• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Expeditious identification and quantification of mycobacteria species in metalworking fluids using peptide nucleic acid probes

Posted on: Saturday, 22 November 2003, 06:00 CST

Abstract

Microbial contamination of metalworking fluids (MWFs), particularly by mycobacteria species, poses a potential health and safety risk for machine tool workers. This hazard can be magnified by undiagnosed or long-term exposure to mycobacteria that is created by the extensive time required to detect mycobacteria in MWF systems (10 to 26 days). This research investigates the use of molecular technology to reduce the mycobacteria detection time in MWFs to within one day. Novel fluorescent peptide nucleic acid (PNA) probes, designed to target the specific 16S rRNA of Mycobacterium parafortuitum, are applied as a means to demonstrate the effectiveness of the approach in MWF The specific hybridization of the PNA probes is detected using epi-fluorescence microscopy and flow cytometry (FCM), and the performance of two PNA probe classes (traditional and beacon) is compared in this application. It is shown that both epi-fluorescence microscopy and FCM can be utilized in conjunction with the PNA probes to specifically identify M. parafortuitum among a mixed microbial community in MWF. It is also shown that the M. parafortuitum labeling methods are insensitive to the synthetic, semi-synthetic, and soluble oil MWF chemistries investigated. In addition, the use of FCM for quantification of M. parafortuitum is demonstrated, and important considerations for applying FCM in the field are described.

Keywords: Mycobacteria, Metalworking Fluids, Peptide Nucleic Acid (PNA), Flow Cytometry, Microbial Detection

1. Introduction

Microbial contamination of metalworking fluids (MWFs) has traditionally been considered a primary source of physical and chemical deterioration that results in quality loss, nuisance odors, and eventual MWF disposal to the environment. In recent years, it has been recognized that significant health and safety risks exist for workers exposed to biologically contaminated MWFs (NIOSH 1998; Kreiss et al. 1997; Rosenman et al. 1997; Rose et al. 1996; Bernstein et al. 1995; Byers 1994; Mattsby-Baltzer et al. 1989; Robertson et al. 1988; Zugerman 1986). These risks include: (1) microbial infection; (2) endotoxin exposure; (3) asthma; (4) dermatitis; and (5) hypersensitivity pneumonias (HP). Each of these hazards is summarized below:

Microbial Infection. MWFs can contain pathogenic or potentially pathogenic microorganisms such as Escherichia coli, Legionella sp., and Mycobacteria sp. Major infectious outbreaks due to MWF bacteria have been reported in the literature, such as the outbreak of illness due to Legionella sp. in an engine plant that affected approximately 200 workers (NIOSH 1998).

Endotoxins. Endotoxins are known to cause chronic bronchitis, asthma, or other general declines in respiratory function. Endotoxins are structural cell wall elements of the most common MWF bacteria (gram negative), and very high concentrations of endotoxins have been occasionally detected in MWFs (NIOSH 1998; Byers 1994; Mattsby-Baltzer et al. 1989).

Asthma. Asthma is a general reduction in airflow through the pulmonary system and is much more prevalent in individuals exposed to MWFs than in the general population (Rosenman et al. 1997). Allergic responses to microbial matter have been isolated as a cause of MWF-induced asthma (NIOSH 1998; Robertson et al. 1988).

Dermatitis. Dermatitis is a general condition referring to an inflammation of the skin. Excessive biocide concentrations, such as those commonly added to MWFs to control microbial "overgrowth," have been linked to allergic contact dermatitis (Byers 1994; Zugerman 1986).

HP. Hypersensitivity pneumonitis (HP, or extrinsic allergic alveolitis) is an allergic disease that can be characterized by pulmonary (alveolar) inflammation, recurrent dyspnea, cough, influenza-like symptoms, and possible long-term fibrosis. HP develops due to repeated exposure and subsequent sensitization to various antigens, including microbial antigens that may arise from bacterial or fungal sources. The disease (or more accurately, the group of diseases) is thought to involve lymphocyte sensitization and a cell-mediated immune response that results in alveolitis (that is, inflammation of the lung alveoli) (Shelton et al. 1999).

Numerous cases of HP have been reportedly linked to MWF exposure (NIOSH 1998; Rose et al. 1996; Bernstein et al. 1995; Muilenberg and Burge 1993; Meredith et al. 1991). Data reported by Kreiss and Cox- Ganser (1997) suggested a correlation between MWF contaminated with fungi and mycobacteria, and the occurrence of HP. The association of HP with MWF is perhaps unsurprising, as it is well known that MWF mists can contain high concentrations of bacterial and fungal spores that have been traditionally associated with HP. Limited data on the subject have suggested that fungi and mycobacteria tend to be present with a higher probability in MWFs that have been linked to the development of HP when compared with MWFs that have not been known to cause HP (Kreiss et al. 1997).

Shelton et al. (1999) provided evidence supporting the hypothesis that aerosolized mycobacteria is a potential cause of MWF-induced HP, which is plausible given the highly antigenic cell wall of mycobacteria. While the fundamental causes for mycobacteria outbreaks in MWF are currently uncertain, it has been hypothesized that shifts in MWF chemistry over the past 20 years (for example, the increased use of synthetic and semi-synthetic MWFs) may be contributing to an increased prevalence of gram-positive bacteria, including mycobacteria. It has also been hypothesized that mycobacteria can proliferate under biocide conditions that are normally detrimental to common metalworking fluid microorganisms (Passman et al. 2002). This is in part because mycobacteria have a highly hydrophobic cell wall composed of mycolic acid that permits the species to withstand adverse physical conditions (such as heat, acidity, salinity), including the application of phenol (Madigan et al. 1999). Due to the prevalence of mycobacteria in MWF systems, and recent discoveries of mycobacteria (e.g., Mycobacterium immunogenum) in MWFs suspected of causing HP (Kreiss et al. 1997; Muilenberg et al. 1993), interest in developing detection methodologies for mycobacteria in MWF has increased significantly. Specifically, there has been a call for rapid mycobacteria detection methods that can be applied within a manufacturing facility (Passman et al. 2002).

