ARTICLE IN PRESS

Systematic and Applied Microbiology 33 (2010) 53–59

Contents lists available at ScienceDirect

Systematic and Applied Microbiology

0723-20

doi:10.1

n Corr

E-m

journal homepage: www.elsevier.de/syapm

Molecular detection and phylogenetic analysis of the alkane1-monooxygenase gene from Gordonia spp.

Fo-Ting Shen a, Li-Sen Young b, Ming-Fang Hsieh c, Shih-Yao Lin c, Chiu-Chung Young c,n

a Center for Environmental Restoration and Disaster Reduction (CERDR), National Chung Hsing University, Taichung, Taiwan, ROCb Department of Biotechnology, College of Health Sciences, Yuanpei University, Hsinchu, Taiwan, ROCc Department of Soil and Environmental Sciences, College of Agriculture and Natural Resources, National Chung Hsing University, 250, Kuo Kuang Rd., Taichung 402, Taiwan, ROC

a r t i c l e i n f o

Article history:

Received 3 November 2009

Keywords:

Alkane 1-monooxygenase gene (alkB gene)

Gordonia

Phylogeny

20/$ - see front matter & 2009 Elsevier Gmb

016/j.syapm.2009.11.003

esponding author. Tel./fax: +886 4 22861495

ail address: ccyoung@mail.nchu.edu.tw (C.-C.

a b s t r a c t

The alkB gene encodes for alkane 1-monooxygenase, which is a key enzyme responsible for the initial

oxidation of inactivated alkanes. This functional gene can be used as a marker to assess the catabolic

potential of bacteria in bioremediation. In the present study, a pair of primers was designed based on

the conserved regions of the AlkB amino acid sequences of Actinobacteria, for amplifying the alkB gene

from the genus Gordonia (20 Gordonia strains representing 13 species). The amplified alkB genes were

then sequenced and analyzed. In the phylogenetic tree based on the translated AlkB amino acid

sequences, all the Gordonia segregated clearly from other closely related genera. The sequence identity

of the alkB gene in Gordonia ranged from 58.8% to 99.1%, which showed higher sequence variation at the

inter-species level compared with other molecular markers, such as the 16S rRNA gene (93.1–99.8%),

gyrB gene (77.5–97.3%) or catA gene (72.4–99.5%). The genetic diversity of four selected loci also

showed that the alkB gene might have evolved faster than rrn operons, as well as the gyrB or catA genes,

in Gordonia. All the available actinobacterial alkB gene sequences derived from the whole genome

shotgun sequencing projects are phylogenetically characterized here for the first time, and they exclude

the possibility of horizontal gene transfer of the alkB gene in these bacterial groups.

& 2009 Elsevier GmbH. All rights reserved.

Introduction

Members of the Gram-positive GC-rich Actinobacteria arepresent in various biotopes, especially in oil or polycyclic aromatichydrocarbon contaminated soils [14,16]. Their catabolic activitiesand rich mycolic acids in the cell wall contribute to thedegradation of the pollutants. Among the Actinobacteria, themetabolically diverse genus Gordonia has been recognized ascontaining valuable microorganisms in recent years. Most of thespecies present in this genus have been isolated for their abilitieseither to degrade environmental pollutants, xenobiotics or otherslowly biodegradable natural polymers [1,30]. The characteriza-tion of the catabolic potential of such types of bacteria seemsimportant when in situ bioremediation has been applied. Theadoption of nucleic acid-based technologies makes it possible toassess the biodegradation potential or microbial diversity inpetroleum hydrocarbon contaminated soils by detecting thecatabolic genes within indigenous microorganisms [2,3,12].Several hydrocarbon catabolic genes encoding for benzenedi- and monooxygenase, toluene di- and monooxygenase, xylene

H. All rights reserved.

.

Young).

monooxygenase, and naphthalene dioxygenase have been studiedand used in the development of oligonucleotide microarrays.However, most of the probes designed were group-specific andmainly focused on Gram-negative bacteria, such as Acinetobacter,Burkholderia, Pseudomonas, Sphingomonas and Gram-positivebacteria such as Rhodococcus[8,9]. The high diversity of catabolicgenes among different bacterial taxa and low availability ofsequence information limit the exploration of numerous catabolicgenes in the environment by hybridization or PCR detectionmethods [12].

