To our knowledge, this is the first study of macrophage– pathogen interactions in which both the macrophages and the pathogens have been genetically modified, such that the host cells do or do not express specific cytotoxic mechanisms, and the bacteria do or do not express a presumptive resistance pathway directed against those mechanisms. By this analysis, noxR1, a novel gene from M. tuberculosis, confers partial resistance to three of the major antimicrobial products of macrophages, the cells ultimately responsible for controlling tuberculosis. However, the greater resistance conferred on noxR1-transformed M. smegmatis in vitro than in macrophages strongly suggests that there are macrophage antimycobacterial products other than RNI, ROI, and H+, and that noxR1 protects against some physiologically relevant stresses but not others. Likewise, the relative selectivity of noxR1's protective effects in vitro and the restriction of the noxR1 gene to a subset of pathogenic mycobacterial species argue that noxR1 is neither a general stress resistance gene nor a housekeeping gene.
Cloned from a prevalent clinical isolate of M. tuberculosis, noxR1 was only identified in the genome of members of the M. tuberculosis complex. noxR1 was absent from the chromosome of mycobacteria considered nonpathogenic or opportunistically pathogenic for humans, including M. smegmatis, M. avium, and M. intracellulare. We do not know if a noxR1-like gene is present in any other organisms, nor whether noxR1 is transcribed naturally by any mycobacteria besides the M. tuberculosis strain we tested. It remains to be determined whether the natural gene may be regulated by environmental conditions, including the stresses against which it confers resistance. It would be of particular relevance to know how much NoxR1 is expressed by M. tuberculosis that resides in phagolysosomes.
A major mystery is noxR1
's mechanism of action. With no homologies or motifs recognized at nucleotide or amino acid levels, the sequence afforded few clues. Because so little NoxR1 appears to be expressed, it is unlikely that its four cysteine residues merely serve to titrate ROI or RNI, as homocysteine is thought to do in Salmonella typhimurium
), or as metallothionein may do when overexpressed in hepatocytes (57
). In E. coli
, the DNA-binding protein encoded by dps
protects DNA from oxidative damage (58
), and noxR1
might work in a similar manner. Its effectiveness in a heterologous mycobacterium whose genome lacks noxR1
may argue against a role as a transcription factor, and its effectiveness in oxyR
- and soxRS
-deficient E. coli
argues against noxR1
activating those two regulons in particular. The oxyR
homolog of M. tuberculosis
contains numerous deletions and frameshifts and is nonfunctional (9
). Perhaps NoxR1 compensates for the loss of OxyR in M. tuberculosis
, similar to the suggested role of AhpC (10
If NoxR1 is an enzyme, the novelty of its sequence suggests that it may work differently than those already known to affect RNI. The latter serve to alter the proportions of various RNI in a mixture. Thus, in vitro, mammalian thioredoxin reductase can catalyze the NADPH-dependent reduction of S
-nitrosoglutathione to glutathione and an NO-like species (61
), whereas superoxide dismutase favors the accumulation of NO at the expense of its conversion to peroxynitrite. At present, the yield of recombinant NoxR1 has been compromised by its apparent autotoxicity upon overexpression, and this has precluded biochemical assays of hypothesized actions.
The physiologic role of NoxR1 cannot be established until it is possible to inactivate noxR1 selectively in M. tuberculosis and compare the growth of the organism in the mammalian host with the growth of isogenic M. tuberculosis to which noxR1 has been restored. Until then, the possibility remains that the actions of noxR1 observed in transformed bacteria are artefacts of over or heterologous expression. Weighing against this concern is that noxR1 conferred a protective phenotype only when expressed at low levels, and did so in diverse species and strains.
There is an urgent need for new antitubercular drugs. Almost all currently used antibacterials manifest their antibacterial activity in pure culture, and new candidates are screened on that basis. Such screens could miss compounds that inhibit a pathway which is dispensable for bacterial growth in vitro but which confers a survival advantage on the pathogen within the infected host. In this sense, inhibitors of NoxR1 may warrant investigation as possible prototypes of a new class of antiinfectious agents.