2.1. Strains, plasmids, media and culture conditions

For the analysis of carotenoids, R. marinus strains TK-1, TK-2, TK-3 and SB-71 were grown on agar plates with modified medium 162 supplemented with 1% NaCl and 0.025 ​g/L adenine. After 48 ​h ​at 65 ​°C, the colonies were transferred to 5 ​mL liquid LB-medium, supplemented with 0.025% adenine, and cultivated at 65 ​°C for 24 ​h in 50 ​mL falcon tubes. The cells were then cultivated in baffled shake flasks at 200 ​rpm increasing the cultivation volume from 50 ​mL to 100 ​mL sequential cultivations, maintaining 10% (v/v) inoculation volume. The harvested cells were then split and transferred to two 500 ​mL bioreactors(Multifors 2, Infors) for cultivation with the following parameters: 65 ​°C, pH 7, 1 VVM aeration, 200 ​rpm stirrer rate cascaded with dissolved oxygen tension (DOT) that was set to 40%. The cultivation was terminated at the onset of the stationary phase, indicated by a sudden increase of DOT, at which the cell culture was set to cool at 10 ​°C. The cells were harvested by centrifugation at 4 ​°C at 10,000 ​g for 10 ​min, after which the pellets were frozen at -80 ​°C and lyophilized until constant weights were reached.

2.2. Design and generation of cloning molecules

Recombinant DNA molecules were designed using the plasmid vector pUC19 and amplified regions of the genome sequences of R. marinus ISCaR-493 (Matís, unpublished) and T. thermophilus HB8 (NC_006461.1). DNA was isolated from the strains using the MasterPure Complete DNA purification Kit (Lucigen). Primers (Table S1) were designed for the amplification of ~1500 bp 5′ and 3′ flanking regions of the genes targeted for deletion as well as the trpB gene (selection marker) and the crtB gene from T. thermophilus. The primers were designed to support HiFi DNA assembly (NEBuilder HiFi DNA Assembly Master Mix, New England BioLabs). The primers Gene5_5_F, Gene5_5_R (5′ flanking region), Gene1_3_F, Gene1_3_R (3′ flanking region), Gene5_Gene14_F and Gene5_Gene14_R (trpB gene) were used to amplify the fragments for the gene deletion cassette used to obtain strain TK-1 (ΔtrpBΔpurAΔcrtYO::trpB). The primers Gene9a_5_F, Gene9a_5_R (5′ flanking region), Gene9a_3_F, Gene9a_3_R (3′ flanking region), Gene9a_Gene14_F and Gene9a_Gene14_R (trpB gene) were used to amplify the fragments for the gene deletion cassette used to obtain strain TK-2 (ΔtrpBΔpurAΔcruF::trpB). The primers Gene9_5_F, Gene9_5_R (5′ flanking region), Gene9_3_F, Gene9_3_R (3’ flanking region), Gene9_Gene13_F, Gene9_Gene13_R (crtB gene from T. thermophilus), Gene9_Gene14_F and Gene9_Gene14_R (trpB gene) were used to amplify the fragments for the gene deletion/insertion cassette used to obtain strain TK-3 (ΔtrpBΔpurAΔcruFcrtB::trpBcrtBT.thermophilus). All following enzymes, kits and cells in this section were obtained from New England BioLabs. Q5 High-Fidelity DNA polymerase was used in all amplifications according to the manufacturer’s instructions. The correct sizes of the resulting fragments were verified by electrophoresis and they were subsequently purified from gel using the Monarch DNA Gel Extraction Kit. The fragments were assembled into SmaI restricted pUC19 vector and introduced into competent E. coli cells (NEB stable) by chemical transformation. Positive clones were confirmed by verifying the presence of the trpB gene by PCR, using Taq DNA polymerase and the primers Gene14_Verify_F and Gene14_Verify_R (Table S1). Vectors were isolated from positive clones using the Monarch Plasmid Miniprep Kit. Linear inserts for R. marinus transformation for the construction of strains TK-1 and TK-2 were obtained by PCR, using the Q5 High-Fidelity DNA polymerase and the primers Gene5_5_F and Gene1_3_R, and TK2_Total_F and TK2_Total_R (Table S1), respectively. A linear insert for R. marinus transformation for the construction of strain TK-3 was obtained by digesting the plasmid with XbaI and KpnI-HF.