As reviewed by Watterson (2000) and Drobiewski et al. (2000), current methods for the detection of mycobacteria include microscopic examination of acid-fast bacilli staining, culturing systems (for example, solid, liquid, radiometric, and nonradiometric), and biochemical analyses (e.g., HPLC, TLC, polymerase chain reaction, nucleic acid sequencing). Some of these methods have been developed into commercial kits for the diagnosis of mycobacteria infection and can offer species-level detection in medical samples: e.g., AccuProbe(R) from Gen-Probe, BACTEC MGIT Systems from Becton Dickinson, and INNO LiPA Mycobacteria from Innogenetics (Hanna et al. 1999; Tortoli et al. 1999; Scarparo et al. 2001). The major disadvantages of these methods include lengthy turnaround time (10-26 days), potential species bias, and large equipment cost. Outsourcing samples for culture-based detection does not address these issues, can be extremely costly (for example, more than $100 per sample), and only slows the detection cycle time. This further reduces the ability to determine cause and effect of microbial outbreaks. Moreover, more modern biochemical assays based on PCR and fluorescent anti- body labeling, while faster than culturing techniques, currently lack sufficient data to support their application to complex fluid matrices such as MWF.

The one to three week lag time associated with currently accepted (culture-based) methods used to detect and quantify mycobacteria in MWF is a major hindrance to controlling infection and eliminating worker exposure. This has prompted investigations into new molecular- based approaches that target the rRNA genome of mycobacteria to rapidly identify different species under "environmental" conditions (e.g., outside of traditional medical applications). In the last decade, many 16S and 23S RNA-targeted DNA probes have been designed to detect various phylogenetic subgroups or species of mycobacteria. More recently, PNA (peptide nucleic acid) probes have been developed for the same purpose (Abe et al. 1999; Drobniewski et al. 2000; Stender et al. 1999), and for the detection of Mycobacterium tuberculosis genes related to antibiotic resistance (Bockstahler et al. 2002). While such molecular probe methods appear to be promising for application to MWF, their effectiveness under the physicochemical conditions of MWFs (for example, ionic strength, optical transmission, nonspecific binding, pH, and so on) is currently unknown.

This paper examines the use of PNA probes to more rapidly identify and quantify mycobacteria in MWFs via fluorescence in situ hybridization (FISH) targeting rRNA sequence composition. rRNA is a convenient target for fluorescent labeling using molecular probes because it is found in all microorganisms, and a typical single cell may contain thousands of copies of rRNA. Due to the sensitivity of rRNA- \targeted probes for identification and quantification of individual species, they have gained much interest and use as a tool for field microbiology studies (Fricker 2000). However, apart from Van der Gast et al. (2001), who demonstrated that rRNA-targeted DNA- based probes coupled with epi-fluorescence microscopy could be used to differentiate various species (e.g., Pseudomonas aeruginosa and Escherichia coli) in MWF, research of rRNA targets in MWF has been scarce in the literature.

This research was performed to accomplish five major objectives: (1) design a novel PNA-based probe targeting a specific region of the 16S rRNA of a representative mycobacteria species (Mycobacterium parafortuitum) and to optimize a corresponding hybridization method; (2) investigate the efficacy of the PNA probe for detecting M. parafortuitum using epifluorescence microscopy and flow cytometry; (3) compare molecular beacon versus traditional PNA probes for utility in this application; (4) quantify the recovery and specificity of the 16S rRNA probes for M. parafortuitum detection in MWF; and (5) determine whether MWF and its contaminants (e.g., particulate, oil, and mixed microbial communities) interfere with the PNA hybridization and detection method.

2. Approach: Fluorescence In Situ Hybridization, PNA, and Flow Cytometry

Fluorescence In Situ Hybridization (FISH) of rRNA-Targeted Molecular Probes

Biomolecular probes are commonly used to distinguish microbial species on the basis of differences in their ribosomal RNA (rRNA) genome. rRNA is found in all microorganisms, and typically a single cell may contain thousands of copies of rRNA, which makes rRNA a convenient target for fluorescent labeling using molecular probes (Stender et al. 2002). rRNA-targeted molecular probes have at least two basic parts: a sequence of base pairs that can bind to complementary regions of the rRNA, and an attached fluorescent "reporter" molecule. A typical fluorescence analysis using rRNA- targeted molecular probes involves four steps: (1) fixation and permeablization of the microbial cell to facilitate transport of the fluorescent probe into the cell; (2) binding (or "hybridization") of the complementary probe sequence to its rRNA targets; (3) washing of the sample and cells to remove unbound fluorescent probe that would interfere with the signal during measurement; and (4) optical detection using an epifluorescence microscope or flow cytometry (FCM).

