Ph1205 sequence analysis and protein structure prediction
The Ph1205 gene derived from P. haitanensis genomic DNA is composed of 3 exons (5’UTR-CDS1, CDS2, and CDS3-3’UTR) (Fig. 1a), with a complete ORF of 1965 bp. It encodes a protein with predicted molecular weight of about 67.18 kDa. Conserved domain search identified Ph1205 belonging to the Golgi pH regulator family N-terminal domain- (GPHR-N, 184 aa-227 aa) and abscisic acid-G-protein coupled receptor (ABA-GPCR, 483 aa-577 aa) domain-containing protein family (G-A family).
Ph1205 contains 9 predicted transmembrane domains (TM1-9) (Fig. 1b). Its N-terminal 14 amino acids, N14, is predicted to be extracellular. Its C-terminus, as long as 74 amino acids, is predicted to be intracellular and contains an extended α helix by secondary structure prediction (Fig. 1b, red underline).
Multiple sequence alignment and evolution analysis of Ph1205 homologous proteins
BLAST analysis showed that Ph1205 shared > 83% amino acid sequence identity with homologous sequences from two red algae of Bangiophyceae, Pyropia yezoensis and Porphyra umbilicalis (Additional file 1: Table S1). But the sequence identity between Ph1205 and those from other selected eukaryotes, including unicellular red algae (Galdieria sulphuraria and Porphyridium purpureum), multicellular red algae Floridophyceae (Gracilariopsis chorda and Chondrus crispus), Chromalveolata, green algae, fungi, plants and animals, are < 37%. Conserved domain search showed that they all belonged to G-A family (Additional file 1: Table S1). However, no orthologs were found in the genomes of unicellular red algae Cyanidioschyzon morulae, some selected prokaryotic cyanobacteria and proteobacteria.
We analyzed TM structures of the eukaryotic G-A family members. The results showed that while their TM distribution patterns were similar, their TM numbers were quite different (Additional file 2: Table S2). Among which, TMs were distributed on both terminals of the sequence, and connected with a large cytoplasmic loop (LCL). In fungi, higher plants and mammals, the TM numbers are relatively consistent, showing the distribution of 5TM-LCL-4TM. However, in some homologous proteins of algae, the predicted TM numbers at both terminals of LCL changed. For example, in red algae, Ph1205 is 5TM-LCL-4TM; while in P. umbilicalis, OSX76732 is 5TM-LCL-1TM; and in G. chorda, PXF46398 is 4TM-LCL-4TM.
We performed multiple sequence alignments among eukaryotic G-A family members. The conserved regions were mainly located in TM domains except TM1 (Fig. 2). GPHR-N domain is located in TM5-CL3. ABA-GPCR domain is located between TM7-TM9 region, and the highly conserved amino acid sites are mainly distributed in TM8-TM9. Additionally, three short conserved motifs were found in the junction region of TM and extracellular loop (EL) (Fig. 2, # mark). For example, [EQ]-[YF]-x-F-[FY] was found in TM9-EL4 junction region. Relatively high frequency of [RK] and [ML] appeared in several TM and cytoplasmic loop (CL) junction regions (Fig. 2, * mark).
In other regions, especially in the two terminal fragments, the homology among G-A family is very low. The homologous proteins of Bangiophyceae, including Ph1205, have the characteristic sequences at both terminals not found in other species. They are longer in length than those derived from animals or plants, and are of less conservatism (Fig. 2). For example, the extracellular N-terminal of Ph1205 (N14 peptide) contains 7 and 4 more amino acids than human GPCR89 and Arabidopsis GTG do, with low identity. Compared to the C-terminus of Arabidopsis GTG, Ph1205 held a longer C-terminal extension fragment of 61 amino acids that may form the tenth predicted helix structure. In addition, some unique long fragments are inserted in loop regions of Bangiophyceae homologous proteins. For example, a long amino acid sequence (23 amino acids) not found in other eukaryotic G-A orthologues was inserted at EL2. Ph1205 also contains unique fragments longer than other eukaryotic members in EL3 and CL3. Furthermore, from several CL or EL regions, e.g. CL3, more amino acid mutations were observed in homologous proteins of red algae or multicellular red algae, but more conservative in those of vesicular algae or green algae in evolution (Fig. 2, orange background).
PROSITE motif analysis identified a conserved ATP-/GTP-binding domain in Arabidopsis GTG (Fig. 2, region 382-411 for GTG1), which is not found in Ph1205 and human GPCR89. Compared with the consensus pattern of GTP-binding motif ([LIV]-G-G-{P}-[FYWMGSTNH]-[SGA]-{PW}-[LIVCAT]-{PD}-x-[GSTACLIVMFY]-x(5,18)-[LIVMFYWCSTAR]-[AIVP]-[LIVMFAGCKR]-K), Ph1205 maintained only seven conserved amino acids (Fig. 2, blue background and @ mark). Other eukaryotic G-A family members except high plant orthologs are also less conserved, especially at the N-terminal two amino acids, which do not conform to [LIV]-G (Fig. 2). In addition, the LCL of Arabidopsis GTG shares a 62%-68% similarity with the degenerate Ras GTPase-activating protein domain ([GSNA]-x-[LIVMF]-[FYCI]-[LIVMFY]-R-[LIVMFY](2)-[GACNS]-[PAV]-[AV]-[LIV]-[LIVM]-[SGANT]-P) (Fig. 2, amino acids 230–243 in GTG1) [14]. Ph1205 has only 28.57% homology in this region (Fig. 2, cyan background and @ mark).
