10PE and PPE Genes: A Taleof Conservation and Diversity

Giovanni Delogu, Michael J. Brennan,and Riccardo Manganelli

Abstract

PE and PPE are two large families of proteins typical of mycobacte-ria whose structural genes in the Mycobacterium tuberculosis complex(MTBC) occupy about 7% of the total genome. The most ancestral PEand PPE proteins are expressed by genes that belong to the same operonand in most cases are found inserted in the esx clusters, encoding a typeVII secretion system. Duplication and expansion of pe and ppe genes,coupled with intragenomic and intergenomic recombination events, ledto the emergence of the polymorphic pe_pgrs and ppe_mptr genes in theMTBC genome. The role and function of these proteins, and particularlyof the polymorphic subfamilies, remains elusive, although it is widelyaccepted that PE and PPE proteins may represent a specialized collectionused by MTBC to interact with the complex host immune system ofmammals. In this chapter, we summarize what has been discovered sincethe identification of these genes in 1998, focusing on M. tuberculosisgenetic variability, host-pathogen interaction and TB pathogenesis.

Keywords

PE • PPE • PE_PGRS • PPE_MPTR • Genetic variability • Genefamily • Polymorphism

G. Delogu (�)Institute of Microbiology, Università Cattolica del SacroCuore, Largo A. Gemelli, 8, 00168, Rome, Italye-mail: giovanni.delogu@unicatt.it

M.J. BrennanAeras, 1405 Research Blvd., Rockville, MD 20850, USAe-mail: mbrennan@aeras.org

R. ManganelliDepartment of Molecular Medicine, University of Padua,Via A. Gabelli, 63, 35121, Padua, Italye-mail: riccardo.manganelli@unipd.it

© Springer International Publishing AG 2017S. Gagneux (ed.), Strain Variation in the Mycobacterium tuberculosis Complex: Its Rolein Biology, Epidemiology and Control, Advances in Experimental Medicine and Biology 1019,DOI 10.1007/978-3-319-64371-7_10

191

192 G. Delogu et al.

10.1 General Features

Sequences encoding PE_PGRS proteins inMycobacterium tuberculosis (Mtb) are thePolymorphic-GC-Rich Sequences (PGRS) thatwere first used in the molecular typing of Mtb(Poulet and Cole 1995). This use occurredprior to the publication of the first Mtb genomesequence of the laboratory strain H37Rv in 1998(Cole et al. 1998). The sequencing was slowedby the homology and redundancy of pe_pgrs, aswell as pe and ppe. Since then, we have learnedmuch about the evolution and cell localizationof certain PE and PPE proteins, but we havemuch more to learn about their function andimmunogenicity.

In 1998, the first Mtb genome sequence high-lighted for the first time the presence of genesgrouped into two large families that were shownto comprise approximately 10% of the genomesize and to account for 7–8% of the Mtb codingpotential (Cole et al. 1998; Brennan and Delogu2002). Based on the presence of conserved Pro-Glu (PE) and Pro-Pro-Glu (PPE) motifs at theirN-terminus, the proteins encoded by the mem-bers of these two gene families were named PEand PPE, respectively. The Mtb H37Rv genomecomprises 100 pe and 69 ppe genes encoding thecorresponding proteins, which are further classi-fied into subfamilies depending on the amino acidsequence at the C-terminus. The 100 proteinsof the PE family share a highly conserved N-terminal domain of about 90–100 amino acidsin length and are further subdivided into thePE and PE_PGRS subfamilies (Gey van Pittiuset al. 2006). PE_PGRS proteins are characterizedby the presence of a polymorphic domain, richin Gly-Gly-Ala/Gly-Gly-Asn amino acid repeats,which can vary in sequence and size. pe_pgrsgenes are found scattered in the genome and aremostly not co-transcribed with other genes. Con-versely, many of the pe genes are adjacent to ppegenes and a number of studies have demonstratedthat these pe/ppe couplets are co-expressed, andat least some of the corresponding proteins areheterodimers that are secreted (Adindla and Gu-ruprasad 2003; Ates et al. 2016). A number of PEproteins, characterized by a unique C-terminaldomain, cannot be included in the PE_PGRS and

PE/PPE couplet subfamilies and can be classifiedas PE unique (Delogu et al. 2008). The 69 PPEproteins are characterized by the presence of ahighly conserved N-terminal domain which isapproximately 180 amino acids in length andthat, similar to the PE domain, seems to playa key role in driving protein localization or se-cretion (Daleke et al. 2011; Dona et al. 2013).Many PPE proteins appear to be co-expressedwith the PE partner (as mentioned) and belongto the PE/PPE pair subfamily; others are encodedby genes found scattered in the chromosome.Twenty of them contain, downstream of the con-served N-terminal PPE domain, the Major Poly-morphic Tandem Repeat (MPTR) region char-acterized by multiple C-terminal repeats of theamino acid sequence motif Asn-X-Gly-X-Gly-Asn-X-Gly (Gey van Pittius et al. 2006).

The discovery of these two protein familiesin 1998 (Cole et al. 1998) prompted many tospeculate on their relevance in Mtb biology and,given their abundance in M. tuberculosis com-plex (MTBC), also about their role in tubercu-losis (TB) pathogenesis. Since then, a numberof studies have started to provide experimentaldata that specifies a function that, however, re-mains mostly vague. In this chapter, we providea general overview of these families and theirimpact on TB pathogenesis, with a special focuson the biological significance of the genetic con-servation as well as diversity of these apparentlyredundant proteins.

10.2 Origin and Expansion

Pe and ppe genes are found in the Mycobacteriumgenus with the fast-growing, non-pathogenic,saprophytic species such as Mycobacteriumsmegmatis or Mycobacterium abscessus, whichare the closest to the last common ancestor of thegenus, having only a few pe and ppe genes. Incontrast, the slow-growing pathogenic mycobac-terial species contain many pe and ppe genes(Gey van Pittius et al. 2006; Sassi and Drancourt2014). Formative studies by Gey van Pittiusand colleagues (2006) aimed at investigatingthe evolutionary history of pe and ppe genes,provided a beginning for defining their genetic

10 PE and PPE Genes: A Tale of Conservation and Diversity 193

relationship in the MTBC complex. Basedon in silico analysis of the available genomesequences, DNA hybridization assays andcomparative genomics, these authors (Gey vanPittius et al. 2006) reconstructed the phylogeneticrelationship of these genes and subdivided boththe pe and ppe genes into five subfamilies. Theprogenitors of these families belong to subfamilyI and are associated with the esx-1 genecluster. Non-pathogenic fast-growing speciestend to have few pe/ppe genes, while slow-growing pathogenic species, such as MTBC,Mycobacterium marinum and Mycobacteriumulcerans, possess a high number of pe/ppe genes,and an accumulation of the most polymorphicgenes (subfamilies V), including pe_pgrs andppe_mptr (Gey van Pittius et al. 2006; Deloguet al. 2008). Expansion of the pe and ppegenes is thought to have been initiated with theduplication of the esx-1 gene cluster, followed bymultiple duplication events (Gey van Pittius et al.2006). The association of esx-1, esx-3 and esx-5with virulence of Mtb (Pym et al. 2002; Simeoneet al. 2009; Serafini et al. 2013; Tufarielloet al. 2016; Bottai et al. 2012), the discoverythat duplication and expansion of pe_pgrs andppe_mptr genes followed the emergence of esx-5(Gey van Pittius et al. 2006), and the elevatedpolymorphisms observed in these genes, wereseen as an indication that PE and PPE proteinsare key for the pathogenesis of TB (Brennan andDelogu 2002; Gey van Pittius et al. 2006).

Comparative genomic studies in smoothtubercle bacilli (STB) strains (see Chap. 2),have also provided useful insights into theevolution of these genes and their contributionin shaping Mtb as a human pathogen. Namouchiand colleagues (2013) identified strong signalsof intragenic and intergenic recombinationevents in pe_pgrs genes that, together withmutation events, were important for generatinggenetic diversity. This was highlighted by thepositive selection demonstrated in STB strainsfor a few genes (Namouchi et al. 2013). Aninteresting example of this is the evolutionof pe_pgrs17 and pe_pgrs18 that emergedas a duplication event in an ancestral strain.This was followed by an accumulation ofgenetic changes that resulted in a polymorphic

region (named 12/40) in pe_pgrs17. Thispolymorphic region, which was conserved duringthe expansion of the Mycobacterium africanum –Mycobacterium bovis lineage, was relocated tothe other paralog pe_pgrs18 only in modernMtb strains by a process of gene conversion(Karboul et al. 2006). Hence, some modernMTBC strains have the 12/40 polymorphismand others do not. It has been proposed thatthe acquisition of the 12/40 polymorphismcoincided with the emergence of the mostsuccessful MTBC strains (Karboul et al. 2006).This implies that these two PE_PGRS proteinsare important for human-Mtb interactions. Ina broader perspective, it has been suggestedthat homologous recombination between geneticsequences emerged by duplication events, andmay have shaped the evolution of pe andppe genes (and particularly of pe_pgrs andppe_mptr genes) in STB. Expansion of thesegene families may have provided the raw materialfor functional innovation as well as adaptationto the human host (Fig. 10.1) (McEvoy et al.2009; Namouchi et al. 2013; Karboul et al. 2008;Supply et al. 2013).

