Difference in Nitrogen Starvation-Inducible Expression Patterns among Phylogenetically Diverse Ammonium Transporter Genes in the Red Seaweed Pyropia yezoensis ()
1. Introduction
Ammonium (
) is a major nitrogen source for higher plants and algae [1] [2] [3] and is used in the biosynthesis of nitrogen-containing compounds such as amino acids and nucleic acids [4] [5]. The influx of extracellular
into cells is mediated by ammonium transporters (AMTs) [6]. In plants, AMTs are encoded by a multigene family comprising the AMT1 and AMT2 subfamilies [2] [7]. AMT1s are responsible for high-affinity
transport in rice (Oryza sativa), wheat (Triticum aestivum) and many other plant species [5] [7] [8] [9] [10]. Genes encoding AMT1s show different expression patterns. For instance, in Arabidopsis thaliana, AtAMT1; 1 is expressed in the roots and leaves [4] and the expression of AtAMT1; 2/1; 3 and AtAMT1; 5 is mostly restricted to roots [11] , while AMT1; 4 shows pollen-specific expression [12]. In addition, AtAMT1; 1 and AtAMT1; 3 are induced by nitrogen starvation, whereas AtAMT1; 2 expression is insensitive to nitrogen deficiency [13]. A recent study showed that the protein encoded by AtAMT2; 1 transfers
from roots to shoots [14] , but the physiological functions of AMT2s are less understood than those of AMT1s.
In the red seaweed Pyropia yezoensis, nitrogen limitation induces severe discoloration in the gametophytic thallus due to a 20% - 50% reduction of the pigment content [15]. Discoloration in thallus decreases its quality as a food [16] [17]. This discoloration can be rescued by increasing the nitrogen concentration in the medium [15] [16]. Of the various nitrogen sources,
is preferentially used over
, urea, and other organic nitrogen sources in P. yezoensis [15] [16] , suggesting the importance of AMTs for nitrogen homeostasis in this species. Genomic and transcriptomic analyses have identified several AMT genes in algae, including the green algae Chlamydomonas reinhardtii [3] [18] and Volvox carteri [19] , the red algae Galdieria sulphuraria [20] and Porphyra umbilicalis [21] , and the diatom Cylindrotheca fusiformis [22]. Of these, expression of one of the AMT genes in Po. umbilicalis and CfAMT1 from Cylindrotheca fusiformis was up-regulated under low-nitrogen conditions [22] [23]. In contrast, AMT genes in P. yezoensis are currently restricted to PyAMT1, whose expression is also induced under nitrogen-deficient conditions.
The P. yezoensis life cycle consists of gametophyte and sporophyte generations [24] [25]. The nitrate transporter gene PyNRT2 shows gametophyte-specific expression [26]. Among three urea transporter genes (PyDUR3.1, PyDUR3.2, and PyDUR3.3) in P. yezoensis, PyDUR3.3 exhibits sporophyte-specific expression that is nutrient deficiency independent; however, nutrient deficiency-inducible expression was observed for PyDUR3.1 and PyDUR3.2, although they have expressed generation independently and gametophyte specifically, respectively [26] [27]. The expression of PyAMT1 is gametophyte specific and is regulated both temporally and by nitrogen deficiency stress [15]. Notably, nitrogen deficiency-inducible expression of PyAMT1 is strongly suppressed by addition of
compared to urea and other amino acid compounds [15].
These findings led us to hypothesize the presence of an AMT gene family in P. yezoensis, possibly with differential regulation of each gene. Here, we demonstrated that there are indeed multiple AMT1 genes in P. yezoensis, with diversity in both phylogenetic relationships with other plant and algal AMTs and expression patterns during the life cycle and under nitrogen-deficient conditions
2. Materials and Methods
2.1. Algal Samples and Culture Conditions
Gametophytes, conchosporangia and sporophytes of Pyropia yezoensis (strain U-51) were maintained in sterilized artificial seawater (SEALIFE; Marinetech, Tokyo, Japan) enriched with ESS2 containing NaNO3 as a nitrogen source, vitamins, and trace metal elements (Table S1) [28]. The algae were grown under 60 μmol photons·m−2·s−1 light in a short-day photoperiod (10 h light/14 h dark) at 15˚C with air filtered through a 0.22-μm filter (Whatman; Maidstone, UK). The culture medium was changed weekly. For nitrogen starvation experiments, gametophytes, conchosporangia and sporophytes were treated with artificial seawater without ESS2 (free of a nitrogen source) for a week. Algal materials were sampled daily after starting the starvation treatment to measure gene expression.
