We assembled by standard PCR and molecular cloning procedures four different constructs expressing TS for this work, introducing the indicated ORFs into the “A” (SmaI/XbaI) site of pIR1SAT, chosen as this site is more highly expressed in amastigotes (Capul et al., 2007a,b) as summarized in Figure 1.; the starting TS gene was TS154/13(Costa et al., 1998). ORFs were amplified with primers bearing an added 5’ SmaI site and a CCACC sequence to enhance expression, and a 3’ XbaI site after the stop codon. WT ORF constructs used the native ATG while the non-secreted constructs removed the first 33 amino acids bearing the predicted signal sequence; these were the aTS+S or aTS−S constructs, respectively. From these, catalytically inactive mutants were generated by site specific mutagenesis, introducing a Y374H mutation into WT (Y343H in the TS−S construct), yielding the iTS+S or iTS−S constructs, respectively. The final DNAs were digested with SmaI and XbaI and the TS-bearing fragments isolated for introduction into SmaI/XbaI-digested pIR1SAT (Figure 1). The relevant regions of all constructs were confirmed by sequencing before use, and the functionality (secretion or activity) was confirmed in transfectants as described in the text. Constructs were digested with SwaI to expose the rRNA segment termini for homologous integration into the parasite prior to use.

L. major strain Friedlin V1 (MHOM/IL/80/ Friedlin) was grown at 26°C in M199 medium (U.S. Biologicals) containing 10% heat-inactivated fetal bovine serum and other supplements as described (Kapler et al., 1990). Leishmania cells were transfected with SwaI-cut DNA by electroporation using high voltage protocol (Robinson and Beverley, 2003). Following transfection, cells were allowed to grow for 16–24 h in M199 medium and then plated on semisolid media containing 1% Nobel agar (fisher) and 100 μg/ml nourseothricin. Individual colonies were picked and grown in liquid medium in same drug concentration as used in plates. Clones were maintained in selective medium and then removed from selection for one passage prior to experiments.

For flow cytometric detection of TS expression, parasites were pelleted, washed with PBS, and fixed by incubation with 1% paraformaldehyde in PBS for 5 min. The fixed parasites were washed again and incubated on ice in methanol for 30 min to permeabilize the cells. After permeabilization, the cells were washed and incubated with sera from T. cruzi-infected animals for 30 min in the dark at 4°C, then washed and incubated with a goat anti-mouse IgG-Alexa Fluor 647 antibody for 30 min in the dark at 4°C. Cells were washed and resuspended in PBS containing 2% fetal bovine serum, and stored protected from light and heat until acquired on a BD FACS Canto II machine. Results were analyzed using FlowJo v7 software (Tree Star, Inc.).

Laemmli buffer was added to parasite lysates and supernatants previously concentrated using 10G spin concentrators (Pierce), and the samples were added to the wells of a 10% acrylamide gel. Recombinant TS protein was used in the assay as a positive control, and Spectra Multicolor Broad Range Protein Ladder was added as a reference (Thermo Fisher). After SDS-PAGE separation and transfer, the nitrocellulose membrane was blocked with 1% milk/PBS for 30 min and then incubated overnight with a rabbit anti-TS antibody diluted in 1% milk/PBS at 4°C. After washing, the blot was incubated with donkey anti-rabbit-HRP diluted in PBS for 1 h, and rinsed. Finally, the membrane was soaked in developer solution with shaking. Once bands appeared, the solution was decanted and the gel rinsed with H₂O.

To generate parasite lysates and supernatants, parasites were cultured overnight in M199 medium at 1 × 10⁶ parasites per well of a 6-well plate. The following day, the parasites were pelleted by centrifugation. Supernatants were aspirated and concentrated using 10 g spin concentrators (Pierce), and the remaining parasite pellet was lysed. TS enzymatic activity was measured as described previously (Schrader and Schauer, 2006; Eickhoff et al., 2016). Supernatants and lysates were incubated with 1 mM 3′-sialyllactose, a donor substrate simulating the sialic acid residues bound to host cell glycoproteins; 0.5 mM 4-methylumbelliferyl-β-D-galactoside (MUGal), a sialyl acceptor substrate; and 100 mM PIPES medium at 20–26°C for 45 min. Samples were added to spin columns containing previously equilibrated Q-sepharose ion exchange resin and washed and eluted HCl. The flow-through was neutralized with Glycine/NaOH buffer. Functional TS activity creates 4-methylumbelliferyl-β-D-sialylgalactoside, the sialylated form of MUGal from the substrates, which results in a fluorescent substrate following acid hydrolysis. The resulting fluorescence correlates with TS activity and is measured by a microplate reader.

