Brucellosis is a major bacterial zoonosis with a worldwide distribution. The causative organisms are Gram-negative bacteria that belong to the class α-proteobacteria. These pathogens can affect a wide variety of mammals, including humans. Eight species are known to infect specific animals: sheep and goats (Brucella melitensis), cows (Brucella abortus), pigs (Brucella suis), dogs (Brucella canis), rodents (Brucella neotomae), sheep (Brucella ovis) and marine species (Brucella pinnipediae and Brucella cetacea) (Refs Reference Bricker1, Reference Nielsen2). Among these, three are virulent for humans (B. melitensis, B. abortus and B. suis). Human brucellosis, also known as Malta Fever, remains endemic in the Mediterranean basin, the Middle East, Central Asia, Africa and Latin America (Ref. Reference Corbel3).

Domestic and wild animals acquire brucellosis through ingestion of contaminated milk or from contact with contaminated tissues and fluids associated with birthing or abortion. These animals potentially serve as a source for reintroduction of brucellosis to additional food animals (Ref. Reference Davis and Elzer4). For these reasons, Brucella spp. are of great concern to the agricultural community worldwide (Ref. Reference Boschiroli, Foulongne and O'Callaghan5). In many industrialised countries, brucellosis in food animals has been effectively controlled through the implementation of vaccination, surveillance and quarantine programmes. However, the high cost of these controls can have a significant impact on the agricultural economy.

Transmission of brucellosis to humans occurs through skin abrasions when in contact with infected animals or by the consumption of unpasteurised dairy products. No human-to-human transmission has been identified; however, people may be infected by inhalation of contaminated dust or aerosols. This accounts for Brucella spp. being one of the most common laboratory-acquired pathogens worldwide (Ref. Reference Bouza6), and for their inclusion in the potential biological weapons list.

Pathogenic Brucella spp. cause abortion in female animals by colonisation of placental trophoblasts, and can cause sterility in male animals (Ref. Reference Corbel3). Human infections with B. abortus or B. melitensis may present in varying forms and can become life-threatening. Although the majority of patients present with general symptoms, such as fever, malaise, sweats and lymphadenopathy and/or hepatosplenomegaly, a more severe form of the disease can also occur, usually accompanied with osteoarticular problems or (in males) epididymoorchitis. More serious and hard-to-treat complications include, in descending order of frequency, spondylitis, neurobrucellosis, liver abscesses and endocarditis. The human disease tends to recur, despite adequate antibiotic regimens.

B. melitensis, B. abortus and B. suis have long been considered as potential biological weapons, and are classified as category B pathogens. In fact, one important bioterrorism-related characteristic of B. melitensis is the very small inoculum needed to induce human disease: 10 to 100 microorganisms are sufficient to cause a debilitating infection for several weeks (Ref. Reference Pappas7). Moreover, as Brucella spp. can be easily aerosolised, this pathogen could be used to develop biological weapons. Compared with other potential biological weapons, the inoculation period with Brucella spp. is relatively protracted, ranging from 9 to 60 days. Thus, the bioweapon potential of Brucella spp. may actually be only of historical importance (Ref. Reference Pappas7).

Three animal vaccines exist to prevent brucellosis: two smooth live vaccine strains – S19 for cattle and Rev 1 for small ruminants – and the rough attenuated strain RB51 for cattle. However, these vaccines cannot be used in humans because they offer only moderate protection along with a considerable risk of side-effects, including induction of antibiotic (rifampicin)-resistant cases of certain strains of the disease (Refs Reference Banai8, Reference Blasco and Diaz9, Reference Moriyon10, Reference Nicoletti11). There is presently no safe and effective vaccine that can be used to prevent the human disease and treatment requires a combination of two antibiotics – doxycycline and rifampicin, or doxycycline and streptomycin – for prolonged periods (Ref. Reference Young12).

Brucella cell biology

Invasion of phagocytic and nonphagocytic cells

Pathogenic Brucella can invade and multiply intracellularly in both phagocytic cells (J744 and THP1 macrophages) (Refs Reference Celli13, Reference Liautard14) and nonphagocytic cells (HeLa, Vero, trophoblast and 3T3 fibroblasts) (Refs Reference Pizarro-Cerda15, Reference Pizarro-Cerda, Moreno and Gorvel16, Reference Delrue17). Several studies show that once inside the cell, B. abortus is surrounded by a membrane-bound compartment to form the Brucella-containing vacuole (BCV). It is able to escape the endocytic pathway to reach a safe place within the cell where the bacteria are protected from host-defence mechanisms to ensure their replication and thus their virulence (Refs Reference Kohler18, Reference Kohler19) (Fig. 1a).

Figure 1.

Model for Brucella spp. trafficking.(Legend; see previous page for figure.) (a) Pathogenic Brucella invades phagocytic and nonphagocytic cells through lipid rafts (1). After lipid-raft-mediated entry, intracellular pathogenic Brucella resides within an early Brucella-containing vacuole (BCV) that interacts with early endosomes (2). These endosomes are marked by positive EEA, Rab5 and TfR expression. Pathogenic Brucella avoids further interaction with the endocytic pathway in an intermediate nonreplicative BCV (3). The bacterium then either interacts and fuses with the endoplasmic reticulum (ER) (4) to generate an ER-derived organelle permissive for Brucella spp. replication, or fails to maintain interaction with the ER and is degraded by fusion with phagolysosomes (5). Once safely in the ER within replicative BCVs (6), Brucella spp. divide rapidly. (b) Model for evasion of lysosomal killing through expression of virulence factors. Brucella spp. entry is dependent on its LPS and the regulatory system BvrR/BvrS at the surface. LPS and other virulence factors, such as the phosphatidylcholine synthase and the cyclic β-1,2 glucan, then play a role in the early stages of intracellular trafficking. Subsequently, interaction and fusion with the ER elements leading to bacterial replication is dependent on the VirB type IV secretion system, since a virB-defective mutant fails to sustain fusion-proficient interactions with the ER and ultimately fuses with lysosomes. Abbreviations: BvrR, Brucella virulence-related regulatory protein; BvrS, Brucella virulence-related sensory protein; CGS, cyclic β-1,2 glucan synthetase; LPS, lipopolysaccharide; PCS, phosphatidylcholine synthase; VirB, VirB type IV secretion system.

