Introduction

The relationship between lichens and reactive nitrogen (N) has been extensively studied in recent decades due to the recognition of the adverse effects of anthropogenic N (Steffen et al. 2015) and to the high responsiveness of lichens to reactive N forms (e.g., Sheppard et al. 2009; Munzi et al. 2014; Maslaňáková et al. 2015). In fact, lichens, except cyanolichens, fulfil their N demand taking ammonium, nitrate, and organic N from atmospheric sources (wet and dry depositions) (Hauck 2010). However, despite recent developments in physiological approaches that provide a deep insight in lichen tolerance to pollution (Piccotto et al. 2011; Álvarez et al. 2012; Munzi et al. 2014), the N response of the individual symbionts and the underpinning mechanisms of N effects are far from being fully explained (Hauck 2010).

The use of “omics” technologies to explore the roles, relationships, and cascade reactions that molecules undergo is the further step necessary to understand the molecular mechanisms underlying lichen response to reactive N. Metabolomics (Freitag et al. 2009) and transcriptomics (Junttila and Rudd 2012) are promising tools to study the molecular facets of lichen physiological response, but have not yet been applied to the study of N effects. Similarly, although proteomic approaches have been shown to have a great potential to study non-model species lacking genetic information (Carpentier et al. 2008), few lichen studies have utilized this approach to investigate lichens or lichen symbionts under stress conditions (Rustichelli et al. 2008; Sanità di Toppi et al. 2008; Gasulla et al. 2013; Nicolardi et al. 2012).

The effects of N excess on lichen proteomic profile have yet to be investigated. To fill this gap, we carried out proteomic analyses of the lichen Cladonia portentosa exposed to various doses of oxidized and reduced N forms. Although Cladonia is usually considered a genus sensitive to N and pollution in general (Nimis and Martellos 2008), in the Whim bog manipulative experimental site, C. portentosa appears to adapt to long-term N wet treatments, surviving in the plots after more than 10 years of N treatment (addition of 3× background kg N ha−1 year−1). This means that this lichen is able to develop mechanisms to cope with increased N availability.

This species was shown to provide coherent biomarkers for N enrichment (Hogan et al. 2010a) and to reflect atmospheric ammonia concentration (Nielsen et al. 2014). In addition, Freitag et al. (2011) found that changes in broad biochemical classes (proteins, amino acids, polysaccharides, and nucleic acids) in the lichen C. portentosa were consistently correlated with mean annual wet inorganic N deposition loads in 25 sites across Britain.

We also tested the role of external supply of phosphorus (P) and potassium (K) (hereafter indicated as PK) on alleviating N toxicity symptoms. In fact, some experiments, both in culture and in the field (Zhang et al. 1990; Feng and Barker 1992), have shown that increased ammonium tolerance in tolerant plant species depends on substantial potassium supply (Szczerba et al. 2008). The working hypotheses of the present work are that (1) different forms and doses of N differentially affect protein production and that (2) PK supply can activate protective mechanisms involving changes in protein expression. The null hypothesis is that proteomic analyses should show different sets of proteins in lichens subjected to different N treatments. Far from being exhaustive, this work contributes to the research on the relationship between lichens and N, encompassing for the first time the changes in the holobiome (lichens and microorganisms growing on/in them) proteome effected by N treatment and deepening the role of different partners of the lichen association in N stress response.

Materials and methods

Nitrogen treatments

At Whim bog experimental site, reactive N treatments have been applied since June 2002 (see Sheppard et al. 2011 for details). Four control plots receive only natural rainfall with a background total inorganic N deposition of 8 kg ha−1 year−1. The other experimental plots receive additional N applied as a spray at rates of 8, 24, and 56 kg N ha−1 year−1. Each treatment is replicated in four plots. Additional treatments of 16 and 64 kg N ha−1 year−1 are supplemented with P and K by adding K2HPO4 at a 1:14 P:N ratio (Sheppard et al. 2004). Treatments are provided in rainfall collected at the site and are only applied during rainfall and when air temperature exceeds 0 °C and wind speed is <5 m s−1 over a 5 min averaged period.

