The use of 18S ribosomal DNA, ITS and rbcL molecular markers to study the genus Dunaliella (Dunaliellaceae) in Iranian samples: A phylogenetic approach

The genus Dunaliella (Dunaliellaceae, Dunaliellales) encompasses bi-flagellate and cell wall-less microalgae that exists in hypersaline environments (Massyuk 1973). Species of this genus are known as the only photosynthetic eukaryotes that can grow in a wide range of salt concentrations, varying from 0.05 to 5.0 M NaCl (Garcia et al. 2007). This microalga was first described by Teodoresco in 1905 (Oren 2005), which was followed by a number of studies aimed at classifying the Dunaliella species. Part of the importance of the Dunaliella taxon results from its high biological and biotechnological potential in the production of antioxidant pharmaceutical pigments (Barzegari et al. 2010). It also raises hopes for the production of the second and third generation of biofuels. Another relevant application of Dunaliella is the production of recombinant proteins, which is to be developed in the future (Dehghani et al. 2018; Dehghani et al. 2017). Dunaliella species and strains have many different characteristics such as varying growth rates, sizes, morphological features as well as various requirements regarding the medium, in addition to varied potential. With such parameters, much attention has been paid to developing more robust and reliable approaches to identify different species of Dunaliella (Borowitzka & Siva 2007). Until recently, several attempts have been made to encourage novel taxonomic methods to identify Dunaliella species. One of the first taxonomic systems for Dunaliella species was based on morphological and physiological characteristics (Massyuk 1973). Different Dunaliella species were cultured under a wide range of salinity and other cultivation factors to study the variability of morphological and physiological features (Massyuk 1973). However, the use of morphological and physiological characteristics is not an effective tool to unambiguously identify all species of the genus Dunaliella (Borowitzka & Siva 2007).

Another well-known taxonomic system of Dunaliella was described based on morphological and biochemical criteria. The taxonomy of Dunaliella species was revised by Borowitzka based on cell length/width, optimum salinity, stigma condition, flagella length, the existence and type of refractile granules, the type of symmetry, cell color, the maximum total carotenoid content, the type of carotenoid and the formation of aplanospores (Borowitzka & Siva 2007). However, it is known that morphological and even physiological features of Dunaliella species vary greatly depending on the developmental stage and culture conditions such as nutrient availability, light intensity and temperature fluctuations (Gomez et al. 1999; Markovits et al. 1993; Riisgard et al. 1980). For instance, Ben-Amotz reported that D. salina lacks canthaxanthin (a type of carotenoid) in the mature form of the species (Ben-Amotz et al. 1982). However, further research showed that canthaxanthin is a major carotenoid in aplanospores of the species (Borowitzka & Siva 2007). Apparently, ultra-structural studies cannot help to discriminate Dunaliella species (Parra et al. 1990). Therefore, new approaches using molecular biological methods are being developed to classify algae, including Dunaliella (Olmos et al. 2009).

Today, different molecular markers such as the 18S rDNA gene, the Internal Transcribed Spacer (ITS1 and especially ITS2), the nuclear-encoded SSU rRNA and chloroplast genomes are frequently used in the taxonomy of green microalgae (Coleman et al. 1994; Mai & Coleman 1997; Olmos et al. 2000; Proschold et al. 2001; Lemieux et al. 2015). Gonzalez et al. (2001) for the first time compared the PCR-RFLP patterns of the ITS region in Dunaliella species and claimed that ITS is a quick and reliable genetic marker for the discrimination of Dunaliella species (Ben-Amotz et al. 2009). Furthermore, the authors studied the ITS diversity in D. salina, D. bardawil, D. tertiolecta, D. parva, D. viridis, D. lateralis and D. peircei and isolated them by cluster analysis (Gonzalez et al. 1998). The conserved and variable regions of the 18S rDNA gene were considered as another critical marker for the identification and classification of eukaryotic organisms (Olsen et al. 1986). Surprisingly, studies of D. parva and D. salina showed that the 18S rDNA gene in these genera contains intron(s), belonging to group I (Wilcox et al. 1992). A set of conserved and specific oligonucleotide primers were designed and successfully used to identify D. salina, D. parva and D. bardawil (Olmos et al. 2009). Further, the rbcL gene (encoding the large subunit of RuBisCo) was also used as an adjunct marker for the classification of various plants and microalgae (Fredericg and Ramírez 1996; Freshwater et al. 1994). Remarkably, the rbcL marker has not been widely used for the classification of Dunaliella species, while this marker can be used to solve many ambiguities in terms of the Dunaliella taxonomy and systematics.

In the present study, the molecular identification of Dunaliella species from two different saline lakes of Iran was performed based on three key markers, including 18S rDNA, ITS, and rbcL markers. These molecular markers represent a newer and more reliable approach, based on which the phylogenetic positions of the Dunalialla species and related strains were analyzed and re-evaluated.