The alkane omega-hydroxylase (AlkB) system in Gram-nega-tive bacteria encodes for a three-component alkane hydroxylasecomplex, consisting of a particular non-heme integral-membranedi-iron alkane 1-monooxygenase (AlkB) and two soluble proteins,rubredoxin (AlkG) and rubredoxin reductase (AlkT) [26]. InRhodococcus, a similar organization in the alkane hydroxylasegene cluster has been found, which includes various homologuesof alkB and rubA followed by the rubB gene [28]. In order to studythe presence of the alkB gene in the members of the genusGordonia and explore the diversity of this catabolic gene, it isessential to establish a PCR-based amplification method. In thepresent study, the degenerate primers for the amplification of thealkB gene from Actinobacteria were developed. A total of 20 alkB

gene sequences of strains affiliated to the genus Gordonia were

ARTICLE IN PRESS

F.-T. Shen et al. / Systematic and Applied Microbiology 33 (2010) 53–5954

phylogenetically characterized. In addition, the genetic variationsamong Gordonia species and the alkB gene heterogeneity inActinobacteria are also discussed.

Materials and methods

Bacterial strains and culture condition

The bacterial strains used in this study are listed in Table 1.Type strains were purchased from the Bioresource Collection andResearch Center (BCRC), Food Industry Research and DevelopmentInstitute, Taiwan, and from Deutsche Sammlung vonMikroorganismen und Zellkulturen (DSMZ), Germany. Otherreference strains were retrieved from our laboratory collection[17,20,29]. All strains were cultivated on tryptic soy agar(HIMEDIA, India) at 30 1C.

Design of oligonucleotide primers

A total of 32 alkB gene sequences retrieved from UniProtKnowledgebase (Swiss-Prot and TrEMBL) and NCBI GenBank werealigned and compared using the CLUSTAL_X 1.83 program(multiple parameters: gap opening: 15, gap extension: 6.66, delaydivergent sequences: 30%, DNA transition weight: 0.5) [25]. TheGonnet series was used for protein weight matrix constructionand CLUSTALW was used for DNA weight matrix construction. APCR primer pair, alkBF and alkBR, was designed based on twoconserved regions in the amino acid sequence of actinobacterialspecies. The forward primer alkBF was 30 mer oligonucleotides(50-ATC AAY RCV GCV CAY GAR YTV GGB CAC AAG-30), targetingamino acid sequence INTAHELGHK (Rhodococcus erythropolis

NRRL B-16531 AlkB2 amino acid (AJ297269) position 154–163).The reverse primer alkBR was 30 mer oligonucleotides (50-SGGRTT CGC RTG RTG RTC RCT GTG NSG YTG-30), targeting amino acidsequence Q(A/P)HSDHHANP (Rhodococcus erythropolis NRRL

Table 1Details of the strains used in the present study with their alkB gene accession

numbers.

Species and strain designation Accession

no.

Isolation source

Gordonia alkanivorans DSM 44369T GU130258 Tar contaminated soil

Gordonia amarae DSM 43392T GU130259 Foaming activated sludge

Gordonia amicalis DSM 44461T GU130260 Garden soil

Gordonia bronchialis DSM 43247T GU130261 Patient sputum

Gordonia desulfuricans DSM

44462T

GU130262 Oil shale spoil heap

Gordonia hydrophobica DSM

44015T

GU130263 Compost biofilter

Gordonia malaquae DSM 45064T GU130264 Wastewater sludge

Gordonia rhizosphera DSM 44383T GU130265 Rhizosphere soil

Gordonia rubripertincta DSM

43197T

GU130266 Soil

Gordonia soli DSM 44995T GU130267 Soil

Gordonia sputi DSM 43896T GU130268 Patient sputum

Gordonia terrae DSM 43249T GU130269 Soil

Gordonia westfalica DSM 44215T GU130270 Fouling tyre water

Gordonia sp. CC-JL2-2 GU130273 Heavy oil contaminated

soil

Gordonia sp. CC-JL4 GU130274 Heavy oil contaminated

soil

Gordonia sp. CC-KS2 GU130275 Mountain soil

Gordonia sp. CC-MJ-39a GU130276 Heavy oil contaminated

soil

Gordonia sp. CC-S2a GU130277 Mountain soil

Gordonia sp. CC-S5-7 GU130278 Mountain soil

Gordonia sp. CC-S5a GU130279 Mountain soil

B-16531 AlkB2 amino acid position 330–339). The regions containamino acid 158 [HELGHK] and amino acid 332 [HSDHH] referringto the first and the third histidine motif, which are conserved inthe class of membrane-bound hydroxylases and desaturases [21].Theoretically, the amplified fragment had a length of 558 bp.

