https://docs.google.com/presentation/d/1Cn7qS2M30VH4OYBUGR8EplkHSSAhIZiah1BLQL39PIM/edit#slide=id.g33f08f69481_46_0

According to the paper (https://www.nature.com/articles/s41467-022-33127-w), Ceres & Demetra are responsible for oxidation of PE, out of which Demetra is capable to oxidize the latter which is a key component for PE degradation.

Interestingly, as author discusses, both proteins share similarity with haemocyanins. Haemocyanins are copper containing proteins that meditate oxidation through their copper core. It can be assumed, that the oxidation in case of Demetra is also mediated by its copper core.

I have considered to take advantage of the following information and have devised different strategies:

1. Duplicate copper containing core to increase oxidation

• Latter can be tested by first repeating the code and then observing stability scores in alpha fold (this approach had been discarded)

1. Replace copper containing core by elements that are better at oxidation (e.g. iron)

• Latter can be done by introducing amino acids from iron containing proteins that share structural similarity or evolutionary history with Demetra (this approach was chosen, as iron is more redox active and might indeed lead to quicker plastic degradation)
• To purify I intend to use a signal peptide that will direct the protein to the membrane, hopefully resulting in “leaked integration in the membrane” leading to protein being expelled in the LB medium.
• Add fluorescent protein tag in the end for testing if protein ends up in the medium
• I’ll use flexible glycine linkages for the fluorescent protein and verify in Alphafold for folding

Apparently, patent had already been filed, but it had not been accepted. Either way, manipulating sequence for academic non-commercial purposes should not be a problem: WIPO - Search International and National Patent Collections

According to the paper there are similarities of Demetra and Ceres to different hemocyanins, especially in specific locations detailed in supplementary file of the paper. Unfortunately, neither blasting these specific regions, nor alignment of Demetra & Ceres with up to 30 random hemocyanin proteins did not reproduce the results mentioned in the paper. Nonetheless, I had decided to proceed with details mentioned in their paper and integrate mutations in the regions that they had described. Bellow is the figure from the supplementary material detailing Demetra to hemocyanin (A) and Ceres to Hemocyanin (B) alignment areas.

I have aimed to introduce iron binding sites or iron mediated oxidizing capacities in Demetra, but before proceeding, I tried multiple alignment by passing Demetra as a subject sequence and MtrA and MtrC family proteins (of Shewanella Oneidensis) as a query sequences, to identify if there already exists similarities that would have made engineering decisions easier. Sadly, no significant similarity was found, so I have decided to proceed by following the similarity to hemocyanin sites mentioned in the paper.

To introduce heme binding sites, first such sites of MtrC were determined. The latter was chosen because it serves as an outer membrane protein and as such must be more relevant for electron transport extracellularly (https://www.rcsb.org/sequence/6R2Q).

>6R2Q_3|Chain C|Decaheme c-type cytochrome, OmcA/MtrC family|Shewanella baltica (62322)
MMNAQTTKIALLLAASAVTMALTGCGGSDGNDGNPGEPGGEPAPAIQILNFTFDKSVITNGVPSVEFTVTNENDLPVVGLQKMRFAAAQLIPQGATGAGNASQWQYFGDETCDVAATCPGTFVDQKNGHYSYTFNMNLTANAKITYNDQLAQRVLIRAYNTPLPDGTQVPNSNAFVDFTADTGAAPTYSRKIVATESCNTCHQDLANVKHGGAYSDVNYCATCHTAGKVGVGKEFNVLVHAKHKDLTLGSLESCQSCHAANDAAPDWGNWSRIPTAATCGSCHSTVDFAAGKGHSQQLDNSNCIACHNSDWTAELHTGKTADKKAVIAQLGMQATLVGQTDDTAVLTVSILDKDGNAIDAATVQDKIKRLETVTNVGPNFPIMGYNKSPGSGAAKIAKDLVKDGALQAGVTLVDGKLVFTTPALPFGTGDTDTAFTFIGLEMCSTGTSLTACTVDSATTSMKAELAFGTKSGNAPSMRHVNSVNFSTCQGCHSDTFEIHKGHHSGFVMTEQVSHAKDANGKAIVGVDGCVACHTPDGTYASGANKGAFEMKLHVIHGEQGVIKECTQCHNDFNLDAFKVKGALATSAGKYTTPI

TATCTSCHAPESIGHGLENMGAIVNGDYVQANQAAQSETCFYCHKPTPTDHTQVKM

According to RCSB databank, following binding sites had been determined in MtrC:

Heme binds multiple times, where there appears to be a pattern of amino acids (violet) until final group of amino acids which bound to heme (green). My aim is to introduce amino acids that are able to form a covalent bond with an iron ion. These are the amino acids from 632 to 650. With gap between 614 to 632.

But after selecting both 632 to 650 and 614 to 632 as yellow, and showing organic (heme sites) as ball representation, it is visible that heme bound sequences are not completely resolved.

After multiple selection I marked amino acids at 5 A from last heme: