Detection, characterization and evolutionary aspects of S54LP of SP (SAP54 Like Protein of Sesame Phyllody): a phytoplasma effector molecule associated with phyllody development in sesame (Sesamum indicum L.)

Abstract

SAP54, an effector protein secreted by phytoplasmas has been reported to induce phyllody. S54LP of SP (SAP54 Like Protein of Sesame Phyllody), a SAP54 ortholog from phyllody and witches’ broom affected sesame (Sesamum indicum L.) was amplified, cloned and sequenced. Comparative sequence and phylogenetic analysis of diverse phytoplasma strains was carried out to delineate the evolution of S54LP of SP. The degree of polymorphism across SAP54 orthologs and the evolutionary forces acting on this effector protein were ascertained. Site-specific selection across SAP54 orthologs was estimated using Fixed Effects Likelihood (FEL) approach. Nonsynonymous substitutions were detected in the SAP54 orthologs’ sequences from phytoplasmas belonging to same (sub) group. Phylogenetic analysis based on S54LP of SP grouped phytoplasmas belonging to same 16SrDNA (sub) groups into different clusters. Analysis of selection forces acting on SAP54 orthologs from nine different phytoplasma (sub)groups, affecting plant species belonging to twelve different families across ten countries showed the orthologs to be under purifying (negative) selection. One amino acid residue was found to be under pervasive diversifying (positive) selection and a total of three amino acid sites were found to be under pervasive purifying (negative) selection. The location of these amino acids in the signal peptide and mature protein was studied with an aim to understand their role in protein–protein interaction. Asparagine residues (at positions 68 and 84) were found to be under pervasive purifying selection suggesting their functional importance in the effector protein. Our study suggests lack of coevolution between SAP54 and 16SrDNA. Signal peptide appears to evolve at a rate slightly higher than the mature protein. Overall, SAP54 and its orthologs are evolving under purifying selection confirming their functional importance in phytoplasma virulence.

Electronic supplementary material

The online version of this article (10.1007/s12298-020-00764-8) contains supplementary material, which is available to authorized users.

Introduction

Sesame (Sesamum indicum L.) is one of the oldest and an important oilseed crop of the Pedaliaceae family. It is grown widely in Africa, Asia and South America for its high quality nutritional seeds (Dossa et al. 2017). It ranks ninth among the top thirteen oilseed crops which make up 90% of the world production of edible oil (Adeola et al. 2010). In addition to oil, sesame seeds are an important source of protein (18–25%), carbohydrates (13.5%), vitamins and minerals also (Bedigian et al. 1985). Sesame oil is particularly valued for containing lignans such as sesmin and sesmolin that protect it from rancidity, thus giving it a long shelf-life inspite of containing high amounts of unsaturated fatty acids (Moazzami et al. 2006; Uzun et al. 2008). Though it is known as the “queen of the oilseeds”, it has often been considered an ‘orphan crop’ as it is not a crop mandated to any international crop research institute (Bedigian and Harlan 1986; Bhat et al. 1999).

Lack of high yielding varieties having resistance to biotic and abiotic stresses, low harvest index, seed shattering, and indeterminate growth habit are some of the important reasons of low sesame productivity (Ashri 1998). Among the biotic stresses, phyllody is one of the major biotic constraints severely affecting sesame yield (Rao and Nabi 2015). In sesame, phyllody disease is associated with phytoplasma, a plant pathogenic Mollicute that causes retrograde mertamorphosis manifested as virescence and phyllody where floral parts are converted to leafy structures resulting in nearly complete loss of fruit and seed formation. Sesame phyllody has been recorded in India, Iran, Iraq, Israel, Burma, Sudan, Nigeria, Tanzania, Pakistan, Ethiopia, Thailand, Turkey, Uganda, Upper Volta and Mexico (Akhtar et al. 2009). In fact, the incidences of phytoplasma associated diseases in sesame and several other crops are increasing in severity and spread mainly attributed to climate change and global warming (Galetto et al. 2011).

Phytoplasmas are insect-transmitted and cause considerable damage to a diverse economically important plants. They release virulence effectors into plants and insects to target host molecules that get unloaded from phloem to access distal tissues and induce changes in the development and defense response of their plant hosts thereby providing fitness advantages to phytoplasma. Effectors released into phloem sieve cells of the host plants cause systemic manipulation at both physiological and morphological levels. Phyllody, witches’ brooms, virescence, bolting, formation of bunchy fibrous secondary roots, sepal hypertrophy etc. are some of the morphological changes in the host plants which in turn evoke insect vector attraction (Bertaccini 2007).

Studies on the role of effectors and their mode of action inside host plants have been gaining importance. These effectors are reported to interact with plant components in the sieve cells and they can also upload from the phloem to interact with target molecules in companion, mesophyll and other plant cells (Sugio et al. 2011a). Phytoplasmas, being intracellular, secrete these effectors via the sec-dependent pathway in which the signal peptide is cleaved off (Kakizawa et al. 2004). Effector molecules and their interacting partners associated with disease development have been identified (MacLean et al. 2014; Sugio et al. 2014; Janik et al. 2017; Kitazawa et al. 2017;Wang et al. 2018; Chang et al. 2018).

One of the candidate effector proteins, SAP 11 (c.10 kDa), has a bipartite NLS sequence of 18 amino acids, which promotes its localization to the plant cell nuclei, in turn affecting expression of various genes, including transcription factors (Bai et al. 2009). SAP11 binds and destabilizes class II CIN (cincinnata) TCP (teosinte-branched, cycloidea, proliferation factor 1 AND 2) (Sugio et al. 2011b). Its linear modular structure facilitates its involvement in three distinct activities i.e. nuclear localization, TCP binding and TCP destaibilization (Sugio et al. 2014). Another efffector molecule, TENGU, a small protein (4.5 kDa), first identified from onion yellows phytoplasma (OY) has been shown to induce dwarfism and witches’ broom symptoms. Down-regulation of auxin-responsive genes was also observed in the tengu-transgenic plants, thus substantiating the role of TENGU in inhibition of auxin related pathways (Hoshi et al. 2009). TENGU has also been reported to be the first virulence factor affecting plant reproduction by perturbation of phytohormone signaling (Minato et al. 2014).

