Introduction

Halophiles are organisms that can thrive in extreme conditions of high salt concentrations. Most of these organisms belong to the archaea or bacteria life domains. Exposed to a hostile environment, halophiles evolved unique biomechanisms, which make them very exciting scientific objects. Probably, the most important is the example of bacteriorhodopsin – a photoactivatable retinal membrane protein with unique properties, including stability at high concentration of salt (up to 5M), protease resistance, tolerance to high temperatures (upto 140 °C when dry)1 and to a broad range of pH (at least 3–10)2. Bacteriorhodopsin has become the most studied membrane protein, often serving as a “model” for all other membrane proteins (see for instance3,4,5,6,7,8).

In recent decades, halophiles have attracted additional scientific interest due to their isolation from Earth’s subsurface halites, in some cases >250 million years old9,10,11,12,13,14,15,16,17,18,19,20. Being buried in brine-filled fluid inclusions, halophiles are able to survive for millennia under conditions of extremely high ionic strength, elevated temperatures and nutrients depletion. Genome studies showed that spore formation is not possible for halophilic archaea21, thus, their survival mechanisms remain unclear22.

Halites were discovered on the surface of Mars by the Opportunity and Spirit rovers and in meteorite samples23,24,25,26,27,28,29. It was shown that they can hypothetically play protective role against radiation30 or biochemical degradation19 and potentially harbour halobacteria31. Thus, sensitive detection of living halophiles in hypersaline medium with minimal influence on their viability is a question of great interest for further exploration of microbial life on Earth and elsewhere.

An additional interest associated with halophiles arises from their role in microbial communities of natural and man-made hypersaline environments32, especially in the case of hydrocarbonoclastic communities, which are potentially applicable for bioremediation of oil pollutions33.

Fluorescent staining can provide a sensitive method for detection of halophilic species (i.e. via flow cytometry) in environmental samples, and can also be used for their analysis via fluorescence microscopy. The main challenge in relation to the fluorescent staining of halophiles is the high ionic strength of the medium and in some cases hostile pH34 of their natural environment, which are inappropriate for conventional antibodies and dyes35. The low permeability of the halobacterial S-layer to macromolecules can additionally complicate immunostaining.

The list of staining approaches for halophiles is currently rather limited. Most of them require fixation (i.e. immunostaining36,37, DAPI38,39, and FISH38,40) or the use of DNA-targeted intercalating dyes (i.e. LIVE&DEAD kit35,39,41,42,43,44,45, Acridine Orange43,46, Hoechst47). The latter approach is unsafe due to interference with the cellular DNA processing machinery, which can result in complete loss of cell viability (for review see48,49). It has been shown that Acridine Orange and its derivatives can additionally stain archaeal phospholipids50. Several other approaches for staining halophiles that have been recently suggested44,51 remain limited to particular applications. Genetic modification of archaea with GFP for its fluorescent visualization44,52 is also worth mentioning, although it is not straightforward for detection purposes. Notably, no dyes for long-term staining without cytotoxic side-effects have been found; the field is dominated by DNA-staining dyes, which has cytotoxicity as an intrinsic limitation.

Here we suggest a new method for staining living halobacteria cells with fluorescent dyes in their growth medium. We show that dyes designed to specifically stain mitochondria in mammalian eukaryotic cells (MitoTracker Orange CMTMRos, MitoTracker Red CMXRos and MitoTracker Deep Red FM) easily permeate the S-layer of Hbt. salinarum and cause stable bright staining of the archaeon. Even without removal of the residual, non-incorporated MitoTracker dye, the signal-to-noise ratio (SNR) is sufficient for clear cell detection and fluorescence microscopy. To test our staining procedure we investigate the conversion of Hbt. salinarum cells into spheroplasts via exposure to an EDTA-containing solution. In the case of MitoTracker Orange CMTMRos we also show that it does not have an effect on the cell’s growth rate. The staining remains bright during long observations (hours to days) and is inherited in cell division. The new method of staining allowed us to detect the development of viable spheres of submicron size (herein called “microspheres” to highlight the difference between them and spheroplasts) from rod-shaped Hbt. salinarum after the exposure to EDTA, which was previously assumed to result only in spheroplast formation. In addition, we demonstrate the cytosolic integrity of giant Hbt. salinarum cells via fluorescence loss in the photobleaching (FLIP) approach. In addition, the dyes were successfully applied to three halophilic microorganisms isolated from environmental samples from hypersaline lakes representing both the bacterial and archaeal domains of life. Therefore, we propose that MitoTracker dyes can be used as an effective and non-harmful dyes for halophile staining.

