Measurement of Cell Concentration in Suspension by Optical Density  

Scott Sutton, Ph.D.
Vectech Pharmaceutical Consultants
This article first appeared in the PMF Newsletter of August, 2006 and is protected by copyright to PMF.  It appears here with permission.

A common issue for the microbiology lab is the determination of starting inoculum concentration.  If the inoculum concentration is determined by plating, the inoculum is several days old before use.  This essay describes the use of turbidity to estimate microbial concentration in a suspension, using the Antimicrobial Efficacy Test as the example.

 Determination of Inoculum for the AET

The compendial antimicrobial efficacy test (AET) requires inoculation of the product with microorganisms to a final concentration of approximately 10⁶ CFU/mL.  Although this seems to be a minor point, it does serve to illustrate some of the inherent difficulties in microbiological testing and the need for experienced and academically trained microbiologists to head the laboratory.

Let’s look at the compendial guidance.  The Pharm Eur (1) instruction on preparing the inoculum for the AET states:

“To harvest the … cultures, use a sterile suspending fluid … Add sufficient  suspending fluid to reduce the microbial count to about 10⁸ micro-organisms per milliliter…Remove immediately a suitable sample from each suspension and determine the number of colony-forming units per milliliter in each suspension by plate count or membrane filtration (2.6.12).  This value serves to determine the inoculum and the baseline to use in the test.  The suspensions shall be used immediately.”

 There are, of course, two problems with these instructions.  The first is that the technician is instructed to use an inoculum of about 10⁸ microorganisms per milliliter and then instructed to determine this by plate count.  Colony forming units (CFU) and cells are two different measures and this will inevitably lead to difficulties as the unfortunate lab worker cannot guarantee the number of cells in the suspension, only the number of CFU found.  However, we can accept the scientific inaccuracy as the numbers will generally work out.  The more serious problem is the instruction to use the plate count CFU for determination of the inoculum for the test, and that the suspension shall be used immediately.  This quite frankly cannot be done.  If you use the suspension immediately, the plate counts are unavailable, if you use the plate counts to set the inoculum, then the suspension is at least a day old.

Contrast these instructions with those in the USP (2) for the same exercise:

“To harvest the … cultures, use sterile saline  … Add sufficient … to obtain a microbial count of about 1 x 10⁸ cfu per mL…[Note: The estimate of inoculum concentration may be performed by turbidimetric measurements for the challenge organisms.  Refrigerate the suspension if it is not used within 2 hours]. 

Determine the number of cfu per mL in each suspension …to confirm the initial cfu per mL estimate.  This value serves to calibrate the size of the inoculum used in the test.”

These USP instructions have the advantage of being physically possible to perform, an obvious advantage to the lab worker.  However, the turbidometric measure of the cells is also only an approximation of CFU.  Thus the instruction to confirm the numbers (after the test is underway) with the plate count is an important control on the test. 

This article will explore the turbidometric approximation for cell numbers, and important controls on the process as well as potential pitfalls to the method.

 Theory

 Light scattering techniques to monitor the concentration of pure cultures have the enormous advantages of being rapid and nondestructive.  However, they do not measure cell numbers nor do they measure CFU.  Light scattering is most closely related to the dry weight of the cells (3). 

Light is passed through the suspension of microorganisms, and all light that is not absorbed is re-radiated.  There is a significant amount of physics involved in this, and those interested are referred to optical treatises, particularly those discussing Huygens’ Principle (a good choice is Light Scattering by Small Particles by H C Van De Hulst).  For our purposes it is enough to say that light passing through a suspension of microorganisms is scattered, and the amount of scatter is an indication of the biomass present in the suspension.  In visible light, this appears “milky” or “cloudy” to the eye (3).  It follows from this that if the concentration of scattering particles becomes high, then multiple scattering events become possible.

Methods

McFarland Turbidity Standards

McFarland standards can be used to visually approximate the concentration of cells in a suspension.  The McFarland Scale represents specific concentrations of CFU/mL and is designed to be used for estimating concentrations of gram negative bacteria such as E. coli.   Note that this estimate becomes uncertain with organisms outside the normal usage as different species of bacteria differ in size and mass, as do yeast and mold.  Use of this method would require calibration and validation.

McFarland Standards are generally labeled 0.5 through 10 and filled with suspensions of Barium salts. (Note - latex bead suspensions are now also available which extend the shelf life of the material).   The standards may be made in the lab by preparing a 1% solution of anhydrous BaCl₂ and a 1% solution of H₂SO₄ – mix them in the proportions listed in the table.  They should be stored in the dark, in a tightly sealed container at 20-25^oC and should be stable for approximately 6 months (4). 

The advantage of the use of these standards is that no incubation time or equipment is needed to estimate bacterial numbers.  The disadvantage is that there is some subjectivity involved in interpreting the turbidity, and that the numbers are valid only for those microorganisms similar to E. coli.  In addition, the values are not in the appropriate range for the AET inoculum and so further dilutions may be required.