All experiments were performed using wild-type BALB/c mice purchased from NCI Charles River Laboratories (Frederick, MD). All mice were housed under specific pathogen free conditions under the care of licensed veterinarians and animal technicians from the Department of Comparative Medicine. Female mice aged 6 to 12 weeks were used for these studies. Small sample sizes of 2–6 mice per group were selected for these experiments based on previous experiences and the exploratory nature of the aims. All protocols and experiments were reviewed and approved by the Saint Louis University Animal Care Committee in adherence to AAALAC guidelines and recommendations.

Leishmania major parasites were passaged through WT BALB/c mice and maintained in culture in M199 medium for less than 10 passages before use. Infectious metacyclics were isolated using a Ficoll gradient (Spath and Beverley, 2001). Log-phase FV1 parasites were resuspended in DMEM, then added over 10% Ficoll overlaying 20% Ficoll in a sterile tube. After a 15 min spin at 2,500 rpm, the top fraction containing metacyclics was aspirated, washed and resuspended in PBS. For vaccinations, L. major parasites in 10 μl PBS were injected intradermally into the hind footpad of mice anesthetized with Ketamine/Xylazine (60 and 5 mg/kg, respectively). The Tulahuèn strain of T. cruzi was passaged through Dipetalogaster maximus insects and BALB/c mice and maintained in culture in LDNT+ medium (to simulate the foregut) and enriched Grace’s insect medium (to simulate the hindgut). For challenges, T. cruzi blood-form trypomastigotes (BFTs) were isolated from the blood of highly infected animals, resuspended in 100 μl of PBS and injected subcutaneously at the base of the tail. Inoculum doses were chosen based on relative virulence of the host strain, titration experiments comparing immunity and pathology resulting after inoculation with various doses (not shown), and prior experience in our lab.

Total spleen cells (SCs) and popliteal lymph node cells (pLNCs) were isolated from mice via mechanical disruption and NH₄Cl lysis of red blood cells. Mixed cellulose ester bottom filter ELISPOT plates (Millipore) were coated with 10 μg/ml of α-IFN-γ clone R46A2 (BD Pharmingen) and blocked with 10% FBS. Total SCs (3 × 10⁵) were seeded to each well along with 1 × 10⁵ control A20 antigen presenting cells (A20 NC), A20 cells transfected with the TS enzymatic domain (A20 TS), or L. major lysate (1 × 10⁶ parasite equivalents/well). After overnight incubation, cells were lysed with H₂O, and plates were incubated with biotin α-IFN-γ clone XMG1.2 (BD Pharmingen) for 2 h at room temperature, then streptavidin conjugated to horseradish peroxidase for 1 h at room temperature. Spots were visualized through the precipitation of 3-amino-9-ethylcarbazole substrate.

Serum samples from vaccinated mice were diluted 1:10 in PBS and then serially diluted 1:3. Diluted serum samples were incubated overnight on NUNC Maxisorp plates previously coated in 5 μg/ml of recombinant TS protein and blocked with 10% FBS. The next day, plates were washed and incubated with 1/5000 dilution of rat α-mouse IgG HRP for 2 h at room temperature, then washed and developed with TMB substrate. Reactions were stopped with H₂SO₄, and plates were read at 450 nm.

Cells were cultured with monensin (GolgiPlug, 1 μg/ml) and Brefeldin A (GolgiStop, 0.67 μg/ml) at 37°C for at least 3 h to trap intracellular cytokines. Cells were then fixed and permeabilized with Foxp3/Transcription Factor Staining Buffer Set (eBioscience), then incubated with fluorochrome-conjugated antibodies directed against intracellular cytokines for 30 min at 4°C. After staining, cells were washed and stored in the dark at 4°C until they were acquired by a BD LSR II flow cytometry.

