**3. Discussion**

In this study, we investigated the distribution of AstaP orthologs from Scenedesmaceae strains. Although AstaP proteins have not been identified except strains that were isolated under high light conditions in our previous study, new AstaP orthologs from Scenedesmaceae strains were identified in the strains from culture collections. All the AstaP orthologs commonly conserved H1 and H2 fasciclin domains and N-terminal hydrophobic signal peptide for secretion; however, the number of glycosylation sites, pI, and molecular weights were found to vary depending on the strain (Supplementary Figure S1). *Cmy. reinhardtii* and *Crl. variabilis* were shown to possess photooxidative stress inducible AstaP orthologs; however, we could not detect the expression of water-soluble carotenoproteins in both strains. Further analysis will be needed to make the function of these gene products clear including their possibilities for carotenoid binding.

Scenedesmaceae strains from the culture collection were isolated under non-stressful conditions, such as in lake (*S. costatus*) [15], bark (*C. vacuolata*) [16], and peat bog (*C. striolata*) [13,17,18]. All the test strains, except *S. obliquus*, were found to express astaxanthin-binding water-soluble protein under the experimental conditions in this study; however, the amounts of production were estimated to be low compared to those of the Ki-4 and Oki-4N. We conclude that AstaP is widely distributed in Scenedesmaceae, and the expression levels of AstaP may correlate with the ability of photooxidative stress tolerance of each strain.

In conclusion, although the functional differences among these AstaP orthologs remain unclear, the presence or absence of glycosylation, the difference in pI, and the presence or absence of signal peptide for secretion may reflect the subcellular localization of AstaPs. In our previous study, Ki-4 AstaP, which is classified as a glycosylated basic pI group, was predicted to be localized at the cell surface [9,19]. On the other hand, Oki-4N AstaP-pinks, which were classified as a non-glycosylated acidic pI group, were suggested to be localized at the endoplasmic reticulum or vacuole but not at the cell surface based on the fluorescent microscopic analysis [10]. Further characterization of AstaP orthologs will clarify the distribution, localization, and specific functions under the photooxidative stress conditions.

#### **4. Materials and Methods**

#### *4.1. Microalgal Strains and Growth Conditions*

Scenedesmaceae strains were obtained from culture collection center as previously described [13]. Strain Ki-4 and Oki-4N were isolated in our laboratory as previously described [9,10]. *Scenedesmus* sp. Oki-4N was named to *Scenedesmus obtusus* Oki-4N based on a high similarity of homology score of 18S rDNA and ITS2 sequence with the morphological identity with *Scenedesmus obtusus* as previously described [10]. The strain name *Coelastrella vacuolata* (synonym: *Chlorella fusca* var *vacuolata* = *Scenedesmus vacuolatus = Graesiella vacuolata*) [13,16,18,20] was used in this study. The composition of the modified A3 medium was the same as previously described [9,10]. Algal strains were cultivated in A3 medium with a 16h/8h light/dark regime at 26 ◦C under low light conditions (60 μmol photons m<sup>−</sup><sup>2</sup> s<sup>−</sup>1) as previously described [18]. Photooxidative stress was induced by the addition of sterilized NaCl when the growth reached an OD750 value of 1.0 cm<sup>−</sup><sup>1</sup> under conditions of high light exposure (~800 μmol photons m<sup>−</sup><sup>2</sup> s<sup>−</sup>1) as previously described [9,10]. Cells did not change color to orange under high light conditions (800 μmol photons m<sup>−</sup><sup>2</sup> s<sup>−</sup>1) without salt, or low light conditions with salt.

#### *4.2. Gel Filtration Column Chromatography*

The water-soluble astaxanthin binding protein fractions were obtained by gel filtration column chromatography as previously described. Briefly, stressed cells were harvested, and 1.0 g of stressed wet cells were suspended in 9.0 mL of 50 mM Tris-HCl buffer at pH 7.5. Cells were broken by a multi-beads shocker (Yasui Kikai, Osaka, Japan), dissolved in Tris-buffer pH 7.5. Cell extracts (CFEs) were ultracentrifuged at 100,000× *g* for 2 h to remove cell debris and lipids. The CFEs were passed through a Sephacryl S-200 HR gel filtration column at a flow rate of 1.0 mL/min (1.6 × 60 cm, GE Healthcare, Chicago, IL, USA). The elution profiles were monitored using a photodiode array detector LaChrome Elite software (Hitachi Ltd., Tokyo, Japan), and fractions of orange eluates were collected. Proteins were concentrated by using Amicon Ultra (Merck, Darmstadt, Germany).

#### *4.3. Pigment Extraction and Identification*

The binding pigments of water-soluble carotenoprotein were extracted using the Bligh–Dyer method as previously described [9,10,19,21]. Pigments were extracted with methanol: chloroform: H2O = 12:5:3 by gentle mixing in a tube. After the addition of chloroform and water (2:3), the organic phase was obtained and evaporated to dryness under N2 gas. The extracted pigments were completely dissolved in acetone and analyzed by C18-HPLC (CAPCEL PAK C18 reversed-phase column, 150 × 4.6 mm, flow rate 1.0 mL/min). The pigment was identified based on the absorption spectra obtained using an HPLC photodiode array detector, HPLC retention times, and molecular masses from high-resolution LC/MS analysis in comparison with those of standard compounds.

#### *4.4. Purification of the Water Soluble Astaxanthin Binding Protein*

The water-soluble astaxanthin binding protein of *S. costatus* was purified by isoelectric focusing. Isoelectric focusing was performed horizontally with Maltiphor II (GE Healthcare, Chicago, IL, USA). Migrated orange band was excised and electrophoresed by SDS-PAGE. As standard markers, Precision Plus ProteinTM standard kit (Bio-Rad, Berkeley, CA, USA) was used.

#### *4.5. Determination of Peptide Sequence*

The N-terminal amino acid sequence was determined by the Edman degradation method using a PPSQ30 peptide sequencer (Shimadzu, Tokyo, Japan) as previously described [9]. Target protein band was obtained using a transferred protein on a PVDF membrane stained with Coomassie Brilliant Blue and digested with trypsin (Promega, Madison, WI, USA).

#### *4.6. Isolation of RNA and Construction of a cDNA Library*

Isolation of RNA and construction of a cDNA library were performed as previously described [9,10]. To prepare a cDNA library, *S. costatus* and *C. vacuolata* cells that were subjected to NaCl stress under high light stress and total RNA were extracted with Trizol reagen<sup>t</sup> (Invitrogen, Waltham, MA, USA). Poly(A)+ mRNA was isolated from total RNA and used to generate a full-length cDNA library with a Smart-Infusion PCR cloning system (Clontech, Palo Alto, CA, USA). The cDNA libraries were sequenced using the Illumina HiSeq 2500 system (Illumina, San Diego, CA, USA) and assembled de novo sequence by CLC Genomics Workbench (Qiagen, Hilden, Germany).