DNA extraction and PCR amplification of the alkB gene

Bacterial genomic DNA from 20 Gordonia strains (Table 1) wasisolated using UltraClean Microbial Genomic DNA isolation kits(MO BIO, USA). Cultures were incubated in tryptic soy broth at30 1C for 48 h and the DNA extraction was carried out accordingto protocols supplied by the manufacturers. These DNA prepara-tions were used as template DNA for the amplification of the alkB

gene by PCR using the primer pair alkBF and alkBR under thefollowing conditions: 25 mL PCR mixture containing 0.2 mM ofeach of the four dNTP, 20 pmol of each primer, 2 mL of extractedDNA and 2 units of ProTaq DNA polymerase (Promega) withappropriate reaction buffer. Amplification was performed in aGeneAmps System 9700 thermal cycler (Applied Biosystem,USA). The reaction conditions for the amplification of the partialalkB gene were as follows: initial denaturation for 5 min at 95 1C,followed by 30 cycles of 30 s at 95 1C, 45 s at 66 1C and 60 s at72 1C, with a final extension of 7 min at 72 1C. All amplificationproducts were checked by electrophoresis on 1% agarose gelsstained with ethidium bromide.

Sequencing of the alkB gene and phylogenetic analysis

The partially amplified alkB gene fragments were purified fromagarose gel using the QIAquick gel extraction kit (Qiagen Inc.,Chatsworth, CA, USA) and directly sequenced or cloned in E. coli

JM109 by using the yT&A cloning vector system (Yeastern Biotech,Taiwan). Colonies were picked, lysed and used to selecttransformants for plasmid extraction, which were sequencedfurther by the alkBF or alkBR primer. Determination of thenucleotide sequence of the PCR products was performed by anautomatic genetic analyzer (ABI PRISM 310, Applied Biosystems,CA, USA) [27]. The alkB gene sequences obtained were depositedin the GenBank nucleotide sequence database (NCBI) underaccession numbers GU130258-GU130279. The translated AlkBamino acid sequences deduced from the alkB nucleotidesequences were aligned and compared with other AlkB sequencesof Actinobacteria available in the GenBank database (1 Dietzia,5 Gordonia, 13 Mycobacterium, 1 Nocardia, 2 Nocardioides,8 Rhodococcus and 1 Tsukamurella strains) using the CLUSTAL_X1.83 program [25] and the aligning parameters described above.

The stability of the phylogenetic trees was tested by using theneighbor-joining, maximum parsimony and maximum likelihoodmethods. Distances and clustering with the neighbor-joining andmaximum parsimony methods were performed by using thesoftware package MEGA (Molecular Evolutionary Genetics Analysis)version 4.0 [24]. The Poisson correction model was used for distanceanalysis. The substitution rates were set the same among the sitesand among the lineages. Gaps were treated by complete deletion.The stability of the phylogenetic trees was also tested using RAxMLversion 7.0 with the General Time Reversible (GTR) substitutionmatrix, according to the maximum likelihood model [22]. Bootstrapvalues based on 1000 replications were listed as percentages at thebranching points. Although the lengths of the aligned AlkB aminoacid sequences were different, the topology of the phylogenetic treereconstructed was similar to that based on the same length ofaligned sequences. The longer length of amino acid sequences ofRhodococcus and Mycobacterium spp. was used in order to give moreaccurate taxonomic characterization of these strains. The 16S rRNA

ARTICLE IN PRESS

F.-T. Shen et al. / Systematic and Applied Microbiology 33 (2010) 53–59 55

gene sequences were also aligned and used in the reconstruction ofthe phylogenetic tree. Sequence distances of the alkB, 16S rRNA, gyrB

or catA genes among Gordonia species were calculated by using thesoftware MegAlign (DNASTAR Inc., Madison, WI, USA). Geneticdiversity of the selected loci among Gordonia species was calculatedby using the DnaSP software package [11].

Results and discussion

Detection and sequencing of the alkB gene in Gordonia

Genomic DNAs isolated from 20 Gordonia strains were used astemplates for the amplification of the alkB gene. By using theprimer pair alkBF and alkBR, the alkB gene fragment (558 bp) wassuccessfully amplified from 17 Gordonia strains representing 10species. Three species, namely G. alkanivorans, G. rhizosphere andG. rubripertincta, gave a negative reaction in the amplification ofthe alkB gene. In order to confirm the absence of the alkB gene inthese three species, we attempted to design another forwardprimer alkBFS based on the alkB gene sequences obtained fromthe present study. The primer alkBFS was 24 mer oligonucleotides(50-GAC GAY CTC GAG CGC TGG CTG TCG-30), targeting amino acidsequence DDLERWLS (Rhodococcus erythropolis NRRL B-16531AlkB2 amino acid position 165–172). By using primer pair alkBFSand alkBR, the amplicon with the corresponding size (525 bp) wassuccessfully obtained in these three Gordonia species. Althoughthe Gordonia strains were originally isolated from variousecosystems, such as oil contaminated soil, wastewater, nativebiotopes or clinical samples (Table 1), all the Gordonia strainstested here possessed the alkB gene. A total of 19 alkB genesequences out of 20 were successfully obtained by directsequencing the PCR product with primer alkBF or alkBR. However,the unclear base recognition of the alkB gene sequence ofG. amarae DSM 43392T was shown when directly sequencingthe PCR amplicon. After the PCR product was cloned and furthersequenced, the alkB gene sequence was successfully obtained.

