Origin and characterization of the B. mallei strains
The five strains of B. mallei used to produce mallein in this work were isolated from clinical cases of glanders in horses and mules in the Brazilian states of Alagoas and Pernambuco. They were obtained from the Laboratório de Bacteriologia do Setor de Veterinária da Universidade Federal Rural de Pernambuco, Brasil and were designated with the numbers 1 to 5. The origin and characterization of the B mallei strains are detailed in Table 1.
Prior to the biochemical and molecular characterization (PCR and Real-Time PCR), the B. mallei strains were passaged through guinea pigs in order to select more virulent clones of better maleogenic potential. The guinea pig passage was done by the inoculation of five animals, one for each strain, by the intraperitoneal route with 1.0 ml of a B. mallei suspension in PBS containing 108 UFC .
The biochemical characterization was performed using the API 20E® kit plus the tests for oxidase, catalase, motility and triple sugar iron (TSI) . The PCR and Real-Time PCR assays were executed according to the OIE Manual of Diagnostic Tests and Vaccines for Terrestrial Animals  using oligonucleotides designed on the basis of the differences in the sequences of the fliP gene of B. mallei and Burkholderia pseudomallei. For the Real-Time PCR, we used the RealQ PCR Master Mix® kit from Ampliqon, Odense, Denmark, and primers and probes from IDT DNA® in an Applied Biosystems ABI 7500® thermocycler. The annealing temperature was changed from 63°C to 62°C to adapt the protocol to our Real-Time PCR kit. The new temperature offered the best compromise between specificity and DNA polymerase activity. As a negative control, we used a B. pseudomallei strain provided by the Laboratório Central de Saúde Pública do Ceará – LACEN-Ce, and the positive control was a B. mallei strain from ATCC (15310).
Production of the mallein pre-concentrate
The five mallein pre-concentrates were produced according to the method of Huitema . This procedure is very similar to that used to produce PPD tuberculin, where the protein purification is achieved by precipitation with trichloacetic acid (TCA). In brief, cultures of B. mallei were grown in Dorset-Henley broth for eight weeks at 37°C and were then steam-inactivated in an autoclave for one hour, tested to confirm the inactivation, centrifuged at 6000 x g for one hour, and the supernatant was filtered through a 0.22 μm PVDF membrane to remove the residual cells. This product, called crude mallein, was further processed by the addition of TCA to concentrate the maleo-proteins by precipitation. The precipitate was washed with 5% saline until the pH of the supernatant rose to 2.4, then with acidified saline (pH 3.0) until the pH of the supernatant rose further to 2.7-2.8. It was then dissolved in a minimum quantity of alkaline solution to produce the mallein pre-concentrate, the pH of which must be approximately 6.7. The protein concentration of this mallein pre-concentrate was estimated spectrophotometrically by the Modified Biuret method using a standard curve of albumin with reading at 544 nm .
Purification of the mallein pre-concentrate
The objective of the purification process was to select the proteins with molecular weights above 350 kDa, which are considered to be more immuno-reactive and specific than those of lower molecular weight . The purification of the mallein pre-concentrate was achieved by ultrafiltration in a Millipore Labscale® TFF system fitted with a Pellicon® XL Ultracel 300 device encasing a regenerated cellulose membrane with a molecular weight limit (MWL) of 300 kDa. The mallein pre-concentrate was diluted in 0.01 Mol/L PBS, pH 7.2-7.4, to a total volume of 500 ml, which is the maximum capacity of the TFF equipment reservoir. The diluted pre-concentrate was ultrafiltered until its volume was reduced to approximately 50 ml (one tenth of the initial volume), and the steps of dilution and ultrafiltration were repeated until we achieved the desired purification. The process was monitored by evaluation of the protein content and the chromatographic profile of the product. The protein content was estimated by the modified Biuret method described above, and the chromatographic profile was established by a High Performance Liquid Chromatography (HPLC) system using an Agilent Zorbax® Bio Series GF-250 column with an internal diameter of 4.6 mm. This methodology employs as mobile phase the same PBS solution used to wash the mallein pre-concentrate in the purification process and uses a flow rate of 0.5 ml/min., an injection volume of 50 μL and a UV detector set to 210 nm. The HPLC equipment was a Shimadzu, model LC-20AT. Prior to injection, the sample was diluted with the mobile phase to yield a protein concentration of 1.0 mg/ml.
