An ELISA-based method for Galleria mellonella apolipophorin-III quantification

Strains and culturing conditions

The heterologous apoLp-III expression was performed in Escherichia coli BL21 Star (DE3) cells (Thermo Fisher Scientific, Waltham, MA, USA). Cells were grown in LB broth (0.5% (w/v) yeast extract, 1% (w/v) gelatin peptone, and 0.5% (w/v) NaCl) at 37 °C. When required, media was supplemented with 2% (w/v) bacteriological agar and 100 μg mL⁻¹ ampicillin (Sigma-Aldrich, St Louis, MO, USA). To challenge G. mellonella, the fungal strains used were Candida albicans wild-type (WT) SC5314, kex2Δ null mutant, C. albicans kex2Δ/KEX2 mutant (strains CNA1 and CNA2, respectively) (Staniszewska et al., 2020), och1Δ null mutant, och1Δ + OCH1 reintegrant control (Bates et al., 2006), Sporothrix schenckii ATCC MYA-4821, Sporothrix brasiliensis ATCC MYA-4823, and Sporothrix globosa FMR-9624 (Castro et al., 2013; Madrid et al., 2009). Propagation of C. albicans strains was performed in Sabouraud broth (1% (w/v) meat peptone and casein, and 4% (w/v) glucose) at 28 °C and 120 rpm. S. schenckii, S. brasiliensis, and S. globosa yeast-like cells were obtained by inoculating conidia in YPD, pH 7.8 broth (1% (w/v) yeast extract, 2% (w/v) gelatin peptone, and 3% (w/v) dextrose), and incubating at 37 °C and 120 rpm for 4 days, as previously reported (Martínez-Álvarez et al., 2017). Infection of G. mellonella with E. coli was performed with strain DH5α (Thermo Fisher Scientific, Waltham, MA, US) grown in LB broth at 37 °C and 120 rpm. To prepare cell homogenates, yeast-like cells were suspended in 1 mL of 5 mM EDTA, 2 mM DTT, 25 mM Tris-HCl, pH 7.5, and lysis was performed with 425–600 µm size glass beads (Sigma-Aldrich, St. Louis, MO, USA), shaking 1 min in a vortex and incubating 1 min on ice, until complete 10 cycles. Then, cell homogenate was centrifuged at 9,485 × g for 10 min, and the supernatant was recovered and stored at −20 °C until use.

Recombinant expression of Galleria mellonella apoLp-III in Escherichia coli

Total RNA was isolated from G. mellonella larvae by TRIzol^® reagent protocol (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s specifications. The cDNA was synthesized and purified by adsorption chromatography following the methodology reported elsewhere (Trujillo-Esquivel et al., 2016). The open reading frame from apoLp-III was amplified by PCR using primer pairs 5′-GAATTCCAAGCGGGCATAGTGCG-3′ and 5′-AAGCTTCTGCTTGCTGGCGGC-3′ (underlined bases indicate EcoRI and HindIII sites, added to the forward and reverse primer sequences, respectively) (Niere et al., 1999). The 510-bp amplicon was cloned into pJET1.2/blunt (Thermo Fisher Scientific, Waltham, MA, USA), and then subcloned into the EcoRI and HindIII sites of pCold I (Takara Bio Inc, Kusatsu, Shiga, Japan), generating pCold-apoLpIII. For recombinant gene expression, the construction was used to transform E. coli BL21 (DE3) Star competent cells (Thermo Fisher Scientific, Waltham, MA, USA) and were grown in LB broth with 100 μg mL⁻¹ ampicillin (Sigma-Aldrich, St. Louis, MO, USA) for 20 h at 37 °C and orbital shaking (180 rpm). Then, 200 mL of fresh LB broth contained in a 2 L Erlenmeyer flask was inoculated with 1.0 mL of the overnight culture and incubated at 37 °C until reaching an OD_600nm = 0.6, then isopropyl-β-D-1-thiogalactopyranoside was added to reach a final concentration of 1.0 mM and further incubated for 20 h at 15 °C. Cells were harvested by centrifuging for 20 min at 1,500 × g and 4 °C and kept at −20 °C until use.