Due to the high diversity of catabolic genes in bacteria,Hamann et al. [7] proposed a need for different sets of primersfor phylogenetically unrelated catabolic genes. The group-specificprimer sets were designed to amplify phylogenetically distinctgroups of alkB genes from the Acinetobacter, Burkholderia,Nocardioides, Pseudomonas and Rhodococcus lineage, which madeit possible to screen known alkB phylogenetic types and resolvein situ genotypic diversity within the alkB lineages [6,14,21]. Theincreasingly available alkB gene sequences in GenBank motivatedus to design a suitable primer set for rapid detection andamplification of the alkB gene from a variety of Gordonia

members. Forward primers alkBF or alkBFS in combination withreverse primer alkBR were successfully designed and verified withpure cultures from Gordonia species, which will be used in theassessment of the catabolic potential of Gordonia in hydrocarboncontaminated sites.

Phylogenetic relationship and sequence identity of the Gordonia

alkB gene

All the 25 Gordonia members were split into two clusters in theAlkB phylogenetic trees, no matter which method was used forthe tree reconstruction (only the phylogenetic tree reconstructedby the neighbor-joining method is shown in Fig. 1a). TheG. bronchialis and G. sputi group in cluster 2 were more closelyrelated to Nocardioides species in the AlkB phylogenetic tree. Allthe available 16S rRNA gene sequences from the actinobacterialstrains listed in Fig. 1a were used to reconstruct the phylogenetic

tree shown in Fig. 1b. In the phylogenetic trees reconstructedbased on the AlkB amino acid or 16S rRNA gene sequences, all theGordonia members could be differentiated from other genera,including Dietzia, Mycobacterium, Nocardia, Nocardioides,Rhodococcus and Tsukamurella. Three environmental isolates,namely CC-JL2-2, CC-KS2, and CC-JL4, grouping together with G.

amicalis DSM 44461T and G. rubripertincta DSM 43197T in the 16SrRNA gene phylogenetic tree were also affiliated to G. amicalis andG. rubripertincta based on AlkB amino acid sequences. Strains TF6,CC-S2a and G. terrae DSM 43249T formed a single cluster in boththe phylogenetic trees. Gordonia sp. IFP 2009 grouped togetherwith G. hydrophobica DSM 44015T in both the phylogenetic trees.G. bronchialis and G. sputi clustered together and were clearlyseparate from all other Gordonia members in the AlkBphylogenetic tree, and this might be related to their isolationsource from sputum samples. Since the sequence identities of thealkB gene between Gordonia cluster 2 (G. bronchialis and G. sputi)and Gordonia cluster 1 (including 11 species) were low (rangingfrom 58.8% to 70.4%), we speculate that their catabolic potentialor the substrate utilization patterns of alkane 1-monooxygenasemight be different. All the AlkB and 16S rRNA gene sequences areavailable as on-line Supplementary material.

When the relationships between various Mycobacterium

species were compared, the topology based on the AlkB aminoacid was also similar to the 16S rRNA gene-based phylogenetictree (Fig. 1a and b). For example, M. avium and M. intracellulare

formed a single cluster, M. marinum and M. ulcerans groupedtogether, and M. austroafricanum was most closely related toM. vanbaalenii, M. gilvum and M. smegmatis. In addition, Rhodococcus

strains belonging to the same species also formed single clustersin the two phylogenetic trees. Consequently, it is proposed thatthe AlkB amino acid sequence could be used to complement theoverall phylogenetic information for these actinomycetes.

The alkB or the 16S rRNA gene sequences of 13 Gordonia

species were aligned, and the percentage identities between thetype strains are listed in Table 2. Sequence similarities of the alkB

gene at the inter-species level ranged from 58.8% (G. soli andG. sputi) to 99.1% (G. alkanivorans and G. rubripertincta), while the16S rRNA gene sequence similarities ranged from 94.8% to 99.7%.When all the available 16S rRNA, gyrB and catA gene sequences ofGordonia type strains were taken into consideration, the 16S rRNAgene sequence similarities between 29 Gordonia species rangedfrom 93.1% to 99.8%, gyrB gene sequence similarities between23 Gordonia species ranged from 77.5% to 97.3%, and the catA genesequence similarities between 11 Gordonia species ranged from72.4% to 99.5% (data not shown). The alkB gene sequences showedhigher variation than other molecular markers such as the 16SrRNA, gyrB or catA genes in the genus Gordonia.