10.3 Regulation of GeneExpression

The first study characterizing the level of ex-pression of pe and ppe genes, in a systematicmanner, is that of Voskuil and colleagues (2004).Using DNA microarrays, they found that, of 169pe and ppe genes, 128 were differentially ex-pressed, in 15 different environmental conditions.No clear conserved pattern of expression wasfound for pe, ppe or even any of their subgroups,with some genes specifically induced in a singlecondition (e.g. pe3 was only induced in acti-vated macrophages), and others differentially ex-pressed in several different conditions (e.g. pe11was differentially regulated in 10 out of 15 differ-ent conditions). Expression of different subsetsof PE and PPE proteins, in response to diverseenvironmental conditions may result in a wideantigen variation, allowing Mtb to evade the hostimmune response. This also suggests that, despitetheir structural similarity, small differences in

194 G. Delogu et al.

Legend:

1Insertion of

pe ppe genesin esx-1region 2

Duplication ofpe ppe genes

within esxcluster

3Duplication ofpe ppe genes

as pairs orsingle gene

4Expansion of pe and ppe genes,

pe_pgrs and ppe_mptr anddiversification.

High level of genetic variabilityamong orthologs in STB

5Stabilization of pe_pgrs and

ppe_mptr genes in Mtb.Most genes showing low level ofgenetic variability and neutral or

purifying selection

Evolution of pe and ppe genes in Mycobacterium and MTBC

PE protein/domain PPE protein/domain Unique domain GRPLI–PGRS domain MPTR domain

Fig. 10.1 Schematic of the PE and PPE proteins showing the expansion of these two families occuring during theevolution that led to MTBC

their sequence may result in diverse functions,so that different sets of proteins are expresseddepending on the environmental conditions en-countered by the bacteria (Delogu et al. 2006).

Differential expression of PE and PPEproteins has also been studied in vivo. Rachmanand colleagues (2006) identified two PE, fourPE_PGRS and three PPE_MPTR induced inlung granulomas from human patients, whileRamakrishnan and colleagues (2000) identifiedtwo PE_PGRS-coding genes of M. marinumpreferentially expressed in macrophages. Inaddition, studying the expression of ppe_mptrgenes throughout the infection of mice, it wasdemonstrated that while some of these genes areexpressed constitutively, others are repressed oralternatively induced during infection (Soldiniet al. 2011).

In vivo expression of PE and PPE proteinswas also demonstrated by proteomics in a studywhich analyzed the bacterial proteome duringinfection of guinea pig lungs. Interestingly,members of the PE and PPE protein family(mostly PE_PGRS and PPE_MPTR) were foundto represent the third most abundant category ofproteins expressed during infection (Kruh et al.2010).

Beyond showing a different pattern ofregulation in response to changing environmentalconditions, pe and ppe genes were also shownto be expressed at different levels in differentmembers of the MTBC (Rehren et al. 2007;Golby et al. 2007) and among different Mtbclinical isolates (Gao et al. 2005). Specifically,Mtb was found to have 18 pe and ppe genesexpressed at a superior level compared to its

10 PE and PPE Genes: A Tale of Conservation and Diversity 195

close relative M. bovis (Rehren et al. 2007).Conversely, 20 pe and ppe genes were foundto be differentially expressed among 10 clinicalisolates of Mtb (Gao et al. 2005).

Several transcriptional factors have been re-ported to be involved in the modulation of pe andppe genes expression (Ahmed et al. 2015; Moha-reer et al. 2011). In particular, (i) the Iron DEpen-dent Regulator IdeR regulates around 11 pe/ppegenes (Gold et al. 2001; Rodriguez et al. 2002);(ii) PhoPR, a two component system involved inthe regulation of genes in the synthesis of cellwall components such as the diacyltrehaloses,polyacyltrehaloses and sulfolipids (Solans et al.2014), also regulates the expression of several peand ppe genes; (iii) Rv0485, which was identifiedin a transposon mutagenesis approach as a regu-lator able to modify the expression of the pe13-ppe18 operon (Goldstone et al. 2009); (iv) finally,some alternative sigma factors are involved in thedifferential expression of PE and PPE proteins(Mohareer et al. 2011). Overexpression of SigBcauses induction of 7 pe_pgrs (Lee et al. 2008),and absence of SigD causes induction of 14pe_pgrs (Raman et al. 2004).

10.4 PE/PPE Couplets and PEand PPE Unique Proteins

The most ancient pe and ppe genes are orga-nized in operons and found within the esx-1 genecluster. A number of other pe/ppe couplets arefound within the esx-2, -3 and -5 gene clus-ters or scattered throughout the genome (Geyvan Pittius et al. 2006). The presence of thesepe/ppe gene pairs in operons was immediatelyindicative of a functional link. Tundup and col-leagues (2006) demonstrated that the pe25/ppe41genes were co-transcribed and their gene prod-ucts were shown to interact. To investigate thestructure of these PE and PPE proteins, Strongand colleagues (2006) co-expressed PE25 andPPE41 in Escherichia coli, demonstrating thatco-expression is required for protein stability.These authors were able to obtain the crystalstructure of the PE25/PPE41 complex and estab-lished that the PE protein interacts with the PPEprotein through conserved apolar residues thatwarrant strong hydrophobic interactions (Strong

et al. 2006). The structure of PE25/PPE41 resem-bles that of the ESXA/B heterodimer, providingfurther support for the observation that PE/PPEheterodimers are substrates for the Type 7 Se-cretion System (T7SS) in mycobacteria (Ateset al. 2016). Secretion of PE/PPE pairs has beendemonstrated in mycobacteria and recent studiesshowed that the EspG5 protein serves as a chaper-one for the PE25/PPE41 heterodimer, that is thendriven for secretion through the ESX-5 T7SS(Ekiert and Cox 2014; Korotkova et al. 2014).PE/PPE couplet proteins are therefore recognizedby the EspG chaperone of the respective T7SSand then translocated through the mycobacterialmembranes to be secreted or to remain associatedwith the capsular layer (Ates et al. 2016).

Inactivation of the ppe25/pe19 genes, insertedin the esx-5 gene cluster in Mtb, caused impairedsecretion and attenuated virulence in mice, impli-cating these PE/PPE proteins in TB pathogenesis(Bottai et al. 2012). The role of these PE/PPEproteins in the context of the T7SS has been thor-oughly reviewed elsewhere (Ates et al. 2016),but the functional role of these proteins in Mtbbiology and their interaction with the other T7SSsubstrates remains to be established.

Co-regulated genes encoding PE/PPE cou-plets are very homologous and are major targetsof the host T cell responses during infection.However, only few of the PE/PPE couplets haveactually been analyzed (Akhter et al. 2012; Geyvan Pittius et al. 2006; Mukhopadhyay and Balaji2011; Fishbein et al. 2015). PPE18, which is apart of a putative PE/PPE couplet, has been im-plicated in the pathogenesis of TB via its abilityto specifically bind to toll-like receptor (TLR)2 and trigger secretion of the anti-inflammatorycytokine IL-10. These are features of fully vir-ulent Mtb (Nair et al. 2009; Bhat et al. 2012).Interestingly, the immunomodulatory propertiesof PPE18 are observed when the protein is eitherexpressed alone or as a recombinant protein in E.coli or M. smegmatis. This raises the question ofwhether heterodimerization with PE13 or PE31is required for proper folding and function aspredicted by bioinformatics studies (Nair et al.2009; Riley et al. 2008). PPE17 is availableon the mycobacterial surface and may deliver aheterologous antigen to the surface, when it isexpressed in the absence of the predicted PE part-

196 G. Delogu et al.

ner (PE11) (Dona et al. 2013). Together with theobservation that a M. smegmatis overexpressingPE11 alone shows enhanced virulence comparedwith the parental strain (Singh et al. 2016), thesefindings suggest that heterodimerization with thepaired PE may not be a requirement for all PPEs(and vice versa).

A significant level of genetic variability inthe ppe18 gene was observed among MTBCstrains, particularly in the region encoding theN-terminal domain of the protein. This regionis involved in the interaction with TLR2 andis usually highly conserved in the PPE proteins(Hebert et al. 2007; Homolka et al. 2016). Withthis background, it would be of interest to in-vestigate the impact of this genetic variability onPPE18-dependent Mtb inflammatory properties.Since PPE18 is one of the most promising vac-cine candidates and it is included in the GSKMtb72F subunit polyprotein (see below in Sect.10.7.1), there has been concern that the geneticvariability observed in Mtb strains could affectvaccine efficacy. Recent data also indicates thatthe insertions/deletions and mutations observedin pe and ppe do not impact the predicted putativeT cell epitopes that cover a wide range of HLAIand II (Mortier et al. 2015).

