| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1,2ARUP Institute for Clinical and Experimental Pathology1 and Department of Pathology2, University of Utah, Salt Lake City, UT 84108, USA
Correspondence Jacob D. Durtschi durtscj{at}aruplab.com
Received March 15, 2005
Accepted June 4, 2005
A membrane-filter-based, fluorescent Gram stain method for bacterial detection in cerebrospinal fluid samples was developed and evaluated as a rapid, sensitive alternative to standard Gram stain protocols. A recently developed, modified version of the aluminium oxide membrane Anopore with low-fluorescence optical properties showed superior performance in this application. Other aspects of the fluorescent Gram stain system that were evaluated include membrane filter selection, strategies to reduce fluorescence fading and the effect of patient blood cells on bacterial detection in the fluorescently stained cerebrospinal fluid samples. The combination of the membrane filter's bacteria-concentrating ability and absolute retention along with high-contrast, fluorescent Gram discriminating dyes enabled rapid bacterial detection and Gram discrimination, with a 1–1.5 order of magnitude increase in the bacterial concentration limit of detection.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
One modified sample preparation method for Gram stains uses the commercially available Cytospin centrifugation system (Shannon Instruments). This method involves sample centrifugation in a disposable funnel apparatus clamped to a microscope slide. Micro-organisms preferentially settle centrifugally onto the slide surface while bulk liquid is absorbed into an integral fibre pad (Pelc, 1982). Other explored methods have employed filtration of the CSF sample through a membrane filter for surface capture of cells followed by imaging on the filter surface. This membrane filtration method is a relatively straightforward method for rapid concentration of samples relative to the classic Gram stain, in which a smaller sample volume is spread on a microscope slide. Membrane filtration also offers convenience and some time savings compared to concentration via centrifugation, which requires more time and user interaction.
Standard Gram staining on filtration surfaces has been demonstrated for bacterial cultures (Bitton et al., 1983; Romero et al., 1988; Saida et al., 1998) as well as bacteria spiked into filtered CSF (Lim et al., 1990). One problem associated with standard Gram staining on membrane filters is the background staining of the membrane filter, which reduces contrast between micro-organisms and the surrounding substrate. This background staining can be significantly higher on a filter substrate compared to a microscope slide. This staining is likely to be related to the large internal surface area of a membrane filter available to retain the stains. A second problem associated with standard Gram staining on membrane filters is the inhomogeneous nature of the filter substrate relative to a microscope slide. Under the transmission light microscope, surfaces can appear rough due to the membrane's inhomogeneous optical features. This extra image detail reduces the ability to discern micro-organism features or differentiate micro-organisms from background debris.
An alternative to a standard Gram stain is fluorescence microscopy and associated fluorescent Gram stains. These fluorescence methods have been developed for bacterial detection and Gram differentiation. One such method exploits the differential bacterial membrane permeability to fluorescent dyes of Gram-positive versus Gram-negative bacteria (Forster et al., 2002; Gunasekera et al., 2003; Mason et al., 1998).
One fluorescent Gram stain kit is the Live BacLight bacterial Gram stain kit (Molecular Probes), which consists of the green SYTO9 dye, the orange hexidium iodide (HI) dye and a proprietary microscope slide mounting fluid. Both SYTO9 and HI are DNA intercalating dyes that are maximally fluorescent when bound to double-stranded DNA and are relatively nonfluorescent in solution. Both dyes can also be excited with blue light near 490 nm. SYTO9 is highly permeant to bacterial membranes and readily labels bacterial DNA with a bright green fluorescence. HI is permeant only to Gram-positive bacteria and has a much higher DNA-binding affinity compared to SYTO9. HI therefore preferentially labels Gram-positive bacterial DNA with a bright orange fluorescence. The Live BacLight system requires that the bacteria have intact, viable membranes to ensure the correct Gram-positive or Gram-negative bacterial membrane permeability, and hence, proper staining. One solution has been to distinctly stain non-viable cells with the fluorescent stain, ethidium homodymer-2 (DEAD red, Molecular Probes).