In the present study the alkane 1-monooxygenases of Gordonia

species have been shown to evolve independently from otherActinobacteria, and we speculate that Gordonia might also differfrom other closely related species in alkane catabolic activity. Theisolates belonging to the genus Gordonia have been shown todegrade a wide range of n-alkanes (C6, C8, C10, C16 and C20) andc-alkanes, and the substrate utilization pattern of short-chainn-alkane is different in Rhodococcus members [10]. Kubota et al.[10] demonstrated that the Acinetobacter isolates degradedlong-chain n-alkane but did not degrade short-chain n-alkane orc-alkanes, and the possibly diverse degrading behavior of variousGordonia species needs to be clarified in the near future.

Genetic variations among Gordonia species

Protein-encoding genes, such as gyrB and catA, have beenshown to evolve faster than rrn operons and provide better

ARTICLE IN PRESS

F.-T. Shen et al. / Systematic and Applied Microbiology 33 (2010) 53–5956

taxonomic resolution among Gordonia [18,19]. Here, we haveattempted to study the gene variation of Gordonia alkB at theinter-species level. From the aligned alkB gene sequence data, thenucleotide polymorphism was analyzed and compared with thatof catA, gyrB and 16S rRNA genes, which were described in ourprevious study [18]. The alkB gene was highly polymorphic and213 positions out of 402 (53.0%) were polymorphic sites, while

Fig. 1. Phylogenetic analysis of 51 actinobacterial strains using the neighbor-joining m

they were 164/378 (43.4%), 369/1165 (31.7%) and 73/1452 (5.0%)for the catA, gyrB and 16S rRNA genes, respectively [18]. Therewere 171 (42.5%) alkB gene parsimony informative sites com-pared to 138 in the catA gene (36.5%), 202 in the gyrB gene (17.3%)and 40 in the 16S rRNA gene (2.8%). The average numbers ofnucleotide differences per site (nucleotide diversity) in alkB, catA,gyrB and 16S rRNA genes were 0.20723, 0.18577, 0.09408 and

ethod based on (a) AlkB amino acid sequences and (b) 16S rRNA gene sequences.

ARTICLE IN PRESS

Fig. 1. (Continued)

F.-T. Shen et al. / Systematic and Applied Microbiology 33 (2010) 53–59 57

0.01479, respectively, indicating that the alkB gene might haveevolved faster than catA, gyrB or 16S rRNA genes in Gordonia.

Tajima’s D-test was used to test the hypothesis that allmutations were selectively neutral in Gordonia. The D-test isbased on the differences between the number of segregating sitesand the average number of nucleotide differences [23]. Tajima’svalue was �0.64976 (insignificant, P40.10) for alkB, �0.96417for gyrB and �1.07432 for 16S rRNA genes. The negative valuessuggest the effect of a stabilizing selection (purifying selection)

event, which might have lowered the frequency of alleles having adeleterious effect on the phenotype, thereby leading to theconservation of genes as a selective pressure against deleteriousvariants.

In the relative synonymous codon usage (RSCU) [15], thecodon preference between the alkB, catA and gyrB genes can beobserved. The pattern of codon preference is more similarbetween the two catabolic genes alkB and catA than the gyrB

gene, for which the same codon preferences in both genes are

ARTICLE IN PRESS

Table 2Sequence identity between the alkB gene nucleotide sequence (upper right) and 16S rRNA gene sequence (lower left).

Type strain Percentage identity (%)

