2 hr at RT and then washed three times with sodium cacodylate buffer. After dehydration with a graded series of ethanol and acetone, the samples were infiltrated through a series of acetone/resin Epon/Araldite mixtures and finally embedded in resin Epon/Araldite mixture (Electron Microscopy Sciences). Ultrathin sections (70 nm) were cut with an ultramicrotome (RMC Boeckeler) and mounted onto 150 mesh formvar-coated copper grids, followed by staining with uranyl acetate for 30 min at RT. The sections were then washed with ultrapure water and stained 10 min with lead citrate and used for observation. Images were taken with a JEOL100 CXII instrument (Japan Electron Optics Laboratories) with SC1000 camera (Model 832, Gatan) and were processed with ImageJ.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
    Application of Microalgae and Fungal-Microalgal Associations for Wastewater Treatment
    1. EJ Egede
    2. H Jones
    3. B Cook
    4. D Purchase
    5. A Mouradov
    (2016)
    In: D Purchase, editors. Fungal Applications in Sustainable Environmental Biotechnology. Springer. pp. 143–181.
  13. 13
  14. 14
  15. 15
  16. 16
    Where the Water Meets the Sky: The Effects of Atmospheric Ozone Pollution on Aquatic Algal and Bacterial Communities
    1. J Heinlein
    (2013)
    Michigan State University.
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
    Mesquite: a modular system for evolutionary analysis
    1. WP Maddison
    2. DR Maddison
    (2009)
    Mesquite: a modular system for evolutionary analysis, 2.6.
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
    FigTree
    1. A Rambaut
    (2007)
    A Graphical Viewer of Phylogenetic Trees.
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
    Paup*. Phylogenetic Analysis Using Parsimony (*and Other Methods)
    1. DL Swofford
    (2002)
    Sinauer Associates, Sunderland, Massachusetts.
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48

Decision letter

  1. Ian T Baldwin
    Senior Editor; Max Planck Institute for Chemical Ecology, Germany
  2. Maria J Harrison
    Reviewing Editor; Boyce Thompson Institute for Plant Research, United States
  3. Maria J Harrison
    Reviewer; Boyce Thompson Institute for Plant Research, United States
  4. Paola Bonfante
    Reviewer; Università di Torino, Italy
  5. Erik Hom
    Reviewer; Unversity of Mississippi, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your work entitled "Algal-fungal symbiosis leads to photosynthetic mycelium" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by a Reviewing Editor and a Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Paola Bonfante (Reviewer #2); Erik Hom (Reviewer #3).

As you will see from the reviews, the reviewers were excited by your discovery of a novel interaction between Nannochloropsis and Mortierella but identified several weaknesses with the experimentation. This led to the consensus decision that the data are not currently strong enough to conclude that these two organisms develop an endosymbiosis. At eLife, we invite revision only if additional work required is fairly minor. Unfortunately, this is not the case here. However, we are excited by the topic and if you are able address the criticisms and include additional data to substantiate the claims of endosymbiosis, we would be pleased to consider reviewing a new manuscript (which would be considered a new submission).

Reviewer #1:

The authors make an interesting discovery that, when grown together, the alga, Nannochloropsis oceanica and the fungus, Mortierella elongata, appear to develop a consortium. The authors provide some evidence for the movement of carbon and nitrogen between symbionts, some evidence of specificity of the interaction, in that it doesn't occur with just any fungus, and also evidence for the growth of N. oceanica within the M. elongate hyphae, although it is not so clear that the hyphae are still living. The data are intriguing, and the system has potential to provide insight into the initiation of new symbioses and how mutualisms develop, as well as possible applications. However, currently, I consider that the data are not strong enough to support the broad claims that are made and the evidence for endosymbiosis is particularly weak. In several cases, essential controls are missing, and strong claims are made without sufficient experimental support.

1) Interactions of Noc with diverse fungi were assessed on the basis of the appearance of aggregates and as aggregates were observed only for this pair of organisms, these data certainly argue for something unique that occurs during the co-cultivation of these two species. However, quantitative data to show growth of the two organisms over time are lacking, even though it is claimed that they can be co-cultivated together for extended periods and that they benefit from the co-cultivation. There is some evidence that N. oceanica shows a small growth benefit during co-culture in -C or -N medium but not -P. However, the baseline viability percentage is different in these experiments (very high in the -P expt), so it is unclear whether the lack of difference in -P is valid. Additionally, the number of cells that were counted was 5 times less in the P experiment. I was unable to find any data to support the claims that Mortierella 'grows in PBS by incorporating algal-derived metabolites' (claimed in subsection “Nutrient-de1ciency and bene1ts of co-cultivation for N. oceanica and M. elonga”). In fact, growth data for Mortierella appears to be lacking entirely.

The observations are interesting, but growth data are essential (and preferably in different nutritional conditions) to substantiate the claims of reciprocal nutrient exchange resulting in a mutualism.

2) The idea that the Noc might be growing inside the hyphae is intriguing. The low power light microscope images certainly show instances of algal cells inside the hyphae. However, the data do not convince me that those hyphae are alive. Mainly because in the EM images, the hyphae that clearly contain Noc are empty of organelles. Additionally, the data to support the claim that the Noc is surrounded by a host cell membrane (Figure 4) is very weak. Chemical fixation approaches do not provide the level of resolution needed to support claims of novel membranes. It is very difficult to visualize the membranes in these images and high pressure freezing should be used. Unfortunately, the videos do not provide clear view of Noc inside living hyphae. If they are indeed living together, with Noc inside the Mortierella hyphae, it should be possible to obtain images of hyphae in which there is clear evidence of cytoplasmic streaming along with living algal cells.

The internalization is interesting but needs additional data to substantiate the claims that both of the organisms are indeed alive and growing. If as the authors comment, that cultures can be fragmented and maintained for many months, then it shouldn't be difficult to demonstrate that the mycelium is alive and that the organisms are co-existing.

Additional points

1) As a general comment, some of the experiments are not clearly described. For example, in Figure 1, aggregation and co-occurrence of the algal and fungal cells was observed. It is stated that the interface is reminiscent of lichens but the features on which this conclusion is based are not described. Do Nostoc cells ever aggregate when grown alone in f/2 medium? Are the surface structures on the Nostoc cells present only when grown in co-culture with Mortierella or is this a general feature of Nostoc cells. At a minimum an image of a Nostoc cells grown in the absence of Mortierella should be included. What is the composition of f/2?

2) Figure 2—figure supplement 1. The PCR tests for contamination are a very good idea but positive controls are missing. Do the 3 hyphal samples show a positive PCR reaction with a fungal gene and likewise the same for the Noc?

3) In the labeling experiments, is the radioactivity normalized with respect to dry weight of the organism or with respect to the dry wt of the molecule being measured (eg protein or lipid or carbohydrate). This is important because the proportions of these macromolecules likely differ in the two organisms.

4) Are there differences in the extent of aggregation in nutrient deficient conditions?

5) It is stated that they can be co-cultivations long term. How long?

Reviewer #2:

The manuscript by Zhi-Yan Du and colleagues describes a novel interaction established between a biofuel-producing alga (Nannochloropsis oceanica) and the fungus Mortierella elongata. The authors offer a detailed description of the association (starting from the algal aggregation eventually leading to algal internalisation) as well as some functional characterization (including the nutrient exchange between the partners). The authors suggest that their discovery may have relevant biotechnological applications, since both the microbes are important for lipid and fuel production.

The manuscript is sure of interest and novel. It is exciting to see that basal fungi like Mucoromycota may interact not only with most of plant lineages, but also with algae. There are however many weak points in the experimental approaches, in the quality of the electron-microscope pictures as well as in the use of symbiosis-related terminology.

