How the Knowledge of Interactions between Meningococcus and the Human Immune System Has Been Used to Prepare Effective Neisseria meningitidis Vaccines

In the last decades, tremendous advancement in dissecting the mechanisms of pathogenicity of Neisseria meningitidis at a molecular level has been achieved, exploiting converging approaches of different disciplines, ranging from pathology to microbiology, immunology, and omics sciences (such as genomics and proteomics). Here, we review the molecular biology of the infectious agent and, in particular, its interactions with the immune system, focusing on both the innate and the adaptive responses. Meningococci exploit different mechanisms and complex machineries in order to subvert the immune system and to avoid being killed. Capsular polysaccharide and lipooligosaccharide glycan composition, in particular, play a major role in circumventing immune response. The understanding of these mechanisms has opened new horizons in the field of vaccinology. Nowadays different licensed meningococcal vaccines are available and used: conjugate meningococcal C vaccines, tetravalent conjugate vaccines, an affordable conjugate vaccine against the N. menigitidis serogroup A, and universal vaccines based on multiple antigens each one with a different and peculiar function against meningococcal group B strains.


Introduction
The immune system protects humans from attack by microorganisms such as bacteria, viruses, protozoa, fungi, parasites, and organisms such as helminths. The skin is the first barrier and its protective action is enhanced by bodily secretions, such as sweat and sebum, which exert a broad antimicrobial activity [1,2]. The mucous membranes are protected by external and internal secretions, such as tears, saliva, and mucus, which contain molecules that can neutralize bacteria. Tissues such as the skin and mucous membranes are populated by immune cells, which can act against the microorganisms that circumvent the first physical and biochemical barriers.
The immune system is very complex and its defensive response is subdivided into innate and adaptive responses [3]. The innate response triggers an immediate, nonspecific, general action and is activated by typical signs of infection.
The adaptive response is able to develop a highly specific, extremely accurate action, which is stored in the so-called immune memory.
This paper provides an overview of the interaction between the immune system and Gram-negative bacteria with particular reference to Neisseria meningitidis in the perspective of developing new vaccines against this pathogen.

Outer Membrane Components.
Over thousands of years, bacteria have developed several mechanisms whereby they can circumvent the immune system. Specifically, Gramnegative bacteria possess a complex of envelopes, which allow the selective passage of nutrients into the cell and the excretion of metabolic waste outside. Structurally, Gramnegative bacteria possess an outer membrane (OM), which, mitogen-activated protein kinase (MAPK) and the transcription of NF-B factor. A different set of NLRs helps to activate caspase-1 and the consequent assembly of inflammasomes [14].
Many bacterial components are able to stimulate the adaptive human immune response. Porins can activate the translation of NF-B in the nucleus of B and DCs, while class I Pilin E induces highly specific antibodies (Abs) and class II induces cross-reacting Abs. Furthermore, complement cascade activation, as well as particularly C3b activation, opsonizes antigens, thereby enabling APCs to activate the adaptive response.