Brucella and autophagy

The occurrence of autophagy in Brucella spp. intracellular trafficking remains a matter of debate. In epithelial cells, B. abortus can be found in autophagosomes (Ref. Reference Pizarro-Cerda15); however, autophagic BCVs have not been detected in macrophages (Ref. Reference Celli13). One hypothesis is that autophagy might be a host-defence mechanism that prevails in nonphagocytic cells to fight invading bacteria. Such a method of attack seems limited to nonphagocytic cells. Indeed, macrophages seem to be efficient at killing Brucella and even with wild-type B. abortus infection, 80–90% of bacteria are killed within the first few hours of infection, which is not the case in epithelial cells. However some bacteria resist macrophage killing and are able to survive and replicate (Ref. Reference Celli13).

Clinical implications

Infiltration of Brucella spp. into professional and nonprofessional phagocytes causes the wide spectrum of clinical symptoms of brucellosis presented and sets important therapeutic limitations, both in terms of antibiotic use and in terms of human vaccine development. Human brucellosis can involve practically any body system, and the clinical manifestations are often considered as an equilibrium between expression of pathogen virulence factors and the host immune response. In humans, the immune response is of a predominantly cellular Th1 type, mediated principally by interferon γ; a humoral component may also be implicated in the early stages. The affinity of the pathogen for the reticuloendothelial system means that bone marrow may serve as the ideal niche for isolating the pathogen.

By hiding intracellularly without interrupting the cell life cycle Brucella spp. manage to largely escape the host immune response: thus, bacterial multiplication can continue for prolonged periods, resulting in a relapsing form of the disease, with months or even years of apparently disease-free periods. The recent recognition of positive results for Brucella spp. gene expression in real-time PCR in apparently healthy individuals, years after disease manifestation and successful treatment, suggests that eradication of the pathogen may not even be feasible (Ref. Reference Navarro55).

Importantly, the acidic environment of the BCV serves as a neutraliser of antibiotics. Thus, macrolides are typically inactive against brucellosis. Furthermore, antibiotics with acknowledged in vitro and in vivo activity against Brucella spp. may show an attenuated intracellular activity, necessitating their prolonged and combined administration. Monotherapeutic regimens in the past have been largely disappointing (Ref. Reference Pappas, Akritidis and Tsianos56), and attempts at shortening the duration of therapeutic regimens have shown a statistically significant higher relapse and failure percentage (Ref. Reference Solera57). This phenomenon underlines the dissociation between in vitro susceptibility testing and in vivo antibiotic performance, as exemplified by numerous cases of treatment failure (Ref. Reference Al Dahouk58).

Research in progress and outstanding research questions

The unique and complex nature of Brucella spp. pathophysiology carries significant issues for the development of a human vaccine. Low LPS virulence and immunogenicity (relative to other Gram-negative pathogens) has hampered the development of live attenuated vaccines. This could be because the mutation of a specific single gene has little effect on virulence unless it is a global regulator, or because two mutations result in too great an attenuation that an immune response is not induced. Recent years have seen various attempts to develop a DNA vaccine for Brucella spp. utilising numerous peptides (Refs Reference Al-Mariri59, Reference Cassataro60, Reference Cassataro61, Reference Munoz-Montesino62, Reference Yu, Hu and Cai63). Most of the studies have shown satisfying results in animal models but no clear potential for further experimentation in humans. The main reason behind this failure is that targeting only the Th1 response may be inadequate. In the latter stages, cytotoxicity may be more important in controlling human infection, whereas in the early bacteraemic stages, a humoral response may limit disease spread and the appearance of focal complications. The development of a bivalent (or even trivalent) vaccine encompassing all these aspects is at present stalled by problems with vectors of DNA vaccines, and by an inability to select the optimal combination of peptides. Moreover, the predominant immunogenicity of LPS and outer-membrane peptides in general might minimise the immunogenic response elicited by other vaccine components (for a review, see Ref. Reference Al-Mariri59).

Study of the interaction between Brucella spp. and the host reveals unique properties, which are mainly focused on the low activation profile of the immune response. The two main factors responsible are the Brucella spp. LPS and the residence of the pathogenic bacteria in safe intracellular compartments, such as the ER. The direct consequence is that these properties affect treatment options. In addition, the fact that it is very difficult to predict the clinical manifestation in humans increases the problems of diagnosis, especially in resource-deprived countries. The disease manifests in vague symptoms, tends to be chronic or relapsing, and needs protracted antibiotic combination treatment for eradication, if the latter is feasible. A hope for the future will be to bring basic science and clinical developments in the field of diagnosis and treatment of brucellosis. Indeed, the complex interplay between the pathogen and host immune response further enforces significant problems in human vaccine development because, at the molecular level, pathogenic Brucella spp. have developed a ‘stealth’ strategy by expressing molecules with low stimulatory activity and toxicity for cells, allowing the bacterium to reach its replication niche before activation of antimicrobial mechanisms by the host. In conclusion, brucellosis is a complex disease, which requires not only optimisation of treatment, but also a good relationship between physicians, patients and molecular bacteriologists.