Recognition of Individual Genes in Diverse Microorganisms by Cycling Primed In Situ Amplification

Abstract

Cycling primed in situ amplification-fluorescent in situ hybridization (CPRINS-FISH) was developed to recognize individual genes in a single bacterial cell. In CPRINS, the amplicon was long single-stranded DNA and thus retained within the permeabilized microbial cells. FISH with a multiply labeled fluorescent probe set enabled significant reduction in nonspecific background while maintaining high fluorescence signals of target bacteria. The ampicillin resistance gene in Escherichia coli, chloramphenicol acetyltransferase gene in different gram-negative strains, and RNA polymerase sigma factor (rpoD) gene in Aeromonas spp. could be detected under identical permeabilization conditions. After concentration of environmental freshwater samples onto polycarbonate filters and subsequent coating of filters in gelatin, no decrease in bacterial cell numbers was observed with extensive permeabilization. The detection rates of bacterioplankton in river and pond water samples by CPRINS-FISH with a universal 16S rRNA gene primer and probe set ranged from 65 to 76% of total cell counts (mean, 71%). The concentrations of cells detected by CPRINS-FISH targeting of the rpoD genes of Aeromonas sobria and A. hydrophila in the water samples varied between 2.1 × 10³ and 9.0 × 10³ cells ml⁻¹ and between undetectable and 5.1 × 10² cells ml⁻¹, respectively. These results demonstrate that CPRINS-FISH provides a high sensitivity for microscopic detection of bacteria carrying a specific gene in natural aquatic samples.

Microscope-based approaches with nucleic acid probes are increasingly popular to gain information on the numbers, kinds, functions, and activities of microorganisms present within individual ecosystems (4, 12, 19, 34). There are two basic categories of nucleic acid probes: phylogenetic and functional.

Phylogenetic probes can be designed to target phylogenetic groups in complex microbial communities. rRNA is the most widely used target molecule. Fluorescence in situ hybridization (FISH) with rRNA-targeted oligonucleotide probes is now a widely accepted method in microbial ecology (4). One drawback of rRNA-targeted FISH, however, is the limited detection of cells with low ribosome content (4), which can represent a considerable fraction of total cells in any environmental sample. Various strategies have been used to overcome this difficulty, including FISH in combination with direct viable counting (20), use of rRNA targeted polynucleotide transcript probes (28), and enzymatic signal amplification (19, 32). Another drawback of FISH is that the degree of resolution obtained with rRNA sequence analysis is not sufficiently discriminatory to permit resolution of intrageneric relationships because of the extremely slow rate of evolution of rRNA. Other gene sequences that provide a higher resolution than that of 16S rRNA, such DNA gyrase (gyrB) or RNA polymerase sigma 70 factor (rpoD), have been used for phylogenetic studies on natural bacterial habitats (42). The higher level of sequence variation in such genes allows differentiation of closely related strains (11, 27). However, the application of FISH targeting these genes is hampered by the low copy number of target molecules. In situ PCR with phylogenetic probes can overcome this difficulty (12, 33).

Functional probes are designed to target genes encoding specific enzymes, which in turn are responsible for specific chemical transformations, toxins, cell growth, cell flagellation, etc. (6, 13, 24, 38). The application of in situ PCR with functional probes also provides a powerful tool for detection of genes or gene products in individual cells. The method, however, poses difficulties for a wide variety of applications to diverse species in the natural environment. First, permeabilization conditions for in situ gene amplification need to be optimized so that all reagents can penetrate the cell without diffusing the amplified products outside. The permeability of microbial cell wall structures is not uniform, and many different permeabilization procedures have been used for in situ PCR (12, 33, 34). Second, in situ PCR is usually carried out on glass slides. For aquatic specimens such as oligotrophic river, lake, marine, and drinking water samples, it is indispensable that target cells be concentrated through filtration prior to quantitative evaluation. For FISH, direct counting on polycarbonate filters was used to rapidly and accurately enumerate total and specific microbes in aquatic samples (23).

The objective of the present study was to describe a new approach to recognize individual genes in a bacterial cell by using cycling primed in situ amplification-fluorescent in situ hybridization (CPRINS-FISH). CPRINS uses one primer and results in linear amplification of the target DNA, which has been used to detect low-copy-number, even single-copy, nucleic acids in eukaryotic tissue (7, 10, 35). Extension products are several kilobases in length and are thus less likely to leak from the cells (16). In addition, coating samples trapped onto polycarbonate filters in gelatin facilitates reliable enumeration of bacterial cells carrying specific genes without diffusing amplified products outside the cell or causing species-selective cell loss and destruction in mixed microbial communities even if extensive permeabilization conditions are used. To improve the specificity and sensitivity of CPRINS, multiply labeled fluorescent probe sets were used for detection of the amplicons by in situ hybridization here. CPRINS-FISH performed on polycarbonate filters allows concentration of target cells in aquatic samples. A phylogenetic probe targeting rpoD and functional probes targeting antibiotic resistance genes were used for CPRINS-FISH analysis with diverse microorganisms.

MATERIALS AND METHODS

Bacterial samples and cell fixation.

