Coregulation of dimorphism and symbiosis by cyclic AMP signaling in the lichenized fungus Umbilicaria muhlenbergii

Significance

Umbilicaria muhlenbergii is the only lichenized species known to be dimorphic in culture. This study showed that yeast-to-pseudohypha transition in U. muhlenbergii is associated with symbiosis and regulated by the cAMP-PKA (protein kinase A) pathway. Treatments with a cAMP diphosphoesterase inhibitor induced pseudohyphal/hyphal growth and differentiation of lichen cortex-like tissues. Two Gα subunits were functionally characterized to show the regulation of dimorphism by UmGPA3. This is a report on identifying genes important for development and symbiosis in lichenized fungi. This is also a report of generating targeted gene deletion mutants in U. muhlenbergii and Lecanoromycetes in general. Because of its relatively fast growth rate and amenability to molecular manipulations, U. muhlenbergii is uniquely suited for studying fungal–algal interactions and initial symbiotic interactions.

Abstract

Umbilicaria muhlenbergii is the only known dimorphic lichenized fungus that grows in the hyphal form in lichen thalli but as yeast cells in axenic cultures. However, the regulation of yeast-to-hypha transition and its relationship to the establishment of symbiosis are not clear. In this study, we show that nutrient limitation and hyperosmotic stress trigger the dimorphic change in U. muhlenbergii. Contact with algal cells of its photobiont Trebouxia jamesii induced pseudohyphal growth. Treatments with the cAMP diphosphoesterase inhibitor IBMX (3-isobutyl-1-methylxanthine) induced pseudohyphal/hyphal growth and resulted in the differentiation of heavily melanized, lichen cortex-like structures in culture, indicating the role of cAMP signaling in regulating dimorphism. To confirm this observation, we identified and characterized two Gα subunits UmGPA2 and UmGPA3. Whereas deletion of UmGPA2 had only a minor effect on pseudohyphal growth, the ΔUmgpa3 mutant was defective in yeast-to-pseudohypha transition induced by hyperosmotic stress or T. jamesii cells. IBMX treatment suppressed the defect of ΔUmgpa3 in pseudohyphal growth. Transformants expressing the UmGPA3^G45V or UmGPA3^Q208L dominant active allele were enhanced in the yeast-to-pseudohypha transition and developed pseudohyphae under conditions noninducible to the wild type. Interestingly, T. jamesii cells in close contact with pseudohyphae of UmGPA3^G45V and UmGPA3^Q208L transformants often collapsed and died after coincubation for over 72 h, indicating that improperly regulated pseudohyphal growth due to dominant active mutations may disrupt the initial establishment of symbiotic interaction between the photobiont and mycobiont. Taken together, these results show that the cAMP-PKA pathway plays a critical role in regulating dimorphism and symbiosis in U. muhlenbergii.

Lichens are symbiotic associations between a fungus (mycobiont) and a photosynthetic partner (photobiont), usually green algae or cyanobacteria. Apart from the two major symbionts, other organisms, including bacteria, fungi, and algae, may be present and have functions in lichens as a stable symbiotic community (1–3). Lichens can thrive in harsh environments and play important roles in the ecosystem by covering ∼8% of the terrestrial surface (4, 5). Based on the morphology of their thalli, lichens can be categorized into different growth types, such as fruticose, foliose, and crustose lichens. Typically, lichens have a cortex or outer layer that consists of highly differentiated, densely aggregated fungal hyphae or symbiotic complex in which individual hyphae are no longer distinguishable (6, 7). Beneath the upper cortex is a layer of photobiont cells that may be arranged in disorder or regularly with interwoven hyphae. For most lichens, the lower surface or cortex may form special structures for tight adherence to the substrate.

Although lichenization occurs in species belonging to polyphyletic groups in different phyla, indicating multiple origins of lichenized fungi, the majority (99%) of them are ascomycetes (8). In fact, ∼46% of ascomycetes are lichen mycobionts (9), and a large majority of them belong to class Lecanoromycetes (8). For most lichenized fungi, their free-living forms are rarely observed in nature (10), although they may be cultured axenically under laboratory conditions. In general, the isolated mycobionts grow very slowly in culture and may require weeks of incubation to form sizable colonies, which are typically compact and pigmented (11). Among all lichenized ascomycetes (>280 genera) that have been isolated, only a few species are known to produce conidia and ascospores in culture (12–14).

Resynthesis of a lichen thallus is a complex process that begins with the physical contact between the mycobiont and photobiont cells. The two symbiotic partners then grow together to form an undifferentiated mass, which further develops into a stratified thallus after a transitional stage (6, 15). The formation of a mature lichen thallus is difficult under laboratory conditions, likely due to the requirement of unknown specific environmental or biological factors in nature. However, the initial interaction between the mycobiont and photobiont and structural development have been reported in several species (16–19). Before the physical contact stage, the mycobiont exhibits selectivity and compatibility toward different phototrophs (20–22). At the interaction stage, fungal hyphae often increase the production of short lateral branches to envelop compatible algal cells (23). Some mycobionts, such as Cladonia, Lecanora, and Xanthoria, form haustoria or haustorium-like structures within photobiont cells (21, 24). However, due to their slow growth rate, lichenized fungi are not ready amenable to molecular genetic studies. To date, no genes responsible for the establishment of symbiosis have been functionally characterized at the molecular level.

