Bioinformatic analysis and functional characterization of CFEM proteins in Setosphaeria turcica

WANG Jian-xia,LONG Feng,ZHU Hang,ZHANG Yan,WU Jian-ying,SHEN Shen,DONG Jin-gao,2,HAO Zhi-min

1 State Key Laboratory of North China Crop Improvement and Regulation/Key Laboratory of Hebei Province for Plant Physiology and Molecular Pathology,College of Life Sciences,Hebei Agricultural University,Baoding 071001,P.R.China

2 College of Plant Protection,Hebei Agricultural University,Baoding 071001,P.R.China

Abstract Common in Fungal Extracellular Membrane (CFEM) domains are uniquely found in fungal extracellular membrane proteins which are important for pathogens.This study identified 13 StCFEM proteins in the genome of Setosphaeria turcica, the hemibiotrophic fungus that causes northern corn leaf blight.Sequence alignment and WebLogo analysis of their CFEM domains indicated that the amino acids were highly conserved and that,with the exception of StCFEM1,2,3,and 6,they contained eight cysteines.Phylogenic analysis suggested that these 13 proteins (StCFEM1–13) could be divided into 2 clades based on the presence of the trans-membrane domain.Six StCFEM proteins with a signal peptide and without a trans-membrane domain were considered as candidate effector proteins.The CFEM domain in the candidate effector proteins could form a helical-basket structure homologous to Csa2 in Candida albicans.Transcriptome analysis suggested that the 13 genes were expressed during fungal infection and a yeast secretion assay revealed that these candidate effectors were secreted proteins.It was also found that StCFEM3,4,and 5 couldn’t affect BAX/INF1-induced programmed cell death(PCD) in Nicotiana benthamiana and while StCFEM12 could suppress INF1-induced PCD,it showed no effect on BAXinduced PCD.This study found that there were 13 members of the S.turcica CFEM protein family and that StCFEM12 was a candidate effector.This study laid the foundation for illustrating the roles of CFEM proteins during the pathogenic processes of phytopathogens.

Keywords:CFEM domain,effector,Setosphaeria turcica,signal peptide,PCD

1.lntroduction

Setosphaeria turcicais a hemibiotrophic fungal pathogen which causes northern corn leaf blight in maize growing areas all over the world (Zhanget al.2012).Characteristic disease symptoms elicited byS.turcicaare disease spots that can develop on all leaves of the maize plant (Huanget al.2020).Setosphaeria turcicahas a serious impact on maize production,particularly in areas with high humidity and moderate temperature (Möoschet al.1999).

Plant pathogens secrete some effector proteins during infection to modulate plant signaling and to suppress plant immunity through different mechanisms (Nicket al.2020).Secreted effectors can be delivered into the host cytoplasm to destroy the immune system (Ronnieet al.2011).Intriguingly,plant resistance is determined by immune receptors that recognize corresponding ligands,including PRRs (pattern recognition receptors) and NLR(nucleotide-binding-leucine-rich repeat) receptors (Wanget al.2020).These fungal effector molecules interfere with PTI (PAMP-triggered immunity) or ETI (Effector-triggered immunity) by inhibiting the translation of receptor proteins(Toruñnoet al.2016) and while PTI and ETI responses are similar,they may differ in their strength.The nature of the delivery mechanism for fungal effectors is also unclear(Rafiqiet al.2012).The strong recognition between effectors and receptor usually results in the hypersensitive response (HR),which occurs at infection sites and prevents pathogenic further expansion (Xuet al.2019).As a typical example,RXLR effector proteins,existing in the hundreds in numerous oomycete genomes,are the best-characterizedPhytophthora infestanseffectors (Baxteret al.2010).In recent years,many studies have been carried out to elucidate the biological functions of effector proteins fromMagnaporthe oryzae,Hyaloperonospora arabidopsidis,Botrytis cinerea,Ustilago maydis,andPuccinia striiformisf.sp.tritici (Loet al.2015).Previous studies showed that Avr1b,the effector protein inPhytophthora sojae,could suppress programmed cell death (PCD) triggered by BAX inNicotiana benthamianacells and could contribute positively to virulence (Douet al.2008a).AvrPiz-t,theM.oryzaeavirulence gene,suppressed PAMP-triggered generation of reactive oxygen species (ROS) and promoted virulence (Parket al.2012).However,little is known about the effectors ofS.turcica.Using bioinformatics methods,60 candidate effector molecules ofS.turcicawere predicted to exist (Yanet al.2017).

