Supporting Information€¦ · Bancerz et al. (2) described a psychro-trophic P. chryogenum isolate expressing a broad substrate range lipase depending on the nitrogen source in the

Supporting InformationMcLean et al. 10.1073/pnas.1419028112SI ResultsPreventing Statin Degradation in P. chrysogenum by Variation of NSources. Unexpectedly, a significant portion of the statins pro-duced is the deacylated form of compactin (ML-236A; Figs. 1Dand 2). To test whether this was due to a dysfunctional com-pactin pathway after its introduction into P. chrysogenum ora competing pathway, the β-lactam–free P. chrysogenum strainDS50652 was cultivated in the presence of compactin. Surprisingly,this leads to a rapid deacylation to ML-236A (Fig. S2C), suggestingthat the presence of an enzymatic activity in P. chrysogenum-degrading compactin, most likely an esterase or a lipase. Although,P. citrinum is also able to produce ML-236A (1), in our shake flaskexperiments, this molecule was barely seen. One of the main dif-ferences in the media used for both species is the nitrogen source:defined (ammonium sulfate) for P. chrysogenum and complex (yeastextract) for P. citrinum. Bancerz et al. (2) described a psychro-trophic P. chryogenum isolate expressing a broad substrate rangelipase depending on the nitrogen source in the medium. Onscreening media with various nitrogen sources for induction, itturned out that some nitrogen sources induced substantial de-acylation of compactin, whereas others lead to only minor de-acylation (Fig. 2B). Urea and ammonium sulfate turned out tobe the best inducers of the compactin deacylation activity: respec-tively, only 2% and 16% of the produced statin is compactin,whereas the remainder was deacylated to ML-236A. Yeast ex-tract, peptone, and lysine resulted in high levels of intact com-pactin (>80%) while producing only low amounts of ML-236A.Therefore, by applying yeast extract or lysine as nitrogen sourcesin shake flask experiments, the deacylation of statins could becontrolled to such an extent that less than 20% of the productwas lost due to degradation.

Identification of Statin Degrading Esterase from P. chrysogenum.P. chrysogenum samples, grown for 96 h with urea (inducinghigh statin degradation) as the sole nitrogen source, were frac-tionated, and the compactin deacylation activity was determined.The fraction with the highest ML-236A:compactin ratio showedtwo distinct bands on SDS/PAGE (Fig. S2B). Selected bandswere cut out of the gel and digested using trypsin. For analyzingpeptide mixtures, a so-called survey scan was used. This method,in which each scan consists of two segments, was defined asfollows: (i) full MS scan analysis, for selecting precursor ions forMS/MS based on defined criteria; and (ii) MS/MS of the selectedions to obtain amino acid sequence information. With this pro-cedure, one precursor ion at a time was selected for MS/MSwithin certain thresholds, which results in MS/MS datasets ofpeptides, all presenting the amino acid sequence of a part ofa protein. All MS/MS spectra were used for protein identifica-tion by database searching using the MASCOT search engine inthe MSDB database.The 40-kDa band was readily identified as Pc15g00720, an-

notated as homologous to 1,4-butanediol diacrylate esterase (48%sequence coverage) BDA1 from Brevibacterium linens (theoreticalmolecular weight of 43 kDa) (3). The 70-kDa band could notbe identified with LC-MS. Reanalysis with MALDI indicatedthat this was Pc22g01380, annotated as β-1,3-exoglucanase fromCochliobolus carbonum (theoretical molecular weight of 86 kDa).The exoglucanase is most likely not involved in compactin de-acylation, whereas the esterase is likely to be involved.To confirm that the esterase, encoded by the gene Pc15g00720,

is a 2-methylbutyric acid esterase, a KO mutant was constructedby inserting a selection marker gene between the promoter and

the terminator. This gene KO was done in a strain derived fromDS17690, wherein the NHEJ pathway was disrupted (4). Elevencorrect transformants were obtained and analyzed for compactindeacylation. After 3 d of cultivation, 0.1 g/L compactin was added,and samples were taken after 4 and 24 h. Although the remainingcompactin in the parent strain was almost zero (1.9%), the levelswere much higher in the mutant strains: 65.8–98.9% (Fig.S2C). These data confirm that the gene Pc15g00720 encodes a2-methylbutyric acid esterase and that by deleting this gene, thespecific esterase activity was decreased significantly.

SI Materials and MethodsUnless stated otherwise, chemicals were purchased from SigmaAldrich. Restriction enzymes and oligonucleotides were fromInvitrogen or New England Biolabs. Routine cloning procedureswere carried out as previously described (5). Synthetic DNA wasordered at GeneArt or DNA2.0.