Meningococcal Genome.
Meningococci have developed several "immunoescape" strategies [23], the molecular bases of which can be understood by taking into account the nature of the Neisserial genome. Progress in the field of molecular biology and the introduction of high-throughput technologies (HTTs) have tremendously advanced our understanding of the complexity of the Neisserial machinery. By using sophisticated approaches such as whole-genome sequencing (WGS) and microarrays, functional genomics investigations have uncovered the mechanisms that facilitate or hinder N. meningitidis growth, colonization, and invasion and have helped to explain its extraordinary intrastrain variation and adaptation to the environment. Other techniques, such as genome-wide association studies (GWAS), have shed light on the pathogen-host interaction and the host's susceptibility to the microbe. Genomics and postgenomics have not only increased our knowledge of the biology and pathogenesis of N. meningitidis but proved to be extremely useful in discovering candidate antigens and in developing effective new vaccines [24].
Being a naturally competent pathogen, N. meningitidis has a highly dynamic, plastic, and flexible genome with a size range of only more than 2,000 kilobases [25]. This genome differs from other microbial genomes in that it lacks some of the typical two-component systems and sigma factors [26]. Despite being relatively small and compact, it has elaborated a variety of mechanisms that contribute to explaining its high adaptability both to host and to environment. Meningococcus is usually polyploid, containing up to 2-5 genomespolyploidy being a sign of virulence-while N. lactamica is monoploid [27]. Neisserial pathogenicity is intrinsically polygenic [28] and is given by a variety of different pathogenic Journal of Immunology Research 3 islands (PAIs or genomic islands, GEIs), including gonococcal genetic islands (GGIs) [29] and a recently discovered meningococcal disease-associated (MDA) island [30].
The nature of the Neisserial chromosome and the presence of extrachromosomal material contribute to explaining an important immunoescape strategy, known as structural or antigenic variation, which consists of camouflage of the Neisserial repertoire expressed. Basically, it can involve horizontal or lateral gene transfer (HGT/LGT) (mainly via transformation and, to a lesser extent, via conjugation and phage transduction) and allelic exchange/rearrangement of genes or gene portions taken up from the environment ( Table 1). In addition, as its genome contains multiple copies of certain genes, for example, opacity factor proteins and pilins [31], homologous intragenic recombination also results in frequent surface structural variation.
Moreover, the pathogen hosts a number of prophages, from the Mu-related family to the phage l-related group and the family of filamentous M13-like phages [27,32]. The most widely studied sites for phage integration are known as duplicated repeat sequence 3 (dRS3) [26], which belong to the family of Neisserial intergenic mosaic elements (NIMEs). Plasmids, such as pJS-A and pJS-B, also play an important role [33].
Another surface modulation occurs via phase variation, a process involving the modulation of gene expression via a mechanism of on/off switching (transition from an expressed state of the gene to an unexpressed one or vice versa). Besides this kind of "functional" phase variation, Neisseria can also undergo a "structural" switch, namely, a transition between two forms of a gene product. The genes, which are involved in this strategy, are termed "contingency genes" [26] and can be coupled and interlinked in structures called phasevarions (phase-variable regulons) [34], which have a regulatory function. Phase variation includes a variety of sophisticated mechanisms [35], such as slipped strand mispairing (SSM) [36], microsatellite instability [37], and reversible insertion of minimal mobile elements (MMEs) [38]. Therefore, these mechanisms can involve single nucleotides (homopolymeric repeats) or complex nucleotidic structures (short tandem repeats), occurring either upstream of a gene in the promoter or within an open reading frame (ORF)/coding sequence (CDS). Changes upstream of a gene result in modulation of its transcriptional efficiency and therefore of its final protein concentration. This is, for example, the case of Opc, porin A (porA), and fetA genes. Alterations within a gene, which insert de novo stop codons, alter the full translation of the gene. An example of this mechanism is provided by the opa genes and the genes coding for adhesins, such as nadA [39]. Phase variation of opa genes has been extensively characterized: they occur in four distinct copies and code for similar, but not identical, proteins. Phase variation can thus involve one copy or another, independently of each other, and can result in eleven variants. In this case, phase variation is therefore equivalent to antigenic variation. Besides opacity factor proteins, phase variation can involve up to hundreds of genes [40]: from the genes coding for pilins [41] or for proteins involved in genome maintenance and DNA repair [42,43] to genes encoding proteins involved in the cell cycle control and regulation [44], autotransporters [45,46], or enzymes like the pilin phosphorylcholine transferase pptA [47] or the glycosylase mutY [48], among others [49,50]; the reader is referred to Table 2, which provides a more detailed overview of the phase-variable genes. Moreover, new mechanisms leading to phase variations have been discovered [51].
It is worth noting that the number of genes involved in phase variation is enormously greater than for any other pathogen studies so far [58]. Some genes are "phasotypes"; that is to say, they play a role in carriage and are downregulated, favouring host persistence [59].
As already mentioned, in some cases antigenic/structural variation and phase variation, albeit conceptually two distinct mechanisms, cooperate in increasing the genetic complexity of the Neisserial genome. Antigenic variation of LPS, for example, can derive from phase variation of one or more enzymes involved in the synthesis of the oligosaccharide chain by SSM, or by modification of LPS, for example, by glycosylation [60][61][62], sialylation [63,64], or acetylation [65,66], which, moreover, confer resistance to neutrophilmediated killing.
Thus, both antigenic and phase variations concur in enabling Neisseria to evade the immune system [26,27].

Meningococcal
Capsule. LPS and themeningococcal capsule (CP) are the two major virulence factors of N. meningitidis. Specifically, the capsule displays a large variability of surface antigens, on the basis of which 13 different N. meningitidis serogroups have been identified. The CP contributes in an important way to the camouflage of the microorganism, which thus can better circumvent the immune system's defences. The clearest expression of this phenomenon is given by the molecular mimicry [67]. This can be seen in the nature of the polysaccharide CP of serogroup B meningococcus, a homopolymer of 2-8-linked sialic acid, which is identical to a neural cell adhesion molecule, NCAM-1 [68]. Moreover, lacto-N-neotetraose (L-NNT) in the lipopolysaccharide of virulent strains is similar to an antigen expressed on red blood cells [69][70][71][72][73]. Further mechanisms of molecular mimicry have been recently discovered and described [74].
During the first phase of infection, meningococcus has to avoid the surface defences of the nasopharynx, such as the peptides secreted at the mucosal surface [90] and IgA secretory Abs [90,196]. To this end, the meningococcus can aggregate into clusters and produce abundant OM vesicles (OMVs), thus managing to hide its surface antigens and to deflect the action of the surface defences from the bacterial cell [196]. In addition, the CP protects Neisseria from

Molecular decoy
FprB has an antigenic subdomain for binding antibodies, which is not essential for the functioning of the autotransporter; it also blebs with OMPs and LPS/LOS distract the immune system, directing the response away from the microbe [94] Immunotype switch LPS immunotype switches from L3 to L8/L1 by lgtA, lgtC phase variation LOS immunotype can contribute to immunoescape [95,96] Phages and prophages The pathogen hosts a number of prophages, from the Mu-related family to the phage l-related group and the family of filamentous M13-like phages [25,30,89] Phase variation High-frequency reversible changes can occur in the length of SSRs (of capsule, LOS, opacity factor, porin, adhesin, invasin, autotransporter, haemoglobin receptor, DNA mismatch repair, and pilin genes, termed as contingency genes and organized in modules called phasevarions) Other repeat sequences can be REP2, CRs, CREEs, and NIMEs Transposon-like elements can play a role Phase variation mediates resistance to antibiotics Phase variation mediates carriage persistence [50,52,59] Pilin conversion and modification Pilin is posttranslationally modified by addition of a glycan, two phosphorylcholines, and a glyceramido acetamido trideoxyhexose residue [97,98]