The Common in fungal extracellular membrane (CFEM)domain is unique to fungi and consists of about 60 aa residues,of which eight are conserved cysteine residues(Reshamet al.2003).The CFEM domain-containing fungal genes have been intensively studied in model fungi,such asColletotrichum graminicola,Candida albicans,B.cinerea,andM.oryzae(Jinget al.2016).CFEM domain effectors can be attached to the fungal cell wall and can be transferred to GPI-anchored site where they can function in the pathogenesis (Yakiret al.2014).For example,the effector geneBcCFEM1inB.Cinereaencoded a CFEM domain containing the putative GPI-anchored site which directly affects pathogenicity,conidial production,and stress tolerance (Zhuet al.2017).While the roles of some effector proteins in the pathogenesis have been characterized,such as Csa2 ofC.albicansand ACI1 ofM.oryzae(Gonget al.2020),there is a distinct lack of systemic identification and functional analysis of the CFEM domain-containing proteins inS.turcica.

This study identified a total of 13 CFEM proteins in theS.turcicagenome.Six of them were considered to be candidate effectors for functional study after taking into account the major criteria for the definition of fungal effector proteins (those with a signal peptide for secretion and lackingtrans-membrane domains).These 6 candidates had varied structures,different transcription levels and diverse cell-death suppression activities in the non-host plantN.benthamiana,and might have important pathogenicity functions.

2.Materials and methods

2.1.Fungal and bacterial strains,plants,and growth conditions

The wild-typeS.turcicastrain,01-23,was inoculated onto potato glucose medium (PDA:2% glucose,20%potato,1.5% agar) and cultured at 25°C for 8–10 days.Yeast YTK12 strain was routinely grown on yeast extract peptone dextrose medium (YPD:1% yeast,2% peptone,2% glucose,0.003% sulfate adenine and 2% agar) and was maintained at 30°C under continuous fluorescent light.TheAgrobacterium tumefaciensstrain GV3101 was used forAgrobacterium-mediated transient expression in plant leaves.Escherichia colistrain DH5α was used for plasmid amplification.Nicotiana benthamianawere grown in a greenhouse at (22±1)°C with a 16-h light period.

2.2.Bioinformatic identification of CFEM domaincontaining proteins in S.turcica

To identify CFEM effectors inS.turcica,the CFEM domaincontaining protein ACI1 inM.oryzaewas used as the query.The JGI Genome Database (https://genome.jgi.doe.gov/portal/) was searched using Basic Local Alignment Search Tool algorithms (BLASTP) with a threshold of E-value<1e–10.All obtained proteins were examined for the presence of CFEM domains using the PFAM tool in the SMART website (http://smart.embl-heidelberg.de/)and sequences without CFEM domains were manually omitted.The signal peptide sequences of these proteins were analyzed using the SignalP 4.1 Server (http://www.cbs.dtu.dk/services/SignalP/).

2.3.Multiple sequence alignments and phylogenetic analysis

To identify conserved amino acids in the StCFEM proteins,the CFEM domain sequences were extracted and a multiple sequence alignment using CLUSTALW was performed (Hanet al.2016).Two CFEM domains,ACI1 (AAN64312.1) inM.oryzaeand Csa2 (XP_713316.1) inC.albicanswere used as the template (Gonget al.2020).The alignment result was modified and drawn with DNAMAN Software and analyzed using WebLogo (http://weblogo.berkeley.edu/logo.cgi).

The StCFEM protein phylogenetic tree was constructed using the neighbor-joining method with the MEGA 7.1 Program (Gonget al.2020).Bootstrap analysis was conducted using the P-distance model with 500 replications.Additionally,the domain architecture of the StCFEM effectors was drawn with iTOL (https://itol.embl.de/) in this study.These domains,including the CFEM domain,were obtained from the bioinformatics analyses from the SMART website.The signal peptide,andtrans-membrane regions were derived from analysis using the SignalP 4.1 Server.