Lichen sampling

Thalli of Cladonia portentosa of 15 cm diameter were collected from the soil immediately outside the manipulated area and placed in the treated plots, two thalli per plot, receiving both oxidized and reduced N forms and PK addition (Table 1). After 6 months, transplanted lichens were collected, transported to the laboratory and carefully cleaned to remove parts of other bog species. To avoid protein degradation, samples were immediately immersed in liquid N and stored at −80 °C until analyzed.

Table 1 Summary of the treatments provided (columns = kg N ha−1 year−1, rows = kg P ha−1 year−1)
Full size table

Chlorophyll a fluorescence

Measurements of the performance index (PI) and the Fv/Fm ratio (Strasser et al. 2000, 2004) of collected lichens were taken as stress indicators. After transfer to the laboratory in paper bags, samples were kept overnight in a climatic chamber to ensure rehydration and were dark-adapted at room temperature for 15 min before measuring fluorescence. Fv/Fm ratio and PI were measured with the Plant Efficiency Analyzer Handy PEA (Hansatech LTD, King’s Lynn, UK) in the youngest part of the thallus (up to 1 cm from the margin). Fifteen replicates were taken for each treatment.

Fungal viability

Mycobiont viability was assessed according to Bačkor and Fahselt (2005) measuring the reduction of triphenyltetrazolium chloride (TTC) to triphenylformazan (TPF) as indicator of dehydrogenase activity (dark respiration). About 15 mg of lichen material was incubated for 20 h in 2 mL of 0.6% TTC and 50 mM phosphate buffer solution at room temperature in the dark. Samples were then removed from the solutions and washed in distilled water. Water-insoluble formazan was extracted with 6 mL of ethanol at 65 °C for 1 h. Tubes were then centrifuged at 4000g for 10 min and absorbance read at 492 nm (Tecan Spectra Rainbow microplate reader) in the supernatant. Results were expressed as absorbance units g−1 (DW). Nine replicates, three thalli per plot, and three replicates per thallus were measured.

Surface pH

After transfer to the laboratory in paper bags, the pH values were measured on the surface of hydrated thalli of C. portentosa with a portable pH meter (Crinson PH25). Fifteen replicates per treatment were used.

Statistics

Significance of differences (P < 0.01) in physiological parameters between treatments and controls was checked by one-way analysis of variance (ANOVA), using the Dunnett test for post hoc comparisons with Statistica 13.0 (Stasoft Inc.). Prior to analysis data not matching a normal distribution (Shapiro–Wilk W test at the 95% confidence interval) was log transformed. Homogeneity of variances was checked with the Levene test.

Protein extraction and determination

Lichen proteins were extracted according to Wang et al. (2006). The protocol is effective in the extraction of proteins from recalcitrant tissues, such as lichens (Nicolardi et al. 2012). Reagents were purchased from Sigma-Aldrich, and samples were processed simultaneously to minimize variability. Protein concentrations were determined using a commercial kit (2-D Quant Kit, GE HealthCare), performing the protocol exactly, as described in the instruction manual, using BSA as reference. Each sample and controls were analyzed in three replicates using a Shimadzu UV-160 spectrophotometer at 480 nm.

Two-dimensional electrophoresis (2DE)

For 2DE analysis, 11-cm IPG Strips with a 3–10 pH gradient (Bio-Rad) were used in combination with 10% Criterion XT gels (Bio-Rad). Strips were rehydrated in the solubilization buffer (40 mM Tris, 8 M urea, 2 M thiourea, 2% CHAPS, and traces of bromophenol blue) to which 18 mM DTT and 20 μL/mL IPG buffer were added. Samples were dissolved at 2 mg/mL in the solubilization buffer. Strips were rehydrated overnight in Immobiline Dry Strip Reswelling Tray (GE HealthCare) and covered with the Dry Strip Cover PlusOne (GE HealtthCare). Strips were run using a Multiphor II apparatus at 300 V for 15 min, from 300 to 3500 V for 1.5 h, and at 3500 V for additional 2.5 h. Strips were stored at −80 °C or used immediately. In both cases, strips were equilibrated for 15 min in equilibration buffer (50 mM Tris–HCl pH 8.8, 6 M urea, 30% glycerol, 2% SDS, bromophenol blue, and 10 mg/mL DTT). Proteins were then separated in the second dimension based on a Bis–Tris buffer system (pH 6.4) that uses discontinuous chloride and Mes or Mops ion fronts to form moving boundaries to stack and then separate denatured proteins by size. Molecular weight standards of the precision series (Bio-Rad) were run in parallel. Gels were stained with Bio-Safe Coomassie Blue, as described in the instruction protocol.