Another fluorescent Gram stain system uses fluorescently labelled wheat germ agglutinin (WGA) that binds to Gram-positive bacterial membrane sites (Holm & Jespersen, 2003; Sizemore et al., 1990). The WGA conjugate combined with a fluorescent counterstain of a second colour constitutes this Gram stain system. Because the WGA system only binds to the exposed surface of Gram-positive bacteria, however, a smaller number of fluorophores are present at each bacteria in comparison to the DNA intercalating dyes used in the Live BacLight system, which permeate the bacterial cytoplasm. As a result, fluorescence levels from the WGA system are lower and visually less distinct than those from the Live BacLight system.
Fluorescence-based Gram stain techniques have been used in imaging not only on microscope slides but also on membrane filters (Lim et al., 1990). The membrane filter of choice for this fluorescence application would ideally exhibit properties including high sample-filtration rate, low plugging potential, a low-fluorescence background and a low non-specific binding affinity to the fluorescent labelling components. With the exception of the fluorescence background levels, standard membrane filters such as the polycarbonate track etch membrane and the aluminium oxide membrane (AOM) Anopore are adequate filters for imaging (Jones et al., 1989). Versions of the track etch membrane with incorporated black dyes to reduce membrane background fluorescence are commercially available. Track etch membranes, however, have some potential drawbacks relative to AOMs, such as lower porosity, which can lead to lower filtration flow rates. Another drawback of track etch is higher adsorption of fluorescent dyes caused by the higher hydrophobicity of track etch relative to AOM. In addition, commercially available darkened track etch membranes are relatively low in fluorescence but do have a noticeable fluorescence background. Although not yet commercially available, a low-fluorescence, black AOM filter has been developed as described by Durtschi et al. (2005). With its low fluorescence level, the black AOM is a good candidate for a filter-based fluorescence Gram stain.
Here we developed a simple Gram stain technique that combines the black AOM membrane filters and the DNA intercalating dyes of the Live BacLight Gram staining kit. We applied this technique to Gram detection of spiked bacteria in pooled CSF to evaluate the performance of our fluorescence-based system. The sharp, high-contrast imaging of the fluorescent Gram stain together with the sample concentration ability of filtration improve bacterial detection in CSF samples by greater than an order of magnitude.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Preparation of membrane filter card for filtration and imaging of samples.
Each 47-mm-diameter round membrane filter was incorporated into a laminate assembly that defined 12 round filtration spots as depicted in Fig. 1. At the top of this layered assembly was an aluminium tape layer, dye-cut into a rounded shape with an array of twelve 4 mm holes, immediately below which was the membrane filter with its filtration surface aligned upward. Beneath the membrane filter was a double-sided adhesive tape dye-cut identical in shape to the aluminium film. Below this adhesive layer was a thin latex layer (cut from laboratory latex gloves). Slits were cut in the latex layer across the full diameters of the 12 filtration regions. The latex layer was intended as a liquid flow valve and evaporation barrier at the bottom surface of the membrane. When vacuum pressure is applied at one of the filtration sites, liquid will flow through the membrane, push against the latex layer and open the slit, allowing flow through the membrane. When vacuum pressure is released, the slit reseals, restricting evaporation and flow on the bottom surface of the membrane. Below the latex layer, a second double-sided adhesive dye-cut layer was added. This assembly was then cut into smaller rectangular pieces each with two filtration areas. Each of the two-filtration-area pieces was then adhered to a thin, anodized aluminium card with holes drilled to match the size and spacing of the laminate assembly filtration areas.
|
Quantification of bacterial concentration based on optical density.
Bacterial cultures of Staphylococcus aureus (ATCC 25923) and Escherichia coli (ATCC 25922) were grown in Luria–Bertani (LB) broth at 37 °C. While in the exponential growth phase, the cultures were placed on ice to retard further growth and serially diluted in sterile LB broth. The optical density at 660 nm (OD660) of higher concentration dilutions was measured on an Amersham Pharmacia Biotech Ultrospec 2100 pro spectrophotometer. Lower concentration dilutions were plated in 50 µl aliquots onto culture plates. Specifically, S. aureus dilutions were plated onto sheep blood agar (SBA) culture plates and E. coli dilutions were plated onto LB culture plates. After a 24 h incubation at 37 °C, plate colonies were counted. Colony counts were used to correlate OD660 readings to actual bacterial concentration.
Bacteria-spiked CSF preparation for sensitivity study.