Molecular techniques and bioinformatics tools were used to identify the samples and re-evaluate the phylogenetic position of Dunaliella species based on the 18S rDNA, ITS and rbcL regions. Specific primers were designed to amplify both the 18S rDNA gene and the ITS region. Amplification results showed an approximately 2500 bp band on the electrophoresis gel (Fig. 2a). Digesting the PCR products with the Hhal restriction enzyme showed different RFLP patterns for the samples (Fig. 2b). Sequencing and blasting against the NCBI database showed high similarity with the sequences of D. salina and D. pseudosalina.

Figure 2

Moreover, the results of rbcL amplification in the Dunaliella species revealed the same size (~900 bp; Fig. 2c) with different RFLP patterns on the electrophoresis gel. Subsequently, these PCR products were sequenced and deposited in the NCBI database and based on the blasting results, the strains were named as D. salina isolate BAK and D. pseudosalina isolate MAH.

Phylogeny based on 18S rDNA

For this purpose, the data were analyzed using the Kimura 2-parameter model and the discrete Gamma distribution based on the numbers of BIC (21263.41) and AICc (20871.44) values. Our ML analysis of the 18S rDNA data supports a big split between D. lateralis strain Nepal and Dunaliella sp. Atacama with all other members of Dunaliella. The results suggest the closer relationship of the species to C. reinhardtii, C. vulgaris, and A. gracilis. In fact, these species are more divergent than other species of the genus Dunaliella. The strains KMMCC 1346 and UTEX LB 2538 of D. bardawil were classified together in the top of the phylogenetic tree in clade A. Moreover, D. salina isolate BAK was grouped with D. pseudosalina isolate MAH through high bootstrap values (90%). The KMMCC 1428 strain of D. salina is also more closely related to clade B. The KU07, KU11, and KU13 strains of D. salina were located together in clade C (100% bootstrap values). In addition, D. parva and D. polymorpha strain CCAP 19/14 were clustered together in clade D (92% bootstrap values; Fig. 3).

Figure 3

ITS based phylogeny

The ITS data were analyzed employing the Kimura 2-parameter model and the discrete Gamma distribution (K2 + G) based on the numbers of BIC (5759.90) and AICc (5421.40) values. Our ITS phylogenetic analyses showed that this marker could not be used solely for the classification of Dunaliella species. However, inconsistent with 18S rDNA, this approach designated D. lateralis strain Nepal as an outgroup in the phylogenetic tree, showing a great divergence with the other Dunaliella species (Fig. 4). The strains CCMP 364 and UTEX 999 of D. tertiolecta, D. bioculata strain UTEX 199, and D. parva are clustered together in clade A. The ITS data strongly support D. bardawil strain UTEX 2538 and D. primolecta strain UTEX as sister strains (clade B). Moreover, Dunaliella sp. SAS11133 and D. salina isolate BAK are grouped together with high support values (98%). Interestingly, D. viridis strain CONC002 and D. polymorpha strain CCAP 19/14 are also clustered together with 100% support values. According to reports based on the 18S rDNA, the KU07, KU11, and KU13, strains of D. salina were grouped together (100% bootstrap values) and they are more closely related to Dunaliella ABRIINW M1/2 (Fig. 4).

Figure 4

Phylogenetic analysis based on rbcL

The rbcL data were analyzed using the Tamura 3-parameter and the discrete Gamma distribution (T92 + G) based on the numbers of BIC (3456.53) and AICc (3145.85) values.

Our phylogenetic analyses based on the rbcL marker using the ML method showed that D. parva, D. pesudosalina isolate MAH and D. tertiolecta strain CCMP 346 are more closely related. Moreover, D. salina UTEX 200 and D. peircei UTEX LB 2192 are located in the same clade (86% bootstrap values), where Dunaliella sp. SAS11133 is closer to the members of this clade (100% bootstrap values). In addition, the rbcL data revealed that the UTEX 2538 and KMMCC 1346 strains of D. bardawil, Dunaliella ABRIINW M1/2, and D. salina isolate BAK are evolutionarily related. The KU07, KU11 and KU13 strains of D. salina are located together in the same clade, and accordingly they are sister strains. Surprisingly, based on the rbcL data, D. acidophila is more closely related to the members of clade D. Moreover, based on the rbcL data, D. viridis strain CONC 002 is highly divergent from the other Dunaliella strains (Fig. 5). To date, the rbcL sequence of D. lateralis strain Nepal has not been reported, and thus its taxonomic position remains unclear.

