Research on Phylogenetic Relationship of Lotus Populations Collected in Thua Thien Hue Province, Vietnam based on the Chloroplast Genome by DNA Barcode

Biodiversity invasions are recognized as one of the most important causes of ecosystem degradation as well as local species community structure and biodiversity loss across the world (Thompson, 1997). Nelumbo nucifera  (N. nucifera) is known by numerous common names like sacred lotus, bean of India and simply lotus. N. nucifera is an aquatic herb with white or red-colored flowers (Maqbool et al., 2019). Since ancient times, the lotus flower has become familiar and close to Vietnamese’s life as well as other countries such as India, China and Japan. Besides, lotus is a valuable biological resource of Thua Thien Hue province, this is one of the indispensable parts of the natural heritage and provides not only local specialties but also services that are related to the ecosystem. Therefore, it is necessary to conserve and manage this species (Long et al., 2020).
       
Recently, there has been increasing interest in molecular techniques that have been widely used to analyze phylogenetic relationships among various breeds/populations (Avise et al., 1987a). DNA barcoding involves the use of a short DNA sequence or sequences from a standardized locus (or loci) as species identification tools (Alam et al., 2020). A DNA sequence from such a standardized gene region can be obtained from a small amount of tissue taken from an unidentified organism and then compared to a library of reference sequences from known species. An ideal DNA barcode should be present in all groups of land plants. it should be short (700-800 bp) and show enough sequence variation to discriminate among species, also it should be easy to amplify and sequenced with a single primer pair (Kress et al., 2009). Different regions from the plastid genome, including trnH-psbA intergenic spacer, rbcL, matK, rpoC1 and rpoB, have been proposed and tested for DNA barcoding of land plants with different level of species identification success depending on the studied group taxa (Long et al., 2020). N. nucifera is an aquatic flowering plant. This perennial usually lives in lakes and ponds.
       
This study was undertaken to test the utility of DNA barcoding to provide the theoretical basis for species identification, genetic diversity analysis of lotus population collected in Thua Thien Hue province, Vietnam. This study shall help make an informed decision on conservation, utilization and exploitation of the lotus genetic resources of the Thua Thien Hue province.

Plant materials
 
The lotus leaf samples (thirty-three samples: white and pink lotus) were collected from 33 different locations in Thua Thien Hue province (Fig 1). These are washed with distilled water and then refrigerated in the dark for further experiments.
 

       
This study was conducted at the Institute of Biotechnology, Hue University, May, 2018.
 
Methods
 
DNA extraction, PCR amplification and sequencing analysis
 
Total DNA was extracted from fresh leaf tissue (lotus leaf samples were stored at 4°C in the dark for about 1 to 2 days to remove part of the starch existing in leaf tissue) by using a CTAB method as described by Sharma et al., (2008).
       
The chloroplast genome DNA barcodes were amplified in a 25 µL reaction volume with universal primer sets presented in Table 1, using My Taq^TM DNA Polymerase (Bioline Reagents Ltd. UK), 0.5 µL primers (10 pmol/µL) and 100 ng DNA template (50 ng/µL). PCR amplification was performed on Applied Biosystems-Life Technologies (Thermo Fisher Scientific Inc. United States). PCR cycling conditions used in this study: 95°C/5 minutes; 30 cycles × (95°C/60 seconds; A°C/30 seconds; 72°C/60 seconds); 72°C/10 minutes (Annealing temperature is shown in Table 1).
 

       
PCR products were tested by electrophoresis on 1% agarose gel in TAE 1X buffer with Ethidium bromide dye and read electrophoresis images by direct UV reading system (UV-transilluminator, Model: Dyna Light). Samples showing a clear single band were sent to Maccrogen Company, Korea and sequenced in both directions with the same primers used for PCR by the method dideoxy terminator method on the ABI PRISM® 3100 Avant Genetic Analyzer (Applied Biosystems). Sequences were uploaded to the GenBank (Accession numbers: rbcL (MN011708 to MN068956); matK (MN011719 to MN068978) and trnH-psbA (MN011730 to MN086252); accD-psaI (MN086253 to MN086285); psbE-petL (MT901764 to MT901796); Rpl32-trnL (MT901731 to MT901763); trnW-psaJ (MT905225 to MT905257); trnSGCU-trnGGCC (MT905258 to MT905290) and ndhA (MZ611976 to MZ612008), respectively.
 