App
It is phase-variable [103] ausI/MspA An autotransporter and a serine protease; it is phase-variable [45,46] Biofilm (and molecules involved in the biofilm synthesis, such as autA or hrpA, or optimizing pathogen survival in biofilm, such as the alpha-peptide of IgA1 protease, adhC, estD) Biofilm protects from macrophages; adhC is involved in S-nitrosoglutathione metabolism and in glutathione-dependent detoxification system; EstD is involved also in Neisseria colonization [104][105][106][107][108] Blebs (with OMPs and LPS/LOS) and SOMVs They protect from neutrophils-mediated killing and NETs; they divert the immune response away from the pathogen [78] Capsule and molecules involved in the capsule synthesis such as kpsC, kpsS It activates TLR2 pathway, it increases serum resistance, and it inhibits the classical pathway of complement [109][110][111] Cas9 and the CRISPR-Cas system CRISPR-Cas9-mediated repression of bacterial lipoprotein expression facilitates evasion of TLR2 by the pathogen; it is involved in gene expression and regulation [112,113] cbpA It mediates zinc piracy and protects from nutritional immunity [93] Cps As a gene, it is involved in the capsule biosynthesis; as RNA, it acts as a thermosensor; Cps gene amplification protects the pathogen [99,114] CrgA It is involved in the regulation of pili and capsule expression; it plays a major role in the infectious cycle of Neisseria [114][115][116] Css As a gene, it is involved in the capsule biosynthesis; as RNA, it acts as a thermosensor [99] ctrA, ctrD As genes, they are involved in the capsule export; as RNAs, they act as thermosensors; IS1301 in the IGR between sia and ctr operons mediates resistance to Abs [99,117,118] cycP It is involved in denitrification metabolism and protects Neisseria from nitrosative stress [90,119,120] dam It is involved in phase variation and modulation [42] dcaC I t i s p h a s e -v a r i a b l e [ 40] dinB A DNA polymerase IV belonging to the SOS regulon: it is involved in phase variation and modulation [42] DNA mismatch repair genes (fpg, mutL, mutS, mutY, recA, recN, uvrD) They are phase-variables; they protect against oxidative stress [42,48,51,121] drg It is involved in phase variation and modulation [42] farA, farB, farR They remove antimicrobial peptides, proteases, lysozyme, and acids from the bacterial cytosol and protect the pathogen [122,123] fbpA, fbpB They are involved in phase variation and modulation [51] Feta It is involved in phase variation and modulation [124][125][126] fHbp (formerly known as GNA1870) It is involved in phase variation and modulation; it protects Neisseria from complement-mediated killing, binding fH [90,127] frpA, frpB, frpC They are phase-variable; they can act as a molecular decoy [124,125,128] funZ It is a site of bacteriophage insertion; it is phase-variable [49] fur It is involved in phase variation and modulation; it tunes the gene expression of virulence genes [102,129] ggt It regulates pathogen growth [130] Ght It is involved in the capsule biosynthesis and in the resistance mechanisms of the pathogen [131,132] gltT It favours meningococcal internalization into human endothelial and epithelial cells; it regulates pathogen growth [133,134] H.8 AAEAP motifs are target for generation of blocking Abs [135][136][137][138] 6 Journal of Immunology Research They are involved in phase variation and modulation [43,[139][140][141] Hfq A RNA chaperone: it is involved in stress response and virulence and is a pleiotropic regulator of protein expression [142] hsdS I t i s p h a s e -v a r i a b l e [ 49] IgA protease It cleaves secretory IgA, hinders Ab binding and function, and may play role in biofilm formation; it cleaves lysosomal LAMP1 in epithelial cells; moreover, it is phase-variable [122,142,143] katA It confers resistance to RNS, including peroxynitrite (PN), protects against ROS, and detoxifies H 2 O 2 [90,102,122,144] Laz A lipid-modified azurin: it protects against hydrogen peroxide and copper toxicity; it promotes Neisseria growth and survival [135,138,145] lbpA, lbpB They are involved in iron acquisition and metabolism; they are phase-variable; moreover, the release of LbpB enables Neisseria to escape from complement-mediated killing [90,122,146] lctP Its inactivation results in C3-mediated cell lysis [102,147,148] lgtA, lgtB, lgtC, lgtD, lgtE, lgtG They are involved in LOS biosynthesis and are phase-variable; for example, lgtA or lgtC phase variation mediates LPS immunotype switch from L3 to L8/L1 [60] LOS/LPS It protects from macrophages; strains of the same species produce different LOS glycoforms [122] lptA It adds a phosphoethanolamine group to lipid A and confers resistance to defensins and cathelicidins [90,149] Lst LOS sialylation (by the enzyme Lst) prevents complement deposition and phagocytosis by neutrophils [122,150] mesJ I t i s p h a s e -v a r i a b l e [ 49] Msf It binds to vitronectin; it increases serum resistance [151] Mip It tunes gene expression [102,152,153] misR, misS They are phase-variable; they are involved in capsule regulation and modification [114,154] mltA (formerly known as GNA33) It tunes gene expression [155] mntA, mntB, mntC They protect against oxidative stress [122,156] modA, modB They are phase-variable [34] msrA, msrB They are involved in the methionine sulfoxide reduction and they repair oxidized proteins [122,157] mtrC, mtrD, mtrE They protect against cationic antimicrobial peptides and toxic hydrophobic molecules [122,158,159] nadA and its regulator nadR It binds to Hsp90, recruits ARF6 and Rab11, and activates human monocytes and macrophages, triggering IFN-gamma and R-848 dependent pathways; it interacts with beta1 integrins; it is phase-variable [39,[160][161][162][163][164][165] nalP An autotransporter protease: it cleaves C3, facilitates degradation of C3b, and enhances Neisserial survival in human serum; it stabilizes the biofilm; moreover, it is involved in the processing of other proteases, such as the proteases which release LbpB, whose release enables Neisseria to escape from complement-mediated killing; NalP processes also App and IgA1 protease; it has an important role in the virulence of the pathogen [24,102,166] Nhba (formerly known as GNA2132) It tunes gene expression [167] nhhA It activates TLR4-dependent and independent pathways; it triggers apoptosis in macrophages; it increases serum resistance and protects from phagocytosis and complement attack; it is essential for colonization [168,169]  nirK It protects Neisseria from nitrosative stress during colonization and invasion [170,171] norB It favours the pathogen growth, enabling utilization and consumption of NO during microaerobic respiration, enhances pathogen survival, protects Neisseria from nitrosative stress during colonization and invasion, decreases and downregulates the production of NO-regulated cytokines, such as TNF-alpha, IL-12, IL-10, CCL5 (RANTES), and CXCL8 (IL-8), and prevents host cell S-nitrosothiol formation [100,119,120,170,172] nspA It binds to factor H and inhibits AP [122,[173][174][175] nsrR It is involved in denitrification metabolism and protects Neisseria against nitrosative stress [176,177] oatW, oatY They tune gene expression [178] Opa It interacts with CEACAM, promoting endothelial cell attachment and upregulating endoglin (CD105) and cooperation with 1 integrins; it elicits innate host defences and actively suppresses adaptive immune responses that would eliminate the pathogen [179][180][181][182][183][184] Opc It binds to vitronectin, it inhibits AP, and it increases serum resistance; it elicits innate host defences and actively suppresses adaptive immune responses that would eliminate the pathogen [179][180][181][182]184] oxyR It regulates catalase expression and is involved in the protection from oxidative stress [185,186] P36 It is involved in Neisserial adhesion. [187] pacA, pacB They are involved in the composition and regulation of peptidoglycan membrane [188] pglA, pglB, pglG, pglH They are phase-variable [60][61][62] Pili They alter the expression levels of human genes known to regulate apoptosis, cell proliferation, inflammatory response, adhesion, and genes for signaling pathway proteins such as TGF-beta/Smad, Wnt/beta-catenin, and Notch/Jagged [189] pilC1 It interacts with mucosal surface and mediates the crossing of the BBB [41,169] PilE, pilS They are involved in nonreciprocal recombination-based antigenic variation [76] PilE, pilV They bind to CD147 for vascular colonization; they mediate also Neisseria internalization [190,191] pilP, pilQ They are involved in pilus biogenesis and outer membrane stabilization [51,192] porA It binds to fH, C3b, C4b, and C4bp (more strongly under hypotonic conditions); it increases serum resistance; it is involved in phase variation [122,139,173] porB It inhibits factor H-dependent AP; it interacts with TLR1 and TLR2 and activates I B, MAPK/MAPKK, and PTK pathways, leading to CD86 upregulation, to IL-6, IL-12, and TNF secretion in B cells and DCs, and to IgB secretion in B cells [122,173] pptA I t i s p h a s e -v a r i a b l e [ 47] Ppx It is an exopolyphosphatase whose mutation protects Neisseria from complement-mediated killing; it interacts with the AP of the complement activation [64] rmpM It is involved in phase variation and modulation [193,194] 8 Journal of Immunology Research sodB, sodC They protect from phagocytosis by human monocytes/macrophages [102] tbpA, tbpB (also known as tbp1, tbp2) They are involved in nutritional immunity [121] TdfF It is involved in intracellular iron acquisition and is found only in genomes of pathogen strains [28] Temperature sensors (such as RNA thermosensors located in the 5 UTRs of genes necessary for capsule biosynthesis, the expression of fHbp, and sialylation of LOS/LPS) Activated by coinfecting pathogens, they recruit mechanisms of resistance and immunity escape [99] tonB It is involved in nutritional immunity, supplying energy to the pathogen [93] Uncharacterized proteins (NGO1686, NMB0741, NMB1436, NMB1437, NMB1438, and NMB1828) They protect from nonoxidative factors, but their mechanisms are still not understood; NMB1436, NMB1437, and NMB1438 are putatively involved in iron metabolism [122,195] Uncharacterized factor (NMA1233) It is involved in phase variation and modulation [26,51] xseB It is involved in phase variation [26] znuD It protects from neutrophils and nutritional immunity [92] Ab: antibody; AP: Alternative Pathway; ARF6: ADP-ribosylation factor 6; App: adhesion and penetration protein; BBB: blood-brain barrier cationic antimicrobial peptides (CAMPs), including cathelicidin [196]. Conversely, the presence of capsular polysaccharide restrains the invasion and colonization of the nasopharyngeal barrier by hiding the adhesins and invasins of the meningococcus [143,158]. On the other hand, the presence of the capsule may allow the microorganism to pass unharmed through the intracellular environment by exploiting the system of cell microtubules, at least in the case of serogroup B N. meningitidis [197]. Moreover, the CP is essential for meningococcal growth and survival in the bloodstream and cerebrospinal fluid, increasing serum resistance. During the different stages of infection, the capsule may hinder or promote the survival of N. meningitidis in the human host; indeed, the microorganism can modulate the production of capsule components, which depends on 4 genes, three of which-siaA, siaB, and siaC-are situated in the cps locus. The siaD gene induces the production of polysialyltransferase, which allows the polymerization of sialic acid. For instance, in the early stages of infection, the production and assembly of sialic acid are downregulated [198]. Another example of polysialyltransferase system is given by oatWY [178].
In addition to the above-mentioned actions, the most important virulence activity of the CP is probably the significant impairment of both Neisserial adherence to DCs and the phagocytic killing of bacteria; indeed, the CP inhibits both the opsonic and the nonopsonic phagocytosis of N. meningitidis [199]. It prevents the formations of effective Abs against N. meningitidis.
CP downregulates both classical and alternative complement pathways and prevents the proper insertion of the membrane attack complex (MAC) [200,201]. LPS also contributes to inhibiting MAC deposition [201,202].
Moreover, CP switching contributes to escaping detection and killing. This is a complex phenomenon due to microevolution and usually involves Neisserial strains expressing sialic acid (e.g., the shift from serogroup B to C, from serogroup C to W-135, from serogroup Y to W-135, and from serogroup Y to B) [79][80][81][82]203]. The molecular basis is provided by the allelic replacement of the sialic acid CP polymerase.