Another ppe gene which has been analyzedbecause of Mtb genetic variability is ppe38,which has been shown to be hypervariable inthe MTBC due to IS6110-recombinations, geneconversion and recombination with the ppe71paralog (McEvoy et al. 2009). In M. marinum, in-activation by transposon mutagenesis of the geneencoding the surface exposed PPE38 protein,results in an attenuated strain defective in phago-cytosis uptake and in the ability to trigger inflam-matory responses in adult zebrafish (Dong et al.2012). The lack of a clear understanding of therole of PPE38 in Mtb biology prevents formulat-ing any hypothesis on the impact of the observedpolymorphisms. Moreover, it would be importantto assess whether the predicted loss of function inppe38 is somehow compensated by other highlyhomologous paralogs (McEvoy et al. 2009).

Other PE and PPE proteins demonstrateimmunomodulatory properties. PPE37, a 473amino acid unique protein, was shown tointerfere with the inflammatory responses

triggered by infected macrophages (Daim et al.2011). In addition, PE3, a 468 amino acidunique protein, was shown to be overexpressedduring the chronic steps of Mtb infection, andimmunization with recombinant PE3 elicitedsignificant levels of protective immune responses(Singh et al. 2013). Also, in a recent report,PE9 and PE10, two proteins encoded byadjacent co-expressed genes, were shown tophysically interact and to promote macrophageapoptosis through interaction with TLR4 (Tiwariet al. 2015), demonstrating the possibility ofheterodimerization between two PE proteins.

PE and PPE domains were shown to serve asimilar functional role in different mycobacterialspecies by mediating secretion of the enzymaticdomain with lipase activity (LipY). For example,in Mtb the N-terminal domain of LipY is a PEdomain, in M. marinum LipY is a PPE domainand in fast-growing mycobacteria it is a signalsequence (Mishra et al. 2008; Daleke et al. 2011).

10.5 PPE_MPTR

PPE_MPTR proteins are characterized by a N-terminal PPE domain followed by a variable-sizeC-terminus. The C-terminal domain (300–4000amino acids) is composed of multiple Asn-X-Gly-X-Gly-Asn-X-Gly repeats and arose from arecent evolutionary event (Gey van Pittius et al.2006). In Mtb H37Rv, there are 20 ppe_mptrgenes, which are commonly scattered throughoutthe chromosome, with few exceptions. For ex-ample, see the locus containing the genes encod-ing PPE_MPTR56-PPE_MPTR55-PE_PGRS50-PE_PGRS49-PPE_MPTR54 (Cole et al. 1998).

10.5.1 Genetic Variability

The repeat structure encoding the MPTRdomain of these PPEs is clearly predicted togo through several variations due to eitherintramolecular recombination or polymeraseslippage. Indeed, McEvoy and colleagues (2012)demonstrated that several ppe_mptr genes weresubjected to extremely high levels of variation.However, macromolecular recombinatorial

10 PE and PPE Genes: A Tale of Conservation and Diversity 197

events were shown to play a minor part in thevariation of these genes. The authors focused,in particular, on six of the most variableppe_mptr genes (ppe_mptr5/6 7/8, 24, 34, 54and 55) and found that most of the mutationswere due to non-synonymous single nucleotidepolymorphisms (nSNPs), frameshifts or in-frameindels. Although these genes show an extremelyhigh variability among strains belonging todifferent phylogenetic lineages, their sequencewas found to be identical in closely relatedisolates suggesting a slow evolution. Anotherexample of a ppe_mptr gene showing an elevatedlevel of variation is ppe_mptr42, whose PCR-RFLP profile was shown to differ in 16% ofisolates (Chakhaiyar et al. 2004).

10.5.2 Potential Role in Pathogenesis

The role of PPE_MPTR proteins in Mtb physi-ology or virulence is still unknown. Three mu-tants missing the genes encoding PPE_MPTR10,PPE_MPTR16 and PPE_MPTR21 were selectedin a high-throughput screening of an M. bovisBCG transposon mutant library enriched in acid-ified phagosomes (Stewart et al. 2005).

The highly repetitive structure of PPE_MPTRsuggests that they might act as dominantantigens. Chakhaiyar and colleagues (2004)for example, demonstrated that PPE_MPTR42was able to elicit a significant humoral responsein patients with relapsed TB, compared to agenerally low T cell response in other TBpatients. Finally, PPE_MPTR34 was shown totrigger functional maturation of human dendriticcells associated with the secretion of IL-10. Thisfacilitates a shift toward a Th2-driven response,which is often ineffective against intracellularpathogens, and is suggestive of an immuneevasion mechanism (Bansal et al. 2010b).

10.6 PE_PGRS

The PE_PGRS proteins share a molecular archi-tecture characterized by the presence at the C-terminus of a polymorphic PGRS domain thatvaries in sequence and size, the presence ofa conserved N-terminal PE domain of 90–100amino acids in length and the existence of a

highly conserved domain that links the PE andPGRS domains. The linker region consistentlyincludes a sequence with the GRPLI motif whichin four proteins is also found as an extra copyin the PGRS domain. A number of PE proteins,lacking the GRPLI and PGRS motifs were pre-viously included in the PE_PGRS family, al-though the lack of these specific amino acidsequences (PGRS and GRPLI motif), suggestthat these should rather be included in the PEunique group. Hence, in Mtb H37Rv, there are 51pe_pgrs genes that are found scattered through-out the genome and that appear to be mostlyexpressed as a monocistronic operon (Deloguet al. 2008).

10.6.1 Protein Structureand Function

The paucity of experimental data on PE_PGRSproteins prevents a satisfactory understandingof their localization in the Mtb cell. However,PE_PGRS33 was found associated with themycobacterial cell wall, with the PGRS domainavailable on the surface of M. bovis BCG andMtb, and accessible for interactions with hostcomponents (Brennan et al. 2001; Delogu et al.2004; Cascioferro et al. 2011; Zumbo et al.2013; Palucci et al. 2016). Functional dissectionof the PE_PGRS33 protein identified the first140 amino acids of the protein as necessaryand sufficient to mediate translocation andlocalization of the protein to the mycobacterialcell wall (Cascioferro et al. 2007). Indeed,these first 140 amino acids of PE_PGRS33,that contain the PE domain and the GRPLImotif, were used as a delivery system to expressa heterologous protein on the mycobacterialsurface (Sali et al. 2010). Proteomic analysis oflipid phase fractions enriched following TritonX-114 extraction, provided further evidence forthe presence of a number of PE_PGRS proteinsin the mycobacterial cell wall (De Souza andWiker 2011). Green-fluorescent protein (GFP)fusions with PE_PGRS33 and PE_PGRS30 havebeen used to monitor protein localization andpolarization in mycobacteria. The data indicatesthat while the first 140 amino acids are sufficientto drive GFP localization to the mycobacterial

198 G. Delogu et al.

surface, polarization can be affected by othersequences found in the PGRS domain or alsoin the region found between the PE domain andthe GRPLI motif (Chatrath et al. 2011, 2014; DeMaio et al. 2014).

A number of PE_PGRS proteins were shownto have immunomodulatory properties, at leastin some cases, probably due to their ability tointeract with TLR2 (Balaji et al. 2007; Bansalet al. 2010a; Chen et al. 2013; Singh et al. 2008;Zumbo et al. 2013; Palucci et al. 2016). Findingsobtained in M. marinum indicate that PE_PGRSproteins are secreted into the media via an ESX-5 dependent mechanism (Abdallah et al. 2009).This suggests that these proteins, similar to otherPE and PPE proteins, can translocate throughthe mycobacterial membranes. However, whenexperiments were performed in Mtb, secretionof PE_PGRS proteins by ESX-5 was not ob-served (Bottai et al. 2012). This implies thatthere might be significant differences betweenMtb and M. marinum species in the translocationof PE_PGRSs.

The available experimental data and the sig-nificant homology observed between the aminoacid sequences of PE_PGRSs, at least in the first90 amino acids of the N-terminal domain and theGRPLI motif, point toward a shared mechanismof protein translocation among PE_PGRSs. Al-though it remains to be determined whether theGRPLI motif is also required for the anchoring ofthe protein on the mycomembrane, or if it has an-other yet undefined role in translocation. Anotherissue concerns the structure of the PE domainof PE_PGRS and whether, similarly to what hasbeen shown for the PE/PPE couplets (Stronget al. 2006), interaction with another PPE pro-tein to form heterodimers is required to warrantprotein stability as predicted by computationalstudies (Riley et al. 2008). Characterization of thestructure of the PE domain of PE_PGRS proteinsmay also provide a better understanding of howPE_PGRSs are localized to the mycobacterialcell wall and the role of the PE domain in thisprocess.

The polymorphic PGRS domain showsthe utmost level of diversity among theparalogs, with significant differences in size andsequence among the PE_PGRS proteins (see

Chap. 9). However, the lack of any significanthomology with other protein sequences inexisting databases is a major obstacle towardunderstanding the functional role of thisdomain. The partial homology with the domainfound in the EBNA protein of Epstein-BarrVirus, prompted some speculation on theirrole in the evasion of the cytotoxic T cellresponses (Cole et al. 1998; Brennan and Delogu2002), though this hypothesis has not beensupported by experimental evidence. A numberof PE_PGRS proteins have been implicatedin the pathogenesis of TB, primarily withevidence showing that, following expressionof PE_PGRS in saprophytic M. smegmatis,enhanced cytotoxity and inflammation wereobserved in in vitro and in vivo models. Thishas been described for PE_PGRS33 (Brennanet al. 2001; Dheenadhayalan et al. 2006; Balajiet al. 2007; Basu et al. 2007; Zumbo et al. 2013),for PE_PGRS16, for PE_PGRS26 (Singh et al.2008), for PE_PGRS30 (Iantomasi et al. 2012;Chatrath et al. 2014), for PE_PGRS11 (Bansalet al. 2010a) and for PE_PGRS17 (Bansal et al.2010a; Chen et al. 2013). In all these instances,the PGRS domain has been identified as a ligand,although it has not been possible to equate aspecific region or sequence within the PGRSdomain accountable for the observed activity.

