There is an urgent need for new targets for development of antibiotics. Pathways that are non-essential because the bacterium has more than one alternative to perform the synthesis have been overlooked in this search. We hypothesise that redundancy targets may be useful, provided one can identify all components of the redundant pathway and block them in parallel.
To test this hypothesis we developed a genome scale model of Salmonella Typhimurium and used this to identify possible redundancy pairs. We then selected two predicted pairs: Spermidine-uptake and spermidine biosynthesis genes (potCD and speE) and ammonia and glutamate dependent asparagine synthetase genes (asnA and asnB). Single and double mutants, and double mutants complemented in trans, were constructed and characterized for growth and virulence. Single mutants were fully virulent or only marginally attenuated, while double mutants were severely attenuated, showing that spermidine and asparagine are essential for virulence of S. Typhimurium. The virulence could be restored by in trans complementation. Corresponding mutants in S. Gallinarum and S. Dublin behaved similarly when tested for virulence in chicken and mice.
Knock out of spermidine biosynthesis in a spermidine-transporter knock-out background caused reduced expression of sseA encoded from SPI-2, suggesting that the Salmonella dependence of spermidine for full virulence relates to the function of SPI-2. The asnA/asnB double mutant grew normally in rich medium but was asparagine auxotrophic in minimal medium, and the attenuation may relate to lack of asparagine at sites of infection.
In conclusion, we have developed an in silico approach to identification of redundancy genes. Using this model, we predicted that spermidine biosynthesis and spermidine transportation as well as asparagine synthesis via asnA and asnB would be essential for infection in S. Typhimurium and this prediction was proven correct. The results suggest that redundancy genes may be useful targets for infection control.