Introduction

The tropical seaweed, Kappaphycus alvarezii (Doty) Doty ex P.C. Silva, is an important marine commodity due to its capacity to produce carrageenan which has economic value (Bindu and Levine 2011). In addition, it is a very important source of carotenoids, proteins, fibres, minerals, essential fatty acids, vitamins and other compounds (Nagarani and Kumaraguru 2012). Kappaphycus alvarezii, which is also known as Eucheuma cottonii, can be described as bushy red algae consisting of numerous round branches with a long, cylindrical thallus and sharp pointed tips (Chan et al. 2013; Bono et al. 2014). Carrageenan are groups of polysaccharides with a mutual structural basis of alternating 4- and 3-linked ß-d-galactopyranosyl units, and it is widely used as a replacing agent for gelatin for its thickening, gelling and stabilizing properties. There are three types of commercially available carrageenan which are kappa, iota and lambda carrageenans. The key variances between the three carrageenan types are the difference in the numbers and positions of the ester sulphate groups on the galactose units with one, two and three sulphate groups per disaccharide, respectively (Necas and Bartosikova 2013). In the presence of salt, particularly potassium, kappa carrageenan forms very strong and rigid gels (Cunha and Grenha 2016) which make it useful for many applications in the food, pharmaceutical, cosmetic and biotechnology industries as a gelling and thickening agent (Yong et al. 2014b). Kappaphycus alvarezii is considered to be the most significant source of κ-carrageenan (Chan et al. 2013).

The current seaweed farming process faces many challenges to meet the increasing industrial demand for carrageenan (Yong et al. 2015). One of the main challenges is the shortage of supply and insufficient materials for the cultivation process due to changes in growing conditions such as water temperature, seasonal changes, pests, destruction by epiphytic filamentous algae and infectious diseases (Luhan et al. 2015). This has caused disruption of seaweed growth, inconsistent yield quality of carrageenan and an increased susceptibility to diseases (Yong et al. 2014a). In order to overcome these problems and increase the productivity of seaweed raw materials, modern biotechnology techniques such as tissue culture has been used to high-quality seedling production (Reddy et al. 2003, 2008; Yong et al. 2014a). Tissue culture or micropropagation is a versatile tool for the multiplication of selected strains to produce specimens of desirable traits in a short period of time. The callus induction may be observed after 2 weeks of culture of the explants from the cut surface as filamentous outgrowths. The callus regeneration throughout the explant surface may become more noticeable and spread all over the cut surface after 2 months. The rate of formation of callus induction using tissue culture method is about 80% (Reddy et al. 2003). This technique was applied in the seaweed cultivation process to achieve good quality seedlings with a greater carrageenan content that are less susceptible to breakage and have a higher survival rate compared to the conventional vegetative cutting method practised by farmers (Yong et al. 2015). Kappaphycus alvarezii was successfully grown in laboratory conditions through tissue culture and then later grown in the sea to maturity for mass production of K. alvarezii (Yong et al. 2014a). Recently published studies showed that tissue-cultured seaweeds have better growth rates and produce higher carrageenan than seaweeds propagated through cuttings (Hayashi et al. 2008; Hurtado et al. 2014; Yong et al. 2014a). However, the technical aspects of macroalgal culture are still very limited compared to higher plants, necessitating further research.

Both the ecological and economic importance of these red algae justifies the need to expand our knowledge of carrageenan production pathways at the molecular level. There have been some reports of the use of proteomic analysis in algae, such as the proteomics studies of Chlamydomonas reinhardtii (Förster et al. 2006), Saccharina japonica (Yotsukura et al. 2010), Ectocarpus siliculosus (Ritter et al. 2010) and the red algae Gracilaria changii (Wong et al. 2006), Gracilaria lemaneiformis (Wang et al. 2016) and Pyropia orbicularis (Ramírez et al. 2014). Currently, the farming practise is to harvest the whole algae and to use a new small cutting when culturing the seaweed back to the sea. The new cuttings are used for the next cycle to ensure higher growth rates and survival percentages. In conclusion, the use of tissue culture techniques in seaweed cultivation can offer seaweed seedlings free from diseases and strains of higher quality and yields of carrageenan as well as higher survival rates, compared to the liquid-cultured K. alvarezii.

This study was performed to compare the difference in protein level between tissue culture derived and liquid culture derived seaweed (seaweed originating from cutting methods). The goal was to evaluate the biological and physiological changes at the molecular level that might play important roles in the production of high-quality planting material with higher carrageenan percentages on an annual basis under a maintained environmental condition (Yong et al. 2014a).

