Cc

LD

a

ARR3A

KCCRS

1

sRfisiatt(tto

otttll

j

0h

Journal of Steroid Biochemistry & Molecular Biology 139 (2014) 33– 44

Contents lists available at ScienceDirect

Journal of Steroid Biochemistry and Molecular Biology

jo ur nal home page: www.elsev ier .com/ locate / j sbmb

holesterol to cholestenone oxidation by ChoG, the main extracellularholesterol oxidase of Rhodococcus ruber strain Chol-4

aura Fernández de las Heras, Julián Perera, Juana María Navarro Llorens ∗

epartment of Biochemistry and Molecular Biology I, Universidad Complutense de Madrid, 28040 Madrid, Spain

r t i c l e i n f o

rticle history:eceived 14 June 2013eceived in revised form0 September 2013ccepted 1 October 2013

eywords:holesterol oxidase

a b s t r a c t

The choG ORF of Rhodococcus ruber strain Chol-4 (referred from now as Chol-4) encodes a putative extra-cellular cholesterol oxidase. In the Chol-4 genome this ORF is located in a gene cluster that includes kstD3and hsd4B, showing the same genomic context as that found in other Rhodococcus species. The putativeChoG protein is grouped into the class II of cholesterol oxidases, close to the Rhodococcus sp. CECT3014ChoG homolog. The Chol-4 choG was cloned and expressed in a CECT3014 �choG host strain in order toassess its ability to convert cholesterol into cholestenone. The RT-PCR analysis showed that choG genewas constitutively expressed in all the conditions assayed, but a higher induction could be inferred when

holestenonehodococcusteroids

cells were growing in the presence of cholesterol. A Chol-4 �choG mutant strain was still able to grow inminimal medium supplemented with cholesterol, although at a slower rate. A comparative study of theremoval of both cholesterol and cholestenone from the culture medium of either the wild type Chol-4 orits choG deletion mutant revealed a major role of ChoG in the extracellular production of cholestenonefrom cholesterol and, therefore, this enzyme may be related with the maintenance of a convenient supply

ccee

of cholestenone for the su

. Introduction

Many bacteria able to degrade different steroid moleculesuch as cholesterol and its derivatives have been described [1–6].ecently the cholesterol catabolic pathway has been partially clari-ed in some actinobacteria such as Mycobacterium and Rhodococcuspecies [7–12], although the physiological roles of many of thenvolved enzymes are not yet fully defined. Cholesterol undergoes

side chain cleavage, either concurrently with the ring degrada-ion (Mycobacterium [13]) or not, (Rhodococcus jostii RHA1 [14,15])o yield 17-keto compounds such as 1,4-androstadiene-3,17-dioneADD) and 4-androstene-3,17-dione (AD) [11]. This first section ofhe catabolic pathway probably begins with the oxidation of choles-erol to yield cholestenone, in a reaction catalyzed by a cholesterolxidase.

The 3�-hydroxysterol oxidase, commonly known as cholesterolxidase (EC 1.1.3.6), is a flavin-dependent bifunctional oxidoreduc-ase [16,17]. This enzyme catalyses two coordinated reactions inhe same active site: the oxidation of a �5-ene-3�-hydroxysterol

o �5-ene-3�-ketosteroid, coupled to the reduction of molecu-ar oxygen to hydrogen peroxide, and the isomerization of theater to yield �4-ene-3�-ketosteroid as the final product [17–19].

∗ Corresponding author. Tel.: +34 913944145; fax: +34 913944672.E-mail addresses: laurfern@bio.ucm.es (L. Fernández de las Heras),

pererag@bio.ucm.es (J. Perera), joana@bio.ucm.es (J.M. Navarro Llorens).

960-0760/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.jsbmb.2013.10.001

ding steps of the catabolic pathway.© 2013 Elsevier Ltd. All rights reserved.

Cholesterol is the best known substrate for cholesterol oxidasesalthough other substrates include �-cholestanol, sitosterol, pre-gnenolone, ergosterol and stigmasterol [20]. This enzyme is uniquein bacteria and is widespread among them, from harmless soilbacteria that use cholesterol as a nutritional source to pathogenicbacteria such as Rhodococcus equi in which this enzyme seemsto play a role as a virulence factor [20,21]. Cholesterol oxidasesare thought to be alcohol oxidases adapted to accommodatethe bulky cholestane frame [17]. Two types of cholesterol oxi-dases have been described so far, one that contains the FADcofactor tightly but non-covalently bound to the enzyme (classI) and another one containing the cofactor covalently linked tothe enzyme (class II) [20,22]. Members of these two groups haveno significant sequence identity and belong to different proteinsuperfamilies: the GMC (glucose/methanol/choline) oxidoreduc-tase family (class I) and the VAO (vanillyl-alcohol oxidase) family(class II). The cholesterol oxidase is a biotechnologically impor-tant enzyme and it has been the subject of recent reviews on itsbiochemical features [17], its physiological functions [23], or itsbiotechnological applications [16,20]. Its three major physiologicalfunctions are: (i) the participation in steroid metabolism; (ii) thepossible involvement in pathogenesis and virulence, a point thatis still controversial [16,24,25]; and (iii) the peculiar intervention

as a biosensor, as described in the polyene macrolide biosynthe-sis [26,27]. Among the biotechnological applications, cholesteroloxidase is being used for blood serum and food cholesterol deter-mination [16,28,29] or for providing valuable intermediates used

3 id Bioc

fsc

iDAtafsitdiittc

2

2

c(iwmcNZ11C04awuei0tcop

fi

TB

4 L. Fernández de las Heras et al. / Journal of Stero

or steroid drug production, such as ADD and AD, two majortarting substrates for the synthesis of anabolic drugs and contra-eptive hormones [30].

In this work, we have searched for cholesterol oxidases involvedn the cholesterol degradation pathway of Chol-4 (CECT 7469;SMZ 45280), a strain isolated from a sewage sludge sample [3].

choG-like ORF was found in a cluster of ORFs possibly related tohe steroid metabolism [31] and has been isolated, cloned and thectivity of the encoded protein evaluated. Results showed that choGrom Chol-4 is induced when growing in cholesterol, that ChoG isecreted and acts outside the cell and that it is the main enzymenvolved in the transformation of cholesterol into cholestenone inhis strain. However, ChoG is not essential for the cholesterol degra-ation pathway in Chol-4, as its �choG mutant was still able to grow

n cholesterol as the sole source of carbon and energy, althoughts absence slowed down the growth of the culture and reducedhe consumption of cholesterol. The major contribution of ChoG tohe catabolic pathway may lie in facilitating the cellular uptake ofholesterol by converting it into cholestenone.

. Materials and methods

.1. Strains, growth conditions and plasmids

Rhodococcus sp. strain CECT3014 was obtained from “Colec-ión Espanola de Cultivos Tipo” (CECT), Valencia, Spain. Chol-4CECT7469, DSMZ45280) was previously obtained and character-zed in our laboratory [3]. Gram-positive bacteria and E. coli strains

ere grown at 30 ◦C or 37 ◦C, respectively, in Luria–Bertani (LB)edium [32] with shaking at 200 rpm, unless otherwise indi-

ated. W minimal salt medium [33] (12.5 mM KH2PO4, 69 mMa2HPO4, 7.5 mM (NH4)2SO4, 0.26 mM MgO, 20 �M CaCO3, 5 �MnSO4, 0.4 �M MgSO4, 3.4 �M FeSO4, 1 �M CuSO4, 1 �M CoSO4,

�M H3BO3) or 457 DSMZ minimal medium (11.1 mM KH2PO4,7.1 mM Na2HPO4, 3.78 mM (NH4)2SO4, 0.8 mM MgSO4, 0.45 mMaCl2, 13 �M EDTA, 0.34 �M ZnSO4, 7.2 �M FeSO4, 0.15 �M MnCl2,.06 �M CuCl2, 0.84 �M CoCl2, 0.08 �M NiCl2, 0.12 �M Na2MoO4,.8 �M H3BO3) were used to grow the bacteria in organic substratess the sole carbon and energy source. W minimal medium has beenidely used for Rhodococcus cultures and therefore it was alsosed for CECT3014 studies. However, Chol-4 did not grow prop-rly in this medium and 457 DSMZ minimal medium was usednstead. Cholesterol was added directly to the minimal medium at.6 mg/mL. Alternatively or when growing on cholesterol plates,he steroid was dissolved at 0.6 mg/mL in 16.5 mM methyl-�-yclodextrin to form inclusion complexes following a modification

f a previously reported method [34] as described [35]. Plates wererepared by adding 1.5% (w/v) agar.

