Deletion of serine/threonine-protein kinase pknL from Mycobacterium tuberculosis reduces the efficacy of isoniazid and ethambutol

Actionable mechanisms of drug tolerance and resistance in Mycobacterium tuberculosis

The emergence of antimicrobial resistance (AMR) across bacterial pathogens presents a serious threat to global health. This threat is further exacerbated in tuberculosis (TB), mainly due to a protracted treatment regimen involving a combination of drugs. A diversity of factors contributes to the emergence of drug resistance in TB, which is caused by the pathogen Mycobacterium tuberculosis (Mtb). While the traditional genetic mutation-driven drug resistance mechanisms operate in Mtb, there are also several additional unique features of drug resistance in this pathogen. Research in the past decade has enriched our understanding of such unconventional factors as efflux pumps, bacterial heterogeneity, metabolic states, and host microenvironment. Given that the discovery of new antibiotics is outpaced by the emergence of drug resistance patterns displayed by the pathogen, newer strategies for combating drug resistance are desperately needed. In the context of TB, such approaches include targeting the efflux capability of the pathogen, modulating the host environment to prevent bacterial drug tolerance, and activating the host anti-mycobacterial pathways. In this review, we discuss the traditional mechanisms of drug resistance in Mtb, newer understandings and the shaping of a set of unconventional approaches to target both the emergence and treatment of drug resistance in TB.

Elimination of PknL and MSMEG_4242 in Mycobacterium smegmatis alters the character of the outer cell envelope and selects for mutations in Lsr2

•

Elimination of pknL and adjacent gene MSMEG_4242 in M. smegmatis produces rough & smooth colonies.

•

All smooth colony mutants have inactivated lsr2 genes.

•

Inactivated lsr2 leads to high expression of MSMEG_4727 and increased synthesis of LOS.

•

Smooth mutants don’t form biofilms, have longer bacilli and increased sliding motility.

•

Rough strains deleted for pknL and MSMEG_4242 form biofilms but have aberrant sliding motility.

•

We propose a phosphorylation cascade of PknB phosphorylating PknL that then phosphorylates Lsr2.

Four serine/threonine kinases are present in all mycobacteria: PknA, PknB, PknG and PknL. PknA and PknB are essential for growth and replication, PknG regulates metabolism, but little is known about PknL. Inactivation of pknL and adjacent regulator MSMEG_4242 in rough colony M. smegmatis mc²155 produced both smooth and rough colonies. Upon restreaking rough colonies, smooth colonies appeared at a frequency of ~ 1/250. Smooth mutants did not form biofilms, showed increased sliding motility and anomalous lipids on thin-layer chromatography, identified by mass spectrometry as lipooligosaccharides and perhaps also glycopeptidolipids. RNA-seq and Sanger sequencing revealed that all smooth mutants had inactivated lsr2 genes due to mutations and different IS1096 insertions. When complemented with lsr2, the colonies became rough, anomalous lipids disappeared and sliding motility decreased. Smooth mutants showed increased expression of IS1096 transposase TnpA and MSMEG_4727, which encodes a protein similar to PKS5. When MSMEG_4727 was deleted, smooth pknL/MSMEG_4242/lsr2 mutants reverted to rough, formed good biofilms, their motility decreased slightly and their anomalous lipids disappeared. Rough delpknL/del4242 mutants formed poor biofilms and showed decreased, aberrant sliding motility and both phenotypes were complemented with the two deleted genes. Inactivation of lsr2 changes colony morphology from rough to smooth, augments sliding motility and increases expression of MSMEG_4727 and other enzymes synthesizing lipooligosaccharides, apparently preventing biofilm formation. Similar morphological phase changes occur in other mycobacteria, likely reflecting environmental adaptations. PknL and MSMEG_4242 regulate lipid components of the outer cell envelope and their absence selects for lsr2 inactivation. A regulatory, phosphorylation cascade model is proposed.