Materials and methods

Sample collection and acclimatization of seaweed in laboratory condition

Acclimatization took about 2 months before it is ready for tissue culture and liquid culture. For acclimatization, seaweed samples obtained from Semporna (4.4833° N, 118.6167° E), Sabah, Malaysia was brought to the laboratory in a polystyrene container. The samples were soaked in seaweed fertilizer in 1:100 ratio (fertilizer: artificial seawater (ASW) (Solis and Draeger 2010) (salinity 31 ppt)) for 2 h prior to acclimatization in controlled laboratory condition (temperature 26–28 °C, phosphate content 1.0 mg L−1, salinity 33 ppt) with water replenished weekly.

Tissue culture method

Axenic culture was done according to Thirunavukkarasu et al. (2011). Acclimatized seaweeds were washed thoroughly under running tap water and were brushed gently under dissecting microscope to remove all dirt. They were then cut into segments of approximately 3–5 mm. The segments were dipped in 70% ethanol for 10 s, preceded by rinsing with sterile distilled water for 10 s (Lawlor et al. 1989). The segments were then dried using sterile paper towel before they were cultured on 1.5% (w/v) agar solidified Provasoli’s Enriched Seawater (PES) medium (Reddy et al. 2003). Approximately 85–90% of the callus forming explants survived. This suggests that the simple sterilization process using 70% ethanol is sufficient to surface sterilize the seaweed explants for callus induction and plant regeneration.

The callus obtained was transferred into embedded culture in 0.4% (w/w) agar solidified PES medium after 2 months of culture to obtain somatic embryogenesis which is later transferred into liquid PES medium. The culture was shaking on a rotary shaker at 100 rpm at 22 ± 1 °C under cool daylight tube lamp 3.8 μmol photons m−2 s−1 with a 12:12 light/dark cycle for 1 month for the development of micropropagules and later transferred into liquid PES medium in aerated culture to further grow the seaweed into young plantlets.

Liquid culture method

Acclimatized seaweeds were washed thoroughly with tap water with soft brush under microscope. The seaweeds were then chopped into 3–5 mm length fragments and transferred into flask containing PES in filtered ASW and were shaking on a rotary shaker at 100 rpm at 22 ± 1 °C under cool daylight tube lamp 3.8 μmol photons m−2 s−1 with light/dark 12:12 cycle for 2 weeks. After that, the seaweeds were transferred into liquid PES medium in aerated culture to further grow the seaweeds prior to transfer to the field.

Seaweed field cultivation

After obtaining weight ≥10 g, the tissue-cultured and liquid-cultured seaweed were relocated to a seaweed farm located in Tanjung Rhu, Langkawi Island (6.3500° N, 99.8000° E), Kedah, Malaysia. Seaweed materials were acclimatized using seawater for 1 week prior to transfer into the open sea. The tissue-cultured and liquid-cultured plantlets were bound in different ropes but in the same contained area in the sea. The seaweed were collected after 60 days and brought back to the research laboratory for analysis.

Carrageenan extraction

Semi-carrageenan was extracted from both tissue-cultured and liquid-cultured plantlets according to Yong et al. (2014a). Seaweed materials were rinsed thoroughly under the tap water and left to dry in oven at 60 °C overnight. About 5–6 g of dried seaweed was ground, and the sample weight (SW) was weighed. Ground samples were placed in pre-heated potassium hydroxide (6% KOH) solution for 30 min. The heated seaweed materials were then rinsed with distilled water several times to remove extra KOH before drying in a 60 °C oven overnight to obtain semi-refined carrageenan weight (CW). Carrageenan yield was calculated following the formula: carrageenan yield (%) = (CW/SW) × 100.

Protein extraction

Protein extraction was carried out using phenol extraction method following Al-Obaidi et al. (2014) with some modification (Al-Obaidi et al. 2016). Approximately 5 g of seaweed materials was grounded using sterilized mortar and pestle in liquid nitrogen. The sample then transferred to an extraction buffer with 1 M Tris–hydrochloride pH 8.3, 5 mM EDTA, 1% DTT and 0.9% sucrose. After short vortex, the sample was centrifuged at 5000×g for 10 min. An equal volume of ice cold phenol was added to the mixture. Samples were vortexed for 10 min at 4 °C and centrifuged for 3 min, 8000 rpm at 4 °C. The upper layer was collected and re-extracted. After phenol extraction, proteins were precipitated overnight with ammonium acetate in methanol. Pelleted proteins were then re-suspended in lysis buffer (Al-Obaidi et al. 2014), and proteins were kept in −80 °C prior to quantitation. Protein quantitation was performed using Bradford assay kit (BioRad, USA) with BSA as a standard.

