The Leptospiral General Secretory Protein D (GspD), a secretin, elicits complement-independent bactericidal antibody against diverse Leptospira species and serovars

Leptospirosis, the most common zoonotic infection worldwide, is a multi-system disorder affecting the kidney, liver, and lungs. Infections can be asymptomatic, self-limiting or progress to multi-organ system failure and pulmonary hemorrhage. The incidence of canine and human leptospirosis is steadily increasing worldwide. At least sixty-four Leptospira species and several hundred lipopolysaccharide-based serovars have been defined. Preventive vaccines are available for use in veterinary medicine and limited use in humans in some countries. All commercially available vaccines are bacterin formulations that consist of a combination of laboratory cultivated strains of different lipopolysaccharide serotypes. The development of a broadly protective subunit vaccine would represent a significant step forward in efforts to combat leptospirosis in humans, livestock, and companion animals worldwide. Here we investigate the potential of General secretory protein D (GspD; LIC11570), a secretin, to serve as a possible antigen in a multi-valent vaccine formulation. GspD is conserved, expressed in vitro, antigenic during infection and elicits antibody with complement independent bactericidal activity. Importantly, antibody to GspD is bactericidal against diverse Leptospira species of the P1 subclade. Epitope mapping localized the bactericidal epitopes to the N-terminal N0 domain of GspD. The data within support further exploration of GspD as a candidate for inclusion in a next generation multi-protein subunit vaccine.


Introduction
Leptospirosis is a potentially life-threatening multi-systemic zoonotic infection of veterinary and human concern that occurs worldwide [1,2]. The incidence of leptospirosis is highest in tropical and subtropical regions where socioeconomic conditions are poor. The genus Leptospira consists of at least 64 pathogenic and saprophytic species that form two pathogenic and saprophytic clades (P and S, respectively). These clades can be further divided into subclades (P1, P2, S1 and S2) [3]. While the virulence potential of many Leptospira species remains to be determined, most well characterized pathogenic species belong to the P1 subclade [3]. Leptospira have also been classified based on the antigenic properties of their lipopolysaccharide (LPS) resulting in the designation of hundreds of serovars.
Leptospira are maintained in nature by a wide array of reservoir hosts [3]. As a consequence of the ability of pathogenic species to establish chronic colonization of the kidneys, leptospires are con-tinually shed in urine. Contact with animals (livestock, wildlife and companion), contaminated soil, and water serve as common sources of infection [4]. The severity of leptospirosis can range from asymptomatic colonization to mild or serious infection with life-threatening clinical manifestations including compromised kidney and liver function and pulmonary hemorrhage [2,5]. Over 1,000,000 cases of human leptospirosis are documented annually with over 60,000 deaths worldwide [6]. While the incidence of leptospirosis is on the rise in companion animals, an accurate count of cases per annum is elusive [7]. In dogs, cases are most common in unvaccinated dogs <6 months of age and <14 lbs in weight with approximately 25% of cases being fatal (American Kennel Club; https://www.akcchf.org/educational-resources/library/an-updateon-canine.html; accessed on 01/07/2021).
In veterinary medicine (small and large animal) vaccination serves as the primary preventative strategy for leptospirosis [8].
Leptospira bacterin vaccines consist of inactivated cell lysates of multiple strains. The antibody response to Leptospira bacterin vaccines is directed largely at LPS. As a result, the protection afforded by bacterins is limited primarily to the serovars represented in the vaccine formulations [9,10]. The dominant circulating serovars in the United States are Icterohaemorrhagiae, Canicola, Grippotyphosa, Pomona, Bratislava, and Autumnalis. In Europe, Icterohaemorrhagiae, Grippotyphosa, Australis, Sejroe and Canicola are dominant [11]. In addition to limited cross-protection [12], concerns have also been raised about the reactogenicity of Leptospira bacterin vaccines, particularly those used in canines [13,14]. Specific protein antigens that convey broad protection remain to be identified (reviewed in [9]). The development of recombinant protein based subunit vaccines represents an attractive alternative to bacterin based vaccines.
In this study we assessed the vaccine potential of the L. interrogans serovar Copenhageni strain Fiocruz L1-130 general secretory protein D (GspD) [15]. GspD is a 67 kDa protein that is predicted to function as a secretin for a leptospiral type 2 secretion (T2S) system [16,17]. Here we demonstrate that GspD is an outer membrane antigen produced during in vitro cultivation and during natural infection in canines. Immunization of rats with recombinant GspD (GspD) elicited potent bactericidal antibody that effectively targeted multiple Leptospira species, strains and serovars of the P1 subclade. Antibody mediated killing was demonstrated to occur through a complement independent mechanism. Epitope mapping revealed that the bactericidal epitopes of GspD reside within its N-terminal 'N0 0 domain [18,19]. The data presented within support the potential use of GspD in a subunit-based vaccine formulation.