The genome of Aster Yellows phytoplasma strain Witches’ Broom (AY-WB) belonging to ‘Candidatus Phytoplasma asteris’ (subgroup 16SrI-A) was mined for the presence of genes encoding secreted proteins based on the presence of N-terminal signal peptides (SP). A total of 56 secreted AY-WB proteins (SAPs) were deduced from the full genome sequence of AY-WB (Bai et al. 2009). Of all the 56 candidate effector genes bioinformatically identified from AY-WB genome sequence, a single effector, SAP54 induced severely altered flower morphology when expressed in model system Arabidopsis (MacLean et al. 2011). It results in abnormal leaf like flowers, typical of symptoms of phyllody and virescence that are characteristic of AY-WB infection.

A SAP54 homolog, sharing 88% amino acid identity to SAP54, was identified in the genome of Onion Yellows phytoplasma strain OY-W (subgroup 16SrI-B) by Maejima et al. (2014). The gene encoding this OY-W effector was designated as PHYL1. The PHYL1 gene encodes a 125 amino acid protein with a 34 amino acid signal peptide at its N-terminus which was predicted to be secreted from phytoplasma cells as a 91 amino acid mature 10.6 kDa protein. Kitazawa et al. (2017) reported that the two phyllogens PHYL1_OY (Maejima et al. 2014) and PHYL1_PnWB (from peanut witches’ broom (PnWB) phytoplasma) (Chung et al. 2013), when expressed in petunia and Nicotiana benthamiana affected floral phenotypes in both the plants. Expression of PHYL1_PnWB in sesame showed phyllody phenotype in every floral organ (Kitazawa et al. 2017).

An ortholog of AY-WB SAP54 from BellVir phytoplasma (16SrIII-J), causing phyllody and virescence symptoms in Bellis perennis was characterized by Fernández et al. (2019). The Bell Vir SAP54 ortholog showed 51% identity with the AY-WB SAP54 (Fernández et al. 2019).

These effectors can serve as potential tools for providing insights into the mechanisms underlying symptom development and can be effective targets for devising disease control measures. However, studies on crop plants such as sesame, an important oilseed crop facing huge yield losses due to phytoplasma associated diseases, are completely lacking. There is a need to investigate the universality of the effector molecules identified in case of the model systems for their presence and efficacy in other plants. Therefore, the present study was carried out to detect and characterize the orthologs of SAP54 in phytoplasma associated with sesame plants showing phyllody and to compare it with its orthologs from other phytoplasma groups and subgroups. In addition, a comparative analysis of these orthologs aimed at studying the evolutionary pressures acting upon these sequences was also carried out.

Materials and methods

Source of phytoplasma

Leaf samples from symptomatic (phyllody affected) sesame plants (Sesame-U.P.) were collected from fields in Baghpat, Uttar Pradesh, India. The collected sesame belonged to Sekhar variety. Samples were also collected from asymptomatic sesame plants (Fig. 1a, b). The samples were collected in triplicates in liquid nitrogen and stored frozen till further use. Phytoplasma infected Catharanthus roseus sample was used as a positive control.

Fig. 1.

Sesamum indicum.a Asymptomatic twig; b symptomatic twig showing phyllody symptom

Genomic DNA extraction and phytoplasma detection using PCR

DNA was extracted from three symptomatic and three asymptomatic sesame leaf samples using CTAB (Cetylmethylammonim bromide) method (Saghai-Maroo et al. 1984). To ensure extraction of phytoplasma enriched genomic DNA, only petioles, midribs and prominent veins of leaf samples were used. Qualitative and quantitative analysis of genomic DNA was done by checking on 1.8% agarose gel and quantifying using QuantiT dsDNA BR assay kit using Qubit fluorometer (Invitrogen, Carlsbad, CA, USA). For the detection of phytoplasma, PCR amplification of the genomic DNA was done using phytoplasma-specific universal 16S rDNA primers P1/P7 (Deng and Hiruki, 1991), followed by nested PCR with primer pair R16F2/R2n (Gundersen and Lee, 1996). The PCR conditions employed while amplification with primers P1/P7 were: an initial denaturation at 94 °C for 3 min, 35 cycles of 94 °C for 1 min, 55 °C for 1 min, 72 °C for 2 min and final extension at 72 °C for 10 min. All PCR conditions were same except a higher annealing temperature of 60 °C while amplication using primer pair R16F2/R2n. The amplified product was electrophoresed on 1.2% agarose gel.

Gel extraction, sequencing and in-silico RFLP

The nested-PCR product was gel extracted and purified using QIAquick gel extraction kit (QIAGEN, Chatsworth, CA, USA). The product of the 16SrDNA was sequenced (2X sequencing coverage) directly using both forward and reverse primers. A total of three samples were given for sequencing on ABI platform. The consensus sequence was analyzed for homology using NCBI BLASTn. In-silico analysis of the obtained 16S rDNA sequence was done using pDRAW 32 1.1.133 DNA analysis software by AcaClone software (http://www.acaclone.com/). Seventeen REs (AluI, BamHI, BfaI, BstUI, DraI, EcoRI, HaeIII, HhaI, HinfI, HpaI, HpaII, KpnI, MboI, MseI, RsaI, SspI and TaqI) were selected for in silico digestion which helped in subgroup classification of the detected phytoplasma.