2.3. Transformation of R. marinus

R. marinus cells were prepared for electroporation as described elsewhere (Bjornsdottir et al., 2005). The electroporation protocol was followed, using the GenePulser Xcell electroporation system (Bio-Rad) with pulses delivered at 20 ​kV/cm 1 ​μg of DNA was used per transformation, in ≤5 ​μL and mixed carefully with 40 ​μL of washed cells. Negative (sterile MilliQ water) and positive (pRM3000) controls were included. Transformed cells were grown on selective agar plates (without tryptophan) for 3–5 days.

2.4. Verification of genotypes using PCR and sequencing

Positive R. marinus clones were verified by PCR, using Taq DNA polymerase. Modifications of strain TK-1 were verified using primers Gene7_Verify_F, Gene7_Verify_R, Gene5_Verify_F, Gene5_Verify_R, Gene4_Verify_F, Gene4_Verify_R, Gene14_Verify_F and Gene14_Verify_R (Table S1). Modifications in strains TK-2 and TK-3 were verified using the primers Gene9a_Verify_F, Gene9a_Verify_R, Gene9b_Verify_F, Gene9b_Verify_R, Gene13_Verify_F, Gene13_Verify_R, Gene14_Verify_F and Gene14_Verify_R (Table S1). Additionally, the modifications were confirmed by sequencing. Several PCR reactions were performed for strains SB-62 and TK-1 using the primers Gene1_Seq_F, Gene1_Seq_R, Gene4_Seq_F, Gene4_Seq_R, Gene5_Seq_F and Gene5_Seq_R (Table S1) and Q5 DNA polymerase. The resulting ampliconswere sequenced using the ABI3730 system (Applied Biosystems, Thermo Fisher Scientific). For strain TK-2, Q5 DNA polymerase and primers Gene9a_Seq_F and Gene9a_Seq_R (Table S1) were used to amplify a 5188 bp region. For strain TK-3, Q5 DNA polymerase and primers Gene9_Seq_F and Gene9_Seq_R (Table S1) were used to amplify a 4686 bp region. Sequencing libraries were made from the amplicons using the Nextera DNA (TK-2) and Nextera Flex (TK-3) methods (Illumina) and libraries were sequenced on the Illumina MiSeq sequencing platform using the V3 2 ​× ​300 cycle chemistry. Obtained fragments were assembled using Geneious.

2.5. Carotenoid extraction

Carotenoids were extracted from R. marinus strains TK-1, TK-2, TK-3, SB-71 and ISCaR-493 using two methods. In the first method, aqueous cell suspensions were sonicated in an ice bath for 5 ​× ​2 ​min, with 1-min rests in-between in order to keep cold sample conditions. After sonication the samples were mixed with ethyl acetate (1:1). In the second method, lyophilized cells were powdered with a glass rod before mixing with dichloromethane (25 ​mL solvent per g freeze-dried cells). All organic phase extracts were vacuum filtered before being dried by rotary evaporation (Heidolph instruments) at 80 ​rpm ​at 40 ​°C. The extracts were reconstituted in 3 ​mL of dichloromethane, filtered through a 0.2 ​μm PTFE syringe filter, flushed with N2 gas and stored at -80 ​°C until mass spectrometry analysis.

Carotenoids were additionally extracted from strains TK-1, TK-2, TK-3, SB-71 and ISCaR-493, for absorbance spectra analysis, by mixing full loops of cells from agar plates with hexane:acetone (1:1) and sonicating in a bath for 20 ​min. This was done in triplicates.

