IMMUNITY TO RHODOCOCUS EQUI INFECTION IN HORSES

Rhodococcus equi is an opportunistic bacterium that commonly infects foals and immunocompromised patients. Due to the large economic losses that it can cause in the fi eld of horse breeding, the microorganism has been studied in details, including its immunological aspect. Within the humoral immunity, the most important immunoglobulins are those of the class G (IgG), produced as a response to the surface antigen associated with virulence (virulence associated protein A, VapA). IgG antibodies provide resistance to pneumonia in foals and have a dose dependent protective eff ect. In addition to them, the protective role of plasma is achieved through various cytokines. Cellular mechanisms are important for killing bacteria within the macrophage. Virulent strains which carry a plasmid with the gene for VapA stimulate the production of interferon gamma (IFN-γ), a key cytokine to kill these bacteria. Th e presence of IFN-γ is crucial for the removal of microorganisms from the lungs and prevention of formation of pulmonary granulomas. For the complete removal of bacteria cooperation of the humoral and cellular immunity is necessary. Particularly signifi cant


Abstract
Rhodococcus equi is an opportunistic bacterium that commonly infects foals and immunocompromised patients.Due to the large economic losses that it can cause in the fi eld of horse breeding, the microorganism has been studied in details, including its immunological aspect.Within the humoral immunity, the most important immunoglobulins are those of the class G (IgG), produced as a response to the surface antigen associated with virulence (virulence associated protein A, VapA).IgG antibodies provide resistance to pneumonia in foals and have a dose dependent protective eff ect.In addition to them, the protective role of plasma is achieved through various cytokines.Cellular mechanisms are important for killing bacteria within the macrophage.Virulent strains which carry a plasmid with the gene for VapA stimulate the production of interferon gamma (IFN-γ), a key cytokine to kill these bacteria.Th e presence of IFN-γ is crucial for the removal of microorganisms from the lungs and prevention of formation of pulmonary granulomas.For the complete removal of bacteria cooperation of the humoral and cellular immunity is necessary.Particularly signifi cant is opsonization, which increases phagolysosomal fusion.Vaccination and

INTRODUCTION
Rhodococcus equi (R. equi) is an intracellular bacterium that survives mechanisms of phagolysosomal fusion, which compromises the eff ectiveness of antibiotic therapy (von Bargen and Haas, 2009).Th e microorganism is present in the soil, on all continents except Antarctica (Prescott, 1991).
It can be found in the soil of 50-95% households which breed horses, as R. equi is present in horses' feces in very high concentrations.In the feces of mares, R. equi is present in concentration from 10 2 to 10 3 colony forming units (CFU) per gram.It is isolated from the feces of foals, starting from the fi rst week of life.In the feces of four weeks old foals it is present in the amount from 10 4 to 10 5 CFU.Such high concentration stays up to 8 and 10 weeks of age, when it starts to decrease and, as foals are maturing, settles to the concentration characteristic for mares (Takai, 1997).Th is is in accordance with the research of optimal doses for the foal experimental challenge.Th e dose of 10 2 CFU gave mild clinical presentation, mimicking natural infection (Sanz et al., 2013).Th e highest excretion by feces, 10 6 to10 8 CFU per gram, appears in the period until 8 weeks of life, precisely at the time when the foals are most receptive to infections (Takai, 1997).
In human population, infections by R. equi oft en aff ect immunocompromised patients (Weinstock and Brown, 2002;Suvajdžić and Považan, 2006;Tuon et al., 2007).Ten percent of infected patients were already being treated with immunosuppressive drugs, which are an integral part of the therapy during organ transplantation and in autoimmune diseases (La Rocca et al., 1998).It is believed that about two-thirds of patients infected with R. equi suff er from HIV (Harvey and Sunstrum, 1991).However, it can also cause diseases in immunocompetent persons (Kedlaya et al., 2001;Suvajdžić, 2004;Suvajdžić et al., 2015).
Managing control and prevention of this disease in veterinary medicine represents an economic category for the single reason that losses which may occur in the case of an epidemic in stables, especially of thoroughbred horses, would be enormous.In addition to general hygiene and sanitation, specifi c active and passive prophylaxis should occupy the place they deserve in the system of overseeing animal health (Muscatello, 2012a).