When appropriate, antibiotics were added at the followingnal concentration: ampicillin at 100 �g/mL; nalidixic acid at

able 1acterial strains (A) and plasmids (B) used in this work.

Characteristics

(A) StrainEscherichia coli DH5�F’ F’/endA1 hsdR17 (rk

− mk+) glnV44 thi-1 recA

deoR (�80dlac�(lacZ)M15)Escherichia coli BL21 (DE3) �DE3 (lacI lacUV5-T7 gene 1 ind1 sam7 ninRhodococcus sp. CECT3014 �choG mutant

Rhodococcus ruber strain Chol-4 Wild type

(B) PlasmidpBluescrip KS II (+) pUC19 derivated, orif1 and oriColE1, Plac , lapUC18 Cloning vector, oriMB1,Plac , lacZ�

pMal C2X oriColE1, rop, malE, lacZ˛, Ptac , lacIq, ApR

pGEM-T Easy Vector Cloning KIT from PCR vector, ApR

pET29a E. coli expression vector, KmR

pTip-QC1 Expression E. coli-Rhodococcus shuttle vect

hemistry & Molecular Biology 139 (2014) 33– 44

20 �g/mL; apramycin at 50 �g/mL; chloramphenicol at 34 �g/mLand kanamycin at 50 �g/mL (in E. coli) or at 200 �g/mL (inRhodococcus cultures). Strains and plasmids used in this study arelisted in Table 1.

For RT-PCR analyses, Chol-4 cultures were grown on 50 mL LBor 457 DSMZ medium with either 0.6 mg/mL cholesterol (addedas a powder, instead of dissolved in cyclodextrin) or 2 mg/mLsodium acetate. Cells were harvested in mid-exponential growthphase (OD600 nm = 0.6) after addition of 5 mL of a 9:1 (v/v) mix ofethanol/phenol and further centrifugation at 5000 rpm and 4 ◦C for15 min. Afterwards, cells were stabilized by addition of RNA protectBacteria Reagent (Qiagen) following the manufacturer’s instruc-tions. Pellets were kept at −80 ◦C until use.

2.2. Isolation, DNA sequencing and in silico analysis

DNA manipulations and other molecular biology techniqueswere essentially as described elsewhere [32]. Genomic DNAextraction was based on the procedure developed for Listeria mono-cytogenes [36].

The pyrosequencing of the genomic DNA from Chol-4 waspreviously performed by LifeSequencing (Valencia, Spain) usingthe Roche 454GS-FLX system ([37], WGS ANGC01000000). TheBioEdit programme [38] was used to BLAST a known ORF againstthe obtained contigs. In silico analyses of sequences were madeusing standard programmes. The sequence of the genomic frag-ment of Chol-4 carrying the choG ORF has been submitted toEMBL Nucleotide Sequence Database (GenBank accession numberFJ842098).

Nucleotide sequences were determined by using a modelAB3730 automated DNA sequencer (Applied Biosystem Inc., Unidadde Genómica, Parque Científico de Madrid, Spain).

Putative signal peptides were predicted by the SignalP 4.0server programme (http://www.cbs.dtu.dk/services/SignalP/)using a model trained on Gram-positive bacteria [39]. Putativedomains in the ORFs were searched with the BLASTs server(http://expasy.org/prosite/). Putative promoters were analyzedusing the Neural Network Promoter Prediction (NNPP), and BPROMprogrammes (http://www.fruitfly.org/seq tools/promoter.html,and http://linux1.softberry.com/berry.phtml, respectively), in allcases with a score value of ≥80%. DNASTAR (Lasergene) pro-grammes were used to analyze sequences, to elaborate cladogramsand to design primers.

2.3. Bacterial transformation, cloning and expression of ChoG

Competent and electrocompetent cells of E. coli DH5�F’ wereprepared and transformed as previously described [32]. Selection oftransformed cells was carried out in LB plates supplemented withappropriate antibiotics. Rhodococcus electrocompetent cells were

Reference

1 gyrA (NalR) relA1 �(lacIZYA- argF) U169 Laboratory collection

5) F− dcm ompT hsdS (rB−mB

+) gal Invitrogen[35]CECT7469-DSMZ45280

cZ�, ApR FermentasThermo ScientificNew England BiolabsPromegaNovagen

or, ApR, PtipA ChlR, repAB (pRE2895) [60]

L. Fernández de las Heras et al. / Journal of Steroid Biochemistry & Molecular Biology 139 (2014) 33– 44 35

Table 2List of primers used in this work.

Primer sequence 5′-3′ Use

CH129 GCTCTAGAGCGGTCCGGTGCTAATCG Amplification of 1.8 kb EcoRI-XbaI R. ruber strain Chol-4 choG region for cloning and expressionCH130 GCGAATTCATGACGATTCGCACCGAACCCCTCH88 CTCCGTCCGCCGAACAGAAAG Amplification of 1.2 kb EcoRI-XbaI (choG up region)CH199 GGTCTAGAGCGGGGCGCAGGTCCAGATCH200 GGTCTAGAGGATGCGCGGACACTACGACAAC Amplification of 1.2 kb XbaI-HindIII (choG down region)CH201 CGTAAGCTTGCGTGCCGTGGCGATCGTGTTCATCH67 CAAGGTCTTCTGCCCGATGCTCT R. ruber choG amplification in RT-PCR, 0.7 kbCH68 GGTCCGGTAGATCGTGCCTTCCTTCH89 CCCAAGCTTGGGCCGTGGCGATCGTGTTCAT With CH88, to check mutant coloniesCH104 GGCGATCTGGTACCCCTTCAC To check mutant coloniesCH105 CGATCCGCTGGTCCGTGTACH190 CGACGGCAGCTGTACGAGACCT R. ruber kstD4 amplification in RT-PCR, 1 kb

kstD5

ptwt(ugww(csdA0t

p3wf

2

ophdcttp

leo1iwtbT(0Dnt

CH191 AACTGCCCGGGACGACCTTGCH192 GCTCGACACCACACCGCTGAA R. ruber

CH193 GCTCTCGCCGCGGTGGTATT

repared and transformed in the same way, but cells were allowedo recover for 4 h at 30 ◦C without shaking before plating. choG ORFas amplified by PCR (Table 2) and cloned into pGEM-T Easy Vec-

or (Promega) and then moved to a pMalC2X or a pET29a vectorNovagen). Expression plasmids with and without a choG ORF weresed to transform E. coli BL21 (DE3) cells (Invitrogen) for heterolo-ous expression. Cultures of E. coli harbouring expression plasmidsere grown (16 ◦C, 200 rpm) in 50 mL LB broth supplementedith ampicillin (100 �g/mL). Isopropyl-�-d-thiogalactopyranoside

IPTG; 0.1 mM) was added for a 4 h induction. choG ORF was alsoloned into a pTip-QC1 E. coli-Rhodococcus shuttle vector. This con-truction was introduced into CECT3014 electrocompetent cellsefective in choG ORF [35] and selection was done in LB/Cm plates.

25 mL culture of a positive clone was grown until an OD600 nm of.6 was reached and then induced for 18 h at 30 ◦C with 1 �g/mLhiostrepton (final concentration).

In all cases, cells were washed twice in 50 mM phosphate bufferH 7.0, frozen at −20 ◦C, sonicated and centrifuged (20,000 × g,0 min, 4 ◦C) in order to collect the supernatant. Protein contentas measured by Bradford assay [40] and protein expression was

ollowed by SDS-PAGE (10%) and Coomassie staining (not shown).