2.4.Protein model analysis of candidate StCFEM effectors

Proteins with an N-terminal signal peptide and notransmembrane domain were candidate effectors.Phyre2 (http://www.sbg.bio.ic.ac.uk/phyre2) was employed to generate predicted structures of these effectors using the amino acid sequences of the 6 candidates.The sequences were blasted with PSI-Blast to find homologies and the Hidden Markov Model (HMM) structure was built.

2.5.Transcriptomic analysis of CFEM protein from S.turcica

Appressoria development was inducedin vitroon an artificial cellophane film using the previously described method (Liet al.2016).The relatively pureS.turcicaconidia were collected from cultures incubated on PDA plates at 25°C for 10 days,and the spore suspension with a concentration of 10 spores µL–1was prepared as the test material.Three pieces of 1.5 cm×1.5 cm cellophane were placed on the agar plate surface.In accordance with Haoet al.(2015),the spore suspension was inoculated onto the cellophane (20 microliters per drop),incubated at 25°C in dark and humid conditions,and samples were collected at 0 h (conidia),3 h(germ tube formation of at least 80%),and 12 h (appressoria formation of at least 80%) post inoculation.Each sample was ground into powder in liquid nitrogen and total RNA was extracted using a Trizol Kit (Thermo Fisher,Beijing).The individual cDNA was synthesized by reverse transcription.Data were obtained from three independent replicates.

The resulting products were connected with adapters,after which size selection and PCR amplification were performed.The constructed library was analyzed using the Illumina HiSeq 4000 Sequencing Platform (BJZX,Beijing,China).The read coverage from a gene was used to calculate its expression level.Adaptor sequences and low-quality reads were initially filtered from the raw data and the remaining clean reads were aligned to the reference genome ofS.turcicausing TopHat2 (Kimet al.2013).Unigene expression was calculated as the FPKM (fragments per kilo bases of exons for per million mapped reads) using the software package Cufflinks.The false discovery rate(FDR) was used to determine the threshold of theP-value in multiple tests.The data used in this experiment can be found in Appendix A,and the RNA-Seq data can be accessed from the NCBI Database using the accession code:SUB7703090(http://www.ncbi.nlm.nih.gov/sra).Heat maps using gene expression level data were drawn using HeatMap (http://www.heatmapper.ca/expression/).

2.6.Yeast secretion assay

The signal peptide sequences of the candidate proteins,StCFEM 3,4,5,10,12,and 13,were amplified by PCR using primers designed with Primer 5.0 Software (Appendix B).The signal peptide sequences were cloned into theEcoRI andXhoI sites of the yeast vector pSuc2t7M13ori and linked to the sucrase geneSuc2,which lacks the initiation codon and signal peptide,to express the SP-SUC2 fusion protein(Konget al.2016).The 6 StCFEMSP signal peptides,the Avr1bSP signal peptide,and the mg87SP signal peptide were inserted into plasmid pSuc2t7M13ori and transformed into yeast strain YTK12.

Transformants were diluted and screened on CMD–W(minus Trp) medium (0.67% yeast N base without amino acids,0.075% W dropout supplement,2% sucrose,0.1%glucose,and 2% agar),Sucrose medium plates (1%yeast extract,2% peptone,2% sucrose,and 2 µg mL–1antimicyn A) and YPRAA medium plates (1% yeast extract,2% peptone,2% raffinose,and 2 µg mL–1antimicyn A).Invertase activity was detected by the reduction of 2,3,5-triphenyltetrazolium chloride (TTC) to the insoluble,red colored 1,3,5-triphenylformazan (TPF).Transformants were cultured in liquid CMD–W medium at OD600of 0.3 and about 1.5 mL of cell suspension was collected and re-suspended in 250 µL of 10 mmol L–1acetic acid-sodium acetate buffer (pH 4.7),500 µL 10% sucrose solution (w/v)and 750 µL sterile distilled water at 37°C for 10 min.After centrifugation at 800×g for 3 min,400 µL of the supernatant was put into a glass test tube containing 3.6 mL 0.1% TTC and incubated at room temperature for 5 min.Observation of colorimetric changes indicated that the signal peptide was functional (Yinet al.2018).