IMMUNITY IN RHODOCOCCUS EQUI INFECTIONS
Rhodococcus equi usually infects foals and very rarely adult horses (Vázquez-Boland et al., 2013).Infections occur sporadically, except in foals, where they can have enzootic character.Th e immune system has a key role in development of infection.In foals, the infection is associated with the period of vanishing of maternal antibodies present in colostrum.In adult animals, accommodation, food and environmental conditions have a signifi cant role in the course of infection, mainly due to their eff ect on the general immunological conditions (Muscatello, 2012b).
Infection of foals occurs within the fi rst week of life.Th e onset of clinical signs occurs approximately in 50 days.Th is also indicates that the incubation period is longer in natural infections, when compared to extremely high inocula sizes used in experiments.(Horowitz et al., 2001).
Th ere are implications that some mares give birth to foals that are particularly sensitive to R. equi infections.Such sensitivity can be explained by low levels of colostrum antibodies, functional immaturity of neutrophils in some foals, or genetic predisposition to infections (Wilks et al., 1982).
Due to the big impact on the health of horses, studies of immune response to infection with R. equi were mostly conducted on horses, in addition to murine models.Many aspects of immunological reactivity of horses were investigated, including both cellular and humoral mechanisms.

HUMORAL IMMUNITY IN HORSES
Th e digestive tract is probably the main source of antigenic stimulation of R. equi in horses.In foals, the level of maternal antibodies decline to the lowest level in the period of eight weeks of age, aft er which foals are actively producing antibodies.Foals with low levels of colostral antibodies, detected by an ELISA test, are particularly vulnerable to rhodococcal pneumonia (Hietala and Ardans, 1987).Th ose with higher levels of antibodies show a milder form of the disease (Martens et al., 1989).Accordingly, the maximum sensitivity of foals occurs in the period between 2 and 6 months of age (Hines et al., 1997).
Th e protective role of antibodies that are produced as the response to the surface antigen virulence-associated protein A (VapA) was shown.Th e G class of immunoglobulins (IgG), created against this antigen was present in bigger concentration when compared to the concentration obtained by immunization with intact but killed R. equi.Also, the increase in the mare serum opsonizing activity is more pronounced in the group that received the VapA antigen compared to the group that received the whole bacteria (Cauchard et al., 2004).
Immunoglobulins created to VapA can be considered a diagnostic parameter in horses' infection, (Sanz et al., 2016), bearing in mind that this antigen is unique to equin R. equi -human or other animal's isolated strains did not posses this antigen (Occampo-Sosa et al., 2007) In the study of IgG to VapA and IgG subclasses, it is shown that VapA-specifi c IgG(T) is the fi rst antibody with the increase of titar.Also, this immunoglobulin is potentially revised as a prognostic parameter in detection of naturally occurring rhodococcal pneumonia (Sanz et al., 2015).
Aft er infection, subclasses of detected antibodies include IgGa, IgGb, IgGc and IgM, with signifi cantly higher levels of IgGa and IgGb subclasses in foals than in horses.(Jacks et al., 2007a) Isotype IgGa is the most important in the prevention of pneumonia in foals (Hooper-McGrevy et al., 2003).
Antibodies against the VapA antigen administered to mice showed a dose-dependent protective eff ect.Aft er exposure to R. equi, the control group without applied antibodies died.Th e group that received a lower dose of the antibodies was partially protected, although the pathohistological examination showed bacteria in the spleen or lungs.In the mice that received the full dose, there were no deaths or signs of the disease, nor was the microorganism found in their tissues (Fernandez et al., 1997).
However, in a study that examined the passive immunization, in a colt with very low titer of antibodies, there have been no clinical manifestations of the disease, while another colt with a high titer developed serious illness with infausto prognosis.Levels of antibodies are not always in correlation with the opsonizing activity.Th erefore, it is assumed that non-specifi c plasma factors, such as lymphokines and interferons, can also impact the protective eff ect of the immune plasma (Martens et al., 1989).