Effector gene amplification, cloning and sequencing

Primers (S54LP of SP_F 5′-C ATG GAG GCC GAATTC ATG TTT AAA ATC AAA AAT AAT TTA-3′ and S54LP of SP_R 5′-GC AGGTCGACGGATCC TTA TTT TCA TCA TTT AAA GTT TTT-3′) were designed from PHYL1 (GenBank: AB812838.1), a homolog of SAP54, identified from Onion Yellows (OY) (16SrI-B) phytoplasma (Maejima et al. 2014). The primers were designed manually such that they contain a 24 bp homology to the PHYL1 sequence, and a 16 bp homology to the linear ends of pGBKT7 vector. Symptomatic sesame plants in which the presence of phytoplasma had been confirmed were used for the amplification of SAP54 ortholog. A total of three symptomatic samples were used for amplification. The PCR product was gel extracted and purified using QIAquick gel extraction kit (QIAGEN, Chatsworth, CA, USA) and Nucleospin gel and PCR clean-up kit (Clontech) respectively. The pGBKT7 vector was linearized by digesting to completion with BamHI and EcoRI enzymes from NEB and purified using the Nucleospin gel and PCR clean-up kit (Clontech). The purified effector gene was inserted in the linearized pGBKT7 vector by in-fusion cloning using In-Fusion Cloning Kit (Clontech) following the In-Fusion Cloning Procedure for Spin-Column Purified PCR Fragments. The effector gene with additional overhangs and linear pGBKT7 were mixed and fused using the In-fusion enzyme. Competent Escherichia coli strain DH5-α cells were transformed with the cloning reaction mixture and plated on LB plates containing kanamycin at a final concentration of 50 µg/ml. Miniprep was done using QIAprep Spin Miniprep Kit (Qiagen) and plasmids were sequenced (3X coverage) on ABI platform using vector specific forward primer (pGBKT7_F 5′-TCA TCG GAA GAG AGT AGT AAC-3′) and gene specific reverse primer (S54LP of SP_R).

Sequence analysis

BLAST analysis was done to find sequence/s with maximum homology to SAP54 ortholog from phyllody affected sesame. All sequences obtained from BLAST result were aligned using CLUSTAL W and a phylogenetic tree was constructed using the maximum likelihood model of MEGA7.0.26 (Kumar et al. 2016). Bootstrapping was performed 1000 times. Sequence analysis and comparisons of the effector gene from sesame and PHYLI were performed at nucleotide and amino acid levels to infer synonymous and nonsynonymous substitutions between two effector proteins encoded in the same subgroup of phytoplasma. SignalP 5.0 was used to confirm the cleavage site of SAP54 ortholog (Armenteros et al. 2019).

Test for selection

Estimation of non-synonymous substitution rates (Ka) and synonymous substitution rates (Ks) across all amino acid sites of twenty SAP54 ortholog sequences from twenty diverse phytoplasma strains belonging to nine different (sub)groups affecting plant species belonging to 12 different families across ten countries (Table 1) was done using DNA Sequence Polymorphism v 6.10.04 software (Rozas et al. 2017). Ka, Ks and Ka/Ks values of full length protein as well as mature peptide region and signal peptide region separately were estimated in pairwise comparisons between nucleotide sequences. Average Ka, Ks and Ka/Ks values of total 190 pairwise sequence comparisons were calculated. Site-specific selection (positive or negative) across each amino acid site was inferred using Fixed Effects Likelihood (FEL) model of DataMonkey, a web-based suite of phylogenetic analysis tools for use in evolutionary biology (Delport et al. 2010).

Table 1.

SAP54 orthologs from phytoplasma strains belonging to nine different (sub)groups affecting plants belonging to twelve different families across ten countries

S. no.	CandidatusPhytoplasma species	16Sr(sub)group	Phytoplasma strain	Place of isolation	GenBank accession
1.	Ca. Phytoplasma asteris	I-B	Sesame phyllody (S54LP of SP)	Delhi, India	MK858224
2.	Ca. Phytoplasma asteris	I-B	Gladiolus witches’-broom (GLAWC)	France	AB862483.1
3.	Ca. Phytoplasma asteris	I-B	Maryland aster yellows (AY1)	Maryland, USA	DQ837760.1
4.	Ca. Phytoplasma asteris	I-L	Aster yellows (AY2192)	Germany	AB862480.1
5.	Ca. Phytoplasma pruni	III-E	Spiraea stunt (SP1)	New York, USA	EF200539.1
6.	Ca. Phytoplasma phoenicium (PPWB)	IX-A	Picris echioides yellows (PEY)	Italy	AB862490.1
7.	Ca. Phytoplasma trifoli	VI-A	Catharanthus phyllody (CPS)	Sudan	AB897827.1
8.	Ca. Phytoplasma pruni	III-A	Clover proliferation (CP)	Italy	AB862489.1
9.	Ca. Phytoplasma asteris	I-C	Leontodon yellows (LEO)	Italy	AB862484.1
10.	Ca. Phytoplasma asteris	I-C	Carrot yellows (CA-76)	Italy	AB862481.1
11..	Ca. Phytoplasma asteris	I-M	Atypic aster yellows (AVUT)	Germany	AB862478.1
12.	Ca. Phytoplasma asteris	I-F	Aster yellows from apricot (A-AY)	Spain	AB862477.1
13.	Ca. Phytoplasma asteris	I-B	Onion yellows (OY-W)	Japan	AB812838.1
14.	Ca. Phytoplasma asteris	I-B	Tomato yellows (TY)	Japan	EF200537.1
15.	Ca. Phytoplasma asteris	I-F	Apricot chlorotic leafroll AY-A(ACLR)	Spain	AB862479.1
16.	Ca. Phytoplasma asteris	I-B	Eggplant dwarf (ED)	Japan	AB862482.1
17.	Ca. Phytoplasma asteris	I-B	Gladiolus witches’-broom (GLAWB)	The Netherlands	AB897828.1
18.	Ca. Phytoplasma asteris	I-B	Primula green (PrG)	UK	AB862485.1
19.	Ca. Phytoplasma asteris	I-B	Oilseed rape virescence (RV)	France	AB862487.1
20.	Ca. Phytoplasma asteris	I-B	Severe western aster yellows (SAY)	USA	AB862488.1

Results

Detection and group and subgroup classification of sesame phytoplasma strain

The symptomatic sample (Sesame-U.P.) was detected positive for the presence of phytoplasma. Nested PCR using phytoplasma 16S rDNA specific primers P1/P7 and R16F2/R2 yielded amplicon size of approximately 1.2 kb (Fig. 2a). An amplicon of same size was obtained in case of phytoplasma infected Catharanthus roseus sample (positive control). No amplification was obtained in the negative control (water) and in the asymptomatic sesame sample collected from the same location. The 16S rDNA sequence of the Sesame-U.P. phytoplasma strain showed 99% identity with strains in 16SrI-B subgroup (Onion yellows phytoplasma, GenBank Accesion Number AP006628 and Sesame phyllody phytoplasma, GenBank Accession Number KF728959). Virtual RFLP profile of the amplified 16SrDNA using seventeen REs generated using pDRAW matched exactly with that of subgroup 16SrI-B strains reported earlier (Lee et al. 1998) (Fig. 2b).

