Protocol

Use of Fluorescence In Situ Hybridization and the daime Image Analysis Program for the Cultivation-Independent Quantification of Microorganisms in Environmental and Medical Samples

Department of Microbial Ecology, Vienna Ecology Centre, University of Vienna, A-1090 Vienna, Austria

Corresponding author (daims{at}microbial-ecology.net)

INTRODUCTION

Conventional cultivation-based methods to measure microbial abundance are unsuitable for quantifying uncultured microorganisms that constitute the majority of microbial life in most environmental or medical samples. This problem is solved by the quantification approach described here, which combines fluorescence in situ hybridization (FISH) with rRNA-targeted probes and digital image analysis. By measuring the areas of probe-labeled biomass in randomly recorded image pairs, an unbiased estimate of the relative biovolume of the population of interest can be obtained. This approach expresses abundance as "biovolume fraction" (relative to the total biovolume of the whole microbial community). This value equals the share of biochemical reaction space occupied by the quantified population and thus can be more relevant ecologically than absolute cell numbers (e.g., a few large cells can contain the same biovolume as many small cells). Another advantage lies in the complete independence of this method from the morphology of the quantified organisms. Regardless of whether the target microbes occur as single cells in plankton samples, as filaments, or as dense aggregates in biofilms, this cultivation-independent method allows the composition of complex microbial communities to be determined.

RELATED INFORMATION

This protocol focuses on the steps that follow the FISH procedure and lead to quantitative results. Differences from standard protocols for rRNA-targeted FISH are emphasized. The procedure presented here uses the daime ("digital image analysis in microbial ecology") software (Daims et al. 2006), which contains functions optimized for this quantification protocol. It is available for Microsoft Windows (2000, XP, and Vista) and for Linux platforms and can be downloaded for free, together with a user’s manual, at http://www.microbial-ecology.net/daime. Additional details on the procedure are available in Daims and Wagner (2007).

MATERIALS

This protocol requires reagents and equipment needed for FISH with rRNA-targeted probes. For details, see, for example, Amann (1995), Pernthaler et al. (2002), and Daims et al. (2005). The following list mainly contains additional items required for the quantitative FISH approach.

Reagents

Agarose (0.5%-1%) (for unfiltered samples only)

Melt the agarose and allow to cool until it is still liquid but warm to the touch before use.

Anti-fadent (e.g., Citifluor AF1; Citifluor Ltd.)

Ethanol (50%, 80%, and 96%)

FISH oligonucleotide probe, population-specific

The probe should target the organism or group of organisms to be quantified and should be labeled with a fluorochrome that is clearly distinguishable from the nucleic acid stain or universal FISH probe mix.

FISH probe mix, universal (for dual hybridization only)

This is an equimolar combination of EUB338 (5'-GCTGCCTCCCGTAGGAGT-3') (Amann et al. 1990), EUB338-II (5'-GCAGCCACCCGTAGGTGT-3') (Daims et al. 1999), and EUB338-III (5'-GCTGCCACCCGTAGGTGT-3') (Daims et al. 1999). It targets most known bacteria and should be labeled with a different fluorochrome than the population-specific probe.

Formamide

Nucleic acid stain (e.g., SYBR Green I) (for nuclear staining only)

Use a stain that labels both DNA and RNA.

Paraformaldehyde (4%, prepared in PBS for bacterial FISH; ice-cold) (for Gram-negative target populations only)

Phosphate-buffered saline (PBS) for bacterial FISH (pH 7.2-7.4) (ice-cold for Steps 2.iv and 2.v)

Samples to be analyzed, environmental or medical

Equipment

Centrifuge

Coverslips

Filters, isopore membrane, polycarbonate, 0.22-µm, 25-mm diameter (Millipore GTTP02500) (for filtered samples only)

Hybridization oven, preset to 46°C

Microscope, epifluorescence or confocal laser scanning (CLSM)

An epifluorescence microscope must be equipped with a charge-coupled device (CCD) camera for recording digital images of probe-labeled microbial cells. A CLSM is needed to quantify biofilm populations without cryosectioning and often yields more reliable quantification results than an epifluorescence microscope.