1 2 3 4 5 6 7 8 9 10 11 12 13

1. Gordonia alkanivorans 74.1 95.9 61.3 72.0 74.7 80.3 93.1 99.1 77.8 62.0 84.6 92.4

2. Gordonia amarae 97.0 72.6 68.0 83.3 86.2 77.5 76.5 73.4 68.9 67.8 71.4 80.7

3. Gordonia amicalis 98.3 96.9 60.5 72.2 73.5 79.8 89.2 95.7 80.0 61.8 86.1 89.8

4. Gordonia bronchialis 97.7 97.0 97.7 69.1 70.2 63.7 65.0 61.3 61.8 82.0 61.6 68.0

5. Gordonia desulfuricans 97.9 97.2 98.4 98.0 83.1 74.4 76.7 72.2 74.7 70.4 71.3 81.3

6. Gordonia hydrophobica 97.6 97.5 97.3 97.1 97.4 77.5 77.3 74.3 73.2 69.1 73.5 81.3

7. Gordonia malaquae 96.0 95.8 96.7 95.7 97.0 97.0 82.0 80.6 75.8 63.9 77.9 82.8

8. Gordonia rhizosphera 96.8 96.7 96.8 98.2 97.2 95.9 95.7 92.7 76.1 66.3 84.6 95.7

9. Gordonia rubripertincta 99.0 97.0 98.9 98.3 98.5 97.9 96.0 97.0 78.0 62.4 84.8 92.0

10. Gordonia soli 97.2 96.8 97.6 98.0 97.4 96.3 95.2 97.5 97.0 58.8 79.0 76.1

11. Gordonia sputi 96.7 96.6 96.5 97.6 97.3 96.7 94.8 97.0 97.2 96.6 61.2 69.7

12. Gordonia terrae 97.6 97.2 97.6 98.3 97.6 97.6 96.0 97.5 98.0 97.3 96.8 84.8

13. Gordonia westfalica 99.7 96.8 98.3 97.8 97.8 97.8 96.5 96.7 99.1 97.2 96.5 98.0

F.-T. Shen et al. / Systematic and Applied Microbiology 33 (2010) 53–5958

listed below: ACC for threonine (T); CAG for glutamine (Q); AAGfor lysine (K); GAC for aspartic acid (D); CGC and CGU for arginine(R). The same codon preferences in three genes among Gordonia

include UUC for phenylalanine (F), AUC for isoleucine (I), CCG forproline (P) and GAG for glutamic acid (E). The codon bias index(CBI) [13] in alkB for all the 13 Gordonia was higher (0.747) thanthat of gyrB (0.378) but lower than that of catA (0.772). The G+Ccontents calculated based on alkB, catA and gyrB coding positions,as well as 16S rRNA gene non-coding positions were 64.9%, 69.6%,65.6% and 57.2%, respectively, which are representative of highG+C mol% Gram-positive bacteria.

Heterogeneous alkB gene in Actinobacteria

Whyte et al. [28] have described the multiple alkane hydro-xylase systems in Rhodococcus strains Q15 and NRRL B-16531 byusing the cloning approach. In their studies, four copies, namelyalkB1, alkB2, alkB3 and alkB4 were analyzed in both theRhodococcus strains. We have attempted to determine thepossibility of multiple copies of the alkB gene in Gordonia closelyrelated genera, such as in Rhodococcus, Mycobacterium andTsukamurella species. After retrieving and analyzing the alkB genefrom 19 genome sequences by in silico analyses, it was possible toobtain the copy number of the alkB gene present in the genomes.There were 4 copies of alkB genes in R. erythropolis PR4, 3 inR. erythropolis SK121, 1 in R. jostii, 1 in R. opacus and 3 in N.

farcinica. In T. paurometabola, 2 Nocardioides strains and Dietzia sp.E1 there was only one copy of the alkB gene. When the genusMycobacterium was considered, 1 or 2 alkB gene copies could befound in 11 Mycobacterium species. Quatrini et al. [14] found 3alkB homologues in Nocardia sp. SoB, while a single alkB gene wasfound in Rhodococcus sp. SoD, Rhodococcus sp. SoF, Gordonia sp.SoCg and Gordonia sp. SoCp, after three or four clones weresequenced, respectively (Table S1). However, they could notexclude the presence of multiple genes that were not detectedwith the PCR and cloning method, and suggested that furtherstudies are needed to clarify if they possess other alkanehydroxylase systems. To date, there are still no records ofmultiple alkB gene sequences in Gordonia members in the publicdatabase (Table S1). In the present study, most of the alkB genesfrom Gordonia species were successfully amplified and directlysequenced, which might raise the possibility of a single copy ofthe alkB gene in these tested Gordonia species. After screening forthe alkB gene in the complete plasmid sequences from members

belonging to suborder Corynebacterineae, only a single alkB genecould be found in a 558 kbp mega-plasmid pROB01 sequence(NC_012520) of Rhodococcus opacus. This alkB gene located in theplasmid showed the highest amino acid sequence identity (82%)with that of the same strain in its chromosome, and thedownstream genes were the rubA (76% between chromosomeand plasmid), rubA (81%) and rubB (54%) genes. The gene encodingfor the alkane 1-monooxygenase was not found in the plasmidsequences from M. abscessus (NC_010394), M. gilvum (NC_009339,NC_009340, NC_009341), M. marinum (NC_010604), M. ulcerans

(NC_005916), N. farcinica (NC_006362, NC_006363), Nocardioides

sp. JS614 (NC_008697), R. erythropolis PR4 (NC_007486,NC_007487, NC_007491) and R. jostii (NC_008269, NC_008270,NC_008271). The higher alkB gene sequence variations betweenspecies might exclude the possibility of horizontal gene transferof the alkB gene in these Actinobacteria, which is an importantcharacteristic useful as a molecular chronometer.