Recombinant protein purification

Induced bacterial cells were resuspended in 5 mL of buffer A (100 mM NaH₂PO₄, 10 mM Tris-HCl, 8 M urea, pH 8.0), and lysis was performed with ≤106 µm size glass beads (Sigma-Aldrich, Burlington, MA, USA), shaking 1 min in a vortex and incubating 1 min on ice, until complete five cycles. Then, the sample was subjected to three cycles of freezing at −70 °C, resting at 35 °C between cycles. The cell homogenate was centrifuged at 9,485 × g for 10 min, then, the supernatant was recovered and loaded onto the Poly-Prep column (Bio-Rad, Hercules, CA, USA) with a 10 mL capacity for sample reservoir, containing 2 mL of TALON Metal Affinity Resin (Jena Bioscience, Jena, Germany). After a 60 min interaction at room temperature, the column was washed with buffer A at pH 8.0, 7.5, and 7.0 (three volumes for each pH solution), and the proteins of interest were eluted with five volumes of buffer A at pH 6.5, five volumes of the same buffer at pH 5.4 and five volumes of buffer A at pH 4.5. For each wash and elution, 2 mL fractions were collected. For purification analysis, 40 µL aliquots of each fraction were subjected to 12% (w/v) polyacrylamide gel and electrophoretically separated under denaturing conditions at 100 V for 90 min. Then, gels were washed three times with deionized water and stained with 0.25 M KCl and 1 mM dithiothreitol to visualize the protein bands and to slice out those of interest, following the procedures described previously (García-Carnero et al., 2021; Hager & Burgess, 1980). The recombinant protein was recovered through passive diffusion from polyacrylamide slices by suspending the gel pieces in PBS and incubating at 4 °C overnight with gentle shaking (120 rpm) (García-Carnero et al., 2021). For purity analysis, protein aliquots were separated by SDS-PAGE in 12% (w/v) gels and stained with silver nitrate as reported (Chevallet, Luche & Rabilloud, 2006). Protein determination was performed with a DC Protein Assay kit (Bio-Rad, Hercules, CA, USA) according to the manufacturer’s instructions.

Generation of polyclonal antibodies against recombinant apoLp-III

A New Zealand rabbit was immunized intramuscularly with an antigenic preparation that contained 150 µg of recombinant apoLp-III purified and the complete Freund’s adjuvant (Thermo Fisher Scientific), following a standard protocol (Mora-Montes et al., 2008). After 2 weeks, one booster dose (150 µg of recombinant apoLp-III purified and the incomplete Freund’s adjuvant) was injected intramuscularly, and the procedure was repeated until three booster doses were completed. A total of 2 weeks after the last booster, the rabbit was bled, the serum was collected, and the globulin fraction was precipitated with 76% (w/v) ammonium sulfate. One Aliquot of 10 mL blood was extracted from the rabbit before the immunization protocol, and the serum was collected and kept at −20 °C until used. Indirect-ELISA was used to title the antibody production, using Clear Flat-Bottom Immuno Nonsterile 96-Well plates (Thermo Fisher Scientific, Waltham, MA, US) coated with 3.0 μg recombinant apoLp-III. Briefly, wells were coated with 3 μg of pure rapoLp-III protein and incubated for 3 h at room temperature. Subsequently, the plate was washed with PBS-0.05% (v/v) Tween 20 3 times (all washes were done in this way) and blocking with 5% (w/v) casein was performed overnight at 4 °C. After washing, different dilutions of anti-apoLp-III or preimmune serum were placed from 1:100 to 1:51,200. It was then incubated for 2 h at room temperature and washing was performed. Subsequently, goat anti-rabbit IgG-HRP (Catalog number A0545; Sigma-Aldrich, St. Louis, MO, USA) was placed at a dilution of 1:2,000, incubated for 2 h at room temperature, and washed. The development was carried out with 100 μL of the substrate 3,3′,5,5′ tetramethylbenzidine (GoldBio, Olivette, MO, USA) solution was added to each well (6 mg mL⁻¹ 3,3′,5,5′ tetramethylbenzidine and 2.6 μL of hydrogen peroxide in 12 mL of a 0.1 M citrate-phosphate solution) and left incubating at room temperature in the dark for 20 min. The reaction was stopped with 50 μL of 3 M H₂SO₄ per well, and the absorbance at 450/540 nm was measured.The title of the antibodies generated was 1:3,200.