Materials and Methods

Cell culture

The Hbt. salinarum strain S9 was used as test sample for halophilic microorganism. The growth medium contained 250 g of NaCl, 20 g of MgSO4·7H2O, 3 g of Na3C6H5O7·2H2O, 2 g of KCl and 10 g of peptone (Helicon, Russia, H-1906-0.5) (per 1 litre), at pH 6.5. Cells were routinely cultured in 100 mL flasks at 37 °C with shaking (200 rpm).

Sample preparation and staining

MitoTracker Green FM (M7514), MitoTracker Orange-CMTMRos (M7510), MitoTracker Red-CMXRos (M7512) and MitoTracker Deep Red FM (M22426) (Thermo Fisher Scientific, USA) were dissolved in DMSO to obtain 1 mM stock solutions. Solutions were stored at −20 °C.

To stain Hbt. salinarum (Fig. 1) or species isolated from environmental samples, MitoTracker solution was added to the growth medium at a 1:1000 dilution (to a final concentration of 1 μM).

Figure 1
figure 1

Living Hbt. salinarum cells in growth medium stained with MitoTracker Orange CMTMRos (A), MitoTracker Red CMXRos (B) and MitoTracker Deep Red FM (C).

Full size image

Confocal fluorescence microscopy

Measurements were performed in thin-bottom 8-well chambered cover glass (Thermo Fisher Scientific, USA, 155409), using an apochromate objective (420792-9800-720, Zeiss) with oil immersion (ISO 8036, Immersol 518F, Zeiss).

MitoTrackers were excited by 561 nm laser for MitoTrackers Orange-CMTMRos and Red-CMXRos, and by 633 nm laser for Deep Red FM. Emission was recorded in λ-mode with 34-channel QUASAR detector unit from Zeiss. All images were processed using Zen 2012 (Zeiss, Germany) and Fiji53 software. Supplementary Video 2 was drift corrected using a StackReg -plugin for Fiji54.

Growth rate measurements

Hbt. salinarum cells were diluted in growth medium to an optical density of approximately 0.05 at 600 nm (OD600nm) (with growth medium used for blank).  5 mL cell cultures were incubated in 15 mL flasks for a week at 37 °C with shaking (200 rpm).

MitoTracker Orange CMTMRos was added twice a day to the first flask (sample), to keep the ratio between number of bacteria and MitoTracker molecules approximately constant and to maintain conditions similar to those in the fluorescence microscopy experiments during the entire observation period, namely, 1 μM of dye per OD600nm unit. No MitoTracker Orange CMTMRos was added to the second flask, so it could serve as a control. The OD600nm was measured for eight days in each flask. Growth medium with and without MitoTracker were used as a reference for sample and control. Aliquots used for the OD600nm determination were returned to the respective flask after each measurement. The growth curves in Fig. 2 represent the average of 3 independent samples.

Figure 2
figure 2

Growth curves of stained and control cells of Hbt. salinarum. The cells were grown with (sample group; downward-pointing triangles) and without (control group; upward-pointing triangles) MitoTracker Orange CMTMRos.

Full size image

MitoTracker durability during cultivation

Hbt. salinarum cells were stained with MitoTracker Orange CMTMRos and after a 24-h incubation in a dye-containing medium (time point “a” in Fig. 3), the cells were sedimented via centrifugation (at 2000 g 37 °C for 5 min) and resuspended into MitoTracker-free growth medium to a final OD600nm of 0.2 (“b”, Fig. 3). Cells were cultured at growth conditions for two days (to OD600nm of 0.8; “d”, Fig. 3) and then diluted again to an OD600nm of 0.2 (time point “e”, Fig. 3). This way, cells always remained in the exponential growth phase. Each day a sample of the cell culture was taken and a fluorescence image was acquired. The dilution-growth cycle was repeated three times. The plot of OD600nm as a function of time from the start of the experiment as well as images obtained using fluorescence microscopy of the samples are shown in Fig. 3. Images were analyzed to estimate SNR. SNR was calculated for 20 cells for each time point as a ratio of difference in the mean signal in the cells and outside divided by the standard deviation of the signal outside the cells (see Fig. 3B).