.4. PCR and RT-PCR

PCR was performed under standard conditions using a 1:1 mixf Expand High Fidelity PCR System Enzyme Mix (Roche) and Pfuolymerase (Biotools). For amplifications of Chol-4 DNA, a specificigh GC buffer (Roche) was also added to the reaction mix. Primers,imethyl sulfoxide, dNTPs and template DNA were added at a finaloncentration of 0.4 �M, 4% (v/v), 200 �M and 1 ng/�L, respec-ively. PCR conditions were: 30 cycles of 1 min at 95 ◦C, 1 min athe desired Tm and 0.5–3 min at 72 ◦C unless stated otherwise. PCRrimers used in this work are listed in Table 2.

RNA samples for RT-PCR experiments were obtained from mid-og exponential phase cultures (OD600 nm 0.7–0.8). Frozen pelletsquivalent to 50 mL of culture were treated with lysozyme (20 �Lf 100 mg/mL stock) in 1 ml Tris–EDTA buffer (10 mM Tris/HCl, pH8,

mM EDTA) at 37 ◦C for 1–12 h. Afterwards, 200 �L of SDS contain-ng proteinase K (160 �L SDS 10% + 40 �L proteinase K 10 mg/mL)

as added and incubated for 10 min at 65 ◦C. Proteins were precipi-ated with 100 �L 5 M NaOH and 100 �L cetyl trimethyl ammoniumromide (0.1 g/mL suspended in 0.7 M NaOH) for 10 min at 65 ◦C.hereafter, total RNA was prepared with the RNeasy Mini KitQiagen) following the manufacturer’s indications. After this, each

.5–1 �g of RNA was treated twice or three times with 5 U of TurboNase RNase-Free (Ambion) in a 100 �L volume for 2 h at 37 ◦C untilo sign of DNA was detected by PCR. RNA samples were precipi-ated with 0.12 volumes of 5 M ammonium acetate, 0.02 volumes

amplification in RT-PCR, 1 kb

of glycogen (5 mg/mL) and 1 volume of isopropanol, washed twicein 70% ethanol and dissolved in water. cDNA was synthesized using6 �g of random primers (Roche) per 10 �g of RNA. Reaction mixturewas incubated with 300 U of SuperScript III Reverse Transcriptase(Invitrogen) for 2 h at 55 ◦C. The resulting cDNA was treated with2 �L of RNase A (10 mg/mL) for 1 h at 37 ◦C, purified using Ultra-Clean PCR Clean-up (MoBio) and recovered in a volume of 50 �L ofwater. cDNA was used as template (25 ng) for PCR reactions (25 �Lfinal volume). Controls without reverse transcriptase (RT−) wereused to detect any contamination of undigested DNA in the RNApreparations.

2.5. Cholesterol oxidase enzymatic assay

Chox indicator plates were prepared with agar, W medium withcholesterol and cyclodextrin (see Section 2.1) containing 0.1 mg/mLo-dianisidine and 1 U/mL peroxidase [41].

A colorimetric assay for measuring cell-free and extracted Choxactivities was performed as described (U.S. Patent No. 4072568)[42].

Cells from 25 mL of thiostrepton induced cultures (Section 2.3)were harvested by centrifugation at 15 min, at 4 ◦C and 5000 rpm.The broth filtrate was kept at 4 ◦C and the pellet was suspendedin 0.05 M potassium phosphate buffer, pH 7.5, containing 5% (v/v)isopropanol and 1% (v/v) Triton X-100. Extracted Chox fractionwas obtained by magnetic stirring of the cellular suspension for30–40 min at room temperature. The enzymatic solution was col-lected by centrifugation (15,000 rpm, 30 min, 4 ◦C) and kept at 4 ◦Cuntil use. The enzymatic activity in both fractions (cell-free andextracted) was measured at 25 ◦C in a 3 mL reaction mix con-taining 0.038 M potassium phosphate buffer, pH 7.5, 0.009% (w/v)o-dianisidine, 0.17 mg/mL cholesterol, 0.33% (v/v) Triton X-100 and10 U of peroxidase. The increase of OD (OD500 nm) for 5–10 min wasrecorded and units of Chox activity were calculated by:

U/mL = (�OD500 nm/ min in sample

− �OD500 nm/ min in blank)V

E· v

where V is the assay total volume (3 mL), v is the sample volume(0.1 mL) and E is the millimolar extinction coefficient of oxidizedo-dianisidine at 500 nm (7.5).

One activity unit will convert 1 �mol of cholesterol to 4-cholesten-3-one per minute at pH 7.5 and 25 ◦C. Data displayedin Section 3.3 show the Chox activity units measured in both the

supernatant (containing the enzyme excreted to the medium) andthe Triton X-100 extracted fraction (that account for the enzymebound to the membrane and intracellular activities) from a 25 mLvolume of cell culture.

3 id Bioc

2

Ufd

aaagicpp

whbAschTdpmkca

2c

Lmmaffa

Fcp

6 L. Fernández de las Heras et al. / Journal of Stero

.6. Mutagenesis of the choG gene

PCR primers used in this experiment are summarized in Table 2.nmarked gene deletion was carried out by conjugative trans-

er of a mutagenic plasmid carrying the sacB selection system asescribed previously for Rhodococcus erythropolis SQ1 [43].

Specific pairs of primers were designed from up (CH88 + CH199)nd downstream (CH200 + CH201) regions of the choG ORF to PCRmplify both choG flanking regions containing EcoRI-XbaI (up)nd XbaI-HindIII (down) restriction sites (see Fig. 4), using Chol-4enomic DNA as template. These amplicons were cloned togethernto a pGEM-T Easy vector to get an EcoRI-HindIII DNA fragmentontaining a truncated choG ORF that was then cloned into theK18mobsacB plasmid [44] to give rise to the mutagenic plasmidK18(U + D).

E. coli S17.1 harbouring pK18(U + D) was used for conjugationith Chol-4 as described before [31]. Chol-4 transconjugants thatave integrated the plasmid pK18(U + D) by homologous recom-ination appeared after 3–4 days at 30 ◦C in LB/Nal/Km plates.ll kanamycin resistant Rhodococcus transconjugants were sucroseensitive because sacB was contained in the integrated plasmid. ThehoG deletion was then achieved as a result of a second spontaneousomologous recombination process within the genome of Chol-4.hus, 2 or 3 colonies of Chol-4 were grown under non-selective con-itions (20 mL LB at 30 ◦C o/n) and subsequently plated on LB agarlus 10% sucrose (w/v). Sucrose resistant colonies grown on thisedium were replica plated on LB agar with or without 200 �g/mL

anamycin and Km sensitive colonies were selected as unmarkedhoG deleted cells. Afterwards, DNA of each colony was obtainednd the deletion of choG verified by PCR.

.7. Monitoring of cholesterol and cholestenone by thin layerhromatography (TLC)

Chol-4 wild-type and its �choG mutant were grown in 3 mLB for two days. Then cells were washed 2–3 times with minimaledium, 50 �L of the washed cells suspended in 50 mL minimaledium supplemented with cholesterol/cyclodextrin at 0.6 mg/mL

nd kept growing at 30 ◦C, 250 rpm. Aliquots of 2 mL were takenrom the culture at fixed times (OD600 nm measured) and their lipidraction was obtained by extraction twice with 2 mL chloroformnd let dried.

ig. 1. Cladogram of some cholesterol oxidases. The cholesterol oxidases studied in this

ompared. Uniprot accession numbers (or NCBI accession numbers in the case of Chol-4 coerformed with MegAlign (Lasergene) using the Clustal V method. Bar: 0.1 substitutions

hemistry & Molecular Biology 139 (2014) 33– 44

When required the 2 mL aliquots were treated as follows: 1 mLwas kept as total culture (T) and the other mL was centrifugedto separate cells (P, pellet) from medium (S, supernatant). Pelletswere suspended in 1 mL of water and T, P and S samples were thenchloroform-extracted and dried.