Due to the sensitivity of rRNA-targeted molecular probes for identification and quantification of specific microbial species, they are currently gaining interest as a tool for field microbiology studies (Stender et al. 2000; Fricker et al. 2000; Gunasekera et al. 2000). Van der Gast et al. (2001) demonstrated that rRNA-targeted DNA-based molecular probes coupled with epi-fluorescence microscopy can be used to differentiate various species (for example, P. aeruginosa and E. coli) in MWF. This paper extends the previous MWF research by investigating: (1) the determination of microbial concentrations for specific species in MWF, (2) the consideration of mycobacteria targets; (3) the use of flow cytometry to detect fluorescence from rRNA targets in MWF, (4) the use of PNA-based probes and molecular beacons, and (5) the consideration of different MWF chemistries.

PNA Probes

As shown in Figure 1, peptide nucleic acid (PNA) probes are DNA mimics comprised of a synthetic, neutral polyamide backbone, which is substituted for the negatively charged phosphate-sugar backbone of DNA. Due to their noncharged backbone and relatively hydrophobic nature, PNA probes can more easily cross the hydrophobic cell membrane of mycobacteria, which expedites PNA protocols relative to traditional DNA probe protocols (Thisted et al. 1999). Also due to the neutrality of PNA, its binding properties to rRNA targets inside the cell are nearly independent of salt concentration, and the probes exhibit high stability under variations in temperature and pH. These characteristics make PNA particularly well suited for MWF applications. Nevertheless, there is no reported application of PNA to the identification and quantification of microbial species in MWFs prior to this investigation.

Figure 1

Peptide Nucleic Acid (PNA) Structure

Molecular Beacon PNA Probes

Molecular beacon (MB) PNA probes are distinct from traditional PNA probes in that they only exhibit visible fluorescence when hybridized to their target (e.g., rRNA). As shown in Figure 2, when MB PNA base pairs bind to their rRNA target sequence, the close association between the fluorescent reporter molecule and quencher molecule is broken, and the reporter molecule is then detectable using techniques such as epi-fluorescence microscopy and FCM. Due to the insensitivity of the detection method to unbound probes (caused by the close proximity of the reporter and quencher molecules in the nonhybridized state), MB PNA probes have the potential to detect low concentrations of microorganisms with higher signal-to-noise (S/N) ratio than traditional PNA probes.

Flow Cytometry (FCM)

Flow cytometry is the measurement of physicochemical characteristics of cells as they flow through an observation cell. Unlike epi-fluorescence microscopy, a flow cytometer aims to observe a single cell at a time rather than the whole sample at once. A multitude of cellular properties can therefore be measured using flow cytometry, including size, shape, species, DNA/RNA content, and protein content. Therefore, FCM has been widely utilized for biomedical, biotechnology, and environmental microbiology research (Shapiro 1995; Porter et al. 1997; De Boer et al. 1999; Davey et al. 1999). Although cellular quantification is the most basic function of a flow cytometer, the technology has not previously been applied to the detection of bacteria and fungi found in MWFs. In fact, nonclinical applications of FCM are very rare. While this is in large measure due to the costs of FCM systems (typical units exceed $100,000), a significant amount of MEMS-based research is currently directed toward lowering this cost by two or more orders of magnitude (Huh et al. 2001; Miyake et al. 1997). Such efforts suggest that FCM could become economically feasible for manufacturing applications in the foreseeable future.

Figure 2

FISH Method Targeting Mycobacterium parafortuitum Using PNA Molecular Beacon

Figure 3 illustrates microbial detection using optical flow cytometry. The process aims to develop a single-file flow of microbes inside a detection flow cell. To achieve this, a microbial sample is injected into a coaxial stream of laminar flow sheath fluid (e.g., particle free water). The mixture is hydrodynamically focused to maintain laminar flow and separation as the cells pass through the observation point. A laser beam is focused at the observation point, and the resulting forward scatter (0.5- 10[degrees]) and side scatter (90[degrees]) signals are detected by photovoltaic photodiodes. If the microbial sample has been labeled using PNA or other bio-specific fluorescent reporters, the laser induces a fluorescence that is separated from scattered radiation and detected by a photomultiplier tube. Consequently, the flow cytometer can distinguish between MWF constituents and microbial cells. The frequency, duration, and intensity of the fluorescent and scatter signals are related to microbial concentration, size, and constitution.

Figure 3

Flow Cytometry Concept

Table 1

rRNA Probes and Sequences

Probes custom synthesized by Applied Biosystems (Foster City, CA)

3. PNA Detection Methodology for Mycobacteria in MWF

Culture Preparation

ATCC (Manassas, VA) cultures of M. parafortuitum (#19686), R aeruginosa (#8707), and E. coli (#11303) were maintained in pure culture under constant log growth conditions with cell densities measured by direct counts and optical density. M. parafortuitum was maintained on Middlebrook 7H9 media, while P. aeruginosa and E. coli were maintained on Luria Burtani media. M. parafortuitum was selected as a representative mycobacteria species due to its genetic and structural similarity to species such as M. immunogenum that have been suspected to be possible causative agents for HP occurrence.

