Structure of SALO, a leishmaniasis vaccine candidate from the sand fly Lutzomyia longipalpis

Background Immunity to the sand fly salivary protein SALO (Salivary Anticomplement of Lutzomyia longipalpis) protected hamsters against Leishmania infantum and L. braziliensis infection and, more recently, a vaccine combination of a genetically modified Leishmania with SALO conferred strong protection against L. donovani infection. Because of the importance of SALO as a potential component of a leishmaniasis vaccine, a plan to produce this recombinant protein for future scale manufacturing as well as knowledge of its structural characteristics are needed to move SALO forward for the clinical path. Methodology/Principal findings Recombinant SALO was expressed as a soluble secreted protein using Pichia pastoris, rSALO(P), with yields of 1g/L and >99% purity as assessed by SEC-MALS and SDS-PAGE. Unlike its native counterpart, rSALO(P) does not inhibit the classical pathway of complement; however, antibodies to rSALO(P) inhibit the anti-complement activity of sand fly salivary gland homogenate. Immunization with rSALO(P) produces a delayed type hypersensitivity response in C57BL/6 mice, suggesting rSALO(P) lacked anti-complement activity but retained its immunogenicity. The structure of rSALO(P) was solved by S-SAD at Cu-Kalpha to 1.94 Å and refined to Rfactor 17%. SALO is ~80% helical, has no appreciable structural similarities to any human protein, and has limited structural similarity in the C-terminus to members of insect odorant binding proteins. SALO has three predicted human CD4+ T cell epitopes on surface exposed helices. Conclusions/Significance The results indicate that SALO as expressed and purified from P. pastoris is suitable for further scale-up, manufacturing, and testing. SALO has a novel structure, is not similar to any human proteins, is immunogenic in rodents, and does not have the anti-complement activity observed in the native salivary protein which are all important attributes to move this vaccine candidate forward to the clinical path.


Conclusions/Significance
The results indicate that SALO as expressed and purified from P. pastoris is suitable for further scale-up, manufacturing, and testing. SALO has a novel structure, is not similar to any human proteins, is immunogenic in rodents, and does not have the anti-complement activity observed in the native salivary protein which are all important attributes to move this vaccine candidate forward to the clinical path.

Author summary
Immunity to sand fly salivary proteins has been shown to confer protection against leishmaniasis in rodent models. Recombinant salivary protein SALO (Salivary Anticomplement of Lutzomyia longipalpis) was shown to protect hamsters against the fatal outcome of visceral leishmaniasis caused by Leishmania infantum and to protect against cutaneous leishmaniasis caused by Leishmania braziliensis. Because of the potential use of this sand fly salivary protein as a component of a vaccine against human visceral leishmaniasis further characterization of SALO needs to be performed as well as a development plan for future scale manufacturing. In this work we present the successful expression and purification of recombinant SALO using Pichia pastoris. SALO from insect saliva inhibits the classical pathway of complement, an activity that may interfere with its role as a vaccine candidate. Here we show that recombinant SALO produced from Pichia pastoris, rSALO (P), does not have the anti-complement activity, and antibodies against rSALO(P) inhibit