CELLULAR IMMUNITY IN HORSES
Bearing in mind the intracellular nature of R. equi, it is believed that cellmediated mechanisms are the most eff ective against this disease (Vázquez-Boland et al., 2013).Aside from the disappearance of maternal antibodies, the decreased killing capacity of macrophages in foals contributes to their susceptibility in age of 1-5 months.(Berghaus et al., 2014) Th e study which compared bronchoalveolar and monocyte-derived macrophages, revealed that the replication of R. equi is greater in bronchoalveolar macrophages, with 3-100 fold increase during 48 hours (Berghaus et al., 2014).Th is is logical, having in mind that predominant route of infection is aerosol inhalation of R. equi (Giguère et al., 2011a).
Rhodococcus is able to survive intracellularly and to replicate in macrophages, both in murine peritoneal, as well as equine alveolar.Th e number of bacteria increases with time aft er the initial phase lag period of the fi rst 6-12 hours of infection, during which the number of intracellular bacteria remains unchanged.During the subsequent 36 hours, the number increases fi ve fold or more, to the inability of quantifi cation, while preserving cell viability.A similar phenomenon was observed with mycobacteria, with the fact that R. equi more rapidly divides.Replication within macrophages is characteristic of R. equi strains that possess VapA (Hondalus and Mosser, 1994).
Th e fi rst step in intracellular killing of pathogens is macrophage activation by interferon gamma.Th e CD4+ and CD8+ T lymphocytes from the lungs are able to produce interferon gamma (INF-γ) (Hines et al., 2003).Pulmonary clearance of the pathogen depends precisely of the increase in the number of pulmonary T cells and IFN-γ production (Kanaly et al., 1995;Hines et al., 2001;Hines et al., 2003).
In studies conducted on mice, it has been shown that there are two phenotypes of CD4+ T lymphocytes, which are referred to as Th 1 and Th 2. Th e classifi cation was based on the cytokines they produce: IFN-γ is characteristic of the Th 1 phenotype, IL-4 of the Th 2. Th e transfer of adequate cell lines, demonstrated that Th 1 phenotype led to the complete removal from the lungs (zero CFU), while in the Th 2 lines led to the development of pulmonary granuloma.Natural killer (NK) cells can also produce INF-γ, but transferring them alone does not lead to clearance of bacteria from the lungs (Kanaly et al., 1996).Th e production of IFN-γ in horses is triggered by VapA.It causes the same later response in horses as well as in mice (Hines et al., 2001;Lopez et al., 2002).Since the protein VapA is encoded by the vapA gene located on the plasmid (Tan et al., 1995), only the virulent strains, carriers of this plasmid, may be removed from the lungs (Hines et al., 2003).
Infection of macrophages by R. equi leads to a rapid translocation of NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells) from cytoplasm into nucleus, resulting in production of large amounts of infl ammatory cytokines.Some of the most important are TNF, interleukin IL-12 and nitric oxide, NO.Th ere is an assumption that these paths of non specifi c immunity are the main reason why the microorganism rarely causes diseases in immunocompetent adult animals (Darrah et al., 2004).
In a study on normal human peripheral blood mononuclear cells it has been demonstrated that they secrete TNF-α, IL-6 and IL-8 when they are stimulated with killed R. equi.Levels of TNF-α were increasing during the fi rst 48 hours, wherein the 90% of secreted cytokines were detected aft er 36 hours.For IL-6, this level was reached aft er only 12 hours.Production of IL-8 was at its maximum at the beginning, and leveled off between 12 and 48 hours (Pece et al., 1997).
Th e eff ect of foal's age on cytokine production is also revised.Th e expression of production signals in bronchoalveolar macrophages for IL-1β, IL-10, IL-12 p40 and IL-8, as well as IL-1β and IL-12 p40 production signals in bronchoalveolar lavage macrophages in 1-3 days old foals is higher than in foals of other age.Th is also indicates exposure to infection with R. equi in very early period of life (Berghaus et al., 2014).
It has been shown that the lack of galectin-3 receptor increases resistance to R. equi infections.Mice with and without this receptor were compared.Macrophages lacking this receptor are less susceptible to R. equi, its reproduction and survival.Th is receptor normally provides a balanced response to nonspecifi c immunity, aff ecting the production of IL-1β by macrophages (Ferraz et al., 2008).