Figure 3
figure 3

Durability of MitoTracker staining during several growth and dilution cycles. (A) Schematic plot of OD600nm change during the experiment: 2-day exponential growth periods correspond to incubation time, while a 4-fold drop corresponds to the dilution. Particular time-points are labelled with lower-case Latin letters (a–j). (B) SNR decrease in the experiment. (C) (a) Microphotograph of cell culture at the day one; (d) two days after staining; (g) four days after staining and (j) six days after staining. Lower-case Latin letters in B and C correspond to the particular measurement time, shown in (A). Halobacteria were stained with MitoTracker Orange CMTMRos. The dye was washed out via centrifugation after staining. Cell cultures were incubated in a dye-free growth medium for 6 days with dilution every 2 days to maintain the active growth phase. The intensity of excitation laser was the same for all images. The detector gain was equal for images d,g,j, but it was reduced for image a to avoid saturation. The variation of the background in each image is due to different distances from the focal plane to the surface of the cover glass. Contrast stretching was applied to achieve the highest contrast of each image.

Full size image

Conversion to spheroplasts

Spheroplasts were prepared as previously described46. Following 15 min of incubation in a staining solution cells in a 2 mL sample of culture (OD600nm 0.8–1.0) were sedimented via centrifugation (at 2000 g 37 °C for 5 min). The supernatant was removed by pipetting. Cells were resuspended in 150 μL spheroplast-forming solution (2 M NaCl, 27 mM KCl, 50 mM Tris hydrochloride pH 8.75, 15% sucrose, 15% glycerol) and transfered to one well of the thin-bottom 8-well chambered cover glass (Thermo Fisher Scientific, USA, 155409). The excitation laser was focused slightly above the bottom of the cover glass. The first measurement (time point 0) was recorded immediately after 15 μL of 0.5 M EDTA was added into the well. The obtained data are shown in Fig. 4 and Supplementary Video 1.

Figure 4
figure 4

Conversion of Hbt. salinarum cells stained with MitoTracker Orange CMTMRos into spheroplasts by EDTA treatment. The entire process lasts approximately 10 min: in the beginning the cells have a clear rod shape (A; t = 0 s). During the conversion, the short cells swell at one end (B, arrows; t = 200 s), while long cells first bend around their midpoint (C, arrows; t = 260 s). Finally, both types of cells form spheroplasts (D, t = 480 s). Cells with intermediate shapes (swollen or bent) are indicated with arrows.

Full size image

For a more detailed observation of the conversion process, the Hbt. salinarum cells were stained with MitoTracker Orange CMTMRos. Cells were concentrated by sedimentation (4000 g for 10 min at room temperature) and entrapped in a 2% agar solution (Helicon, Russia, H0102) with growth medium. The mixture was distributed on the bottom of microscopy dish to form a thin layer. After solidification the gel was washed 3 times with a spheroplast forming solution and then observed by fluorescence microscopy for up to 10 hours in an excess of spheroplast forming solution containing 50 mM EDTA (see Fig. 5 and Supplementary Video 2).

Figure 5
figure 5

Appearance of microspheres during conversion of Hbt. salinarum cells into spheroplasts by EDTA treatment. Hbt. salinarum cells were stained with MitoTracker Orange CMTMRos, entrapped in 2% agar and exposed to EDTA in a spheroplast-forming solution, as described in Materials and Methods. In addition to the spheroplasts formation, the generation of microspheres (marked by arrows) was observed. In contrast to rod-shaped cells and spheroplasts, microspheres are mobile in agar, but fluctuate only in proximity of spheroplasts originating from the same rod-shaped cells. This can be explained by fluctuation inside cavities formed in the agar after the contraction of the rod-shaped cells.

