Genome-Wide Analysis of Putative G-Quadruplex Sequences (PGQSs) in Onion Yellows Phytoplasma (Strain OY-M): An Emerging Plant Pathogenic Bacteria

Abstract

Phytoplasma, an emerging plant pathogen is an endocellular obligate parasite of plant phloem tissues with highly reduced genomes and low GC content. They contain a minimal set of genes essential for survival as an intracellular parasite. The role of G-Quadruplexes in pathogenicity has been reported in a variety of microbial pathogens. Detailed investigation on the genome wide occurrence and distribution of Putative G-Quadruplex forming Sequences (PGQSs) in the AT-rich genome of Onion yellows phytoplasma (strain OY-M) was carried out. Relative enrichment and depletion of these putative secondary structures in different genomic regions of OY-M was investigated with an aim to unravel their association with functionally important genomic locations. PGQSs density of 0.4407/Kbp was detected in the genome of OY-M phytoplasma, which is significantly higher than the average PGQSs density (0.136/Kbp) reported for other members of its phylum, namely Tenericutes. A non-random distribution of PGQSs across the length of the genome was observed. Putative promoter regions of OY-M were found to be particularly enriched in PGQSs followed by genic regions. The repeat rich regions were identified to have minimum PGQSs density. Presence of PGQSs in important genes such as those involved in secretory pathways of virulent factors, transport related functions, rRNA and tRNA was particularly intriguing. Our study reports for the first time a detailed investigation on the genome-wide locations of putative G-Quadruplexes in phytoplasma and highlights the need to further investigate their role in the metabolism and also in the mechanism of pathogenicity.

Introduction

G-Quadruplexes (G4s) are non-canonical, stable nucleic acid secondary structures held together by G–G base pairs [1]. Guanine bases (Gs) have intrinsic tendency to self-associate and form four-stranded helical structures [2]. The G-tetrads stack over one another to form the G4 structure stabilized by monovalent cations. The G-tetrads are held together by loops arising from the intervening mixed-sequence nucleotides. G4s can exist in several topological forms that depend on the conformation of glycosidic bonds of guanine bases, stabilizing monovalent cation and the number of G-tetrads stacking over one another [3].

The propensity of G-rich sequences to form secondary structures was first observed in 1910 by Bang [4] when he detected a concentrated solution of guanylic acid forming a polycrystalline gel. Therefore, though discovered 100 years ago, these G-rich sequences have evoked the scientists during the last two decades only to unravel their role in an array of organisms [5] ranging from viruses to human beings. More than 700,000 G4s have been detected in vitro in the human genome [6].These secondary structures target the vital machineries of replication, transcription and translation [2, 7]. Role of G-Quadruplexes in epigenetic modification has also been reported [1]. Harris and Merrick [3] outlined the roles of G-Quadruplexes in pathogen biology. In addition, G-Quadruplexes have been indicated to play role in antigenic variations in several pathogens and have also been implicated in silencing two major human viruses (HIV and EBV), thereby facilitating immune evasion and maintaining chronic infections [3]. Perrone et al. [8] reported G-Quadruplex mediated transcriptional regulation in long terminal repeat (LTR) promoter of HIV-1.

Comparative analysis of genome wide distribution of G-Quadruplexes in prokaryotes was first done by Rawal et al. [9]. Comprehensive genome wide analyses have been reported for E. coli [10] and human herpes virus [11] genomes. Recently, Ding et al. [12] conducted bioinformatic analysis in all representative microorganism genomes. Out of the total 1464 bacterial genomes (belonging to 12 phyla) analysed, genomes of E. coli, C. botulinum, M. acetivorans and H. pylori have considerably lower number of G4 forming sequences. Bartas et al. [13] reported a comprehensive survey of 1627 bacterial genomes for PGQSs density (no. of PGQSs/Kbp). At the subgroup level, they observed the highest mean density of 6.626 PQSs/Kbp in the subgroup Deinococcus–Thermus (which included eighteen different strains), whereas the lowest mean density of 0.136/Kbp was found in the subgroup Tenericutes (which included 52 different strains).

