There are more than 180 species of mycoplasma bacteria, and their prevalence in cell culture across many fields have been well studied. Due to a lack of a cell wall, mycoplasmas are very small (<0.3 µm), can be easily transferred in contaminated cell media, and infect eukaryotic cells by adhering to the host cell membranes due to specialized adhesin-rich attachment organelles for fusion. Mycoplasma contamination can cause a variety of problems for cell culture including affecting the host cell physiology, metabolism, productivity and safety. There are additional potential pathogenic risks to patients if the injectable biologic is contaminated. There are a variety of detection methods used, and if detected, cultures and manufactured materials are disposed of and laboratories are shut down for decontamination. Although the potential issues associated with mycoplasma contamination are well known, there has been limited work investigating the possible early cell culture process parameters that may be diagnostic of an infection, especially in a biomanufacturing-representative process.
In “Impacts of intentional mycoplasma contamination on CHO cell bioreactor cultures,” the research team intentionally contaminated their process vessels to look at how standard process and metabolic parameters were affected. Using six single-use rocker bags, DG44 CHO line cells expressing an IgG1 were cultured in perfusion mode. Two rocker bags were controls (uncontaminated) and four bags were spiked with M. arginini mycoplasma at different time points during the process and at varying levels of inoculation. Process analytics were collected daily for all the rocker bags, including viable cell density (VCD), pH, glutamine, glucose, lactate, glutamate, ammonia, amino acids, and titer to probe if there was any correlation to the mycoplasma growth to the CHO culture metabolism.
The study revealed that the general growth rate of M. arginini, and therefore, the CHO culture’s contamination level was independent of the inoculum density. This suggests that the growth profile of the mycoplasma is more dependent on competition (or limitations) for media components as CHO cells are growing within the same bioreactor. This hypothesis was supported by the stalling of the CHO growth profile in all bioreactors that were infected with M. arginini. Despite the decrease in CHO growth, CHO cell viability did not correlate well with the presence of the M. arginini.
The levels of certain nutrients and metabolites gave the indication that the contaminated cultures differed metabolically during the processes. The levels of glucose and glutamine were significantly higher at the end of the process with the high inoculum M. arginini bioreactors compared to the controls. The one low level inoculated bioreactor had slightly higher levels of glucose compared to the controls, but the levels of glutamate were not significant enough to indicate a difference. The final ammonia levels of all the M. arginini inoculated cultures all were greater than 10 mM, while the control systems had levels below 10 mM. As a result, an unexplained rise in ammonia levels from a culture could indicate contamination by M. arginini. Also, glutamate levels were higher in the inoculated cultures consistent with the mycoplasma consumption of glutamine resulting in the excretion of glutamate and ammonia into the bioreactors. Because lactate levels correlated to the CHO growth it could not be used, on its own, as a distinguishing marker of mycoplasma contamination in this study. Interestingly, the levels of arginine dropped more than 90% after inoculation which could be correlated to the higher levels of ammonia in those cultures. The authors noted that arginine is significant in metabolic processes of mycoplasma and can also lead to the production of ornithine, ATP and CO2, all of which may be worthwhile to monitor in future studies. Concerning other process variables, it was noted that the levels of dissolved oxygen and CO2 began to increase in the M. arginini bioreactor. Finally, the cell specific productivity early in the cultures was not significantly different between control and inoculated cultures since even infected cultures still produced IgG. Therefore, this variable was not a reliable indicator of early infection of mycoplasma like M. arginini.
Since mycoplasmas compete with host cells for nutrients, this study provided insight into potential process parameters that might be worth more routine monitoring. Traditional mycoplasma detection approaches like staining, cultures, immunofluorescence, ELISA or PCR all take significant time to complete. Additionally, these approaches are conducted after the process is finished, so they cannot offer a near-real time indication that a culture is contaminated. Although this study only looked at a single mycoplasma strain (M. arginini), certain metabolic markers like the accumulation of ammonia, depletion of arginine and dissolved oxygen levels may be used as an early indicator that a process may be contaminated.
By: Glenn A. Harris, Director of Integrated Life Sciences Platforms