Fresh cultures of S. aureus and E. coli were generated by the inoculation of LB broth with single bacterial colonies from bacterial culture plates. The cultures were incubated overnight at 37 °C. The cultures were then placed on ice to retard further growth. Culture dilutions of approximately 10-fold were prepared to reach the more sensitive range of the OD660 versus bacterial concentration curves. OD660 values were then measured and bacterial concentrations were determined based on the OD660 versus concentration curves. The initial stock concentrations were calculated to be 6.8 x 108 bacteria ml–1 and 3.5 x 108 bacteria ml–1 for E. coli and S. aureus, respectively. Stock bacterial solutions were then serially diluted into LB broth to create a dilution series of 3 x 108, 1 x 108, 3 x 107, 1 x 107, 3 x 106, 1 x 106, 3 x 105, 1 x 105, 3 x 104 and 1 x 104 bacteria ml–1. These dilutions were stored on ice to retard bacterial growth. Twenty-fold dilutions of these series were made into a stock of pooled, refrigerated CSF to create E. coli-spiked CSF and S. aureus-spiked CSF in concentrations of 1.5 x 107, 5 x 106, 1.5 x 106, 5 x 105, 1.5 x 105, 5 x 104, 1.5 x 104, 5 x 103, 1.5 x 103 and 5 x 102 bacteria ml–1. Prior to the bacterial spike, pooled CSF was tested via both standard and fluorescent Gram staining and found to be free of bacteria (data not shown). The spiked CSF samples were then refrigerated until use (testing was completed within 4 h of spiked sample preparation).
Standard Gram stain procedure.
All standard Gram staining was performed using the Cytospin centrifugation system (Shannon Instruments). Gram staining reagents were purchased as a kit (Protocol Gram stain kit, Fisher Scientific). The Gram stain microscope slide preparation began with the attachment of the disposable Cytofunnel sample chamber to a microscope slide. One hundred and fifty microlitres of 22 % bovine serum albumin (BSA) in deionized water solution was added to the Cytofunnel assembly, which consisted of the Cytofunnel, the glass microscope slide and the associated metal clamp. Two hundred microlitres of bacteria-spiked CSF was then added to the assembly. The assembly was placed in the Cytospin centrifuge for 5 min at 2000 r.p.m. The assembly was then disassembled and the slide was placed on a 200 °C hot plate for approximately 1 min until dry and thoroughly heated. Crystal violet solution was pooled on the slide for 30 s and then shaken off. Iodine solution was then pooled on the slide for 30 s and shaken off. Decolourizing solution was squirted onto the vertically held slide until the visible front of crystal violet staining had moved well past the area of bacterial fixation (approximately 10 s). The slide was briefly rinsed in a stream of deionized water and then shaken off. The safranin counterstain solution was pooled on the slide for 1 min followed by another deionized water rinse. The slide was patted dry with disposable absorbent wipes. Total processing time was approximately 14 min. Immersion oil was added directly to the microscope slide surface, which was then analysed on a Spencer binocular transmission light microscope, with a x100 oil-immersion objective and x10 eyepieces with 18 mm apertures.
Filter-based fluorescent Gram stain procedure.
Live BacLight kit fluorescent staining solution consisted of a 667-fold dilution of the kit's SYTO9 stock solution and a 667-fold dilution of the kit's HI stock solution in LB broth. All steps in this procedure were performed at room temperature. Staining solution was mixed fresh each day. Vacuum tubing was attached to the back of the aluminium card with a vacuum pressure of approximately 17–34 kPa. The initial step of the membrane-based Gram staining was the vacuum filtration of a 200 µl bacteria-spiked CSF sample through the 3.7-mm-diameter membrane filter area. Filtration times were typically 30–60 s. The vacuum pressure was removed immediately after liquid filtration was complete to avoid drying of bacteria. An approximately 15–20 µl droplet of fluorescent staining solution was quickly added to the filter area (with the vacuum off). This was incubated for 8 min, a time that was approximately the minimum required for proper Gram-specific colouration. Over the course of the 8 min incubation the droplet did not dry or leak through the membrane significantly. The staining solution was then filtered through the membrane via vacuum. A drop of the Live BacLight kit mounting fluid was put on the membrane, followed by a microscope coverslip. Total processing time was approximately 10 min. The membrane and sample assembly was then ready for analysis on a Zeiss binocular epifluorescence microscope, 485 band pass excitation filtered, 520 long pass emission filtered, with a x100 oil-immersion objective and x10 eyepieces with 18 mm apertures.