Bacterial strains and cultivation
Leptospira isolates were cultured at 30°C in Ellinghausen-McCul lough-Johnson-Harris (EMJH) medium supplemented with Probumin vaccine grade BSA (Millipore) as previously described [20]. Growth was monitored by wet-mounts using dark-field microscopy. All isolates assessed in vitro or in silico are described in Table 1. Note that strain virulence was not assessed as part of this study.

Phylogenetic analyses
GspD amino acid sequences from diverse Leptospira isolates were aligned using Mega X [21]. GspD homologs from Leptonema illini, Turneriella parva, and Escherichia coli K-12 served as outliers. A phylogenetic tree was constructed in Mega X using the neighborjoining tree algorithm with 500 bootstraps, a poisson model of amino acid substitution, and pairwise deletion of gaps [21]. Mega X was also used to calculate percent amino acid identity and similarity values. GspD accession numbers are listed in Table 1.

Cloning, expression and purification of recombinant proteins
PCR and cloning were performed using standard conditions as previously described [22]. LipL32, gspD, qlp42 and gspD gene fragments were PCR amplified from L. interrogans serovar Copenhageni strain Fiocruz L1-130 genomic DNA using Phusion polymerase (Thermo Scientific) and primers that harbor restriction sites for cloning ( Table 2). Signal sequences predicted by SignalP-5.0 [23] were excluded from the amplified lipL32 and qlp42 sequences. The amplicons were purified using QIAquick PCR Purification kits (Qiagen), digested with the appropriate restriction enzymes (New England Biolabs), ligated with linearized pET-45b(+), transformed into E. coli BL21(DE3) cells, and protein production induced with 1.0 mM Isopropyl b-d-1-thiogalactopyranoside [24]. All recombinant proteins were produced with an N-terminal hexa-histidine tag. Gene sequences were verified on a fee for service basis (Gene-wiz). Recombinant proteins that separated with the soluble fraction (Qlp42 and LipL32) were purified by nickel affinity chromatography using an ÄKTA purification platform as previously described [25]. Recombinant Treponema denticola Factor H binding protein B (FhbB), generated as part of an earlier study, served as control in several experiments [26]. Recombinant proteins that partitioned in the insoluble fraction (GspD and GspD subfragments F1 through F5) were purified using urea-denaturing conditions and then dialyzed stepwise into 1X PBS [24]. Protein concentrations were determined by bicinchoninic acid assay (BCA; Pierce).

Generation of antisera
Antisera against recombinant GspD, Qlp42 and LipL32 were generated as previously described [27]. In brief, Sprague Dawley rats were immunized (intraperitoneally) with 50 mg of recombinant protein in Freund's complete adjuvant (Sigma-Aldrich) with 2 boosts of 25 mg recombinant protein in Freund's incomplete adjuvant (Sigma-Aldrich) delivered 2 weeks apart. One week after the final boost, rats were sacrificed, blood collected via cardiac puncture and the sera harvested using Grenier Bio-one Vaccuette Z Serum Sep Clot Activator tubes (Grenier). All animal use protocols were reviewed and approved by the VCU IACUC and followed the Guide for the Care and Use of Laboratory Animals (Eighth edition).

