Cytochrome b shows signs of adaptive protein evolution in Gerbillus species from Egypt

Abstract

Background

Amino acid polymorphisms in the mitochondrial cytochrome b (cytb) gene of four Gerbillus species have been investigated for their geographical distribution and possible functional significance. The sequences were obtained from a total of 20 specimens representing four species of genus Gerbillus collected from Siwa Oasis, Dabaa, Wadi El Natron, El Faiyum, and Baltim in Egypt.

Results

Our results identified a group of amino acid variant polymorphisms that were useful for both species taxonomic and biogeographic assignments. The results demonstrated that amino acid variants L>F173 (Leucine>Phenylalanine), A>M203 (Alanine>Methionine), and I>V221(Isoleucine>Valine) were specific to G. andersoni, while the variant V>M283 (Valine>Methionine) was only specific to G. andersoni from Baltim. The variants, L>P263 (Leucine>Proline) and M>T311 (Methionine>Threonine) were specific only to G. amoenus collected from El Faiyum. Compared to other amino acid variants, L>P263 was remarkably less frequent, and it was predicted using PROVEAN database tool to have non-neutral effects.

Conclusion

Amino acid polymorphisms within the cytochrome b gene could be assigned to specific geographic locations. They might prove suitable to track accumulated and recent environmental changes as they could represent signs of adaptive evolution.

Background

The Egyptian gerbils of the genus Gerbillus are among the more diversified rodent genera of arid and semi-arid regions of the old world. Occupying diverse habitats, these rodents present a useful model for investigating historical environmental changes that affected the arid, old world belt and drove the distribution and morphological diversification in mammals. In Egypt, the distribution and systematics of this genus and related rodent genera have been comprehensively covered (Osborn and Helmy 1980). In this genus, the molecular assessments have brought recent insights in a group of rodents where morphological studies have always dominated (Musser and Carleton 2005; Ndiaye, Chevret, Dobigny, & Granjon, 2016; Ndiaye et al. 2016b).

At the molecular levels, the most widely used population genetic indicator in animals is mitochondrial DNA (Avise et al. 1987). One of the important mitochondrial genes for survival of organisms is the cytochrome b (cytb) gene. This gene has proved particularly useful for discerning phylogenetics and taxonomic relations (Castresana 2001; Cook et al. 1999; Irwin et al. 1991; Kuwayama and Ozawa 2000; Kocher et al. 1989; Lau et al. 1998).

The cytb gene encodes a membrane-bound molecule, a central catalytic subunit of ubiquinol cytochrome c reductase (bc1 complex or complex-III), that is present in the respiratory chain of mitochondria (Howell 1989). Given the exception of protozoans missing mitochondria, all eukaryotes need this class of reduction-oxidation enzyme, and subsequently cytb, for energy metabolism (Hauska et al. 1983; Trumpower 1990; Widget and Cramer 1991). Further studies have evaluated the physiochemical changes which resulted from the molecular evolution of cytb functional domains in pocket gophers and cetartiodactyls (McClellan and McCracken 2001).

Although many scientific studies on the assessment of cytb amino acid residues at levels of both structure and function (Esposti et al. 1993) as well as physiochemical changes (McClellan and McCracken 2001) already date back to the end of the twentieth century, it is reasonable for investigation as they become more feasible with freely available annotated collection of mtDNA sequences. Some methods have been employed to discover signs of adaptation, among them identifying local declines in diversity, indicating selective sweeps in different populations of species (Andolfatto 2001; Nielsen 2005). Possibly, recent and established online database tools as SIFT (Ng and Henikoff 2003) and PolyPhen-2 (Adzhubei et al. 2013) that have been implemented for evaluating non-synonymous amino acid changes are also helpful in this process. In consistence with previous tools, prediction and assessment of single amino acid substitution using PROVEAN (Choi et al. 2015) become more attainable.

The present study aims to address the relationships between intra- and interspecies-specific mitochondrial cytb amino acid variants in genus Gerbillus from different geographical regions in Egypt. Specific amino acid variants will be assigned to different localities of the Western Desert of Egypt.