Genomic DNA Isolation. Fungi were cultivated for 48 h at 25 °C.Cells were washed once in 0.85% NaCl and frozen in liquid ni-trogen. Cells were ground and resuspended in an equal volumeof phenol:chloroform:isoamyl alcohol (25:24:1). The mixture wasvortexed vigorously, and the aqueous phase was extracted twicewith fresh phenol:chloroform:isoamyl alcohol. Finally, chromo-somal DNA was precipitated with ethanol and air dried.

Reintroduction of the Penicillin Gene Cluster. Gene cluster engi-neering was done as described previously (6).

Analysis of Compactin Gene Cluster Transformants. The PCR-amplified fragments spanning the complete compactin genecluster were cotransformed to P. chrysogenum with a ble ex-pression cassette encoding for a protein that mediates phleo-mycin resistance. This cassette was isolated as a 1.4-kb SalIfragment from pAMPF7 (7). Transformants were selected onmineral medium supplemented with 50 μg/mL phleomycin and1 M saccharose. Phleomycin-resistant colonies were restreakedon fresh phleomycin plates without the saccharose and grownuntil sporulation. Colony PCR was used to determine thepresence of one or more compactin gene fragments. To thisend, a small piece of colony material was suspended in 50 μLTris-EDTA (TE) buffer (5) and incubated for 10 min at 95 °C.To pellet the cell debris, the mixture was centrifuged for 5 minat 1,000 × g. The supernatant (5 μL) was used as a template forPCR amplification (SUPER TAQ; HT Biotechnology Ltd).The PCR-amplified patterns were analyzed on E-gels (In-vitrogen) and used as an indication for integrated compactingenes. First, the presence of the 18-kb fragment was checked.Approximately one quarter of the colonies checked (112/480;∼23%) had the 18-kb fragment stably integrated. Subsequently,the presence of the other two fragments (14 and 6 kb) was veri-fied. Forty-five of the 18 kb-positive transformants also had bothother parts of the compactin gene cluster and thereby qualified asputative compactin production strains.

Compactin Deacylation Assay. P. chrysogenum was cultivated inmineral medium (25 mL) for 96 h at 25 °C. Statins were added inconcentrations ranging from 0.10 to 0.15 g/L. After 24 and 48 h,broth samples (0.25 mL) were mixed with an equal volume ofmethanol, followed by vigorous vortexing (at least 1 min persample) and analyzed. Care was taken that all material in thetube was brought into suspension, and no clumps remained during

McLean et al. www.pnas.org/cgi/content/short/1419028112 1 of 13

www.pnas.org/cgi/content/short/1419028112

extraction. The tubes were centrifuged at 16,000 × g for 5 min, andthe clear supernatant was transferred to HPLC sample tubes.

Identification of Statin Degrading Esterase from P. chrysogenum:Fractionation of Activity. Cell-free extracts (CFE) of P. chrysogenumsamples, grown for 96 h with either yeast extract (little statin deg-radation observed) or urea (inducing statin degradation) as solenitrogen source, were prepared by mortaring an ∼240-mg freezedried sample on ice and dissolving in 15 mL cold 50 mMTris·HCl buffer, pH 8. After dissolving, the samples werecentrifuged for 10 min at 20,000 × g and 4 °C. The supernatant(CFE) was stored at −20 °C in aliquots. Before use, a sample wasthawed on ice (generally, samples were always kept on ice unlessstated otherwise). The sample was precipitated by slowly adding106 mg ammonium sulfate up to 20% saturation (106 mg/mL) at0 °C. After a 30-min precipitation, the sample was centrifugedfor 10 min at 20,000 × g and 4 °C. The pellet was discarded, andthe supernatant was precipitated by slowly adding 175 mg am-monium sulfate up to 50% saturation (291 mg/mL) at 0 °C. Aftera 30-min precipitation, the sample was centrifuged for 10 min at20,000 × g and 4 °C. The supernatant was discarded, the pelletwas resuspended in 1 mL 20% saturated ammonium sulfate, andthe solution was purified by hydrophobic interaction chroma-tography (HIC). The HIC column (1 mL butyl FF HiTrap) wasequilibrated with 20% saturated ammonium sulfate on an Äktapurifier. After applying ∼0.8 mL of the precipitated sample (yeastextract grown, as well as the urea grown sample) on the column,a gradient of 20% saturated to 0% saturated ammonium sulfatewas initiated with a flow rate of 1 mL/min. Fractions of 1 mLwere collected.

CompactinAssay.Tenmicroliters (2mg/mL stock solution in ethanol)compactin was added to a 100-μL sample and incubated at 37 °Cfor 2 h. Fifty microliters of the incubated sample was extracted with50 μL methanol and centrifuged for 5 min at 18,000 × g. Fiftymicroliters of the supernatant was analyzed by HPLC.