Fig. 2.

R16F2/R16R2 gene profile. a agarose gel (1.2%) showing amplification product (~ 1.2 kb) of nested PCR using phytoplasma-specific 16S rDNA primer pairs P1/P7 and R16F2/R2. Lanes: 1(asymptomatic sesame), 2, 3, 4, 5 (phyllody affected sesame) and 6 (water). M-100 bp DNA ladder; b virtual RFLP profile obtained from in silico digestions of R16F2/R16R2 gene fragment from phytoplasma strain producing phyllody symptoms in sesame (Sesamum indicum) in Baghpat, U.P. Following 17 restriction enzymes were used in the simulated digestions: AluI, BamHI, BfaI, BstUI, DraI, EcoRI, HaeIII, HhaI, HinfI, HpaI, HpaII, KpnI, MboI, MseI, RsaI, SspI and TaqI. MW-100 bp Plus DNA ladder, Fermentas

Effector amplification, cloning and miniprep

Amplification carried out using primers designed from PHYL1 (GenBank: AB812838.1), a homolog of SAP54, identified from OY (16SrI-B) phytoplasma (Maejima et al. 2014) resulted in an amplicon of expected size, i.e. 410 bp in DNA samples from symptomatic sesame plant in which the presence of phytoplasma had already been confirmed. No amplification was observed in the asymptomatic sample and water (Fig. 3a). Transformed E. coli colonies were observed on LB plates containing kanamycin selection. Colony PCR confirmed the amplicon fusion into pGBKT7 vector and its successful transformation into E. coli. (Fig. 3b). Miniprep followed by plasmid PCR with vector specific forward primer and gene specific reverse primer reconfirmed effector gene fusion with the double digested vector (Fig. 3c).

Fig. 3.

Effector gene gel electrophoresis. a Agarose gel (1.2%) showing ~ 410 bp band of S54LP of SP from phyllody affected sesame obtained by PCR with PHYL1 specific primers (S54LP of SP_F and S54LP of SP_R) designed with additional overhangs for infusion ligation in pGBKT7 vector. Lanes: 1(water), 2 (phyllody affected sesame) and 3 (asymptomatic sample). M- 100 bp DNA ladder (Fisher Scientific); b Agarose gel (1.2%) showing products of colony PCR of transformed E. coli (DH5α) confirming ligation of the S54LP of SP in pGBKT7 vector; c agarose gel (1.2%) showing product of plasmid (isolated from E. coli DH5-α) PCR using vector specific forward primer (pGBKT7_F) and gene specific reverse primer (S54LP of SP_R). M-100 bp DNA ladder (Fisher Scientific)

Sequence analysis of sesame phytoplasma effector

Homology search using BLASTn showed 15 sequences displaying similarity index ranging from 93 to 99% (Table 2). The amplicon obtained from phyllody affected sesame (OY subgroup) from India (U.P.) in this study exhibited maximum (99%) sequence identity with that of SAP54 orthologs identified from Gladiolus witches broom (GLAWC) from France and Maryland aster yellows (AY1), both of which belong to species Ca. P. asteris, Onion Yellows (16SrI-B) subgroup (GenBank: AB862483.1 and DQ837760.1 respectively). Therefore, the effector gene obtained was denoted as S54LP of SP (SAP54 Like Protein of Sesame Phyllody) gene hereafter. The nucleotide sequence of S54LP of SP has been submitted to GenBank as S54LP of SP_Sesame-U.P. and it has been assigned an accession number MK858224. Strikingly, S54LP of SP gene shares same level of identity (99%) with SAP54 orthologs from phytoplasma strains belonging to16SrI-L and 16SrIII-E groups but only 97% nt sequence identity with PHYL1(GenBank: AB812838.1), a SAP54 ortholog encoded in the OY-W genome (16SrI-B subgroup) from which the primers were designed.

Table 2.

BLASTn result of effector gene (S54LP of SP) from phyllody affected sesame displaying similarity ranging from 93 to 99% with other SAP54orthologs

S. no.	Phytoplasma strain	16Sr Group	Accession no.	Identity (%)	E-value	Query cover (%)
1.	Gladiolus witches’-broom phytoplasma phyl1 gene for phytoplasmal effector causing phyllody symptoms 1, complete cds	IB	AB862483.1	99	0.0	100
2.	Maryland aster yellows phytoplasma strain AY1 hypothetical protein genes, partial and complete cds	IB	DQ837760.1	99	0.0	100
3.	Aster yellows Phytoplasma AY2192 phyl1 gene for phytoplasmal effector causing phyllody symptoms 1, complete cds	IL	AB862480.1	99	0.0	100
4.	Spiraea stunt phytoplasma variable mosaic gene cluster, partial sequence	IIIE	EF200539.1	99	0.0	100
5.	Candidatus Phytoplasma phoenicium phyl1 gene for phytoplasmal effector causing phyllody symptoms 1, complete cds	IXA	AB862490.1	98	0.0	100
6.	Candidatus Phytoplasma trifolii phyl1 gene for phytoplasmal effector causing phyllody symptoms 1, complete cds	VIA	AB897827.1	98	0.0	100
7.	Candidatus Phytoplasma pruni phyl1 gene for phytoplasmal effector causing phyllody symptoms 1, complete cds	IIIA	AB862489.1	98	0.0	100
8.	Leontodon yellows phytoplasma LEO phyl1 gene for phytoplasmal effector causing phyllody symptoms 1, complete cds	IC	AB862484.1	98	0.0	100
9.	Carrot yellows Phytoplasma CA-76 phyl1 gene for phytoplasmal effector causing phyllody symptoms 1, complete cds	IC	AB862481.1	98	0.0	100
10.	Peach yellows phytoplasma PYR phyl1 gene for phytoplasmal effector causing phyllody symptoms 1, complete cds	IB	AB862486.1	98	1e−180	100
11.	Apricot atypic aster yellows Phytoplasma AVUT phyl1 gene for phytoplasmal effector causing phyllody symptoms 1, complete cds	IM	AB862478.1	98	5e−180	100
12.	Apricot aster yellows Phytoplasma A-AY phyl1 gene for phytoplasmaleffector causing phyllody symptoms 1, complete cds	IF	AB862477.1	98	5e−180	100
13.	Onion yellows phytoplasma OY-W PHYL1 gene for phytoplasmal effector causing phyllody 1, complete cds	IB	AB812838.1	97	1e−174	100
14.	Tomato yellows phytoplasma variable mosaic gene cluster, partial sequence	IB	EF200537.1	97	1e−174	100
15.	Aster yellows witches’-broom phytoplasma AYWB, complete genome	IA	CP000061.1	93	1e−156	100