Figure 5

Combined phylogenetic analysis based on 18S rDNA, ITS, and rbcL

The data were analyzed employing Tamura-Nei and the discrete Gamma distribution (TN93 + G) based on the numbers of BIC (22 520.49) and AICc (22107.66) values. The ML phylogenetic analysis based on the combined 18S rDNA, ITS and rbcL sequences showed that D. lateralis strain Nepal is highly divergent compared to other Dunaliella species. Thus, in the presence of three different genera of microalgae (i.e. C. reinhardtii, C. vulgaris, and A. gracilis), D. lateralis strain Nepal is placed as an outgroup. Moreover, D. bioculata strain UTEX 199, D. parva and D. salina strain KMMCC 1428 are evolutionary related (clade A). D. pesudosalina isolate MAH and D. salina isolate BAK as well as D. peircei UTEX LB 2192 and D. salina UTEX 200 are clustered together in clade B and C, respectively. Further, the CCMP 364 and UTEX 999 strains of D. tertiolecta and D. primolecta strain UTEX 1000 are clustered together with high bootstrap values (clade D). In addition, KU07, KU11 and KU13 strains of D. salina, Dunaliella ABRIINW M1/2, and Dunaliella sp. SAS 11133 are grouped together with 100% support values (Fig. 6).

Figure 6

So far, several studies have been performed using different molecular markers for the Dunaliella classification. The taxonomy of Dunaliella species was revised by the ITS2 secondary structure and compensatory base changes (CBCs). Consequently, D. primolecta UTEX 1000 and D. bioculata UTEX 199 were renamed as D. tertiolecta (Assuncao et al. 2012). This suggestion is consistent with our combined analysis with the18S rDNA, ITS and rbcL markers, confirming that D. primolecta UTEX 1000 and D. bioculata UTEX 199 are clustered with D. tertiolecta strains. Furthermore, using the ITS marker, Gonzalez et al. (2001) proposed that D. percei UTEX 2192 should be renamed to D. tertiolecta. The results obtained in this work also verify these revisions in terms of the ITS data due to the grouping of D. percei UTEX 2192 with D. tertiolecta strains.

Another ambiguous case is the phylogenetic position of D. bardawil UTEX 2538. Based on the morphological features of D. bardawil UTEX 2538, it should be considered as D. salina (Borowitzka & Huisman 1993). However, all the analyses presented here revealed that D. bardawil UTEX 2538 was classified with D. bardawil strain KMMCC 1346.

The taxonomic position of D. salina UTEX 200 is also very unclear. According to studies of the ITS-RFLP and ITS sequences, this strain is more closely related to D. pseudosalina CONC 010 (Gonzalez et al. 2001). For comparison, recent physiological studies showed that D. salina UTEX 200 is more similar to D. viridis rather than D. pseudosalina CONC 010 (Cifuentes et al. 2001). Interestingly, phylogenetic studies using morphological characteristics identified D. salina UTEX 200 as a synonym of D. viridis (Borowitzka & Siva 2007). In the ITS tree, this strain is closer to D. bardawil strain KMMCC 1346, while other data showed that D. salina UTEX 200 is closer to D. peircei strain UTEX LB 2192.

Dunaliella sp. ABRIINW M1/2 has a different 18S rDNA arrangement with respect to the intron insertion site compared to the other Dunaliella strains (Hejazi et al. 2010). As regards the ITS2 secondary structure, it is believed that Dunaliella sp. ABRIINW M1/2 should be renamed to D. viridis (Assuncao et al. 2012). Consistently, the ITS data showed that Dunaliella sp. ABRIINW M1/2 is close to D. viridis strain CONC002, while the rbcL analysis revealed that the microalga is more closely related to the D. bardawil strains.

The 18S rDNA analysis shows that D. salina isolate BAK and D. pseudosalina isolate MAH are evolutionary close (90% bootstrap values). Based on the previous morphological studies, D. pseudosalina is larger than D. salina and accumulates a remarkable amount of canthaxanthin (a carotenoid that is not found in D. salina). It is believed that D. pseudosalina is phylogenetically close to D. salina and D. parva (Ben Amotz et al. 1982; Massyuk 1973). Our present data support traditional findings regarding D. pseudosalina and D. salina. Although rbcL data suggest that D. pesudosalina isolate MAH is related to D. parva and D. tertiolecta CCMP 364, D. salina isolate BAK is phylogenetically closer to D. bardawil strains. Furthermore, our findings confirm that KU07, KU11 and KU13 strains of D. salina are similar and are therefore considered as one strain.

Based on the former morphological and molecular data, D. lateralis strain Nepal shows a high divergence in relation to the other Dunaliella species. The ITS1 and ITS2 phylogenetic analyses showed that freshwater microalga D. lateralis strain Nepal is clearly different from the Dunaliella strains. Therefore, D. lateralis strain Nepal is not considered to be a member of the genus Dunaliella (Assuncao et al. 2012; Gonzalez et al. 2001). This concept is also supported by ultra-structural studies and the presence of contractile vacuoles (Borowitzka & Siva 2007; Melkonian & Preisig 1984). Accordingly, the presented analyses strongly support the notion that D. lateralis strain Nepal is clearly placed outside the Dunaliella phylogenetic tree.

