Blackwell Science, LtdOxford, UKEMIEnvironmental Microbiology 1462-2912Blackwell Science, 20024771774Short communicationFluorimetric determination of %GCJ. M. Gonzalez and C.

Saiz-Jimenez

Environmental Microbiology (2002) 4(11), 770773

Brief report A uorimetric method for the estimation of G+C mol% content in microorganisms by thermal denaturation temperature
J. M. Gonzalez* and C. Saiz-Jimenez Instituto de Recursos Naturales y Agrobiologia, CSIC, Apartado 1052, 41080 Sevilla, Spain. Summary G+C mol% content in microorganisms is one of the recommended characteristics for the standard description of bacterial species. In this study we present a novel uorimetric method to estimate the G+C mol% content in microorganisms. Doublestranded DNA was specically stained with SYBR Green I, and its thermal denaturalization was followed by measuring a decrease in uorescence using a real-time PCR thermocycler. Unlike most previous determinations of G+C mol%, in this study only DNA from microorganisms with an available completely sequenced genome were used to prepare the calibration curves. Calibration curves showed a linear relationship between G+C mol% content and melting temperature and they were performed both in the absence and presence of 30% formamide. This protocol proves to be a rapid and inexpensive method to estimate DNA base ratios of novel microorganisms. Introduction The information contained in the DNA is encoded by four nitrogenated bases A, C, G and T (adenine, cytosine, guanine and thymine respectively). The proportion of G+C, or DNA base ratio (moles per cent of G+C), is considered part of the standard description of bacterial taxa (Vandamme et al., 1996; Rosell-Mora and Amann, 2001). Mol% G+C content varies between 24 and 76% in the bacterial world (Torsvik et al., 1995; Vandamme et al., 1996). Several techniques have been used for assessing the fraction of G+C in the genome of microorganisms. A direct method based on the sequencing of the whole genome from the query microbial species is a straightforward procedure; however, this would represent a long and unfeasible project only for the purpose of obtaining the G+C fraction in the DNA of a microbial species. The two most common approaches are either by high-performance liquid chromatography (HPLC) or by thermal denaturalization techniques. High-performance liquid chromatography techniques are usually considered accurate but require the set up of a HPLC system specically for this purpose which is expensive and only worthwhile if these determinations are performed with high frequency (Tamaoka and Komagawa, 1984). Measurements of absorbance during the thermal denaturalization of DNA have been used as a rapid alternative to estimate the content of G+C in genomic DNA from microorganisms. This strategy requires the availability of a spectrophotometer with a thermal controller. Melting temperatures of DNA molecules and their percentages of G+C follow a linear relationship (Marmur and Doty, 1962; De Ley et al., 1970). Fluorimetric determinations of DNA denaturalization should translate in a sensitive and simple method for assessing G+C mol% content in microorganisms. Because of the widespread use of quantitative polymerase chain reaction (PCR) techniques, real-time PCR thermocyclers are becoming common in most laboratories. Herein, we propose an easy, rapid, high-throughput, uorimetric technique to estimate percentage G+C content in DNA samples. The method uses a uorescent, double-strand specic dye and the melting temperature software and hardware capabilities of modern quantitative, real-time PCR thermocyclers. We obtained calibration relationships between GC mol% and melting temperature using microorganisms with completely sequenced genomes. Results and discussion
Received 1 August, 2002; accepted 10 September, 2002. *For correspondence. E-mail jmgrau@irnase.csic.es; Tel. (+34) 954624 711; Fax (+34) 954624 002.

