Scholarly article on topic '
                        Mycobacterium smegmatis
 is a suitable cell factory for the production of steroidic synthons'

Mycobacterium smegmatis is a suitable cell factory for the production of steroidic synthons Academic research paper on "Industrial Biotechnology"

CC BY
0
0
Share paper
Academic journal
Microbial Biotechnology
OECD Field of science
Keywords
{""}

Academic research paper on topic " Mycobacterium smegmatis is a suitable cell factory for the production of steroidic synthons"

microbial biotechnology

Mycobacterium smegmatis is a suitable cell factory for the production of steroidic synthons

Beatriz Gaian,1# Iria Uhía,1'2'* Esther García-Fernandez,1 Igor Martínez,1 Esther Bahíllo,3 Juan L. de la Fuente,3 Jose L. Barredo,3 Lorena Fernández-Cabezón1 and José L. García1*

1 Department of Environmental Biology, Centro de Investigaciones Biológicas, Consejo Superior de Investigaciones Cienóficas, Ramiro de Maeztu 9, 28040 Madrid, Spain.

2MRC Centre for Molecular Bacteriology and Infection, Department of Medicine, Imperial College London, London SW7 2AZ, UK.

3Department of Biotechnology, Gadea Biopharma, Parque Tecnoloógico de Leoón, Nicostrato Vela s/n, 24009 León, Spain.

Summary

A number of pharmaceutical steroid synthons are currently produced through the microbial side-chain cleavage of natural sterols as an alternative to multistep chemical synthesis. Industrially, these synthons have been usually produced through fermentative processes using environmental isolated microorganisms or their conventional mutants. Mycobacterium smegmatis mc2155 is a model organism for tuberculosis studies which uses cholesterol as the sole carbon and energy source for growth, as other mycobacterial strains. Nevertheless, this property has not been exploited for the industrial production of steroidic synthons. Taking advantage of our knowledge on the cholesterol degradation pathway of M. smegmatis mc2155 we have demonstrated that the MSMEG_6039 (kshB1) and MSMEG_5941 (kstD1) genes encoding a reductase component of the 3-ketosteroid ^-hydroxylase (KshAB) and a ketosteroid D'"-dehydrogenase (KstD), respectively, are indispensable enzymes for

Received 21 June, 2016; revised 19 September, 2016; accepted 21

September, 2016. *For correspondence. E-mail: jlgarcia@cib.csic.es;

Tel. +34 918373112; Fax +34 915360432.

Microbial Biotechnology (2016) 0(0), 000-000

doi:10.1111/1751-7915.12429

#These authors contributed equally to this work.

Funding Information

The technical work of A. Valencia is greatly appreciated. We thank I. Calvillo for assistance in the analytical GC/MS experiments. This study was supported by grants from the Ministry of Science and Innovation (BFU2006-15214-C03-01, BFU2009-11545-C03-03) and Ministry of Economy and Competitiveness (BIO2012-39695-C02-01).

the central metabolism of cholesterol. Therefore, we have constructed a MSMEG_6039 (kshB1) gene deletion mutant of M. smegmatis MS6039 that transforms efficiently natural sterols (e.g. cholesterol and phytos-terols) into 1,4-androstadiene-3,17-dione. In addition, we have demonstrated that a double deletion mutant M. smegmatis MS6039-5941 [DMSMEG_6039 (DkshB1) and DMSMEG_5941 (DkstD1)] transforms natural sterols into 4-androstene-3,17-dione with high yields. These findings suggest that the catabolism of cholesterol in M. smegmatis mc2155 is easy to handle and equally efficient for sterol transformation than other industrial strains, paving the way for valuating this strain as a suitable industrial cell factory to develop a la carte metabolic engineering strategies for the industrial production of pharmaceutical steroids.

Introduction

Androstenedione (4-androstene-3,17-dione; AD) and androstadienedione (1,4-androstadiene-3,17-dione; ADD) are key intermediates of microbial steroid metabolism (García et al., 2012). These compounds belong to the 17-keto steroid family and are used as the starting materials for the preparation of different pharmaceutical steroids (Malaviya and Gomes, 2008; Donova and Egorova, 2012; Garcia et al., 2012). These synthons can be produced by microbial side-chain cleavage of cholesterol or phytosterols as an alternative to multi-step chemical synthesis based on digoxigenin, a steroid found exclusively in the flowers and leaves of Digitalis plants, as a starting material. Industrially, AD and ADD have been produced through fermentative processes using wild microorganisms that have been subsequently modified and optimized by conventional mutagenic procedures (Donova et al., 2005; Andor et al., 2006; Donova and Egorova, 2012; Garcia et al., 2012). Mycobacterium spp. NRRL 3805B and 3683B capable of forming AD and ADD from sterols, respectively, are examples of these mutants used at industrial scale (Donova et al., 2005; Donova and Egorova, 2012). A drawback of the AD and ADD industrial production based on these wild strains is the usual and concomitant accumulation of unwanted byproducts which hinder downstream processes. The use of molecularly defined mutants has been envisioned to avoid such drawbacks, but the lack of genetic data on the microbial catabolism of steroids has hampered the

© 2016 The Authors. Microbial Biotechnology published by John Wiley & Sons Ltd and Society for Applied Microbiology.

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and

reproduction in any medium, provided the original work is properly cited.

construction of genetically engineered mutants so far. Nevertheless, some attempts have been conducted to construct site-directed mutant strains of Rhodococcus (i.e. the best characterized cholesterol-degrading organism) to produce AD, ADD and 9a-hydroxy-4-androstene-3,17-dione (9OH-AD) from natural sterols (e.g. cholesterol or phytosterols), but these mutants have not been used at industrial scale yet (van der Geize et al., 2000, 2001a,b, 2002a,b, 2008; Wilbrink et al, 2011; Yeh et al, 2014).

To create a la carte mutants able to produce AD and ADD from cholesterol or phytosterols we have tested Mycobacterium smegmatis mc2155 as a model strain

based on our current knowledge on sterol catabolism in this microorganism (Fig. 1) (Garcia et al., 2012). The 3-ketosteroid 9a-hydroxylase (Ksh) has been proposed as the key enzyme for ring-B opening, and therefore, the removal of this activity should render ADD as end-product. The Ksh enzymes of Rhodococcus and Mycobac-terium have been characterized as two-component monooxygenases, composed of an oxygenase (KshA) and a ferredoxin reductase (KshB) (van der Geize et al., 2002a; Capyk et al., 2009, 2011; Petrusma et al., 2009, 2012, 2014; Hu et al, 2010; Bragin et al., 2013; Penfield et al., 2014). Therefore, theoretically, a deletion mutant of one of the Ksh encoding genes, i.e. kshA or kshB,

Cholesterol

Side chain oxidation

3ß-HSD

5-cholesten-3-one

4-cholesten-3-one

i h [> h

h j h "KstD f^T H j h

AD ADD

KshAB KshAB

h j) h

T oh| h KstD f^t o^ h

9OH-AD 9OH-ADD

& y ho-

H HsaAB * fs HsaC

3,4-DHSA

4,9-DSHA HsaD

Pyruvate Propionyl-CoA

HsaEFG

Central metabolism -

FadE30?

