1. Introduction
Bacterial biofilms are structures consisting of extracellular polymeric substances and single or multiple species of bacteria that form clusters and adhere to surfaces [
1,
2]. A major proportion of the biofilm, up to 97%, consists of bound water [
3]. Compared to planktonic bacteria, bacteria in biofilms better survive harsh environmental conditions, and consequently biofilm bacteria show a 10–1,000 times higher resistance towards antibiotics than planktonic bacteria [
4]. This has become a major healthcare concern as antibiotic-resistant bacterial strains like methicillin-resistant
Staphylococcus aureus (MRSA) are appearing. Multiple factors contribute to this resistance, including physical and chemical diffusion barriers [
5], reduced sensitivity towards antibiotics due to the slow growth rate of bacteria in biofilms [
6,
7,
8], as well as the structural heterogeneity within biofilms and development of biofilm-specific biocides-resistant bacteria phenotypes [
4,
9]. Furthermore, biofilms are often associated with implants or other biomaterials as such biocompatible materials also provide an ideal substrate for biofilm formation. Understanding biofilms and the antibacterial methods for their removal or inhibition is thus critical in the development of functional biomaterials.
To be able to determine the microbial antibiotic susceptibility, or assess the effectiveness of other bactericidal treatments on bacteria in biofilm, it is essential to be able to determine the amount of viable bacteria in the biofilm, or at least the relative reduction in viable bacteria due to the treatment. There are several means for quantification of bacterial viability, but these are generally much better suited for evaluating planktonic cultures. Unfortunately, one cannot transfer results from tests performed on planktonic bacteria to biofilm bacteria since it is known that, for example, bacteria in biofilm have an inherent lack of susceptibility to antibiotics compared to planktonic cultures of the same bacteria. The classical means of determining bacterial viability is counting the number of colony forming units (cfu) after plating cultures. Using this method to assess biofilm viability can lead to significant errors, since there is a high degree of aggregation due to the presence of the biofilm matrix. Procedures such as sonication or the use of matrix degrading enzymes can be used to separate bacteria from the matrix or the surface to which they are attached, but have the disadvantage that the viability of the bacteria may be affected. Additionally, the bacteria cells must be re-suspended in order to perform the cfu counting. The Calgary Biofilm Device [
10] is a popular method that utilizes such procedures in determining the microbial antibiotic susceptibility of bacteria in biofilm, but does not detect the number of bacteria in the biofilm. Staining techniques such as crystal violet or live/dead staining [
11] provide valuable information, but also have inherent limitations. Crystal violet provides a good measure of biofilm mass; however, it does not give a measure of biofilm viability as it stains both the bacteria cells and the extracellular matrix. Live/dead staining is used to quantify planktonic bacteria in combination with, for example, flow cytometry [
11], and is useful for imaging biofilms, but is not ideal for high-throughput quantification of biofilm viability as it must be used in conjunction with a laser scanning confocal microscope in a time consuming process where only a small section of the biofilm can be assessed at a time.
Metabolic assays are excellent candidates for quantification of bacterial viability in biofilm. These assays are indirect methods based on the detection of metabolic products produced by bacteria and have the advantage of being able to assess viability without sample manipulation since these assays generally do not require the removal of the biofilm from the adherent surface. A number of different indicators are used for the detection of bacterial metabolic activity. For example, the resazurin assay (also known as the Alamar Blue assay) is based on the reduction of resazurin, a blue dye that can be reduced by metabolically active cells to pink resorufin, which is fluorescent [
12]. Thus, fluorescent measurements of a resazurin assay containing a biofilm can be used to quantify the viability of the biofilm [
13,
14]. Similarly, an assay based on the conversion of non-fluorescent fluorescein diacetate (FDA) into the highly fluorescent fluorescein has been used for the quantification of biofilm mass [
15,
16] and viability [
17]. Another strategy is the use of a pH indicator to measure the change in pH of a viability assay due to the production of acids by bacterial metabolism [
18,
19]. Part of the reason for the existence of the variety of metabolic assays is the fact that often the assays are strain specific or work better with certain types of bacteria.
However, a significant limitation of these metabolic assays that has not been adequately addressed is the fact that bacteria in biofilm do not have the same metabolic activity as planktonic bacteria. In order to correlate the metabolic signal of the assay containing the biofilm in question, one must know how many viable bacteria in such a biofilm produce a given signal. Typically, assays performed on biofilm are calibrated against their planktonic cultures, with the assumption that the metabolic rates are similar. However, this introduces significant errors since it is known that the metabolic rates of bacteria can differ greatly between their planktonic and biofilm form [
4,
20,
21]. For example, it has been shown that when using the resazurin metabolic assay,
Staphylococcus aureus planktonic bacterial suspensions generate a larger signal than that from the same concentration of corresponding biofilm bacteria [
22].