Phytoplasmas (phylum Tenericutes) are unculturable cell-wall less plant pathogenic bacteria which are obligate parasites of the phloem tissues. They are associated with a variety of diseases in several economically important plants worldwide. They are characterized as highly reduced AT-rich genomes with limited metabolic abilities but possess a highly functional transport system [14]. Phytoplasmas produce an arsenal of effectors which target host plant proteins leading to diverse developmental alterations in the host plants [15].

Their considerably small (i.e., 680–1600 Kbp) and repeat-rich genomes with very low GC-content (21–28%) is intriguing. The role of G-Quadruplexes in pathogenicity in case of microbes such as Neisseria gonorrhoeae and Trypanosoma brucei [3] and an unusually low GC content in phytoplasma evoked us to examine the presence of PGQSs in the whole genome of Onion Yellows phytoplasma (strain OY-M) which belongs to 16SrI-B group. The study also focusses on their occurrence and distribution in the genic, putative regulatory and repeat-rich regions of the genome. Phytoplasma releases effectors into the sieve cells of the host via Sec-dependent secretion system which is highly conserved and indispensable for its viability [16]. Thus, the presence of PGQSs in the secretory pathway genes (secA, secE and secY) of OY-M has also been studied. To the best of our knowledge, this is the first detailed genome-wide survey of the occurrence and distribution of PGQSs in the phytoplasma genome.

Materials and Methods

Genome Retrieval

Whole genome sequence of Onion yellows phytoplasma (strain OY-M) (NC_005303.2) was retrieved from NCBI GenBank [17].

Mapping of PGQSs

Putative G-Quadruplex forming Sequences (PGQSs) were mapped using QGRS mapper, a web-based server for predicting G-Quadruplexes in nucleotide sequences [18]. We have used C++ implementation of QGRS mapping algorithm. The following parameters which are the default parameters of QGRS mapper [18] were used: minimum G-tetrad = 2; loop length = 1–10; maximum length of PGQSs = 45 nts for all motifs with ≥ 3 tetrads and 30 nts for motifs with 2 tetrads.

Whole Genome

The genome was searched for PGQSs and their density (number of PGQSs/Kbp) was calculated. In order to analyze the pattern of occurrence of PGQSs (random vs non-random), the whole genome of OY-M phytoplasma was divided into eleven equal-sized genomic intervals of 80 Kbp each and PGQSs were searched in each interval on both the strands using QGRS mapper. The number of genes present in each genomic interval on both the strands was also calculated using a custom shell script.

Genic Regions

The genic regions were extracted from the OY-M genome using a custom shell script and searched for PGQSs and density of PGQSs was calculated.

Putative Regulatory Regions

In order to avoid operons, genes with a distance of minimum 200 bp between them were selected. Up to 2000 bp upstream regions of these selected genes were searched for PGQSs and PGQSs density was calculated.

Repeat Rich Regions

The repeat rich regions of the OY-M genome was identified using DACCOR (https://github.com/Integrative-Transcriptomics/Daccor) [19] which is used for detection, characterization, and reconstruction of repetitive regions in bacterial genomes. PGQSs were mapped in these regions and PGQSs density was calculated.

Estimation of GC Contents and Their Correlation with PGQSs Densities in Various Regions

GC contents of whole genome, genic regions, putative regulatory regions and repeat rich regions of the Onion yellows (OY-M) phytoplasma was calculated using shell script and the correlation between GC% and PGQSs densities in different regions of the phytoplasma genome was calculated.

PGQSs Mapping in the Highly Conserved and Essential sec Genes of the Sec Protein Translocation System in OY-M

Gene sequences (including 1000 bp upstream (UPS) and downstream (DS) flanking regions) of secA, secE, secY were retrieved from KEGG and presence of PGQSs in these regions was analyzed and compared.