2-hydroxy-2,4-hexadienoic acid +

IpdAB FadD3

5OH-HIP-CoA HIP-CoA

Fig. 1. Proposed pathway for cholesterol degradation in Mycobacterium smegmatis. Cholest-4-en-3-one or any of the subsequent metabolites from degradation of the side-chain up to (and including) AD may undergo a dehydrogenation reaction to introduce a double bond in the position 1, leading to compound cholest-1,4-diene-3-one in the case of cholest-4-en-3-one, or to the corresponding 1,2-dehydro derivatives for other molecules. The side-chain degradation of this compounds will be identical to that of the cholest-4-en-3-one to the common intermediate 9a-hydroxyandrosta-1,4-diene-3,17-dione (9OHAD). 3b-hydroxysteroid dehydrogenase (3P-HSD), 3-ketosteroid-D1-dehydrogenase (KstD), 3-ketos-teroid-9a-hydroxylase (KshAB), 3-hydroxy-9,10-secoandrosta-1,3,5(10)-trien-9,17-dione (3-HSA), 3,4-dihydroxy-9,10-secoandrosta-1,3,5(10)-trien-9,17-dione (3,4-HSA), 4,5,9,10-diseco-3-hydroxy-5,9,7-trioxoandrosta-1(10),2-diene-4-oic acid (4,9-DSHA), 3aa-H-4a(3'-propionic acid)-7ap-methylhexahydro-1,5-indanedione (HIP), 3aa-H-4a(3'-propionic acid)-5a-hydroxy-7b-methylhexahydro-1-indanone (5OH-HIP), 3-hydroxy-9,10-secoandrosten-1,3,5(10)-trien-9,17-dione dyoxigenase (HsaC), 4,5,9,10-diseco-3-hydroxy-5,9,7-trioxoandrosta-1(10),2-diene-4-oic acid hydroxylase (HsaD), 2-hydroxy-2,4-hexadienoic acid hydratase (HsaE), 4-hydroxy-2-hydroxy-2-ketovalerate aldolase (HsaF), acetaldehyde dehydrogenase (HsaG), HIP-CoA transferase (LpdAB and FadD3), FadE30 (Acyl-CoA dehydrogenase).

should accumulate ADD. On the other hand, 3-ketoster-oid A1-dehydrogenase (KstD) is the key enzyme that transforms AD into ADD and thus, strains lacking KstD should theoretically accumulate 9OH-AD. Finally, double mutants in Ksh and KstD should render AD as the main product for sterol degradation (Fig. 1) (Garcia et al., 2012).

Nevertheless, these previous assumptions cannot be always easily confirmed by gene inactivation since cholesterol catabolic pathway usually presents functional redundancy for some steps of the pathway; this means that one catalytic step can be carried out by several homologous enzymes. Two KstD enzymes were reported in Rhodococcus erythropolis SQ1 (van der Geize et al., 2000) and accordingly, targeted disruption of only one of the kstD genes did not result on the accumulation of intermediates, but the strain lacking both dehydrogenase activities was able to convert AD into 9OH-AD stoichiometrically (van der Geize et al., 2002b). Similarly, Fernández de las Heras et al. (2012) have demonstrated the existence of three KstD activities in Rhodococcus ruber strain Chol-4 and showed that the triple KstD mutant was able to convert AD into 9OH-AD. However, these mutants were unable to accumulate 9OH-AD from cholesterol or phytosterols (van der Geize et al., 2001a; Fernandez de las Heras et al., 2012). Brzostek et al. (2009) have identified six putative KstD enzymes within the M. smegmatis genome and a targeted disruption of one of them (KstD1) resulted in partial inactivation of the cholesterol degradation pathway and the consequent accumulation of AD. More recently, Wei et al. (2010) have identified and deleted a kstD gene (named ksdDM) in Mycobacterium neoaurum NwIB-01, a strain isolated from a sterol contaminated soil that naturally accumulates ADD from phytosterols. Interestingly, the resulting deleted mutant (named NwIB-2) was able to accumulate AD from phytosterols, but ADD is still present in the culture medium, suggesting that this strain contains other KstD enzymes (Wei et al., 2010). Three homologues of KstD have been characterized in M. neoaurum ATCC 25795 and thus, single deletion of these genes failed to result in a stable and maximum accumulation of 9OH-AD due to residual KstD activities (Yao et al., 2014).

On the other hand, van der Geize et al. (2002a, 2008) have demonstrated that Ksh mutants of R. erythropolis SQ1 deleted in kshA or kshB genes can produce ADD from AD, but surprisingly they do not accumulate AD, ADD or other metabolites during sterol conversion. Interestingly, the degradation of phytosterols was not impaired in a kshA- mutant and rates of degradation were comparable to those of the parent strain, suggesting that Rhodococcus has alternative enzymes to catab-olize phytosterols. Moreover, the kshB- mutant failed to

cleave the side-chain of sterols and, although phytosterols were oxidized to their stenone derivatives, they were not metabolized further, suggesting that kshB could be also involved in the degradation of sterol side-chain in Rhodococcus (van der Geize et al., 2002a). The situation can be more complex taking into account that some strains of Rhodococcus may contain up to five KshA homologous proteins, each displaying unique steroid induction patterns and substrate ranges, confirmed that the 9a-hydroxylation can take place at different steps of steroid oxidation (Petrusma et al., 2011).

On the basis of our knowledge on cholesterol metabolism in M. smegmatis (Uhia et al., 2012) we detected that this organism has several important differences with the sterol catabolism in Rhodococcus and we anticipated that these differences might render M. smegmatis as a suitable cell factory for the metabolic manipulation of this pathway. The results presented below demonstrate this hypothesis and show that M. smegmatis is an ideal cell factory to develop metabolic engineering strategies for the industrial production of AD, ADD or other steroid intermediates using natural sterols as feedstock.

Results

Identification of Ksh and KstD enzymes in M. smegmatis

In contrast to the frequent observation that the components of bacterial multicomponent oxygenases are encoded in a single operon, the kshA and kshB genes encoding Ksh activity in Rhodococcus (van der Geize et al., 2002a) or in M. tuberculosis (Cole et al., 1998) are located far from each other in the genome. Nevertheless, using the sequence of the annotated kshA and kshB genes from Rhodococcus (van der Geize et al., 2002a, 2008) and M. tuberculosis (Capyk et al., 2009) as probes, we have localized the corresponding ortho-logues in M. smegmatis (Table 1). This analysis revealed that there are at least two genes encoding putative oxygenase components, named kshA1 and kshA2, and two genes encoding putative reductase components, named kshB1 and kshB2. This means that M. smegmatis could produce theoretically four different Ksh hydroxylases, i.e. Ksh1 (KshA1B1), Ksh2 (KshA2B2), Ksh3 (KshA1B2), Ksh4 (KshA2B1). However, a detailed analysis of the upstream sequences of these genes showed that only the MSMEG_5925 (kshA1) and MSMEG_6039 (kshB1) genes are preceded by promoters inducible by cholesterol, having the consensus operator sequence for the binding of the KstR repressor, one of the regulators of cholesterol catabolism (named KstR regulon) (Kendall et al., 2007). Moreover, microarray expression experiments carried out on M. smegmatis indicate that kshA1 and kshB1 are induced 1.9-fold and 7.0-fold, respectively, in cholesterol

Table 1. In silico analysis of M. smegmatis mc2 155 genome.