The bacterial strains used in the present study are as follows: Aeromonas caviae ATCC 15468, Aeromonas enteropelogenes ATCC 49803, Aeromonas hydrophila ATCC 7966, Aeromonas jandaei ATCC 49568, Aeromonas media ATCC 33907, Aeromonas schubertii ATCC 43700, Aeromonas sobria ATCC 43979, Agrobacterium tumefaciens R1000, Bacillus subtilis 168, Brevundimonas diminuta IFO 3140, Comamonas testosteroni IAM 12419, Empedobacter brevis NCTC 11099, Enterobacter gergoviae JCM 1234, Escherichia coli O157:H7 ATCC 43888, Escherichia coli DH10, Escherichia coli JM109, Klebsiella oxytoca ATCC 13182T, Pseudomonas putida ATCC 12633, Ralstonia eutropha KT1, Sphingobacterium thalpophilum NCTC 11429. The pUC19 plasmid encoded ampicillin-resistance gene was transformed to E. coli JM109. The plasmid pBBR122 (MoBiTec, Göttingen, Germany) was transformed to C. testosteroni, E. coli DH10, E. gergoviae, P. putida, and R. eutropha. A. hydrophila, A. media, A. sobria, B. diminuta, C. testosteroni, E. brevis, S. thalpophilum, P. putida, and R. eutropha were grown in Luria-Bertani medium (1% tryptone, 0.5% yeast extract, 0.5% NaCl) at 30°C. Other strains were grown in Luria-Bertani medium at 37°C.

Cells were harvested by centrifugation at 8,000 × g for 10 min at 4°C and washed twice with phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 8.1 mM Na₂PO₄, 1.5 mM KH₂PO₄ [pH 7.2]). Cells were suspended in 4% paraformaldehyde in PBS for 16 h at 4°C. After fixation, cells were filtered through gelatin [0.1% gelatin, 0.01% CrK(SO₄)₂]-coated polycarbonate white filter (0.2-μm pore size, 25-mm diameter; ADVANTEC, Tokyo, Japan) and rinsed twice with filtered deionized water. Fixed cells were stored at −20°C until use.

DNA extraction.

Cells were harvested by centrifugation at 8,000 × g for 10 min at 4°C, washed twice with PBS, and suspended in sterile deionized water. Cell suspension was frozen in liquid nitrogen and then thawed at 60°C three times (total) in succession. The suspension was centrifuged at 8,000 × g for 10 min at 4°C, and the supernatant was used for PCR amplification.

Oligonucleotides and polynucleotides.

The nucleotide sequences of the ampicillin resistance (ampR) gene, chloramphenicol acetyltransferase (CAT) gene, rpoD gene, and 16S rRNA gene were obtained from GenBank (Release 138.0). Multiple-sequence alignment was carried out with CLUSTAL X 1.81 (37). Oligonucleotide primers for PCR and CPRINS and polynucleotide probes were designed in the present study (Table 1) with the exception of the universal 16S rRNA primer (15). Jan-Sobf primer (5′-CCGAAGATATGGCGCCGACT-3′) and Med-Sobf primer (5′-CGGACGACGTGGCGCCGACC-3′) were also designed as a forward primer for PCR in order to examine the specificity of CPRINS primer for A. sobria (Sobr). Jan-Sobf and Med-Sobf were complementary to rpoD sequences of A. jandaei, A. media, respectively. Med-Hydr primer (5′-TATTCAAACCAGGCAGTTTCG-3′), and Sch-Hydr primer (5′-TACTCGAACCAGGCGGTCTCG-3′) were designed as a reverse primer for PCR in order to examine the specificity of CPRINS primer for A. hydrophila (Hydf). Med-Hydr and Sch-Hydr were complementary to rpoD sequences of both A. jandaei-A. media and A. schubertii, respectively. The specificities of the primer and probe sequences were verified against NCBI nucleotide databases by using the BLAST program (2). All primers and polynucleotide probes were purchased from Texas Genomics Japan Co., Ltd. (Tokyo, Japan), and probes were labeled with Alexa Fluor 546 at guanine by using the ULYSIS Alexa Fluor 546 nucleic acid labeling kit (Molecular Probes) and the manufacturer's recommended procedures.

TABLE 1.