2.2. Quantification of Photosynthetic Pigments
Gametophytes and sporophytes were treated with N-free (ESS2-free) seawater for 3, 5 or 7 days to observe discoloration. For recovery from discoloration, gametophytes and sporophytes discolored for a week were transferred into seawater supplied with 500 μM of NH4Cl, NaNO3, or urea and then cultured for a further week. Discolored and recovered samples (0.1 g fresh weight per sample) were used to calculate chlorophyll a (Chl a) contents according to Seely et al. [29] and phycoerythrin (PE) and phycocyanin (PC) contents as described by Beer and Eshel [30].
2.3. Identification and Characterization of AMTs
Unigenes annotated as putative AMTs were selected from our transcriptome analyses of P. yezoensis [31] , and their identity was confirmed by comparison of predicted amino acid sequences with those of known AMTs by a BLAST search (https://blast.ncbi.nlm.nih.gov/Blast.cgi) after identification of full-length open reading frames (ORFs) with the ORF finder (https://www.ncbi.nlm.nih.gov/orffinder/). The ProtParam tool
(https://web.expasy.org/protparam/) was used to predict the molecular weights, theoretical isoelectric point (pI), and grand average of hydropathicity (GRAVY). The location of the ammonium transporter (AMT) domain was identified with Pfam (http://pfam.xfam.org/search#tabview=tab0), and transmembrane helices in the conserved AMT domain were predicted using a SMART search (http://smart.embl-heidelberg.de). In addition, three-dimensional structures of AMT domain-containing proteins were predicted with the Phyre2 Server (http://www.sbg.bio.ic.ac.uk/phyre2/html/).
2.4. Phylogenetic Analysis
AMTs used for the phylogenetic analysis were obtained from GenBank, genome and EST databases and our unpublished transcriptome analyses are listed in Table S2 with their accession numbers and gene IDs. These included AMTs from Streptophyta (Arabidopsis thaliana, https://www.arabidopsis.org; Physcomitrella patens, https://genome.jgi.doe.gov/Phypa1_1/Phypa1_1.home.html), Rhodophyta (Porphyra umbilicalis, https://phytozome.jgi.doe.gov/pz/portal.html#; Porphyra purpurea, https://www.ncbi.nlm.nih.gov/sar/SRX100230; Porphyridium purpureum, http://cyanophora.rutgers.edu/porphyridium/; Cyanidioschyzon merolae, http://merolae.biol.s.u-tokyo.ac.jp; Galdieria sulphuraria, http://plants.ensembl.org/Galdieria_sulphuraria/Info/Index), Chlorophyta (Chlamydomonas reinhardtii, https://genome.jgi.doe.gov/Chlre4/Chlre4.home.html; Volvox carteri f. nagariensis, https://www.uniprot.org/proteomes/UP000001058), and Heterokontophyta (Phaeodactylum tricomutum, https://genome.jgi.doe.gov/Phatr2/Phatr2.home.html; Cylindrotheca fusiformis, https://www.uniprot.org/uniprot/?query=cylindrotheca+fusiformis&sort=score). A neighbor-joining phylogenetic tree was constructed with MEGA 7 software (https://www.megasoftware.net) using ClustalW to align the AMT amino acid sequences.
2.5. Total RNA Extraction and cDNA Synthesis
Total RNA was separately extracted from gametophytes, conchosporangia and sporophytes using the RNeasy Plant Mini Kit (Qiagen, Hilden, Germany) and then treated with DNase (TURBO DNA-free TM kit, Invitrogen, Carlsbad, USA) to remove genomic DNA contamination. Then, first-stand complementary DNA (cDNA) was synthesized from 300 ng of total RNA with the PrimeScript 1st strand cDNA Synthesis Kit (TaKaRa Bio, Kusatsu, Japan) according to the manufacturer’s instructions. Before being used as a template in quantitative PCR (qPCR) analyses, the quality of the cDNA was evaluated by amplification of the P. yezoensis 18S rRNA gene with its gene primer set (Table S3) [27] via PCR reactions with Phusion high-fidelity DNA polymerase with GC buffer (Biolabs, Massachusetts, USA) according to the manufacturer’s instructions. The thermal cycling parameters consisted of an initial denaturation step 98˚C for 30 s, and 30 cycles of 98˚C for 10 s, 60˚C for 30 s and 72˚C for 20 s, and a final extension step 72˚C for 5 mins.