Personal computer equipped with suitable image analysis software (e.g., the daime program)

Petri dish lid, plastic, on ice (for unfiltered samples only)

Scissors (for filtered samples only)

Slides, microscope

Support filters, nitrocellulose, 0.45-µm, 25-mm diameter (Sartorius) (for filtered samples only)

Vacuum filtration system, glass (GV 025 series; Whatman) (for filtered samples only)

Waterbath preset to 48°C

METHOD

Sample Preparation

1. Filter or centrifuge samples with low microbial cell density (e.g., oligotrophic lake water) to concentrate the cells before fixation. Resuspend centrifuged cells in PBS for bacterial FISH.

See Glöckner et al. (1996) for a protocol to filter and subsequently fix planktonic cells.

2. Fix the samples containing the microbes to be quantified:

For Gram-positive target populations:

i. Mix the samples 1:1 with ice-cold 96% ethanol.
Analyze ethanol-fixed samples on the same day or store at -20°C for, at most, 72 h before quantification. Lysis of Gram-negative cells in the sample can bias quantification results even if the targeted population is Gram-positive.

For Gram-negative target populations:

ii. Mix one volume of the sample with three volumes of ice-cold 4% paraformaldehyde.

iii. Incubate for 3-12 h at 4°C. Do not freeze.
Longer fixation times or higher temperatures can render the envelopes of Gram-negative cells impermeable to oligonucleotide probes.

iv. Pellet the cells by centrifugation at 15,000g for 5 min at 4°C. Resuspend the pellet in ice-cold PBS for bacterial FISH. Repeat two to three times to remove residual paraformaldehyde.

v. Resuspend the final pellet in ice-cold PBS for bacterial FISH. Add an equal volume of 96% ethanol.
These samples can be stored for several months at -20°C.

Fluorescence In Situ Hybridization

The basic procedure for FISH with rRNA-targeted probes has been described in detail by Amann (1995) and Daims et al. (2005). This protocol focuses on the differences between the standard qualitative procedure and quantitative FISH. Hybridized and dried samples can be stored for several weeks at -20°C before evaluation by fluorescence microscopy.

3. Process the samples for FISH:

For filtered samples:

i. Cut the filters into small pieces.

ii. Place the filter pieces onto microscope slides.
Proceed to Step 6.

For unfiltered samples:

iii. Apply 5-10 µL of resuspended fixed sample onto a microscope slide.

iv. Dry the slide for ~10 min at 46°C until all liquid has evaporated.

v. Apply another 5-10 µL of fixed sample on top of the first layer. Repeat Step 3.iv.

vi. Repeat Step 3.v until a thick, dry layer of biomass is obtained on the microscope slide.
With a dense sample such as activated sludge, two to three iterations are usually needed to immobilize the required amount of biomass on the slide.

vii. Dip the slide horizontally in molten agarose for ~5 sec.

viii. Immediately place the slide (biomass side up) onto a precooled Petri dish lid (on ice). Allow the agarose to solidify.

ix. Wipe off excess agarose from the bottom of the slide only.

4. Dehydrate the samples by incubating the slides through a graded series of ethanol:

i. 50% ethanol for 3 min.

ii. 80% ethanol for 3 min.

iii. 96% ethanol for 3 min.

5. Dry the slides for ~15 min at 46°C.
The agarose should be a very thin, dry layer on top of the samples resembling a "glue" that sticks to the slide and prevents biomass detachment during the remainder of the procedure.

6. Perform FISH as described by Amann (1995), Daims et al. (2005), or (for filtered samples) Glöckner et al. (1996).
Use a formamide concentration appropriate for the population-specific probe. The universal FISH probe mix can be used with 0%-60% formamide.
See Troubleshooting.

For dual hybridization:

i. Hybridize the samples with the population-specific FISH oligonucleotide probe and the universal FISH probe mix simultaneously.

For nucleic acid staining:

ii. Hybridize the samples with the population-specific probe.

iii. Wash the samples for 10-15 min at 48ºC as described in Daims et al. (2005).
The washing buffer composition depends on the stringency of the hybridization.

iv. After washing, apply 20 µL of nucleic acid stain SYBR Green I (10,000-fold dilution) to each sample.

v. Incubate in the dark for 10 min at room temperature.

vi. Wash the slides briefly in water. Air-dry.

Microscopy and Image Acquisition

7. Mount the samples:

i. Apply 2 drops of anti-fadent to the microscope slide.

ii. Place a coverslip on top of the drops and the biomass.

iii. Wait 5-10 min until the anti-fadent has spread and penetrated the biomass.