Protein Genea M. smegmatis mc2 155 Protein (aa length) Identityb (%) KstR-regulon operator Induction foldc

KshA MSMEG_5925 383 67 Yes 7.00

MSMEG_2870 386 53 No 1.00

KshB MSMEG_6039 353 63 Yes 1.88

MSMEG_2893 351 56 No 1.21

KstD MSMEG_5941 566 65 Yes 13.00

MSMEG_4864 587 39 No -1.46

MSMEG_2869 558 40 No -1.17

MSMEG_2867 522 31 No -1.14

MSMEG_4870 546 30 No -1.35

MSMEG_5835 547 17 No 1.14

a. Gene identifications correspond to the last annotation of M. smegmatis genome and are different to the annotations described by Brzostek et al. (2005).

b. Identities were established using the proteins of R. jostii RHA1 as reference.

c. Induction fold was calculated comparing gene expression in M. smegmatis cultured in cholesterol verus glycerol containing media (Uhia et al., 2012).

with respect to glycerol (Table 1) (Uhia et al., 2012). On the contrary, kshA2 and kshB2 are not induced by cholesterol under the tested conditions (Table 1) (Uhia et al., 2012). These results strongly suggest that most probably only Ksh1 encoded by kshA1B1 is involved in the metabolism of cholesterol in M. smegmatis.

Among the six putative kstD genes identified in the M. smegmatis genome by Brzostek et al. (2009) that might encode the KstD-like activity only the MSMEG_5941 (kstDI) gene is controlled by a promoter containing the consensus KstR operator sequence and is differentially expressed in the presence of cholesterol (Table 1) (Uhia et al., 2012). This observation suggests that in M. smegmatis most probably only kstDI was involved in the catabolism of cholesterol.

Therefore, based on these analyses we anticipated the hypothesis that in M. smegmatis a single deletion of kshBI (or kshAI) and a double deletion of kshBI (or kshAI) and kstDI might be sufficient to generate mutants able to accumulate ADD or AD, respectively, when cultured in the presence of sterols.

Construction of the DkshB1 M. smegmatis mutant

As mentioned above, to eliminate the two-component Ksh1 hydroxylase activity (KshA1B1) in M. smegmatis, we assumed that it was enough to delete the MSMEG_6039 (kshBI) gene encoding the reductase component, thus we constructed the MS6039 mutant (DkshBI) (Fig. S1). Ksh1 activity can be eliminated either by deleting kshBI or kshAI (or both), but we decided to test the single deletion of kshBI instead of kshAI, because in the absence of the corresponding oxygenase component, the reductase component of mul-ticomponent oxygenases usually acts as a gratuitous scavenger of reduction power, generating futile cycles that might reduce the efficiency of the biotransformation

process (Galan et al., 2000; Blank et al., 2010) and therefore, in this sense, it can be more convenient to suppress the reductase instead the oxygenase component. However, it is also true that other reductase enzymes present in the cell could replace their function. Nevertheless, we also decided to delete kshBI because the effect of the deletion of the kshB homologous gene has already been analysed in Rhodococcus (van der Geize et al., 2002a,b) and we would like to compare the performance of M. smegmatis having the same gene deletion.

The engineered MS6039 mutant (DkshBI) was unable to efficiently grow in cholesterol or phytosterols as a sole carbon and energy source when compared with the wild-type strain (Fig. 2A). However, the mutant perfectly grows using glycerol as a carbon source (Fig. 2B). To confirm that the deletion of kshBI was responsible for the observed phenotype, we transformed the MS6039 mutant with the plasmid pMV6039 harbouring the kshBI gene. Figure 2D shows that the complemented MS6039 (pMV6039) strain recovered the capacity to grow in sterols as a sole carbon and energy source. These results confirmed that kshBI plays an essential role in the catabolism of sterols in M. smegmatis. In addition, considering that the MS6039 mutant is unable to efficiently grow in sterols even at long incubation periods, we can conclude that the KshB1 reductase activity of Ksh1 cannot be replaced by other mycobacterial reductases, i.e. KshB2 or other KshB-like reductases, in the tested conditions, either because they are not expressed in these conditions or because they cannot interact with the KshA1 oxygenase component.

These experiments also revealed that wild-type M. smegmatis is able to use efficiently different phytos-terols as a carbon and energy source (Fig. 2C), and this is remarkable because the capacity of this strain to grow

1.2 1 0.8 0.6 0.4 0.2 0

1.2 1 1 -0.8 0.6 0.4 -0.2 ; 0 0

40 60 Time (h)

20 40 60 Time (h)

3.5 3 2.5 2 1.5 1 0.5 0

20 40 60 80 Time (h)

Time (h)

Fig 2. Growth curves of M. smegmatis mc2155, mutants and complemented strains cultured in shake flasks with different carbon sources. (A) Strains cultured with 0.4 g r1 of cholesterol (mc2155 (squares), MS6039 (circles) and MS6039-5941 (triangles)); (B) Strains cultured with 18 mM glycerol (mc2155 (squares), MS6039 (circles) and MS6039-5941 (triangles)); (C) Strains cultured with 0.4 g l_1 of phytosterols (mc2155 (squares), MS6039 (circles) and MS6039-5941 (triangles)); (D) Gene complementation of MS6039 mutant strain cultured with 1.8 mM of cholesterol (mc2155 (pMV261) (diamonds), MS6039 (pMV261) (squares) and MS6039 (pMV6039) (circles)). The data reported are the averages of three different assays.

on a mixture of phytosterols has not been well documented in the literature so far.

On the other hand, the slight growth of MS6039 in the presence of sterols suggested that the mutant was able to degrade their side-chain and predicted a possible accumulation of metabolic intermediates in the culture medium. As expected, Figure 4 shows the accumulation of ADD in the culture medium of MS6039 mutant when cultured in shake flasks, whereas M. smegmatis wildtype strain completely mineralizes cholesterol (or phytos-terols) and does not accumulate any intermediate in these culture conditions (data not shown). The ADD molar yield for the transformation of cholesterol or phy-tosterols was 91% for both feedstocks. Nevertheless, we have also detected the presence of small amounts of 22-hydroxy-23,24-bisnorchol-1,4-dien-3-one (1,4-HBC, 20-hydroxymethylpregna-1,4-dien-3-one) as a by-product (Fig. 4C and D).

In addition, we have tested the production of ADD from phytosterols in 5- and 20-l stirred jar bioreactors using higher concentrations of phystosterols (20 g r1) as substrate to study the industrial potential of MS6039 mutant strain. By carrying several replicates of these experiments in different operational conditions, we concluded that between 55% and 70% of added phytosterols are consumed and the molar yield to ADD varies from 67% to 80%, depending on the bioreactor scale. AD was also obtained in a yield from consumed phytos-terols between 22% and 30%; other by-products, such as 1,4-HBC were detected as traces (Fig. S2). These

experiments carried out at high sterol concentrations showed a significant contamination of unconverted AD suggesting that KstD activity is a bottleneck for the complete transformation of phytosterols into ADD at industrial scale. This problem can be overcome by changing the operational conditions (e.g. preinoculation growth, inoculation conditions, bioreactor agitation configuration, aeration flow, composition of culture medium, pH control, substrate additions, etc.) in combination with the overexpression of kstD genes in MS6039 that lead to a significant increase in ADD/AD ratio (Gadea personal communication).