COOPERATION OF HUMORAL AND CELLULAR IMMUNITY
Th e best descriptions of important interactions between cell-mediated immunity and humoral immunity come from in vitro studies that have investigated the mechanisms of killing R. equi by macrophages (Hietala and Ardans, 1987).
Alveolar macrophages in foals experimentally exposed to R. equi phagocyte and kill both non opsonized and opsonized microorganisms more effi ciently than alveolar macrophages in foals that were not exposed to the bacteria.Macrophage activity in foals which were already in contact with R. equi, was similar to the macrophage activity in adult horses.However, the rate of killing of opsonized R. equi was slightly lower in foals than in horses.In all tested alveolar macrophages, opsonization of R. equi signifi cantly increases phagolysosomal fusion when compared to nonopsonized bacteria (Hietala and Ardans, 1987;Giguère et al., 2011).
Th e use of plasma with antibodies also decreased bacterial viability in the extracellular matrix, but not in the macrophages (Dawson et al., 2011).

IMMUNOPROPHYLAXIS AND IMMUNOTHERAPY IN HORSES
Conventional therapy with antibiotics, with some being more eff ective than others (Giguère et al., 2015), is practically impossible without immunological support, with nature taking its course (Berghaus et al., 2011).Scientifi c attempts, with various approaches to increasing antibiotics in target sites included streptolysin O (Horohov et al., 2011;Gurel et al., 2013) and liposomal forms of antibiotic (Burton et al., 2015).In this way, we can "help nature" with immunomodulating support.
Field research conducted by Magnusson in 1923 and1924 failed to justify vaccination with an inactivated culture of R. equi.Vaccination of foals with isolates of R. equi, which were inactivated with formalin, failed to protect animals from a strong intratracheal exposure to the bacteria (Magnusson, 1923).Lopez et al. (2008) tested a live attenuated strain of R. equi which was a ribofl avin auxotroph on a mouse and a colt.Th is strain has kept the virulence plasmid.However, in a vaccine it is important to balance immunogenicity and attenuation.It is shown that such a vaccine must increase immunogenicity to demonstrate its success in neonatal foals.In order to make disease prevention more eff ective on endemic farms, a vaccine should be able to induce an immune response very early in life (Lopez et al., 2008).Currently, there is no vaccine, but new candidates are being tested (Cauchard et al., 2013).
Other attempts with immunostimulans containing Propionibacterium acnes showed no eff ect as imunomodulators in therapy in foals (Sturgill et al., 2011).
Pneumonia in foals can be prevented by vaccination of pregnant mares.Th ese vaccines contained VapA and exoenzim.Th e result of this research was the decline of mortality rates in cases of rhodococcal pneumonia on farms that have vaccinated mares from 3% to 1.2% (Becú et al., 1997).
Studies on a few endemic farms in Japan, United States and Argentina have shown that administration of hyperimmune plasma in foals led to protection in the earliest period of life.Some of the recommendations for prevention of pneumonia are: separation of foals in clean and ventilated paddocks, serological testing on 30th and 45th days aft er birth and immunization with plasma immediately aft er birth (Higuchi et al., 1999), no later than second day of life (Giguère et al., 2011a;Vázquez-Boland et al., 2013) Unspecifi c measures, such as feces removal from paddocks, avoiding spreading manure on pastures, irrigation of holding pens, foaling at pastures-gave unsatisfying results (Giguère et al., 2011b)

CONCLUSION
Only alive and virulent R. equi bacteria can cause immunity in mice.Avirulent live strains and the strains that were killed by heat did not show this activity (Takai et al., 1999).Eff ective immunization against intracellular bacterial pathogens requires the use of living microorganisms, rather than inactivated, in order to initiate a cellular immune response (Collins, 1988).
In passive immunization, the quality of donor's plasma is crucial: immunized donor's plasma protects signifi cantly from severe illnesses when compared to non immunized donor's plasma (Martens et al., 1989).In vitro reaction of passively immunized foals was similar to actively immunized mares (Becú et al., 1997).
In our country there is no immunoprophylaxis or immunotherapy for R. equi.Would it be reasonable to take into consideration collecting indigenous isolates, determining antigens and creating a vaccine that covers the antigenic determinants (including VapA) present in our area?Serum production on the same isolates could be professionally, scientifi cally and economically justifi ed.

AKNOWLEDGMENTS
Th is work has been fi nancially supported by a grant from Ministry of Education, Science and Technological Development, Republic of Serbia, Project numbers III 46012 and III 41012.
Th e authors wish to thank Milica Radovanović for her revision of the English language.