Introduction
Sand flies are vectors of Leishmania parasites, causal agents of the neglected tropical disease (NTD) leishmaniasis, which is ranked among the most important NTDs in terms of global disease burden [1] and is re-emerging at alarming rates due to the ongoing conflicts in the Middle East and North Africa [2]. To date, there is no effective or licensed vaccine against human leishmaniasis, despite studies demonstrating the cost-effectiveness of developing such biotechnologies for use in resource-poor settings [3,4].
Sand flies deliver saliva into the skin of the host while probing for a blood meal. The saliva of blood feeding arthropods, including sand flies, has a number of potent bioactive molecules, such as anticoagulants, vasodilators, and inhibitors of platelet aggregation that assist in acquiring blood meals. In the case of sand flies, some of these bioactive components also modify the immunological environment at the host skin site of bite, favoring the establishment of Leishmania infection in the host [5,6]. Other biological activities of sand fly saliva have been reported and the proteins responsible for these effects have been identified [7], including Lundep, an endonuclease that destroys neutrophil traps [8]; salivary yellow proteins that bind biogenic amines [9]; and recently SALO (Salivary Anti-complement from Lutzomyia longipalpis), an inhibitor of the classical pathway of complement [10].
Although sand fly saliva was shown to exacerbate Leishmania infection, immunity to sand fly saliva protects against Leishmania infection [6,11]. The protection is correlated to the induction of a Th1 cellular immune response. Recently, a number of salivary proteins have emerged as vaccine candidates against cutaneous leishmaniasis, including PpSP15 [12] and Linb11 [13] in rodent models, and PdSP15 in non-human primates [14]. For visceral leishmaniasis, the salivary proteins LJM17 and LJL143 were shown to induce a Th1 immune response in dogs [15]. Immunization with DNA plasmid coding for the salivary protein SALO (formerly known as LJM19) was shown to protect hamsters against the fatal outcome of visceral leishmaniasis [16] and this protection was correlated with the induction of a Th1 cellular immune response [16]. Recently, a combination vaccine comprised of recombinant salivary protein SALO and a genetically modified Leishmania donovani resulted in a strong protection against visceral leishmaniasis [17], further suggesting SALO as an important component for a visceral leishmaniasis vaccine.
In this study we present the production and crystal structure of SALO genetically engineered in yeast as part of efforts to develop it as a recombinant vaccine for leishmaniasis.

Ethics statement
All animal procedures were reviewed and approved by the National Institute of Allergy and Infectious Diseases (NIAID) Animal Care and Use Committee, under animal protocol LMVR4E, and handled in accordance to the Guide for the Care and Use of Laboratory Animals and with the NIH OACU ARAC. Further, the animal protocol is in full accordance with 'The guide for the care and use of animals' as described in the US Public Health Service policy on Humane Care and Use of Laboratory Animals 2015 (http://grants.nih.gov/grants/olaw/ references/phspolicylabanimals.pdf).

Cloning, expression and purification of recombinant SALO in Pichia pastoris
DNA coding for SALO without signal peptide was codon optimized based on Pichia pastoris usage preference and subcloned into Pichia secretory expression vector pPICZαA (Invitrogen) using EcoRI/XbaI restriction sites. The correct insert sequence and reading frame of recombinant plasmid was confirmed by double-stranded sequencing using vector flanking primers αfactor and 3'AOX-1 and then transformed into Pichia pastoris X-33 by electroporation. The expression of rSALO(P) was induced with 0.5% methanol at 30˚C for 72 hours and the highest expression clone was chosen for making seed stock with 20% glycerol. The large-scale expression of hexa histidine tagged rSALO(P) was induced with methanol in a 10L fermentor and purified by immobilized metal affinity chromatography (see S1 Text).
Cloning and production of recombinant SALO in HEK293-F Cloning, expression and purification of SALO was performed as previously described [10]. Briefly, DNA coding for SALO without the signal peptide and containing a C-terminal hexahistidine tag was synthesized by Eurofins genomics (Huntsville, AL). The synthesized gene was subcloned into the VR2001-TOPO expression vector. The transfection into HEK 293-F cells and expression of rSALO(H) was performed at the Protein Expression Laboratory at the Frederick National Laboratory for Cancer Research (Frederick, Maryland). The supernatant was recovered after 72 hours, concentrated, and buffer exchanged into PBS pH 7.4 using a 10K Amicon concentrator device (Millipore). The protein was purified by immobilized metal ion affinity chromatography in the same buffer and eluted with imidazole.

Sand fly Salivary Gland Homogenate (SGH) preparations
Salivary glands were dissected from Lu. longipalpis sand flies obtained from the Vector Molecular Biology Section, LMVR, NIAID, NIH as previously described [18].

Antibodies to rSALO
Six to eight weeks old female Balb/c mice were injected intradermally in the ear three times every 15 days with 2 μg of rSALO(P) mixed (1:1 volume) with Magic Mouse Adjuvant (Creative Diagnostics, Shirley, NY) as recommended by the manufacturer. Fifteen days after the last inoculation, blood was collected to obtain the rSALO antiserum.