2.6. Carotenoid analysis

The carotenoid extracts were analyzed by means of supercritical fluid chromatography – mass spectrometry (MS) in an Ultra Performance Convergence Chromatography system (UPC2) coupled to a quadrupole – orthogonal acceleration time-of-flight tandem mass spectrometer (XEVO-G2 Q-TOF) with an electrospray ion source, both instruments from Waters (Mildford, MA, USA). Both systems were controlled, and all data was analyzed, with MassLynx™ (v 4.1, SCN 77; Waters). The chromatographic method used was based on a modification of the method described in (Jumaah et al., 2016). Briefly, the chromatographic separation was achieved using Acquity UPC2 Torus 1-Aminoanthracene (100 ​mm ​× ​3 ​mm, 1.7 ​μm) column from Waters. The mobile phase consisted of (A) CO2 and (B) methanol in a gradient elution analysis programmed as follows: 0–0.5 ​min, 5% (B); 0.5–2.5 ​min, 5–15% (B); 2.5–8 ​min, 15% (B); 8–9 ​min, 15-5% (B); and 9–10 ​min, 5% (B). The flow rate was 1.5 ​mL ​min-1, column temperature was kept at 56 ​°C and backpressure at 160 ​bar. Ammonium formate (10 ​mM) in methanol was used as a make-up solvent at a flowrate of 0.5 ​mL ​min-1. The total run-time of the program was 10 ​min.

UV-Vis spectra were recorded in the range of 200–500 ​nm by the diode array detector (DAD). The electrospray ionization (ESI) ion source was operated in positive mode and full-scan MS spectra were obtained by scanning the range m/z 50–1000. The mass spectrometer was calibrated with sodium formate. Centroid mode data was collected after mass correction during acquisition using an external reference comprising of 10 ​μL/min solution of leucine-enkephalin (2 ​ng/μL). The capillary and cone voltage were set at 3 ​kV and 40 ​V, respectively. Nitrogen was used as both cone gas (50 ​L/h) and desolvation gas (800 ​L/h). The source and desolvation temperature were set at 150 and 300 ​°C, respectively. Simultaneous acquisition of exact mass at high and low collision energy, MSE (where E represents collision energy), was used to obtain full scan accurate mass fragment, precursor ion, and neutral loss information. The collision energy in function 1 (low energy) was off while in function 2 (high energy) and the collision energy ranged between 15 and 60 ​V. MS/MS analysis of the carotenoids was performed with the quadropole set at m/z 536 in MSEmode.

The absorbance spectra (350–600 ​nm) of the carotenoids of strains TK-1, TK-2, TK-3, SB-71 and ISCaR-493, extracted with hexane:acetone (1:1), were analyzed in a benchtop spectrophotometer (MULTISKAN Sky, Thermo Scientific), using 1 ​cm cuvette. This was done to investigate the profile of the spectra and Amaxvalues of obtained peaks.

Lyophilized TK-3 ​cells from shake flask cultivations were weighed before organic solvent extraction and reconstitution in chloroform. The extracts were then analyzed by a spectrophotometer at 485 ​nm. The lycopene concentration was calculated using Beer’s law with the molar attenuation coefficient of 150855 ​L ​mol-1 cm-1 for lycopene in chloroform (Naviglio et al., 2008). Lycopene, α-, β- and γ-carotene standards were of analytical grade quality (Supelco).

3.1. Bioinformatic analysis of genes encoding carotenoid biosynthetic enzymes

Several strains of the thermophilic bacterium R. marinus produce carotenoids derived from γ-carotenoid. Previous gene homology searches for known carotenoid biosynthetic genes using the genomic data of the type strain R. marinus DSM 4252T, resulted in the discovery of genes encoding homologues to enzymes in the carotenoid biosynthesis pathways from other species. This allowed the identification of a gene cluster, with two adjacent operons, one small and one larger (Bjornsdottir et al., 2011; Ron et al., 2018). In this work, the bioinformatic analysis of genomic sequence data was expanded, enabling further identification of the enzymes in the biosynthetic pathway for carotenoid production in R. marinus (Fig. 1).