Full size image

Recovery of rod-shaped Hbt. salinarum from spheroplasts and microspheres

Hbt. salinarum cells were stained with MitoTracker Orange CMTMRos in the growth medium and then converted to spheroplasts as described above. Microspheres were isolated from the mixture by filtration through a 0.45-µm PTFE filter (4555, Pall Corporation, Russia). Sample purity was confirmed by fluorescent microscopy (see Supplementary Figure 2C). In total, 500 µL of sample containing filtered microspheres was mixed with 2 mL of rich growth medium (250 g/L NaCl, 2 g/L MgSO4·7H2O, 3 g/L KCl, 3 g/L Na3C6H5O7, 0.2 g/L CaCl2, 5.0 g/L tryptone (Serva, Germany, 48647.01), 2.0 g/L yeast extract (Organotechnie, France, 19512), 1.25 g/L glycerol, 1 mg/L ZnSO4·7H2O, 50 mg/L FeSO4·7H2O, 0.3 mg/L MnSO4·H2O, pH 7.0.) with 15% sucrose and incubated at 37 °C for 10 days in the dark.

Spheroplasts were isolated from the microsphere-containing mixture by centrifugation in a spheroplast forming solution at 20000 g for 20 min at room temperature. At these conditions, spheroplasts were dominantly sedimented, while microspheres remained in the supernatant. Centrifugation was repeated 4 times. Each time, the supernatant was discarded and the pellet was resuspended in fresh spheroplast forming solution. Examination by confocal fluorescence microscopy (see Supplementary Figure 2B) confirmed that both the supernatant and the resuspended pellet after the last centrifugation were free of microspheres. After the 4th centrifugation the pellet was resuspended in rich growth medium with 15% sucrose and incubated for 8 days.

After the incubation both samples were examined by DIC microscopy (Supplementary Figure 2D and E).

FLIP of giant rod Hbt. salinarum cells

After 15–20 days of incubation at growth conditions Hbt. salinarum cells were stained with MitoTracker Orange CMTMRos and then incubated for several (1–4) days without stirring at room temperature. Rod-shaped cells longer than 30 µm were observed by fluorescence microscopy (see Fig. 6 and Supplementary Video 4) and iteratively bleached (several seconds each cycle) using a high-intensity (100% power) 561-nm laser in the region (typically, 10 × 10 µm2) close to one of the rod’s ends, in intervals of approximately 30 s.

Figure 6
figure 6

Bleaching of stained giant rod Hbt. salinarum cells. The cytosol integrity of the cells (approximately 40 µm long in this case) was demonstrated using FLIP of MitoTracker Orange CMTMRos in Hbt. salinarum cells. The bleaching region is marked with a white square, the intensity of the fluorescence signal is shown as a heatmap (pseudocolour, intensity increases from purple to red according to the scale below image). The loss of fluorescence intensity along the entire length of the cell shows that the majority of the dye can freely diffuse along the cell, demonstrating the existence of one cellular compartment not separated by septal membranes.

Full size image

Isolates from environmental samples

Four samples of extremely halophilic microorganisms were isolated from various hypersaline lakes using standard microbiological techniques55 and then identified as follows: Halorubrum sp. – euryarchaeota from the Alikes salt lake (Kos island, Greece); Halomonas sp. and Haloferax sp. – proteobacteria and euryarchaeota from the Elton salt lake (Volgograd region, Russia), respectively; and Salicola sp. – proteobacteria from the Chott el Djerid salt lake (Tunisia).

The cultures were cultivated in a nutrient medium composed of 250 g/L NaCl, 20 g/L MgSO4·7H2O, 3 g/L KCl, 3 g/L Na3C6H5O7, 0.2 g/L CaCl2, 5.0 g/L tryptone (Serva, Germany, 48647.01), 2.0 g/L yeast extract (Organotechnie, France, 19512), 1.25 g/L glycerol, 1 mg/L ZnSO4·7H2O, 50 mg/L FeSO4·7H2O,0.3 mg/L MnSO4·H2O, pH 7.0. The medium was sterilized at 121 °C for 30 min.