The results contrast with the data obtained with the equivalent kshB- mutant of R. erythropolis strain SQ1 (RG4 mutant), which only accumulates sitostenone from p-sitosterol, because apparently, the RG4 mutant does not degrade the side-chain of p-sitosterol. To explain this performance, it was proposed that the KshB reductase from R. erythropolis was not only involved in the production of 9OH-AD, but also in sterol side-chain removal (van der Geize et al., 2002b). In M. smegmatis, the absence of KshB1 does not hinder the complete degradation of the sterol side-chain (see Discussion).

Construction of the DkshB1 and DkstD1 M. smegmatis double mutant

According to our previous genomic analysis, it should be possible to produce AD from sterols in M. smegmatis by eliminating the KstD activity in the MS6039 mutant.

Therefore, to test this hypothesis, we constructed the M. smegmatis MS6039-5941 double mutant (DkstD1, DkshB1) by deleting the MSMEG_5941 (kstDI) gene in the MS6039 (DkshBI) mutant (Fig. S1). As expected, the MS6039-5941 (DkshB1-DkstD1) double mutant was unable to efficiently grow in cholesterol or phytosterols as the sole carbon and energy source, but the double mutant grows in glycerol at the same rate than the wildtype strain (Fig. 2B).

Accordingly to our predictions, when the MS6039-5941 double mutant was cultured in minimal media containing glycerol as carbon and energy source and cholesterol (or phytosterols) as feedstock, the culture yields a large accumulation of AD in shake flasks (Fig. 4A and B). The AD molar yields for sterol transformations were 90% and 84% when cholesterol or phytosterols were used as substrates respectively. Interestingly, in these culture conditions, small amounts of 22-hydroxy-23,24-bisnorchol-4-en-3-one (4-HBC, 20-hydroxymethylpregna-4-en-3-one) were accumulated during the biotransformation (Fig. 4C and D).

These results support the hypothesis that although M. smegmatis has six putative KstD dehydrogenases, the KstDI enzyme encoded by the MSMEG_5941 gene is the main KstD used for sterol catabolism in this bacterium. This observation contrasts with other cholesterol-degrading bacteria, where several KstD dehydrogenases are involved in the metabolism of sterols (van der Geize et al., 2000, 2002b; Fernández de las Heras et al., 2012; Yao et al, 2014).

The MS6039-5941 double mutant was tested in 2-l jar bioreactor in the presence of 10 g Г1 of phytosterols. An almost complete transformation of phytosterols into AD (88-90%), 4-HBC (10-11%) and very small amounts of ADD and 1,4-HBC were detected in several replicates of this experiment (Fig. S3).

Interestingly, the accumulation of small amounts of ADD observed in the bioreactor experiments suggests that one of the other identified kstD2, kstD3, kstD4, kstD5 or kstD6 homologous genes (Table 1) is somewhat active in the mutant. Thus, to further improve the AD production yield, this residual KstD activity should be identified and eliminated. Moreover, in this sense, it will be also important to understand why the three last carbons at C-17 are not efficiently converted into propionyl-CoA, rendering 4-HBC and 1,4 HBC as by-products of the pathway.

Discussion

The first remarkable finding presented above is the experimental demonstration that M. smegmatis mc2155 is able to efficiently metabolize mixtures of phytosterols as carbon and energy sources through the same

catabolic pathway utilized to mineralize cholesterol (Fig. 1) (Uhia et al., 2012). This finding is important because it supports the proposal of considering M. smegmatis as a useful cell factory for the industrial production of steroidic synthons from raw sterols. Surprisingly, in spite of the large accumulated knowledge on M. smegmatis mc2155 biology, this organism has not been used as a cell factory for this industrial purpose so far. The only case reported in the literature is an antibiotic resistant mutant deposited as M. smegmatis VKPM Ac-1552 in the Russian National Collection of Industrial Microorganisms (VKPM) that appears to transform ster-ols into AD (Russian Federation Patent no. 2 126 837 (1999), reviewed in Donova (2007)).

The observation that the growth of M. smegmatis mc2155 in sterols is impaired by a deletion of the kshB1 gene (MSMEG_6039) suggests that KshB1 is the main reductase component of the two-component KshAB 9a-hydroxylase in the cholesterol degradative pathway of this organism. Apparently, this reductase activity is not redundant in M. smegmatis and thus, it cannot be replaced by similar enzymes, as it has been demonstrated for other key enzymes of the pathway (Uhia et al., 2011). In addition, the kshBI gene does not appear to be critical for the degradation of the side-chain of sterols in M. smegmatis, as suggested in Rhodococ-cus (van der Geize et al., 2002a), and therefore, its absence does not impair an efficient transformation of sterols into AD or ADD in the mutant strain.

On the other hand, the significant accumulation of AD in the MS6039-5941 (DkshB1-DkstDI) double mutant ascribes a fundamental role to the kstD1 gene (MSMEG_5941) in the catabolism of cholesterol. Nevertheless, in this mutant, we have observed some enzyme redundancy because we were able to detected small amounts of ADD when high concentrations of phytos-terols are transformed. Therefore, we can conclude that at least one of the other five putative KstD enzymes identified in M. smegmatis mc2155 (Table 1) can also replace this function, but this alternative KstD activity is very low and does not appear to fulfil the relevant function observed in other cholesterol-degrading strains (van der Geize et al., 2000, 2002b; Fernindez de las Heras et al., 2012; Yao et al, 2014).

The results presented in this work suggest that the metabolism of sterols in M. smegmatis mc2155 concerning to the two central/key enzymes investigated, i.e. KstD and Ksh, is less redundant than in Rhodococcus. Perhaps, the low redundancy of these central enzymes can explain why many strains currently used at industrial scale to transform phytosterols into AD, ADD or 9OH-AD are mycobacterial mutants obtained by conventional mutagenesis, i.e. because single mutations like those produced in this work might render a producer strain.

Genomic analyses of some of these industrial mutants have confirmed this hypothesis (data not shown).

Another interesting difference observed between the sterol catabolism of Mycobacterium and Rhodococcus is the presence of 4-HBC and 1,4-HBC alcohols as metabolic by-products in the cultures of our mutants (Figs 3C, D and 4C, D). It has been described that the cultures of equivalent Rhodococcus mutants only accumulate the corresponding acids (Wilbrink et al., 2011; Yeh et al., 2014). The RG32 mutant of Rhodococcus rhodochrous DSM43269, a mutant completely devoid of Ksh by inacti-vation of five kshA genes, was able to produce very small amounts of ADD from ß-sitosterol (7% molar) but large amounts of 3-oxo-23,24-bisnorchola-1,4-dien-22-oic acid (1,4-BNC) (67% molar) (Wilbrink et al, 2011). A kshB- mutant of Rhodococcus equi USA-18 was also able to produce ADD and 1,4-BNC from sterols in similar molar ratios (Yeh et al., 2014).

In this regard, 4-HBC alcohol was detected many years ago together with AD during phytosterol transformations performed by the industrial strain of Mycobac-terium sp. NRRL B-3805 (Marsheck et al., 1972). This alcohol was postulated as a side reaction product but

not as a physiological intermediate of the major pathway leading to production of C-17-ketonic products (Marsheck et al., 1972). It has been proposed that 4-HBC and 1,4-HBC derive from 4-BNC and 1,4-BNC, respectively, by the action of a carboxyl-reductase by one or two consecutive enzymatic steps (Szentirmai, 1990; Xu et al., 2016). Nevertheless, these carboxyl-reductases have not been identified in mycobacteria yet (Xu et al., 2016). Assuming this hypothesis, these enzymes should not be active, or very low active, in Rhodococcus, since only the 4-BNC/1,4-BNC acids are detected in this organism (Yeh et al., 2014).