1. Download : Download high-res image (384KB)
2. Download : Download full-size image

Fig. 1. Proposed carotenoid biosynthetic pathway in R. marinus, based on (Ron et al., 2018), with added information obtained in this study. The molecular structures and abbreviated enzyme names are shown, phytoene synthase (CrtB), 1’,2’-hydratase (CruF), phytoene desaturase (CrtI), glycosyltransferase(CruC), acyltransferase (CruD), lycopene cyclase (CrtY) and carotene ketolase (CrtO).

Here, we report the identification of the cruF gene in the gene cluster (Fig. 2), which likely acts as a 1′,2′-hydratase (Sun et al., 2009). This gene was not identified in the previous work because it is fused together with the phytoene synthase (crtB) gene (Fig. 2A, gene 9) and together they were annotated as a phytoene synthase. These genes are usually separate, while cruF in R. marinus is without a stop codon and fused with the downstream crtB in one open reading frame (ORF). The fused genes, along with the carotenoid genes acyltransferase(cruD) and phytoene desaturase (crtI) (Fig. 2A, genes 9, 11 and 12, respectively) are located in the smaller operon in the gene cluster. Two additional genes are also present in the smaller operon, encoding products showing homology to hypothetical proteins (gene 10) and a MerR family transcriptional regulator from T. thermophilus (31% similarity) (gene 8).

1. Download : Download high-res image (530KB)
2. Download : Download full-size image

Fig. 2. The carotenoid gene cluster in the R. marinus stain SB-62 (ΔtrpBΔpurA) (A) and three different genetically modified strains: mutant TK-1 (ΔtrpBΔpurAΔcrtYO::trpB) (B), mutant TK-2 (ΔtrpBΔpurAΔcruF::trpB) (C) and mutant TK-3 (ΔtrpBΔpurAΔcruFcrtB::trpBcrtBT.thermophilus) (D). Genes involved in the carotenoid biosynthesis are colored grey. Gene deletions and insertions were verified using PCR (1–14) and the sequencing of PCR products (a–g). Pictures showing gel electrophoresis following PCR are shown to the right side of the gene clusters. Lane M1: 100 bp DNA ladder. Lane M2: 1 ​kb DNA ladder. Unedited electrophoresis pictures and sequencing results are shown in the supplementary file (Figs. S1–S11).

Two genes in the larger operon were also identified based on sequence homology, with high sequence similarity to the known carotenoid genes encoding lycopene cyclase (crtY) (50%) (Fig. 2A, gene 2) and carotene ketolase (crtO) (60%) (Fig. 2A, gene 3). Five other genes (Fig. 2A, genes 1, 4–7) were found in this larger operon, but since they lack sequence homology to known genes in carotenoid biosynthetic pathways, their role in the pathway remains unknown. They were annotated as hypothetical protein (gene 1), NAD dependent epimerase (gene 4), FAD dependent oxidoreductase (gene 5), which showed slight homology (36% similarity) to crtI, deoxyribodipyrimidine photo-lyase (gene 6) and short-chain dehydrogenase/reductase (gene 7). Several known carotenoid biosynthetic genes belong to the crtI gene family, including crtO(gene 3, above) and also 3′,4′-desaturase, an enzyme for which the corresponding gene is not identified in R. marinus. The native structures of the carotenoids in R. marinus suggest that a 3′,4′-desaturase is a part of the biosynthetic pathway, but since the crtI homology of gene 5 is low and there is no other evidence so far that it encodes a 3′,4′-desaturase, this step in the pathway is still labeled as unknown (Fig. 1).