Major and Minor Adhesion
Mechanisms of N. meningitidis. N. meningitidis possesses a multifaceted system of adhesins, such as pili and other systems of adhesion (i.e., opacity-associated proteins OpA and OpC). Adhesion is probably a coordinated process mediated first by pili, which are composed of several proteins; the most important of these is Pilin E (PilE) [204], but Pilin C (PilC) [205] and the secretin Pilin Q (PilQ) [206] have also been described. PilE is the main constituent of the Neisseria type IV pilus. In 1984, Diaz et al. identified proteins I and II as the main components of the type IV pili and noted that Abs against protein I were highly specific [207]. Subsequently, Pilin E was classified as Journal of Immunology Research 9 belonging to class I and class II. Class I Pilin E is highly variable, while class II Pilin E is highly conserved [208,209]. For this reason, class II Pilin E has been suggested as a candidate antigen for a vaccine against meningococci [147]. The regulation of pilin genes is quite complex [97,115,116,190].
Although the interactions between type IV pili and cellular receptors are not completely known, they may interact with a membrane cofactor protein, named CD46 receptor, and with alpha 1 and alpha 2 integrins [211]. However, it is known that pili contribute to the aggregation of Neisserial cells [212]. This fact, associated with the ability of pili to act as a signalling protein by interacting directly with the 2adrenergic receptor, contributes to the depletion of junction proteins, thus helping meningococcus to pass through the epithelial and endothelial cells and, subsequently, to cross the blood-brain barrier (BBB) [75,213].
Although pili are essential to the first phase of Neisserial adhesion to the cells of the airways, other adhesion molecules, such as LPS and porin A, intervene to strengthen the microbial bond. In particular, OpcA and OpcC appear to be very important; indeed, OpcA binds carcinoembryonic antigen cell adhesion molecules (CEACAMs), heparan sulphate proteoglycan (HSPG) and integrins [179][180][181][182][183]214]. Opc proteins can combine with HSPGs and, through vitronectin and fibronectin, with their integrin targets. Furthermore, Opa proteins are able to elicit innate human defences that favour the survival of N. meningitidis, while actively suppressing adaptive immune responses that would eliminate the pathogen [184]. The variability of the expression of different Opa proteins could play a major role in prolonging the state of infection by circumventing the humoral host immune response [215].
The adhesion and penetration protein (app) [103], which is a member of the autotransported family of secreted proteins, owes its name to its ability to adhere to human cells, thereby favouring the entry of Neisseriae. To circumvent the immune system, meningococci possess formidable machineries that allow them to secrete proteins in different manners; in particular, Neisseriae mainly use the autotransporter pathway (also known as type V secretion system) [216]. The first-described protein belonging to the autotransporter superfamily was an IgA protease from N. gonorrhoeae [217]. MspA (meningococcal serine protease A) is another putative autotransporter protein. Not all strains of Neisseria gonorrhoeae/meningitidis possess the gene for MspA/AusI (also known as NMB1998). However, Turner et al. [218] demonstrated that convalescent sera from subjects affected by serogroup B invasive disease recognized the MspA antigen. NhhA (Neisseria hia/hsf homologue A, also known as Hsf) and Neisserial adhesin A (NadA) also belong to the autotransporters. Nhha contributes to bacterial adhesion by binding heparin sulphate and laminin. In addition, through the activation of caspase, NhhA increases the rate of macrophage apoptosis [168,219]. NadA, which is expressed by 50% of virulent strains [160], but only by 5% of the strains isolated from carriers, is of particular interest because it is one of the components of the 4CMenB (Bexsero) vaccine and binds beta1 integrins [220].