The accumulation of these C-22 intermediates suggests that the elimination of the last isopropyl group of the sterol side-chain seems to be highly dependent of the 9a-hydroxylation by Ksh. We assume that in the absence of 9a-hydroxylation, the enzymes responsible for removing the last propionic acid of the side-chain at C-17, i.e. the postulated acyl-CoA-dehydrogenase, enoyl-CoA-hydratase and aldol-lyase enzymes (Garcia et al., 2012), do not work efficiently. This inefficiency can be caused because they are feedback inhibited by the accumulation of AD and/or ADD, or because the real

Fig 3. Production of ADD by the MS6039 mutant in shake flasks. ADD is represented by triangles. (A) 9 mM glycerol + 0.4 g r1 of cholesterol (circles) used as substrates; (B) 9 mM glycerol + 0.4 g r1 of phytosterols (diamonds) used as substrates; (C) Analysis by CG-MS of the products after 96 h of incubation on phytosterols. (1) ADD, (2) cholestenone (internal standard), and (3) 1,4-HBC. (D) Chemical structure and fragmentation pattern of 1,4-HBC. The data reported are the averages of three different assays.

0.45 0.4 0.35 0.3 j 0.25 M 0.2 0.15 0.1 0.05 0

20 40 60 80 Time (h)

0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0

20 40 60 80 Time (h)

Fig 4. Production of AD by the MS6039-5941 mutant in shake flasks. AD is represented by squares. (A) 9 mM glycerol + 0.4 g r1 of cholesterol (circles) used as substrates; (B) 9 mM glycerol + 0.4 g r1 of phytosterols (diamonds) used as substrates; (C) Analysis by GC/MS of the products after 96 h of culture on phytosterols. (1) AD, (2) cholestenone (internal standard) and (3) 4-HBC. (D) Chemical structure and fragmentation pattern of 4-HBC. The data reported are the averages of three different assays.

substrates of these enzymes are the 9a-hydroxy derivatives of 1,4-BNC-CoA or 4-BNC-CoA. In this sense, it has been demonstrated that the best substrate of Ksh from M. tuberculosis is 4-BNC-CoA and 1,4-BNC-CoA (Capyk et al., 2011), suggesting that 9a-hydroxylation occurs before the release of the last propionic acid at C-17. If the previous 9a-hydroxylation of the steroid is critical to be recognized as a substrate for some of the enzymes involved in the release of the side-chain, this might explain why in Rhodococcus the absence of KshB impairs the complete degradation of the sterol side-chain (van der Geize et al., 2002a). Then, most probably the homologous enzymes of Mycobacterium are efficient enough on non-hydroxylated substrates and thus, the 9a-hydroxylation of sterols is not so critical to allow the complete side-chain degradation. In M. neoaurum ATCC 25795, the accumulation of 4-HBC and 1,4-HBC was only detected after the deletion of the hsd4A gene coding for a dual-functional enzyme with both 17p-hydro-xysteroid dehydrogenase and p-hydroxyacyl-CoA dehy-drogenase activities. However, these compounds are not the substrates of the Hsd4A in vitro (Xu et al., 2016).

This means that some partial or complete blockage of the side-chain degradation might cause the accumulation of 4-HBC or 1,4-HBC as by-products. This blockage can be caused not only by specific mutations, but also by retro-inhibitions of the enzymes due to the accumulation of certain intermediates when the cells are cultured with high concentrations of sterols.

During the course of this work, Xu et al. (2015) have described a double (DkshAI, DkstDI) mutant of Mycobac-terium sp. (apparently a derivative of M. neoaurum NwIB-02 (DkstDI) (Wei et al, 2010) that accumulates AD, ADD, 1,4-HBC and 4-HBC at different concentrations depending of fermentation temperature when cultured on phytosterols. This result confirmed that mycobacteria accumulate C-22 alcohols instead of the corresponding C-22 acids. Interestingly, they have also demonstrated that the residual KstD activity is not active on 9OH-AD (Xu et al., 2015) suggesting that a D1-dehydrogenation by KstD precedes Ksh hydroxylation. Remarkably, in contrast with our MS6039-5941 (AkshB1-AkstD1) double mutant, this M. neoaurum mutant still retains other Ksh and KstD active isoenzymes, because this mutant completely metabolized high amounts

of the sterol nucleus (39.8% by moles), which considerably reduces the final conversion yield of the process (Xu et al., 2015). This result strongly reinforces our proposal that M. smegmatis can be considered as a good cell factory to produce steroid synthons by metabolic engineering.

The finding that MS6039 and MS6039-5941 mutants can produce large amounts of ADD and AD from natural sterols, respectively, constitutes an experimental demonstration that metabolic engineering can be implemented in a model mycobacterial system like M. smegmatis mc2155 to generate pharmaceutical steroid synthons at industrial scale using a rational metabolic engineering approach. In fact, these engineered strains are already competing at industrial scale, i.e. under industrial operational conditions, with the existing strains isolated from environmental sources that were modified for their industrial use long time ago by conventional mutagenic procedures. Thus, our results lay the foundations for the valuation of M. smegmatis as a useful tool to develop a la carte engineered strains as cell factories to transform with high efficiency natural sterols into valuable pharmaceutical steroids.

Experimental procedures

Chemicals

Cholest-4-en-3-one was purchased from Fluka (Steinheim, Germany). ADD and AD were purchased from TCI

America. Chloroform and glycerol were purchased from Merck (Darmstardt, Germany). Cholesterol, N,O-bis(tri-methylsilyl) trifluoroacetamide (BSTFA), gentamicin, pyridine, Tween 80 and tyloxapol were from Sigma (Steinheim, Germany). Oligonucleotides were from Sigma-Genosys.

Bacterial strains and culture conditions

The strains as well as the plasmids used in this work are listed in Table 2. M. smegmatis mc2155 and its mutant strains were grown at 37°C in an orbital shaker at 200 r.p.m. Middlebrook 7H9 broth medium (Difco) containing 10% albumin-dextrose-catalase supplement (Becton Dickinson, New Jersey, USA) and 0.05% Tween 80 was used as rich medium. 7H10 agar (Difco) plates supplemented with albumin-dextrose-catalase were also used for solid media. Middlebrook 7H9 broth medium (Difco) without albumin-dextrose-catalase supplements containing 18 mM glycerol was used as a minimal medium. The minimal culture media used for the production of AD or ADD contained a mixture of 9 mM glycerol and 0.4 g Г1 cholesterol, or a mixture of 9 mM glycerol and 0.4 g Г1 of phytosterols. Commercial phytosterols (provided by Gadea Biopharma, Spain) contained a mixture of different sterols (w/w percentage): brassicasterol (2.16%), stigmasterol (8.7%), campesterol (36.8%) and ß-sitosterol (54.4%). Cholesterol and phytosterols were

Table 2. Bacterial strains and plasmids used in this study.