 Approximate E. coli concentrations on McFarland Scale

Spectrophotometer

 The spectrophotometer method  measures turbidity directly.  The best case (i.e. most sensitive) would be to have a narrow slit and a small detector so that only the light scattered in the forward direction would be seen by the detector.  This instrument would give larger apparent absorption readings than other instruments.

As should be obvious, each spectrophotometer used must be independently calibrated for use in estimating microbial concentrations.  Not only is the apparent absorption affected by the width of the instrument’s slit, the condition of the filter, and the size and condition of the detector, but also each time the lamp is changed the calibration needs to be repeated as different bulbs may vary in total output.

The correlation of absorption to dry weight is very good for dilute suspensions of bacteria (5), and this relationship seems to hold regardless of cell size (although the relationship of absorption to CFU does not).  However, in more concentrated suspensions this correlation (absorption to dry weight) no longer holds.  The linear range of absorption to estimated CFU is of limited scope and for this reason the calibration study must demonstrate the linear range of the absorbance vs CFU values and the relevant values. 

Procedure

As there are a variety of different instruments, there cannot be one single procedure.  In general, the spectrophotometer can be set at a wavelength of 420 – 660 nm.  This wavelength must be standardized and may need to be adjusted specifically to the material being tested.  Different vegetative cells, bacterial spores and spores of Aspergillus niger may not have the same maximal absorbance wavelength.

It is important to have the cells in known physiological state of growth.  That is to say, as the cell size varies with phase of growth (lag, log, stationery) the approximate relationship between absorbance and CFU will also vary.  A recommended practice might be to pass a single well-isolated colony twice on overnight cultures surface streaks from the refrigerated stock, harvesting the rapidly growing culture from the second passage for preparation of vegetative cells.  This also will serve to minimize a source of variability for the AET (6).

A second source of concern might be the cuvette used for the measurement – care must be taken to maintain the correct orientation of the cuvette, and to protect it from damage that could affect the passage of light.  Finally, it is necessary to blank the spectrophotometer (adjust the absorbance reading to zero) using a standard, either water or the suspending fluid, and maintain this practice.

 Calibration

It must be stressed that this calibration should be done for all organisms.  The size of the organism, any associated pigments, the preparation of the suspension, and other factors all influence the readings.  This calibration study should also be rechecked after changing the bulb on the light source, and should be reevaluated throughout the life of the light bulb.

The calibration itself is simple to perform.  Prepare a concentrated solution of the organism, grown under the conditions that will be used for the test.  Make a series of dilutions to cover the range of absorption measurements of interest; 5 to 8 dilutions are recommended.  Immediately take the spectrophotometer readings in sequence, and then take a confirmatory reading of the first in series to confirm that no growth has occurred.  The dilutions are then immediately plated for viable count (serial dilution of the suspensions will be necessary).   Graph the relationship between the absorbance and the CFU/mL after the plate counts are available and use values in the linear range of this graph.

As there are several factors that can affect this curve (quality of lamp output, size of slit, condition of filter, condition of detector, microorganism characteristic, etc) this calibration should be confirmed when the conditions of the assay change.

Conclusions 

The use of optical density to estimate CFU in a suspension is possible, if basic precautions are taken.  It is important to control:

• The physiological state of the organism

• The species of the organisms (i.e. don’t calibrate the instrument using E. coli and expect the numbers to work for Candida albicans)

• The nature and condition of the equipment

 Despite the inherent inaccuracy of the method, if the procedure is adequately controlled and calibrated the estimation of microbial numbers by optical density (either by McFarland Standards or spectrophotometrically) is sufficiently accurate for use in preparing inocula for QC testing and offers the overwhelming advantages of being rapid, low cost and non-destructive.

References

1. EP. 2006. 5.1.3 Efficacy of Antimicrobial Preservation. Pharm Eur. 5.0:447-449.
2. USP. 2006. <51> Antimicrobial Effectiveness Testing United States Pharmacopeia 29:2499-2500
3. Koch, AL. 1994. “Growth Measurement” IN: Methods for General and Molecular Bacteriology Gerhardt, P et al (ed) American Society for Microbiology, Washington, DC. p. 248-277.
4. Smibert, RM and NR Kreig. 1994 “Phenotypic Characterization” Section 25.4.9 IN: Methods for General and Molecular Bacteriology Gerhardt, P et al (ed) American Society for Microbiology, Washington, DC. p. 607-654.
5. Koch, AL. 1970. Turbidity Measurements of Bacterial Cultures in Some Available Commercial Instruments. Anal Biochem 38:252-259
6. Gilbert, P. et al. 1987. Inocula for Antimicrobial Sensitivity Testing: a Critical Review. J Antimicrob Chemother. 20:147-154.

Consulting with Scott Sutton