As a general comment (and the authors can or cannot agree) the comparison with lichens is sometimes misleading: the organisation of a lichen is very diverse, with small hyphae which surround algal cells, and very often these hyphae penetrate inside algal cells producing pegs, or intracellular haustoria (see for example Honegger, 1986). The system here described is fully diverse (Noc is very small in diameter, when compared to Mortierella hyphae), and recalls other interactions where algae, as eukaryotic endosymbionts proliferate inside heterotrophic protists like Paramecium, or Hydra. The result of these interactions is a photosynthetic association, where no pseudotissues are produced (differently from lichens). The authors could give a look at Angela Douglas work (2009) or to the extensive review by Nowack and Melkonian, (2010), where these symbioses are illustrated. The main conclusion is that the first function performed by eukaryotic endosymbionts when are involved in stable interactions with living protists is photosynthesis. This general concept could also help the authors to better characterize the functionality of the association they describe (a green-photosynthetic mycelium).

Essential revisions:

- The quality of the pictures revealing the two partner interaction is not fully satisfying. Figure 1 panels C,E,D are very poor. Both hyphae and algal cells seem to be collapsed with material present at the surface which is very difficult to interpret (see for example the surface of the Noc in the panel D). Is this the result of a preparation artefact? What about a control algal cell, which is maintained in the absence of the fungus? is its surface smooth? Are the warts/projections produced in the presence of the fungus? In my opinion a control experiment is missing. N oceanica is usually described with a smooth surface. Also, the legend has to be carefully checked. If the authors say that Noc cells are captured, this implies an active mechanism by the fungus (as for fungi which trap nematodes.). By contrast, the experiments suggest an aggregation (see below).

- Aggregation of algal cells. There are reports demonstrating that N.oceanica can easily aggregate in the presence of bacterial strains or of bio flocculants (Wang et al., 2012). In some cases, the active molecules which act as bioflocculants have been identified (Wan et al., 2012) as proteoglycans.

Since genomics and metabolomics data are available for Mortierella elongata, can the Authors provide some experimental support to the aggregation they describe? Can they provide a time course experiment? The author could treat Nocs with Mortierella exudates just to see whether the aggregation occurs.

- Nutrient transfer experiments are well developed and accurately described. The supplemental material provides many interesting details. However, some aspects are not fully clear. Nutrient experiments: Figure 2 is not easy to interpret. If I well understand Figure 2 A (left), only a reduced quantity of labelled carbon is moving to the fungus (less than 1 radioactivity dpm/mg), while in the text this is described as a relevant quantity. Which is the comparison term to define "relevant" the radioactivity value? Is the difference between the labelled glucose found in attached vs free Noc cells significantly different? In panel C, the same experiment is repeated, but the radioactivity value in the fungus is much higher (12,7%). In addition, why N experiments are represented in a different way?

- Mechanisms underlying the carbon transfer. The results clearly demonstrate that carbon is moving from the alga to the fungus. Does Mortierella genome give some suggestion on the underlying mechanism? Presence of glucose transporter? On the other hand, the results showing a moving of C from the fungus to the alga even in the absence of a physical contact are very difficult to understand. Noc is photosynthetic: have the environmental conditions an impact on its photosynthetic activity? have the Authors checked some photosynthetic parameters under these conditions?

- Nutrient conditions: Noc responds to short and long term N starvation activating specific molecular responses (Dong et al., 2012). Have the Authors checked these recovery mechanisms when Noc cells are maintained in the presence of Mortierella?

- Specificity experiment. This experiment is very interesting, also thinking of the reports which show how bacteria can aggregate the small Noc cells (see previous comment). In my opinion, a couple of information is missing: which is the behaviour of the original Mortierella strain, the strain which contains Mycoavidus? have the Authors checked some lichenised fungi? It would be interesting to see the behaviour of Rhizopus, a related Mucoromycota which hosts B. rhizoxinica. This bacterium enters inside Rhizopus hyphae following modalities which recall those described for Mortierella-Noc, using chitinase to degrade the fungal wall at the tip (Moebius et al., 2014).

In addition, I would comment the negative results on Saccharomyces. Hom and Murray, 2014 clearly demonstrate that the interactions between the yeast and the alga depends on the environmental conditions! the environment (for example, nutrient starvation) is the driving force for the mutualistic association. In this context, the authors should also consider the very interesting results illustrated in Li Chien et al., 2017, where a synthetic platform is developed by associating different yeasts to photosynthetic cyanobacteria.

- Subsection “Long-term co-cultivation leads to internalization of N. oceanica within M. elongata hyphae”: WGA is a lectin expected to bind to N-acetylglucosamine, the monomer of chitin. Does Noc contain this fungal-wall component?

Looking at the confocal pictures, (Figure 4—Figure supplement 1) no doubt that the WGA is staining the fungal walls, but the algal walls are not labelled (Figure 4—figure supplement 1, panel B), only the division septum between two dividing cells shows a fluorescence. By contrast, the red chlorophyll fluorescence clearly allows to identify both the isolated algal cells and those among the hyphae. In the last panel (on the right) the chloroplast is clearly seen inside the hypha, suggesting the algal internalisation. I would suggest to re-write the description.

Figure 4 has probably to be reorganized in order to allow an easier reading: First the DIC images clearly showing the algal cluster inside at the tip of the hypha, and then a couple of TEM images, selecting the best: I would suggest Figure 4—figure supplement 2 the first two pictures from the left, since the fungus seems to be alive in all the other images, the hyphae are empty: no organelle, no membranes, suggesting that the algae are entering in an empty niche…By contrast in this Figure 4—figure supplement 2 picture, it seems that some fungal membranes are present. And then the magnification of Figure 4 panel C to show the ultrastructure of Noc. However, the pictures do not solve the question whether Noc cells are surrounded by the fungal membrane. Again, the Authors can check their images with algae living inside unicellular protists. For example, the beautiful pictures from Song et al., (2017).

Reviewer #3:

This manuscript documents a very exciting finding: the endosymbiosis of Nannochloropsis algal cells by a Mortierella fungus in the Mucoromycota, a phylum that is coming under greater scrutiny in relation to the evolution of plant-fungal associations. The text is well written, and the experiments described are compelling and demonstrate the intracellular association of the alga (via light and electron microcopy) and the exchange of carbon and nitrogen between the alga and fungus. While the mechanisms underlying the (vertical?) transmission and maintenance of this new co-culture 'induced' symbiosis are not explored, this work establishes the basis for further research.

A few concerns/questions:

1) In the Abstract, it is written: "This symbiosis begins with chemotactic attraction…". It's not clear to me what evidence there is for this claim. Is this statement based on the aggregation of algal cells at the tips of fungal hyphae? Can one safely conclude that there was chemotactic attraction based on this?

2) Is there any series of images or a movie that be provided that shows how the dense clustering of the algal cells at hyphal tips changes or progresses over time (Discussion section)? Perhaps from lower density to higher density?

3) Video 4 is referenced (subsection “Long-term co-cultivation leads to internalization of N. oceanica within M. elongata hyphae”) as providing evidence that both algal and fungal cells can be passaged through fragmentation and remain viable. This video shows a focal hypha with endosymbiotic algae (within the context of a larger "green tissue") time-lapsed over 6 days (not months of co-culture) but does not show any fragmentation and/or passaging. Was the wrong file uploaded? Is there another video or set of figures/images that can actually be used to support the stated claim?

