gene diversity in populations of cyanobacterium

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

3.7. Discussion

This is the first molecular study of the cpmA gene diversity in populations of cyanobacterium Chroococcidiopsis from extreme habitats. The major findings of this study are low nucleotide diversity, low genetic differentiation among populations, high gene diversity and strong selection at the cpmA locus of Chroococcidiopsis sp.

3.7.1. Low levels of nucleotide diversity & low population differentiation globally; possibly contributed by a high migration rate

Low levels of nucleotide diversity in the globally distributed Chroococcidiopsis sp.

Despite the harsh environment and global distribution of the sampling locations, both the intra-population (Table 3.5) and inter-population diversity (Table 3.6) of Chroococcidiopsis sp. were quite low. The nucleotide diversity (Ï€) averaged only 0.0034 for the species as a whole. The theta (q) value, another indicator of genetic diversity within a population, ranged between 0.004 and 0.007 and was also rather low.

The average π value (0.021) of three housekeeping genes (atpD, glnII and recA) of, a Gram negative soil bacterium Bradyrhizobium was nearly six times higher than that of gene cpmA of Chroococcidiopsis sp. (average π = 0.0034) (Vinuesa et al., 2005). These genes of Bradyrhizobium are essential for nucleotide binding and catalysis, nitrogen stress response and DNA repair. Likewise, the average number of nucleotide differences for the cpmA gene of Chroococcidiopsis sp. (K = 1.74) were ten to fifteen-fold lower than that for the above genes of Bradyrhizobium (average K = 10.94) (Vinuesa et al., 2005). A comparison of the intra-population nucleotide diversity in gene cpmA sequences with that of the seven housekeeping genes of a cyanobacterium, Microcystis aeruginosa (Tanabe et al., 2007), yields similar results. The average nucleotide diversity reported for the genes of M. aeruginosa was 0.023, ranging from 0.013 (gene recA) to 0.043 (gene pgi); this is more than ten-fold higher than for the cpmA gene (Table 3.5).

Housekeeping genes are genes which are essential for any organism's activity. They are thought to be among the most conserved (Jordan et al., 2002). The above data indicate that at least some circadian genes are likely more conserved than the housekeeping genes.

Low inter-population differentiation globally

The very low inter-population differentiation of the globally distributed Chroococcidiopsis sp. (average GST and FST for the species were -0.0007 and 0.0079, respectively) observed in our study is quite surprising. For example, the above mentioned Bradyrhizobium, which was sampled at much smaller geographical scale (across Canarian Islands), had an average GST = 0.061 (Vinuesa et al., 2005). Furthermore, globally sampled strains of a cyanobacterium, Mastigocladus laminosus, showed quite high differentiation at several nitrogen metabolism loci (FST = 0.60) (Miller et al., 2007).

The low GST and FST values for the cpmA gene of Chroococcidiopsis sp. indicate that the majority of the variation resides within populations. The significant difference among the FST values between Chroococcidiopsis and M. laminosus is probably due to the differences in spatial isolation of the populations. Populations of M. laminosus were suggested to have low gene flow due to the dispersal barriers, whereas Chroococcidiopsis sp. seems to have a high inter-population gene flow (mean Nm = 31.25).

Similar low interpopulation differentiation despite the huge distribution area may be found in other species. For example, eukaryote Pinus sylvestris (Scots pine) is a main forest species distributed across Eurasia, from Spain to Kamchatka Peninsula in Russia. However, it showed very low nucleotide diversity in several functional genes, which was explained by low mutation rate and high gene flow (Dvornyk et al., 2002a).

The high migration rate is likely one of the most important factors for the low genetic diversity and differentiation among the populations found in this study. Other likely causes include high conservation of the cpmA gene (section 3.7.3) and an expanding population of Chroococcidiopsis sp. (section 3.7.4).

3.7.2. Large number of rare haplotypes in spite of the low nucleotide diversity

Despite the low nucleotide diversity, the haplotype diversity in Chroococcidiopsis sp. was high (H = 0.774). Similar high gene diversity (H = 0.951) was reported in a cyanobacterium Microcystis aeruginosa genotyped at seven housekeeping loci (Tanabe et al., 2007). These cyanobacteria, Chroococcidiopsis sp. and M. aeruginosa have similar high gene diversity. However, the average nucleotide diversity (π = 0.0034) for the cpmA gene of Chroococcidiopsis sp. was almost seven-fold lower than that reported for the multiple genes of M. aeruginosa (average π = 0.023) (Tanabe et al., 2007).