2.7.Bax/lNF1 suppression assay

The whole cDNA sequence of the candidate proteins StCFEM 3,4,5,and 12 was amplified by PCR with primers designed with Primer 5.0 Software (Appendix C) and cloned into theSmaI andClaI sites of pGR107,a recombinant potato X virus vector with a 35S promoter.The final recombinant vectors were transformed intoA.tumefaciensGV3101,which was cultured in LB (Luria-Bertani) medium with 25 mg L–1rifampicin and 50 mg L−1kanamycin.To infect tobacco plants the following procedure was used.Recombinant strains were washed three times with acetosyringone (AS) buffer (10 mmol L–1MgCl2,10 µmol L–1(acetosyringone) AS,10 mmol L–12-(N-morpholino)ethanesulfonic acid (MES),pH5.6) and infiltrated into leaves of 4-week-oldN.benthamianaat an OD600of 0.5 using syringes without needles attached.The strains expressing StCFEM proteins were infiltrated into leaves 24 h prior to the infiltration of the strain carrying Bax/INF1.The disease symptoms in theN.benthamianaleaves were noted at 4–5 days post inoculation (dpi) until BAX/INF1-induced cell death was visible.

3.Results

3.1.Bioinformatics identification of CFEM protein repertoire in S.turcica

The Pfam Database is a huge collection containing many multiple sequence calibrations and a “hidden” Markov model:covering many protein families.In this study,13 CFEM proteins (StCFEM1–13) were identified in theS.turcicagenome through BLASTP analysis using ACI1 fromM.oryzaeas query and the presence of CFEM domains in these proteins was further verified by SMART analysis.The protein ID of these proteins in theS.turcicagenomed database is shown in Table 1.The amino acid length of these proteins ranged from 96 (StCFEM13) to 758 (StCFEM11).Every CFEM protein except StCFEM3 contained a CFEM domain positioned near its N-terminus.

Table 1 Identification of the Setosphaeria turcica Common in Fungal Extracellular Membrane (CFEM) protein family

3.2.Bioinformatics analysis of CFEM domain

A neighbor-joining phylogenetic tree was constructed based on the amino acid sequences of the StCFEM proteins(Fig.1).These CFEM proteins could be divided into 2 clades,one containing 7 proteins (StCFEM1,2,6,7,8,9,and 11) which contained sometrans-membrane regions,and the other containing 6 proteins (StCFEM3,4,5,10,12,and 13) which had notrans-membrane domains.Fungal effectors are generally involved in the transport of proteins within cells and are guided by signaling peptides.All proteins,except StCFEM1 and 6,contained an N-terminal signal peptide and were predicted to be secreted proteins.

Fig.1 Phylogenetic analysis of Common in Fungal Extracellular Membrane (CFEM) proteins in Setosphaeria turcica.A neighborjoining phylogenetic tree was constructed based on the amino acid sequences of StCFEM effectors.The numbers at nodes represent the percentage of their occurrence in 1 000 bootstrap replicates;the nodes supporting more than 50% are shown in the result.The red box represents the length and localization of the signal peptide,the yellow box represents the CFEM domain,and the blue box represents the presence of a trans-membrane domain.

To further analyze and elucidate the conserved amino acids in the CFEM domains,a multiple sequence alignment was performed (Fig.2-A) with 2 well-known CFEM domaincontaining proteins,Csa2 fromC.albicansand ACI1 fromM.oryzae,used as references.The length of CFEM domain was conserved in all selected proteins at about 60 aa (ranged between 54 and 74 aa).Most of CFEM domains usually contained 8 characteristically spaced cysteines which formed 4 disulfide bonds to help stabilize the whole structure of the domain,however,a few proteins contained domains with fewer conserved cysteines,such as StCFEM6 which had 6,and StCFEM1,2,and 9 which owned 7.A conserved aspartic residue was found in 10 CFEM domains,which was important for haem binding.This study found other relatively conserved amino acids in these CFEM domains including proline (at position aa 6),valine,and leucine.The different amino acid composition of these CFEM domains may result in the formation of various structures,and serve multiple functions during plantpathogen interactions.WebLogo analysis indicated that the amino acids of the CFEM domains from theseS.turcicaderived proteins werehighly conserved and contained 8 cysteines.