2.6. Gene Expression Analysis
Primers for qPCR were designed using Primer Premier 5 software (http://www.premierbiosoft.com) and are listed in Table S3. To confirm the sizes of amplified products and applicability of primers, a mixture of three cDNA samples was used with all primer sets for PCR with Phusion high-fidelity DNA polymerase and GC buffer (Biolabs, Massachusetts, USA) according to the manufacturer’s instructions. PCR products were checked by agarose gel electrophoresis. Primer sets that amplified DNA bands with expected sizes were employed for the qPCR. qPCR was carried out in a total volume of 20 μl containing 10 μl of 2 × SYBR Premix Ex Taq GC, 0.4 μl of ROX Reference Dye, 2 μl of cDNA template, and 0.4 μl (10 μM) of each primer, using the SYBR Premix Ex Taq GC kit (Takara Bio, Kusatsu, Japan). The thermal cycling parameters consisted of 95˚C for 5 min and 40 cycles of 94˚C for 30 s, 60˚C for 30 s and 72˚C for 20 s. The dissociation curve was generated by heating from 60 to 95˚C to check for specificity of amplification using an Applied Biosystems 7300 real-time PCR system (Life Technologies, Carlsbad, USA). The data were examined with one-way ANOVA, and the level of significance was defined at P < 0.05.
3. Results
3.1. A multiplicity of P. yezoensis AMT Genes
Based on the functional annotation in our P. yezoensis transcriptome analysis [31] , we identified six unigenes (CL1839, CL3739, Unigene15210, CL1882. Contig5, CL1882. Contig6 and Unigene24155) as candidate P. yezoensis AMT (PyAMT) genes, in addition to the known PyAMT1 [15]. These six additional unigenes contained predicted open reading frames (ORFs) encoding 484, 654, 589, 616, 690, and 522 amino acid products (Figure 1) with molecular masses of 51.2, 67.8, 58.67, 63.68, 72.06, and 53.78 kDa, respectively, and all contained the conserved AMT domain (Table S4). In addition, CL1839, CL3739, Unigene15210, CL1882. Contig5, CL1882. Contig6 and Unigene24155 showed 55.04%, 32.42%, 38.07%, 27.77%, 25.36%, and 16.88% identity to PyAMT1, respectively. Based on these findings, we concluded that all of the candidates are likely to be AMTs. However, the product of Unigene 24155 had 12 predicted transmembrane (TM) helixes in contrast to the products of the other candidate genes, which had 11 TM helixes (Table S5), suggesting a structural difference in Unigene24155 from the other genes. Other structural characteristics of these gene products, including a total number of atoms and theoretical pI, are listed in Table S4.
The crystal structure of the ammonium transporter protein in Escherichia coli (AmtB) revealed that two phenylalanine residues, Phe107 and Phe215, block the hydrophobic
conduction pore, and two highly conserved histidine residues, His168 and His318, maintain the shape of the central pore involved in
transport [32] [33]. As shown in Figure 1, these four sites were highly conserved in P. yezoensis AMTs. In addition, in consistent with Kakinuma et al. [15] , a tripeptide sequence, Phe-Gly-Phe (Tyr/Asn), indicating AMT identity, was found in all of the PyAMTs (Figure 1). Moreover, three-dimensional structures predicted in silico for all of P. yezoensis AMT domain-containing proteins were similar to those of known AMTs. Indeed, crystal structures of 5 PyAMTs (PyAMT1, CL1839, Unigene 15210, CL1882. Contig5, CL1882. Contig6) structurally resembled that of the AMT template c5aexB (Saccharomyces cerevisiae MEP2), while CL3739 was quite similar to the other AMT template c5aezA (Candida albicans MEP2) (Figure 2). Moreover, three-dimensional structures of Unigene24155 were closely related to that of the template c3hd6A (human rhesus
![]()
Figure 1. Conservation of AMT identity among six PyAMT1s and PyRh. Amino acid residues conserved in all seven proteins and in six proteins are highlighted by black (with white characters) and gray backgrounds, respectively. The boxes indicate the Phe-Gly-Phe (Tyr/Asn) triplet conserved in AMTs, as well as phenylalanine and histidine residues corresponding to residues involved in
binding and transport in E. coli EcAmtB. The amino acid numbers are indicated on the right.
protein, Rh, structurally related to AMT) (Figure 2). Taken together with characteristics in primary sequences (Figure 1), these findings highly suggested functional
-transport and Rh activities of AMT domain-containing proteins from P. yezoensis.