Results and Discussion

Density of PGQSs in the Genome of Strain OY-M

The genome of OY-M is AT-rich with a low GC content of 27.80%. A total of 376 PGQSs were detected in the genome comprising of 174 and 202 PGQSs on the positive and negative strands respectively (Fig. 1). The number of PGQSs on the positive strand ranges from 5 to 32 in different class intervals with an average value of 15.82. Similarly, on the negative strand, the number of PGQSs ranges from 6 to 28 in different class intervals with an average value of 18.36 (Fig. 2). Thus, a non-random distribution of PGQSs was observed in the genome of OY-M. Density of PGQSs (number of PGQSs per Kbp) would be useful in revealing the important pattern, if any, with reference to their enrichment or depletion in a particular region of the genome. Due to significant variations in the sizes of different regions (genic, regulatory and repeat-rich) of the genome, PGQSs density would thus be important to assess any association between their occurrence and function. Despite its low GC content, the phytoplasma displays a considerably higher number of PGQSs with a density of 0.4407/Kbp (Table 1).

Fig. 1.

Putative G-Quadruplex forming Sequences (PGQSs) on positive and negative strands in different genomic intervals (80 Kbp each) in the genome of OY-M phytoplasma

Fig. 2.

Pattern of variation in the PGQSs with the number of genes in each class interval on the OY-M: a positive strand; b negative strand; c PGQSs on positive and genes on negative strand; d PGQSs on negative and genes on positive strand

Table 1.

Size (Kbp), GC%, number of Putative G-Quadruplex forming Sequences (PGQSs) and PGQSs densities (PGQSs/Kbp) in different genomic regions of Onion yellows phytoplasma (strain OY-M)

Genomic regions	Size (Kbp)	GC%	No. of PGQSs	PGQSs density (PGQSs/Kbp)
Whole genome	853.092	27.80	376	0.4407
Genic region (including both repetitive and non-repetitive regions of the genome)	677.476	28.88	226	0.3336
Repeat rich region	207.880	24.68	33	0.1587
Putative regulatory region	125.898	24.00	43	0.3415

Bartas et al. [13] showed the highest PGQSs density of 14.213 PGQS/Kbp in Thermus oshimai which belongs to the subgroup Deinococcus-Thermus (Terrabacteria group). They analysed that the lowest density of 0.013 PGQS/Kbp was found in Lacinutrix venerupis (subgroup Bacteroidetes/Chlorobi, FCB group). At the subgroup level, the lowest mean density of PGQSs was found in Tenericutes (0.136/Kbp), which also has the lowest average GC content (28%).The highest mean PGQSs density was observed in Deinococcus-Thermus (6.626/Kbp). Both of these subgroups fall under the Terrabacteria group. Phytoplasmas also belong to the phylum Tenericutes and thus, were expected to have very low PGQSs densities as found in other members of this group, as reported by Bartas et al. [13]. Therefore, strikingly high PGQSs density in the genome of OY-M phytoplasma is intriguing.

Correlation Between Number of Genes and Number of PGQSs

Lack of correlation was observed between the number of PGQSs and the number of genes on the same strand whereas a weak positive but marginally significant correlation (r = 0.5778 at P < 0.1) existing between the number of PGQSs on the positive strand of genome of strain OY-M and the number of genes on the negative strand might be of biological significance. A weak positive (r = 0.2768) but statistically insignificant correlation exists between the number of PGQSs on negative strand and number of genes on the positive strand. Thus, it appears that the number of PGQSs on a strand is not a function of the number of genes on that strand but the type of genes present. The genomic class interval between 240 and 320 Kbp has predominance of genes for ribosomal proteins and non-coding RNAs (tRNAs and 5srRNA, 16SrRNA and 23srRNA) on the positive strand and has a comparatively higher number of PGQSs. Wallgren et al. [20] reported enrichment of G4-motifs on rDNA sequences in S. cerevisiae. G rich nature of non-template DNA repeats coding for rRNA both within the exons and the spacers has also been reported by Hanakahi et al. [21]. Highest PGQSs frequency in the upstream and downstream regions of tRNA genes followed by gene body of tRNA and rRNA in bacteria was revealed by Bartas et al. [13]. Interestingly, non-coding RNAs have been reported to play important role in the regulation of pathogenicity by controlling virulence gene expression in Staphylococcus aureus and Vibrio cholera [22].