200 �L chloroform were added to all previously dried samplesand 10 �L of the solution were applied on aluminium TLC Silicagel 60 F254 sheets (Merck). Chromatography was performed usinghexan:ethylacetate (10:4, v/v) as solvent and spots were revealedby UV exposure (�254 nm). Afterwards the TLC plate was sprayedwith a 20% sulphuric acid solution (v/v) followed by heating 5 minat 100 ◦C. Standard control samples of cholestenone, cholesterol,AD and ADD at 1 �g were also included.

3. Results and discussion

3.1. Identification of the choG gene of Chol-4

In a search for cholesterol oxidase genes in the genome of Chol-4, we found 4 chox ORFs. Two of them may encode putative proteins(P571 and P577, with 571 and 577 amino acids, respectively) thatbelong to the class I of cholesterol oxidases (Fig. 1). P577 keeps 82%amino acid sequence identity with the cholesterol oxidase ChoD ofR. equi 103S (Uniprot: E4WKM2) and P571 keeps 79% amino acidsequence identity with ChoD of R. jostii RHA1 (Uniprot: Q0S8P0).As displayed in Appendix A, P571 ORF from Chol-4 is flanked bygenes encoding a flavin monooxygenase and a hypothetical pro-tein, a genomic context that is conserved in R. jostii RHA1 (Uniprot:Q0S8P1 and Q0S8N9, respectively), R. opacus (Uniprot: C1B9V7and C1B9V9) and R. equi 103S (Uniprot: E4WET8 and E4WEU0),among other bacteria of this genre. Chol-4 P577 ORF is flanked byan inosine 5-monophosphate dehydrogenase gene and a GMP syn-thase (GuaA) gene (Appendix A), a gene cluster related to the IMPmetabolism that is highly conserved in Rhodococcus strains.

On the other hand, P591 (a 591 amino acids-protein that keeps69% amino acid sequence identity with a FAD-dependent oxi-doreductase of R. equi 103S (Uniprot: E4WCF5)) and ChoG (anextracellular protein of 581 amino acids; NCBI: FJ842098) belong

to the class II of cholesterol oxidases (Fig. 1). The P591 ORF fromChol-4 is flanked by a hypothetical protein and an arabinan endo-1,5-alpha-l-arabinosidase (Appendix A), a genomic context thatdoes not seem to be conserved in other strains.

work and other proteins representative of the two classes of this enzyme [20] arentigs in which the choxs are located) appear between brackets. The cladogram was

per amino acid.

id Bioc

Roitsg

apol(sttsCowcpiiCcedccd

dgkgtooho

3

eo[cyastgiw

Rgtaplgig

L. Fernández de las Heras et al. / Journal of Stero

ChoG shows 78% amino acid sequence identity to ChoG fromhodococcus sp. CECT3014, reported to be the main cholesterolxidase of this strain [35]. It displays a 60% amino acid sequencedentity with P591 and only 12% amino acid sequence identity withhe other two Chox found in Chol-4. This work is focused on thetudy of this ORF and has led to its identification as the Chol-4 choGene.

The choG ORF of Chol-4 has a high GC content (70%)nd would encode a 62.8 kDa protein with an isoelectricoint of 5.9. SignalP 4.0 algorithm revealed the presencef a cleavage site for signal peptidases in which the mostikely cleavage site was located between residues 36 and 37VHA-VP), suggesting that ChoG is a secreted protein. Thisignal [MTIRTEPLRLSRRGFLAAGAGALAATALGGWTPVHAVPA] con-ains most of the (S/T)RRxFLK consensus of the twin-arginineranslocation pathway signal peptide [45] similarly to the majorecreted protein of R. equi, cholesterol oxidase ChoE [46]. Chol-4hoG contains the consensus FAD binding site of class II cholesterolxidases: R101PRGAMHGW109, S178VAGSLA184 and L244GRTFV249

here the underlined amino acids support the hydrogen bondingontact to the cofactor [17]. The residues Glu311, Glu475 and Arg477,roposed to be involved in the catalysis of the steroid substrate

n class II cholesterol oxidases (for a review [17]), are also presentn Chol-4 ChoG and in the same position (Appendix B). Moreover,hol-4 ChoG contains a FAD binding domain (pfam01565) and aholesterol oxidase substrate-binding domain (pfam09129) with-values of 3.40e−07 and 4.15e−147, respectively. This pfam09129omain is involved in the formation of the roof of the active siteavity and allows the catalysis of oxidation and isomerization ofholesterol to 4-cholesten-3-one [conserved domains at the NCBIatabase; http://www.ncbi.nlm.nih.gov/Structure/index.shtml].

When present in Rhodococcus species genomes (R. equi, R. pyri-inivorans, R. erythropolis), choG always occurs close to two otherenes also involved in the steroid metabolism, namely the genestD3 (encoding a 3-keto-5�-steroid �1-dehydrogenase) and theene hsd4B (encoding a 2 enoyl acyl CoA hydratase), giving riseo the kstD3-hsd4B-choG cluster (Appendix A). Such a cluster alsoccurs in Chol-4 genome. It should be noted that choG is absent inther Rhodococcus strains that, however, do have the cluster kstD3-sd4B and are able to grow in cholesterol (e.g., R. jostii RHA1 or R.pacus B4, Appendix A).

.2. Transcriptional analysis of Chol-4 choG

Rhodococcal cholesterol oxidase is reported to be an induciblenzyme, being up-regulated by the presence of the sterol side chainr repressed by androstenedione or testosterone intermediates16,23]. However, these published induction data depend on theholesterol oxidase activity estimation rather than on mRNA anal-sis. In fact, there are so far only a few published transcriptional databout cholesterol oxidase expression: choG gene from Rhodococcusp. CECT3014 and cho2 gene from G. cholesterolivorans are reportedo be transcribed at a moderate basal level although an additionalene induction in cells growing on cholesterol was also observedn the former [35,47]. In R. jostii RHA1 no induction by cholesterol

as found in any of its putative cholesterol oxidases [4,10].The expression of choG of Chol-4 was analyzed by means of

T-PCR (Fig. 2). RNA samples were prepared from cultures of cellsrowing in 457 DSMZ medium supplemented with either choles-erol or an alternative carbon source. Specific primer pairs (Table 1nd Fig. 4A) were designed to search for choG mRNA. Resultsointed out that choG transcripts could be detected in both cultures,

eading to the conclusion that there is an expression of the choGene under both conditions (Fig. 2). These results also showed anmportant qualitative increase in choG expression when cells wererowing on cholesterol. This is consistent with the expression of

hemistry & Molecular Biology 139 (2014) 33– 44 37

choG from Rhodococcus sp. CECT3014, which has been proposed tobe controlled by two different promoters: a choG promoter locatedjust upstream the gene and driving a constitutive gene expression,and a second one, a steroid-regulated promoter located upstreamthe putative kstD3-hsd4B-choG operon [35].

In silico studies have been performed in the upstream genomicregion of Chol-4 choG and no putative promoter was found withinthe 500 base pairs upstream choG, although it has to be noted thatthere is no specific programme for Gram-positive bacteria pro-moter prediction and data must be taken carefully.

On the other side, the co-transcription of the whole gene clusterkstD3-hsd4B-choG from a promoter upstream kstD3 gene in Chol-4 cultures grown in the presence of cholesterol is being studied(article in preparation). Preliminary results show a similar situationto that observed in Rhodococcus sp. CECT3014 [35].

3.3. Cloning and expression of Chol-4 ChoG

Cholesterol oxidases from different sources have beenexpressed in E. coli using pUC19 [48], pET28a [49] or pET24b[50], being the optimization of the codon usage for E. coli requiredin most cases, and in Streptomyces lividans using a pIJ702 vector[51].

In our case, the cloning and expression of Chol-4 ChoG in E. coliusing pMal or pET29 systems only resulted in the accumulation ofthe protein in the non-soluble cellular fraction. However, we man-aged to express Chol-4 choG in a closer strain, namely Rhodococcussp. CECT3014, by using the E. coli-Rhodococcus shuttle vector pTip-QC1, which is non-functional in Chol-4 but it is able to replicate inRhodococcus sp. CECT3014.