16S rRNA Probes

Table 1 provides the nucleotide sequence information for the novel M. parafortuitum targeted PNA probes developed for this research. The M. parafortuitum probe targets a region of the 16S rRNA sub-unit previously found to be distinctive for mycobacteria species in Stender et al. (1999). Table 2 shows the 16S rRNA base sequences for a subset of the microorganisms that were used to develop the probe. A considerable amount of genetic distinction can be observed in this region of the 16S genome of M. parafortuitum, even when compared with other mycobacteria species such as M. tuberculosis (20% distinct), M. chelonae (7% distinct), and M. bovis (20% distinct). The degree of heterogenicity was also calculated for common MWF microorganisms such as Pseudomonas aeruginosa (13% distinct), Bacillus subtilis (27% distinct), and Escherichia coli (40% distinct). It should be noted that the nucleic acid sequence used to detect M. parafortuitum is 100% identical to M. fortuitum, which is a pathogenic mycobacteria species not investigated here. It is therefore expected that the same probes and procedures discussed here would yield similar results for M. fortuitum. If desired, it is readily possible to develop a probe to distinguish between M. parafortuitum and M. fortuitum using a similar 16S rRNA detection approach to that investigated here.

Table 2

16s rRNA Sequences Alignments Performed Using CLUSTALW (v1.81) Software

All probes used in this research were dissolved in 0.1 % trifluoroacetic acid and frozen at -20[degrees]C in 200 [mu]L aliquots (100 mM) for no longer than three months before use. The probes were synthesized with a lysine linker (to facilitate solubilization) and a fluor\escein fluorescent marker. DABCYL 4-(4'- dimethylaminophenylazo)benzoic acid, a non- fluorescent chromophore, served as the quenching molecule in MB 1227.

Hybridization Procedure for epi'Fluorescent Microscopy Detection (epiFISH)

The epiFlSH method was modified from Trusted et al. (1999) and was executed as follows. 25 [mu]L of cells were smeared onto a glass slide, dried, flame fixed, and immersed in ethanol. The cells were dried and treated with 4% paraformaldehyde and then preincubated for 30 minutes in a hybridization solution (10% w/v dextran sulfate, 30% v/v formamide, 0.1% w/v sodium pyrophosphate, 0.2% w/v polyvinyl- pyrrolidone, 0.2% w/v ficoll, 1 mM disodium EDTA, 0.1% v/v Triton X- 100, 10 mM tris-HCl at pH 7.5). 10 [mu] of PNA probe (50 nM) was added to the slide, incubated in a rotating hybridization oven at 56[degrees]C for two hours, and then immersed for 30 minutes in wash solution (5 mM Tris, 15 mM NaCl, 0.1% (v/v) Triton X-The labeled cells were viewed using a Zeiss D-7082 epi-fluorescent microscope.

Hybridization Procedure for FCM Detection fcmFISH)

The primary difference between epiFISH and fcmFISH is that in epiFISH the microbial cells were initially fixed to a glass slide as necessary for microscopic observation. fcmFISH requires that the cells are ultimately suspended in aqueous matrix, leading to significant differences relative to epiFISH. For fcmFISH, the hybridization method developed to detect marine Cyanobacteria (Worden et al. 2000) was modified and executed as follows. 2 mL of cells were collected via centrifugation (13,000x g for 10 minutes) and resuspended in 2 mL phosphate buffer saline (PBS) to wash the cells of growth medium. The cells were then collected (13,000x g for 10 minutes) and resuspended in 500 [mu]L of ethanol to kill the organisms, fix the cell wall, and dissolve hydrophobic extracellular entities. After fixation, the cells were permeabilized in 4% paraformaldehyde (200 [mu]L) for 60 minutes at room temperature. The cells were then collected (again, 13,000 xg for 10 minutes) and resuspended in a mixture of 150 [mu]L hybridization solution and 50- 100 nM PNA probe and incubated in hybridization solution at 56[degrees]C for 2 hours. 2 mL wash solution (pre-warmed) was added to the mixture and incubated at 56[degrees]C for 50 minutes before final resuspension of the cells in 2 mL PBS. The samples were analyzed using a BectonDickenson (Franklin Lakes, NJ) FACSCalibur Flow Cytometer.

4. Epi-Fluorescence Microscopy for M. Parafortuitum Detection

Observation

The epi-fluorescence microscopy images in Figure 4 demonstrate the specificity of the epiFISH method. A fluorescent response was observed for the M. pamfortuitum cultures whereas no response was observed from the E. coli and P. aeruginosa cultures. A duplicate control sample for each epiFISH sample was stained using acridine orange (AO, 500 nM), which will label all viable microorganisms present. Figure 4 illustrates that, for M. parafortuitum, a similar number of cells are responding (orange/ red) for both the AO and PNA probes, thereby demonstrating that a high percentage of the target microorganisms are detected by the PNA probe using the epiFISH method. For E. coli and P. aeruginosa, a negligible percentage of the cells treated with Myco 1227 were detected after application of the epiFISH method, demonstrating the high specificity of the Myco 1227 PNA probe designed during this research.

Figure 5 provides a 100x composite photo of M. parafortuitum cells labeled with the Myco 1227 probe and observed using the epiFISH method. The photo illustrates the morphological characteristics (e.g., rod-shaped clustering) of the cells, providing a secondary confirmation of the organism type. To demonstrate the efficacy of the identification approach for mixed microbial communities, 10 random identification experiments were performed in blind-experiment fashion. In each case, a trained microscope operator was able to positively determine the presence or absence of M. parafortuitum in a mixed culture of E. coli and P. aeruginosa.

Figure 4

40x epiFluorescent Microscope Images

(a) Results with epiFISH probe, (b) Results with acridine orange control

Additional experiments were performed to compare the performance of the more traditional Myco 1227 probe with the molecular beacon MB 1227 probe (Table 1). The results were, on the whole, indistinguishable between the two different types of probes, except it was observed that MB 1227 fluorescence was qualitatively dimmer than Myco 1227, even after modifying epiFISH parameters. This observation was later verified quantitatively fcmFISH (see section 5). When using MB 1227, washing was still useful to increase signal- to-noise ratio (diffuse fluorescent residue was observed if the washing step was omitted), even if the probe concentration was diluted considerably. Extensive optimization with the epiFISH protocol for use with the molecular beacons did not yield significant improvements in performance. Consequently, the beacon form of the PNA probe was not considered to be a major advantage in this application.

