Brucellosis is a zoonotic disease caused by bacteria of the genus Brucella, that also entails large economic losses. The main infection source for humans are ruminants, in which the disease induces abortion and other reproductive disorders. In most affected regions, vaccinating animals is the only realistic strategy to control the infection. The attenuated strain B. melitensis Rev1 is the only recommended vaccine against ovine and caprine brucellosis, but it has several drawbacks, such as virulence when administered in pregnant animals, serological interference, possibility of infecting humans, and resistance to treatments including streptomycin, that urge the search for new vaccines. The smooth lipopolysaccharide (S-LPS) is a well-known virulence factor and the immunodominant antigen in diagnostic tests, thus, its modifications represent an interesting approach in the research for a safe vaccine. Among S-LPS biosynthesis pathways, the Wzm/Wzt two-component system is the responsible of transporting the O-PS synthesised onto the cytoplasmic side of the inner membrane to the periplasm, for its subsequent assembly to the core-lipid A to form the S-LPS. Blocking this pathway leads to the generation of rough LPS (R-LPS) mutants capable of accumulating O-PS within the bacteria, that may be decisive for an effective immune response against the infection. However, the biological implications of each transporter protein are still unknown. In particular, the O-PS blockage could lead to modifications in bacterial envelope, since the O-PS transport requires the participation of bactoprenol as a carrier molecule, that is also involved in the biosynthesis of other cellular structures. Aiming to deepen the knowledge about this system and its potential application in vaccines development, in Chapter 1 of this thesis, it is presented the construction and characterisation of single and double Δwzm/Δwzt mutants derived from Rev1. In addition, for comparative purposes, ∆wzm mutants were constructed from B. abortus 2308 and S19 and studied together with a 16M∆wzm mutant, derived from the B. melitensis 16M virulent strain previously developed in the Animal Health Group of the Institute of Agrobiotechnology (IdAB). As a result, the mutants lacking outer O-PS exhibited changes in gene expression, antigenic properties of inner O-PS, and phenotypic changes associated with outer membrane and/or cell wall modifications. In addition, Rev1 ∆wzm/∆wzt mutants were highly susceptible to polymyxins and other antibiotics used in human brucellosis treatment, particularly, to streptomycin; and they showed a large attenuation in mouse. Among them, Rev1∆wzm stood out for inducing a transient peak of splenomegaly, being highly immunogenic, and generating effective protection against B. melitensis and B. ovis virulent infections in murine model. In Chapter 2, prior to studying the safety of Rev1∆wzm in the natural host, it was determined the ability of this vaccine candidate to grow in Farrell (FM) and CITA (CM) selective culture media, currently recommended for Brucella primary isolation. The mutant strong susceptibility to polymyxins and the synergy between colistin and vancomycin observed led to the severe inhibition of the vaccine candidate in both selective media. To solve this, the named Brucella Selective Medium (BSM) was reformulated and evaluated with more than 1,700 ovine field samples, being able to improve the previous media performance. Furthermore, by exploring the productivity of BSM for other brucellae, it was observed that both BSM and CM inhibited the growth of a B. abortus bv1 strains collection. The analysis of the possible causes led to the development of a second selective medium called Brucella Selective Improved Culture (BruSIC), which was permissive for all the Brucella strains analysed, and as inhibitory as CM for the usual contaminant bacteria in veterinary samples. In addition, BruSIC offers the advantage of not requiring bovine serum to isolate serum-dependent Brucella strains, simplifying medium preparation, lowering costs and avoiding the use of compounds from animal sources. In Chapter 3, it was analysed the safety and immune response induced in lambs inoculated with Rev1∆wzm, demonstrating the activation of antibodies development without interfering in the standard serological tests for brucellosis diagnosis. These results, as well as others achieved by the research group in adult cattle and pregnant sheep, make this mutant a promising vaccine candidate to control B. melitensis and B. ovis infections in small ruminants. The content of these three thesis chapters has been published in the scientific journals Frontiers in Microbiology (Frontiers group) and Microbiology Spectrum (ASM).