We present a scheme for quantitative determination of biofilm viability and show that this information could be used to determine, for example, the effect of antibacterial treatments or the performance of a bactericidal implant surface [
23,
24,
25]. The specific growth rate of biofilm bacteria is determined by a series of metabolic assays, and this information is used to calibrate the results of biofilm bacteria viability testing. We apply the method for quantification of a
Streptococcus mutans biofilm.
2. Theory
A metabolic assay is used to assess the number of viable bacteria in the assay media through the production of metabolites that are subsequently detected or measured. Normally the relationship between the starting bacterial population and the signal level (or time to achieve a certain signal level) in an assay is empirically determined by producing a standard curve from a series of bacteria cultures with different starting concentrations. This poses no difficulty with planktonic bacterial cultures since it is easy to determine the starting concentration; however, this is not the case with bacteria in biofilm form since it is practically impossible to know the starting number of viable bacteria in a biofilm.
The amount of metabolite produced by a biofilm, or a planktonic culture, in a viability assay will depend on both the metabolic activity of the individual bacteria and the number of bacteria. If the assay indicator (e.g., color of the pH indicator or the amount of fluorescence) is measured shortly after the biofilm is placed in the assay, the measurement corresponds to the initial number of viable bacteria in the biofilm since the bacteria have not had time to multiply. However, signal levels may be too small to be detected reliably if the bacteria population is relatively small. The detection range of the method can be greatly extended by allowing the culture to multiply until the population of bacteria is large enough to produce a measureable signal. In this case, the growth rate of the bacteria plays an important part in the measured signal. An example of this type of viability assay is one incorporating phenol red, a pH indicator that changes the color of the assay media when the bacteria have produced enough acid metabolites.
An analytic estimation of the time for a bacterial culture to produce a certain signal level can be determined by the following analysis. In this analysis, we consider the assay to contain an excess of sustenance that allows the bacteria to exist in the exponential growth phase. In this case, we can represent the number of bacteria at time
t by the Equation (1)
![Jfb 03 00418 i001]()
Where
![Jfb 03 00418 i002]()
is the initial number of bacteria,
![Jfb 03 00418 i003]()
is the specific growth rate, and the time for the bacterial population to double (
i.e., the generation time) is given by
![Jfb 03 00418 i004]()
.
If
![Jfb 03 00418 i005]()
is the amount of metabolite produced per bacterium per unit of time, then
![Jfb 03 00418 i006]()
is the amount of metabolite produced in the time interval
dt at time
t. In order to calculate the accumulated amount of metabolite at time
t we integrate
dm:
![Jfb 03 00418 i007]()
where
v is the integration variable. If the
![Jfb 03 00418 i008]()
(
i.e., the bacterial population is constant), then Equation (2) simplifies to
![Jfb 03 00418 i009]()
as expected (this can be proved by taking the Maclaurin series of
![Jfb 03 00418 i010]()
before setting
![Jfb 03 00418 i008]()
in Equation (2)).
If a certain amount of metabolite is required to produce a signal (e.g., a specific amount of metabolite is required to change the color of the phenol red assay to yellow) we can write the relationship between the time,
![Jfb 03 00418 i011]()
, it takes a starting amount of bacteria,
![Jfb 03 00418 i012]()
, to produce the required amount of metabolite,
![Jfb 03 00418 i013]()
:
![Jfb 03 00418 i014]()
solving for
![Jfb 03 00418 i012]()
gives
![Jfb 03 00418 i015]()
for
![Jfb 03 00418 i016]()
,
Equation (5) can also be expressed in terms of
![Jfb 03 00418 i011]()
:
![Jfb 03 00418 i018]()
Thus, if we plot
![Jfb 03 00418 i011]()
as a function of
![Jfb 03 00418 i019]()
we find that the slope of the line corresponds to the inverse of the specific growth rate of the bacteria. This plot corresponds to the standard curves that are used to calibrate metabolic viability assays; different known starting concentrations of bacteria are placed in the assay and the times for the required amount of metabolite to be produced are recorded and plotted against the starting concentrations of bacteria. In fact, if we look at standard curves that have been produced in other studies with viability assays incorporating indicators such as phenol red or resazurin [
14,
19], we can observe the logarithmic relationship between time and initial bacterial concentration as indicated by Equation (6).
The fact that we show that the slope of the line in the standard curve is directly related to the specific growth rate suggests that these viability assays can be calibrated by other means that provide the specific growth rate of the bacteria in the assay media. With only knowledge of the bacterial growth rate, we cannot correlate the result of a viability assay to an exact number of viable bacteria; however, we can use it to quantify the relative number of viable bacteria between different samples. This would be of great use in determining the log reduction in antibiotic susceptibility testing or other types of antimicrobial treatments.