Methods

DNA techniques

Analysis was carried out on 20 specimens from different localities in Egypt (Table 1). Specimens were classified based on morphological taxonomic characters described by Harrison and Bates (1991) and Osborn and Helmy (1980). Four species have been identified: Gerbillus campestris, G. andersoni, G. amoenus, and G. gerbillus. Molecular analyses were done in the Laboratory of Molecular Biology, Zoology Department, Faculty of Science, Al Azhar University, Cairo. Genomic DNA was isolated from femoral muscle, using QIAamp® DNA Mini Kit.

Table 1 List of specimens with their GenBank accession numbers of the sequences of cytb (italicized numbers refer to original sequences in this study), geographic origin, haplotype code, and literature source

The complete cytb was amplified using Thermocycler GeneAmp 9700, USA. Pairs of primers, L14723 (5′-ACCAAT GAC ATG AAA AAT CAT CGT T-3′) and H15915 (5′-TCT CCATTT CTG GTT TAC AAG AC-3′), as described by Ducroz (1998), were used to target the cytb gene of 1140 bp size. The polymerase chain reaction (PCR) thermal program was set as follows. Initial denaturation at 94 °C for 1 min. Then 35 cycles of denaturation at 94 °C for 1 min, annealing at 62 °C for 1 min, and extension at 72 °C for 1 min. Final extension step at 72 °C for 10 min. PCR amplifications were carried out in 25 μl reaction volume including 0.5 μl of each primer (10 pmole/μl), 12.5 μl 2× PCR Master Mix solution (i-Taq™), 10.5 H₂O and 1 μl of template DNA. PCR products were purified using the QIAquick PCR purification kit and prepared for automated sequencing (3500 Genetic Analyzer, Applied Biosystems) relying on one of the primers utilized for the amplification. The acquired new sequences were all deposited in GenBank under serial accession numbers KX786151 to KX786155 and KX792465 to KX792479 after sequence quality assessments (Table 1).

Genetic data analysis

Sequence and database analysis

MEGA 7.0.14 software (Kumar et al. 2016) was used to align and proofread generated sequences from the sequencer. Conflicting DNA bases within DNA sequences were verified against the associated chromatograms. Accordingly, the cytochrome b sequences were generated for subsequent database analysis.

Basic Local Alignment Search Tool (BLAST)

To check and identify generated sequences, each was blast searched as a query through NCBI (National Center for Biotechnology Information) Blastn tool (www.ncbi.nlm.nih.gov/BLAST/). Sequences with best hits were retrieved and used as outgroups (Table 1) for further comparison to cytb sequences from the current study.

Multiple sequence alignment (MSA)

The nearly complete cytb gene sequences, 20 samples generated in this study were aligned against 24 sequences downloaded from GenBank (www.ncbi.nlm.nih.gov/genbank) of various well characterized Gerbillus species as a reference, and Sekeetamys calurus was used as an outgroup. Initial analysis, using DnaSP v.510 (Librado and Rozas 2009), of the total 44 cytb gene sequences gave 37 haplotypes represented as Hap1, Hap2, …, and Hap37 that will stand for future alignment.

Multiple sequence alignment was performed for DNA and protein sequences. The Bioinformatics Resource Portal (http://web.expasy.org/translate/) was used to translate the DNA sequence of cytb gene into its respective protein. Functional effect predictions of non-synonymous single nucleotide polymorphisms (nsSNPs) were achieved using PROVEAN (Protein Variation Effect Analyzer) (http://provean.jcvi.org/) (Kumar et al. 2014), which measures the damaging effect of variations in protein sequences (Choi et al. 2012). The prediction is based on the change in the similarity of the sequence to related protein sequences in a MSA by a delta alignment score of the reference and the variant carrying protein sequence with respect to the alignment of homologous sequences (Choi 2012). Variants with score equal to or below − 2.5 are considered as deleterious nsSNP.