Unlike all other cultivated lichen-forming ascomycetes that grow only as hyphae, the isolated mycobiont of the lichen, Umbilicaria muhlenbergii, grows by budding as yeast cells in axenic cultures (25). U. muhlenbergii is a foliose lichen that usually grows on bare stones at high altitudes. Its photobiont partner is mainly Trebouxia jamesii, a green algal species (26). The unicellular yeast cells of U. muhlenbergii mycobiont grow by budding and form typical yeast-like colonies (25). Although U. muhlenbergii lives in the yeast form in culture, it exists in the hyphal or pseudohyphal form in the lichen thallus, indicating that this lichenized fungus must undergo the yeast-to-hypha transition during the establishment of symbiosis. U. muhlenbergii, like many other species in the genus Umbilicaria, can grow in extreme environmental conditions and often produces heavily melanized thalli. Ascospores formed in apothecia (sexual fruiting bodies) are its preferred propagules for dispersal instead of the vegetative reproducing mixtures of mycobiont and photobiont cells (27). Other than ascospores, the U. muhlenbergii lichen thalli also can produce spermatia on spermatiophores in spermogonia (asexual fruiting bodies). Unicellular spermatia are rod like (4.8 × 2.4 µm) and hyaline (28). Although some Umbilicaria species produce thalloconidia that are often darkly pigmented and multicellular (multiseptate) directly from scleroplectenchymatous tissues or protruding rhizinomorphs of the lower cortex of lichen thalli, U. muhlenbergii is one of the Umbilicaria species that does not form thalloconidia (29–32). Even the formation of spermatia and spermogonia observed in the symbiotic state (lichen thalli) has not been observed in culture, and their biological functions in the life cycle of U. muhlenbergii remain to be characterized. U. muhlenbergii is a widely distributed fungus in northern Asia and North America (32). However, to date, only one U. muhlenbergii strain isolated from China has been observed to undergo dimorphic transition in axenic cultures (25), and the underlying mechanism regulating dimorphism is not clear.

Whereas dimorphism is known to be related to pathogenesis in some fungal pathogens, U. muhlenbergii is the only lichenized fungus known to be dimorphic under laboratory conditions and thus, is uniquely suitable for studying the relationship between and coregulation of dimorphism and symbiosis. In this study, we show that nutritional and environmental stresses as well as algal cells of its photobiont induced the yeast-to-pseudohypha transition in U. muhlenbergii. Treatment with cAMP and IBMX (3-isobutyl-1-methylxanthine; an inhibitor of cAMP diphosphoesterase) also stimulated pseudohyphal growth and the differentiation of lichen cortex-like structures in culture, indicating a regulatory role of cAMP signaling (SI Appendix, Fig. S1). Whereas mutants deleted of the UmGPA3 Gα gene that functions upstream from the cAMP-PKA pathway were defective in the yeast-to-pseudohypha transition, expressing dominant active alleles of UmGPA3 enhanced pseudohyphal growth and had a detrimental effect on T. jamesii algal cells in close contact with fungal pseudohyphae. These results indicate that the cAMP-PKA pathway plays a critical role in regulating dimorphism and the initial symbiotic interaction with its photobiont cells in the Lecanonormycete U. muhlenbergii.

Results

Isolation of Mycobiont and Photobiont.

The lichen U. muhlenbergii normally has a circular thallus with umbilici at the center of the lower surface for substrate attachment and often forms apothecia on mature thalli (Fig. 1A). Beneath the highly differentiated upper cortex is the algal layer and medullary layer (Fig. 1A). Close contacts between branching hyphae of the mycobiont and algal cells of the photobiont were observed at the interface of these two layers (Fig. 1B). After surface sterilization and removal of the top layers, U. muhlenbergii strain JL3 was isolated from the medullary layer of lichen thalli collected at Tulaopoding Mountain, China. As previously reported (25), strain JL3 grew by budding as yeast cells (3.0 to 4.5 × 5.0 to 6.5 µm) in potato dextrose agar cultures (SI Appendix, Fig. S2). After incubation at 25 °C for 14 d, strain JL3 formed whitish colonies with a smooth margin. BLASTn searches and phylogenetic analysis with the rRNA (ribosomal RNA)-ITS (internal transcribed spacer) sequence (ITS1 + 5.8S rRNA + ITS2) amplified with primers ITS4 and ITS5 (33) confirmed that it is a U. muhlenbergii isolate (SI Appendix, Fig. S3). The rRNA-ITS sequence of U. muhlenbergii strain JL3 is identical to that of the published strain (25, 34).

Fig. 1.

Lichen thallus, colony morphology, and dimorphism of U. muhlenbergii. (A) The lichen thallus and its cross-section to show the upper cortex (UC), photobiont layer (P), medulla layer (M), and lower cortex (LC). Apothecia (sexual fruiting bodies) are formed on the upper surface. The undersurface has the umbilicus (a central holdfast) for attaching to rocks. (B) The cross-section of a lichen thallus examined by scanning electron microscopy. Hyphae of the mycobiont (MY) wrap around algal cells of the photobiont (PH). (C) Colonies of U. muhlenbergii on regular and diluted potato dextrose agar (0.2× PDA). (D) Yeast cells grown on PDA and pseudohyphae grown on 0.2× PDA were examined by DIC (differential interference contrast) and scanning electron microscopy (SEM). (E) Colonies and pseudohyphae of U. muhlenbergii on BBM plates. Pseudohyphae and conidia were examined after staining with CFW. (F) Colonies, close-up view of colony margins, and pseudohyphae of U. muhlenbergii cultured on PDA with 0.5 or 0.8 M sorbitol. Fuzzy surfaces and margins are related to pseudohyphal growth. (G) Pseudohyphae induced by 0.5 M sorbitol were examined by DIC and scanning electron microscopy (SEM). (H) Size and morphology of yeast cells, pseudohyphae, and conidia of U. muhlenbergii. The incubation time was 14 d for regular PDA, 0.2× PDA, or PDA with 0.5/0.8 M sorbitol and 1 mo for BBM.

We also isolated the green algal photobiont T. jamesii from the same specimens and verified its identity by sequencing the rRNA-ITS region (SI Appendix, Fig. S3). Algal cells of T. jamesii are spherical and 10 to 15 µm in diameter. Another green algal species with a different colony morphology and bigger algal cells than T. jamesii was repeatedly isolated from the same lichen thallus (SI Appendix, Fig. S2), which was not observed in the earlier report (27). Analysis of the rRNA-ITS sequences showed that it is an Elliptochloris species closely related to Elliptochloris subsphaerica (SI Appendix, Fig. S3). Elliptochloris species have been reported as endolichenic algae present in various lichens (3, 35). In comparison with T. jamesii, this Elliptochloris isolate had a faster growth rate and required higher light intensity for growth.