Isoelectric focusing (IEF) and SDS-PAGE

For the first dimension, IEF was carried out using PROTEAN i12 (BioRad, USA) by loading 400 μg using 13 cm immobilized pH gradient (IPG) strips (pH 4–7, GE Healthcare, UK). Passive rehydration was performed for 12 h at 20 °C. Once the passive rehydration was completed, the strips were ready for IEF run. IEF was performed at 500 V for 1 h followed by 1000 V constant for 1 h, 11,300 V for 2.30 h and 3400–7400 V for 55 min which made a total run of 25,000 V for 5 h and 25 min. The IPG strips were incubated for 15 min with shaking in freshly prepared equilibration buffer-1 (6 M urea, 50 mM Tris–HCl pH 8.8, 30% glycerol, 2% SDS and 100 mg mL−1 DTT). Second equilibration step was performed using equilibration buffer-2 (6 M urea, 50 mM Tris–HCl, pH 8.8, 30% glycerol, 2% SDS and 250 mg L−1 IAA) for 15 min. After rinsing in SDS gel running buffer, the IPG strips positioned on the top of the 12% SDS gels. The SDS gels were run at 20 °C until the dye front ran out of the gels; protein spots of gels were stained with CBB (G-250) staining (Neuhoff et al. 1988).

Image analysis

The stained gels from two experiments with four biological replicates each were scanned using GS-800 Calibrated Imaging Densitometer (BioRad, USA). The scanned images were analysed using Progenesis (Nonlinear Dynamics Ltd., UK) same spot analysis software. The protein expression profiles of tissue culture seaweed were used as a reference.

Protein identification

Tryptic digestion

Spots of interest were excised from the gel followed by washing steps. After that, the spots were destained three times with a solution containing 100 mM NH4HCO3 and 50% acetonitrile (ACN) for 20 min each at 25 °C, followed by 100% ACN. After that, trypsin (12.5 mg mL−1 in freshly prepared 50 mM ammonium bicarbonate buffer) was added onto the gel spots and incubated overnight at 30 °C. Peptides were collected using 50% ACN for 15 min and 100% ACN for another 15 min. Finally, peptides were re-suspended with 0.1% trifluoroacetic acid (TFA). Peptides were desalted using Zip-tip C18 (Milipore) and spotted on the AB SCIEX MALDI plate (Opti-TOFTM 384 well insert).

MALDI TOF/TOF mass spectrometry analysis and database search

Mass spectrometry was performed using matrix-assisted laser desorption/ionization time of flight tandem spectrometry (MALDI TOF/TOF) (ABI 4800 Plus). Proteins were identified as a result of accumulation of MS/MS spectra by manual searching using sequence databases implemented in the MASCOT search engine against red seaweed database (19,858 sequences; 6,932,644 residues). MASCOT search were achieved using the following criteria: carbamidomethyl-fixed modifications and mass values were set as monoisotopic with ±100 ppm of peptide mass tolerance and ±0.2 Da fragment mass tolerance with 1 maximum missed cleavages. Significance was considered for candidates with protein score of CI % or ion CI % of more than 95.

Functional analyses using IPA software

Functional analyses of the proteomics data were further performed using the Ingenuity Pathways Analysis (IPA) software to predict protein networks related to the alteration of proteins abundance in response to the different tissue culture methods. Details of the significantly altered proteins, their quantitative expression values (fold change difference) and p values were introduced into the IPA software. Identifiers were mapped against its corroborating protein object and were overlaid onto a universal molecular network developed from information provided in the Ingenuity Knowledge Base. Network of proteins algorithm was generated based on their connectivity. Right-tailed Fischer’s exact test was used to calculate a p value indicating the probability that each biological function assigned to the network and canonical pathway is due to chance alone.

Verification of MALDI TOF mass spectrometry data by using ELISA analysis

Proteins, namely β-amylase, NAD-dependent sugar epimerase and B-phycoerythrin which were aberrantly expressed in the tissue-cultured K. alvarezii were selected for ELISA analyses. This assay employs the quantitative sandwich enzyme immunoassay technique. Briefly, proteins were extracted as mentioned in the previous method using phenol extraction method following (Al-Obaidi et al. 2014) with some modification (Al-Obaidi et al. 2016). The abundance of protein was detected against β-amylase (Agrisera, AS09 380), NAD-dependent sugar epimerase (Abcam, ab155226) and B-phycoerythrin (Agrisera, AS08 279). Briefly, 100 μL of standards or protein samples was added into the ELISA microplate pre-coated with antibodies specific for the selected proteins. The selected proteins (5 ng mL−1) were probed with an anti β-amylase, NAD-dependent sugar epimerase and B-phycoerythrin antibodies with 1:2000 dilution ratio. The reaction mixtures were incubated for 2 h at 37 °C. After removing any unbound substances, 100 μL of biotin-conjugated antibody specific for β-amylase, NAD-dependent sugar epimerase and B-phycoerythrin was added to the wells and the reactants were incubated for 1 h at 37 °C. The plates were washed before adding 100 μL of avidin-conjugated horseradish peroxidase (HRP) followed by 1 h incubation at 37 °C. Following a wash step to remove any unbound avidin-enzyme reagent, 90 μL of tetramethylbenzidine (TMB) substrate solution was added to the wells and colour develops in proportion to the amount of the selected proteins bound in the initial step. The colour development was stopped by adding 50 μL of stop solution, and the intensity of the colour was measured at 450 nm. A standard curve was created to determine the concentration of protein abundances present in the samples.