The high gene diversity in M. aeruginosa (Tanabe et al., 2007) was rather unexpected, because it was thought that clonal organisms such as bacteria should experience selective sweeps at most loci resulting in low gene diversity (Atwood et al., 1951). In order to explain such high gene diversity, Tanabe et al. suggested the existence of several ecologically distant populations of M. aeruginosa (Tanabe et al., 2007). Higher genetic diversity is maintained if more ecologically distant populations exist within a single species, as each distant population falls into its own sequence-based cluster (Cohan, 2002). This possibility was also supported by several distinct clusters observed in the phylogenetic tree of M. aeruginosa (Tanabe et al., 2007). However, this scenario is unlikely for the cpmA gene in Chroococcidiopsis sp, because differentiation among the populations of Chroococcidiopsis sp. is very low and no resolved clades are observed in the phylogenetic tree (Figure 3.4).

The high number of rare haplotypes may be an indicator of a rapidly expanding population, and is not necessarily an indicator of high genetic diversity (Slatkin & Hudson, 1991), as further discussed in section 3.7.4.

3.7.3. High conservation of the cpmA gene

The previous micro-site study on the cpmA gene (Yau & Dvornyk, unpublished data) of cyanobacterium Nostoc linckia from the "Evolution Canyon" revealed no polymorphism in the cyanobacterial populations from both the stressful and the temperate slopes. On the other hand, genome-wide studies of N. linckia from the same location showed significantly higher genetic diversity in the populations from the stressful slope as compared to those from the temperate slope (Krugman et al., 2001; Satish et al., 2001). The high conservation of the cpmA gene was also supported by the low nucleotide diversity observed in the gene on a global scale.

Higher conservation for a gene usually occurs if that gene is essential according to the "knockout-rate" prediction (Jordan et al., 2002). This is because there is stronger purifying selection acting on essential genes than that for less essential genes, which are more functionally redundant. Stronger purifying selection leads to lower rates of substitution, and, ultimately, to a higher degree of conservation of a gene; the level of conservation is generally related to the functional importance of that particular gene (Jordan et al., 2002).

The high conservation for circadian genes is supported by a study of the kai genes of the N. linckia that revealed evidence for purifying selection (Dvornyk et al., 2002b). Purifying selection was detected on the cpmA locus from our findings by dN/dS ratio, Tajima's D, and Fu and Li's D* and F* values. This also supports the assumption of high conservation of circadian genes (mentioned in detail in section 3.7.5).

3.7.4 Low genetic diversity in Chroococcidiopsis sp. explained by population expansion and/or genetic bottlenecks

Chroococcidiopsis sp. had low nucleotide diversity and generally low level of differentiation among the populations at the cpmA gene. This can also be a result of one or more recent genetic bottlenecks followed by expansion of the population.

Expanding population with one or more recent genetic bottlenecks

At present, Chroococcidiopsis sp. has a large population size and is dominant in most extreme environments (Friedmann, 1980). Chroococcidiopsis sp. is one of the evolutionarily oldest cyanobacteria, based on the morphological resemblance to certain Proterozoic microfossils (Friedmann & Ocampo-Friedmann, 1994). Low nucleotide diversity, low differentiation, and multiple low frequency haplotypes are consistent with a possible population expansion (Bodkin et al., 1999; Slatkin & Hudson, 1991).

We observed the negative values of Tajima's D, and Fu and Li's D* and F* test statistic in all populations. The parameters of above two neutrality tests, yielded the statistically significant (P ≤ 0.05) average for the species: D = -2.6443, D* = 4.7781 and F* = -4.7860). This represents an excess of low frequency polymorphism.