8. Observe the sample by epifluorescence or confocal microscopy.
Use a magnification of 400X (or lower) to ensure that each image contains enough biomass to obtain statistically meaningful results.
See Troubleshooting.

9. Acquire test images of the signal for each FISH probe (population-specific and universal/nucleic acid stain). Select regions with cells stained by both probes, with approximately the same area in the images of either probe signal (i.e., the cells should be congruent in the images).
See Troubleshooting.

10. Adjust the detector settings of the CLSM or the exposure time of the CCD camera to register signals from both probe types.
See Troubleshooting.

11. When all adjustments have been made, take digital images of 20-30 fields of view (FOVs) at randomly chosen positions:

i. Move the slide in the x and y dimensions, and change the focal plane (z dimension) to select random FOVs.
Each analyzed FOV should contain some microbial biomass.

ii. Do not select the FOVs by looking for the population to be quantified.
Such nonrandom selection seriously biases the results.

12. For each randomly chosen FOV, capture one image of the population-specific FISH probe signal and one image of the universal FISH probe (or nucleic acid stain) signal.

13. Store the paired images from each FOV with numbered file names on disk as uncompressed TIFF files, 8 bits per pixel (monochrome) or 32 bits per pixel (RGB color).
These formats are read by most image analysis programs. Choose file name freely, as long as the image pairs are numbered and named consistently according to the corresponding probe and the FOVs analyzed.

Image Analysis

The following portion of the protocol uses functions contained in the daime software. If using another program, adapt the respective steps accordingly.

14. Start the daime program. Import the images of the population-specific probe signal and of the universal probe signal (or nuclear stain) as two separate image "series" (batches).
Refer to the daime user manual for details on importing TIFF images. The images are sorted automatically based on the numbered file names.

15. Choose the option "Stereology:Biovolume fraction" in the "Analysis" menu.

16. In the following dialog, indicate which image series contains the population-specific probe signal and which series contains the universal probe signal (or nucleic acid stain). Click "OK."
The images will be segmented automatically, that is, objects (single cells and biomass aggregates) will be detected by the computer.

17. Use the "Object Editor" window to exclude individual objects detected in each image from further analysis:
This step is necessary if the images contain artifacts such as fluorescent plant material, which should not be quantified. The daime manual explains how to exclude objects.
See Troubleshooting.

i. Be sure to exclude artifacts in both images of the affected image pair (FOV).
The "Object Editor" has a button to switch between the two images of each pair.

ii. When all necessary changes are made, click "OK."
The program will now measure the areas of all objects in all images, except objects that have manually been excluded in the previous step.

18. Consult the dialog for the results of the analysis, including:

i. The biovolume fraction of the quantified microbial population in percent (relative to the total biovolume of all detected microbes).

ii. Statistics that allow the accuracy of the quantification to be assessed.
Refer to the daime manual for details.
See Troubleshooting.

19. If desired, export results as text (ASCII) files for further evaluation by other software.

20. Click "OK" to close the dialog screen.

TROUBLESHOOTING

Problem: Biomass detaches from the microscope slide during FISH despite being covered with agarose.

[Step 6]

Solution: Coat glass slides with gelatin or poly-L-lysine for better adhesion of biomass to the slide surface. See Daims et al. (2005) for details.

Problem: There is nonspecific fluorescence signal from autofluorescent material or from nonspecifically detected microbes.

[Step 8]

Solution: These signals can bias the quantification results. Reevaluate putatively specific FISH probes in silico by comparing the probe sequences to current rRNA sequence databases. If necessary, adjust the hybridization stringency or design more specific probes. See Wagner et al. (2003) for additional solutions for nonspecific fluorescence after FISH.

Problem: Fluorescence signals are very dark, or no fluorescence is visible at all.

[Step 8]

Solution: Consider the following:

1. If the target population is present but not detected by FISH, check whether the sample has been fixed correctly. (Use paraformaldehyde fixation for Gram-negative target populations only.)

2. The limited accessibility of thick cell envelopes to FISH probes can be improved by ethanol fixation and subsequent enzymatic pretreatment.

3. Low ribosomal content within the cells can result in low fluorescence intensity. Signal amplification techniques such as CARD-FISH (Pernthaler et al. 2002) can be used to increase the strength of the signal.