For planktonic bacterial suspensions one could use turbidity (optical density, OD) measurements to determine the specific growth rate, since this is a standard method of determining the specific growth rate of bacteria cultures. One could argue, however, that this is unnecessary since a standard curve can be readily prepared with know bacterial concentrations when dealing with planktonic bacteria. However, this is not the case with biofilm since it is practically impossible to know the starting number of viable bacteria in a biofilm. Instead, one could perform a series of assays with, for example, a varying number of sample disks covered with biofilm. Plotting this relative number of biofilm bacteria versus the time required to produce the required signal would allow one to calculate the specific growth rate using Equation (6), and thus be able to quantitatively determine the log reduction in, for example, antibiotic susceptibility tests.
If the above technique using a varying number of sample surfaces covered with biofilm is not possible or practical, another means of determining the specific growth rate is required. Returning to Equation (3), we can see that for a fixed initial number of bacteria
![Jfb 03 00418 i012]()
there is an exponential relationship between the amount of metabolite required to produce a signal and the corresponding time
![Jfb 03 00418 i011]()
for values of
![Jfb 03 00418 i020]()
. Thus, if a series of assays containing the same initial amount of biofilm
![Jfb 03 00418 i012]()
are performed using different volumes (
i.e., varying
![Jfb 03 00418 i013]()
), a plot of
![Jfb 03 00418 i011]()
versus
![Jfb 03 00418 i021]()
can be made and the specific growth rate can be extracted from the slope of the line for values of
![Jfb 03 00418 i020]()
:
![Jfb 03 00418 i022]()
where the slope = ln(10)/
μ. Once the specific growth rate is known, this information can be used to calibrate the results from antibacterial testing on the biofilm and thus quantitatively determine the log reduction. A schematic overview of the method alternative in which the volume of the assay media is varied is displayed in
Figure 1.
Figure 1.
Schematic diagram of the method for determining the specific growth rate of biofilm bacteria. The specific growth rate μ is extracted from a plot of the time required for color switch of Metabolic Activity Assay (MAA) broth containing a set amount of biofilm bacteria as a function of MAA broth volume. Included is a laser scanning confocal microscopy image of a S. mutans biofilm using live/dead staining (L13152, Invitrogen, Eugene, UR, USA).
Figure 1.
Schematic diagram of the method for determining the specific growth rate of biofilm bacteria. The specific growth rate μ is extracted from a plot of the time required for color switch of Metabolic Activity Assay (MAA) broth containing a set amount of biofilm bacteria as a function of MAA broth volume. Included is a laser scanning confocal microscopy image of a S. mutans biofilm using live/dead staining (L13152, Invitrogen, Eugene, UR, USA).
3. Experimental Section
3.2. Specific Growth Rate of Planktonic S. mutans by OD Measurements
Planktonic S. mutans bacteria cultured to exponential phase in BHI broth were inoculated to 12 mL of BHIS broth and cultured statically at 37 °C. Every 30 min the culture was mixed well and a 300 μL aliquot was removed and turbidity measurements to assess the bacterial concentration were made using a multimode microplate reader (Infinite M200, Tecan) in absorbance mode at 600 nm. The specific growth rate of planktonic S. mutans in BHIS broth was calculated from these OD600 measurements as a function of time based on Equation (1).
3.4. Specific Growth Rate of Planktonic S. mutans by MAA Measurements
An MAA standard curve for planktonic S. mutans was determined by measuring the time for color change as a function of initial S. mutans concentration. Fourteen wells containing 2 mL of MAA broth were inoculated with S. mutans at bacterial concentrations ranging from 10 to 108 cfu/mL. The sensitivity of the MAA, which is about 10 cfu/mL, is limited by deviations in diluting very small populations of cells. Incubation was performed at 37 °C without shaking, and color of the MAA broth was checked every 20 min.
The initial S. mutans concentrations used for the MAA standard curve were set according to OD, which had been correlated to concentration by cfu counting. To ensure accurate correlation between OD and concentration, a 10 ml sample of the media containing bacteria was sonicated for 3 s at 20% intensity in order to separate the bacteria (Vibra-cell CV33 with tapered microtip 6300418, Sonics & Materials, Inc., Newtown, PA, USA) before culturing on an agar plate for cfu counting. Due to the possibility that the ultrasound treatment killed a small proportion of the cells, a comparison of the number of live cells before and after the ultrasonic treatment was made using a live/dead bacterial viability assay (L13152, Invitrogen, Eugene, OR, USA). It was found that 3% were killed due to the sonication, and thus a correction factor was applied to the OD-cfu correlation.
Additionally, a series of ten MAA were performed on planktonic
S. mutans (initial number of bacteria
![Jfb 03 00418 i012]()
= 5 × 10
5 cfu for all assays) in assay volumes ranging from 3 to 200 mL.
3.5. Standard Curves of Biofilm S. Mutans by MAA Measurements
A series of 6 MAA was performed on S. mutans in biofilm form in which the initial number of bacteria was controlled by varying the number of disks cultured with S. mutans biofilm from 1 to 32 disks in an assay volume of 10 mL. Finally, a series of 8 MAA was performed in assay volumes ranging from 1 to 50 ml in which each assay contained a single disk cultured with S. mutans biofilm.