Cultivation in liquid medium was carried out at 38.5 °C and 150 rpm using a Unimax 2010 orbital platform shaker (Heidolph, Germany) with illumination (Philips TL-D 18W/33–640 lamps) for 7 days in 100 mL conical flasks with 5 mL of inoculum added to each flask.

The halophilic microorganisms from the environmental samples were identified by comparative phylogenetic analysis of their 16S rRNA gene sequences. For this purpose DNA was extracted using the Wizard technique combined with the modified Birnboim-Doli method56. The concentration of DNA was 30–50 µg/mL, RNA was present in trace quantities (<1 %).

A universal primer system (Univ11f-Univ1492r for bacteria and 8fa-1492r for archaea; Evrogen, Russia) was used for PCR and further sequencing of PCR-fragments of 16S rRNA genes57,58. PCR products were purified by electrophoresis in a 1% agarose gel followed by extraction with the Wizard PCR Preps DNA Purification System (Promega, Madison, WI, USA) as described in the manufacturer’s protocol.

PCR-fragments of 16S rRNA genes were sequenced by Sanger’s method using a Big Dye Terminator v.3.1 reagent kit (Applied Biosystems, Inc., USA) on an ABI PRIZM 3730 system (Applied Biosystems, Inc., USA) according to standard instructions from the manufacturer. Sequence similarities were identified in BLAST59.

16S rRNA sequences were deposited to the GenBank database (MF148853 – Haloferax sp., KY781161 –Halorubrum sp., KY781162 – Halomonas sp., KY781163 – Salicola sp.).

Results and Discussion

Staining of Hbt. salinarum under growth conditions

To find an appropriate method to stain Hbt. salinarum we started with the idea that an important intrinsic property of a halobacterial cell is its negative membrane potential. This led us to try to stain halobacteria with MitoTracker cationic probes, which are normally used to stain mitochondria. It is the mitochondrial membrane potential which drives cationic MitoTrackers60 into organelles. Halobacteria have membrane potentials of the same sign and can potentially be stained using the same approach.

Indeed, our first experiments showed that living Hbt. salinarum cells can be stained in their growth medium using three MitoTracker dyes (MitoTracker Orange CMTMRos, MitoTracker Red-CMXRos, and MitoTracker Deep Red FM), as described in Materials and Methods and observed using a confocal fluorescence microscope (see Fig. 1). Fluorescence signal accumulated from cells is much higher than the background level (SNR >100) and it can be inferred that free dye is drawn into the cells. Apparently, MitoTrackers are driven by membrane potential in the same way as in mitochondria.

We have also tried staining halobacteria with MitoTracker Green FM, which has been shown to stain mitochondria regardless of membrane potential61 (data not shown). In this case, no accumulation of the dye in the cells was observed. This additionally supports our hypothesis that membrane potential plays a crucial role in the staining of halophiles with MitoTracker dyes.

Effects on growth rate and staining durability

To determine whether staining has a negative effect on growth rate, we studied cells stained with MitoTracker Orange CMTMRos (see Materials and Methods for details). Figure 2 shows the growth curve of the stained culture compared with control non-stained one. MitoTracker staining does not alter the normal cell growth rate (with doubling time in the range of 1.2 ± 0.3 days). Cell size and shape remain for at least three generations as observed with a fluorescence microscope (data not shown).

The absence of an effect of MitoTracker dyes on growth rate is highly desirable for a wide range of biological studies and is not shared by the most popular staining alternatives. LIVE&DEAD kit staining is the only approach that was systematically considered in terms of its applicability to living halophiles in their native environment35. One of the two examined cell lines (Halobacterium sp. NRC-1) exhibited a two-fold decline in CFU (colony forming units) in the presence of the staining reagent35. For the other line (H. dombrowskii H4), LIVE&DEAD staining has been shown to have almost no effect on cell viability35. However, in both cases cells incubated with dyes for long and short time periods were compared, and therefore, a rapid response occurring immediately after staining cannot be excluded and therefore can be a substantial experimental limitation.

