Comparison of RNA expression profiles on generations of Porphyra yezoensis (Rhodophyta), based on suppression subtractive hybridization (SSH)

Porphyra yezoensis is one of the most important edible seaweeds commercially cultivated mainly in China, Japan and Korea. It has a dimorphic life cycle which consists of macroscopical haploid foliose blade phase and microscopic diploid filamentous conchocelis phase [1]. Conspicuous differences exist in the two generations, such as morphology, cell structure, biochemistry, physiology, and so on. Seed-stock material for Porphyra cultivation consists of carpospores (also called zygotespores) [2] spread into oyster shells in the spring, conchocelis germinated and grow in the shells at the onset of conchospores release in the autumn. The conchospores settle on net twine, cultivate in the sea water and develop into blade thalli mainly in winter and early spring. In a mature thallus of Porphyra yezoensis, four kinds of cells are developed. From the basal to the top, there are rhizoid cells, vegetative cells, male gametes and female gamete [3].

The developmental process of Porphyra yezoensis has been studied thoroughly, but the mechanism is still ambiguous. Very little of the genetic expression has been noticed. What is the process and essential happened in the period from vegetative cells to format sexual cells? What is the difference between the male gametes and female gametes? What is the difference between the foliose blade phase and conchocelis phase?

Nikaido et al (2000) first constructed the normalized and size-selected cDNA libraries of leafy gametophyte Porphyra yezoensis and generated a total of 10,154 expressed sequence tags (EST) [4]. Only 934 EST groups were classified based on the annotations of their homologous protein entries in the public databases. The reason for this feature could be that fewer sequences from red and green algae have been registered in the public DNA databases compared to those from higher plants. Asamizu et al (2003) compared the expressed sequence tags (EST) differences between the gametophytic and sporophytic generations of Porphyra yezoensis Ueda [5]. Among the 10,625 EST groups, only 1013 (22.5%) groups were classified as ESTs that commonly occurred in both generations, whereas a large proportion of EST groups were identified as being unique to either the gametophyte (43.1%) or the sporophyte (34.3%). A statistical significance test revealed 89 and 112 highly expressed gene candidates in the gametophyte and the sporophyte respectively.

Kakinuma et al (2006) have constructed subtracted cDNA libraries enriched for differentially expressed transcripts in vegetative and reproductive thalli of Porphyra, and randomly selected 1,152 cDNAs from each subtracted library [6]. BLAST analysis showed that the cDNAs represented 63 and 59 unique clones for the vegetative and reproductive subtracted libraries respectively. Interestingly, some of the cDNAs isolated from the reproductive subtracted library were homologous to genes encoding protein kinases, GTP-binding protein, and heat shock proteins involved in signal transduction and the molecular chaperon system.

The SSH method is designed to selectively amplify differentially expressed transcripts while suppressing the amplification of abundant transcripts, thus eliminating the need to separate single- and double-stranded molecules. In addition, SSH normalizes target transcripts to approximately equal abundance [7, 8].

To identify gene candidates related to the morphological and physiological differences between the gametophytic and sporophytic phases of Porphyra yezoensis, suppression subtractive hybridization (SSH) method is conducted to generate large-scale expressed sequence tag (EST). In this study, the carpospores instead of the filamental conchocelis were chosen as sporophyte material. The vegetative (gametophyte) and carpospores (sporophyte) cells live in the same blade of thallus so that the difference between the samples could be neglected.

The morphological and physiological differences between the leafy gametophyte and the filamentous sporophyte in the life cycle of Porphyra species are so great that many studies have been performed. Gametophyte- and sporophyte-specific cDNAs of Porphyra purpurea which encode proteins such as elongation factors, serine protease-like proteins, polysaccharide-binding proteins, and lipoxygenases had been isolated by differential screening and subtraction of phase-specific cDNA libraries (Liu et al., 1994, 1996) [11–15].

For single direct partial sequencing of anonymous cDNA clones of Porphyra purpurea performed by Liu et al was not available in obtaining enough genetic information. Lee et al. constructed the EST library of P. yezoensis thullus [16]. Among 190 ESTs generated, 81 sequence tags carried partial cDNA with significant amino acid sequence similarity to known genes, but the other 109 clones were not homologous to previously identified genes. At the same time, EST analysis has been performed to identify candidate genes related to the morphological and physiological differences between the gametophytic and sporophytic generations, and 20779 cDNAs have been identified [4, 5]. However, because these are mainly new genes, without homologous to known ones, the genes which regulate the life history process of Porphyra thalli remain poorly understood.