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for submitting your article "Algal-fungal symbiosis leads to photosynthetic mycelium" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a member of our Board of Reviewing Editors, and the evaluation has been overseen by a Reviewing Editor and Ian Baldwin as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Maria J Harrison (Reviewer #1); Paola Bonfante (Reviewer #2); Erik Hom (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

The authors have made an interesting discovery that the alga, Nannochloropsis oceanica and the fungus, Mortierella elongata, are capable of living together in a mutualistic association which includes growth of Nannochloropsis within the Mortierella hyphae. This is a unique study has potential to provide insights into the emergence of mutualisms and possibly endosymbiosis. All three reviewers appreciate the authors' efforts to add new data and agree that the main claims are now well supported. The revisions outlined below are essential but should not be too onerous as they are largely text edits to ensure that all claims are fully supported by the data presented and that the assumptions being made are clearly stated.

Essential revisions:

Figure 4G. The authors have tempered their comments about the fungal membranes, and we appreciate this. However, the TEM Pictures (Figure 4) do not fully solve the question: are the algal cell surrounded by a fungal-derived membrane? We feel that this cannot be concluded from the data presented. In many cases, it is apparent that the Mortierella hyphae are empty and the fact that the fungus grows when put in a new plate simply means that some hyphae are alive and capable of re-starting their growth. But these are not necessarily the hyphae with Noc in them. So, unless you have data to show this, we request that you delete the sentence that says that 'putative fungal membranes surround internalized algal cells.'

We really appreciate the new scanning images showing the Nannochloropsis cell wall structure. And indeed, they reveal that the alga alone has a smooth surface, as reported in literature. However, in order to conclude that the extensions exist underneath the smooth outer coat, transmission electron microscopy of cross sections are needed. It is possible that the extensions have been elicited by the contact with the fungus. Consequently, we request that the conclusions be modified and that this point is discussed (unless you can include cross sections that show the extensions below the coat).

Abstract: The first sentence does not make sense (…the coevolution of land plants and lichens). The abstract has been corrected following previous suggestion in some documents, but not in the PDF "merged new version". In addition, I believe that the term coevolution requires two members. (for example coevolution between plants and fungi, between fungi and insects.). In conclusion: the sentence has to be re-written.

Discussion section: hyphal tips are among the least developed tissues. Please note that tissue means: groups of cells that have a similar structure and origin and act together to perform a specific function. Therefore, the term tissue cannot be use for a part of a cell (hypha in this case)..In this context I would write: the hyphal tips are the least differentiated portions of a mycelial network

Subsection “Nutrient-de1ciency and bene1ts of co-cultivation for N. oceanica and M. elongate”: Somewhere here, we think it's proper to acknowledge that the linolenic acid marker for biomass was standardized under replete, monoculture conditions. The authors' have not ruled out the possibility that marker correspondence with biomass may break down under the conditions of co-culture, especially since lipid compositional remodeling is certainly possible and not unprecedented in symbioses (e.g., between plants and arbuscular mycorhizal fungi). We think it right to add a simple, honest sentence that states that insignificant changes in C18:3 vs.biomass are assumed for any physiological changes that might be experienced by M. elongata in coculture. (I think it probably is insignificant, but one should be precise and not speculate.) This applies to Figure 3—figure supplement 2 as well.

The nutrient exchange experiments are nicely performed and presented, but a discussion of their biological meaning is missing. M.elongata is a strong saprotroph and does not need the carbon coming from the alga. On the other hand, it also releases N to the alga. So, what is the benefit for the fungus? The only benefit seems to be the increased fungal biomass; however, this is contingent on the point noted above. The C18:3 quantification could mirror a different metabolism (lipid store) more than an active growth. Please add a point of discussion including the C and N conditions of the media in which these experiments were performed.

Subsection “Long-term co-cultivation leads to internalization of N. oceanica within M. elongata hyphae”: for algae that are inside live fungi, how can you be sure they are really alive based only on these data? Free/extracellular algae might be accessible to SYTOX Green, but algae in live fungi may never be exposed to SYTOX Green, yes? (The live fungi would exclude it, so one may not be able to tell intracellular viability of the algae.) We do not feel this is a major point because the EM data indicate that algae are alive, but it might not be possible to determine true viability of intracellular algae with exclusionary dyes and this should be acknowledged.

Abstract – Long term co-cultivation is a subjective term. The length of time observed should be inserted here.

Figure 3—figure supplement 2. Again, details of the statistical tests used should be added.

Figure 4 G -What is Mo? It is missing from the legend.

Subsection “Long-term co-cultivation leads to internalization of N. oceanica within M. elongata hyphae”: We suggest rewording to be more conservative in claims: "The results is consistent with the notion that both fungal host and algae inside are alive (Figure 4—figure supplement 5), although DIC microscopy…surrounded by fungal organelles and what appear to be lipid droplets…"

Changed "is consistent with" for "showed", added "although" before "DIC microscopy, removed "the" before "fungal organelles", and changed "and what appear to be" for "such as". Lipid droplets are not fungal organelles…. As a side note: how does one know that these are lipid droplets? Do they stain with Nile Red? This is stated but not justified.

Discussion section: We suggest rewording to be more conservative in claim: "green hyphae appear to remain alive after 2-months…". Changed "appear to remain" for "remained".

Please insert the nature of the statistical tests used in the legends and/or in the Materials and methods section.

https://doi.org/10.7554/eLife.47815.030

Author response

We first want to thank you for your constructive reviews of our manuscript and say that we found your comments valuable. Following your advice, we carried out many additional experiments that were suggested, including a series of new scanning electron microscopy experiments to detail the cell wall interaction between the two cell types, PCR controls for isolation of algae from fungi, photosynthesis/chlorophyll assays of cocultured algae, growth/biomass assays of co-cultured algae and fungi, light and confocal microscopy and viability assays of both algae and fungi in green hyphae. We have prepared new figures and revised all of the figures for clarity, including new transmission electron micrographs showing the endo-microalgae surrounded by fungal membranes and vesicles. We feel your suggestions and these additions have greatly improved our new submission and hope that you and the reviewers concur.

Reviewer #1:

The authors make an interesting discovery that, when grown together, the alga, Nannochloropsis oceanica and the fungus, Mortierella elongata, appear to develop a consortium. The authors provide some evidence for the movement of carbon and nitrogen between symbionts, some evidence of specificity of the interaction, in that it doesn't occur with just any fungus, and also evidence for the growth of N. oceanica within the M. elongate hyphae, although it is not so clear that the hyphae are still living. The data are intriguing, and the system has potential to provide insight into the initiation of new symbioses and how mutualisms develop, as well as possible applications. However, currently, I consider that the data are not strong enough to support the broad claims that are made and the evidence for endosymbiosis is particularly weak. In several cases, essential controls are missing, and strong claims are made without sufficient experimental support.

We have carried out additional experiments to support the findings and we have toned down strong language.

1) Interactions of Noc with diverse fungi were assessed on the basis of the appearance of aggregates and as aggregates were observed only for this pair of organisms, these data certainly argue for something unique that occurs during the co-cultivation of these two species. However, quantitative data to show growth of the two organisms over time are lacking, even though it is claimed that they can be co-cultivated together for extended periods and that they benefit from the co-cultivation. There is some evidence that N. oceanica shows a small growth benefit during co-culture in -C or -N medium but not -P. However, the baseline viability percentage is different in these experiments (very high in the -P expt), so it is unclear whether the lack of difference in -P is valid. Additionally, the number of cells that were counted was 5 times less in the P experiment. I was unable to find any data to support the claims that Mortierella 'grows in PBS by incorporating algal-derived metabolites' (claimed in subsection “Nutrient-de1ciency and bene1ts of co-cultivation for N. oceanica and M. elonga”). In fact, growth data for Mortierella appears to be lacking entirely.