Results

Carrageenan production between liquid-cultured and tissue-cultured K. alvarezii

After 2 months of growing in the open sea, significant changes in the carrageenan concentration were observed between liquid-cultured and tissue-cultured K. alvarezii. Higher concentration of carrageenan in the tissue-cultured K. alvarezii (69.9 ± 4.8%) was obtained in comparison to liquid-cultured seedlings (59.2 ± 2.7%). Results obtained are the mean with standard deviation of four replicates. Student’s t test was used to compare carrageenan yield and quality of tissue-cultured and liquid-cultured K. alvarezii. Differences were considered significant at p ≤ 0.05.

Analyses of the proteins induced in tissue-cultured K. alvarezii

To identify proteins specifically showing a different expression profile in tissue-cultured versus liquid-cultured K. alvarezii, 2-DE profiles were compared. By using Coomassie blue staining method, 785 spots were obtained from tissue-cultured K. alvarezii compared to 741 spots obtained from the liquid-cultured sample. Image analysis of the gels produced from the two conditions revealed 54 spots that showed significant differences with a fold change >1.5 and p < 0.05 (ANOVA). Among these protein spots, 45 spots were successfully identified (Fig. 1). Out of these, 13 were downregulated and 32 were upregulated in tissue-cultured K. alvarezii compared to the liquid-cultured samples (Supplementary Table 1).

Fig. 1
figure 1

Representative 2-DE gel image of seaweed total protein. Numbered proteins were identified by MALDI-TOF/TOF and shown in Supplementary Table 1

Full size image

MALDI-TOF/TOF identification

The corresponding protein spots with significant changes in the expression profiles were digested by trypsin and afterward analysed by mass spectrometry using the MALDI TOF/TOF (ABI 4800 Plus) mass spectrometer system. Out of the 54 spots that showed a significant change in the expression profile between the two conditions, 45 spots were successfully identified by MS/MS analysis representing 37 proteins with known functions.

Functional classification

Functional classification and subcellular location were predicted using Gene Ontology (GO) information obtained from the universal protein database (http://www.uniprot.org/). The 45 proteins identified were classified according to their molecular function (Fig. 2a). The group with the highest protein number was related to ATP binding, representing 25% of the differentially expressed proteins. According to their subcellular location, the proteins were grouped, 55% of the proteins were of non-predicted cellular locations, and 29% were of chloroplast located proteins (Fig. 2b).

Fig. 2
figure 2

Pie charts of a molecular function and b subcellular location of proteins identified and differentially expressed in Kappaphycus alvarezii

Full size image

Network interactions and top functional analyses of the significantly expressed proteins

The IPA identified “energy production, carbohydrate metabolism” as the top associated linkage 28 of the differentially regulated proteins with different cellular and molecular functions, with a score of 21 (Table 1, Fig. 3). A minimum score of 2 indicates at least a 99% of confidence that a network is not being generated by random chance, greater confidence indicated by higher scores. “cellular energy production, carbohydrate metabolism” came second, and scoring “small molecule biochemistry, molecular transport” was ranked third with a score of 2.

Table 1 A summary of Ingenuity Pathway Analysis (IPA) showing the top associated networks and top biological functions affected by the abundance of proteins expressed in tissue-cultured Kappaphycus alvarezii
Full size table
Fig. 3
figure 3

Predicted canonical pathway of altered abundance proteins of tissue-cultured Kappaphycus alvarezii. Lines between proteins represent known interactions. Nodes in red indicate upregulated proteins while those in green represent downregulated proteins. Various shapes of nodes represent functional class of proteins. Different arrow shapes represent different types of interactions. Edges are displayed with various labels that describe nature of relationship between the nodes. Names of genes corresponding to the abbreviations are as follows: ACT actin, FtsH cell division protein FtsH, CYP cytochrome P450, DYNC1I1 dynein intermediate chain, GAPDH glyceraldehyde-3-phosphate dehydrogenase, HSP heat shock protein, PRDX peroxiredoxin, v-SNARE putative vesicle transport v-SNARE, RPL16 ribosomal protein L16, TERA transitional endoplasmic reticulum ATPase, EIF4A translation initiation factor eIF4 subunit A