We also isolated the mycobiont from lichen thalli of U. muhlenbergii collected from Smoky Mountain, NC. Similar to the Chinese isolates, the US isolate (US1) also grew as yeast cells on regular PDA media and is in the same clade with known U. muhlenbergii isolates based on phylogenetic analysis (SI Appendix, Fig. S3).

Nutrient Limitation Induces Pseudohyphal Growth in U. muhlenbergii.

Because of its slow growth rate, we normally examined yeast cells and colony morphology of U. muhlenbergii after incubation for 14 d (Fig. 1C). We noticed that colony morphology changed in old cultures. In 1-mo-old cultures, colonies of U. muhlenbergii became brown and wrinkled. Microscopical examination showed that some cells had undergone the yeast-to-hypha transition, and pseudohyphae were visible.

We hypothesized that pseudohyphal growth was induced by nutrient starvation in old cultures. To test this hypothesis, we cultured U. muhlenbergii on dilute PDA medium. On 0.2× PDA plates, colonies became brown and had an uneven surface after incubation for 14 d (Fig. 1C). A large number of yeast cells were found to have switched to pseudohyphal growth. Daughter cells were not separated from mother cells and elongated, forming pseudohyphae that had budding tip cells (Fig. 1D).

To further test that nutrient limitation induces pseudohyphal growth, we cultured U. muhlenbergii on the inorganic ion medium BBM (Bold's Basal Medium) (36) used for algal cultures. Due to the lack of a carbon source, fungal growth on BBM was limited, but pseudohyphal development was observed as early as 5 d (Fig. 1E). After incubation on BBM for 1 mo, pseudohyphae became extremely long and produced unicellular, rod-shaped conidia (1.5 to 2.0 × 4.0 to 5.0 μm) that were smaller than typical yeast cells (Fig. 1 E and H). Conidia were produced and aggregated at the tip or branching site of long pseudohyphae, which was not observed in old cultures on 0.2× PDA.

Hyperosmotic Stress Also Induces Pseudohyphal Growth.

In aged cultures, U. muhlenbergii may develop pseudohyphae in response to hyperosmotic stress caused by the desiccation of PDA media. To test this hypothesis, we first assayed growth on PDA supplemented with 1 M sorbitol, 1 M NaCl, or 20% sucrose. After incubation for 10 d, pseudohyphal growth was observed on PDA with 1 M sorbitol, but the overall fungal growth was limited, and many cells appeared to be dead or devoid of cytoplasm (SI Appendix, Fig. S4). Similar results were obtained when U. muhlenbergii was cultured on PDA with 1 M NaCl or 20% sucrose. These results indicated that the osmotic stress caused by 1 M sorbitol, 1 M NaCl, or 20% sucrose induces pseudohyphal growth but is harmful to U. muhlenbergii.

We then assayed for pseudohyphal growth on PDA with 0.5 or 0.8 M sorbitol. After incubation for 14 d, the overall growth of U. muhlenbergii was much more robust on medium with 0.5 M sorbitol than on medium with 0.8 M sorbitol, confirming its sensitivity to hyperosmotic stress. Colonies with fuzzy margins and an uneven surface were observed in both conditions, although 0.8 M sorbitol appeared to be more effective in inducing pseudohyphal growth (Fig. 1F). Pseudohyphae induced by 0.5 or 0.8 M sorbitol were longer and more highly branched than pseudohyphae formed on old PDA cultures (Fig. 1F). In comparison with normal yeast cells, cellular compartments in pseudohyphae were narrower but longer, and normally, only the tip cell still grew by budding. Lateral growth from some compartments resulted in the branching of pseudohyphae (Fig. 1G).

When stained with 4′,6-diamidino-2-phenylindole and Calcofluor white (CFW) to visualize the nucleus and cell wall, the unicellular yeast cells of U. muhlenbergii had a single nucleus. Pseudohyphae induced by nutrient starvation or osmotic stress also had one nucleus in each compartment (SI Appendix, Fig. S5). These results indicated that, although the compartments in pseudohyphae were longer than normal yeast cells of U. muhlenbergii, they were still uninucleate. Pseudohyphae appeared to grow by budding at the tip, and each septum was associated with a constriction at the budding site of the growing tip cell.

Oxidative and Cell Wall Stresses Fail to Induce Pseudohyphal Growth.

We tested the impact of oxidative and cell wall stress on pseudohyphal growth in U. muhlenbergii. This fungus appears to be hypersensitive to all these stresses (SI Appendix, Fig. S4). In the presence of 1 mM H₂O₂, growth was limited, and most of the cells were dead (empty) after incubation for 10 d. On PDA containing 1.5 mM Congo Red, U. muhlenbergii grew slightly better, but many dead (empty) yeast cells also were observed. Unlike hyperosmotic stress, stimulation of pseudohyphal growth by oxidative and cell wall stresses was not observed.

Pseudohyphal Growth Is Induced by T. jamesii Cells.

Because U. muhlenbergii is in the hyphal form in the lichen thallus, we tested the effect of algal cells on the yeast–hypha dimorphic transition. When cells of T. jamesii were mixed with U. muhlenbergii cells and coincubated on PDA, attachment of yeast cells to algal cells and induction of pseudohyphal growth were observed (Fig. 2A). Pseudohyphae were visible around algal cells after cocultivation for 72 h (Fig. 2A), and the resulting fungal–algal cell aggregates could not be easily separated by a gentle touch. When yeast cells of U. muhlenbergii were resuspended to 5 × 10³/mL in the supernatant of BBM cultures of T. jamesii after removing algal cells by centrifugation and cultured on PDA plates, pseudohyphal growth as shown in Fig. 2A was not observed after incubation for 72 h, indicating the importance of physical contact between fungal and algal cells.

Fig. 2.