We successfully cloned the PCR-amplified Chol-4 choG gene inE. coli and then transferred either the pTip-choG recombinant plas-mid or the empty vector (control sample) into a Rhodococcus sp.CECT3014 �choG mutant strain [35] by electroporation. Positiveclones were selected in LB/Cm plates and confirmed by PCR. Thecholesterol oxidase test showed that cells transformed with theChol-4 choG gene were able to develop a brown colour indicatingthat the ChoG protein was expressed in the CECT3014 host cellsand was acting as an extracellular cholesterol oxidase (Fig. 3A).

It has been reported that the extracellular cholesterol oxidasecan occur in a secreted form and/or in a cell-surface-associatedform, both of them being products of the same gene [23,35]. Thecholesterol oxidase activity was assayed in both cell-free super-natant and Triton X-100 cell extracts of Rhodococcus sp. CECT3014�choG cultures harbouring either pTip-choG recombinant plasmidor the empty vector (Fig. 3B). Results quantitatively confirmed theexpression of Chol-4 ChoG in cells transformed with the pTip-choGplasmid but not with the pTip empty vector. Furthermore, the assayprovided a measurement of the relative oxidase activity of both, sol-uble and cellular fractions: only 20% of the total ChoG cholesteroloxidase activity is found in the LB culture medium (approximately0.161 Chox units per mg). This percentage is reported to be around30% in cultures of wild type Rhodococcus sp. CECT3014 [35], 42%in R. erythropolis [52] or 47% in Rhodococcus sp. GK1 [53] althougha quantitative comparison of the individual activity levels amongthem obviously cannot be done, given the multicopy occurrenceof choG and the inducible promoter that leads its expression inCECT3014 �choG pTip-choG cells.

3.4. Construction of a Chol-4 choG deletion mutant

Chol-4 cells are able to develop a brown colour when growing

in a cholesterol oxidase test plate (Fig. 4C) indicating the presenceof an extracellular enzyme. This extracellular activity was detectedin the supernatants of Chol-4 cultures growing in the presence ofcholesterol. The oxidase activity was assayed in these supernatants

38 L. Fernández de las Heras et al. / Journal of Steroid Biochemistry & Molecular Biology 139 (2014) 33– 44

Fig. 2. Transcription analysis of choG gene of Chol-4. Agarose gel electrophoresis of specific RT-PCR amplification products from Chol-4 choG transcripts. RNA samples wereisolated from Chol-4 cultures grown on minimal medium M457 supplemented with either sodium acetate (NaAc) or cholesterol (CHO). 25 ng of cDNA and CH104 and CH105primers (Table 2 and Fig. 4A) were used as indicated in Section 2.4. A PCR fragment of 0.7 kb was expected. RT−: negative controls (no retrotranscriptase included in thereaction) were run to exclude DNA contamination in RNA samples. The expression of kstD4 and kstD5, two ortholog genes of those coding for a FAD-binding dehydrogenase(NCIB: WP 017680123) and a fumarate reductase (NCIB: WP 017679800) from R. ruber, respectively, is also shown as internal control.

Fig. 3. Chol-4 ChoG activity assays using Rhodococcus sp. CECT3014 �choG mutant as host strain. (A) Cholesterol oxidase test plate in W minimal medium of Rhodococcuss from Ca ase acp Trito

ais

aFdbts0trbcohi

p. CECT3014 �choG mutant cells electroporated with pTip-QC1 harbouring choG

fter being electroporated with an empty pTip-QC1 vector (3). (B) Cholesterol oxidTip constructions. Activity was measured in both the supernatant and the 1% (v/v)

t different times of the growth curve, the maximum oxidase activ-ty being detected at the range from 2 to 3 OD600 nm units (data nothown).

To unequivocally assign the extracellular activity of this strain, choG deletion mutant of Chol-4 was prepared (see Section 2.6 andig. 4). By using the sacB selection system, an unmarked choG geneeletion was performed and the final construction was confirmedy PCR analysis: PCR with CH88-CH89 primers yielded amplifica-ion fragments of 3.6 and 2.4 kb in the wild type and the mutanttrain, respectively; PCR with CH104-CH105 primers produced a.7 kb amplification fragment in the wild type cell and no amplifica-ion in the mutant strain (Fig. 4B). Phenotypic analysis of this strainevealed that the �choG mutant had lost the ability to developrown colour in the cholesterol oxidase test plate (Fig. 4C), indi-

ating that choG gene codes for the only extracellular cholesterolxidase in this strain, similarly to the choG gene of CECT3014, thatas been reported to be the main extracellular cholesterol oxidase

n Rhodococcus sp. CECT3014 [35].

hol-4 (1). Control samples were the same cells before being electroporated (2) ortivity assayed in cultures of Rhodococcus sp. CECT3014 �choG mutant harbouring

n X-100 extracted fraction, as described in Section 2.5.

3.5. Characterization of the Chol-4 choG deletion mutant

The growth curve of Chol-4 in minimal medium supplementedwith cholesterol and cyclodextrin yielded a doubling time of 4.8 h(Table 3), in contrast to the 11.9 h described for strain CECT3014[35]. The addition of cyclodextrin to the culture medium improvedthe growing parameters of strains as it is described that cyclodex-trin facilitates the transport of the steroid substrate through themicrobial cell wall [54]. The Chol-4 �choG mutant strain lost itsextracellular cholesterol oxidase activity (Section 3.4) but was stillable to grow in minimal medium supplemented with cholesterol asthe only source of energy and carbon, reaching a similar maximumOD600 nm to the wild type strain (Table 3). It has also been reportedthat choD mutants of several Mycobacterium species [55–57] are

able to grow in cholesterol as the only source of carbon and energy,which, in turn, is consistent with the ability of R. jostii RHA1 to growon cholesterol, even though its cells lack the choG ORF, as we havementioned before (Section 3.1).

L. Fernández de las Heras et al. / Journal of Steroid Biochemistry & Molecular Biology 139 (2014) 33– 44 39

Fig. 4. Construction and verification of the Chol-4 �choG mutant. (A) Scheme of the construction of the deletion mutant. The upstream and downstream flanking regions(dashed boxes) of choG were PCR amplified and cloned together. These constructions were transferred into Chol-4 cells by conjugation and, after several selection processes,double recombinants (the deletion mutants) were obtained. Primers used in the colony PCR carried out to confirm the correct deletion in the mutant cells are indicatedas arrows. (B) Verification by colony PCR of the deletion in the choG gene of Chol-4 mutant. A 0.8% (w/v) agarose gel was run for 50 min at 92 V. The expected sizes of theamplicons were “CH88 + CH89′′: 3.6 kb in the wild type, 2.4 kb in the deletion mutant; “CH104 + CH105′′: 0.7 kb in the wild type, no amplification in the mutant. The GeneR rol ox�

io[thaA

fstroisf“i

TG

TN

uler 1 kb Ladder (Fermentas) was used as a molecular weight marker. (C) CholestechoG mutant (2).

However, while the growth curves of strain CECT3014 culturesn these conditions showed a very similar time-dependent patternf cholesterol use in both the wild type strain and the �choG mutant35], the Chol-4 �choG mutant strain showed a higher doublingime than the wild type (14 h, Table 3, Appendix C). On the otherand, such a difference was not appreciated between the wild typend the Chol-4 mutant when growing on cholestenone (Table 3,ppendix C).

Biotransformation of cholesterol in Chol-4 and its mutant wasollowed by TLC. Chloroform extract of the wild type culturehowed two spots which matched with cholestenone and choles-erol. We therefore made TLC semi-quantitative analysis of theemoval rates of both cholesterol and cholestenone in culturesf both the wild type and the �choG mutant of Chol-4 growingn cholesterol as the only carbon source (Fig. 5). The wild typetrain eliminated most of both the cholesterol and the cholestenone

rom the medium before 46 and 54 h, respectively, according to acholesterol-to-cholestenone” catabolic pathway. The correspond-ng behaviour of the �choG mutant revealed a much slower but

able 3rowth data of Chol-4 and its derived �choG mutant strain.