SDS/PAGE.The samples were precipitated by adding 20% (vol/vol)trichloroacetic acid to a sample for a final concentration of10%, and the sample was kept on ice for 1 h. The samples werecentrifuged for 10 min at 18,000 × g and 4 °C. The pellet waswashed with 80% ethanol and again centrifuged for 10 min at18,000 × g and 4 °C. The supernatant was discarded, and the pelletwas dried in air for 5 min. Five microliters of sample buffer wasadded to the pellet and heated for 10 min at 70 °C. Two micro-liters of reducing agent and 13 μL milli-Q water were added to thesample to a total volume of 20 μL, and it was again heated for5 min at 70 °C. SDS/PAGE electrophoresis was performed usingMops buffer with a 4–12% Bis-Tris PAGE gel.

Deletion of the Compactin Esterase, pc15g00720. The promoter andthe terminator were PCR amplified (for oligonucleotides, seeTable S5) from P. chrysogenum genomic DNA (strain Wisconsin54-1255) using Phusion Hot-Start Polymerase (Finnzymes). Bothfragments were ∼1,800 bp long, containing a 14-bp tail suitable forthe STABY cloning method (Eurogentec), and were processed asdescribed previously (8). The resulting PCR fragments were mixedand used to transform a P. chrysogenum strain with the hdfA genedeleted (3). In this strain, the NHEJ pathway is disturbed, andtherefore, the random integration of DNA is drastically reduced.As the combined PCR fragments should recombine also to forma functional amdS selection marker gene (i.e., the so-called bi-partite or split-marker method), correctly targeted integrantsshould undergo a triple homologous recombination event.

Isolation of a Compactin Hydroxylase CYP105AS1 from A. orientalis:Discovery of Compactin Hydroxylating Bacteria.As a first approach,∼100 different bacteria, mainly Actinomycetes bacteria from

DSM’s internal strain collection, were tested for their ability tohydroxylate compactin. All bacteria were cultivated in shakeflasks using standard 2× yeast-tryptone (YT) medium and in-cubated with 0.1–0.5 g/L hydrolyzed compactin (compactin wasdissolved in ethanol to a final concentration of 20 mg/mL).NaOH was added from a 4 N stock to a final concentration of0.1 N. To open the lactone, the solution was heated at 50 °Cfor 1–2 h. Twenty-five milliliters of an overnight culture werecentrifuged (5 min, 3,000 × g). The pellet was washed and re-suspended in 10 mL 50 mM Tris·HCl, pH 7.5, followed by ad-dition of 50 μL toluene for 10 min at room temperature topermeabilize the cells. The volume was adjusted to 50 mL withTris·HCl, and cells were washed and resuspended in 10 mLTris·HCl. After overnight incubation with compactin, the su-pernatants were subsequently analyzed for formation of hydrox-ylated compactin using LC-MS. Initially, four species showedhydroxylation of compactin: S. carbophilus, A. orientalis, Actino-kineospora riparia, and Pseudonocardia alni. However, for thelatter two bacterial strains, this positive result proved difficult toreproduce. Conversion rates for S. carbophilus and A. orientalis were23–86% (at 0.5–0.1 mg/mL compactin) and 8–100%, respectively.

Generation of a Genomic Library of A. orientalis. A. orientalis strainNRRL18098 was grown overnight at 28 °C in a medium con-taining (g/L) glucose (10), yeast extract (5), starch (20), calciumcarbonate (1), and casamino acids (0.5), using baffled shakeflasks. Cells were harvested by centrifugation (15 min at 10,000 × g)and resuspended in 5 mL 50 mM Tris·HCl and 50 mM EDTA,pH 8. After addition of 100 μL lysozyme (100 mg/mL) and 40 μLproteinase K (20 mg/mL), the suspension was incubated for30 min at 37 °C. Subsequently, 6 mL Nuclei Lysis Solution(Promega) was added. Incubation for 15 min at 80 °C, followedby 30 min at 65 °C, led to almost complete lysis. FollowingRNase treatment (RNase A, 10 μL of 100 mg/mL stock), 2 mL ofProtein Precipitation Solution (Promega) was added, and themixture was vortexed for 20 s and incubated on ice for 15 min.After centrifugation (15 min at 3,000 × g), the supernatant wasmixed with 0.1 volume of sodium acetate (3 M, pH 5) and2 volumes of ethanol (96%). The precipitated DNA was trans-ferred with a Pasteur pipette and dissolved in 500 μL of 10 mMTris (pH 8). A second proteinase K treatment (10 μL of 20 mg/mLper 200-μL sample; incubation for 30 min at 37 °C) andphenol:chloroform extraction to remove remaining proteins wereperformed because the genomic DNA was not pure enough afterisolation (A260/A280 = 1.7). Five hundred microliters of phenol:chloroform:isoamyl alcohol (25:24:1) were added, vigorously vor-texed, and centrifuged for 5 min at 18,000 × g. The water phasewas transferred and extracted with 500 μL of chloroform:isoamylalcohol (24:1) to remove any traces of phenol. Next the waterphase was mixed with sodium acetate (0.1 volume) and ethanol(2 volumes) to precipitate the DNA. The DNA was washedwith 70% cold ethanol and dissolved in 500 μL of TE buffer. Intotal, 134 μg of purified genomic DNA with an A260/A280 of1.85 was obtained.To construct the genomic library, 50 μg of genomic DNA was

restricted with Sau3AI. Fragments of size between 3 and 4 kb wereisolated from an agarose gel and extracted using a QiaQuick ex-traction kit (Qiagen). The fragments were ligated in pZErO andtransformed to E. coli DH10B (Invitrogen). The average size ofthe DNA inserts of the library was 3.8 kb.

