OD misses low cell concentrations

Optical methods such as OD₆₀₀ offer a convenient approach to frequent tracking of bacterial cultures. Despite these advantages, the sensitivity is relatively low and OD₆₀₀ is often insufficient for tracking changes during lag, deceleration and early-exponential growth stages.

This study aimed to investigate the lower limit of detection (LLOD) and lower limit of quantification (LLOQ) for OD₆₀₀ and compare these with the BactoBox® method.

Take-home messages

• Depending on dilution BactoBox® has at least 50–500× better lower limit of quantification (LLOQ) compared to OD₆₀₀.
• For the OD₆₀₀ method, samples at ≤10⁷ cells/mL are indistinguishable from blank measurements.

The complete investigation and supporting data are presented below.

Study design

We used a stationary stage culture of E. coli (ATCC 25922). Stationary stage was chosen to minimize drift in OD₆₀₀ and BactoBox® measurements across different dilutions. Cells in stationary stage are often relatively small. Therefore LLOD and LLOQ for the OD600 method may be worse than if the culture was investigated in early exponential stage where cells are usually bigger.

• Cultivation was done by transferring a Vitroids™ disc to 50 mL LB medium in a 250 mL baffled Erlenmeyer flask with vented cap. Culture was incubated 12 hours at 37 °C and 200 RPM.
• Cell concentration of the final culture was determined using triplicate BactoBox® measurements.
• The culture was diluted according to the below dilution scheme using dilution with sterile-filtered 1×PBS to a final volume of 10 mL.
• OD₆₀₀ and BactoBox® measurements were done using triplicate technical repeats. The dilution containing 6.5×10⁸ cells/mL was only assessed using OD₆₀₀.
• OD₆₀₀ measurements were done using an Implen™ diluphotometer and standard cuvettes.
• Blank measurements for OD600 were done using 1×PBS. Blank measurements for BactoBox® were done using BactoBox® diluent filtered through a 0.2 µm syringe filter.

BactoBox® quantifies 50–500× lower concentrations than OD₆₀₀

Limit of detection and quantification is assessed by determining the "noise level" of a sample that contains no analytes (a blank sample). The grey bands on the plot represents the "blank band", i.e. the arithmetic mean value of the blank measurements ± standard deviation (σ). The dashed yellow and red lines represent the LLOD and LLOQ, respectively. X-axis represents concentration calculated from the known dilution factor relative to the 12 hour culture.

Lower limit of detection (LLOD) and lower limit of quantification (LLOQ) investigated for OD₆₀₀ and BactoBox®. Each data point represents arithmetic mean ± standard deviation. The dashed red lines represent LLOQ whereas the dashed yellow lines represents LLOD. The lavender line represents BactoBox® lower limit defined by the technical specifications. The concentrations on the x-axis are calculated from the dilution factors relative to the original culture. Prior to BactoBox® measurements, the bacterial stocks were further diluted to hit the right conductivity; the labels on the BactoBox® plot show which dilution was used for the given stock concentrations.

Sensitivity of the OD₆₀₀ method

The OD₆₀₀ measurement for the culture diluted to 1×10⁸ cells/mL is above the LLOD, but it is still not safely above the LLOQ. The OD measurement for the culture diluted to 5×10⁷ cells/mL is in principle beyond the LLOD, but the error bands are hugging the yellow border.

The table summarizes the key metrics. In OD units the LLOQ is 0.123 AU. Converting to a cell concentration requires interpolation. We used the three highest data points from the OD measurements for a linear interpolation in the linear space (not log transformed, results not shown). By this approach LLOQ is ~2×10⁸.

Method	Upper blank band (mean + σ)	LLOD (mean + 3σ)	LLOQ (mean + 10σ)
OD600	0.003 + 0.012 = 0.015	~0.04 AU ~5×107 cells/mL	~0.123 AU ~2×108 cells/mL

Sensitivity of the BactoBox® method

The technical specification for BactoBox® uses a lower limit of detection for the measured sample of 30,000 cells/mL, i.e. 3×10⁴ cells/mL. This is represented by the dashed lavender line in the above figure. The lavender line sits safely beyond the upper blank band, LLOD and LLOQ. This demonstrates that the specified BactoBox® lower limit of detection is reliable.

In practice, to hit the right conductivity, most cultures must be diluted prior to BactoBox® measurements. This affects the practical sensitivity when performing BactoBox® measurements.

• A dilution factor of 1:100 is the typical lowest dilution factor when using standard diluent. In this case the practical lower limit for the BactoBox® method is 3×10⁶ cells/mL. This is ~50× better than the LLOQ of the OD₆₀₀ method.
• A dilution factor of 1:10 is often possible if sterile-filtered low conductivity water is used instead of BactoBox® diluent. This improves practical lower limit for BactoBox® to 3×10⁵ cells/mL. This is ~500× better than the LLOQ of the OD₆₀₀ method.

We compare OD₆₀₀'s statistical LLOQ against BactoBox®'s specified lower limit, because that is the number you work to in practice. The results are summarized in the table below.

Method	Upper blank band (mean + σ)	LLOD (mean + 3σ)	LLOQ (mean + 10σ)
BactoBox® 1:100	6.2×10⁵ cells/mL	9.3×10⁵ cells/mL	2.0×10⁶ cells/mL
BactoBox® 1:10	6.2×10⁴ cells/mL	9.3×10⁴ cells/mL	2.0×10⁵ cells/mL

BactoBox® shows the increase that OD₆₀₀ misses

The figure below represents OD₆₀₀ and BactoBox® result as a function of the theoretical concentration. As previously discussed, for OD₆₀₀ the samples with low bacterial concentrations are indistinguishable from the results of blank measurements on sterile-filtered 1× PBS. In addition the the error bars are very pronounced at low bacterial concentration. In short, OD₆₀₀ is not reliable for cultures containing low concentrations of bacteria.

In contrast, BactoBox® reliably measures concentrations for all investigated samples. The correlation is linear (R² ≈ 1.000) with a slope of 1.00. Furthermore the measurements are precise with narrow error bars throughout the entire measurement range.

OD₆₀₀ and BactoBox® results at different bacterial concentrations. The labels represent stock concentration, dilution-adjusted mean ± σ, and dilution factor in case the stock was diluted prior to measurement. Grey band represents arithmetic mean of blank measurements ± σ. The dashed red line represents lower limit of quantification (LLOQ) whereas the dashed yellow line represents lower limit of detection (LLOD). The theoretical concentrations on the x-axis are calculated from the dilution factors relative to the original culture. OD₆₀₀ measurements were done directly on the bacterial stocks with no dilution prior to measurement.

Implications

This study demonstrates that BactoBox® has significantly better lower limit of quantification (LLOQ) compared to OD₆₀₀. This has practical implications.

BactoBox® is reliable when the culture has low concentrations of bacteria. This results in more reliable growth tracking throughout lag and acceleration growth stages.

The additional measurement points throughout the early-exponential growth stage enables high-confidence determination of growth rates. As an example an initial seeding concentration of 1×10⁶ cells/mL is readily detectable with BactoBox® measurements. With OD₆₀₀ you would have to wait for the cell concentration to get significantly higher than the blank background at which point growth may already begin to be limited by reduced nutrient and oxygen availability. More data at the early part of the exponential stage results in high-confidence determination of growth rates.