D. acidophila strain CCAP 19/35 (other fresh water Dunaliella) is classified within the subgenus Dunaliella based on the ITS and rbcL studies (Assunçao et al. 2012), while the morphological data did not confirm this position (Borowitzka & Siva 2007). The ITS analysis revealed that D. acidophila strain CCAP 19/35 is clustered within the subgenus Dunaliella, nonetheless, it showed greater divergence from the other members of Dunaliella. D. acidophila strain CCAP 19/35 is genetically similar to Dunaliella strains, especially D. salina strains from the clade D concerning the rbcL data.

In fact, because of the variable morphology as well as the type and content of carotenoids under different environmental conditions, the classical methods cannot provide a reliable and proper approach to the classification of Dunaliella species. These characteristics may mislead us in terms of the identification and classification of different Dunaliella species. Moreover, 18S rDNA, ITS (ITS1 and ITS2) and rbcL genes are newer and more efficient markers for the taxonomy of microalgae. However, as mentioned above, these markers cannot serve as a fully reliable tool for phylogenetic and taxonomic approaches to Dunaliella species. We therefore believe that these markers should be used together to assess the phylogenetic position of this genus. Similar to several reports, our findings showed that the D. lateralis strain shows high divergence in relation to the other Dunaliella strains (Fig. 6).

Based on morphological, physiological and ITS approaches, D. tertiolecta and D. primolecta are clustered together as the Tertiolectae section (Oren 2010). As shown in Figure 6, the technique confirmed that D. primolecta strain UTEX 1000 is clustered with the other D. tertiolecta strains. In addition, depending on some physiological and biochemical features (e.g. salinity tolerance), it is suggested that D. peircei UTEX 2192 and D. parva (AC: M62998.1) were incorrectly named (Cifuentes et al. 2001). Accordingly, on the basis of the ITS marker, Gonzalez et al. (2001) stated that D. parva is misidentified and this strain should be named as D. viridis. However, because of high biodiversity within the ITS sequences of the mentioned strains (Oren 2010), this marker cannot be solely used to classify the strains. The method presented in this study is contrary to a later report, so that D. parva is clustered with D. salina strain KMMCC 1428 and D. bioculata strain UTEX LB 199 (Fig. 6). Similarly, based on the ITS marker, D. peircei UTEX 2192 is more closely related to the section Tertiolectae and is considered as D. tertiolecta (Gonzalez et al. 2001). The presented method revealed that D. peircei UTEX 2192 is more closely related to D. salina (Fig. 6). More specifically, due to the lack of a precise description and, more importantly, any other available strain for the Peirceinae section, the determination of the phylogenetic position of this strain is problematic (Oren 2010).

According to the morphological variability and physiological traits, D. bioculata is identified as a form of D. viridis (Massyuk 1973). Further studies proposed that D. bioculata UTEX 199 could belong to the Tertiolectae section (Oren 2010). In comparison, our approach showed that D. bioculata UTEX 199 is closer to D. parva (Fig. 6).

Traditionally, the section Dunaliella includes three species: D. parva, D. pseudosalina, and D. salina (Oren 2010). However, the ITS based analysis showed that D. pseudosalina is closer to D. viridis (Gonzalez et al. 2001; Oren 2010). The present study showed that D. pseudosalina isolate MAH and D. salina isolate BAK are clustered together with high bootstrap values, even though D. parva is more divergent from these strains. The clustering of D. parva into the Dunaliella section appears questionable (Oren 2010). Importantly, due to some morphological characteristics (color changeability) of D. salina UTEX LB 200, this strain should not be considered as D. salina (Oren 2010), while the presented analysis confirmed the report that D. salina UTEX 200 is more closely related to D. salina.

Some previous reports suggested that D. bardawil should be considered as D. salina or its variety (Oren 2010). On the other hand, the method presented in this paper confirms the grouping of D. bardawil strain UTEX 2538 from the USA and D. bardawil strain KMMCC 1346 from Korea (Fig. 6).

In addition, the present method offers a more reliable system for accurate phylogenetic analysis of the Dunaliella genus. Therefore, Dunaliella sp. ABRIINW M1/2 is more closely related to D. viridis and should be renamed as a variety of D. viridis. Strains KU07, 11 and 13 of D. salina are certainly one strain of D. salina. Further, Dunaliella sp. SAS11133 should be considered as D. viridis. Furthermore, D. salina strain KMMCC 1428 is more closely related to D. parva and should also be considered as D. parva (Fig. 6). Further attention and research are obviously necessary to shed more light on the phylogeny and taxonomy of Dunaliella.