Chol-4 strain Minimal medium with 0.6 mg/mL cholesterol,16.5 mM cyclodextrin

Maximum OD (A600) Doubling time

WT strain 4.8 ± 0.4 5.4 ± 1.3

�choG mutant 4.2 ± 0.7 14.1 ± 3.2

hese data are the result of 3–6 independent experiments.o growth was observed in minimal medium with or without cyclodextrin.

idase test plate in 457 DSMZ minimal medium of Chol-4 wild-type (1) and Chol-4

still complete removal of cholesterol from the culture medium anda total absence of cholestenone accumulation, suggesting a majorinvolvement of the ChoG oxidase in the extracellular transforma-tion of cholesterol into cholestenone and explaining the fact thatthe mutant can still grow on cholesterol although at a much slowerspeed (Table 3). Less active intracellular cholesterol oxidases maybe responsible for this reaction. The evaluation of cholestenone inboth supernatant and cellular fractions of wild type cells (Fig. 5C)showed that the accumulation of this intermediate occurs mostlyin the exterior of the cell, depicting a situation where ChoG quicklyoxidizes cholesterol to cholestenone that accumulates outside thecell. Cholestenone transport may act as a limiting speed step in thisinitial phase of the pathway. The lack of ChoG in the �choG mutantstrain leads the cholesterol catabolism to rely upon its transportinside the cell prior its oxidation by secondary intracellular oxi-dases, which reduces the amount of available cholestenone inside

the cells and may explain the slow growth rate of this strain. Therelative involvement of the cholesterol uptake and its intracellu-lar oxidation on the growing pace of the Chol-4 �choG mutant is

Minimal medium with 0.6 mg/mLcholestenone, 16.5 mM cyclodextrin

(h) Maximum OD (A600) Doubling time (h)

2.7 ± 0.4 2.8 ± 0.52.3 ± 0.5 3.5 ± 0.7

40 L. Fernández de las Heras et al. / Journal of Steroid Biochemistry & Molecular Biology 139 (2014) 33– 44

Fig. 5. Thin-layer chromatography-based monitoring of cholesterol and cholestenone occurrence in Chol-4 cultures. Wild type Chol-4 (WT) and its �choG mutant (�choG)were grown in minimal medium with cholesterol and cultures were analyzed at different incubation times. Initial OD600 nm values were 0.1 and 0.2 for the wild type andthe mutant strain respectively. (A) TLC plates irradiated with short-wavelength UV light (254 nm) in which cholestenone and other UV absorbing intermediaries are easilyobserved. (B) TLC plates stained with 20% H2SO4 in order to see cholesterol spots. Standards (S) used are: 1: 1,4-androstadiene-3,17-dione (ADD); 2: 4-androstene-3,17-dione(AD); 3: cholesterol; 4: cholestenone; 5: cyclodextrin. (C) TLC-based cholestenone monitoring in supernatants and pellets of some aliquots of Chol-4 wild type culturesgrowing in cholesterol. TLC plates were irradiated with short-wavelength UV light (254 nm). T: total culture; S: supernatant; P: pellet; CN: cholestenone (standard sample).Results are representative of three independent experiments.

id Bioc

ao((t

s5(R(Ch6Ris

4

gcscrslctliCoastim

L. Fernández de las Heras et al. / Journal of Stero

question not yet defined. Whatever this relationship is, growthf the Chol-4 wild type in minimal medium with cholestenone0.6 mg/mL cholestenone, 16.5 mM cyclodextrin) is a bit fasterdoubling time of 2.8 ± 0.5 h; Table 3) than its cultures with choles-erol (doubling time of 5.4 ± 1.3 h; Table 3).

There have been some attempts to describe steroid transportersuch as the mce4 locus (that takes up cholesterol, 5-�-cholestanol,-�-cholestanone and �-sitosterol) or the camABCD/camM genesthat would work on cholate metabolites) [5,58,59] in R. jostiiHA1, but there is no data available for cholestenone transportalthough it is via the mce4 transporter in M. smegmatis, J.L. Garcia,IB, personal communication). Chol-4 contains in its genome thesd4AsupABmce4ABCDEF cluster (NCBI: ANGC01000018) that keeps2–92% amino acid sequence identity with the operon described for. jostii RHA1 [5]. This mce4 system might be putatively involved

n the uptake of cholesterol and cholestenone in Chol-4. Furthertudies on this subject are in progress.

. Conclusions

Among the 4 putative cholesterol oxidase ORFs found in theenome of Chol-4, choG is the only gene coding for the extra-ellular activity displayed by Chol-4. ChoG from Chol-4 has beenuccessfully expressed in Rhodococcus sp. CECT3014, where it couldomplement a chromosomal choG mutation and led the cells toegain an extracellular cholesterol oxidase phenotype. The expres-ion of choG seems to be the result of both a constitutive basalevel transcription plus a second induced process. Mutation ofhoG did not prevent the cells from growing in cholesterol ashe single source of energy and carbon, although they grow at aower rate. ChoG oxidase acts in the extracellular medium, yield-ng cholestenone that will be further transformed inside the cell.hol-4 �choG mutant is deprived of this activity and relies onlyn the intracellular enzymes to produce cholestenone. Cholesterolccumulates in the mutant strain culture medium because of its

low catabolic removal. The transport of cholesterol inside the cell,ogether with the probably less efficient intracellular oxidase activ-ties may give rise to the slower growth rate displayed by the

utant strain respect to that of the wild type cells.

Fig. A1.

hemistry & Molecular Biology 139 (2014) 33– 44 41

Acknowledgments

L.F.H. was in receipt of a scholarship from the ComplutenseUniversity of Madrid. This project was funded by project BFU2009-11545-C03-02 from the Spanish Ministry of Education andScience.

Appendix A.

The nucleotide sequence accession numbers are: Chol-4 [Gen-Bank ID: FJ842098], R. equi 103S [GenBank ID: FN563149], R.pyridinivorans AK37 [GenBank ID: AHBW01000027], R. erythro-polis PR4 [GenBank ID: AP008957], R. opacus B4 [GenBank ID:AP011115], R. jostii RHA1 [GenBank ID: CP000431]. Abbrevia-tions: kstD3: 3-ketosteroid-�1-dehydrogenase; kst4D (or tesI):3-ketosteroid-�4-(5�)-dehydrogenase; hsd4B: 2-enoyl acyl-CoAhydratase, putative dehydrogenase; MP: membrane protein;kshA: 3-ketosteroid-9�-hydroxylase; hsaE: 2-hydroxypenta-2,4-dienoate hydratase; merR: MerR family transcriptional regulator;nlpC/P60: NlpC/P60 superfamily papain-like enzyme; hyp: hypo-thetical protein. Homologous genes are marked using the samecolour.

Appendix B.

Alignment of the ChoG protein sequences of Rhodococcus sp.CECT3014 and Chol-4. Conserved amino acids and conservedsubstitutions are depicted. Signal peptide residues appear under-lined. FAD-binding sequences are in bold font. Catalytic residuesappear with an asterisk and in bigger size.

Appendix C.

Time course of growth of both wild type (WT) and �choG mutantChol-4 in 457 minimal medium supplemented with 0.6 mg/mL

cholesterol (CHO) or cholestenone (CN) and 16.5 mM cyclodex-trin. Cultures in minimal medium plus cyclodextrin were usedas a control (CD). A representative experiment of 4–8 replicas isshown.

42 L. Fernández de las Heras et al. / Journal of Steroid Biochemistry & Molecular Biology 139 (2014) 33– 44

Fig. B1.

Fig. C1.

id Bioc

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

L. Fernández de las Heras et al. / Journal of Stero

eferences

[1] K. Arima, M. Nagasawa, M. Bae, G. Tamura, Microbial transformation of sterols.Part I. Decomposition of cholesterol by microorganisms, Agric. Biol. Chem. 33(1969) 1636–1643.

