Biosci. Biotechnol. Biochem., 69 (10), 1995–1998, 2005 Note macerans Possesses Two Types of 16S rDNA Copies in a Genome with a Length Difference of Twelve Base Pairs

y Yoshikatsu HAMASAKI,1 Yoshihisa WATANABE,1 Satoshi KOTOURA,1; Hidetaka FUCHU,2 Masaaki SUGIYAMA,1 Katsuhito HASHIZUME,3 and Hidetoshi MORITA4

1Central Research Institute, Marudai Food Co., Ltd., 21-3 Midori-cho, Takatsuki, Osaka 569-8577, Japan 2Product Development Department, Umeya Co., Ltd., 1-12-14 Mori, Tanabe, Wakayama 646-0023, Japan 3Dragon Genomics Center, Takara Bio Inc., 7870-15 Sakura-cho, Yokkaichi, Mie 512-1211, Japan 4School of Veterinary Medicine, Azabu University, 1-17-71 Fuchinobe, Sagamihara, Kanagawa 229-8501, Japan

Received April 28, 2005; Accepted June 28, 2005

Two Paenibacillus macerans strains, JCM 2500T and ‘‘convenience’’ food containing Chinese-style food, we MCRI 12, exhibited two types of 16S rDNA copies in unexpectedly discovered a new intra-genomic hetero- their genomes, accompanied by a length difference of geneity of Paenibacillus macerans. Paenibacillus spp. 12 bp at positions 203 to 214 (Escherichia coli number- are gram-positive, spore-forming rods belonging to a ing). The long-type sequences were newly identified for class of of the phylum .6) These P. macerans 16S rDNA, and the copy numbers were have been isolated from a variety of sources, different between the two strains. Both types of 16S including soil, water, the plant rhizosphere, food, rRNA were expressed in each strain, and it was diseased insect larvae, and clinical specimens.7) Some predicted that the polymorphism at this position is Paenibacillus spp. have shown intra-genomic 16S rDNA located in helix H10, based on a comparison with the heterogeneity,1,7) but it has not previously been found in E. coli 16S rRNA secondary structure model. P. macerans. We tried to identify MCRI 12 based on 16S rDNA Key words: Paenibacillus macerans; 16S rDNA; 16S sequencing. The strain was grown on nutrient agar rRNA; polymorphism (Difco Laboratories, Detroit, MI) for 48 h at 30 C, and in nutrient broth (Difco) for 24 h at 30 C. Bacterial Acceptance of 16S rRNA as an appropriate tool for DNA samples for sequencing were isolated from a the identification and classification of bacteria has been single colony of the bacteria. Bacterial DNA extracts based on the fact that multiple copies of rRNA genes were prepared using PrepMan ultra reagent (Applied (rDNA) present in a single organism are identical or Biosystems, Foster City, CA). Partial fragments of 16S nearly identical in nucleotide sequences. Most strains rDNA were selectively amplified by PCR using univer- possess multiple 16S rDNA showing no intra-genomic sal primers: 5F (50-TGGAGAGTTTGATCCTGGCT- heterogeneity. When more than a single 16S rDNA from CAG-30) as the forward primer and 531R (50-TACC- one bacterial genome has been analyzed, the sequences GCGGCTGCTGGCAC-30) as the reverse primer.8) PCR determined have generally been found to be identical, or was performed in 50 ml of reaction mixture containing to differ from each other by less than 1% of their 1 ml bacterial DNA extract as the template, 2.5 mM 1) nucleotide positions. Although mutations such as MgCl2, 0.2 mM dNTP, 1 mM of each primer, and 1.25 U transition or transversion are introduced by misincorpo- of AmpliTaq Gold DNA polymerase in GeneAmp gold ration or misrepair by DNA polymerase during DNA buffer (Applied Biosystems). PCR conditions were replication,2) certain cases of intra-genomic heterogene- consistent with the instructions provided with the ity of 16S rDNA cannot be fully explained by these MicroSeq 500 16S rDNA bacterial sequencing kit phenomena. For example, Ochrobactrum intermedium,3) (Applied Biosystems), and sequencing was carried out Streptomyces spp.,2) Thermobispora bispora,4) and with an ABI Prism 310 Genetic Analyzer (Applied Thermomonospora chromogena5) show heterogeneity, Biosystems). differing from those introduced by misincorporation or Direct sequencing gave an ambiguous result. In misrepair. forward sequencing, the sequence at positions 5 to 202 In the process of identification of strain MCRI 12 (E. coli numbering) was determined, but the rest (culture collection of the Marudai Central Research indicated double sequencing signals. Similarly, the Institute), isolated from a commercially available pouch sequence at positions 215 to 531 (E. coli numbering)

y To whom correspondence should be addressed. Tel: +81-72-661-2552; Fax: +81-72-661-2598; E-mail: kotoura [email protected] 1996 Y. HAMASAKI et al.