Probes and primers designed in this study

Name	Targeta	Primer or probe	Nucleotide sequence (5′-3′)
AmpR20f	Ampicillin-resistance gene	Primer	GTGTCGCCCTTATTCCCTTT
AmpR840r	Ampicillin-resistance gene	Primer	GGCACCTATCTCAGCGATCT
Amp1	Ampicillin resistance gene	Probe	CTTCTGCGCTCGGCCCTTCCGGCTGGCTGGTTTATTGCTGATAAATCTGG
Amp2	Ampicillin resistance gene	Probe	TGAGCGTGGGTCTCGCGGTATCATTGCAGCACTGGGGCCAGATGGTAA
Amp3	Ampicillin resistance gene	Probe	CACGACGGGGAGTCAGGCAACTATGGATGAACGAAATAGACAGATCGCTGAG
CAT39f	CAT gene	Primer	ATCCCAATGGCATCGTAAAG
CAT649r	CAT gene	Primer	CCTGCCACTCATCGCAGTA
CAT372Lf	CAT gene	Probe	GGCGTGTTACGGTGAAAACCTGGCCTATTTCCCTAAAGGGTTTATTGAG
CAT437Lf	CAT gene	Probe	CAGCCAATCCCTGGGTGAGTTTCACCAGTTTTGATTTAAACGTGGCCAATATGG
CAT519Lf	CAT gene	Probe	GGGCAAATATTATACGCAAGGCGACAAGGTGCTGATGCCGCT
CAT569Lf	CAT gene	Probe	AGGTTCATCATGCCGTCTGTGATGGCTTCCATGTCGGCAGAATG
CAT617Lf	CAT gene	Probe	ATGAATTACAACAGTACTGCGATGAGTGGCAGGGCGGGG
Sobf	rpoD gene of A. sobria	Primer	CCGAAGATGTGGCGCCTACT
Sobr	rpoD gene of A. sobria	Primer	TACTCAAACCAGGCACTATCA
Sob1	rpoD gene of A. sobria	Probe	GAAGATGTGGCGCCTACTGCCACCCATATCGGCTCTGAACTCAGCG
Sob2	rpoD gene of A. sobria	Probe	AGAAGTTTGGCGAGCTGCGCGCCCAGTATGAGGTAACCCGCCTCT
Sob3	rpoD gene of A. sobria	Probe	AGATGATCTGGCTGATGAGGACGAGGAAGAGGACGAAGATGAAGACG
Sob4	rpoD gene of A. sobria	Probe	GGATGGCGACAACTCGGATGATGAAGGTGACAGTGGCCCGGATCC
Sob5	rpoD gene of A. sobria	Probe	AACCTTCGTCGCTGCCTTCACTAACAACGAGTGTGATAGTGCCTG
Hydf	rpoD gene of A. hydrophila	Primer	AACCAGGTACAGAGTTCCGTC
Hydr	rpoD gene of A. hydrophila	Primer	TACTCGAACCAGGCAGTTTCG
Hyd1	rpoD gene of A. hydrophila	Probe	CAAGCAGATAGGTAATCGCTTCCGGGTACTCGGCGACGGAACTCTG
Hyd2	rpoD gene of A. hydrophila	Probe	GGGATCGATGAAGCCGGAAATGATGTCGGACAGACGCAGCTGTTCCG
Hyd3	rpoD gene of A. hydrophila	Probe	TCGGAGCCGATGTGAGTGGCGGTCGGTGCGACGTCGT
Hyd4	rpoD gene of A. hydrophila	Probe	CGTCGGAGTCATCACCATCACCGTCTTCGTCTTCATCTTCGTCTTC
Hyd5	rpoD gene of A. hydrophila	Probe	GCTGAATGGAGAGACGAGTTACTTCGTACTGGGCGCGCAGATCACCGAA
EUB933	Conserved 16S rRNA gene	Primer	GCACAAGCGGTGGAGCATGTGG
EUB1387Gr	Conserved 16S rRNA gene	Primer	CCCCCCGTGCCCCCGCCCCGCCCGCCGCGCGCGGCGGGCGGCCCGGGAACGTATTCACCG
EUB1387Gf	Conserved 16S rRNA gene	Probe	CGGTGAATACGTTCCCGGGCCGCCCGCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGG

^arpoD gene, RNA polymerase sigma factor gene.

PCR.

The PCR mixture containing PCR buffer (10 mM Tris-HCl [pH 9.0], 50 mM KCl, 1% Triton X-100), 2.0 mM MgCl₂ (2.5 mM for EUB933f and EUB1387Gr primers), 0.2 mM concentrations of each deoxynucleoside triphosphate, 0.4 μM concentrations of the primers, and 2.5 U of AmpliTaq Gold (Applied Biosystems) was made up with DNA-free water. PCR cycles consisted of a hot start at 95°C for 9 min, denaturation at 94°C for 1 min, and annealing and extension at 72°C for AmpR20f and AmpR840r primers, at 65°C for CAT39f and CAT649r primers, or at 64°C for Sobf and Sobr primers or Hydf and Hydr primers for 2 min. Amplification was repeated for 20 to 35 cycles with a thermal cycler (PTC-200; MJ Research, Inc.). Annealing temperatures for ampR gene and CAT gene were optimized by increasing it stepwise by 2°C using template DNAs extracted from gram-negative strains (C. testosteroni, E. coli, E. gergoviae, K. oxytoca, P. putida, and R. eutropha) without the ampR gene or CAT gene as negative controls. Annealing temperatures for rpoD genes of A. hydrophila and A. sobria were optimized by using template DNAs extracted from other nontarget Aeromonas spp. (A. caviae, A. enteropelogenes, A. jandaei, A. media, and A. schubertii) and E. coli as negative controls. In order to confirm the specificity of Sobr and Hydf primers, (i) a combination of Sobr and other forward primers (Jan-Sobf, Med-Sobf) and (ii) a combination of Hydf and other reverse primers (Jan-Hydr, Sch-Hydr) were used for PCR with template DNAs from A. hydrophila, A. jandaei, A. media, A. schubertii, A. sobria, and E. coli as negative controls.