All the alkB gene partial sequences obtained in this study wererelated to the alkB2 type [28]. In R. erythropolis, alkB2 forms part ofa cluster with rubredoxin and rubredoxin reductase genes, and isthe only one for which heterologous expression in E. coli andPseudomonas was obtained. Similarly, the AlkB2 system ofGordonia sp. TF6 that is able to degrade n-alkanes from C5 toC16, was cloned and expressed in E. coli, which produced theconversion of n-alkanes with 5–13 carbons to their correspondingalcohols [4]. The catabolic patterns of the monooxygenase inGordonia species toward various alkane substrates will be neededin order to clarify the metabolic diversity of the alkane hydro-xylase systems present in this unique actinobacterial clade.

The phylogenetic analysis based on multilocus sequences hasproved to be more accurate and robust for species delineationwithin bacterial genera such as Streptomyces [5]. Here, weobtained the alkB gene sequences from Gordonia type strainsand environmental isolates, which, in combination with othermolecular markers, such as the 16S rRNA, gyrB and catA genes,might provide a valuable tool for differentiation of Gordonia

members at the inter-species or intra-species level.

Acknowledgements

This research work was kindly supported by grants from theMinistry of Economic Affairs and in part by the Ministry ofEducation, Taiwan, R.O.C. under the ATU plan.

ARTICLE IN PRESS

F.-T. Shen et al. / Systematic and Applied Microbiology 33 (2010) 53–59 59

Appendix A. Supporting information

Supplementary data associated with this article can be foundin the on-line version at doi:10.1016/j.syapm.2009.11.003.

References

[1] Arenskotter, M., Broker, D., Steinbuchel, A., 2004. Biology of the metabolicallydiverse genus Gordonia. Appl. Environ. Microbiol. 70, 3195–3204.

[2] Chandler, D.P., Brockman, F.J., 1996. Estimating biodegradative gene numbersat a JP-5 contaminated site using PCR. Appl. Biochem. Biotechnol. 57/58,971–982.

[3] Fleming, J.T., Sanseverino, J., Sayler, G.S., 1993. Quantitative relationshipbetween naphthalene catabolic gene frequency and expression in predictingPAH degradation in soils at town gas manufacturing sites. Environ. Sci.Technol. 27, 1068–1074.

[4] Fujii, T., Narikawa, T., Takeda, K., Kato, J., 2004. Biotransformation of variousalkanes using the Escherichia coli expressing an alkane hydroxylase systemfrom Gordonia sp. TF6. Biosci. Biotechnol. Biochem. 68, 2171–2177.

[5] Guo, Y., Zheng, W., Rong, X., Huang, Y., 2008. A multilocus phylogeny of theStreptomyces griseus 16S rRNA gene clade: use of multilocus sequence analysisfor streptomycete systematics. Int. J. Syst. Evol. Microbiol. 58, 149–159.

[6] Hamamura, N., Fukui, M., Ward, D.M., Inskeep, W.P., 2008. Assessing soilmicrobial populations responding to crude-oil amendment at differenttemperatures using phylogenetic, functional gene (alkB) and physiologicalanalyses. Environ. Sci. Technol. 42, 7580–7586.

[7] Hamann, C., Hegemann, J., Hildebrandt, A., 1999. Detection of polycyclicaromatic hydrocarbon degradation genes in different soil bacteria bypolymerase chain reaction and DNA hybridization. FEMS Microbiol. Lett.173, 255–263.

[8] He, Z., Gentry, T.J., Schadt, C.W., Wu, L., Liebich, J., Chong, S.C., Huang, Z., Wu,W., Jardine, P., Criddle, C., Zhou, J., 2007. GeoChip: a comprehensivemicroarray for investigating biogeochemical, ecological and environmentalprocesses. ISME J. 1, 67–77.

[9] Iwai, S., Kurisu, F., Urakawa, H., Yagi, O., Kasuga, I., Furumai, H., 2008.Development of an oligonucleotide microarray to detect di- and monoox-ygenase genes for benzene degradation in soil. FEMS Microbiol. Lett. 285,111–121.

[10] Kubota, K., Koma, D., Matsumiya, Y., Chung, S.Y., Kubo, M., 2008. Phylogeneticanalysis of long-chain hydrocarbon-degrading bacteria and evaluation oftheir hydrocarbon-degradation by 2,6-DCPIP assay. Biodegradation 19,749–757.

[11] Librado, P., Rozas, J., 2009. DnaSP v5: a software for comprehensive analysisof DNA polymorphism data. Bioinformatics 25, 1451–1452.