Most modern real-time PCR thermocyclers allow to perform melting curve experiments. We used an iQ iCycler

2002 Blackwell Science Ltd

Fluorimetric determination of %GC 771 real-time thermocycler (Bio-Rad, Hercules, CA) to obtain melting curves of genomic DNA from a number of microorganisms and calculated their melting temperatures. Eleven prokaryotic strains with completely sequenced genomes were used in this study (Table 1). The G+C mol% of these strains ranged from 30.9% to 66.6%. Melting temperatures were calculated in the presence and absence of 30% formamide. In the absence of formamide, melting temperatures for the microbial species used in this study ranged between 70C and 90C (Table 1). In the presence of 30% formamide the range of melting temperatures for the microorganisms used in this study was approximately between 50C and 70C (Table 1). SYBR Green I (Molecular Probes, Eugene, Oregon) is a uorescent dye showing high uorescence when it binds to double-stranded DNA. Thus, using this dye, the doublestranded DNA molecules can be exclusively quantied during the denaturalization experiments. SYBR Green I shows maximum uorescence at excitation and emission wavelengths of 497 and 520 nm respectively. These peaks are coincident with the FAM lter set available in the standard conguration of any real-time PCR thermocycler. The G+C mol% from these microorganisms showed a positive relationship with the melting temperature (Tm) of their total genomic DNAs (Fig. 1). In the absence of formamide, the obtained regression line (n = 11, r 2 = 0.99, P << 0.001) was: %GC = 1.98 Tm 106.91 (Fig. 1A). In the presence of 30% formamide, the obtained relationship G+C mol% versus melting temperature (n = 11, r 2 = 0.98, P << 0.001) was: %GC = 1.99 Tm 71.08 (Fig. 1B). No signicant differences were found between the regression coefcients in the presence and absence of 30% formamide. These relationships can be used to estimate the G+C mol% content of novel microorganisms. Standard methods for estimating genotypic parameters, such as G+C mol% content from thermal denaturation curves and DNADNA hybridization, have been based on indirect determinations because controls showing the actual percentages of G+C content were not available for comparisons in the pre-genomic era. In this study, we calibrated our method against microorganisms with completely sequenced genomes. Unlike previous procedures for G+C mol% estimates using thermal denaturation curves, the use of completely sequenced genomes represents a direct method for the calibration of the proposed protocol. After a genome sequencing project has been completed, it has been observed that the G+C mol% content previously estimated from indirect methods could be up to several per cent units different from the actual G+C mol% value. As an example, we present the cases of Pyrococcus furiosus and P. horikoshii, two species of hyperthermophilic Archaea, whose genomes have been recently sequenced. Pyrococcus furiosus (Fiala and Stetter, 1986) actually has a G+C mol% of 40.8% whereas the initially reported value of G+C mol% was 38%. Highperformance liquid chromatography based method for G+C mol% estimates are exempt of this problem and the reported estimate for P. horikoshii was 42% (Gonzlez et al., 1998). Pyrococcus horikoshii (Gonzlez et al., 1998) has an actual G+C mol% content of 41.9% and the initially reported value obtained from melting curves was 44%. The use of actual percentage G+C-values from control microorganisms to be used in the calibration relationships

Table 1. Microbial species used in this study showing the strain name, G+C mol% content from their genome sequences and the melting temperatures of their genomic DNA obtained in the absence and presence of 30% formamide.a G+C mol% 30.93 35.33 40.77 41.88 43.52 44.71 48.58 50.79 53.81 66.56 66.63 Tm ( SD) 30% formamide 51.2 53.0 57.6 56.7 58.0 58.7 61.0 61.6 62.9 69.4 69.0 (0.5) (0.1) (0.1) (0.0) (0.3) (0.0) (0.1) (0.1) (0.3) (0.1) (0.3)

Species Clostridium acetobutylicumT Lactococcus lactis lactis Pyrococcus furiosusT Pyrococcus horikoshiiT Bacillus subtilis Pyrococcus abyssiT Archaeoglobus fulgidusT Escherichia coli Corynebacterium glutamicum Pseudomonas aeruginosa Deinococcus radioduransT

Strain ATCC 824, CECT 508 IL1403, EMR4, CECT 4433 DSM 3638 DSM 12428 Strain 168, CECT 461 Strain GE5 DSM 4304 K12, MG1655, CECT 433 ATCC 13869, CECT 77 PAO1, CECT 4122 R1, CECT 833

