FIGURE 15.3 Schematic illustration of combined FISH and microautoradiography. Refer to text for explanation.

For FISH-MAR, a sample is incubated with a radioactive tracer as described above, but the fixed and sectioned sample is hybridized to fluorescent rRNA-targeted probes before the film emulsion is applied. The in situ hybridization requires that the sample is dehydrated with ethanol in order to make the cytoplasmic membranes of microbial cells permeable for the oligonucleotide probes. During the hybridization and the subsequent washing steps, soluble radioactive tracer that has been imported into the cells without incorporation into macromolecules is washed out of the cells through the perforated cytoplasmic membranes and is therefore not detectable by MAR. Therefore, FISH-MAR is restricted to substrates that are incorporated into biomass. Labeled substrate molecules, which adhere to the outer surfaces of microbial cells although they are not used by the microorganisms, can lead to false-positive results of FISH-MAR. The misinterpretation of MAR signals, which are caused by adhesion and not by the uptake and incorporation of a substrate, can be avoided by negative control experiments with pasteurized sample material.70

FISH-MAR has been used to study, for example, the in situ physiology of Thiothrix species involved in activated sludge bulking,71 the uptake of lipids by the filamentous organism Microthrix parvicella,72 and the use of inorganic and organic carbon sources by uncultured Nitrospira-like nitrite-oxidizing bacteria in activated sludge and biofilm.31 FISH-MAR has also revealed that marine planktonic archaea take up amino acids from ocean water.73 A recently developed extension of FISH-MAR allows quantifying the incorporated amount of labeled substrate and has been applied to measure the kinetics of acetate incorporation by filamentous bacteria in activated sludge.74

Stable-isotope probing (SIP) of DNA75,76 uses stable isotope tracers to investigate the substrate usage of environmental microorganisms. The rationale of this approach is that bacteria, which can use a particular 13C-labeled carbon source, will incorporate a fraction of the imported 13C atoms into nucleic acids. An environmental sample is incubated with a 13C-labeled substrate and DNA is extracted from the sample afterwards. The 13C-labeled DNA ("heavy DNA") is separated from unlabeled 12C ("light") DNA by density gradient centrifugation. The heavy DNA

fraction is used as template for 16S rRNA-specific PCR, and the amplified 16S rRNA genes are cloned and sequenced. The obtained sequences are subjected to phylo-genetic analyses in order to identify the organisms that used the labeled substrate. This technique is applicable on a wide spectrum of environmental samples, but relatively long incubation times and high tracer concentrations are required for sufficient stable-isotope labeling of DNA. During the incubation, bacteria that can use the 13C-labeled substrate will most likely incorporate a fraction of the 13 C label into waste compounds, which are excreted and may be metabolized by other bacteria that could not directly use the original labeled substrate ("cross-feeding"). The addition of relatively large amounts of labeled substrates may cause an enrichment of organisms that can use these substrates. Cross-feeding between different bacterial populations and shifts of the microbial population structure become more likely during prolonged incubation periods and can falsify the results of SIP. Manefield et al.77 demonstrated that shorter incubation times are possible if ribosomal RNA instead of DNA is extracted and analyzed. The stable isotope is incorporated into rRNA whenever the cell produces new ribosomes, while DNA is labeled only during genome replication prior to cell division. However, both variants of SIP require PCR-amplification and cloning of rRNA genes. If the heavy nucleic acid fraction is contaminated prior to PCR with light nucleic acid, the rRNA genes of bacteria that used the labeled substrate are amplified together with the rRNA genes of bacteria that did not use it. These genes cannot be distinguished during the following phylo-genetic analyses. Unfortunately, such contamination can never be ruled out. We have in collaboration with Linda Blackall at the University of Queensland demonstrated that SIP can be combined with FISH-MAR in order to overcome some drawbacks of these techniques.78 In the first step of the combined approach, SIP is applied on an environmental sample. New, highly specific rRNA-targeted probes are then designed based on 16S rRNA gene sequences, which were retrieved from the heavy nucleic acid fraction. These new probes are used to identify their target organisms in FISH-MAR experiments with the same sample and substrate that was used for SIP, but for FISH-MAR the substrate is radioactively labeled. This combination of methods is particularly useful for studying phylogenetically heterogeneous microbial guilds. Such functional groups of organisms are difficult to analyze by FISH-MAR if one has no clue which of their members occur in an environmental sample, and which rRNA-targeted probes should be used to detect these organisms in situ. The combined approach solves this problem because candidate organisms that may use the offered substrates are identified by SIP. Therefore, one must perform only a few FISH-MAR experiments with specific probes that target these organisms. FISH-MAR is independent from PCR, works with shorter incubation times than SIP, and therefore is used to verify the results that were obtained by SIP. However, this combination of SIP and FISH-MAR is restricted to environmental samples that are amenable to FISH.