Strains or plasmids

Genotype and/or description

Source or reference

Strains Mycobacterium smegmatis mc2155 MS6039 MS6039-5941 mc2155 (pMV261) MS6039 (pMV261) MS6039 (pMV6039) Escherichia coli DH10B

Plasmids pJQ200x

pJQ6039

pJQ5941

pGEM®-T Easy pGEMT6039

pMV261 pMV6039

ept-1, mc26 mutant efficient for electroporation

M. smegmatis mc2155 AMSMEG_6039

M. smegmatis mc2155 AMSMEG_6039 AMSMEG_5941

mc2155 strain harbouring plasmid pMV261

MS6039 strain harbouring plasmid pMV261

MS6039 strain harbouring plasmid pMV6039

F-, mcrA, A (mrrhsdRMS-mcrBC), ®80d/acZAM15, DlacX74, deoR, recA1, araD139, A(ara-leu)7697, galU, galK, k-, rpsL, endA1, nupG

Suicide vector used to perform allelic exchange

mutagenesis in Mycobacterium, P15A ori,sacB, Gmr, xylE pJQ200x containing one fragment upstream and

another downstream of MSMEG_6039 gene pJQ200x containing one fragment upstream and

another downstream of MSMEG_5941 gene E. coli cloning vector; Ampr pGEMT-Easy harbouring the MSMEG_6039

gene encoding KshB1 from M. smegmatis mc2 155 Mycobacterium/E. coli shuttle vector with the kanamycin resistance aph and the promoter from the hsp60 gene from M. tuberculosis pMV261 harbouring the MSMEG_6039 gene encoding KshB1 from M. smegmatis mc2 155

Snapper et al. (1990)

This study

This study

This study

This study

This study

Invitrogen

Jackson et al. (2001)

This study

This study

Promega This study

Stover et al. (1991)

This study

dissolved in 10% tyloxapol prior to its addition to the minimal medium when assayed in flasks. Due to the low solubility of cholesterol and phytosterols, stock solutions were warmed at 80°C in agitation, sonicated in a bath for 1 h and then autoclaved. Gentamicin (5 ig/ml) was used for selection of M. smegmatis mutant strains when appropriate.

Bioreactor experiments were performed using the culture media previously described (Herrington and Spas-sov, 2003) containing a vegetal oil to dissolve phytosterols and corn steep liquor as supplementary carbon and nitrogen sources. The bioconversion experiments were performed in stainless steel (20-l) or glass (2- and 5-l) jar bioreactors with efficient stirring. The entire process was performed at 37°C.

Escherichia coli DH10B strain was used as a host for cloning. It was grown in rich LB medium at 37°C in an orbital shaker at 200 r.p.m. LB agar plates were used for solid media. Gentamicin (10 ig ml-1), ampicillin (100 ig ml-1) or kanamycin (50 ig ml-1), were used for plasmid selection and maintenance in this strain.

Gene deletions

The knock-out strains of M. smegmatis named MS6039 and MS6039-5941 were constructed by homologous recombination using the pJQ200x plasmid, a derivative of the suicide pJQ200 vector that does not replicate in Mycobacterium (Jackson et al., 2001). The strategy consist, for each gene, in generating two fragments of ~700 bp, the first one containing the upstream region and few nucleotides of the 5'end of the gene and the second one containing the downstream region and few nucleotides of the 3'end of the gene, that are amplified by PCR using the oligonucleotides described in Table 3 and M. smegmatis genomic DNA as template (Fig. S1). M. smegmatis genomic DNA extraction was performed as described (Uhia et al., 2011). The two fragments generated were digested with the corresponding enzymes and cloned into the plasmid pJQ200x using E. coli DH10B competent cells as described (Uhia et al., 2011).

Plasmid DNA from E. coli DH10B recombinant strains was extracted using the High Pure Plasmid Purification Kit (Roche, Basel, Switzerland), according to the manufacturer's instructions. This procedure was performed for genes MSMEG_6039 and MSMEG_5941, generating the pJQ6039 and pJQ5941 plasmids respectively. Plasmid pJQ6039 was electroporated into competent M. smeg-matis mc2155 to obtain strain MS6039. Plasmid pJQ5941 was electroporated into competent MS6039 cells to obtain MS6039-5941 strain. Single cross-overs were selected on 7H10 agar plates containing gentam-icin and the presence of the xylE gene encoded in pJQ200x was confirmed by spreading catechol over the single colonies of electroporated M. smegmatis. The appearance of a yellow coloration indicates the presence of the xylE gene. Colonies were also contra-selected in 10% sucrose. A single colony was grown in 10 ml of 7H9 medium with 5 ig ml-1 gentamicin up to an optical density of 0.8-0.9 and 20 il of a 1:100 dilution was plated onto 7H10 agar plates with 10% sucrose to select for double cross-overs. Potential double cross-overs (sucrose-resistant colonies) were screened for gentamicin sensitivity and the absence of the xylE gene. The mutant strains MS6039 and MS6039-5941 were analysed by PCR and DNA sequencing to confirm the deletions of MSMEG_6039 and MSMEG_5941 genes.

Construction of pMV6039 plasmid

To isolate the kshBI gene from M. smegmatis mc2 155 genomic DNA was extracted and amplified by PCR using the primers MSMEG_6039F (CGGAATTCTGACCTAAG GAGGTGAATGTGACTGATGAGCCCCTGGG) and MS MEG_6039R (CGAAGCTTCTATTCGTCGTAGGTGAC TTCG). The amplified fragment of 1092 bp was cloned into the commercial plasmid pGEM®-T Easy to generate pGEMT6039 plasmid that was transformed in E. coli DH10B competent cells. The cloned fragment was further digested with EcoRI and HindIII to clone into pMV261, a shuttle plasmid that replicates in E. coli and Mycobac-terium generating the plasmid pMV6039 that was

Table 3. Primers used in this study.

Primer Sequence (50-30)a Use

MSMEG_6039 up F ctagctcgagccagttgtgcacaccgatg MSMEG_6039 deletion (amplification of upstream region)

MSMEG_6039 up R ctagactagtgcagcgtcaggaactggc MSMEG_6039 deletion (amplification of upstream region)

MSMEG_6039 down F ctagactagtggtgcacatggagatcaacg MSMEG_6039 deletion (amplification of downstream region)

MSMEG_6039 down R ctagtctagagtaccagtcgatcggtgtc MSMEG_6039 deletion (amplification of downstream region)

MSMEG_5941 up F ctagactagtgatgttgcgaatgtcgatgtc MSMEG_5941 deletion(amplification of upstream region)

MSMEG_5941 up R ctagtctagaccaccacaacgtcgtactcc MSMEG_5941 deletion(amplification of upstream region)

MSMEG_5941 down F ctagtctagaccatgacattcggttacctgg MSMEG_5941 deletion(amplification of downstream region)

MSMEG_5941 down R ctaggagctcgcagggagatctcgaaatcg MSMEG_5941 deletion (amplification of downstream region)

a. Restriction sites are underlined.

transformed into E. coli DH10B to generate the recombinant strain E. coli DH10B (pMV6039). Once the sequence of the pMV6039 was checked it was used to transform electrocompetent cells of M. smegmatis MS6039 generating the M. smegmatis MS6039 (pMV6039) recombinant strain.