Screening the A. orientalis Genomic Library for Compactin Hydroxylation.The E. coli transformants with A. orientalisDNA were plated on LBmedium with kanamycin. Four thousand separate colonies weretransferred into 96-well microtiter plates (MTP) containing 0.2 mL2× YT medium with kanamycin, using a colony-picking automatedsystem. The MTPs were incubated for 48 h at 25 °C and 500 rpm.Five microliters of the culture were used to inoculate fresh medium



in deep well MTPs. The remainder of the master plates was mixedwith 10% glycerol and frozen at −80 °C. The freshly inoculateddeep well MTPs were incubated for 48 h at 25 °C and 500 rpm. Thecells were spun down and resuspended in 250 mL 2× YT mediumcontaining 200 mg/L pretreated (hydrolyzed and neutralized)compactin and 2 g/L glucose, and the pH was adjusted by 50 mMphosphate buffer to 6.8. The plates containing the cell-substrate mixwere incubated for 48 h at 30 °C and 280 rpm. After this conversionperiod, 0.35 mL methanol was added per well, and the productswere extracted for 1 h at 280 rpm. The cells were spun down, and100-μL samples of the supernatants were analyzed by LC-MS.

Identification of Compactin Hydroxylating Enzymes of A. orientalisNRRL18098. Clones showing compactin hydroxylation weretracked back on the initial master plates. Using the same growthand assay conditions, the compactin hydroxylation activity wasverified. Next, the positive clones were grown in 50-mL shakeflask cultures containing LB medium with kanamycin at 37 °C.Plasmid DNA was isolated using a Qiagen Midiprep PlasmidIsolation Kit (Qiagen). The obtained plasmids were sequencedat Baseclear. The sequences of the plasmid inserts were checkedfor ORFs using National Center for Biotechnology Information(NCBI) BLAST software.

Cloning of the A. orientalis CYP Gene in pBAD-DEST49. The CYP genewas PCR amplified from A. orientalis NRRL 18098 chromosomalDNA using the oligos attB1-ATGAGAGTAGACTCCGAAAATand attB2-CTATGCATCCCATGCAACG. The resulting 1.2-kbDNA band was gel extracted using a ZymoResearch Gel ExtractionKit (ZymoResearch). Subsequently, the purified DNA fragmentwas incubated with the pDONR221 vector (Invitrogen) usingstandard conditions as described in the corresponding Gatewayprotocols. E. coli TOP10 cells (Invitrogen) were used. The ENTRYvector was formed harboring the CYP105AS1 gene. A subsequentGateway reaction in the presence of the destination vector pBAD-DEST49 (method applied as stated in the Invitrogen Gatewaymanual) resulted in the expression vector pBAD-DEST49-CYP105AS1. Sequence errors were excluded by sequence analysis.

Generation, Expression, and Purification of CYP105AS1 (WT) and P450PravaMutants in E. coli. The heme domain (P450Prava CYP105AS1) fromthe fusion (P450Prava-RhF) construct and point mutants I95T,

Q127R, A180V, L236I, and A265N in different combinations werecreated using the QuikChange Lightning Multi Site-Directed Mu-tagenesis kit (Agilent). Proteins were produced in E. coli TOP10(Life Technologies) and purified by Ni-IDA affinity, ion exchange,and gel filtration chromatography as described previously (9).

Interaction of WT and Mutant CYP105AS1 Proteins with Compactin andPravastatin. Measurements of binding constants (Kd values) forWT and mutant CYP105AS1 enzymes (including P450Prava) withcompactin and pravastatin were performed by optical titrations,as described previously (9). Concentrated stocks of the ligands(∼30 mM) were prepared in DMSO. The Kd values were deter-mined as described previously using Origin software (OriginLab).

Cultivation of P. chrysogenum in Microtiter Plates and Shake Flasks.Fungal growth in MTP or shake flasks was performed in a syn-thetic medium containing (g/L) glucose (5); lactose (80); urea(4.5); (NH4)2SO4 (1.1); Na2SO4 (2.9); KH2PO4 (5.2); K2HPO4(4.8); and 10 mL/L of a trace element solution A containingcitric acid (150); FeSO4:7H2O (15); MgSO4:7H2O (150); H3BO3(0.0075); CuSO4:5H2O (0.24); CoSO4:7H2O (0.375); ZnSO4:7H2O(5); MnSO4:H2O (2.28); and CaCl2:2H2O (0.99). The pH beforesterilization was 6.5. The cultures were incubated at 25 °C in anorbital shaker at 280–400 rpm for 160 h.