Pseudohyphal growth of U. muhlenbergii induced by its photobiont T. jamesii. (A) Time course assays of the initial interaction of U. muhlenbergii yeast cells with its photobiont T. jamesii. (B) Pseudohyphae of U. muhlenbergii after coculturing with T. jamesii for 2 wk. (C) Pseudohyphae induced by algal cells were stained with CFW and examined by epifluorescence microscopy. (D) A mixed culture of U. muhlenbergii and T. jamesii cells after incubation for 14 d on PDA.

After coincubation for 14 d, long pseudohyphae with growing tip cells were formed (Fig. 2B). Some of these pseudohyphae had the tip cell growing away from algae, although their basal ends aggregated around T. jamesii cells (Fig. 2B). The close contact between fungal and algal cells was observed after staining with CFW (Fig. 2C). Interestingly, compartments of pseudohyphae tended to vary significantly in size and morphology in the hyphal–algal aggregates but became more uniform when pseudohyphae were not in close contact with algae (Fig. 2 B and C). After incubation for 14 d, the surface of fungal–algal cocultures had a hairy appearance that is related to pseudohyphal growth of U. muhlenbergii (Fig. 2D). Enhanced pigmentation and pockets of green algal cells also were observed in some parts of the cocultures (Fig. 2D). These results indicated that pseudohyphal growth of U. muhlenbergii was induced by algal cells of T. jamesii during their early interactions, but they failed to establish a stable symbiotic relationship as in the lichen thallus under this laboratory condition.

Algal Cells of Elliptochloris Fail to Induce Pseudohyphal Growth in U. muhlenbergii.

To determine whether the Elliptochloris strain isolated in this study induces dimorphic transition in U. muhlenbergii, we also coincubated mixtures of these two organisms as described above. In their cocultures, pseudohyphal growth and changes in yeast cell morphology were not observed in U. muhlenbergii cells in close contact with Elliptochloris cells (SI Appendix, Fig. S6), and they could be easily separately by a gentle touch or washing with sterile water. Algal cells of this Elliptochloris isolate grew faster that T. jamesii but failed to induce the yeast-to-hypha morphological transition in U. muhlenbergii, indicating a specific recognition between the mycobiont and T. jamesii cells.

Exogenous cAMP and IBMX Stimulate the Development of Pseudohyphae.

Because of the conserved role of the cAMP-PKA pathway in intracellular signaling and nutrient sensing in fungi, we assayed the effect of exogenous cAMP on U. muhlenbergii. The addition of 10 mM cAMP affected colony morphology and stimulated the elongation of yeast cells that was visible as early as 48 or 72 h on PDA plates (Fig. 3A). After incubation for 14 d, colonies of U. muhlenbergii often had a bumpy or uneven surface and some areas with increased pigmentation due to pseudohyphal growth (Fig. 3A).

Fig. 3.

Effects of cAMP or IBMX on colony morphology and cellular differentiation. (A) Colony and cell morphology of U. muhlenbergii after incubation on PDA with 10 mM cAMP for 14 d. Pseudohyphae were observed as early as 48 h and became longer after incubation for 72 h. In 14-d-old cultures, pseudohyphae became more abundant but generally not longer than eight compartments. (B) Cultures of U. muhlenbergii on PDA treated with IBMX were examined after incubation for 7 and 14 d. The filter paper placed on the left side of the U. muhlenbergii colony contained 25 nmol IBMX. Bottom shows a close-up view of the surface and margin of 14-d-old cultures. (C) Cellular differentiation induced by IBMX treatment was examined by DIC (differential interference contrast) microscopy. Individual cells were no longer visible in the heavily melanized and differentiated tissues at 14 dpt. (D) One-month-old PDA cultures 1 cm away from the IBMX filter were examined for surface differentiation, hyphal and pseudohyphal growth, and aggregates of melanized cells. SEM, scanning electron microscopy.

We then tested the effect of IBMX, an inhibitor of cAMP diphosphoesterase, on U. muhlenbergii. Discs of Whatman filter paper containing 25 nmol of IBMX (dissolved in dimethyl sulfoxide) were placed on one side of a streak of U. muhlenbergii cells across the surface of 10-mL PDA plates. As IBMX diffused in the medium, it formed a descending concentration gradient across the U. muhlenbergii colony (Fig. 3B). Stimulation of colony pigmentation (Fig. 3B) and pseudohyphal growth (Fig. 3C) were observed at 7 d post–3-isobutyl-1-methylxanthine treatment (dpt). At 14 dpt, the surface of U. muhlenbergii colony near the filter paper (0.5 cm away from the filter paper) became darkly pigmented with a bumpy and hairy appearance. When the dark-pigmented clumps (tissue-like structures) were crushed, aggregates of pseudohyphae or hyphae and heavily melanized spherical cells were observed (Fig. 3 B and C). These spherical cells are likely related to differentiated cells or compartments of pseudohyphae.

After incubation for 1 mo, IBMX induced more significant changes in U. muhlenbergii, particularly in areas near the filter paper. The bumpy colonies were heavily melanized and covered with protruding hyphae or long pseudohyphae (Fig. 3D). Microscopic examination showed that the highly differentiated, hardened tissues consisted of a dense epidermal layer with heavily melanized pseudohyphae or yeast cells, in which individual compartments or cells were often no longer distinguishable. Furthermore, hyphal filaments without obvious constrictions also were often observed in these highly differentiated tissues. Pseudohyphal and hyphal growth on the surface of IBMX-treated colonies also was observed by scanning electron microscopy (Fig. 3D). Therefore, IBMX was more effective in inducing pseudohyphal and hyphal growth than cAMP. The development of an epidermal layer of highly differentiated and melanized hyphal compartments or yeast cells on the clumps of IBMX-treated colonies was, similar to the cellular differentiation of U. muhlenbergii, induced by the photobiont in the lichen thalli.

The UmGPA3 Gene Regulates Pseudohyphal Growth.