Results and discussion

Multiple sequence alignment

Cytb DNA sequence

The alignment of cytb DNA sequences, from the current study and database, became more accurate after using a codon-wise approach which facilitated assuring positions of possible insertions and deletions. Codon-wise sequence alignment of cytb haplotypes from populations of different Gerbillus species, together with outgroup haplotypes, exposed most of the segregation sites which were probably enough for speciation events between different taxa.

Out of the 942 nucleotides used in the multiple alignments of cytb haplotypes, 282 variable sites compared to 660 conserved sites. From the variable sites, 38 positions were found as singletons while the remaining 244 variable sites were parsimony informative.

Cytb amino acid sequence

Multiple alignment of amino acid sequences is a standard technique for visualizing the relationships between residues in a collection of evolutionary related proteins. In the current study, multiple alignment of the cytb amino acid residues from Gerbillus species gave the chance to deepen the understanding of the evolutionary relationship between some species within the generic name Gerbillus.

Out of 313 deduced amino acid residues used in the alignment of cytb haplotypes (Fig. 1), 27 polymorphic sites were compared to 286 conserved sites. Out of these polymorphic sites, 19 sites were parsimony informative sites. Species-specific variations as well as those shared within the genus Gerbillus appear to be geographically/habitat related and could possibly be assigned to species biogeographic history. Possibly, these variations can be used to explain evolutionary relationships, as well as the type and effect of natural selection which might act differently in different geographic regions.

Fig. 1

Multiple sequence alignment of cytb protein residues haplotypes. “Hap” indicates haplotype; (.) represent the same amino acid as in the consensus sequence; (/ /) stands for sequences shared between all haplotypes. G.am: Gerbillus amoenus; G.na: Gerbillus nanus; G.an: Gerbillus andersoni; G.py: Gerbillus pyramidum; G.ge: Gerbillus gerbillus; D.ca: Dipodillus campestris (also common as Gerbillus campestris). Haplotypes with gray background are these from the current study compared to those from the database

DNA-based amino acid polymorphisms

The major species-specific and region-restricted amino acid variants within the cytb sequence of G. amoenus and its closely related congeneric species are compiled in Table 2 and Fig. 2. This evaluation of the multiple sequence alignment of cytb amino acid sequences proved helpful for exploring a group of variants, that are species specific, in more details. In addition, more understanding was achieved after using the online PROVEAN (Protein Variation Effect Analyzer) tool for filtering variants that possibly have an impact on the biological function of cytb protein.

Table 2 Intraspecific variation within different species of the Gerbillus species

Fig. 2

Cytb amino acid variants within and between species. G & S indicates Gerbillus & Sekeetamys; na, ca, am, an, cam, ge, & py, represent nanus, calurus, amoenus, andersoni, campestris, gerbillus, & pyramidum, respectively. Asterisk (*) is used to indicate variants which are geographically assigned

In general, among polymorphic sites that were identified from the alignment, 8 amino acid variants V>I2, I>L106, I>V176, L>M222, F>L226, I>L229, L>S237, and I>T295 were specifically portraying the outgroup genus Sekeetamys calurus (Hap37), compared to 15 variants V>A30, I>N31, L>F223, L>P263, M>T311, L>F173, A>M203, I>V221, V>M283, V>M30, A>S203, V>L30, F>Y217, I>T229, and V>I283 assigned specifically to species members from genus Gerbillus and were enough for separating this genus from its closest outgroup S. calurus, while 8 amino acid variants were in common between both genera (Fig. 2, Table 2).

In more details, amino acid variants V>A30 (Valine>Alanine), I>N31 (Isoleucine>Asparagine), L>F223 (Leucine>Phenylalanine) L>P263 (Leucine>Proline), and M>T311 (Methionine>Threonine) showed a restriction to G. amoenus, i.e., only 5 amino acid variant positions are specifically characterizing G. amoenus members from their closest species. Further, amino acid variants and similar scale of variations that showed a specificity to the other species are assorted in Table 2 and Fig. 2. Therefore, this possibly introduces a usable utility with a respective variation scale for each species. That way, collectively with inclusion of more genes, we demand further investigation to extend this scale of variation to cover more species and genera and develop a utility as a novel trend to probably solve and/or reduce recent controversies of molecular and morphologic taxonomy of genus Gerbillus reported by Musser and Carleton (2005) and Ndiaye et al. (2016a, 2016b).