3.2. Genetic modifications to alter the carotenoid production

Three different genetic modifications were performed in the carotenoid gene cluster of R. marinus strain SB-62 (ΔtrpBΔpurA) (Fig. 2). The target genes were deleted and replaced with the selective marker trpB (encoding the tryptophan synthase beta chain), using double crossover homologous recombination using linear insertion cassettes. In the first modification, a 5890 bp region was deleted from the larger operon, resulting in the R. marinus mutant TK-1 (ΔtrpBΔpurAΔcrtYO::trpB) (Fig. 2B). This region includes the genes crtY (gene 2) and crtO (gene 3) which encode the enzymes responsible for the 4-keto β-ionone ring modifications displayed on the left side of the carotene backbone (Fig. 1), as well as gene 5, where the deduced amino acid sequence is showing slight homology to crtI (see section 3.1). The other two modifications involved knocking out the cruF gene (1′,2′-hydratase), which encodes one of the enzymes that modify the right side of the carotene backbone (Fig. 1). In the first cruF modification, only the cruF part of the gene was deleted, leaving the crtBpart intact, resulting in the R. marinus mutant TK-2 (ΔtrpBΔpurAΔcruF::trpB) (Fig. 2C). The crtB part of the gene includes a start codon, which raises the question if it could encode a functional enzyme without the cruF part. In the second cruF modification, both the cruF and crtB parts of the gene were deleted and the crtB gene from Thermus thermophilus strain HB8 was inserted, resulting in the R. marinus mutant TK-3 (ΔtrpBΔpurAΔcruFcrtB::trpBcrtBT.thermophilus) (Fig. 2D). This modification was designed in case the native CrtB enzyme in TK-2 would be non- or low functioning after the deletion of the cruF part.

Successful genetic modifications were verified by amplifying target regions from the genomes of putative mutants, and sequencing. PCR reactions amplifying several genes in the cluster (genes 1, 4, 5, 7, 9a (cruF), 9b (crtB), 13 and 14 in Fig. 2), using primers designed to bind inside each gene, were performed for all three mutants and strain SB-62. For strain SB-62 all reactions resulted in the expected sized amplicons, except for the crtB gene from T. thermophilus and the trpB gene (Fig. 2A). The presence of an amplicon using primers for amplifying the crtB gene from T. thermophilus in R. marinus was not expected. However, in all R. marinus strains the same sized unknown amplicon (~300 bp) was obtained and was subsequently sequenced. This showed that the amplification occurred from a location on the chromosome distant from the carotenoid gene cluster (Fig. S9). This gene did not show homology to crtB from T. thermophilus. The primers used to amplify the crtB gene from T. thermophilusdid however show homology to this region, which explains the unexpected amplification. The trpB gene was not amplified in strain SB-62 which was to be expected since it has the trpB gene deleted, enabling us to use it as a selective marker.

For strain TK-1 (ΔtrpBΔpurAΔcrtYO::trpB), the PCR results showed two main differences compared to strain SB-62 (Fig. 2B). First, genes 1, 4 and 5 were not amplified, which indicated the successful deletion of the region. Second, the reaction for the trpB gene (gene 14) did result in the expected sized amplicon, indicating the insertion of the selection marker. To further verify this, additional PCR reactions (a – f in Fig. 2) were performed for strains SB-62 and TK-1 and the products were sequenced (Figs. S5 and S6).