The Gα subunits of the trimeric G proteins that function upstream from the cAMP-PKA pathway have been characterized in different fungi for roles in morphogenesis and yeast–hypha dimorphism (37, 38). A BLASTP search of the U. muhlenbergii genome (34) with three Magnaporthe oryzae Gα subunits identified three orthologs on contigs 70, 79, and 232, respectively. They were named as UmGPA1 (GenBank accession no. MT50998), UmGPA2 (GenBank accession no. MT509979), and UmGPA3 (GenBank accession no. MT509978) in this study. Sequence alignment and phylogenetic analysis showed that UmGpa3 is homologous to Ustilago maydis Gpa3 and Candida albicans Gpa2 that are involved in regulating dimorphism and pathogenesis. UmGpa2 is in the same clade with U. maydis Gpa2 and C. albicans Gpa1 (SI Appendix, Fig. S7).

Based on the functions of orthologs characterized in other fungi, such as U. maydis Gpa3 and C. albicans Gpa2 (37, 39), we hypothesized that UmGPA3 may play a more important role in dimorphism than the other Gα subunits in U. muhlenbergii. To determine its function, the Umgpa3 gene knockout mutant was generated by homologous recombination using the split-marker approach (SI Appendix, Fig. S8) (40). After screening hygromycin-resistant transformants generated in five independent transformations, three independent Umgpa3 deletion mutants (gpa3-1, gpa3-3, and gpa3-4) were identified and confirmed by PCR with anchor primers. All three Umgpa3 deletion mutants had the same phenotype, although only data of mutant strain gpa3-1 are presented below. The growth rate of the ΔUmgpa3 mutant was similar to that of the wild type in PDB (potato dextrose broth) medium (SI Appendix, Fig. S9), suggesting that deletion of UmGPA3 gene had no effect on yeast growth.

After incubation on 0.2× PDA for 10 d, the wild type formed colonies with melanized areas on which pseudohyphae or hyphae protruded from the surface (Fig. 4A). Under the same conditions, the ΔUmgpa3 mutant formed colonies with a smooth surface, likely due to its defects in the yeast-to-hypha transition. In 10-d-old liquid BBM cultures, the majority of the wild-type cells were pseudohyphae (Fig. 4B). However, the ΔUmgpa3 mutant still grew by budding as yeast cells (Fig. 4B), indicating that deletion of UmGPA3 affected the yeast-to-hypha transition on nutrient-limited media.

Fig. 4.

Defects of the ΔUmgpa3 mutant in pseudohyphal growth. (A) Ten-day-old 0.2× PDA cultures of the wild type and the ΔUmgpa2 and ΔUmgpa3 deletion mutants were examined for pigmentation and colonial growth. (B) Ten-day-old BBM cultures of the same set of strains were examined for pseudohyphal growth. Unlike the wild type and ΔUmgpa2 mutant, the ΔUmgpa3 mutant was defective in the yeast–pseudohypha transition. (C&D) The morphology of colonies (C) and cells (D) of the indicated strains cultured on PDA with 0.5 M sorbitol for 10 d. The wild type and ΔUmgpa2 formed wrinkly, fuzzy colonies and produced branching pseudohyphae. Pseudohyphal growth was not observed in smooth colonies of the ΔUmgpa3 mutant. (E) PDA cultures of the ΔUmgpa3 mutant treated with IBMX were examined for pseudohyphal growth and colony surface after incubation for 4 or 14 d.

We also assayed the defect of ΔUmgpa3 mutant in dimorphic transition induced by hyperosmotic stress. On PDA with 0.5 M sorbitol, the wild-type strain formed wrinkly colonies with fuzzy margins, but ΔUmgpa3 colonies had a smooth appearance (Fig. 4C). Microscopic examination of cells from colony margins showed that the wild type had long and branching pseudohyphae induced by 0.5 M sorbitol. Under the same conditions, no pseudohyphal growth was observed in the ΔUmgpa3 mutant (Fig. 4C and SI Appendix, Fig. S10).

To further verify that UmGPA3 functions upstream from the cAMP-PKA pathway, 2.5 mM IBMX was added to PDA cultures of the ΔUmgpa3 mutant. Pseudohyphal growth induced by IBMX was visible in the ΔUmgpa3 mutant after incubation for 4 d (Fig. 4E). After incubation for 14 d, the highly differentiated and melanized pseudohyphae were observed in the hairy clumps of the IBMX-treated cultures (Fig. 4E). Therefore, we conclude that UmGPA3 plays an important role in regulating pseudohyphal growth in U. muhlenbergii, and the defect of the ΔUmgpa3 mutant in yeast-to-hypha transition can be rescued by treatment with IBMX.

The UmGPA2 Gene Partially Regulates Pseudohyphal Growth.

The orthologs of UmGPA2 have overlapping functions with other Gα subunits in Cryptococcus neoformans and Neurospora crassa (41, 42). To determine its function, we also generated the Umgpa2 deletion mutant by gene replacement. After incubation for 1 mo on 0.2× PDA, colony melanization and pseudohyphal growth were still observed in ΔUmgpa2 cultures but to a lesser content in comparison with the wild type (Fig. 4A). In liquid BBM cultures, pseudohyphal growth was still observed in the ΔUmgpa2 mutant, but the length of its pseudohyphae was shorter than those of the wild type (Fig. 4B). When cultured on PDA with 0.5 M sorbitol for 10 d, the ΔUmgpa2 mutant also produced shorter pseudohyphae and less wrinkled colonies than the wild type (Fig. 4 C and D). These results indicate that UmGPA2 is not essential for the yeast-to-hypha transition but may play a minor role in the growth of pseudohyphae.

The ΔUmgpa3 Deletion Mutant Is Defective in Response to Algal Cells.