Data analysis
 
Raw sequences for each region were assembled and edited using BioEdit v7.2.5. Edited sequences were then aligned by ClustalW in MEGA X and the non-overlapping sequence regions at the 5'- and 3'- ends were trimmed (Kumar et al., 2018). The evolutionary history was inferred by using the Maximum Likelihood method and Tamura-Nei model (Tamura and Nei, 1993). The tree with the highest log likelihood (-12544.64) is shown. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. Evolutionary analyses were conducted in MEGA X (Kumar et al., 2018). The barcode sequences were queried against the GeneBank database (NCBI) using the Nucleotide BLAST algorithm.
       
The seven parameters including the number of separate polymorphic sites (S), the total number of mutant sites (Eta), number of haplotypes (h), haplotype diversity (Hd), the average number of nucleotide differences (k), nucleotide diversity (Pi) and Minimum number of recombination events (Rm) are considered as a polymorphic measurement in the population. Neutrality is tested based on five methods namely (Tajima’s D test (Tajima, 1989), Fs, Fu’s statistic (Fu, 1997); S, Strobeck’s statistic (Stroreck, 1987); D* and F*, Fu and Li’s statistics (Fu and Li, 1993) were used to DNAsp 6.0 software (Rozas et al., 2017).

Sequence characteristics of the barcode
 
The nine DNA barcodes of the lotus chloroplast genome showed high success rates for PCR amplification and sequencing using specific single primer pair. Sequence analysis results showed that the total length of nine chloroplast DNA regions of the two lotus study populations ranged from 8892-8921 bp (white lotus) and 8861-8916 bp (pink lotus) (Table 2). BLAST results on the National Center for Biotechnology Information (NCBI) showed that the obtained nucleotide sequence is very similar to that of the lotus species N. nucifera (Accecssion number: KF009944.1).
 

       
Sequence characteristics of a total of nine regions are present in Table 2 shows the nine sequences the within mean group distance in each lotus population was 0.180±0.900 × 10^-3 (white Lotus) and 3.390±0.450 × 10^-3 (pink Lotus). G+C content contained in the nine genetic regions is between different lotus samples was low, ranging from 0.334-0.335 (Table 2). The appearing percentage of each nucleotide type showed that Adenine (A) and Timin (Uracin) accounted for the highest proportion and there was no high difference between studied lotus samples and ranged from 32.863 to 33.180% (A) and 33.546 to 33.676% (T), reaching an average of 33.085 and 33.595, respectively. Meanwhile, the lowest proportions were Cysteine (C) and Guanidin (G), ranging from 17.207% to 17.454% (C) and from 15.971% to 16.138% (G), reaching an average of 17.266% (C) and 16.054% (G) of the lotus samples studied.
       
Besides, analysis of a total of nine genomic regions of two lotus populations using DNAsp 6.0 software showed that there are five separate polymorphic positions (S), create five mutant positions (Eta) for the white lotus population; eighty-four separate polymorphic positions (S) create eighty-five mutant positions (Eta) for the pink lotus population and eighty-five separate polymorphic positions (S) create eighty-six mutant positions (Eta) for the total of both lotus populations. A total of nine genomic regions with these differences were divided into three types of haplotypes (h) for the white lotus population; seven types of haplotypes (h) for the pink lotus population and nine types of haplotypes (h) for the total of both lotus populations. The haplotype diversity coefficient (Hd) accounts for 0.618±0.104 (white lotus); 0.818±0.050 (pink lotus) and 0.869±0.026 (total of both lotus populations), the average number of nucleotide differences (k) was 156.364×10^-2 (white lotus); 2908.225×10^-2 (pink lotus) and 2210.985×10^-2 (total of both lotus population). The nucleotide diversity coefficient (Pi) accounts for 0.180±0.050×10^-3 (white lotus), 3.280±0.830×10^-3 (pink lotus) and 2.550±0.720×10^-3 (total of both lotus population), respectively. The minimum number of recombination events (Rm=1) for the pink lotus population but not expressed in the white lotus population. All the indicators were processed with statistical significance p<0.05 (Table 2).
       