Spot analysis

Images of gels were captured using the Fluor-S Multi-Imager (Bio-Rad). The exposure time was 5–7 s for gels stained with Coomassie. Analysis of spots in 2DE gels, derived by three replicates of singles NO3 − and NH4 + treatments, was performed using the Spot Detection Wizard of PDQuest (Bio-Rad) by selecting the weakest protein spot and the larger protein clusters. Subsequently, spot analysis was improved manually by adding unidentified spots and by removing incorrect signals. We first created a so-called Master (virtual) gel, which represents all spots as detected in all samples. Spots were then matched to determine qualitative and quantitative differences. Further analysis of spots was done using the spot and match set tools. The intensity of protein spots was normalized in relation to the total abundance of effective spots. After normalization and background subtraction, gels from control and treated samples were used to create a match set, which allowed the differential expression analysis between treated and control samples. The protein spots were selected according to their relative concentration trend. The fundamental criterion for removing any arbitrary or random selection was that spots should exhibit a difference of relative intensity between maximum and minimum values of at least a factor of 2. In particular, we selected those spots showing a bell-shaped trend (direct or reverse), namely, spots whose relative intensity increased at the beginning of treatment and then decreased at the end or exhibiting an inverse behavior. Other marked spots showed a continuous increase in their relative concentration or decreased constantly from beginning to end of treatment. Theoretical pI and molecular mass were eventually determined using the free web-based tool “Compute pI/Mw”.

Protein identification by mass spectrometry

Protein identification was performed as previously described (Hellman et al. 1995; Soskic et al. 1999). Spots of interest were manually excised, destained in 2.5 mM ammonium bicarbonate and 50% (v/v) acetonitrile, and then dehydrated in acetonitrile. Gel pieces were rehydrated in trypsin solution (Sigma Aldrich), and in-gel protein digestion was performed by an overnight incubation at 37 °C. 1.25 μL of each protein digest was directly spotted onto the MALDI target, air-dried, covered with 0.75 μL of matrix solution (a solution of 5 mg/mL alpha-cyano-4-hydroxycinnamic acid in 50% (v/v) acetonitrile and 0.5% (v/v) trifluoroacetic acid), and allowed to dry again. Acquisition of mass spectra was performed using an Ultraflex III MALDI–TOF/TOF mass spectrometer (Bruker Daltonics, Billerica, MA, USA). Spectra were analyzed by the Flex Analysis software v.3.0. Peptide mass fingerprinting (PMF) database searching was carried out in NCBInr for C. portentosa and/or Swiss-Prot/TrEMBL databases set for fungi and/or Viridiplantae (Green Plants) and/or all entries using the Mascot (Matrix Science Ltd., London, UK, http://www.matrixscience.com) online available software. The search settings were: 100 ppm as mass tolerance, trypsin as the digestion enzyme with one allowed missed cleavage, carbamidomethylation of cysteine as a fixed modification, and oxidation of methionine as a variable modification. Identifications were accepted based on the number of matched peptides, the extent of sequence coverage, and the probabilistic score.

Peptide digests for which the identification was uncertain were further subjected to peptide sequencing by tandem mass spectrometry. MS/MS analysis was performed on the above-cited instrument. The MS/MS database searching was carried out in NCBInr and/or Swiss-Prot/TrEMBL databases using the online-available MASCOT MS/MS ion search software. The search parameters were: fungi and/or Viridiplantae (Green Plants) and/or all entries as taxonomy, trypsin specificity, one missed cleavage allowed, peptide precursor mass tolerance: ±100 ppm, fragment mass tolerance: ±0.6 Da, peptide precursor charge state: +1, carbamidomethylation of cysteine as a fixed modification, and oxidation of methionine as a possible modification. Only peptides with individual ion scores P < 0.05 were significant.