Phylogenetic analysis

Phylogenetic interrelationship of S54LP of SP with 15 orthologs obtained from BLASTn using maximum likelihood method clustered S54LP of SP(16SrI-B) with SAP54 orthologs from diverse groups i.e., 16SrI-B, 16SrI-L and 16SrIII-E groups with a high bootstrap support value of 75 (Fig. 4). An ortholog belonging to 16SrIX-A group also grouped with S54LP of SP, though with a lower bootstrap value of 63. It is noteworthy that orthologs from OY-W and TY (Tomato yellows phytoplasma), both belonging to 16SrI-B, were grouped in different clusters.

Fig. 4.

Phylogenetic tree constructed using Maximum likelihood analyses of S54LP of SP orthologs obtained from BLASTn result. Bootstrap analyses were done in 1000 replicates (Tamura-Nei model). Phylogenetic analyses were performed using MEGA 7.0.26 software

Comparative sequence analysis of PHYL1 and S54LP of SP

PHYL1 and S54LP of SP both retrieved from phytoplasma belonging to 16Sr-I-B consists of 125 amino acids which are secreted into the host cells as 91 amino acids mature proteins, the signal peptide of 34 amino acids being cleaved off before secretion. The cleavage site of S54LP of SP was searched using SignalP 5.0. Comparison of full length (378 bps) ORFs of PHYL1 and S54LP of SP genes showed a total of thirteen SNPs (Fig. 5a). Out of total thirteen SNPs, ten (76.92%) were non-synonymous and three (23.08%) were synonymous. It was further observed that 40% of the non- synonymous SNPs were present in the signal peptide (SP) and sixty percent in the mature (secreted portion) protein (Fig. 5b). Thus, 40% of total amino acid changes occur in the signal peptide and 60% in the mature protein. The properties of all the amino acids, except one, also changed either from hydrophobic to hydrophilic and vice versa or from acidic to basic and vice versa (Table 3).

Fig. 5.

Clustal W alignment of aPHYL1 and S54LP of SP nucleotide sequences showing thirteen SNPs; b PHYL1 and S54LP of SP amino acid sequences showing ten nonsynonymous substitutions

Table 3.

Single nucleotide polymorphisms, synonymous and nonsynonymous substitutions and changes in properties of amino acids in SAP54 Like Protein of Sesame Phyllody (S54LP of SP) from phyllody affected sesame

S. no.	SNPs	Position	Change	Codon change	Amino acid change	Position of amino acid change	Synonymous/non synonymous substitution	Change in property of amino acid
1.	A	36	G - A	TTG → TTA	Lys → Lys	–	Synonymous	–
2.	G	40	T- G	TCG → GCG	Ser → Ala	S14A	Non synonymous	Hydrophilic–hydrophobic
3.	A	55	G-A	GCT → ACT	Ala → Thr	A19T	Non synonymous	Hydrophobic–hydrophilic
4.	A	60	C-A	TTC → TTA	Phe → Leu	F20L	Non synonymous	Hydrophobic–hydrophobic
5.	A	79	G-A	GCT → ACT	Ala → Thr	A27T	Non synonymous	Hydrophobic–hydrophilic
6.	G	199	A-G	AAA → GAA	Lys → Glu	K67E	Non synonymous	Basic–acidic
7.	C	231	A-C	AGA → AGC	Arg → Ser	R71S	Non synonymous	Basic–Neutral
8.	C	252	T-C	AAT → AAC	Asp → Asp	–	Synonymous	–
9.	A	264	C-A	AAC → AAA	Asp → Lys	N88K	Non synonymous	Acidic-Basic
10.	C	282	T-C	GCT → GCC	Ala → Ala	–	Synonymous	–
11.	A	310	C-A	CAT → AAT	His → Asp	H104N	Non synonymous	Basic-Acidic
12.	A	322	G-A	GAA → AAA	Glu → Lys	E108K	Non synonymous	Acidic-Basic
13.	C	344	A-C	AAA → ACA	Lys → Thr	K115T	Non synonymous	Basic-Neutral

Comparison of GLAWC and S54LP of SP nucleotide sequences showed a single synonymous SNP that was present in the signal peptide. Thus, the mature protein sequences of GLAWC (IB) and S54LP of SP (IB) were 100% identical.

Selection analysis

Polymorphism exists in SAP54 orthologs from diverse strains

Multiple sequence alignments of twenty amino acid sequences using MUSCLE (MEGA7.0.26) showed seventeen polymorphic sites, seven in the signal peptide region and ten in the mature protein (Fig. 6a). MEGA 7.0.26 identified a total of thirteen out of seventeen variable sites to be parsimony informative sites. The remaining four sites are singletons. Amino acid substitutions involved different chemical classes of amino acids, such as hydrophobic amino acids, including phenylalanine(F), leucine (L), alanine (A), proline (P) and isoleucine (I) and polar amino acids, including, asparagine(N), serine(S), tyrosine(Y), threonine(T), histidine (H), lysine (K) and arginine (R). Site specific amino acid conservation and relative frequency of each amino acid has been depicted in the weblogo (Fig. 6b). Analysis of amino acid composition shows the orthologs to be particularly enriched (19.7%) in asparagine (N) residues followed by leucine (16%) and lysine (10.9%) (Online Resource 1_Fig. S1). Cysteine (C) and tryptophan (W) residues are totally lacking in the effector molecules.