Tajima's D, and Fu and Li's D* and F* are sensitive tests for the frequency distribution of segregating sites. In a constant-size neutral equilibrium population, Tajima's D is expected to be nearly zero (Tajima, 1989). In terms of population size changes, negative values of Tajima's D are observed if the population has experienced one or more genetic bottlenecks and/or is increasing in size (Tajima, 1989). A rapidly increasing population would result in many haplotypes occurring at low frequencies, but with low overall nucleotide diversity (Slatkin & Hudson, 1991). This may be the case observed for the globally distributed Chroococcidiopsis sp. populations (Section 3.7.2).

Expansion of a population usually occurs after one or more genetic bottlenecks causes a decline in genetic diversity in the population (Robert, 1987). A population bottleneck occurs, for example, when a small number of individuals' groups attempt to find a new population (also known as the 'founder effect'). This may result in a significant loss of genetic variation (Leberg, 1992). However, our findings indicate low genetic variation for all populations that are geographically quite distant; therefore the likelihood of one or more recent genetic bottlenecks in all locations simultaneously is low.

Chroococcidiopsis sp. is more often found in extreme environments than in the temperate ones (Friedmann & Ocampo-Friedmann, 1994). It was proposed that this results from the inability of Chroococcidiopsis sp. to compete with more discriminating, specialized, or aggressive species abundant in temperate environments (Friedmann & Ocampo-Friedmann, 1994). It is possible that, low variation in Chroococcidiopsis sp. could be more suited for extreme environments. Due to this possibility, the species may not be able to compete with organisms of higher variation in moderate environments.

The long evolutionary history of Chroococcidiopsis sp. and its abundance in most extreme environments do not support the possibility of recent bottlenecks. No such cases were reported for other cyanobacteria as well (Mes et al., 2006). Therefore, more studies of multiple genes are required from Chroococcidiopsis sp. in order to study the population size changes.

3.7.5. Selection as a factor of evolution at the cpmA locus of Chroococcidiopsis sp.

Purifying selection is the most common type of selection observed. For example, only purifying selection was detected for housekeeping genes of two strains of Bradyrhizobium sp. (dN/dS ratio averaging 0.108). Even at a global scale, housekeeping genes of another bacterium Bacillus cereus, only exhibited purifying selection (dN/dS = 0.031) (Priest et al., 2004). However, in some cases, positive selection was also detected. For example, Mes et al. (2006) found that the petB and kaiC genes experience this type of selection at some sites. Our findings indicate strong selective pressure upon the cpmA locus. Interestingly, both positive and purifying selections were detected in the different populations of Chroococcidiopsis sp.

Positive selection in populations AC, AH and SH

The dN/dS ratio was above 1 in three populations, AC, AH and SH. The dN/dS test requires a rather strong signal in order to detect selection and is quite conservative (Plotkin et al., 2004). Therefore the respective ratio is rarely above 1 signifying positive, diversifying selection. For example, a large-scale database search performed by Endo et al. (1996) identified only 17 genes that are under positive selection out of 3,595 functional genes.

Although rare, examples of positive selection have been reported. For instance, 55 protein coding genes has been listed by Yang and Bielawski (2000) and two genes of cyanobacteria has been reported elsewhere (Mes et al., 2006) detected by the dN/dS ratio. Positive selection results in selection of favorable mutations or selective sweeps lowering genetic diversity (Galtier et al., 2000). A comprehensive examination of 12 protein coding genes of cyanobacteria revealed positive selection at the kaiC gene of Microcoleus chthonoplastes and the rbcX gene of Anabaena and Aphanizomenon sp. (Mes et al., 2006). The kaiBC operon, which encodes two core circadian genes, kaiB and kaiC, was also reported to be under positive selection in another cyanobacterium, Nostoc linckia (Dvornyk et al., 2002b).

Purifying selection in populations AE, CH and TH

The negative Tajima's D, and Fu and Li's F* and D* (Table 3.7) suggest strong purifying selection acting on the cpmA gene in the studied populations. For the populations, AE, CH and TH, as well as the average of the species, the dN/dS ratio was below 1 implying purifying selection. Purifying selection is common in which deleterious mutations are eliminated to preserve functional genetic features due to being conserved over time. These functional genetic features can include protein coding genes or regulatory sequences.