Another beneficial property of MitoTracker staining is its durability. The staining persists during washing (removal of supernatant), dilution of stained cells in a MitoTracker-free medium and cultivation of cells (see Materials and Methods, MitoTracker durability during cultivation). As depicted in Fig. 3, the staining quality of Hbt. salinarum cells remains satisfactory and provides cell fluorescence well above background through at least 3 cycles of incubation/dilution.

Conversion of Hbt. salinarum to spheroplasts

To validate the staining quality, imaging of a well-known process of halobacterial cell wall removal by EDTA was carried out. In this process, rod-shaped halobacteria stained with MitoTracker Orange CMTMRos were converted to spherical particles called spheroplasts. After several minutes of incubation in a 10 mM EDTA-containing buffer, rod-shaped cells were converted to spheroplasts. The conversion process is depicted in Fig. 4 and Supplementary Video 1.

The shorter individual cells swell at one end (Fig. 4B) and the remaining rod-like part of the bacterium merges with this roundish end-structure (Supplementary Video 1). In this process, cells undergo the same conversion steps as previously described based on phase-contrast microscopy of Hbt. salinarum62. The longer cells bend around their midpoint in the beginning and then undergo the same process (Fig. 4C). The resulting spheroplasts are clearly visible with SNR >100. Each rod gives rise to one or two spheroplasts depending on its length. In previous studies concerned with osmotic conversions42,62,63 the average number of spheres produced (observed by phase contrast microscopy62,63, electron microscopy42 or fluorescence microscopy based on LIVE&DEAD staining42) varied from one to four, which provides evidence that different protocols (salinity change, pH change, EDTA, other cell culture conditions before conversion etc.) can result in different conversion physiology.

In addition to MitoTracker Orange CMTMRos, MitoTracker Red-CMXRos and MitoTracker Deep Red FM were used to study the conversion process, which resulted in similar observations. The quality of final spheroplast staining was equally good for all three MitoTracker dyes and is shown in Supplementary Figure 1.

Microspheres formed during spheroplast conversion

Recently, it was shown that rod-shaped Halobacterium species can convert to microspheres (approximately 0.4 µm in diameter) upon a decrease in water activity below 0.7542. The term water activity is widely used to define the availability of water for hydration of materials. A value of 1.0 indicates pure water, 0.75 approximately corresponds to 4 M NaCl buffer. One rod forms three to four microspheres that remain viable for years and proliferate into normal rods when supplied with proper nutrients. Microspheres with similar properties form in fluid inclusions in laboratory-grown halites as well as in salt deposits isolated from millions-year-old minerals. Presumably, microspheres represent a dormant form of haloarchaea with higher resilience, which allows them to survive in unfavourable environmental conditions22.

Surprisingly, the MitoTracker staining approach revealed similar microspheres to coexist with spheroplasts after the EDTA conversion (see Supplementary Figure 2A) according to the standard protocol described in Materials and Methods (Conversion to spheroplasts). To identify the origin of the microspheres, we recorded the conversion of Hbt. salinarum to spheroplasts, trapped in a 2% agarose gel (see Supplementary Video 2 and Fig. 5). Single microspheres are formed from some halobacteria while they lose their S-layer during spheroplast conversion. Restoration of rod-shaped Hbt. salinarum from spheroplast solution has been well described in previous studies and is widely used for Hbt. salinarum transfection46,64. Since microspheres were not previously observed in a spheroplast solution, the question arises whether restored rod-shaped Hbt. salinarum cells originate from microspheres, spheroplasts or both. To tackle this question we separated spheroplasts and microspheres as described in Materials and Methods (Recovery of rod-shaped Hbt. salinarum from spheroplasts and microspheres). Interestingly, after cultivation for 10 days in nutrient-rich medium, both samples restored rod-shaped halobacteria species (see Supplementary Figure 2D and E). This result implies that microspheres formed from halobacteria during spheroplast conversion possess all that is necessary for growing into normal rod-shaped halobacteria. Similar to the case of fluid inclusion, microspheres may serve as a halobacterial back-up to thrive in unfavourable conditions.

