LOV Domain

LOV stands for light oxygen voltage. It is a sensor protein that detects the presence of blue light (365 nm). In its wild type form it is used by higher plants, fungi, and bacteria. In higher plants LOV controls phototropism and chloroplasts relocation. In this form it absorbs blue light (365 nm) and in the wild state flavin mononucleotides (FMN) link to cysteine. This results in LOV not being able to efficiently emit green light (495 nm) due to FMN.

We choose to modify LOV as our anaerobic environment and growth indicator. We choose LOV over green fluorescent proteins (GFP) due to the fact that GFPs are completely dependant on molecular oxygen to glow. However, due to FMN, LOV cannot release green light (495 nm).

Thomas Drepper found a solution to this problem. Drepper and his fellow researchers realized the effects of FMN and found away to remove it. By eliminating the LOV domain's cysteine amino acid, FMN had nothing to bind to.

Following Drepper's model we replaced the cysteine amino acid at position 62 of the Bacillus subtilis-derived LOV domain with an alanine:

001 MASFQSFGIP GQLEVIKKAL DHVRVGVVIT DPALEDNPIV YVNQGFVQMT GYETEEILGK
061 NARFLQGKHT DPAEVDNIRT ALQNKEPVTV QIQNYKKDGT MFWNELNIDP MEIEDKTYFV
121 GIQNDITKQK

This is where our models departed. Instead of optimizing our gene for E.coli, we choose to optimize codon usage for S. oneidensis. We used the online [http://www.jcat.de/ Java Codon Adaptation Tool] to create a DNA sequence which has been optimized for our use in the following ways:

• Optimized for S. oneidensis codon usage.
• Avoids use of the four 3A Assembly restriction enzymes: EcoRI, SpeI, XbaI, PstI.
• Avoids internal prokaryotic ribosome binding sites.

When using these parameters, JCAT produces the following nucleic acid sequence:

ATGGCTTCTTTCCAATCTTTCGGTATCCCAGGTCAATTAGAAGTTATCAA     50
AAAAGCTTTAGATCACGTTCGTGTTGGTGTTGTTATCACTGATCCAGCTT     100
TAGAAGATAACCCAATCGTTTACGTTAACCAAGGTTTCGTTCAAATGACT     150
GGTTACGAAACTGAAGAAATCTTAGGTAAAAACGCTCGTTTCTTACAAGG     200
TAAACACACTGATCCAGCTGAAGTTGATAACATCCGTACTGCTTTACAAA     250
ACAAAGAACCAGTTACTGTTCAAATCCAAAACTACAAAAAAGATGGTACT     300
ATGTTCTGGAACGAATTAAACATCGATCCAATGGAAATCGAAGATAAAAC     350
TTACTTCGTTGGTATCCAAAACGATATCACTAAACAAAAAGAATACGAAA     400
AATTATTAGAA

Finally, we added a TAA stop codon to the end of the sequence to ensure termination of the translation.

We used the sequence alignment tool to confirm that this was in fact the correct gene. The result show an exact match.

100.0% identity in 137 residues overlap; Score: 712.0; Gap frequency: 0.0%

Intended     1 MASFQSFGIPGQLEVIKKALDHVRVGVVITDPALEDNPIVYVNQGFVQMTGYETEEILGK
Translated   1 MASFQSFGIPGQLEVIKKALDHVRVGVVITDPALEDNPIVYVNQGFVQMTGYETEEILGK
               ************************************************************

Intended    61 NARFLQGKHTDPAEVDNIRTALQNKEPVTVQIQNYKKDGTMFWNELNIDPMEIEDKTYFV
Translated  61 NARFLQGKHTDPAEVDNIRTALQNKEPVTVQIQNYKKDGTMFWNELNIDPMEIEDKTYFV
               ************************************************************

Intended   121 GIQNDITKQKEYEKLLE
Translated 121 GIQNDITKQKEYEKLLE
               *****************

We used an E. coli codon usage application to see how well it will do in the bacteria. This analysis shows that there are multiple occurrences of the codon TTA (for leucine) which is poorly expressed in E. coli. The next most common Leucine codon in Shewanella is CTG, which is also well tolerated in E. coli. The following sequence replaces every instance of the Leucine TTA codon with CTG:

atggcttctttccaatctttcggtatcccaggtcaactggaagttatcaaaaaagctctg
M  A  S  F  Q  S  F  G  I  P  G  Q  L  E  V  I  K  K  A  L
gatcacgttcgtgttggtgttgttatcactgatccagctctggaagataacccaatcgtt
D  H  V  R  V  G  V  V  I  T  D  P  A  L  E  D  N  P  I  V
tacgttaaccaaggtttcgttcaaatgactggttacgaaactgaagaaatcctgggtaaa
Y  V  N  Q  G  F  V  Q  M  T  G  Y  E  T  E  E  I  L  G  K
aacgctcgtttcctgcaaggtaaacacactgatccagctgaagttgataacatccgtact
N  A  R  F  L  Q  G  K  H  T  D  P  A  E  V  D  N  I  R  T
gctctgcaaaacaaagaaccagttactgttcaaatccaaaactacaaaaaagatggtact
A  L  Q  N  K  E  P  V  T  V  Q  I  Q  N  Y  K  K  D  G  T
atgttctggaacgaactgaacatcgatccaatggaaatcgaagataaaacttacttcgtt
M  F  W  N  E  L  N  I  D  P  M  E  I  E  D  K  T  Y  F  V
ggtatccaaaacgatatcactaaacaaaaagaatacgaaaaactgctggaataa

We also added the following terminal restriction sites to allow our LOV gene to be comparable with the 3A assembly process.

 Prefix
5' GTTTCTTCGAATTCGCGGCCGCTTCTAGAG[part] 3'
Suffix
5' [part]TACTAGTAGCGGCCGCTGCAGGAAGAAAC 3'

To check over our work we also ran the sequence through the http://tools.neb.com/NEBcutter2/index.php NEB Cutter to ensure that the 3A assembly restriction sites are correctly located at the ends of the sequence and not elsewhere within the open reading frame. The results show that the restriction sites are present where required and not elsewhere.

This is the BioBrick-formatted, Shewanella-optimized, gene we ordered from the sequencing company for insertion into the pSB1C3 plasmid:

>LOV
GTTTCTTCGA ATTCGCGGCC GCTTCTAGAG atggcttctt tccaatcttt 
cggtatccca ggtcaactgg aagttatcaa aaaagctctg gatcacgttc 
gtgttggtgt tgttatcact gatccagctc tggaagataa cccaatcgtt 
tacgttaacc aaggtttcgt tcaaatgact ggttacgaaa ctgaagaaat 
cctgggtaaa aacgctcgtt tcctgcaagg taaacacact gatccagctg 
aagttgataa catccgtact gctctgcaaa acaaagaacc agttactgtt 
caaatccaaa actacaaaaa agatggtact atgttctgga acgaactgaa 
catcgatcca atggaaatcg aagataaaac ttacttcgtt ggtatccaaa 
acgatatcac taaacaaaaa gaatacgaaa aactgctgga ataaTACTAG 
TAGCGGCCGC TGCAGGAAGA AAC