Fig.2 Multiple amino acid sequence alignment and WebLogo of Common in Fungal Extracellular Membrane (CFEM) domains from 13 proteins.CFEM domains of Csa2 from Candida albicans,and ACI1 from Magnaporthe oryzae were used as positive control.A,alignment was performed using ClustalW and the result was modified with DNAMAN.Conserved amino acid are highlighted in black and gray and marked with the corresponding letters.B,for WebLogo,letters of prominent size indicate a relatively high degree of conservation.

Fungal pathogen effectors play important roles in pathogenicity and usually contain signal peptide and do not havetrans-membrane regions.According to these criteria,this study identified 6 CFEM proteins (StCFEM3,4,5,10,12,and 13) which could be considered as candidate effectors and hence were selected for further analysis.

3.3.Helical-basket shape detected in the structural model of the CFEM domains

The amino acid sequences of the CFEM domains from the 6 selected candidate effectors were uploaded into Phyre2 for modeling construction and protein homology analysis.The results showed that each sequence showed a high similarity with Csa2 (c4y7sC inC.albicans),which was the only CFEM protein possessing a three-dimensional (3D)crystal structure.Each CFEM domain has been modelled with more than 95% confidence to Csa2.The structure of the CFEM domain in Csa2 protein is comprised of 6 α-helices and one β-strand which form a helical-basket shape with an elongated N-terminal loop as handle.The models of all selected CFEM domains also contained a helical basket like structure similar to Csa2 (Fig.3).Among the 6 CFEM proteins,2 (StCFEM5 and 10) contained 5 α-helices,3(StCFEM3,12,and 13) contained 3 α-helices and 1 β-strand,and 1 (StCFEM4) contained 4 α-helices.

Fig.3 Analysis of structural models of StCFEM effectors from the Phyre2 Server.The amino acid residues of these effectors have been modelled on the crystal structure of the Common in Fungal Extracellular Membrane (CFEM) domain in Csa2 from Candida albicans with more than 90% confidence.The amino acid sequence of each effector is shown in the top line.The sequence on the next line shows residues colored according to a property-based scheme:yellow (A,S,T,G,P:small/polar),green (M,I,L,V:hydrophobic),red (K,R,E,N,D,H,Q:charged) and purple (W,Y,F,C:aromatic+cysteine).Green helices represent an α-helix,blue arrows indicate β-strands,and faint lines indicate coils.

3.4.Transcriptomic analysis of CFEM domaincontaining protein expression

The effector is secreted by pathogenic fungi during infection,and it is more likely that the infection stage is closely related to overall fungi pathogenicity.There are 4 stages,including the conidial stage,germing stage,appressorial formation and invasive hyphal growth,duringS.turcicainfection.Samples at different stages were subjected to RNA-Seq.Each sample generated 43.01 million clean reads,each with a GC content of 46.44%.The average coverage of transcriptome data relative to the whole genome (S.turcica)was 90%.The Q30 of the raw data ranged from 93.19 to 93.28%,indicating that they were high-quality reads worthy of further analysis.Based on different stages of the developmental process of the invasive structure,this studycompared the differential expression of 13 CFEM proteincoding genes by transcriptome analysis (Fig.4).All these genes were expressed in different degrees during the whole invasive process.Compared to the other genes,StCFEM4,5,and 12,which were predicted to be secreted proteins,were highly expressed throughout the whole infection processes,especially in the appressorial formation stage.StFEM3 was highly expressed in the appressorial formation and germing stage.StCFEM10 was stably expressed during all infection stages,but the expression level was relatively low.StCFEM13 was lowly expressed mainly in the appressorial formation stage and minimally expressed in the other stages.