![]()
Figure 2. High degree of similarity in three-dimensional structures between AMT domain-containing proteins from P. yezoensis and known AMTs. Templates c5aexB and c5aezA indicate three-dimensional structures of AMTs (MEP2s) from Saccharomyces cerevisiae and Candida albicans, respectively, while c3hd6A represents a three-dimensional structure of the human rhesus protein (Rh). Cov means the coverage percent of AMT domain-containing proteins to the corresponding templates.
3.2. Phylogenetic Classification of P. yezoensis AMTs into AMT1 and Rhesus Protein Subfamilies
To explore what type(s) of AMTs these unigenes encode, we performed a phylogenetic analysis with other full-length amino acid sequences of known AMTs from algae, plants, animals and bacteria. All of the P. yezoensis sequences except for Unigene24155 were placed in the plant AMT subfamily 1 clade. Therefore, we considered these unigenes to encode AMT1s and designated them as PyAMT1.2 (CL1839), PyAMT1.3 (CL3739), PyAMT1.4 (Unigene15210), PyAMT1.5 (CL1882. Contig5), and PyAMT1.6 (CL1882. Contig6). Unigene24155 shared 27.56% and 24.70% identity with CrRh1 and CrRh2 from the green alga Chlamydomonas reinhardtii, respectively, but only 16.88% with PyAMT1. Our analysis placed this protein in the Rh clade, which is phylogenetically divergent from both AMT1 and AMT2 clades (Figure 3). Rh proteins are homologues of AMT proteins that were first identified in human erythroid cells [34] [35]. The predicted 12 transmembrane helixes of the Unigene24155 protein is in accordance with Rh proteins of other organisms [36] [37]. Thus, we designated Unigene24155 as PyRh.
Our phylogenetic analysis also indicated that algal AMTs were mostly classified into the AMT1 subfamily, which is distantly related to the AMT2 subfamily clade. In addition, AMT1s from red algae, green algae, diatoms, and land plants formed independent clades. Thus, an ancient algal AMT1 may have existed prior to the divergence of red and green algae, although the origin of AMT2s is unclear.
The six PyAMT1s were subdivided into three different Rhodophyta clades, each of which contained pairs of PyAMT1s: PyAMT1 and PyAMT1.2, PyAMT1.3 and PyAMT1.4, and PyAMT1.5 and PyAMT1.6 (Figure 3). In addition, the three Rhodophyta clades also contained pairs of AMT1s from the red seaweed Porphyra umbilicalis: The PyAMT1/1.2 clade with Pum0126s0003.1 (OSX77964.1), Pum1775s0001.1 (OSX69172.1), Pum0463s0020.1 (OSX72158.1), and Pum0165s0019.1 (OSX77025.2); the PyAMT1.3/1.4 clade with Pum0027s0002.1 (OSX80976.1) and Pum1656s0001.1 (OSX69292.1); and the PyAMT1.5/1.6 clade with Pum0022s0083.1 (OSX81363.1) (Figure 3). Moreover, AMT1s from Chlorophyta were separated into three different clades, independent from each other and from the three clades of Rhodophyta. Thus, expansion and divergence of the ancient algal AMT1 gene into three groups occurred independently in red and green algae after their establishment.
3.3. Differences in Temporal Expression Patterns of PyAMT1 and PyRh Genes during the Life Cycle
Figure 4 shows the relative transcript abundance of PyAMT1 and PyRh genes in the gametophyte, conchosporangium and sporophyte tissues under normal growth conditions. When life cycle specificity of gene expression was compared between the gene pairs found in the three phylogenetic clades, PyAMT1 was specifically expressed in the gametophyte, while the expression of PyAMT1.2 was found in both the sporophyte and conchosporangium. By contrast, PyAMT1.3 and PyAMT1.4 exhibited the same sporophyte-dominant expression pattern. PyAMT1.6 was expressed constitutively, whereas transcripts of PyAMT1.5 were not detectable at any stage (data not shown). Thus, two of the gene pairs in the same clade did not share the same expression pattern. We did not find evidence of PyRh expression at any point in the life cycle (data not shown).
![]()
Figure 3. Neighbor joining-based phylogenetic tree of AMTs from terrestrial plants and algae. Boxes indicate PyAMTs and PyRh. The bootstrap values with 1000 replicate over 70% are indicated at the nodes of the tree. The DDBJ/EMBL/GenBank accession numbers of AMTs and Rhs used in the phylogenetic analysis are listed in Table S2.
![]()
Figure 4. Differences in temporal expression patterns of PyAMTs in the life cycle of P. yezoensis. The relative mRNA levels of PyAMT1s were normalized with the reference gene 18S rRNA. Error bars indicate the standard deviation of triplicate experiments (n = 3), and different letters on bars indicate significant differences at P < 0.05 tested with ANOVA. G, gametophyte; C, conchosporangium; S, sporophyte.
3.4. Induction of and Recovery from Discoloration
Gametophyte and sporophyte tissues maintained in the ESS2-containing seawater were transferred to seawater without ESS2 and cultivated for an additional 3, 5 or 7 days. As results, discoloration was initially observed after 3 days and gradually strengthened until 7 days in both gametophytes and sporophytes (Figure 5). Correspondingly, the contents of photosynthetic pigments Chl a, PE and PC were decreased respectively in gametophytes from 1.37 to 0.40, 6.42 to 3.18, and 1.27 to 0.35 mg∙g−1 FW and in sporophyte from 1.49 to 0.58, 5.24 to 1.99, and 0.48 to 0.18 mg∙g−1 FW (Figure 6). To examine recovery from discoloration, 7-day-discolored gametophytes and sporophytes were treated with nutrition-deficient medium containing 500 μM NH4Cl, NaNO3, or urea for a week. As shown in Figure 5, discoloration was recovered visibly, which was supported by the increase in the contents of Chl a, PE and PC to the levels corresponding to those in non-discolored gametophytes and sporophytes (Figure 6). These findings suggested essential roles of the AMT activity for recovery from discoloration in P. yezoensis.
3.5. Diversity in the Nutrition Starvation-Inducible Pattern of PyAMT1 Subfamily Genes during the Life Cycle
We examined the nitrogen deficiency-inducible expression of the PyAMT1 genes. As shown in Figure 7, all PyAMT1 genes displayed nitrogen deficiency-inducible expression without alterations in their life cycle stage specificity. However, the expression patterns differed among the genes. For instance, transient induction in gametophytes was observed for PyAMT1, PyAMT1.2, and PyAMT1.3, whereas expression of PyAMT1.2 and PyAMT1.4 gradually increased in both
![]()
Figure 5. Artificial induction of and recovery from discoloration in Pyropia yezoensis. Gametophytes (left) and sporophytes (right) were treated with N-free (ESS2-free) seawater for a week, and sampling was performed at 3, 5 and 7 days to observe discoloration. Control indicates samples cultured in ESS2-enriched seawater. In addition, discolored gametophyte and sporophytes produced in N-free seawater for 7 days were transferred into ESS2-less seawater supplied with 500 μM of NH4Cl, NaNO3, or urea and then cultured for an additional 7 days to examine the recovery from discoloration. Scale bars in left and right panels indicate 1 cm and 50 µm in gametophytes and 1 cm and 500 µm in sporophytes.
![]()
Figure 6. Changes in the contents of chlorophyll a, phycoerythrin and phycocyanin in discolored and recovered P. yezoensis. Culture conditions were identical to those in Figure 5. C means control samples. Error bars indicate the standard deviation of triplicate experiments (n = 3), and different letters on bars indicate significant differences at P < 0.05 tested with ANOVA.
conchosporangia and sporophytes. Moreover, the PyAMT6 expression gradually increased in gametophyte tissue, and transient expression of PyAMT6 was observed in conchosporangia and sporophytes. Transcripts of PyAMT1.5 and PyRh remained undetectable in all tissues evaluated, even under nitrogen-deficient conditions (data not shown).