Results

Physiological parameters

Table 2 shows the results of the photobiont and mycobiont viability tests and of pH measurements. Values of Fv/Fm and PI in the treated samples were not different from the control. Similarly, the dehydrogenase activity tended to be higher when thalli were treated with the lowest N dose, but was significantly different from the control only for 16 NO3 − PK. Likely, the lack of a clear pattern was due to the wide intraspecific variability observed in each treatment.

Table 2 Mean values (±standard deviation) of the chlorophyll a fluorescence parameters Fv/Fm and PI (n = 15), the surface pH (n = 15), and the mycobiont viability expressed as absorbance at 492 nm on a dry weight basis (n = 9) for each treatment
Full size table

The pH was the only parameter clearly affected by the N supply. As expected from their chemical characteristics, the reaction of the treatment chemicals with H2O in a humid environment led to the alkalinization and acidification of the lichen surface with NO3 − and NH4 +, respectively.

Proteomic analysis

The proteomic profile obtained by 2DE electrophoresis identified 12 spots from NO3 − treatments (Fig. 1; Table 3) and 15 spots for NH4 + treatments (Fig. 2; Table 3), whose abundance differed at least two-fold between treatments.

Fig. 1
figure 1

Master gel obtained from the PDQuest-based comparison of 2D electrophoresis gels of NO3-treated samples (control NO3 −, 16 NO3 −, 32 NO3 −, 64 NO3 −, 16 NO3 − PK, and 64 NO3 − PK). The virtual image of gel represents all spots as detected in all samples tested. Molecular mass standards are indicated on the left, while the pH range is at the top. Equivalent protein contents (200 µg) were loaded in each sample. Spots analyzed by MALDI–TOF MS are circled and numbered

Full size image
Table 3 Proteins identified by MALDI–TOF and/or MALDI–TOF/TOF mass spectrometry
Full size table
Fig. 2
figure 2

Master gel obtained from the PDQuest-based comparison of 2D electrophoresis gels of NH4-treated samples (control NH4 +, 16 NH4 +, 32 NH4 +, 64 NH4 +, 16 NH4 + PK, and 64 NH4 + PK). The virtual gel represents all spots as detected in all samples tested. Molecular mass standards are indicated on the left, while the pH range is at the top. Equivalent protein contents (200 µg) were loaded in each sample. Spots analyzed by MALDI–TOF MS are circled and numbered

Full size image

The comparison of gels showed a clear difference in protein composition between treated samples and controls as shown by the relative abundance of individual spots (Fig. 3).

Fig. 3
figure 3

Graphs of relative abundance of individual spots in the different samples analyzed. Spots are numbered exactly as shown in the figures of master gels. Spots from 1 to 12 are analyzed in NO3 − treatment, and spots from 13 to 27 are identified in NH4 + treatment. Each spot is presented with the value of its relative quantification. The first four columns indicated the progressive treatments (control, 16, 32 and 64) while the last two columns indicate the two conditions at 16 + PK and 64 + PK. Intensity of spots is reported in y-axis as integrated density (that is the sum of the pixel values in the image or selection is the equivalent to the product of the area and mean grey value). As example is reported only in the first spot

Full size image

Proteins matching fungal peptides were considered to belong to the mycobiont, while those matching plant peptides were considered to belong to the photobiont, and it was assumed that these made up almost the total lichen biomass.

Spots 14 and 17, 25 and 26, and 4 and 10 are couples of different spots identified as the same proteins. They have a similar or complementary pattern, suggesting that they are isoenzymes with a compensatory activity. Alternatively, they can be specific modifications of a single protein, or modifications induced by the extraction in the analysis phase. In any case, the couples were considered together.

To simplify the description of results, proteins have been clustered in six physiologically important functional categories contributing to different key cellular functions: respiration, photosynthesis, protein synthesis, stress (chaperone/folding/oxidation), regulation, and secretion.