Cadieux and colleagues (2011), establishedthat PE_PGRS33 co-localizes to the mitochon-dria of transfected cells, a phenomenon depen-dent on the linker GRPLI region and the PGRSdomain, but not the PE domain. Furthermore,although all constructs localizing to the mito-chondria did induce apoptosis, only the wild-typePE_PGRS33 plus its own PE domain inducedprimary necrosis as well as apoptosis, signifyinga potentially important role for the PE domainin this phenomenon. Considering the importanceof primary necrosis in Mtb dissemination dur-ing natural infection, PE_PGRS33 may play acrucial role in TB pathogenesis. Interestingly,other PE_PGRS proteins tested (PE-PGRS1, -18and -24) did not co-localize to the mitochondriaof transfected cells, suggesting this is a phe-nomenon that may be unique to PE_PGRS33.

Functional and immunological character-ization of PE_PGRS11 provided interesting

10 PE and PPE Genes: A Tale of Conservation and Diversity 199

insights into the multifunctional role of theseproteins during Mtb infections. PE_PGRS11is a 584 amino acid protein, which has ahighly conserved unique domain of 314 aminoacids that encode a functional phosphoglyceratemutase (Chaturvedi et al. 2010) downstream ofa relatively small PGRS domain (�100 aa).The enzyme was properly active when thefull length PE_PGRS11 was expressed inM. smegmatis and provided protection fromoxidative stress. Interestingly, PE_PGRS11 wasalso capable of binding TLR2 inducing anti-apoptotic signals. It has been proposed that thissurface exposed protein is engaged in the rescueof pathogenic Mtb-infected lung epithelial cellsfrom oxidative stresses (Chaturvedi et al. 2010).The short PGRS domain could be responsiblefor the interaction with TLR2, and recent data,obtained with PE_PGRS33, indicates that evena small PGRS region, containing few repeats,can activate the TLR2-dependent entry intomacrophages (Palucci et al. 2016). Hence,PE_PGRS11 is an example of the complexityof the functional potential of PE_PGRSs that canbe seen as proteins with moonlighting properties.It would be of interest to explore whether thismoonlighting property is a common trait for otherPE_PGRSs, and it is expected that investigationof similar PE_PGRS proteins showing a largeunique C-terminal domain (>100 aa in length)downstream of the PGRS (such as PE_PGRS11,PE_PGRS16, -17, -18, -30, -35, and -39) mayhelp address this issue (Delogu et al. 2008).

The unique C-terminal domain found inPE_PGRS30 is highly homologous to the proteinencoded by Rv3812 (which is still namedPE_PGRS62 although it is missing two keyfeatures of PE_PGRSs, that is the GRPLI motifand the GGA-GGN repeats). Interestingly, theunique � 300 amino acid long C-terminaldomain in PE_PGRS30, but not the PGRSdomain, has been shown to be dispensable forthe full virulence of Mtb (Iantomasi et al. 2012).

The functional characterization of the PGRSdomain remains elusive. The identification ofthe sequences within PGRS that interact withTLR2 or mediate other functions are unknown,although they likely include one or more of theGGA-GGN repeats.

Expression of a number of PE_PGRSproteins in M. smegmatis, or overexpression ofPE_PGRSs in Mtb, results in changes in colonymorphology and cell ultrastructure, suggestingsignificant alteration of the mycobacterialcell wall. However, it is not clear whetherparts of the PGRS domain are embedded inthe mycomembrane, or whether the observedchanges in ultrastructure are simply a resultof protein hindrance. The finding that underphysiological conditions, where the PE_PGRS isexpressed under the control of its own promoterin Mtb, the PGRS domain mediates proteinpolarization (De Maio et al. 2014; Chatrathet al. 2014) suggesting that this domain isdirectly interacting with other mycobacterialcomponents found in the mycomembrane or inthe periplasmic space.

It is expected that, despite the significant ho-mology, a specific and unique function for allor most of the PE_PGRSs will be established.Differences in function between PE_PGRSs willlikely be due to specific amino acids sequencesfound in the PGRS domain or regions found atthe C-terminus of the protein. Based on the infor-mation available at this time, the protein specificintercalating sequences, of different sizes andpatterns found between the GGA-GGN repeats,are likely to provide a specific function to eachPE_PGRS.

10.6.2 Genetic Variability and Impacton Virulenceand Pathogenesis

Demonstration that the polymorphic PGRS do-main is available on the mycobacterial surface(Brennan et al. 2001; Banu et al. 2002; Deloguet al. 2004) and that it can be the target of thehost humoral response (Singh et al. 2001; Deloguand Brennan 2001; Cohen et al. 2014) led furthersupport to the hypothesis that PE_PGRSs wereinvolved in antigenic variation. Sequencing ofthe pe_pgrs33 gene in hundreds of different Mtbclinical isolates demonstrated a number of poly-morphisms, mostly occurring in the PGRS region(Talarico et al. 2005, 2007) and characterizedby the presence of single nucleotide polymor-

200 G. Delogu et al.

phisms (SNP)s, insertions or deletions. Interest-ingly, deletion of large sequences in the PGRSregion of pe_pgrs33 resulted in an attenuation ofthe ability of PE_PGRS33 to elicit TNF’ (Basuet al. 2007). Also, alleles with large deletions inthe pgrs region were found in Mtb strains isolatedfrom patients with absence of cavitation in thelung (Talarico et al. 2007). Moreover, Mtb strainsshowing pe_pgrs33 alleles significantly differentfrom the dominant one, were associated with TBmeningitis and negative PPD skin tests in chil-dren (Wang et al. 2011). These studies supportthe notion that variation in PE_PGRS33 couldhave consequences for the clinical features ofTB. A remarkable level of genetic variation wasalso observed in PE_PGRS16 and PE_PGRS26,though the lack of an understanding of the func-tion of these proteins in TB pathogenesis pre-vents a full assessment of the implications ofthese polymorphisms (Talarico et al. 2008).

Whole-genome sequences for many Mtbstrains was not used to investigate sequencevariation in pe_pgrs (and ppe_mptr) genes,since the GC-rich sequences are not reliablyresolved by Next-Generation- Sequencing andoften contain many sequence errors (McEvoyet al. 2012). A more systematic approach wastaken by Copin and colleagues (2014), whoSanger sequenced 27 different pe_pgrs genesin a collection of 94 Mtb clinical isolates whichwere representatives of five phylogeographiclineages, and provided a broad overview on thegenetic variability in this gene family (see Chap.9). While pe_pgrs genes as a group were foundto show more diversity than the rest of the MTBCgenome, great differences in sequence diversityfor different pe_pgrs genes were observed,with some genes under diversifying selectionand others under purifying selection. Thisprompted the authors to conclude that distinctselection pressures act on individual genes.Genetic diversity was mostly observed in thepgrs sequences, with SNPs and indels, that whenresulting in frameshift mutations were apparentlymostly removed by purifying selection. The dataobtained support the hypothesis that membersof the pe_pgrs family have a distinct and non-redundant function and that pe_pgrs genes aremostly conserved in Mtb (Copin et al. 2014).

T cell immune responses play a key role inprotection against TB. It has been assumed thatgenetic variation in Mtb serves to evade or ma-nipulate host immune responses to avoid killing,promote disease and assure transmission (Comaset al. 2010). PE_PGRS proteins are enriched inT cell epitopes that appear to be restricted tothe PE domain, which displays a low level ofpolymorphisms. In contrast, although the PGRSdomain is polymorphic, it harbors only a fewpredicted T cell epitopes, prompting the authorsto exclude that T cell recognition is the drivingforce of sequence diversity in pe_pgrs genes (seeChap. 9) (Copin et al. 2014). Hence, it remains tobe determined what biological forces are drivingthe genetic diversity observed in the few pe_pgrsgenes showing positive selection. Moreover, ofthe genetic variations observed in those genesthat appear under purifying selection, what arethe consequences of Mtb-host interaction andTB pathogenesis? Since most of the PE_PGRSproteins for which we have experimental datahave been shown to interact with TLRs, it canbe postulated that pe_pgrs genetic variation mayhave an impact on the immunomodulatory prop-erties of the Mtb strains.