In this work, among the evaluation outputs of multiple sequence alignment of cytb protein is the assignment of a group of amino acid variants to specific biogeographic regions. The variants V>A30 in haplotype 5 (Hap5, current study), I>N31 in haplotype 8 (Hap.8, database), L>P263 and M>T311 in haplotypes 1–4 (Hap1–4, current study), V>M283 in haplotypes 13–16 (Hap13–16, current study), and V>L30 in haplotype 36 were respectively assigned to Wadi el Natroun Egypt, Libya, El Faiyum Egypt, Baltim Egypt, and Tunisia (Table 2, Fig. 1). With these different localities, characterized by certain amino acid variants, we can record signs of interaction between environmental changes and wild life of diverse habitat overtime in Egypt in order to better track future changes and answer questions relating to population level.

Database prediction of different amino acid variants using PROVEAN allowed speculation of possible effects on the biological function of cytb protein in different localities. The PROVEAN predicted that the effects of cytb amino acids were mostly of neutral biological function except for amino acid variants L>P263 in G. amoenus collected from El Faiyum (current study) and L>S237 in S. calurus from Egypt (database) which were deleterious. These amino acid variants that gave neutral PROVEAN scores are given in Table 2. We are aware that any scores of in silico prediction tools should be taken with caution and might not actually represent the true effect of variants on biological function. The predictive accuracy of these tools was recently studied by Leong et al. (2015) who concluded that they must never be relied on as a final arbiter of pathogenicity but that they should rather be assessed as raising or lowering probabilities. Also, reports by Choi et al. (2012) have underlined that low scores are given to those amino acid residues found in conserved regions or domains while high scores were found in non-conserved regions when searched against the database of related sequences. PROVEAN has a higher specificity score compared to other efficient and commonly used tools like SIFT and PolyPhen-2 (Choi et al. 2012).

Interestingly, unlike other amino acid variants that are predicted to give a neutral PROVEAN score, L>P263 (Leucine>Proline), with its relatively small prevalence as a minor allele (Fig. 2, Hap4) in and specific to the El Faiyum depression, is predicted by PROVEAN to carry a deleterious functional effect (Table 2) with a score of − 4.513. Based on this predicted effect, it may be expected that G. amoenus members carry this “P” allele as a short-lived polymorphism, which does not persist in the population long enough to become fixed and rather become lost overtime due to lack of fitness and further fixation of the alternative “L” allele, possibly because of random genetic drift as similarly explained by Ohta (1992a, 1992b) who studied amino acid polymorphisms in both Drosophila and human mtDNA. However, a relatively similar study (Rand and Kann 1996) on one of the mitochondrial genes gave a different point of view, which suggested that some pressure of adaptive divergence might be acting in conflict with forces removing or rather eliminating variation. Interestingly, this might better describe the running scene of adaptive divergence in El Faiyum Depression. With the low sample size and one out of five gerbils at that location carrying the allele with the non-neutral, and hence possibly deleterious or possibly advantageous effect on the function of the vital cytb protein, we might either have sampled by chance a very rare individual, or the allele frequency might be indeed rather high. The latter scenario would argue against a deleterious effect of L>P263 and thus favor the hypothesis of positive adaptive evolution.

Conclusions

We evaluated the importance of multiple sequence alignment of cytochrome b amino acid sequences of the genus Gerbillus from different habitat types in Egypt. This alignment proved useful for assigning specific amino acid variants within and/ or between species and to specific geographic locations of varying environmental conditions. Herewith, the provided data and analysis approach will be a useful starting point for further investigation on the geographic distribution of cytb sequence variants and their possible role in environmental adaptation of Gerbillus spp. in Egypt.