Selective neutrality tests affected by population size changes, possible cause of contradiction between dN/dS and neutrality test values

The amount and pattern of DNA variation in a population may be affected by natural selection. However, other than natural selection, changes in population size and structure are main factors that affect DNA variation (Tajima, 1989). dN/dS ratio above 1 for three of the populations implied positive selection. However, negative values of Tajima's D, and Fu and Li's D* and F* for all populations implied purifying selection in our study. Tajima's, and Fu and Li's test indices, are powerful but can be affected by changes in population size and therefore can give inaccurate interpretation of selection (Tajima, 1989; Fu and Li, 1996).

In an expanding population, there would be a tendency for Tajima's D values to be negative and vice versa (Tajima, 1989). For example, the dN/dS ratio above 1 of the AMA1 (Apical Membrane Antigen 1) gene in Plasmodium falciparum indicates positive selection; however, the observed positive values of Tajima's D, and Fu and Li's D* and F* suggest balancing selection acting upon the gene (Polley and Conway, 2001). Thus the two approaches of tests for selection provided contradictory results. This contradiction between neutrality test values and dN/dS ratio was also present in our findings. The difference in the results obtained from these two approaches was explained by a recent reduction in the P. falciparum population, as changes in population size can cause skewed neutrality test values (Polley and Conway, 2001).

Mes et al. (2006) obtained negative values of Tajima's D in the kaiC and petB genes of a cyanobacterium Microcoleus chthonoplastes. These genes are essential for circadian regulation and photosynthesis, respectively. The negative D may indicate either purifying selection or expanding population size, or both. However, the dN/dS ratio was higher for locus kaiC than petB (Mes et al., 2006). Since, dN/dS ratio is rather selective and uninfluenced by population size changes; this suggests that these two genes from M. chthonoplastes are under different selective pressures. This was confirmed by the significant evidence found for positive selection at locus kaiC by McDonald-Kreitman tests, which was not found for the petB locus. This finding was deemed to be important as it allowed the separation of gene-specific forces (such as positive selection) and genome-specific forces (population size changes and genetic bottlenecks) (Mes et al., 2006).

We discovered similar differences in selective pressures for different populations, identified by the neutrality test values and dN/dS ratio. However, data on other genes from Chroococcidiopsis sp. are not available to distinguish between gene-specific forces and genome-specific forces on the population. Therefore, in order to distinguish between selective pressures and population size changes in Chroococcidiopsis sp., more genes need to be studied. The M. chthonoplastes data set is an excellent example of a study on multiple genes from a single cyanobacterium which identified forces that affect genomic diversity without the knowledge on population size changes.

Considering the power of the dN/dS ratio as it stays unaffected by population changes and the limitations of the neutrality tests used, it can be assumed that the cpmA locus in Chroococcidiopsis sp. is most likely under selective pressures for all populations; some with rare positive selection (AC, AH and SH populations) .

Significant neutrality test values rejecting the neutral theory model at equilibrium for populations TH, CH and the average of the species, supporting high conservation of the cpmA gene

Significant values of Tajima's D, as well as Fu and Li's F* and D* (Table 3.7) imply that the populations TH and CH, as well as the overall species are not randomly evolving for the cpmA locus. Significant negative Tajima's D, and Fu and Li's F* and D* usually suggest either balancing selection operating on the population or a demographically expanding population (Tajima, 1989). However, the dN/dS ratio clearly indicates positive selection for the three populations, AC, AH and SH. It is possible that, even though positive selection is operating in some populations, it is not strong enough to overcome the effects of purifying selection in the average of the species. Since skewing of the Tajima's D and Fu and Li's F* and D*, and expansion of population is indeed a possibility, the overestimation of the neutrality test values can not be ruled out.

The strong purifying selection observed on the species can also be an indicator of higher conservation of the cpmA gene and circadian genes proposed in section 3.7.3. Comparisons of several functional genes from other bacteria with the cpmA gene of Chroococcidiopsis sp. revealed low nucleotide diversity and differentiation. Although genetic diversity was higher for multiple genes in Microcystis aeruginosa, Tajima's D was observed to be almost close to zero for all genes signifying weak selective forces acting upon those genes (Tanabe et al., 2007). This could imply the cpmA gene of Chroococcidiopsis sp. are more conserved, and therefore have more selective pressure than the functional nitrogen metabolism genes in M. aeruginosa.