Fig. 6.

Polymorphism across SAP54 orthologs. a Multiple sequence alignment of amino acid sequences of 20 SAP54 orthologs from phytoplasmas belonging to nine different (sub)groups affecting plant species belonging to twelve different families across 10 countries. Amino acids which are identical across all strains are indicated by dots. b Weblogo indicating the degree of conservation and the relative frequency of each amino acid at a particular site (http://weblogo.threeplusone.com/create.cgi)

SAP54 orthologs are under purifying selection

Ka and Ks was calculated across the entire ORF sequences of SAP54 orthologs. Pairwise sequence comparisons of twenty full length nucleotide sequences of the orthologs showed that only 16 out of 190 pairs have Ka values greater than Ks (i.e. ω = Ka/Ks > 1). This shows that purifying/stabilizing selection acts on SAP54 orthologs from diverse phytoplasma strains. The silent mutation rate (Ks) of SAP54 orthologs from diverse phytoplasma groups was found to range from 0 substitution/site to 0.0731. The non-silent rate (Ka) was slightly lower, ranging between 0 substitution/site to 0.0403. The Ka/Ks fluctuated in the range of 0 to 0.0766 to 4.485714. Out of total 190 pairwise sequence comparisons, only 16 have Ka/Ks values greater than 1(diversifying or positive selection). A total of 72 sequence comparisons have Ka/Ks values in the range 0–0.5 (strong purifying selection) while 74 have Ka/Ks values > 0.5 and ≤ 1 (weak purifying selection).

A pair-wise comparison of Ka/Ks varies from 0 to 0.2288 to 4.4857 across 190 combinations for all the 20 SAP54 orthologs with the highest value (4.4857) in case of two pairs viz. CP (16SrVIA) versus AVUT (16SrIM) and CP (16SrVIA) versus PrG (16SrIB). A detailed analysis of intragroup Ka/Ks values between S54LP of SP (16SrIB) from sesame and its ten orthologs from phytoplasmas belonging to the same group, i.e., 16SrIB but infecting different host plants showed that half of the pairs have Ka/Ks values of 0 and the rest have values ranging from 0.2357 (strong purifying selection) to 0.7767 (relaxed purifying selection). This observation is indicative of the fact that evolution of SAP54 orthologs is independent of the phytoplasma phylogeny based on 16SrDNA.

Comparative analysis of selection acting on the signal peptide and the mature protein regions

Deviation in amino acid substitution pattern across different regions of the SAP54 orthologs was analysed by calculating Ka and Ks of the signal peptide and the mature protein separately. Pairwise sequence comparisons of the signal peptide region of 20 SAP54 orthologs showed only 8 out of total 190 pairs to be under diversifying selection. Similarly, only 3 out of 190 pairs showed ω > 1 in case of mature protein. The average Ka/Ks value of all pairwise sequence comparisons of signal peptide region is higher (0.5950) than that of the mature protein (0.5306). This result suggests that signal peptides might be under relaxed purifying selection as compared to mature proteins.

Comparative analyses of signal peptide and mature protein showed relatively elevated values for different selection parameters in case of signal peptide. Out of total 190 pairwise sequence comparisons, 147 (77.37%) have Ka values greater in case of signal peptide as compared to that of mature protein (Online Resource 1_Fig. S2a). It was further observed that 139 (73.16%) out of total 190 pairwise comparisons also have greater Ks values in case of signal peptides (Online Resource 1_Fig. S2b). A total of 121 out of 190 pairs have Ka/Ks values higher in signal peptide (Online Resource 1_Fig. S2c). This trend further elucidates the fact that signal peptides are under relaxed purifying selection in contrast to mature protein which is under stringent purifying selection.

SAP54 amino acid sites are under pervasive purifying selection

Site specific selection (positive or negative) estimated using fixed effects likelihood (FEL) approach using a p value threshold of 0.1 showed the presence of pervasive positive/diversifying selection at one site (in the signal peptide) and negative/purifying selection at three sites.

A total of 12 sites were found to have beta values (non-synonymous substitution rate at a site, dN) higher than the rate estimate under the neutral model (Online Resource 1_Fig. S3a). Out of these 12, the fourth site has the highest dN value of 36.268 which is statistically significant too. Thus the fourth site was considered to be under pervasive positive selection. Similarly, 7 sites were estimated to have alpha values (synonymous substitution rate at a site, dS) higher than the rate estimate under the neutral model (Online Resource 1_Fig. S3b). A total of 3 out of total 7 sites have statistically significant alpha values and thus were considered to be under pervasive negative selection.

The fourth site (I) is positively selected with a likelihood-ratio test (LRT) of 2.738. The fifth (K), sixty-eighth (N) and eighty-fourth (N) sites are negatively selected with LRT values 4.521, 8.123 and 3.065 respectively. The log L values under the Nucleotide GTR model and Global MG94xREV models are − 667.12 and − 643.24 respectively (Table 4). log likelihood scores are strongly negative for both the models, which means both the models fit the data well. dN/dS ratio for background is 0.286 and dN/dS ratio for test is 0.460. This indicates that the SAP54 orthologs are under purifying selection.

Table 4.