The class interval of 640–720 Kbp on the negative strand has comparatively higher number of PGQSs (Fig. 2a–d). This region too is rich in non-coding RNAs and also houses genes for secA subunit of protein translocation and ABC type amino acid transport system. The secA gene in Onion yellows phytoplasma strain OY-M encodes an 835 amino acid protein and shows high degree of similarity with secA gene in Bacillus subtilis [16]. This protein has ATPase activity which mediates translocation of proteins across the cytoplasmic membrane [23]. Thus, role of G-Quadruplexes in transporter-related genes of phytoplasma may also be speculated.

Correlation Between PGQSs Densities in Different Genomic Regions of OY-M and Their Respective GC Contents

A comparison between different regions of the OY-M genome revealed highest PGQSs density (0.3415/Kbp) in the putative regulatory regions of OY-M followed by genic regions (0.3336/Kbp) and repeat-rich regions (0.1587/Kbp) whereas the highest GC content was observed in the genic regions (28.88%) followed by repeat rich regions (24.68%) and putative regulatory regions (24.00%) (Table 1, Fig. 3). A weak positive but statistically insignificant (r = 0.5143) correlation between PGQSs density and GC% in different genomic regions of OY-M was obtained.

Fig. 3.

Densities of Putative G-Quadruplex forming Sequences (PGQSs/Kbp) in various genomic regions in the OY-M phytoplasma

In other words, despite low GC content in the putative regulatory regions, their PGQSs density is strikingly high. In fact, it is noteworthy that PGQSs densities in the genic and regulatory regions are rather similar despite their significantly different GC contents. Similarly, PGQSs density in regulatory regions is more than twice of that in the repeat regions though their GC contents are comparable. Enrichment of PGQSs in the regulatory regions (Table 2) is implicative of regulatory roles of G-Quadruplexes in phytoplasma. Perrone et al. [24] detected the presence of G-Quadruplexes in the promoter regions of Mycobacterium tuberculosis and has reported association of most stable G-Quadruplexes to the promoter regions of highly functional genes. Marsico et al. [25] also revealed enrichment of PGQSs around the transcription start sites and promoter regions in human and mouse and even in case of Trypanosoma genome. However, other eukaryotic species (C. elegans, zebrafish and Drosophila) showed reduced levels of PGQSs in the promoters, 5′UTRs, exons and 3′UTRs, and enrichment of PGQSs in non-coding RNAs. Earlier, a survey of > 61,000 ORFs across 18 prokaryotes have also shown enrichment of PGQSs in the promoter regions of genes involved in vital functions such as transcription, secondary metabolite biosynthesis, transport and catabolism and signal transduction [9].

Table 2.

Interval-wise number of Putative G-Quadruplex forming Sequences (PGQSs) in the putative promoters of Onion yellows phytoplasma (strain OY-M) on positive and negative strands

Genomic intervals (Kbp)	No. of PGQSs in putative regulatory regions (up to − 2000 bp upstream) of OY-M
	Positive strand	Negative strand
1-80	1	1
80.001-160	1	2
160.001-240	0	0
240.001-320	1	1
320.001-400	2	0
400.001-480	4	5
480.001-560	2	6
560.001-640	6	2
640.001-720	0	3
720.001-800	2	2
800.001-880	1	1

PGQSs density in the genic regions of OY-M was calculated as 0.3336/Kbp. Interestingly, the genomic intervals rich in rDNAs and transporter related genes have higher number of PGQSs. Eddy and Maizels [26] reported that different functional classes of genes in humans are associated with either very low or very high frequency of PGQSs. In humans, tumor suppressor genes have extremely low PGQSs densities, whereas proto-oncogenes are highly enriched in PGQSs. Thus, distinct functions of genes are driving forces for selection of genomic structures and regulation of their expression.