Three different types of unused MWF (synthetic, semi-synthetic, and soluble oil) were examined to determine the impact of MWF matrix interactions on the epiFISH labeling procedure. Dilutions of 5% and 10% of each MWF concentrate were tested. A collection of M. parafortuitum was removed from pure culture during log-phase growth and incubated in MWF for four hours (at least one generation time) prior to epiFlSH. It was found that the ability to distinguish M. parafortuitum from P. aeruginosa or E. coli was unaffected by the initial presence of MWF, which was effectively removed by the first washing step of epiFlSH. The high speed of centrifugation, combined with the density of the cells, formed a clean pellet that was not obviously affected by the presence of oils, particulate, or other MWF constituents. Subsequent steps of the epiFlSH method were found to proceed unaffected by the initial presence of MWF for the ranges of oil concentration, ionic strength, and pH found in the three MWFs investigated here.

Figure 5

Composite Images of M. parafortuitum Clusters (100x oil immersion)

The impact of sustained M. parafortuitum growth in MWFs on the epiFISH method was also investigated during this research. The semi- synthetic formulation, which was the most hospitable media for mycobacteria growth, was used to culture M. parafortuitum for 10 to 20 generations prior to epiFISH analysis. Extended growth of M. parafortuitum in semi-synthetic MWFs was not observed to impact the efficacy of the epiFlSH detection approach. In other words, results analogous to Figures 4 and 5 were achieved regardless of growth in nutrient broth or MWF. This suggests that common MWF matrices do not interfere with the epiFISH method to a significant degree.

Discussion

During the development of the epiFISH and fcmFISH methods, the hybridization temperature was found to be the most critical parameter to control. The optimized hybridization temperature was 56[degrees]C for Myco 1227, which is near to the theoretical hybridization temperature (63[degrees]C) as calculated using the method developed by Worden et al. (2000). At lower temperatures (e.g., 50[degrees]C), instances of non-specific binding were observed for P. aeruginosa. For E. coli, hybridization of Myco 1227 was not observed under any conditions. Presumably, the differences in nonspecific binding behavior of P. aeruginosa versus E. coli were due to the fact that P. aeruginosa rRNA contains a sequence that is only two base pairs distinct from the target of the probe, while E. coli has a six base pair distinction (Table 2). This is consistent with the fact that a higher degree of stringency (e.g., increasing temperature) is necessary to keep slightly mismatched complementary sequences from hybridizing. At temperatures well above 56[degrees]C, the epiFISH method was less effective, or totally ineffective, for M. parafortuitum detection.

5. Quantification of M. Parafortuitum Using Flow Cytometry Probe Functionality and Cell Counting

The fcmFISH method was applied to M. parafortuitum, E. coli, and P. aeruginosa cultures under log growth conditions. Plots of fluorescence detection versus 90[degrees] scatter intensity for M. parafortuitum (left column) and E. coli (right column) samples are shown in Figure 6. These data demonstrate the appreciable detection of M. parafortuitum (detection defined as >10 intensity units). In contrast, E. coli is not detected using Myco 1227. A similar response was observed for P. aeruginosa. These data indicate that the Myco 1227 probe specifically detects M. parafortuitum when the fcmFISH protocol is applied.

For the quantification of M. parafortuitum cells, a count was operationally defined by considering the analysis of an unstained (control) M. parafortuitum population exhibiting log-phase growth characteristics. Based on these data, any particle demonstrating greater than 101 fluorescence intensity units on the y-axis and less than 102 side-scatter intensity units on the x-axis was considered to be a mycobacteria cell. Figure 6 shows that greater than 95% of FCM observable particles taken from the log-phase growth culture (14 hours growth) of M. parafortuitum are detected with the fcmFISH approach.

Figure 6

Flow Cytometric Analysis of M. parafortuitum and E. coli Cells With and Without fcmFISH with Myco 1227 Probe

rRNA Detection Target Considerations

To demonstrate conclusively that the probe is binding to the rRNA of M. parafortuitum, the fcmFISH method was modified to include a ribonuclease (RNase) digest. RNase was added after cell permeablization and prior to adding the PNA probe. Because RNase breaks down RNA, including rRNA, it was expected that the inclusion of RNase in the fcmFISH protocol would eliminate the rRNA targets of the PNA probe and no fluorescence response would be observabl\e by FCM analysis. The data illustrated in Figure 7 indicate that Myco 1227 specifically bound to rRNA, as no signal was detected by FCM after RNase treatment.

DeLong et al. (1989) observed that cellular ribosome content tends to increase proportionally with growth rate. Because Myco 1227 targets rRNA, it should be possible to correlate the intensity of the fluorescence signal observed via FCM with the rRNA content. Consistent with this anticipated trend, variations in fluorescence intensity were observed as a function of cellular growth stage during this research.

