In Situ Reverse Transcription-PCR for Monitoring Gene Expression in Individual Methanosarcina mazei S-6 Cells

Abstract

An in situ reverse transcription-PCR protocol for detecting specific mRNA in Methanosarcina mazei S-6 is described. This method allowed us to detect heat shock-induced increases in the intracellular levels of the transcript of the universal stress gene dnaK. The cell walls of paraformaldehyde-fixed cells were permeabilized by a thermal cycling procedure or by lysozyme treatment, and the cellular DNA was removed with DNase. The cells were subjected to a seminested reverse transcription-PCR protocol in which a digoxigenin-labeled primer was used. Detection of the reporter molecule was based on the 2-hydroxy-3-naphtoic acid-2′-phenylanilide phosphate–Fast Red detection system and binding of anti-digoxigenin–alkaline phosphatase conjugate. Fluorescence in permeabilized cells increased after a heat shock compared to fluorescence in non-heat-shocked cells, and the increase corresponded to an increase in the level of the dnaK transcript.

The pace of research to examine the molecular biology of methanogens has accelerated in recent years, but the techniques and methods used for anaerobic systems are not yet as advanced as the techniques and methods used for other microbial systems. This is in part due to the laborious routines that must be used when these strict anaerobes are handled. Shuttle and integration vector systems that have great potential have been described for the genus Methanosarcina (7, 18) and for the genus Methanococcus (4). Introduction of reporter gene systems, a technique based on the plasmids constructed by Metcalf and coworkers (18), is a technique that could be used to manipulate Methanosarcina strains. Reporter gene technology has also been used in promoter studies of Methanococcus maripaludis (3, 6). A system to study gene expression at the single-cell level, which does not require genetic manipulation, has recently been developed. In this in situ reverse transcription-PCR (in situ RT-PCR) technique, specific messengers are amplified by RT-PCR inside whole cells and are made available for detection either by hybridization with specific labeled probes or by direct incorporation of a label during the PCR. Fluorogenic substrates converted by antibody conjugate bound to a reporter molecule can be visualized with appropriate microscopic cameras and filters, and advanced software allows workers to determine the relative level of RNA inside an individual cell (8, 9, 12, 20). This principle has been applied to both bacterial and eukaryotic systems, but to our knowledge no archaeal in situ RT-PCR systems have been described until now. Transcription of the dnaK gene of Methanosarcina mazei S-6, which we have studied previously (5, 14, 15), is strongly induced under heat shock conditions, and we decided to use dnaK as a target gene for developing an in situ RT-PCR technique for Methanosarcina cells.

MATERIALS AND METHODS

Strains and growth conditions.

M. mazei S-6 was used as the test organism in this study. Cultures were grown at 37°C in 10 ml of BA medium as described by Angelidaki et al. (1), with the following exceptions: the medium was supplemented with yeast extract (2 g/liter), tryptone (2 g/liter), and trimethylamine (80 mM), and the vitamin solution was omitted. After autoclaving (140°C, 20 min), CaCl₂ · H₂O (0.5 g/liter), MgCl₂ · 6H₂O (1.0 g/liter), and Na₂S · 9H₂O (0.25 g/liter) were added from separate sterile solutions. Exponentially growing cultures were used as inocula (10%, vol/vol).

Cell fixation.

After approximately 36 h of incubation, the cells were in the early exponential phase. Cells were fixed either after no treatment (controls) or after a heat shock treatment consisting of 45°C for 30 min. Cells were harvested by centrifugation (4,000 × g, 2 min) and fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS) buffer (10× PBS contained [per liter] 80 g of NaCl, 2 g of KCl, 11.5 g of Na₂HPO₄ · 7H₂O, and 2 g of KH₂PO₄; pH 7.2) at 4°C overnight. The cells were washed twice in 1× PBS buffer and then processed immediately.

Cell wall permeabilization.

The following different strategies were used to permeabilize cell walls after fixation: (i) treatment with different concentrations of dimethyl sulfoxide (DMSO) (0, 0.1, 0.5, and 1.0%, vol/vol) in PBS buffer on ice for 15 min, followed by three washes in cold PBS buffer; (ii) resuspension in 1× PBS buffer and five cycles consisting of 1 min at 94°C and 1 min at 4°C (or on ice); (iii) incubation in saponin (0.1% [wt/vol] in PBS buffer) and sodium azide (0.1%, wt/vol) twice (15 min each) (throughout the rest of the procedure saponin [0.1%, wt/vol] was added to all solutions); (iv) protoplasting by incubation in 0.85% sucrose for 15 min on ice and two subsequent washes in PBS buffer; and (v) treatment with 0.01, 0.1, 0.5, 1.0, and 3.0 mg of lysozyme per ml in lysozyme buffer (100 mM Tris [pH 7.5], 50 mM EDTA). The samples were incubated for 30 min at room temperature and washed three times in PBS buffer.