Fig. 1. ClustalW Alignment of Short and Long Types of 16S rDNA Partial Sequences in Paenibacillus macerans Strains JCM 2500T and MCRI 12, and Specific Primers for Sequencing of Polymorphism Points. Asterisks represent identity between all sequences, and dashes indicate gaps introduced to optimize the alignment. Arrow numbers are positions of E. coli numbering. Forward specific primers (A) and reverse specific primers (B) are listed, with complemented sequences of reverse primers in parentheses. 50-end positions of the primers are indicated as a (Pm191Fa1, Pm191Fa2, and Pm191Fb), b (Pm200F), and c (Pm232Ra and Pm232Rb). was determined in reverse sequencing, but the forward identity of over 99% to P. macerans IAM 12467T part showed double signals. This result suggested the without any single nucleotide polymorphisms. The cells existence of two types of 16S rDNA copies with local of MCRI 12 were gram-positive, spore-forming rods. Its polymorphism containing a gap at positions 203 to 214 biochemical properties were consistent with those of (E. coli numbering). The BLAST program was used to P. macerans according to biochemical tests using the compare the present 16S rDNA sequences with se- API 20E and API 50CHB systems (data not shown).9) quences deposited in the GenBank (http://www.ncbi. Consequently, MCRI 12 was identified as P. macerans. nlm.nih.gov/) and DDBJ (http://www.ddbj.nig.ac.jp/) The long-type sequence was quite similar to the short- databases. We carried out a homology search using the type sequence except at the polymorphism position, BLAST program with the sequences determined in while the particular 18 bp in the long-type sequence was forward and reverse sequencing, and discovered a absent from all Paenibacillus spp. 16S rDNA sequences similarity with P. macerans IAM 12467T (accession available in data banks, except for a single sequence no. AB073196) and Paenibacillus sp. R-7487 (accession obtained from Paenibacillus sp. R-7487. no. AY397771). These two sequences showed different For the P. macerans type strain JCM 2500T (obtained sequences at position 203 to 214 (E. coli numbering) from the Japan Collection of Microorganisms), direct with a length difference of 12 base pairs, indicating sequencing after PCR amplification using primer 5F- identity with double sequencing signals in MCRI 12. 531R did not show double sequencing signals. Never- Based on these results, we designed other primers for theless, PCR amplification products using primer confirmation of sequencing of the polymorphism points. Pm200F-531R, anchored in the particular 18 bp for the The specific primers for sequencing of polymorphism long-type sequence, were also obtained from JCM points are shown in Fig. 1A and B. The target products 2500T (data not shown). Moreover, two sequences were certainly obtained by PCR using the primer sets similar to those of MCRI 12 were also identified in 5F-Pm232Ra, 5F-Pm232Rb, Pm191Fa1-531R, JCM 2500T, using the specific primer sets for the Pm191Fa2-531R, and Pm191Fb-531R. Direct sequenc- polymorphism points (Fig. 1). The DDBJ accession nos. ing with the primer sets showed two types of 16S rDNA, for the 16S rDNA sequences of JCM 2500T are with different sequences at positions 203 to 214 (E. coli AB162431 and AB162432. numbering), as expected (Fig. 1). Alignment was per- The copy number of 16S rDNA was determined by a formed using the ClustalW program (http://www.ddbj. Southern blotting experiment using a 16S rDNA probe nig.ac.jp/). The DDBJ accession nos. for the 16S rDNA on genomic DNAs digested with HindIII and/or EcoRI sequences of MCRI 12 are AB162433 and AB162434. (restriction sites for both HindIII and EcoRI are absent A BLAST analysis of the short-type sequence in P. macerans 16S rDNA) (Fig. 2A). Genomic DNA containing 6 bp at the polymorphism position showed was purified with a G NOME DNA kit (Qbiogene, P. macerans Possesses Two Types of 16S rDNA Copies 1997 Intra-genomic heterogeneity of 16S rDNA appears as double sequencing signals in direct sequencing, such as that found by Teyssier et al.3) and Marchandin et al.10) for O. intermedium and Veillonella spp. respectively. In the present study, double sequencing signals appeared in direct sequencing with universal primers for MCRI 12. At first, only single sequencing signals