Mammalian cells function optimally in a stable environment, where temperature, metabolic gases, and pH are controlled within tight margins via homeostasis. In human tissues, pH is maintained via the primary physiological buffer system, CO₂/HCO₃− [1, 2]. However, exposure to excess atmospheric CO₂ could decrease the ratio of bicarbonate to CO₂, resulting in a shift towards a more acidic environment [3]. To understand the impact of this shift on human health, we must first examine how extracellular pH changes affect cellular performance. The acid–base balance of mammalian fluids is regulated by a suite of processes that maintain pH within a narrow range for optimal for cellular function (termed physiological pH) [4]. In vitro, extracellular pH is moderated by four processes; gas exchange at the atmosphere/media interface, media buffering capacity, partial pressure of carbon dioxide, and cellular metabolism [5]. Such processes operate simultaneously, making it difficult to control extracellular pH in cell culture within optimal ranges [6, 7]. Standardized techniques, including the use of exogenous buffers and CO₂ incubators, that attempt to maintain physiologically relevant conditions during cell culture are commonly used. However, broad departures from CO₂, O₂, and pH setpoints are frequently encountered in routine experiments [6, 7], raising concerns over how such instabilities introduce experimental artifacts and affect reproducibility [2, 5]. Previous studies focusing on the effects of pH deviations on cellular performance typically use exogenous acid and/or bases (e.g., HCl, NaOH, non-volatile buffers) to adjust pH [2, 5], limiting the physiological relevance of the results obtained.

Here we developed a method, relying solely on the physiological buffer system, to address this research gap. We demonstrate that this method tightly maintains nominal pH levels within ± 0.013 units (SE) and sustains levels within physiological ranges (defined 7.35–7.45 for blood cells or physioxia) during prolonged cultures (hours to days). We utilized this approach to analyze the transcriptome of the human GM12878 cell line—a widely benchmarked cell line—cultured under different medium pH levels (6.8, 7.0, 7.2, and 7.4) during 72 h of culture.

We modified the automatic control scripts and gas sparging system of the DASbox® Mini bioreactor system to maintain pH and physioxia using only gases and physiological buffers. Briefly, pH control was achieved by sparging CO₂ or N₂, when pH was higher or lower than the setpoint, respectively, in an automated feedback loop. Similarly, dO₂ was stably maintained using pure O₂ sparging (see Methods). These gases were delivered in a medium buffered solely with the physiological CO₂/HCO₃− buffer system. GM12878 cultures were stably maintained at four different pH levels, ranging from the physiological pH 7.4 (pH 7.25–7.45 in mammalian arterial blood and irrigated tissues [4]) to 6.8, a value routinely observed in batch culture [5]. Across the four pH treatments, dO₂ was constantly maintained at physiological dO₂ levels (~85%) (Fig. 1A). The modified bioreactor setup permitted the stable control of temperature (37°C ± 0.001°C, n = 146), pH, and dO₂ levels throughout the experiment (Fig. 1B and 1C). Custom-made diffusers were used to enhance rates of gas diffusion in the medium (Fig. 1G). The diffusers created microbubbles of the gases, which quicken the rate of gas/medium equilibration by increasing the contact area between the liquid medium and the gases. Gas sparging rates that maintained the setpoints are shown in Fig. 1D–F. Cell performance and transcriptome responses were examined every 24 h.