To determine whether deletion of UmGPA3 affects its response to the photobiont, yeast cells of the wild-type, ΔUmgpa2, and ΔUmgpa3 mutant strains were cocultured with algal cells of T. jamesii and observed every 12 h (Fig. 5). After 36 h of coculturing, yeast cells of the wild type that were associated with algal cells began the yeast-to-hypha transition, resulting in the formation of pseudohyphae that became more distinctive by 72 h (Fig. 5). Similar results were obtained with the ΔUmgpa2 mutant. Under the same conditions, yeast cells of the ΔUmgpa3 mutant closely associated with algal cells mainly grew by budding. Occasionally, chains of three yeast cells were observed, but these cells, unlike pseudohyphae, separated shortly thereafter. Pseudohyphae with the tip cell growing away from algal cells were observed in the wild type and ΔUmgpa2 mutant but never in the ΔUmgpa3 mutant (Fig. 5A).

Fig. 5.

Defects of the ΔUmgpa3 mutant in interactions with T. jamesii cells. (A) Time course assays of the interaction of the wild type and the ΔUmgpa2 and ΔUmgpa3 deletion mutants with the photobiont T. jamesii cells. Pseudohyphal growth was induced by T. jamesii in the wild type and ΔUmgpa2 mutant (marked with arrows). The ΔUmgpa3 mutant grew as yeast cells aggregated around algal cells but failed to form elongated pseudohyphae. (B) Percentages of colonies with extensive pseudohyphae in cocultures of T. jamesii cells with yeast cells of the wild type and ΔUmgpa2 or ΔUmgpa3 mutant after incubation for 3 d. Mean and SD were calculated with data from three independent replicates, with at least 100 cell clusters examined in each replicate. Asterisks * and ** indicate significant differences based on one-way ANOVA with the LSD (least significant difference) t test analysis (P < 0.01).

After cocultivation for 3 d, individual colonies consisting of fungal and algal cells became visible. Whereas the vast majority (95.8%) of the wild-type U. muhlenbergii colonies-associated algal cells were in the pseudohyphal form, all of the ΔUmgpa3 mutant colonies remained in the yeast form. Under the same conditions, 35.8% of ΔUmgpa2 mutant colonies underwent morphological transition (Fig. 5B). After cocultivation for 10 d, colonies with obvious pigmentation appeared on the wild-type cultures, but ΔUmgpa3 mutant colonies remained nonpigmented and dome shaped (SI Appendix, Fig. S11). For the ΔUmgpa2 mutant, the extent of changes in colony morphology and pigmentation was between those of the wild type and the ΔUmgpa3 mutant strains (SI Appendix, Fig. S11). These results indicated that UmGPA3 plays a major role, but UmGPA2 also has a minor role in regulating the induction of pseudohyphal growth by algal cells of T. jamesii in U. muhlenbergii.

Dominant Active Mutations in UmGPA3 Promote Pseudohyphal Growth.

The G142V and G355L mutations are known to result in dominant active mutations in Gpa2 of C. albicans (37). Sequence alignment showed that these two amino acid residues are highly conserved in UmGPA3 (G45 and G208) and its orthologs from other ascomycetes (SI Appendix, Fig. S7). To further characterize its function in pseudohyphal growth, we generated the UmGPA3^G45V and UmGPA3^Q208L alleles and transformed them into the wild type. The resulting UmGPA3^G45V and UmGPA3^Q208L transformants grew normally as yeast cells on PDA medium. The dominant active mutations in UmGPA3 had no effect on growth (SI Appendix, Fig. S9). However, colony melanization and pseudohyphal growth were observed in 20-d-old PDA cultures of these two transformants. Whereas the wild type grew as yeast cells, the UmGPA3^G45V and UmGPA3^Q208L transformants, particularly the latter, had abundant pseudohyphae (Fig. 6A). These results indicated that expressing these dominant UmGPA3 alleles enhanced pseudohyphal growth and may stimulate yeast-to-hypha transition under conditions not inducible in the wild type.

Fig. 6.

Effects of dominant active mutations in UmGPA3 on pseudohyphal growth. (A) Twenty-day-old cultures of the wild type and transformants expressing the indicated dominant active alleles of UmGPA3 on PDA were examined for pseudohyphal growth. In comparison with the wild type that had limited pseudohyphal growth, the UmGPA3^Q45V and UmGPA3^Q208L transformants had more abundant and longer branching pseudohyphae, particularly the latter. (B) Ten-day-old cultures of the wild type and UmGPA3^Q45V or UmGPA3^Q208L transformant grown on PDA with 0.5 M sorbitol were examined for colony margins (Upper) and pseudohyphal growth (Lower). (C) Microscopic examination of colony edges of the indicated strains cultured on BBM plates for 20 d. Whereas the wild type had limited pseudohyphal growth, the UmGPA3^Q45V and UmGPA3^Q208L transformants, particularly the latter, produced long branching pseudohyphae. Aggregates of conidia (marked with arrows) were observed in all three strains.

When cultured on PDA with 0.5 M sorbitol, colonies of the UmGPA3^G45V and UmGPA3^Q208L transformants, particularly the latter, had a bumpier surface and more hairy margins than wild-type colonies after incubation for 10 d (Fig. 6B). More robust pseudohyphal growth was observed in the UmGPA3^G45V and UmGPA3^Q208L transformants than in the wild type (Fig. 6B). When cultured on BBM agar, long branching hyphae and pseudohyphae as well as aggregates of conidia were observed on the margins of 20-d-old cultures of the UmGPA3^G45V and UmGPA3^Q208L transformants (Fig. 6C). Under the same conditions, the wild type also produced conidia but had shorter pseudohyphae (Fig. 6C). Clusters of pseudohyphae protruding from the margins of UmGPA3^Q208L colonies were longer and more abundant than those in the UmGPA3^G45V transformants. These results indicated that the Q208L mutation may have a more significant dominant active effect on UmGPA3 than the G45V mutation, which is similar to reports in C. albicans (37). However, conidia were observed in all these transformants, indicating that the G45V and Q208L mutations have no effect on conidiation.

Dominant Active Mutations in UmGPA3 Affect the Establishment of the Initial Symbiotic Interaction.