We selected an active T. cruzi TS gene (Costa et al., 1998) and inserted its 678 aa catalytic domain into the strong Leishmania expression vector pIR1SAT (Capul et al., 2007a), inserting a CCACC sequence conferring high translatability upstream of the start codon, as described in the methods (Figure 1A). To explore TS secretion when expressed heterologously and its impact on the immune response, we expressed a truncated TS lacking the presumptive 33 amino acid secretory domain, and to explore the requirement for TS enzymatic activity, we engineered a catalytic site inactivating point mutation (Y374H for WT) into both the WT and truncated non-secretory TS. These constructs are termed pIR1SAT-aTS+S (WT TS – active+signal sequence), pIR1SAT-aTS−S (active, lacking signal sequence), pIR1SAT-iTS+S (inactive+signal sequence) or pIR1SAT-iTS−S (inactive lacking signal sequence).

Each construct was digested with SwaI and transfected into L. major; using the fully virulent strain FV1, a partially attenuated strain CC1, or the avirulent dihydrofolate reductase, thymidylate synthetase double knock-out CC1 (dhfr-ts⁻ /CC1; Figure 1D). SwaI digestion exposes sequences targeting the construct into the parasite rRNA gene locus, where high levels of expression of active mRNA arises from the strong rRNA promoter and processing to mature TS-expressing mRNAs by the parasites’ trans-splicing and polyadenylation machinery (Capul et al., 2007a, 2007b). Successful transfection and integration into the rRNA locus were confirmed by PCR tests (not shown) and numerous clonal lines were obtained, of which several were taken for subsequent analysis. Formally these transfectants were termed SSU:IR1SAT-aTS+S, etc., to mark the integration status.

We screened the transfected FV1 parasites for the expression of TS using flow cytometry, western blotting, and enzyme activity assays. High levels of intracellular TS expression were detected by flow cytometry and by western blotting within parasites expressing the non-secreted forms of TS (aTS−S; Figures 2A,B). Conversely, much less intracellular TS was detected in parasites expressing Tc TS retaining its native secretion signal (aTS+S; Figures 2A,B). These data confirmed the functionality of the TS signal sequence when expressed in Leishmania, and the loss of secretion in its absence.

TS enzymatic activity was determined in parasite cellular lysates or culture supernatants. As expected, enzymatic activity was seen only from transfectants expressing the “active” TS mutants (SSU:IR1SAT-aTS+S, -aTS−S) but not the inactive mutants (Figures 2C,D). For the active TS, activity was seen mostly in the supernatant for the WT secreted TS (SSU:IR1SAT-aTS+S), and mostly in the cell lysate for the WT non-secreted TS (SSU:IR1SAT-aTS−S; Figures 2C,D). These data indicate that our recombinant Lm Tc expression constructs predominantly express TcTS with the predicted properties (secreted or not, catalytically active or not).

The rTS-Lm vaccines based on the parent Friedlin strain FV1 generated above were expected to cause some pathology in mice. However, they also presented us the opportunity to quickly assess the immunogenicity of a persistent leishmania-based T. cruzi vaccine in an animal model, even though such a persistent parasite backbone would not be amenable for use in humans. We asked whether expression of the Tc TS forms could induce TS-specific immune responses in vivo, and ultimately, protection against a T. cruzi challenge. We first vaccinated wild-type BALB/c mice with 1,000 FV1 strain L. major parasites transfected with the empty pIR1SAT vector as a negative control (NC), aTS+S, or aTS−S via intradermal injection into the hind footpad. Some animals developed severe footpad swelling and required euthanasia (not shown). Representative surviving mice were sacrificed 3 months after vaccination for immune studies. From these mice, total SCs and draining pLNCs were isolated and re-stimulated in IFN-γ ELISPOT assays to measure the development of antigen-specific responses.

