Culture filtrate helps, but can’t fully rescue, tiny tuberculosis samples

Tuberculosis diagnosis and research rely on the ability to grow Mycobacterium tuberculosis (Mtb) from patient samples, yet some bacterial forms, called differentially culturable (DC) phenotypes, slip through routine tests. DC Mtb can reduce the sensitivity of sputum culture and may be linked to worse treatment outcomes for tuberculosis patients. Accurately counting these hard-to-grow bacteria is therefore a pressing research question. Current laboratory approaches often combine Most Probable Number (MPN) assays with supplementation using culture filtrate (CF), a preparation thought to help dormant or slow-growing Mtb resume replication. Those methods rest on an assumption that each bacterial cell has the same chance of growing, regardless of how many neighbors are present—a point that had not been directly tested for Mtb. To examine that assumption and to measure how CF changes growth, researchers led by corresponding author Digby F. Warner performed a systematic dilution experiment. They made a half-logarithmic dilution series of Mtb spanning 70,000 to 0 CFU/mL, and then cultured each inoculum either in standard 7H9 medium or in medium supplemented with CF, tracking growth over time by optical density. This setup allowed the team to compare growth probability and timing across a wide range of starting densities.

The study used a controlled dilution series and optical density monitoring to follow whether and when cultures became positive. When starting inocula were at or above 2,000 CFU/mL, those cultures were 33 times more likely to show growth than inocula at or below 700 CFU/mL. Adding culture filtrate (CF) changed the picture: CF increased the odds that a low-density inoculum would grow by five-fold, and it shortened the time-to-positivity by 294 hours, roughly 12 days, compared with 7H9 medium alone. However, CF’s ability to promote growth was not uniform across densities. Its growth-promoting effect diminished as the inoculum density rose and was effectively negligible at the highest tested density, 70,000 CFU/mL. Beyond effects on growth probability and timing, CF altered bacterial cell physiology: cultures grown with CF produced shorter bacilli that were less likely to incorporate the mycomembrane probe DMN-trehalose. These combined findings show both a strong density-dependence in Mtb culturing and broader physiological changes when CF is present.

The results carry several practical and scientific implications. First, Mtb is much harder to culture from very low starting numbers, and culture filtrate only partially compensates for that difficulty. Because growth probability varies with starting density, the standard assumption used in unsupplemented MPN assays—that each cell is equally likely to grow—appears flawed for Mtb, meaning those assays may systematically underestimate colony-forming units (CFU) in samples with many low-density bacteria. Second, culture filtrate does more than simply ‘‘resuscitate’’ DC Mtb: its effects are density-dependent and it changes cell shape and mycomembrane properties, as shown by differences in DMN-trehalose incorporation. That suggests CF influences multiple aspects of mycobacterial physiology and population behavior, not just awakening dormant cells. Taken together, the findings argue for improved methods to detect and quantify DC Mtb phenotypes. Better tools could strengthen clinical care by revealing bacteria missed by current sputum culture and help researchers uncover the factors that govern mycobacterial replication at different population densities.