Cultivation of P. chrysogenum in 10-L Fed-Batch Fermentation. Fed-batch cultivation was carried out as described previously (10). Thepreculture contained 500 mL medium [per kilogram: dextrose,10 g; Bacto yeast extract (Difco), 4 g; Bacto peptone, 4 g; MgSO4:7H2O, 0.5 g; KH2PO4, 2 g; K2HPO4, 4 g; pH 6.8) in a 2,000-mLshake flask, was inoculated with P. chrysogenum, and was cultivatedfor 75 h at 21 °C and 250 rpm.The main fermentation medium contained (g/kg) citrate, 15;

FeSO4, 1.5; MgSO4, 15; CuSO4, 0.024; ZnSO4, 0.15; MnSO4,0.23; H3BO3, 0.00075; CaCl2, 0.01; CoSO4, 0.037; Na2SO4, 5.4;KH2PO4, 7.7; and dextrose, 14.2; pH 4.5. CaCO3 (0.2 g) wasadded after sterilization of the medium. The full preculture wasused for inoculation. The feed started based on CO2 measurement.If all glucose and organic acids were consumed, the CO2 profilecollapses. Once the CO2 concentration dropped below 0.4%, thefeed (45% glucose and 1.3% ammonium sulfate) was started.

1. Serizawa N, et al. (1983) 6 Alpha-hydroxy-iso-ML-236B (6 alpha-hydroxy-iso-compac-tin) and ML-236A, microbial transformation products of ML-236B. J Antibiot (Tokyo)36(7):918–920.

2. Bancerz R, Ginalska G, Fiedurek J, Gromada A (2005) Cultivation conditions andproperties of extracellular crude lipase from the psychrotrophic fungus Penicilliumchrysogenum 9′. J Ind Microbiol Biotechnol 32(6):253–260.

3. van den Berg MA, et al. (2008) Genome sequencing and analysis of the filamentousfungus Penicillium chrysogenum. Nat Biotechnol 26(10):1161–1168.

4. Snoek ISI, et al. (2009) Construction of an hdfA Penicillium chrysogenum strain im-paired in non-homologous end-joining and analysis of its potential for functionalanalysis studies. Fungal Genet Biol 46(5):418–426.

5. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory Manual(Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY), 2nd Ed.

6. van den Berg MA, Westerlaken I, Leeflang C, Kerkman R, Bovenberg RAL (2007) Func-tional characterization of the penicillin biosynthetic gene cluster of Penicilliumchrysogenum Wisconsin54-1255. Fungal Genet Biol 44(9):830–844.

7. Fierro F, Kosalková K, Gutiérrez S, Martin JF (1996) Autonomously replicating plasmidscarrying the AMA1 region in Penicillium chrysogenum. Curr Genet 29(5):482–489.

8. Veiga T, et al. (2012) Impact of velvet complex on transcriptome and penicillin Gproduction in glucose-limited chemostat cultures of a β-lactam high-producing Pen-icillium chrysogenum strain. OMICS 16(6):320–333.

9. McLean KJ, et al. (2009) The structure of Mycobacterium tuberculosis CYP125: Mo-lecular basis for cholesterol binding in a P450 needed for host infection. J Biol Chem284(51):35524–35533.

10. Douma RD, et al. (2010) Intracellular metabolite determination in the presence ofextracellular abundance: Application to the penicillin biosynthesis pathway inPenicillium chrysogenum. Biotechnol Bioeng 107(1):105–115.



Fig. S1. Schematic overview of the cloning procedure for the compactin biosynthetic gene cluster.



Fig. S2. Deletion of Pc15g00720 encoding a compactin esterase. (A) Analysis of the CFE of P. chrysogenum to determine the esterase active fraction, afterincubation with compactin. (B) SDS/PAGE gel of the fractions 18–24 and 28. The two marked areas were cut from gel and identified with LC-MS. (C) Compactinhydrolysis in various P. chrysogenum Pc15g00720 deletion strains, after 24 h of incubation. The sum of compactin and ML-236A is set at 100%. 1–11, individualtransformants; Pc, P. chrysogenum DS17690.



Fig. S3. Sequence alignment of A. orientalis CYP105AS1 (WT) with other known CYP enzymes: Streptosorangium roseus (NCBI ref YP_006881683.1),Streptomyces venezuelae (NCBI ref YP_003338846.1), and S. carbophilus P450sca-2. The red box indicates the conserved alcohol/acid pair that likely stabilizesdioxygen and acts to protonate it during catalysis. The blue box highlights the cysteine thiolate and surrounding heme-binding motif.