Total protein extraction from G. mellonella moths

Moths were obtained from an in-house colony previously established and maintained on a diet based on corn bran and honey (Clavijo-Giraldo et al., 2016) G. mellonella moths were incubated at −20 °C for 20 min. Immediately after, one moth was submerged in 500 μL PBS, pH 7.4, and vigorously homogenized with a Potter. Samples were incubated on ice for 5 min, and an additional 500 µL of PBS was added and shaken in a vortex for 30 s. The samples were centrifuged for 10 min at 8,000 × g and the supernatants were recovered. This was further centrifuged for 10 min at 15,000 × g and the supernatants were recovered in new tubes. The protein aliquots were stored at −20 °C until used.

Immunodetection of native apoLp-III by western blotting

Aliquots containing 40 µg of recombinant apoLp-III or total protein from G. mellonella moths were separated by SDS-PAGE in 12% (w/v) polyacrylamide gels at 120 V for 90 min, and then electrotransferred onto a nitrocellulose membrane Protran Premium 0.45 µm NC (Sigma-Aldrich, St. Louis, MO, USA) using a Mini Trans-Blot Cell system (Bio-Rad, Hercules, CA, US), and transfer buffer (25 mM Tris, 192 mM glycine, and 20% (v/v) methanol). Membranes were blocked with 5% (w/v) casein in PBS-Tween 20 at 0.05% (v/v) overnight at 4 °C and then incubated for 2 h at room temperature with the anti-apoLp-III antibodies at a dilution of 1:3,200, washed with PBS-Tween 20 at 0.05% (v/v) incubated for 2 h with the goat anti-rabbit IgG-HRP antibody (Catalog number A0545; Sigma-Aldrich, St. Louis, MO, USA) at a 1:2,000 dilution, and the membrane was washed again with PBS-Tween 20 at 0.05% (v/v). Color development was carried out with 1 mg mL⁻¹ 3, 3′-diaminobenzidine (Sigma-Aldrich, St. Louis, MO, USA), and 1% (v/v) hydrogen peroxide (Sigma-Aldrich, St. Louis, MO, US). The membrane was dried and the image was captured with the ChemiDoc^™ MP (Bio-Rad, Hercules, CA, USA). As a control of the antibodies, an immunoblotting was performed with the pre-immune serum as the first antibody.

Infection assays in Galleria mellonella

Yeast-like cells were inoculated using cell suspensions adjusted at 2 × 10⁹ cells mL⁻¹ when it was used C. albicans WT, mutants and reintegrants strains; or 1 × 10⁷ cells mL⁻¹ when it was used S. schenckii, S. brasiliensis and S. globosa as reported (García-Carnero et al., 2020); while bacteria were inoculated from a suspension at 1 × 10¹¹ cell mL⁻¹ (Zdybicka-Barabas & Cytryńska, 2011). Groups containing 10 larvae were injected with 10 µL of the cell suspensions with a Hamilton syringe equipped with a 26-gauge needle in the last left pro-leg, previously sanitized with 70% (v/v) ethanol. As a control, one group was inoculated with PBS. Larvae were housed and incubated at 37 °C as reported (García-Carnero et al., 2020). After inoculation, larvae were incubated for 2, 8, 12, or 24 h, and the hemolymph was extracted (20 µL per larva), mixed in batch with 500 µL of cold PBS buffer with sodium citrate at 3.8% (w/v), giving a final dilution of the larva tissue of 1:3.5, and stored at −20 °C until used.