Tm ( SD) 70.0 71.3 74.8 74.5 76.1 75.9 78.6 80.3 80.6 87.3 88.2 (0.2) (0.2) (1.5) (0.0) (0.4) (0.2) (0.1) (0.2) (0.1) (0.2) (0.3)

a. Bacterial DNA extraction was performed following the method described by Marmur (1961) and Sambrook et al. (1989). Archaeal DNA was extracted as described by Charbonnier and Forterre, 1989). Thermal denaturalization was performed in 0.1 standard saline citrate with approximately 5 mg genomic DNA (De Ley et al., 1970) and SYBR Green I at a nal dilution of 1 : 100000. Thermal conditions consisted in a ramp from 25C to 100C at 1C min-1. Fluorescence measurements were performed at each step during this ramp. Calibration curves were obtained from the melting temperatures (Tm) of total genomic DNA and the G+C mol%. G+C mol% was obtained from the whole genome sequence of the microorganisms used in this study, including every genetic element, as [G+C]/[A + T + C + G] 100. Genome sequences were obtained from Entrez Genomes (http://www.ncbi.nlm.nih.gov/PMGifs/Genomes/micr.html) at NCBI. Tm was calculated from the minimum value of the slope tangent to the melting curve of uorescence versus temperature. Least-square linear regression analyses and slope comparisons were performed according to Sokal and Rohlf (1981). 2002 Blackwell Science Ltd, Environmental Microbiology, 4, 770773

772 J. M. Gonzalez and C. Saiz-Jimenez

Fig. 1. A. Calibration curve of G+C mol% versus melting temperature (Tm) for a number of microorganisms with their genomic sequence available. B. A calibration curve of the same strains in the presence of formamide (30%) is also shown.

between G+C mol% and melting temperature is an important factor for accurate G+C mol% results from thermal denaturation curves. The use of HPLC techniques for estimates of the DNA base ratio (Tamaoka and Komagawa, 1984; Meshbah et al., 1989) provide accurate results although the method involve the use of a dedicated HPLC set up not available in non-specialized laboratories. A relationship between G+C mol% content and melting temperature is useful for estimating percentage GC content of novel microorganisms during their taxonomic classication and is universal for the Prokaryotes (De Ley et al., 1970). A previous G+C mol% versus melting temperature relationship gave a regression coefcient of 2.44 (De Ley et al., 1970). This slope is signicantly higher than the one resulting from our study and these differences might be a consequence of the percentage GC values estimated for the species used in the calibration (see above paragraph) and the methods employed to obtain the melting curves; former studies measured absorbance using automatic recording thermal spectrophotometers (De Ley et al., 1970) whereas we have measured uorescence emitted by a double-stranded DNA-specic dye on a real-time PCR thermocycler. A previous percentage GC versus melting temperature calibration curve (De Ley et al., 1970) would imply the possibility of melting curves reaching temperatures over the boiling point of water for microorganisms with high GC content (>70%). In this study, we have showed that the G+C mol% estimates for these microorganisms with extremely high percentage GC

can also be performed in the presence of 30% formamide. The presence of 30% formamide decreased the melting temperature signicantly (by about 18C; Fig. 1) (Vandamme et al., 1996). Thus, %G+C estimates from thermal denaturation curves may be performed for any Prokaryote. Prokaryotic taxonomists agree that a reliable classication of a microorganism can only be achieved through its study by a number of different techniques in what is generally known as the polyphasic approach (Vandamme et al., 1996). This implies that both the genomic information and the phenotype of a microorganism should be investigated. The DNA base ratio was the rst genomic technology applied to prokaryote taxonomy (Lee et al., 1956) and has proved to be a useful and routine approach of distinguishing between species with a similar phenotype (Goodfellow and ODonnell, 1993). Empirically, it has been shown that organisms that differ by more than 10 mol% do not belong to the same genus and that species are within a range of 5 mol% (Wayne et al., 1987). Well-dened species usually narrow this range to 3% (Stackebrandt and Liesack, 1993; Vandamme et al., 1996). Species comparisons using the DNA base ratio can only be used to discriminate between different strains, as similar DNA base ratios do not necessarily imply phylogenetic similarity (Rosell-Mora and Amann, 2001). Today, the DNA base ratio, or mol% G+C content, is one of the recommended characteristics for the standard description of prokaryotic species. This study proposes an
2002 Blackwell Science Ltd, Environmental Microbiology, 4, 770773