To assess the evolutionary relationship among these proteins, we constructed phylogenetic tree (Fig. 3). The protein from unicellular red alga G. sulphuraria appears to be the earliest origin of G-A family, thereafter they form two clades: one is for red algae, including multicellular red algae and unicellular P. purpureum. Their branching situations are consistent with the evolutionary relationship of their genomes [17-19]. Another gradual evolutionary branch has formed two large clusters, one large group includes mainly homologous proteins of fungi, amoeba and metazoans, which share an ancestral homologous sequence with that of cryptoalgae Guillardia theta and that of a collar flagellate Monosiga brevicollis. Another large group includes mainly the homologous sequences of the green plant kingdom, including green algae, and terrestrial plants. The evolution of the homologous sequences of the brown algae Ectocarpus siliculosus and Oomycetes, lies between the two clades of red algae and other eukaryotes. However, several members of diatoms are in parallel with homologous sequences of G. sulphuraria on the evolutionary tree, being an independent branch.
Analysis of subcellular localization of Ph1205 protein
In order to analyze subcellular localization of Ph1205 protein, we performed three experiments. First, we transfected the HEK293T cells with pCI recombinant plasmids carrying Egfp or the fused Ph1205-Egfp. When only Egfp was expressed in HEK293T cells, Egfp green fluorescence was distributed throughout the cytoplasm (Fig. 4a). However, when the fused Ph1205-Egfp complex was expressed, Egfp green fluorescence was mainly concentrated on the cellular membrane. The green fluorescence was overlapped very well with the red fluorescence derived from the membrane probe Dil with a Pearson correlation coefficient (r) of 0.98. These results indicate that Ph1205 can be selectively localized to the cellular membrane (Fig. 4a).
Then we transfected HEK293T cells with pCI-Ph1205 and immunochemically hybridized cells with N14 polyclonal antibody, produced by a N-terminus peptide of Ph1205, followed by detection with FITC-linked secondary antibody. Fig. 4b showed that the green fluorescence was also concentrated in the cell membrane, and overlapped mainly with the red fluorescence derived from Dil-red with the Pearson correlation coefficient (r) of 0.98, further confirming that the recombinant Ph1205 is localized to the cellular membrane of HEK293T cells (Fig. 4b).
We then prepared protoplast of P. haitanensis and similarly probed the protoplast with the primary antibody of N14. The results showed that the green fluorescence only existed at the edge of P. haitanensis cells and the spontaneous red fluorescence was seen inside P. haitanensis cells, clearly confirming that Ph1205 is indeed localized on the membrane of P. haitanensis cells (Fig. 4c).
The interactions between Ph1205 and candidate compounds
We determined whether the membrane-localized Ph1205 can act as a receptor for binding with candidate compounds. Oligoagar, JA, SA, 1-octen-3-ol, and BL, that have been studied in algae and can cause the resistant responses of algae [3, 4, 6, 20-22], were selected. Phytohormones, including ABA, gibberellic acid (GA), IAA and Flg22, were also selected. Their binding kinetics to N14 polypeptide of Ph1205 were analyzed by FortéBio Octet system. Fig. 5a showed that the N14 polypeptide could interact with SA, oligoagar, JA and Flg22. Except for Flg22, the binding and dissociation of SA, oligoagar and JA showed dose-dependent changes. Their binding constants (Kon), dissociation constants (Koff) and affinity constants (Kd) were listed in Table 1. Binding rates of both SA and Flg22 to N14 were fast, but there was no significant difference in the dissociation rate among three compounds. The Kd values reflected that the binding affinity of SA to N14 peptide was the highest, reaching 5.13E-05, followed by that of Flg22. JA displayed only a weaker binding affinity for N14. Because oligoagar is a mixture with no precise molecular weight, its binding affinity constant could not be calculated. However, oligoagar exhibited a quite high binding affinity to N14. At high concentrations, the binding thickness with the probe reached 0.6 nm.
Table 1. The affinity constants between N14 peptide of Ph1205 and candidate compounds.
Compounds
|
Kon (ms-1)
|
Koff (s-1)
|
Kd (M)
|
Full R2
|
Salicylic acid
|
4. 66E+03
|
2. 39E-01
|
5. 13E-05
|
0. 9889
|
Jasmonic acid
|
4. 88E+01
|
5. 99E-01
|
1. 23E-02
|
0. 9788
|
Flg22
|
3. 82E+03
|
6. 43E-01
|
1. 68E-04
|
0. 6553
|
We also tested their binding to HEK293T cells transfected with pCI-Ph1205. These compounds bound to the probed HEK293T cells (Fig. 5b), but were not completely consistent with their binding to N14 described above. Oligoagar, Flg22 and JA were bound with the cells in a concentration-dependent manner while JA binding was still weaker. The dissociation of oligoagar at moderate and high concentrations were better. Flg22 showed higher binding to probed HEK293T cells. At high concentrations, the probe thickness could reach 0.88 nm. However, SA no longer showed a concentration-dependent binding trend.