Quantification of Galleria mellonella apoLp-III by enzyme-linked immunosorbent assay

Clear Flat-Bottom Immuno Nonsterile 96-Well plates (Thermo Fisher Scientific,Waltham, MA, USA) were sensitized with 100 μL per well of serial dilution of the polyclonal antibodies anti-apoLp-III. The sera was diluted in 50 mM Tris-HCl, 150 mM NaCl, 1% (w/v) bovine serum albumin, pH 7.4. The plate was incubated for 2 h at 37 °C, then it was washed three times with 300 μL per well of PBS-Tween 20 at 0.05% (v/v). and then blocked with 300 μL of porcine skin gelatin in TBS at 1% (w/v) per well and incubated overnight at 37 °C. Then, plates were washed three times with PBS-Tween 20 at 0.05 % (v/v), 100 μL of the hemolymph was placed in each well and incubated at room temperature for 2 h and washed three times with PBS-Tween 20 at 0.05% (v/v). Next, 100 µL of anti-apoLp-III antibody diluted 1:400 were placed in each well and incubated for 2 h at room temperature, washed with PBS-Tween 20 at 0.05% (v/v), 50 μL of a dilution 1:2,000 of the goat anti-rabbit IgG-HRP (Catalog number A0545; Sigma-Aldrich, St. Louis, MO, USA) were placed and incubated at room temperature for 1 h. After washing with PBS-Tween 20 at 0.05% (v/v), 100 μL of the substrate 3,3′,5,5′ tetramethylbenzidine (GoldBio, Olivette, MO, USA) solution was added to each well (6 mg mL⁻¹ 3,3′,5,5′ tetramethylbenzidine and 2.6 μL of hydrogen peroxide in 12 mL of a 0.1 M citrate-phosphate solution) and left incubating at room temperature in the dark for 20 min. The reaction was stopped with 50 μL of 3 M H₂SO₄ per well, and the absorbance at 450/540 nm was measured. As a positive control, 100 μL of pure recombinant apoLp-III at 1.25 pg µL⁻¹ were included instead of the hemolymph. As an extra control, the same ELISA assay was realized, but with pre-immune serum dilutions and 100 μL of pure recombinant apoLp-III at 1.25 pg µL⁻¹ as antigen.

Statistical analysis

Infection assays in Galleria mellonella were performed in triplicate, as well as, the sandwich ELISA-like. The Dunnett’s one-way analysis of variance was used to compare the apoLp-III concentration of all larvae groups vs. the control group, or Tukey’s multiple comparison tests when comparing all pairs of groups. The latter was also used to compare ELISA readings using different antibody concentrations. All P values less than 0.05 are considered significant. The GraphPad Prism^® 6.0 program was used for the statistical analysis (GraphPad Software, La Jolla, CA, USA).

Heterologous expression and purification of apoLp-III in Escherichia coli

To generate an immunodetection system capable of detecting G. mellonella apoLp-III, we first generated a recombinant version of this protein in a bacterial system, as this has been previously performed and the recombinant product retained functional activity (Niere et al., 1999), suggesting that posttranslational modifications, if present in the functional polypeptide are dispensable for both lipid binding and opsonin activity (Niere et al., 1999). Since the native polypeptide contains a putative signal peptide at the N-terminus, we did not include the first 16 amino acids in the recombinant protein, and instead, the expression vector adds a TEE motif, a His tag, and a factor Xa site element. As a consequence, the recombinant version of apoLp-III is expected to be a 21 kDa polypeptide, instead of the 18-kDa native protein (Niere et al., 1999). When bacteria were transformed with pCold-apoLp-III and gene expression was induced with 1.0 mM isopropyl-β-D-1-thiogalactopyranoside for 20 h at 15 °C, a differential 21-kDa protein band was observed in cell homogenates of induced bacteria (Fig. 1A). Since this protein was absent in the cell homogenates of non-transformed cells or cells harboring the empty pCold-I vector (Fig. 1A), it was suggested that recombinant apoLp-III was generated. The recombinant protein was purified by metal affinity chromatography as detailed in ‘Material and Methods’, and the aliquots enriched with the recombinant protein were analyzed by denaturing SDS-PAGE in 12% (w/v) gels. Since both Coomassie blue and silver stainings showed only one protein band in the purified fractions (Figs. 1B and 1C), we considered the protein preparation at a purity level enough to use it for antibody generation. It is worth noting that since no 2D-electrophoretic analysis was performed, we cannot claim the protein was purified to homogeneity.