Full size image

Figure 3 represents the predicted molecular linkage between the proteins listed under the top associated networks, which connected them to the top four predicted biological functions analyses that are listed in Table 1. The IPA identified “energy production” (p < 4.51 × 10−6) and “carbohydrate metabolism” (p < 1.17 × 10−5) as the two top functions that linked to the altered expression of proteins, namely ATP-dependent Clp protease, ATP synthase, B-phycoerythrin alpha chain, C-phycocyanin beta chain, cytochrome P450, ferredoxin–NADP+ oxido-reductase, FtsH, heat shock protein, lysidine synthase, peptidyl-prolyl cis–trans isomerase, peroxiredoxin, photosystem I, pyruvate dehydrogenase, pyruvate kinase, R-phycoerythrin alpha chain, R-phycoerythrin beta chain, ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO), transitional endoplasmic reticulum ATPase, transketolase, vacuolar-type H+-ATPase, v-SNARE, actin, β-amylase, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), glucose-6-phosphate isomerase, NAD-dependent sugar epimerase, tetratricopeptide repeat (TPR) domain containing protein and WD40-repeat containing protein. “Cellular function and maintenance” (p < 5.57 × 10−4) was ranked third which linked nine proteins, namely dynein intermediate chain, EIF4A, multiprotein-bridging factor 1, peroxiredoxin, protein kinase, ribonuclease, ribosomal protein L16, thiazole synthase and vacuolar-type H+-ATPase. “Small molecule biochemistry” (p < 3.51 × 10−3) was listed as the top fourth function linking four proteins including GCN5-like N-acetyltransferase, heat shock protein, transketolase and tRNA dihydrouridine synthase. The top associated networks suggest that tissue-cultured K. alvarezii affects the top functions by changing the expression of those proteins, converging mainly on photosynthesis, polysaccharide synthesis, plant growth and stress tolerance.

Detection of altered protein abundance in tissue-cultured and liquid-cultured K. alvarezii using ELISA

Figure 4 shows significantly higher levels of β-amylase, NAD-dependent sugar epimerase and B-phycoerythrin in the tissue-cultured K. alvarezii compared to the liquid-cultured samples. All proteins were selected as they showed significant alteration of protein abundances in 2-DE profiles, and they were listed in the top canonical pathways using the IPA software.

Fig. 4
figure 4

Enzyme-linked immunosorbent assay (ELISA) analyses of the β-amylase, NAD-dependent sugar epimerase and B-phycoerythrin antibodies level in the tissue-cultured and liquid-cultured Kappaphycus alvarezii. ELISA analyses were done according to manufacturer’s protocols. Bars not sharing the same superscript letter indicate significant difference at p < 0.05

Full size image

Discussion

Production of carrageenan in the red seaweed, K. alvarezii, is a complicated process, and understanding this process at the molecular level has been a main goal. The proteomics approach can be applied as a molecular tool to detect protein expression, and it is considered appropriate platform for exploring the molecular mechanisms of organism adaption to different environmental circumstances (Sun et al. 2012). Comparing the proteome of K. alvarezii from different cultivation methods may help to obtain more comprehensive understanding of the carrageenan biosynthesis process of seaweeds. This study considered the first report of the K. alvarezii proteome, focusing on the carrageenan biosynthesis in different cultivation methods.

When the differentially regulated proteins were subjected to IPA analyses, energy production and carbohydrate metabolism were listed as the top biological functions. Studies reported that factors to increase the production of high-quality carrageenan in seaweeds include increasing the rate of photosynthesis, refining polysaccharide synthesis and manipulating the nutrient supply (Shalaby 2011; Wang et al. 2014).