Bacterial identification and quantification from stained samples.
Slides were evaluated by three individuals who were blinded to the bacterial concentrations. Samples were ranked by bacterial count using the following 0 to 4 scale: 0, no bacteria found; 1, less than 1 bacteria per field of view (FOV); 2, 1–5 bacteria per FOV; 3, 6–30 bacteria per FOV; 4, greater than 30 bacteria per FOV. The FOV for all microscope configurations in our study was a 0.18 mm diameter area. This diameter is set by the 18 mm eyepiece aperture value divided by the x100 objective power. Digital images were taken of typical FOVs from prepared and stained samples with original bacterial concentrations of 1.5 x 106, 5 x 105 and 1.5 x 105 bacteria ml–1. Images were collected through the microscope eyepiece tube (eyepiece removed) with a Nikon 5000 CoolPix Camera and an associated eyepiece tube adapter with an 18 mm aperture (Edmond Industrial Optics). Camera settings were full manual mode, 2 s exposure and 3.6 F stop. Total bacterial counts for each of these images were made manually. These counts were then used as relative bacterial surface concentration indicators.
Comparison of background fluorescence for three membrane types used for bacterial staining with the BacLight kit.
Three membrane types were compared for use in membrane-filter-based fluorescent Gram staining. These membrane types were the unmodified 0.2 µm absolute cut-off AOM (47 mm Anodisk), the electroless nickel deposited, blackened version of this AOM described above and a commercially available, 0.4 µm pore size, blackened, polycarbonate track etch membrane (Millipore). These three filter types were incorporated into the filtration disposable, and the fluorescent Gram staining procedure was performed using bacteria-spiked CSF as samples as described above. The membranes were imaged and digitally photographed on the above-described fluorescence microscope. Image background fluorescence values were determined from these images with the MaxIm DL image processing software version 3.10 (Diffraction Limited). Image processing began with the conversion of red/green/blue colour images to grey-scale by averaging the three colour-intensity values for each pixel. Regions of images were then selected that were free from micro-organisms or particulate debris. This collection of pixels was then averaged for the image from each membrane type and compared as an indicator of unwanted background fluorescence.
Use of anti-fade agent Trolox to reduce fluorescence fading.
A popular anti-fade agent, Trolox (Sigma), was added to evaluate its effect on the fluorescence fade (photo-bleaching) rate of the SYTO9 and HI dyes from the Live BacLight kit. Manders et al. (1999) recommended a Trolox concentration of 0.1 mM for live cell fluorescence imaging. We evaluated this concentration and a concentration of 1.0 mM to determine the effects of high Trolox levels. Bacteria-spiked samples were fluorescently stained and prepared as before. The fluorescent staining solutions used in this section, however, were spiked with 0.0, 0.1 or 1.0 mM Trolox. The prepared samples were imaged under the fluorescence microscope and evaluated for fluorescence fade rate under the maximum light-source intensity. Since the orange HI dye was the most susceptible to fading, this dye was used as an indicator of anti-fade properties. The exposure time needed to fade the orange colour from the Gram-positive bacteria leaving only a relative green fluorescence remaining for a given view field was measured for each of the Trolox concentrations.
Evaluation of white and red blood cells in fluorescently stained bacterial samples.
The effect of imaging bacteria in the presence of white and red blood cells was evaluated by spiking 5 µl of fresh whole blood into 200 µl samples of bacteria-spiked pooled CSF. These samples were fluorescently stained and prepared as before. Increased filtration time of the blood-spiked samples compared to the pooled CSF samples was noted. Fluorescence microscope images were taken and qualitatively evaluated to determine the effect of white or red blood cells on bacterial detection and classification.
|
RESULTS AND DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Other track etch membrane issues seen in our studies show the potential advantages of the blackened AOM. One such issue was the relatively high track etch membrane interaction with the fluorescent dyes used for bacterial staining. Over time, the track etch membrane adsorbs the dyes that were used and could cause reduced or improper bacterial staining. This is particularly apparent in our simplified protocol, in which bacterial staining takes place on the membrane surface rather than in solution. On-membrane staining could potentially allow dye adsorption on the membrane before adequate bacterial dye uptake. The track etch membrane is also very flexible unlike the rigid nature of the AOM membrane. The flexible nature of the track etch material led to vacuum filtration-induced-membrane bowing. This bowed, sloped membrane surface was present during imaging and hindered focus across the microscope's entire FOV. This membrane flexibility is a potential problem for simple flow-through disposable systems that are well-suited to rapid Gram stain diagnostics. Membrane flexibility is clearly less problematic if the track etch membrane is mounted before imaging on a rigid surface such as a microscope slide.