3.7.6 Contrast among microbial communities in a micro-site due to varying levels of environmental stress

Cyanobacteria are remarkable for their ability to flourish in diverse environments with fluctuating stress factors, such as temperature, water availability and light intensity (Schopf, 1996). A study on the cyanobacterial circadian genes under contrasting environments with long-term microclimatic stress revealed that permanent ecological stress may result in a higher mutation rate and higher genetic polymorphism (Dvornyk et al., 2002b). From our findings it could not be deduced if stress caused higher polymorphism in the cpmA gene of Chroococcidiopsis sp., as samples that were used in our study were only from stressful sites and not from temperate. However, variation was observed in the polymorphism level in the cpmA gene sequences of six different populations from four sites. This indicates differences in the stress levels that these six populations are exposed to, even for populations within a habitat.

Antarctic populations; from same habitat yet with different selective forces on the cpmA locus

The three populations from the Antarctic Balham Valley used in our study were AC, AE and AH. They differ by their microbial community structure (Table 3.2). Although their nucleotide diversity (p), gene diversity (Hd), K and theta (q) values were similar, the dN/dS ratio was quite different. The AE population had the lowest dN/dS (0.37) as compared to those of the AC and AH populations (1.06 and 1.64, respectively).

Although, the three Antarctic populations were from the same habitat with similar environmental conditions, UV radiation is assumed to be highly different. This is because AE consisted of colonies inhabiting pore spaces deep within rocks, whereas AC and AH consisted of colonies inhabiting the surface and underneath the rocks (Figure 3.1). Considering the localization of these three types of microbial communities, it could be assumed that the least solar radiation would be received by AE, followed by AH and AC.

Previous studies on cyanobacteria under stress showed that UV radiation is one of the major stress factors in stressful environments (Dillon et al., 2002). Also, recombination frequencies and mutation rate tend to increase under stressful conditions (Hoffmann & Hercus, 2000). Therefore, if endolithic communities were exposed to lower stress levels than the chasmolithic and hypolithic communities, they should have lower polymorphism. K and theta (q) values support this assumption, as these were found to be somewhat lower for AE population compared to that of AC and AH.

From the dN/dS ratio, AE populations were detected to be under strong purifying selection, whereas AC and AH both were detected to be under strong positive selection. Our data suggests that, there is selective removal of deleterious alleles in AE population by purifying selection that could result in a decrease in variation. On the other hand, AC and AH populations were detected to have amino acid changes fixed at a higher rate than synonymous changes by positive selection, allowing adaptive molecular evolution. It is thought that evolutionary rates are higher in more stressful environments (Nevo, 1997). Therefore it could be due to the exposure of higher UV stress, that AC and AH populations are under positive selection for cpmA locus. This allows these populations to adapt to the harsh environment by gaining fitness of the protein, CpmA.

It was reported from a study on microbial diversity in Antarctic Balham Valley by Pointing et al. (2009), that even though the three microbial communities (AC, AE and AH) were present in the same habitat, they were highly different from each other (with an average FST value of 0.088), which supports our observation in differences in polymorphism levels in the cpmA gene for the three microbial communities. In addition, our findings reveal different selective forces acting upon the three microbial communities in the Antarctic Balham Valley. We have not found data in the available literature that report different selective forces for same locus. This is why our finding can deem to be important in understanding of the link between environmental stress and selective pressure.

3.8 Conclusion

Compared with other housekeeping genes of bacteria, nucleotide diversity and differentiation was rather low in the cpmA gene of Chroococcidiopsis sp. A high gene flow and/or population expansion with one or more genetic bottlenecks, supported by large Nm value, are considered to be plausible reasons behind the low genetic differentiation among globally separated populations. Chroococcidiopsis sp. were detected to be under strong selective forces (with both rare positive selection and purifying selection) for the same locus. A strong purifying selection found for the whole species, and low variation in the cpmA gene allowed us to propose a high conservation of the cpmA gene. In addition, our findings reveal different selective forces acting upon three microbial communities in the Antarctic Balham Valley, on the same locus. More molecular studies of multiple genes from Chroococcidiopsis sp. populations and/or from other cyanobacterium species globally, are required to confirm the high migration rate and/or population size expansion.