SDS-PAGE, immunoblot, and dot-blot analyses
Cell lysates (0.3 OD 600 units per lane) were separated using AnykD Criterion Precast Gels (Bio-Rad) and transferred to PVDF membranes using the Transblot Turbo System per the manufacturer's protocol (Bio-Rad). Dot-blots were generated by spot application of 500 ng of GspD, LipL32, and FhbB (negative control) on to nitrocellulose membranes (22 mM pore size; BioRad) followed by air drying. Immunoblots and dot-blots were incubated with blocking buffer (PBS with 5% nonfat dried milk; 0.2% Tween 20; 2 h). Microscopic agglutination test (MAT) positive canine sera (1:200 dilution) were added (1 h), the blots were washed and rabbit anti-canine IgG HRP-conjugated secondary antibody (1:20000 dilution; Novus Biologicals) was added. IgG binding was detected using ECL substrate (Bio-Rad) and images were captured using a ChemiDoc Touch Imaging System (Bio-Rad). Note that the immunoblot and dot-blot images were cropped to generate the figures.

Bactericidal assays
Bactericidal activity of anti-GspD antisera was assessed as previously described [27,29] with some modifications. In brief, 4 mL of mid-log phase Leptospira cultures (density of~100 cells per field of view under 400Â magnification) were combined with 8 mL of EMJH media. Four mL of heat-inactivated (HI; 56°C, 30 min) rat anti-GspD antiserum, 4 mL of complement certified normal human serum (NHS; Innovative) or 4 mL of HI-NHS was added. All assays were performed in triplicate. Cells incubated in EMJH, NHS, and rat pre-immune (PI) serum served as negative controls. The samples were incubated at 30°C for 2 h, and the average number of live cells per 5 fields of view determined by visual counting under dark-field microscopy. Percent killing was calculated by determining the decrease in the number of live cells in the treatment groups versus the NHS + rat pre-immune serum negative control. Statistical analyses were conducted using Student's t test.

Analysis of GspD conservation amongst Leptospira species and strains
Evolutionary distances and phylogenetic relationships among GspD sequences were determined through pairwise comparisons. A phylogenetic tree is presented in Fig. 1A. The Leptospira GspD sequences assessed divide into three distinct phyletic clusters that are well-supported by bootstrap analyses. Amino acid identity and similarity values are presented in Fig. 1B. Identity values among Leptospira P1 subclade isolates ranged from 90.1 to 100%. GspD sequences derived from isolates of subclades P2 and S1 display approximately 75 and 60% identity with P1 subclade isolates, respectively. Hence, while GspD is conserved within a subclade, divergence between subclades is significant.

In vitro expression and subcellular localization of GspD
In vitro expression of GspD by diverse P1 strains was demonstrated by screening immunoblots of Leptospira whole cell lysates with anti-GspD antiserum. A single protein of approximately 67 kDa was detected in all strains tested with the exception of the saprophyte, L. biflexa strain Patoc I, which belongs to subclade Fig. 1. Phylogenetic analyses and pairwise sequence comparisons of GspD amino acid sequences. Panel A presents a phylogenetic tree constructed using GspD amino acid sequences as detailed in the text. Representative GspD sequences from Leptospira subclades P1, P2, and S1 (as indicated) were included. Leptonema illini, Turneriella parva and Escherichia coli str. K-12 served as outgroups. The results of bootstrap analyses are indicated on each node and the scale bar is shown in the lower left. Note that the numbers (No.) in parentheses that follow each isolate designation correspond to those used in the distance matrix table (Panel B). Panel B presents the GspD percent amino acid similarity and identity values (lower and upper quadrants, respectively). The subclade classification of each isolate of origin is indicated along the top of panel B. All accession numbers can be found in Table 1.