[2] O. Drzyzga, J.M. Navarro Llorens, L. Fernández de las Heras, E. García Fernández,J. Perera, Gordonia cholesterolivorans sp. nov., a cholesterol-degrading actino-mycete isolated from sewage sludge, Int. J. Syst. Evol. Microbiol. 59 (Pt 5) (2009)1011–1015.

[3] L. Fernández de las Heras, E. García Fernández, J. María Navarro Llorens, J. Per-era, O. Drzyzga, Morphological, physiological, and molecular characterizationof a newly isolated steroid-degrading actinomycete, identified as Rhodococcusruber strain Chol-4, Curr. Microbiol. 59 (5) (2009) 548–553.

[4] J.M. Mathieu, W.W. Mohn, L.D. Eltis, J.C. LeBlanc, G.R. Stewart, C. Dresen, K.Okamoto, P.J. Alvarez, 7-Ketocholesterol catabolism by Rhodococcus jostii RHA1,Appl. Environ. Microbiol. 76 (1) (2010) 352–355.

[5] W.W. Mohn, R. van der Geize, G.R. Stewart, S. Okamoto, J. Liu, L. Dijkhuizen,L.D. Eltis, The actinobacterial mce4 locus encodes a steroid transporter, J. Biol.Chem. 283 (51) (2008) 35368–35374.

[6] X. Yang, N.M. Nesbitt, E. Dubnau, I. Smith, N.S. Sampson, Cholesterolmetabolism increases the metabolic pool of propionate in Mycobacterium tuber-culosis, Biochemistry 48 (18) (2009) 3819–3821.

[7] J.E. Griffin, J.D. Gawronski, M.A. Dejesus, T.R. Ioerger, B.J. Akerley, C.M. Sassetti,High-resolution phenotypic profiling defines genes essential for mycobacterialgrowth and cholesterol catabolism, PLoS Pathog. 7 (9) (2011) e1002251.

[8] S.T. Thomas, B.C. Vanderven, D.R. Sherman, D.G. Russell, N.S. Sampson, Path-way profiling in Mycobacterium tuberculosis: elucidation of cholesterol-derivedcatabolite and enzymes that catalyze its metabolism, J. Biol. Chem. 286 (51)(2011) 43668–43678.

[9] I. Uhía, B. Galán, S.L. Kendall, N.G. Stoker, J.L. García, Cholesterol metabolism inMycobacterium smegmatis, Environ. Microbiol. Rep. 4 (2) (2012) 168–182.

10] R. van der Geize, K. Yam, T. Heuser, M.H. Wilbrink, H. Hara, M.C. Anderton, E.Sim, L. Dijkhuizen, J.E. Davies, W.W. Mohn, L.D. Eltis, A gene cluster encodingcholesterol catabolism in a soil actinomycete provides insight into Mycobac-terium tuberculosis survival in macrophages, Proc. Natl. Acad. Sci. U.S.A. 104 (6)(2007) 1947–1952.

11] K.C. Yam, S. Okamoto, Adventures in Rhodococcus – from steroids to explosives,Can. J. Microbiol. 57 (3) (2011) 155–168.

12] J.L. García, I. Uhía, B. Galán, Catabolism biotechnological applications of choles-terol degrading bacteria, Microb. Biotechnol. 5 (6) (2012) 679–699.

13] J.K. Capyk, I. Casabon, R. Gruninger, N.C. Strynadka, L.D. Eltis, Activity of 3-ketosteroid 9�-hydroxylase (KshAB) indicates cholesterol side chain and ringdegradation occur simultaneously in Mycobacterium tuberculosis, J. Biol. Chem.286 (47) (2011) 40717–40724.

14] C. Dresen, L.Y. Lin, I. D’Angelo, E.I. Tocheva, N. Strynadka, L.D. Eltis, Aflavin-dependent monooxygenase from Mycobacterium tuberculosis involvedin cholesterol catabolism, J. Biol. Chem. 285 (29) (2010) 22264–22275.

15] K.Z. Rosloniec, M.H. Wilbrink, J.K. Capyk, W.W. Mohn, M. Ostendorf, R. van derGeize, L. Dijkhuizen, L.D. Eltis, Cytochrome P450 125 (CYP125) catalyses C26-hydroxylation to initiate sterol side-chain degradation in Rhodococcus jostiiRHA1, Mol. Microbiol. 74 (5) (2009) 1031–1043.

16] L. Pollegioni, L. Piubelli, G. Molla, Cholesterol oxidase: biotechnological appli-cations, FEBS J. 276 (23) (2009) 6857–6870.

17] A. Vrielink, S. Ghisla, Cholesterol oxidase: biochemistry and structural features,FEBS J. 276 (23) (2009) 6826–6843.

18] A.G. Smith, C.J. Brooks, The substrate specificity and stereochemistry, reversibil-ity and inhibition of the 3-oxo steroid �4-�5-isomerase component ofcholesterol oxidase, Biochem. J. 167 (1) (1977) 121–129.

19] A.G. Smith, C.J. Brooks, The mechanism of the isomerization of cholest-5-en-3-one to cholest-4-en-3-one by cholesterol oxidase [proceedings], Biochem. Soc.Trans. 5 (4) (1977) 1088–1090.

20] N. Doukyu, Characteristics, biotechnological applications of microbial choles-terol oxidases, Appl. Microbiol. Biotechnol. 83 (5) (2009) 825–837.

21] M.K. Hondalus, Pathogenesis and virulence of Rhodococcus equi, Vet. Microbiol.56 (3–4) (1997) 257–268.

22] N. Croteau, A. Vrielink, Crystallization and preliminary X-ray analysis of choles-terol oxidase from Brevibacterium sterolicum containing covalently bound FAD,J. Struct. Biol. 116 (2) (1996) 317–319.

23] J. Kreit, N.S. Sampson, Cholesterol oxidase: physiological functions, FEBS J. 276(23) (2009) 6844–6856.

24] A. Brzostek, B. Dziadek, A. Rumijowska-Galewicz, J. Pawelczyk, J. Dziadek,Cholesterol oxidase is required for virulence of Mycobacterium tuberculosis,FEMS Microbiol. Lett. 275 (1) (2007) 106–112.

25] Y. Pei, C. Dupont, T. Sydor, A. Haas, J.F. Prescott, Cholesterol oxidase (ChoE) isnot important in the virulence of Rhodococcus equi, Vet. Microbiol. 118 (2006)240–246.

26] J.F. Aparicio, J.F. Martin, Microbial cholesterol oxidases: bioconversion enzymesor signal proteins? Mol. Biosyst. 4 (8) (2008) 804–809.

27] M.V. Mendes, E. Recio, N. Anton, S.M. Guerra, J. Santos-Aberturas, J.F. Mar-tin, J.F. Aparicio, Cholesterol oxidases act as signaling proteins for thebiosynthesis of the polyene macrolide pimaricin, Chem. Biol. 14 (3) (2007)

279–290.

28] J.J. Ahire, A.A. Bhat, J.M. Thakare, P.B. Pawar, D.G. Zope, R.M. Jain, B.L. Chaud-hari, Cholesterol assimilation and biotransformation by Lactobacillus helveticus,Biotechnol. Lett. 34 (1) (2012) 103–107.

[

[

hemistry & Molecular Biology 139 (2014) 33– 44 43

29] L. Chenfeng, T. Yixin, W. Longgang, J. Wenming, C. Yili, Y. Shengli, Bioconversionof yolk cholesterol by extracellular cholesterol oxidase from Brevibacterium sp.,Food Chem. 77 (2002) 457–463.

30] N. Dogra, G.N. Qazi, Steroid biotransformation by different strains of Micrococ-cus sp., Folia Microbiol. (Praha) 46 (1) (2001) 17–20.

31] L. Fernández de las Heras, R. van der Geize, O. Drzyzga, J. Perera, J. Maria NavarroLlorens, Molecular characterization of three 3-ketosteroid-�1-dehydrogenaseisoenzymes of Rhodococcus ruber strain Chol-4, J. Steroid. Biochem. Mol. Biol.132 (3–5) (2012) 271–281.

32] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning: A Laboratory Manual,Cold Spring Harbor Laboratory Press, NY, 1989.

33] M. Seto, K. Kimbara, M. Simura, T. Hatta, M. Fukuda, K. Yano, A novel trans-formation of polychlorinated biphenyls by Rhodococcus sp. strain RHA1, Appl.Environ. Microbiol. 61 (9) (1995) 3353–3358.

34] U. Klein, G. Gimpl, F. Fahrenholz, Alteration of the myometrial plasma mem-brane cholesterol content with �-cyclodextrin modulates the binding affinityof the oxytocin receptor, Biochemistry 34 (42) (1995) 13784–13793.

35] L. Fernández de las Heras, V. Mascaraque, E. García Fernandez, J.M. Navarro-Llorens, J. Perera, O. Drzyzga, ChoG is the main inducible extracellularcholesterol oxidase of Rhodococcus sp. strain CECT3014, Microbiol. Res. 166(5) (2011) 403–418.

36] R.K. Flamm, D.J. Hinrichs, M.F. Thomashow, Introduction of pAM�1 into Listeriamonocytogenes by conjugation and homology between native L. monocytogenesplasmids, Infect. Inmun. 44 (1984) 157–161.

37] L. Fernández de Las Heras, S. Alonso, A. de la Vega de Leon, D. Xavier, J. Perera,J.M. Navarro Llorens, Draft genome sequence of the steroid degrader Rhodococ-cus ruber strain Chol-4, Genome Announc. 1 (3) (2013) e00215–e313.

38] T.A. Hall, BioEdit a user-friendly biological sequence alignment editor andanalysis program for windows 95/98/NT, Nucl. Acids Symp. Ser. 41 (1999)95–98.

39] T.N. Petersen, S. Brunak, G. von Heijne, H. Nielsen, SignalP 4.0: discriminatingsignal peptides from transmembrane regions, Nat. Methods 8 (2011) 785–786.

40] M.M. Bradford, A rapid and sensitive method for the quantitation of micro-gram quantities of protein utilizing the principle of protein-dye binding, Anal.Biochem. 72 (1976) 248–254.

41] Y. Nishiya, N. Harada, S.I. Teshima, M. Yamashita, I. Fujii, N. Hirayama, Y.Murooka, Improvement of thermal stability of Streptomyces cholesterol oxi-dase by random mutagenesis and a structural interpretation, Protein Eng. 10(3) (1997) 231–235.

42] J. Kreit, G. Lefebvre, A. Elhichami, P. Germain, M. Saghi, A colorimetric assay formeasuring cell-free and cell-bound cholesterol oxidase, Lipids 27 (6) (1992)458–465.

43] R. van der Geize, G.I. Hessels, R. van Gerwen, P. van der Meijden, L.Dijkhuizen, Unmarked gene deletion mutagenesis of kstD, encoding 3-ketosteroid �1-dehydrogenase, in Rhodococcus erythropolis SQ1 using sacB ascounter-selectable marker, FEMS Microbiol. Lett. 205 (2) (2001) 197–202.

44] A. Schafer, A. Tauch, W. Jager, J. Kalinowski, G. Thierbach, A. Puhler, Smallmobilizable multi-purpose cloning vectors derived from the Escherichia coliplasmids pK18 and pK19: selection of defined deletions in the chromosome ofCorynebacterium glutamicum, Gene 145 (1) (1994) 69–73.

45] M. Muller, Twin-arginine-specific protein export in Escherichia coli, Res. Micro-biol. 156 (2) (2005) 131–136.

46] C. Barbey, A. Budin-Verneuil, S. Cauchard, A. Hartke, C. Laugier, V. Pichereau,S. Petry, Proteomic analysis and immunogenicity of secreted proteins fromRhodococcus equi ATCC 33701, Vet. Microbiol. 135 (3–4) (2009) 334–345.

47] O. Drzyzga, L. Fernández de las Heras, V. Morales, J.M. Navarro Llorens, J. Perera,Cholesterol degradation by Gordonia cholesterolivorans, Appl. Environ. Micro-biol. 77 (14) (2011) 4802–4810.

48] J. Navas, B. Gonzalez-Zorn, N. Ladron, P. Garrido, J.A. Vazquez-Boland, Identi-fication and mutagenesis by allelic exchange of choE, encoding a cholesteroloxidase from the intracellular pathogen Rhodococcus equi, J. Bacteriol. 183 (16)(2001) 4796–4805.

49] Y. Xin, H. Yang, X. Xia, L. Zhang, C. Cheng, G. Mou, J. Shi, Y. Han, W. Wang, Affinitypurification of a cholesterol oxidase expressed in Escherichia coli, J. Chromatogr.B: Anal. Technol. Biomed. Life Sci. 879 (13–14) (2011) 853–858.

50] F. Volonte, L. Pollegioni, G. Molla, L. Frattini, F. Marinelli, L. Piubelli, Produc-tion of recombinant cholesterol oxidase containing covalently bound FAD inEscherichia coli, BMC Biotechnol. 10 (2010) 33.

51] Y. Murooka, T. Ishizaki, O. Nimi, N. Maekawa, Cloning and expression of a Strep-tomyces cholesterol oxidase gene in Streptomyces lividans with plasmid pIJ702,Appl. Environ. Microbiol. 52 (6) (1986) 1382–1385.

52] M. Sojo, R. Bru, D. López-Molina, F. García-Carmona, J.C. Argüelles, Cell-linkedand extracellular cholesterol oxidase activities from Rhodococcus erythropolis.Isolation and physiological characterization, Appl. Microbiol. Biotechnol. 47 (5)(1997) 583–589.

53] A. Elalami, J. Kreit, A. Filali-Maltouf, J. Boudrant, P. Germain, Characterization ofa secreted cholesterol oxidase from Rhodococcus sp. GK1 (CIP 105 335), WorldJ. Microbiol. Biotechnol. 15 (1999) 579–585.

54] R. Kumar, J.S. Dahiya, D. Singh, P. Nigam, Biotransformation of cholesterol usingLactobacillus bulgaricus in a glucose-controlled bioreactor, Bioresour. Technol.78 (2) (2001) 209–211.

55] A. Brzostek, A. Rumijowska-Galewicz, B. Dziadek, E.A. Wojcik, J. Dziadek, ChoD,HsdD can be dispensable for cholesterol degradation in mycobacteria, J. SteroidBiochem. Mol. Biol. 134 (2013) 1–7.

56] T.V. Ivashina, V.M. Nikolayeva, D.V. Dovbnya, M.V. Donova, Cholesterol oxidaseChoD is not a critical enzyme accounting for oxidation of sterols to 3-keto-4-ene

4 id Bioc

[

[

[

194 (24) (2012) 6720–6727.

4 L. Fernández de las Heras et al. / Journal of Stero

steroids in fast-growing Mycobacterium sp. VKM Ac-1815D, J. Steroid Biochem.Mol. Biol. 129 (1–2) (2012) 47–53.

57] I. Uhía, B. Galán, V. Morales, J.L. García, Initial step in the catabolism of choles-

terol by Mycobacterium smegmatis mc2 155, Environ. Microbiol. 13 (4) (2011)943–959.

58] U. Haussmann, D.A. Wolters, B. Franzel, L.D. Eltis, A. Poetsch, Physiological adap-tation of the Rhodococcus jostii RHA1 membrane proteome to steroids as growthsubstrates, J. Proteome Res. 12 (3) (2013) 1188–1198.

[

hemistry & Molecular Biology 139 (2014) 33– 44

59] K. Swain, I. Casabon, L.D. Eltis, W.W. Mohn, Two transporters essential for reas-similation of novel cholate metabolites by Rhodococcus jostii RHA1, J. Bacteriol.

60] N. Nakashima, T. Tamura, Isolation and characterization of a rolling-circle-type plasmid from Rhodococcus erythropolis and application of the plasmidto multiple-recombinant-protein expression, Appl. Environ. Microbiol. 70 (9)(2004) 5557–5568.