Fig.4 Heat map showing expression levels of Common in Fungal Extracellular Membrane (CFEM) protein-coding genes.A bacterial spore suspension with a concentration of 10 spores µL–1 was seeded on the sterile cellophane film with a 20 µL drop,and incubated at 25°C under the protection of light and moisture.Inoculation materials were collected at 0 h (conidia),3 h (bud tube forming >80%),and 12 h (apposition forming rate >80%) after inoculation.The cDNA was sent to Beijing Group Biotechnology Co.,Ltd.,China,for sequencing.The sample was analyzed using the Illumina HiSeq 4000 Sequencing Platform.Unigene expression was calculated as the FPKM using the software package Cufflinks.Heat maps were drawn from gene expression level data using R with the darker the color,the higher the expression level.

3.5.Functional evaluation of the signal peptide of 6 candidate effectors

A yeast secretion system described previously (Jacobset al.1997;Huet al.2014) was employed to verify the function of the signal peptide from above candidate effectors.The signal peptide of StCFEM3,4,5,10,12,and 13,mg87SP(the signal peptide of mg87),and Avr1bSP (the signal peptide of Avr1b) were introduced into pSuc2t7M13ori and transformed into a yeast strain,YTK12,which lacks a secreted invertase.Transformants were streaked onto CMD–W plates,YPRAA plates and sucrose plates,which only support the growth of yeast with a functional secreting invertase.In CMDW medium (with sucrose,and where yeast growth seems seen even in the absence of a secreted invertase) and in the absence of tryptophan(Trp),the negative control (pSUC2-mg87SP),the positive control (pSUC2-Avr1bSP),and pSUC2-StCFEMsSP all grew well when the TRP1 cassette was in the vector.The transformants containing StCFEMsSP grew normally on the YPRAA and sucrose plates.Compared to Avr1bSP,YTK12 and the transformants containing negative mg87SP failed to grow on the YPRAA and sucrose plates (Fig.5-A).

The enzyme activity of secreted invertase was detected by the reduction of 2,3,5-triphenyltetrazolium chloride(TTC) (Fig.5-B).The transformants containing Avr1bSP and StCFEMSP could turn the reaction reagents red,while the strain containing mg87SP as the negative control and the mock one couldn’t,indicating that the signal peptide of Avr1b and StCFEMs could mediate invertase secretion.These results showed that the StCFEM signal peptides were functional,confirming the accuracy of the SignalP prediction,and strongly suggested that StCFEMs were secreted proteins.

Fig.5 Functional evaluation of the signal peptide of 6 candidate effectors.A,the secretion function of the signal peptide of candidate effectors were confirmed using a yeast secretion system.The function of signal peptides of 6 candidate effectors was identified using the Saccharomyces cerevisiae strain YTK12 containing the pSuc2t7M13ori vector,Avr1b,mg87 or CFEMSP.Yeast transformants were cultured on CMD–W medium (on which the yeast grew well without invertase secretion),sucrose plates,and YPRAA medium (yeast only grew with active invertase secretion).B,the invertase activity was detected by reduction of 2,3,5-triphenyltetrazolium chloride(TTC) to the insoluble,red colored 1,3,5-triphenylformazan(TPF).Transformed cells were collected in tubes to detect invertase activity.The negative control is the YTK12 strain carrying empty vector,and the positive control is the YTK12 strain carrying the signal peptide of Avr1b.WT,mg87,Avr1b,St3,St4,St5,St10,St12,and St13 represent YTK12,mg87SP,StCFEM3SP,StCFEM4SP,StCFEM5SP,StCFEM10SP,StCFEM12SP,and StCFEM13SP,respectively.

3.6.StCFEM effectors suppress programmed cell death in N.benthamiana

Some fungal effector proteins secreted into plant cells can induce or suppress plant cell death,which contributes positively to pathogen virulence.The effects of StCFEM3,4,5 and 12,which were actively expressed during the invasive process,on BAX/INF1-induced programmed cell death (PCD) ofN.benthamiana,which resembles the defense-related HR in plant cells,was assayed to identify whether these StCFEMs possessed the effector activity.HR suppression is regarded as one of the criteria for suppression of plant immunity.BAX/INF1-induced PCD inN.benthamianawas used as the positive control.The results showed that StCFEM3,4,5,and 12 did not suppress Bax-induced PCD inN.benthamiana(Fig.6),and could not abolish Bax function.StCFEM3,4,and 5 did not suppress INF1-induced PCD yet.However,few symptoms of cell death were observed following StCFEM12 infiltration.The results indicated that StCFEM12 could suppress INF1-induced cell death inN.benthamianaleaves.Therefore,it is concluded that StCFEM12 is a potential effector protein inS.turcica.