We further examined the expression of the PyAMT1 genes to determine the effects of nitrogen recovery on their expression. For these experiments, we selected PyAMT1.2 and PyAMT1.4, whose expression continually increased in sporophytes under nitrogen deficiency. When discolored sporophytes produced by 7-day culture in the ESS2-less medium were transferred to ESS2-less medium containing 500 μM NH4Cl and further cultured for 3 days, the expression of the two genes was strongly down-regulated within 24 h (P < 0.05; Figure 8). In addition, the same effect was observed in ESS2-less medium containing 500 μM
![]()
Figure 7. Differences in nitrogen starvation-induced expression patterns among PyAMT1 genes. Samples of gametophytes, conchosporangia, and sporophytes treated with ESS2-free seawater were collected every day for 7 days to examine the expression of the PyAMT1 genes by qRT-PCR. The relative mRNA levels were normalized with the reference gene 18S rRNA. Error bars indicate the standard deviation of triplicate experiments (n = 3), and different letters on bars indicate significant differences at P < 0.05 tested with ANOVA. G, gametophyte; C, conchosporangium; S, sporophyte. The numbers on the X-axis indicate the duration of culture under nitrogen-deficient conditions (days).
NaNO3 or 500 μM urea (Figure 8). These results were consistent with the previous observation [15] that PyAMT1 gene expression is down-regulated by the addition of inorganic and organic nitrogen sources.
4. Discussion
A
is an important nitrogen source that is transported via AMTs, which are ubiquitous plasma membrane proteins and are classified into three subfamilies,
![]()
Figure 8. Suppression of the nitrogen starvation-induced expression of PyAMT1 genes via recovery with different nitrogen sources. Sporophyte samples grown in ESS2-less seawater for 7 days (∆N) were treated with seawater containing 500 μM of NH4Cl, NaNO3, or urea for 3 days. Sporophytes cultured in nitrogen-supplied seawater were collected every day for 3 days, and the expression of PyAMT1.2 and PyAMT1.4 was examined by qRT-PCR (upper and lower, respectively). The relative mRNA levels were normalized with the reference gene 18S rRNA. Error bars indicate the standard deviation of triplicate experiments (n = 3), and different letters on bars indicate significant differences at P < 0.05 tested with ANOVA. S, sporophytes cultured in ESS2-containing seawater. The numbers on the X-axis indicate the duration of culture under nitrogen-deficient conditions (days).
AMT1, AMT2 and Rh [38]. Although the structure, expression patterns and physiological roles of AMT genes have been well studied in animals and land plants [2] [13] [34] [39] , algal AMT genes remain poorly understood. Here, we report the presence of the AMT1 and Rh gene subfamilies in P. yezoensis and the diversity in their expression patterns.
Our phylogenetic analysis demonstrated the diversity of the AMT1 subfamily, consisting of independent phylum-specific clades (Figure 3). Land plants have their own AMT1 subfamily, with five genes in Arabidopsis [4] [11] [12] [13] , at least 10 genes in rice [9] , and 23 genes in wheat [8]. Although AMT1s from land plants formed a single clade, it was separated from the algal AMT1 subfamily, in which the unicellular green algae C. reinhardtii and the red seaweed Po. umbilicalis have 11 and 7 AMT1 genes, respectively [18] [21] , in addition to the 6 PyAMT1 genes. AMTs in Po. umbilicalis have been annotated as AMT1, AMT2, and AMT3 by Brawley et al. [21]. Despite these names, our phylogenetic analysis indicated that all of the AMTs in Po. umbilicalis are AMT1 subfamily members (PuAMT1s) as are other algal genes (Figure 3). Moreover, the six PyAMT1s and seven PuAMT1s were subdivided into three groups (Figure 3). The existence of multiple independent AMT1 clades in P. yezoensis and Po. umbilicalis is the distinguishing characteristic of AMT1s from Bangiales, since AMT1s from land plants formed a single clade (Figure 3). These findings imply that an ancient red algal AMT gene may have diversified into three genes prior to the separation of Pyropia and Porphyra, and then further diversification occurred independently for each of the three genes in these species.
Similar to Rhodophyta, AMT1s of Chlorophyta were divided into three groups, although their phylogenetic positions were different from the Rhodophyta clades (Figure 3). This result is in agreement with the report that three subfamilies of CrAMT1s have been established in Chlamydomonas [3]. Therefore, a two-step diversification of algal AMT1s has been proposed: An early expansion of the ancient algal gene into three variants prior to the divergence of Chlorophyta and Rhodophyta and a late duplication of each of the three variants after the divergence of the green and red lineages.