SCs and pLNCs from mice vaccinated with NC rLm, aTS+S rLm, or aTS−S rLm produced IFN-γ in response to co-culture with L. major lysate, indicating that rLm vaccination induced Leishmania-specific memory T cell responses (Figures 3A,B). In addition, SCs and pLNCs from mice inoculated with aTS+S rLm produced IFN-γ after stimulation with A20 APCs transfected with the TS enzymatic domain of TS, indicating successful induction of TS-specific T cell immunity in these animals (Figures 3A,B). However, TS-specific IFN-γ responses were not produced by SCs from mice vaccinated with either NC rLm or aTS−S rLm, and only minimal IFN-γ responses to TS were detected among the pLNCs of mice given aTS−S rLm (Figures 3A,B).

While only mice inoculated with aTS+S rLm developed large numbers of TS-specific IFN-γ-producing SCs, vaccination with either aTS+S or aTS−S rLm resulted in the generation of high titers of TS-specific IgG antibody compared to control vaccination (Figure 3C). These studies demonstrate that vaccination with aTS+S rLm expressing TS can induce T cell and antibody responses to both Leishmania antigens and T. cruzi TS in vivo. Vaccination with aTS−S rLm resulted in similar immune responses to L. major and high TS-specific IgG titers, but did not result in substantial numbers of SCs that could produce IFN-γ in response to TS, perhaps because of decreased delivery of TS antigen to draining lymph nodes or induction of alternative inflammatory profiles not measured, such as Th17 cell responses.

To study protection, the remaining rLm vaccinated mice were challenged with 5,000 highly virulent T. cruzi blood-form trypomastigotes recovered from infected animals. Two of the three mice vaccinated with either aTS+S rLm or aTS−S rLm survived long-term after T. cruzi challenge, compared to none of the mice vaccinated with NC rLm. These results indicate that both TS-expressing vaccines can protect against T. cruzi challenge, and suggest a protective role for TS-specific antibodies as the common denominator induced by rLm vaccines (Figures 3C,D). T cell IFN-γ production is an important endpoint, because activated CD8+ T cells are generally considered the ultimate effector cells for the control of T. cruzi infection. However, survival among mice vaccinated with aTS−S rLm despite the lack of IFN-γ responses detectable by ELISPOT suggests that alternative mechanisms, such as induction of sufficiently strong TS-specific antibody titers capable of neutralizing blood form trypomastigotes, may be enough to protect against future challenge.

Our vaccination experiments using the virulent FV1 rLm demonstrate that recombinant TS-expressing Lm parasites can indeed induce protection against T. cruzi infection, but this strain retains pathogenicity and causes significant skin lesions at the site of injection, which necessitated the sacrifice of many test mice prior to T. cruzi challenge.

The promising experiments above were performed using a virulent rLm background, necessitating the use of a low dose immunization protocol which nonetheless occasionally yielded strong pathology from L. major itself. Previous studies have shown that the CC1 strain of Lm is less virulent and less likely to cause severe pathology in the host than the FV1 strain (Titus et al., 1995). In addition, a mutant CC1 strain containing a double knock-out for both the dihydrofolate reductase (dhfr) and thymidylate synthetase (ts) genes (dhfr-ts⁻ CC1) has already been developed and characterized as a safe and effective vaccine against virulent L. major infection in mice (Titus et al., 1995).

To test whether the CC1 background on which the dfr-ts- mutant exists can also serve as an effective vehicle for TS vaccination, we inoculated mice in the hind footpad with 10⁶ CC1 NC, aTS+S, or aTS−S rLm parasites. This dose was chosen as it led to little Leishmania-induced pathology (data not shown). Approximately one month afterward, representative mice from each group were sacrificed for immune studies. TS-specific IFN-γ T cell responses were detected in both SCs (Figure 4A) and pLNCs (Figure 4B) of mice vaccinated with either aTS+S or aTS−S rLm compared to those vaccinated with NC rLm, although stronger T cell responses were detected in mice vaccinated with aTS+S rLm, consistent with the FV1 vaccination data. However, both aTS+S rLm and aTS−S rLm generated similarly high TS-specific antibody titers (Figure 4C), again consistent with the FV1 vaccination data. Surprisingly, despite inducing lower TS-specific IFN-γ responses compared with aTS+S rLm vaccination, aTS−S rLm conferred the greatest survival benefit following a normally lethal T. cruzi challenge (Figure 4D), suggesting that other responses not measured, including Th17 type cytokines, may mediate protection. In contrast, a trend for increased immunopathology in aTS+S rLm-vaccinated animals was seen, as some mice needed to be humanely euthanized due to severe necrotizing lesions at the rLm inoculation site.

