New strategy to improve quality control of Montenegro skin test at the production level

Introduction: The production of the Montenegro antigen for skin test poses difficulties regarding quality control. Here, we propose that certain animal models reproducing a similar immune response to humans may be used in the quality control of Montenegro antigen production. Methods: Fifteen Cavia porcellus (guinea pigs) were immunized with Leishmania amazonensis or Leishmania braziliensis, and, after 30 days, they were skin tested with standard Montenegro antigen. To validate C. porcellus as an animal model for skin tests, eighteen Mesocricetus auratus (hamsters) were infected with L. amazonensis or L. braziliensis, and, after 45 days, they were skin tested with standard Montenegro antigen. Results: Cavia porcellus immunized with L. amazonensis or L. braziliensis, and hamsters infected with the same species presented induration reactions when skin tested with standard Montenegro antigen 48-72h after the test. Conclusions: The comparison between immunization methods and immune response from the two animal species validated C. porcellus as a good model for Montenegro skin test, and the model showed strong potential as an in vivo model in the quality control of the production of Montenegro antigen.


INTRODUCTION
Cutaneous leishmaniasis (CL) is a serious social and public health problem, because it can result in sequels 1 .Early diagnosis and a suitable treatment are the best tools to control the disease 1 .Immunological methods have been largely used as a screening tool for diagnosis 1,2 .One method widely used in Brazil is the Montenegro skin test (MST) based on the delayed-type hypersensitivity reaction (DTH) 2 .The in vivo manifestation of cellular immune response is an induration that can be measured and semi-quantified by skin tests 3 .The method has been widely used as a complementary leishmaniasis diagnosis because of its high sensitivity and specificity.Furthermore, MST is an easy method to perform, low cost, does not require sophisticated equipment, and can be performed in loco 4 .
The Montenegro antigen available in Brazil is provided by the Centro de Produção e Pesquisa de Imunobiológicos (CPPI) in Paraná State, Southern Brazil.The production is authorized and inspected by the Agência Nacional de Vigilância Sanitária (ANVISA) that establishes the standard evaluation of internal testing for antigen production (RDC 59/2000; RDC 167/2004) [5][6][7] .During the production process, methods for the qualitative and quantitative control of the produced antigens are necessary to evaluate the antigen efficiency and to validate the lots of antigen [5][6][7] .Currently, this analysis is performed in in vivo systems that are exposed to the antigen in order to evaluate the biological response to exposure [5][6][7] .This control is performed in CL human patients 5 .Such approach demands clinical cases of CL and poses ethical questions, making the quality control of antigen a complicated process.This study aimed to address these issues in order to establish an experimental model capable to replace the current in vivo model in the quality control process of Montenegro antigen production.

METHODS
This study was divided in two stages.To evaluate the immune response of Cavia porcellus to standard Montenegro antigen in order to establish this species as an experimental model capable of replacing the current in vivo process used in the quality control of Montenegro antigen production.Secondly, to validate C. porcellus as a suitable animal model, Mesocricetus auratus, which is considered a susceptible bio model for infection with Leishmania sp. 8,9, was chosen.

Evaluation of Cavia porcellus immune response to skin test
Fifteen adult male guinea pigs, albino strain, weighing 250-350g, were maintained in the animal facility, housed in polypropylene cages, and fed a balanced diet with water and food ad libitum.The animals were maintained in groups of five animals.The animal number for this study was calculated as follow: N = (Z α/2 ?(δ/E)) 2 where Z α/2 : 1.96; E: standard error (5%); δ: standard deviation = amplitude/4.The standard deviation was 9.5.Our aim was to minimize the number of animal used in the experiment while maintaining statistically relevant data (according to the 3 Rs concept) 10 .
The experiments were conducted in four steps (Figure 1).The first step consisted in the preparation of Leishmania antigen to immunize C. porcellus.Promastigotes forms of L. amazonensis (MHOM/BR/73/M2269) and L. braziliensis (MHOM/BR/94/ M2903) from the Centre de Ressources Biologiques des Leishmania, Montpellier, France, were cultivated until a concentration of 10 7 parasites/mL was reached.After washing three times with saline solution at 0.9, 0.3, and 0.9%, the pellet was stored at -20°C until further use.The soluble Leishmania antigens were prepared as described by Szargiki et al. 11 .Briefly, the promastigotes were defrosted and centrifuged at 4°C for 15 min at 800 x g.The pellet was diluted in sterile distilled water in a volume equal to half of the pellet volume.The cell suspension was disrupted by heat shock, alternating between -80°C and 50°C for 5 min, and further sonicated six times at 30% of potency at intervals of 1 min in an ice bath.The extracted solution was centrifuged at 14,000 x g at 4°C for 60 min and lyophilized.The antigen emulsions (L.amazonensis and L. braziliensis) were prepared by diluting 20.5mg of each lyophilized antigen in 5.1mL of sterile water, and 3.3mL of this solution was added to 3.3mL of incomplete Freund's adjuvant (IFA).The final concentration of the antigen emulsion was 4mg/ mL.The guinea pigs were divided into three groups with five animals in each.The groups were as follows: G1: immunized with L. amazonensis antigen emulsion; G2: immunized with L. braziliensis antigen emulsion; and G3: inoculated with IFA emulsion and saline solution (negative control group).The guinea pigs were intramuscularly inoculated with 1mL (4mg/ mL) of antigen emulsion or IFA.The inoculum (0.5mL) was intramuscularly injected in two sites of guinea pig chest.After 30 days, C. porcellus was skin tested with standard Montenegro antigen (CPPI -lot 01/12-registration number: 80151040004).For the skin test process, the guinea pigs were depilated, and four inoculation sites were delimited.The inoculations with disposable syringes for each site were as follows: 0.1 or 0.2mL of standard Montenegro antigen, or 0.1 or 0.2mL of intradermally inoculated vehicle.The vehicle was represented by phenol saline solution in which the Montenegro antigen was diluted.The guinea pigs were skin tested with this solution to discriminate unspecific reaction or cross reactivity to it.The Montenegro antigen was inoculated at a concentration of 4µg/0.1mL.The readouts of the reaction were performed at 24, 48, and 72h after the intradermal inoculation by measuring the diameter of the indurations using a caliper.