As reported before, increasing rate of photosynthesis directly affected the production of carrageenan (Shalaby 2011). In IPA analyses, the altered abundance of heat shock protein (HSP), photosystem I (PSI) and RuBisCO affected the photosynthesis process in K. alvarezii. The process of photosynthesis consists of two phases. The first phase related to the conversion of light into chemical energy (as ATP). The ATP produced was then used in the second phase of the reactions for carbon fixation (Wu et al. 2013). Photosystems comprise a group protein complex where light absorption processes take place (Morosinotto and Bassi 2012). Oxygenic photosynthesis uses two photosystems subunit (I and II) positioned in the thylakoid membranes of chloroplasts in algae (Hasan et al. 2011). Both photosystems are highly significant for light harvesting and photosynthetic process activity regulation (Caffarri et al. 2014). In this study, expression of PSI subunit II is upregulated. Other than PSI, HSP was also shown to be upregulated in tissue-cultured samples in comparison to the liquid-cultured one. Heat shock protein was reported to protect the electron transport and oxygen evolution of the photosystem, and its upregulation in tissue-cultured samples was significantly positively correlated with photosynthetic thermo tolerance and secures functioning of photosynthesis under heat stress in the brown alga Fucus serratus (Jueterbock et al. 2014). The expression profile of PSI indicates increasing in the photosynthesis rate in tissue-cultured K. alvarezii which might have indirect involvement in carrageenan synthesis (Shalaby 2011).

The peptidyl-prolyl cis–trans isomerase (PEP) is an abundant protein family in the thylakoid lumen (Galat 2003). Several members of the cyclophilin subfamilies catalyse cis–trans isomerisation of the peptidyl-prolyl bond. A previous study reported that PEP in the red seaweed Chondrus crispus was upregulated during abiotic stress which may suggest a role for this enzyme during the synthesis of carrageenan (Collén et al. 2013). Phycobilisomes (PBSs) are known to be major photosynthetic units that provide K. alvarezii and other red algae with the capability of light-harvesting and energy migration (Apt et al. 1995). In cyanobacteria, the pigments related to light-harvesting are chlorophyll-a, carotenoids and phycobiliproteins. Phycobiliproteins are usually divided into three main groups according to their structure: phycocyanins (purplish blue), allophycocyanin (blue) and phycoerythrins (red) (Ting et al. 2002). Phycobiliproteins represent the most abundant soluble proteins and the major light-harvesting complex for photosynthesis (Guan et al. 2013). Therefore, the upregulation of phycocyanin and phycoerythrin expressions in this study may further increase photosynthetic rate, thus increasing the production of the polysaccharide carrageenan.

Other than the upregulated proteins, two proteins involved in light absorption of photoautotrophs, namely FtsH and ATP-dependent Clp protease, were downregulated in this study. The ATP-dependant Clp protease and FtsH likely carry out most proteolysis in the plastid; Clp protease has been found in the stroma while FtsH was found in the thylakoid membrane (Olinares et al. 2011). Previously, upregulation of the FtsH and Clp protease was reported to have a role in photosystem proteins degradation during leaf ageing and abiotic stress (Majeran et al. 2010). Proteases including FtsH subunits play significant role as a molecular chaperone in chloroplasts (Chi et al. 2012). The efficient harvesting of solar energy is the first important step in photosynthesis process. However, extra solar energy can be destructive when the amount of absorbed light exceeds the limits. Photosynthetic organisms have evolved a number of photoregulatory processes including photoinhibitory quenching. Hence, the downregulation of FtsH and Clp proteases could be linked to re-arrangement of the photosynthetic system in response to different light conditions that balance excitation between photosystem II (PSII) and photosystem I (PSI) in order to optimize photosynthetic electron transport (Roach and Krieger-Liszkay 2014). Other than regulation of light absorption, chloroplast maintenance also influences the rate of photosynthesis. Proteins peroxiredoxin (Prx) and v-SNARE, which were involved in chloroplast antioxidant defence and transport, were upregulated in this study. Peroxiredoxin plays a significant role in antioxidative defence during abiotic stress conditions. They catalyse the detoxification of peroxides, thus reducing the damaging effects caused by ROS to the plant (Tripathi et al. 2009). In addition to detoxifying peroxides, Prx is also known to play a role in scavenging for reactive nitrogen species (RNS) especially peroxynitrite (Sakamoto et al. 2003). Previously, the downregulation of Prx in Arabidopsis thaliana leads to a reduction in chlorophyll content and the photosynthetic process (Baier and Dietz 1999). In addition, the decrease in chlorophyll-fluorescence parameters in Prx-knockout A. thaliana suggests a probable role in defending photosynthesis (Lamkemeyer et al. 2006) which may suggest a role of Prx in chloroplast antioxidant defence during photosynthesis in seaweed. Similar to this study, upregulation of v-SNARE expression was reported to be involved in photosynthesis, including preservation of the photosynthetic machinery, as this support system is dependent on such vesicle transport (Zheng et al. 1999; Khan et al. 2013).