DNase treatment.

Cells were subjected to DNase treatment in order to avoid PCR amplification of the chromosomal DNA. Cells were resuspended in 100 μl of DNase buffer (40 mM Tris, 6 mM MgCl₂; pH 7.5) after permeabilization and were digested for 30 min at 37°C with 20 U of RNase-free DNase I (Amersham Pharmacia Biotech GmbH, Freiburg, Germany). The reaction was stopped by incubation at 80°C for 5 min, followed by two washes in PBS. To confirm that this treatment removed template DNA, the PCR products originating from cells permeabilized and treated with DNase prior to either an amplification procedure that included an RT step followed by PCR or a standard PCR amplification procedure were visualized in stained agarose gels as described below.

Oligonucleotides.

The following three oligonucleotide primers that were specific for the dnaK gene of M. mazei S-6 (17) were designed and used for a seminested analysis (20): dnaKf (positions 516 to 535 relative to the start codon; sense orientation; 5′-TGGAGGCGGAACCTTCGATG-3′), dnaKr (positions 1248 to 1267 relative to the start codon; antisense orientation; 5′-GGACTCCTGCCTGAATTGCTGC-3′), and dnaKi (positions 1047 to 1066 relative to the start codon; antisense orientation; 5′-TTTACCTCTCCGCCCAGGACTC-3′). dnaKi was 5′ labeled with either biotin or digoxigenin (DIG).

Standard PCR.

DNA was extracted from M. mazei S-6 by using a Wizard genomic DNA purification kit (Promega, Madison, Wis.). The two DNA fragments (dnaKf-dnaKr and dnaKf-dnaKi) were amplified with a Progene thermal cycler (Techne, Cambridge, United Kingdom) by using the following thermal cycling program: initial denaturation at 94°C for 60 s; 25 cycles consisting of denaturation at 94°C for 30 s, annealing at 60°C for 45 s, and amplification at 72°C for 45 s; and a final a 7-min hold step at 72°C in a standard PCR (2). The identity of the PCR product was confirmed by sequencing (results not shown) performed with a model ABI377 automated sequencer (P.E. Biosystems) by using the reverse PCR primer as the sequencing primer, a Dye ET Terminator sequencing kit (Amersham-Pharmacia) and the methods recommended by the manufacturer.

In situ PCR procedure.

A GeneAmp EZ rTth RNA PCR kit from Perkin-Elmer (Birkerød, Denmark) was used to amplify intracellular dnaK mRNA. Pelleted cells were resuspended in 50 μl of the PCR mixture, which consisted of EZ buffer (25 mM Bicine, 115 mM potassium acetate, 8% [wt/vol] glycerol; pH 8.2), 600 μM dATP, 600 μM dGTP, 600 μM dCTP, 600 μM dTTP, 1 μM forward primer, 1 μM reverse primer, 4 mM manganese acetate, and 5 U of rTth DNA polymerase. The seminested protocol used was adapted from the protocol described by Tolker-Nielsen et al. (20). The first step was a one-tube RT, followed by PCR amplification that was performed by using primers dnaKf and dnaKr and the following temperature profile: 42°C for 5 min, ramp from 42 to 60°C over a 10-min period, 60°C for 30 min, 20 cycles consisting of 30 s at 94°C and 2 min at 60°C, and then finally 7 min at 60°C. The cells were harvested by centrifugation at 8,000 × g for 2 min and resuspended in fresh PCR solution. For the second PCR step primers dnaKf and dnaKi were used with the following temperature program: 94°C for 45 s, followed by 5, 10, 15, or 20 cycles (see below) consisting of 94°C for 30 s and 60°C for 2 min, and finally a 7-min hold at 60°C. In the initial PCR analysis only one PCR round was performed; the number of cycles was 30, and primers dnaKf and dnaKi were used.

Analysis of PCR products.

The PCR products obtained in the standard PCR and the PCR products present in the supernatant after the first and second PCR cycles of the in situ procedure were electrophoresed in 1.2% agarose-TAE gels at 40 V. The gels were stained with ethidium bromide, visualized under UV light (MacroVue UV-25; Hoefer Pharmacia Biotech, San Francisco, Calif.), and photographed with a Polaroid MP4+ instant camera system.