Fluorescence fading in membrane-based fluorescence Gram staining
Fluorescence fading (i.e. photobleaching), another concern of fluorescence imaging, did appear to be a potential problem for this Gram stain application, in which extended microscope viewing may be desirable. To address this issue, we evaluated the addition of an accepted, oxygen-scavenging, anti-fade agent, Trolox, to our fluorescent dye solution (Manders et al., 1999). A typical concentration of Trolox, 0.1 mM, as well as no Trolox and a relatively high Trolox concentration, 1.0 mM, were compared. A Trolox concentration of 1.0 mM was too high and, within 15 min of sample incubation, caused bacteria to take on a faded and contracted appearance with improper stain colouration. This apparent bacterial degradation occurred with or without exposure to microscope illumination and was clearly unacceptable. The high Trolox concentration was not further evaluated.
The lower Trolox concentration of 0.1 mM was compared to no Trolox based on the fade rate of the orange HI dye, which appeared to be far more sensitive to light than the green SYTO9 dye. The lengths of time under high microscope excitation for the orange HI colour to fade until only green fluorescence was visually detectable were 10 s and 40 s for the 0.0 mM and 0.1 mM Trolox concentrations, respectively. When the light intensity of the microscope was reduced by approximately 70 %, the respective fade times were approximately 30 s and 5 min of illumination exposure. The addition of Trolox at 0.1 mM clearly reduced the fluorescence fade rate compared to no addition of Trolox. Trolox appears to extend the usable viewing time by at least a factor of four in our configuration. In addition, illumination intensity has an obvious impact on fade rate. These results for fade rate also indicate that fluorescent stain assays, such as this, can be very useful as a rapid diagnostic with reduced applicability if extended viewing is required.
Effect of red and white blood cells on bacterial detection with membrane-based fluorescence Gram staining
Turbid CSF samples are sometimes encountered in the course of Gram staining. The turbidity is most often caused by white blood cells associated with an immune response or red blood cells (RBCs) associated with accidental introduction of blood into a sample. Although standard Gram stained bacteria can be visualized in the presence of red or white blood cells, excessive concentrations of blood cells can lead to multiple cell layers on the standard Gram staining surface when prepared with the Cytospin system, which can then obscure bacterial visualization. The effect of these cells on the fluorescent stain method was also studied.
Fresh blood-spiked samples of CSF were prepared using our fluorescent Gram stain method and evaluated. The resulting images showed that white blood cells stained a relative bright green, which would potentially mask any green-stained, Gram-negative bacteria overlapping the white blood cells (see Fig. 2D). RBCs appeared to interfere very little with bacterial staining. Bacterial fluorescence intensity appeared to be unchanged in the presence of RBCs and the RBCs did not appear to fluoresce beyond extremely faint green outlines. This faint fluorescence did not appear to obstruct bacterial detection or Gram classification. These images showed that whole blood contaminated samples (without high white blood cell concentrations) do appear to allow adequate bacterial detection. High concentrations of white blood cells, however, will likely obstruct bacteria. In addition, bacteria that are present inside phagocytic white blood cells, which can be visualized by Gram stain methods, may be difficult or impossible to identify using the fluorescence method due to the strong green fluorescence of the cells. In this study, we did not have samples that exhibited intracellular bacteria by the standard Gram stain method. We were therefore unable to evaluate the ability of the fluorescent Gram stain method to detect intracellular bacteria.
|
Comparison of standard and membrane-based fluorescence Gram stain methods for bacterial detection
We compared the fluorescent Gram stain method with a common Gram stain protocol for low-bacterial-count detection sensitivity. We applied both staining methods to an identical dilution series of bacteria-spiked pooled CSF samples. Because the components of the fluorescent Live BacLight Gram staining kit used in our studies have been previously evaluated for a range of bacterial strains (Forster et al., 2002; Gunasekera et al., 2003; Holm & Jespersen, 2003; Mason et al., 1998), we limited our study to one Gram-positive and one Gram-negative bacteria, S. aureus and E. coli, respectively.