SAP54 sites under pervasive positive or purifying selection as detected by Fixed Effects Likelihood (FEL) method

Position	Amino acid	Selection	Synonymous substitution rate at a site (alpha)	Non-synonymous substitution rate at a site (beta)	The rate estimate under the neutral model (alpha = beta)	LRT (Likelihood ratio test)	p value
4	I	Pervasive Positive/Diversifying	6.718	36.268	15.737	2.739	0.098
5	K	Pervasive Negative/Purifying	52.048	0.000	4.244	4.521	0.033
68	N	Pervasive Negative/Purifying	60.296	0.000	8.000	8.125	0.004
84	N	Pervasive Negative/Purifying	12.796	0.000	2.879	3.066	0.080

Discussion

In the present study, phytoplasma associated with sesame plants showing phyllody and witches’ broom symptoms collected from Baghpat, Uttar Pradesh, India was found to be belonging to 16Sr-I B group on the basis of sequence similarity as well as virtual RFLP profiling. However, S54LP of SP, an ortholog of SAP54 in phyllody affected sesame plants, obtained from these symptomatic plants by using primers from PHYL1 (its ortholog encoded in OY-W genome belonging to 16SrI-B) showed only 97% sequence identity. Interestingly, higher (99%) sequence similarity was observed with the SAP54 orthologs from Phytoplasma strains belonging to 16SrI-L and 16SrIII-E groups. Thus, it can be inferred that SAP54 evolution is independent of the species-level phylogeny based on 16Sr DNA. It has been observed that phytoplasmas from phylogenetically distant groups but having the same host plant species can have highly similar SAP54 because of horizontal gene transfer and the evolution of SAP54 is also predicted to be dependent on the MTF (MADS Domain Transcription Factors) repertoire of its host plant range (Orlovskis 2017) as diverse16Sr (sub-) groups of phytoplasmas have been reported to infect same or closely related plant species (Lee et al. 2004).

The effector gene S54LP of SP was sequenced with 3X coverage. Although, the presence of the start and stop codons of S54LP of SP were not determined in this study, significantly high query coverage and percentage identity with other orthologs of SAP54 and low E-values of S54LP of SP when BLASTn was performed suggests S54LP of SLP to be a functional ortholog of SAP54. No insertion or deletion was observed in S54LP of SP sequence. Single nucleotide polymorphism (SNPs) observed were also not significant to make it a non-functional or pseudogene. Moreover, none of the SAP54 orthologs identified so far have been designated as pseudogenes. All the above reasons suggest S54LP of SLP to be a functional ortholog of SAP54 and not a pseudogene.

Detailed comparative sequence analysis of PHYL1 and S54LP of SP showed presence of 13 SNPs wherein nonsynonymous substitutions were three fold higher than the synonymous ones. These nonsynonymous mutations were present in the signal peptide as well as in the region corresponding to the mature protein. As deduced from the nature of amino acids, the signal peptide of S54LP of SP contains the typical three-partite structure but the preferential AxA residues that have been considered essential for the activity of Signal peptidase I (Auclair et al. 2012) were absent. Instead, a VxG motif was present which was conserved across all the orthologs except AY2192 strain (IL), in which AxG motif was present. It is noteworthy that the canonical AxA sequence was not present in any of the 20 S54LP of SPorthologs analysed in the present study at the cleavage site. Therefore, analysis of the S54LP of SPamino acid sequences revealed that the site of cleavage of signal peptide i.e., after 34th amino acid is non- canonical. In fact, Payne et al. (2012) have reported that contrary to the earlier assumption, alanine preference at the cleavage site is common but not universal on the basis of proteome data from 32 bacterial and archaeal organisms from nine phyla. Further computational analysis of ~ 1500 genomes from major evolutionary clades also revealed the replacement of canonical signal peptide with other motifs.

Interestingly, an AXA motif was present before the 43^rd site of S54LP of SP, but it was not indicated to be the cleavage site for obtaining the mature protein. It is noteworthy that Maejima et al. (2014), in their truncation experiment, have demonstrated an essential role of the 42nd amino acid residue in induction of phyllody.

Identifying regions under diversifying or purifying selection is helpful in recognizing domains critical for effector gene’s virulence and its interaction with the host machinery (Rohmer et al. 2004). Ka/Ks ratio gives an estimate of the level of selective constraint (i.e., the strength of purifying selection) acting on a particular protein. Phytoplasma S54LP of SP gene appears to be evolutionarily conserved (average Ka/Ks for full length, ω = 0.617), indicating the orthologs to be under purifying (negative) selection. Similar reports of bacterial effectors/virulent factors under purifying selection have been found in case of non -plant pathogenic bacteria. The streptokinase (ska) gene of Stretococcus pyogenes, a human pathogenic bacteria, has evolved under purifying selection with an average ω ratio of 0.449 (Kalia and Bessen 2004). However, specific codons (∼ 4%) corresponding to the beta domain are under strong positive selection (ratio, > 5). During infection, it is the beta domain of streptokinase that makes direct molecular interactions with mammalian host plasminogen.

However, in case of phytoplasma S54LP of SP gene, 2.4% of total codons are under pervasive purifying/negative selection and only 0.8% (codons corresponding to the signal peptide region) are under pervasive positive selection. Rümpler et al. (2015) proposed SAP54 to be a coiled-coil protein possessing a double-helical structure separated by a short interhelical (kink) region. The codons under purifying selection mainly correspond to the coiled -coil helix and the kink region of SAP54 gene. The protein interaction domain of MIKC-TYPE MADS domain TFs, i.e., the K domain is a highly conserved region and might be the reason for stringent purifying selection acting on phytoplasma S54LP of SP ortholgs as a whole and particularly the coiled-coil domain that has been proposed as the interacting domain by Rümpler et al. (2015). The crystal structure of PHYL1_OY, an effector protein from ‘Candidatus Phytoplasma asteris’ strain onion yellows (OY) has recently been elucidated by Iwabuchi et al. (2019). They revealed the monomeric nature of PHYL1_OY. The structure comprised of two alpha-helices interconnected with a loop in a coiled-coil fashion. Distributions of hydrophobic amino-acid residues are highly conserved among phyllogens and the hydrophobic amino acid residues are oriented towards the interior of the protein. Structural remodeling revealed that SAP54 and PHYL1_PnWB homologs of PHYL1_OY share similar structures. The crystal structure of PHYL1_PnWB, a phyllogen from Peanut witches’ broom (PnWB) phytoplasma was first proposed by Liao et al. (2019) using X-ray crystallography. They have reported PHYL1_PnWB to fold into an α-helical hairpin like structure, the two helices being connected by a loop. It was proposed that the amino acids L60 and Y64 located on α-helix 2 are crucial for the degradation of SEP3. The hydrophobic amino acid residues are reported to be exposed on the surface of the protein plying vital biological roles.