Production of polyclonal anti-apoLp-III antibodies

The purified protein was used as an antigen to generate rabbit anti-apoLp-III antibodies. To assess the ability of these antibodies to detect native apoLp-III in the G. mellonella hemolymph, a G. mellonella moths homogenate was prepared and used in a western blot analysis. Despite the hemolymph electrophoretic analysis showing several protein bands of different molecular weights, the antibodies only recognized one 18-kDa protein band, a molecular weight expected for native apoLp-III (Figs. 2A and 2B). As a positive control, we also detected the recombinant apoLp-III, identifying a robustly detected protein band (Fig. 2B). As negative controls, the western blot analyses were performed using the preimmune serum instead of the primary antibody, or the secondary antibody was omitted (Figs. 2C and 2D). The western blot with the preimmune serum showed a faint detection of two protein bands in the homogenate, with molecular weights of 50 and 31 kDa (Fig. 2C). A closer analysis of the results shown in Fig. 2B indicated that these are also barely detected in the homogenate by the anti-apoLp-III antibodies; however, this detection could be associated with antibodies already circulating in the rabbit sera. Nevertheless, densitometric analysis of the bands shown in Fig. 2B indicated that more than 98% of the protein detected in the hemolymph was the native apoLp-III. The western blot assay where no secondary antibody was included showed no protein band detected (Fig. 2D). Next, using ELISA as described in the Material and Methods section, the antibody sensitivity was analyzed, using different antibody dilutions and recombinant apoLp-III concentrations. When 1.25 or 0.5 µg recombinant protein was used as antigen and different dilutions of the antibodies, concentration-dependent readings were observed (Fig. 3). Except for the readings obtained with the dilution 1:100, the readings with 0.1 µg of recombinant apoLp-III were similar to those of the system with no recombinant protein, indicating our system cannot detect this amount of recombinant apoLp-III (Fig. 3). Since the reading with a dilution of 1:1,600 and 0.5 µg recombinant protein was the last that gave a statistically different absorbance this was considered the detection limit (Fig. 3).

Generation of a system to immunodetect Galleria mellonella apoLp-III

The polyclonal anti-apoLp-III antibody was used to generate a Sandwich-like ELISA system to detect native apoLp-III in the G. mellonella hemolymph, as described in the Materials and Methods section. The specificity of this system was analyzed by assaying different cell homogenates of microorganisms, and we included E. coli, C. albicans, S. schenckii, S. brasiliensis, and S. globosa. In addition, we also analyzed the detection of an irrelevant recombinant protein from S. schenckii, the peptide SPSK_06559 (Giosa et al., 2020). In all cases, readings were similar to the negative control where only PBS was included, with average readings at 450 nm of 0.154 ± 0.036, which contrast with the average reading of the positive control wells containing 1.25 µg rapoLp-III (OD450 nm 1.1635 ± 0.029).