GC/MS analyses

To perform GC/MS analysis, culture aliquots (0.2 ml) were extracted twice at various extents of incubation with an equal volume of chloroform. Previously to its extraction, 10 il of a solution of 10 mM cholesterol (when phytosterols were used as substrate) or 10 mM cholestenone (when cholesterol was used as substrate) dissolved in chloroform were added to the aliquots as internal standards. The chloroform fraction was concentrated by evaporation and the trimethylsilyl ether derivatives were formed by reaction with 50 il of BSTFA and 50 il of pyridine and heating at 60°C for 45 min. Calibration standards were derivatized in the same way. The GC/MS analysis was carried out using an Agilent 7890A gas chromatograph coupled to an Agilent 5975C mass detector (Agilent Technologies, Palo Alto, CA, USA). Mass spectra were recorded in electron impact (EI) mode at 70 eV within the m/z range 50-550. The chromatograph was equipped with a 30 m x 0.25 mm i.d. capillary column (0.25_m film thickness) HP-5MS (5% diphenyl 95% dimethylpolysiloxane from Agilent Technologies). Working conditions in the sample were as follows: split ratio (20:1), injector temperature, 320°C; column temperature 240°C for 3 min, then heated to 320°C at 5°C min-1.For quantification of the peak area, the quantitative masses were 329 + 458 m/z for cholesterol; 343 + 384 m/z for cholestenone; 382 + 472 m/z for campesterol; 394 + 484 m/z for stigmasterol; 357 + 486 m/z for ß-sitosterol; 244 + 286 m/z for AD and 122 + 284 m/z for ADD in selected ions of monitoring. EI mass spectra and retention data were used to assess the identity of compounds by comparing them with those of standards in the NIST Mass Spectral Database and commercial standards (NIST 2011).

High-performance liquid chromatography analyses

To perform high-performance liquid chromatography (HPLC) analysis, samples (1 g) were withdrawn and extracted with 10 mL of ethyl acetate during two hours in a magnetic stirrer. One aliquot was centrifugated 10 min at 10 000 r.p.m. and supernatant was diluted 1:10 in acetonitrile. Samples were filtered (0.2 im pore size) prior to the chromatographic analysis. The HPLC analysis was carried out using a Waters liquid chromatograph with a PDA detector system. A Phenomenex

C18 column (Nucleosil C18, 100 A, 250 x 4.6 mm, 5 im particles) was employed and the temperature of the column was fixed in 50°C. Working conditions were as follows: the mobile phase was a mixture water: acetonitrile: acetic acid (48: 52: 0.1 v/v); flow rate was 1.1 mL min-1; the injection volumen was 10 il. Peaks were monitorized at 240 nm and calibrations were performed using highly purified standards of each compound.

Conflict of interest

None declared.

References

Andor, A., Jekkel, A., Hopwood, D.A., Jeanplong, F., Ilkoy, E., Konya, A., et al. (2006) Generation of useful insertion-ally blocked sterol degradation pathway mutants of fast-growing mycobacteria and cloning, characterization, and expression of the terminal oxygenase of the 3-ketosteroid 9a-hydroxylase in Mycobacterium smegmatis mc2155. Appl Environ Microbiol 72: 6554-6559.

Blank, L.M., Ebert, B.E., Buehler, K., and Bühler, B. (2010) Redox biocatalysis and metabolism: molecular mechanism and metabolic network analysis. Antioxid Redix Signal 13: 349-394.

Bragin, E.Y., Shtratnikov, V.Y., Dovbnya, D.V., Schelkunov, M.I., Pekov, Y.A., Malakho, S.G., et al. (2013) Comparative analysis of genes encoding key steroid core oxidation enzymes in fast-growing Mycobacterium spp. strains. J Steroid Biochem Mol Biol 138: 41-53.

Brzostek, A., Sliwinski, T., Rumijowska-Galewicz, A., Kor-ycka-Machata, M., and Dziadek, J. (2005) Identification and targeted disruption of the gene encoding the main 3-ketosteroid dehydrogenase in Mycobacterium smegma-tis. Microbiology 151: 2393-2402.

Brzostek, A., Pawelczyk, J., Rumijowska-Galewic, A., Dzi-adek, B., and Dziadek, J. (2009) Mycobacterium tuberculosis is able to accumulate and utilize cholesterol. J Bacteriol 191: 6584-6591.

Capyk, J.K., D'Angelo, I., Strynadka, N.C., and Eltis, L.D. (2009) Characterization of 3-ketosteroid 9a-hydroxylase, a Rieske oxygenase in the cholesterol degradation pathway of Mycobacterium tuberculosis. J BiolChem 284: 99379946.

Capyk, J.K., Casabon, I., Gruninger, R., Strynadka, N.C., and Eltis, L.D. (2011) Activity of 3-ketosteroid 9a-hydroxy-lase (KshAB) indicates cholesterol side chain and ring degradation occur simultaneously in Mycobacterium tuberculosis. J Biol Chem 286: 40717-40724.

Cole, S.T., Brosch, R., Parkhill, J., Garnier, T., Churcher, C., Harris, D., et al. (1998) Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393: 537-544. Erratum in: Nature 396: 190.

Donova, M.V. (2007) Transformation of steroids by Acti-nobacteria: a review. Appl Biochem Microbiol 43: 1-14.

Donova, M.V., and Egorova, O.V. (2012) Microbial steroid transformations: current state and prospects. Appl Microbiol Biotechnol 94: 1423-1447.

Donova, M.V., Gulevskaya, S.A., Dovbnya, D.V., and Pun-tus, I.F. (2005) Mycobacterium sp. mutant strain producing 9a-hydroxyandrostenedione from sitosterol. Appl Microbiol Biotechnol 67: 671-678.

Fernandez de las Heras, L., van der Geize, R., Drzyzga, O., Perera, J., and Navarro-Llorens, J.M. (2012) Molecular characterization of three 3-ketosteroid-D(1)-dehydrogenase isoenzymes of Rhodococcus ruber strain Chol-4. J Steroid Biochem Mol Biol 132: 271-281.

Galan, B., Diaz, E., Prieto, M.A., and Garcia, J.L. (2000) Functional analysis of the small component of the 4-hydroxyphenylacetate 3-monooxygenase of Escherichia coli W: a prototype of a new Flavin:NAD(P)H reductase subfamily. J Bacteriol 182: 627-636.

Garcia, J.L., Uhia, I., and Galan, B. (2012) Catabolism and biotechnological applications of cholesterol degrading bacteria. Microbial Biotechnol 5: 679-699.

van der Geize, R., Hessels, G.I., and Dijkhuizen, L. (2000) Targeted disruption of the kstD gene encoding a 3-ketosteroid M1-dehydrogenase isoenzyme of Rhodococ-cus erythropolis strain SQ1. Appl Environ Microbiol 66: 2029-2036.

van der Geize, R., Hessels, G. and Dijkhuizen, L. (2001a) Preparing genetically modified organism unable to degrade the steroid nucleus, useful for producing corticos-teroid intermediates, comprises inactivating the dehydro-genase genes. Patent WO 200131050.

van der Geize, R., Hessels, G.I., van Gerwen, R., van der Meijden, P., and Dijkhuizen, L. (2001b) Unmarked gene deletion mutagenesis of kstD, encoding 3-ketosteroid D1-dehydrogenase, in Rhodococcus erythropolis SQ1 using sacB as counter-selectable marker. FEMS Microbiol Lett 205: 197-202.

van der Geize, R., Hessels, G.I., van Gerwen, R., van der Meijden, P., and Dijkhuizen, L. (2002a) Molecular and functional characterization of kshA and kshB, encoding two components of 3-ketosteroid 9a-hydroxylase, a class IA monooxygenase, in Rhodococcus erythropolis strain SQ1. Mol Microbiol 45: 1007-1018.

van der Geize, R., Hessels, G.I., and Dijkhuizen, L. (2002b) Molecular and functional characterization of the kstD2 gene of Rhodococcus erythropolis SQ1 encoding a second 3-ketosteroid D1-dehydrogenase isoenzyme. Microbiology 148: 3285-3292.

van der Geize, R., Hessels, G.I., Nienhuis-Kuiper, M., and Dijkhuizen, L. (2008) Characterization of a second Rhodococcus erythropolis SQ1 3-ketosteroid 9a-hydroxy-lase activity comprising a terminal oxygenase homologue, KshA2, active with oxygenase-reductase component KshB. Appl Environ Microbiol 74: 7197-7203.