Respiration

The mitochondrial malate dehydrogenase NAD dependent regulates the levels of NADH and oxaloacetate, ensuring adequate levels of NADH. Increase of this protein was identified in treatments with both NO3 − and NH4 + and corresponded to spot 1, spot 25, and spot 26. These spots showed a significant homology with the malate dehydrogenase of fungi and plants. The level of spot 1 was high in controls but decreased considerably after the onset of N exposure and was only detected in 32NO3 − and 64NO3 − samples. Spots 25 and 26 only accumulated to low levels in controls but increased considerably in NH4 + treatments.

The NADH–quinone oxidoreductase moves electrons from NADH, via FMN and iron–sulfur (Fe–S), towards quinone in the respiratory chain. The electron acceptor for the enzyme is ubiquinone. It couples the redox reaction to the translocation of protons, conserving energy in the redox proton gradient. Spot 21 showed homology with oxidoreductase of Gloeophyllum trabeum, a plant fungal pathogen. This protein was expressed in the control, increased in the 16NH4 + treatment, but disappeared in all other treatments, except the 64NH4 + + PK treatment, where it accumulated.

The ATP synthase mitochondrial membrane (F1–F0 ATP synthase complex or V) produces ATP from ADP in the presence of a trans membrane proton gradient generated by the complex of electron transport in the respiratory chain. Spot 9 in NO3 − treatments resembled the ATP synthase beta subunit of different organisms, such as yeast. The function of this enzyme, which falls with the highest N dose, may be closely related to the enzymes described above.

Photosynthesis

Both spots 2 and 15 were identified as chloroplastic oxygen-evolving enhancer protein 1, which stabilizes the cluster of manganese, the primary site of water splitting. The expression of this protein is similar in controls and at the lowest concentration of NO3 − but decreases at the highest concentration and in the presence of PK, while the enzyme is much more expressed in the NH4 + treatments than in controls but decreases with PK, notably the 64 treatment.

Spots 3 and 19 were identified as chloroplastic ATP synthase subunit beta which produces ATP from ADP in the presence of a trans membrane proton gradient. The catalytic sites are hosted primarily by the beta subunits. We noted an increase in the enzyme’s amount at intermediate doses of NO3 −, while the enzyme decreased with increasing NH4 +.

The chloroplastic fructose–bisphosphate aldolase 1 is not directly involved in photosynthesis, but in carbohydrate degradation and glycolysis, mediating the conversion of d-glyceraldehyde 3-phosphate and glycerone phosphate from d-glucose. This protein was identified only in NO3 − treatments and corresponds to spot 5. The expression was highest at the intermediate dose of NO3 −.

Protein synthesis

Spot 22 corresponded to prolyl-tRNA synthetase of Bradyrhizobium japonicum, a legume-root nodulating microsymbiotic nitrogen-fixing bacterium species. This enzyme catalyzes the attachment of an amino acid to its cognate transfer RNA molecule in a highly specific two-step reaction. This enzyme showed much higher levels in both N treatments than in the control but decreased in PK treatments.

Stress (chaperone/folding/oxidation)

The first protein identified in the stress functional group is an enzyme belonging to the superfamily of peptidyl–prolyl cis/trans-isomerases (PPIases), which possess chaperone activity and can fold proteins into active configuration by catalyzing slow cis/trans isomerization on proline–peptide bonds. This protein was present in both N treatment forms. In response to NO3 −, sequence correspondence with spot 7 was identified, while in NH4 + treatments, spot 24 corresponded to this enzyme from different organisms, such as fungi and yeast. In NO3 − treatments, this enzyme showed a bell-shaped trend. In NH4 + treatments, a different trend was noted, with lower expression in the treatments than in the control, but an increase following the increased dose of treatment.