N. meningitidis: Avoidance Mechanisms against Reactive Oxygen Species (ROS), Reactive Nitrogen Species (RNS), and
Antimicrobial Peptides (AMPs). When in contact with the mucosa of the nasopharynx, N. meningitidis can implement several strategies of adhesion, but it must overcome many barriers of innate immunity. We have already mentioned how the capsule allows bacteria to mitigate the activity of DCs. However, it must elude other substances, such as the reactive oxygen species and reactive nitrogen species produced by macrophages and the antimicrobial peptides produced by activated neutrophils. As already mentioned, the capsule protects N. meningitidis from LL-37 cathelicidin, but LPS also contributes to the resistance of the bacterium against this cathelicidin [156].
Furthermore, the toxic action of ROS is neutralized by the secretion of enzymatic proteins, such as catalase and superoxide dismutase [144,149,221]. The gene that codes for catalase is katA and is regulated by OxyR [185,186,222], while SodB and SodC code for superoxide dismutase [223]. Laz, a lipid-modified azurin, protects the pathogen against H 2 O 2 and cupper toxicity [135-138, 145, 176].
In addition, N. meningitidis possesses genes, which encode enzymes able to exert a denitrification action, such as aniA, CycP, nirK, nsrR, and norB. They favour the growth of the pathogen, enabling utilization and consumption of NO during microaerobic respiration, enhance pathogen survival, and protect Neisseria from nitrosative stress during colonization and invasion by preventing host cell S-nitrosothiol formation. Moreover, they reduce and downregulate the production of NO-dependent cytokines, such as TNF-alpha, IL-12, IL-10, CCL5 (RANTES), and CXCL8 (IL-8) [100,101,119,120,130,157,170,172,177,224]. Contemporarily, already at the level of the mucosa, the microorganism must resist the complement system.
Another interesting mechanism is the strategy whereby N. meningitidis escapes the attempts of the host to sequester nutrients essential for growth and survival of the pathogen. This process has been termed "nutritional immunity" [131,139]. The microbe is endowed with OM receptors (such as HmbR, HpuA or HpuB, TbpA or TbpB, and TdfF) for acquiring iron and other important metals [93,124,125,128,129,140,195,225]. ZnuD is a high-affinity zinc uptake receptor, which plays an important role in enabling the pathogen to evade neutrophil-mediated killing [226,227]. CbpA, a receptor for calprotectin, a protein released by neutrophils during inflammatory processes, is upregulated when N. meningitidis suffers from zinc limitation [226,227]. Further examples of metabolic enzymes involved in nutritional immunity are glutamate transporters or molecules taking part in the carbon cycle [132][133][134][228][229][230].