However, RuBisCO was downregulated in this study. It is a bifunctional enzyme known to catalyse carbon dioxide fixation and oxygenation (Wong et al. 2006) in order to create a competitive metabolic pathway between photosynthesis and photorespiration (Chen et al. 1988; Raines 2011). The competition between oxygen and carbon dioxide takes place at the active site of the RuBisCO large subunits, and the enzyme is readily interconverted between the activated and inactivated forms. In Rhodophyta and Phaeophyta, both large and small RuBisCO subunits were reported to be encoded in the chloroplast genome (Wang et al. 2011). The upregulation of the RuBisCO large chain affected the growth rate of K. alvarezii (Tee et al. 2015). Hence, downregulation of RuBisCO would likely protect the chloroplast from degradation, thus improving the rate of photosynthesis. Other than PSI, HSP and RuBisCO which regulated the photosynthesis process, the other significant proteins are also involved indirectly to increase the photosynthesis rate in K. alvarezii. IPA analyses showed that the altered abundance of PBSs including phycocyanins and phycoerythrins was shown to have regulated the light response which elevated the photosynthetic rate in K. alvarezii (Fig. 3). Another protein that has shown downregulation in this study was ferredoxin–NADP+ oxido-reductase (FNR). It is an enzyme, which has been identified in various organisms including plants and algae (Goss and Hanke 2014). In addition, FNR was reported in plants suffering from different abiotic stress (Mulo 2011). The FNR reported previously to be produced from the thylakoids in the plants affected by drought stress (Lehtimäki et al. 2010). The downregulation of FNR in this study showed that the cultured K. alvarezii in our study was not exposed to any stress condition, thus inhibiting the regulation of FNR to respond to oxidative stress signalling. In addition to its role in photosynthesis, this enzyme reported previously to be involved in the production of NADPH which involved in the synthesis of actively natural products with potential pharmaceuticals uses (Nielsen et al. 2013). Cytochrome P450 monooxygenases (P450s) are enzymes reported to have a role in the production of bioactive natural products (Hamdane et al. 2008). An important role of P450s includes the synthesis of essential compounds such as fatty acids, steroid, vital nutrients, antibiotics and plant-related defence compounds (Denisov et al. 2005). Cytochrome P450s reported previously to be involved in the biosynthesis of many plant terpenoids (Bohlmann and Keeling 2008). This might also apply to biosynthesis of carrageenan, as the upregulation of cytochrome P450 would induce carrageenan production.

In addition to those proteins, the vacuolar type H+-ATPase (V-ATPases) in plants is a large multimeric enzyme complex whose function in proton transportation (Elston et al. 1998). The enzyme previously reported for being responsible for the de-acidification of the cytosol (Magnotta and Gogarten 2002). The V-ATPases play a key role in the maintenance of vacuolar homeostasis in plant cells. They are also involved in plant defences against environmental stress (Perera et al. 1995). The upregulation V-ATPase in our study might provide those aforementioned protections to the cultured K. alvarezii.

Transketolase (TK) is a universal amphibolic enzyme catalysing reactions in the Calvin cycle and the oxidative pentose pathway producing erythrose 4–phosphate, leading to phenylpropanoid metabolism. Thus, any change in its abundance is expected to affect both photosynthetic carbon assimilation and secondary metabolism in plants (Henkes et al. 2001). A recent study showed that early glycolytic step is unessential in the rice cells–M. oryzae interaction, whereas TK is essential for this process (Wilson and Talbot 2009; Fernandez et al. 2014). Hence, the downregulation of TK in our study might reflect a similar function in K. alvarezii, thus conferring further protection.

Carbohydrate metabolism was listed as in the top network as well as the second top biological function by IPA software (Table 1). Actin, β-amylase, GAPDH, glucose-6-phosphate isomerase, NAD-dependent sugar epimerase, pyruvate dehydrogenase, TPR domain containing protein and WD40-repeat containing protein were upregulated, while pyruvate kinase was downregulated. Carrageenan is a polysaccharide made in the red algae cell wall (McKim 2014). In plants, many bioactive products are generated from components of the cell wall (Pauly and Keegstra 2010). Manipulation of plant cell wall biosynthesis has the possible role to improve the production of biofuel compounds (Furtado et al. 2014). Actin is an important protein with an important role in cell division and other different cellular processes (An et al. 1999). Previously, it was described that a number of algal cells, including red algae, encode different actin genes (Wu et al. 2009). Cytoskeletal depolymerisation studies revealed the role of actin in the movement of Golgi stacks (Nebenführ et al. 1999). The upregulation of actin protein in tissue-cultured sample may indicate the role of cell wall component in the production of bioactive compound such as carrageenan (Crowell et al. 2009; McFarlane et al. 2014).