Fig.6 Suppression of BAX/INF-triggered cell death on Nicotiana benthamiana leaves.Agroinfiltration was performed on tobacco leaf with Agrobacterium tumefaciens carrying BaX/INF1,StCFEM,or BaX/INF:StCFEM.BAX/INF was used as a positive control and in combination with candidate Common in Fungal Extracellular Membrane (CFEM) to determine the hypersensitive response.The results were recorded 5 days post inoculation (dpi).

4.Discussion

Pathogens secrete proteins called effectors that may act either inside or outside plant cells,interact with various host targets to block PTI or ETI and lower plant immunity to assist pathogen survival (Douet al.2008b).Current efforts are being made to identify the pathogenic function of effectors from filamentous phytopathogens (fungal and oomycete).For example,someP.infestanseffectors or their targets can suppress INF1-triggered cell death (ICD) to regulate plant pathogens (Qinet al.2020).A big breakthrough for the effector field has been the identification of the conserved amino acid motifs RxLR in oomycete pathogens which have been proven to be conserved in some plant pathogens.Recent research concluded that PpE4 was a virulence RXLR effector ofPhytophthora parasiticawhich triggered non-specific cell death in a variety of plants,suggesting that it is recognized by the plant immune system (Huanget al.2019).PexRD2 and Pi22926,2 RXLR effectors fromP.infestans,promotedP.infestanscolonization by targeting MAP3K proteins that acted in parallel with the same immune signal transduction pathway to mediate infection (Renet al.2019).There is,however,no such universal motif for fungal effectors.Fungal effector proteins in pathogenic fungi have been increasingly studied in recent years.For example,a recent study suggested that Sta1 in the biotrophic fungusUstilago maydiswas a novel kind of effector,whose deletion resulted in a dramatic reduction in pathogenicity (Tanakaet al.2020).An essential virulence effector,SCRE1,was identified inUstilaginoideavirens,which could inhibit host immunity (Zhanget al.2020).These studies suggested that fungal pathogenic effector proteins were crucial in plant infection.

The CFEM family is a critical fungal protein group and elucidated CFEM effector function is essential for pathogen–host interactions.Some new CFEM proteins have been identified from different fungi since the first discovery of CFEM proteins inCoccidioides immitis.CFEM domains consist of about 60 aa with a total of 8 cysteines,which are frequently found in extracellular fungal proteins (Zhanget al.2017).Until now,many CFEM domain proteins have been identified in numerous fungal pathogens,and the number of domains in each protein varies from one to more than 10 (Zhanget al.2015).The numbers of CFEM in different phyla of fungi were computed and their pathogenesis was examined (Zhanget al.2015).The results showed that there were 10 CFEM domains in10 proteins ofNeurorspora crassa,20 in 20 proteins ofM.oryzaeproteins,and 9 in 6 proteins ofC.albicans,1 in 1 protein ofSaccharomyces cerevisiae,and none inSicyopterus japonicusandRhizopus oryzae.They found that the number of CFEM domains in pathogenic species was significantly larger than that in nonpathogenic fungi.The current study found that there were 13 CFEM domains in13 proteins ofS.turcica.Hence,it was proposed that CFEM domains inS.turcicawere more likely to be associated with pathogenicity.In addition,it was found that most CFEM domains located near the N-terminus of the corresponding protein,while the one in StCFEM3 located at the C-terminus,which was seen in other fungi as well such asC.graminicola(Jinget al.2016).Therefore,it is concluded that these particular structures may play multiple functions during the complex pathogenic processes.