[12] Milcic-Terzic, J., Lopez-Vidal, Y., Vrvic, M.M., Saval, S., 2001. Detection ofcatabolic genes in indigenous microbial consortia isolated from a diesel-contaminated soil. Bioresour. Technol. 78, 47–54.

[13] Morton, B.R., 1993. Chloroplast DNA codon usage: evidence for selection atthe psbA locus based on tRNA availability. J. Mol. Evol. 27, 273–280.

[14] Quatrini, P., Scaglione, G., de Pasquale, C., Riela, S., Puglia, A.M., 2008.Isolation of Gram-positive n-alkane degraders from a hydrocarbon-contaminated Mediterranean shoreline. J. Appl. Microbiol. 104, 251–259.

[15] Sharp, P.M., Tuohy, T.M.F., Mosurski, K.R., 1986. Codon usage in yeast: clusteranalysis clearly differentiates highly and lowly expressed genes. NucleicAcids Res. 14, 5125–5143.

[16] Shen, F.T., Ho, M.J., Huang, H.R., Arun, A.B., Rekha, P.D., Young, C.C., 2008.Molecular detection and phylogenetic characterization of Gordonia species inheavily oil-contaminated soils. Res. Microbiol. 159, 522–529.

[17] Shen, F.T., Huang, H.R., Arun, A.B., Lu, H.L., Lin, T.C., Rekha, P.D., Young, C.C.,2007. Detection of filamentous genus Gordonia in foam samples using genus-specific primers combined with PCR-denaturing gradient gel electrophoresisanalysis. Can. J. Microbiol. 53, 1–7.

[18] Shen, F.T., Lin, J.L., Huang, C.C., Ho, Y.N., Arun, A.B., Young, L.S., Young, C.C.,2009. Molecular detection and phylogenetic analysis of the catechol1,2-dioxygenase gene from Gordonia spp. Syst. Appl. Microbiol. 32,291–300.

[19] Shen, F.T., Lu, H.L., Lin, J.L., Huang, W.S., Arun, A.B., Young, C.C., 2006.Phylogenetic analysis of members of the metabolically diverse genusGordonia based on proteins encoding the gyrB gene. Res. Microbiol. 157,367–375.

[20] Shen, F.T., Young, C.C., 2005. Rapid detection and identification of themetabolically diverse genus Gordonia by 16S rRNA-gene-targeted genus-specific primers. FEMS Microbiol. Lett. 250, 221–227.

[21] Smits, T.H.M., Rothlisberger, M., Witholt, B., van Beilen, J.B., 1999. Molecularscreening for alkane hydroxylase genes in Gram-negative and Gram-positivestrains. Environ. Microbiol. 1, 307–317.

[22] Stamatakis, A., 2006. RAxML-VI-HPC: maximum likelihood-based phyloge-netic analyses with thousands of taxa and mixed models. Bioinformatics 22,2688–2690.

[23] Tajima, F., 1989. Statistical method for testing the neutral mutationhypothesis by DNA polymorphism. Genetics 123, 585–595.

[24] Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: molecularevolutionary genetics analysis (MEGA) software version 4.0. Mol. Biol. Evol.24, 1596–1599.

[25] Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997.The CLUSTAL_X windows interface: flexible strategies for multiplesequence alignment aided by quality analysis tools. Nucleic Acids Res. 25,4876–4882.

[26] van Beilen, J.B., Wubbolts, M.G., Witholt, B., 1994. Genetics of alkaneoxidation by Pseudomonas oleovorans. Biodegradation 5, 161–174.

[27] Watts, D., MacBeath, J.R., 2001. Automated fluorescent DNA sequencing onthe ABI PRISM 310 Genetic Analyzer. Methods Mol. Biol. 167, 153–170.

[28] Whyte, L.G., Smits, T.H.M., Labbe, D., Witholt, B., Greer, C.W., van Beilen, J.B.,2002. Gene cloning and characterization of multiple alkane hydroxylasesystems in Rhodococcus strains Q15 and NRRL B-16531. Appl. Environ.Microbiol. 68, 5933–5942.

[29] Yassin, A.F., Shen, F.T., Hupfer, H., Arun, A.B., Lai, W.A., Rekha, P.D., Young,C.C., 2007. Gordonia malaquae sp. nov., isolated from sludge of a wastewatertreatment plant. Int. J. Syst. Evol. Microbiol. 57, 1065–1068.

[30] Young, C.C., Lin, T.C., Yeh, M.S., Shen, F.T., Chang, J.S., 2005. Identification andkinetic characteristics of an indigenous diesel-degrading Gordonia alkanivor-ans strain. World J. Microbiol. Biotechnol. 21, 1409–1414.