CRE-mediated signal transduction in Ph1205-expressing HEK293T cells induced by candidate Compounds
After binding to its ligand, GPCR activates intracellular G-protein and triggers the downstream gene expression via the regulation of CRE present in the cells [23, 24]. Thus, by taking the advantage of the endogenous CRE in cells, we co-transfected the HEK293T cells with PCI-Ph1205 and Dual-Luciferase reporter systems and further screened the compounds binding to Ph1205 (Fig. 5c). Both SA and oligoagar aroused significant signal responses (P < 0.05), with a concentration-dependent trend. At high concentration, SA stimulated the highest signal response, consistent with the SA-N14 interaction (P < 0.01), while treatments of HEK293T cells with Flg22 and JA also caused the increasing fluorescence signals thought not significant.
Transcriptional changes of Ph1205 regulated by candidate compounds
Both JA and oligoagar significantly promoted the mRNA level of Ph1205 gene (Fig. 6). When thalli were stimulated by JA for 0.5 h at 50 μM, Ph1205 transcription was increased by 10.85-fold (P < 0.01). Oligoagar at 50 μg mL-1 for 3 h significantly up-regulated overall Ph1205 expression by 13.45-fold (P < 0.01). However, both SA and Flg22 hardly changed but even down-regulated its transcription.
Transcriptome patterns of P. haitanensis stimulated by SA and oligoagar
The transcriptome sequencing of P. haitanensis after being stimulated with SA and oligoagar was conducted. Comparing two sets of transcriptome data revealed that their effects on types of GO biological processes and KEGG pathways were quite similar (Fig. 7a). For instance, in the GO classification, the three mostly enriched biological process were all Metabolic processes, Cellular process, and Signal-organism process. The mostly enriched in Cellular components were Membrane, and Membrane part, etc. In KEGG Pathway Kierarchy1, the mostly significant differential pathways in Cellular processes were Transport and catabolism. In Metabolism, the mostly significant differential pathways were Carbohydrate metabolism and Energy metabolism. However, comparing the changes in gene transcription profiles between two groups revealed that the up- and down-regulation trends induced by SA and oligoagar were quite different, and many genes have shown even the opposite changes. For instance, in the SA treatment group, the up-regulation pathways were enriched in Carbon fixation photosynthetic organisms, Carbon metabolism, and Pentose phosphate pathway, etc. In the oligoagar treatment group, these pathways were just down-regulated. Thus, we chose Carbon fixation photosynthetic organisms to conduct a detailed comparison. Fig. 7b showed that most of the genes were reversely regulated when P. haitanensis was treated with SA and oligoagar. The heat-map showed this opposition.
There were also significant differences in the expression profiles of the transcriptional factors after being stimulated by these two compounds (Additional file 3: Table S3). For example, after stimulation with oligoagar at 200 μg mL-1 for 1 h, the transcription factors that were significantly up-regulated were mainly those in the C2H2 family and the TIG family. Pha005926 of C2H2 family gene was up-regulated by 247.28-fold (Log2FC = 7.95, q < 0.01); and TIG family gene, Pha005519, was upregulated by 26.72-fold (Log2FC = 4.74, q < 0.01); Under SA stimulation, Pha005926 was even significantly down-regulated by 4.23-fold (Log2FC = -2.08, q < 0.01).
Referring to the SA signaling pathway in plants (http://www.yeastrc.org/pdr/viewGONode.do? acc=GO:0009862), we retrieved genes in this pathway from P. haitanensis genome (Additional file 4: Table S4). However, after SA stimulation, almost no significant transcriptional differences were seen, indicating that the SA signaling pathway in P. haitanensis has not been evolved as that in plants. We also performed homologous search of the G-protein of P. haitanensis and related genes involved in cAMP signaling pathway. No information about heterotrimeric Gα and β/γ-subunit homology in P. haitanensis genome was obtained, but many members of other Ras-like GTPase family were found (Additional file 5: Table S5). After stimulation by SA and oligoagar at 100 μM for 1 h, these genes did not appear to be differentially expressed, except for one ribosome-associated GTPase EngA gene, Pha007322, whose transcript was significantly up-regulated by 1.74-fold (Log2FC = 0.80, q < 0.05). Furthermore, the cAMP signaling pathway of P. haitanensis was also found imperfect (Additional file 5: Table S5). For example, no adenylate cyclase gene and cAMP-dependent protein kinase gene were detected, but two genes encoding adenylate cyclase-associated protein (CAP) genes, Pha000384 and Pha006544, and one gene encoding the conserved domain of catalytic domain of Ser/Thr kinases, Pha009109, were found while they were not differentially regulated by either SA or oligoagar.