The observations are interesting, but growth data are essential (and preferably in different nutritional conditions) to substantiate the claims of reciprocal nutrient exchange resulting in a mutualism.

Because P was not shown to be important in the interaction and is not a major part of this research we have removed the result from the manuscript and prepared a new Figure 3 and new Figure 3—figure supplement 1.

Quantification of fungal biomass is technically challenging. We did perform time-lapse microscopy and we found that the fungal hyphae kept growing in PBS buffer in the presence of the algae partner, whereas the fungi alone in PBS didn’t show any obvious new growth. We provide a side-by-side video to show the difference (Video 1).

In this new submission we have made great efforts to quantify the growth of the algae and fungi during co-cultivation. We have found that Mortierella elongata has major unique fatty acids such as the α-linolenic acid (C18:3) that are only a minor component in Nannochloropsis. Therefore, it can serve as a biomarker to assess fungal abundance. We onserved that the ratio of C18:3 to biomass in the Mortierella fungus is very consistent following the incubation in f/2-N medium (Figure 3—figure supplement 2A). Thus, we used C18:3 as a marker to quantify the fungal biomass within algae-fungi aggregates. We measured the C18:3 content and total biomass of algae-fungi aggregates. We did observe increase in the biomass of both algae and fungi following co-cultivation in f/2-N medium (nutrient-deprived medium for both fungi and algae), which was not observed when partners were incubated alone in the same medium (Figure 3—figure supplement 2B).

2) The idea that the Noc might be growing inside the hyphae is intriguing. The low power light microscope images certainly show instances of algal cells inside the hyphae. However, the data do not convince me that those hyphae are alive. Mainly because in the EM images, the hyphae that clearly contain Noc are empty of organelles. Additionally, the data to support the claim that the Noc is surrounded by a host cell membrane (Figure 4) is very weak. Chemical fixation approaches do not provide the level of resolution needed to support claims of novel membranes. It is very difficult to visualize the membranes in these images and high pressure freezing should be used. Unfortunately, the videos do not provide clear view of Noc inside living hyphae. If they are indeed living together, with Noc inside the Mortierella hyphae, it should be possible to obtain images of hyphae in which there is clear evidence of cytoplasmic streaming along with living algal cells.

The internalization is interesting but needs additional data to substantiate the claims that both of the organisms are indeed alive and growing. If as the authors comment, that cultures can be fragmented and maintained for many months, then it shouldn't be difficult to demonstrate that the mycelium is alive and that the organisms are co-existing.

We understand the reviewers’ concerns. To check whether the algae cells are living inside the fungal hyphae and whether the fungal hypha are themselves living (have cellular contents), we have performed additional microscopy including DIC light, confocal, and electron microscopy. Our new data demonstrate that algal cells are alive within living fungal cells.

In this revision, we have replaced the TEM images in Figure 4 with one showing that intact algal cells are inside the fungal hypha, and they are surrounded by fungal cytoplasm including organelles (AG77 and Noc, Figure 4F and G; NVP64 and Noc, Figure 4—figure supplement 2C).

Confocal SYTOX green microscopy images of Figure 4—figure supplement 5 A-C demonstrate that both internalized algae and host cells are alive. The SYTOX probe enters dead cells highlighting them and the signal is obvious when dead cells are used as controls (Figure 4—figure supplement 5D to H). Several dead algae cells were observed in the aggregates, but none were observed inside the fungal hyphae. We then used DIC light microscope to check whether the fungal hyphae were hollow, or whether they had cytoplasm and organelles. We have observed that the fungal hyphae contain cytoplasm and the algal cells within fungal hyphae are surrounded by fungal vesicles, including lipid droplets, which are present in living cells (Figure 4—figure supplement 6). We recorded real-time videos to show the 3D content of the green hyphae and Video 5 is a representative one. We also note that cytoplasmic streaming is common in young actively growing cells in low carbon f/2 medium, but after 2-3 weeks cytoplasmic streaming is not evident in older parts of the mycelial network. Hence, it is not unexpected that no cytoplasmic streaming was observed in the mature fungal hyphae over one-month of co-cultivation.

Essential revisions:

1) As a general comment, some of the experiments are not clearly described. For example, in Figure 1, aggregation and co-occurrence of the algal and fungal cells was observed. It is stated that the interface is reminiscent of lichens but the features on which this conclusion is based are not described. Do Nostoc cells ever aggregate when grown alone in f/2 medium? Are the surface structures on the Nostoc cells present only when grown in co-culture with Mortierella or is this a general feature of Nostoc cells. At a minimum an image of a Nostoc cells grown in the absence of Mortierella should be included. What is the composition of f/2?

We thank the reviewer for these comments. We have revised the manuscript accordingly to provide more detailed descriptions. To clarify, we are working with Nannochloropsis (Noc), a unicellular microalga rather than the cyanobacterium Nostoc. We demonstrate through additional controlled experiments that the morphology and aggregation of Nannochloropsis (Noc) cells differ significantly between mono-culture and co-culture with living Mortierella. Following prolonged incubation after stationary phase in f/2 medium Nannochloropsis (Noc) cells can aggregate when grown alone, but in comparatively small clusters (up to dozens of cells) (Figure 1—figure supplement 1A). Through scanning electron microscopy (SEM) we demonstrate that Noc cells incubated as a monoculture have a smooth outer cell wall surface (Figure 1D; Figure 1-S1A), whereas Noc cells co-cultured with living Mortierella cells exhibit a fibrous extension from their cell walls (Figure 1C, E and F; Figure 1—figure supplement 1D). Specifically, the thin outer layer breaks away from Noc cells when incubated with Mortierella, and the fibrous extensions beneath become exposed (Figure 1—figure supplements1D). Large pieces of the broken membrane are evident in Noc cells (yellow arrows, Figure 1—figure supplement 1D), while the others have smaller residual fragments (Figure 1E and F). We also examined the free Noc cells in the supernatant of Noc-Mortierella co-culture, and we found the cells have a partially damaged cover membrane (not as smooth as in the control) but the fibrous extensions were not exposed (Figure 1—figure supplement 1C).

As in control algal mono-cultures, the thin outer membrane covering Noc cells is also evident when co-cultured with non-interactive fungi (Clavulina sp. PMI390 and Morchella americana GB760 – Figure 3—figure supplement 3 and Figure 3—figure supplement 4) or with dead Mortierella cells (Figure 1—figure supplement 1B). The fibrous extensions seem to be important in the physical interaction of Noc and AG77 cells. We have included these new findings in the manuscript. We have also included the composition of f/2 medium in the method section, which is a common and widely used general enriched medium for growing marine algae.

2) Figure 2—figure supplement 1. The PCR tests for contamination are a very good idea but positive controls are missing. Do the 3 hyphal samples show a positive PCR reaction with a fungal gene and likewise the same for the Noc?

We have performed PCR tests with positive and negative PCR controls. The three hyphal samples showed a positive PCR reaction for the fungal marker gene EF1a, and the two algal samples had a positive amplification of the algal gene AURE04. There was no PCR amplification for either negative control reactions (Figure 2—figure supplement 1D and E).

3) In the labeling experiments, is the radioactivity normalized with respect to dry weight of the organism or with respect to the dry wt of the molecule being measured (eg protein or lipid or carbohydrate). This is important because the proportions of these macromolecules likely differ in the two organisms.

For the labelling experiments we used total dry weight of the cells. We now have clear description of this in the Method section and clarify this in the figure legend.

4) Are there differences in the extent of aggregation in nutrient deficient conditions?