Figure 8 illustrates the response of M. parafortuitum cells to the fcmFISH method after allowing the log phase cells to incubate in growth medium over a period ranging from 24 to 192 hours. Depending on the growth stage of the culture, a variable percentage of background particulate that is below the intensity threshold (e.g., dead cells, cellular debris, and weakly fluorescing cells) is observed. In the log phase of growth, most of the particulate picked up by FCM are fluorescently labeled cells. After two days, the FCM plot illustrates a decrease in the percentage of fluorescently labeled cells and a decrease in the fluorescence intensity. This was attributed to lower expression levels of rRNA at later stages of growth, and therefore fewer targets per cell for Myco 1227 hybridization.

Because M. parafortuitum was observed to reach a stationary phase of growth after approximately 24 hours, it is perhaps unsurprising that only approximately 5% of the particulate material observed via fcmFISH after eight days is identifiable mycobacteria. By this time, a large percentage of cells are likely to be dead and mixed with cell debris and extracellular material. To better understand the composition of the population (dead vs. alive) after eight days without substrate replacement, the general nucleic acid probe Syto 62 (Molecular Probes, Eugene, OR) was added to the fcmFISH method. It was found that, of the background particulate, less than 10% responded to the general nucleic acid probe, signifying that most of the background was extraneous cell debris or extracellular material.

Figure 7

M. parafortuitum Response to fcmFISH Method Without RNase Treatment (left) and With RNase Treatment (right)

Figure 8

Results of FCM Analyses on Pure M. parafortuitum Culture with fcmFISH Applied After 24, 48, 96, and 192 Hours

fcmFISH as a Quantification Tool

Due to the dependence of FCM signal measurements on growth stage, the cells used for the fcmFISH experiments in this research were consistently observed at mid-log phase growth, approximately 14 hours after the start of a new generation. At this time, there were 5 x 10^sup 6^ to 1 x 10^sup 7^ cells per milliliter (mL) in the culture as determined by optical density (OD). As seen in Figure 9, FCM usually consistently indicated approximately one order of magnitude fewer cells present. This difference was likely due to the loss of cells generated by the centrifugation and resuspension steps involved with the fcmFISH method.

Figure 9

Number of M. parafortuitum Observed via FCM Under Different E. coli vs. Mycobacteria Ratio

A correlation coefficient of 0.98 was found between OD and FCM measurements, demonstrating the viability of fcmFlSH as a counting technology. Somewhat lower correlation coefficients were observed for cell populations that were older, likely due to less rRNA in individual cells and the fact that OD does not distinguish between live and dead cells. Moreover, the coefficient of variation for FCM typically observed (0.1) was significantly lower than that for the direct microscope counts (0.7) on which the OD calibration curve for M. parafortuitum was based. The low coefficient of variation, coupled with the fact that direct microscopic counts and OD cannot be easily used for the quantification of individual species within mixed cultures, emphasize the utility of fcmFISH as a tool for microbial enumeration in field applications.

Influence of MWF and Mixed Cultures on fcmFISH

M. parafortuitum recovery using fcmFISH was also investigated in the presence of a mixed microbial culture. To start, mixtures of E. coli and M. parafortuitum were produced at variable ratios ranging from 100x to 0.1x E. coli : M. parafortuitum in decade increments. Figure 9 illustrates the efficient recovery of M. parafortuitum over this range (coefficient of variation = 0.15), with results indicating a constant recovery within statistical error. This indicates that the fcmFISH procedure has the potential to detect mycobacteria in MWF even among a dense biological background.

An additional experiment was performed to confirm the specific targeting and quantification of M. parafortuitum in the presence of a dense E. coli population using the fcmFISH method. 1 x 10^sup 6^ cells/mL of mycobacteria (1x) and 1 x 10^sup 7^ cells/mL of E. coli (10x) were mixed in equal volumes. In one sample, the E. coli sample was stained using Syto 62 (red) probe prior to mixing with mycobacteria, and the mixed culture was subjected to fcmFISH without the addition of Myco 1227 probe (Figure 10a). In the second sample, E. coli (unstained) were mixed with mycobacteria and subjected to the unmodified fcmFISH method including Myco 1227 probe (Figure 10b). The FCM quantification data indicate the total ratio of E. coli (Figure 10a) to mycobacteria (Figure 10b) is about 8.9, which signifies that E. coli and mycobacteria are conserved in approximately equal proportions after application of the fcmFISH method.

After demonstrating the efficacy of fcmFISH for quantifying M. parafortuitum cultures maintained in nutrient broth, fcmFISH quantification of M. parafortuitum isolated from each of the three MWFs was compared to an identical experiment performed using epiFISH. Similar to epiFISH detection, it was observed that the MWF does not interfere significantly with the use of fcmFISH as a preparation for M. parafortuitum detection via FCM regardless of whether the mycobacteria was mixed with the MWF, or whether the culture was maintained in culture (as with the semi-synthetic MWF in section 4). Figure 11 illustrates that while there is some variation in the amount of recovery observed for each MWF, there is no statistically observable difference in the amount of M. parafortuitum recovered from any particular class of MWF chemistry.

Figure 10

Red Fluorescence of E. coli

(a) Separated from green fluorescence of M. parafortuitum, (b) Used to calculate ratio of E. coli to M. parafortuitum

6. Conclusions

This research has analyzed the use of a novel PNA-based rRNA probe for the detection of M. parafortuitum in MWFs. Flow cytometry (FCM) and epi-fluorescent microscopy protocols were developed and optimized using a fluorescent peptide nucleic acid (PNA) molecule, yielding the ability to qualitatively and quantitatively enumerate mycobacteria within MWF in a single day. It was found that the Myco 1227 PNA probe designed during this research could specifically recognize M. parafortuitum in the presence of a dense biological background, simulating a realistic MWF contamination occurrence. Based on these results, it was determined that the PNA probe confers a high degree of specificity in its target region of 16S rRNA (down to a two-base pair sequence change). Moreover, the optimized methods developed for FCM and epi-fluorescence microscopy were found to be insensitive to three common MWF chemistries used in machining operations.

