Dynamics associated with submerged microbial cultivation: Important considerations when running turbidimetric bioassays

by Ethan Levy Ph.D.   

Micra Biotechnologies Inc.

Turbidimetric assays as published in the CFR-21 during the early 1970s remain in use to the present day for qualifying potencies  of antibiotic drugs headed for release into the public domain. Over the years, many misconceptions have developed with regard to methodology and how much over simplification can be effected. As a consequence, biased antibiotic potencies are generated among certain antimicrobial agents with results which may be contributing to  todays trends in AMR. In order to address this dilemma, I would like to review some basic concepts and misconceptions associated with turbidimetric bioassays 

including the absolute requirement for an antibiotic dose response which is proportional to the observed optical response.  Despite all of the pitfalls  associated with photometric assays linearity in dose response is the  minimal criteria needed to assure validity of potency results.

Turbidimetric bioassays when compared to plate or diffusion assays, follow different paths in their dose response relationships. With turbidimetric assays detection of inhibition relies on several important  developments which occur during incubation. These include changes in culture optical properties resulting from increases in cell size and number. During submerged cultivation, due to the presence of multiple asynchronously growing subpopulations (Figure 1.0) culture optical properties may affect  standard and sample absorbance measurements to different extents.  During incubation, due to the diversity of sub populations which are interacting  with a given antimicrobial, test antibiotics will exert their action only on strains  which are in a particular susceptible phase in their growth cycle. 

Different antibiotics have different mode of actions acting at different phases  of the organism growth cycle. Thus at any given time, only a fraction of the microbial population will respond to the test antimicrobial with consequential kill and lysis.  The remainder of cells will continue to grow variably at their respective rates.  Some strains may be at the same lag phase as those in the zero antibiotic control tubes.  Additionally, consequential to the variable rates in the elongation and division cycles, physical factors including light scatter (reflection and refraction), birefringence, cell clumping and opacity will further contribute to the lack of proportionality in the drug/organism dose response, considering the development  of  absorbance over time. Younger cells occurring during lag phase, are larger and more varied in size than older cells. Due to the initial notable changes in cell size, photometric measurements become more affected by cell mass and not cell number. It is therefore important to recognize that absorbance is non linear with respect to cell concentration which can potentially result in biased and inaccurate potency values.

The above attributes,  while in theory may also affect plate or diffusion assays, are generally not considered to interfere with bioassays on performed solid  media. With plate assays the prerequisite for zone formation is determined by the development of the critical cell mass. Additionally plate assays,  in most cases are scored following an overnight incubation while turbidimetric assays are read within the same day. By comparison photometric assays require  a much shorter incubation period and optimal differential turbidities are achieved

i.e., across all the standard dose response points (A-E), 3-4 within hours.

Recognizing the dynamic considerations associated with photometric bio-assays, it is important that standard dose response concentrations be comprised in ratios of small increments so that the overall optical response falls within the linear range of the dose response curve (Figure 2.0). Assay results are only applicable when the measured response is linearly proportional with the dosages of the active principle. Any departure from the linearity at the upper or lower extremes of the standard curve would invalidate  the assay. Another way of looking at this is,  the greater and steeper the dose response slope, the shorter the potency range of the assay. Doses which follow small geometric progressions e.g., dose ratios of 1.14:1 to 1.40:1, where points do not differ by increments of more than 10-15 % in response should be adhered to.

By way of example, a straight line relationship may extend from 0.01-0.025 units/ml for penicillin G when plotted against S. aureus i.e., when the antibiotic response is scored relative to the non antibiotic containing control. Departure from the narrow, straight-line dose response region (Figure 2.0 C-D) towards the extremities in which there is either no or minimal growth (Figure 2.0 A-B)  to the extremity where close to the optimal measurable

turbidity has been reached (Figure 1. E-F) will result in grossly inaccurate assays. Regions where measurements of absorbance no longer share the same response rate over time are inaccurate. In essence performing turbidimetric runs with wide dose response increments 

will likely preclude whole sub populations of organisms. Hence the 0.64, 0.80, 1.00, 1.25 and 1.56 mcg/ml dose formats typically utilized for plate assays are not suited for turbidimetric assay.