We did not see any obvious difference in the Noc-AG77 aggregates following nutrient deficient conditions.

5) It is stated that they can be co-cultivations long term. How long?

We define long-term co-culture as lasting between one to three months. Both Noc and AG77 are fast growers. Noc cells can double within 48 hours and AG77 can fill a plate or extend across a flask within 2-3 days after inoculation.

Reviewer #2:

The manuscript by Zhi-Yan Du and colleagues describes a novel interaction established between a biofuel-producing alga (Nannochloropsis oceanica) and the fungus Mortierella elongata. The authors offer a detailed description of the association (starting from the algal aggregation eventually leading to algal internalisation) as well as some functional characterization (including the nutrient exchange between the partners). The authors suggest that their discovery may have relevant biotechnological applications, since both the microbes are important for lipid and fuel production.

1) The manuscript is sure of interest and novel. It is exciting to see that basal fungi like Mucoromycota may interact not only with most of plant lineages, but also with algae. There are however many weak points in the experimental approaches, in the quality of the electron-microscope pictures as well as in the use of symbiosis-related terminology.

We have improved our electron-micrographs, and reassessed terminology following the reviewer’s comments.

2) As a general comment (and the Authors can or cannot agree) the comparison with lichens is sometimes misleading: the organisation of a lichen is very diverse, with small hyphae which surround algal cells, and very often these hyphae penetrate inside algal cells producing pegs, or intracellular haustoria (see for example Honegger, 1986). The system here described is fully diverse (Noc is very small in diameter, when compared to Mortierella hyphae), and recalls other interactions where algae, as eukaryotic endosymbionts proliferate inside heterotrophic protists like Paramecium, or Hydra. The result of these interactions is a photosynthetic association, where no pseudotissues are produced (differently from lichens). The authors could give a look at Angela Douglas work (2009) or to the extensive review by Nowack and Melkonian, (2010), where these symbioses are illustrated. The main conclusion is that the first function performed by eukaryotic endosymbionts when are involved in stable interactions with living protists is photosynthesis. This general concept could also help the authors to better characterize the functionality of the association they describe (a green-photosynthetic mycelium).

We thank the reviewer for these suggestions and agree. We have revised and included these in the Discussion section.

Major comments:

3) The quality of the pictures revealing the two partner interaction is not fully satisfying. Figure 1 panels C,E,D are very poor. Both hyphae and algal cells seem to be collapsed with material present at the surface which is very difficult to interpret (see for example the surface of the Noc in the panel D). Is this the result of a preparation artefact? What about a control algal cell, which is maintained in the absence of the fungus? is its surface smooth? Are the warts/projections produced in the presence of the fungus? In my opinion a control experiment is missing. N oceanica is usually described with a smooth surface. Also, the legend has to be carefully checked. If the authors say that Noc cells are captured, this implies an active mechanism by the fungus (as for fungi which trap nematodes.). By contrast, the experiments suggest an aggregation (see below).

We thank the reviewer for these suggestions.

Following these suggestions, we have carried out controlled experiments to assay algal cells grown in mono-culture in f/2 medium, and also with non-living Mortierella and non-compatible fungi. As mentioned above, we found that these Noc cells alone have a smooth surface (Figure 1D; Figure 1—figure supplement 1A) when grown in mono-culture or with non-living Mortierella or non-compatible fungi (e.g. Morchella, Clavulina spp.). Fibrous extensions were only observed in the Noc cells co-cultured with live Mortierella cells (Figure 1C, E and F; Figure 1—figure supplement 1D). The fibrous extensions are covered by a thin layer of relatively smooth membrane. When incubated with Mortierella this layer breaks up exposing the fibrous extensions below (Figure 1—figure supplement 1D). Some Noc cells still have big pieces of the broken membrane (yellow arrows, Figure 1—figure supplement 1D), while the others have small residues (Figure 1E and F).

Scholz et al., 2014 performed super high resolution Cryo-EM and they suggested that the fibrous extensions were on the outer layer of Nannochloropsis gaditana, a related algal species. However, according to our data, it appears the fibrous extensions are exposed after the cover membrane is broken. In fact, the cover membrane is also visible in the Cryo-EM images by Scholz et al., 2014. The authors may have missed the membrane as a part of the Noc cell wall because of how they processed and prepared their samples (Figure 2B and C in Scholz et al., 2014). Further research on N. gaditana also showed that the cells have a smooth surface by SEM (Jazzar et al., 2015).

We agree with the reviewer’s comments on the legend and have also changed the terminology as an ‘aggregation’ rather than captured cells.

4) Aggregation of algal cells. There are reports demonstrating that N.oceanica can easily aggregate in the presence of bacterial strains or of bio flocculants (Wang et al., 2012). In some cases, the active molecules which act as bioflocculants have been identified (Wan et al., 2012) as proteoglycans.

Since genomics and metabolomics data are available for Mortierella elongata, can the Authors provide some experimental support to the aggregation they describe? Can they provide a time course experiment? The author could treat Nocs with Mortierella exudates just to see whether the aggregation occurs.

We thank the reviewer for these suggestions. We think the important findings in this manuscript are the details of the physical cell wall interaction and mutualism between the algae and fungi, as well as long-term co-culture that led to the formation of green hyphae. In terms of the algae flocculation by fungi, we have recently published on the ability of Mortierella to bioflocculate Nannochloropsis as an efficient, cost effective, and environment friendly approach to harvest microalgae compared to conventional harvesting methods such as centrifuge and chemical flocculation. We provide data including a short time course flocculation and lipid/fatty acid profiling results, as well as other oil productivity data for biotechnological purposes (Du et al., 2018). No aspect of nutrient exchange or green hyphae observed after longer term incubations were discussed in the Biotechnology for Biofuels paper, as these data are the substance of this submission. Determining the composition of the algal outer wall, and the means by which Mortierella affects this, is part of a larger ongoing JGI supported project, and beyond the scope of the current study.

5) Nutrient transfer experiments are well developed and accurately described. The supplemental material provides many interesting details. However, some aspects are not fully clear. Nutrient experiments: Figure 2 is not easy to interpret. If I well understand Figure 2 A (left), only a reduced quantity of labelled carbon is moving to the fungus (less than 1 radioactivity dpm/mg), while in the text this is described as a relevant quantity. Which is the comparison term to define "relevant" the radioactivity value? Is the difference between the labelled glucose found in attached vs free Noc cells significantly different? In panel C, the same experiment is repeated, but the radioactivity value in the fungus is much higher (12,7%). In addition, why N experiments are represented in a different way?

First of all, we thank you for your comment and have tried to more clearly describe the presented data and figures. One key point is that the majority of transferred 14C-carbon ended up in the lipid fraction of the fungal cells. The later relative values in Figure 2 panels C and D, are the total transferred 14C-carbon in different kinds of samples such as non-labelled fungi with physical contact to labelled algae (Figure 2 panel C #1, 12.7%) and with physical contact to heat-killed-labelled algae (Figure 2 panel C #3, 1.3%) compared to the total radioactivity of prelabelled samples such as the 14C-labelled algae (Figure 2 panel C, 100%). These analyses were aimed to test whether physical contact and live cells are essential to the 14C-carbon transfer. We have revised the text content and the figure to make it easier to understand. We add “100%” after the 14C-Noc and 14C-AG77 in Figure 2 panels C and D.

Compared to the carbon exchange experiments that use 14C-carbon that can be measured with a scintillation counter and readily presented, the nitrogen exchange experiment was conducted using 15N, a stable, non-radioactive isotope of nitrogen in combination isotope ratio mass spectrometry to calculate [15N/(15N+ 14N), mol/mol]. The latter experiments are admittedly more complicated than the carbon labelling results (Figure 2—figure supplement 3J and K). To make these results better understandable to the general audience, we summarize the results in a simple figure (Figure 2 panel E), and provide more details in the supplementary files.