The NAD–nucleotide–sugar epimerase are activated monosaccharaides which directly used by glycosyltransferases for the biosynthesis of various polysaccharides (Yin et al. 2011). All of the NAD-dependent sugar epimerase families are generally active in plant and algal cells having carbohydrate-rich cell walls, which are involved in the synthesis of cell wall polysaccharides, which are considered the major components of plant biomass (Reiter 2008). The UDP-glucose 4-epimerase (UGE) isoforms that are tightly bound to NAD+ could have a role in the biosynthesis of stress-related carbohydrates or during catabolic cell wall turnover. The increased levels of epimerase either could be used for glycolysis reaction or involve the production of cellulose and starch production; this could explain the upregulation of the NAD epimerase and UGE in the tissue-cultured sample and the possible role of these two enzymes as precursor the carrageenan synthesis (Goulard et al. 2001; Pan et al. 2010). A recent study showed that β-amylase plays a main role in starch degradation (Nagler et al. 2015). Downregulation of β-amylase in potato leaves using antisense methods resulted in a starch-excess phenotype compared to wild-type plants (Scheidig et al. 2002). Our results showed upregulation of β-amylase which might be an important factor to help the seaweed cope with the stress during the tissue culture growth (Kaplan and Guy 2004).

Another cell wall-related polysaccharide enzyme, WD40-repeat (WDR) proteins often function in the formation of lignocellulosic biomass of cell wall found to be upregulated in tissue-cultured sample. The enzyme may play important role in the polysaccharides synthesis process that requires the assembly of membrane enzyme complexes, vesicle trafficking and the cytoskeleton interactions (Zhong et al. 2004; Allan and Ratajczak 2011; Guerriero et al. 2016; Sparks et al. 2016). Knockout of WDR protein in Arabidopsis causes phenotypes related to a reduced secondary cell wall thickness in fibres and xylem vessels, with consequent decrease in stem strength (Zhong et al. 2004). Other than those aforementioned proteins, there are other significantly regulated proteins that are associated with the maintenance and protection of K. alvarezii, thus might indirectly affect the carrageenan production. The third top biological functions in IPA analyses showed the altered abundance of dynein intermediate chain, EIF4A, multiprotein-bridging factor 1, peroxiredoxin, protein kinase, ribonuclease, ribosomal protein L16, thiazole synthase and vacuolar-type H+-ATPase which affected the cellular function and maintenance in K. alvarezii. Other than dynein, more protein showed high level of expression in tissue-cultured seaweed such as proteins called translation initiation factors (eIFs) that were also involved in cellular maintenance (Muñoz and Castellano 2012). The phosphorylation of several eIFs is involved in triggering translation process. A previous study showed the involvement of different eIFs phosphorylation in controlling translation initiation (Boex-Fontvieille et al. 2013; Dutt et al. 2015). In addition to eIF4A, ribosomal protein L16 (RPL16) was also involved in regulation of abiotic stress tolerance in plants. Similar to our findings, RPL16 was upregulated in brown algae, Sargassum fusiforme (Liu et al. 2016) and Fucus vesiculosus (Pearson et al. 2010) that undergo desiccation stress and rehydration to increase the abiotic stress tolerance of the plants (Liu et al. 2016). Our data showed an upregulation pattern in dynein, eIF4A and RPL16 expressions, which implies the potential of these proteins in regulating cellular function and maintenance in our tissue-cultured K. alvarezii. In this study, it was evident that the alteration of protein abundance in tissue-cultured, compared to liquid-cultured, K. alvarezii showed in the 2-DE profile was corroborated by changes in abundance of the corresponding proteins analysed using ELISA. Three proteins, namely β-amylase, NAD-dependent sugar epimerase and B-phycoerythrin, showed a consistent pattern of upregulation in ELISA analyses hence validating the 2-DE profiles.

In conclusion, in this work, the changes in the protein expression of K. alvarezii cultivated using tissue culture compared to liquid culture using a comparative gel-based proteomic approach were investigated. The changes in the protein profiling led to the identification of different proteins that might be involved directly or indirectly in carrageenan synthesis during two different propagation conditions. Ingenuity Pathway Analyses were applied to identify functions of proteins related to the carrageenan biosynthesis pathway of seaweed including energy production and carbohydrate metabolism. These functions play a vital role in the production of carrageenan in seaweed. Some proteins were upregulated, such as actin, β-amylase, NAD-dependant sugar epimerase and TPR domain containing protein, and might be directly involved in the production of carrageenan in K. alvarezii cell walls. In addition, this study provided some basic information on the molecular mechanisms behind the synthesis of carrageenan and the effect of propagation methods of red algae, especially K. alvarezii, on carrageenan synthesis. More molecular studies on the transcriptional level will be helpful in further understanding how changes in RNA level lead to the production of higher quality and quantity of carrageenan using seaweed micropropagation.