Fig. S4. Unrefined density in the CYP105AS1 (WT) structure cocrystallized with compactin. Omit electron density present in the CYP105AS1 active site is shownand occurs when the CYP is cocrystallized or soaked in the presence of compactin. A compactin molecule (as present in the P450Prava mutant active site) isdepicted in gray for comparison.



Fig. S5. General spectroscopic features of CYP105AS1 (WT) and P450Prava (#10 mutant in Table S2). (A) UV-Vis spectra of the P450Prava mutant (4.2 μM)showing the oxidized (black line) spectrum with Soret maximum at 418 nm, the Fe(II) dithionite-reduced spectrum (blue) with Soret at 409 nm, and the Fe(II)COspectrum (cyan) with Soret at 448 nm. (Inset) The spectral difference between substrate-free, ferric forms of the WT (magenta) and P450Prava mutant (black)enzymes (both at 4.2 μM), with a broader spectrum and small HS shoulder at 395 nm for the P450Prava enzyme, whereas WT is predominantly LS. (B) Compactinbinding titration spectra of the P450Prava enzyme, showing the transition from the resting form (thick black line, Soret at 418 nm) to the fully HS sub-strate-bound species (red, Soret at 393 nm), with intermediate spectra in dashed black lines. (Inset) Difference spectra generated by subtraction of the sub-strate-free spectrum from those collected following each successive addition of compactin (colors as before). (C) Spectral binding curve for compactin with theP450Prava enzyme. Data (ΔA388 – ΔA421 vs. [compactin]) were fitted to the Morrison equation to yield a Kd = 1.38 ± 0.41 μM (cf. WT compactin Kd = 29.3 ±1.1 μM; Table S3).

Fig. S6. Indirect effects of the Q127R mutation. The hydrogen bonding network involving Q127R at the D-helix/G-helix interface is shown. The motion of Y166accompanying the open-to-closed conformational transition is likely affected by the Q127R mutation. Color coding is as in the main text with CYP105AS1 (inblue), the P450Prava (#10 mutant; Table S2) structure (in green), and the compactin shown in CPK (yellow carbons, C6 in cyan).



Fig. S7. Fed batch fermentation of P. chrysogenum strain PRA424. The picture shows the pravastatin titer development during fermentation and constantproductivity until 200 h after feed start.

Table S1. Penicillin production after reintroduction of thepenicillin biosynthetic gene cluster

Strain No. penicillin gene clusters PenicillinV (g/L)

DS17690 6–7 (1) 2.4DS50652 0 (2) 0AFF254 Transformant 1.99AFF257 Transformant 1.28AFF274 Transformant 0.96AFF295 Transformant 1.73

DS17690 (1), high producing penicillin strain; DS50652 (2), zero copy de-rivative of DS17690; AFFxxx, individual transformants obtained after trans-formation of DS50652 with the penicillin gene cluster.

1. van den Berg MA, et al. (2008) Genome sequencing and analysis of the filamentous fungus Penicillium chrysogenum. Nat Biotechnol 26(10):1161–1168.2. Harris DM, et al. (2009) Exploring and dissecting genome-wide gene expression responses of Penicillium chrysogenum to phenylacetic acid consumption and penicillinG production.

BMC Genomics 10:75.



Table

S2.

Statin

productionleve

lsofva

rioustran

sform

ants

Strain

Compactingen

esP4

50ML-23

6A(m

g/L)

Compactin

(mg/L)

6-ep

i-Prav

astatin

(mg/L)

Prav

astatin

(mg/L)

Totalstatin

(mg/L)

Rem

arks

P.citrinum

WT(N

RRL80

82)

—0

190

019

SeeFig.2A

P.citrinum

WT(N

RRL80

82)

P450

sca-2

04.5

00

4.5

P.ch

rysogen

um

mlcA-m

lcH,mlcR

—21

441

00

062

4Se

eFig.2A

;onretran

sform

ationofthisstrain

with

either

P450

sca-2orP4

50sca-2in

combination

withS.

carbophilu

sFe

rrodoxinsca-2an

dreductase,

no(6-epi)