10.7 Practical Implications

10.7.1 Vaccines

Although one-third of the human populationworld-wide is infected with Mtb, we still do notunderstand what antigens are needed or whatimmune responses should be elicited to makea TB vaccine effective. A number of studiesshowed that PE and PPE proteins are the maintarget of T cell responses in humans and animalsinfected with Mtb, in different stages of infectionand in different disease states (Dillon et al.1999; Delogu and Brennan 2001; Vordermeieret al. 2012). Detailed analysis at the peptidelevel indicates that T cell epitopes are mainlyfound in the conserved N-terminal domainsof these proteins (Lindestam Arlehamn et al.2013; Copin et al. 2014) and that also cross-reactivity exists among the different PE and PPEantigens (Vordermeier et al. 2012). Moreover, T

10 PE and PPE Genes: A Tale of Conservation and Diversity 201

cell responses directed against some of the PEand PPE proteins were shown to elicit protectionin animal models (Skeiky et al. 2000) and PPE18emerged as one of the most promising vaccinecandidates and is currently included in the GSKvaccine Mtb72F (Skeiky et al. 2004). Mtb72fis a polyprotein subunit fusion vaccine thatwhen adjuvanted with AS01B was able to elicitprotection in several animal models, when eitheradministered alone, or in a prime-boost strategyin previously BCG immunized animals (Tsenovaet al. 2006; Reed et al. 2009). Results of phaseI/II clinical trials showed that Mtb72f is safe andimmunogenic in humans (Spertini et al. 2013).Another polyprotein subunit vaccine whichexhibited good protection in animal models isID93, which is composed of four Mtb antigens,one of which is PPE_MPTR42 (Bertholet et al.2010).

The results obtained with these two vaccinesindicates that PPE18 and PPE_MPTR42are immuno-dominant, likely because theycross-react with many PPEs expressed byMtb (Vordermeier et al. 2012). Indeed, miceimmunized with a Mtb attenuated mutant lackingppe25 and pe19, were able to mount significantT cell responses when stimulated with peptidescorresponding to shared epitopes of both PEand PPE proteins, demonstrating cross-reactivityof these antigens (Sayes et al. 2012). Thesedata, together with other experimental evidence,which showed that protective activity in animalmodels is elicited by a number of other PE andPPE proteins (Romano et al. 2008; Parra et al.2006; Vipond et al. 2006; Singh et al. 2013),highlights the vaccine potential of PE and PPEproteins.

As suggested by Comas and colleagues(2010), the immune responses to major antigensof Mtb are evolutionarily conserved to thebenefit of the pathogen. Although the PPE andPE families were not included in that study,subsequent findings that many T cell epitopesin PE and PPE proteins are detected in theconserved regions (Copin et al. 2014), raisesthe same concerns that have been highlighted forother Mtb immune-dominant antigens (Comaset al. 2010).

Several PE and PPE proteins, and primarilyPE_PGRS, PPE_MPTR, PE-unique and PPE-unique, are surface exposed antigens that directlyinteract with host components to modulate innateimmune responses or exert other specific func-tions. Since the genes encoding these proteinsshowed a level of conservation greater than ex-pected, and some of these proteins play a key rolein TB pathogenesis, proteins of these familiesmay represent a suitable target for a humoralresponse. In principle, antibodies specifically di-rected against the surface available domains mayneutralize the activity of these proteins and pro-vide some level of protection. Antibodies di-rected against the PGRS domain of PE_PGRS33were observed in mice and humans with activeTB (Singh et al. 2001; Delogu and Brennan 2001;Cohen et al. 2014), although immunization witha DNA vaccine that elicits antibodies againstthe PGRS domain did not provide a meaningfullevel of protection in mice (Delogu and Brennan2001). It remains to be seen whether the anti-bodies developed following infection, or thoseinduced by the PE_PGRS33 vaccine, are capableof properly binding the critical epitopes involvedin the interaction with host components. A betterfunctional characterization of these proteins is re-quired before ruling out a potential activity of an-tibodies against these surface molecules. PGRS-containing vaccines could represent a promisingtarget of a protective humoral response againstMtb (Delogu et al. 2014).

10.7.2 Drug Targets

The lack of functional information on the roleof PE and PPE proteins have so far impededstudies aimed at targeting these proteins withdrugs. Meszaros and colleagues (2011) suggestedthat PE and PPE should be considered idealcandidates as drug targets because they arespecific for mycobacteria. Also, many of theseproteins are found only in Mtb, and they maybe less hostile to normal microbiota, which areoften affected by the long lasting traditional anti-TB treatment (Wu et al. 2013). Some of thePE and PPE proteins, and primarily PE_PGRS,

202 G. Delogu et al.

were shown to be overrepresented in the lungtissue of infected guinea pigs (Kruh et al. 2010),implying that they may impact TB disease.In addition, these proteins are available onthe surface and in the mycobacterial cell walland are more accessible to drugs compared tomycobacterial cytoplasmic targets, that need topass the highly impermeable mycobacterial cellwall.

Computational studies have proposed that adrug, capable of inhibiting or displacing the hy-drophobic interaction between PE and PPE do-mains, may block important processes requiredfor Mtb virulence (Meszaros et al. 2011). Theyalso indicate that even the disordered domainfound at the C-terminus of many PE and PPEproteins may be a valid drug target. As recentlyproposed (Ahmed et al. 2015), biological andmolecular characterization of these PE and PPEmay be instrumental for identifying potentialdrug targets. These drugs could then block theinteraction between surface available PE and PPEand host receptors such as TLR2, and could beused for treating TB.

10.7.3 Diagnostics

Several studies have commenced to characterizethe cellular and humoral immune responses toPE and PPE proteins found in Mtb infected sub-jects. A main goal in TB diagnosis is the iden-tification of immunological biomarkers that canhelp distinguish between Mtb infected subjects(LTBI), and patients with active TB as well asnormal subjects. In a recent study, stimulationof human whole-blood with PE35 and PPE68,using procedures routinely used for Interferon-” release assays, but where IL-2 was assessed,often discriminated between active and LTBI pa-tients (Pourakbari et al. 2015). Robust CD8 T cellresponses, directed against PE_PGRS33, the PE-unique Rv3812 and PPE46, were also detectedin LTBI subjects. A better immunological char-acterization of these proteins, particularly thoseincluded in the RD regions which are missingin the vaccine strain, BCG (see Chap. 8), maylead to the development of improved Interferon-

” release assays with improved prognostic value.Moreover, serological diagnosis of TB may po-tentially be improved by the use of a numberof PE and PPE proteins for which a robust an-tibody response has been observed, such as forPE_PGRS33, PPE41 (Choudhary et al. 2003),PPE55 (Singh et al. 2005), PPE57 (Zhang et al.2007), and PPE17 (Khan et al. 2008).

10.8 Conclusion

Despite a growing interest in PE and PPE pro-teins since their discovery, their role in Mtbphysiology and pathogenicity is still elusive. Insummary, in the last few years we have learnedthat: (i) PE/PPE couplets are substrates of theESX T7SSs whereby they are secreted or surfaceexposed; (ii) several PE and PPE proteins havethe ability to modulate the host immune systemby directly engaging with TLRs; (iii) PE and PPEproteins are also the main targets of the adaptivehost T cell response; (iv) PE and PPE domainscan serve as shuttles to deliver PGRS and MPTRdomains to the cell surface, or other unique do-mains, some of which possess enzymatic activityas LipY or PE_PGRS11. While the role of thePE and PPE domains has been associated withprotein localization, the biological significance ofthe structural redundancy observed in PE_PGRSsand PPE_PMTRs is still a matter of debate. Like-wise, the consequences of the differential regu-lation of these abundant genes in Mtb during theinfectious process are unclear. A major puzzle isthat, despite the abundance of these genes and thepresence of a degree of genetic polymorphismshigher than the average Mtb genome, most pe andppe genes, including the polymorphic pe_pgrsand ppe_mptr, are under neutral or purifyingselection. This suggests that the correspondingproteins have, despite the presence of repeatmotifs, structural and/or functional constraintsand that they are critical for TB pathogenesis. Abetter understanding of the role and function ofthe PE and PPE proteins will require clarificationof the impact of the contradictory forces of bothconservation and diversity acting on pe and ppegenes.