6) Mechanisms underlying the carbon transfer. The results clearly demonstrate that carbon is moving from the alga to the fungus. Does Mortierella genome give some suggestion on the underlying mechanism? Presence of glucose transporter? On the other hand, the results showing a moving of C from the fungus to the alga even in the absence of a physical contact are very difficult to understand. Noc is photosynthetic: have the environmental conditions an impact on its photosynthetic activity? have the Authors checked some photosynthetic parameters under these conditions?

We agree these are interesting and obvious remaining questions, but decided that to address the mechanism of carbon transport should be the next step requiring additional metabolic, transcriptomic experiments, and a transformation system for the fungus in place. We believe these experiments go beyond the current manuscript, which already contains a large amount of original data.

Regarding C transfer from fungus to algae we report that Mortierella fungi release organic carbon and nitrogen to the environment/medium when they are incubated alone in the f/2 medium (Figure 3D and E). This would allow the unattached algae to receive carbon and nitrogen from the co-cultured fungi (Figure 2).

Environmental conditions are sure to impact photosynthesis. We analyzed the chlorophyll content that can be extracted using acetone:DMSO solvents, as this can be used as a proxy of photosynthetic potential. The chlorophyll content decreases when Noc cells are stressed. We did see a significantly higher chlorophyll content in the Noc cells co-cultured with AG77 fungi than in the Noc cells alone serving as control after 10-days-prolonged incubation in f/2 medium, indicating that the Noc cells with Mortierella fungi have higher photosynthetic potential (Figure 3—figure supplement 1A).

7) Nutrient conditions: Noc responds to short and long term N starvation activating specific molecular responses (Dong et al., 2012). Have the Authors checked these recovery mechanisms when Noc cells are maintained in the presence of Mortierella?

We observed increased viability of Noc cells following long term (10 days) N starvation in the presence of Mortierella fungi (Figure 3A to C), which release organic carbon and nitrogen to the medium (Figure 3D and E). The Noc cells without Mortierella were significantly more stressed, evident by chlorophyll degradation following prolonged incubation (Figure 3—figure supplement 1A).

8) Specificity experiment. This experiment is very interesting, also thinking of the reports which show how bacteria can aggregate the small Noc cells (see previous comment). In my opinion, a couple of information is missing: which is the behaviour of the original Mortierella strain, the strain which contains Mycoavidus? have the Authors checked some lichenised fungi? It would be interesting to see the behaviour of Rhizopus, a related Mucoromycota which hosts B. rhizoxinica. This bacterium enters inside Rhizopus hyphae following modalities which recall those described for Mortierella-Noc, using chitinase to degrade the fungal wall at the tip (Moebius et al., 2014).

In addition, I would comment the negative results on Saccharomyces. Hom and Murray, 2014 clearly demonstrate that the interactions between the yeast and the alga depends on the environmental conditions! the environment (for example, nutrient starvation) is the driving force for the mutualistic association. In this context, the authors should also consider the very interesting results illustrated in Li Chien et al., 2017, where a synthetic platform is developed by associating different yeasts to photosynthetic cyanobacteria.

We thank the reviewer for these suggestions. We have tested the wild-type Mortierella elongata AG77, which carries the endobacterium Mycoavidus cysteinexigens. We observed significantly reduced viability of Noc cells with the WT AG77 compared to the cured AG77, as well as severe chlorophyll degradation and ROS accumulation in the Noc cells when co-cultured with the WT AG77. However, we have repeated the experiments several times and the results were not consistent. This could result from fluctuations in bacterial activity. While intriguing, the presence of the endobacteria makes the symbiosis more complicated and is not the focus of this research. Thus, we removed the WT AG77 data and we have revised the Discussion section. Although we did not screen specifically for Rhizopus or lichen fungi, we did screen a phylo-diverse selection of fungi including relatives of these organisms. We found the phenotype we report here to be unique to Mortierella.

9) Subsection “Long-term co-cultivation leads to internalization of N. oceanica within M. elongata hyphae”: WGA is a lectin expected to bind to N-acetylglucosamine, the monomer of chitin. Does Noc contain this fungal-wall component?

Yes. It has been reported that Noc has the N-acetylglucosamine in the cell wall (Scholz et al., 2014), and we do know that Mortierella is efficient at degrading this substrate.

10) Looking at the confocal pictures, (Figure 4—Figure supplement 1) no doubt that the WGA is staining the fungal walls, but the algal walls are not labelled (Figure 4—figure supplement 1, panel B), only the division septum between two dividing cells shows a fluorescence. By contrast, the red chlorophyll fluorescence clearly allows to identify both the isolated algal cells and those among the hyphae. In the last panel (on the right) the chloroplast is clearly seen inside the hypha, suggesting the algal internalisation. I would suggest to re-write the description.

We thank the reviewer for these suggestions. The Figure 4—figure supplement 1B is not just a regular confocal image focused on a single panel. It’s a stacked confocal image which shows the 3D-view of the WGA and chlorophyll signals. That’s the reason why the WGA signal is all over the cells instead of just a circle. The division septum did show a stronger signal. We lowered the gain of WGA to show the inside chlorophyll signal. Otherwise, they will be bright green cells. We have a video of a rotating 3D-model to show how the confocal images are stacked, as well as for the Figure 4—figure supplement 1C (Video 2). We have rewritten the description as requested.

11) Figure 4 has probably to be reorganized in order to allow an easier reading: First the DIC images clearly showing the algal cluster inside at the tip of the hypha, and then a couple of TEM images, selecting the best: I would suggest Figure 4—figure supplement 2 the first two pictures from the left, since the fungus seems to be alive..in all the other images, the hyphae are empty: no organelle, no membranes, suggesting that the algae are entering in an empty niche…By contrast in this Figure 4—figure supplement 2 picture, it seems that some fungal membranes are present. And then the magnification of Figure 4 panel C to show the ultrastructure of Noc. However, the pictures do not solve the question whether Noc cells are surrounded by the fungal membrane. Again, the authors can check their images with algae living inside unicellular protists. For example, the beautiful pictures from Song et al., (2017).

We thank the reviewer for the suggestions. We first replaced the TEM images in Figure 4 with the ones showing that intact algal cells are inside the fungal hypha and they are surrounded by fungal membranes and organelles (AG77 and Noc, Figure 4F and G; NVP64 and Noc, Figure 4—figure supplement 2C). The previous TEM images showing algal cells inside fungal hypha with clear ultrastructure of chloroplast and some other algal organelles were moved to the Figure 4—figure supplement 2B. The algal cells have unique and distinguishable chloroplast and thylakoid membranes. For the fungal organelles, we include a TEM control of the fungi incubated alone in f/2 medium to compare the algal and fungal organelles (Figure 4—figure supplement 2A). Some fungal mycelia are hollow after the harsh steps of TEM sample preparation as the other images of co-cultured fungal tissue.

Unfortunately, we were unsuccessful with Cyro-EM to get higher resolution images to address whether a fungal membrane surrounds the algal cells. However, we have used DIC light and confocal microscopy with SYTOX green and demonstrated algae cells are living inside living fungal cells. Figure 4—figure supplement 5A to C is a representative example that both the algae and hyphae are alive. The SYTOX probe enters only the dead cells and the signal is obvious in the dead cell controls (Figure 4—figure supplement 5D to H).