prava

statin

was

obtained

P.ch

rysogen

um

mlcA-m

lcH,mlcR

CYP1

05AS1

75.9

3237

9.4

15.8

503.1

SeeFig.3B

P.ch

rysogen

um

mlcA,mlcC,mlcF-mlcH

—0

00

00

Tran

sform

ationwithoutmlcB,mlcD-m

lcE,

andmlcR

P.ch

rysogen

um

mlcA-m

lcH

—0

00

00

Tran

sform

ationwithoutmlcR

P.ch

rysogen

um

mlcA-m

lcD,mlcF-mlcH,mlcR

—0

00

00

Tran

sform

ationwithoutmlcE

P.ch

rysogen

um

mlcA-m

lcD,mlcF-mlcH

—0

00

00

Tran

sform

ationwithoutmlcEan

dmlcR

P.ch

rysogen

um

mlcA-m

lcH,mlcR

CYP1

05AS1

-CpO

69.8

70.4

540.3

13.9

694.4

SeeFig.3B

P.ch

rysogen

um

mlcA-m

lcD,mlcF-mlcH,

P pcb

C-m

lcE,

P pcb

C-m

lcR

P450

Prava

383

6026

260

729

mlcEan

dmlcRunder

controlofstrong

P.ch

rysogen

um

promoter

P.ch

rysogen

um

ΔPc15

g00

720

mlcA-m

lcD,mlcF-mlcH,

P pcb

C-m

lcE,

P pcb

C-m

lcR

P450

Prava

141

3837

672

888

mlcEan

dmlcRunder

controlofstrong

P.ch

rysogen

um

promoter



Table

S3.

EvolutionofCYP1

05AS1

(WTP4

50)into

P450

Prava

AoCYPmutant

Position

Prav

astatin

(%)

Relativeen

zyme

activity

WTCYP1

05AS1

31.0

Firstround:Error-pronePC

R#1

I233

T2

4.3

#2I95T

151.2

#3L2

36P

190.5

#4A18

0T22

1.1

#5A38

8T18

1.7

#6A18

0V48

1.0

#7L2

36I

A26

5V21

1.2

Seco

ndround:Site

saturationco

mbined

witherror-pronePC

R#8

F24L

A18

0LL2

36I

A26

5C94

1.1

#9P6

9SA18

0LL2

36I

A26

5V96

1.4

#10

I95T

Q12

7RA18

0VL2

36I

A26

5N86

0.3

#11

A18

0LE1

88K

L236

IA26

5V95

1.3

#12

H13

2RA18

0LL2

36I

A26

5VV29

5MF3

62L

961.6

#13

I95A

A18

0FI233

TT2

86A

910.6

#14

I95T

A18

0VL2

36I

A25

5VA26

5L93

0.6

#15

I95F

A18

0V91

0.6



Table S4. Properties of WT CYP105AS1 and P450Prava mutant proteins

AA mutation

P450Prava proteins

WT P1a P1b P2 P3 P4a P4b P450Prava

I95T + + + + + +L236I + + + + + +A180V + + + +Q127R + +A265N + +Soret peak (α, β) (nm) 419 (570,537) 419 (570,537) 419 (570,537) 418 (570,536) 415 (570,535) 418 (570,536) 418 (570,536) 418 (570,536)eSoret (mM−1·cm−1) 117 117 117 113 106 107 107 103Kd of compactin (μM) 29.3 ± 1.1 19.4 ± 1.8 16.1 ± 1.9 8.35 ± 0.5 2.03 ± 0.3 1.83 ± 0.4 1.96 ± 0.3 1.38 ± 0.4% HS + compactin 47.1 51.3 52.6 67.0 100 100 100 100Kd of pravastatin (μM) 54.9 ± 1.7 54.6 ± 1.9 54.1 ± 1.8 30.7 ± 1.7 11.2 ± 0.9 9.26 ± 1.4 9.87 ± 1.7 6.65 ± 0.6% HS + pravastatin 41.9 38.8 40.7 62.1 100 100 100 100

The AA modifications present in each mutant are indicated. A summary of mutant substrate/product binding and spectroscopic features is given.



Table S5. Oligonucleotides used

Target DNA Gene(s) Forward primer Reverse primer

Amplification of the compactincluster18-kb fragment (left 10 kb) mLcH GGGGACAACTTTGTATAGAAAAGTTGAAGGATGAC-

TATTCCAGTGATTAGCAC

GAGAAGACGAAACTCGTGCTTTGAGTG

mLcGmLcF5′mLcA

18-kb fragment (right 8 kb) 3′mLcA CACTCAAAGCACGAGTTTCGTCTTCTC GGGGACTGCTTTTTTGTACAAACTTGAAGGGAG-

TACTTGTGTCCACGTCGTTGmLcC18-kb fragment (internal

6 kb)‘mLcA’ GTGGTAGGCGGCCAGGTAGAAC CAGCATCTTCGTGGAGGTGCGC

14-kb fragment mLcB GGGGACAAGTTTGTACAAAAAAGCAGGCTAACCCG-

CCTTCCGACTACATATCCACAATC

GGGGACCACTTTGTACAAGAAAGCTGGGTACTC-

AGGAATGAATCAGATCAACATTCmLcD6-kb fragment mLcE GGGGACAGCTTTCTTGTACAAAGTGGAAGTATCAGG-

ATTGATGCCTGAAACATC

GGGGACAACTTTGTATAATAAAGTTGAGATCT-

GCTGGTAGACTAGAGCCTGCCmLcRDeletion of the compactin

esterase, pc15g007201.8-kb promoter CCTTCGCCGACTGATCCAAAAGTGAGACCAATGC AATGATATTGAGAGACAGGG