Functional studies of CFEM proteins showed that these proteins were closely associated with pathogenicity.For example,inM.oryzae,PTH11,an appressorium signaling effector molecule,was closely related to pathogenicity (Deng and Dean 2008) and MoCDIP2 induced apoptosis of host cells,which was the only earliest identified CFEM effector inM.oryzae(Songkumarn and Pattavipha 2013).The CFEM effector AG2 regulated mycelium growth and pathogenicity inC.simmitis(Penget al.1999).It has been confirmed that the secreted protein expressed in the infection stage was more likely to affect the pathogenicity of the pathogen (Elliset al.2009).The present study determined that 13 CFEM proteins inS.turcicawere expressed during the infection stage by transcriptomic analysis,and that StCFEM3,4,5,10,12,and 13 were all expressed at the infection stage.It was therefore proposed that these StCFEMs may function as pathogenic factors inS.turcica.Previous study inFusarium oxysporumshowed that during the sporulation and hypha infection stages,expression of CFEM_DR was significantly higher than that seen in other stages.This protein was known to play important pathogenic roles in colonization and infection of hosts (Linget al.2015).Zhuet al.(2017) concluded that the transcription levels of BcCFEM1 fromB.cinereaincreased from 12 to 36 hpi,after which they decreased but were still higher than at 0 hpi,indicating that this effector gene may play significant roles in the early stages of infection.Transcription analysis of 22 CFEM domain-containing pathogenic effectors ofC.graminicolashowed that they were expressed highly in all 3 stages of the infection period (Jinget al.2016).The current research found that the StCFEM3,4,5,and 12 expression was significantly higher in the appressorial stage than in other stages,suggesting that they may play an important role in the appressorial formation and/or later infection stages.

The first step in identifying an effector is to determine whether it can be secreted extracellularly.Since secreted proteins play an important role in pathogenesis,almost all existing research in the field of plant pathogens focused on the secretome.A number of virulence factors have been identified,such asPst(Xuet al.2019),P.viticola(Konget al.2016),andVerticillium dahliae(Liuet al.2013).It is generally thought that secreted proteins are required for successful disease development and for determining hostpathogen compatibility.In this study,the secretion activity of the signal peptide of StCFEM3,4,5,10,12,and 13 were identified,indicating that these StCFEMs belonged to secretory proteins.However,not all secreted proteins are associated with pathogenicity,and studies have shown that secreted proteins expressed at the infection stage are more likely to affect pathogenicity and more likely to be effector than [XXX] (Jinget al.2016).Hence,StCFEM3,4,5,and 12,which specifically expressed in the appressorial stage,were used to do Bax/INF1 suppression assay.This study finally obtained a specific CFEM protein,StCFEM12,that could suppress programmed cell death (PCD) triggered by ETI.In summary,StCFEM12 was identified as an CFEMcontaining effector ofS.turcica.The functions of these CFEM proteins need to be illustrated in future research to lay a solid foundation for exploring new strategies for disease prevention and control.This study provides insights into the regulation of CFEM family byS.turcica,which will pave the way for understanding the molecular mechanism by which pathogen effectors manipulate immunity.

5.Conclusion

This study identified 13 StCFEM proteins containing CFEM domains in theS.turcicagenome.The CFEM domain in these effectors could form a helical-basket structure with different conserved amino acids.Six of these proteins were candidate effectors due to the presence of a signal peptide and the lack of anytrans-membrane domains.Moreover,structural model analysis suggested that CFEM effectors fromS.turcicawere highly similar to Csa2 (c4y7sC) inC.albicans.The results supporting the functionality of the signal peptide from 6 StCFEM proteins indicated that StCFEM3,4,5,10,12,and 13 were secreted proteins.Transcriptome analysis suggested that StCFEMs(StCFEM3,4,5,10,12,and 13) was specifically expressed in the pathogen infection stage.Furthermore,compared with other genes,StCFEM3,4,5,and 12 were highly expressed in the appressorial stage.Finally,it was found that StCFEM12 could suppress PCD on tobacco leaves induced by INF1.

Acknowledgements

This work was supported by the grants from the Natural Science Foundation of Hebei Province,China (C2020204172 and C2018204120),the China Agriculture Research System of MOF and MARA (CARS-02-25) and the National Natural Science Foundation of China (31601598 and 31901827).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on http://www.ChinaAgriSci.com/V2/En/appendix.htm