10 PE and PPE Genes: A Tale of Conservation and Diversity 203

References

Abdallah AM, Verboom T, Weerdenburg EM, Gey vanPittius NC, Mahasha PW, Jimenez C et al (2009) PPEand PE_PGRS proteins of Mycobacterium marinumare transported via the type VII secretion system ESX-5. Mol Microbiol 73:329–340

Adindla S, Guruprasad L (2003) Sequence analysis corre-sponding to the PPE and PE proteins in Mycobacteriumtuberculosis and other genomes. J Biosci 28:169–179

Ahmed A, Das A, Mukhopadhyay S (2015) Immunoreg-ulatory functions and expression patterns of PE/PPEfamily members: roles in pathogenicity and impact onanti-tuberculosis vaccine and drug design. IUBMB Life67:414–427

Akhter Y, Ehebauer MT, Mukhopadhyay S, Hasnain SE(2012) The PE/PPE multigene family codes for viru-lence factors and is a possible source of mycobacterialantigenic variation: perhaps more? Biochimie 94:110–116

Ates LS, Houben EN, Bitter W (2016) Type VII secretion:a highly versatile secretion system. Microbiol Spectr 4

Balaji KN, Goyal G, Narayana Y, Srinivas M, ChaturvediR, Mohammad S (2007) Apoptosis triggered byRv1818c, a PE family gene from Mycobacterium tu-berculosis is regulated by mitochondrial intermediatesin T cells. Microbes Infect 9:271–281

Bansal K, Elluru SR, Narayana Y, Chaturvedi R, PatilSA, Kaveri SV et al (2010a) PE_PGRS antigens ofMycobacterium tuberculosis induce maturation and ac-tivation of human dendritic cells. J Immunol 184:3495–3504

Bansal K, Sinha AY, Ghorpade DS, TogarsimalemathSK, Patil SA, Kaveri SV et al (2010b) Src homology3-interacting domain of Rv1917c of Mycobacteriumtuberculosis induces selective maturation of humandendritic cells by regulating PI3K-MAPK-NF-kappaBsignaling and drives Th2 immune responses. J BiolChem 285:36511–36522

Banu S, Honore N, Saint-Joanis B, Philpott D, PrevostMC, Cole ST (2002) Are the PE-PGRS proteins ofMycobacterium tuberculosis variable surface antigens?Mol Microbiol 44:9–19

Basu S, Pathak SK, Banerjee A, Pathak S, BhattacharyyaA, Yang Z et al (2007) Execution of macrophage apop-tosis by PE_PGRS33 of Mycobacterium tuberculosisis mediated by Toll-like receptor 2-dependent releaseof tumor necrosis factor-alpha. J Biol Chem 282:1039–1050

Bertholet S, Ireton GC, Ordway DJ, Windish HP, Pine SO,Kahn M et al (2010) A defined tuberculosis vaccinecandidate boosts BCG and protects against multidrug-resistant Mycobacterium tuberculosis. Sci Transl Med2:53ra74

Bhat KH, Ahmed A, Kumar S, Sharma P, MukhopadhyayS (2012) Role of PPE18 protein in intracellular survivaland pathogenicity of Mycobacterium tuberculosis inmice. PLoS One 7:e52601

Bottai D, Di LM, Majlessi L, Frigui W, Simeone R, SayesF et al (2012) Disruption of the ESX-5 system ofMycobacterium tuberculosis causes loss of PPE proteinsecretion, reduction of cell wall integrity and strongattenuation. Mol Microbiol 83:1209

Brennan MJ, Delogu G (2002) The PE multigene family:a ‘molecular mantra’ for mycobacteria. Trends Micro-biol 10:246–249

Brennan MJ, Delogu G, Chen Y, Bardarov S, Kriakov J,Alavi M, Jacobs WR Jr (2001) Evidence that mycobac-terial PE_PGRS proteins are cell surface constituentsthat influence interactions with other cells. Infect Im-mun 69:7326–7333

Cadieux N, Parra M, Cohen H, Maric D, Morris SL,Brennan MJ (2011) Induction of cell death after lo-calization to the host cell mitochondria by the My-cobacterium tuberculosis PE_PGRS33 protein. Micro-biology 157:793–804

Cascioferro A, Delogu G, Colone M, Sali M, StringaroA, Arancia G et al (2007) PE is a functional domainresponsible for protein translocation and localizationon mycobacterial cell wall. Mol Microbiol 66:1536–1547

Cascioferro A, Daleke MH, Ventura M, Dona V, DeloguG, Palu G et al (2011) Functional dissection of the PEdomain responsible for translocation of PE_PGRS33across the mycobacterial cell wall. PLoS One 6:e27713

Chakhaiyar P, Nagalakshmi Y, Aruna B, Murthy KJ,Katoch VM, Hasnain SE (2004) Regions of highantigenicity within the hypothetical PPE major poly-morphic tandem repeat open-reading frame, Rv2608,show a differential humoral response and a low Tcell response in various categories of patients withtuberculosis. J Infect Dis 190:1237–1244

Chatrath S, Gupta VK, Dixit A, Garg LC (2011) TheRv1651c-encoded PE-PGRS30 protein expressed inMycobacterium smegmatis exhibits polar localizationand modulates its growth profile. FEMS Microbiol Lett322:194–199

Chatrath S, Gupta VK, Garg LC (2014) The PGRS domainis responsible for translocation of PE_PGRS30 to cellpoles while the PE and the C-terminal domains localizeit to the cell wall. FEBS Lett 588:990–994

Chaturvedi R, Bansal K, Narayana Y, Kapoor N, SukumarN, Togarsimalemath SK et al (2010) The multifunc-tional PE_PGRS11 protein from Mycobacterium tuber-culosis plays a role in regulating resistance to oxidativestress. J Biol Chem 285:30389–30403

Chen T, Zhao Q, Li W, Xie J (2013) Mycobacterium tu-berculosis PE_PGRS17 promotes the death of host celland cytokines secretion via Erk kinase accompanyingwith enhanced survival of recombinant Mycobacteriumsmegmatis. J Interf Cytokine Res 33:452–458

Choudhary RK, Mukhopadhyay S, Chakhaiyar P, SharmaN, Murthy KJ, Katoch VM, Hasnain SE (2003) PPEantigen Rv2430c of Mycobacterium tuberculosis in-duces a strong B-cell response. Infect Immun 71:6338–6343

204 G. Delogu et al.

Cohen I, Parada C, Acosta-Gio E, Espitia C (2014) ThePGRS Domain from PE_PGRS33 of Mycobacteriumtuberculosis is Target of Humoral Immune Response inMice and Humans. Front Immunol 5:236

Cole ST, Brosch R, Parkhill J, Garnier T, Churcher C, Har-ris D et al (1998) Deciphering the biology of Mycobac-terium tuberculosis from the complete genome se-quence [see comments] [published erratum appears inNature 1998 Nov 12;396(6707):190]. Nature 393:537–544

Comas I, Chakravartti J, Small PM, Galagan J, NiemannS, Kremer K et al (2010) Human T cell epitopes ofMycobacterium tuberculosis are evolutionarily hyper-conserved. Nat Genet 42:498–503

Copin R, Coscolla M, Seiffert SN, Bothamley G, Suther-land J, Mbayo G et al (2014) Sequence diversity in thepe_pgrs genes of Mycobacterium tuberculosis is inde-pendent of human T cell recognition. MBio 5:e00960–e00913

Daim S, Kawamura I, Tsuchiya K, Hara H, KurenumaT, Shen Y et al (2011) Expression of the Mycobac-terium tuberculosis PPE37 protein in Mycobacteriumsmegmatis induces low tumour necrosis factor alphaand interleukin 6 production in murine macrophages.J Med Microbiol 60:582–591

Daleke MH, Cascioferro A, de Pundar K, Ummels R,Abdallah AM, van der Wel N et al (2011) ConservedPE and PPE protein domains target LipY lipases ofpathogenic mycobacteria to the cell surface via ESX-5. J Biol Chem

De Maio F, Maulucci G, Minerva M, Anoosheh S, PalucciI, Iantomasi R et al (2014) Impact of protein domainson PE_PGRS30 polar localization in Mycobacteria.PLoS One 9:e112482

De Souza GA, Wiker HG (2011) A proteomic view ofmycobacteria. Proteomics 11:3118–3127

Delogu, G, Brennan, MJ (2001) Comparative ImmuneResponse to PE and PE_PGRS Antigens of Mycobac-terium tuberculosis. Infect Immun 69: 5606–5611

Delogu G, Pusceddu C, Bua A, Fadda G, Brennan MJ,Zanetti S (2004) Rv1818c-encoded PE_PGRS pro-tein of Mycobacterium tuberculosis is surface exposedand influences bacterial cell structure. Mol Microbiol52:725–733

Delogu G, Sanguinetti M, Pusceddu C, Bua A, BrennanMJ, Zanetti S, Fadda G (2006) PE_PGRS proteins aredifferentially expressed by Mycobacterium tuberculo-sis in host tissues. Microbes Infect 8:2061–2067

Delogu G, Cole ST, Brosch R (2008) The PE and PPEprotein families of mycobacterium tuberculosis. In:Kaufmann SH, Rubin E (eds) Handbook of tuberculo-sis. Wiley-VCH Verlag GmbH%Co. KGaA, Weinheim,pp 131–150

Delogu, G, Manganelli, R, Brennan, MJ (2014) Criticalresearch concepts in tuberculosis vaccine development.Clin Microbiol Infect 20 Suppl 5: 59–65

Dheenadhayalan V, Delogu G, Brennan MJ (2006) Ex-pression of the PE_PGRS 33 protein in Mycobac-terium smegmatis triggers necrosis in macrophagesand enhanced mycobacterial survival. Microbes Infect8:262–272

Dillon DC, Alderson MR, Day CH, Lewinsohn DM, ColerR, Bement T et al (1999) Molecular characterizationand human T-cell responses to a member of a novelMycobacterium tuberculosis mtb39 gene family. InfectImmun 67:2941–2950

Dona V, Ventura M, Sali M, Cascioferro A, Provvedi R,Palu G et al (2013) The PPE domain of PPE17 isresponsible for its surface localization and can be usedto express heterologous proteins on the mycobacterialsurface. PLoS One 8:e57517

Dong D, Wang D, Li M, Wang H, Yu J, Wang C et al(2012) PPE38 modulates the innate immune responseand is required for Mycobacterium marinum virulence.Infect Immun 80:43–54