Validation of Cavia porcellus as an animal model for skin test
To validate the proposed animal model, the same experimental design was applied to gold hamster (M.auratus), capable of developing the disease through infection with L. amazonensis or L. braziliensis.Eighteen male, weighing 140-160 g and 6-8 weeks old, were used.The animals were housed in polypropylene cages and fed with water and food ad libitum.The animals were maintained in groups of six.The animal number for this study was calculated as previously mentioned.
For experimental infection, promastigotes forms of L. amazonensis and L. braziliensis were separately cultivated in biphasic brain heart infusion culture medium and incubated at 24ºC.Promastigotes cultured until stationary phase (5 days) were used to infect the hamster groups.Briefly, the parasites were harvested, diluted in ultrapure water (5x), counted, and adjusted to a concentration of 1 x 10 7 parasites in 0.1mL of saline solution for inoculation.Three groups with six animals each were formed as follows: GH1: infected with 1 x 10 7 L. amazonensis; GH2: infected with 1 x 10 7 L. braziliensis; and GH3: saline solution (negative control group).The hamsters were intradermally inoculated in the right hind paw with 0.1mL of saline solution containing 1 x 10 7 parasites.After 45 days, when initial lesions were observed, hamsters were skin tested with standard Montenegro antigen.First, they were depilated, and the inoculation was performed as described above.The skin test solutions were intradermally inoculated using an antigen concentration of 4µg/0.1mL.The readouts of the reaction were performed at 24, 48, and 72h after measuring the diameter of the indurations using a caliper.

Statistical analysis
The homogeneity of variance analysis and the Kolmogorov-Smirnov test for the normal condition of variable evaluation were performed using the Statistica 7 software.A Kruskal-Wallis one-way analysis with a Dunn post-test was performed to analyze the significant difference among groups using GraphPad Prism 6 software, assuming a significance level of 5%.

Ethical considerations
This study was approved by the Research Ethics Committee of the Federal University of Paraná (Process n. 101328/2015-69).
In C. porcellus immunized with IFA and saline solution (negative control group), induration reactions were not observed.Statistical analysis showed that there was no significant difference between the induration reactions with inoculation of 0.1 and 0.2mL of Montenegro antigen in the guinea pigs immunized with L. amazonensis or L. braziliensis (p-value < 0.01) after 24, 48, and 72h.
In hamsters infected with saline solution (negative control group), induration reactions were not observed.Statistical analysis showed that, in the group infected with L. amazonensis, there was a considerable difference between induration reactions resulting from inoculation with 0.1 and 0.2mL of Montenegro antigen at 24 and 48h.After 72h, there was no significant difference between induration reactions with 0.1 and 0.2mL of Montenegro antigen (p-value = 0.01).For the group infected with L. braziliensis, there was no significant difference between induration reactions with 0.1 and 0.2mL of Montenegro antigen after 24, 48, and 72h (p-value = 0.04).

DISCUSSION
A volume of 0.1mL of Montenegro antigen is the standard amount used to inject patients when performing the skin test for cutaneous leishmaniasis 12 .In this study, the higher level of 0.2mL was tested to evaluate if such increase in volume would show a better induration reaction.The results showed that both injected volumes (0.1 and 0.2mL) promoted induration reactions, and the difference between these reactions was not considerable.However, the skin test reaction to the 0.2-mL Montenegro antigen inoculation was different in some animals