How Meningococcus Circumvents the Complement System.
Three pathways can activate the complement functions, namely, the classic pathway, the alternative pathway, and the lectin pathway. All three of these pathways contribute to the transformation of C3 to C3b [231].
The alternative pathway acts by comparing self-with nonself-antigens and is activated by anything that differs from the markers of host cells. Specifically, factor H recognizes hostassociated molecular patterns (HAMPs). Properdin, first identified in 1959, is another protein that can directly activate the alternative pathway of the complement system [232]. It has been demonstrated that properdin deficiency favours recurrent episodes of N. meningitidis infection [233].
Meningococci produce three different variants (1, 2, and 3) of a protein that binds factor H. This protein, named fHbp (factor H binding protein) or GNA1870, blocks activation of the alternative pathway of the complement system. Indeed, by surrounding themselves with fHbp, N. meningitidis cells capture and inactivate factor H. Thus, it is easy for the microorganism to survive and reproduce, especially in the bloodstream and cerebrospinal fluid. Hence, it is important that the 4CMenB (Bexsero) vaccine contains variant 1 of this protein, which is often expressed by virulent meningococcal strains [228]. Another vaccine (Trumenba), recently licensed in the USA, contains recombinant variants 1 and 2 of fHbp from N. meningitidis serogroup B, A and B subfamilies (A05 and B01, resp.) [234,235]. The proteins are produced by exploiting an advanced genetic engineering technique, using E. coli as a vector.

The Adaptive Immune Response against Neisseria meningitidis.
Microorganisms such as N. meningitidis are able to change many exposed surface proteins, while the polysaccharides, which constitute the capsule, are T-independent (TI-2) antigens and can activate B cells directly, without the intervention of the MHC. However, TI-2 antigens do not induce an efficient secondary response and do not induce the production of avid immunoglobulins. Rather, they induce the production of short-lived Abs belonging to the IgM class [236].
In addition, it is important to consider that, in infants and children, the development of the immune system is a dynamic process, which begins in utero and continues for months and even years after birth. This explains why many components of the immune system are inefficient or partially efficient in infancy and early childhood [237][238][239]. For this reason, most cases of meningitis and sepsis from N. meningitidis occur under the age of 4 years and particularly in the first year of life.
The critical role of bactericidal Abs against the exposed surface antigens of N. meningitidis has been demonstrated by several studies. Indeed, Goldschneider et al. [240] showed that only individuals without bactericidal Abs contracted the clinical disease. In addition, the successful therapeutic use of immune sera, which markedly reduced lethality when first implemented by Flexner [241], has shown the central role of these Abs in protecting against invasive disease. The opsonic activity of Abs is also very important in the protection against and the recovery from meningococcal disease, as is demonstrated by the role of neutrophils, macrophages, and DCs in combating N. meningitidis. It is also well known that the cerebrospinal fluid of patients contains large numbers of neutrophils full of microorganisms. These clusterings of neutrophils are known as neutrophil extracellular traps (NETs) and massively release cathepsin G. N. meningitidis circumvents these traps by blebbing spontaneously released OMVs (SOMVs). Other strategies that the pathogen exploits are modification of lipid A of LPS with phosphoethanolamine protected and upregulation of ZnuD [92].
The adaptive immune response has been studied in carriers and during both the invasive period leading to clinical meningitis and the convalescence period. T cell response is "two-faced"; while proinflammatory T cells may indeed blunt the invasive power of the pathogen, the induction of the Treg response, which is able to limit virulence, carries the price of the reduced effectiveness of the protective response, especially in children [242]. During infection, increased meningococcal antibody titres can be detected from the 4th day, peaking at the end of the third week or the beginning of the fourth week and showing a correlation with the severity of the disease and the age of the patient. In the acute period of the disease, the number of T cells generally drops, while that of B cells increases; by the end of the second week, IgG levels decline and IgM levels rise [243]. In particular, abnormalities in T cell response can be detected, such as an elevated percentage of CD25+ and HLA-DR+ T cells, an increase in CD4+ CD45R+ (suppressor-inducer) cells, with subsequent expression of activation antigens, and a decrease in CD4+ CDw29+ (helper-inducer) cells [244]. During convalescence, an age-associated Th response can be observed: specifically, a Th1 response (low IL-10/IFN-ratio) and a highly proliferative Th2 response (higher IL-10/IFN-ratio) can be detected in younger and older patients, respectively [245]. Generally, a significant CD4+ T central memory response, with serum bactericidal antibodies, a marker of protective immunity, can be found [246]. However, the above-mentioned age-related mucosal T effector/memory cell response may also be present without bactericidal antibodies [247].