Gametogenesis of Porphyra thalli was induced by changing the photoperiod and water temperature. Kakinuma et al. (2006) have constructed subtracted cDNA libraries enriched for differentially expressed transcripts in vegetative and reproductive thalli, and randomly selected 1,152 cDNAs from each subtracted library [17]. Results of the dot blot analyses used for identification of differentially expressed cDNAs and BLAST analysis of nucleotide and deduced amino acid sequences showed that the cDNAs represented 63 and 59 unique clones for the vegetative and reproductive. Some of the cDNAs isolated from the reproductive subtracted library were homologous to genes encoding protein kinases, GTP-binding protein, and heat shock proteins involved in signal transduction and the molecular chaperon system. However, it is possible that changes in expression of these genes may only be a response to the change in temperature and photoperiod and are unrelated to gametogenesis.

To identify genes involved in the Porphyra yezoensis asexual sporulation, Kitade, et al. (2008) compared the gene expression profiles derived from gametophytes and sporophytes using cDNA macroarray, which includes 4,896 nonredundant expressed sequence tag (EST) groups. Candidate genes were screened by two different macroarray data analyses combined with reverse transcription-PCR (RT-PCR) analysis and Northern analysis. RT-PCR analysis revealed that nine genes were expressed in gametophyte specific manner, and two genes were expressed only in gametophytes that formed archeospores [18].

The cDNA microarray provides a high capacity and credibility system for the investigation of Porphyra yezoensis genes expression profile. For example, the different generations of P. yezoensis 467 functional gene clones were obtained and dotted onto the slides by cDNA microarray which coated by poly-lysine. The analysis of the normalized data showed that 55 genes were excessively expressed in gametophyte generation, 21 of which were similar to the genes of known or putative function; 86 genes were excessively expressed in sporophyte generation, and among them, 24 were similar to the genes of known or putative functions [19].

Compared with previous works on Porphyra yezoensis generation related gene expression, we constructed the difference expression libraries by the suppression subtractive hybridization (SSH) method The GO annotation showed that 58% cellular components of carpospores were located in cytoplasm and nucleus; while vegetative cells were mainly located in cytoplasm and nucleus, especially located in Golgi apparatus, and high expression on plastid, ribosome, endoplamic reticulum. In the thallus of vegetative growth, photosynthesis and carbohydrate assimilation, RNA transcription and protein translation, and material transport are the main biological processes. According to the cultivation manner of P. yezoensis in China, the thalli are aerial exposure on every six hours during low tide. The extreme tolerance to desiccation of high intertidal P. yezoensis is demonstrated and blades survive a loss of up to 85-95% of their water during the daytime low tide to become crisp sheets. Temperature (winter freezing stress at high latitudes and heat stress in summer) and highlight stress often accompanies the cultivation period. The genes in response to heat and oxidative stress are expressed at large amount.

The carpospores express more genes in transcription, transport, carbohydrate metabolism, particular in signal transduction, DNA and protein modification, protein and nucleotide metabolism. The carpospores tends to be more active in development, rather in growth so that the signal transduction and a series of metabolisms happened in order to get enough energy for releasing and drilling into calcium carbonate shell. The expressed sequence tags of filamentous sporophyte of Porphyra haitanensis analysis studied by Pang et al. also found the transport and transcription, responsing to oxidation genes expression, and so on, similar to this research [24].

Subtractive hybridization is the first technique to be widely used for the purpose of identifying differentially expressed genes on a global scale. Advantages of the technique include the ability to isolate genes with no prior knowledge of their sequence or identity and the use of common molecular biology techniques that do not require specialized equipment or analyses. Several limitations of the original protocols, such as requirements of large quantities of RNA and bias toward abundant genes, have been overcome by incorporation of PCR into the suppression subtractive hybridization (SSH) technique. However, SSH remains applicable only on pair-wise treatment comparisons and must be replicated with the tester and driver reversed to identify gene expression changes in both directions. Additionally, subtractive hybridization does not provide a quantitative measure of expression differences and is most efficient at identifying genes that are completely absent, rather than expressed less abundantly, in the driver sample.

Though the suppression subtractive hybridization (SSH) technique can overcome the bias toward abundant genes more effective than subtractive hybridization, there are still four genes obtained in this research, such as ribosomal protein L41(HS), 50S ribosomal protein L9(HS), ribosomal protein S11 (HS572889 and HS572733) and ATP-binding protein (HS572892 and HS572609) which are identified on both vegetative cells and carpospores. These genes are too abundant to remove their mRNAs so that they can be amplified by the subsequent PCR process. This is another limitation of SSH.

Eighteen genes have not been annotated in this study. Due to the insufficient research on functional genes and unclear function of many sequences, further investigations are needed.

According to the information of GO annotation, the vegetative cells tend to growth and self- protection, while the carpospores tends to be more active in development. Interpretation of the differentially expressed genes revealed new insight into the molecular processes of the generation alternation of Porphyra yezoensis. Due to the insufficient research on functional genes and unclear function of many sequences, further investigations are needed.