In both N forms, we identified disulfide isomerase corresponding to three spots, spot 12 in NO3 − treatment and spots 14 and 17 in NH4 + treatment. The protein-disulfide isomerase participates in the folding of proteins containing disulfide bonds and it may be involved in glycosylation, prolyl hydroxylation, and triglyceride transfer. For all spots, a correlation was found with the enzyme of Humicola insolens. In NO3 − treatments, this protein was only detected in the 64NO3 −-PK sample, whereas with NH4 + treatments, two different spots corresponding to this protein were found. The expression of these two spots is complementary (spot 14 expressed and spot 17 not expressed in 64 NH4 + treatment) that globally indicated an increased level of this enzyme in treatments with respect to the control.

A further protein identified was the major protein of the Woronin body of the dimorphic fungus Paracoccidioides brasiliensis Pb18. This is the main protein in hexagonal crystals of bodies of Woronin, a peroxisomal vesicle that seals the septum pores in response to cellular damages. The protein can interact with the 78 k subunit of the NADH–ubiquinone oxidoreductase. Significant reductions in this protein occurred at the lowest and highest N doses.

Thiol-specific antioxidant protein 3 or cysteine peroxiredoxin has peroxiredoxin activity and was identified in spot 11 with NO3 − treatments, but only in the presence of PK.

Spot 18 in NH4 + treatments was identified as cell division protein 48 (Cdc48). This chaperone protein is involved in spindle breakdown, degradation of ubiquitinated (conjugated to the protein ubiquitin) proteins, also in the endoplasmic reticulum-associated degradation (ERAD) pathway, extraction of incompletely synthesized nascent chains destined to proteosomal degradation, and protein export from the endoplasmic reticulum to the cytoplasm. It has been suggested (Wolf and Stolz 2012) that cdc48 contributes to endoplasmic reticulum membrane maintenance under exposure to stress. This protein showed a bell-shaped response trend to N dose but was highly expressed with 16NH4 + + PK.

Regulation

The protein classified as Ran-specific GTPase-activating protein 1, which stimulates the GTPase activity in the presence of RNA1, was found to correspond with three different spots, two of them in NO3 − treatments (spot 4 and spot 10) and one (spot 16) in NH4 + treatments. The expression level of spot 4 and spot 16 decreased with more treatment, while the opposite effect was found for spot 10, whose expression increased during treatment.

Secretion

Clathrin light chain was identified in Talaromyces marneffei as corresponding to spot 13 in NH4 + treatment. This protein is the main protein of the coating of coated vesicles, active in intracellular transport. The expression was low in the control, but rapidly increased with treatment dose falling back to low values in the presence of PK.

Four spots were identified without specific correspondence to databases; in fact, spots 6, 8, 20, and 27 had homology with hypothetical proteins or with proteins, whose function is putative. In this case, it is impossible to determine the functional significance of this protein in either treatment. Table 4 and Fig. 4 summarize the changes in protein expression caused by the addition of NO3 − or NH4 +.

Table 4 Summary of changes in protein expression caused by the addition of NO3 − or NH4 +
Full size table
Fig. 4
figure 4

Simplified summary of changes in protein expression caused by the addition of NO3 − or NH4 + (black upregulation, brown downregulation, and violet no defined pattern). In each rectangle representing the lichen, the green block is the alga and the orange block is the fungus. The figures show the main organules, namely, chloroplasts, mitochondria, Golgi apparatus and endoplasmic reticulum, and enzymes responding to N treatments. NH + 4 ammonium, NO − 3 nitrate, aa amino acids, GS/GOGAT glutamate synthase/glutamine oxoglutarate aminotransferase, ATPase ATP synthase subunit beta, CLA clathrin light chain, CDP48 cell division control protein 48, FBA fructose-bisphosphate aldolase, MDH malate dehydrogenase, NADHox NADH–quinone oxidoreductase, OEP1 oxygen-evolving enhancer protein 1, PDI protein-disulfide isomerase, PPIases peptidyl–prolyl cis–trans isomerase, RGTPase Ran-specific GTPase-activating protein 1, WB Woronin body major protein

Full size image

Discussion

External supply of different forms of N caused changes in the protein profile of C. portentosa (Fig. 4). The fungus seemed to respond more than the algal partner, with the majority of the under/over expressed proteins belonging to the fungus. Similarly, Hogan et al. (2010b) showed an increase in fungal extracellular phosphatase activity in the outer regions of the thallus of C. portentosa with increased N and N:P ratio.