Herrington, E.J. and Spassov, G. (2003) Process for fermentation of phytosterols to androstadienedione. Patent WO 2003/064674.

Hu, Y., van der Geize, R., Besra, G.S., Gurcha, S.S., Liu, A., Rohde, M., et al. (2010) 3-ketosteroid 9a-hydroxylase is an essential factor in the pathogenesis of Mycobac-terium tuberculosis. Mol Microbiol 75: 107-121.

Jackson, M., Camacho, L.R., Gicquel, B., and Guilhot, C. (2001) Gene replacement and transposon delivery using the negative selection marker sacB. In Mycobacterium Tuberculosis Protocols, Vol. 54, Parish, T., and Stocker, N.G. (eds). Totowa, NJ: Humana Press, pp. 59-75.

Kendall, S.L., Withers, M., Soffair, C.N., Moreland, N.J., Gurcha, S., Sidders, B., et al. (2007) A highly conserved transcriptional repressor controls a large regulon involved in lipid degradation in Mycobacterium smegmatis and Mycobacterium tuberculosis. Mol Microbiol 65: 684-699.

Malaviya, A., and Gomes, J. (2008) Androstenedione production by biotransformation of phytosterols. Bioresour Technol 99: 6725-6737.

Marsheck, W., Kraychy, S., and Muir, R. (1972) Microbial degradation of sterols. Appl Microbiol 23: 72-77.

Penfield, J.S., Worrall, L.J., Strynadka, N.C., and Eltis, L.D. (2014) Substrate specificities and conformational flexibility of 3-ketosteroid 9a-hydroxylases. J Biol Chem 289: 25523-25536.

Petrusma, M., Dijkhuizen, L., and van der Geize, R. (2009) Rhodococcus rhodochrous DSM 43269 3-ketosteroid 9a-hydroxylase, a two-component iron-sulfur-containing monooxygenase with subtle steroid substrate specificity. Appl Environ Microbiol 75: 5300-5307.

Petrusma, M., Hessels, G., Dijkhuizen, L., and van der Geize, R. (2011) Multiplicity of 3-ketosteroid-9a-hydroxy-lase enzymes in Rhodococcus rhodochrous DSM 43269 for specific degradation of different classes of steroids. J Bacteriol 193: 3931-3940.

Petrusma, M., Dijkhuizen, L., and van der Geize, R. (2012) Structural features in the KshA terminal oxygenase protein that determine substrate preference of 3-ketosteroid 9a-hydroxylase enzymes. J Bacteriol 194: 115-121.

Petrusma, M., van der Geize, R., and Dijkhuizen, L. (2014) 3-Ketosteroid 9a-hydroxylase enzymes: Rieske non-heme monooxygenases essential for bacterial steroid degradation. Antonie Van Leeuwenhoek 106: 157-172.

Snapper, S.B., Melton, R.E., Mustafa, S., Kieser, T., and Jacobs, W.R. Jr (1990) Isolation and characterization of efficient plasmid transformation mutants of Mycobacterium smegmatis. Mol Microbiol 4: 1911-1919.

Stover C.K., de la Cruz V.F., Fuerst T.R., Burlein J.E., Benson L.A., Bennett L.T., et al. (1991) New use of BCG for recombinant vaccines. Nature 351: 456-460.

Szentirmai, A. (1990) Microbial physiology of sidechain degradation of sterols. J Ind Microbiol 6: 101-115.

Uhia, I., Galan, B., Morales, V., and Garcia, J.L. (2011) Initial step in the catabolism of cholesterol by Mycobacterium smegmatis mc2155. Environ Microbiol 13: 943-959.

Uhia, I., Galan, B., Kendall, S.L., Stoker, N.G., and Garcia, J.L. (2012) Cholesterol metabolism in Mycobacterium smegmatis. Environ Microbiol Rep 4: 168-182.

Wei, W., Wang, F.Q., Fan, S.Y., and Wei, D.Z. (2010) Inac-tivation and augmentation of the primary 3-ketosteroid-M1-dehydrogenase in Mycobacterium neoaurum NwIB-01: biotransformation of soybean phytosterols to 4-andros-tene- 3,17-dione or 1,4-androstadiene-3,17-dione. Appl Environ Microbiol 76: 4578-4582.

Wilbrink, M.H., Petrusma, M., Dijkhuizen, L., and van der Geize, R. (2011) FadD19 of Rhodococcus rhodochrous

DSM 43269, a steroid-Coenzyme A ligase essential for degradation of C-24 branched sterol side chains. Microbiology 158: 3054-3062.

Xu, X.W., Gao, X.Q., Feng, J.X., Wang, X.D., and Wei, D.Z. (2015) Influence of temperature on nucleus degradation of 4-androstene-3, 17-dione in phytosterol biotransformation by Mycobacterium sp. Lett Appl Microbiol 61: 63-68.

Xu, L.Q., Liu, Y.J., Yao, K., Liu, H.H., Tao, X.Y., Wang, F.Q., et al. (2016) Unraveling and engineering the production of 23,24-bisnorcholenic steroids in sterol metabolism. Sci Rep 6: 21928.

Yao, K., Xu, L.Q., Wang, F.Q., and Wei, D.Z. (2014) Characterization and engineering of 3-ketosteroid-M1-dehydro-genase and 3-ketosteroid-9a-hydroxylase in Mycobac terium neoaurum ATCC 25795 to produce 9a-hydroxy-4-androstene-3,17-dione through the catabolism of sterols. Metab Eng 24: 181-191.

Yeh, C.H., Kuo, Y.S., Chang, C.M., Liu, W.H., Sheu, M.L., and Meng, M. (2014) Deletion of the gene encoding the reductase component of 3-ketosteroid 9a-hydroxylase in Rhodococcus equi USA-18 disrupts sterol catabolism, leading to the accumulation of 3-oxo-23,24-bisnorchola-

1,4-dien-22-oic acid and 1,4-androstadiene-3,17-dione. Microb Cell Fact 13: 130.

Supporting information

Additional Supporting Information may be found online in the supporting information tab for this article:

Fig. S1. Construction of the mutant strains MS6039 and MS6039-5941. The polylinker restriction sites of the suicide plasmid pJQ200x are indicated (B, BamHI; Sp, SpeI; X, XbaI; N, NotI; Bs, BstXI; S, SacI).

Fig. S2. Production of ADD from phytosterols by the MS6039 mutant in 5-L jar bioreactor. Analysis by HPLC of the transformation products at 120 h of culture. (1) solvent front; (2) ADD; (3) AD; (4) 1,4-HBC.

Fig. S3. Production of AD from phytosterols by the MS6039-5941 mutant in 2-L jar bioreactor. Analysis by GC/ MS of the transformation products at 96 h of culture. (1) AD; (2) ADD; (3) 4- HBC; (4) 1,4-HBC; (5) campesterol; (6) stigmasterol; (7) ß-sitosterol.