For EUB933f and EUB1387Gr primers, the annealing temperature was initially set at 65°C and was then decreased by 0.5°C every cycle until it was 55°C; then, 10 additional cycles were carried out at 55°C (15). Denaturing was carried out at 94°C for 1 min, primer annealing was performed by using the scheme described above for 1 min, and primer extension was performed at 72°C for 3 min. The final extension step was 7 min at 72°C.

After amplification, 5 μl of each PCR mixture was analyzed by electrophoresis in 9% polyacrylamide gel at 150 V for 45 min. After electrophoresis, the gel was stained with 0.5 μg of ethidium bromide ml⁻¹ for 15 min. The gel was scanned with Foto/Eclipse (FOTODYNE).

Cell wall permeabilization.

Permeabilization was used prior to CPRINS-FISH for bacterial cells trapped on gelatin-coated polycarbonate filter. To avoid cell loss during extensive cell wall permeabilization, the filters with bacterial cells were coated in gelatin (29). The filters were soaked in gelatin solution, containing 0.1% gelatin and 0.01% CrK(SO₄)₂, at ca. 50°C for 5 min, dried face up on a glass slide at room temperature (ca. 25°C), and subsequently vacuum dehydrated. Filters were incubated in a lysozyme solution (10 mg of lysozyme [Nacalai Tesque, Inc., Kyoto, Japan] ml⁻¹, 100 mM Tris-HCl [pH 8.2], 50 mM EDTA) at 37°C for 60 min. Filters were rinsed with sterile deionized water, dehydrated in 99% ethanol, and vacuum dried. Each filter was cut into 16 sections and subjected to CPRINS-FISH.

CPRINS-FISH.

A 1/16 section of the filter was transferred to a microtube (0.2 ml) and immersed in 100 μl of the CPRINS buffer, containing PCR buffer (10 mM Tris-HCl [pH 9.0], 50 mM KCl, 1% Triton X-100), 2.0 mM MgCl₂ (2.5 mM for EUB1387Gr primer), a 0.2 mM concentration of each deoxynucleoside triphosphate, 0.4 μM primer, 0.5 M Betain, and 2.5 U of AmpliTaq Gold (Applied Biosystems). CPRINS cycles consisted of a hot start at 95°C for 9 min, denaturation at 94°C for 1 min, and annealing and extension at 72°C for AmpR840r primer, at 65°C for CAT649r primers, or at 64°C for Sobr primer or Hydf primer for 2 min. Amplification was repeated for 20 to 35 cycles with a thermal cycler (PTC-200; MJ Research, Inc.). For EUB1387Gr primers, the annealing temperature was initially set at 65°C and was then decreased by 0.5°C every cycle until it was 55°C; then, 10 additional cycles were carried out at 55°C. Denaturing was carried out at 94°C for 1 min. Primer annealing was performed by using the scheme described above for 1 min, and primer extension was performed at 72°C for 3 min. The final extension step was 7 min at 72°C.

After the amplification, filters were rinsed with 0.1% Nonidet P-40 PBS at room temperature for 5 min and dehydrated in 99% ethanol and vacuum dry. Filters were soaked in 100 μl of hybridization buffer (for the ampR gene, CAT gene, and 16S rRNA gene, the buffer consisted of 1 M Betain, 20 mM Tris-HCl [pH 8.8], 10 mM KCl, 10 mM NH₄SO₄, 4 mM MgSO₄, and 0.1% Triton X-100; for the rpoD gene, the buffer consisted of 0.9 M NaCl, 20 mM Tris-HCl, 0.01% sodium dodecyl sulfate, 5 mM EDTA, and 30% formamide) containing 5 to 20 ng of Alexa Fluor 546-labeled probe. After 1 min of heat treatment at 90°C, the filters were incubated at 70°C for 30 min and then washed with washing buffer (0.1% Nonidet P-40 in PBS) at 37°C for the ampR gene, CAT gene, and 16S rRNA gene or at 70°C for rpoD genes for 10 min. Hybridization and washing temperatures for both ampR and CAT genes were optimized by increasing it stepwise by 5°C using gram-negative cells (C. testosteroni, E. coli, E. gergoviae, K. oxytoca, P. putida, and R. eutropha) without target gene as negative controls. For the 16S rRNA gene, the same temperatures were used, and the utility was confirmed using laboratory strains. Hybridization and washing temperature for rpoD genes were optimized by increasing it stepwise by 5°C using nontarget Aeromonas cells (A. caviae, A. enteropelogenes, A. jandaei, A. media, and A. schubertii) as negative controls. The formamide concentration in hybridization buffer was also optimized by increasing it stepwise by 10%.

Finally, filters were counterstained with 1 μg of DAPI (4′,6′-diamidino-2-phenylindole) ml⁻¹ for 10 min. In order to exclude the possibility of nonspecific probe binding to cell structures other than target DNA in the target cells, (i) FISH using laboratory strains without amplification of target DNA and (ii) CPRINS-FISH targeting green fluorescent protein gene (gfp) using laboratory strains that did not carry the gfp gene were performed.

rRNA-targeted FISH.