Recently, the high sensitivity and substrate specificity of isotope techniques have been combined with the high degree of parallelization, which is offered by DNA microarrays. The isotope array.79 contains specific rRNA-targeted probes. Following the incubation of an environmental sample with a radioactive substrate, rRNA is extracted from the sample and is labeled with a fluorescent dye. The rRNA

is then hybridized to the microarray, which afterward is scanned for fluorescence and for radioactivity. The pattern of fluorescing spots on the array shows which probe-target organisms are present in the sample, while radioactive probe spots indicate which of these organisms incorporated the radioactive tracer into their ribosomal RNA.

DNA microarrays that target functional genes are promising tools for monitoring specific physiological groups of microorganisms. In a pioneering study, Wu et al.80 designed and evaluated a microarray, which detected some functional genes of nitrification, denitrification, and methane oxidation. Dennis et al.81 applied another functional gene array to monitor the induction of genes coding for the degradation of resin acid. Further development of DNA microarray technology will most likely improve the specificity and the sensitivity of microarrays, and will eventually make it possible to quantitatively analyze, with the efficiency of a high-throughput technique, the expression patterns of numerous functional genes in complex environmental samples.

One could assume that combining certain fluorescent life stains and FISH with rRNA-targeted probes makes it possible to detect and identify metabolically active cells in situ in environmental samples. Tetrazolium dyes, for example, are reduced and form insoluble formazan crystals in living cells. Staining with the tetrazolium dye 5-cyano-2,3-tolyl-tetrazolium chloride (CTC)82 can be combined with FISH because the fluorescent formazan crystals remain in the cells during the fixation and hybridization steps.83 Unfortunately, CTC and other life stains fail to detect all active bacterial cells in environmental samples.83,84

The ribosome content of microbial cells can also act as an indicator for metabolic activity. Fluctuations of the cellular ribosome content are detected by measuring and comparing the fluorescence intensity of cells stained by FISH with rRNA-targeted probes. This method is suitable for monitoring the physiological state of bacterial species like Escherichia coli, which show a correlation of ribosome content and growth rate.85 However, in marine Synechococcus cells, such a correlation exists only at medium growth rates.86 Interestingly, at the highest growth rates, the Synechococcus cells contained fewer ribosomes than at lower growth rates. Chemo-lithoautotrophic ammonia-oxidizing bacteria maintain a high ribosome content even during starvation and when ammonia oxidation is chemically inhibited.29,63 Thus, the ribosome content should not be used to measure metabolic activity unless a clear correlation between activity and ribosome content has been demonstrated for the organisms being studied. This can be achieved by using FISH-MAR, which allows monitoring activity and ribosome content simultaneously. However, for this purpose, at least one substrate must be known that will be incorporated by most active cells of the investigated bacterial population.

Primary transcripts (precursors) of ribosomal RNA, which are transient intermediates of ribosome synthesis, seem to be better indicators for bacterial metabolic activity than mature ribosomes. The bacterial 16S, 23S, and 5S rRNA genes are separated by intergenic spacer regions. These spacers are transcribed together with the rRNA genes, but are degraded when the primary transcript is processed during ribosome maturation. Cangelosi and Brabant87 designed oligonucleotide probes, which targeted the spacer regions between the rRNA genes of Escherichia coli, and applied these probes in FISH experiments with pure cultures of E. coli. Probe-conferred fluorescence after FISH with the spacer-targeted probes indicated that the stained cells contained rRNA precursors and had been producing new ribosomes prior to fixation for FISH. The fluorescence intensity mirrored the intracellular concentration of rRNA precursors and the degree of ribosome-producing activity. These experiments with E. coli showed that fluctuations of the precursor rRNA level correlated better with changes of the growth phase than the less pronounced shifts of the mature ribosome content. However, the magnitude of the fluctuations in the precursor rRNA pool depended on the kind of the applied growth-limiting conditions.87 Thus, the results of FISH with spacer-targeted probes must be carefully interpreted in order to determine how different stress factors affect the metabolic activity of bacteria.

Spacer-targeted probes do not stain inactive bacterial cells because the rRNA precursors have been degraded when the cells fall into dormancy. These probes also allow observing relatively rapid activity changes. Schmid et al.88 monitored, by using spacer-targeted probes, a rapid decrease of the activity of anaerobic ammonium-oxidizing bacteria while these organisms were being exposed to oxygen. Oerther et al.89 designed spacer-targeted probes specific for Acinetobacter and observed that the activity of Acinetobacter cells in activated sludge increased while the sludge was being mixed with wastewater that was high in nutrients. Probes targeting the ribosomal RNA can be used together with spacer-targeted probes in the same FISH experiment, and the same image analysis methods can be applied to quantify cells detected by spacer-targeted probes and cells detected by rRNA-targeted probes. However, the design of spacer-targeted probes, which cover larger phylogenetic groups, is more difficult because the nucleotide sequences of the intergenic spacers are less conserved than the sequences of the ribosomal RNAs. A spacer-targeted probe may already fail to detect some strains of one target species due to strain-specific sequence differences at the probe-binding site. On the other hand, the lower degree of sequence conservation in the spacer regions allows developing strain-specific probes, which make it possible to monitor the activity of very closely related bacteria in the same environmental sample.

0 0

Post a comment