A B
compared to the group immunized with L. braziliensis.These results showed that 0.1-mL Montenegro antigen inoculation is the best volume of antigen for skin test in the C. porcellus model.The positivity criteria established for skin test in guinea pigs in the present study were induration reactions ≥ 0.5cm and readouts at 48-72h after the intradermal test.Some studies are in accordance with the positivity criteria established in the present study [13][14][15] .The readouts after 24 h were not considered in the test evaluation, because 24h characterized the initial stage of the immunology response in DTH reactions, representing the cell migration to the injection site 3 .The influx of T cells reaches the maximum levels at 48-72h, indicating the time to determine the result of skin test 3 .
The animal model chosen for this study was C. porcellus, because it is a well-established animal model to study DTH in different types of diseases and is also used in the investigation of several infectious diseases [16][17][18][19][20] .Furthermore, some studies on the genetics of guinea pig have shown immunological analogies between these species 21 .For instance, guinea pig genes that encode for major histocompatibility complex (MHC) proteins are homologous to human proteins, and genetic expression pattern of interferon-γ (IFN-γ) and interleukin 12 (IL-12) are similar in the two species 21 .
The skin test in hamsters showed that they developed induration reaction for Montenegro antigen in both infected groups (L.amazonensis and L. braziliensis), irrespective of the inoculum volume.The skin test volume of 0.2mL promoted higher induration reaction in hamsters than volume 0.1mL in both the infected groups, but induration responses were not statistically different.The induration response difference was considerable just in hamsters in the L. amazionensis group after 48h.However, this result was in accordance with induration reaction developed in guinea pigs injected with the same Montenegro antigen volumes.As in guinea pigs, there was no considerable difference between induration reactions promoted by 0.1 or 0.2mL of Montenegro antigen.These results in hamsters validated the experimental model proposed in this work.
The guinea pigs proposed as a model for quality control have many advantages compared to hamsters.Firstly, in all infected groups (L.amazonensis, L. braziliensis, and negative control saline), hamsters presented lesions resulting from the phenol in the vehicle solution, which were not observed in guinea pigs.This could be a result of the more sensitive skin of the hamsters.Furthermore, a sensitive skin makes the intradermal injection difficult, because the needle can easily punch the skin or produce lesions, leading to difficult readouts.Secondly, a volume of 0.2mL of Montenegro antigen was required for the skin test in hamsters in order to observe induration reactions.Thirdly, hamster fur grows faster than that of guinea pigs, causing difficulties in performing the readouts after 48 and 72h.Fourthly, induration reaction response in hamsters was heterogeneous between the animals, while in guinea pigs this response was homogeneous, reducing the standard deviation and making the test more reliable.Fifthly, hamsters have strait dorsal compared to guinea pigs, making it difficult to organize the skin test inoculation sites.Furthermore, guinea pigs need be immunized with dead parasites (parasite proteic extract) and do not develop the disease, while hamsters need to be infected (live parasites) to respond to skin test 8,9,14 .Once immunized, the risk of contamination is reduced, and the time needed to develop an immune response is faster than that of hamsters (guinea pigs = 30 days; hamsters = minimum of 45 days).Thus, guinea pigs are better animal models to be used in an industrial qualitative control process.Moreover, they are already used as bio models for skin test antigen evaluation in tuberculosis 19 , leprosy 22 , and immunodeficiency virus type 1 infection 20 .Additionally, the process of experimental infection in hamsters is time consuming and requires well-trained professionals 8,9 .Observations made during experimental tests with hamsters also confirm that guinea pigs are a suitable animal model for Montenegro antigen quality control process.
We conclude that the experimental tests performed confirm that this experimental model may be used to replace the current human in vivo model in quality control process of Montenegro antigen production.The induration reaction was observed in all animals in both immunized groups.The establishment of this experimental model for skin tests represents a good alternative for qualitative and quantitative control process of antigen production.

FIGURE 2 -
FIGURE 2 -A.Montenegro skin test in Cavia porcellus previously immunized with Leishmania amazonensis.a: Induration reactions against standard Montenegro antigen with an inoculation of 0.1mL.b: Induration reactions against Montenegro antigen with an inoculation of 0.2mL.B. Reaction to Montenegro antigen in Cavia porcellus previously immunized with Leishmania braziliensis.c: Induration reactions against standard Montenegro antigen with an inoculation of 0.1mL.d: Induration reactions against standard Montenegro antigen with an inoculation of 0.2mL.The diameter of indurations was measured 24, 48, and 72h after the antigen inoculation.Each value corresponds to individual induration reaction of the five animals in each group.MST: Montenegro skin test.

FIGURE 3 -
FIGURE 3 -A.Montenegro skin test in Mesocricetus auratus infected with Leishmania amazonensis.a: Induration reactions against standard Montenegro antigen with an inoculation of 0.1mL.b: Induration reactions against standard Montenegro antigen with an inoculation of 0.2mL.B. MST in Mesocricetus auratus infected with Leishmania braziliensis.c: Induration reactions against standard Montenegro antigen with an inoculation of 0.1mL.d: Induration reactions against standard Montenegro antigen with an inoculation of 0.2mL.The diameter of indurations was measured 24, 48, and 72h after skin test.Each value corresponds to individual induration reaction of the six animals in the group.MST: Montenegro skin test.

TABLE 1 :
Diameter in mm of induration after Montenegro skin in Cavia porcellus immunized with Leishmania amazonensis or Leishmania braziliensis.

TABLE 2 :
Diameter of induration after reaction against Montenegro antigen in Mesocricetus auratus infected with Leishmania amazonensis or Leishmania braziliensis.