The results test neutrality shown in Table 3 with D, Li’s D* and F* values are negative for the white lotus population (D= -0.32197, D*= -0.76384 and F*= -0.73647, Not significant, p>0.10), shows an excess of rare nucleotide site variants and recently emerging haplotypes compared to what would be expected under a neutral model and either population expansion or background selection has occurred concerning for to the evolution of the lotus population under study (Fu and Li, 1993). While the D value of the pink lotus population and the combination of the total of both lotus populations yielded a positive value was with not significant p>0.10, this indicates the evolution of the pink lotus population is studied and both total lotus populations may have been suffered a recent bottleneck or we may have evidence for overdominant selection at this locus. Besides, the values of Fu and Li’s D* and F* of the pink lotus population and the combination of the white and pink lotus populations (Not significant: p>0.10), indicated that the study population has very few individuals show large differences in when compared to other individuals in the population (Table 3). Also, the results of the Fu’s Fs test, based on the distribution of haplotypes, showed positive values for the lotus populations as evidence for a deficiency of alleles, as would be expected from a recent population bottleneck or we may have evidence for overdominant selection at this population.
 

       
Following Fu’s Fs test, the hypothesis of natural evolution was significantly rejected for all regions of the lotus population under study. Strobeck’s S, the probability of obtaining equal or fewer haplotypes based on gene frequency and mutation rate is low in the white lotus population (Strobeck’s S statistic=0.446) and does not occur in the pink lotus population and total of both lotus populations (Table 3). These results are not consistent with deviation from neutrality due to either selection or population expansion.
 
Researchers often use Fst to assess gene flow, a higher Fst value indicates a lower level of gene flow (Nm) and higher genetic differentiation among populations (Hedrick, 2005). Fst reflects the level of inbreeding within populations (Wright, 1984) or the extent to which populations are differentiated (Hartl and Clark, 2007). The presence of genetic structure is an outcome of limited gene flow and a high level of genetic drift within each reproductively isolated group. Fst values below 0.05 indicate negligible genetic differentiation, whereas values greater than 0.25 indicate high genetic differentiation within the analyzed population (Weir, 1996). Fst values of the lotus populations in Thua Thien Hue, Viet Nam were significant but weak (Fst=0.00233; p<0.05, Table 4). This may be because primitive and highly conservative of the chloroplast genome in this species. This genetic distance implies that the kinship between unrelated individuals of the same ancestry relative to the lotus study population is equivalent to the kinship between individuals in a randomly mating lotus population. The presence of genetic structure is an outcome of limited gene ow and a high level of genetic drift within each reproductively isolated group. This study indicated that the two population genetic structure of lotus collection from different places in Thua Thien Hue, Vietnam are stable in both spatial and temporal scales. The random dispersal and their year-round spawning behavior might inhibit the population differentiation.
 
@table4
 
Phylogenetic analysis
 
The Phylogenetic tree was built based on Maximum Likelihood method (bootstrap=1000). The result is shown the first cluster groups a population of pink lotus (SH02, SH05, SH06, SH07 and SH09); The second cluster includes both the pink and white lotus samples and they are divided into 3 different sub-clusters. Sub-clusters one includes 7 pink lotus samples (SH04, SH10, SH11, SH12, SH20, SH21 and SH22), Sub-clusters two includes 11 white lotus (ST01-ST11) and 1 pink lotus (SH01) samples and Sub-clusters three includes the rest of the pink lotus samples (SH03, SH08, SH13-19). The tree topology is supported by a good bootstrap value. The differences between the two pink and white lotus populations were not found in the total nine regions of barcodes. Although, the two lotus populations have a different flower color, shared the same haplotype for the nine markers of the barcodes region, which are considered the most variable coding and non-coding regions of the plastid genome (Chase et al., 2007) (Fig 2).
 

Although the chloroplast genome contains many noncoding regions, relatively few have been exploited for interspecific phylogenetic and intraspecific phylogeographic studies. In our recent evaluation of the phylogenetic utility of non-coding chloroplast regions, we found the most widely used non-coding regions are among the least variable, but the more variable regions have rarely been employed (Long et al., 2020). To explore the potential variability of previously unexplored regions, we used nine gene regions in the study of thirty-three lotus samples that were collected in Thua Thien Hue province. The result presented in the study indicates that despite PCR and sequencing efficiency, unfortunately, this region can not be considered as an effective white and pink lotus variety of N. nucifera species barcode. Analyses involving this sequence showed only 0.562% polymorphism in the studied taxa. Our results also indicate the need to use a different region, e.g., the ITS region, trnH-psbA, rbcL, matK, rpoC1 and rpoB, to correctly identify white and pink lotus varieties of N. nucifera species. All of the new sequences have been deposited in GeneBank.
 

Dang Thanh Long was funded by Vingroup Joint Stock Company and supported by the Domestic Master/ PhD Scholarship Programme of Vingroup Innovation Foundation (VINIF), Vingroup Big Data Institute (VINBIGDATA), code No: VINIF.2020.TS.119.

Authors have declared that no competing interests exist.