For strains TK-2 (ΔtrpBΔpurAΔcruF::trpB) and TK-3 (ΔtrpBΔpurAΔcruFcrtB::trpBcrtBT.thermophilus), the cruF part of gene 9 was not amplified, while gene 14 was, indicating the expected substitution of cruF for trpB. Additionally, an amplicon of the expected size for gene 13 (crtB from T. thermophilus) was observed in strain TK-3. An identical amplicon was observed from T. thermophilus (Fig. S1). The native crtB gene (second half of gene 9) was deleted from strain TK-3 and the amplicon seen in strain SB-62 (~600 bp) was not amplified in strain TK-3. However, a smaller unknown amplicon (~400 bp) was obtained and was subsequently sequenced. The results showed that the amplification did not occur from the carotenoid gene cluster (Fig. S8), but in a gene on the chromosome distant from the cluster. This gene did not show homology to the crtB gene, but the primers used to amplify crtB in R. marinus did show homology to this region, which explains the unexpected amplification. To further verify the successful modifications in strains TK-2 and TK-3, additional PCR reactions were performed (g in Fig. 2) and the resulting products were sequenced (Figs. S10 and S11).

3.3. Analysis of carotenoids by UHPSFC-DAD-QTOF/MS and spectrophotometry

Absorbance spectra of carotenoids from strains TK-1 (ΔtrpBΔpurAΔcrtYO::trpB), TK-3 (ΔtrpBΔpurAΔcruFcrtB::trpBcrtBT.thermophilus) and ISCaR-493 extracted with hexane:acetone (1:1), showed Amax for strain ISCaR-493 at 478 ​nm, for TK-1 at 488 ​nm and for TK-3 at 472 ​nm, with an additional peak at 502 ​nm (Fig. 3 and Figs. S17-S19). The spectra for TK-3 is identical to previously published spectra for lycopene (Britton et al., 2004). Absorbance between 350 and 600 ​nm was neither observed for strain TK-2 (ΔtrpBΔpurAΔcruF::trpB) nor SB-71 (ΔtrpBΔpurAcrtBI’::trpB).

1. Download : Download high-res image (364KB)
2. Download : Download full-size image

Fig. 3. Carotenoid structures identified in R. marinus strain ISCaR-493 (A), mutant TK-1 (ΔtrpBΔpurAΔcrtYO::trpB) (B) and mutant TK-3 (ΔtrpBΔpurAΔcruFcrtB::trpBcrtBT.thermophilus) (C). The top row shows chemical formulas of identified carotenoids, corresponding theoretical masses and detected masses by MS. The middle row shows absorbance spectra from 350 to 600 ​nm of carotenoids isolated with acetone:hexane (1:1) and Amax values of obtained peaks. Graphs with both axes are shown in Figs. S17-S19. The bottom row shows the chemical structures of identified carotenoids: Native carotenoids (4-keto 2‘-hydroxy β,ψ-carotene acyl glycoside), modified lycopene (2’-hydroxy ψ,ψ-carotene acyl glycoside) and lycopene. For more information on R1 and R2, refer to Fig. 1.

Extracts from R. marinus strains TK-1, TK-2, TK-3, SB-71 and ISCaR-493 were analyzed by UHPSFC-DAD-QTOF/MS. By DAD, it was not possible to detect any peaks between 400 and 600 ​nm for strains TK-2 and SB-71. Moreover, the mass spectrometer did not detect any masses corresponding to the native carotenoids nor masses corresponding to common C40 carotenoids. It was therefore concluded that TK-2 (ΔtrpBΔpurAΔcruF::trpB), like SB-71 (ΔtrpBΔpurAcrtBI’::trpB), does not produce carotenoids or produces them in very low concentrations. This was expected for SB-71 since the genes crtB and crtI (Fig. 1) were knocked-out (Bjornsdottir et al., 2011). For TK-2, these results indicated that the crtB part of the cruF-crtB gene (Fig. 2, gene 9) does not encode an active enzyme without the cruF part.