The following rRNA-targeting oligonucleotide probes were used for FISH: EUB338 (5′-GCTGCCTCCCGTAGGAGT-3′), which is complementary to a conserved region of most bacterial 16S rRNA molecules; NON338 (5′-ACTCCTACGGGAGGCAGC-3′), which is a negative control probe (3). Bacterioplankton in environmental samples was fixed with a final concentration of 4% paraformaldehyde and trapped onto polycarbonate white filters (0.2-μm pore size, 25-mm diameter; ADVANTEC). For EUB338 and NON338 probes, hybridization and washing were performed as described by Alfreider et al. (1).

Microscopy.

Filters were mounted in immersion oil for observation by epifluorescence microscopy (E-400; Nikon, Tokyo, Japan) with the Nikon filter sets UV-2A and HQ:CY3 for UV and green excitation, respectively. Image were acquired by a cooled charge-coupled device camera (Cool Snap; Roper Photometrics) and stored as digital files. More than 1,500 DAPI-stained objects were counted per sample.

Freshwater samples.

Surface river water samples were taken from Takayama (34°52′12"N, 135°28′60"E) and Takiue (34°51′10"N, 135°28′49"E) in the Minohgawa River, Kuwazu in the Inagawa River (34°46′47"N, 135°25′42"E) in the northern part of Osaka, Japan (19, 41). Takayama and Takiue are surrounded by forest and oligotrophic sites. At these sites, the river is narrow and shallow. The water is not exposed to domestic or industrial effluents. Kuwazu is located in an industrial area. The site is considered to be polluted by total organic carbon. Pond water was taken from the garden of the Graduate School of Pharmaceutical Sciences, Osaka University (34°48′49"N, 135°31′30"E). The pond is 10 by 15 m square, about 1.2 m deep, and eutrophic.

Samples were fixed with a final concentration of 4% paraformaldehyde at each sampling site. After fixation, 2- to 20 ml-portions were filtered through gelatin [0.1% gelatin, 0.01% CrK(SO₄)₂]-coated polycarbonate white filter (0.2-μm pore size, 25-mm diameter; ADVANTEC) and rinsed twice with filtered deionized water. The filters were stored in −20°C until used. For negative control in CPRINS-FISH with environmental samples, CPRINS-FISH targeting the gfp gene was performed. There was no gfp in freshwater samples used in the present study.

Statistical evaluation.

Statistical evaluation was carried out with Microsoft Excel XP software. The percentage of DAPI counts detected by CPRINS-FISH with universal 16S rRNA gene primer and probe was compared to that detected by FISH with EUB338 probe in environmental samples by using the Student t test.

RESULTS AND DISCUSSION

Structure of primer and probe for CPRINS-FISH.

CPRINS uses one primer resulting in linear amplification of target DNA with extension products that are several kilobases in length and thus less likely to leak from cells (16). Extensive permeabilization conditions can be used. Although the technique has proven to be extremely sensitive, it has not been widely used in the field of microbiology because direct incorporation of labeled nucleotide during DNA amplification in cells results in high background and nonspecific binding. Multiply labeled fluorescent probes were used to hybridize amplicons and improve specificity and sensitivity in the present study (Fig. 1A). Each probe ranged from 39 to 54 mer in length and was labeled with Alexa Fluor 546 at the guanine bases. That is, each probe contained approximately 7 to 19 fluorescent molecules. When CPRINS was repeated up to 30 cycles and three to five probes were used for hybridization, amplicons in an individual cell would be labeled with more than 1,500 fluorescent molecules. This number of fluorescent molecules was thought to be sufficient for single cell detection under epifluorescence microscopy.

FIG. 1.

Systematic representation of the primer and probes for CPRINS-FISH. (A) Multiply labeled probes, whose lengths are 39 to 54 mer, are hybridized with amplicon initiated by CPRINS primer at different target regions. (B) Cytosine clamp primer initiates strand synthesis. Then, the guanine clamp probe is hybridized with amplicon at the cytosine clamp primer region. Circles indicate fluorescent molecules labeled at guanine bases.

Another strategy for probe design used a cytosine clamp primer for amplification of the conserved 16S rRNA sequence (Fig. 1B). This primer consisted of 20 mer of target sequence and 40 mer of cytosine-rich sequence. After CPRINS was repeated for target bacteria, the complementary guanine clamp probe, which was labeled with a total of 32 fluorescent molecules at the guanine bases, was hybridized with the cytosine clamp primer region of the amplicon. The advantage of the cytosine clamp primer/guanine clamp probe system is that it requires as little as 20 mer of sequence for the target, which allows detection of the target gene with a short specific region but still incorporates enough fluorescent molecules at the guanine bases in the probe.

CPRINS-FISH targeting antibiotic resistance gene.

In order to establish the basic protocol of CPRINS-FISH, multicopy genes were targeted at the first step. E. coli cells harboring ampR gene on a high-copy-number plasmid pUC19 and gram-negative strains harboring CAT gene on a medium-copy-number plasmid pBBR122 were used. After optimizing PCR conditions with ampR gene primer sets, target genes were linearly amplified for 20 cycles with single primer at single cell level. Then, three probes (Amp1, Amp2, and Amp3 [Table 1]) were hybridized. Hybridization signals were observed in E. coli JM109 cells carrying plasmid pUC19, which encodes the ampR gene but not in E. coli O157:H7 cells (Fig. 2). CPRINS yielded a low-level background because amplicons were prevented from leaking outside cells.