EJA. Schuler and RT. Marconi
Vaccine: X 7 (2021) 100089 S1 ( Fig. 2A). While it is possible that L. biflexa does not produce GspD in vitro, an alternative possibility is that its GspD protein, which shares only 60% amino acid identity with the L. interrogans serovar Copenhageni strain Fiocruz L1-130 GspD protein used to generate the sera, is antigenically distinct. In light of the conservation of GspD at the intra-clade level, the difference in signal intensity among P1 subclade strains observed in the immunoblot most likely suggests variable levels of expression in vitro (Fig. 2A). The subcellular localization of GspD was determined through Triton X-114 extraction and phase partitioning of L. interrogans serovar Copenhageni strain Fiocruz L1-130. The resulting fractions were immunoblotted and screened with anti-GspD, anti-LipL32 and anti-Qlp42 antisera (Fig. 2B). GspD was detected in the whole cell lysate, the detergent insoluble phase (protoplasmic cylinder), and the detergent soluble phase (outer membrane). Based on earlier studies that demonstrated that GspD homologs of other bacterial species interact with both the inner and outer membrane [18,19,32], the detection of GspD in both the detergent insoluble and soluble phases was expected and is consistent with its putative function as a secretin in type 2 secretion [17]. As controls, identical immunoblots were screened with anti-LipL32 and anti-Qlp42 antisera. LipL32 and the Qlp42 have been demonstrated to partition with the outer membrane and aqueous phases, respectively [28,[33][34][35][36]. Both proteins partitioned as expected.

GspD is expressed during natural infection in client-owned canines
Serum from forty-one MAT positive and nine MAT negative client owned dogs were screened for IgG to GspD using a dot-blot format. Recombinant LipL32 and FhbB (T. denticola FhbB [26]) served as positive and negative control detection antigens, respectively. Of the forty-one MAT positive dogs, 92.7% were IgG positive for antibody to GspD and all were immunoreactive with LipL32, a known surface exposed in vivo antigen (Fig. 3) [28,33]. MAT negative sera were IgG negative for antibody to all proteins tested and all sera were antibody negative for FhbB. The detection of antibodies to GspD in canines is consistent with the detection of anti-GspD antibodies in the sera of confirmed human leptospirosis patients [37]. It can be concluded that GspD is actively produced during Leptospira infections in canines.

Anti-GspD antiserum is bactericidal and kills through a complement independent mechanism
Bactericidal assays revealed that anti-GspD antibody can kill diverse species and strains of Leptospira of the P1 subclade (Fig. 4). Antibody dependent killing (60-74%) occurred independent of complement ( Fig. 4) with no significant difference in killing in the presence of active NHS vs HI-NHS (P = .41). The mechanism of killing by anti-GspD antibody remains to be determined but could be linked to the critical functions mediated by secretin proteins and T2SS in general. The binding of antibody to GspD may interfere with the passage of effector molecules from the cell into the extracellular milieu [38]. Importantly, anti-GspD antibody effectively killed all strains tested including those of differing serotypes suggesting that GspD or domains derived from it, may be good candidates for inclusion in a multi-protein subunit based vaccine.

Localization of the bactericidal epitopes of GspD
To localize the immunodominant epitopes of GspD that are presented during natural infection, GspD fragments harboring 20 amino acid overlaps were generated (Table 2) and screened with MAT positive canine sera by ELISA (Fig. 5A). The sera reacted most strongly with full length GspD and the F1 fragment (residues 1-135). IgG binding to other fragments was low and variable among individual dogs. The difference in the level of antibody binding to F1 versus all other fragments was significant (P 0.0001). The results demonstrate that the immunodominant epitopes of GspD reside within a segment of the N-terminal N0 domain within amino acids 1 through 135.
To verify that the epitopes of GspD that elicit bactericidal antibody reside within the F1 fragment, bactericidal assays were performed with anti-GspD antisera absorbed with full length GspD, GspD sub-fragments or FhbB prior to conducting the assay (Fig. 5B). Bactericidal activity was significantly reduced by both full length GspD or the F1 fragment (P < 0.01), but not by adsorption with fragments F2, F3, F4, F5 or the irrelevant FhbB antigen. Hence, it can be concluded the bactericidal epitopes of GspD reside within the first 135 amino acids which comprise a majority of the N0 domain. The N0 domain extends from the outer membrane across the periplasm where it interacts with substrates and inner membrane components [19]. Antibody binding to the N0 domain may disrupt T2S and thus explain the basis for the complement independent bactericidal activity of anti-GspD antibody. A BLASTP search using GspD amino acid residues 1 through 135 as the query revealed that the N0 domain of GspD is highly conserved among isolates of subclade P1. This is noteworthy as subclade P1 isolates exhibit greater genome diversity at the intra-clade level than isolates belonging to subclades P2, S1, and S2, and are most commonly associated with severe clinical manifestations of leptospirosis [3].