Detection of in situ PCR products.

Two different systems were used to detect the in situ PCR products, depending on the label used for primer dnaKi.

For biotin detection the Tyramidine Signal Amplification (TSA) direct system (DuPont, Nen Research Products) was used. Cells were washed twice in 100 μl of wash buffer (100 mM Tris [pH 7.5], 150 mM NaCl) and once in 100 μl of TBN buffer (5 mg of blocking reagent per ml in wash buffer) and were incubated for 30 min in TBN buffer at room temperature. The cells were resuspended in 100 μl of a horseradish peroxidase-conjugated antifluorescein antibody solution (1:50 in TBN buffer) and incubated for 2 h at room temperature in the dark. The cells were washed twice in TNT buffer (0.05% Tween 20 in wash buffer), resuspended in 100 μl of a tetramethylrhodamine-tyramide solution (0.05 mg of tetramethylrhodamine-tyramide per ml in DMSO diluted 1:50 with 1× amplification diluent [DuPont, Nen Research Products]), incubated for 5 min in the dark, and washed twice in TNT buffer.

For DIG detection we used a 2-hydroxy-3-naphtoic acid-2′-phenylanilide phosphate (HNPP) fluorescent detection set (Boehringer Mannheim GmbH, Mannheim, Germany). Following a PCR, cells were washed twice in blocking buffer 1 (0.5% [wt/vol] blocking reagent [Boehringer Mannheim] in buffer [100 mM Tris-HCl, 250 mM NaCl; pH 7.2]) and then incubated for 1 h in antibody solution (anti-DIG–alkaline phosphatase conjugate diluted 1:500 in blocking buffer). The cells were washed three times with wash buffer 1 (0.05% Tween 20 in buffer 1) and twice with detection buffer (100 mM Tris-HCl, 100 mM NaCl, 10 mM MgCl₂; pH 8.0). Five milligrams of Fast Red TR (Boehringer Mannheim) was dissolved in 200 μl of MilliQ H₂O. Ten microliters of the resulting Fast Red TR solution was mixed with 10 μl of HNPP in 1 ml of detection buffer. The mixture was passed through a 0.2-μm-pore-size syringe filter before it was used. Pelleted cells were resuspended in 100 μl of the HNPP-Fast Red TR mixture and incubated at room temperature in the dark for 30 min. The cells were finally washed twice with wash buffer.

H₂O₂ treatment.

Samples were treated with H₂O₂ to quench endogenous peroxidase activity when we tested the TSA direct kit. Cells were incubated in 0.9 or 3% H₂O₂ in 10% methanol in PBS for 15, 60, and 120 min. The cells were washed twice in PBS before they were subjected to TSA direct detection.

Microscopic examination.

Microscopic examinations were performed as previously described (20) with an Axioplan microscope (Carl Zeiss, Oberkochen, Germany). A type Ph3 Plan-NEOFLUAR 63×/1.25 oil objective (Carl Zeiss) was used for phase-contrast microscopy, and a Plan-NEOFLUAR 63×/1.25 oil objective (Carl Zeiss) was used for epifluorescence microscopy. A slow-scan charged-coupled device was mounted on the microscope and used to capture the images digitally. Fluorescence was visualized with no. 15 a filter set (type BP546/12 excitation filter, type 580 dichronic filter, and type LP 590 emission filter; Carl Zeiss).

Solutions.

All of the solutions used for RNA analysis were prepared by using diethyl pyrocarbonate-treated MilliQ H₂O.

RESULTS AND DISCUSSION

An in situ RT-PCR system for M. mazei S-6, which was capable of detecting differences at the dnaK transcript level in heat-shocked and non-heat-shocked cells, was designed. Transcripts of dnaK were amplified by using DIG-labeled primers and subsequently were detected by binding of anti-DIG alkaline phosphatase, which was visualized by using the HNPP fluorescence detection system (Boehringer Mannheim).

M. mazei S-6 was harvested and fixed in 4% paraformaldehyde in PBS overnight at 4°C. This treatment made the cells sufficiently resistant to withstand the thermal cycling procedure in the PCR and allowed them to maintain a coccoid shape. Appropriate permeabilization of the cell membrane is a key part of the in situ RT-PCR technique.

