Fluorescent dyes were assessed with regard to their ability to discriminate between viable and non-viable cells of hyperthermophilic archaea and bacteria. Using bis-(1,3-dibutylbarbituric acid) trimethine oxonol (DiBAC4(3)), a membrane potential-sensitive probe, a safe and rapid discrimination of viable cells was possible by fluorescence microscopy. Single viable individuals, identified by DiBAC4(3), were selectively isolated from mixtures of viable and dead cells by the use of a laser microscope (`optical tweezers') and grown in pure culture.
Hyperthermophilic archaea and bacteria with optimal growth temperatures above 80°C represent the organisms at the upper temperature border of life [1–3]. A new micromanipulation technique has been developed recently for the selective isolation of single cells of specific hyperthermophiles. Single cells, optically trapped by the use of a strongly focused infrared laser beam (`optical tweezers'), were separated from a mixed culture under visual control and grown in pure culture (here we use the term `selected cell cultivation' for this isolation method) . `Selected cell cultivations' can only be successful if the trapped cell is alive and divides after isolation. Therefore, the assessment of the physiological state of each single cell before trapping and isolation could be most helpful to enhance cultivation efficiency. A variety of fluorescent dyes are known as indicators for the detection of cell viability [5–8]. Different principles usually apply, frequently based on nucleic acid binding assays or on respiratory chain activity. Methods that rely on the combined effect of esterase activity and membrane integrity or the ability to maintain the transmembrane electrochemical potential are also employed .
The aims of this study were to develop a universal approach for the rapid discrimination of viable and non-viable hyperthermophiles using fluorescent dyes and the application of this approach in the isolation of living single cells by the use of a laser microscope .
2Materials and methods
Comparative studies were carried out with the bacteria Thermotoga maritima DSM 3109 and Aquifex pyrophilus DSM 6858 and the archaea Thermoproteus tenax DSM 2078, Pyrobaculum organotrophum DSM 4185, `Thermofilum librum' strain V24N, Desulfurococcus mucosus DSM 2161, `Pyrodictium sp.' strain SNP6, `Thermococcus sp.' strain RR2, `Thermococcus sp.' strain LPC5, Pyrococcus furiosus DSM 3638, Methanothermus fervidus DSM 2088, Methanococcus igneus DSM 5666 and Methanopyrus kandleri DSM 6324 . The organisms were from the culture collection of our institute and were cultivated as previously described (see ref. in ). For the experiments, cells from the exponential phase of growth were used. Dead cells were obtained by autoclaving exponentially growing cultures (20 min, 121°C, 200 kPa). These autoclaved cultures could no longer be successfully transferred (1% inoculation).
The esterase-substrate fluorescent probes 5(6)-carboxyfluorescein diacetate (CFDA; stock solution (s.s.) in ethanol), 2′,7′-dichlorohydrofluorescein diacetate (H2DCFDA; s.s. in ethanol), 5-(and-6)-sulfofluorescein diacetate sodium salt (SFDA; s.s. in ethanol/H2O, 60/40, v/v), calcein acetoxymethyl ester (calcein AM; s.s. in DMSO) and calcein blue acetoxymethyl ester (calcein blue AM; s.s. in DMSO) were used at a final concentration of 10 μM (from 10 mM stock solutions). CFDA was purchased from Sigma (St. Louis, MO, USA), H2DCFDA, SFDA, calcein AM and calcein blue AM from MoBiTec (Göttingen, Germany). Viable cells are fluorescent and can be discriminated from non-fluorescent dead cells.
The membrane potential-sensitive probes rhodamine 123 (s.s. in ethanol), bis-(1,3-dibutylbarbituric acid) trimethine oxonol (DiBAC4(3); s.s. in ethanol), 3,3′-dihexyloxacarbocyanine iodide (DiOC6(3); s.s. in DMSO), tetramethylrhodamine methyl ester perchlorate (TMR; s.s. in DMSO), WW 781 triethylammonium salt (WW 781; s.s. in ethanol), Di-8-ANEPPS (s.s. in ethanol) and bis-(3-propyl-5-oxoisoxazol-4-yl) pentamethine oxonol (oxonol VI; s.s. in ethanol) were used at a final concentration of 0.5 μM (from 0.5 mM stock solutions). Rhodamine 123 was obtained from Sigma, DiBAC4(3), DiOC6(3), TMR, WW 781, Di-8-ANEPPS and oxonol VI from MoBiTec. Dead cells are fluorescent and can be discriminated from non-fluorescent viable cells.
All staining procedures were carried out in an anaerobic chamber. The cultures and the probes were mixed thoroughly and examined by microscopic inspection using a phase-contrast microscope (Nikon Mikrophot EPI FL, Tokyo, Japan), equipped with an oil immersion objective (Plan 100/1.25) and a mercury lamp (maximum power: 100 W). The filter sets for CFDA, H2DCFDA, SFDA, calcein AM, rhodamine 123, DiBAC4(3) and DiOC6(3) were EX450, DM510 and BA520 (B-2A, Nikon), for calcein blue AM EX330–380, DM400 and BA420 (UV-2A, Nikon), for TMR, WW 781 and oxonol VI EX546, DM580 and BA590 (G-1B, Nikon) and for Di-8-ANEPPS EX436, DM455 and BA460 (BV-1A, Nikon).