We then used DIC light microscopy and have observed that the algal cells inside fungal hyphae are surrounded by fungal vesicles and lipid droplets, which are usually presented in live cells (Figure 4—figure supplement 6). Given that the samples are usually not in the same focal panel at high magnification, we provide videos to show the 3D content of the green hyphae and Video 5 is a representative one.

Reviewer #3:

This manuscript documents a very exciting finding: the endosymbiosis of Nannochloropsis algal cells by a Mortierella fungus in the Mucoromycota, a phylum that is coming under greater scrutiny in relation to the evolution of plant-fungal associations. The text is well written, and the experiments described are compelling and demonstrate the intracellular association of the alga (via light and electron microcopy) and the exchange of carbon and nitrogen between the alga and fungus. While the mechanisms underlying the (vertical?) transmission and maintenance of this new co-culture 'induced' symbiosis are not explored, this work establishes the basis for further research.

Essential revisions:

1) In the Abstract, it is written: "This symbiosis begins with chemotactic attraction…". It's not clear to me what evidence there is for this claim. Is this statement based on the aggregation of algal cells at the tips of fungal hyphae? Can one safely conclude that there was chemotactic attraction based on this?

We have rephrased this. The Nannochloropsis-Mortierella interaction begins with the flocculation of algal cells with the mycelium and the loss of the outer cell wall covering in the algal photobiont.

2) Is there any series of images or a movie that be provided that shows how the dense clustering of the algal cells at hyphal tips changes or progresses over time (Discussion section)? Perhaps from lower density to higher density?

This is a great point. We have spent tremendous efforts to try to address this, but it is a formidable challenge. We can observe the green hyphae in the long-term co-culture and green hyphae are distinguishable under DIC microscope (Figure 4B to E, Figure 4—figure supplement 4). However, to record the progress through time-lapse is extremely difficult. This is due to the fact that the process occurs over weeks, and the focal plane is quite restricted with light microscopy. Given it is a living growing system (3D), we are not able to predict where in the colony these hyphae will develop, and what point of the 3D colony to focus on over days. For instance, Figures 4-S3C to F were taken as time-lapse images over 3 days at a randomly selected region. It took us thousands of hours of microscopy time to observe this. To complicate matters, the algae and fungi form aggregates after co-culture (Figure 4—figure supplement 1D and E), and if the cells are too dense in number, visibility is obstructed. Nonetheless, we were very lucky to record a series of images showing a group of algal cells attached to the hyphal tip growing inside the hyphae (Figure 4—figure supplement 3A and B).

3) Video 4 is referenced (subsection “Long-term co-cultivation leads to internalization of N. oceanica within M. elongata hyphae”) as providing evidence that both algal and fungal cells can be passaged through fragmentation and remain viable. This video shows a focal hypha with endosymbiotic algae (within the context of a larger "green tissue") time-lapsed over 6 days (not months of co-culture) but does not show any fragmentation and/or passaging. Was the wrong file uploaded? Is there another video or set of figures/images that can actually be used to support the stated claim?

The reviewer is correct. Video 4 shows the growth and dividing of algal cells within the fungal hyphae, not fragmentation. However, the co-culture in the video was fragmented (vegetative propagated) from an older mother colony, as is commonly done with fungal mycelium in the lab. We have revised the manuscript to read: “While there is no indication that algae are transmitted vertically through fungal reproductive structures, the algal cells remain viable (growing and dividing) over months of co-culture (Video 4).”

[Editors’ note: The responses to the re-review follow.]

Essential revisions:

Figure 4G. The authors have tempered their comments about the fungal membranes, and we appreciate this. However, the TEM Pictures (Figure 4) do not fully solve the question: are the algal cell surrounded by a fungal-derived membrane? We feel that this cannot be concluded from the data presented. In many cases, it is apparent that the Mortierella hyphae are empty and the fact that the fungus grows when put in a new plate simply means that some hyphae are alive and capable of re-starting their growth. But these are not necessarily the hyphae with Noc in them. So, unless you have data to show this, we request that you delete the sentence that says that 'putative fungal membranes surround internalized algal cells.'

We have deleted the sentence.

We really appreciate the new scanning images showing the Nannochloropsis cell wall structure. And indeed, they reveal that the alga alone has a smooth surface, as reported in literature. However, in order to conclude that the extensions exist underneath the smooth outer coat, transmission electron microscopy of cross sections are needed. It is possible that the extensions have been elicited by the contact with the fungus. Consequently, we request that the conclusions be modified and that this point is discussed (unless you can include cross sections that show the extensions below the coat).

We added an explanation to the Results section as follows: “While it is possible that the fibrous extensions have been elicited by the contact with the fungus, remnant pieces of the outer coat covering the underlying extensions are evident in our observations, as shown in Figure 1E and F and Figure 1—figure supplement 1D. Therefore, it seems likely that these extensions are present underneath the outer smooth layer.”

Abstract: The first sentence does not make sense (…the coevolution of land plants and lichens). The abstract has been corrected following previous suggestion in some documents, but not in the PDF "merged new version". In addition, I believe that the term coevolution requires two members. (for example, coevolution between plants and fungi, between fungi and insects.). In conclusion: the sentence has to be re-written.

We agree with the reviewers and have changed ‘coevolution’ to ‘evolution’.

Discussion section: hyphal tips are among the least developed tissues. Please note that tissue means: groups of cells that have a similar structure and origin and act together to perform a specific function. Therefore, the term tissue cannot be use for a part of a cell (hypha in this case). In this context I would write: the hyphal tips are the least differentiated portions of a mycelial network

We agree, and the sentence has been re-written as suggested.

Subsection “Nutrient-de1ciency and bene1ts of co-cultivation for N. oceanica and M. elongate”: Somewhere here, we think it's proper to acknowledge that the linolenic acid marker for biomass was standardized under replete, monoculture conditions. The authors' have not ruled out the possibility that marker correspondence with biomass may break down under the conditions of co-culture, especially since lipid compositional remodeling is certainly possible and not unprecedented in symbioses (e.g., between plants and arbuscular mycorhizal fungi). We think it right to add a simple, honest sentence that states that insignificant changes in C18:3 vs.biomass are assumed for any physiological changes that might be experienced by M. elongata in coculture. (I think it probably is insignificant, but one should be precise and not speculate.) This applies to Figure 3—figure supplement 2 as well.

We agree, and we have added the discussion as suggested. Subsection “Nutrient -deficiency and benefits of co-cultivation for N. oceanica and M. elongata.”.

The nutrient exchange experiments are nicely performed and presented, but a discussion of their biological meaning is missing. M.elongata is a strong saprotroph and does not need the carbon coming from the alga. On the other hand, it also releases N to the alga. So, what is the benefit for the fungus? The only benefit seems to be the increased fungal biomass; however, this is contingent on the point noted above. The C18:3 quantification could mirror a different metabolism (lipid store) more than an active growth. Please add a point of discussion including the C and N conditions of the media in which these experiments were performed.

The co-culture of nutrient exchange experiments was performed in f/2 medium containing 14N (2.5 mM NaNO3) and 12C (20 mM NaHCO3 and ambient CO2). Thus, the alga should be the only resource of organic carbon for the fungus. Since M. elongata is a strong saprotroph, we did test whether the alga (as well as the fungus) was dead in the co-culture (Figure 2—figure supplement 3) and we also fed the fungus with heat-killed 14C-labeled alga and fed the alga with heat-killed 14C-labeled fungus. These results have been summarized in the Figure 2C and discussed in subsection “Carbon and Nitrogen Transfer between N. oceanica and M. elongata.”.