To test the effect of dominant active mutations in UmGPA3 on symbiosis, the interaction between algal and yeast cells were monitored on PDA. In the wild-type strain, pseudohyphal growth was induced by algal cells of T. jamesii. Green, healthy algal cells in close contact with pseudohyphae of U. muhlenbergii were observed at 72 h (Fig. 7A). In the UmGPA3^G45V and UmGPA3^Q208L transformants, pseudohyphal growth also was observed, and the UmGPA3^Q208L transformants appeared to have more robust pseudohyphal growth. However, after 72 h or longer, algal cells associated with pseudohyphae of the UmGPA3^G45V and UmGPA3^Q208L transformants often were collapsed and empty (Fig. 7A). Algal cells in close contact with U. muhlenbergii appeared to be damaged at the contact site by the penetration or intrusion of fungal growth (Fig. 7A). In repeated assays, ∼30% of the algal cells associated with pseudohyphae of UmGPA3^G45V and UmGPA3^Q208L transformants were not healthy or dead and lost chlorophyll autofluorescence by 72 h (Fig. 7 B and C). Under the same conditions, only 6.6% of algal cells in cocultures with the wild-type U. muhlenbergii lost chlorophyll autofluorescence (Fig. 7C), likely due to natural death of algal cells. These results indicate that these two dominant active mutations in UmGPA3 may disrupt the initial establishment of their symbiotic relationship.

Fig. 7.

Effects of dominant active mutations in UmGPA3 on interaction with algal cells. (A) Time course assays of the interaction of T. jamesii cells with the wild-type U. muhlenbergii and transformants expressing the indicated dominant active mutant alleles of UmGPA3. The same field was observed every 12 h. Dominant active mutations in UmGPA3 resulted in the collapse of algal cells after coincubation for 72 h. (B) Viability of T. jamesii cells decreased when cocultured with transformants expressing dominant active UmGPA3 alleles. After cocultivation for 3 d, algal cells in close associations with pseudohyphae were examined for cellular morphology and chlorophyll fluorescence (559-nm excitation, exposure time: 2,000 ms) by DIC (differential interference contrast) and epifluorescence microscopy. (C) Percentage of damaged algal cells with no or faint chlorophyll fluorescence in mycobiont–photobiont clusters in cocultures of T. jamesii cells with yeast cells of the wild type and Umgpa3^G45V or Umgpa3^G208L transformant after incubation for 3 d. Mean and SD were calculated with data from three independent replicates, with at least 100 algal cells examined in each replicate. *Significant differences between the wild type and Umgpa3^G45V or Umgpa3^G208L transformant based on one-way ANOVA with the LSD (least significant difference) t test analysis (P < 0.01).

Discussion

The yeast–hyphal dimorphism occurs to fungi belonging to different phyla and is usually stimulated by environmental factors (43). In fungal pathogens, including the human pathogen C. albicans and corn smut fungus U. maydis, the morphological transformation occurs during infection and is necessary for adaptation to host cells (44–47). Similar to a previous report (25), in this study, we isolated the U. muhlenbergii mycobiont that grew as yeast cells in culture from lichen thalli collected in China and the United States, indicating that dimorphism is not an isolate-specific phenomenon but a common characteristic of U. muhlenbergii, a member of Lecanoromycetes, the largest class of lichen-forming fungi.

Although U. muhlenbergii normally grew as unicellular yeast cells, pseudohyphal growth was induced by nutrient limitation in aging cultures or by growth on the inorganic medium BBM without a carbon source. Longer pseudohyphae bearing conidia were observed in BBM cultures, likely due to severe nutritional deficiency. In Saccharomyces cerevisiae, nutrient limitation induces filamentation that is associated with the yeast to hyphal transition, although morphological changes are slightly different between the haploid and diploid cells (48, 49). U. muhlenbergii is a haploid fungus, and BBM medium lacks carbon and nitrogen sources. Conidiation was not induced by cAMP or IBMX treatment. Conidiation in axenic cultures has not been reported in U. muhlenbergii, but its lichen thalli produce hyaline unicellular, rod-like spermatia that are formed in spermogonia (28). Because they are formed under different conditions and by different tissues, the relationship between conidia and spermatia is not clear.

Like fungal pathogens that undergo dimorphic changes during infection, U. muhlenbergii also responded to its compatible photobiont with the dimorphic transition. Pseudohyphal growth appeared to begin in yeast cells in close contact with T. jamesii cells as early as 36 h of cocultivation. As the time of cocultivation increased, clusters of pseudohyphae tended to envelop algal cells, and algal cells inside appeared to be compressed. Although it is not clear whether haustoria are formed, there was an interface and close association between mycobiont and photobiont cells. We noticed that pseudohyphae and algal cells in close contact could not be easily separated by dispersing in sterile distilled water. Furthermore, pseudohyphae growing away from algal cells differed from those in close contact with algal cells in compartment length and morphology. In similar cocultivation assays with the Elliptochloris isolate, pseudohyphal growth was not induced, and the mixture of U. muhlenbergii and Elliptochloris cells could be easily separated by dispensing in water, indicating a specific recognition of photobiont cells by the mycobiont to induce pseudohyphal growth. In two dimorphic plant pathogenic fungi, U. maydis and Taphrina deformans, mating between compatible yeast cells is associated with the formation of dikaryotic hyphae for penetration and infectious growth (50, 51). In U. muhlenbergii, the yeast-to-pseudohypha transition may be necessary for the establishment of symbiosis and require the presence of compatible photobiont cells.

Similar to its role in the regulation of dimorphism in S. cerevisiae and several other fungi (52, 53), the well-conserved cAMP-PKA pathway was found to be important for regulating pseudohyphal growth and possibly the establishment of the symbiotic state in U. muhlenbergii. Because deletion of the regulatory or catalytic subunits of PKA often results in severe growth defects in other fungi (54), in this study we characterized two Gα subunits and found that the ΔUmgpa3 mutant was defective in the dimorphic transition induced by stress or the presence of algal cells but still responded to IBMX for pseudohyphal growth. Furthermore, pseudohyphal growth was enhanced in transformants expressing the UmGPA3^G45V and UmGPA3^Q208Ldominant active alleles. UmGpa3 is orthologous to MagB of M. oryzae and GzGpa3 of Fusarium graminearum, which are the major Gα subunits for regulating plant infection processes (55–57). Its orthologs also are important for virulence in human pathogenic fungi and regulate the yeast-to-pseudohypha/hypha dimorphic transition in S. cerevisiae and C. albicans (37, 38, 41). In U. maydis, Gpa3 regulates both virulence and mating that are related to the dimorphic switch (39). Our results indicated that, like fungal pathogens, UmGpa3 functions upstream from the cAMP-PKA pathway for regulating pseudohyphal growth and symbiosis in U. muhlenbergii.