OY-M has a circular genome [17] and is rich in repeat regions unlike most other prokaryotes. The genome shows a discontinuous GC skew implicative of frequent recombination events and genome instability [27]. Though the GC content in the repeat rich regions (24.68%) of OY-M is slightly higher than that in the regulatory regions, the PGQSs density (0.1587/Kbp) is considerably lower as compared to regulatory regions or other genomic regions of OY-M This might be due to the fact that the repeat rich regions in the genome are mainly comprised of multiple redundant copies of core genes related to DNA recombination and replication. A correlation between positions of recombination breakpoints and distribution of PGQSs in the genome of HSV-1 was reported implicating the possible roles of G4s in the evolution of these viruses [28].The potential mobile units (PMUs) present in genome of OY-M characterized by large repetitive regions have been suggested to promote intrachromosomal recombination [27]. The PMUs contain several copies of genes related to DNA recombination and replication implicative of their propensity to transpose within the genome [29].Our study reports the presence of PGQSs in the PMUs of OY-M thus, the role of G-Quadruplexes in intramolecular recombination in case of OY-M also appears likely and worth investigating as has already been done in case of HSV-1 [28].

Presence of PGQSs in the sec Genes, secA, secE and secY

Phytoplasmas secrete effector molecules directly into the host cells via Sec-dependent pathway which is considered one of the most highly conserved protein secretion pathways ubiquitous in all domains of life [30]. It consists of three components; a protein targeting component, a motor protein and a membrane integrated conducting channel, the SecYEG translocase [31]. Genes encoding for SecA, SecY and SecE have been identified in ‘Candidatus Phytoplasma asteris’ OY-M strain, ‘Candidatus Phytoplasma asteris’ AY-WB strain, ‘Candidatus Phytoplasma mali’ AT strain and ‘Candidatus Phytoplasma australiense’ [16, 27, 32, 33]. Also secA protein has been expressed in plants infected with eight phytoplasma strains belonging to four distinct 16S groups [34] further confirming that the Sec system is widely conserved among phytoplasmas.

Analysis of PGQSs in these genes shows presence of PGQSs only in the putative regulatory regions of secY and not in the gene body. secE too has PGQSs in the upstream and downstream flanking regions but is devoid of its presence in the gene body. In contrast, secA gene shows the presence of PGQSs in the upstream regulatory region, gene body as well as in downstream regions in case of OY-M (Table 3).

Table 3.

Table showing number of PGQSs in different regions of highly conserved and vital genes secA, secE and secY involved in Sec translocation pathway of Onion yellows phytoplasma (strain OY-M)

S. no.	UPS (Upstream region)	Gene	DS (Downstream region)
secA	2	2	1
secE	1	–	1
secY	3	–	1

Conclusion

Despite having an AT rich genome, average density of PGQSs in OY-M genome is considerably high. PGQSs show a nonrandom distribution in Onion Yellow Phytoplasma (strain OY-M) which was not found to be a function of GC content. Further, significantly high PGQSs density in the regulatory regions in comparison to genic and repeat rich regions is noteworthy. Presence of PGQSs in the genes involved in vital functions such as seceretory pathway and transport related functions further highlights the importance of these secondary structures. Presence of G4s in different genomic locations of phytoplasma throws light on their probable role in pathogenicity as in case of other microbial pathogens [3], further implying that G4s can act as potential targets of therapeutic agents to control diseases. The findings obtained provide an important platform for further investigations on the role of G-Quadruplexes and will help in elucidating the mechanism of their functioning in prokaryotes in general and pathogenic bacteria with a skewed GC ratio in particular. Further, the results obtained also highlight the need for the experimental validation of the structures predicted as well as the precise mechanisms by which they affect the important metabolic processes of phytoplasma.

Compliance with Ethical Standards

Conflict of interest

Both authors declare that they have no conflict of interest.

Ethical Approval

This manuscript is original and has not been published elsewhere in any form or language. The authors discussed the result, read and approved the final article. The authors confirm that there are no ethical issues in publication of the manuscript.

Research Involving Human Participants and/or Animals

This study does not involve studies with human participants or animals performed by any of the authors.

Informed Consent

This article is original and has not been submitted or published elsewhere.