Discrimination between Mycobacterium tuberculosis and Mycobacterium bovis using Fourier transform infrared...

Area of Science:

• Microbiology
• Spectroscopy
• Bioinformatics

Background:

• Zoonotic tuberculosis (TB), primarily caused by Mycobacterium bovis (Mbo), impedes global TB eradication efforts.
• Distinguishing Mbo from Mycobacterium tuberculosis (Mtb) is crucial for effective diagnostics and surveillance.
• Fourier Transform Infrared Spectroscopy (FT-IRS) offers potential for rapid microbial identification.

Purpose of the Study:

• To evaluate FT-IRS using the IR Biotyper® system for differentiating clinical isolates of Mbo and Mtb.
• To compare bacterial inactivation methods for FT-IRS analysis.
• To assess the utility of FT-IRS for distinguishing other tuberculous mycobacteria, such as Mycobacterium africanum (Maf).

Main Methods:

• Two inactivation protocols (paraformaldehyde and heat) were tested on Mtb and Bacillus Calmette-Guérin (BCG) strains.
• FT-IRS analysis was performed on Mtb and Mbo isolates, focusing on polysaccharide, protein/carbohydrate, and lipid spectral regions.
• Machine learning algorithms, including Linear Discriminant Analysis, UPGMA dendrogram, and artificial neural networks, were employed for spectral data analysis.

Main Results:

• Heat inactivation was found to be more efficient for biomass recovery and ease of use.
• Clear spectral separations between Mtb and Mbo were observed, particularly in the polysaccharide region.
• An artificial neural network classifier achieved 99% accuracy in distinguishing Mbo from Mtb.
• FT-IRS also demonstrated the ability to differentiate Mycobacterium africanum (Maf) strains.

Conclusions:

• FT-IRS is a highly effective, rapid, and accurate tool for differentiating Mbo and Mtb.
• The method shows significant potential for improving zoonotic TB diagnostics and surveillance.
• FT-IRS can be extended to differentiate other tuberculous mycobacteria like Maf, broadening its diagnostic applicability.