1.8-kb terminator TATAATTGGCGCTGAGTATG CCTTCGCCGACTGAACCAATATCTTCTGGGTCTC

PCR verification of theintegrated compactin genes18-kb fragment CACAGGAATCACAGCAGAACAGTCATC TCCCATTTGCTGTTGATGGAGCAGC

14-kb fragment GATCTGAGATGTCACATGCGTGTAGATAGAC CAATTGATCTTCTCTCGTGGCAAAGAG

6-kb fragment TGGTTGCGAAGGCTGCAAAGAC TGTACACGCTGACCTCGCATATGAAG

niaA CACAGAGAATGTGCCGTTTCTTTGG TCACATATCCCCTACTCCCGAGCC

PCR amplification ofA. orientalis P450 genesCYP105AS1 cmpH GGCTAGGAGGAATTAACCCATATGAGAGTAGACTCC CTTTAAAGCTGGGACTAGTCGCATCCCATGCAAC

Error prone PCR amplificationof A. orientalis P450 gene,first roundCYP105AS1 cmpH GGGGACAAGTTTGTACAAAAAAGCAGGCTATGAG-

AGTAGACTCCGAAAAT

GGGGACAGCTTTCTTGTACAAAGTGGCTATGC-

ATCCCATGCAACG

Site-saturation PCRamplification of A. orientalisP450 mutants, second roundPosition 1-left GGGGACAAGTTTGTACAAAAAAGCAGGCTAGGAGGA-

ATTAACCATG

GGGTGCGTCCATGTTNNNGAACCAGCCGGGCGC

Position 1-right GCGCCCGGCTGGTTCNNNAACATGGACGCACCC GGGGACCACTTTGTACAAGAAAGCTGGGTCTA

Position 2-left GGGGACAAGTTTGTACAAAAAAGCAGGCTAGGAGGA-

ATTAACCATG

TGTCTTGTCCACACTNNNCATGATCGTGGTCTG

Position 2-right CAGACCACGATCATGNNNAGTGTGGACAAGACA GGGGACCACTTTGTACAAGAAAGCTGGGTCTA

Position 3-left GGGGACAAGTTTGTACAAAAAAGCAGGCTAGGAGG-

AATTAACCATG

GAGCAAAAGCAACGCNNNGTTCGTCAGCTCTTC

Position 3-right GAAGAGCTGACGAACNNNGCGTTGCTTTTGCTC GGGGACCACTTTGTACAAGAAAGCTGGGTCTA

Position-4-left GGGGACAAGTTTGTACAAAAAAGCAGGCTAGGAGG-

AATTAACCATG

CCCGGCGACGAGCAANNNCAACGCGATGTTCGT

Position-4-right ACGAACATCGCGTTGNNNTTGCTCGTCGCCGGG GGGGACCACTTTGTACAAGAAAGCTGGGTCTA

Position-5-left GGGGACAAGTTTGTACAAAAAAGCAGGCTAGGAGGA-

ATTAACCATG

GTCCGGGCTGTCCAGNNNAGCGATCTGCTCCGG

Position-5-right CCGGAGCAGATCGCTNNNCTGGACAGCCCGGAC GGGGACCACTTTGTACAAGAAAGCTGGGTCTA

Position-6-left GGGGACAAGTTTGTACAAAAAAGCAGGCTAGGAGG-

AATTAACCATG

CCAAACCCCGTACGCNNNGGACTTCTCGCGCAG

Position-6-right CTGCGCGAGAAGTCCNNNGCGTACGGGGTTTGG GGTCAGGTGGGACCACCGCGCTACTGCCGCCAGG



Table S6. Crystallographic data and model refinement parameters

CYP105AS1 PDB 4OQS P450Prava-compactin PDB 4OQR

Data collectionSpace group H3 P212121

Cell dimensionsa, b, c (Å) 133.9, 133.9, 73.0 39.4, 86.8, 139.6

Resolution (Å) 29.5–2.04 46.6–1.8 (1.85–1.80)Rmeas 8.1 (43.2) 5.9 (32.5)I/σI 12.3 (1.91) 14.2 (2.71)Completeness (%) 100.0 (100.0) 99.84 (99.94)Redundancy 7.26 5.2RefinementResolution (Å) 30–2.04 30–1.80No. reflections 29,496 42,324

Rwork/Rfree 20.0/24.8 17.5/21.8B-factors (Å2) 26.6 16.4RMS deviationsBond lengths (Å) 0.024 0.025Bond angles (°) 2.25 1.97

Data are shown for the substrate-free CYP105AS1 (WT) structure and for the P450Prava (#10 mutant inTable S3) in complex with compactin.



Supporting Information€¦ · Bancerz et al. (2) described a psychro-trophic P. chryogenum isolate expressing a broad substrate range lipase depending on the nitrogen source in the

Documents