Immunization of mice with rSALO and DTH measurements
Six to eight weeks old female Balb/c mice were injected intradermally in the right ear three times every 15 days with 2 μg of rSALO(P) (without adjuvant). The endotoxin level of rSALO (P) was 0.00127 Endotoxin Units per injection. Delayed type hypersensitivity response or skin immune response was measured in the ear of C57Bl/6 mice as previously described [19,20]. The mouse ear thickness and redness were used as an indicator of a cell-mediated immune response to rSALO(P) [19,20]. Briefly, ear thickness from the from the dorsal to the ventral portion of the ear was measured using a Digital Vernier caliper (Mitutoyo Corp.) at 24 and 48 h following intradermal injection of rSALO(P). Measurements were taken for five mice in each group and repeated at least twice.

Statistical analysis
Statistical analysis was performed using the GraphPad Prism software. Multiple groups were analyzed using one-way analysis of variance followed by Tukey's multiple-comparison test.

Size-Exclusion Chromatography and Multi-Angle Light Scattering (SEC-MALS)
The average molecular weight of the SALO protein was determined by SEC-MALS. The system consisted of an Agilent 1260 Infinity series HPLC, coupled with a UV detector (Agilent), a miniDAWN triple-angle light-scattering detector (Wyatt Technology), and an Optilab rEX differential Refractive Index (dRI) detector (Wyatt Technology). 40 μg of SALO was loaded into a TSK gel Super SW2000 column (TOSOH Biosciences, King of Prussia, PA) and eluted at 0.35 ml/min isocratically with Tris-HCL pH 8 for 30 min. Protein constants were 0.185 mL/ g and 0.911 mL/(mgÁcm) for dRI and UV detectors, respectively. Data collection and analysis was done with Wyatt's ASTRA 6.1.1 software.
Crystallization and data collection rSALO(P) was buffer exchanged and concentrated to 24 mg/ml in 50 mM Tris HCl pH 8.0 using a 5K MW cutoff centrifugal concentrating device (Millipore). The initial protein concentration was confirmed by measurement of OD 280 prior to setting up crystallization experiments. Crystallization conditions were screened using commercial screens from Hampton Research at 298K. Crystals were grown by vapor diffusion in sitting drops, which were equilibrated against well containing 0.5 ml crystallization solution. Drops were prepared by mixing 1.5 μl of protein solution with an equal volume of crystallization solution. No crystals were obtained for protein produced in mammalian cells, possibly because of the presence of the Nterminus vector derived sequence. rSALO(P) crystallized within 16 hours from a precipitant solution containing 0.02M calcium chloride, 30% v/v MPD and 0.1 M sodium acetate pH 4.6. Larger crystals with dimensions 0.8 mm X 0.5 mm X 0.3 mm were obtained within 48 hours by setting up larger drops using a ratio of 4 μl of protein to 1.5 μl of the same precipitant solution.
Since crystals grew in solutions that contained adequate cryoprotectant, they were flashcooled directly in a stream of N 2 gas at 113 K prior to collecting diffraction data. X-ray diffraction data were collected at the Baylor College of Medicine core facility (Rigaku HTC detector, Rigaku FR-E+ SuperBright microfocus rotating anode generator, with VariMax HF optics) using the Crystal Clear (d Ã trek) package [21]. Data was integrated using MosFLM and scaled with SCALA [22]. Crystallographic data is shown in Table 1.

Structure determination
The structure of rSALO(P) was solved using single-wavelength anomalous dispersion with the anomalous signal from sulfur at Cu-K alpha wavelength. FA values were calculated using the program SHELXC [23]. Based on an initial analysis of the data, the maximum resolution for substructure determination and initial phase calculation was set to 1.94 Å. The location of 89 atoms (C, S, N, O) were automatically determined using the program SHELXD [23] and based on the results of this automated search 82.08% of the model was built using the program ARP/ wARP [24,25]. Since the difference between R factor and R free remained unreasonably high, the structure was subsequently refined in a lower symmetry space group with a dimer in the asymmetric unit. The final model was obtained by iterative manual model building cycles using the program Coot [26] followed by structure refinement with REFMAC5 [27] [28] and PHENIX [29]. Structural figures were generated using PyMOL [30]. The refined coordinates and structure factors have been deposited in the RCSB protein databank under accession code 4LU2.