FIG. 2.

CPRINS-FISH of the ampR gene in fixed cells of E. coli JM109 pUC19 (A and D) and E. coli O157:H7 ATCC 43888 (C and F) and a fixed-cell mixture of E. coli JM109 pUC19 and E. coli O157:H7 ATCC 43888 (B and E). (A, B, and C) Under UV excitation, all DAPI-stained cells were visualized. (D, E, and F) Under green excitation, only cells with ampR gene-amplified products emitted red fluorescence of Alexa Fluor 546-labeled probe.

CPRINS-FISH targeting CAT gene on pBR122 was carried out for gram-negative strains. pBBR122 has a broad host range and can be replicated in diverse gram-negative strains. The pBBR122 was introduced to different genus in the present study, and applicability of our technique was examined with the different genus. The copy number of pBBR122 was less than that of pUC19 and thus, to improve sensitivity, target genes were amplified for 35 cycles and five probes were hybridized with the single-stranded amplicons. Positive hybridization signals were observed in C. testosteroni, E. coli, E. gergoviae, P. putida, and R. eutropha cells harboring pBBR122 but not with the same strains without the plasmid or other gram-negative strains (Table 2). CPRINS-FISH was performed with the strains harboring pBBR122 under the same permeabilization and thermal conditions. Many different permeabilization procedures have been used for in situ PCR/reverse transcription-PCR and in situ LAMP to detect todC1 and nahA in Pseudomonas putida (6, 12), groEL, tsf in Salmonella enterica serovar Typhimurium (13), the phenol hydroxylase gene in Ralstonia eutropha (34), stx₂ gene in E. coli (21, 24), and flhDC in Serratia liquefaciens (38). Impermeability of the cell wall to enable the polymerase to access the cell causes false-negative results, while diffusion of labeled amplicon and its adhesion to cells gives false-positive results. The permeabilization used here for CPRINS was more severe than that used for in situ PCR (33) because the amplicon of CPRINS is larger and resists leaking from the cells. Consequently, the same permeabilization conditions could be used to visualize the same gene in different microorganisms. Because the amplicon in CPRINS is single-strand DNA, the time for hybridization with multiply labeled probes is brief (only 30 min). As a result, the total reaction time for experiment by CPRINS-FISH, including permeabilization, DNA amplification, FISH, and washing, was 8 h. This is much shorter than the time required for other procedures such as chromosomal painting (22), chromosome targeted FISH (25), and RING-FISH (43), which detect specific sequence in a single bacterial cell.

TABLE 2.

Specificity of CPRINS-FISH targeting CAT gene on pBBR122

Strain	Plasmid (marker)	Presence (+) or absence (−) of signal
Comamonas testosteroni R5	pBBR122 (CAT)	+
Enterobacter gergoviae JCM 1234	pBBR122 (CAT)	+
Escherichia coli DH10	pBBR122 (CAT)	+
Pseudomonas putida ATCC 12633	pBBR122 (CAT)	+
Ralstonia eutropha KT1	pBBR122 (CAT)	+
Comamonas testosteroni R5		−
Enterobacter gergoviae JCM 1234		−
Escherichia coli DH10		−
Escherichia coli ATCC 43888		−
Escherichia coli JM109	pUC19 (KanR)	−
Escherichia coli JM109	pT7GFP (ampR, gfp)	−
Klebsiella oxytoca ATCC 13182T		−
Pseudomonas putida ATCC 12633		−
Ralstonia eutropha KT1		−

CPRINS-FISH targeting rpoD gene.

The protocol of CPRINS-FISH was established with functional probes targeting antibiotic resistance (ampR and CAT) genes, and then its applicability to phylogenetic identification was examined. The RNA polymerase sigma factor is responsible for promoter-specific transcription initiation on RNA polymerase, which is essential for cell survival in bacteria (18). Horizontal transmission of the gene may be as rare as that of rRNA genes (31, 42). Phylogenetic analysis using the rpoD sequences of Aeromonas spp. provided higher resolution than one using 16S rRNA sequences and allowed easy design of CPRINS primers and probes.

The specificities of the rpoD-targeted primers and probes used in the present study were summarized in Table 3. CPRINS primers (Sobr and Hydf) have at least three mismatches against nontarget Aeromonas spp. and other bacterial species. In order to determine thermal conditions, PCR amplification of rpoD gene was attempted using DNA from Aeromonas spp. Sobf primer was used as a forward primer in PCR for A. sobria, while Hydr primer was used as a reverse primer in PCR for A. hydrophila. The primer set Sobf and Sobr amplified specifically the rpoD sequences (457 bp) of A. sobria, and the primer set Hydf and Hydr amplified the rpoD sequences (588 bp) of A. hydrophila (Table 3).

TABLE 3.