Experiments for `selected cell cultivations' were carried out by the use of a laser microscope as described recently . For the discrimination between viable and dead cells in the isolation unit, the inverted microscope was equipped with a mercury lamp (maximum power: 100 W) and the fluorescein filter set EX450–490, DM510 and BA520 (Zeiss) in addition.
3Results and discussion
3.1Esterase-substrate fluorescent probes
In an attempt to discriminate between live and dead cells of hyperthermophiles, five esterase-substrates were tested at room temperature and at the physiological growth temperatures of Thermotoga maritima (85°C), `Thermococcus sp.' strain RR2 (80°C) and Pyrobaculum organotrophum (100°C). However, the experiments at high temperature were not successful, because the fluorescence signal of living cells was either not detectable or, in the case of CFDA, viable and dead cells showed the same fluorescence intensity. At room temperature, only calcein AM and calcein blue AM permitted differentiation between viable and non-viable cells of Thermotoga maritima and `Thermococcus sp.' strain RR2. Viable and inviable cells of Aquifex pyrophilus, Thermoproteus tenax, Pyrobaculum organotrophum, `Thermofilum librum', Desulfurococcus mucosus, `Pyrodictium sp.' strain SNP6 and Pyrococcus furiosus could not be clearly discriminated using calcein AM and calcein blue AM as probes. Our studies with esterase-substrate based assays indicate that these dyes are not suitable for the assessment of the physiological state in hyperthermophiles. Recently, similar results were reported for mesophilic bacteria, where fluorogenic esters were found to be unsuitable as a universal discrimination method .
3.2Membrane potential-sensitive probes
Seven membrane potential-sensitive probes were tested in Thermotoga maritima, `Thermococcus sp.' strain RR2 and Pyrobaculum organotrophum at room temperature (Table 1). With DiBAC4(3), a clear discrimination between viable and dead cells of these hyperthermophiles was possible (Table 1). Additional studies showed that DiBAC4(3) permits a clear distinction between viable and non-viable cells of Aquifex pyrophilus, Thermoproteus tenax, `Thermofilum librum', Desulfurococcus mucosus, `Pyrodictium sp.' strain SNP6, Pyrococcus furiosus, Methanococcus igneus, Methanopyrus kandleri and Methanothermus fervidus (Fig. 1). During our experiments, additional favorable properties of DiBAC4(3) became evident. Dead cells could be easily identified by an intensive, yellow-green fluorescence, which remained unbleached for at least 20 s. The staining method was rapid, because clear discrimination was possible immediately after adding the dye. Taking into account the phylogenetic and physiological diversity of the organisms used in this study, DiBAC4(3) can be considered to be a dye universally suitable for hyperthermophiles. Beside the detection of cell viability in cultures, this fluorescent dye could also be applied to study hyperthermophilic communities in nature. DiBAC4(3) in combination with flow cytometry could allow a rapid interrogation or sorting of individuals from populations and thus a fast enumeration and physical characterization of cells from high temperature ecosystems [9–11].
Table 1. Discrimination between viable and dead cells of hyperthermophiles using membrane potential-sensitive probes at room temperature
`Thermococcus sp.' isolate RR2
+: viable and dead cells can be easily discriminated based on fluorescence intensity; (+): viable and dead cells can hardly be discriminated; −: no discrimination possible; /: different results obtained after 30 min and 6 h incubation with the probe. aCationic dye. bAnionic dye. cNeutral dye.
In further experiments with DiBAC4(3) stained cultures, `selected cell cultivations' of Thermotoga maritima and of `Thermococcus sp.' LPC5 were possible, demonstrating that the cell viability was not affected by the membrane potential-sensitive probe and irradiation with light from a mercury lamp (100 W; wavelength 450–490 nm). Based on these results, we tried to isolate single viable cells with the `optical tweezers' from a 1:1 mixture of viable and dead cells of Thermotoga maritima, using DiBAC4(3) as the discriminating dye. Eight `selected cell cultivation' experiments were carried out with an efficiency of 100%, indicating that the application of DiBAC4(3) can significantly improve laser-based isolation technology. Recently, a novel strategy for the isolation of hyperthermophiles, predicted by 16S rRNA sequences, was developed and successfully applied . The use of DiBAC4(3) can complement this approach and contribute significantly in `selected cell cultivations' of novel hyperthermophiles either from cultures or directly from natural samples.
We wish to thank Dr. K.O. Stetter for stimulating discussions and for proposing the topic. Furthermore, we are grateful to C.F. Brunk for critically reading the manuscript. This work was supported by a grant of the European Community (EC Generic Project `Biotechnology of Extremophiles', Contract BIO-CT93-02734) and the Fonds der Chemischen Industrie.