This is to be expected due to the large biomass of the fungal partner in the symbiotic association (Smith and Douglas 1987) and its role as a barrier between the algal cells and the external environment. In fact, the fungus absorbs most of the N from external sources (Pavlova and Maslov 2008), being nutrients primarily dissolved in the fungal apoplastic continuum and then absorbed by the different lichen symbionts (Dahlman et al. 2004). The direct effect of the supplied N on the fungal cortex is also supported by the changes in pH values observed on the lichen surface (Table 2).

The algal partner only showed different levels of the enzymes responsible for energetic metabolism and photosynthesis, such as increased subunit beta of the ATP synthase with both N forms and fructose–bisphosphate aldolase treated with NO3 − at intermediate N doses. Trebouxia and Asterochloris sp. algae are characterized by a large chloroplast that almost completely fills the cell. Thus, it is not surprising that the observed changes were found amongst the photosynthetic proteins of the photobiont. Alteration of the photosynthetic pathway, including changes in the levels of ATP synthase, is one of the most common effects of metal contamination in higher plants (Bah et al. 2010; Kosová et al. 2011), supporting the idea that producing higher levels of energy represents a generic response to stress.

Another enzyme linked to photosynthesis, the oxygen-evolving enhancer, was downregulated with NO3 − and upregulated with NH4 + (spots 2 and 15). This is understandable if we believe that NH4 + can interfere with water splitting, since ammonia acts as a water analogue in the water oxidizing cycle (Evans et al. 2005) and that this effect can be counterbalanced by stabilizing the manganese cluster (Offenbacher et al. 2013).

The fact that algae functionality was not compromised by the N addition, is supported at the physiological level by the analysis of chlorophyll a fluorescence, and commonly used as a vitality index in lichens (Munzi et al. 2012, 2014). In fact, Fv/Fm and PI in our samples were not affected by N addition (concentration or form).

When analyzing the lichen response to the distinct concentrations of N addition, NO3 −-treated samples showed a poorly defined pattern, except for enzymes responsible for energetic metabolism, which followed a bell-shaped curve with the highest expression at the intermediate concentration. Several other proteins had their lowest level of expression in response to the highest N dose irrespective of the N form (Fig. 3). This pattern suggests the existence of a threshold of N concentration/dose beyond which the homeostasis mechanisms change or just stop working, provoking the down regulation of proteins upregulated at lower N concentrations. This observation finds support in the findings of Carreras et al. (1998) which showed a fertilizing effect of traffic-related N pollutants at low concentrations in Usnea sp. Similarly, Munzi et al. (2013) also found that when the species-specific N tolerance is exceeded, the toxic effects over-rode the positive feedback between N availability and photosynthesis. It is also compatible with the observation that repeated exposures can exacerbate the toxic effects of N (Sheppard et al. 2009).

Like NO3 −, NH4 + treatments in C. portentosa increased the expression of ATP synthase (chloroplast) (spot 19), ATP synthase subunit beta (mitochondrial) (spot 9), and particularly malate dehydrogenase, mitochondrial (spots 25 and 26), which are all part of pathways for energy production.

Taking into consideration that the incorporation of NH4 + into organic compounds requires less energy than that of NO3 − (Raven et al. 1992), this higher demand for energy must be linked to other cellular activities than NH4 + assimilation. Coherent with it, a marked increase in post-transcriptional regulation and folding (spots 14 and 17), and transport/secretion and degradation (spots 13 and 18) protein in the fungus exposed to NH4 + was observed.

Lichen species with the same photobiont can have different tolerance degrees to N (Nimis and Martellos 2008; Pirintsos et al. 2009; Munzi et al. 2014) as shown also for other stresses like heavy metals (Branquinho et al. 2011). Together with the fact that the fungus mediates the interaction between the external environment and the photobiont, this suggests that the mycobiont and not the photobiont is primarily responsible for the tolerance degree to N. In this sense, the altered protein regulation observed in the mycobiont in response to N treatments can be part of specific mechanisms responsible for N tolerance in lichens supported by a higher energy production by the photobiont. Accordingly, the lack of significant changes in the dehydrogenase activity (Table 2) confirmed that the functionality of the fungal partner was not impaired by the amount of treatment supplied.