Specificity of the rpoD-targeted primers and probes determined by PCR or CPRINS-FISH

Species	No. of mismatches with primer and probes for A. sobria						PCRa	CPb	No. of mismatches with primer and probes for A. hydrophila						PCRc	CPd
	Sobr	Sob1	Sob2	Sob3	Sob4	Sob5			Hydf	Hyd1	Hyd2	Hyd3	Hyd4	Hyd5
A. sobria	0	0	0	0	0	0	+e	+	5	8	9	9	12	8	−	−
A. hydrophila	4	9	7	8	10	6	−	−	0	0	0	0	0	0	+	+
A. media	4	8	8	8	10	6	−	−	4	9	6	3	11	6	−	−
A. jandaei	4	6	5	7	7	5	−	−	3	8	7	7	9	8	−	−
A. schubertii	5	12	8	15	13	8	−	−	3	6	8	6	13	14	−	−
E. coli	12	12	15	31	31	22	−	−	5	9	15	8	22	15	−	−

^aPCR targeting rpoD gene of A. sobria.

^bCPRINS-FISH targeting rpoD gene of A. sobria.

^cPCR targeting rpoD gene of A. hydrophila.

^dCPRINS-FISH targeting rpoD gene of A. hydrophila.

^e+, positive; −, negative.

Using combination of Sobr and other forward primers (Jan-Sobf and Med-Sobf) or combination of Hydf and other reverse primers (Jan-Hydr and Sch-Hydr), the specificities of each CPRINS primer (Sobr and Hydf) were further tested with template DNAs from A. jandaei, A. media, and A. schubertii. The Jan-Sobf, Med-Sobf, Jan-Hydr, and Sch-Hydr primers were complementary to rpoD sequences of A. jandaei, A. media, or A. schubertii. Sobr and Hydf primers have three to five mismatches against these species. There was no amplification when a combination of Sobr and other Aeromonas-targeted forward primers or a combination of Hydf and other Aeromonas-targeted reverse primers was used. Under the current conditions, each CPRINS primer was thought to be able to discriminate the number of mismatch more than three bases.

The specificities of the rpoD-targeted primers were further confirmed by CPRINS-FISH (Fig. 3). Probes Sob1, Sob2, Sob3, Sob4, and Sob5 were used for detection of rpoD gene of A. sobria. The probes were completely complementary to sequences of rpoD gene of A. sobria and have at least five mismatches against nontarget bacterial species (Table 3). The thermal condition used for CPRINS was the same as for PCR and resulted in positive hybridization signals for only A. sobria. No signal was observed with the nontarget species (Fig. 3 and Table 3). CPRINS-FISH was performed also with rpoD of A. hydrophila. Probes Hyd1, Hyd2, Hyd3, Hyd4, and Hyd5 were used for detection of rpoD of A. hydrophila. The probes have more than three mismatches against nontarget bacterial species (Table 3). Positive hybridization signals were observed only in A. hydrophila. The utility of CPRINS-FISH with these primers and probes was confirmed using nontarget Aeromonas spp. (Table 3) and other genera as follows: α-Proteobacteria (A. tumefaciens and B. diminuta), β-Proteobacteria (C. testosteroni and R. eutropha), γ-Proteobacteria (E. coli, E. gergoviae, and P. putida). Consequently, CPRINS-FISH targeting rpoD gene enables the detection of bacterial species at the single cell level and facilitates studies on species diversity and the abundance or interaction of bacteria at species level in complex microbial communities.

FIG. 3.

Discrimination of A. sobria ATCC 43979 from mixed Aeromonas spp. by CPRINS-FISH. (A and D) A. sobria ATCC 43979; (C and F) A. hydrophila ATCC 7966; (D and E) mixture of A. sobria ATCC 43979 and A. hydrophila ATCC 7966. (A, B, and C) Under UV excitation, all DAPI-stained cells were visualized. (D, E, and F) Under green excitation, only target species emitted red fluorescence of Alexa Fluor 546-labeled probe.

Applicability of CPRINS-FISH to environmental samples.

In order to examine the applicability of CPRINS-FISH to freshwater bacterioplankton, CPRINS-FISH with a universal 16S rRNA gene primer and probe was developed. The utility of CPRINS-FISH with this primer and probe was confirmed by using phylogenetically distant genera as follows: α-Proteobacteria (A. tumefaciens and B. diminuta), β-Proteobacteria (C. testosteroni and R. eutropha), γ-Proteobacteria (A. hydrophila, A. sobria, E. coli, E. gergoviae, and P. putida), Bacteroides-Flavobacterium-Cytophaga phylum (E. brevi and S. thalpophilum), and B. subtilis, and then it was applied to environmental samples.

The densities of heterotrophic bacterioplankton in the oligotrophic river samples (Takayama, Takiue), eutrophic river sample (Kuwazu), and pond sample (Osaka University) were 9.3 × 10⁴, 1.1 × 10⁵, 4.6 × 10⁵, and 4.5 × 10⁶ cells ml⁻¹, respectively. By coating the samples in gelatin, significant cell loss was not detected during the experimental steps of CPRINS-FISH, as shown in Table 4. In contrast, a decline in cell numbers was observed in samples that were not coated (data not shown). Samples that were not treated with lysozyme showed lower detection rates (less than one-third of DAPI-stained cells) and hybridization signals after CPRINS-FISH. Representative photomicrographs of bacterioplankton in pond water sample detected by CPRINS-FISH are shown in Fig. 4A and D. The detection rates of bacterial cells by CPRINS-FISH with universal 16S rRNA gene primer and probe were higher (mean, 71%; range, 65 to 76% of total cell counts) than those by standard FISH with monolabeled EUB338 probe (mean, 43%; range, 25 to 59% of total cell counts) in all samples (Osaka University, P < 0.01; Takayama, P < 0.01; Takiue, P < 0.01; Kuwazu, P < 0.01) (Table 4). However, 25 to 35% of the DAPI-stained cells in the samples still could not be detected by CPRINS-FISH with the universal primer and probe. These cells may be impermeable to Taq DNA polymerase (97 kDa) or perhaps had a sequence mismatch against primer. Cells that are impermeable to enzyme may be resistant to lysozyme treatment. Certain bacterial genera such as Actinobacteria and Staphylococcus were not sensitive to these enzymes. The application of other enzymatic or chemical treatments may be required for in situ detection of such bacteria (30).

