Engineered biochemical pathway for the biosynthesis of retinal in E. coli The genes β-carotene biosynthesis (ctrEIBY) and cleavage (blh) are engineered in E. coli . Farnesyl diphosphate, the starting molecule of the pathway is naturally built in E. coli

Proteorhodopsin needs all-trans retinal to be active. In many bacteria that naturally produce proteorhodopsin (i.e. SAR86), retinal is synthetized starting from β-carotene. These bacteria have a cluster of genes that includes in addition to proteorhodopsin, the operon for β-carotene production ctrEIBY and blh that encodes for β-carotene 15,15’-Dioxygenase^[1].

We built two different devices for the production of retinal and one for the production of β-carotene. We used parts newly built by us and other extracted by the Registry (kit 2012).

Devices for the production of retinal To provide blh and β-carotene synthesis to E. coli, we designed the new part BBa_K1604020 starting from BBa_K274210. We finally merged the two together to create part BBa_K1604022.

Our β-carotene device was built with the arabinose promoter from UniTN 2012 (BBa_K731201) and the β-carotene part of Cambridge 2009 extracted from the registry.

Cells transformed with this device were grown and induced for 24 hours with 5 mM arabinose. After the cells were centrifuged, we compared the two pellets to discover that also the uninduced sample showed a very bright orange pellet. The araC-pBAD promoter is very tightly regulated and has a very little basal expression. By looking at the sequence of the operon, we discovered a possible internal promoter in the ctrE gene, the first gene of the pathway.

Bacterial pellet confirms β-carotene production in transformed cells A: Pellet collected from BBa_K731201 transformed and induced (control). B-C: pellet collected from BBa_K1604020 (β-carotene producers) induced (B) and not induced (C). BBa_K1604020 transformed cells produced β-carotene both in presence and absence of the arabinose induction.

Our goal was to synthetize retinal and provide it to proteorhodopsin. Our blh producing device was synthetized by Genescript^[2] with a sequence optimized for E. coli expression. We subcloned it into pSB1C3 and submit it to the registry (BBa_K1604021). It was not possible to characterize blh in vivo by itself, due to the low solubility of β-carotene and the fact that the cells would not uptake the molecule. To overcome this problem we decided to produce endogenous β-carotene. Blh, under the control of a constitutive promoter J23100, was characterized both in co-transformed cells with BBa_K1604020 (araC-pBAD-ctrEBI) and in a complete device containing the complete pathway (BBa_K1604022).

β-carotene is absent in blh-expresing cells A: Pellet collected from BBa_K1604020 (β-carotene producer) transformed cells induced with arabinose. B-C-D: Pellet collected from BBa_K1604022 (retinal producer) transformed cells not induced (B), induced with arabinose 5mM (C), and with both arabinose 5 mM and Fe (D); Control cells (BBa_K731030) induced with arabinose and Fe²⁺(E);

Where is our retinal?

Happy for this result we extracted with acetone the pigments^[3] (carotenoids and retinoids) from cells expressing BBa_K1604022 and run them on HPLC reverse phase column. β-carotene was disappeared, however we did not detect any peak with the corresponding retention time of retinal.

HPLC profile HPLC profiles between reference (mixture of pure β-carotene and retinal), extract from BBa_K1604020 induced cells and BBa_K1604022 not induced cells. The pigments were extracted incubating the pellet with 2.5 mL of acetone for 10 min at 50C. The samples were concentrated with N2, methanol was added to reach a final volume of 500 uL. The different samples were run on the HPLC Agilent 1100 on a Agilent Eclipse XDB C8 3.5 uM (4.6 mmx150 mm) reverse phase column. Eluent used were, buffer A MeOH/H2O 7/3 + 12 mmM acetate, buffer B MeOH + 12 mmM acetate at 08 mL/min. The gradient used was 35% of buffer B up to 100% in 40 minutes.

We also measured the absorption spectrum of the extracted pigments. We used pure retinal and β-carotene for reference, which absorb at 373 nm and 456 nm respectively. Also this test confirmed the disappearance of β-carotene, but did not show proof of retinal synthesis. The same behavior was observed in co-transformed cells.

Absorption spectra of cellular extracts An UV-Visible assay was performed on β-carotenoids extracted from cells with acetone. Absorption spectrum of cells transfected with BBa_K1604020 (β-carotene producers) (yellow) resembles the spectra of a β-carotene reference sample. Conversely, cells transformed with BBa_K1604022 (red) and retinal reference do not show such analogy.

We think that β-carotene is being cleaved to form retinal (as shown by the evident loss of color, and the absence of a peak in the HPLC or UV-VIS spectrum), and it is immediately taken by the cell to enter different biochemical pathways. The biosynthesis of retinal involves the formation of intermediate molecules that could also be used by E. coli in different metabolic reactions.