Other Immunoescape Strategies.
Temperature fluctuation plays an important, although underscored, role in microbial pathogenesis, colonization, invasion, and host evasion. In contrast to mammals that maintain constant body temperature, pathogens' and other animals' temperature oscillates on a daily basis. Loh and collaborators [99] have identified the molecular bases of this temperature-dependent strategy. They have studied three RNA thermosensors located in the 5 untranslated regions (UTRs) of genes involved in the CP biosynthesis, the expression of fHbp, and sialylation of LOS/LPS. Increased temperature (e.g., during inflammation by coinfecting pathogens, such as influenza virus) "alarms" the meningococcus and triggers its defence mechanisms against human immune killing. This could be a key determinant for the transformation of a symbiont pathogen into a virulent one. However, the precise nature of this mechanism remains elusive.
Clustered regularly interspaced short palindromic repeats-(CRISPR-) Cas9 is another molecular tool that Neisseria can use in order to divert immune surveillance. It is an intrinsically ambivalent device, since, on one hand, being involved in gene expression and regulation, it restricts the possibility of editing the Neisseria genome via HGT/LGT or the insertion of exogenous nucleic material, and therefore it limits the microbial variability. On the other hand, CRISPR-Cas9-mediated repression of bacterial lipoprotein expression facilitates evasion of TLR2 by the pathogen [112].
Another mechanism is the molecular decoy, with which the microbe deceives the human immune system. For example, fprB has an antigenic subdomain for binding antibodies, which is not essential for the functioning the autotransporter [94].
The example of fprB is useful to show how Neisseria can use concurrently the previously described immunoescape strategies: fprB is subject to a high degree of antigenic variation, is a phase-variable gene, is involved in nutritional immunity, and moreover exploits a molecular decoy. Neisserial carbohydrates, mimicking host carbohydrates, circumvent immune system and at the same time exploit their mimicry to recruit fH [91].

Polysaccharide Vaccines.
As at least six meningococci are pathogenic in humans, the development of meningococcal vaccines has been a challenge. The first step towards solving the problem was to find a common denominator among antigens that showed high variability. In 1969, Gotschlich et al. [248] correctly put the first imperfect piece of the puzzle in place by demonstrating the possibility to extract the capsule polysaccharide. However, although high molecular weight groups A and C meningococcal polysaccharides proved immunogenic in adults [249][250][251], important limitations of this vaccine emerged in subsequent years. Indeed, while both group A and group C vaccines proved effective in USA and Italian recruits [252,253] after one administration, they displayed only short-term efficacy in older children and adults. Moreover, vaccines against serogroup C did not prevent disease in infants, and the efficacy of group A vaccines in children under 1 year of age was unclear. The immune response occurred 10 days after vaccination. In schoolchildren and adults, one dose of these vaccines seemed to provide protection for at least 3 years [254]. These findings are explained by the characteristics of the antigens contained in the vaccines. Indeed, polysaccharides with repeating epitopes induce an immune response with the following characteristics [255,256]: (i) The response occurs between the ages of 3 and 18 months but is variable; in children less than 2 years of age, the response is usually poor; the affinity maturation of the Abs does not occur.
(ii) The immunologic memory is not stimulated.
(iii) Almost all (>90%) Abs produced belong to the IgM class and are produced in the spleen [257,258]. Furthermore, several studies have suggested that vaccination with large amounts of polysaccharides induces immune tolerance towards these antigens [259].

Polysaccharide Conjugate Vaccines.
Three immunogenic carrier proteins are generally used inpolysaccharide conjugate vaccines, namely, diphtheria or tetanus toxoid, CRM197 (a nontoxic variant of diphtheria toxin obtained by molecular biology), and a complex of outer membrane protein (OMP) mixture from N. meningitidis. The toxoids were selected as carriers, firstly because of their immunological potency and secondly because if the recipient had previously been immunized with the toxoid, a booster effect was predictable. Moreover, under particular conditions, suppressive effects can also occur. The conjugate vaccine against N. meningitidis serogroup C had great success in several countries (UK, Netherlands, Spain, Italy, etc.) [260]. Today, tetravalent conjugate vaccines ACWY are also available in both the USA and Europe [260][261][262].