Giant rod Hbt. salinarum cells

Most Hbt. salinarum cells observed in our experiments are rod-shaped with an approximate length of 9 µm (see Supplementary Video 3). Мuch longer rod-shaped cells were occasionally found with a length of up to 45 µm. Notably, the length of a cell correlated with its motility: longer cells were less motile. The fraction of giant rod cells was higher when cells were incubated for a long time in a stationary phase without dilution (15–20 days). These observations are in good agreement with earlier data showing long and immobile Hbt. salinarum cells in surface-adherent biofilms (see Fig. 1 in65 and Fig. 5 in43).

Little is known about the morphology of giant rod halobacterial cells. We observed that long rod cells split into several smaller ones during spheroplast formation (Fig. 4 and in Supplementary Video 1). Thus, it could have been expected that giant rod cells consist of several smaller individual cells that remain concatenated after division. To test whether the giant rods are single cells and whether their cytosols are not divided into sub-compartments, we investigated the diffusion of MitoTracker dyes inside single giant rod cells. As described in greater detail in Materials and Methods (FLIP of the giant rod Hbt. salinarum cells), we bleached the MitoTracker dye on one end of the giant rod stepwise, and recorded the reduced fluorescence intensity along the whole cell (Fig. 6 and Supplementary Video 4). Notably, the fluorescent dye is not completely bleached using this approach, which shows that a fraction of the dye does not diffuse freely but remains attached to some cell compartments. The binding of the dye’s chloromethyl group to thiol groups of immobile intracellular (or membrane-associated) proteins is a likely explanation61.

Isolates from environmental samples

To check the general applicability of our MitoTracker dye staining approach for other halophile species we tested four strains of microorganisms, isolated from environmental samples from hypersaline lakes: Halorubrum sp. – euryarchaeota from the Alikes salt lake at the Kos island, Greece; Halomonas sp. and Haloferax sp. – proteobacteria and euryarchaeota from the Elton salt lake, Volgograd region, Russia; and Salicola sp. – proteobacteria from the Chott el Djerid salt lake, Tunisia. These selected halophile species belong to different domains of life: 2 are bacteria and 2 are archaea. The samples were stained with three different MitoTracker dyes, similar to Hbt. salinarum, and were examined using a confocal fluorescence microscope. In all cases, samples showed bright homogeneous staining of cells, as expected (see Fig. 7). This success provides evidence that the staining mechanism is general and not restricted to Hbt. salinarum or even to a single domain of life. Thus the suggested staining technique can be applied for the detection and analysis of yet unknown microorganisms in extremely halophilic environments or in extraterrestrial samples.

Figure 7
figure 7

MitoTracker dyes staining of halophilic microorganisms (Halomonas sp., Halorubrum sp., Haloferax sp., Salicola sp.) isolated from hypersaline lakes.

Full size image

Conclusions

MitoTracker dyes of three different colours (MitoTracker Orange CMTMRos, MitoTracker Red CMXRos, MitoTracker Deep Red FM) were shown to stain five halophilic microorganisms (three archaea and two bacteria) under growth conditions with high staining efficiency and low residual free dye level in the medium. The staining procedure does not require a washing step.

Staining quality of MitoTrackers is sufficiently good for studying the morphology of halophiles. The high SNR is suitable for morphological studies, cell counting and detection, which is important for future applications. Staining persists during the harsh spheroplasts conversion protocol, which implies that MitoTracker dyes can be used even when severe washing procedures are required. The MitoTracker dye staining does not affect growth rate and is inherited through several cell generations, making it a good choice for long-term studies, which require non-cytotoxic staining of live cells.

Our observations suggest that MitoTracker dyes are drawn into the cell by the membrane potential. Subsequently, they become anchored inside the bacteria via chloromethyl group binding to thiol groups of intracellular proteins as was previously described for MitoTrackers66.

Overall, MitoTracker dyes are an ideal choice for staining of halophiles and have many advantages compared with the previously applied staining techniques. Using this new approach, we demonstrated the formation of viable microspheres during Hbt. salinarum spheroplast conversion, and cytoplasmic continuity of giant rod Hbt. salinarum species.