Prediction of T cell epitopes
T-cell epitope was predicted for full SALO sequence using the program NetMHC II release 2.2 [31]. The program was set to default parameters that allow identification of 15-mer amino acid peptides with predicted binding affinity below 50 nM to MHC II alleles [32]. For predicted epitopes for the same MHC II allele with sequence length overlap higher than 50%, the peptide with the highest affinity score was kept. Graphs were built using in-house Perl scripts.

Results
SALO produced from Pichia pastoris, rSALO(P), is pure and monodisperse SALO was produced in Pichia pastoris to establish a feasible process for a product and clinical development path. This includes testing the immunogenicity of this salivary protein and resolving its crystal structure. Recombinant SALO was expressed as a soluble protein with a vector derived EF on the amino terminus using Pichia pastoris after 72 hours of methanol induction. Typical yields of rSALO(P) by a single immobilized metal affinity chromatography purification step were~1g/L, which is 500 times higher than the 2.0 mg per L for rSALO(H) produced in HEK293 cells. Purified rSALO(P) appeared to be~99% pure ( Fig 1A). The electrophoretic mobility of~15kDa is likely due to the charge of the molecule and not due to posttranslational modifications because the molecular weight of SALO determined by SEC-MALS (11.8 kDa) is close to the theoretical molecular mass of 11.9 kDa ( Fig 1B). rSALO(P) elutes at 18.2 min as a single, monodisperse peak with a calculated molecular weight of 11.8 kDa ( Fig  1B) which agrees with the theoretical molecular mass (11.9 kDa) of monomeric rSALO. Recombinant SALO produced from HEK cells, rSALO(H), elutes off the sizing column as two overlapping peaks (Fig 1C). The main peak at 18.3min is~83.9% of all the protein components has a molecular weight of 12.6 kDa, and a minor overlapping peak (~15.5%) at 17.8 min with molecular weight of 13.7 kDa (Fig 1C). The theoretical molecular weight of rSALO(H) is 12.2 kDa. Thus, rSALO(P) is pure and exclusively monomeric in solution (monodisperse), which will simplify the downstream process for the production of a recombinant biologic for clinical development. Antibodies against rSALO(P) inhibit the anti-complement activity of sand fly salivary gland homogenate It was previously shown that rSALO(H) and SGH of Lu. longipalpis containing SALO inhibited the classical pathway of complement [10]. rSALO(P) did not inhibit the classical pathway of complement, in contrast to rSALO(H) which inhibits the classical pathway of complement (S1 Fig). Nevertheless, antibodies produced against rSALO(P) neutralized Lu. longipalpis SGH anti-complement activity in a dose dependent manner (Fig 2A). Importantly, antibodies raised against rSALO(P) recognized both rSALO(P) and rSALO(H), and a single band from Lu. longipalpis SGH, by Western blot (Fig 2B). Furthermore, rSALO(P) had similar immune recall responses as rSALO(H) (S2A and S2B Fig)

SALO is a small helical protein
The crystal structure of SALO was refined with a dimer in the asymmetric unit in the space group P 4 2 with statistics shown in Table 1. We chose the dimer because refining the structure of SALO as a monomer in a higher symmetry space group (P 4 2 2 1 2), resulted in >12% difference between R Factor and R Free and increased disorder in loop regions. The SALO dimer ( Fig  3A) appears to be crystallographic and PISA analysis shows no appreciable buried surface area at the dimer interface. SALO is~80% alpha helix and~20% loop. Each SALO monomer has an overall topology comprised exclusively of helices, stabilized by disulfide bonds and connected by short loops (Fig 3B). SALO also has large segregated exposed charged regions (Fig 3C).
Predicted SALO T cell epitopes are exposed on the surface of the protein There are three predicted human T cell epitopes in SALO, EDCENIFHDNAYLL (peptide 1), VAKIIRECIAQVSTQ (peptide 2) and KFSEIYDCYMKKKIC (peptide 3), ( Table 2). All three epitopes are located on surface exposed helices (Fig 4).
The structure of SALO is unique and could not have been predicted from any known protein structures. SALO is comprised entirely of helices and belongs to the all-alpha protein class with EF hand like fold. Pfam analysis using PDBSum (http://www.ebi.ac.uk/pdbsum/) reveals that the C-terminal of SALO (residues 50-105) contains the Pfam domain family PF01395,  otherwise known as the odorant-binding domain of insect proteins. Members of this family have limited sequence identity and their proposed shared function is to bind insect pheromones or odorants [33,34]. Additional studies are required to clarify if SALO can indeed bind odorants. The structure of the salivary protein PdSP15 from the sand fly Phlebotomus duboscqi has been reported, and like SALO, its structure is all helices connected by loops [35]. While SALO only shares 19.6% sequence similarity to PdSP15, their C-terminal odorant binding domains superpose quite well (Fig 5A-5C). Additionally, secondary structure alignment reveals a series of conserved residues including disulfide bonds that connect the central helices (Fig 5C & 5D). Interestingly, two of the three predicted T-cell epitopes (peptide 2 and peptide 3) are located in the structurally conserved C-terminus odorant-binding domain (Fig 5D). It remains unknown what roles these conserved residues play in odorant binding or the functions of SALO and similar proteins.