The apoLp-III response when G. mellonella larvae were infected with E. coli or C. albicans was previously analyzed by western blotting using antibodies against G. mellonella apoLp-III and a heterologous antibody against apolipophorin I and apolipophorin II from Bombyx mori (Stączek et al., 2018). They did not recognize apolipophorin III. In addition, antibodies against apoLp-III used in this study were obtained against G. mellonella apoLp-III. This analysis indicated that apoLp-III levels tended to increase in the first hour after injection of the pathogen and then decrease after 24 h of inoculation (Stączek et al., 2018). To validate our system, we used both pathogens to challenge G. mellonella larvae and then quantified the apoLp-III levels using our ELISA system. Our results indicated that the control group of larvae injected only with PBS showed a significant increment in apoLp-III levels when compared to untouched larvae (P = 0.009), referred to here as the naïve group (Fig. 4). At short post-inoculation times, such as 8 and 12 h apoLp-III levels significantly increased in the hemolymph of infected insects, an observation that applies to both groups, larvae challenged with E. coli or C. albicans (Fig. 4). As reported, at 24 h post-inoculation of either bacteria or yeast cells, apoLp-III levels reduced and were like those observed in larvae challenged only with PBS (P = 0.1076). A similar situation was observed at 48 h post-infection (P = 0.2074), suggesting that a second wave of apoLp-III increment was unlikely (Fig. 4). One possible explanation for this observations is the humoral response by this opsonin is no longer relevant at this time points because the shift to a cellular response mediated by hemocytes.

To further validate our system, we analyzed the apoLp-III levels stimulated in larva infected with C. albicans mutant strains with known virulence defects in both murine and G. mellonella models. The strains included in this validation were mutants lacking one or the two KEX2 alleles, which code for a Golgi-resident serine protease involved in the classic protein secretory pathway, and both mutant strains showed an avirulent phenotype when injected in G. mellonella larvae (Gómez-Gaviria et al., 2020). We also included the C.albicans och1∆ null mutant strain, which lacks a Golgi-resident α-1,6-mannosyltransferase involved in the synthesis of the N-linked glycan outer chain, and whose virulence is attenuated in this insect model (Bates et al., 2006; Pérez-García et al., 2016). Interestingly, both KEX2 mutants failed in inducing apoLp-III at 8 and 12 h post-infection, times associated with high lipoprotein levels in the hemolymph of larvae infected with the wild-type control strain (Fig. 5). Similar results were observed with the och1∆ null mutant strain, but the complementing mutant strain, where one copy of the disrupted gene was placed back into the null mutant background, restored the ability to induce apoLp-III at the same level as the wild-type control strain (Fig. 6). Therefore, our system to detect and quantify apoLp-III is at least useful to analyze the stimulation of this peptide by C. albicans.

We also explored the usefulness of the apoLp-III detection system when G. mellonella larvae were inoculated with Sporothrix yeast-like cells, as these larvae have been validated to study the virulence of species of the Sporothrix pathogenic clade (Clavijo-Giraldo et al., 2016; Lozoya-Pérez et al., 2020). When S. brasiliensis, S. schenckii, or S. globosa yeast-like cells were inoculated in larvae, the three fungal species stimulated the production of apoLp-III (Fig. 7). Alternatively, the increment of apoLp-III could be due to releasing of this protein stored in fat body cells. In the groups infected with S. brasiliensis and S. schenckii the apoLp-III levels incremented after 2 h of inoculation and this was further induced at 4 h post-inoculation (Fig. 7). However, when compared to the apoLp-III levels at 4 and 8 h post-inoculation this increment was not significant (P = 0.181 and 0.417 for S. brasiliensis and S. schenckii, respectively). For the case of S. globosa, the apoLp-III quantification at 2h post-inoculation was significantly higher when compared to those levels stimulated with PBS (P = 0.0457) but this did not further increment at 4 or 8 h post-inoculation (Fig. 7). When the apoLp-III stimulation by the three fungal species was compared, the data indicated that S. brasiliensis stimulated the highest levels of this protein at 2 and 4 h post-inoculation, while S. globosa was the lowest (Fig. 7). At 8 h post-inoculation, the only statistically significant difference was observed between larvae stimulated with S. brasiliensis and S. globosa (P = 0.0143). Collectively, these data indicate that our apoLp-III detection system allowed us to generate species-specific stimulation patterns for the Sporothrix species under analysis.