On the other side, it is well known in higher plants that energy-dissipating cycles, such as the production of non-functional proteins, are a main symptom of N excess (Britto and Kronzucker 2002) and in particular of NH4 + toxicity (Britto et al. 2001). The activation of a futile cycle with production and degradation/loss of proteins and a consequent dissipation of energy in lichens treated with NH4 + would be in agreement with our finding and would explain data from the literature indicating that NH4 + is more harmful to lichens than NO3 − (Munzi et al. 2009; Pirintsos et al. 2009; Hauck 2010).

One of the toxic effects of N on lichens is cell membrane damage and consequent leakage of important cellular cations, i.e., K and magnesium (Mg) (Munzi et al. 2013). The loss of essential cations, such as K, calcium, and Mg, is a well-documented change induced in higher plants by NH4 + nutrition (Britto and Kronzucker 2002), while providing K can counterbalances NH4 + toxicity, for example, improving cell membrane stability and thus the ability to control the concentration of the osmoticum (Premachandra et al. 1991). However, the effect of external PK supply in our experiment was not generic and influenced differently the responses of different enzymes, some of them randomly. This is also predictable, since K is known to play an active role in many different physiological processes (Szczerba et al. 2008) and in the activation of different enzymes and stabilization of protein synthesis in plants (Suelter 1970).

Finally, a putative outer membrane autotransporter (spot 20), the hypothetical protein DSC_01120 (spot 8), a predicted oxidoreductase related to nitroreductase, and a regulator of protein synthesis (spot 22) seem to belong, respectively, to an amoeba the first and to bacteria the others. Since the experiment has been carried out in an open field, the presence of microorganisms growing on lichen surface is not unexpected. Even considering the uncertainties in the identification of these proteins, it is interesting that their pattern presented two of the main characteristics observed in the lichen: the enzymes responsible for protein transportation and synthesis were upregulated with increasing NH4 + like in the fungus and presented a threshold at the highest N concentration. Conversely, the oxidoreductase drastically decreased with PK fertilization (spot 8). Although the presence of the lichen microbiome (bacteria, fungi, algae, and other microorganisms) could be considered a confounding element in our study, actually it provided an insight on the functional response of the whole lichen complex, the holobiome. In fact, even if mainly composed of the mycobiont and photobiont, lichens are complex micro-ecosystems, where, as in higher plants, associated bacteria have recently been found to be functional contributors to robustness of the lichen association (Cernava et al. 2015), although their functions are still largely unknown in comparison with the fungal and algal partners (Grube et al. 2015). Another example of this complexity is provided by a recent study (Spribille et al. 2016) that indicates two unrelated fungi found in the cortex being responsible for phenotypic variations in lichen species.

Conclusions

This work provides evidence at the molecular level that different N forms affect different metabolic pathways in lichen species. In particular, the most relevant changes in protein expression were observed in the fungal partner, with NO3 − mostly affecting the energetic metabolism and NH4 + affecting transport and regulation of proteins and the energetic metabolism much more than NO3 − did. Due to the sensitivity of the method used, only the greatest changes in proteome were highlighted, likely leading to underestimate the role of the photobiont. However, although a much finer regulation in proteins expression can be expected, these results provide a deeper insight into the effects of N on lichens than previously provided. That contributes to understand previous ecological and physiological observations and to explain why in nature different lichenized fungi with the same photobiont species can have different degrees of tolerance to N.

Author contribution statement

SM, CB, and LJS planned and designed the research. SM, LJS, and IDL conducted fieldwork. LJS and IDL managed the experimental site. GC, LP, AG, and LB performed proteomic analyses. SM, CB, CC, GC, and LP analyzed and interpreted the data. SM, CC, GC, and LP wrote the manuscript. All authors revised the manuscript and provided editorial advice.