Subsection “Long-term co-cultivation leads to internalization of N. oceanica within M. elongata hyphae”: for algae that are inside live fungi, how can you be sure they are really alive based only on these data? Free/extracellular algae might be accessible to SYTOX Green, but algae in live fungi may never be exposed to SYTOX Green, yes? (The live fungi would exclude it, so one may not be able to tell intracellular viability of the algae.) We do not feel this is a major point because the EM data indicate that algae are alive, but it might not be possible to determine true viability of intracellular algae with exclusionary dyes and this should be acknowledged.

Excellent point and we acknowledge the short comings of this experiment. Our assumption that the algae inside the fungus are alive are mostly based on the fact that divide and to some extent that they are how chlorophyll fluorescence, which in the long run indicates an assembled photosynthetic membrane.

Abstract – Long term co-cultivation is a subjective term. The length of time observed should be inserted here.

The time has been added: “over a month”.

Figure 3—figure supplement 2 – Again, details of the statistical tests used should be added.

The details of the statistical tests have been added.

Figure 4 G -What is Mo? It is missing from the legend.

Mo stands for ‘Mortierella organelles’. This has been added in the legends of Figure 4G and Figure 4—figure supplement 2.

Subsection “Long-term co-cultivation leads to internalization of N. oceanica within M. elongata hyphae”: We suggest rewording to be more conservative in claims: "The results is consistent with the notion that both fungal host and algae inside are alive (Figure 4—figure supplement 5), although DIC microscopy…surrounded by fungal organelles and what appear to be lipid droplets…"

Changed "is consistent with" for "showed", added "although" before "DIC microscopy, removed "the" before "fungal organelles", and changed "and what appear to be" for "such as". Lipid droplets are not fungal organelles…. As a side note: how does one know that these are lipid droplets? Do they stain with Nile Red? This is stated but not justified.

We have amended the text. We stained and visualized the lipid droplets with BODIPY by confocal microscopy (Figure 4—figure supplement 6). Lipid droplets are distinguishable because of their color (blue green) and size under light microscope.

Discussion section: We suggest rewording to be more conservative in claim: "green hyphae appear to remain alive after 2-months…". Changed "appear to remain" for "remained".

Amended as suggested.

Please insert the nature of the statistical tests used in the legends and/or in the Materials and methods section.

The statistical tests have been added in the legends as suggested.

https://doi.org/10.7554/eLife.47815.031

Article and author information

Author details

  1. Zhi-Yan Du

    1. Department of Energy-Plant Research Laboratory, Michigan State University, East Lansing, United States
    2. Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, United States
    3. Department of Plant Biology, Michigan State University, East Lansing, United States
    Contribution
    Conceptualization, Formal analysis, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-7646-2429
  2. Krzysztof Zienkiewicz

    1. Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, United States
    2. Department of Plant Biochemistry, Albrecht-von-Haller-Institute for Plant Sciences, Georg-August-University, Göttingen, Germany
    3. Centre of Modern Interdisciplinary Technologies, Nicolaus Copernicus University in Toruń, Toruń, Poland
    Contribution
    Data curation, Formal analysis, Investigation, Visualization, Methodology, Writing—original draft
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-8525-9569
  3. Natalie Vande Pol

    Department of Microbiology and Molecular Genetics, Michigan State University, East Lansing, United States
    Contribution
    Formal analysis, Investigation, Writing—original draft
    Competing interests
    No competing interests declared
  4. Nathaniel E Ostrom

    1. Department of Integrative Biology, Michigan State University, East Lansing, United States
    2. DOE Great Lakes Bioenergy Research Center, Michigan State University, East Lansing, United States
    Contribution
    Data curation, Formal analysis, Methodology, Writing—original draft, Writing—review and editing
    Competing interests
    No competing interests declared
  5. Christoph Benning

    1. Department of Energy-Plant Research Laboratory, Michigan State University, East Lansing, United States
    2. DOE Great Lakes Bioenergy Research Center, Michigan State University, East Lansing, United States
    3. Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, United States
    4. Department of Plant Biology, Michigan State University, East Lansing, United States
    Contribution
    Supervision, Funding acquisition, Methodology, Writing—original draft, Writing—review and editing
    For correspondence
    benning@msu.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-8585-3667
  6. Gregory M Bonito

    1. Department of Microbiology and Molecular Genetics, Michigan State University, East Lansing, United States
    2. DOE Great Lakes Bioenergy Research Center, Michigan State University, East Lansing, United States
    3. Department of Plant, Soil and Microbial Sciences, Michigan State University, East Lansing, United States
    Contribution
    Conceptualization, Formal analysis, Supervision, Funding acquisition, Methodology, Writing—original draft, Project administration, Writing—review and editing
    For correspondence
    bonito@msu.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-7262-8978

Funding

U.S. Department of Energy (DE-FG02-91ER20021)

  • Christoph Benning
  • Zhi-Yan Du

U.S. Department of Energy (DE-SC0018409)

  • Nathaniel E Ostrom
  • Gregory M Bonito

U.S. Department of Energy (DE-FC02-07ER64494)

  • Nathaniel E Ostrom
  • Christoph Benning

European Union Seventh Framework Programme (FP7/2007-2013 n° [627266])

  • Krzysztof Zienkiewicz

National Science Foundation (DEB 1737898)

  • Zhi-Yan Du
  • Natalie Vande Pol
  • Gregory M Bonito

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank T James, J Tiedje, S Y He, B Sears, R Roberson and F Trail for critical discussion and review of this manuscript. We acknowledge J Uehling (Duke) and A Desirò (MSU) for detection and clearing of endobacteria from the fungal isolates used in this research and R Vilgalys (Duke) for use of the genome reference isolate of Mortierella elongata AG77. We thank A Albin, C Flegler, A Withrow, M Frame at the MSU Center for Advanced Microscopy for technical assistance of microscopy. We thank D Weed and S Hamilton (Kellogg Biological Station, MSU) for carbon and nitrogen analyses. We thank H Gandhi (Department of Integrative Biology, MSU) for technical assistance of 15N analyses. We thank J Alvaro (Hope College, Michigan) for assistance in the preliminary test of interaction between algae and fungi. We thank D Schnell (MSU) for providing [14C]-sodium bicarbonate and E Poliner and E Farre for providing Aureo4 primers. This work was supported in part by a grant from the Chemical Sciences, Geosciences, and Biosciences Division, Office of Basic Energy Sciences, Office of Science, U.S. Department of Energy (DE-FG02-91ER20021) to CB. This material is based upon work supported by the Great Lakes Bioenergy Research Center, U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research under Award Numbers DE-SC0018409 and DE-FC02-07ER64494. KZ received funding from the People Programme (Marie Curie Actions) of the European Union's Seventh Framework Programme FP7/2007−2013/under REA grant agreement n° [627266] supporting KZ. The EU is not liable for any use that may be made of the information contained therein. GB and CB are grateful to AgBioResearch for financial support.

Senior Editor

  1. Ian T Baldwin, Max Planck Institute for Chemical Ecology, Germany

Reviewing Editor

  1. Maria J Harrison, Boyce Thompson Institute for Plant Research, United States

Reviewers

  1. Maria J Harrison, Boyce Thompson Institute for Plant Research, United States
  2. Paola Bonfante, Università di Torino, Italy
  3. Erik Hom, Unversity of Mississippi, United States

Publication history

  1. Received: April 19, 2019
  2. Accepted: June 17, 2019
  3. Version of Record published: July 16, 2019 (version 1)

Copyright

© 2019, Du et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 3,964
    Page views
  • 531
    Downloads
  • 1
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)

Further reading

    1. Ecology
    Jakub Witold Bubnicki et al.
    Research Article Updated
    1. Ecology
    Laure-Anne Poissonnier et al.
    Research Article