The multiplicity of AMT1 genes in Chlorophyta and Rhodophyta points to functional divergence of these genes in seaweeds. This hypothesis is supported by our gene expression analyses indicating differences in temporal and nitrogen stress-inducible expression patterns for the PyAMT1 genes (Figure 7 and Figure 8). PyAMT1, PyAMT1.3, and PyAMT1.4 commonly exhibited a transient increase in their expression under nitrogen-deficient conditions, although differences were observed in their life cycle stage-specific expression. Thus, it seems that functional divergence in PyAMT1s might allow for functional specialization over a range of
concentrations, which would enable P. yezoensis to react appropriately to a wide range of
concentrations in the environment. In the future, functional analysis of each PyAMT1 focused on
uptake under nitrogen deficiency conditions should help us to understand their transport capacity and how P. yezoensis responds and adapts to nitrogen deficiency stress during its life cycle.
Nitrogen deficiency results in leaf senescence in land plants [40] [41]. In leaf senescence, nitrogen is reused to support plant growth and reproduction by reallocating from aging leaves to younger tissues [8] [9] and degradation of chlorophylls and following discoloration of tissues occur in aging leaves [42] [43]. Thus, loss of nitrogen and degradation of photosynthetic pigments are responsible for leaf senescence [40] [44]. In macroalgae, nitrogen plays an important role in producing amino acids and photosynthetic pigments such as Chl a, PE and PC [45]. As shown in Figure 5 and Figure 6, nitrogen deficiency-induced discoloration by reducing these three photosynthetic pigments in P. yezoensis. Given the simple architecture of the thallus and conchocelis and the fact that discoloration was observed throughout the entire organism (Figure 5), reallocation of nitrogen is not responsible for the discoloration in P. yezoensis, suggesting that the mechanism behind the discoloration caused by nitrogen starvation is different from that in land plants, although loss of photosynthetic pigments in discoloration is common.
The discoloration in P. yezoensis thalli was rescued and the mRNA level of PyAMT1 was down-regulated by an increase in the concentrations of not only inorganic but also organic nitrogen sources [15]. Similarly, nitrogen deficiency-inducible discoloration and the expression of PyAMT1.2 and PyAMT1.4 in the sporophyte were also strongly repressed after addition of NH4Cl, NaNO3, and urea in the culture medium (Figure 8). Since inorganic and organic nitrogen sources are metabolized into
to assimilate the nitrogen into cellular components [46] [47] , NaNO3 and urea might increase the intracellular
contents and similarly affect the expression of PyAMT1 genes. These findings indicated that discoloration in P. yezoensis during nitrogen starvation is induced by the decreased extracellular nitrogen content, although leaf senescence in land plants occurs as a result of nitrogen reallocation between different tissues [48] [49].
Another distinguishing characteristic of algal
transporters is the presence of Rh proteins, such as PyRh in P. yezoensis and contig2015.11 in Porphyridium purpureum (Figure 3), which contain the conserved AMT domain but are distantly related to the AMT1 and AMT2 subfamilies [36] [50]. Rh split from AMT in archaeal species and coexists in microbes and invertebrates, but not in fungi, vascular plants, and vertebrates [50]. To date, Rh has not been reported in algae except for CrRh1 and CrRh2 from the green alga C. reinhardtii [37] [51] ; our findings reveal the presence of Rh in red algae (Figure 3). Although the expression of CrRh1 and CrRh2 is regulated by CO2 [37] [51] , the expression of PyRh was not detected during the life cycle in P. yezoensis nor under the nitrogen-deficient conditions (data not shown). Thus, little is known about the physiological functions of Rh proteins in algae.
In conclusion, algal
transporters are divided into the AMT1 and Rh subfamilies. The AMT1 subfamily of P. yezoensis consists of three groups containing genes whose expression patterns differ temporally and are nitrogen deficiency-dependent during the life cycle. These findings are novel for algal
transporters, and future work should elucidate the functions of each member of the AMT1 and Rh subfamilies, which could help clarify the unique algal strategies of response and acclimation to nitrogen deficiency stress by phylogenetically independent and diverse AMT1s and Rhs during the life cycle.
Acknowledgements
We are grateful to Mr. Masahiro Suda and Mr. Ryunosuke Irie for their supporting for laboratory culture of P. yezoensis gametophytes and sporophytes. Chengze Li was supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan and by the China Scholarship Council.