In case of other plant pathogenic bacteria such as Pseudomonas syringe, type III secretome has been reported to have the largest repertoire of T3SEs with remarkable diversity (McCann and Guttman 2008). Evolutionary analyses of Type III effector/chaperone gene families of Pseudomonas syringae showed ten (77%) out of total thirteen ancient gene families to evolve under purifying selection, suggesting selective pressure to maintain presumed virulence function of these important genes. Also, only 36% of the total horizontally transferred T3SEs showed signs of being under positive selection (Rohmer et al. 2004).

In a recent study by Yamaguchi et al. (2019); evolutionarily conserved virulence factor, CbpJ, has been identified from Streptococcus pneumonia. Selection pressure analysis was done to quantify the severity of negative selection pressures on genes encoding the pneumococcal choline-binding proteins (CBPs). They found particularly strong selective constraints on the gene cbpJ, indicating its importance in bacterial virulence. Thus, comparative enrichment of codons under negative selection is indicative of the measure of virulence and success in case of bacteria. As domains under purifying/negative selection do not allow amino acid changes, they can be harnessed to promote development of resistance.

The signal peptide region of S54LP of SP was shown to be under relaxed purifying selection as compared to the mature protein. This observation might partly be responsible for the wide host range of this phytopathogenic bacteria. Similar reports of signal peptides evolving under relaxed purifying selection or at rates higher than their flanking mature proteins have been found in case of Frankia, Mycobacterium and Streptomyces (Thakur et al. 2013). A genome-wide analysis of secretory proteins in prokaryotes and eukaryotes by Li et al. (2009) also showed that signal peptides evolve faster than mature proteins. Also, an over representation of leucine residues in the S54LP of SP signal peptide is observed. A survey using COPASAAR showed leucine repeats to be the most frequent type of Single Amino Acid Repeats (SAARs) in a wide range of species covering all three kingdoms (Eukaryota, Archaea and Bacteria) (Depledge and Dalby 2005). Analysis of signal peptides of secreted and type 1 membrane proteins of over 100 eukaryotic species by Labaj et al. (2010) showed high frequency/overrepresentation of leucine repeats.

Further, all the four, out of total 125 amino acids, found to be under pervasive positive or negative selection as detected by Fixed Effects Likelihood (FEL) method are coded by AT-rich codons. Thus, AT-rich codons contribute more significantly to the evolutionary rate in case of S54LP of SP gene in phytoplasma, which interestingly has an AT-rich genome too. There exists a certain level of correlation between the composition of amino acid and the rate of evolution in bacteria (Du et al. 2018). Phytoplasma has an AT-rich genome and therefore, the evolution of S54LP of SP gene across diverse strains involves selection pressure acting on amino acids coded by only AT-rich codons. It has been shown that in case of AT-rich genomes, the usage of AT-rich amino acids increases i.e., mostly the AT-rich amino acids are negatively selected. Three out of four sites were found to be under pervasive purifying selection. These sites may be important to the basic structure of S54LP of SP and are thus highly conserved. Recently, Du et al. (2018) confirmed the correlation between amino acid composition and evolutionary rate Ka/Ks in bacteria revealing that GC content and richness of amino acids influence their contributions to evolutionary rates. There also exists a certain degree of correlation between the overall gene expression level and the fraction of sites under strong constraint (Marek and Tomala 2018). Lowly expressed genes tend to have fewer sites under strong constraint as compared to highly expressed genes. Though such studies in case of phytoplasma remain elusive, these observations would help in ascertaining the relative level of expression of SAP54 vis-a vis other genes.

A closer look at the individual amino acids constituting S54LP of SP further revealed that asparagine (at positions 68th and 84th) is a functionally important amino acid residue as it is under strong pervasive purifying selection. Interestingly, significant role of asparagine residues at specific sites has been established in stablising protein–protein interactions viz. antigen–antibody complexes by substituting them with other amino acids (Yokota et al. 2010).

Fraser et al. (2002) reported that proteins with more number of interacting partners have slower evolution rates because a major proportion of the protein is involved in protein functions. MacLean et al. (2014) have identified the interacting partners of SAP54 by performing a yeast two-hybrid screen against an Arabidopsis seedling library. SAP54 has been shown to interact with Type II MADS-domain transcription factors (MTFs) AGAMOUS-LIKE 12 (AGL12), MADS AFFECTING FLOWERING1 (MAF1), and SEPALLATA3 (SEP3). Yeast two-hybrid and co-immunoprecipitation experiments also confirmed the interaction of SAP54 with two isoforms of RAD23 namely, RAD23C and RAD23D, which act as shuttle proteins targeting ubiquitylated proteins to the proteasome for degradation. Thus, SAP54 orthologs appear to have a number of potential interacting partners, which further substantiates stringent purifying selection to act on the secreted part and its slow evolution rate. Since SAP54 orthologs interact with highly conserved MTFs and RAD23 proteins (MacLean et al. 2014), substitution of any kind in the effector molecule is likely to perturb the protein interaction and thus they are removed by selection (Hirsh and Fraser 2001). However, studies on SAP54 interacting partners and their mechanism of interaction in plant hosts are rather scanty so far. Further studies need to be done to elucidate this important interaction and the underlying mechanism.

S54LP of SP evolves under stringent purifying selection which elucidates its vital role in phytoplasma virulence. The slow evolution of the mature protein region of S54LP of SP is also indicative of the highly conserved nature of the host target. Asparagine residue has been suggested to play an essential role in S54LP of SP interaction with the host target.

Use of effector molecules is being considered as an important strategy for developing resistance against pathogens. Therefore, identification, structural, functional characterization and their evolutionary dynamics form an important area of research. ‘Effectromics’ holds significant promise not only in resistance breeding but understanding basic plant responses particularly in case of pathogens like phytoplasma where dramatic alterations of developmental processes is involved.

Electronic supplementary material

Below is the link to the electronic supplementary material.