TABLE 4.

Detection rates by FISH and CPRINS-FISH in freshwater samples^a

Sample site	Detection rate (% DAPI) with universal probes primer or probe		% of remaining cellsb	Cells ml−1 (SD)c
	FISH	CPRINS-FISH		A. sobria	A. hydrophila
Osaka University (pond)	59 (2.3)	76 (1.1)	106 (11)	9.0 × 103 (3.6)	<4.5 × 103
Takayama (river)	29 (4.0)	70 (5.7)	86 (14)	6.3 × 102 (1.5)	2.3 × 102 (0.4)
Takiue (river)	25 (5.3)	65 (3.7)	104 (19)	2.1 × 102 (0.1)	1.8 × 102 (0.4)
Kuwazu (river)	57 (6.2)	72 (3.1)	114 (11)	1.0 × 103 (0.2)	5.1 × 102 (0.8)

^aValues in parentheses indicate standard deviations of triplicate samples.

^bThe recovery of bacterioplankton cells on gelatin-coated polycarbonate filter during all experimental steps of CPRINS-FISH was determined by comparing cell number on the filter before and after CPRINS-FISH.

^cThe concentration of bacterial cells detected by CPRINS-FISH targeting the rpoD gene of A. sobria or A. hydrophila.

FIG. 4.

Photomicrographs of bacterioplankton in a pond water sample detected by CPRINS-FISH targeting the universal 16S rRNA gene (A and D), in a river water sample collected from Takiue detected by CPRINS-FISH targeting the rpoD gene of A. sobria (B and E), and in a river water sample collected from Kuwazu detected by CPRINS-FISH targeting ampicillin resistance gene (C and F). (A, B, and C) Under UV excitation, all DAPI-stained cells were visualized. (D, E, and F) Under green excitation, only cells detected by CPRINS-FISH emitted red fluorescence of Alexa Fluor 546-labeled probe.

A. sobria and A. hydrophila in freshwater samples were also visualized by CPRINS-FISH (Fig. 4B and E). The concentrations of cells detected by the CPRINS-FISH targeting rpoD genes of A. sobria and A. hydrophila in river and pond water samples varied between 2.1 × 10³ and 9.0 × 10³ cells ml⁻¹ and between undetectable and 5.1 × 10² cells ml⁻¹, respectively (Table 4). Aeromonas spp. are ubiquitous and widely isolated from natural mineral waters, coastal waters, river water, and lake water (5, 9, 39, 40). The detection of Aeromonas in freshwater samples is consistent with our results: Aeromonas spp. may be common in freshwater systems and sometimes constitute a considerable cell fraction (8, 40). Aeromonas spp. are opportunistic pathogens that are at the same time infectious and enterotoxigenic (17). The presence of Aeromonas spp. in drinking water is a potential risk (14). Analysis of the distribution and diversity of the harmful bacteria in aquatic ecosystems will increase our understanding of their roles in nature and make a contribution which will help prevent disease outbreaks caused by such bacteria.

Bacteria carrying antibiotic resistance genes were detected by CPRINS-FISH (Fig. 4C and F), but the abundance is <0.1% of the DAPI-stained cells in the samples. The advantage of CPRINS-FISH prior to rRNA-targeted FISH is that it enables the detection of a single-copy gene in individual cells in aquatic environments. Numerous reports have demonstrated that lateral gene transfer has significantly contributed to the acquisition of new genetic traits, e.g., bacterial antibiotic resistance (26, 36). CPRINS-FISH has the potential to become a valuable tool in quantitative determination of gene movement among bacteria at the single-cell level.

Conclusion and future prospect.

The method presented here offers the possibility to address the specific gene in a bacterial cell in situ, regardless of copy number or metabolic activity. The high sensitivity of CPRINS-FISH was demonstrated easily by targeting a single copy gene on a chromosome. rpoD is a common housekeeping gene that all bacteria possess. Therefore, CPRINS-FISH with rpoD-targeted primer and probes can be used for phylogenetic identification and monitoring of target bacteria at the species level for various microorganisms. In addition, in situ DNA amplification on a polycarbonate filter allowed effective concentration of target cells from aquatic samples and enhanced the quantitative analysis. Thus, it can be applied to various aquatic samples such as marine, lake, drinking water, and water distribution systems. We anticipate that this gene-targeting approach will lead to further understanding of the composition and function of microbial communities and their dynamics in the natural environment.