Improving accuracy of cell and chromophore concentration measurements using optical density

Abstract

Abstract Background UV–vis spectrophotometric optical density (OD) is the most commonly-used technique for estimating chromophore formation and cell concentration in liquid culture. OD wavelength is often chosen with little thought given to its effect on the quality of the measurement. Analysis of the contributions of absorption and scattering to the measured optical density provides a basis for understanding variability among spectrophotometers and enables a quantitative evaluation of the applicability of the Beer-Lambert law. This provides a rational approach for improving the accuracy of OD measurements used as a proxy for direct dry weight (DW), cell count, and pigment levels. Results For pigmented organisms, the choice of OD wavelength presents a tradeoff between the robustness and the sensitivity of the measurement. The OD at a robust wavelength is primarily the result of light scattering and does not vary with culture conditions; whereas, the OD at a sensitive wavelength is additionally dependent on light absorption by the organism’s pigments. Suitably robust and sensitive wavelengths are identified for a wide range of organisms by comparing their spectra to the true absorption spectra of dyes. The relative scattering contribution can be reduced either by measurement at higher OD, or by the addition of bovine serum albumin. Reduction of scattering or correlation with off-peak light attenuation provides for more accurate assessment of chromophore levels within cells. Conversion factors between DW, OD, and colony-forming unit density are tabulated for 17 diverse organisms to illustrate the scope of variability of these correlations. Finally, an inexpensive short pathlength LED-based flow cell is demonstrated for the online monitoring of growth in a bioreactor at culture concentrations greater than 5 grams dry weight per liter which would otherwise require off-line dilutions to obtain non-saturated OD measurements. Conclusions OD is most accurate as a time-saving proxy measurement for biomass concentration when light attenuation is dominated by scattering. However, the applicability of OD-based correlations is highly dependent on the measurement specifications (spectrophotometer model and wavelength) and culture conditions (media type; growth stage; culture stress; cell/colony geometry; presence and concentration of secreted compounds). These variations highlight the importance of treating literature conversion factors as rough approximations as opposed to concrete constants. There is an opportunity to optimize measurements of cell pigment levels by considering scattering and absorption-dependent wavelengths of the OD spectrum.