were found for JCM 2500T because of the small number of long-type copies in this strain; by contrast, MCRI 12 possesses many more of these genes. However, when the direct sequencing results for JCM 2500T were examined carefully, feeble double sequencing signals were iden- tified. It is believed that the ratio of the numbers of the different copies in a 16S rDNA polymorphism is reflected in the signal intensity found in direct sequenc- ing. We amplified 16S rRNA molecules selectively by RT-PCR for short- and long-type sequences using primer sets 5F-Pm232Ra and 5F-Pm232Rb respectively on total-RNA extracts (Fig. 3). Total bacterial RNA was isolated and purified using the SV total RNA isolation system (Promega, Madison, WI), and reverse tran- scription-PCR (RT-PCR) was performed with the AccessQuick RT-PCR system (Promega). Non-reverse- Fig. 2. Southern Blot Analysis of Short- and Long-Type Copies of transcribed RNA was used as the PCR template for the 16S rDNA (A) and Only Long-Type Copies (B) in Each Genome of Paenibacillus macerans JCM 2500T (lanes J) and MCRI 12 (lanes negative control. RT-PCR products for both short- and M). long-type 16S rRNAs were obtained from each strain, Shown are Southern blots of genomic DNA digested by restriction and we concluded that both types of 16S rRNA are enzyme, using only HindIII (H), EcoRI (E), and both HindIII and expressed in each strain. Prediction of RNA secondary EcoRI (H þ E). Hybridizations were performed with probes of 5F- structure by energy minimization was performed with 531R PCR products (A) and Pm200F (B). Positions of molecular weight markers are indicated to the left. the on-line Mfold server at http://mfold.burnet.edu.au/. The polymorphism position was predicted to be located in helix H10, based on comparison with the E. coli 16S Carlsbad, CA). DNAs digested with HindIII and/or rRNA secondary structure model. Veillonella spp. show EcoRI were subjected to electrophoresis for 1.5 h at 50 V a difference in sequence but not a difference in length at in a 0.8% agarose gel in TBE using Mupid gel electrophoresis (Advance, Tokyo). Electrophoresis gels were transferred to Hybond Nþ nylon membranes (Amersham Biosciences, Little Chalfont, UK) by alka- line blotting. 16S rDNA digoxigenin (DIG)-labeled probes were obtained by PCR using 50-end DIG-labeled primers of 5F-531R (Kurabo Industries, Osaka, Japan). Hybridization of the probes was detected with a DIG nucleic acid detection kit (Roche Diagnostics, Mann- heim, Germany). At least eight 16S rDNA hybridizing fragments (rrs operons) were found in each strain according to the fragment patterns of double digestion with HindIII and EcoRI for JCM 2500T, and of EcoRI digestion for MCRI 12. The copy number of the long- type sequence was determined using the oligonucleotide probe corresponding to the particular 18 nucleotides Fig. 3. Detection of the Two Types of 16S rRNA by RT-PCR in 0 (Fig. 2B). A 5 -end DIG-labeled probe of Pm200F Strains JCM 2500T and MCRI 12. (Kurabo) was used as the oligonucleotide probe. MCRI Lanes S, use of primers 5F-Pm232Ra for short-type copies; lanes 12 had five long-type copies based on the double L, use of primers 5F-Pm232Rb for long-type copies; lane N, a digestion fragment patterns obtained with HindIII and negative control performed on the RNA of strain MCRI 12 without T reverse transcription and with primers 5F-Pm232Rb (negative EcoRI. JCM 2500 had two copies of the long type, controls were also obtained for strain JCM 2500T and for primers according to the fragment patterns of HindIII and EcoRI 5F-Pm232Ra, but are not shown); lane M, 100 bp DNA ladder as a digestion. molecular weight marker. 