4. See Wagner et al. (2003) for a review of factors leading to problems with FISH, and their solutions.

Problem: It is difficult to ensure that the same biomass particles (cells or cell aggregates) are congruent (i.e., have the same area) in the image of the specific probe signal and the image of the universal probe signal.

[Step 9]

Solution: Adjust the detector settings of the CLSM or the camera to improve the situation. Once the settings look satisfying, take all image pairs with exactly the same settings and avoid subsequent changes. Cells do not need to be totally congruent; a small deviation (e.g., 5%-10% incongruence) usually is acceptable. In the "Results" dialog, the daime program indicates the average congruence of the same biomass particles in the images (in percent). Values above 90% should be considered acceptable.

Problem: Agarose exhibits autofluorescence that hampers image acquisition by epifluorescence or confocal microscopy.

[Step 10]

Solution: Use a lower concentration of molten agarose, or use a different batch (or brand) of agarose.

Problem: Some regions in the images of probe-labeled microbes in biofilms look blurred, and these blurred regions interfere with image analysis.

[Step 17]

Solution: If a CLSM is used to record the images, reduce the diameter of the confocal pinhole to obtain sharper images. If an epifluorescence microscope is used and no CLSM is available, consider cryosectioning the samples to obtain sharp images of thin sections.

Problem: The microbial target population is distributed very heterogeneously in the sample. Therefore, only some of the randomly chosen FOVs actually contain this population, and the quantification results might not be very reliable.

[Step 18]

Solution: Take more image pairs of additional randomly chosen FOVs to increase the number of available images that contain the target population to be quantified.

DISCUSSION

The approach described in this protocol quantifies the abundance of specific microbial groups in terms of biovolume fraction, which is a relative value. However, sometimes absolute cell numbers must be quantified, for example, to infer per-cell substrate turnover rates. Cells in planktonic samples often can be counted manually by fluorescence microscopy, but this can be extremely tedious or prohibitive with dense biofilms, where most microbes grow in tight aggregates. Automated counting by image analysis of such clustered cells often is not possible either, because the computer does not correctly recognize the densely packed individual cells. These problems can be overcome by extending the described quantification method, using an internal standard to derive absolute cell numbers from measured biovolume fractions (Daims et al. 2001).

REFERENCES

Amann RI. 1995. In situ identification of micro-organisms by whole cell hybridization with rRNA-targeted nucleic acid probes. In Molecular microbial ecology manual (eds. ADL Akkermans et al.), pp. 1–15. Dordrecht, The Netherlands, Kluwer Academic.

Amann RI, Binder BJ, Olson RJ, Chisholm SW, Devereux R, Stahl DA. 1990. Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Appl Environ Microbiol 56: 1919–1925.[Abstract/Free Full Text]

Daims H, Wagner M. 2007. Quantification of uncultured microorganisms by fluorescence microscopy and digital image analysis. Appl Microbiol Biotechnol 75: 237–248.[Medline]

Daims H, Brühl A, Amann R, Schleifer KH, Wagner M. 1999. The domain-specific probe EUB338 is insufficient for the detection of all Bacteria: Development and evaluation of a more comprehensive probe set. Syst Appl Microbiol 22: 434–444.[Medline]

Daims H, Ramsing NB, Schleifer KH, Wagner M. 2001. Cultivation-independent, semiautomatic determination of absolute bacterial cell numbers in environmental samples by fluorescence in situ hybridization. Appl Environ Microbiol 67: 5810–5818.[Abstract/Free Full Text]

Daims H, Stoecker K, Wagner M. 2005. Fluorescence in situ hybridization for the detection of prokaryotes. Molecular microbial ecology 213–239.

Daims H, Lücker S, Wagner M. 2006. daime, a novel image analysis program for microbial ecology and biofilm research. Environ Microbiol 8: 200O–213.

Glöckner FO, Amann R, Alfreider A, Pernthaler J, Psenner R, Trebesius K, Schleifer KH. 1996. An in situ hybridization protocol for detection and identification of planktonic bacteria. Syst Appl Microbiol 19: 403–406.

Pernthaler A, Pernthaler J, Amann R. 2002. Fluorescence in situ hybridization and catalyzed reporter deposition for the identification of marine bacteria. Appl Environ Microbiol 68: 3094–3101.[Abstract/Free Full Text]

Wagner M, Horn M, Daims H. 2003. Fluorescence in situ hybridisation for the identification and characterisation of prokaryotes. Curr Opin Microbiol 6: 302–309.[Medline]