Ekiert DC, Cox JS (2014) Structure of a PE-PPE-EspGcomplex from Mycobacterium tuberculosis revealsmolecular specificity of ESX protein secretion. ProcNatl Acad Sci U S A 111:14758–14763

Fishbein S, van Wyk N, Warren RM, Sampson SL (2015)Phylogeny to function: PE/PPE protein evolution andimpact on Mycobacterium tuberculosis pathogenicity.Mol Microbiol 96:901–916

Gao Q, Kripke KE, Saldanha AJ, Yan W, Holmes S,Small PM (2005) Gene expression diversity amongMycobacterium tuberculosis clinical isolates. Microbi-ology 151:5–14

Gey van Pittius NC, Sampson SL, Lee H, Kim Y, vanHelden PD, Warren RM (2006) Evolution and ex-pansion of the Mycobacterium tuberculosis PE andPPE multigene families and their association with theduplication of the ESAT-6 (esx) gene cluster regions.BMC Evol Biol 6:95

Golby P, Hatch KA, Bacon J, Cooney R, Riley P, Allnutt Jet al (2007) Comparative transcriptomics reveals keygene expression differences between the human andbovine pathogens of the Mycobacterium tuberculosiscomplex. Microbiology 153:3323–3336

Gold B, Rodriguez GM, Marras SAE, Pentecost M, SmithI (2001) The Mycobacterium tuberculosis IdeR is adual functional regulator that controls transcription ofgenes involved in iron acquisition, iron storage andsurvival in macrophages. Mol Microbiol 42:851–865

Goldstone RM, Goonesekera SD, Bloom BR, SampsonSL (2009) The transcriptional regulator Rv0485 mod-ulates the expression of a pe and ppe gene pair andis required for Mycobacterium tuberculosis virulence.Infect Immun 77:4654–4667

Hebert AM, Talarico S, Yang D, Durmaz R, Marrs CF,Zhang L et al (2007) DNA polymorphisms in the pepAand PPE18 genes among clinical strains of Mycobac-terium tuberculosis: implications for vaccine efficacy.Infect Immun 75:5798–5805

Homolka S, Ubben T, Niemann S (2016) High sequencevariability of the ppE18 gene of clinical Mycobac-terium tuberculosis complex strains potentially impactseffectivity of vaccine candidate M72/AS01E. PLoSOne 11:e0152200

Iantomasi R, Sali M, Cascioferro A, Palucci I, ZumboA, Soldini S et al (2012) PE_PGRS30 is required forthe full virulence of Mycobacterium tuberculosis. CellMicrobiol 14:356–367

10 PE and PPE Genes: A Tale of Conservation and Diversity 205

Karboul A, Gey van Pittius NC, Namouchi A, VincentV, Sola C, Rastogi N et al (2006) Insights into theevolutionary history of tubercle bacilli as disclosed bygenetic rearrangements within a PE_PGRS duplicatedgene pair. BMC Evol Biol 6:107

Karboul A, Mazza A, Gey van Pittius NC, Ho JL,Brousseau R, Mardassi H (2008) Frequent homologousrecombination events in Mycobacterium tuberculosisPE/PPE multigene families: potential role in antigenicvariability. J Bacteriol 190:7838–7846

Khan N, Alam K, Nair S, Valluri VL, Murthy KJ,Mukhopadhyay S (2008) Association of strong im-mune responses to PPE protein Rv1168c with activetuberculosis. Clin Vaccine Immunol 15:974–980

Korotkova N, Freire D, Phan TH, Ummels R, Creek-more CC, Evans TJ et al (2014) Structure of theMycobacterium tuberculosis type VII secretion systemchaperone EspG5 in complex with PE25-PPE41 dimer.Mol Microbiol 94:367–382

Kruh NA, Troudt J, Izzo A, Prenni J, Dobos KM (2010)Portrait of a pathogen: the Mycobacterium tuberculosisproteome in vivo. PLoS One 5:e13938

Lee JH, Karakousis PC, Bishai WR (2008) Roles of SigBand SigF in the Mycobacterium tuberculosis sigmafactor network. J Bacteriol 190:699–707

Lindestam Arlehamn CS, Gerasimova A, Mele F, Hender-son R, Swann J, Greenbaum JA et al (2013) MemoryT cells in latent Mycobacterium tuberculosis infectionare directed against three antigenic islands and largelycontained in a CXCR3CCCR6C Th1 subset. PLoSPathog 9:e1003130

McEvoy CR, van Helden PD, Warren RM, Gey van PittiusNC (2009) Evidence for a rapid rate of molecularevolution at the hypervariable and immunogenic My-cobacterium tuberculosis PPE38 gene region. BMCEvol Biol 9:237

McEvoy CR, Cloete R, Muller B, Schurch AC, van HeldenPD, Gagneux S et al (2012) Comparative analysis ofMycobacterium tuberculosis pe and ppe genes revealshigh sequence variation and an apparent absence ofselective constraints. PLoS One 7:e30593

Meszaros B, Toth J, Vertessy BG, Dosztanyi Z, Simon I(2011) Proteins with complex architecture as potentialtargets for drug design: a case study of Mycobacteriumtuberculosis. PLoS Comput Biol 7:e1002118

Mishra KC, de Chastellier C, Narayana Y, Bifani P, BrownAK, Besra GS et al (2008) Functional role of the PEdomain and immunogenicity of the Mycobacteriumtuberculosis triacylglycerol hydrolase LipY. Infect Im-mun 76:127–140

Mohareer K, Tundup S, Hasnain SE (2011) Tran-scriptional regulation of Mycobacterium tuberculosisPE/PPE genes: a molecular switch to virulence? J MolMicrobiol Biotechnol 21:97–109

Mortier MC, Jongert E, Mettens P, Ruelle JL (2015)Sequence conservation analysis and in silico humanleukocyte antigen-peptide binding predictions for theMtb72F and M72 tuberculosis candidate vaccine anti-gens. BMC Immunol 16:63

Mukhopadhyay, S, Balaji, KN (2011) The PE and PPEproteins of Mycobacterium tuberculosis. Tuberculosis(Edinb) 91: 441–447

Nair S, Ramaswamy PA, Ghosh S, Joshi DC, Pathak N,Siddiqui I et al (2009) The PPE18 of Mycobacteriumtuberculosis interacts with TLR2 and activates IL-10induction in macrophage. J Immunol 183:6269–6281

Namouchi A, Karboul A, Fabre M, Gutierrez MC, Mar-dassi H (2013) Evolution of smooth tubercle Bacilli PEand PE_PGRS genes: evidence for a prominent role ofrecombination and imprint of positive selection. PLoSOne 8:e64718

Palucci I, Camassa S, Cascioferro A, Sali M, AnooshehS, Zumbo A et al (2016) PE_PGRS33 Con-tributes to Mycobacterium tuberculosis Entry inMacrophages through Interaction with TLR2. PLoSOne 11:e0150800

Parra M, Cadieux N, Pickett T, Dheenadhayalan V, Bren-nan MJ (2006) A PE protein expressed by Mycobac-terium avium is an effective T-cell immunogen. InfectImmun 74:786–789

Poulet S, Cole ST (1995) Characterization of the highlyabundant polymorphic GC-rich-repetitive sequence(PGRS) present in Mycobacterium tuberculosis. ArchMicrobiol 163:87–95

Pourakbari B, Mamishi S, Marjani M, Rasulinejad M,Mariotti S, Mahmoudi S (2015) Novel T-cell assaysfor the discrimination of active and latent tuberculosisinfection: the diagnostic value of PPE family. MolDiagn Ther 19:309–316

Pym AS, Brodin P, Brosch R, Huerre M, Cole ST (2002)Loss of RD1 contributed to the attenuation of the livetuberculosis vaccines Mycobacterium bovis BCG andMycobacterium microti. Mol Microbiol 46:709–717

Rachman H, Strong M, Ulrichs T, Grode L, SchuchhardtJ, Mollenkopf H et al (2006) Unique transcriptomesignature of Mycobacterium tuberculosis in pulmonarytuberculosis. Infect Immun 74:1233–1242

Ramakrishnan L, Federspiel NA, Falkow S (2000)Granuloma-Specific Expression of MycobacteriumVirulence Proteins from the Glycine-Rich PE-PGRSFamily. Science 288:1436–1439

Raman S, Hazra R, Dascher CC, Husson RN (2004) Tran-scription regulation by the Mycobacterium tuberculosisalternative sigma factor SigD and its role in virulence.J Bacteriol 186:6605–6616

Reed SG, Coler RN, Dalemans W, Tan EV, DeLa CruzEC, Basaraba RJ et al (2009) Defined tuberculosisvaccine, Mtb72F/AS02A, evidence of protection incynomolgus monkeys. Proc Natl Acad Sci U S A 106:2301–2306

Rehren G, Walters S, Fontan P, Smith I, Zarraga AM(2007) Differential gene expression between Mycobac-terium bovis and Mycobacterium tuberculosis. Tuber-culosis (Edinb) 87:347–359