Because of the increased safety profile of the double knock-out, mutant dhfr-ts⁻ Lm strain, it has the most potential to be applicable in humans. To test the dhfr-ts⁻ rLm as T. cruzi vaccines, we inoculated WT BALB/c mice with 10⁶ log-phase WT or dhfr-ts⁻ CC1 NC or aTS+S parasites by injection into the hind footpad. Approximately 5 weeks later, representative mice were euthanized for immune studies. No IFN-γ responses to TS stimulation were detected by ELISPOT assay among immune cell populations isolated from mice vaccinated with dfhr-ts- rLm expressing NC or aTS+S (not shown).

In our experiments, dhfr-ts⁻ CC1 rLm parasites could be recovered from the footpads of vaccinated animals on day 3 after injection but not on day 7, when WT CC1 parasites are routinely recovered, indicating a limited timeframe of dhfr-ts⁻ CC1 rLm parasite persistence as described previously (Titus et al., 1995). We hypothesized that the limited persistence of dhfr-ts⁻ CC1 rLm parasites may limit the amount of TS antigen secreted and able to prime relevant immune cells, so we investigated a prime-boost vaccination strategy to amplify TS-specific immune responses. We vaccinated mice with 10⁶ log-phase dhfr-ts⁻ CC1 NC or dhfr-ts⁻ aTS+S CC1 parasites once a week for 3 weeks, then boosted these mice with another dose 3 days prior to sacrifice at 4 weeks for immune studies (Figure 5A). Repeated prime-boost vaccination with dhfr-ts⁻ aTS + S did not generate high titers of TS-specific IgG antibodies compared to vaccination with control parasites (Figure 5B). We also did not detect any TS-specific responses among SCs by IFN-γ ELISPOT assays. However, in ICS assays, we measured some increased expression of IFN-γ among CD8+ T cells recovered from mice vaccinated with dhfr-ts⁻ aTS+S rLm (Figures 5C,D).

None of the mice vaccinated with dhfr-ts⁻ CC1 rLm developed lesions despite multiple injections, consistent with prior studies where very high inocula were required (Titus et al., 1995), while all mice receiving WT CC1 vaccination developed significant swelling within 2 months, with some developing ulcerative lesions (although not to the degree seen with FV1 infections). Thus, we confirmed the avirulent and nonpathogenic nature of the dhfr-ts⁻ CC1 strain compared to the WT CC1 strain. However, the data suggest this strain is not immunogenic enough to induce responses to the recombinant T. cruzi antigen that it encodes. The results of these experiments indicate that future work should explore methods of increasing the immunogenicity of the partially attenuated dhfr-ts⁻ CC1 strain to render it both a safe and effective vehicle for the delivery of vaccine antigens.

We inoculated mice with equal numbers of WT control L. major FV1 parasites or recombinant parasites expressing the aTS+S, aTS−S, or iTS+S enzymes. Mice inoculated with rLm expressing active TS, regardless of whether the TS was secreted or non-secreted, developed lesions 1 week earlier than mice injected with WT L. major not expressing TS (day 14 compared to day 21, Figure 6A). After lesions appeared, however, the progression was not significantly different between the groups. The kinetics of lesion development were associated with increased expression of sialic acid on the surface of promastigote parasites expressing active TS (Figure 6B). Although the surface sialylation was studied on promastigote recombinants, we expect similar sialylation on amastigotes because of the nature of the expression system used. In contrast, mice inoculated with iTS+S rLm only developed lesions on day 21, similar to WT FV1 parasites, indicating that the expression of enzymatically inactive versions of TS does not result in earlier lesion development.