In U. muhlenbergii, unlike the ΔUmgpa3 mutant, the ΔUmgpa2 deletion mutant still underwent dimorphic transition in response to nutrient or osmotic stress and algal cells. However, pseudohyphal growth in the ΔUmgpa2 mutant was not as robust as in the wild type. These results indicate that UmGPA2 may play a minor role in pseudohyphal growth. Therefore, it will be important to generate the ΔUmgpa2 ΔUmgpa3 double mutants to characterize the functional relationship between these two Gα subunits during dimorphism and symbiosis. Unfortunately, hygromycin resistance is the only selectable marker that is effective for transformant selection in this lichen-forming fungus, although this is a report of generating targeted gene deletion mutants in U. muhlenbergii and lichenized fungi in general. The UmGPA1 gene was not characterized in this study because it is more distantly related to UmGPA3 than UmGPA2 (SI Appendix, Fig. S7), and its orthologs have been shown in M. oryzae, F. graminearum, and other fungi to be dispensable for growth and development (55–57).

Although lichen thalli formed on the surface of rocks can tolerate desiccation and other stress, we found that U. muhlenbergii is sensitive to hyperosmotic, oxidative, and cell wall stresses, which may be related to physiological differences between fungal cells in axenic culture and lichen thalli. Nevertheless, moderate osmotic stress such as 0.5 M sorbitol induced pseudohyphal growth. In fungi, responses to hyperosmotic stress are normally regulated by the high-osmolarity glycerol response (HOG) pathway (58, 59). In U. muhlenbergii, the HOG pathway may cross-talk with the cAMP-PKA pathway for regulating the yeast-to-hypha transition. In Aspergillus fumigatus, the HOG and cAMP-PKA pathways coregulate the mobilization and utilization of storage carbohydrates for cell wall biosynthesis (60). In S. cerevisiae, filamentation occurs in HOG pathway mutants under osmotic stress (61). In U. muhlenbergii, activation of the HOG pathway by osmotic stress may somehow activate the cAMP-PKA pathway and induce pseudohyphal growth. Because the mating MAP kinase pathway is known to be involved in dimorphism in S. cerevisiae and C. albicans (62, 63), it may also cross-talk with the cAMP-PKA or HOG pathway in the yeast-to-pseudohypha transition in U. muhlenbergii. The genome sequence of U. muhlenbergii (34) has the conserved mating MAP (mitogen-activated protein) kinase pathway, although its mating system and mating behavior have not been characterized.

In cocultivation with T. jamesii, the UmGPA3^G45V and UmGPA3^Q208L transformants switched rapidly to pseudohyphal growth upon contact with algal cells. Interestingly, many algal cells associated with pseudohyphae of these transformants appeared to be collapsed and dead after 72-h coculturing. The dominant active mutations in UmGPA3 likely resulted in defects in the proper regulation of pseudohyphal growth in U. muhlenbergii, and the death of algal cells may be directly caused by aggressive growth of the UmGPA3^G45V and UmGPA3^Q208L transformants. It is also possible that algal cells of T. jamesii normally signal back to U. muhlenbergii to turn off UmGPA3-mediated cAMP signaling, which is affected by these dominant active mutations. In M. oryzae, the G42R mutation in MagB stimulates appressorium formation but significantly reduces its virulence on rice plants (64). In F. graminearum, the Q20L mutation in GzGPA2 also reduces its virulence on wheat heads (56). Similar to these fungal pathogens, dominant active mutations in UmGPA3 may stimulate pseudohyphal growth of U. muhlenbergii but have a negative impact on the establishment of its symbiotic relationship with T. jamesii. Therefore, it will be important to determine the cellular interaction between U. muhlenbergii and T. jamesii and the mechanism responsible for algal cell death caused by the G45V and Q208L mutations in UmGPA3 in future studies. In some lichens, other than wrapping around algal cells, the mycobiont may penetrate and form small haustoria for nutrient uptake (15, 21). If U. muhlenbergii penetrates and forms haustoria in T. jamesii cells, it will be important to determine whether these two dominant active mutations affect the proper regulation of haustorium formation and cause the death of algal cells due to aggressive fungal growth.

This is a report on functional characterization of genes in lichen-forming or lichenized fungi. The approaches developed in this study can be used to further characterize the intimate interaction between the mycobiont and photobiont, particularly at the early stages. Because of its relatively fast growth rate in culture in comparison with other lichenized fungi and amenability to molecular manipulations, U. muhlenbergii is uniquely suitable for studying the regulation of dimorphism and symbiotic interactions with photobiont cells. Furthermore, our data showed that the cAMP-PKA pathway plays an important role in regulating the yeast-to-hypha transition, and proper regulation of pseudohyphal growth is critical for the establishment of symbiotic relationship in U. muhlenbergii, a species in Lecanoromycetes that contains majority of lichen-forming fungi.

Materials and Methods

The materials and methods described in detail in SI Appendix include methods for the isolation and identification of mycobiont and algal strains, culture conditions for inducing dimorphic changes in U. muhlenbergii, cocultivation assays with fungal and algal cells, and generation of the ΔUmgpa2 and ΔUmgpa3 deletion mutants and transformants expressing the UmGPA3^G45V and UmGPA3^G208L alleles. The origins of lichen samples, sequences, and primers (SI Appendix, Table S1) used in this study also were presented in SI Appendix.

Data Availability.

All study data are included in the article and SI Appendix.