To protect the proteins during the freeze-drying process, the purified malleins, retentate of the TFF process, were mixed with equal volumes of a buffered 14% glucose solution . This product was filtered through a 0.22 μm PVDF membrane, freeze-dried in 1.0 ml volumes and identified with the same number (1 to 5) as the B. mallei strain from which they originated. Prior to use, the purified malleins were rehydrated and diluted in a sterile 0.5% phenol solution to a final protein concentration of 1.0 mg/ml.
The permeate of the TFF process (proteins with molecular weights below 350 kDa) was concentrated using the same TFF equipment fitted with a Pellicon® XL Biomax 5 device encasing a polyethersulfone membrane with MWL of 5 kDa. This concentrate was mixed with equal volumes of the same buffered glucose solution used to dilute the retentate, filtered and freeze-dried.
Characterization of the purified malleins
The purified malleins were characterized by HPLC using the methodology described above. As a molecular weight reference for the purified fraction (proteins with molecular weights above 350 kDa) we used bovine thyroglobulin (SIGMA), which has a molecular weight between 660 and 690 kDa. As a reference for the removed fraction (proteins with molecular weights below 350 kDa), we used bovine albumin (SIGMA – Fraction V), which has a molecular weight of 67 kDa.
Potency and specificity tests of the purified malleins
To test the potency of the purified malleins, we used four mixed-breed, non-pregnant mares between 4 and 10 years of age. The animals were sensitized with a 25% packed volume inactivated suspension of B. mallei prepared with 3 of the 5 strains used to produce the malleins, according to the method described by Verma et al. . The three strains used to produce the suspension were selected on the basis of their virulence as evaluated by inoculation in guinea pigs (Table 1). The cell suspension was diluted in PBS and Freund’s incomplete adjuvant at a ratio of 1:1.5:2.5, respectively, and emulsified. The animals were inoculated with 1.0 ml of the emulsion, with 0.5 ml injected intramuscularly and 0.5 ml injected subcutaneously. A second dose was administered 30 days after the first, and the first potency test was performed 44 days after the second dose. Additional reinforcement doses were administered 30 days after each potency test and the animals were reutilized 30 days after each reinforcement dose to test the next purified mallein. On the whole, the four sensitized animals were used five times, to test the five purified malleins. To evaluate the efficiency of the sensitization, serum samples were collected immediately before the first dose of sensitizing emulsion, 30 days after the first and second doses, and immediately before each test. The sera were tested for the presence of anti-B.mallei antibodies by the CFT.
The purified malleins were tested in the same order as their number identification, from 1 to 5. The potency of the purified malleins was evaluated by comparison with a standard mallein (Pasteur Institute, Bucharest, Romania). For each horse out of the four used, four 6 cm2 sites were prepared by shaving the skin at each side of the neck, with two sites on an upper line and two on a lower line, vertically and laterally separated from each other by 6 cm. The thickness of the skin fold at each site was measured with a caliper and recorded, and 0.1 ml of mallein, at a concentration of 1.0 mg/ml, was inoculated intradermally at a point in the center of each site in a systematic way such that on the right side of the neck, the purified mallein was inoculated at points 1 and 4, and the standard was inoculated at points 2 and 3, and the order was inverted on the left side of the neck. Forty-eight hours after the inoculation of the malleins, the thickness of the skin at each site was measured again, and the intensity of the reaction was registered as the difference, in millimeters, between the two readings. At each individual test, the two malleins were inoculated in the four sensitized animals and the mean of the reactions (16 individual readings) produced by the purified mallein was compared with the mean of the reactions (16 individual readings) produced by the standard mallein.
To evaluate the specificity of the experimental malleins and the standard, the same protocol was applied at each test to four non-sensitized horses (negative controls), and the means of the reactions produced by each reagent in this group were compared to the means of the reactions produced by the same reagent in the sensitized group.
The data were analyzed statistically by a split plot ANOVA, with the reagents (purified and standard malleins) in the plot and the points of inoculation (1 to 4) in the subplot. For the potency test, we used a randomized blocks experimental design, and for the specificity test, we used a fully randomized experimental design.
The use of animals in this experiment had the approval of the Ethics Committee in Animal Experimentation of the Universidade Federal de Minas Gerais, under the reference number of 167/2011.
In an attempt to better understand the activity of the purified antigens in a real scenario with naturally infected animals, purified mallein number 2 was submitted to a field test with the standard mallein involving 15 suspected glanderous horses from an endemic region. The comparison was performed by the simultaneous inoculation of 0.1 mL of the purified mallein and the standard at a protein concentration of 1.0 mg/mL, intradermally, in the right and left lower eye-lid, respectively, of each animal. Reading of the results was done 48 hours after the inoculation and it was considered positive a marked lower eye-lid swelling accompanied or not by purulent secretion and conjunctivitis.