Results were reported on a semi-quantitative scale from 0 to 4 indicating no bacteria present, less than 1 bacteria per FOV, 1–5 bacteria per FOV, 6–30 bacteria per FOV and greater than 30 bacteria per FOV, respectively. The results, in Fig. 3, indicate that the fluorescence method showed 1–1.5 orders of magnitude higher concentrations of detectable bacteria across all samples. Bacteria were not detectable by standard Gram stain when the bacterial concentration was below 5000 bacteria ml–1, whereas the fluorescence method allowed detection down to our lowest sample concentration, 500 bacteria ml–1. Although the absolute detection limit of the fluorescence method was not determined, these experiments indicate that the fluorescence method showed at least one order of magnitude improvement in the absolute detection of bacteria in low-bacterial-count CSF samples. The visual contrast between the bacteria and background was significantly higher for the fluorescence method than for the standard method. This contrast clearly increased the ease of bacterial identification. However, some particulate debris with varying levels of fluorescence was present in the pooled CSF samples as shown in Fig. 2(B, C). Debris fluorescence and morphology was significantly different to that of the bacteria and, hence, debris did not appear to significantly hinder bacterial detection.
|
To generate quantitative data on the ability of the filtration-based, fluorescent Gram stain to concentrate bacterial samples, fluorescent and Cytospin-based Gram stain images from three intermediate sample concentrations were collected and the bacteria counted. The fields of view of the fluorescence and standard light microscopes were the same size, which allowed total bacterial counts from images to indicate the relative concentrating abilities of the filtration-based fluorescence method versus the Cytospin-based Gram stain method. Results are shown in Fig. 4. The mean bacterial count increase of the fluorescence method relative to the standard Gram stain method was 38-fold.
|
The Cytospin method's decreased sensitivity for bacterial detection in this study likely stems from three main causes: the method's poorer bacterial concentrating ability, its poorer bacterial retention during the staining procedure and possibly the poorer visual contrast of the standard Gram stains used in the Cytospin method compared to the fluorescent stains. Bacterial concentration in the Cytospin method, which concentrates bacteria onto a microscope slide, is significantly less effective than in the filtration method, which concentrates bacteria directly onto a much smaller surface area of filtration membrane. Bacterial retention in the standard Gram stain method can suffer due to poor and inconsistent bacterial adhesion to the microscope slide, which can lead to bacterial loss during the staining and rinsing procedure (Romero et al., 1988). The fluorescence, filtration-based method alleviates this possible problem because all fluids flow through the membrane and therefore ensure retention of all bacteria on the filter surface. Visual contrast, as previously discussed, does appear to be lower with standard Gram stains versus fluorescent staining. This improved optical contrast in the fluorescence method may increase the likelihood that bacteria are identified and counted.
As a visual reference of fluorescent bacterial staining free of CSF and its associated background, Fig. 2(A) shows an image of fluorescent-Gram-stained mixed E. coli and S. aureus with no CSF. Images of the fluorescent Gram-stained E. coli and S. aureus in CSF are shown in Fig. 2(B) and Fig. 2(C), respectively.
The filtration-based fluorescent Gram stain proposed here appears to have some significant advantages over our reference Gram stain system, the Cytospin system. The fluorescent Gram stains also have some drawbacks and concerns relative to standard Gram stains. White blood cells fluoresce strongly and can obscure intracellular as well as extracellular bacteria. In addition, intracellular bacteria have a high likelihood of being dead or compromised, which would likely lead to improper staining with the live Gram staining fluorescent dyes we have used. The current studies provide a foundation for future development and evaluation of alternative fluorescent staining methods that are compatible with intracellular bacteria and that do not require live bacteria. The fluorescence system does, however, appear to have significant advantages over the Cytospin system and other standard alternatives. These alternatives include direct smearing of a small CSF sample volume onto a microscope slide for conventional Gram stain or the centrifugation of a larger CSF sample followed by a microscope slide smear of the resulting centrifugation pellet. These methods do not have the combined procedural simplicity and bacterial-concentrating ability of the filtration-based fluorescent Gram stain. These filtration advantages along with the bright, high-contrast staining of the fluorescent dyes provide important elements of a potentially superior alternative to standard Gram staining of clinical CSF samples.
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