1998 Y. HAMASAKI et al. the same helix10) and Ochrobactrum intermedium 3) Teyssier, C., Marchandin, H., Simeon De Buochberg, exhibits a polymorphism with a length difference of 42 M., Ramuz, M., and Jumas-Bilak, E., Atypical 16S nucleotides for helix H9, a near neighbor of H10.3) rRNA gene copies in Ochrobactrum intermedium strains Based on the E. coli 16S rRNA secondary structure reveal a large genomic rearrangement by recombination model, a polymorphism accompanied with a long length between rrn copies. J. Bacteriol., 185, 2901–2909 (2003). difference was found for helix H10 in the present study, 4) Wang, Y., Zhang, Z., and Ramanan, N., The actino- in contrast with previously detected 16S rRNA poly- mycete Thermobispora bispora contains two distinct morphisms. In the Database of Ribosomal Cross- types of transcriptionally active 16S rRNA genes. J. 11) Links, there are no known interactions with rRNA Bacteriol., 179, 3270–3276 (1997). or ribosomal proteins listed for helix H10. Furthermore, 5) Yap, W. H., Zhang, Z., and Wang, Y., Distinct types of H10 is located far from the active center of the 30S rRNA operons exist in the genome of the actinomycete subunit, which is located on the intersubunit side, Thermomonospora chromogena and evidence for hori- including the decoding center, and is instead positioned zontal transfer of an entire rRNA operon. J. Bacteriol., near the spur or foot of the 30S subunit on the solvent 181, 5201–5209 (1999). side.12) We found no difference in certain phenotypical 6) Ash, C., Priest, F. G., and Collins, M. D., Molecular traits such as colony aspect, morphology, and growth identification of rRNA group 3 bacilli (Ash, Farrow, Wallbanks and Collins) using a PCR probe test. Proposal properties between strains JCM 2500T and MCRI 12. for the creation of a new genus Paenibacillus. Antonie For the reasons stated above, it is believed that Van Leeuwenhoek, 64, 253–260 (1993). extensions of H10 have little effect on ribosome 7) Bosshard, P. P., Zbinden, R., and Altwegg, M., Paeni- function; this speculation fits with the low conservation turicensis sp. nov., a novel bacterium harbour- within this area between 16S rRNA from all bacteria in ing heterogeneities between 16S rRNA genes. Int. J. the Gutell laboratory database (http://www.rna.icmb. Syst. Evol. Microbiol., 52, 2241–2249 (2002). utexas.edu/). 8) Hall, L., Doerr, K. A., Wohlfiel, S. L., and Roberts, G. In bacterial type strains, large scale intra-genomic D., Evaluation of the MicroSeq system for identification heterogeneity is generally recognized as an unusual of mycobacteria by 16S ribosomal DNA sequencing and property, though it has been found for Thermobispora its integration into a routine clinical mycobacteriology bispora4) and Thermomonospora chromogena.5) In the laboratory. J. Clin. Microbiol., 41, 1447–1453 (2003). 9) Heyndrickx, M., Vandemeulebroecke, K., Scheldeman, present study, two types of 16S rDNA copies were found P., Kersters, K., de Vos, P., Logan, N. A., Aziz, A. M., from P. macerans, including a long-type sequence Ali, N., and Berkeley, R. C., A polyphasic reassessment which was newly identified from the type strain. Since of the genus Paenibacillus, reclassification of Bacillus minority copies of 16S rDNA are not sequenced by lautus (Nakamura 1984) as Paenibacillus lautus comb. direct sequencing using universal primers, 16S rDNA nov. and of Bacillus peoriae (Montefusco et al. 1993) as copies in a single bacterium might be more heteroge- Paenibacillus peoriae comb. nov., and emended de- neous than recognized at present. For identification scriptions of P. lautus and of P. peoriae. Int. J. Syst. using 16S rDNA information, the possible existence of Bacteriol., 46, 988–1003 (1996). other bacteria possessing large-scale intra-genomic 10) Marchandin, H., Teyssier, C., Simeon De Buochberg, heterogeneities should receive further attention. M., Jean-Pierre, H., Carriere, C., and Jumas-Bilak, E., Intra-chromosomal heterogeneity between the four 16S rRNA gene copies in the genus Veillonella: implications References for phylogeny and . Microbiology, 149, 1493– 1501 (2003). 1) Nubel, U., Engelen, B., Felske, A., Snaidr, J., 11) Baranov, P. V., Kubarenko, A. V., Gurvich, O. L., Wieshuber, A., Amann, R. I., Ludwig, W., and Shamolina, T. A., and Brimacombe, R., The Database of Backhaus, H., Sequence heterogeneities of genes encod- ribosomal cross-links: an update. Nucl. Acids Res., 27, ing 16S rRNAs in Paenibacillus polymyxa detected by 184–185 (1999). temperature gradient gel electrophoresis. J. Bacteriol., 12) Wimberly, B. T., Brodersen, D. E., Clemons, W. M., Jr., 178, 5636–5643 (1996). Morgan-Warren, R. J., Carter, A. P., Vonrhein, C., 2) Ueda, K., Seki, T., Kudo, T., Yoshida, T., and Kataoka, Hartsch, T., and Ramakrishnan, V., Structure of the 30S M., Two distinct mechanisms cause heterogeneity of 16S ribosomal subunit. Nature, 407, 327–339 (2000). rRNA. J. Bacteriol., 181, 78–82 (1999).