Discussion
Immunization with sand fly salivary protein SALO protects against leishmaniasis (visceral and cutaneous), either as a DNA vaccine or as a recombinant protein [16,37]. Furthermore, this vaccine candidate was recently shown to inhibit the classical pathway of complement [10]. To move this vaccine candidate towards the clinical path, we solved the structure of SALO and expressed it in P. pastoris. Further, we developed a process for pilot production of SALO. Results from the current work demonstrate the feasibility of expressing SALO in P. pastoris for future scale manufacturing. We previously showed that rSALO(H) has anti-complement activity. In this work, we reproduce this finding (S1 Fig) and demonstrate that rSALO(P) lacks anti-complement activity, and is monomeric. Interestingly, both native SALO from salivary glands and rSALO(H) form multi-species as previously shown by Western blot [10]. It is possible that some or one of these multiple species are required for anti-complement activity, and that the single species observed in rSALO(P) may not be the active form of the protein.
Regardless, rSALO(P) has the desirable features of a vaccine candidate: it is monomeric, monodisperse and does not have anti-complement activity while retaining its immunogenicity. After immunization, both rSALO(H) and rSALO(P) induced a robust delayed hypersensitivity response in mice. Of note, though rSALO(P) lacks anti-complement activity, antibodies against it inhibit the anti-complement activity from Lu. longipalpis sand fly salivary gland homogenate, suggesting that the overall structure of SALO is conserved regardless of the expression source. Structure of SALO, a sand fly salivary protein vaccine candidate Structure of SALO, a sand fly salivary protein vaccine candidate Our structural analyses also reveal that SALO does not share any appreciable tertiary or quaternary structural similarity to any known mammalian protein families. Furthermore, SALO is not found in any other insect vectors or other organisms [38], displaying appreciable sequence homology only to proteins found in New World sand flies of the genus Lutzomyia and Nyssomyia [39]. Interestingly, our current studies reveal that SALO and PdSP15, another salivary vaccine candidate that was previously shown to protect non-human primates against vector-transmitted L. major infection [14], have conserved structural features in their odorant binding protein domain, which contains two of the three predicted CD4 + T cell epitopes, strongly suggesting that the odorant binding protein domain may be relevant for their immunogenic properties. Both SALO and PdSP15 produce a robust cellular immune response that is protective against leishmaiasis [14,16]. In light of our current findings, further studies are necessary to determine the importance of the odorant binding protein domain in the antigenicity of these salivary proteins.
In summary, this work demonstrates that rSALO(P) is suitable for further scale-up, manufacturing, and testing as a vaccine candidate against leishmaniasis. The structure of SALO is novel and unique to sand flies with no resemblance to any protein sequence or structure from humans. rSALO(P) retains its immunogenicity and importantly it lacks anti-complement activity, overcoming a potential obstacle for its development as a vaccine. The attributes of recombinant rSALO(P) and its feasibility for future large-scale production make this molecule an attractive target as a component of a Leishmania vaccine for humans. C57bl/6 mice were immunized in the ear with 2 μg of recombinant SALO produced in P. pastoris (p) or SALO produced in HEK cells (m). Two weeks (A) or 4 weeks (B) after the last immunization 2 μg of recombinant protein was injected and induration and redness in the ear was measure at 48 hours. Naïve mice (control) were injected with PBS. The data represents the mean ± standard deviation of a representative experiment with 5 mice per group of two independent experiments (ANOVA and Tukey test). (PDF)