Conclusions
In this study we have demonstrated that the Leptospira secretin protein, GspD, is expressed in vitro and antigenic during natural infection in canines. Recombinant GspD delivered to rats elicited a high titer IgG antibody response. In vitro assays demonstrated that anti-GspD antibodies are bactericidal and kill through a complement-independent mechanism. The epitopes that elicit bactericidal antibodies were localized within the within the N0 domain of GspD that spans the N-terminal 135 residues. Anti-GspD antibody displayed significant killing activity against antigenically diverse subclade P1 strains including L. interrogans serovars Canicola, Copenhageni and Pomona, L. kirschneri serovar Grippotyphosa, L. noguchii serogroup Autumnalis, and L. borgpetersenii serogroup Ballum. While GspD is conserved amongst subclade P1 isolates, the GspD sequences of other subclades show significant divergence. Hence, a GspD vaccine formulation may need to be multivalent, including the F1 fragment of at least one P1 and P2 subclade isolate. While this study was in review, it was demonstrated that L. interrogans serovar Canicola (LOCaS46) (subclade P1) full length GspD conveyed partial protection to hamsters against lethal challenge with the homologous strain of Leptospira [39]. The data presented here further support the potential utility of GspD as a component of a subunit vaccine for leptospirosis.
A few studies have sought to develop recombinant chimeric antigens that can elicit antibody against multiple Leptospira antigens [40][41][42][43][44]. An OmpL1-LipL32-LipL21 chimeric conferred significant protection against lethal homologous challenge with L. interrogans sv. Lai in hamsters [45] and a LigAc-LenA-LcpA-Lsa23 significantly reduced leptospiral renal burden [46]. However, the potential for these chimeric protein vaccinogens to elicit protection against heterologous strains has not yet been assessed. Recently, a recombinant chimeric epitope based vaccine antigen, referred to as a chimeritope, was successfully developed as a vaccine antigen for canine Lyme disease. The sequences encoding the immunodominant and variable L5 and H5 epitopes from several Borreliella burgdorferi outer surface protein C variants were linked to form a single contiguous gene sequence. Protein expressed from this chimeritope encoding sequence, designated as Ch14, was demonstrated to elicit broadly cross reactive and protective antibodies [27,47]. The Ch14 chimeritope is one of two recombinant antigens that comprise the commercially successful canine Lyme disease vaccine, VANGUARD Ò crLyme (Zoetis) [47,48]. The remarkable diversity of the Leptospira suggests that a chimeritope approach may offer a path to a broadly protective vaccine antigen for both human and veterinary leptospirosis. Future work will seek to   4. Anti-GspD antisera is bactericidal and kills diverse strains through a complement independent mechanism. The bactericidal activity of rat anti-GspD antiserum against diverse Leptospira strains (as indicated in the figure key) was assessed as described in the methods. Abbreviations are as follows: NHS (complement certified normal human serum), HI-NHS (heat inactivated normal human sera), and Rat PI (rat pre-immune serum). Equivalent numbers of cells were incubated with NHS, HI-NHS, rat PI, or anti-GspD antisera as indicated below the figure. Percent killing was determined by visual examination of cell numbers per field of view using dark field microscopy as detailed in the text. Differences in cell counts per field of view in the negative control (NHS + rat PI) versus in the treatment groups (NHS + Anti-GspD and HI-NHS + Anti-GspD) were assessed using the two-tailed student's t-test (* indicates P .05).
incorporate epitopes several leptospiral antigens into a single recombinant chimeritope protein.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.