Two of the native carotenoids were detected in the TK-1 (ΔtrpBΔpurAΔcrtYO::trpB) extract. The first detected carotenoid (tR 2.48) had a mass (m/z 910.672, Δ3.2 ​mDa, Δ3.5 ​ppm) matching the β,ψ-carotenoid acyl glucoside without the 2′-hydroxyl group previously identified in R. marinus.MS/MS fragmentation confirmed this by detection of the carotenoid fragment (C40H55) but not the 4-keto β-ionone ring fragment (m/z 203.1436) found in salinixanthin (Ron et al., 2018). Moreover, the characteristic m/z 28 pattern, resulting from acyl group length variation, was detected for C11 (m/z 882.637, Δ2.4 ​mDa), C13 (m/z 910.672, Δ3.2 ​mDa), C15 (m/z 938.703, Δ2.9 ​mDa), and C17 (m/z 966.732, Δ0.8 ​mDa). The other detected carotenoid (tR 2.83 ​min) in the TK-1 extract had a mass identical to that of the native 2′-hydroxyl version of the above mentioned β,ψ-carotenoid acyl glucoside (m/z 926.660 Δ4.1 ​mDa). This carotenoid followed a similar pattern of acyl length loss and MS/MS fragmentation as the native non-hydroxylated β,ψ-carotenoid acyl glucoside identified in a previous study (Ron et al., 2018). Salinixanthin in both the 2′-hydroxyl and non-hydroxylated forms was detected in TK-1. These results suggest that all the native modifications displayed on the right side of the carotene backbone in Fig. 1, including the C-3′,4′ double bond, are still present in TK-1. This refutes the hypothesis that gene 5 encodes a 3′,4′-desaturase (see section 3.1). The results also showed that no keto group was present on the β-ring on the left side of the carotene backbone (Fig. 1), confirming the deletion of the crtO gene. Carotenoids with a linear ψ-end and a β-ring have the exact same mass and the deletion of the crtY gene could therefore not be confirmed by mass spectrometry alone. However, the change from a β-ring and ψ-end should increase the wavelength of the absorption spectrum, due to the lower contribution of the β-ring to the bathochromic shift (Krinsky et al., 2004). The absorption spectra of the carotenoids extracted from ISCaR-493 and TK-1 (Fig. 3) show a clear shift towards higher wavelengths for the TK-1 carotenoids, suggesting the presence of a ψ-end and therefore the deletion of the crtY gene.

The extract of R. marinus strain TK-3 (ΔtrpBΔpurAΔcruFcrtB::trpBcrtBT.thermophilus) showed a sole peak (tR 2.75 ​min) by DAD (λ450 nm). This peak had a mass (m/z 536.428 ​± ​0.010) corresponding to C40H56 carotenoids, such as α-carotene, β-carotene, γ-carotene and lycopene. By comparing the tR of the peak of TK-3 to these carotene standards, TK-3 had the same tR as lycopene (Figs. S12 and S13). In addition, MS/MS fragmentation spectra were analyzed. The carotenes γ-carotene and lycopene have acyclic ψ-ends, which can be determined by removal of an isoprene unit, resulting in a [M-69]+• fragment of m/z 467.368 (van Breemen et al., 2012). This fragment could be detected in the standards γ-carotene and lycopene but also in the TK-3 MS/MS spectrum at tR 2.75 ​min. Moreover, absorption spectra of the TK-3 extract (Fig. 3) shows an identical wavelength profile as established spectra for lycopene (Britton et al., 2004). From these results it can be concluded that lycopene is produced by TK-3 as the sole product from the carotenoid biosynthetic pathway. Lycopene was quantified spectrophotometrically using Beer’s law in TK-3 extracts in a separate shake flask experiment. The result was a yield of 0.49 ​± ​0.01 ​g/kg cell dry weight (CDW), which corresponded to 0.14 ​± ​0.004 ​mg/L cultivation volume. In conclusion, the results suggest that the cruF-crtB gene was successfully deleted from the genome of R. marinus and that the heterologous crtB gene from T. thermophilus encodes an active enzyme in strain TK-3. Additionally, and somewhat surprisingly, the only active carotenoid enzymes in TK-3 seem to be CrtB and CrtI, leaving the other enzymes, such as CrtY, non-functional.