Riley R, Pellegrini M, Eisenberg D (2008) Identifyingcognate binding pairs among a large set of paralogs:the case of PE/PPE proteins of Mycobacterium tuber-culosis. PLoS Comput Biol 4:e1000174

206 G. Delogu et al.

Rodriguez GM, Voskuil MI, Gold B, Schoolnik GK,Smith I (2002) ideR, An essential gene in mycobac-terium tuberculosis: role of IdeR in iron-dependentgene expression, iron metabolism, and oxidative stressresponse. Infect Immun 70:3371–3381

Romano M, Rindi L, Korf H, Bonanni D, Adnet PY,Jurion F et al (2008) Immunogenicity and protectiveefficacy of tuberculosis subunit vaccines expressingPPE44 (Rv2770c). Vaccine 26:6053–6063

Sali M, Di SG, Cascioferro A, Zumbo A, Nicolo C, DonaV et al (2010) Surface expression of MPT64 as afusion with the PE domain of PE_PGRS33 enhancesMycobacterium bovis BCG protective activity againstMycobacterium tuberculosis in mice. Infect Immun78:5202–5213

Sassi M, Drancourt M (2014) Genome analysis revealsthree genomospecies in Mycobacterium abscessus.BMC Genomics 15:359

Sayes F, Sun L, Di LM, Simeone R, Degaiffier N, Fi-ette L et al (2012) Strong immunogenicity and cross-reactivity of Mycobacterium tuberculosis ESX-5 typeVII secretion: encoded PE-PPE proteins predicts vac-cine potential. Cell Host Microbe 11:352–363

Serafini A, Pisu D, Palu G, Rodriguez GM, ManganelliR (2013) The ESX-3 secretion system is necessary foriron and zinc homeostasis in Mycobacterium tubercu-losis. PLoS One 8:e78351

Simeone R, Bottai D, Brosch R (2009) ESX/type VIIsecretion systems and their role in host-pathogen in-teraction. Curr Opin Microbiol 12:4–10

Singh KK, Zhang X, Patibandla AS, Chien P Jr, LaalS (2001) Antigens of Mycobacterium tuberculosis ex-pressed during preclinical tuberculosis: serological im-munodominance of proteins with repetitive amino acidsequences. Infect Immun 69:4185–4191

Singh KK, Dong Y, Patibandla SA, McMurray DN, AroraVK, Laal S (2005) Immunogenicity of the Mycobac-terium tuberculosis PPE55 (Rv3347c) protein dur-ing incipient and clinical tuberculosis. Infect Immun73:5004–5014

Singh, PP, Parra, M, Cadieux, N, Brennan, MJ (2008) Acomparative study of host response to three Mycobac-terium tuberculosis PE_PGRS proteins. Microbiology154: 3469–3479

Singh SK, Tripathi DK, Singh PK, Sharma S, Srivas-tava KK (2013) Protective and survival efficacies ofRv0160c protein in murine model of Mycobacteriumtuberculosis. Appl Microbiol Biotechnol 97:5825–5837

Singh, P, Rao, RN, Reddy, JR, Prasad, R, Kotturu, SK,Ghosh, S, Mukhopadhyay, S (2016) PE11, a PE/PPEfamily protein of Mycobacterium tuberculosis is in-volved in cell wall remodeling and virulence. Sci Rep6: 21624

Skeiky YA, Ovendale PJ, Jen S, Alderson MR, Dillon DC,Smith S et al (2000) T cell expression cloning of a My-cobacterium tuberculosis gene encoding a protectiveantigen associated with the early control of infection.J Immunol 165:7140–7149

Skeiky YA, Alderson MR, Ovendale PJ, Guderian JA,Brandt L, Dillon DC et al (2004) Differential immuneresponses and protective efficacy induced by compo-nents of a tuberculosis polyprotein vaccine, Mtb72F,delivered as naked DNA or recombinant protein. JImmunol 172:7618–7628

Solans L, Gonzalo-Asensio J, Sala C, Benjak A, Up-lekar S, Rougemont J et al (2014) The PhoP-dependent ncRNA Mcr7 modulates the TAT secretionsystem in Mycobacterium tuberculosis. PLoS Pathog10:e1004183

Soldini S, Palucci I, Zumbo A, Sali M, Ria F, ManganelliR et al (2011) PPE_MPTR genes are differentiallyexpressed by Mycobacterium tuberculosis in vivo. Tu-berculosis (Edinb)

Spertini F, Audran R, Lurati F, Ofori-Anyinam O, ZyssetF, Vandepapeliere P et al (2013) The candidate tuber-culosis vaccine Mtb72F/AS02 in PPD positive adults:a randomized controlled phase I/II study. Tuberculosis(Edinb) 93:179–188

Stewart GR, Patel J, Robertson BD, Rae A, Young DB(2005) Mycobacterial mutants with defective control ofphagosomal acidification. PLoS Pathog 1:269–278

Strong M, Sawaya MR, Wang S, Phillips M, Cascio D,Eisenberg D (2006) Toward the structural genomicsof complexes: crystal structure of a PE/PPE proteincomplex from Mycobacterium tuberculosis. Proc NatlAcad Sci U S A 103:8060–8065

Supply P, Marceau M, Mangenot S, Roche D, Rouanet C,Khanna V et al (2013) Genomic analysis of smoothtubercle bacilli provides insights into ancestry andpathoadaptation of Mycobacterium tuberculosis. NatGenet 45:172–179

Talarico S, Cave MD, Marrs CF, Foxman B, Zhang L,Yang Z (2005) Variation of the Mycobacterium tuber-culosis PE_PGRS 33 gene among clinical isolates. JClin Microbiol 43:4954–4960

Talarico S, Cave MD, Foxman B, Marrs CF, Zhang L,Bates JH, Yang Z (2007) Association of Mycobac-terium tuberculosis PE PGRS33 polymorphism withclinical and epidemiological characteristics. Tubercu-losis (Edinb) 87:338–346

Talarico S, Zhang L, Marrs CF, Foxman B, Cave MD,Brennan MJ, Yang Z (2008) Mycobacterium tubercu-losis PE_PGRS16 and PE_PGRS26 genetic polymor-phism among clinical isolates. Tuberculosis (Edinb)88:283–294

Tiwari B, Ramakrishnan UM, Raghunand TR (2015)The Mycobacterium tuberculosis protein pair PE9(Rv1088)-PE10 (Rv1089) forms heterodimers and in-duces macrophage apoptosis through Toll-like receptor4. Cell Microbiol 17:1653–1669

Tsenova L, Harbacheuski R, Moreira AL, Ellison E,Dalemans W, Alderson MR et al (2006) Evaluation ofthe Mtb72F polyprotein vaccine in a rabbit model oftuberculous meningitis. Infect Immun 74:2392–2401

Tufariello JM, Chapman JR, Kerantzas CA, Wong KW,Vilcheze C, Jones CM et al (2016) Separable roles forMycobacterium tuberculosis ESX-3 effectors in iron

10 PE and PPE Genes: A Tale of Conservation and Diversity 207

acquisition and virulence. Proc Natl Acad Sci U S A113:E348–E357

Tundup S, Akhter Y, Thiagarajan D, Hasnain SE (2006)Clusters of PE and PPE genes of Mycobacteriumtuberculosis are organized in operons: evidence thatPE Rv2431c is co-transcribed with PPE Rv2430c andtheir gene products interact with each other. FEBS Lett580:1285–1293

Vipond J, Vipond R, Allen-Vercoe E, Clark SO, HatchGJ, Gooch KE et al (2006) Selection of novel TB vac-cine candidates and their evaluation as DNA vaccinesagainst aerosol challenge. Vaccine 24:6340–6350

Vordermeier HM, Hewinson RG, Wilkinson RJ, Wilkin-son KA, Gideon HP, Young DB, Sampson SL (2012)Conserved immune recognition hierarchy of mycobac-terial PE/PPE proteins during infection in natural hosts.PLoS One 7:e40890

Voskuil MI, Schnappinger D, Rutherford R, Liu Y,Schoolnik GK (2004) Regulation of the Mycobac-terium tuberculosis PE/PPE genes. Tuberculosis (Ed-inb) 84:256–262

Wang J, Huang Y, Zhang A, Zhu C, Yang Z, Xu H (2011)DNA polymorphism of Mycobacterium tuberculosisPE_PGRS33 gene among clinical isolates of pediatricTB patients and its associations with clinical presenta-tion. Tuberculosis (Edinb) 91:287–292

Wu J, Liu W, He L, Huang F, Chen J, Cui P et al(2013) Sputum microbiota associated with new, re-current and treatment failure tuberculosis. PLoS One8:e83445

Zhang H, Wang J, Lei J, Zhang M, Yang Y, Chen Y, WangH (2007) PPE protein (Rv3425) from DNA segmentRD11 of Mycobacterium tuberculosis: a potential B-cell antigen used for serological diagnosis to distin-guish vaccinated controls from tuberculosis patients.Clin Microbiol Infect 13:139–145

Zumbo A, Palucci I, Cascioferro A, Sali M,Ventura M, D’Alfonso P et al (2013) Functionaldissection of protein domains involved in theimmunomodulatory properties of PE_PGRS33of Mycobacterium tuberculosis. Pathog Dis 69:232–239