Vaccines against N. meningitidis Serogroup B.
The most difficult problem was to prepare a vaccine against N. meningitidis serogroup B strains. Indeed, the maximum expression of camouflage is found in the capsule of these strains. Specifically, the polysaccharide of serogroup B meningococcal strains is a homopolymer of sialic acid residues and has structural similarities to brain glycoproteins. For this reason, it was impossible to prepare a polysaccharide antimeningococcus B vaccine. To overcome this obstacle, OMVs-vaccines were developed and used during epidemics caused by N. meningitidis serogroup B strains. However, given the high variability of the proteins, such as porins, present in OMVs, these vaccines were effective only against very specific epidemic hypervirulent strains [263,264]. In order to develop a vaccine against meningococcus B, other strategies were therefore implemented. The most promising results were obtained through reverse vaccinology. This involves identifying the antigens for the vaccine not in the classic way-that is, from the components of the bacterium-but instead from the genes that express the proteins with the best characteristics to be good antigens for the vaccine. To obtain a universal vaccine against serogroup B meningococcal strains, complex bioinformatics software was also applied. Following the complete sequencing of the meningococcus B genome [265], researchers at Novartis Vaccines and Diagnostics identified 600 ORFs, which expressed proteins that are exposed on the surface of the bacterial cell. Subsequently, 350 proteins were successfully expressed in E. coli, purified, and used to immunize mice. Later, 28 novel protein antigens able to elicit Abs with bactericidal activity were identified. Finally, three of these 28 proteins were selected, namely, NHBA (GNA2132, fused with GNA1030 protein), fHbp (fused with GNA2091 protein), and nadA. NHBA (Neisseria heparin binding antigen), which is present in virtually all strains, binds heparin, which may increase the serum resistance of bacteria. fHbp (factor H binding protein) binds factor H, which enables the bacterium to survive in theblood [266,267], thereby blocking the alternative pathway of the complement system. nadA (Neisserial adhesin A) promotes adherence to and invasion of human epithelial cells [161,162]. In addition, the vaccine developed by Novartis Vaccines and Diagnostics contains a fourth component, namely, the vesicle of the OM from the New Zealand strains, which contain porin 1.4. Theoretically, this vaccine should elicit bactericidal Abs against the following: (i) NHBA, thus increasing the bactericidal activity of the serum.
(ii) fHbp, thus exposing N. meningitidis to the alternative pathway of the complement system.
(iii) NadA, thus hindering the adherence of N. meningitidis to epithelial cells.
(iv) Porin A 1.4 and other components of the mixture of proteins contained in OMVs; indeed, OMVs can enhance the immune response by functioning as a conjugate complex of proteins or, rather, a complex of adjuvants.
Several clinical trials have demonstrated the immunogenicity and safety of this vaccine, also named 4CMenB, in infants, children, adolescents, and adults [268,269]. Consequently the vaccine has been approved by the EMA and named Bexsero [270]. It has also been approved by several national drug agencies (such as FDA), including the Agenzia Italiana del Farmaco (AIFA, Italian National Agency for Drugs) [271].
Owing to the wide variability of Neisserial antigens, a particular laboratory test (MATS) has been developed to estimate the potential effectiveness of the vaccine. Studies conducted worldwide have shown the potential effectiveness of Bexsero, which has been estimated at 87% in Italy [85].
Trumenba was approved for individuals aged 10 to 25 years in the USA in October 2014 [85]. The potential of the vaccine antigen was tested in the laboratory [272] and on a murine model [273,274]. The immunogenicity, safety, and tolerability of this vaccine were investigated in a randomized controlled trial in infants aged 18-36 months, who were subdivided into three cohorts (receiving 20-, 60-, and 200g rLP2086 dose, resp.) and matched with two control groups: one vaccinated against hepatitis A virus (HAV) and the other with a saline placebo. After the vaccination cycle, seroconversions against Neisserial strains expressing LP2086 variants homologous to the vaccine antigens were found in 61.1-88.9% of toddlers and against strains expressing heterologous LP2086 variants in 11.1-44.4%. Adverse reaction rates were negligible and the vaccine proved to be safe and well tolerated [275]. However, another randomized phase 1/2 clinical study found high fever rates in toddlers receiving one 20-or 60-g rLP2086 dose (64% and 90%, resp.) [276].
In a randomized study performed in the USA and Europe in a sample of adolescents (11-18 years of age), Trumenba proved to be highly immunogenic. The proportion of vaccinees with human serum bactericidal activity (h-SBA) titres with a ≥4-fold rise against hypervirulent Neisserial strains with different variants of fHbp was in the range of 75-100%. In another randomized clinical study, carried out in Australia, the safety and immunogenicity of the vaccine were assessed in 60 healthy adults (18-40 years of age) who received 120 g doses at 0, 1, and 6 months. The percentage of seroprotected vaccinees was 94.3% against the homologous strains and 70-94.7% against the heterologous strains. The vaccine was well tolerated [277,278].
As fHbp is also expressed by Neisserial serogroups other than B, the anti-fHbp Abs elicited by rLP2086 might exert a bactericidal effect on meningococci, such as those against N. meningitidis serogroup C, as proved by Harris and collaborators [279] and by Konar and colleagues [280]. Moreover, there is some evidence that Trumenba could, at least in part, have an effect on carriage and reduce the risk of acquiring some hypervirulent strains [281].

New
Vaccines. The currently available Neisseria vaccines, described in the previous paragraphs, are reported and summarized in Table 3.
Reverse vaccinology has proved to be a promising approach, enabling researchers to develop the effective vaccine Bexsero. New highly integrated approaches, which combine genomics with postgenomics, are leading to nextgeneration vaccines. A combination of reverse and forward vaccinology techniques, such as immunoproteome investigation via combined cell surface immunoprecipitation and mass spectrometry (MS) [153], and new bioinformatics strategies, such as the protectome approach [292], are promising in identifying highly conserved motifs in known bacterial protective antigens and using them for the design of effective universal vaccines [293,294].

Conclusions
The development of effective vaccines against meningococcal disease has been a long and hard struggle. Early efforts yielded only partial results, with the creation of polysaccharide vaccines [295]. Subsequent research, however, led to the production of the conjugate vaccines [296]. Today, we have the conjugate meningococcal C vaccines [297], an affordable conjugate vaccine against N. meningitidis serogroup A (MenAfriVac) [298,299], the tetravalent conjugate vaccines [270], and, finally, two "universal" vaccines against meningococcal group B strains [300,301]. The critical rate of coverage required in order to eliminate the disease is probably not among the highest [302]. Indeed, the conjugate vaccine for serogroup C has resulted in dramatic reductions of cases of the disease [303] and created herd immunity that seems to have had a significant effect on the carrier status of adolescents and young adults. Thus, the prospect of dominating this very serious disease lies decidedly in the medium term.

Conflict of Interests
R. Gasparini, D. Panatto, N. L. Bragazzi, P. L. Lai, A. Bechini, M. Levi, P. Durando, and D. Amicizia declare that they have no conflict of interests regarding the publication of this paper.