Single Methanosarcina cells are fragile and lyse readily in low-ionic-strength solutions. The cell envelope of these cells is very different from the cell envelopes of bacteria (13). Therefore, it was necessary to examine different methods for permeabilizing the cells. The simplest permeabilization technique tested was a thermal cycling procedure that consisted of five shifts between 94 and 2 to 4°C, accomplished with a thermal cycler or a heating block and an ice bath. Significant differences in the amounts of PCR product were observed in response to heat shock after both the first and second rounds of thermal cycling, when a fraction of the PCR supernatant was electrophoresed in a gel (Fig. 1). Without thermal cycling there was no difference in the amounts of transcript on the stained gels. The heat-cold permeabilization treatment also resulted in a difference in fluorescence between the heat shock-induced cells and the cells that were not heat shocked, when the populations were analyzed microscopically (Fig. 2). The heat cycling permeabilization technique, however, was found to be sensitive to variations in growth and possibly cell surface and was not generally applicable. Therefore, we decided to search for a permeabilization method which resulted in a higher level of reproducibility. It has been suggested that DMSO is a universal permeabilizing agent, and this compound worked well with Trichodesmium cells (16). DMSO treatment did indeed facilitate transport across the cell envelope, as judged by the amounts of PCR product in the reaction supernatants (Fig. 1), but the intensity of the signal obtained for the induced and noninduced cells was not satisfactory. Saponin permeabilization, protoplasting, and lysozyme treatments were also tested. The first two methods did not result in any signal in the cell populations and were not used subsequently. Treatment with lysozyme, which is used at low concentrations to permeabilize gram-negative bacteria (8, 10, 20), gave very encouraging results. The following five lysozyme concentrations were tested: 0.01, 0.1, 0.5, 1.0, and 3.0 mg/ml. The signal intensity obtained with cells treated with 0.5 mg of lysozyme per ml for 30 min at room temperature was substantially (5- to 10-fold) greater than the signal intensity obtained with cells treated with the other concentrations of lysozyme tested (data not shown). Signals, as determined by supernatant analyses, were also strongly dependent on the lysozyme concentration (Fig. 3). Lysozyme is rarely used with methanogens due to the chemical composition of the cell walls of these organisms (11), but the effect of lysozyme described here should be examined in other methanogenic systems.

FIG. 1.

Analysis of supernatant fractions by using DMSO, heat-cold cycling, or no treatment for permeabilization. Ten-microliter portions of the supernatants from PCR mixtures were electrophoresed in an agarose gel and stained with ethidium bromide. (A) Supernatants from the first round of PCR. (B) Supernatants from the second round of PCR. The treatments used are indicated below the lane numbers. Abbreviations: H.S., heat shock at 45°C for 30 min; PCR1, first round of PCR (performed with primers dnaKf and dnaKr), including an RT step and 20 cycles of standard amplification; PCR2, second round of PCR (performed with primers dnaKf and dnaKi) with 5 or 20 cycles as indicated (for the samples in lanes 15 through 17 RT was included); Perm., type of permeabilization treatment (D, DMSO treatment; T, heat-cold treatment; N, no permeabilization treatment); +, used; −, not used.

FIG. 2.

Detection of dnaK in M. mazei S-6 cells. Cells were permeabilized by thermal cycling and subjected to RT-PCR and a seminested PCR by using 10 cycles in the second round. See text for details. (A and C) Phase-contrast photomicrographs. (B and D) Epifluorescence photomicrographs. The cells in panels A and B were heat shocked for 30 min at 45°C.

FIG. 3.

Cell permeabilization with lysozyme. Fixed cells were treated with lysozyme at concentrations of 0.01 mg/ml (lanes 1 through 4), 0.1 mg/ml (lanes 5 through 8), 0.5 mg/ml (lanes 9 through 12), 1.0 mg/ml (lanes 13 through 16), or 3.0 mg/ml (lanes 17 through 20). Ten-microliter portions of supernatants obtained from the first round of PCR (lanes 1, 2, 5, 6, 9, 10, 13, 14, 17, and 18) or the second round of PCR (lanes 3, 4, 7, 8, 11, 12, 15, 16, 19, and 20) were electrophoresed in an agarose gel. Odd lanes contained samples from cells that had not been heat shocked. Even lanes contained samples from cells that had been heat shocked at 45°C for 30 min.