Additionally, FCM analysis of mycobacteria over the course of eight days illustrated the potential for the method to be applied as a means to determine the relative growth stage of the population. This makes the method potentially useful for quickly developing anti- mycobacteria biocide strategies after a contamination event has occurred in a MWF system. Taken together, the evidence presented suggests that probes similar to the PNA probes investigated in this research, coupled with FCM or epi-fluorescence microscopy, are a viable means to significantly increase the speed by which mycobacteria species can be detected and eliminated from MWF systems.

Figure 11

Number of M. parafortuitum Observed via FCM After Cells Removed from Metalworking Fluids and Subjected to fcmFISH Method

Acknowledgments

This research was funded by the United States National Science Foundation under DMI-0093514. The authors also wish to thank Professor Peter Adriaens and Ms. Anna Khijniak of the University of Michigan at Ann Arbor for conducting the flow cytometry tests during this investigation.

This paper is an original work and has not been previously published except in the Transactions of NAMRI/SME, Vol. 31, 2003.

References

Abe, C.K. et al. (1999). "Simple and rapid identification of the Mycobacterium tuberculosis complex by immunochromatographic assay using anti-MPB64 monoclonal antibodies." Journal of Clinical Microbiology (v37), pp2693-3697.

Bernstein, D. et al. (1995). "Machine operators lung: a hypersensitivity pneumonitis disorder associated with exposure to metalworking fluid aerosols." Chest (v108), pp626-641.

Bockstahler, L.E. et al. (2002). "Peptide nucleic acid probe detection of mutations in Mycobacterium tuberculosis genes associated with drug resistance." Biotechniques (v32), pp508-515.

Byers, J.P. ed. (1994). Metalworking Fluids. New York: Marcel Dekker.

Davey, H.M. et al. (1999). "Variable selection and multivariate methods for the identification of microorganisms by flow cytometry." Cytometry (v35), pp162-168.

De Boer, E. et al. (1999). "Methodology for detection and typing of foodborne microorganisms." Int'l Journal of Food Microbiology (v50), pp11-130.

DeLong, E.F. et al. (1989). "Phylogenetic stains: ribosomal RNA- based probes for the identification of single cells." Science (v243), pp1360-1363.

Drobniewski, F.A. et al. (2000). "Differentiation of Mycobacterium tuberculosis complex and nontuberculous mycobacterial liqu\id cultures by using peptide nucleic acid-fluorescence in situ hybridization probes." Journal of Clinical Microbiology (v39), pp1032-1035.

Fricker, C.R. (2000). "From media to molecules: new approaches to the detection of micro-organisms in water." Water, Air, and Soil Pollution (v123), pp35-41.

Gunasekera, T.S. et al. (2000). "A flow cytometry method for rapid detection and enumeration of total bacteria in milk." Applied and Environmental Microbiology (v66), pp1228-1232.

Hanna, B.A. et al. (1999). "Multicenter evaluation of the BACTEC MGIT 960 system for recovery of mycobacteria." Journal of Clinical Microbiology (v37), pp748-752.

Hartmans, S. et al. (1992). "The Genus Mycobacterium - Nonmedical." In Balows, A. et al., The Prokaryotes, 2nd ed. (Vol. II). New Yourk: Springer-Verlag.

Huh, D. et al. (2001). "Air-liquid two-phase microfluidic system for low-cost, low-volume, and low-power micro flow cytometer." 5th Int'l Symp. on Micro Total Analysis Systems ([mu]TAS), Monterey, CA, Oct. 21-25, 2001.

Kido, C. et al. (2000). "Rapid and simple detection of PCR product DNA: a comparison between southern hybridization and fluorescence polarization analysis." Gene (v259), pp123-127.

Kreiss, K. et al. (1997). "Metalworking fluid - associated hypersensitivity pneumonitis: a workshop summary." American Journal of Industrial Medicine (v32), pp423-432.

Madigan, M.T. et al. (1999). Biology of Microorganisms. Englewood Cliffs, NJ: Prentice Hall.

Mattsby-Baltzer, I. et al. (1989). "Microbial growth and accumulation in industrial metalworking fluids." Applied and Environmental Microbiology (v55), pp2681-2689.

Meredith, S.K. et al. (1991). "Occupational respiratory disease in the United Kingdom 1989: a report to the British Thoracic Society and the Society of Occupational Medicine by The SWORD Project Group." British Journal of Internal Medicine (v48), pp292-298.

Miyake, R. et al. (1997). "Investigation of sheath flow chambers for flow cytometers (micro machined flow chamber with low pressure loss)" Trans, of the Japan Society of Mechanical Engg. (v62, n597), ppl06-113.

Muilenberg, M. and Burge, H.A. (1993). "Hypersensitivity pneumonitis and exposure to acid-fast bacilli in coolant aerosols." Journal of Allergy Clinical Immunology (v91), pp311-320.

NIOSH (1998). Criteria for a Recommended Standard: Occupational Exposure to Metal-working Fluids, National Institute of Occupational Safety and Health, Cincinnati, OH.

Ohiso, I. et al. (2000). "A fluorescence polarization assay using oligonucleotide probes for the rapid detection of verotoxin- producing Escherichia coli." Journal of Biotechnology (v81), pp15- 24.