Fluorimetric determination of %GC 773 easy, rapid, and inexpensive method for estimating the G+C mol% of microorganisms, which could be performed by any researcher interested in the classication of novel microorganisms.
Goodfellow, M., and ODonnell, A.G., (eds), London: Academic Press, pp. 354. Lee, K.Y., Wahl, R., and Barbu, E. (1956) Contenu en bases purique et pyrimidiques des acides deoxyribonucleiques des bacteries. Ann Inst Pasteur 91: 212224. Marmur, J. (1961) A procedure for the isolation of deoxyribonucleic acid from microorganisms. J Mol Biol 3: 208218. Marmur, J., and Doty, P. (1962) Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature. J Mol Biol 5: 109118. Meshbah, M., Premachandran, U., and Whitmann, W. (1989) Precise measurement of the G+C content of deoxyribonucleic acid by high performance liquid chromatography. Int J Syst Bacteriol 39: 159167. Rosell-Mora, R., and Amann, R. (2001) The species concept for prokaryotes. FEMS Microbiol Rev 25: 3667. Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory. Sokal, R.R., and Rohlf, F.J. (1981) Biometry, 2nd edn. New York: W.H. Freeman. Stackebrandt, E., and Liesack (1993) Nucleic acids and classication In Handbook of New Bacterial Systematics. M. Goodfellow and A.G.ODonnell, (eds). Academic Press, London, pp. 151194. Tamaoka, J., and Komagawa, K. (1984) Determination of DNA base composition by reversed-phase highperformance liquid chromatography. FEMS Microbiol Lett 25: 125128. Torsvik, V.F.L., Daae and Goksyr (1995) Extraction, purication, and analysis of DNA from soil bacteria. In Nucleic Acids in the Environment: Methods and Applications. Trevors, J.T., and van Elsas, J.D., (eds). New York: Springer Verlag, pp. 2948. Vandamme, P., Pot, B., Gillis, M., de Vos, P., Kersters, K., and Swings, J. (1996) Polyphasic taxonomy, a consensus approach to bacterial systematics. Microbiol Rev 60: 407 438. Wayne, L.G. et al. (1987) Report of the ad hoc committee on reconciliation of approaches to bacterial systematics. Int J Syst Bacteriol 37: 463464.

Acknowledgements
The authors thank Drs Luz Candenas de Lujan and Francisco Pinto for the use of the iQ iCycler and their valuable assistance and comments. We are greatly appreciated to the CECT (Spanish Culture Collection) for providing with the bacterial species needed for the calibration curves. The authors acknowledge funding from the European projects COALITION (EVK4-CT-199920001) and CATS (EVK4CT-200000028). J.M.G. thanks funding from the Spanish Ministry of Science and Technology, Ramn y Cajal programme.

References
Charbonnier, F.P., and Forterre (1989) Purication of plasmids from thermophilic and hyperthermophilic Archaea. In Archaea, a Laboratory Manual Thermophiles Robb, F.T., and Place, A.R., (eds), Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press, pp. 8790. De Ley, J., Cattoir, H., and Reynaerts, A. (1970) The quantitative measurement of DNA hybridization from renaturation rates. Eur J Biochem 12: 133142. Fiala, G., and Stetter, K.O. (1986) Pyrococcus furiosus sp. nov. represents a novel genus of marine heterotrophic archaebacteria growing optimally at 100C. Arch Microbiol 145: 5661. Gonzlez, J.M., Masuchi, Y., Robb, F.T., Ammerman, J.W., Maeder, D.L., Yanagibayashi, M., et al. (1998) Pyrococcus horikoshii sp. nov., a hyperthermophilic archaeon isolated from a hydrothermal vent at the Okinawa Trough. Extremophiles 2: 123130. Goodfellow, M., and ODonnell (1993) Roots of bacterial systematics. In Handbook of New Bacterial Systematics

2002 Blackwell Science Ltd, Environmental Microbiology, 4, 770773