To assess the level of PCR product that genuinely originated from the RT-produced cDNA, the fixed cells were treated with DNase. The sufficiency of this treatment was tested, and no PCR product was formed during a standard PCR performed with DNase-digested cells (Fig. 4). DNase treatment of the cells resulted in better differentiation between stressed and nonstressed cells, as determined by an analysis of the supernatant fractions (Fig. 1) when it was assumed that dnaK is supposed to be overexpressed under heat shock conditions. When the DNase treatment was omitted, the level of PCR amplification from chromosomal DNA was great and minimized the differences that resulted from different mRNA levels. We found that the DNase step was necessary to distinguish between induced and noninduced cells, which was consistent with the results obtained with bacterial populations (8, 9).

FIG. 4.

Cells of M. mazei were fixed and treated with DNase as described in the text. Ten microliters of the PCR product was visualized in a stained agarose gel; all lanes are from same gel, and irrelevant lanes were removed. Lanes 1 and 2 are replicates, and lanes 3 and 4 are replicates. After DNase treatment the cells in lanes 1 and 2 were subjected to RT followed by PCR; the RT step was omitted for the cells in lanes 3 and 4. Lane 5 contained a no-polymerase control, which allowed us to judge background staining.

We found that decreasing the number of cycles in the second PCR round from 20 to 10 was advantageous when we wanted to detect differences in the transcript levels. When 20 cycles were used, the difference between induced and noninduced cells seemed to disappear (Fig. 1). Less than 10 cycles resulted in faint signals for cell populations, both induced and noninduced, that were examined microscopically (data not shown). When we used a single primer set (primers dnaKf and dnaKi) and a single PCR step (but a larger number of amplification cycles), a higher background signal was obtained. We assumed that the PCR products produced in the first step blocked some potential sites prone to nonspecific binding. Hodson et al. (8) and Tolker-Nielsen et al. (20) found that the seminested approach reduced formation of nonspecific products, and although only a single, very distinct band was formed under standard PCR conditions with each of the primer pairs (primers dnaKf and dnaKr and primers dnaKi and dnaKr) and DNA template, the seminested approach may reduce critical nonspecific amplification in the second round, in which the labeled primer is introduced.

In order to obtain the most sensitive system possible, biotin labeling and DIG labeling were compared. Biotin was detected with a TSA direct kit (DuPont). However, this approach could not be used with M. mazei S-6 as the background signal was high. High levels of endogenous peroxidase activity can cause a high background value. Incubation in the presence of 3, 1, and 0.3% H₂O₂ in methanol for 10, 15, and 30 min, respectively, before immunostaining is recommended by the manufacturer in order to circumvent this problem. We used the highest recommended concentration of H₂O₂ and increased the incubation time, but even this treatment did not decrease the background value noticeably. Some background signal was also present when no biotin-labeled primer was added. Extended blocking did not result in decreased background values. We believe that nonspecific binding of the antibody or the fluorescent tyramide to the cell envelope or to other cell components causes the problem. Therefore, we concluded that the TSA direct system cannot be used for the organism which we studied.

DIG labeling proved to be superior for studies of Methanosarcina cells. The HNPP-Fast Red system (Boehringer Mannheim) gave specific signals. By incubating preparations with and without labeled primer, we determined that the background signal of the HNPP-Fast Red system was acceptably low.

A seminested PCR with 10 PCR cycles in the second round performed with DIG-labeled primers and detection of the reporter molecule with the HNPP-Fast Red system (Boehringer Mannheim) seemed to provide the most sensitive response in our system. Using this protocol, we identified clear and reproducible differences in the signals of stressed and nonstressed cells. The stressed cells exhibited greater fluorescence, indicating that dnaK transcription increased under these conditions. This is consistent with the results of previous studies in which a heat shock resulted in an increase in the level of the dnaK transcript, as determined by Northern and mRNA slot blotting (5, 14). However, the fact that the difference in signal strength between stressed and nonstressed cells depends on the treatment and number of cycles used in the PCR means that controls must be used carefully.

Using in situ PCR techniques along with identification of organisms by fluorescent in situ hybridization techniques provides a way to specifically determine which organisms express which genes under certain conditions. In order to be able to examine communities, internal standards must be developed, and cell permeabilization procedures probably will have to be tailored to each community investigated. In contrast to traditional mRNA analysis techniques (e.g., Northern blotting), in situ RT-PCR may reveal heterogeneous gene expression in a microbial population (9, 19), and this technique may provide a more detailed picture of the physiological state of a population. In conclusion, the applicability of the method described here to M. mazei S-6 indicates that this method should be useful for studies of microbial populations and their functions.