Passman, F.J. et al. (2002). "Reassessing the health risks associated with employee exposure to metalworking fluid microbes." Lubrication Engg. (v58), pp30-38.

Porter, J. et al. (1997). "Go with the flow - use of flow cytometry in environmental microbiology." FEMS Microbiology Ecology (v24), pp93-168.

Robertson, B. et al. (1988). "Occupational asthma due to oil mists." Thorax (v43), pp200-205.

Rose, C. et al. (1996). "Biopsy-confirmed hypersensitivity pneumonitis in automobile production workers exposed to metalworking fluids: Michigan 1994-1995." MMWR (v45), pp606-610.

Rosenman, K.D. et al. (1997). "Work-related asthma and respiratory symptoms among workers exposed to metalworking fluids." American Journal of Industrial Medicine (v32), pp325-331.

Scarparo, C. et al. (2001). "Direct identification of mycobacteria from MB/BacT Alert 3D bottles: comparative evaluation of two commercial probe assays." Journal of Clinical Microbiology (v39), pp3222-3227.

Shapiro, H.M. (1995). Practical Flow Cytometry. New York: Alan R. Liss Inc.

Shelton, B.G. et al. (1999). "Mycobacterium sp. as a possible cause of hypersensitivity pneumonitis in machine workers." Emerging Infectious Diseases (v5), pp270-273.

Stender, H. et al. (1999). "Fluorescence in situ hybridization assay using peptide nucleic acid probes for differentiation between tuberculous and nontuberculous Mycobacterium species in smears of Mycobacterium cultures." Journal of Clinical Microbiology (v37), pp2760-2765.

Stender, H. et al. (2000). "Rapid detection, identification, and enumeration of Pseudomonas aeruginosa in bottled water using peptide nucleic acid probes." Journal of Microbiological Methods (v42), pp245-253.

Stender, H. et al. (2002). "PNA for rapid microbiology." Journal of Microbiological Methods (v48), pp1-17.

Thisted, M. et al. (1999). "Application of peptide nucleid acid probes for in situ hybridation." in Peptide Nucleid Acids-Protocols and Applications, P.E. Nielsen et al., eds. Portland, OR: Horizon Scientific Press.

Tortoli, E. et al. (1999). "Use of BACTEC MGIT 960 for recovery of mycobacteria from clinical specimens: multicenter study." Journal of Clinical Microbiology (v37), pp3578-3582.

Van der Gast, C.J. et al. (2001). "Identification and characterisation of bacterial populations of an in-use MWF by phenotypic and genotypic methodology." Int'l Biodeterioration & Biodegradation (v47), pp113-123.

Watterson, S.A. (2000). "Modern laboratory diagnosis of mycobacterial infections." Journal of Clinical Pathology (v53), pp727-732.

Weber, G. (1953). "Rotational Browninan motion and polarization of the fluorescence of solutions." Advances in Protein Chemistry (v8), pp415-459.

Worden, A.Z. et al. (2000). "in situ hybridization of Prochlorococcus and Synechococcus (marine cyanobacteria) spp. with rRNA-targeted peptide nucleic acid probes." Applied and Environmental Microbiology (v66), pp284-289.

Zugerman, C. (1986). "Cutting fluids: their use and effects on the skin." Occupational Medicine: State of the Art Reviews (v1), pp245-258.

Steven J. Skerlos, Laura A. Skerlos, Carlos A. Aguilar, and Fu Zhao, Dept. of Mechanical Engineering, University of Michigan, Ann Arbor, Michigan, USA. E-mail: skerlos@umich.edu

Authors' Biographies

Professor Steven J. Skerlos (www.umich.edu/~skcrlos) has served as an assistant professor of mechanical engineering at the University of Michigan at Ann Arbor since January 2000. He is director of the Environmental and Sustainable Technologies (EAST) laboratories, and his research focuses on the environmentally conscious design and manufacturing of engineering systems. He has a bachelor of science degree in electrical engineering with highest honors and a PhD in industrial engineering from the University of Illinois at Urbana Champaign. Professor Skerlos is a member of SME, ASME, STLE, the International Society of Industrial Ecology, and the Society of American Environmental Engineering and Science Professors.

Laura A. Skerlos received a master of science degree in molecular biology from the University of Illinois at Urbana-Champaign in 1999. Her research thesis enabled the first genetic description of nonculturable rumen protozoa through use of molecular tools. Currently, she is an associate research scientist in antibacterial pharmacology at Pfizer Global Research and Development in Ann Arbor, MI.

Carlos A. Aguilar is a research assistant in the Environmental and Sustainable Technologies Lab in the Dept. of Mechanical Engineering at the University of Michigan at Ann Arbor. He received his BS in mechanical engineering in December 2002. His research interests are in the design, fabrication, and modeling of microfluidic systems and bio-micro-electro-mechanical systems (Bio- MEMS). He is currently working on the detection and control of microorganisms in metalworking fluids and the development of recyclable metalworking fluids. He is a member of ASME and fellow of the University Research Opportunity Program (UROP).

Fu Zhao is currently a PhD candidate in the Dept. of Mechanical Engineering at the University of Michigan at Ann Arbor. He received his MS (1996) and BS (1993), both in thermal engineering, from Tsinghua University, China. His research interests are in environmentally benign design and manufacturing. He is currently researching the recycling of metal-working fluids using microfiltration to reduce health and environmental risks in the machine tool industry.

Copyright Society of Manufacturing Engineers 2003

More News in this Category